Corrosion and Corrosion Control, 4th Ed

CORROSION AND CORROSION CONTROL An Introduction to Corrosion Science and Engineering FOURTH EDITION R. Winston Revie Senior Research Scientist CANMET Materials Technology Laboratory Natural Resources Canada Herbert H. Uhlig Former Professor Emeritus Department of Materials Science and Engineering Massachusetts Institute of Technology A JOHN WILEY & SONS, INC., PUBLICATION CORROSION AND CORROSION CONTROL CORROSION AND CORROSION CONTROL An Introduction to Corrosion Science and Engineering FOURTH EDITION R. Winston Revie Senior Research Scientist CANMET Materials Technology Laboratory Natural Resources Canada Herbert H. Uhlig Former Professor Emeritus Department of Materials Science and Engineering Massachusetts Institute of Technology A JOHN WILEY & SONS, INC., PUBLICATION Copyright © 2008 by John Wiley & Sons, Inc. All right reserved Published by John Wiley & Sons, Inc., Hoboken New Jersey Published simultaneously in Canada No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax (978)750-4470, or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at http://www.wiley.com/go/permission. Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages. For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at (800) 762-2974, outside the United States at (317) 572-3993 or fax (317) 572-4002. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic formats. For more information about Wiley products, visit our web site at www.wiley.com. Library of Congress Cataloging-in-Publication Data: Uhlig, Herbert Henry, 1907– Corrosion and corrosion control : an introduction to corrosion science and engineering / Herbert H. Uhlig, R. Winston Revie.—4th ed. p. cm. Includes bibliographical references and index. ISBN 978-0-471-73279-2 (cloth) 1. Corrosion and anti-corrosives. I. Revie, R. Winston (Robert Winston), 1944– II. Title. TA462.U39 2008 620.1′1223–dc22 2007041578 Printed in the United States of America 10 9 8 7 6 5 4 3 2 1 CONTENTS Preface xvii 1 1 1 2 5 6 6 6 6 7 7 9 9 11 13 15 18 19 19 1 DEFINITION AND IMPORTANCE OF CORROSION 1.1 Definition of Corrosion 1.1.1 Corrosion Science and Corrosion Engineering 1.2 Importance of Corrosion 1.3 Risk Management 1.4 Causes of Corrosion 1.4.1 Change in Gibbs Free Energy 1.4.2 Pilling–Bedworth Ratio References General References Problems ELECTROCHEMICAL MECHANISMS 2.1 The Dry-Cell Analogy and Faraday’s Law 2.2 Definition of Anode and Cathode 2.3 Types of Cells 2.4 Types of Corrosion Damage References General References Problems THERMODYNAMICS: CORROSION TENDENCY AND ELECTRODE POTENTIALS 3.1 Change of Gibbs Free Energy 3.2 Measuring the Emf of a Cell 3.3 Calculating the Half-Cell Potential—The Nernst Equation 3.4 The Hydrogen Electrode and the Standard Hydrogen Scale 3.5 Convention of Signs and Calculation of Emf 2 3 21 21 22 22 24 25 v vi CONTENTS 3.6 3.7 3.8 3.9 3.10 Measurement of pH The Oxygen Electrode and Differential Aeration Cell The Emf and Galvanic Series Liquid Junction Potentials Reference Electrodes 3.10.1 Calomel Reference Electrode 3.10.2 Silver–Silver Chloride Reference Electrode 3.10.3 Saturated Copper–Copper Sulfate Reference Electrode References General References Problems Answers to Problems 28 28 30 33 34 35 36 36 37 38 38 40 43 43 44 45 47 48 49 50 50 50 51 53 53 54 56 58 58 63 66 68 71 73 75 4 THERMODYNAMICS: POURBAIX DIAGRAMS 4.1 Basis of Pourbaix Diagrams 4.2 Pourbaix Diagram for Water 4.3 Pourbaix Diagram for Iron 4.4 Pourbaix Diagram for Aluminum 4.5 Pourbaix Diagram for Magnesium 4.6 Limitations of Pourbaix Diagrams References General References Problems Answers to Problem KINETICS: POLARIZATION AND CORROSION RATES 5.1 Polarization 5.2 The Polarized Cell 5.3 How Polarization Is Measured 5.3.1 Calculation of IR Drop in an Electrolyte 5.4 Causes of Polarization 5.5 Hydrogen Overpotential 5.6 Polarization Diagrams of Corroding Metals 5.7 Influence of Polarization on Corrosion Rate 5.8 Calculation of Corrosion Rates from Polarization Data 5.9 Anode–Cathode Area Ratio 5.10 Electrochemical Impedance Spectroscopy 5 CONTENTS vii 5.11 Theory of Cathodic Protection References General References Problems Answers to Problems 77 79 80 80 82 83 83 84 88 89 90 92 95 96 97 99 100 103 108 108 109 111 112 113 115 115 116 116 120 120 127 129 131 138 138 138 6 PASSIVITY 6.1 Definition 6.2 Characteristics of Passivation and the Flade Potential 6.3 Behavior of Passivators 6.3.1 Passivation of Iron by HNO3 6.4 Anodic Protection and Transpassivity 6.5 Theories of Passivity 6.5.1 More Stable Passive Films with Time 6.5.2 Action of Chloride Ions and Passive–Active Cells 6.6 Critical Pitting Potential 6.7 Critical Pitting Temperature 6.8 Passivity of Alloys 6.8.1 Nickel–Copper Alloys 6.8.2 Other Alloys 6.9 Effect of Cathodic Polarization and Catalysis References General References Problems Answers to Problems IRON AND STEEL 7.1 Introduction 7.2 Aqueous Environments 7.2.1 Effect of Dissolved Oxygen 7.2.2 Effect of Temperature 7.2.3 Effect of pH 7.2.4 Effect of Galvanic Coupling 7.2.5 Effect of Velocity on Corrosion in Natural Waters 7.2.6 Effect of Dissolved Salts 7.3 Metallurgical Factors 7.3.1 Varieties of Iron and Steel 7.3.2 Effects of Composition 7 1 Electrochemical Dissolution 8.5 Radiation Damage 8.3 Adsorption-Induced Localized Slip 8.1 Introduction 9.4.3.2 Remedial Measures 8.6 Rate of Crack Growth (Fracture Mechanics) 8.2 Film-Induced Cleavage 8.3 Mechanism of Corrosion Fatigue 8.4 Stress Sorption 8.7 Fretting Corrosion 8.2 Stress-Corrosion Cracking 8.7.1 Mechanism of Hydrogen Damage 8.1 Critical Minimum Corrosion Rates 8.6 Corrosion Fatigue 8.1 Mechanism of Fretting Corrosion 8.3.7.3 Mechanism of Stress-Corrosion Cracking of Steel and Other Metals 8.2 Types of Atmospheres 9 .4.3.6.2 Remedial Measures References General References Problems Answers to Problems ATMOSPHERIC CORROSION 9.2.1 Iron and Steel 8.3.3 Effect of Heat Treatment 7.5 Initiation of Stress-Corrosion Cracking and Critical Potentials 8.3.viii CONTENTS 7.4 Hydrogen Damage 8.6.4 Steel Reinforcements in Concrete References General References Problems Answers to Problems 142 143 145 147 147 148 149 149 150 151 156 157 158 158 158 161 162 166 167 170 172 173 177 178 179 180 182 184 185 188 190 190 191 191 192 8 EFFECT OF STRESS 8.1 Cold Working 8.3.3.6.2 Effect of Metal Flaws 8. 4 Corrosion-Product Films Factors Influencing Corrosivity of the Atmosphere 9.1 Introduction 11.10.4.2 Gases in the Atmosphere 9.3 Moisture (Critical Humidity) 9.6 Oxide Properties and Oxidation 11.3 Thermodynamics of Oxidation: Free Energy–Temperature Diagram 11.8 Hot Ash Corrosion 11.1 Particulate Matter 9.4.5 Remedial Measures References General References Problems 192 195 196 197 199 201 202 203 204 205 205 206 207 208 210 211 212 212 215 215 216 218 218 220 223 224 227 229 229 230 231 231 232 233 234 234 10 CORROSION IN SOILS 10.1 Three Equations of Oxidation 11.10 Oxidation of Copper 11.4 Stress-Corrosion Cracking 10.CONTENTS ix 9.4.5 Wagner Theory of Oxidation 11.1 Internal Oxidation 11.13.1 Reactive Element Effect (REE) 11 .2 Factors Affecting the Corrosivity of Soils 10.3 9.1 Introduction 10.9 Hot Corrosion 11.10.3.2 Reaction with Hydrogen (“Hydrogen Disease”) 11.11 Oxidation of Iron and Iron Alloys 11.5 Remedial Measures References General References OXIDATION 11.13 Oxidation-Resistant Alloys 11.3 Bureau of Standards Tests 10.12 Life Test for Oxidation-Resistant Wires 11.4.4 Protective and Nonprotective Scales 11.7 Galvanic Effects and Electrolysis of Oxides 11.2 Initial Stages 11.1 Pitting Characteristics 10. 1 Sacrificial Anodes 13.2 Sources of Stray Currents 12.2 Chromium–Iron Alloys 11.8 Economics of Cathodic Protection 13.1 Overprotection 13.5 Soil-Resistivity Measurement 12.2 Brief History 13.3 How Applied 13.7 Criteria of Protection 13.2 Doubtful Criteria 13.7.6.3.7.9 Anodic Protection References General References 13 .3 Quantitative Damage by Stray Currents 12.5 Furnace Windings References General References Problems Answers to Problems 235 236 236 237 237 239 239 240 241 241 242 244 245 246 246 247 247 247 249 251 251 252 253 254 255 257 258 259 260 260 262 262 263 263 265 265 12 STRAY-CURRENT CORROSION 12.3 Position of Reference Electrode 13.7.4 Detection of Stray Currents 12.13.6 Means for Reducing Stray-Current Corrosion References General References Problems Answers to Problems CATHODIC PROTECTION 13.13.1 Introduction 13.1 Introduction 12.6 Anode Materials and Backfill 13.4 Combined Use with Coatings 13.x CONTENTS 11.13.4 Nickel and Nickel Alloys 11.5 Magnitude of Current Required 13.1 Potential Measurements 13.3 Chromium–Aluminum–Iron Alloys 11.13. 6 Chromium-Plated Steel for Containers 14.3.7 Plastic Linings 15 16 .CONTENTS xi Problems Answers to Problems 266 267 269 269 271 272 272 274 274 276 277 279 280 281 282 285 285 286 286 288 288 289 289 289 291 293 293 294 295 296 296 296 298 299 14 METALLIC COATINGS 14.3.5.7 Aluminum Coatings References General References INORGANIC COATINGS 15.5 Tin Coatings 14.3 Zinc Coatings 14.2 Portland Cement Coatings 15.3 Requirements for Corrosion Protection 16. Oils.6.1 Vitreous Enamels 15.6 Filiform Corrosion 16. and Greases from the Surface 16.3.3.4 Cadmium Coatings 14.2 Classification of Coatings 14.1 Cleaning All Dirt.2 Painting of Aluminum and Zinc 16.5.2 Paints 16.3 Specific Metal Coatings 14.3.1 Nickel Coatings 14.1 Theory of Filiform Corrosion 16.2 Complete Removal of Rust and Mill Scale 16.3 Chemical Conversion Coatings References General References ORGANIC COATINGS 16.4 Metal Surface Preparation 16.5 Applying Paint Coatings 16.1 Introduction 16.1 Wash Primer 16.4.3.2 Lead Coatings 14.4.1 Methods of Application 14.3. 3.2 Classes and Types 19.2.2. STAINLESS STEELS 19.3.1 Brief History 19.3 Mechanisms References General References ALLOYING FOR CORROSION RESISTANCE.4 Slushing Compounds 17.and Cold-Water Treatment 18.2 Applications of Passivators 17.1 Deaeration and Deactivation 18.1 Applications of Pickling Inhibitors 17.5 Vapor-Phase Inhibitors 17.2.1 Cooling Waters 18.3 Pickling Inhibitors 17.2 Boiler-Water Treatment for Corrosion Control 18.4 Pitting and Crevice Corrosion 19.1 Inhibitor to Reduce Tarnishing of Copper References General References TREATMENT OF WATER AND STEAM SYSTEMS 18.5.3 Intergranular Corrosion 19.3.2.2 Passivators 17.2.1 Introduction 19.3 Boiler-Water Treatment 18.2.2.5 Stress-Corrosion Cracking and Hydrogen Cracking 19.1 Boiler Corrosion 18.1 Introduction 17.xii CONTENTS References General References 300 301 303 303 304 304 308 310 312 313 313 314 315 316 317 317 321 322 323 323 326 328 330 331 17 INHIBITORS AND PASSIVATORS 17.2.3.6 Cracking of Sensitized Austenitic Alloys in Polythionic Acids 18 19 333 333 335 336 337 343 350 354 359 .1 Mechanism of Passivation 17.2.2 Hot.2 Stainless Steels 19. 2 Dealloying/Dezincification 20.2 Coatings 22.2.1.3 Stress-Corrosion Cracking (Season Cracking) 20.2.2 Corrosion in Water and Steam 21.1 Corrosion in Natural Waters 20.4 Condenser Tube Alloys Including Copper–Nickel Alloys References General References Problems Answers to Problems ALUMINUM AND ALUMINUM ALLOYS 21.4 Corrosion Characteristics 21.1 Introduction 22.1.1.1.1 Stress-Corrosion Cracking References General References MAGNESIUM AND MAGNESIUM ALLOYS 22.1.3.1 Stress-Corrosion Cracking 22.3 Magnesium Alloys 22.2 Copper Alloys 20.2.5 Galvanic Coupling 21.1 Copper–Zinc Alloys (Brasses) 20.2 Aluminum Alloys 21.CONTENTS xiii 19.7 Galvanic Coupling and General Corrosion Resistance References General References 361 362 365 367 367 369 371 371 372 374 378 379 381 381 381 383 383 384 384 387 388 392 393 394 396 397 399 399 399 400 402 403 404 20 COPPER AND COPPER ALLOYS 20.3 Effect of pH 21.1 Clad Alloys 21.2.2 Magnesium 22.3.1.2.1 Copper 20.2.1 Aluminum 21.4 Summary 21 22 . 6. 5% Fe 23. 15% Cr. 30% Cu 23.1 General Behavior 23.3.2 Zirconium Alloys 24 25 26 .3 Pitting and Crevice Corrosion 25.3. 30% Mo. 16% Cr. 20% Fe. 22% Cr.3.4 Intergranular Corrosion and Stress-Corrosion Cracking References General References Problem ZIRCONIUM 26.6 Ni–Cr–Mo System: Alloy C—54% Ni.2 Cobalt Alloys References General References TITANIUM 25.3.1 Introduction 23. 5% Fe 23.2 Titanium Alloys 25.1 Introduction 24.1 Introduction 26.2 Ni–Cu System: Alloy 400—70% Ni.7 Ni–Fe–Cr System: Alloy 825—Ni. 2% Cu 23.3 Nickel Alloys 23.2 Nickel 23.3 Ni–Cr–Fe System: Alloy 600—76% Ni.3. 7% Fe 23. 4% W.5 Ni–Cr–Fe–Mo–Cu System: Alloy G—Ni.5% Mo. 22% Cr References General References COBALT AND COBALT ALLOYS 24. 16% Mo.3.4 Ni–Mo System: Alloy B—60% Ni. 31% Fe.xiv CONTENTS References General References 405 405 407 407 408 411 411 414 414 415 416 416 417 417 418 419 419 420 423 423 425 425 427 429 430 432 434 434 435 435 436 23 NICKEL AND NICKEL ALLOYS 23.3.1 Titanium 25. 3 Summary References General References APPENDIX 29.7 Derivation of the Equation Expressing Weight Loss by Fretting Corrosion 29.1 Activity and Activity Coefficients of Strong Electrolytes 29.1 Introduction 27.2 Current Density Equivalent to a Corrosion Rate of 1 gmd 28 29 456 458 461 467 469 470 471 474 475 475 .8 Conversion Factors 29.1 Additional Conversion Factors 29.8.5 Derivation of the Equation for Potential Drop along the Soil Surface Created by Current Entering or Leaving a Buried Pipe 29.2.2 Derivation of Stern–Geary Equation for Calculating Corrosion Rates from Polarization Data Obtained at Low Current Densities 29.6 Derivation of the Equation for Determining Resistivity of Soil by Four-Electrode Method 29.4 Derivation of Potential Change along a Cathodically Protected Pipeline 29.2.1 The General Equation 29.2 Corrosion Behavior References General Reference LEAD 28.1 Introduction 28.8.3 Derivation of Equation Expressing the Saturation Index of a Natural Water 29.1 Lead–Acid Battery 28.CONTENTS xv 26.3 Behavior in Hot Water and Steam References General References 437 439 440 441 441 441 443 443 445 445 446 447 448 449 449 451 451 27 TANTALUM 27.2 Corrosion Behavior of Lead and Lead Alloys 28. 9 Standard Potentials 29.xvi CONTENTS 29.10 Notation and Abbreviations References Index 476 476 478 479 . The reliability of materials used in nuclear waste disposal must be sufficient so that that the safety of future generations is not compromised. so that “zero” really does mean “zero” and “no” means “no. oil and natural gas are transmitted across continents using high-pressure steel pipelines that must operate for decades without failure. innovative approaches. and analysis wherever possible. not only xvii . Notwithstanding the many years over which university. The ability to manage corrosion is a central part of using materials effectively and efficiently to meet these challenges. so that neither the groundwater nor the air is unnecessarily polluted. explanation. when necessary. operation. Senior management of some organizations has adopted a policy of “zero failures” or “no failures. water. clean water to drink. rather than passively observing the smorgasbord of information presented. and failure occurs.” In translating this management policy into reality. Although the teaching of corrosion should not be regarded as a dismal failure.PREFACE The three main global challenges for the twenty-first century are energy. the galvanic series in seawater is presented with the potential range of each material. Materials reliability is becoming ever more important in our society. high-profile corrosion failures continue to take place. for example. For example. A second objective is to engage students. particularly in view of the liability issues that develop when reliability is not assured. experimental trials. the student can see how anodic/cathodic effects can develop. and continuing education courses in corrosion have been available. in this new edition. and. sufficient energy to ensure a reasonable standard of living. and maintenance of nuclear power plants. In design. and clean air to breathe. and air—that is. management of corrosion is critical. it has certainly not been a stellar success providing all engineers and technologists a basic minimum “literacy level” in corrosion that would be sufficient to ensure reliability and prevent failures. One objective of preparing the fourth edition of this book is to present to students an updated overview of the essential aspects of corrosion science and engineering that underpin the tools that are available and the technologies that are used for managing corrosion and preventing failures. Considering the potential ranges that can be involved. rather than only as a qualitative list. college. safety is compromised. The main emphasis is on quantitative presentation. so that they are active participants in understanding corrosion and solving problems.” engineers and others manage corrosion using a combination of well-established strategies. passivity. have reduced air pollution in the United States. In this edition. magnesium (Chapter 22). there is the dual importance of thermodynamics (In which direction does the reaction go? Chapters 3 and 4) and kinetics (How fast does it go? Chapter 5). and titanium (Chapter 25) to provide more information on these metals and their alloys than in the previous editions. This is primarily a textbook for students and others who need a basic understanding of corrosion. For example. alloys are identified with their UNS numbers as well as with their commonly used identifiers. Although the basic organization of the book is unchanged from the previous edition. very useful tools that indicate the thermodynamic potential–pH domains of corrosion. researchers. the EPA Lead and Copper rule enacted in the United States in 1991. numerous examples throughout this book illustrate advances that have been made in corrosion engineering of automobiles. Advances in protecting cars and trucks from corrosion must also be viewed in the context of reducing vehicle weight by using magnesium. A consideration of the relevant Pourbaix diagrams can be a useful starting point in many corrosion studies and investigations. The book is also a reference and starting point for engineers. there have been many advances in corrosion. the next 23 chapters expand upon the fundamentals in specific systems and applications and discuss strategies for protection. and the problems are integrated into the book by presenting them at the ends of the chapters. but also when materials that are nominally the same are coupled. advances in alloys for application in aggressive environments. and other lightweight materials in order to decrease energy usage (increase the miles per gallon. There are separate chapters on aluminum (Chapter 21). compliant coatings have been developed (Chapter 16). and immunity to corrosion. there is in this edition a separate chapter on Pourbaix diagrams. aluminum. and technologists requiring specific information. environmental concerns and regulations are presented in the context of their impact on corrosion and its control—for example. and advances of industry in response to public demand. The industrial developments in response to the Clean Air Act. For consistency with current practice in metallurgical and engineering literature. as well as in this book. For this reason. or kilometers per liter. including alloys both old and new. After establishing the essential basics of corrosion in the first five chapters. Throughout this book. enacted in 1970. of gasoline) and reduce greenhouse gas emissions. some new numerical problems have been added.xviii PREFACE when different materials form a couple. and also because many students have a fascination with cars. including advances in knowledge. As always in corrosion. with some effect on atmospheric corrosion (Chapter 9). To answer the question from students about why so . Since the third edition of this book was published. To meet the requirements of environmental regulations and reduce the use of organic solvents. The book includes discussion of the main materials that are available. consumer demand for corrosion protection of automobiles has led to a revolution of materials usage in the automotive industry. An instructor’s manual is also being prepared. discussions of new alloys that have been developed to meet demands for increasing reliability notwithstanding the increased structural lifetimes that are being required in corrosive environments of ever-increasing severity. hence. with their suggestions and ideas for future editions. discussed in Chapters 8 and 26. and.PREFACE xix many alloys have been developed and are commercially available. with whom it has been my privilege to work for the past nearly 30 years. Uhlig (1907–1993). In the discussion of stainless steels (Chapter 19). and most of these were published in 2000 and later. Greta Uhlig. there are numerous references to further sources of information.I. I would like to thank the Uhlig family for their generosity and hospitality during five decades. as well as the information on critical pitting potential (CPP). as well as in the traditional print format. and to help reduce the huge economic losses and dismaying wastage of . reviews. and papers in journals. the concept of critical pitting temperature (CPT) is introduced. I would like to thank Bob Esposito and his staff at John Wiley & Sons. AC impedance measurements (Chapter 5). I would like to acknowledge Mrs. In addition. Experience has been invaluable in using the book in a corrosion course in the Department of Mechanical and Aerospace Engineering at Carleton University in Ottawa. new tools for presenting books have become available. It would be a delight for me to hear from readers of this book. This edition includes introductory discussions of risk (Chapter 1). this book is being published as an electronic book. which Glenn McRae and I developed along with other members of the Canadian National Capital Section of NACE International. In particular. other books. At the end of each chapter. Inc. throughout the book. Lastly.T. I would like to quote from the Preface of the first edition of this book: If this book stimulates young minds to accept the challenge of continuing corrosion problems. Perhaps nowhere are the demands for reliability more challenging than in nuclear reactors. Throughout the book. I would like to acknowledge my many friends and colleagues at the CANMET Materials Technology Laboratory. there is a list of “General References” pertinent to that chapter. for their encouragement with this book and also with the Wiley Series in Corrosion. acknowledgments for specific material are provided throughout the book. the contributions of individual elements to endow alloys with unique properties that are valuable for specific applications are discussed. Corrosion Laboratory in the 1960s and 1970s. The important problem of corrosion of rebar (reinforced steel in concrete) is discussed in Chapter 7 on iron and steel. In addition to new technologies and new materials for managing corrosion. beginning when I was a student in the M. I would also like to thank the many organizations and individuals who have granted permission to use copyright material. who continues to encourage initiatives in corrosion education in memory of the late Professor Herbert H. Ellingham diagrams (Chapter 11). including handbooks. Winston Revie . Ottawa. Indeed.xx PREFACE natural resources caused by metal deterioration. Canada September 2007 R. it will have fulfilled the author’s major objective. this remains the main objective today. In some instances. but is described as erosion.1 DEFINITION AND IMPORTANCE OF CORROSION 1. Inc. wood may split or decay.1 Corrosion Science and Corrosion Engineering Since corrosion involves chemical change. galling. corrosive wear. Nonferrous metals. but do not rust.1 DEFINITION OF CORROSION Corrosion is the destructive attack of a metal by chemical or electrochemical reaction with its environment. “Rusting” applies to the corrosion of iron or iron-base alloys with formation of corrosion products consisting largely of hydrous ferric oxides. granite may erode. in this book. Winston Revie and Herbert H. 1 . Deterioration by physical causes is not called corrosion. by R. or fretting corrosion. but the term corrosion. and Portland cement may leach away. Uhlig Copyright © 2008 John Wiley & Sons. Nonmetals are not included in this definition of corrosion. or wear. 1. as described by the following terms: corrosion–erosion. corrode. therefore. Plastics may swell or crack. an understanding of Corrosion and Corrosion Control.1. is restricted to chemical attack of metals. chemical attack accompanies physical deterioration. the student must be familiar with principles of chemistry in order to understand corrosion reactions. Because corrosion processes are mostly electrochemical. bridges. tests and develops new and better paints. For example. Direct losses include the costs of replacing corroded structures and machinery or their components. Economic losses are divided into (1) direct losses and (2) indirect losses. in turn. synthesizes corrosion-resistant alloys. such as condenser tubes. Safety is a critical consideration in the design of equipment for nuclear power plants and for disposal of nuclear wastes. pipelines. and human. Direct losses include the extra cost of using corrosion-resistant metals and alloys instead of carbon . Other examples are (a) repainting structures where prevention of rusting is the prime objective and (b) the capital costs plus maintenance of cathodic protection systems for underground pipelines. corrosion engineers. The corrosion engineer. marine structures. boilers. prescribes proper dosage of corrosion inhibitors. metallic containers for toxic chemicals. Corrosion can compromise the safety of operating equipment by causing failure (with catastrophic consequences) of. tanks. bridges. the corrosion engineer uses cathodic protection on a large scale to prevent corrosion of buried pipelines. rebuilding corroded equipment requires further investment of all these resources— metal. that result from the corrosion of piping. metal components of machines. ships. since structure and composition of a metal often determine corrosion behavior. and so on. The corrosion scientist. turbine blades and rotors. mufflers. as well as the accompanying economic losses. aim to reduce material losses. develops better criteria of cathodic protection. 1. In addition. outlines the molecular structure of chemical compounds that behave best as inhibitors. airplane components. Furthermore. To reduce the economic impact of corrosion. Loss of metal by corrosion is a waste not only of the metal. energy. and conservation. and metal roofing. and automotive steering mechanisms. The corrosion scientist studies corrosion mechanisms to improve (a) the understanding of the causes of corrosion and (b) the ways to prevent or at least minimize damage caused by corrosion. on the other hand. and the human effort that was used to produce and fabricate the metal structures in the first place. the water. Sizable direct losses are illustrated by the necessity to replace several million domestic hot-water tanks each year because of failure by corrosion and the need for replacement of millions of corroded automobile mufflers.2 IMPORTANCE OF CORROSION The three main reasons for the importance of corrosion are: economics. water. safety. and recommends heat treatment and compositional variations of alloys that will improve their performance. Both the scientific and engineering viewpoints supplement each other in the diagnosis of corrosion damage and in the prescription of remedies. but also of the energy. or recommends the correct coating. including necessary labor. applies scientific knowledge to control corrosion. with the support of corrosion scientists. pressure vessels. for example.2 DEFINITION AND IMPORTANCE OF CORROSION electrochemistry is also important. the student should be familiar with the fundamentals of physical metallurgy as well. there are also the costs of galvanizing or nickel plating of steel. or water occur through a corrodedpipe system until repairs are made.000 or more per day for power purchased from interconnected electric systems to supply customers while the boiler is down. but a brief survey of typical losses of this kind compels the conclusion that they add several billion dollars to the direct losses already outlined. Studies of the cost of corrosion to Australia. made necessary by partial clogging of water mains with rust. In each country studied. Similarly.000 or more per hour in lost production. the cost of corrosion is approximately 3–4 % of the Gross National Product [2]. A further example is provided by internalcombustion engines of automobiles where piston rings and cylinder walls are continuously corroded by combustion gases and condensates.000. approximately $276 billion in the United States. The replacement of a corroded tube in an oil refinery may cost a few hundred dollars. 3. representing losses that may amount to billions of dollars. and of dehumidifying storage rooms for metal equipment. increased pumping capacity. according to a recent study [1]. Shutdown. Japan. Indirect losses are more difficult to assess. A small amount of copper picked up by slight corrosion of copper piping or of brass equipment that is otherwise durable . 2. in the United States. replacement of corroded boiler or condenser tubes in a large power plant may require $1. costs many millions of dollars per year. Losses of this kind cost the electrical utilities in the United States tens of millions of dollars annually. It has been estimated that. causing an explosion. of adding corrosion inhibitors to water. or because of the clogging of pipes with rust necessitating increased pumping capacity. Contamination of Product. Loss of Product. gas. Antifreeze may be lost through a corroded auto radiator. Losses of oil. Losses sustained by industry and by governments amount to many billions of dollars annually. Corrosion processes can impose limits on the efficiencies of energy conversion systems.IMPORTANCE OF CORROSION 3 steel where the latter has adequate mechanical properties but not sufficient corrosion resistance. Loss of Efficiency. and other countries have also been carried out. Loss of critical dimensions leading to excess gasoline and oil consumption can be caused by corrosion to an extent equal to or greater than that caused by wear. or 3. 4. but shutdown of the unit while repairs are underway may cost $50. Great Britain. Loss of efficiency may occur because of diminished heat transfer through accumulated corrosion products. or gas leaking from a corroded pipe may enter the basement of a building. Examples of indirect losses are as follows: 1. It has been estimated that about 25–30% of this total could be avoided if currently available corrosion technology were effectively applied [1]. The economic factor is a very important motivation for much of the current research in corrosion.1% of the Gross Domestic Product (GDP). soft waters that pass through lead piping are not safe for drinking purposes. pipelines transporting oil and gas at high pressure. more reliable estimates of equipment life can be made. Overdesign. Lead equipment. The U. . is not permitted in the preparation of foods and beverages because of the toxic properties imparted by very small quantities of lead salts. In the event of loss of health or life through explosion. condenser tubes. Similarly. the indirect losses are still more difficult to assess and are beyond interpretation in terms of dollars. boilers. unpredictable failure of chemical equipment. On recognizing the cause. or wreckage of airplanes.5 million in one year because the caps perforated by a pitting type of corrosion. thereby allowing bacterial contamination of the contents. water tanks. 5. although it is difficult to arrive at a reasonable estimate of total losses. Traces of metals may similarly alter the color of dyes. The poisonous effects of small amounts of lead have been known for a long time. For example. Bureau of Food and Drugs. The symptoms were called in his time “dry bellyache” and were accompanied by paralysis of the limbs. oil-well sucker rods are normally overdesigned to increase service life before failure occurs by corrosion fatigue. The disease originated because New England rum distillers used lead coil condensers. Indirect losses are a substantial part of the economic tax imposed by corrosion. Equipment is often designed many times heavier than normal operating pressures or applied stresses would require in order to ensure reasonable life. If the corrosion factor were eliminated. Benjamin Franklin [4] warned against possible ill effects of drinking rain water collected from lead roofs or consuming alcoholic beverages exposed to lead. using metal caps on glass food jars. Copper salts accelerate rancidity of soaps and shorten the time that they can be stored before use. Overdesign is common in the design of reaction vessels. otherwise durable.S. A cannery of fruits and vegetables once lost more than $1 million in one year before the metallurgical factors causing localized corrosion were analyzed and remedied. 1786. and the expense of recovering a lightweight rod after breakage would be lower. Another company. With adequate knowledge of corrosion. There would be further savings because less power would be required to operate a lightweight rod. or automobiles through sudden failure by corrosion of critical parts. and marine structures. the Massachusetts Legislature passed an act outlawing use of lead for this purpose. lost $0. for example. trains. In a letter to Benjamin Vaughn dated July 31. permits not more than 1 ppb of lead in bottled drinking water [3].4 DEFINITION AND IMPORTANCE OF CORROSION may damage an entire batch of soap. oil-well sucker rods. losses would be cut at least in half. and design can be simplified in terms of materials and labor. Another form of contamination is spoilage of food in corroded metal containers. Low Risk. A simplified approach to risk management. risk management must be included in the design stage. of the occurrence. On the other hand. R. Managing corrosion is an essential aspect of managing risk. Managing risk is an important part of many engineering undertakings today. Engineering design must include corrosion control equipment. repair. P. risk. and then. Moderate Risk. Consequence 4 3 2 1 A B C D Probability Extreme Risk. Risk controls may be justified. Firstly. failures with low consequence may be tolerated more frequently.3 RISK MANAGEMENT In general. Consequence is typically measured in financial terms—that is. clean-up. where both consequence and probability are high. and so on. Figure 1. the total cost of a corrosion failure.1 shows a simplified approach to risk management. that is. Maintenance must be carried out so that corrosion is monitored and significant defects are repaired. of an occurrence multiplied by the consequence. . High-Consequence Risk. including the cost of replacement. the risk of a corrosion-related failure equals the probability that such a failure will take place multiplied by the consequence of that failure.1. after operation starts. Any type of failure that occurs with high consequence must be one that seldom occurs. such as cathodic protection systems and coatings. R = P ×C Hence. so that risk is managed during the operational lifetime. is defined as the probability. downtime. Figure 1. Risk controls required. indicating qualitatively the areas of high risk. maintenance must be carried out so that risk continues to be managed. Some risk controls required. Extensive risk controls must be applied. C.RISK MANAGEMENT 5 1. The role of the change in Gibbs free energy is discussed in detail in Chapter 3. Supplement to . and n is the number of metal atoms in a molecular formula of scale.6 DEFINITION AND IMPORTANCE OF CORROSION 1. REFERENCES 1. respectively. Nevertheless. two parameters are mentioned: the change in Gibbs free energy and the Pilling–Bedworth ratio [5]. ΔG. A Pilling–Bedworth ratio greater than 1 is not sufficient to predict corrosion resistance. the Pilling– Bedworth ratio is a parameter that can be used to predict the extent to which oxidation may occur.1 Change in Gibbs Free Energy The change in Gibbs free energy. On the other hand. Koch.2 Pilling–Bedworth Ratio Although many factors control the oxidation rate of a metal. Gerhardus H.4.8. the Pilling– Bedworth ratio is 1. A film of such a corrosion product would be expected to contain cracks and pores and be relatively nonprotective. there are exceptions and limitations to the predictions of the Pilling–Bedworth ratio. the volume of the corrosion product scale is greater than the volume of the metal from which the scale formed. Y. Thompson.3. the volume of the corrosion product is less than the volume of the metal from which the product formed. H. if Md/nmD > 1. protective of the underlying metal. In this introductory chapter. n = 2. Reactions occur in the direction that lowers the Gibbs free energy. The Pilling–Bedworth ratio indicates whether the volume of the corrosion product is greater or less than the volume of the metal from which the corrosion product formed. The more negative the value of ΔG. which tends to form a nonprotective oxide.4 CAUSES OF CORROSION The many causes of corrosion will be explored in detail in the subsequent chapters of this book. for Al2O3. H. of the metal. so that the scale is in compression. the greater the tendency for the reaction to go. and J. Paul Virmani. If Md/nmD >> 1. Michiel P. The Pilling–Bedworth ratio is Md/nmD. 1. which forms a protective oxide and corrodes very slowly in most environments. of the corrosion product scale that forms on the metal surface during oxidation. Payer. the ratio is 0. m and d are the atomic weight and density.4. For aluminum. 1. Corrosion Costs and Preventive Strategies in the United States. whereas for magnesium. the scale that forms may buckle and detach from the surface because of the higher stresses that develop. where M and D are the molecular weight and density. and these are discussed in Chapter 11. Brongers. respectively. If Md/nmD < 1. for example. Neil G. for any chemical reaction indicates the tendency of that reaction to go. New York. Benjamin Franklin’s Autobiographical Writings. in Uhlig’s Corrosion Handbook. G. A manufacturer provides a warranty against failure of a carbon steel product within the first 30 days after sale. Payer. including the cost of environmental clean-up.fda. M. Virmani. S.gov/fdac/features/1998/198_lead. http://www. 1945. Wiley. Metals 29. 959–967. Oxford. Federal Highway Administration. N. Penguin Books. W. R. 2003. 13B. Gerd Gigerenzer. Kruger. 2000. VA. N. N. Economics of corrosion. 529 (1923). 2003. Report No. and Protection. 1051 (2006).. in Uhlig’s Corrosion Handbook. V. Koch. and Resources. Corrosion: Materials. Report No. OH. Bhaskaran. Corrosion Cost and Preventive Strategies in the United States. H. P.K. Geoff Davies. McLean VA. V. Viking Press. N. and M. Brongers. England. editor. in ASM Handbook. New York. Winston Revie. Rengaswamy. Thompson. Oxford. H. and J. and M. J. Elsevier. 2003. Corrosion: Fundamentals. Revie. editor. in ASM Handbook. pp. Materials Performance. Is the cost of corrosion really quantifiable? Corrosion 62 (12). Palaniswamy. pp. editor. ASM International. Bedworth. July 2002. Department of Transportation. 2nd edition. Virmani. Inst. J. GENERAL REFERENCES R. Ghali. 621–628. 13A. New York. Out of 1000 sold. 2nd edition.html Carl Van Doren. Practical Solutions. G. N. Learning to Live with Uncertainty. Biezma and J. ASM International. 2002. Federal Highway Administration. 2005. H. 10 were found to have failed by corrosion during the warranty period. Jayachandran. 3rd edition. 3. London. 2007. downtime. U. McLean. Verink. Pilling and R. pp. Kent Muhlbauer.. and replacement. Reckoning with Risk. Koch. Payer. FHWA-RD-01-156. Vol. G. U. G. Elsevier. 4. Chichester. Ohio. Testing. Vol. E. Sastri. P. loss of product. Global cost of corrosion—A historical review. S. Materials Park. H. M. 11–25. p. W. M. Brongers. R.S. 3–10. and J. PROBLEMS 1. P. Corrosion Prevention and Protection. 5. .K.PROBLEMS 7 2. March 2002. Materials Park. Cost of metallic corrosion. R. Wiley. 671. Thompson.000. Techniques. Total cost of replacement for each failed product is approximately $100. Elboujdaini. FHWA-RD-01-156. 2000. Y. 2004. San Cristóbal. H. H. Y. Pipeline Risk Management Manual: Ideas. Wiley. pp. P. Materials for Automobile Bodies. repair. D. Direct costs of corrosion in the United States. U. E. 30 (2006). is this additional cost justified based on the risk calculation in (a)? (c) Calculate the minimum percentage of failures caused by residual salt contamination at which the additional cost of $4100 for testing and removal of these salts is justified. in dollars.] (a) Calculate the risk due to this type of failure assuming that 20% of failures are caused by residual salt contamination. (b) If a corrosion-resistant alloy would prevent failure by corrosion. Monetizing the risk of coating failure. is an incremental cost of $100 to manufacture the product using such an alloy justified? What would be the maximum incremental cost that would be justified in using an alloy that would prevent failures by corrosion? 2. (b) If the cost of testing and removal of contaminating salts is $4100. The cost of reworking a failed lining of a specific tank has been estimated at $174. Between 15% and 80% of coating failures have been attributed to residual salt contamination. [Reference: H. Materials Performance 45(5).000. Peters. Linings of tanks can fail because of salt contamination of the surface that remains after the surface is prepared for the application of the lining.8 DEFINITION AND IMPORTANCE OF CORROSION (a) Calculate the risk of failure by corrosion. . with the electrical energy being supplied by chemical reactions at both electrodes. however. both surrounding the carbon electrode.2 ELECTROCHEMICAL MECHANISMS 2. They differ in this way from gaseous ions. Uhlig Copyright © 2008 John Wiley & Sons. the greater the amount of zinc that corrodes. Inc. for example. which are not hydrated. In aqueous media. Corrosion and Corrosion Control. by R.1). corrosion processes are most often electrochemical. chemical reduction occurs. It is common practice. the corrosion reactions are similar to those that occur in a flashlight cell consisting of a center carbon electrode and a zinc cup electrode separated by an electrolyte consisting essentially of NH4Cl solution* (Fig. but their number is not well-defined. and at the zinc electrode (negative pole) oxidation occurs. Zn2+·nH2O. to omit mention of the appended H2O molecules and to designate hydrated zinc ions. Winston Revie and Herbert H. †Ions in aqueous solution attach themselves to water molecules. with metallic zinc being converted into hydrated zinc ions. An incandescent light bulb connected to both electrodes glows continuously. The relationship is *The function of carbon granules for conduction and manganese dioxide as depolarizer. 9 . need not concern us at this point. At the carbon electrode (positive pole). as Zn2+. 2.† The greater the flow of electricity through the cell.1 THE DRY-CELL ANALOGY AND FARADAY’S LAW As described in Chapter 1. Any metal surface. is a composite of electrodes electrically short-circuited through the body of the metal itself (Fig. as Michael Faraday showed in the early nineteenth century. like iron. On short-circuiting the cell with a low-resistance metallic connector. 2. The value of k for zinc is 3.10 ELEC TROCHEMIC AL MECHANISMS Figure 2.1) where I is the current in amperes (A). localaction cells are able to function and are accompanied by chemical conversion of the metal to corrosion products. quantitative. Dry cell. local-action current and corrosion are not observed.2). So long as the metal remains dry. This is the relationship now known as Faraday’s law: Weight of metal reacting = kIt (2. may . the coulomb being defined as the amount of electricity represented by 1 A flowing for 1 s. But on exposure of the metal to water or aqueous solutions. Local-action current. Current of this kind is called local-action current. but acts only to heat up the surroundings.1.39 × 10−4 g/C (gram per coulomb). The slow consumption of zinc occurring on open circuit is accounted for largely by activity of minute impurities. and k is a constant called the electrochemical equivalent. in other words. similar to the situation for zinc. but when the cell is left disconnected (open circuit). t is in seconds (s). of course. these impurities assume the same role as carbon and allow the flow of electricity accompanied by corrosion of zinc. the zinc may remain intact for years. and the corresponding cells are called local-action cells. produces no useful energy. Local-action current. the zinc cup perforates by corrosion within a matter of hours. embedded in the surface of zinc. or alkalies. with the positive and negative electrode areas interchanging and shifting from place to place as the corrosion reaction proceeds. This matter is discussed in Section 7. Whenever impurities in a metal constitute the electrodes of local-action cells. for example. and positive electrodes are areas exposed to oxygen.2. a physician in Bologna. salt solutions. showing schematic arrangement of local-action cells. as was done many years ago when the electrochemical theory was first proposed.DEFINITION OF ANODE AND C ATHODE 11 Figure 2. that pure metals do not corrode at all.2 DEFINITION OF ANODE AND CATHODE A combination of two electrical conductors (electrodes) immersed in an electrolyte is called a galvanic cell in honor of Luigi Galvani. as might be expected. because impurities now enter predominantly as electrodes of local-action cells. established before anything was known about the nature of electricity. and high-purity zinc resists dilute hydrochloric acid much better than does commercial zinc. purified aluminum and magnesium are much more resistant to corrosion in seawater or in acids than are the commercial varieties of these metals. This direction of current flow follows an arbitrary convention. their removal.3. however. and is employed today . Accordingly. it is not correct to assume. account for the corrosion of metals exposed to water. appreciably improves corrosion resistance. A galvanic cell converts chemical energy into electrical energy. who published his studies of electrochemical action in 1791. local-action cells are also set up when there are variations in the environment or in temperature. A difference in rates is observed in acids. positive current flows through the metallic path from positive electrode to negative electrode. the negative electrodes are commonly portions of the iron surface itself covered perhaps by porous rust (iron oxides). Italy. high-purity iron in air-saturated water corrodes at essentially the same rate as impure or commercial iron. Accordingly. However. As we will see later. acids. 2. With iron or steel in aerated water. Metal surface enlarged. On short-circuiting such a cell (attaching a low-resistance wire to connect the two electrodes). Examples of cathodic reactions are H+ → 1 H2 − e− 2 Cu 2 + → Cu − 2e − Fe3+ → Fe2+ − e − all of which represent reduction in the chemical sense. without designating the sign of the carrier. consequently. where I is the current in amperes. Similarly. known as ions (electrically charged atoms or groups of atoms). the cathode is the positive pole. I = E/R. go from negative to positive pole. and Sn. Corrosion of metals usually occurs at the anode. as in electroplating—reduction occurs at the electrode connected to the negative pole of the external current source. The current carried by each ion depends on its mobility and electric charge. not to . positive current is always implied. The electrode at which chemical oxidation occurs (or + electricity leaves the electrode and enters the electrolyte) is called the anode. The total of positive and negative current in the electrolyte of a cell is always exactly equivalent to the total current carried in the metallic path by electrons alone. is the cathode. It is perhaps best. such as Al. which corrode rapidly on exposure to either acids or alkalies. when current is impressed on a cell from a generator or an external battery—for example. however. Electrons. Within the electrolyte. therefore. whereas the anode is the negative pole. the electrode connected to the positive pole of the generator is the anode. Ohm’s law—that is. alkaline reaction products forming at the cathode can sometimes cause secondary corrosion of amphoteric metals. Pb. or electrons. move in a metal. However.12 ELEC TROCHEMIC AL MECHANISMS despite contemporary knowledge that only negative carriers. Zn. In galvanic cells. Whenever current is said to flow. to current flow in electrolytes as well as in metals. and R the resistance in ohms—applies precisely. Nevertheless. and this electrode. under conditions with which we are presently concerned. E the potential difference in volts. current is carried by both negative and positive carriers. Examples of anodic reactions are Zn → Zn 2 + + 2e − Al → Al 3+ + 3e − Fe2 + → Fe3+ + e − These equations represent oxidation in the chemical sense. The electrode at which chemical reduction occurs (or + current enters the electrode from the electrolyte) is called the cathode. of course. opposite to the imaginary flow of positive carriers. These reactions tend to bring the two solutions to the same concentration. For example. For example. but instead to remember the cathode as the electrode at which current enters from the electrolyte and remember the anode as the electrode at which current leaves to return to the electrolyte. however. a copper pipe connected to an iron pipe. because of differing absolute surface areas of what were previously differing crystal faces. hence. Examples of dissimilar electrode cells include: the dry cell (discussed at the beginning of this chapter).. is called a differential aeration cell. There are two kinds of concentration cells. grain-boundary metal in contact with grains. Dissimilar Electrode Cells. SO2− ). and the other deaerated *The various crystal faces of a metal. The most corrodible planes of atoms react first.. the latter eventually are the only faces exposed regardless of the original orientation. Cu2+ + 2e− → Cu) on the other electrode (cathode). H+. 1.* 2.3). Fe2+) and are always positively charged whether current is drawn from or supplied to the cell. the (100) face of iron is the least reactive crystallographic plane [2]. on shortcircuiting the electrodes.e. in dilute nitric acid. These are cells with two identical electrodes. .g. although initially exhibiting different potentials (tendencies to corrode). Cations are ions that migrate toward the cathode when electricity flows through the cell (e. The first is called a salt concentration cell. each in contact with a solution of different composition.. leaving behind the least corrodible planes. if one copper electrode is exposed to a concentrated copper sulfate solution. the electrolyte around one electrode being thoroughly aerated (cathode). The second kind of concentration cell. and a bronze propeller in contact with the steel hull of a ship. tend to achieve the same potential in time when exposed to an environment that reacts with the metal [1]. but varies with environment.T YPES OF CELLS 13 remember anode and cathode as negative and positive electrodes. The most corrosion resistant crystal face of any metal is not always the same. Concentration Cells. Cl−. Similarly. 4 2.g. and a single metal crystal of definite orientation in contact with another crystal of different orientation. This situation is true whether current is impressed on or drawn from the cell.e. a metal containing electrically conducting impurities on the surface as a separate phase. 2. This may include two iron electrodes in dilute sodium chloride solution. OH−. and another to a dilute copper sulfate solution (Fig. copper dissolves (i. anions are always negatively charged (e.3 TYPES OF CELLS There are three main types of cells that take part in corrosion reactions. Dissimilar electrode cells also include coldworked metal in contact with the same metal annealed. Cu → Cu2+ + 2e−) from the electrode in contact with the dilute solution (anode) and plates out (i. which in practice is the more important. The corrosion rates continue to differ. or vice versa.. 2. 2.5) and at the water line—that is. such as seawater. The difference in oxygen concentration produces a potential difference and causes current to flow (Fig. Differential Temperature Cells. The amount of oxygen reaching the metal that is covered by rust or other insoluble reaction products is less than the amount that contacts other portions where the permeable coating is thinner or nonexistent. Differential aeration cells can also lead to localized corrosion at pits (crevice corrosion) in stainless steels. and the areas of lower oxygen concentration (inside the crevice) are anodic with respect to areas of higher oxygen concentration (outside crevices). Figure 2.3. immersed . aluminum. Differential aeration cells can also cause pitting damage under rust (Fig. Salt concentration cell.14 ELEC TROCHEMIC AL MECHANISMS Figure 2. at the water–air interface (Fig.6). This type of cell accounts for the pronounced damage at crevices. (anode) by.4). each of which is at a different temperature. 2. Differential aeration cell. at the interface of two pipes that are coupled together and at threaded connections. which is called crevice corrosion. Components of these cells are electrodes of the same metal. 3. for example. nickel. The oxygen concentration is lower within crevices. Crevices are common in many engineering designs—for example.4. bubbling nitrogen through the solution. and other passive metals that are exposed to aqueous environments. 5]. and the copper electrode at the lower temperature is the anode [3]. in an electrolyte of the same initial composition. depending on aeration. However. copper deposits on the hot electrode and dissolves from the cold electrode. but after several hours. These cells are found in heat exchangers. 2. Water-line corrosion. In copper sulfate solution. Lead acts similarly. the copper electrode at the higher temperature is the cathode. For iron immersed in dilute aerated sodium chloride solutions. most types of corrosion. resulting. immersion heaters. the hot electrode is anodic to colder metal of the same composition. but corrosion products are not necessarily observable and metal weight loss need not be appreciable . In engineering practice.T YPES OF CORROSION DAMAGE 15 Figure 2. showing differential aeration cell. On short-circuiting the cell. but for silver the polarity is reversed. corrosion damage occurs in other ways as well. occur by electrochemical mechanisms. Figure 2.4 TYPES OF CORROSION DAMAGE Corrosion is often thought of only in terms of rusting and tarnishing.6. cells responsible for corrosion may be a combination of these three types. boilers. Less is known about the practical importance and fundamental theory of differential temperature cells than about the cells previously described. Differential aeration cell formed by rust on iron. and whether the two metals are short-circuited. in failure by cracking or in loss of strength or ductility.5. stirring rate. In general. and similar equipment. for example. the polarity may reverse [4. with some exceptions. g.. These classifications are as follows: A.005 ipy)—Metals in this category have good corrosion resistance to the extent that they are suitable for critical parts.005 to 0. metals are classified into three groups according to their corrosion rates and intended application. Conversion tables are given in the Appendix. or Uniform Attack.05 ipy)—Usually not satisfactory.8. piping. for uniform attack. valve seats. C. Depth of pitting is sometimes expressed by the pitting factor. or 0. springs. mils (1 mil = 0. Conversion of mm/y to gmd or vice versa requires knowledge of the metal density. General Corrosion. fixed area of metal.13 mm/y. 25 mdd. Rates of uniform attack are reported in various units. These represent timeaveraged values. If the area of attack is relatively larger and not so deep.005 ipy. and bolt heads. and milligrams per square decimeter per day (mdd). <0. This type of corrosion includes the commonly recognized rusting of iron or tarnishing of silver. Section 29. for example. pump shafts and impellors.g. Other units that are frequently used include inches penetration per year (ipy).. B. Duration of exposure should always be given when corrosion rates are reported because it is often not reliable to extrapolate a reported rate to times of exposure far exceeding the test period. aluminum) represents a greater actual loss of metal thickness than the same weight loss for a heavy metal (e. The five main types of corrosion classified with respect to outward appearance or altered physical properties are as follows: 1. acting as anode. Generally. lead). the initial corrosion rate is greater than subsequent rates [6]. These units refer to metal penetration or to weight loss of metal. for tanks.7). when subjected to high-velocity liquids.5 mm/y (>0. valve bodies. “Fogging” of nickel and high-temperature oxidation of metals are also examples of this type. with accepted terminologies being millimeters penetration per year (mm/y) and grams per square meter per day (gmd). whereas stainless steels immersed in seawater characteristically corrode with formation of deep pits.5 gmd. excluding any adherent or nonadherent corrosion products on the surface. with the rate of corrosion being greater at some areas than at others.001 inch) per year (mpy). 2.05 ipy)—Metals in this group are satisfactory if a higher rate of corrosion can be tolerated. A given weight loss per unit area for a light metal (e. undergo a pitting . Many metals. Iron buried in the soil corrodes with formation of shallow pits. A pitting factor of unity represents uniform attack (Fig.15 mm/y (<0. Steel.15 to 1. 2. the ratio of deepest metal penetration to average metal penetration as determined by the weight loss of the specimen. >1.5 mm/y (0. for example. the resultant pits are described as deep. the pits are called shallow. 2. 0. corrodes at a relatively uniform rate in seawater of about 0. for example. This is a localized type of attack. Pitting. For handling chemical media whenever attack is uniform.16 ELEC TROCHEMIC AL MECHANISMS to result in major damage. If appreciable attack is confined to a relatively small. usually leads to a series of pits at the metal interface. Dealloying. Fig. and may appear undamaged except for surface tarnish. which results from slight relative motion (as in vibration) of two substances in contact.. Dealloying is the selective removal of an element from an alloy by corrosion. Copper and brass condenser tubes. are subject to this type of attack.4). an alloy of Au–Ag containing more than 65% gold resists concentrated nitric acid as well as does gold itself. Intergranular Corrosion. Sketch of deepest pit in relation to average metal penetration and the pitting factor. leaving a porous residue of copper and corrosion products (Fig. 261). which is the formation and collapse of vapor bubbles at a dynamic metal–liquid interface—for example. dezincification. with aluminum corroding preferentially. or a water hammer occurs. 4. 20. sometimes appearing as a honeycomb of small relatively deep fissures (see Uhlig’s Corrosion Handbook. one or both being metals. 12. for example. editor. Wiley. This type of corrosion causes a sequence of pits. The alloy so corroded often retains its original shape. p. Parting is similar to dezincification in that one or more reactive components of the alloy corrode preferentially. 3. Copper-base alloys that contain aluminum are subject to a form of corrosion resembling dezincification. resulting in loss of strength and ductility. 2nd edition. type of corrosion called impingement attack. Metal-oxide debris usually fills the pits so that only after the corrosion products are removed do the pits become visible. New York. 2000. Fretting corrosion. leaving a porous residue that may retain the original shape of the alloy. Parting is usually restricted to such noble metal alloys as gold–copper or gold–silver and is used in gold refining.g. Revie. on addition of silver to form an alloy of approximately 25% Au–75% Ag. One form of dealloying. R.7. Cavitation–erosion is the loss of material caused by exposure to cavitation. or sometimes corrosion-erosion. Dezincification. For example. but its tensile strength and ductility are seriously reduced. and Parting. W. causing the pipe to split open. Dezincified brass pipe may retain sufficient strength to resist internal water pressures until an attempt is made to uncouple the pipe. However. is a type of attack occurring with zinc alloys (e. Grain-boundary . This is a localized type of attack at the grain boundaries of a metal.T YPES OF CORROSION DAMAGE 17 Figure 2. reaction with concentrated HNO3 forms silver nitrate and a porous residue or powder of pure gold. in rotors of pumps or on trailing faces of propellers. yellow brass) in which zinc corrodes preferentially. 3rd revised edition. Tragert and W. If a metal. U. intergranular corrosion can occur because. brass. is in contact with large areas of grain acting as cathode. Massalski. Almost all structural metals (e. 210. As knowledge accumulates regarding the specific media that cause cracking and regarding the limiting stresses necessary to avoid failure within a given time period. Both stress-corrosion cracking and cracking caused by absorption of hydrogen generated by a corrosion reaction follow this definition. S. Barrett and T. nickel alloys. If a metal cracks when subjected to repeated or alternate tensile stresses in a corrosive environment.K. Distinguishing differences between the two types of cracking are discussed in Section 8.. Soc. Duralumin. subject to a constant tensile stress and exposed simultaneously to a specific corrosive environment. W. The observed cracks are intergranular or transgranular. for example.g. depending on the metal and the damaging environment. The attack is often rapid. At elevated temperatures. called the fatigue limit or endurance limit. J. carbon. The stress may be residual in the metal. will not fail by fatigue even after a very large. Highly stressed metal structures must be designed with adequate assurance that stress-corrosion cracking will not occur. under some conditions. 2. Principles and Data. Fortunately.18 ELEC TROCHEMIC AL MECHANISMS material of limited area. it will be possible to design metal structures without incidence of stress-corrosion cracking. C. or it may be externally applied. In the absence of a corrosive environment. nickel sulfide can form and cause catastrophic failures [7]. Robertson. stainless steels. titanium alloys.4. the metal stressed similarly. Structure of Metals: Crystallographic Methods. 86 (1955). The types of environment causing corrosion fatigue are many and are not specific. which proceeds without regard to whether the metal is stressed. p. when nickel-base alloys are exposed to sulfur-bearing gaseous environments. magnesium alloys. Failures of this kind differ basically from intergranular corrosion. D. or the necessary stresses are sufficiently high to limit failures of this kind in engineering practice. Improperly heat-treated 18-8 stainless steels or Duralumin-type alloys (4% Cu–Al) are among the alloys subject to intergranular corrosion. This type of attack is usually called sulfidation. 1980. B. the failure is called stress-corrosion cracking. either the damaging environments are often restricted to a few chemical species.and low-alloy steels. Electrochem. Pergamon Press. Oxford. it is said to fail by corrosion fatigue. cracks immediately or after a given time. A true endurance limit does not commonly exist in a corrosive environment: The metal fails after a prescribed number of stress cycles no matter how low the stress. phases of low melting point form and penetrate along grain boundaries.. but at values below a critical stress. acting as anode. number of cycles. penetrating deeply into the metal and sometimes causing catastrophic failures. or infinite. Cracking. and many others) are subject to stress-corrosion cracking in some environments. . REFERENCES 1. 5. 102. as from cold working or heat treatment. 4. 2000.PROBLEMS 19 3. OH. E. Bopinder Phull.6 Pitting Factor 1 2 9. D. Evaluating intergranular corrosion. 13A. N. 2005. Corros. GENERAL REFERENCES P.2 Calculate maximum penetration in millimeters for each material at the end of one year. Laboratory corrosion tests on three alloys in an industrial waste solution show the following results: Material A B C Density of Material (g/cm2) 2. Derive the general relation between mm/y and gmd. Revie and N. . Corrosion: Materials. ASM International. Corrosion 2. 7. 2003. ASM International. 140 (1950). Materials Park.7 9. Cambridge.. in Uhlig’s Corrosion Handbook. Greene. Wiley. Sci. MA. 9. W. D. V. Elliott. Vol. What is the rate in mm/y? If this corrosion rates applies to lead. Vol. 570. PROBLEMS 1. 261 (1946). H. 1950. Simpson. R. Uhlig and O. S. Magnesium corrodes in seawater at a rate of 1. Department of Chemical Engineering. New York. in ASM Handbook. Noss. Verink. Corrosion: Fundamentals.0 7.8 Weight Loss (gmd) 40 62 5. what is the corresponding rate in mm/y? 3. 5. in ASM Handbook.B. 13B. Jr. 2nd edition.45 gmd.I. H. 97–109. 631–646. Corrosion 6. Designing to prevent corrosion.T. M. pp. pp. 6. Berry. p.. Materials Park. Testing. thesis. OH. 2. Gallery of corrosion damage. and Protection. 755 (1969). . 700 J The reaction tendency is less. The more negative the value of ΔG.3 THERMODYNAMICS: CORROSION TENDENCY AND ELECTRODE POTENTIALS 3. 21 . ΔG. is measured by the Gibbs free-energy change. For example. On the other hand. Uhlig Copyright © 2008 John Wiley & Sons. Inc. by R. consider the following reaction at 25 °C: 1 Mg + H 2 O (l) + O2 (g) → Mg(OH)2 (s) 2 ΔG = −596. Or we can say that the corrosion tendency of copper in aerated water is not as pronounced as that of magnesium. we have Corrosion and Corrosion Control. Winston Revie and Herbert H. we have 1 Cu + H 2 O (l) + O2 (g) → Cu(OH)2 (s) 2 ΔG = −119. including the reaction of a metal with its environment. Finally.1 CHANGE OF GIBBS FREE ENERGY The tendency for any chemical reaction to go.600 J The large negative value of ΔG° (reactants and products in standard states) indicates a pronounced tendency for magnesium to react with water and oxygen. the greater the tendency for the reaction to go. indicating that the reaction has no tendency to go at all.3 CALCULATING THE HALF-CELL POTENTIAL—THE NERNST EQUATION Based on thermodynamics [1]. it is essential that any current that flows in the circuit is sufficiently small that the cell is not polarized—that is. an equation can be derived to express the emf of a cell in terms of the concentrations of reactants and reaction products. the greater the tendency for the overall reaction of the cell to go.2 MEASURING THE EMF OF A CELL The emf of a cell. does not corrode in aqueous media to form Au(OH)3. depending on various factors described in detail later in this book. the reaction rate may be rapid or slow. The general reaction for a galvanic cell is . >1012 Ω (ohms). at least 1012 Ω. The term ΔG can be converted from calories to joules by using the factor 1 cal = 4. a current of only 10−12 A flows.e. and the reading of the known emf indicates the exact emf of the cell. 3. E. It should be emphasized that the tendency to corrode is not a measure of reaction rate. when ΔG is positive. 3. A large negative ΔG may or may not be accompanied by a high corrosion rate. that the emf of the cell is not changed because of the current flow.184 absolute joules. must be used. In making these measurements.. so that. This applies to any of the types of cells described earlier. when the emf of the cell is exactly balanced by the known emf). correspondingly. Accordingly. the emf of the cell can be opposed with a known emf until no current flows through a galvanometer in series with the cell. where n is the number of electrons (or chemical equivalents) taking part in the reaction. is defined by ΔG = −nFE.700 J The free energy is positive. sensitive galvanometers of high input impedance. it can be stated with certainty that the reaction will not go at all under the particular conditions described. the greater the value of E for any cell. as set up in the laboratory or in the field. the tendency for a metal to corrode can also be expressed in terms of the electromotive force (emf) of the corrosion cells that are an integral part of the corrosion process. At exact balance (i. can be measured using a voltmeter of high impedance. using a potentiometer. Alternatively. For this reason.500 C/eq). but. if the potentiometer and cell are unbalanced to the extent of 1 V. Since electrical energy is expressed as the product of volts by coulombs (joules. J). and F is the Faraday (96. Such a current is not sufficient to polarize (temporarily alter the emf of) the cell.22 THERMODYNAMICS: CORROSION TENDENCY AND ELEC TRODE POTENTIALS 3 3 Au + H 2 O (l) + O2 (g) → Au(OH)3 (s) 2 4 ΔG = +65. and gold. In view of the electrochemical mechanisms of corrosion. no current flows through the cell. If ΔG is negative. the relation between ΔG in joules and emf in volts. we have the expression ΔG − ΔG = RT ln q r aQ ⋅ aR l m aL ⋅ aM (3. there is no tendency for it to go. r moles of substance R. it follows that ΔG° = −nFE°. for this reaction is given by the difference in molal free energy of products and reactants. where GQ represents the molal free energy of substance Q. The corresponding change of Gibbs free energy. and so on.16). where the symbol G° indicates standard molal free energy: ΔG = (qGQ + rGR + ) − (lGL + mGM + ) (3.2) If aL is the corrected concentration or pressure of substance L. when all the activities of reactants and products are equal to unity. and T is the absolute temperature (degrees Celsius + 273.5) On the other hand.1) A similar expression is obtained for each substance in the standard state or arbitrary reference state. the logarithm term becomes zero (ln 1 = 0). Subtracting (3. and so on. react to form q moles of substance Q. the difference of free energy for L in any given state and in the standard state is related to aL by the expression l l (GL − GL ) = lRT ln aL = RT ln aL (3. we have .3) where R is the gas constant (8.4). ΔG = (qGQ + rGR + ) − (lGL + mGM + ) (3. and ΔG = ΔG°.C ALCUL ATING THE HALF-CELL POTENTIAL—THE NERNST EQUATION 23 lL + mM + → q Q + rR + meaning that l moles of substance L plus m moles of substance M. Since ΔG = −nFE. where E° is the emf when all reactants and products are in their standard states (activities equal to unity).1) and equating to corresponding activities. and so on.314 J/deg-mole). ΔG = 0. Corresponding to (3. Hence ΔG = − RT ln K (3. ΔG.2) from (3. and q r aQ ⋅ aR l m aL ⋅ aM =K where K is the equilibrium constant for the reaction. called its activity.4) When the reaction is at equilibrium. 3. which expresses the exact emf of a cell in terms of activities of products and reactants of the cell. Similarly. for water. Section 29.1. For example.303. it is convenient to assume arbitrarily that the standard potential for the reaction .500 C/eq. except for very dilute solutions.1. the latter being a pure solid and equal.0592 V. and definitions and rules applying to the use of activity coefficients are outlined in Section 29.6) This is the Nernst equation. Table 3. approximated at ordinary pressures by the pressure in atmospheres.9. and in the Appendix. The activity of a pure solid is arbitrarily set equal to unity. the activity is set equal to unity.24 THERMODYNAMICS: CORROSION TENDENCY AND ELEC TRODE POTENTIALS E=E − q r RT aQ ⋅ aR ln l m nF aL ⋅ aM (3. If L is a gas. its activity is equal to its fugacity. of a dissolved substance L is equal to its concentration in moles per thousand grams of water (molality) multiplied by a correction factor. The activity coefficient is a function of temperature and concentration and. γ. to unity. Then. Values of activity coefficients for various electrolytes are given in the Appendix. (Zn) is the activity of metallic zinc.2.8. This coefficient appears frequently in expressions representing potentials or emf. Measured or calculated values of standard potentials φ° at 25 °C are listed in several reference books [2–4]. called the activity coefficient. T = 298. the coefficient 2. for the electrode reaction Zn 2 + + 2e − → Zn we have φZn = φZn − RT (Zn) ln 2F (Zn 2 + ) (3. must be determined experimentally. Since the emf of a cell is always the algebraic sum of two electrode potentials or of two half-cell potentials.303 RT/F at 25 °C becomes 0.1. therefore. it is convenient to calculate each electrode potential separately.7) where (Zn2+) represents the activity of zinc ions (molality × activity coefficient). with concentration essentially constant throughout the reaction. and F = 96. aL. Table 29. the value of the coefficient RT/F is multiplied by the conversion factor 2. and φZn is the standard potential of zinc (equilibrium potential of zinc in contact with Zn2+ at unit activity). Since it is more convenient to work with logarithms to the base 10. The activity.314 J/deg-mole. Section 29. Some values are listed in Section 3.8) (3.2 K.4 THE HYDROGEN ELECTRODE AND THE STANDARD HYDROGEN SCALE Since absolute potentials of electrodes are not known. from the value of R = 8. sometimes expressed as φH or φ(s. which is not separately measurable. The potential of the electrode equals zero if the hydrogen ion activity and the pressure of hydrogen gas in atmospheres are both unity. for example.CONVENTION OF SIGNS AND C ALCUL ATION OF EMF 25 Figure 3. the accepted value is −0.e. The half-cell potential for any electrode expressed on this basis is said to be on the normal hydrogen scale or on the standard hydrogen scale. for which the potential is also reversible to (in equilibrium with) hydrogen ions. The hydrogen electrode potential is measured by immersing a piece of platinized platinum in a solution saturated with hydrogen gas at 1 atm (Fig.1. Hence. or. refers to the emf of a cell with two electrodes—zinc and the standard hydrogen electrode: . by the glass electrode. are with reference to the hydrogen electrode. 2 H + + 2e − → H 2 is equal to zero at all temperatures. Hence.763 V.h. This is the standard hydrogen potential. of a zinc and hydrogen electrode in zinc salt solution of known activity of Zn2+ and H+. the standard potential. 3. φH2 = 0 − pH RT ln +2 2 (H ) 2F (3. more conveniently. By measuring the emf of a cell made up.5 CONVENTION OF SIGNS AND CALCULATION OF EMF In accord with the foregoing discussion.1). 3. φ°. All values of electrode potentials. the half-cell potential for any electrode is equal to the emf of a cell with the standard hydrogen electrode as the other electrode.10) where pH2 is the fugacity of hydrogen in atmospheres and (H+) is the activity of hydrogen ions.9) (3.). therefore. for zinc can be calculated. Hydrogen electrode. the standard potential of zinc. for products and reactants in their standard states. H + . or Zn 2 + + H 2 → Zn + 2H + standard emf = −0. If current flows within the cell from right to left.26 THERMODYNAMICS: CORROSION TENDENCY AND ELEC TRODE POTENTIALS Pt. Cu. as if it were cathode (whether it is or not is clarified later): Figure 3. H 2 .763 V (3. H +. the standard reduction potential for zinc is opposite in sign to the standard oxidation potential for zinc.13) the corresponding reaction is Zn + 2H+ → Zn2+ + H2. Zn2+. Zn. . Zn (3.2. equals 0.763 V (3. correspondingly. Zn 2 + . ΔG°. This designation of sign has the advantage of conforming to the physicist’s concept of potential defined as the work necessary to bring unit positive charge to the point at which the potential is given. indicating that the reaction is thermodynamically possible.2. and. on short-circuiting a cell. Pt Standard emf = 0. Zn 2 + . Thus. It is said to be negative to the hydrogen electrode. the positive value indicates that the reaction is not thermodynamically possible as written. It was agreed at the 1953 meeting of the International Union of Pure and Applied Chemistry that the reduction potential for any half-cell electrode reaction would be called the potential. 3. the foregoing conventions of sign dictate the direction regarding spontaneous flow of electricity. then the emf is positive. the standard emf is positive.11) The corresponding reaction. It also has the advantage of corresponding in sign to the polarity of a voltmeter or potentiometer to which an electrode may be connected. and ΔG° is negative. somewhat simplified.12) The free-energy change. On the other hand. Cu2+. is written arbitrarily subtracting the left-hand reduction reaction from the right-hand reduction reaction. we can first write the reduction reaction of the left electrode. H 2 . In setting up the emf of a cell. To calculate the emf of the cell shown in Fig. Cu. Copper–zinc cell. positive current through the electrolyte within the cell flows from left to right. the emf is negative. for the cell Zn.763 × 2F joules. zinc has a negative reduction potential and is also the negative pole of a galvanic cell of which the standard hydrogen electrode is the other electrode. Clearly. the left electrode is anode and the right electrode is cathode. If. 18) is a simplification of the actual reaction. A more exact approach would include calculation of the liquid-junction potential between CuSO4 and ZnSO4 and elimination of single-ion activities. Cu2+ plates out on the copper electrode. This fixes the true polarity of the cell. In other words.0592 (Cu 2 + ) log 2 (Zn 2 + ) (3.CONVENTION OF SIGNS AND C ALCUL ATION OF EMF 27 Cu 2 + + 2e − → Cu and φCu = 0.763 V (3. the corresponding half-cell potentials. which does depend on amount of substance reacting): emf = φZn − φCu = −1. as positive (cathode) and the right electrode.0592 1 log 2 (Cu 2 + ) (3. multiplying by any factor has no effect on either emf or φ° values.18)* The emf is obtained by adding.0592 1 log 2 (Zn 2 + ) (3.15) The reduction reaction for the right electrode is Zn 2 + + 2e − → Zn and φZn = −0. the emf E is −1.17) φ = −0.15) and (3. algebraically.14) is subtracted from (3.337 V (3. . if necessary.16) Reaction (3. Although reversing a reaction changes the sign of potential. as negative (anode). and the reaction is. not spontaneous as written.763 − 0.16). therefore. and the zinc electrode corrodes.100 − 0. with the left electrode. This results in the tentative reaction for the cell.18) is positive. by a numerical factor so that the total electrons are canceled out. (3. but goes instead in the opposite direction. when current is drawn from the cell. the free-energy change of (3. *Reaction (3.337 − φ = 0. current flows spontaneously from right to left within the cell. since the tendency for a reaction to go is independent of the amount of substance reacting (in contrast to the total free-energy change.17). Since the emf is negative. From the relation ΔG = −nFE. for which values are not measurable.100 V.14) 0. Zn. Cu. multiplying.19) If the activities of Cu2+ and Zn2+ are chosen to be the same. Cu + Zn 2 + → Cu 2 + + Zn (3. H2O → H+ + OH−.614 pH 7.01 × 10 −14 = 7. the solution is alkaline.28 THERMODYNAMICS: CORROSION TENDENCY AND ELEC TRODE POTENTIALS Similarly. for the half-cell reaction 2H+ + 2e− → H2. we have φ H2 = −0.27 7. with the pressure of hydrogen equal to 1 atm. the pH of pure water is less than 7 (Table 3.7 THE OXYGEN ELECTRODE AND DIFFERENTIAL AERATION CELL The oxygen electrode can be represented by platinized platinum immersed in an electrolyte saturated with oxygen. Therefore.6 MEASUREMENT OF pH Hydrogen ion activity is commonly expressed.77 6.115 × 10−14 0.0 If (H+) exceeds (OH−). in terms of pH. it is possible to calculate the activity of either the hydrogen ion or the hydroxyl ion from the ionization constant.0592 pH Since pure water contains equal concentrations of H+ and OH− in equilibrium with undissociated water. as in acids. above 25 °C. defined as pH = − log (H+ ) Hence. the pH of pure water at 25 °C is − log 1.916 9. The pH of strong acids can be negative. for convenience. polarity.008 2.01 × 10−14. If the pH is greater than 7. the value of which at 25 °C is 1.293 1.00 6. the emf.47 7. the ionization constant of H2O is greater than at 25 °C. and the pH of strong alkalies can be greater than 14.1. At temperatures above 25 °C. This electrode is particularly important to corTABLE 3. Ionization Constant KW and pH of Pure Water at Various Temperatures Temperature (°C) 0 10 25 40 60 KW 0. therefore. and spontaneous reaction can be determined for any cell for which half-cell reactions and standard potentials are known. 3.1). the pH is less than 7.51 . 3. it is useful to know the direction of expected potential change as.22) (3. for example.2 atm) → O2 (1 atm) 4 4 Subtacting.401 V (3. Eqs. Eqs.23) (3. (3. φ 2 − φ1 = −0.21 4 (3.0592 log Subtracting. (3. Instead.THE OX YGEN ELEC TRODE AND DIFFERENTIAL AER ATION CELL 29 rosion studies because of its role in differential aeration cells in the mechanism of crevice corrosion and pitting. Nevertheless. is positive (cathode) and the right-hand electrode.24)–(3. The observed potential tends to be less noble than the calculated reversible value.25) 1 1 H 2 O + O2 (0.2 4 (OH − ) 0.26) is positive.2 atm) + e − → OH − 2 4 φ 2 = 0. unlike the hydrogen electrode reaction. however. The potentials of the left.24) (3.21) φ = 0.401 − 0.20) This reaction.2 atm (right). and hence the measured potential may vary with time and is not reproducible.0592 log (OH − ) 11 4 (3.0103 V 14 0.22). and hence the reaction is not spontaneous as written.0592 log 11 4 0. The equilibrium for such an electrode is expressed as 1 1 H 2 O + O2 + e − → OH − 2 4 and φO2 = 0. positive electricity flows spontaneously within the cell from right to left. the left-hand electrode.26) (3. (3.and right-hand electrodes.401 − 0.22). Therefore. is negative .0592 = log 0. is not strictly reversible under practical conditions of measurement. 1 1 O2 (0.2 = −0.401 − 0. are as follows: 1 1 H 2 O + O2 (1 atm) + e − → OH − 2 4 φ1 = 0.23).0592 log (OH − ) 14 pO2 (3.27) The negative value of emf indicates that ΔG for (3. (3. For illustration. consider two oxygen electrodes in an aqueous environment: one in contact with O2 at 1 atm (left) and the other with O2 at 0.25)–(3. respectively.24). when oxygen pressure is altered. perhaps even more active than iron.1. this oxide acts as an oxygen electrode. 3. Since unit activity corresponds in some cases to impossible concentrations of metal ions because of restricted solubility of metal salts.0. The polarity of the iron– tin couple under these conditions reverses sign. the Emf Series has only limited use for predicting which metal is anodic to another. When a similar cell is made of iron electrodes instead of platinum. causing the potential of tin to become much more active. and the electrode in contact with higher-pressure oxygen tends to be the cathode. The ratio of Sn2+ to Fe2+ within the can must be very small for the reversal of polarity to occur. be larger than the small value calculated for two platinum electrodes. the electrode in contact with lower-pressure oxygen tends to be the anode. The operating emf of such a cell is much larger than that of the cell made up of platinum electrodes. In practice. Of two metals composing a cell. an adherent.27 V. certain constituents of foods combine chemically with Sn2+ ions to form soluble tin complexes. This is an appreciable value for resultant flow of current and accompanying corrosion at the anode. but the emf would. for example. when in contact with aerated solution. and the partial pressure of oxygen at the cathode equal to that of air (0. is noble to iron according to the Emf Series.440 − 0. The more negative values correspond to the more reactive metals (Table 3. This expresses the fact formulated earlier that. in any differential aeration cell.8 THE EMF AND GALVANIC SERIES The Emf Series is an orderly arrangement of the standard potentials for all metals. as can be calculated from φ° values for iron and tin in accord with the following reaction: .2). provided that the ion activities in equilibrium are both unity. the pH of water at the cathode equal to 7. the operating emf of the corresponding cell is 1. Tin.30 THERMODYNAMICS: CORROSION TENDENCY AND ELEC TRODE POTENTIALS (anode). the anode is the more active metal in the Emf Series. with the value being given by E = −0. in general.401 − 12 pO2 0.440 V). But at anodic areas. and the electrode acts as an iron electrode (φ° = −0.0592 log 2 ( Fe2 + )(OH − )2 (3. This is also the normal galvanic relationship of tin to iron in tinplate exposed to aerated aqueous media. Fe2+ forms. Reactions of this kind greatly lower the activity of Sn2+ ions with which the tin is in equilibrium.28) If the ferrous ion activity at the anode is assumed equal to 0.2 atm). But on the inside of tin-plated iron containers (“tin cans”) in contact with food. especially as approximated by an iron oxide film on iron. the emf is less than this because of the irreversible nature of the oxygen electrode. electrically conducting oxide of iron forms at cathodic areas. Position in the Emf Series is determined by the equilibrium potential of a metal in contact with its ions at a concentration equal to unit activity. 403 −0.2 0.342 −0.2.93 −3.18 −1. φ°.30) .2 −0.91 ∼−1.0592 (Sn 2 + ) log 2 ( Fe2 + ) (3.440 − 0.53 −1. Electromotive Force (Emf) Series Electrode Reaction Au3+ + 3e− = Au Pt2+ + 2e− = Pt Pd2+ + 2e− = Pd Hg2+ + 2e− = Hg Ag+ + e− = Ag Hg 2+ + 2e − = 2 Hg 2 Cu+ + e− = Cu Cu2+ + 2e− = Cu 2H+ + 2e− = H2 Pb2+ + 2e− = Pb Sn2+ + 2e− = Sn Mo3+ + 3e− = Mo Ni2+ + 2e− = Ni Co2+ + 2e− = Co Tl+ + e− = Tl In3+ + 3e− = In Cd2+ + 2e− = Cd Fe2+ + 2e− = Fe Ga3+ + 3e− = Ga Cr3+ + 3e− = Cr Zn2+ + 2e− = Zn Cr2+ + 2e− = Cr Nb3+ + 3e− = Nb Mn2+ + 2e− = Mn Zr4+ + 4e− = Zr Ti2+ + 2e− = Ti Al3+ + 3e− = Al Hf4+ + 4e− = Hf U3+ + 3e− = U Be2+ + 2e− = Be Mg2+ + 2e− = Mg Na+ + e− = Na Ca2+ + 2e− = Ca K+ + e− = K Li+ + e− = Li Standard Potential.854 0.87 −2.521 0.250 −0.987 0.70 −1.63 −1.37 −2.80 −1.136 ∼−0.763 −0.74 −0.1 −1.342 0.THE EMF AND GALVANIC SERIES 31 TABLE 3.66 −1.000 −0.29) (3.50 ∼1.53 −0.85 −2.440 −0.71 −2.789 0. in volts at 25 °C 1.136 − 0.336 −0.277 −0.05 Fe2 + + Sn → Sn 2 + + Fe and E = 0.800 0.126 −0. tend to increase the corrosion rates of many metals by reducing the metal-ion activity. These films shift the measured potential in the noble direction. In this state. the ratio (Sn2+)/(Fe2+) must be less than 5 × 10−11 for tin to become active to iron. chromium becomes anodic to iron. the reverse polarity occurs. Hence. to form specific surface films. log (Sn 2 + ) 2 = −0. commonly exhibit passivity in aerated aqueous solutions. and strong alkalies.. such as EDTA. although normally near zinc in the EMF Series. but also on their relative areas and the extent to which they are polarized (see Chapter 5). chromium becomes the cathode and current flow accelerates the corrosion of iron. Because of the limitations of the Emf Series for predicting galvanic relations.32 THERMODYNAMICS: CORROSION TENDENCY AND ELEC TRODE POTENTIALS The cell reverses polarity when E = 0. since only in this state is true equilibrium attained. The potential difference of the polarized electrodes and the conductivity of the corrosive environment determine how much current flows between them. The damage incurred by coupling two metals depends not only on how far apart they are in the Galvanic Series (potential difference). The potentials that determine the position of a metal in the Galvanic Series may include steady-state values in addition to truly reversible values. especially the transition metals of the periodic table. because of surface films. represents a nonequilibrium state in which the metal. on the contrary. the metal is said to be passive (see Chapter 6). when coupled with iron. a specific Galvanic Series exists for each environment. Although only one Emf Series exists. The passive state. The metal acts like an oxygen electrode instead of like chromium. cyanides.27 (Fe2 + ) 0. depending on whether they are active or passive. Some metals occupy two positions in the Galvanic Series. therefore.0592 or. and also because alloys are not included. 3. In the active state (e. hence. Another factor that alters the galvanic position of some metals is the tendency. especially in oxidizing environments. is no longer in normal equilibrium with its ions. alloys and passive metals are included. that is. there can be several Galvanic Series because of differing complexing tendencies of various environments and differences in tendency to form surface films. hence.304 × = −10. Many metals. whereas in the Emf Series only the active positions are possible. and the relative positions of metals in such Series may vary from one environment to another. in hydrochloric acid). In general. This small ratio can occur only through formation of tin complexes resulting from certain organic substances in foods.g. . chromium. thereby shifting metal potentials markedly in the active direction. behaves galvanically more like silver in many air-saturated aqueous solutions because of a passive film that forms over its surface. The Galvanic Series for metals in seawater is given in Fig. Complexing agents in general. Hence. The Galvanic Series is an arrangement of metals and alloys in accord with their actual measured potentials in a given environment.3. the Galvanic Series has been developed. Galvanic series in seawater [5].asminternational. The potential difference is called the liquid . All rights reserved.org.3. potential differences also arise whenever two solutions of different composition or concentration come into contact.9 LIQUID JUNCTION POTENTIALS In addition to potential differences between two metals in an electrolyte. (Reprinted with permission of ASM International®. www.) 3.LIQUID JUNCTION POTENTIALS 33 Figure 3. 00 junction potential. If the HCl solutions discussed previously are saturated with KCl. For example. except perhaps for strongly acidic or basic solutions.01 N and 0.87 mV 8. the dilute aqueous solution acquires a positive charge with respect to the concentrated solution. so that most of the current across the boundary is carried by K+ and Cl− ions.20 6. and the LiCl solution is positive. but the derivation of such calculations is relatively complex even for simple junctions.00 − 6. such as the chloride ion.01 N 33.92 2.1 N (−)HCl KCl NH4Cl NaCl (+)LiCl 35. mobility at infinite dilution is 36 × 10−4 and 7. Use of a saturated KCl solution whenever liquid junctions are formed is one practical approach to minimizing liquid junction potentials. 3.65 mV 8.57 0.65 − 8. in a junction formed between dilute and concentrated hydrochloric acid. Characteristic Liquid Junction Potentials of Salt Solutions [6] Electrolyte 0.87 6.34 THERMODYNAMICS: CORROSION TENDENCY AND ELEC TRODE POTENTIALS TABLE 3. Values at 0.92 mV. Hence. Interest is usually focused on the reaction that occurs at one electrode only. Characteristic values are given in Table 3.3. and its sign and magnitude are determined by the relative mobility of ions and their concentration differences across the liquid junction.00 Concentration 0. respectively.89 2. For potassium chloride. the observed value represents a tendency for simultaneous reactions to occur at both electrodes of the cell. the values are small and are not of great concern for most corrosion measurements. the liquid junction potential is decreased greatly.63 0. assigning zero arbitrarily to LiCl. For example. H+ ions move with greater velocity than Cl− ions. the NH4Cl solution is negative.10 REFERENCE ELECTRODES In the measurement of emf.87 = 26. liquid junction potentials between dilute and concentrated KCl are small in comparison with HCl junctions.3. for the junction LiCl(0.9 × 10−4 cm/s. On the other hand.1 N). The criterion of com- . the value is 0. hence. Potentials of junctions formed between salt solutions of the same concentration having a common ion. are additive.1 N) : NH4Cl(0.1 N) is equal to 35.78 mV with KCl solution positive and HCl solution negative. Calculations of liquid junction potentials can be made on the basis of certain assumptions.1 N) : KCl(0. the mobilities of K+ and Cl− are similar.1 N are nearly alike and. the potential of the junction HCl(0.92 = −6. which is only slightly soluble in potassium chloride solution.1 Calomel Reference Electrode The calomel reference electrode has long been a standard reference electrode used in the laboratory.4. or calomel). A type of calomel reference electrode. p. Biosensors: an Introduction. Potentials on 2 the standard hydrogen scale for various KCl concentrations are listed in Table 3. as discussed in the following paragraphs. 56. that has a fixed value of potential regardless of the environment in which it is used. Pure mercury covers a platinum wire sealed through the bottom of a glass tube. called a reference electrode. It consists of mercury in equilibrium with Hg 2+ . the activ2 ity of which is determined by the solubility of Hg2Cl2 (mercurous chloride. the latter filling the cell. 3. Any change in emf is the result of a change in potential of the electrode under study and not of the reference electrode.REFERENCE ELECTRODES 35 plete cathodic protection based on potential measurements is an example. (B. The 0.4. Eggins.10. Measurements of this kind are made by using an electrode.4. The saturated KCl electrode is convenient to prepare. Figure 3. The mercury is covered with powdered mercurous chloride. The activity of Hg 2+ depends on the concentra2 tion of KCl since the solubility product (Hg 2 + )(Cl − )2 is a constant. Copyright Wiley and Teubner. Reference electrodes use stable reversible electrode systems.) . 1996. Details on procedures for preparing reference electrodes are available elsewhere [7]. The half-cell reaction is Hg 2 Cl 2 + 2e − → 2 Hg + 2Cl − φ = 0.1 N electrode has the lowest temperature coefficient.268 V One design of electrode is illustrated in Fig. but the potential is somewhat more sluggish in responding to temperature changes than are the unsaturated KCl electrodes. 3. as shown in Fig.6 × 10−4 The values cited neglect the liquid-junction potential at the KCl boundary. and the temperature coefficient is 7 × 10−4 V/°C [8]. the following equilibrium is established: AgCl + e − → Ag + Cl − φ = 0.6. When the silver–silver chloride electrode is immersed in a chloride solution. Alternatively. .342 V For saturated copper sulfate. A schematic of a silver–silver chloride reference electrode is shown in Fig. a silver wire can be chloridized as just described.88 × 10−4 −2. for example.1 N KCl.222 V Like the calomel electrode.4. The accuracy of this electrode is adequate for most corrosion investigations. The half-cell reaction is Cu 2 + + 2e − → Cu φ = 0. the potential is more active the higher the KCl concentration.36 THERMODYNAMICS: CORROSION TENDENCY AND ELEC TRODE POTENTIALS TABLE 3. the potential is 0.2 Silver–Silver Chloride Reference Electrode The silver–silver chloride reference electrode. 3.3 Saturated Copper–Copper Sulfate Reference Electrode The saturated copper–copper sulfate reference electrode consists of metallic copper immersed in saturated copper sulfate.3 × 10−4 V/°C.10. Potentials of Calomel Reference Electrodes KCl Concentration 0. 3. increases the absolute value an average of several millivolts. even though it falls somewhat below the precision obtainable with the calomel or silver chloride electrodes. can be prepared by electroplating with silver a platinum wire sealed into a glass tube. and the temperature coefficient is −4. 3.1 N 1.2416 Temperature Coefficient (V/°C) −0. Potentials for other concentrations of KCl can be obtained by substituting the corresponding mean ion activity of Cl− into the Nernst equation.10.288 V.75 × 10−4 −6.3338 0. The silver coating is then converted partly into silver chloride by making it anode in dilute hydrochloric acid.0 N Saturated KCl φ (V) 0. It is used primarily in field measurements where the electrode must be resistant to shock and where its usual large size minimizes polarization errors.316 V.2801 0. which in the case of strong acids. the value is 0. In 0. 3.5. More details on electrode preparation are available elsewhere [7]. Antelman. Latimer. New York. G. K. W. Plenum Press. The Encyclopedia of Chemical Electrode Potentials.) Figure 3. Prentice-Hall. M. Eggins. and L. NJ. Chichester. 2. 1978. 57. 3. McGraw-Hill.5. Caroli. Randall. . Milazzo and S. Schematic of a silver–silver chloride reference electrode. for example. The Oxidation States of the Elements and Their Potentials in Aqueous Solutions. See. (B.6. S. 1982. Copper–saturated copper sulfate reference electrode. M. New York. Wiley. p. M. 1996.REFERENCES 37 Figure 3. G. Tables of Standard Electrode Potentials. Pitzer. 1961. Englewood Cliffs. Biosensors: an Introduction. 1952. Thermodynamics. 2nd edition. Brewer. Wiley and Teubner. Lewis. 4. REFERENCES 1. Assume that the corrosion products are H2 and Ni(OH)2. the solubility product of which is 1. 13A. 198–213.38 THERMODYNAMICS: CORROSION TENDENCY AND ELEC TRODE POTENTIALS 5.6 × 10−16. 7. Calculate the theoretical tendency for nickel to corrode (in volts) in deaerated water of pH = 7. 4. Calculate the half-cell potential of the hydrogen electrode in a solution of pH = 7 and partial pressure of H2 = 0. Washington. p.C. G.. New York. Wiley. D. Testing. Potential measurements with reference electodes. D. J. Materials Park. Wiley. 3rd edition. Chemical Sensors and Biosensors. Taylor & Francis. Gaskell. Corrosion: Fundamentals. OH 2003. in Uhlig’s Corrosion Handbook. Designing to prevent corrosion. ASTM G82–98(2003). Calculate the theoretical tendency of zinc to corrode (in volts) with evolution of hydrogen when immersed in 0. and Protection. Calculate the exact half-cell potential for the Ag–AgCl electrode in 1 M NaCl. pp. Verink. in ASM Handbook. 13–16. ASM International. 8. 5. 10.01 M ZnCl2 acidified to pH = 2. Corrosion: Fundamentals. in ASM Handbook. Protopopoff. 2nd edition. S. J. ASM International. Dover. Ewing. The Copper–Copper Sulfate Half-Cell for Measuring Potentials in the Earth. OH. 1995. Pennsylvania. MacInnes.5 M ZnCl2 is +0. Materials Park. New York. D. H. Eggins. p. Vol. Janz. 2003. 2. U.01 M ZnCl2. American Gas Association. Reference Electrodes: Theory and Practice. temperature is 25 °C unless otherwise stated. 99. 6. PROBLEMS In these problems. GENERAL REFERENCES B. 13A. Testing. Chichester. New York. R. The Principles of Electrochemistry. Hack. The emf of a cell made up of zinc (anode) and hydrogen electrode (cathode) immersed in 0. 2000. 1961. 1939. pp. 1961. D. E. ASTM International. 1. What is the pH of the solution? 6. New York. 7.590 V. Vol. E. Introduction to the Thermodynamics of Materials. A. 3. and Protection. Academic Press. . Calculate the value for 2..5 atm at 40 °C. Calculate the exact half-cell potential of zinc in 0. D. R.303 RT/F at 50 °C. Committee Report. p. 2002. P. Ives and G. 1939. West Conshohocken.K. Evaluating galvanic corrosion. Reinhold. 563. Standard Guide for Development and Use of a Galvanic Series for Predicting Galvanic Corrosion Performance. when CN− activity = 1. (a) Calculate the emf of the following cell: Pt. O2 (0. Calculate whether silver will corrode with hydrogen evolution in deaerated KCN solution.1 atm). What is the polarity and which electrode is anode? 11. (a) Calculate the emf of a cell made up of a hydrogen electrode ( pH2 = 1atm ) and an oxygen electrode ( pO2 = 0. to form Cu2+ (activity = 0. H 2 O. Ag + (activity = 0. pH = 0. Cu(CN)− + e − → 2CN − + Cu 2 10.01). Pt.) 14. (a) Calculate the emf of a concentration cell made up of copper electrodes in 0.PROBLEMS 39 8.446 V . What is the reaction and at what activity ratio Zn2+/Cu2+ will the reaction stop? 17. What is the corrosion tendency in volts? 15.5 atm ) in 0. 16.1). (a) Calculate the emf of the following cell at 40 °C: O2 (1 atm). (b) What is the polarity and which electrode is anode? (Assume that the oxygen electrode is reversible. Zinc is immersed in a solution of CuCl2.1 M CuSO4 and 0. calculate whether copper will corrode in deaerated KCN (activity CN− = 0. Ag 0 (b) Write the spontaneous reaction for the cell. What is the corrosion tendency in volts? (b) Similarly.0 and Ag(CN)− activity is 2 0.5 M CuSO4. Pt (b) What is the polarity and which electrode is anode? 13.1M CuCl2 to form solid AgCl. Which electrode tends to corrode when the cell is short-circuited? φ = −0. 12.1) and H2 (1 atm). (a) What is the emf of a cell made up of a copper electrode and a hydrogen electrode ( pH2 = 2 atm) in copper sulfate (activity Cu2+ = 1) of pH = 1? (b) What is the polarity of the cell and which electrode is anode? 9. Fe2 + (activity = 0.5) of pH = 10. pH = 9. Calculate whether silver will corrode if immersed in 0. neglecting the liquid-junction potential. Fe3+ (activity = 0. the 2 activity of which is 10−4. (b) Write the spontaneous reaction of the cell and indicate which electrode is anode. (a) Calculate whether copper will corrode in deaerated CuSO4. Calculate the emf of a cell made up of iron and lead electrodes in a solution of Fe2+ and Pb2+ of equal activity.001).001.5 M NaOH. assuming that Cu(CN)− is formed. −0. 4. (b) 0.440 V.] 24. A copper storage tank containing dilute H2SO4 at pH = 0.405 V. Calculate the maximum Cu2+ contamination of the acid in moles Cu2+ per liter.827 V.54 V 19. −0.40 THERMODYNAMICS: CORROSION TENDENCY AND ELEC TRODE POTENTIALS 18.28. 10.8 × 10−15. 21. . 9. in deaerated water with Fe(OH)2 as the corrosion product. 0. [Solubility product Fe(OH)2 = 1. calculate φ° for Fe3+ + 3e− → Fe. Which electrode corrodes on short-circuiting the cell? (Assume that HPbO− forms as corrosion product and that its activity is 0. (a) 0. −0. 3.1 M FeCl2. 0. What is the corresponding contamination if the hydrogen partial pressure is reduced to 10−4 atm? Answers to Problems 0.0641 V. φ° = 0. 11. Calculate the pressure (fugacity) of hydrogen required to stop corrosion of iron immersed in 0.1M CuSO4 electrode is anode. pH = 3. and Fe3+ + e− → Fe2+. 7.1 is blanketed with hydrogen at 1 atm. also assume 2 that iron is passive and that its potential approximates the oxygen electrode. (a) No.233 V.208 V.111 V. 3.1. (b) Ag is negative and is the anode.709 V. −0. Calculate the emf as in Problem 17 in an air-saturated alkaline solution of pH = 10.] 23.401 V.426 V. Calculate the pressure of hydrogen. (b) Yes. as in Problem 21. 6. 20.0 × 10−14. (a) 0. [Solubility product Cd(OH)2 = 2. Calculate the pressure of hydrogen required to stop corrosion of cadmium at 25 °C in deaerated water.307 V. Given O2 + 2H2O + 4e− → 4OH−. 8.056 V. −0. φ° = −0. 1. 22. H2 electrode is negative and is the anode.) HPbO− + H 2 O + 2e − → Pb + 3OH − 2 φ = −0. Given Fe2+ + 2e− → Fe.771 V. 2. 5. 0.0096 V. calculate φ° for O2 + 4H+ + 4e− → 2H2O. φ° = 0. with Cd(OH)2 as the corrosion product. 16. 14.6 × 10−12. 0.42 atm.019 V. Pb. 15. 42 atm. −0. (a) 0. 18. Fe.0155 V. 1.26 × 1010 atm. 1. −0. 20. 0. 13. 21. No. 22.6 × 10−8 mole/liter.84 V. 2.225 V.45 × 1037.1 atm O2 electrode is negative and is the anode. 19. emf = −0.045 V.036 V. 24. 2.23 V. 17. (b) 0. .314 V. H2 electrode is negative and is the anode. Yes.PROBLEMS 41 12. 1. 0. 23. 1. . and potential–pH diagrams are essential tools. by cathodic protection) and/or by adjusting the pH in specific domains identified using Pourbaix diagrams. By controlling potential (e. where the metal itself is the stable phase. the value of Pourbaix diagrams is their usefulness in identifying potential–pH domains where corrosion does not occur—that is.g. From a corrosion engineering perspective. In general. They have the advantage of showing at a glance specific conditions of potential and pH under which the metal either does not react (immunity) or can react to form specific oxides or complex ions. known as Pourbaix diagrams. Pourbaix diagrams indicate the potential–pH domain in which each species is stable. are now available for most of the common metals [1].1 BASIS OF POURBAIX DIAGRAMS M. by R. in the potential–pH domain labeled “Fe (immunity). Uhlig Copyright © 2008 John Wiley & Sons. which relate to the electrochemical and corrosion behavior of any metal in water. Inc. thermodynamics is an excellent starting point for many corrosion studies. For this reason.. For example. Pourbaix Corrosion and Corrosion Control. These diagrams. 43 .” iron is stable. that is.4 THERMODYNAMICS: POURBAIX DIAGRAMS 4. Winston Revie and Herbert H. and no corrosion is predicted. Pourbaix devised a compact summary of thermodynamic data in the form of potential–pH diagrams. it may be possible to prevent corrosion from taking place. 4. from the surface of an immersed electrode.4 0 –0. water is stable. Pourbaix. below which hydrogen is evolved. S. a liberation of hydrogen and alkalization 0 2 4 6 8 10 12 14 16 Figure 4.6 –2 0 2 4 6 8 10 12 14 16 liberation of oxygen and acidification b thermodynamically stable region of water at 1 atm. from the Nernst equation. Atlas of Electrochemical Equilibria in Aqueous Solutions. 2nd English edition. b.4 –0. p.8 0. For this equilibrium.1. a. and the hydrogen line. predictions must be tested experimentally and validated before using them. Between these two lines. Pourbaix diagram for water at 25°C. 100.H.303 RT 1 log 2 2F (H+ ) –2 2 E(V) 1. showing the oxygen line. The Pourbaix diagram for water is presented in Fig.E. oxygen is evolved in accord with the reaction H2O → 2 O2 + + − 2H + 2e . 1 4.) 0. the relationship between potential and pH is. Above line b.8 –1. φ = φ − 2. (M.1.2 POURBAIX DIAGRAM FOR WATER Each line of a Pourbaix diagram represents conditions of thermodynamic equilibrium for some reaction. As every good engineer knows.2 –1. and predictions arrived at using Pourbaix diagrams are no exception to this general rule.44 THERMODYNAMICS: POURBAIX DIAGR AMS diagrams continue to be developed for interpreting corrosion studies in specific systems of engineering importance [2]. copyright NACE International 1974 and CEBELCOR.6 1.2 Potential (volts. however. above which oxygen is evolved.) . that is. then φ = −0. a horizontal line on the Pourbaix diagram. For this equilibrium. R = 8. 4.2. Fe2+ + 2e− → Fe. T = 298. using the Nernst equation. we have φ = 1.2).0296 log (Fe2 + ) If (Fe2+) is taken as 10−6.0592 pH Above line b. the relationship between potential and pH is φ = φ − 2. defined by this equation. we obtain φ = −0. hydrogen gas is evolved from the surface of an immersed electrode. A vertical line involves H+ or OH−.9). 2Fe3+ + 3H2O → Fe2O3 + 6H+. A horizontal line represents a reaction that does not involve pH.229 (Appendix. For this equilibrium.0592 pH 2 2F (H + ) With φ° = 0 (Section 3. but not electrons. In Figure 4. Table 3.2 K. the vertical line separating Fe3+ from Fe2O3 corresponds to this reaction. water is stable.3 POURBAIX DIAGRAM FOR IRON The Pourbaix diagram for iron is presented in Fig. Using the Nernst equation for this equilibrium. neither H+ nor OH− is involved. water is stable. we have K= (H + )6 ( Fe3+ )2 log K = 6 log (H+ ) − 2 log ( Fe3+ ) . Between lines a and b.229 − 0. and F = 96. for example.440 + 0. we obtain φ = φ − 2. as in the reaction. hydrogen is evolved in accord with the reaction 2H+ + 2e− → H2. Below line a.8.303 RT 1 log = −0.2.POURBAIX DIAGR AM FOR IRON 45 With φ° = 1. represented by this equation.500 C/eq. Below this line. Section 29.0592 pH Below line a.617 V. oxygen is evolved at the surface of an immersed electrode. 4.314 J/deg-mole.303 RT 1 log nF (Fe2 + ) φ = −0. and corrosion is expected to take place without any protection afforded by a surface oxide film. For example. A sloping line involves H+. and electrons.. considering Fe. at pH < 1.2.72 − 3 pH Taking (Fe3+) = 10−6.76). at pH > 1. we obtain log K = 1. the vertical line at pH 1. Fig. In the Pourbaix diagram for iron. as a protective film. For this reaction. Fe2O3 is the stable phase.e. Fe3O4. Pourbaix diagram for the iron–water system at 25°C.76). ferric ions in solution are stable. and this oxide. we have φ = φ − 2. we have pH = 1.2. 4.76. To the left of this line (i. 2Fe3+ + 3H2O → Fe2O3 + 6H+. would be expected to provide some protection against corrosion..e. To the right of this line (i.76 represents the equilibrium reaction. log K = −6 pH − 2 log (Fe3+ ) Since ΔG° = −RT ln K and ΔG° = −8240 J/mole.43 log (Fe3+ ) = −0. and Fe2O3 as the only solid substances.303 RT (Fe2 + )2 log nF (H + )6 . OH−. the sloping line separating Fe2+ from Fe2O3 represents the reaction Fe2O3 + 6H+ + 2e− → 2Fe2+ + 3H2O.46 THERMODYNAMICS: POURBAIX DIAGR AMS Figure 4. 4). is the stable phase between about pH 4 and 9 [3].0296 ( Fe2 + ) φ = 0. Thus the horizontal line at −0. in general. Actually.0592 0. above which passivity of iron is observed in media such as sulfuric or nitric acid.44 + (0. Indeed.728 − 0.4 POURBAIX DIAGRAM OF ALUMINUM The Pourbaix diagram of aluminum. To the left of this line. This value in some instances (e. When any reaction involves ions other than H+ or OH−.62 V. the passive film (Definition 1. This diagram also predicts the amphoteric nature . indicates that hydrargillite. the lines separating the phases are shifted (as can be seen in Figs.2 represents the equilibrium.3. Soluble ferrates ( FeO2− ) can form in alkaline 4 solutions at very noble potentials. this film is considered to be responsible for the successful use of aluminum in many structural applications. we obtain φ = 1. that the activity equals 10−6.6 V at pH = 0. Section 6.0592 log (Fe2 + ) − 0. φ = −0. Fe in aerated H2O) differs from that of the bulk solution. 4.POURBAIX DIAGR AM OF ALUMINUM 47 Since φ° = 0.3 and 4. it is assumed.728 − 0.g. Section 6. intersecting 0. If a value other that 10−6 is used.728 V and n = 2. parallels line a and b.0592 log (Fe2 + )2 + log (H + )6 2 2 φ = 0. presented in Fig. This is correct only insofar as passivity is accounted for by a diffusion-barrier oxide layer (Definition 2.082 − 0.62 V means that iron will not corrode below this value to form a solution of concentration > 10−6 M Fe2+ in accord with Fe → Fe2+ + 2e−. Soluble hypoferrites (HFeO− ) can form in very alkaline solutions within a 2 restricted active potential range. The fields marked Fe2O3 and Fe3O4 are sometimes labeled “passivation” on the assumption that iron reacts in these regions to form protective oxide films.1) may not be any of the equilibrium stoichiometric iron oxides. 4. Fe2O3 + 6H+ + 2e− → 2Fe2+ + 3H2O. 4. but the stable field is not well-defined. The pH values in Pourbaix diagrams are those of solution in immediate contact with the metal surface.1776 pH Taking (Fe2+) = 10−6. Al2O3·3H2O. For this reason.728 − 0. we get φ = 0.1). as is further discussed in Chapter 6. To the right of this line..059/2)log (10−6) = −0. Fe2+ is a stable species in solution.1776 pH This line in Fig. the Flade potential. Fe2O3 is a stable phase that is expected to form a surface oxide film that protects the underlying metal from corrosion. 4. 3. a stable film of Mg(OH)2 protects magnesium from corrosion. Pourbaix. Pourbaix diagram for the aluminum–water system at 25°C [2]. magnesium reduces water. with corrosion under highly acidic and highly alkaline conditions and protection by an oxide film between about pH 4 and 9.48 THERMODYNAMICS: POURBAIX DIAGR AMS Figure 4. The potential–pH domain in which magnesium is stable is well below the domain in which water is stable. 2nd English edition. magnesium cannot be prepared by electrolysis of its aqueous solutions.4) contrasts with that of aluminum in that Mg2+ is the stable species in solution over most of the potential–pH range [4]. Between pH about 8. At pH greater than about 11.) of aluminum. (M. which would evolve hydrogen without forming magnesium.5. Mg2+ + 2e− → Mg. .5 POURBAIX DIAGRAM OF MAGNESIUM The Pourbaix diagram of magnesium (Fig. and hydrogen evolution is the cathodic reaction. Behavior of aluminum is discussed further in Chapter 21. p. therefore. 171. 4. a protective film of oxide or hydroxide may provide some protection. 4.5.5 and 11. copyright NACE International 1974 and CEBELCOR. Because of the very active value of the equilibrium potential of the reaction. Atlas of Electrochemical Equilibria in Aqueous Solutions. they convey no information on rates of reaction—that is. for example. Pourbaix diagram for the magnesium–water system at 25°C [3].6 LIMITATIONS OF POURBAIX DIAGRAMS Since Pourbaix diagrams are based on thermodynamic data. whether any possible reaction proceeds rapidly or slowly when the energy changes are favorable. such as SO2− or Cl−.4. the diagrams do clearly outline the nature of the stoichiometric compounds into which any less stable compounds may transform whenever equilibrium is achieved. copyright NACE International 1974 and CEBELCOR. With these limitations. they do not indicate 4 the detailed conditions under which nonstoichiometric metal compound films are possible. . some of these films are important in determining observed corrosion rates (see Chapter 6). the diagrams usefully describe the equilibrium status of a metal either as immersed in acids or alkalies or when a given potential is impressed on it. However. p. (M.) 4. the corrosion of iron–nickel alloys cannot be accurately predicted by superimposing conventional Pourbaix diagrams for iron and nickel [5]. Pourbaix. 2nd English edition.LIMITATIONS OF POURBAIX DIAGR AMS 49 Figure 4. Similarly. but they provide no measure of how effective such barrier films may be in the presence of specific anions. The data indicate the conditions for which diffusion-barrier films may form on an electrode surface. 141. Atlas of Electrochemical Equilibria in Aqueous Solutions. Conventional Pourbaix diagrams do not include alloys. 2nd English edition.06 ppm iron in each of the following potential ranges. TX. W. Pourbaix diagrams for multielement systems. GENERAL REFERENCES Marcel Pourbaix. 2000. and V. D. 1974. Wiley. J.2). 625 (2007). 1. 2. 125–136. 2nd edition. editor. Pelton. in Uhlig’s Corrosion Handbook. translated from the French by James A. M. Belgium. and A. −0. Thompson. Wiley. H.. Texas. editor.4 V. 171. 125–136. National Association of Corrosion Engineers. W. Pelton.2). Simplified procedure for constructing Pourbaix diagrams.50 THERMODYNAMICS: POURBAIX DIAGR AMS REFERENCES 1. Kaye. E. Potential ranges: (a) (b) (c) (d) (e) φ < −0. J. W. list equilibrium reactions corresponding to lines separating (a) Fe and Fe3O4. Wiley. New York. in Uhlig’s Corrosion Handbook.. J. translated from the French by James A. 4. p. R. 3. 2. D. New York. T. Kaye. Revie. Verink.7 V. Atlas of Electrochemical Equilibria in Aqueous Solutions. 2nd English edition. W. in Uhlig’s Corrosion Handbook.2 V < φ < +0. 141. Pourbaix diagrams for multielement systems. C. M. 2000. L.6 < φ < −0. and Centre Belge d’Etude de la Corrosion (CEBELCOR). C. National Association of Corrosion Engineers. W. 2nd edition.4 V < φ < −0. W. Revie. 4. determine the direction of each of the following four reactions in solution of pH 7 containing 0. pp. T.80 V < φ . Bale. New York. Houston. W. 4. Franklin. −0. Atlas of Electrochemical Equilibria in Aqueous Solutions. −0. p. Revie. 2nd edition. (b) Fe3O4 and Fe2O3. 1974. pp. Corrosion 63 (7).. Bale. Brussels. R. Referring to the Pourbaix diagram for iron (Fig. Ref. and A. M. Muñoz-Portero. 5. Brussels. Houston. +0. 2000. Thompson.2 V. and (c) Fe2+ and Fe2O3. Calculate slopes ∂φ/∂pH in each case and compare with the diagram. Jr. H. pp. 111–124. Using the Pourbaix diagram for iron (Fig. R. PROBLEMS 1. García-Antón. Pérez-Herranz. D. Guiñón.8 V. Franklin. Marcel Pourbaix. editor. Belgium. Ibid. and Centre Belge d’Etude de la Corrosion (CEBELCOR). PROBLEMS 51 Reactions: Fe2 + + 2e − → Fe 2 H + + 2e − → H 2 Fe2 O3 + 6 H + + 2e − → Fe2 + + 3H 2 O O2 + 4 H + + 4e − → 2 H 2 O Answers to Problem 1. (a) Fe3O4 + 8H+ + 8e− → 3Fe + 4H2O ∂φ/∂pH = −0.059 (b) ∂φ/∂pH = −0.18 . . Winston Revie and Herbert H.5 KINETICS: POLARIZATION AND CORROSION RATES 5. In general.. Instead. we must know the equilibrium state of the system before we can appreciate the various factors that control the rate at which the system tends toward equilibrium. that is. a fundamental approach to nonequilibrium states. The measured potential of such an electrode is altered to an extent that depends on the magnitude of the external current and its direction. Inc. It must not be concluded. Uhlig Copyright © 2008 John Wiley & Sons. along with calculation of corrosion rates. 53 . aluminum) nevertheless react so slowly that they meet the requirements of a structural metal and may actually be more resistant in some media than other metals that have inherently less tendency to react. An electrode is not at equilibrium when a net current flows to or from its surface. begins with the primary consideration that equilibrium has been disturbed. however. by R. therefore. the rate of corrosion. Some metals with pronounced tendency to react (e. that considerations of equilibria are irrelevant to the study of corrosion. In practice.1 POLARIZATION Thermodynamics—the equilibrium between metals and their environments—has been discussed in the previous chapters. we are concerned primarily with rates of corrosion—that is. The direction Corrosion and Corrosion Control. The concept of corrosion tendency is based on thermodynamics.g. however. with kinetics. . 5. exchange current density. of necessity. the anode is always more cathodic in potential and the cathode always becomes more anodic. with the difference of potential between the anode and cathode becoming smaller as current is increased. the electrodes of which are connected to a variable resistance R. the measured potential difference falls below 1 V because both electrodes polarize. hence.1. voltmeter V. Important concepts of electrode kinetics that will be introduced in this chapter are the corrosion potential (also called the mixed potential and the rest potential).1. opposes the flow of current. is called polarization. On complete short-circuiting (very small external resistance). maximum current flows and the potential difference of copper and zinc electrodes becomes almost zero. as shown in Fig. The potential difference (emf) of zinc and copper electrodes of the cell without current flow is about 1 V. The potential change caused by net current to or from an electrode.2 THE POLARIZED CELL Consider a cell made up of zinc in ZnSO4 solution and copper in CuSO4 solution (the Daniell cell). When current flows in a galvanic cell. corrosion current density. The interpretation of corrosion processes by superimposing electrochemical partial processes was developed by Wagner and Traud [1]. If a small current is allowed to flow through the external resistance. the reader should refer to specialized texts listed at the end of the chapter. For more detailed discussion of electrode kinetics. Polarized copper–zinc cell. Figure 5. 5. measured in volts. and ammeter A.54 KINETICS: POL ARIZATION AND CORROSION R ATES of potential change always opposes the shift from equilibrium and. elementary and directed toward application of corrosion science. The science of electrode kinetics has made possible many advances in the understanding of corrosion and the practical measurement of corrosion rates. Electrode kinetics is the study of reaction rates at the interface between an electrode and a liquid. The treatment of electrode kinetics in this book is. for example. whether the current is impressed externally or is of galvanic origin. and Tafel slope. The voltage continues to fall as the current increases. 500 C/eq. b – e. Obviously. On shortcircuiting. the polarization of copper is given by the difference of potential e – d. The maximum current is equivalent to (65.2. only if a method is introduced for reducing the polarization of zinc or copper. or I1(Re + Rm). Polarization diagram for copper–zinc cell. the current becomes maximum. φ. Imax.38/2)Imax/F grams zinc corroding per second.3. Similarly. 5. Similarly. the polarization of zinc in volts is given by the difference between the actual potential of zinc at b and the thermodynamic value a or φZn. the zinc electrode polarizes along curve abc. any factor tending to increase polarization will decrease current through the cell and decrease the corresponding corrosion rate of zinc.38/2 is the equivalent weight of zinc. Re. The effect of net current flow on voltage of the Daniell cell can be represented by plotting a polarization diagram—that is. although they can approach each other very closely if anodes and cathodes are closely spaced in media of moderate . of copper and zinc electrodes (as described in Section 5. “How Polarization Is Measured”) with total current. The corrosion rate of zinc can exceed the indicated equivalent corrosion rate.2. and 65.2. The cathodic reaction corresponds to the identical chemical equivalents of copper depositing per second on the cathode. as shown in Fig. the polarization curves can never actually intersect. where Imax is in amperes. and the potential difference of both electrodes decreases to a minimum equal to ImaxRe. thereby reducing the slopes of abc or def. In Fig. or both. The potential difference of the polarized electrodes. in series. 5. I. a graph showing potential. is equal to the current i1 multiplied by the total resistance of both the external metallic resistance Rm and the internal electrolytic resistance.THE POL ARIZED CELL 55 Figure 5. Imax. At a value of current through the ammeter equal to I1. causing an approach to intersection at a larger value of I. and the copper electrode polarizes along curve def. Then Rm can be neglected. The thermodynamic potentials (no current flow through the cell) are given by φZn and φCu. F is equal to 96. In order to achieve uniform current density at B. The probe L. The measured potential of a corroding metal.3(a). The emf of cell B–R is recorded for each value of current as read on ammeter A. the distance of the tip of the Luggin capillary from the electrode being studied [the distance between L and B in Fig. in which the working electrode (the electrode being studied). and hence the corrosion rate per unit area can always be expressed as a current density. and thermometer are all included [2]. the mixed potential of both polarized anodes and cathodes.0400 A/m2.3 HOW POLARIZATION IS MEASURED Referring to Fig. showing a two-compartment cell separated by a sintered glass disk. 5.3(b) shows the type of cell that is commonly used in corrosion studies to measure polarization. with subsequent extrapolation to zero distance.3(a)] can have a significant effect on IR drop that arises because of the current flow through the electrolyte. the relative anode–cathode area ratio for the corroding metal must also be known. the corrosion rate of anodic areas on a metal surface is proportional to Icorr. with B in the center and two auxiliary electrodes in the outer compartments. is known as the corrosion current. This correction. The Luggin capillary can disturb the distribution of the current applied to electrode (B) in Fig. Referring to Fig. assume electrode B to be polarized by current from electrode D. allowing sufficient time for steady-state conditions. gas inlet and outlet. 5. Imax. φcorr. a corrosion rate of 1 gmd is equivalent to 0. 5. In general. By Faraday’s law. whether anode or cathode.8. is recorded in volts with reference to half-cell electrode R for various values of current density. There will always be a finite potential difference accompanying an observed flow of current. of reference cell R (or of a salt bridge between R and B) is placed close to the surface of B. G. Luggin capillary. thereby minimizing extraneous potentials caused by IR drop through the electrolyte. For zinc.2. particularly in low conductivity solutions. Electrolytic cells that account for the corrosion of metals are analogous to the short-circuited cell. is also referred to as the corrosion potential. Figure 5. 5. since polarization data are usually obtained under conditions where the electrode surface is all anode or all cathode. sometimes called the Luggin capillary or salt bridge. The potentials are often converted to the standard hydrogen scale.0342 A/ m2. the corresponding value is 0. as shown in . Potentials can be measured with probe L adjusted at various distances from B.56 KINETICS: POL ARIZATION AND CORROSION R ATES to good conductivity. Values for other metals are listed in the Appendix (Section 29. we can calculate the corrosion rate of a metal if data are available for the corrosion potential and for the polarization behavior and thermodynamic potential of either anode or cathode. Polarization of B.2). two (for uniformity of current flow) counter electrodes.3(a). In addition. a three-compartment cell can be used. For Fe corroding to Fe2+. The value. 5. Icorr. HOW POL ARIZATION IS MEASURED 57 (a) (b) Figure 5. This correction. (Copyright ASTM INTERNATIONAL. is needed only when the measurements require highest accuracy. however. does not include any high-resistance reaction product film on the surface of the electrode. (a) Cell for measuring polarization. as in distilled water.3. . Reprinted with permission. (b) Schematic diagram of commonly used polarization cell [2]. when the current densities are unusually high.) the next paragraph. or when the conductivity of the electrolyte is unusually low. This value is negligible in establishing the critical minimum current density for adequate cathodic protection.3.58 KINETICS: POL ARIZATION AND CORROSION R ATES 5. where i is the current density.0592 log (Cu 2 + )s 2 (5.337 + 0.1 Calculation of IR Drop in an Electrolyte The resistance of an electrolyte solution measuring l cm long and S cm2 in cross section is equal to l/κS ohms.05 Ω−1 cm−1. or more active.0592 log (Cu 2 + ) 2 (5. the smaller the surface concentration of copper ion. the IR drop in volts equals il/κ. then the potential φ1. In practice. In some soft waters. Hence. is the concentration polarization.05 = 0. 5. activation polarization. thus the larger the corresponding polarization.0592 (Cu 2 + ) log 2 (Cu 2 + )s (5. The potential φ2 of the electrode becomes φ 2 = 0. 1. Concentration Polarization (Diffusion Overpotential). Infinite concentration polarization is approached when (Cu2+)s approaches zero at the electrode surface. The difference of potential.337 + 0. where κ is the specific conductivity. κ = 0. or the smaller the value of (Cu2+)s. φ2 − φ1.1 A/m2)(the magnitude of current density that might be applied for cathodically protecting steel) produces an IR drop correction for a 1-cm separation of probe from cathode equal to (1 × 10−5 V)/0. and IR drop.2 mV. a current density of 1 × 10−5 A/cm2 (0. the corresponding IR drop equals 1 V/cm. the potential of the polarized cathode is less noble. where κ may be 10−5 Ω−1 cm−1. If copper is made cathode in a solution of dilute CuSO4 in which the activity of cupric ion is represented by (Cu+2). is given by the Nernst equation φ1 = 0.4 CAUSES OF POLARIZATION There are three types of electrode polarization: concentration polarization.3) The larger the current. copper is deposited on the electrode. however.2) Since (Cu2+)s is less than (Cu2+). equal to φ 2 − φ1 = − 0. therefore. than in the absence of external current. in absence of external current.1) When current flows. For seawater. . the corresponding current density that results in this limiting lower value of (Cu2+)s is called the limiting current density. thereby decreasing surface concentration of copper ions to an activity (Cu2+)s. 5) where D is the diffusion coefficient for the ion being reduced. If iL is the limiting current density for a cathodic process and i is the applied current density. another electrode reaction establishes itself at a more active potential than corresponds to the first reaction. the potential moves to that for hydrogen evolution. In the case of copper deposition. that is. This is shown by the plot of φ2 − φ1 versus i in Fig.4. The limiting current density (A/cm2) can be evaluated from the expression iL = DnF c × 10 −3 δt (5. 2H+ + 2e− → H2. n and F have their usual significance. for example. and hydrogen gas is liberated while copper is simultaneously plated out.4. t is the transference number of all ions in solution except the ion being reduced (equal to unity if Figure 5. Dependence of concentration polarization at a cathode on applied current density. minus infinity. .4) As i approaches iL. it can be shown [3] that φ 2 − φ1 = − RT i ln nF iL − i (5.05 cm in an unstirred solution). φ2 − φ1 approaches −∞.CAUSES OF POL ARIZATION 59 polarization can never reach infinity. δ is the thickness of the stagnant layer of electrolyte next to the electrode surface (about 0. instead. 5. Activation polarization is caused by a slow electrode reaction. The most important example is that of hydrogen ion reduction 1 at a cathode. the following reactions are thought to occur in sequence: H+ + e − → Hads where Hads represents hydrogen atoms adsorbed on the metal surface. η = φmeansured − φ reversible At a platinum cathode.2 × 10−5 cm2/s. averages about 1 × 10−5 cm2/s. Should the copper electrode be polarized anodically. opposite to the direction of potential change when the electrode is polarized as cathode. is the difference between the measured potential and the corrosion (or mixed) potential.02 nc (5. is defined as polarization (potential change) of a corroding electrode caused by flow of an applied current. concentration of copper ion at the surface is larger than that in the body of solution. This relatively rapid reaction is followed by another reaction.3 × 10−5 and 5. 2. Since D for all ions at 25oC in dilute solution. adsorbed hydrogen atoms combining to form hydrogen molecules and bubbles of gaseous hydrogen: 2 Hads → H 2 *Overvoltage. the limiting current density is approximated by iL = 0.60 KINETICS: POL ARIZATION AND CORROSION R ATES many other ions are present). the limiting upper value for concentration polarization corresponds to formation of saturated copper salts at the electrode surface. is the difference between the measured potential and the thermodynamic. and c is the concentration of diffusing ion in moles/ liter. the polarization is called hydrogen overpotential. potential. Overpotential is defined as the polarization (potential change) of an equilibrium electrode that results from current flow across the electrode–solution interface. respectively (infinite dilution).3) changes sign. namely. For this reaction. on the other hand. that is. . Activation Polarization. or reversible. H+ + e− → 2 H2. The ratio (Cu2+)/ (Cu2+)s then becomes less than unity and φ2 − φ1 of (5. ε = φmeasured − φcorr. Overpotential. ε. For a copper anode. The reaction at the electrode requires an activation energy in order to proceed. overvoltage. that is.* that is. In other words. This limiting value is not as large as for cathodic polarization where the Cu2+ activity approaches zero. η. so that the corresponding values of iL are higher. D equals 9. concentration polarization at an anode polarizes the electrode in the cathodic or noble direction.6) For H+ and OH−. for example. except for H+ and OH−. hydrogen overpotential values are negative and. but the values at a given current density are much smaller than those for O2 or H2 evolution. or both. such as iron. . Although usually listed as positive. Activation polarization is also characteristic of metal-ion deposition or dissolution. The anion associated with the metal ion influences metal overpotential values more than in the case of hydrogen overpotential. The exchange current density. Tafel [Z. for example. i1. it is given by η. the difference between measured and equilibrium potentials.1). but. Chem. but varies with metal. through a metal-reaction product film on the surface. nickel.CAUSES OF POL ARIZATION 61 This reaction is relatively slow. Phys. A typical plot of activation polarization or overpotential for H+ discharge is shown in Fig. At the equilibrium potential for the hydrogen electrode (−0. 641 (1904)]. The controlling step in the reaction is not known precisely. overpotential is zero. correspondingly. the smaller the corresponding overpotential. i0. 3. who first proposed a similar equation to express hydrogen overpotential as a function of current density. and zinc. oxygen overpotential values are positive on the φ scale. 50.5. IR Drop. but it is larger for the transition metals. represents the current density equivalent to the equal forward and reverse reactions at the electrode at equilibrium. or dehydration of the hydrated ion as it enters the lattice. copper. in accord with the Tafel equation*: η = β log i i0 where β and i0 are constants for a given metal and environment and are both dependent on temperature. where I is the *Named after J. Overpotential may also occur with Cl− or Br− discharge. such as silver. The value may be small for nontransition metals. cobalt. it is probably a slow rate of hydration of the metal ion as it leaves the metal lattice. At applied current density. in some cases. and its rate determines the value of hydrogen overpotential on platinum. i. and environment. This contribution to polarization is equal to iR. The controlling slow step of H+ discharge is not always the same. Polarization measurements include a so-called ohmic potential drop through a portion of the electrolyte surrounding the electrode. and chromium (see Table 5.059pH). 5. The larger the value of i0 and the smaller the value of β. An ohmic potential drop always occurs between the working electrode and the capillary tip of the reference electrode. current density. Activation polarization. η. Pronounced activation polarization also occurs with discharge of OH− at an anode accompanied by oxygen evolution: 1 2OH − → O2 + H 2 O + 2e − 2 The polarization corresponding to this reaction is called oxygen overpotential. increases with current density. 10 0.10 0.40 (Stern) 0.20(Bockris) 0.68(Bockris) 1 mA/cm2 = 10 A/m2.47 0.1 N H2SO4 0.05 9 × 10−8 4 × 10−9 5 × 10−9 0.12 0.40 0.10 1.09 0.1 N NaOH 0.15 0.16 0.75 0. Academic Press.68 2 10−2 10−2 10−1 10−1 5 × 10−3 8 × 10−3 4 × 10−3 10−3 10−3 10−2 10−3 2 × 10−3 1 × 10−2 10−5 10−6 10−5 10−4 10−3 1.81 0. Conway.13 0.02 0.12 0.12 0. J. 1952. Overpotential Values η = β log Temperature (°C) i i0 β (volts) i0 (A/m2) η at 1 mA/cm2 (V)a Metal Solution Hydrogen Overpotential Pt (smooth) Pd Mo Au Ta W Ag Ni Bi Nb Fe Cu Sb Al Be Sn Cd Zn Hg 20 25 20 20 20 20 20 20 20 20 20 20 16 25 20 20 20 20 20 20 16 20 20 20 20 20 1 N HCl 0. 1095 (1966).1. Soc.12 0.15 0. Electrochem.36 0. editor. 113.12 0.31 0.16 Pb Oxygen Overpotential Pt (smooth) Au 20 20 20 0.12 0.10 0.12 N NaOH 1 N HCl 1 N HCl 1 N HCl 4% NaCl pH 1–4 0. New York.03 0.01–8 N HCl 0. New York.34 0. Source: Data from B. Modern Aspects of Electrochemistry.72 0. J.60 0. 609 (1955). Bockris. 1954.40 0.1 N HCl 0. J. Electrochem.10 0.1 N NaOH 0. E. Elsevier.47 Metal Overpotential (Deposition) Zn Cu Fe Ni a 25 25 25 25 1 M ZnSO4 1 M CuSO4 1 M FeSO4 1 M NiSO4 0.12 10 0.6 N HCl 1 N HCl 1 N HCl 1 N HCl 5 N HCl 0.2 0.12 0.03 0.04 0.22 0.70 0.11 0.1 N NaOH 0.10 0.62 KINETICS: POL ARIZATION AND CORROSION R ATES TABLE 5.1 N HCl 0. Kita. 102.10 0.08 0.1 N HCl 0.16 1.94 1.1 N NaOH 0.80 0.44 0. Soc.12 0.6 × 10−7 7 × 10−9 2 × 10−9 3 × 10−11 2 × 10−9 0.10 0.1 N H2SO4 0. M.15 1.05 0.12 0.20(Bockris) 0.15 N NaOH 2 N H2SO4 2 N H2SO4 1 N HCl 1 N HCl 1 N HCl 1 N H2SO4 0.15 0.11 0.12 0.10 0.60(Bockris) 0.00 0.12 0. Electrochemical Data.45 0. Stern.1 N HCl 0. H.10 0. .30 0.20 0.2 10−4 2 × 10−5 0.05 0. hydrogen overpotential is the difference of potential between a cathode at which hydrogen is being evolved. Note: Concentration polarization decreases with stirring. and R. The product. current density. φmeasured. hydrogen overpotential includes only the activation polarization term corresponding to the reaction 2H+ + 2e− → H2. whereas concentration polarization and activation polarization usually decay at measurable rates.3. Hydrogen overvoltage as a function of current density.5 HYDROGEN OVERPOTENTIAL The polarization term that controls the corrosion rate of many metals in deaerated water and in nonoxidizing acids is hydrogen overpotential. therefore. Ideally.HYDROGEN OVERPOTENTIAL 63 Figure 5. The iR contribution can be calculated as described in Section 5. . represents the value in ohms of the resistance path of length l cm and specific conductivity κ. decays simultaneously with shutting off the current. H 2 overpotential = φmeasured − (−0. whereas activation polarization and iR drop are not affected significantly. and a hydrogen electrode at equilibrium in the same solution. equal to l/κ. that is. In accord with the previously discussed definition of polarization. iR.059pH) Hydrogen overpotential. is measured in the same way as polarization is measured.5. but reported values often include iR drop and sometimes concentration polarization as well. 5.1. which varies with potential and which. This assumption is consistent. β. with an observed value . appears to be recombination of adsorbed hydrogen atoms. and above the exchange current density. η = β log i i0 is obeyed within the region of applied current density. at i >> ir + icorr. and several other metals.4–0. Greater area and improved catalytic activity of a rough surface account for the observed effect. the overpotential for a noncorroding metal can be represented by a modified Tafel equation. but with a slope of opposite sign.5). Roughening of the surface. This equation describes the small observed slope of dη/d log i for small values of impressed current density. i. The Tafel equation.64 KINETICS: POL ARIZATION AND CORROSION R ATES The absolute values of hydrogen overpotential for a given metal decrease with: 1. Decreasing current density. Although hydrogen overpotential values usually differ in acid compared with alkaline media. 3. Stern [4] showed that for a corroding metal we have η = β log i + ir + icorr i0 where ir is the reverse current for the reaction H2 → 2H+ + 2e−. Values of ir determined from measured values of η also follow the Tafel equation. Similarly. The slow step in the discharge of hydrogen ions on platinum. where α is a constant and the other terms have their usual significance. from kinetic considerations. below the limiting current density for concentration polarization. 2.3RT/αF. Cu. For metals that corrode by hydrogen evolution. is equal to i0. values are not greatly sensitive to pH within either range.6 for Fe. i0. Increasing temperature (i0 increases). η = β log i + ir i0 This equation holds from zero to finite values of i (Fig. The term β is equal to 2. The term α is approximately 2 for platinum and palladium and is approximately 0. Hg. Ni. 5. i. or palladium. at equilibrium. decreasing hydrogen overpotential is one factor accounting for increase of corrosion as the temperature is raised. with the slope increasing as i approaches icorr + ir and reaching the true Tafel slope. as described earlier. A sandblasted surface has a lower hydrogen overpotential than a polished surface. may induce spontaneous cracking. If a catalyst poison. perhaps forming arsenides.. β = 0. deposits out at cathodic areas. having a high hydrogen overpotential. It is also possible [4] that the latter compounds. causing hydrogen embrittlement (loss of ductility). instead.05% As2O3 in 72% H2SO4).1. leads to a low value of hydrogen overpotential. but when present in larger amount (e.* The increased concentration of surface hydrogen favors entrance of hydrogen atoms into the metal lattice.g. either existing in the steel as separate phases or formed as a surface compound by reaction of iron with H2S or phosphorus compounds in solution. the value of α is approximately 0. copper). it retards the rate of formation of molecular H2 and increases the accumulation of adsorbed hydrogen atoms on the electrode surface.5 and. H2SO4). the slow discharge of the hydrated hydrogen ion is apparently the slow step: H + → Hads − e − For many metals at high current densities. Fe → Fe2+ + 2e− (reduce activation polarization). Similarly.g. such as mercury and lead. and. The slow discharge step may.. but presence of sulfur or phosphorus in steels is observed instead to increase the rate. .HYDROGEN OVERPOTENTIAL 65 of α = 2. Tin salts have the same inhibiting effect and have been used to protect steels from attack by pickling acids during descaling operations. stimulate the anodic dissolution reaction. correspondingly. 0. either by combination with itself or with H+. probably because elementary arsenic. be the reduction of H2O. such as mercury or lead. this slow discharge step is also the controlling reaction. This increase probably results from the low hydrogen overpotential of ferrous sulfide or phosphide. The rapidity with which Hads combines to form H2. like hydrogen sulfide or certain arsenic or phosphorus compounds. it is an effective corrosion inhibitor. To account for these values. the proposal has been made that the slow step in the hydrogen evolution reaction on iron is probably H+ + Hads → H 2 − e − The same slow step may apply to various metals having intermediate values of hydrogen overpotential (e. called hydrogen cracking (see Section 8. or alter the anode– cathode area ratio. in addition. arsenious oxide in small amount accelerates corrosion of steel in acids (e. For metals of high hydrogen overpotential. whereas a poor catalyst.4). is added to the electrolyte. nickel. Catalyst poisons increase absorption of hydrogen whether the metal *Increase of hydrogen overpotential normally decreases the corrosion rate of steel in acids. For iron. such as platinum or iron. accounts for a high value of overpotential. is affected by the catalytic properties of the electrode surface. 1 H 2 O → OH − + H 2 − e − 2 The reduction of water as the controlling reaction applies generally to metals in alkaline solutions at high and low current densities. in some stressed high-strength ferrous alloys. A good catalyst.g.. iron. In the absence of hydrogen sulfide. depending on the type of measurements to be made. Values of β. Slight general corrosion of the tubing produces hydrogen. The polarization procedure is then repeated. the first measurement is usually that of the corrosion potential. The current between working and counter electrodes is controlled. can be used. when the applied current. sometimes called Evans diagrams. As can be seen from the table. Various electronic circuits for potentiostat design have been presented in the corrosion literature and their applications to corrosion studies have been discussed [6]. Evans at the University of Cambridge in England.66 KINETICS: POL ARIZATION AND CORROSION R ATES is polarized by externally applied current or by a corrosion reaction with accompanying hydrogen evolution. Iappl. and of η at 1 mA/cm2 for H+ discharge are given in Table 5. Using the . A gas bubbler in used for atmospheric control—for example. 5. consisting of a power supply. are graphs of potential versus log current or log current density. 5. but the effect. ammeter. general corrosion occurs. are more susceptible to hydrogen cracking than are lowstrength steels. i0. a portion of which enters the stressed steel to cause hydrogen cracking. To establish a polarization diagram.6.6 POLARIZATION DIAGRAMS OF CORRODING METALS Polarization diagrams of corroding metals. resistor. They were originally developed by U. and potentiometer. Values also change with concentration of electrolyte. The design of a 1-liter electrochemical cell used in many corrosion laboratories [2] is shown in Fig. For this reason. in addition to the electrode being studied (the “working” electrode). but hydrogen enters the lattice in either case. a galvanostatic circuit. but with Iappl reversed. the usual electrodes that are employed.3(b). is zero. but without hydrogen cracking.1. tending to blister low-strength steels instead. is not large. some oil-well brines containing H2S are difficult to handle in low-alloy steel tubing subject to the usual high stresses of a structure extending several thousand feet underground. who recognized the usefulness of such diagrams for predicting corrosion behavior [5]. φcorr. are the reference electrode and the inert counter (or auxiliary) electrode that is usually made of platinum. High-strength steels. because of their limited ductility. Alternatively. R. values of hydrogen overpotential vary greatly with the metal. The working electrode is then polarized either anodically or cathodically to establish one of the dashed lines in Fig. 5. to obtain the second dashed line. The measurements are usually made using a potentiostat—an instrument that automatically maintains the desired potential between the working and reference electrodes by passing the appropriate current between the working and counter electrodes. comparatively. In experimentally establishing a polarization diagram. and the potential of the working electrode with respect to the reference electrode is measured. to deaerate the solution or to saturate the solution with a specific gas. 5. that is. for a given area of metal surface.POL ARIZATION DIAGR AMS OF CORRODING METALS 67 Figure 5.6. the complete polarization diagram is then constructed. ia. Polarization diagram. In terms of corrosion current. but ia = ic only if Aa = Ac—that is. In an aerated neutral or basic aqueous solution. In general.. M → Mz+ + ze−.e. similarly. and. potentiostatically) or continuously (i. as shown in Fig. In this system. whereas in a deaerated acid.6 for metal M. ic depends on Ac. The corresponding current density.e. Having established φ versus log Iappl on the more noble and the more active sides of the corrosion potential. and the reduction reaction may be symbolized as Rn+ + ne− → R. the reduction reaction could be 2H+ + 2e− → H2. Aa. if 50% of the surface is anodic and 50% is cathodic. the chemical equivalents of metal going into solution at the anodic sites are equal to the chemical equivalents of reduction products produced at cathodic sites. potentiodynamically). Ia = Ic. the general relation for any anode–cathode area ratio Iappl = Ia − Ic can be translated to corresponding current densities iappl = ia − ic. potentiostat. polarization may be carried out either in potential steps (i. that is anode. the . at anodic areas depends on the fraction of metal surface. the oxidation reaction may be the dissolution of metal. the anodic and cathodic currents are equal in magnitude. Under the latter condition. the reduction reaction could be O2 + 2H2O + 4e− → 4OH−.. When the electrode is polarized at sufficiently high current densities to shift the potential more than approximately 100 mV from the corrosion potential. For any corroding metal. Ia/Aa = ia and Ic/Ac = ic. similarly. This situation is described as mixed control. for cathodic polarization. electrodes are in close proximity to each other. When polarization occurs mostly at the anodes. where a lead sulfate film covers the anodic areas and exposes cathodic impurities. φA. If the anodic area of a corroding metal is very small. Under anodic control. Examples are zinc corroding in sulfuric acid and iron exposed to natural waters. the corrosion rate is said to be cathodically controlled. The corrosion current is then controlled by the IR drop through the electrolyte in pores of the coating. The corrosion potential is then near the thermodynamic anode potential. When polarization occurs mostly at the cathode. and. The extent of polarization depends not only on the nature of the metal and electrolyte.7). depending on the direction of Iappl. An example occurs with a porous insulating coating covering a metal surface. for anodic polarization. i0a. 5. the equal rates of oxidation and reduction reactions are expressed as a current density. Similarly. For local-action cells on the surface of a metal. are usually negligible [7].7 INFLUENCE OF POLARIZATION ON CORROSION RATE Both resistance of the electrolyte and polarization of the electrodes limit the magnitude of current produced by a galvanic cell. for the reaction Mz+ + ze− → M is determined. the exchange current density i0c for the cathodic reaction is determined. iappl ≈ ia. the metal surface acts either as all anode or all cathode. iappl ≈ ic. Resistance control occurs when the electrolyte resistance is so high that the resultant current is not sufficient to appreciably polarize anodes or cathodes.4). by extrapolating from the Tafel region to the reversible potential φc. It is common for polarization to occur in some degree at both anodes and cathodes.” Section 5. but also on the actual exposed area of the electrode. for example. By extrapolating from either the anodic or cathodic Tafel region to the corrosion potential φcorr. resistance of the electrolyte is usually a secondary factor compared to the more important factor of polarization. by porous surface films. Other examples are magnesium exposed to natural waters and iron immersed in a chromate solution. Although the latter condition often closely approximates the true situation. the corrosion reaction is said to be anodically controlled (see Fig. the corrosion rate icorr can be determined for the condition that Aa = Ac (anode–cathode area ratio = 1). or “back” reactions. a more exact approximation to the corrosion rate would require information about the actual anode–cathode area ratio. Tafel slopes can then be determined (see paragraph “Activation Polarization. there may be considerable anodic polarization accompanying corrosion. By extrapolating from the anodic Tafel region to the reversible (equilibrium) anode potential. caused. that is. the corrosion potential is close to the thermodynamic potential of the cathode. 5. Accordingly. A practical example is impure lead immersed in sulfuric acid. the exchange current density. consequently. where ic = ia. even .68 KINETICS: POL ARIZATION AND CORROSION R ATES reverse. such as copper. Polarization diagram for corroding metal when anode area equals one-half of cathode area. corresponding polarization curves are shown in Fig. the anode–cathode area ratio is also an important factor in determining the corrosion rate. 5. Figure 5. If current density is plotted instead of total corrosion current.7. Types of corrosion control. for example. .8. Consequently.8. as.INFLUENCE OF POL ARIZATION ON CORROSION R ATE 69 Figure 5. when the anode area is half the cathode area. though measurements show that unit area of the bare anode polarizes only slightly at a given current density. leading *Mercury in aqueous solutions acts first as a mercury electrode. For example. The steady-state corrosion rate of pure cadmium is appreciably higher than that of cadmium amalgam. Any conducting surface on which H+ ions are discharged acts as a polarized hydrogen electrode and can be so considered in evaluating a corrosion cell. it is the high hydrogen overpotential of mercury that limits the corrosion rate of amalgams in nonoxidizing acids. . Consequently. The amalgam polarizes anodically very little.* and zinc atoms act as anodes of corrosion cells. and the corrosion reaction is controlled almost entirely by the rate of hydrogen evolution at cathodic areas. In other words. An important experiment that illustrates the electrochemical mechanism of corrosion was performed by Wagner and Traud [1]. mercury atoms of the amalgam composing the majority of the surface apparently act as cathodes.70 KINETICS: POL ARIZATION AND CORROSION R ATES Figure 5. or hydrogen electrodes. but. They measured the corrosion rate of a dilute zinc amalgam in an acid calcium chloride mixture and the cathodic polarization of mercury in the same electrolyte.9. A piece of platinum coupled to the amalgam considerably increases the rate of corrosion because hydrogen is more readily evolved from a low-overpotential cathode at the operating emf of the zinc–hydrogen electrode cell. The very low corrosion rate and the absence of appreciable anodic polarization explain why amalgams in corresponding metal salt solutions exhibit corrosion potentials closely approaching the reversible potential of the alloyed component. all mercury ions in solution are deposited before H+ is discharged. the corrosion potential of cadmium amalgam in CdSO4 solution is closer to the thermodynamic value for Cd → Cd2+ + 2e− than is observed for pure cadmium in the same solution.9). The current density equivalent to the corrosion rate was found to correspond to the current density necessary to polarize mercury to the corrosion potentiual of the zinc amalgam (Fig. 5. Polarization diagram for zinc amalgam in deaerated HCl. when cathodically polarized. using (5.0 11. the Tafel equation expresses cathodic polarization behavior.) −0.2.203 −0. Then.7) from which icorr and the equivalent corrosion rate can be calculated.8 CALCULATION OF CORROSION RATES FROM POLARIZATION DATA The corrosion current can be calculated from the corrosion potential and the thermodynamic potential if the equation expressing polarization of the anode or cathode is known.24 4% NaCl + HCl pH = 1 pH = 2 0.015 icorr (μA/cm2) Calculated 10. iron.11 10. the surface of the metal is probably covered largely with adsorbed H atoms and can be assumed.059 pH.06 0.g. For corrosion of active metals in deaerated acids. Typical values are given in Table 5.2 Observed 11.084 0. The thermodynamic potential is −0. and if the anode–cathode area ratio can be estimated. In general. therefore. This situation also holds for the steady-state potentials of more noble metals (e.2 0.1 M Malic acid pH = 2. mercury) which undergo corrosion in the presence of dissolved oxygen.4 1.5 1.5 11. zinc. 5.C ALCUL ATION OF CORROSION R ATES FROM POL ARIZATION DATA 71 to greater deviation of the measured corrosion potential from the corresponding thermodynamic value. nickel.10 0. copper.083 i0 (μA/cm2) 0.1 11. φcorr = − 0..2.g. S..201 β (V) 0.093 0.158 −0. for example.059pH (volts.1 M Citric acid pH = 2. Comparison of Calculated and Observed Corrosion Currents for Pure Iron in Various Deaerated Acids [4] Solution φHcorr + 0. the steady-state potential of any metal more active than hydrogen (e.3 . Stern [4] showed that calculated corrosion rates for iron. and if icorr 1 is sufficiently larger than i0 for H+ → 2H2 − e−. The measured value tends to be more noble than the true value.172 −0.7) and employing empirical values for β and i0.E.059 pH + β log ( icorr i0 ) (5.100 0. were in excellent agreement with observed rates. TABLE 5.H. cadmium) in an aqueous solution of its ions deviates from the true thermodynamic value by an amount that depends on the prevailing corrosion rate accompanied by H+ discharge. to be mostly cathode.100 0. Under other assumptions.12 V. and βa is between 0. Also. Equation (5.72 KINETICS: POL ARIZATION AND CORROSION R ATES Subsequently.9) is shown by data summarized in Fig. for which Δφ is not more than about 10 mV. The observed corrosion current.10. straight lines are shown representing values calculated on the basis of several assumed β values within which most of the empirical data lie. (5.8) and (5.3 Δφ (5. It can also be used to . Section 29.8). now known as the Stern–Geary equation. respectively. with Iappl is essentially linear (for derivation. corresponding to data on corrosion of nickel in HCl and on corrosion of steels and cast iron in acids and in natural waters. Stern showed. The linear polarization method has obvious advantages in calculating instantaneous corrosion rates for many metals in a wide variety of environments and under various conditions of velocity and temperature.3R ⎝ βc + βa ⎠ Iappl ⎛ βcβa ⎞ 2.9) have been widely used with considerable success for determining the corrosion rates of various metals in aqueous environments at ambient and elevated temperatures. Stern and Geary [8] derived the very important and useful equation. extends over six orders of magnitude.06 V.9) Although values of β are relatively well known for H+ discharge. which need not be known.2). that the majority of reported β values are between 0.12 V. Some exchange current densities for Fe3+ → Fe2+ − e− on passive surfaces are included because the same principle applies in calculating i0 for a noncorroding electrode as in calculating Icorr for a corroding electrode. and (5. If βc is known to be 0.8) simplifies to I corr = βa Iappl 2. remains essentially constant and conditions otherwise at the surface of the corroding metal are largely undisturbed.06 and 0. the anode–cathode area ratio. The general validity of Eqs. (5. see the Appendix. the calculated corrosion current is within at least 20% of the correct value. Equations (5. 5. they are not as generally available for other electrode reactions. Δφ.3Δφ ⎝ βc + βa ⎠ (5. I corr = or I corr = 1 ⎛ β cβ a ⎞ 2. and Iappl/Δφ is the polarization slope (the reciprocal of the polarization resistance. Under conditions of slight polarization. If corrosion is controlled by concentration polarization at the cathode. for which the change of potential. R = Δφ/Iappl) in the region near the corrosion potential. as when oxygen depolarization is controlling. the corrosion rate can be calculated to at least a factor of 2.8) where βc and βa refer to Tafel constants for the cathodic and anodic reactions.06 and 0. for example.7). however. Mansfeld and Oldham [10] have developed a set of equations that provides less error than Eq. Section 29. A correction is required if an IR drop is involved in the measurement. Relation between polarization slope at low applied current densities and observed corrosion or exchange current densities [9]. Reprinted with permission. The Stern–Geary equation has been modified to minimize errors under particular conditions.8) when this situation applies (see Appendix. 5. (5. (Copyright ASTM INTERNATIONAL. (5.) evaluate inhibitors and protective coatings. Errors in Eq.8) are especially significant in systems in which the corrosion potential is close to one of the reversible potentials—that is.2).10. is outside the Tafel region.ANODE– C ATHODE AREA R ATIO 73 Figure 5.9 ANODE–CATHODE AREA RATIO The reduction of H+ at a platinum cathode at a given rate is accompanied by the simultaneous reverse reaction at a lower rate for H2 oxidizing to H+. as well as for detecting changes of corrosion rate with time. The . Aa is the fraction of surface that is anode such that Aa + Ac = 1.13) into (5.12) Since Aa = 1 − Ac. and i0a and i0c are the exchange current densities for the anode and cathode reactions. respectively.2. Then ′ ′ βc log I corr = φc − φcorr + βc log Ac i0 c and φcorr = φa + βa log I corr − βa log Aa i0 a Substituting (5.10) where subscripts a and c refer to anode and cathode. Section 29. For a corroding metal.13) (5. φ′ is the polarized potential. and since the observed polarization of anodic or cathodic sites depends. in part. Hence. Stern [11] developed the general relation for which anodic and cathodic polarization follows Tafel behavior and IR drop is negligible. we obtain φ′ = φc − βc log c and φ′ = φa + βa log a Ia Aai0 a (5. φ is the thermodynamic potential. the anode and cathode reactions are distinct and different. I is the current per unit metal surface area. Using a typical polarization diagram. on the area over which oxidation or reduction occurs. an anode–cathode area ratio exists. the distance between anode and cathode may vary from atomic dimensions to a separation of many meters. At steady state. we get . as illustrated in the Appendix. we obtain log I corr = φc − φa βa βc + log Aa i0 a + log Ac i0 c βc + βa βc + βa βc + βa (5. φc = φa = φcorr and I a = I c = I corr . At equilibrium.1. Correspondingly.12). respectively.14) (5. 29. the anode–cathode area ratio is an important factor in the observed corrosion rate.74 KINETICS: POL ARIZATION AND CORROSION R ATES oxidation and reduction reactions are assumed to take place on the same surface sites. one is not merely the reverse reaction of the other. the oxidation reaction can take place only on sites of the metal surface that are different from those on which reduction takes place. the forward and reverse reactions are equal and the equivalent current density is called the exchange current density. For this situation. β is the Tafel slope.11) Ic Ac i0 c (5. Fig. (a) Electrical equivalent circuit model used to represent an electrochemical interface undergoing corrosion in the absence of diffusion control.15) (5. impedance determines the amplitude of current for a given voltage. then the maximum corrosion rate occurs at Ac = 2—that is. W is the Warburg impedance [13]. If.3(βc + βa) Ac Aa (5.16) equals zero. the response of an electrode to alternating potential signals of varying frequency is interpreted on the basis of circuit models of the electrode/electrolyte interface. the corrosion rate is less.3(βc + βa ) (1 − Ac) 2.ELEC TROCHEMIC AL IMPEDANCE SPEC TROSCOPY 75 d log I corr βa 1 βc 1 =− + dAc 2.16) Maximum Icorr occurs at d log Icorr/dAc = 0—that is. In electrochemical impedance spectroscopy (EIS). βc = 1 βa. 5. such as inhibition and coatings. and Rs is the solution resistance [15]. when the numerator of (5. In an alternating-current circuit. Rp is the polarization resistance.11. Impedance is the proportionality factor between voltage and current. at an anode– cathode area ratio of unity. This occurs when Ac = βc/(βc + βa). . Rp is the polarization resistance.10 ELECTROCHEMICAL IMPEDANCE SPECTROSCOPY Electrochemical methods based on alternating currents can be used to obtain insights into corrosion mechanisms and to establish the effectiveness of corrosion control methods. At any other anode–cathode area ratio.11 shows two circuit models that can Cdl Rs Rp (a) Cdl Rs W Rp (b) Figure 5. Figure 5. Cdl is the double layer capacitance. (b) Electrical equivalent circuit model when diffusion control applies. as is frequently observed. reaching zero at a ratio equal to either zero or infinity.3(βc + βa) Ac = βc(1 − Ac) − βa Ac 2. that depends on the circuit parameters. and t is the time (seconds) [12].. the resulting sinusoidal current waveform is out of phase with the applied potential signal by an amount.11(b). and the phase angle. Cdl (the capacitance of the double layer). is the frequency-dependent proportionality factor in the relationship between the voltage signal and the current response. alternatively. Z(ω). At each frequency. Z is the impedance (ohm-cm2).22) E ′ + jE ′′ I ′ + jI ′′ Z ′′ Z′ Z = (Z ′)2 + (Z ′′)2 In electrochemical impedance analysis. The simplest model for characterizing the metal–solution interface. sometimes called the Warburg impedance. When diffusion control is important. the impedance of the corroding metal is analyzed as a function of frequency. 5. by the real component. E = E0 sin(ωt). The mathematical convention for separating the real and imaginary components is to multiply the magnitude of the imaginary component by j [ = −1] and report the real and imaginary values as a complex number. A sinusoidal potential change is applied to the corroding electrode at a number of frequencies. is added in series with Rp.e.20) (5.21) (5. i = i0sin (ωt + θ). or. showing −Z″ versus Z′) and two dif- . with the result that.76 KINETICS: POL ARIZATION AND CORROSION R ATES be used for analyzing EIS spectra. In electrochemical impedance spectroscopy. frequency is zero). the sum of Rs and Rp is measured.18) (5. the phase angle. ω. Impedance is a complex number that is described by the frequencydependent modulus. Fig. the circuit with the smallest impedance dominates. In parallel electrical circuits. and Rp (the polarization resistance). The equations for electrochemical impedance are [13] E = Ereal + Eimaginary = E ′ + jE ′′ I = I real + I imaginary = I ′ + jI ′′ Z = Z ′ + jZ ′′ = tanθ = 2 (5. the impedance of the capacitor approaches infinity. When direct-current measurements are carried out (i.17) where E is the voltage signal. Z (ω) = E(ω) i(ω) (5. under these conditions. Z′. If Rs if significant. ZD. θ. 5. θ. |Z|. another element. includes the three essential parameters.11(a). i is the current density.19) (5. Z″. The electrochemical impedance. Rs (the solution resistance). as shown in Fig. and the imaginary component. The current amplitude is inversely proportional to the impedance of the interface. three different types of plots are commonly used: Nyquist plots (complex plane. the corrosion rate is underestimated. 5.14. 5.13. beyond the corrosion potential to the thermodynamic potential of the anode. 5.) ferent types of Bode plots.11(a) [15].THEORY OF C ATHODIC PROTEC TION 77 Figure 5. if polarization of the cathode is continued. indicating that the only process occurring is charge transfer. is one of the most effective engineering means for reducing the corrosion rate to zero. One example of each type of plot is shown in Figs. This is the basis for cathodic protection of metals.12 illustrates electrochemical impedance data for cast 99.2 sodium borate in the Nyquist format. discussed further in Chapter 13.12. as shown in Fig. both electrodes attain the same potential and no corrosion of zinc can occur.13. Cathodic protection is accomplished by supplying an external current to the corroding metal that is to be protected. 5. using external current. at very low frequencies. From Fig.13 shows Bode magnitude and phase angle plots for a system with the equivalent circuit shown in Fig. Current leaves the auxiliary anode (composed of any metallic or nonmetallic . 5. Z = Rs (5. (Reproduced by permission of ECS.12 and 5.2 sodium borate presented in the Nyquist format [14].23) 5. Z = Rs + Rp At very high frequencies. Figure 5.9% magnesium in pH 9. it can be seen that. Cathodic protection. showing the impedance magnitude versus log frequency and showing phase angle versus log frequency.9% magnesium in pH 9. it is clear that. This figure shows a single capacitive semicircle.11 THEORY OF CATHODIC PROTECTION Observing the polarization diagram for the copper–zinc cell in Fig.24) (5.2. The Electrochemical Society. Figure 5. Electrochemical impedance data for cast 99. passes through the electrolyte.) Figure 5. The metal cannot corrode as long as the current is applied to maintain this potential.org.13. B. and the equilibrium potential. The corroding metal is fully protected—and the corrosion rate is zero—when it is polarized to the open-circuit anode potential.14. the thermodynamic potential. www.11(a) [15]. . Bode magnitude and phase angle plots showing the frequency dependence of electrochemical impedance for the equivalent circuit model in Fig. 5. (Reprinted with permission of ASM International®. conductor).15. enters the metal that is to be protected. 5. where Iappl is the applied current necessary for complete protection. and returns to the source of direct current.asminternational. also known as the reversible potential. Cathodic protection by impressed current. All rights reserved. The polarization diagram is shown in Fig.78 KINETICS: POL ARIZATION AND CORROSION R ATES Figure 5. ASTM. and the applied current for complete cathodic protection equals Iappl. and may damage amphoteric metals and coatings. Inman. of the anode. however. A. 2nd edition. Evans. φA. As the applied current e–b is increased. J. Z. J. O’M. West Conshohocken. N. and the corrosion current b becomes smaller. Gamboa-Aldeco. 2. Reddy. Elektrochem. Greene.15. Stern. REFERENCES 1. Should the applied current fall below that required for complete protection. Bockris. C. D. 4. Modern Electrochemistry. Polarization diagram illustrating principle of cathodic protection. K. Electrochem. R. If the metal is polarized slightly beyond the open-circuit potential. Rensselaer Polytechnic Institute. Mansfeld and R. For example. if the corrosion potential is moved from φcorr to a in Fig. J. 102. Vol. therefore. 208.REFERENCES 79 Figure 5. by applied current e–b. 3. Franklin Inst. When a coincides with φA. 44. Troy. Kluwer Academic/Plenum Publishers. M. Fundamentals of Electrodics. Soc. . 5. 21 (1979). Net current flows from the electrolyte to the metal. the potential a moves to more active values. the corrosion rate remains zero. Current in excess of the required does no good. the impressed current is kept close to the theoretical minimum. and M. the corrosion current becomes zero. N. PA.. 1246–1247. In practice. some degree of protection nevertheless occurs. Traud. 6. the corrosion current decreases from Icorr to b. ASTM Standard G5-94 (Reapproved 1999). 2A. NY. metal ions cannot enter the solution. ibid. 391 (1938). 5. 45 (1929). Wagner and W. Experimental Electrode Kinetics. F. pp. 2000.15. New York. Corrosion 35 (1). U. 663 (1955). hence. V. 1965. 609. Vol. Buchheit. pp. Kelly. Theory. New York. Soc.80 KINETICS: POL ARIZATION AND CORROSION R ATES 7. Marcel Dekker. edited by M. 414 (1990). R. Modern Electrochemistry. 2nd edition. Barsoukov and J. Electrochemical Materials Science. 2003. Shoesmith. J. 2003. Scully. Electrochem. Plenum Press. 56 (1957). Wiley. Standard Practice for Verification of Algorithm and Equipment for Electrochemical Impedance Measurements. PROBLEMS 1. New York. Kluwer Academic/Plenum Publishers. H. pp. C. Silverman. Wiley. editor. editors. The potential of an iron electrode when polarized as cathode at 0. Oldham. ASM International. Sci. Testing Mater. p. M. in Comprehensive Treatise of Electrochemistry. A Century of Tafel’s Equation: A Commemorative Issue of Corrosion Science. 97. in ASM Handbook. M. 59. and Applications. R. Smyrl. Scully and R. Wiley. Electrochem. 1179. O’M. D. in Electrochemical Techniques in Corrosion Science and Engineering. Plenum Press. 329t (1958). W. 125. K. 2000. 1280 (1959). Am. New York. Staehle. 9. New York.001 A/cm2 is −0. 15. Geary. The pH of the electrolyte is 4. M. Makar and J. 787 (1971). Kelly. December 2005.. 2008. 2005. 2nd edition.0. 43–45. Proc. Corrosion: Fundamentals. GENERAL REFERENCES E. 2003. PA. Macdonald. G. New York. D. Practical corrosion prediction using electrochemical techniques. J. N. L. Mansfeld and K. Corrosion Science 47(12). Experiment. New York. Electrochemical Impedance Spectroscopy. 1981. R. Revie. Methods for DeterminIng Aqueous Corrosion Reaction Rates. Stern. Orazem and B. Weisert. R. Corros. O’M. Bockris et al. 4. see also Ref. 13A. West Conshohocken. in Electrochemical Techniques in Corrosion Science and Engineering. 11. 3. 13. J. 14. 104.916 V versus 1 N calomel half-cell. Corrosion 14. R. editors. M. OH. Soc. 2nd edition. W. Mansfeld. Vol. 1054–1055. 6. Fontana and R. 10. F. Marcel Dekker. Tribollet. Impedance Spectroscopy. W. G. in Uhlig’s Corrosion Handbook. Stern and A. R. New York. J. editor. What is the value of the hydrogen overpotential? . J. p. p. G. 137. and R. Soc. Stern and E. Vol. R. 11. Corrosion studies of rapidly solidified magnesium alloys. 73. J. G. Kruger. Bockris and A. p. 163. Scully. 1976. G. in Advances in Corrosion Science and Technology. Reddy. Testing. ASTM. Kelly. 2000. New York. and Protection. ASTM G 106-89 (Reapproved 1999). p. 12. J. Materials Park. 8. F. PROBLEMS 81 2. The potential of a cathode at which H+ discharges at 0.001 A/cm2 is −0.92 V versus Ag–AgCl in 0.01 N KCl at 25°C. (a) What is the cathode potential on the standard hydrogen scale? (b) If the pH of the electrolyte is 1, what is the value of hydrogen overpotential? 3. The potential of a platinum anode at which oxygen evolves in an electrolyte of pH 10 is 1.30 V with respect to the saturated calomel electrode. What is the value of oxygen overpotential? 4. The potential of a copper electrode on which Cu2+ deposits from 0.2 M CuSO4 versus 1 N calomel electrode is −0.180 V. What is the polarization of the electrode in volts? Is the electrode polarized in the active or noble direction? 5. Closely separated short-circuited zinc and mercury electrodes are immersed in deaerated HCl of pH = 3.5. What is the current through the cell if the total exposed area of each electrode is 10 cm2? What is the corresponding corrosion rate of zinc in gmd? (Corrosion potential of zinc versus 1 N calomel half-cell is −1.03 V.) 6. The corrosion potential of mild steel in a deaerated solution of pH = 2 is −0.64 V versus saturated Cu–CuSO4 half-cell. The hydrogen overpotential (volts) for the same steel follows the relation 0.7 + 0.1 log i, where i = A/cm2. Assuming that approximately all the steel surface acts as cathode, calculate the corrosion rate in mm/y. 7. Derive an expression for the slope of corrosion rate versus pH for a dilute cadmium amalgam in a deaerated cadmium ion solution. Neglect concentration polarization and assume that the amalgam is approximately all cathode. 8. The potential of platinum versus saturated calomel electrode when polarized cathodically in deaerated H2SO4, pH = 1.0 at 0.01 A/cm2 is −0.334 V and at 0.1 A/cm2 it is −0.364 V. Calculate β and i0 for discharge of H+ on platinum in this solution. 9. The corrosion rate of iron in deaerated HCl of pH = 3 is 3.0 gmd. Calculate the corrosion potential of iron in this acid versus 0.1 N calomel electrode. Assume that the entire iron surface acts as a cathode. 10. The linear polarization slope dφ/dI at low current densities for iron in a corrosive solution equals 2 mV/μA-cm2. Using Eq. (5.8), calculate the corrosion rate in gmd. Assume βa = βc = 0.1 V. 11. Anodic activation overpotential η for small applied current density i follows the relation η = ki. Derive the value of k in terms of the exchange current density i0, assuming βa = βc = 0.1 V. 82 KINETICS: POL ARIZATION AND CORROSION R ATES 12. Determine the corrosion potential and corrosion rate of an iron pipe carrying 1 N sulfuric acid at 0.2 m/s at 25°C. Assume that the entire iron surface acts as cathode, that Tafel slopes are ±0.100 V, and that the exchange current densities for Fe/Fe2+ and for hydrogen on iron are 10−3 and 10−2 A/m2, respectively. 13. Determine the corrosion potential and corrosion rate of zinc in 1 N hydrochloric acid. Assume that the entire zinc surface acts as cathode, that Tafel slopes are ±0.100 V, and that the exchange current densities for zinc and for hydrogen on zinc are 0.1 and 10−4 A/m2, respectively. 14. Show the complex-plane impedance spectra for the following electrochemical cells: (a) The cell is represented by a single electrical resistance of value, R (Ω). (b) The cell is represented by a capacitance of value, C (F). (c) The cell is represented by an equivalent circuit with a resistance, Rs (Ω), in series with a capacitance, Cs (F). (d) The cell is represented by an equivalent circuit with a resistance, Rp (Ω), in parallel with a capacitance, Cp (F). (e) The cell is represented by the circuit shown in Fig. 5.11(a). Answers to Problems 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. −0.40 V. (a) −0.58 V; (b) −0.52 V. 0.90 V. −0.188 V; active direction. 2.3 × 10−7 A; 0.0067 gmd. 0.134 mm/y. dicorr/dpH = −2.3(0.059icorr)/β. β = 0.03 V; i0 = 8 × 10−4 A/cm2. −0.719 V. 2.73 gmd. k = 0.0217/i0. −0.17 V (S.H.E.); 0.58 mm/y. −0.53 V (S.H.E.); 29.9 mm/y. 6 PASSIVITY 6.1 DEFINITION A passive metal is one that is active in the Emf Series, but that corrodes nevertheless at a very low rate. Passivity is the property underlying the useful natural corrosion resistance of many structural metals, including aluminum, nickel, and the stainless steels. Some metals and alloys can be made passive by exposure to passivating environments (e.g., iron in chromate or nitrite solutions) or by anodic polarization at sufficiently high current densities (e.g., iron in H2SO4). As early as the eighteenth century, it was observed that iron reacts rapidly in dilute HNO3, but is visibly unattacked by concentrated HNO3 [1]. Upon removing iron from the concentrated acid and immersing it into the dilute acid, a temporary state of corrosion resistance persists. In 1836, Schönbein [2] defined iron in the corrosion-resistant state as “passive.” He also showed that iron could be made passive by anodic polarization. Around the same time, Faraday [3] performed several experiments showing, among other things, that a cell made up of passive iron coupled to platinum in concentrated nitric acid produced little or no current, in contrast to the high current produced by amalgamated zinc coupled to platinum in dilute sulfuric acid. Although passive iron corrodes only slightly in concentrated HNO3, and similarly amalgamated zinc corrodes only slightly in Corrosion and Corrosion Control, by R. Winston Revie and Herbert H. Uhlig Copyright © 2008 John Wiley & Sons, Inc. 83 84 PASSIVIT Y dilute H2SO4, Faraday emphasized that a low corrosion rate is not alone a measure of passivity. Instead, he stated, the magnitude of current produced in the cell versus platinum is a better criterion, for on this basis iron is passive, but not zinc. This, in essence, defined a passive metal, as we do today, as one that is appreciably polarized by a small anodic current. Later investigators deviated from this definition, however, and also called metals passive if they corrode only slightly despite their pronounced tendency to react in a given environment. This usage brought about two definitions of passivity: Definition 1. A metal is passive if it substantially resists corrosion in a given environment resulting from marked anodic polarization. Definition 2. A metal is passive if it substantially resists corrosion in a given environment despite a marked thermodynamic tendency to react. C. Wagner [4] offered an extension of Definition 1, the essence of which is the following: A metal is passive if, on increasing the electrode potential toward more noble values, the rate of anodic dissolution in a given environment under steadystate conditions becomes less than the rate at some less noble potential. Alternatively, a metal is passive if, on increasing the concentration of an oxidizing agent in an adjacent solution or gas phase, the rate of oxidation, in absence of external current, is less than the rate at some lower concentration of the oxidizing agent. These alternative definitions are equivalent under conditions where the electrochemical theory of corrosion applies. Thus, lead immersed in sulfuric acid, or magnesium in water, or iron in inhibited pickling acid, would be called passive by Def. 2 based on low corrosion rates, despite pronounced corrosion tendencies; but these metals are not passive by Def. 1. Their corrosion potentials are relatively active, and polarization is not pronounced when they are made the anode of a cell. Examples of metals that are passive under Definition 1, on the other hand, include chromium, nickel, molybdenum, titanium, zirconium, the stainless steels, 70%Ni–30%Cu alloys (Monel), and several other metals and alloys. Also included are metals that become passive in passivator solutions, such as iron in dissolved chromates. Metals and alloys in this category show a marked tendency to polarize anodically. Pronounced anodic polarization reduces observed reaction rates, so that metals passive under Definition 1 usually conform as well to Definition 2 based on low corrosion rates. The corrosion potentials of metals passive by Definition 1 approach the open-circuit cathode potentials (e.g., the oxygen electrode); hence, as components of galvanic cells, they exhibit potentials near those of the noble metals. 6.2 CHARACTERISTICS OF PASSIVATION AND THE FLADE POTENTIAL Suppose iron is made anode in 1N H2SO4 and is arranged so that, as the potential is gradually increased, the corresponding polarizing current reaches the value CHAR ACTERISTICS OF PASSIVATION AND THE FL ADE POTENTIAL 85 required to maintain the prevailing potential with respect to a reference electrode. In the laboratory, this is usually accomplished using a potentiostat, which controls the potential between the “working” electrode and the reference electrode by automatically adjusting the current between the “working” electrode and an inert electrode, usually platinum, that is called the “auxiliary” or “counter” electrode. The resulting polarization curve, shown in Fig. 6.1 [5], is called a potentiostatic polarization curve. Galvanostatic polarization is an alternative to potentiostatic polarization. In galvanostatic polarization measurements, the current between working and counter electrodes is controlled, and the potential between working and reference electrodes is automatically adjusted to the value required to maintain the current. Galvanostatic measurements can be made using, for example, the circuit shown in Fig. 5.3(a) in Section 5.3. The galvanostatic polarization curve for iron in 1N H2SO4 is shown in Fig. 6.2. The potentiostatic polarization curve provides considerably more information about passivity than does the galvanostatic curve. Referring to Fig. 6.1, iron is active at small current densities and corrodes anodically as Fe2+ in accord with Faraday’s law. As current increases, a partially insulating film forms over the electrode surface, composed probably of FeSO4. At a critical current density, icritical, of about 0.2 A/cm2 (higher on stirring or on lowering pH of the environment), the current suddenly drops to a value orders of magnitude lower, called the passive current density, ipassive. At this point, the thick insulating film dissolves, being replaced by a much thinner film, and iron becomes passive. The value of ipassive decreases with time, diminishing to about Figure 6.1. Potentiostatic anodic polarization curve for iron in 1N H2SO4. 86 PASSIVIT Y Figure 6.2. Galvanostatic anodic polarization curve for iron in 1N H2SO4. 7 μA/cm2 in 1N H2SO4. In other electrolytes, ipassive may be higher or lower than in 1N H2SO4. The true critical current density for achieving passivity of iron in absence of an insulating reaction-product layer is estimated to be about 10– 20 A/cm2, as shown by short-time current–pulse measurements. On further gradual change of the potential, the current density remains at the above low value, and the corrosion product is now Fe3+. At about 1.2 V, the equilibrium oxygen electrode potential is reached, but oxygen is not evolved appreciably until the potential exceeds the equilibrium value by several tenths of a volt (oxygen overpotential). The increased current densities in the region labeled “transpassive” represent O2 evolution plus Fe3+ formation. When the applied potential is removed, passivity decays within a short time in the manner shown in Fig. 6.3. The potential first changes quickly to a value still noble on the hydrogen scale, and then it changes slowly for a matter of seconds to several minutes. Finally, it decays rapidly to the normal active potential of iron. The noble potential arrived at just before rapid decay to the active value was found by Flade [6] to be more noble the more acid the solution in which passivity decayed. This characteristic potential, φF, was later called the Flade potential, and Franck established it to be a linear function of pH [7]. His measurements in acid media, combined with later data by others, provide the relation at 25°C: φ F ( volts, S.H.E.) = +0.63 − 0.059 pH (6.1) This reproducible Flade potential and its 0.059 pH dependence is characteristic of the passive film on iron. A similar potential–pH relation is found for the passive film on chromium, for Cr–Fe alloys,* and for nickel, for which the stan*When activated cathodically, the Flade potential of chromium and stainless steels follows the relation n(0.059 pH), where n may be as high as 2. For self-activation, n is 1 [8]. CHAR ACTERISTICS OF PASSIVATION AND THE FL ADE POTENTIAL 87 Figure 6.3. Decay of passivity of iron in 1N H2SO4 showing Flade potential, ϕF. dard Flade potentials (pH = 0) are less noble than for iron, in accord with more stable passivity. Stability of passivity is related to the Flade potential, assuming the following schematic reaction to take place during anodic passivation: M + H 2 O → O ⋅ M + 2 H + + 2e − (6.2) where −φF is the potential for the reaction, and O·M refers to oxygen in the passive film on metal M whatever the passive film composition and structure may be. The amount of oxygen assumed to be combined with M has no effect on present considerations. It follows, as observed, that φF = φF + 0.059 log (H + )2 = φ F − 0.059 pH 2 (6.3) The positive value of φ F for iron (0.63 V) indicates considerable tendency for the passive film to decay [reverse reaction of (6.2)], whereas an observed negative value of φ F = −0.2 V for chromium indicates conditions more favorable to passive-film formation and, hence, greater stability of passivity. The value of φ F for nickel is 0.2 V. For chromium–iron alloys, values range from 0.63 V for pure iron to increasingly negative values as chromium is alloyed, changing most rapidly in the range 10–15% chromium and reaching about −0.1 V at 25% chromium. The corresponding φ F values are shown in Fig. 6.4. 88 PASSIVIT Y Figure 6.4. Standard Flade potentials for chromium–iron alloys and chromium [7–9]. The potential (P in Fig. 6.1) at which passivity of iron initiates (passivating potential) approximates, but is not the same as, the Flade potential because of IR drop through the insulating layer first formed and because the pH of the electrolyte at the base of pores in this layer differs from that in the bulk of solution (concentration polarization). These effects are absent on decay of passivity. 6.3 BEHAVIOR OF PASSIVATORS* The same Flade potential is obtained whether iron is passivated by concentrated nitric acid or is anodically polarized in sulfuric acid, indicating that the passive film is essentially the same in both instances. In fact, when iron is passivated by immersion in solutions of passivators, whether chromates (CrO2− ), nitrites 4 (NO− ), molybdates (MoO2− ), tungstates (WO2− ), ferrates ( FeO2− ), or pertechne2 4 4 4 tates (TcO− ), the corresponding Flade potentials are close to values obtained 4 *See also Chapter 17. BEHAVIOR OF PASSIVATORS 89 otherwise [10, 11]. Hence, it can be concluded that the passive film on iron is essentially the same regardless of the passivation process. The amount of passivefilm substance, as determined by coulometric and other measurements, is approximately 0.01 C/cm2 in all cases. Passivation proceeds by an electrochemical mechanism, with some exceptions discussed later. The standard Flade potential of iron passivated by chromates (φ F = 0.54 V) is less noble than for iron passivated by HNO3 (φ F = 0.63 V). One explanation [11] is that chromates adsorb on the passive film more strongly than do nitrates, thereby reducing the overall free energy of the system and increasing the stability of the passive film. Other passivators presumably adsorb similarly, but with differing energies of adsorption. Passivators are reduced at cathodic areas at a current density equivalent to a true current density at anodic areas equaling or exceeding icritical (10–20 A/cm2) for passivation of iron. The passivator is reduced over a large cathodic area of the metal surface to an extent not less than that necessary to form a chemically equivalent passive film at small residual anodic areas. The small passive areas, in turn, adsorb passivator, thereby becoming noble to adjoining passive or nonpassive areas, causing passivity to spread. When the passive film is complete, it acts as cathode over its entirety, and further reduction of passivator proceeds at a much slower rate, equivalent to the rate of continuous passive film breakdown, or to ipassive. Since breakdown is accelerated by presence of chlorides and by elevated temperatures, consumption of passivator is also increased correspondingly. Passivators as a group are inorganic oxidizing agents that have the unique property of reacting only slowly when in direct contact with iron, but they are reduced more rapidly by cathodic currents. For this reason, they can adsorb first on the metal surface, with each site of adsorption adding to the cathodic area. The higher the concentration of passivator, the more readily it adsorbs, and the smaller the residual anodic areas become; this situation obviously favors increased anodic polarization and ultimately passivation. It requires about 0.5–2 h for the passive film to form completely when iron is immersed in 0.1% K2CrO4, with the shorter time being characteristic of aerated solution while the longer time is typical of deaerated solution [12, 13]. 6.3.1 Passivation of Iron by HNO3 In nitric acid, the cathodic depolarizer (passivator) is nitrous acid, HNO2. This must form first in sufficient quantity by an initial rapid reaction of iron with HNO3. As nitrous acid accumulates, anodic current densities increase, eventually reaching icritical. Passivity is then achieved, and the corrosion rate falls to the comparatively low value of about 2 gmd [14]. If urea is added to concentrated HNO3, passivation is interrupted because urea reacts with nitrous acid in accord with (NH 2)2CO + 2 HNO2 → 2 N 2 + CO2 + 3H 2O (6.4) 90 PASSIVIT Y thereby decreasing the nitrous acid concentration. The rate of reaction is, nevertheless, sufficiently below that of HNO2 production so that passivity can still occur, although periodic breakdown and formation of the passive film usually result. Hydrogen peroxide added to concentrated nitric acid also causes periodic breakdown and formation of passivity, probably by oxidizing HNO2 to HNO3 [14]. The peroxide, of itself, is not as efficient a cathodic depolarizer as HNO2, and hence the passive film in its presence can repair itself only when the momentary surface concentration of HNO2 formed by reaction of iron with HNO3 is sufficiently high. After passivity is achieved, the surface concentration of HNO2 is diminished by reaction with peroxide below that needed to maintain passivity, whereupon the cycle is repeated. If iron is first immersed in dilute chromate solution for several minutes, it remains passive in concentrated nitric acid without the initial reaction to form HNO2. The passive film is preformed by chromate, and nitrous acid is no longer necessary as depolarizer to reach icritical, but is needed only in concentration sufficient to maintain the film already present. 6.4 ANODIC PROTECTION AND TRANSPASSIVITY The electrochemical nature of the passivation process is the basis on which anodic polarization leads to passivation. The anodic polarization that is required for passivation can be achieved by applying current or by increasing either the cathodic area or the cathodic reaction rate. High-carbon steels, for example— which contain areas of cementite (Fe3C) that act as cathodes—are more readily passivated by concentrated nitric acid than is pure iron. For this reason, nitrating mixtures of sulfuric and nitric acids are best stored or shipped in drums of steel of the highest carbon content consistent with required mechanical properties. Similarly, stainless steels that may lose passivity in dilute sulfuric acid retain their corrosion resistance if alloyed with small amounts of more noble constituents of low hydrogen overpotential, or of low overpotential for cathodic reduction of dissolved oxygen, such as Pd, Pt, or Cu [15]. The polarization diagram corresponding to increased passivity by low overpotential cathodes is shown in Fig. 6.5, and the corresponding improved corrosion resistance to sulfuric acid is shown in Fig. 6.6. Because titanium, unlike 18–8 stainless steel, has a low critical current density for passivity in chlorides as well as in sulfates, passivity in boiling 10% HCl is made possible by alloying titanium with 0.1% Pd or Pt [16]. Pure titanium, on the other hand, corrodes in the same acid at very high rates (see Fig. 25.2, Section 25.3). Based on the same principle, the corrosion resistance of metals and alloys with polarization curves having active–passive transitions can be greatly improved by an impressed anodic current initially equal to or greater than the critical current for passivity. The potential of the metal moves into the passive region 1). .9) because the current flow is in the direction opposite to that which is used in cathodic protection. Whereas cathodic protection can.5.ANODIC PROTEC TION AND TR ANSPASSIVIT Y 91 Figure 6. 6. so that the current density and accompanying corrosion rate correspond to the low value of ipassive. Figure 6. Corrosion rates in sulfuric acid of 18–8 stainless steel alloyed with copper or palladium. 360-h test.6. be applied to both passive and nonpassive metals. (Fig. Section 6. anodic protection is applicable only to metals that can be passivated when anodically polarized (see Definition 1. depending on overvoltage of cathodic areas (differing cathodic reaction rates). Polarization diagram for metal that is either active or passive. in principle. 20°C [15]. This process is called anodic protection (see Section 13.1). Expressed another way. and move the corrosion potential into the transpassive region. This theory is sometimes referred to as the oxide-film theory. This second point of view is called the adsorption theory of passivity. adsorbed oxygen decreases the exchange current density (increases anodic overvoltage) corresponding to the overall reaction M → Mz+ + ze−. hence.5) is less noble than the potential for the oxygen evolution reaction O2 + 4 H + + 4e − → 2 H 2 O φ = 1. The second holds that metals passive by Definition 1 are covered by a chemisorbed film—for example.5 intersect the anodic curve at a still more noble potential.92 PASSIVIT Y If the cathodic polarization curves of Fig. is greatly increased over the corrosion rate at less noble potentials within the passive region.5 THEORIES OF PASSIVITY There are two commonly expressed points of view regarding the composition and structure of the passive film. metal oxide or other compound that separates metal from its environment and that decreases the reaction rate. Even less than a monolayer on the surface is observed to have a passivating effect [17]. Such a layer displaces the normally adsorbed H2O molecules and decreases the anodic dissolution rate involving hydration of metal ions. The first holds that the passive film (Definition 1 or 2) is always a diffusion-barrier layer of reaction products—for example.23 V (6. 6. Transpas4 sivity accounts for an observed increase of corrosion rate with time for 18–8 stainless steels in boiling concentrated nitric acid. Examples of protective films that isolate the metal from its environment are (a) a visible lead sulfate film on lead immersed in H2SO4 and (b) an iron fluoride film on steel immersed in aqueous HF. in which corrosion products 2− accumulate. of oxygen. . the corrosion rate of. for example. within the transpassive region.30 V (6. but also with chromium. There is no question on either viewpoint that a diffusion-barrier film is the basis of passivity of many metals that are passive only by Definition 2. but increased corrosion does occur in alkaline solutions that favor formation of ferrate. and the corrosion products become 2− Cr2O7 and Fe3+. for which the potential for the reaction 2 Cr2 O7 − + 14 H + + 12e − → 2Cr + 7 H 2 O φ = 0. FeO2−. Transpassivity occurs not only with stainless steels. it is suggested that the film cannot act primarily as a diffusion-barrier layer. 6. stainless steel.6) Appreciable corrosion in the transpassive region does not occur for iron in sulfuric acid (oxygen evolution is the primary reaction). in particular Cr2O7 . On the nontransition metals. If the surface is abraded. The characteristic high energies for adsorption of oxygen on transition metals correspond to chemical-bond formation. the corresponding potential of which is calculated (see Problem 2 in Chapter 6) using the observed heat and estimated entropy of adsorption of oxygen on iron in accord with the reaction [21] O2 ⋅ Oads on Fe + 6 H + + 6e − → 3H 2O + Fe surface φcalc = 0. evidence has been developed to show that the passive film formed on iron in borate solution is an anhydrous. But such oxides are less important in accounting for passivity than the chemisorbed films that form initially and continue to form on metal exposed at pores in the oxide. which are called physically adsorbed. that is. including those responsible for passivity. which also contains uncoupled electrons (hence its slight paramagnetic susceptibility) resulting in electron-pair or covalent bonding supplementary to ionic bonding. Observed values of the Flade potential are consistent with a chemisorbed film of oxygen on the surface of iron. Furthermore. they contain electron vacancies or uncoupled electrons in the d shells of the atom. Surface analytical techniques that can be used to study films on metals include Auger electron spectroscopy (AES). Metals and alloys in this category have been the source of extended debate and discussion on the mechanism of passivity over the past 150 years. the films are usually invisible. hence. 19]. a property that favors adsorption of the environment because metal atoms tend to remain in their lattice. The adsorption theory emphasizes that the observed Flade potential of passive iron is too noble by about 0. local high temperatures generated at the surface produce a detectable oxide. whereas oxide formation requires metal atoms to leave their lattice. about 2 to 3 nm thick. especially O2.57 V (6. but this is not the passive film. This situation reverses for multilayer adsorbed oxygen films (lower energy of bonding to the metal) that revert in time to metal oxides. The uncoupled electrons account for strong bond formation with components of the environment. On transition metals. crystalline spinel with a “γ-Fe2O3/Fe3O4-like” structure [19]. and any chemisorbed films on the metal surface are short-lived. oxides tend to form immediately. such films are called chemisorbed in contrast to lower-energy films. based on marked anodic polarization.6 V to be explained by any of the known iron oxides in equilibrium with iron. X-ray photoelectron spectroscopy (XPS). and others [18.THEORIES OF PASSIVIT Y 93 For metals that are passive by Definition 1.7) . such as copper and zinc. Low-energy electron-diffraction (LEED) techniques are used to detect adsorbed films. Using these advanced techniques. The adsorption theory derives support from the fact that most of the metals that are passive by Definition 1 are transition metals of the periodic table. the initially chemisorbed oxygen is generally more stable thermodynamically than is the metal oxide [20]. transition metals have high heats of sublimation compared to nontransition metals. secondary ion mass spectrometry (SIMS). 500 × 5. because of their pronounced affinity for oxygen. can occur by direct chemisorption of oxygen from the air or from aqueous solutions. It is the decomposed passive film that is presumably isolated in experiments designed to remove the passive film from iron.4 nm thick (based on apparent area). d = 5. CrO2−. Similarly. which is in reasonable agreement with the calculated value.12) = 5. in 4 2 NaOH solution [14]. On breakdown of passivity. passivity of chromium and the stainless steels. oxygen in air can adsorb directly on iron and passivate it in aerated alkaline solutions.8) Calculations show that the same reaction will not go for chromium or the stainless steels.1). or also in near-neutral solutions if the partial pressure of oxygen is increased sufficiently. This oxidation does not occur with active iron. = 159. (6. by concentrated nitric acid. The agreement is still better if account is taken of adsorbed water displaced from the metal surface by the passive film during the passivation reaction.01 C/cm2 of adsorbed oxygen on iron reacts with underlying metal in accord with O2 ⋅ Oads on Fe + 2 Fe + 3H 2O → 2 Fe(OH)3 or Fe2O3 ⋅ 3H 2O (6. the adsorbed passive film may be represented above by O2·Oads on Fe. .5–10 nm). on which the passive film is more stable.12 g/cm3) is formed equal to a minimum of (0. as cited in Eq. CrO−.07 nm) over which one layer of oxygen molecules (r = 0.012 C/cm2 of passive-film substance in accord with the reaction O2⋅Oads on Fe + Fe + OH − + CrO− + H 2O → Fe(OH)3 + CrO2− 2 4 ΔG = −138. According to the adsorption theory.94 PASSIVIT Y The measured chemical equivalents of passive-film substance (about 0.12 nm) is chemisorbed. hence.9) A layer of Fe2O3 (mol.57 V.000 J (6. 0. corresponds to φobs = 0.63 V . and the equivalents of oxygen so adsorbed were found [22] to be of the same order of magnitude as the equivalents of passive film formed on iron when passivated either anodically. Further confirming evidence that the passive film on iron contains oxygen in a higher-energy state than corresponds to any iron oxide is obtained from the ability of the passive film to oxidize chromite. The maximum oxidizing capacity corresponds to 0. this displacement presumably involving a greater free energy change than the adsorption of water on the passive film itself [8].7)/(6 × 96. 0. to chromate. or by exposure to chromates. The value of the Flade potential.7.01 × 159. The hydrated oxide would be thicker. wt. This value compares in magnitude with measured values for the thickness of the decomposed passive film (2.01 C/cm2) correspond (roughness factor = 4) to one atomic layer of oxygen atoms (r = 0. 6.1 More Stable Passive Films with Time Flade observed that the longer iron remains in concentrated nitric acid. This initial adsorbed layer is found to be more stable thermally than the oxide. becoming a mobile species within the adsorbed oxygen layer. along with schematic structure of a thicker passive film containing additional metal ions and protons in nonstoichiometric amounts. The initial passive film grows into a multilayer adsorbed M·O·H structure. indicate that the first-formed adsorbed film consists of a regular array of oxygen and nickel ions located in the same approximate plane of the surface. At increasing oxygen pressures. but the adsorbed (passive) film remains intact at pores in the oxide. probably consisting of O2. A schematic structure of the first-formed adsorbed passive film is shown in Fig. Low-energy electron-diffraction data for nickel single crystals [23]. NiO. In other words. with the coexisting opposite charges tending to stabilize the adsorbed film. for example. Frankenthal [17] noted that the film on 24% Cr–Fe thickened and became more resistant to cathodic reduction when the alloy was passivated for longer times in 1N H2SO4 at potentials noble to the passivating potential (P in Fig.7. . showed that the passive film on 18–8 stainless steel contains H2O. the longer the passive film remains stable when the iron is subsequently immersed in sulfuric acid [6]. the film is stabilized by continued exposure to the passivating environment. particularly in the noble potential range.1). and such nuclei then grow laterally to form a uniform oxide film. several adsorbed layers. Okamoto and Shibata [24]. although less than monolayer quantities of oxygen (measured coulometrically) suffice to passivate this alloy in this solution. Similarly. the stoichiometric oxide is nucleated at favorable sites on the metal surface. 6. The observed stabilizing effect is likely the result of positively charged metal ions entering the adsorbed layers of negatively charged oxygen ion and molecules. Eventually.THEORIES OF PASSIVIT Y 95 6. form on top of the first layer and result in an amorphous film.7 [20]. Additional metal ions in time succeed in entering such a film. Schematic structure of initial passive films containing less or more than monolayer amounts of adsorbed oxygen. which can Figure 6. Protons from the aqueous environment are also incorporated.5. 96 PASSIVIT Y be considered to be an amorphous nonstoichiometric oxide. nickel. The potential difference between such areas is 0. accompanied by cathodic protection of the metal area immediately surrounding the anode. Halogen ions have less effect on the anodic behavior of titanium. which decreases the rate of metal dissolution. other halogen ions—break down passivity or prevent its formation in iron. tungsten. The effect is so pronounced that iron and the stainless steels are not readily passivated anodically in solutions containing an appreciable concentration of Cl−. that deep pitting is much more common with passive metals than with nonpassive metals. It is evident. with the preferred surface sites being determined perhaps by small variations in the passive-film structure and thickness. In other words. 6. and zirconium. Passivity of these metals may continue in media of high chloride concentration. which lose passivity. Cl− adsorbs on the metal surface in competition with dissolved O2 or OH−.5 V or more. etc. . Cl− favors hydration of metal ions and increases the ease with which metal ions enter into solution. opposite to the effect of adsorbed oxygen. cobalt. such as SO2−. Alternatively. From the perspective of the oxide-film theory. because of the possibility of passive–active cell formation. the less likely it is that another pit will initiate nearby. 4 Cl− may colloidally disperse the oxide film and increase its permeability. if pitting occurs at all. and Fe–Cr alloys. Minute anodes of active metal are formed surrounded by large cathodic areas of passive metal. hence. and the stainless steels. chromium. Ta. This fixes the anode in place and results in pitting corrosion. High current densities at the anode cause high rates of metal penetration. This behavior is sometimes explained by formation of insoluble protective Ti. Mo. the greater the current flow and cathodic protection at any pit. to a lesser extent. tantalum. in accord with the noble critical potentials of these metals above which pitting is initiated. However. Instead. Also. the metal continues to dissolve at high rates in both the active and passive potential ranges. the true situation is probably related to the high affinity of these metals for oxygen.5. molybdenum. On the other hand. according to the adsorption theory. Cl− penetrates the oxide film through pores or defects easier than do other ions. Breakdown of passivity by Cl− occurs locally rather than generally. and the resulting cell is called a passive– active cell. basic chloride films. chromium.. in contrast to the behavior of iron. Once in contact with the metal surface. making it more difficult for Cl− to displace oxygen of the passive film.2 Action of Chloride Ions and Passive–Active Cells Chloride ions—and. It differs markedly in protective properties from the stoichiometric oxide into which it may eventually convert. adsorbed chloride ions increase the exchange current (decrease overvoltage) for anodic dissolution of the above-mentioned metals over the value prevailing when oxygen covers the surface. the observed number of deep pits per unit area is usually less than that of smaller shallow pits. 6 CRITICAL PITTING POTENTIAL When iron is anodically polarized potentiostatically in 1N H2SO4 to which sodium chloride is added (>3 × 10−4 moles/liter). However. 6. and it moves toward more noble values with increasing pH and with lower temperatures [25]. The CPP for 18–8 moves toward less noble values as the Cl− concentration increases. Addition of 3% NaNO3 to 10% FeCl3.8). whereas in the absence of NaNO3 the alloy is observed to corrode by pitting within hours [26]. . and NaClO4 (Fig. instead of oxygen being evolved. accompanied by random formation of pits (Fig. no matter how long the time.1M NaCl showing increasingly noble values of the critical pitting potential with additions of Na2SO4. The added salts under these conditions become effective inhibitors. the metal corrodes locally with formation of visible pits.1N NaCl within the lower passive region. analogous to its behavior in Na2SO4 solution. Similarly. The 3% NaNO3 addition moves the CPP to a value more noble than the open-circuit cathode potential or the reversible potential for Fe3+ + e− → Fe2+. the apparent transpassive region is shifted to lower potentials. NaNO3. 6. there is a rapid increase in the current. The CPP is also moved in the noble direction by adding extraneous salts to the NaCl solution—for example. If sufficient extraneous salt is added to shift the CPP to a value more noble than the prevailing corrosion potential. But above a critical potential. if 18–8 stainless steel is anodically polarized in 0. for example.CRITICAL PIT TING POTENTIAL 97 6. Na2SO4. the alloy remains passive.8). 25°C. called the critical pitting potential (CPP). Figure 6. has prevented pitting of 18–8 stainless steel or any appreciable weight loss for a period of more than 25 years.8. pitting does not occur at all on exposure of the stainless steel to the aqueous salt mixture. Potentiostatic polarization curves for 18–8 stainless steel in 0. 62 0.1. however. as that value needed to build up an electrostatic field within the passive or oxide film sufficient to induce Cl− penetration to the metal surface [34].) = −0. s.h. from one point of view. to a lesser extent. corresponding to increased resistance to pitting [30.088 log (Cl−) + 0.1. depending on size and charge.1N NaCl are listed in Table 6. contaminating the oxide and making it a better ionic conductor favoring oxide growth. φcritical (V. It is observed that an increase in the chromium content of stainless steels— and.37 0.504. The critical potential has been explained.) Ala Ni Zr 18–8 stainless steelb 30% Cr–Fe 12% Cr–Fe Cr Ti a Reference [29] [30] [31] [30] [30] [30] [32] [33] −0. s. Pitting corrosion has been reported in long-term exposure at potentials below the CPP [28]. 200°C) b φcritical (V. where (Cl−) is activity of Cl−.28 0. 31]. for example. The noble critical potentials for chromium and titanium [more noble than the oxygen electrode potential in air (0. CPPs of several metals in 0. values of which were derived from anodic polarization curves. because CPPs are the result of short-term experiments. so that pitting of titanium.98 PASSIVIT Y TABLE 6. is observed in concentrated hot CaCl2 solution despite its immunity to pitting in seawater. an increase of nickel content—shifts the critical potential to more noble values.46 0.) = −0.26 0. it is usually confirmed by holding the potential at the critical value for 12 hours or more and observing absence of pits under a low-power microscope.0 (1N NaCl) ≈1.2 atm) + 4H+ (10−7) + 4e− → 2H2O] indicate that these metals are not expected to undergo pitting corrosion in aerated saline media at normal temperatures. Critical Pitting Potentials in 0.e.1 V below the CPP in 0.124 log (Cl−) − 0. Most of the data were obtained by allowing 5 minutes or more at a given potential and observing whether the resultant current increases or decreases with time. either .0 >1.h.20 >1.1N NaCl at 25°C Potential (V. the critical potentials become more active.0 (1N NaCl. Leckie [27] polarized 18–8 stainless steel at a constant 0. Eventually.8 V) in accord with O2(0.1N NaCl for 14 weeks without observable pitting. they should be used with caution in predicting immunity to pitting corrosion in long-term service. The CPP is the most noble potential for which the current decreases or remains constant. Other anions may also penetrate the oxide. At elevated temperatures and high Cl− concentrations.e. however.e.168. Along the same lines. CPPs can be used to compare the relative susceptibility of different alloys to pitting corrosion.h. s. compared to adsorbed oxygen. Alternatively. compete with 4 Cl− for sites on the passive surface. The metal has typically greater affinity for oxygen than for Cl−. or cathodic polarization of local shielded areas]. which do not break down the passive film or cause pitting.D. The same relation applies to inhibition of pitting of aluminum. making it necessary to shift the potential to a still more noble value in order to increase Cl− concentration sufficient for successful exchange with adsorbed oxygen. The equation can be derived assuming that ions adsorb competitively in accord with the Freundlich adsorption isotherm [26]. pitting is predicted not to occur. pitting could then initiate independent of whether the overall prevailing potential is above or below the critical value. the more Cl− in aqueous solution adsorbs 2− on metallic Cr. or cations of the oxide undergo dissolution at the electrolyte interface. Extraneous anions. Physik. The observed induction period is the time required for successful competitive adsorption at favored sites of the metal surface.) 215. Cl− cannot displace adsorbed oxygen so long as the passive film remains intact. reduced oxygen or depolarizer concentration at a crevice (crevice corrosion). The induction period preceding pitting is related to the time required for the supposed penetration of Cl− through the oxide film.7 CRITICAL PITTING TEMPERATURE Pitting tendency increases with increasing temperature. the concentration of Cl− ions at the metal surface increases. 25 (1960)] reported a quantitative study showing that the more noble the applied potential.CRITICAL PIT TING TEMPER ATURE 99 the oxide is undermined by condensation of migrating vacancies. The relation between minimum anion activity necessary to inhibit pitting of 18–8 stainless steel in a solution of given Cl− activity follows the relation log (Cl−) = k log (anion) + const.g. as well as time for penetration of the passive film. (D. Should passivity break down because of factors other than those described [e. The critical pitting temperature (CPT) is defined as the temperature below which an alloy does not *Rosenfeld and Maximtschuk [Z. This is in contrast to the usual procedure of cathodic protection. 6. pitting results.. But under conditions of uniform passivity for the entire metal surface.R. such as NO3− or SO2− . but. application of cathodic protection to avoid pitting corrosion need only shift the potential of the metal below the critical value. eventually reaching a value that allows Cl− to displace adsorbed oxygen.* Below the CPP. Chem. Adsorbed Cl−. and they also showed that added SO 4 or OH− displaces adsorbed Cl−. hence. in both cases. results in lower anodic overvoltage for metal dissolution. which requires polarization of a metal to the much more active open-circuit anode potential. with the result that the critical pitting potential decreases as temperature increases. as the potential is made more noble. which accounts for a higher rate of corrosion at any site where the exchange has taken place. . the critical potential is explained in terms of competitive adsorption of Cl− with oxygen of the passive film (adsorption theory). Several other alloy systems exhibit critical compositions for passivity. provided that ≥12 wt.9–6. It is found that the passive property of chromium is conferred on alloys of Cr–Fe. The CPT can be used to evaluate and compare pitting susceptibilities of alloys [35]. Typical corrosion. the passive film becomes more stable with increasing chromium content of the alloy. Examples of approximate critical compositions Figure 6. in contrast to iron and copper. Eng.8 PASSIVITY OF ALLOYS Several metals. and they remain bright and tarnish-free for years. 533 (1926). which corrode or tarnish in short time. CPTs are usually measured in an aggressive Cl− solution.% Cr are known as the stainless steels. Chem.% Cr is present. as was first described by Tammann [36]. Note in Fig. 6. and critical-current density behavior of Cr–Fe alloys is shown in Figs. illustrating why >12% Cr–Fe alloys are selfpassivating. In addition. Chappell. Whitman and E. Ind. such as chromium. Corrosion rates of chromium–iron alloys in intermittent water spray at room temperature. American Chemical Society.9.100 PASSIVIT Y pit regardless of potential and exposure time. Copyright 1926. This value is so low that corrosion currents in aerated aqueous media easily achieve or exceed this value. 6. Iron-base alloys containing a minimum of 12 wt. such as a FeCl3 solution. 6. 18.11. are naturally passive when exposed to the atmosphere. potential.] .11 that icritical for passivation of Cr–Fe alloys at pH 7 reaches a minimum at about 12% Cr in the order of 2 μA/ cm2. [Reprinted with permission from W. 6th Meeting. London. Uhlig. Electrochem. Carr. and Kinetics.10. Olivier.11. and P. Schneider. N. International Committee on Electrochemistry. Potentials of chromium–iron alloys in 4% NaCl. Critical current densities for passivation of chromium–iron alloys in deaerated 3% Na2SO4 at pH 3 and 7. p. Proceedings. Trans. . Soc.PASSIVIT Y OF ALLOYS 101 Figure 6. room temperature.] Figure 6. 79. 314. Poitiers. data in 10% H2SO4. Thermodynamics. 25°C [8]. Butterworths. The Electrochemical Society. 111 (1941). 1955. [Reprinted with permission from H. Copyright 1941. are from R. respectively. It . and Cr–Ni–Fe [38]. critical compositions based on corrosion data usually vary with the environment to which the alloys are exposed. has been described both by the oxide-film theory and by the adsorption theory. on the other hand. but not below. With iron. it is considered that. Critical alloy compositions for passivity have also been observed in three. The cathodic reaction rate in an aerated solution when controlled by oxygen reduction. In 33% HNO3. 8%Cr–Co. passivity is stable.0. is given by the diffusion current idiff corresponding to the reaction 2 O2 + − − H2O → 2OH − 2e . Hence. On the other hand. The corrosion rate. the critical pH is approximately 10. Whether the alloy favors a chemisorbed or reaction-product film depends on the electron configuration of the alloy surface.4 and 5. By the adsorption theory. decreasing pH and increasing temperature increase icritical. but immediately reacts below the critical composition to form a less protective or nonprotective oxide film.11. but not stoichiometric Cr2O3 itself. of 18% Cr. but not below.g. for example. Fe–Ni–Mo. on the other 1 hand.39 A/m2). For an unstirred air-saturated solution. chromium) may form protective oxides (e. This effect of environment is to be expected because whether an alloy becomes passive depends on whether the corrosion rate equals or exceeds the specific icritical for passivity in the given environment. it can be shown that idiff = 0. the critical chromium concentration for passivity of Cr–Fe alloys decreases to 7% Cr. Fe–Cr– Ni–Mo [37].. the values of which depend on the anion. a critical pH exists for each alloy composition at which icritical = idiff and above which. for example. in turn. Critical compositions based on icritical values are not sensitive to the pH of the electrolyte in which the polarization data are measured (Fig.039 mA/cm2 (0.1M Na2SO4 [40] are 1. The structure of the passive film on alloys. It has been suggested that protective oxide films form above the critical alloy composition for passivity.and four-component alloys—for example. 8% Ni.. and the fact that the passive film on stainless steels can be reduced cathodically. as with passive films in general. The observed critical pH values. As seen in Fig. only in aqueous solutions of about pH 7 is the critical composition equal to 12%. 6. and 14% Cr–Ni. The electron configuration theory describes the specific alloying proportions corresponding to a favorable d-electron configuration accompanying chemisorption and passivity.102 PASSIVIT Y determined from plots of icritical versus alloy composition are 35% Ni–Cu. in close agreement with the calculated values. in the presence of water. Cr2O3) above a specific alloy content. 6. in particular on d-electron interaction. No quantitative predictions have been offered based on this point of view. The preferential oxidation of passive constituents (e. for which icritical is much higher than for the stainless steels.11). but nonprotective oxide films form below the critical composition. and of 12% Cr stainless steels in 0.g. increasing chromium content reduces icritical. remains unexplained. oxygen chemisorbs on chromium–iron alloys above the critical composition corresponding to passivity. whereas in FeSO4 solution it increases to 20% Cr. It can be shown from first principles [39] that icritical = K(H+)λ where K and λ are constants. 15% Mo–Ni. depends on factors of metal and environment affecting the rate of both cathodic and anodic reactions. marine fouling organisms are much less successful in establishing themselves on the surface of nonpassive nickel–copper compositions because the Figure 6.6 d-electron vacancy per atom (as measured magnetically). Corrosion rates of copper–nickel alloys in aerated 3% NaCl. Similarly. a nontransition metal containing no d-electron vacancies. 6. Practically. when alloyed with copper.12.13). Corrosion pitting in seawater is observed largely above 40% Ni because pit growth is favored by passive–active cells (see Section 6. in the range of high nickel compositions.I. 80°C.8.1 Nickel–Copper Alloys Nickel. 6.% Ni. by corrosion pitting behavior in seawater (Fig. confers passivity on the alloy above approximately 30–40 at. [41–43] or by decay (Flade) potentials (Fig.13). why the properties of nickel dominate above those of copper in the nickel–copper alloys at >30–40% Ni. and. but not Monel (70% Ni–Cu).). and such cells can operate only when the alloy is passive—that is. by measured values of icritical and ipassive (Fig. this distinction is observed in the specification of materials for seawater condenser tubes in which pitting attack must be rigorously avoided.T. 48-h tests (M. more quantitatively. Corrosion Lab. 6. 6. 6.12 and 6. containing 0. Initiation of passivity beginning at this composition is indicated by corrosion rates in sodium chloride solution (Figs.14). . The cupro nickel alloys are used (10–30% Ni).PASSIVIT Y OF ALLOYS 103 suggests the nature of electron interaction that determines which alloyed component dominates in conferring its chemical properties on those of the alloy—for example.15) [44] in 1N H2SO4.5). J. Soc. Values of critical and passive current densities obtained from potentiostatic anodic polarization curves for copper–nickel alloys in 1N H2SO4. Nav.14. The Electrochemical Society. Am. Figure 6.13. 53.104 PASSIVIT Y Figure 6. 25°C [42]. Copyright 1961. LaQue. Eng. 29 (1941)]. Behavior of copper–nickel alloys in seawater [F. (Reproduced with permission.) . Potentiostatic anodic polarization behavior of nickel–copper alloys in 1N H2SO4 (Fig. Pure copper behaves like Alloy D [36]. latter corrode uniformly at rates that release enough Cu2+ to poison fouling organisms. In other words.PASSIVIT Y OF ALLOYS 105 Figure 6. This observation is interpreted as a filling of vacancies in the d band of electron energy levels of nickel by electrons donated by copper. LaQue and W. *The minimum concentration of Cu2+ required to poison marine organisms corresponds to a corrosion rate for copper of about 0. Electrochem. the magnetic saturation moment. if copper and nickel atoms are considered to be alike. to be just filled by electrons from copper at 60 at. 6.5 gmd [F. they no longer contain d-electron vacancies. In this connection. Clapp. except that copper contains one more electron per atom than nickel. Soc. 25°C (two time scales). that is.15. Polarization curves of alloys containing >70% Cu or <30% Ni resemble those of pure copper. Physicists [45] have made the assumption that.14) establishes that the passive current density largely disappears above 60% Cu and vanishes completely at about 70% Cu. 6. but not otherwise. as observed.15) confirm that a passive film is formed on anodically passivated alloys containing >40% Ni. 103 (1945)].001 ipy or 0. hence. such alloys are not passive. then the 0. Trans.13).* But on passive nickel–copper compositions. 87. 6. Potential decay curves (Fig. Potential decay curves for nickel–copper and nickel–copper–zinc alloys in 1N H2SO4. fouling organisms in general can gain a foothold and flourish (Fig. becomes zero at >60% Cu. alloys containing copper above the critical composition lose their transition-metal characteristics.6 delectron vacancy per nickel atom is expected.% Cu. for which the overall corrosion rate is much less. . which is also a function of d-electron vacancies in the alloy. % Ni) = 1× at.% Ni. accounting for the measured 0. Non-transition-metal additions of valence >1 should shift the critical composition for passivity to higher percentages of nickel. By plotting 1× at. The relevant equations become at. the uncoupled electrons of any single nickel atom tend to couple with uncoupled electrons of neighboring atoms.% Ni (2 − 0. then the vacancies per nickel atom can be set equal to 2 − (2 − 0.% Ni) + v at. This value corresponds closely to the observed value derived from magnetic saturation data. with nickel–copper alloys and noting the specific compositions at which icritical and ipassive merged or at which Flade potentials disappeared. corresponding to two d-electron vacancies or to the equivalent of two uncoupled d-electrons in the third shell of the atom.% Cu = (100 − at. The alloy loses its transition-metal characteristics beginning at the composition for which the total number of d-vacancies equals the total number of donor electrons (one per copper atom). In the gaseous state. it was necessary to assume approximately 80% instead of 100% donation of valence electrons.% Ni/100.% Ni (2 − 0. one zinc atom of valence 2 or one aluminum atom of valence 3 should be equivalent in the solid solution alloy to two or three copper atoms.16 [46].014 at.% Ni (2 − 0. The maximum number of d-electrons that can be accommodated is 10.014 at. whereas transition-metal additions should have the opposite effect. For example. corresponding to two vacancies for 0% Ni and 0.% Ni) = 1 × at.% Y and at. This results in a smaller number of electron vacancies in the solid compared to the gas. respectively.014 at.% Z for transition-metal additions.% Cu where n is the number of electrons donated per atom of Y and v is the number of vacancies introduced per atom of Z. If plotted instead with v at.014 at.106 PASSIVIT Y One can also start with the alternative assumption that the two atoms maintain their individuality and that the vacancies per atom of nickel are a function of alloy concentration [46]. a straight line is predicted of unit negative slope for nontransition-element additions.% Cu− at. a unit positive slope should result.% Y. the critical composition is found to be 41 at.% Cu + n at.6 vacancy for 100% Ni.6) at.% Cu. 6. This has been confirmed experimentally [47]. If we assume that the intercoupling of delectrons increases with proximity of nickel atoms in the alloy and is a linear function of nickel concentration.% Ni) and solving. A plot for the alloying additions so far studied is shown by data summarized in Fig. Setting at.% Ni) with n at.6 vacancy or 0. In order for the line to pass through the origin with unit slope. or transition elements Z. or at.6 uncoupled electron per nickel atom.% Z = 1 × at.% Ni (2 − 0. This model was checked by alloying small amounts of other nontransition elements Y. nickel has the configuration of 3d84s2. corresponding to copper: 3d104s. In the process of condensing to a solid and forming the metallic bond. This means that valence electrons from copper and . and zinc shown in Fig.16 are 4. For iron and cobalt. or electrons donated by alloying additions. and 2.8 electron donor per copper atom in binary nickel–copper alloys. aluminum. respectively. Plot of excess electrons or d-electron vacancies in nickel–copper alloys at their critical compositions versus electron vacancies. respectively.% instead of 41 at. 6. which is close to the value 1. . therefore.% Z/100.* This value is consistent with the composition at which ipassive and icritical intersect in Fig. Values of vg for iron and cobalt are 4 and 3. the critical nickel composition below which the d-band is filled becomes 35 at.014 × 35 = 1. unit slope. For gallium. between observed and predicted critical *The calculated number of d vacancies per Ni atom at this composition equals 2 − 0. 3.51.16. the number of vacancies per atom is equated to vg − (vg − vs) at. in accord with their normal valence. Assuming 0. respectively. 6. Values of n for germanium. 39]. There are no observed phase changes or major discontinuities in the thermodynamic properties of nickel–copper alloys at 60–70% Cu.PASSIVIT Y OF ALLOYS 107 Figure 6. but other explanations are also possible.% as calculated earlier.6 assumed earlier [33.14. a valence of 2 is in better accord with the other data. reflecting perhaps the known tendency of gallium to form chemical compounds having a valence lower than 3. whereas chemisorption on any metal is known to be favored by an unfilled d-band configuration [48].7. from other nontransition elements are presumably donated in major part to nickel. The good agreement.2 and 1. and for vs they are 2. but not entirely. where vg and vs are d-electron vacancies per atom in the gas and solid. which occurs for alloys containing less than 12% Cr. the corrosion behavior of the alloy is like that of iron. the latter being available to fill vacancies in the d-band of the metal. catalytic properties also depend on ability of a metal or alloy to chemisorb certain components of its environment.. the d-electron vacancies of chromium are assumed to fill with electrons from alloyed iron [41]. to chlorides) so long as the d-band of energy levels for molybdenum remains unfilled. and passive iron are cathodically polarized. is no longer able to chemisorb oxygen or become passive. 2– 3% Mo). Hydrogen so dissolved is partly dissociated into protons and electrons. according to the adsorption theory. In Type 316 stainless steel (18% Cr.% Mo [51]. but by iron at lower nickel compositions [50]. that . useful simplifying assumptions can be made. passivity is destroyed. 6. Similarly. and the alloy behaves more like chromium. the most passive component of an alloy is assumed to be the acceptor element. can also be related to the contribution of electrons from nickel or cobalt to the unfilled d-band of chromium [49]. It is not surprising. the d-electron vacancies of chromium are unfilled. therefore. for stainless steels. In the ternary Cr–Ni–Fe solid solution system. but also an adsorbed structure of the passive film. stainless steels. therefore. electrons are donated to chromium mostly by nickel above 50% Ni. equal to 14% Cr and 8% Cr.2 Other Alloys Because present-day theory of the metallic state does not treat the situation.108 PASSIVIT Y alloy compositions supports not only an effect of electron configuration on passivity.g. At this ratio or above. Accordingly. Nevertheless.9 EFFECT OF CATHODIC POLARIZATION AND CATALYSIS When chromium. In addition. The critical compositions for passivity in the Cr–Ni and Cr–Co alloys. For example. 6. 10% Ni. Above 12% Cr. respectively. This occurs by reduction of the passive oxide or adsorbed oxygen film. molybdenum alloys retain in large part the useful corrosion resistance of molybdenum (e.8. the weight ratio of Mo/Ni is best maintained at or above 15/85. A transition metal containing sufficient hydrogen. corresponding to the observed critical ratio for passivity in the binary molybdenum–nickel alloys equal to 15 wt. which tends to share electrons donated by the less passive components. the electron configuration of alloys made up of two or more transition metals with relation to their passive behavior is not as well understood as for the copper– nickel system. according to whichever viewpoint of passivity is adopted. At the critical composition at which vacancies of Cr are apparently filled. passive properties imparted by molybdenum appear to be optimum. for example. Often. hydrogen ions discharged on transition metals tend to enter the metal as hydrogen atoms. because the d-band is filled. favorable collisions of gaseous H with adsorbed H are less probable.17. the filled dband is less favorable to hydrogen adsorption. and the reaction rate decreases. The similarity to passive behavior of copper–nickel alloys. Gold.17 [53]. Correspondingly. hence. which contains 0. REFERENCES 1. This behavior is explained by filling of the d-band by electrons of dissolved hydrogen. is charged with hydrogen cathodically. 4th Symposium on Passivity. H. The parallel conditions affecting passivity and catalytic activity support the viewpoint that the passive films on transition metals and their alloys are chemisorbed. and magnetic measurements confirm that the d-band is just filled at the critical gold concentration. it loses its efficiency as a catalyst for the ortho–para hydrogen conversion [52]. The catalytic effect of copper–nickel alloys as a function of composition for the reaction 2H → H2 is shown in Fig. p. editors. supplies electrons to the unfilled d-band of palladium.REFERENCES 109 Figure 6. Passivity of Metals. which also decreases above 60 at. Princeton. History of passivity. Catalytic efficiency of nickel–copper alloys for 2H→H2 as a function of alloy composition (efficiency increases with values of γH) [53]. 1978. Electrochemical Society.) transition metals are typically good catalysts and that the principles of electron configuration in alloys favoring catalytic activity are similar to those favoring passivity.6 d-electron vacancy per atom in the metallic state.% Cu. as a result of which the metal no longer chemisorbs hydrogen. Above 60 at. Frankenthal and J. . For example. Proceedings. R. (Reproduced with permission of The Royal Society of Chemistry. Uhlig. the catalytic efficiencies of palladium–gold alloys resemble palladium until the critical concentration of 60 at. a non-transition metal. experiments and theories. 6. 1. NJ. can be noted. Kruger.% Au is reached. when palladium.% Cu. at which composition and above the alloys become poor catalysts. 117 (1955). Pogg. King. . K. Soc. Gilman. 26. Uhlig and J. 99. 113. Austenitic and Ferritic Stainless Steels. editor. J. 23. Physik. J. 57. Electrochem. 44. Gatos and H. Soc. J. 35. 262 (1953). 65 (1997). Surface Science 52. M. 20. 18. Revie. J. Soc. 127. Z. F. R. Rocha and G. 216 (1953). Kolotyrkin. 836 (1961). 8. Mater. 27. J. 5. 14. Corros. 2026 (1959) 9. Uhlig. Z. 26. Bohlmann. 115. Chem. 751 (1965). 62. Experimental Researches in Electricity. Electrochem. (with E. Z. 6. Böhni and H. in Uhlig’s Corrosion Handbook. Frankenthal. J. Leckie and H. J. C. F. editor. Uhlig and S. New York. 4A. Mears. 63. Soc. 664 (1975). Uhlig. Uhlig. Arch. J. 4. Wiley. P. Simpson. Uhlig. Electrochem. Second Symposium on Fresh Water from the Sea. Kaesche. H. 108. Schönbein. 15. May. 76. Tomashov. 1967. 36. 513 (1911). Greene. Uhlig and P. A. 475 (2003). Soc. Leckie. 169. pp. J. M. 562 (1955). King and H. J. 279 (1965). Stern. N. Chem. Wiley. Böhni. Jr. 542 (1967). J. 151 (1928). C. New York. G. Corrosion 59(6). 319 (1964). 77 (1953). Flade. 906 (1969). J. 620. Y. Electrochem. 2000. Graham. 229t (1958). Uhlig. Electrochem. 12. Burns. Horvath and H. Corros. H. Anorg. Sci. O’M. 13. Z. 2000. Chem. Bockris. 37. O’Connor. 390 (1836). Soc. 19. Corros. J. and N. H. 32. N. 2nd edition. F. 30. 325 (1967). Athens. Poser and E. ibid. Soc. Electrochem. 21. J. Vol. J. 759 (1959). Soc. 257 (1923). H. Revie. and J. 10. Nature 206. Z. C. 8. B. Revie. Baker. New York. Surface Sci. H. Corrosion 20. Naturforschung. Faraday. 5. 7. Stern and H. 106. Corrosion 19. Hoar. Baker. 1460 (1975). Sotter). Soc. H. (F. and M. 102. 24. 5. Hall. H. 7. Chem. Eisenhüttenw. W. Electrochem. 114. Wagner. Localized Corrosion of Passive Metals. R. 36. Elektrochem. 289t (1964). Rothwell. 2nd edition. 175–176. Greece. G. 31. 1350 (1965). Z. Bishop. 33. Cartledge and R. Perform. M. D. F. R. H. Streicher.. II. 3. 791 (1968). Lennartz. Sci. T. Electrochem. Uhlig and T. Hackerman. 117. Phys. Soc. Allgem. C. Metallk. 29. O’M. G. 22. Sci. W.D. Okamoto and T. Phys. 398 (1937). M. Soc. in Uhlig’s Corrosion Handbook. 261t (1963). 16. Soc. R. H. and B. 106. W. Uhlig. Corrosion 14. H. Chem. 122. 1262 (1966). Uhlig. N. H.) 34. Phys. Dover. Shibata. R. 116. MacRae. Electrochem. 25.R. and G. 378 (1949). H. J. G. 626 (1958). 100. 34. 1 (1959). Wissenberg. reprinted 1965. 17. Electrochem. Phys. Bockris. U. J. H. Electrochem. H. Chem. J. 11. 973 (1957). 28. Electrochem. J. Physik. Revie. G.110 PASSIVIT Y 2. Appl. 250 (1952). 1152 (1970). A. 61. 87 (1962). Ann. W. Tammann. Lord. Franck. p. Bonhoeffer. M. R. 1. 13A. R. D. Bond and H. Passivity. 61–67. G. 2003. Takahashi. 900 (1968). Uhlig. Uhlig. J. Soc. and Properties Of Thin Oxide Layers. 108. Hardy and J. Revie. Wiley. Linnett. Wiley. 1999. NJ. Soc. Frankel. 2nd edition. Testing. Mansfeld and H. 8. in ASM Handbook. Marcus and V. J. 51. H. 1997. in ASM Handbook. and R. Kittel. Isaacs. and Protection. Electrochem. Electrochem. New York. Elektrochem. 27 June–1 July 2005. Kruger. Acta 16. 43. 2000. 1939 (1971). J. 2006. 13A. 38. Soc. Discussions Faraday Soc. 62. A Selection of Papers from the 9th International Symposium. Rodda. editor. 9. Soc. 85. P. Soc. pp. and J. 1964. 40. Pennington. M. 2003. H. Electrochem. J. Electrochem. H. 650 (1963). OH. 107. Uhlig. Eley. H. 66. 1999. in Uhlig’s Corrosion Handbook. 117. Uhlig. Pitting corrosion. Electrochim. Vol. Marcus. 2nd edition. and Protection. 327 (1961). P. Maurice. Uhlig. 47. ASM International. 115. R. 70 (1968). Luo. M. F. C. Passivity and Localized Corrosion. S. P. R. editors. and H. NJ. Seo. F. Uhlig. 53. New York. Uhlig. W. Electrochem. H. May 9–15. Electrochem.GENER AL REFERENCES 111 37. Feller and H. Passivation of Metals and Semiconductors. J. A. H. A. Introduction to Solid State Physics 3rd edition. Butterworths. Testing. in Uhlig’s Corrosion Handbook. 50. H. H. Faraday Soc. Electrochemical Society. and M. J. Janik-Czachor. 39. 44. Seo. 49. Amsterdam. 173–190. editors. 581. Feller and H. MacDougall. 236–241. 110. 172 (1950). Uhlig. F. 45. W.C. G. Sci. 41. 447 (1970). J. Pennington. 42. Washington. Ives. 700 (1958). . V. Trans. 108. Corrosion: Fundamentals. GENERAL REFERENCES H. M. A. Corrosion 24. editors. Mansfeld and H. Leckie. Osterwald and H. 107.. 307 (1944). 48. Localized corrosion of passive metals. Alberta. Materials Park. Trapnell. Uhlig. Natishan. Electrochem. Kruger. 488 (1960). D. J. Soc. Proceedings of the Eighth International Symposium. 8. Chemisorption. 864 (1960). Couper and D. OH. J. H. 2001. B. editor. Uhlig. Passivity of Metals and Semiconductors. pp. 165–171. Kelly. J. Electrochem. ASM International. J. H. Jasper. editors. Passivity. NJ. Elsevier. B. H. Feller. Corros. Canada. 2000. P. p. L. Uhlig. New York. Bond. Vol. 52. Soc. Mansfeld and H. Hayward and B. W. Materials Park. France. pp. Macagno. 46. An International Symposium in Honor of Professor Norio Sato. 427 (1970). S. Trans. 1968. Revie. Uhlig. 515 (1961). Wiley. p. M. Electrochem. 864 (1960). 377 (1969). pp. Paris. Corrosion: Fundamentals. Electrochemical Society. Electrochemical Society. Soc. 107. Pennington. Böhni. Z. J. Soc. Proceedings of the Symposium on Passivity and Its Breakdown. H. 5.112 PASSIVIT Y PROBLEMS 1. calculate the apparent free energy of formation of the passive film per gram-atom of oxygen. 0. ΔG° = 0.2 kJ/mole H2O. 3. allowing a pit to initiate. for H2O (l). and (c) chemisorbed oxygen.001M ZnSO4(pH = 3. . a = k1(anion)1/n1.2atm). and for H+. O 2(0. ΔH° = −314 kJ/mole O2. where (anion) is the minimum activity of anion necessary to inhibit pitting of a passive metal in a chloride solution of activity (Cl−). Zn assuming that passive stainless steel acts as a reversible oxygen electrode. For chemisorption of oxygen on Fe. (D for O2 at 25°C = 2 × 10−5 cm2/s) (Hint: Equate limiting diffusion current for reduction of oxygen to critical current density for passivity.). A pit in 18–8 stainless steel exposed to seawater grows to a depth of 6. for Fe3O4. From the standard Flade potential for iron. Calculate the absolute emf of the cell: stainless steel.) 2.0 kJ/mole Fe2O3. The steady-state passive current density for iron in 1N H2SO4 is 7 μA/cm2.0). Calculate the standard Flade potential at 25°C for iron. Derive the relation log(Cl−) = k log(anion) + const. (Data: ΔG° of formation for Fe2O3 = −741. Do the same for nickel and chromium. ΔG° = −237.5 cm in one year. To what average current density at the base of the pit does this rate correspond? 6.2 is (a) Fe2O3. How many atom layers of iron are removed from a smooth electrode surface every minute? 7.2) in Section 6. Calculate the minimum concentration of oxygen in milliliters per liter necessary to passivate iron in 3% Na2SO4. assuming that the passive-film substance represented schematically as O·M in Eq. Assume that the amount of ion a adsorbed per unit area follows the Freundlich adsorption isotherm. (6. where k1 and n1 are constants. ΔS° = −193 J/ mole O2-°C (estimated from ΔS° value for chemisorbed N2. What is the change of emf per 1 atm O2 pressure? 4. (b) Fe3O4. ΔG° = −1014 kJ/ mole Fe3O4. and that at a critical ratio of adsorbed Cl− to adsorbed anion the passive film is displaced by Cl−. also for 12% Cr–Fe alloy. 57 V.5. 2.PROBLEMS 113 Answers to Problems 1. 5.800.032 V/atm at 0. 0. 6 × 10−3 A/cm2. 4. 0. For iron. −115 kJ/g-atom O.32 mL/liter. 1. 6. (a) −0.2 atm O2 pressure. (c) 0. 3.051 V. 31. (b) −0. 0. .90 V.085 V. . the equilibrium potential of the hydrogen electrode becomes more noble. as shown by lack of pronounced polarization when iron is made anode employing an external current.7 IRON AND STEEL 7. this reaction proceeds slowly in alkaline or neutral aqueous media. by R. as described earlier. as pH is reduced. At anodic areas. the rate is usually controlled by the cathodic reaction. relates corrosion to a network of short-circuited galvanic cells on the metal surface. φC − φA becomes larger. and the resulting corrosion current increases. 115 . is much slower (cathodic control). the cathodic reaction is H+ → 1 H2 − e− 2 (7. in general. Metal ions go into solution at anodic areas in an amount chemically equivalent to the reaction at cathodic areas. which. Inc. hence. Winston Revie and Herbert H.1 INTRODUCTION The electrochemical theory of corrosion. Uhlig Copyright © 2008 John Wiley & Sons. The corrosion Corrosion and Corrosion Control.2) This reaction proceeds rapidly in acids. the following reaction takes place: Fe → Fe2+ + 2e − (7. In deaerated solutions. On the other hand. When iron corrodes.1) This reaction is rapid in most media. we obtain 1 Fe + H 2 O + O2 → Fe(OH)2 2 (7. is less than 0.1 Effect of Dissolved Oxygen 7. that is. highpurity iron continues to corrode in acids. In . in accord with 1 1 Fe(OH)2 + H 2 O + O2 → Fe(OH)3 2 4 (7.3). access to dissolved oxygen converts ferrous oxide to hydrous ferric oxide or ferric hydroxide. The rate of hydrogen evolution at a specific pH depends on the presence or absence of low hydrogen overpotential impurities in the metal. although white when the substance is pure. for example.5) Hydrous ferric oxide is orange to reddish brown in color and comprises most of ordinary rust. The color of Fe(OH)2. It exists as nonmagnetic αFe2O3 (hematite) or as magnetic γFe2O3. but at a measurably lower rate than does commercial iron.2. using the reaction H2O → H+ + OH−. The pH of saturated Fe(OH)2 is about 9. In neutral or near-neutral water at ambient temperatures.1) and (7. is normally green to greenish black because of incipient oxidation by air.1. the metal surface itself provides sites for H2 evolution.2. with the α form having the greater negative free energy of formation. Hence rust films normally consist of three layers of iron oxides in different states of oxidation. independent of the presence or absence of impurities in the metal.3) Dissolved oxygen reacts with hydrogen atoms adsorbed at random on the iron surface.4) Hydrous ferrous oxide (FeO·nH2O) or ferrous hydroxide [Fe(OH)2] composes the diffusion-barrier layer next to the iron surface through which O2 must diffuse.116 IRON AND STEEL rate of iron in deaerated water at room temperature. For pure iron. A magnetic hydrous ferrous ferrite. The oxidation reaction proceeds as rapidly as oxygen reaches the metal surface. a process called depolarization: 1 2 H + + O2 → H 2 O − 2e − 2 (7. The cathodic reaction can be accelerated by dissolved oxygen in accord with the following reaction. At the outer surface of the oxide film.2 AQUEOUS ENVIRONMENTS 7.5.005 mm/y (0. dissolved oxygen is necessary for appreciable corrosion of iron. Saturated Fe(OH)3 is nearly neutral in pH. Adding (7. Fe3O4·nH2O. 7.1 Air-Saturated Water. greater stability. often forms a black intermediate layer between hydrous Fe2O3 and FeO. hence. so that the surface of iron corroding in aerated pure water is always alkaline.1 gmd). g. the critical concentration of oxygen above which corrosion decreases again. (5.6) where D is the diffusion coefficient for dissolved oxygen in water (cm2/s). is about 12 mL O2/liter (Fig.1).45 mm/y [1]. Since the diffusion rate at steady state is proportional to oxygen concentration. beyond a critical concentration. corresponding to a corrosion rate of 0. we obtain i = 38.1. The concentration of dissolved oxygen in airsaturated. it follows from (7. Apparently. In distilled water. airsaturated fresh water is [cf. in the presence of halide ions. Section 6.1 to 0.E.4 V in oxygen-saturated H2O (28 mL O2/liter).5). 7.05 cm. more oxygen reaches the metal surface than can be reduced by the corrosion reaction—the excess. tending to be higher the greater the relative motion of water with respect to iron. Eq.2..2 Higher Partial Pressures of Oxygen.H.AQUEOUS ENVIRONMENTS 117 air-saturated water. of steel in stagnant. 6. at crevices). 7. the initial corrosion rate may reach a value of about 10 gmd. particularly at higher temperatures. passive–active cells are established in the event that passivity breaks down locally (e.2) [3].5 gmd. is available to form the passive film. the corrosion rate at room temperature is negligible both for pure iron and for steel.1.3) that the corrosion rate of iron is also proportional to oxygen concentration. as shown by potentials of iron in air-saturated H2O of −0. In the absence of dissolved oxygen. c is the concentration of oxygen. at higher partial pressures. therefore. the theoretical corrosion current density. The decrease of corrosion rate is caused by passivation of iron by oxygen. Such breakdown is accompanied by severe pitting.6 × 10−6 A/cm2.4 to −0. and c = 8/32 × 10−3 mole/liter at 25 °C. Although increase in oxygen concentration at first accelerates corrosion of iron. This value increases with dissolved salts and with temperature. i (A/cm2). it is found that. the corrosion drops again to a low value [2]. In the absence of diffusion-barrier films of corrosion products on the surface.5 V versus S. n = 4 eq/mole. increasing the cathodic reaction rate by higher oxygen concentration increases polarization of anodic areas until the critical current density for passivity is reached (see Fig. the critical oxygen concentration reaches the value for airsaturated water (6 mL O2/liter) and is still less for more alkaline solutions.500 C/eq. and 0. Using D = 2 × 10−5 cm2/s.4] i= ( DnF )c × 10 δ −3 (7. δ = 0. This behavior limits the practical use of high partial oxygen pressures as a means of reducing . natural fresh waters at ordinary temperatures is 8–10 ppm. and the other terms have their usual significance. Section 5. or at a critical pressure of oxygen where passivity is on the verge of either forming or breaking down. Because passivity accompanies higher oxygen pressures. At pH of about 10. δ is the thickness of the stagnant layer of electrolyte next to the electrode surface.0–2. and it decreases with an increase in velocity and pH. F = 96. The steady-state corrosion rate may be 1. This rate diminishes over a period of days as the iron oxide (rust) film is formed and acts as a barrier to oxygen diffusion. (Reproduced with permission.1. enhanced erosion corrosion. have been reported to undergo MIC. process chemicals. and in such media increased oxygen pressure results in an increased corrosion rate. foods. Thiobacilli [sulfur-oxidizing bacteria (SOB). MIC has been documented to take place in seawater. and hydrogen embrittlement [4]. In appreciable concentration of chlorides. Microbiologically influenced corrosion can cause various forms of localized corrosion. soils. 48-h test.118 IRON AND STEEL Figure 7. All engineering alloys. hydrocarbon fuels. MIC can also be caused by other types of microorganisms—for example. with the exception of titanium and high chromium–nickel alloys. passivity of iron is not established at all. are a very common cause of corrosion and have been extensively studied. As a result of MIC. stress corrosion cracking.1. including pitting. Effect of oxygen concentration on corrosion of mild steel in slowly moving distilled water. and it can occur at very high rates. enhanced galvanic corrosion. Furthermore. saliva. human plasma. including microalgae. including both bacteria and fungi. fresh water.2. and sewage [4]. dealloying.) corrosion of steel.3 Microbiologically Influenced Corrosion. organisms that cannot be seen individually with the unaided human eye. 7. bacteria. as in seawater. and fungi [4]. active only in anaerobic (oxygenfree) environments. . corrosion can occur at locations where it would not be predicted. Microbiologically influenced corrosion (MIC) is corrosion that is caused by the presence and activities of microorganisms—that is. 25 °C [2]. Copyright 1955. The Electrochemical Society. Although sulfate-reducing bacteria (SRBs). distilled/demineralized water. which oxidize sulfur compounds to sulfuric acid] and other acid-producing microorganisms. They thrive only under anaerobic conditions in the pH range of about 5. and are found in many waters and soils. which is normally adsorbed on the metal surface and which the bacteria use in reducing SO4.g. Certain varieties multiply in fresh waters and in soils containing sulfates. Their relation to an observed accelerated corrosion rate in soils low in dissolved oxygen was first observed by von Wolzogen Kühr in Holland [7]. Sulfate-reducing bacteria easily reduce inorganic sulfates to sulfides in the presence of hydrogen or organic matter. others flourish in brackish waters and seawater. Analysis of a rust in which sulfate-reducing bacteria were active shows this approximate ratio of oxide to sulfide. the action of sulfatereducing bacteria as the cause of corrosion in a water initially free of sulfides can be detected by adding a few drops of hydrochloric acid to the rust and noting the smell of hydrogen sulfide. and still others are stated to exist in deep soils at temperatures as high as 60–80 °C (140–175 °F). therefore. Qualitatively. When the acidic products of bacterial action are trapped at the biofilm–metal interface. probably act essentially as depolarizers. The bacteria. one equivalent of Fe2+ enters solution to form rust and FeS.AQUEOUS ENVIRONMENTS 119 Microbes can adhere to metal surfaces forming a biofilm. 5]. Desulfovibrio desulfuricans)..5. For each equivalent of hydrogen atoms they consume.9) (7. and they are aided in this process by the presence of an iron surface.11) Ferrous hydroxide and ferrous sulfide are formed in the proportion of 3 moles to 1 mole. leading to corrosion [4.5–8. consisting of a community of microorganisms.8) (7. measuring about 1 × 4 μm. A possible reaction sequence can be outlined as follows: Anode: 4 Fe → 4 Fe2 + + 8e − Cathade: 8 H 2 O + 8e − → 8 Hads on Fe + 8OH − 8 Hads + Na 2SO4 ⎯bacteria → 4 H 2O + Na 2S ⎯⎯⎯ Na 2S + 2H 2CO3 → 2 NaHCO3 + H 2S (7. These high rates have been traced to the presence of sulfate-reducing bacteria (e. their impact on corrosion is intensified [6]. The aid that iron provides in this reduction is probably to supply hydrogen. Although iron does not corrode appreciably in deaerated water.10) Summary: 4 Fe + 2 H 2O + Na 2SO4 + 2 H 2CO3 → 3Fe(OH)2 + FeS + 2 NaHCO3 (7. . The bacteria are curved.7) (7. the corrosion rate in some natural deaerated environments is found to be abnormally high. the effect of pH may be different in a hard water. In an open vessel. whereas municipal water using similar wells. water on corrosion of iron at room temperature is shown in Fig. on the other hand. the rate increase is more than double for every 30 °C rise in temperature.2. Midwest caused failure of a galvanized water pipe 50 mm (2 in. 7. the rate increases with temperature to about 80 °C (175 °F) and then falls to a very low value at the boiling point (Fig. 7. for example. 7. water-cooled rolling mills. allowing dissolved oxygen to escape. but fungi may be capable of growing under such conditions [10]. but microorganisms are capable of becoming resistant to specific chemicals after long-term use. Chlorination is used to eliminate bacteria that cause corrosion. or soft.12) Severe damage by sulfate-reducing bacteria has occurred particularly in oil-well casing. buried pipelines. such as the reduction of H2S: 2 H 2S + 2e − → 2 HS− + H 2 (7. but this treatment can produce byproducts that are environmentally unacceptable [10].8).) in diameter by action of sulfate-reducing bacteria. and the corrosion rate continues to increase with temperature until all the oxygen is consumed.3 Effect of pH The effect of pH of an aerated pure. The rate for iron corroding in hydrochloric acid. Within 2 years. In a closed system. well water in the U. was much less corrosive.2). but which was chlorinated beforehand. When corrosion is attended by hydrogen evolution. Addition of certain biocides can be beneficial. oxygen cannot escape.2 Effect of Temperature When corrosion is controlled by diffusion of oxygen. Aeration of water reduces activity of anaerobic bacteria since they are unable to thrive in the presence of dissolved O2. 9]. The lower corrosion rate above 80 °C is related to a marked decrease of oxygen solubility in water as the temperature is raised.3 [12]. A combination of low temperature and low humidity is one approach to controlling the growth of bacteria. however.120 IRON AND STEEL In addition to the cathodic reaction listed in (7.2. in which a protective film of CaCO3 forms on the metal surface. approximately doubles for every 10 °C rise in temperature. and pipe from deep water wells. Regular cleaning is a good practice to prevent biofilm formation and subsequent corrosion. .S. Eradication of microbial populations may be achieved by combining several chemicals or by increasing the concentration of a biocide [10]. other cathodic reactions could be considered [8. the corrosion rate at a given oxygen concentration approximately doubles for every 30 °C (55 °F) rise in temperature [11]. and this effect eventually overshadows the accelerating effect of temperature alone. 7. 168. and V. 3rd edition. Regardless of the observed pH of water within this range. the . The major diffusion barrier of hydrous ferrous oxide is continuously renewed by the corrosion process.) Figure 7. (Reprinted with permission from F. McGraw-Hill. 665 (1924).3. Speller. as will be discussed in Section 7.AQUEOUS ENVIRONMENTS 121 Figure 7. p. Chem. 1951. Within the range of about pH 4–10. Altieri. Effect of pH on corrosion of iron in aerated soft water. N. In obtaining the data shown in Fig. Russell. the corrosion rate is independent of pH and depends only on how rapidly oxygen diffuses to the metal surface. Effect of temperature on corrosion of iron in water containing dissolved oxygen. Corrosion.1. Causes and Prevention.6. R. Ind. Whitman. 7.3. [Reprinted with permission from W. American Chemical Society]. sodium hydroxide and hydrochloric acid were used to adjust pH in the alkaline and acid ranges.2. room temperature [12]. New York. Eng. Copyright 1924.2. 16. some investigators. but that is still a relatively low rate. the corrosion rate depends only on the rate of diffusion of oxygen to the available cathodic surface. anodes and cathodes may be several feet apart.9 V. In seawater. the ferrous oxide film is dissolved. using different natural waters and different chemicals to control pH.003–0.5 in water of pH < 10. the surface pH falls. † This equivalence is valid in waters of low conductivity only if anode and cathode areas are in close proximity. Although the rate of formation of FeO− in concentrated alkalies at room temperature is 2 low. and iron is more or less in direct contact with the aqueous environment.† In addition.1 mm/y (0. This behavior has been attributed to reduced buffer capacity of HCO− (an inhibitor) as pH increases. passivity is disrupted.2.0 gmd). In the region of pH 4–10. acting as cathode.1. The corrosion rate correspondingly increases slightly to 0. The fact that Fe2+ is complexed by OH− in strong alkalies to form FeO− with accompanying reduc2 tion in activity of Fe2+ accounts for the observed active potential of iron. to a noble value of 0. 3 causing a pH at local sites that is lower than the pH that would otherwise exist in a solution of saturated hydrous ferrous oxide.4 to −0. The reader may wish to consult the review by Matsushima on this matter [13]. This was established in an experiment of Whitman and Russell [15]. pH < 4. the potential of iron changes from an active value of −0. the observed pH of which is about 9. an increase in alkalinity of the environment raises the pH of the iron surface. If the alkalinity is markedly increased. for example. to 16 N NaOH (43%).05–2.1 V in 1 N NaOH. with an accompanying decrease in the corrosion rate. have observed that the corrosion rate does change with pH in the pH range 6–9. as explained in Section 7. All oxygen reaching the copper surface. In the absence of dissolved oxygen. the reaction proceeds with hydrogen evolution forming sodium hypoferrite. and the potential achieves the very active value of −0. The corrosion rate correspondingly decreases because iron becomes increasingly passive in the presence of alkalies and dissolved oxygen. Confirming the occurrence of passivity by Definition 1 in Section 6.1. Above pH 10. Na2FeO2 [14]. The extent of the cathodic surface is apparently not important. Massachusetts. all oxygen reaching cathodic areas of the iron surface itself produced *On the other hand.122 IRON AND STEEL surface of iron is always in contact with an alkaline solution of saturated hydrous ferrous oxide. The total weight loss of these specimens compared to control specimens not plated was found to be the same. The increased rate of reaction is then the sum of both an appreciable rate of hydrogen evolution and oxygen depolarization. who exposed steel specimens that were copper-plated over three-quarters of the surface to tap water of Cambridge.* Within the acid region. was reduced in accord with the reaction 1 H 2 O + O2 → 2OH − − 2e − 2 and an equivalent amount of Fe2+ was formed at the iron surface acting as anode. caused by marked polarization of probably both anodic and cathodic areas. In this region.5. iron corrodes with formation of soluble sodium ferrite (NaFeO2).2. the rate becomes excessively high at boiler temperatures. . in part. and the corrosion reaction is established. Many laboratory and service data obtained with a variety of irons and steels support the validity of this conclusion [16]. This means that whether a high. or cold-rolled mild steel is exposed to fresh water or seawater. any small variation in composition of a steel and its heat treatment.g.e. the factors governing its corrosion in media of low pH are less important than those in natural waters and soils.. as well as in food cans where fruit and vegetable acids corrode the container with accompanying hydrogen evolution. in turn. depends on the hydrogen overpotential of various impurities or phases present in specific steels or irons. there are certain applications where such factors must be considered—for example. is a phase of low hydrogen overpotential. Because cementite. Since iron is not commonly used in strongly acid environments. Less is known about the effect of impurities and metallurgical factors on the corrosion rate in the very alkaline region (pH ∼14) where corrosion is again accompanied by hydrogen evolution. However. 1–2% Ni. In the passive region (pH ∼10–13).AQUEOUS ENVIRONMENTS 123 an equivalent amount of Fe2+. Thus. Mo. and velocity of the water alone determine the reaction rate. is not expected to be pronounced. . by the rate of hydrogen evolution. mixed control). for example. In the acid range. which is the case within pH 4–10. wrought iron. In general. etc. has no bearing on corrosion behavior. Oxygen concentration. a low-carbon steel corrodes in acids at a lower rate than does a high-carbon steel. low-alloy steel (e. oxygen is not controlling. it follows that so long as oxygen diffusion through the oxide layer is controlling.). The latter. in steam-return lines containing carbonic acid. cast iron. the total amount of iron corroding was the same regardless of whether copper was plated over part of the specimens. heat treatment affecting the presence and size of cementite particles can appreciably alter the corrosion rate. These observations answer the once vociferous argument that wrought iron. The rate becomes sufficiently high in this pH range to make anodic polarization a possible contributing factor (i. Nevertheless. temperature. as well as the effect of metallurgical factors.. Fe3C. Furthermore. all the observed corrosion rates in a given environment are essentially the same. Similarly. any condition producing a large cathode-to-anode area ratio facilitates achievement of passivity and increases the stability of the passive state once it is achieved. Therefore. provided that the diffusion-barrier layer remains essentially unchanged.1. or whether it is cold worked or annealed. the effect on passivity by impurities in their usual concentrations.or low-carbon steel. cold-rolled steel corrodes more rapidly in acids than does an annealed or stress-relieved steel because cold working produces finely dispersed low-overpotential areas originating largely from interstitial nitrogen and carbon. penetration of iron in the case of the plated specimens was four times that of the bare specimens. pH < 4. Mn. A few typical data are summarized in Table 7. These facts are important because almost all natural waters fall within the pH range 4–10. is supposedly more corrosion resistant than steel. which is often the more important reason.10. The sensitivity of the corrosion rate of iron or steel in nonoxidizng acids to dissolved oxygen concentration is shown . Causes and Prevention.0034 0. 2.2. greater accessibility of oxygen to the metal surface on dissolution of the surface oxide favors oxygen depolarization. there is more available H+ to react with and dissolve the barrier oxide film using a weak acid compared to a strong acid. various specimens tempered at 300–800 °C Distilled water Distilled water Distilled water 65 °C 65 °C 65 °C 0. Speller.0036 0. N. Corrosion. dissolution of the oxide occurs at a higher pH.2% Ni Wrought iron Not stated Not stated Not stated Not stated Seawater Seawater Seawater Seawater 0. in strong acids. In weaker acids.11 0.005 0.39 0. the corrosion rate of iron increases accompanied by hydrogen evolution at pH 5 or 6. such as acetic or carbonic acids.0014 0.004 0.124 IRON AND STEEL TABLE 7. New York.06.1.091 0.084 Effect of Carbon 0. 0. 7. in fact.13 0. hence. Corrosion Rates of Various Steels when Oxygen Diffusion Is Controlling % Carbon Treatment Environment Temperature of Test Corrosion Rate ipy mm/y Effect of Heat Treatment 0.39 Cold drawn.1 Corrosion of Iron in Acids.3.036 0.05 0. This difference is explained [12] by the higher total acidity or neutralizing capacity of a partially dissociated acid compared with a totally dissociated acid at a given pH.13 0. 1951.086 0. We have seen that.041 Source: F. In other words.13 0. the diffusion-barrier oxide film on the surface of iron is dissolved below pH 4.0033 0. The increased corrosion rate of iron as pH decreases is not caused by increased hydrogen evolution alone.038 0.10 0.005 0.0015 0. annealed 500 °C Normalized 20 min at 900 °C Quenched 850 °C.005 0.32 Not stated Not stated Not stated 3% NaCl 3% NaCl 3% NaCl Effect of Alloying 0. at a given pH.34% Cu 0.39 0. such as hydrochoric and sulfuric acids.13 Room Room Room 0.0016 0. McGraw-Hill. in which the corrosion rate is higher—inhibit the corrosion reaction.79 0. such as nitric acid. Traces of oxygen in dilute H2SO4—or substantial amounts in more concentrated acid. and an effect of velocity is no longer observed (Fig.0 N H2SO4 at 25 °C was found to be 41. thereby decreasing corrosion.79 0. the inhibiting effect of dissolved oxygen is observed within a critical velocity range. the corrosion rate of steel in the presence of air in 0. For zone-refined iron.1 12. the ratio of corrosion rate with oxygen present to corrosion rate with oxygen absent is 87.2. At rest. This effect is observed.2. . allowing more oxygen to reach the metal surface. Potential and polarization measurements indicate that oxygen in small concentrations at the metal surface increases cathodic polarization. Accordingly.0 gmd [18]. for reasons previously stated (Fig. In addition. Effect of Dissolved Oxygen on Corrosion of Mild Steel in Acids [17] Acid Corrosion Rate (mm/y) Under O2 6% acetic 6% H2SO4 4% HCl 0. therefore. In general. 7.5). In more concentrated acids.2 Ratio by data of Table 7. oxygen acts mainly as a depolarizer. both aerated and deaerated.0043 N H2SO4 at a velocity of 3. In the absence of dissolved oxygen. the ratio is almost unity. This result is expected because hydrogen overpotential (activation polarization) is insensitive to velocity of the electrolyte. the ratios are larger the more dilute the acid.2 9. independent of dissolved oxygen.04% HCl 1.14 40 87 12 15 71 1. for which the corrosion rates are high and the diffusion of oxygen is impeded by rapid hydrogen evolution (Fig. The important depolarizing action of dissolved oxygen suggests that the velocity of an acid should markedly affect the corrosion rate. the ratio of corrosion rates is about 12 [19]. In 6% acetic acid at room temperature. 7.16 0.5 gmd.2% Co–Fe alloy in 1. with the critical velocity becoming higher the more rapid the inherent reaction rate of steel with the acid. whereas in hydrogen-saturated acid the rate was 68.5) [20]. Similar effects are shown by pure 9.AQUEOUS ENVIRONMENTS 125 TABLE 7. the rate of hydrogen evolution is so pronounced that oxygen has difficulty in reaching the metal surface. In oxidizing acids. 7.7 m/s (12 ft/s) is the same as that in 5 N acid at the same velocity.2% HNO3 13. in which diffusion of oxygen is impeded to a lesser extent. which act as depolarizers and for which the corrosion rate is. the average corrosion rate in aerated 1. depolarization in more concentrated acids contributes less to the overall corrosion rate than in dilute acids.9 46 Under H2 0. increasing the rate.9 9. only hydrogen evolution occurs at cathodic areas. Hence. particularly with dilute acids.4) [18]. Relative motion of acid with respect to metal sweeps away hydrogen bubbles and reduces the thickness of the stagnant liquid layer at the metal surface.0 N H2SO4. in higher concentrations. Corrosion of 9.4. 23 ± 2 °C.. 25 °C [18]. Effect of velocity on corrosion of mild steel (0. showing an inhibiting effect of dissolved oxygen. 18 mm diameter. and 0.009% C steel under nitrogen [20].12% C) in 0.5. 45-min test. rotating spec.33 N sulfuric acid under air and oxygen [19].126 IRON AND STEEL Figure 7. Figure 7. 56 mm long.2% Co–Fe alloy in 1 N H2SO4. . is given by p = p0 1 + ( Ac Aa ) (7. the general relation between penetration. therefore. Conductivity of the electrolyte and geometry of the system enter the problem because only that part of the cathode area is effective for which resistance between anode and cathode is not a controlling factor.2. 7. All more noble metals accelerate corrosion similarly. the situation is shown in Fig. the minimum shifts to higher velocities the higher the carbon content of the steel because the corrosion rate and the accompanying hydrogen evolution both increase with carbon content. also showed that the actual penetration of iron increases when iron is coupled to a more noble metal. The corresponding galvanic currents are given by projecting the intersection of anode–cathode polarization curves to the log I axis. it may be several decimeters. on lead) acts as a barrier to diffusion of oxygen or when the metal is a poor catalyst for reduction of oxygen. In soft tap water. the increased corrosion caused by coupling can be considerable. In the case of coupled metals exposed to deaerated solution for which corrosion is accompanied by hydrogen evolution.g. thereby impeding oxygen diffusion to the metal surface. of a metal having area Aa coupled to a more noble metal of area Ac. any metal on which hydrogen discharges acts as a hydrogen electrode with an equilibrium potential at 1 atm hydrogen pressure of −0. except when a surface film (e.AQUEOUS ENVIRONMENTS 127 For aerated acid. Figure 7.7. in seawater. When a corroding metal is coupled to a more noble metal of variable area. Slope 1 represents polarization of a noble metal area having high hydrogen overpotential. The critical distance is greater the larger the potential difference between anode and cathode.4 Effect of Galvanic Coupling The classic experiment of Whitman and Russell. the minimum rate occurs at higher velocities the more concentrated the acid because the rate of hydrogen evolution is more pronounced. the critical distance between copper and iron may be 5 mm. Similarly. increased area of the more noble metal also increases corrosion of the less noble metal. If the anode of area Aa is coupled to the more noble . p (proportional to corrosion rate).13) where p0 is the normal penetration of the uncoupled metal.059pH volt. For the situation where diffusion of a depolarizer is controlling. which showed that weight loss of iron coupled to copper is the same as if the entire surface had been iron. This experiment. 7.6 shows polarization curves for an anode that polarizes little in comparison to a cathode at which hydrogen is evolved (cathodic control).. Slopes 2 and 3 represent metals with lower hydrogen overpotential. If the ratio of areas Ac/Aa is large. where log current density is plotted instead of log total current. provides information about the effect of coupling on the corrosion rate of the less noble component of a couple. In general. . Effect of hydrogen overpotential of cathode on galvanic corrosion in deaerated nonoxidizing acids.7. (a) Large cathode coupled to small anode.128 IRON AND STEEL Figure 7. Figure 7.6. (b) Large anode coupled to small cathode. Effect of anode–cathode area ratio on corrosion of galvanic couples in deaerated nonoxidizing acids. Aa 7. enough oxygen may reach the surface to cause partial passivity. for hydrogen ion I discharge on the noble metal. and the corrosion rate increases without a Figure 7.] . passivity is not established at any velocity.14) where φgalv is the corrosion potential (S.) of the galvanic couple (measured at a large distance compared to dimensions of the couple). Should the velocity increase still further.8). 48-h tests [21]. respectively. Ind. Effect of velocity on corrosion of mild steel tubes containing Cambridge water. At sufficiently high velocities. and galv = iA. 7.059 pH A + log c i0 β Aa (7. Chem.AQUEOUS ENVIRONMENTS 129 metal of area Ac. If this happens. [Reprinted with permission from R. 19. In the presence of high concentration of Cl−.E. and relative motion of the water at first increases the corrosion rate by bringing more oxygen to the surface.8. 65 (1927). the pH is usually too high for hydrogen evolution to play an important role. American Chemical Society. The maximum corrosion rate preceding passivity occurs at a velocity that varies with smoothness of the metal surface and with impurities in the water. the rate decreases again after the initial increase (Fig. 21 °C.galv . the mechanical erosion of passive or corrosion-product films again increases the rate. β and i0 are the Tafel constant and exchange current density. Copyright 1927.galv = −φ galv − 0. Eng.. then the galvanic current density at the anode produced by coupling can be shown to be log iA. Russell et al. as in seawater.H.5 Effect of Velocity on Corrosion in Natural Waters In natural fresh water.2. 7. from increased metal loss in laboratory tests using seawater compared to fresh water. Damage can be reduced by operating rotary pumps at the highest possible head of pressure in order to avoid bubble formation. and on the water-cooled side of dieselengine cylinders [25]. on the trailing faces of propellers and of water-turbine blades. aeration of water cushions the damage caused by collapse of bubbles. or cavitation damage.9). For turbine blades. as is experienced with glass or plastics or when damage to a metal occurs in organic liquids. Cavitation attack is usually controlled by design and materials selection [26]. may involve both chemical and mechanical factors. for example. 27]. To reduce damage of diesel-engine cylinder liners. The surface becomes deeply pitted and appears to be spongy (Fig. If conditions of velocity are such that repetitive low-pressure (below atmospheric) and high-pressure areas are developed. it is used as a facing for water-turbine blades. The part played by chemical factors is apparent.2. bubbles form and collapse at the metal–liquid interface. decrease at any intermediate velocity (Fig.130 IRON AND STEEL Figure 7. high-density polyethylene has a cavitation–erosion resistance similar to that of nickel-based and titanium alloys [26.1 Cavitation–Erosion. There is a wide range of polymers with good resistance to cavitation–erosion and excellent resistance to corrosion—for example. Since 18–8 stainless steel is one of the relatively resistant alloys.5.10). . The damage. Damage from this cause may be purely mechanical. This phenomenon is called cavitation. Cavitation–erosion occurs typically on rotors of pumps. The damage can be reproduced in the laboratory using metal probes undergoing forced high-frequency mechanical oscillations normal to the metal surface. or like a honeycomb-like structure [24]. particularly if protective films are destroyed. The same behavior is expected at elevated temperatures that preclude the possibility of passivity by dissolved oxygen. however. The damage to a metal caused by cavitation is called cavitation–erosion. addition of inhibitor to the cooling water has proved effective [28]. allowing corrosion to proceed at a higher rate. 7.9. Effect of velocity on corrosion of steel in seawater [22]. 7. it is important to understand why the rate first increases.) 7.2. Oxygen solubility in water decreases continuously with sodium chloride concentration.11. Cavitation–erosion damage to cylinder liner of a diesel engine [23]. The initial rise appears to be related to a change in the protective nature of the diffusion-barrier rust film that forms on corroding iron. The corrosion rate first increases with salt concentration and then decreases. 1950. and then decreases. with the value falling below that for distilled water when saturation is reached (26% NaCl). reaching a maximum at about 3% NaCl (seawater concentration).10. Since oxygen depolarization controls the rate throughout the sodium chloride concentration range. 7.6 Effect of Dissolved Salts The effect of sodium chloride concentration on corrosion of iron in air-saturated water at room temperature is shown in Fig. explaining the lower corrosion rates at the higher sodium chloride concentrations. In distilled water having low conductivity. (Copyright NACE International. anodes and cathodes must be located .AQUEOUS ENVIRONMENTS 131 Figure 7. Consequently. Nitrates appear to be slightly less corrosive than chlorides or sulfates at low concentrations (0.g. on the other hand. SrCl2) are slightly less corrosive than alkalimetal salts. This film provides an effective diffusion-barrier film. relatively near each other. which are salts that hydrolyze to form acid solutions. Above 3% NaCl. but not necessarily at higher concentrations [30]. instead. Any Fe(OH)2 so formed does not provide a protective barrier layer on the metal surface. 7. room temperature (composite data of several investigations). Hence. CaCl2. Alkaline-earth salts (e.. etc.132 IRON AND STEEL Figure 7. Acid salts.. the conductivity is greater. KI.25 N). iron corrodes more rapidly in dilute sodium chloride solution because more dissolved oxygen can reach cathodic areas. from their specific effect on the Fe(OH)2 diffusion-barrier layer. LiCl. Chlorides appear to be slightly more corrosive in the order Li. the corrosion rate decreases. Na2SO4.) affect the corrosion rate of iron and steel in approximately the same manner as sodium chloride.11.1] 1 O2 + H 2 O → 2OH − − 2e − 2 (7.3) and (7. additional anodes and cathodes can operate much further removed one from the other. (7. KCl. Na.15) are always in the proximity of Fe2+ ions forming at nearby anodes. In sodium chloride solutions. these substances diffuse into the solution and react to form Fe(OH)2 away from the metal surface. Eqs. NaBr.4). or perhaps from the different adsorptive properties of the ions at a metal surface resulting in differing anode–cathode area ratios or differing overvoltage characteristics for oxygen reduction. Effect of sodium chloride concentration on corrosion of iron in aerated solutions. hence. NaOH does not react immediately with FeCl2 formed at anodes. the continuing decreased solubility of oxygen becomes more important than any change in the diffusion-barrier layer. cause corrosion with combined hydrogen evolution and oxygen depolarization at a rate .2–0. The small differences for all these solutions may arise. OH− ions forming at cathodes in accordance with [cf. hence. for example. Alkali-metal salts (e. Sec. At such cathodes. and K [29].g. resulting in a film of Fe(OH)2 adjacent to and adherent to the metal surface. whereas in Boston water (5 ppm Ca2+. but produce a higher corrosion rate than corresponds to their pH (at NH4Cl concentrations >0. If the concentration of such salts is high. sodium tetraborate (Na2B2O7). Ammonium nitrate in high concentration is more corrosive (as much as eight times more) than the chloride or sulfate. depending on the source and location of the water.3). In addition to favoring passivation of iron by dissolved oxygen.1 Natural-Water Salts. 7. a cold-worked mild steel reacting much more rapidly than one quenched from elevated temperature.6. This indicates that the reaction is not diffusion-controlled. The complex formed in this case has the structure [Fe(NH3)6](NO3)2 [31]. CuCl2. Since coupling mild steel to an equal area of platinum has no effect on the rate. but depends instead on the rate of metal-ion formation at the anode and perhaps also to some extent on the rate of depolarization at the cathode. sodium silicate (Na2SiO3). a galvanized-iron hot-water tank was observed to last 10–20 years before failing by pitting in Chicago Great Lakes water (34 ppm Ca2+. They may. 3 In the presence of excess NH3. thereby reducing activity of Fe2+ and increasing the tendency of iron to corrode. NaNO2. HgCl2. Oxidizing salts are either (a) good depolarizers. otherwise. and sodium carbonate (Na2CO3). in part because of the depolarizing ability of NO− . 7. also inhibit corrosion below pH 10 and may provide better inhibition above pH 10 than NaOH or Na2CO3. For example. NH4Cl) are also acid. Metallurgical structure affects the rate.3. a similar tank lasted only one to two years. . act as corrosion inhibitors. Ammonium salts (e. the reaction is apparently anodically controlled. 157 ppm dissolved solids). Examples of the second class are Na2CrO4. and K2FeO4. They passivate iron in the presence of dissolved oxygen in the same manner as NaOH (Fig. Section 7. For many years before the causes were clearly understood.2. The differences in properties accounting for an oxidizing salt being either a depolarizer or a passivator are discussed in Chapter 17. common to some synthetic fertilizer solutions. on this account.g. Examples of the first class are FeCl3. Increased corrosivity is accounted for by the ability of NH4+ to complex iron ions. Alkaline salts. 43 ppm dissolved solids). Natural fresh waters contain dissolved calcium and magnesium salts in varying concentrations. MnCl2. and sodium hypochlorite. NiSO4. the corrosion rate in ammonium nitrate at room temperature may reach the very high value of 50 mm/y (2 ipy). They represent the most difficult class of chemicals to handle in metal equipment. it was recognized that a soft water was more corrosive than a hard water. Examples of such salts are AlCl3. the water is called hard. they may form corrosion-product layers of ferrous or ferric phosphates in the case of Na3PO4. and therefore corrosive.. with such layers acting as more efficient diffusion barriers than hydrous FeO. which hydrolyze to form solutions of pH > 10. it is called soft. Examples of such salts are trisodium phosphate (Na3PO4). and FeCl2.AQUEOUS ENVIRONMENTS 133 paralleling that of the corresponding acids at the same pH value. or analogous compounds in the case of Na2SiO3. or (b) passivators and efficient inhibitors. KMnO4.05 N) [30].2. pH.134 IRON AND STEEL The two parameters that control corrosivity of soft waters are the pH and the dissolved oxygen concentration. whereas a negative value indicates that it will not form.3). no such protective film of CaCO3 can form. .or oversaturated waters tended to form a protective film of CaCO3 on iron. When pH > pHs.3) and nomographs have been prepared to obtain values of pHs for waters varying widely in composition and at various temperatures [33]. and total dissolved salt concentration. Section 29. and concentration of dissolved solids in the water.2. Using certain simplifications. Since only near. *For a derivation of this and a more exact equation. also known as the Langelier index. Charts (see Appendix. alkalinity. In hard waters. Ks′ is the solubility 3 3 product of calcium carbonate [(Ca 2+ )(CO2−)] . exists at which the water is in equilibrium with solid CaCO3.or oversaturation. But hardness alone is not the only factor that determines whether a protective film is possible.3 in the Appendix. the concentration of calcium ions 3 (Ca2+) is in moles/1000 g H2O. and alk (alkalinity) represents the equivalents per liter of titratable base to the methyl orange end point (often reported as ppm CaCO3) according to the relation − (alk) + (H + ) → 2CO2− + HCO3 + OH − 3 The letter p refers to the negative logarithm of all these quantities. This film retards diffusion of dissolved oxygen to cathodic areas. or Saturation index = pH measured − pH s A positive value of the saturation index indicates a tendency for the protective CaCO3 film to form. an estimate of the corrosivity of a water was established through analytical criteria for under. Langelier showed that the value of pHs—at which a water is in equilibrium with solid CaCO3 (neither tends to dissolve nor precipitate)—can be calculated from the relation* 2+ pH s = (pK 2 − pKs′) + pCa + palk ′ (7. a value of pH. the natural deposition on the metal surface of a thin diffusion-barrier film composed largely of calcium carbonate (CaCO3) protects the underlying metal. For given values of hardness.16) ′ where K 2 is the ionization constant (H + )(CO2−)/(HCO−) . however. supplementing the natural corrosion barrier of Fe(OH)2 mentioned earlier (Section 7. In soft water. the deposition of CaCO3 is thermodynamically possible. see Section 29. Ability of CaCO3 to precipitate on the metal surface also depends on total acidity or alkalinity. The saturation index. is defined as the difference between the measured pH of a water and the equilibrium pHs for CaCO3. Langelier divided natural fresh waters into two groups: those oversaturated with CaCO3 and those undersaturated [32]. given the symbol pHs. 2–5. originally organized by Nordell.1 32.3 44.6–21. 35]. .6 1.6 6.or undersaturation can also be obtained in the laboratory by measuring the pH of a water before and after exposure to pure CaCO3 powder for a time adequate to achieve equilibrium.1 1.7 1.5 1.8 1.0 1. 35] Total Dissolved Solids (ppm) 50–300 400–1000 A 0.4–16.4–50.1 2.1 2.1 B 2.9 10. 3.0 2.3.5 2.2 1. PSI = 2 pH s − pH e pH e = 1.4 1. pHs = (9.9 2. After studying hundreds of concentrated cooling waters. according to Puckorius and Brooke.2–26.3 + A + B) − (C + D).5 2.6 1. Obtain values of A.8–43.7–63. C. for calculating the Langelier saturation index are presented in Table 7. 2.9 2.54 CaCO3 precipitates at PSI < 6. An estimate of over. O.3 2.3 64.3 1.3 1.2 2.6 1.2–36.7 0.5 1. Puckorius and Brooke [36] proposed a new index.7 37.4–71.8 1.7 17.2 2.9 1. B.1 1.0. An increase in pH corre- TABLE 7.7 2.6 0.3 2.7 1.0 Temperature (°C) 0–1.4 1.4 2.1 2.9 1. the practical saturation index (PSI).7–8.AQUEOUS ENVIRONMENTS 135 Tabulated data.1 0.3 [34. and D from this table.3 14.6 M.2–81.2 2.5 2.6 56.4 1.8 1.4 2.4 2.7 27.3 1.7 1.2 1.1–55.0 51.3 2.6 2.1 2. Langelier saturation index = pH–pHs.1 72. Data for Calculating the Langelier Saturation Index [34.0–13.9 3.0 1. Alkalinity (ppm as CaCO3) 10–11 12–13 14–17 18–22 23–27 28–35 36–44 45–55 56–69 70–88 89–110 111–139 140–176 177–220 230–270 280–350 360–440 450–550 560–690 700–880 890–1000 D 1.0 1.5 1.1 21.0 2.2 Ca Hardness (ppm as CaCO3) 10–11 12–13 14–17 18–22 23–27 28–34 35–43 44–55 56–69 70–87 88–110 111–138 139–174 175–220 230–270 280–340 350–430 440–550 560–690 700–870 880–1000 C 0.6 2.8 2.8–31.8 0.2 1.485 × log (total alkalinity) + 4. 136 IRON AND STEEL sponds to undersaturation. Because the composition of the water is altered by any CaCO3 taken up or precipitated during the test, the final pH is not necessarily the same as the calculated pHs. Alternatively, the water after saturation with CaCO3 can be titrated with acid, with the increase in required milliliters of acid compared to that required by the original water being a measure of undersaturation (marble test) [37]. Any fresh water can be placed in one of the following categories: Saturation Index Positive Zero Negative Saturation Level Oversaturated with respect to CaCO3 In equilibrium Undersaturated with respect to CaCO3 Characteristic of Water Protective film of CaCO3 forms Corrosive Chicago water, for example, has an index of 0.2, whereas the value for Boston water is −3.0. A soft water with negative saturation index can usually be treated with lime, sodium carbonate, or sodium hydroxide, in order to raise the saturation index to a positive value, thereby making the water less corrosive. For very soft waters, the required pH may be too high for uses such as tap water unless Ca2+ ions are added simultaneously. A saturation index of +0.1 to about +0.5 is considered to be satisfactory to provide corrosion protection. At higher values of the saturation index, excessive deposition of CaCO3 (scaling) can occur, particularly at elevated temperatures. Waters with a slightly negative Langelier index may deposit CaCO3 because of pH fluctuations. In general, for waters of negative saturation index, the more negative the index, the more corrosive the water. The relationship between the Langelier index and the corrosion rates of water pipes from a field test at a city water works is presented in Fig. 7.12 [38, 39]. At above-room temperatures, the saturation index may become more positive, and, in any case, the rate of CaCO3 deposition is higher, should there be any tendency toward deposition. Hence, corrosion protection at all temperatures without excessive scaling requires adjustment of water composition to a saturation index that is at least constant over the entire operating temperature range. Powell et al. [33, 40] showed that, for any specific alkalinity, there is a corresponding pH value at which the decrease of measured pH with temperature is almost exactly compensated by a decrease in the factor ( pK 2 − pKs′ ) . Under these condi′ tions, the saturation index is nearly constant with temperature, and scaling tends to be the same in hot or cold water (Table 7.4). The amount of possible scaling, of course, depends on the value of the saturation index. The alkalinity and the pH of waters can be adjusted to bring them into the proper composition range using Ca(OH)2, Na2CO3, NaOH, H2SO4, or CO2. AQUEOUS ENVIRONMENTS 137 Figure 7.12. Relationship between the Langelier index and the corrosion rates of water pipes. (Reproduced from Fujii [39] with permission of Japan Association of Corrosion Control.) TABLE 7.4. Alkalinity–pH Limits for Uniform Scale Deposition at Various Temperatures [40] Alkalinity (ppm as CaCO3) 50 100 150 200 pH Measured at Room Temperature 8.10–8.65 8.60–9.20 8.90–9.50 8.90–9.70 Limitations of Saturation Index. If a natural water contains colloidal silica or organic matter (e.g., algae), CaCO3 may precipitate on the colloidal or organic particles instead of on the metal surface. If such is the case, the corrosion rate will remain high even though the saturation index is positive. For waters high in dissolved salts, such as NaCl, or for waters at elevated temperatures, the CaCO3 film may lose its protective character at local areas, resulting in corrosion pitting. Furthermore, if complexing ions are added to a chemically treated water, such as polyphosphates, which retard precipitation of CaCO3, the saturation index may no longer apply as an index of corrosivity. Other than these exceptions, the saturation index is a useful qualitative guide to the relative corrosivity of a fresh water in contact with metals for which the corrosion rate depends on diffusion of dissolved oxygen to the surface, such as iron, copper, brass, and lead. The index does not apply to corrosivity of a water 138 IRON AND STEEL in contact with passive metals that corrode less the higher the surface concentration of oxygen, such as aluminum and the stainless steels. 7.3 METALLURGICAL FACTORS 7.3.1 Varieties of Iron and Steel As discussed in Section 7.2.3, the corrosion rate of iron and steel in natural waters is controlled by diffusion of oxygen to the metal surface. Hence, whether a steel is manufactured by the Bessemer, oxygen furnace, or open-hearth process, as well as whether it is a wrought iron or a cast iron, makes little or no difference to the corrosion rate in natural waters, including seawater [16]. The same statement applies to corrosion in a variety of soils, because factors determining the corrosion rate underground are similar to those of total submersion. In general, therefore, the least expensive steel or iron of a given cross-sectional thickness having the required mechanical properties is the one that should be specified for those environments. On the other hand, in the acid range of pH (approximately <4) and probably also in the extreme alkaline range (>13.5) where impurities play a role in the hydrogen evolution reaction, differences in manufacture affect the corrosion rate. A relatively pure iron corrodes in acids at a much lower rate than an iron or steel high in residual elements, such as carbon, nitrogen, sulfur, and phosphorus. The high nitrogen content of Bessemer steel makes it more sensitive than open-hearth steels to stress-corrosion cracking in hot caustic or nitrate solution. For this reason, open-hearth steel is usually specified for boilers. Cast iron in natural waters or in soils corrodes initially at the expected normal rate, but may provide much longer service life than steel. Other than the factor of considerable metal thickness common to cast structures, this advantage occurs because cast iron, composed of a mixture of ferrite phase (almost pure iron) and graphite flakes, forms corrosion products that, in some soils or waters, cement together the residual graphite flakes. The resulting structure (e.g., water pipe), although corroded completely, may have sufficient remaining strength, despite low ductility, to continue functioning under the required operating pressures and stresses. This type of corrosion is called graphitic corrosion. It occurs only with gray cast iron (or “ductile” cast iron containing spheroidal graphite) but not with white cast iron (cementite + ferrite). Graphitic corrosion can be reproduced in the laboratory over a period of weeks or months by exposing gray cast iron to a very dilute sulfuric acid that is renewed periodically. 7.3.2 Effects of Composition Composition of an iron or steel within the usual commercial limits of carbon and low-alloy steels has no practical effect on the corrosion rate in natural waters or METALLURGIC AL FAC TORS 139 soils (see Table 7.1, Section 7.2.3; Table 10.1, Section 10.3). Only when a steel is alloyed in the proportions of a stainless steel (>12% Cr) or a high-silicon iron or high-nickel iron alloy for which oxygen diffusion no longer controls the rate, is corrosion appreciably reduced. For atmospheric exposures, the situation is changed because the addition of certain elements in small amounts (e.g., 0.1–1% Cr, Cu, or Ni) has a marked effect on the protective quality of naturally formed rust films (see Chapter 9). Although carbon content of a steel has no effect on the corrosion rate in fresh waters, a slight increase in rate (maximum 20%) has been observed in seawater as the carbon content is raised from 0.1 to 0.8% [41]. The cause of this increase is probably related to greater importance of the hydrogen evolution reaction in chloride solution (with complexing of Fe2+ by Cl−) supplementary to oxygen depolarization as the cathodic surface of cementite (Fe3C) increases. In acids, the corrosion rate depends on the composition as well as the structure of the steel and increases with both carbon and nitrogen content. The extent of the increase depends largely on prior heat treatment (see Section 7.3.3) and is greater for cold-worked steels (Fig. 8.2, Section 8.1). Statistical techniques have been used to investigate the effects of minor alloying elements on the corrosion characteristics of commercial carbon and low-alloy steels in 0.1 N H2SO4 at 30 °C [42]. For the particular steels studied, corrosion rates were increased with increasing carbon content, especially in the range 0.5–0.7% C, and by phosphorus. Both alloyed sulfur and phosphorus markedly increase the rate of attack in acids. These elements form compounds that apparently have low hydrogen overpotential; in addition, they tend to decrease anodic polarization so that the corrosion rate of iron is stimulated by these elements at both anodic and cathodic sites. Rates in deaerated citric acid are given in Table 7.5 [43]. In strong acids, the effects of these elements are still more pronounced [44] (see Fig. 7.13 and Table 7.5). Sulfide inclusions have been found to act as initiation sites for pitting corrosion of mild steels in neutral-pH solutions [45, 46]. On the other hand, sulfur content has been found to have no significant effect on corrosion rates in acids of steels containing more than 0.01% Cu [42]. TABLE 7.5. Corrosion Rates of Iron Alloys in Deaerated Citric Acid and in 4% NaCl + HCl at 25 °C [43] 0.1 M Citric Acid pH = 2.06 Pure iron (0.005% C) +0.02% P +0.015% S +0.11% Cu +0.10% Cu + 0.03% P +0.08% Cu + 0.02% S 2.9 gmd 16.5 70.6 4.1 37.6 3.2 4% NaCl + HCl pH = 1 3.0 gmd 100 283 39 60.6 18.6 140 IRON AND STEEL Figure 7.13. Effect of alloyed phosphorus, sulfur, and silicon in iron on corrosion in deaerated 0.1 N HCl, annealed specimens, 25 °C [44]. (Reproduced with permission. Copyright 1965, The Electrochemical Society.) Measurements of hydrogen permeation rates through cathodically polarized mild steels containing elongated (FeMn)S inclusions show that H2S produced at the metal surface by dissolution of inclusions promotes hydrogen entry into the steels. Permeation rates increase with sulfur content of the steel in the range 0.002–0.24% S, but only where the H2S supply can be maintained by dissolution of inclusions [47]. Elongated inclusions, most commonly manganese sulfide inclusions, provide sites for initiation of hydrogen-induced cracking (HIC), also called stepwise cracking (SWC), in steels that are used for high-pressure pipelines transporting oil and natural gas [48, 49]. HIC is discussed further in Section 8.4. Arsenic is present in some steels in small amounts. In quantities up to 0.1%, it increases the corrosion rate in acids (less than sulfur and phosphorus); in large amounts (0.2%), it decreases the rate [44]. Manganese, in amounts normally present, effectively decreases acid corrosion of steel containing small amounts of sulfur. Inclusions of MnS have low electrical conductivity compared to FeS; in addition, manganese reduces the solid solubility of sulfur in iron, thereby probably restoring the anodic polarization of iron which is lowered by the presence of sulfur [50]. Silicon only slightly increases the corrosion rate in dilute hydrochloric acid (Fig. 7.13). Copper alloyed with pure iron to the extent of a few tenths of 1% moderately increases the corrosion rate in acids. However, in the presence of phosphorus or sulfur, which are normal components of commercial steels, copper counteracts METALLURGIC AL FAC TORS 141 the accelerating effect of these elements. Copper-bearing steels, therefore, usually corrode in nonoxidizing acids at lower rates than do copper-free steels [51]. The data of Table 7.5 for several relatively pure iron alloys show that 0.1% Cu reduces corrosion in 4% NaCl + HCl when 0.03% P or 0.02% S is present in the iron, but not for the phosphorus alloy in citric acid. Addition of 0.25% Cu to a low-alloy steel caused reduction of the corrosion rate from 1.1 to 0.8 mm/y in 0.5% acetic acid–5% sodium chloride solution saturated with hydrogen sulfide at 25 °C [52]. These particular relations apply only to the specific compositions and experimental conditions reported; they are not necessarily general. Steels containing a few tenths of 1% of copper are more resistant to the atmosphere, but show no advantage over copper-free steels in natural waters or buried in the soil where oxygen diffusion controls the rate. An amount up to 5% chromium (0.08% C) was reported to decrease weight losses in seawater at the Panama Canal [53] at the end of one year. A sharp increase in rates was observed between 2 and 4 years; after 16 years, the chromium steels lost 22–45% more weight than did 0.24% C steel. Depth of pitting was less for the chromium steels after one year, but comparable to pit depth in carbon steel after 16 years. Hence, for long exposures to seawater, low-chromium steels apparently offer no advantage over carbon steel. By comparison, however, low-alloy chromium steels (<5% Cr) have improved resistance to corrosion fatigue in oil-well brines free of hydrogen sulfide. An amount up to 5% nickel (0.1% C) did not appreciably change weight losses of steels exposed to seawater in tests at the Panama Canal [53] extending up to 16 years. Depth of pits, although less for one year of exposure, was greater for long-exposure times for the nickel steels compared to 0.24% C steel (77% deeper for 5% Ni steel after 8 years). Low-nickel steels (<5% Ni) have improved resistance to corrosion fatigue in oil-well brines containing hydrogen sulfide [54]. Nickel also decreases the corrosion rate of steels exposed to alkalies, and the effect increases with nickel content [55]. 7.3.2.1 Galvanic Effects through Coupling of Different Steels. Unimportant though low-alloy components may be in determining overall corrosion rates in waters or in soils, composition is nevertheless of considerable importance to the galvanic relations and consequent corrosion of steels coupled to each other. For example, because of increased anodic polarization, a low-nickel, lowchromium steel is cathodic to mild steel in most natural environments. The reason is obvious from relations shown in Fig. 7.14. Both mild steel and a lowalloy steel, not coupled, corrode at about the same rate, icorr, established by the limiting rate of oxygen reduction. When coupled, however, the potentials of both steels, which are originally different, become equal to φgalv. The corrosion rate of mild steel, estimated from the extension of its anodic polarization curve, is now increased to i2, and that of the low-alloy steel is decreased to i1. The precise value of φgalv depends on the relative areas of the two steels. Accordingly, steel bolts and nuts used to couple underground mild-steel pipes, or a weld rod used for steel plates on the hull of a ship, should always be of a low-nickel, low-chromium 142 IRON AND STEEL Figure 7.14. Polarization diagram showing effect of coupling of a low-alloy steel to mild steel on subsequent corrosion rates. steel or similar composition that is cathodic to the major area of the structure (small cathode, large anode). Should the reverse polarity occur, serious corrosion damage would be caused quickly either to the bolts or to the critical area of weld metal [56, 57]. Cast iron is initially anodic to low-alloy steels and not far different in potential from mild steel. As cast iron corrodes, however, especially if graphitic corrosion takes place, exposed graphite on the surface shifts the potential in the noble direction. After some time, therefore, depending on the environment, cast iron may achieve a potential cathodic to both low-alloy steels and mild steel. This behavior is important in designing valves, for example. The trim of valve seats must maintain dimensional accuracy and be free of pits; consequently, the trim must always be chosen cathodic to the valve body making up the major internal area of the valve. For this reason, valve bodies of steel are often preferred to cast iron for aqueous media of high electrical conductivity. 7.3.3 Effect of Heat Treatment A carbon steel quenched from high temperatures has a structure called martensite—a supersaturated solution of carbon in iron—a single metastable phase with carbon in solid solution in interstitial positions of the body-centered tetragonal lattice of iron atoms. Random distribution of carbon atoms accompanied by electronic interaction of carbon atoms with neighboring iron atoms limits their effectiveness as cathodes of local-action cells; consequently, in dilute acid the STEEL REINFORCEMENTS IN CONCRETE 143 corrosion rate of martensite is relatively low. Interstitial carbon reacts in large part with acid to form a hydrocarbon gas mixture (accounting for the odor of pickled steel) and some residual amorphous carbon, which is observed as a black smut on the steel surface. Heating martensite to low temperatures (below 727 °C) and then air-cooling (called tempering) results in the formation of tempered martensite, which consists of α-iron and many dispersed particles of cementite (i.e., iron carbide, Fe3C). This two-phase structure of α-iron and cementite sets up galvanic cells that accelerate the corrosion reaction. The amount of finely divided cementite that forms depends on the tempering time and temperature; for a 2-h heat treatment of 0.95% C steel, the maximum cementite forms at about 400 °C (300 °C for 0.07% C steel). After tempering at this temperature, cementite, acting as cathode, offers maximum peripheral surface adjoining ferrite, and galvanic action during subsequent exposure to an aqueous environment is at a maximum. Above this temperature, cementite coalesces to larger particle size, and the subsequent corrosion rate is lower. The particles of cementite are now large enough to resist complete dissolution in acid and can be detected in the residue of corrosion products. At the same time, there is a corresponding decrease in hydrocarbon gas formation. On slowly cooling a carbon steel from the austenite (face-centered cubic lattice) region, above 727 °C, cementite, in part, assumes a lamellar shape, forming a structure called pearlite. This structure again corrodes at a comparatively low rate because of the relatively massive form of cementite formed by decomposition of austenite compared with smaller-size cementite particles resulting from decomposition of martensite. Corrosion rate increases as the size of iron carbide particles decreases. Pearlitic structures corrode faster than spheroidized ones, and steels containing fine pearlite corrode more rapidly than those with coarse pearlite [42]. The importance of both the amount of cementite acting as cathode and its state of subdivision supports the electrochemical mechanism of corrosion. The rate of corrosion is cathodically controlled and, hence, depends on hydrogen overpotential and interfacial area of cathodic sites. In practice, an effect of heat treatment on corrosion is seldom observed, because oxygen diffusion controls the rate in the usual environments. However, in handling acid oil-well brines, marked localized corrosion is sometimes found near welds or “up-set” ends of steel oil-well tubing. This observed increase in corrosion encircling a limited inner region of the tube is called ring-worm corrosion. It is caused by heat treatment incidental to joining and fabrication techniques, and it can be minimized by a final heat treatment of the tubing or by addition of inhibitors to the brine [58]. 7.4 STEEL REINFORCEMENTS IN CONCRETE Steel-reinforced concrete is an industrially very important system in which steel is exposed to an alkaline (pH > 12) environment and, hence, is normally passive. 144 IRON AND STEEL Nevertheless, corrosion of steel reinforcements does occur in structures such as concrete bridge decks and parking garages. These problems have been studied for many years and result from the breakdown of passivity caused, for example, by salt in the environment or in the concrete [59]. Passivity may also be lost after several years if air diffuses through the concrete to the reinforcements, converting alkaline Ca(OH)2 to less alkaline CaCO3. Corrosion processes for steel in concrete are illustrated in Fig. 7.15 [60]. There are several methods that can be used to control corrosion of steel reinforcements in concrete. First, the design of the structure should provide for drainage of salt-containing waters away from the reinforced concrete. Second, concrete of adequate thickness, high quality, and low permeability should be specified to protect the reinforcements from the environment. Third, chloride content of the concrete mix should be kept to a minimum. For further protection, the steel reinforcements can be epoxy-coated. In many parts of North America, steel reinforcements used in bridge decks are now epoxy-coated as a standard construction procedure. Cathodic protection is also being used, both with impressed current anodes and with sacrificial anodes [61]. (See Chapter 13.) Steel reinforcements in concrete, being passive, are noble in potential with respect to steel outside the concrete that is galvanically coupled to the reinforcements. The measured potential difference is in the order of 0.5 V [62]. The effect of large cathode area and small anode area has caused premature failures of buried steel pipe entering a concrete building [63]. In this situation, use of epoxycoated reinforcements (to coat the cathode) or insulated couplings should prove beneficial. Figure 7.15. Corrosion process for steel in concrete [59]. (Reprinted with permission of John Wiley & Sons, Inc.) REFERENCES 145 There is evidence that high-strength steel reinforcements, such as rod or bar, heat-treated to high hardness (Rockwell C > 40) and subjected to a tensile stress, have failed by SCC or hydrogen cracking (see Chapter 8) through exposure to moisture penetrating the concrete coating. Heat-treating processes sometimes produce a decarburized (hence softer) surface layer. In this situation, SCC may be delayed until the outer metal layer undergoes general corrosion, exposing the SCC-susceptible harder core underneath. Either moisture must be excluded from contact with such steels, or a steel of lower strength should be specified. REFERENCES 1. I. Matsushima, in Uhlig’s Corrosion Handbook, 2nd edition, R. W. Revie, editor, Wiley, New York, 2000, p. 529. 2. H. Uhlig, D. Triadis, and M. Stern, J. Electrochem. Soc. 102, 59 (1955). 3. H. Uhlig, J. Electrochem. Soc. 108, 327 (1961). 4. B. J. Little and J. S. Lee, Microbiologically influenced Corrosion, Wiley Series in Corrosion, R. W. Revie, series editor, Wiley, New York, 2007, p. 1. 5. J.-D. Gu, T. E. Ford, and R. Mitchell, in Uhlig’s Corrosion Handbook, 2nd edition, edited by R. W. Revie, Wiley, New York, 2000, pp. 349–365, 915–927. 6. Ref. 4, p. 41. 7. C. von Wolzogen Kühr, Water and Gas 7, 26, 277 (1923); Corrosion 17, 293t (1961). 8. Ref. 4, pp. 28–30. 9. R. Schweisfurth and E. Heitz, Mikrobiologische Materialzerstoerung und Materialschutz, DECHEMA, Frankfurt am Main, Germany, 1989. 10. Ref. 5, pp. 359–360. 11. G. Skaperdas and H. Uhlig, Ind. Eng. Chem. 34, 748 (1942). 12. W. Whitman, R. Russell, and V. Altieri, Ind. Eng. Chem. 16, 665 (1924). 13. I. Matsushima, Ref. 1, pp. 536–538. 14. R. Scholder, Angew. Chem. 49, 255 (1936). 15. W. Whitman and R. Russell, Ind. Eng. Chem. 16, 276 (1924). 16. F. LaQue, Am. Soc. Testing Mater. Proc. 51, 495 (1951). 17. W. Whitman and R. Russell, Ind. Eng. Chem. 17, 348 (1925). 18. A. P. Bond, Sc.D. thesis, Department of Metallurgy, M.I.T., Cambridge, MA, 1958. 19. W. Whitman, R. Russell, C. Welling, and J. Cochrane, Ind. Eng. Chem. 15, 672 (1923). 20. Z. Foroulis and H. Uhlig, J. Electrochem. Soc. 111, 13 (1964). 21. R. Russell et al., Ind. Eng. Chem. 19, 65 (1927). 22. F. LaQue, in Corrosion Handbook, H. H. Uhlig, editor, Wiley, New York, 1948, p. 391. 23. F. Speller and F. LaQue, Corrosion 6, 209 (1950). 24. J. Postlethwaite and S. Nesic, in Uhlig’s Corrosion Handbook, 2nd edition, R. W. Revie, editor, Wiley, New York, 2000, p. 261. 25. Ibid., p. 262. R. 34. Mater. 205–220. L. Bacon. 54. R. Water Works Assoc. 85. Wiley. 38. CORROSION/99. Electrochem. Corrosion 47. 31. 28. 808 (1946). and A. Oxford. Powell. 1936. Uhlig. 40. J. D. C. Santyr. H. Southwell and A. Corrosion Resistance of Metals and Alloys. New York.. 287. Copson. J. 58. 29. 45. Malik. Corros. Y. Corrosion 2. 537 (1980). and J. Langelier. O. Hirzel. Greene. Brigham. Jarman. Lill. Williams and M. editors. Gainer and G. L. 57. Borgmann. NACE International. R. C. Phelps. 585 (1969). G. Bosei Kanri (Rust Prevention and Control) 27. pp. 270 (1912). New York. Q. 29. 278 (1946). Eng.. C. Mech. 1963. 842 (1945). 240. National Association of Corrosion Engineers. 44. Vol. Am. Bauer. 1. 38. N. 42. 813 (1938). J. 3rd edition. Soc. Electrochem. 112. Basic Eng. 1. Can. 36. Angell. H. Donepudi. 348 (1980). 267–269. Y. 39. Am. Water Treatment for Industrial and Other Uses. 9 (1965). W. R. 2nd edition. 27. 51. in Corrosion. Corrosion 3. R. J. F. E. Kröhnke. L. Hoover. 37. Parkins. 55. p. Corrosion 36. Sci. E. 102. Puckorius and J. S. 1177 (1965). Corrosion 21. Powell. Nordell. 30. 30. 156 (1982). Leith and A. 143. Prot. 663 (1955). U. J. Kim. TX. Manuel. Lill. 60. J. Reinhold. 9. 56. Am Soc. 311 (1988). Revie. 2nd International Congress on Metallic Corrosion. Ref. Trans. and E. M. 1966. Yoshino. and R. p. B. 49. and J. Revie. A. Ind. J. E. 47. Corrosion 36. 280 (1991). D. E. 1500 (1936). Paper No. O. Chem. W. Die Korrosion metallischer Werkstoffe. editor. M. F. Brooke. December. P. 50. Sci. 415 (1947). Corrosion 38. Ref. Eng. in Uhlig’s Corrosion Handbook. 1211 (1964). S. p. Alexander. T. and G. 48. 52. M. P. Chappell. 334. 41. 534. Masing. 1. McLean. B. Eng. H. 2000. pp. Fujii. Z. S. 7. R. T. Wescott. 14 (1970). V. Wilde. 8:197–8:207. Houston. Corros. Ind. Pumphrey. pp. Holmberg. Am. S. Wranglen. 814 (1937). J. p. Mech. Leipzig. B. C. M. Cleary and N. P.. Houston. W. 35. Electrochem. C. S. 625 (1980). Soc. Elboujdaini. 1961. LaQue and H.K. Eng. 53. p. Bacon. J. Butterworth-Heinemann. 821 (1967). H. Schick and H. 1999. G. Water Works Assoc. 27(4). 9. 43. Metall. p. 535. Komp. H. 111. Shreir. R. Uhlig. Iron Steel Inst.-Z. M. B. Burstein. Reinhold. Wallwork.146 IRON AND STEEL 26. editors. 46. Uusitalo in Proc. Foroulis and H. Chem. 33. R. 28. 1994. . Garner. 37. and G. Corrosion 36. L. 2nd edition. Soc. W. Thompson. Water Works Assoc. J. 795 (1960). Texas. Stern. 85 (1983). L. 1802 (1938). Wang. 32. New York. 812. Ibid. pp.-D. . Wiley. and David Whiting. p. R. J. 2nd edition. pp. Ahluwalia. editor. Revie. 2007. Microbiologically influenced Corrosion. Nmai. Vrable.2 (3. Neal S. Corrosion of Steel in Concrete: Understanding. editor. Report for NACE International. and R. Hoboken. 60. J. 2nd edition. Cambridge. Microbiologically Influenced Corrosion Handbook. 2nd edition. in ASM Handbook. Kearns and B. West Conshohocken. PA. Ford. Taylor & Francis. W. P. PROBLEMS 1. E. 61. Gu. Little. P. Mater. J. NJ. pp. Corrosion of ferrous metals. ASTM. New York. New York. W. Little and J.0065 ipy (0. in Uhlig’s Corrosion Handbook. Houston. 1994. 515–600. conducting solution in which uncoupled iron corrodes at 0. Ibid. 62. Revie. Wiley Series In Corrosion. 373–376. R. W. series editor. pp. Berke. series editor. Mater. Texas. R. 581–600. Wiley. 1–53. in Uhlig’s Corrosion Handbook. Microbiologically Influenced Corrosion Testing. Microbial degradation of materials: General processes. R. 2007. W. Mitchell. 2005. 63. Revie. Matsushima.org/nace/images_ndx_story/ highwaybridgegapanalysis. Highway Bridges Gap Analysis. J. pp. 1996. Perf. August 14. Five iron rivets. Corrosion inspection and MonitorIng. Wiley. Uhlig’s Corrosion Handbook. pp. editor. STP 1232. 0.PROBLEMS 147 59. S. New York. private communication. Investigation and Repair. The sheet is immersed in an aerated. 2007. 46(1). R. 2000. GENERAL REFERENCES H. Wiley. 13B: Corrosion: Materials. Broomfield. ASM International. Editors. in Uhlig’s Corrosion Handbook. Wiley. R. PA. B. J.pdf. Revie. 51 (1982). Revie. editors. London. stirred. Philadelphia. Broomfield. W. W. 2000. New York. Lee. J.165 mm/y). OH. T.. Charles K. J. Wiley Series In Corrosion. R. Materials Park. R. Woodhead Publishing. 21(3). P. Corrosion of steel in concrete. STP 1276. S. I. 582. are inserted in a copper sheet of 8 ft2 (7430 cm2) exposed area. Revie. 26 (2007). England. ASTM.5 in. 349–365. J. 2nd edition. 2007. Techniques to Assess the Corrosion Activity of Steel Reinforced Concrete Structures. Roberge. 2nd edition. J. Wiley. Vol. editor. Broomfield. Borenstein. http://www. W. P. 2000. Broomfield.nace. 2000.2 cm2) of total exposed area each. Perf. 1994. 581–600. New York. Edward Escalante. 1 atm). MI −2.3 gmd. 29. Slower corroding steel is anode. with the same dimensions? 2. Two steels immersed in dilute deaerated sulfuric acid corrode at different rates.4. Water entering a steel pipeline at the rate of 40 liters/min contains 5.50 mL O2 per liter (25 °C. (7.2 2.16) in Section 7.0 7.2. (b) 25 °C (77 °F) using chart Fig.3 2. When they are arranged as a cell.5 .90 8.0 ipy (77 mm/y). a small potential difference is observed between them. Assuming that all corrosion is concentrated at a heated section 30 m2 in area forming Fe2O3.2 −0. Water 4.1 Boston.3. (Note that the calcium concentration is listed in the chart as ppm CaCO3.6 −0.7 Total Solids (ppm) 43 157 153 Answers to Problems 1. IL Detroit. Calculate the saturation index of each of the following waters at: (a) 25 °C (77 °F) using Eq.2 0. 4.7 0. what is the corrosion rate in gmd? 3.148 IRON AND STEEL (a) What is the corrosion rate of the rivets in ipy (mm/y)? (b) What is the corrosion rate of an iron sheet in which five copper rivets are placed. (c) 75 °C (167 °F) using charts. IL Detroit. 3. 2. Water leaving the pipe contains 0. Saturation Index (Using Charts) 25 °C 75 °C −2. MA Chicago. 31.6. Which is anode? Illustrate by a polarization diagram. MI Alkalinity (ppm CaCO3) 13 120 73 Ca (ppm) 5 34 26 pK 2 − pK S ′ ′ 2.2 pH 6. Section 29.15 mL O2 per liter.1.) Water Boston. (a) 3. MA Chicago. Many authors have traditionally ascribed this effect to residual stress within the metal. 8. however.1 [1]. The observed increase of rate is apparently caused instead by segregation of carbon or nitrogen atoms at imperfection sites produced by plastic deformation rather than by the presence of the imperfections themselves (Fig. Corrosion and Corrosion Control. hence. cold working increases the corrosion rate severalfold. which serves to increase the corrosion tendency.8 EFFECT OF STRESS 8. a cold-worked commercial steel corrodes in natural waters at the same rate as an annealed steel. this intuitive concept is probably not correct [2]. In acids. But the residual energy produced by cold working. as shown in Fig. 8. with the latter either increasing or decreasing corrosion depending on the particular crystal faces that lie parallel to the metal surface. 149 . Such sites exhibit lower hydrogen overpotential than does either cementite or iron [1] and probably constitute the most important factor. Inc.1 COLD WORKING As discussed in Section 7. (2) the increased surface area of cementite lamellae fractured by the coldwork process. by R. as measured in a calorimeter (usually <30 J/g). and (3) the preferred orientation of ferrite grains. Entering to a lesser extent are (1) increased surface area of metal exposed at slip steps. is less than sufficient to account for an appreciable change in free energy.2).2. Winston Revie and Herbert H. Uhlig Copyright © 2008 John Wiley & Sons. ) Subsequent heat treatment of cold-worked steel at 100°C induces additional diffusion of interstitial carbon atoms to imperfections in the metal lattice. unlike mild steel.C. S. A severely cold-worked. or .2 STRESS-CORROSION CRACKING Stress-corrosion cracking (S. 8.1 and 8. Hence. The Electrochemical Society. accompanied by precipitation of carbides or nitrides of particle size that increases with heat treatment temperature. on the other hand.076% C steel (85% reduction of thickness) and zone-refined iron (50% reduction of thickness) on corrosion in deaerated 0.C. (Reproduced with permission. interstitial impurities does not occur.C. deaerated hydrochloric acid at no higher rate than when annealed (Figs. leading to a corresponding reduction of corrosion rate. When commercial nickel is severely cold-worked. corrodes in dilute. Copyright 1964.1N HCl.2) [1].. that is. heat treatment above 100°C results in cathodic sites of higher hydrogen overpotential and of decreasing peripheral area.C.C. thereby increasing the cathodic area of low hydrogen overvoltage and also accelerating the corrosion rate. lacking sufficient carbon or nitrogen.C. 25°C [1]. Effect of heat treatment of cold-worked 0. 8. the usual ductility of the material (e. The tensile stress can be applied or residual.1.g. failures appear to be brittle. Heat treatment at higher temperatures reduces the density of the imperfection sites. is simultaneously stressed in tension and exposed to an environment that causes S. On a macroscopic level. which suggests that similar formation of low-overpotential cathodic sites derived from segregated.C. it does not corrode appreciably more rapidly in acids [2]. when stressed in air) is considerably reduced.) is a type of cracking that occurs when a material that is susceptible to S. zone-refined (pure) iron.150 EFFECT OF STRESS Figure 8.C. or subsequently annealed.C. Generation of the environment that causes S. [3] 2. Copyright 1964.g. such as deformation (e.C. with the result that there is a multiplicity of possible S. occurs. mechanical. or a combination of the two. Stress-corrosion cracks can be intergranular or transgranular.C. models. Table 8.1N HCl. control measures that can be implemented.1 lists some alloy–environment combinations in which S.C. until failure occurs Depending on the metal–environment combination and the stressing condition.C. theories. Residual stresses result from fabrication processes.C. on corrosion in deaerated 0.STRESS-CORROSION CR ACKING 151 Figure 8. For this reason.C.2. 8. forming of a pipe) and welding.C.2.C.. there are three stages in the stress-corrosion cracking process: 1. These matters will be summarized in this section. Propagation of S. (Reproduced with permission.C.) both. Effect of carbon in steels cold-rolled 50%. Initiation of S. In general. the time to failure can vary from minutes to many years. has led to a very large number of hypotheses. The Electrochemical Society.C. 25°C [1]. occurs.C. 3. Stress-corrosion cracking is caused by the interaction of metallurgical. and environmental factors.C. and controversies on the mechanisms by which S.1 Iron and Steel When mild steel is stressed in tension to levels near or beyond the elastic limit and is exposed simultaneously to hot concentrated alkaline or hot concentrated . In addition.C. the complexity of S. inspection of stressed metals that are exposed to a corrosive environment during service is essential to establish whether cracks have initiated and to develop mitigation procedures before failure occurs. I− H2O. S. nitrates added in amounts equivalent to 20–40% of NaOH alkalinity act as inhibitors of S. 0.C.C. of boilers has become less common.1. Stress-corrosion cracking of steel was first encountered in a practical way in riveted steam boilers. whereas at lower temperatures silicates may act as inhibitors [4]. 1% V Aluminum alloys Zirconium alloys Magnesium alloys <40 at.2– 1. 1% Mo. until the concentration of alkali suffices to induce S.% Au–Cu a Environment NO− .C. [4] and are used for this purpose in practice. for example. OH− 3 H2O Cl−. Br− NH3. Its occurrence has not been eliminated entirely. This is one of several illustrations that can be cited regarding the necessity to understand thoroughly the fundamentals of a laboratory test before extrapolating results to large-scale applications.C..C. and boiler waters are normally treated with alkalies to minimize corrosion.152 EFFECT OF STRESS TABLE 8.0% PbO.C.C. Some Alloy–Environment S.000 psi (>1240 MPa) yield strength or approximately >Rockwell C 40 hardness value. however.C.C.C. On the other hand. or NaNO3 are also accelerators [4]. OH−. amines Cl−. addition of 2% NaOH to boiling nitrate solutions may inhibit cracking [5]. aqua regia >180. NaCl solutions Cl− Cl− FeCl3.. Stresses at rivets always exceed the elastic limit.C. The mechanism is quite different from that described in Section 8. nitrate solutions [such as boiling 60% Ca(NO3)2 + 3% NH4NO3]. may be established at welded sections of boilers or in tanks used for storing concentrated alkalies. austenitic α-Brass Titanium alloys 8% Al. At the higher temperatures and pressures common to boiler conditions. 7] industries in handling hot nitrate solutions. With the advent of welded boilers and with improved boiler-water treatment. sometimes accompanied by explosion of the boiler. Br−. Laboratory tests in boiling 50% NaOH have shown that at atmospheric pressure. Because alkalies were recognized as one of the causes. it cracks along intergranular paths by S.1. failures of this kind were first called caustic embrittlement.C. Crevices between rivets and boiler plate allow boiler water to concentrate. stressed steel may also fail in contact with . Systems Metal Mild steel High-strength steelsa Stainless steels. In addition to failures in the chemical and nuclear-chemical [6. KMnO4. The presence of silicates in alkaline boiler water above 225°C accelerates S. Na2CrO4. because significant stresses. Includes high-strength martensitic and precipitation-hardening stainless steels. than is oil-tempered wire having equal mechanical properties.C. on further heating (in the unstressed state) at 445°C (830°F) for 70 h or at 550°C (1020°F) for 3 h. bridge after 12 years in service [8].3). Ohio. stress-relieved in the range of 400–650°C (750–1200°F) and subsequently stressed.C. 8. presumably containing trace amounts of ammonium nitrate from the atmosphere.C. There was evidence in this case that the steel cable was unusually susceptible to S. Explosions have been reported involving cracks in steel tanks containing compressed illuminating gas. of pipeline steels have been reported [13.01N solutions of NaCl. 0. had accumulated and concentrated. (NH4)2SO4. which acts as an inhibitor [9]. Heat treatment of cold-rolled mild steel at 600°C (1110°F) for 1/2 h.C.C. however. Therefore.C. 14]. or at lower temperatures for correspondingly longer times induces susceptibility again. Along these lines. . by avoiding air contamination of the NH3. but can be made resistant by tempering 1 at 250°C (480°F) for 2 h (Fig. it was found that severely cold-worked mild steel (0. or by addition of about 0. Cathodic polarization in the latter solution prevented damage. Based on tests in boiling nitrate solution [15].C. in 0. it is recognized in practice that cold-drawn steel wire is more resistant to S.C.STRESS-CORROSION CR ACKING 153 nitrates at room temperature.06% C. or in 0. The problem was solved by removing traces of HCN and moisture from the gas. Stress-corrosion cracking of steel also occurs when the metal is in contact with anhydrous liquid ammonia at room temperature—for example. plastically deformed steel. Cracking can be avoided by stress-relief heat treatment of the steel. Subsequent tests conducted at the U. Mild steel quenched at 900–950°C (1650–1740°F) is susceptible. The failures at stresses below the elastic limit were transgranular and were traced to small amounts of HCN contained in the gas [12]. Both transgranular and intergranular S. Transgranular S. but they can also be transgranular. This resistant state. at coldformed heads and at welds of steel tanks used to contain the liquefied gas.C.2% H2O.7% C steel cables of the Portsmouth. Intergranular cracking of stressed steel has also been reported in contact with SbCl3 + HCl + AlCl3 in a hydrocarbon solvent [10].3).2% C steel to an aqueous mixture of carbon dioxide and carbon monoxide at 100 psi [11]. for example.C. Underground steel pipelines that transmit oil and gas and that operate at high internal pressures have been reported to fail by S. was reported at room temperature on exposure of 0. is temporary. No cracking occurred after similar tests in distilled water.C. or NaOH. Such failure occurred. the steel becomes susceptible again and remains susceptible.C. NaNO2.001% N) is resistant to S. at 445°C (830°F) for 48 h. (Fig. or at 200°C (390°F) for 48 h. as well as for correspondingly shorter times at higher temperatures. Cracks are intergranular.S. even if the steel exposed to nitrates is highly stressed after heat treatment.1– 0. The cables cracked at their base where rain water.01N NH4NO3 or NaNO3 at room temperature. National Bureau of Standards (now the National Institute for Standards and Technology) showed that specimens of the 1 1 cable stressed in tension failed within 3 2 –9 2 months when immersed in 0. is more rather than less susceptible as a result of heat treatment. 8. Effect of heat treatment of mild steel after quenching or cold rolling (70% reduction of thickness) on resistance to stress-corrosion cracking in boiling nitrate solution [15]. Pure iron is immune to S. Iron (>0.002% C) [16] or mild steel (0. 70 h at 445°C) allows slowmoving lattice imperfections within the grain to migrate to the grain boundaries.g.. produces immunity by destroying continuous grain-boundary paths. accounting for increased resistance or immunity.3..C. by generating imperfections that have high affinity for carbon and that deplete any continuous paths along which carbon atoms tend to segregate. and. of steel in nitrate solutions include the following: .org. (Reprinted with permission of ASM International®.154 EFFECT OF STRESS Figure 8.asminterna tional. transporting carbon atoms with them.C.06% C) quenched from about 925°C contains a sufficient concentration of carbon atoms along the grain boundaries to induce susceptibility. 250°C for 2 h) randomly nucleates iron carbides.C. Low1 temperature heating (e.g. more important. Longer heating at higher temperatures (e. and the steel again becomes susceptible. Cold working.) The overall evidence suggests that grain boundaries become suitable crack paths only when carbon (or nitrogen) atoms (not Fe3C) segregate at grainboundary regions. Remedial measures to avoid S. All rights reserved. on the other hand. which rob the grain boundaries of carbon. www.C. In accelerated tests. mechanical. or quenched from this temperature range and then tempered at about 250°C 1 (480°F) for 2 h. the addition of 3% NaCl or 2% sodium acetate to boiling 60% Ca(NO3)2 + 3% NH4NO3. Bohnenkamp [18] reported maximum susceptibility of various carbon steels (0. on metallurgical. 3. is resistant.3).C.66 to −0. Parkins [17] found similar protection in hot nitrate solution.75 V (S. control are presented in Table 8.C.1 V either noble or active to this range. Sodium nitrate has already been mentioned as a practical inhibitor for steel exposed to boiler waters.3. Both anodic and cathodic protection were found to be effective (see Fig.) with orders-of-magnitude longer life at potentials at 0. or at higher temperatures for shorter times.2 [22]. and are not dissolved by corrosion.C.2% C are resistant [20]. Inhibitors. exhibit improved resistance to S. Heat Treatment. Compressive stresses are produced at the surface of the metal that are effective in avoiding damage so long as the compressive layers are continuous.003–0. . Cathodic Protection. This resistance is temporary and short-lived (200 h) at temperatures on the order of 400°C (750°F).C. or niobium plus tantalum [19]. which react preferentially with carbon and nitrogen.11% C) in 33% NaOH boiling at 120°C at −0. 8.C. at 108°C was found to inhibit cracking (>200 h) [21] of mild steel. Severe Cold Working. and environmental variables results in three main avenues for S. Buffer ions. remain intact. titanium.H. 2.g.. but it is predicted to reach the order of thousands of hours at 300°C (570°F) or below [15]. 4.C. Crude quebracho extract and waste sulfite liquor are also used. Schroeder and Berk [4] found that cathodic polarization of steel stressed in hot sodium hydroxide–sodium silicate solution greatly delayed or prevented cracking. Steels containing small alloying additions of aluminum.C. Cold rolling to >50% reduction of thickness imparts relative immunity to a stressed mild steel in boiling nitrate solution. Alloyed additions of <2% Ni increase susceptibility of low-carbon steels in nitrates. >1% Cr or Mo decrease susceptibility. for thousands of hours.. 6. The dependence of S. Surface Peening or Shot Blasting. Mechanical: Reduce the applied stress and/or stress-relief anneal to reduce the residual stress. 100–200°C (200–400°F)].C.7. The various approaches to S. 5.E. but are not immune.C. Section 8.C. Special Alloys.STRESS-CORROSION CR ACKING 155 1. control: Metallurgical: Use an alloy that is resistant to S. Mild steel slowly cooled at 900–950°C (1650–1740°F). Furnace-cooled (pearlitic) steels of >0. are useful 4 because they avoid high concentrations of OH− in concentrated boiler water. such as PO3− . The resistant state is predicted to persist at low temperatures [e. are susceptible. sensitized at 650°C for 1 h) fails intergranularly in a wide variety of electrolytes regardless of whether it is stressed. 2. but the opposite situation applies to mild steel.C.. Fe2(SO4)3. depending on the system). Pure alloys. S.3 MECHANISMS OF STRESS-CORROSION CRACKING OF STEEL AND OTHER METALS The distinguishing characteristic of S. Chloride ions.1 in Section 8.C. it was understood that pure metals did not undergo S. . of pure copper under aggressive laboratory conditions has been reported [25. For many years. however. occurs in the presence of NH3 or amines.C. 8. The much more rapid S. Typical specific environments for several metals are listed in Table 8.C.156 EFFECT OF STRESS TABLE 8. Stress-Corrosion Cracking Control Measures [22] Mechanical Avoid stress concentrators Relieve residual stresses Introduce surface compressive stresses Reduce operating stresses Design for nondestructive inspection Metallurgical Change alloy composition Change alloy structure Use metallic or conversion coating Environmental Modify the environment Apply anodic or cathodic protection Add inhibitor Use organic coating Modify temperature Environmental: Change the environment or eliminate the environment from contact with the metal surface by using a coating.C.2. Furthermore. apply electrochemical protection (cathodic or anodic. on the other hand. or BiCl3 solutions [23]. an improperly heat-treated 18–8 stainless steel (e. of 18–8 austenitic stainless steels (containing 74% Fe). is the requirement of tensile stress acting conjointly with a specific environment. 26]. The damaging anions have no clear relation to general corrosion rates of the metals that they affect.C.C. despite well-defined corrosion paths along grain boundaries [24].C.2.C.C. The necessity for a tensile stress is illustrated by the sensitivity of 70% Cu–30% Zn brass to intergranular corrosion (not requiring stress) in a variety of electrolytes.. Following are five characteristics of S. Cu–Au. such as dilute H2SO4. General resistance or immunity of all pure metals.: 1.C.. cause S.C. and Mg–Al. Specificity of damaging chemical environments—not all environments that cause corrosion also cause S.C. but not nitrate ions.C.C. but the same heat-treated steel stressed in tension and placed in a boiling solution of magnesium chloride fails transgranularly by S. which is usually also intergranular. such as Cu–Zn.C.g. C.C. Notwithstanding the absence of an accepted generalized model of S. Successful use of cathodic polarization to avoid initiating S. CuAl2 precipitated from a 4% Cu–Al alloy) along grain boundaries or along slip planes. to the potential of grains.3. stressed in tension. Inhibiting effect of various extraneous anions added to damaging environments. Many models have been proposed to explain the mechanisms of S.C.MECHANISMS OF STRESS-CORROSION CR ACKING OF STEEL AND OTHER METALS 157 3.C. thereby exposing fresh anodic material to the corrosive medium. If the crack growth rate is controlled by the rate of dissolution at the crack tip. M is the atomic weight of the metal. Dix [29] proposed that galvanic cells are set up between metal and anodic paths established by heterogeneous phases (e. 1% V) that otherwise cracks in 3. or 1% sodium acetate.C.C.C. but each of these models has its own limitations in that it can be used to explain S. and d is the density . or active. of mild steel in boiling nitrates [21].1 Electrochemical Dissolution Electrochemical dissolution is the basis for the film rupture model. Any metal of large grain size.g. is more susceptible to S. The film rupture model was used to explain transgranular S. in a limited number of metal–environment systems. Cl− or acetates inhibit S. of 18–8 stainless steel (>200 h) [27]. da/dt: da M = ia × dt zFd where ia is the anodic current density. the applied stress effectively ruptures brittle oxide films at the tip of the crack. When the alloy. 5... For example. 4. at atomic levels.C.C. F is Faraday’s constant.C. β and γ brass (>40% Zn) may crack in water alone.C. An appreciable effect of metallurgical structure. Supporting this mechanism was a measured potential of metal at grain boundaries that was negative. in addition.C. 8. or 3.5% NaI added to magnesium chloride solution boiling at 130°C (33 g/100 mL) inhibits S. is exposed to corrosive environments.or transgranular. of copper in 1M NaNO2 solution [30]. Similarly.C. 1% Mo. whether failure is inter. In 1940.5% NaCl solution at room temperature [28]. and SO2− or NO− can be used 3 4 to inhibit a titanium alloy (8% Al. the ensuing localized electrochemical dissolution of metal opens up a crack.C.C.C. also called the slip–dissolution model. 2% NaNO3. but α brass (70% Cu. a few models are summarized in the following paragraphs. cathodic polarization prevented S. Also. Furthermore. for example. z is the valency of the solvated species.C. than is the same metal of small grain size. ferritic stainless steels (body-centered cubic) are much more resistant to Cl− than are austenitic stainless steels (face-centered cubic). 30% Zn) fails in NH3 or an amine.C. then the following equation can be used to estimate the crack growth rate. 8.3. Below the critical value of −0. Anodic polarization induces shorter cracking times the more noble the controlled potential. as well as the discontinuous nature of crack propagation [32. of 18–8 stainless steel initiates when exposed to magnesium chloride solution boiling at 130°C with and without inhibiting anion additions [27].C. this process repeats itself. This model can also be used to explain the crack arrest markings and cleavagelike facets on the fracture surface. A brittle crack initiates in this layer.2 Film-Induced Cleavage The film-induced cleavage model of S. cleavage-like fracture. and S.4).C. although more research into surface films and brittle fracture is required before this model can be thoroughly evaluated [34].158 EFFECT OF STRESS of the metal.H. with chemisorption facilitating the nucleation of dislocations at the crack tip and promoting the shear processes that result in brittle. 2. hydrogen embrittlement. 8.. Indeed.C.3 Adsorption-Induced Localized Slip The proponents of this model reasoned that similar fracture processes occur in liquid-metal embrittlement. Addition of various salts. providing some basis for the credibility of this equation.5 indicate the critical potentials noble to which S. 4. is based on the following five steps: 1. extends the observed cracking times. on the other hand.3.C.E.145 V (S. the alloy becomes essentially immune (Fig. 33]. The crack growth rate represented by this equation is an upper limit of the crack growth rate by continuous dissolution since neither dissolution nor cracking is likely to take place continuously. 8. 8. to the magnesium chloride solution shifts the critical potential to more noble values. A thin surface layer forms on the surface.4 Stress Sorption Figures 8. When the amount of . a correlation has been established for some systems between crack growth rates and dissolution current densities on bare surfaces. 5. but rather will be interrupted by film growth at the crack tip until film rupture takes place [31].3. After this crack becomes blunt and stops growing.). 3. The brittle crack continues to propagate in the metal substrate.4 and 8. cathodic polarization.C. such as sodium acetate.C. It was found that cleavage fracture occurs by alternate slip at the crack tip and formation of voids ahead of the crack tip [35–37]. The brittle crack crosses the film–metal-substrate interface with little loss in velocity. Effect of applied potential on time to failure of stressed moderately cold-rolled 18–8 stainless steel in magnesium chloride solution boiling at 130°C [27]. damaging components are also specific. The name suggested for this mechanism is stress-sorption cracking. increasing the tendency of the metal . Effect of applied potential on time to failure of stressed moderately cold-rolled 18–8 stainless steel in magnesium chloride solution with sodium acetate additions. Copyright 1969. Copyright 1969.C. Since chemisorption is specific. It was proposed [38. no longer occurs (Fig..C.MECHANISMS OF STRESS-CORROSION CR ACKING OF STEEL AND OTHER METALS 159 Figure 8. in general. The surface energy of the metal is said to be reduced. The Electrochemical Society.) Figure 8. (Reproduced with permission.5). The Electrochemical Society.5. boiling at 130°C (2% sodium acetate addition is inhibiting) [27]. 8.C. proceeds not by electrochemical dissolution of metal.4. but instead by weakening of the cohesive bonds between surface metal atoms through adsorption of damaging components of the environment.) added salt shifts the critical potential to a value that is noble to the corrosion potential. 39] that S. S.C. (Reproduced with permission. is interpreted as that value above which damaging ions adsorb on defect sites and below which they desorb. one being that large-scale mercury boilers can be constructed of carbon steel. in that only certain combinations of liquid metals and stressed solid metals result in intergranular failure (Table 8. Susceptibilitya of Solid Metals to Embrittlement by Liquid Metals [44] Liquid → Steel Copper alloys Aluminum alloys Titanium alloys a Li C C NC NC Hg NC C C C Bi NC C NC NC Ga NC — C NC Zn C — C NC C = cracking. apparently.160 EFFECT OF STRESS to form a crack under tensile stress. Stress-sorption cracking is the basic mechanism applying to stress cracking of plastics by specific organic solvents [41. 42] and to liquid-metal embrittlement—the cracking of solid metals by specific liquid metals. water adsorbed on glass understandably reduces the stress needed to cause fracture. .6). It is also the mechanism proposed earlier by Petch and Stables [43] to account for stress cracking of steel induced by interstitial hydrogen (see Section 8. which is also shifted in the noble direction by the presence of extraneous anions (Section 6. cracking no longer occurs. thereby making it necessary to apply a more positive potential in order to reach a surface concentration of the damaging species adequate for adsorption and cracking. Hence.C. hence. but this reduction of bond strength does not occur for iron in this application. The mechanism of competitive adsorption parallels that describe earlier applying to the critical pitting potential. Inhibiting anions. Whenever the inhibiting anion drives the critical potential above the corrosion potential.4). The mechanism of liquid-metal embrittlement is analogous to that of S. because catastrophic intergranular failure would result. Adsorption of any kind that reduces surface energy should favor crack formation. Such a potential. compete with damaging species for adsorption sites. but not of titanium or its alloys or of brass. This specificity has important consequences. this is because the damaging ion can no longer adsorb. which of themselves do not initiate cracks. It is.C.3) [44].C. can be either active or noble to the corrosion potential.C.3. the mechanism is related to the Griffith criterion [40] of crack formation in glass and similar brittle solids. TABLE 8. in principle. NC = no cracking. possible for adsorbed mercury atoms to reduce the metal bond strength within grain-boundary regions of stressed titanium or brass adequate to cause failure. for which the strain energy of the stressed solid must be greater than the total increased surface energy generated by an incipient crack. The critical potential for S. 7.3. occurs. Precracked specimen.). The susceptible range of potentials is still more restricted for slightly cold-rolled steel.6) [47].C. values of critical potentials. [45] Metal Solution Critical Potential (V.C.C.C. WQ Al. For deeper cracks. occurs.5 Initiation of Stress-Corrosion Cracking and Critical Potentials For various metals and solutions. For example. but much less so outside this range (Fig. Failure times are most rapid at potentials between those at which S.113 −1. Some Critical Potentials for Initiation of S.4. according to the stress-sorption viewpoint.H. S. 130°C MgCl2. Since the corrosion potential is more noble TABLE 8. 1% Tab Mild steel 63% Cu.) −0.4. 20% Ni stainless steel.C. in ammonium carbonate (170 g/liter) at 70°C within the potential range −0. are listed [45] in Table 8. 20.5. 2% Cb. Section 20. 7% Al. 130°C 0. 110°C 1M (NH4)2SO4. adsorption of damaging ions on mobile defect sites is optimum. room temperature 60% Ca(NO3)2. WQa 18% Cr.MECHANISMS OF STRESS-CORROSION CR ACKING OF STEEL AND OTHER METALS 161 8.095 18% Cr.C. 5. room temperature 3% NaCl.128 −0.04 −0. accounted for by metallic shielding effects within the crack and by solution composition changes resulting from factors like accumulation of anodic dissolution products within the crack. In the relatively narrow potential range over which S. cracks initiated to a depth not greater than 0.C.5% Mg Ti.C. In other words.H. 2. but also at potentials not far above (noble to) such a range. the imposed potential necessary to stop propagation is increasingly negative (active). b . For some metal–solution combinations.E.) stop propagating at a potential 5 mV or less below the critical value [46].C. 3% NH4NO3.013–0. immediately above which (or noble to) S.11 −1.1 −0. cold-rolled 36% reduction area 25% Cr.C. 154°C MgCl2.26 to −0. 0. water-quenched from 1050°C. 8% Ni stainless Steel. the environmental conditions required to initiate a crack are also those required for crack propagation.C.025 cm (0. Based on data obtained for 18–8 stainless steel in MgCl2 at 130°C. S. is avoided by polarizing not only below a certain critical potential or potential range. 130°C MgBr2. 8% Ni stainless steel. initiates.05M CuSO4.35 V (S.01 in. 8.3. pH 6. room temperature WQ.145 −0.005–0. c See Fig.5M NaCl. 37% Zn brassc a MgCl2.2.055 +0.C.E. mild steel undergoes pronounced intergranular S.5% Zn. such as PbO.H.6.C. and aluminum alloys. indicating greater sensitivity of failure time to change of applied stress. the susceptible potential range may be overlapped under conditions of cathodic protection accompanied by risk of S.162 EFFECT OF STRESS Figure 8.C. for smooth specimens of austenitic and martensitic stainless steels. brass.C.C.3. However.. 70°C [47].C. the higher the rate.E. either anodic or cathodic polarization extends time to failure. the subsequent rate of crack growth depends on prevailing stress. Effect of applied potential on stress-corrosion cracking of mild steel in 170 g ammonium carbonate per liter.).C.H. that is.71 V.)] and well removed from the susceptible range. 8.7) [48].8 support the concept of .] [0. the slope changes to one that is relatively shallow at low values of stress.C. carbon steels. For some metals. does not occur within 200 h or more unless dissolved O2 or an oxidizing agent. ammonium carbonate is not listed among the usually encountered damaging environments. A similar restricted potential versus time-to-failure curve is exhibited by mild steel in 35% NaOH at 85–125°C (caustic embrittlement) (Fig.90 V (S. S. if carbonates are present. 8.6 Rate of Crack Growth (Fracture Mechanics) Under conditions favorable to initiation of S. The linear relation and usually greater resistance of small-grain-size alloys is shown for brass in Fig. An empirical linear relation is found between applied stress and log time to fracture by S. however.C. 8. Since the corrosion potential is at −0. 8. [Reprinted with permission from Corrosion 32 (2).8 [49]. 57 (1976). Copyright NACE International 1976. the higher the stress. It will be noted.16 V(S. that neither the empirical relation expressing fracture time versus applied stress nor the typical data of Fig. is present that shifts the corrosion potential into the maximum susceptibility range at or near −0.E. In this event. Effect of applied potential on failure times of 0.) . No. (Data of Morris [49] replotted. 34% Zn brass exposed to ammonia.] Figure 8.Figure 8. 430 (1972). [Reprinted with permission from Corrosion 28.8. (11). Relation of applied stress to time to fracture of 66% Cu.09% C mild steel at three temperatures in 35% sodium hydroxide solution [48]. Copyright NACE International 1972.7. A stress intensity factor of 20 kg mm−3/2 corresponds to a crack 1 mm deep in a solid subject to a uniform stress of 10. at surface defects that are sufficiently deep.9. These changes plus increased local stress sometimes combine to initiate S. At low levels of K1. 8. A threshold value. Additionally. K1S.C. A low surface tensile stress means only that time to fracture is relatively long.. the rate cannot be assumed to be zero unless conditions are such that S.C. crack growth rate is very low.99 − 0. K1 takes into account the load and specimen geometry for a Mode 1 of crack opening. and W is specimen width.41 W + 18.C. does not occur. 8. K1 is given by the following expression [51]: K1 = Pa1/ 2 ⎡ a a ⎢1. Surface defects acting as effective stress raisers include corrosion pits and fatigue cracks.1/2 or kg mm−3/2. in short-time tests. a crack growth rate of 10−10 m/s corresponds to 0. three regions are observed (Fig. the composition and pH of solution within the defect can be altered by electrochemical transport through operation of differential aeration cells. or essentially zero. it is there that a crack grows most rapidly.C.C. K1.1 nm/s. From a practical standpoint. crack growth rate is highly sensitive to K1 (unlike region II). and temperature. Mode 1 type of crack opening. is designated at which the measured S. For reference. viscosity.70 W BW ⎣ ( ) ( )+ 2 ⎤ ⎥ ⎦ where P is applied uniform load.C. For the specimen geometry shown in Fig. B is specimen thickness. In region II.1 kg/mm2.9.10).C. a very low rate requiring years to measure.C.C. The dimensions of K1 are ksi in.C. . Since stress at a surface is highest at the root of a notch or imperfection.C.164 EFFECT OF STRESS a threshold stress below which S. If the stress intensity factor K1 is plotted versus a measured S. The actual stress near the tip of a crack of length a in a stressed elastic solid can be calculated in terms of the elastic stress intensity factor.C. the electrochemical potential differs from that at the surface. crack velocity. is not operative. Figure 8. as well as accelerate crack growth [50]. within region I. the crack velocity is independent of K1 but is strongly dependent on solution pH. also. For high-strength aluminum alloys. there is no crack growth in dry air whatever the value of K1.. then a critical crack depth.C.) of the metal. Crack velocities in mercury. the crack grows at an increasing rate until failure occurs. in general. region I is shifted to lower stress intensities the higher the humidity.C. show a similar dependence on stress intensity. Effect of stress intensity at a crack tip on stress-corrosion crack velocity.3% Cr.10. It is reasonable that some aspects of the crack mechanism overlap in the two different environments. Also. exists above which the stress intensity factor exceeds K1S.5% Zn) in NaCl solution and also in liquid mercury (liquid-metal embrittlement) at room temperature [52]. acr. 5. ⎞ ⎝ Y. Based on the previous expression for K1. ⎠ 2 .S.11 shows region I and II crack velocities for 7075 Al alloy (1. With increase of relative humidity.C.0% Cu. If one assumes that the applied stress is on the order of the yield strength (Y. although orders of magnitude higher than in aqueous solution. and region II is shifted to higher velocities [53].25⎛ S.2–2. 2. Figure 8.MECHANISMS OF STRESS-CORROSION CR ACKING OF STEEL AND OTHER METALS 165 Figure 8.1–2.9% Mg.S. the following approximate relation holds: K1 acr = 0. metallurgical factors affecting crack growth rates in one environment are similarly effective in the other. 0.C. Under this condition. ) because of its appearance] Stress-oriented hydrogen-induced cracking (S. also called hydrogen stress cracking Hydrogen blistering Hydrogen-induced cracking (H. and the higher the value of Y.C.. 8.) [also known as stepwise cracking (S. Dependence of crack velocity of 7075 Al alloy on stress intensity factor K1 [52].C.I.H.W. a stressed high-strength carbon steel or a martensitic stainless steel .S.11.4 HYDROGEN DAMAGE Hydrogen damage is a generic term that includes: • • • • • • Hydrogen embrittlement (loss of ductility) Hydrogen cracking. when stressed.) Sulfide stress cracking (S.C.166 EFFECT OF STRESS Figure 8.C. crack on exposure to a variety of corrosive aqueous solutions with no evidence that the damaging solutions need be specific. The lower the value of K1S.S.I.) Some metals. For example. the smaller the allowable depth of surface flaws leading to failure.C.O.C. The cracks tend to be mostly transgranular. in which hydrogen is more soluble than in ferrite and the diffusion rate is lower—are immune under most conditions of exposure [57]. Steels containing interstitial hydrogen are not always damaged.HYDROGEN DAMAGE 167 immersed in dilute sulfuric or hydrochloric acid may crack within a few minutes. bainite. but cracking usually takes place only under conditions of sufficiently high applied or residual tensile stress. cracking is caused by hydrogen atoms entering the metal interstitially either through a corrosion reaction or by cathodic polarization. such steels are cathodically protected under these conditions. 18–8 and 14% Mn steel (face-centered cubic). as is shown by the formation of visible blisters containing hydrogen when ductile metals are cathodically polarized or exposed to certain corrosive media. Failures of this kind are called hydrogen stress cracking or hydrogen cracking. This is contrary to the behavior of austenitic stainless steels in boiling magnesium chloride. . many stressed high-strength steels (e. more hydrogen escapes as molecular H2 and less adsorbed H is available to enter the metal. Carbon steels are especially susceptible to hydrogen cracking when heattreated to form martensite.C.1 Mechanisms of Hydrogen Damage The mechanism of hydrogen cracking has been explained by development of internal pressure [54] on the assumption that interstitial atomic hydrogen is released as molecular hydrogen at voids or other favored sites under extreme pressure.. they may follow former austenite grain boundaries [55]. sometimes observed during pickling in sulfuric acid or after electroplating. The failures have the outward appearance of S. Failures have also occurred within hours when martensitic 12% Cr-steel self-tapping screws (cathode) were applied in contact with an aluminum roof (anode) in a moist atmosphere.g. or martensite [56]. which retard the above reaction. Cracking of steel springs. They almost always lose ductility (hydrogen embrittlement). carbon steels or 9% Ni steels) have failed from this source within days or weeks after exposure to oil-well brines or to natural gas containing H2S [54]. but if the alloy is cathodically polarized. a less ductile metal would crack instead. In all these cases.. cracking is intensified.C. but are less so if the structure is pearlitic. When H2S is the poison. failures are described as sulfide stress cracking (S. Steels are less susceptible to hydrogen cracking above room temperature. 8. In martensite. Carbon steel heat-treated to form a spheroidized carbide structure is less susceptible than pearlite. is another example. Hence. If catalyst poisons that favor entrance of hydrogen into the metal lattice.). contrary to the effects of catalyst poisons. Such an effect certainly occurs. Austenitic steels—for example. cracking still occurs or happens in shorter time. Under similar conditions. such as sulfur or arsenic compounds.C.S.4. with iron becoming a better catalyst for the reaction Hads + H+ + e− → H2. In practice. are added to the acids. 4% C steel as a function of hydrogen content. Troiano. The delay time is only slightly dependent on stress. it decreases with increasing hydrogen concentration in the steel and with increase in hardness or tensile strength [58]. For small concentrations of hydrogen. The crack propagates discontinuously because plastic deformation occurs first. J. and then hydrogen dif- . Delayed fracture times and minimum stress for cracking of 0.168 EFFECT OF STRESS Figure 8.12. The critical stress decreases with increase in hydrogen concentration. Delay in fracture apparently results because of the time required for hydrogen to diffuse to specific areas near a crack nucleus until the concentration reaches a damaging level. These effects are shown in Fig. Hydrogen atoms preferably occupy such sites because they are then in a lowerenergy state compared to their normal interstitial positions. and finally subjected to a static stress [59].] An interesting characteristic of hydrogen cracking is a specific delay time for appearance of cracks after stress is applied. and A. and delayed failure in steel. fracture may occur some days after the stress is applied. A critical minimum stress exists.4% C) charged with hydrogen by cathodic polarization in sulfuric acid. These specific areas are presumably arrays of imperfection sites produced by plastic deformation of metal just ahead of the crack. below which delayed cracking does not take place in any time. crack initiation. AIME 212. Specimen initially charged cathodically. The hydrogen concentration was reduced systematically by baking. Trans. Hydrogen. 531 (1958). 8. then cadmium-plated to help retain hydrogen.12 for SAE 4340 steel (0. Morlet. Johnson. [Figure 5 from H. baked at 150°C for varying times to reduce hydrogen content [59]. and hence it lowers the critical minimum stress and shortens the delay time. Also. in boiling 3% NaCl.HYDROGEN DAMAGE 169 Figure 8. hydrogen embrittlement and cracking are minimized because diffusion of hydrogen is too slow. to the contrary. in practice. they break prematurely despite more-than-adequate test strength.C. failure times by hydrogen cracking (cathodic polarization). the effect of applied potential on failure times shows that. Below −110°C or at high deformation speeds.C.13) with resistance to cracking between these potentials. whereas failure at or below −1.1 V (Fig. A short notch in a steel surface favors plastic deformation at its base.C.C.02 V and below a potential of about −1. it is important that high-strength steel bridge cables be cold-drawn in order to avoid failure by S. whereupon the crack propagates one step further. steels fail by S. because the critical potential is shifted noble to the corrosion potential. The Electrochemical Society. 8. in moist air. Rc46.C.) fuses to imperfection arrays produced by deformation. Accordingly. In the absence of cold work. It is sometimes assumed that S. of high-strength steels of hardness >Rc40 (Table 8. In support. (Reproduced with permission.. as has happened in the failure of bridges . Failure in the upper potential region is better interpreted as S. whereas resistance to hydrogen cracking decreases.C.1 V corresponds to hydrogen cracking [60].C.40 ± 0. cracking occurs only above (noble to) a critical potential of −0. Copyright 1975. However.C.13 Effect of applied potential on time to failure of 4140 low-alloy steel. are longer the higher the temperature.1) in water or moist air is caused by hydrogen resulting from the reaction of H2O with iron.C. in H2O above room temperature in shorter times than at room temperature. cold working of high-strength steels improves resistance to S. in boiling 3% NaCl [60].C. C. Mg alloys. leading to increased internal pressure.I.. such as those occurring in maraging steels) by separation from the plastically deformed matrix. The significant effect of adsorbed H2O rather than interstitial hydrogen as a cause of cracking of high-strength steels probably also applies to high-strength martensitic and precipitation-hardening stainless steels. in final failure. [53].C. Internal cracks are nucleated where stress intensity is highest. Formation of hydrogen atoms at the steel surface and adsorption on the surface. MnS—provide internal flaws where hydrogen blisters often nucleate.170 EFFECT OF STRESS in the United States and other countries as well. cracks appear at right angles.I. similar to H. Eventually. such as at elongated inclusions oriented parallel to the rolling direction. can occur in the absence of any applied or residual stress in the steel other than the hydrogen pressure in traps.O. The blisters contribute to pipeline failures by H.2 Effect of Metal Flaws Nuclei for hydrogen cracking of steels form at the interface of precipitated phases (e. In steels.4. in S.C. and linkage of separate cracks. Whereas in H. failure can occur by S.H. the blister cracks tend to form in an array that is per- . Fe3C or intermetallic compounds. Accumulation of hydrogen atoms at hydrogen traps. thereby greatly improving resistance to hydrogen cracking of specimens stressed parallel to the rolling direction. nonmetallic inclusions—for example. Nuclei of perhaps Fe3C in low-carbon 10% Ni–Fe alloys can be oriented by cold rolling.I.g. H.C. 8.I. linking the cracks that previously developed at the elongated inclusions. Al alloys. 3. but not of specimens stressed at right angles [61]. crack initiation and propagation.C. On the other hand.C. pure carbon–iron alloys crack readily. a surfacedecarburized (hence softer surface) high-strength steel does not fail in boiling water or 3% NaCl solution. requires only a steel with a susceptible microstructure (typically elongated sulfide inclusions) and sufficient hydrogen to cause cracking. Under the influence of an applied stress. H.C. Diffusion of hydrogen atoms into the steel substrate. elliptically and spherically shaped inclusions are less damaging.Ti alloys. all of which are sensitive to failure in the presence of moisture.I.I. and β and γ brass. 2. such as voids around inclusions in the steel matrix. but is readily hydrogen-cracked when cathodically polarized.O. The role of internal lattice flaws on hydrogen cracking explains in part the resistance of stressed pure iron to cracking despite its becoming hydrogen-embrittled. The small amount of hydrogen produced by the H2O–Fe reaction has no effect on the hard steel core. except that the morphology of cracking is different..H.I. the blister cracks form at widely distributed sites and then link in a stepwise pattern. which occurs in three steps: 1. Furthermore. Similarly. by adding calcium to the steel composition. for example.or high-strength steels exposed to oil-well brines containing H2S. any processing of steel that minimizes sharp internal surfaces at the inclusion interface.I. It has been suggested [64]. corresponding to a yield strength of about 90 ksi [63]. or with copper. alloying with a small percentage of Pt or Pd.) pendicular to the applied stress.H. reduces the tendency toward cracking. (b) Stress-oriented hydrogen-induced cracking (S. for example by avoiding low rolling temperatures.14.H. and these individual blister cracks link leading to failure. Figure 8. Marked susceptibility to cracking in this environment makes it necessary to specify steels below the strength level at which failure occurs.I.C. Surface flaws may influence the hydrogen cracking (or sulfide cracking) of moderate. Increased resistance to H.HYDROGEN DAMAGE 171 2 mm (a) (b) Figure 8. Any factors that reduce hydrogen absorption in the corrosion process are beneficial.O.14 presents illustrations of H. which catalyze formation of molecular hydrogen at the steel surface [62].). and S.I.O.5. which forms an insoluble sulfide film that provides some protection at pH above about 4. (a) Hydrogen-induced cracking (H.C.C. can also be obtained by reducing the sulfur content in the steel and by controlling the shape of sulfides—for example.C.I.C. Strength of a steel is proportional to its hardness.I. based on threshold values of stress intensity .). The empirically determined maximum hardness is specified as Rockwell C 22 (Rc22). (Courtesy of Malcolm Hay. C.5 RADIATION DAMAGE Radiation damage is the change of properties of a material caused by exposure to ionizing radiation.S.C. or fission fragments in nuclear fuel material [65]. In general. In acids. During radiation. called “temperature spike. irradiated steel (but not pure iron) would presumably have a greater increase in rate than would irradiated nickel. neutrons.and high-strength steels.C. on this basis. radiation damage leads to increased hardness. and deformation behavior of the material as well as on the chemistry and electrochemistry of the solution. as intergranular S. 8. Indeed.C. At greater depths. are. such as HNO3 and H2O2. or formation of localized displacement spikes during radiation.S. metals for which the corrosion rate is controlled by oxygen diffusion should suffer no marked change in rate after irradiation. caused by clusters of interstitials and vacancies [66].C.C. irradiation-assisted stress-corrosion cracking (I.A. in H2S environments.C.. Austenitic stainless steels often become more sensitive to S. corresponding with the general experience of the oil industry that. in which many atoms move into interstitial positions. and dislocations are produced.C. which is less sensitive to cold working.C. in which few or no atoms leave their lattice sites. of austenitic stainless steels is related to chromium depletion along grain boundaries. and it has been observed. The porous structure produced by dealloying has been described as similar to that found in irradiated materials [66].) is a cause of failure of core components in both boiling water reactors (BWR) and pressurized water reactors (PWR). and displacement spikes. are not readily avoided in practice. which are calculated to be still less tolerable for harder steels. a local temperature rise. surface flaws become more important to cracking the higher the strength of the steel. in I. Small-size flaws of this order. they might be expected to become more sensitive after irradiation. the effect of radiation may be expected to parallel that of cold work. The complexities of S. There are two kinds of spikes: thermal spikes. That is. microchemistry.C. interstitial atoms. which have a secondary effect on corrosion. In addition to the grain-boundary chromium depletion that can result from sensitization during . I. affect both low..” may occur. gamma rays. steels of hardness >Rc22 should be avoided. and these increase the diffusion rate of specific impurities or alloyed components. Except for chemicals produced in the environment by radiation. heavy-particle radiation. Lattice vacancies.172 EFFECT OF STRESS factors for steels exposed to aqueous H2S solutions. that Rc22 corresponds to a critical surface-flaw depth of about 0.5 mm (0.02 in).A. flaws are likely to develop quickly into major cracks. such as X-rays. on the other hand. after cold working.C.C. Metals exposed to intense radiation in the form of neutrons or other energetic particles undergo lattice changes resembling in many respects those produced by severe cold work.A.S. internal flaws. on the other hand. compounded by the effects of radiation on microstructure. in austenitic stainless steels and nickel-base alloys [67]. In copper and nickel at room temperature. Accelerated corrosion of Zircaloys induced by irradiation is not observed above 400°C (750°F).15. [68.6 CORROSION FATIGUE A metal that progressively cracks on being stressed alternately (reverse bending) or repeatedly is said to fail by fatigue. on the other hand.2) [69]. which have a maximum carbon content of 0. has also been used successfully [70]. In a corrosive environment. chromium depletion in the grain boundary can also occur because of neutron irradiation.2. In general. further reducing resistance to S. type 347 stainless steel. the shorter is the time to failure. 8.3.C. One materials solution to mitigate I. Cox [72] stated that both fast neutron irradiation and presence of dissolved oxygen or an oxidizing electrolyte must be present simultaneously for any observed acceleration of corrosion to occur in high-temperature water.2. in nuclear reactors is to use nuclear-grade stainless steels. Of course.2) at 250°C in dilute uranyl sulfate solution containing small amounts of H2SO4 and CuSO4 was very much increased by reactor irradiation [71]. In a review of the subject. a true endurance limit exists which is approximately half the tensile strength. For steels. 69]. a design without crevices. the corrosion of a zirconium alloy (Zircaloy-2.100% to maintain strength at the lower carbon content compared to the non-nuclear grades (see Table 19. such as type 316NG and type 304NG.15).3.3 and 19.C. Niobiumstabilized.1). see Section 26. In other words. 8. The greater the applied stress at each cycle. 8.C.S. The effect of irradiation on corrosion of some uranium alloys is considerable.2 and Fig. A number of cycles at the corresponding stress to the right of the upper solid line results in failure.2. A plot of stress versus number of cycles to failure.2. but no failure occurs for an infinite number of cycles at or below the endurance limit or fatigue limit. a corrosive environment can decrease the fatigue properties of any engineering alloy. a 3% Cb–U alloy having moderate resistance to water at 260°C disintegrated within 1 h after irradiation. Cracking of a metal resulting from the combined action of a corrosive environment and repeated or alternate stress is called corrosion .). called the S–N curve. Furthermore. but not necessarily for other metals. 19. meaning that corrosion fatigue is not dependent on material and environment [73]. Section 19.CORROSION FATIGUE 173 welding (see Figs. failure at a given stress level usually occurs within fewer cycles.020% and a nitrogen content of 0.060 to 0.G. Section 19. and a true fatigue limit is not observed (Fig. is shown in Fig.C. failure occurs at any applied stress if the number of cycles is sufficiently large. 19. Frequency of stress application is sometimes also stated because this factor may influence the number of cycles to failure. The fatigue strength of any metal. which can provide sites for localized corrosion (see Sections 2. The effects have been explained in terms of changes in the physical properties of the protective oxide film. is the stress below which failure does not occur within a stated number of cycles. with low carbon called 347NG. For example. good corrosion design is also essential—for example. Figure 8.16). The damage is almost always greater than the sum of the damage by corrosion and fatigue acting separately. Corrosion-fatigue crack through mild steel sheet. 8.15. Fatigue cracks are similarly transgranular (exception: lead and tin). resulting from fluttering of the sheet in a flue gas condensate (250×). but rarely is there evidence of more than . fatigue. They are often branched (Fig.174 EFFECT OF STRESS Figure 8.16. and several cracks are usually observed at the metal surface in the vicinity of the major crack accounting for failure. S–N curve for steels subjected to cyclic stress. Corrosion-fatigue cracks are typically transgranular. in water at room temperature and have shorter fatigue life in moist air compared to dry air.C.C.C. always represents a measure of corrosion fatigue. the shaft of a ship propeller slightly out of line will operate satisfactorily until a leak develops. but had considerable effect on rate of crack propagation. resistance of a metal to corrosion fatigue is . used to pump oil from underground. resulting in eventual parting and failure of the shaft. but contrary to the situation for copper. In early tests. in part. Stainless steels and nickel or nickel-base alloys are also better than carbon steels.. whereas steels of lower strength. Pipes carrying steam or hot liquids of variable temperature may fail similarly because of periodic expansion and contraction (thermal cycling). In general. Steel oil-well sucker rods. fatigue life of oxygen-free. which do not undergo S. in water. This is seen in the behavior of a high-strengh low-alloy steel fatigued in both the absence and presence of moisture. The main effect was attributed to oxygen.. this had little effect on initiation of cracks. but pitting is not a necessary precursor to failure. therefore. Fresh waters and particularly brackish waters have a greater effect on the corrosion fatigue of steels than on that of copper. 8. Despite use of high-strength medium-alloy steels and oversized rods. general chemical environments. corrosion pits may form at the metal surface at the base of which cracks initiate. In some environments. Fatigue life of pure aluminum was also affected by air. which neither chemisorbs oxygen nor oxidizes. For example.C.C. Corrosion fatigue of steel occurs in fresh waters. In corrosion fatigue. Steels of greater than about 1140 MPa (165 ksi) yield strength (Rc37) are subject to S. but for 70–30 brass it was 26% [74]. Cracks may then develop within a matter of days. contrary to the situation for S.C. fatigue cracking is supplemented by S. Corrosion fatigue is a common cause of unexpected cracking of vibrating metal structures designed to operate safely in air at stresses below the fatigue limit. with the general rule that the higher the uniform corrosion rate. failures from this source are a loss to the oil industry in the order of millions of dollars annually. water vapor in absence of air was equally effective. with the latter being a more corrosion-resistant metal. the shorter the resultant fatigue life.17) [76].C. allowing water to impinge on the shaft in the area of maximum alternating stress. For mild steel the increase was only 5%. Wire cables commonly fail by corrosion fatigue. high-conductivity (OFHC) copper at 10−5 mmHg air pressure was found to be 20 times greater than at 1-atm air pressure.CORROSION FATIGUE 175 one major crack. The usual fatigue test conducted in air on a structural metal is influenced by both oxygen and moisture and. and so on. for which only certain ion–metal combinations result in damage. Gold. have the same fatigue life (Fig. Aqueous environments causing corrosion fatigue are numerous and are not specific.C. In later tests [75]. have limited life because of corrosion fatigue resulting from exposure to oil-well brines. with the latter occurring under conditions of constant stress. had the same life whether fatigued in air or in a vacuum. seawater. the fatigue strength for copper in a partial vacuum was found to be increased 14% over that in air. combustion-product condensates. 176 EFFECT OF STRESS Figure 8. and with aeration. and fatigue strength at 107 cycles in 93% relative humidity air. fatigue limit in dry air. are found to vary with rate of stressing. yield strength. Unlike the fatigue limit in air. besides varying with environment. A few values of corrosion fatigue strength determined by McAdam [77] in fresh and brackish waters are listed in Table 8.5 and similar data are as follows: 1. These values. Conclusions from data of Table 8. with temperature. 2. Tensile strength. hence. (With kind permission of Springer Science and Business Media.17.5. Medium-alloy steels have only slightly higher corrosion fatigue strength than carbon steels. they are useful only for qualitative comparison of one metal with another. . There is no relation between corrosion fatigue strength and tensile strength. they are not usually reliable for engineering design. of 4140 steel heat-treated to various hardness values [76].) associated more nearly with its inherent corrosion resistance than with high mechanical strength. 78% Cu.000 9.96%.57 0. multiply by 6890.71 0.2% Mn. 8% Ni.86 0.4%. 99.000 36. 8.000 50. 140 ppm NaCl. 0.02 0. 21% Ni.51 0. quenched and tempered 17% Cr.000 50. 67.02 0.00 0.1 Critical Minimum Corrosion Rates In order for the corrosion process to affect fatigue life.000 10.500 26.55 0. annealed 760°C Cupro-nickel. c Severn River water.65 0.800 5.000 23.9% Cr.500 7.70 1. the corrosion rate must exceed a minimum value.000 22. tempered Brass. 17 ppm MgCl2. Corrosion fatigue strength of all steels is lower in salt water than in fresh water.500 0.6.000 21. 98%.8% Cr.1% V. Corrosion-resistant steels.36 0. particularly steels containing chromium. 0.000 35.1% C steel.77 0.5% Ni.000 20.000 35.52 0.000 5. annealed 3. have higher corrosion fatigue strength than other steels. 60–40.000 33. 1.45 0.000 18. annealed 760°C Monel. hot-rolled Nickel.36 0.000 42. with about one-sixth the salinity of seawater.59 0.11% C steel.000 29. 5.000 2.64 0.95 1. translating uniform .09% C steel.000 23.95 1.38 To convert from psi to MPa.000 16. 0. Such rates are conveniently determined by anodic polarization of test specimens in deaerated 3% NaCl. annealed a b 25.50 0.800 6. residual stresses are deleterious.700 17.000 50.3% C steel.000 21.000 7. 0. 0. quenched and tempered 1. 2 ppm CaSO4.5% C steel.71 0.000 42.000 18.000 10. annealed 13.2 0. 4.500 28. annealed 650°C Aluminum. 3. annealed Aluminum.100 3.000 49.000 18.2% C steel. annealed 0. Fatigue Limit and Corrosion Fatigue Strength of Various Metals [77] Metal Fatigue Limit in Air (psia) Corrosion Fatigue Strength (psia) Damage Ratio (Corrosion Fatigue Strength)/ (Fatigue Limit) Well Water Salt Water Well Waterb Salt Waterc (107–108 cycles at 1450 cycles/min) 0. hard Duralumin.700 18.36 0.5% Cu. 200 ppm CaCO3. 29.5% Ni. Heat treatment does not improve corrosion fatigue strength of either carbon or medium-alloy steels.000 25. 98. annealed 0.5.CORROSION FATIGUE 177 TABLE 8.16% C steel. annealed 760°C Copper.900 10.500 19. Inhibitors are also effective [82.6. 8. Various values at 30 cycles/s (1800 cycles/min) are listed in Table 8.g.2 Remedial Measures There are several means available for reducing corrosion fatigue. Critical Minimum Corrosion Rates. 30 Cycles/s i μA/cm2 a 0.50 0.H.8 1.. and the fatigue life regains its value in air [78]. 4–15 mdd). the uniform corrosion rate falls below the critical rate. 25°C. 8.2 100 To convert μA/cm2 to A/m2.70 <0.E. In the case of mild steel.H. whereas the value required to attain zero corrosion rate lies 40 mV more active at −0. The existence of a critical rate in 3% NaCl explains why cathodic protection of steel against corrosion fatigue requires polarizing to only −0. made by B.0 2.178 EFFECT OF STRESS TABLE 8.50 0. The average value of 0. Sacrificial coatings (e.0 2.49 V (S. the fatigue life of copper is observed to be about the same in air as in fresh and saline waters (Table 8. Haigh in about 1916. and (c) heat treatment (hardness).8 mdd) is less than the uniform corrosion rates of steels in aerated water and 3% NaCl (1–10 gmd. But at pH 12.5 gmd. thorough deaeration of a saline solution restores the normal fatigue limit in air (Fig. One of the very first observations and diagnoses of corrosion fatigue.5 78 76 76 76 79 80 2.5).63 0.). For copper.6. For the steels.18% C mild steel 4140 low alloy steel Rc20 Rc37 Rc44 Nickel OFHC copper a Reference gmd 0. divide by 100. zinc or cadmium electrodeposited on steel) are effective because they cathodically protect the base metal at defects in the coating.18) [81]. the critical rate of 28.53 V. and galvanizing .E.) accomplishes the same result. current densities by Faraday’s Law into corrosion rates below which the fatigue life is unaffected.6. It is expected that such values increase with cyclic frequency of the test.5 2. Cathodic protection to −0.30 28.5 gmd (285 mdd) is much higher than the uniform corrosion rates in aerated water and 3% NaCl (0. involved premature failure of steel towing cables exposed to seawater. critical corrosion rates are independent of (a) carbon content. 83].49 V (S.4–1. 10–100 mdd). (b) applied stress below the fatigue limit. These measured current densities are found to be independent of the total area of specimen surface exposed to anodic currents. hence.58 gmd (5. 8. presumably. they owe their effectiveness to exclusion of the environment [85]. or otherwise introducing compressive stresses. corrosion resistance of a metal is usually more important than high tensile strength in .6. The basic effect of the corrosion process is to accelerate plastic deformation accompanied by formation of extrusions and intrusions. Below the fatigue limit. or silver on steel are also said to be effective without reducing normal fatigue properties. which protrude above the metal surface (extrusions).18. copper. For this reason. which in turn impedes the fatigue process. Effect of dissolved oxygen concentration in 3% NaCl. 25°C.3 Mechanism of Corrosion Fatigue The mechanism of fatigue in air proceeds by localized slip within grains of the metal caused by alternating stress. resulting in slip steps at the metal surface. Shot peening the metal surface.) provided greatly increase life in this application [84]. All rights reserved. Continued slip produces displaced clusters of slip bands. is beneficial. In addition. damage by corrosion fatigue—a conjoint action of corrosion and fatigue—is greater than the damage caused by the sum of both acting separately. 8.asminternational.19). work hardening accompanying each cycle of plastic deformation eventually impedes further slip. Organic coatings are useful if they contain inhibiting pigments in the prime coat. (Reprinted with permission of ASM International®. on fatigue behavior of 0. Adsorption of air on the clean metal surface exposed at slip steps probably prevents rewelding on the reverse stress cycle. lead. corresponding incipient cracks (intrusions) form elsewhere (Fig.18% C steel [81].org. Electrodeposits of tin. www.CORROSION FATIGUE 179 Figure 8. Specimen plated with silver after test and mounted at an angle to magnify surface protuberances by about 20×. they are also not immune to corrosion fatigue. and. Metall. . Since pure metals are not immune to uniform corrosion. The fatigue process in copper as studied by electron metallography. Wood and H. 182 (1962).19. Soc. Overall magnification about 500×. AIME 224. [Figure 3 from W. it may lead to failure by fatigue or corrosion fatigue. The mechanism of the slip dissolution process [86] takes place in the following steps: • • • • Diffusion of the active species to the crack tip Rupture of the protective film at the slip step Dissolution of the exposed surface Nucleation and growth of the protective film on the bare surface 8.] establishing optimum resistance to corrosion fatigue.7 FRETTING CORROSION Fretting corrosion is another phenomenon that occurs because of mechanical stresses. Extrusions and intrusions in copper after 6 × 105 cycles in air. Trans. in the extreme. Bendler.180 EFFECT OF STRESS Figure 8. leads to similar damage. because dimensional accuracy is lost. The rapid conversion of metal to metal oxide may. The corrosion products are exuded from the faying surfaces.FRET TING CORROSION 181 Figure 8.20. as when one roll moves slightly faster than another roll in contact. subject to slight relative slip. are largely composed of αFe2O3 plus a small amount of iron powder [87]. It is at such pits that fatigue cracks eventually nucleate. for example. by formation of pits. Continuous slip. in the case of steel. in itself. Wear corrosion and friction oxidation are terms that have also been applied to this kind of damage. or corrosion products may cause clogging or seizing. one or both being metal. cause malfunction of machines. Damage by fretting corrosion is characterized by discoloration of the metal surface and. Weight loss of mild steel versus mild steel by fretting corrosion in air [87]. In the case . It is defined as damage occurring at the interface of two contacting surfaces. that caused by vibration. The slip is usually oscillatory as. in the case of oscillatory motion. which. called asperities.1 Mechanism of Fretting Corrosion When two surfaces touch. Fretting corrosion is also increased by increased slip. Because the oxides are unable to escape during oscillatory slip. the ball-bearing races of the wheels became badly pitted by fretting corrosion. connecting rods. contacts of electrical relays. These effects are depicted in Fig. king pins of auto steering mechanisms. 8. is obviously not electrochemical. Relative slip of the surfaces causes asperities to . where fluid flow generates vibrations [92]. shrink fits. the products are NiO and a small amount of nickel. but in nitrogen no frequency effect is observed. Fretting corrosion is frequently the cause of failure of suspension springs. High vibration levels and other types of mechanical stress result in small relative movement between parts of operational aircraft systems [89]. therefore. they are Cu2O with lesser amounts of CuO and copper [88]. One of the first examples of fretting corrosion was recognized when automobiles were shipped some years ago by railroad from Detroit to the West Coast.7. Fretting corrosion is also a problem in nuclear reactors. particularly on heat-exchanger tubes and on fuel elements.2 times as much in the case of iron) than the metal from which the oxide forms. their accumulation increases the stress locally. but not moisture. and many parts of vibrating machinery. The mechanism. Increased load increases damage. provided that the surface is not lubricated. Note that the initial rate of metal loss during the run-in period is greater than the steady-state rate. for copper. Because of vibration. damage is less in moist air compared to dry air and is much less in a nitrogen atmosphere. it can cause derailment and other damage [90]. but could be avoided if the load on the wheels was relieved during shipment. and atmospheric corrosion of the substrate increases the contact resistance to intolerable levels [91].182 EFFECT OF STRESS of nickel for continuous-slip experiments. accounting for the tendency of pits to develop at contacting surfaces because corrosion products—for example. Relative motion between the contacts removes the gold plate. thereby accelerating damage at specific areas of oxide formation. Electrical contacts made of electroplated gold coatings can fail by fretting corrosion. so that the automobiles were not operable. It may cause discoloration of stacked metal sheets during shipment. Increase in frequency for the same number of cycles decreases damage. where the surface protudes. bolt and rivet heads. 8. Another example of fretting corrosion can occur with railroad car wheels that are shrink-fitted onto their axles. αFe2O3—occupy more volume (2. Also. Fretting corrosion occurs on airframe structural surfaces that move against each other in a corrosive environment. Damage increases as temperature is lowered.20. variable-pitch propellers. Damage was worse in winter than in summer. jewel bearings. If failure occurs causing the wheel to come off the axle. Laboratory experiments [87] have shown that fretting corrosion of steel versus steel requires oxygen. contact occurs only at relatively few sites. An equation for weight loss W of a metal surface undergoing fretting corrosion by oscillatory motion has been derived [93] (Appendix. which in turn is wiped off. and k2 are constants. Idealized model of fretting action at a metallic surface.20: W = (k0 L1/ 2 − k1L) C + k2 lLC f where L is the load. The first two terms of the right-hand side of the equation represent the chemical factor of fretting corrosion. by less Figure 8.21. causing a certain amount of wear by welding or shearing action. Section 29. This is demonstrated by increased damage at below-room temperatures. electrical resistance between the surfaces is at first low. Any metal particles eventually are converted partially into oxide by secondary fretting action of particles rubbing against themselves or against adjacent surfaces. f. It is found that either the mechanical or chemical factor may dominate in accounting for damage depending on specific experimental conditions. in the case of metal. The last term is the mechanical factor independent of frequency. forming another fresh metal track (Fig.7) on the basis of the model just described. These become smaller the higher the frequency. k1. through which metal particles are dislodged. then becomes high and remains so). which accounts reasonably satisfactorily for data of Fig. the metal surface after an initial run-in period is fretted by oxide particles moving relative to the metal surface rather than by the mating opposite surface originally in contact (hence. f. . which. This is the chemical factor of fretting damage. f is the frequency. Fretting corrosion is not a high-temperature oxidation phenomenon. l is the slip. or it may mechanically activate a reaction of adsorbed oxygen with metal to form oxide. corresponding to less available time for chemical reaction (or adsorption) per cycle. but proportional to slip and load. In addition. Also.21). In nitrogen. or it may oxidize superficially. and W is independent of frequency.FRET TING CORROSION 183 rub a clean track on the opposite surface. 8. asperities plow into the surface. and k0. C is the number of cycles. This is the mechanical factor. the mechanical factor alone is operable. immediately becomes covered with adsorbed gas. the debris is metallic iron powder. The next asperity wipes off the oxide. 8. Low-viscosity oils. which melts at 80°C. but the beneficial effects tend to be temporary because the lubricant is eventually displaced by surface movement. by steel being badly fretted in contact with polymethacrylate plastic. Because of their relatively poor strength. thereby reducing damage.. At low temperatures.184 EFFECT OF STRESS damage at high frequencies. Tin-. In addition. Combinations of stainless steels tend to be worst. but not higher [94]. Polytetrafluoroethylene (Teflon) has a low coefficient of friction and reduces damage. Combination of a Soft Metal with a Hard Metal. thereby avoiding slip at the interface. Design of Contacting Surfaces to Avoid Slip Completely (e. At sufficiently high loads.indium-.g. for which surface temperatures are highest. More fundamental research is needed to help clarify these details of the general mechanism. Brass in contact with steel tends to be less damaging than steel versus steel. Low-viscosity oils diffuse more readily to the clean metal surface produced by oscillatory slip. Cements exclude air from the interface.. particularly in combination with a phosphate-treated surface. particularly if baked onto the surface. by the fact that oxide produced in fretting corrosion of iron is αFe2O3 and not the hightemperature form. Intentional design to completely avoid slip is not always easy to accomplish. and hence the surface could have reached temperatures of this order. . lead. furthermore. 3.2 Remedial Measures 1. silver-. 2. Use of Cobalt-base Alloys. because damage is presumably caused by relative movement approaching the order of atomic dimensions. Increased load is effective in this direction if it is high enough to prevent slip. hydrated αFe2O3 is probably less abrasive than the anhydrous oxide. otherwise. finally. Use of Elastomer Gaskets or Materials of Low Coefficient of Friction. The effect of moisture when adsorbed on the metal surface may be that of a lubricant.7. Molybdenum sulfide is effective as a solid lubricant. and cadmium-coated metals in contact with steel have been recommended. grit blasting. a soft metal may yield by shearing instead of sliding at the interface. and. 4. soft metals serve to exclude air at the interface. or otherwise roughening the surface). rubber cement to the faying surfaces). materials of this kind are expected to be effective only at moderate loads.g. damage is greater presumably because O2 can adsorb more rapidly or more completely than at high temperatures. damage is worse. 5. Rubber absorbs motion. These are effective especially in the presence of liquid water or aqueous solutions (see Chapter 24). Use of Lubricants. 8. Fe3O4. 6. can be helpful in reducing damage if the load is not too high. Application of Cements (e. 24. 964 (1965). 50 (1977). Intercrystalline Cracking of Boiler Steel and Its Prevention. V. H. Berk. 785 (1953). 25. 1944. Uhlig. 26. 1945. R. M. 95 (1985). Robertson. Corrosion 24. 577. MH-2–95. p. 154. 149 (1952). S. 22. 7(1). Mag. 6. Jones. J. 1956. D. in Corrosion– Deformation Interactions. Revie. 13. Prot. MA. 9. Stahl u. Chambreuil-Paret. R. in Proceedings. 1969. 20. in Symposium on Stress-Corrosion Cracking of Metals. Kim. 547 (1970). J. and R. M. 1247–1248. TX. and B. W. Corrosion 21. Wiley. R. European Federation of Corrosion Publications. Treseder and A. and B. Philos. 27. M. Delafosse. Corros.. in Symposium on Stress-Corrosion Cracking of Metals. Parkins. Elayaperumal. W. 1941. 16. Bureau of Mines Bulletin 443. Wachter. Electrochem. New York. Buchholtz and R. Metals 56. Cambridge. London. Soc. R. Eisen 62. 172. 12. Sava. C. D. R. C. W. W. Electrochem. N. A. E. No. W. 522 (1964). 3. 1948. 518 (1995). in Uhlig’s Corrosion Handbook. pp. H. Newman. 125 (1965). 7. 1966. Talerman. J. PA. Nagata. Kowaka and S. Wilkins. 433. P. 2nd edition. Calgary. Uhlig and E. in Uhlig’s Corrosion Handbook. K. 19. Wiley. Corrosion 51(7). Stahl u. H. 18. National Association of Corrosion Engineers. GPO. Parkins. Harrison. p. A Brown. W. Electrochem. Wiley. Pusch. 8. Soc. 11.REFERENCES 185 REFERENCES 1. Philadelphia. Department of Metallurgy and Materials Science. M. editor. M. 229 (1974). Uhlig. Streicher. 5. 374. Cook. H. Magnin. New York. 15. Charles and R. The Institute of Materials. 2000. Parkins. A. 116. Washington. Château. Lynes.S.I. Uhlig. R. Corrosion 5. 12. Z. 10. 4. T. p. Mater. pp. p. . H. Corrosion 30. Trans. 21. N. Baerlecken and W. 173 (1969). in Stress Corrosion Cracking and Embrittlement. Sieradski and R. K. editor. M. K. 361 (1963). R. A 51 (1). Alberta. Parkins. 17. D. Schroeder and A. p. and M. A. editor. N. Bohnenkamp. 383 (1949). M. ASTM–AIME. Wilde. p. J. National Energy Board of Canada. p.. 21 (1942). Public Inquiry Concerning Stress Corrosion Cracking on Canadian Oil and Gas Pipelines. 2000. Long and H. Am. Foroulis and H. Wiley. 437. 21. 2nd edition. U. Eisen 73. 427 (1968). 2000. Revie. Iron Steel Inst. Jr. 23. Soc. Uhlig and J. 111. Pollard. Hirsch. 10. 112.. Parkins. 14. Wiley. Bayle. Uhlig. 200. ASTM–AIME. Sci. E. Conference Fundamental Aspects of Stress Corrosion Cracking. H. Corrosion 33(2). R. 191–194. Report of the Inquiry. 36 (1968). D. thesis. Soc. PA. J. A. 2. Scheil. R. E. Philadelphia. November 1996. New York. Holzworth et al.. Houston. N. New York. 2nd edition. National Energy Board. editor. 1997.T. Canada. Agarwala. in Corrosion Handbook. editor. in Uhlig’s Corrosion Handbook. Copson. New York. H. Revie. H. J.C. Shehata. Mater. Dienes. H. 50. see also S. Corrosion 32. Wang. R. Houston. W.. Am. Weinstein. 11 (1940). 6. The Metallurgical Society. pp. 701 (1979). Corrosion: Fundamentals. Lynch. Interscience. ASTM–AIME. ASM International. H. McCaughey. 43. p. 1981. and B. October 1988. W. 13A. 126. pp. Trans. The Institute of Metals. D. TX. Nature 169. 53. and W. Kluwer Academic Publishers. N. 51. CDI ’96. 29. M. A. R. p. in Uhlig’s Corrosion Handbook. A. Lynch. 41. T. Thompson and I. 144. M.-Z. P. 1960. Morris. 863. and E. Fischer. Materials Park. 15 (10). R. Coleman. Lynch. Scientific Affairs Division. Revie. Uhlig. Markus. ibid. 1–17. New York. Astronaut. Y. 49. Jones. I. 10. Massoumzadeh. Soc. TX. W. 1. Reinhold. Corrosion 28. Uhlig. R. J. 9. Am. Dix. 47. Revie. T. p. F. Aeronaut. Res. 2003. J. S. 326 (1968). Kunststoffe 55. Broek. 273 (1984). Mazille and H. Petch and P. OH. Inst. M. E. Hydrogen-induced cracking and sulfide stress cracking. N. 39. 33. E. M. Stress corrosion crack initiation processes: Pitting and microcrack coalescence. H. Ives. Staehle. A. 842 (1952). J. 4th revised edition. 191 (1979). Scully. editor. Uhlig. A. Rostoker. in Physical Metallurgy of Stress Corrosion Fracture. Mater. D. 37. 56 (1976).-H. Hydrogen Effects in Metals. M. in Symposium on Stress-Corrosion Cracking of Metals. 52. The Netherlands. Embrittlement by Liquid Metals. in Environment-Induced Cracking of Metals. Electrochem. Dix. p. J. Uhlig. Lynch. 305. Eng. 85. Brussels. European Federation of Corrosion Publications. Philadelphia. 256 (1930). pp. 1629 (1973). K. Elementary Engineering Fracture Mechanics. 3329 (1985). Pednekar. editors. London. 1959. Electrochem. 298 and 302. p. North Atlantic Treaty Organization. 40. Vol. 45. J. Beck and M. S. Stables. Gangloff and M. Proceedings of the First International Conference on Environment-Induced Cracking of Metals. in Corrosion–Deformation Interactions. 1945. Min. 20. New York. E. P. 1990. 137. A. H. J. 1 (1986). Trans. editors. G. in Theory of Stress Corrosion Cracking in Alloys. 00379. Metall. 1997. Sci. S. H. 46. R. Agrawal. Min. Brown. Parkins. 32. 44. 323. Second International Conference on Corrosion–Deformation Interactions in Conjunction with EUROCORR ’96. Rostoker. and M. 120. 31. 42. Surf. 206. 35. K. p. editor. 48. Thomson. P. B. Speidel. Testing. Eng. Philos. J. 427 (1972). N. 205–220. Lin and R. Wiley. Sci. Paskin. Electrochem. 463 (1981). 34. R. NACE International. Mears. Jr. 30. Bernstein. W. Met. 362. 54. in ASM Handbook. editor. P. Metall.186 EFFECT OF STRESS 28. J. 1971. Elboujdaini. 453 (1965). Chaung. Sci. R. Inst. Houston. and H. A220. Elboujdaini. . 2000. Appl. 491 (1961). Number 21. CORROSION/2000. R. and R. National Association of Corrosion Engineers. P. P. Parkins. Hixson and H. (Gr. J. D. 89. Sieradzki. 1986. Uhlig. Soc. 2000. 36. T. Paper No. Acta Metall. Stress corrosion cracking. Newberg and H. H. Britain) 9. S. M. 2nd edition. Griffith. R. Blackburn. W. Rhodin. 38. 163 (1920). New York. p. and Protection. Asher et al. Corrosion: Fundamentals. Am Soc. L. 102 (1927). J. in Environment-Induced Cracking of Metals. p. H. A. Materials Park. 65. Stress corrosion cracking behavior of alloys in aggressive nuclear reactor core environments. Corrosion 36. Nucl. October 1988. Metall. McAdam. Materials Park. 13C. Inst. 386–404. National Association of Corrosion Engineers. J. Hyatt. R. Materials Performance 12(3). p. 74. Sopwith. D. p. 382–396. Wadsworth. editor. Trans. Fig. Proceedings of the First International Conference on Environment-Induced Cracking of Metals. and H. p. editors. Interscience. Uhlig. S. Sci. 625 (1980). 73. 10. editor. pp. Min. R. Vol. Gangloff and M. Trans. 122. Testing Mater. New York. OH. Corrosion 63 (1). 183 (1944). Horn. Cox. A. NACE Committee T-1F. 66. Am. pp. Andresen. Vol. P. Houston. 67. A. Inst. 695 (1970). editors.-Z. in Uhlig’s Corrosion Handbook. in The Effects of Radiation on Materials. Metals 52. 2003. R. G. 56. M. Am. TX. Wiley. Was and P. in ASM Handbook. Simnad. J. Asphahani and H. Vol. 75. 528 (1958). Houston. Corrosion: Environments and Industries. H. 437t (1957). 19 (2007). 41 (1973). Rassweiler and W. Johnson. 6. Elsevier. and E. M. Carter and M. 70. 64. 174 (1975). 1990. T. Uhlig. N. New York. P. editors. 1977. OH. and A. 2952 (1972).REFERENCES 187 55. 2000. p. Jones. Busby. Matsushima. 68. G. 76. OH. Corrosion 13. 62.. Ives. Robertson. 54 (1960).. ASM International. H. in Physical Metallurgy of Stress Corrosion Fracture. 71. TX. National Association of Corrosion Engineers. Corrosion: Environments and Industries. R. 129–133. and R. New York. 2nd edition. 1 (1968). Bruemmer. Environment-induced crack growth processes in nickel-base alloys. 311. H. October 1988. Gordon. 68. 221. G. 551. L. Mater. J. 93 (1932). Staehle et al.. 1959. Am. Uhlig. Grube. Corros. D. Ford. Metall. Was. New York. National Association of Corrosion Engineers. Ref. Houston. ASM International. . Gangloff and M. 63. H. Corrosion 26. 1958. Jr. R. Ives. 289–290. ASM International. Revie. Reinhold. 215 (1970). Gough and D. Metall. I. Soc. 61. Soc. F. See also R. Andresen. and Protection. 28. TX. S. in Environment-Induced Cracking of Metals. Effect of irradiation on stress-corrosion cracking and corrosion in light water reactors. R. M. Corrosion in boiling water reactors. ASM Handbook. 354. 284. Uhlig. editors. Eng. 2006. 27 (II). C. and P. p. P. Trans. Eng. P. 158. C. 58. Trans. 13A. Schuetz and W. Lee and H. 3. Jones and S. Troiano. p. in ASM Handbook. Inst. Testing. W. pp. B. 13C. M. 1023. Y. Harwood et al. Marquez. Troiano. 59. 60. H. 347. 1959. 77. in Stress Corrosion Cracking and Hydrogen Embrittlement of Iron-Base Alloys. 2006. 212. Kim. B. H. 72. Proc. J. B. Proceedings of the First International Conference on Environment-Induced Cracking of Metals. J. B. Morlet. Wang. 69. Electrochem. in Internal Stresses and Fatigue In Metals. 57. pp.. Materials Park. Bastien. Wilde. J. Min. editor. Metals 49. J. Corrosion fatigue of type 4140 high strength steel. Phelps. Rhodin. 1990. Gangloff and M. 395 (1954). Corrosion. 90. TX. 92. Am. 215–244. Devereux et al. H. Appl. Ko. 27 (1981–1982). in Corrosion Fatigue. B.K. Proceedings of the First International Conference on Environment-Induced Cracking of Metals. 84. D. 81. A. Modelling stress corrosion cracking and corrosion fatigue. editors. Environment-Induced Cracking of Metals. Uhlig. A. Mech. O. 13A. p. 85. 449 (1968). J. W. GENERAL REFERENCES Stress-Corrosion Cracking C. thesis. Ives. 270. Department of Materials Science and Engineering. Chem. Corrosion: Fundamentals. American Society for Testing and Materials. R. Mech.. PA. M. R. Uhlig. Wang. 94. S. H. K. Wright. . 324. National Association of Corrosion Engineers. Gough. February 1975. Uhlig. 49 (1932). M. TX.. 1972. J. F. Corrosion: Fundamentals. Materials Park. Weckworth. Suresh. 88. 110 (1933). H. p. Martin. (published by ASME) 21. 24.T. J. H. Southampton. 1956. editors. ASM International. London. Trans. TX. M. Staehle. and Protection. Wright. Adey. 1990. L. Chemistry and Electrochemistry of Corrosion and Stress Corrosion Cracking: A Symposium Honoring the Contributions of R. STP 780. H. 658. P. 2nd edition. MA. H. Eng. Cambridge Solid State Science Series.K. 363–368. Hofmann. Inst. J. Cambridge. Section 5 in Simulation of Electrochemical Processes. Houston. B. H. 13A. pp. R. K. OH. Duquette and H. Fink and U. 2003. p. 1991. Institution of Mechanical Engineers. 87. p. Mech. in Uhlig’s Corrosion Handbook. 556 (1952–1953). Houston. Brebbia. 17 (1932). ASM International. Testing. 93. H. 89. 231. in ASM Handbook. Metals 61. M. Drozdowicz. Bauccio. W. TMS. 401 (1954). Am. and Protection. Uhlig. 79.I. Gadd. 2005. Trans.. Materials Park. 82.D. Vol. 1987. in International Conference on Fatigue of Metals. Uhlig. U. ASM International. W. editor. Glaeser and I. E. Revie. and M. Stress Corrosion Cracking Control Measures.. D. Ph. OH. 13. U. Warrendale. Cambridge. A. pp. Inst. V. 2001. WIT Press. 80. Metals 62. Metallk. OH. Testing. in ASM Handbook. in Materials Evaluations under Fretting Conditions. Houston. National Association of Corrosion Engineers. 86.-Z. October 1988. 2000. 86–105. 91. L. Materials Park. pp. p. P. in ASM Handbook. 882. Asphahani. Soc. Anorg. Wear 74. L. 1977. Philadelphia. Appl. I-Ming Feng and H. G. Jones. editor. 1981. G. 2003. 21. Cambridge University Press. and R. National Association of Corrosion Engineers. Tromp. Brown. Proc. editors.188 EFFECT OF STRESS 78. Soc. Metals 49. 1B. Vol. Allg. R. Y. DeGiorgi. Duquette and H. New York. 839 (1969). Z. Z. 210. 1030. Antler and M. Fatigue of Materials. 83. Vol. R. p. Wiley. OH. Materials Park. 1992. pp.. 2003. 13C. in ASM Handbook. Fretting wear. editor. D. Fatigue of Materials. 992–998. Wiley. Andresen. Materials Park. 2002. Revie. New York. West Conshohocken. and P. Vol. in ASM Handbook. Magnin. New York. B. OH. U. Elboujdaini. 1981. 221–232. Cambridge. Elliott III. Parkins. 2nd edition. Symposium on Effects of Radiation on Materials. 2000. . W. Revie. R. 13A. Waterhouse. W. 1129–1137. 2nd edition. Revie. Wiley. OH. Fretting Fatigue: Current Technology and Practices. New York. 1997. American Society for Testing and Materials. Nice. in Uhlig’s Corrosion Handbook. Materials Park. Kane. R. R. R. pp. Fretting Fatigue. PA. 2nd edition. Vol. in Uhlig’s Corrosion Handbook. Ohio. editors. Hydrogen-induced cracking and sulfide stress cracking. Materials Park. Test methods for wet H2S cracking. Y. 367–380. Hoeppner. Vol. R. R. 2000. and Protection. 1996. S. Suresh. Corrosion: Environments and Industries. Wiley. 2nd edition. 2004. 2000. in Uhlig’s Corrosion Handbook. Materials Evaluation under Fretting Conditions. Lubrication. New York. pp. Corrosion: Fundamentals. B. and Charles B. 2nd edition. 2000. 386–414. M. Vol. West Conshohocken. Hydrogen Damage B. Friction. ASM International. London. Corrosion: Environments and Industries. 191–204. Pergamon. Wang. STP 1367. 1981. and Wear Technology.-Z. Second International Conference on Corrosion-Deformation Interactions. editors. Oxford. Applied Science Publishers. Fatigue and Corrosion Fatigue S. Stress corrosion cracking. editor. OH. pp. B. R. 1972. M. Chandrasekaran. J. Elboujdaini. Busby. G. 242–256. pp. Allen et al. ASM International. Special Technical Publication 780. Waterhouse. Was. N. Corrosion fatigue. R. Waterhouse. Proceedings. editor. 22nd Int. Boston. Hydrogen damage. pp. Philadelphia. R. in ASM Handbook. Wiley. L. 1998. pp. ASM International. ASTM. Testing. The Institute of Materials. pp. Failure Analysis and Prevention. France. Craig. ASTM International. ASM International. pp. Vol. Materials Park. editor. 205–220. 11. CDI ’96. R. 18. STP 1475. V.K. 2006. Effect of irradiation on stress-corrosion cracking and corrosion in light water reactors. Effects of Radiation T. 922–940. 13C. London. 2006. 2006. 2000. in Uhlig’s Corrosion Handbook. in ASM Handbook. PA.. Corrosion in petroleum refining and petrochemical operations. Fretting Corrosion. Fretting Corrosion David W. Fretting wear failures.GENER AL REFERENCES 189 T. editor. W. ASM International. Cambridge University Press. Corrosion-Deformation Interactions. Revie. in ASM Handbook. W. editor. 2.2 mm/s. . The residual energy of a cold-worked steel as determined calorimetrically is 21 J/g.006 V. New York. (a) What effect on SCC of 18–8 results from coupling to nickel (φcorr = −0.8. what does your answer indicate with regard to the electrochemical mechanism of crack growth? 2. below what value of pH at 130°C does the cathode potential of 18–8 become more noble than φcritical? Answers to Problems 1.190 EFFECT OF STRESS PROBLEMS 1. 3. To what corrosion current density does this rate correspond? If this rate is typical.145 V (S.E. Engell and Bäumel (T. Physical Metallurgy of Stress Corrosion Fracture.). 354) reported that cracks in iron stressed in boiling calcium nitrate solution propagate discontinuously at the rate of 0. Calculate the increased tendency in volts for iron to corrode when cold worked.H. p. N.) 3. 545 A/cm2. (Assume that the entropy change is negligible. The critical potential for stress-corrosion cracking of moderately cold-worked 18–8 stainless steel in deaerated MgCl2 solution at 130°C is −0. Interscience.18 V)? (d) On adding HCl? Why? (c) Under (b). 1959. Rhodin. (c) 1. 0. airborne particles of any composition).1 INTRODUCTION In the absence of moisture. Also. lead (Pb). not only on the moisture content of air. Incidence of corrosion by the atmosphere depends. steel parts abandoned in the desert remain bright and tarnish-free for long periods of time. nitrogen dioxide (NO2). Improvements in air quality help to mitigate atmospheric corrosion. carbon monoxide (CO). as will be discussed in this chapter. in climates below the freezing point of water or of aqueous condensates on the metal surface. by R. iron exposed to the atmosphere corrodes at a negligible rate. which have been established for ozone (O3). but also on the particulate matter content and gaseous impurities that favor condensation of moisture on the metal surface. For example. rusting is negligible. 191 . however. Ambient air quality in the United States has improved dramatically since the Clean Air Act was enacted in 1970. Ice is a poor electrolytic conductor. The Environmental Protection Agency (EPA) monitors air quality and compares the data with the National Ambient Air Quality Standards (NAAQS).9 ATMOSPHERIC CORROSION 9. Corrosion and Corrosion Control. sulfur dioxide (SO2). Winston Revie and Herbert H. the corrosion process cannot proceed without an electrolyte. Inc. Uhlig Copyright © 2008 John Wiley & Sons. hence. and particulate matter (PM. 3 CORROSION-PRODUCT FILMS Corrosion-product films formed in the atmosphere tend to be protective. and various suspended salts are encountered. galvanized iron performs well in rural atmospheres. with large differences in corrosivity. and organic acids. This is true to a lesser extent of pure iron. compared to the copper-bearing or low-alloy steels. Recognition of marked differences in corrosivity has made it convenient to divide atmospheres into types. The major types are marine. and contaminants. Atmospheric Corrosive Gases in Outdoor Urban Environments [1] Gas H2S SO2 NH3 HCl NO2 O3 RCOOH Approximate Concentration Range (ppbv) 0.5–100 0.1) [3]. There are also subdivisions. atmospheric corrosion rates vary markedly around the world. specimens of steel located 24 m (80 ft) from the ocean. corroded about 12 times more rapidly than similar specimens located 240 m (800 ft) from the ocean [2].192 ATMOSPHERIC CORROSION 9. for which the rate is relatively high. hence.9–600 0. relative corrosion behavior of metals changes with location. On the other hand. air is laden with increasing amounts of sea salt. Acids that can form from these gases include sulfuric acid (H2SO4). where salt water spray is frequent.2–700 3–1000 1–90 0. but is relatively less resistant to industrial atmospheres. Rust films on the latter steels tend TABLE 9.5–300 0. urban. nitric acid (HNO3). Approaching the seacoast. At industrial areas. elsewhere because a protective film of lead sulfate forms on the surface. NH3. NO2. SO2. or better than. such as wet and dry tropical. specimens exposed to a marine atmosphere corrode at greatly differing rates depending on proximity to the ocean. tropical. 9. such as formic acid (HCOOH) and acetic acid (CH3COOH). lead performs in an industrial atmosphere at least as well as. At Kure Beach.5–30 .1. hence. temperature. 9. North Carolina. A metal that resists corrosion in one atmosphere may lack effective corrosion resistance elsewhere. for example. in particular NaCl. industrial. Also. and rural.1 [1]. that is. H2S. arctic. Approximate concentration ranges of corrosive gases in urban areas are presented in Table 9. the corrosion rate decreases with time (Fig. which are more resistant.2 TYPES OF ATMOSPHERES Atmospheres vary considerably with respect to moisture. weathering steels have very low values of n. Larrabee.1) where p can be expressed as specimen weight loss (g/m2). and k and n are constants that depend on the metal and on the atmospheric conditions (i.1 follow the relation p = kt n (9. Data of Fig. Wiley. editor.CORROSION-PRODUC T FILMS 193 Figure 9. Within the experimental error of such determinations. H. and 55% Al–Zn coatings on steel [5–7].2% C steel.1% Cr and 0.2) [4]. that is. Atmospheric corrosion of steels as a function of time in an industrial environment (C. . climatic and pollution factors) at the test site. n = 1. 124) to be compact and adherent. In addition to carbon steels. The data in Table 9.2 also show the beneficial effect of alloying elements.5 to 1. in the low-alloy steel compared to the 0. aluminized.1. in Corrosion Handbook. In comparison with carbon steels. as can be seen from data obtained by the American Society for Testing and Materials for various metals exposed 10 or 20 years to several atmospheres (Table 9. the corrosion product film is not protective. Uhlig. applies for a linear rate law. the rate for a 20-year period is about the same as that for a 10-year period.. 9. The corrosion rate eventually reaches steady state and usually changes very little on further exposure. Values of n typically range from about 0.e. p. 1948. New York. This is characteristic of other metals as well. primarily 1.4% Cu. or as specimen penetration (μm). whereas on pure iron they are a powdery loose product. The constant. during time t (years). this relation is also found to apply to atmospheric test data for galvanized. H. 4 μm/year. 0. Larrabee. A value of n = 0.050 0.029 0. p/t.009 0. 1. Extrapolation up to 20–30 years from tests of 4 years is possible with reasonable confidence [9].03% Ni.052 0.04% S. 0./year = 0.5 corresponds to a parabolic rate law. Over the longer term.226 0.043 — — 20 Years 0.09 1 mil/year = 0. 1% Mn (Monel) Zinc (99. Corrosion 9.112 0.006 0. 32% Cu.1% C.2% P.032 0.02% P. values cited are one-half reduction of specimen thickness [C.017 0.018 0.019 0.193 0.023 0.005 0.004 0.021 0.063 0.4% Cu) a b Atmosphere La Jolla. This bilogarithmic law can be very useful in predicting long-term atmospheric corrosion damage based on tests of shorter duration.013 — 0.1).007 0. can be calculated as log ( p ) = A + B − 1 log t t ( ) (9.042 — — 20 Years 0. 0. New Jersey (near New York City).05% Cu. Kearney.2. the linear bilogarithmic law may be expressed as log p = A + B log t (9. PA (Rural) 10 Years 0. 259 (1953)]. 0.05% S.194 ATMOSPHERIC CORROSION TABLE 9.006 0.48 0.017 0. From (9.069 0.028 0.068 — — State College.02% Cr) Low-alloy steelb (0.025 0.128 0.001 in.015 0.2% C steelb (0.and 20-Year Exposure Times. 1 usually less than 2 [8].053 0. Average Atmospheric Corrosion Rates of Various Metals for 10. 0.2) where A = log k and B = n.144 0.052 0. CA (Marine) 10 Years 0.0%) 0. The mean corrosion rate. changes in the environment are likely to be more significant than deviations from the model.047 0.047 0.054 0.006 0.0254 mm/year = 25. 0. P.016 0. 0.02% Ni.202 0.1% Cr.069 — — 20 Years 0.091 0. 2% Fe.034 0.001 0.3) .9%) Zinc (99.007 0.218 — — 0. The atmospheric behavior of a specific metal at a specific location can be described using the two parameters A and B.062 0. mils/yeara (American Society for Testing and Materials) [4] Metal New York City (Urban Industrial) 10 Years Aluminum Copper Lead Tin Nickel 65% Ni.044 0. 0.003 0. Soil data for steel are averaged for 44 soils.15 2. compared to a partially sheltered exposure. Comparison of Atmospheric Corrosion Rates with Average Rates in Seawater and in Soilsa Environment Steel Rural atmosphere Marine atmosphere Industrial atmosphere Seawater Soil 1 Corrosion Rate (gmd) Zinc 0. Uhlig. but where rain cannot wash the surface. 29 soils. Romanoff.4 FACTORS INFLUENCING CORROSIVITY OF THE ATMOSPHERE In all except the most corrosive atmospheres. Circ. for copper. the average corrosion rates of metals are generally lower when exposed to air than when exposed to natural waters or to soils. This advantage presumably would not extend to uncontaminated atmospheres.029 0.017 0. H. 1948.07 — 0. and so FeSO4 may oxidize directly to Fe2O3. from Symposium on Atmospheric Exposure Tests on Non-Ferrous Metals. may corrode more rapidly than specimens fully exposed. 11-year exposure. 579. 10-year exposure. zinc. H.29 0. 8-year exposure–from Underground Corrosion. absorbed by rust will continue to accelerate corrosion. D. Corrosion 9. for example. This fact is illustrated by the data of Table 9.3% copper steel.031 0. 1946.5 a Atmospheric tests on 0. 259 (1953).3 Copper 0. Atmospheric rates for zinc and copper. M.C. .5 0. 7 2-year exposure.014 0. Direct exposure of a metal to rain may. from C.3 for steel.5) Intermediate formation of ferric sulfate has not been demonstrated. for zinc. and between log dp/dt and log t. Larrabee. 1957. Wiley. ASTM. there are also linear relationships between log p/t and log t. be beneficial. Seawater data from Corrosion Handbook. 12 soils. National Bureau of Standards.FAC TORS INFLUENCING CORROSIVIT Y OF THE ATMOSPHERE 195 and the instantaneous corrosion rate. 12-year exposure. Nevertheless.0 0. 9. dp/dt. editor.10 1.8 0.032 0. rust contaminated in this way catalyzes the formation of more rust. can be expressed as log ( dp ) = A + log B + (B − 1) log t dt (9. perhaps by means of the cycle H 2SO4 + O2 O2 + H 2SO4 H 2O 1 3 1 2 2 Fe ⎯⎯⎯⎯⎯ FeSO4 ⎯4 → ⎯⎯⎯⎯⎯ Fe2 (SO4)3 ⎯2 → Fe2O3 + H 2SO4 → ⎯⎯ 2 2 2 1 1 1 1 (9. New York. The reason for this is that sulfuric acid. Metal surfaces located where they become wet or retain moisture. and copper in three atmospheres compared to average rates in seawater and a TABLE 9. therefore.3.4) When there is a linear relationship between log p and log t. Washington. organic chemicals. Microclimatic conditions (e. and moisture (critical humidity). It was estimated that more than 35.1 Particulate Matter Particulate matter (PM) is the term for a mixture of solid particles and liquid droplets in air. who exposed specimens of iron to an indoor atmosphere.g. atmospheric corrosion of passive metals (e.000 kg of dust per km2 (100 tons per square mile) settled every month over an industrial city [15]. Specific factors influencing the corrosivity of atmospheres are particulate matter. the engineer or architect would be well advised to review air-quality data for a corrosion assessment at the design stage. and soil or dust particles. the Environmental Protection Agency (EPA) compiles data on ambient outdoor-air quality from information obtained at air-monitoring stations. Industrial atmospheres carry suspended particles of carbon and carbon compounds. PM consists of several components. metal oxides. Because of the importance of trace constituents on atmospheric corrosion behavior. reaching 1000 mg/m3 or more [14]. In the United States in 2006. the unscreened specimens showed rust and appreciable gain in weight. revised in 2006. some specimens being entirely enclosed by a cage of single-thickness muslin measuring several inches larger in size than the specimen. 9. are set at 150 μg/m3 as the 24-h average for coarse particulates (2. and other . gases in the atmosphere. The importance of atmospheric dust was established in the early experiments of Vernon [11]. whereas the muslin-screened specimens showed no rust whatsoever and had gained weight only slightly. with higher values for an industrial atmosphere. particulate matter influences the corrosion rate in an important way.5-μm diameter) [12]. Particulate matter is found near roads and some industries. sunlight. proximity to a highway where deicing salts are used) can be different from macroclimates and should be considered. In the United States. the average concentration of fine particulates amounted to about 13 μg/m3 [13]. After several months. in smoke and haze. In addition. (NH4)2SO4. In the 1950s. east versus west exposure. where concentrations of the more common gases found in trace amounts in the atmosphere are measured continuously. Particulate matter can be a primary contaminant of many atmospheres. H2SO4. aluminum and stainless steels) tends to be more uniform and with less marked pitting than corrosion in waters or soils.. NaCl.196 ATMOSPHERIC CORROSION variety of soils. including acids.to 10-μm diameter) and are set at 35 μg/m3 as the 24-h average for fine particulates (≤2. wind direction.4. it can be directly emitted from sources such as forest fires. it was estimated that the average city air contained about 2 mg/m3.g. metals.5. In contact with metallic surfaces.. The NAAQSs for particulate matter. but form particles in the atmosphere. Particle pollution is controlled by reducing directly emitted particles and by reducing emissions of pollutants that are gases when emitted. to avoid disappointment at the operational stage. and it can be formed when gases emitted from power plants and automobiles react in the air. Variation of average sulfur dioxide content of New York City air with time of year [17]. Steel specimens rust in a carbon-dioxide-free atmosphere as readily as in the normal atmosphere.2 Gases in the Atmosphere The carbon dioxide normally present in air neither initiates nor accelerates corrosion. probably by favoring a more protective rust film. Since fuel consumption is higher in winter. 9. they form an electrolyte on the metal surface. it was estimated that about 1. In New York City in the 1950s. oil.2.2). sulfur dioxide contamination is also higher (Fig. and chloride ions (Cl−). Early experiments by Vernon showed that the normal carbon dioxide content of air actually decreases corrosion [16]. Sulfur dioxide originates predominantly from the burning of coal. and ozone is a powerful oxidizing agent. ozone (O3). Important atmospheric pollutants are sulfur dioxide (SO2). This amount was equivalent to burdening the atmosphere with an average of 6300 tons of H2SO4 every day and has been reduced by limiting the allowable sulfur content of fuels burned within city limits. is less apt to cause corrosion than is air heavily laden with dust. depending on magnitude and direction of the prevailing winds. or. Marine atmospheres contain salt particles that may be carried many miles inland. particularly if the dust consists of water-soluble particles or of particles on which H2SO4 is adsorbed. and gasoline. These substances. because of their hygroscopic nature. Sulfur dioxide and nitrogen oxides form corrosive acids.4. nitrogen oxides (NOx). combined with moisture. initiate corrosion by forming galvanic or differential aeration cells. therefore. The tarnish films are composed of Ag2S and a mixture of Cu2S + CuS + Cu2O. 9.FAC TORS INFLUENCING CORROSIVIT Y OF THE ATMOSPHERE 197 salts. A trace amount of hydrogen sulfide in contaminated atmospheres causes the observed tarnish of silver and may also cause tarnish of copper. respectively. It is also obvious from this cause that the average sulfur dioxide content of the air (and Figure 9.5 million tons of sulfur dioxide were produced every year from burning coal and oil [17]. . Dust-free air. 003 5–10 8–16 0. In the industrial atmosphere of Altoona. Copper exposed to industrial atmospheres forms a protective green-colored corrosion product called a patina. and J. galvanized steel sheets [0.2. D. and iron. .005 0. and transportation sources contribute most of the remainder.020 0. 9.001 0.002 corresponding corrosivity) falls off with distance from the center of an industrial city. but remain reddish-brown on the opposite side. Alley. and stainless steels. (Table 9. 17) City Distance Miles Kilometers Detroit Philadelphia–Camden Pittsburgh St. In comparison with the data in Fig. A copper-covered church steeple on the outskirts of a town may develop such a green patina on the side facing prevailing winds from the city.023 0.006 0.381 kg zinc per m2.4). A high sulfuric acid content of industrial and urban atmospheres shortens the life of metal structures (see Tables 9. Meller.030 0. forming a protective basic copper sulfate film.004 0.3).001 0. CuSO4·3Cu(OH)2. 9.010 0.6 years [19].C. 1. on which the corresponding films are less protective. cadmium. Near the seacoast. composed in part of basic copper chloride. where sulfuric acid is less readily available.4.S. the engineering and industrial responses to the U. Pennsylvania. The EPA estimates that ambient SO2 levels have decreased by more than 50% since 1983. rust appeared only after 14.2 and 9.C. such as lead.029 0.012 0.1 mil thick)] began to rust after 2. Sherrick. Variation of SO2 Content of Air with Distance from Center of City (H.001 0. J. is more resistant than nickel or 70% Ni–Cu alloy.001 Parts per Million 10–15 15–20 20–25 25–30 16–24 24–32 32–40 40–48 0.012 0. Louis Washington. Pennsylvania.2.111 0. whereas in the rural atmosphere of State College. quoted in Ref. 0–5 0–8 0. The effect is most pronounced for metals that are not particularly resistant to sulfuric acid. such as Washington.009 0. such as zinc. About 10% of SO2 emissions result from industrial processes.012 0.198 ATMOSPHERIC CORROSION TABLE 9. Clean Air Act have resulted in a major decline in the levels of air pollutants.060 0.048 0.015 0.004 0. a similar patina forms.016 0. Copper.021 0.030 0.01 ppm. the annual average concentration of SO2 in New York City air in 2006 was about 0.028 mm thick (1.018 0. and the NAAQS for SO2 is 0.4 years. composed mostly of basic copper sulfate.018 0. D.018 0. aluminum. and it is clear that this effect is not as pronounced in the case of a residential city.014 0. It is less pronounced for metals that are more resistant to dilute sulfuric acid. Although the SO2 levels in the air of many major centers of population around the world may be similar to the data in Fig.25 oz zinc per ft2.03 ppm (annual arithmetic mean) [18]. nickel. 0. but is sensitive to sulfuric acid of industrial atmospheres (Table 9. The cell B is placed about 1 m above ground level with the surface inclined at 45 °. 9. it is apparent that. forming a surface tarnish composed of basic nickel sulfate. copper. 9. designed the device shown in Fig. Calibration using data from weight-loss experiments at the same site shows that the electrochemical technique is suitable for estimating short-term variations in corrosion rates [26]. Pennsylvania (Table 9. Copper did not fail. Nickel is quite resistant to marine atmospheres. and about 20 times higher than in the rural atmosphere of State College. An important factor determining susceptibility to atmospheric corrosion of a metal in a particular environment is the percentage of time that the critical humidity is exceeded [23]. California. This period of time is called the “time of wetness.2). Surface moisture from either precipitation or condensation is the cell electrolyte. nickel. the relative humidity must be reduced to values much lower than 100% in order to ensure that no water condenses on the surface. In very early studies. . Experimental values for the critical relative humidity are found to fall. in an uncontaminated atmosphere at constant temperature.2.” It is determined by measuring the potential between a corroding metal specimen and a platinum electrode [23. Vernon discovered that a critical relative humidity exists below which corrosion is negligible [21].2). An electronic integrator automatically integrates cell currents over extended periods of time.4. appreciable corrosion of a pure metal surface would not be expected at any value of relative humidity below 100%. because of normal temperature fluctuations (relative humidity increases on decrease of temperature) and because of hygroscopic impurities in the atmosphere or in the metal itself.FAC TORS INFLUENCING CORROSIVIT Y OF THE ATMOSPHERE 199 Industrial atmospheres can cause S. however. Practically.3. but copper alloys containing >20% Zn failed on exposures for up to 8 years [20]. 9. between 50% and 70% for steel. 24].3). 26]. Typical corrosion behavior of iron as a function of relative humidity of the atmosphere is shown in Fig.C. mostly accounted for by presence of nitrogen oxides (see Section 20. Corrosion in the industrial atmosphere of New York City is about 30 times higher than in the marine atmosphere of La Jolla. in general. a critical humidity may not exist [22]. To estimate atmospheric corrosion rates.4 [25. Reviews have been prepared summarizing the atmospheric corrosion standards and testing procedures of the American Society for Testing and Materials [27–30].3 Moisture (Critical Humidity) From previous discussions. and zinc.C of copper-base alloys. Atmospheric corrosion testing is important to the suppliers of metals and to the engineers and architects who use metals under atmospheric conditions. In a complex or severely polluted atmosphere. Kucera et al. 26]. (Copyright ASTM INTERNATIONAL. General arrangement of electrochemical device for measurement of atmospheric corrosion: A is a zero resistance ammeter. C is an external emf.3.200 ATMOSPHERIC CORROSION Figure 9. showing critical humidity [21]. [25. B is an electrochemical cell with (a) electrodes and (b) insulators. the circuit of which is shown on the right. 55 days’ exposure. Corrosion of iron in air containing 0. Reprinted with permission. Figure 9.01% SO2.4.) . 1).. as is apparent from their satisfactory use over many years as .09% P. These are used for construction of buildings and bridges and for architectural trim. 0. Corrosion rates of structural steels in tropical atmospheres (e. in general. This protective measure is effective except perhaps when corrosion is caused by acid vapors from nearby unseasoned wood or by certain volatile constituents of adjacent plastics. Such steels do not have any advantage when buried in soil or totally immersed in water.. If the presence of unusually hygroscopic dust or other surface impurities is suspected. [33] confirmed that the typical inner adherent rust layer consists mostly of δ-FeOOH. Copper additions are more effective in temperate climates than in tropical marine regions. and rural atmospheres. thereby saving appreciable amounts in maintenance costs over the life of the structure. These are discussed in Chapter 17. Heating air—or. 0. The usefulness of low-alloy steels to resist atmospheric corrosion through formation of protective rust films has resulted in the development of weathering steels. and chromium are particularly effective in reducing atmospheric corrosion. Use of Organic. Use of Alloys. copper. 0. Panama) were found.4% Mn.4% Cu. alternate drying and wetting favors its protective qualities. because the corresponding rust films formed under continuously wet conditions are no more protective than those formed on carbon steels. nickel.g. It consists of amorphous δ-FeOOH. Keiser et al. or paints.REMEDIAL MEASURES 201 9. better still.5). and 16. Inorganic. to be about two or more times higher than in temperate atmospheres (e. 3. phosphorus. Coatings are discussed in Chapters 14. 0.3% Ni. 2. North Carolina) mainly because of the higher relative humidity and higher average temperatures. Kure Beach.g.2. [32] proposed that an inner cohesive protective rust film is formed on low-alloy steels after long atmospheric exposure (industrial or urban).3. 15. chromium and nickel additions combined with copper and phosphorus are effective in both locations (Table 9. A typical commercial composition is as follows: 0. the value should be reduced still further. reducing the moisture content—can reduce relative humidity. When alloyed with steel in small concentrations.09% C. 4.8% Cr. Their more noble corrosion potentials compared to carbon steels may make them useful in certain galvanic couples (see Section 7. It is only by a process of alternate wetting and drying that the rust continues to be protective. Reduction of Relative Humidity. the formation of which is catalyzed by copper and phosphorus on the steel surface. urban.5 REMEDIAL MEASURES 1. Stainless steels and aluminum resist tarnish in industrial. or Metallic Coatings. without the necessity of painting. Misawa et al. Lowering the relative humidity to 50% suffices in many cases. 0. Use of Vapor-Phase Inhibitors and Slushing Compounds. 2 0. 2003. pp. in ASM Handbook.2 0.2 0. Revie. Vol. Pearson.0 1 Cr 1.1 0.1 0. 495 (1951). in Uhlig’s Corrosion Handbook. 276–284.15 0. Corrosion: Fundamentals. Legault and V. P.2 0.24 0.19 20. 2nd edition.076 9.2 0. REFERENCES 1. Soc.11 0. 433 (1978).008 0.52 0. 1955. 5.6 9.2 0. Testing Mater. 5% Fe.45 0. C. PA.1 0. 1946. 51.5-year exposure.08 0. American Society for Testing Materials. 13 (1979).14 0.09 0.8 2. Philadelphia.5 7.7 0.4 0. Larrabee and S. L.07 0. New York.6 3. Wiley. 15% Cr. % P Cu Other Loss of Thickness mm mils Industrial Atmosphere (Kearney. Corrosion 34. L. Atmospheric corrosion. M. Tullmin and P. 8. in Proceedings.7 3.4 0. 1962. Mater. Proc. 6. Effect of Low-Alloy Components on Atmospheric Corrosion of Commercial Steel Sheet (Eight-Year Exposure) Steel C Composition. Townsend and J. K. Atmospheric Corrosion. Also previous Symposium. 7.202 ATMOSPHERIC CORROSION TABLE 9.01 0. Testing. LaQue. Wiley. R. NJ) [19] Carbon Copper bearing Low chromium Low nickel 0.4 1. editor.2 0. Symposium on Atmospheric Corrosion of Non-Ferrous Metals.02 0. C. Perf.1 0.02 0.5 Ni Temperate Marine Atmospherea(Kure Beach. Zoccola.2 0. F.5 Ni Tropical Marine Atmosphere (Panama Canal Zone) [31] Carbon Copper bearing Low chromium Low nickel a 0. 4% W) is very resistant to tarnish in marine atmospheres.23 0.069 0.051 8. Butterworths. 13A.03 0. W. Kane. R. 41.2 Cr 2. . Veleva and R.20 0. 17% Mo. and Protection. OH. Materials Park.25 0.02 0. D. NC) [19] Carbon Copper bearing Low chromium Low nickel 0. STP 175. Graedel. Am. 315–317. London.5. making it a useful alloy for reflectors on board ship.048 0.7 0. p. 4.1 0.04 0. First International Congress On Metallic Corrosion.5 5. ASM International. 3.24 0.1 7. R. 18(10). 203–204. 2000.9 2. pp.004 0.02 0. architectural trim of buildings.2 0. Coburn. 2. pp. Roberge.03 0.4 17.0 1 Cr 1. New York.3 0.1 Ni 0. Leygraf and T. Hastelloy C (54% Ni.0 4. H. 2000. 2nd edition. R. Jacobs. 1517 (1956). Philadelphia. J. p. Corros. STP 646.gov/air/airtrends/pm. 435t (1958). ASTM International. Mansfeld. 1978. Kodama. Electrochem. 2005. S. Vol. Suetaka. 7. 29. J. 2005. W. Werks. 1459 (1979). 48. 492 (1957). Vol. 1982. Ailor. P. and R. V. 279 (1975). p. 24.epa. W. 1971. New York. GENERAL REFERENCES W. 1982. Corrosion 31. 255 (1931) and 31. West Conshohocken. p. W. Outdoor atmospheres. Kyuno. editor. http://www. Ailor. C. 1668 (1935). J.html 13. Trans. editor. pp. P. 31. Sinclair. Eng. 26. ASM International. London. Wiley. The linear bilogarithmic law for atmospheric corrosion. ASM International. p. B. 12. Sci. 35 (1971). B. in ASM Handbook. Sixth European Congress On Metallic Corrosion. D. editor. 24. 1974. p. Indoor atmospheres. Vernon. Griffin. Corrosion in marine atmospheres. Faraday Soc. T. Baboian. January 4.epa. 33. in Ref. Brown. News. 259 (1953). W. 17. Corros. Chem. Heidersbach.gov/oar/particlepollution/naaqsrev2006. Vernon. 42–60. 563 (1980). Faraday Soc. London. Soc.. Misawa. New York. V. H. 27. 129. Kutzelnigg. R. Forgeson. pp. Keiser. 2003. Popplewell and T. Shimodaira. . W. in Corrosion Tests and Standards: Application and Interpretation.epa. 107–121. 189. T. Philadelphia. 20. 22.html 14. and R. Wiley. Materials Park. in ASM Handbook. In Eurocorr ’77. 2006. in Ref. C. editor. Vernon. Atmospheric Corrosion. 13B. Materials Park. 18. W. West Conshohocken. 21. 23. 127. Chem. Electrochem. 31 (1933). Mattsson. Ailor. Sci. 10. 126. 13C. in Atmospheric Corrosion. and A. 16. 113 (1927). STP 558. Ailor. 113 (1927). F. 349–361. S. Rice. and S. Kucera and M. 27. in Atmospheric Corrosion. 251 (1983). 11. Collin. M. OH. 31. 29. 15. Corrosion: Materials. 23. American Society for Testing and Materials. Wiley. Lawson. T. http://www. J. 2nd edition. Eng. Corrosion: Environments and Industries. pp. H. Corrosion 9. Sereda. pp. D. Alexander. H. p. OH. Gearing. in Corrosion In Natural Environments. 11. 28. 7. p. Pourbaix. in Atmospheric Factors Affecting the Corrosion of Engineering Metals. Baboian. editor. 23. A. R. W. Trans. 1977.gov/air/airtrends/sulfur. Soc. P. Kucera and E. Dean. 239. 139. 23. Larrabee. 343–348. New York. Korros. editor. 1668 (1935). p. C.GENER AL REFERENCES 203 9. American Society for Testing and Materials. 8. ASTM International. L. Tremoureux. in Corrosion Tests and Standards: Application and Interpretation. 23. 64. 1982. Trans. ibid. Phipps. 32. 30. Corrosion 14. Trans. Faraday Soc. Coburn. 25. 195.html 19. Southwell. Greenburg and M. Society of Chemical Industry. 1977. http://www. Veleva and R. Roberge. OH. M. R. Wiley. and 40 years of exposure. . Vol. Corrosion inspection and Monitoring. in terms of penetration per year. and 40 years of exposure. assuming that environmental conditions do not change significantly during this time period. editor. and Protection. editor. Testing. Leygraf. L. Corrosion: Fundamentals. Roberge.204 ATMOSPHERIC CORROSION C. and C. at 5. Assuming the bilogarithmic law applies. pp. (c) The instantaneous corrosion rate. Corrosion of metal artifacts displayed in outdoor environments. over a 40-year period. Roberge. S. Atmospheric Corrosion. (c) The instantaneous corrosion rate. Wiley-Interscience. in Corrosion Mechanisms in Theory and Practice. Kane. 2nd edition. 2007. D. the penetration is 130 μm. 2002. PROBLEMS 1. Johnson. 2006. Wiley. R. Leygraf. New York. 13A. 289–305. Wiley Series In Corrosion. P. 2000. A carbon steel is exposed to a highly polluted. Selwyn and P. C. Qiu. Henriquez. 301–311.2). and after 4 years. New York. M. Revie. Hedberg. 196–209. series editor. calculate the following: (a) The constants A and B in Eq. over a 40-year period. Marcel Dekker. the penetration is 190 μm.2). After one year of exposure the penetration is 25 μm. J. Leygraf and T. assuming that environmental conditions do not change significantly during this time period. Marcus. in ASM Handbook. ASM International. (9. Corrosion 63(8). Revie. W. W. Materials Park. H. Atmospheric corrosion. P. 715 (2007). Atmospheric corrosion. After one year of exposure. Materials Park. 25. (b) The average corrosion rate. R. at 5. (9. 305–321. Gil. pp. in terms of penetration per year. in Uhlig’s Corrosion Handbook. pp. Corrosion: Environments and Industries. R. C. A weathering steel is exposed to the same environment as the carbon steel in Problem 1. New York. pp. Assuming the bilogarithmic law applies. 2nd edition. 2003. OH. New York. (b) The average corrosion rate. Graedel. Atmospheric corrosion. L. industrial-marine environment. in ASM Handbook. Vol. Molecular in situ studies of atmospheric corrosion. 13C. 25. 2000. in terms of penetration per year. J. Tullmin and P. P. ASM International. in terms of penetration per year. R. 2. calculate the following: (a) The constants A and B in Eq. and after four years the penetration is 28 μm. such as a pipeline [2]. retarding evaporation. Inc. Fluid flow through soil is controlled by the permeability of the soil. Capillary action in fine-grained soil can draw water up. Winston Revie and Herbert H. Corrosion and Corrosion Control. whereas bulk water can flow through porous soil. refineries. for example. and sometimes replaced.10 CORROSION IN SOILS 10.3 million miles) of pipelines crossing the United States. fine-grained soils high in clay. depends on the size distribution of the solid particles in the soil. keeping the soil water-saturated. in turn.7 million kilometers (2. maintained.1 INTRODUCTION There are more than 3. allows good drainage and access of atmospheric oxygen to a depth greater than. Underground corrosion is of major importance and results in a significant portion of pipeline failures [1]. In soils. Because of corrosion. preventing drainage. and ports to customers [1]. for example. Uhlig Copyright © 2008 John Wiley & Sons. water and gas occupy the spaces between solid particles. these pipelines must be regularly inspected. Some of this water is bound to mineral surfaces. 205 . and restricting oxygen access from the atmosphere to a buried structure. and these spaces can constitute as much as half the volume of dry soil. which. transporting natural gas and hazardous liquids from sources such as wells. Coarse-grained sand. by R. however. For example. for example. Galvanic effects of coupling iron or steel of one composition to iron or steel of a different composition are important. because they are under conditions of total immersion (Section 7. In most soils. groundwaters in tropical climates tend to be very acidic. is subject to graphitic corrosion. sulfate-reducing bacteria may be active. but not elsewhere. some clay soils buffer the groundwater pH [2]. a copper-bearing steel. a mild steel. Hence. (2) electrical conductivity or resistivity. 10. . pH. in some respects. Another factor to be considered is that. because of specific differences in soil composition. Both the soil and the climate influence the groundwater composition. cold working or heat treatment do not affect the rate. moisture content. for example. The situation is more complex. Minor composition changes and structure of a steel. A metal may perform satisfactorily in some parts of the country. Corrosion rates underground have been measured using the Stern–Geary linear polarization method (Section 5. Gray cast iron in soils. In addition. in poorly aerated soils containing sulfates. including depolarizers or inhibitors. and (5) pH. A porous soil may retain moisture over a longer period of time or may allow optimum aeration. For example. in assessing the corrosion rates of footings of galvanized-steel towers used to support power lines [5]. the behavior on total immersion in water.206 CORROSION IN SOILS The electrochemical corrosion processes that take place on metal surfaces in soils occur in the groundwater that is in contact with the corroding structure. a cast iron water pipe may last 50 years in New England soil. vary to a marked degree with the type of soil. these organisms often produce the highest corrosion rates normally experienced in any soil. The influence of soil on groundwater composition is discussed in detail elsewhere [3]. (4) moisture. The former method has been useful.2 FACTORS AFFECTING THE CORROSIVITY OF SOILS Among the factors that control corrosivity of a given soil are (1) porosity (aeration). Groundwater in desert regions can be high in chloride and very corrosive. and wrought iron are found to corrode at approximately the same rate in any given soil. Localized corrosion of this kind is obviously more damaging to a pipeline than a higher overall corrosion rate occurring more uniformly. and so on. corrosion in soils resembles atmospheric corrosion in that observed rates. In other respects.1). particularly if not well-aerated. On the other hand. The corrosion behavior of iron and steel buried in the soil approximates. observed corrosion takes the form of deep pitting. a low-alloy steel. because corrosion products formed in an aerated soil may be more protective than those formed in an unaerated soil.8) as well as weight loss [4].3. and both factors tend to increase the initial corrosion rate. but only 20 years in the more corrosive soil of southern California. Each of these variables may affect the anodic and cathodic polarization characteristics of a metal in a soil. (3) dissolved salts. although usually higher than in the atmosphere. are not important to corrosion resistance.2. as well as in water. Macrogalvanic cells or “longline” currents established by oxygen concentration differences. the beneficial effect of aeration extends to soils that harbor sulfate-reducing bacteria because these bacteria become dormant in the presence of dissolved oxygen. it was concluded that the corrosivity of soil towards ferrous metals could be estimated by measuring the resistivity of the soil and the potential of a platinum electrode in the soil with respect to a saturated calomel reference electrode [7]. A soil containing organic acids derived from humus is relatively corrosive to steel. A few typical corrosion rates averaged for . H.) below the surface showed greater corrosion at greater depth for five soils.). confirmation was obtained through 5-year tests in England [8]. is. They showed similarity in corrosion rates in a given soil for various kinds of iron and steels. or by dissimilar surfaces on the metal become more important when electrical conductivity of the soil is high. zinc.H. These tests continued until 1955 and now constitute the most important source of information on soil corrosion available [6]. make the soil very corrosive. lead. apart. Anodes and cathodes may be thousands of feet. but showed the reverse trend for two soils. or <0. But conductivity alone is not a sufficient index of corrosivity.43 V if the soil is clay] were considered to provide a suitable environment for sulfate-reducing bacteria and to be aggressive. and copper. Borderline cases by these two criteria were resolved by considering that a water content of more than 20% causes a soil to be aggressive. anodic and cathodic polarization characteristics of a corroding metal in the soil are also a factor. but also indirectly through the influence of oxygen reacting with and decreasing the concentration of the organic complexing agents or depolarizers naturally present in some soils. Logan of the National Bureau of Standards (now the National Institute of Standards and Technology). A poorly conducting soil.BUREAU OF STANDARDS TESTS 207 Aeration of soils may affect corrosion not only by the direct action of oxygen in forming protective films. or even miles. The measured total acidity of such a soil appears to be a better index of its corrosivity than pH alone. by soils of differing composition.3 BUREAU OF STANDARDS TESTS The most extensive series of field tests on various metals and coatings in almost all types of soils were begun in 1910 by K. less corrosive than a highly conducting soil. In this regard. 10. Bureau of Standards tests [6] on steel specimens exposed 6–12 years and buried 30–120 cm (12–48 in. High concentrations of sodium chloride and sodium sulfate in poorly drained soils. in accordance with the principles of electrode kinetics discussed in Chapter 5. in general. whether from lack of moisture or lack of dissolved salts or both. Soils of low resistivity (<2000 Ω-cm) were considered to be aggressive.E. From studies at the National Physical Laboratory in the United Kingdom. such as are found in parts of southern California. but not invariably. Corrosion rates tend to increase with depth of burial near the soil surface.40 V (S. Soils in which the potential at pH 7 was low [<0. being one-half that of iron. Type 304 stainless steel was not pitted or was only slightly pitted (<0. in three soils. In 5-year tests carried out in Great Britain. above 6% Cr. Increase in chromium content of steel decreases observed weight loss in a variety of soils. however. Pitting is not pronounced.1. In 14-year tests. at least one specimen was perforated [0.3. copper 8 times. 10.4–0.1 Pitting Characteristics Because the dimensions of all field test specimens were on the order of inches up to about 1 ft (3–30 cm). on average. commercially pure aluminum was severely pitted in four soils (0. even in cinders in which the zinc coating was destroyed within the first two years.15 mm. For 4.1. 12% Cr and 18% Cr steels were severely pitted.or 5-year exposures. depth of pitting increases. . we see that copper.1.. Zinc coatings are effective in reducing weight loss and pitting rates of steel exposed to soils. It is expected. that pitting would also occur with this alloy in longer 1 time tests since pitting occurs in seawater within about 2 2 years. 4 to >63 mils). In 10-year tests in 45 soils. the rate is comparatively higher than in most other soils. Buttonwillow. Actual depth of pits in a given time is found to increase with size of test specimen.6 mm.15 mm (6 mils). corrodes at about one-sixth the rate of iron.95 kg/m2 (3. Referring again to Table 10. In poorly aerated soils or soils high in organic acids. 8% Ni.8 mm (16–32 mils) thick] by pitting.1 oz/ft2) effectively reduced (but did not prevent) corrosion. Type 316 stainless steel did not pit in any of the 15 soils to which the alloy was exposed for 14 years.8 oz/ft2) based on one side of the specimen (0. the symbol > means that the thickness of test specimen was completely penetrated by pitting at the end of the exposure period. the reported pitting rates represent minimum rather than maximum values. for example. data are listed for two soils relatively corrosive to steel and for one that is relatively noncorrosive. in tidal marsh. but. In later tests extending up to 13 years. coatings of 0. Pitting in some of these soils penetrated the test specimen thickness.e. In addition. but was virtually unattacked in a fifth soil [9]. a coating of 0. Lead also corrodes less on the average than does steel. The 18% Cr. Cinders constitute one of the most corrosive environments.85 kg/m2 (2. However. but. nor was weight loss appreciable in 10 out of 13 soils. showing the large variations in corrosion rate from one soil to another. and lead 20 times higher than the average rates in 13 different soils. California) in which some penetration of the base steel could be measured.208 CORROSION IN SOILS many soils are listed in Table 10. accounting for an average maximum penetration greater than the average given in Table 10. probably because cathodic area per pit increases (i. Zinc also pitted in some soils to an extent greater than the specimen thickness. For San Diego soil. <6 mils).13 mm or 5 mils thick) protected steel against pitting with the exception of one soil (Merced silt loam.1 to >1. the corrosion rate may be much higher (four to six times) than the average. The rate for copper is normal in otherwise corrosive San Diego soil. with the maximum depth reaching less than 0. corrosion rates of steel and zinc in cinders was 5 times. 07 0.19 — — 36 — — Soil Open Hearth Iron.001 in. Copper. 12-Year Exposure mils 0.10 21 >132 1.025 mm) for Total Exposure Period Average corrosion rates in g m−2 d−1 (gmd) Wrought Iron. 12-Year Exposure gmd 0.2 years) 0.10 23 0.09 209 . Adobe.6 years) 0.08 Montezuma clay. Lead. and Zinc in Soils (National Bureau of Standards [6]) BUREAU OF STANDARDS TESTS Maximum Penetration in mils (1 mil = 0.3 >53 (12 soils) 0.02 13 0.45 70 (44 soils) 0.95 100 0. 12-Year Exposure Copper.02 <6 (13.1.16 1.53 0.37 Merrimac gravelly sandy loam.45 61 (44 soils) mils gmd mils gmd mils gmd Bessemer Steel.06 10 (9.07 <6 (29 soils) Tidal marsh. Elizabeth. Corrosion of Steels.34 0. 12-Year Exposure mils gmd Average of several soils 90 >145 28 0.43 >137 80 1. Norwood. = 0.052 >32 (21 soils) <6 <6 0. New Jersey 1. San Diego 1. 8-Year Exposure Lead.47 59 (44 soils) 1.TABLE 10.013 19 Zinc. 11-Year Exposure gmd mils 0. Massachusetts 0. This difference is sometimes sufficiently great to make it worthwhile rotating a pipeline 180 ° after a given period of exposure in order to increase pipe life.. thus accounting for higher current densities at the pits. the greater the tendency for the pitting rate to fall off with time. In addition to this factor.210 CORROSION IN SOILS anode/cathode area ratio decreases). The reaction in aerated near-neutral solution at the pipe surface is O2 + 2 H 2O + 4e − → 4OH − (10.e. Stress-corrosion cracking of cathodically protected pipelines originates on the outer pipe surface..3) .) is a well-known cause of underground oil and gas transmission pipeline failures [11]. most commonly at areas where the coating is disbonded (i. Hydrogen ions. or decaying vegetation may then react with the hydroxide forming bicarbonates and carbonates [13]: CO2 + OH − → HCO− 3 HCO− + OH − → CO2− + H 2O 3 3 (10. consisting of sodium hydroxide and carbonate/sodium bicarbonate solutions. no longer attached to the surface of the steel pipe). producing an air space between the pipe and the soil. and k and n are constants. the pitting rate approaches a constant value.4 STRESS-CORROSION CRACKING Stress-corrosion cracking (S.E.1) Carbon dioxide from either the air. It has been reported [10] that values of n for steels range from about 0.). corresponding to −0.53 V (S. because of the pipe settling. 10. or penetration is proportional to time.9 for a poorly aerated soil.g. to 0. if present. The rate at which pits grow in the soil under a given set of conditions tends to decrease with time and follows a power-law equation P = ktn.C. long-line currents or macrocells. tends to become detached. and water containing dissolved oxygen migrate to the cathodic pipe surface (via pores or defects in the coating). cations (e. Pitting on the bottom side results from constant contact with the soil. Pits tend to develop more on the bottom side of a pipeline than on the top side.2) (10. In most engineering structures where cathodic protection is used to protect steel from general corrosion. Na+).C. Cathodic protection of underground pipelines can result in the accumulation of alkali at the pipe surface [12]. surrounding water. increase pit depth over the values obtained on small specimens where such cells do not operate. whereas the top side. where P is the depth of the deepest pit in time t. the steel is polarized to a potential of −0.85 V versus Cu/CuSO4. As n approaches unity.H.1 for a well-aerated soil. The smaller the value of n. 8. Section 8. Paint coatings of thickness normal to atmospheric protection fail within months when exposed to the soil. or copper.31 and −0.) thick afforded excellent protection to steel in 11-year exposure tests. in a more active potential range. of steel is maximum in the region of −0. the maximum susceptibility occurs near −0.C.46 V (S. A soil high in organic acids can be made less corrosive by surrounding the metal structure with limestone chips.7 V (S. 3.). Organic and Inorganic Coatings.C.C. for as many years as cathodic protection is adequately maintained.C. whereas in ammonium carbonate solution at 70 °C. Alteration of Soil. Coatings based on bituminous compositions offer excellent service but are now being supplemented by others—for example. that occurs in high-pH solutions. buried pipelines and tanks are protected by a combination of cathodic protection and an organic coating.E. steel.6 and 8.5 REMEDIAL MEASURES 1. Accumulation of NaOH can also cause S.3. . The effectiveness of zinc coatings has already been mentioned.C. A layer of chalk (CaCO3) surrounding buried pipes has been used in some soil formations likely to produce microbiologically influenced corrosion [16]. On the other hand.5). In addition to the intergranular S. In general. are found to be practical. 10. 4.REMEDIAL MEASURES 211 The potential range at which stress-corrosion cracks form in carbonate solutions is reported as noble to the optimum cathodic protection potential and active to the corrosion potential. whereas carbonates would be the more likely cause in less active potential regions.E.000 psi yield strength) linepipe steel developed intergranular cracks between −0.H. In 35% NaOH at 85 °C.) in 1 N Na2CO3 + 1 N NaHCO3 solution at 40 °C [14]. with reinforcing pigments or inorganic fibers to reduce cold flow. Bureau of Standards tests showed that soft rubber coatings 6 mm (0. Metallic Coatings. one high-strength (541 MPa.C. This combination protects steel against corrosion in all soils.) (see Figs. 2. 78. both effectively and economically. Cathodic Protection.E. They provide effective protection at reasonable cost. From these potential relations.3 V (S. thickly applied coatings of coal tar. one could assume that NaOH is more likely to be the cause of failure in sections of pipeline that are cathodically overprotected. fusion-bonded epoxy resin typically 1 mm thick applied at the pipe mill.H.H.7. for example. Such coatings deteriorate more rapidly when galvanically coupled to large areas of bare iron.25 in. S. in which case insulating couplings can be useful. transgranular cracking has been found to occur in dilute aqueous solutions of near neutral pH [15]. Schwerdtfeger and R. New York. 1989. Morgan. Arnold. U. Natl. R. Escalante. London. 1957. Underground Corrosion. Payer.S. Wiley. New York. 3rd edition. National Institute of Standards and Technology. McKinney. 1977. and B. C. J. H. Beavers. pp. The Corrosion and Oxidation of Metals. NACE. NISTIR 7415. Manuele. 8. Revie. in Underground Corrosion. GENERAL REFERENCES E. Analysis of Pipeline Steel Corrosion Data from NBS (NIST) Studies Conducted between 1922–1940 and Relevance to Pipeline Management. 14. G.212 CORROSION IN SOILS REFERENCES 1. National Bureau of Standards. 1–2. W. Philadelphia. R. 46–49. Berry. J. W. U. Wilmott and T. National Energy Board. D. editor. reprinted with a Preface by John A. Analysis of Pipeline Steel Corrosion Data from NBS (NIST) Studies Conducted between 1922– 1940 and Relevance to Pipeline Management. J. London. Sixth European Congress On Metallic Corrosion. Richard E. E. Metall. Can. Barlo. 39. 5. M. 10. Wiley. 741. Iron and Steel Institute. 1952. 1952. 1981. J. in Eurocorr ’77. May 2. N. 1978. R. Porter. Booth. Houston. Corrosion/78. 15. M. editor. Aquatic Chemistry: Chemical Equilibria and Rates in Natural Waters. 579. 22. Report MH-2-95. E. editor. Stumm and J. pp. Escalante. editor. 16. 12. 1981. Microbiological Corrosion. 235 (1983). TX. 64. 205–211. Ref. R. New York. in Uhlig’s Corrosion Handbook. J. Hewitt and W. L. Hudson and G. R. 123–132. National Institute of Standards and Technology. Department of Commerce. Parkins. 191–194. 2nd edition. Ricker. in Corrosion of Buried Metals. Texas. Calgary. 2000. B. pp. M&B Monographs CE/1. R. Revie. 9. J. Special Technical Publcation No.S. Wiley. Underground Corrosion. A. Acock. November 1996. Revie and R. 13. American Society for Testing and Materials. 4. Iron and Steel Institute Special Report 45. H. U. J. W. Special Report 45. H. 7. May 2. Romanoff. . p. R. Mills & Boon.) 65C. U. in Uhlig’s Corrosion Handbook.S. London. London. Evans. p. R. 6. Department of Commerce.S. Ricker. Q. Bur. 1971. T. Circ. R. Philadelphia. Society of Chemical Industry. Houston. 2. 171 (1961). Jack. pp. P. 1960. Moore. Ramsingh. W. London. 2000. Jones. E. 6. R. Public Inquiry Concerning Stress Corrosion Cracking on Canadian Oil and Gas Pipelines. Std. 2007. 11. Paper No. pp. 1996. Alberta. Fessler. NISTIR 7415. Report of the Inquiry. 329–348. W. 272. ASTM. Gilbert and F. J. Special Technical Publication 741. 3. 2007. London. National Association of Corrosion Engineers. 1977. (U. in Corrosion of Buried Metals. see also the reexamination and analysis of the original data. pp. 2nd edition. P. Wilmott and T. 2000. J. E.GENER AL REFERENCES 213 R. Revie. W. U. New York. 2nd edition. Department of Commerce. 2007. NISTIR 7415. Morgan. Jack. pp. . R. New York. M. Aquatic Chemistry: Chemical Equilibria and Rates in Natural Waters. W. editor.S. 1996. Wiley. Analysis of Pipeline Steel Corrosion Data from NBS (NIST) Studies Conducted between 1922–1940 and Relevance to Pipeline Management. in Uhlig’s Corrosion Handbook. Ricker. R. May 2. 329–348. J. National Institute of Standards and Technology. Stumm and J. Wiley. 3rd edition. . For example. Because of the present trend to increase process temperatures. to improve hightemperature oxidation resistance. 215 . which identifies the role of small additions of reactive elements.1 INTRODUCTION In high-temperature corrosion. Winston Revie and Herbert H. chemical synthesis) results in increased efficiency. such as yttrium. Cr2O3. Uhlig Copyright © 2008 John Wiley & Sons. energy conversion. Inc. it is more important than ever for engineers to understand Corrosion and Corrosion Control. Al2O3—that form between the metal and the environment are of critical importance.. by R. 2]. or alumina. hafnium. and cerium. thereby helping to achieve sustainability and improving economics. and it also increased the combustion efficiency (decreased fuel consumption) by 20% and reduced the nonburned hydrocarbons emission [1. zirconium. but also to reduce emissions. not only to increase efficiency. One of the major advances in corrosion science over the past few decades has been the characterization of the reactive element effect (REE). lanthanum. partially insulating the combustion chamber on a Diesel engine increased the piston surface temperature from 450 °C to 900 °C. there is a high priority for designers and engineers to increase process temperatures. waste incineration.g.11 OXIDATION 11. the protective oxide films—usually either chromia. Because increasing the temperature of most high-temperature processes (e. hence. or volatile compounds. the first-formed adsorbed film consists of atomic oxygen. 1377 (1968). or silver tarnishing in a contaminated atmosphere. Because the free energy of adsorption of atomic oxygen exceeds the free energy of dissociation of oxygen. through observation of the metal surface in H2O–H2 mixtures at elevated temperatures. such as oxygen. reaction products can be solid films or scales (a scale is a thick film). The basics of these matters will be explored in this chapter. Soc. When a metal is exposed at room or elevated temperatures to an oxidizing gas. that oxygen adsorbed on an 18% Cr stainless steel is thermodynamically more stable than the metal oxide. or both must diffuse through the scale in order for the reaction to continue. see A. For copper oxidizing in O2. chromia (Cr2O3). Acta Metall. but along different paths.5). which occurs when a metal is exposed to water or to the soil. Sci (Paris) 242. and (3) growth of a continuous oxide film. This is sometimes called “dry” corrosion. liquids. at *J. sulfur. 1118 (1965). Pignocco and G. Acad. 11. for nickel. migrate through solid oxides. As was discussed earlier in relation to the passive film (Section 6. E. For analogous data on iron. sulfides.2 INITIAL STAGES The processes that take place at a clean. and they diffuse simultaneously with electrons. respectively. the environment. When a protective scale—such as alumina (Al2O3). the metal. 16. Bénard [C. corrosion may occur in the absence of a liquid electrolyte. Any additional layers of adsorbed oxygen are still less strongly bonded. . J. Usually ions. Low-energy electron-diffraction data show that some metal atoms enter the approximate plane of adsorbed oxygen to form a relatively stable twodimensional structure of mixed O ions (negative) and metal ions (positive).* Continued exposure to low-pressure oxygen is followed by adsorption of O2 molecules on metal atoms exposed through the first adsorbed layer. The resultant structure is usually more stable on transition than on nontransition metals [4]. in contrast to “wet” corrosion. 1724 (1956)] showed. this initial adsorbed partial monolayer is thermodynamically more stable than the metal oxide. for example.216 OXIDATION the principles of high-temperature corrosion. Moreau and J. or halogens. and the methods that are being used to improve the performance of engineering materials at high temperatures. Pellissier. whereas NiO decomposes accompanied by oxygen dissolving in the metal. it is adsorbed without dissociation to its atoms. rather than atoms. the limitations of materials. 112. Electrochem. (2) formation of oxide nuclei that grow laterally. reactive metal surface exposed to oxygen obey the following sequence: (1) adsorption of oxygen. the solid electrolytes are Cu2O and Ag2S. the adsorbed film resists decomposition up to the melting point of nickel [3]. the reaction product can be considered an electrolyte. In dry corrosion. Since the second layer of oxygen is bonded less energetically than the first layer. R. and the outer layers. or silica (SiO2)—forms. Hondros. The migrating ions are not hydrated. or halides. * and then they grow more Figure 11. more sites are brought into play. multilayer adsorbed oxygen on metal M eventually favors transformation to a crystalline stoichiometric oxide. aided by rapid surface diffusion of M and O ions (Fig. 525 °C. ledges. with the corresponding diffraction pattern being that of an amorphous structure. 20 s (17. *Limited perhaps by the electron tunneling distance [4] below which electrons can penetrate the metal-oxide–interface barrier without first acquiring the usual activation energy. When conditions are favorable. Black lines are bands of imperfections (stacking faults) in the copper lattice (Brockway and Rowe [5]). Because the free energy of adsorption per mole of oxygen decreases with amount of oxygen adsorbed (the O–substrate bond becomes weaker). Because the amount of adsorbed oxygen on any likely site increases with oxygen pressure. since multilayer adsorption decreases at elevated temperatures. such as at surface vacancies. the density of nuclei decreases with increase of temperature [6. eventually become mobile.3 Pa).1) [5]. whereas ΔG for 2nM + (nO2)·O·Mads → (2n + 1)MO to form the oxide becomes correspondingly more negative. 7]. Nuclei grow rapidly to a certain height on the order of nanometers. . ΔG for O·Mads + nO2 → (nO2)·O·Mads becomes less negative per mole O2 as n increases. Similarly.INITIAL STAGES 217 elevated temperatures.1.600×). 11. oxide nuclei form at specific sites of the surface. oxide formation nucleates preferably at surface sites where multilayer adsorption is favored. and the density of nuclei increases. Therefore. In other words. or other imperfections. Nuclei of Cu2O formed on copper surface at 10−1 mmHg oxygen pressure (13. H. extrapolating to the scales of CO/CO2. shown in Fig. or whether it contains cracks and pores and is relatively nonprotective.2. at 1000 °C. respectively. 11. the stresses set up in the oxide may cause it to crack and detach—called spalling—and the oxidation rate increases. 11. a continuous film of Cu2O forms within 6 s. the reactant is 1 mole of O2. by drawing straight lines through the points labeled O. Using transmission electron microscopy and computer modeling. When the continuous film reaches a thickness in the order of several hundred nanometers. Growth of the resulting thin continuous oxide film may be controlled by transfer of electrons from metal to oxide [9]. Eventually. we can estimate the equilibrium values of oxygen partial pressure. and through the point on the metal/metal oxide line at the temperature of interest. that is. for any oxide can be read directly off the vertical axis of the graph. From the diagram. The latter situation prevails so long as the oxide remains continuous. For example. the crystal structure of the oxide is aligned with the structure of the metal. can be used to predict the conditions under which a metal is oxidized or a metal oxide is reduced [11]. at a critical thickness. and C.4 PROTECTIVE AND NONPROTECTIVE SCALES The rate of reaction in the later stages of oxidation depends on whether the thick film or scale remains continuous and protective as it grows. and pO2. 11. On copper exposed to 1 mm Hg (133 Pa) oxygen pressure at 550 °C. or by migration of metal ions in a strong electric field set up by negatively charged oxygen adsorbed on the oxide surface [10].3 THERMODYNAMICS OF OXIDATION: FREE ENERGY–TEMPERATURE DIAGRAM Graphical data of the standard Gibbs free energy of formation of oxides versus temperature.2.218 OXIDATION rapidly in a lateral rather than in a vertical direction. For every reaction in Fig. ΔG °. the standard Gibbs free energy of formation of NiO is approximately −250 kJ for two moles of NiO. 11. it has been shown that the metal–scale interface is very tightly bonded as a result of epitaxial relations between the metal and the scale. diffusion of ions through the oxide may become rate controlling instead. so that the reactions are 2x 2 M + O2 = M x O y y y The standard Gibbs free energy of formation. as well as CO/CO2 and H2/H2O ratios in contact with metals and their oxides. as the crystalline oxide grows on the metal substrate. H2/H2O. at 1 atm O2. Because reaction-product films . the time required is still less. at least for thin scales [8]. asminternational. D is the density of scale. or in compression. Md/ nmD.PROTEC TIVE AND NONPROTEC TIVE SC ALES 219 Figure 11.org. that is.) are usually brittle and lack ductility. This situation. m and d are the atomic weight and density of metal. depends on whether the volume of reaction product is greater or less than the volume of metal from which the product forms. www. where M is the molecular weight. favoring fracture. All rights reserved. and n is the number of metal . (Reprinted with permission of ASM International®. the initiation of cracks depends in some measure on whether the surface film is formed in tension. respectively. in turn.2. Standard Gibbs free energy of formation of oxides as a function of temperature [11]. it depends on the Pilling–Bedworth ratio. favoring protection. 1 Three Equations of Oxidation The three main equations that express thickness y of film or scale forming on any metal within time t are (1) the linear.6 0. Under these conditions. which tend to form nonprotective oxides. whereas for aluminum and chromium. and (3) the logarithmic.0 1. the scale is formed in tension and tends to be nonprotective. the ratios are 1.3 0. the oxide. and high thermodynamic stability.4.1.8 1.6. a protective scale is predicted to form. then compressive stresses can develop because of the constrained space available for the oxide to occupy. for example. but it becomes nonprotective above a certain thickness. the ratios are 0. when this ratio is less than unity. protective scales are expected.1 lists the Pilling–Bedworth ratios for several metals.64 and 0. which tend to form protective oxides. is expected to be protective.7 1. low vapor pressure and high melting temperature of the oxide. except at high temperatures (approximately >800 °C). On the other hand. where it volatilizes. oxide is protective during the early stages of oxidation. if the scale forms at the metal–oxide interface by migration of oxide ions. Similarly. scale that forms at the oxide–gas interface by migration of metal ions outward is not constrained to occupy the volume of the metal that was oxidized. thick scales of TiO2 on titanium are nonprotective. slow oxide growth rate. Table 11.3 and 2. Cr2O3 scale on pure chromium at 1100 °C.220 OXIDATION TABLE 11.9 1. In addition to the Pilling–Bedworth ratio. protection by an oxide depends on adherence of the oxide to the substrate. The growth mechanism of the oxide is also relevant to oxide protection.7 1.7 1. On the other hand. at high temperatures the rapid growth of oxide may cause compressive stresses to become so great that the scale spalls.0 [13]. . Although the Pilling–Bedworth ratio may be greater than 1. and hence. For example.0. respectively.79. WO3. Pilling–Bedworth Ratios Metal Chromium Cobalt Titanium Iron Copper Nickel Aluminum Magnesium Calcium Lithium Oxide Cr2O3 CoO TiO2 FeO CuO NiO Al2O3 MgO CaO Li2O Pilling–Bedworth Ratio 2. 11. For calcium and magnesium. with a Pilling–Bedworth ratio of 2. Typically.6 atoms in a molecular formula of scale substance (n = 2 for Al2O3) [12]. (2) the parabolic.8 0. If Md/nmD >1. so that a protective scale would be predicted. respectively. Tungsten has a ratio of 3. t. in accord with the parabolic equation. such as when the controlling reaction rate is constant at an inner or outer phase boundary of the reaction-product scale. as a function of time of oxidation. the thickness of scale. on the ratio Md/nmD and the relative thickness of the film or scale. for such metals. for example. or dy/dt = k and y = kt + const. When the rate of formation of the outer scale becomes equal to that of the inner scale. plotted with time. therefore. the rate of oxidation is constant. the linear equation may also hold even though the latter ratio is greater than unity. The particular equation that applies depends. tungsten first oxidizes at 700–1000 °C. when the environment reaches the metal surface through cracks or pores in the reaction-product scale. the ratio Md/nmD is usually less than unity. 11. forming an outer porous WO3 layer and an inner compact oxide scale [14]. is linear (Fig. the linear equation is obeyed. as. in part. y. For the linear equation.3).PROTEC TIVE AND NONPROTEC TIVE SC ALES 221 Figure 11. the diffusion of ions or migration of electrons through the scale is controlling. Hence. y.3. and the rate. Equations expressing growth of film thickness. This equation holds whenever the reaction rate is constant at an interface. is inversely proportional to scale thickness. For the parabolic equation. for example. dy k ′ = dt y or y2 = 2k ′t + const . In special cases. Hence. where k is a constant. t. including Cu. On this basis. it is found that dy k ′′ = dt t or y = k ′′ ln t ( const + 1) This is called the logarithmic equation. Sn. Data obeying this equation can also be represented. It is often difficult to distinguish between the logarithmic and the inverse logarithmic equations because of the limited range of time over which data can be accumulated for thin-film behavior. and the inverse logarithmic equation [10] results: 1 = const − k ln t y This relation has been reported to hold for copper and iron oxidized at low temperatures [17]. Ni. When the thickness of the film exceeds the maximum thickness of the spacecharge layer. changes the oxidation rate. iron. First reported by Tammann and Köster [15]. for thin-film behavior. a linear relation is obtained (Fig. The preponderance of electric charge of usually negative sign in oxides near the metal surface. If. Al. Correspondingly. For relatively thin protective films. or magnetic transition (Curie temperature). and cobalt. the parabolic equation applies. migration of ions controls the rate. diffusion or migration through the film then usually becomes controlling. This situation also makes it difficult to evaluate other types of equations that have been proposed. metals forming protective films first follow the logarithmic equation and then the parabolic (or linear) equation. chromium. by a two-stage logarithmic equation where an initial lower rate is followed by a final higher rate .222 OXIDATION Accordingly. the logarithmic equation has been found to express the initial oxidation behavior of many metals. Mn. y3 = kt + const. lattice transition. no longer affect the rate. This equation holds for protective scales corresponding to Md/nmD > 1 and is applicable to the oxidation of many metals at elevated temperatures. Pb. Cd. Therefore. if y is plotted with log (t + const). such as grain orientation. when it contains a space charge throughout its volume. as when metals oxidize initially or when oxidation occurs at low temperatures. Zn. nickel. such as grain orientation and Curie temperature.3). similar to the electrical double layer in aqueous electrolytes. a linear relation is obtained. Ti. such as copper. as has been observed [16]. The logarithmic equation can be derived on the condition that the oxidation rate is controlled by transfer of electrons from metal to reaction-product film [9] when the latter is electrically charged—that is. or with log t for t >> const. such as the cubic equation. has been demonstrated experimentally. Fe. and the factors just mentioned. 11. the rate of ion migration is an exponential function of the field strength. and Ta. any factor changing the work function of the metal (energy required to extract an electron). if y2 is plotted with t. in many cases. and the prevailing electric field within the film is considered to be set up by gaseous ions adsorbed on the outer surface. with either equation apparently applying equally well. who noticed that when an iron surface was painted with a slurry of Cr2O3. and A is a constant. outward diffusion of metal ions occurs in preference to diffusion of the larger oxygen ions inward. T is the absolute temperature. by quantitative studies. the greencolored layer. 11. migrate through Ag2S. after oxidation. The oxidation rate for thin. For these metals. Reaction occurring at the outer rather than the inner oxide surface was first reported by Pfeil [18]. 11. . was buried in the middle or lower layers of iron oxide. reacting there with oxygen. in contact with the sulfur. that Ag+ ions. He [20] derived an expression for the parabolic rate constant that has the following simplified version [21]: Figure 11.or thick-film conditions increases with temperature. neither gained nor lost weight. next to the silver. 220 °C (Wagner [19]). Wagner also showed that by assuming independent migration of Ag+ and electrons.4. oxidize. in other words. he showed that.3).4). R is the gas constant. 11. whereas the top pellet. and nickel. after 1 h at 220 °C. 1-h test. Wagner [19] showed. and not S2− ions.5 WAGNER THEORY OF OXIDATION When certain metals. the observed reaction rate could be calculated from independent physical chemical data. diffused through the Cr2O3 marker and reacted with oxygen at the gas–oxide interface. gained weight chemically equivalent to the loss in weight of metallic silver. Formation of silver sulfide from silver and liquid sulfur. The higher rate is ascribed to formation of a diffuse space-charge layer overlying an initially constant charge-density layer [9]. Similarly. By placing two weighed pellets of Ag2S over silver. Iron ions. it is found that metal ions migrate through the oxide to the outer oxide surface.WAGNER THEORY OF OXIDATION 223 [16] (Fig. obeying the Arrhenius equation reaction − rate constant = A exp ΔE ( −RT ′ ) where ΔE′ is the activation energy. the bottom pellet. on top of which was molten sulfur (Fig. zinc. such as copper. sites where electrons are missing. together with electrons. O2−. n2. p. There are two types of semiconducting oxides. and y is the thickness. κ is the mean specific conductivity of film substance. Some metals—for example. Conductivity increases with a slight shift from stoichiometric proportions of metal and oxygen and with an increase of temperature. shift of stoichiometric proportions takes the form of a certain number of missing metal ions in the oxide lattice called cation vacancies. represented by . in a Cu2O lattice is an example of a positive hole. The excellent agreement between theory and observation confirms the validity of the model suggested by Wagner for the oxidation process within the region for which the parabolic equation applies. and electron transference numbers. anion. their conductivity lies between insulators and metallic conductors. in part. The constant. Equation (11.6 × 10−6 eq/ cm-s. via anion vacancies in the corresponding outer oxides. that is. and F is the Faraday. and Cr2O3. NiO. k. appears in the relation dw A =k dt y where dw/dt is the rate of formation of film substance in equivalents/second. FeO. CoO.3 × 104 Pa). Bi2O3. a process that is equivalent to the reverse migration of Cu+ and electrons. At the same time. ⊕.and n-types (p = positive carrier. excess metal ions exist in interstitial positions of the oxide lattice. respectively. These migrate to the metal surface.1) holds for diffusion of either cations or anions or both in the reaction-product scale. within the reaction-product film. that migrate during oxidation . n = negative carrier).5. 11. the calculated and observed values of k are 6 × 10−9 and 7 × 10−9 eq/cm-s.6 OXIDE PROPERTIES AND OXIDATION Metal oxides are commonly in the class of electrical conductors called semiconductors. 11. an equivalent number of positive holes. In the p-type. titanium and zirconium— oxidize. Cu2+. For n-type oxides. by migration of oxygen ions. and it is these. cation vacancies and positive holes are formed at the outer O2–oxide surface. namely. A is the area. A model for the Cu2O lattice is shown in Fig. Oxides of p-type are Cu2O. For copper oxidizing at 1000 °C and 100-mm oxygen pressure (1.1) where E is the emf of the operating cell derived either from potential measurement or free-energy data. to maintain electrical neutrality.224 OXIDATION k= En3 (n1 + n2 )κ F (11. form—that is. and n3 are the mean cation. A cupric ion. n1. The calculated value of k for silver reacting with sulfur at 220 °C is 2–4 × 10−6 compared to the observed value of 1. respectively. During the oxidation of copper. Lattice defects. Hence. the concentration of positive holes must increase in order to maintain electrical neutrality. if a few singly charged Li+ ions are substituted for doubly charged Ni2+ ions in the NiO lattice. Wagner showed that the law of mass action can be applied to the concentration of interstitial ions and electrons. and also to cation vacancies and positive holes. TiO2. for Cu2O.5) to the outer oxide surface.5) These relations lead to predictions regarding the effect of impurities in the oxide lattice on oxidation rate. to maintain equilibrium for the reaction . Examples of n-type oxides are ZnO.5. the corresponding equilibrium is 1 ⎯ → int ZnO ← Zn 2+ + 2e − + O2 ⎯ 2 and 2 C Zn2+ ⋅ C e− = int (11. and Al2O3. the equilibrium relations are 1 ⎯ → O2 ← Cu 2O + 2Cu − + 2 ⊕ ⎯ 2 and 2 2 2 C Cu− ⋅ C ⊕ = const ⋅ p1/2 O (11. (Fig. for example.3) For ZnO. Hence. 11. CdO.OXIDE PROPERTIES AND OXIDATION 225 Figure 11.2) (11.4) const 1 pO/22 (11. and incorporation of Cr3+ increases cation vacancy concentration. where x connotes a number between 0 and 1. Incorporation of Ni2+ ions into the Cr2S3 scale decreases cation vacancy concentration.0 3.0 Parabolic Rate Constant.3). possibly because a scale composed of Cr2O3 forms instead of NiO. with the rate of sulfidation being less than that of pure chromium for Cr–Ni alloys containing >20% Cr.2. copper oxidizes at higher rates the higher the oxygen pressure. in accord with prediction [25]. outward diffusion of Cr3+ occurs in a scale composed of Cr2S3. In a sulfur atmosphere. up to 2% Cr alloyed with nickel accelerates reaction at 600–900 °C [24]. 1 atm O2 Weight Percent Cr 0 0. the scale is heterogeneous. the rate decreases again. the concentration of positive holes decreases. Data of Wagner and Zimens [23] showing the effect of alloyed chromium on oxidation of nickel are given in Table 11. which alters the rate of ion migration apart from factors described previously. and the oxidation rate increases. it is apparent that higher partial pressures of oxygen for ptype semiconductors must be accompanied by higher concentrations of vacancies and holes at the O2–oxide interface. This is accompanied by a decrease in oxidation rate [22] because the rate is controlled by migration of cation vacancies. In Cr–Ni alloys containing >40% Cr. At 10% Cr. consisting of both nickel and chromium sulfides. Hence. if small amounts of Cr3+ are added to NiO. thereby decreasing the reaction rate to a value below that for pure chromium. (11. From Eq. At intermediate chromium compositions. On the other hand. the concentration of cation vacancies correspondingly increases.1 × 10−10 14 × 10−10 26 × 10−10 31 × 10−10 1.3 1.2. Oxidation of Nickel Alloyed with Chromium.226 OXIDATION 1 ⎯ → O2 ← NiO + Ni 2− + 2 ⊕ ⎯ 2 as expressed by 2 2 C Ni 2− ⋅ C ⊕ = const ⋅ p1/2 O the concentration of cation vacancies must decrease. Outward diffusion of Ni2+ occurs through cation vacancies in Ni1−xS. k (g2cm−4s−1) 3.0 10. for reasons paralleling those pertaining in the situation with O2.5 × 10−10 . 1000 °C. TABLE 11. Ag+ ions diffuse over the surface of the tantalum. The rate of .6). or graphite.7 GALVANIC EFFECTS AND ELECTROLYSIS OF OXIDES The electrochemical nature of oxidation at elevated temperatures suggests that galvanic coupling of dissimilar metals should affect the rate.4% Li 0. with the latter supplying electrons that hasten conversion of silver to AgI. in fact. for example. 390 °C. and. 11. which is an n-type semiconductor. it was found [28] that. 11. 1 atmO2 Parabolic Rate Constant. Al3+ decreases the rate. The reaction of silver with gaseous iodine at 174 °C. Ag2S forms a tightly adherent coating over the gold surface.5)]. Li+ increases the oxidation rate of zinc. A second example is given by nickel completely immersed in molten Na–K borax to a depth of about 3 mm at 780 °C in 1 atm O2 (Fig.1% Al Zn + 0. Data showing these effects are presented in Table 11. when silver coated with a porous gold electroplate is exposed to sulfur vapor at 60 °C.3. (11.GALVANIC EFFEC TS AND ELEC TROLYSIS OF OXIDES 227 TABLE 11. which play a major role in semiconductor properties. contrary to its effect in NiO.3 [26]. and any further decrease brought about by increasing oxygen pressure has little effect on their concentration gradient referred to the metal surface where ZN 2+ concentration is highest. which is mainly an ionic conductor..g. platinum. In addition. the electron concentration decreases in order to preserve neutrality. int Therefore. AgI. such effects do occur [27]. forms on silver at a rate limited by transport of electrons across the AgI layer. int By similar reasoning. alloying elements present in large percentages (e. to the usual film of AgI on the silver. traces of impurities. therefore. k (g2cm−4h−1) Zn pure Zn + 0. This increase in concentration facilitates the diffusion of ZN 2+ .8 × 10−9 1 × 10−11 2 × 10−7 If small amounts of Li+ are added to ZnO. and the concentration of interstitial zinc ions increases in accord with the law of mass action [Eq. When silver is coupled to tantalum. is accelerated by contact of the silver with tantalum. >10% Cr–Ni) affect the oxidation rate by a gross alteration of the actual composition and structure of the film in addition to any effects on semiconducting properties.7). Analogously. 11. also appreciably affect rates of oxidation of metals protected by semiconductor films. The oxidation rate of zinc is almost independent of oxygen pressure because the concentration of interstitial zinc ions is already low at the O2–oxide interface. hence. Oxidation of Zinc. the compound spreads progressively over the tantalum surface (Fig. On the other hand. these investigators decreased the oxidation rate at 880 °C when iron was made cathodic.6. O2. borate melt.7 V under the conditions cited previously.5 A/cm2. Instead. Ni—illustrating accelerated oxidation of nickel through contact with platinum. The dependence of oxidation on ion migration within a reaction-product layer suggests that the oxidation rate should be affected by an applied electric current. If nickel is coupled to platinum or silver gauze. [29] By wrapping a platinum wire around oxidized iron and passing a current of 1.7. Effect of coupling tantalum to silver on reaction of iodine vapor with silver (Ilschner-Gensch and Wagner [27]). probably by supplying Fe2+ ions. Nickel under these conditions corrodes even more rapidly than if exposed to pure oxygen at the same temperature because a protective NiO scale does not form. and the platinum acts as an oxygen electrode. . and they accelerated the rate when iron was made anodic. which are oxidized to Fe3+ by oxygen near the electrolyte surface. This was first shown to be the case by Schein et al. Addition of 1% FeO to borax increases the oxidation rate still more. Galvanic cell—Pt. Figure 11. oxidation under these conditions is low because of limited access of oxygen from the gaseous phase. corrosion of the nickel is accelerated by a factor of 35–175 for 1-h exposure. with the trivalent ions then being reduced again either at the cathode or by local cell action at the nickel anode. Similar relations were shown by Jorgensen for the oxidation of zinc in oxygen at 375 °C [30]. the open-circuit potential difference between platinum and nickel is 0. the latter reaching above the liquid borax surface.228 OXIDATION Figure 11. Ni2+ dissolves in the borax electrolyte. In the case of vanadium-containing oils. Eutectic combinations of oxides reduce the melting point still further.p. The melting points of V2O5.9 HOT CORROSION Hot corrosion is the accelerated oxidation of a material at elevated temperatures induced by a thin film of fused salt deposit [36]. Additions to oil of substances (e. operating at temperatures below the melting point of the liquid phase avoids catastrophic oxidation. High-vanadium oils are found only in certain wells. and 294 °C. The solid residual ash of such oils may reach 65% V2O5 or higher. respectively. In the late 1960s. and the damage caused by the ash is not different from that observed when vanadium is alloyed with heatresistant materials. Also.g. in which corrosion is commonly controlled by the diffusion of oxygen to the metal surface. accordingly. compared to 1527 °C for Fe3O4. Similar accelerated oxidation in air was reported for stainless steels containing a few percent molybdenum or vanadium [32]. It has been suggested that the cause is a lowmelting oxide phase that acts as a flux to dislodge or dissolve the protective scale. is not easily removed. alloys high in nickel are more resistant because of the high melting point of NiO (1990 °C). 2800 °C) are found. or hot ash corrosion [35]. The vanadium is present in the oil as a soluble organic complex and. Damage of this kind is sometimes called catastrophic or accelerated oxidation. or as little as 0. It is called ‘hot corrosion’ because. the soluble oxidant is SO3 2− ( S2 O7 ) in the fused salt. the melting point of which has been found to be as low as 500 °C.8 HOT ASH CORROSION An 8% Al–Cu alloy. powdered dolomite. it shares some similarities with aqueous atmospheric corrosion. MoO3. some of which are located in South America. oxidizing in air at 750 °C in the presence of MoO3 vapor from an adjoining molybdenum wire not in contact. was found to react at an abnormally high rate [31]. The condensed liquid film deposits .HOT CORROSION 229 11. during the Viet Nam conflict. the presence of sulfur and sodium compounds in the oil and the catalytic effect of V2O5 for converting SO2 to SO3 result in a scale containing Na2SO4 and various metal oxides. hence. being caused by a thin electrolyte film.p. hot corrosion caused severe corrosion of gas turbine engines of military aircraft during operation over seawater. are 658 °C. and B2O3. The oxidation products were voluminous and porous. An analogous accelerated damage is found in boiler tubes or gas turbine blades operating at high temperatures in contact with combustion gases of crude oils high in vanadium [34]. or powered Mg) that raise the melting point of the ash by forming CaO (m. 795 °C. for example. In hot corrosion. calcium or magnesium soaps. Similarly.04% boron [33].. to be beneficial. Sulfide formation results from the reaction of the metallic substrate with a thin film of fused salt of sodium sulfate. 2570 °C) or MgO (m. 11. but hot corrosion of marine turbine blades may occur even with use of very low-sulfur fuels [38].g. but may also involve galvanic effects. e.e. the presence of Cr2O3 films may be related to the increased volatility and lesser tendency of Na2SO4 to deposit on the alloy surface at higher temperatures [38. Supplied with a liquid electrolyte in which oxygen and metal ions migrate rapidly. The accelerated attack was repeated in the laboratory [37] by applying a thin coating of Na2SO4 to either 30% Cr–Co or 30% Cr–Ni alloys and oxidizing at 600–900 °C (1110–1650 °F) in O2 at 1 atm containing 0. porous network of an electronically conducting oxide. at 650 °C [37]. A spongy. Metallic sulfides often form inclusions or a network of sulfides along grain boundaries of the alloy.15% (SO2 + SO3). It was first identified in marine propulsion gas turbines where service temperatures of first-stage blades and vanes were in the range 650–700 °C (1200–1300 °F). In metallographically examining corroded components. the vapor pressure of Na2SO4 in the vapor phase exceeds its equilibrium partial pressure at the substrate temperature) or by physical deposition (solid or liquid salt detaches from an upstream component and attaches to a hot substrate on impact) [36].230 OXIDATION either by chemical deposition (i. oxide particles were observed to be dispersed in the adherent salt film.. forming a film of Cu2O. such a cell accounts for an accelerated oxidation process far exceeding the rate for a metal reacting directly with gaseous oxygen through a continuous oxide scale. and the base metal acts as an anode. Oxidation products consisted mostly of once-liquid mixtures of Na2SO4 + CoSO4 (or NiSO4). but precipitates as non-protective particles within the salt film [36].. a fluxing mechanism may apply. in accord with practical experience. 11. High-chromium alloys are more resistant to hot corrosion than are low-chromium alloys. Damaging deposits of Na2SO4 may originate alone from sea salt contamination of intake air. in which an otherwise protective oxide scale on the substrate surface dissolves at the oxide/salt interface. Sulfur dioxide and trioxide of oil combustion products also contribute. The Fe3O4 sponge acts as an oxygen electrode of large area. The rate varies with crystal face. 39]. As a result. reproduces the cell described previously containing a platinum cathode and a nickel anode in contact with molten borax. It has been suggested that lower oxidation rates above 750 °C result from formation of protective Cr2O3 films unable to form at lower temperatures. In turn. such as Fe3O4 filled with a liquid electrolyte. The fluxing mechanism of hot corrosion has been thoroughly reviewed by Rapp [36]. Ilschner-Gensch and Wagner suggested [27] that the mechanism is not one of melting point or fluxing alone.10 OXIDATION OF COPPER Copper oxidizes in air at low temperatures (<260 °C) in accord with the two-stage logarithmic equation. the corresponding attack is then called sulfidation. . The highest oxidation rates occurred at 650–750 °C. Hot corrosion may occur in the absence of metallic contaminants in the fuel. Alloying elements that are particularly effective for improving oxidation resistance at high temperatures are aluminum. Internal oxidation is not observed. with cadmium-. On the other hand. aluminum. Above a critical concentration of the alloying component. silicon. copper oxidizes with precipitation of oxide particles within the body of the metal as well as forming an outer oxide scale. 750 °C. In accord with a diffusion-controlled mechanism. at 256 °C. 500 °C. The subject was reviewed by Rapp [45]. Oxidation within the metal is called subscale formation or internal oxidation. and much more rapidly than that for nickel or the heat-resistant Cr–Fe alloys. Between about 260 °C and 1025 °C. such as for alloys of sodium–lead. 11. 450 °C. 1050–1100 °C. the Cu2O film is overlaid by a superficial film of CuO.10. 25–30% Cr–Fe (0. manganese. 11.1% C). heat treatment with nitrogen or helium increases the rate because adsorbed oxygen (traces from gas or metal) favors submicroscopic facets of predominantly (100) orientation [40]. Maximum improvement by aluminum additions occurs at about 8% [43]. Internal oxidation is usually not pronounced for any of the iron alloys. beryllium. or zinc-based alloys. The mechanism is apparently one of oxygen diffusing into the alloy and reacting with alloying componens of higher oxygen affinity than that of the base metal before the alloying components can diffuse to the surface. and magnesium–tin [44]. aluminum–tin. Similar behavior is found for many silver alloys. beryllium. tin. a 2% Be–Cu alloy oxidizes in 1 h at 1/14 the rate of copper [42]. for example. Cast tough-pitch copper containing . Ni. 8–10% Cr–Fe (0. 800 °C. presumably favoring the (111) orientation. titanium.g. and magnesium. Fe.2 Reaction with Hydrogen (“Hydrogen Disease”) The tendency of copper to dissolve oxygen when the metal is heated in air leads to rupture of the metal along grain boundaries by formation of steam when the metal is subsequently heated in hydrogen. Oxidation changes from logarithmic to parabolic behavior above 400–500 °C. iron. in general. the depth of subscale grows in accord with the parabolic equation [44]. a compact protective layer of the component oxide tends to be formed at the external surface which thereafter suppresses internal oxidation.1% C). Only Cu2O forms in air above 1025 °C. tin-. This is shown by the following temperatures [41].1 Internal Oxidation When alloyed with small percentages of certain metals (e..10. Copper oxidizes at a rate slightly higher than that for iron. lead-. but without formation of an outer scale. below which the scaling losses in air are less than approximately 2–4 g m−2h−1: Cu.OXIDATION OF COPPER 231 decreasing in the order (100) > (111) > (110). Heat treatment of polycrystalline copper with hydrogen at 300–450 °C decreases the oxidation rate in oxygen at 200 °C because submicroscopic surface facets are formed by adsorbed hydrogen. A few exceptions have been noted. and zinc). . Unfortunately. it decomposes at room temperature into Fe3O4 plus Fe. 9% Fe3O4. 49]. consequently. Typical oxidation behavior is shown in Fig. Data reported by various investigators are not in good agreement. but they may become moderately so. some of these drawbacks of aluminum–iron alloys are overcome. Oxygenfree silver heated in hydrogen for 1 h at 850 °C (1550 °F) is not embrittled or damaged. Subsequently.11 OXIDATION OF IRON AND IRON ALLOYS In the low-temperature region (approx 250 °C). The mechanism is the same as that applying to copper. Some dissolved hydrogen undoubtedly escapes before oxygen can diffuse into the silver. and the tendency to form aluminum nitride that causes embrittlement have combined to limit their application as oxidation-resistant materials. and the proportions of these layers change as the temperature or oxygen partial pressure changes. grow to form a uniform film of oxide. hence diminishing subsequent damage.232 OXIDATION free Cu2O is very sensitive to this type of damage. The good oxidation resistance of the chromium–iron alloys. however. decreasing in the order (100) > (111) > (110) [47].8. The most efficient alloying elements for improving oxidation resistance of iron in air are chromium and aluminum. Below 570 °C (1058 °F). if any is formed above this temperature. apparently consisting of Fe3O4. Oxidation of iron in the parabolic range is complicated by formation of as many as three distinct layers of iron oxide. the poor mechanical properties of aluminum–iron alloys. and it becomes blistered or loses ductility if later heated in hydrogen above 500 °C (925 °F). 11. Oxygen-free coppers are not susceptible. The oxide nuclei. At 600 °C in 1 atm O2 for 100 min. 11. but not as severe as. are not subject to similar damage when heated in hydrogen. In combination with chromium. Instances are on record where damage has been caused by hydrogen at temperatures as low as 400 °C (750 °F). the characteristic loss when silver containing oxygen is heated in hydrogen [46]. An 8% Al–Fe alloy is reported to have the same oxidation resistance as a 20% Cr–80% Ni alloy [51]. Use of these elements with additional alloyed nickel and silicon is especially effective. a three-layer scale is composed of 90% FeO. and less than 1% Fe2O3. FeO is unstable and. However. loss of ductility occurs that is similar to. should they be heated at any time in oxygen or in air. Gold and platinum dissolve little or no oxygen and. it is reported that the scale is composed of two layers: an inner FeO layer equal in weight to an outer Fe3O4 layer [50]. the sensitivity of their protective oxide scales to damage. α Fe2O3 nucleates and covers the Fe3O4 layer [48. combined with acceptable mechanical properties and ease of fabrication. the oxidation rate of iron is sensitive to crystal face. when it is heated immediately afterward in air at the same temperature. accounts for their wide commercial application. probably because of variations in the purity of iron used for oxidation tests—particularly its carbon content. At 900 °C for 100 min. Silver similarly dissolves oxygen when heated at elevated temperatures in air. Berlin (1956). depends not only on the diffusion-barrier properties of reaction-product scales. 220 h [E.) Improvement in oxidation resistance of iron by alloying with aluminum or chromium probably results from a marked enrichment of the innermost oxide scale with respect to aluminum or chromium. especially on long exposures. Springer. but also on the continuing adherence of such scales to the metal. Scales that are otherwise protective often spall (become detached) during cooling and heating cycles because their coefficient of expansion differs from that of the metal. Vol. from chemical analysis. 815. 11. These inner oxides resist ion and electron migration better than does FeO. an ASTM accelerated test [53] for oxidation resistance of wires calls for a cyclic heating period of 2 min at a specific temperature followed by a cooling period of 2 min. The middle oxide scales are known. The life of a wire in this test is measured as the time to failure or the time to reach a 10% .5% C.8. and electron-microprobe analyses confirm marked enrichment of chromium in the oxide adjacent to the metal phase in the case of chromium–iron alloys [52].LIFE TEST FOR OXIDATION-RESISTANT WIRES 233 Figure 11. Houdremont. Effect of alloyed chromium on oxidation of steels containing 0. to be so enriched. For chromium–iron alloys. I. Fig. Hence. the enriched oxide scale is accompanied by depletion of chromium in the alloy surface immediately below the scale. 677]. Handbuch der Sonderstahlkunde. Alternate heating and cooling results in much shorter life than would occur if the wire is heated continuously. (With kind permission of Springer Science and Business Media. This situation accounts for occasional rusting and otherwise poor corrosion resistance of hot-rolled stainless steels that have not been adequately pickled following high-temperature oxidation.12 LIFE TEST FOR OXIDATION-RESISTANT WIRES The merit of a particular alloy for resisting high-temperature environments. p. Chromstähle. 11.% or less of a reactive element.13 OXIDATION-RESISTANT ALLOYS 11.13. zirconium.9. Life of heat-resistant alloy wires in ASTM life test as a function of temperature (20% Cr–Ni and 5–26% Cr–Fe).9. lanthanum. such as yttrium. hafnium. or cerium.) increase in electrical resistance. is added to . 11.3RT Typical test results for life of several alloys are given in Fig. An equation related to the Arrhenius reactionrate equation expresses the dependence of life on temperature [54]: log life(h) = ΔE ′ + const 2.1 Reactive Element Effect (REE) The reactive element effect (REE) is obtained when 1 wt. (Copyright ASTM International.234 OXIDATION Figure 11. in air of 100% relative humidity at 25 °C (Brasunas and Uhlig [54]). The life of the most resistant alloys is more sensitive to temperature change (higher activation energy ΔE′) than that of alloys with inherently shorter life. Reprinted with permission. TABLE 11. when combined with silicon and nickel. the reactive element was found in the middle of the oxide scale. 56]. By using secondary ion mass spectrometry (SIMS) to establish the location of the reactive element in the oxide scale. The effect can be obtained when the reactive element is added as an alloy addition or when a coating of the reactive element oxide is applied to the surface of the alloy.OXIDATION-RESISTANT ALLOYS 235 high-temperature alloys containing chromium or aluminum. Approximate Upper Temperature Limits for Exposure of Cr–Fe Alloys to Air % Cr in Cr–Fe Alloys 4–6 9 13 17 27 Maximum Temperature in Air °C 650 750 750–800 850–900 1050–1100 °F 1200 1375 1375–1475 1550–1650 1925–2000 . The approximate upper temperature limits for exposure to air are presented in Table 11. The 12% Cr–Fe alloy is used for steam turbine blades because of excellent oxidation resistance and good physical properties. they are used for valves of internalcombustion engines. Addition of 1% yttrium to a 25% Cr–Fe alloy was reported to extend the upper limit of oxidation resistance to about 1375 °C (2500 °F) [58.4. are beneficial to oxidation resistance of chromium and chromium alloys. 11. Addition of the reactive element inhibits the diffusion of Cr and results in growth primarily by the inward diffusion of oxygen [56. Commercial chromiaforming heater alloys now all contain small amounts of one or more reactive elements. The 9–30% Cr alloys are used for furnace parts and burners. suggesting that inward oxygen transport in chromium oxide is the dominant transport mechanism. 57]. Rare-earth metal additions. 59]. The useful life of high-temperature alloys is typically limited by spalling of the oxide from the metal surface. Reactive elements greatly improve the spalling resistance of both chromia-forming and alumina-forming alloys. including gas turbine alloys [60].13.2 Chromium–Iron Alloys The 4–9% Cr alloys are widely used for oxidation resistance in oil-refinery construction. such as yttrium.4. in general. and yttrium has become a commonly used reactive element addition to high-temperature alloys [55. Applying a 4-nm-thick coating of ceria (CeO2) or yttria (Y2O3) has been found to increase the oxidation resistance of chromium and iron–chromium alloys. and sometimes other alloying elements. caused in part by aluminum nitride formation. combining the oxidation-resistant properties imparted by chromium and aluminum. but similarly has excellent physical properties.13.. probably in part because they decrease the tendency of protective oxides to spall.5% Al. nickel and high-nickel alloys are subject to intergranular attack. Nickel and high-nickel alloys tend to oxidize along grain boundaries when subject to alternate oxidation and reduction. It is a heat-resistant alloy that combines excellent oxidation resistance with good physical properties at both low and elevated temperatures. 5% Al. 2% Co alloy (Trade name: Kanthal A) is resistant up to 1300 °C. The 16% Cr. Similarly. Drawbacks in their properties are poor hightemperature strength and tendency toward embrittlement at room temperature after prolonged heating in air. and can be used in air up to a maximum temperature of about 1100 °C (2000 °F). Some heat-resistant thermocouple wires consist of nickel alloys. Electrical heating units for some stoves are fabricated of this alloy in tubular form. 0. this alloy (one U. 11. thorium. Also. this alloy is often used for construction of carburizing or nitriding furnaces. Alloying with chromium reduces this tendency. for example. among other applications. 1% Si. Oxidation resistance of the commercial alloy is considerably improved by the addition of calcium metal deoxidizer during the melting process. the 30% Cr. cerium) are also beneficial. The 10% Cr–Ni alloy (Chromel P) and 2% Al.3 Chromium–Aluminum–Iron Alloys The chromium–aluminum–iron alloys have exceptional oxidation resistance. 5. For this reason. and rare-earth metals (e. An electric heating wire of the 20% Cr–Ni alloy inside the tube is insulated from the outer sheath by powdered MgO. for furnace windings and parts and for electric-resistance elements. nickel is not usefully resistant to such atmospheres above about 315 °C . Because of its high nickel content and good strength (nickel does not readily form a carbide or nitride). the 24% Cr. Consequently. 7% Fe. This explanation was also mentioned earlier as applying to the beneficial effect of the rare-earth yttrium when alloyed with the chromium–iron alloys.5% Si alloy (Trade name: Megapyr) resists oxidation in air up to at least 1300 °C (2375 °F).g. They are used. 2% Mn. trade name: Nichrome V) is resistant in air to a maximum temperature of about 1150 °C (2100 °F).S.236 OXIDATION 11. is readily fabricated or welded. in contact with sulfur or sulfur atmospheres at elevated temperatures. bal. which is said to avoid oxidation of the alloy along grain boundaries. Small amounts of zirconium.4 Nickel and Nickel Alloys The good oxidation resistance of nickel is improved by adding 20% Cr. furnace windings must be well-supported and are usually corrugated in order to allow for expansion and contraction. Ni alloy (Alumel) can be exposed to air at a maximum temperature of about 1100 °C (2000 °F). 76% Ni alloy (Trade name: Inconel 600) is slightly less resistant to oxidation than the 20% Cr–Ni alloy.13. Vacuum 65. in Metals Handbook. 4. p. 1. pp. Brockway and A. Rep. MacRae. such furnace windings are blanketed in a protective atmosphere of hydrogen. A. Part 1. High-Temperature Oxidation of Metals. Their susceptibility to hot corrosion and sulfidation has already been discussed. 193. Because molybdenum oxidizes in air.5 Furnace Windings Twenty percent chromium–nickel and various chromium–aluminum–iron alloys are in common use as furnace windings. Mishra. New York. Proceedings. 11. OH. 2003. Cabrera and N. Saltsburg et al. 147. Z. Molybdenum-wound furnaces are operated to at least 1500 °C (2725 °F). in Fundamentals of Gas–Surface Interactions. New York. This alloy performs better than pure platinum because of higher strength and a lower rate of grain growth. Uhlig. several percent aluminum. Surface Sci. J. 2. 12. UetikonZuerich. 14. 4. Per Kofstad. 13. 53. Academic Press. 541 (1956). 6th International Symposium on High Temperature Corrosion and Protection of Materials. H. 2nd edition. Wiley. Desk Edition. 1998. High temperature corrosion and protection of materials 6. Duret. Materials Park. Chim. REFERENCES 1. 12. 319 (1964). J. pp. editors. Coddet. Hejwowski and A. A single crystal of the same dimensions as the resistance wire cross section tends to shear easily and cause failure. L. Switzerland. . Forum 154. 534 (1923). 63. Weronski. 325 (1967). Inst. Phys. 2004. 252–254. 1966. ASM International. Rowe. Les Embiez. 7. and a few hundredths percent yttrium. H. OH. 7. Robert A.13. and Protection. 5. Bedworth. 163 (1949). For best resistance to sulfur-containing environments. Trans Tech Publications. Materials Park. May 2004. 799 (1959). 6.. Elektrochem. Mater. F. Phys. 13A. Gronlund. a 10% Rh–Pt alloy can be used up to at least 1400 °C (2550 °F). 3. H. Testing. ASM Handbook. Sci. Corrosion: Fundamentals. ASM International. 11. T. Bénard. Gronlund. 718. 9. H. N. To achieve higher temperatures in air. N. 660 (1956). Sci. Oudar. H. Acta Metall. B.REFERENCES 237 (600 °F). Corros. J. Progr. F. J. Applied coatings of aluminum or of aluminum– chromium–yttrium increase resistance to attack. 10. Metals 29. 119 (1994). 427 (2002). Rapp. C. 8. and M. 1967. Gas turbine blades are essentially nickel-base or cobalt-base alloys containing substantial amounts of chromium. Pilling and N. Vol. France. iron-base alloys should contain high chromium and low nickel. 228–229. Uhlig. p. Mott. p. Melford. 14. and J. Grauer and W. Electrochem. Phys. T. 167 (1950). Z. 46. 29. Swanson and H. Hauffe. Grünewald. Ref. 17. J. Parker. 105. Pfeiffer. Hoar and L. F. 19 (3). Washington. 118. Wagner. D. 9. 21B. 33. Allg. Uhlig. 50. Thomas. 44..238 OXIDATION 15. J. No. P. 55 (1965). Gilroy and J. Schein. Min. C. Z. Uhlig. P. 34. R. M. 51. J. Soc. 21. Electrochem. Corros. Eng. 301 (1966). 31. Tammann and W. C. 40. Ilschner-Gensch and C. Köster. C. Perf. Sci. 196. J. Luthra and D. K. 113–127. P. 30. 196 (1922). 520 (1929). Metall. 47. Mitt. 32. Corrosion 23. 163. Bergman. 153–173. Cohen. P. 28. Corrosion 9. Corros. 41. 1 (1933). editors. Nishimura. 635 (1958). 623 (1947). Acta Metall. 23. 382 (1965). Metals 44. 110. Soc. Shores. 851 (1938). Rathenau and J. Soc. J. Rapp. 147. F. Goodenough and M. 42. LeBoucher. Metal Progr. 257 (1961). Solid State Chemistry of Energy Conversion and Storage. Soc. 100. Hauffe and H. 1325 (1971). 1977. 547 (1947).C. pp. 44 (2). M. Rapp. 49. AIME 221. 1. Chem. Brasunas and N. Pickett. Z. 7. J.. Trans. K. 6. E. 19. Am. 25 (1933). 127. Min. 24. Sci. C. Electrochem. J. Zastawnik. C. Wagner and K. 142 (1961). A. Sci. Metall. Gensch and K. (1980). 123. 72 (1967). L. J. Krupp 1. H. 455 (1938). Min. Gwathmey. K. Eng. Acad. 205. G. November. . Metallurgia 42. D. Mater. F. Phys. T. Zimens. Metall. Campbell and U. 36. American Chemical Society. 390 (1952). Simnad. Soc. Jr. Electrochem. Birchenall. 1117 (1952). Demo. Soc. Z. Corrosion 21. Faraday Soc. 889 (1951). 20. Sci. Elliott. J. Monkman and N. 6. Tech. 1951. Trans. 4157 (1961). Soc. H. W. 34. Inst. 111 (1959). Werber. Robert A. J. Advances in Chemistry Series. A. Trans. J. N. 461 (1963). Grünewald. 39. Acta Chem. Electrochem. Rhines et al. and A. Lawless. 460 (1953). Electrochem. Corros. Lustman. Chem. C. Iron Steel Inst. 353 (1960). Metall. Price. G. Am. Elektrochem. Eng. 198. 25. 40B. 107. Cleveland. Wood and D. Corros. 850 (1946). 43. Trans. A. Martin and E. 269 (1943). J. J. R. Ziegler. 45. Trans. Sci. 91. 152. J. 18. Faraday Soc. 22. 56. Chem. Meijering. American Society for Metals. Mayne. 655 (1938). D. 209 (2002). Wagner. 267 (1932). 48. Anorg. Metals 3. Wagner. B. and P. Wagner. Electrochem. Grant. Whittingham. 26. 867 (1938). Soc. Atom Movements. MacNairn. 111. Davies.. C. C. Wagner and K. 52. Feitknecht. Fry. T. 35. 5. and C. Iron Steel Inst. 34. Grant. Mrowec. 501 (1964). Kyoto 9. Egan and A. G. Phys. (Paris) 252. Pfeil. Chem. Min. 37. Metall. 427 (1950). J. Sewell and M. pp. 2202 (1980). 318 (1942) R. Jorgensen. Z. Trans. 198. 27. and N. 38. B. LaCombe.. N. 119. Scand. S. Mendizza. 47 (1966). Soc. OH. Wagner and K. 16. Proceedings from Materials Solutions Conference 2001. OH. GENERAL REFERENCES N. M. ASTM B76-90(2007). Kofstad. B. ASM International. 58. . Materials Science and Engineering. May 1952. Meier. Wright. London. National Association of Corrosion Engineers. Interscience. Elsevier. Pettit. R. 1961. 71. 60. 57. I. High-temperature corrosion-related failures. Trans Tech Publications. 54. Uetikon-Zuerich. 1989. Wright. 1966. 129 (1989). Uhlig. E. Les Embiez. I. Tortorelli. p. A. G. 182. G. Introduction to the High Temperature Oxidation of Metals. McGurty. Calculate Md/nmD for aluminum forming Al2O3 and for sodium forming Na2O. 55. 2002. ASTM. 2003. A. Fox and J. 11. Standard Test Method for Accelerated Life of Nickel–Chromium and Nickel–Chromium–Iron Alloys for Electrical Heating. in ASM Handbook. West Conshohocken. Y. Birks. and P. Corrosion 59 (6). pp. Wiley. Vol. 56. J. Proceedings of the 6th International Symposium on High Temperature Corrosion and Protection of Materials. editors. H. Proceedings. 2006. G. H. Cambridge University Press. J. John Stringer Symposium on High Temperature Corrosion. Engi. May 1989. 475 (2003). Pieraggi. 207. ASTM Bulletin. Materials Park. 108. Graham. Materials Park. PA. Houston. R. Stringer. ASTM.K. P. 1988. U. F. High-Temperature Oxidation of Metals. PROBLEMS 1. S. and R. Indicate whether the oxides are protective. B. 2nd International Symposium on High Temperature Corrosion of Advanced Materials and Coatings. Harwood. G. Brasunas and H. Podor. November 15. Pond and D. M. Philadelphia. 2003. 2004. Meier. P. Les Embiez. Switzerland. New York. 490 (1961). editor. Electrochem. France. J.PROBLEMS 239 53. A. A120/121. editors. Mater. P. A. Galerie. 59. 359 (1968). Steinmetz. High Temperature Corrosion. Semchyshen and J. ASM International. Shifler. Kofstad. p. Failure Analysis and Prevention. OH. Rapp. Soc. in Refractory Metals and Alloys. High Temperature Corrosion and Protection of Materials 6. Hou. May 2004. A120. Materials Park. TX. Viswanathan. J. OH. B. 2nd edition.. P. Cambridge. New York. 868–880. Corrosion 24. 1983. Fellen. and F. Sci. No. France. ASM International. editors. in John Stringer Symposium on High Temperature Corrosion. R. deS. Proceedings from Materials Solutions Conference 2001. High Temperature Corrosion. Pint. Section 11. chromium.18 g/cm3) that results from the oxidation of 1 cm3 of iron. Assume that NiO in borax is saturated.7 4. (b) Rate is proportional to po21/6.24 g/cm3).8 −137.1 cm3. Compare with reported values as follows: Temperature (K) 1000 1100 ΔG ° (kJ/mole) −146. Using the Gibbs free energy of formation versus temperature diagram. (c) Same as above.7 (Section 11. but.6]? (b) Make the same calculation for nickel.1cm3 Calculated ΔG ° = −135 kJ/mole (a) Rate is proportional to po21/8. (b) 2.7 cm3.28. estimate (a) the minimum pressures of oxygen required to oxidize aluminum. (c) 2. (b) the maximum ratios of H2/H2O required to produce the oxides in (a). 4.95 g/cm3) that results from the oxidation of 1 cm3 of iron. Fig.58 (a) 1.7). 3. 1. Estimate the free energy of formation of NiO at 780 °C from the cell illustrated in Fig. at 600 °C. 5.240 OXIDATION 2. (b) Similarly. and SiO2.3). and silicon to Al2O3. calculate the volume of Fe3O4 (d = 5. for Fe2O3 (d = 5. (a) Calculate the volume of FeO (d = 5. (a) What is the quantitative relation [use Eq.2. Copper oxidizes within the parabolic range at a higher rate the higher the oxygen pressure. 0. and (c) the maximum ratios of CO/CO2 required to produce the oxides in (a). (11. Cr2O3. Answers to Problems 1. 11. 2. 11. 3. . Damage by alternating current (ac) is less than that by direct current (dc). either because their magnitude is small or because their duration is short. or partially reversible. Under some conditions. This probably means that rates of reversible.1 INTRODUCTION Stray electric currents are those that follow paths other than the intended circuit. they cause corrosion at areas where the currents leave to enter the soil or water. anode–cathode reactions in the aerated solution are symmetrical with respect to alternating applied potentials. Uhlig Copyright © 2008 John Wiley & Sons. Winston Revie and Herbert H. Jones [2] reported that in 0. or they may be any extraneous currents in the earth. but severalfold higher (but still low) in a deaerated solution. with the resultant corrosion usually being greater for lower frequency and less for higher-frequency currents. If currents of this kind enter a metal structure. pipelines can incur considerable corrosion damage as a result of telluric currents—that is.12 STRAY-CURRENT CORROSION 12. natural earth currents are not important from a corrosion standpoint. the increased corrosion rate of a carbon steel caused by a 60-cycle current density of 300 A/m2 is negligible in an aerated solution. by R. Inc. currents induced in the steel pipeline by changes in the geomagnetic field of the earth [1]. 241 . largely because of the hydrogen evolution Corrosion and Corrosion Control. whereas in the deaerated solution the reactions are not symmetrical.1N NaCl. Usually. Sources of ac stray currents are usually grounded ac power lines or currents induced in a pipeline by parallel power lines.1. cathodic protection systems. and copper in common environments. similarly. Section 6. but the critical current density remained about the same.5M Na2SO4. Stray-current corrosion of a buried pipe. and electroplating plants. aluminum in dilute salt solutions at 15 A/ m2 (1. 12. grounded electric dc power lines. An example of dc stray current from an electric street railway system in which the steel rails are used for current return to the generating station is shown in Fig. electric welding machines. pH 7. The effect of unsymmetrical reactions (Faradaic rectification) is observed especially in the ac corrosion of metals that are passive (mostly by Definition 1. They also found that the corrosion potential was shifted several tenths of a volt in the active direction by an ac current density of 500 A/m2. and at 100 A/m2 (9. .242 STR AY-CURRENT CORROSION reaction. These results suggest that passivity lost during the cathodic cycle. can account for unexpectedly high corrosion rates of otherwise corrosion-resistant metals in aqueous media or in soil.2 SOURCES OF STRAY CURRENTS Sources of dc stray currents are commonly electric railways. It has been estimated that for metals like steel. Because of poor Figure 12. Feller and Rückert [4] found that the passive region of potentiostatic polarization curves for nickel in 1N H2SO4 completely disappeared and that high anodic current densities persisted throughout the noble potential region. In a study of 1-V. if not restored during the anodic cycle. 12.3 A/ft2) as much as 31% of the corrosion damage caused by equivalent dc current densities. 60-cycle alternating current (ac) causes less than 1% of the damage caused by an equivalent dc current [3]. Chin and Fu [5] found analogous behavior of mild steel in 0. reaching the same order as the critical current density (no passive region) at a 60-cycle current density of 2000 A/m2.4 A/ft2) suffers 5%. The Flade-potential region simultaneously was shifted in the active direction.1. 54cycle ac current superimposed on dc current. It has been reported that stainless steels subjected to ac current electrolysis are corroded [4]. lead. The passive current density increased with increasing superimposed ac current.1). for example. accelerating penetration of the pipe. combined with poor insulation of rails to the earth. since the resistance of any conductor per unit length equals ρ/A. Stray-current damage to a ship by a welding generator. Assuming that the cross-sectional area of water is 10 times that of the steel pipe. can produce damage similar to that by the railway illustrated in Fig. Current flowing long a water pipe (e.SOURCES OF STR AY CURRENTS 243 bonding between rails.2. For iron. 12. . This small current leaving the Figure 12. In this case. Another example of stray-current corrosion is illustrated in Fig. it is seen that if current through the pipe is 1 A. when located in the neighborhood of an unprotected pipeline.. to the contrary. some of the return current enters the soil and finds a low-resistance path. Also. but the problems of stray-current corrosion originating from metropolitan railway transit systems continue [6]. since ac currents leaking to ground cause less stray-current damage. used as an electric ground) usually causes no damage inside the pipe because of the high electrical conductivity of steel or copper compared to water. only about 10−8 A flows through the water. respectively.2. If B coats the pipe. ρM = 10−5 Ω-cm. it is better to place the generator on board ship and bring ac power leads to the generator. such as a buried gas or water main. 12. then the ratio of current carried by a metal pipe compared to that carried by the water it contains is equal to ρWAM/ρMAW. which is an understandable layman’s reaction to any corrosion problem. is harassed by corrosion failures because the service pipe of his house is anodic with respect to the rails. where ρ is the resistivity and A is the crosssectional area. The owner of a household water service pipe at A benefits by cathodic protection and experiences no corrosion difficulty. ρW may be 104 Ω-cm. Street railways have now in large part been replaced by other forms of transportation. cathodically protected structures requiring high currents. The basic rule: Never coat the anode. and for a potable water. but owner B. matters are made worse because all stray currents now leave the pipe at defects in the insulating coating at high-current densities. A welding motor generator located on shore with grounded dc lines to a ship under repair can cause serious damage to the hull of the ship by current returning in part from the welding electrodes through the ship and through the water to the shore installation.1.g. where the subscripts W and M refer to water and metal. Weight Loss of Metals by Stray-Current Corrosion Metal Equivalent Weight Weight of Metal Corroded per Ampere-Year Pounds Fe Cu Pb Zn Al 55.4 33.244 STR AY-CURRENT CORROSION Figure 12. 12.38/2 26.85/2 63. At high current densities in some environments. If seawater is transported instead.8 10.2/2 65. with ρW = 20 Ω-cm. the ratio of currents is 2 × 105. oxygen may be evolved.7 2.5 Kilograms 9.3).8 74.57/2 207. corrosion is now focused on the water side of the joint where any current that persists leaves the pipe to enter the water.3. if a high-resistance joint exists between two sections of a buried pipe.3 QUANTITATIVE DAMAGE BY STRAY CURRENTS In general. where such currents leave the pipe and enter the soil.5 23. Effect of current flowing along a buried pipeline on corrosion near insulated couplings. At low current densities. If insulating joints are installed in the above-mentioned pipe in order to reduce stray-current pickup.1 10. 12. .1 22.1. even in this case. Weight losses of typical metals for the equivalent of 1 A flowing for one year are listed in Table 12. Or. local-action corrosion supplements stray-current corrosion.1. However.98/3 20. indicating that. stray-current corrosion of the outer surface may be appreciable. TABLE 12. most of the current is carried by the metallic pipe and there is very little stray-current corrosion on the inner surface.6 6. reducing the amount of metal corroding per faraday of electricity.9 pipe and entering the water causes negligible corrosion. corrosion may be more pronounced on the side where current enters the soil (Fig. the amount of metal corroding at anodic areas because of stray current can be calculated using Faraday’s law. which are relatively steady. stray currents can often be detected and their origin can be traced to the generator source for which the load. which are all soluble in excess alkali. we have Δφ = 0. we obtain Δφ = ρj y2 + h2 ln 2π h2 (12.734ρj (12. form in variable amounts per faraday.. The amount of cathodic corrosion is not readily estimated. h the depth of pipe below the surface. 12. see Appendix.m.2) An outer reference electrode (+) with respect to a (−) reference electrode directly over the pipe corresponds to current entering the pipe.m. may be damaged at cathodic areas where alkalies accumulate by electrolysis. and y the distance along the soil surface over which the potential difference is measured. This is in contrast to galvanic or cathodic protection currents. are larger at 7 to 9 a. Section 29. varies in a pattern similar to that of the measured potential change. by recording the potential of a corroding system with respect to a reference electrode over a 24-h period. Hence. Formulas for such compounds may also vary with conditions of formation. This damage is in addition to damage at anodic areas. observing whether the potential of the corroded structure also fluctuates at the same frequency.5). and zincates (Na2ZnO2). aluminum. a street railway is suspected. if stray currents. hence. . depending on diffusion rates. temperature.g. the protective current can be turned off and on briefly at regular intervals. day or night. Plumbites (NaHPbO2). lead. If Δφ is the measured potential difference. ρ the resistivity of soil. If interference by a cathodic protection system is suspected. In general. and other factors. the compounds hydrolyze at lower values of pH some distance away from the cathode to form insoluble metal oxides or hydroxides. tin. stannates (Na2SnO3).4 DETECTION OF STRAY CURRENTS Stray currents may fluctuate over short or long intervals of time parallel to the varying load of the power source.. as indicated by potential measurements. The magnitude of current leaving (or entering) a buried pipe from whatever source can be calculated by measuring the potential difference between a position on the soil surface directly over the pipe and a position on the soil surface some distance away and at right angles to the pipe. zinc) corrode in alkalies as well as in acids and. and again at 4 to 6 p.1) where j is the total current entering or leaving the pipe surface per unit length (for derivation. Thus. If y is chosen equal to 10h.DETECTION OF STR AY CURRENTS 245 Amphoteric metals (e. aluminates (NaH2AlO3). .246 STR AY-CURRENT CORROSION Figure 12. Then ρ= 2 πaΔφ I (12.1. 12. In Fig. Δφ. 12. the bond may include a resistor just sufficient to avoid large change of potential of the unprotected system when the cathodic protection current is turned on and off. of the two inner reference electrodes (e. Section 29. Bonding.6 MEANS FOR REDUCING STRAY-CURRENT CORROSION 1. 12. and the potential difference. At the same time. In the event of a cathodic protection system causing the damage.5 SOIL-RESISTIVITY MEASUREMENT The resistivity of a soil can be measured by the four-electrode method with each electrode arranged in a straight line and each separated by the distance a (Fig. The resistor avoids major damage to the unprotected system. 12. Steady current I from a battery is passed through the two outer metal electrodes. which in effect serves to protect the adjoining system in addition to the protected structure.4). The measurement is usually repeated with current direction reversed in order to cancel out any stray currents. it avoids the necessity of a large increase in protective current caused by current flowing through the bond. If stray currents for some reason should periodically reverse direction. Cu−CuSO4) is measured simultaneously.g. rectifying units (diodes) are inserted into the bonding connection. stray-current corrosion is completely avoided by placing a low-resistance metallic conductor between service pipe B and the rails at C. This is called bonding the rail and pipe systems.6). Four-electrode method for measuring soil resistivity. thereby ensuring that any current flow is in that direction that is nondamaging to the structure..4.3) where ρ is the resistivity of a uniform soil measured to an approximate depth equal to a (for derivation. see Appendix. (a) A direct current of 10 A enters and leaves a steel water pipe of 2-in. in Peabody’s Control of Pipeline Corrosion. Insulating Couplings. PROBLEMS 1. REFERENCES 1. 12. Calculate the current carried by the steel and that carried by the water. Present and Future of Rail Transit Systems. 67 (1969). Szeliga. H. in Uhlig’s Corrosion Handbook. If a source of dc current is inserted between the intentional anode and the pipe such that current flows in the soil in a direction opposite to that of the stray current. NACE International. Such couplings are often useful for minimizing stray-current damage.3. 4. wall thickness. 3. D. outside diameter and 0. TX. 27 (Sept. 428 (1978). Stray current analysis. Feller and J. Chin and T. Stray Current Corrosion: The Past. editor. containing water with resistivity of 104 Ω-cm. Corrosion 34. W. 1994. Fitzgerald III. 5. They are less useful if voltages are so large that current is induced to flow around the insulating joint. 2001. D. Cathodic protection is installed whenever the intentional anode is not sufficient to overcome all corrosion caused by stray currents. Jones. M. By installing one or more insulating couplings.1 is not feasible. 751 (1998). an intentional anode of scrap iron may be buried in the direction of the rails and attached by a copper conductor to point B. Corrosion 54(9). D. Mater. 2nd edition. Stray currents then cause corrosion only of the intentional anode. 8(9).1 becomes a less favorable path for stray currents. H. . which is easily replaced at low cost. 12. Houston. 58. Stray current corrosion. Kish. Perf. L. 635 (1967). 12. R. H. causing corrosion as depicted in Fig.25-in. pp. Assume that the resistivity of the pipe equals 10−5 Ω-cm. 3. Fu. editor. J. 2. Metallk. pp. Michael J. Revie. Corrosion 35. R. Rückert. editor. Hewes. the pipeline in Fig. 1079–1087. 2nd edition. F. Boteler and W. the arrangement is equivalent to cathodically protecting the pipe. GENERAL REFERENCES J. Wiley. Szeliga. Intentional Anodes and Cathodic Protection. H. 2000. Texas. 1981). 20 (9). Seager. G. 211–236. Prot. Houston. Bianchetti. If a bond from B to C in Fig. New York. NACE International. 6. Mater.PROBLEMS 247 2. 514 (1979). Z. (a) Does current flow from or to the pipe? How many amperes per linear foot of pipe? (b) What is the percent error in calculated current if it is found later that the pipe is buried 7 ft instead of 6 ft below the surface? 4. what distance between electrodes must be specified in order to apply the simplified formula: ρ= 1000Δφ I 6. (a) If a spherical electrode of radius a is half-buried in earth of resistivity σ. Assuming that a pipe is buried h meters underground in soil with a resistivity of 3000 Ω-cm. and soil resistivity measures 3000 Ω-cm. The potential difference between two Cu−CuSO4 reference electrodes located on the soil surface over the pipe and a point at right angles 60 ft distant. calculate the current if the pipe conveys seawater of resistivity 20 Ω-cm. plot ΔE versus y for a constant current. (a) A direct current of 10 A passes through external clamps 2 m apart attached to a copper pipe of 5-cm outside diameter containing water of conductivity 10−4 Ω−1 cm−1.35 cm. diameter is buried 6 ft underground. j. entering the pipe per linear length of pipe in meters. in diameter and 2 ft long. calculate the total weight if the pipe contains seawater of conductivity equal to 0. If the pipe wall thickness is 0. What is the initial corrosion rate in ipy caused by this current? 7. A stray current of 0. A long pipeline of 8-in. The electrode over the pipe is negative to the other. is 1. 2. (b) Similarly. 5.7 A leaves a section of buried steel pipe 2 in. In measuring resistivity by the four-electrode method.05 Ω−1 cm−1.248 STR AY-CURRENT CORROSION (b) Similarly. calculate the total weight of the copper in grams per year that corrodes internally at the positive clamp because of this current. (c) What weight of copper would in theory corrode in one year if an insulating joint were inserted between the clamps? 3.25 V. what is the resistance R to earth as a function of distance D from the electrode center? (b) At what distance is the resistance within 99% of maximum? . 22 ft). (a) 5. (a) 0.04 × 105 g. (b) 0.025. (c) 1.0 × 10−5. (b) 100a. 6. σ 1 1 7.33 ipy. 3. (b) 6. 5. − 2π a D ( ) . 1.PROBLEMS 249 Answers to Problems 2. (a) R = .4%. 0.0173 A/ft.59 m (5. . looked at another way. Pitting corrosion can be prevented in passive metals. corrosion is reduced essentially to zero.1 INTRODUCTION Cathodic protection is probably the most important of all approaches to corrosion control. Or. by R.13 CATHODIC PROTECTION 13. Inc. 251 . Using an externally applied electric current. and corrosion currents no longer flow. such as steel. there is no tendency for metal ions to enter into solution. copper. As discussed in Section 5. the mechanism of cathodic protection depends on external current that polarizes the entire surface to the thermodynamic potential of the anode. such as the stainless Corrosion and Corrosion Control. and brass. at a high enough value of external current density. Winston Revie and Herbert H. The surface becomes equipotential (cathode and anode potentials become equal). Uhlig Copyright © 2008 John Wiley & Sons. a net positive current enters the metal at all regions of the metal surface (including anodic areas). A metal surface that is cathodically protected can be maintained in a corrosive environment without deterioration for an indefinite time. hence. against corrosion in all soils and in almost all aqueous media. Cathodic protection can be applied in practice to protect metals. also known as galvanic cathodic protection. lead.11. There are two types of cathodic protection: impressed current cathodic protection (ICCP) and sacrificial anode cathodic protection (SACP). but the overall cathodic protection of seagoing ships was not explored again until about 1950. which supplies a sufficient concentration of copper ions to poison fouling organisms (see Section 6. the fitting of zinc slabs became traditional on all Admiralty vessels. 18–8 stainless steel). Nor does the protective current extend into electrically screened areas. A cathodically protected. Because fouling reduced the speed of ships under sail. successfully protected the iron work of buoys by attaching zinc blocks. or dezincification of brass. .C. corrosion fatigue of most metals (but not fatigue). even though the water box may be adequately protected. It can be used to avoid S. aluminum).g. contrary to the behavior of unprotected copper. After Davy’s death in 1829.1).U. Vernon and Mr. about *Communication of Dr. of water tanks) is not affected. but. the corrosion rate of copper sheathing was appreciably reduced. H. stainless steels.g. but not hydrogen cracking of such steels. Cathodic protection can be used effectively to eliminate stress-corrosion cracking (e. mild steel. a method first patented in France in 1836 and in England in 1837 [3]. When wooden hulls were replaced by steel.. of high-strength steels. The first application of impressed electric current for protection of underground structures took place in England and in the United States. of brass. because the impressed current cannot reach metal areas that are out of contact with the electrolyte.252 C ATHODIC PROTEC TION steels or aluminum.H.C. Cathodic protection was incidental to the mechanism of protecting steel sheet coated by dipping into molten zinc (galvanizing)(see Section 14. He recommended cathodic protection of coppersheathed ships. Corrosion above the water line (e. and Robert Mallet. his cousin.* These slabs provided localized protection. Kenworthy with H. However. it was shown that cathodic protection of ships is feasible and can make possible appreciable savings in maintenance costs.3. especially against the galvanic effects of the bronze propeller. cathodically protected copper is subject to fouling by marine organisms.3). employing sacrificial blocks of iron attached to the hull in the ratio of iron to copper surface of about 1 : 100. W. produced a zinc alloy particularly suited as a sacrificial anode. Sir Humphry Davy [1] reported in 1824 that copper could be successfully protected against corrosion by coupling it to iron or zinc. hull also reduces fuel costs of ship operation. magnesium. such as the interior of water condenser tubes (unless the auxiliary anode enters the tubes).g.8. By using antifouling paints in combination with anticorrosion paints.. the British Admiralty decided against the idea. L. this time by the Canadian Navy [2]. of Duralumin. as Davy had predicted.2 BRIEF HISTORY As a result of laboratory experiments in salt water.. unfortunately. hence smooth. In practice. intergranular corrosion (e. 13. Edmund Davy (Professor of Chemistry at the Royal Dublin University). the practice of zinc coating of steel was apparently described in France as early as 1742 [4]. in 1840. J. offshore oil-drilling structures. Since then. the applied voltage must be raised. submarines. In soils or waters of high resistivity. Sketch of cathodically protected pipe. water tanks.HOW APPLIED 253 1910–1912 [3]. cathodic protection of high-pressure oil and gas pipelines is a government-mandated regulatory requirement. marine piling. The applied voltage is not critical—it need only be sufficient to supply an adequate current density to all parts of the protected structure. A sketch of a cathodically protected buried pipeline is shown in Fig. in this way. 13. current flows from the electrode through the electrolyte to the structure.1. Or.3 HOW APPLIED Cathodic protection requires a source of direct current and an auxiliary electrode (anode) usually of iron or graphite located some distance away from the protected structure. The direct current (dc) source is connected with its positive terminal to the auxiliary electrode and its negative terminal to the structure to be protected. auxiliary anode.1. 13. Cathodic protection is also applied to canal gates. chemical equipment. In many parts of the world. . and rectifier. when the extremities of a long pipeline are to be protected by a single anode. and now tens of thousands of miles of buried pipeline and cables are effectively protected against corrosion by this means. the general use of cathodic protection has spread rapidly. condensers. Figure 13.4). parking garages. and other reinforced concrete structures (see Section 7. the applied voltage must be higher than in environments of low resistivity. bridge decks. the danger of stray-current damage (interference problems) to adjoining metal structures is reduced.. Sacrificial metals used for cathodic protection consist of magnesium-base and aluminum-base alloys and. They are useful particularly when electric power is not readily available. The opencircuit potential difference of magnesium with respect to steel is about 1 V (φH for magnesium in seawater = −1.254 C ATHODIC PROTEC TION The source of current is usually a rectifier supplying low-voltage dc of several amperes.2.1 Sacrificial Anodes If the auxiliary anode is composed of a metal more active in the Galvanic Series than the metal to be protected. Figure 13.3 V). zinc. particularly in high-resistivity soils. 13. Even in periods of calm. Sacrificial anodes serve essentially as sources of portable electrical energy. a galvanic cell is set up with current direction exactly as described in the previous section. some degree of protection of steel persists temporarily because of the inhibiting effect of alkaline electrolysis products at the cathode surface. so that only a limited length of pipeline can be protected by one anode. as shown in Fig. or in situations where it is not convenient or economical to install power lines for the purpose. 13. Motor generators have been used. Cathodically protected pipe with sacrificial anode. . This low voltage is sometimes an advantage over higher impressed voltages in that danger of overprotection to some portions of the system is less. and the electrode is called a sacrificial anode. the rectifier) can then be omitted.e.3. The impressed source of current (i. to a lesser extent. and since the total current per anode is limited. Windmill generators are employed in areas where prevailing winds are dependable.2. although maintenance is troublesome. For offshore applications. By using high-purity magnesium containing about 1% Mn. For cathodic protection of offshore platforms. the advantage of a higher potential (with higher current output per anode) is obtained [9]. A special chemical environment high in chlorides surrounding the electrode can be provided in order to avoid passivity. whereas aluminum anodes are characterized by reliable long-term performance [8]. A disadvantage of aluminum as a sacrificial anode is that it tends to become passive in water or in soils with accompanying shift of potential to a value approaching that of steel. magnesium anodes have not been popular since the 1980s. when tested in seawater for 254 days.1% Sn and 5% Zn [6. Magnesium anodes are often alloyed with 6% Al and 3% Zn to reduce pitting-type attack and to increase current efficiency. alloying aluminum with 0. where the currents naturally set up by the magnesium–iron couple reach higher values than in waters of low conductivity (soft waters).1N NaCl is −1. Such rods may increase the life of a steel tank by several years. High-purity zinc is usually specified in order to avoid significant anodic polarization with resultant reduction of current output caused by accumulation of adherent insulating zinc reaction products on commercial zinc. because of improvements in aluminum and zinc anodes [8].) compared to −0. antimony.COMBINED USE WITH COATINGS 255 The potential of zinc is less than that of magnesium (φH in seawater = −0. is only a temporary measure. made from aluminum-zinc alloys. Some sacrificial aluminum anodes contain about 0. however.5 V for pure aluminum. 13. passivity can be avoided by alloying additions. use of mercury has been banned in most locations because of pollution concerns [8]. The degree of protection is greater in waters of high conductivity. For example. or mercury.6% Zn. 0. but at somewhat lower efficiency in most soils. The corrosion potential of the 0. however. current output per anode is also less. This tendency is less pronounced in zinc of high purity. The observed efficiency of magnesium anodes averages about 500 A-h/lb compared to a theoretical efficiency of 1000 A-h/lb. such an environment.3.1% Sn followed by heat treatment at 620 °C for 16 h and water quenching to retain the tin in solid solution very much decreases anodic polarization in chloride solutions [5].2 V (S. such as tin.E. particularly if the rod is renewed as required. Another composition containing 0. A sketch of a magnesium anode rod installed in a steel hot-water tank is shown in Fig. This alloy operates at a current efficiency in seawater similar to the first-mentioned alloy.H. indium. hence. called backfill. are the preferred material [8]. Aluminum operates theoretically at a voltage between magnesium and zinc. Magnesium anodes may be consumed before the structure has reached the end of its lifetime.1% Sn alloy in 0.8 V). 13. operated at a current efficiency of 94% (1270 Ah/lb).06% Fe. In seawater.04% Hg. 7].4 COMBINED USE WITH COATINGS The distribution of current in a cathodically protected steel water tank is not ideal—too much current may flow to the sides and not enough to the top and . aluminum anodes. and 0. since the total required current is less than that for an uncoated tank. it is expedient to use an insulating coating. the magnesium anode lasts longer. and this .256 C ATHODIC PROTEC TION Figure 13. since the protective current flows preferably to exposed metal areas. Better distribution is accomplished by using an insulating coating (e.g. wherever located.. a partially protective coating may form on steel that consists largely of CaCO3 precipitated by alkalies generated as reaction products at the cathode surface. which are precisely the areas needing protection. bottom. are also useful in distributing the protective current and in reducing total current requirements. Cathodically protected hot-water tank with magnesium anode.3. if adherent. The insulating coating need not be pore-free. In the general application of cathodic protection using either impressed current or sacrificial anodes. Also. an organic coating at ordinary temperatures or a glass coating at elevated temperatures). In hard waters. A similar coating forms gradually on cathodically protected surfaces exposed to seawater (more rapidly at high current densities). Such coatings. and z is the resistance of pipe coating per unit area (for derivation. 5. The potential decay. Using an impressed current at higher applied voltages. the higher must be the impressed current density for protection. Considering a finite pipeline for which a/2 is half the distance to the next point of bonding.15 in Section 5. one magnesium anode can protect as much as 8 km (5 miles) of a coated pipeline. along an infinite pipeline measured from the point of attachment to the dc source having potential EA is expressed as an exponential relation with respect to distance.1) Both Ex and EA represent differences between polarized potential with current flowing and corrosion potential in absence of current. 13. RL. ⎡ 2 πrRL Ex = EB cosh ⎢ ⎣ kz ( ) (x − a )⎤⎥⎦ 2 12 (13.11 that the applied current density must always exceed the current density equivalent to the measured corrosion rate in the same environment.5 MAGNITUDE OF CURRENT REQUIRED The current density required for complete protection depends on the metal and on the environment. Hence.MAGNITUDE OF CURRENT REQUIRED 257 combination is the accepted practice today. For example. the distribution of current to a coated pipeline is much improved over that to a bare pipeline.2) Cathodic protection is optimum within a specific potential range (see Section 13. is a good electrical conductor. and coating resistance. and potential Ex at a/2 = EB. For a bare pipeline. see Appendix. increases. Ex.4). x. It can be seen from Fig. and the total length of pipeline protected by one anode is much greater. z. the corresponding distance might be only 30 m (100 ft). so that the length of pipeline protected by one anode increases as the metallic pipe resistance. Note that Ex becomes zero at x = ∞. decreases.7). one anode might protect as much as 80 km (50 miles) of a coated pipeline. This equation is derived by assuming that polarization of the cathodically protected surface is a linear function of current density. When the corrosion rate is cathodically controlled and the corrosion potential approaches the open-circuit anode potential. and the resistivity of the soil is localized only within the region of the pipeline or the electrodes. along the pipeline in accord with ⎡ 2 πrRL Ex = EA exp ⎢ − kz ⎣ ( ) 12 ⎤ x⎥ ⎦ (13. the required current density is . The limiting length of pipe protected per anode is imposed not by resistance of the soil. k is a constant. Since the earth. taken as a whole. but by the metallic resistance of the pipeline itself. RL is the resistance of pipe of radius r per unit length. the greater the corrosion rate. Section 29. the total current and required number of anodes are less. 015 0.258 C ATHODIC PROTEC TION TABLE 13. platinum-clad copper. Scrap iron is consumed at the rate of 15–20 lb/A-year and must be renewed periodically. Graphite anodes are consumed at a lower rate. which have also been recommended. 2% Ag−Pb.7). or platinized niobium have been recommended as corrosionresistant anodes using impressed current [10–12]. only orders of magnitude can be given. For protection of structures in seawater.1. the total required current. The precise requirements of current density for complete protection can be determined in several ways.015 0. platinized titanium. the required current can be considerably greater than the corrosion current. In the absence of such measurements. If the protective current induces precipitation of an inorganic scale on the cathode surface.15 (initial) 0. the 2% Ag−Pb anodes are estimated to last more than 10 years. These are listed in Table 13.1 for steel exposed to various environments. only the total current density per apparent unit area is less.6 ANODE MATERIALS AND BACKFILL Auxiliary anodes for use with impressed current are usually composed of scrap iron or graphite. 13. not exceeding perhaps 2 lb/A-year. both initially and in subsequent higher electrical power costs because of the noble potential and accompanying high overpotential of oxygen (or Cl2) evolution on graphite compared to an active potential and lower overvoltage for the reaction Fe → Fe2+ + 2e−. The advantages and disadvantages of graphite apply in similar measure to 14% Si−Fe alloy anodes and magnetite anodes. However. and it is still greater for corrosion reactions that are anodically controlled.05 A/ft2 35 0. such as in hard waters or in seawater. Graphite is also fragile compared to scrap iron and must be installed with greater care. the most important is potential measurement of the protected structure (see Section 13. falls off as the scale builds up.003 0.005 only slightly greater than the equivalent corrosion current. Whereas sacrificial magnesium anodes require replacement approximately every 2 years.5 0.03 (final) 0.001–0. the current density remains the same as before scale formation. But for mixed control.05 0. at exposed areas of metal.15 0. as described earlier.01–0. and the 90% Pt–10% Ir anodes still . Orders of Magnitude of Current Density Required for Cathodic Protection of Steel Environment Sulfuric acid pickle (hot) Soils Moving seawater Air-saturated water (hot) Moving fresh water A/m2 400 0. Graphite costs more than scrap iron. 5. A suitable backfill may consist of approximately 20% bentonite (an inorganic colloid used for retention of moisture). In general. Accordingly.6. to formation of colloidal metal particles [13.ANODE MATERIALS AND BACKFILL 259 longer [11]. as well as increasing conductivity of the immediate environment. Sometimes. The latter ions are unstable and react in part with water in accord with 2Mg + + 2 H 2O → Mg(OH)2 + Mg 2+ + H 2 Hence. The main disadvantages are waste of electric power and increased consumption of auxiliary anodes. Whereas auxiliary anodes need not be consumed in order to fulfill their purpose.4) is caused. aluminum anodes are sometimes used for impressed current systems. the observed yield of a magnesium anode is only about one-half the 1000 A-h/lb calculated on the basis of Mg2+ formation. 14] or. about half the magnesium corroding anodically appears as Mg(OH)2 and half as MgCl2. For impressed current systems. hydrogen embrittlement of the steel (loss of ductility through absorption of hydrogen). being a conductor. and 5% Na2SO4. such as Mg(OH)2. Additional lesser side reactions may also take place at the same time. additional disadvantages result if so much hydrogen is generated at the protected structure that blistering or disbonding of organic coatings. sacrificial anodes are consumed not less than is required by Faraday’s law in order to supply an equivalent electric current. 75% gypsum. but for magnesium it is appreciable. accompanied by hydrogen evolution in about the amount expected according to this reaction [15]. In fresh waters.1 Overprotection Moderate overprotection of steel structures usually does no harm. lake. carries part of the current. In the extreme. this consists of surrounding the anode with a thick bed of coke and adding a mixture of perhaps 3 or 4 parts gypsum (CaSO4·H2O) to 1 part NaCl. reducing in some measure consumption of the anode itself. the observed rate of consumption is greater than the theoretical. The coke backfill. perhaps more important. to initial formation of univalent magnesium ions [15]. in dilute sodium chloride. Backfill may not be required if the anode is immersed in a river bed. backfill has the advantage of reducing resistance of insulating corrosion-product films. Because the effective resistivity of soil surrounding an anode is confined to the immediate region of the electrode. the backfill is packaged beforehand in a bag surrounding the anode. or the ocean. or hydrogen cracking (see Section 8. Damage to steel by hydrogen absorption is particularly apt to occur in environments containing sulfides [16] for reasons discussed in Section 5. with the cause being ascribed to localaction currents on the metal surface. For zinc the difference is not large. . it is common practice to reduce local resistance by backfill. 13. For magnesium anodes. so that anode and backfill can be placed simultaneously into position in the soil. excess alkalies generated at the surface of overprotected systems damage the metals by causing increased attack rather than reduction of corrosion. A section of buried pipeline is cleaned. For example. but the critical potential for complete protection (see below) shifts to more active values [17]. A piece of absorbent paper soaked in potassium ferricyanide solution is brought into contact. Aluminum can be cathodically protected against pitting by coupling it to zinc [18] used as a sacrificial anode. such as aluminum. but if coupled to magnesium. After exposure to the soil for a period of weeks or months. After a relatively short time. 13. examination of the paper for the blue color of ferrous ferricyanide indicates that cathodic protection is not complete. whereas absence of blue indicates satisfactory protection. Or the hull of a ship can be inspected at regular intervals for depth of pits. and it is the criterion generally accepted and used by corrosion engineers. and both the cable and surface between coupon and pipe are overlaid with coal tar. A weighed metal coupon shaped to conform to the outside surface of a buried pipe is attached by a brazed connecting cable. noting that leaks per year drop to a small number or to zero after cathodic protection is installed. It is also possible to check effectiveness of protection by short-time tests. Coupon Tests. 13. It is based on the fundamental concept that optimum cathodic protection is achieved when the protected structure is . including overprotection. is obtained through measuring the potential of the protected structure.260 C ATHODIC PROTEC TION In the case of amphoteric metals.7 CRITERIA OF PROTECTION The effectiveness of cathodic protection in practice can be established in more than one way. of the cleaned coupon is a measure of whether cathodic protection of the pipeline is complete.1 Potential Measurements A criterion that indicates degree of protection. exposing bare metal. This measurement is of greatest importance in practice. the weight loss. including the following measures: 1.7. and the soil is returned in place. It was shown that cathodic protection of lead continues into the alkaline range of pH. lead. Colorimetric Tests. the observed number of leaks in an old buried pipeline is plotted against time. and several criteria have been used in the past to prove whether protection is complete. zinc. 2. Both the coupon and the colorimetric tests are qualitative and do not provide information about whether just enough or more than enough current is being supplied. if any. and tin. overprotection may result with consequent damage to the aluminum. 91 V versus copper-saturated CuSO4. we obtain φ H = −0.). Section 3.1). This potential for steel. The latter potential lies within the passive range and is less noble the higher the Cl− concentration. for nickel it is about 0. is equal to −0. in accord with previous discussions of oxide-film composition on iron exposed to aqueous media (see Section 7. The observed more-active galvanic potential of cadmium is plausibly explained by the known tendency of Cd2+ to form complex ions that lower Cd2+ activity below the value corresponding to saturated Cd(OH)2.059 log (Fe2+ ).H. Since passive metals corrode uniformly at low rates. therefore.CRITERIA OF PROTECTION 261 polarized to the open-circuit anode potential of local-action cells. as determined empirically.8). which has a more favorable corrosion potential of about −0.59 V (S.85 V versus the coppersaturated CuSO4 half-cell. provided that the external solution in contact does not alter the natural pH of the surface metal hydroxide.4 V is not suited as a sacrificial anode for cathodically protecting aluminum.E. Values of surface pH for other corroding metals are specific to the metal. is about −0.H.78 V versus copper-saturated CuSO4 compared to the calculated value for a Pb(OH)2 film on lead equal to −0.E. or −0. but by pitting corrosion at high rates. which is noble to iron.59 V.).3. the critical potential in 3% NaCl is 0.6). and is in essential agreement with the empirical value.E. Using the Nernst equation. the calculated value comes closer to the empirical value.21 V. Coupling of the latter metals to a suitable area of either iron or zinc.53 V (S. the criterion of protection differs from that just described. equivalent to a potential difference of −0. with formation of plumbites. can effectively protect them cathodically in seawater against initiation *This relation holds for a saturated solution of Fe(OH)2 in water. In alkaline media. This pH is observed at an iron surface between an external pH of 4–10. in accord with Fe(OH)2 → Fe2+ + 2(OH−).23 V. in 3% NaCl the value for aluminum is −0. The theoretical open-circuit anode potential for iron can be calculated assuming that the activity of Fe2+ in equilibrium is determined by the solubility of a covering layer of Fe(OH)2. but only to a value more active than the critical potential at which pitting initiates (see Section 6.5. For 18–8 stainless steel. iron with a corrosion potential in seawater of about −0.2. 3. the natural pH of which is 9.* The potential so calculated is −0. contrary to an observed galvanic potential less noble than iron (see Fig.† Other calculated values are listed in Table 13.54. unlike zinc.44 + 0.H. .8 V. † A similar situation applies to Cd assumed to be coated with a film of Cd(OH)2 (solubility product = 2 × 10−14).).3. The empirical value for lead. For passive metals. as discussed in Section 7. 2 where ( Fe2+ ) = solubility product (OH −)2 The value of (OH−) can be estimated assuming that its concentration at equilibrium is twice that of (Fe2+). Hence. cathodic protection of metals like aluminum and 18–8 stainless steel depends on polarizing them not to the usual thermodynamic anode potential.2. known only approximately [19]. The calculated value of φH is −0.45 V (S. but the relation (2M2+) = (OH−) also applies. 5 × 10−17 4. This criterion is not exact and in many situations leads to under. 13. oxygen depolarization control). Components of practical structures.91 −0. such as ships and offshore oil-drilling platforms. This position is chosen because cathodic protection currents flow mostly to the lower surface and are minimal to the upper surface of a pipe buried a few feet below the soil surface. breaks may occur when concentration polarization or IR effects through partially protective surface films become appreciable. Such breaks may. occur in some environments when the applied current is just equal to or slightly greater than the corrosion current (e.6 × 10−19 4.g. In practice. even where adequate precautions are exercised otherwise..16 −1.59 Iron Copper Zinc Lead −0. It has also been suggested that polarization of the structure should proceed to breaks in slope of the voltage versus current plot.16 −0.59 0. Calculated Minimum Potential φ for Cathodic Protection Metal φ° (volts) Solubility Product. Contrary to what is sometimes erroneously supposed.3 V more active than the corrosion potential.8 × 10−15 1.2. a compromise position is chosen at the soil surface located directly over the buried pipe. Any IR drop through corrosionproduct films or insulating coatings will persist. discontinuities in slopes of polarization curves have no general relation to the anode or cathode open-circuit potentials of the corroding system.7. are sometimes designed to take advantage of galvanic couples of this kind.2 Doubtful Criteria Criteria have sometimes been suggested based on empirical rules—for example.3 Position of Reference Electrode The potential of a cathodically protected structure is determined ideally by placing the reference electrode as close as possible to the structure to avoid an error caused by IR drop through the electrolyte.440 0.27 of pitting corrosion. of course.262 C ATHODIC PROTEC TION TABLE 13.93 −0. 13.337 −0. polarizing a steel structure 0. but in other environments. tending to make the measured potential more active than the actual potential at the metal surface.or overprotection.126 −0. breaks of this kind in polarization measurements have varied causes and are of doubtful value as criteria of cathodic protection. in principle.25 −0. .7.763 −0. As Stern and Geary showed [20]. M(OH)2 1. for buried pipelines.2 × 10−15 φH (volts) φH versus Cu−CuSO4 Reference Electrode (volts) −0. and assured public safety. Anodic protection has found application in handling sulfuric acid [22]. corrosion fatigue. recommended because currents do not penetrate remote areas.g. The guarantee that no leaks will develop on the soil side of a cathodically protected buried pipeline makes it economically feasible. One important advantage is that the gates can continue to operate without the necessity of periodic long shutdowns for repairs caused by corrosion. In addition.5% the cost of replacing the gates. The Panama Canal gates. such as iron and stainless steels. but the method is also applicable to other acids (e. 6. lack of corrosion on the soil side makes it possible to specify thinner wall pipe adequate to withstand internal pressures and to avoid any extra thickness as a safety factor against corrosion. and pitting of various structural metals are substantial in terms of extended lifetime.9 ANODIC PROTECTION As mentioned in Section 6. for example. These potentials differ by the IR drop through the soil and through coatings. and of 0.5M Na2SO4 [23]. The further economic advantages. the cost of cathodic protection is far less than for any other means offering equal assurance of protection. It has been shown to be effective for increasing the resistance to corrosion fatigue of various stainless steels in 0.2). through avoiding stress-corrosion cracking. with the initial costs of installation being less than 0. This saving alone is substantial. and hence. a ship that is cathodically protected can operate. to transport oil and high-pressure natural gas across entire continents. some metals. the potential measured at a remote position is a compromise potential at some value between that of the polarized structure and the polarized auxiliary or sacrificial anode. IR drop effects are avoided. usually electronically. Actually.4. in principle. tends to be more active than the true potential of the structure. can also be protected by making them anodic and shifting their potential into the passive region of the anodic polarization curve (see Fig. in other instances. 13. for longer periods between dry-docking. The potential measured at a remote location. Similarly. 13. therefore.8 ECONOMICS OF CATHODIC PROTECTION For buried pipelines. by an instrument called the potentiostat. are protected by using impressed current. phosphoric acid) and to alkalies and some salt solutions.1. Engineering installations have been described for anodically protecting mild .19% C steel in 10% NH4NO3 [24]. in 10% H2SO4 or 10% NH4NO3. resulting in a structure that may be underprotected. enhanced reliability..ANODIC PROTECTION 263 The reference electrode is sometimes located at a position remote from the pipeline. Practical application of anodic protection and use of the potentiostat for this purpose were first suggested by Edeleanu [21]. Section 6. for example. thereby saving thousands of dollars annually. The passive potential is automatically maintained. if Cl− should contaminate the electrolyte. far exceeding similar throwing power obtained in cathodic protection. Corrosion 25. . since for cathodic protection the current cannot be less than the normal equivalent corrosion current in the same environment. which is reported to occur at room temperature in 10N H2SO4 + 0. Since passivity of iron and the stainless steels is destroyed by halide ions. Corrosion rates corresponding to these current densities are 0. Acello and N. which shift the corrosion potential into the passive region in the absence of an applied current. The cause has been ascribed to high electrical resistance of the passive film. but this is probably not correct. See S. silver. For anodic protection.).1 μA/cm2) for Type 316 stainless steel to 0. it has been reported [22] that unusual throwing power (protection at distances remote from the cathode or in electrically screened areas) is obtained. 286t (1962).E. Current densities to initiate passivity. Scully.264 C ATHODIC PROTEC TION steel against uniform corrosion in NH4NO3 fertilizer mixtures [25]. or copper-base alloys. Anodic protection of aluminum exposed to high-temperature water has been shown to be feasible (see Section 21.2). Anodic protection is applicable only to metals and alloys (mostly transition metals) which are readily passivated when anodically polarized and for which ipassive is very low. are relatively high. because measurements have shown that such resistances are typically low. Greene. But currents for maintaining passivity are usually low. copper. S2O8 . is prevented by anodically polarizing the alloy to 0.7M oxalic acid at temperatures up to 50 °C (120 °F) [27]. It is not applicable. the possible danger of pitting becomes a consideration even if the passive current density remains acceptably low. It is typical of anodic protection that corrosion rates. For stainless steels. Fe3+).5N NaCl. carbon steel in 86% spent sulfuric acid at temperatures up to 60 °C (140 °F) [26].g. with the orders of magnitude being 10−3 A/m2 (0.H.02–2. The cause instead may be related to the corrosion-inhibiting properties of anodic corrosion products released by stainless 2− 2− steels in small amounts (e. with 6 A/m2 being required for Type 316 stainless steel in 67% H2SO4 at 24 °C (75 °F). although small. contrary to the situation for cathodic protection of steel. this value of current density corresponds to the rather high corrosion rate for the active state of the alloys. Harston and J. *Stress-corrosion cracking of Type 304 stainless steel.. Corrosion 18. for example. Cr2O7 .15 A/m2 (15 μA/cm2) for mild steel [22]. and carbon steel in 0.7 V (S.* Titanium. both in 67% H2SO4. On the other hand. is passive in the presence of Cl− (low ipassive) and can be anodically protected without danger of pitting even in solutions of hydrochloric acid. In the latter case. anodic protection of these metals is not possible in hydrochloric acid or in acid chloride solutions for which the current density in the otherwise passive region is very high. to zinc. icritical. however. the required current densities in corrosive acids are usually much lower than those for cathodic protection. are not reduced to zero. J. magnesium. cadmium.1. 493 (1969). it is only necessary to operate in the potential range below the critical pitting potential for the mixed electrolyte.5 gmd. Also.1–0. which has a very noble critical pitting potential over a wide range of Cl− concentration and temperature. Prot. L. F. 36 (Dec. K. 18. 539 (1956). D. Electrochem. Nichols. ibid. Materials Park. J. Fahrney. 10. Perrigo. K. 21. 104–106. pp. 1 (1967). Stern and A. 627t (1956). and D. H. Soc. 56 (1957). J. Southampton. Kerrich. 7. 16. 1061–1078. 58t (1960). ibid. 14. Bradley. Hutchison. R. Eng. Werner. 17. Corrosion 6. Brandt. Wiley. Burns and W. (with E. G. Soc. U. J. Christie. Bruckner and K. M. 68 (1957). Conger. J. Prot.K. H. Soc. 1967. Hoey and M. 103. 73 (1966). 242–246. Davy. M. Electrochem. 24. Corrosion 15. G. R. 2nd edition. 107. DeGiorgi. Gage. Protective Coatings for Metals. 7. Corrosion 12. Am Inst. Sperry. Robinson. Cowley. Prot. Preiser and B. 581t (1959). Prot. 328–346 (1824–1825). Corrosion 15. Benedict. H. 114. P. pp. Chem. 104. 5. 73. Adey. Soc. Engineering of cathodic protection systems. 116. and J. 8. 4. Groover. Smart III. Roy. 26. W. Phys. 6. Johnson. 777 (1967). Prot. G. Electrochem. Mater. 20.. Mater. 3rd edition. p. 8. 37. 27. Soc. (with G. 13C. 98. O. ibid. R.GENER AL REFERENCES 265 REFERENCES 1. D. Soc. 2000. Marsh and E. Pryor. 18. Burgbacher. 3c (1951). J. Lynes. Mater. 245 (1958). 3. A. Corros. Prot. ASM International. Chem. WIT Press. 3. Banks and M. Soc. 105. 31 (February 1969). R. Hays. Jansson. Corrosion: Environments and Industries. in ASM Handbook. Metallurgia 50. Cathodic protection systems. 33 (1968). 596t (1959). Mater. J. Greenblatt. George. Electrochem. 7. Lennox. A. Tytell. 22. Electrochem. Barnard (with G. 114 (1951). . 911 (1941). Mater. Geary. 1965). Bruckner and O. Mears and H. 223 (1968). editors. in Uhlig’s Corrosion Handbook. W. Trans. L. 246 (1953). Keir. New York. W. Schashl. Section 1 in Simulation of Electrochemical Processes. GENERAL REFERENCES C. 9. J. 232 (1950). H. 23. 26 (1968). and R. 19. 960 (1960). Hutchison. J. J. Brebbia. 1–56. 113 (1954). 319 (1969). Zinck) Corrosion 15. Reinhold. J. 2. Br. 76t (1959). Barnard. 46 (1966). W. ibid. Vol. Trans. and P. Edeleanu. Philos. Corrosion 15. 591t (1959). 2006. Cohen. Christie and J. H. J. Peterson. W. OH. Corrosion 13. Nature 173. and J. 4. 389t (1959). (Leipzig) 234. Mater. 11. 114. pp. Spähn. Marine corrosion protection. Fitzgerald III. and R. 2005. Myles. Z. 125t (1962). C. M. Newport. S. and N. V. 9. T. and J. 15. 739 (1954). 25. 12. New York. 151–158. Christie). Electrochem. Riggs. Greenblatt). Corrosion 16. 5. 7. G. Heidersbach. M. D. 13. 5. Corrosion 15. J. in ASM Handbook. Riggs. Anodic protection. what interval of time is required between filling and emptying the tank to ensure minimum corrosion of outlet steel water piping? (Solubility of oxygen in inlet water. in ASM Handbook. OH. Smart III. Plenum Press. and J. 2006. Marine corrosion protection. 73–78. 2003.266 C ATHODIC PROTEC TION R. pp. 25 °C. as required for complete cathodic protection. pp. 2001. compare applied current. (b) The potential of the pipe with the reference electrode located far from and at right angles to the pipe. Heidersbach. 855–870. calculate the minimum initial current density (A/m2) necessary for complete cathodic protection. OH. . H. Johnson. Anodic Protection. 1981. Sketch the polarization diagram of a buried steel pipe connected to a sacrificial magnesium anode. Locke. Corrosion: Environments and Industries. and Protection. Current between a magnesium anode and a 50-gal steel tank filled with airsaturated hot water is 100 mA. J. Vol. Brandt. OH. S. Testing. Materials Park. E. Bianchetti. 2. 13A. 4. Jr. assuming Zn(OH)2 as corrosion product on the surface. Materials Park. and Protection. = 6 mL/liter. R.) 3. Cathodic protection. (b) The corrosion rate is cathodically controlled. pp. Locke. (Solubility product of Zn(OH)2 = 4. Indicate the following on the diagram: (a) The potential near the pipe surface with respect to the open-circuit potentials of anodic and cathodic areas of the pipe.) 5. Heidersbach. Vol. Theory and Practice in the Prevention of Corrosion. PROBLEMS 1. D. 2nd edition. New York. Peabody’s Control of Pipeline Corrosion.5 gmd. NACE International. L. TX. (c) The change in potential of the pipe proceeding toward the vicinity of the magnesium anode. and C. Testing. Materials Park. Corrosion: Fundamentals. ASM International. ASM International. Calculate the minimum potential versus Cu−CuSO4 reference electrode to which zinc must be polarized for complete cathodic protection. 13A. ASM International. 13C. 2003. Corrosion: Fundamentals. Houston. 851–854. Vol. Assuming that all corrosion is by oxygen depolarization. Disregarding local-action currents. H. in ASM Handbook. C. From sketches of polarization diagrams of a corroding metal. R. to the normal corrosion current when: (a) The corrosion rate is anodically controlled.5 × 10−17. O. Iron corrodes in seawater at a rate of 2. editor. pH of steel surface = 9. 4.5. φH = −0.9 h.) are evolved per square centimeter per day from a steel surface maintained at the critical potential for cathodic protection.T. 7. 5.042 V. 7. γ0. You have been asked to review a proposal to cathodically protect a structure in an aqueous environment with a pH of approximately 7. how many milliliters H2 are evolved if cathodic protection is increased to a potential 0.105 log i/(1 × 10−7).PROBLEMS 267 6.25 V 49. φo = −0. 0. −0. Calculate the minimum potential versus saturated calomel electrode to which copper must be polarized in 0. where i = A/cm2.0019 mL/cm2-day.13 mm/y.1 V more active than the critical value? [Data: Assume hydrogen overpotential (volts) on steel = 0. . Under what circumstances can cathodic protection (either sacrificial protection or impressed current protection) be used to protect an automobile from corrosion? Answers to Problems 2.) 8. 11. (b) 0. The structure will be constructed of a high-strength steel that is very susceptible to hydrogen cracking. Discuss the factors to be considered in deciding on how to protect the structure from corrosion without causing failure by hydrogen cracking. What minimum current must be applied to the couple in order to avoid corrosion of both iron and copper? (Corrosion rate of uncoupled iron in seawater is 0. A copper bar of 300-cm2 total exposed area coupled to an iron bar 50-cm2 area is immersed in seawater.342 V.59 V? (b) Similarly. 3.] 10.018 mL/cm2-day. 9.8 × 10−3 A. (a) How many milliliters H2 (S. 8. −1.P.634 V.01M In2(SO4)3 solution for complete cathodic protection? [Data: In3+ + 3e− → In.142. 6.1M CuSO4 for complete cathodic protection.] 9.01M In2(SO4)3 = 0. (a) 0.10 A/m2. 0. To what minimum potential versus saturated calomel electrode must indium be polarized in 0. . gold. the production rate that is required. electroplating. as in galvanizing steel sheet. The selection of a coating process for a specific application depends on several factors. Uhlig Copyright © 2008 John Wiley & Sons. most commonly zinc. or a batch process—for example. by R. the number of parts being produced.1 METHODS OF APPLICATION Metal coatings are applied by dipping. and fasteners [1]. the anticipated lifetime of the coated material. In electroplating. the substrate. Winston Revie and Herbert H. these coatings can also be decorative. and silver. Corrosion and Corrosion Control. such as tin–zinc. chromium. spraying. usually steel. tin. cadmium. metal is made the cathode in an aqueous electrolyte from which the coating is deposited. with a metallic luster after polishing [2]. or base. A wide of variety of coatings can be applied by electroplating—for example. galvanizing fabricated parts. bolts. Hot dipping is carried out by immersing the metal on which the coating is to be applied. 269 . cementation. and diffusion. nuts. but also aluminum and aluminum–zinc alloys.14 METALLIC COATINGS 14. nickel. Hot dipping can be either a continuous process. Although the primary purpose of electroplating coatings is to achieve corrosion resistance. including the corrosion resistance that is required. copper. zinc. and environmental considerations. as well as alloys. Inc. in a bath of the molten metal that is to constitute the coating. CrCl2. Electrogalvanizing is the electroplating of zinc on either iron or steel. brass. a plasma at a temperature of about 12. These coatings can be made adherent and of almost any desired thickness. Nickel coatings of this kind are called electroless nickel plate. powder feedstock is melted and carried by the flame onto the workpiece. the velocity of the particles that are sprayed onto the substrate to form the coating. and a stream of air propels the molten material onto the workpiece. rod. with the three main variables in each type being the temperature of the flame. . Ion implantation is a process of producing thin surface alloy coatings by bombarding the metal with ions in vacuum. for example. on impact. the feedstock is rapidly heated and propelled toward the substrate where. titanium. Coatings are also sometimes produced by gas-phase reaction. and the nature of the material that is to form the coating (i. or liquid) [4.” In all cases. There are several types of thermal spraying. The material that is to form the coating is called the “feedstock. and so on. In flame-wire spraying. the flame melts the wire.270 METALLIC COATINGS zinc–nickel. Ti. in a bath of molten calcium containing some of the coating metal in solution [6].e. Thermally sprayed coatings tend to be porous. aluminum. results in formation of a chromium–iron alloy surface containing up to 30% Cr in accord with the reaction 3 3 CrCl 2 + Fe → FeCl 3 + Cr (alloyed with Fe) 2 2 Similar surface alloys of silicon–iron containing up to 19% Si can be prepared by reaction of iron with SiCl4 at 800–900 °C (1475–1650 °F). 5]. Coatings are also produced by electroless plating—that is. For example. allowing the metal to diffuse into the base metal. In plasma spraying. under an inert atmosphere. and nickel alloys [2]. Electroplated zinc-coated sheet is widely used for exposed automobile body panels because of its uniform coating thickness and surface characteristics compared to hot-dip zinc-coated.000 °C is formed. wire. powder. and the plasma stream carries the powder feedstock to the workpiece [4]. Sometimes. nickel. Aluminum and zinc coatings can be prepared in this way. Diffusion coatings of chromium. In thermal spraying of metal coatings. Such coatings of.. when volatilized and passed over steel at about 1000 °C (1800 °F). pores are filled with a thermoplastic resin in order to increase corrosion protection. although porosity can be controlled by optimizing the process variables [4]. and they can be applied on already fabricated structures. gold alloys. by chemical reduction of metal–salt solutions. In flame-powder spraying. it consolidates forming an adherent coating. Coatings range in thickness from 4 to 14 μm [3]. Cementation consists of tumbling the work in a mixture of metal powder and a flux at elevated temperatures. with the precipitated metal forming an adherent overlay on the base metal. B. a gun is used that simultaneously melts and propels small droplets of metal onto the surface to be coated. can also be prepared by immersing metal parts. bronze. galvanic action at the base of a pore or scratch becomes an important factor in determining coating performance.g.. 8]. nickel. 14. As the names imply. to delay access of water to the underlying metal. The degree of Figure 14.. as a result. . copper. zinc. Sketch of current flow at defects in noble and sacrificial coatings.g. 14. cadmium) and.1. metal coatings can be divided into two classes. consequently.CL ASSIFICATION OF COATINGS 271 Cr. Sometimes. they provide a barrier between the corrosive environment and the metal substrate. the base metal is cathodically protected (Fig. it is important that noble coatings always be prepared with a minimum number of pores and that any existing pores be as small as possible. which provide only barrier protection. 14. noble coatings (e.2).2 CLASSIFICATION OF COATINGS All coatings provide barrier protection. in addition to barrier protection. For sacrificial coatings (e. or chromium) on steel are noble in the Galvanic Series with respect to the base metal.g.. From the corrosion standpoint. that is. which.1 and 14. in certain environments. the pores are filled with an organic lacquer. coatings tend to become damaged during shipment or in use. At exposed pores. corrosion of the base metal does not occur. lead. the direction of galvanic current accelerates attack of the base metal and eventually undermines the coating (Figs. namely. As long as adequate current flows and the coating remains in electrical contact. or a second metal. also aluminum and tin on steel. is diffused into the coating at elevated temperatures (e. all commercially prepared metal coatings are porous to some degree. and sacrificial coatings.1). This usually means increasing the thickness of coating. Therefore. also provide cathodic protection. Furthermore. silver. noble coatings. of lower melting point. the direction of galvanic current through the electrolyte is from coating to base metal. zinc or tin into nickel). however. or Y have specialized applications for wear and high-temperature oxidation resistance [7. The metal is plated either directly on steel or sometimes over an intermediate coating of copper. contrary to the situation for noble coatings. In general.1 Nickel Coatings Nickel coatings are usually prepared by electroplating. The copper underlayer is used to facilitate buffing of the surface on which nickel is plated. in seawater. and also to reduce the required thickness of nickel (which costs more than copper) for obtaining a coating of minimum porosity. zinc protects steel several decimeters or feet removed from the zinc. The automotive industry uses nickel as an underlayer for microcracked chromium to protect steel [9]. is not of great importance.3 SPECIFIC METAL COATINGS 14.2. of course. However. the longer cathodic protection continues.) wide may begin to show rust at the center. Crack resulted from cyclic stressing in a corrosion-fatigue test [H. For zinc coatings on steel in waters of low conductivity. 14. Undermining of nickel electrodeposit on steel by galvanic corrosion in 3% NaCl solution (100×). such as distilled or soft waters. The area of base metal over which cathodic protection extends depends on the conductivity of the environment. whereas cathodic current densities in waters of low conductivity fall off rapidly with distance from the anode. Fässler. porosity of sacrificial coatings. 17. because copper is softer than steel. Spähn and K.3. a coating defect about 3 mm (1/8 in. This difference in behavior results from adequate current densities for cathodic protection extending over a considerable distance in waters of high conductivity. Werkst Korros. which is a good conductor. .272 METALLIC COATINGS Figure 14. therefore. 321 (1966)]. the thicker the coating. such coatings are composed mostly of nickel. and temperature [11]. In this way. electrodeposited from a conventional nickel plating bath.013 mm (0. Optimum protection is obtained with “microcracked” chromium.8–1.008–0. overlies a matte coating two to three times thicker. The density of cracks in chromium deposits varies from 0 to more than 120 cracks/mm. for auto bumpers) for use near the sea coast or in industrial environments. so that the bright nickel acts as a sacrificial coating. appearance of rust is delayed and the multiple coating system is more protective for a given thickness of chromium and nickel [12]. in reality. a certain minimum thickness is recommended.6 mil)/h in the form of a nickel–phosphorus alloy [13]. coatings tend to lose their specular reflectivity because a film of basic nickel sulfate forms that decreases surface brightness [10]. depending on exposure conditions. This thin coating.” although. particularly by industrial atmospheres.0008 mm. a coating 0. in turn.02– 0. a very thin (0. current density. For use in the chemical industry. Nickel salts are reduced to the metal by sodium hypophosphite solutions at or near the boiling point. thereby decreasing depth of penetration that would otherwise result from attack at fewer sites. In the phenomenon called fogging. electrodeposited with many inherent microscopic cracks.0003–0. which includes a small amount of sulfur. For indoor exposures.5 mil) may be specified.SPECIFIC METAL COATINGS 273 Because porosity is an important factor in the life and appearance of nickel coatings. 0. a coating 0. To minimize fogging. This thin chromium overlayer has led to the term “chrome plate. act as catalysts for the reaction so that deposits can be built up .03 mil) coating of chromium is electrodeposited over the nickel.015 mm (0.04 mm (0.25 mm (1–10 mils) thick. but thinner coatings are usually adequate for dry or unpolluted atmospheres. For outdoor exposures.3–0. nickel coatings commonly range from 0.5 mil) is considered adequate for many applications.g. Nickel is sensitive to attack. depending on bath chemistry. The chromium deposit usually overlies a thin nickel coating formed from a plating bath containing additives (usually sulfur compounds) that improve surface brightness (bright nickel). Any pit beginning under the chromium coating tends to grow sidewise rather than through the nickel layers.01–0. Various metal surfaces. The usual range of phosphorus content in coatings of this kind is 7–9%. A typical solution is the following: NiCl2·6H2O Sodium hypophosphite Sodium hydroxyacetate pH 30 g/liter 10 g/liter 50 g/liter 4–6 This particular formulation deposits nickel at the rate of about 0. The thicker coatings in this range (or still thicker) are desirable (e. The bright nickel.. including nickel. The many cracks in chromium favor initiation of corrosion at numerous sites. Electroless nickel plate is produced primarily for the chemical industry.025–0. is anodic to the underlying nickel coating of lower sulfur content. cadmium. When 3–15% tin is incorporated.274 METALLIC COATINGS to relatively heavy thicknesses. In the hot-dipping operation. . and Zn–Ni [17]. such as Zn–Fe. the pores tending to fill with corrosion product. The coating produced in the batch process is thicker than that produced in the continuous process. the overall corrosion rate in short time tests is lowest within the pH range 7–12 (Fig.2 Lead Coatings Lead coatings on steel are usually formed either by hot dipping or by electrodeposition. 14.3 Zinc Coatings Coatings of zinc. Corrosion resistance of the nickel– phosphorus alloy is stated to be comparable to electrolytic nickel in many environments. whether hot-dipped or electroplated. are called galvanized coatings. In acid or very alkaline environments. Lead coatings are not very protective in the soil. 14. and the term electrogalvanized is used to differentiate electroplated from hot-dipped coatings. except when seawater spray comes into direct contact with the surface. Specific additions are made to the commercial solution in order to increase the rate of nickel deposition and also for coating glass and plastics. The electroplating process has been used to explore new zinc alloy coatings. Lead and lead–tin coatings are used to enhance the corrosion resistance of iron castings to H2SO3 and H2SO4 [16]. bismuth. Zn–Co. with controlled thickness. and it has clearly distinguishable alloy layers.3.1 lists the ranges of typical atmospheric corrosion rates in each of the three types of atmospheres. most likely lead sulfate (PbSO4) which stifles further reaction [15].3). the resulting lead–tin alloy is called terne. should this be desirable. the inside surfaces of tubes and other recessed areas [14]. Zinc coatings are relatively resistant to rural atmospheres and also to marine atmospheres. a few percent tin are usually incorporated to improve bonding with the underlying steel. the major form of attack results in hydrogen evolution. “Terne. Metals on which the coating does not deposit include lead. The coating produced by hot dipping is bonded to the underlying steel by a series of Zn–Fe alloys. because of the poisonous nature of small quantities of lead salts (see Section 1. with a layer of almost pure zinc on the outside surface. Table 14. marine. tin solder. 14. rural. Electroless nickel plating is useful for uniformly coating. The phosphorus content makes it possible to harden the coating appreciably by low-temperature heat treatment—for example. and antimony.2). Coatings of lead or lead–tin alloy are resistant to atmospheric attack.” was originally used to differentiate such coatings from bright tin plate. The continuous process results in a thinner coating with a very thin alloy layer at the coating–steel interface [17].” which means “dull.3. Lead coatings must not be used in contact with drinking water or food products. and urban/ industrial [18]. 400 °C (750 °F). In aqueous environments at room temperature. a coating 5 mils (0.5–8 2–16 Figure 14. thus. aerated solutions. Roetheli. .3.13 mm) thick will protect steel against the appearance of rust for about five years. and W.SPECIFIC METAL COATINGS 275 TABLE 14. Effect of pH on corrosion of zinc. zinc reacts rapidly to form soluble zincates in accord with Zn + OH − + H 2O → HZnO− + H 2 2 In seawater. Littreal.5.03 mm) of zinc equaling about one year of life.1. with each mil (0. 30 °C [B.2–3 0. Atmospheric Corrosion Rates of Zinc [18] Type of Atmosphere Rural Marine (outside the splash zone) Urban and industrial Range of Corrosion Rate (μm/year) 0. zinc coatings are effectively resistant for protecting steel against rusting. 73 (1932)]. Metals and Alloys 3. G. Cox. Above about pH 12. see Section 8. whereas those high in chlorides and sulfates decrease the reversal tendency [23]. This reversal of polarity leads to zinc having the characteristics of a noble coating instead of a sacrificial coating. compared to zinc coatings. A 15-year service test on piping carrying Baltimore City water at a mean temperature of 46 °C (115 °F) and maximum of 80 °C (176 °F) confirmed that pitting of galvanized pipe was 1. but this is not necessarily true in aerated waters. perform in aerated waters as an oxygen electrode. is noble to both zinc and iron. hence.4 Cadmium Coatings Cadmium coatings are produced almost exclusively by electrodeposition. a galvanized coating under these circumstances can induce pitting of the base steel. In one series of tests. but not overlapping the still more active hydrogen cracking potential range. the potential of which.3. but is related to formation of ZnO instead under conditions where the reverse polarity occurs [24].7 as deep as those in black iron pipe. but cadmium retains a bright metallic . like mill scale on steel. being a semiconductor. the presence of HCO− and NO−. Cadmium is more expensive than is zinc. whereas Cl− and SO2− favor formation of hydrated reaction 4 products instead. in deaerated hot or cold waters in which an oxygen electrode does not function. By maintaining a potential below the critical potential for S. because oxygen is absent. the current in a zinc–iron couple decreased to zero after 60–80 days.4. zinc is always anodic to iron. reversal of polarity between zinc and iron occurs at temperatures of about 60 °C (140 °F) or above [19–21]. more reliably protect the steel against failure by cracking in moist atmospheres. corresponding to shorter life of the galvanized pipe. in water or dilute sodium chloride. It was found that waters high in carbonates and nitrates favor the reversal in polarity. under those conditions for which zinc is anodic to steel. and a slight reversal of polarity was reported [25]. indicating. pits in galvanized pipe were only 0. The difference in potential between cadmium and iron is not as large as that between zinc and iron.. The latter compound conducts electronically. 14. hence. a beneficial effect of galvanizing [22]. Accordingly.2–2 times deeper than that in black iron pipe (not galvanized) of the same type.C. Apparently. The potential of zinc relative to steel is apparently related to the formation of porous Zn(OH)2 or basic zinc salts. therefore. stimulates 3 3 formation of ZnO.4–0. In cold water. This trend is less pronounced with high-purity zinc. which are insulators.1). The lower potential difference provides an important advantage to cadmium coatings applied to high-strength steels (Rockwell C > 40. It can. cadmium coatings. aided by elevated temperatures. the current output of zinc as anode decreases gradually because of insulating corrosion products that form on its surface. however. on which insulating coatings have less tendency to form.276 METALLIC COATINGS In many aerated hot waters.C. At room temperature. in this case. cathodic protection of steel by an overlayer of cadmium falls off more rapidly with size of coating defects. Tin layer. 0.00008 mil) thick Tin coatings so thin are.002 μm (0. and. The nontoxic nature of tin salts makes tinplate ideal for use with beverages and foods.012 mil) thick 4. cadmium coatings must not come into contact with food products. Steel sheet approximately 200–300 μm (7. the typical tinplate product consists of five layers [28]: 1.08 μm (0.5 Tin Coatings Tinplate is a low-carbon steel strip product coated on both sides with a thin layer of tin. either in chromic acid (H2CrO4) or in sodium dichromate (Na2Cr2O7).3. for this reason. Zinc salts are less toxic in this respect. After electroplating the tin on the steel substrate. The tinplate is then passivated. hence. disposal is difficult. and galvanized coatings are tolerated for drinking water. it is essential that the tin act as a sacrificial coating in order to avoid perforation by pitting of the thin-gage steel on which the tin is applied. . cadmium. however. Cadmium should be specified only after thorough investigation of the implications and alternatives [26].3 μm (0.SPECIFIC METAL COATINGS 277 appearance longer. the tinplate is given a heating cycle that melts the tin coating. Oil film about 0. Because of toxicity considerations and regulations concerning use of cadmium and disposal of wastes containing cadmium. very porous. naturally. 14. finally. Because the coefficient of friction of cadmium is less than that of zinc.004 mil) thick 3. and. Furthermore. and. ensures better electrical contact and solderability. resists attack by strong alkalies. tinplate is now produced universally by electrodeposition. it is corroded by dilute acids and by aqueous ammonia. Otherwise.00008 mil) thick 5. cadmium is preferred for fastening hardware and connectors that are repeatedly taken on and off.002 μm (0. unlike zinc. creating a layer of tin–iron intermetallic compound at the tin–steel interface. effective cadmium replacements are being identified wherever possible [27]. finds use in electronic equipment.7 mils) thick 2. Several million tons of tinplate are produced each year. it is more resistant to attack by aqueous condensate and by salt spray. but they also are not recommended for contact with foods. Passivation film about 0. Like zinc. Because electrodeposited tin is more uniform than is hot-dipped tin and can be produced as thinner coatings. Cadmium salts are toxic. and most of it is used to manufacture many billions of food containers. a thin layer of lubricant is applied. consequently. On each side. cadmium coatings exposed to the atmosphere are not quite as resistant as zinc coatings of equal thickness. Because solutions used for plating cadmium are toxic.8–11. As supplied to the can maker. In aqueous media. This condition usually applies. a layer of tin–iron intermetallic compound about 0. Applications in which corrosion products of cadmium can enter the environment must be avoided. with accompanying corrosion of the tin. Any further corrosion of tin is limited by the hydrogen evolution reaction (HER). In the absence of other depolarizers.278 METALLIC COATINGS On the outside of a tinned container. foods usually contain various organic substances.e. tin is almost always anodic to iron. The reason for this behavior is not firmly established.3) (14. Because the tin coating is the anode. reduction of hydrogen ions on tin is not significant. the corrosion reaction for tin is Sn → Sn 2+ + 2e − Two possible reduction reactions are O2 + 2 H 2O + 4e − → 4OH − and 2 H + + 2e − → H 2 (14. resulting in a change in corrosion potential of tin in the active direction (see Section 3. by reaction (14.. Because of the presence of organic depolarizers. some of which are complexing agents. which are known to inhibit corrosion of iron in acids. On the inside. it is this reaction—reduction of hydrogen on the steel substrate at pores in the tin coating—that controls detinning.2). tin is cathodic to iron. Corrosion of tin usually proceeds slowly over the years of useful life of the can.1) takes place. It is also . increase hydrogen overpotential. until the oxygen is consumed. thereby favoring reduction of organic substances at the iron cathode. As tin dissolves. so that oxygen reduction takes place.440 V for iron. so that both reactions—reduction of hydrogen on the steel and dissolution of tin—accelerate [29].1). but it may be related to the fact that Sn2+ ions. Foods low in inhibitor and high in depolarizer substances may cause more rapid corrosion of food containers than. loss of tin by reaction (14. however. When the tin coating has all corroded. On the inside of the can. as mentioned previously. it is observed that subsequent corrosion usually occurs by hydrogen evolution. the space between the top of the contents of the can and the lid). corrosion of the inside tin coating of food containers may occur with little or no hydrogen evolution. Stannous ions formed continuously at the iron surface during corrosion of the tin layer do not maintain sufficient concentration once the tin layer has dissolved. In addition to acids or alkalies as incidental and natural components. tin cathodically protects the base steel. however. more steel may become exposed.2) (14. in accord with the standard potential of tin equal to −0. highly acid foods. This fortunate reversal of potential occurs because stannous ions. where as hydrogen reduction can occur on the steel substrate. for example. Because tin has high hydrogen overpotential.136 V compared to −0. Sn2+.1) Oxygen inside the can is usually limited to that in the headspace (i.8). and others act either as corrosion inhibitors or as cathodic depolarizers. are complexed by many food products that greatly reduce Sn2+ activity. reaction (14. and hence. freedom from undercutting of the organic coating on the inside. but.6 Chromium-Plated Steel for Containers Because of price advantage. corrodes tin slowly (1. such as HCl and H2SO4.44 V).2 V). limiting its use in some application to can ends. Section 8. the corrosion potential of chromium-plated steel. Subsequent low-temperature heat treatment.1). which is always porous to some degree. but most often by the composition and structure of the base steel. Hence. it is not readily soldered. in galvanic couples of the two metals. The amount of hydrogen accumulated within the lifetime of the container is determined not only by the tin coating thickness. but they become more than 10 times higher at the boiling point. and the chemical nature of food in contact.008–0.1–2. and resistance to filiform corrosion on the outside of the container. NH4OH and Na2CO3 have little effect. good adhesion. chromium is polarized below its Flade potential. incidental or intentional. is determined largely by concentration of dissolved oxygen in the acid. However.5–6. 14. As with acids.1). making the contents suspect. especially in acid media. The hydrogen overpotential of tin is so high that corrosion accompanied by hydrogen evolution is limited. Tin is an amphoteric metal. followed by chromium oxide (5–40 mg Cr/m2) and an organic top coat. is more active than that of either passive . the temperature. but chromium has a strong tendency to become passive ( φ F = 0. for example. since bacterial decomposition also causes gas accumulation. The advantages include good storage ability. indicating appreciable increase of attack with temperature [30]. Deaerated acetic acid.5 gmd. the rate of attack is rapid with formation of sodium stannate. may increase or decrease the rate (see Fig. which is standard procedure for strengthening the container walls.1. Corrosion in dilute nonoxidizing acids. 0.5. Only the high price of tin accounts for its replacement for distilled water piping by other materials. either hot or cold. in concentrated NaOH. It does not corrode in soft waters and has been used for many years as piping for distilled water. Hence. reacting with both acids and alkalies.1 Properties of Tin.3.SPECIFIC METAL COATINGS 279 possible that the potential difference of the iron–tin couple favors adsorption and reduction of organic depolarizers at the cathode.3–0.4 μin. Eventual failure of containers is by socalled hydrogen swells. so that it exhibits an active potential. However. the rate is increased by aeration. Chromium (φo = −0. whereas these processes do not take place at lower potential differences.3. 8. but it is relatively resistant to neutral or near-neutral media.74 V) is more active in the Emf Series than is iron (φo = −0. such as aluminum. in which an appreciable pressure of hydrogen builds up within the can. The rate of hydrogen evolution is increased by cold working of the steel (see Section 8.01 μm).0 gmd). steel may be coated with a combined thin chromium layer (0. rates are only 0. 14. resistance to sulfide staining. the potential of chromium in aqueous media is usually noble to that of steel. In aerated fruit juices at room temperature. The metallic chromium layer contributes resistance to undercutting of the organic coating. hence. increases resistance to the flow of electric current produced by galvanic cells. as well as increased concentrations of corrosive chemicals. Sprayed coatings are commonly sealed with organic lacquers or paints to delay eventual formation of visible surface rust. In one series of tests conducted in an industrial atmosphere.280 METALLIC COATINGS chromium or the underlying steel [31]. 14. coated with aluminum) by hot dipping or spraying and. As a result. An aluminum coating under these conditions is sacrificial and cathodically protects steel. such as sulfuric acid. seals pores in the metal coating. They are unaffected by temperatures up to 480 °C (900 °F). such as hot dip aluminumcoated type 409 stainless steel.08–0. materials resistant to high-temperature oxidizing conditions. the coatings become refractory. but with dependence of its active potential on passive–active behavior rather than on formation of metal complexes. At still higher temperatures. Hot-dipped coatings are used for oxidation resistance at moderately elevated temperatures. Under these conditions. since 1975. Al has been reported to have excellent resistance to marine and industrial atmospheres. bal.3.7 Aluminum Coatings Steel is aluminized (i. with optimum adhesion occurring for a hydrated oxide thickness [32] equivalent to 20 mg/m2 (2 mg/ft2). In seawater and in some fresh waters.5% Si.e. It also protects against oxidation at elevated temperatures. The organic coating. Molten baths of aluminum for hot dipping usually contain dissolved silicon in order to retard formation of a brittle alloy layer. chromium acts as a sacrificial coating in the same manner as does a tin coating. in turn. of catalytic converters in automobiles has resulted in increased exhaust gas temperatures. and reduces the amount of iron salts entering the contents of the container and affecting flavor or color [33]. in the exhaust gas stream. An Al–Zn alloy coating consisting of 44% Zn. In soft waters. an application that is limited by higher costs of aluminum compared to zinc coatings and by variable performance. aluminum exhibits a potential that is positive to steel. such as for oven construction and for automobile mufflers. A chromium oxide coating enhances adhesion of the organic top coating. with metal temperatures up to 760 °C (1400 °F). or hot-dipped [38].2 mm).08 mm) thick showed an average life of 12 years in comparison with 7 years for zinc. it acts as a noble coating. . the potential of alu4 minum becomes more active and the polarity of the aluminum–iron couple may reverse. are being used in automotive exhaust systems [35–37]. sprayed coatings 3 mils (0. with the latter coatings being sprayed. The use. especially those containing Cl− or SO2− . up to 870 °C (1600 °F). The usual thickness of sprayed aluminum is 3–8 mils (0. but continue to be protective up to about 680 °C (1250 °F) [34]. 1. to a lesser extent. Aluminum coatings are also used for protecting against atmospheric corrosion. by cementation.. electrodeposited. 13A. 3. Vernon. Std. 18. editor. and G. Mooney. T. Trans. A. in Corrosion of Metals Processed by Directed Energy Beams. 888–889. Testing. Ibid. 6. NY. 1–21. 405–422. T. Corrosion: Fundamentals. Natl. PA. TX. Materials Park. T. in ASM Handbook. Materials Park. Materials Park. in Uhlig’s Corrosion Handbook. 779. Warrendale. 15. Inst. ASM International. 1987. J. (U. in ASM Handbook. G. Materials Park. editors. Materials Park. M. 801. C. Corrosion: Materials. 14. in Handbook of Environmental Degradation of Materials. OH. Testing. Vol. Houston. Corrosion. ibid. 2nd edition. ASM International. Hubler. C. Townsend. editor. ASM International. NACE International. G. 2003. LaQue. OH. J. 93. Moore. 31-1. E. 803–813. W. 13A. William Andrew Publishing. X. Berndt. Al2O3. as well as resistance to sulfur-containing atmospheres. and Protection. p. Corrosion: Fundamentals. Performance 21(7). ASM International. E. Testing. OH. 5. ASM International. such as those encountered in oil refining. 780. Testing.S. 7. 127 (1964). Vol. 16. G. 13A. PA. 8. in Automotive Corrosion and Protection. Dorfman. in ASM Handbook. pp. OH. Spence. M. Metal Finishing 41. Preece and J. Proceedings of the CORROSION/91 Symposium. Vol. and Protection. pp. 13B. R. Preece. Jones. 2003. C. Mooney. M. Carter. 12. E. S. 121 (1932). Dearnaley. pp. They are sometimes used to protect gas turbine blades (nickel-base alloys) from oxidation at elevated temperatures. 2005. pp. Mattsson. REFERENCES 1. Corrosion: Fundamentals. J. 17. and a small amount of NH4Cl as flux in a hydrogen atmosphere at about 1000 °C (1800 °F). Cemented aluminum coatings on steel are not useful for resisting aqueous environments. 13. Corrosion: Fundamentals. 9. G. 1–20. Bur. in ASM Handbook. L. p. p. editor. and Protection. 4. 13A. P. Langill. Mater. 777. Lashmore. Ayers. p. Vol. Brenner and G. C. 11. OH. New York. and T. Kutz. R. in ASM Handbook. 2000. and Protection. . F. Metals 48. in ASM Handbook. 2003. The Metallurgical Society. McCafferty. 2. Vol. Inst. 13. Berndt and C. W. in Ion Implantation Metallurgy. 772–785. M. R Clayton and C. 9 (1982). Vol. p. 1980.) 39. H. 10. The Metallurgical Society of AIME. Mooney. A. An aluminum–iron surface alloy forms that imparts useful resistance to high-temperature oxidation in air up to 850–950 °C (1550–1750 °F). Norwich. p. Mazia. J. K. Hirvonen. 1982. Metal Progr.. D.REFERENCES 281 Cemented coatings (“Calorizing”) are produced by tumbling the work in a mixture of aluminum powder. 117 (1968). ASM International. 1991.. R. OH. Warrendale. 871. pp. Wiley. D. 2003. 385 (1947). Zhang. 2005. J. pp. p. Revie. 49. Materials Park. T. Riddell. Prot. 247 (1939). G. pp. 59 (1941). and N. Roters and F. Corros. PA. Linde. editors. 13. 13A. 772–813. in ASM Handbook. B. 38. 36. Wiley. Corrosion: Fundamentals. Lead and lead alloys. H. Materials Park. Cadmium elimination. p. Vol. J. Ludwigsen. in Uhlig’s Corrosion Handbook. 25. p. 2005. Grainger and J. 2000. ASM International. 30. Kamm. Society of Automotive Engineers. Vol. Materials Park. Engineering Coatings—Design and Application. ASM International. G. Soc. Ingle. p. 1299 (1969). Bonilla. 29 (1968). 2000. H. W. R. R. 5. Soc. editors. 71–106. Warrendale. C. 1987. pp. 34. S. P-78. W. McKirahan and R. Zinc-Based Steel Coating Systems: Metallurgy and Performance. J. National Steel Corporation. W. New York. England. p. Blunt. J. Warrendale. Ind. ASM International. Gilbert. Materials Park. 28. pp. E. Hudson. OH. ASM International. Trans. Goodwin. Materials Park. OH. Arch. Prevention of Corrosion of Motor Vehicle Body and Chassis Components. PA. in ASM Handbook. Gonser and J. 971 (1985). Cambridge. p. in ASM Handbook. Wiley. ASM Handbook. ASM International. Electrochem. 858–860. editor.282 METALLIC COATINGS 19. Ashworth. Douthett. Corrosion: Materials. T. K. Materials. . editor. and corrosion effects on exhaust system life. 76. A. Eisenstecken. OH. 237 (1945). G. 2005. Zhang. 767–792. 327. 29. Electrochem. Wiley. Krauss and D. OH. G. P. Trans. Schikorr. 13C. Glass and V. SAE Paper 780921. 2nd edition. Soc. (London) 1961. Testing. H. William Andrew Publishing. Willey. Strader. 31. 25. Matlock. 1981. Chem. in ASM Handbook. 13B. Corrosion: Materials. Vol. R. 87. TMS. 116. Metal coatings. 20. 1983. X. in Uhlig’s Corrosion Handbook. pp. R. Vol. M. Soc. Corrosion. 1998. 1978. M. Sci. P. 519–520. Electrochem. 780. ASM Handbook. 13B. Murphy. 2nd edition. 918–924. 99. New York. 22. OH. 2nd edition. 413. 3 (January 7). Surface Engineering. pp. 24. pp. 1990. 16 (1952). 23. GENERAL REFERENCES F. Revie. editor. design. Patterson. and Protection. Vucich. in Corrosion Handbook. Electrochem. Uhlig. Materials Park. 37. 26. New York. Ref. Prutton. G. Hoxeng and C. 1948. J. 330 (1949). Vol. 35. R. K. Vol. Automotive Exhaust System Corrosion. W. Corrosion: Environments and Industries. 21. Eisenhüttenw 15. 32. 27. Materials Park. 1994. p. 14. 7(12). 2006. 186. Corrosion 5. 33. Society of Automotive Engineers. OH. Revie. in Designing for Automotive Corrosion Prevention. 1016. private communication with HHU. ASM International. Mater. Revie. 887–904. Wiley. 2nd edition. R. England. Chichester. W. editor. 2000. L. G. Tin and tinplate. W. in Uhlig’s Corrosion Handbook. Pawlowski. in Uhlig’s Corrosion Handbook. . 1995. P. editor. R. pp. Zinc. Revie. New York. Zhang. 2000. pp.GENER AL REFERENCES 283 T. 2nd edition. 853–862. The Science and Engineering of Thermal Spray Coatings. Wiley. X. Murphy. Wiley. New York. . 15 INORGANIC COATINGS 15. Vitreous enamel coatings are used mostly on steel. brass. mild alkalies. Winston Revie and Herbert H. The glasses are composed essentially of alkali borosilicates and can be formulated to resist strong acids. or both. In addition to decorative utility. Although pores in the coating are permissible when cathodic protection supplements the glass coating. for other applications the coating must be perfect and without a single defect. This means that glass-lined vessels for the food and chemical industries must be maintained Corrosion and Corrosion Control. Several coats may be applied. Their highly protective quality results from virtual impenetrability to water and oxygen over relatively long exposure times. 285 . and porcelain enamels are all essentially glass coatings of suitable coefficient of expansion fused on metals. Uhlig Copyright © 2008 John Wiley & Sons. but some coatings are also possible on copper. glass linings. Glass in powdered form (as glass frits) is applied to a pickled or otherwise prepared metal surface. Their use in cathodically protected hot-water tanks has already been mentioned in Section 13. by R. and aluminum. and from their durability at ambient and above-room temperatures. vitreous enamels protect base metals against corrosion by many environments. Inc.4. then heated in a furnace at a temperature that softens the glass and allows it to bond to the metal.1 VITREOUS ENAMELS Vitreous enamels. Vitreous enamels are also used to protect against high-temperature gases (e. decorative building panels. Open tanks are easily repaired.g. In addition. 15. Failure occurs eventually by formation of a network of cracks in the coating—called crazing—through which rust appears. Portland cement coatings are used to protect cast iron and steel water pipe on the water or soil side or both.2 PORTLAND CEMENT COATINGS Portland cement coatings have the advantages of low cost. The ensuing reaction produces a network of porous metal phosphate crystals firmly bonded to the steel surface (Fig.1). and ease of application or repair.. oil tanks. They include special coatings.and cold-water tanks.. Enameled steels exposed to the atmosphere last many years when used as gasoline pump casings.). 15. Portland cement coatings are used on the interior of hot. appliances. however.286 INORGANIC COATINGS free of cracks and other defects.. There is evidence that small cracks in coldwater piping are automatically plugged with a protective reaction product of rust combining with alkaline products leached from the cement. Portland cement may be attacked. or by spraying. ZnH2PO4 plus H3PO4).g. If damage occurs. plumbing fixtures. and they have long life when exposed to soils. with an excellent record of performance.” “Bonderizing”) are produced by brushing or spraying. The coatings can be applied by centrifugal casting (as for the interior of piping). Susceptibility to mechanical damage and cracking by thermal shock are the main weaknesses of glass coatings. which forms when steel containers are filled with hydrofluoric acid (>65% HF). a coefficient of expansion (1. which forms when lead is exposed to sulfuric acid. or NO− ) to speed the reaction. but cement compositions are now available with improved resistance to such waters. Accelerators are sometimes added to the phosphating solution (e.2–1 in. onto a clean surface of steel. repairs can sometimes be made by tamping gold or tantalum foil into the voids. A disadvantage of Portland cement coatings is the sensitivity to damage by mechanical or thermal shock. ClO− . thick coatings are usually reinforced with wire mesh. and iron fluoride. by troweling.3 CHEMICAL CONVERSION COATINGS Chemical conversion coatings are protective coatings formed in situ by chemical reaction with the metal surface. advertising signs. Usual thickness ranges from 5 to 25 mm (0. and chemical storage tanks. such as oils. In sulfate-rich waters. The coatings are usually cured for 8–10 days before exposure to nonaqueous media. 15.2 × 10−5/°C).0 × 10−5/°C) approximating that of steel (1. a cold or hot dilute manganese or zinc acid orthophosphate solution (e. Phosphate coatings on steel (“Parkerizing. such as PbSO4. and so on. in airplane exhaust tubes). They are also used to protect against seawater and mine waters. 3 3 . Cu2+.g. by troweling fresh cement into cracked areas. some protection is obtained. In corrosion-resistant automobile bodies.CHEMIC AL CONVERSION COATINGS 287 Figure 15. phosphate coatings are impregnated with oils or waxes. or chlorates. like phosphate coatings. New York. nitrates. Sometimes. consist mostly of Fe3O4 and. edited by P. Weiss and G.025 mm (0. Although it is the thinnest coating in the paint system [about 3 μm (0.1 mil) thick]. as is often done with oxidized gun barrels. Such coatings. They are useful mainly as a base for paints. which may provide some protection against rusting. The oxide so formed must be hydrated to improve its protective qualities by exposing anodized articles to steam or hot water for several minutes. Cheever. blue.g. it is the anchor for the subsequent layers [1].. The resultant coating of Al2O3 may be 0. for example. Scanning electron photomicrograph of phosphated type 1010 mild steel. brown. especially if they contain corrosion inhibitors. the first—and probably the most important—layer of the paint system is the phosphate coating.1.1–1 mil) thick. Phosphate coatings do not provide appreciable corrosion protection in themselves. or black in color. are not protective against corrosion. ensuring good adherence of paint to steel and decreasing the tendency for corrosion to undercut the paint film at scratches or other defects. When rubbed with inhibiting oils or waxes. Elsevier. 1968). a process called sealing. acidic zinc phosphate + sodium nitrate accelerator applied at 65 °C for 1 min (Symposium on Interface Conversion for Polymer Coatings. dilute sulfuric acid) at current densities of 100 or more A/m2.0025–0. by immersion in hot concentrated alkali solutions containing persulfates. Oxide coatings on aluminum are produced at room temperature by anodic oxidation of aluminum (called anodizing) in a suitable electrolyte (e. Improved . Oxide coatings on steel can be prepared by controlled high-temperature oxidation in air or. 750–754. The oxide coating may be dyed various colors. which is otherwise difficult to paint without special surface preparation. TX. Townsend and J. Houston. . Baboian. Zoccola. Similar coatings have been recommended for Zn–Al coatings [3] and for cadmium coatings on steel.. Wiley. Singer. GENERAL REFERENCES Automotive Corrosion and Protection. 125. 1992. W. OH. Vol.. 411–437. Grambow. This coating has been recommended for surface cleaning or as a base for finishing treatments [2]. in Uhlig’s Corrosion Handbook. R. and M. Anodized coatings provide a good base for paints on aluminum.g. R. Corrosion of Glass. R. ASM International. 2nd edition. editor. New York. A zinc chromate surface is produced that imparts a slight yellow color and that protects the metal against spotting and staining by condensed moisture. Materials Park. 8 mL/liter) at room temperature and then rinsing and drying. Electrochem. Anodizing provides aluminum with some degree of improved corrosion resistance. REFERENCES 1. Soc. L. Emley. Principles of Magnesium Technology. E. 2000. pp. NACE International.g. 13A. pp. 20–8. 1992. Ostermiller. Testing. 2003. but the additional protection is not spectacular. L. Pergamon Press. H. 200 g/liter) acidified with sulfuric acid (e. editor. 2. 1290 (1978). Revie. It also increases the life of zinc to a modest degree on exposure to the atmosphere. NACE International. Coatings of MgF2 on magnesium can be formed by anodizing the metal at 90–120 V in 10–30% NH4HF2 solution at room temperature. 1966. in Automotive Corrosion and Protection. editor. either in the anodizing bath or afterward. New York. in ASM Handbook. Porcelain enamels. and Protection. 3. Corrosion: Fundamentals. and it is certainly less than proportional to oxide thickness. L. pp. R.288 INORGANIC COATINGS corrosion resistance is obtained if sealing is done in a hot dilute chromate solution. Chromate coatings are produced on zinc by immersing the cleaned metal for a few seconds in sodium dichromate solution (e. Piepho. J. B. Baboian. TX. Houston. by R. half of which is estimated to be used for corrosion protection.2 PAINTS The total value of paints. Uhlig Copyright © 2008 John Wiley & Sons. and lacquers produced in the United States in 2006 amounted to about 21 billion (21 × 109) dollars [1]. 16. $500 million Corrosion and Corrosion Control. have been developed with reduced volatile organic compounds. were used. new coating products. Painted systems require continued maintenance to ensure integrity of corrosion protection. Solvent vapors from coating operations undergo photochemical oxidation in the atmosphere and contribute to smog. alternative processes are now being used in place of the traditional vapor degreasing in which chlorofluorocarbons and 1.1 INTRODUCTION In the United States. varnishes. 289 . In order to reduce this pollution.l-trichloroethane. in 1995. In addition. For example. termed compliant coatings. Inc.16 ORGANIC COATINGS 16. the Clean Air Act has had a major influence on the formulation and application of organic coatings.1. both being class 1 ozone-depleting chemicals. the Port Authority of New York and New Jersey began a 20-year. Winston Revie and Herbert H. Fe2O3) or other compounds. alkalies. The new paint system. Paints are a mixture of insoluble particles of pigment suspended in a continuous organic or aqueous vehicle. their life is relatively short. Vinyl resins have good resistance to penetration by water. and durability. consisting of three coats. Lacquers consist of resins dissolved in a volatile thinner. use of aqueous vehicle paints is preferred.. by comparison. over use of paints employing volatile organic solvents. Because of the need to reduce atmospheric pollution. because of favorable cost. such as ZnCrO4. Examples of synthetic resins include phenol-formaldehyde formulations.290 ORGANIC COATINGS program to repaint all bridges. These resins are the basis of plastic mixtures that. More ade- . fast-drying properties. Silicone and polyimide resins are useful at still higher temperatures. and a volatile thinner. they oxidize and polymerize to solids. for resisting a variety of corrosive media. and urethane finish. BaSO4. to seal leaks temporarily in ferrous or nonferrous piping. a process that can be hastened by small amounts of catalysts. Alkyd resins. in general. Epoxy resins are also resistant to alkalies and to many other chemical media and have the distinguishing property of adhering well to metal surfaces. Varnishes consist generally of a mixture of drying oil. are quickly saponified and disintegrated by alkaline reaction products formed at a cathode. Pb3O4. which withstand boiling water or slightly higher temperatures and are used in the chemical industry in the form of multiple coats. Pigments usually consist of metallic oxides (e. dissolved resins. Synthetic resins are now more often used as vehicles or components of vehicles. one reason being that mechanical damage to thin coatings by contact with the soil is difficult to avoid. on addition of a suitable catalyst. or they may polymerize through application of heat or by addition of catalysts. solidify in place within a short time. The latter property presumably derives from many available polar groups in the molecule. have found wide application for protecting the metal surfaces of machinery and home appliances. particularly for continuous contact with water or where resistance to acids.g. except possibly for short periods of time on the order of one year or less. The existing. Linseed. clays. Tests have shown that. These resins may dry by evaporation of the solvent in which they are dissolved. TiO2. for example. PbCO3. They are useful. whether in waters or in soil. Coatings based on coal tar or on fusion-bonded epoxy have been found to be far more practical. The vehicle may be a natural oil—linseed or tung oil. In hot water. they sometimes contain pigments as well.and tung-oil paints. the usual linseed-tung oil paints are not durable for metal structures totally immersed in water. is expected to last 25 years with “moderate maintenance” [2]. or higher temperatures is required. and in some cities required. Paints. Similarly. such as lead. for example. manganese. or cobalt soaps. baked on. When these drying oils are exposed to air. and so on. life is still shorter. epoxy intermediate. for this purpose. are not useful for protecting buried structures. lead-based paint was removed by abrasive blasting down to the bare metal. Their resistance to alkalies makes them useful for painting structures that are to be protected cathodically. zinc primer. instead. The solubility of zinc tetroxychromate is reported to be 2 × 10−4 mole/liter [3]. by brushing) to the metal surface (e. All paints are permeable. only relatively few actually do the job that is required.3 REQUIREMENTS FOR CORROSION PROTECTION To protect against corrosion. Diffusion.. tends to be accelerated by electroosmosis whenever the coated metal is made the cathode of a galvanic couple or is cathodically protected. 16. solubility relations being just 4 right to release at least the minimum concentration of the ion (>10−4 mole/liter) for optimum inhibition of steel. Inhibit against Corrosion. but their better performance as a diffusion barrier applies only to well-adhering multiplecoat applications that effectively seal pores and other defects.g. the inhibiting ion is CrO2−. Pigments incorporated into the prime coat (the coat immediately adjacent to the metal) should be effective corrosion inhibitors. Because of the expense of a multiple-coat system of this kind. many applications for fresh water or seawater make use. For zinc chromate. Particularly effective in this regard are pigments having the shape of flakes oriented parallel (e. By far the majority of paints perform best for protecting metals against atmospheric corrosion. micaceous or flaky hematite.. but red lead appears to be best of the lead compounds. Water reaching the metal surface then dissolves a certain amount of pigment. yet not soluble to a degree that they are soon leached out of the paint. of thick coal-tar coatings. which is released in just sufficient 4 amounts to passivate steel. on the other hand. Among pigments that have been recommended for prime coats. It is likely that lead oxides and hydroxides of other compositions also have inhibiting properties in this regard.g. is not nearly . on the other hand. in some degree. aluminum powder).REQUIREMENTS FOR CORROSION PROTECTION 291 quate protection extending up to several years at ordinary temperatures can be obtained by applying four or five coats of a synthetic vehicle paint. and this is their main function. making the water less corrosive. Provide a Good Vapor Barrier. Lead chromate. The inhibiting ion in the case of red lead is probably PbO4− . The diffusion path through a paint film is normally increased by incorporating pigments. Some vehicles are less permeable than others. Corrosion-inhibiting pigments must be soluble enough to supply the minimum concentration of inhibiting ions necessary to reduce the corrosion rate. a good paint should meet the following requirements: 1. protecting it against rusting by water reaching the metal surface. to water and oxygen. 2. as is done in the chemical industry. Effective pigments for which performance has been established in many service tests include (1) red lead (Pb3O4) having the structure of plumbous ortho-plumbate (Pb2PbO4) and (2) zinc chromate (ZnCrO4) and basic zinc chromate or zinc tetroxychromate. Sacrificial protection by zinc-rich coatings requires intimate contact with the substrate. it is advantageous to apply several coats rather than one. its ability to reflect infrared and ultraviolet radiation) and the type of vehicle used play some part. probably because pores are better covered by several applications and also because evaporation or dimensional changes during polymerization are better accommodated by thin films. A coating of this kind protected steel in seawater against rusting at a scratch for 1–2 years. The minimum amount of zinc dust pigment required to provide cathodic protection depends on several factors. including Zn particle size. A reasonable cost of paint should be gauged by its performance. Vehicles for zincrich paints to accommodate such a large fraction of pigment include chlorinated rubber. Paints pigmented with zinc dust using essentially an aqueous sodium silicate vehicle (called inorganic zinc-rich paint). or an organic vehicle. present either inadvertently or intentionally. rust appeared after 1–2 days and 10–20 days.292 ORGANIC COATINGS soluble enough (solubility = 1. The sulfate and chloride content of a commercial ZnCrO4 (or ZnMoO4) pigment must be low so that it can passivate the metal surface. the performance of good-quality paints used for corrosion protection is largely determined by the thickness of the final paint film. they are relatively ineffective for this purpose in the presence of high concentrations of chlorides. Zinc molybdate has been suggested as an inhibiting pigment for paints [4]. nature of the vehicle. in turn. the amount of pigment in the dried paint film should account for 95% of its weight. 3. Such paints are sometimes used over partly rusted galvanized surfaces because they also adhere well to zinc. and so these coatings are always used as primers. are also useful as prime coats. Commercial formulations of lead chromate sometimes contain lead oxides. It is less toxic than chromates. polystyrene. such as in seawater. respectively. In achieving a given coating thickness. The rate of deterioration of a paint depends on the particular atmosphere to which it is exposed.4 × 10−8 mole/liter) and acts only as an inert pigment. whereas with 86% and 91% Zn. epoxy. and polyurethane. being white instead of the characteristic yellow of chromates. Other things being equal. and the amount of ZnO and other pigments that may be present [6]. Provide Long Life at Low Cost. The color of the top coat (i. Because inhibiting pigments passivate steel. . A paint system lasting 5 years justifies double the cost of paint if the more expensive paint provides 35% longer life or lasts short of 7 years (labor to paint cost ratio of 2 : 1). It has been reported [5] that in order to ensure good electrical contact between zinc particles and with the base metal.e. depends on the amount of atmospheric pollution and on the amount of rain and sunshine.. It probably also depends on the extent to which insulating coatings form on zinc particles before the paint is applied (age of paint). which. but rust should first be removed. which may impart a degree of inhibition. the function of the zinc being to cathodically protect the steel in the same manner as galvanized coatings. A well-prepared surface is the foundation on which the paint system is built. they may leave chloride residues on the metal surface that can later initiate corrosive attack. or spraying. In addition.1) or. Oils. Suitable solutions contain one or more of the following substances: Na3PO4. in which the work is suspended in the vapor of the boiling solvent. Chlorinated solvents. In other words. 16.] containing about 30–75 g/liter (4–10 oz/gal) alkali.1 Cleaning All Dirt. One such solvent is Stoddard solvent. ensuring that adequate chemical inhibitors are added and maintained in the chlorinated solvent in order to avoid catastrophic corrosion (see Section 21.or perchlorethylene). Na2O·nSiO2. in somewhat more dilute concentration. If the metal is free of mill scale and rust. and it is not particularly toxic.METAL SURFACE PREPAR ATION 293 16. Alkaline Solutions. borax. Adequate surface preparation consists of two main processes. This factor is. Mineral spirits. to avoid an explosive reaction.4. on the other hand.4 METAL SURFACE PREPARATION Many test results and a great deal of engineering data over many years have shown that the most important single factor influencing the life of a paint is the proper preparation of the metal surface. may also be sprayed onto the surface.p. naphtha. Electrolytic cleaning in alkaline solutions is also sometimes used. sufficiently high to minimize fire hazard. and so on.1. ethers. a petroleum base mineral spirit with a flash point of 40–55°C (100–130°F). The hot solution. although nonflammable. Care must be exercised in the vapor degreasing of aluminum. .4. NaOH. Solvents. in the extreme. are relatively toxic and contribute to pollution. a poor paint system on a properly prepared metal surface usually outperforms a better paint system on a poorly prepared surface. chlorinated solvents. Cleaning may be done by immersing the work in the hot solution [80°C (180°F) to b. They are more efficient in this particular function than solvents. more important than the quality of the paint that is applied. this process makes use of the mechanical action of hydrogen gas and the detergent effect of released OH− at the surface of the work connected as cathode to a source of electric current. alcohols. They are used largely for vapor degreasing (tri. but perhaps less effective for removing heavy or carbonized oils. a final rinse in water and in a dilute chromic–phosphoric acid ensures both removal of alkali from the metal surface. generally.or metaphosphate and a wetting agent. are applied by dipping. and sometimes sodium pyro. and Greases from the Surface Initial cleaning can be accomplished by using solvents or alkaline solutions. brushing. Aqueous solutions of certain alkalies provide a method of removing oily surface contamination that is cheaper and less hazardous than the use of solvents. Na2CO3. The poor performance of paints on weathered steel exposed to an industrial atmosphere was shown in tests reported by Hudson [9] (Table 16. Other methods of removing mill scale include flame cleaning. It would.2 Complete Removal of Rust and Mill Scale Rust and mill scale are best removed by either pickling or sandblasting. and also temporary protection against rusting. or a 10–20% H3PO4 is used at temperatures up to 90°C (190°F). scale is removed by high-velocity particles impelled by an air blast or by a high-velocity wheel. which can then be further removed by wire brushing.4. 4–6 oz/ gal) or chromic–phosphoric acid. but has the advantage of producing a phosphate film on the steel surface that is beneficial to paint adherence.1). The latter acid is more costly.g. Oxide next to the metal surface is dissolved. alumina.. by which scale spalls off the surface through sudden heating of the surface with an oxyacetylene torch. or HCl alone is used at lower temperatures.3) at a temperature of 65–90°C (150–190°F) for an average of 5–20 min. Blasting. in fact. But these procedures are less satisfactory than complete removal of scale and rust by pickling or blasting. particularly after electrolytic action has occurred between metal and scale by aqueous solutions that have penetrated to the metal surface. is dipped into an acid (e. or sometimes of steel grit. loosening the upper Fe3O4 scale. refractory slag. Hexavalent chromium (Cr6+) is a potent human toxin and known cancer-causing agent [7]. Fragmentation of mill scale allows paint to become dislodged.294 ORGANIC COATINGS which would otherwise interfere with good bonding of paint. Some pickling procedures. A relatively nontoxic pickle for steel consists of hot 3–10% ammoniated citric acid followed by a dilute alkaline sodium nitrite solution to minimize superficial rusting before application of paint [8]. 16. silicon carbide. 3–10% H2SO4 by weight) containing a pickling inhibitor (see Section 17. the natural rusting of the surface dislodges scale. Sometimes. which serves to prevent rusting of the surface before the prime coat is applied. Weathering for several weeks or months is also possible. Pickling. Sometimes. call for a final rinse in dilute H3PO4 in order to ensure removal from the metal surface of residual chlorides and sulfates that are damaging to the life of a paint coating. . cleaned as described previously. Use of chromates is less popular than formerly because their toxicity places restrictions on disposal of spent solutions. The relatively long life of paint shown for intact mill scale would probably not be achieved in actual service. sodium chloride is added to the sulfuric acid. Blast materials usually consist of sand. or rock wool byproducts. the final dip is a dilute solution of chromic (30–45 g/liter. The metal. be difficult to keep large areas and various shapes of mill scale from cracking before or after painting. for example. Using this procedure. Only in unusual cases should paint be applied over a damp or wet surface because poor bonding of paint to steel results under these conditions. Tougher and thicker materials have been developed.2 2. is that it results in a uniform. and the application processes have been improved. A second prime coat can be applied after the first has dried. and environmentally acceptable [12]. For many years it has been standard practice to coat automobile bodies and electric appliances with phosphate before painting. automated. . there have been vast improvements in the corrosion performance of automobiles. or a sequence of top coats can follow.5 APPLYING PAINT COATINGS The prime coat should be applied to the dry metal surface as soon as possible after the metal is cleaned in order to achieve a good bond. During the past 30 years.3 16. if necessary.APPLYING PAINT COATINGS 295 TABLE 16. for example. The advantages of a phosphate coat are a better bond of paint to metal and good resistance to undercutting of the paint film at scratches or other defects in the paint at which rust forms and progresses beneath the organic coating. but also by the Clean Air Act of 1970 [with regulations regarding volatile organic content (VOC)] and subsequent revisions.1. the vehicle is immersed in the electrodeposition bath for typically 2–3 minutes [11]. One of the most important advances has been the development of the cathodic electrodeposition priming process.0 1. introduced in 1976 and. England 2 Coats Red Lead + 2 Coats Red Iron Oxide Paint Intact mill scale Weathered and wire-brushed Pickled Sandblasted 8. the metal should first be given a phosphate coat (see Section 15. efficient. by 1985.5 10. compared to spraying. Following phosphate pretreatment. Better still. fully implemented by most vehicle manufacturers in their assembly plants. A total of four coats with combined thickness of not less than about 0.2 4.4 2 Coats Red Iron Oxide Paint 3. thin coating [about 25 μm (1 mil) thick] with coverage on both exterior and interior cavity surfaces.3 9. this process is controllable. Effect of Surface Preparation of Steel on Life of Paint Coatings [9] Surface Preparation Durability of Paint (years) in Sheffield. In addition.3). and a film is deposited. Today. can be delayed for a short while. nearly 100% of the cars and trucks manufactured in the world are primed using the cathodic electrodeposition process [11]. in which case the prime coat. Electric current is applied. One of the advantages of this process. driven not only by the demands of the public for increased corrosion resistance.6 6.13 mm (5 mils) is considered by some authorities to be the recommended minimum for steel that will be exposed to corrosive atmospheres [10]. or a wash primer should be used as a prime coat. is a suitable inhibiting pigment in the prime coat. or zinc-rich paints can be used. Zinc or galvanized surfaces are also difficult to paint and should be phosphated beforehand. In this case. was developed during World War II in order to facilitate the painting of aluminum. the filaments.6 FILIFORM CORROSION Metals with organic coatings may undergo a type of corrosion resulting in numerous meandering thread-like filaments of corrosion product. instead of two. which provides its own surface treatment. It has been described by several investigators and reproduced in the laboratory [16]. and the synergistic improvement is greater for mild environments [14]. Red lead is not recommended. just before using. A coating system consisting of painted galvanized steel provides a protective service life up to 1. Zn and ZnO apparently react beforehand with organic acids of the vehicle. corresponding to the presence of . in general. because of the galvanic interaction of aluminum with metallic lead deposited by replacement of lead compounds.5 times that predicted by adding the expected lifetimes of the paint and the galvanized coating in a severe atmosphere. Subsequently. The prime coat should. This is sometimes known as underfilm corrosion. Otherwise. unless the wash primer is used. contain zinc chromate as an inhibiting pigment. It has the advantage of providing in one operation. on steel are typically 0. 16. which. The thread itself is red in color. or threads. zinc (galvanized steel).1 Wash Primer A wash primer. 16. but not red lead. a phosphating treatment of the metal and the application of the prime coat.296 ORGANIC COATINGS 16.2 Painting of Aluminum and Zinc Paints do not adhere well to aluminum without special surface treatment. One wash primer solution consists of approximately 9% polyvinyl butyral and 9% zinc tetroxychromate by weight in a mixture of isopropanol and butanol as one solution.5 mm wide. it was also found advantageous as a prime coat for steel and several other metals. and aluminum.% H3PO4 in isopropanol and water in the weight ratio of four of the pigmented solution to one of the latter [13].1).1–0. and the head is green or blue. characteristic of Fe2O3. zinc chromate. is mixed with a solution of 18 wt. and it was called filiform corrosion by Sharmon [15] (Fig. ensuring that aluminum soaps and other compounds do not form at the paint–metal interface to weaken the attachment of paint to metal. It has proved to be an effective prime coat on steel. According to reported descriptions. 16.5. phosphating or anodizing is suitable. WP1. Paints pigmented with zinc dust plus ZnO (zinc-rich paints) can also be used satisfactorily as a prime coat. As with aluminum. forming a good bond with the metal.5. The mixture must be used within 8–24 h after mixing. 840 h). [Reprinted with permission. M. (b) Clear varnish on steel. (a) Lacquered tin can. Copyright NACE International 1953. 279 (1953). Corrosion 9. and R.] .H. Bruhn. Filiform corrosion.(a) (b) Figure 16. D.1.. Laiderman. 10× (86% R. 1×. Van Loo. Oxygen continues to diffuse through the film. The precipitated oxide assumes a typical V shape because more alkali is produced in the region between the head and the body (more O2). as well as at the rear of the head. leaving this portion relatively dry.* The liberated OH− ions probably play an important role in undermining the film in view of the well-known ability of alkalies to destroy the bond between paints and metals (cathodic delamination). since it is less hygroscopic than ferrous salt solutions. 16. The mechanism appears to be a straightforward example of a differential aeration cell. In addition. but threads never cross each other. water again diffuses out through the film. This sets up a differential aeration cell with the cathode and accumulation of OH− ions in all regions where the film makes contact with the metal. Hence. compared to lower concentrations in the center of the head. preferably deaerated.2. aluminum. *The accumulation of alkali at the periphery can be demonstrated by placing a large drop of dilute sodium chloride solution (1–5%). containing a few drops of phenolphthalein and about 0. reacting with Fe2+ to form FeO·nH2O.1% K3Fe(CN)6 on an abraded surface of iron. the threads may broaden to form blisters. At 100% relative humidity. as is stated to be the case for paraffin [17]. The anode is located in the central and forward portions of the head attended by formation of Fe2+. zinc. 65–95%). If a head should approach another filament body. the periphery turns pink and the center turns blue . they probably also occur under opaque paint films.6. is oxidized by O2 to Fe2O3·nH2O.4 mm/day in random directions.298 ORGANIC COATINGS ferrous ions. The head is made up of a relatively concentrated solution of ferrous salts. they diffuse toward the center of the head. and chromium-plated nickel. Filiform corrosion occurs independent of light.. Although threads are visible only under clear lacquers or varnishes. Oxygen also diffuses through the film and reaches higher concentrations at the interface of head and body and at the periphery of the head. They may not form at all if the film is relatively impermeable to water. If a head approaches another thread. and bacteria. Fe2O3 exists predominantly. in turn. Behind the Vshaped interface. including steel.g. the previous participation of the film in filiform growth will have depleted the film of organic and inorganic anions necessary to the accumulation of high concentrations of ferrous salts in the head and also of cations necessary to build up a high pH at the periphery. 16.1 Theory of Filiform Corrosion Various schematic views of a filiform thread or filament are shown in Fig. as was shown by analysis [17]. metallurgical factors in the steel. which. it either glances off at an angle or stops growing. water tends to be absorbed from the atmosphere in this region of the thread. They have been observed under various types of paint vehicles and on various metals. Each thread grows at a constant rate of about 0. and. Within a few minutes. This type of corrosion takes place on steel only in air of high relative humidity (e. serving to keep the main portion of the filament cathodic to the head. magnesium. compared to the periphery of the head. the filament would stop growing. Schematic views of filiform filament on iron showing details of differential aeration cell causing attack.2. Furthermore. Wholly adequate remedies have not yet been found.PL ASTIC LININGS 299 Figure 16. 16. plus still greater abundance of oxygen. and perhaps more important. the previous accumulation of OH− added to that being formed. Phosphate surface treatments and chromate prime coats of paint serve to limit filiform corrosion. but apparently do not prevent its occurrence. This serves to discourage further growth of the filament in the direction of the old filament. . alkalies. and corrosive liquids and gases in general can be obtained by bonding a thick sheet of plastic or rubber to a steel surface. tends to ensure that the old filament body remains cathodic.7 PLASTIC LININGS Protection against acids. encouraging the approaching anode to veer off in another direction. a situation sometimes observed. If the head should lose its electrolyte because of delaminated film at the old thread which it approaches. Blume. P. Nylen and E. 168. Sci. exposed to the soil. Neoprene. REFERENCES 1. editor. Of Commerce. 5. 168. Chem. However. 20–20. Bureau of Census. 11. and all organic solvents up to about 250°C (480°F). 1965. Current Industrial Report MA325F(06)-1. and HNO3. R. Baboian. L. gaseous chlorine.C. such as are common in the chemical industry. J. P. Hudson. J. issued June 2007.C. gasket material. 5. Singer. Am. Ross et al. tape can become disbonded with time.panynj. 120. 18. including HF. (London) 55. This compound is used. T. and in contact with the groundwater that leaks into such disbondments. NACE International. . and M. Perf. and vinylidene chloride (Saran) are examples of materials 1 that are so used. J. J. Slow decomposition into HF and fluorinated hydrocarbons begins at temperatures above 200°C (400°F). 7. ibid. Ostermiller. Hudson. R. Port Authority of New York and New Jersey website: http://www. 165 (1951). 106 (1970).S. Chem. for linings. particularly to protect buried metal structures. Corros. Ind. J. As a material. Sunderland. Houston. Dept. The expense of such coatings usually excludes their use for any but severely corrosive environments. (with W. 9. L. Johnson). Such tape finds practical use for coating pipe and auxiliary equipment. 6704 (1998). In Automotive Corrosion and Protection. 20–9. Modern Surface Coatings. Eng. U.) 4. Mater. Toxic gases may also be liberated by heat generated during machining of the plastic.. p.html 3. 18 (1977). Plastic coatings of vinyl or polyethylene can be applied as an adhesive tape. 689. 8. including pipe connections and valves.. Interscience. Corros. H2SO4.300 ORGANIC COATINGS Rubber. Table 1. D. Its extreme inertness makes bonding to any surface difficult. pp. 2. and diaphragm valves. 1992. Washington. tending to creep readily under stress. Br.C. 153 (1951). Lay and A. Mayne. Chem. Soc. November 14. 1960. 3 (1961). This sequence of events has resulted in failure by S. TX. leaving the pipe uncoated. Paint and Allied Products: 2006. (p. 10. Levina. with the mixture of gases being highly toxic [19]. W. Iron Steel Inst. typically. 58. It resists concentrated boiling alkalies. not cathodically protected since the tape material is nonconductive. A thickness of 3 mm ( 8 in. New York. 505 (1978). It successfully resists aqua regia and boiling concentrated acids. L. Teflon is not strong. J. It is said to react only with elementary fluorine and with molten sodium.) or more ensures a relatively good diffusion barrier and protects the base metal against attack for a long time. of some underground highpressure oil and natural gas pipelines [18]. Piepho. 6. One of the most stable plastics in terms of resisting a wide variety of chemical media is tetrafluorethylene (Teflon).gov/ CommutingTravel/bridges/html/painting. News. 16(3). 621 (1944). 2nd edition. P. W. B. in ASM Handbook. Revie. Corrosion: Fundamentals. and composites for corrosion control. H. Organic coatings and linings. Vol. Revie. Materials Park. Wiley. C. Boca Raton. Corrosion Control through Organic Coatings.I. 2000. ASM International. Wiley. in Uhlig’s Corrosion Handbook.T. Baboian. and Protection. Hahin and R. R. 837–844. Testing. elastomers. pp. OH. G. Report of the Inquiry. Hare. Corrosion: Fundamentals. 965–1022. Materials Park. 2nd edition. 1126 (1952). 17. pp. Revie. M. pp. Bulletin. editor. 18. 13A. New York. OH. T. November 1996. (London) 46. ASM International. 1994. in ASM Handbook. 2006. D. Vol. pp. Vol. 2nd edition. 13A. ASM International. 2003. Corrosion 15. W. New York. Environmental regulation of surface engineering. pp. Smukowski. K. Norsworthy. R. 14. TX. 15. OH. Vol. J. Occupational Medical Service. 2nd edition. John J. Paint systems. Using plastics. editor. 13A. New York. 1023–1039. 2003. pp. Slabaugh and M. and Protection. GENERAL REFERENCES A. Tator. 16. Corrosion: Fundamentals. Thomson and R. Vol. OH. R. and Protection. 13. and Protection. pp. 153. 1041–1059. In ASM Handbook. 5. p. New York. Whiting. ASM International. R. J. Materials Park. in Uhlig’s Corrosion Handbook. Eng. Ind. editor. Chem. W. Corrosion control of steel by organic coatings. Wiley.GENER AL REFERENCES 301 12. Forsgren. 817–833. National Energy Board. 1151–1164. 2000. 911–917. . Materials Park. 2003. in Automotive Corrosion and Protection. pp. B. R. Public Inquiry Concerning Stress Corrosion Cracking on Canadian Oil and Gas Pipelines. 2000. in ASM Handbook. R. K. Campion. Calgary. W. 46. Grotheer. Taylor & Francis. December 1961. Nature. 2003. 1014 (1954). CRC Press. editor. Langill. Chem. 2000. B. Surface Engineering. in Uhlig’s Corrosion Handbook. Report MH-2-95. Revie. 1992. P. Testing of polymeric materials for corrosion control. pp. C. ASM International. FL. 311t (1959). 800. Tator. Testing. 19. in ASM Handbook. C. Materials Park. 21-2–21-3. 13A. Houston. Corrosion: Fundamentals. editor. Wiley. OH. Vincent. in Uhlig’s Corrosion Handbook. Testing. 248–256. Buchheit. L. Alberta. NACE International. W. R. Sharmon. Testing. Khaladkar. Selection and use of coatings for underground or submersion service. . therefore. They represent. the discussion in this chapter pertains to steel. Passivators are usually inorganic oxidizing substances (such as chromates. and (3) vaporphase inhibitors. There are several classes of inhibitors. by R. are usually organic substances that have only slight effect on the corrosion potential. being more efficient in this regard than most of the nonpassivating types.or centivolts. In general. Uhlig Copyright © 2008 John Wiley & Sons. such as the pickling inhibitors. Nonpassivating inhibitors. nitrites. the best inhibitors available for certain metal–environment combinations. changing it either in the noble or active direction usually by not more than a few milli. the passivatingtype inhibitors reduce corrosion rates to very low values. 303 . Except where noted. conveniently designated as follows: (1) passivators. The practice of corrosion inhibition is greatly influenced by new regulations that have been developed because of toxicity and environmental effects resulting Corrosion and Corrosion Control. (2) organic inhibitors.1 INTRODUCTION An inhibitor is a chemical substance that.17 INHIBITORS AND PASSIVATORS 17. effectively decreases the corrosion rate. and molybdates) that passivate the metal and shift the corrosion potential several tenths volt in the noble direction. Winston Revie and Herbert H. including slushing compounds and pickling inhibitors. Inc. when added in small concentration to an environment. 2 PASSIVATORS 17.1.304 INHIBITORS AND PASSIVATORS from industrial effluents.2. in applications where toxicity.1. Cr6+. For example.1 Mechanism of Passivation The theory of passivators has already been discussed in part. Hence. according to this viewpoint. must be at least chemically equivalent to the amount of passive film Figure 17. Only those ions can serve as passivators that have both an oxidizing capacity in the thermodynamic sense (noble oxidation–reduction potential) and that are readily reduced (shallow cathodic polarization curve). environmental damage. Polarization curves that show effect of passivator concentration on corrosion of iron. with the former reducing too sluggishly to achieve the required high value of icritical. SO2− or ClO− ions are not passivators for iron. a known carcinogen. nor are NO− ions compared to NO− . and pollution caused by these chemicals are important considerations. especially those that contain hexavalent chromium. initiating high current densities that exceed icritical for passivation at residual anodic areas. and should be referred to again. 17. in Section 6. 17. there is a trend to replace some widely used inhibitors. 3 2 because nitrates are reduced less rapidly than are nitrites. An oxidizing substance that reduces sluggishly does not induce passivity (dotted cathodic polarization curve) (schematic).3. 4 4 because they are not readily reduced. Passivators in contact with a metal surface act as depolarizers. as illustrated in Fig. The extent of chemical reduction on initial contact of a passivator with metal. . The critical concentration for CrO2− .. at threads of a pipe or in crevices). it is also essential to maintain as low a concentration of these ions as possible. thereby raising the critical passivator concentration to higher values as well. and eventually the cathodic polarization curve intersects the anodic curve in the active region instead of in the passive region alone (Fig. The equation assumes an adsorbed passive-film structure. A concentration of 10−3 M Na2CrO4 is equivalent to 0. and avoiding crevices and surface films of grease and dirt. for example. For the passive film on iron. this is on the order of 0. The total quantity of equivalents of chemically reduced chromate is found to be of this order. . Since consumption of passivators increases in the presence of chloride and sulfate ions. Below this concentration. the active potential at such areas in galvanic contact with passive areas elsewhere of noble potential promotes localized corrosion.1). The amount of chromate reduced in the passivation process was arrived at from measurements [1] of residual radioactivity of a washed iron surface after exposure to a chromate solution containing radioactive 51Cr. but the same reasoning applies whatever the structure. equivalent in the absence of dissolved oxygen to the value of ipassive [<0.015 C passive-film substance/cm2). at the active areas through the action of passive–active cells. consistent with ipassive that also decreases with time.5 × 10−7 eq or 0. the critical concentration of CrO2− and NO− is about 4 2 10−2 M [4. The following reaction applies.g. For this reason. and it is probably also the same for other passivators acting on iron. Lower concentrations of passivator correspond to more active values of the oxidation– reduction potential. Iron oxide and chromate reduction products slowly accumulate. such as higher H+ activity. or WO2− is about 10−3 to 4 2 4 4 −4 10 M [2–5]. Chloride ions and elevated temperatures increase icritical as well as ipassive.016% or 160 ppm. that less chromate is consumed as exposure time continues. as has been discussed earlier.3 μA/cm2 (<3 mA/m2)] based on observed corrosion rates of iron in chromate solutions.PASSIVATORS 305 formed as a result of such reduction. At 70–90°C. The rate of reduction increases with factors that increase ipassive. passivators behave as active depolarizers and increase the corrosion rate at localized areas. Should the passivator concentration fall below the critical value in stagnant areas (e. 17. it is important to maintain the concentration of passivators above the critical value at all portions of the inhibited system by stirring. and the presence of Cl−. such as pits. It is found. MoO2− .01 C/cm2 (100 C/m2) of apparent surface area. the concentration of passivator must exceed a certain critical value. higher temperatures. rapid flow rates. NO− . assuming that all reduced chromate (or dichromate) remains on the metal surface as adsorbed Cr3+ or as hydrated Cr2O3: 2 Cr2O7− + 8 H + + Fesurface → 2Cr 3+ + 4 H 2O + O2 ⋅ Oads on Fe ΔG = −406 kJ Residual radioactivity accounts for 3 × 1016 Cr atoms/cm2 (1. or pitting. 5]. in practice. For optimum inhibition. Reduction of passivator continues at a low rate after passivity is achieved. These are all nonoxidizing substances requiring dissolved oxygen in order to inhibit corrosion.14%) sodium benzoate.5% sodium benzoate solutions. 10] (NaPO3)n (Fig. iron phosphate.2. below this value the hydrogen evolution reaction.25 nm2. hence. and sodium polyphosphate [9. 0.01 M (0. to induce passivity.5 containing 17 ppm NaCl occurred at and above 0.5% sodium benzoate when coupled to gold.1. on which the reduction of oxygen is apparently not retarded. The mechanism of inhibition in the case of sodium polyphosphate solutions may depend in part on the ability of polyphosphates to interfere with oxygen reduction on iron surfaces. requires dissolved oxygen [6]. thereby. evidence of protective film formation of the diffusion-barrier type on cathodic . The mechanism of passivation is similar to that described in Sections 7.g.05% sodium benzoate or 0. there is. Na2O · nSiO2. decreases reduction of oxygen at cathodic areas. As little as 5 × 10−4 M sodium benzoate (0. In this category are alkaline compounds (e. whereas in the absence of O2. and 0. apparently through facilitating the adsorption of dissolved oxygen. NaOH. and 0.5. but inhibition is not observed in deaerated water. protection is supplemented by diffusion-barrier films of iron silicate.3 and 7. 0.12.3. producing passivity. sodium cinnamate [8] (C6H5 · CH · CH · COONa). or.007%) effectively inhibits steel in aerated distilled water [11]. a very small amount of benzoate on the metal surface increases adsorption of oxygen. pH 6. 7] (C6H5COONa).16 monomolecular layer of benzoate [0. Clearly. alternatively. and so on. presumably becomes dominant. for example. Gatos [12] found that optimum inhibition of steel in water of pH 7. 17. thereby decreasing the probability of a reaction between dissolved O2 and adsorbed H. Passivation of iron by molybdates and tungstates. Na3PO4. dissolved oxygen may help create just enough additional cathodic area to ensure anodic passivation of the remaining restricted anode surface at the prevailing rate of reduction of MoO2− or of 4 WO2− .2% sodium cinnamate. contrary to the situation for chromates and nitrites. High concentrations of OH− displace H adsorbed on the metal surface. icritical is not achieved. only 0. The steady-state corrosion rate of iron in aerated 0. In addition to passive films of this kind.2. The excess oxygen is then available to adsorb instead. 4 Sodium benzoate [6.07. This effect is specific to the cathodic areas of iron because iron continues to corrode in 0. for which benzoate ions have no inhibiting effect. Na2B4O7). In this case.8. oxygen is properly considered the passivating substance.306 INHIBITORS AND PASSIVATORS Some substances indirectly facilitate passivation of iron (and probably of some other metals as well) by making conditions more favorable for adsorption of oxygen. making it easier for dissolved oxygen to adsorb and. is only 0. whereas in deaerated solution the rate is 0. and inhibition was observed in all the solutions. Inhibition occurs only above about pH 5. both of which inhibit in the near-neutral pH range.1. and the passive film of oxygen is destroyed. By using radioactive 14C as tracer.001 gmd.2) are further examples of nonoxidizing compounds that effectively passivate iron in the nearneutral range.073 gmd. respectively.. roughness factor 3] was found on a steel surface exposed 24 h to 0.3. Other factors enter as well. The Electrochemical Society. Calcium. Protection of these metals apparently results largely from formation of relatively thick diffusion-barrier films of insoluble metal . using.2. A lesser degree of inhibition can be obtained with the nontransition metals. Copyright 1955. (Reproduced with permission. and Pb. 17.1. chromates. 17. their anodic polarization curves have the shape shown in Fig. iron. and zinc polyphosphates are better inhibitors than the sodium compound.5% NaCl solutions containing several hundred parts per million calcium polyphosphate [13].) areas. Cu. At low concentration of dissolved oxygen.PASSIVATORS 307 Figure 17. Zn. 25°C [9]. 48-h test. corrosion of iron is accelerated by sodium polyphosphate because of its metal-ion-complexing properties (Fig. for example. allowing passivity to be established and then maintained at low current densities. In line with the theory of passivators just described. and such diffusion-barrier films probably account for the observed inhibition of steel exposed to as high as 2.2). transition metals are those expected and found to be inhibited best by passivators. Effect of oxygen concentration on sodium polyphosphate as a corrosion inhibitor of iron showing beneficial effect of dissolved O2 and Ca2+. such as Mg. 1). 17.062 0.04–0.2 Applications of Passivators Chromates are applied mostly as inhibitors for recirculating cooling waters (e. The concentration of Na2CrO4 used for this purpose is about 0.039 0.040 0. 14-day tests Na2Cr2O7 · 2H2O (g/liter) → % NaCl 0 Temperature (°C) 20 75 95 20 75 95 20 75 95 20 20 0.266a 0. and Temperature on Corrosion of Mild Steel [14] Velocity of Specimen: 0.430 0.002 0.120a 0.5 by addition of NaOH. in the past..178a 0. Periodic analysis (colorimetric) is required to ensure that the concentration remains above the critical level (10−3 M or 0. The pH is adjusted.579 0.441a 0. with the higher concentrations being employed at higher temperatures or in fresh waters of chloride concentration above 10 ppm.5–3.014 0.015 0.034 0.785 2.055 0. of internal combustion engines.356a 0.022 0.001 0. Chlorides.131a 0.010 0.37 m/s.0 a Pitted.308 INHIBITORS AND PASSIVATORS TABLE 17.002 0.0 g Na2Cr2O7/liter adjusted with NaOH to form CrO2−).007 0.1. if necessary.001 0.668 1.020 0.000 0. Chromates are not adequate inhibitors for the hot concentrated brine solutions that.1 0.004 0. The reduction in corrosion rate is not as pronounced as when chlorides are absent [14] (see Table 17.539 0.905 0.042 0.606 0. to 7. There is also the possibility that adsorption of CrO2− on the metal surface contributes in some degree to the lower reaction rate 4 by decreasing the exchange current density for the reaction M → M2+ + 2e−.037 0.001 0. and any reduction that occurs apparently results from formation of a surface diffusion barrier of chromate reduction products and iron oxides. Effect of Chromate Concentration. and cooling towers).185 0 0.039 0.004 0.529 0.5 22.120a 0. chromates mixed with oxides.5–9.030 0. In the presence of 4 so large a Cl− concentration.2%.g.05 3.0 Corrosion Rate (mm/y) 0.016% Na2CrO4 at room .002 0.000 0.066 0.1) does not take place. passivity of the kind discussed under Definition 1 (Section 6. An inhibiting mechanism similar to that for nontransition metals in contact with passivators probably also applies to steel in concentrated refrigerating brines (NaCl or CaCl2) to which chromates are added as inhibitors (approximately 1.5 1.705 0. were sometimes mistakenly proposed as antifreeze solutions for engine cooling systems.011 0.2.158 0. rectifiers. 2. They are used to inhibit cutting oil–water emulsions employed in the machining of metals (0. and so they must be used with caution and with full regard to disposal requirements. it releases dissolved water which. unlike chromates. Chromates (Cr6+) are toxic and carcinogenic and cause a rash on prolonged contact with the skin. is higher than the critical concentration in distilled water.1–0. where water is a very minor phase.02 0. they have little tendency to react with alcohols or ethylene glycol. they decompose. in contact with large quantities of oxygen dissolved in the gasoline (solubility of O2 in gasoline is six times that in water). Sodium nitrite enters the water phase and effectively inhibits rusting. Room Temperature % NaNO2 0.0 0.0. Nitrites are used as inhibitors for antifreeze cooling waters (see Section 18.0 0.10 Corrosion Rate (mm/y) 0. on reaching lower temperatures underground. The corrosion rates of steel in contact with water–gasoline mixtures containing increasing amounts of NaNO2 are listed in Table 17.04 0. pH 9.1 [14]. 14-Day Exposure. forming voluminous rust products that may clog the line. gasoline is corrosive to steel because. sufficient nitrite or chromate solution may be continuously injected to give a 2% concentration in the water phase [17]. corrodes steel. Chromates used for the same purpose have the disadvantage that they tend to react with some constituents of the gasoline. In pipelines transporting gasoline and other petroleum products. They are not so well-suited to cooling tower waters because they are gradually decomposed by bacteria [16].1) because. Corrosion rates of mild steel as a function of chromate and chloride concentration at various temperatures are shown in Table 17. but there may be adequate protection against pitting for the treatment of very large volumes of water employing cooling towers [15]. Nitrites are inhibitors only above about pH 6.2.PASSIVATORS 309 temperature). Combinations of chromates with polyphosphates or other inhibitors may permit the concentration of chromates to fall below the critical level. because of impurities present in the water. forming volatile nitric oxide and nitrogen TABLE 17.0 .02 0. and Gasoline.2 [18].08 0.06% or 7 × 10−3 M which. The minimum amount of NaNO2 for effective inhibition is 0. In more acid environments. This reduction in chromate concentration results in some sacrifice of inhibiting efficiency.2%). Corrosion Rates of Mild Steel in Sodium Nitrite Solutions Containing Gasoline [18] Rotating Bottle Tests Using Pipeline Water.11 0.06 0. In this connection. 4 and 5. contrary to the situation for chro4 mates [3–5] (Table 17. These include the organic N. Compounds serving as pickling inhibitors require. although the corrosion rate may be appreciably reduced (Fig. and both reactions tend to be retarded.1 V). the corrosion potential of steel is not greatly altered (<0. they tend to induce pitting at concentrations near the critical in presence of Cl− or SO2− ions. on addition of an inhibitor to an acid. or vice versa. amine. S.3). a favorable polar group or groups by which the molecule can attach itself to the metal surface. For example. cathodic polar- . In common with other passivators. the corrosion of iron in 1N hydrochloric acid was found to be inhibited by derivatives of thioglycolic acid and 3-mercaptopropionic acid to an extent that varied systematically with chain length [20]. Stagnant Solutions Critical Concentrations NaCl Na2CrO4 NaNO2 200 ppm 500 50 ppm 100 500 12 ppm 30 210 460 >2000 Na2SO4 55 ppm 120 20 55 450 a See also Refs. probably no more than a monolayer in thickness.3. and electric charge of the molecule play a part in the effectiveness of inhibition. 17. For inhibitors that adsorb better at increasingly active potential as measured from the point of zero surface charge (potential of minimum ionic adsorption). 17. For example.310 INHIBITORS AND PASSIVATORS TABLE 17. shape. Whether a compound adsorbs on a given metal and the relative strength of the adsorbed bond often depend on factors such as surface charge of the metal [21]. both iodide and quinoline are reported to inhibit corrosion of iron in hydrochloric acid by this mechanism [19].3).3 PICKLING INHIBITORS Most pickling inhibitors function by forming an adsorbed layer on the metal surface. orientation. In this regard. which essentially blocks discharge of H+ and dissolution of metal ions. 4 nitrates are less sensitive to Cl− than to SO2− . Some inhibitors block the cathodic reaction (raise hydrogen overpotential) more than the anodic reaction. Critical Concentrations of NaCl or Na2SO4 above which Pitting of Armco Iron in Chromate or Nitrite Solutions Occurs [11]a 5-Day Tests. but adsorption appears to be general over all the surface rather than at specific anodic or cathodic sites. 25°C. The size. and OH groups. Hence. by and large. peroxide. PICKLING INHIBITORS 311 Figure 17.3. Polarization diagram for steel corroding in pickling acid with and without inhibitor. ization in presence of the inhibitor provides better protection than either the equivalent cathodic protection or use of an inhibitor alone. Antropov [22] demonstrated this for iron and zinc in sulfuric acid containing various organic inhibitors. The anion of the pickling acid may also take part in the adsorbed film or double-layer structure, accounting for differing efficiencies of inhibition for the same compound in HCl compared to H2SO4. For example, the corrosion rate of steel at 20°C inhibited with 20 g/liter quinoline is 26 gmd in 2N H2SO4, but only 4.8 gmd in 2N HCl, whereas the corrosion rates in the absence of inhibitor are 36 and 24 gmd, respectively [23]. In addition, specific electronic interaction of polar groups with the metal (chemisorption) may account for a given compound being a good inhibitor for iron, but not for zinc, or vice versa. The latter factor in certain cases may be more important than the steric factor (diffusion-barrier properties) of a closely packed, oriented layer of high-molecular-weight molecules. This is shown by the outstanding inhibition provided by the simple molecule carbon monoxide, CO, dissolved in HCl to which 18–8 stainless steel is exposed [24]; in 6.3N HCl at 25°C, the inhibition efficiency is 99.8%, where the inhibition efficiency is defined as 0 Percent inhibition efficiency = (rateno inhib − ratewith inhib) × 100 /rateno inhib (17.1) Very effective inhibition of corrosion of iron is also provided by a small amount of iodide in dilute H2SO4 [25]. Both CO and iodide chemisorb on the metal surface, interfering mostly with the anodic reaction [26]. Kaesche [27] showed that 10−3 M KI is a much more effective inhibitor for iron in 0.5M Na2SO4 solution at pH 1 (89% efficient) compared to pH 2.5 (17% efficient), indicating that adsorption of iodide, particularly in this pH range, is pH-dependent. The interaction of factors just mentioned, plus perhaps others, enter into accounting for the fact that some compounds (e.g., o-tolylthiourea [28] in 5% 312 INHIBITORS AND PASSIVATORS H2SO4) are better inhibitors at elevated temperatures than at room temperature, presumably because adsorption increases or the film structure becomes more favorable at higher temperatures. Others (e.g., quinoline derivatives) are more efficient in the lower-temperature range [28]. Typical effective organic pickling inhibitors for steel are quinolinethiodide, o- and p-tolylthiourea, propyl sulfide, diamyl amine, formaldehyde, and p-thiocresol. Inhibitors containing sulfur, although effective otherwise, sometimes induce hydrogen embrittlement of steel in the event that the compound itself or hydrolysis products (e.g., H2S) are formed that favor entrance of H atoms into the metal (see Section 5.5). In principle, inhibitors containing arsenic or phosphorus can behave similarly. 17.3.1 Applications of Pickling Inhibitors Pickling inhibitors used in practice are seldom pure compounds. They are usually mixtures, which may be a byproduct, for example, of some industrial chemical process for which the active constituent is unknown. They are added to a pickling acid in small concentration, usually on the order of 0.01–0.1%. The typical effect of concentration of an inhibitor on the reaction between steel and 5% H2SO4 is illustrated [28] in Fig. 17.4, showing that, above a relatively low concentration, presumably that necessary to form an adsorbed monolayer, an additional amount of inhibitor has little effect on further reducing the rate. Inhibitors are commonly used in the acid pickling of hot-rolled steel products in order to remove mill scale. The advantages of using an inhibitor for this purpose are (1) saving of steel, (2) saving of acid, and (3) reduction of acid fumes Figure 17.4. Effect of inhibitor concentration on corrosion of 0.1% C steel in 5% H2SO4, 70°C [28]. VAPOR-PHASE INHIBITORS 313 caused by hydrogen evolution. Inhibited dilute sulfuric or hydrochloric acid is also used to clean out steel water pipes clogged with rust, to clean boiler tubes encrusted with CaCO3 or iron oxide scales, and to activate oil wells underground, the inhibitor protecting steel oil-well tubing. For example, boiler scale can be removed by using 0.1% hexamethylene tetramine in 10% HCl at a maximum temperature of 70°C (160°F) [29]. 17.4 SLUSHING COMPOUNDS Slushing compounds are used to protect steel surfaces temporarily from rusting during shipment or storage. They consist of oils, greases, or waxes containing small amounts of organic additives. The latter are polar compounds that adsorb on the metal surface in the form of a closely packed oriented layer. In this respect, the mechanism of inhibition by organic additives is similar to that of inhibition by pickling inhibitors, except that for use as slushing compounds, additives must suitably adsorb in the near-neutral range of pH, whereas pickling inhibitors adsorb best in the low pH range, calling for somewhat different properties. Whether an oil or a wax is chosen as the vehicle depends on (1) the relative length of time protection is desired, the wax usually providing longer life and (2) the ease of removal before the protected machine part is put into service, with an oil being easier to wipe off or dissolve in solvents. The thickness of an applied coat varies from 0.1 mm to more than 2.5 mm (5 to more than 100 mils). Suitable organic additives for use in slushing compounds include organic amines, zinc naphthenate, various oxidation products of petroleum, alkali and alkaline-earth metal salts of sulfonated oils, and various other compounds [30]. A substance that has been used successfully is lanolin, obtained from wool scouring; its active constituents are various high-molecular-weight fatty alcohols and acids. Sometimes, lead soaps are added to slushing compounds, with these soaps reacting to form relatively insoluble PbCl2 with any NaCl from perspiration transferred to steel surfaces through handling. 17.5 VAPOR-PHASE INHIBITORS Substances of low but significant vapor pressure, the vapor of which has corrosion-inhibiting properties, are called vapor-phase inhibitors. They are used to protect critical machine parts (e.g., ball bearings or other manufactured steel articles) temporarily against rusting by moisture during shipping or storage. They have the advantage over slushing compounds of easy application, with the possibility of immediate use of the protected article without first removing a residual oil or grease film. They have the disadvantgage of accelerating the corrosion of some nonferrous metals, discoloring some plastics, and requiring relatively effective sealing of a package against loss of the inhibiting vapor. The latter requirement is relatively easily achieved, however, by using wrapping paper impregnated 314 INHIBITORS AND PASSIVATORS on the inside surface with the inhibitor and incorporating a vapor-barrier coating on the outside. The mechanism of inhibition has not been studied in detail, but it appears to be one of adsorbed film formation on the metal surface that provides protection against water or oxygen, or both. In the case of volatile nitrites, the inhibitor may also supply a certain amount of NO− that passivates the surface. 2 Detailed data have been presented for dicyclohexylammonium nitrite [31], which is one of the most effective of the vapor-phase inhibitors. This substance is white, crystalline, almost odorless, and relatively nontoxic. It has a vapor pressure of 0.0001 mmHg at 21°C (70°F), which is about one-tenth the vapor pressure of mercury itself.* One gram saturates about 550 m3 (20,000 ft3) of air, rendering the air relatively noncorrosive to steel. The compound decomposes slowly; nevertheless, in properly packaged paper containers at room temperature, it effectively inhibits corrosion of steel over a period of years. However, it should be used with caution in contact with nonferrous metals. In particular, corrosion of zinc, magnesium, and cadmium is accelerated. Cyclohexylamine carbonate has the somewhat higher vapor pressure of 0.4 mmHg at 25°C and its vapor also effectively inhibits steel [32]. The higher vapor pressure provides more rapid inhibition of steel surfaces either during packaging or on opening and again on closing a package, during which time the concentration of vapor may fall below that required for protection. The vapor is stated to reduce corrosion of aluminum, solder, and zinc, but it has no inhibiting effect on cadmium, and it increases corrosion of copper, brass, and magnesium. Ethanolamine carbonate and various other compounds have also been described as vapor-phase inhibitors [23, 32]. A combination of urea and sodium nitrite has found practical application, including use in impregnated paper. The mixture probably reacts in the presence of moisture to form ammonium nitrite, which is volatile, although unstable, and conveys inhibiting nitrite ions to the metal surface. 17.5.1 Inhibitor to Reduce Tarnishing of Copper Benzotriazole (BTA), C6H5N3, is a corrosion inhibitor widely used for copper and its alloys. Typical protective films formed on copper are reported to consist of various multilayers. Film thickness is normally controlled by the pH of the solution, ranging from less than 0.01 μm in near-neutral and mildly basic solutions to 0.5 μm in acidic solutions [33]. For a wide variety of conditions [34], this inhibitor is reported to form a Cu+ surface complex following its adsorption on Cu2O that is present on copper surfaces. *I. Rosenfeld et al. (Symposium sur les Inhibiteurs de Corrosion, University of Ferrara, Italy, 1961, p. 344) report the lower value of 0.00001 mmHg obtained for dicyclohexylammonium nitrite carefully purified by multiple recrystallization from alcohol. REFERENCES 315 REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. H. Uhlig and P. King, J. Electrochem. Soc. 106, 1 (1959). W. Robertson, J. Electrochem. Soc. 98, 94 (1951). S. Matsuda and H. Uhlig, J. Electrochem. Soc. 111, 156 (1964). A. Mercer and I. Jenkins, Br. Corros. J. 3, 130 (1968). A. Mercer, I. Jenkins, and J. Rhoades-Brown, Br. Corros. J. 3, 136 (1968). M. Pryor and M. Cohen, J. Electrochem. Soc. 100, 203 (1953). D. Brasher and A. Mercer, Br. Corros. J. 3, 120 (1968). F. Wormwell, Chem. Ind. (London) 1953, 556. H. Uhlig, D. Triadis, and M. Stern, J. Electrochem. Soc. 102, 59 (1955). G. Hatch, Mater. Prot. 8, 31 (1969). S. Matsuda, D. Sc. thesis, Department of Metallurgy, M.I.T., Cambridge, MA, 1960. H. Gatos, in Symposium sur les Inhibiteurs de Corrosion, University of Ferrara, Italy, 1961, p. 257. G. Butler and D. Owen, Corros. Sci. 9, 603 (1969). B. Roetheli and G. Cox, Ind. Eng. Chem. 23, 1084 (1931). H. Kahler and P. Gaughan, Ind. Eng. Chem. 44, 1770 (1952). T. P. Hoar, Corrosion 14, 103t (1958). J. Bregman, Inhibition in the petroleum industry, in Proceedings of the Third European Symposium on Corrosion Inhibitors, University of Ferrara, Italy, 1970, p. 365. A. Wachter and S. Smith, Ind. Eng. Chem. 35, 358 (1943). K. Rajagopalan and G. Venkatachari, Corrosion 36, 320 (1980). M. Carroll, K. Travis, and J. Noggle, Corrosion 31, 123 (1975). L. Antropov, Kinetics of Electrode Processes and Null Points of Metals, Council of Scientific & Industrial Research, New Delhi, 1960, pp. 48–82. L. Antropov, Inhibitors of metallic corrosion and the phi-scale of potentials, First International Congress on Metallic Corrosion, Butterworths, London, 1961, p. 147. I. Putilova, S. Balezin, and V. Barannik, Metallic Corrosion Inhibitors, G. Ryback, translator, Pergamon Press, New York, 1960, p. 63. H. Uhlig, Ind. Eng. Chem. 32, 1490 (1940). K. Hager and M. Rosenthal, Ordnance 35, 479 (1951). K. Heusler and G. Cartledge, J. Electrochem. Soc. 108, 732 (1961). H. Kaesche, in Symposium sur les Inhibiteurs de Corrosion, University of Ferrara, Italy, 1961, p. 137. T. Hoar and R. Holliday, J. Appl. Chem. 3, 502 (1953). W. Cerna, Proceedings of the Seventh Annual Water Conference, Pittsburgh, PA, 1947. H. Baker and W. Zisman, Ind. Eng. Chem. 40, 2338 (1948). A. Wachter, T. Skei, and N. Stillman, Corrosion 7, 284 (1951). E. Stroud and W. Vernon, J. Appl. Chem. 2, 178 (1952). I. Ogle and G. Poling, Can. Metall. Q. 14, 37 (1975). D. Chadwick and T. Hashemi, Corros. Sci. 18, 39 (1978). 316 INHIBITORS AND PASSIVATORS GENERAL REFERENCES B. P. Boffardi, Corrosion inhibitors in the water treatment industry, in ASM Handbook, Vol. 13A, Corrosion: Fundamentals, Testing, and Protection, ASM International, Materials Park, OH, 2003, pp. 891–906. R. L. Martin, Corrosion inhibitors for oil and gas production, in ASM Handbook, Vol. 13A, Corrosion: Fundamentals, Testing, and Protection, ASM International, Materials Park, OH, 2003, pp. 878–886. S. Papavinasam, Corrosion inhibitors, in Uhlig’s Corrosion Handbook, 2nd edition, R. W. Revie, editor, Wiley, New York, 2000, pp. 1089–1105. S. Papavinasam, Evaluation and selection of corrosion inhibitors, in Uhlig’s Corrosion Handbook, 2nd edition, R. W. Revie, editor, Wiley, New York, 2000, pp. 1169–1178. J. D. Poindexter, Corrosion inhibitors for crude oil refineries, in ASM Handbook, Vol. 13A, Corrosion: Fundamentals, Testing, and Protection, ASM International, Materials Park, OH, 2003, pp. 887–890. V. S. Sastri, Corrosion Inhibitors: Principles and Applications, Wiley, Chichester, 1998. K. L. Vasanth, Vapor phase corrosion inhibitors, in ASM Handbook, Vol. 13A, Corrosion: Fundamentals, Testing, and Protection, ASM International, Materials Park, OH, 2003, pp. 871–877. 18 TREATMENT OF WATER AND STEAM SYSTEMS 18.1 DEAERATION AND DEACTIVATION In accord with principles described in Sections 7.1 and 7.2, corrosion of iron is negligible at ordinary temperatures in water that is free of dissolved oxygen. An effective practical means, consequently, for reducing corrosion of iron or steel in contact with fresh water or seawater is to reduce the dissolved oxygen content. In this way, corrosion of copper, brass, zinc, and lead is also minimized. Removal of dissolved oxygen from water is accomplished either by chemically reacting oxygen beforehand, called deactivation, or by distilling it off in suitable equipment, called deaeration. Deactivation can be carried out in practice by slowly flowing hot water over a large surface of steel laths or sheet contained in a closed tank called a deactivator. The water remains in contact long enough to corrode the steel, and, by this process, most of the dissolved oxygen is consumed. Subsequent filtration removes suspended rust. Water so treated is much less corrosive to a metal pipe distribution system. Deactivators of this kind have been employed in some buildings, but, since use of deactivation requires regular attention in addition to periodic renewal of the steel sheet, this approach to Corrosion and Corrosion Control, by R. Winston Revie and Herbert H. Uhlig Copyright © 2008 John Wiley & Sons, Inc. 317 318 TREATMENT OF WATER AND STEAM SYSTEMS oxygen removal is usually too cumbersome compared with the use of chemical inhibitors or deaeration. Deactivation of industrial waters (not potable waters, because the chemicals used are toxic) is possible by using oxygen scavengers (strong reducing agents), such as sodium sulfite (Na2SO3) and hydrazine (N2H2), which function as corrosion inhibitors by removing the oxygen. Since hydrazine has been identified as a possible carcinogen, there have been considerable efforts to replace it. Sodium sulfite scavenges oxygen in accord with the reaction 1 Na 2 SO2 + O2 → Na 2 SO4 2 (18.1) for which Na2SO3 reacts with oxygen in the weight ratio of 8 : 1 (8 kg Na2SO3 to 1 kg O2). The reaction is relatively fast at elevated temperatures, but slow at ordinary temperatures. It can be accelerated by adding catalysts, such as Cu2+ or Co2+, salts [1, 2]. The rapid decrease with time of dissolved oxygen in the San Joaquim, California, river water after treatment with 80 ppm Na2SO3 (0.67 lb/1000 gal) plus copper or cobalt salts is shown in Fig. 18.1. Water so treated using CoCl2 as catalyst was found by Pye [1] to be noncorrosive to a steel heat-exchanger system that, without treatment, had previously suffered serious corrosion and loss of heat transfer. Tests showed a reduction in corrosion rate from 0.2 mm/y (0.008 ipy, pitting factor = 7.4) before treatment to 0.004 mm/y (0.00016 ipy) afterward. Hydrazine (N2H4), supplied as a concentrated aqueous solution, also reacts with dissolved oxygen, according to Figure 18.1. Effect of cobalt and copper salts on reaction rate of Na2SO3 with dissolved oxygen at room temperature [1]. [Reprinted from Journal AWWA 39 (11) (November 1947), by permission. Copyright 1947, American Water Works Association.] DEAER ATION AND DEAC TIVATION 319 N 2 H 4 + O2 → N 2 + 2 H 2O (18.2) in the weight ratio of 1 : 1. This reaction is slow at ordinary temperatures, but can be accelerated by using catalysts (e.g., activated charcoal, metal oxides, alkaline solutions [3] of Cu2+ and Mn2+) and by raising the temperature. At elevated temperatures, decomposition occurs slowly at 175 °C (345 °F) and more rapidly at 300 °C (570 °F), producing ammonia [4]: 3N 2 H 4 → N 2 + 4 NH 3 (18.3) The normal reaction products—nitrogen, water, and a small amount of NH3—are all volatile, and, unlike sulfite addition, no dissolved solids accumulate in the treated water. Two additional oxygen scavengers are carbohydrazide, (NH2NH)2CO, and diethylhydroxylamine (DEHA), (C2H5)2NOH; the reactions with oxygen are (NH 2 NH)2 CO + 2O2 → 2 N 2 + 3H 2O + CO2 and 4(C 2 H 5)2 NOH + 9O2 → 8CH 3COOH + 2 N 2 + 6H 2O (18.5) (18.4) Both these oxygen scavengers have the additional advantage that they form protective films on both iron and copper [5]. Deaeration is accomplished by spraying water or flowing it over a large surface, countercurrent to steam. Oxygen distills off and also some dissolved carbon dioxide (Fig. 18.2). Water is heated in the process and is suitable, therefore, as feedwater for boilers. Steam deaerators of this kind are standard equipment for all high-pressure stationary boilers. On the other hand, if the water is to be used cold, dissolved gases are distilled off by lowering the pressure, employing a mechanical pump or steam ejector instead of a countercurrent flow of steam. This is called vacuum deaeration. Equipment of this kind has been designed to deaerate several million gallons of water per day. In principle, it is more difficult and more expensive to remove the last traces of dissolved oxygen by distillation compared to the first 90–95%, and it is more difficult at low temperatures than at high temperatures. To achieve a low enough oxygen level in cold water, it is often necessary to use multiple-stage vacuum treatment. Fortunately, acceptable levels of dissolved oxygen for corrosion control in cold water are higher than in hot water or in steam. Allowable levels established through experience are given in Table 18.1 [6]. 320 TREATMENT OF WATER AND STEAM SYSTEMS Figure 18.2. Sketch of one type of steam deaerator. TABLE 18.1. Approximate Allowable Oxygen Concentration in Deaerated Water for Corrosion Control in Steel Systems Maximum Oxygen Concentration ppm Cold water Hot water Low-pressure boilers (<1.7 MPa, <250 psi) High-pressure boilers 0.3 0.1 0.03 0.005 mL/liter 0.2 0.07 0.02 0.0035 In hot-water systems. and it is also effective for reducing corrosion of hot-water systems.1). Zinc orthophosphate is effective in controlling corrosion of steel. nontoxic chemicals. Under other conditions (e.2.5 (see Section 7. Hot-Water Heating Systems. corrosion is negligible once the dissolved oxygen is consumed. must be taken into account in arriving at the proper proportions of added chemicals. such as polyphosphates.g. which causes scaling. and it also eliminates blue staining by water that has passed through copper or brass piping. For the treatment to be successful. Soft waters. In the presence of silicate. This treatment reduces “red water” caused by suspended rust resulting from corrosion of ferrous piping. Municipal Water Supplies. but it is clear that dissolved calcium and magnesium salts have an effect and that some protection may result alone from the alkaline properties of sodium silicate.6. lead. hard waters of positive saturation index are relatively noncorrosive and do not require treatment of any kind for corrosion control. and silicates. and brass is also reduced by this treatment. are readily contaminated with toxic quantities of lead salts on passing through lead piping. Sodium hydroxide induces similar passivity and corresponding low corrosion rates at the somewhat higher pH range of 10–11. on the other hand. Raising the saturation index is a potentially useful means for reducing the corrosion rate in either flowing or stagnant portions of the distribution system. This treatment requires addition of lime [Ca(OH)2] or both lime and soda ash (Na2CO3) to the water in amounts that raise the saturation index to about +0. copper. The conditions under which protection exists or is optimum are not entirely understood.HOT. pH 8). Usually. These are closed steel systems in which the initial corrosion of the system soon consumes dissolved oxygen. and lead.5. The hydrogen reaction can be minimized by treating the water with NaOH or Na3PO4 to a pH of 8. orthophosphates. with a characteristic odor due to traces of hydrocarbon gases originating from the reaction of carbides in steel with water.7 gmd [7]. the water must be low in colloidal matter and in dissolved solids other than calcium salts. A continuing minor reaction of steel with water produces hydrogen.2 HOT.1–0. Zinc polyphosphate has been used for more than 60 years in controlling corrosion of steel in municipal water systems.AND COLD-WATER TREATMENT 321 18. passivity of iron may be observed at pH 10 with an accompanying reduction of the corrosion rate to 0. Sodium silicate treatment in the amount of about 4–15 ppm SiO2 is sometimes used by individual owners of buildings in soft-water areas. Chemical treatment of potable waters is limited to small concentrations of inexpensive. the possibility of excess deposition of CaCO3. Corrosion of copper. and cause blue staining of bathroom fixtures by copper salts originating from slight corrosion of copper or brass piping.AND COLD-WATER TREATMENT 1. apparently containing SiO2 . cause rapid accumulation of rust in ferrous piping. a protective diffusion-barrier film is formed. 2.. Domestic or industrial hot-water heaters of galvanized steel through which hot aerated water passes continuously are not protected reliably in all types of water by nontoxic chemical additions such as silicates or polyphosphates. chromates must be handled. Adjustment of the saturation index to a more positive value. 2000 ppm sodium chromate (0. Otherwise. 18.2%) can be used to reduce damage by cavitation-erosion as well as by aqueous corrosion (see Section 7. cathodic protection or use of nonferrous metals. as discussed earlier. which produce an oily protective coating. such as copper or 70% Ni–Cu (Monel). but this also limits the available number of suitable inhibitors. Recirculating cooling waters. 16 ppm Cl−). For diesel or other heavy-duty engines.2. Sulfonated oils. and soldered components of such systems. is sometimes helpful. Adjustment of the saturation index to a more positive value is sometimes a practical possibility. additions must be made periodically in order to maintain the concentration above the critical. In such small concentrations. polyphosphates are not toxic.2% (or the equivalent amount of Na2Cr2O7·2H2O plus alkali to pH 8). used. a marked decrease in the corrosion rate was observed. brass. In the United States. Often. a protective coating or metals more corrosion resistant than steel must be used.5.322 TREATMENT OF WATER AND STEAM SYSTEMS and consisting perhaps of an insoluble iron silicate. Many proprietary inhibitor mixtures are on the market that are usually dissolved beforehand in methanol or in ethylene glycol in order to simplify the packaging problem. or wells) usually cannot be treated chemically. Chromates cannot be used in the presence of antifreeze solutions. If larger quantitites of sodium silicate were added to raise the pH of the water to 10 or 11. may be treated with sodium chromate (Na2CrO4) in the amount of 0. but water disposal may continue to be a problem because of the need to avoid accumulation of phosphates in rivers and lakes.1 Cooling Waters Once-through cooling waters (usually obtained from rivers. as for engine-cooling systems. 44 ppm Ca. which specifically inhibits corrosion of copper and at the same time removes the accelerating influence of . Sometimes.1). aluminum. the range in which passivity of iron occurs. and disposed of appropriately and with regard to regulatory requirements.04–0. both because of the large quantities of inhibitors required and because of the problem of water pollution. no inhibition was obtained in the laboratory at the same SiO2 content using Cambridge tap water (pH 8. However. 10 ppm Mg.3. is the best or only practical measure. because of their tendency to react with organic substances. Because of the toxicity of hexavalent chromium (Cr+6). and mercaptobenzothiazole. lakes. borax (Na2B4O7·10H2O) is a common ingredient. copper. Chromates inhibit corrosion of steel. Since chromate is consumed slowly. Laboratory tests in distilled water at 25 °C showed a reduction in the corrosion rate of iron on the order of 85–90% when sodium silicate was added to bring the pH to 8 [7]. additions of about 2–5 ppm sodium or calcium polyphosphate are made to help reduce corrosion of steel equipment.2. 7% borax. 0. Other nontoxic. The pH value is adjusted to 5–6 in order to minimize pitting and tubercle formation.06% Na2HPO4.. Sodium polyphosphate is often used in a concentration of about 10–100 ppm. Windage losses (loss of spray by wind) must be carefully avoided because chromates are toxic. has also been reported to be a useful component of inhibitor formulations for use in recirculating aqueous systems [9. they favor algae growth. in the presence of calcium or magnesium ions. One specification calls for a final concentration in the antifreeze solution of 1.3. However. which is less toxic than sodium chromate. but polyphosphates in low concentration are not toxic. chromatefree inhibitor formulations are based on mixtures of sorbitol. The steam may afterward pass to a superheater of higheralloy steel at higher temperature than the boiler itself. Unlike chromates. and the required optimum amount of inhibitor is less than that for chromates. but they consist essentially of a low-carbon steel or low-alloy steel container for water that is heated by hot gases.3 BOILER-WATER TREATMENT 18. sometimes with added zinc salts to improve inhibition. and as the sulfate and chloride concentrations increase through evaporation of the water. Toxicity also makes it difficult to dispose of chromate solutions whenever it becomes necessary to reduce the concentration of accumulated chlorides and sulfates. with the latter being added specifically to protect aluminum. causing scale formation on the warmer parts of the system. and 0. Polyphosphates decompose slowly into orthophosphates. numerous inhibitor systems have been developed as alternatives—for example. benzotriazole or tolytriazole. chromates are the most reliable from the standpoint of efficient inhibition. For industrial water cooled by recirculation through a spray or tray-type tower.1 Boiler Corrosion Steam boilers are constructed according to various designs. and water-soluble phosphates [9]. Nontoxic inhibitor formulations containing mixtures of azoles and watersoluble phosphates (e. are sometimes added to borax.g. Corrosion inhibition with polyphosphates is less effective than that by chromates. precipitate insoluble calcium or magnesium orthophosphate. chromates tend to cause pitting or may cause increased galvanic effects at dissimilar metal couples. organic phosphonic acids. 10]. which necessitates the addition of algaecides to the water. as well as scale deposition. the critical concentration is high. which are effective in alkaline waters and are biodegradable [8]. 18.1% mercaptobenzothiazole. Because of pollution problems that can be caused by chromates. which. Sodium molybdate. For maximum heat . Ethanolamine phosphate is also used as an inhibitor for engine-cooling systems containing ethylene glycol.BOILER-WATER TREATMENT 323 dissolved Cu2+ on corrosion of other portions of the system. disodium phosphate and sodium tripolyphosphate) have been developed [9]. less frequently.324 TREATMENT OF WATER AND STEAM SYSTEMS transfer. The history and causes of corrosion in power-station boilers have been reviewed by Mann [11]. Observation of the detector is Figure 18. The steam. and 90-day tests. after doing work or completing some other kind of service. 60-. a series of boiler tubes is usually used. or seawater. case histories of corrosion and its prevention in industrial boilers have been presented by Frey [12]. adjustment of this screw regulates a slight leak of hot boiler water in the region where the specimen is subject to maximum tensile stress and where boiler water evaporates. brackish.3. Embrittlement detector which. Some boilers are equipped with an embrittlement detector by means of which the chemical treatment of a water can be evaluated continuously in terms of its potential ability to induce stress-corrosion cracking (Fig. A boiler water is considered to have no embrittling tendency if specimens do not crack within successive 30-.3) [13]. through the inner surface of the tubes. eventually reaches a condenser that is usually constructed of copper-base alloy tubes. 18. Steam is cooled on one side of the tubes by water passing along the opposite side of a quality ranging from fresh to polluted. when attached to an operating boiler. A specimen of plastically deformed boiler steel is stressed by setting a screw. detects tendency of a boiler water to induce stress-corrosion cracking. The condensed steam then returns to the boiler and the cycle repeats. with the hot gases passing either around the outer surface or. . causing pitting or sometimes grooving of the boiler tubes. On the steam side.6) The mechanism of this reaction. Ferrous hydroxide eventually decomposes. on the other hand. indicate that growth of the oxide obeys the parabolic equation [14].8) Any factors that disturb the protective magnetite layer on steel. a reaction occurs between H2O and Fe. the specific damaging chemical factor of excess OH− concentration is discussed later. it is observed that rate of hydrogen evolution is momentarily high after a boiler is started up again. at a rate depending on temperature.. before the boiler is damaged. especially if vanadium-containing oils are used as fuel. in line with the mechanism depending on ion and electron migration through solid reaction products as described in Chapter 11. usually in a localized region.g. In this regard. hence. Hence. thereby exposing fresh metal. Differential contraction of the oxide and metal causes some degree of spalling of the oxide. indicates that Fe3O4 is formed only below about 570 °C (1060 °F) and that above this temperature. if necessary.BOILER-WATER TREATMENT 325 a worthwhile safety measure because the tendency toward cracking is more pronounced in the plastically deformed test piece than in any portion of the welded boiler. water treatment can be corrected. Accordingly.8 and 11.9. experiments show that Fe(OH)2 is the initial reaction product and not Fe3O4 [15]. resulting in a protective film of magnetite (Fe3O4) as follows: 3Fe + 4 H 2O → Fe3O4 + 4 H 2 (18. The mechanism of corrosion in this temperature range follows that described for anode and cathode interaction on the plane of the metal surface in contact with an electrolyte. Corrosion of boiler and superheater tubes is sometimes a problem on the hot combustion gas side. either chemically or mechanically. FeO forms instead. mechanical damage. induce a higher rate of reaction. may take place each time the boiler is cooled down. so far as it is understood. the rate is diffusion-controlled. The latter then decomposes on cooling to a mixture of magnetite and iron in accord with 4 FeO → Fe3O4 + Fe (18.7) Measurements of hydrogen accumulation in boilers as a function of time. as well as laboratory corrosion rate determinations. into magnetite and hydrogen in accord with a reaction first described by Schikorr (Schikorr reaction)[16] 3Fe(OH)2 → Fe3O4 + H 2 + 2 H 2O (18. with hydrogen production then falling to normal values . since modern boiler practice ensures removal of dissolved oxygen from the feedwater. At lower temperatures (e. room temperature to about 100 °C) and probably at higher temperatures before a relatively thick surface film develops. This matter is discussed in Sections 11. 1.1). Control of scale formation usually requires removing all calcium and magnesium salts by various means. Carbonic acid is corrosive to steel in the absence of dissolved oxygen and more so in its presence [18].3. 18. The latter type of failure may take place without any major loss of tube wall thickness. as described in the following paragraphs. For corrosion control. usually causing pitting of the boiler tubes and general attack elsewhere. including use of ion-exchange resins or adding substances to the water that favor precipitation of sludges rather than adherent continuous scales on the metal surface. or so-called plug-type oxidation occurs. concentrations of silica and silicates in feedwaters must also be reduced in order to minimize volatilization of SiO2 with steam. Removal of oxygen is accomplished by steam deaeration of the water. Furthermore. followed by addition of a scavenger.005 ppm O2) analyzable by chemical methods (e. leading eventually to stress rupture of the tube. which accentuates local temperature rise.. For high-pressure boilers. particularly if the water is acidified before the deaeration process to liberate carbonic acid from the dissolved carbonates. these types of damage are not observed [17]. Final oxygen concentration is usually held below values (<0. carbonates dissociate as follows: . such as sodium sulfite or hydrazine (see Section 18. Pitting usually results.g. Details are discussed in standard references on boiler-water treatment. followed by microfissuring along grain boundaries and eventual blowout of the affected tube.326 TREATMENT OF WATER AND STEAM SYSTEMS presumably after a reasonable thickness of oxide has again built up at damaged areas. If steam is used for power production. the basic treatment consists of removal of dissolved gases. Conditions of boiler operation that lead to metal oxide or inorganic deposits (from the boiler itself or from condenser leakage) on the water side of boiler tubes cause local overheating accompanied by additional precipitation of solutes from the water. however. which interfere with heat transfer. In the absence of deposits within boiler tubes. causing formation of damaging deposits on turbine blades. any remaining dissolved oxygen in the feedwater combines quantitatively with the metals of the boiler system.2 Boiler-Water Treatment for Corrosion Control Feedwaters of boilers are chemically treated both to reduce corrosion of the boiler and auxiliary equipment and to reduce formation of inorganic deposits on the boiler tubes (scaling). hydrogen resulting from the H2O–Fe corrosion reaction may enter the steel causing decarburization. but addition of alkali to boiler water limits any corrosion caused by carbon dioxide to the boiler itself by converting dissolved carbon dioxide to carbonates. and addition of inhibitors. Deaeration is accompanied by some reduction of carbon dioxide content. At prevailing boiler temperatures. Removal of Dissolved Oxygen and Carbon Dioxide. addition of alkali. the Winkler method). with the carbon dioxide again being available for further corrosion. Corrosion of iron by water at 310 °C (590 °F) at various values of pH measured at 25 °C [E. Since both steel and copper alloys are common in boiler systems. but attack of copper-base alloys is negligible in the absence of oxygen. . Steel return-line systems suffer serious corrosion. it accumulates increasingly in the boiler with each addition of feedwater unless an occasional blowdown (intentional release of some boiler water) is arranged.4 that excess alkali can be damaging to a boiler in that the corrosion rate increases rapidly as pH is increased above 13. Hall. Trans. Feedwater for a high-pressure boiler is treated to a minimum pH (measured at room temperature) of 8. 2. if the carbon dioxide content of boiler water is high. The copper-alloy condenser system also suffers corrosion should dissolved oxygen be present together with carbon dioxide.2. therefore. 597 (1939)].5. Eng. where it decomposes into Fe(OH)2 + CO2. Since carbon dioxide is not used up in the corrosion process. with the optimal pH range being 9. Soc. Soluble FeCO3 is formed. Partridge and R.2– 9. Am. Addition of Alkali.9) bringing hot carbonic acid into contact with the condenser and return-line systems. Mech. For copper alloys.5 to minimize corrosion of steel. It is apparent from Figure 18. The Figure 18.BOILER-WATER TREATMENT 327 Na 2CO3 + H 2O → CO2 + 2NaOH (18. 61. the preferred pH range is 8. Alkali addition to boiler waters is standard practice for most high-pressure boilers in the United States and abroad. and it returns with the condensate to the boiler.5–9.2 is recommended [19].8–9. a compromise range of 8.4. it is held advisable to add buffer ions. 3. When any of these is added to boiler water in sufficient amount. Such ions are also useful in avoiding similar high OH− concentrations that lead to stress-corrosion cracking of any portion of the boiler under high residual or applied stress. Structural formulas for three neutralizing-type amines.5–10. beneath a cracked oxide scale. neutralizers. or at a hot spot on a scaled tube surface.5). such as is formed between riveted plates.5. and dioctadecylamine are typical of the filming-type inhibitors. Operating by a different principle. whereas the other amines are actually. which protect against corrosion by building up a protective organic film on the condenser surface. There are two categories of volatile amines used for this purpose: (1) neutralizing amines and (2) filming amines.328 TREATMENT OF WATER AND STEAM SYSTEMS Figure 18. benzylamine. 18. hexadecylamine. it neutralizes carbonic acid and raises the pH of steam condensate to an alkaline value. Corrosion caused by dissolved carbon dioxide in steam condensate is minimized by adding a volatile amine to the boiler water. volatile octadecylamine. which limit the increase in pH that a water can achieve no matter how concentrated it becomes. and morpholine (Fig. For this reason. as mentioned previously. Advances in this area have been assisted by the development . The first can be minimized by addition of phosphates. The filming amines more nearly fit the definition of an inhibitor. The first group includes cyclohexylamine. stress-corrosion cracking and returnline corrosion. such as PO3− 4 (Na3PO4).0 was more effective in reducing corrosion of high-pressure boiler tubes under a variety of operating conditions than either NaOH or NH3 treatment. may reach OH− concentration levels above the safe range. Goldstein and Burton [17] reported that 5–10 ppm phosphate at pH 9. thereby making the condensate less corrosive. Addition of Inhibitors. for the most part. at a weld. 18. namely. It is possible to add inhibitors for controlling two kinds of corrosion in boiler systems.3 Mechanisms Considerable progress has been made in elucidating the mechanisms and causes of boiler corrosion. danger is not so much that the initial pH of the boiler water may be too high as that accidental concentration of an alkaline boiler water at a crevice.3. whereas the inner protective layer consists of densely packed crystallites 0. for example. Alkali Additions. There are many interacting factors— including feedwater composition. Oxygen is damaging primarily because it initiates pitting. Such corrosion is sufficient to return small amounts of copper salts to the boiler followed by deposition of metallic copper. which is a natural constituent of seawater. or a fundamental investigation of corrosion mechanisms. . The limiting safe concentration of dissolved O2 is reported to depend inversely on the Cl− concentration in the boiler water. dissolved neutral chlorides are not a corrosion problem in the absence of dissolved O2. Leakage will vary with design and age of the condenser system. possibly through action of differential aeration cells.2 μm in diameter that adhere tightly to the underlying metal. The optimum alkalinity of a boiler water depends. the question remains whether pitting of the boiler is initiated by copper contamination of boiler waters. are nevertheless effective in causing attack of the condenser system. accounting for higher corrosion rates. NH4OH is less effective than NaOH + Na3PO4 as a neutralizing agent in the quantities of each usually used in boiler-water treatment. In this regard. and boiler operation. The outer layer consists of loosely packed crystals >1 μm in diameter.05–0. Statistical analysis of many boilers. is required in order to obtain a final answer.BOILER-WATER TREATMENT 329 of experimental techniques for studies at high temperature and high pressure [20]. The effects of dissolved O2 are more complex. Potential–pH. seawater is used for cooling. diagrams have been developed for some systems at elevated temperature. on the extent to which contaminants slowly leak into the boiler from the cooling water of the condenser system (usually at the interface of condenser tubes and tubesheet). iron and iron alloys form a characteristic double oxide layer [22]. In other words. In solutions of too high or too low pH. and the presence of dissolved O2 is not a problem in sufficiently pure water [11]. in part. Fe3O4. varying from one boiler to another—that determine whether oxygen and copper contamination in specific instances are damaging. the protective layer of magnetite dissolves or is undermined. ensuing effects on the boiler depend on the composition of a specific cooling water. however. hydrolyzes to HCl and causes acid attack of the boiler. In high-temperature aqueous environments. condenser leakage. Whether a particular boiler is damaged by a given water treatment is not evidence in itself that the treatment. even if not damaging to the steel boiler directly. especially if CO2 or NH3 is also present in the condensate. is good or bad. Periodic sodium hydroxide additions to the boiler water neutralize the acid and prevent such attack [23]. Although the condensers may not be damaged appreciably by such corrosion. so that equilibria under specific potential and pH conditions may be more readily predicted [21]. in general. or Pourbaix. Removal of dissolved oxygen is generally agreed upon as a necessary step in all boiler-water treatment. magnesium chloride. with the layers in deoxygenated solutions consisting of magnetite. boiler design. Traces of O2. If. 2003. 730. 13A. H. Mech. on the other hand. Trans. corrosive concentrations of NaOH are usually built up by evaporation of alkaline boiler water at various crevices where liquid flow is impeded and heat transfer is restricted. P. 299 (1956). NH3 is a possible cause of stress-corrosion cracking of copper-base alloys in the condenser system. Callow. is damaging because it may slowly dissolve the magnetite film in accord with Fe3O4 + 4 NaOH → 2 NaFeO2 + Na 2 FeO2 + 2 H 2O (18. B. Corrosion inhibitors in the water treatment industry.4 shows. Oxygen Scavengers. phosphate additions avoid the high values of pH in concentrated boiler water that are necessary for reactions (18.10) and (18.330 TREATMENT OF WATER AND STEAM SYSTEMS Excess alkali.4 MPa (1800 psi) steam pressure. p. 78. Corrosion: Fundamentals. 3. Water Works Assoc. J. Testing. REFERENCES 1. (London) 1957. and Protection. The use of sodium sulfite as a scavenger in highpressure boilers is limited by the decomposition of sulfites at high temperatures into sulfides or SO2. D. Eng. as Fig. Soc. Chem. It is said to be satisfactory below 12. Appl. Wetton. Pirt. Gillett. 171 (1966).12) The use of hydrazine is not subject to similar objections. and W. Also. Ind. 39. 892. . 4. Gaunt and E. One possible path for decomposition is the following: 4 Na 2SO3 + 2 H 2O → 3Na 2SO4 + 2 NaOH + H 2S (18. concentrated alkali reacts directly and more rapidly with iron to form hydrogen and sodium ferroate Fe + 2 NaOH → Na 2 FeO2 + H 2 (18. In the absence of conditions that favor concentration of alkaline solutes. Pye. 5. In addition.10) forming sodium hypoferrite or sodium ferroate (Na2FeO2) and sodium ferrite (NaFeO2). Am. S. 18. Am. Furthermore. Boffardi. 1121 (1947). should oxygen accidentally contaminate the condensate. Chem. D. Vol. M. but its slow reaction with oxygen and partial decomposition into NH3 should be taken into account in calculating proper dosages.11). both of which are soluble in hot concentrated NaOH. OH. Baker and V. 2. 16. ASM International. Marcy. J. Materials Park. As mentioned earlier. damage by corrosion is expected to be minimal.11) Reactions of this kind account in part for pitting and grooving of boiler tubes and are those that account for the excessive corrosion rate of iron at high values of pH. in ASM Handbook. 2003. Park Ridge. Houston. 6 (1977). 105. New York 1980. West Conshohocken. Masterson. 748 (1942). Robitaille and J. Corrosion in fossil fuel power plants. Matsuda. Italy. Jones and R. Linnenbom. in High Temperature High Pressure Electrochemistry in Aqueous Solutions. 33 (1933). 891–906. 75 (1969). High Temperature High Pressure Electrochemistry in Aqueous Solutions. 1960. Allgm. Industrial Water Treatment Chemicals and Processes: Developments since 1978. pp. ASM International. The chemistry of steam-generator corrosion. Vol.. 7. Corros. H. Mech. Skaperdas and H. 1971. National Association of Corrosion Engineers. 83(27). TX.. P. Ind. 13A. in Proceedings of the Fourth International Congress on Metallic Corrosion. V. 12. p. de G. 1976. Electrochem. 49 (1981). J. Corrosion: Fundamentals. editors. TX. pp. ASTM International. in ASM Handbook. Fontana and R. 985–1010. P. Mater. MA. S. p. 13. Am. 1976. Materials Park. thesis. 10. Townsend. Chem. Collie. Ref. Goldstein and C. . 36–42. Engrs. New York. M. (London) 1969. p. Chem. 13. National Association of Corrosion Engineers. G. Schikorr. 893. Ulrich. pp. Soc. Case histories of corrosion in industrial boilers. 65 (1929). H. Trans. 22. P. H. 12. Staehle. TX. Università degli Studi di Ferrara. 308. Z. D. Perf. 20. Berlin. editors. Hoar. editor. Ind. p. 212. 20(2). Frey. NJ. Wiederholt. Corrosion. D. D. in Passivierende Filme und Deckschichten. Uhlig. J. D. 20(2).. 1972. See ASME Consensus on Operating Practices for the Control of Feedwater and Boiler Water Quality in Modern Boilers. Mann. 23. Springer. Noyes Data Corp. 9. Plenum Press. Garnsey. 5. G. 8. Elektrochem. GENERAL REFERENCES B. 1987. Burton. E. Vol. 1956. 49 (1981). Mater.T. Brit. ASM International. 322 (1958). de G. p. Boffardi. 17. 18. M. OH. Houston. Houston. Cambridge. 34. Standard Practice for Assessing the Tendency of Industrial Boiler Waters to Cause Embrittlement (USBM Embrittlement Detector Method). Perf. editors. Jones and R. Department of Metallurgy. Bilek. Soc. M. 16. 34. Corrosion inhibitors in the water treatment industry. PA. Staehle. 15. 477. Materials Park. 77 (1976). United Engineering Center. Anorg. in ASM Handbook. p. Mann. J.I. 177. G. 217. Anlaufschichten. Z. Dobson. K. Vol. 19. and G.GENER AL REFERENCES 331 6. Testing. 619. Fischer.D. ASTM D807-05. 14. Chemical Technology Review No. Sc. D. J. Hauffe. 1983. 11. Advances in Corrosion Science and Technology. OH. and Protection. and W. G. Series A 91. Chem. editors. in Third European Symposium on Corrosion Inhibitors. 35. 1979. Staehle. Combustion 52(2). National Association of Corrosion Engineers. Frey. 36 (1980). Eng. Castle. 7. Chem. T. Mann. 21. Eng. 1261. R. in ASM Handbook. OH. Combustion 52(2). G. 19 (1980). Latanision and D. Vol. Vol. Boiler water treatment. Gabrielli. Corrosion: Environments and Industries. pp. ASM International. Corrosion: Environments and Industries. OH. 2006. B. Materials Park. in ASM Handbook. Mitton. Materials Park. R. 2006. 229–235. Holcomb. Corrosion in supercritical water—Ultrasupercritical environments for power production. Corrosion in supercritical water—Waste destruction environments. ASM International. . 13C.332 TREATMENT OF WATER AND STEAM SYSTEMS G. 13C. Navitsky and F. 236–245. pp. M. and chemical cleaning of drum-type utility steam generators. feedwater treatment. R. Inc. The environment. In general.1 INTRODUCTION The development of alloys for controlling corrosion in specific aggressive environments is certainly one of the great metallurgical developments of the twentieth century. 333 . The alloy of specific chemical composition and metallurgical structure. The alloy/environment combination. by R. 2. 4. whether it is sufficiently aggressive to break down the protectiveness of the surface film.19 ALLOYING FOR CORROSION RESISTANCE. Winston Revie and Herbert H. the type and rate of corrosion that propagates after initiation has occurred. The basis upon which alloys resist corrosion and the causes of corrosion susceptibility of alloys are explored in this and subsequent chapters. if not. the corrosion behavior of alloys depends on the interaction of: 1. thereby initiating localized corrosion. 3.11). STAINLESS STEELS 19. in addition to the Corrosion and Corrosion Control. The beneficial effect of alloyed chromium or aluminum on the oxidation resistance of iron has already been described (Section 11. Uhlig Copyright © 2008 John Wiley & Sons. The film on the alloy surface. controlling whether the film selfrepairs after breakdown and. 10 at. copper no longer corrodes preferentially. A mechanism involving dissolution of the alloy and subsequent redeposition of gold is not likely because dissolved gold in any amount is not detected.7 64. the same reaction limits apply. X-ray examination of the residual depleted porous layer shows that interdiffusion of gold and copper takes place to form solid–solution alloys varying in composition from that of the original alloy to pure gold. hence. the alloy corrodes. Corresponding alloy compositions vary with the corrosive medium. or nickel on atmospheric resistance (Section 9. iron.% Au occurs by injection of divacancies at the corroding alloy surface that readily diffuse into the alloy at room temperature and tend to fill with copper atoms.2).0 50 Wt. Rapid quenching of certain alloys consisting of.5–75.7 .5 74.% Au for one or more weeks to nitric acid resulted in measurable attack [3]. Above the critical gold composition for parting.9 62.% Au 46. This behavior of noble-metal alloys is known as parting and is probably similar in mechanism to dezincification of copper–zinc alloys (see Section 20.% Au 32 49. At higher temperatures. Pt–Cu. whatever the environment.5 Ag–Au Alloys At. it is possible that conditions for divacancy formation are no longer favorable. in strong acids. STAINLESS STEELS beneficial effect of small alloying additions of copper. Alloys of gold with copper or silver retain the corrosion resistance of gold above a critical alloy concentration called the reaction limit by Tammann [1].g. For example. Below the reaction limit. for example.% of the noble-metal component (Table 19. Instead.2.5).334 ALLOYING FOR CORROSION RESISTANCE. TABLE 19. leaving a residue of pure gold either as a porous solid or as a powder.0–49.5–25. Pickering and Wagner [2] proposed that the predominant mechanism of parting in a gold–copper alloy containing. The latter diffuse in the opposite direction toward the surface.1). In other words. for example. for example. or nickel base and containing one or more metalloid additions. exposure of the gold–silver alloys containing more than 50 at.1.. or perhaps the vacancies tend to fill with a higher proportion of gold compared to copper atoms. where they enter into aqueous solution.5 75.3) H2SO4 24.5 50 50 49–50 Wt. chromium.% Au (NH4)2S2 H2CrO4 HNO3 (specific gravity 1. but reaction limits in general. Various other noble-metal alloys (e. cobalt.2 48.5 63.2–51.5 75. such as phosphorus. where it alone undergoes dissolution.8–63. Reaction Limits [1] Corrosive Medium Cu–Au Alloys At. with the exception that some attack may initiate above the reaction limit composition after long exposure times. and Pt–Ag) also exhibit reaction limits in HNO3. the evidence supports solid-state diffusion of copper from within the alloy to the surface.% Au 50. at 100°C. tend to fall between 25 and 50 at. Pt–Ni. similar to the passivity of pure chromium itself. The alloying element may reduce either icritical [e. alloys have been designed with glass-formation cooling rates as low as 1 K/s [5].14.8)].3).4 and 25. without defects. In principle. 6. As a result of their amorphous state. icritical is reduced to such small values that any small corrosion current exceeds icritical and accounts for initiation of passivity in aerated aqueous solutions. The noble metal stimulates the cathodic reaction and.g. Cooling from the liquid at rates of about 106 K degrees per second “freezes” the atoms in nearly the same positions that they occupied in the liquid [4].5% Cr. respectively. Alloying is an especially effective means for improving corrosion resistance if passivity results from the combination of a metal. For >12% Cr–Fe alloys. nickel alloyed with copper (Fig. intermetallic compounds) can induce passivity in multicomponent alloys by the same mechanism. second phases of any kind (e. allowing passivity to establish itself in less concentrated acid than is the case for pure iron. Some of the glassy-alloy compositions are especially resistant to acid media. or ipassive [e. The stainless steels are ferrous alloys that contain at least 10. so that the corrosion resistance may be much greater than that of crystalline alloys of the same composition. Any breakdown of this film is likely to result in localized corrosion.. and 12% Cr. which resist sulfuric acid at concentrations and temperatures otherwise extremely corrosive.STAINLESS STEELS 335 silicon.8. Cr–Ni. increases the anodic current density to the critical value for passivation. Carbon in steel can act similarly by creating cathodic sites of cementite (Fe3C) on which reduction of HNO3 (or HNO2 in HNO3) proceeds rapidly. the corrosion rate may be reduced by orders of magnitude. the critical composition is 30–40% Ni. in this way. otherwise active. 19. A wide range of stainless steels have been developed for applications in .2 STAINLESS STEELS Stainless steels depend on the integrity of a thin passive film to maintain low corrosion rates in aqueous solutions. and Cr–Fe alloys.1)]. Section 6. For Ni–Cu alloys. multiple phases. Stainless steels are passive in many aqueous media.g. carbon.11. Similarly. 14%. may be above or below 12%. Practical examples are palladium or platinum alloyed with stainless steels or with titanium (see Sections 6. More recently. at pits and crevices. these alloys do not have the long-range order that is characteristic of crystalline materials.g. FeCl3). for Cr–Co.. Instead. Chromium. Stress-corrosion cracking may also occur. as explained in Section 6. produces amorphous materials that are commonly referred to as glassy alloys. passivity usually occurs at and above a composition that is specific to each alloy and that depends on the environment.. and grain boundaries. if present.8. it is 8%. including HCl and oxidizing solutions (e. or boron.g. if passivity results from a noble metal added in small amount to an active metal or alloy.. chromium alloyed with iron (Fig. In homogeneous single-phase alloys. these materials are chemically and structurally homogeneous. Section 6. with a normally passive metal. 6. and they are the most important of the passive alloys. on the other hand. largely because the accompanying high carbon content impaired corrosion resistance.. His alloys were obtained by direct reduction of the mixed oxides. OH. 1949. in Stainless Steels. was apparently first described by Monnartz [9] of Germany. however. In France in 1821. STAINLESS STEELS many environments. producing what today is called ferrochromium (40–80% Cr). Berthier prepared some steels using ferrochromium as a component. Cleveland. *Based largely on the account by Carl Zapffe. was below that required for passivity. In 1904. The commercial utility of the Cr–Fe hardenable stainless steels as possible cutlery materials was recognized in about 1913 by H. England. The maximum chromium content. and W). This research included a description of the beneficial effect of oxidizing compared to reducing environments on corrosion resistance. apparently. Mo. pp. Faraday in England published a report [6] on the corrosion resistance of various iron alloys they had prepared. He studied the metallurgical structure and mechanical properties of the Cr–Fe alloys and also of the Cr–Fe–Ni alloys that are now called austenitic stainless steels. J. he noticed that the 12% Cr–Fe alloys did not etch with the usual nitric acid etching reagents and that they did not rust over long periods of exposure to the atmosphere. Maurer and B. it was this report in which the chromium–iron alloys were first mentioned. American Society for Metals. and the principles for successful use of these important alloys are described.1 Brief History* In 1820. who began his research in 1908 and published a detailed account of the chemical properties of Cr–Fe alloys in 1911. Ti. Although a variety of chromium–iron alloys was produced in subsequent years by several investigators who took advantage of the high-strength and highhardness properties imparted by chromium. the properties and behavior of stainless steels are reviewed. Berthier [7]. whose attention had been drawn to the work of Stodart and Faraday.2. Strauss of the Krupp Steel Works. were first exploited in Germany in 1912–1914 based on research of E. . The alloys were brittle and were high in carbon. In this section. 5–25. the inherent corrosion resistance of the alloys was not observed. Brearley of Sheffield.336 ALLOYING FOR CORROSION RESISTANCE. Stodart and M. V. but the chromium content again was too low to overlap the useful passive properties characteristic of the stainless steels. the necessity of maintaining low carbon contents. Guillet [8] of France produced low-carbon chromium alloys overlapping the passive composition range. initiating at a minimum of 12% Cr. The austenitic Cr–Fe–Ni stainless steels. and they narrowly missed discovering the stainless steels. But recognition of the outstanding property of passivity in such alloys. and they had no value as structural materials. Looking for a better gun-barrel lining. 19. found that iron alloyed with considerable chromium was more resistant to acids than was unalloyed iron.g. and the effects of small quantities of alloying elements (e. Martensite is produced by a shear-type phase transformation on cooling a steel rapidly (quenching) from the austenite region (face-centered cubic structure) of the phase diagram. Typical applications include cutlery. Uses include trim on autos and applications in synthetic nitric acid plants. Ferritic stainless steels are body-centered cubic in structure and are magnetic. as well as of the martensitiic stainless steels. Ferritic. the structure is body-centered cubic.2 Classes and Types There are five main classes of stainless steels. the AISI numbers remain in use today. phase for pure iron is the stable phase existing below 910 °C. existing as a stable or metastable structure depending on composition. They can be hardened moderately by cold working.STAINLESS STEELS 337 The 18% Cr. precipitation-hardenable. Austenitic. 1. and corrosion properties. C. the alloys are ferritic at all temperatures up to the melting point. A listing of many stainless steels that are produced commercially is given in Table 19. which. It is the characteristically hard component of quenched carbon steels. For low-carbon Cr−Fe alloys. above this Cr content. Although AISI stopped issuing designations for new stainless steels several decades ago. the hightemperature austenite (or gamma) phase exists only up to 12% Cr. In stainless steels. Ferritic steels are named after the analogous ferrite phase. magnetic. and tools. and duplex. exists as a stable structure between 910 °C and 1400 °C. 3. phase. Each class consists of several alloys of somewhat differing composition having related physical. designated in accord with their crystallographic structure.2. or γ. 2. austenitic. These were given type numbers by the American Iron and Steel Institute (AISI). for pure iron. Martensitic. 8% Ni (18–8) austenitic stainless steel is the most popular of all the stainless steels now produced. Co. Alloyed nickel is largely responsible for the retention of austenite on quenching the commercial Cr–Fe–Ni alloys from high temperatures. and it is readily deformed. The five main classes of stainless steels are martensitic. or relatively pure-iron component of carbon steels cooled slowly from the austenite region. The Unified Numbering System (UNS) was introduced in the 1970s to include all alloys. It is the major or only phase of austenitic stainless steels at room temperature. and the alloys are magnetic.2. The name of the first class derives from the analogous martensite phase in carbon steels. or alpha. steam turbine blades. and N also contribute . This phase is face-centered cubic and nonmagnetic. with increasing nickel content accompanying increased stability of the retained austenite. but not by heat treatment. Austenitic stainless steels are named after the austenite. Alloyed Mn. ferritic. including stainless steels. 19. The ferrite. 12 max 0.2.20 max 0.5 — 440B S44003 16. Magnetic.0 11.0 1.0 Turbine quality 403 410 414 416 S40300 S41000 S41400 S41600 11.04 0.5 1.0 1.0 16.15 0. Types and Compositions of Stainless Steels Composition (%) C Mn (max) P (max) S (max) Si Other Elements Remarks 338 AISI Type or Common Name UNS No.04 0.15 max 0.TABLE 19.15 min 0.5 — 416Se S41623 12.0 1.25 1. nonseizing Easy machining.5–13.0 1.0 1.6–0.04 0.03 0.0 1. Magnetic.15 Se min 405 430 430F S40500 S43000 S43020 11.04 0.0 — 0. STAINLESS STEELS 430FSe S43023 16. 0.04 0.15 Se min 0.04 0.25 0.06 1.0 — 440C S44004 16.06 0.0 1.75 Mo max Highest attainable hardness Class: Ferritic—Body-Centered Cubic.0 1.0–17.15 min 0.5–13. nonseizing .04 0.15 max 0.0 16.1–0.0 15.0–14.03 0.6 Mo max 0.5 11.03 0.0–18.06 1.5 16.15 max 0.0 — — 1.75 0.0–18.75–0.0 0.0–14.0 1.0–18.2 1.0 1.0 0.06 0.25 0.0 — 1.95–1.03 1.0–18.15 max Over 0.03 0.03 1.03 0.06 0.04 0.6 Mo max 0.0–14. nonseizing 420 431 440A S42000 S43100 S44002 12.5 12.5–14.75 Mo max 0.03 0.25–2.0–18. Cr Ni Class: Martensitic—Body-Centered Tetragonal.0–18. Heat-Treatable 0.75 Mo max 0.0 0.5–13.0 — Easy machining.0 1.03 1.0 — — — ALLOYING FOR CORROSION RESISTANCE.12 max 1.03 0.08 max 0.0 1.12 max 0. nonseizing Easy machining.0 1. not Heat-Treatable 0.25 0.25–2.15 max 0.0 1.3 Al.04 0.95 0.0 1.0 0.0 1.04 0.0 1.06 0.0 — Easy machining.0 1. Composition (%) Cr Ni 446 S44600 23.045 0.0 12.0–6.0 1.15 max 0.0 2.03 1.15 max 5.25 max 0.045 0.0 17.03 0.0 Resistant to hightemperature oxidation Easy machining.0–10. nonseizing Easy machining.0–14.0 22.0 10.0 Mo Ti: 5 × C min .0–10.03 max 0.0 2.0–22.0–14.03 0.03 0.0–26.5 1. Nonmagnetic.25 max 0.0 22.0–12.08 max 0.04 0.045 0.03 0.5 4.0 8.03 0.0 24.0–18.06 0.0 9.0–3.03 max 0.045 0.0 19.0 10.045 0.0–10.0 308 309 309 S 310 310 S 314 316 316L 317 321 S30800 S30900 S30908 S31000 S31008 S31400 S31600 S31603 S31700 S32100 19.0 1.0 1.03 1.08 max 0.20 max 0.0 19.15 max 0. not Heat-Treatable 0.0–19.0 1.20 0.03 1.0 2.08 max 2.20 0.045 0.0–22.25 N max 0.0 8.0–19.0–10.08 max 0.0–18.0 2.5–3.0–15.0 16.0–19.0–4.045 0.0 2.0 10.0 17.0 2.0 2.15 min 1.15 max 0.0 1.0 2.0 12.0–15.0–20.03 0.5 1.08 max 0.045 0.0 3.03 0.08 max 0.0 0.0–24.20 max 1.0 1.0 0.0 8.0–19.0 2.03 0.03 0.0 0.0 0.0 24.5–10 2.0 1.0 23.5 7.0 1.0–26.0 0.045 0.0 17.06 1.0–19.0 S30300 17.0 8.03 0.045 0.5 8.0 2.045 0.0–20.0 2.0–10.0 1.0 1.045 0.0–21.0 0.0 11.045 0.0 18.03 1.045 0.15 Se min 303 Se S30323 17.0–20.0–3.0–3.0–24.0–18.0 10.25 N max 201 202 301 302 302B S20100 S20200 S30100 S30200 S30215 16.0 2.03 0.6 Mo max 0.0 1.0 Mo 2.08 max 0.03 0.15 max 0.0 0.0 8.0–8.0 18.0 2.15 max 0.0 17.0–27.0 1.0 19.045 0.0 16.0 2.AISI Type or Common Name C 0.03 0.0 17.045 0.0–19.5–5.0–19.0 0.0–12.15 max 2.0 Extra-low carbon 339 Stabilized grade 2.12 max 2.0–26.0 2.0 0.0 303 0.03 0.03 0.0 0.25 N max Resistant to hightemperature oxidation Mn (max) P (max) S (max) Si Other Elements Remarks UNS No.5 0.0–22.0–18.5–7.0 16. nonseizing Extra-low carbon Lower rate of work hardening than 302 or 304 304 304L 305 S30400 S30403 S30500 18.06 0.5–13.0 6.0 — STAINLESS STEELS Class: Austenitic—Face-Centered Cubic.0–15.0–12.0 Mo 3. 04 1.03 1.20 Co max Class: Precipitation-Hardenable 0.5–3.2 N 0.0 2.0 0.2 N 3.5 3.04 2.0 6.0–19.01 1.15–0.03 max 2.00 15–5 PH S15500 14.0–8.0 0.75–1.0 Stabilized grade Stabilized grade 0.50–1.05–0.5–4. Continued 340 Composition (%) C 0.00–15.50 Al Class: Daplex—Austenitic/Ferritic 0.0 Cb + Ta: 10 × C min Cb + Ta: 10 × C min 0.0–19.5 0.0–4.0 348 S34800 17.20–0.0 9.04 1.6 Cu 0.0 2.5 ALLOYING FOR CORROSION RESISTANCE.03 max 0.2.5 Cu 0.0–23.TABLE 19.5 Mo 0. Cr Ni 347 S34700 17.1 Ta max when radiation conditions require low Ta Mn (max) P (max) S (max) Si Other Elements Remarks AISI Type or Common Name UNS No.5 2304 S32304 21.75 0.45 Nb 0.07 1.5–6.5–24. STAINLESS STEELS Zeron 100 S32760 24.0–18.0–5.00 2.09 1.0–13.05–0.03 1.00 Cu 0.00 W 0.0 9.08 max 2.0 0.05–0.0 6.0 4.0–13.0–26.014–0.50–1.00 0.08 max 0.045 0.0 2205 S32205 21.30 N .03 1.02 0.0 1.0 0.6 Mo 0.04 0.045 0.03 max 1.03 0.50–7.00 0.0 0.0 Mo 0.5 3.03 0.50 17–7 PH S17700 16.04 0.50–5. and various structural units for the food and chemical industries.g. the descriptor. on the other hand. 347. consisting of a continuous ferrite matrix with austenite islands (Fig. using low-temperature heat treatment [e. when hardening of the alloy is best done after machining operations.g.g. Alloys 305. 310. the highest resistance to uniform corrosion is obtained with the nickel-bearing austenitic types. duplex stainless steels are used in many applications.g. These Cr–Fe alloys contain less nickel (or none) than is necessary to stabilize the austenite phase.2) transform in part to the ferrite phase (hence. such as aluminum or copper. Precipitation Hardenable. 303. but not by heat treatment. see Table 19. 304.STAINLESS STEELS 341 to the retention and stability of the austenite phase. 302B. Zeron 100) have been developed [11]. “metastable”) having a bond-centered-cubic structure that is magnetic. and... is achieved by including in the composition a balance of elements that stabilize austenite (e. Duplex. Austenitic stainless steels can be hardened by cold working. chromium and molybdenum). architectural and automobile trim.g.1 [10]). good erosion and wear resistance. including pressure vessels. storage tanks (e. and 309 S.. 314) are essentially stable austenitic alloys and do not transform to ferrite or become magnetic when deformed. 302. On cold working. 316. The mixed austenite– ferrite microstructure. 310 S. and low thermal expansion. duplex stainless steels of higher molybdenum content (e. 348. high resistance to chloride stress-corrosion cracking. 316L.. the metastable alloys (e.. They are applied whenever the improved corrosion resistance imparted by alloyed nickel is desirable. but not otherwise. and heat exchangers [10]. that produce high hardness through formation and precipitation of intermetallic compounds along slip planes or grain boundaries. work harden at a relatively low rate and become only slightly magnetic. For . The 22% chromium duplex stainless steel was developed in the late 1970s and was later modified and designated as Alloy 2205 (UNS designation S32205).. 301. 304L. Uses of austenitic stainless steels include general-purpose applications. if at all. The precipitation-hardening stainless steels achieve high strength and hardness through low-temperature heat treatment after quenching from high temperatures. 5. 201. For seawater service. 4. 308. in addition. 303Se. 19. the highest nickel-composition alloys in this class are more resistant than the lowest nickel compositions. more important. for phosphoric acid). 202. 309.. Alloys containing higher chromium and nickel (e. Because of good mechanical properties. 480°C (900°F)] rather than a high-temperature quench as is required in the case of the martensitic stainless steels. they contain alloying elements. This transformation also accounts for a marked rate of work hardening. typically with the austenite:ferrite ratio ∼ 60 : 40. The duplex stainless steels contain both austenite and ferrite.g. and. In general. in general.g. nickel and nitrogen) and those that stabilize ferrite (e. or. 321. 342 ALLOYING FOR CORROSION RESISTANCE.2—for example. *For additional details on heat treating stainless steels. Typical analyses include 26% Cr. locally or generally.1. . Materials Park. provided they retain passivity.025% N. their corrosion rates tend to be higher than those of the active nickel-containing austenitic stainless steels of equivalent chromium and molybdenum contents [12]. in NaCl solutions. have optimum corrosion resistance as quenched from the austenite region. they are very hard and brittle. 316L. STAINLESS STEELS Figure 19. Alloyed titanium or columbium is sometimes included in the first two alloys in order to raise the tolerable carbon and nitrogen contents. pp.02% N.) optimum corrosion resistance. Ductility is improved by annealing (650–750°C for 403. 28% Cr. stainless steels listed in Table 19. In this state. 4% Mo. <0. 410. Heat Treating. see ASM Handbook. OH. 4. 1% Mo.01% C. <0. and crevice corrosion.01% C. 317) have improved corrosion resistance to chloride-containing environments. 2% Mo. 1991. <0. <0. Their corrosion properties are generally better than those of the conventional ferritic. 416.* Recently developed ferritic stainless steels of controlled purity contain molybdenum additions and low carbon and nitrogen. HNO3. and various organic acids. Should they lose passivity. 769−792.025% C. <0. (Reprinted with permission of John Wiley & Sons. The ferritic stainless steels have optimum corrosion resistance when cooled slowly from above 925°C (1700°F) or when annealed at 650−815°C (1200−1500°F). Typical microstructure of a duplex stainless steel. <0. austenitic alloys must be quenched (rapidly cooled by water or by an air blast) from about 1050–1100°C (1920–2000°F). The martensitic stainless steels. dilute nonoxidizing acids. and some austenitic. The molybdenum-containing austenitic alloys (316. Vol.02% N. ASM International. on the other hand. Inc. The grains are elongated in the rolling direction (∼300×) [10]. All these alloys can be quenched from above 925 °C without loss of corrosion resistance. 18% Cr. B. become susceptible to hydrogen cracking. The degree of *After transformation.2. . the tempering range 450–650°C (840–1200°F) should be avoided if possible.3 Intergranular Corrosion Improper heat treatment of ferritic or austenitic stainless steels causes the grain boundaries. from the transformation of martensite containing interstitial carbon into a network of chromium carbides attended by depletion of chromium in the adjoining metallic phase. otherwise resistant. composition differences are established between the two phases setting up galvanic cells that accelerate corrosion. quenched austenitic stainless steel (face-centered cubic) of 18% Cr. Composition gradients. in part.2. also. 19. if a similar alloy containing a mixture of austenite and ferrite is briefly heated at. the sensitizing temperature range is approximately 425–875°C (800–1610°F). C. In general.06% C [13]. Phase changes brought about by cold working the metastable austenitic alloys are not accompanied by major changes in corrosion resistance. However.3% C.STAINLESS STEELS 343 416Se. 680–750°C for 440 A. of the same chromium and nickel composition but with lower carbon and nitrogen content [14].5–3% Ni [15]. 8% Ni composition has approximately the same corrosion resistance as quenched ferritic stainless steel (body-centered cubic). 19. 600°C. stainless steels containing less than this amount of nickel are sensitized in the temperature range typical of nickel-free ferritic steels. the alloys.1 Austenitic Stainless Steels. the transition in sensitizing temperatures for steels containing 18% Cr occurs at about 2. for example.3.* Furthermore. and 420). Corrosion of this kind leads to catastrophic reduction in mechanical strength. Cold working any of the stainless steels usually has only minor effect on uniform corrosion resistance if temperatures are avoided that are high enough to permit appreciable diffusion during or after deformation. are more important in establishing uniform corrosion behavior than are structural variations of an otherwise homogeneous alloy. For austenitic alloys. 650–730°C for 414. 620–700°C for 431.2–0. to become especially susceptible to corrosion. as well as 650°C for a similar steel containing 0. whatever their cause. Resistance to pitting and rusting in 3% NaCl at room temperature reaches a minimum after tempering a 0. They are very much different for the ferritic and the austenitic stainless steels. In this respect. whereas those containing more nickel respond to the temperature range typical of the austenitic stainless steels. This statement is true for metals and alloys in general. The specific temperatures and times that induce susceptibility to intergranular corrosion are called sensitizing heat treatments. 13% Cr stainless steel at 500°C. ferritic and martensitic stainless steels on cold working tend to become more susceptible to hydrogen cracking. Decrease of corrosion resistance probably results. which separate the individual crystals. but at some sacrifice of corrosion resistance. Slow cooling through the sensitizing temperature range or prolonged welding operations induce susceptibility. with the alloy corroding along grain boundaries at a rate depending on severity of the environment and the extent of sensitization. on the other hand. (Reprinted with permission of ASM International®. whereas alloying additions of molybdenum increase the time [16]. with a few minutes at the higher temperature range of 750°C (1380°F) being equivalent to several hours at a lower (or still higher) temperature range (Fig. on exposure to a corrosive environment. but rapid cooling avoids damage.2) [16. 0.06% C is affected less. whereas a similar alloy containing 0. STAINLESS STEELS damage to commercial alloys caused by heating in this range depends on time.asminternational. 19. 0. In seawater. 19. Sensitizing temperatures are reached some millimeters away from the weld metal itself. Arc welding. in which the metal is rapidly heated by a momentary electric current followed by a naturally rapid cooling. 11. failure of an austenitic stainless steel weld—called weld decay— occurs in zones slightly away from the weld rather than at the weld itself (see Fig. The extent of sensitization for a given temperature and time depends very much on carbon content.0% Ni. Damage occurs only on exposure to a corrosive environment. and for 0. can cause damage. especially when heavy-gage material is involved. with the latter being at the melting point or above. www. austenitic stainless steels must be quenched from high temperatures.05% C. The higher the nickel content of the alloy. Spot welding. with the effect being greater the longer the heating time.2% Cr. a sensitized stainless-steel sheet may fail within weeks or Figure 19. the shorter the time for sensitization to occur at a given temperature. Hence.03% C the alloy heat treated similarly may suffer no appreciable damage on exposure to a moderately corrosive environment. Effect of time and temperature on sensitization of 18.3). the alloys become slightly stronger and slightly less ductile.org. Hence.) . Because precipitation of carbides accompanies sensitization. does not cause sensitization. All rights reserved. The physical properties of stainless steels after sensitization do not change greatly. 17].1% C or more may be severely sensitized after heating for 5 min at 600°C. An 18–8 stainless steel containing 0.344 ALLOYING FOR CORROSION RESISTANCE.2.05% N stainless steel [16]. 3.3. normally present in commercial alloys to the extent of a few hundredths percent. This reaction depletes the adjoining alloy of chromium to the extent that the grain-boundary material may contain less than the 12% Cr necessary for passivity. is less effective than carbon in causing damage (Fig.STAINLESS STEELS 345 Intergranular corrosion Weld Figure 19. but within (or somewhat above) the sensitizing temperature range it rapidly diffuses to the grain boundaries where it combines preferentially with chromium to form chromium carbides (e.2. Since intergranular corrosion of austenitic stainless steels is associated with carbon content. If the alloy is rapidly cooled through the sensitizing zone.02% C) is relatively immune to this type of corrosion [18]. At high temperatures [e. After sensitization. Nitrogen. months.2 Theory and Remedies.g.. 19. specimen was exposed to 25% HNO3. On the other hand. with the grains constituting large cathodic areas relative to small anodic areas at the grain boundaries.4) [19]. carbon is almost completely dispersed throughout the alloy. Example of weld decay (2×). 19. a low-carbon alloy (<0. failure occurs within hours. if the . M23C6. in which M represents the presence of some small amount of iron along with chromium). Electrochemical action results in rapid attack along the grain boundaries and deep penetration of the corrosive medium into the interior of the metal. The affected volume of alloy normally extends some small distance into the grains on either side of the boundary itself. which is used as an accelerated test medium..g. but in a boiling solution containing CuSO4·5H2O (13 g/liter) and H2SO4 (47 mL concentrated acid/liter). carbon does not have time either to reach the grain boundaries or to react with chromium if some carbon is already concentrated at the grain boundaries. 1050°C (1920 °F)]. causing apparent grain-boundary broadening of the etched surface. The chromium-depleted alloy sets up passive–active cells of appreciable potential difference. showed no appreciable depletion in nickel and an increase in iron content. perhaps in part because nitrides precipitate more generally throughout the grains or they form islands along grain boundaries. All specimens sensitized at stated temperatures for 217 h [19]. Microprobe scans of sensitized stainless steels have indicated chromium depletion and nickel enrichment at grain boundaries [21].4% and 83%.7% Cr in the alloy that had corroded from grain-boundary regions. In this respect. respectively. Carbides.22% C stainless steel for 10 days. on the other hand. interrupting a continuous path along which the corrosive agent can proceed [19]. using cold concentrated sulfurous acid acting on sensitized 18% Cr.12% C stainless steel has demon- . Intergranular corrosion measured by change in electrical resistance of 18–8 stainless steels containing nitrogen or carbon immersed in 10% CuSO4 + 10% H2SO4. For example. Copyright 1945. corrosion products of grain-boundary alloy obtained by choosing corrosive conditions that avoid attack of carbides show a lower chromium content than corresponds to the alloy. Schafmeister [20]. The Electrochemical Society. it is less effective than carbon in causing damage. form continuous paths of chromiumdepleted alloy.8% Ni. chromium again diffuses into the depleted zones. carbides isolated from grain boundaries of sensitized stainless steels have shown an expected high chromium content. reestablishing passivity of the grain boundaries and eliminating susceptibility to preferred attack. 0.346 ALLOYING FOR CORROSION RESISTANCE. Other experimental data can be cited that supports the chromium-depletion mechanism of intergranular corrosion of austenitic stainless steels. STAINLESS STEELS Figure 19. (Reproduced with permission.) alloy remains within the sensitizing temperature zone an especially long time (usually several thousand hours).8% Ni. Along the same lines. Accompanying analyses for Ni and Fe of 8. 12. Although nitrogen forms chromium nitrides.4. 0. Radioactive C14 introduced into an austenitic 18% Cr. 8. found only 8. after welding operations. they can be welded or otherwise heated within the sensitizing zone without marked susceptibility to intergranular corrosion. At high temperature. 347. If. Except as noted below.g.STAINLESS STEELS 347 strated enrichment of carbon at grain boundaries observable immediately after quenching from 1000 °C to 1350 °C.. the weld rod usually contains niobium rather than titanium. Niobium. It is not always a possible treatment. These alloys can be welded or otherwise heated in the sensitizing temperature range with much less resultant susceptibility to intergranular corrosion. Reduction of Carbon Content.g. precipitated carbides are dissolved. carbon is restrained from diffusing to the grain boundaries. they are not immune...g. There are at least three effective means for avoiding susceptibility to intergranular corrosion: 1. Ti or Nb) having higher affinity for carbon than does chromium. 2.03% C) are designated by letter L—for example. In such a situation. . during welding the stabilized grades of stainless steel. 3. Carbon content can be reduced in the commercial production of stainless steels. This heat treatment converts available carbon to stable carbides of titanium or niobium in a temperature range of lower carbon solubility compared to that of the usual higher quench temperatures. in part. This treatment is recommended. is lost by oxidation to a lesser extent. however. 22]. Heat Treatment at 1050−1100°C (1920−2000°F) Followed by Quenching. the region of the base metal adjacent to the weld is heated to temperatures at which the carbides of titanium and niobium dissolve [above about 1230°C (2250°F)] and the cooling rate is sufficiently rapid to prevent the formation of the stabilizing carbides. on the other hand. Addition of Titanium or Niobium (Columbium). because of the size of the welded structure or because of a tendency of the alloy to warp at high temperatures. By alloying austenitic alloys with a small amount of an element (e. on the reaction rate of chromium with carbon already at grain boundaries and not alone on the time required for additional carbon to diffuse to grain boundaries. Types 321. 348). a narrow region immediately adjacent to the weld is susceptible to intergranular corrosion. Optimum resistance to intergranular corrosion at 675°C (1250°F) is obtained by a prior carbide stabilizing heat treatment for several hours at about 900°C (1650°F) [17. However. Alloys of low carbon content (e. then these steels can be sensitized if they are subsequently heated in the temperature range in which chromium carbides precipitate. Alloys of this kind are called stabilized grades (e. for example. This indicates that time for sensitization may depend. the latter tends to oxidize at elevated temperatures with the danger of its residual concentration becoming too low to stabilize the weld alloy against corrosion. and rapid cooling prevents their re-formation. <0. and any which is already at the boundary reacts with titanium or columbium instead of with chromium. but at extra cost. Types 304L and 316L. In welding operations. 430. Silicon causes increased intergranular attack in the intermediate range of 0. High-purity alloys are immune. and 446 stainless steels. When polarized anodically in 1 N H2SO4 or in 27% HNO3 at 40°C. being more than 10 times higher for a 78% Ni. are necessary additions to the boiling nitric acid. 20% Ni. intermetallic compound. Grain-boundary attack of 19% Cr.g. 9% Ni stainless steel is most rapid after quenching from 1100°C to 1200°C. if it occurs. STAINLESS STEELS which. it is less pronounced on quenching from either 900°C or 1400°C [26].1−2% for the 14% Cr. but in these alloys the phase forms at a rate so low that it is important only when the alloys are used in the temperature range in which sigma phase is formed [23]..5 N K2Cr2O7). or manganese added in small concentrations have no effect.348 ALLOYING FOR CORROSION RESISTANCE. The necessity of strongly oxidizing conditions and presence of phosphorus for intergranular attack was confirmed for a quenched low carbon 20% Cr. intergranular attack occurred only at very noble potentials in the transpassive region. whereas silicon and phosphorus (>100 ppm) are damaging. Sigma phase may also form in Types 310. 0. Alloyed carbon. hence. which results in preferential direct attack on the sigma phase particles in preference to the molybdenum-depleted zones surrounding them [23]. Fe alloy compared to a similar 10% Ni alloy (70-h test). quenched from 1050°C.2. nitrogen. bal. Stress is not necessary.1–0. This type of intergranular corrosion can be prevented by selecting the welding process parameters to avoid temperatures at which carbides of titanium and niobium dissolve and by using a carbidestabilizing heat treatment as described above.002% P in the alloy. The rate of attack increases with the amount of alloyed nickel [25]. austenitic stainless steels. Similar intergranular attack of 15% Cr. such attack was not observed at any potential. NaOH. 17% Cr. 14% Ni stainless steel. Sigma phase is a hard. 19. Mn7+. In strongly oxidizing media (e. oxygen.3 Intergranular Corrosion of Nonsensitized Alloys. This trend is opposite to the beneficial effect of nickel on the resistance of stainless steels to stress-corrosion cracking. with a tetragonal structure that can precipitate at grain boundaries in Type 316 stainless steel when heated in the range 540–1000 °C (1000–1800°F). Sigma phase impairs corrosion resistance in highly oxidizing nitric acid environments. or Ce4+. and the presence of . oxidizing ions. The mechanism of corrosion may be based on the high molybdenum content of sigma phase. boiling 5N HNO3 + Cr6+). in larger or smaller amounts the alloy is not susceptible [27]. The attack occurs only in the transpassive region. such as Cr6+ (0. brittle. With only 0.3. bal. Ni (Inconel 600) in hightemperature water (350°C) or steam (600–650°C) [25] or of stabilized 18–8 stainless steel in potassium hydroxide solutions (pH 11) at 280°C [29] has been reported. including stabilized grades. undergo slow intergranular attack [24]. or dissolved lead (from heat-transfer tube–tube sheet leakage)].1% P stainless steel [28]. Contaminants in the water [such as traces of dissolved O2. 6% Fe. is called knife-line attack. This matter is of prime interest in view of the common use of Inconel 600 and stainless steels in the construction of nuclear reactors for power production. rich in chromium and molybdenum. the high-temperature 0. A type 430 ferritic stainless steel containing only 0. The sensitizing range for ferritic stainless steels lies above 925°C (>1700°F).01% FeCl3 at 340°C. as described previously. minimize the observed intergranular corrosion of welds exposed to boiling 65% HNO3. Increased nickel content favors intergranular corrosion [32] of 18–20% Cr stainless steels at 200–300°C in water containing Cl− and dissolved oxygen. The overall results confirm that intergranular attack is the result of specific impurities in the alloy segregating at grain-boundary regions.STAINLESS STEELS 349 crevices at which superheating occurs accompanied by concentration of solute accelerate S. Unlike the latter medium. whether high or low (16–28% Cr). The same accelerating media (i. behaved similarly.3%) and phosphorus (>0. Laboratory tests indicate that such failures also occur in longer times in relatively pure water [31] (see also Section 23. however.05% C. Chromium content of ferritic steels. The extent of their damaging effect depends on the chemical environment to which the alloy is exposed.3). These temperatures are the opposite of those applying to austenitic stainless steels. 14% Ni stainless steel exposed to 0. This behavior may be explained by the observed marked reactivity of titanium carbides. Niobium additions.009%C was still susceptible [34].C. whereas in austenitic steels the damage by weld decay localizes some small distance away from the weld. analogous to results in HNO3 + Cr6+. 19. but not to boiling 65% HNO3 [34]. Only on decarburizing lowcarbon steels containing 16% or 24% Cr in H2 at 1300°C for 100 h was immunity obtained in the CuSO4–H2SO4 test solution [35]. boiling CuSO4–H2SO4 or 65% HNO3) produce intergranular corrosion in either class. however. but not niobium carbides. In welded sections. a low-carbon (∼0. [30].002%) 25% Cr–Fe alloy was reported to be immune [36]. . it was stated.01% FeCl3 solution (21-day exposure) caused intergranular corrosion [33] of the nonsensitized stainless steel when it contained >0. hence. the observed damage in such instances is better described as intergranular corrosion rather than as stresscorrosion cracking. and immunity is restored by heating for a short time (approximately 10–60 min) at 650–815°C (1200–1500°F). damage to ferritic steels occurs to metal immediately adjacent to the weld and to the weld metal itself.e. but not titanium additions. Silicon (>0..3.4 Ferritic Stainless Steels. cause intergranular corrosion. has no appreciable influence on susceptibility to intergranular corrosion [34]. Similarly. Addition of titanium in the amount of eight times carbon content or more provided immunity to the CuSO4 test solution. Similar to the situation for austenitic steels.023%) in 14% Cr.2. lowering the carbon content is helpful.3. was effective in conferring immunity to nitric acid. as it does in boiling HNO3–Cr6+. It was reported [37]. with HNO3 along grain boundaries where such carbides are concentrated [38]. stress is not always necessary. and the extent and rapidity of damage are similar.C. Although applied stress has been stated to increase the observed attack in various media. but the critical carbon content is very much lower. that columbium additions (8× C + N content). and only heat treatment. and by their titanium or niobium content which. thereby restoring the grain-boundary alloy to normal stainless-steel composition. if present. some more effective than others. in solution at higher temperatures. Intergranular corrosion of ferritic stainless steels is best explained on the basis of the chromium-depletion theory that is generally accepted for the austenitic stainless steels [39].5 years. STAINLESS STEELS The Mo–Cr stainless steels of controlled purity. Cu2+. Susceptibility to pitting and crevice corrosion is greater in the martensitic and ferritic steels than in the austenitic steels. reacts preferentially with carbon and nitrogen. 317) are still more resistant to seawater. In the absence of passivity. This holds even though.01% C. But accompanying high diffusion rates of chromium account for restored immunity to intergranular corrosion on heating the alloy for several minutes in the range 650–815°C (rather than weeks or months required to restore sensitized austenitic stainless steels). This is accounted for by their molybdenum content.2. which slows down diffusion of carbon and nitrogen. all the stainless steels tend to corrode at specific areas and to form deep pits. develop visible pits within hours. For example. and chromium in the ferritic body-centered-cubic lattice compared to the austenitic face-centered-cubic lattice. chromium carbides and nitrides. act as pitting inhibitors when added to chloride solutions.. nitrogen. 316L. general corrosion may be appreciable.6. nonoxidizing metal chlorides (e.4 Pitting and Crevice Corrosion In environments containing appreciable concentrations of Cl− or Br−. pitting does not occur. Many extraneous anions. Ions such as thiosulfate 2− ( S2O3 ) may also induce pitting. depleting the adjoining alloy of chromium to resultant compositions below stainless-steel requirements and which corrode at correspondingly higher rates than the grains. although in some instances containing more than 0. addition of 3% NaNO3 to a 10% FeCl3 solution completely inhibited pitting . such as in deaerated alkali-metal chlorides. The austenitic 18–8 alloys containing molybdenum (types 316. These solutions have sometimes been used as accelerated test media to assess pitting susceptibility. Differences in sensitization temperatures and times compared to the austenitic steels are explained by the much higher diffusion rates of carbon. Stainless steels exposed to seawater develop deep pits within a matter of months. Accordingly. crevice corrosion and pitting of these alloys eventually develop within a period of 1–2. Stainless steels exposed at room temperature to chloride solutions containing active depolarizing ions. are immune to intergranular corrosion. precipitate rapidly (within seconds) below 950°C along grain boundaries.g. or Hg2+. in acid environments. in which stainless steels otherwise remain essentially passive.350 ALLOYING FOR CORROSION RESISTANCE. described earlier. however. it decreases in the latter alloys as the nickel content increases. with the pits usually initiating at crevices or other areas of stagnant electrolyte (crevice corrosion). or oxidizing metal chlorides at low pH. 19. such as Fe3+. SnCl2 or NiCl2). as mentioned in Section 6. H. in which the integrity of the passive film is breeched at localized areas and pits form.E. which partially shield the surface from access to oxygen. 2. In aerated 4% NaCl solution at 90°C.41 V) are too active. In addition to the critical pitting potential (CPP) (noble to which pits initiate). hence. Increasing amounts of alloyed chromium. with their effectiveness in this regard decreasing in the order − − OH − > NO3 > SO2− > ClO4 . because of the increasing acidity inside the pit.8 V) or the ferric–ferrous potential (φ° = 0. at a rate that is often found to increase with time.1 Theory of Pitting. Initiation. reducing any tendency for localized corrosion. addition of 8 g NaOH per liter was found to eliminate pitting [40]. as discussed earlier (Section 6. Similarly. pitting in 3% NaNO3 + 10% FeCl3 is not observed. stable pitting does not take place at any potential. or for other reasons.15 V) and the chromic–chromous potentials (φ° = −0. Propagation of the pit. The initiation of pits on an otherwise fully passive 18–8 surface. and rhenium in stainless steels also shift the critical potential in the noble direction. so long as passivity does not break down at crevices because of dissolved oxygen depletion. 19.) in 3% NaCl]. addition of alkalies inhibits pitting. localized corrosion does not initiate in inhibited solutions no matter how long the time. Hence. Stable pitting occurs at temperatures above the CPT [42].. over a period of at least 25 years. 4 molybdenum.g. Below this temperature. nickel. in neutral chlorides (e. as well as avoiding general attack. The oxygen potential in air at pH 7 (0. Similarly.5. In sufficient concentration.2. requires that the corrosion potential exceed the critical potential [0. at which temperature the pitting rate of 18–8 is highest.4. The movement of flowing seawater tends to keep the entire surface in contact with aerated water and uniformly passive.STAINLESS STEELS 351 of 18–8 stainless steel. Other anions shift the critical potential similarly. Pits develop more readily in a stainless steel that is metallurgically inhomogeneous. Pitting corrosion is usually considered to consist of two stages: 1. pitting of 18–8 stainless steels is not observed in deaerated stannic or chromic chloride solutions.77 V) is sufficiently noble to induce pitting. In refrigerating brines.2). The CPT has been explained as the . hence. the nitrate ion shifts the critical potential to a value that is noble to the ferric–ferrous oxidation–reduction potential. the pitting tendency of an austenitic steel increases when the alloy is heated briefly in the carbide precipitation (sensitization) range. critical pitting temperatures (CPT) have been measured.21 V(S. NaCl solutions). addition of 1% Na2CO3 was effective for at least 5 years [41]. However. the stannic–stannous (φ° = 0. accounting for increased resistance to pitting. Pitting resulting from crevice corrosion is also favored whenever a stainless steel is covered by an organic or inorganic film or by marine fouling organisms. 19. An 18–8 stainless-steel sheet exposed to seawater for one year was observed to develop an elongated narrow pit reaching 60 mm (2. At the same time. Ni2+. The resultant high current density accompanies a high corrosion rate of the anode (pit) and. . such as SO2− . The factors accounting for rates of corrosion at crevices follow the same principles as those described for pit growth. which. Through flow of current. inducing breakdown of passivity wherever the products come into contact with the alloy surface. 19. resulting in formation of a deep groove under the FeCl2 stream within a few hours (Fig.5).5 in. NO− ) reinitiate passivity on entering a pit. a passive–active cell is set up of 0. dissolved oxygen or passivator 4 ions (e. STAINLESS STEELS temperature below which the rapid dissolution required for pitting cannot take place. The measured pH of the confined anodic corrosion products of 18–8 stainless steel in 5% NaCl at 200 A/m2 (0. This accounts for a shape of pits elongated in the direction of gravity.6). The high Cl− concentration and low pH ensure that the pit surface remains active. The mechanism of growth was demonstrated in the laboratory [43] by continuously flowing a fine stream of concentrated FeCl2 solution over an 18–8 surface slightly inclined to the vertical and totally immersed in FeCl3. by hydrolysis. however.6 V potential difference. Extraneous anions. as is often observed in practice. the high specific gravity of such corrosion products causes leakage out of the pit in the direction of gravity.) from its point of origin (Fig. and Cr3+ chlorides. at the same time. on the other hand. Successful repassivation 3 depends on factors such as pit geometry and stirring rate. A pit stops growing only if the surface within the pit is again passivated. A similar groove does not form on an iron surface.5 [43]. account for an acid solution (Fig. chloride ions transfer into the pit forming concentrated solutions of Fe2+. bringing the pit and the adjacent alloy to the same potential. Once a pit initiates.6). 19.g. Passive–active cell responsible for pit growth in stainless steel exposed to chloride solution. The higher the electrolyte conductivity and the larger the cathode area outside the crevice.5. have no effect. does not depend Figure 19. The initiation of crevice corrosion.02 A/cm2) is 1.5−0. the higher the rate of attack at the anode.352 ALLOYING FOR CORROSION RESISTANCE. polarizes the alloy surface immediately surrounding the pit to values below the critical potential. because a passive–active cell is not established.. by oxygen depletion in the crevice caused by slow uniform alloy corrosion. Cathodic protection effectively avoids crevice corrosion. This contrasts with the more lenient requirement of polarizing below the critical potential in order to avoid pit initiation. Additions of molybdenum to 18–8 stainless steel (type 316) and palladium additions to titanium (see Fig. . 25. pit began at crevice formed between bakelite rod and interior surface of hole. thereby establishing a still larger potential difference between active metal in the crevice and passive metal outside. on exceeding the critical pitting potential.) exposed to Boston Harbor seawater for 1 year. as well as of chlorides. nitrates. 4 hr. for example. accounting for the observation that crevice corrosion occurs in solutions of sulfates. Instead. Alloy additions that are effective in helping retain passivity in the presence of both low dissolved oxygen concentration and acid corrosion products help reduce or avoid crevice attack. provided that the alloy surrounding the crevice is polarized to the open-circuit potential of the active (nonpassive) alloy surface within the crevice. This mechanism for initiation of crevice corrosion indicates that chlorides are not essential to its occurrence. Section 25. it depends only on factors influencing the breakdown of passivity within the crevice. This breakdown may occur. analogous to cells operating in pitting corrosion.2) are in this category. (b) Artificial elongated pit formed by flowing 50% FeCl2 in a fine stream over 18–8 stainless steel immersed in 10% FeCl3. and so on. followed by setting up a differential aeration cell that results in accumulation of acid anodic corrosion products in the crevice. acetates.STAINLESS STEELS 353 (a) (b) Figure 19. These changes in electrolyte composition eventually destroy passivity.2. (a) Elongated pit in 18–8 stainless steel specimen 75 × 125 mm (3 × 5 in.6. C. Morrill. relatively concentrated chloride that hydrolyzes to a slightly acid pH.3 Reducing or Avoiding Crevice Corrosion. Avoid stagnant solutions. Although at low stress levels. Damaging environments that cause cracking may differ for austenitic. An impressed current can be used. A boiling. can cause cracking of thick sections of stressed 18–8 within hours.2. To avoid S. 3 3.C. given sufficient time in a critical environment. and to minimize pitting. Compressive stresses. NaCl solutions). A *In aerated 4% NaCl. Cathodically protect. STAINLESS STEELS 19. stainless steel can be coupled to an approximately equal or greater area of zinc. .354 ALLOYING FOR CORROSION RESISTANCE. or 18–8 propellers on steel ships. Austenitic stainless steels used to weld mild-steel plates. operate at the lowest temperature possible. [H. Avoid crevices between metals or between metals and nonmetals.2. Reduce oxygen concentration of chloride environments (e. Ensure uniform composition of electrolyte at all regions of the metal surface. there is. no practical minimum stress below which cracking will not occur. Circulate. stainless steels. 3. time to cracking may be long. are not damaging.4. There are three ways to reduce or avoid crevice corrosion: 1. such as FeCl2 or MgCl2. 875 (1941)]. iron. or aluminum [44]. do not pit. Periodically remove contaminating surface films using alkaline cleaners with stainless-steel wool or the equivalent. Add extraneous anions (e. For austenitic steels.2. 2. in general. Eng.7). In seawater. OH− or NO− ) to chloride environments.5 Stress-Corrosion Cracking and Hydrogen Cracking In the presence of an applied or residual tensile stress.4.g. Approaching such a potential reduces the corrosion rate.g. 2. but not to zero. Chem. compared to martensitic or ferritic. the two major damaging ions are hydroxyl and chloride (OH− and Cl−). to the contrary. For type 304.g.* 19. the shorter is the time to failure. or in good conducting media (e. operating temperatures in NaCl solutions should be kept below 60–80°C.. 33. 19. The higher the tensile stress. ideally to the potential of corroding active metal at the crevice. There are four principal ways of reducing or avoiding pitting corrosion: 1. Ind. and aerate electrolytes in contact with stainless steels. 19. use of sacrificial iron and similar less-noble-metal anodes have proved useful [45]. seawater). stir.. Cathodically protect at a potential below the critical pitting potential. stainless steels may crack transgranularly when exposed to certain environments (Fig. maximum weight loss by pitting occurs at 90°C.2 Reducing or Avoiding Pitting Corrosion.. Uhlig and M. 4. but its availability hastens damage. as does the presence of oxidizing ions.7. cracking of 18–8 usually occurs not in alkaline boiler water.4. An undulating path accounts for disconnected cracks as viewed in one plane [47]. Stress-corrosion cracking of 18–8. cracking of austenitic stainless steels is observed only if dissolved oxygen or other oxidizing agent is present [46].5% chlorides.5 and 19. to values that are noble to the critical potential for stress-corrosion cracking. for example. but .8). and the amount of chloride needed to cause damage can be extremely small (Fig. such as Fe3+.) solution of concentrated MgCl2 boiling at about 154°C. 19.STAINLESS STEELS 355 Figure 19.02–0. Another contributing factor is a shift of the corrosion potential in the presence of oxygen. (Copyright ASTM International. is used as an accelerated test medium. Note that the cracks for this environment start at a pit. 19. on the other hand. Cracking by alkaline solutions requires relatively high concentrations of OH−. The presence of dissolved oxygen in such solutions is not necessary for cracking to occur. if small pits form initially at which chlorides concentrate (Figs. Pitting is not a preliminary requirement for initiation of cracks. the cracking tendency is accentuated by concentrated FeCl2 and analogous metal chlorides within the pits. but not in its absence. 100°C (250×). In these instances. hence. oxygen may induce stress-corrosion cracking in sodium chloride solutions because pitting occurs when it is present.1).2. Cracking of 18–8 stainless tubes in heat exchangers has been observed in practice after contact with cooling waters containing 25 ppm Cl− or less.7) (see Section 19. stress-corrosion cracking can occur regardless of whether corrosion pits develop. and cracking has also been induced by small amounts of chlorides contained in magnesia insulation wrapped around stainlesssteel tubes [47]. type 304 stainless steel exposed to calcium silicate insulation containing 0. Reprinted with permission. For this situation. In NaCl and similar neutral solutions. Hence. these steels being sufficiently resistant to warrant their practical use in preference to austenitic stainless steels in chloride-containing solutions. When alloyed with more than 1. pH 10.g. both higher and lower nickel contents resist cracking up to 200 h [50]. They (e. rather within the splash zone above the water line where dissolved alkalies concentrate by evaporation. 68. When annealed at 815°C for 1 h. There is no evidence that transgranular stress-corrosion cracking occurs in pure water or pure steam. at and above 1. J. the 18% Cr–Fe.6.356 ALLOYING FOR CORROSION RESISTANCE. Eckel. The effect of nickel is partly explained by an observed shift of potentials such that. 50 ppm PO3− . 242–260°C (467–500°F). 0.5% Ni. Failures can occur in such solutions in the absence of dissolved oxygen [48]. in general. Soc. are susceptible to transgranular stress-corrosion cracking in MgCl2 boiling at 130°C.003% C stainless steels. Am. STAINLESS STEELS Figure 19. 1–30 days’ exposure [W. Cracking of stressed. Eng. 4 Williams and J. Nav. nickel-free ferritic steels. 93 (1956)]. does not occur in the chloride media described previously.5% Ni. type 430) are also resistant to 55% Ca(NO3)2 solution boiling at 117°C and to 25% NaOH solution boiling at 111°C [49]. the corrosion potential of the cold-rolled material becomes noble . Relation between chloride and oxygen content of boiler water on stresscorrosion cracking of austenitic 18–8-type stainless steels exposed to steam phase with intermittent wetting. only the alloy containing 2% Ni fails.8.. coldrolled. 4. Severely cold-worked 18–8 austenitic stainless steels may also fail under conditions that would damage the martensitic types [52]. when heattreated to high hardness values. Also. All these facts suggest that the martensitic steels under these conditions fail not by stress-corrosion cracking but by hydrogen cracking (see Section 8. Martensitic stainless steels and also the precipitation-hardening types. Increasing the amount of nickel in austenitic stainless steels shifts the critical potential for stress-corrosion cracking in MgCl2 solution in the noble direction more rapidly than it shifts the corresponding corrosion . life of ultra-high-strength 12% Cr martensitic stainless steels was in the order of 10 days or less [57]. hardened martensitic stainless-steel propellers coupled to the steel hull of a ship have failed by cracking soon after being placed in service. even when not coupled to other metals [53–55]. As mentioned in Section 8. 19. or oxidation products of phosphorus or selenium.2. sulfides accelerate damage. Austenitic stainless steels are immune to both hydrogen blistering and cracking. arsenic compounds. accelerates failure. the martensitic steels. hence cracking occurs.9) [58]. In slightly or moderately acidic solutions. 19.5. particularly in the presence of sulfides. Such failures have been demonstrated in laboratory tests [51]. and immersed in aqueous media. practical instances have been noted of self-tapping martensitic stainless-steel screws cracking spontaneously soon after being attached to an aluminum roof in a seacoast atmosphere. Similarly. in salt spray. and since the alloy on cold working undergoes a phase transformation to ferrite. rather than protecting against cracking. are very sensitive to cracking. and exposed to a marine atmosphere.400 MPa) yield strength. the observed effect is probably another example of hydrogen cracking. especially at high current densities. contrary to the situation for austenitic steels. It is generally accepted that ultra-high strength steels crack by a hydrogen embrittlement mechanism [55]. Austenitic stainless steels containing more than about 45% Ni are immune to stress-corrosion cracking in boiling MgCl2 solution and probably in other chloride solutions as well (Fig. for which only specific anions are damaging.1 Metallurgical Factors. Stressed to about 75% of the yield strength. Martensitic stainless-steel blades of an air compressor [56] have failed along the leading edges where residual stresses were high and where condensation of moisture occurred. have been reported to crack spontaneously in the atmosphere. The reverse order of potentials applies to alloys of lower nickel content [50].4). heattreated to approximately >200. Here again.000 psi (>1. The specific anions of the acid make little difference so long as hydrogen is evolved. Galvanic coupling of active metals to martensitic stainless steels may also lead to failure because of hydrogen liberated on the stainless (cathodic) surface. cathodic polarization.STAINLESS STEELS 357 to the critical potential. The more ductile ferritic stainless steels undergo hydrogen blistering instead when they are cathodically polarized in seawater. Edeleanu and Snowden [48] noted that high-nickel stainless steels were more resistant to cracking in alkalies. 10) [60]. the alloys become more resistant [59]. whereas additions of carbon decrease susceptibility (Fig. Stabilizing additions effective in preventing intergranular corrosion. hence. Austenitic alloys do not crack above 45% Ni [58]. the alloys resist stress-corrosion cracking regardless of applied potential.358 ALLOYING FOR CORROSION RESISTANCE. At and above approximately 45% Ni. 19. type 310). nitrogen is the element largely responsible for stress-cracking susceptibility.g. For austenitic steels that are resistant to transformation on cold working (e. The effect is related to alloy imperfection structure rather than to any shift of either critical or corrosion potential [59]. such as titanium or columbium. STAINLESS STEELS Figure 19. Stress-corrosion cracking of 15–26% Cr–Fe–Ni alloy wires in boiling 42% MgCl2.. potential. but metallurgical factors.9. become more important. have no . such as unfavorable dislocation arrays or decreasing interstitial nitrogen solubility. indicating that not environmental. Low-nickel or nickel-free alloys are ferritic and do not crack. in acidified chloride solutions.4. Stress-corrosion cracking of cold-rolled 19% Cr.C. One function of polythionic acids is to act as cathodic depolarizers.C. it was found that rapid intergranular cracking occurred in the presence of a tensile stress (but not otherwise). Chloride S. can occur at lower temperatures [61. 20% Ni austenitic stainless steels in boiling MgCl2 (154 °C) as affected by carbon or nitrogen content [60]. but. The cause was traced to polythionic acids (H2SxO6 where x = 3. 19. nor do alloying additions such as 2–3% Mo in type 316 stainless steel. beneficial effect on stress-corrosion cracking. occurs primarily above about 90°C (190°F).STAINLESS STEELS 359 Figure 19.6 Cracking of Sensitized Austenitic Alloys in Polythionic Acids During shutdown of certain 18–8 stainless-steel oil-refinery equipment operating in the carbide precipitation temperature range.C.5) formed by reaction of residual metallic sulfide films on the equipment surface with moist air at room temperature [63–66].2.10. Such acids can be produced in the laboratory by bubbling H2S through water saturated with SO2. S. 62]. thereby stimulating dissolution of chromium-depleted grain-boundary material. Another possible function is that their cathodic reduction products (H2S or analogous compounds) stimulate absorption to interstitial hydrogen by chromium-depleted .C. It is important to the mechanism of failure that these acids are readily reduced cathodically by operating corrosion cells. g.H.g. but the one prevailing at the crack tip. which is the primary catalyst poison stimulating absorption of atomic hydrogen by the alloy. use of stabilized stainless steels avoids intergranular cracking of the kind described. High concentrations of alkalies occurring initially or incidentally through concentration at crevices or in vapor zones are damaging. However.E. An acid environment is necessary to failure because reaction with alloy supplies the required hydrogen. Coupling of stressed 18–8 to a small area of nickel (φcorr = −0. e. I−. In acid environments. This potential range lies above the values associated with hydrogen ion discharge and would seem to rule out hydrogen cracking. ferritic alloys may become embrittled by hydrogen or they may . 4 d. The latter alloy along grain boundaries. by electrochemical dissolution along an active path exposed through application of stress [68]. it also favors formation of H2S (rather than HS− or S2−).) [68]. eliminate Cl−. Substitute ferritic alloys (e. To Reduce or Eliminate Cracking Austenitic Stainless Steels a.001% C stainless steel) (Fig.04–0. is subject to hydrogen cracking when stressed. STAINLESS STEELS alloy.360 ALLOYING FOR CORROSION RESISTANCE.128 V (S. b. The function of >2 ppm sulfur as Na2S or as cathodic reduction 2− products of sulfites ( SO2− ) or of thiosulfates ( S2O3 ) to induce hydrogen cracking 3 of a 0.H. ferritic in structure. Add extraneous ions (e.C. unlike the grains austenitic in structure that are resistant. However. which can be appreciably more active. In neutral or slightly alkaline chloride environments.10). acetates).g.04% N in the case of 20% Cr. the relevant potential is not the one measured at the alloy surface. The critical potential of 18–8 stainless steel in MgCl2 at 130 °C is −0. 3 c... whereas sulfuric acid is much less effective in causing cracking because SO2− is not similarly reduced. NO−.C. if present) to the lowest possible value (<0. Add buffering ions (e. because SO2− is readily reduced at cathodic sites forming H2S 3 or similarly acting reduction products.77% C high-strength steel and also of ferritic and martensitic stainless steels was demonstrated in seawater [67]. In practice. 20% Ni. or (as a porous nickel coating) in water containing 50 ppm Cl− at 300°C [69]. Avoid high concentrations of OH−. 0. Use alloys containing >45% Ni or reduce the nitrogen content (and other detrimental impurities.).E. It is expected that polythionic acids would act in an analogous manner. The mechanism of failure has been described as S. PO3−).. 19. 4 Cracking in polythionic acids is most pronounced in the potential range 0. Alternatively. type 430 or controlled purity Cr–Mo steels). similar to that by polythionic acids.18 V) prevents cracking in this medium. the mechanism may be one of progressive hydrogen cracking along the grain boundaries of sensitized alloy. Aqueous SO2 solutions are also found to cause intergranular cracking of sensitized 18−8.34 V (S. eliminate dissolved oxygen and other oxidizing ions. Cathodically protect. . 3. Whether used in handling chemicals or exposed to the atmosphere.4. Avoid galvanic coupling to more active metals. This includes metals and alloys like silver..2. Also resistant to cold or hot sulfuric acid if anodically protected. stainless steels are best employed under fully aerated or oxidizing conditions. stainless steels are resistant to: 1. Martensitic. aluminum in environments in which it remains passive. crevice corrosion may cause pitting and localized rusting. Operate below 60–80°C (140–175°F). 19. Precipitation-Hardening. 70% Ni−Cu alloy. Avoid excess current if cathodic protection is applied. Otherwise. 10%) and at boiling temperatures if Fe3+.4).2. and. or Pd are alloyed (see Section 6. c. or nitric acid is added as an inhibitor [71]. except under stress in hot concentrated caustic solutions. f. hardness should be below about Rockwell C 40. Pt. 4.STAINLESS STEELS 361 blister when galvanically coupled to more active metals in certain environments. 76% Ni. minimum susceptibility to hydrogen cracking occurs after tempering for 2 h at 260°C (500°F) [70]. Temper martensitic or precipitation-hardening steels to the lowest possible hardness values. As discussed in Section 19. they can be coupled successfully to metals that are either passive or inherently noble. Nitric acid. which favor the passive state. 16% Cr.g. Many organic acids.7 Galvanic Coupling and General Corrosion Resistance Since stainless steels are passive and exhibit a noble potential. Very dilute sulfuric acid at room temperature when aerated. Cu2+. 4 5. Austenitic stainless steels cooled too slowly through the sensitizing temperature zones tend to rust in the atmosphere. In brief. or at lower temperatures if small amounts of Cu. Sulfurous acid (in the absence of SO2− or Cl−). silver solder. or Ferritic Stainless Steels a. b. nickel. including almost all food acids and acetic acid (but not boiling glacial acetic acid). 2. copper. the alloy surface should always be kept clean and free of surface contamination. Alkalies. usually. Exposed to the atmosphere. 7% Fe alloy. over a wide range of concentrations and temperatures. higher concentrations (e. Maximum susceptibility of types 410 and 420 stainless steels to cracking in salt spray or to hydrogen cracking occurs after tempering for 2 h at 425–550°C (800–1000°F). 3. 4. W. Steigerwald. D. Stainless steels are not resistant to: 1. and E. 2005. 3. Electrochem. Corrosion of wrought stainless steels. editor. 13. Trans. p. 6. These and type 430 are used for auto trim. Metallurgie 8. Corrosion 33. Pickering and C. de Gruyter. 7. Scully and A. Guillet. Lehrbuch der Metallkunde. in Korrosion und Korrosionsschutz. Ann. 202. R. 412–414. Chapman. pp. R. STAINLESS STEELS 6. Lit. 350 (1905). 9. HBr. Stodart and M. J. Rev. Vol. A.g. . REFERENCES 1. 651–666. formic. Ann. Pickering. except for brief periods or when cathodically protected.g. 114. Grubb. Vol. pp. especially fixing solutions containing thiosulfates (pitting occurs). Corrosion: Materials. 143 (1968). 476–489. 745 (1964). Chim. L. OH. Soc. Duplex stainless steels. 2. J. Testing. J. pp. Materials Park. Some organic acids. 253 (1822).. in ASM Handbook. Leipzig.362 ALLOYING FOR CORROSION RESISTANCE. and NaOCl). type 304) in waters containing Cl− plus O2 at temperatures above 60–80°C. H. Wagner.g. Materials Park. and J. 11. Types 302 and 304 have been used successfully as architectural trim of storefronts and buildings (e. 2000. J. Met. Dilute or concentrated HCl. 155 (1904). Corrosion: Materials. 6. Fritz. and G. Soc. and Protection. and lactic acids. F. 2003. 2nd edition. Dundas. Arts 9. Effects of metallurgical variables on aqueous corrosion. Mines 6. R. Q. 13A. also not resistant to salts that hydrolyze to these acids. Soc. 3. 428–433. L. 115. Photographic solutions. Tödt. Seawater. 279 (1977). pp. J. 8. M. Corrosion: Fundamentals. P. 2. R. 17. 161 (1911). J. Noël. 12. 13B. 10. ASM International. Lucente. J. in Uhlig’s Corrosion Handbook. CuCl2. J. Electrochem. J. 332 (1906). Iron Steel Inst. Phys. editors. 5. Roy. Truman. Corrosion of amorphous metals. Materials Park. 2. Berthier. 1932. DeBold. W. T. Berlin. F. FeCl3. in ASM Handbook. pp. P. Oxidizing chlorides (e. Perry. 259. in ASM Handbook. Lizlovs. 698 (1967). Revie. H. Philos. 1955. Vol. ASM International. Bond. Faraday. 573 (1821). 319 (1820). the Chrysler and Empire State buildings in New York City). 13B. Tammann. 1. 2005. OH. 5. Katz. 4. 56–57. ASM International. Sci. G.. J. Monnartz. 55 (1821). Wiley. and HF. The atmosphere. OH. 112. Erbing Falkland. including oxalic. New York. Stressed austenitic alloys (e.. H. HgCl2. A. 38. P. Duncan. Tofaute. 24 (1968). p. 22. 24. 39. 20. Soc. 30. 3705 (2007). 106. 1965. 21–35.. Rev. Philadelphia. Wanklyn and D. Ref. 15. 2000. 16. 36. 41. 42. Houston. Frankenthal and H. Soc. September 1969. Butterworths. 33. in Proceedings of the First International Congress on Metallic Corrosion.. Streicher. E. de Pourbaix. R. Rutherford. Uhlig. Binder. Frankenthal and F. A. J. Tofaute. Am. P. Bain. Hinnüber. H. Am. Coriou et al. Mansfeld. Austenitic and ferritic stainless steels. American Society for Testing And Materials. and J. 8. 19. Trans. H. 1996. Hochmann. in Corrosion and Corrosion Protection. J. J. 87. 21. 405 (1936–37). H. 143 (1967). Y. P. H. Coriou et al. 411 (1940). E. J. Metals 46. Pennington. A. Beck. Mater. Wiley. 49(9). 89 (1964). Soc. Metals 30. 48. P. Upp. Am. 28. 759 (1958). Schafmeister and W. J. Corrosion 22. H. Krupp 3. 798 (1964). Metall. 352–359. New York. H. 223 (1935). Trans. B. R. D. and M. 613. Trans. Corrosion 7. Kasparova. 161 (1959).REFERENCES 363 14. 37 (1969). P. Stahl Eisen 72. Eng. W. editors. Ernst and R. 26. 481 (1933). Östberg. Brown. Schafmeister. 23 (1973). in Fundamental Aspects of Stress Corrosion Cracking. 734 (1951). Nucl. Arch. 65. Lula. Mater. 1969. 34. H. Soc. 1972. New York. Sci. Corrosion 24. Jones. Kolotyrkin and O. Corros. 615. 2000. 32. Streicher. National Association of Corrosion Engineers. 17. 613. Lacombe. 120. Austenitic and ferritic stainless steels. Z. Metals 50. 37. 419 (1951). W. Inst. . 1962. R. R. NJ. H. in Proceedings of the Fourth International Congress on Metallic Corrosion. Mitt. See also Ref. 1301 (1949). editor. Corrosion 24. pp. 275. Soc. Corros. H. Electrochem. 18. 2. 4. Pickett. M. W. Copson and G. Uhlig. 1981. Am. Streicher. See also M. Houdremont and W. Revie. C. 29. R. Am. Newman. p. Soc. Uhlig. Chem. J. F. Eisenhüttenw. Stahl Eisen 84. Metals (Steel Treating) 21. A. 188. Electrochem. Metals 41. Electrochem. 2nd edition. 25. CORROSION/96. Electrochemical Society. M. Bäumel. W. Anorg. Allgem. Lizlovs. Sci. Houston. 539 (1952). Economy. in Advances In Technology of Stainless Steels and Related Alloys. H. 40. M. and G. Fontana. Armijo. Uhlig. J. 22. 140. Trans. Houston. Soc. TX. 1305 (1969). 154 (1959). J. J. Armijo. Lena. in Uhlig’s Corrosion Handbook. 23. p. 10. Met. Electrochem. Hübner. Tech. Corrosion Sci. NACE. Revie. Trans. 16. 947 (1942). Arnvig and A. H. Ebling and M. R. 43. 55 (1968). Texas. Trans. 309 (1930). p. National Association of Corrosion Engineers.. Matthews and H. 116. Baümel et al. J. Pickering. 7. Franks. and A. 197 (1954). Aborn. Prot. Bond and E. and G. Amsterdam. editor. p. Armijo. Soc. Trans. pp. 186. R. Strauss. 35. A. TX. and R. J. Wiley. Scheil. Paper 437. 193 (1945). Kiefer. A. 27. C. Min. in Uhlig’s Corrosion Handbook. 280 (1966). London.-E. Soc. Am. p. and J. Bisgård. 31. 2nd edition. Schottky. and G. 67. and Protection. Australia. quoted in Fracture. Vol. Uhlig. p. Corrosion 25. G. T. Copson. Bond. Bull. Rhodin. Edeleanu and P. E. 1967. STAINLESS STEELS 44. and Protection. 13A. J. 68. 57. 117. Craig. 23(4). H. J. 69. and Protection. 63. 493 (1969). T. F. Corrosion in petroleum refining and petrochemical operations. Academic Press. p. Materials Park. editor. 345t (1963). Corrosion: Fundamentals. 351t (1955). C. pp. 56. 99. Am. Phull. Nakamura. Tuthill. E. Am. Kane. A. Eng. 116. Corrosion: Environments and Industries. 38 (1984). 62. H. Corrosion 38. Corrosion 19. 54. 406 (1957). Soc. Harston and J. H. New York. TX. editor. Electrochem. 22(1). I. P. Iron Steel Inst. in ASM Handbook. A. H. Soc. B. in ASM Handbook. Testing Mater. 373. in Physical Metallurgy of Stress Corrosion Fracture. and H. in ASM Handbook. H. R. 59. Griess. 1962. New York. Chapman. J. Trans. 1975. Corrosion 20. 247. New York. Mater. 58. Lillys and A. 2006. ASM International. 186. W. American Society for Testing and Masterials. D. Newberg and H. Soc. p. 18 (1970). 13A. Perf. 2003. Metals 48. 1969. Vol. p. Samans. Evaluating stress-corrosion cracking. LaQue. J. p. Snowden. 13A. Uhlig. 416. 745 (1964). 16(1). 2003. Mater. 981 (1972). Elliott. Iron Steel Inst. in Fundamental Aspects of Stress Corrosion Cracking. Macdonald. 319. p. in Proceedings of the Sixth International Congress on Metallic Corrosion. 61. 307 (1954). Dana. Takahashi. OH. Testing. D. Metal Progr. H. 49. Mater. 1959. Philiadelphia. 60. Uhlig. J. Testing.364 ALLOYING FOR CORROSION RESISTANCE. p. Phelps and R. 225. Trans. Uhlig. Cobb and H. 3. 65. Bloom. OH. Berry. 45. Williams and J. H. Neumann and J. National Association of Corrosion Engineers. OH. Uhlig and J. 327 (1956). Melbourne. 915. Lee and A. Nehrenberg. Houston. p. Testing. 46 (1957). Corrosion 11. 1948. Vol. 70. Materials Park. Liebowitz. 991–992. in Corrosion Handbook. Mears. 64. p. Soc. Cragnolino and D. Australia. Australasian Corrosion Association. 53. H. and H. C. Sydney. Phelps. 514. J. 46. in Stress Corrosion Testing. B. 119. Nishida. 50. H. Sava. 52. Dundas. 2003. Perf. in ASM Handbook. Nav. W. editor. 256t (1964). J. P. 202. Interscience. J. J. Eckel. 486 (1950). 13 (1952). 656. 48. L. Badger. Wiley. Soc. C. 62. Uhlig. 68. 22 (1977). 1971. Vol. ASM International. Eng. 13C. First International COngres On Metallic Corrosion. K. OH.. Perf. Corrosion: Fundamentals. Butterworths. Soc. Am. Truman. Vol. 51. P. R. Uhlig. F. Corrosion: Fundamentals. Scully. ASM International. Matsushima. J. . 57. H. Jr. Electrochem. 48 (1983). ASM International. Lee and H. 406 (1982). R. 604. Special Technology Publication No. 93 (February 1956). Materials Park. 71. Electrochem. 55. Marshall. Soc. 425. p. London. Materials Park. Automot. 47. 66. 398. pp. ASM International. New Delhi. Glassy Alloys J. Corrosion of cast stainless steels. Wiley. R.GENER AL REFERENCES 365 GENERAL REFERENCES B. Corrosion: Materials. editors. New York. Austenitic and ferritic stainless steels. OH. W. Lucente. editor. OH. Corrosion: Materials. Martensitic stainless steels. 2005. Sedriks. Fundamental Aspects of Corrosion Films in Corrosion Science. D. . J. D. 13B. Materials Park. pp. Corrosion of wrought stainless steels. 2005. in Uhlig’s Corrosion Handbook. Revie. A. pp. Vol. in ASM Handbook. editor. 2005. Mitigation and Monitoring. R. R. Blair. S. in Uhlig’s Corrosion Handbook. Corrosion of Austenitic Stainless Steels: Mechanism. Wiley. Wiley. 1996. Khatak and B. Scully and A. pp. A. Vol. 2nd edition. 2000. in Uhlig’s Corrosion Handbook. F. 651–666. pp. J. 1991. Streicher. Stainless Steels M. Craig. T. W. 13B. 601–650. Narosa Publishing House. 2nd edition. Corrosion of Stainless Steels. Revie. R. Raj. Corrosion of amorphous metals. ASM International. Revie. 54–77. Materials Park. and J. Grubb. New York. DeBold. Grubb. Plenum Press. 13B. 2002. New York. 2000. M. 2000. ASM Handbook. 2nd edition. New York. 2nd edition. in ASM Handbook. ASM International. Duplex stainless steels. M. 78–87. OH. editor. Erbing Falkland. Fritz. J. 667–676. Vol. F. W. Corrosion: Materials. New York. Materials Park. H. L. Wiley. . a Cu2O film develops that is covered by a thin film of CuO. for example. considering that passivation occurs by formation of the oxides Cu2O and CuO [2]. as. if complexes are formed. On the other hand.1 COPPER Cu 2 + + 2e − → Cu φ° = 0. Copper and its alloys.20 COPPER AND COPPER ALLOYS 20. which is formed as the film thickness increases [1]. immunity. In Fig. 367 . although toxic to fouling organisms. Comparatively. by R. such as seaweed and shellfish. and passivation of copper at 25°C are shown. between Cu+ and Cl− in a chloride solution. Equilibrium relations at the metal surface indicate that the reaction Cu + Cu2+ W 2Cu+ is displaced far to the left (see Problem 1. p. Uhlig Copyright © 2008 John Wiley & Sons. the theoretical potential–pH domains of corrosion. the continuous depletion of Cu+ by conversion to CuCl − favors the 2 univalent ion as the major dissolution product.337 V Copper dissolves anodically in most aqueous environments forming the divalent ion Cu2+.1. when copper is heated in air at elevated temperatures. undergoing severe attack by sulfide if conditions allow the growth of sulfate reducing bacteria [3. 20. are subject to microbiologically influenced corrosion. 4]. Winston Revie and Herbert H. 381). Inc. Corrosion and Corrosion Control. 389.2).) Figure 20. Cu2O and CuO [2]. Atlas of Electrochemical Equilibria in Aqueous Solutions. and it is thermodynamically inert with no tendency to corrode in water and in nonoxidizing acids free of dissolved oxygen. immunity.g. excellent electrical and thermal conductivity. 20. 2nd English edition. and ease of soldering or brazing. It is noble to hydrogen in the Emf Series. (M. NH + ).. assuming that passivation results from the formation of the oxides. Copper is also characterized by sensitivity to corrosion by high-velocity water and by aqueous solutions. CN−. Pourbaix. In oxidizing acids and in aerated solutions of ions that form copper complexes (e. The rate increases with dissolved oxygen content. Copyright NACE International 1974 and CEBELCOR. p.368 COPPER AND COPPER ALLOYS Figure 20. Copper is a metal widely used because of good corrosion resistance combined with mechanical workability. and passivation for copper. whereas in air-free high-velocity water .2. Theoretical potential–pH domains of corrosion. corrosion can be 4 severe.1. Longitudinal cross section of undercut pits associated with impingement attack of condenser alloy by salt water (7×) (flow of water from left to right). called impingement attack (Fig. The rate of corrosion increases with dissolved oxygen concentration and temperature. For this reason. a water containing carbonic or other acids nevertheless may corrode copper and copperbase alloys sufficiently to cause blue staining of bathroom fixtures and cause an increase in the corrosion rate of iron.001–0.COPPER 369 up to at least 7.S. a hot water in Michigan. Aluminum is an exception because of its sensitivity to extremely small amounts of Cu2+ in solution. with normal corrosion being sufficient to release copper ions in concentrations that poison marine life. zeolite softened with resultant high concentration of NaHCO3. In groundwaters. with the usual water treatment being inadequate to reduce Cu2+ pickup to nondamaging levels.5 m/s (25 ft/s). It is one of the very few metals that remains free of fouling organisms. It is possible to avoid the damaging effects of Cu2+ contamination of water by using copper piping coated on the inside surface with tin (tinned copper). Environmental Protection Agency Lead and Copper Rule. implemented in 1991.1 Corrosion in Natural Waters Copper is resistant to seawater. the oxide film on copper is found to be a mixture of Cu2O and CuO [6. The same water unsoftened. corrosion resistance depends on the presence of a surface oxide film through which oxygen must diffuse in order for corrosion to continue.0 gmd (0. copper tanks for holding distilled water or for water storage on ships .3 mg L−1 [10].002 ipy) in quiet water and somewhat higher in moving water.5–1. In seawater and in fresh waters. forming numerous small galvanic cells. Treatment of acid waters or waters of negative saturation index with lime or with sodium silicate reduces the corrosion rate sufficiently to overcome both staining and accelerated corrosion of other metals. the rate of corrosion may show a maximum as temperature is increased because the solubility of oxygen decreases with temperature [5]. In an open system. Visible light markedly retards the rate of oxide formation [6]. Accelerated corrosion in this instance is caused by a replacement reaction in which copper metal is deposited on the base metal. the rate increases by 11/2–2 times. leading to a high corrosion rate. the corrosion rate increases with increasing salinity (concentration of dissolved salts) [5]. or aluminum surfaces with which such water comes into contact. except aluminum. Even if the corrosion rate is not excessive and the copper is durable. on the other hand. impingement attack is negligible. 7]. The U. Copper piping used to transport well water in homes in some areas of New Jersey [9] developed pinhole leaks (pits) within 10 years. galvanized steel. with the corrosion rate in temperate climates being about 0. was not nearly as corrosive because a protective film of CaCO3 containing some silicate was deposited on the metal surface. limits total copper in potable water to a maximum of 1.1. 20. The film is easily disturbed by high-velocity water or is dissolved by either carbonic acid or the organic acids that are found in some fresh waters and soils. For example. In tropical climates. Upon exposure to distilled water at room temperature. perforated copper water pipe within 6–30 months [8]. as high as 0. The tin coating must be pore free in order to avoid accelerated attack of copper at exposed areas. 2. 3. sacrificial anodes of aluminum are sometimes used. and other nonoxidizing acids. H3PO4. In addition to the effects just described. Copper is especially suited to convey soft. whereas mounds around the pits may contain basic copper sulfate.26 mm/year (62 mdd). may result in sufficient copper to stain plumbing fixtures [11]. whether hot or cold. Hard waters form a protective film of calcium compounds on the copper surface that protects against corrosion. the corrosion rate of copper is significant. above 60°C.370 COPPER AND COPPER ALLOYS should be tinned [11]. In the second category are soft waters containing small amounts of manganese salts. hot or cold. particularly in certain areas of Great Britain [15]. To protect copper hot-water tanks from pitting. type 1 pitting.051–0. supplemented in some cases by turbulent flow. The first category. With these waters. in waters with low bicarbonate/ sulfate ratio. the types of pits that can develop on copper have been divided into categories [12. Copper piping is usually satisfactory for transporting seawater and hard waters. < 1. acetic acid. Type 1 pitting has been largely eliminated by improvements in the tube manufacturing processes [11]. Seawater. corrosion products within the pit are Cu2O plus a small amount of chloride.16 mm/year (12–7 mdd). In this respect. 13]. usually occurs in the hottest part of the system. usually takes place in hard or moderately hard waters and is usually restricted to the cold portions of the system. dilute H2SO4. Periodic cleaning of the piping or tubing usually prevents corrosion by such deposits. pitting. copper is resistant to: 1. In this case. In soft waters. and the pipe surface is covered by a black scale rich in manganese oxide. and in carbonated waters in general. since tin (or copper–tin intermetallic compounds) is cathodic to copper. pitting nevertheless may occur in waters of certain compositions or as a result of pipe fabrication methods. Deaerated. beneath which is a highly cathodic film of carbon [14] that was the residue of the lubricant used during pipe manufacture. particularly those containing appreciable amounts of free carbon dioxide. designated as Type 2 pitting. Fresh waters. which initiates impingement attack. hot or cold. . Pits of this type contain a higher proportion of cuprous chloride than do pits formed in soft water. The initial corrosion rate in distilled water may be 0. In summary. Even when the copper surface is free of deposits. whereas the rate in an aggressive supply water. a pitting type of corrosion may be induced in waters having relatively good conductivity if dirt or rust from other parts of the system accumulate on the copper surface. Differential aeration cells are formed. aerated waters that are low in carbonic and other acids. The surrounding surface scale is largely CaCO3 stained green. The resulting corrosion is sometimes called deposit attack. and mounds consist largely of cupric carbonate. dry Br2 at 25°C. 2. Corrosion failures of brasses usually occur by dezincification. In corrosive waters (high in O2 and CO2. 20. as shown in Fig. High-velocity aerated waters and aqueous solutions. iodine below 375°C [16]. The tendency for brass to corrode in these ways. Muntz metal. 4. varies with zinc content.2. brasses are used in preference to copper for condenser tubes.5–4% Si have better physical properties than copper and similar general corrosion resistance. 5. and some sulfur compounds. Manganese bronze is also similar. Hydrogen sulfide. <65°C (<150°F).1 Copper–Zinc Alloys (Brasses) Copper–zinc alloys have better physical properties than copper alone. 13] below 1. Atmospheric exposure. Pitting is usually caused by differential aeration cells or high-velocity conditions. FeCl3 and Fe2(SO4)3. dry Cl2 below 150°C. pitting. Various names have become attached to brasses of different zinc content. for example. Copper is not resistant to: 1. 20. NH4OH (plus O2). 40% Zn−Cu.3.COPPER ALLOYS 371 4.2 COPPER ALLOYS Tin bronzes are copper–tin alloys noted for their high strength. Cu(NH 3)2+. losing only 20% of the corresponding weight loss of copper after 16 years [17]. It can normally be avoided by keeping the brass surface clean at all times and by avoiding velocities and design geometry that lead to impingement attack. Oxidizing acids. containing about 1% . hot concentrated H2SO4. dry HF below 600°C. Oxidizing heavy metal salts. is used primarily for condenser systems that use fresh water (e.4 m/s (8 ft/s). 3. the velocity should be kept [8. the velocity should be kept below 2. or stress-corrosion cracking. Substituted NH3 4 compounds (amines) are also corrosive. and they are also more resistant to impingement attack. Great Lakes water) as coolant.2 m/s (4 ft/s). except for pitting attack. Naval brass is a similar composition. 5. and aerated nonoxidizing acids (including carbonic acid). 20. for example. The copper– silicon alloys containing 1. sulfur.g. Halogens under specific conditions—for example. HNO3. hence. A complex ion. Alloys containing more than 5% Sn are especially resistant to impingement attack. forms. These compounds are those that cause stress-corrosion cracking of susceptible copper alloys. Seawater immersion tests at Panama showed that 5% Al−Cu was the most resistant of the common copper-base alloys. low in Ca2+ and Mg2+). fluorine below about 400°C. but containing 1% Sn. in less corrosive waters.. as one type of dealloying. called layer type (Fig. Yellow brass. The alloy gradually dezincifies in seawater and in soft fresh waters. or uniformly over the surface.2. . antimony. and impingement attack with increasing zinc content in copper–zinc alloys (brasses). In brasses. Dezincification of manganese bronze propellers in seawater is avoided to some extent by the cathodic protection afforded by the steel hull. 20. Red brass. a plug of dezincified alloy may blow out. dezincification takes place either in localized areas on the metal surface. Trends of dezincification. is relatively immune to dezincification. 20. was defined in Section 2. or phosphorus still further retards the rate of dezincification. each of tin. called inhibited admiralty metal. the resulting alloy. This tendency is retarded by addition of 1% Sn. Brass so corroded retains some strength. but is more susceptible to impingement attack than yellow brass. Addition of small amounts of arsenic. is used for a variety of applications where easy machining and casting are desirable.5). it is used for ship propellers. Because dezincified areas are porous. Other types of dealloying include the selective dissolution of copper in copper–gold alloys. 20. is used in seawater or fresh-water condensers. but has no ductility. 15% Zn−Cu. leaving a hole. 30% Zn−Cu. iron. and. among other applications.372 COPPER AND COPPER ALLOYS Figure 20. for plug type.4. called plug type (Fig.2 Dealloying/Dezincification Dezincification. the corresponding alloy being called admiralty metal or admiralty brass.4).3. Layer-type dezincification in a water pipe may lead to splitting open of the pipe under conditions of sudden pressure increase. stress-corrosion cracking. and lead. Plug-type dezincification in brass pipe (actual size).4.5.COPPER ALLOYS 373 Figure 20. Figure 20. Layer-type dezincification in brass bolts (actual size). . Looking at the situation metallurgically. The detailed mechanisms of dealloying have recently been reviewed [19]. antimony. as mentioned earlier. or phosphorus. are: 1. by small alloying additions of tin plus a few hundredths percent arsenic. . Also. For example. Requirements for cracking include the presence of water. conditions favoring thermogalvanic action between overheated local areas and adjoining colder areas of heat exchanger tubes can cause attack of this kind at the overheated zones. based on preexisting interconnected paths of like elements in the binary alloy and effects of curvature on dissolution potential [24. 2. 25]. constituted early evidence for a volume diffusion mechanism of zinc transport from the bulk alloy to the surface [26]. an extension of the surface diffusion mechanism. visible in the completely or incompletely dezincified layer. The surface diffusion mechanism. The four main mechanisms that researchers have developed to account for the processes by which one metal in an alloy is removed by corrosion.3 Stress-Corrosion Cracking (Season Cracking) Virtually all copper alloys. copper corrodes preferentially. without dissolution of gold. Any one of these mechanisms may apply in specific instances of dealloying. The volume diffusion mechanism. especially if acid. brasses that contain 15% or less zinc are usually immune. twin bands in brass. can be made to crack in ammonia [27].374 COPPER AND COPPER ALLOYS plugs may be covered on the outside surface with corrosion products and residues of evaporated water. In the gold–copper alloy system.2. 3. 22]. as well as pure copper. 20. and (3) porous inorganic scale formation. ammonia. leaving a porous residue of gold–copper alloy or pure gold. leaving behind a porous metal. Although inhibited admiralty brass resists dezincification. according to which the alloy corrodes and the more noble metal (copper in copper–zinc alloys) is then redeposited to form a porous outer layer [20]. dezincification of α-brasses (up to 40% Zn) can be reduced. based on selective dissolution of the less noble element and volume diffusion of both elements in the solid phase [21. Conditions of the environment that favor dezincification are (1) high temperatures. (2) stagnant solutions. The percolation model of selective dissolution. in which only the less noble element (zinc in copper–zinc alloys) dissolves and the remaining more noble metal is rearranged by diffusion on the surface and nucleation of islands of almost pure metal [23]. 4. Attack is reduced by using inhibited or scaling (positive saturation index) waters [18]. The ionization–redeposition mechanism. Cracking through the grains (transgranular) may occur in specific test solutions or if the alloy is severely deformed plastically. According to one source. When an α brass is subjected to an applied or residual tensional stress in contact with a trace of NH3 or a substituted ammonia (amine). 29]. most likely by reducing the dissolved oxygen concentration [27. . premature failure of yellow brass brackets in the humidifier chamber of an air-conditioning system was traced to this cause [30]. and cracking occurs at ambient temperature.C. Only trace amounts of ammonia are required in many cases. Traces of nitrogen oxides may also cause stress-corrosion cracking. 28].6). both types of cracking were originally called season cracking because of the resemblance of stress-corrosion cracks in bar stock to those of seasoned wood.6. Intergranular stress-corrosion cracking of brass (75×) (specimen stored 1 year). Small concentrations of hydrogen sulfide inhibit S. Stresscorrosion cracking failures of copper pipe under elastomeric insulation have been attributed to wet ammoniacal environments [27. In England. in the presence of oxygen (or another depolarizer) and moisture.COPPER ALLOYS 375 air or oxygen.C. it cracks usually along the grain boundaries (intergranular) (see Fig. 20. In one instance of this kind. particularly during the monsoon season. of brass in petroleum refinery process streams. the origin of the term is ascribed to the fact that years ago brass cartridge cases stored in India were observed to crack. and tensile stress in the metal. probably because such oxides are converted to ammonium salts on the brass surface by chemical reaction with the metal. The air had passed through an electrostatic dust Figure 20. C. S. Similar cracking of stressed brass could be reproduced in the laboratory over a period of days by using a spark discharge in air of 100% relative humidity. with times to failure being longer at higher pH and considerably longer at lower pH values.C.e. with the latter settling down as dust on the brass parts. Since the corresponding corrosion potential is 0. Mattsson [32] observed that minimum cracking time of 37% Zn−Cu brass in a solution of 1 mole NH 3 + NH + .003M Cu2+ must be present in the test solution for spontaneous cracking to occur.C. but also many more sulfate particles than Los Angeles air.3.095 V (S..7 shows the effect of applied potential [34] on S.C.005M NaBr or 0. 23% Zn−Cu alloy (nickel brass) parts of Central Office Telephone Equipment in Los Angeles occurred within two years for similar reasons [31]. only moisture is required to cause failure [35]. Br−..C.H. in the absence of other metal ammonium complexes. but at more noble potentials.04M NaCl to the Mattsson test solution inhibits spontaneous cracking. plus 0. and.E. indicating that sulfates may act as inhibitors. If cracks do not appear with 15 min. the equivalent of more than 0. Figure 20. 45–50% Zn−Cu) having a β or β + γ structure.26 V. specific gravity 1. . does not stress-corrosion crack.376 COPPER AND COPPER ALLOYS precipitator. when cracking does not occur. Annealed brass.). if not subject to a high applied stress. in turn. Also. the high voltage field of which generated traces of nitrogen oxides.g. Johnson and Leja [33] reported stress-corrosion cracking of brass in alkaline cupric citrate and in tartrate solutions at pH values in which complexing of Cu2+ is pronounced. formed corrosion products on the brass surface which were found by analysis to contain a high proportion of NH4. stresscorrosion crack through the grains (i. High-zinc brasses (e.C. transgranularly).42) per liter of water. The critical potential below which cracking does not occur is 0. Similar failures have not been observed as frequently in New York City. causing intergranular cracking of the stressed brackets. addition of more than 0. These. the alloy is probably free of damaging stresses. Mercury is released and penetrates the grain boundaries of the stressed alloy. unlike α brasses. The latter observations are related to a shift of potentials by Cu2+. Whether residual stresses in cold-worked brass are sufficient to cause stress-corrosion cracking in an ammonia atmosphere can be checked by immersing brass in an aqueous solution of 100 g mercurous nitrate [Hg2(NO3)2] and 13 mL nitric acid (HNO3. where the air contains less nitrate. Pollution of Los Angeles air accounts for abnormally high concentrations of nitrogen oxides and suspended nitrates. Stress-corrosion cracking also occurs on substituting Cd2+ or Co2+ in place of Cu2+ additions. Stress-corrosion cracking of 12% Ni. on simple immersion of brass is spontaneous. of 37% Zn−Cu brass in a solution similar to that proposed by Mattsson. or Cl− such that. the corrosion potential in each case lies active to the critical potential for S.05 mole CuSO4 per liter occurred at pH 4 7. Low-zinc brasses are more resistant than high-zinc brasses. It is found that. No alloying additions in small amounts effectively provide immunity to this type of failure in brasses. According to the generally accepted film rupture mechanism. it has been noted that the grain boundaries of polycrystalline brasses are more active in potential than the grains. but not in FeCl3 solution. alternatively. CdSO4. 1M (NH4)2SO4. S.05M CuSO4. Both high-purity alloys and single crystals of α brass crack when stressed in NH3 atmospheres [36]. in any event. It has been proposed. The Electrochemical Society.) The mechanism of stress-corrosion cracking in brasses has been the subject of much study. initiates by film rupture as a result of plastic deformation at the crack tip. and repassivation is inhibited by further plastic deformation at the crack tip [39–41]. (Reproduced with permission. The resulting composition gradient favors galvanic action between such areas and the grains. Copyright 1975. that a brittle oxide film forms on brasses that continuously fractures under stress exposing fresh metal underneath to further oxidation [38]. In support of an electrochemical mechanism. as measured in NH4OH. in which stress-corrosion cracking does not occur [37]. Effect of applied potential on time to failure of 37% Zn−Cu brass in 0. or CoSO4.C.C. that zinc atoms aided by plastic deformation segregate preferentially at grain boundaries. The crack propagates by localized anodic dissolution. There is evidence. accounting for slow .5 at room temperature [34]. pH 6.COPPER ALLOYS 377 Figure 20.7. leading to rapid crack formation.C. heating at 350°C (660°F) for 1 h may be effective. Avoiding Contact with NH3 (or with O2 and Other Depolarizers in the Presence of NH3).C.C. 0. 20. It is difficult to guarantee that there will be no contact with NH3. Aluminum brass may also pit readily in unpolluted. Susceptibility to stress-corrosion cracking of brasses can be minimized or avoided by four main procedures: 1. bal.. because the concentration of oxygen is extremely low. brass condenser tubes do not crack in boiler-water condensate containing NH3. one of the cupro-nickel alloys (10–30% Ni. or by coating brass with a sacrificial metal (e.g. may also favor adsorption of complex ammonium ions within a specific potential range. Hence. subject to plastic deformation. in α brass. Although segregated zinc may be essential to the observed intergranular corrosion of brasses. but recrystallization and some loss of strength of the alloy result. it is probable that the defect structure of grain boundaries or of slip bands is more important to stresscorrosion cracking. For 30% Zn−Cu brass. or beryllium [43]. Similar effects can occur along slip bands (transgranular cracking). seawater. Because ammoniacal S. 29].2. Muntz metal.378 COPPER AND COPPER ALLOYS intergranular attack in a variety of corrosive media without the necessity of an applied stress (intergranular corrosion). arsenic. 3. can occur at relatively low stress levels. and admiralty metal (inhibited) are frequently used. in part. antimony. 4. admiralty metal. 2% Al. For polluted waters. Cathodic protection to a potential below the critical value can be provided by either impressed current. 2.04% As) are used. Plastics containing or decomposing to traces of amines are a continuing source of damage to unannealed brass. failure of copper-base alloys by cracking can occur when copper is alloyed not only with zinc.C. copper. zinc). according to which a brittle crack that initiates in a thin surface film propagates into the ductile matrix and eventually blunts and arrests. To explain the mechanism of transgranular S. For brackish or seawater. but stagnant. after which the process repeats itself [39. Using H2S as an Inhibitor [27. On the other hand. has similarly caused cracking of brass. but also with a variety of other elements. Cathodic Protection. 45]. nickel. Fertilizer washing from farm land. 76% Cu. . aluminum. Cu) and aluminum brass (22% Zn. The mechanism may. stress relief may not be adequate in all cases [27]. involve reaction with available free oxygen. such as silicon. phosphorus [42]. But such areas. Newman and Sieradzki have proposed the film-induced cleavage model. or air over fertilized soil. 44. Stress-Relief Heat Treatment. because of the small traces of ammonia that cause cracking.4 Condenser Tube Alloys Including Copper–Nickel Alloys For fresh waters. cupro-nickel alloys are preferred over aluminum brass because the latter is subject to pitting attack. Wiley. Piping and Air-Conditioning. pp. OH. M. May. especially the 10% Ni−Cu alloy. Cupro-nickel alloys are especially resistant to highvelocity seawater when they contain small amounts of iron and sometimes manganese as well. Testing. Brussels. 1. July. New York. Little and J. Heating. W. It is found that supplementary protective films are formed on condenser tube surfaces when iron is contained in water as a result of corrosion products upstream or when added intentionally as ferrous salts. . pp. the amount of alloyed iron is usually less (e. Wiley. 2000. S. and E. Materials Park. Materials Park. Vol. pp.. Obrecht. 2nd edition. for the analogous 30% Ni composition. 109–116. p. 184. It appears that sulfides interfere with formation of the usual protective oxide films [47]. in ASM Handbook. 645. 13A. Stott. p. Krger. 6. REFERENCES 1. Evaluating microbiologically influenced corrosion. M. 8. Mental and Elemental Nutrients.0% Mn maximum) [46]. pp. 125–133 (1960). Also see M. 847 (1959). September. ASM International. Corrosion of copper and copper alloys. Corrosion: Fundamentals. 0. 7. March.40–0. 2000. second English edition. Vol. J. Sci. pp. CT. Shanley. April. 115–122. Accordingly. 5. p. A detailed general account of the behavior of copper–nickel alloys. pp. pp. J. A. editor. p. 106.75% Mn maximum. Ref. 481 (1980). the optimum iron content is about 1. 9. New York. 189t (1962). J. J. Atlas of Electrochemical Equilibria in Aqueous Solutions. C. Cohen. p. Corrosion of copper and copper alloys. 105–113. April. A. C. TX. and Protection. B. Wiley. R. January. Revie. 137. 131–137. in ASM Handbook. 4. 755.REFERENCES 379 Aluminum brass resists high-velocity waters (impingement attack) better than does admiralty metal.g. F. The susceptibility of Cu−Ni alloys to corrosion increases in aerated sulfidepolluted seawater. For the 10% Ni cupro-nickel alloy.75%. Quill. or compared to any of the 30% Zn−Cu brasses. Pourbaix. 129–134 (1961). 1974.70% Fe accompanied by 1. 750–753. Houston. Lee. Corrosion: Materials. the beneficial effect of iron alloyed with copper–nickel alloys is considered to result from similar availability of iron in the formation of protective films. 2003. D. 3. It is pertinent that the 30% Ni−Cu alloy is relatively resistant to stress-corrosion cracking compared to the 10 or 20% Ni−Cu alloys [42]. 10. Keats Publications. Pfeiffer. C. New York.0–1. 2. in seawater is given by Stewart and LaQue [48]. and CEBELCOR. 2007. 1975. 330. Verink. p. Soc. National Association of Corrosion Engineers. Obrecht and L. 13B. Sequeira. 389. New Canaan. Electrochem. OH. with 0. ASM International. 165–169. 2005. R. Corros. Microbiologically Influenced Corrosion. in Uhlig’s Corrosion Handbook. Corrosion 18. 20. C. pp. Hummel. TX. . Ereneta. 12. 19. OH. H. 633 (1954). Lenox and P. Plenum Press. Materials Park. 1981. 1969.380 COPPER AND COPPER ALLOYS 11. Pickering and C. 2003. Mater. Wagner. Electrochem. 16. Publication 15. London. 15. J. and J. 49. 42. Hermance. 345 (1950). Soc. 178 (1966). 7. Metals 87. H. Stress Corrosion Testing. 115. Rapp. in ASM Handbook. Pugh. p. 36. Alexander. 6 (1980). 21. 31. Sieradzki. McCall. 22. C. B370 (2002). p. 1992. Corrosion: Fundamentals. G. Corcoran. 39 (1922). McKinney and H. Sieradzki. K. Soc. Rowlands. 118. pp. ASTM–AIME. 5. 274. Advances in Corrosion Science and Technology. Testing. Polushkin and H. Staehle. A. Mag. 171 (1981). National Association of Corrosion Engineers. 698 (1967). ASM International. Corros. 1976. Mater. Jones. 3rd edition. J. November (1995). Abrams. 202. Campbell. and Protection. Corros. A. 29. 18. J. Elliott. ASM International. Arnold. Sansone. Min. Vol. Hummer. J. Br. Materials Park. 13. 287–293. 114. p. Appl. 24. D. 28. C. 3. R. and Protection. Electrochem. 32. Johnson and J. 343 (1975). A. Special Technical Publication No. 13A. 15. ASHRAE J. Electrochem. H. 39. Mattsson. 1985. 23. Prot. 122. 20. R. H. Soc. Gupta. Houston. Mattsson. editor. Prot. N. 227–228. R. An Introduction to Metallic Corrosion. 27. Soc. H. 507 (1979). 25. Leja. 14. Movrin. in ASM Handbook. 38. 161. 26. Campbell. Soc. Mater. p. Corrosion 22. 214 (1945). J. McIntyre and C. Metals 77. Eng. Fontana and R. S. p. 279 (1960–1961). E. Takano. in Stress-Corrosion Cracking. 37. Jr. H. Uhlig. Robertson. 34. 61 (1968). U. New York. 21 (1964). Corrosion: Fundamentals. H. A. see also S. Metall. S. J. A 43. Philadelphia. Ref. E. 103. 917–918. Liang. Soc. E. 30. Chem. OH. W. Sci. 7. p. 19. 425. Inst. Materials selection for corrosion control. pp. 380 (1959). Electrochim. Vol. J. and A. R. 747. Shuldener. 2868 (1993). Materials Technology Institute of the Chemical Process Industries. Vasiljevic. R. D. Acta 3. Prot. P. P. Electrochem. Dimitrov. 67. J. J. Beavers. H. G. 4. H. Vol. Electrochem. 2003. Philadelphia. J. K. Bakish and W. Craig. editor. R. C. PA. J. American Society for Testing and Materials. J. Am. Shimodaira and M. p. Forty and G. N. V. W. 1. P. 143 (1968). Electrochem. Philos. Dillon. 13A. 33. 35. J. Edmunds in Symposium on Stress Corrosion Cracking of Metals. M. 320 (1956). Guidelines for Preventing Stress Corrosion Cracking in the Chemical Process Industries. Testing. Soc. and A. Pickering. Effects of metallurgical variables on dealloying corrosion. ASM International. N. and W. B. 1967. 1945. Daniel and R. K. E. pp. Evans. 149. Inst. Erlebacher.. Uhlig and J. 41 (1968). Hough. in Proceedings of Conference Fundamental Aspects of Stress Corrosion Cracking. Trans. Materials Park. Chem. Bailey. 114. Sedriks. 140. Southwell. OH. 17. Wing. 13A. 44. Revie. As in Problem 1. which valence ion predominates when Cu is dissolved anodically? (Data: Cu2+ + 2e− → Cu. 71 (1989). Symposium on Internal Stresses in Metals and Alloys. H. pp. and K. Min. N. Bur. 48(2). from the analogous equilibrium constant. PROBLEMS 1. OH. Corrosion of copper and copper alloys. Stress-corrosion cracking. Stress-corrosion cracking of copper alloys. OH. W. Jones. (Data: Cr2+ + 2e− → Cr. ASM International. Newman. 47. editor. Beavers. 185. Cr3+ + 3e− → Cr. 42.S. H. p. φ° = −0. 45. August 20. Tracy.91 V. φ° = −0. 46. Cu+ + e− → Cu. . in ASM Handbook. A. Cr2+ predominates.PROBLEMS 381 40. 2nd edition. Stand. Cohen. Materials Park. B. Testing. Champion. C.) 2. Scr. ASM International. Magnani. T. Eiselstein. J. Sci. D. Wiley. in ASM Handbook. Thompson and A. 259 (1952). Metall. 23. in Stress-Corrosion Cracking. GENERAL REFERENCES J. London. Corrosion: Fundamentals. OH. 1948. R. 6.74 V. pp. and Protection. Corros. 729–765. Corrosion 36. Caligiuri. 2000. Eng. Military Specification MIL-C-15726 E (Ships). 2005. Sequeira. ASM International. 223 (1983). Vol. 468. Jones. 362.337 V. 211–231. Syrett. and R. F. New York. R. C. A. Corrosion: Materials. φ° = 0. Corrosion 8. Calculate the equilibrium constant at 25°C for the reaction Cu + Cu2+ → 2Cu+. estimate. editor. pp. H. Since equilibrium tends to be maintained between ions and metal. 100 (1949). Vol. C.1 × 10−7. A. in Uhlig’s Corrosion Handbook. Metall. φ° = 0. 1992. 2003. Materials Park. 125–163. L. 23. Natl. Shahrabi. 41. 48.521 V. R. Stewart and F. p. LaQue. 43. which valence ion predominates when Cr in the active state is dissolved anodically. S. Materials Park. Sieradzki. 99 (1952). Logan. 1965. Cu2+ 2. 260 (1980). W. R. Corrosion of copper and copper alloys. A.) Answers to Problems 1. Institute of Metals. 13B. L. Res. U. . It is usually assumed that the passive film is composed of aluminum oxide. Inc. and passivation can be seen from Fig. Uhlig Copyright © 2008 John Wiley & Sons.1 ALUMINUM Al 3+ + 3e − → Al φ = −1. The theoretical potential–pH domains for immunity. which.21 ALUMINUM AND ALUMINUM ALLOYS 21. is estimated at about 2–10 nm (20–100 Å) in thickness. 4. It is very active in the Emf Series (Section 3. its presence is not necessary to achieve passivity. combined with good electrical and thermal conductivity. with the exception of magnesium. but becomes passive on exposure to water. all these impurities. the high-purity metal is much more corrosion-resistant than commercially pure Corrosion and Corrosion Control.71 g/cm3) having good corrosion resistance to the atmosphere and to many aqueous media.4). corrosion. by R.2) of aluminum is active with respect to the hydrogen electrode. tend to be cathodic to aluminum. indicating that the Flade potential (Section 6. The observed corrosion behavior of aluminum is sensitive to small amounts of impurities in the metal.8).3 (Section 4.7 V Aluminum is a lightweight metal (density = 2. 383 . Winston Revie and Herbert H. In general. Although oxygen dissolved in water improves the corrosion resistance of aluminum. for air-exposed aluminum. 1. especially if the high-strength inner alloy should become sensitized through fabrication procedures or by accidental exposure to high temperatures. is usually more resistant than aluminum alloys.2. . Aluminum piping of high-purity metal or of type 1100 has been used satisfactorily for distilled water lines over many years.* Some of the aluminum alloys produced commercially in the United States are listed in Table 21. similar to that provided by a zinc coating on steel.1) including an observed critical potential below which pitting corrosion does not initiate [3. 21. 4]. 2]. in turn. Extraneous anions in dilute chloride solutions act as pitting inhibitors by shifting the critical pitting potential to more noble values [4]. 21. The mechanism is analogous to that described for the stainless steels (Section 19. a high-strength alloy may be sandwiched between pure aluminum or one of the more corrosion-resistant alloys (e. largely to obtain increased strength. which.g. for distilled water or for a water from which heavy metal ions have been removed. The following ions are effective in decreasing order: nitrate > chromate > acetate > benzoate > sulfate. thereby both initiating pitting and. metallurgically bonded at the two interfaces. The clad structure. which all contain traces of heavy metal ions.1.1. The presence of iron as an impurity in lower-grade metal avoids this type of attack or raises the temperature at which it occurs (>200°C for type 1100) [1. aluminum is satisfactory.. therefore.384 ALUMINUM AND ALUMINUM ALLOYS aluminum. it is alloyed. *One exception is known with reference to intergranular disintegration of high-purity aluminum in steam or pure water at temperatures above 125°C (260°F).4. For a complete list. being efficient cathodes. provides cathodic protection to the inner alloy by sacrificial action of the outer layers. the reference given in the table should be consulted.1 ppm) or of Fe3+ in water react with aluminum. In order to make use of the good corrosion resistance of pure aluminum. by galvanic action. stimulating pit growth.1 Clad Alloys Pure aluminum is soft and weak. For this reason. the less noble coatings also protect against intergranular corrosion and stress-corrosion cracking. The copper or iron. 1% Mn–Al) active in the Galvanic Series with respect to the inner alloy. On the other hand.2 Corrosion in Water and Steam Aluminum tends to pit in waters containing Cl−. shift the corrosion potential in the noble direction to the critical potential. aluminum is not a satisfactory material for piping to handle potable or industrial waters. In addition to cathodically protecting against pitting corrosion. Traces of Cu2+ (as little as 0. depositing metallic copper or iron at local sites. particularly at crevices or at stagnant areas where passivity breaks down through the action of differential aeration cells. This combination is called alclad or clad. 10 0.10 0.15 Alloy AA No.2–2.c UNS No.15 0. D. March 2003.10 0. Washington.2 0.. Si + Mg.40–1.10 0. 6-5–6-6. d 45–65% of the Mg content.20–0.7 0.05–0.35 0.3 1. Composition Limits for Some Wrought Aluminum Alloysa Maximum Amounts Specified (%)b Fe 0.1–6. 3XXX.10 0.7 0..10 0.10 0.6 0.9 d 1100 2017 2024 3003 3004 5050 5052 5056 6053 6061 6063 7072 7075 A91100 A92017 A92024 A93003 A93004 A95050 A95052 A95056 A96053 A96061 A96063 A97072 A97075 0.10 0.8 1.2–2. 2XXX.2–1.8 4.30 0.8 0.7 0.10 0.20 0.6 1.20 a b Aluminum Standards and Data 2003. Zn.10 0.5–5. Composition in percent by weight maximum except when composition ranges are given.10 0.20 0.7 0.6 0.05 0.25 0.5 1.0–1.35 0.10 0.15–0.05–0.8–4. Si 0.8 0. 6XXX.8 0.35 0.35 0.04–0.2–0. 7XXX.4 0.25 0.5 3.10 0. Inc.15 0.8–1.8–1. c Code for major alloying element: 1XXX.8 2. Mn.50 0.10 0. 385 .40 0. 5XXX. 4XXX.25 0. The Aluminum Association.7 0.10 1.10 0.2 0.25 0.10 0.30 0.5 0.20 0.40 0.1 0. >99.15–0.40 0.30–0.9 0.10 0.9 1. Mg.25 0.15–0.0 0.1–1.30 0.20 3.ALUMINUM TABLE 21. Cu.C.00% Al.0–1.3 5.95 (Si + Fe) 0.40–0. Si.25 0.1–2.50 0.1–1.35 0.15 0.25 0.05–0.9 0.10 0.8–1.5–4. pp.40 0.10 2.28 Cu Mn Mg Cr Zn Ti 0.1.0 0.45–0.18–0.40–0.7 (Si + Fe) 0.05–0.40 0.50 0.10 0. . For example. in the absence of crevices.). both of which are cathodic to aluminum.E.4). the corresponding weight losses for 16 years were higher—347 and 103 g/m2— and the deepest pits measured 2.386 ALUMINUM AND ALUMINUM ALLOYS In uncontaminated seawater in the tropical environment of Panama.8 mm (0.006–0. At 150°C the rate in distilled water is about 0.84 and 2.079 in. with the attack in all these instances being uniform and without marked pitting. the corrosion rate at 350°C for type 1100 aluminum alloyed with 1% Ni is about 1. which polarized the couple to approximately −0. but the value at 70°C is still very low.H. These field tests support the concept of critical potentials below which.11 in. for 368 days pitted or not according to their corrosion potentials [6]. the corrosion rate of type 1100 in deaerated distilled water increases. whereas those alloys of more active potential (−0. . 0. the total weight losses were 67 and 63 g/m2. equal to −0.0 mm (0. These potential ranges can be compared with the critical pitting potential of pure aluminum measured in 3% NaCl.8% Mg.). and the corresponding observed deepest pits were 0. but usually only after several days of exposure. presumably because a more protective oxide film forms within the higher temperature range.85 V (S.6).).4 to −0.15–0. 0. At 500–540°C. It is important to note that this catastrophic corrosion does not occur at the cited high temperatures if the aluminum is coupled to a sufficient area of stainless steel or of zirconium.004 gmd in the presence of 10−3N H2O2 and still lower by 2 ppm K2Cr2O7.033 and 0. the rate becomes more rapid by a factor of 10–20 times. After 16 years of exposure. This rate is reduced to 0.039 in.) each. Various commercial aluminum alloys immersed in seawater at Key West. At the same location in fresh water. the oxidation rate of pure aluminum in steam is appreciably lower than at 300–450°C [8]. and at 200°C it is 0. for smallsize test panels [5]. 275°C). being about 0. causing rapid failure.04 gmd.E. The same beneficial effect is obtained by small alloying additions of nickel and iron.25 gmd.3.7 to −1.g. and corrosion proceeds mostly along grain boundaries. Alloys with corrosion potentials ranging from −0. respectively.6 V (S.45 V (see Section 6.1). The beneficial effect of cathodic areas can be explained [7] as one of anodic passivation or anodic protection of aluminum in the same manner as alloying additions or coupling of platinum or palladium to stainless steels or to titanium passivate these metals in acids (Section 6. Coupling of susceptible alloy panels to a smaller-area panel of an active aluminum alloy (see Section 13.) (most of which contained some alloyed copper) showed a mean depth of pitting equal to 0. At 315°C. commercially pure aluminum (1100) or 0.2% Cu–Al alloy (6061-T) corroded at decreasing rates with time.007 gmd [1].H.8 gmd [1]. succeeded in largely preventing pitting over the same period of time. probably contaminated with heavy metals.99 mm (0. As the temperature increases. complete disintegration to Al2O3 may occur within a few hours. At higher temperatures (e.6% Si. neither aluminum nor its alloys undergo pitting corrosion. Florida.0 V) were essentially not pitted. ALUMINUM 387 21. Effect of pH on corrosion of commercially pure aluminum (1100). The reason for this difference is that Al3+ is readily complexed by OH−.3 Effect of pH Aluminum corrodes more rapidly both in acids and in alkalies compared to distilled water. when using sulfuric acid for adjustment of pH in the acid region. 1–2 ppm NaCl (J.1 provides data at 70–95°C [1]. 68 ppm CaSO4. Ruther). pH. was adjusted with H2SO4 or NaOH.1. Corrosion rates of aluminum in the alkaline region greatly increase with pH. measured at room temperature. . occurs between pH 4. unlike iron and steel.5. showing that the minimum rate.1. which remain corrosion-resistant. Draley and W. forming AlO− in accord 2 with Figure 21. Most solutions contained 1–10 × 10−5 N H2O2. aerated solutions. with the rates in acids depending on the nature of the anion. At room temperature. the minimum rate occurs in the pH range approximating 4–8.5 and 7. Figure 21. 30 ppm MgSO4. CCl4.1. aluminum surfaces in contact with wet concrete may evolve hydrogen visibly. the most important of these constituents was the fungus Cladosporium resinae [11].e. overprotection must be avoided in order to ensure against damage to the metal by accumulation of alkalies at the cathode surface. Mercury ions present in solution in only trace amounts similarly accelerate corrosion.4 Corrosion Characteristics Aluminum is characterized by (1) sensitivity to corrosion by alkalies and (2) pronounced attack by traces of copper ions in aqueous media. Lime. formation of an aluminum amalgam). The rate of attack can be appreciable in either dilute or concentrated alkalies.g. In addition. whereas for iron a similar reaction forming NaFeO2 and Na2FeO2 requires concentrated alkali and high temperatures.. Subsequent corrosion has been explained on the basis of water-soluble organic acids produced by the metabolism of various microorganisms and also on the basis of oxygen depletion beneath the growing microbial layer (differential aeration cell). In the presence of moisture. Fungi produce highly corrosive organic acids that corrode aluminum fuel tanks. The growth rate is controlled by temperature and is greatest at about 30–35°C. producing intolerably high rates of attack. A drop of mercury in contact with an aluminum surface rapidly breaks down passivity accompanied by amalgamation (i. Until recently.388 ALUMINUM AND ALUMINUM ALLOYS 3 Al + NaOH + H 2O → NaAlO2 + H 2 2 (21. 18].1) This reaction proceeds rapidly at room temperature. . and propylene dichloride) [9. For this reason. the amalgamated metal quickly converts to aluminum oxide. aluminum is subject to rapid attack by (3) mercury metal and mercury ions and (4) anhydrous chlorinated solvents (e. 21. but continues if the concrete is kept moist or contains deliquescent salts (e. The bacteria isolated from aircraft jet fuel tanks have been found to be closely related to Bacillus. Fresh Portland cement contains lime and is also corrosive. Temperature and the availability of water are reported to be the principal factors determining whether and where microbial growth occurs. with the resulting fungus adhering to the tank lining. and they do have the potential to cause microbiologically influenced corrosion (MIC) [11. CaCl2). Ca(OH)2. 10]. causing perforation of piping or sheet. Growth initiates in water deposits at the interface with the fuel. The corrosion rate is reduced when the cement sets.. when aluminum is cathodically protected. 17. and some of the strongly alkaline organic amines (but not NH4OH) are corrosive. ethylene dichloride. The changes in jet fuel and jet fuel additives that have taken place in recent years have also brought about a shift in the microbial community in aircraft fuel. Aluminum tanks in aircraft fuel systems are subject to corrosion resulting from the growth of microbiological constituents in kerosene fuel [11–16].. hence.g. precedes the rapid rate (Fig. 21. An induction time of about 55 min. 21. is about twice that in the anhydrous solvent (Fig. this solution may create new problems.2) forming hexachloroethane.1 Reaction with Carbon Tetrachloride. The reaction with CCl4. Severe accidents have been caused by the rapid reaction of aluminum with anhydrous chlorinated solvents. distilled CCl4 containing 0. following a prolonged induction time. Copyright 1952. during which corrosion is negligible. C2Cl6. On the other hand.99% Al in boiling.2. The Electrochemical Society. 3750 gmd (20 ipy) [19]. Figure 21.3). with evolution of heat.) . or even in the case of aluminum used as a container at room temperature for mixed chlorinated solvents. biocides may selectively enrich a population of bacteria capable of biocide resistance [12].2). The corrosion rate of aluminum in water-saturated CCl4. Mn or Mg (30 h for type 5052). This induction time is lengthened either by the presence of water in the CCl4 (480 min) or by some alloying additions—for example. has been shown to be 2 Al + 6CCl 4 → 3C 2Cl 6 + 2 AlCl 3 (21. for example. for example. (Reproduced with permission.1. The corrosion rate for 99. such as in the degreasing of castings. 21.0011% H2O [19]. Should the temperature reach the melting point of aluminum. addition of AlCl3 or FeCl3 to CCl4 decreases the induction time to zero without an appreciable effect on the corrosion rate.ALUMINUM 389 Although MIC may be controlled by adding a biocide to the fuel.4. Weight loss as a function of time for 99. the reaction may proceed explosively.99% Al in boiling anhydrous CCl4 is very high—for example. in ball milling of aluminum flake with carbon tetrachloride (CCl4). ) In practice. which presumably react preferentially with and destroy free radicals (e. quinones. Effect of water on corrosion rate of 99.390 ALUMINUM AND ALUMINUM ALLOYS Figure 21.. The Electrochemical Society. Water itself may react with and destroy initiating species until the water is consumed by chemical reaction or until sufficient AlCl3 forms. volatile inhibitors are often added to chlorinated solvents in order to suppress the fast reaction. is the following: CCl 4 →iCCl 3 + iCl Al + iCl → AlCl AlCl + CCl 4 → AlCl 2 + iCCl 3 AlCl 2 + CCl 4 → AlCl 3 + iCCl 3 AlCl 3 + CCl 4 → CCl +[AlCl 4]− 3 CCl +[AlCl 4]− → AlCl 3 + iCCl 3 + iCl 3 2iCCl 3 → C 2Cl 6 i Initiation Propagation Cl + iCCl 3 → CCl 4 Termination Subsequent investigations by others have confirmed the presence of ·CCl3 and various of the proposed intermediate reacting species [21. ·CCl3 and ·Cl).g. (Reproduced with permission.3. and amines. Copyright 1952. The latter are considered to play an important part in the reaction mechanism [20]. for example. Examples include various ketones.22]. A proposed sequence of intermediate reactions. Hence.99% Al in boiling CCl4 [19]. the induc- . In summary. In addition to analytical evidence for ·CCl3. vacuum treatment of aluminum specimens results in a shorter induction time of 5 min compared to 55 min not treated. HF. 8. along with the same observed corrosion rate in the vapor phase as in the boiling liquid are a strong arguments that the reaction does not follow an electrochemical mechanism.4). 2. such as HCl and HBr (dilute or concentrated). tartaric. Highest rates of attack are for the boiling 1–2% acid. 2. . as well as strong alkalies. Excellent resistance to rural. Substances like AlCl3 decrease the induction time by forming a complex with CCl4 that later dissociates into free radicals taking part in the chain reaction. Hot or cold acetic acid. Aluminum is resistant to citric. Ninety-nine percent acetic acid is not appreciably corrosive. lesser resistance to marine atmospheres. Sulfur. Lime and fresh concrete are corrosive. and industrial atmospheres. Alkalies. and H2S. Strong acids. Formic acid and Cl− contamination also increase attack. NaOH and the very alkaline organic amines. Fluorinated refrigerant gases. >80% up to about 50°C (120°F) [23] (see Fig. H3PO4. 6. in anhydrous CCl4. Distilled water. 4. but not to methyl chloride or bromide. H2SO4 (satisfactory for special applications at room temperature below 10%). aluminum (type 1100) is resistant to the following: 1. 7. Aluminum is used for mining of sulfur. 21. alloying elements like magnesium are considered to react preferentially with free radicals. but removal of the last 0. Fatty acids. Aluminum equipment is used for distillation of fatty acids. urban. oxalic. and formic. 3. The behavior of alloying constituents in general (often opposite to their effect in aqueous environments). (the species that probably accounts for the red color of CCl4 reacting with Al). Corrosion by soap solutions can be inhibited by adding a few tenths percent of sodium silicate (not effective for strong alkalies). HClO4. Hot or cold NH4OH. 5.5% H2O increases attack over a hundredfold. for example. sulfur atmospheres. By the same token. Nitric acid. probably by reducing the H2O content of the oxide film. Atmospheric exposure. and malic acids. Aluminum is not resistant to the following: 1. Similarly. accounting for the observed improved corrosion resistance of magnesium–aluminum alloys to CCl4 compared to pure aluminum. the observed lack of effect of galvanic coupling or of an impressed nominal voltage on the rate.ALUMINUM 391 tion time is extended when water is present in the solvent. such as Freon. and trichloroacetic acids. the ease with which many added organic substances suppress the fast reaction (free radicals are very reactive) also supports the free-radical mechanism. mine waters or waters previously passing through copper.) 21. 8. Contact with wet woods. 5. 6. Corrosion rates of commercially pure aluminum (1100) in nitric acid. or ferrous piping). can be used in . Any wood impregnated with copper preservatives is especially damaging. see Section 10. especially when trace amounts of heavy metal ions are present. or butyl alcohols at elevated temperatures. in particular beech wood [24].1. Anhydrous ethyl. Chlorinated solvents. 4.g. Pitting occurs at crevices and surface deposits. propyl.4.5 Galvanic Coupling Although cadmium has the nearest potential to that of aluminum in many environments. bolts. A trace of water acts as an inhibitor [23]. brass. 7. (For behavior in soils. Mercury and mercury salts.) 3. Waters containing heavy metal ions (e. and cadmium-plated steel screws. room temperature [23]. trim. and so on.3..392 ALUMINUM AND ALUMINUM ALLOYS Figure 21. Seawater. (Copyright NACE International 1961. ALUMINUM ALLOYS 393 direct contact with aluminum. so that manganese additions would not be expected to counteract the harmful effects of these elements on corrosion behavior. Zn. aluminum may be anodic or cathodic to steel. which act as efficient cathodes. Copper is in solid solution above the homogenizing temperature of about 480°C (900°F) and. Should the alloy be quenched from solid–solution temperatures into boiling water.1). can be coupled to magnesium without damage to either metal [26]. types 2017 and 2024) contain several percent copper. copper.2 ALUMINUM ALLOYS The usual alloying additions to aluminum in order to improve physical properties include Cu. in this way reducing the harmful influence on corrosion of small quantities of alloyed iron present as an impurity [27]. and nickel. because galvanic currents are reduced in the absence of iron. but it is also usually satisfactory. and zinc accelerates the corrosion of aluminum. remains in solution. Zinc is anodic to. cadmium plating is forbidden on most or all parts in many countries [25]. after quenching should it be heated (artificially aged) above 120°C (250°F).7. manganese may actually improve the corrosion resistance of wrought and cast alloys. in seawater) is so high that aluminum may be overprotected cathodically with resultant damage to aluminum. and nickel impurities. High-purity aluminum. In this connection. deriving their improved strength from the precipitation of CuAl2 along slip planes and grain boundaries. cathodically protects. Zinc is somewhat further removed in potential. One reason is that the compound MnAl6 forms and takes iron into solid solution. however. the compound CuAl2 forms preferentially . copper. Precipitation takes place slowly at room temperature with progressive strengthening of the alloy. depending on small composition differences of the water. Si. In fresh waters. aluminum against initiation of pitting in neutral or slightly acidic media (see Section 13. the polarity reverses. In alkalies.g. Magnesium is anodic to aluminum. because of the resulting damage to aluminum..g. and.. Mg. It is imperative that aluminum never be coupled to copper or copper alloys. and Mn. nevertheless. In rural areas. The Duralumin alloys (e. coupling of steel to aluminum is usually satisfactory. on quenching. Of these. No such incorporation occurs in the case of cobalt. 21. Tin coatings are reported to be satisfactory. but the potential difference and the resultant current flow when the metals are coupled (e. it is reported. Aluminum is damaged in this respect to a lesser extent when alloyed with magnesium. or. cadmium is now recognized as a biocumulative poison with toxic effects similar to those of mercury and lead. The compound (MnFe)Al6 settles to the bottom of the melt. hence. but in seacoast areas attack of aluminum is accelerated. it is also important to avoid contact of aluminum with rain water that has washed over copper flashing or gutters because of the damage caused by small amounts of Cu2+ dissolved in such water. makes it more susceptible to this type of failure. To avoid such failures. and a marked susceptibility to intergranular corrosion. and realistic. Hence.C. Simulation in the laboratory is accomplished by controlled intermittent spray with 5% NaCl containing added acetic acid to pH 3 at 35–50°C (95–120°F) [28]. and Mn). Should a Duralumin alloy. It occurs in various types of aluminum alloys in addition to the copper-bearing series. be stressed in tension in the presence of moisture. . in heat-treatment procedures. Proper heat treatment may alleviate such attack. Exfoliation is a related type of anodic path corrosion in which attack of rolled or extruded aluminum alloy results in surface blisters followed by separation of elongated slivers or laminae of metal. accounting for grain boundaries anodic to the grains. it is better practice to aim at a slightly overaged rather than an underaged alloy.5% Mg alloy. a 4. as described previously. Following these guidelines has led to successful use of. 21.1 Stress-Corrosion Cracking Pure aluminum is immune to stress-corrosion cracking (S.. Microstructures that are resistant to S.2% Cr to the alloy [31]. a process that is promoted by addition of 0..5%. it may crack along the grain boundaries. especially when magnesium is added in amounts >4. For structural components in automotive applications. maximum susceptibility was observed at times somewhat short of maximum tensile strength [30]. 34] recommend alloys with a maximum of 3% Mg where there is exposure for long periods to temperatures greater than about 75°C (167°F). are those either with no precipitate along grain boundaries or with precipitate distributed as uniformly as possible within grains [32].C. for example. but at some sacrifice of physical properties.C. the thermal exposure of the part during its lifetime should be established. Cu.g. Sensitizing the alloy by heat treatment.394 ALUMINUM AND ALUMINUM ALLOYS along the grain boundaries. This results in a depletion of copper in the alloy adjacent to the intermetallic compound. the Aluminum Association guidelines [33. slow cooling (50°C/h) from the homogenizing temperature is necessary supposedly in order to coagulate the β phase (Al3Mg2). thereby eliminating susceptibility to this type of corrosion.2.C. The cause in this instance was ascribed to pipe fabrication methods and presence of excess amounts of impurity elements (e.C. severe exfoliation corrosion has been experienced in water irrigation piping constructed of type 6061 alloy [29]. Exfoliation is commonly experienced on exposure of susceptible aluminum alloys to marine atmospheres. If an alloy of higher Mg content is being considered. on the other hand.). In practice. Magnesium alloyed with aluminum increases susceptibility to S. In aging tests at 160–205°C (320–400°F). Fe. alloy 5182. followed by room-temperature aging. In practice.C. fullcomponent testing should be carried out as part of a detailed materials assessment to evaluate and qualify the specific alloy under consideration. Prolonged heating (overaging) restores uniform composition alloy at grains and grain interfaces. the alloy is quenched from about 490°C (920°F). contained in the surface oxide film are sufficient to cause cracking. Edeleanu [37] showed that cathodic protection stopped the growth of cracks that had already progressed into the alloy immersed in 3% NaCl solution.ALUMINUM ALLOYS 395 in structural components of the Honda NSX with no known record of field failures caused by SCC [35. This behavior paralleled that of a Duralumin alloy cited previously. Traces of H2O. and such alloys are now used.C. 36].C. including S. On this basis. and grain boundary precipitates in the initiation and propagation of cracks [40–42]. probably because more grain-boundary area of elongated grains along which cracks propagate comes into play. . 43]. A similar mechanism probably applies to the copper–aluminum alloy series.C. practical structures are sometimes shot peened. specific composition ranges and heat treatments for these alloys are usually chosen with the intent of minimizing susceptibility to S.1). Accordingly.C.C. hydrogen embrittlement.C.C. These conditions. involving several time-dependent interactions that are difficult to evaluate independently. for example. for example. but instead was made up of magnesium atoms segregating at the grain boundary before the intermetallic compound formed.C. on Boeing 777 aircraft [41]. such as the roles of anodic dissolution. susceptibility to S.C. Susceptibility of any of the wrought alloys is greatest when stressed at right angles to the rolling direction (in the short transverse direction). suggest that a possible mechanism of failure may be one of stress-sorption cracking caused by adsorbed water or organic molecules on appropriate defect sites of the strained alloy. Nevertheless.C.C. Edeleanu proposed that the susceptible material along the grain boundary which caused cracking was not the equilibrium β phase responsible for hardness. however.C. decreased on continued aging of the alloy because the separated β phase consumed the original segregated grain-boundary material responsible for susceptibility. hence.C. occurred before maximum hardness values were reached. As mentioned earlier. High concentrations of zinc in aluminum (4–20%) also induce susceptibility to cracking of the stressed alloys in the presence of moisture. carefully baked-out specimens in dry air do not fail [38].. Many high-strength aluminum alloys are available (some are listed in Table 21. maximum susceptibility to S. Service temperatures—especially those above room temperature—that can cause artificial aging sometimes induce susceptibility followed by premature failure in the presence of moisture or sodium chloride solutions. of aluminum alloys is complex.. Understanding has been sufficient. cladding of alloys can serve to cathodically protect them from either intergranular corrosion or S. On aging the alloy at low temperatures. the mechanism of S. plus the susceptibility of a high-strength aluminum alloy (7075) to failure in organic solvents [39].C. Oxygen is not necessary. to permit new aluminum alloys to be developed with high strength combined with improved resistance to corrosion. Compressive surface stresses are effective for avoiding S. Solution heattreatment temperature affects stress-corrosion susceptibility by altering the grain-boundary composition as well as the alloy metallurgical microstructure [42. nor is a liquid aqueous phase required. and J. Stern and H. Dante. Park. (1957). W. pp. T. 105–126. J. 26. 18. 2007. R. 13A. Biofouling 221(5/6). Gilbert and D. and M. T. Corrosion of pure aluminum. Draley and W. Soc. Phillips. Phys.F. Hedrick. (1962). Whitaker. 16. J. J. Corrosion of aluminum alloy 2024 by jet fuel degrading microorganisms. Houston. Metallurgia 68. 381. (1963). T. R. Corrosion 34. Canterbury. McNamara. Maurice. E. p. J. Wolf. 687–690. A. Butterworths. 189t (1958). 27 (1970). Perry. (1978). R. 2. New York. P. 1014. Uhlig. 9. Mitchell. R. Leard. in ASM Handbook. pp. H. Harter. B. Aluminum and aluminum alloys. London. K. Ruther. Materials Prot. Soc. 03568. TX. Bearce. (1952). Paper No. B. D. 257 (2005). 14. Met. Z. T. Chem. M. G. 2nd edition. Electrochem. E. editor. 21. 2003. editor. 1970. 7. J. Electroplated coatings. 5291 (1980). Prot. 157–172. 2000. Ellialtioglu. ASM International. in Microbial Aspects of the Deterioration of Materials. Greenblatt. 8. T. A. Electrochem.-D. 99. 10. Corrosion 17. 9(1). Wiley. J. Corrosion 14. Academic Press. 100. Böhni and H. M. (1969). H. Wiley. 34. 906. Hart and J. and J. NACE International. Corrosion 12. 2000. and C. in Uhlig’s Corrosion Handbook. 87. J. Farmer. 23. 543. Alexander. J. 12. (1952). Stern and H. 19. White. . Microbiologically Influenced Corrosion. Prot. Vol. Revie. Revie. W. Phys. Materials Park. M. American Elsevier. in Microbial Aspects of Metallurgy. A. Soc.396 ALUMINUM AND ALUMINUM ALLOYS REFERENCES 1. Electrochem. Archer and M. Uhlig. London. 27. D. Wolfram. 2nd edition. Elphick. C. 106. H. Mater.. C. Rev. 11. pp. Leard. 80. R. 1961. and J. 222. Videla. 1. Bothwell. Corrosion 39. Jr. and R. (1969). 159. (1983). Cook. 22. CORROSION/2003. 480t (1956). The corrosion of aluminum and aluminum alloys in high temperature water. R. Perry IV. 25t (1961). 4.. Godwin. M. Lovelock. 195–201. 72(10). (1953). 389. Ghali. and Protection. Southwell. 30. and R. 4. 161. Lee. New York. 305. J. 8. in Uhlig’s Corrosion Handbook. L. Little and J. Macmillan. Kaesche. Horst. Mater. 429. Metal Ind. J. pp. New York. Mooney. (1965). Peterson. 923–924. edited by R. R. Chem. Hamstead. T. Microbiological corrosion of metals. Uhlig. Binger. New York. Dante. McNamara. 25. 13. editor. Jr. Pearson and H. Groover. Miller. (London) 2. 25. 15. p. (1965). Testing. E. 26. Mitchell. (1959). 1975. H. MacLennan. 5. D. OH. J. pp. Elder. S. Salvarezza. Electrochem. R. 778. A. 3. D. 2003. C. N. and H. 441t. W. M. J. 6. Gu. Soc. R. R. Mitchell. and T. and W. 17. R. 20. Hummer. de Mele. N. Lennox. 24. J. Ford. Wiley. Corrosion 21. Corrosion: Fundamentals. 116. in Proceedings of the First International Congress on Metallic Corrosion. C. Vol. Houston. TX. P. 1992. Revie. (1954). 33. 7. Yahalom. 132–133. 509. Paxton and R. 1988. 311–345. Agarwala. 1990. Eng. editor. 41. and A. 1991. Inst. in Proceedings of the First International Conference on Environment-Induced Cracking of Metals. Komatsu et al. pp. R. p. Corrosion of aluminum and aluminum alloys. Corrosion 22. ASM International. 311–345. 105–106. 1998. see also U. 38. in ASM Handbook. Metal Progr. J. B. Metall. Publication AT7. (1951). Sprowls. J. 40. in Fundamental Aspects of Stress Corrosion Cracking. 34. J. G. 13B. 13C. Corrosion in the military. 673. in ASM Handbook. Evans. J. TX. National Association of Corrosion Engineers. Corrosion: Materials. Vol. J. Zahavi and Y. in Proceedings of the First International Conference on Environment-Induced Cracking of Metals. 2006. 1969. ASM International. Am. 31. . Metals 80. ASM International. Aluminum for Automotive Body Sheet Panels. 216 (1946). Wiley. Staehle et al. S. Z. W. 1181. pp.. (1981). Sci. (1966). pp. Inst. Corros. 42. Vol. Aluminum Association. Aluminum: The Corrosion Resistant Automotive Material. Corrosion: Environments and Industries. National Association of Corrosion Engineers. Aluminum and aluminum alloys. Corrosion of aluminum components in the automotive industry. 105–106. N. 2000. Houston. Vol. 13C. P. OH. Electrochem. 1988. Society of Automotive Engineers. E. Materials Park. W. Corrosion: Materials. Arnold. New Methods for Corrosion Testing of Aluminum Alloys. 21. The Corrosion and Oxidation of Metals. OH. London. Aluminum Association. Holroyd. 1960. Environment-induced cracking of high-strength aluminum alloys. OH. 2nd edition. P. NACE-10. TX. p. Y. J. N. pp. OH. STP 1134. H. Ghali. editors. 129. pp. 297. Gangloff and M. 166. 2003. Trans. Joshi. Materials Park. 32. C. Metall. Wasserman. Brenner.. Courval. V. H. Lifka and D. G. pp. J. Gangloff and M. in Uhlig’s Corrosion Handbook. Materials Park. National Association of Corrosion Engineers. editors. Publication AT3. 187. H. B. 39. 674. S. 2005. G. editor. New York. Procter. Soc. 65. 43. Environment-induced cracking of high-strength aluminum alloys. in ASM Handbook. GENERAL REFERENCES V. M. B. Holroyd. R. editors. 1990. M. G. 677–716. Houston. Corrosion of aluminum and aluminum alloys. Shastry. R. Technical Paper 910548. 2001. 36. 112. ASM International. 13B. 34. 37. Ives. Philadelphia. Min. Levy. Edeleanu. p. 29. Agarwala and G. NACE-10. 30. 545. Ugiansky. (1942). Ives. in ASM Handbook. Corrosion: Environments and Industries. Kaufman. 35. Robertson. (1982). 2006. R. Materials Park. ASTM.GENER AL REFERENCES 397 28. Kaufman. Application of all aluminum automotive body for Honda NSX. Materials Park. Corrosion of aluminum and aluminum alloys. ASM International. 13A. and Protection. Materials Park. 13B. 1992. ASM International. G. 689–691. R. . ASM International. Snodgrass and J.398 ALUMINUM AND ALUMINUM ALLOYS J. Testing. OH. pp. Vol. editor. Corrosion: Fundamentals. Moran. OH. Vol. OH. 233–249. 2003. Corrosion resistance of aluminum alloys. H. Materials Park. 2003. Stress-corrosion cracking of aluminum alloys. Jones. in ASM Handbook. pp. in ASM Handbook. 95–124. J. in Stress-Corrosion Cracking. Kaufman. Corrosion: Materials. pp. high stiffness/weight ratio. 399 . Winston Revie and Herbert H. including light weight. cases for laptop computers and cameras). high strength/weight ratio. engine blocks). In the original Volkswagen Beetle.. in the automotive industry (transmission casings. Uhlig Copyright © 2008 John Wiley & Sons. magnesium alloys are used in consumer applications (e.34 V Corrosion and Corrosion Control. by R. and ease of machinability [1]. 22. The corrosion behavior of magnesium and its alloys and the principles behind successful use of these materials are discussed in this chapter.2 MAGNESIUM Mg 2+ + 2e − → Mg φ = −2.g.22 MAGNESIUM AND MAGNESIUM ALLOYS 22. it is possible. by appropriate alloy selection combined with good engineering design and correctly applied surface treatment and protection schemes. Today. Inc. to use these materials successfully and obtain numerous benefits. and in the aerospace industry (helicopter transmission and rotor housings). magnesium alloys were successfully used for the engine and for the transmission casings.1 INTRODUCTION Although magnesium and its alloys have long been perceived as rapidly corroding. similar to the situation for aluminum (Section 21. for nickel and copper. Section 4.7 g/cm3) of all structural metals. and zirconium sometimes being used as well. An additional mechanism may involve formation of an insoluble intermetallic compound that includes iron in its structure. above . Magnesium corrodes over a wide range of potential and pH. it is the corrosion resistance of magnesium alloys that is of primary interest. to a lesser extent. 22. This low-density film allows electrolyte to contact the underlying metal. Alloys containing 2–10% aluminum with minor additions of zinc and manganese. This very low density makes magnesium and magnesium alloys useful wherever low weight is important. which is about twice the rate for iron.4. but commercial magnesium corrodes at a rate 100–500 times higher. The potential–pH diagram for magnesium and water (Fig. There are two major alloying systems: 1. 4. passivity. This is heavier than magnesium. Distilled magnesium corrodes in seawater. zinc.400 MAGNESIUM AND MAGNESIUM ALLOYS Several investigators have obtained evidence that magnesium dissolves anodically to a large extent as Mg+ and that this ion in solution reduces water to form Mg2+ plus H2 (see Section 13. magnesium is alloyed with aluminum. having about one-fourth the density of iron and one-third that of aluminum. if not required. the film ruptures because of the higher molar volume of magnesium hydroxide compared to metallic magnesium. respectively (Fig.2). Their effect becomes marked above a critical concentration called the tolerance limit [3]. At high pH. at the rate of 0. and silver.25 mm/y (0. in this way reducing the Fe content of the cast metal. To enhance strength and resistance to corrosion. In addition. and immunity. for example. nickel and copper. Magnesium has the lowest density (1.01 ipy). 22. Corrosion resistance of magnesium depends on purity of the metal even more than the corrosion resistance of aluminum depends on its purity. Additions of manganese or zinc to the metal raise these limits to higher values.017%. it settles to the bottom of the melt. these alloys are used up to 95–120°C (200–250°F). similarly.0005% and 0. resulting in the constant exposure of fresh metal [2].3 MAGNESIUM ALLOYS Because unalloyed magnesium is not used extensively for structural applications. hence. Manganese has been reported to surround iron particles and to render them inactive as local cathodes [4]. the magnesium hydroxide corrosion product film is only partially protective. manganese. thorium.6). with minor additions of cerium. The impurities in commercial magnesium largely responsible for the higher rate are iron and. rhenium.1). For iron.5) indicates the theoretical domains of corrosion. Magnesium is also the most active metal in the Emf Series that is used for structural purposes.1%. it is 0. the tolerance limit is 0. with visible hydrogen evolution. in the region identified as passive. lithium. and J. and containing zirconium. alternate immersion.MAGNESIUM ALLOYS 401 Figure 22. these alloys provide improved elevated temperature properties compared to those in the first group. but not aluminum. Nelson. Corrosion of magnesium in 3% NaCl. thorium. Hanawalt. . 147. zinc. 2. Peloubet. which provides grain refinement and improved mechanical properties. and silver). showing tolerance limit for iron and beneficial effect of alloyed zinc and manganese [Figure 4 from J. 281 (1942)]. Trans. Alloys containing various elements (rare earths. Metals Div. 16 weeks. Inst.1. Corrosion studies of magnesium and its alloys. AIME. C. which temperature range both corrosion resistance and strength deteriorate. and susceptibility increases with Al concentration. however. For this reason. Design to prevent galvanic corrosion is.C. letters that are sometimes appended to the alloy designation are assigned chronologically and usually indicate alloy improvements in purity [5]. thorium. urban. In the ASTM system for designating magnesium alloys. Z.C.g.1 Stress-Corrosion Cracking The magnesium alloys with greatest susceptibility to stress-corrosion cracking (S. in magnesium alloys is usually transgranular with secondary cracks. Magnesium–zinc alloys have intermediate susceptibility.C. Stress-corrosion cracking of magnesium alloys occurs in many environments.C. for example.C. therefore. The letter corresponding to the element present in greater concentration is listed first.C. dissolved oxygen accelerates S. Alloy purity. and deaeration retards it. rare earths. seawater. does not prevent galvanic corrosion that occurs if the magnesium alloy is coupled. In saline solutions. Although several mechanisms for S. attributed to a grain-boundary phase of Mg17Al12 that causes galvanic attack of the adjacent matrix [8]. essential. and coastal areas. 22. rounded off to the nearest whole number. the first two letters indicate the principal alloying elements: A. hydrogen embrittlement is most likely a factor [8]. occurs. E. including distilled water. if the elements are in equal concentration. In addition. the corrosion rate is controlled by the concentration and distribution of the critical elements. and atmospheric environments in rural.g. iron or steel) has been found to reduce the galvanic corrosion of the magnesium to one-tenth the rate.C. which create cathodic sites of low hydrogen overpotential.402 MAGNESIUM AND MAGNESIUM ALLOYS Some alloys in both systems that are commonly available are listed in Table 22. to a steel bolt. All commonly used metals cause galvanic corrosion of magnesium in saline solutions. silver. Welded components of magnesium alloys are normally stressrelieved after welding to reduce the residual stresses that develop during welding.C.. aluminum. K. results in an even greater reduction in the galvanic corrosion rate. H. manganese. As in most systems in which S. both applied and residual stresses are important. or branching. changing from 3% NaCl to tap water). and initiation usually occurs at pits. a reduction in the conductivity (e. for example. industrial. zirconium.C. S. AZ91 indicates the alloy Mg–9Al– 1Zn. Q.C. Intergranular S.C. In addition. nickel.1. M. these elements are usually controlled at low impurity levels.) are the Mg–Al alloys..3. and alloys that contain neither aluminum nor zinc are the most resistant [7]. zinc. and copper. they are listed alphabetically. In addition to the anode/cathode area ratio. In water. of magnesium alloys has also been reported. the conductivity and composition of the medium in which a couple is immersed are controlling factors in the rate of galvanic corrosion. . of magnesium alloys have been proposed.C. iron. The letters are followed by numbers that indicate the nominal compositions of these alloying elements. but zinc plating the cathode (e. Resistance to S.5 — — UNS M15100 M10500 M10600 M11310 M11610 M11630 M11800 M11910 M12331 — M13310 M13320 M18220 M18210 M16410 M16630 M16400 M16600 Product Formb W C C W W C C. or chromates [11]. 10]. .1. may also be increased by using shot peening and other processes that produce compressive residual stresses at the surface [8]. Molded plastic caps were also found to be effective.MAGNESIUM ALLOYS 403 TABLE 22. in helicopter drive system components.5 — — — — Zr — — — — — — — — 0.2 0.5 — 0.5 5. wrought alloy.5 0. W C C C C C C. heads of bolts made from AZ91D were protected from galvanic corrosion by using a nylon coating that was electrostatically applied. 80Sn–20Zn) electroplated on magnesium and given a chromate treatment were found to decrease the corrosion rate of magnesium by more than 90% in salt spray testing [9.5 0. in aerospace applications where extended component life and low maintenance costs are required.0 Mn 1.5 0.3 0.5 — — — 2 1 1. In light-truck applications.C. W C C W C. C. 6.5 0.2 0. W Adapted from Ref.2 0. Zinc alloy coatings (e.g. for example. Anodized magnesium castings are used successfully.5 4.2 0.3.0 6. 22. In general. Anodizing treatments result in porous coatings that provide excellent bases for subsequent painting.5 0. Nominal Compositions of Some Magnesium Alloysa Alloy Number ASTM M1 AM50 AM60 AZ31 AZ61 AZ63 AZ80 AZ91 EZ33 ZM21 HK31 HZ32 QE22 QH21 ZE41 ZE63 ZK40 ZK60 a b Element (%) Al — 5 6 3 6 6 8 9 — — — — — — — — — — Zn — — — 1 1 3 0.5 1 2.5 Th — — — — — — — — — — 3 3 — 1 — — — — Re — — — — — — — — 2.1 2 — — 4.C.2 Coatings Coatings provide an important strategy for protecting magnesium alloys from corrosion. all surfaces of magnesium alloys exposed to corrosive environments must be coated. Various protective anodized coatings are available as produced in electrolytes composed mainly of fluorides.5 2 0. phosphates.5 0. W.5 0. W C.5 2.5 0.2 — 1 — — — — — — — — Ag — — — — — — — — — — — — 2.5 2.2 0. cast alloy.. and so on. the automotive industry is reducing vehicle weight to the extent possible.094 in.4 SUMMARY Limitations in the use of magnesium alloys for structural parts are as follows: 1.) per year in the temperate marine atmosphere of Kure Beach. hence. compared to 0.2% Mn alloy in 5% HF is 2. .1 mm (0.05 gmd. 1. Pure aluminum is also satisfactory. The rate in 48% NaOH−4% NaCl is 0. avoiding subsequent high applied stress in order to avoid stress-corrosion cracking in moist atmospheres. clutch and brake pedal supports.2 gmd (high-purity Mg). The corrosion rate of 8% Al. types 5050. for example. One of the more recent developments in the automotive industry is the application of Alloy AM50 for the front-end support assembly for light-duty Ford trucks [12]. retractable headlight assemblies. 2. Sensitivity to stress-corrosion cracking in moist air.g. 0. Magnesium. North Carolina [13]. Tendency of magnesium and its alloys to corrode when galvanically coupled to other metals. including protection against galvanic corrosion. Aluminum alloys containing some magnesium (e. Fasteners that are specially designed using nylon or plastic washers. Alkalies. A protective film of magnesium fluoride forms. 2. for underbody and wheel applications. engine compartment grills. 4. can be applied in contact with magnesium.004 in. the rate increases appreciably..3 gmd. and clutch and transmission housings. Large heavy automobile parts are particularly attractive candidates for replacement by magnesium. Pitting may occur at the water–air interface. Distilled water (same precautions as for 1). in 48% HF. is resistant to the following: 1. The controlled purity 3% Al. >2% HF.404 MAGNESIUM AND MAGNESIUM ALLOYS Driven by the need to reduce both gasoline consumption and greenhouse gas production. Most alloys must be stress relieved. however. magnesium must be insulated from all dissimilar metals. insulating washers must be used under bolt or screw heads in order to extend the path of electrolyte resistance and thereby minimize galvanic action.4 mm (0. 22. it is 0.) per year. in brief. magnesium is resistant to alkalies. Underbody wax-type coatings may provide additional protection. Unlike aluminum. Atmospheric exposure. comprehensive protection schemes are essential. In general.5% Zn–Mg alloy exposed for eight years to the tropical marine atmosphere at Panama lost 2. and one approach being used is increasing the magnesium content of automobiles. Magnesium alloys are also being used for air cleaner covers. 5052. Above 60°C (140°F). are used to help control galvanic corrosion of magnesium [9]. 3. The extent of corrosion protection required depends on the severity of the environment. sleeves. and 5056) are least affected by alkalies generated by a magnesium–aluminum couple and. Alexander. Prot. ASM International. Waters containing traces of heavy metal ions. 3. 1966. Emley. J. Eng. R. 793–830. Testing. D. Materials Park. p. turpentine. B. Corrosion: Materials. pp. 793. Metals Div.. 205. 2000. Revie. pp. 147. . Ref. R. Stress-corrosion cracking of magnesium alloys. Inst. Hanawalt. New York. 2. 1999. Peloubet. A.g. 1992. or dimethylglyoxime (DMG). 3.. Ref. pp. New York. SAE International. H. 692–705. L. 8. 2. Magnesium and Magnesium Alloys. 2. Wiley. ASM International. Materials Park. C. 30 (1965). Detroit. 9. 7. E. Ref. A. Leaded gasoline mixtures. Michigan. 4(12). in Uhlig’s Corrosion Handbook. Shaw. Revie. p. editor. in Stress-Corrosion Cracking. C. 12. Magnesium and magnesium alloys. 5. Corrosion of magnesium and magnesium-base alloys. Seawater. 1995. 251–263. Magnesium and magnesium alloys. 13. 273 (1942). Avedesian and H. New York. ASM Specialty Handbook. 11. pp. in ASM Handbook. The anhydrous gas is not corrosive. H2CrO4. W. E. 5. 693. R. 1. in Uhlig’s Corrosion Handbook. 2nd edition. Metall. Inorganic or organic acids and acid salts (e. Materials Park. editors. Miller. Ruden. Jr. and Protection. Jones. Corrosion resistance of magnesium alloys. editor. p. OH. NH4 salts). and J. Galvanic compatibility of coated steel fasteners with magnesium. GENERAL REFERENCES M. 6. OH. 2003. Principles of Magnesium Technology. Min. H. 222–223. 10. 218–219. editor. B. REFERENCES 1. 1. in ASM Handbook. 2. 2003. paper no. ASM International. 2nd edition. Corrosion 10. 4. OH. 950429. magnesium methylate forms. and C. p. 796. 813. 2000. Freon (CCl2F2) plus water. The higher alcohols are resisted satisfactorily. p. W. International Congress and Exposition. Hawke and T. Hummer. The reaction can be inhibited by (NH4)2S. PA. Ref. George. Ghali. Nelson. M. Methanol (anhydrous). Robinson and P. K. Materials Park. W. Warrendale.GENER AL REFERENCES 405 Magnesium is not resistant to the following: 1. 13B. pp. E. Shaw and R. ASM International. Ghali. Pergamon Press. 4. Corrosion: Fundamentals. Vol. A. 6. 13A. Southwell. 182 (1954). OH. Wiley. Mater.. Wolfe. C. Baker. Vol. TMS. editor. O. Powell. in ASM Handbook. Vol. OH. W. W. and B. Warrendale. OH. Pekguleryuz and L. 2005. H. Butterworth-Heinemann. B. K. TMS. Luo. in Stress-Corrosion Cracking. R. 251–263. Materials Park. Corrosion: Fundamentals. PA. I. 2006. Corrosion resistance of magnesium alloys. H. 692–696. 2003. Magnesium Technology in the Global Age. pp. Lightweight Electric/Hybrid Vehicle Design. Mackenzie. PA. R. A. F. Neelameggham. 1992. editor. Shaw. Canadian Institute of Mining. Canada. Oxford. Metallurgy and Petroleum.406 MAGNESIUM AND MAGNESIUM ALLOYS R. Warrendale. Jones. ASM International. 2001. . Materials Park. Fenton. 2004. 13A. and Protection. Montreal. M. N. Alan A. SAE International. Hodkinson and J. Magnesium Technology 2005. Stress-corrosion cracking of magnesium alloys. Kaplan. R. Miller. pp. Magnesium Technology 2004. Testing. ASM International. 407 . Uhlig Copyright © 2008 John Wiley & Sons. it can be alloyed with many other elements while maintaining its ductile face-centered-cubic structure. Winston Revie and Herbert H. power production.1 INTRODUCTION Nickel has many applications in industry where corrosion resistance is required. energy conversion. Inc. These new nickel-base alloys were designed by developing and applying fundamental knowledge on the role of individual alloying elements and by improving thermomechanical processes to optimize alloy metallurgy. ternary. Corrosion and Corrosion Control. first. petrochemical. Nickel is an excellent base on which to design alloys because. As a result. Today. Alloy selection from the large number of alloys available depends on the specific properties that are required.2 list the major alloying elements and the effects that they have on properties of alloys intended for aqueous corrosion and for hightemperature application. the element itself is moderately corrosion resistant. over the past 100 years. and many other industries. second. Tables 23.23 NICKEL AND NICKEL ALLOYS 23. and. supercritical water. by R. and other nickelbase alloys were developed to meet the increasingly demanding requirements in many industries for resistance to aqueous corrosion and to high-temperature oxidation conditions. respectively [1]. pharmaceutical. a wide range of binary.1 and 23. pulp and paper. nickel-base alloys are used in chemical process. waste incineration. Provides resistance to reducing (nonoxidizing) corrosive media. Fe Because of the increased cost of these alloys compared to stainless steels. Detrimental to thermal stability in high Ni–Cr–Mo alloys. which shows. Provides solid solution strengthening. Si High silicon (>4%) improves resistance to oxidizing mineral acids. Enhances localized corrosion resistance. Other Features Enhances corrosion resistance in mild reducing media and in alkali media. 23.250 V Nickel is active in the Emf Series with respect to hydrogen.2 NICKEL Ni 2+ + 2e − → Ni φ = −0. Detrimental in certain corrosive environments. and passivation of nickel at 25°C are shown in Fig. Ta Carbon stabilizer.1. Provides solid solution strengthening. Provides resistance to oxidizing corrosive media. they are used when stainless steels are not suitable or when product purity and/or safety are critically important [1]. The theoretical potential–pH domains of corrosion. Austenitic stabilizer. Improves HAZ and intergranular corrosion resistance.C. Ti. for example.408 NICKEL AND NICKEL ALLOYS TABLE 23.C. detrimental to thermal stability in higher Ni–Cr–Mo alloys. thermal stability. immunity. Improves chloride stresscorrosion cracking (S. Alloying Elements and Their Major Effects in Alloys for Aqueous Corrosion Resistance Alloying Element Ni Main Features Provides matrix for metallurgical compatibility with alloying elements. 23. Reduces cost by replacing nickel and enhances scrap utilization. Cr Mo W N Behaves similarly to Mo. sulfuric acid.1 [2]. Improves resistance to seawater and sulfuric acid. Cu Enhances localized corrosion resistance. and nitric acid. but noble with respect to iron.). Nb.C. Improves thermal stability and fabricability. Enhances localized corrosion resistance and chloride S.C. . Enhances resistance to environments containing H2SO4 and HF. and mechanical properties. but less effective. Provides matrix for metallurgical compatibility with various alloying elements. economical substitute for nickel. detrimental to oxidation resistance. improves sulfidation resistance. for example. Nickel does not react rapidly with dilute nonoxidizing acids (e. Ni Y. Improves strengthening by age hardening. Reduces chlorination resistance. oxidation resistance adversely affected. improves creep strength. Co Increases solubility of carbon in nickel-base alloys. detrimental to sulfidation resistance. Re that nickel can be cathodically protected in acid or neutral solution by controlling the potential below about −0. increasing resistance to carburization. carburization resistance. sulfidation and carburizing resistance. detrimental to nitriding and fluorination resistance Improves oxidation. Improves adherence and spalling resistance of oxide layer.H. Improves strength. Increases short-term creep strength. detrimental to nitriding resistance. that on chromium (for . improves oxidation. but the passive film is not as stable as. hence. helps nitriding resistance. It is thermodynamically inert to (will not corrode in) deaerated water at room temperature. sulfidation. and chlorination resistance. Improves carburization. and hence. detrimental to oxidation resistance at higher temperatures. Slight positive effect on high temperature strength and creep. detrimental to nonoxidizing chlorination resistance. improves solid solution strength. beneficial to carburization resistance. Al Mo W Nb C Ti Mn Detrimental to nitriding resistance. increases solubility of nitrogen. Independently and synergistically with Cr improves oxidation resistance. improves age hardening effects.2. Helps improve sulfidation resistance. detrimental to nitriding resistance. in which Ni(OH)2 forms as a corrosion product. H2SO4 and HCl) unless dissolved oxygen is present. Similar to molybdenum. Synergistically acts with chromium to improve high temperature degradation. nitriding. Improves high temperature strength.g.4 V versus S. Alloying Elements and Their Major Effects in High Temperature Alloys Alloying Element Cr Si Main Features Improves oxidation resistance.E.. nitriding. Reduces rate of sulfur diffusion. Other Features Improves sulfidation and carburization resistance. It is passive in contact with many aerated aqueous solutions. may be beneficial in carburizing.NICKEL 409 TABLE 23. Copyright ANCE International 1974 and CEBELCOR.410 NICKEL AND NICKEL ALLOYS Figure 23. Only silver and possibly zirconium are more resistant. a soluble complex. Nickel is outstandingly resistant to hot or cold alkalies. or with sulfates in the presence of organic or other chemically reducing impurities. . and passivation for nickel at 25°C [2]. It also resists fused NaOH. When subject to alternate oxidizing and reducing atmospheres at elevated temperatures. Nickel is attacked by aerated aqueous ammonia solutions. immunity. lowcarbon nickel being preferred for this application in order to avoid intergranular attack of the stressed metal. It corrodes by pitting when exposed to seawater (passive–active cells). 5 min at 875°C (1100°F) is advisable.2 V ) [3]. (M. p. Pourbaix. Fused salts contaminated with sulfur. Atlas of Electrochemical Equilibria in Aqueous Solutions.0001 ipy). 2nd English edition. It also fails along grain boundaries in the presence of sulfur-containing environments above about 315°C (600°F). stress relief anneal. may damage nickel or high-nickel alloys in this manner. It is also attacked by strong hypochlorite solutions with formation of pits. forming as a corrosion product.06 gmd (0. Nickel has excellent resistance to oxidation in air up to 800–875°C (1500– 1600°F) and is often used in service at still higher temperatures.) nickel. Nickel is not subject to stress-corrosion cracking except in high concentrations of alkali or in fused alkali. Flade potential φ F = 0. Nickel exposed to boiling 50% NaOH corrodes at the rate of 0.1. Theoretical potential–pH domains of corrosion. 2+ Ni(NH 3)6 . 334. it tends to oxidize along grain boundaries. g. composed of basic nickel sulfate (fogging). HNO3). Fogging is minimized by a thin chromium electroplate over nickel. Sulfur or sulfur-containing reducing environments. There is good resistance to oxidation in air at elevated temperatures.g. Hence. iron).75% Fe. Above about 60–70 at. In industrial atmospheres.. 3. HNO3 and H2CrO4) by supporting the passivation process. 6. bal. 2.1 General Behavior Adding copper to nickel improves. and deep pits form when the alloys are . aerated conditions that ensure uniform passivity. are used for seawater condenser tubes.3. 5. nickel is resistant to the following: 1. is best used under rapidly moving.8. however. The 70% Ni–Cu alloy (Monel). Also. a nonprotective film forms. nonoxidizing acids. including hydrofluoric acid). Alkalies. FeCl3. in line with the resistance of copper to pitting. The critical minimum chromium content [4] obtained from critical current densities for anodic passivation in sulfuric acid is 14 wt. It does not pit under conditions that provide cathodic protection. including fused alkalies. and pits tend to be shallow rather than deep.. 10–30% Ni. which still further improves resistance to impingement attack.3 NICKEL ALLOYS 23. Oxidizing acids (e. These alloys are more sensitive than nickel to attack by Cl− and by HCl.. on the other hand. pits in stagnant seawater and. the alloys lose the passive characteristics of nickel and tend to behave more like copper (see Section 6. CuCl2.. Resistance is improved if acids are deaerated. Cu alloys (cupro-nickels) do not pit in stagnant seawater and are resistant to corrosion in rapidly flowing seawater. the tendency of nickel–copper alloys to form pits in seawater is less pronounced than for nickel. resistance under reducing conditions (e. such as when the alloy is coupled to a more active metal (e. Seawater. Dilute nonoxiding inorganic and organic acids. appreciably improved resistance to impingement attack.% Cu (62–72 wt. in seawater. 23. and K2Cr2O7). >315°C (>600°F).1) retaining.g.NICKEL ALLOYS 411 In summary. Oxidizing salts (e..% Cu).% Cr. Nickel is not resistant to the following: 1. to a moderate degree. 3.g. Additions of chromium to nickel impart resistance to oxidizing conditions (e. The atmosphere. 2. Aerated ammonium hydroxide. Alkaline hypochlorites. 4. Such alloys containing a few tenths to 1.g. hot or cold. therefore. hence. all lie within 2 mV of a platinized platinum electrode in the same solution [5]. or chromium are given in Table 23.8% Mo in 5% H2SO4. poor workability). iron. . tend to be more active than their Flade potentials [5]. By alloying nickel with both molybdenum and chromium. but it is not improved with respect to oxidizing media (e. These are also difficult to work. One such alloy. tungsten behaves similarly to molybdenum.4). Because the binary nickel–molybdenum alloys have poor physical properties (low ductility. corrodes at 1/12 the rate of nickel in deaerated 10% HCl at 70°C. sulfuric acid over a wide range of concentrations. 80% Ni (see Section 11. Alloys of this kind. such as in hydrochloric acid. and the Ni–Mo–Fe and Ni–Mo–Cr alloys are difficult to work or fabricate. but increases anodic polarization instead. Notwithstanding an active corrosion potential. whereas the Ni–Cr alloys are less readily. Chromium also imparts to nickel improved resistance to oxidation at elevated temperatures. One important commercial alloy contains 20% Cr. they do not pit in the strong acid media to which they are usually exposed in practice. The Ni–Cu alloys are readily rolled and fabricated. and the rate decreases still further with increasing molybdenum content (1/100 the rate for nickel at 25% Mo).13. the alloys are not passive in the sense of Definition 1 (Section 6. Polarization measurements show that molybdenum alloyed with nickel has little effect on hydrogen overpotential. however. Alloying nickel with molybdenum provides. for example. As an alloying element in nickel. For example. The mechanism of increased anodic polarization is related most likely to a sluggish hydration of metal ions imparted by molybdenum. Resistance of such alloys to hydrochloric and sulfuric acids is better than that of nickel. HNO3). the corrosion rate of the alloy is anodically controlled. which also contains a few percent iron and tungsten (Alloy C). or alternatively to a porous diffusion-barrier film of molybdenum oxide.1). for example.412 NICKEL AND NICKEL ALLOYS exposed to stagnant seawater. The corrosion potentials of such alloys in acids. Since the Ni–Mo–Fe alloys have active corrosion potentials and do not. improved resistance under reducing conditions. the alloy containing 15% Mo. They are designed to resist.3. whether aerated or deaerated. hence. but they mark an improvement over the binary alloys. molybdenum. an alloy is obtained resistant to oxidizing media imparted by alloyed chromium. other elements. hydrogen-saturated. as well as to reducing media imparted by molybdenum. rather than to formation of a passive film typical of chromium or the passive chromium–nickel alloys. among other media. establish passive–active cells. but is less effective [6]. Some commercial Cr–Ni–Fe–Mo alloys corresponding in composition to high-nickel stainless steels also contain a few percent copper. the corrosion potentials of nickel alloys containing 3–22. despite improved resistance to Cl−.g. Nominal compositions of some commercial nickel-base alloys containing copper. are added to form ternary or multicomponent alloys. is immune to pitting and crevice corrosion in seawater (10-year exposure) and does not tarnish appreciably when exposed to marine atmospheres. corrode more rapidly in hydrochloric acid than do the nickel–molybdenum alloys that do not contain chromium.. to a marked degree. bMinimum.5 0.0a — — — 1.35a Al: 0.1 0.5a 19.4 Cu: 0.5 — — 21.5 1.015 0.5 1.5a Cu: 0.5 — a UNS No. 32 58. Ni–Fe–Cr 825 N08825 42 a Maximum.5 0.0 1.12 0.5 Ni–Cr 29.0b Ni–Cr–Fe– Mo–Cu G G-3 N06007 N06985 Bal. Bal.5 2. Bal.5 1. ASM International.5 — — — 6.5 8 Bal. Source: Adapted from Metals Handbook.TABLE 23.0 3.5 0.0a 2.5 0.2 0.0 6.2 Al: 0.0 1.5 5. Bal. Al: 0.5 21 30 1.0 V: 0.02 0.0 0.9 Ni–Mo B B-2 B-3 N10001 N10665 N10675 Bal.5 2. 10 a Common Name of Alloy — 1.5a 3. 413 .5 0.0a 1. ASM International. Desk Edition.38 Cu: 0.5 0.35a V: 0.35a V: 0.02 0. Ti: 0.01 0.0 — 0.5 — — — 2.5a Cu: 31.4.0 3.5 1.5 Ni Ni–Cu 400 N04400 66.5 16 22 16 23 22. Materials Park.5 1. OH.01 0. Nb: 2.08 0.0 15.05 0.2 0.05 0.5a Cu: 0. Ti: 0.25 Al: 0. 2001.5a Cu: 2.0a 1. Bal Bal.0 22. For detailed alloy specifications.6 Al: 0.0 Cu: 2.0 0.015 0.0 2.0 1. Materials Park.0b 0.5a 4 3 4 — 6 3 5 1.3 9 5.08 0. for example.3.4.2a.0. 1998.0 2. p.1 0.0 1. 4th edition.5 19. Ni–Cr–Mo C C-22 C-276 59 N10002 N06022 N10276 N06059 Bal. 610.05 0. 2nd edition.1 0.08 0. OH.5 2.0 0.08 0.5 7 16 13 16 16 3 625 N06625 61 Nb + Ta: 3. Typical Compositions (%) of Commercial Nickel-Base Alloys Mo Cr W Fe Other Coa Sia Mna Ca NICKEL ALLOYS Alloy System — 0. — — 2.15 0.1 — 22. 65b Ni–Cr–Fe 600 800H 690 N06600 N08810 N06690 Bal. Worldwide Guide to Equivalent Nonferrous Metals and Alloys.0 0.5 28 28.0 1.0 0. see. Ti: 0.5 21.0 1.0a 1.2a. but is not as good. in moist aerated hydrofluoric acid and in hydrofluorosilicic acid vapor. many nickel– copper and copper–nickel alloys are possible. Boiler water in contact with the outside of the tubes is all-volatile treated (minimum dissolved solids and dissolved O2.3 Ni–Cr–Fe System: Alloy 600—76% Ni. 30% Cu Because nickel and copper in all proportions form solid solutions. Hence.001 ipy). It is outstandingly resistant to unaerated HF at all concentrations and temperatures up to boiling. and NH3. Typically.13.2 Ni–Cu System: Alloy 400—70% Ni.g. SO2. In nitric acid. 3. 120°C (248°F) is 0. and HNO3]. The alloy is susceptible to S.g. except in concentrated hot caustic solutions. reactor coolant circulates through the tubes at 315°C (600°F) exiting 30–35°C lower in temperature. Since Alloy 400 is resistant to high-velocity seawater. HNO3. it is often used for valve trim and pump shafts. to accelerate the cathodic reaction (H+ or O2 reduction) to the point where the anodic current density reaches or exceeds the critical value for anodic passivation. CuSO4..C.4).008 ipy)(23-h test) [8]. of tubes tend to occur at inlet-end sections above the tube sheet in crevice and sludge buildup areas [11]. 16% Cr. but susceptibility can be minimized by deaeration of the environment and by stress relief annealing the alloy component [10]. this alloy tends to pit in seawater and also in oxidizing metal chlorides..C. namely. Fe2(SO4)3.414 NICKEL AND NICKEL ALLOYS The function of alloyed copper is the same as that of alloyed palladium in titanium mentioned in Section 6.025 mm/y (0. It is also resistant to all concentrations of NH4OH at room temperature. [Rate in N2-saturated 35% HF.8 mm/y (0.C. Br2. mine waters. 23.4. Resistance to alkalies is good. such as FeCl3. which is still widely used today. its corrosion behavior is similar to that of nickel in many ways.C. It is also used for industrial hot freshwater tanks and for equipment in the chemical industry. With about 31% copper in a nickel matrix. Pressurized water reactors of nuclear-power plants use mill-annealed Alloy 600 tubes for heat exchangers. slightly alkaline using NH3). The first commercial nickel–copper alloy was Alloy 400 (commonly known as Monel). and H2CrO4) nor to wet Cl2. Like the stainless steels. 23.] Resistance to alkalies is good. in general. laboratory tests of stressed alloy exposed to hot NaOH . CuSO4. Washings from such areas test alkaline and are high in sodium.3. Alloy 400 is not resistant to oxidizing media (e. when air saturated. 7% Fe Alloy 600 resists oxidizing aqueous media [e.3. except in hot concentrated caustic solutions or aerated NH4OH.15 ipy) [9]. It resists boiling sulfuric acid in concentrations less than 20%. FeCl3.20 mm/y (<0. including red fuming acid. resistance is best above 20% HNO3. Alloy 600 is used extensively where oxidation resistance at elevated temperatures is required (see Section 11. with the corrosion rate being <0. but improved relative to nickel in other ways [7]. as the stainless steels. Thinning and intergranular S. C. 23.8 mm/y. 15] (see also Section 19. resistance is also good. Alloy 690.2 mm/y. it is subject to intergranular attack in nonoxidizing acids (e. Through heat treatment.046%) [13]. with the highest rate for the pure acid applying to the boiling 86% H3PO4 (0. acetic.C. But whether or not chromium depletion occurs is not related to the cause of intergranular S. Alloy 600 fails in 10% NaOH at 315°C (600°F) independent of carbon content (0.3. [16]. but containing only 0. followed by rapid cooling in air or water. in NaOH solutions [12].03 ipy).C. Heat treatment to avoid such attack consists of annealing at 1150–1175°C (2100–2150°F). The rate of corrosion is 0. In phosphoric acid.02 ipy) in boiling 20% HCl. and hydrochloric) when heated in the range 500–700°C (930–1300°F).23 mm/y (0.1% NaCl [14].009 ipy) in boiling 10% HCl.C. and an alloy (18% Cr. Alloy B and commercial alloys of similar composition are resistant to hydrochloric acid of all concentrations and temperatures up to the boiling point. Accordingly. hot or cold.NICKEL ALLOYS 415 solutions (290–365°C) are used as an accelerated measure of resistance to S. In various organic acids.5 mm/y (0.007 ipy). The improved resistance also extends to an environment of relatively pure water at 345°C (650°F) [12]. 0.C.05 mm/y (0. The . Both phosphorus and boron have been suspected [13.C. 5% Fe The original Alloy B (N10001) was invented in the 1920s.3. The extent of S. Nickel–molybdenum alloys must not be used where oxidizing conditions exist.006–0. in boiling MgCl2 solutions.C. however. formic. Contamination of nonoxidizing acids with oxidizing ions (such as Fe3+ or Cu2+) causes a large increase in corrosion rate [19].C. The long incubation time typically required for cracks to initiate in pure water (several months) supports the view that specifically damaging elements must reach a critical concentration by slow diffusion to grain boundaries in order for the alloy to become susceptible.C. 77% Ni) approximating the Alloy 600 composition. as may occur in the heat-affected zone (HAZ) of a weld. Alloy 600 is resistant to S. of the alloy in water at 350°C is unaffected by. has been found to provide satisfactory service. and remains intergranular in. discontinuous carbide precipitates form at grain boundaries accompanied at 650°C (but not at 700°C) by chromium depletion. 0.4 Ni–Mo System: Alloy B—60% Ni.C. rate of attack is low for all concentrations and temperatures. <0.2. 30% Mo. Because of the high carbon content of Alloy B. Because of the lack of chromium. the presence of 0. in operational steam units..C.3). remains susceptible in water at 350°C (660°F) [14].002 ipy) in 37% HCl at 65°C (150°F) [17]. Resistance to boiling sulfuric acid is good up to 60% H2SO4 (<0.g. and 0.C. with no reported failures by intergranular S. Alloy B is not resistant to oxidizing conditions (for example. HNO3) or to oxidizing metal chlorides (such as FeCl3) [18]. Such tests show that heat treatment of Alloy 600 at either 650°C for 4 h or at 700°C for 16 h or more considerably improves resistance to S. The latter situation accounts for intergranular corrosion of the alloy by 25% HNO3 and by polythionic acids.002% C. but with greater resistance to HAZ attack and better weldability.C.3. These alloys are resistant to both oxidizing and reducing environments. Because of its chromium content. For this . but not at higher temperatures [5 mm/y (0. but are higher above this concentration. Resistance to acetic acid is excellent. aqueous corrosion resistance in oxidizing environments. weld HAZs undergo severe intergranular corrosion.01 ipy). 23. followed by rapid cooling in air or water [22]. 16% Mo. Alloy G-3 is widely used for downhole tubulars in the oil and gas industry [20. 4% W.5 mm/y (0. 22% Cr. 15% Cr. 21]. for this reason. crevice corrosion. The lower carbon content results in slower kinetics of carbide precipitation. they are easily formed and are weldable. Alloy G-3 has replaced Alloy G in most applications. and oxidation resistance. Alloy G was developed during the 1960s. Resistance is excellent to wet or dry Cl2 at room temperature. chromium. 23. 5% Fe The Ni–Cr–Mo alloys are widely used within the chemical process industry. B-2.025 mm/y (0. HNO3–H2SO4 mixed acids. H2CrO4. In addition. but it is difficult to fabricate and can crack during manufacturing operations. Alloy C has good resistance to hydrochloric acid at room temperature [0. 2% Cu The high nickel content in these alloys provides corrosion resistance in reducing environments. and CuCl2. Corrosion rates fall below 0. They resist chloride-induced pitting. and the higher molybdenum content improves the localized corrosion resistance. contains less carbon and is less sensitive to intergranular corrosion resulting from intermediate heat treatment. Alloy G-3 is an improvement on Alloy G.25 mm/y). Alloy C has outstanding resistance to pitting or crevice corrosion in seawater. with carefully controlled additions of iron. These embrittlement problems were resolved by developing Alloy B-3.416 NICKEL AND NICKEL ALLOYS low-carbon alloy. Alloy C-276.25 mm/y (0. FeCl3.25 mm/y) and for boiling phosphoric acid up to at least 50% acid (0. Alloy C is resistant to such oxidizing media as HNO3.001 ipy) for 37% HCl]. and S.3. When heated in the temperature range 500–700°C (930–1300°F). rates are high.5 Ni–Cr–Fe–Mo–Cu System: Alloy G—Ni. the alloy tends to corrode intergranularly. which can be avoided by a final heat treatment at 1210– 1240°C (2210–2260°F). and manganese [18.02 ipy) in <50% HNO3 at 65°C (150°F). 20% Fe. It is recommended for boiling sulfuric acid up to 10% acid (0.6 Ni–Cr–Mo System: Alloy C—54% Ni. introduced in the 1960s. 65°C (150°F)]. The chromium content contributes to strength.2 ipy) for 20% HCl. 6. In boiling 40% formic acid. In boiling >15% HNO3. Fe2(SO4)3. is less susceptible to intergranular attack and HAZ corrosion. Alloy C also resists wet or dry SO2 up to about 70°C (155°F)[17]. pitting in wet Cl2 may occur at higher temperatures.5% Mo.C. the rate is 0. As a result. 19]. 6. Crum. N. p. Wiley.REFERENCES 417 reason.C. Uhlig. 12. Corrosion in pressurized water reactors. Uhlig.. This alloy was developed from Alloy 800 by increasing the nickel content and by adding molybdenum (3%). The high chromium content in these alloys. and CEBELCOR. 515 (1961). 22% Cr The Ni–Fe–Cr system bridges the gap between the high-nickel austenitic stainless steels and the Ni–Cr–Fe alloy system. ASM International. Houston. . 33. editor. these alloys have been found to have superior crevice corrosion resistance in seawater [24]. 1. Wiley. Uhlig. 15. 1948. P. 14. Corrosion. M.. and titanium (0. Ref. Corrosion 35. imparts excellent corrosion resistance of these alloys to nitric acid [23]. 271. In addition. W. Vol. 107. p. Pement. 31% Fe. 836. 13. Electrochem. Electrochem. 3. 8. p. Bond and H. in Uhlig’s Corrosion Handbook.. 273. Z..3. Atlas of Electrochemical Equilibria in Aqueous Solutions. Nickel and nickel alloys. respectively. Electrochem.. P. C. Brussels. H. M. 352. 2. pp. 106. 831.. Alloy 825 is an upgrade from the 300 series stainless steels and is used to obtain resistance to localized corrosion and S. editor. 2006. in Corrosion Handbook. D. 21]. in ASM Handbook. Uhlig. Ref. 9. p.9%) for improved corrosion resistance in many media [20. 834. 650 (1963). 22% and 23% for C-22 and 59. 40 (1982). 5. Alloy C-22 and Alloy 59 were developed for applications where resistance to corrosion under highly oxidizing conditions is required. Combrade. 46 (1983). Scott and P. 1974. 2000. OH. P. Airey and F. Staehle et al. J. 7. H. TX. 369. 16. 1. J. Pessall et al. New York. p. Friend. New York. R. 4. 13C. REFERENCES 1. 2nd edition. and H. editor. 10. Corrosion 39. H. 367. Pourbaix. National Association of Corrosion Engineers. Feller. J. Ibid. R. Alloy C-276 can be used in most applications in the as-welded condition without severe intergranular attack. J. TX. Corrosion: Environments and Industries. p. J. in Proceedings of Conference Fundamental Aspects of Stress Corrosion Cracking. Osterwald and H. 110. Ibid. H. Materials Park. Bond. Soc. Coriou et al. Agarwal. p. 23. G.C. 835. copper (2%). p. Soc. 488 (1960). 100 (1979). Soc. National Association of Corrosion Engineers. W. Revie. Domian et al. 334. 1969. Houston. 11. 26 (1977).7 Ni–Fe–Cr System: Alloy 825—Ni. Corrosion 38. H. Materials Park. F. Corrosion in boiling water reactors. 2005. P. 22. J. Crook. 1997. pp. Corrosion: Environments and Industries. ASM International. Materials Park. American Nuclear Society. 19. R. 18. 19. pp. in ASM Handbook. Corrosion of nickel and nickel-base alloys. C. 13A. 2nd edition. M. W. ASM International. Corrosion: Environments and Industries. Gordon. Corrosion: Materials. pp. Wiley. Corrosion of nickel and nickel-base alloys. 1. and R. M. Vol. Combrade. Corrosion: Environments and Industries. Vol. Ref. in Uhlig’s Corrosion Handbook. p. 13C. 19. 236. editor. . Vol. Ref. P. Corrosion: Environments and Industries. ASM International. New York. in ASM Handbook. Was. 840. Corrosion resistance of stainless steels and nickel alloys. Corrosion in the Nuclear Industry B. OH. G. ASM International. L. in ASM Handbook. 2005. Vol. 339–340. 831–851. 21. p. Ref. OH. Materials Park. pp.418 NICKEL AND NICKEL ALLOYS 17. 2006. Samans. 231. Corrosion: Fundamentals. ASM International. Environmental Degradation of Materials in Nuclear Power Systems—Water Reactors. Horn. Vol. Manufacturer’s literature. 697–702. in ASM Handbook. OH. Nickel and nickel alloys. and Protection. 13B. 228–251. p. 336 (1966). A. B. 2006. OH. Gordon. 2006. Sedriks. 24. 2006. 19. GENERAL REFERENCES D. 20. in ASM Handbook. C. Revie. ASM International. OH. Tisinai. 1. 13C. Vol. 23. Materials Park. A. Ref. 362–385. Ref. p. in ASM Handbook. Introduction to corrosion in the nuclear power industry. ASM International. Meyer. La Grange Park. p. 13B. Scott and P. Proceedings. and P. and G. 238. Corrosion: Materials. 13C. pp. p. 230. pp. J. 2006. 13C. Materials Park. Materials Park. P. Busby. 2000. in ASM Handbook. Effect of irradiation on stress-corrosion cracking and corrosion in light water reactors. Materials Park. Testing. Ford. IL. OH. M. andresen. Eighth International Symposium. 386–414. Corrosion in pressurized water reactors. 837. pp. 341–361. P. M. Vol. Corrosion 22. OH. Crook. Agarwal. the metal is rapidly attacked by oxidizing acids and salts (for example.2. Section 3.H. Inc.8). [1]. 419 .1 INTRODUCTION Co2+ + 2e − → Co φ = −0. It resists hot reducing sulfur-bearing atmospheres better than nickel does. cobalt can be cathodically protected in acid or neutral solution by polarizing to potentials active to −0.277 V The standard potential of cobalt is close to that of nickel.1 [1]. by R. Like nickel. and passivation of cobalt in aqueous solution are presented in Fig. immunity. cobalt is relatively resistant to oxidation in air.24 COBALT AND COBALT ALLOYS 24. Winston Revie and Herbert H. Uhlig Copyright © 2008 John Wiley & Sons. hot or cold. Cobalt is also corroded by aerated aqueous ammonia solutions. HNO3 and FeCl3).E. 2+ Co(NH 3)6 . Corrosion and Corrosion Control. it resists alkalies. 24.5 V versus S. At elevated temperatures. being 27 mV more active (Table 3. The theoretical potential–pH domains of corrosion. As this figure indicates. forming a soluble complex—for example. although not as well as does nickel. 1. Alloys for high-temperature use 3. Alloys for wear resistance 2. for resisting corrosive environments and for high-temperature strength and hardness. Pourbaix. and passivation of cobalt at 25°C [1].420 COBALT AND COBALT ALLOYS Figure 24. Copyright NACE International 1974 and CEBELCOR. (M. iron. Alloying elements that are fcc stabilizers (such as nickel. to erosion by high-velocity liquids. Atlas of Electrochemical Equilibria in Aqueous Solutions. and implant materials for the human body. The cobalt-base alloys. cobalt is usually alloyed with chromium for applications where the alloys have practical advantages over similar nickel. Chromium-bearing cobalt alloys can be divided into three groups [2]: 1. 326. whereas above 417°C.or iron-base alloys. it is face-centered cubic (fcc). pressure gauges. nozzles. At temperatures up to 417°C. p.) 24. Alloys for resistance to aqueous corrosion and wear Cobalt alloys were developed in the early 1900s by Elwood Haynes of Kokomo. immunity. and carbon) lower the transformation . and to cavitation damage. cobalt has a hexagonal close packed (hcp) structure. pressure seats. are better resistant to fretting corrosion. for example. The alloys are used in cutting tools operating in aggressive chemical media. Indiana. steam valves and valve seats.2 COBALT ALLOYS Being less plentiful and more expensive than nickel. bushings. Theoretical potential–pH domains of corrosion. 2nd English edition. 4%—the alloy is resistant. behavior. a property that extends to Alloy 25 and MP35N as well. in aerated chlorides at temperatures below 60–80°C. susceptibility results [5]. Failures in 5% NaCl–0. saturated with H2S (NACE solution [6]). Vitallium.2%. chromium. Alloys of cobalt that contain both fcc and hcp stabilizers have transformation temperatures that are a complex function of the alloy composition [3]. Alloying cobalt with chromium reduces the current density. raise the transformation temperature. They typically contain chromium as well as molybdenum and/or tungsten. and tungsten. and tungsten [3].C. or sometimes by relatively rapid uniform dissolution. Accordingly. but by replacing most nickel with cobalt (and the incidental reduction of molybdenum to 6% and increase of chromium to 30%). but in 10% H2SO4 or HCl.C. which can be carbides of chromium. having a very noble critical pitting potential is also resistant to pitting in dilute NaCl solutions.C.C. making them relatively resistant to both reducing and oxidizing conditions. On the other hand. hot or cold. depend on temperature and the prevailing uniform corrosion rate to produce hydrogen. This heat treatment improves strength. The alloy containing 10–12% Cr is negligibly attacked by 10% HNO3. Applying stress to alloys in a metastable fcc structure at ambient temperature can partially transform them to hcp. In boiling 50% NaOH. When severely cold-worked and cathodically polarized at room temperature in 5% H2SO4 plus As2O3. as in Vitallium. but not above.1% acetic acid. Similar failures are also expected when the alloys are coupled in the same acid to a more active metal like iron. passivity is lost. This susceptibility does not include Vitallium exposed to saline solutions at 37°C (body temperature) in which the alloy is resistant. The alloys resist pitting in FeCl3 at room temperature. and corrosion rates become excessively high.COBALT ALLOYS 421 temperature. in MgCl2 solution at 153–154°C. but may also .. which simulates deep sour gas well environments. except Alloy 6B because of its high carbon content of 1. resembling the effect of added iron rather than the beneficial effect of added nickel. 0.C.C. The situation is analogous to the observed resistance of 18–8 (types 304 and 316) stainless steels to S. with the resultant galvanic action paralleling cathodic polarization. molybdenum. failures in this solution by hydrogen cracking (also sometimes called sulfide stress cracking) usually occur only for the cold-worked alloys that are subsequently heat-treated. At room temperature. molybdenum.C. with the required minimum current density of 5000 A/m2 (500 mA/cm2) being 14 times higher than that for nickel [4]. Alloy 25 is also susceptible. but not by HNO3. High resistance to abrasion results from the precipitation of carbides. Some commercial cobalt-alloy compositions are listed in Table 24. Cobalt can be anodically passivated in 1N H2SO4. the stressed alloys are susceptible to failure by hydrogen cracking. Alloying of Co–Cr alloys with molybdenum or tungsten reduces attack by H2SO4 or HCl.1. At a lower carbon content—for example. hcp stabilizers. all of the stressed cobalt alloys fail by S.C. the MP35N alloy resists S. with the 10% Cr alloy requiring only 10 A/m2 (1 mA/cm2) to become passive. Substantial additions of cobalt to chromium plus molybdenum (or tungsten) alloys are detrimental to S. 5a 4 2. Source: Data from ASM Handbook.007a Ti 1a N 0. pp. ASM International.75a 1a 0.1 1.5 1 1.7 Mo W Ni Fe C Mn Si Co bal.1.4 3a 0.15a 0. bal. Nominal Compositions (%) of Cobalt Alloys Cr 30 1.3 — bal.025a 0.25 0.08 — COBALT AND COBALT ALLOYS a Maximum. 13B.422 TABLE 24. B 0.5a 25 R30605 0.5 9. — Aqueous corrosion + wear resistance 21 MP35N Ultimet Vitallium R30021 R30035 R31233 — 1a 0. 2005. 172. 165. bal.75 35 9 — 3a 1a 3 — 0. Corrosion: Materials.15a 0. Other — Alloys for Common Name UNS No. Materials Park.8 0.06 0.4a bal. . bal.75 5 6 — — 2 — 2. OH. Wear resistance 20 — 15 10 3a 0.5 6B R30016 High temperatures 27 20 26 30 5. Vol. Davis. Electrochem. REFERENCES 1. OH. Materials Park. New York. 4. Soc. Austin et al. 13B. In this event. GENERAL REFERENCES Corrosion of cobalt and cobalt-base alloys. 164–176. Revie. in high-velocity (244 m/s) hot brines common to geothermal wells. L. p. in laboratory cavitation-erosion tests in distilled water. ASM Specialty Handbook.GENER AL REFERENCES 423 increase uniform corrosion rates in acids sufficient to generate hydrogen in amounts necessary to induce cracking. Proceedings of the Conference on Production and Applications of Less Common Metals. Crook and H. 718. P. W. Uhlig. Fretting Corrosion. p. Brussels. R. The LLL Geothermal Energy Program Status Report. but chloride S. W. prevents cracking. 9. 488 (1960). Houston. 2nd edition. 1982. Waterhouse [7] showed that a Vitallium screw mounted in a metal plate in saline solution and subjected to variable stress so as to cause slight rubbing of the screw head was damaged less than stainless steel. Ref. 5. cobalt alloys exhibited superior resistance [8]. and Their Alloys. 2nd edition. P. NACE International. Silence. Wiley. Also. M. 7. Corrosion: Materials. OH. Richards. Uhlig and A. 326. Mater. p. J. pp. Lawrence Livermore Laboratory. in ASM Handbook. Nickel. pp. Materials Park. and CEBELCOR. Livermore. April 1977. A. TX. 8. as in cathodic protection. The Metals Society. p. 2005. Similarly. may displace hydrogen cracking as the failure mechanism in the higher-temperature range. Evaluation of Pipeline and Pressure Vessel Steels for Resistance to Hydrogen-Induced Cracking. A. Atlas of Electrochemical Equilibria in Aqueous Solutions.C. The advantage of cobalt-base alloys to reduce fretting corrosion contributes to the advantage of using Vitallium in the human body. Perf. 2. Cobalt Alloys. Alloy 25 and MP35N resisted corrosion-erosion better than C-276 and much better than 26% Cr–1% Mo stainless steel [9]. P. J.C. editors. R. 717–728. coupling of the alloys to a more active metal. L. 107. Pergamon Press. TX. New York. Houston. 2000. California. Cobalt. R. Vitallium and Alloys 6B and 25 lost only 1/3 to 1/14 the weight loss of similar specimens of nickel-base C-276 and iron-base type 304 stainless steel. 2nd English edition. 2. 2000. 59. 6. Asphahani. 2000. . London. editor. 1972. ASM International. 717. Silence. in Uhlig’s Corrosion Handbook. editor. Bond and H. New York. Pourbaix. in Uhlig’s Corrosion Handbook. Crook and W. Cobalt alloys. Wiley. editor. TM0284-2003.. 1974. Vol. Standard Test Method. Revie. 3. Crook and W. 18(11). 9 (1979). R. National Association of Corrosion Engineers. Waterhouse. ASM International. H. Failure by hydrogen cracking usually diminishes as the temperature is raised (less hydrogen enters the metal and more escapes as gas). . 2 V more active than iron (Table 3. density = 4. 25. and HF [3]. NaOH. about 1. Uhlig Copyright © 2008 John Wiley & Sons. The theoretical potential–pH domains of corrosion.2.1 [1]. It has been reported that.25 TITANIUM 25. typically less than 10 nm thick. by R.5 g/cm3]. immunity.63 V Titanium is very active in the Emf Series. It was developed for the aircraft industry because of its high strength-to-weight ratio and excellent corrosion resistance. The metal is readily passivated in aerated aqueous Corrosion and Corrosion Control.p. Inc. 425 . = 1668°C (3053°F). This film. In this figure. it is now widely used in chemical process equipment. Titanium is a metal of relatively high melting point [m. Section 3. in the active state. tightly adherent oxide film that forms instantaneously in the presence of an oxygen source [2]. including hot concentrated HCl. is very chemically resistant and is attacked by very few substances. H2SO4. passivation is attributed to a film of TiO2. titanium may oxidize to form Ti3+ ions in solution [4]. The excellent corrosion resistance of titanium in many environments results from a hard.8). and passivation of titanium in aqueous solution are presented in Fig. Winston Revie and Herbert H.1 TITANIUM Ti 2+ + 2e → Ti φ = −1. 219. or wet chlorine. assuming passivation by TiO2 [1].8). Titanium is markedly resistant to oxidizing media in the presence of Cl−. In the passive state.3. Semiconducting properties of the passive film are generally attributed to oxygen anion vacancies and Ti3+ interstitials that function as donor states and give n-type semiconducting characteristics to the oxide.) solutions. Atlas of Electrochemical Equilibria in Aqueous Solutions.05 V). The Flade potential region (see Section 6. including dilute acids and alkalies. 3. (M. indicating stable passivity [5. It resists alkalies at room temperature. the average composition of which is TiO2. Pourbaix. Although titanium resists attack by hot dilute NaOH or by dilute H2O2. titanium is covered with a nonstoichiometric oxide film [4]. Only in strong acids or alkalies does passivity break down. and passivation of titanium at 25°C. Copyright NACE International 1974 and CEBELCOR. It is better resistant to nitric acid at elevated temperatures than are the stainless steels. 6]. Section 3. 2nd English edition. accompanied by appreciable corrosion. Theoretical potential–pH domains of corrosion. p.426 TITANIUM Figure 25. aqua regia at room temperature. heavy metal chlorides (FeCl3). for example. but is attacked by hot concentrated or fused sodium hydroxide. immunity. The galvanic potential of titanium in seawater is near the noble value for stainless steels (see Fig. combining the two produces high initial .1.2) is relatively active (φ F ≈ −0. Fig.5 1 % H2O2 1 1 + 0.1% Pd alloy.1. Ion implanting titanium .1% Ni in titanium or 0.35 + 1% Na2SiO3 + 0. The 8% Mn–Ti alloy is especially sensitive in this regard. also improve corrosion resistance.5% Ti metal by X-ray analysis) [9].2.5–28% NO2 and not more than 1. Small alloying additions of palladium.5% (NaPO3)6 1 + 0.2 TITANIUM ALLOYS Some commonly used titanium alloys are listed in Table 25.5% NaCl acidified to pH 1. IV.05% (NaPO3)6 696 169 22 0. the β. Test Period 1. it is furthermore reported that the corrosion rate is lowest for the basal plane of the hexagonal close-packed titanium lattice. boiling. On exposure of the metal for 1 h or more. in contact with liquid oxygen is reported to be sensitive to detonation by impact [10].0 gmd mm/y corrosion rates of more than 50 mm/y (2 ipy) [7. resistance is improved by orders of magnitude in the presence of small amounts of Cu2+ or Fe3+ (0. In the dry state.5 mm/y for 0.2) [13. for example.8 0. and it explodes when abraded in contact with concentrated HNO3. presumably as sponge.5% (NaPO3)6 0. Some data are given in Table 25.1.5 h % NaOH 2 2 0. containing 0. 25. Groups III. 25. which also shows the inhibiting properties of added sodium hexametaphosphate and sodium silicate. 0. Titanium. a dark substance forms over the surface (97. either in the environment or alloyed with titanium. body-centered cubic phase. Groups I and II contain a single phase. Corrosion Rates of Commercial Titanium in Alkaline-Peroxide Solutions [8] 60°C. the α phase.15 mm/y in 10% HCl. Group I alloys are the commercially pure grades. which has a hexagonal close-packed (hcp) structure. whereas Group II alloys contain palladium for improved corrosion resistance in reducing media. stressed or unstressed. or ruthenium also effectively reduce the corrosion rate in boiling 10% HCl (2. In this solution. platinum.0 56 14 1. to fuming HNO3 containing 2. Although corrosion resistance is poor in boiling HCl or H2SO4 (114 mm/y in boiling 10% HCl).TITANIUM ALLOYS 427 TABLE 25. Small amounts of nickel. Beta alloys are used predominantly in the aerospace industry because of their high strength/density ratios [2]. this surface film is pyrophoric when scratched.25% H2O. 14]. and V contain increasing concentrations of alloying elements and of the second phase. 8].02 mole Cu2+ or Fe3+ per liter) [11].2 ppm Ni2+ added to the solution are reported [12] to establish passivity in boiling 3. 8 Ni 8 Al.428 TITANIUM TABLE 25. Some Commonly Used Titanium Alloys [2] Common Alloy Designation UNS Number ASTM Grade Nominal Composition (%) Group I Commercially Pure Titanium Grade 2 R50400 2 0. 6 Cr.12 O. 4 Zr. low Pd R52400 R52402 7 16 0.3 Mo. 4 Mo 5 6 Al. 1 V.2.2. 0. Corrosion of titanium in boiling 10% HCl as a function of Fe3+ and Cu2+ concentration and alloyed palladium or platinum.15 Pd 0.05 Pd Group III Other Alpha and Near-Alpha Alloys Grade 12 Ti 8–1–1 R53400 R54810 12 — 0. 0. 4 V Figure 25. Pd Grade 2.12 O. 0. 1 Mo Group IV Alpha–Beta Alloys Ti 6–4 R56400 Group V Beta Alloys Beta C R58640 19 3 Al. . 8 V.12 O Group II Low Alloy Content Titanium with Pd Additions Grade 2. 6% H2O is present.3 PITTING AND CREVICE CORROSION Titanium is characterized by marked resistance to pitting and crevice corrosion in seawater. 21]. but instances have been recorded in hot seawater. These alloying elements or oxidizing ions accelerate the cathodic reaction (Cu2+. Titanium does not undergo crevice corrosion in seawater at room temperature. showed that titanium loses passivity in anhydrous 1 N HCl in CH3OH. Attack of this kind. At high Cl− concentrations and elevated temperatures.3). Fe3+. penetration of hydrogen into the metal accompanied by metal embrittlement occurs only above about 80°C (175°F) [18]. also. Resistance to pitting is accounted for by a very noble critical pitting potential of about 12–14 V in dilute chloride solutions at room temperature [20.PIT TING AND CREVICE CORROSION 429 surfaces with palladium reduces the corrosion rate [15] by a factor of 1000 in boiling 1 M H2SO4. Cl− was not necessary. Anodic passivation (or anodic protection) is also practical in HCl.9 and 1. pitting corrosion is observed in hot concentrated CaCl2 and similar solutions [22].8 V. such attack is more commonly observed in acid and neutral rather than in alkaline chloride solutions. In laboratory tests.0025 mm/y). Mansfeld [24]. a surface film of titanium hydride may form. When titanium is cathodically polarized or coupled to a more active metal in an acid like HCl. surface alloying with palladium by an electrospark technique or by electrochemical deposition with subsequent annealing are similarly effective [16]. Susceptibility was related to acid anodic corrosion products 4 . Br−. Initiation of pitting in titanium is more pronounced in Br− and I− than in Cl− solutions. 25. and that the pitting potential becomes increasingly noble as more H2O is added. the critical potential becomes much more active (Fig. following earlier Russian investigations. Griess [26] showed that titanium undergoes crevice attack in 1 M NaCl containing dissolved oxygen at 150°C (300°F). crevice attack was also observed in hot solutions of I−. Ni2+ reduction. and some other acids using a small impressed current [17]. such as under an asbestos gasket at 95–120°C (200–250°F) [25]. does not occur at temperatures below about 95°C (200°F). because of restricted diffusion of hydrogen in titanium at room temperature. respectively [23]. No pitting or appreciable weight loss occurred when titanium was buried in a variety of soils for eight years [19]. hence. The critical pitting potential in 1 M Br− and 1 M I− solutions at room temperature are 0. Addition of 1% Mo to titanium is effective in shifting the critical potential to more noble values above 125°C (260°F). However. it is reported.4). For this type of corrosion. 25. with accompanying greater resistance to pitting in this temperature range. and SO2− . but that passive behavior as indicated by polarization curves is restored when >0. H2SO4. or increased rate of H+ discharge) to a level at which the corresponding anodic current density reaches or exceeds the critical value for anodic passivation (see Section 6. Only very slight uniform weight loss occurs over periods of exposure of several years (<0. In deaerated NaCl in which differential aeration cells are not established. Cl2. the 0. or I− [31].3. Critical pitting potentials of commercial titanium and 1% Mo–Ti alloy in 1 M NaCl as a function of temperature [22]. 27]. Similar corrosion of commercially pure titanium occurs in methanol solutions containing Br2. Water acts as an inhibitor. Hence. or I2. or Br−. which resists breakdown of passivity in acids.1% Pd–Ti alloy. . where its resistance to pitting and corrosion-erosion make it a useful material. Cl−. Various case histories of crevice and pitting corrosion observed in practice have been described [28]. Titanium tubing can be used successfully for seawater-cooled heat exchangers [29].to 16-h tests). crevice corrosion is suppressed.4 INTERGRANULAR CORROSION AND STRESS-CORROSION CRACKING Intergranular corrosion of titanium (and various of its alloys) occurs in fuming nitric acid at room temperature (3. was found to resist crevice corrosion better than did titanium [26. Addition of 1% NaBr acts as an inhibitor [30]. accumulating in the crevice which reached a pH as low as 1.430 TITANIUM Figure 25.0. 25. C. Nevertheless. and I− are unique in causing S. (or intergranular corrosion?) usually along grain boundaries [33–35]. OH−.g.2–0. [40].C. about 100 ppm KNO3 is effective) [37. or I−. 4% V–Ti alloy has failed by stress-corrosion cracking (S.C. Successful anodic protection against S. lowering the aluminum content) that favor the β phase increase resistance to S. . in distilled water (e.C.E.C. SO2− . Br−.g. and ClO− do not cause 3 4 4 failure and. 2% Nb. 1% Ta–Ti alloy in 3% NaCl at −1. their behavior simulates the effect of extraneous anions in the case of austenitic stainless steels (see Section 19..H. S2−.76 V (S.1 V (S. 1% Mo. in the presence of halide ions.. F. may inhibit crack propagation in some alloys susceptible to S. Leckie [39] reported protection of 7% Al. the specificity of the environment. suggesting similarities in the mechanisms.C.4% O) and of various alloys including the 8–1–1 alloy.C.C. The 8–1–1 alloy is a mixture of mostly α (hexagonal close packed) and some β phase (body-centered cubic). On the other hand.C. of the 8–1–1 alloy was reported [38] in the presence of Cl− but not in the presence of Br− or I−. B.2. Various anions of the group just listed also act as inhibitors of S. Beck [38] found that cathodic polarization of the 8–1–1 alloy to −0. The sensitive effect of alloy structure. when a sharp surface flaw is present.) prevents failure in the presence of Cl−. Pure titanium is resistant to this type of failure. and the effects of extraneous anions and applied potential are similar to effects observed in the stainless steels (see Section 8. but less potassium nitrate (10 ppm) was needed to inhibit cracking in methanol as compared to water [37]. NO− . the mechanisms of cracking of titanium alloys are still being discussed. Various titanium alloys.) in liquid N2O4 within 40 h at 40°C (105°F) [32].H. Failure of the 8–1–1 alloy also occurs in pure methanol (CH3OH).E. the composition of the phase is also a determining factor.C. Heat-treatment procedures and composition changes (e.C. at which hydrogen evolution was copious.C. from fingerprints) at 260°C (500°F) or higher. adding a small amount of Cl− to distilled water or to methanol did not increase velocity of crack propagation. because the β phase of several other titanium alloys is found to be susceptible to S.C. [36] showed that titanium alloys in seawater.C. 38]. tend to undergo transgranular S.C.6). in this respect.3 V. 1% V–Ti (8–1–1) heated in air in contact with moist sodium chloride (e. Interestingly. F−. Br−.) and also at −1. with debate continuing to take place regarding the relative importance of anodic dissolution and hydrogen reduction processes.C. undergo S. A slight excess of NO (or the presence of H2O) inhibits such failure. of commercial titanium high in oxygen content (0. including 8% Al. It has been observed that aqueous solutions of Cl−.g.3). The stressed alloy was also found to be sensitive to cracking in the pure nonaqueous solvents CCl4 and CH2Cl2. with observed cracks proceeding across grains of the α alloy but with the β phase failing by ductile fracture. despite otherwise excellent corrosion resistance. Brown et al. However. to the contrary.C..C.INTERGR ANUL AR CORROSION AND STRESS-CORROSION CR ACKING 431 The 6% Al. OH. Pourbaix. 2005. it forms at lower temperatures when hydrogen is cathodically discharged. Atlas of Electrochemical Equilibria in Aqueous Solutions.g. Fe3+ and Cu2+) or other oxidizing substances (e. Houston. Hydrochloric and sulfuric acids. hot or cold (e. Wiley. Molten salts (e. 2. titanium is resistant to the following: 1. titanium may ignite.. 8. Corrosion of titanium and titanium alloys.g. 0. 2nd English edition. J. REFERENCES 1. Known to resist pitting and crevice attack for five years and longer. editor. Seawater. Wet chlorine (>0. Titanium hydride forms rapidly above 250°C (480°F).432 TITANIUM In brief. Been and J.9% H2O. FeCl3. or dilute alkalies plus H2O2. 42 m/s) [41]. >10% formic. 219.. Oxalic.g. Concentrated hot alkalies. Titanium and titanium alloys. 6. less in flowing gas) [42]..g. or hydrogen. 2nd edition. R. 5. and CEBELCOR.005 mm/y. Titanium is not resistant to the following: 1. Schutz. M. Fluorine. all concentrations and temperatures up to boiling. nitrogen. 4. 7. or hydrogen at elevated temperatures leads to embrittlement. Absorption of oxygen. W. 4. 4. Oxidizing solutions. 2. 2000. K2Cr2O7 and NaNO3) or when alloyed with platinum or palladium. Boiling CaCl2. anhydrous acetic [43] acids. Aqueous hydrogen fluoride. 864.C. 2. and K2Cr2O7). 3. Revie. Corrosion: Materials. Brussels. New York. >55% [44]. in ASM Handbook.C. 1974. ASM International. Hypochlorites. R. 3. . except when dilute.0008 mm/y.5-year exposure. Materials Park. Oxidation in air occurs above 450°C (840°F) with formation of titanium oxides and nitrides. LiCl. in Uhlig’s Corrosion Handbook. but not to fuming HNO3. p. CuCl2. High-temperature exposure to air. including high-velocity conditions (0. Ignition temperature decreases with increased air pressure sometimes causing localized combustion of titanium-alloy blades in the compressor section of gas-turbine engines [45]. 13B. 3. Grauman. Vol. and fluorides). W. S. or in moderate concentrations when inhibited by oxidizing metal ions (e. TX.. National Association of Corrosion Engineers. 253. NaCl. p. Nitric acid. CuSO4. High oxygen concentration in titanium may induce susceptibility to S. In dry Cl2. 5. nitrogen. p. and N. J. 397. T. 8. Prot. and T. Posey and E. Also S.. in Proceedings of the Fifth International Congress on Metallic Corrosion. Fennel and N. Soc. Rittenhouse. Philadelphia. Boyd. Green and R. Prot. 96 (1968). 40. p. Battelle Memorial Institute. 19. 88. Chernova. Soc. 13. Hubler. A. E. Fedoseeva. 6. Trans. Sci.M. 23. Athens. Soc. Sedriks. K. Wissenberg. 1152. 30. 16. 29. Dugdale and J. 1. Special Technical Publication No. Staehle et al. Am. 11. A.REFERENCES 433 4.. T. M. 21. Corrosion 35. Sanderson and M. 29 (April 1969). Mansfeld. PA. Tomashov. Special Technical Publication No. Report No. September 1966. 262 (1953). DMIC Memo 218. Jackson and W. J. Bohlmann. 31. 2. Soc. Kiefer and W. 386 (1973). 14. Trans. J. 319 (1982). Corrosion 29. 644. 18. Ind. Thomas and K. 10. 116. Covington. Mater. Zashchita Metallov. Nakamoto. 378 (1979). J. editors. Corrosion 36. 1961. . L. 51(1). Kelly. 12(4). in Ref. Mater. Beck. 57. Assoc. Nobe. J. Bishop. J. 218–226. Mater. Metals 51. J. I. 125. Electrochem. 239 (1960). pp. 1892 (1975). N. Wissenberg. 432. 363 (1976). Tech. F. American Society for Testing and Materials. 32. Latanision. Phys. 397 (1964). J. 99. 25. Suzuki and Y. E. Am. 12. J. 28 (1979). Soc. 6(10). American Society for Metals. 26. Vol. 106. American Society for Testing and Materials. 1748 (1969). pp. Sigalovskaya et al. Corrosion 23. Columbus. Metals Handbook. Boyd. OH. 63. Chem. Griess. Chem. 650. Soc. Jr. 324 (1969). Hackerman. Harple. Corros. Stern and H. 14. Perf. Cobb and H. Electrochem. 1968. 74 (1953). C. Soc. 8th edition. B. Cotton. 17. Cotton. 20. G.. Electrochem. Mater. TX. Takamura. 306 (1967). Hall. TX. 28. Stalter. Corrosion 25. OH. Electrochem. Soc. Houston. Battelle Memorial Institute. Second Symposium On Fresh Water from the Sea. Columbus. J.. Greece. Stern and C. Houston. 18(4). in Localized Corrosion. M. 5. Jackson and W. p. 1412 (1971). 15. 33. Morioka. 20(6). J. in Applications Related Phenomena in Titanium Alloys. 62A (1968). November 1957. J. 1974. 755 (1959). 13 (1952). National Association of Corrosion Engineers. 22. Corrosion of Ti. 871 (1951). p. TAPPI. (London) 1958. 106. T. 1966. 1974. National Association of Corrosion Engineers. M. R.L. McMaster. Sedriks and J. 68. Modern Aspects Electrochem. Philadelphia. J. J. Jr. 118. 9. Perf. 24. Metals 52. 313–317. 4. Uhlig. Corrosion 31. Stress Corrosion Cracking of Titanium. Electrochem. G. Stern and H. Pulp and Paper Ind. 23 (1981). Corrosion 24. Electrochem. 27. 759 (1959). J. Metals Progr. McCafferty and G. May 1967. 60 (1975). 8. T. J. A. 7. F. Cleveland. J. 201 (1980). Green. OH. Shiobara and S. Rideout et al. Romanoff. 22 (1967). See also K. N. Titanium and titanium alloys. G. TX. Gegner and W. OH). 551 (1967). Trans. Report NBS 1R-79-1616. 36. Been and J. International Nickel Co. Revie. Columbus. T.. Wilson. Center. ASM International. H. Schulz. CAB. 385 (1967). 43. editor. Washington. 319 (1982). 6.. Risch.. Leckie. Adams and E. Welsch. Turner. T. Electrochem. Corrosion 23. Mellway and M. in ASM Handbook. 1962. 114. 40. Von Tiesenhausen. 1969. H. 45. J. Blackburn. Strobridge et al. Chem. T. NRL Memo. 41. F. Collins. 307 (1965). Houston. H. 44. and H. June 1980. 252–299. 187 (1967). 326 (1968). Boulder. pp. 14. PROBLEM 1. 1994. W. S. Kelly. Corrosion of titanium and titanium alloys. 863–885. 13B. Materials Park. editors. ASM International. Mod. 35. Corrosion 21. R. Aeronaut. July 1965. 3101 (1970). Am. New York. Titanium Alloys. Corrosion 15.. p. Inst. 2000. show the effect of alloying with palladium or platinum and explain the reduction of corrosion rate in the alloyed material compared to pure titanium. Materials Park. Corrosion: Materials. private communication. 42. . in Proceedings of Conference Fundamental Aspects of Stress Corrosion Cracking. W. Corrosion 25. R. E. Soc. OH. 2nd edition. J. Issue No. 2005. 1. Materials Properties Handbook. Brown et al. Staehle et al. 84. GENERAL REFERENCES J. 337 (1969). Using the polarization curve for titanium in sulfuric acid. K. Corrosion 23. editors. R. National Association of Corrosion Engineers. 38. E. Beck and M. Gray. Grauman. Battelle Columbus Labs. Newcomber. Boyer. Gray and J. May. in Uhlig’s Corrosion Handbook.-Ing. P. B. Beck. Wiley. Aspects Electrochem. W. 39. Vol. National Bureau of Standards. Naval Research Laboratory. Astron. D. July 1979 (Metals and Ceramic Inf.C. pp. 341t (1959). R. Electrochemical behavior of titanium.. Colorado. Marine Corrosion Studies (Third Interim Report of Progress). 691.434 TITANIUM 34. R. Tourkakis. R. Report 1634. T. H. OH. 88 (1967). and E.-Tech. Johnston. 39. Metall. 37. Kleinman. Winston Revie and Herbert H. however. 26. = 1852 °C (3366°F). The reaction is slow enough. N2. Corrosion and Corrosion Control. hexagonal close-packed (hcp) structure.1 INTRODUCTION Zirconium is an active metal in the Emf Series. α-zirconium. at higher temperatures. normally exhibiting very stable passivity. which transforms to body-centered cubic (bcc). It reacts with air to form both zirconium oxides and nitrides. α-zirconium absorbs nitrogen in solid solution up to 25 at.p. Inc. At temperatures below ∼860 °C.8 wt. β-zirconium. with the metal dissolving up to 29 at. Similarly. An unusual property is the high solid solubility of the metal for oxygen. by R. and H2.%) according to the oxygen–zirconium phase diagram [2].1. density = 6. As illustrated in Fig.26 ZIRCONIUM 26. 435 . The metal [m. zirconium is passive over a very wide range of potential and pH [1].7 wt.45 g/cm3] reacts readily at elevated temperatures with O2. Uhlig Copyright © 2008 John Wiley & Sons.%). is the stable form. to permit hot rolling at 600–750 °C (1100–1400°F) [3].% (6.% (4. a transition or “breakaway” point is observed in the corrosion rate (see below) after a specific time of exposure. and passivation of zirconium at 25 °C. The . Commercial zirconium. The outstanding corrosion property of zirconium is its resistance to alkalies at all concentrations up to the boiling point. 2nd English edition. titanium.2 0.H. the metal must be low in carbon (<0. to a lesser extent.8 –1. contains up to 4.02% max) has a low thermal neutron capture.6 –2 immunity 0 2 4 6 8 10 12 –2. In this respect.2 ZIRCONIUM ALLOYS Because of good corrosion resistance in both acids and bases.5% hafnium.436 ZIRCONIUM –2 2 1.06%) for optimum resistance.8 Potential (volts. which is controlled by a very stable oxide.2 –1.) 26. this passive oxide is 2–5 nm thick [4].4 –0.2 0.4 0 –0. At ambient temperature. Zirconium is resistant to hydrochloric and nitric acids at all concentrations and to <70% H2SO4 up to boiling temperatures.1. which is difficult to separate because of the similar chemical properties of zirconium and hafnium.6 1. The pure metal low in hafnium (0. 227. p.6 –2 –2. as used primarily for corrosion resistance in the chemical industry [4]. In boiling 20% HCl. Copyright NACE International 1974 and CEBELCOR. In HCl and similar media. It also resists fused sodium hydroxide. immunity.6 1.) 0. zirconium alloys are widely used in chemical plants.4 14 16 pH Figure 26. Pourbaix. making it useful for nuclearpower applications.4 –0. it is distinguished from tantalum and. S. Atlas of Electrochemical Equilibria in Aqueous Solutions.2 –1.8 –1. Theoretical conditions of corrosion.4 –2 a b 0 2 4 6 8 10 12 14 16 2 1.E. which are attacked by hot alkalies. assuming passivation by ZrO2·2H2O [1].4 0 –0. The presence or absence of hafnium has no effect on the corrosion resistance. (M.8 corrosion passivation corrosion 0. H.E.45 V in 0. Zircaloy-2. several new zirconium alloys have been developed. which is higher than the initial rate. but not necessarily immunity [15]. similar to that of commercial titanium.C. 0.g. or mixed intergranular/transgranular [8]. an alloy used in nuclear reactors.C. 17]. the presence of NO2 does not have a significantly damaging effect [14]. and Zircaloy-4 (Zr. zirconium alloys are used as nuclear fuel cladding and structural fuel assembly components (see Section 8.C. commercially pure zirconium. Cold work and irradiation hardening have been reported to increase susceptibility. transgranular. but also a delayed hydride cracking process. nor is it resistant to HF or fluosilicic acid. materials reliability does have a significant effect on the economics of nuclear power plants. but not when a small amount of water is added [7]. and there is considerable incentive to develop a full understanding of the mechanisms of corrosion of zirconium alloys in reactors and to develop alloys that are resistant to both irradiation and corrosion in reactors [18].C..5% Nb. such as Zirlo (Zr–1.C.2% Fe. Although zirconium alloys do have good resistance to S.5). they are susceptible in many environments. Cracking can be intergranular. To help improve the corrosion resistance of Zircaloy.0045 ipy) [5]. Critical pitting potentials of 0.1% Cr. Stress-corrosion cracking may also occur in FeCl3 and CuCl2 solutions [13]. in >20% HNO3 at 25 °C with the maximum cracking rate in 70–90% HNO3.E. suggests that stress may not be necessary and that the failure is perhaps better described as intergranular corrosion. In slow strain rate tests. 1. 1. 26. weldments should be heat-treated to reduce residual stresses. is subject to S. Notwithstanding the progress so far. Unlike the situation for titanium. In general.5% Sn. whereas the higher-strength alloys are more susceptible. Zirconium is not resistant to oxidizing metal chlorides (e. Unalloyed zirconium has been reported to be resistant to S.38 V (S.) in 1 N NaCl and 0..C.3 BEHAVIOR IN HOT WATER AND STEAM The good resistance of zirconium to deaerated hot water and steam is of special importance in nuclear-power applications. 0.C.C. 0. 0. not only S.0% Nb–1. 0. because of its low yield strength. This behavior. Zircaloy-2 (Zr.05% Ni).C.C.5% Sn.0% Sn–0. The intense radiation within the reactor core accelerates degradation by increasing the rates of corrosion and hydriding.34 V (S. C-ring) tests of commercial zirconium and some zirconium alloys in 70% HNO3 up to the boiling point showed high resistance to SCC. It undergoes intergranular S.11 mm/y (0. can occur at highly stressed locations [8–11]..) in 5% NaCl] [12].1% Fe. Both zirconium and Zircaloy-2 are subject to SCC in iodine vapor (a major fission product of uranium) at 300–350 °C [16. The metal or its alloys can be exposed .1% Cr) undergo S.1% Fe). In the nuclear industry. In Zr–2. Constant strain (U-bend.C. is usually less than 0. FeCl3. in chlorides at 25 °C at potentials noble to the breakdown potential of the air-formed oxide film [0.BEHAVIOR IN HOT WATER AND STEAM 437 final rate.H. in anhydrous methyl or ethyl alcohol containing HCl.1 N NaCl [6] indicate that the metal is vulnerable to pitting in seawater.C. pitting occurs). This “transition” phenomenon is reported to occur for pure or impure zirconium after a weight gain on the order of 3. Phosphoric acid.04%) [20].2. including fused caustic. but after a certain time of exposure. acetic. Hydrogen formed by the decomposition of H2O during reaction appears to exert a dominant role. But were this the mechanism.438 ZIRCONIUM in general for prolonged periods without pronounced attack at temperatures below about 425 °C (800°F). especially that portion which dissolves in the metal.5% tin in combination with lesser amounts of iron. X-ray data for the oxides that form in H2O show a monoclinic modification of ZrO2 either before or after the breakaway time. The oxidation behavior of Zircaloy-2 in water and steam is shown in Fig. but tin in the presence of small amounts of alloyed iron. may occur under slow strain rate conditions [14]). It has been suggested on this basis that trivalent nitrogen ions in the ZrO2 lattice increase aniondefect concentration. Sulfuric acid. <70%. boiling. Adding to the complexity is the observation that alloyed tin appreciably shortens corrosion life of zirconium in water. Embrittlement of metal and higher corrosion rates occur above boiling temperatures under pressure. 6. causing higher oxidation rates [19]. Boiling formic. and similar additional accelerated oxidation may occur at still higher weight gains [19]. 5. Marker experiments indicate that oxidation proceeds by diffusion of oxygen ions toward the metal–oxide interface (anion defect lattice).C. the rate suddenly increases. The damaging effect of nitrogen in this respect is offset by alloying additions of 1. thereby increasing the diffusion rate of oxygen ions. Hydrochloric acid. or. 26. 3. which is not the case. all concentrations up to boiling point. 2. However. all concentration up to boiling point. . oxidation in oxygen would also be affected. an increased corrosion rate does not occur when the metal oxidizes in oxygen except for much longer times and much thicker oxide films. nickel. or citric acids. chromium again increases corrosion resistance. including red fuming acid (S. lactic. and chromium. Nitric acid. but with some evidence that the initial oxide is a tetragonal modification [20]. Alkalies. The rate of attack is characteristically low at first. zirconium is resistant to the following: 1. It occurs at lower temperatures if the zirconium is contaminated with nitrogen (>0.0 g/m2. all concentrations up to boiling point. to a lesser extent. In brief.5–5. <55% H3PO4.005%) or with carbon (>0.5–2. The mechanism accounting for the transition phenomenon is not well understood. boiling. ranging from minutes to years depending on temperature. It has been explained on the basis of cracks forming in the oxide because of stresses accumulating as the oxide thickens. 4.C. with the combination overcoming the detrimental effect of alloyed nitrogen. nickel. 3.2. 2nd English edition. Boiling CaCl2. >55% [21]. Jr. nitrogen. 2. National Association of Corrosion Engineers. 227.. Atlas of Electrochemical Equilibria in Aqueous Solutions. 7. M. B. Hansen. showing transition points [19]. p. Oxygen. 200 °C (explodes) [22]. Trichloroacetic or oxalic acids. p. Wet chlorine. 8. . p. Hurford. editors. Aqua regia. 1079. 2. Constitution of Binary Alloys. Zirconium is not resistant to the following: 1. and hydrogen at elevated temperatures. New York. REFERENCES 1. Kerze.REFERENCES 439 Figure 26.g. 1974. FeCl3 and CuCl2). 1955. 1958. Hydrofluoric acid and fluosilicic acid. 5. 3. and CEBELCOR. M. 4. Lustman and F. McGraw-Hill.. 6. Houston. Oxidizing metal chlorides (e. Pourbaix. boiling. Brussels. TX. 263. New York. in The Metallurgy of Zirconium. McGraw-Hill. W. Carbon tetrachloride. Corrosion of Zircaloy-2 in high-temperature water and steam. Revie. A. pp. 39. F. 224–240. ibid. in ASM Handbook. 13. Gegner and W. pp. H. Beavers. B. editor. R. ASM International. 79 (1977). R. 173. Te-Lin Yau. R. pp. Takamura. D. International Atomic Energy Agency. 2006. Huang and W. in Uhlig’s Corrosion Handbook. 22. Nucl. 18. 241 (1991). 13C. B. Fracture Mech. Vol. Revie. Corrosion 37. 7. 1987. Trans. Mills. Kuhn. 1976. P. Kolotyrkin. J. B. Zirconium alloy corrosion. 13B. Kerze. 908. K. 15. R. Coleman. editors. 10. A. 29 (1966). Cox and J. Cox. editor. 2005. 159 (1978). Wiley. pp. J. 20. Wilson. Vienna. Lustman and F. C. 406–408. 16. Cox. 1. Electrochemical Society. 1 (1990). Mori. Cox. Thomas. Oxidation of zirconium and its alloys. Huang. Corrosion 15. 21. 207 (1972). H. Lemaignan. M. B. E. and W. New York. 366 (1974). Cox. Wiley. Eng. Fontana and R.. J. Corrosion: Environments and Industries. p. OH 2006. W. pp. Corrosion 28. Princeton. ASTM STP 939. 167 (1983). OH. in ASM Handbook. 261t (1963). Corrosion 34. 14. 300–324. 4. 415–420. 608–640. 6. Coatings Corros. 1974. Plenum Press. 292 (1981). J. Corrosion 22. 275. J. C. Cox. Corrosion of zirconium and zirconium alloys. 39. New York. B. editors. in Corrosion Problems in Energy Conversion and Generation. Wood. in ASM Handbook. 17. 8. Gangloff et al. 157 (1973). B. Corrosion 19. Corrosion: Environments and Industries. and J. GENERAL REFERENCES B. B. p. 2049 (1991). Irradiation effects on corrosion of zirconium alloys. Sutherlin. Corrosion 35. Corrosion of zirconium alloy components in light water reactors. Materials Park. 2nd edition. J. Zirconium alloy corrosion. Cheadle. McGraw-Hill. Corrosion: Materials. Materials Park. C. ASM International. 5. New York. NJ. Corrosion 15. 19. Shimose. A 22A. Boyd. Vol. 1998. 341t (1959). W. 2000. New York. C. Corrosion 29. Cox. 170. Met. p. Materials Park. Griess. ASM International. W. 905–914. 103t (1959). Jr. Corrosion 18. 2nd edition. IAEA-TECDOC-996 (1998) Waterside corrosion of zirconium alloys in nuclear power plants. Vol. in Uhlig’s Corrosion Handbook. OH. Archer and M. 33. Corrosion 39. Tedmon. Mater. 905–906. Vol. 1955. . See also C. 2000. 9. 33t (1962). F. Cox. Rev. Cox. 5. in The Metallurgy of Zirconium.440 ZIRCONIUM 4. W. B. B. B. Busby. Advances in Corrosion Science and Technology. 316 (1979). 11.. Prevention of delayed hydroide cracking in zirconium alloys. W. Staehle. editor. Te-Lin Yau and R. pp. and T. pp. Mills and F. 13C. J. Ref. 12. Ambler. Harter. Tantalum is attacked by alkalies and by hydrofluoric acid. Inc. and a low passive current that is insensitive to Cl−. 441 .3 mm (0. Corrosion and Corrosion Control. 27. by R. density = 16. as expected from the theoretical potential–pH domains of immunity and passivation shown in Figure 27. HNO3.6 g/cm3] exhibits the most stable passivity among known metals.013 in. permitting varied applications of the metal despite high cost..g.) thick. Liners of tantalum sheet may average only 0. H2SO4 concentrators and HCl absorption systems).2 CORROSION BEHAVIOR High corrosion resistance to acids makes tantalum useful for special applications in the chemical industry (e.p. HCl.1 INTRODUCTION Tantalum [m. Corrosion resistance of this order suggests a Flade potential much more active than the hydrogen electrode potential. Tantalum retains passivity in boiling acids (e.g.1 [1]. It is readily embrittled by hydrogen at room temperature when the metal is cathodically polarized.. Winston Revie and Herbert H.27 TANTALUM 27. Uhlig Copyright © 2008 John Wiley & Sons. or H2SO4) and in moist chlorine or FeCl3 solutions above room temperature. = 3000 °C (5430°F). H. tantalum is resistant to the following: 1.4 –0. Platinum can be attached to tantalum by riveting.6 –2 14 16 pH 0. Nitric acid. For example.2 immunity –1. even though tantalum is not embrittled by concentrated hydrochloric acid at boiling temperatures. Hydrochloric acid.8 –1.4 vol. The benefit of coupled platinum extends to avoiding embrittlement caused by hydrogen generated through a corrosion reaction.E. The damage by cathodic hydrogen can be avoided by coupling tantalum to a very small area of a low-overpotential metal.2 0. Theoretical potential–pH domains of immunity and passivation for tantalum at 25 °C [1]. welding.) 0 2 4 6 8 10 12 14 16 2 1.4 0 –0.) or when coupled in an electrolyte to a metal more active in the Galvanic Series. it becomes embrittled at 190 °C (375°F) under pressure.6 b 1.8 –1. Pourbaix.8 passivation 0.000.6 1.4 –0. corrodes generally in anhydrous CH3OH–Br2 mixtures at room temperature and undergoes pitting corrosion in CH3OH. This damage does not occur if an area ratio of platinum to tantalum is provided on the order of at least 1 : 10.8 0. S. Metal embrittled by cathodic hydrogen or by that entering at elevated temperatures can be restored to normal properties only by heating in vacuum. p. despite its pronounced passivity in aqueous media. 4% H2O. .4 0 –0.% concentrated HCl at potentials slightly noble to the corrosion potential in this solution. all concentrations up to boiling point. or electrodeposition.2 –1. Hydrogen ions then discharge on the platinum instead of entering the tantalum lattice. Atlas of Electrochemical Equilibria in Aqueous Solutions.442 TANTALUM –2 2 1. Copyright NACE International 1974 and CEBELCOR. 2nd English edition. all concentrations up to and above boiling point. 2. Palit and Elayaperumal [4] stated that tantalum is subject to pitting in CH3OH + 0. such as platinum [2]. (M.2 Potential (volts.1. In brief.6 –2 –2 0 2 4 a 6 8 10 12 Figure 27. 254. 10% Br2. Fischer [3] reported that tantalum. Fischer. Aqua regia. Materials Park. Houston. 9. 13B. oxalic. 18. 2nd English edition. 1974. and acetic. Resistant to all concentrations up to and. Sulfuric acid. OH. above boiling temperatures. Tantalum is not resistant to the following: 1. . 4. Krupp. 169 (1978). Corros. and CEBELCOR. Fuming acid attacks tantalum at room temperature. Attack occurs at lower temperatures when the acid is contaminated with HF (>4 ppm). ASM International. Oxidizing metal chlorides. hot or cold. 7. rate = 0. for example.GENER AL REFERENCE 443 3. National Association of Corrosion Engineers. Fuming sulfuric acid. Corrosion 17. pp.09 mm/y (0. M. 5. Vol. Chromic acid. p. trace amounts. with 5% NaOH. Atlas of Electrochemical Equilibria in Aqueous Solutions. Tech. hot or cold (e. Cathodic hydrogen causes embrittlement at room temperature. Embrittlement occurs. REFERENCES 1. in ASM Handbook.g. 2. G. GENERAL REFERENCE Corrosion of tantalum and tantalum alloys. 254. 100 °C. all concentrations (except fuming) <175 °C (<350°F). Oxidation in air becomes appreciable above 250 °C (500°F). in some cases. W. 3. Corrosion: Materials. 3.. 2. 4. Sci. Elayaperumal. or hydrogen at elevated temperatures. FeCl3 and CuCl2). 2005. Bishop and M. 125 (1964). nitrogen. 8. Halogen gases [wet or dry Cl2 up to 150 °C (300°F). Brussels. Hydrofluoric acid and fluorides. Alkalies. Organic acids: lactic. TX. 337–353. Pourbaix. 4. 5. Phosphoric acid. Stern. 6.0035 ipy). Palit and K. Oxygen. Methanol solutions of Br2 or HCl. Mitt. For 85% H3PO4 at 225 °C. 22. Br2 up to 175 °C (350°F)]. 379t (1961). C. . 28 LEAD 28. therefore. 445 . and H2CrO4) by reason of thick diffusion-barrier coatings (Definition 2. As shown in the diagram. immunity. acetic and formic acids). becoming passive in many corrosive media that form insoluble lead compounds (e. Corrosion and Corrosion Control. The theoretical potential–pH domains for corrosion. HF. and passivation are shown in Figure 28.3 V.g. Inc. H3PO4. In these acids.1 [1]. H2SO4. however. lead can be cathodically protected by controlling the potential to less than −0. The metal is.8 V. Lead is used.4 to −0. Section 6. In alkaline solution. in the chemical industry as lining and piping..1 INTRODUCTION Pb2+ + 2e − → Pb φ = −0.126 V Lead is a relatively active metal in the Emf Series.. corrosion resistance is good provided relative velocity of metal and acid is below the value that causes erosion of protective films. Winston Revie and Herbert H. by R. in acid or neutral solution. potential control for cathodic protection is below −0.1). corroded by dilute nitric acid and by several dilute aerated organic acids (e. depending on the pH [1]. Uhlig Copyright © 2008 John Wiley & Sons.g. including waters that have been in contact with fresh Portland cement. immunity. Nevertheless. depending on aeration. the corrosion rate of lead is usually under anodic control [2].4 pH Figure 28. Pourbaix. It is attacked. and passivation for lead at 25 °C [1]. S.4 –2 0 2 4 corrosion 6 8 10 12 14 a 0 –0. 2nd English edition. the toxic properties of trace amounts of lead salts make it man- .2 CORROSION BEHAVIOR OF LEAD AND LEAD ALLOYS Being amphoteric. lead is corroded by alkalies at moderate or high rates. for example.2 –1. temperature. particularly damp timber including western cedar.8 immunity –1.2 0.446 LEAD –2 2 E(V) 1.8 0. by calcium hydroxide solutions at room temperature.8 Potential (volts. It is also durable for use in contact with fresh waters. This cause of corrosion can be prevented by fully drying the wood. Theoretical potential–pH domains of corrosion. but corrosion resistance of the alloy in some media is below that of pure lead.4 –0. treating it to prevent contact with moist air.6 –2 –2. p.1.) 0.2 –1.8 –1. Copyright NACE International 1974 and CEBELCOR. however. Organic acids harmful to lead can be leached from wood structures.6 1.6 1. (M. oak.H.4 –0. or inserting a moisture barrier between wood and lead [3].2 0.) 28. Alloying with 6–12% Sb increases strength [only at temperatures <120 °C (<250°F)] of the otherwise weak metal. and Douglas fir. An alloy of 2% Ag–Pb is used as a corrosion-resistant anode in impressed current systems for cathodic protection of structures in seawater (see Section 13. lead resists corrosion in many environments because the products of corrosion are insoluble and form self-healing protective films. 490.4 0 2 4 6 8 10 12 14 16 2 passivation b 1. Because of these protective films.6 –2 –2. Lead is resistant to seawater.6) [4].E. and concentration.4 corrosion 0 –0. Atlas of Electrochemical Equilibria in Aqueous Solutions. and the use of its alloys.1).. In applications where thermal cycling occurs. the corrosion rate may exceed that of steel in some soils (e. In the absence of dissolved oxygen. decreases the current. (28. particularly to industrial atmospheres in which a protective film of lead sulfate forms. 28.g. for soft potable waters. (28. For example. The U. Soluble silicates. Buried underground. . corrosion can also lead to shorter battery lifetime and reduced capacity to produce energy. The lead–acid battery uses lead grids at both the anode and the cathode. which lead to reduced battery lifetime and capacity. Following are the reactions that take place during discharge of a lead–acid battery [8]: At the anode: At the cathode: The net reaction: Pb → Pb2+ + 2e − PbO2 + 4 H + + 2e − → Pb2+ + 2 H 2O Pb + PbO2 + 4H + → 2Pb2+ + 2 H 2O (28. which results in energy production. Lead is resistant to atmospheric exposures. (28. the corrosion rate in waters or dilute acids is either negligible or very low.2.1) (28. In addition to this positive and beneficial aspect of corrosion in energy generation. limits lead in potable water to a maximum of 15 ppb [5].1) is part of the overall electrochemical reaction by which electrical energy is generated using the lead–acid battery. For the destructive forms of corrosion. and causes a deterioration of battery performance. Environmental Protection Agency (EPA) Lead and Copper Rule. Lead is used in solder alloys (typically with 5%tin. the overvoltage should be as low as possible.1 Lead–Acid Battery Lead–acid batteries are a principal use of lead. For corrosion by Eq. those containing organic acids). also act as effective corrosion inhibitors. but in soils high in sulfates the rate is low. which are components of many soils and natural waters. Corrosion of lead at the cathode has been reported to cause loss of capacity of batteries [9]. carbonated beverages. and a small amount of silver to increase strength) for joints of automobile radiators [7]. and all food products. the high coefficient of expansion (30 × 10−6/°C) of lead may cause intergranular cracking due to fatigue or corrosion fatigue.S.3) Corrosion of lead in Eq. The rate of corrosion in aerated distilled water is high (approximately 9 gmd) and the rate increases with concentration of dissolved oxygen [6].CORROSION BEHAVIOR OF LEAD AND LEAD ALLOYS 447 datory to exclude its use. corrosion during battery storage reduces the quantity of lead available for energy production and may also produce corrosion products that inhibit the corrosion that takes place by Eq. Passivation of the lead anode increases the anodic overvoltage. the overvoltage should be as high as possible [10]. implemented in 1991.2) (28.1). 3 SUMMARY In summary.08 mm/y (0. containing typically about 5% antimony. c. d. Lead is not resistant to the following: 1. <100 °C (but only dry Br2 and at lower temperatures). Alkalies. HNO3. Commercial H3PO4 (hot or cold). Furthermore. 4. . b. toxic stibine (SbH3) is produced. Chlorine. have been used for many years [13]. lead is resistant to the following: 1. Lead–antimony alloys. HF <60–65% (room temperature).0005 ipy). 5. on overcharge. Alloying elements are added to improve mechanical properties. tin. the rate for <80% H2SO4. Many strong acids. SO2. <96%. the rate is 0. 4. the rate in caustic alkalies is considered tolerable. 3. titanium. Many aerated organic acids. Atmospheric exposures.08 ipy). 14].448 LEAD Various alloys of lead are used as support grids for positive and negative plates of lead–acid batteries. e. 28. Seawater (0. room temperature). 3. Lead–calcium–tin alloys used as grid material for maintenance-free batteries have been reported to be susceptible to intergranular corrosion at large grain sizes [16]. These alloys develop an adherent protective corrosion product layer. a. 15]. boiling. 2.01 mm/y. HF. room temperature. H2CrO4 (as used in chromium plating). <70%. SO3. the deposition of antimony on the negative plate lowers the hydrogen overpotential [13. 2. is <2 mm/y (<0. Other possible alloying elements that have been studied include magnesium. H2SO4. for 20% H2SO4. but anodic corrosion of these alloys can limit battery lifetime [11. For some chemical applications. 0. 12]. however. 14]. H2SO3. 6. and bismuth [14. gaseous. Concentrated H2SO4 (>96%. For these reasons. alloys of lead with calcium. wet or dry. particularly industrial. H2S. and aluminum are also being used [13. HCl. boiling.003 ipy) [17]. in ASM Handbook. Corrosion of lead and lead alloys. Goodwin. Alhassan. 127. N. 130. Bialacki. 2003. 14. 2003. 2nd edition. Ref. London. Lead and lead alloys. Rose and J. 2000. Anodes for batteries.GENER AL REFERENCES 449 REFERENCES 1. 2nd English edition. Hampson. OH. 13B. R. 17. 2003. and I. J. S. pp. Windisch. 16. Corrosion: Fundamentals. Wiley. F. 203. editor. 11. and K. Soc. Environmental Protection Agency website: http://www. 171 (1993). Electrochem. . in Uhlig’s Corrosion Handbook. Atlas of Electrochemical Equilibria in Aqueous Solutions. TX. Vol. Materials Park. M. Electrochem. 19–21. Switzerland. Young. and Protection. S. 767–792. Essentials of Lead–Acid Batteries. 9 (1995). Staehle. 2nd edition. Karaikudi.epa. Appl. NACE International. C. F. 2060 (1981). 1976. Electrochem. Houston. and Protection. editor. OH. Elsevier. M. 179 (1994). Materials Park. 8. New York. 13B. R. Testing. Anodes for batteries. A. C. 2000. and CEBELCOR. Corrosion: Materials. A. Sutula. Revie. 13A. J. 91 (1980). Metal Bulletin Limited. TX. 2003. C. 2. Vol. 2. Peters. p. S. OH. Lectures on Electrochemical Corrosion. in ASM Handbook. Revie. 170–177. R. F. translator. Corrosion: Materials. Dacres. Pavlov. Soc. 37. in ASM Handbook. D. S. 770. Green.. p. W. 1974. E. J. 860. 170–177. Corrosion: Fundamentals. Corrosion of lead and lead alloys. Windisch. E. New York. Houston. p. Electrochem. Peters. 13A. OH. 9. Lausanne. Cathodic protection. Hampson. N.gov/safewater/lcmr/ 6. p. 19. and Protection. p. 1797 (1983). 195–204. Ref. Vol. ibid. p. 768. W. Reamer. ASM International. J. 1456 (1980). Heidersbach. 490. J. Ref. Pourbaix. Soc. pp. Wiley. F. H. 2006. J. in ASM Handbook. R. editor. 2. R. pp. Pavlov. third English edition. N. in ASM Handbook. ASM International. W. Materials Park. ibid. 10. 771–775. D. 13. ASM International. ASM International. 5. ASM International. S. 775. Corrosion: Fundamentals. 128. Vol. V. pp. 3. Testing. P. Society for Advancement of Electrochemical Science and Technology. 48. 1977. 91. Power Sources 46. in Uhlig’s Corrosion Handbook. Kelly. M. Peters. in Lead–Acid Batteries: A Reference and Data Book. 2. Angres. Lead and lead alloys. and K. in Proceedings of the Fifth International Conference on Lead. National Association of Corrosion Engineers. Hampson. 13A. Materials Park.. 4. p. pp. Materials Park. 2003. India. and K. Goodwin. Ruetschi. 53. 1995. J. J. GENERAL REFERENCES S. J. Brussels. Pourbaix. Testing. Alhassan. Kelly. Vol. 7. 12. pp. 15. OH. p. . by R.29 APPENDIX 29. assuming total dissociation. Uhlig Copyright © 2008 John Wiley & Sons. By definition. and γ the activity coefficient. μ = μ° + RT ln a− + RT ln a+ = μ° + RT ln γ − M− + RT ln γ + M+ = μ° + 2 RT ln γ ± M Corrosion and Corrosion Control. For a binary electrolyte of molality M. Winston Revie and Herbert H. μ = μ° + RT ln a a = γM where γ → 1 as M → 0. μ° the partial molal free energy in the standard state (a = 1).1 ACTIVITY AND ACTIVITY COEFFICIENTS OF STRONG ELECTROLYTES Let μ be the partial molal free energy. a the activity. Inc. 451 . M the molality. pp. Calculate the emf of the following cell: Zn. Modern Electrochemistry. 0. 4 μ = μ° + RT ln γ 5 (2 M )2 (3M )3 ± = μ° + 5RT ln γ ± M (2 2 × 33)1/ 5 For the general case in which one mole of electrolyte dissociates into ν1 moles of cation and ν2 moles of anion.123 − = 2.1) Zn 2 + + 2e − → Zn Cl 2 + 2e − → 2Cl − Tentative reaction: (29.452 APPENDIX where M− = M+ = M. Bockris and A. Pt(Cl 2 .2) 0.) Activity coefficients are listed in Table 29.01M ).360 V (29. Ltd. and γ± is the mean ion activity coefficient (it is impossible to measure the activities of individual ions). 2nd edition. Vol. J.5 atm) γ ± for 0. O’M.01)(0.763 + 1. Plenum.71)[(0.01M ZnCl 2 = 0.1) Zn + Cl 2 → Zn 2 + + 2Cl − The corresponding Nernst equation is E = 0.1. Davies. Reddy. 1967. W.02)(0. we have La 2 (SO4 )3 → 2La 3+ + 3SO2 − 4 μ = μ° + 2 RT ln γ ± MLa 3+ + 3RT ln γ ± MSO2− 4 Since MLa 3+ = 2 M and MSO2− = 3M .. For an electrolyte such as lanthanum sulfate of molality M.0592 (Zn 2+ )(Cl − )2 log 2 pCl 2 (29. K. ZnCl 2 (0.2)– (29.71 (Table 29. Example. 1998.1) (29.29 V 0. N. George Newnes.5 .71)]2 log 2 0. New York.3) φ° = −0.360 − = 2. Electrochemistry.763 V φ° = 1. 1: Ionics. London. 251–257.0592 (0. the mean activity is given by v v a± = γ ± M (v11 v22 )1 /( v1+ v2 ) (See also: C. Activity Coefficients of Strong Electrolytes (M = molality) 453 .ACTIVIT Y AND ACTIVIT Y COEFFICIENTS OF STRONG ELECTROLY TES TABLE 29.1. 1.454 TABLE 29. Continued APPENDIX . ACTIVIT Y AND ACTIVIT Y COEFFICIENTS OF STRONG ELECTROLY TES 455 . 29. Ic. we have − I appl = I c − I a For anodic polarization.3) is spontaneous (ΔG is negative). Ia. is accompanied by decreasing cathodic current. Current relations are shown in Fig.1 for the corroding metal polarized as anode by means of an external current to a potential φ.4) Figure 29.5) (29. zinc is anode (−). 29. we obtain (29.456 APPENDIX [Reaction (29. Iappl = I a − I c Similarly. because of the relation. Polarization diagram for corroding metal polarized anodically from φcorr to φ. Increasing anodic current. Icorr. Also assume that concentration polarization and IR drop are negligible.2 DERIVATION OF STERN–GEARY EQUATION FOR CALCULATING CORROSION RATES FROM POLARIZATION DATA OBTAINED AT LOW CURRENT DENSITIES Assume that the corrosion current.1. . for cathodic polarization. occurs at a value within the Tafel region for both anodic and cathodic reactions. and platinum is cathode (+)]. for which Iappl changes sign. 6) where Aa is the fraction of area that is anode. .8) can be approximated by 1 1 Iappl = 2.3Δφ (29. if the metal is polarized an equal amount in the cathodic direction. Note that i0a in the Tafel equation refers to current per unit local area (not total area). we have Δφ = −βc log or I c = I corr 10 − Δφ/βc then Iappl = I corr (10 Δφ/βa − 10 − Δφ/βc ) Expressed as a series.3 x + (2.10) (29.9) where R is the polarization resistance. hence. we have 10 x = 1 + 2.7) If Δφ/βc and Δφ/βa are small.3 x)2 + 2! (29. determined experimentally by applying both anodic and cathodic currents to the electrode and measuring the polarization.8) and I a = I corr 10 Δφ/βa Ic I corr (29.10) is the slope of the φ versus I graph at the corrosion potential. Likewise.3R ⎜ βa + βc ⎟ 2. higher terms can be neglected and (29. (29.3I corr Δφ ⎛ + ⎞ ⎝ βc βa ⎠ or I corr = I appl ⎛ βaβc ⎞ 1 ⎛ βaβc ⎞ = ⎜ ⎠ ⎝ ⎠ ⎝ βa + βc ⎟ 2.DERIVATION OF STERN-GEARY EQUATION FOR CALCUL ATING CORROSION R ATES 457 φ − φcorr = Δφ = βa log = βa log Ia I corr Ia I − βa log corr ioa Aa ioa Aa (29. and i0a is the exchange current density for the anodic reaction. The value of R used in Eq. Icorr/Aa is the current density to which i0a applies. 10) is subject to error. and Δφ = I appl βc βa 2.3Δφ 2.2.12) indicating. Mansfeld and Oldham [1] considered this case and .458 APPENDIX Figure 29. This situation is equivalent to a large or infinite value of βc in (29. then i0a can be substituted for Icorr in Fig.10) is the Stern–Geary equation. the corrosion current equals the limiting diffusion current (Fig.2).11) If a noncorroding metal (minimum local action) that polarizes only slightly (high value of i0a) is polarized anodically at moderate current densities. 29. when it approaches either the reversible anode or cathode potential. for example.3R (29.1. “back-reactions” become appreciable and (29. Polarization diagram for metal corroding under control by oxygen depolarization.1 The General Equation If φcorr falls outside the Tafel region.2.3ioa βa + βc (29. Equation (29. (29. 29. Should the cathodic reaction be controlled by concentration polarization. as occurs in corrosion reactions controlled by oxygen depolarization.10) becomes I corr = βa Iappl β = a 2. Under these conditions. that anodic polarization of many metals at low current densities is a linear function of applied current. 29. as observed.10). 3.4).15) with respect to φ.3. and bcr = βcr/2. anodic polarization—results in I appl = I a − I ar − ( I c − I cr ) (29.14) ) ( ) ( ) ( ) According to electrode kinetics. (29. Letting Δφa = φcorr − φa and Δφc = φc − φcorr. choosing Iappl and Ia with the same sign—that is.3.15) becomes I corr = Ioa exp = I oc exp Δφa ⎡ − na F Δφa ⎤ 1 − exp ⎢ ⎥ ⎦ ba ⎣ RT c c ( ) ⎡1 − exp − n F Δφ ⎤ ( RT )⎦⎥ ⎢ ⎣ (29.13) where Iar and Icr are the back-reactions for the anodic and cathodic reactions.16) At the corrosion potential. we obtain 1 1 nF + = a ba bar RT and 1 1 nF + = c bc bcr RT where na and nc are constants related to the transfer coefficient. we have I appl = Ioa exp ( φ − φa φ − φa φ − φc φ − φc − I oa exp − − I oc exp − + I oc exp bcr ba bar bc (29. we get ∂I appl I oa bnF n F (φa − φ) ⎤ φ − φa ⎡ = exp 1 + a a − 1 exp a ⎥ ⎢ ⎦ ba ba ⎣ RT RT ∂φ I bnF n F (φ − φ c ) ⎤ φ −φ ⎡ + oc exp c 1 + c c − 1 exp c ⎢ ⎥ ⎦ bc bc ⎣ RT RT ( ) ( ( ) ( ( ) ( ) Δφc bc ) ( ) ( ) ) (29.17) (29.10) starts with modifying (29.3.18) . η = b ln(I/I0) and letting ba = βa/2.DERIVATION OF STERN-GEARY EQUATION FOR CALCUL ATING CORROSION R ATES 459 formulated the required modifications. Using these relations to eliminate bar and bcr. Because Iappl changes sign at potentials above compared to below φcorr. it follows that Iappl = Ioa exp φ − φa ⎡ n F (φa − φ) ⎤ 1 − exp a ⎢ ⎥ ⎦ ba ⎣ RT φ −φ ⎡ n F (φ − φ c ) ⎤ 1 − exp c − I oc exp c ⎢ ⎥ ⎣ ⎦ RT bc ( ( ) ) ( ( ) ) (29. respectively. Introducing the Tafel equation expressed in natural logarithms. bar = βar/2. The derivation of the more exact form of (29. Iappl = 0 and Ia = Ic = Icorr.15) Differentiating (29. bc = βc/2. we obtain 1 1 1 1 nF = I corr ⎛ + + − c ⎞ ⎜ ⎝ ba bc Δφc 2 RT ⎟ ⎠ R (29.21) If Δφc = 0.19) gives 1 1 nF ⎧1 = I corr ⎨ + + a R ⎩ ba bc RT ⎡exp na F Δφa − 1⎤ + nc F ⎢ ⎥ ⎣ ⎦ RT RT ⎡exp nc F Δφc − 1⎤ ⎫ (29.1 V.18) into the second part of (29. Δφa = 0.5.21) is 1/Δφc = 200 V−1. These correction terms are large compared to 1/ba + 1/bc = 46 V−1 (assuming βa = βc = 0. The corresponding correction term from (29.2 V−1.19) and (29. which is negligible.23) .1 V).. the corrosion potential is close to the reversible potential of the metal). and nc is assumed to be 0. whereas the approximated correction term in (29. T = 298 °K.16) becomes 1 ⎛ ∂I appl ⎞ I bnF − na F Δφa ⎤ Δφa ⎡ = oa exp 1 + a a − 1 exp =⎜ ⎟ ⎢ ⎥ ⎦ R ⎝ ∂φ ⎠ φ = φcorr ba ba ⎣ RT RT I Δφc + oc exp bc bc c c c c ( ) ( ) ( ( ) ( ( ) −1 ) ( ⎡1 + b n F − 1 exp − n F Δφ ⎤ ⎢ ⎥ ⎣ RT RT ) ⎦ ( ) (29.17) into the first part of (29. the correction term as given by (29.22) If ncΔφc < 2RT/F (i. the corrosion potential is close to the reversible potential of the reduction reaction). the corresponding correction term from (29.001 V−1. Additional approximations [1] that could be used if the corrosion potential is close to one of the reversible potentials are as follows.20) equals 0.20) is 3.20) to 1 1 1 1 1 ⎞ = I corr ⎛ + + + ⎜ ⎝ ba bc Δφa Δφc ⎟ ⎠ R (29.20) is 190 V−1. the corresponding correction term approaches infinity. If naΔφa < 2RT/F (i. On the other hand. Should Δφc = 0.460 APPENDIX and (29. If the correction terms involving Δφa and Δφc can be neglected. (29.19) Substituting (29.5. If at the same time.e.005 V.20) becomes the Stern–Geary equation. 1 1 1 1 nF = I corr ⎛ + + − a ⎞ ⎜ ⎝ ba bc Δφa 2 RT ⎟ ⎠ R (29. Small values of nFΔφ/RT allow the exponential terms to be expanded. simplifying (29.20) was developed by Mansfeld and Oldham [1] and is equivalent to the one derived originally by Wagner and Traud [2].e. then F/RT = 39 V−1 and ncFΔφc/RT = 0..1.20) ⎢ ⎥ ⎬ ⎣ ⎦ ⎭ RT ) −1 Equation (29. if either Δφa or Δφc approaches zero.5 V and na = 0. equivalents of added acid equal equivalents of carbonate and bicarbonate.6. (29.10). 29.20) are presented in Fig. Relative errors in corrosion currents calculated by use of (29. ba = RT/F.21). bc = 2 RT/F. The errors that result from each of these approximations to (29.3 [1]. depending on pH of the water: (alk) + (H + ) → 2(CO2 − ) + (HCO− ) + (OH − ) 3 3 (29.22). where na = 2.3.26) . (29. plus OH− or minus H+.20) [1] (with permission from Pergamon Press).2. nc = 1.3 DERIVATION OF EQUATION EXPRESSING THE SATURATION INDEX OF A NATURAL WATER In accord with the discussion of saturation index in Section 7.1.25) Assume that salts of weak acids other than carbonic acid are absent.DERIVATION OF EQUATION EXPRESSING THE SATUR ATION INDEX 461 Figure 29. 29. we have K s′ = (Ca 2 + )(CO2 − ) 3 K2 = ′ 2 (H + )(CO3 − ) − (HCO3 ) (29. Then when a water is titrated. instead of the exact equation (29. and φc − φa = Δφa + Δφc = 10 RT/F were assumed. and (29.24) (29.23). we obtain (CO2 − ) = 3 K2 (alk) ′ ( H + ) 1 + 2 K 2 /( H + ) ′ (29.30) into (29. we obtain (CO2 − ) = 3 From Eq.5 z2 μ1/2. (29. −log γ = 0.30) and substituting (29.25). (CO2 − ) = 3 Therefore. then K s = K s′γ 2 .29) into (29. if Ks is the true activity product. Hence.29) was derived by Langelier [3] on the assumption that K s′ and K 2 are based on concentrations (moles/liter) rather than on activities. we obtain . or titratable equivalents of base per liter (by titrating with acid) obtained by using methyl orange as indicator.24).5 and 10. Concentrations of (H+) and (OH−) are small between pH values 4.462 APPENDIX where (alk) = alkalinity. From Eq. where γ± refers to the mean ion activity coefficient for CaCO3. but also with total dissolved ′ solids because of the effect of ionic strength of a solution on activities of specific ions.3 and may be neglected. The ± activity coefficient was approximated by Langelier using the Debye–Hückel theory. we get (Ca 2 + ) K2 (alk) ′ = K s′ ( H + ) 1 + 2 K 2 /( H + ) ′ (29. (29.31) Taking logarithms of both sides and using the notation log 1/α = pα. Substituting (29.28) K 2 (HCO− ) ′ 3 (H + ) (29. (HCO− ) = 3 (alk) 1 + 2 K 2 /( H + ) ′ (29.24). Accordingly.27).29) (alk) − (HCO− ) 3 2 (29. where μ is the ionic strength and z is the valence.27) Equation (29.26). referring to (29. concentrations of CO2− and HCO− obtained by titration can be equated to 3 3 corresponding concentrations of these species in K s′ and K 2 . ′ For example. K s′ ′ and K 2 vary not only with temperature. 13.3 2K2 ′ : (H + )s ( 2K2 ′ (H+ )s ) as a Function of pH 10.05 ( ) 0. and the temperature.7 0.2. 2. To use the chart.4 0.g. 29. the actual pH of water at the higher temperature should be used. at 25°C.5.8 × 10 −11 . 20°C. 29.09 9. 1. The saturation index is then the algebraic difference between the measured pH of a water and the computed pHs: Saturation index = pH measured − pH s (29.29 s 9. 1. total dissolved solids in ppm.5 [5].48. or MgSO4).19. we must know the alkalinity of a water and calcium ion concentration calculated as ppm CaCO3..33) To calculate the saturation index at above-room temperature.32) where pHs is the pH of a given water at which solid CaCO3 is in equilibrium with its saturated solution. A chart for the same purpose as prepared by Powell. the value is 2. 50°C.1 0.2. This effect increases values of ( pK 2 − pK s′ ). typical values as a function of pHs in the alkaline ′ range at 25°C are given in Table 29. . which gives values for two waters of differing alkalinity. the increasing ionic strength of the solution depresses the activity of other ions in solution.47 2K2 ⎞ ′ pH s = pK 2 − pK s′ + p(Ca 2 + ) + p(alk) + log ⎛ 1 + ′ ⎜ ⎟ ⎝ (H + )s ⎠ (29. 25°C. ′ 2. A nomogram for obtaining pHs of a water at various temperatures and dissolved-solids content was constructed by C.54. Na2SO4. Bacon. This can be estimated from the room temperature value by using Fig. at a total dissolved-solids content of 100 ppm.17 9. and Lill [5] is reproduced in Fig. For ′ example.0 0.DERIVATION OF EQUATION EXPRESSING THE SATUR ATION INDEX 463 TABLE 29.96. Values of ( pK 2 − pK s′ ) decrease with increasing temperature as follows: 0°C. and for 500 ppm it is 2.04. Values of log 1 + pHs: log 1 + 10. Hoover [4]. The last term is ordinarily small and can be omitted when pHs < 9. In the presence of other salts (e. Based on the value K 2 = 4.4. NaCl. and temperature is expressed in °F. (“Ca” and “alkalinity” are expressed as ppm CaCO3. Bacon. Chart for calculating saturation index (Powell.464 APPENDIX Figure 29. and Lill [5]).) .4. 5. . Bacon.DERIVATION OF EQUATION EXPRESSING THE SATUR ATION INDEX 465 Figure 29. (b) For water of 100 ppm alkalinity (methyl orange end point). Values of pH of water at elevated temperatures (Powell. (a) For water at 25 ppm alkalinity (methyl orange end point). and Lill [5]). 5. Continued .466 APPENDIX Figure 29. or Ex = k1ix ′ (29.6.DERIVATION OF POTENTIAL CHANGE ALONG A CPP 467 29. the potential change along the pipe is given by dEx = − RL I x dx (29. ix is the current density at pipe surface at distance x from point of bonding. true current density increases with z. we obtain d 2 Ex = RL (2 πrix ) dx 2 (29.35). By Ohm’s law. or Figure 29.36) At small values of ix. and z is the resistance of pipe coating per unit area. RL is the resistance of metallic pipe per unit length. Combining (29.37) Assuming that the resistance z per unit area of coating is inversely proportional to the total cross-sectional area of pores per unit area.35) (29. Ix is the total current in pipe at distance x.34) where RL is the resistance of metallic pipe per unit length. in the pipe per unit length at x is equal to the total current entering the pipe at x. . r is the radius of pipe.34) and (29. Sketch of buried pipe cathodically protected by anodes distance a apart. 29. or dI x = −2πrix dx where r is the radius of the pipe. Then the change of current. Ix. returning to the anode by way of the pipe (Fig.6).4 DERIVATION OF POTENTIAL CHANGE ALONG A CATHODICALLY PROTECTED PIPELINE Assume that current enters a pipe from the soil side through a porous insulating coating. E is the difference between measured and corrosion potentials of pipe. polarization of the pipe surface is a linear function of the ′ true current density ix at the base of pores in the coating. and the potential Ex at a/2 = EB.39) ( ) (29.37) and (29.45) . or half the distance to the next point of bonding (see Fig.38).42) ( ) 1/ 2 a⎤ 2⎥ ⎦ (29. if the length of pipeline to be protected is a/2. On the other hand.38) (29.35).39) in (29. Introducing this boundary condition.41) into (29.40) ( ) 1/ 2 ⎤ x⎥ ⎦ (29.44) The total current at x = 0 (sometimes called drainage current) multiplied by 2. 29.36). we obtain d 2 Ex R 2 πr Ex = L 2 dx kz For an infinite pipeline.43) The current in an infinite pipeline at any point x is given by substituting the derivative of (29. or (dEx/dx)x=a/2 = 0. we obtain ⎡ 2 πrRL Ex = EB cosh ⎢ ⎣ kz and ⎡ 2 πrRL EA = EB cosh ⎢ − kz ⎣ ( ) ( x − a )⎤⎥⎦ 2 1/ 2 (29. Substituting (29. we get Ex = kzix where k = k1k2. or Ix = rR ( 2πkz ) L 1/ 2 EA ⎡ 2 πrRL exp ⎢ − RL kz ⎣ ( ) 1/ 2 ⎤ x⎥ ⎦ (29.6). to account for current flowing from either side of the pipeline to the point of bonding.41) for the boundary conditions Ex = 0 at x = ∞. and Ex = EA at x = 0. this differential equation has the solution ⎡ 2 πrRL Ex = EA exp ⎢ − kz ⎣ (29. equals IA = 2 EA 2 πrRL RL kz ( ) 1/ 2 (29. it follows that current in the pipeline at a/2 = 0.468 APPENDIX ix = k2 zix ′ Combining (29. for a finite pipe.48) Figure 29. . Then the component of current density along the soil surface equals iy = 2j y 2πR R (29. less current flows in the h direction toward the soil surface than in other directions.7). Sketch of buried pipe at which current enters or leaves.47) But for a pipe buried h cm below the soil surface.46) 29.DERIVATION OF THE EQUATION FOR POTENTIAL DROP ALONG THE SSC 469 Similarly. i= j 2πR (29. causing potential drop along soil surface.5 DERIVATION OF THE EQUATION FOR POTENTIAL DROP ALONG THE SOIL SURFACE CREATED BY CURRENT ENTERING OR LEAVING A BURIED PIPE For a deeply buried pipe. The computation for current flow in this situation is approximated by assuming an image pipe located h cm above the soil surface that supplies an equal current j (Fig. 29.7. we have IA = ( )( 2 EB RL 2 πrRL kz ) 1/ 2 ⎡ a 2 πrRL sinh ⎢ kz ⎣2 ( ) 1/ 2 ⎤ ⎥ ⎦ (29. 300 (1952). [See also. are at a distance 3a apart. Two reference electrodes.51) where Δφ is in volts. C and D.3 log 101 = 0. By Ohm’s law. are located distance a apart and also a from A and B.6 DERIVATION OF THE EQUATION FOR DETERMINING RESISTIVITY OF SOIL BY FOUR-ELECTRODE METHOD Two metal electrodes.] 29. Corrosion 8. and j in A/cm. R.8). we have Δφ = ∫0 y (29. .49) y h 2 + y2 ρj ρj y ρj ydy = ∫0 dy = ln 2π h2 πR 2 π h 2 + y2 (29. Figure 29. Δφ = ρj 2. respectively (see Fig. 29. A and B. Howell. or y = 10h.50) If we take y = 10 times distance h.734ρj 2π (29. Sketch of four-electrode arrangement for determining soil resistivity. ρ in Ω-cm.470 APPENDIX The same result can be obtained by assuming half-cylindrical current distribution in the soil multiplied by the factor y/R to account for zero current flowing to the pipe in the h direction (y = 0) and an increasing current density iy as distance y increases.8. the potential gradient along the soil surface is dφ j yρ = dy πR R Since R2 = h2 + y2. In the fretting process. gas from the atmosphere adsorbs rapidly. By Ohm’s law. in turn. Section 8.54) Similarly.52) ∫ 2 dr = 2π r − r 2 1 r1 2 πr r2 Iρ − Iρ 1 ( 1 ) (29. Behind each asperity on the track of clean metal. current density in the soil = I/2πr2. Hence.DERIVATION OF THE EQUATION EXPRESSING WEIGHT LOSS 471 Assuming hemispherical symmetry at distance r from A or B.1). scrapes the oxide film off and leaves.55) [See also S. The next asperity. the potential difference Δφ2 between C and D because of current entering electrode B is Iρ/4πa. The average time during which oxidation occurs is t. Δφtotal = Δφ1 + Δφ 2 = Iρ 2 πa or ρ = 2 πa Δφ I (29.56) The corresponding amount of oxide W removed by one asperity plowing out a path l long and c wide depends on the amount of oxide formed in time t. 221 (1953). for mathematical convenience. Ewing and J. a track of clean metal behind. with each asperity plowing out a path of clean metal of width averaging c and of length depending on distance of travel.7 DERIVATION OF THE EQUATION EXPRESSING WEIGHT LOSS BY FRETTING CORROSION Assume n asperities or contact points per unit area of metal (or oxide) surface which. the asperities move over a plane surface of metal at linear velocity υ.4. Hutchinson. Then t= s υ (29. the logarithmic equation is obeyed (see Section 11.7. moving in the same path as the first.1): . 8.21. dφ I = ρ dr 2 πr 2 Δφ = (29. For thin-film oxidation. followed in time by formation of a thin oxide film. The average diameter of asperities is c and the average distance apart is s (Fig. Corrosion 9. are circular in shape.53) Because current leaves A.] 29. the potential difference Δφ1 between C and D is Δφ1 = − Iρ 1 1 Iρ − = 2 π 2a a 4 πa ( ) (29. 57). the linear displacement from the midpoint of travel at time t. a reaction that is activated mechanically by asperities moving over the metal surface. we can also consider the situation of oxygen adsorbing rapidly as physically adsorbed gas. A linear rate of gas adsorption suggests that the amount of oxygen reaching the clean metal surface as physically adsorbed gas may actually be controlling. the average velocity is given by *This argument is not compelling. followed by conversion at a slower rate to chemisorbed oxygen atoms. reacts with underlying metal to form metal oxide.57) [6]. corresponding to increased rate and extent of physical adsorption at lower temperatures. 2l is the total length of travel in any one cycle and x. The chemisorbed oxygen. in turn.57) where τ and k are constants. Chemisorption limits the amount of oxide that is formed in such a process.61) Therefore. Hence. rather than its conversion to chemisorbed oxygen atoms.* Substituting (29. Preliminary to oxidation. this is related to constant angular velocity by the expression dθ = 2 πf dt (29. because the logarithmic term is eventually expanded and only the first term is used. the rate of chemisorption following an equation identical in form to that of (29. is given by x= and dx l dθ = υ = − sin θ dt 2 dt (29.60) l cosθ 2 (29. The rate of chemisorption. whichever process applies.472 APPENDIX W = clk ln ( τt + 1) (29. This is equivalent to a linear rate of oxidation or of gas adsorption with time.56) into (29. usually decreases as the temperature is lowered.59) If f represents constant linear frequency. . we obtain W = clk ln s ( υτ ) + 1 (29. on the other hand. This possibility is given support by the observed increase of fretting weight loss as the temperature is lowered. the form of the final equation is essentially the same.58) Assuming that relative motion of the two surfaces is sinusoidal. This condition applies particularly to high loads (small values of s). for n contacts or asperities per unit area of interface.62) Hence. the total area of contact. high-frequency f. Therefore.63). l = 0. The term pm is approximated by three times the elastic limit. and large value of slip l. is important now because of the “tearing out” or welding action taking place during mechanical wear. weight loss W per cycle caused solely by oxidation is s Wcorr = 2 nlck ln ⎛ + 1⎞ ⎝ 2lf τ ⎠ (29. rather than the width. Empirical values of the constant τ for iron range from 0. hence. weight loss per cycle is given by Wmechanical = 2k ′n ( ) πl c 2 2 (29.000 psi.06 s. is equal [7] to the load L divided by the yield pressure pm. the logarithmic term can be expanded according to ln( x + 1) = x − x2 x3 + − 2 3 where x is equal to s/(2lfτ). term and the mechanical term Wtotal = Wcorr + Wmechanical (29. in contrast to scraping off of chemical products from the surface.01 cm. as discussed previously. for mild steel. on the average. and s (distance between .63) To this must be added loss of metal by wear because each asperity. Wmechanical = 2k ′ lL = k2 lL pm (29.66) Returning to (29. When the latter expression is much smaller than unity. Assuming reasonable values for τ = 0. For n circular asperities.06 to 3 s. f = 10 cps.DERIVATION OF THE EQUATION EXPRESSING WEIGHT LOSS 473 V=− πlf ∫ sin θdθ 0 π π = 2lf (29. or corrosion. The area of asperity. A shearing off of asperities without welding also leads to wear that depends on total area of contact.64) But nπ(c/2)2.65) where k2 is a constant equal to 2k′/pm. digs below the oxide layer and dissipates metal in an amount proportional to the area of contact of the asperities and the length of travel. The total wear or metal loss per cycle is the sum of the oxidation. pm equals 100 kg/mm2 or 140. the square and higher terms can be omitted. The linear rate reasonably approximates the actual state of affairs for very short times of adsorption or oxidation.70 8.0431 0. a linear rate of oxidation or of gas adsorption on clean iron.2 0. Since the number of asperities along one edge of unit area is equal to n .5 24. whenever experimental conditions approximate those cited and higher terms of the logarithmic expansion can be neglected.0422 0.87 7. Therefore.7 23.365/Density 0. and (29.74 × Density 7.57 8.75 8. (29. then s/(2lfτ) = 0.5) Density (g/cm3) 2. recalling that nπ(c/2)2 = L/pm.0408 0.67). from the very start.365/density to obtain mm/y. c.74 × density (g/cm3) to obtain grams per square meter per day (gmd).95 7.68) Combining (29. where k/τ is the reaction-rate constant.67) This expression is equivalent to assuming. or Wcorr = where k0 = k pm π τ 2 and k1 = 4 k pm π τ k0 L1/ 2 k1 L − f f (29.2 23.0 23.6 19.135 0.65).96 8. we have the final expression for fretting as measured by weight loss corresponding to a total of C cycles: Wtotal = (k0 L1/ 2 − k1 L) C + k2 lLC f (29.0521 .0407 0. Also.474 APPENDIX asperities) = 10−4 cm. we obtain Wcorr = ncks fτ (29. Metal Aluminum Brass (red) Brass (yellow) Cadmium Columbium (niobium) Copper Copper–Nickel (70–30) Iron Iron–silicon (Duriron) (84–14.65 8.0417 0.66).6 24.0426 0.008.68). Multiply gmd by 0.0 2. the terms n.47 8.5 21.69) 29. and s can be eliminated from (29.0464 0. it follows that s + c is approximated by 1/ n .8 CONVERSION FACTORS Multiply millimeters penetration per year (mm/y) by 2.40 24. CONVERSION FACTORS 475 Metal Lead (chemical) Magnesium Nickel Nickel–copper (Monel) (70–30) Silver Tantalum Tin Titanium Zinc Zirconium Density (g/cm3) 11.35 1.74 8.90 8.84 10.49 16.6 7.30 4.51 7.13 6.49 2.74 × Density 31.1 4.77 24.4 24.2 28.7 45.5 20.0 12.4 19.5 17.8 0.365/Density 0.0322 0.210 0.0410 0.0413 0.0348 0.0220 0.0500 0.0809 0.0512 0.0562 Multiply inches penetration per year (ipy) by 696 × density to obtain milligrams per square decimeter per day (mdd). Multiply mdd by 0.00144/density to obtain ipy. 29.8.1 Additional Conversion Factors Multiply ipy mdd mA/cm2 A/m2 psi Å by 25.4 0.1 10 0.093 6.89 × 103 0.1 to obtain mm/y gmd A/m2 A/ft2 Pa (pascals) nm (nanometer) 29.8.2 Current Density Equivalent to a Corrosion Rate of 1 gmd 1 gmd = 1.117n / W amperes per square meter where W is the gram atomic weighta Reaction Al → Al3+ + 3e− Cd → Cd2+ + 2e− Cu → Cu2+ + 2e− Fe → Fe2+ + 2e− Fe → Fe3+ + 3e− Pb → Pb2+ + 2e− Mg → Mg2+ + 2e− Ni → Ni2+ + 2e− Sn → Sn2+ + 2e− Zn → Zn2+ + 2e− a A/m2 0.124 0.0199 0.0352 0.0400 0.0600 0.0108 0.0919 0.0381 0.0188 0.0342 Multiply A/m2 by 0.1 to obtain mA/cm2. 476 APPENDIX 29.9 STANDARD POTENTIALS 25°C, see also Section 3.8. φ° (volts) FeO2 − + 8 H + + 3e − → Fe3 + + 4 H 2O 4 Co3+ + e− → Co2+ PbO2 + SO2 − + 4 H + + 2e − → PbSO4 + 2 H 2O 4 NiO2 + 4H+ + 2e− → Ni2+ + 2H2O Mn3+ + e− → Mn2+ PbO2 + 4H+ + 2e− → Pb2+ + 2H2O Cl2 + 2e− → 2Cl− Cr2O2 − + 14 H + + 6e − → 2Cr 3+ + 7 H 2O 7 O2 + 4H+ + 4e− → 2H2O Br2(l) + 2e− → 2Br− Fe3+ + e− → Fe2+ O2 + 2H+ + 2e− → H2O2 I2 + 2e− → 2I− O2 + 2H2O + 4e− → 4OH− Hg2Cl2 + 2e− → 2Hg + 2Cl− AgCl + e− → Ag + Cl− SO2 − + 4 H + + 2e − → H 2SO3 +H 2O 4 Cu2+ + e− → Cu+ Sn4+ + 2e− → Sn2+ AgBr + e− → Ag + Br− Cu(NH3)2+ + e− → Cu + 2NH3 Ag(CN)− + e − → Ag + 2CN − 2 PbSO4 + 2e − → Pb + SO2 − 4 HPbO− + H 2O + e − → Pb + 3OH − 2 2H2O + 2e− → H2 + 2OH− Zn(NH 3 )2 + + 2e − → Zn + 4 NH 3 4 a 1.9 1.82 1.685 1.68 1.51 1.455 1.3595 1.33 1.229a 1.0652 0.771 0.682 0.5355 0.401 0.2676 0.222 0.17 0.153 0.15 0.095 −0.12 −0.31 −0.356 −0.54 −0.828 −1.03 At pH 7, 0.2 atm O2, φ = 0.81 V. Source: Data from Oxidation Potentials, W. Latimer, Prentice-Hall, Englewood Cliffs, NJ, 1952 (with permission). 29.10 NOTATION AND ABBREVIATIONS activity activity coefficient ampere angstrom atmosphere atom percent calorie a γ A Å atm at.% cal NOTATION AND ABBREVIATIONS 477 centimeter corrosion current corrosion current density corrosion potential corrosion potential of a galvanic couple coulomb critical crack depth current current density decimeter density electromotive force (emf) equivalent exchange current density faraday foot Flade potential gallon Gibbs free energy gram gram calorie grams per square meter per day hour inch inch penetration per year joule kilogram kilometer liter mean ion activity coefficient megapascal meter microampere micrometer (micron) milliampere milligram milligrams per square decimenter per day milliliter millimeter millimeters penetration per year millivolt minute molal solution nanometer normal solution cm Icorr icorr φcorr φgalv C acr I i dm d E eq i0 F ft φF gal G g cal gmd h in. ipy J kg km L γ± MPa m μA μm mA mg mdd mL mm mm/y mV min M nm N 478 APPENDIX ohm ounce overpotential overvoltage parts per million pascal potential potential on the standard hydrogen scale pound pound per square inch pressure relative humidity saturated calomel electrode second standard hydrogen electrode stress-corrosion cracking Tafel slope volt weight percent yield strength Ω oz η ε ppm Pa φ φH, φ(S.H.E.) lb psi p R.H. S.C.E. s S.H.E. S.C.C. β V wt.% Y.S. REFERENCES 1. F. Mansfeld and K. Oldham, Corros. Sci. 11, 787 (1971); F. Mansfeld, in Advances in Corrosion Science and Technology, Vol. 6, M. Fontana and R. Staehle, editors, Plenum Press, New York, 1976, p. 163. 2. C. Wagner and W. Traud, Z. Elektrochem. 44, 391 (1938). 3. W. Langelier, J. Am. Water Works Assoc. 28, 1500 (1936). 4. C. Hoover, J. Am. Water Works Assoc. 30, 1802 (1938). 5. S. Powell, H. Bacon, and J. Lill, Eng. Chem. 37, 842 (1945). 6. F. Stone, in Chemistry of the Solid State, W. Garner, editor, Butterworths, London, 1955, p. 385. 7. F. P. Bowden and D. Tabor, The Friction and Lubrication of Solids, Oxford University Press, New York, 1950. INDEX KEY: T refers to tables. F refers to figures. A Abbreviations, 476–478 Activity, coefficient: defi nition, 24 rules, 451–452 tables, 453–455 Aircraft: aluminum alloys in, 395 Boeing 777, 395 fretting corrosion, 182 hot corrosion, 229–230 microbiologically influenced corrosion, 388–389 stress-corrosion cracking, 395 titanium, 425 Aluminum, 383–398 anodic protection, anodizing, 287–288, 296 atmospheric corrosion, 194T, 383, 391 cathodic protection of, 252, 260, 261, 384, 395 chlorinated solvents, corrosion in, 389–391 inhibitors for, 390 mechanism of, 390–391 water, effect of, 390F coatings, 280–281 corrosion characteristics, 384–392 by microbiological constituents in fuel, 388–389 galvanic coupling of, 392–393 nitric acid, corrosion rates in, 392 oxide coatings on, 287–288, 383, 386 painting of, 296 passivity of, 83, 383, 384, 386, 388 soil corrosion, 208 steam, corrosion in, 384 vapor degreasing of, 293 waters, corrosion in, 384–386 pH, effect of, 387 pitting, 98, 99 Aluminum alloys, 393–395 anodes in impressed current cathodic protection, 259 clad, 384, 395 compositions, 385T exfoliation of, 394 intergranular corrosion, Al–Cu alloys, 393–394 sacrificial anodes, 254–255 stress-corrosion cracking, 384, 394–395 Amorphous alloys, 335 Amphoteric metals, 12, 47, 79, 245, 260, 279, 446 Anion, defi nition, 13 Anode: defi nition, 11–13 materials, for cathodic protection, 258–260 sacrificial, 144, 251, 252, 254–257, 263, 354, 370, 378 Anode–cathode area ratio, 56, 67–69, 71–75, 123, 132, 208–210 magnesium, galvanic corrosion of, 402 tantalum, 442 Anodic protection, 263–264 by alloying, 90–92 of aluminum, 264, 386 theory of, 90–92 Anodizing: aluminum, 287–288, 296 magnesium, 403 Atmospheric corrosion, 191–204 avoiding, 201–202 Corrosion and Corrosion Control, by R. Winston Revie and Herbert H. Uhlig Copyright © 2008 John Wiley & Sons, Inc. 479 480 INDEX comparison with seawater and soils, 195T corrosion product fi lms, protection by, 192–195 corrosion rates, steels, 193F alloying, effect of, 202T various metals, 194T critical humidity, 199–200 factors influencing, 195–200 gases in atmosphere, effect of, 197–199 sulfur dioxide content of air, 197–198 moisture, effect of, 199–200 particulate matter, effect of, 196–197 time of wetness, 199 types of atmospheres, 192 Automobiles, 4, 196, 295 aluminum alloys in, 394–395 body panels: electrogalvanizing, 270 painting, 295 phosphate coating, 287 engines, 3 Ford, 404 fretting corrosion, 182 Honda, 395 lead in, 447 magnesium in, 399, 404 nickel coatings, 272 mufflers, 2, 280 painting, 295 stainless steels in, 341 Volkswagen, 399 B Backfi ll, for cathodic protection, 255, 259 Bacteria, 4, 118–120, 206–207, 279, 298, 309, 367, 388–389 Blister, 66 Blister cracking, 170–172 Boiler corrosion, 323–330 alkali addition, effect of, 327–328 embrittlement detector, 324–325 inhibitors for, 328 mechanisms of, 328–330 oxygen, dissolved, effect of, 329 oxygen scavengers, 318–319, 330 Boiler water, treatment of, 326–328 alkali addition, 327–328 carbon dioxide removal, 326–327 inhibitor addition, 328 oxygen removal, 326 phosphate addition, 328 Brasses, 371–378 naval, red, yellow, 371–372 see also Copper–zinc alloys Bridges, 2 cathodic protection for, 253 steel reinforcements in, 144 failure by stress-corrosion cracking, 153, 169–170 painting, 290 weathering steels for, 201 Bronzes, 13, 252, 270, 371–372 C Cadmium, 70–71, 168, 178, 184, 198, 231, 264, 271, 274, 276–277, 288, 314, 392–393, toxic properties of, 277, 392–393 Catalytic properties and passivity, 108–109 hydrogen overpotential, relation to, 65 Cathode, defi nition, 11–13 Cathodic protection, 251–267 anode materials for, 258–260 backfi ll, use of, 255, 259 coatings, combined use with, 255–257 criteria of protection, 260–263 doubtful, 262 potential measurements, 260–262 current, magnitude required, 77–79, 257–258 economics of, 263 history of, 252–253 how applied, 253–254 overprotection, effect of, 259–260 pitting, used to avoid, 261–262, 354 potential decay along pipeline, 257 derivation of formula for, 467–469 steel reinforcements in concrete, 144, 253 stress-corrosion cracking, to avoid: in aluminum alloys, 384, 395 in brasses, 378 in stainless steels, 360 in steel, 155–156, 162 in titanium alloys, theory of, 77–79 Cation, defi nition, 13 Caustic embrittlement, 152 Cavitation-erosion, 130–131 in diesel engine liner, 131F Cells: Daniell, 54–55 galvanic, 11–12, 22, 26, 54, 68, 84, 115, 143, 157, 197, 207, 254, 280, 343, 369 local-action, 10–11, 68, 142, 261 passive–active, 96, 280, 352 reference, 34–37 calomel, 35–36 copper–copper sulfate, 36–37 silver–silver chloride, 36–37 INDEX 481 types, in corrosion reactions, 13–15 concentration, 13–14 differential aeration, 13–15, 28–30, 164, 197, 298–299, 329, 353, 370, 371, 384, 388, 430 differential temperature, 14–15 dissimilar electrode, 13 Cement, Portland, as cause of corrosion: of aluminum, 388, 391 of lead, 446 as protective coating for steel, 286 steel reinforcements, failure of, 143–145 galvanic effects of, 144 Chromates, see Passivators Chromium–iron alloys: critical composition for passivity, 100–101 critical current densities for passivity, 101F oxidation, elevated temperatures, 232–235 effect of yttrium, 234–235 maximum temperatures in air, 235T potentials of, in 4% NaCl, 101F Flade, 88F see also Stainless steels Chromium–nickel alloys: commercial alloys, composition, 413T corrosion characteristics, 411–417 critical composition for passivity, 102 intergranular corrosion, 415 oxidation, elevated temperatures, 226T, 236–237 Clean Air Act, 191, 198, 289, 295 Coatings: compliant, 289 inorganic, 285–288 chemical conversion, 286–288 chromate, 288 oxide, 287–288 phosphate, 286–287 as preparation for painting, 295–296 Portland cement, 286 vitreous enamel, 285–286 metallic, 269–283 aluminum, 280–281 application, methods of, 269–271 cadmium, 276–277 chromium, 279–280 classification, 271–272 lead, 274 nickel, 272–274 electroless, 273–274 fogging of, 273 thickness of, 272–273 tin, 277–279 zinc, 274–276 atmospheric corrosion, 198 seawater corrosion, 275 soil corrosion, 208 polarity reversal in hot waters, 276 see also Steel, galvanized organic, 289–301 cathodic protection, combined use with, 255–257 filiform corrosion, 296–299 pigments, inhibiting, 291–292 plastic, 299–300 prime coat, 291–292, 294, 295, 296, 299, requirements for, 291–292 rules for applying, 295 surface preparation, 293–295 effect on life, 295T Cobalt and cobalt alloys, 419–423 cavitation-erosion, 420, 423 compositions, 422T corrosion characteristics, 419–423 fretting corrosion, 420, 423 Cold working, effect on corrosion: iron and steel, 149–150 nickel, 150 pure iron, 150F Complexes, chemical, effect on Emf, 30–32 Concrete, 143–145, 253, 388, 391 Condenser tube alloys, 378–379 Converson factors: corrosion rates, 474–475T to current densities, 475T Copper, 367–381 atmospheric corrosion, 194T, 195T, 197–199 blue staining, caused by, 321, 369 corrosion characteristics, 367–371 deposit attack, 370 hydrogen, reaction with, 231–232 impingement attack, 17, 368–372, 379, 411 inhibitor for tarnish of, 314 oxidation, high temperature, 230–232 pitting of, 369–370, 371, 378 seawater corrosion, 195T, soil corrosion, 195T, 208, 209T stress-corrosion cracking, 371, 374 waters, corrosion by, 369–371 Copper–nickel alloys, for condenser tubes, 378–379 see also Nickel alloys, copper–nickel, Copper–zinc alloys, 371–378 dezincification, 17, 252, 334, 371–374 effect of Zn content, 372F mechanism of, 334, 374 impingement attack, effect of Zn content, 372F 482 INDEX stress-corrosion cracking, 374–378 effect of applied potential, 377F effect of Zn content, 372F mercurous nitrate test, 376 prevention, 378 Corrosion: alloying, effect of, 333–335 cost, 2–5 current and corrosion rate, 56, 475T importance of, 2–5 direct and indirect losses, 2–5 products, protective, in atmospheric corrosion, 192–195 rates, calculation of, from polarization data, 71–73, 456–461 classification for handling chemical media, 16 control of, anodic, cathodic, 68–69 derivation of formula for, 456–461 polarization, influence on, 68–71 scientist and engineer, 1–2 tendency and free energy, 21–22 types, 15–18 Corrosion-erosion, defi nition of, 17 Corrosion fatigue, 173–180 appearance of, 174 (Photo) defi nition, 18, 173–174 effect of oxygen, moisture, 175 mechanism of, 179–180 minimum corrosion rate for, 178T prevention, 178–179 strength, values of, 177T Crazing, 286 Crevice corrosion, 14, 29, 99, 117, 173, 342, 350–351, 352–354, 361, 412, 416–417, 429–430, 432 Critical pitting potential, 97–99, 160, 264, 351– 354, 384–386, 421, 429, 437 Critical pitting temperature, 99–100, 351–352 Crystal face, effect on corrosion, 13, 149 effect on oxidation, 230–231, 232 Current density: critical for passivity, 85–86, 90, 117, 242, 411 for Cr–Fe alloys, 101F for Ni–Cu alloys, 104F exchange, 61 limiting, 58–60 passive, 85 for Ni–Cu alloys, 104F D Deactivation of water: defi nition, 317 by hydrazine, 318–319 by ion-exchange resins, 326 by sodium sulfite, 318, 326 Dealloying, 17, 118, 172, 372–374 Deaeration of water, 178, 317–319, 326, 402, 414 defi nition, 319 oxygen concentration, maximum allowable for steel systems, 320T Depolarization: in acids, 124–125 by oxygen, 72, 116, 122, 124, 131–133, 139, 262, 458 Deposit attack, 370 Dezincification: Cu–Zn alloys, 334, 371–374 defi nition, 17 mechanism, 374 Diesel engines, 130, 215, 322, E Electrochemical equivalent, 10 Electrochemical impedance spectroscopy, 75–77 Electrochemical mechanism: of corrosion, 9–18, 22, 70–71, 143 for corrosion of iron and steel, 115–116 of passivation, 89 of pitting, 96, 351–353 of stress-corrosion cracking, 157–158, 190, 377 Electrode: copper–copper sulfate, 36–37 hydrogen, 24–25 oxygen, 28–30 reference, 34–38 standard or normal hydrogen, 24–25 Electron configuration theory, 102–108 Emf: calculation of, 25–28 relation to Gibbs free energy, 22 Series, defi nition of, 30 Table, 31T Endurance limit, 18, 141 Environmental Protection Agency (EPA), 191, 196, 369, 447 Exfoliation, 394 F Faraday, value of, 22 Faraday’s law, 9–11, 56, 85, 178, 244, 259, Fatigue, affected by air, moisture, 175 limit, defi nition of, 18, 173 strength, defi nition of, 173 see also Corrosion fatigue INDEX 483 Filiform corrosion, 279, 296–299 photos, 297 theory of, 298–299 Films, oxide: isolation of, 94 passive, oxidizing properties, 93–94 thickness of: on aluminum, 383 on iron, 93–94 Flade potential: aluminum, 383 defi nition, 86 chromium, 279 chromium–iron alloys, 88F iron, 47, 84–89, 93–94, 112 nickel, 410 nickel–copper alloys, 106 nickel–molybdenum alloys, 412 tantalum, 441 titanium, 426 Fogging, 16, 273, 411 Fouling, marine: Cu 2+ concentration to prevent, 103–105, 252, 369 of Cu–Ni alloys, 103–105, 104F of stainless steel, 351 Fracture mechanics, 162–166 Fretting corrosion, 180–184 aircraft, 182 automobiles, 182 cobalt alloys, 420, 423 defi nition, 17, 180–181 equation for weight loss, 183 derivation of, 471–474 factors causing, of steel mechanism, 182–184 nuclear reactors, 182 prevention, 184 railroad car wheels, 182 Fugacity, 24 Furnace windings, 236, 237 G Galvanic coupling: of aluminum, 391–393 in aerated media, 127 in deaerasted media, 127–129 of different steels, 141–142 in soils, 206 effects of, at elevated temperatures, 227–228 hydrogen cracking of stainless steels, caused by, 357 hydrogen embrittlement of Ta, caused by, iron–gold, 306 in natural waters and soils, 141–142 and passivity of stainless steels, 361 Galvanic Series: defi nition, 32 in seawater, 33T Galvanized steel, see Steel, galvanized Glass, cavitation–erosion, 130 coatings, 256, 285–286 electroless nickel coatings on, 274 stress-sorption cracking, 160 Glass electrode, 25 Glassy alloys, 335 Gmd, conversion factors of mm/y, 474–475 defi nition, 16 Gold, 17, 22, 109, 227, 232, 269–270, 286, 334, 372–374, coatings, electrical contacts, 182 coupled to iron, 306 fatigue, 175 Graphitic corrosion, 138, 142, 206 Gun barrels, protecting, 287 H Humidity: critical in atmospheric corrosion, 196, 199–201 and fi liform corrosion, 298 and growth of bacteria, 120 and stress-corrosion cracking, 165, 376 Hydrogen: cracking, 65–66, 166–172 and cadmium coatings, 276 and cathodic protection, 252, 259, 267 of cobalt alloys, 421–423 mechanism, 167–170 metal flaws, effect of, 170–172 of stainless steels, 354–361 of steel, 166–172 of steel reinforcements in concrete, 145 disease, in copper, 231–232 in silver, 203 electrode, 24–25, 25F standard, 25 embrittlement, 54, 158, 166–169 of aluminum alloys, 395 and cathodic protection, 259 of magnesium alloys, 402 caused by microbiologically influenced corrosion, 118 caused by pickling inhibitors, 312 of Ta, of ultra-high-strength steels, 357 484 INDEX overpotential, 60–61, 62T, 63–66, 70, 80–81, 90, 116, 123, 125–129, 139, 143, 149– 150, 267, 278–279, 310, 402, 412, 448 standard scale, defi nition of, 25 I Impingement attack: of condenser alloys, 379 of copper, 368–370 of copper–tin alloys, 371 of copper–zinc alloys, 371–372 defi nition, 16–17 photo, 368 Inhibition efficiency, defi nition, 311 Inhibitors, 2, 3, 303–316 for aluminum, 384, 390, 392 classification, 303 for corrosion fatigue, 178–179 and critical pitting potential, 97 defi nition, 303 for diesel-engine cylinder liners, 130 for engine cooling systems, 323 evaluation of, 73 in foods, 278 for industrial waters, 318 boiler waters, 152, 328 cooling waters, 322–323 for iron and steel, 130, 133, 143 for lead, 447 mechanism of, 304–308 nontoxic, 323 oxygen, as inhibitor in acids, 125 in phosphate coatings, 287 pickling, 310–313 applications of, 294, 312–313 concentration, effect of, 312F hydrogen embrittlement, caused by, 312 pigments for paints, 291–292 slushing compounds, 313 sodium: benzoate, 306, 384 borate, 133, 306, 322 polyphosphate, 306–307, 323 silicate, 133 for soil corrosion, 206 for stainless steels, 157, 361 pitting corrosion, 350–351 for stress-corrosion cracking, of brass, 376, 378 of steel, 152–156, 157 of stainless steels, 157 of titanium alloys, 157, 430–431 for tarnishing of copper, 314 toxicity, 303–304, 309 in vapor degreasing aluminum, 293 vapor phase, 201 see also Passivators Insulation, stress-corrosion cracking associated with, 355, 375 Intergranular corrosion: in copper–zinc alloys, 156, 378 defi nition, 17–18 in aluminum alloys, 252, 384, 394, 395 in high purity aluminum, 384 in lead alloys, 448 in nickel-base chromium alloys, 415, 416 in stainless steels, 252, 343–350, 358 in titanium, 430–431 in zirconium, 437 Ion implantation, 270 Ionization constant of water, various temperatures, 28T ipy, conversion factors to mdd, 474–475 defi nition, 16 IR drop, 56–57, 63, 68, 73, 74, 88, 262–263, 456 calculation of, in electrolytes, 58 as cause of polarization, 61–63 Iron, 115–148 acids: citric, rates in, 71T, 139T dissolved oxygen: effect of, 124–125 inhibition by, 125–127 HCl + NaCl, rates in, 139T velocity, effect of, 125–127 anaerobic bacteria, effect of, 118–120 anodic protection, 90, 263–264 atmospheric corrosion, 191–204 cast, coupled to steel, 142 graphitic corrosion, 138, 142, 206 soil corrosion, 138, 206 water pipe, 286 in water, 138 composition, effects of, 138–142 heat treatment, effect of, 142–143 hydrogen cracking, 166–172 delay time and H2 content, 168F metallurgical factors, effect of, 138–143 oxidation, elevated temperatures: in oxygen, 232–234 in steam, 323–326 effect of pH, 327F oxygen, dissolved, effect of, 116–118 pH, effect of, 120–123 salts, dissolved, effect of, 131–138 seawater, rates in, 124T, 195T 399–400 as sacrificial anode. 251. 216–217. 172–173. 414–415 nickel–copper. 317. 120 velocity. 108. 348 L Lead. 445–448 of Pb-acid battery grids. in seawater. 447–448 in differential temperature cells. 234F. 447 Liquid-metal embrittlement. 182. EPA. 104F stress-corrosion cracking. 266 Luggin capillary. 286 coatings of. 103. 92. 151–156 effect of: alkalies. 411–412 nickel–chromium–iron. 205–213 stress. 22–24 Nickel. 10–11. in seawater.INDEX 485 soil corrosion. 125–127 in natural waters. 341– 342. 57F M Magnesium. defi nition of. chemical conversion coatings on. 259. 408–411 Flade potential. derivation of. in 1N H2SO4. 400 for iron. 413T critical pitting potential. 445–449 and air quality. 208. 198–201 corrosion characteristics. 236–237 intergranular. 151–155 silicates. 237. 236–237. 379 nickel–molybdenum. 414–415 in pressurized water reactors. effect of in aqueous media. 153 cold work. 198 cathodic protection of. stray current corrosion. 321. 179. 192. 348. 271. 411 corrosion fatigue. 261. 409T. 378–379. 153 nitrates. 160. 353. 403–404 corrosion characteristics: alloys. 261 current. 153–155 illuminating gas (HCN). 105 passive current densities. 15 fatigue cracks in. 351. 209T see also Steel K Knife-line attack. 96. 123. 408–411 atmospheric corrosion. 245 toxic properties of. 348. 236–237. 209T. 402–403 tolerance limit. 129–130 wrought. 138. 10. 152 temperature. 244T. 103F critical current densities. 149–150 stress-corrosion cracking. 229. 256F stress-corrosion cracking. 410 Nickel alloys. 410 alloyed with Cr. 231 in paints. 172 nickel–chromium. 260. 206. 104F passivity. 127 polarization of. 158. 56. 400–403. 98T irradiation-assisted stress-corrosion cracking. 254–255 efficiency of. 226T. 350. woods. Nitrites. 296 passivation of. 4. 207. 415–417. 259 mdd. 274 in copper alloys. 103–108 pitting. 194T. see Passivators Nuclear reactors. elevated temperatures. conversion factors to ipy. 403T coatings on. 437 O Overpotential: defi nition. 404–404 pure magnesium. on steel. 411–417 compositions. 255 for hot water tanks. 393 in water. 415–416 behavior in acids. 446 Lead and Copper Rule. 165 Local-action cells. 236. 174 hydrogen overpotential of. 153 liquid ammonia. 184. 378–379 corrosion in 3% NaCl. 68 soil corrosion. 191 atmospheric corrosion. 262T. 151–152 carbon monoxide–carbon dioxide. 194T. 421 N Nernst equation. 414 atmospheric corrosion. F fouling of. 16 Molybdenum. 60 . 344. 177T in condenser tubes. 290. 62T. 295T. 124T. 372 corrosion characteristics. 410 oxidation. 68. 184. 412. corroded by. 142. 236–237. effect of: in acids. 401F univalent ion. 65 oxidation. 412. 291–292. 399–406 alloy compositions. 84. 194T cupro-nickel. 474–475 defi nition of. 242. 84. effect of cold work. 369. 137. 215. effect of. 198–199 pH: aluminum. 28 iron. 133. 219–220 Pipelines. 323 critical concentration. 84–87 theories of. 274–275 Pilling and Bedworth rule. 28T zinc. 216–218 internal. 92–94. effect of. effect of concentration. 224–225 P Paints. n-type. 467–470 microbiologically influenced corrosion of. 44T Overvoltage. 62T oxygen. various temperatures and Cl− concentrations. 90. 89–90 in sulfuric acid. 6. effect of. 310T effect on steel. 93–96 of iron: in nitric acid. 322–323 critical concentrations of chlorides and sulfates. 96 catalytic properties. 153. 306–307. 253–255. 303–304 see also Inhibitors Passivity. 206 stray-current corrosion of. 92. 227T Oxidation-resistant alloys. 308–310 chromates. 96–100 defi nitions. 234–235 Oxides. 99. 92–96 adsorption. 100–108 critical compositions. effect of. 321 disposal. 387–388 defi nition. 223–224 Zn alloyed with Al. 88–90. 210 soil corrosion of. 236–237 coupled to Pt in borax. 108–109 cathodic plolarization. 4 cathodic protection of. 28 values for water at various temperatures. Li. properties. 227–228 initial. 234–237 Wagner theory. 60 Oxidation. 108 chlorides. 303–310 applications. 233–234 nickel and nickel alloys. 306. benzoate. 102. 308T nontoxic alternatives. 322 silicate. 83–113 of alloys. action of. 309 for treatment of municipal waters. 215–240 catastropic. 309–310 corrosion of steel. 108–109 oxide-film. 321–322 theory of. 235–236 life test. 232–233 iron–chromium alloys. 205. 96. 378–379 . 89. 241–244. relation to. 309 pitting of. 295–296 see also Coatings. 306. defi nition. 305 defi nition. 304. 383 on iron. 210–211. 309T critical concentrations of chlorides and sulfates. 322–323 toxicity. 94. ASTM. hot ash. defi nition. 211. 63–66 factors affecting. 224–227 p-. 310T polyphosphate. 125. 84 fi lm thickness: on aluminum. 384 borate. 334 Passivators. 61 values of. 384–386 in condenser alloys. 306 sodium. 108 electron configuration. 88 oxygen as. 300 Pitting: in aluminum. organic Parting. 323 combined with chromates. 227–228 reactive element effect. 152. 303 examples of. 140. 120–123 measurement. 306 for treatment of potable waters. 61 values of. 60–61. 303–306. 231 iron and iron alloys. 96. 17. on copper. 100–102 breakdown of. 64–65 values of. 257– 263. 289–291 applying. 322 nitrite. 63–66 mechanism of. 303–310 for cooling waters. 133. 220–223 galvanic effects. 170 inhibitors.486 INDEX hydrogen. 120 hydrogen-induced cracking of. 102 Patina. 304–308 toxicity. 247 stress-corrosion cracking of. thin fi lm. 2. 234–237 Reactive element effect. 229 of copper. 62T metal deposition and dissolution. 102. 230–232 equations expressing rate. 105. 33–34 characteristic values. 369–370 critical potential for. 321–322 Schikorr reaction. normal. 34T sign of. 87 pH. 476T definition. 61–63 in soil. 246 formula for. 367–368 iron. 48–49. 206–207 measurement. 16 elongated. 31T. 24–26 drop along soil surface by current entering pipe. defi nition. 280. 86 as calculated. for iron. 97. 68–71 defi nition. 25 liquid-junction. used for. 184. 334T Reactive element effect. by water treatment. 461–463 effect of temperature. 116 defi nition. 43–44 copper. 58. 99–100 defi nition. 97–99 critical temperature for. 291 . 22. 158–162. 254–256. 43–51 aluminum. value for. 25–28 standard. 469–470 Flade. 192–195. 201 S Saturation index: alteration of. relation to stability of. 54 definition. effect on corrosion of. dependence on. 96–99. 417F derivation of equation for. 44–45 zirconium. 85F Potential: calculation of. value for. 86 iron. 386. 73–75 calculation of. protection by. 334–335 Seawater: aluminum in. 352. 1 iron. treatment of. 467–469 formula. 355–358 decay along pipeline: derivation of formula. 353F factor. 137–138 and municipal waters. 61–63 corrosion rates: anode–cathode areas: effect of. 134–136. for iron. 242 Pourbaix diagrams. 56–58 potentiostatic. 257 defi nition. 258T. 245. 441–442 titanium. 66–67. 15F. in atmospheric corrosion. 93. 45–47 lead. 172–173 Reaction limit. 14. 470–471 Ring-worm corrosion. 408–410 tantalum. 58–63 activation. 25 Potentiometer. 24 standard hydrogen. 263 Potentiostatic polarization. 334 values. 272. 25 standard. 85. chromium. 279 chromium–iron alloys. 60–61 concentration. coatings. 47–48. 71–73 relation to. 383 basis of. 85. 7–8 Rust: composition. 16 in stainless steels. 234–237 Resistivity: IR drop in electrolytes. gold alloys. 54 galvanostatic. 136 limitations of. 393 cathodic protection in. 325 Season cracking. 86–87 hydrogen. 435–436 R Radiation damage. 58–60 IR drop. 66 critical: for pitting.INDEX 487 in copper. 445–446 limitations. 26. 299–300 tetrafluorethylene. 410 passivity. 5. 56 measurement. 22–28 corrosion. 419–420 water. 321 calculation of. 143 Risk. 97–99 for stress-corrosion cracking. value for. 136. derivation of. 86F how measured. coal-tar coatings. 88F definition. 392 galvanic coupling. 49 magnesium. 88–89 nickel. 350–354 theory of. 300 Polarization: causes of. 112 in chromates. 400 nickel. 261. 215. 66 Potentiostat. 351–353 Plastics. 274–275 zirconium in. 286 and stray current corrosion. 361 heat treatment. 357. 351. 11. 127 galvanic series in. 207. corrosion in. effect of. 90–91. 358. 208–210 rates. 195T. effect of. 231. 351–353 soils. 426. 103. 393–394 stainless steels. corrosion in. 334. 104F. 246 stress-corrosion cracking. 98. 269. 342–343 sensitizing. 411 nickel–chromium alloys in. 414 nickel–chromium–molybdenum alloys in. 97–99 critical temperature. 356F . 412. 15. 345–347 ferritic types. boiler corrosion. 361. 369–370. cold worked. 344F ferritic. 350–351 theory of. 353F. 357 avoiding. 400–402. 170. 360 damaging ions. 354 defi nition. 348–349 theory of. 357 galvanic coupling. hydrogen cracking. 156. as cause of. 118 lead in. 123. 343–347. 206–207 pitting. 16. 209T resistivity measurements. 130F effect of nickel. 216. 229 copper in. for optimum corrosion resistance. 357. 172. 372 corrosion of aircraft over. 410. 341. 343–349 avoiding. 124T. 324. 446. 118–119 nickel in. 352. 232. 354–355 in polythionic acids. 178 deaeration. 129. 335 galvanic coupling. 167. 179. 141 paints for. 141. 354–361 austenitic: avoiding. 208 effect of flow velocity. 99–100 inhibiting against. 223–224. 358F nitrogen. 130 corrosion fatigue. 343 austenitic. 184. 90 pitting. 195T. 14 corrosion rate of steel in. 343–345. 344. 131–132. 431 zinc in. 17. 357–359 nickel. corrosion in. 329 cavitation–erosion. 350–351 avoiding. 16 corrosion rate of magnesium in. 350–354 avoiding. 33F IR drop in. 350 passivity. 405 stress-corrosion cracking. 446 magnesium in. 343 compositions. 349 history. 371 copper–nickel alloys in. 32. 361–362 crevice corrosion.488 INDEX copper–aluminum alloys in. 195T. 414 stainless steel in. 32. 271. 354. 357 intergranular corrosion. 347–348 effect of carbon and nitrogen. 205–213 avoiding. 429–430. 197. 263–264 classes and types of. 447 Slushing compounds. 207 factors affecting. 100. 208. 139 effect of chromium. 362 steel in. 195T. 359F oxygen dissolved. 354 critical potential. 348 of nonsensitized alloys. 167. 361 18–8. 208 depth of burial. Pt. 437 Sensitization: aluminum alloys. 412. 400. 384. 402 microbiologically influenced corrosion in. 338–340T corrosion resistance. 408T. 355. 227–228. 337–343 cold work. nickel–chromium–iron alloy in. 343–350 austenitic types. 335–365 anodic protection of. effect of. 19 galvanic coupling in. 411. effect of alloyed Cu. 350. 211 in cinders. effect on. carbon. 349–350 theory of. 378–379 copper–zinc alloys in. 61. 244 titanium in. Pd. 360. 360–361 critical potential in MgCl2. 175. 359–360 metallurgical factors. 313 Soils. 410. 432 titanium alloys in. 264. 138. general. 58 iron in. 317 effect of carbon content. 359–361 Silver. 336–337 hydrogen cracking. 208 stress-corrosion cracking. 416–417 nickel–copper alloys in. 346F knife-line attack. 210–211 Stainless steels. 448 lead alloys for cathodic protection. 350–351. 292 Portland cement coatings on. 246–247 damage by. corrosion in. inside of. 71. 209T first patent. 171–172 Stress-sorption cracking. 141–142 galvanized. ac vs. see Stainless steels weathering. 18 initiation of. 357 and stress-corrosion cracking of steel. 64. 207. 202T Bessemer. 206. 354. 431 of titanium alloys. 275T in soils. 274–276 corrosion of: atmospheric. 274. 124T. 458–461 Stray-current corrosion. 166. 327 corrosion: of aluminum in. 180–184 galvanic coupling. 165–166 defi nition. 162. 402–403 in stainless steels. 374–379 critical crack depth. 241–249 avoiding. 421 effect of surface flaws. 177T oxide coatings on. oxygen-furnace. residual. 198. 18. 161–162 of iron and steel. 155. 156–166 of pipelines. 209T. 195T. 241 Terne plate. 252 painting of. 206. 376 and corrosion fatigue. 72–73 derivation of. 177 and hydrogen-induced cracking. 226. 277–279 . 139 cold rolling. 253 soils. 270 Tin: atmospheric corrosion. 206 atmospheric corrosion. 241–242 to amphoteric metals. critical potentials for. 138–141 corrosion fatigue. 140. 140–141. 143–145 cathodic protection. 149 in brass. effect of. 436. 457. 227. 194T coatings. 441–442 passivity. 166 Stern–Geary equation. 231 Sulfidation. in hot water. 156T in zirconium alloys. 237. 287 pickling of. 171–172 T Tafel equation. 242–244 Stress. 244 detection of. effects of. 210 stress-corrosion cracking in. 441–442 Telluric currents. 162–166 of stainless steels. 228F and glass coatings. 143 low alloy. 124T in acids. 409T Sulfide stress cracking. 151. 193F. formation of. 193F. 170 in magnesium alloys. 144. 151 in seawater. 163F. 138–142. avoiding. 245 pipe. 294 reinforcements in concrete. 354–361 surface flaws. 193F. 386 of iron in. 166–167. effect on corrosion. 327. 153 of copper–zinc alloys. 201 copper-bearing. 437–438 return-line corrosion. 210–211 rate of crack growth. 206 atmospheric corrosion. 124T in hydrochloric acid. 149–150 composition. 328 Steel: atmospheric corrosion. 61. 356–357. 395 Subscale formation. 441–443 galvanic effect. 155 corrosion characteristics. 208. effect of. 201 see also Iron Stepwise cracking. 459 Tantalum. 344. 202T corrosion in aqueous media. 243–244 sources of. in water. 245 quantitative values. 194T. 276 heat treatment. 138 carbon content. 208. 158–161. 286 hydrogen embrittlement. 138–142 corrosion fatigue. 456–458 general corrosion. 96. 394–395 of bridge cable. 151–156 of magnesium alloys. 173–180 fretting corrosion. 437 Stress-corrosion: of aluminum alloys. 123. 361 weld decay.INDEX 489 ferritic and martensitic. 431 Stress intensity factor. dc. 345F Steam: condenser corrosion. 274 Thermal spraying. 194T. 402–404 mechanisms. 323–326 of zirconium in. 441–443 as alloying element. effect on corrosion. 210–211 stainless. effects of. 296 pitting of. in 3% NaCl. 164–166. 230. 292. corrosion by. effect of. 211 Zirconium. 425–434 alloy compositions. effect on corrosion. 296 Water. 275F -rich paints. 435–440 corrosion characteristics. 279 Titanium. 270 oxidation of. 428F effect of alloyed Pt or Pd. treatment of. 437 transition point. critical potentials for. 439F water. 93 Transpassivity. 30–32 corrosion characteristics. 435–439 pitting in chlorides. 430F seawater. 427 Flade potential. 429–430 stress-corrosion cracking of alloys. see also Steel. 195T. 223–224 and Traud experiment. 439F effect of alloyed tin. 108–109 overpotential for deposition or dissolution. 234–235. 345 (Photo) definition. 173 steam.490 INDEX complexes. 208. 209T. defi nition. 436 radiation. behavior in. reaction with. iodides. 288 coatings on steel. 436. 437–438. 428 corrosion characteristics. 199 chromate coatings on. 438 stress-corrosion cracking. 348 weld decay. 220. corrosion rates in inhibited. 437 Pourbaix diagram. 429–430 in bromides. 317–332 boiler. resistance to. 195T. 274–276. 412. 431 cathodic protection against. fuming. 215. 208. 227T painting of. 142 Vanadium. 437–439 . 436 in alkali–hydrogen peroxide solution. 96. 296 seawater corrosion. 207. 275 soil corrosion. galvanized electrogalvanizing. 425–432. 431 Transition metals: catalytic properties. high temperature. 431 inhibiting ions for. 438. 426 hydrochloric acid. 296 pH. 321–322 municipal supplies. 61 passive properties. effect on. 221. 428F nitric acid. 92 Tungsten. 236–237. 427 pitting and crevice corrosion. 84 theory of oxidation. 195T. 421. V Valves. 271 Z Zinc: atmospheric corrosion. 427–429. 321–322 Welds: knife-line attack. in O2 . 229 W Wagner: defi nition of passivity. 194T. 70 Wash primer. corrosion of. 323–328 cooling. 429–430 pitting potentials in chlorides at elevated temperatures. 344 Y Yttrium. 322–323 hot-water heating systems. 198. as cause of catastrophic oxidation.