Shelf Life Prediction Of Medical GlovesPresented for the ASTM WG Committee for “Medical Glove Expiration Dating Guidance” By Uday Karmarkar Akron Rubber Development Laboratory, Inc. [email protected] Motivation • Shelf Life Prediction of rubber Medical Products • Clients’ demand for reliable prediction of part shelf life in the laboratory • Lack of incorporation of aging effects and chemical degradation • Assessment is often qualitative - based on quality tests which are very misleading • Address the nonlinear aging effects (Current FDA/ASTM/ISO Taskgroups) Importance • • • • • • New shelf life prediction criteria New material development Safety, security and liability Material changes due to market trends Can reduce the number of product tests Cost of unexpected failure is high Major Issues for Shelf Life • What is the Mechanism of Aging ? • How is Aging defined ? • Can the Activation Energy be defined so that the temperature dependence of the failure can be found ? • Are the aging mechanisms the same at different temperatures ? • Can an accelarated test be run in the laboratory ? • Are the mechanisms the same in the laboratory as in storage ? Goals • Need a scientific methodology for quantitative life prediction of elastomers • Shelf life prediction • Establishing the effectiveness of the existing models . Challenges • • • • • • Establishing accelerated aging techniques Establishing the degradation process/mechanisms Developing measurement techniques Establishing a basic “practical” model Establishing accuracy and reliability Establishing a degradation mechanism or aging mode • Developing a good methodology which covers all thin elastomeric materials . “Natural” Aging Environmental Factors Alkali/Acid Elastomer Products Electricity Energy Oil/Solvent Radiation Air Ozone Wind Stress (Static & Dynamic) Light Water Corrosive mist Environment Corrosion Dust salt water Mechanical Stress Microbial Heat Molecular Chemical Transitions Additives Other Physical Chemical Mechanical Electrical Structural Properties Life Predictions . Degradation Process/Mechanism Thermal Thermo-Oxidative Photo Photooxidative Ozone Random Chain Scission Depolymerization Crosslinking Mechanical Hydrolytic Chemical High Energy Radiation Side Group Elimination Substitution Plasticizer Loss Filler Bonding Change . swelling and analysis by mass spectrophotometer) • Failure mode evaluation by two different techniques • Incorporating all the aging conditions similar to storage .Accelerated Aging • The basic requirement for reliable accelerated aging is that the course of degradation is identical at the storage and test conditions • Measure minimum two structural parameters at a given temperature. physical properties and chemical analysis (extraction. Tensile Test Tensile Strength .Degradation Monitoring Tests • • • • • • • • • Modulus .Tensile Test FTIR (Field sample & Lab aged) Crosslink Density (Wet Chemistry) GC/MS Specific Gravity .Tensile Test Tear Strength Strain Energy (“Work to Break”) .Tensile Test Ultimate Elongation . Life Prediction Models • Arrhenius Life Prediction • Time Temperature Superposition • ARDL P & K Method (Activation Energy successive averaging technique) . Temperature Plot Decreasing Temperature 100 75 % Property Retention 50 25 0.1 1 10 Log hours 100 1000 .Time . Arrhenius Approach Life Limit Service Rating Temperature Log Time Elevated Temperature Test Points 1 / Temperature (Deg K) Increasing Temperature Typical Arrhenius Shelf Life Prediction Plot for Elastomeric Products . for every 10 Deg C that temperature rises. the first order chemical reaction rate doubles” .E/ RT + ln A “In General.Arrhenius equation • The Arrhenius Equation has the basic form k = A e -E / RT where • A = proportionality constant • e = base for natural logarithms • E = Activation Energy • R = the gas constant • T = Absolute Temperature ln k = . Arrhenius fit : Practical Issues • An Arrhenius fit provides the worst case estimate of shelf life • Arrhenius Approach prediction will be better with lower temperature test data • Lower temperature data need more testing time • The basic assumption is that the dominant mode of failure is identical at all temperatures and stress levels . Time .31432 J/K). R= gas constant ( 8.1/ T(age) } where Ea is the activation energy.Temperature Superposition at = exp { Ea ( 1 / T(ref) .1 1 10 Log hours 100 1000 . Tref and Tage are the reference and aging temperatures respectively in Kelvin 100 70 60 50 80 Decreasing Temperature 75 % Property Retention 50 25 0. Temperature Superposition : Practical Issues • Need to calculate Activation Energy • Assume the activation energy as 84 .Time .117 KJ/mole • High sensitivity on “cutoff” . ARDL “P & K” Method Activation energy (E) is averaged with a successive segment approach Predicted Data Points E8= E7+E6+E5+E4/4 E7=E6+E5+E4+E3/4 E6=E5+E4+E3+E2/4 Ln (1 / seconds ) E4 E3 E2 E5=E4+E3+E2+E1/4 Calculated Test Data Points E1 1 /Temperature (1 / K ) . Successive Activation Energy Extrapolation : Practical Issues • This method is sensitive to changes in activation energy • Predictions are made depending upon the actual data available • This method follows the change in Activation Energy . 97 use Arrhenius extrapolation Case II : R2 < 0.97 use Successive Averaging technique .ARDL Aging Procedure Deg C 30 min 80 x 70 x 60 x 50 x 2 hrs x x x x 4 hrs x x x x 8 hrs x x x x 1 week 2 weeks 4 weeks 6 weeks 8 weeks 10 weeks 24 weeks x x x x x x x x x x x x x x x x • • • • Number of test specimens: 13 ( Tensile Test) Calculate coefficient of variation: CV = s/x (100%) Typical requirement: Maximum CV = 15% Correlation Coefficient (R2) of Arrhenius plot Case I : R2 > 0. Natural Rubber Compounds . 00300012 60 Time Hrs 1/ T in K Deg Celsius Data at 0.0029127 70 Time Hrs 1/ T in K Deg Celsius % Property Retention 95 90 85 80 70 55 25 5 8 17 91 300 810 1000 1500 2700 Data at 0.00275 90 1/ T in K Deg Celsius Data at 0.0028303 80 Time Hrs 8 17 29 37 70 91 100 1/ T in K Deg Celsius % Property Retention 95 90 80 70 60 30 10 8 17 91 120 130 190 370 600 Data at 0.003093 50 1/ T in K Deg Celsius Time Hrs % Property Retention 8 10 13 17 95 70 35 0 % Property Time Hrs % Property Retention 95 90 85 80 70 60 40 15 8 17 91 300 6000 10000 13000 16000 Retention 95 90 85 80 60 45 30 15 .Natural rubber .Gloves Compound 1 Data at 0. NR Glove Compound 1: Raw Data 100 90 80 70 60 % Property Retention 50 40 30 20 10 0 -1 0 90 80 70 60 50 D D D D D e g re e s e g re e s e g re e s e g re e s e g re e s C C C C C e ls i u s e ls i u s e ls i u s e ls i u s e ls i u s 10 100 1000 10000 T im e in h o u r s . Time for aging to 70 % retention of Tensile Strength Temperature Deg C Aging Time (hours) 50 1342 60 810 70 130 80 37 90 10 . 0 0 3 5 20 25 30 80 90 53227 1 0 9 6 1 3 = 1 1 .5 2 y rs s7 1 1 6 8 2 8 = 6 .Shelf Life Prediction Curve : NR Compound 1 T e s t D a ta A v e r a g e S lo p e P r e d ic t e d D a ta A r r h e n iu s F it A v e r a g e S lo p e P r o je c t io n T e c h n iq u e T e m p e r a t u r e in D e g r e e s C e ls iu s 60 50 40 70 A r r h e n iu s F it -2 1 .0 0 2 8 0 .0 -1 7 .5 -1 7 .0 7 y rs 24887 s6 s 7 = (s 3 + s 4 + s 5 + s 6 ) / 4 s 6 = (s 2 + s 3 + s 4 + s 5 ) / 4 s 5 = (s 1 + s 2 + s 3 + s 4 ) / 4 s5 5 s4 4 s 4 = y 5 -y 4 / x 5 -x 4 s 3 = y4 -y 3 / x 4 -x 3 1342 810 5671 ln (1 / seconds ) taken to reach 70 % Property Retention s3 3 s2 2 130 s 2 = y 3 -y 2 / x 3 -x 2 37 s 1 = y 2 -y 1 / x 2 -x 1 s1 1 10 0 .0 0 .0 -1 6 .0 0 3 2 0 .0 -1 0 .0 -1 4 .5 -1 9 .0 0 2 9 0 .0 -2 0 .5 -1 1 .5 -1 2 .0 -1 3 .0 -1 5 .0 -1 9 .5 -2 0 .0 0 2 7 1 / T e m p e r a tu r e in K e lv in Time in hours ( to reach 70 %) .0 0 3 0 0 .5 -1 4 .0 0 3 1 0 .5 -1 3 .5 -1 5 .5 -1 8 .0 0 3 3 0 .5 -1 0 .0 -1 8 .0 -1 2 .5 -1 6 .0 0 3 4 0 .0 -1 1 . 52 6.07 Arrhenius Approach ARDL P & K Method .Predicted Shelf Life: NR Glove Compound 1 Years 11. Shelf Life Prediction 110 100 90 % Strength Retention 80 70 60 50 40 30 20 10 0 0.1 10 Time (hours) 1000 100000 90 Deg C 80 Deg C 70 Deg C 60 Deg C 50 Deg C .NR: Glove Compound 2: Raw Data Gloves . Time for Aging to 75 % Retention of Tensile Strength Temp DegC Aging Time in Hrs 90 30 80 30 70 30 60 30 50 30 . 0025 -2.00 -18.9749 Linear (Test Data) 0.NR Glove Compound 2 : Arrhenius Plot Gloves .0031 0.00 -16.Arrhenius Plot 0.003 0.00 -12.0032 0.0026 0.00 -8.00 -20.00 Ln ( 1/Seconds taken to reach 75 % of the origina strength) -4.00 -6.0029 0.00 Temperature (1 / K) Test Data y = -10105x + 15.0028 0.00 -10.0034 .0027 0.0033 0.00 -14.65 2 R = 0.00 0. Predicted Shelf Life : NR Glove Compound 2 Years 3.31 1.15 Arrhenius Approach ARDL P & K Method . 3 65.7 Data at 0.NR Glove Compound 3 Data at 0.2 82.4 .2 60.00300012 60 Time Hrs 336 504 672 840 1/ T in K Deg Celsius % Property Retention 81.0028303 80 Time Hrs 72 168 336 504 1/ T in K Deg Celsius % Property Retention 75.8 Data at 0.4 43.8 71 57.2 20 1.002912734 70 Time Hrs 168 336 504 672 1/ T in K Deg Celsius % Property Retention 75.4 85. NR: Compound 3: Raw Data 100 80 60 % Property Retention 40 20 6 0 D e g re e s C e lc iu s 7 0 D e g re e s C e lc iu s 8 0 D e g re e s C e lc iu s 0 100 1000 T im e in h o u rs . 035 266.6422 94.53996 .Time for Aging to 75 % Retention of Tensile Strength Temperture 0C Aging Time in Hrs 60 70 80 708. 0 -1 7 .0 0 2 8 s 2 = y 3 -y 2 / x 3 -x 2 s1 s 1 = y 2 -y 1 / x 2 -x 1 1 s2 2 2 6 6 .0 0 3 2 0 .5 7 s6 = ( s4 + s5 + s6 ) / 3 s5 5 6 3 9 9 .5 4 1 / T e m p e r a t u r e in K e lv in Time in hours ( to reach 75 %) .5 -1 9 .0 -1 2 .0 -1 8 .5 -1 3 .5 0 .4 s6 8 0 4 8 4 .Shelf Life Prediction Curve : NR Compound 3 T e s t D a ta A v e r a g e S lo p e P r e d i c t e d D a t a A r r h e n iu s F it A v e r a g e S lo p e P r o je c t io n T e c h n iq u e T e m p e r a t u r e in D e g r e e s C e ls iu s 20 -2 1 .0 0 3 3 0 .0 0 3 5 0 .0 0 2 9 0 .5 -1 5 .0 3 5 s 4 = (s 2 + s 3 + s 4 ) / 3 s4 4 2 0 5 5 .0 s 5 = (s 3 + s 4 + s 5 ) / 3 -1 6 .0 0 3 0 0 .5 25 30 40 50 60 70 A r r h e n iu s F it 80 ln (1 / seconds ) taken to reach 75% Property Retention -2 0 .7 3 4 2 y rs 21871 9 4 .1 1 4 3 y rs 4 1 5 5 0 = 4 .0 s 3 = y 4 -y 3 / x 4 -x 3 -1 4 .0 -2 0 .0 -1 9 .5 -1 6 .5 -1 8 .0 -1 3 .7 1 4 4 8 3 1 = 5 .5 -1 4 .6 4 s3 3 7 0 8 .0 -1 5 .0 0 3 1 0 .5 -1 7 .0 0 3 4 0 . 73 .Predicted Shelf Life : NR Compound 3 Shelf Life Technique Arrhenius Approach ARDL P & K Method Years 5.11 4. Inc .Real time data • “ Shelf life Estimate Technique for rubber Medical Gloves” Wunan Huang. Maxxim Medical. Wunan Huang. Maxxim Medical.Shelf Life Prediction . Inc 110 100 90 % Strength Retention 80 70 60 50 40 30 20 10 0 1 100 Time (hours) 10000 70 Deg C 50 Deg C 37 Deg C . 0032 0.00 Ln ( 1/Seconds taken to reach 75 % of the origina strength) -4.8x + 13.548 2 R = 0.00 -20.00 -12.00 -18.00 -14.00 Temperature (1 / K) Test Dat a .775 0.Arrhenius Plot 0.0033 -10.00 y = -9523.00 0.0029 0.0026 0.003 0.0025 -2.00 -8.0028 0.00 -6.00 -16.0031 0.0027 0. 22 5.53 5.Maxxim Medical Study Shelf Life Technique Arrhenius Approach ARDL P & K Method Real Time Study Years 6.Shelf Life Prediction .6 . *** total slope based on regression analysis of the data from all the temperatures.Shelf Life Prediction .43 78.53 yrs K” method Arrhenius Approach** 101. .6 yrs* study ARDL “P & 110.22 yrs * Note: * Real .7 6.7 101.7 101.98** 5.36 101. ** Average of previous three segments. KJ/mole 800C– 700C 700C–600C 600C– 500C 500C– 400C 400C– 300C Shelf Life.7 101. years Real – time 5.68 114.16** 97.time shelf life.Maxxim Medical Study Shelf Life Prediction Technique Segment Activation Energy.7 101. .Latex Product Stability Medical Gloves Aging temperature ( 0C) 80 70 60 50 Time (w eeks) 1 3 10 24 • This time .temperature matrix is established based on ARDL Client Shelf Life Prediction Studies. Conclusion • Successive Activation Energy extrapolation along with better obtained property retention data seems to be a good methodology/model for shelf life prediction of elastomers • It is important to define aging modes/mechanisms • Correlation between storage and lab data should be established based on a minimum of two testing techniques. preferably one physical test and one chemical test • Each application should be studied separately for established shelf life prediction methodology . Conclusion (cont..) • Quantitative shelf life prediction for materials will be a good guide for future specification development • Sensitivity of the shelf life may be evaluated by modifying the threshold value by ± 10% • Need more test data for accurate shelf life prediction for thermoplastic elastomers(preferably in 5 degree increments) • Shelf life prediction as an engineering tool is in the infancy stage of development • More applications related research is needed to establish the effectiveness of the package on shelf life prediction .