Gaussian 03 Manual

Gaussian 03 Online ManualLast update: 19 September 2003 • • • • • • • • Introduction o About Gaussian 03 o Gaussian 03 Citation o Additional Citation Recommendations Using the G03W Program Running Gaussian 03 o Configuring the Gaussian Environment o Setting Up the Default Route File o Efficient Use of Gaussian o Running Test Jobs o Program Limits Preparing Input Files o About Gaussian Input o Job Types o Model Chemistries o Basis Sets o The Title Section o Molecule Specifications o Multi-Step Jobs Gaussian 03 Keywords Gaussian 03 Utilities Additional Information About Z-Matrices References Gaussian 03 Online Manual Last update: 4 April 2003 Gaussian 03 Capabilities Gaussian has been designed with the needs of the user in mind. All of the standard input is free-format and mnemonic. Reasonable defaults for input data have been provided, and the output is intended to be self-explanatory. Mechanisms are available for the sophisticated user to override defaults or interface their own code to the Gaussian system. The authors hope that their efforts will allow users to concentrate their energies on the application of the methods to chemical problems and to the development of new methods, rather than on the mechanics of performing the calculations. The technical capabilities of the Gaussian 03 system are listed in the subsections below. Fundamental Algorithms • • • • Calculation of one- and two-electron integrals over any general contracted gaussian functions. The basis functions can either be cartesian gaussians or pure angular momentum functions, and a variety of basis sets are stored in the program and can be requested by name. Integrals may be stored in memory, stored externally, or be recomputed as needed [20,21,22,23,24,25,26,27,28]. The cost of computations can be linearized using fast multipole method (FMM) and sparse matrix techniques for certain kinds of calculations [29,30,31,32,33,34]. Transformation of the atomic orbital (AO) integrals to the molecular orbital basis by "in-core" means (storing the AO integrals in memory), "direct" means (no integral storage required), "semi-direct" means (using some disk storage of integrals), or "conventional" means (with all AO integrals on disk). Use of density fitting to speed up the Coulomb part of pure DFT calculations [35,36]. Numerical quadrature to compute DFT XC energies and their derivatives. Energies • • • • Molecular mechanics calculations using the AMBER [37], DREIDING [38] and UFF [39,40] force fields. Semi-empirical calculations using the CNDO [41], INDO [42], MINDO/3 [43,44], MNDO [43,45,46,47,48,49,50,51,52], AM1 [43,48,49,53,54], and PM3 [55,56] model Hamiltonians. Self-consistent field calculations using closed-shell (RHF) [57], unrestricted open-shell (UHF) [58], and restricted open-shell (ROHF) [59] Hartree-Fock wavefunctions. Correlation energy calculations using Møller-Plesset perturbation theory [60] carried to second, third [61], fourth [62,63], or fifth[64] order. MP2 calculations • • • • • • • • use direct [21,65] and semi-direct methods [23] to use efficiently however much (or little) memory and disk are available. Correlation energy calculations using configuration interaction (CI), using either all double excitations (CID) or all single and double excitations (CISD) [66]. Coupled cluster theory with double substitutions (CCD)[67], coupled cluster theory with both single and double substitutions (CCSD) [68,69,70,71], Quadratic Configuration Interaction using single and double substitutions (QCISD) [72], and Brueckner Doubles Theory (BD) [73,74]. A non-iterative triples contribution may also be computed (as well as quadruples for QCISD and BD). Density functional theory [75,76,77,78,79], including general, user-configurable hybrid methods of Hartree-Fock and DFT. See this page for a complete list of available functionals. Automated, high accuracy energy methods: G1 theory [80,81], G2 theory [82], G2(MP2) [83] theory, G3 theory [84], G3(MP2) [85], and other variants [86]; Complete Basis Set (CBS) [87,88,89,90,91] methods: CBS-4 [91,92], CBS-q [91], CBS-Q [91], CBS-Q//B3 [92,93], and CBS-QCI/APNO [90], as well as general CBS extrapolation; the W1 method of Martin (with slight modifications) [94,95,96]. General MCSCF, including complete active space SCF (CASSCF) [97,98,99,100], and allowing for the optional inclusion of MP2 correlation [101]. Algorithmic improvements [102] allow up to 14 active orbitals in Gaussian 03. The RASSCF variation is also supported [103,104]. The Generalized Valence Bond-Perfect Pairing (GVB-PP) SCF method [105]. Testing the SCF wavefunctions for stability under release of constraints, for both Hartree-Fock and DFT methods [106,107]. Excited state energies using the single-excitation Configuration Interaction (CISingles) method [108], the time-dependent method for HF and DFT [109,110,111], the ZINDO semi-empirical method [112,113,114,115,116,117,118,119,120], and the Symmetry Adapted Cluster/Configuration Interaction (SAC-CI) method of Nakatsuji and coworkers [121,122,123,124,125,126,127,128,129,130,131,132,133,134,135]. Gradients and Geometry Optimizations • • • • Analytic computation of the nuclear coordinate gradient of the RHF [136], UHF, ROHF, GVB-PP, CASSCF [137,138], MP2 [22,23,139,140], MP3, MP4(SDQ) [141,142], CID [143], CISD, CCD, CCSD, QCISD, Density Functional, and excited state CIS energies [108]. All of the post-SCF methods can take advantage of the frozen-core approximation. Automated geometry optimization to either minima or saddle points [136,144,145,146,147,148], using internal or cartesian coordinates or a mixture of coordinates. Optimizations are performed by default using redundant internal coordinates [149], regardless of the input coordinate system used. Automated transition state searching using synchronous transit-guided quasiNewton methods [150]. Reaction path following using the intrinsic reaction coordinate (IRC) [151,152]. • • • • • Two- or three-layer ONIOM [153,154,155,156,157,158,159,160,161,162,163] calculations for energies and geometry optimizations. Simultaneous optimization of a transition state and a reaction path [164]. Conical intersection optimization using state-averaged CASSCF [165,166,167]. IRCMax calculation which locates the point of maximum energy for a transition structure along a specified reaction path [168,169,170,171,172,173,174,175,176]. Classical trajectory calculation in which the classical equations of motion are integrated using analytical second derivatives [177,178,179,180] using either: o Born Oppenheimer molecular dynamics (BOMD) [177,178,179,180,181,182] (see [183] for a review) [184,185,186,187,188]. This can be done using any method for which analytic gradients are available, and can optionally make use of Hessian information. o Propagation of the electronic degrees of freedom via the Atom Centered Density Matrix Propagation molecular dynamics model [188,189,190]. This method has similarity and differences to the related Car-Parrinello approach [191]. See the discussion of the ADMP keyword for details. This can be done using the AM1, HF, and DFT methods. Frequencies and Second Derivatives • • • • • • • Analytic computation of force constants (nuclear coordinate second derivatives), polarizabilities, hyperpolarizabilities, and dipole derivatives analytically for the RHF, UHF, DFT, RMP2, UMP2, and CASSCF methods [25,139,192,193,194,195,196,197,198,199], and for excited states using CIS. Numerical differentiation of energies or gradients to produce force constants, polarizabilities, and dipole derivatives for the MP3, MP4(SDQ), CID, CISD, CCD, and QCISD methods [143,200,201,202]. Harmonic vibrational analysis and thermochemistry analysis using arbitrary isotopes, temperature, and pressure. Analysis of normal modes in internal coordinates. Determination of IR and Raman intensities for vibrational transitions [193,194,196,200,203]. Pre-resonance Raman intensities are also available. Harmonic vibration-rotation coupling [204,205,206,207]. Anharmonic vibration and vibration-rotation coupling [204,206,207,208,209,210,211,212,213,214]. Anharmonic vibrations are available for the methods for which analytic second derivatives are available. Molecular Properties • Evaluation of various one-electron properties using the SCF, DFT, MP2, CI, CCD and QCISD methods, including Mulliken population analysis [215], multipole moments, natural population analysis, electrostatic potentials, and electrostatic potential-derived charges using the Merz-Kollman-Singh [216,217], CHelp [218], or CHelpG [219] schemes. • • • • • • • • Static and frequency-dependent polarizabilities and hyperpolarizabilities for Hartree-Fock and DFT methods [220,221,222,223,224,225]. NMR shielding tensors and molecular susceptibilities using the SCF, DFT and MP2 methods [226,227,228,229,230,231,232,233,234,235]. Susceptibilities can now be computed using GIAOs [236,237]. Spin-spin coupling constants can also be computed [238,239,240,241] at the Hartree-Fock and DFT levels. Vibrational circular dichroism (VCD) intensities [242]. Propagator methods for electron affinities and ionization potentials [243,244,245,246,247,248,249]. Approximate spin orbit coupling between two spin states can be computed during CASSCF calculations [250,251,252,253,254]. Electronic circular dichroism [255,256,257,258,259] (see [260] for a review). Optical rotations and optical rotary dispersion via GIAOs [261,262,263,264,265,266,267,268,269,270,271]. Hyperfine spectra: g tensors, nuclear electric quadrupole constants, rotational constants, quartic centrifugal distortion terms, electronic spin rotation terms, nuclear spin rotation terms, dipolar hyperfine terms, and Fermi contact terms [272,273,274,275,276,277,278,279]. Input can be prepared for the widely used program of H. M. Pickett [280]. Solvation Models All of these models employ a self-consistent reaction field (SCRF) methodology for modeling systems in solution. • • Onsager model (dipole and sphere) [281,282,283,284], including analytic first and second derivatives at the HF and DFT levels, and single-point energies at the MP2, MP3, MP4(SDQ), CI, CCD, and QCISD levels. Polarized Continuum (overlapping spheres) model (PCM) of Tomasi and coworkers [285,286,287,288,289,290,291,292,293,294,295,296,297,298,299,300,301,302,30 3] for analytic HF, DFT, MP2, MP3, MP4(SDQ), QCISD, CCD, CCSD, CID, and CISD energies and HF and DFT gradients and frequencies. o Solvent effects can be computed for excited states [298,299,300]. o Many properties can be computed in the presence of a solvent [304,305,306]. o IPCM (static isodensity surface) model [307] for energies at the HF and DFT levels. o SCI-PCM (self-consistent isodensity surface) model [307] for analytic energies and gradients and numerical frequencies at the HF and DFT levels. Technical Support Information Last update: 24 March 2003 The current required citation for Gaussian 03 is the following (presented in two formats for convenient cutting and pasting): Normal Name Order Gaussian 03, Revision A.1, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgomery, Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, C. Gonzalez, and J. A. Pople, Gaussian, Inc., Pittsburgh PA, 2003. Last Name First Gaussian 03, Revision A.1, Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, Jr., J. A.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; and Pople, J. A.; Gaussian, Inc., Pittsburgh PA, 2003 Replace “Revision A.1” with the identifier for the revision of the program that you actually use. A paper describing the scientific capabilities of Gaussian 03 is in preparation. Once it is published, this reference should be cited thereafter. The advances presented for the first time in Gaussian 03 are the work of M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgomery, Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, A. D. Daniels, O. Farkas, A. D. Rabuck, K. Raghavachari and J. V. Ortiz. Gaussian 03 Online Manual Last update: 19 September 2003 In general, we recommend citing the original references describing the theoretical methods used when reporting results obtained from Gaussian calculations, as well as giving the citation for the program itself. These references are given in the discussions of the relevant keywords. The only exceptions occur with long established methods such as Hartree-Fock theory which have advanced to the state of common practice and are essentially self-citing at this point. In some cases, Gaussian output will display the references relevant to the current calculation type. Gaussian also includes the NBO program as link 607. If this program is used, it should be cited separately as: NBO Version 3.1, E. D. Glendening, A. E. Reed, J. E. Carpenter, and F. Weinhold. The original literature references for NBO can also be cited [12,13,14,15,16,17,18,19]. Gaussian 03 Online Manual Last update: 4 April 2003 Using the G03W User Interface • • • • • • • Getting Started Menus and Toolbars Batch Processing of Gaussian Job Files Converting PDB and other Files Customizing the G03W Interface Setting G03W Execution Defaults Utility Programs Included with G03W Last update: 19 September 2003 Gaussian 03 Online Manual This chapter explains the Windows approach to the Gaussian program, and gets you up and running with a simple example. INPUT MADE EASY Every complete set of instructions processed by Gaussian is called a job step. A file containing one or more jobs steps is called a job file. Gaussian job files have the 3 letter extension of GJF in the Windows environment. Job files that are composed of multiple jobs steps can have individual steps that are dependent on, or make reference to, previous job steps within the file. In addition, job files may have multiple job steps that have nothing to do with the other steps contained therein. Beyond multiple job step files, G03W can process batches of job files, through the use of a Batch Control & Batch Control File. While job steps may be stored in files, G03W allows simply entering your job step into an on screen form (called the Job Entry Form). From here you can begin processing the job step, and/or save what you've typed in to a GJF file. PROCESSING OF JOB STEPS AT THE PRESS OF A BUTTON. Once you have a job step in memory, you can begin, pause, resume and/or kill the processing of that step (or group of steps) from buttons on the Toolbar or menu items. You can even use your favorite editor to edit the input and view the output right from inside of G03W. VIEW GAUSSIAN OUTPUT TWO WAYS When processing jobs, G03W displays the current output in an on screen, scrollable area, while writing the output to a user defined file. Even if you minimize G03W down to an icon, the processing of the job steps is viewable, as the title of the icon continues to update the current status. FILE CONVERSIONS INTEGRATED Through the use of the NewZMat utility, you can convert to and from numerous chemistry file formats, and automatically load the results into your favorite editor, or into Gaussian itself for processing. CUSTOMIZE GAUSSIAN TO THE WAY YOU WORK Taking advantage of the full range of possibilities in the environment, G03W lets you setup your preferences about editors, directories, colors, fonts, warnings, questions and messages, and default behavior with normal and batch processing. LIKE DRAG & DROP ? G03W if a fully Drag & Drop-aware program. Select a GJF file in the file manager, drag it over the top of a non-processing Gaussian window or icon, and drop the file. Gaussian will load the file, and if you've customized it to do so, begin processing. Select several GJF files and drop them on Gaussian, and Gaussian builds a Batch Control File with your selections and loads it (and possibly starts processing them). Gaussian 03 Online Manual Last update: 2 October 2003 Menus and Toolbars Main Window • • • • • File Menu Process Menu Utilities Menu View Menu Main Window Toolbar Job Edit Window • • • • • File Menu Edit Menu Set-Start Menu Check Route Menu Job Edit Window Toolbar Additional Jobs Steps Window • • • • Step Menu View Menu Check Route Menu Job Step Window Toolbar Main Window: File Menu The File menu allows you to create and access Gaussian 03W input files and to set program preferences. New: Create new Gaussian 03W input (residing only in memory until it is explicitly saved to disk). Open: Open an existing Gaussian 03W input file. The extension of a Gaussian 03W input file is .GJF. The Open menu item may also be used to load an existing batch control file. The batch facility is described later in this section. Finally, it may be used to open a PDB file for conversion (this process is discussed later). Modify: Edit the current input, via the Existing File Job Edit window. Preferences: Set Gaussian 03W preferences. Preferences are described in a separate section later in this document. Exit: Exit from Gaussian 03W. You will be prompted whether to save any unsaved new or modified input files as well as any unsaved changes to the preferences. Main Window: Process Menu The Process menu allows you to manipulate executing jobs. All of its items have equivalent icons in the Job Processing window (described later in this section). Begin Processing: Begin executing the currently loaded input. Pause: Immediately suspend the currently executing job. Pause ® Next Link: Suspend execution of the currently executing job after it completes the current link. (The Gaussian 03 program is divided into a series of modules known as links. Different links perform different parts of the calculation, and the various links execute sequentially, making up the total job.) Resume: Restart execution of a paused job. Kill Job: Immediately abort the currently executing job. If a batch is running, the next job in the batch (batches are formally defined later in this section) will begin executing (unless the End Batch Run on Error preference is set). End Batch: Stop executing the current batch when the current job finishes. Kill Batch: Immediately abort the currently executing job and terminate batch processing without running any more jobs. Main Window: Utilities Menu The Utilities menu gives you access to the batch and file conversion facilities and other utilities provided with Gaussian 03W. We’ll consider them in detail later in this manual. Edit Batch List: Edit the currently loaded batch control file (extension .BCF), via the Edit Batch List window (described later). If no batch control file is loaded, then a new batch list is created and any currently loaded input is erased from memory. NewZMat: Convert files using the NewZMat utility. After selecting this option, you designate the file to be converted from the Open File dialog box. The NewZMat File Conversion window then appears (described later in this document). CubeGen: Generate a cube file for use in a visualization program. You will be prompted for all necessary information. CubMan: Manipulate or transform one or more existing cube files. You will be prompted for all necessary information. FreqChk: Retrieve frequency and thermochemistry data from a checkpoint file. After selecting this option, you designate the checkpoint file to be used with the Open File dialog box. FormChk: Convert a binary checkpoint file to an formatted (ASCII) version. After selecting this option, you designate the checkpoint file to be used with the Open File dialog box. UnFchk: Convert a formatted checkpoint file back to its G03W binary format. After selecting this option, you designate the checkpoint file to be used with the Open File dialog box. ChkChk: Display information about the contents of a checkpoint file. After selecting this option, you designate the checkpoint file to be used with the Open File dialog box. ChkMove: Convert a binary checkpoint file to a form suitable for moving it to another kind of computer system. After selecting this option, you designate the checkpoint file to be used with the Open File dialog box. C8603: Convert a binary checkpoint file from a previous Gaussian version to the Gaussian 03 format. External PDB Viewer: View the current molecular structure with an external PDB viewing program. The program to use is specified in the preferences (described later in this document). Main Window: View Menu The View menu controls the appearance of the window and enables you to invoke an external text editor. The default settings of the various display options may also be controlled via preferences. The editing options also have icon equivalents (described later in this section). Toolbar: Toggles the display of the toolbar portion of the window. When the toolbar is visible, this item is checked. Main Window: Help Menu The Help menu follows standard Windows conventions.Processing Output: Toggles the display of the Output Display area of the window. including the program version and the serial number of this copy: Start current job. Editor: Invoke the external editor (which editor is used is defined in the preferences). When the status bar is visible. Resume executing paused job. Immediate kill current job and batch. this item is checked. When the Output Display area is visible. Note that an executing job must be paused before invoking an editor on its output file. Terminate the current job. . Edit the current Batch Control File (or create new one). Immediately pause job. Contents: Display the table of contents for the on-line help. End the current batch after the current job completes. Pause after the current link. Status Bar: Toggles the display of the status bar portion of the window. About: Display an informational window about this version and copy of Gaussian 03W. Open external editor. Editor -> Output File: Invoke the external editor on the current output file. which shows a brief description of the current menu item. this item is checked. entering the desired input. Load: Load an existing input file (extension . Save Job: Save the current input to its original file (you will be prompted for a filename if it is newly created input). Save Job As: Save the current input to a file that you specify. If you select the Load option without changing the contents of the filename field. No warning is given about any unsaved changes. Abandon Data: Exit from this window. External Editor: Invoke the external editor on the current input. and then saving it. Exit & Run: Return to the Job Processing window and begin executing the current input (not automatically saved to disk). replacing any current input. Cut. then you will be prompted for the file to load. If it is blank. The loaded file replaces any current input (after prompting for needed saves). and Delete. If the filename field is filled in. discarding all input and changes. Job Edit Window: File Menu The File menu allows you to load and save Gaussian 03 input files.GJF). Paste. this file will be loaded. Some of its options have equivalent icons (described later in this section). Job Edit Window: Check-Route Option . Exit: Return to the Job Processing window. It also has this additional option: Clear Form: Erase all information in all sections of the window. You can create a new input file from this form by selecting Clear Form. Current input is retained but is not automatically saved. The external editor is specified via the preferences. then the current input will revert to the last-saved form on disk (provided that you answer No to the save prompt).Edit G03W output file with external editor. Copy. Job Edit Window: Edit Menu The Edit menu includes the standard Windows Edit menu options: Undo. and rearrange the order of job steps. Discard all input and return to main window. Title Section. The contents of the % Section.This item runs the Check Route utility on the current input (described later in this document). Save all current input to disk. Return to main window and start job. . The default is the main (first) step. and Charge & Multipl. Specify the starting job step. Return to main window. Load an input file (replacing current file). Select the starting step by double clicking on the desired step. There is an equivalent icon for this option (described later). remove. Exit from the window by choosing Close from the window’s System menu (reached via the close bar in its upper left corner). They may be edited as desired as the additional areas are filled in. There is an equivalent icon for this option (described later). Run the Check Route utility. Job Edit Window: Set-Start Option This option enables you to set the starting job step for this input file (additional job steps are discussed later in this section). Additional Jobs Steps Window: Step Menu The Step menu is used to create. Add Step: Create a new job step after the current one. areas from the main job are automatically copied to the new step. Its items also have equivalent icons (described later in this section). Load From File: Replace the current step with the job stored in an external file (you will be prompted for the filename). Prev Step: Move to the previous step in this job. Go to previous job step. Additional Jobs Steps Window: Check-Route Item This item runs the Check Route facility on the current input step (described in a separate section later in this document). Next Step: Move to the next step (higher numbered) in the job. If the file contains more than one job step itself and the current step is the last job step. Exit: Return to the Job Edit window.Delete Step: Remove the current step from the job. There is an equivalent icon for this menu item (described later in this section). If the file contains multiple job steps and the current step is not the last step in the job. Reorder: Change the order of the job steps using the Re-Ordering Data window (described in a separate section later in this document). and an error message will be displayed. . Move to a specific job step. as the current step. Additional Jobs Steps Window: View Menu The Additional Jobs Steps Window menu allows you to move among the additional jobs steps within the current job. Choose Step: Move to the job step number that you specify. Go to next job step. then only the first step from the file will be loaded. then all steps from the file will be loaded in their current order. Run the Check Route utility. Gaussian 03 Online Manual Last update: 19 September 2003 Batch processing in G03W is implemented through the use of the Batch Control system and BCF files. BCF files are also automatically created if a group of files are dropped onto the G03W form or icon from an appropriate file manager. resumed. edit. Batch processing can be paused. or when batch data is entered directly. load and generate BCF files from this same editor. . Multiple GJF files can be processed when in batch mode. The Edit Batch Window Double clicking on a filename in either the input or output list box allows editing of the individual elements in the list. and reorder entries in the batch list. and at exit. You access this feature via the Utilities=>Edit Batch menu item or via the corresponding toolbar icon: . Add Button: Adds an input/output file pair to the list. Delete Button: Removes the currently highlighted input/output file pair. Any and all modifications you have made to the batch control system are saved in memory. you are reminded if you have not saved them to a file. Reorder Button: Allows the user to reorder the data in the list using the Reorder Data dialog (see below). delete. Set-Start Button: Sets the starting file to process in the batch. The built-in batch list editing features allow you to add. Lastly. Return to the Job Edit window. This mode is entered automatically whenever a BCF file is loaded. you can control certain aspects of batch processing via Process Preferences selections. You can also save. specify starting entry. ended and killed through menu and toolbar process controls. The file extension will be adjusted as the user selects conversion parameters under output options. in their old order. Save As: This menu item saves the contents of the list to a new filename. and implement the new orderin Edit Batch Window: File Menu New: This menu item clears the batch list and prepare memory for a new list typed in. standard job processing mode is set. Double-Clicking on an item in the top list box moves it to the bottom list box which holds the new order. The top list box contains those items (Batch Filename data or Additional Job Step names) that can be reordered. Once your choices are highlighted. If not.Reorder Data This form allows for the reordering of list based data. hold down the Shift (select a range) or Control (select specific) key while clicking on your choices. Use the FIND FILE button to quickly select a different conversion source file. Save: This menu item saves changes to the already loaded file. After selecting an appropriate file. To move a group of items from one list box to another. . Gaussian 03 Online Manual Last update: 19 September 2003 Use this command to translate from one chemistry file format to another. If there are any entries in the list. Double-Clicking on an item in the bottom list box (New Order) moves it to the top list box (Old Order) and places it there in its original order. Open: This menu item loads a BCF file. G94W stays in batch processing mode. Exit: This menu item exits the Edit Batch area. the dialog box appears for conversion. Preliminary conversion parameters are preset depending on the file extension of the filename selected. Generate File Filename: The system attempts to build an appropriate filename for the selected source file. The generated file will be created in the same directory as the source file. Only when all the items in the Old Order list box are in the New Order list box can you press OK. pressing the appropriate GROUP button will transfer the items. and load a converted file into memory or an external editor. Editor->Generated File: Tells the system to load the newly generated file into the user defined external editor for modification and display. consult the Gaussian 03 User's Reference. gray background). and the fonts to use for both input and output. and you can make these choices permanent via the Display Preferences section of the Preferences form. you may or may not want to be prompted for the name of the output file.Load Converted File as Job: Tells the system to load the newly generated file into memory for further processing by Gaussian. Output Options: This button allows user control over the NewZMat Output Parameters. Gaussian 03 Online Manual Last update: 2 October 2003 Customizing the G03W Interface G03W allows you to configure to your tastes many aspects of the user interface. Input Options: This button allows user control over the NewZMat Input Parameters. . and how to save complicated jobs (jobs which are a conglomeration of multiple files) from the Edit Preferences section of the Preferences form. This will only happen if the file conversion was successful. the foreground and background colors to use for the output display area. Processing Output Area and Status Bar via the View Menu on the main form. Other Options: This button allows user control over the NewZMat Other Parameters. FILES AND MESSAGES: You can choose how you want to be prompted concerning over-writing existing files. The control for this is found under the Process Preferences section. each time you run. The file is not loaded into Gaussian memory. These menu items will change the size and shape of the main form. how often to look into the run-time output file and display any new contents. including visual aspects and operating procedures. Edit Generated File: Tells the system to load the newly generated file into memory. On the display preferences form you can choose to see an hourglass when the a link has control of the CPU. whether or not to have a Motif-like look to Gaussian (raised or lowered 3D controls. and display it for editing. Ext. VISUAL PREFERENCES: You can choose actively to display or not to display the toolbar. In addition. For more information about NewZMat. Bin Path: This edit area tells G03W where the link executables exist on your system. If this edit is empty. and how to handle messages. output and errors during batch processing. what happens when a file or set of files is dropped on G03W. (which can take up lots of disk space). WARNING: Having incorrect information will cause all jobs to fail at the first link.e. If left blank. the system will assume no scratch directory is present.CONTROL OVER EVENTS: You can define what happens when a file is loaded (i. If left blank.EXE if no other editor has been defined. the editor can be called from the View menu with the output file. This function will fill in the edit area with your selection. All these options are controlled from the Process Preferences section. ASCII Editor Fill in this edit area with the fully qualified path and filename of the text editor you prefer to you use. the default for GJF files is the directory where the input file was found. . the output filename defines where it goes. the system looks in the directory where you last loaded a file from (in the current session). The scratch path entry tells the system where you want temporary files to be created and re-created. This editor will be available from the edit form menus and the View menu. During the initial installation. after a job has successfully run. The optional output path tells the system where the default should be to create output files. do you jump into the internal editor or not). The BIN PATH entry tells Gaussian where to find its links. or from the toolbar button. It is highly recommended that you have a scratch directory. In addition. the ASCII Editor is preset to NOTEPAD. or from the toolbar button. DEFAULT LOGIC: You can also deal with multiple operating paths by setting the default path information on the main Preferences dialog. for BCF files. This information is filled in by the initial installation program and should normally not be altered. Scratch Path: This edit area tells G03W where the scratch files should be created. Find File: Use this button to quickly locate your preferred editor executable. and all temporary files will be created in the same directory as the input file (if there is one) or the current working directory (if there is no input file). by overwriting the same files. as this will reduce the impact of multiple Gaussian job runs. The input path tells the system where it should look first to find files. text. (Default OFF). Edit: The edit button allows control over the file editing elements of the interface. or in the same directory that the input file was found in. (An indicator of both processing and multitasking). . (See Edit Preferences ). (Default: 15secs). (Default ON). After a file is loade. If this edit is empty. Use System Colors: Toggle whether or not to use the colors defined in the current Windows system color scheme. Use this command to adjust the visual elements of the G03W interface to your tastes: Cursor Indication of Processing: This switch toggles whether or not the cursor should be changed to an hourglass while a link has the CPU.Output Path: This edit area tells Gaussian where you would like all output files to be created. list boxes. (Default ON). then the current working directory is used until a file is loaded. Range 23600 seconds. for aspects of screen display (edits. Process: The process button allows control over the Gaussian Job Step processing elements of the interface. Show Output File Area at Startup: Toggle whether or not to view the output of jobs run when the program first opens. then the output file will be created either where you specify it. Display: The display button allows control over the visual elements of the interface. (Default ON). the directory where the loaded file was found. and display it in the output display area. Output File Scan Time: Set the time (in seconds) that the front-end should wait to scan the output file for new information. (Default ON). If this is edit is empty. Show Status Bar at Startup: Toggle whether or not to view the Status Bar at the bottom of the window when the program first opens. Motif Look: Toggle whether to use a gray background and add height or depth to on screen controls. becomes the default. Input Path: This edit area tells Gaussian you have a preferred default input path to search for GJF files. (See Display Preferences ). (See Process Preferences ). Show ToolBar at Startup: Toggle whether or not to view the toolbar when the program first opens. Output Font: This button displays the font selection box for the output display area. whether or not this toggle button is checked. (meaning Name. and over-writes any previous files (dangerous). Save each step to an individual file (filename is created with the step number). The first toggle button controls whether the interface queries the user for a choice when this condition exists.scrollbars. Multi-Step Job File Saves: When the contents of memory comprises multi-step jobs. The second option provides notification only when a file in memory is being saved to a different filename. style and size) and the text color must be anything but white. only fixed width fonts are available in this area. Any normal font can be used. Use this command to adjust the file I/O elements of the G03W interface to your tastes: File OverWrite Warnings: Select whether you want notification that you are about to write over an existing file. Colors may not be set for this text edit area. In addition.. (Default . whether the user loaded steps from multiple files or not.. the steps may be saved in one of three combinations: • • • Save the steps back to their original files (DEFAULT). Save all the steps to a single file.) Note: Motif Look overrides the color control for window backgrounds. you must fully select a font. (Default OFF). Output Background: This button displays the color selection screen to allow the user to set a color for the background of the output display area.Dark Blue R:0 B:64 G:0). Keep in mind that a color should also be selected for the text (see Output Font below) that will allow seeing the text. Input Font: This button displays the font selection box for the input displays (any edit area on the input forms). • • • The first option provides notification anytime this would occur. Since the information in the output assumes a fixed font (terminal like) display. . The last option never bothers the user with notification.etc. you may select a text color if the Use System Colors switch (above) is off. and that new filename already exists. Note: to see an example in the Sample window. (Default ON). (Default OFF). The format of the file is the same on all computer systems. C:\G03W\scratch).Route.. the batch start entry value is set to the file that caused the error. or to assume default behavior. this file is named Default. (Default ON). Show File On Load: Toggles whether or not to display the contents of a file after its loaded. (Default OFF). (Default ON).Rou. Note: If this feature is active and an error occurs while processing a batch. it is sometimes necessary for performance reasons to override some of the defaults built into the program. This can be done by creating a site customization file. Such questions include file overwrite warnings and non-fatal system errors. Minimize Until End / Error: Toggles whether Gaussian should become an ICON while processing batch jobs. (Default OFF). (See Drag & Drop in your Windows manual). (Default ON). Scan Output During Batch: Toggles whether or not to display the output of the currently processing job in the output display area when processing batches of jobs.g. the Gaussian defaults file is Default. and it is located in the Gaussian 03W scratch subdirectory (e. Run Dropped Files: Toggles whether or not to immediately run a file or list of files dropped on Gaussian by a file manager. On Unix systems. End Batch Run on Error: Toggles whether to halt batch processing when an error occurs. Gaussian 03 Online Manual Last update: 6 October 2003 Depending on the characteristics of a particular computer system.Use this command to adjust the job processing elements of the G03W interface to your tastes: Query Output Name: Toggles whether or not to ask the user the name and directory of the output file to create. If an error occurs or the end of the batch is reached. and this feature is active. residing in $g03root/g03. or to skip to the next job in the batch and keep going. Prompt Messages: Toggles whether or not ask questions of the user when processing batches. Under Windows. . then Gaussian will re-display itself in an open state. and have the same form as normal route section commands. For example.Route contains: -#. especially for MP2. Commands listed in Default. For example.Route file. which affects the results of many kinds of calculations. the rule is that only options which do not affect the outcome of a calculation (i. SCF=Conven. All sites will want to specify the amount of scratch disk space available via the MaxDisk keyword in the Default.MaxDisk=800MB This line will have the effect of limiting disk usage in the semi-direct algorithms to the specified amount. Route Defaults These parameters are introduced by -#. For example.Route change only the defaults. Default. Thus. do not change the values of any predicted quantities) are allowed in the file.e.Route Limitations Not all route section keywords are honored in the Default. will be honored. Thus. Keep in mind that the more disk space is available.line in the file. if the Default. Some suitable limit should be defined for your configuration. In general. they are overridden by anything specified in the route section of an input file.. this line will set the default SCF algorithm to the conventional (non-direct) algorithm: -#. then the conventional MP2 algorithm will be used.Route file. the faster the evaluation. then the direct algorithm will be used. which changes only the integral storage algorithm.directive enforces a default dynamic memory limit. the following line sets default memory use to 32 MB: . the following line sets MaxDisk to 800 MB: -#. However.MP2=NoDirect and the route section contains the MP2 keyword.SCF=Conventional There may be more than one -#. Memory Defaults It is often the case that Gaussian jobs which unwisely use excessive memory can cause severe difficulties on the system.The following subsections describe the types of information which can be supplied in the defaults file. if the route section contains the MP2=Direct keyword. while Int(Grid=3). The -M. will be ignored. MaxDisk=10GB User Defaults Files .MaxDisk=400MB • On a powerful workstation with 8 processors and 1 GB of memory. The default site name is GINC. The value may also be followed by KB. Clearly. MB. For example. the program defaults to execution on only a single processor. all 8 processors should be used by default. more memory should be given to each job: -M. or a substantial decrease in performance will result. and parallel processing is supported in your version of Gaussian. For example.-M. Site Name The site name may be specified by the directive. the default algorithms and memory allocation are fine. you may specify the default number of processors to use in the Default.4000000 Note that this limit can be bypassed with the %Mem Link 0 command. The default memory size is 6 MW.as the site name to be used in archive entries generated by Gaussian. the following command sets the default number of processors to 4: -P. -#. GB.EXPCONS Typical Default Settings Here are reasonable default settings for various machine configurations: • For a small workstation with 64 MB memory and 1 GB of disk. The %NProcShared Link 0 command can be used to override the default for a specific job. KW.64MW -P. the number of processors requested should not exceed the number of processors available.Route file. the following line sets the site name to EXPCONS: -S. MW or GW to indicate units other than words. Number of Processors If your computer system has multiple processors. which sets -S. Also. MaxDisk is all that need be specified. being used for large jobs.8 -#.4 Normally. The utilities are discussed in alphabetical order within this chapter. Standalone cube generation utility. and options specified in the route section of the job take precedence over both of them. Prints frequency and thermochemistry data from a checkpoint file. after moving it from a different type of computer system). Displays the route and title sections from a checkpoint file. subtracted.g. Standalone molecular mechanics program. Settings in the local file take precedence over those in the site-wide file. Determines memory requirements for frequency calculations. and so on). Conversion between a variety of molecular geometry specification formats. Route section syntax checker and non-standard route generation. freqchk* freqmem gauopt ghelp mm newzmat* testrt* unfchk* . Most utilities are available for both UNIX and Windows versions of Gaussian. Performs optimizations of variables other than molecular coordinates. On-line help for Gaussian. Manipulates Gaussian-produced cubes of electron density and electrostatic potential (allowing them to be added. The following lists the available utilities and their functions (starred items are included on the Gaussian 03W Utilities menu): c8603 chkchk* cubegen* cubman* formchk* Converts checkpoint files from previous program versions to Gaussian 03 format. Gaussian checks the current working directory for a file of this name when a job is initiated. pressure and scale factor can be specified for the thermochemistry analysis. Converts a binary checkpoint file into an ASCII form suitable for use with visualization programs and for moving checkpoint files between different types of computer systems. be sure to consult the release notes accompanying the program for information pertaining to specific operating systems.Gaussian users may set their own defaults by creating their own Default. temperature. Alternate isotopes. However. Gaussian 03 Online Manual Last update: 4 April 2003 Utility Programs This page discusses various utility programs included with Gaussian 03. Convert a formatted checkpoint file back to its binary form (e..Route file. in either interactive or batch mode. If the environment variable is unset. Specifying the locations of the various scratch files. and our discussion will examine the remaining three items on the list. Gaussian 03 Online Manual Last update: 10 October 2003 Running Gaussian This page describes the operating system commands required to execute Gaussian on Unix-based computer systems. These are the internal input files used by the program.rwf The Two-Electron Integral file: name. these files are given a name generated from the process ID of the Gaussian process. In this page. Its value should be set to the desired amount of memory in words. You may also see files of the form name. Initiating program execution.int The Two-Electron Integral Derivative file: name.d2e By default. and they are stored in the scratch directory. The final section lists the component links of the Gaussian 03 program. designated by the GAUSS_SCRDIR environment variable (UNIX).inp in this directory. Gaussian uses several scratch files in the course of its computation. Running Gaussian involves the following activities: • • • • Creating Gaussian input describing the desired calculation. See the additional instructions accompanying the program for the equivalent information for other operating systems. the location defaults to the current working directory of the Gaussian process. They include: • • • • The Checkpoint file: name.GAUSS_MEMDEF Environment Variable The GAUSS_MEMDEF environment variable may be used to increase the memory available to utilities which do not offer such an option themselves. Specifying resource requirements. . This discussion assumes that the program has already been installed. we will assume that a basic Gaussian input file has been created.chk The Read-Write file: name. . the value may be followed by KB. . you may wish to save the checkpoint file for later use in another Gaussian job. Gaussian will automatically generate unique filenames for any loc which specifies a directory only. Note that 1 MB = 10242 bytes = 1. . to restart a failed job. directory specifications (without filenames) must include a terminal slash. you may want to split the scratch files among several disk locations. providing an explicit name and/or location for it. megawords or gigawords. Here is the syntax for the %RWF command: %RWF=loc1.000 bytes). and/or the Integral Derivative file among two or more disks (or file systems). and so on. Splitting Scratch Files Across Disks An alternate syntax is provided for splitting the Read-Write file.chk. By default. a command like this one will specify an alternate directory location as well as filename: %Chk=/chem/scratch2/water If disk space in the scratch directory is limited. and so it is the one for which an alternate location is most often specified.By default. or by KW. and each size is the maximum size for the file segment at that location. This may be accomplished by naming the checkpoint file. but space is available elsewhere on the system. In this case. MW or GW to indicate units of kilowords.loc2.000. MB or GB. However. which is placed at the beginning of the input file (before the route section-see chapter 3 for details). the sizes are in units of words. The following commands allow you to specify the names and locations of the other scratch files: %RWF=path %Int=path %D2E=path Read-Write file Integral file Integral Derivative file In general. the read-write file is by far the largest. via a %Chk command within the Gaussian input file. where each loc is a directory location or a file pathname. MB or GB (without intervening spaces) to designate KB.size1. the Integral file. However. these files are deleted at the end of a successful run. respectively.size2.048. overriding the usual generated name and causing the file to be saved at job conclusion. for use by a visualization program. respectively. Here is an example: %Chk=water This command..576 bytes (not 1. gives the checkpoint file the name water. the file will reside in the current directory. On UNIX systems. and the third will be given the name my_job. 800 MB.A value of -1 for any size parameter indicates that any and all available space may be used.profile C shell Bourne shell Note that the g03root environment variable must be set up by the user. these commands specify a name for the checkpoint file. Files to be saved go here. unnamed scratch files are deleted at the end of the Gaussian run. Due to limitations in current UNIX implementations. However. The %NoSave command may be used to change this default behavior.60MW./temp/s0/my_job. and an alternate name and directory location for the readwrite file. and unlimited. Gaussian will generate names for the first two segments. When this directive is included in an input file. For example. and a value of 0 says to use the current size of an existing segment./scratch/. as it will attempt to extend a file segment beyond all remaining disk capacity on these systems. named scratch files whose directives appear in the input file before %NoSave will be deleted at the end of a run (as well as all unnamed scratch files). for which it is the default. Saving and Deleting Scratch Files By default. the file will be retained. -1 should be used with caution. the following directive splits the Read-Write file across three disks: %RWF=/dalton/s0/.800MB. These files are: $g03root/g03/bsd/g03. it is customary to include lines like the following within the . For example.login $g03root/g03/bsd/g03.-1 The maximum sizes for the file segments are 480 MB.login or . -1 is useful only for the last file specified.profile file for Gaussian users: . and named files are saved. Initialization Files The Gaussian system includes initialization files to set up the user environment for running the program.login files: . and cause only the checkpoint file to be saved at the conclusion of the Gaussian job: %RWF=/chem/scratch2/water %NoSave %Chk=water Files to be deleted go here. using it will also have the side effect of keeping any additional file segments included in the list from ever being used. respectively. Note that the directory specifications include terminal slashes. Thus. if the % directive naming the file appears after the %NoSave directive. 6 megawords are used. The %Mem command controls the amount of dynamic memory to be used by Gaussian. the g03 command is used to execute Gaussian 03 (see below).profile files: g03root=location export g03root .profile Once things are set up correctly.login . the following command also sets the amount of dynamic memory to 64 MB: %Mem=64MB Even larger allocations may be needed for very large direct SCF calculations-at least 3N2 words. where N is the number of basis functions. Gaussian 03 may be run interactively using one of two command styles: . MW. Once all input and resource specifications are prepared. Frequency and post-SCF calculations involving f functions should be given 6 MWords if possible. MB. This can be changed to n double-precision words by specifying: %Mem=n For example. you are ready to run the program. If Gaussian is being used on a machine with limited physical memory. GB or GW (no intervening spaces) to denote other units. the following command sets memory use to 64 million bytes: %Mem=8000000 The value given to %Mem may also be followed by KB. the default algorithms as well as the default memory allocation should be set appropriately during installation. $g03root/g03/bsd/g03. See this page for more details on using Gaussian efficiently.. KW.e.setenv g03root location source $g03root/g03/bsd/g03. Warning: Requesting more memory than the amount of physical memory actually available on a computer system will lead to very poor performance. a direct SCF with less than 500 basis functions) does not improve performance on most systems. Using more than 6 million words for moderate-sized calculations (i. For example. so that the default of 48 MB is not available. By default. For example. Either form of command can be forced in the background in the same manner as any shell command using &. Finally.g03 job-name g03 <input-file >output-file In the first form.log.0 1.log echo "$file Done with status $status" >> Status end echo "All Done. and it maintains a log of its activities in the file Status: #!/bin/csh echo "Current Job Status:" > Status foreach file ($argv) echo "Starting file $file at `date`" >> Status g03 < $file > $file:r.com and writes its output to jobname. the following script runs all of the Gaussian input files specified as its command line arguments. The latter are lacking full route sections. g03 commands like those above may be included in a shell script. " All lines preceding the string following the << symbols are taken as input to the g03 command. loops may be created to run several Gaussian jobs in succession.log %Chk=water #RHF/6-31G(d) water energy 0 O H H 1 1 1 1.0 END echo "Job done." >> Status The following more complex script creates Gaussian input files on-the-fly from the partial input in the files given as the script's command line arguments.0 2 120. Scripts and Gaussian Scripts designed to run Gaussian 03 may also be created in several ways (we will use the C shell in these examples). Secondly. actual Gaussian input may be included in the script using the << construct: #!/bin/csh g03 <<END >water. When job-name is not specified. the program reads input from job-name. their route sections consist of simply a # sign or a # line . First. and these can be redirected or piped in the usual UNIX fashion. the program reads from standard input and writes to standard output. submits an input file to a batch queue.p) SP Guess=Read Geom=AllCheck END echo "$file Done with status $status" >> Status end # end of foreach echo "All Done. defined in the initialization files. It includes the latter by exploiting the Gaussian 03 @ include file mechanism: #!/bin/csh echo "Current Job Status:" > Status foreach file ($argv) echo "Starting file $file at `date`" >> Status g03 <<END> $file:r. and does not affect the run-time priority. The subg03 command." >> Status Batch Execution with NQS Gaussian may be run using the NQS batch facility on those UNIX systems that support it. The NQS log file is sent to job-name. The script creates a two-step job for each partial input file-a Hartree-Fock optimization followed by an MP2 single point energy calculation-consisting of both the literal commands included in the script and the contents of each file specified at script execution time. -p n can be used to set the priority within the queue to n. Any other parameters are taken to be NQS options.containing special keywords needed for that molecular system. a file like the following should be created (with filename name. It has the following syntax: subg03 queue-name job-name [-scrdir dir1] [-exedir dir2] [-p n] The two required parameters are the queue and job names.com and output goes to job-name. but no method. or calculation type.out -eo # QSUB -lt 2000 -lT 2100 # QSUB -lm 7mw -lM 7mw g03 <name. In particular. respectively. basis set. Input is taken from jobname.log.job): # QSUB -r name -o name.log %Chk=$file:r # HF/6-31G(d) FOpt @$file/N --Link1-%Chk=$file:r %NoSave # MP2/6-31+G(d.com . This is priority for initiation (1 being lowest). To submit an NQS job from an interactive session.batch-log. just as for interactive runs. The optional parameters -scrdir and -exedir are used to override the default scratch and executable directories. and causes errors to be included in the output file. The first line names the running job. names the output file. builds list of links to execute. of energies to compute polarizabilities & hyperpolarizabilities EF optimization using analytic gradients EF numerical optimization (using only energies) Follows reaction path using the intrinsic reaction coordinate (IRC) Numerical self-consistent reaction field (SCRF) Post-SCF SCRF Trajectory calculations Controls ONIOM calculations ADMP calculations Counterpoise calculations . $ qsub name. diff. The memory parameters are used both for initial scheduling of your job for execution and by the program to determine dynamic memory use. and initializes scratch files Reads title and molecule specification FP optimization Berny optimizations to minima and TS.job and the output would be placed in your current working directory. (for example. The following table lists the component programs of Gaussian 03—known as links— along with their primary functions: L0 L1 L101 L102 L103 L105 L106 L107 L108 L109 L110 L111 L113 L114 L115 L116 L117 L118 L120 L121 L122 Initializes program and controls overlaying Processes route section. This job would then be submitted by issuing the command. The time parameters are different to allow addition of job control for cleanup. archiving the checkpoint file in the event that the job exceeds its time limit).where name should be replaced with a name that is appropriate to your calculation. STQN transition state searches MS optimization Numerical differentiation of forces/dipoles to obtain polarizability/ hyperpolarizability Linear-synchronous-transit (LST) transition state search Potential energy surface scan Newton-Raphson optimization Double numerical differentiation of energies to produce frequencies Double num. all direct methods. SCRF) Iteratively solves the SCF equations using direct minimization Performs an ROHF or GVB-PP calculation Quadratically convergent SCF program MC-SCF Population and related analyses (including multipole moments) 1-electron properties (potential. UHF & ROHF. and checks variables Generates basis set information Calculates overlap. kinetic. calculates symmetry.L202 L301 L302 L303 L308 L310 L311 L314 L316 L319 L401 L402 L405 L502 L503 L506 L508 L510 L601 L602 L604 L607 L608 L609 L701 L702 L703 L716 L801 L802 L804 L811 L901 L902 Reorients coordinates. and potential integrals Calculates multipole integrals Computes dipole velocity and Rx∇ integrals Computes spdf 2-electron integrals in a primitive fashion Computes sp 2-electron integrals Computes spdf 2-electron integrals Prints 2-electron integrals Computes 1-electron integrals for approximate spin orbital coupling Forms the initial MO guess Performs semi-empirical and molecular mechanics calculations Initializes an MCSCF calculation Iteratively solves the SCF equations (conven. and field gradient) Evaluates MOs or density over a grid of points Performs NBO analyses Non-iterative DFT energies Atoms in Molecules properties 1-electron integral first or second derivatives 2-electron integral first or second derivatives (sp) 2-electron integral first or second derivatives (spdf) Processes information for optimizations and frequencies Initializes transformation of 2-electron integrals Performs integral transformation (N3 in-core) Integral transformation Transforms integral derivatives & computes their contributions to MP2 2nd derivatives Anti-symmetrizes 2-electron integrals Determines the stability of the Hartree-Fock wavefunction . field. profile include: . computes various properties (including NMR) Iteratively solves the CP-MCSCF equations Computes analytic CI-Singles second derivatives Computes 1-electron integral derivatives Computes dipole derivative integrals 2-electron integral derivative contribution to Fx 2 PDM and post-SCF derivatives MP2 second derivatives Finalizes calculation and output Gaussian 03 Online Manual Last update: 19 September 2003 Gaussian locates executables and creates scratch files in directories specified by several environment variables . the user is responsible for creating two of them: • • g03root : Indicates the directory where the g03 directory resides (i. The environment variables created by g03. SCF stability Computes fifth order quantities (for MP5.login and g03.e. See this page for more details. However. RPA and Zindo excited states. All Gaussian users need to execute the appropriate Gaussian initialization file within their UNIX shell-specific initialization file.. QCISD(TQ) and BD(TQ)) Old MP4 and CCSD Reoptimizes the wavefunction Iteratively solves the CPHF equations.L903 L905 L906 L908 L909 L913 L914 L915 L916 L918 L1002 L1003 L1014 L1101 L1102 L1110 L1111 L1112 L9999 Old in-core MP2 Complex MP2 Semi-direct MP2 OVGF (closed shell) OVGF (open shell) Calculates post-SCF energies and gradient terms CI-Singles. GAUSS_SCRDIR : Indicates the directory which should be used for scratch files. The Gaussian initialization files are responsible for initializing other aliases and environment variables as needed. the directory above it). On Unix systems. G03BASIS : The directory which contains files specifying the standard Gaussian internally stored basis sets. Scratch files are deleted automatically when a job completes successfully or dies cleanly by default. and into which temporary archive files should be placed if the main archive is unavailable. GAUSS_ARCHDIR : Specifies the directory in which the main site-wide archive file is kept. Gaussian generates unique scratch file names based on the process ID when no name has been specified by the user.d/rc3.g.Rou. leftover files may accumulate in the scratch directory. The following subsections describe the types of information which can be supplied in the defaults file. Consequently. Gaussian 03 Online Manual Last update: 6 October 2003 Depending on the characteristics of a particular computer system. However.• • • GAUSS_EXEDIR : Specifies the directories in which the Gaussian images are stored. this file is named Default. clearing the scratch directory should also be done before NQS is started. This can be done by creating a site customization file. This environment variable is provided for convenience and is designed for use with the @ include mechanism. and to have that scratch directory cleared at system boot time by adding an rm command to the appropriate system boot script (e.g. it is sometimes necessary for performance reasons to override some of the defaults built into the program. ensuring that no jobs are using the directory when it is cleared. It defaults to $g03root/g03/arch if unset. This mechanism is designed to allow multiple Gaussian jobs to execute simultaneously using a common scratch directory.. If the NQS batch system is in use. /etc/rc or one of the files under /etc/rc. Under Windows. An easy method for avoiding excessive clutter is to have all users share a common scratch directory. By default it includes the main directory $g03root/g03 and several alternate directories.. as well as some additional basis sets in the form of general basis set input. residing in $g03root/g03. scratch files are not deleted when a job is killed externally or otherwise terminates abnormally. Route Defaults . C:\G03W\scratch). and it is located in the Gaussian 03W scratch subdirectory (e.d). the Gaussian defaults file is Default. The format of the file is the same on all computer systems. Scratch File Considerations On UNIX systems.Route. GB.Route file. while Int(Grid=3). will be honored. The -M.Route Limitations Not all route section keywords are honored in the Default. the following line sets default memory use to 32 MB: -M. SCF=Conven. this line will set the default SCF algorithm to the conventional (non-direct) algorithm: -#.Route file. Thus. if the route section contains the MP2=Direct keyword. they are overridden by anything specified in the route section of an input file.4000000 Note that this limit can be bypassed with the %Mem Link 0 command. especially for MP2. if the Default.e. Default. which changes only the integral storage algorithm. then the direct algorithm will be used. For example. Thus. KW.Route change only the defaults. However. will be ignored. MB. All sites will want to specify the amount of scratch disk space available via the MaxDisk keyword in the Default. For example. Keep in mind that the more disk space is available.These parameters are introduced by -#.and have the same form as normal route section commands. Memory Defaults It is often the case that Gaussian jobs which unwisely use excessive memory can cause severe difficulties on the system.Route contains: -#. For example. which affects the results of many kinds of calculations. the following line sets MaxDisk to 800 MB: -#. do not change the values of any predicted quantities) are allowed in the file. the rule is that only options which do not affect the outcome of a calculation (i.SCF=Conventional There may be more than one -#.line in the file. Commands listed in Default. The value may also be followed by KB. In general. the faster the evaluation.. Some suitable limit should be defined for your configuration. The default memory size is 6 MW. MW or GW to indicate units other than words. . then the conventional MP2 algorithm will be used.MP2=NoDirect and the route section contains the MP2 keyword.directive enforces a default dynamic memory limit.MaxDisk=800MB This line will have the effect of limiting disk usage in the semi-direct algorithms to the specified amount. The default site name is GINC. the number of processors requested should not exceed the number of processors available.8 -#.Route file. all 8 processors should be used by default. you may specify the default number of processors to use in the Default. which sets -S. being used for large jobs. Gaussian checks the current working directory for a file of this name when a job is initiated. and parallel processing is supported in your version of Gaussian. more memory should be given to each job: -M. MaxDisk is all that need be specified. or a substantial decrease in performance will result. the following command sets the default number of processors to 4: -P. Also.Route file. the following line sets the site name to EXPCONS: -S.as the site name to be used in archive entries generated by Gaussian.4 Normally. -#. Settings in the local file take precedence over those Gaussian 03 Online Manual Last update: 8 July 2004 .MaxDisk=10GB User Defaults Files Gaussian users may set their own defaults by creating their own Default. Clearly.Number of Processors If your computer system has multiple processors. The %NProcShared Link 0 command can be used to override the default for a specific job. the program defaults to execution on only a single processor.MaxDisk=400MB • On a powerful workstation with 8 processors and 1 GB of memory. For example. Site Name The site name may be specified by the directive. the default algorithms and memory allocation are fine. For example.64MW -P.EXPCONS Typical Default Settings Here are reasonable default settings for various machine configurations: • For a small workstation with 64 MB memory and 1 GB of disk. for many sites. given in the following table: Highest Angular Momentum Basis Function f functions g functions h functions i functions j functions 4 MW 4 MW 9 MW 23 MW ~60 MW 4 MW 5 MW 16 MW 38 MW 4 MW 9 MW 27 MW 4 MW 5 MW 10 MW 28 MW ~70 MW Job Type SCF Energies SCF Gradients SCF Frequencies MP2 Energies . GB. Note that some defaults have changed with Gaussian 03 to reflect current typical problem sizes. MB. no other special actions are required for overall efficient program use. KW. MW or GW (without intervening spaces) to specify units of kilo-. For users or sites who routinely run very large jobs. an understanding of the possibilities and tradeoffs can help you to achieve optimal performance.MaxDisk=available-disk • where the amount of available memory and disk are specified as indicated. the following defaults placed in the Default. In general. The default algorithms used in Gaussian are generally designed for longer jobs. and M is a minimum value that depends on the job type. however. Before proceeding. the program attempts to select the most efficient algorithm given the memory and disk constraints imposed upon it.Route file is set up. let us emphasize two very important points: • • The default algorithms selected by the program give good performance for all but very large jobs. The default memory size is 6MW. mega. Since Gaussian does offer a wide choice of algorithms.available-memory -#. the default units for each are 8-byte words. Once the Default.or giga.Gaussian has been designed to work efficiently given a variety of computer configurations.bytes or words. and either value may be followed by KB.Route file will produce good general performance: -M. Defaults used in earlier versions of the program were designed for small jobs of under 100 basis functions. Estimating Calculation Memory Requirements The following formula can be used to estimate the memory requirement of various types of Gaussian jobs (in 8-byte words): M + 2NB2 where NB is the number of basis functions used in the calculation. related information may be found in reference [572]. and Recomputation of Integrals One of the most important performance-related choices is the way in which the program processes the numerous electron repulsion integrals.608 bytes). For example.048. Transformation. Use four Linda workers (one per multiprocessor). not just good overall performance. a 300 basis function HF geometry optimization using g functions would require about 5.388. these two parallelization methods can be combined.. The remainder of this chapter is designed for users who wish to understand more about the tradeoffs inherent in the various choices in order to obtain optimal performance for an individual job. Memory required by each multiprocessor. if the value from the table is 10 MW and you want to use four shared memory processors. Memory Requirements for Parallel Calculations When using multiple processors with shared memory. on a 32-bit system. In Gaussian 03. set %Mem to be at least 40 MW. the amount of memory specified in %Mem should be equal to or greater than the value from the preceding table. those performed via Linda). Additional.MP2 Gradients MP2 Frequencies 4 MW 6 MW 6 MW 10 MW 16 MW 28 MW 38 MW For example. They also reflect the use of uncontracted higher angular momentum functions-f and above-which is the default type. Techniques for both very large and small jobs will be covered. Thus. a good estimate of the memory required is the amount of memory from the preceding table for each processor.e.2 MW (~42 MB) of memory. Use two shared memory processors on each multiprocessor Storage. Larger amounts of memory may be required for derivatives of contracted high angular momentum functions. There are five possible approaches to handling two-electron repulsion integrals implemented in Gaussian: . For distributed memory calculations (i.576 words (= 8. The values in the table are for 32-bit computer systems. you would use the following directive in order to run a job on 8 CPUs located on four twoheaded shared memory multiprocessors (assuming that the memory value from the table is 10 MW): %Mem=20MW %NProcLinda=4 %NProcShared=2 computer. they would need to be doubled for 64-bit systems. Note that 1 MW = 1. MO The AO integrals are generated once and stored externally..e. direct methods are the only choice when memory and disk are exhausted and consequently are inevitably used for the largest calculations. the direct method is the default for SCF calculations in Gaussian . In-Core The AO integrals are generated once and stored in canonical order in main memory (i. At least two of these approaches are available for all methods in Gaussian. This is the approach used by conventional SCF calculations. SCF Energies and Gradients The performance issues that arise for SCF calculations include how the integrals are to be handled. It might seem that direct SCF would be preferred only when disk space is insufficient.7 or better. but allows the integrals to be processed using simple matrix operations and no I/O. this is not the case in practice. SCF calculations use the direct algorithm. In addition. but does potentially involve additional computational effort. The various options and tradeoffs for each method are described in the following sections. while conventional SCF scales in practice as N3. other savings are possible that compensate for this additional effort. MO quantities are stored temporarily on disk in whatever size chunks fit in the available disk space. including zeroes). and consequently is very fast. then transformed to the molecular orbital basis. This is the approach used by earlier versions of Gaussian for all correlated energy methods. This does not require O(N4) internal or external storage. a point is reached fairly quickly where recomputing the integrals (really. Direct The AO integrals (and possibly integral derivatives) are recomputed as needed. Because of the use of cutoffs. Semi-Direct The AO integrals (and possibly integral derivatives) are recomputed as needed. In any case. In some cases. Integral Storage By default. Consequently. This requires large amounts of memory. only those . However. and which alternative calculation method to select in the event that the default procedure fails to converge. The transformed (MO) integrals are also stored externally. The default method for a given job is chosen to give good performance on small to medium sized molecules. the cost of direct SCF scales with molecular size as N2. In contrast to earlier versions of the program.5 [572].AO The two-electron integrals over the atomic orbitals (AO integrals) are generated once and stored externally on disk. and it varies from machine to machine. and N4/4 + 500. Where this crossover occurs depends on how fast the integral evaluation in direct SCF is. Direct SCF Procedure In order to speed up direct HF calculations. This step is omitted by default if any transition metal atoms are present. full accuracy of . so performance can be improved by ensuring that enough memory is available to hold all the density and operator matrices at once. there are 2Norb operators. For MCSCF. SCF=InCore can be requested explicitly. the most efficient strategy is to do an in-core SCF as long as it is feasible. Generally. on modern computer systems. the conventional algorithm is virtually never a good choice on such systems. where Nactive is the number of orbitals in the active space. However. The change to direct SCF as the default algorithm in Gaussian 98 was made in consideration of these facts.6 GB for a 200 basis function job. and use the direct algorithm from that point on. GVB and MCSCF calculations can also be done using direct or in-core algorithms [405]. the iterations are done in two phases: • • The density is converged to about 10-5 using integrals accurate to six digits and a modest integration grid in DFT calculations. The number of operators can be quite large for larger MCSCF active spaces.000 words of memory are necessary for closed-shell in-core SCF. Otherwise. This corresponds to about 100 MB for a 100 basis function job. so the crossover in efficiency is at a larger number of basis functions. the integrals will be evaluated more than once per iteration. SCF=Conven keyword is only needed on small memory computer systems like obsolete PCs.integrals that are needed) actually consumes less CPU time than relying on external storage. For GVB. Consequently: • • Cutoffs are less effective than for Hartree-Fock.1 GB for a 300 basis function job (closed-shell). In the event of difficulties. Direct SCF calculations that have enough memory to store the integrals are automatically converted to in-core runs. and 8. about N4/8 + 500. This step is terminated after 21 iterations even if it is not fully converged.000 words for UHF or ROHF in-core SCF. 1. In-core SCF is also available. This approach is substantially faster than using full integral accuracy throughout without slowing convergence in all cases tested so far. in which case the job will be terminated if insufficient memory is available to store the integrals. The density is then converged to 10-8 using integrals accurate to ten digits. The primary difference is that many Fock operators must be formed in each iteration. allowing up to a total of 64 cycles total for the two steps. there are Nactive(Nactive-1)/2 + 1 operators. Memory requirements are similar to the open-shell Hartree-Fock case described above. where Norb is the number of orbitals in GVB pairs. EDIIS [559] uses energies for extrapolation. multipole moments. it is usually the first choice if convergence problems are encountered.the integrals throughout can be requested using SCF=NoVarAcc. combined with CDIIS and dynamic damping of early SCF iterations. handles the latter phases of SCF convergence. including relative energies. These are the available alternatives if the default approach fails to converge (labeled by their corresponding keyword): SCF=Fermi Requests temperature broadening during early iterations [562]. Since it combines linear minimizations with the Newton-Raphson algorithm suggested by Bacskay. Problem Convergence Cases The default SCF algorithm now uses a combination of two Direct Inversion in the Iterative Subspace (DIIS) extrapolation methods EDIIS and CDIIS. SCF=QC is about twice as expensive as conventional SCF. and previously troublesome SCF convergence cases now almost always converge with the default algorithm. electrostatic potentials. . whichever comes first. at the expense of additional CPU time. a modification of the default SCF approach is used: • • The integrals are done to only 10-6 accuracy except for all-electron (non-ECP) calculations involving molecules containing atoms heavier than argon. Conventional SCF single points and all jobs other than single points use tight convergence of 10-8 on the density. which performs extrapolation based on the commutators of the Fock and density matrices. See the discussion of the SCF keyword for more details. This is sufficient accuracy for the usual uses of single-point SCF calculations. The SCF is converged to either 10-4 on both the energy and density. it is guaranteed to reach a stationary point eventually. Gaussian 03 offers Fermi broadening and damping in combination with CDIIS (including automatic level shifting). and it dominates the early iterations of the SCF convergence process. Typically. CDIIS. and electrostatic potential derived charges. or to 10-5 on the energy. See the discussion of the SCF keyword for more details. The tighter convergence can be applied to single-point direct SCF by requesting SCF=Tight. but not for complex or ROHF. It can be used for RHF and UHF. Since SCF=QC is reliable and can be used for direct SCF. population analysis. Single-Point Direct SCF Convergence In order to improve performance for single-point direct and in-core SCF calculations. based on the method of Bacskay [563]. SCF=QC This is quadratically convergent SCF. For the few remaining pathological convergence cases. This new algorithm is very reliable. A stability calculation can be used to verify that a proper SCF solution has been obtained (see the Stable keyword). as for direct SCF. SCF=DM This is the older steepest descent algorithm of Seeger [564]. for which there are fewer alternatives. Note also that you should verify that the final wavefunction corresponds to the desired electronic state. but even for 100 basis functions. AO The CPHF equations are solved using the written-out AO integrals. The petit (symmetry reduced) list can be used. especially as to the symmetries of the occupied orbitals. These approaches all tend to force convergence to the closest stationary point in the orbital space. Guess=Alter can be used to modify the orbitals selected for occupation. which may not be a minimum with respect to orbital rotations. the direct algorithm is preferred above 100 basis functions or so on vector machines and must be used when disk is exhausted on scalar machines. Hence. This is the only method used . SCF=NoDIIS This implies conventional SCF using the old 3 and 4 point extrapolation. direct frequencies are only about 40% slower than conventional (AO). Since cutoffs are not as effective for direct CPHF as for direct SCF. Note that merely increasing the number of SCF cycles for the default algorithm is rarely helpful. SCF=(MaxCyc=N) Increases the total number of SCF iterations to N. This algorithm is the default. the crossover to direct being faster is higher. It is not usually a good choice for RHF and UHF. SCF Frequencies Four alternatives for integral processing are available for Hartree-Fock second derivatives: Direct The coupled perturbed Hartree-Fock (CPHF) equations are solved using integrals that are recomputed every iteration. especially when using Guess=Alter.Guess=Alter Sometimes convergence difficulties are a warning that the initial guess has occupied the wrong orbitals. but not for direct HF or DFT. It can be used for complex HF. This may be the optimal choice for jobs of up to about 100 basis functions. The guess should be examined. MO The CPHF equations are solved using transformed integrals. but is sometimes helpful for ROHF. More than 64 cycles may be needed for convergence. during in-core frequencies. This keeps the disk storage down to the same modest amount as for directO(N3). 3 x Number-of-atoms) and if only some of the matrices can be held in memory.in most other electronic structure programs and is a bit faster than using the AO basis for small cases. The critical element of this decision making is the value of MaxDisk. link 1002 must form updates to all the derivative Fock matrices. to run optimally. This is the fastest available method when it can be used. In-Core The integrals are stored in memory in canonical order. Freq=Raman produces Raman intensities by numerical differentiation for DFT and MP2 frequency calculations. but most of the decision-making is done automatically by the program. then specifying SCF=(InCore. Computing pre-resonance Raman intensities (with CPHF=RdFreq) will approximately double the job's CPU requirements. While frequency calculations can be done using very modest amounts of memory. they can be suppressed by specifying Freq=NoRaman. MP2 Energies Four algorithms are available for MP2. If N4/8 disk is available. HF frequency calculations include prediction of the infrared and Raman vibrational intensities by default. Link 1110 requires 3NAN2/2 words of memory. and in-core is being used only for speed. The IR intensities add negligible overhead to the calculation. performance on very large jobs will be considerably better if enough memory is available to complete the major steps in one pass. Using this option does not change the calculation's disk requirements.e. it will compute the integral derivatives more than once. Link 1002 requires 3NAN2 words. in every iteration of the CPHF solutions. Memory requirements are the same as for in-core SCF. the integrals are computed once by each link that needs them. Similarly.. This algorithm is selected using SCF=InCore. The freqmem utility program returns the optimal memory size for different parameters of frequency calculation (i. but the Raman intensities add 10-20%. This option is not available for DFT methods. If the Raman intensities are not of interest. plus a constant amount for the integral derivatives to run optimally.Pass) will cause the integrals to be stored on disk (on the read-write file) after they are computed for the first time. the amount required to perform the major steps in a single pass). which should be set according to your particular system configuration (see . It is selected by specifying CPHF=MO in the route section. Link 1110 must form a "skeleton derivative Fock matrix" for every degree of freedom (i. but it will increase the CPU time for the job. plus a constant amount. By default. and then read from disk rather than be recomputed by later steps. but it is basically a waste of disk space..e. No external storage is required.directive and MaxDisk keyword in the Default. p. This was the default algorithm in Gaussian 88 and earlier versions. While the new (semi-direct) algorithm can function well for very large N in modest memory. If no value is specified for MaxDisk. The half-transformed integrals (ip|λσ) over one or more occupied orbitals i are sorted on disk. This is best accomplished with -M. direct. Conventional The AO integrals are written out and transformed. and d functions. This is a good method only for machines with large amounts of physical memory. FullDirect The AO integrals are recomputed as needed during evaluation of E(2). then the MO integrals are antisymmetrized to produce E(2). This method can function in as little as O(N2) memory and N3 disk and is usually the optimal choice. when the direct. The MP2=Conven keyword forces this conventional MP2 algorithm. This is very fast if sufficient memory is available. which does in-core SCF and MP2. the SCF phase can be either conventional. which is slower on all machines. This algorithm can be specified with MP2=InCore. Doing so allows the program to decide between the various available algorithms. and in-core MP2 algorithms are used. The default is direct or in-core SCF. This algorithm is selected with Use=L903. Use=L903 The AO integrals are written out. It is specified with MP2=SemiDirect. It indicates the maximum amount of disk space available in words. can be run in very small memory and may be needed on low-end machines. Gaussian will assume that enough disk is available to perform the calculation with no redundant work. specifying the amount of available memory and disk is by far the most important way of optimizing performance for MP2 calculations. Thus. The algorithms available for MP2 energies are: Semi-Direct The AO integrals are generated as needed. In-Core The AO integrals are generated once and stored in canonical order in memory. In addition. selecting the optimal one for your particular system configuration.Route file (although MaxDisk and %Mem may be included in the input file). which may not be the case for larger runs.Route file. This was an option in Gaussian 88 if the energy but not the gradient was desired. then the transformation and formation of E(2) are done in memory. The old code. It is specified with MP2=FullDirect. N4/4 memory is required. semi-direct. . N3/2 memory is necessary. either in the route section or in the Default. The number of integral evaluations depends on the amount of memory available.chapter 3). it does have a fixed minimum memory requirement of about one million words for basis sets containing only s. or in-core. The transformation recomputes the AO integrals as needed and leaves only the minimum number of MO integrals on disk (see below). If f functions are used. The AO integrals may be written out for use in the SCF phase of the calculation or the SCF may be done directly or in-core. and how the remaining terms are computed: • • • z$$f$$>The default in Gaussian is a semi-direct algorithm. The modern methods compute the integral derivatives at least twice. which was the default in Gaussian 90. for small systems (50 basis functions and below) on scalar machines. the conventional algorithm is somewhat faster. The new methods require no more disk space for gradients than for the corresponding energies. MP2 Frequencies Only semi-direct methods are available for analytic MP2 second derivatives. The remaining terms are computed by recomputing AO integrals. QCISD. once in the E2 phase and once after the CPHF step. etc. involves storing the AO integrals on disk.MP2 Gradients The choices for MP2 gradients are much the same as for MP2 energies. how many are stored. reading them back during the transformation. except: • • • The conventional algorithm requires the storage of the two-particle density matrix and therefore uses considerably more disk than if only energies are needed. CCSD. or in-core E2. direct. The integral derivative evaluation during E2 in the new algorithms requires extra main memory if higher than f functions are used.) all require that some transformed (MO) integrals be stored on disk and thus (unlike MP2 energies and gradients) have disk space requirements that rise quartically with the size of the molecule. As a result. Higher Correlated Methods The correlation methods beyond MP2 (MP3. CISD. the default is to do direct or in-core SCF and then dynamically choose between semi-direct. A full transformation is performed if MaxDisk supplies sufficient disk for doing so. The conventional algorithm. MP4. The default of six million words should be increased for larger jobs. and forming all of the MO two-electron integrals except those involving four . As for the MP2 energy. several alternatives as to how the transformed integrals are generated. eight million words should be provided for computer systems using 64-bit integers. however. This will be faster than other approaches unless the computer system's I/O is very slow. These reduce the disk storage required below what a conventional algorithm requires. MP2 frequency jobs also require significant amounts of memory. There are. a partial transformation is done and some terms are computed in the AO basis. it is crucial for a value for MaxDisk to be specified explicitly for these types of jobs. InCore Requests that the AO Raffenetti combinations be held in memory. Integral Storage Excited states using CI with single excitations can be done using five methods (labeled by their corresponding option to the CIS keyword). CCD. the judicious use of the restart facilities can improve the economy of CIS and TD calculations. CISD.Route file. this is done. the job will fail. either within the route section or via a system wide setting in the Default. QCISD(T). The following points summarize the effect of MaxDisk for post-SCF methods: • • • CID. the program assumes that disk is abundant and performs a full transformation by default.virtual orbitals. Note that only the first two options are available for the TD method: Direct Solve for the specified number of states using iterative diagonalization. but obey MaxDisk in avoiding larger storage requirements. . Thus. CCSD. CISD. forming the product vectors from two-electron integrals computed as needed. BD. Excited State Energies and Gradients In addition to integral storage selection. and QCISD energies also have a fixed storage requirement proportional to O2N2. QCISD densities and CCSD gradients have fixed disk requirements of about N4/2 for closed-shell and 3N4/4 for open-shell. In-core is quite efficient. However. If a post-SCF calculation can be done using a full integral transformation while keeping disk usage under MaxDisk. with a large factor. If MaxDisk is left unset. if not. it is appropriate only on very slow machines like legacy PCs. and BD(T) energies have fixed disk requirements proportional to ON3 which cannot be limited by MaxDisk. If MaxDisk is not set and sufficient disk space is not available for a full transformation. CCSD(T). CID. This approach is used automatically if there is sufficient memory available. This procedure can be requested in Gaussian by specifying Tran=Conven in the route section. This algorithm reduces memory and disk requirements to O(N2). but is only practical for small molecular systems or large memory computers as N4/4 words of memory are required. CCD. The four virtual terms were computed by reading the AO integrals. which recovers whatever generalized density was stored for the current method (presumably CIS) and repeats the population analysis. forming the product vectors from written-out AO integrals.. but that dipole moments and other properties at the CIS level are known to be much less accurate if the one-particle density is used (i. since a CPHF calculation must be performed for each state. This is the fastest method and is the default. using Density=(Check. then SCF=Restart should be specified in addition to CIS=Restart or TD=Restart.e. To do this. The more occupied orbitals. but requires O2V2 memory and O3V3 CPU time. This is a slow method and is never the best choice. This is of limited use for smaller calculations. the CIS density can be recovered from the checkpoint file. which may be performed in the MO basis. This algorithm is an efficient choice up to about 150 basis functions. the sooner the direct algorithm should be used. as the final SCF wavefunction is not moved to its permanent location (suitable for Guess=Read) until the entire job step (or optimization step) completes. If a direct CIS job is aborted during the CIS phase. Accordingly. it is practical only for very small molecular systems and for debugging purposes. Separate calculations are required to produce the generalized density for several states. AO Solve for the specified number of states using iterative diagonalization. and the excited states have already been found. ICDiag The entire CIS Hamiltonian matrix is loaded into core and diagonalized. CIS Excited State Densities If only density analysis is desired.447].Current) Guess=Only. Restarting Jobs and Reuse of Wavefunctions CIS and TD jobs can be restarted from a Gaussian checkpoint file. The minimum disk required is about 4O2N2 (6O2N2 for open-shell). Note that the one-particle (unrelaxed) density as well as the generalized (relaxed) density can be examined. except for comparison with the correct density and with other programs that cannot compute the generalized density. depending on the number of occupied orbitals. as new integrals and transformation must be done. forming the product vectors using MO integrals. Since only integrals involving two virtuals are needed (even for gradients) an attempt is made to obey MaxDisk. but is invaluable for direct CIS.MO Solve for the specified number of states using iterative (Davidson) diagonalization. the use of the CIS one-particle density is strongly discouraged. This produces all possible states. if the orbital relaxation terms are neglected) [108. first solve for all the states and the density for the first excited state: . Consequently. g. CASSCF Frequencies . This is useful if specific bond pairs are to be included. since localization separates electron pairs. For singlets. The MO.NStates=N) Density=Current for states M=2 through N. neither are the excited states produced by CIS or TD [573]. There are several possible tactics: • • • Use the standard delocalized initial guess orbitals. Use Guess=Only to inspect the orbitals and determine whether any alterations are required before running the actual calculation. if the active space consists of all p electrons. CASSCF Efficiency The primary challenge in using the CASSCF method is selecting appropriate active space orbitals. It is most useful for complex systems in which it is not clear which electrons are most poorly described by doubly-occupied orbitals. Pitfalls for Open-Shell Excited States Since the UHF reference state is not an eigenfunction of S2. using a route section of the form: CIS=(Read. This is sometimes sufficient. Stability Calculations Tests of Triplet and Singlet instabilities of RHF and UHF and restricted and unrestricted DFT wavefunctions can be requested using the Stable keyword.416]. Then do N-1 additional runs.Root=M. this requires that one has coaxed the UHF run into converging to a broken symmetry wavefunction (normally with Guess=Mix).# CIS=(Root=1. Use the natural orbitals from the total density from a UHF calculation (CASUNO) [415. Use localized initial guess orbitals. and InCore options are available. The default is Direct. Direct stability calculations can be restarted as described above for CIS.NStates=N) Density=Current if N states are of interest. so that the converged active space can be checked to ensure that the desired electrons have been correlated before proceeding. AO. e. In all cases. which request the corresponding algorithm. Direct. There are additional considerations in solving for CASSCF wavefunctions for excited states (see the discussion of the CASSCF keyword for details). a single-point calculation should be performed before any optimization. such as tests/rs6k for the RS/6000 files. Note that some test jobs are intended for fast hardware and are quite expensive on smaller. 194. The file $g03root/g03/tests/tests. Note that you should run the test jobs from a separate directory to prevent them from clobbering the reference output. A command file is provided which runs ranges of test jobs automatically (described below). 155. Tests 1.log The utility gau-machine returns the system name on all UNIX platforms (i. Examples • The script submit.com. along with their corresponding output files.idx lists what each test job does.CASSCF frequencies require large amounts of memory. However. and 302 cover a range of Gaussian capabilities. 28. 296. Output files are in a separate subdirectory under $g03root/g03/tests for each machine.csh can be used to run test jobs. Rename Existing Default. Running Gaussian Test Jobs An extensive set of test jobs for Gaussian are provided. slower computer systems. Note that certain settings in this file can cause some test jobs to fail. If you build the program from source code. 94.. you will need to make sure that they run with the program's built-in default settings. Increasing the amount of available memory will always improve performance for CASSCF frequency jobs (the same is not true of frequency calculations performed with other methods). Therefore. These calculations also require O2N2 disk space. we recommend that you run a few of the test jobs to verify that the program has been built correctly. Test job input files have names of the form testnnn. . you'll need to rename both the site-wide Default. You can extract this information using the following commands: $ cd $g03root/g03/tests/`gau-machine` $ grep "cpu time" *. It accepts two parameters: the numbers of the first and last jobs to run (by default. a keyword corresponding to the type of computer on which you are running).Route file (located in the $g03root/g03 directory) as well as any individual version of the defaults file that you may have prior to running any test job. The input files are found in directory $g03root/g03/tests/com. You do not need to run test jobs for binary distributions. and the reference output files provided with Gaussian indicate how long the jobs can be expected to take.Route File Before Running Test Jobs If you choose to run some or all of the Gaussian test jobs. all of the tests are run).e. it is not usually necessary to run the entire test suite. . In addition. and a maximum of 20000 Z-matrix centers (atoms. Basis Set Limitations Throughout the Gaussian 03 system. cd /chem/newtests $ ln -s $g03root/g03/tests/com . This page outlines the various size limitations that exist within Gaussian 03. using the directory /chem/newtests as the test job executor area and test job 28 as an example: $ mkdir /chem/newtests. After each test job finishes. These limitations occur in the form of fixed dimension statements and algorithm design limitations. The differences that appear should be limited to non-substantive items. verify that it completed successfully. Z-matrix Limitations There are restrictions on the size of a Z-matrix.csh m n & The final command runs test m through n. dimensioning requirements limit the total number of basis functions that can be used in a few of the older of the energy evaluation procedures. For example: $ $g03root/g03/tests/d1 m n The d1 script filters out insignificant differences from the output files for test jobs m through n and pipes the remaining output through more. The main restriction is imposed within the integral evaluation programs and limits the number of primitive gaussian functions and how they are combined into atomic orbital basis functions. and dummy atoms). $ mkdir `gau-machine` $ $g03root/g03/tests/submit. and their overall effect is to limit the size and types of calculation that can be performed. the maximum number of variables that can be specified in an optimization is unlimited for Berny optimizations but must not exceed 50 for Murtaugh-Sargent or Opt=EF optimizations (30 for Fletcher-Powell optimizations). the maximum number of variables and the maximum number of atoms within a calculation. compare its current output with the reference output using the d1 script. ghost atoms. Then.• The following commands illustrate the recommended procedure for running a test job. These are set consistently for a maximum of 20000 real atoms (including ghost but not dummy atoms). basis set limitations manifest themselves in two ways. Secondly. UHF. ROHF. Complex HF calculations are limited to 180 basis functions. DFT. and σ of the two-electron integral (μν| λσ) are packed into a computer word as 8-bit quantities in the UNIX version. the maximum number of primitive d-shells is 20000. disk space limitations force the use of direct methods before the following limits are reached. These limits apply only when two-electron integrals are written out and can be avoided entirely by using SCF=Direct (which is the default in Gaussian 03). The GVB program is limited to 100 paired orbitals. The maximum degree-of-contraction allowed is 100. the reader must have some understanding of the concepts presented in discussion of the Gen keyword (input of nonstandard bases). SCF=DM is limited to 255 basis functions. Normally. or BD calculations using the default algorithms. or CASSCF is also specified). CI. This in effect limits the number of basis functions to 255 under UNIX for conventional calculations in this mode. MP. SCF and Post-SCF Limitations There are only a few other links which have additional dimensioning limits. There is no further restriction for RHF. The remaining restrictions are in some of alternative programs which must be specifically requested. Stable=Complex. In the terminology introduced there. the maximum number of contracted shells is 20000.Integral Program Limitations To understand fully the limitations in the integral programs. the two linearized suffixes (μν) and (λσ) (where (μν=(μ(μ-1)/2)+ν) are packed into a word. although the preferred SCF=QC can be used with direct SCF and imposes no dimensioning limits. λ. and as 16bit quantities in the UniCOS version. This imposes a theoretical limit of 361 basis functions for conventional calculations on the 32-bit computer systems. . CC. Link 903 (in-core MP2) requires O(N3) words of main memory. ν. and are also limited to f functions. the suffixes μ. The other major restriction that appears in the integral programs is in the manner in which integral labels are packed. the limitations are as follows: the maximum total number of primitive shells is 60000. and complex MP2 calculations are effectively limited by a requirement of O(N3) words of main memory. QCISD. When the Raffenetti modes are selected (for SCF=Conventional except when Tran=Conventional. When the conventional integral storage procedure is selected (in contrast to the Raffenetti ("PK") storage modes [574]). which is not a restriction in practice. These limits do not apply to direct calculations. NBO Dimensions NBO is dimensioned for 200 atoms and 10000 basis functions. the maximum number of primitive f-shells and higher is 20000. 137 H 0.464 1.0 In this job. which requests a single point energy calculation on water: # HF/6-31G(d) water energy 0 1 O -0. model chemistry and other options (blank line terminated). The input section following the molecule specification is used by the Opt=ModRedundant keyword. Optional additional sections: Additional input needed for specific job types (usually blank line terminated). Link 0 section Route section Title section Molecule Specification section Add a bond and an angle to the internal coordinates used during the geom. third. The basic structure of a Gaussian input file includes several different sections: • • • • • Link 0 Commands: Locate and name scratch files (not blank line terminated).0 0. and fourth sections. Title section: Brief description of the calculation (blank line terminated). the route and title sections each consist of a single line. Here is an example of such a file. The molecule specification section begins with a line giving the charge and spin multiplicity for the molecule: 0 charge (neutral molecule) and spin multiplicity 1 (singlet) in this case. Many Gaussian 03 jobs will include only the second.Gaussian 03 input consists of a series of lines in an ASCII text file.0 0. This job requests a geometry optimization.464 0. The following input file illustrates the use of Link 0 commands and an additional input section: %Chk=heavy #HF/6-31G(d) Opt=ModRedundant Opt job 0 1 atomic coordinates … 3 8 2 1 3 opt.441 -0. The charge and spin multiplicity line is followed by lines describing the location of each atom in the molecule. and it serves to add an .177 H -0. this example uses Cartesian coordinates to do so. Molecule specification: Specify molecular system to be studied (blank line terminated). Route section (# lines): Specify desired calculation type.143 Route section Title section Molecule specification 0. Molecule specifications are discussed in more detail later in this chapter. . Gaussian 03 Input Section Ordering Section Keywords Final blank line? . commas. . in this case... Link 0 commands were introduced in the last chapter and are discussed individually in the penultimate section of this chapter. Options to keywords may be specified in any of the following forms: keyword = option keyword(option) keyword=(option1. In general. All keywords and options may be shortened to their shortest unique abbreviation within the entire Gaussian 03 system. Spaces. along with the keywords associated with each one. the Conventional option to the SCF keyword may be abbreviated to Conven. the table below lists all possible sections that might appear within a Gaussian 03 input file. Appending /N to such commands will prevent the included file's contents from being echoed at the start of the output file. CBSExtrap(NMin=6). The contents of an external file may be included within a Gaussian 03 input file using the following syntax: @filename.. option2.additional bond and angle in the internal coordinates used in the geometry optimization. The job also specifies a name for the checkpoint file. which may appear anywhere on a line. or spaces may optionally be included before and/or after it. This holds true whether or not both Conventional and Convergence happen to be valid options for any given keyword. option2. Note that some options also take values. The remaining input sections are discussed in the subsequent subsections of this introductory section. Separate comment lines may appear anywhere within the input file. This causes the entire file to be placed at the current location in the input stream. the option name is followed by an equals sign: for example.) • • • • • Multiple options are enclosed in parentheses and separated by any valid delimiter (commas are conventional and are shown above). For convenience. Gaussian input is subject to the following syntax rules: • • • Input is free-format and case-insensitive. tabs. but not to Conv (due to the presence of the Convergence option). Multiple spaces are treated as a single delimiter. The equals sign before the opening parenthesis may be omitted. or forward slashes can be used in any combination to separate items within a line. Thus. Comments begin with an exclamation point (!)..) keyword(option1. QST3) IRC=ReadIsotopes CPHF=RdFreq Opt=FCCards Opt=ReadError ADMP and BOMD no yes yes yes yes yes yes yes yes yes yes for both yes yes yes yes yes no yes yes yes yes yes yes yes no yes yes Gen. ExtraBasis Massage ExtraBasis. Pseudo=Cards. QST3) Geom=Connect or ModConnect Opt=(ModRedun. GenECP.Link 0 commands Route Section (# lines) Extra Overlays Title section Molecule specification Modifications to coordinates Connectivity specifications 2nd title and molecule specification Modifications to 2nd set of coordinates Connectivity specifications for 2nd set of coordinates 3rd title and initial TS structure Modifications to 3rd set of coordinates Connectivity specifications for 3rd set of coordinates Atomic masses Frequency of interest Initial force constants (Cartesian) Accuracy of energy & forces BOMD/ADMP input (1 or more sections) Basis set specification Basis set alterations % commands all ExtraOverlays all all Opt=ModRedundant Geom=Connect or ModConnect Opt=QST2 or QST3 Opt=ModRedun and QST2 or QST3 Geom=Connect or ModConnect and Opt=ModRedun and QST2 or QST3 Opt=QST3 Opt=(ModRedun. ECP specification GenECP Density fitting basis set specification Extra Density Basis Background charge distribution Charge Finite field coefficients Field=Read Symmetry types to combine Guess=LowSymm Orbital specifications (separate α & Guess=Cards β) Orbital alterations (separate α & β) Guess=Alter . Stable Test wavefunction stability. There are three key components to this specification: • • • The job type The method The basis set The following table lists the job types available in Gaussian 03: • • • • • • • • • • • • • • SP Single point energy. atomic masses PROAIMS/Pickett output filename Guess=Permute SCRF=Read SCRF=COSMORS CASSCF=StateAverage CASSCF=Spin GVB Pop=ReadRadii or ReadAtRadii Prop=Read or Opt Cube Pop=NBORead ReadWindow options OVGF=ReadOrbitals Freq=ReadIsotopes Output=WFN or Pickett no yes no no no no yes yes yes no yes yes no no The route section of a Gaussian 03 input file specifies the type of calculation to be performed. ReArchive Extract archive entry from checkpoint file only. recompute population analysis. Scan Potential energy surface scan.Orbital reordering (separate α & β) PCM solvation model input Filename for COSMO/RS Weights for CAS state averaging States of interest for spin orbit coupling # Orbitals/GVB pair Alternate atomic radii Data for electrostatic properties Cube filename (& Cards input) NBO input Orbital freezing information OVGF orbitals to refine Temperature. Density=Checkpoint Recompute population analysis only. IRCMax Find the maximum energy along a specific reaction path. IRC Reaction path following. . Guess=Only Print initial guess only. Volume Compute molecular volume. Opt Geometry optimization. pressure. Freq Frequency and thermochemical analysis. ADMP and BOMD Direct dynamics trajectory calculation. Force Compute forces on the nuclei. Polar Polarizabilities and hyperpolarizabilities. the default calculation type is usually a single point energy calculation (SP). However. Prop Electrostatic-potential derived charges: Pop=Chelp. Anharmonic]) Hyperpolarizabilities: Freq. the Opt keyword is optional and is the default. In the latter case. For example. Polar Ionization potentials via propagator methods: OVGF IR and Raman spectra: Freq Pre-resonance Raman spectra: Freq CPHF=RdFreq Molecular orbitals: Pop=Regular Multipole moments: Pop NMR shielding and chemical shifts: NMR NMR spin-spin coupling constants: NMR=SpinSpin Optical rotations: Polar=OptRot CPHF=RdFreq .In general. a route section of the form: method2/basis2 // method1/basis1 may be used to request an optimization calculation (at method1/basis1) followed by a single point energy calculation (at method2/basis2) at the optimized geometry. Predicting Molecular Properties The following table provides a mapping between commonly-desired predicted quantities and the Gaussian 03 keywords that will produce them: • • • • • • • • • • • • • • • • • • • • Atomic charges: Pop Dipole moment: Pop Electron affinities via propagator methods: OVGF Electron density: cubegen Electronic circular dichroism: TD Electrostatic potential: cubegen. Note that Opt Freq calculations may not use this syntax. the following route section requests a HF/6-31G(d) geometry optimization followed by a single point energy calculation using the QCISD/6-31G(d) model chemistry: # QCISD/6-31G(d)//HF/6-31G(d) Test In this case. Opt may be combined with IRCMax in order to specify options for the optimization portion of the calculation. only one job type keyword should be specified. When no job type keyword is specified within the route section. ChelpG or MK Frequency-dependent polarizabilities/hyperpolarizabilities: Polar CPHF=RdFreq High accuracy energies: CBS-QB3. W1U Hyperfine coupling constants (anisotropic): Prop Hyperfine spectra tensors (incl. the geometry optimization is automatically followed by a frequency calculation at the optimized structure. G2. VibRot[. G3. The exceptions to this rule are: • • Polar and Opt may be combined with Freq (although SCRF may not be combined with Opt Freq). g tensors): Freq=(VCD. TD Vibration-rotation coupling: Freq=VibRot Vibrational circular dichroism: Freq=VCD The combination of method and basis set specifies a model chemistry to Gaussian. Zindo. . This is usually accomplished via two separate keywords within the route section of the input file. along with the job types for which each one may be used. specifying the level of theory. frequencies. numerical calculations are often available for unchecked methods (see the discussion of the specific keyword in question for details).• • • • • Polarizabilities: Freq. Note that the table lists only analytic optimizations. and polarizability calculations. Polar Thermochemical analysis: Freq UV/Visible spectra: CIS. Every Gaussian job must specify both a method and basis set. The following table lists methods which are available in Gaussian. although a few method keywords imply a choice of basis set. 313. Most method keywords may be prefaced by R for closed-shell restricted wavefunctions. These are accessed via the 6-31G(d') and 6-31G(d'. if no basis set keyword is included in the route section.. and so on. The exceptions consist of a few methods for which the basis set is defined as an integral part of the method. and MP2 energies.312] 4-31G [317. All molecular mechanics methods.322.324.318. MINDO3 and MNDO and PM3 semi-empirical energies and gradients.320.314.329] (note that the basis sets for P. S.327]. CBS and W1 methods. they are listed below: • • • All semi-empirical methods. Compound model chemistries: all Gn. ONIOM and IRCMax jobs require multiple method specifications.318.323. 6-311G: Specifies the 6-311G basis for first-row atoms and the McLean-Chandler (12s.320] 6-31G [317. ROHF.321.52111) basis sets for second-row atoms [328.9p) (621111.315.326] 6-31G†: Gaussian 03 also includes the 6-31G† and 6-31G†† basis sets of George Petersson and coworkers. and including more than one of them will produce bizarre results.316] 6-21G [311. note that analytic ROMP2 gradients are not yet available. defined as part of the Complete Basis Set methods [88. RO is available only for Hartree-Fock.312. The following basis sets are stored internally in the Gaussian 03 program (see references cited for full descriptions). listed below by their corresponding Gaussian 03 keyword (with two exceptions): • • • • • • • STO-3G [309.310] 3-21G [311. Most methods require a basis set be specified. f functions may also be added: e.325.g. HF is assumed. or RQCISD.If no method keyword is specified. only a single method keyword should be specified. including ZINDO for excited states.p') keywords. U for unrestricted open-shell wavefunctions. UMP2.319.319. 6-31H(d'f). AM1. In general. However. to which single or double diffuse functions may also be added. and Cl are those called "negative ion" basis sets by McLean . The form model2 // model1 described previously may be used to generate an automatic optimization followed by a single point calculation at the optimized geometry. there are exceptions: • • • CASSCF may be specified along with MP2 to request a CASSCF calculation including electron correlation. However. all Density Functional methods. then the STO-3G basis will be used. they are given as options to the corresponding keyword. or RO for restricted open-shell wavefunctions: for example. 1h 7s.344.2h.356. D95V: Dunning/Huzinaga valence double-zeta [335]. the basis set of Blaudeau and coworkers for Ca and K [322].. cc-pV6Z: Dunning's correlation consistent basis sets [368.2p.5p.1d 4s. these were deemed to give better results for neutral molecules as well).357.• • • • • • • • • • • • • and Chandler.2p. The SDD. and all three keywords are equivalent for these atoms.350.1f 5s.363.369.342].1h cc-pV6Z 6s.5p.364. triple.367]. Los Alamos ECP plus MBS on NaBi [340.365.3d.3p.1d 4s.347. cc-pV5Z.1f 4s. MDF.1f cc-pVQZ 4s. quintuplezeta and sextuple-zeta.2g.4d. Curtiss and coworkers for the other elements in the third row [324.353.1g 5s. cc-pVTZ.6p.2p.339]. CEP-4G: Stevens/Basch/Krauss ECP minimal basis [337.g.1f 5s. These basis sets have had redundant functions removed and have been rotated [373] in order to increase computational efficiency.2d. SHC: D95V on first row. The following table lists the valence polarization functions present for the various atoms included in these basis sets: cc-pVDZ 2s.352.339].1p 2s. quadruple.3p. Los Alamos ECP plus DZ on Na-Bi [340.360.1h not available 7s. MDF28 for the MDF potential replacing 28 core electrons). cc-pVQZ.4p.2f.4d.1d 3s. MHF. LanL2DZ: D95V on first row [335].3f. and the 6-311G basis set of McGrath.351.334].1g cc-pV5Z 5s. SHF.1p 3s.4f.333.370.3p.4p. using the scaling factors of Raghavachari and Trucks [332]. respectively).331] all electron basis set for the first transition row.4d. cc-pVDZ.371.4p.3f.359. These basis sets include polarization functions by definition.366.338.2g.310] on first row.1i not available Atoms H He B-Ne Al-Ar .349. Note that the number of core electrons must be specified following the form (e. Goddard/Smedley ECP on second row [335.372] (double.339].341.346.358.2f.3d.1d cc-pVTZ 3s.2d.338.355. Also known as SEC. SDF.1g 6s. Note that there is only one CEP basis set defined beyond the second row.354.341.2f.336].5p.3p.3d.5d. LanL2MB: STO-3G [309.3d.2d. the WachtersHay [330.362.348.1g 6s. MC-311G is a synonym for 6-311G.2f. Note that Raghavachari and Trucks recommend both scaling and including diffuse functions when using the Wachters-Hay basis set for first transition row elements. CEP-31G: Stevens/Basch/Krauss ECP split valance [337. CEP-121G: Stevens/Basch/Krauss ECP triple-split basis [337. the 6-311+G form must be specified to include the diffuse functions.2d.6p.3g.3f.4p.36 1. SDDAll: Selects Stuttgart potentials for Z > 2. D95: Dunning/Huzinaga full double zeta [335]. MWB forms may be used to specify these basis sets/potentials within Gen basis input.345.342]. SDD: D95V up to Ar [335] and Stuttgart/Dresden ECPs on the remainder of the periodic table [343.338.2g. Note that (d. UGBS1P adds a p function for each s.p).1f not available not available not available • These basis sets may be augmented with diffuse functions by adding the AUGprefix to the basis set keyword (rather than using the + and ++ notation-see below).388]. When the AUG. Adding a single polarization function to 6-311G (i.384. a d function for each p and so on. • UGBS. 6-311G(d)) will result in one d function for first and second row atoms and one f function for first transition row atoms. one p. and one f diffuse functions on B through Ne and Al through Ar. one diffuse function of each function type in use for a given atom is added [368. EPR-III is a triple-zeta basis set including diffuse functions. double dpolarizations and a single set of f-polarization functions. and one d. for example-and that the 3-21G* basis set has polarization functions on second row atoms only.381. DGDZVP2 and DGTZVP basis sets used in DGauss [387. one d.1)/[6.2p) designates the 6-31G basis set supplemented by diffuse functions. Li. etc. as are multiple polarization functions [390].2)/[4.5.4p. TZV and TZVP of Ahlrichs and coworkers [374. • EPR-II and EPR-III: The basis sets of Barone [377] which are optimized for the computation of hyperfine coupling constants by DFT methods (particularly B3LYP). • MIDI! of Truhlar and coworkers [376].5p. • The DGDZVP. .386]. • MTSmall of Martin and de Oliveira.383.382.p) and ** are synonymous-6-31G** is equivalent to 6-31G(d.380. UGBS1P. UGBS2P and UGBS3P: The universal Gaussian basis set of de Castro. Also in this case the spart is improved to better describe the nuclear region: (6.2.1] for B to F.1)/[7. and UGBS3P adds a p. and Na do not have diffuse functions defined within these basis sets. The latter three keyword forms have an additional 1.1d 6s. Adding Polarization and Diffuse Functions Single first polarization functions can also be requested using the usual * or ** notation.1] for B to F. and supplemented by 2 sets of p functions on hydrogens. 3 sets of d functions and one set of f functions on heavy atoms. the elements He. For example. 2 or three polarization functions for each function in the normal UGBS basis set (i.3d.1] for H and (10.).prefix is used to add diffuse functions to the cc-pV*Z basis sets.2. defined as part of their W1 method (see the W1U keyword) [94]. one d. The MidiX keyword is used to request this basis set.369]. The keyword syntax is best illustrated by example: 6-31+G(3df. and one p diffuse functions on hydrogen atoms.385.2] for H and (11..Ga-Kr 5s.e. a d and f function for each p. UGBS2P adds a p and d function for each s. d and f for each s.7.379.2. • SV. Be.4. the AUG-cc-pVTZ basis places one s.375].1)/[4. Jorge and coworkers [378. However.e. SVP. Mg. The + and ++ diffuse functions [389] are available with some basis sets. EPR-II is a double zeta basis set with a single set of polarization functions and an enhanced s part: (6. Gacc-pV(DTQ5)Z included in definition Kr Basis Set Applies to cc-pV6Z H. Thus while a D95** calculation on water has 26 basis functions. B-Ne. and a 6-31G** calculation on the same system has 25 functions.p) 6-31G H-Kr (3df. La-Bi LanL2DZ H. La-Bi SDD. C. adding a diffuse function to the 6-311G basis set will produce one s. I. one p. S-Cl.since d functions are already present for the valence electrons in the latter.3pd) 6-311G H-Kr (3df. B-Ne included in definition included in definition included in definition included in definition included in definition Diffuse Functions + ++ ++ ++ ++ added via AUGprefix added via AUGprefix SV H-Kr SVP H-Kr TZV and TZVP H-Kr MidiX H. Br EPR-II. Li-Ba. EPR-III H. O. When a frozen-core calculation is done using the D95 basis. F . there will be 24 orbitals used in a frozen-core post-SCF calculation involving either basis set. N. Al-Ar. C-F. B.3pd) D95 H-Cl except Na and Mg (3df. Similarly.p) SHC H-Cl * CEP-4G H-Rn * (Li-Ar only) CEP-31G H-Rn * (Li-Ar only) CEP-121G H-Rn * (Li-Ar only) LanL2MB H-Ba.3pd) D95V H-Ne (d) or (d. SDDAll all but Fr and Ra H-He. The following table lists polarization and diffuse function availability and the range of applicability for each built-in basis set in Gaussian 03: Polarization Functions STO-3G H-Xe * 3-21G H-Xe * or ** 6-21G H-Cl (d) 4-31G H-Ne (d) or (d. and one d diffuse functions for third-row atoms. both the occupied core orbitals and the corresponding virtual orbitals are frozen. The ChkBasis keyword indicates that the basis set is to read from the checkpoint file (defined via the %Chk command). 6-31G††. if you want to add basis functions for Xe from the 3-21G basis set to the 6-311 basis set via the ExtraBasis keyword. all d functions must be 5D or 6D. 6-31G. for example. Cartesian d functions will be used.2. 6-21G. the Xe basis functions will be pure functions. C-F. . Gen and GenECP keywords. Other basis sets may also be input to the program using the ExtraBasis and Gen keywords. Note that basis functions are generally converted to the other type automatically when necessary. the basis set explicitly specified in the route section always determines the default form of the basis functions (for Gen. Cartesian basis functions: • • • All of the built-in basis sets use pure f functions. These keywords also apply to all higher functions (g and beyond). these are 5D and 7F). Likewise. Within a job. Al-Ar. See the individual descriptions of these keywords later in this chapter for details. when a wavefunction is read from the checkpoint file for use in a calculation using a basis consisting of the other type [391]. Cartesian f functions). Similarly. 7F and 10F: Use 7 or 10 f functions (pure vs. Al-Ar UGBS(1. Cartesian Basis Functions Gaussian users should be aware of the following points concerning pure vs. if you add basis functions for a transition metal from the 6311G(d) basis set via ExtraBasis to a job that specifies the 6-31G(d) basis set in the route section. The preceding keywords may be used to override the default pure/Cartesian setting. When using the ExtraBasis. pure functions will be used unless you explicitly specify 6D in the route section in addition to Gen. 4-31G. Sc-Zn H. if you use a general basis set taking some functions from the 3-21G and 6-31G basis sets. For example. respectively. the exceptions are 3-21G. D95 and D95V. Most also use pure d functions. Issues Arising from Pure vs. Cartesian d functions).3)P Additional Basis Set-Related Keywords The following additional keywords are useful in conjunction with these basis set keywords: • • 5D and 6D: Use 5 or 6 d functions (pure vs. respectively.UGBS MTSmall DGDZVP DGDZVP2 DGTZVP H-Lr H-Ar H-Xe H-F. CEP-31G. and all f and higher functions must be pure or Cartesian. 6-31G†. this section might contain the compound name. This approach expands the density in a set of atom-centered functions when computing the Coulomb interaction instead of computing all of the two-electron integrals. He and B through Ne. In addition. and any other relevant information. relative energies and molecular properties. no fitting set is used. DGA1 is available for H through Xe. but is not interpreted in any way by the Gaussian 03 program. where MaxTyp is the highest angular momentum in the AO basis and MaxVal is the highest valence angular momentum. but this is typically more functions than are needed. Auto=All. PAuto generates all products of AO functions on one center instead of just squares of the AO primitives. density fitting sets can be generated automatically from the AO primitives using Auto. The DGA1 and DGA2 fitting sets [387. Typically. N is the maximum angular momentum retained in the fitting functions. Since archive entries resulting from calculations using a general basis set or the ReadWindow keyword do not contain the original input data for these options. it is strongly recommended that the title sections for these jobs include a complete description of the basis set or frozen-core selection used. or you may select one of the built-in fitting sets.388] are available in Gaussian. It provides significant performance gains for pure DFT calculations on medium sized systems too small to take advantage of the linear scaling algorithms without a significant degradation in the accuracy of predicted structures. and optionally retrieved from the checkpoint file (use ChkBasis to do so). as in this example: # BLYP/6-31G(d)/Auto Note that the slashes are required when a density fitting basis set is specified. the electronic state. The default is Max(MaxTyp+1. The title section cannot exceed five lines and must be followed by a terminating blank line. defined in full with the Gen keyword. In the latter case. its symmetry. The Job Title Section This section is required in the input.36. It appears in the output for purposes of identification and description.Density Fitting Basis Sets Gaussian 03 provides the density fitting approximation for pure DFT calculations [35. Density fitting basis sets may be augmented with the ExtraDensityBasis keyword. .2*MaxVal). The desired fitting basis set is specified as a third component of the model chemistry. By default. or Auto=N. Gaussian 03 can generate an appropriate fitting basis automatically from the AO basis. and DGA2 is available for H.392]. as in an ONIOM calculation. as Cartesian coordinates.. The most general format for the line within it is the following: Element-label[–Atom-type[–Charge]][(param=value[. the entry 0 1 is appropriate. either as Cartesian coordinates or as a Z-matrix definition. or as a mixture of the two (note that Cartesian coordinates are just a special case of the Z-matrix). while the second uses internal coordinates. For the syntax being described here. We'll begin by considering the initial and final items. and especially ^G. Nuclear parameters for this atoms are specified in the parenthesized list. . It is needed only when additional parameters follow the normal data. For a radical anion..The following characters should be avoided in the title section: @ # ! . The remainder of the molecule specification gives the element type and nuclear position for each atom in the molecular system. Thus. for a neutral molecule in a singlet state._ \ all control characters. The optional format-code parameter in the second line specifies the format of the Z-matrix input. There are several ways in which the nuclear configuration can be specified: as a Zmatrix. Lines of both types may appear within the same molecular specification. and possibly an optional molecular mechanics atom type and partial charge. this code is always 0. The remainder of the line contains information about the atom's location. The entire molecule specification (and title section) may be omitted by including Geom=AllCheck in the route section. The first line of the molecule specification section specifies the net electric charge (a signed integer) and the spin multiplicity (a positive integer). The following are the basic formats for specifying atoms within the molecule specification (omitting all of the optional items): Element-label x y z Element-label [n] atom1 bond-length atom2 bond-angle atom3 dihedral-angle [format-code] Although these examples use spaces to separate items within a line. The first form specifies the atom in Cartesian coordinates. .])] Atom-position-parameters Each line contains the element type. This input section specifies the nuclear positions and the number of electrons of α. This is the only molecule specification input required if Geom=CheckPoint is used.and βspin. any valid separator may be used. -1 2 would be used. n is an optional parameter related to freezing atoms during optimizations using ONIOM or (rarely) ones not performed using redundant internal coordinates (see ONIOM for details). and then go on to discuss the remaining items. 02 -0.2 H3 H3 H3 H6 H6 120. The position of the current atom is then specified by giving the length of the bond joining it to atom1.92 1.5 H3. and so on. the remaining items on each line are Cartesian coordinates specifying the position of that nucleus. H8. Here is another Z-matrix form for this same molecule: 0 C1 C2 H3 H4 H5 H6 H7 H8 1 C2 C2 C2 C1 C1 C1 ACCH ACCH ACCH ACCH ACCH ACCH C1 RCC C1 RCH C1 RCH C1 RCH C2 RCH C2 RCH C2 RCH Variables: RCH = 1.39 -0.51 0.00 -0.00 0. If the elemental symbol is used.C2. In the second form.5 RCC = 1. 120.111.-120.1.-120. atom2. Here are two molecule specification sections for ethane: 0 C C H H H H H H 1 0.1.120.88 0.88 0. -120.1.2 H4. the other atoms' line numbers within the molecule specification section may be used for the values of variables.111.H6. where the charge and spin multiplicity line is line 0).C1.180. Note that the first three atoms within the Zmatrix do not use the full number of parameters. C2. the angle formed by this bond and the bond joining atom1 and atom2.39 -0. -120. The version on the left uses Cartesian coordinates while the one on the right represents a sample Z-matrix (illustrating element labels).C1.C2.C2. and the dihedral (torsion) angle formed by the bond joining atom2 and atom3 with the plane containing the current atom.111. atom3 are the labels for previously-specified atoms which will be used to define the current atoms' position (alternatively.111.C1.C2.C1.Element-label is a character string consisting of either the chemical symbol for the atom or its atomic number.2.1 ACCH = 111.51 0.2. In the first form. it may be optionally followed by other alphanumeric characters to create an identifying label for that atom.00 1.88 0.88 0.H3.C1.1.1. only at the fourth atom are there enough previously-defined atoms for all of the parameters to be specified.2.1.39 1. H7. .H6.00 0. C3.1.1. H5. atom1.02 0. H6.1.111.1. A common practice is to follow the element name with a secondary identifying integer: C1.C2. atom1 and atom2.111.1 C1 C2.120.00 1. this technique is useful in following conventional chemical numbering.1.51 -0.00 0.00 -0.H3. 180.C1.2.51 -1.2.H3.52 -0.C1.C2.1.1.92 0.92 1. 1434584477.0.9860944599.1 5 7 0 0 -0. Note that it specifies TV as the atom symbol.000000 0.733871 -0.0. see the examples for the Opt keyword for more details.0.2.3005160849 C.1.9131400756 H. Specifying Periodic Systems Periodic systems are specified with a normal molecule specification for the unit cell.5258773194.5015912465. This Z-matrix form may be used at any time.4.0.e. as are the C-C bond distances and the C-C-H bond angles.3762843235 H.-0.0..0316702826 H.1. The following molecule specification could be used for a two-dimensional PBC calculation on BN: 0.0.772234452.-0.p)/Auto SCF=Tight neoprene.8372739404.0.0.635463 -0.9237098673.6209825739.-0. indicating the replication direction(s).-0.0220081655.1.5737182157.0966261209 H.-0.1. For example.-0. The only additional required input are one. the following input specifies a one-dimensional PBC single point energy calculation for neoprene: # PBEPBE/6-31g(d. the literal bond lengths and angle values have been replaced with variables.0.9267226529.-1.-0.-1.1714181332.4060180273.0792380182.733871 .3523143977.0. In the latter case.5790889876.0. The C-H bond distances are all specified by the same variable.-1. and it is required as the starting structure for a geometry optimization using internal coordinates (i.2791284293 C.0.2511397907.915962512 H. Variable definitions are separated from the atom position definitions by a blank line or a line like the following: Variables: Symmetry constraints on the molecule are reflected in the internal coordinates.7876398696 TV.8477468928.3627869487. The values of the variables are given in a separate section following the specification of the final atom.0.-0. [-CH2-CH=C(Cl)-CH2-] optimized geometry 0 1 C.9206168644.5112729831 The final line specifies the translation vector.1548899113.752511583 Cl.-2.In this Z-matrix.7653394047.635463 0.-1.000000 0.770750797 C. Opt=Z-matrix). the variables indicate the items to be optimized. two or three translation vectors appended to the molecule specification (with no intervening blank line).-1. 635463 0.825000 2.000000 2. located at the origin. the program will automatically use the corresponding actual exact isotopic mass (e.000000 5.825000 1. The following items may be included in the list of parameters: • • • • • Iso=n: Isotope selection. and Gaussian uses the value 17.99916).650000 Specifying Isotopes and other Nuclear Parameters Isotopes and other nuclear parameters can be specified within the atom type field using parenthesized keywords and values.467642 -1.825000 1.0 0.000000 2. GFac=n: Nuclear g-factor.000000 0.237500 1.237500 0.429118 0.000000 0.412500 4.000000 0.475315 -1.000000 1.541855 0.g.412500 4.000000 5.000000 2.000000 0..412500 0.000000 0.237500 4. QMom=n: Nuclear quadrupole moment.825000 0.000000 2.825000 2.000000 Finally.000000 0.237500 4.412500 4. ZEff=n: Effective charge.000000 1.000000 1.000000 0.000000 Here is the molecule specification for a graphite sheet: 0 1 C C TV TV 0.650000 0.412500 1.000000 0.219952 0.403026 0.635463 0.Spin=3) 0. here is the molecule specification that could be used for a three-dimensional PBC calculation on gallium arsenide: 0 1 Ga Ga Ga Ga As As As As TV TV TV 0. Molecular Mechanics Atom Types . 18 specifies 18O.650000 0.000000 2.412500 4. If integers are used to specify the atomic masses.000000 0.0 The line specifies a 13C atom with a nuclear spin of 3/2 (3 * 1/2).000000 0. Spin=n: Nuclear spin.237500 5.7 5 TV TV 0 0 0 0 0.000000 2. in units of 1/2.000000 0.000000 2.133447 0. as in the following example: C(Iso=13.000000 0.825000 0.467642 4. and the ESR g tensor and the electronic spin-molecular rotation hyperfine tensor (NMR Output=Pickett). This parameter is used in spin orbit coupling (see CASSCF=SpinOrbit).237500 1.0 0. . Here are some examples: C-CT C-CT-0. Such calculations differ slightly from ones requested with Massage in previous versions of Gaussian in that they include the grid points from the ghost atoms in DFT XC quadrature.0 Isotope specifications .32 O-O--0.0 2. This requests a counterpoise calculation. Nuclear parameters can also be defined. The new way is a more consistent superposition correction and also easier to use. Atom types and optional partial charges can be specified for each atom.5. Note that counterpoise calculations can also be requested with the Counterpoise keyword. Specifies an SP3 aliphatic carbon atom with a partial charge of 0..5 Specifies an SP3 aliphatic carbon atom.ReadIsotopes) Frequencies at 300 K charge and spin 300.32. Multiple Gaussian jobs may be combined within a single input file.1(Spin=3) Specifying Ghost Atoms An atom with mechanics type Bq (e.Molecule specifications for molecular mechanics calculations may also include atom typing and partial charge information. as in these examples: C-CT(Iso=13) C-CT--0. with its normal basis functions and numerical integration grid points but no nuclear charge or electrons. The input for each successive job is separated from that of the preceding job step by a line of the form: --Link1-- Here is an example input file containing two job steps: %Chk=freq # HF/6-31G(d) Freq Frequencies at STP Molecule specification --Link1-%Chk=freq %NoSave # HF/6-31G(d) Geom=Check Guess=Read Freq=(ReadFC.g. "O-Bq") is set up as a ghost [393] of the corresponding atom. Specifies a carbonyl group oxygen atom with a partial charge of -0. Gaussian 03 Keywords Online Help TOC # Archive CASSCF Charge CNDO CPHF Dreiding Frozen Core Options Frequency GFInput Hartree-Fock IOp MaxDisk MP* Keywords Opt PM3 Prop ReArchive SCF Stable Test UFF Zindo Obsolete Keywords References AM1 BD CBSExtrapolate CID Constants DensityFit Amber BOMD CCD CIS Counterpoise Density Functional Methods ExtraBasis Force Geom GVB Integral LSDA MNDO ONIOM PBC Pressure QCISD Scan Sparse Temperature Transformation W1U Program Development Keywords ADMP B3LYP CBS Keywords ChkBasis Complex Density ExtendedHuckel External Field G* Keywords GFPrint Huckel IRC MINDO3 Name Output Polar Pseudo SAC-CI SCRF Symmetry TestMO Units Link 0 Commands FMM Gen Guess INDO IRCMax MM NMR OVGF Population Punch Scale SP TD TrackIO Volume Non-Standard Routes # .line. and then again at 300 K and 2 atmospheres.This input file computes vibrational frequencies and performs thermochemical analysis at two different temperatures and pressures: first at 298.15 K and 1 atmosphere. Note that a blank line must precede the --Link1-. . For most jobs. the calculation defaults to HF/STO-3G SP. ALTERNATE FORMS #N Normal print level. Fictitious masses for the electronic degrees of freedom are set automatically [188] and can be small enough that thermostats are not required for good energy conservation.179.178.396]). in which the Kohn-Sham molecular orbitals. ADMP belongs to the extended Lagrangian approach to molecular dynamics using Gaussian basis function and propagating the density matrix.395. ADMP This keyword requests a classical trajectory calculation [177.189. this is the default. CP calculations are usually carried out in a plane wave basis (although Gaussian orbitals are sometimes added as an adjunct [394.190]. all of the information can be placed on this first line. ψi.180] using the Atom Centered Density Matrix Propagation molecular dynamics model [188. The best known method of this type is Car-Parrinello (CP) molecular dynamics [191].The route section of a Gaussian job is initiated by a pound sign (#) as the first non-blank character of a line. it is not necessary to use pseudopotentials on hydrogen or to use Deuterium rather than hydrogen in the dynamics. If no keywords are present in the route section. as well as covergence information in the SCF. but overflow to other lines (which may but need not begin with a # symbol) is permissible. #P Additional output is generated. The route section must be terminated by a blank line. This method provides equivalent functionality to Born-Oppenheimer molecular dynamics (see the BOMD keyword) at considerably reduced computational cost [188]. Unlike plane wave CP. This includes messages at the beginning and end of each link giving assorted machine-dependent information (including execution timing data). The remainder of the section is in free-field format. #T Terse output: output is reduced to essential information and results. are chosen as the dynamical variables to represent the electronic degrees of freedom in the system. It can be applied to molecules. NuclearKineticEnergy is a synonym for this option. no K-integration). The other alternative is Choleski. the maximum number of steps will default to the number specified in the original calculation... Len. The Morse parameter data is used to determine the vibrational excitation of diatomic fragments using the EBK quantization rules. First.ADMP can be performed with semi-empirical.] Optional initial Cartesian velocities (ReadVelocity and ReadMWVelocity options) Entire section is repeated NTraj times Optional Morse params. clusters and periodic systems. MaxPoints=n Specifies the maximum number of steps that may be taken in each trajectory (the default is 50). and the Morse curve parameters De (Hartrees) and Be (Angstroms-1). Be . HF. PBC calculations use only the Γ point (i. It consists of the atomic symbols for the two atoms.e.] [Atom1. ADMP calculations can accept some optional input: [Initial velocity for atom 1: x y z Initial velocity for atom 2: x y z . the energy at that distance (E0 in Hartrees).. This input subsection is terminated by a blank line. for each diatomic product Terminate subsection with a blank line. in Angstroms). NKE=N Set the initial nuclear kinetic energy to N microHartrees. OPTIONAL INPUT Although most jobs will not require it. and pure and hybrid DFT models (see availability section below for more details). Initial velocity for atom N: x y z . . the initial velocity for each atom is read if the ReadVelocity or ReadMWVelocity option is included. One complete set of velocities is read for each requested trajectory computation. Lowdin Use the Löwdin basis for the orthonormal set. the bond length between them (Len.. Morse parameter data may also be specified for each diatomic product. De. Atom2. Each initial velocity is specified as a Cartesian velocity in atomic units (Bohr/sec) or as a mass-weighed Cartesian velocity (in amu1/2*Bohr/sec). If a trajectory job is restarted.. which uses the Cholesky basis and is the default. E0.. respectively.. By default.0001 femtoseconds. Restart Restart an ADMP calculation from the checkpoint file. ElectronMass=N Set the fictitious electron mass to |N/10000| amu (the default is N=1000. Note that the velocities must have the same symmetry orientation as the molecule. If you require a spin-unrestricted wavefunction. HF and DFT methods. Semi-empirical. If N<0. DensityKineticEnergy is a synonym for this option. BandGap Whether to diagonalize the Fock matrix in order to report the band gap at each step. This option suppresses the fifth-order anharmonicity correction. core functions are weighted more heavily than valence functions. EMass is a synonym for this option.DKE=N Set the initial density kinetic energy to N microHartrees. FullSCF Do the dynamics with converged SCF results at each point. resulting in a fictitious mass of 0. Int=MNDO or Int=PM3 keyword is required for ADMP jobs using semi-empirical methods. . StepSize=n Sets the step size in dynamics to n*0. Note that options set in the original job will continue to be in effect and cannot be modified. This option suppresses the fifth-order anharmonicity correction. ReadVelocity Read initial Cartesian velocities from the input stream. The Int=AM1. ReadMWVelocity Read initial mass-weighted Cartesian velocities from the input stream. You may also specify alternative isotopes for ADMP jobs using the standard method.1 amu). include the UHF keyword as well. Note that the velocities must have the same symmetry orientation as the molecule. The default is NoBandGap. then uniform scaling is used for all basis functions. 74415774 r3 1.10000 femptosec Ficticious electronic mass = 0. r1 1. the parameters used for the job are displayed in the output: TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ ------------------------------------------------------------------INPUT DATA FOR L121 General parameters: Maximum Steps = 50 Random Number Generator Seed = 398465 Time Step = 0.NDtrC = 1 Multitime step .00000 hartree Initial electr. KE scheme = 0 Multitime step . energy = 0. energy = 0.09413376 a 114.81897892 b 49. kin. = True Initial nuclear kin.08562961 Final blank line At the beginning of an ADMP calculation. Rotate density with nuclei = False = 12 = True = True The molecular coordinates and velocities appear at the beginning of each trajectory step (some output digits are truncated here to save space): Cartesian coordinates: .NDtrP = 1 No Thermostats chosen to control nuclear temperature Integration parameters: Follow Rxn Path (DVV) Constraint Scheme Projection of angular mom.BOMD The following sample ADMP input file will calculate a trajectory for H2CO dissociating to H2 + CO. starting at the transition state: # B3LYP/6-31G(d) ADMP Geom=Crowd Dissociation of H2CO --> H2 + CO 0 C O H H 1 1 r1 1 r2 2 a 1 r3 3 b 2 180.10000 hartree Initial electr.15275608 r2 1.10000 amu MW individual basis funct. 4313539D+00 I= 4 X= 4. and numerical frequencies. summary Kinetic (au) 0. summary information is displayed in the output for each time step in the trajectory: Trajectory summary for trajectory Energy/gradient evaluations Hessian evaluations Trajectory Time (fs) 0.1305362D+00 I= 3 X= 2. AM1 This method keyword requests a semi-empirical calculation using the AM1 Hamiltonian [43.0000000D+00 Z= -1.091965532835 NIter= 10..000000 .0000852 Delta A (h-bar) 0.401.3068948D+12 Z= -2. No basis set keyword should be specified.I= 1 X= -1.739540 .1971360D-01 Y= 0.3305097D+14 Y= -3.0 or other visualization software to display the trajectory path in three dimensions.0500312 -113.0922004D+12 Z= 1.0488505D+13 Y= 6.0000000D+00 Z= -2. and z components of the dipole moment): Energy= -.398.0000000000000000 0.0000000000000021 You can also use GaussView 3. Dipole moment= .0000000D+00 Z= -3.4729976D+13 Z= 1.4220839D+14 TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ After the trajectory computation is complete.0000531 0.0368385D+12 Y= 1.402].000000 -.4547606D+13 Y= -6.1000000 0.0483488 -113.49.0000000000000003 0.300000 .0983706 0.100000 0.1794401D+13 Z= 2.0000000 0.53. Energies.0000150 0.2951936D+14 I= 3 X= -3.0478570D+00 I= 2 X= -1.5350603D-01 Y= 0.0000000D+00 Z= 1.400.0995307 0.399.397.0469941 1 51 51 Delta E (au) 0.0495469 -113.200000 0.48.8527270D+14 I= 4 X= -1.0000000000000009 0.1971360D-01 Y= 0.54.0970481 Potent (au) -113.000000 0.0344227D+00 MW Cartesian velocity: I= 1 X= -4. "analytic" gradients. y..8718451D+00 Y= 0. The AM1 energy appears in the output file as follows (followed by the x.4109897D+14 I= 2 X= 4. . these are referred to as hard-wired parameters. UFF: The UFF force field as described in [39]. They were implemented for use in ONIOM calculations. CHARGE ASSIGNMENT-RELATED OPTIONS Unless set in the molecule specification input. We use this current version from the AMBER web site (amber. Molecular Mechanics Methods There are three molecular mechanics methods available in Gaussian. No basis set keyword should be specified with these keywords. no charges are assigned to atoms by default when using any molecular mechanics force field. Soft parameters are ones specified . The actual parameters (parm96.dat) have been updated slightly since the publication of this paper.edu).e. but they are also available as independent methods. UnCharged Assign QEq charges for all atoms which have charge zero (i. Note that energy differences computed from the values in semi-empirical calculations are in Hartrees and may be compared directly with energy differences computed from jobs using other methods. UnTyped Assign QEq charges only to those atoms for which the user did not specify a particular type in the input.scripps. PARAMETER PRECEDENCE OPTIONS Terminology: Gaussian contains built-in parameter sets for the built-in force fields listed above. The following force fields are available: AMBER: The AMBER force field as described in [37].The energy is as defined by the AM1 model. DREIDING: The DREIDING force field as described in [38]. Options are available to estimate charges at the initial point using the QEq algorithm under control of the following options for any of the mechanics keywords: QEq Assign charges to all atoms using the QEq method [40]. all atoms which were untyped or which were given a type but not a charge in the input). by the user in the input stream for the current job (or a previous job when reading parameters from the checkpoint file). HardFirst Read additional parameters from the input stream. with hard-wired parameters having priority over the read-in. INPUT CONVENTIONS . read-in parameters are used only if there is no corresponding hard-wired value. By default. FirstEquiv If there are equivalent matches for a required parameter. ChkParameters Read parameters from the checkpoint file. The following options specify other ways of dealing with multiple matches. the hard-wired parameters are the only ones used. unless HardFirst is also specified. Any non-standard (soft) parameters present in the checkpoint file are used with higher priority than corresponding hard-wired parameters. use the first one found. Modify Read modifications and additions to the parameter set (after it has been constructed from hard and/or soft parameters). LastEquiv If there are equivalent matches for a required parameter. with soft (read-in) parameters having priority over the hard-wired values. soft ones. ignoring hard-wired parameters. SoftOnly Read parameters from the input stream and use only them. The default is to abort if there are any ambiguities in the force field. even when the latter contains an exact match for the same item. Use SoftFirst if you want to override hard-wired parameter matches. when no relevant option is given. NewParameters Ignore any parameters in the checkpoint file. Hence. it is possible for more than one parameter specification to match a given structure. HANDLING MULTIPLE PARAMETER SPECIFICATION MATCHES Since parameters can be specified using wildcards. Note that wildcards matches within the hardwared parameter set take precidence over soft parameters. SoftFirst Read additional parameters from the input stream. use the last one found. In equations. Consult the AMBER paper [37] for definitions of atom types and their associated keywords. angles are in degrees.0 for pairs separated by one or two bonds. There are a number of ways to implement the calculation of non-bonded interactions. We follow a two-step procedure.32 Specifies an SP3 aliphatic carbon atom with a partial charge of 0. For these methods. Function equivalencies to those found in standard force fields are indicated in parentheses. and frequencies.32. ONIOM. Analytic energies. using a factor 0. distances are in Angstroms. energies are in Kcal/mol and charges are in atomic units. R refers to distances and θ refers to angles.0 for pairs that are separated by three bonds). we subtract out the contributions that . but they are not required. we can use computationally efficient (linear scaling) algorithms. First. In this step. the program will attempt to determine atom types automatically. However. gradients. C-CT-0. interactions between pairs of atoms that are separated by three bonds or less are usually scaled down (in most force fields. Geom=Connect GENERAL MOLECULAR MECHANICS FORCE FIELD SPECIFICATIONS Unless otherwise indicated. without taking the scaling into account. In the second step.0 and 1. we calculate the interactions between all pairs.5. O-O--0.5 Specifies a carbonyl group oxygen atom with a partial charge of -0.AMBER calculations require that all atom types be explicitly specified using the usual notation within the normal molecule specification section: C-CT Specifies an SP3 aliphatic carbon atom. They are indicated by a 0 or an asterisk. and some value between 0. Wildcards may be used in any function definition. the non-bonded (Vanderwaals and electrostatic) interactions are evaluated for every possible pair of atoms. Atom types and charges may also be provided for UFF and DREIDING calculations. In MM force fields. 0 for acceptor type. In the soft force field input. NonBon V-Type C-Type. but were included in the first step. VDW94 Atomic-pol NE Scale1 Scale2 DFlag Atomic-pol Atomic polarizability (Angstrom3). V-Cutoff C-Cutoff VScale1 VScale2 VScale3 CScale1 CScale2 CScale3 V-Type is the Vanderwaals type: 0 No Vanderwaals 1 Arithmetic (as for Dreiding) 2 Geometric (as for UFF) 3 Arithmetic (as for Amber) 4 MMFF94-type Vanderwaals .0. the computer time for this step scales again linearly with the size of the system. NE Slater-Kirkwood effective number of valence electrons (dimensionless). you can specify just the non-bonded master function NonBon. and the NBTerm entry is used for the subsequent subtraction of the individual pairs.0 for donor type atom. Vanderwaals parameters. MMFF94 electrostatic buffering Buf94 Atom-type Value Non-bonded interaction master function. 2. used for NBDir and NBTerm (See MMFF94 below for MMFF94-type Vanderwaals parameters). Since this involves only pairs that are close to each other based on the connectivity. to make things easier. otherwise 0. the NBDir function entry corresponds to the calculation of all the pairs. which is automatically expanded into the actual functions NBDir and NBTerm during pre-processing. the overall algorithm is the more efficient than the alternatives. Although at first sight it seems that too much work is done. This function will be expanded into pairs and a direct function (NBDir and NBTerm) before evaluation of the MM energy. Scale2 Scale factor (dimensionless). DFlag 1. However.should have been scaled. VDW Bond-length Well-depth MMFF94 type Vanderwaals parameters (used for NBDir and NBTerm). Scale1 Scale factor (Angstrom1/4). and C-Cutoff as above. NBDir V-Type C-Type V-Cutoff C-Cutoff V-Type. and C-Scale as above. If any scale factor < 0. V-Scale. C-Type.C-Type is the Coulomb type: 0 No Coulomb 1 1/R 2 1/R2 3 1/R buffered (MMFF94) V-Cutoff and C-Cutoff are the Vanderwaals and Coulomb cutoffs (respectively): 0 No cutoff >0 Hard cutoff <0 Soft cutoff VScale1-3 are Vanderwaals scale factors for 1 to 3 bond separated pairs. C-Type.2 scaling is used (as for Amber). V-Cutoff. Atomic single bond radius AtRad Atom-type Radius Effective charge (UFF) EffChg Charge GMP Electronegativity (UFF) EleNeg Value Step down table .0. V-Cutoff. the 1/1. Coulomb and Vanderwaals single term cutoffs NBTerm Atom-type1 Atom-type2 V-Type C-Type V-Cutoff C-Cutoff V-Scale C-Scale V-Type. Coulomb and Vanderwaals direct (evaluated for all atom pairs). C-Cutoff. CScale1-3 are Coulomb scale factors for 1 to 3 bond separated pairs. Xi and Xj are GMP electronegativity values defined with EleNeg.12*Zi*Zj/(Rij3) Electronegativity correction: Ri*Rj*[Sqrt(Xi) .PropC*lnBO)*(Ri + Rj) + Ren Force constant: k = 664. Harmonic stretch III (UFF [1a]): k*(R-Rij)2 Equilibrium bond length Rij = (1 . it is determined on-the-fly) PropC Proportionality constant Ri and Rj are atomic bond radii defined with AtRad. Zi and Zj are the effective atomic charges defined with EffChg. Morse stretch I (Amber): DLim*(e-a(R-Req)-1)2 where a = Sqrt(ForceC/DLim) MrsStr1 Atom-type1 Atom-type2 ForceC Req DLim ForceC Force constant Req Equilibrium bond length DLim Dissociation limit .Sqrt(Xj)]2/(Xi*Ri + Xj*Rj) HrmStr3 Atom-type1 Atom-type2 BO PropC BO Bond order (if <0. Harmonic stretch I (Amber [1]): ForceC*(R-Req)2 HrmStr1 Atom-type1 Atom-type2 ForceC Req ForceC Force constant Req Equilibrium bond length Harmonic stretch II (Dreiding [4a]): ForceC*[R-(Ri+Rj-Delta)]2 HrmStr2 Atom-type1 Atom-type2 ForceC Delta ForceC Force constant Delta Delta Ri and Rj are atomic bond radii specified with AtRad.Table Original-atom-type Stepping-down-type(s). Sqrt(Xj))2/(Xi*Ri + Xj*Rj) MrsStr3 Atom-type1 Atom-type2 BO PropC BO Bond order (if <0. Xi and Xj are GMP electronegativity values defined with EleNeg.12*Zi*Zj/Rij3 Electronegativity correction: Ren = Ri*Rj*(Sqrt(Xi) . Quartic stretch I (MMFF94 [2]): (Req/2)*(R-ForceC)2*[1+CStr*(R-ForceC+(7/12)*CStr2*(R-ForceC)2] QStr1 Atom-type1 Atom-type2 ForceC Req CStr ForceC Force constant (md-Angstrom-1) Req Equilibrium bond length (Angstrom) CStr Cubic stretch constant (Angstrom-1) Atomic torsional barrier for the oxygen column (UFF [16]) UFFVOx Barrier Atomic sp3 torsional barrier (UFF [16]) UFFVsp3 Barrier . Zi and Zj are the effective atomic charges defined with EffChg. it is determined on-the-fly) PropC Proportionality constant Ri and Rj are atomic bond radii defined with AtRad.Morse stretch II (Dreiding [5a]): DLim*exp[-a(Ri+Rj-Delta)]-1)2 where a = Sqrt(ForceC/DLim) MrsStr2 Atom-type1 Atom-type2 ForceC Delta DLim ForceC Force constant Delta Delta DLim Dissociation limit Ri and Rj are atomic bond radii defined with AtRad. Morse stretch III (UFF [1b]): A1*A3*(exp[-a(R-Rij)]-1)2 where a = Sqrt(k/[BO*PropC]) Equilibrium bond length Rij = (1 .PropC*lnBO)*(Ri + Rj) + Ren Force constant k = 664. 12*Zi*Zk*(3*Rij*Rjk*(1-cos(θeq2))-cos(θeq)*Rik2)/Rik5 UFFBnd3 Atom-type1 Atom-type2 Atom-type3 θeq BO12 BO23 PropC θeq Equilibrium angle BO12 Bond order for Atom-type1–Atom-type2 (when <0. it is determined on-the-fly) PropC Proportionality constant Ri. C1 = -4*C2*cos(θeq) and C0=C2*(2*cos(θeq2)+1) Force constant: k = 664. Rj and Rk are atomic bond radii defined with AtRad. Zj and Zk are effective atomic charges defined with EffChg.Atomic sp2 torsional barrier (UFF [17]) UFFVsp2 Barrier Harmonic bend (Amber [1]): ForceC*(T-θeq)2 HrmBnd1 Atom-type1 Atom-type2 Atom-type3 ForceC θeq ForceC Force constant (in kcal/(mol*rad2) θeq Equilibrium angle Harmonic Bend (Dreiding [10a]): [ForceC/sin(θeq2)]*(cos(θ)-cos(θeq))2 HrmBnd2 Atom-type1 Atom-type2 Atom-type3 ForceC θeq ForceC Force constant θeq Equilibrium angle Dreiding Linear Bend (Dreiding [10c]): AForceC*(1+cos(θ)) LinBnd1 Atom-type1 Atom-type2 Atom-type3 ForceC ForceC Force constant UFF 3-term bend (UFF [11]): k*(C0 + C1*cos(θ))+C2*cos(2θ) where C2=1/(4 * sin(θeq2)). Zi. Xi. it is determined on-the-fly) BO23 Bond order for Atom-type2–Atom-type3 (when <0. Xj and Xk are GMP electronegativity defined with EleNeg. . Zi.12*Zi*Zk*(3*Rij*Rjk*(1-cos(Per2))-cos(Per)*Rik2)/Rik5 UFFBnd2 Atom-type1 Atom-type2 Atom-type3 Per BO12 BO23 PropC Per Periodicity: 2 for linear. Zero bend term: used in rare cases where a bend is zero. it is determined on-the-fly) PropC Proportionality constant Ri. BO12 Bond order for Atom-type1–Atom-type2 (when <0. ZeroBnd Atom-type1 Atom-type2 Atom-type3 Cubic bend I (MMFF94 [3]): (ForceC/2)*(1+CBend*(θ-θeq))*(θ-θeq)2 CubBnd1 Atom-type1 Atom-type2 Atom-type3 ForceC θeq CBend ForceC Force constant (in md*Angstrom/rad2) θeq Equilibrium angle CBend "Cubic Bend" constant (in deg-1) MMFF94 Linear Bend (MMFF94 [4]): ForceC*(1+cos(θ)) LinBnd2 Atom-type1 Atom-type2 Atom-type3 ForceC ForceC Force constant (md) Amber torsion (Amber [1]): Σi=1. This term is needed for the program not to protest about undefined angles.UFF 2-term bend (UFF [10]): [k/(Per2)]*[1-cos(Per*θ)] Force constant: k = 664.4 (Magi*[1+cos(i*θ-I(i+4))])/NPaths AmbTrs Atom-type1 A-type2 A-type3 A-type4 PO1 PO2 PO3 PO4 Mag1 Mag2 Mag3 Mag4 NPaths . Rj and Rk are atomic bond radii defined with AtRad. Xj and Xk are GMP electronegativity defined with EleNeg. it is determined on-the-fly) BO23 Bond order for Atom-type2–Atom-type3 (when <0. 3 for trigonal. Zj and Zk are effective atomic charges defined with EffChg. Xi. 4 for square-planar. UFF torsion with atom type-based barrier height (UFF [16]): [V/2]*[1-cos(Period*PO)* cos(Period*θ)]/NPaths where V=Sqrt(Vj*Vk) UFFTor1 Atom-type1 Atom-type2 Atom-type3 Atom-type4 Period PO NPaths .PO1-PO4 Phase offsets Mag1. determined on-the-fly).. it is determined on-the-fly) NPaths Number of paths (when <0. determined on-the-fly. UFF torsion with bond order based barrier height (UFF [17]): [V/2]*[1-cos(Period*PO)* cos(Period*θ)]/NPaths where V = 5*Sqrt(Uj*Uk)*[1+4.Mag4 V/2 magnitudes NPaths Number of paths (if < 0. When zero or less. determined on-the-fly). Dreiding torsion (Dreiding [13]): V*[1-cos(Period*(θ-PO))]/(2*NPaths) DreiTrs Atom-type1 Atom-type2 Atom-type3 Atom-type4 V PO Period NPaths V Barrier height V PO Phase offset Period Periodicity NPaths Number of paths (if < 0. it is determined on-the-fly) Uj and Uk are atomic constants defined with UFFVsp2.18*Log(BO12)] UFFTorB Atom-type1 Atom-type2 Atom-type3 Atom-type4 Period PO BO12 NPaths Period Periodicity PO Phase offset BO12 Bond order for Atom-type1–Atom-type2 (when <0. UFF torsion with constant barrier height (UFF [15]): [V/2]*[1cos(Period*PO)*cos(V*θ)]/NPaths UFFTorC Atom-type1 Atom-type2 Atom-type3 Atom-type4 Period PO V NPaths Period Periodicity PO Phase offset V Barrier height V NPaths Number of paths.. When zero or less.0. Vj and Vk are atomic constants from UFFVOx.0. it is removed. Period=6. When zero or less.0.Period Periodicity PO Phase offset NPaths Number of paths. During processing. with the following parameters: • • • If there are three atoms bonded to the third center and the fourth center is H. Otherwise. then these values are used: V=4.0.0. Vj and Vk are atomic constants defined with UFFVsp3. UFF torsion with atom type based barrier height (UFF [16]) (differs from UFFTor1 in that the atomic parameter that is used): [V/2]*[1cos(Period*PO)*cos(Period*θ)]/NPAths where V=Sqrt(Vj*Vk) UFFTor2 Atom-type1 Atom-type2 Atom-type3 Atom-type4 Period PO NPaths Period Periodicity PO Phase offset NPaths Number of paths. OldTor Atom-type1 Atom-type2 Atom-type3 Atom-type4 Improper torsion (Amber [1]): Mag*[1+cos(Period*(θ-PO))] ImpTrs Atom-type1 Atom-type2 Atom-type3 Atom-type4 Mag PO Period Mag V/2 Magnitude PO Phase offset Period Periodicity Three term Wilson angle (Dreiding [28c].0. Wilson Atom-type1 Atom-type2 Atom-type3 Atom-type4 ForceC C1 C2 C3 . and at least one of them is H. and NPaths=-1. UFF [19]): ForceC*(C1 + C2*cos(θ) + C3*cos(2θ)) averaged over all three Wilson angles θ.0. Period=3. and NPaths=-1. these values are used: V=1. but the fourth center is not H. determined on-the-fly. PO=0. it is replaced with DreiTRS.0. determined on-the-fly. If there are three atoms bonded to the third center. PO=0. Dreiding special torsion for compatibility with Gaussian 98 code. g. HrmStr-1. HrmStr-2 and so on).g. For dihedral angles.00 ≤ bond order < 1..50 ≤bond order < 2. Req23 Equilibrium bond lengths θeq Equilibrium angle USING SUBSTRUCTURES Substructures may be used to define different parameter values for a function for distinct ranges of some geometrical characteristic.50 The following substructures apply to functions for bond angles (values in degrees): First substructure: • • • -1 -2 -3 0 ≤ θ ≤ 45 45 < θ ≤ 135 135 < θ ≤ 180 Second substructure: • -i-n Number of atoms bonded to the central one. The following substructures apply to functions related to bond stretches: • • • -1 -2 -3 Single bond: 0.50 Triple bond: bond order ≥ 2. C3 Coefficients Harmonic Wilson angle (MMFF94 [6]): (ForceC/2)*(θ2) summed over all three Wilson angles θ.50 Double bond: 1. ForceC2 Force constants (in md/rad) Req12. Use a zero for the first substructure to specify only the second substructure. C2. one or two substructures may be used (e. Substructure numbers are appended to the function name.ForceC Force constant C1. HrmWil Atom-type1 Atom-type2 Atom-type3 Atom-type4 ForceC ForceC Force constant Stretch-bend I (MMFF94 [5]): (ForceC1*(R12-Req12)+ForceC2*(R32-Req23))*(θ-θeq) StrBnd1 Atom-type1 Atom-type2 Atom-type3 ForceC1 ForceC2 Req12 Req23 θeq ForceC1. . separated by a hyphen (e.. AmbTrs-1-2). 30 ≤ bond order ≤ 1.00 ≤ bond order < 1.50 C_2 500. Not all job types may be archived.0 180.08 C_2 350.0 1. See the discussion of the individual keywords for such limitations. Rearchive.70) Amide central bond (priority over resonance) None of the above Here is some simple MM force field definition input: HrmStr1 HrmStr1-1 HrmStr1-2 HrmBnd2 DreiTrs-1 DreiTrs-2 H_ C_2 C_2 * * * C_2 360. archive entries are still listed at the end of the output file (from which they may be extracted at a later time if desired).0 C_2 C_2 * 5. Test .First substructure: • • • • -0 -1 -2 -3 Skip this substructure (substructure "wildcard") Single central bond: 0.0 2.0 Archive This keyword directs Gaussian to place the results from the calculation into the site archive (job results database) if the job completes successfully.0 -1.0 180.50 Double central bond: 1.0 C_2 C_2 * 45.50 Triple central bond: bond order ≥ 2.50 Second substructure: • • • -i-1 -i-2 -i-3 Resonance central bond (1. In this case. The GAUSS_ARCHDIR environment variable specifies the location of the archive files. The Test keyword may be used to suppress automatic archiving.0 120.0 -1. Archiving is also disabled by default whenever the IOp keyword is used to set internal program options.0 2.0 1.50 ≤ bond order < 2. NoTest is a synonym for Archive.0 1.40 C_2 * 50. the Archive keyword can override this. .1. Note. TQ Requests a Brueckner Doubles calculation with triples and quadruples contributions [64] added.Here is a sample archive entry. however. T Requests a Brueckner Doubles calculation with a triples contribution [73] added. BD This method keyword requests a Brueckner Doubles calculation [73.1.447e-04\PG=C02V [C2(O1). that it does not include quantities which can be rapidly recomputed from them (such as thermochemistry results for a frequency calculation). For those job types which cannot be archived. sections are separated by multiple backslashes. the following line will appear in the output file in place of the archive entry: This type of calculation cannot be archived. and the entry ends with an at sign (@). It also contains the molecule specification or optimized geometry and all of the calculation's essential results. BD-T is a synonym for BD(T). Fields within the archive entry are separated by backslashes.1. date.1\O\H.SGV (H2)]\\@ The lines of the archive entry are wrapped without regard to word breaks. The archive entry records the site.2.\H.1\HF=-74. B3LYP See DFT Methods below.\\V ersion=IBM-RS6000-G03RevA.9490523\RMSD=5.1.120.74]. and program version used for the calculation. as it appears at the conclusion of a Gaussian 03 output file: 1\1\GINC-JANIS\SP\RHF\STO-3G\H2O1\MJFRISCH\24-Oct-2004\0\\#T TEST POP=NONE\\Water single point energy\\0. . as well as the route section and the title section for the job. user. following the final correlation iteration: DE(CORR)= -.55299518D-01 E(CORR)= -. T4(aaa)= . Analytic energies. This is the default calculation mode. and numerical frequencies. The BD energy appears in the output labeled E(CORR). MaxCyc=n Specifies the maximum number of cycles. BOMD .62349557751D-04 BD(TQ) = -. Disk space used for TT scratch files : 512 words E5TTaaa = .75019628089D+02 The energy is given in Hartrees.44495423D-04 DE5 = -.24701500D-04 E5TQ2 = .00000000D+00E5TTaab = -.00000000D+00 T4(aab)= -.FC This indicates "frozen-core.40349028D-04 T4(abb)= -.00000000D+00 E5TT = -. numerical gradients. See FC for full information. If triples (or triples and quadruples) were requested. labeled BD(TQ).68473650D-05 EQQ2 = -.12350750D-04 E5TTbbb = . the energy including these corrections appears after the above: Brueckner Doubles with Triples and Quadruples (BD(TQ)) ======================================================== Saving the triples amplitudes on disk. The final energy appears in the last line.40349028D-04 T4(bbb)= .12350750D-04 E5TTabb = -. using 192 words of disk.10 seconds.75019771137D+02 The section gives information about the computation of the non-iterative triples and quadruples correction." and it implies that inner-shells are excluded from the correlation calculation.00000000D+00 Time for triples= . R2. this value will be zero (a blank line is also allowed). Note that the ADMP method provides equivalent functionality at substantially lower computational cost at the Hartree-Fock and DFT levels. R4.DPert.SBeta. When the number of dissociation paths is greater than zero.. initial Cartesian coordinates and velocities may be read in..178. Alternatively.. REQUIRED INPUT All BOMD jobs must specify the number of dissociation paths. The selection of the initial conditions using quasi-classical fixed normal mode sampling and the final product analysis are carried out in the same manner as in the classical trajectory program VENUS [404].. IFragNAtoms information ... ITest.186. The implementation in Gaussian 03 [184. The latter uses a fifth-order polynomial or a rational function fitted to the energy. .. the trajectory is integrated for a fixed number of steps. . gradient and Hessian at the beginning and end points of each step. JAtom. the full BOMD job input has the following general structure: NPath dissociation paths (maximum=20) IFrag1.Ef. as specified by the program default of 100 or the value of the MaxPoints option. (SimAnneal) [Mode-num.182]. G5. In this case. times [R1. IAtom.] times [Estart.] normal mode energies (NSample) [Initial velocity for atom 1: x y z velocities Number of Fragmentation Repeated NPath Optional stopping Repeated NPath Optional Optional initial Optional initial .IFlag] simulated annealing params.179. and no other BOMD input will be used.187] extends the usual methodology by using a very accurate Hessian based algorithm that incorporates a predictor step on the local quadratic surface followed by a corrector step. This method for generating the correction step enables an increase in the step size of a factor of 10 or more over previous implementations.. see [403] for an extended review article). If NPath is set to -1. for many jobs. R6 criteria (ReadStop option) .DelE.This keyword requests a classical trajectory calculation [177. R3. VibEng(Mode-num).180] using a BornOppenheimer molecular dynamics model (first described in [181.. the dissociation pathways will be detected automatically and a gradient criteria (Hartree/Bohr) will be used in place of the usual fragmentation pathway and stopping criteria. . Note that fragment information for each path must begin on a new line..Initial velocity for atom 2: x y z or ReadMWVelocity) . Ef is the Fermi energy (wavenumbers): all modes corresponding to a frequency in wavenumbers below Ef will be enhanced. (ReadVelocity Entire section is Optional Morse Terminate The input line(s) following NPath define the fragmentation information for each path. for each diatomic product . Len. distance between atoms IAtom and JAtom > R6 (0) Otherwise. a value of zero for any parameter turns off testing for the corresponding stopping criteria. DPert is the size of the random perturbation.. Initial velocity for atom N: x y z . whole those above Ef will be reduced. The stopping criteria tests are defined as follows (default parameter values are in parentheses): • • • • • • • Minimum distance between the centers of mass for any pair of fragments > R1 (18) Minimum distance between atoms located in different fragments > R2 (20) Maximum distance between the center of mass and any atom in the same fragment < R3 (0) The maximum distance between any pair of atoms in the same fragment < R4 (0) Interfragment gradient < G5 (10-6) If ITest=1.e. It has the . Be params. SBeta is the Fermi-Dirac inverse temperature (1/Hartrees). Up to six stopping criteria may be specified for each path. distance between atoms IAtom and JAtom < R6 (0) All distances are specified in Bohr. Atom2. A trajectory is terminated when all of the active criteria are satisfied.. Parameters for simulated annealing/fragmentation follow the stopping criteria in the input stream when the SimAnneal option is specified: • • • • • • Estart is the desired initial kinetic energy (Hartrees). IFlag determines the algorithm for applying an energy perturbation for simulated annealing (i. The reverse will happen if SBeta is negative. The value in each position specifies that the corresponding atom belongs to the specified fragment number (i. but the ones for any individual path may be continued over as many lines as necessary.e...] repeated NTraj times [Atom1. E0. adding/removing energy from the eigenmodes)..] subsection with a blank line. However.. De. and the units of the gradient G5 are Hartrees/Bohr. Stopping criteria are specified next when the ReadStop option is specified. atom i belongs to fragment number IFragi). DelE is the energy gain/loss in Hartrees. N4]]) Defines the phase for the transition vector such that forward motion along the transition vector corresponds to an increase in the specified internal coordinate. and the Morse curve parameters De (Hartrees) and Be (Angstroms-1). three atom numbers specify an angle bend and four atoms define a dihedral angle. Morse parameter data can be specified for each diatomic product. (You can explicitly specify the forward direction using the Phase option. 2 (combination of 0 and 1). Phase=(N1. the maximum number of steps will default to the number specified in the original calculation.following possible values: 0 (weigh each eigencomponent according to its frequency). This input subsection is terminated by a blank line. This option suppresses the fifth-order anharmonicity correction. MaxPoints=n Specifies the maximum number of steps that may be taken in each trajectory (the default is 100). and 10 (ignore any nearby transition state). Note that the velocities must have the same symmetry orientation as the molecule. the bond length between them (Len. The Morse parameter data is used to determine the vibrational excitation of diatomic fragments using the EBK quantization rules. the initial velocity for each atom is read if the ReadVelocity or ReadMWVelocity option is included. If VibEng < 0. ReadMWVelocity Read initial mass-weighted Cartesian velocities from the input stream. If two atom numbers are given. respectively. the coordinate is a bond stretch between the two atoms. . then the initial velocity is in the reverse direction. One complete set of velocities is read for each requested trajectory computation. The next part of the input specifies how much energy is in each normal mode when the NSample option is used. 1 (add DelE in a random fashion). Finally. For each mode.N2 [. add all energy along that mode). Each initial velocity is specified as a Cartesian velocity in atomic units (Bohr/sec) or as a mass-weighed Cartesian velocity (in amu1/2*Bohr/sec). Note that the velocities must have the same symmetry orientation as the molecule. ReadVelocity Read initial Cartesian velocities from the input stream.) Next. the energy at that distance (E0 in Hartrees). in Angstroms). If a trajectory job is restarted.N3[. It consists of the atomic symbols for the two atoms. This option suppresses the fifth-order anharmonicity correction. VibEng is the translational energy in kcal/mol in the forward direction along the transition vector. 00 (if near a transition state. Additional parameters are read in. The default is Fixed normal mode energy unless RTemp was specified. NSample=N Read in initial kinetic energies for the first N normal modes (the default is 0). StepSize=n Sets the step size in dynamics to n*0. Sample=type Specifies the type of sampling. SCF. Restart Restart a trajectory calculation from the checkpoint file. in which case Local mode sampling is implied. ReadStop Read in alternative stopping criteria. and Local. RTemp=N Specifies the rotational temperature. . All semi-empirical. CIS. Only one of ReadVelocity. GradOnly requests that only gradients be done and that the Hessian be updated all the time (full second derivatives are not computed). Microcanonical. Using the Update keyword causes the program to perform Hessians update for n gradient points before doing a new analytic Hessian. ReadMWVelocity and SimAnneal can be specified. Update=n By default BOMD does second derivatives at every point. The energies for the remaining modes are determined by thermal sampling by default. You may also specify alternative isotopes for BOMD jobs using the standard method.0001 femtoseconds. as described above. NTraj=N Compute N trajectories. Note that options set in the original job will continue to be in effect and cannot be modified. Fixed. where type is one of these keywords: Orthant.SimAnneal Use simulated annealing (the initial velocity is randomly generated). CASSCF. MP2 and DFT methods. The default is to choose the initial rotational energy from a thermal distribution assuming a symmetric top (the temperature defaults to 0 K). 8 bohr. one or two stopping criteria are sufficient. In most cases. any atom in the fragment is less than 2. in the direction of the products (the forward direction is characterized by an increase in the larger C-H distance).RTemp=300.145 kcal/mol.0 11.3 bohr from the center of mass of the fragment.94603924 Final blank line Note that all six stopping criteria are used here only for illustrative purposes. The trajectory will be stopped if the distance between the centers of mass of H2 and CO exceed 13 bohr.15275608 r2 1.49458169 2.ADMP The following sample BOMD input file illustrates many of the available options. they were computed in a previous calculation.145 C O -112.73482237 0. the gradient for the separation of the fragments is less than 0.19500473 1.3 2. and the two hydrogens belong to fragment 2. . There is one fragmentation pathway: C and O belong to fragment 1. all atoms in a fragment are less than 1. It will calculate a trajectory for H2CO dissociating to H2 + CO.0000005 hartree/bohr.8 1 5. r1 1.NSample=1. The calculation will be carried out at 300 K. The Morse parameters for H2 and CO are specified to determine the vibrational excitation of the product diatomics. starting at the transition state. The initial kinetic energy along the transition vector is 5.ReadStop Geom=Crowd HF/3-21G dissociation of H2CO --> H2 + CO 0 C O H H 1 1 r1 1 r2 2 a 1 r3 3 b 2 180. the closest distance between H2 and CO exceeds 11 bohr.5 0.12895435 0.74415774 r3 1.24078955 H H -1.5 bohr from all other atoms in the fragment.12295984 0. # HF/3-21G BOMD(Phase=(1.09413376 a 114.0000005 1 1 3 12.0 1.81897892 b 49. and the distance between atoms 1 and 3 is greater than 12.08562961 1 1 1 2 2 13.09329898 1. Stopping criteria are also specified in this example job.3). 89754 6 3168.859 0 2.99064 5 2026.000D-05 Hartree Reaction Path 1 **************** Fragment 1 center 1 ( C ) 2 ( O ) Fragment 2 center 3 ( H ) 4 ( H ) Termination criteria: The CM distances are larger than 13.0 K = 1.00D-07 Hartree/bohr Distance between atom center 1 ( C ) and 3 ( H ) is GE 12. the parameters used for the job are displayed in the output: TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ-TRJ ------------------------------------------------------------------INPUT DATA FOR L118 ------------------------------------------------------------------General parameters: Max.1950047 1.52987 .7348224 0.330 0 1.182 0 1.9460392 --------------------------------------------------------------------- The initial kinetic energies for the normal modes appears at the beginning of each trajectory step: ------------------------------------------------------Thermal Sampling of Vibrational Modes Mode Wavenumber Vib.At the beginning of a BOMD calculation..476 0 1.14500 2 837.4945817 2.1229598 0.0 bohr The max atomic and CM distances in frags are shorter than 1.800 bohr Morse parameters for diatomic fragments: E0 Re De Be C O -112.250 sqrt(amu)*bohr Sampling parameters: Vib Energy Sampling Option Vib Sampling Temperature Sampling direction Rot Energy Sampling Option Rot Sampling Temperature Start point scaling criteria .1289544 0.0 K = Forward = Thermal distribution (symmetric top) = 300. = 100 Total Number of Trajectories = 1 Random Number Generator Seed = 398465 Trajectory Step Size = 0..0932990 1.# Energy (kcal/mol) ------------------------------------------------------1 -2212.59137 4 1392. points for each Traj.19702 3 1113.689 0 4. quant.500 bohr The change of gradient along CM is less than 5.2407896 H H -1. = Thermal sampling = 300.3 bohr The max atomic distances in fragments are short than 2.761 5.000 bohr The min atomic distances among fragments are larger than 11. 0468302 0.0000000000000000 -0.6)-on a closed-shell system.0000144 0.. Since 2 occupied orbitals were included.195.169296 2.0388912 0. all reported at each time step. the lowest 4 virtual orbitals would become part of the active space.0000000000053006 -0. the output includes geometrical parameters for the atoms in each fragment. Thus.0000000000045404 The information is given for each time step in the trajectory. An MC-SCF calculation is a combination of an SCF computation with a full CI involving a subset of the orbitals. summary information is displayed in the output: Trajectory summary for trajectory Energy/gradient evaluations Hessian evaluations Trajectory Time (fs) 0.0293490 -113.M). and the relative mass-weighted velocities for each fragments and between fragments.------------------------------------------------------- After the trajectory computation is complete.98. the 2 highest occupied MOs would be included. the active space would consist of: • • Enough occupied orbitals from the guess to provide 4 electrons.405]. the active space is defined assuming that the electrons come from the highest occupied orbitals in the initial guess determinant and that the remaining orbitals required for the active space come from the lowest virtuals of the initial guess. You can also use GaussView 3. Enough virtual orbitals to make a total of 6 orbitals. Note that options may be interspersed with N and M in any order. for a 4electron. Thus.0 or other visualization software to display the trajectory path in three dimensions. In addition. The number of electrons (N) and the number of orbitals (M) in the active space for a CASSCF must be specified following the keyword: CASSCF(N. .0214192 -113.138.0000091 0. CASSCF This method keyword requests a Complete Active Space Multiconfiguration SCF (MCSCF) [97..0000000 0. By default.000000 1. summary Kinetic (au) Potent (au) 0. 6-orbital CAS-specified as CASSCF(4. this subset is known as the active space. the distances between fragments.137.0407383 -113.0582248 1 76 76 Delta E (au) Delta A (h-bar) 0.161873 . Above 8 orbitals. Calculations on excited states of molecular systems may be requested using the NRoot option. and the subsequent job will include Guess=(Read. You need to include Pop=Regular in the route section of the preliminary job in order to include the orbital coefficient information in the output (use Pop=Full for cases where you need to examine more than just the few lowest virtual orbitals). CAS is a synonym for CASSCF. You may also choose to view the orbitals in a visualization package such as GaussView 3. a full Hartree-Fock single point calculation may be done. Alternatively. We strongly recommend that you study the cited references before attempting to run production CASSCF calculations (this is especially true for CASSCF MP2). Note that a value of 1 specifies the ground state. [308]. A brief overview of the CASSCF method is given in chapter 9 (exercises 5 and 6) and appendix A of Exploring Chemistry with Electronic Structure Methods.100. you may use Pop=NBOSave to save the NBOs.0. not the first excited state (in contrast to usage with the CIS keyword). 2nd ed.411. Alternatively. In Gaussian 03.409. VARIATIONS • • An MP2-level electron correlation correction to the CASSCF energy may be computed during a CASSCF calculation by specifying the MP2 keyword in addition to CASSCF within the route section [101].Permute) in order to retrieve and then modify the computed initial guess from the checkpoint file. Normally. algorithmic improvements make an active space of up to about 14 orbitals feasible [99. CASSCF calculations use a direct algorithm to avoid disk storage of integrals. a 4 electron.408.407. the CASSCF code automatically uses this new direct method for matrix elements.Similarly. By default. See reference [138] for a detailed discussion on the choice of an active space. 6 orbital CAS on a triplet would include the highest 3 occupied orbitals (one of which is doubly occupied and two singly occupied in the guess determinant) and the lowest 3 virtual orbitals. A prior run with Guess=Only can be used to quickly determine the orbital symmetries (see the first example below). . A conventional algorithm may be selected by including SCF=Conven in the route section. which are often the best choice for starting CAS orbitals. Use #P in the route section to include the final eigenvalues and eigenvectors in addition to the energy and one-electron density matrix in the CASSCF output. Note: CASSCF is a powerful but advanced method with many subtleties. See this page for a discussion of efficiency considerations for CASSCF calculations.410.102]. Guess=Alter or Guess=Permute is necessary to ensure that the orbitals which are selected involve the electrons of interest and that they are correlated correctly.412]. Example applications are discussed in references [406. The option defaults to the ground state (j=1). Conical intersections and avoided crossings may be computed by including Opt=Conical in the route section of a CASSCF job (see the examples) [165.253. so that an excited state is obtained when j > 1. specified on a separate input line. RASSCF calculations partition the molecular orbitals into five sections: the lowest lying occupieds (doubly occupied in all configurations). the RAS3 space of weakly occupied MOs and the remaining unoccupied orbitals. The Restricted Active Space variation (RASSCF) [103] is now supported [104].8 (no trailing blank line). StateAverage is not allowed in combination with Opt=Conical or CASSCF=SpinOrbit. In order to compute the spin orbit coupling. the active space in CASSCF calculations is divided into three parts in a RAS calculations. The method used in Gaussian 03 is based on reference [254]. and then effective charges are used that scale the Z value for each atom to empirically account for 2 electron effects. Approximate spin orbit coupling between two spin states can be computed during CASSCF calculations by including the SpinOrbit option [250. It is available for the elements H through Cl. See the discussion of the RAS option for the methods for specifying these values.254. and allowed configurations are defined by specifying the minimum number of electrons that must be present in the RAS1 space and the maximum number that may be in the RAS3 space. the integrals are computed in a oneelectron approximation involving relativistic terms." StateAverage Used to specify a state-averaged CASSCF calculation.252. the RAS1 space of doubly occupied MOs. note that such calculations will be state-averaged by default. The state specified by NRoot is referred to as the "state of interest. Implies a state-averaged CASSCF calculation. NRoot=j Requests that the jth root of the CI be used.166. Thus. It is selected via the RAS option. This option requires the weighting for the various states to be input in format nF10. Finally.167]. All states up to NRoot are averaged.414]. This value can be specified for each atom via the molecule specification nuclear parameters list. in addition to the total number of electrons in the three RAS spaces. . both of which perform state-averaged calculations by default. SpinOrbit Compute approximate spin orbit coupling between two states.251.• • • • • • State-averaged CASSCF calculations may be performed using the StateAverage and NRoot options to specify the states to be used.413. the RAS2 space containing the most important orbitals for the problem. FullDiag Requests the use of the full (Jacobi) diagonalization method for the CI matrix instead of Lanczos or Davidson iterations.StateGuess=k) sets C(k) to 1. Note that the two CASSCF keyword parameters specify the size of the entire active space: RAS1 + RAS2 + RAS3 (see the examples).. For example. The default diagonalization method is most efficient if the size of the CI problem is greater than about 50. The StateGuess option (below) may be used to change this default. In such cases. DavidsonDiag Requests the use of the Davidson diagonalization method for the CI matrix instead of the Lanczos iterations.0. and to c particles in the d orbitals in the RAS3 space (i.RAS=(a. the starting vector is initialized in j+1 positions. excitations from RAS1 into RAS2 or RAS3) in the b orbitals in the RAS1 space. otherwise. CASSCF(…. and when one knows nothing at all about the CI eigenvector (in the latter case. The full Jacobi diagonalization method must be used if quadratic convergence is required (see the QC option below).g. where j is the value given to the NRoot option (or its default value). or the user can identify one or more dominant components in the eigenvector from the onset of the calculation. I=1..c. if the CI eigenvector is dominated by configuration k. C(Ind(I)). By default. (Ind(I).. StateGuess=1 is appropriate if the CI vector is dominated by the SCF wavefunction). NoFullDiag suppresses the use of the full diagonalization method. However.e. Thus. running a preliminary calculation to determine the orbital symmetries may be required. NZ).e.. k may also be set to the special value Read.b. specify FullDiag for calculations involving more than 6 active orbitals) StateGuess=k Set the starting vector for the Lanczos method to configuration k.g. Thus.d) Requests a RASSCF calculation which allows up to a holes (i. . Lanczos is the default for NRoot values of 1 or 2. setting the StateGuess option to k will generate a good starting vector (e. The positions correspond to the lowest j+1 energy diagonal elements of the CI Hamiltonian. which says to read in the entire eigenvector from the input stream (format: NZ. excitations from RAS1 or RAS2 into RAS3). The central requirement for this vector is that it not be deficient in the eigenvector that is required. The default is full diagonalization if there are 6 or fewer active orbitals. the minimum number of electrons in RAS2 is 2b-a. Davidson is the default. via the initial trail vector. This usually results in good convergence for the lowest j roots. if the coefficient of configuration k is exactly zero (e. this option can be useful for selecting a configuration of the correct symmetry for a desired excited state (different from that of the ground state). OrbRot OrbRot includes and NoCPMCSCF excludes the orbital rotation derivative contributions from the CP-MC-SCF equations in an Opt=Conical calculation. Often. some of the natural orbitals which have modest occupation are not the important ones for the process of interest.UNO) Guess=Read. one of QC and RFO should be specified. It implies NoFullDiag. . one normally runs one job which does a UHF calculation with Pop=NaturalOrbital. and a single-point CASSCF(…. then that eigenvector will be missing. The UNO guess must be used with caution. and the calculation will converge to a higher state. This is the default for CAS calculations involving 10 or more orbitals. this entire process depends on the user being able to coax the UHF wavefunction to converge to the appropriate broken spin-symmetry (non-RHF) result. and finally an optimization can be run with CASSCF(…. and then examines the resulting orbitals. This option is needed to locate a conical intersection/avoided crossing between a singlet state and a triplet state. This option should be used with caution. Consequently. HWDet Use Hartree-Waller determinants instead of Slater. For singlets.UNO) Guess=(Read. The orbitals which belong in the active space are then selected. NPairs=n Number of GVB pairs outside of the CAS active space in a CAS-GVB calculation [417].by symmetry) in the desired root. The resulting converged orbitals are then examined to verify that the correct active space has been located. OrbRot is the default. it works well only with a very good guess. Alter) calculation is performed. UNO Requests that the initial orbitals for the CAS be produced from the natural orbitals generated from a previous UHF calculation [415. Normally used with Guess=Read. RFO Requests the RFO quadratic step. unless the entire valence space is being correlated (which is usually prohibitively expensive). At most.416]. QC Requests a quadratically convergent algorithm for the CAS. SlaterDet Use Slater determinants in the CASSCF calculation. Only one of QC and RFO should be specified. 3-cyclobutadiene. analytic gradients.29536 0.34716 0.00000 30 4 C 2PX 0. Analytic polarizabilities may not be performed with the CASSCF method. We include Pop=Reg to obtain the molecular orbital output in the population analysis section: # HF/3-21G Guess=Only Pop=Reg Test The molecule being investigated is 1.34716 0. therefore.37752 0.16911 0. and analytic and numerical frequencies.37752 -0. so π orbitals will have significantly non-zero coefficients in the X direction.24339 0. We are going to run a 4x4 CAS.29536 0. The following route section illustrates one method of quickly examining the orbitals in order to determine their symmetries and any alterations needed to produce the desired initial state.00000 0.00000 7 3PX 0.21750 0.24339 16 V 0. You can restart a CASSCF calculation by specifying SCF=Restart in the route section.00000 0. MP2.00000 16 3PX 0.29536 0. Guess.21750 15 V 0. Here are the relevant coefficients for orbitals 10 and 1316: Molecular Orbital Coefficients 10 13 O O 3 1 C 2PX 0.00000 12 2 C 2PX 0.Energies.00000 34 3PX 0. we see that only orbitals 14 and 15 are of the correct type. a singlet with D2h symmetry.00000 0. orbitals 13 through 16 will comprise the active space. The molecule lies in the YZ-plane. We want all four orbitals to be π orbitals.00000 0. Opt=Conical.16911 0.16911 0.34716 -0.00000 21 3 C 2PX 0.00000 25 3PX 0.00000 0.37752 -0. In order to restart a CASSCF optimization.24339 -0.21750 -0.00000 0. When we examine these orbitals.00000 .00000 14 O 0. Pop. so there will be four orbitals in the active space: 2 occupied and 2 virtual. Use CASSCF Polar=Numer.21750 -0.16911 0.37752 0.34716 -0. Preliminary Examination of the Orbitals (Guess=Only). CASSCF may not be combined with any semi-empirical method.29536 0. SCF We will consider several of the most important uses of the CASSCF method in this section.24339 -0. The HOMO is orbital 14.00000 0. the keywords CASSCF Opt=Restart Extralinks=L405 must be included in the job's route section. Orbital 10 is clearly also a π orbital. symbolic density matrix: 2 3 0. We have found our four necessary orbitals.. In either case. D2H.564187D-06 4 0.00D-05 ITN= 10 MaxIt= 64 E= -152. Interchange orbitals 16 and 19.130613D-05 0.. the two "occupied" orbitals have values less than 2. then that orbital was empty throughout the calculation.3-Cyclobutadiene Singlet. If we look at higher virtual orbitals. and there is probably a problem with the CASSCF calculation. DO AN EXTRA-ITERATION FOR FINAL PRINTING The value of E for the final iteration is the predicted energy: -152. there were no excitations into or out of the orbital in question. . and the other two orbitals in the active space have non-zero occupancies. When we run this CASSCF calculation on cyclobutadiene. we will find that orbital 19 is also a π orbital. so things are fine. and can now use Guess=Alter to move them into the active space. It appears in the CASSCF output as follows: TOTAL -152. If any of these values is (essentially) zero.345450D-05 4 0.182680D+01 0.13 16.139172D-05 3 0. Pi 4x4 CAS 0 1 molecule specification 10.8402826495 hartrees in this case.836259 .4)/3-21G Guess=Alter Pop=Reg Test 1.191842D+01 2 -0.820965D-01 The diagonal elements indicate the approximate occupancies for each successive orbital in the active space. energy at each iteration ITN= 9 MaxIt= 64 E= -152.415187D-05 0. Here is the input file for the CASSCF calculation: # CASSCF(4. if any of them is essentially 2. In our case.19 Interchange orbitals 10 and 13.172679D+00 0..00D-05 . It is also important to examine the one-electron density matrix. similarly. then that orbital was doubly occupied throughout the CAS.98D-06 Acc= 1..327584D-06 MCSCF converged.17D-05 Acc= 1.8402826495 DE=-3. we will obtain a prediction for the energy. CASSCF Energy and the One-Electron Density Matrix. which appears next in the output: Final one electron 1 1 0.8402786733 DE=-1. The string EUMP2 labels the energy.15310383973610D+03 Electron correlation-corrected energy. in this case. and an α electron in orbital 15. the α electron from orbital 14 has been excited to orbital 15. the value is -153. in configuration 3. the β electron in orbital 13 remains there. and run the CAS 0. The following two-step job illustrates one method for studying excited state systems using the CASSCF method. The first step assumes that a preliminary Hartree-Fock single point calculation has been done in order to examine the orbitals. there is a β electron in orbital 13. When you run a CASSCF calculation with correlation (CASSCF MP2 in the route section). . and the electron order in the output is: α α β β). This is a 4x4 CAS. it takes advantage of the initial guess computation done by that job.1 orbital alterations . from lowest to highest.. so the primary basis function output indicates that there is an α and b electron in both orbitals 13 and 14 (the numbers refer to the orbitals in the active space. CAS Configuration Information. in the following format: PRIMARY BASIS FUNCTION= 1 2 1 1 3 2 1 2 1 3 2 3 2 SYMMETRY TYPE = 0 2 SYMMETRY TYPE = 0 The first line indicates the electron assignments for the reference configuration.2635549296D+00 EUMP2 = -0.Alter) Pop=NaturalOrbital Test Geom=Check Alter the guess so that the three LUMOs are all the desired symmetry. In configuration 2..CASSCF MP2 Energy. an α (from 13) and β electron in orbital 14. The beginning of the CASSCF output lists the configurations. the α electron in orbital 13 remains there. as does the β electron in orbital 14. Similarly.1038397361 hartrees. E2 = -0. which it retrieves from the checkpoint file: %chk=CAS1 # CASSCF(2. the following additional lines will appear in the CASSCF output (with the first one coming significantly before the second): MP2 correction to the MCSCF energy is computed Indicates a CASSCF MP2 job.4) 6-31+G(D) Guess=(Read. Using CASSCF to Study Excited States. 77417416 . The first excitation energy for the system will then be computed by taking the energy difference between the two states (see exercise 5 in chapter 9 of Exploring Chemistry with Electronic Structure Methods [308] for a more detailed discussion of this technique).31874441E-02 .1 The second job step uses the NRoot option to CASSCF to specify the first excited state.000000D+00 . 0..48879229 .553225D-01 . Predicting Conical Intersections. 2nd state is 2 Transition Spin Density Matrix 1 2 1 . Spin Orbit Coupling. If the two eigenvalues (the first entry in the lines labelled with a state number) are essentially the same..0000000 cm-1 .0000000 cm-1 magnitude in y-direction= . it is an avoided crossing. 0.45467877 -0.4.NRoot=2) 6-31+G(D) Guess(Read) Pop(NaturalOrbital) Geom=Check Test Excited state calculation 0.141313D+01 2 .000000D+00 magnitude in x-direction= . The optimized structure will be that of the conical intersection or avoided crossing.0503161 -154.. Including Opt=Conical keyword in the route section changes the job from an optimization of the specified state using CASSCF to a search for a conical intersection or avoided crossing involving that state. Here is the output from a CASSCF calculation where the spin orbit coupling has been requested with the Spin option (the coupling is between the state specified to the NRoot option and the next lower state): **************************** spin-orbit coupling program **************************** Number of configs= 4 1st state is 1 Identifies the two states between which the spin orbit coupling is computed. and it is a conical intersection.. Otherwise.0501151 0.--Link1-%chk=CAS1 %nosave # CASSCF(2. Distinguishing between these two possibilities may be accomplished by examining the final eigenvalues in the CASSCF output for the final optimization step (it precedes the optimized structure): FINAL EIGENVALUES AND EIGENVECTORS VECTOR EIGENVALUES CORRESPONDING EIGENVECTOR state 1 2 energy -154.16028934E-02 0.72053292 -0. then the energies of the two states are the same... No basis set should be specified with any of these keywords. Montgomery Jr. Ochterski. All of these distinct steps are performed automatically when one of these keywords is specified. and Z components.4)) 6-31G(d) If this molecule is a neutral singlet. CBS-q {Petersson. 1994 #304. CBS-Q {Ochterski. the RAS2 space will have 9 to 13 electrons in all configurations. followed by its total magnitude. The spin orbit coupling is broken down into X.18).2. Montgomery Jr. and RAS3 with 4 orbitals. Thus. Petersson. 1999 #503. Y.18. Petersson.. 2000 #794}. CBS-Q//B3 {Montgomery Jr.RASSCF(1. 1996 #307}. 0-3 electrons in all configurations.e. Here is an example RASSCF calculation route section: # CAS(16.2016070 cm-1 total magnitude= 55. then this route defines the following spaces: RAS1 with 2 orbitals.. 1996 #307. These methods are complex energy computations involving several to many pre-defined calculations on the specified system.magnitude in z-direction= 55. 1981 #301. 2000 #794}.. 1996 #307. 3 or 4 electrons in all configurations. Montgomery Jr.e. . which in this case is 55. 2000 #794} and CBS-APNO {Ochterski. same orbitals as for a regular CAS(16. Montgomery Jr. 12 electrons in the reference configuration. Montgomery Jr.3. and the final computed energy value is displayed in the output.. 1991 #303.2016070 cm-1 MCSCF converged. 1999 #503. Spin orbit coupling. The keywords refer to the modified version of CBS-4 {Ochterski. Petersson. RAS2 with 12 orbitals. respectively. 1991 #302. 1991 #303} (i.. RASSCF example. 1996 #307} methods. CBS-4M CBS-Lq CBS-Q CBS-QB3 CBS-APNO These method keywords specify the various Complete Basis Set (CBS) methods of Petersson and coworkers for computing very accurate energies {Nyden. 1988 #338.. The orbitals taken from the reference determinant for the active space are (assuming a spin singlet) the 8 highest occupieds and 10 lowest virtuals: i.2016070 cm-1. Lq for "little q"). ReadIsotopes Specify alternate temperature. the CBS-4O keyword requests the earlier parametrization. If the previous calculation did not complete. 18 specifies O18. respectively. CBS-APNO is available for first row atoms only. If integers are used to specify the atomic masses.99916). and scale are the desired temperature. CBS-Q and CBS-QB3 are available for first and second row atoms. the program will automatically use the corresponding actual exact mass (e. and the most abundant isotopes).g.. The new version. arranged in the same order as they appeared in the molecule specification section. the ReadIsotopes option is retained for rerunning completed calculations under different conditions (see the examples). it will be completed. and an optional scale factor for frequency data when used for thermochemical analysis (the default is the value defined by the selected method). CBS-4M. and/or isotopes (the defaults are 298.Either of the Opt=Maxcyc=n or QCISD=Maxcyc=n keywords may be used in conjunction with any of the these keywords to specify the maximum number of optimization or QCISD cycles.. where temp. The CBS-4 model chemistry has also been updated with both the new localization procedure and improved empirical parameters {Montgomery Jr. This information appears in a separate input section having the format: temp pressure [scale] isotope mass for atom 1 isotope mass for atom 2 . The remaining lines hold the isotope masses for the various atoms in the molecule. and Gaussian uses the value 17. pressure. isotope mass for atom n Values must be real numbers. pressure. Restart Restart from the checkpoint file from a previous CBS calculation. However. CBS-4M. . 1 atmosphere. CBS-Lq. Energies only.. (M referring to the use of Minimal Population localization) is recommended for new studies.15 K. pressure. 2000 #794}.. You should specify alternative isotopes for CBS jobs using the standard method. 011929 -.043444 The temperature and pressure are given first. All of the energies are in hartrees. 6081 (1991) G.0 isotope specifications CBSExtrapolate . Al-Laham.066592 CBS-Q Free Energy= -39.e. The following two-step job illustrates the method for running a second (very rapid) CBS calculation at a different temperature.The output from each step of a CBS method calculation is included in the output file. A. Montgomery.065647 Pressure= E(Thermal)= DE(MP2)= DE(MP34)= DE(Int)= 1.018702 . JCP 94. Tensfeldt & J. Petersson and M.150000 . The final line gives the CBS-Q enthalpy (including the thermal correction for the specified temperature) and the Gibbs free energy computed via the CBS-Q method (i.069447 -39. Here is the output from a CBS-Q calculation on CH2 (triplet state): Complete Basis Set (CBS) Extrapolation: G.. Temperature= E(ZPE)= E(SCF)= DE(CBS)= DE(QCI)= DE(Empirical)= CBS-Q (0 K)= CBS-Q Enthalpy= 298.004204 CBS-Q Energy= -39. Rerunning the Calculation at a Different Temperature...002781 -. This job computes the CBS-4 energy at 298. Petersson.114652 -.000000 . 6091 (1991) additional references . JCP 94.0 1.019690 -. A.15 K and then again at 300 K: %Chk=cbs # CBS-4 Test CBS-4 on formaldehyde 0 1 molecule specification --Link1-%Chk=cbs %NoSave # CBS-4(Restart.ReadIso) Geom=AllCheck Test 300.936531 -.15 K by default). followed by the components terms of the CBS-Q energy. the CBS-Q energy including the frequency job free-energy correction).005891 -39. The second-to-last line gives the CBS-Q energy values (reading across): at 0 K and at the specified temperature (298. The final section of the file contains a summary of the results of the entire run. CBS Summary Output.016835 -38. T. with or without diffuse functions).89.p) or (3df. BoysLocal Use Boys localization [419. NRPopLocal Newton-Raphson population localization. REQUIRED OPTION NMin=N Specifies N as the minimum number of pair natural orbitals. .590) grid.This keyword requests a general Complete Basis Set extrapolation of the MP2 energy [87. ADDITIONAL OPTIONS MinPopLocal Use localization based on populations in minimal basis [92]. an alternate grid can be specified with the Int=Grid keyword. NRBoysLocal Newton-Raphson Boys localization.p) polarization functions (again. See the description of the Integral keyword for a full discussion of the available grids. 6-31G†† and 6-311G** basis sets (with or without diffuse functions). The first can be specified with the NMin option. This is the default.420. The method requires two parameters: the minimum number of pair natural orbitals and the integration grid. NRMinPopLocal Use 2nd order minimal population analysis. or an error will result. Single point energy calculations only. The default integration grid is the (99.327]. The integration portion is a small part of the total CBS extrapolation computation. using any electron correlation method. NMin must be specified in all other cases. PopLocal Use population localization as described in reference [418]. NoLocal Do not use any localization.88. so this relatively large grid was chosen. and to 10 for the 6-311G basis set with (2df.421]. and it defaults to 5 for the 6-31G**. Int=Grid.69. T1Diag Computes the T1 diagnostic of T. or both single and double substitutions for CCSD [68. CC and QCID are synonyms for CCD. Analytic energies and gradients for CCD and CCSD. FC All frozen core options are available with CCD and CCSD. MaxCyc=n Specifies the maximum number of cycles for CCSD calculations. Transformation. MP4. J. and numerical frequencies for all methods.70. QCISD . using double substitutions from the Hartree-Fock determinant for CCD [67].71]. Lee and coworkers [423](CCSD only). T Include triple excitations non-iteratively [72] (CCSD only).422] calculations. E4T Used with the T option to request inclusion of triple excitations for both the complete MP4 and to form CCSD(T). numerical gradients for CCSD(T). Conver=N Sets the convergence calculations to 10-N on the energy and 10-(N+2) on the wavefunction. CBS keywords CCD CCSD These method keywords request a coupled cluster [67. CCSD-T is a synonym for CCSD(T). The default is N=7 for single points and N=8 for gradients. CCSD(T)= -.425].54979226D-01 . charge is the charge.y. Only valid for single point calculations. in this format: x y z charge 0.0 A [ρ B] where x.. By default.0 is a fixed field. Bohrs Indicates that input charge locations are specified in Bohrs..75019717665D+02 E(CORR)= -. The charge distribution is made up of point charges [424. the charges are read from the input stream.z are the coordinates in the input orientation (in the units specified by the Units keyword) and defaulting to Angstroms). . one per line.75019641794D+02 The CCSD energy is labeled E(CORR). and the energy including the non-iterative triples contribution is given in the final line. 0. Charge The Charge keyword requests that a background charge distribution be included in the calculation. and the remaining items are parameters in the following equation for the additional electron repulsive term: Angstroms Indicates that input charge locations are specified in Angstroms.The Coupled Cluster energy appears in the output as follows (following the final correlation iteration): DE(CORR)= -. Single point energies. Units To perform geometry optimizations in the presence of background charges.0 1.0 2.0 -2. Here is an example: # RHF/STO-3G Opt=Z-Matrix Charge NoSymm Water. and is useful in compound jobs involving general basis sets by allowing them to have only one . Not valid with semi-empirical methods or PBC.StandardOrientation Indicates that the input charges are specified in the standard orientation rather than the input orientation.1 O H 1 R1 H 1 R2 2 A1 Variables: R1=1. optimizations and frequencies.1 ChkBasis The ChkBasis keyword requests that the basis set be read from the checkpoint file. Use the %KJob=L301 Link 0 command to quickly determine the standard orientation for a molecule. STO-3G.0 2. %KJob.0 1.2 2. Check Reads the background charge distribution from the checkpoint file.0 A1=105.0 2. 2.0 R2=1. point charges 0. you must use Opt=Z-Matrix NoSymm keywords and define the input geometry either in traditional Zmatrix coordinates or symbolic Cartesian coordinates. Cartesian functions. GenECP. no basis set keyword should be specified with ChkBasis. and RdBasis are all synonyms for ChkBasis. ChkBasis will also retrieve any density fitting basis in the checkpoint file. Note.copy of the basis set in the input stream (see the discussion of the Gen keyword below). . ExtraDensityBasis The following route section will retrieve the basis set and density fitting set (if any) from the checkpoint file and use them for the current job: # BLYP/ChkBasis The following route section will retrieve only the basis set from the checkpoint file. that ChkBasis can be used to retrieve whatever basis set exists in a checkpoint file. ECP's specified in the basis set are also retrieved. By default. ExtraBasis. See the examples for other possibilities. Gen. however. Pseudo.143. regardless of how it was originally specified. CheckPointBasis. and an automatically generated density fitting basis will be used: # BLYP/ChkBasis/Auto The following route section will retrieve only the density fitting basis from the checkpoint file: # BLYP/6-31G(d)/ChkBasis CID CISD These method keywords request a Hartree-Fock calculation followed by configuration interaction with all double substitutions (CID) or all single and double substitutions (CISD) from the Hartree-Fock reference determinant [61. CIDS and CI are synonyms for CISD. Of course. as are the choices for pure vs.202]. ReadBasis. MaxCyc=n Specifies the maximum number of cycles for CISD calculations. Norm(A)-1 gives a measure of the correlation correction to the wavefunction. analytic gradients.10129586D+01 E(CI)= -. Norm(A) is the factor by which to divide the wavefunction as given above to fully normalize it.75009023292D+02 The output following the final CI iteration gives the predicted total energy. Conver=N Sets the convergence calculations to 10-N on the energy and 10-(N+2) on the wavefunction. Note that the wavefunction is stored in intermediate normalization. The default is N=7 for single points and N=8 for gradients.FC All frozen core options are available with CID and CISD.48299990D-01 . Thus: . that is: where Ψ0 is the Hartree-Fock determinant and has a coefficient of 1 (which is what intermediate normalization means). the coefficient of the HF configuration is thus 1/Norm(A). The second output line displays the value of Norm(A). Energies. and numerical frequencies. Transformation The CI energy appears in the output as follows: DE(CI)= NORM(A) = -. Triplets Solve only for triplet excited states. . 50-50 Solve for half triplet and half singlet states. Chapter 9 of Exploring Chemistry with Electronic Structure Methods [308] provides a detailed discussion of this method and its uses. for which it is the default. The default is the first excited state (N=1). Note that Density cannot be used with CIS(D).427]. CIS jobs can include the Density keyword. This option only affects calculations on closed-shell systems. STATE SELECTION OPTIONS Singlets Solve only for singlet excited states. The CIS(D) keyword and option is used to request the related CIS(D) method [426. CIS CIS(D) The CIS method keyword requests a calculation on excited states using single-excitation CI (CI-Singles) [108]. This option only affects calculations on closed-shell systems. You can also follow a CIS job with a CIS(D) job to compute the excitation energies for additional states (see the examples). This option only affects calculations on closed-shell systems. this keyword causes the population analysis to use the current (CIS) density rather than its default of the HartreeFock density. the coefficient of singly-excited determinantΨi→a is Tia/Norm(A). without options.The coefficient of the Hartree-Fock determinant in the fully normalized wavefunction is then 1/Norm(A). and so on. Root=N Specifies the "state of interest" for which the generalized density is to be computed. . an initial guess for one basis set cannot be used for a different one. The default is N=4 for single points and N=6 for gradients. Conver=N Sets the convergence calculations to 10-N on the energy and 10-(N+2) on the wavefunction. This is the default algorithm in Gaussian 03. AO Forces solution of the CI-Singles equations using the AO integrals. except for small molecules on systems having very limited disk and memory. CIS=Direct should be used only when the approximately 4O2N2 words of disk required for the default (MO) algorithm are not available. the default is 3 singlets and 3 triplets).NStates=M Solve for M states (the default is 3). NStates gives the number of each type of state for which to solve (i. DENSITY-RELATED OPTION AllTransitionDensities Computes the transition densities between every pair of states.e. Add=N Read converged states off the checkpoint file and solve for an additional N states. Restart Restarts the CI-Singles iterations off the checkpoint file. If 50-50 is requested. PROCEDURE. unlike for SCF. or for larger calculations (over 200 basis functions).AND ALGORITHM-RELATED OPTIONS FC All frozen core options are available with CIS and CIS(D). Direct Forces solution of the CI-Singles equation using AO integrals which are recomputed as needed. Read Reads initial guesses for the CI-Singles states off the checkpoint file. Note that. Also implies SCF=Restart. The AO basis is seldom an optimal choice. The transformation attempts to honor the MaxDisk keyword. NStates cannot be used with this option.. This option implies Read as well. thus further moderating the disk requirements. MO Forces solution of the CI-Singles equations using transformed two-electron integrals. avoiding an integral transformation. IVOGuess. After the first iteration. additional initial guesses are made. MaxDiag=N Limits the submatrix diagonalized in the Davidson procedure to dimension N. There are no special features or pitfalls with CI-Singles input. An SCF is followed by the integral transformation and evaluation of the ground-state MP2 energy. EqSolv Whether to perform equilibrium or non-equilibrium PCM solvation. MaxDisk. NonEqSolv is the default. analytic gradients. maximum delta is 0. is the default. Information about the iterative solution of the CI problem comes next. The change in excitation energy and wavefunction for each state is printed for each iteration (in the #P output): Iteration 3 Dimension 27 Root 1 not converged. Useful when using non-standard routes to do successive CI-Singles calculations. note that at the first iteration. This is mainly a debugging option. which uses improved virtual orbitals. NoIVOGuess Forces the use of canonical single excitations for the guess. Energies.002428737687607 . and analytic frequencies for CIS. and energies for CIS(D). This is mainly a debugging option. Transformation. to ensure that the requested number of excited states are found regardless of symmetry. TD. MaxDavidson is a synonym for this option.RWFRestart Restarts the CI-Singles iterations off the read-write file. Density CIS Output. ZINDO. DEBUGGING OPTIONS ICDiag Forces in-core full diagonalization of the CI-Singles matrix formed in memory from transformed integrals. Output from a single point CI-Singles calculation resembles that of a ground-state CI or QCI run. one new vector is added to the solution for each state on each iteration. BB:AA.Read.0008 8 -> 9 0. => This is the "state of interest.769540624626156 Change is -0. When the CI has converged. the excitation energy. E(Cis) = .NStates=6) The same procedure will work using CIS(D) in the follow-up job.700631883679401 Change is -0.841115226789293 Change is -0. and/or second-order corr.047396173133051 The iterative process can end successfully in two ways: generation of only vanishingly small expansion vectors. the sum of the squares of the expansion coefficients is normalized to total 1/2 (as the beta coefficients are not shown).011232152003400 Root 3 : 8. For open shell calculations. CNDO . maximum delta is 0. maximum delta is 0. Finding Additional States. the oscillator strength. Normalization. Next. For closed shell calculations. the results on each state are summarized. The following route will read the CIS results from the checkpoint file and solve for 6 additional states beyond the second state: # CIS=(D. or negligible change in the updated wavefunction. the normalization sum is 1.113.Root 2 not converged. the results are displayed.” Total Energy.Root=2. and the largest coefficients in the CI expansion (use IOp(9/40=N) to request more coefficients: all that are greater than 10N ): Excitation energies and oscillator strengths: symmetry excitation energy oscillator strength Excited State 1: Singlet-A" 3.03 nm f=0.013107675296678 Root 3 not converged.BB> singles matrix: ***************************************************************** The transition dipole moments between the ground and each excited state are then tabulated. Excitation is from orbital 8 to orbital 9 This state for opt.696894498 CIS energy is repeated here for convenience.030654755631835 Excitation Energies [eV] at current iteration: Root 1 : 3. beginning with this banner: ***************************************************************** Excited States From <AA. including the spin and spatial symmetry.7006 eV 335.69112 CI expansion coefficients for each excitation.001084398684008 Root 2 : 7. Analytic energies for Hartree-Fock and MP2 only. Note that energy differences computed from the values in semi-empirical calculations are in Hartrees and may be compared directly with energy differences computed from jobs using other methods. and z components of the dipole moment): Energy= -19.000000 -. No basis set keyword should be specified. "analytic" gradients. SCF Constants . Complex This keyword allows the molecular orbitals to become complex. y. analytic HF gradients.739540 The energy is as defined by the CNDO model.This method keyword requests a semi-empirical calculation using the CNDO Hamiltonian [41]. and numerical frequencies. It may only be used for closed-shell singlet states. Energies. Dipole moment= . The CNDO energy appears in the output file as follows (followed by the x. and numerical HF frequencies.000000 .887711334547 NIter= 10. The OldConstants keyword is a synonym for Constants=1979.803242 x 10-10 ESU [432] = 1. mostly from [431].022413996 m3 [428] . Raw Constants.602176462 x 10-19 Coulombs [428] Planck's constant (h) = 6.99792458 x 1010 cm-sec-1 [431] Boltzman constant (k) = 1. from [429.15 K = 0. CURRENT VALUES Here are summarized various conversion factors and physical constants used by Gaussian 03 in converting from standard to atomic units. The constants which are stored directly in the program are: 1 Bohr (a0) = 0.Specifies which set of physical constants to use.02214199 x 1023 [428] 1 Calorie = 4.62606876 x 10-34 Joule-secs [428] Avogadro's number (NA) = 6.430].03599976 [428] Molar volume of ideal gas at 273.66053873 x 10-27 kilograms [428] 1 Electron charge (e) = 4. All quantities used in calculations inside Gaussian are in atomic units. the conversion factors are used only when processing input or generating printed output.184 Joules [431] 1 Hartree = 4. This is the default 1986 Constants used in Gaussian 88 through Gaussian 98.5291772083 Å [428] 1 Atomic mass unit (amu) = 1. Note that using an older set should only be done in order to compare results with earlier versions of Gaussian. 1998 Constants from [428] and references therein.380603 x 10-23 Joules-degree-1 [428] Inverse fine structure constant (α-1) = 137.3597482 x 10-18 Joules [429] Speed of light (c) = 2. 1979 Constants used in Gaussian 80 through Gaussian 86. 82573 x 10-3 cm-2-atm-1 at STP 1 Hartree-1/2-Bohr-1-amu-1/2 = 219474.434] may be computed using the Counterpoise keyword. which can be used on an energy calculation. The facility also requires an additional integer to be placed at the end of each atom specification indicating which fragment/monomer it is part of.Proton rest mass = 1.002319304386 (dimensionless) [428] Conversion Factors.2561 km-mol-1 = 5.2114 eV 1 Bohr-electron = 2.541746 x 10-18 esu-cm 1 Debye2-angstrom-2-amu-1 = 42.8880 electron mass 1 Electron volt (eV) = 23.06055 kcal-mol-1 1 Hartree = 627.1527 electron mass 1 Atomic mass unit (amu) = 1822. optimization or frequency calculation or BOMD.478352 x 10-30 C-m Counterpoise Counterpoise corrections [433.648777 x 10-41 C2-m2-J-1 Dipole moment = 1 Bohr-electron = 8.5095 kcal-mol-1 = 27.541746 Debye = 2.910938 x 10-30 kg Proton mass = 1836.28476362 x 10-24 J-T-1 [428] Free electron g-factor = 2.142206 x 1011 V-m-1 Electric polarizability: 1 au = 1. The Counterpoise keyword takes an integer value specifying the number of fragments or monomers in the molecular structure. The following useful conversion factors can be derived from the above: Electron mass = 0. .6 cm-1 Electric field: 1 au = 5.67262158 x 10-27 kg [428] Electron magnetic moment = 9. 92 2. frag.1. NewBq is a synonym for NewGhost. the first atom in the Z-matrix must be given in Cartesian coordinates.RO1H. The format of the corresponding input line in this case is: total-charge.RO2H2.180. 2 multiplicity An example counterpoise optimization using ECPs: # hf/lanl2dz counterpoise=2 nosymm opt test HBr + HF...90.2 X.2.1 H.2 Z-matrix variables. This is the default and the recommended method.. OldGhost Requests old-style ghost atoms..1.3. 1-charge.00 0. Also.X3O H.00 0.1.2.00 1 2 2 1 Note that the Z-matrix input requires a 0 after the dihedral angle value/variable (to indicate that the final angle is a dihedral) prior to the fragment number.1.73 3.3.2 H.0..52.3.77 9 0.HOX3. using Cartesian coordinates for such jobs makes specifying fragment numbers in the input much more straightforward.0. This option is only useful for comparison with previous results. 2 charge.0.0.00 0..HOX3. frag.00 0.1.H8OX.2 O.RO2H1.1 X.00 9 0. Counterpoise Input.H7OX.0.43 0.0.1 multiplicity.ROO. total-spin.RO1H.6. # MP2/6-31G Counterpoise=2 Opt Counterpoise with Cartesian 0.180. The preceding Z-matrix also illustrates the use of fragment-specific charge and spin multiplicity specifications.2.-90.1 structures begin here O.1. frag.0..1.2.0. frag. optimization with counterpoise correction using ECP basis 0 1 H -0.00 0.1. OldBq is a synonym for OldGhost. Here are examples using a Z-matrix (left) and Cartesian coordinates (right): # MP2/6-31G Counterpoise=2 Opt Counterpoise with Z-matrix 0. 0..1..17 1 0.2. Clearly.5.1.NewGhost Requests new-style ghost atoms for which integration grid points for DFT quadrature are included.046866 0.0.2.0.1 1 0.0.586860 1 .0 H.1.0.6.3. UltraFine. This option is relevant for Freq and Polar jobs.584835 0. Grid=grid Specify the integration grid for the CPHF portion of the calculation. The default is a static frequency calculation. This option causes the desired frequency to be read from the input stream. -0.443. Here is some sample output from a Counterpoise calculation: Counterpoise: corrected energy = -2660.801000 2. Other units may be specified by including a suffix. Otherwise.396755 0.440. See the discussion of Integral=Grid for full details on grid specification. then that grid is also used for the CPHF. This is the default for static perturbations.331864 0. then SG1 is used for the CPHF. respectively.438. when the latter uses the SG1 or Fine grid.083831739527 Counterpoise: BSSE energy = 0. 0. The default grid used depends on the one used for integral evaluation.739275 3. and so on) or a specific grid.Br F H -0. reading in the frequencies for the electromagnetic field perturbation. . RdFreq Perform frequency-dependent CPHF. The syntax is the same as for the Int=Grid option.442.439. NonEqSolv is the default for dynamic (non-zero frequency) perturbations.641534 1 2 2 Counterpoise Output. CPHF This keyword selects the algorithm used for solving the CPHF equations [435. If any specific grid is specified to the Integral keyword.436. EqSolv Use equilibrium solvation. 0.441.110)).003902746890 These lines give the corrected energy and basis set superposition errors. and when UltraFine is used for the integrals. SG1 is the default grid for Polar=OptRot and Freq=Anharmonic.444]. The argument to this option may be a grid keyword (Fine. the Coarse grid is used for the CPHF (a pruned (35. The default units for this value are Hartrees. one of cm (cm-1) and nm (wave numbers).437. 443]. The default is N=9 for CPHF=Separate and N=10 for CPHF=Simultaneous (the default). Canonical Canonical CPHF. but is slightly less accurate. This is the default.445. MOD Use MOD orbital derivatives for SAC-CI gradients (which uses configuration selection). The default is as large a space as memory permits. which does them separately. AO Solve CPHF in the atomic orbital basis [436. but to treat them together if there are only electromagnetic perturbations. The NoZVector keyword says to use the full 3 x NAtoms CPHF for post-SCF gradients. MO Solve in the molecular orbital basis.439. This is the default. This is faster than using separate spaces. the default. The default is to treat them separately if nuclear perturbations are also being done.442. Separate Use a separate expansion space for each variable in the CPHF (the opposite of Simultaneous). MaxInv=N Specifies the largest reduced space for in-core inversion during simultaneous solution (up to dimension N). ZVector Use the Z-Vector method [140. Allowed and the default if Hartree-Fock 2nd derivatives are not also requested. Larger reduced problems are solved by a second level of DIIS. SCF Density .Simultaneous Use one expansion space for all variables.446] for post-SCF gradients. XY Treat real and imaginary perturbations together. Conver=N Set the CPHF convergence criterion to 10-N. The opposite is NoXY. .M) Use the CIS transition density between state M and state N. CI Use the generalized density corresponding to the CI energy. MP4(SDQ). AllTransition Use all available CIS transition densities. and the CASSCF density for CAS jobs).447]. QCISD. SCF Use the SCF density. This is the default when no option is given to Density. only the density for the state of interest (see the examples below for a method of doing the former). Note that this option does not produce densities for all of the excited states in a CI-Singles calculation.e. Transition=N or (N. population and other analysis procedures use the SCF density (i. CCSD. QCI Use the generalized density corresponding to the QCI (or coupled cluster) energy. These are based on the Z-Vector [140. HF is a synonym for SCF.446. and hence yield multipole moments which are the correct analytical derivatives of the energy. The Density keyword without an option is equivalent to Density=Current. CC is a synonym for QCI.By default. CID and CISD and SAC-CI methods are available. Current Use the density matrix for the current method.445. the Hartree-Fock density for post-SCF methods. the DFT density for DFT jobs. MP3. M defaults to 0. CCD. The unrelaxed densities at second order (not the same as MP2) can also be used but are not recommended. The options of the Density keyword select which density to analyze. MP2 Use the generalized density corresponding to the second-order energy.. The generalized densities for the MP2. which corresponds to the ground state. All Use all available densities. This is allowed for population analysis but not for electrostatics or density evaluation. This is not the same as the MP2 density.Root=N) Density=Current Pop=CHelpG # Guess=Read Geom=AllCheck This route picks up the converged CIS and CIS wavefunction from the checkpoint file.. Guess. ChkBasis The following route section specifies a CI-Singles calculation which predicts the first six excited states of the molecule under investigation. Note that this is not the same as the density resulting from CIS(Root=N. and performs the necessary CPHF calculation to produce the relaxed density for state N. and its use is discouraged! [447] CIS=N Use the total unrelaxed CIS density for state N. DensityFit . Checkpoint Recover the density from the checkpoint file for analysis. SCF.RhoCI Use the one-particle density computed using the CI wavefunction for state N. Rho2 Use the density correct to second-order in Møller-Plesset theory.. and retrieves the basis set from the checkpoint file. which is to be preferred [447]. Implies Guess=Only ChkBasis: the calculation does not recompute new integrals. and its use is discouraged! Chapter 9 of Exploring Chemistry with Electronic Structure Methods discusses this issue [308].. which is then used in the population and other analyses.) Density=Current. This is not the same as the CI density [447]. The population and other analyses will use the CIS density corresponding to the lowest excited state: %Chk=benzene # CIS(NStates=6)/6-31+G(d. and so on.p) Density=Current Pop=CHelpG The following route section may be used to rerun the post-CIS analyses for the other excited states: %Chk=benzene # CIS(Read. 457.455. NonIterative is the default except for ADMP.449] models (see also [448. InvToler=N Set the tolerance for a non-trivial eigenvalue of the generalized inverse of the fitting matrix to 10-N.452. Convergence=N Specifies 10-N as the convergence criterion for iterative solution of the fitting equations.199] are available for all DFT models. Energies [78]. The same optimum memory sizes given by freqmem are recommended for the more general models.Controls density fitting for the Coulomb problem.76. Applies only to DFT calculations using pure (non-hybrid) functionals.454. DenFit is a synonym for this keyword. . The default is 10-6 for ADMP and 10–9 for the BOMD.461] for discussions of DFT methods and applications). ExtraDensityBasis. optimizations.456. and frequency calculations to model systems in solution. Gen.460. discussed here. Density fitting basis sets are specified as part of the model chemistry within the job's route section. The self-consistent reaction field (SCRF) can be used with DFT energies. Implies Iterative. See the discussion here for details.198.458. Pure DFT calculations will often want to take advantage of density fitting.448. analytic gradients.459. and true analytic frequencies [197.453. ChkBasis Density Functional (DFT) Methods Gaussian 03 offers a wide variety of Density Functional Theory (DFT) [75.450.451. Iterative Controls whether a generalized inverse is formed or the fitting equations are solved iteratively. the specific functionals available in Gaussian 03 are given. In density functional theory. and EC[P] is the correlation functional. Note: Polarizability derivatives (Raman intensities) and hyperpolarizabilities are not computed by default during DFT frequency calculations. -1/2<PK(P)> The exchange energy resulting from the quantum (fermion) nature of electrons. with EX[P] given by the exchange integral -1/2<PK(P)> and EC=0.∇ρβ(r))dr where the methods differ in which function f is used for EX and which (if any) f is used for EC.The next subsection presents a very brief overview of the DFT approach.∇ρα(r). Hartree-Fock theory is really a special case of density functional theory. Gaussian supports hybrid methods in which the exchange functional is a linear combination of the Hartree-Fock exchange and a .1/2<PK(P)> where the terms have the following meanings: V The nuclear repulsion energy. the exchange-correlation functional. In addition to pure DFT methods. Use Freq=Raman to request them.ρβ(r). the exact exchange (HF) for a single determinant is replaced by a more general expression. P The density matrix. <hP> The one-electron (kinetic plus potential) energy 1/2<PJ(P)> The classical coulomb repulsion of the electrons. The functionals normally used in density functional theory are integrals of some function of the density and possibly the density gradient: EX[P] = ∫f(ρα(r). Following this. which can include terms accounting for both exchange energy and the electron correlation which is omitted from Hartree-Fock theory: EKS = V + <hP> + 1/2<PJ(P)> + EX[P] + EC[P] where EX[P] is the exchange functional. BACKGROUND In Hartree-Fock theory. the energy has the form: EHF = V + <hP> + 1/2<PJ(P)> . The final subsection surveys considerations related to accuracy in DFT calculations. Comb. Gill 96: The 1996 exchange functional of Gill [469.77].77]. Keyword: Used Alone: N/A.Form: B. Correlation Functionals. usually used when this exchange functional is used without a correlation functional [75. Keyword: Used Alone: HFB. often referred to as Local Spin Density (LSD) correlation [475] (functional III in the paper). listed by their corresponding keyword component: • • VWN: Vosko. Form: S Xαρ4/3 with the empirical coefficient of 0. Comb.76. Comb. In some cases. Comb. Comb. Form: MPW. VWN V(VWN5): Functional V from the 1980 paper which fits the Ceperly-Alder solution to the uniform electron gas (this is the functional recommended in the paper) [475]. Form: PW91.7. Form: G96. OPTX: Handy's OPTX modification of Becke's exchange functional [474].467]. PBE: The 1996 functional of Perdew. Becke 88: Becke's 1988 functional. The following exchange functionals are available in Gaussian 03: • • • • • • • • Slater: ρ4/3 with theoretical coefficient of 2/3. also referred to as Local Spin Density exchange [75. KEYWORDS FOR DFT METHODS Names for the various pure DFT models are given by combining the names for the exchange and correlation functionals.465. Barone's Modified PW91: The Perdew-Wang 1991 exchange functional as modified by Adamo and Barone [468]. Form: XA. Keyword: Used Alone: N/A. and Nusair 1980 correlation functional(III) fitting the RPA solution to the uniform electron gas.470]. Proposed functionals lead to integrals which cannot be evaluated in closed form and are solved by numerical quadrature. Exchange Functionals. which includes the Slater exchange along with corrections involving the gradient of the density [462]. The following correlation functionals are available. Wilk. Comb.functional integral of the above form. Comb. Keyword: Used Alone: N/A. Comb. Perdew-Wang 91: The exchange component of Perdew and Wang's 1991 functional [463. Keyword: Used Alone: N/A. Keyword: Used Alone: N/A. Keyword: Used Alone: XAlpha.472].464. Form: O. Form: PBE. standard synonyms used in the field are also available as keywords. The combination forms are used when one of these exchange functionals is used in combination with a correlation functional (see below).466. . Keyword: Used Alone: HFS. Burke and Ernzerhof [471.76. Some other software packages with DFT facilities use the equivalent of SVWN5 when "LSDA" is requested. HCTH147 to HCTH/147. PL (Perdew Local): The local (non-gradient corrected) functional of Perdew (1981) [478]. BLYP requests the Becke exchange functional and the LYP correlation functional. Standalone Functionals. along with his 1981 local correlation functional [479].467].464. which include a mixture of Hartree-Fock exchange with DFT exchange-correlation. Three hybrid functionals. and Parr which includes both local and non-local terms [476.484]. HCTH93 to HCTH/93. Check the documentation carefully for all packages when making comparisons. PW91 (Perdew/Wang 91): Perdew and Wang's 1991 gradient-corrected correlation functional [463. are available via keywords: • • Becke Three Parameter Hybrid Functionals. HCTH/*: Handy's family functional including gradient-corrected correlation [482. LSDA is a synonym for SVWN.465. For example. P86 (Perdew 86): The gradient corrections of Perdew. The following functionals are self-contained and are not combined with any other functional keyword components: • • VSXC: van Voorhis and Scuseria's τ-dependant gradient-corrected correlation functional [481].466. Hybrid Functionals. PBE: The 1996 gradient-corrected correlation functional of Perdew. Yang. Correlation Functional Variations. All of the keywords for these correlation functionals must be combined with the keyword for the desired exchange functional.472]. and is known in the literature by its synonym LSDA (Local Spin Density Approximation). HCTH refers to HCTH/407. The following correlation functionals combine local and non-local terms from different correlation functionals: • • VP86: VWN5 local and P86 non-local correlation functional.• • • • • • LYP: The correlation functional of Lee. and HCTH407 to HCTH/407. Note that the related HCTH/120 functional is not implemented. B95 (Becke 95): Becke's τ-dependent gradient-corrected correlation functional (defined as part of his one parameter hybrid functional [480].483. These functionals have the form devised by Becke in 1993 [79]: A*EXSlater+(1-A)*EXHF+B*ΔEXBecke+ECVWN+C*ΔECnon-local . Burke and Ernzerhof [471. V5LYP: VWN5 local and LYP non-local correlation functional.477]. SVWN requests the Slater exchange and the VWN correlation functional. depending on whether an exchange functional is desired or not. Bradley and Tozer's modification to B97: B972 [488]. and any scaling is accomplished using P3 and P4. 1372 (1993)). Becke's 1998 revisions to B97 [486. The 1997 hybrid functional of Perdew. MPW1PW91. In one variation. as implemented by Adamo and Barone [480. Burke and Ernzerhof [472].5*EXLSDA + 0. B3P86 specifies the same functional with the non-local correlation provided by Perdew 86. P1 is usually set to either 0.5*ΔEXBecke88 + ECLYP Note that these are not the same as the "half-and-half" functionals proposed by Becke (J. The program also provides other. The B1B95 keyword is used to specify Becke's one-parameter hybrid functional as defined in the original paper [480]. and VWN functional III for local correlation (not functional V). Phys. uses modified Perdew-Wang exchange and Perdew-Wang 91 correlation [468 ]. Becke One Parameter Hybrid Functionals. Chem. similar one parameter hybrid functionals. Half-and-half Functionals. which should be used when only local exchange is desired. and it implements equation 2c in reference [487]. 98.0. These functionals are included for backward-compatibility only.0 or 1. Gaussian 03 can use any model of the general form: P2EXHF + P1(P4EXSlater + P3ΔExnon-local) + P6EClocal + P5ΔECnon-local The only available local exchange method is Slater (S). the correlation functional used is actually: C*ECLYP+(1-C)*ECVWN In other words. You specify the values of the six parameters with various non-standard options to the program: • IOp(3/76=mmmmmnnnnn) sets P1 to mmmmm/10000 and P2 to nnnnn/10000. B3LYP uses the non-local correlation provided by the LYP expression.• • • • • • • • • • • • • • • where A. Another version.5*EXHF + 0. Wilson. User-Defined Models. B. Any combinable non-local exchange functional and combinable correlation functional may be used (as listed previously). Handy. the LYP correlation functional is used (as described for B3LYP above). The keyword is PBE1PBE. VWN is used to provide the excess local correlation required. B1LYP. This functional uses 25% exchange and 75% correlation weighting. and C are the constants determined by Becke via fitting to the G1 molecule set.5*EXLSDA + ECLYP BHandHLYP: 0. The keyword is B98. Tozer and coworkers modification to B97: B971 [482]. There are several variations of this hybrid functional. .487].5*EXHF + 0. Note that since LYP includes both local and nonlocal terms. which implement the following functionals: BHandH: 0. since LYP contains a local term essentially equivalent to VWN.485]. and B3PW91 specifies this functional with the non-local correlation provided by Perdew/Wang 91. We do not recommend using any smaller grid in production DFT calculations. Note also that it is important to use the same grid for all calculations where you intend to compare energies (e. and so on). The output from a BLYP calculation is labeled similarly: . Thus in addition to the sources of numerical error in Hartree-Fock calculations (integral accuracy. Note that all values must be expressed using five digits. after 5 cycles The item in parentheses following the E denotes the method used to obtain the energy. IOp(3/76=1000005000) sets P1 to 1.• • IOp(3/77=mmmmmnnnnn) sets P3 to mmmmm/10000 and P4 to nnnnn/10000. DenFit The energy is reported in DFT calculations in a form similar to that of Hartree-Fock calculations. heats of formation.U. ADMP calculations. analytic gradients. Stable. Larger grids are available when needed (e. IOp. the accuracy of DFT calculations also depends on number of points used in the numerical integration. Here is the energy output from a B3LYP calculation: SCF Done: E(RB+HF-LYP) = -75.g. Here is a route section specifying the functional corresponding to the B3LYP keyword: # BLYP IOp(3/76=1000002000) IOp(3/77=0720008000) IOp(3/78=0810010000) ACCURACY CONSIDERATIONS A DFT calculation adds an additional step to each major phase of a Hartree-Fock calculation. tight optimization of certain kinds of systems). Int=Grid. This grid greatly enhances calculation accuracy at minimal additional cost.. adding any necessary leading zeros. For example. An alternate grid may be selected by including Integral=(Grid=N) in the route section (see the discussion of the Integral keyword for details). The "fine" integration grid (corresponding to Integral=FineGrid) is the default in Gaussian 03. and analytic frequencies.g. computing energy differences. IOp(3/78=mmmmmnnnnn) sets P5 to mmmmm/10000 and P6 to nnnnn/10000. This step is a numerical integration of the functional (or various derivatives of the functional).0 and P2 to 0. CPHF convergence). TD.5.3197099428 A. SCF convergence. Energies. but they are also available as independent methods. They were implemented for use in ONIOM calculations.scripps. all atoms which were untyped or which were given a type but not a charge in the input).dat) have been updated slightly since the publication of this paper.e. no charges are assigned to atoms by default when using any molecular mechanics force field.. PARAMETER PRECEDENCE OPTIONS Terminology: Gaussian contains built-in parameter sets for the built-in force fields listed above. We use this current version from the AMBER web site (amber.edu).2867073414 A. UnCharged Assign QEq charges for all atoms which have charge zero (i. after 5 cycles Molecular Mechanics Methods There are three molecular mechanics methods available in Gaussian. UnTyped Assign QEq charges only to those atoms for which the user did not specify a particular type in the input. DREIDING: The DREIDING force field as described in [38]. CHARGE ASSIGNMENT-RELATED OPTIONS Unless set in the molecule specification input.U. The actual parameters (parm96.SCF Done: E(RB-LYP) = -75. Options are available to estimate charges at the initial point using the QEq algorithm under control of the following options for any of the mechanics keywords: QEq Assign charges to all atoms using the QEq method [40]. these are referred to as hard-wired parameters. No basis set keyword should be specified with these keywords. Soft parameters are ones specified by the user in the input stream for the current job (or a previous job when reading . The following force fields are available: AMBER: The AMBER force field as described in [37]. UFF: The UFF force field as described in [39]. parameters from the checkpoint file). By default, when no relevant option is given, the hard-wired parameters are the only ones used. HardFirst Read additional parameters from the input stream, with hard-wired parameters having priority over the read-in, soft ones. Hence, read-in parameters are used only if there is no corresponding hard-wired value. Note that wildcards matches within the hardwared parameter set take precidence over soft parameters, even when the latter contains an exact match for the same item. Use SoftFirst if you want to override hard-wired parameter matches. SoftFirst Read additional parameters from the input stream, with soft (read-in) parameters having priority over the hard-wired values. SoftOnly Read parameters from the input stream and use only them, ignoring hard-wired parameters. ChkParameters Read parameters from the checkpoint file. Any non-standard (soft) parameters present in the checkpoint file are used with higher priority than corresponding hard-wired parameters, unless HardFirst is also specified. NewParameters Ignore any parameters in the checkpoint file. Modify Read modifications and additions to the parameter set (after it has been constructed from hard and/or soft parameters). HANDLING MULTIPLE PARAMETER SPECIFICATION MATCHES Since parameters can be specified using wildcards, it is possible for more than one parameter specification to match a given structure. The default is to abort if there are any ambiguities in the force field. The following options specify other ways of dealing with multiple matches. FirstEquiv If there are equivalent matches for a required parameter, use the first one found. LastEquiv If there are equivalent matches for a required parameter, use the last one found. INPUT CONVENTIONS AMBER calculations require that all atom types be explicitly specified using the usual notation within the normal molecule specification section: C-CT Specifies an SP3 aliphatic carbon atom. C-CT-0.32 Specifies an SP3 aliphatic carbon atom with a partial charge of 0.32. O-O--0.5 Specifies a carbonyl group oxygen atom with a partial charge of -0.5. Consult the AMBER paper [37] for definitions of atom types and their associated keywords. Atom types and charges may also be provided for UFF and DREIDING calculations, but they are not required. For these methods, the program will attempt to determine atom types automatically. Analytic energies, gradients, and frequencies. ONIOM, Geom=Connect GENERAL MOLECULAR MECHANICS FORCE FIELD SPECIFICATIONS Unless otherwise indicated, distances are in Angstroms, angles are in degrees, energies are in Kcal/mol and charges are in atomic units. Function equivalencies to those found in standard force fields are indicated in parentheses. In equations, R refers to distances and θ refers to angles. Wildcards may be used in any function definition. They are indicated by a 0 or an asterisk. In MM force fields, the non-bonded (Vanderwaals and electrostatic) interactions are evaluated for every possible pair of atoms. However, interactions between pairs of atoms that are separated by three bonds or less are usually scaled down (in most force fields, using a factor 0.0 for pairs separated by one or two bonds, and some value between 0.0 and 1.0 for pairs that are separated by three bonds). There are a number of ways to implement the calculation of non-bonded interactions. We follow a two-step procedure. First, we calculate the interactions between all pairs, without taking the scaling into account. In this step, we can use computationally efficient (linear scaling) algorithms. In the second step, we subtract out the contributions that should have been scaled, but were included in the first step. Since this involves only pairs that are close to each other based on the connectivity, the computer time for this step scales again linearly with the size of the system. Although at first sight it seems that too much work is done, the overall algorithm is the more efficient than the alternatives. In the soft force field input, the NBDir function entry corresponds to the calculation of all the pairs, and the NBTerm entry is used for the subsequent subtraction of the individual pairs. However, to make things easier, you can specify just the non-bonded master function NonBon, which is automatically expanded into the actual functions NBDir and NBTerm during pre-processing. Vanderwaals parameters, used for NBDir and NBTerm (See MMFF94 below for MMFF94-type Vanderwaals parameters). VDW Bond-length Well-depth MMFF94 type Vanderwaals parameters (used for NBDir and NBTerm). VDW94 Atomic-pol NE Scale1 Scale2 DFlag Atomic-pol Atomic polarizability (Angstrom3). NE Slater-Kirkwood effective number of valence electrons (dimensionless). Scale1 Scale factor (Angstrom1/4). Scale2 Scale factor (dimensionless). DFlag 1.0 for donor type atom, 2.0 for acceptor type, otherwise 0.0. MMFF94 electrostatic buffering Buf94 Atom-type Value Non-bonded interaction master function. This function will be expanded into pairs and a direct function (NBDir and NBTerm) before evaluation of the MM energy. NonBon V-Type C-Type, V-Cutoff C-Cutoff VScale1 VScale2 VScale3 CScale1 CScale2 CScale3 V-Type is the Vanderwaals type: 0 No Vanderwaals 1 Arithmetic (as for Dreiding) 2 Geometric (as for UFF) 3 Arithmetic (as for Amber) 4 MMFF94-type Vanderwaals C-Type is the Coulomb type: 0 No Coulomb 1 1/R 2 1/R2 3 1/R buffered (MMFF94) V-Cutoff and C-Cutoff are the Vanderwaals and Coulomb cutoffs (respectively): 0 No cutoff >0 Hard cutoff <0 Soft cutoff VScale1-3 are Vanderwaals scale factors for 1 to 3 bond separated pairs. CScale1-3 are Coulomb scale factors for 1 to 3 bond separated pairs. If any scale factor < 0.0, the 1/1.2 scaling is used (as for Amber). Coulomb and Vanderwaals direct (evaluated for all atom pairs). NBDir V-Type C-Type V-Cutoff C-Cutoff V-Type, C-Type, V-Cutoff, and C-Cutoff as above. Coulomb and Vanderwaals single term cutoffs NBTerm Atom-type1 Atom-type2 V-Type C-Type V-Cutoff C-Cutoff V-Scale C-Scale V-Type, C-Type, V-Cutoff, C-Cutoff, V-Scale, and C-Scale as above. Atomic single bond radius AtRad Atom-type Radius Effective charge (UFF) EffChg Charge GMP Electronegativity (UFF) EleNeg Value Step down table Table Original-atom-type Stepping-down-type(s). Harmonic stretch I (Amber [1]): ForceC*(R-Req)2 HrmStr1 Atom-type1 Atom-type2 ForceC Req ForceC Force constant Req Equilibrium bond length Harmonic stretch II (Dreiding [4a]): ForceC*[R-(Ri+Rj-Delta)]2 HrmStr2 Atom-type1 Atom-type2 ForceC Delta ForceC Force constant Delta Delta Ri and Rj are atomic bond radii specified with AtRad. Harmonic stretch III (UFF [1a]): k*(R-Rij)2 Equilibrium bond length Rij = (1 - PropC*lnBO)*(Ri + Rj) + Ren Force constant: k = 664.12*Zi*Zj/(Rij3) Electronegativity correction: Ri*Rj*[Sqrt(Xi) - Sqrt(Xj)]2/(Xi*Ri + Xj*Rj) HrmStr3 Atom-type1 Atom-type2 BO PropC BO Bond order (if <0, it is determined on-the-fly) PropC Proportionality constant Ri and Rj are atomic bond radii defined with AtRad. Xi and Xj are GMP electronegativity values defined with EleNeg. Zi and Zj are the effective atomic charges defined with EffChg. Morse stretch I (Amber): DLim*(e-a(R-Req)-1)2 where a = Sqrt(ForceC/DLim) MrsStr1 Atom-type1 Atom-type2 ForceC Req DLim ForceC Force constant Req Equilibrium bond length DLim Dissociation limit Morse stretch II (Dreiding [5a]): DLim*exp[-a(Ri+Rj-Delta)]-1)2 where a = Sqrt(ForceC/DLim) MrsStr2 Atom-type1 Atom-type2 ForceC Delta DLim ForceC Force constant Delta Delta DLim Dissociation limit Ri and Rj are atomic bond radii defined with AtRad. Morse stretch III (UFF [1b]): A1*A3*(exp[-a(R-Rij)]-1)2 where a = Sqrt(k/[BO*PropC]) Equilibrium bond length Rij = (1 - PropC*lnBO)*(Ri + Rj) + Ren Force constant k = 664.12*Zi*Zj/Rij3 Electronegativity correction: Ren = Ri*Rj*(Sqrt(Xi) - Sqrt(Xj))2/(Xi*Ri + Xj*Rj) MrsStr3 Atom-type1 Atom-type2 BO PropC BO Bond order (if <0, it is determined on-the-fly) PropC Proportionality constant Ri and Rj are atomic bond radii defined with AtRad. Xi and Xj are GMP electronegativity values defined with EleNeg. Zi and Zj are the effective atomic charges defined with EffChg. Quartic stretch I (MMFF94 [2]): (Req/2)*(R-ForceC)2*[1+CStr*(R-ForceC+(7/12)*CStr2*(R-ForceC)2] QStr1 Atom-type1 Atom-type2 ForceC Req CStr ForceC Force constant (md-Angstrom-1) Req Equilibrium bond length (Angstrom) CStr Cubic stretch constant (Angstrom-1) Atomic torsional barrier for the oxygen column (UFF [16]) UFFVOx Barrier Atomic sp3 torsional barrier (UFF [16]) UFFVsp3 Barrier Atomic sp2 torsional barrier (UFF [17]) UFFVsp2 Barrier Harmonic bend (Amber [1]): ForceC*(T-θeq)2 HrmBnd1 Atom-type1 Atom-type2 Atom-type3 ForceC θeq ForceC Force constant (in kcal/(mol*rad2) θeq Equilibrium angle Harmonic Bend (Dreiding [10a]): [ForceC/sin(θeq2)]*(cos(θ)-cos(θeq))2 HrmBnd2 Atom-type1 Atom-type2 Atom-type3 ForceC θeq ForceC Force constant θeq Equilibrium angle Dreiding Linear Bend (Dreiding [10c]): AForceC*(1+cos(θ)) LinBnd1 Atom-type1 Atom-type2 Atom-type3 ForceC ForceC Force constant UFF 3-term bend (UFF [11]): k*(C0 + C1*cos(θ))+C2*cos(2θ) where C2=1/(4 * sin(θeq2)), C1 = -4*C2*cos(θeq) and C0=C2*(2*cos(θeq2)+1) Force constant: k = 664.12*Zi*Zk*(3*Rij*Rjk*(1-cos(θeq2))-cos(θeq)*Rik2)/Rik5 UFFBnd3 Atom-type1 Atom-type2 Atom-type3 θeq BO12 BO23 PropC θeq Equilibrium angle BO12 Bond order for Atom-type1–Atom-type2 (when <0, it is determined on-the-fly) BO23 Bond order for Atom-type2–Atom-type3 (when <0, it is determined on-the-fly) PropC Proportionality constant Ri, Rj and Rk are atomic bond radii defined with AtRad. Xi, Xj and Xk are GMP electronegativity defined with EleNeg. Zi, Zj and Zk are effective atomic charges defined with EffChg. UFF 2-term bend (UFF [10]): [k/(Per2)]*[1-cos(Per*θ)] Force constant: k = 664.12*Zi*Zk*(3*Rij*Rjk*(1-cos(Per2))-cos(Per)*Rik2)/Rik5 UFFBnd2 Atom-type1 Atom-type2 Atom-type3 Per BO12 BO23 PropC Per Periodicity: 2 for linear, 3 for trigonal, 4 for square-planar. BO12 Bond order for Atom-type1–Atom-type2 (when <0, it is determined on-the-fly) BO23 Bond order for Atom-type2–Atom-type3 (when <0, it is determined on-the-fly) PropC Proportionality constant Ri, Rj and Rk are atomic bond radii defined with AtRad. Xi, Xj and Xk are GMP electronegativity defined with EleNeg. Zi, Zj and Zk are effective atomic charges defined with EffChg. Zero bend term: used in rare cases where a bend is zero. This term is needed for the program not to protest about undefined angles. ZeroBnd Atom-type1 Atom-type2 Atom-type3 Cubic bend I (MMFF94 [3]): (ForceC/2)*(1+CBend*(θ-θeq))*(θ-θeq)2 CubBnd1 Atom-type1 Atom-type2 Atom-type3 ForceC θeq CBend ForceC Force constant (in md*Angstrom/rad2) θeq Equilibrium angle CBend "Cubic Bend" constant (in deg-1) MMFF94 Linear Bend (MMFF94 [4]): ForceC*(1+cos(θ)) LinBnd2 Atom-type1 Atom-type2 Atom-type3 ForceC ForceC Force constant (md) Amber torsion (Amber [1]): Σi=1,4 (Magi*[1+cos(i*θ-I(i+4))])/NPaths AmbTrs Atom-type1 A-type2 A-type3 A-type4 PO1 PO2 PO3 PO4 Mag1 Mag2 Mag3 Mag4 NPaths PO1-PO4 Phase offsets Mag1...Mag4 V/2 magnitudes NPaths Number of paths (if < 0, determined on-the-fly). Dreiding torsion (Dreiding [13]): V*[1-cos(Period*(θ-PO))]/(2*NPaths) DreiTrs Atom-type1 Atom-type2 Atom-type3 Atom-type4 V PO Period NPaths V Barrier height V PO Phase offset Period Periodicity NPaths Number of paths (if < 0, determined on-the-fly). UFF torsion with constant barrier height (UFF [15]): [V/2]*[1cos(Period*PO)*cos(V*θ)]/NPaths UFFTorC Atom-type1 Atom-type2 Atom-type3 Atom-type4 Period PO V NPaths Period Periodicity PO Phase offset V Barrier height V NPaths Number of paths. When zero or less, determined on-the-fly. UFF torsion with bond order based barrier height (UFF [17]): [V/2]*[1-cos(Period*PO)* cos(Period*θ)]/NPaths where V = 5*Sqrt(Uj*Uk)*[1+4.18*Log(BO12)] UFFTorB Atom-type1 Atom-type2 Atom-type3 Atom-type4 Period PO BO12 NPaths Period Periodicity PO Phase offset BO12 Bond order for Atom-type1–Atom-type2 (when <0, it is determined on-the-fly) NPaths Number of paths (when <0, it is determined on-the-fly) Uj and Uk are atomic constants defined with UFFVsp2. UFF torsion with atom type-based barrier height (UFF [16]): [V/2]*[1-cos(Period*PO)* cos(Period*θ)]/NPaths where V=Sqrt(Vj*Vk) UFFTor1 Atom-type1 Atom-type2 Atom-type3 Atom-type4 Period PO NPaths Period Periodicity PO Phase offset NPaths Number of paths. When zero or less, determined on-the-fly. Vj and Vk are atomic constants defined with UFFVsp3. UFF torsion with atom type based barrier height (UFF [16]) (differs from UFFTor1 in that the atomic parameter that is used): [V/2]*[1cos(Period*PO)*cos(Period*θ)]/NPAths where V=Sqrt(Vj*Vk) UFFTor2 Atom-type1 Atom-type2 Atom-type3 Atom-type4 Period PO NPaths Period Periodicity PO Phase offset NPaths Number of paths. When zero or less, determined on-the-fly. Vj and Vk are atomic constants from UFFVOx. Dreiding special torsion for compatibility with Gaussian 98 code. During processing, it is replaced with DreiTRS, with the following parameters: • • • If there are three atoms bonded to the third center and the fourth center is H, it is removed. If there are three atoms bonded to the third center, and at least one of them is H, but the fourth center is not H, then these values are used: V=4.0, PO=0.0, Period=3.0, and NPaths=-1.0. Otherwise, these values are used: V=1.0, PO=0.0, Period=6.0, and NPaths=-1.0. OldTor Atom-type1 Atom-type2 Atom-type3 Atom-type4 Improper torsion (Amber [1]): Mag*[1+cos(Period*(θ-PO))] ImpTrs Atom-type1 Atom-type2 Atom-type3 Atom-type4 Mag PO Period Mag V/2 Magnitude PO Phase offset Period Periodicity Three term Wilson angle (Dreiding [28c], UFF [19]): ForceC*(C1 + C2*cos(θ) + C3*cos(2θ)) averaged over all three Wilson angles θ. Wilson Atom-type1 Atom-type2 Atom-type3 Atom-type4 ForceC C1 C2 C3 ForceC Force constant C1, C2, C3 Coefficients Harmonic Wilson angle (MMFF94 [6]): (ForceC/2)*(θ2) summed over all three Wilson angles θ. HrmWil Atom-type1 Atom-type2 Atom-type3 Atom-type4 ForceC ForceC Force constant Stretch-bend I (MMFF94 [5]): (ForceC1*(R12-Req12)+ForceC2*(R32-Req23))*(θ-θeq) StrBnd1 Atom-type1 Atom-type2 Atom-type3 ForceC1 ForceC2 Req12 Req23 θeq ForceC1, ForceC2 Force constants (in md/rad) Req12, Req23 Equilibrium bond lengths θeq Equilibrium angle USING SUBSTRUCTURES Substructures may be used to define different parameter values for a function for distinct ranges of some geometrical characteristic. Substructure numbers are appended to the function name, separated by a hyphen (e.g., HrmStr-1, HrmStr-2 and so on). The following substructures apply to functions related to bond stretches: • • • -1 -2 -3 Single bond: 0.00 ≤ bond order < 1.50 Double bond: 1.50 ≤bond order < 2.50 Triple bond: bond order ≥ 2.50 The following substructures apply to functions for bond angles (values in degrees): First substructure: • • • -1 -2 -3 0 ≤ θ ≤ 45 45 < θ ≤ 135 135 < θ ≤ 180 Second substructure: • -i-n Number of atoms bonded to the central one. For dihedral angles, one or two substructures may be used (e.g., AmbTrs-1-2). Use a zero for the first substructure to specify only the second substructure. 509.0 2.506.0 1.507].00 ≤ bond order < 1.0 C_2 C_2 * 45.50 Double central bond: 1.First substructure: • • • • -0 -1 -2 -3 Skip this substructure (substructure "wildcard") Single central bond: 0.70) Amide central bond (priority over resonance) None of the above Here is some simple MM force field definition input: HrmStr1 HrmStr1-1 HrmStr1-2 HrmBnd2 DreiTrs-1 DreiTrs-2 H_ C_2 C_2 * * * C_2 360.505.0 120.08 C_2 350.50 Triple central bond: bond order ≥ 2. No basis set keyword should be specified.504.0 Huckel This method keyword requests an extended Hückel calculation [503. Hoffmann Requests an Extended Huckel calculation using the default parameter set from the Huckel group.50 ≤ bond order < 2.0 180.40 C_2 * 50.50 C_2 500.0 180. Guess Requests an Extended Huckel calculation using the modified parameters used for Guess=Huckel [508. ExtendedHuckel is a synonym for this keyword.0 -1.0 -1.50 Second substructure: • • • -i-1 -i-2 -i-3 Resonance central bond (1.0 2.510]. .30 ≤ bond order ≤ 1.0 1. Muller Requests an Extended Huckel calculation using parameters collected by Edgar Muller.0 C_2 C_2 * 5.0 1. 000000 0. Guess=Huckel External Requests a calculation using an external program. and the name of the file which should be read in after the external program completes (the output file).544 -0. "analytic" gradients and numerical frequencies. A text file is produced with the current structure. but can also be used to conduct geometry optimizations using Gaussian's optimizer with external programs providing the function values and derivatives.968836513622 NIter= 0.637 -0. This mechanism is primarily intended to facilitate the use of external programs to provide the low-level calculations in ONIOM calculations.000000 0.043 Energy= -5. Dipole moment= 0. The energy appears in the output file as follows (followed by the x. y.-1.558 -0. Convert the results into a standard text form for recovery by Gaussian.352 The energy is as defined by this semi-empirical model. This script is expected to: • • • Convert the text file into input for another program. The script is passed two parameters: the name of the file Gaussian has prepared as input for the external program (the input file). Note that energy differences computed from the values in semi-empirical calculations are in Hartrees and may be compared directly with energy differences computed from jobs using other methods.000000 0. Gaussian uses a standardized interface to run an external program to produce an energy (and optionally a dipole moment or forces) at each geometry. and a script named Gau_External is run. and z components of the dipole moment): Huckel eigenvalues -.Energies. INPUT FILE FORMAT .245 -0. Run that program. For these reasons. and attempting to do so will result in an error. consult the description for that keyword before deciding to use this one. The remaining lines specify the atomic number. All basis functions specified with this keyword are added to the ones in the basis set specified in the route section. The second section is present only if first derivatives or frequencies were requested. It cannot be used to replace a definition within a built-in basis set.12 Repeated as needed. using any of the valid formats (which are described in detail in the discussion of the Gen keyword). 1=first derivatives. Format: 3D20. coordinates and molecular mechanics charge for each atom.The input file has the following format: #atoms derivatives-requested atomic# x y z MM-charge atom. These basis functions appear in a separate section in the input stream. In the latter case. Gen is often easier to use than ExtraBasis. charge & spinlow charge & spinmedium charge & spinhigh Repeated for each The first line specifies the number of atoms in the molecule. ExtraDensityBasis is ignored if no density fitting basis is specified in the route. and the final section is present only if frequencies were requested. and the molecule's charge and spin multiplicity. OUTPUT FILE FORMAT The output file is in fixed format. the Hessian is given in lower triangular form: αij. ExtraBasis is most useful for supplying basis functions for elements undefined in a standard basis set. and has the following information: energy dipole-moment(xyz) force on atom (xyz) Hessian (xyz) Format: 4D20. 2=second derivatives). i=1 to N. . what derivatives are to be computed (0=energy only.12 Format: 3D20. ExtraBasis ExtraDensityBasis These keywords indicate that additional basis functions are to be added to the basis set or density fitting basis set specified in the route section for the calculation (respectively).12 Repeated for each atom. j=1 to i. 00 0.p) ExtraBasis ..4380000000D-01 **** 0.Gen.00 0.1000000000D+01 **** ! here are some extra fitting functions.7500000000D+00 0.5 **** h 0 spd 1 0. Pseudo. GFInput. and f & g fitting functions 0.32 **** Frozen Core Options .000000000000 0.000000000000 0.1 cl h.00 0.29 ! here are some extra AO polarization functions cl 0 F 1 1.1.1000000000D+01 0.1000000000D+01 **** h 0 p 1 1..1000000000D+01 The following job supplies additional functions for both the basis set and for density fitting: #p rblyp/6-31g*/dga1 extrabasis extradensitybasis 6d HCl using the internally stored 6-31g* AO basis & DGA1 fitting set.p) basis set along with an additional diffuse function on all of the carbon atoms: # HF/6-31G(d. GenECP. cl 0 f 1 1. title section molecule specification C 0 SP 1 1. adding f functions to the AO basis.1612777588D+00 0. GFPrint The following job uses the 6-31G(d.1.5 g 1 1. FreezeInnerNobleGasCore In post-SCF calculations. the next to largest noble gas core is frozen. FrzINGC and FC1 are synonyms for this option. Gaussian 03 adds some additional options to the ones already available in the program [489]. RW The "read window" option means that specific information about which orbitals are retained in the post-SCF calculation will be given in the input file. then the highest m orbitals are retained. FC This indicates "frozen-core. FreezeG2 Freeze orbitals according to the G2 convention: d orbitals of main group elements are frozen. if the value for the last orbital is negative (-n). the outermost core orbitals are retained. then the highest |m| occupied and lowest |m| virtual orbitals are retained. Note that FC. Full This specifies that all electrons be included in a correlation calculation. but the outer sp core of 3rd row and later alkalai and alkalai earth elements are kept in the valence. depending on where it is used. A value of zero indicates the first or last orbital. followed by a blank line. The additional input section consists of a line specifying the starting and ending orbitals to be retained. then the highest n orbitals are frozen. If m is positive and n is omitted. This is the default calculation mode. except that the outer s and p core orbitals of 3rd row and later alkalai and alkalai earth atoms are not frozen (in accord with the G2/G3 conventions). Here are some examples for a calculation on C4H4: . If the value for the first orbital is negative (-m). FrzNGC is a synonym for this option. If m is negative and n is omitted. RW and Window are mutually exclusive. It is which is equivalent to FreezeG2 for the 6-31G and 6-311G basis sets and to FreezeNobleGasCore for all other basis sets.These options specify which inner orbitals are frozen in post-SCF calculations. That is." and it implies that inner-shells are excluded from the correlation calculation. FreezeNobleGasCore In post-SCF calculations the largest noble gas core is frozen. n defaults to 0. Full. Two lists are read for unrestricted calculations. and specifies the magnitude of the Fermi contact perturbation in the second format. . 5 of which will be frozen. Window=(m[.g.: 2 7-10 14 Field The Field keyword requests that a finite field be added to calculation. Field requires a parameter in one of these two formats: M±N or F(M)N where M designates a multipole.0.0 freezes the 4 core orbitals and keeps all virtual orbitals (equivalent to FC if the basis has a single zeta core). 5. there will be 14 occupied orbitals. but takes its input as parameters in the route section rather than from the input stream. if this is a closed shell calculation. since C4H4 has 28 electrons. terminated by a blank line.-4 freezes the four core orbitals and the highest four virtual orbitals. e.22 retains orbitals 6 through 22 in the post-SCF. and F(M) designates a Fermi contact perturbation for atom M (following the ordering in the molecule specification section of the input file). This is the appropriate frozen-core for a basis with a double-zeta core. ReadWindow is a synonym for RW.n]) Performs the same function as the ReadWindow option.0 is equivalent to Full. For example. A range of orbitals can be specified. -6 retains orbitals 9 through 20. so the post-SCF calculation will involve 9 occupied orbitals (orbitals 6-14) and 8 virtual orbitals (orbitals 15-22). ListWindow Causes a list of orbitals to freeze (omit from post-SCF calculations) to be read from the input stream. 5. 6. N*0.0001 specifies the magnitude of the field in atomic units in the first format. ChkWindow The window read in during a previous job is recovered from the checkpoint file. In Gaussian 03. the field can either involve electric multipoles (through hexadecapoles) or a Fermi contact term. Field=X+10 applies an electric dipole field in the X direction of 0. EChk Extracts only the three electric dipole field components from the checkpoint file. All parameters are in the input orientation. Chk is a synonym for Checkpoint. The field specification parameter may be placed among any other options as desired. Single point energy. Checkpoint Reads the 35 multipole components from the checkpoint file. RWF Takes the 35 multipole components from the read-write file. be careful of the choice of sign convention when interpreting the results. while Field=XXYZ-20 applies the indicated hexadecapole field with magnitude 0. OldRead Reads the coefficients of 35 electric multipole components from the input stream. geometry optimizations. Read Reads the coefficients of 34 electric multipole components from the input stream in free format.0027 times the spin density on atom 3. Note that the coefficients are those of the Cartesian operator matrices. LIMITATIONS .Thus. ERWF Extracts only the three electric dipole field components from the read-write file.0020 au and direction opposite to the default (which is determined by the standard orientation). Field=F(3)27 applies a perturbation of 0. and Force and Scan calculations. Similarly. Archiving is disabled when Field is specified.001 au.10 (the first component is a charge). in the old style format (including the monopole term): using format 3D20. 46 . use Guess=NoSymm whenever using Field with GVB. Here is an example using a Zmatrix: # RHF/3-21G Field=x+60 Opt=Z-Matrix NoSymm Z-Matrix optimization 0 C H H H H 1 1 1 1 1 B1 B2 B3 B4 A1 A2 A3 D1 D2 B1 B2 B3 B4 2 2 2 A1 A2 A3 3 3 D1 D2 1.0 y3=-0.000015 -119.070000 109. depending on whether the selected field breaks molecular symmetry.471231 120.070000 1.46 x3=0.471203 109. To be safe.12 x2=0.75 z2=-0.0 z1=0.471203 109. the finite field will or will not lead to correct numerical derivatives.999993 Here is an example using symbolic Cartesian coordinates: # HF/6-31G(d) Opt=Z-Matrix Field=z-50 NoSymm Symbolic Cartesian coordinates optimization 0 1 O 0 H 0 H 0 x1 y1 z1 x2 y2 z2 x3 y3 z3 x1=0.Note that if symmetry is left on during a GVB calculation.75 z3=-0. you must use Opt=Z-Matrix NoSymm keywords and define the input geometry either in traditional Zmatrix coordinates or symbolic Cartesian coordinates.070000 1.070000 1. To perform geometry optimizations in the presence of an electric field.0 y1=0.0 y2=0. The use of FMM is automated in Gaussian 03. By default. Levels=N Specifies the number of levels to use in the FMM. the exact results will vary from case to case (compact systems show the least speedup. FMM is enabled for nonsymmetric molecules with 60 atoms or more for both Hartree-Fock and DFT. BoxLen=N Sets the minimum box length (size) to N/10 Bohrs. Gaussian 03 generally turns on the FMM facility when using it provides even a modest performance gain (say. FMM is enabled for Hartree-Fock and hybrid DFT above 240 atoms and for pure DFT above 360 atoms.491. LMax=N Specifies the maximum order multipole. such as nearly linear polypeptides and long carbon nanotubes.490.FMM Force the use of the fast multipole method [29.33.30.492] if possible.31. .2x). Tolerance=N Specifies the accuracy level as 10-N. N is 30. For molecules with high symmetry. Thus. The default is 25 (or 15 when SCF=Sleazy is used). Of course. The default values for N are 8 for single point energy calculations and 10 for other calculation types. For molecules with low (but non-zero) symmetry. but the defaults are very unlikely to enable FMM when it has a negative effect on performance and are also as unlikely to fail to enable it when it would be worth a factor of 1. stretched out linear ones the most). The NoFMM keyword may be used to prevent this facility from being used. The default is 8 for molecules and is adjusted dynamically for PBC. intermediate thresholds are used.5x or more. You will begin to see substantial performance improvements (2x or better) with another factor of two in system size. 1. For a molecule with no symmetry.32. users are unlikely to need to control FMM by hand except for some very unusual special cases. AllNearField Turn on all near-field in FMM. CASSCF. The dipole moment is also computed (as a proper analytic derivative of the energy for MP2.447].01 Å. StepSize=N Sets the step size used in numerical differentiation to 0. the gradient of the energy). CISD. It is the default for all methods for which analytic gradients are unavailable. all DFT methods. MP3. Restart Restarts numerical evaluation of the forces. StepSize is valid only in conjunction with EnOnly. Sparse Force This calculation type keyword requests a single calculation of the forces on the nuclei (i. Analytic gradients are available for all SCF wavefunctions.. Note that this procedure exhibits some numerical instability. The forces on the nuclei appears in the output as follows (this sample is from a calculation on water): . EnOnly Compute the forces by numerically differentiating the energy once. For other methods. MP4(SDQ). gradients and frequencies for HF. MP2. CC. pure and hybrid DFT. This keyword may also be used within method specifications for ONIOM layers.e.0001*N. the forces are determined by numerical differentiation. The default step size is 0. CCD. CID.Energies. QCISD. so care must be taken that an optimal step size is specified for each case. CCSD. QCI and CI) [202. The units are Angstroms by default unless Units=Bohr has been specified. CIS. SAC-CI and all semiempirical methods. 023346514 3 1 . computing 3-21G frequencies at a STO-3G optimized geometry produces meaningless results. CCD. For example.000000000 . it is meaningless to compute frequencies at any geometry other than a stationary point for the method used for frequency determination. and by double numerical differentiation for those methods for which only energies are available.031211490 ------------------------------------------------------------------Internal Coordinate Forces (Hartree/Bohr or radian) Cent Atom N1 Length/X N2 Alpha/Y N3 Beta/Z ------------------------------------------------------------------1 O 2 H 1 -. The recommended practice is to compute frequencies following a previous geometry . Freq This calculation type keyword computes force constants and the resulting vibrational frequencies. UHF. the force constants are determined analytically if possible (for RHF. Intensities are also computed.***** AXES RESTORED TO ORIGINAL SET ***** ------------------------------------------------------------------Center Atomic Forces (Hartrees/Bohr) Number Number X Y Z ------------------------------------------------------------------1 8 -. MP4(SDQ). and all semi-empirical methods). all DFT methods.052127033 RMS . CIS.000000000 -. MP2. CID. QCISD. or vice-versa. Vibrational frequencies are computed by determining the second derivatives of the energy with respect to the Cartesian nuclear coordinates and then transforming to massweighted coordinates. but are restored to the original (Zmatrix) set of axes before printing (as noted in the output). and CASSCF). The forces are followed in each case by their maximum and root-mean-square values.052127033 ------------------------------------------------------------------MAX .028780519 2 1 .023347( 1) 3 H 1 -.023347( 2) 2 -.049849321 . By default. by single numerical differentiation for methods for which only first derivatives are available (MP3. This transformation is only valid at a stationary point! Thus. This is followed by the corresponding derivatives with respect to the internal coordinates (lengths and angles used in the Z-matrix) when internal coordinates are in use.054412682 J The forces are determined in the standard orientation. It is also incorrect to compute frequencies for a correlated method using frozen-core at a structure optimized with all electrons correlated.088273( 3) ------------------------------------------------------------------MAX . CISD.088272874 RMS .046711997 .003137324 .000000000 -. Use the Stable keyword to test the stability of Hartree-Fock and DFT wavefunctions. This option also computes optical rotations (see Polar=OptRot).optimization using the same method. the frequency analysis is performed and the results of the calculation are archived as a frequency job. This may be accomplished automatically by specifying both Opt and Freq within the route section for a job. It may be specified for DFT and MP2 calculations in order to produce Raman . when numerical differentiation is required (or requested with Freq=Numer). The keyword Opt=CalcAll requests that analytic second derivatives be done at every point in a geometry optimization. This option is valid for Hartree-Fock and DFT methods. Pre-resonance Raman intensities may be computed by specifying a Raman option..g. and also including CPHF=RdFreq within the route and specifying the desired frequency in the input file (see the examples for additional information). Note also that the coupled perturbed Hartree-Fock (CPHF) method used in determining analytic frequencies is not physically meaningful if a lower energy wavefunction of the same spin multiplicity exists. polarizabilities must be explicitly requested using the Polar keyword (e. The VCD option may be used to compute the vibrational circular dichroism (VCD) intensities in addition to the normal frequency analysis at the Hartree-Fock and DFT levels [242]. Therefore. the ReadIsotopes option is retained for rerunning completed calculations under different conditions (see the examples). Frequency-dependent polarizabilities and hyperpolarizabilities may similarly be computed by including CPHF=RdFreq within the route (subject to their usual availability restrictions). Raman Compute Raman intensities in addition to IR intensities. polarizabilities are also computed automatically. VCD Compute the vibrational circular dichroism (VCD) intensities in addition to the normal frequency analysis [242]. Once the requested optimization has completed all the information necessary for a frequency analysis is available. FREQUENCY CALCULATION VARIATIONS When frequencies are done analytically. However. You should specify alternative isotopes for frequency jobs using the standard method. QCISD Freq Polar). This is the default for HartreeFock. . NNRaman Do polarizability derivatives by numerically differentiating the analytic polarizability with respect to nuclear coordinates. and anharmonic vibrationalrotational couplings if VibRot is also specified [207. it is equivalent to NRaman.214]. pressure. This is the default if Raman is requested along with CPHF=RdFreq.209. at minimal computational cost.213.intensities by numerical differentiation of dipole derivatives with respect to the electric field.493.210.495]. DFT.494. Available input options are documented following the examples.208. and the normal mode and thermochemical analysis be repeated.210. saving 10-30% in CPU time.212. HPModes Include the high precision format (to five figures) vibrational frequency eigenvectors in the frequency output in addition to the normal three-figure output. NoRaman Skips the extra steps required to compute the Raman intensities during Hartree-Fock analytic frequency calculations. then Raman is equivalent to NNRaman for all methods.493.495]. This is the default for CIS. VibRot Analyze vibrational-rotational coupling [206. presumably using a different temperature.494. Note that since the basis set is read from the checkpoint file. ReadFC Requests that the force constants from a previous frequency calculation be read from the checkpoint file.211.209. DFT. For these methods. If the Raman option was specified in the previous job.208. If CPHF=RdFreq is used.211.207. CIS and MP2. NRaman Do polarizability derivatives by numerically differentiating the analytic dipole derivatives with respect to an electric field. This option is only available for methods with analytic second derivatives: Hartree-Fock. and MP2 if Raman is requested but CPHF=RdFreq is not. no general basis should be input. then do not specify it again when using this option. or isotopes. anharmonic frequencies [206. ReadAnharm Read an input section with additional parameters for the vibration-rotation coupling and/or anharmonic vibrational analysis (VibRot or Anharmonic options). Anharmonic Do numerical differentiation along normal modes to compute zero-point energies. N also specifies the step-size in the electric field. IntModes is a synonym for this option. For the projection. It is the default and only choice for those methods for which no analytic derivatives are available. and 0. the default is 0. and all DFT methods.005 for GVB and CASSCF Freq=Numer. compute the projected frequencies for vibrations perpendicular to the path. 0. EnergyOnly is a synonym for EnOnly. Numerical This requests that the second derivatives of the energy are to be computed numerically using analytically calculated first derivatives. This option is available only for RHF. and it is the default for those cases. The default is 0. Projected For a point on a mass-weighted reaction path (IRC). CASSCF. If Freq=Numer and Polar=Numer are combined. For Freq=Anharmonic or Freq=VibRot. Restart This option restarts a numerical frequency calculation after the last completed geometry (analytic frequency calculations are not restartable).01 Å for Freq=EnOnly. This computation is not meaningful at a minimum. Cubic Requests numerical differentiation of analytic second derivatives to produce third derivatives. It can be used with any method for which gradients are available and is the default for those for which gradients but not second derivatives are available. Freq=Numer can be combined with Polar=Numer in one job step. Analytic This specifies that the second derivatives of the energy are to be computed analytically. the geometry should be specified to at least 5 significant digits. Note that this computation is very sensitive to the accuracy of the structure and the path [496].0001*N (in Angstoms unless Units=Bohr has been specified). . A failed numerical frequency job may be restarted from its checkpoint file by simply repeating the route section of the original job.025. adding the Restart option to the Freq keyword. the gradient is used to compute the tangent to the path.InternalModes Print normal modes as displacements in redundant internal coordinates. MP2. Step=N Specifies the step-size for numerical differentiation to be 0. UHF. No other input is required. Accordingly.001 Å for Hartree-Fock and correlated Freq=Numer. EnOnly This requests double numerical differentiation of energies to produce force constants. CIS. such as transition states. the thermodynamic functions are corrected. CCD. If any normal modes are identified as internal rotation. See the discussion of the Opt keyword for details on the input format. pressure.. By default. Thus. the program will automatically use the corresponding actual exact isotopic mass (e. additional vibrational/internal rotation analyses may be performed by specifying Freq=(ReadFC. hindered or free. and/or isotopes (the defaults are 298. and Gaussian uses the value 17.g. and the most abundant isotopes). for which this is the same as Opt=ModRedundant. isotope mass for atom n Must be real numbers. MP4(SDQ). If the force constants are available on a previously generated checkpoint file. for use with InternalModes). DFT. CIS and CASSCF methods.e. this information can appear in a separate input section having the format: temp pressure [scale] isotope mass for atom 1 isotope mass for atom 2 . Note that the same coordinates are used for both optimization and normal mode analysis in an Opt Freq.. CCSD and QCISD. Analytic frequencies are available for the HF. the set of redundant internal coordinates may need to be altered via the Geom=Modify keyword. CISD. may have a specific bonding pattern not automatically recognized. MP2.15 K. pressure. The remaining lines hold the isotope masses for the various atoms in the molecule. and scale are the desired temperature. pressure. Opt=(CalcAll.99916). ModRedundant Read-in modifications to redundant internal coordinates (i. this option uses temperature.. arranged in the same order as they appeared in the molecule specification section. and scale factor specified in the route section. and an optional scale factor for frequency data when used for thermochemical analysis (the default is unscaled). CID. Since Opt=CalcAll automatically performs a vibrational analysis on the optimized structure. . If integers are used to specify the atomic masses. The identification of the rotating groups is made possible by the use of redundant internal coordinates. Because some structures. HinderedRotor). Numerical frequencies are available for MP3. 18 specifies O18.HinderedRotor Requests the identification of internal rotation modes during the harmonic vibrational analysis [497]. where temp. Alternatively. redundant internal coordinates must be used for the HinderedRotor option to function properly. ReadIsotopes Specify alternate temperature. HinderedRotor) may also be used. 1 atmosphere. pressure.. Specifying #P in the route section produces some additional output for frequency calculations. βXXZ.49920497D-16 -1.10991697D+01 #P also produces a bar-graph of the simulated spectra for small cases. The basic components of the output from a frequency calculation are discussed in detail in chapter 4 of Exploring Chemistry with Electronic Structure Methods [308].55019858D-14 -7.11369582D-17 2. You may be surprised to see output that looks like it belongs to a geometry optimization at the beginning of a frequency job: GradGradGradGradGradGradGradGradGradGradGradGradGradGradGrad Berny optimization. βXYY.. βXXY. and will automatically perform a frequency analysis at the final structure. βYYY. which will complete the optimization if the geometry is determined not to have fully converged (usually. βYZZ.07727337D+01 4. Of most importance are the polarizability and hyperpolarizability tensors (they still may be found in the archive entry in normal print-level jobs). βYYZ.37312183D-16 -6. only one additional optimization step is needed).72397709D-16 3.27350412D-14 4.αZZ and βXXX.17080415D-16 5. If you think this concern is applicable.e. The next step is printed at the end of a frequency calculation so that such problems can be identified. respectively (i.Polar. in the standard orientation: Dipole = 2. They are presented in lower triangular and lower tetrahedral order. αXX. This is done so that the quadratic optimization step can be computed using the correct second derivatives. Opt.39281319D-01 Polarizability= 7.09052038D-14 -2. Occasionally an optimization will complete according to the normal criterion using the approximate Hessian matrix.83427191D-01 1. βZZZ).αXY. which performs geometry optimizations.40402516D-13 -1. . αXZ.60008472D-15 6. Initialization pass.80285860D+00 -3.62729494D+00 HyperPolar = 3. use Opt=CalcAll instead of Freq in the route section of the job. given the full second derivative matrix near a stationary point.08796953D-16 -6. but the step size is actually larger than the convergence criterion when the correct second derivatives are used. Stable Frequency Output. βXZZ. αYY. αYZ. Link 103. βXYZ.26773439D-01 -1.66133815D-16 -9. is executed at the beginning and end of all frequency calculations. Thermochemistry analysis follows the frequency and normal mode data. The zero-point energy output in Gaussian has been expanded over that produced by older versions: Zero-point correction= .023261 (Hartree/Particle) Thermal correction to Energy= .026094 Thermal correction to Enthalpy= .027038 Thermal correction to Gibbs Free Energy= .052698 Sum of electronic and zero-point Energies=-527.492585 E0=Eelec+ZPE Sum of electronic and thermal Energies= -527.489751 E= E0+ Evib+ Erot+Etrans Sum of electronic and thermal Enthalpies=-527.488807 H=E+RT Sum of electronic and thermal Free Energies=-527.463147 G=H-TS The raw zero-point energy correction and the thermal corrections to the total energy, enthalpy, and Gibbs free energy (all of which include the zero-point energy) are listed, followed by the corresponding corrected energy. The analysis uses the standard expressions for an ideal gas in the canonical ensemble. Details can be found in McQuarrie [498] and other standard statistical mechanics texts. In the output, the various quantities are labeled as follows: E (Thermal) CV S Q Contributions to the thermal energy correction Constant volume molar heat capacity Entropy Partition function The thermochemistry analysis treats all modes other than the free rotations and translations as harmonic vibrations. For molecules having hindered internal rotations, this can produce slight errors in the energy and heat capacity at room temperatures and can have a significant effect on the entropy. The contributions of any very low frequency vibrational modes are listed separately so that if they are group rotations and high accuracy is needed, their harmonic contributions can be subtracted from the totals, and their correctly computed contributions included. Expressions for hindered rotational contributions to these terms can be found in Benson [499]. The partition functions are also computed, with both the bottom of the vibrational well and the lowest (zero-point) vibrational state as reference. Pre-resonance Raman. This calculation type is requested with one of the Raman options in combination with CPHF=RdFreq. The frequency specified for the latter should be chosen as follows: • • Determine the difference in frequency between the peak of interest in the Raman spectrum and the incident light used in the experiment. Perform a TD calculation using a DFT method in order to determine the predicted location of the same peak. • Specify a frequency for CPHF=RdFreq which is shifted from the predicted peak by the same amount as the incident light differs from the observed peak. Pre-resonance Raman results are reported as additional rows within the normal frequency tables: Harmonic frequencies (cm**-1), IR intensities (KM/Mole), Raman scattering activities (A**4/AMU), depolarization ratios for plane and unpolarized incident light, reduced masses (AMU), force constants (mDyne/A), and normal coordinates: 1 B1 Frequencies -- 1315.8011 Red. masses -1.3435 Frc consts -1.3704 IR Inten -7.6649 Raman Activ -0.0260 Depolar (P) -0.7500 Depolar (U) -0.8571 RamAct Fr= 1-0.0260 Dep-P Fr= 1-0.7500 Dep-U Fr= 1-0.8571 RamAct Fr= 2-0.0023 Dep-P Fr= 2-0.7500 Dep-U Fr= 2-0.8571 Vibration-Rotation Coupling Output. If the VibRot option is specified, then the harmonic vibrational-rotational analysis appears immediately after the normal thermochemistry analysis in the output, introduced by this header: Vibro-Rotational Analysis at the Harmonic level If anharmonic analysis is requested as well (i.e., VibRot and Anharmonic are both specified), then the anharmonic vibrational-rotational analysis results follow the harmonic ones, introduced by the following header 2nd order Perturbative Anharmonic Analysis Anharmonic Frequency Calculations. Freq=Anharmonic jobs product additional output following the normal frequency output. (It follows the vibration-rotation coupling output if this was specified as well.) We will briefly consider the most important items within it here. This output displays the equilibrium geometry (i.e., the minimum on the potential energy surface), followed by the anharmonic vibrationally averaged structure at 0 K: Internal coordinates for the Equilibrium structure (Se) Interatomic distances: 1 2 3 4 1 C 0.000000 2 3 4 O H H 1.220000 1.080000 1.080000 O2-C1-H3=120. O2-H3-H4= 62.0127 0.000000 1.993088 0.000000 1.993088 1.870615 0.000000 Interatomic angles: O2-C1-H4=120. H3-C1-H4=120. Dihedral angles: H4-C1-H3-O2= 180. Internal coordinates for the vibr.aver. structure at 0K (Sz) Interatomic distances: 1 2 3 4 1 C 0.000000 2 O 1.223954 0.000000 3 H 1.093363 2.007355 0.000000 4 H 1.093363 2.007355 1.894824 0.000000 Interatomic angles: O2-C1-H3=119.9442 O2-C1-H4=119.9442 H3-C1-H4=120.1116 O2-H3-H4= 61.8377 Dihedral angles: H4-C1-H3-O2= 180. Note that the bond lengths are slightly longer in the latter structure. The anharmonic zero point energy is given shortly thereafter in the output, preceded by its component terms: Zero Point Terms Harmonic ZPE (cm-1) Sum(Xij) (cm-1) 3rd der.Anh.E0 (cm-1) 4th der.Anh.E0 (cm-1) Vibr.Rot.E0 (cm-1) Anharmonic ZPE (cm-1) = = = = = = 6339.70913 -79.34418 -24.91960 23.36569 -4.77806 6254.03298 The anharmonic frequencies themselves appear just a bit later in this table, in the column labeled E(anharm): Vibrational Energies and Rotational Constants (cm-1) Mode(Quanta) E(harm) E(anharm) Aa(z) Ba(x) Equilibrium Geometry 9.560323 1.288616 Ground State 6339.709 6254.033 9.425702 1.283838 Fundamental Bands (DE w.r.t. Ground State) 1(1) 3180.793 3008.554 9.244416 1.283898 2(1) 1839.248 1805.679 9.432233 1.280472 3(1) 1661.905 1625.622 9.467760 1.288838 4(1) 1315.801 1292.782 7.968990 1.271489 5(1) 3292.300 3172.585 9.311674 1.282911 6(1) 1389.371 1365.996 10.859898 1.285869 Ca(y) 1.135528 1.125877 1.123734 1.118196 1.123277 1.126802 1.124406 1.119543 The harmonic frequencies are also listed for convenience. Rerunning a Frequency Calculation with Different Thermochemistry Parameters. The following two-step job contains an initial frequency calculation followed by a second thermochemistry analysis using a different temperature, pressure, and selection of isotopes: %Chk=freq # HF/6-31G(d,p) Freq Test Frequencies at STP molecule specification --Link1-%Chk=freq %NoSave # HF/6-31G(d,p) Freq(ReadIso,ReadFC) Geom=Check Test Repeat at 300 K 0,1 300.0 1.0 16 2 3 ... Note also that the freqchk utility may be used to rerun the thermochemical analysis from the frequency data stored in a Gaussian checkpoint file. ADDITIONAL INPUT FOR FREQ=READANHARM This input is read in a separate section which can contain the following keywords: Fermi Also perform a vibrational averaging of isotropic hyperfine couplings. PrintGeom Print the geometries at which properties for vibrational averaging are computed. TolFre=x Minimum frequency difference (cm-1) for Fermi and Darling-Dennison resonances (default 10.0). Must be a real number. TolCor=x Threshold (cm-1) on Coriolis couplings (default 10-3). Must be a real number. ScHarm=x Scaling factor for linear scaling of harmonic frequencies (1.0 x 10-5 for B3LYP/631+G(d)). Must be a real number. By default, the value from the normal Scale keyword is used. G1 G2 G2MP2 G3 G3MP2 G3B3 G3MP2B3 These method keywords request the Gaussian-1 (more colloquially known as G1) [80,81], Gaussian-2 (G2) [82], and Gaussian-3 (G3) [84] methods for computing very accurate energies. G2MP2 requests the modified version of G2 known as G2(MP2), which uses MP2 instead of MP4 for the basis set extension corrections [83], and is nearly as accurate as the full G2 method at substantially reduced computational cost. G3MP3 requests the similarly modified G3(MP2) method [85]. The G3 variants using B3LYP structures and frequencies [86] are requested with the G3B3 and G3MP2B3 keywords. All of these methods are complex energy computations involving several pre-defined calculations on the specified molecular system. All of the distinct steps are performed automatically when one of these keywords is specified, and the final computed energy value is displayed in the output. No basis set keyword should be specified with these keywords. Either of the Opt=Maxcyc=n or QCISD=Maxcyc=n keywords may be used in conjunction with any of the these keywords to specify the maximum number of optimization or QCISD cycles, respectively. You should specify alternative isotopes for these jobs using the standard method. However, the ReadIsotopes option is retained for rerunning completed calculations under different conditions (see the examples). ReadIsotopes Specify alternate temperature, pressure, and/or isotopes (the defaults are 298.15 K, 1 atmosphere, and the most abundant isotopes). This information appears in a separate input section having the format: temp pressure [scale] isotope mass for atom 1 isotope mass for atom 2 ... isotope mass for atom n Must be real numbers. where temp, pressure, and scale are the desired temperature, pressure, and an optional scale factor for frequency data when used for thermochemical analysis (the default value for the corresponding model is used if scale is omitted or set to 0.0); these values must be real numbers. The remaining lines hold the isotope masses for the various atoms in the molecule, arranged in the same order as they appeared in the molecule specification section. If integers are used to specify the atomic masses, the program will automatically use the corresponding actual exact mass (e.g., 18 specifies O18, and Gaussian uses the value 17.99916). Restart Resume a partially-completed calculation from its checkpoint file. When used in combination with ReadIso, this option allows for the rapid computation of the energy using different thermochemistry parameters and/or isotope selections. StartMP2 Assume that the specified checkpoint file contains the results of a Hartree-Fock frequency calculation at the HF/6-31G* optimized structure, and begins the G2 calculation from that point (implies Geom=AllCheck). Calculation Summary Output. After all of the output for the component job steps, Gaussian prints a table of results for these methods. Here is the output from a G2 calculation: Temperature= E(ZPE)= E(QCISD(T))= DE(Plus)= G1(0 K)= G1 Enthalpy= E(Delta-G2)= G2(0 K)= G2 Enthalpy= 298.150000 .020511 -76.276078 -.010827 -76.328339 -76.324559 -.008275 -76.332054 -76.328274 Pressure= E(Thermal)= E(Empiric)= DE(2DF)= G1 Energy= G1 Free Energy= E(G2-Empiric)= G2 Energy= G2 Free Energy= 1.000000 .023346 -.024560 -.037385 -76.325503 -76.303182 .004560 -76.329219 -76.306897 The temperature and pressure appear first, followed by the various components used to compute the G2 energy. The output concludes with the G2 energy at 0 K and at the specified temperature (the latter includes a full thermal correction rather than just the zero-point energy correction), and (in the final output line) the G2 theory predictions for the enthalpy and Gibbs free energy (both computed using the thermal-corrected G2 energy). (Note that the same quantities predicted at the G1 level are also printed in this summary section.) The energy labels thus have the following meanings (G2 is used as an example): G2 (0 K) Zero-point-corrected electronic energy: E0 = Eelec + ZPE G2 Energy Thermal-corrected energy: E = E0 + Etrans + Erot + Evib G2 Enthalpy Enthalpy computed using the G2 predicted energy: H = E + RT G2 Free Energy Gibbs Free Energy computed using the G2 predicted energy: G = H - TS Rerunning the Calculation at a Different Temperature. The following two-step job illustrates the method for running a second (very rapid) G2 calculation at a different temperature. This job computes the G2 energy at 298.15 K and then again at 300 K: %Chk=formald # G2 Test G2 on formaldehyde 0 1 molecule specification --Link1-%Chk=formald %NoSave # G2(Restart,ReadIso) Geom=Check Repeat at 300 K 0,1 300.0 1.0 isotope specifications Gen GenECP A set of "standard" basis sets is stored internally in Gaussian (see the "Basis Sets" section earlier in this chapter); these basis sets may be specified by including the appropriate keyword within the route section for the calculation. The Gen keyword allows a userspecified basis set to be used in a Gaussian calculation. It is used in the place of a basis set keyword or a density fitting basis set keyword. In this case, the basis set description must be provided as input (in a separate basis set input section). Gen may be used in a completely analogous way to specify an alternate density fitting basis set (see the examples). The GenECP variation may be used to read in both basis functions and ECPs; it is equivalent to Gen Pseudo=Read. It is designed for use in ONIOM calculations in which you want to use a general basis set with ECPs within one ONIOM layer. The GFPrint keyword may be used to include the gaussian function table within the output file. The GFInput keyword may be used to have the table printed in a form which is suitable for input to Gen. The ExtraBasis keyword may be used to make additions to standard basis sets. Similarly, the ExtraDensityBasis keyword may be used to make additions to standard density fitting basis sets BASIS FUNCTION OVERVIEW A single basis function is composed of one or more primitive gaussian functions. For example, an s-type basis function φμ(r) is: N is the number of primitive functions composing the basis function, and it is called the degree-of-contraction of the basis function. The coefficients dιμ are called contraction coefficients. The quantities αιμ are the exponents, and f is the scale factor for the basis function. The maximum degree-of-contraction permitted in Gaussian is 100. A shell is a set of basis functions φμ with shared exponents. Gaussian supports shells of arbitrary angular momentum: s, p, d, f, g, h, and so on. An s-shell contains a single s-type basis function. A p-shell contains the three basis functions pX, pY, and pZ. An sp-shell contains four basis functions with common gaussian exponents: one s-type function and the three p-functions pX, pY and pZ. A d-shell may be defined to contain either the six second-order functions (dX2, dY2, dZ2, dXY, dXZ, dYZ), or the five "pure d" basis functions (d z2-r2, dx2-y2, dxy, dxz, dyz). Likewise, an fshell may contain either the 10 third-order gaussians or the 7 "pure f" functions. Higher order shells function similarly. Note that the contraction coefficients in a shell must be the same for all functions of a given angular momentum, but that s and p contraction coefficients can be different in an sp-shell. A scale factor is also defined for each shell. It is used to scale all the exponents of primitives in the shell. The program has the ability to convert between the two types of functions [391]. Consider the series of basis sets STO-3G, 6-31G, and 6-311G(d) for the carbon atom. With the STO-3G, basis there are two shells on a carbon atom. One is an s-shell composed of 3 primitive gaussian functions (which are least-squares fit to a Slater 1s orbital). The other sp-shell is a least-squares fit of 3 gaussians to Slater 2s and 2p orbitals with the constraint that the s and p functions have equal exponents. These expansions are the same for all atoms. Only the scale factors for each shell differ from atom to atom. For carbon atoms, the 1s- and 2sp-shells have scale factors of 5.67 and 1.72, respectively. The 6-31G basis on a first row atom has three shells. One shell is a contraction of six primitive s-type gaussians. The second shell is a combination of three primitive sp-shells. The third shell consists of a single sp-function. These functions were optimized for the atom. Scale factors of 1.00, 1.00, and 1.04, respectively, for each shell for carbon were then determined by molecular calculations. As its name implies, the 6-311G(d) basis has 5 shells: an s-shell with 6 primitives, 3 sp-shells with 3, 1, and 1 primitives, and an uncontracted d-shell. All shells are "unscaled" (have unit scale factor). BASIS SET INPUT FORMAT External basis sets are read into Gaussian by specifying Gen (for general basis) in the route section. The keywords 5D, 6D, 7F, and 10F are used to specify use of Cartesian or pure d and f (and higher) functions; the defaults are 5D and 7F. All d-shells in a calculation must have the same number of functions. Similarly, f- and higher shells must either be all Cartesian or all pure. Defining a shell. External basis input is handled by the routine GenBas in Link 301. The basic unit of information that it reads from the basis set input section is the shell definition block. A shell definition block, together with the global specification of pure vs. Cartesian functions, contains all necessary information to define a shell of functions. It consists of a shell descriptor line, and one or more primitive gaussian lines: IType NGauss Sc α1 d1μ α2 d2μ ... αN dNμ Shell descriptor line: shell type, # primitive gaussians, and scale factor. Primitive gaussian specification: exponent and contraction coefficient. There are a total of NGauss primitive gaussian lines. IType defines the shell type and shell constraint and may be S, P, D, SP, SPD, F, G, ..., for an s-shell, p-shell, d-shell, sp-shell, f-shell, g-shell, and so on. NGauss specifies the number of primitive gaussian shells (the degree of contraction) for the shell being defined. The shell scale factor is given by Sc (i.e., all primitive exponents are scaled by Sc2). The subsequent NGauss primitive gaussian lines define the exponents αk and contraction coefficients, dkμ. Each line provides the exponent for one primitive, followed by its contraction coefficient (or s and p coefficients for an sp-shell). A second format also exists to specify a shell as a least-squares gaussian expansion of a Slater orbital. This is requested by a shell descriptor line of the form STO, IOrb, NGauss, Sc. IOrb is one of 1S, 2S, 2P, 2SP, 3S, 3P, 3SP, 3D, 4SP, and specifies which expansion is requested. Note that 2SP requests the best least-squares fit simultaneously to S and P slater orbitals and is not equivalent to separately specifying the best S and the best P expansions. NGauss is the same as above. Gaussian expansions of Slater functions having from 1 to 6 primitives are available. Sc is the scale factor and hence the exponent of the slater function being expanded. No primitive gaussian lines are required after a shell descriptor line requesting an STO expansion. Defining the basis for an atom or atom type. One customarily places at least one, and often several, shells on any given nuclear center ("atom"), via a center definition block. A center definition block consists of a center identifier line, and one shell definition block for each shell desired on the center(s) specified. It is terminated by a line with either asterisks or plus signs in columns 1 through 4: c1 c2 ... 0 IType NGauss Sc α2 d2μ ... αN dNμ ... IType NGauss Sc α2 d2μ ... αN dNμ **** Center identifier line: specifies applicability for these shells. First shell definition block. Additional shell definition blocks. Final shell definition block. Separator: terminates the center definition block. The center identifier line specifies a list of centers on which to place the basis functions in the center definition block, terminated by a 0. It can contain one or more integers, which are used to indicate the corresponding atom(s) in the molecule specification; more commonly, it contains a list of atomic symbols to refer to all atoms of a specific type. Center numbers and atomic symbols may be freely intermixed within a single center identifier line. To help detect input mistakes, if a center definition block specifies an atom that is not present in the molecule, the run is aborted. If the center is preceded by a minus sign (e.g. -H), the basis set information is simply skipped if no atom of that type is present in the molecule specification (the terminal zero may also be omitted in this case). The latter syntax is intended for creating basis set include files that specify a standard basis set for many atoms; once built, it can be included in its entirety in the input stream when the basis set is desired, via the include (@) function (as described earlier in this chapter). A center or atom type may be specified in more than one center definition block. For example, in the Gaussian 03 basis set directory—$g03root/g03/basis on UNIX systems —there is one file which specifies 6-31G as a general basis set (631.gbs), and another file containing d exponents which would be included as well to specify 6-31G* (631s.gbs). Every atom from H through Cl is specified in both files, and in practice both of them would be included (most often along with additional basis set specifications for those atoms in the molecule for which the 6-31G basis set is not available). Drawing on Pre-Defined Basis Sets in Gen Input. Gaussian adds flexibility to general basis set input by allowing them to include pre-defined basis sets within them. Within a center definition block for an atom type (or types), an entire shell definition block may be replaced by a line containing the standard keyword for a pre-defined basis set. In this case, all of the functions within the specified basis set corresponding to the specified atom type(s) will be used for all such atoms within the molecule. The SDD, SHF, SDF, MHF, MDF, MWB forms may be used to specify Stuttgart/Dresden basis sets/potentials within Gen basis input. Note that the number of core electrons must be specified. Here is a portion of the Gen input corresponding to the 6-31+G(d) basis set: H 0 S 3 1.00 0.1873113696D+02 0.3349460434D-01 0.2825394365D+01 0.2347269535D+00 0.6401216923D+00 0.8137573262D+00 S 1 1.00 0.1612777588D+00 0.1000000000D+01 **** C 0 S 6 1.00 0.3047524880D+04 0.1834737130D-02 0.4573695180D+03 0.1403732280D-01 0.1039486850D+03 0.6884262220D-01 0.2921015530D+02 0.2321844430D+00 0.9286662960D+01 0.4679413480D+00 0.3163926960D+01 0.3623119850D+00 SP 3 1.00 0.7868272350D+01 -0.1193324200D+00 0.1881288540D+01 -0.1608541520D+00 0.5442492580D+00 0.1143456440D+01 SP 1 1.00 0.1687144782D+00 0.1000000000D+01 D 1 1.00 0.8000000000D+00 0.1000000000D+01 **** C 0 Applies to all carbons. SP 1 1.00 0.4380000000D-01 0.1000000000D+01 **** Applies to all hydrogen atoms. Applies to all carbons. 6-31G functions. 0.6899906660D-01 0.3164239610D+00 0.7443082910D+00 0.1000000000D+01 Polarization function. Diffuse function. 0.1000000000D+01 The following Gen input uses the 6-31G(d,p) basis set for the carbon and hydrogen atoms and the 6-31G†† basis set for the fluorine atoms in the molecule, and places an 1 cl h. 6D is specified because the default for general basis input is 5D but the 6-31g* basis is defined to use 6D 0.00 0.2421450000D+03 0. The following example uses general basis set input to specify both the basis set and the density fitting basis set.2518010000D+05 0.gbs/N Note that .1403419883D-01 0.1.6909739426D-01 0.4830339599D+00 0.00 0.extra function only on center number 1 (which happens to be the first carbon atom in the molecule specification for 1. # RBLYP/GEN/GEN 6D HCl: reading in 6-31g* AO basis and DGA1 fitting set. 0.1000000000D+01 The following job uses the Gaussian include file mechanism to specify the basis functions for chromium: # Becke3LYP/Gen Opt Test HF/6-31G(*) Opt of Cr(CO)6 molecule specification C O 0 6-31G(d) **** @/home/gwtrucks/basis/chrome.8604740000D+03 0.7733490000D+02 0.3398559718D+00 .1-difluoroethylene): C H 0 6-31G(d.29 ! here are the 6-31g* basis sets for Cl and H cl 0 S 6 1.1832959848D-02 0.2374519803D+00 0.p) **** F 0 6-31G(d'.2624700000D+02 0.4380000000D-01 **** Place a diffuse function on just one carbon atom.1000000000D+01 0.3780350000D+04 0.gbs is the conventional extension for basis set files (for gaussian basis set).p') **** 1 0 SP 1 1.1. 1000000000D+01 **** h 0 S 1 1.1000000000D+01 S 1 1.4203770000D+00 1 1.7500000000D+01 0.3071371894D-01 -0.3279510723D+00 0.00 0.8137573261D+00 0.1000000000D+01 0.4500000000D+02 0.2560000000D+03 0.00 0.1000000000D+01 S 1 1.6400000000D+02 0.1000000000D+01 0.1000000000D+01 S 1 1.1000000000D+01 0.00 0.1000000000D+01 S 1 1.2347269535D+00 0.5893533634D+00 0.7500000000D+00 -0.1000000000D+01 0.1125280694D+00 0.1000000000D+01 0.1000000000D+01 SPD 1 1.4535271000D+00 0.1000000000D+01 0.1612777588D+00 **** 6 1.2500000000D+00 0.00 0.1000000000D+01 S 1 1.00 0.00 0.00 0.00 0.4096000000D+04 0.1000000000D+01 S 1 1.1144270000D+01 0.1298800286D+00 0.00 0.00 0.1426570000D+00 1 1.6401216923D+00 S 1 1.1000000000D+01 0.7435077653D+00 0.3031770668D-01 0.1429931472D-01 0.2297391417D-02 -0.00 0.00 0.1060184328D+01 0.1024000000D+04 0.3349460434D-01 0.00 0.1000000000D+01 SPD 1 1.4501632776D-01 0.00 0.1000000000D+01 0.1000000000D+01 0.4000000000D+01 0.2518280280D+00 -0.1000000000D+01 0.1000000000D+01 SPD 1 1.3989400879D-02 0.1500000000D+01 0.5452150000D+01 0.00 0.1378340000D+02 0.00 0.1000000000D+01 ! here are the DGA1 fitting sets for Cl and H cl 0 S 1 1.2048000000D+05 0.1000000000D+01 .4917650000D+03 0.2225880000D+01 3 1.6158925141D-01 0.3186490000D+01 0.2825394365D+01 0.3741530000D+02 0.1169840000D+03 0.1000000000D+01 SPD 1 1.2000000000D+02 0.SP SP SP D **** h 0 S 3 1.1000000000D+01 0.3235723331D+00 0.00 0.1873113696D+02 0.00 0.4652062868D+00 0.1000000000D+01 0.2521540556D+00 -0. Step=N Retrieves the structure produced by the Nth step of a failed or partial geometry optimization (it is not valid for a successful optimization). Only the charge and multiplicity are read from the input stream. ExtraDensityBasis.S **** 1 1.1000000000D+01 If you wanted to specify the density fitting basis set with general basis set input. then you would use a route section like this one (substituting the appropriate basis set for your problem): # RBLYP/6-31G(d. This action is safe since Gaussian will abort the job if an optimization fails. It also controls what geometry-related information is printed and use of internal consistency checks on the Z-matrix. This option is not valid with Modify but may be combined with ModRed. Geom=Checkpoint may be used by a later job step to retrieve the geometry optimized during an earlier job step from the checkpoint file. ITEM SELECTION OPTIONS Checkpoint Causes the molecule specification (including variables) to be taken from the checkpoint file. Pseudo Geom The Geom keyword specifies the source of the molecule specification input.p)/Gen 6D ExtraBasis.00 0. GFPrint. the charge and multiplicity. May be combined with the ModRedundant option if you want to retrieve and alter the molecule specification in a checkpoint file using redundant internal coordinate-style modifications. By default. and consequently subsequent job steps which expect to use the optimized geometry will not be executed. Thus. The Geom keyword is not meaningful without at least one item selection option. GFInput. Step=Original recovers the . For example. Geom may be used to specify an alternate input source. and the title section to be taken from the checkpoint file. it is read from the input stream. as described previously. only the route section and any input required by keywords within it need be specified when using this option.3000000000D+00 0. AllCheck Causes the molecule specification (including variables). initial starting geometry. Check or AllCheck to retrieve and modify a geometry from a checkpoint file. The ModLargeRedundant variation uses the minimal setup for Opt=Large. K Remove the coordinate and kill all related coordinates containing this coordinate. When combined with the AllCheck option. Modification specifications for redundant coordinates have the same format as the input for the ModRedundant option of the Opt keyword (we summarize these formats only briefly here. N3 and N4 are atom numbers or wildcards. Action is an optional one-character code letter indicating the coordinate modification to be performed. S n stp Perform a relaxed potential energy surface scan. only the geometry modifications input is needed. F Freeze the coordinate in the optimization. and increment the coordinate by stp a total of n times. ModRedundant Modify the current geometry (regardless of its coordinate system) using redundant internal coordinate modifications before performing the calculation. When used with Check or Step. and +=Value increments the coordinate by Value. a Hessian updated message in the log file means that the corresponding step is available in the checkpoint file. This option may be used to modify a geometry specified in the input file using these features even when some calculation type other than an optimization is to be performed. AllCheck or Modify. It may also be combined with Step. and the second contains alterations to the retrieved geometry. two input sections will be read: the first contains the charge and multiplicity. Set the initial value to Value (or its current value). N2. (numbering begins at 1 and any dummy atoms are not counted.) Value gives a new value for the specified coordinate. see the discussion of the Opt keyword for a full description): [Type] N1 [N2 [N3 [N4]]] [[+=]Value] [Action [Params]] [[Min] Max]] N1. performing an optimization from each resulting starting geometry. sometimes followed by additional required parameters (the default action is to add the specified coordinate): • • • • • • B Add the coordinate and build all related coordinates. It may not be used for periodic boundary calculations. . This option is used for restarting geometry optimization from intermediate points. R Remove the coordinate from the definition list (but not the related coordinates). A Activate the coordinate for optimization if it has been frozen. Note that not all steps are always present in the checkpoint file. It must be combined with one of Checkpoint. D Calculate numerical second derivatives for the row and column of the initial Hessian for this coordinate. new-value is an optional new value to be assigned to it. Value. then the variable's status remains the same as it was in the original molecule specification. where the fourth atom is used to determine the 2 orthogonal directions of the linear bend. and the second contains alterations to the retrieved geometry. Min and Max define a range (or maximum value if Min is not given) for coordinate specifications containing wildcards. the coordinate type is determined automatically from the number of atoms specified): • • • • • • X Cartesian coordinates. specifying the X. Min and Max are each pairs of numbers. O Out-of-plane bending coordinate for a center (N1) and three connected atoms. in the initial Hessian to dv. Note that in Gaussian 03. Connect Specify explicit atom bonding data via an additional input section (blank line-terminated) following the geometry specification and any modification to it. In this case.Z coordinates. Modification specifications for geometry optimizations using Z-matrix coordinates have the following form: variable [new-value] [A|F|D] where variable is the name of a variable in the molecule specification. and the final item is a one-letter code indicating whether the variable is to be active (i. B Bond length A Valence angle D Dihedral angle L Linear bend specified by three atoms (or if N4 is -1) or by four atoms. Value. Modi is the shortest valid abbreviation for this keyword. An asterisk (*) in the place of an atom number indicates a wildcard.e. Modify Specifies that the geometry is to be taken from the checkpoint file and that modifications will be made to it. The Action is taken only if the value of the coordinate is in the range.• • H dv Change the diagonal element for this coord. A total of two input sections will be read: the first contains the charge and multiplicity. Type can be used to designate a specific coordinate type (by default. specifying the two orthogonal bending components.Y. ordered the same as in the molecule specification. the code letter D requests numerical differentiation be performed with respect to that variable and activates the variable automatically.. Min and Max are each triples of numbers. If the code letter is omitted. using the following syntax: N1 Order1 [N2 Order2 …] . This option requires one line of input per atom. optimized) or frozen. In this case. 0 respectively.0 This input section is terminated by a blank line.0 5 2.0 -1. OldRedundant Use the Gaussian 94 redundant internal coordinate generator.0 removes a bond. with bond orders of 1.0 and 2. InitialHarmonic is a synonym for this option.0 9 -1. For example.0 respectively: 8 4 1. and the Order's are the bond order of the corresponding bond. Connectivity modifications use the following syntax: M N1 Order1 [N2 Order2 …] where M is the atom number. Bond orders involving dummy atoms are discarded. This option requires an additional input section (blank line-terminated) following the geometry specification and any modification to it.0 and 2. For example. this input specifies that the current atom is bonded to atoms 4 and 5.where the N's are atoms to which the current atom is bonded. with force constant n/1000 Hartree/Bohr2.0 ZMConnect Read connectivity using the atom numbering specified in the Z-matrix (including dummy atoms). and the Order's are the bond order of the corresponding bond. OUTPUT-RELATED OPTIONS . IHarmonic=n Add harmonic constraints to the initial structure with force constant n/1000 Hartree/Bohr2. A bond order of -1. this input specifies that atom 8 is bonded to atoms 4 and 5. RHarmonic is a synonym for this option. the N's are atoms to which that atom is bonded. CHarmonic is a synonym for this option.0 5 2. and removes any bond to atom 9: 8 4 1. ModConnect Modify the connectivity of the atoms in the molecule specification (or retrieved from the checkpoint file). ReadHarmonic=n Add harmonic constraints to an additional structure read in the input stream (in the input orientation). ChkHarmonic=n Add harmonic constraints to the initial structure saved on the checkpoint file with force constant n/1000 Hartree/Bohr2. with bond orders of 1. and NoAngle may be specified. The default is not to print unless some atoms are specified by Cartesian coordinates or an optimization in redundant internal coordinates is being performed. KeepDefinition Retains the definition of the redundant internal coordinates (the default). CAngle Requests printing of interatomic angles using distance cutoffs to determine bonded atoms. Only one of Dihedral. CAngle.Distance Requests printing of the atomic distance matrix (which is the default for molecules with fewer than 50 atoms). CDihedral Requests printing of dihedral angles using distance cutoffs to determine connectivity. and NoDihedral may be specified. The default is to retain them in symbolic form for the Berny algorithm. Crowd Crowd activates and NoCrowd turns off a check which aborts the job if atoms are closer . the new modifications are appended to any earlier Opt=ModRedundant input before the coordinate system is updated. Dihedral Specifies printing of dihedral angles using connectivity information from the Z-matrix to decide which atoms are bonded (the default is not to print). GEOMETRY SPECIFICATION AND CHECKING OPTIONS KeepConstants KeepConstants retains and NoKeepConstants discards information about frozen variables. CDihedral. using the Z-matrix to determine which atoms are bonded. Angle Requests printing of the interatomic angles. Its opposite is NewDefinition. NoDihedral suppresses this output. If used with Geom=Modify. The default is not to print unless at least one atom is specified using Cartesian coordinates. Only one of Angle. NoAngle suppresses this output. and to discard them for older optimization algorithms (which don't understand them anyway). NoDistance suppresses this output. NewRedundant Rebuilds the redundant internal coordinates from the current Cartesian coordinates. PrintInputOrient Include the table giving the Cartesian coordinates in the input orientation. This is done by default only if a full optimization is requested using the Berny algorithm (Opt=Z-matrix).than 0. Independent Independent activates and NoIndependent turns off a check on the linear independence of the variables specified in a Z-matrix. the check is done at the initial point. Guess=Read. and can thus be used in adding to or modifying standard basis sets. GFPrint GFPrint . MODEL BUILDER OPTIONS ModelA. Gen. but not at later points of an optimization. the job will be aborted. This option is implemented only for H through Ne. By default. ModelB These options specify that model builder [500] connectivity information will be read and used to construct a symbolic Z-matrix.5 Å. and in some cases will not generate a symbolic Z-matrix with the correct symmetryconstrained number of variables. Opt=ModRedundant GFInput The GFInput ("Gaussian Function Input") output generation keyword causes the current basis set to be printed in a form suitable for use as general basis set input. If geometry optimization has been requested and this problem occurs. Print Turns on additional printing by the model builder facility. 0 times the QEq value. The default is 7. a Harris guess is used (see below).This output generation keyword prints the current basis set and density fitting basis set in tabular form. OldHuckel Use the old Huckel guess (pre-Gaussian 03) instead of CNDO or the updated Huckel. Gen. INDO Use the Gaussian 98 default guess: INDO for first-row systems. This is the default unless atoms heavier than Xe are present. which is the default when atoms heavier than Xe are present. The variant GFOldPrint keyword prints the basis set information in the Gaussian format. By default. Guess is not meaningful without an option. AM1 Do an AM1 calculation for the initial guess (currently only works with sparse matrix code). Huckel Requests that a Huckel guess be generated. . CNDO for secondrow.Always) causes later steps in a geometry optimization to generate a new guess at each point and compare the energies with the density from the old point and the new guess and take the better. Harris Diagonalize the Harris functional [501] for the initial guess. Guess=(AM1. GFInput Guess This keyword controls the initial guess for the Hartree-Fock wavefunction.and Huckel for third-row and beyond. RdScale Read in the scale factor on atomic hardnesses used in iterative extended Huckel. the β section is required even though it is empty (and vice-versa). Both sections are always required. By default. Alter Indicates that the orbitals selected for occupation in the Hartree-Fock wavefunction should not be those of lowest energy.g. projected onto the current basis set. the SCF results from the last point are used for the guess at the next point. The alteration sections consist of a set of transpositions indicating that one of these occupied orbitals is to be replaced by one of the other (virtual) orbitals. Separate permutation lists for α and β orbitals must be specified (on separate lines) for open shell systems. DensityMix[=N] Whether to mix occupied and virtual orbital contributions in forming the initial guess density. the first specifying transpositions ofα orbitals.Core Requests that the core Hamiltonian be diagonalized to form the initial guess. even if only α transpositions are needed. and then the specified alterations are made. Each such transposition is on a separate line and has two integers N1 and N2 (free format. Ranges (e. the occupied orbitals are selected as those with lowest eigenvalues for the one-electron Hamiltonian used in the initial guess programs. Always Requests that a new initial guess be generated at each point of an optimization. N defaults to -3 (use Huckel eigenvalues to decide which orbitals to mix). For UHF calculations. The TCheck option says to attempt to read a guess from the checkpoint file. and all orbitals not listed are put in after the listed orbitals in their original order. Thus. Normally. Guess=Core is most commonly used for atomic calculations. This option may be combined with Alter. The numbers of the generated guess orbitals are given in the order in which they should be used in the SCF. in which case the orbitals are read from the checkpoint file. Permute Read in a permutation of orbitals in the initial guess. two such orbital alteration sections are required. Read Requests that the initial guess be read from the checkpoint file (Guess=Read is often specified along with Geom=Checkpoint). The second blank line to indicate an empty β section must be included. Checkpoint is a synonym for Read. . The list of orbital transpositions is terminated by the blank line at the end of the input section. 712) can be used. but to generate a new one if necessary. and the second specifying transpositions of β orbitals. separated by spaces or a comma as usual) indicating that orbital N1 is to be swapped with orbital N2. This option can be used to read a complete initial guess from the input stream by replacing every orbital. which includes Guess(Read. with the new groups separated by 0. It may fail in unusual cases. This enables the orbitals (and possibly but not necessarily the total wavefunction) to have lower symmetry than the full molecular point group. Localized orbital analysis of a converged SCF wavefunction may then be done using a second job step. such as when a wavefunction is used as a guess for a system with a different stoichiometry. This is the default. Synonymous with SCF=NoSymm and Symm=NoSCF. The numbers correspond to the order in which the representations are listed by Link 301 in the output file (see the examples subsection below).Mix Requests that the HOMO and LUMO be mixed so as to destroy α-β and spatial symmetries. in which case Guess=NoTranslate should be specified. which is read in. This option is available only for GVB calculations. NoSymm Requests that all orbital symmetry constraints be lifted. to allow lowered symmetry of the wavefunction. Note that irreducible representations are combined before orbital localization is done and that localized orbitals retain whatever symmetry is kept. LowSymm Requests that irreducible representations of the molecular point group be combined in the symmetry information used in the N3 steps in the SCF. This is useful in producing UHF wavefunctions for singlet states. Cards Specifies that after the initial guess is generated. Since this input section is always exactly one line long.Only) and Pop=Full in its route section. Occupied and virtual orbitals are localized separately. be translated to the current atomic coordinates. The option expects a single line of input (in the format 16I2) giving the numbers of the irreducible representations to combine. The replacement orbitals . some or all of the orbitals will be replaced with ones read from the input stream.Local. the list itself must be terminated by a 9. Translate Translate requests that the coordinates of the atoms used to produce a guess. Local Requests that orbitals be localized using the Boys method [421]. Guess=NoSymm removes all orbital symmetry constraints without reading any input. it is not terminated by a blank line. where it is often necessary for calculations on symmetric systems (see the discussion of the GVB keyword below for an example using this option). and the irreducible representations (after possible merging using LowSymm or NoSymm) are not mixed. Read) may also be used to produce population and other post-calculation analyses from the data in a checkpoint file.N) Orbital to replace (0=end. and it is the default for non-PBC calculations. This is the default for periodic boundary conditions calculations if Guess=Alter is not specified. Guess=Only may be specified with CASSCF in order to obtain information on the number of CI configurations in the CAS active space (as well as the initial orbitals).Read) Prop will cause electrostatic properties to be calculated using the wavefunction in the checkpoint file. -1=replace all orbitals in order). New orbital in the format specified in the first line. enclosed in parentheses. The remainder of the section contains one or more instances of the following: IVec (A(I. these options alone will produce a population analysis using the wavefunction in the checkpoint file.are placed in the input section following the guess alteration commands. The replacement orbitals input section (the α replacement orbitals section for UHF) begins with a line specifying the Fortran format with which to read the replacement orbital input. NoFock disables this behavior. Guess(Only. Guess(Only. For example. Note that the amount of orbital information that is printed is controlled by the Pop keyword. This option is useful for saving localized orbitals. See the examples section for sample replacement orbital input.I=1. Print Print the initial guess. Fock Reuse Fock matrices rather than orbitals when reading from previous results on the rwf or chk files. Save Save the generated initial guess back into the checkpoint file at the conclusion of a Guess=Only run. For example: (4E20.8). Only Guess=Only functions as a calculation type keyword and requests that the calculation terminate once the initial guess is computed and printed. For example. This option is useful in preliminary runs to check if configuration alteration is necessary. The format for the line containing IVec is Fortran I5. Alpha Use alpha orbitals for both alpha and beta guess during Guess=Read. Guess=Only may not be used with semi-empirical methods. The β orbital replacement section for UHF calculations differs only in that it omits the initial format specification line. For UHF.IVec). . there are separate α and β replacement orbital input sections. if any. This must be accomplished via a separate job step specifying this option as well as Check. RESTRICTIONS Guess=Only may not be used with semi-empirical methods.Read) is contradictory and will lead to unpredictable results.e. Geom.Alter) and Guess=(Read.. First. By default. Pop Transposing 2 Orbitals with Guess=Alter. and the optimization did not take a small step as flagged by variable 4 in ILSW. alterations are read once and the same interchanges are applied at each geometry). Conversely. new initial guess when reading orbitals from the RWF (i. Guess=(Always. ForceAbelianSymmetry Force the initial guess orbitals to transform according to irreps of the Abelian point group. no alteration of configuration was requested. Only.Extra Do an extra. Use NoExtra to disable this feature. Refer to the input sections order table at the beginning of this chapter to determine the ordering of the input sections for combinations of options like Guess=(Cards. See the discussion of the Population keyword for details. This option may be useful for very. during geometry optimizations). These options may be combined in any reasonable combination. Thus Guess=(Always. and Read. this is done if the default Harris guess is allowed. a Guess=Only calculation is run to determine whether any alter instructions are needed to obtain the desired electronic state. This example finds the UHF/STO-3G structure of the 2A1 excited state of the amino radical. very large HF or DFT calculations using the sparse matrix facility. The HF/STO-3G theoretical model is used by default: # Guess=Only Test Amino radical test of initial guess .Alter) work as expected (in the former case. NaturalOrbitals Include natural orbitals in the checkpoint file. Sparse Perform a sparse SE calculation for the initial guess.Alter). NoForceAbelianSymmetry is the default. 75. Alpha Orbitals: Occupied (A1) (A1) (B2) (B1) (A1) Virtual (A1) (B2) Beta Orbitals: Occupied (A1) (A1) (B2) (A1) Virtual (B1) (A1) (B2) <S**2> of initial guess= . The expectation value of S2 for the unrestricted initial guess is printed.0 Here is the orbital symmetry summary output from the job. the electron configuration in the initial guess is a12a12b22a12b1. since no SCF has been performed. . yielding a 2B1 wavefunction. since we want to model the 2A1 excited state.7544 Since a doublet state is involved. This is indeed the ground state of NH2.g. Note that the orbital energies printed in a Guess=Only job are simply -1.0 for the occupied orbitals and 0.03 hnh 120. In this case. a full semi-empirical energy calculation can be performed specifying the desired method (e. which comes immediately before the population analysis in the output: Initial guess orbital symmetries.03 hnh 120. Guess=Alter may also be used to accomplish this.0 for the virtual orbitals.0 2 n h 1 nh h 1 nh 2 hnh nh 1. Here is the input for the geometry optimization # UHF/6-31G(d) Opt Guess=Alter Pop=Reg Test Amino radical: HF/6-31G(d) structure of 2-A1 state 0 2 n h 1 nh h 1 nh 2 hnh Variables: nh 1. it is close to the pure doublet value of 0. we will need to alter this initial orbital configuration: a β electron must be moved from orbital 4 to orbital 5 (the electron configuration is then a12a12b22b12a1). From the orbital symmetries.0 Blank line ends the molecule specification section. α and β orbitals are given separately. If the actual orbital energies are desired. Returning to our consideration of the amino radical. INDO). 4 5 Blank line ends the α section(empty in this case). Transpose orbitals 4 and 5. End of the β alteration section. Note that an extra blank line-line 12-is necessary to indicate an empty α alteration section. The final two lines then constitute the β alteration section. The initial guess program prints a list of orbitals that were interchanged as a result of the Alter option: Projected INDO Guess. NO ALPHA ORBITALS SWITCHED. PAIRS OF BETA ORBITALS SWITCHED: 4 5 The eigenvalue of S2 is printed for the UHF wavefunction. The value which results if contamination of the wavefunction from the next possible spin multiplicity (quartets for doublets, quintets for triplets, etc.) is removed is also printed: Annihilation of the first spin contaminant: S**2 before annihilation .7534, after .7500 Although this calculation does in fact converge correctly to 2A1 state, it sometimes happens that the order of orbital symmetries switches during the course of the SCF iterations. If the orbital symmetries of the final wavefunction are different from those in the initial guess (whether or not you are using Guess=Alter), we recommend using the direct minimization routine, specified with the SCF=QC or SCF=DM keywords, which usually holds symmetry from one iteration to the next. Reordering Orbitals with Guess=Permute. This option is often is the easiest way to perform a complex modification of the initial guess, as in this example: # CASSCF/6-31G(d,p) Opt Guess=Permute Pop=Reg Test CAS job 0 1 molecule specification 1-60 65 63 64 66 68 67 61-62 69 Specify new ordering. Here we have rearranged orbitals 61-68. Listing the final orbital (69) is not really necessary, but it help to make the input easier to understand for humans. Reading in Orbitals with Guess=Cards. Some or all of the orbitals may be replaced after the initial guess is generated using Guess=Cards. Here is some sample input for this option, which replaces orbitals 1 and 4 (note that the format for the third and following lines is specified in line 1): (3E20.8) 1 0.5809834509E+00 0.1724432549E-02 0.1639966912E-02 -0.4538843604E-03 0.6038992969E-04 4 0.7700779642E-13 -0.4479190461E-12 0.6441113412E-12 -0.1190754528E-11 -0.2567325943E+00 0 0.4612416518E+00 0.1282235396E-14 -0.9146282229E-15 0.6038992958E-04 -0.1131035471E-03 0.1240395916E-12 -0.1478805861E-13 -0.3119296374E-14 0.2567325943E+00 -0.1459733219E+00 -0.6437319952E-04 0.5417658499E-13 -0.6407549694E-13 -0.1131035485E-03 -0.3110890228E-12 0.5807753928E+00 0.1554735923E+00 0.1459733219E+00 An orbital number of zero ends the replacement orbital input. GVB This method keyword requests a perfect-pairing General Valence Bond (GVB-PP) calculation. GVB requires one parameter: the number of perfect-pairing pairs to split; for example: GVB(4). This parameter may also be specified with the NPair option. The natural orbitals for the GVB pairs are taken from occupied and virtual orbitals of the initial guess determinant (described below). INPUT FOR GVB CALCULATIONS Normally most of the difficult input for a GVB-PP calculation involves specifying the initial guess. (Link 401). This often includes alteration of orbitals to ensure the correct identification of high-spin, perfect-pairing, and closed-shell orbitals and possible reduction of SCF symmetry to account for the localized orbitals which usually represent the lowest energy solution for GVB-PP. The GVB program reads the number of orbitals in each GVB pair (in format 40I2). The number of lines read is fixed (and normally 1), so no terminating blank line is needed. For a molecule having spin multiplicity S, N GVB pairs, and n1, ..., nN orbitals in each pair, orbitals from the initial guess are used in the following manner by the GVB program: • • • The S-1 highest occupied orbitals in the initial guess, which would have been singly occupied in an ROHF calculation, become high-spin orbitals. The next lower N occupied orbitals, which would have been doubly occupied in an ROHF calculation, become the first natural orbitals of the GVB pairs. Any remaining orbitals occupied in the guess stay closed-shell. • • The lowest n1-1 virtual orbitals become natural orbitals 2 through n1 of the first GVB pair, then the next n2-1 orbitals are assigned to pair 2, and so on. The GVBPP scheme does not allow an orbital to be shared by more than one GVB pair. Any remaining (virtual) orbitals from the initial guess become virtual orbitals in the GVB calculation. Generally Guess=Alter is required to ensure that guess occupied orbitals, which will be used as first natural orbitals, match up with the correct guess virtual orbitals which will become the corresponding higher natural orbitals. Often it is helpful to start off with Guess=(Local,Only), examine the orbitals to determine alteration requirements, then do Guess=(Local,Alter) and GVB(NPair=N,Freeze) to allow the higher natural orbitals to become more appropriate. Finally the full calculation can be run with Guess=Read and all orbitals optimized in the GVB. If there is any confusion or concern with the orbitals breaking symmetry, the calculation should be done with Symm=NoSCF and initially with Guess=Local. In fact, this approach is generally recommended except for those very expert users. If the number of orbitals in a pair is negative, the root of the CI to use for that pair and the pair's initial GVB coefficients are read in format (I2,5D15.8). This is useful if a 1Σ or 1Δ state is being represented as a GVB pair of the form x2 ± y2. NPair Gives the number of perfect-pairing pairs. GVB(N) is equivalent to GVB(NPair=N). NPair=0 is acceptable and results in a closed-shell or spin-restricted SCF calculation. InHam=N Read in N Hamiltonians (Fock operators, sets of coupling coefficients). This option may be combined with perfect-pairing pairs. Each Hamiltonian is read using the following syntax (format in parentheses): NO Fj (AJ(I), I =1,NHam) (AK(I), I =1,NHam) # of orbitals in current Hamiltonian (I5) Occup. # (1.0=closed-shell) (D15.8) J coefficients (5D15.8) K coefficients (5D15.8) Combining several orbitals with the same AJ and AK coefficients into one "shell" is not currently supported, so NO is always 1. The ham506 utility can be used to generate averaged Hamiltonians for the common case of spherical averaging in atomic calculations. The Hamiltonian coefficients are described in Bobrowicz and Goddard [105]. A good introduction to the qualitative interpretation of GVB wavefunctions can be found in the review article by Goddard and Harding [502]. OSS Do a two electron, two orthogonal orbital open-shell singlet. This option may be combined with perfect-pairing pairs. OpenShellSinglet is a synonym for OSS. Freeze Freeze closed-shell and open-shell orbitals, and first natural orbitals of GVB pairs, allowing only 2nd and higher orbitals to vary. This option is useful for starting off difficult wavefunctions. Energies, analytic gradients, and numerical frequencies. Here is a GVB(3/6) calculation performed on singlet methylene: # GVB(3)/6-31G(d) Guess=(Local,LowSym,Alter) Pop=Full Test GVB(3) on CH2 molecule specification 1 4 0 2 3 9 2,3 2 2 2 Guess=LowSym input Guess=Alter input GVB input Each of the 3 valence electron pairs is split into a GVB pair. A preliminary Guess=Only calculation was performed to determine the localized orbitals and what alterations would be required. The perfect pairing GVB method includes the effects of intra-pair correlation but not those of inter-pair correlation. Consequently, GVB electrons pairs tend to be localized. In the case of singlet methylene, the carbon lone pair is localized even at the Hartree-Fock level. The canonical Hartree-Fock orbitals for the C-H bonds are delocalized into linear combinations (C-H1 + C-H2) and (C-H1 - C-H2) having A1 and B2 symmetry, respectively. In order to allow the localization in the guess to produce separate bond pairs, these two irreducible representations must be combined. Similarly, the GVB calculation itself must be told not to impose the full molecular symmetry on the orbitals, which would force them to be delocalized. Combining the A1 and B2 representations and combining the A2 and B1 representations causes the calculation to impose only Cs symmetry on the individual orbitals, allowing separate GVB pairs for each bond. Since the resulting pairs for each bond will be equivalent, the resulting overall wavefunction and density will still have C2v symmetry. The Guess=LowSym keyword specifies that the irreducible representations of the molecular point group will be combined in the symmetry information used in a GVB calculation. It takes a single line of input consisting of giving the numbers of the irreducible representations to combine, where the numbers correspond to the order in which the representations are listed in the output file (they appear just after the standard orientation). For example, here is the output for a molecule with C2v symmetry: There There There There are are are are 4 0 1 2 symmetry symmetry symmetry symmetry adapted adapted adapted adapted basis basis basis basis functions functions functions functions of of of of A1 A2 B1 B2 symmetry. symmetry. symmetry. symmetry. Thus for C2v symmetry, the order is A1, A2, B1, B2, referred to in the Guess=LowSym input as 1 through 4, respectively. A zero separates groups of representations to be combined, and a nine ends the list. Thus, to combine A1 with B2 and A2 with B1, thereby lowering the SCF symmetry to Cs, the appropriate input line is: 1 4 0 2 3 9 Since this information always requires exactly one line, no blank line terminates this section. The order of orbitals generated after localization by the initial guess in the first job step was C-1s C-H1 C-H2 C-2s for the occupied orbitals, and C-2p C-H1* C-H2* for the lowest virtual orbitals. Hence if no orbitals are interchanged, the C-2s lone pair would be correctly paired with the unoccupied p-orbital, but then the next lower occupied, C-H2, would be paired with the next higher virtual, C-H1*. So either the two bond occupied orbitals or the two bond virtual orbitals must be exchanged to match up the orbitals properly. Finally, the one line of input to the GVB code indicates that there are 2 natural orbitals in each of the 3 GVB pairs. HF This method keyword requests a Hartree-Fock calculation. Unless explicitly specified, RHF is used for singlets and UHF for higher multiplicities. In the latter case, separate α and β orbitals will be computed [57,58,59]. RHF, ROHF or UHF can also be specified explicitly. SCF single point energy calculations involving basis sets which include diffuse functions should use the SCF=Tight keyword to request tight SCF convergence criteria. Energies, analytic gradients, and analytic frequencies for RHF and UHF and numerical frequencies for ROHF. The Hartree-Fock energy appears in the output as follows: SCF Done: E(RHF) = Convg = S**2 = -74.9646569691 .6164D-03 .0000 A.U. after 4 cycles -V/T = 2.0063 The second and third lines give the SCF convergence limit and the expectation value of S2. Huckel This method keyword requests an extended Hückel calculation [503,504,505,506,507]. ExtendedHuckel is a synonym for this keyword. No basis set keyword should be specified. Hoffmann Requests an Extended Huckel calculation using the default parameter set from the Huckel group. Muller Requests an Extended Huckel calculation using parameters collected by Edgar Muller. Guess Requests an Extended Huckel calculation using the modified parameters used for Guess=Huckel [508,509,510]. Energies, "analytic" gradients and numerical frequencies. The energy appears in the output file as follows (followed by the x, y, and z components of the dipole moment): Huckel eigenvalues -- -1.245 -0.637 -0.558 -0.544 -0.043 Energy= -5.968836513622 NIter= 0. Dipole moment= 0.000000 0.000000 0.000000 0.352 The energy is as defined by this semi-empirical model. Note that energy differences computed from the values in semi-empirical calculations are in Hartrees and may be compared directly with energy differences computed from jobs using other methods. Guess=Huckel INDO Requests a semi-empirical calculation using the INDO Hamiltonian [42]. No basis set keyword should be specified. Energies, "analytic" gradients, and numerical frequencies. The INDO energy appears in the output file as follows (followed by the x, y, and z components of the dipole moment): Energy= -19.034965532835 NIter= 10. Dipole moment= .000000 .000000 -.739540 The energy is as defined by this semi-empirical model. Note that energy differences computed from the values in semi-empirical calculations are in Hartrees and may be compared directly with energy differences computed from jobs using other methods. Integral The Integral keyword modifies the method of computation and use of two-electron integrals and their derivatives. INTEGRATION GRID SELECTION OPTION Grid=grid Specifies the integration grid to be used for numerical integrations. Note that it is very important to use the same grid for all calculations where you intend to compare energies (e.g., computing energy differences, heats of formation, and so on). The parameter to this option is either a grid name keyword or a specific grid specification. If a keyword is chosen, then the option name itself may be optionally omitted (i.e, Integral(Grid=FineGrid) and Integral(FineGrid) are equivalent). "Pruned" grids are grids that have been optimized to use the minimal number of points required to achieve a given level of accuracy. Pruned grids are used by default when available (currently defined for H through Kr). The default grid is a pruned (75,302) grid, having 75 radial shells and 302 angular points per shell, resulting in about 7000 points per atom; the value FineGrid is used to specify this grid. Other grids may be selected by giving an integer value N as the argument to Grid. Grid=UltraFine requests a pruned (99,590) grid. It is recommended for molecules containing lots of tetrahedral centers and for computing very low frequency modes of systems. Other special values for this parameter are CoarseGrid, which requests a pruned version of the (35,110) grid, and SG1Grid, a pruned version of (50,194). Note, however, that the FineGrid has considerably better numerical accuracy and rotational invariance than these grids, and they are not recommended for production calculations [511]. Pass0Grid requests the obsolete pruned (35,110) grid once intended for pass 0 of a tight SCF calculation. Specific grids may be selected by giving an integer value N as the argument to Grid. N may have one of these forms: • • A large positive integer of the form mmmnnn, which requests a grid with mmm radial shells around each atom, and nnn angular points in each shell. The total number of integration points per atom is thus mmm*nnn. For example, to specify the (99,302) grid, use Int(Grid=99302). The valid numbers of angular points are 38, 50 [512], 72 [513], 86, 110 [512], 146, 194, 302 [514], 434 [515], 590, 770, and 974 [516]. If a larger number of angular points is desired, a spherical product grid can be used. A large negative integer of the form -mmmnnn, which requests mmm radial shells around each atom, and a spherical product grid having nnn θ points and 2*nnn φ points in each shell. The total number of integration points per atom is therefore • 2*mmm*nnn2. This form is used to specify the (96,32,64) grid commonly cited in benchmark calculations: Int(Grid=-96032). Note, that any value for nnn is permitted, although small values are silly (values of nnn < 15 produce grids of similar size and inferior performance to the special angular grids requested by the second format above). Large values are expensive. For example, a value of 200100 would use 2*200*100*100 or 4 million points per atom! RELATIVISTIC CALCULATIONS DKH Requests a Douglas-Kroll-Hess 2nd order scalar relativistic calculation [517,518,519,520] (see [521,522] for an overview). This method uses a Gaussian nuclear model [523]. DKH2 and DouglasKrollHess are synonyms. NoDKH and NonRelativistic request a non-relativistic core Hamiltonian, which is the default. DKH0 Requests a Douglas-Kroll-Hess 0th order scalar relativistic calculation RESC Requests a RESC scalar relativistic calculation INTEGRAL FORMAT OPTION Raff Raff requests that the Raffenetti format for the two-electron integrals be used. This is the default. NoRaff demands that the regular integral format be used. It also suppresses the use of Raffenetti integrals during direct CPHF. This affects conventional SCF and both conventional and direct frequency calculations. CNDO Do calculation in main code using CNDO/2 ints. INDO Do calculation in main code using INDO/2 ints. ZINDO1 Do calculation in main code using ZINDO/1 ints. ZINDOS Do calculation in main code using ZINDO/S ints. ALGORITHM SELECTION OPTIONS SSWeights Use the weighting scheme of Scuseria and Stratmann [524] for the numerical integration for DFT calculations. This is the default. BWeights Use the weighting scheme of Becke for numerical integration. NoSComp Turn off symmetry blocking of MO 2-electron integrals. NoSymmComp is a synonym for NoSComp. DPRISM Use the PRISM algorithm [27] for spdf integral derivatives. This is the default. Rys1E Evaluate one-electron integrals using the Rys method [525,526,527], instead of the default method. This is necessary on machines with very limited memory. Rys2E If writing two-electron integrals, use Rys method (L314) [192,525,526,527]. This is slower than the default method, but may be needed for small memory machines and is chosen by default if regular (non-Rafenetti) integrals are requested (by the NoRaff option). Berny Use Berny sp integral derivative and second derivative code (L702). Pass Pass specifies that the integrals be stored in memory via disk, and NoPass disables this. Synonymous with SCF=[No]Pass, which is the recommended usage. Symm NoSymm disables and Symm enables the use of symmetry in the evaluation and storage of integrals (Symm is the default). Synonymous with the keywords Symm=[No]Int, which is the recommended usage. NoSP Do not use the special sp integral program (L311) when writing integrals to disk. RevDagSam Reverse choice of diagonal sampling in Prism. CPKS1Mat Don't use CPKS multiple-matrices code. SquareLoops Forces square loops. SqLoops is a synonym for this option. NoJEngine Forbid use of special Coulomb code. FofCou Use FoFCou even when it would not otherwise be used. NoFoFCou forbid uses of FoFCou. RevRepFock Reverse choice of Scat20 vs. replicated Fock matrices. NoSchwartz Turn off Schwartz cutoffs in FMM/NFx. NoMPCut Turn off MP-based cutoffs in FMM/NFx. NoDFTCut Turn off extra DFT cutoffs. LTrace Trace Linda transactions. SplitSP Split AO S=P shells into separate S and P shells. NoSplitSP is the default. SplitSPDF Split AO S=P=D and S=P=D=F shells into S=P, D, and F. NoSplitSPDF is the default. SplitDBFSP Split density S=P shells into separate S and P shells. NoSplitDBFSP is the default. SplitDBFSPDF Split density S=P=D and S=P=D=F into S=P, D, and F. NoSplitDBFSPDF is the default. NoGather Forbid use of gather/scatter digestion, even when processing small numbers of density matrices. Splatter is a synonym for this option. ForceNuc Do nuclear-electron Coulomb with electron-electron. ECPAcc=N Set ECP accuracy parameter to N. PreComputeXCGridPoints is a synonym for this option.) used for allocation. SepJK Do J and K in HF/hybrid DFT separately for testing. SeqXC Set up for parallel 2 electron integral evaluation but then do not run in parallel (for debugging). etc. UnconAOBasis Uncontract all the primitives in the AO basis. NoDMRange Do not the density matrix in assigning FMM NF/FF ranges.NoSqrtP Turn off use of Sqrt(P) in density-based cutoffs. PCXCP Precompute XC quadrature parameters (number of significant functions. to save the work of recalculating the weights. PCXCGrid Precompute XC quadrature parameters and store both the weight and coordinates for each grid point. NoPreComputeXC is a synonym for this option. but do not store information about individual grid points. By default. NoPCXC Do not precomputed grid information for DFT XC quadrature. Seq2E Set up for parallel 2 electron integral evaluation but then do not run in parallel (for debugging). PCXCWt Precompute XC quadrature parameters and store weights for each point. UncontractAOBasis is a synonym for this option. BigAtoms Make all atom sizes large in XC quadrature. UncontractDensityBasis is a synonym for this option. . PreComputeXCWeights is a synonym for this option. UnconDBF Uncontract all the primitives in the density fitting basis. Sqrt(P) is included in ranges when only Coulomb and not exchange is being computed. PreComputeXCParameters is a synonym for this option. BigShells Make all shell sizes large in XC quadrature. NoXCTest Skip tests of numerical accuracy of XC quadrature. D2EBufSize=N Sets the integral derivative buffer size to N words. LinMIO Convert to linear storage in FoFCou for testing. The default value (which is machinedependant) is generally adequate. The default is to precompute for molecules but not for PBC. Only allowed for single point calculations and Polar=Restart. Useful only in debugging new derivative code. RevDistanceMatrix Reverse choice of whether to precompute distance matrix during numerical quadrature. Both the integral file and checkpoint file must have been preserved from a previous calculation. SCF IOp . NoSymAtGrid Do not use (Abelian) symmetry to reduce grid points on symmetry-unique atoms. WriteD2E Forces the integral derivative file to be written in HF frequency calculations. INTEGRAL FILE-RELATED OPTIONS ReUse Use an existing integral file. BUFFER SIZE OPTIONS IntBufSize=N Sets the integral buffer size to N integer words. They are also documented on our web site: www. Thus the significance of a particular option applies to all the component links in one pass through the overlay. this applies to keyword combinations like Opt Freq and to inherently multi-step methods such as G2 and the CBS methods. Since setting internal options can have arbitrary effects on the calculation.com/iops. The execution of each overlay of Gaussian 03 is controlled by options (numbered from 1 to 50).htm. archiving is disabled by use of this keyword. IOp values explicitly set in the route section are not passed on to the second and subsequent automatically-generated job steps. Overlay 1 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 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 88 89 90 91 92 94 95 96 97 98 101 102 103 104 105 106 107 108 109 110 111 112 113 114 Overlay 2 9 10 11 12 13 14 15 16 17 18 19 20 29 30 40 41 Overlay 3 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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 67 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 Overlay 4 . with 0 being the default.. The value of an option is held unchanged throughout execution of all of the links in one overlay. if you want to specify an alternate grid for a DFT optimization+frequency job. The full list of Gaussian 03 options is given in the Gaussian 03 IOps Reference. Each option may be assigned an integer value. you must use an option to the Int=Grid keyword rather than an explicit IOp value..Ov2/Op2=N2.gaussian. . The syntax is: IOp(Ov1/Op1=N1.) which sets option number Opi to the value Ni for every occurrence of overlay Ovi.The IOp keyword allows the user to set internal options (variables in system common /IOp/) to specific values. For example. 5 6 7 8 9 10 11 13 14 15 16 17 18 19 20 21 22 23 24 25 26 28 29 31 33 34 35 36 37 38 43 44 45 46 47 48 60 61 62 63 64 65 66 67 68 69 71 72 80 81 82 110 Overlay 5 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 35 36 37 38 39 40 41 42 43 45 47 48 49 50 51 52 53 55 56 57 58 59 60 61 62 63 64 65 70 71 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 Overlay 6 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 70 71 72 73 74 75 76 77 78 79 80 81 82 Overlay 7 6 7 8 9 10 11 12 13 14 15 16 18 25 29 30 31 32 40 41 42 43 44 45 52 53 64 65 70 71 72 74 75 76 77 87 Overlay 8 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 27 28 29 30 31 32 35 36 38 39 40 41 42 43 44 45 46 47 Overlay 9 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 25 26 27 28 30 31 36 37 38 40 41 42 43 44 45 46 47 48 49 60 61 62 70 71 72 73 74 75 81 82 83 84 85 86 Overlay 10 5 6 7 8 9 10 11 13 14 15 16 17 18 19 20 21 22 28 29 30 31 32 45 46 47 48 60 61 62 63 72 72 74 75 76 77 78 79 Overlay 11 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 26 27 28 29 30 31 32 33 39 40 41 42 43 45 46 53 60 61 62 63 70 71 75 Overlay 9999 5 6 7 8 9 10 11 12 13 14 15 16 17 18 33 . L109) = Min(40. L109. L107. L114: MAXIMUM NUMBER OF STEPS (OR NUMBER OF STEPS FOR AN LST SCAN). L114: CONVERGENCE ON THE FIRST DERIVATIVE AND ESTIMATED DISPLACEMENT FOR THE OPTIMIZATION RMS FIRST DERIVATIVE . 0 Default value for single-points: 10**-5 in L116. L113.NVAR+10) (L103. 0 NSTEP = Max(20.LT. L109.Overlay 1 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 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 88 89 90 91 92 94 95 96 97 98 101 102 103 104 105 106 107 108 109 110 111 112 113 114 IOp(1/5) L103 MODE OF OPTIMIZATION 0 FIND LOCAL MINIMUM 1 FIND A SADDLE POINT N FIND A STATIONARY POINT ON THE ENERGY SURFACE WITH N NEGATIVE EIGENVALUES OF THE 2ND DERIVATIVE MATRIX L107: MODE OF SEARCH 0 LOCATE THE MAXIMUM IN THE LST PATH. CONFV. 1 SCAN THE LST PATH. CONVX=4*CONVF -1 ConvF = 1/600 HARTREE/BOHR OR RADIAN 0 CONVF = 0.NVAR+10) (L102. L112. L113. L112. . L105. L105. IOp(1/6) L102. 10**-7 in L117.0003 HARTREE/BOHR OR RADIAN N CONVF = N*10**-6 L116. L114) N NSTEP = N IOp(1/7) L103. RMS EST. L117: Convergence on electric field/charges -1 Default value for optimizations: 10**-7.NVar+20) (L113. N 10**-N. DISPLACEMENT .LT. L112) = Min(20. L103. L105. = 0. scale for transition states. 4000 Apprx. for the grad RHO or Vne trajectories.3 Bohr or Radian (L103. I . 00 Whether to scale or search the sphere when reducing the step size to the trust radius (Default search for minima. Use only the Center of NuclearCharge 4 Use Interlocking Spheres N0 Order of Adam's-Bashforth-Moulton (ABM) predictor-corrector method to use in solving diff. L121: Time step. N DXMAXT = 0. 20000 Use density. default 0.2 Bohr or Radian (L105). 1 2 No. IV . 3000 Apprx.Do III and Allow Surface To "Relax" in Solution if no spheres N0000 Whether to evaluate densities using orbitals or density matrix.). . 10000 Use MOs. Read or CalcFC). Default is III for Tomasi (interlocking spheres) and IV for general surface. = 0. Default is to use density. N000 Which approximation to make. L109. 0 DXMAXT = 0.0001 fs. 0 Which type of basin to use to partition the density isosurface. But No Compensation. Default is 4.1 BOHR OR RADIAN (L103. 0 Whether to update trust radius (DXMaxT. III .IOp(1/8) L103. L112: MAXIMUM STEP SIZE ALLOWED DURING OPT.1 IOp(1/9) L103: Use of Trust radius. no for TS. eqns.Don't Do Self-Polarization or "Compensation" 2000 Apprx.3 Bohr or Radian (L113. default Yes). N00 Number of small steps per ABM step to be used in starting ABM and when "slow down" is needed in ABM. L114). Estm or UnitFC). 1000 Apprx. Yes. II . Default is 5.Do Self-Polarization and Compensation. max is 9.Do-Self Polarization. Default is Yes for minima. = 0. Default is 4 1 GradVne 2 GradRho 3 Don't Use Basins.01 * N L117: General control. N*0. J=1. .6) (L103 only).FC(I. L117: Whether to delete points which are too close together: 0 1 No Yes. L109. How close to get to the isosurface in search.10 20 Scale.0) (L103 only). End with a blank card.J). L112. Read ((FC(I.F20. 0 1 YES.I).0**-N L121: Whether to read in initial velocities: 0 1 2 3 Default (same as 1) Generate random initial velocity Read in initial cartesian velocity (Bohr/sec) Read in initial MW cartesian velocity (sqrt(amu)*Bohr/sec) IOp(1/10) L103. L107: WHETHER TO MAINTAIN SYMMETRY ALONG THE SEARCH PATH.0D-6 (N=20) 2. Bohr.J. 0 N Approx 1. 0 1 2 3 Use defaults (not valid for L109). and radians).J). NO.NVAR) (8F10. L113.05 Angstroms) -N Yes. (5I3. using a (10**-N Angstroms) criteria. L105.I=1. Read from checkpoint file in internal coordinates. using a default criteria (0. Search. Read I. L114: Input of initial Hessian: All values must be in atomic units (Hartree. Use unit matrix throughout. (not valid for L109).8) from input stream.010 1. 6 Read cartesian forces followed by cartesian force constants (both in format 6F12. only recognized by 103). 1 2 DON'T TEST. TEST. 0 N 0.000 + N*(0. 0 DEFAULT (TEST for z-matrix or cartesian TS but not for LST/QST or for minimum). L117: Scaling Factor for Determining Overlaps of VDW atoms -1 0 N Turn off scaling Default is 1.1/N IOp(1/12) .001) Step size for ABM method in Trudge for isodensity method. IOp(1/11) L103: TEST OF CURVATURE.05 (N=2) 0. Use unit matrix (default for L105.4 Second derivative matrix calculated analytically. Estimate force constants using valence force field. 7 8 9 10 Use semiempirical force constants. 5 Read cartesian forces and force constants from the checkpoint file are convert to internal coordinates. BOMB THE JOB IF THE SECOND SECOND DERIVATIVE MATRIX HAS THE WRONG NUMBER OF NEGATIVE EIGENVALUES. followed by a blank line. L114. 0 Default (0 for TS. only in L103 and L115).L113. no quadratic step). number of bad steps to allow before attempting a linear minimization (i. D2Corr (Old.L103: OPTIMIZATION CONTROL PARAMETERS 0 1 USE DEFAULT VALUES READ IN NEW VALUES FOR ALL PARAMETERS (SEE INITBS) IOp(1/13) L103. 1 2 3 4 5 6 7 8 9 Powell (not in L103). otherwise BFGS). BFGS (not in L103) BFGS. 1 for minima). Powell for L113 and L114). only in L103 and L115). D2Corr (BFGS) D2Corr (Bofill Powell+MS for transition states). 7 for L103 TS. L121: Multi-time step parameter (NDtrC. D2Corr (No update. use initial Hessian).NDtrP) 0 NN No multi-time stepping Iterate density constraints NN times per step MM00 Do gradient once every MM steps IOp(1/14) L103: Max.. D2Corr and L115. safeguarding positive definateness (not inL103 or L115) D2Corr (New. D2Corr (New if energy rises.L115: Type of Hessian Update: 0 Default (9 for L103 minimization.e. . L114: MINIMUM ALLOWABLE MAGNITUDE OF THE EIGENVALUES OF THE SECOND DERIVATIVE MATRIX.N Allow N -. Do full optimization in redundant internal coord. THE SIZE OF THE EIGENVALUE IS REDUCED TO THE MAXIMUM. . IOp(1/15) L103. AND PROCESSING CONTINUES. SIMMILAR TO IOp(16) 0 N EIGMIN = 0. 10 20 30 40 Do optimization in cartesian coordinates.1 * N IOp(1/17) L103.1 * N IOp(1/16) L103.L113.L113.0001 EIGMIN = 1. IF THE LIMIT IS EXCEEDED. / N IOp(1/18) L103: Coordinate system. Do optimization in Z-matrix coordinates.0 HARTREE / BOHR**2 OR RADIAN**2 EIGMAX = 0.L114: MAXIMUM ALLOWABLE MAGNITUDE OF THE EIGENVALUESOF THE SECOND DERIVATIVE MATRIX.L109: ABORT IF DERIVATIVES TOO LARGE -1 or 0 N No force test at all. No optimization will occur. FMAXT = 0. Do full optimization in pruned distance matrix coords. 0 N EIGMAX = 25. 0 Proceed normally 1 Second derivatives will be computed as directed on the variable definition cards.linear only starts with the N+1st. LINEAR AND STEEPEST DESCENT. Linear as usual. Quadratic if curvature is correct. Read the AddRedundant input section for each structure. in L103.50 100 1000 2000 3000 Do full optimization in redundant internal coords with large molecular tools. Quadratic if curvature is correct. STEEPEST DESCENT AND LINEAR ONLY WHEN ESSENTIAL. 4000000 Microiterations for pure MM. Default (2000000). IOp(1/19) L103: SEARCH SELECTION 0 2 3 4 5 6 7 Default (same as 6). RFO if not. RFO without linear. RFO and linear. No linear search. . 1000000 Skip MM atoms in internal coordinate definitions and do microiterations the old way. RFO if not. done in L402. 2000000 Include MM atoms in internal coordinate definitions (no microiterations). Do not define H-bonds Define H-bonds with no related coordinates (default) Define H-bonds and related coordinates 10000 Reduce the number of redundant internals 20000 Define all redundant internals 100000 0000000 Old definition of redundant internals. in L120. 3000000 Skip MM atoms in internal coordinate definitions and do microiterations the new way. EIGMAX AND EIGMIN. 0 1 Yes. First-order simultaneous optimization. Newton-Raphson only. 0 NORMAL MODE. .L114: Search Selection: 0 P-RFO OR RFO STEP ONLY (DEFAULT) 1 P-RFO OR RFO STEP FOR "WRONG" HESSIAN OTHERWISE NEWTONRAPHSON IOp(1/20) L101.L113. Assume reactant order equals product order. DXMAXT. 1 EXPERT MODE: CERTAIN CUTOFFS USED TO CONTROL THE OPTIMIZATION WILL BE RELAXED. IOp(1/22) L107: Whether to reorder coordinates for maximum coincidence.8 9 10 11 13 Newton-Raphson and linear. L109. THESE INCLUDE FMAXT. L106. L110: INPUT UNITS 0 1 2 3 ANGSTROMS DEGREES BOHRS DEGREES ANGSTROMS RADIANS BOHRS RADIANS IOp(1/21) L103. GDIIS and linear GDIIS only.L114: EXPERT SWITCH. L113. L108. Yes. VECTOR BACKWARD DIRECTION AND GENERATE S. VECTOR FORWARD DIRECTION AND READ S. VECTOR BOTH DIRECTIONS AND GENERATE S. No.001 degree. 0 1 2 Energy only. round angles within 0. 0 1 2 Default (Yes). VECTOR 8F10.2 Read in a re-ordering vector from the input. VECTOR 8F10.6 BOTH DIRECTIONS AND READ S. . L115: KIND OF SEARCH: 0 1 2 3 4 5 6 7 BOTH DIRECTIONS AND GENERATE SEARCH VECTOR FORWARD DIRECTION AND GENERATE S. VECTOR 8F10. VECTOR 8F10. IOp(1/25) Wether SCRF is used with numerical polarizability: 0 No.6 IOp(1/23) L112: Derivative availability. Energy + Forces + Force constants IOp(1/24) Whether to round tetrahedral angles.6 BACKWARD DIRECTION AND READ S.6 FORWARD DIRECTION AND READ S. Energy + Forces. 1 Input one structure. Gradient 10**(MM). K = 1-9 Interpolation of initial guess of TS between R and P (TS=0. either initial guess of the minimizing structure or transition structure without QST. gradient 1. The union of the two redundant coordinates are taken as the redundant coords for the TS. the field in /Gen/ must be cleared each time. gradient 1.d-7). L= 2 Input 2 structures. The first one is reactant the second one is the product. default J=5) J=1 J=2 J=3 J=4 LST constraint in internals QST constraint in internals LST constraint in distance matrix space QST constraint in distance matrix space .d-4) Fine grid accuracy for DFT (Energy 1. The third one is the initial guess of the transition structure.e. Standard grid accuracy for DFT (Energy 1. IOp(1/26) Accuracy of function being optimized: -NNMM Energy 10**-(NN). -1 0 1 2 3 Read in values Default (same as 1). 1000*I+100*J+10*K+L) Transition state searching using QST and redundant internal coordinates L= 0. R and P are used to guide the QST optimization of the TS. The values of the TS coord are estimated by interpolating the sturcture of R and P.1 Yes. the second one is the product.1*(10J)*P. R and P are used to guide the QST optimization of the TS. L= 3 Input 3 structures. Normal accuracy for HF (energy and gradient both 1.d-5.1*J*R + 0.d-6) IOp(1/27) = IJKL (i.d-7. the first one is the reactant. Default. same as -1. Default (same as 100).05) IOp(1/28) L103: Number of translations and rotations to remove during redundant coordinate transformations: -2 -1 0 N 0. default QSTrad = 0.g. By model builder. model A. . Abort job if model builder generates a z-matrix with too many variables. By model builder. IOp(1/29) L101: SPECIFICATION OF NUCLEAR CENTERS 0 1 2 3 4 5 BY Z-MATRIX BY DIRECT COORDINATE INPUT (must set IOp(29) in L202). 7 10 000 100 200 1000 Get all input (title. 6 Get Z-matrix from the checkpoint file. Normal (6 or 5 for linear molecules). QSTRad = 0.I = 0-9 Control parameters for climbing phase of QST (e. Read optimization flags in format 50L1 after the z-matrix. GET Z-MATRIX AND VARIABLES FROM THE CHECKPOINT FILE. Do not abort job if model builder generates a z-matrix with too many variables. Print details of the model building process. GET CARTESIAN COORDINATES ONLY FROM THE CHECKPOINT FILE. N. structure) from the checkpoint file. model B. charge and multiplicity. but read new values for some variables from the input stream.01*I. Purge flags except the frozen variables. same as 10000. IOp(1/33) L101: L102 L103 L106 L109 L110 L113 L114 0 1 OFF ON DEBUG PRINT . 100000 Generate a symbolic z-matrix using all Cartesians if none is present on the checkpoint file (a hack to make IRCs work with Cartesian input). It also suppresses error termination on large gradients. 200000 Same as one. 0 1 DON'T PUNCH. Do not retain symbolic constants.2000 3000 4000 5000 00000 10000 20000 Set all optimization flags to optimize. PUNCH. 0 1 YES NO IOp(1/32) TITLE CARD PUNCH CONTROL. Rebuild the coordinate system. but retain the redudant internal coordinate definitions. IOp(1/30) L103: ARE THE READ-WRITE FILES TO BE UPDATED? THIS OPTION IS SET FOR THE LAST CALL TO 103 IN FREQUENCY CALCULATIONS IN ORDER TO PRESERVE THE VALUES OF THE VARIABLES FOR ARCHIVING. (2+3) Purge all flags but keep the coordinate definition. Mark Z-matrix constants as frozen variables rather than wired-in constants. Default. 0 1 NORMAL CHECKPOINT OF OPTIMIZATION. IOp(1/38) Entry control option (currently only by L106. L103. L108. IOp(1/36) CHECKPOINT. 1 FIRST POINT OF A RESTART. and L112 but not L102. 0 NORMAL OPTIMIZATION. ET. L109. . GET GEOMETRY. Initial entry. FROM THE CHECKPOINT FILE. N>1 . In L103: Initial entry of guided optimization using N levels. and L105). 1 THIS IS THE LAST POINT AT WHICH ANALYTIC SECOND DERIVATIVES WILL BE DONE.IOp(1/34) L101 L102 L103: DEBUG + DUMP PRINT 0 1 OFF ON IOp(1/35) RESTART (L102-L112). L107. 0 1 Continuation of run. SUPPRESS CHECKPOINTING. WAVEFUNCTION. L111. DELETE THE D2E FILE AND THE BUCKETS AND TRUNCATE THE READ/WRITE FILES. IOp(1/37) D2E CLEANUP (obsolete) 0 NO CLEANUP. L110. 001 au in L111). L111).0. N Use step-size of 0. 0. (L106.01 Angstrom in L110. Read from checkpoint file. L110. IOp(1/40) L113. The default. 0.N0 In L106: differentiate Nth derivatives once. Path step size in L115. Just update. Recalculation the Hessian every N steps. 1 for others). IOp(1/41) Step number of optimization from which to take geometry. In L106: differentiate wrt electric field. 000 100 200 In L106: differentiate wrt nuclear coordinates. electric field au in L111). IOp(1/39) Step size control for numerical differentiation.0001*N (angstroms in L106. -1 for the initial geometry . Read from input stream. L109. 0. In L106: differentiate wrt field and nuclear.0001*N atomic units everywhere. L114: Hessian recalculation. -1 Read stepsize (up to 2 for L106.001 Angstroms in L106.005 A in L109. -1 0 N Pick up analytic second derivatives every time. L110. 0 Use internal default (0. free-format. L109. -N>1 Use step-size of 0. In L110 and L111: differentiate energy N times. execpt for CalcAll. L116: Whether to read initial E-field: 0 1 2 Start with 0. Rad) or (Symbol. terminated by a blank line.IOp(1/42) L103. Initial radius of spheres to be placed around attractors to "capture" the gradient trajectories. centers will be on atoms Read-in centers and radii on cards Force Merz-Kollman radii (Default) Force CHELP (Francl) recommended radii.Rad). L117: Cutoff to be used in evaluating densities. Force CHELPG (Breneman) recommended radii.0D-10 1. terminated by a blank line.Rad). Default is 6. 200 Read in replacment radii for selected atoms as pairs (I.1 N. 0 N 1.0D-N IOp(1/43) L116: Extent of Reaction Field. 0 NM 0. 0 1 2 3 Dipole Quadrupole Octapole Hexadecapole L117: How to define Radii 0 1 2 10 20 30 Default is 11 Use internally stored Radii. 100 Read in replacement radii for selected atom types as pairs (IAn.M = NM/10 . L115: Number of points along the reaction path in each direction. The final radius is then automatically optimized separately for each atom. IOp(1/46) Order of multipoles in numerical SCRF: 0 1 2 Dipole Quadrupole Octapole . This parameter should be chosen with the parameter Cont in mind 0 NM 10. 0 1 Do not read isotopes.used in Trudge only to eliminate. Read Isotopes. from the outset.0 au N.M au = NM/10 L121: Seed for random number generator (ISeed) -1 0 N Use system time initialize iseed (Note each run will give different results) Use default seed value (ISeed = 398465) Set random number seed to N IOp(1/45) Read isotopes in L115. Mass-weighted.IOp(1/44) IRC Type 0 1 2 3 Default (same as 3). Cartesian. L117: Maximum distance between a nucleus and its portion of the isosurface . points which clearly lie in another basin. Internal. No electrostatics included in the model systems . 0 N Default: 50000 N.3 Hexadecapole. Default: 1 2/1 unimolecular/bimolecular reaction. CIOp(1/49) Options to IRC path relaxation (IJKL) L K J I 2/1 dont/do optimize reactant structure. IOp(1/47) Number of redundant internal coordinates to allow for. IOp(1/48) IRCMax control. Default: 1 3/2/1 dont/QST3/QST2 optimize TS structure (for QST input). Default: unimolecular IOp(1/52) L101 and L120: Type of ONIOM calculation: 0/1 2 3 00 10 20 One layer. normal calculation Two layers Three layers Default (20) Include electrostatics in model systems using MM charges. Default: 1 2/1 dont/do optimize product structure. 1 20 Do IRCMax Include zero-point energy. IOp(1/54) .100 Do full square for testing. gradients. energy. Integrate energy Integrate energy and gradient Integrate energy. and hessian Restore point MM from RWF but do not create a new model system. NN0 Save necessary information (some rwf's. IOp(1/53) L120: Action of each invocation of L120: 0 1 Do nothing Set up point MM on rwf from initial data 2 Set up point MM on rwf from initial data and restore point MM on chk file if ONIOM data is present there. gradient. MM000 Calc Level High ||| Mid ||| Low SML 1--3--6 system size 2--5--8 4--7--9* Next point to do is MM. The default is MK charges. 3 4 5 6 7 Restore point M from data on the rwf. NN = MaxLev**2 + 1 (currently 17) to restore real system. N000 Use atomic charge type N-1 during microiterations. hessian) of point NN of the ONIOM grid. Yes. read from input stream Yes. L115: IRC optimization. Read modifications. otherwise 3). Yes. Amber allowing any symbol. No. Use displacements to find the next geometry. read from rwf file. generate connectivity. IOp(1/56) Set of atom type names to parse: 0 1 2 3 Accept any. . for use with parameters in input stream. Yes. IOp(1/55) L103: Options for GDIIS: ICos*1000+IChkC*100+IMix*10+Method form. use gradients to find the next geometry. read from checkpoint file.Whether to recover initial energy during IRCMax from chk file: 0 1 No. Yes. IOp(1/57) Whether to produce connectivity: 0 1 2 3 4 5 10 Default (4 if reading geom from chk file and connectivity is there. Amber. 0 1 Default. Dreiding/UFF. No. Normal: leave the rwf set up for the low-level calculation on the real system. NN < 50. IOp(1/63) . IOp(1/60) Interpret extra integer and fp values in z-matrix as scan information. NN NN fragments. but with NBasis and NBsUse for the high-level calc on the model system. Old version. IOp(1/59) Update of coordinates in L103 0 1 2 Default (1 for large opt. 2 MOMM: leave the rwf set up for the real system. 0 1 2 Default (No). IOp(1/62) Counterpoise control. including dummy atoms. 2 for regular) New versions. IOp(1/61) How ONIOM should leave the rwf at the end of each geomtry: 0 1 Default (1). Useful for treating the full system as having electrons only on the QM atoms. IOp(1/58) IRCMax control in L115.100 Connectivity input is in terms of z-matrix entries. Yes. Dreiding. AMBER. Use only soft. 10000 Use the first when there are equivalent matches. MM2 (NYI). 0000 Do not read modifications to parameter set. MM3 (NYI). MMFF (NYI). Use only hard-wired. Quartic fitting field (NYI). Use soft and hard-wired.lone fragments IOp(1/64) Molecular mechanics force field selection: 0 1 2 3 4 5 6 7 000 100 200 300 None. 00000 With soft parameters.Step in counterpoise calculation: MNN NN = 0 1-NFrag M = order of derivatives (1=Energy. abort when different parameters match to the same degree. hard-wired has priority. . Lowest 2 digits then have no meaning. Use soft and hard-wired. soft has priority. 1000 Read modifications to parameter set. UFF.2*NFrag -. 2=Gradient. Supermolecule Fragments with ghost atoms NFrag+1 . IOp(1/67) Source of MM parameters. Default (100) Do for atoms which have charge specified or defaulted to 0. IOp(1/65) Control of which terms are included in MM. or to use the values already there. If IOp(67)=3. then the default is to apply soft parameters with higher priority. Do for all atoms regardless of typing. 0 1 10 100 1000 10000 Do all (default) Non-bonded Stretching Bending Torsion Out-of-plane IOp(1/66) Whether to generate QEQ charges. corresponding to the 'classes' in FncInf. 1==> 221) Do QEq. Don't do QEq.20000 Use the last when there are equivalent matches. Default (10) Do for atoms which were not explicitly typed. 0 1 2 00 10 20 000 100 200 Default (2. . Do for all atoms regardless of initial charge. over-written the values in AtChMM. Copy from chk file. Pick up non-standard parameters from chk file. L121: Lagrangian constrain method for ADMP (ICType) Half*Gamma*Tr[(P*P-P)**2] + Lambda*[Tr(P)-Ne] + Eta*Tr(P*P-P) 0 Default Same as 7 if no Mass-Weighting (IOp(76) < 0) Same as 10 if MassWeighting (IOp(76) > 0) 1 2 3 Use Lambda and Eta only. Gamma. (Gamma=0) Use Lambda. IOp(1/70) L118 Type of sampling (Nact) 0 1 2 3 4 Defalt (same as 3) Orthant sampling Microcanonical normal mode sampling Fixed normal mode energy Local mode sampling ( now only Nact = 0 or 3 OK ) IOp(1/71) Whether to print out input files for each structure along an IRC: 0 1 No. Eta. IOp(1/72) L103: Algorithm choice for microiterations. Gamma = .2 Use Lambda. else 1. Yes. Constraints for scalar Mass case: . Generate here. Gamma.0 1 2 3 Default: 2 if reading geom from chk file. reading from input if requested by IOp(64). Eta. Gamma = 1. IOp(1/73) L103: NInit for microiterations.6 5: 0. The default is 500. etc.0 1: 0.0 IOp(1/75) ADMP control flag (ICntrl) 0 1 2 3 Standard ADMP Read converged density at every step Fix the nuclear coordinates Test time reversability (MaxStp must be even) .0 2: 0.2. Constraints for tensorial Mass: 8-11 Mass-weighting constraints. Documentation maybe found in DVelV1.4 Use exact constraint Sum(ij)[Vij*(P**2-P)ij] 5-7 Iterative Scheme same as 4. Different initial guesses. L121: Initial Kinetic energy of the Nuclei (EStrtC) 0 Default (. all layers are scaled by at least as much as ones farther out. IJKLMN 6th through 1st nearest neighbors of current layer scaled by I*0. J*0. 10 is default. 0 ==> 5 (no scaling).4 4: 0. The actual factors used are: 0: 1.0 Hartree IOp(1/74) Charge scaling for charge embedding in ONIOM.2. M L0 Factor for charges one bond away from link atom Factor for charges two bonds away from link atom K00 Factor for charges three bonds away from link atom IJ etc.8 6-9: 1.1 Hartree) N>0 N*micro-Hartree N<0 0. 7 is default for scalar mass case.2 3: 0. Alpha and Beta each get half this energy) and Option Number to compute initial kinetic energy.0001 AMU. full elsewhere . If IOp(76)<0.0001 AMU BoxMas=0 Box coordinates not propagated (default).00 10 20 Default (20). currently same as N=0 above) IOp(1/78) Sparse in L121 -N Sparse here with cutoff 10**(-N).e.0 Hartree) N>0 N*micro-Hartree IWType is used to figure out how the initial velocity is is computed (in gnvelp). XXXX If PBC: Mass of Box Coordinates (BoxMas) = XXXX*.0001 AMU MW core functions more than valence functions. Do not read stopping parameters. Alpha and Beta each get half this energy) 0 Default (0. IOp(76)<0 YYYY*.XXXXZYYYY = Ficticous electron mass (EMass) YYYY Default (1000) IOp(76)>0 YYYY*.. Format of Input: XXYYYY (six digits) IWType = XX N = YYYY (For UHF. IOp(1/77) Initial Kinetic energy of the density matrix (EStrtP) (For UHF. Use uniform scaling for all basis functions (Note YYYY > 9999 makes no sense) Z Mass-weighting option. If XXYYYY < 0 : Initial velocity = 0. Read stopping parameters from input. IOp(1/76) +/. Z is meaningless.0 Hartee (i. but only for the first XXXXX simulation steps. 1000000 Velocity scaling. IOp(1/81) Nuclear KE thermostat in ADMP -. 11XXXXX Velocity scaling. Yes. IOp(1/82) Temperature for nuclear KE thermostat in L121. IOp(1/84) Differentiation of frequency-dependent properties.1 to surpress the 5th order correction after surface hop has been made in Trajectory Surface Hopping calculations. Needs also IOp(10/80=1) Nuclear Kinetic Energy Thermostat Option. No. L118: .eq. 0 1 2 Default (No). IOp(1/80) L106: 0/1/2 Cartesian/Normal mode/Internal coordinate differentiation. Use sparse fixed form IOp(1/79) IRCMax convergence in L115 Stopping criteria in L118 and L121.0 1 Use full matrices or spase based on standard settings. IOp(1/83) Whether to read in frequencies for electric and magnetic perturbations. (Currently only Velocity scaling is implemented) 0 No Thermostat. 2 is NYI. if thermostating in only required during equilibration. . all the way through the simulation.temperate is checked and scaled every IOp(81) steps. (This options is useful. Diagonalizae Fock matrix to get band gap.etc: 0: SCF 1: MP first order . Print eigenvectors as well. Integrate densities specified by following digits: Density to use from gridpoint 1 Density to use from gridpoint 2 M000 etc. No. IOp(1/87) ONIOM integration of density.0 N No. IOp(1/85) Band gap calculation in PBC ADMP: 0 1 2 Default (No). K. 0 1 2 Default (1). etc.M.L. Mask for which properties on file 721 will be differentiated. Integrate current densities. Print tensors and eigenvalues. IOp(1/86) Printing for NMR for ONIOM. evolution. 0 1 2 K0 L00 Do not integrate. IOp(1/91) Thermostat Option. Average energy (in microhartree) to be maintained during Simulation. 2 Read isotopes from input. IOp(1/89) Maximum allowed deviation from average nuclear KE during ADMP. in Kelvin. Read isotopes from chk. IOp(1/92) . IOp(1/90) To read in the velocity in cartesian coordinates Nuclear Kinetic Energy Thermostat Option. 3 4 Read isotopes from rwf. for backwards compatibility. as required by IOp(80). 4 if geometry read from chk) Use most abundant isotopes. The temperature and pressure are read first.2: MP2 3: MP3 4: MP4 5: CI one-particle 6: CI 7: QCI/CC 8: Correct to second order IOp(1/88) Whether to read in atomic masses (isotopes): 0 1 Default (1 if geometry read from input. direct /w MM Hessian incore. direct /w raw MM Hessian incore. 96. Quadratic micro-iterations. full diagonalization. IOp(1/101. N4 IOp(1/105) . 98) IOp(94): IOp(95): IOp(96): Davidson control for quadratic micro-iterations (see MMOpt2) RFO/Davidson control for quadratic micro-iterations (see MMOpt2) Davidson control for coupled QM/MM macro step (see MMOpt2) IOp(97): RFO/Davidson control for coupled QM/MM macro step (see MMOpt2) IOp(98): Control of quadratic micro-iterations and coupled QM/MM quadratic macro step. 95. Coupled macro step.Maximum allowed deviation from average nuclear KE specified in IOp(81). Also in microhartree. fully direct. direct /w prepared Hessian incore. N3. 104) Phase control in L115 and L118: N1. 97. <0 0 1 2 3 4 5 10 20 30 40 50 Do not use dynamic convergence criteria for the micro-iterations. IOp(1/94. fully direct. Coupled macro step. 102. Quadratic micro-iterations. direct /w full Hessian incore. Quadratic micro-iterations. Quadratic micro-iterations. Default(15). N2. Coupled macro step. Regular non-coupled macro step. full diagonalization. Regular micro-iterations. Coupled macro step. 103. 0001 IOp(1/107) Error tolerance for DVV time step correction (Error) .04 (N=400) v0 is set to N*0.Reaction direction 00 10 Default (Same as 10) Forward direction 20 Reverse direction Damped-Velocity Verlet (DVV) options for Dynamic Reaction Path Following 0 1 2 00 10 20 000 100 200 300 0000 1000 2000 Default (Same as 2) Use DVV Do not use DVV Default (Same as 10) Follow the rxn path in the forward direction Follow the rxn path in the reverse direction Default (Same as 200) Time step correction not used Time step correction used but not to recalculate current DVV step Time step correction used and current DVV step recalculated Default (Same as 1000) Use DVV stopping criteria Do NOT use DVV stopping criteria IOp(1/106) Damping constant for DVV Dynamic Rxn Path following (v0) 0 N Default v0=0. 0001 IOp(1/108) Gradient magnitude for DVV stopping criteria (Crit1) 0 N Default (N=15) N*0.003 (N=30) Error=N*0. Effectively disables the scaling Default (50) Scale up or down to maximum change in a variable of N/1000 IOp(1/112) .0 N Default Error=0.0001 IOp(1/109) Force-Velocity angle for DVV stopping criteria (Crit2) 0 N Default (90 Degrees) Use N Degrees IOp(1/110) Scaling of rigid fragment steps during microiterations. 0 1 2 -n Do not scale Scale with 1/NRA (NRA = number of atoms in fragment) Scale with 1/Sqrt(NRA) Scale with 1/n IOp(1/111) Step-size to use with steepest descent when L103 is having trouble: -N -1 0 N Scale up to RMS step of N/1000 if DXRMS is less. 0 N Default (1 unless specified by IOp in overlay 7 or read in). Do not print the angle matrix. unless read in). 0 N Default (1 atomosphere. Default: same as 20.Temperature for thermochemistry. . IOp(1/114) Scale factor for harmonic frequencies for use in thermochemistry and harmonic vibration-rotation analysis. IOp(1/113) Pressure for thermochemistry. Overlay 2 9 10 11 12 13 14 15 16 17 18 19 20 29 30 40 41 IOp(2/9) Printing of distance and angle matrices. Print distance matrix. N/1000 atmospheres. using z-matrix connectivity if possible. N/1000000. Print the angle matrix. unless read in). 0 1 2 00 10 20 30 Default: same as 2. N/1000 degrees. Do not print the distance matrix. 0 N Default (standard temperature. Use cutoffs instead of the z-matrix for determining which angles to print. 0 1 2 Default (print if 50 atoms or less) Print Don't print IOp(2/12) CROWDING ABORT CONTROL 0 1 2 3 Default (same as 1). using the z-matrix for connectivity info. Do not abort the run regardless of 0 distances. using a distance cutoff for connectivity info. Print dihedral angles. Print dihedral angles. 0000 Default: print only for small cases 1000 Do not print the cartesian coordinates in the input orientation 2000 Do print the cartesian coordinates in the input orientation IOp(2/10) TETRAHEDRAL ANGLE FIXING 0 1 2 Default (don't test). .000 100 200 300 Default: same as 100. DO NOT TEST FOR SUCH ANGLES. ANGLES WITHIN 0.471 WILL BE SET TO ACOS(-1/3).001 DEGREE OF 109. Do not print dihedral angles. Abort the run for zero atomic distances only Abort the run if any atoms are within 0.5 A. IOp(2/11) PRINTING OF Z-MATRIX AND RESULTANT COORDINATES. 12) IOp(2/14) Internal coordinate linear independance. -1 Turns on symmetry. ATOMS WILL NOT TAKE THE ATOMIC NUMBERS. SO THEY ARE NOT PUNCHED. for PBC. 10 Compute nuclear forces as well as second derivatives for the test. but is useful for debugging the derivative transformation routines.e. IN 'ATOMS' FORMAT (3E20.3E20. same as 0 for molecules but turns on assignment of space group ops. test the variables for linear independance and abort the job if they are dependant. THE ATOMIC NUMBERS AND COORDINATES ARE PUNCHED IN FORMAT (I2. Perform the test. Do not perform the test. 2 YES.. but do not abort the job. . 0 Leave symmetry in whatever state it is presently in. IOp(2/15) SYMMETRY CONTROL.IOp(2/13) PUNCH COORDINATES.12). 0 1 2 Default (same as 2). 3 If internal coordinates are in use. NOTE. 0 NO 1 YES. this is not a full optimization). IN FORMAT SUITABLE FOR COORD INPUT TO Gaussian. This is not correct for the linear independance check. 100 Abort the job if the number of z-matrix variables is not exactly the number of degrees of freedom (i. then confirm symmetry with tight cutoffs. IOp(2/16) action taken if the point group changes during an optimization. IOp(2/17) Tolerance for distance comparisons in symmetry determination.d-8). .1 Unconditionally turn symmetry off. 3 Don't even call symm. 10**-N. symmetrize the resulting coordinates. Keep going. the molecule is not oriented. 5 Recover the previous symmetry operations from the rwf. currently 1. 4 Call Symm once with loose cutoffs. Turn on symmetry operations for PBC. 10**N. Default (10) Do re-orientation for PBC. use same tolerance for orientation. using the new symmetry. 0 N>0 N<0 Default (determined in the symmetry package. Suppress re-orientation for PBC. 0 1 2 3 Abort the job. but get symmetry info from the chk. but then disable further use of symmetry. Keep going and leave symmetry on. and will determine the framework group. 6 00 10 20 100 Same as 5. However. using the old symmetry. Note that symm is still called. 2 Bring the molecule to a symmetry orientation. and confirm that the new structure has the same symmetry. Keep going and leave symmetry on. Yes. but remove z-matrix. currently 1.d-7). IOp(2/29) Update of coordinates from current Z-matrix. as Hollerith string. 0 N>0 N<0 Default (determined in the symmetry package. format (50I2) IOp(2/40) Save (initial) structure and possible constraints in rwf 698: 0 1 Default (No). IOp(2/20) Number (1-3 for X-Z) of axis to help specify which subgroup of the type specified in IOp(19) to use. use same tolerance for orientation. 10**N. 0 1 2 3 Default (1) No. . Yes. 10**-N. IOp(2/30) Read in vector of atom types (for debugging).IOp(2/18) Tolerance for non-distance comparisons in symmetry determination. 0 1 No Yes. Yes. IOp(2/19) Largest allowed point group. The same numbers are used for all basis sets.52111) for second ATOMS. Read in structure from input stream. N/1000 Hartree/Bohr**2 Overlay 3 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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 67 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 IOp(3/5) TYPE OF BASIS SET. MacLean-Chandler (12s. 6 LANL ECP basis sets. IOp(2/41) Force constants for Harmonic constraints. (VARIOUS IMPLEMENTATIONS MAY OMIT SECOND ROW ATOMS.6-31G MINIMAL STO-NG (VALENCE FUNCTIONS ONLY) EXTENDED LP-N1G (VALENCE BASIS FOR CORELESS HARTREE-FOCK PSEUDOPOTENTIALS) EXTENDED 6-311G (UMP2 FROZEN CORE OPTIMIZED) BASIS for first row. SPLIT VALENCE N-21G (OR NN-21G) BASIS FOR FIRST OR SECOND ROW MINIMAL STO-2G TO STO-6G EXTENDED 4-31G. . IOp(6) selects options. 7 GENERAL--SEE ROUTINE GenBas FOR INPUT INSTRUCTIONS. Default (no constraint unless reading constraint from chk file).5-31G.2 3 Pick up structure from rwf698 on the chk file.) SEE IOp(6) FOR DETERMINATION OF THE NUMBER OF GAUSSIANS IN THE INNER SHELL. -2 -1 0 N Read in force constants for each cartesian coordinates No constraints. 0 1 2 3 4 row. 5 USE IOp(8) TO SELECT 5D/6D. whether intended for use in expanding AOs (IOp(5)) or in expanding the density (IOp(82)).9p)-->(631111. Si 6-31+G(df. Literature citations in SDDPot. 26 Coreless: Li.6-31+g(d. NOTE IF IOp(5)=3 AND IOp(6)=8 . IOp(6) specifies type. useful only for density basis sets.6}Z) and augmented if IOp(7)=10. S. 22 EPR-III basis sets. Type selected by IOp(6) for H-Ar.Ne 6-311+g(3d2f) on Na . 24 G3large basis set. Literature citations in CEPPot.Large APNO basis set 15 CBS basis #6 -. type selected by IOp(6) (=0-4 for V{D. 9 Stevens/Basch/Krauss/Jasien/Cundari ECP basis sets for H-Lu. B-Ne 2MWB. rest LANL1MB. IOp(3/6)=5 for MTsmall basis set.Ne 6-311++g(3d2f) on Na .Ar 13 CBS basis #4 -. IOp(3/6) NUMBER OF GAUSSIAN FUNCTIONS N STO-NG. 17 Stuttgart/Dresden ECP basis sets.p) on H .B. LP-31G FOR LI. 20 MIDI! basis sets.p) on P. 23 UGBS basis set.5.MG. Type selected by IOp(6).T.Ar 11 CBS basis #2 -. Cl 14 CBS basis #5 -. He 6-311+G(2df) on Li . 21 EPR-II basis sets. selected by IOp(6) 28 Auto-generated.BE. 19 Ahlrichs TZV basis sets. 18 Ahlrichs SV basis sets.Q. DEFAULT OPTIONS IOp(6)=0 IF IOp(5)=0: N=3 STO-3G IF IOp(5)=1: N=4 4-31G IF IOp(5)=2: N=3 STO-3G (VALENCE) IF IOp(5)=3: N=3 IF IOp(5)=5: N=3 .p) on H.8 Dunning/Caltech basis sets. 25 G3MP2large basis set. use daggers if any polarization 12 CBS basis #3 -.AL LP-41G FOR OTHER ROW1 AND TWO ATOMS.2p) on H .STO-NG-VALENCE.Be 2SDF.6-31+G(d.N-31G. 10 CBS basis #1 -. 27 DGauss basis sets.Core correlation basis set 16 Dunning cc basis sets.6-311++G(2df. N-21G.NA.6-31G.LP-N1G. otherwise LANL1). J. Dunning valence double-zeta. When IOp(5)=6 (LANL basis and potentials) this selects the type: 0 1 2 LANL1 ECP. they are read along with the basis set information. otherwise LANL1). This option is useful when doing general basis geometry optimizations or properties using a wavefunction on the checkpoint file. THIS OPTION IS USED TO CONTROL WHERE THE BASIS IS TAKEN FROM: 0 READ GENERAL BASIS FROM THE INPUT STREAM.WHEN IOp(5)=7 (GENERAL BASES). SHC ECP on second row). WAG basis (Dunning VDZ on first row. When IOp(5)=9 (CEP basis) this option selects the type (H-Ar only): 0 1 CEP-4G. MBS. 85. 1662 (1981) and J. Chem. Chem. where one wants to modify the basis differently during different steps. 3 Same as 1. Phys. MBS. See Rappe. LANL2 ECP (where available. 3 LANL2 ECP (where available. When IOp(5)=8 (Dunning bases) this option selects the type: 0 1 2 Dunning full double-zeta. for density basis (generated here from 2) 1x Read from the alternate file and remove functions/ECPs for inactive atoms. DZ. 85. If non-standard ECPs are in use. Smedley. for density basis (generated here from 1) 4 Same as 2. Used for counterpoise calculations. Phys. 2 Read the general basis from the checkpoint file. 1 READ THE GENERAL BASIS FROM THE RW-FILES AND MERGE WITH THE COORDINATES IN BLANK COMMON TO PRODUCE THE CURRENT BASIS. and Goddard. LANL1 ECP. DZ. . CEP-31G. 3546 (1981). D-FUNCTIONS ON HEAVY ATOMS (2ND ROW ONLY FOR 3-21G). Functions from extended Huckel theory.2 CEP-121G. 3 ONE SET OF D-FUNCTIONS AND ONE SET OF F-FUNCTIONS ON HEAVY ATOMS (indicates an extra tight 2df with ccp basis sets. 2 2 D-funcs. along with LP-31G potential. on heavy atoms (scaled up/down by a factor of 2 from the standard single D value). 0 1 NONE. When IOp(5)=17 (Stuttgart/Dresden ECP bases) this option selects the type according to: 6 SDD 7 SDD for Z > 18. 4 TWO SETS OF D-FUNCTIONS AND ONE SET OF F-FUNCTIONS ON HEAVY ATOMS. When IOp(5)=7 (DGauss basis sets): 1 2 3 4 5 DGDZVP DZVP2 DGTZVP DGA1 (fitting basis) DGA2 (fitting basis) IOp(3/7) DIFFUSE AND POLARIZATION FUNCTIONS. D95 and no ECP otherwise. VSTO-4G basis for 1st row. When IOp(5)=26 (Coreless basis) this selects the choice of basis (the same ECPs are used regardless): 0 1 2 3 Default (3) Primitives which match the ECPs. . 1 2 . 2d. 2 SETS OF P-FUNCTIONS ON HYDROGENS.p polarization basis Tight d for VnZ+1 (W1 theory) A SET OF DIFFUSE SP FUNCTIONS ON HEAVY ATOMS.d. Augment non-hydrogens only (cc basis sets only). Three sets of p-functions and one set of d-functions. FORCE 6D. 0 SELECTION DETERMINED BY THE BASIS N-31G 6D/7F N-311G 5D/7F N-21G* 5D STO-NG* 5D LP-N1G* 5D LP-N1G** 5D GENERAL BASIS 5D/7F FORCE 5D. CBS-Q d(f). IOp(3/8) SELECTION OF PURE/CARTESIAN FUNCTIONS. 400 TWO SETS OF P-FUNCTIONS AND ONE SET OF D-FUNCTIONS ON HYDROGENS.p) -.5 6 7 8 9 10 20 100 200 Three sets of d functions. P-FUNCTIONS ON HYDROGENS.d. Three sets of d functions and one set of f-functions.2d on 2nd and later atoms. 500 600 700 1000 Three sets of p-functions. A DIFFUSE S FUNCTION ON HYDROGENS. 1d on 1st row atoms. 300 ONE SET OF P-FUNCTIONS AND ONE SET OF D-FUNCTIONS ON HYDROGENS. Three sets of d functions and two sets of f-functions. Li-Ne=0. Split S=P AO basis shells into separate S and P shells. AO shells into S=P. Massage the data in Common /B/ and Common /Mol/. FORCE 10F.65*I-4. the values for H-Ar can be determined by Slater's rules: H=1. Do not split S=P AO shells.2. IOp(3/9) Where 308 should store dipole velocity integrals. shells.325*(IA-1). 10000 Split S=P=D= 20000 Do not split AO S=P=D 100000 Uncontract the AO basis. 200000 Uncontract the density basis 300000 Uncontract both basis sets. DEFAULTS STO-NG STANDARD SCALE-FACTORS.95)/3 . Na-Ar=(0. Store in RWF N. D. 0 -1 N Usual place (572). He=1. Read in general basis data in addition to setting up a standard basis.10 20 FORCE 7F. F. Write over the dipole length integrals (518). Add ghost atoms to /B/ so that every shell is on a separate center. For VSTO-nG. IOp(3/10) Modification of internally stored bases (default 12000): 0 1 10 100 1000 2000 None.7. 50 15.56 13.05 2.90 2.74 1.48 3.66 8.50 1.ATOM H HE LI BE B C N O F NE NA MG AL SI P S CL A 1S 1.95 2.53 14.31 5.64 10.88 3. 2686 (1963) OUTER SHELL HAS BEEN SELECTED ON THE BASIS OF NUMEROUS OPTIMIZATION STUDIES ON VARIED SMALL MOLECULES.PHYS.75 1.36 4.59 12.CHEM.55 2.70 1.67 7.67 6. 38.26 6.43 17.10 2.80 1.83 5.68 4.15 1. N-31G (ALSO N-31G* AND N-31G**) STANDARD SCALE-FACTORS HYDROGEN 1S 1S* .25 2.68 5.24 1.75 1.33 INNER SHELLS ARE BEST ATOM VALUES J.69 2.70 1.72 1.61 11.47 16.79 6.69 3.40 2SP 3SP 0.65 9.90 4. 99 0.01 LP-N1G SCALE=1.00 1.12 1.00 3SP 0.00 2SP 1.98 0.6 VALUE C-NE 0.0 FOR LI-AR (INNER AND OUTER) STANDARD POLARIZATION EXPONENTS FOR N-31G* AND N-31G** BASES ATOM H LI BE B 1.00 3SP* 1.00 2SP 1.4 0.99 1.H 1.98 0.20 1.00 1.8 STANDARD POLARIZATION EXPONENTS FOR STO-NG* BASIS.00 1.2 0.1 0.00 1. ATOM VALUE .01 1.00 2SP* 1.00 1.00 1.98 1.00 1.98 1.00 SECOND ROW ATOMS ATOM P S CL 1S 1.03 1.00 1.15 FIRST ROW ATOMS ATOM 1S B C N O F 1.00 0.02 1.04 0. MG AL-CL 0. but not yet accepted by them.NA. CNDO/2. . SUITABLE FOR USE WITH THE OPEN SHELL (UHF) SCF. INDO/2. 0 REGULAR INTEGRAL FORMAT IS USED. 1 RAFFENETTI '1' INTEGRAL FORMAT IS USED. to account for sparkles. IOp(3/14) Addition of electrostatic integrals to core hamiltonian. SUITABLE FOR USE WITH OPEN SHELL RHF SCF AND THE POST-SCF PROCEDURES. IOp(3/13) Nuclear center whose Fermi contact terms are to be added to the core hamiltonian. ZINDO/S.09 0.multiply moments by fudge factor for charged species. 3 RAFFENETTI '3' INTEGRAL FORMAT. 9 USE ILSW TO DECIDE BETWEEN RAFFENETTI 1 AND 2.39 IOp(3/11) CONTROL OF TWO-ELECTRON INTEGRAL STORAGE FORMAT. SCRF calculation -. ZINDO/1. 0 -1x No. IOp(3/12) Flag for semi-empirical runs. 2 RAFFENETTI '2' INTEGRAL FORMAT. The magnitude is specified by IOp(3/15). CAN ONLY BE USED WITH THE CLOSED SHELL SCF. translation vectors and d functions properly: 1 2 3 MNDO/AM1. yyz.xyz field second derivatives.xxzz.xxyy.yyzz. octopole. xxx. -1 Yes. starting with electric field.yyyz.xzz. read if general basis.zzzy. and the electric field is stored in Gen(2-4).-6 Read coefficients of field.xz. ECPs if defined with the basis set.yz electric field gradient. Read components of electric field only from /Gen/.zzzx.0001. 1-34 Just component number n in the above order with magnitude given by IOp(3/15). N N * 0. yzz. No.yyyy. xxy. (These correspond to dipole.xxyz. read 12 cards with x.z components of electric field. IOp(3/15) Magnitude of electric field.xxxy. IOp(3/17) SPECIFICATION OF PSEUDOPOTENTIALS -1 Read potential in old format. Read components of moments off rwf 521 on chk file.10).xy.yyxz. . and xxxx.zz. xxxz. zzxy field third derivatives in format (3D20.zzzz. Read components of moments off rwf 521. and hexadecapole perturbations).y.xyy. -5 -4 -3 -2 Read components of electric field only from /Gen/ on checkpoint file. up through 34 elements (hexadecapoles) in free format.xxz. quadrupole. blank terminated.zzz. followed by xx. IOp(3/16) Pseudopotential option 0 1 2 Default. The nuclear repulsion energy is also modified appropriately.yy. yyyx.yyy. Yes. SDD combination Dresden/Stuttgart potentials . Stevens/Basch/Krauss CEP potentials.MDF Dresden/Stuttgart potentials .MHF (second set) Dresden/Stuttgart potentials . based on IOp(3/5).MWB (third set) Pseudopotentials for all coreless basis. .MHF (first set) Dresden/Stuttgart potentials . Alternative potentials for coreless basis.SDD for Z > 18. unused READ IN FROM CARDS (SEE PINPUT FOR DETAILS) Dresden/Stuttgart potentials .MWB (first set) Dresden/Stuttgart potentials . USE INTERNALLY STORED 'CORELESS HARTREE-FOCK' Goddard/Smedley SECE/SHC potentials.MWB (second set) Dresden/Stuttgart potentials . LANL1 potentials.0 1 2 3 4 5 6-7 8 9 10 11 12 13 14 15 16 17 18 19 20 Default. D95V. no ECP otherwise.SDF Dresden/Stuttgart potentials . IOp(3/18) PRINTING OF PSEUDOPOTENTIALS 0 PRINT ONLY WHEN INPUT IS FROM CARDS or if GFPrint was specified.SHF Dresden/Stuttgart potentials . Dresden/Stuttgart potentials . LANL2 potentials. N integer words. IOp(3/21) Size of buffers for integral derivative file.CHF). PRE-CUTOFFS DESIGNED FOR THE 6-31G* BASIS. 16384 integer words on VAX. 55296 words on Cray). . IOp(3/20) Size of buffers for integral file. IOp(3/23) Disable use of certain basis functions. No longer used.1 2 PRINT DON'T PRINT IOp(3/19) SPECIFICATION OF SUBSTITUTION POTENTIAL TYPE 0 DONT USE ANY SUBSTITUTION POTENTIALS N REPLACE THE STANDARD POTENTIAL OF THIS RUN (EG. IOp(3/22) CONTROL OF THE PRE-CUTOFF IN THE TWO-ELECTRON D-INTEGRAL PROGRAM. N N integer words. 0 Default (Machine dependant. WITH A SUBSTITUTION POTENTIAL OF TYPE N WHEREVER SUCH A SUBSTITUTION POTENTIAL EXISTS. 0 N Default (3200 integer words). Used only in L312. 0 1 NO PRE-CUTOFF. 0 Use all basis functions. LINK NUMBER. 0 1 10 100 Default (don't print). Print as GenBas input. DO ALL INTEGRALS AS WELL AS POSSIBLE in L314. IOp(3/25) NUMBER OF LAST TWO-ELECTRON INTEGRAL LINK. 0 1 >0 WE ARE NOT USING TWO-ELECTRON INTEGRALS. USE OLD very inaccurate CUTOFFS IN LINK 311. terminated by a blank line. Print in more readable format. 1 Read in a list of basis function numbers in Format (10I5). Print old-fashioned table. .0. 0 1 2 10 DEFAULT. Direct SCF. IOp(3/26) ACCURACY OPTION. and set their dialgonal core Hamiltonian elements to +100. -1 We are re-using integrals produced earlier in the current calculation. use the /IBF/ already on the RWF. DO ALL INTEGRALS AS WELL AS POSSIBLE in L311. -2 Use integrals from a previous job read /IBF/ from the checkpoint file. TEST. INTEGRALS ARE COMPUTED TO 10**-10 ACCURACY. 1000 Print shell coordinates. STO-3G. TEST. IOp(3/24) Printing of gaussian function table. IOp(3/31) USE OF SYMMETRY IN COMPUTING GRADIENT (Obsolete). 1 TWO-ELECTRON INTEGRAL SYMMETRY IS TURNED ON.L311 does nothing. 0 1 2 Default. IOp(3/27) HANDLING OF SMALL TWO-ELECTRON INTEGRALS. 0 TWO-ELECTRON INTEGRAL SYMMETRY IS TURNED OFF. 10**-N. use IsAlg. HOWEVER.20 Sleazy. d and higher elsewhere. IOp(3/28) Special SP code control. Use looser cutoffs in L314. IF THEY DON'T. DISCARD INTEGRALS WITH MAGNITUDE LESS THAN 10**-N. THE SET2E WILL INTERROGATE ILSW TO SEE IF THE SYMMETRY RW-FILES EXIST. SP integrals in link 311. NOTE. All integrals with d's -. IOp(3/30) CONTROL OF TWO-ELECTRON INTEGRAL SYMMETRY. 0 N DISCARD INTEGRALS WITH MAGNITUDE LESS THAN 10**-10. SYMMETRY HAS BEEN TURNED OFF ELSEWHERE. IOp(3/29) Accuracy in L302: 0 N Default (10**-12). IOp(3/32) . AND SET2E WILL ALSO TURN IT OFF HERE. PRINT TWO-ELECTRON INTEGRALS IN STANDARD FORMAT. PRINT ONE-ELECTRON INTEGRALS. Yes. Yes. Yes. and reduce expansion space if linear dependence is found (NYI). CONTROL WORDS PRINTED (AS USUAL). IOp(3/33) INTEGRAL PACKAGE PRINTING. Yes. and use Schmidt orthogonalization to reduce expansion space. 0 1 NO DUMP. 2 ADDITIONALLY.Whether to check the eigenvalues of the overlap matrix: 0 1 2 3 4 5 Default (4). IOp(3/35) . 3 ADDITIONALLY. COMBINATION OF 1 AND 3. PRINT TWO-ELECTRON INTEGRALS IN DEBUG FORMAT. using SVD to reduce expansion space. No. IOp(3/34) DUMP OPTION. COMMON/B/ IS DUMPED AT THE BEGINNING OF EACH INTEGRAL LINK. COMBINATION OF 1 AND 4. THE INTEGRALS ARE PRINTED (STANDARD FORMAT). 0 1 3 4 5 6 NO INTEGRALS ARE PRINTED. No. Do not compute absolute overlaps.LAST INTEGRAL DERIVATIVE LINK (No longer used in overlay 3). Compute absolute overlap over both contracted and over primitive functions. IOp(3/38) Algorithm for 1e integrals: . Octopole. Hexadecapole. 0 WHATEVER LINK STARTS WRITING THE INTEGRAL DERIVATIVE FILE SHOULD ALSO CLOSE IT. Dipole. IOp(3/36) Maximum order of multipoles to compute in L303: -1 0 1 2 3 4 00 10 20 30 None Default (dipole). IOp(3/37) Whether to sort integrals in L320. N IS THE NUMBER OF THE LAST TWO-ELECTRON INTEGRAL DERIVATIVE PROGRAM. Compute absolute overlap over contracted functions. 0 1 2 Default (No). Quadrupole. Default (same as 20). Yes. Rys. PRISM. Do only overlap and not other 1e integrals. neglect three center one-electron integrals. PRISM. same as 1. Default in 308. IOp(3/40) Neglect of integrals: 0 1 2 3 10 20 30 keep all integrals. neglect 2e integrals with diatomic differential overlap. Leave alone. neglect 1e integrals with diatomic differential overlap. Explicit spdf code. 0 1 Initialize. 0 1 No NDDO NDDO . same as 1. neglect four center integrals. IOp(3/39) Initialization of force and force constant rwfs. IOp(3/41) Various semi-empirical methods. neglect three center two-electron integrals as well.0 1 2 00 10 20 Default in 302. 100000 Do CNDO/2. 200000 Do INDO/2. followed ((HDiag(J. GNDDO/1 parametrization.I=1. Use the real overlap matrix. Default (unit overlap matrix).J=1.00 Default use of NDDO beta parameters (arithmetic mean for indo parameters. . Use the unit matrix for the overlap. 10 20 000 100 by Arithmetic mean in NDDO.I). Geometric mean in NDDO.I=1.12). Read parameters for atomic numbers 1-18 in the order Scale (D20. followed by ((Beta(J. geometric mean for NDDO/1 or read-in parameters).18) 200 300 400 500 0000 1000 00000 10000 20000 Read parameters from rwf.I). 1100000 Do Harris functional scaling atomic densities for current charge and multiplicity.3). Default parameters (same as 5).3).18) (Format 3D20.J=1. Use Slater's rules scale factors. 300000 Do ZINDO/1. 1000000 Do Harris functional. Original INDO/2 Beta and HDiag Parameters. 400000 Do ZINDO/S.12). Read parameters from chk. Use STO-3G scale factors. 1200000 First-order XC. neglect four center transformed integrals. Force units of Bohr for coordinates. IOp(3/43) Solvent charge distribution to add to Hamiltonian: 0 1 2 3 4 5 10 20 None Read charges and DBFs from input stream in input orientation Read from RWF. 0 1 keep all integrals. Read from Chk. the perturbation is computed separately and stored in the third and fourth matrices in the core Hamiltonian rwf. for testing. If negative.1300000 Second-order XC (NYI). IOp(3/44) integral rejection using L318. . Read charges and DBFs from input stream in standard orientation Force units of Angstroms for coordinates. 1400000 Regular SCF with separate K. 1500000 J as usual but NDDO for K. Load all the integrals into memory. Read the integrals sequentially. Same as 1. IOp(3/42) How to form NDDO core hamiltonian in L317: 0 1 2 Default (same as 1). Use expanded matrix logic for PBC exact exchange. .T and S unchanged. 2 Reverse choice of whether to precompute distance matrix during numerical quadrature. Bit flags: 0 1 If bit 0 is set (use AllowP array) then read in a list of allowed paths.2 3 4 5 6 neglect four center and 3 center (ab|ac) integrals.0) integrals. only 1e. neglect four center and three center (0. Use unit matrix (for debugging). Do not transform 2e integrals. NDDO approximation -. Order of multipoles in SCRF for L303.no (ab|xx) and no NDDO on 2e and V ints only -. IOp(3/45) Transformation matrix in L318. -1 Force use of only the OS path for all calculations. yes IOp(3/47) Flags for use in PRISM and CalDFT throughout the program. 0 1 2 use S**-1/2. IOp(3/46) Whether to abort the job if badbas detects an error: 0 1 2 Default (yes). just orthogonalize functions on the same center. no. The default is diagonal only on vector machines.3 4 Skip consistency checks for XC quadrature Do not do extra work to use cutoffs better. Force use of FoFCou. . currently only affects CalDFT. Force all near field in FMM. currently only in PrismC. Forbid use of gather/scatter digestion even for small numbers of density matrices. Force single matrix code in CPKS. No longer used. Turn off use of Sqrt(P) in density-based cutoffs. 6 7 8 9 10 11 Trace input and output using Linda/subprocess. 13 14 15 16 17 18 19 20 21 22 23 24 Turn off Schwartz during FMM/NFx calculations. Reserved for more control of scatter/gather. Use Mura radial grid instead of Euler-2 grid. Turn on angular offsets in XC grid generation. Force square loops. 12 Reverse normal choice of Scat20 vs. Default is to use replicated matrices only on Fujitsu and NEC. even if not doing FMM. Turn off vFMM. replicated Fock matrices. Forbid use of FoFCou. Turn off MP-based cutoffs in FF part of NFx. Do allocation for parallel 2e integrals but run sequentially. 5 Reverse normal choice of diagonal/canonical sampling in Prism and PrmRaf. Do nuclear contribution in FoFCou even for non-PBC Do not use special Coulomb algorithm in FoFCou. Uncontract all shell pairs. IOp(3/48) Options for FMM: RRLLNNTTWW RR: LL: NN: TT: WW: Range (default 2) LMax (default from tolerance) Number of levels (default 8) Tolerance (default 18) IWS (default 2).25 26 27 28 29 Do allocation for parallel XC but run sequentially. IOp(3/49) More options for FMM: 1 2 4 8 16 32 64 128 Indicates whether FMM can be used by FoFCou. Do not symmetry reduce grid points on unique atoms. Turn on FMM print. Split primitives for better boxification. Default UseUAB/Use 256. Make all atoms large in XC quadrature. Make all shells large in XC quadrature. . Convert to sparse storage under FoFCou for testing. Turn on use of precomputed XC weights. Apply symmetry to derivative distributions (NYI). Do not save as many multipole expansions as possible in memory. ECP integrals are evaulated in L302. based on geometry (but minimum for molecules 3. no NFx NFx range NN (R+n with n=NN-1). IOp(3/52) Turn off normal evaluation of ECP integrals. 1024 Set up for parallel FMM but run loops sequentially. No Cutoffs 10**-N. if 128 set. Turn off parallelism in FMM (does not use parallel logic).256 512 UseUAB. IOp(3/53) Accuracy in ECP integral evaluation: 0 -1 N Default. Default box length. IOp(3/51) Parameters for NF exchange and box length (MMMMNN): 00 NN 0000xx Bohr).0 MMMMxx Box length MMMM/1000 Bohr. 4096 Force FMM on. 0 1 Default: if needed. IOp(3/54) . 2048 Do not default to FMM. so L302 does not do ECP ints. 8192 Set by PsmSet to indicate whether the NAtoms test for defaulting FMM was passed. Old routines will be used. IOp(3/57) No. IOp(3/58) Cholesky control options. No Yes. cutoff 5 x 10 ** (N+100) Yes. of core electrons for Stuttgart/Dresden ECP's.Type of core density to use with ECPs: -1 0 1 2 None Default (None) Non-relativistic Relativistic IOp(3/55) Use of sparse storage: N<-100 -3 -2 -1 0 N Yes. crude accuracy (5x10**-5) Yes. intermediate accuracy (5x10**-7) Yes. IOp(3/59) Threshold for throwing avay eigenvectors of S: . default accuracy (10**-10). cutoff 10**(-N) IOp(3/56) Cutoff for intermediate matrices during sparse operations: 0 N 100 times smaller than storage cutoff. 10**(-N). XX00 F(Mu. Mu. IOp(3/62) Maximum allowed error in S over orthogonalized basis functions: 0 N Default (10**-9). (Defaults to 15 angstroms). . Orthogonalize and remove primitives with 0 coefficients. Yes. IOp(3/60) Control of orthogonalization and simplification of ccp basis sets. Nu are basis functions on the same atom. Lambda are basis functions on different atoms. (Defaults to no F(Mu. Only active if IOp(3/39)=0. Orthogonalize and remove primitives with 0 or small coefficients.Nu) atom--atom cutoff criterion (angstroms) Mu. 10**-(N). IOp(3/63) Debug option to test point charge FMM. 0 1 No.Lambda) atom--atom cutoff criterion (angstroms).Nu) cutoff). IOp(3/61) Sparse Semiempirical Hamiltonian Cutoffs in L302: XX F(Mu. 0 1 2 Default (1). IOp(3/64) Set value for ILSW derivative flag.0 N Default (10**-4) 10**-N. IOp(3/70) SCRF flag. Leave alone. IOp(3/67) Electric-field dependent functions: 0 1 2 3 Default (on if already present in basis read from rwf or chk. Set to N. Old logic for NRecip=N. About N points. with standard values.-2 -1 0 N Set to zero Set to -1. with read-in values. Yes. IOp(3/65) Number of k-points: -1 N -N Just Gamma point. otherwise off). 0 Default (1) . set to N. No. IOp(3/66) Over-ride setting of NThInc in lineary dependence cutoff: -1 0 N 0 Don't change. Yes. IOp(3/74) Type of exchange and correlation potentials: -24 O3LYP. Read setting from checkpoint. Flag for macroiterations (IPCM). C-PCM.1 2 3 4 5 0100 1000 2000 2100 2200 2300 3000 4000 Use defaults. IOp(3/73) ONIOM system flag (for SCRF setup). Onsager. IOp(3/72) Solvent type flag (for SCRF setup). Read from rwf. . Read setting from the input stream. IVC-PCM Cramer/Truhlar solvation model. Read setting from checkpoint and modify them by reading from the input stream. IEF-PCM. SCI-PCM. 10000 Generate COSMOTHERMO output. D-PCM. IOp(3/71) IDeriv level flag (for SCRF setup). HCTH147. BA1PBE. PBE1PBE mPW3PBE. Becke3 using VWN/LYP for correlation. PBE3PBE. . mPW1PBE.5 HF + 0.5 LSD Do only coulomb part. B97-2. B1LYP. LG1LYP. HCTH93. Becke "Half and Half" with LYP/VWN correlation. mPW91PW91. B98. B97-1. Becke 3 with Perdew 86 correlation. mPW1LYP. Becke3 with Perdew 91 correlation. BA3PBE.-23 -22 -21 -20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 HCTH407. B1B95. skip exchange-correlation. Becke "Half and Half": 0. Vosko-Wilk-Nusair method 5 correlation. Hartree-Fock-Slater exchange (Alpha = 2/3). X-alpha exchange (alpha= 0. Perdew 81 + Perdew 86 correlation.00 00 01 02 03 04 05 06 07 08 09 10 11 18 19 20 100 200 300 400 500 600 700 Default. Lee-Yang-Parr correlation. LG exchange. Perdew 81 correlation. same as 100. PW91 exchange Gill 96 exchange . No correlation.7) Becke 1988 exchange. VWN 80 (LSD) correlation VWN 80 (LSD) + Perdew 86 correlation OS1 correlation PW91 PBE VSXC Bc96 VWN5+P86 LYP+VWN5 for scaling KCIS Hartree-Fock exchange. JJJ angular points. IOp(3/75) Number of radial and angular points in numerical integration for DFT: 0 IIIJJJ Use a special grid designed for efficiency (default). B97-1 coefficients. -10 -9 -8 -7 -6 -5 -4 O3LYP coefficients. IOp(3/76) Mixing of HF and DFT. III radial points. same as B1B95. B97-2 coefficients.72 c=0.0 and 1.800 900 1000 1100 1200 PW86 exchange mPW exchange PBE exchange BA exchange VSXC exchange So 100 is Hartree-Fock. with a=0. 200 is Hartree-Fock-Slater.81. Coefficients of 0.8 b=0. B98 coefficients. 205 is Local Spin Density. Note that Becke actually used Perdew correlation rather than LYP. mPW91PW91 coefficients. HCTH coefficints. respectively. -3 -2 -1 Becke "Half and Half" 0.5 HF + 0.0 for DFT and HF. .5 Xc + Corr Coefficients of 0 and 0 (no exchange). and 402 is BLYP. Becke 3 coefficients: aLSD + (1-a)HF + b(dBx) + VWN + c(LYP-VWN). IOp(3/77) Mixing of local and non-local exchange: -1 0 0 for both. In L510. Use Savin weights. IOp(3/80) . Sign is applied to the local term. 1 to set up for CAS-MP2 or 2 to do spin-orbit calculation.0 Default: pure HF. DFT or mixed in accord with IOp(3/76) Mixture of MMMM/10000 DFT exchange and NNNNN/10000 MMMMMNNNNN HF exchange. 0 -1 -2 -3 Default (SS weights) Use SS weights. IOp(3/79) Range cutoff in Becke weights. N Use Becke weights with cutoff N Bohr. Default (coefficients of 1 and zero as determined by IOp(3/42) MMMMMNNNNN MMMMM/10000 non-local plus NNNNN/10000 local. Default (coefficients of 1 and zero as determined by IOp(3/42) MMMMMNNNNN MMMMM/10000 non-local plus NNNNN/10000 local. IOp(3/78) Mixing of local and non-local correlation: -1 0 0 for both. Sign is applied to the local term. -M<-3 Use SS weights with XCal = M/1000. Use Becke weights with default cutoff of 30 au. 0 Keep all functions with angular momentum up to MaxTyp+1. IOp(3/84) Equivalent of IOp(3/7) for density basis. None (static limit). IOp(3/83) Equivalent of IOp(3/6) for density basis. Negative to turn off screening of basis functions and grid points. 3 Gross-Kohn form. For auto-generated basis sets: -1 Keep all generated functions. 0 1 Default (same as 1). For auto-generated basis sets: 0 Default (22) . -1 0 N None. IOp(3/82) Fitting density basis set for Coulomb in DFT. 1000000000 turns of microbatching logic. Same numbering of basis sets as for AO basis. 2 Also static limit. See comments for IOp(3/5) and IOp(3/6) 28=Generate automatically from AO basis. including 7=General basis. Default (-1). where MaxTyp is the highest AO angular momentum. but returning zero for imaginary response contributions.Range for microbatching in DFT. for debugging. IOp(3/81) Frequency-dependence (if any) for XC functional. N Discard functions with L>=N. 100 Add ghost atoms to /B/ so that every shell is on a separate center. IOp(3/87) Discard density basis functions based on angular momentum: 0 N No. Pure. Massage the data in Common /B/ and Common /Mol/. 0 1 10 None. Discard basis functions with angular momentum >= N. Do not split shells Split F and higher shells away from S=P=D. Use only AO primitives squared in fititing basis. . Read in general basis data in addition to setting up a standard basis. IOp(3/88) Modification of internally stored density basis. 0 1 2 Default (pure for read-in basis). IOp(3/85) Pure vs. Discard density basis functions with angular momentum >= N. Cartesian functions in density basis. IOp(3/86) Discard basis functions based on angular momentum: 0 N No.1 2 10 20 Use all products of AOs. This is also done if requested in IOp(3/10). Cartesian. . Solve iteratively with no preconditioning Solve iteratively with diagonal preconditioning. 5 6 7 1xx 2xx 3xx 4xx 5xx 6xx 1xxx Iterative.. D. Form A' over neutral distributions via direct products. 4 Use fits. IOp(3/90) . 1 otherwise). IOp(3/89) Set up for density fitting. density shells into S=P. Do not split S=P density shells.1000 2000 Split S=P density basis shells into separate S and P shells. solving iterative with stored A. 1 2 3 Do not use density fits. . Use fits. shells. Solve non-iterative using precomputed A'^-1.. F. no formation of A.. Form A' over neutral distributions via multiplies by A. solving iterative with direct products.. Put all functions into a single block in forming the preconditioning matrix. forming Z = modified A^-1. Use fits. 20000 Do not split density S=P=D. 0 Default (102 if a fitting set has been included and pure DFT is being used. 10000 Split S=P=D=. with A formed to generate preconditioning. Form inverse matrix once. Solve iteratively with symmetric block-diagonal preconditioning. Solve iteratively with non-symmetric block-diagonal preconditioning.. Douglass-Kroll-Hess 0th order. default NN=06. IOp(3/91) Scalar relativistic core Hamiltonian: 0 1 2 3 4 Default (1) Non-relativistic. DBFs have normalization as for AOs. AOs have J-normalization.Thresholds for density fitting MMNN 10**(-MM) on iterative solution. unless scalar relativistic) . DBF coefficients are for raw primitives. AO coefficients are for raw primitives. IOp(3/93) Nuclear charge distribution: 0 Default (1. IOp(3/92) Whether read-in basis sets are in terms of normalized primitives? 0 1 2 3 10 20 30 Default (12). default MM=09. Douglass-Kroll-Hess 2nd order. AOs have normalization as for AOs. 10**(-NN) on generalized inverse. DBFs have J-normalization. RESC. for debugging. 0 N -M default (100). -N 0 N At least N cells in each direction. IOp(3/94) Range of PBC cells in Bohr. At least N cells total. al Very tight single s Gaussians. IOp(3/96) Number of PBC cells for DFT: 0 N As many as look significant. Based on range estimate (IOp(3/94)). for debugging. Include nuclear charge distributions in DBF set.1 2 3 4 10x Point nuclei Single s Gaussians using formula of Quiney et. N Bohr. Same as 2 but exponents are 100x smaller. IOp(3/95) Minimum number of PBC cells. Mxxx Use method M to handle nuclear charges during density fitting. Multiply usual range by M. IOp(3/97) Number of PBC cells for exact exchange: 0 As many as look significant. At least N. . 0 N Default. IOp(3/102) . and point coordinates. No more than N matrices. storing only grid parameter Yes. IOp(3/98) Maximum number of density matrices in PBC.N At least N. weights. No Yes. storing grid parameters and weights. storing grid parameters. IOp(3/101) Maximum range of cells -N 0 N No more than N in each direction No limit. -1 otherwise). Yes. IOp(3/100) Minimum Number of PBC cells for PBC-MP2 0 N Same as for HF exchange. N. IOp(3/99) Whether to set up precomputed quadrature grid in L302: 0 -1 1 2 3 Default (2 if doing DFT. No more than N total. based on number of cells having overlap with cell 0. 1000). Default is Max(NDBF/2. Overlay 4 5 6 7 8 9 10 11 13 14 15 16 17 18 19 20 21 22 23 24 25 26 28 29 31 33 34 35 36 37 38 43 44 45 46 47 48 60 61 62 63 64 65 66 67 68 69 71 72 80 81 82 110 IOp(4/5) Type of guess: . IOp(3/103) Maximum number of vectors allowed in expansion space during iterative density fitting. Do not reuse. Negative to use projected equations rather than least-squares. Default is Max((1000. Read from chk file. 0 N Default (60) N atoms for the C1 case.Number of density fittings solutions to save from previous SCF iterations. to compensate for the loss of some symmetry with FMM. IOp(3/104) Maximum number of iterations during iterative density fitting. Reuse. This number is scaled up appropriately if symmetry is in use. IOp(3/106) Override default number of atoms threshold for turning on FMM (for debugging). 0 1 2 3 Default (re-use if present). Default is 6 (using 5 previous solutions plus the current right-hand side to generate the initial guess). IOp(3/105) Re-use of PBC cell data.NDBF+100). otherwise 2).3. This uses the Harris functional unless atoms heavier than Xe are present. Huckel guess (only valid for NDDO-type methods). 1000 Use the simultaneous optimization recipe: S**-0. Read intermediate SCF results which are on the checkpoint file. Projected ZDO guess. .4 above. 100 Convert Guess=Check to Guess=Restart or to generating guess depending on what if anything is on the checkpoint file. 9 Read generalized density specified by IOp(38) from the rwf file & generate natural orbitals from it.2. FORCE PROJECTED GUESS.1 corresponding to 1. Guess from model Hamiltonian. 8 Read generalized density specified by IOp(38) from the chk file & generate natural orbitals from it. 6 7 Renormalize and orthogonalize intermediate SCF results which are on the RWF. IOp(4/6) FORCED PROJECTION WHEN GUESS IS READ FROM CARDS (401). in which case Huckel is used.5 * V. chosen via IOp(11). 20000 Re-use orbitals not Fock matrices. IGuess values of 10-14 are generated internally and are the sparse versions of 0 and 5-8. 5 Renormalize and orthogonalize the coefficients which are currently on the readwrite files. EVEN WHEN BASES ARE IDENTICAL. 1 2 3 4 Read guess from the checkpoint file. EVEN WHEN BASES ARE IDENTICAL.3. 0 1 FORCE PROJECTED GUESS. 00000 Default (1 for PBC without alter.0 Default. Note that variable IGuess here has 4. 10000 Re-use Fock matrices instead of orbitals.2. 402. COMPLEX RHF. Suppress orthogonalization. IOp(4/8) ALTERATION OF CONFIGURATION (401). Check MOs for othornormality. Don't check MOs for othornormality. REAL RHF. 1 READ IN PAIRS OF INTEGERS in free format INDICATING WHICH PAIRS OF MO'S ARE TO BE INTERCHANGED.403). 2 Read in a permutation of the orbitals.2 00 10 20 000 100 200 SUPPRESS PROJECTION. COMPLEX. PAIRS ARE READ UNTIL A BLANK CARD IS ENCOUNTERED. 0 DO NOT ALTER CONFIGURATION. USE ILSW TO DETERMINE. REAL UHF. REAL ROHF. BUT USE ILSW TO DECIDE WHETHER RHF/UHF. . COMPLEX UHF. Default MO checking (yes). IOp(4/7) SCF CONSTRAINTS (401. -1 0 1 2 3 4 5 6 Ignore ILSW and determine on the fly. Default orthogonalization (perform) Schmidt orthogonalize guess orbitals. Initial guess orbital symmetries are assigned. as represented by the symmetry adapted basis functions produced by link 301. 0 Mix the HOMO with the LUMO. IOp(4/10) Orbitals to mix to form complex guess (401). 10 20 100 200 Localize all occupied orbitals together and all virtual orbitals together Localize the orbitals within the selected or defaulted symmetry. TWO SETS OF DATA AS DESCRIBED ABOVE WILL BE EXPECTED. 5 (Use symmetry in SCF if possible. . 4 use the full abelian point group. NOTE IF THE CONFIGURATION IS ALTERED ON AN OPEN SHELL SYSTEM. This option can cause the symmetry adapted basis function common blocks to be modified. SECOND FOR BETA. 1000 Force the guess orbitals to have the Abelian symmetry. Do not assign orbital symmetries in full symmetry. IOp(4/9) SCF symmetry control (401). Pick up the symmetry mixing information from the Alteration read-write file. Use alpha orbitals for guess for both alpha and beta.10 100 READ ALTERATION INFORMATION FROM THE READ-WRITE FILE. but do not assign initial guess abelian symmetries). Assign orbital symmetries for printing in full symmetry. 0 Default. These are read before any orbitals and before alteration commands. 2 3 Use no symmetry in the SCF. same as 104 1 Read groups of irreducable representations to combine in the SCF. FIRST FOR ALPHA. Use bare core matrix. Old Huckel. Use UltraFine and 10^-8 in Harris functional.4 in order of preference). converted to IGuess=3 and IZDO=3 here. For unprojected single diagonalization guess: 0 1 2 3 000 100 200 300 Default (same as 1). . ALSO. 6 Harris. same as 2. Use Harris Functional. Reading is terminated by a blank card. Best available (6. NOTE THE SAME CONSIDERATIONS FOR OPEN SHELL SYSTEMS WHICH APPLIED IN IOp(8) APPLY HERE. Iterative extended Huckel. Use SG1 and 10^-6 accuracy in Harris guess Use FineGrid and 10^-8 in Harris functional. CNDO. Dress core Hamiltonian with QEq-based density.1 Read from cards (2I3) pairs of integers indicating which pairs of orbitals are to be mixed. New Huckel. IOp(4/11) Type of Guess (401): For iterative ZDO Guess: -1 0 1 2 3 4 5 Force old path using old Huckel. Default. INDO. the replacements are read before the corresponding alterations (thus the order is alpha orbitals. beta orbitals. LUMO = LUMO + HOMO (ALPHA) AND LUMO = LUMO . IOp(4/16) .vector in the specified format. 1 Yes. Note that if alter is also specified. If IVec is -1.HOMO (BETA). 0 1 NO MIXING. (Vector(I). n0000 Force IDoV=n in HarFok. alpha alterations. Input is terminated by IVec=0. 1000 Save energy in Gen(43) for Harris functional.vector to replace. N Use multiplicity N. beta alterations). read one card with the format for the orbitals.974) and 10^-12 in Harris functional. MMMM00000 Use functional MMMM IOp(4/13) MIXING OF ORBITALS (401).NBasis) -. followed by zero or more sets of IVec (I5) -.400 500 Use user's IRadAn and 10^-8 in Harris functional. NOTE THAT THIS WILL USUALLY DESTROY BOTH SPACIAL AND ALPHA/BETA SYMMETRY. This is useful for generating guesses for open-shell singlets or unusual spin states involving orthogonal orbitals by treating them as high-spin in the guess (which only does UHF). IOp(4/14) Reading of specific orbitals (401). For alpha orbitals. IOp(4/15) Spin-state for initial guess (401). For beta orbitals. Use (199. all NBasis vectors follow. the same format as for alpha is used. THE MIXING IS DONE AFTER ANY ALTERATIONS. 0 No. 0 Use multiplicity in /Mol/.I=1. IOp(4/18) Number of orbitals in CI in 402. and determine an overall rotation to provide to the read. IOp(4/17) Number of open-shell orbitals (not electrons) in 402. . CIOp(4/19) L402: Spin change in CI (default based on multiplicity). N. 3 Translate to the current atomic coordinates. L405: Trucation level for excitations -. Use the basis functions as is.Whether to translate basis functions of read in guess (401). Translate to the current atomic coordinates. IOp(4/20) Type of model (402): (This is also tested in 401 to see whether atomic number greater than 102 are special flags).default full CAS. 0 1 2 Default (same as 2). MINDO/3. Number of electrons in the CAS space. 0 1 2 3 Default (AM1). CNDO.in orbitals. 0 N #open electrons. INDO. L405: Number of orbitals in the CAS space. Default is number of open shells. Pulay. 3/4 point on unless Pulay or Camp-King. Camp-King. AM1. Just list configurations. Use determinant basis with Sz=b/2. No Camp-King. . Use pseudo-diagonalization.4 5 MNDO. IOp(4/22) Derivatives? (402). 0 Default (no Pulay. IOp(4/21) SCF type (402). No 3/4. No Pulay (DIIS). 0 No. No pseudo-diagonalization. 1 2 10 20 100 200 1000 2000 3/4. use pseudodiagonalization). 10000 Write unformatted file (NDATA) of symbolic matrix elements. Use slater determinants. 100000 Write formatted file of symbolic matrix elements. Flags for MCSCF (L405): 1 10 100 1000 Read options from input stream. no Camp-King. RHF/UHF from IOp(5). Single determinant. (For Opt=MNDOFC). 403). ROHF (NYI). 0 1 2 Default (Same as 1). More flags for MCSCF: IFlag2 IOp(4/23) Number of iterations (402. eigenvalues.1 2 12 100 1 Yes. NDiag in L405. N. . but don't convert from Lowdin orbitals. and ILSW on the rwf (402). IOp(4/25) Wavefunction (402). 0 N Default. Restart 2nd derivatives. (For Opt=MNDOFC). 2nd derivatives. Don't update. Update second force array instead of first. NRow in L405. IOp(4/24) Whether to update orbitals. Update. Do 1st derivatives analytically if possible. Update. /Mol/. (For straight semiempirical calculations). 0 1 2 3 10 Default (don't update). mix all equally. IOp(4/26) Whether to mix orbitals in generated guess density: 0 No -3 Yes.MNDO.05 au (according to ZDO) of the HOMO & virtuals within 0. IOp(4/31) Root to solve for in CI (402) (Default is 1).AM1). Equal occupations of the lowest N virtuals and high N occupieds. Yes. Closed-shell 1/3 CI (only for MINDO3. IOp(4/29) NC in L405. General CI.15 au. mix valence orbitals and an equal number of virtuals. 0 1 2 NO PRINTING. mix valence occupieds with 0.3 4 5 -N Biradical 1/2 CI (only for MINDO3.AM1).10 binary switches in L405. -2 -1 N Yes. with N microstates read in. default 10**-7). IOp(4/28) SCF Convergence (10**-N. PRINT EVERYTHING. General CI. IOp(4/33) PRINTING OF GUESS. . using specified orbitals. PRINT THE MO COEFFICIENTS.MNDO. Old Si parameters. 0 1 No. IOp(4/38) Generalized density to use for natural orbitals: N Density number N. Old S parameters. Copy on disk is used. IOp(4/36) ZIndo reformating. reformat ZIndo integrals and wfn into rwf.IOp(4/34) DUMP OPTION. Yes. IOp(4/35) Overlap matrix. 0 1 NO DUMP. TURN ON ALL POSSIBLE PRINTING. Overlap assumed to be unity. . 0 1 2 Default (copy on disk is used). IOp(4/37) Selection of old MNDO parameters in L402: 0 1 2 Defaults. IOp(4/43) IDiEij = Switch for direct matel calculation. No pairs. implicitly. A flag is automatically sent to L510 to tell it to compute the remaining atels directly. and sets up L510 to compute extra matels etc. This is a normal GVBCAS calculation. but is smaller subsequently. since there can be many thousands in a large CAS. This type of computation can only be done in a CAS comp. 1 For direct route. IOp(4/44) 1 Prepare input for Mp2 implies IOp(21)=10 Slater Det. Eij's calculated here and stored on disk. n There are n pairs: 2*n extra orbitals and electrons will be added into the active space later. This occupies as much space as a full CAS in this link. option generates data for use in MC-SCF generation of zeroth order H note: for b=0 ie no unpaired spins forces use of Clifford Algebra Spinors instead of simple determinants c2IOp(4/45) Ipairs = number of GVB pairs in GVBCAS. normal CAS calculation. This is the GVBCAS test mode. but all configurations are printed. -n There are n pairs: 2*n orbitals and electrons of the specified CAS are to be considered to be GVB type orbitals when generating configs / matels. Also L510 must use Lanczos. L510 will execute normally. 0 Default. The configurations will not be listed unless see below. with all matels calculated here and stored on disk. L405 performs a CAS on the inner space.0 For normal route. This will be the only way to print configs in a direct matel calc. Configs printed as normal. 2 As option 1. IOp(4/46) CI basis in CASSCF: 1 Hartree-Waller functions for singlets . and IRanGd. Yes. -1 0 N No. Turn on non-bonded terms. IOp(4/63) Flags for which terms to include in MM energy: 0 1 2 10 100 Default (111111) Turn on all terms. Default (-1 if sparse is turned on) Yes. use the Lewis dot structure to generate a sparse guess directly. Use threshold 10**-N. IRanWt. Turn on all terms. r**-1 Coulomb. r**-2 Coulomb. Use diagonal guess density. Use Lewis dot structure guess density.2 3 10 Hartree-Waller functions for triplets Slater Determinants Write SME on disk IOp(4/47) Convert to sparse storage after generating guess. Turn on inversions/improper torsions . Yes. IOp(4/60-62) IOp(60-62) Over-ride standard values of IRadAn. IOp(4/48) Whether to do (sparse) Conjugate Gradient methods in 402: 0 1 2 No. 10**-N. 10**-N. 0 1 N N Default: Yes. 0 N Default (no cutoff). Turn on angle bending. No variable thresholds. IOp(4/65) Tighten the zero thresholds as the SCF calculation proceeds. Eps = N / 1000. Don't do QEq. Yes. initial threshold 5 x 10 ** (N+100) IOp(4/66) Dielectric constant to be used in MM calculations. 2 otherwise. 1==> 221) Do QEq. IOp(4/67) Whether to use QEq to assign MM charges.0. 100000 Turn on bond stretches. initial threshold 10**(-N) N<-100 Yes. . initial threshold 5x10-5. 0 1 2 Default (211 if UFF. 0 N Eps = 1.1000 10000 Turn on torsions. IOp(4/64) Cutoff for MM non-bonded term. Just generate orbitals from the Harris guess. Do for all atoms regardless of initial charge. Default (10 for PBC. 10**(-N). Do the extra guess and store as the initial Fock matrix. Default (200) Do for atoms which have charge specified or defaulted to 0. IOp(4/72) Irreps to keep in MCSCF CI-wavefunction. Save the Harris guess as an initial Fock matrix. Do the extra guess regardless. Do not do the extra guess. .00 10 20 000 100 200 Default (20) Do for atoms which were not explicitly typed. and not a small geometry step. Harris guess. Do for all atoms regardless of typing. IOp(4/69) Whether to do a new additional guess in addition to reading orbitals from the rwf: 0 1 2 3 00 10 20 Default: yes if no Guess=Alter. 20 otherwise). IOp(4/71) Write out AM1 integrals in 402 0/1 No/Yes. IOp(4/68) Convergence criterion for microiterations in L402: 0 N Default. 0 All List of up to 8 irrep numbers to include. .Nu) atom--atom cutoff criterion (angstroms) Mu. IJKLMNOP IOp(4/80) The maximum conjugate gradient step size (MMNN) 0000 MMNN No maximum step size Step size of MM.Nu) cutoff). IOp(4/82) Conjugate-Gradient Parameters MM NN00 cycles). MM=01 turns DIIS off) NN00 on the F(Mu. (defaults to no F(Mu. (MM=00 defaults to 20 cycles. (defaults to 15 angstroms). (defaults to 3 Don't use CG DIIS Use CG DIIS.Lambda) atom--atom cutoff criterion (angstroms) Mu.NN IOp(4/81) Sparse SCF Parameters MM Maximum number of SCF DIIS cycles. (defaults to 4 CG cycles). 00000 10000 000000 100000 0000000 Maximum Number of CG cycles per SCF iteration. Polak-Ribiere CG minimization Fletcher-Reeves CG minimization Use diagonal preconditioning in Conjugate-Gradient. PP0000 F(Mu. Maximum Number of purification cycles per CG iteration. Lambda are basis functions on different atoms. Nu are basis functions same atom. Overlay 5 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 35 36 37 38 39 40 41 42 43 45 47 48 49 50 51 52 53 55 56 57 58 59 60 61 62 63 64 65 70 71 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 IOp(5/5) Direct SCF control (L502. . Use option NNNNN in control of 2e integral calculation. Compute 2e integrals. L508). Compute 2e integrals and store in-core. Read the integrals off disk OR Incore (Acording to Memory) Compute 2e integrals(Using FoFDir). Form full Fock matrix every time. Form delta-F each iteration -. Default -.only in L502.incremental Fock matrix formation only for direct SCF. 0 1 2 3 4 NNNNNx 0000000 1000000 2000000 Default (same as 1). Read the integrals off disk. L510: Direct MCSCF control (L510). Compute 2e integrals (Use TrnDir + FoFDir = 4 Can.1000000 No preconditioning. Calcs).(How to Obtain the Integrals) 0 1 2 3 Incore or Direct(FoFDir) according to available Memory. IOp(4/110) Scaling of rigid fragment steps during microiterations. Compute 2e integrals and forbid in-core. : ICntrl = Algorithm control: 0 1 2 3 4 5 6 7 8 10 20 30 40 100 200 300 400 500 Default for MCSCF is (1522). forbidding incore Force conventional Something obsolete NNNNNx use option NNNNN in control of 2e integral calculation. BraKet up to L=8. Do not compute operator matrices. rest not done here. NNNNN=ICntrl with values as below. Rys f. GVB: DA==>FJ. HGP df (no sp done here at all). The default for first derivatives. Sleazy (10**-6) Cutoffs. Rys df (for debugging). BraKet only. Rys f (no sp done here at all). Force HGP only.4 5 6 Force FoFDir. Cutoffs for high accuracy.FK. Compute CIS operators Compute WTilda terms. Cutoffs for 10**-10 accuracy. HGP d. . No cutoffs. Compute SCF Fock matrices. Force Rys only. The default for integrals or second derivatives. HGP sp. HGP spd. Load R1 and R2. and R03. Load R1 and R03. Load R1. 1000000 Form full Fock matrix every time.incremental Fock matrix formation only for direct SCF. Compute second derivatives. 700 Compute raffenetti integrals: IOpCl=0 IOpCl=1 IOpCl=2 IOpCl=3 IOpCl=4 1000 2000 3000 4000 Load R1.600 Compute regular integrals and load into R03. 2000000 Form delta-F each iteration -. 10**-N. Compute forces. Make derivative Fock matrices Make derivative Fock matrices and form contributions to polarizability derivatives (ie 6 sets of forces will be returned in FXYZ. 0 N 10**-8. R2. IOp(5/6) Convergence (RMS density except in L506 (SQCDF). and L510 (Energy)). and 3 extra sets of densities must be supplied in PA. 5000 10000 0000000 Default -. in canonical form if IOpcl=0 and square form if IOpcl=1. Do not compute forces. L508(rms rotation gradient). Compute forces using including CIS 2PDM terms.B).only in L502. Load R2 and R03. except 10**-7 for PBC. NMatS is used as the dimension of R0 if IOpcl=1. L510: CONVERGENCE CRITERIUM (ACC) FOR THE ENERGY IN THE MCSCF 0 N Acc = 10**(-8) Acc = 10**(-N) . except 512 in L503 and L508. Scaled steepest descent. Options other than the standard SCF ones: IOp(5/8) SELECTION OF THE PROCEDURE OF DIRECT MINIMIZATION (L503). TAU (G18. L510: MAXIMUM NUMBER OF ITERATIONS TO BE DONE (MaxIt) 0 N -1 MaxIt=64 (Default Value) MaxIt=N It does only a CI calculation. 0 128. TAU (G20.5) SCALING FACTOR FOR SUBSEQ. 0 1 STEEPEST DESCENT WITH SEARCH PARAMETERS DEFAULT STEEPEST DESCENT WITH SEARCH PARAMETERS READ (SEE BELOW) 2 CLASSICAL SCF (ROOTHAAN'S METHOD OF REPEATED DIAGONALIZATION 4 5 CONJUGATE GRADIENTS WITH SEARCH PARAMETERS DEFAULT CONJUGATE GRADIENTS WITH SEARCH PARAMETERS READ: MAX.5) Search method (L508).5) Q (G20. NUMBER OF SEARCH POINTS (I1) INITIAL STEPSIZE. 0 1 2 Default (123).IOp(5/7) Maximum number of iterations. NUMBER OF SEARCH POINTS (I1) MIN. Steepest descent. . CRITERION READ IN (FORMAT G16. coefficients are frozen at initial values (L504: causes coefficients to be read in in order 11 12 22). Read Fock matrix restriction matrix. Do a linear search only if the energy goes up after the initial step. IFrzCI. Take pure NR steps. Default handling of wrong curvature (switch direction). IOp(5/9) SWITCH TO CLASSICAL SCF AFTER DENSITY MATRIX HAS ACHIEVED A CERTAIN CONVERGENCY (L503 only). Do a full linear search to locate a minimum. read damping coefficients. even if curvature is wrong. CRITERION DEFAULT 10(**-3) YES.10) Number of pair iterations (L504. . Use Levy damping. Flags for L510: 1 10 100 1000 10000 1000 IRdF2. Default linear search (full search). 0 1 2 NO YES.3 00 10 20 000 100 200 Quadratic convergence (after rotation gradient is sufficiently small). Read in damping factors from cards. L506). Read unformatted symbolic matrix elements from NDATA instead of rwf. Reverse direction if curvature in NR step direction is wrong. -1 None. freeze CI coefficients after 1st iteration. Number of GVB pairs (L506). and if energy DIIS is turned on.0 L510: 5. Yes.001*N Dynamic level shift to a default goal (same as -200) No level shifting. 1 THREE-POINT EXTRAPOLATION IS INHIBITED. BUT THE PROGRAM WILL STILL PERFORM FOUR-POINT EXTRAPOLATION WHEN POSSIBLE.L502. (IE. If non zero. IOp(5/10) IVShft Level shifting: -N -2 -1 Dynamic level shifting to achieve a gap of -0. 0 BOTH 3-POINT & 4-POINT EXTRAPOLATION PERFORMED WHEN APPLICABLE. -1 for sparse diagonalization replacements. L503 has only 4 point. the number of orbitals in each pair is read in format (30I2). L505). N Shift by 0. IOp(5/12) Whether to allocate only two N**2 arrays for RHF: 0 1 2 Default (No). 0 Default: -200 for diagonalization calculations. DISABLED). Each pair consists of the highest available occupied from the guess (after high spin orbs are accounted for) and the lowest .001*N IOp(5/11) 3 and 4 Point Density extrapolation control (L501. No. 100 ADMP.available virtuals.-For testing purposes. Use Generalized energy-weighted density routines regardless. .(Calculates directly Nact*(Nact+1)/2 Fock matrices by contracting the AO integrals with the Density matrices. pair coefficients are read. L510: MCSCF flags: 2 Generate MOs using UHF natural orbitals.turned on automatically in FoFDir100 INFC Number of Frozen Core Orbitals XX000 IRdNT. 1 Continue the run even on non-convergence. Number of rows in an initial transformation of MO. Turn the current RHF run into a uhf run at the end of this link. 10 IRdNLp. IOp(5/13) Action on convergence failure (L502): 0 Default (2). 2 Terminate on non-convergence. Just recompute band structure from stored real-space Fock matrix. later cycles: transform the density from L121 before calculating the energy and Fock matrices. The ILSW flag for convergence failure is set. Terminate after computing the 2e terms at the first iteration. otherwise standard initial values are used. 200 1000 ADMP. first cycle: use initial AO densities. (More input from cards See below) IOp(5/14) Special functions in L502: 0 1 10 20 None. If <0. 2 CALCULATION IS PERFORMED. 0 1 2 ON BESSEL CRITERION ON STRONGER INDIVIDUAL-OVERLAP CRITERION OFF L510: Flags for MCSCF: 1 10 100 1000 10000 Skip valence-valence Fock matrix elements. Skip valence-virtual Fock matrix elements. Use full diagonalization method rather than Lanczos. IOp(5/15) Apply Abelian symmetry constraints on orbitals. 1 CALCULATION IS BYPASSED. Skip core-valence Fock matrix elements. use IOp(17)). AND THE SYSTEM RW-FILES FOR THE APPROPRIATE DENSITY MATRICES ARE UPDATED (USEFUL IF ONE WANTS A POPULATION ANALYSIS). Skip core-valence Fock matrix elements. 10000 Fit the converged density even if fitting is not in use during the SCF. .2000 Do not use GEW routines even for CP. 0 CALCULATION IS PERFORMED (PROVIDED OF COURSE THAT ENOUGH SPACE EXISTS IN THE RW-FILES). L503). CONTROL OF ANNIHILATION OF SPIN CONTAMINANTS (L502). CONTINGENT ON SPACE. 100000 State average density matrices. REORDERING OF THE ORBITALS (MAINTAINING CONTINUITY OF THE WAVEFUNCTION ALONG THE SEARCH PATH. (Obsolete. Also redoes the fit at the end even if using fits during SCF. Sign Matrix Method. CONTROLS THE AUTOADJUSTMENT OF TAU (L503). No. 00 10 20 Default (use Abelian symmetry in diagonalization). 4 for sparse). keep overall wavefunction the same as the initial guess. Use Abelian symmetry in diagonalization. . keep occupation of each irrep the same as the initial guess. Yes. 0 1 DONE TAU IS KEPT FIXED IOp(5/16) Diagonalization method (L502): 0 -N 1 2 3 4 5 6 7 8 Default (1 for full matrices. CEM. CGDMS. Do not use Abelian symmetry in diagonalization. SNRDMS. 3 Yes.0 1 2 Default (1 for L502. Pseudo-diagonalization with real diagonalization every Nth cycle. 2 for L501 and L506). KyDiag. but doing the minimal amount of orbital switching to accomplish this. PDM. DiagD. Pseudo-diagonalization whenever possible. Force formation of the Fock matrix using sparse storage. DEGEN) M.O.14) Selection of virtual orbitals (L506). INHIBIT PERFORMANCE OF MINIMIZATION OF ALTERNATE WAVEFUNCTION PROVIDED BY SECOND ORDER PROCEDURES (L503).(ILzVec=-1) See below C(1) = 1. 0 1 Virtuals obtained by diagonalization of hamiltonians. SET TO ZERO IF Abs(F(I.0 IOp(5/17) CONDITION OFF-DIAGONAL TERMS OF THE Fock MATRIX (L503). Lanczos starting vector in L510: -1 0 N Read in eigenvector. DEGEN=2. 0 1 NO YES Selection of OCBSE vectors (L506). By orbital least change.D-10.J)). DELETE COUPLING TERMS BETWEEN ALMOST DEGENERATE (DELTA E .LE.0 C(N) = 1.D-5 FUZZY AND DEGEN READ IN (2D20. VECTORS 0 1 FUZZY=1. By energy least change.LE. Use of symmetry (in L502 and L508) and linear equation convergence (in L508): .1xx 2xx Force formation of the Fock matrix using full storage.FUZZY. 0 1 2 By eigenvalue. Virtuals obtained by Schmidt orthogonalization to occupieds. Force the density matrix to have full symmetry at the first iteration. L510: MCSCF flags. 10 If 2E symmetry is on. choose between replicating integrals and symmetrizing the Fock matrix based on whether the current density matrix is symmetric. 20 If 2E symmetry is on. Bad for 1st order method. symmetrize them to ensure that they are exactly symmetric. Don't orthogonalize. 2nd order iteration at end. symmetrize Fock matrices and require proper density matrix symmetry. Just schmidt. Choose LinEq convergence based on orbital gradient. . replicate integrals so that density matrices and wavefunctions need not be symmetric. Always use tight convergence. 1012 for 508). 0 1 2 5 10 100 200 Orthogonalize C. Force the density matrix to have full symmetry at every iteration. 30 If 2E symmetry is on. in preparation for CPMCSCF. Lowdin orthogonalize C+O and V. 40 100 200 Same as 30 in 502 but 20 in 508. then Schmidt. Don't use natural orbitals each iteration.V by separate Lowdin. then Schmidt. Tighten convergence by an extra factor of 10. 0000 Default (1000) 1000 If the density matrices pass the symmetry test. Use full 2nd order convergence.O. 2000 Do not symmetrize the density matrices.0 1 2 3 Default (1032 for 502. J)).O. 30000000 Do SA and prepare for Gradient of Energy difference. Warning!! should be used only for large jobs where Hessian does not fit in memory. Default (-1 unless reoptimizing during Stable=Opt). Turn off damping.LE. Read CI vector and use it every iteration. 500 2nd order iteration using RFO type step + level shift and prepare for non-direct CPMCSCF 10000 100000 Attempt to control root flipping in CI.8) 20000000 Do SA and prepare for SA-CPMCSCF. Dynamic selection of density damping based on band gap and DIIS error.(the weights 8F10.D-7 . L503: CUTOFF CRITERIA IN SYMMETRY DETERMINATION OF M. SYMMETRY IS DETERMINED IF LARGEST OFF-DIAGONAL M. 10000000 Use State Average density matrices. SPAN=5.SPAN ARE CONSIDERED TO BE ZERO 0 STHRS=1.300 2nd order iteration using RFO type step + level shift. N/100 new density. 400 Prepare for CPMCSCF(FREQ): Direct method with no storage of Hessian.(IRdCIV) 1000000 Use full diagonalization method rather than Lanczos.STHRS ELEMENTS Abs(F(I. FOCKMATRIX ELEMENT Abs(F(I.O.S. 40000000 Do SA and prepare for SA Second Derivative Computation (terms involving 2nd order orbital rotation derivatives not included) IOp(5/18) L502: Mixing when doing damping: -3 -2 -1 0 N MO damping at all iterations.GE.D-4. (100-N)/100 old density.J)). 10**-N. scaled steepest descent is used): 0 N Default (1.IBuf2E). Save generated integrals on disk (file 610).T.d-2). Force Lower-triangular in-memory storage. L502. otherwise 1.1 STHRS AND SPAN READ IN (2D20. . do not convert to in-core if direct and enough memory for in-core is available.ne. Force Square in-memory storage. IOp(5/19) Over-ride integral storage control (L501. 0 1 NO YES IOp(5/20) Final non-DIIS iteration (L501.END 2I2) (L503). L508): -1 0 Choose the best given amount of memory available. (READ ONE CARD WITH START. L502. 2 3 4 1x 2x 3x Force allocation for 1 or 2 buffer case conventional case (VV.14) Damping (L506) Maximum rotation gradient for Newton-Raphson in L508 (above this value. 2 if possible. L506. L504). Do not save integrals (same as 0x). Force computation of raff 1 and 2 integrals even for RHF. PRINT F(1). 1 Forbid in-core: force re-reading of integrals even if they fit in 2 buffers if conventional. 0 Default (no). Extrapolation control in L506. 0 1 ABORT RUN VIA LNK1E. IRdNLp. MCSCF flags: 2 10 Generate MOs using UHF natural orbitals. No. just quit when extrapolated convergence is reached. Orbital rotation control (L506). IOp(5/22) Use of DIIS extrapolation (L501. L502. CONTINUE RUN. do a final unextrapolated diagonalization after convergence is reached. 1 No. L504). maximum virtual mixing for MO damping: For density damping: 0 N Default (Damp if error > 0. ACTION IF OTEST DETECTS PROBLEMS (L503). IOp(5/21) DIIS error for density damping. .001) Damp if error > 10**-N For MO damping: 0 N Default. Maximum N/1000 virtual component.1 2 Yes. 0 Default (1042) for calculations using diagonalization (2) for calculations using sparse diagonalization replacements. no more than 1/3 virtual component for any occupied at each iteration. N Rotations are turned on when SQCDF is below 10**(-N). . used (DIIS error/10^-N) for weight of energy DIIS in method 4. high spin and first natural orbitals. 0 Optimize all orbitals. Orbital mixing control in L506. Yes. mixture of energy and commutator. deciding on the fly between the two forms. otherwise. 0 1 2 Read from input stream. L509). Optimize only 2nd and higher naturals. IOp(5/24) Orbital freezing (L506). Mxxxx Use print level M in DIIS. so that pair and hamiltonian information can be reused (L506. with "FON" version of Fermi broadening. Yes. IOp(5/23) Flag for later points of an optimization. Nxxx Switch from energy to commutator when error is 10^(-N) in method 3. Regular DIIS Energy-based mixing Energy DIIS when DIIS error has increased significantly or is above threshold 40 Energy DIIS when DIIS error has increased significantly. with Fermi broadening as well.2 3 4 5 10 20 30 Yes. Read from rwf. with "pFON" version of Fermi broadening. 1 Freeze all closed. Yes. 1xx Use energy DIIS when commutator gives huge coefficients. Read from chk. IOp(5/26) Type of calculation (L504). IOp(5/27) Whether to do closed-shell calculation in L502. GVB as CAS(2. an open-shell singlet is assumed. closed for Multip=1.2) GVB(1/2) Orthogonal open-shell singlet.2) Excited singlet as 2nd root of CAS(2. Apply rotations sequentially. Force closed shell.2). Force UHF. If -1. Closed/Open control for L511: 0 1 2 Default. 0 1 2 Default (Yes. 3 2 1 0 -1 -2 3rd root of CAS(2. IOp(5/28) . If zero. error if Multip>1.IOp(5/25) Rotation application (L506). the unpaired orbitals are assumed to be high spin. No Yes (used for RHF direct SCF). Number of hamiltonians to read in (L506). 0 1 Default (exponentiate rotation angles). ROHF Triplet (a debugging option). if mulitplicity 1). It will never use Raffenetti for SCF. N Integrals with degree of contraction greater than or equal to N are done are regular integrals. L510): 0 Default (No). . All integrals are done as regular integrals. Default: let FoFDir decide. Yes. IOp(5/32) Sleazy SCF (L502. IOp(5/30) Whether to symmetrize final orbitals using abelian symmetry operations (L502. not needed in L506).L510: Root of CI to use in MCSCF (IState) 0 -1 N Defaults to Istate=1 Read IState from cards (see below) IState = N IOp(5/29) Use of rafinetti integrals during direct SCF. IOp(5/31) How many vectors to form at a time during microiterations in L508 (NYI) and L509: 0 N Default (3 in L509). 0 1 2 Default (Yes). L505. N. -1 0 1 All integrals done as Raffenetti. No. 0 ONLY SUMMARY RESULTS ARE PRINTED (WITH POSSIBLE CONTROL FROM THE 'NO. 2 3 4 5 No. 00000 Use general DBF logic only if the DBF rwf is present. O. then one iteration with next grid up. doing an inexpensive pass 0 and then full accuracy in pass 1. BUT AT THE END OF EACH ITERATION. Yes. 2 SAME AS IOp(33)=1. 00 N00 I000 No longer used. IOp(5/33) PRINT IOp(33) PRINT OPTION. 1 THE EIGENVALUES AND THE M. 10000 Force use of 1c instead of general DBF logic. Decide between 1 and 4 based on details of the calculation.1 Yes. 3 4 SAME AS IOp(33)=2. No longer used. use loose integral cutoffs. BUT ADDITIONALLY THE DENSITY MATRIX IS PRINTED.PRINT' OPTION). COEFFICIENTS ARE PRINTED AT THE END OF THE SCF.) . BUT ALL MATRIX TRANSACTIONS ARE PRINTED (BEWARE!!! MUCH OUTPUT EVEN ON SMALL MOLECULES. SAME AS IOp(33)=3. Thresholds similar to DGauss for convergence and integrals. 0=normal 1=Linear approximation to Xc. 20000 Force use of general DBF logic. The default is CoarseGrid for iterations and SG1 for final energy. 6 Do iterations with sleazy XC grid. Use approximation I. convergence on either energy or density and always do incremental Foc formation. do integrals at same accuracy as convergence until final iteration. Default (every 20 for direct). do integrals 3 digits more accurately than current convergence. IOp(5/36) Whether to checkpoint after every SCF cycle. IOp(5/37) Frequency at which to do full Fock formation instead of incremental (L502). IOp(5/35) Whether basis is orthonormal (L501. Don't checkpoint. Yes. No. Every Nth cycle. 0 1 2 Default (No). 3 Yes. No. then 2 digits more accurately. IOp(5/38) Whether to vary integral cutoffs during direct SCF: 0 1 2 Default (same as 1). Checkpoint. Yes. . 0 1 2 Default (checkpoint only if direct). L502). -1 0 N Do not do incremental Fock formation.IOp(5/34) DUMP OPTION. REGULAR SYSTEM DEFAULTS APPLY HERE. 4 5 6 Converge to 10**-5 with integrals good to 10**-6 first. only used now for Onsager and control of details of SCIPCM -N 0 N Multipoles of order N. REQUIRE 'NOFULLDIAG' Remember: the first digit indicating the type of function to be used. decide based on details of problems. Default is DON'T EXCLUDE any integral yy000 Use cutoff = 10**(-yy) on the product Integr*DenMat. then full convergence. IOp(5/39) New On-Fly symbolic matrix element generator. Default is DON'T EXCLUDE any integral 100000 Lanczos 200000 Use sum of the first IState roots of a Reduced Hamiltonian as guess for Use IState-th root of a Reduced Hamiltonian as guess for Lanczos 300000 Save first IState roots on disk for Davidson (this option will automatically call Davidson instead of Lanczos) 1000000 2000000 Print S**2 Print S**2 and its orbital components IOp(5/40) Use of reaction field. VarAcc forbidded because of guess=read. 1 2 3 Hartree-Waller functions for singlets Hartree-Waller functions for triplets Slater Determinants xx0 Use cutoff = 10**(-xx) on integral value to exclude contributions. Multipoles of order N. must be set. VarAcc allowed. increment field in Gen(2-4) No. store field in Gen(2-4) . allows different default actions for PBC. convergence only on maximum). IOp(5/41) Whether to converge on maximum density change as well or instead of RMS: 0 N 22. 40000 Same as 3.e. L510: Davidson options. but re-use 1e matrix instead of surface terms. 20000 Update surface every iteration in pass 1 only. -1 Maximum allowed changed is same as RMS (i. WARNING !!! ratio (zz+k-1)/k must be equal to n. 30000 Update surface on pass 2 iterations only.. 10000 Update surface every iteration.00000 Default (same as 10000). Default=60 Number of vectors provided in input BEWARE !!! Davidson will execute zz updating per iteration.50) yy00 zz0000 Maximum dimension of iterative subspace. Maximum allowed change is 10**N larger than RMS. number specified in nroot=n. 50000 Update surface and restart DIIS when within 10**-2 of convergence. Option xx is used also by Lanczos if IOp(39)=10000n or 20000n xx Maximum dimension of reduced Hamiltonian used as guess Default=Min(NSec. -2 Converge only on RMS density change. Default=IVEC k000000 Reduction factor between number of guess vectors provided and number of vectors wanted at the end (1<=k<=9). Default=1 (no reduction) . N0 Converge on energy to 10**(N)*RMS-density-accuracy Also control of iterative diagonalization in L510. IOp(5/47) . Uses MC-SCF density to compute B88 + LYP energy (These are hard-wired since they were the only choices that gave sensible results) 2 Replaces diagonal elements of MC-SCF CI with B88 + LYP energy IOp(5/45) Numerical Derivative Coupling calculation(for testing) 0 1 No Yes (Needs NonStd root and two cards in input stream): i3 the other vector which coupled with iVec. If negative reads the vector from rwf.8 f10. value must be 0<=ll<=20 Default=0 IOp(5/42) Number of orbitals to localize in L510 1 n Localize all active orbitals Localize first n (strongly occupied!) orbitals IOp(5/43) L509: Whether 5th order terms are treated explicitly 0 1 2 Default: set to 1 All 5th order terms are treated implicitly (Debug option) 5th order GG and WG terms are explicitly computed in L715 L510: DFT corrections to MCSCF on last iteration 0 No 1 Yes. If positive reads vector from input 4f20. 10 Include the CSF contribution to the orbs for the DerCpl.7: the displacement in geometry in internals in Angs.ll0000000 Davidson iteration at which to scale back the number of vectors WARNING !!! For overflow reasons. 0 1 2 10 20 Default (11. L510: Option for using reorthogonalization procedure in Lanczos 0 1 No Yes IOp(5/49) Use of sparse storage and Conjugate Gradient optimization instead of N**2 memory and diagonalization. 1 to set up for CAS-MP2 or 2 to do spin-orbit calculation. Diagonalization Conjugate gradient. L510: Option for using lanczos in CPMCSCF calculations 0 1 2 No Yes Use lanczos except for the last iteration IOp(5/50) L510: Option for setting the maximum number of lanczos iterations in Direct CPMCSCF IOp(5/51) . Linear storage (only in Fock formation if diagonalization). 1 2 Prepare data for Mp2 (l906 obsolete) Compute transition spin density and SO coupling IOp(5/48) Options to be passed to CalDFT: N Control flag for CalDFT is N. or 22 if sparse is set in ILSW).In L510. Square storage (only in Fock formation if CG). N>0 N<0 Solvent type N. Speeds up convergence in CP-MCSCF 1 Do not Canonicalize (turn this on to maintain compatability with previous versions of code. N words. default parameters Dielectric constant |N|/1000 IOp(5/55) How many HOMOs and LUMOs to solve for after CG: 0 N None.L510: Canonicalize MC-SCF orbitals by diagonalization of Core and Virt Fock operators. -1 0 N None Default. L510: configuration cutoff for mp2 0 i . N N/1000 Bohr. L510: see below IOp(5/56) A0 for Onsager SCRF. also none. 0 Yes canonicalize. N of each. .) IOp(5/52) Amount of memory to allocate to stashing integrals.1 Float(1/i) IOp(5/53) PCM input and solvent type. 0) to (0. which is usefull when a trajectory is started at a degenerate region. The state averaging will be switched off (graduately) when the degenerate region is left.050 IOp(57) Threshhold for switching off. this can cause jobs sensitive to SA to fail to converge The 0 1 (Default) No action Turn on new more flexible SA options: The old method switches .0-1. L510: See below IOp(5/55-58) L510: Switching on a State Averaged calculation graduately.5). In that case.) When IOp(55).5-0. the calculation is started with coefficients (0.0-1.001 . IOp(56) Threshhold for switching on.L510: See below IOp(5/57) First iteration at which to level shift and do FON. Default is 0. the number of steps for switching on/off is -IOp(55). (Or reversed. instead of a State Averaged second derivatives calculation.0).lt.075 IOp(58) When set.1 unless doing stable=opt.0. 0 Default .5-0.5). Energy difference larger than IOp(57)*0. Usefull for optimizations or trajectory calculations where only a part of the surface is (nearly) degenerate. IOp(55) Number of steps over which the SA coefficients are brought from (0.0) (default) 2: Allways SA second derivatives THESE OPTIONS MUST BE SET IDENTICAL FOR OVERLAY 10! (only IOp(10/55) needs to be set in overlay 10) IOp(5/59) IOp(59) SA on when you reach certain energy gap. Default is 0. 1: Normal computation when (0. regardless of weather that continues to decrease after the switch point.001 . when the SA calculation graduately is switch off. Energy difference smaller than IOp(56)*0.0-1. in link 1003 a normal frequency calculation is performed when the optimization is in a region of (0. then start after instability searches. 0001 (checks Offdiagonal element).1 >limit Note: trajectory option set in overlay 10: IOp(10/99)=n -. .8 >limit 7 0. IOp(5/65) Over-ride NFx parameter.new method checks the gap each cycle and will decide to increase or decrease the SA based on that result. IRanWt. 0 1 Use global default. IOp(5/64) Over-ride default value of FMFlags 0 N No.5/0. for a hypothetical example with nroot=2: cycle old SA new SA Egap 4 0.62) Override standard values of IRadAn.5 0. Yes.7 <limit 6 0.2/0.1/0. and IRanGd.3/0. only used with IOp(97)=11 IOp(5/60 . Turn off FMM here regardless.6 0.8 <limit 5 0. 0 N No.9 >limit 8 crash 0.3/0. use N. use N.8 0. IOp(5/63) Whether to do FMM.4/0.Threshold for a determining an adiabatic hop threshold=n*0.0/0. Yes.2/0.2/0. IOp(5/70) Maximum Initial temperature for FON (non-PBC).7 0. or temperature for broadening (PBC and IOp(74)=[1-4]xx). -2 -1 0 N None Start at a high temperature (limited only by DIIS error). Use LT method (interpolation) Occupy entire points (used together with broadening) Full points for insulators. Use Cholesky basis for CP orthonormal transform. IOp(5/74) Type of k-point integration: 0 1 2 3 9 10 20 90 Default (911). 6000K for PBC) N degrees IOp(5/71) Number of steps to apply simulated annealing (L502): 0 N Default -. Occupy lowest NE at each k point regardless of the energies Improved LT with quadratic corrections Original LT method No concern for corrections .10 steps FON / 20 steps PFON N steps IOp(5/73) Options for ADMP: 0 1 2 Default (2) Use Lowdin basis for CP orthonormal transform. Default (3000K = 10 milliHartrees for non-PBC. temperature broadening for metals. 0 Default.100 200 300 400 500 900 Smearing Marzari method I Smearing Marzari method II First order Hermite-Gaussian of Paxton and Methfessel Gaussian smearing Classical Fermi-Dirac broadening No broadening (this will be Gaussian broadening with small T) IOp(5/75-78) Number of alpha electrons. IOp(5/79) Range around Fermi level where temperature distribution will be applied if broadening is turned on for PBC. Maximum Number of purification cycles per CG iteration. beta electrons.NN MMNN IOp(5/81) Conjugate-Gradient Parameters MM NN00 cycles).8) Step size of MM. (defaults to 4 CG cycles). (defaults to 3 Don't use CG DIIS Use CG DIIS . alpha orbitals. IOp(5/80) The maximum conjugate gradient step size -1 0 No maximum step size Default maximum (. a value will be chosen in ZInLT1. 00000 10000 Maximum Number of CG cycles per SCF iteration. and beta orbitals for fractional occupation. G. No. -1 0 N Default for first step (128). IOp(5/82) C. No preconditioning. Limit is N cycles. IOp(5/85) Over-riding of maximum cycles for XQC. Restart using Fock matrix in L502. Yes. 0 1 No. IOp(5/86) Strategy options . Convergence criterion 0 N Defaults to 10**(-7) 10**(-N) IOp(5/83) Maximum SCF DIIS vectors 0 N Default (20).000000 100000 0000000 1000000 Polak-Ribiere CG minimization Fletcher-Reeves CG minimization Use diagonal preconditioning in Conjugate-Gradient. Use SCF DIIS with N vectors IOp(5/84) Restart in L509. Just continue as usual if energy goes up. Level shift to a maximum of the Goal.1) 6000 Level shift only if the HOMO-LUMO gap is zero or insignificant (>-0. set in FoFDir/FoFCou/CalDSu based on accuracy part of IOp(5). 200000 Retain 3 and 4 point extrap. if DIIS is on. Level shift as much as necessary for HOMO>LUMO. . Turn on dynamic level shift from the beginning Turn on dynamic level shift only after FON is over. Default (1 except during Stable=Opt. Reduce DIIS space when energy goes above the lowest energy. Level shift only if the HOMO-LUMO gap is zero.000000 0 1 2 3 4 10 20 100 200 1000 2000 3000 4000 5000 Default (101100). then 4). The energy is only checked after FON has been turned off.1). Reduce DIIS space whenever energy is above the lowest energy. Reduce DIIS space when energy rises from previous cycle. up to twice the goal N0000 No longer used. Level shift to a maximum of 2*Goal. Typically 10^-10 for molecules and 10^-12 for periodic systems. 100000 Turn off 3 and 4 point extrapolation if DIIS is on. IOp(5/87) Accuracy criterion in Fock matrix formation: 0 Default. Turn off level shift after energy rises. Keep level shift after energy rises. Level shift only if the HOMO-LUMO gap is zero or insignificant (>-0. 1 Yes. IOp(5/91) Control option for chebyshev sparse control. variable=kPrint 0 1 2 Print only summary information Print the a(t) vector and probability for each csf Print almost everything for debugging 3 Print everything for debugging warning! this is a lot of stuff and you will only be able to do a few cycles IOp(5/89) Linearly dependent basis control for PBC. this and ZFormV should be moved to L302. separate Coulomb and NFx exchange for PBC). separate FoFDir for exchange. IOp(5/90) Whether to generate sparse guess here. do preliminary AM1 calculation. FoFCou for Coulomb. separate FoFCou/NFx for exchange. IOp(5/88) L510: controls the amount of printing. IOp(5/92) Whether to use FoFDir or FoFCou for exact exchange: 0 Default: normal processing based on FMM for non-PBC. L510: flag hopping controls starting and stoping options (x=0 or x=1) . 2 Yes. 1 2 FoFCou for Coulomb.N 10**-N. do preliminary AM1 calculation and compare with guess from previous step in geometry optimization. L510: has different meaning depending on if you are using IOp(97)=22 or IOp(97=23) If IOp(97)=22.25 in GS) and if Zyy=125.01) Determines a variable Ulimt which lies between 01 and 99 Z Determines if this probability is halved after the first hop default Zyy=25.variable=iBack for first hop being up: xxxx0 Hopping down (forward) xxxx1 Hopping up (backward) variable=iStNow xxx0x xxx1x variable=iSpace xx0xx xx1xx Use Energy gap criteria to start timedep Start timedep imediately Use full space CI basis Use reduced space in projection of alpha variable=iEnd number of cycles to carry on before stopping the timedep code after exiting the hopping region x0xxx Default stop 6 iterations later xNxxx Stop time dep on cycle N after exit (if IOp(92) is negative then stop immediately) variable=iFcTD to stop the time dependent code on cyle Z 0xxxx No effect Nxxxx Stop time dep on cycle N allowable values 1-9 IOp(5/93) Number of initial iterations for which damping is allowed: 0 N Default (10). threshold for a hop up (after a hop down) = 0.01) or the lower state (=(1-x)*0. N iterations.75 in ES (or 0.125 in GS . where: Zyy yy Threshold for a hop down determined by probability of being on the upper state (=x*0. Value = xxxzyy. threshold =0. wait time is xxx*0.04 switch timdep on. default yyy=5 ie deltaE>0. on the model system L510: Threshold for turning propagation method on and off yyy xxx The first three digits determine the energy gap for turning off The last three digits determine the energy gap for turning on Threshold =xxx*0.01 (checks energy gap).xxx How long to wait before checking for a hop after going through a hop. which determines what type of gradient calculation to do: xxN Choose the basis variable=mBasis 0 (Default) same as 1 1 Use a(t) basis 2 Use mcscf basis orthogonal to a(1) and a(2) 3 Use currnet mcscf vectors to check the code must be used with mHTest=2 Do with diagonalisationof Ecc variable=mTDGrd: 0 (Default) same as 1 1 Only check Ecc if MCSCF energies are almost degenerate 2 Force check on Ecc by diagonalising it xNx N00 Testing options.05 switch timdep off only used with IOp(97)=22 and IOp(97)=23 IOp(5/95) Option for using Davidson in RFO calculations . variable=mHTest 0 (Default) no testing 1 Testing construction of ecc 2 Testing construction of the ecc or exx portion of hessian using the mcscf vector information 3 Calculate but do not use TD gradient IOp(5/94) PCM/ONIOM calculation 0 1 2 Standard PCM calculation PCM/ONIOM calcn. value=xxx. on the real system PCM/ONIOM calcn. variable=IWait. default xxx=4 ie deltaE<0.1 fempto seconds If IOp(97)=23. Hopping threshold during trajectories with L510. Yes. Ras1 orbitals are doubly occupied. 0 N Use default. Ras3 orbitals are empty. and the maximum number of electrons in the Ras3 space: zzyyxxww. but the hop is determined by the diabatic or adiabatic criteria (whichever determines a hop first). Ras2 and Ras3. if available). We also need to define the maximum number of holes in the Ras1 space (ie the number of electrons that can be excited out of the Ras1 subspace. Needs . Vector following or Root following hop alone (Needs option 80) 11 Make a hop based on the secular equation (adiabatic hop) this option also includes a hop decision based on the vector following method (diabatic hop) (Needs option 99 and 80) 21 Debugging Option: Propagation of the wavefunction is switched on. No. RAS control in L510.0 1 Yes No use Lanczos (not recommended) IOp(5/96) Over-ride IRadAn for CPHF-like step in L509. 0 1 2 10 Default (Yes. and for pass 0 grid in L502. Use grid N. The CAS active space is subdivided into three RAS active subspaces. where ww xx yy zz Number of Ras1 orbitals Maximum number of holes in Ras1 Number of Ras3 orbitals Maximum number of electrons in Ras3 IOp(5/97) Whether to update precomputed grid data with timing information. In the reference space. Ras1. No. 0 -1 Default (10) No test N>0 Abort of SDif is larger than N.12.8 for 1. IOp(5/102) . No. the molecule continues in a mixed state (Needs option 71 and 80) IOp(5/98) Whether to save eigenvalues and orbitals at all k-points. Yes. 0 Default: 32. 22 Propagation of the wavefunction. for detailed control of the propagation conditions see options 92-96. IOp(5/99) Grid for numerical k-integration in FT-LT method.2.option 80 and option 99. Yes. hop is based on the probability of being in a specific state (Needs also option 80) (detailed control is determined by options 92-96) 23 Propagation of the wavefunction but no hopping. 0 1 2 Default (No).3d IOp(5/100) Tight convergence during CGDMS: 0 1 2 Default (No). IOp(5/101) SDif test on numerical accuracy of PBC diagonalization. N. . 0 NO ADDITIONAL CENTERS. L510 Notes These options must be set in multiple links: L1003 iop(97) yes yes yes no no no yes L510 L118 iop(55-59) yes iop(80) yes These options must be set for the following links: IOp l118 l510 l1003 55-58 no yes yes 80 yes no yes 97 no yes yes 98 no no yes 99 no no yes Overlay 6 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 70 71 72 73 74 75 76 77 78 79 80 81 82 IOp(6/15) SPECIFICATION OF ADDITIONAL CENTERS. 1 READ ADDITIONAL CENTERS. IF MORE THAN ONE OF THESE IS REQUESTED.Maximum number of configurations for CAS-MP2: 0 N Default (1000). EVALUATE THE PROPERTIES ONLY AT EACH ATOMIC CENTER. ONE CARD PER CENTER WITH THE X. Y AND Z COORDINATES IN ANGSTROMS (FREE FORMAT). THE LISTS ARE IN SEPARATE INPUT SECTIONS IN THE ORDER LISTED BELOW. ZO 0. An arbitrary list of points. STARTING AT EACH POINT. B. Two forms of specifications are allowed: A.XZ-EFG YZ-EFG Note that either form of grid should be specified with respect to the standard orientation of the molecule.output unit and coordinates of one corner of grid.Z1 N2.Z-coord. Three cards are required: KTape.Y1. . -. potential and field (NEFG=2).X1.KTape The coordinates of N points in Angstroms will be read unit LTape in format (3F20.XX-EFG YY-EFG. Only one card is needed: N. with N2 values in each record. LOCATED THE NEAREST STATIONARY POINT IN THE ELECTRIC. 8 Do potential-derived charges.LTape. containing: X-coord.XO.Y-field.NEFG. Thus if NEFG=3 for each point there will be 4 cards written per point.ZZ-EFG. or potential.number of increments and vector.12). Evenly spaced rectangular grid. it defaults to 51.2 READ IN COORDINATES AS FOR 1. and field gradient (NEFG=1) will be computed and written along with the coordinates to unit KTape in format (4F20. N1.Y-coord. POTENTIAL. 4 Read in a set of cards specifying a grid of points at which the electric potential will be computed.Y2.12).Z2 -.Z-field.Potential X-field.number of increments and vector.YO. field. The potential (NEFG=3). If KTape is -. N1 records will be written to unit KTape.XY-EFG.X2. Exact is already in z-matrix orientation. COMPUTE ONLY THE NUCLEAR CONTRIBUTION. 1 Do all points to full accuracy. 0 Use full accuracy in calculations at specific points. IOp(6/20) How to do electrostatic-potential derived charges: . use as-is. This is like IOp(9) in L9999. IOp(6/17) DEBUGGING CONTROL (L602). 0 1 2 -N COMPUTE ALL CONTRIBUTIONS TO SELECTED PROPERTIES. Exact is still in standard orientation. same as 1.16 32 128 Constrain the dipole in fitting charges. IOp(6/18) Whether to update dipole rwf 0/1 yes/no. do not default cube. 0 1 2 Default. IOp(6/16) Cutoffs in L602. Read in centers at which to evaluate the potential from the rwf. IOp(6/19) Whether to rotate exact polarizability before comparing with approximate (which will be calculated in the standard orientation). Read grid. COMPUTE ONLY THE CONTRIBUTION OF SHELL N. COMPUTE ONLY THE ELECTRONIC CONTRIBUTION. so rotate. but use sleazy cutoffs in mapping a grid of points. 100 Read in replacement radii for selected atom types as pairs (IAn. L604): -1x +1x Read density matrices from .0 -1 1 2 3 00 10 20 30 Default (1).Rad). IOp(6/21) Operation of L603: 0 1 2 Default (same as 2). terminated by a blank line. 10000 Use only active atoms in the fit. Read density matrices from .chk file. Optimize density basis set. 1000 Fit united atoms (heavy atoms only) rather than all atoms. Read a list of points at which to fit. Read in density basis functions and compute populations. CHELPG point selection.Rad).chk file. L602. Merz-Kollman point selection CHELP point selection. Default radii are those defined with the selected method. 200 Read in replacment radii for selected atoms as pairs (I.Rad) or (Symbol. Force CHELP (Francl) recommended radii. . terminated by a blank line. IOp(6/22) Selection of density matrix (currently only in L601. Force CHELPG (Breneman) recommended radii. Force Merz-Kollman radii. one per line. Density values and gradients. IOp(6/23) Density values to evaluate over grid in L604: 0 1 2 3 Default (same as 3). IOp(6/25) Whether to compute coulomb self-energy in L601: 0 1 No. gradients and divergence. Transition density between the states given by IOp(29) and IOp(30). Default (Yes). IOp(6/24) Frozen core: -N 0 1 2 Freeze N orbitals. N. Yes.-5 -4 -3 -2 All available transition densities.ge. Density values. classically (including self terms .0 Use the density matrix for method N (see Link 1 for the numbering scheme). . No. Density for the excited state given by IOp(29).requires 2e integrals. O(N**4)). Use all available density matrices. Yes. -1 Use the density matrix for the current method. or the HF density if the one for the current method is not available. Density values. IOp(6/26) Which density to use in L602 and L604: 0 1 2 3 4 Default (same as 1). quantum mechanically (no self terms . Yes: mark as well. Do Mulliken populations. and only available for HF. IOp(6/28) Mark SCF density as current density. Total. Beta. Spin. Do Minimal population analysis.requires 2e integrals. Alpha. 0 1 No: save SCF density. Don't do Mulliken populations. O(N**5)). Do bonding Mulliken Populations. but do not mark. IOp(6/30) .2 Yes. IOp(6/29) Excited state to use if requested by IOp(22). Don't do bonding Mulliken Populations. IOp(6/27) Choice of population analysis: 0 1 2 10 20 100 Default (12). where MItLoc MItLoc*NOrb*(NOrb-1)/2 is the maximum number of iterations in localization of (spin)orbitals (1.. Store only alpha NOs. Transition denstiy between state IOp(29) and state N.9. IDcInt IPrLoc Localized (spin)orbitals with atomic occupancies less than 0. default 6)... IOp(6/31) Whether to determine natural orbitals from densities: 0 1 2 3 4 5 No.s.2nd excited state for transition density: 0 N Transition denstiy between state IOp(29) and g. 0: Print the atomic occupancies of localized (spin)orbitals (default).9. IOp(6/32) CONTROL PARAMETERS FOR COVBON in L609 (NOT TO BE CHANGED UNDER MOST CIRCUMSTANCES): 10000*MItLoc+1000*ITlLoc+100*IDcInt+IPrLoc. using total density. 0 Make this a scratch file. default 9). Yes.. Use spin density.**(-ITlLoc) is the convergence criterion for (spin)orbital localization (1. ITlLoc 10. using alpha and beta separately for UHF..01*IDcInt are interpreted as lone pair MOs rather than bond MOs (1. default 10). L605. 1: Do not print the atomic occupancies. L606: naming of RPAC interface file. Yes. ..99. Store only beta NOs. 10 Compute energies of electrostatic interactions between AIMs (IDoPot). Compute atomic orbitals in molecule (IDoAOs). ring points. 200000 If necessary. which is a default anyway (IHwAug=0). Note that analytical force constants have to be available! Compute localized orbitals and bond orders (IDoLoc). 100000 If necessary. Note that IDoSRe should be set to 1 in order to obtain correct results! Also note that analytical polarizabilities have to be available but force constants have to be absent! 2000 as 10000 20000 Compute derivatives of atomic properties with respect to nuclear displacements well (IDoNuD).11' IOp(6/35) WHAT TO DO: 0 1 2 4 Determine attractors. augment valence electron densities with relativistic core contributions. Compute kinetic energies and multipole moments of AIMs (IDoPrp). attractor interaction lines. Determine zero-flux surfaces (IDoZrF). This precludes calculations of atomic property derivatives with respect to nuclear displacements. Compute charges of AIMs (IDoAtC). augment valence electron densities with nonrelativistic core contributions (IHwAug=1). 100 200 Compute atomic overlap matrices (IDoAOM). and cage points. . 400 Include zero-flux surface relaxation terms in all atomic matrix elements (IDoSRe) 1000 Compute derivatives of atomic properties with respect to electric field (IDoSeP).1 Name this file 'rpac. Compute other atomic matrix elements (IDoAMa). where INoZer 0: Ignore (spin)orbitals with zero occupancies (default). No other properties can be calculated. default 7). Length of Fourier expansion for the trial path (1. default 2).999.400000 Abort if pseudopotentials have been used (IHwAug=3). default 20).9. 1000000 Reduce accuracy so atomic charges can be computed more rapidly (IQuick). RNGPNT. 2000000 Use numerical instead of analtyic integration.. 1: Do not ignore (spin)orbitals with zero occupancies. AND CAGPNT in L609 (NOT TO BE CHANGED UNDER MOST CIRCUMSTANCES): 1000000*MxBpIt+100000*SBpMax+1000*NGrd+LookUp. 3000000 Use numerical instead of analtyic integration and use reduced cutoffs. IOp(6/37) CONTROL PARAMETERS FOR ATINLI.. IOp(6/36) CONTROL PARAMETERS FOR NEGLECT OF ORBITALS AND PRIMITIVES in L609: 10000*INoZer+100*IPrNDe+IPrNAt.**(-IPrNAt) in integrations over atomic basins (099. SBpMax NGrd LookUp Maximum number of iterations in trial path determination (1. IOp(6/38) . This option sets IPrNDe=5. default 100). default 7). IPrNDe electron Neglect primitive contributions below 10.. and IEpsIn=100. default Maximum value of the control sum (1...99.**(-IPrNDe) in evaluations of density and its derivatives (0 99.. IPrNAt=5... IPrNAt Neglect primitive contributions below 10. Number of grid points in critical point search (1.99. where MxBpIt 10). . INInGr 10*INInGr is the initial number of grid points in theta and phi in the adaptive integration subroutine (1.9. IEpsIn 0. default 2). IRScal IRScal is the scaling factor in the nonlinear transformation used in the intersection search (1. default 2). IEpsSf IEpsSf is the safety factor used for patches with surface faults in the adaptive integration subroutine (1. 10.........CONTROL PARAMETERS FOR ZRFLUX AND OIGAPI in L609 (NOT TO BE CHANGED UNDER MOST CIRCUMSTANCES): 100000*INStRK+10000*IHowFa+1000*IGueDi+100*IPraIn+10*IRScal+IRtFSe INStRK 10*INStRK is the number of steps in the Runge-Kutta integrations along gradient paths (1.*IRtFSe is the safety factor used in the intersection search (1...9.... default 6). default 2).9. IRtFSe 2).. default IOp(6/39) More CONTROL PARAMETERS FOR ZRFLUX AND OIGAPI in L609 (NOT TO BE CHANGED UNDER MOST CIRCUMSTANCES): 1000000*IToler+100000*INInGr+10000*INInCh+1000*IEpsSf+10*IEpsIn+INTrig IToler 10.0001*IEpsIn is the target for integration error (1.9. default 2).9...9. .99. IGueDi 10.**(-IGueDi) is the initial displacement from the critical point in the Runge-Kutta integrations (1.**(-5-IToler) is the tolerance for the intersection search (1. default 5).9. default 5).*IPraIn is the cut-off for zero-flux surfaces (1.. default 2). default 2). IPraIn 10. default 6).9. INInCh 5+INInCh is the initial number of sampling points in the intersection search (1....9. IHowFa IHowFa is the maximum distance in the Runge-Kutta integrations along gradient paths (1..9. INTrig 10*INTrig is the number of sine and cosine functions in the trial function for surface sheets (1. points per unit area. 3 Read the deletion energy produced by a previous run with IOp(40)=2 and print it. whose energy can then be computed by one of the SCF links. IOp(6/44) Type of calculation in L604: ..01*N.don't read input. Read input data to control NBO analysis. IOp(6/43) Increment between layers in MK charge fit. IOp(6/41) Number of layers in esp charge fit. 0 1 Default NBO analysis . 2 Delete selected elements of NBO Fock matrix and form a new density. N layers.9. default 2). 0 N Default (4). 0 N Default (0.. IOp(6/42) Density of points per unit area in esp fit. must be >=4. This link must have been invoked with IOp(40) = 0 or 1 prior to invoking it with IOp(40)=2. IOp(6/40) Control of link 607. 0 N Default (1).4/Sqrt(#layers)) 0. 50 is recommended. 0 -1 N Default . N*0.10**-3 Read from input. same as 2. 1 digit better than desired acuracy for volumes). N 10**-N . IOp(6/46) Threshold for molecular volume integration. Compute the molar volume Evaluate the density over a cube of points Evaluate MO's over a cube of points Skip header information in cube file.for tight accuracy. IOp(6/47) Scale factor to apply to van der Waals radii for the box size during volume integration: 0 N Default.for debugging. IOp(6/48) Use of cutoffs 0 Default (10**-6 accuracy for cubes. IOp(6/45) Number of points per bohr**3 for Monte-Carlo calaulation of molar volume -1 0 N Read from input Default (20) N points .0 1 2 3 10 Default. N*10**-4.01 . -N>4 Grid using 1000 / N points/Bohr. Fine grid. on non-SCF densities. Medium grid. IOp(6/52) Number of radial integration points in L609: 0 N Default (100). Coarse grid.IOp(6/49) Approximate number of points per side in cube in l602/l604: 0 N -1 -2 -3 -4 Default (80) N points Read from cards. on non-SCF densities. No. Yes. IOp(6/51) Whether to apply Extended Koopman's Theorem (EKT): 0 N -1 -2 Default (No). up to N IPs and EAs. Yes. 6 points/Bohr. all possible IPs and EAs. . 3 points/Bohr. 12 points/Bohr. IOp(6/53) Distribution of radial points in L609: 0 N Default (cubic) Polynomial of order N. N. IOp(6/54) Maximum number of domains. 0 N Default (100000). N. IOp(6/55) Number of inner angular points in numerical integration in L609: -1 0 N 0 (no inner sphere) 302 N point Lebedev grid (see AngQad). IOp(6/56) Whether to read in density matrix from input stream in L608. 0 1 No. Yes. IOp(6/57) Whether to generate data over a grid using the total SCF density: 0 1 2 3 No. Yes, read in name for output file. Yes, also read in name for input file with a different grid and compare. Output in the form of data statements. IOp(6/58) Grid to use in generating tables of density and potential. Must be an unpruned grid. 0 Default (99001). IOp(6/59) Approximations to Exc -1 0 1 2 3 Test superposition of atomic densities using L608: Do correct energies. Do correct energies and 0th order approximation Do correct energies and 0th-1st order approximations Do correct energies and 0th-2nd order approximations IOp(6/60-62) Over-ride standard values of IRadAn, IRanWt, and IRanGd. IOp(6/63) Suppress number of electrons test in XC quadrature in L608 (for debugging with small grids): 0 1 2 Default (do test). Suppress test. Do test as usual. IOp(6/64) Natural Chemical Shielding Analysis: 0 1 2 3 No. Yes, of isotropic value. Yes, of diagonal tensor elements and isotropic value. Yes, of all tensor components. IOp(6/65) Threshold for printing of NCS contributions. -1 0 Zero. Default (1 pmm). N N/1000 ppm IOp(6/70) Control of L610. IOp(6/71) XC functional in L610. IOp(6/72) Whether to read isotopes for hyperfine interractions and do hyperfine terms in L602: 0 1 2 3 4 Default (1). Yes, if open-shell, NMR data is available, and other terms are being computed No. Yes, regardless of other terms. Yes, reading isotopes IOp(6/73) Whether to save orbitals from NBO: 0 1 2 3 10 Default (No). Save NBOs in place of regular MOs. Save NLMOs in place of regular MOs. Save NLMO occupieds and NBO virtuals. Suppress re-orthogonalization. IOp(6/74) Whether to use Gaussian connectivity in choosing Lewis structure for NBO. 0 1 Default (use if present and choose is selected in NBO input). Use. 2 Don't use. IOp(6/75) model for CM2 charges. IOp(6/76) Threshold for linear dependence in L607. 0 N Default (1.D-6). 10**(-N). IOp(6/77) Restraint in charge fitting in L602: 0 -1 N None. 2.d-4 N * 10^-5. IOp(6/78) Use MOs instead of density in AtmTab. 0 1 2 Default (2). Use density. Use MOs. IOp(6/79) Whether to calculate Hirshfeld charges. 0 1 2 Default (No). Yes. No. IOp(6/80) Whether to calculate Lowdin charges and Mayer bond orders. 0 1 2 Default (No). Yes. No. IOp(6/81) Print kinetic energy of orbitals? 0 1 2 3 Default (yes, if doing other orbital results). Yes, for the top 5 occupieds and lowest 5 virtuals. No. Yes, for all orbitals. IOp(6/82) Tensors for hyperfine spectra. 0 Default, compute if there are 100 or fewer atoms 1 Compute QEq tensors and for open-shell systems compute isotropic and anisotropic splitting tensors. 2 Do not compute tensors. Overlay 7 6 7 8 9 10 11 12 13 14 15 16 18 25 29 30 31 32 40 41 42 43 44 45 52 53 64 65 70 71 72 74 75 76 77 87 IOp(7/6) operation of link 705 (NYI). 0 1 2 Default (12). Do not the ecp contribution to the SCF forces. Form the ecp contribution to the SCF forces. 10 20 Do not form 1e derivative matrices. Increment the 1e derivative matrices with ecp terms. IOp(7/7) USE OF INTERNAL COORDINATES. 0 1 YES NO 2 Yes, but neglect first derivatives in conversion of second derivatives to internal coordinates. IOp(7/8) Harmonic frequency calculation: 0 1 2 3 10 20 30 Nxx Mxxx No. Yes, with most common isotopes. Yes, with read-in isotopes. No. Print higher precision normal modes. Print normal mode displacements in redunant internals. Print both HP modes and internal displacements. Default scale factor is #N (1=HF, 1/1.12, 2=CBS4=0.91671, 3=CBSQ=0.91844) If M=1, only harmonic thermochemistry. If M=2, do hindered rotor analysis. If M=3, Read hindered rotor parameters from input. IOp(7/9) Whether to rotate derivatives back to the z-matrix orientation. 0/1 yes/no. IOp(7/10) First/second derivative control. 0 1 2 do only first derivatives. do only second derivatives. do both. IOp(7/11) control of integral derivative algorithm: 0 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Default use IsAlg to decide. Scalar Rys SPDF. Berny SP, Scalar Rys DF. Old vector Rys SPDF. Berny SP, old vector Rys DF. FoFDir: Rys spdf. Berny SP, FoFDir Rys df. FoFDir: HGP sp, Rys df. Berny SP, FoFDir Rys df (same as 7). FoFDir: HGP spd, Rys f. Berny SP, FoFDir HGP d Rys f. FoFDir: HGP spdf. Berny SP, FoFDir HGP df. FoFDir: PRISM spdf. FoFDir: Berny SP, PRISM df. IOp(7/12) Selection of density matrix. 0 Usual SCF density. N Use generalized density number N for both the one-electron integral derivatives and the corresponding 2PDM terms. IOp(7/13) Contraction with two-particle density matrices: 0 1 2 3 4 5 Default (same as 1). Use HF 2PDM. Use external 2PDM. Use both HF and external 2PDM. Generate 2PDM from CIS square 1PDM (for debugging) Generate 2PDM from CIS square 1PDM and use HF/Z 2PDM as well. 6 Contract with external 2PDM derivatives. The types of derivatives are given by IOp(15). 7 Form derivative 2PDM from CIS and HF derivative density matrices. The types of derivatives are given by IOp(15) 10 Leave the external 2PDM on the disk instead of deleting it. 0-5 imply use of the generalized density in L701, while 6-7 imply use of the generalized density derivatives in L701. IOp(7/14) State for CIS gradients. Defaults to 1. IOp(7/15) The nature of the perturbation(s). 0 IJK Default (1st order nuclear and electric field). Nuclear Kth order. Electric field Jth order. Magnetic Field Ith order. 1000 Generate simulated density derivatives. Only 1, 10, and 11 are valid in overlay 7. IOp(7/16) Number of translations and rotations to remove during redundant coordinate transformations: -2 -1 0 N 0. Normal (6 or 5 for linear molecules). Default, same as -1. N. IOp(7/18) Derivative accuracy option: 0 1 2 10 20 100 200 Compute to 10**(-8) accuracy. DO AS ACCURATELY AS POSSIBLE in L702. USE THE ORIGINAL 'BERNY' VALUES in L702. DO AS ACCURATELY AS POSSIBLE in L703. Use sleazier cutoffs in L703. DO AS ACCURATELY AS POSSIBLE in L708. Use sleazier cutoffs in L708. IOp(7/25) Type of derivatives available. 0 1 2 10 First. Second. Third. Read derivatives from checkpoint file (in Z-matrix orientation). IOp(7/28) SKIP OPTION TO DEFER INTEGRAL EVALUATION TO L703. 0 2 COMPUTE AS NORMAL. DO ALL GRADIENT INTEGRALS IN L703 IOp(7/29) MODE OF USE OF L716. 0 1 2 6 00 Normal, same as 2. Normal + Generate estimated initial force constants. Normal NUCLEAR REPULSION ONLY (USEFUL FOR TESTING). Default method for initial force constants IOp(7/30) USE OF SYMMETRY IN OVERLAY 7: 0 1 USE (SUBJECT TO AVAILABILITY). DON'T USE. IOp(7/31) Handling of forces contributions. 0 Just use the forces in IRWFX. 1 Compute HF forces from D2E file and increment both FX and FXYZ (non-O11 PSCF grad and HF freq). 00 Use FX in conversion of force constants to internal coordinates. (HF freq, PSCF freq=numer). 10 Use FXYZ in conversion of forces constants to internal coordinates (PSCF opt with HF 2nd deriv). IOp(7/32) PUNCH OPTION. 0 None. 1 Punch energy in format D24.16, forces and lower triangular force constants in format 6F12.8. 2 Punch nuclear coordinate derivatives. Forces are punched in 3D20.12 format, one card per atom. Force constants and third derivatives are punched in 4E20.12 format in compressed form. 3 Punch energy, coordinates, and derivatives in cartesians and redundant internals. 4 Punch energy, coordinates, and derivatives in redundant internals only in compressed form. 5 Punch energy, first and second derivatives in both cartesian and internal coordinates. 1x Do punch only if second derivatives are available. IOp(7/40) Neglect of integrals (only option 1 works in Overlay 7): 0 1 2 3 10 20 30 Keep all integrals. Neglect four center integrals. Neglect three center two-electron integrals as well. Neglect 2e integrals with diatomic differential overlap. Neglect three center one-electron integrals. Neglect 1e integrals with diatomic differential overlap. Do only overlap and not other 1e integrals. IOp(7/41) NDDO flag. 0 Evaluate integrals correctly. with forces done the usual way for CIS or MP2 2nd derivatives. (debugging option: compute fifth order WG and GG terms in L715) IOp(7/44) Handling of an applied electric field.true. -1 0 1 Do not add electric field terms to forces.1 Apply NDDO approximation. Update forces for the self-consistent reaction field (SCRF) method 2 Update forces for a uniform electric field. IOp(7/43) 2nd order simultaneous optimization flag. IOp(7/42) 1PDM: 0 N Use SCF total density. 0 1 2 3 Do not project.d-6.true. IOp(7/45) Controlling the projection of the reaction path. Project using the Newton-Raphson step. The point is a stationary point.false. 0 1 2 . Use generalized density N. IOp(7/52) . (other 2nd derivative options must also be set appropriately) . Update forces for a uniform electric field. . Project using forces if the RMS force is larger than 1. Project the reaction path and compute 3N-7 frequencies. IOp(7/53) Convert forces over shells to field-dependent dipole and forces over atoms (for debugging): 0 1 10 No. IRanWt. IOp(7/64) Type of simulated spectrum in output. 0 1 Use global default. Turn off FMM here regardless. Lines Lorenzians Both IOp(7/65) Harmonic constraints with respect to initial structure during geometry optimization. No. Yes.Whether ECP integrals should be done in L701 as usual. -1 No. . IOp(60-62) IOp(60-62) Over-ride standard values of IRadAn. Compute optimimum lambdas. and IRanGd. 0 1 Yes. 0 1 2 3 Default (1). IOp(63) Whether to do FMM. IOp(7/71) Do vibrational 2nd order perturbation: 0 No 1 Yes. Yes. IOp(7/75) Threshold for printing redundant internal contributions to normal mode displacements: 0 N Default (10%) 10**-N . No. IOp(7/70) Do vibro-rotational analysis: 0 1 2 Default (No). Yes. if ref structure is present and has non-zero force constants). Yes.0 1 Default (Yes. IOp(7/72) Read additional parameters for anharmonic computations 0 1 No Yes IOp(7/74) Non-equilibrium PCM gradients: 0 1 No. Currently lots of hacks to determine where we are in the process instead of different values of this option. Turn off 1c logic for 1c DBF case. Omit subtraction and do P(Fit)*Jx*P. . IOp(7/76) Over-ride use of FoFCou in L703: 0 1 2 Normal processing. Copy fit density over real density and do P(Fit)*Jx*P(Fit). 10**(-N).-1 Zero (all printed). 10^-10 for molecules. 10^-12 for PBC. Clear real density and do -1/2 P(Fit)*Jx*P(Fit). IOp(7/77) Debuging options for DBFs: 0 1 2 3 4 Normal processing. The threshold is automatically lowered for each mode until 90% of the absolute displacements are included. Force FoFCou. IOp(7/87) Accuracy in FoFDir/FoFCou/CalDSu: 0 N Default. Prohibit FoFCou. Overlay 8 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 27 28 29 30 31 32 35 36 38 39 40 41 42 43 44 45 46 47 IOp(8/5) Whether to pseudocanonicalize ROHF orbitals. otherwise same as 4.-1 0 Yes. No. BUCKETS FOR STABILITY: (IA/JB). BUCKETS FOR CID OR MP3: (IJ/AB). IOp(8/7) SCF convergence test. QCISD. Full transformation if this is consistent with MaxDisk. CISD. THE COMPLETE SET OF TRANSFORMED INTEGRALS. otherwise same as 3.(IJ/KL). IOp(8/9) Debug control (L802): .(IA/JB). 0 1 2 Default (No). (IK/KL). BD: (IJ/AB). 0 1 Test that SCF has convergd. No.(IJ/AB). Full transformation if this is consistent with MaxDisk. 0 1 2 BUCKETS FOR MP2: (IA/JB). IOp(8/8) Whether to delete MO integrals in L811. 3 BUCKETS FOR SEMI-DIRECT MP4DQ. 4 5 6 7 CISD or MP4SDQ or MP4SDTQ. BUT INCLUDES (IA/BC).(IJ/KA). Yes.(IA/JB). IOp(8/6) Bucket selection. Do not test SCF convergence (mainly used for testing). 20000 Do symmetry compress transformed integrals (buckets) (This will cause windowed MOs. Force semi-direct transformation. 00000 Default (10000) 10000 Do not symmetry compress transformed integrals.. Generate and test RInt3 array (L804).. Force sorting for output bucks. Accumulate MP2 force constant terms in direct fashion Write the MO basis first derivative ERI's to disk Force fully in-Core algorithm (L804 only). Force semi-direct mode 3 if IOp(6)=3. Force semi-direct mode 4 if IOp(6)=3. virt-rep1. Eigenvalues and symmetry assignment vectors will be put in correspondence with vectors. Force output bucket in Core antisymmetrization.occ-rep2... reordered in the order of representations like occ-rep1.virtrep2.. L811): 0 1 2 3 10 20 30 100 200 1000 2000 3000 4000 Operate normally. (Upper triangle of symmetry compressed integrals for IOp(6)=5 or 4 only!) . Force transformed integrals in Core algorithm. Force N orbitals per pass. Direct Transformation Control (L804. Force semi-direct mode 2..) 30000 Symmetry compress transformed integrals only if RHF. Force semi-direct mode 1.0 -N Operate normally. except that the outer sp electrons of 3rd row and later alkalai and alkalai earth elements are retained. 3 The next to the largest noble gas core is frozen. Use orbital energies to choose core orbitals. a card is read in indicating the start and the end. 90 Use all MOs. 91 The window is specified by IOp(37-38). The window is recovered from file 569 on the checkpoint file. core virtuals are also frozen. 92 93 94 000 10x 20x 30x The window is recovered from rwf 569. 4 The largest noble gas core and main group d's are frozen. Default. G2 frozen-core: the largest noble gas core and main group d orbitals are frozen. Use overlap with atomic core orbitals from Core Ham to choose core orbitals. eigenvalues and symmetry assignment vectors according to ther representations IOp(8/10) Window is selected as follows: -N 0 N 1 2 Use the top N occupieds and lowest N virtuals. IOp(8/11) . same as 4. A negative value for the end deletes the top virtuals. 1 <= N <= 89 selects frozen-core type N: The largest noble gas core is frozen. Use overlap with atomic core orbitals from Harris to choose core orbitals. If IOp(37) is 0. Read a list of orbitals to freeze. Default (200). For basis sets with double-zeta cores.100000 Reorder MOs. 2 Kill the job if any mo coefficients are greater than 1000. orbital energy. kill a frozen-core job if there there is significant core-valence mixing in the canonical orbitals 00 10 Default.0 or the smallest difference between occupied and virtual orbital energies is less than 0. Output transformed integrals for DRT-CI calculation. Read 'old-style' drt input cards. Used to select output mode. e. IOp(8/14) Control of drt input. if there is only 1 electron or 1 virtual spin-orbital. 0 1 Default. and number of electrons test. Also. IOp(8/12) Calculation of frozen-Core contributions. Suppress such a test (CPHF may still be done for such a case). 0 1 Output to Gaussian system buckets.g. 0 1 Take necessary input from Gaussian data structures. IOp(8/13) Control of output. same as 10. IOp(8/15) .001. 20 Kill the job if there is no correlation energy. 1 Calculate 2 J ..K over deleted orbitals and add to Core-Hamiltonian. 0 No. This is done when IOpCl=0 or when IOpCl=1 and the calculation is rohf or gvb. same as 2 Just print a warning message.MO coefficient. 3 4 5 6 Non-canonical.x. 0 1 2 Write DRT output to RW-files.x = -1/2 Sij. (only if integral derivative file is being written) Excitation level for SDGUGA-CI. Uij. same as 3.x = 0. Non-canonical.x = -Sab.EjSij. Canonical occupieds.x) / (Ei-Ej) Note that this blows up for degenerate orbitals and is intended primarily for debugging. Uij. Write DRT output to Fortran unit 'drttap'.x = -Sij.x Uji.x = (Fij.x/2 Canonical virtuals. IOp(8/17) Specification of integral block size for GUGA CI programs. Uij. IOp(8/16) Maximum number of orbitals per pass in L811. Uab. Integral Block Size = N.x/2 .x = -Sij. 0 N Default let program decide. Uij. 2 Canonical. Non-canonical.Control of DRT output.x .x = -1/2 Sij. Do both. Excitation level = N. 0 N Default excitation level = 2. Uij. except canonical in frozen-active blocks.x. IOp(8/18) Which type of derivative transformation to do in L811: 0 1 Default. N Transform N orbitals (after frozen Core) as occupieds (i. N words. Nuclear Kth order.. 0 No.e. Electric field Jth order. IOp(8/22) These options control the in-Core post-SCF link. IOp(8/27) Maximum amount of disk to use in L804 and L811: 0 N Unlimited. L805. Look there for more information. Magnetic Field Ith order. set NOA=NOB=N for purposes of transformation).e. MO derivative times integral term. IOp(8/28) Hack number of occupieds for full ci using links 921 or 922: -1 Transform all orbitals (after freezing Core) as occupieds (i. IOp(8/20) Which terms to include in L811: 0 1 10 Default (same as 11). set NOA=NOB=NROrb in transformation)..IOp(8/19) The nature of the perturbation(s) in L811: 0 IJK Default (1st order nuclear and electric field). IOp(8/29) . MO times integral derivative term. -1 Set up /Orb/ for a full window but then blank the wavefunction coefficients in L804. -1 0 1 N Do one batch. but use multi-batch logic.. Default (same as 1). Do a single atom at a time (minimum disk usage). N triplets. 0 Default (-1) N N evaluations and hence N coarse tiled batches (1. This will determine the number of times AO integrals and derivatives are evaluated unless overridden by IOp(31). Test window.Maximum number of perturbations per batch in L811: (only applies if integral deriv file is written) -3 Do not use batching logic. subject to the limit imposed by MAXDISK (IOp(27)). IOp(8/31) . -1 Use as much as needed for maximum efficiency. -2 Use an amount which is similar to the maximum disk usage in other parts of the MP2 frequency code. Set up for full but zero Core MOs. Set up /Orb/ as indicated by IOp(10)..6 are the currently implemented options) IOp(8/30) Type of window. -2 Do as many in a batch as can be overlapped with sorting space for half transformed integrals. Requested disk usage. This only applies if the integral derivatives are not stored. independant of MAXDISK. 0 1 Default. -3 Use as much as desired. order g i. IOp(8/36) Whether to update force constants with the MP2 product of MP2 integral derivatives term (only applies if integral derivative file is not written). All. . Yes. 0 1 NO YES IOp(8/32) Whether to do CI in the interacting space only. IOp(8/35) Output format for closed-shell and debugging control: (only for when integral derivative file is written) 0 Default (consistent with integrals for open-shell. i<=jab alpha-beta only for closedshell). 0 N Let the program choose (make it as large as possible) maximum fine tile batch size.PERFORM PRIMITIVE POST-SCF OPERATIONS (NOT CURRENTLY FUNCTIONAL). 0 1 Default (Yes). Explicit control of the fine tile batch size (largely a debugging option. Interacting only. up to 9. 10 Do extra debugging computations. 1 Store only the unique AB integral derivatives (gO2V2/4. only for no-Ix routines).=j a<=b) for closed-shell 2 AA and AB consistent with integrals. 0 1 2 Default (all spin-eigenfunctions). IOp(8/39) Localized orbital method adopted in SAC/SAC-CI. IOp(8/40) Handling of ROHF window: 0 Default (2). Boys + population method. No. No. but save MOs. 1 Use ROMP2 approach. Yes. (Yes if Ix is not stored. forming pseudo-canonical alpha and beta orbitals and doing UHF transformation. . no otherwise).2 No. Take reference MOs from disk if available. IOp(8/38) Integers specifying the window to use. No. 00 Default on whether to make Poo and Pvv for MP2. 10 20 Yes. 0 1 2 Default. transforming only alpha orbitals. Boys method. 2 Treat as RHF. IOp(8/41) Transformation of spin-orbitals (alpha only) within occupied and unoccupied orbital subspaces by minimum orbital-deformation (MOD) method. No localization. 0 1 2 3 Default. Population method. K-point at which to apply correction.3 00 10 20 No for the 1st geometry of opt. Force not to rotate MOs. Rotate MOs to compute the displaced overlap. N -N Use N points for MP2. use unity as rotation. Use N points and set up for band gap correction. Yes. yes otherwise. 0 N Use the k-point for which the hoco is highest and luco is lowest. 0 1 2 00 10 000 100 No Yes. . Force to rotate MOs. (for SAC-CI single point calculation) Use orbital energies in ordering Don't use orbital energies in ordering Use second moments in ordering Don't use second moments in ordering 0000 Use dipole moments in ordering 1000 Don't use dipole moments in ordering IOp(8/43) Number of Laplace points to use. If C1. IOp(8/42) Whether to reorder MOs during potential surface exploration. IOp(8/44) K-point specification for MP2 band correction. IOp(8/45) Type of quasiparticle job: 0 -1 1 Band gap. Full list. Overlay 9 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 25 26 27 28 30 31 36 37 38 40 41 42 43 44 45 46 47 48 49 60 61 62 70 71 72 73 74 75 81 82 83 84 85 86 IOp(9/5) . Electron affinity. but blank contributions from inactive atoms. 3 Full list. Active atoms. No difference from 2 for overlay 8.MOs are canonical HF orbitals. IOp(8/47) Whether 804/811 should generate results compressed over active atoms: 0 1 2 Default (2). 1 Input orbitals are not canonical HF and pseudocanonical orbitals must be generated here for the post-SCF. 0 Default . -N N-th occupied band at the k-point for which the hoco is highest (by default) or at kpoint specified by IOp(44) N N-th virtual band at the k-point for which the luco is lowest (by default) or at kpoint specified by IOp(44) IOp(8/46) Indicates special case of non-HF calculation. Ionization potential. 13) Max N cycles. DOUBLE AND QUADRUPLE SUBSTITUTIONS. L914: MAXIMUM NUMBER OF EXPANSION VECTORS IN DAVIDSON SCHEME 0 N 200 VECTORS N VECTORS . COUPLED CLUSTER THEORY WITH SINGLE AND DOUBLE SUBSTITUTIONS. 1 2 CID. 3 MP4(DQ). FOURTH ORDER PERTURBATION THEORY IN THE SPACE SINGLE. CI WITH ALL DOUBLE SUBSTITUTIONS. DOUBLE. 4 5 MP4(SDQ). THIRD ORDER PERTURBATION THEORY. TRIPLE AND QUADRUPLE SUBSTITUTIONS.D18. FOURTH ORDER PERTURBATION THEORY IN THE SPACE DOUBLE AND QUADRUPLE SUBSTITUTIONS. 7 CCSD. CONFIGURATION INTERACTION WITH ALL SINGLE AND DOUBLE SUBSTITUTIONS. FULL FOURTH ORDER PERTURBATION THEORY IN THE SPACE OF SINGLE. MP4(SDTQ). COUPLED CLUSTER THEORY WITH DOUBLE SUBSTITUTIONS.METHOD 0 CIDS. 8 9 QCISD. IOp(9/6) L913: CRITERIA FOR TERMINATION OF THE ITERATION 0 -1 N DEFAULT CONVERGENCE CRITERION AND MAXCYCLE READ IN MAXCYCLES AND CONVERGENCE CRITERION (I2. MP3. 6 CCD. BD. IOp(9/7) UPDATE THE ENERGY IN COMMON/GEN/ 0 1 2 7 YES. UMP43. UMP43.4) for first iteration Use DD[1-3]R and UMP4xR (closed-shell) on 1st iteration Original code for 2nd and later iterations. UMP41. Use DD[1-3]R and UMP4xR (closed-shell). ECID IN CID. UMp41U. Original code (DD1.**** NOTE: WHEN EXPANSION VECTORS EXCEED THE MAXIMUM. YES. Use DD1. same as 1. DD4RQ (closed-shell). Original routines. .3. UMP42. Default. Term and method selection for debugging in 906.2. NO IOp(9/8) L902: Constraint on output wavefunction for stability calculations (see link 902). Slava routines. DD4UQ Use DD1R. WITH EUMP3.3. UMP42. default 1 in 907 and 10 in 919. UMP41R. fast and R where possible). L913: Whether to use fast routines: 000 1 2 10 20 30 40 000 100 200 Default (no Slava. DAVIDSON RESTARTS WITH CURRENT EIGENVECTORS AS INITIAL GUESSES. WITH EMP4(SDQ) OR EMP4(DQ) IF SINGLES ARE NOT AVAILABLE. WITH THE CORRELATION ENERGY.2. AND EUMP4 IN MP4 CALCULATIONS YES. Number of roots in 907 and 919. ECISD IN CISD EUMP3 IN MP3. L914: State of interest: 0 N WE ARE NOT DOING GRADIENTS.The defaults are 22 for RCI. and 31 for UQCI. L914 gradient: 10**-6 wfn.. Force direct computation of contributions.e. 10**-5 wfn. follow /Orb/. Only active in L914. Normal production of intermediates (in-core if possible). N IOp(9/10) Test flag in link 902 Whether to do "fake" frozen-core (i. note number of frozen core and virtual and reset /Orb/ for full. 11 for UCI. Yes. FP OR CIS-MP2 WE ARE INTERESTED IN THE NTH EXCITED STATE IOp(9/9) Convergence criterion (on energy for L913. 42 for RQCI. wavefunction for L914). and store full /Orb/ back on disk. 10**-N. L913 gradient: 10**-8 energy. Yes. Force direct computation of contributions. 0 Default: L913 single point: 10**-7 energy. with a full transformation). For AO usage (NYI here). . L914 single point:n: 10**-4 wfn. 10**-6 wfn. IOp(9/11) Flags for Green's function calculations: 0 1 2 00 Normal use of MO integrals. 0 1 2 3 No. IOp(9/13) Symmetry constraint of output wavefunction from stable=opt: 0/1 yes/no. but save the amplitudes. Spin projection control in L913: 0 1 2 Default (1) Do basic projection. BD(T). Link 909 only.10 Force use of sort for intermediates. Include triples? IOp(9/12) Test flag in l902. EMax. IOp(9/14) Non-iterative corrections: ICNonI 0 1 2 3 4 No.QCISD(T). but with two ranges on the same line for open-shell. 1000 10000 Force N**3 algorithm in GFSCMA. and pole strength warning level on one line. 100 Read window of MOs to refine in the same format as 801. IOp(9/15) . Fourth and fifth order singles and triples . Read EMin. Test flag in l902. but do E4T as well. Same as 2. Same as 2. Fourth-order triples (NYI). default: in-core if possible. included Z-amplitudes if necessary. suppress in-core storage. but also force use of the fully outof-core algorithm in Tran4D. using fully direct methods if possible. L914: Control of in-core integrals for W(Tilda): -6 -3 0 1 Force in-core storage. 0 M Default (same as -3) Use disk storage for partially transformed integrals handling M occupieds at once. L913. Do Lagrangian in L906. -2 Force a single integral evaluation (two for UMP2) using disk-based algorithm. . -6 Force the fully in-core algorithm. otherwise spilling to disk.Type of derivative information generated: 0 1 2 None. -1 Force in-memory algorithm (fully direct MP2. 2N**3 words for derivatives). Use AO integral algorithm (L914 only). requires 2OVN words of memory for E2. -5 Try to minimize integral evaluations as for -3. but also force use of the fully out-ofcore algorithm in Tran4D. IOp(9/16) L906: Control of (Semi-) Direct MP2: -N Do a maximum of (-N-6) occupieds per pass. -3 Try to minimize integral evaluations. -4 Force a single integral evaluation as for -2. using the fully out of core allgorithm. Do AO derivatives and Lagrangian in L906. Do gradient in L913. 0. ALWAYS. (IOp(5)). Use BFGS. NOTE THAT FOR PERTURBATION METHODS (METHOD=2. 1. Use old extrapolation for CI. ALWAYS.4.IOp(9/17) Auto-adjustment of tau in L918. Use RLE. Functional to use in L914. 1 2 W(0)/A0.DE = W(0)/A0.GT. FOR METHOD = 0.5) DE IS NOT REALLY NEEDED SINCE THE WAVE FUNCTION FORMED NEVER GETS USED. IOp(9/19) EXTRAPOLATION. Reset RLE for Z iterations. QCISD using RLE. 0 1 2 3 4 5 00 10 100 Default: CI using old extrapolation.E. 0 USE DE DEPENDING ON THE METHOD USED. Use DIIS.->DE = 0. FOR METHOD . Do not extrapolate. Use scaled A as guess for Z.1->. IOp(9/18) ITERATION SCHEME: DE= (IN A(S)=W(S)/(DE-DELTA(S)) I. IN THE FORMATION OF A NEW WAVE FUNCTION. Use A as guess for Z.3. IOp(9/20) . Localize both. 0 1 Yes. No (used in HF second derivative calculations). Read in configurations. IOp(9/23) Localization of orbitals in L919. -1 0 N Read in factor in format D20. IOp(9/22) Conversion factor in L919. 0 1 2 3 00 10 20 000 100 None. Rettrup-Davidson RPA. Localize occupieds. Localize virtuals. Default of 10**-8. . 1000 In-core method. IOp(9/21) Guess for eigenvector of y-matrix in link 902. Jorgensen-Linderberg Hermetian RPA.Whether to update the total energy with the MP2 energy in L901.10. 10**-N. Default (same as 10). Choose configurations by simple truncation. 0000 Out-of-core method. IOp(9/25) PRINT PAIR CONTRIBUTION AND WEIGHT TO CORRELATION ENERGY 0 1 2 3 4 NO YES. AT THE END OF CI YES. The functional is given by IOp(17). SUM(S) A(S)**2 = 1 (ALL S) NOTE: PERTURBATION THEORETICAL RESULTS ARE VALID WITH NORM=0 ONLY IOp(9/27) . AT EACH CYCLE YES. Correction to CIS in L914: 0 -2 -1 1 2 No CIS-DFT (in primitive energy code) CIS-MP2 (in primitive in-core program) CIS-MP2 (in MO Basis disk routine) CIS-DFT (in production code).00000 Singlet states. 10000 Triplet states. Maximum order of perturbation theory in L921 and L922. AT ONE CYCLE GIVEN BY INPUT (I3) YES. AT FIRST CYCLE AND AT END IOp(9/26) NORMALIZATION OF THE WAVEFUNCTION 0 1 NORMALIZED TO A(0)=1. Compute the CI one-particle density matrix. N words. IOp(9/28) PRINTING OF DOMINANT CONFIGURATIONS. 0 -3 -2 -1 Default (print coefficients 0. N Scan the 'A' vector and print all coefficients having coefficients greater than 0.J.Maximum amount of disk to use in L906: -1 0 N No disk. Lee . 22 otherwise). force fully direct method by default.1 and above). IOp(9/31) Print vectors and matrices in 902 and 918 0/1 no/yes. 10 Compute the density correct to second order (NOT the same as the density corresponding to the MP2 energy). Scan the 'A' vector and print all coefficients. Do not print coefficients. Use as much disk as needed for a single pass. IOp(9/30) Calculation of the one-particle density matrices: 00 1 2 Default (21 for CI. IOp(9/36) Compute the T1 Diagnostic of T.0001*N. Do not form the CI one-particle density matrix. 20 Do not compute the density correct to second order. Print all coefficients every iteration. Default: energy and gradient. the smallest dimension one and the default is three.IOp(9/37) Maximum dimension for the QCISD extrapolation. the only dimension is IOp(39) L913: Type of convergence test 0 . the previous job indicates wavefunction not just expansion vectors has converged). Same for DIIS extrapolation. IOp(9/38) Minimum dimension for the QCISD extrapolation. The maximum dimension is eight. For BFGS extrapolation. For BFGS extrapolation. The maximum dimension is 25.default size is ten. For DIIS extrapolation. L914: Pick out guesses from restart file or othogonalize guesses to the states already on restart file (IOp49 must be set to 1 or 2 for this option to be valid) 0 N -1 Just take guess from restart file Make N additional orthogonal guesses to those present Read which N states to use (free format integers) *** WARNING: The states on the restart file MUST be orthogonal to the convergence requested (ie.Converge on gradient only Convergence on gradient is for extrapolated CI and QCISD procedures.Converge on energy and gradient 3 . IOp(9/40) Reference wavefunction for MP2 in L906: .Converge on energy only 2 . 1 . AO basis. DEFAULT IS: 3 (RHF REFERENCE STATE) (UHF REFERENCE STATE) MATRIX AA. HF. THRESHOLD FOR PRINTING EIGENVECTOR COMPONENTS in L914: 0 N ITHR = 1 ITHR = N WHERE THRESHOLD = GFLOAT(10)**(-ITHR) IOp(9/41) L914: NUMBER OF STATES TO SEEK WHEN USING DAVIDSON.0 1 2 Default (HF). In-core. MO Mapped to 3. 333. FORMAT I5 ON LAST CARD BEFORE EOF IOp(9/42) METHOD AND MATRIX BLOCKS TO WORK ON in L914 (See below) -NNN 1 2 3 0 333 BITS 1 Mapped directly to NNN below. OR NUMBER OF STATES TO PRINT OUT INFORMATION FOR WHEN USING DODIAG: 0 N -N DEFAULT TO 2 LOWEST N STATES READ IN PRINCIPLE COMPONENT OF N GUESSES (DAVIDSON).BB METHOD -- . 222. CASSCF. or 20 as appropriate. or 30 as appropriate. Mapped to 2. IOp(9/44) Density matrix control for filling RWF 633 in L914: 0 1 2 Same as 2 Do densities of each excited state Do densities and transition densities from ground 3 Do densities. then half the number of states at the second iteration.BB AB BA AA. BASIS -- IOp(9/43) How to handle subsequent Davidson Iterations in L914: 0 If this is not a restart.O.BB AB BA |-> FORCE DAVIDSON IN A. 0 Use Phycon to convert to eV's . transition densities from ground. Force Davidson not to half the number of states at iteration 2. and transitions densities among all excited states IOp(9/45) Debug option for comparing previous results in L914. BASIS --|-> FORCE DODIAG TO FIND ALL ROOTS --|-> FORCE DAVIDSON IN M. If this is a restart.NYI NYI 2 20 200 3 30 300 10 AB 100 BA AA. 1 2 Force Davidson to half the number of states at iteration 2.O. then don't. IOp(41) Converge on Ci Amplitudes for N lowest states IOp(9/47) Control of Davidson Iterations in L914: 0 1 2 Usual Don't do any iterations (guess=print) Stop after first iteration IOp(9/48) RESTRICTION ON TYPES OF ROOTS (DAVIDSON RHF ONLY) 0 1 2 3 GUESS ONLY SINGLETS Same as 0 GUESS BOTH SINGLETS AND TRIPLETS GUESS ONLY TRIPLETS NOTE: A SINGLET GUESS MAY RESULT IN A TRIPLET ROOT IN EXTREME CASES (SMALL NUMBER OF ROOTS SOUGHT) IOp(9/49) INITIAL GUESS VECTORS 0 1 MAKE A GUESS BASED ON DIAGONAL ELEMENTS USE GUESS VECTORS ALREADY ON RWF .1 Use old conversion to eV's IOp(9/46) Control of Davidson Convergence in L914: <0 0 N Use Ortvec convergence only Converge on the number of roots . IOp(9/70) 1 to force TDHF in L914. 0 1 No. IOp(9/71) Whether to do an extra iteration after Davidson convergence. No. no for stability).2 3 4 USE GUESS VECTORS ALREADY ON CHK GENERATE GUESSES FROM CIS DENSITIES on CHK GENERATE GUESSES FROM CIS DENSITIES on RWF IOp(9/60-62) Over-ride standard values of IRadAn. Yes. Yes. IOp(9/73) Whether to do non-equilibrium solvation in L914: 0 1 2 Default (Yes. 0 1 2 Default (No). Yes. and IRanGd. use equilibrium. No. IOp(9/72) Whether to computed frequency-dependant polarizabilities. if doing excited states. IOp(9/74) Over-ride default choice of frequency dependence of the XC functional in L914: . IRanWt. Save amplitudes. Get the lowest energy value between CBS(N) and CBS(NVirt). 0 -N N Default . IOp(9/81) Minimum number of Pair Natural Orbitals (PNO) to start the extrapolations from. IOp(9/75) Whether to save amplitudes and integrals in L906: 0 1 2 Save only if doing second derivatives (SqS12 set). NStart. Population Boys+Population.0 N Use default value.p')).e. Boys. Save amplitudes and integrals. Use 10**(-N) IOp(9/83) Localization Method. 0 N Use the default. -1 0 1 2 3 No localization. Calculate the extrapolated value at N only. Default (4). IOp(9/82) Convergence tolerance for CBS localization. i. Use form N (see IOp(88) in overlay 5). 6-31+G(d'.5 (assuming CBS-4 calculations. . Yes. Yes. Do 2nd order. No. save them. 0 1 No. 0 NO . intended for visualization). IOp(9/85) Flags for SAC-CI IOp(9/86) Whether L906 should generate data compressed to active atoms during mp2 frequencies with ONIOM: 0 1 2 Default (2). Localize core even if not needed. Overlay 10 5 6 7 8 9 10 11 13 14 15 16 17 18 19 20 21 22 28 29 30 31 32 45 46 47 48 60 61 62 63 72 72 74 75 76 77 78 79 IOp(10/5) CALCULATION OF FIRST DERIVATIVES OF POST-SCF ENERGIES. Only implemented for closed-shell and UHF. No localization. don't save (default).4 5 10 100 Minimal population. IOp(9/84) Save CBS localized orbitals to RWF (this will overwrite the SCF orbitals. No contributions to the force constants are done here). Setup for external processing of W and Z. 00 Default: use new Px/Wx digestion code if possible. do D2 E(SCF) / D R(I) D R(J) 2 Setup For MP2 2nd Derivatives (i.1 2 3 4 5 6 7 8 9 00 10 20 30 CALC.e. D E(CISD) / D R Calc. 100 200 Compute F1 and S1 derivative terms here. Available for RHF and UHF. . IOp(10/6) Calculation of the second derivatives of the SCF energy. Z-Vector method. D E(CIS) / D R Calc. D E(CCD) / D R Calc. Yes. Test Z-Vector using full CPHF. 0 1 No. D E(CCSD/QCISD) / D R Calc. save as little data as possible.just set up here unless doing HF 2nd derivatives simultaneously. D E(BD) / D R Calc. D E(MP3) / D R Calc. D E(MP4) /D R Default CPHF usage (Z-vector unless HF D2E) Full 3*NAtoms CPHF. D E(MP2) / D R CALC. Partially coded but NYI for high-spin ROHF. Don't process any derivative terms here. D E(CID) / D R CALC. 000 Default derivative processing . 1. 0 N Default: 1. 3 Invert the A matrix directly. Use new Px/Wx code but save both S1 and F1 over MOs.D-9. 0 1 No. possibly reverting to the old (one variable at a time) method in the secondary solution. Compute dipole derivatives using only electric field CPHF and F(x) matrices. 1000 Set up for GIAO MP2 calculation.A) contributions. 2 Solve all equations together. IOp(10/10) . Use new Px/Wx code and don't save S1 but do save F1. except 1. The max element is tested against 10* this value. IOp(10/7) RMS CONVERGENCE ON C1(I. IOp(10/8) Selection of linear equation solution method.D-11 for Z-Vector CPHF.10 20 30 100 Use old Px/Wx digestion code.D-N. 0 1 Default (same as 2). Yes. Expand each variable in a separate expansion space. IOp(10/9) Whether to compute Born-Oppenheimer corrections. 20000 Do hyperpolarizabilities for second-harmonic generation. 10000 Do DFT 3rd derivatives. Always use DIIS. 1 Save vectors at end. IOp(10/11) Largest matrix for direct inversion in LinEq2. 0 Default (1st order nuclear and electric field). IOp(10/15) What to do with expansion vectors from the linear equations. 0 -1 N Default invert directly if there is enough memory. =2 otherwise). Magnetic Field Jth order. never invert directly. N. . Don't update polarizability. IOp(10/13) The nature of the perturbation(s). IJKL Nuclear Lth order.0).eq. IOp(10/14) Whether to update dipole and polarizability derivatives. Force 2nd order cphf for polarizability derivatives. 0 1 2 10 20 100 Default (yes if IOp(5). Update dipole. Nuclear magnetic moment Ith order.Control of CPMCSCF during avoided crossing/conical intersection searches. 0 Default (=1 if IOp(8)=1 and electric field only and no derivatives are being computed. Don't update dipole Update polarizability. Electric field Kth order. for debugging frozen-core with integrals over the full window. and S2 off the AO 2PDM.2 3 00 10 20 Delete vectors at end of each CPHF. 3 Save as 2. which is used only for one term in polarizability derivatives and for which the accuracy requrirements are less stringent. Convergence is 10**(-N) for max and rms. IOp(10/17) Frozen-core: 0 1 Default (use AO 2PDM for Lagrangian only if orbitals are frozen in /Orb/). Default (Use old vectors if available). 2 Convert /Orb/ to full. but leave the full version of /Orb/ on the disk. This option is normally used to pass 1st order electric field results to the second invokation of 1002 during frequency calculations. IOp(10/18) Whether to do correct or approximate CPHF. . S1. Pass vectors from 1st to 2nd order CPHF. Use old vectors if available. This may be acceptable in the electric field second order CPHF. 0 CPHF is done correctly. but use of electric field expansion vectors for nuclear coordinate CPHF can cause errors of up to 1 cm**-1 with current tolerances. Note that because of numerical instabilities in the simultaneous solution method. IOp(10/16) Convergence in secondary linear equations (only for simultaneous solution). reusing old expansion vectors for new B vectors can reduce accuracy. but delete at end of link. Ignore old vectors. C2. Do C1. 0 N Use standard machine tolerance (MDCutO) on maximum and rms. Disables use of symmetry to reduce the size of the CPHF problem here. even if present. 2 The U-matrices are set to zero.. Compute the 2e integrals when needed. Read the 2e integral files. Yes. independent of convergence criteria. Yes. Note that the appropriate rwf (588) must be present in any case. i.e. Use option MN in control of 2e integral calculation. uncoupled Hartree-Fock is used). .e. 4 MNx Default (decide on the fly). IOp(10/19) Whether overlap (S1) terms must be included. IOp(10/21) Whether to store Uai. IOp(10/20) How to handle 2e integral contributions: 0 1 2 3 direct. MO if possible. 3 Only a single set of products AX are computed. Spq. No. even if MO ones are available. 0 1 Default (No). Simultaneous solution is implied.1 The A-matrix is neglected. Force use of AO integrals. 0 1 2 Default (yes). and hence the U-matrices are set equal to the Bmatrices (i. force AO or Don't use integrals. and full MO Fock matrix derivatives in permanent rwfs. IOp(10/22) Which multipole (electric field) perturbations to include? Only used if J part of IOp(13) is non-zero. Octopole. IOp(10/30) In-core storage of 2e integrals: 0 Default . IOp(10/28) State for CPMCSCF: 0 N Default (ground state).2 10 No. IOp(10/29) Use of rafinetti integrals during direct SCF. Quadrupole (electric field gradient. -N All integrals done as Raffenetti if there are N or more matrices. Save magnetic MO deriviatives. N Integrals with degree of contraction greater than or equal to N are done are regular integrals. All integrals are done as regular integrals. . Dipole (uniform electric field). all 6 cartesian components. Hexadecapole. all as regular if there are less than N. 0 1 2 3 4 Default.do if possible in direct calculation. 0 1 Default: let FoFDir decide. Nth excited state. Uniform electric field (dipole) only. Yes. IOp(10/31) Whether to use symmetry to reduce the number of CPHF equations: 0 1 2 Default (yes). No.gauge origin coincident with the nucleus of the integrated atomic regions. IOp(10/45) Type of Gauge Transformations to perform to calculate the current distribution within the molecule. . No. Yes. Default (16 if doing magnetic CPHF). 2 Use IGAIM method . Whether to read D2E file in link 1003: 0 1 2 Default (No). and hence the molecule's other magnetic properties. Yes. No. IOp(10/32) Whether to apply interchange in link 1004: 0 1 2 Default (No). Force recomputation. recover ints if available on rwf 610. -1 0 None.1 2 Force in-core storage. 1 Use single gauge origin .the gauge used to calculate the angular momentum perturbed wavefunctions. IOp(10/60-62) Over-ride standard values of IRadAn. IOp(10/48) Whether to operate only over perturbations involving active atoms. IOp(10/47) Whether to do spin-spin coupling constants. IOp(10/46) Whether to calculate dipole and rotational strengths (VCD). 0 1 2 3 Default (For nuclear.the coordinates of which are read in (in Angstroms). Use single gauge origin . Don't compress. 0 1 2 Default (No) Yes. compress if overlay 11 did). but blank contributions for inactive atoms. Compress. Use GIAOs. 0 1 2 3 No (Default) Yes No Do only optical rotational strength.4 8 16 Use CSGT method. No. and IRanGd. Don't compress. IOp(10/63) . IRanWt. Do not update. use more aggressive cutoffs in Xc integration. 0 N Default (1000). unless both electric and magnetic properties are requested). IOp(10/73) Maximum number of CPHF cycles.Changing defaults. Use global cutoffs. read in frequencies. . lower cutoffs suitable only for CPHF/CPKS. 1 2 10 20 Turn off FMM here regardless. calculation is performed. 0 Default: Use FMM if turned on globally. Use FMM if turned on globally. Use local. Yes. Yes. 00 Update frequency-dependent property file if frequency-dep. IOp(10/72) Whether to do frequency-dependent properties: 0 1 2 3 4 Default (No. IOp(10/74) Whether to do non-equilibrium solvation. N. Yes. with formalism for frequency-dependent XC response. 10 20 Update regardless. use more aggressive cutoffs in integrals and FMM unless doing NFx. No. 0 1 2 Default: Only if frequency-dependent. 0 1 2 00 Default (1). Canonical MO derivatives. 0 1 2 Default (No). MOD orbital derivatives. IOp(10/78) Whether to solve CPHF equations for MOD method. IOp(10/77) Test CPHF results by checking the CPHF equations using the complete MO Fock and density derivatives. Print eigenvectors as well. 0 N Use global value for this job step. No. Yes. 0 1 2 Default (1). Default (20). IOp(10/76) Over-ride general choice of exchange-correlation frequency dependence. . Use type N (see IOp(88) in overlay 5). No. Yes. Print tensors and eigenvalues. IOp(10/75) Print during NMR. Solve with SimEqn. IOp(10/79) Stop the link at selected points. YES.10 20 Solve using DiagD. 0 1 NO. restart point NN. 2 Yes. IOp(11/6) IFHFFX: WHETHER OR NOT TO CONTRACT INTEGRAL DERIVATIVES WITH HARTREE-FOCK DENSITY MATRIX TERMS TO PRODUCE HARTREE-FOCK TWO-ELECTRON CONTRIBUTION TO THE FORCES. PRODUCE A D2E FILE. for testing restarts. also contracted electric field density matrix derivatives to form the twoelectron integral derivative contribution to the polarizability derivatives. 0 1 NO. YES. IOp(11/7) IFTPDM: WHETHER OR NOT TO CONTRACT INTEGRAL DERIVATIVES WITH A 'READ-IN' TWO-PARTICAL DENSITY-MATRIX. 0 1 DO NOT PRODUCE A D2E FILE. . Overlay 11 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 26 27 28 29 30 31 32 33 39 40 41 42 43 45 46 53 60 61 62 63 70 71 75 IOp(11/5) IFWRT: DERIVATIVE INTEGRAL WRITE OPTION. MNN Stop at pass M (default 1). Yes. and SCF energy in L1110: 0 1 No. ADD IN CONTENTS OF FX2. but generate and write out the HF 2PDM here for debugging purposes. NO. 0 NONE. MERELY SET THE ARRAY TO ZEROES. CONTAINS I2*100+I1*10+I0. NONE. I2 0 1 PROCESSING OF TPDM CONTRIBUTIONS.2 Yes. IOp(11/11) Control of integral derivative algorithm: 0 2 Default use IsAlg to decide. Lagrangian. I1 PROCESSING OF TWO-ELECTRON HARTREE-FOCK CONTRIBUTIONS. IF NOT THERE. IOp(11/9) IDOUT: FIRST-DERIVATIVE OUTPUT OPTION. 3 Form 1/2(F+H) term in link 1110. YES. Yes. IOp(11/10) Whether to compute Fock matrices. 2 TAKE HF CONTRIBUTIONS FROM F1 (A LA IFF1). 0 1 No. 1 TAKE HF CONTRIBUTIONS FROM FX1 (A LA IFHFFX). I0 0 1 WHETHER OR NOT TO USE THE CONTENTS OF IRWFX. . (forms the 1/2(F-H) term in link 1110). IOp(11/8) IFF1: WHETHER OR NOT TO COMPUTE F1 OVER AO'S. Scalar Rys SPDF. FoFDir: Prism spdf. . Illegal here. Illegal here. Old gOV3 I/O algorithm. Use L(x) and Ux*I.3 4 5 6 7 8 9 10 11 12 13 Illegal here. Illegal here. Use Ix. Illegal here. IOp(11/13) Flags for L1112: 0 1 2 00 10 20 Default for Ix==>Sx (same as 1). same as 1. Formation of Ux*I*T terms. Illegal here. Illegal here. N Use generalized density number N for both the one-electron integral derivatives and the corresponding 2PDM terms. N**4 I/O algorithm. default. Illegal here. IOp(11/12) Selection of 1PDM in L1102 and L1110: 0 Usual SCF density. Illegal here. Illegal here. 0 controls output of 'old' format.0 load fxyz from rw-files if it exists. . i4*10000+i3*1000+i2*100+1i*10+i0 i0 .ne. i3 . Force O2V2 method. Also compute HF contribution to the dipole moment.000 100 200 300 Formation of Fx*T*T terms: default is to choose based on available memory. i2 . IOp(11/15) Controls output of derivatives to rw-files.ne.0 forces out-of-core algorithm IOp(11/16) Mode of operation of L1102. 0000 Default Ix*T algorithm (1) 1000 Force new algorithm.ne. Force old N**5 I/O algorithm. IO(11/14) The nature of the perturbation(s). Magnetic Field Ith order.0 calculate one-electron contribution. Nuclear Kth order. 2000 Force old algorithm. 0 1 10 Default: compute dipole derivative matrices only. 0 IJK Default (1st order nuclear and electric field). Electric field Jth order. Also compute dipole derivative integral contribution to the HF dipole derivatives. i1 . Use (2g+O)V2 memory algorithm even if O2V2 memory is available.1 calculate nuclear contribution. i4 .ne.eq. and S2 off the AO 2PDM.IOpCl+1) in rwf 1001.NAt3. 2 Compute the 2e integrals when needed. 0 N No. No. IOp(11/19) Whether to delete MO integrals after 1112: 0 1 2 Default (Yes). IOp(11/18) Save AO 2PDM from L1111. C2. MO if possible.NTT) and includes factors (2-Delta(ij))(2-Delta(kl)). Do C1.NBasis. Form the derivative integral contribution to the Lagrangian as well. IOp(11/20) How to handle 2e integral contributions in L1112: 0 1 Default (same as 1). Save the AO 2PDM on rwf N. It doesn't include any normalization factor. 3 10 Save as 2. but leave the full version of /Orb/ on the disk. It is (NTT. for debugging frozen-core with integrals over the full window. S1.IOp(11/17) Frozen-core in L1111: 0 1 Default (use AO 2PDM for Lagrangian only if orbitals are frozen in /Orb/). Read the 2e integral files. This is stored on disk as RL(NBasis. . This link must have been built with the non-dummied version of FoFDir and associated integral routines. 2 Convert /Orb/ to full. Yes. even if MO ones are available. N Integrals with degree of contraction greater than or equal to N are done are regular integrals. IOp(11/21) Size of buffers for integral derivative file. IOp(11/26) PROGRAM ACCURACY OPTION. 0 1 Default: let FoFDir decide. Contract with density matrix to form dipole derivative contributions. IOp(11/22) In-core option in 1112. . 0 N Default (Machine dependent.3 Force use of AO integrals. IOp(11/23) Use of rafinetti integrals during direct term in L1112: -N All integrals done as Raffenetti if there are N or more matrices. 0 DO INTEGRALS ECOMOMICALLY TO 10**(-10) ACCURACY. see DSet2E). IOp(11/24) Output of 1102: 00 1 10 Default (01). All integrals are done as regular integrals. Store dipole derivative matrices on disk. MNx Use option MN in control of 2e integral calculation. N integer words. all as regular if there are less than N. 627. IOp(11/29) What to do: 1 10 100 Transform 1PDM and Lagrangian from MO to AO. Compute HF DMs. Transform 2PDM from MO to AO. and 628). 0 N RETAIN INTEGRALS GE 10**(-10) IN THE D2E FILE (IF SELECTED) AND/OR 10**(-10) IN THE INTEGRAL HEAP IF IFF1=1 AND MODE=2. . The sorted 2PDM is left in file 602.1 'TEST' OPTION BYPASS CUTOFFS. Default (RWFs 626. Sort AO 2PDM into shell order. and N+2 (2PDM). RWFS N (1PDM). Compute CID 2PDM. IOp(11/28) Location or generation of MO 1 and 2 PDMs for L1111: -7 -6 -5 -4 -3 -2 -1 0 N Compute QCISD 2PDM Compute CCD 2PDM Compute CIS 2PDM Compute CISD 2PDM. IOp(11/27) INTEGRAL RETENTION PARAMETER. RETAIN INTEGRALS GE 10**(-N). the double-length AO 2PDM is expected in file 1001. If back transformation has not been requested. Compute MP2 2PDM. N+1 (W). 1000 10000 20000 Suppress writing alpha. beta. Note that if the is also to be left behind. . Compute 2PDM only. Compute Density Only. NO PRINTING. Generate replicated 2PDM copies for testing. it will be over 6d/10f and have the HGP d and f scale factors in it. No. and spin density rwfs. no density or W. Compute Density and W Only. Form and sort the 2PDM derivatives rather than the 2PDM. IOp(11/33) IPRINT IOp(33) 0 PRINT OPTION. Gradient. IOp(11/32) Whether to do 2PDM or just Lagrangian in L1111: 0 1 2 3 4 Compute Full Gradient Compute Full Gradient (Same as Default). 1 10 Energy.200 2PDM Form the contribution of the 2PDM to the forces right here. IOp(11/31) Whether to use symmetry in Rys integral derivatives in L1110: 0 1 Yes. IOp(11/30) What to compute using integrals or D2E file. Neglect four center integrals. IOp(11/40) Neglect of integrals (only option 1 works in Overlay 10): 0 1 2 3 10 20 30 Keep all integrals. Do only overlap and not other 1e integrals. but blank contributions for inactive atoms. IOp(11/41) NDDO flag. Neglect three center two-electron integrals as well. Compute over active atoms only. Neglect 1e integrals with diatomic differential overlap.1 2 PRINT COMPUTED FIRST-DERIVATIVES. 0 1 Evaluate usual integrals. Evaluate matrices in the NDDO approximation. Compute over the full list of atoms. Neglect three center one-electron integrals. Compute over the full list of atoms. Neglect 2e integrals with diatomic differential overlap. IOp(11/42) Compressed file formats. IOp(11/39) Compression of derivative matrices: 0 1 2 3 Default (2). . PRINT F1 MATRICES. Compressed Sx but separate H1 and F1.0 1 2 3 Default: compressed. Save gradients to disk. but using the out of core algorithms. Force compressed form. Yes. Yes. . smallest possible number of passes. 0 1 N Default. needed for non-canonical methods. IOp(11/45) Force NAt3 instead of NAt3+3 storage of matrices (for debugging): 0 1 No. IOp(11/46) Whether to include orbital rotation gradient terms for SAC-CI. IOp(11/43) Batching in overlay 11. N is 0/1/2 for default/in-core/out-of-core. 0 1 2 No. Do at least one pass. Force expanded form. For Rys in L1110. IOp(11/53) Convert forces over shells to field-dependent dipole and forces over atoms (for debugging): 0 1 No. Do at least N passes. Convert 1PDM to canonical representation. Yes. 0 1 2 Default (No). IOp(11/63) Whether to do FMM. No. and IRanGd. Only works using 1C shell pairs for the density basis and only with cartesian functions. Print tensors and eigenvalues. Print eigenvectors as well. . IOp(11/71) Debugging option for DBF derivatives: 0 1 2 Normal processing. Ignore fitting density and just process real density in L1110. Turn off FMM here regardless. IOp(11/70) Whether to allow cavity to move in PCM derivatives. IOp(11/75) Print during NMR. 0 1 2 Default (1).IOp(11/60-62) IOp(11/60-62) Over-ride standard values of IRadAn. 0 1 Use global default. IRanWt. Copy fitting density over real density. .Overlay 9999 5 6 7 8 9 10 11 12 13 14 15 16 17 18 33 IOp(9999/5) CONTROLS HANDLING OF THE CHECKPOINT FILE: 0 THE RUN IS AN OPTIMIZATION OR FREQUENCY RUN. PolyAtom output in working precision to Fortran unit 8. No GVB2P5 trans file. DELETE THE RESTART INFORMATION IF THE RUN IS FINISHING NORMALLY (I. SAVE THE PERMANENT INFORMATION (MOS. GVB2P5 trans file to unit 14. IF THE ERROR TERMINATION ILSW BIT IS NOT SET).. if a new version was not generated in this step). WFN file output WFN file output with magnetic orbital derivatives. Archive data from the checkpoint file. 1 FILE.) ON THE CHECKPOINT Do not write anything to the checkpoint file. but don't remove extra data (i.e. 2 3 4 Restart a multi-step job. THE RUN IS NOT AN OPTIMIZATION. BASIS SET INFO ETC. SO BOTH THE PERMANENT AND RESTART FILES ARE IN THE CHECKPOINT FILE. 5 Save data on the chk file.E. IOp(9999/6) Controls output of Fortran unformatted files for other programs: 0 1 00 10 100 200 No PolyAtom output. recovering data from the checkpoint file and figuring out which job step to run next and whether it needs restart if an optimization or numerical frequency. 3D20. Yes. IOp(9999/9) Controls archiving of dipole moment and other electic field derivatives. Atomic numbers and coordinates in format (I3.e. 0 1 2 Archive all as is. 0 1 2 Nothing. No. except for archiving from the chk file. and isotopes during multi-step energy calculations: 0 1 2 Default (same as 1).300 1000 WFN file output with GIAO magnetic orbital derivatives. Yes. pressure.12). . 4 Derivatives (forces and force constants) in format (2X. formatted output to unit 7). Title. use defaults. IOp(9999/8) Reading temperature. IOp(9999/10) Controls punching of assorted information (i.3D20. but rotates to z-matrix orientation first. Archive all. Use natural orbitals in WFN file IOp(9999/7) Controls whether MOs are written to the polyatom integral tape in LANL style. Don't archive. These are in the Z-matrix orientation. 0 1 No..12). An input deck for HONDO. Read a list of atoms to use in the Pickett input. 32 The molecular orbitals. in format suitable for guess=cards. 64 128 256 512 1024 2048 A GAMESS input deck. . IOp(9999/11) Which type of database to update: 0 1 2 3 Default (3). New format. Output hyperfine tensors as input to Pickett's program (sent to the output file). Both. in the standard orientation. Use natural orbitals in WFN file. Yes. Old format. Yes. IOp(9999/13) Whether this is the end of the job step: 0 1 Default (Yes). This is independant of normal archiving to the main file. A WFN file for PROAIMS. IOp(9999/12) Flag for coordinate optimization: 0 1 No.8 16 The archive entry. The natural orbitals generated by link 601. remove /ZMat/ and /ZSubst/ from the rwf and chk files. IOp(9999/15) Act as though in multi-step job type IOp(15). . and update rwfs. Yes. DEBUG PRINT. Yes. IOp(9999/33) CONTROLS DEBUG PRINT: 0 1 NO DEBUG PRINT.2 3 No. Include N virtual orbitals. IOp(9999/18) How many virtual orbitals to include in the WFN file. Go back to Link 1. Include all virtual orbitals. IOp(9999/16) Treat the job as type (Info(7)) given by IOp(16). IOp(9999/17) Treat as MSJDon=IOp(17) step in a multi-step job. IOp(9999/14) Whether to attempt to express the final optimized structure in terms of the input z-matrix: 0 1 2 3 Yes if there are 20 or fewer atoms. 0 -1 N Default (None). No. the coordinate is a bond stretch between the two atoms. You must provide these to the calculation in some way. By default. Note that one of RCFC. and the path can be followed in one or both directions from that point. designated by up to four atom numbers. CalcAll and FCCards must be specified.1 amu1/2 bohr along the path. and then specify IRC=RCFC in the route section. The usual method is to save the checkpoint file from the preceding frequency calculation (used to verify that the optimized geometry to be used in the IRC calculation is in fact a transition state). and four atoms define a dihedral angle. The initial geometry (given in the molecule specification section) is that of the transition state.152]. three atom numbers specify an angle bend. . The other possibilities are providing the force constants in the input stream (IRC=FCCards) and computing them at the beginning of the IRC calculation (IRC=CalcFC). IRC calculations require initial force constants to proceed. IRC studies are not currently archived. PATH SELECTION OPTIONS Phase=(N1 N2 [N3 [N4]]) Defines the phase for the transition vector such that "forward" motion along the transition vector corresponds to an increase in the specified internal coordinate. The geometry is optimized at each point along the reaction path such that the segment of the reaction path between any two adjacent points is described by an arc of a circle.IRC This method keyword requests that a reaction path be followed [151. Cartesians or internals coordinates. it can be defined explicitly using the Phase option. CalcFC. the forward direction is defined as the direction the transition vector is pointing when the largest component of the phase is positive. an IRC calculation steps 6 points in mass-weighted internals in the forward direction and 6 points in the reverse direction. and so that the gradients at the end points of the arc are tangent to the path. You should specify alternative isotopes for IRC jobs using the standard method. IRC calculations accept Z-matrices or Cartesian coordinates as molecule specifications and uses these coordinates in following the reaction path. The path can be computed in mass-weighted internals. If two atom numbers are given. By default. in steps of 0. ReadCartesianFC is a synonym for RCFC.Forward Follow the path only in the forward direction. StepSize=N Step size along the reaction path. The format is Z-matrix (FFF(I).6). RCFC Specifies that the computed force constants in Cartesian coordinates from a frequency calculation are to be read from the checkpoint file. read as (8F10. CalcAll Specifies that the force constants be computed at every point. MW is a synonym for MassWeighted. ReadVector Read in the vector to follow. I=1. MaxPoints=N Number of points along the reaction path to examine (in each direction if both are being considered). The default is 20.01 amu1/2-Bohr. . The default is 6. MaxCyc=N Sets the maximum number of steps in each geometry optimization. COORDINATE SYSTEM SELECTION OPTIONS MassWeighted Follow the path in mass-weighted internal (Z-matrix) coordinates (which is equivalent to following the path in mass-weighted Cartesian coordinates). in units of 0. Internal Follow the path in internal (Z-matrix) coordinates without mass weighting Cartesian Follow the path in Cartesian coordinates without mass weighting. CalcFC Specifies that the force constants be computed at the first point. Reverse Follow the path only in the reverse direction. The default is 10.NVAR). This is the default. CASSCF. CID. CCD. This option can be used to read force constants recovered from the Quantum Chemistry Archive using its internal FCList command. where NAt3 is the number of Cartesian coordinates. CIS. all DFT methods. RESTART OPTION Restart Restarts an IRC calculation which did not complete.16) Cartesian forces (lines of format 6F12. MP2. This option is necessary if a very small step size along the path is requested. OPTIMIZATION ALGORITHM-RELATED OPTION VeryTight Tightens the convergence criteria used in the optimization at each point along the path. Each step is introduced by this banner line (where "IRC" has replaced "Grad"): IRC-IRC-IRC-IRC-IRC-IRC-IRC-IRC-IRC-IRC-IRC-IRC-IRC-IRC-IRC-IRC-IRC . but for which additional points along the path are desired.I=1. and all semi-empirical methods. If both FCCards and ReadIsotopes are specified.I). or restarts an IRC calculation which did complete.FCCards Reads the Cartesian forces and force constants from the input stream after the molecule specifications. the masses of the atoms are input before the energy. CCSD.8) Force constants (lines of format 6F12. Opt. MP4(SDQ).J=1. The format for this input is: Energy (format D24. CISD. Cartesian gradients and the Cartesian force constants.NAt3).I). MP3. Scan. HF.8) The force constants are in lower triangular form: ((F(J. IRCMax The output for each step of an IRC calculation is very similar to that from a geometry optimization. QCISD. IRCMax . Coord: Angstroms.08274 6 -40.09946 Once the entire IRC has completed.39205 NET REACTION COORDINATE UP TO THIS POINT = 0.15640 -0.08193 4 -40.16235 -0.49831 1. and Degrees) -------------------------------------------------------------------ENERGY RX.Optimized point # 1 Found.0047 ! ! CH2 1.08353 10 -40.0082 ! -------------------------------------------------------------------RADIUS OF CURVATURE = 0.15914 -0.08145 2 -40.00000.15552 0.15649 0.08328 11 -40.16418 1.16837 -0.08300 7 -40.16486 0.00000 1. the program prints a table summarizing the results: -------------------------------------------------------------------SUMMARY OF REACTION PATH FOLLOWING: (Int.86318 1.As the optimization at each point completes.49759 1.39938 1.29820 1.10833 1.0827 -DE/DX = 0.80000 The initial geometry appears in the middle of the table (in this case.03788 1.25158 1. -.08276 -------------------------------------------------------------------TOTAL NUMBER OF GRADIENT CALCULATIONS: 28 TOTAL NUMBER OF POINTS: 10 AVERAGE NUMBER OF GRADIENT CALCULATIONS: 2.26036 0.08232 5 -40.18068 1. as point 6).19985 1.30200 0.0143 ! ! HH 0.17324 0.19914 1.49968 0.91500 1.12245 1.09260 1.39764 1.54387 0.45133 0.0008 ! ! HCH 106. It can be identified quickly by looking for a reaction coordinate value of 0. ---------------------------! Optimized Parameters ! ! (Angstroms and Degrees) ! --------------------------------------! Name Value Derivative information (Atomic Units) ! -------------------------------------------------------------------! CH1 1.16957 0.76567 1.34481 0. the optimized structure is displayed: Optimization completed.80711 1.8632 -DE/DX = -0.09946 1.39854 0.09990 1.29975 1.21116 1.08349 9 -40.74371 1.3448 -DE/DX = 0.COORD CH1 HH CH2 1 -40.08330 8 -40.73360 1.16542 -0.15999 0.08164 3 -40.96924 1.207 -DE/DX = -0. this calculation type finds the maximum energy along a specified reaction path.1 amu1/2 bohr until the maximum of the MP2/6-31G(d) energy (including the HF/3-21G* ZPE scaled by 0. The Zero option will produce the data required for zero curvature variational transition state theory (ZC-VTST) [169. and the total MP2/631G(d) + ZPE energy of the TS. PATH SELECTION OPTIONS Forward Follow the path only in the forward direction. You should specify alternative isotopes for IRCMax jobs using the standard method. the imaginary frequency for tunneling (fit to MP2/6-31G(d) + ZPE).170. all real vibrational frequencies (HF/3-21G*).175.Performs an IRCMax calculation using the methods of Petersson and coworkers [168. The position along the HF/3-21G* IRC for this MP2/6-31G(d) TS will then be optimized.169.173. ZC-VTST OPTIONS Zero Include the zero-point energy in the IRCMax computation.6).Zero.p) reaction path where the B3LYP/6-31G(d.p):HF/6-31G(d.173.171.NVAR).p) energy is at its maximum.176]. The format is Z-matrix (FFF(I).175. . The output includes all quantities required for the calculation of reaction rates using the ZC-VTST version of absolute rate theory: TS moments of inertia.Stepsize=10) This job will start from the HF/3-21G* TS and search along the HF/3-21G* IRC with a stepsize of 0. separated by a colon: IRCMax(model2:model1).170. I=1. ReadVector Read in the vector to follow.172. Taking a transition structure as its input. REQUIRED INPUT IRCMax requires two model chemistries as its options.176]. read as (8F10.p)) This calculation will find the point on the HF/6-31G(d.174.91671) is bracketed.174. Reverse Follow the path only in the reverse direction. Consider the following route: # IRCMax(MP2/6-31G(d):HF/3-21G*. Here is an example route section: # IRCMax(B3LYP/6-31G(d. The default is 10.01 amu1/2-Bohr. COORDINATE SYSTEM SELECTION OPTIONS MassWeighted Follow the path in mass-weighted internal (Z-matrix) coordinates (which is equivalent to following the path in mass-weighted Cartesian coordinates). This option is necessary if a very small step size along the path is requested. in units of 0. MW is a synonym for MassWeighted. This option can be used to read force constants recovered from the . CONVERGENCE-RELATED OPTION VeryTight Tightens the convergence criteria used in the optimization at each point along the path. for each optimized point on the path [496].MaxPoints=N Number of points along the reaction path to examine (in each direction if both are being considered). StepSize=N Step size along the reaction path. This option is valid only for reaction paths in mass-weighted internal coordinates. Cartesian Follow the path in Cartesian coordinates without mass weighting. This is the default. The default is 20. CalcFC Specifies that the force constants be computed at the first point CalcAll Specifies that the force constants be computed at every point. Freq Calculate the projected vibrational frequencies for motion perpendicular to the path. FCCards Reads the Cartesian forces and force constants from the input stream after the molecule specifications. MaxCyc=N Sets the maximum number of steps in each geometry optimization. The default is 6. Internal Follow the path in internal (Z-matrix) coordinates without mass weighting. IRC. For example. If both FCCards and ReadIsotopes are specified. Analytic gradients are required for the IRC portion of the calculation (model1 above). some implement the functional specified by the SVWN5 keyword.NAt3).8) The force constants are in lower triangular form: ((F(J. It is equivalent to SVWN. Note that LSDA is not uniquely defined in the literature. RESTART OPTION Restart Restarts an IRC calculation which did not complete. Opt. Cartesian gradients and the Cartesian force constants. or restarts an IRC calculation which did complete. using the Slater exchange functional and the VWN correlation functional for the DFT calculation.8) Force constants (lines of format 6F12. the masses of the atoms are input before the energy. while others use a correlation functional of Perdew. While Gaussian offers this keyword for convenience. but for which additional points along the path are desired.16) Cartesian forces (lines of format 6F12. many differing but related methods are referred to using this term.I).J=1.I). it is probably better practice to specify the exact functional desired. .I=1. Any non-compound energy method and basis set may be used for model2. Freq LSDA This method keyword request a Local Spin Density Approximation calculation. Other programs offering an LSDA method may use somewhat different functionals.Quantum Chemistry Archive using its internal FCList command. The format for this input is: Energy (format D24. In fact. see DFT Methods for full details on specifying and using Density Functional Methods in Gaussian. where NAt3 is the number of Cartesian coordinates. MP4. so the effects of this keyword vary: • • • • • • • SCF energy. if not. it is crucial for a value for MaxDisk to be specified explicitly for these types of jobs. MB. CCSD(T). BD. but obey MaxDisk in avoiding larger storage requirements. Click here for a detailed discussion of the efficient use of disk resources in Gaussian calculations. which must be at least 2ON2. MP2 energies and gradients obey MaxDisk. CCSD. only cubic in the size of the system) and is not usually a limitation. Normally. QCISD densities and CCSD gradients have fixed disk requirements of about N4/2 for closed-shell and 3N4/4 for open-shell. the Stingy. QCISD(T). CI-Singles energies and gradients in the MO basis require about 4O2N2 words of disk for a limited set of transformed integrals. a partial transformation is done and some terms are computed in the AO basis. CCD. The value may optionally be followed by a units designation: KB. Additional scratch space is required during the transformation and this is limited as specified by MaxDisk. If the calculation can be done using a full integral transformation while keeping disk usage under MaxDisk. NoStingy and VeryStingy options are not needed. and frequency calculations use a fixed amount of disk. and CCSD(T) calculations all now look at MaxDisk.MaxDisk The MaxDisk keyword specifies the amount of disk storage available for scratch data. this is set for a site in the site-wide Default.Route file. CISD. Thus. Analytic MP2 frequencies attempt to obey MaxDisk. this is done. and QCISD energies also have a fixed storage requirement proportional to O2N2. CCD. but have minimum disk requirements. gradient. CID. the program now assumes that disk is abundant and performs a full transformation by default (in contrast to Gaussian 94 where a partial transformation was the default in such cases). MW or GB. QCISD(T). CISD. Since MP2 obeys MaxDisk as much as possible. with a large factor. If MaxDisk is not set and sufficient disk space is not available for a full transformation. MP3. GB. . CID.Route file. and BD(T) energies have fixed disk requirements proportional to ON3 which cannot be limited by MaxDisk. KW. either within the route section or via a system wide setting in the Default. in 8-byte words. If MaxDisk is left unset. This disk requirement can be eliminated entirely by performing a direct CI-Singles calculation by using CIS=Direct. QCISD. This is quite small. CCSD. Not all calculations can dynamically control their disk usage. the job will fail. dat) have been updated slightly since the publication of this paper. . y.000000 -. Note that energy differences computed from the values in semi-empirical calculations are in Hartrees and may be compared directly with energy differences computed from jobs using other methods. The MINDO3 energy appears in the output file as follows (followed by the x. Molecular Mechanics Methods There are three molecular mechanics methods available in Gaussian.44]. Dipole moment= .080309984532 NIter= 10. "analytic" gradients. and numerical frequencies.000000 . The following force fields are available: AMBER: The AMBER force field as described in [37].MINDO3 This method keyword requests a semi-empirical calculation using the MINDO3 Hamiltonian [43. We use this current version from the AMBER web site (amber. and z components of the dipole moment): Energy= -. but they are also available as independent methods. respectively). No basis set keyword should be specified. The actual parameters (parm96. DREIDING: The DREIDING force field as described in [38]. Energies. They were implemented for use in ONIOM calculations.739540 The energy is as defined by this semi-empirical model.scripps.edu). UFF: The UFF force field as described in [39]. Restricted open shell (RO) wavefunctions are limited to optimizations using the Fletcher-Powell and pseudoNewton-Raphson methods (the FP and EnOnly options to Opt. No basis set keyword should be specified with these keywords. .. HardFirst Read additional parameters from the input stream. Note that wildcards matches within the hardwared parameter set take precidence over soft parameters. these are referred to as hard-wired parameters. SoftOnly Read parameters from the input stream and use only them. soft ones. no charges are assigned to atoms by default when using any molecular mechanics force field. UnCharged Assign QEq charges for all atoms which have charge zero (i. SoftFirst Read additional parameters from the input stream.CHARGE ASSIGNMENT-RELATED OPTIONS Unless set in the molecule specification input. ChkParameters Read parameters from the checkpoint file. with hard-wired parameters having priority over the read-in. when no relevant option is given. even when the latter contains an exact match for the same item. By default. Hence. all atoms which were untyped or which were given a type but not a charge in the input). Use SoftFirst if you want to override hard-wired parameter matches.e. ignoring hard-wired parameters. PARAMETER PRECEDENCE OPTIONS Terminology: Gaussian contains built-in parameter sets for the built-in force fields listed above. Soft parameters are ones specified by the user in the input stream for the current job (or a previous job when reading parameters from the checkpoint file). read-in parameters are used only if there is no corresponding hard-wired value. Any non-standard (soft) parameters present in the checkpoint file are used with higher priority than corresponding hard-wired parameters. the hard-wired parameters are the only ones used. with soft (read-in) parameters having priority over the hard-wired values. UnTyped Assign QEq charges only to those atoms for which the user did not specify a particular type in the input. Options are available to estimate charges at the initial point using the QEq algorithm under control of the following options for any of the mechanics keywords: QEq Assign charges to all atoms using the QEq method [40]. unless HardFirst is also specified. and frequencies. The default is to abort if there are any ambiguities in the force field. Geom=Connect . For these methods. but they are not required.5. Consult the AMBER paper [37] for definitions of atom types and their associated keywords. use the first one found. HANDLING MULTIPLE PARAMETER SPECIFICATION MATCHES Since parameters can be specified using wildcards.NewParameters Ignore any parameters in the checkpoint file.5 Specifies a carbonyl group oxygen atom with a partial charge of -0. FirstEquiv If there are equivalent matches for a required parameter.32 Specifies an SP3 aliphatic carbon atom with a partial charge of 0. The following options specify other ways of dealing with multiple matches. Analytic energies. INPUT CONVENTIONS AMBER calculations require that all atom types be explicitly specified using the usual notation within the normal molecule specification section: C-CT Specifies an SP3 aliphatic carbon atom. C-CT-0. O-O--0. Atom types and charges may also be provided for UFF and DREIDING calculations. LastEquiv If there are equivalent matches for a required parameter. Modify Read modifications and additions to the parameter set (after it has been constructed from hard and/or soft parameters). ONIOM. use the last one found.32. gradients. it is possible for more than one parameter specification to match a given structure. the program will attempt to determine atom types automatically. Wildcards may be used in any function definition. distances are in Angstroms. They are indicated by a 0 or an asterisk. . the NBDir function entry corresponds to the calculation of all the pairs. In this step. VDW Bond-length Well-depth MMFF94 type Vanderwaals parameters (used for NBDir and NBTerm). without taking the scaling into account. which is automatically expanded into the actual functions NBDir and NBTerm during pre-processing. the computer time for this step scales again linearly with the size of the system. interactions between pairs of atoms that are separated by three bonds or less are usually scaled down (in most force fields. we subtract out the contributions that should have been scaled.0 for pairs separated by one or two bonds. Although at first sight it seems that too much work is done. In equations. Since this involves only pairs that are close to each other based on the connectivity. using a factor 0. VDW94 Atomic-pol NE Scale1 Scale2 DFlag Atomic-pol Atomic polarizability (Angstrom3). we calculate the interactions between all pairs. but were included in the first step. In the soft force field input. R refers to distances and θ refers to angles. NE Slater-Kirkwood effective number of valence electrons (dimensionless). energies are in Kcal/mol and charges are in atomic units. First. In MM force fields. and some value between 0. We follow a two-step procedure. the non-bonded (Vanderwaals and electrostatic) interactions are evaluated for every possible pair of atoms. and the NBTerm entry is used for the subsequent subtraction of the individual pairs. Vanderwaals parameters. to make things easier. the overall algorithm is the more efficient than the alternatives. Function equivalencies to those found in standard force fields are indicated in parentheses. we can use computationally efficient (linear scaling) algorithms.0 and 1. angles are in degrees. However. However. In the second step. used for NBDir and NBTerm (See MMFF94 below for MMFF94-type Vanderwaals parameters). There are a number of ways to implement the calculation of non-bonded interactions. Scale1 Scale factor (Angstrom1/4).GENERAL MOLECULAR MECHANICS FORCE FIELD SPECIFICATIONS Unless otherwise indicated.0 for pairs that are separated by three bonds). you can specify just the non-bonded master function NonBon. 2 scaling is used (as for Amber). If any scale factor < 0. V-Cutoff. C-Type.0 for acceptor type. Coulomb and Vanderwaals single term cutoffs NBTerm Atom-type1 Atom-type2 V-Type C-Type V-Cutoff C-Cutoff V-Scale C-Scale .0.Scale2 DFlag Scale factor (dimensionless).0. the 1/1. and C-Cutoff as above. CScale1-3 are Coulomb scale factors for 1 to 3 bond separated pairs. NonBon V-Type C-Type. otherwise 0. Coulomb and Vanderwaals direct (evaluated for all atom pairs).0 for donor type atom. This function will be expanded into pairs and a direct function (NBDir and NBTerm) before evaluation of the MM energy. 1. V-Cutoff C-Cutoff VScale1 VScale2 VScale3 CScale1 CScale2 CScale3 V-Type is the Vanderwaals type: 0 No Vanderwaals 1 Arithmetic (as for Dreiding) 2 Geometric (as for UFF) 3 Arithmetic (as for Amber) 4 MMFF94-type Vanderwaals C-Type is the Coulomb type: 0 No Coulomb 1 1/R 2 1/R2 3 1/R buffered (MMFF94) V-Cutoff and C-Cutoff are the Vanderwaals and Coulomb cutoffs (respectively): 0 No cutoff >0 Hard cutoff <0 Soft cutoff VScale1-3 are Vanderwaals scale factors for 1 to 3 bond separated pairs. 2. NBDir V-Type C-Type V-Cutoff C-Cutoff V-Type. MMFF94 electrostatic buffering Buf94 Atom-type Value Non-bonded interaction master function. Atomic single bond radius AtRad Atom-type Radius Effective charge (UFF) EffChg Charge GMP Electronegativity (UFF) EleNeg Value Step down table Table Original-atom-type Stepping-down-type(s).Sqrt(Xj)]2/(Xi*Ri + Xj*Rj) . Harmonic stretch III (UFF [1a]): k*(R-Rij)2 Equilibrium bond length Rij = (1 . and C-Scale as above. Harmonic stretch I (Amber [1]): ForceC*(R-Req)2 HrmStr1 Atom-type1 Atom-type2 ForceC Req ForceC Force constant Req Equilibrium bond length Harmonic stretch II (Dreiding [4a]): ForceC*[R-(Ri+Rj-Delta)]2 HrmStr2 Atom-type1 Atom-type2 ForceC Delta ForceC Force constant Delta Delta Ri and Rj are atomic bond radii specified with AtRad.V-Type.12*Zi*Zj/(Rij3) Electronegativity correction: Ri*Rj*[Sqrt(Xi) . C-Cutoff.PropC*lnBO)*(Ri + Rj) + Ren Force constant: k = 664. C-Type. V-Scale. V-Cutoff. Morse stretch III (UFF [1b]): A1*A3*(exp[-a(R-Rij)]-1)2 where a = Sqrt(k/[BO*PropC]) Equilibrium bond length Rij = (1 .Sqrt(Xj))2/(Xi*Ri + Xj*Rj) MrsStr3 Atom-type1 Atom-type2 BO PropC BO Bond order (if <0. Morse stretch I (Amber): DLim*(e-a(R-Req)-1)2 where a = Sqrt(ForceC/DLim) MrsStr1 Atom-type1 Atom-type2 ForceC Req DLim ForceC Force constant Req Equilibrium bond length DLim Dissociation limit Morse stretch II (Dreiding [5a]): DLim*exp[-a(Ri+Rj-Delta)]-1)2 where a = Sqrt(ForceC/DLim) MrsStr2 Atom-type1 Atom-type2 ForceC Delta DLim ForceC Force constant Delta Delta DLim Dissociation limit Ri and Rj are atomic bond radii defined with AtRad. it is determined on-the-fly) PropC Proportionality constant Ri and Rj are atomic bond radii defined with AtRad. Zi and Zj are the effective atomic charges defined with EffChg. Xi and Xj are GMP electronegativity values defined with EleNeg.HrmStr3 Atom-type1 Atom-type2 BO PropC BO Bond order (if <0. . Xi and Xj are GMP electronegativity values defined with EleNeg. it is determined on-the-fly) PropC Proportionality constant Ri and Rj are atomic bond radii defined with AtRad.12*Zi*Zj/Rij3 Electronegativity correction: Ren = Ri*Rj*(Sqrt(Xi) .PropC*lnBO)*(Ri + Rj) + Ren Force constant k = 664. Zi and Zj are the effective atomic charges defined with EffChg. Quartic stretch I (MMFF94 [2]): (Req/2)*(R-ForceC)2*[1+CStr*(R-ForceC+(7/12)*CStr2*(R-ForceC)2] QStr1 Atom-type1 Atom-type2 ForceC Req CStr ForceC Force constant (md-Angstrom-1) Req Equilibrium bond length (Angstrom) CStr Cubic stretch constant (Angstrom-1) Atomic torsional barrier for the oxygen column (UFF [16]) UFFVOx Barrier Atomic sp3 torsional barrier (UFF [16]) UFFVsp3 Barrier Atomic sp2 torsional barrier (UFF [17]) UFFVsp2 Barrier Harmonic bend (Amber [1]): ForceC*(T-θeq)2 HrmBnd1 Atom-type1 Atom-type2 Atom-type3 ForceC θeq ForceC Force constant (in kcal/(mol*rad2) θeq Equilibrium angle Harmonic Bend (Dreiding [10a]): [ForceC/sin(θeq2)]*(cos(θ)-cos(θeq))2 HrmBnd2 Atom-type1 Atom-type2 Atom-type3 ForceC θeq ForceC Force constant θeq Equilibrium angle Dreiding Linear Bend (Dreiding [10c]): AForceC*(1+cos(θ)) LinBnd1 Atom-type1 Atom-type2 Atom-type3 ForceC . Zj and Zk are effective atomic charges defined with EffChg. Xi. 4 for square-planar.12*Zi*Zk*(3*Rij*Rjk*(1-cos(θeq2))-cos(θeq)*Rik2)/Rik5 UFFBnd3 Atom-type1 Atom-type2 Atom-type3 θeq BO12 BO23 PropC θeq Equilibrium angle BO12 Bond order for Atom-type1–Atom-type2 (when <0. Rj and Rk are atomic bond radii defined with AtRad. BO12 Bond order for Atom-type1–Atom-type2 (when <0. C1 = -4*C2*cos(θeq) and C0=C2*(2*cos(θeq2)+1) Force constant: k = 664. Xj and Xk are GMP electronegativity defined with EleNeg. UFF 2-term bend (UFF [10]): [k/(Per2)]*[1-cos(Per*θ)] Force constant: k = 664. it is determined on-the-fly) PropC Proportionality constant Ri. Zi. Zj and Zk are effective atomic charges defined with EffChg. it is determined on-the-fly) BO23 Bond order for Atom-type2–Atom-type3 (when <0.ForceC Force constant UFF 3-term bend (UFF [11]): k*(C0 + C1*cos(θ))+C2*cos(2θ) where C2=1/(4 * sin(θeq2)). Xj and Xk are GMP electronegativity defined with EleNeg.12*Zi*Zk*(3*Rij*Rjk*(1-cos(Per2))-cos(Per)*Rik2)/Rik5 UFFBnd2 Atom-type1 Atom-type2 Atom-type3 Per BO12 BO23 PropC Per Periodicity: 2 for linear. This term is needed for the program not to protest about undefined angles. Xi. Zero bend term: used in rare cases where a bend is zero. it is determined on-the-fly) PropC Proportionality constant Ri. it is determined on-the-fly) BO23 Bond order for Atom-type2–Atom-type3 (when <0. 3 for trigonal. Rj and Rk are atomic bond radii defined with AtRad. Zi. ZeroBnd Atom-type1 Atom-type2 Atom-type3 Cubic bend I (MMFF94 [3]): (ForceC/2)*(1+CBend*(θ-θeq))*(θ-θeq)2 CubBnd1 Atom-type1 Atom-type2 Atom-type3 ForceC θeq CBend . Dreiding torsion (Dreiding [13]): V*[1-cos(Period*(θ-PO))]/(2*NPaths) DreiTrs Atom-type1 Atom-type2 Atom-type3 Atom-type4 V PO Period NPaths V Barrier height V PO Phase offset Period Periodicity NPaths Number of paths (if < 0.ForceC Force constant (in md*Angstrom/rad2) θeq Equilibrium angle CBend "Cubic Bend" constant (in deg-1) MMFF94 Linear Bend (MMFF94 [4]): ForceC*(1+cos(θ)) LinBnd2 Atom-type1 Atom-type2 Atom-type3 ForceC ForceC Force constant (md) Amber torsion (Amber [1]): Σi=1.Mag4 V/2 magnitudes NPaths Number of paths (if < 0... determined on-the-fly). determined on-the-fly. When zero or less. determined on-the-fly).4 (Magi*[1+cos(i*θ-I(i+4))])/NPaths AmbTrs Atom-type1 A-type2 A-type3 A-type4 PO1 PO2 PO3 PO4 Mag1 Mag2 Mag3 Mag4 NPaths PO1-PO4 Phase offsets Mag1. UFF torsion with constant barrier height (UFF [15]): [V/2]*[1cos(Period*PO)*cos(V*θ)]/NPaths UFFTorC Atom-type1 Atom-type2 Atom-type3 Atom-type4 Period PO V NPaths Period Periodicity PO Phase offset V Barrier height V NPaths Number of paths. UFF torsion with bond order based barrier height (UFF [17]): . and NPaths=-1. PO=0.0.0. these values are used: V=1. determined on-the-fly. Dreiding special torsion for compatibility with Gaussian 98 code. but the fourth center is not H.0. PO=0. and at least one of them is H. UFF torsion with atom type-based barrier height (UFF [16]): [V/2]*[1-cos(Period*PO)* cos(Period*θ)]/NPaths where V=Sqrt(Vj*Vk) UFFTor1 Atom-type1 Atom-type2 Atom-type3 Atom-type4 Period PO NPaths Period Periodicity PO Phase offset NPaths Number of paths.0. Period=6. it is determined on-the-fly) NPaths Number of paths (when <0. . Vj and Vk are atomic constants defined with UFFVsp3. Vj and Vk are atomic constants from UFFVOx. it is removed. then these values are used: V=4.0. Period=3.0. determined on-the-fly. If there are three atoms bonded to the third center. it is determined on-the-fly) Uj and Uk are atomic constants defined with UFFVsp2. and NPaths=-1. it is replaced with DreiTRS. When zero or less. During processing.0. Otherwise. with the following parameters: • • • If there are three atoms bonded to the third center and the fourth center is H.18*Log(BO12)] UFFTorB Atom-type1 Atom-type2 Atom-type3 Atom-type4 Period PO BO12 NPaths Period Periodicity PO Phase offset BO12 Bond order for Atom-type1–Atom-type2 (when <0.0. When zero or less.[V/2]*[1-cos(Period*PO)* cos(Period*θ)]/NPaths where V = 5*Sqrt(Uj*Uk)*[1+4. UFF torsion with atom type based barrier height (UFF [16]) (differs from UFFTor1 in that the atomic parameter that is used): [V/2]*[1cos(Period*PO)*cos(Period*θ)]/NPAths where V=Sqrt(Vj*Vk) UFFTor2 Atom-type1 Atom-type2 Atom-type3 Atom-type4 Period PO NPaths Period Periodicity PO Phase offset NPaths Number of paths. . HrmWil Atom-type1 Atom-type2 Atom-type3 Atom-type4 ForceC ForceC Force constant Stretch-bend I (MMFF94 [5]): (ForceC1*(R12-Req12)+ForceC2*(R32-Req23))*(θ-θeq) StrBnd1 Atom-type1 Atom-type2 Atom-type3 ForceC1 ForceC2 Req12 Req23 θeq ForceC1. Wilson Atom-type1 Atom-type2 Atom-type3 Atom-type4 ForceC C1 C2 C3 ForceC Force constant C1. UFF [19]): ForceC*(C1 + C2*cos(θ) + C3*cos(2θ)) averaged over all three Wilson angles θ.00 ≤ bond order < 1. separated by a hyphen (e. ForceC2 Force constants (in md/rad) Req12.g. C3 Coefficients Harmonic Wilson angle (MMFF94 [6]): (ForceC/2)*(θ2) summed over all three Wilson angles θ. The following substructures apply to functions related to bond stretches: • -1 Single bond: 0. C2.OldTor Atom-type1 Atom-type2 Atom-type3 Atom-type4 Improper torsion (Amber [1]): Mag*[1+cos(Period*(θ-PO))] ImpTrs Atom-type1 Atom-type2 Atom-type3 Atom-type4 Mag PO Period Mag V/2 Magnitude PO Phase offset Period Periodicity Three term Wilson angle (Dreiding [28c]. HrmStr-1. Substructure numbers are appended to the function name. Req23 Equilibrium bond lengths θeq Equilibrium angle USING SUBSTRUCTURES Substructures may be used to define different parameter values for a function for distinct ranges of some geometrical characteristic.50 . HrmStr-2 and so on). g.00 ≤ bond order < 1.50 ≤ bond order < 2.50 Double central bond: 1. AmbTrs-1-2).70) Amide central bond (priority over resonance) None of the above Here is some simple MM force field definition input: HrmStr1 HrmStr1-1 HrmStr1-2 HrmBnd2 DreiTrs-1 DreiTrs-2 H_ C_2 C_2 * * * C_2 360.0 180.30 ≤ bond order ≤ 1.0 120.0 -1. For dihedral angles.50 C_2 500. Use a zero for the first substructure to specify only the second substructure.0 1.50 The following substructures apply to functions for bond angles (values in degrees): First substructure: • • • -1 -2 -3 0 ≤ θ ≤ 45 45 < θ ≤ 135 135 < θ ≤ 180 Second substructure: • -i-n Number of atoms bonded to the central one.50 ≤bond order < 2.0 180.0 -1.0 2. First substructure: • • • • -0 -1 -2 -3 Skip this substructure (substructure "wildcard") Single central bond: 0.50 Second substructure: • • • -i-1 -i-2 -i-3 Resonance central bond (1.50 Triple central bond: bond order ≥ 2.• • -2 -3 Double bond: 1.0 C_2 C_2 * 5. one or two substructures may be used (e.40 C_2 * 50.0 2.50 Triple bond: bond order ≥ 2.0 1.08 C_2 350.0 C_2 C_2 * 45.0 MNDO ..0 1. 0908412558735 NIter= 10.52.46.65]. truncated at second-order for MP2 [21.66]. third order for MP3 [61. MP2 MP3 MP4 MP5 These method keywords request a Hartree-Fock calculation (RHF for singlets.140].50.49.47. y. and z components of the dipole moment): Energy= -. Note that energy differences computed from the values in semi-empirical calculations are in Hartrees and may be compared directly with energy differences computed from jobs using other methods. "analytic" gradients.23.45.000000 -.142]. Energies.54].739540 The energy is as defined by this semi-empirical model. and analytic frequencies are available for MP2 [25]. Restricted open shell (RO) wavefunctions are limited to optimizations using the Fletcher-Powell and pseudoNewton-Raphson methods (FP and EnOnly. and numerical frequencies.51. Dipole moment= .139. The MNDO energy appears in the output file as follows (followed by the x. MP3 and MP4(SDQ) [141. AVAILABLE ALGORITHMS FOR MP2 There are four basic algorithms for MP2 calculations and for producing transformed (MO) integrals on disk: .22.23.48. fourthorder for MP4 [62].000000 .25. UHF for higher multiplicities) followed by a Møller-Plesset correlation energy correction [60].This method keyword requests a semi-empirical calculation using the MNDO Hamiltonian [43. and fifth-order for MP5 [64]. Analytic gradients are available for MP2 [22. respectively). No basis set keyword should be specified. which uses both main memory and external (disk) storage as available [23]. Conventional. and the only method for generating MO integrals on disk in Gaussian 90. Just specifying MP4 defaults to MP4(SDTQ). The available disk can be specified via the MaxDisk keyword. It is seldom a good choice on any but the smallest computer systems. Note that selection of the direct or semi-direct MP2 and transformation algorithms is separate from selecting direct SCF (which is the default SCF algorithm in Gaussian 03). Thus. MP4(SDQ) for single.• • • • Semi-Direct. either in the route section or (preferably) in the Default. which requires no external storage beyond that for the . or MP4(SDTQ) for full MP4 with single. This method requires O3V3 disk storage and scales as O4V4 in cpu time. which uses no external storage by recomputing the integrals as needed during the transformation. and so specifying MP5 defaults to a UMP5 calculation. direct.Route file. and semi-direct algorithms based on available memory and disk. Direct. VARIATIONS OF MP4 MP4(DQ) is specified to use only the space of double and quadruple substitutions. triple and quadruple substitutions [62. This is the default algorithm. ALGORITHM SELECTION OPTIONS (MP2 METHODS) Note: The appropriate algorithm for MP2 will be selected automatically based on the settings of %Mem and MaxDisk. in which all the AO integrals are generated and stored in main memory. This was the only method available in Gaussian 88. which stores the transformed integrals on disk. FROZEN-CORE OPTIONS (POST-SCF METHODS) FC The frozen-core options for defining inner-shells to be excluded from the correlation calculation are valid with these keywords. double and quadruple substitutions. The E(2) calculation or transformation then recomputes integrals as needed in the form required for vectorization.63]. then used without storing them externally. double. The default is to decide between the in-core. See the discussion here for details. these options are almost never needed. LIMITATIONS FOR MP5 The MP5 code has been written for the open shell case only. In-core. FullDirect Forces the "fully direct" algorithm. 32068082D-02 UMP4(DQ)= E4(SDQ)= -.75017782203D+02 -.3906492545D-01 EUMP2= -. except for machines with very large main memory and limited disk. Transformation. labeled as EUMP2: E2= -. and numerical frequencies. MP4(SDTQ) and MP5: Analytic energies. InCore Forces the in-memory algorithm.55601167D-04 E3= -. V=number of virtual orbitals. Requires a minimum of 2OVN words of main memory (O=number of occupied orbitals. HF. This is very fast when it can be used.33238377D-02 UMP4(SDQ)= -. ROMP2 is available for energies only. The MP2 energy appears in the output as follows.75017899233D+02 . and analytic frequencies.04 seconds.75003727493390D+02 Energies for higher-order Møller-Plesset methods follow. SCF. MP4(DQ) and MP4(SDQ): Energies. Direct Requests some sort of direct algorithm.10847902D-01 EUMP3= E4(DQ)= -. SemiDirect Forces the semi-direct algorithm. It is normally used in conjunction with SCF=InCore. analytic gradients. MP3. The choice between in-core.75014575395D+02 -. MP4(T)= -.SCF. N=number of basis functions). Here is the output from an MP4(SDTQ) calculation: Time for triples= . and numerical frequencies. fully direct and semidirect is made by the program based on memory and disk limits and the dimensions of the problem. numerical gradients. but requires N4/4 words of memory. analytic gradients. MaxDisk Energies. This is seldom a good choice. MP2: Energies. NoInCore prevents the use of the in-core algorithm. 240. Rearchive NMR This properties keyword predicts NMR shielding tensors and magnetic susceptibilities using the Hartree-Fock method.233] (a slight variation on the CSGT method) and the Single Origin method.229.241].239. the default for the user name is the operating system-level login name of the user who runs the job. for both shielding tensor and magnetic susceptibilities. . via the SpinSpin option.75017954834D+02 The energy labelled EUMP3 is the MP3 energy.33794389D-02 UMP4(SDTQ)= -. all DFT methods and the MP2 method [232. Name=RChavez). Archive. On UNIX systems. and the various MP4-level corrections appear after it.g. with the MP4(SDTQ) output coming in the final line (labeled UMP4(SDTQ)).234. It takes the desired username as its parameter (e. Structures used for NMR calculations should have been optimized at a good level of theory. Spin-spin coupling constants may also be computed during an NMR job [238. Name This keyword specifies the username that is stored in the archive entry for the calculation.230]. Note that CSGT calculations require large basis sets to achieve accurate results.E4(SDTQ)= -.227.228. NMR shielding tensors may be computed with the Continuous Set of Gauge Transformations (CSGT) method [231. Magnetic susceptibilities may also be computed with both GIAOs [236. Gaussian also supports the IGAIM method [231.235] and the Gauge-Independent Atomic Orbital (GIAO) method [226.528]..237] and CGST. Test. This keyword is of most use to Gaussian users who also use the Browse Quantum Chemistry Database System.233. 0000 ZZ= 187.3406 Anisotropy = 5.4233 For this molecular system.0000 2 H Isotropic = 23. All Compute properties with all three of the SingleOrigin.4092 ZX= .0000 ZY= .9397 XX= 27. GIAO Compute NMR properties using the GIAO method only. IGAIM Use atomic centers as gauge origins. NMR may be combined with SCRF.2745 ZX= . This calculation type has a computational cost of about twice that of computing vibrational frequencies. CSGT Compute NMR properties using the CSGT method only.0000 XY= . This is the default.0000 ZY= .0000 ZZ= 20. In Gaussian 03. the values for all of the atoms of a given type are equal.0000 YY= -62.0670 XZ= .0000 YZ= . .SpinSpin Compute spin-spin coupling constants in addition to the usual NMR properties. DFT and MP2 methods.5514 XZ= . PrintEigenvectors Display the eigenvectors of the shielding tensor for each atom SCF. so we have truncated the output after the first two atoms. IGAIM. It is available only for Hartree-Fock and DFT methods. and CSGT methods.3287 YX= . SingleOrigin Use a single gauge origin.0000 YZ= .0000 Anisotropy = 194.0000 YY= 24. This method is provided for comparison purposes but is not generally recommended.7345 XX= 48. Here is an example of the default output from NMR: Magnetic properties (GIAO method) Magnetic shielding (ppm): 1 C Isotropic = 57.4143 YX= .0000 XY= . 000000D+00 2 0. those in the model system) and those in Cartesian coordinates (typically. the ONIOM optimization procedure is enhanced in Gaussian 03 to use microiterations [163] and an optional quadratic coupled algorithm [162]. the molecular system being studied is divided into two or three layers which are treated with different model chemistries.The additional output from spin-spin coupling computations appears as follows: Total nuclear spin-spin coupling K (Hz): 1 2 1 0. Medium and High layers. Electronic embedding incorporates the partial charges of the MM region into the quantum mechanical Hamiltonian.or three-layer ONIOM [153. ONIOM(MO:MM) calculations can take advantage of electronic embedding.158. any conventional calculation can be viewed as a one-layer ONIOM.156.000000D+00 Total nuclear spin-spin coupling J (Hz): 1 2 1 0.155.157. The layers are conventionally known as the Low. For ONIOM(MO:MM) jobs.159]. the atoms only in the MM layer) in order to produce more accurate steps (the latter can be requested with Opt=QuadMacro).154. By default. and the J matrix gives the values taking the job's specific isotopes into account (whether explicitly specifed or the default isotopes). The results are then automatically combined into the final predicted results. atoms are placed into the High layer.000000D+00 The various components of the coupling constants precede this section in the output file. It displays the matrix of isotropic spin-spin coupling between pairs of atoms in lower triangular form.432614D+03 0. REQUIRED INPUT .147308D+02 0. ONIOM This keyword requests a two. This technique provides a better description of the electrostatic interaction between the QM and MM regions and allows the QM wavefunction to be polarized. In this procedure. (From a certain point of view.000000D+00 2 0. The K matrix gives the values which are isotope-independant. The latter takes into account the coupling between atoms using internal coordinates (typically.) Layer assignments are specified as part of the molecule specification (see below). if only two parameters are specified. Layer is a keyword indicating the layer assignment for the atom. The other optional parameters specify how the atoms located at a layer boundary are to be treated. Gaussian 03 determines bond distances between atoms and their link atoms by scaling the original bond distance (i. Medium and Low. using UFF for the Low layer. up to three scale factors may be specified (in the order low. via additional parameters on each line according to the following syntax: atom coordinate-spec layer [link-atom [bonded-to [scale-fac1 [scale-fac2 [scale-fac3]]]]] where atom and coordinate-spec represent the normal molecule specification input for the atom. then you must explicitly set the first scale factor to 0. Gaussian 03 does not define them automatically or provide any defaults. then the program will determine the corresponding scale factor in the normal way. Thus. In this case. the scale factors specify the link atom bond distance in the model system when calculated at the low and high levels (respectively). If a scale parameter is explicitly set to 0. then all three scale factors will use that value. in the real system). For a two-layer calculation. one of High. then both scale factors will use that value.The two or three desired model chemistries are specified as the options to the ONIOM keyword. Low (the final one may obviously be omitted). If it is omitted. this route section specifies a threelayer ONIOM calculation. if only one parameter is specified. if you want to change only the second scale factor (model system calculated at the medium level).0. medium. All of these scale factors correspond to the g-factor parameter as defined in reference [158].. AM1 for the Medium layer. Link atoms are necessary when covalent bonding exists between atoms in different layers in order to saturate the (otherwise) dangling bonds.0. You use link-atom to specify the atom with which to replace the current atom (it can include atom type and partial charge and other parameters). you can also specify these scale factors explicitly. Note: All link atoms must be specified by the user. The distinct models are separated by colons. For a three-layer ONIOM. For a three-layer ONIOM. Medium. then the third scale factor will use the second value. For example. Gaussian will attempt to identify it automatically. in the order High. and HF for the High layer: # ONIOM(HF/6-31G(d):AM1:UFF) Atom layer assignment is done as part of the molecule specification. for a three-layer ONIOM.e. the third scale factor will have the same value as the second parameter unless it is explicitly . However. extended to allow separate values for each ONIOM calculation level. For a two-layer ONIOM. using scaling factors which the program determines automatically. The bonded-to parameter specifies which atom the current atom is to be bonded to during the higher-level calculation portion. In general. if only one parameter is specified. high). Medium and Low levels): cRealL sRealL [cIntM sIntM [cIntL sIntL [cModH sModH [cModM sModM [cModL sModL]]]]] For 3-layer ONIOM=SValue calculations. PER-LAYER CHARGE AND SPIN MULTIPLICITY Multiple charge and spin multiplicity pairs may also be specified for ONIOM calculations. then the third pair defaults to the same values as in the second pair.. the first character is one of: Real. in this second context.. The fourth pair applies only to ONIOM=SValue calculations. it defaults to the values for the real system at the low level. and Mod=Model system. M and L for the High. For two-layer ONIOM jobs. the charge and spin multiplicity items default according to the following scheme.assigned a non-zero value (i. and the second character is one of: H. the format for this input line is: chrgreal-low spinreal-low [chrgmodel-high spinmodel-high [chrgmodel-low spinmodel-low [chrgreal-high spinreal-high]]] where the subscript indicates the calculation for which the values will be used.. where the number in each cell indicates which pair of values applies for that calculation in the corresponding :circumstances: Charge & Spin Defaults # Pairs Specified (SValue only) 1 2 3 4 5 6 7 8 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 1 2 3 3 3 3 3 3 1 2 2 4 4 4 4 4 1 2 2 4 5 5 5 5 1 2 2 4 5 6 6 6 1 2 2 2 2 2 7 7 1 1 1 1 1 1 1 8 1 1 1 1 1 1 1 8 Calculation Real-Low Int-Med Int-Low Model-High Model-Med Model-Low Int-High Real-Med Real-High 9 1 2 3 4 5 6 7 8 9 . up to three additional pairs may be specified: . If the final pair is omitted for an S-value job. all levels will use those values. cIntH sIntH [cRealM sRealM [cRealH sRealH]] Defaults for missing charge/spin multiplicity pairs are taken from the next highest calculation level and/or system size. Values and defaults for three-layer ONIOM calculations follow an analogous pattern (in the subscripts below. When only a single value pair is specified. when only a subset of the six or nine pairs are specified. 0.e.0 has the same meaning as an omitted value). Thus. Int=Intermediate system. If two pairs of values are included. NoCompress performs the calculation without compression. and they are specified in the usual manner (see the examples). ChkBas. the values of i through n must be monotonically decreasing. Density fitting sets may also be used when applicable. then numerical frequencies will be computed for all models. Additional charge and spin multiplicity pair(s) may be specified for the additional calculations (see below). NMR calculations may be performed with the ONIOM model.2n. Thus. and so on. ScaleCharge=ijklmn Specifies scaling parameters for MM charges during electronic embedding in the QM calculations. The default value is 500 (i.. those two bonds away use 0. 123500 and 500 are all equivalent. 555500). ONIOM can also perform CIS and TD calculations for one or more layers. Sparse and NoFMM keywords may also be specified for relevant models. However. Compress Compress operations and storage to active atoms during ONIOM second derivative calculations. Pseudo=Read.e. . MK Specifies that Merz-Kollman-Singh (see Population) approximate charges be used during geometry optimization microiterations with electronic embedding. Atoms bonded to the inner layers use a scale factor of 0. gradients and frequencies. even when analytic frequencies are available. It is the default. SValue Requests that the full square be done for testing. ScaleCharge implies EmbedCharge. The integers are multiplied by 0. 555500. NoEmbedCharge is the default.EmbedCharge Use MM charges from the real system in the QM calculations on the model system(s). this is the default. Blank does the uncompressed calculation but then discards contributions from inactive atoms (which are currently non-zero only for nuclear moment perturbations: shielding and spin-spin coupling tensors). Energies. and the largest value among them is used for all parameters to its left. The Gen. Note that if any of the specified models require numerical frequencies.2 to obtain the actual scale factors.2m. to produce substituent values ( S-values) for the S-value test [160]. 1.25 C-CA--0.1 C-CA--0.1 5 -0.419886 -0.B3.0 H.703834 -3.1 H-HA-0. The other atoms are explicitly placed into the Medium and Low layers.A3.A5.0 H.5.2.-D2. Opt=QuadMacro Molecule Specifications for ONIOM Jobs.578498 -0.4.D1.B1 H.B7.779173 L -0. Note that the Z-matrix specification must include the final 0 code indicating the Z-matrix format when ONIOM input is included.25 C-CA--0.B5.778723 H -0.B2.25 C-CA--0.787462 -2.609095 -3.8.615995 0.0 C.5.4.0 H.1.187368 -1.0.0. Here is an input file for a two-layer ONIOM calculation using a DFT method for the high layer and Amber for the low layer.331033 -2. Molecular Mechanics keywords.25 C-CA--0.778859 L H-HC-0.A6.748381 -5.1.A1 C.180.5.B5.779093 L -0.4. Note that atom types are used for both the main atom specifications and the link atoms: # ONIOM(B3LYP/6-31G(d):Amber) Geom=Connectivity 2 layer ONIOM job 0 1 0 1 0 1 model-low C-CA--0.B4.779093 H -0.5.779093 L H-HA-0.4.841116 -0.1 3 .619477 -1.A2.2.A4. Here is a simple ONIOM input file: # ONIOM(B3LYP/6-31G(d.778569 H -0.3.766244 -1.779356 L H-HA-0.D2.25 H-HA-0.607871 0.785438 -0.-D1.779321 L -0.264565 -2.841116 -1.779336 L -0.0 variable definitions M H L H M M L L L The High layer consists of the first three atoms (placed there by default).326965 -4.A6.p):AM1:UFF) Opt Test 3-layer ONIOM optimization 0 1 C O.0 H. The molecule specification includes atom types (which are optional with UFF but required by Amber).1.25 Charge/spin for entire molecule (real system).4.8.B7.1.540960 -2.5.640022 -5.1 3 -0.4.1.180.0 H.586452 -0.B6.0.1.1 H-HA-0.Geom=Connect.25 C-CA--0. model system-high level & 0 0 0 0 0 0 0 0 0 0 -4.A4.180.2. 0 8 9 10 15 1.519523 -1.231354 -0.0 H 0 0.1 H-HA-0.5 11 14 1.1 L L H-HC-0. then the corresponding atom is frozen during geometry optimizations: C -1 0.0 12 13 14 15 1.5 12 1.637238 -0.C-CA--0. Note that it contains connectivity information for use with Geom=Connect.0 6 13 1.0 16 17 18 This input file was created by GaussView.778608 -0.1 L L H-HA-0.5 17 1.5 16 1. ONIOM optimizations can take advantage of the optional second field within molecule specifications. If it is set to -1.0 0. Here is an example of a complex ONIOM route section: # ONIOM(BLYP/6-31G(d)/Auto TD=(NStates=8):UFF) This example uses density fitting for the DFT high layer time-dependent excited states calculation.0 3 4 1. This job also illustrates the use of multiple charge and spin multiplicity values for ONIOM jobs.0 15 18 1.584286 1.5 11 1.25 H-HA-0.532954 -0.9 ..610773 2.0 0.295622 -6.25 H-HA-0. A Complex ONIOM Route.670662 0. We thank Prof.168671 -0.778881 -0.645844 1.779522 H L H-HA-0. .1 H-HA-0.515342 -3.0 0.832597 1.5 4 5 1.5 6 1.832665 0.25 C-CA--0. This example should be used for illustrative purposes only.778487 -0.1 C-CA--0.5 8 1.779059 -0. This field defaults to 0 if omitted.1 H-HA-0.0 2 3 1.0 7 10 1. Freezing Atoms During ONIOM Optimizations.1 L L 5 11 11 1 2 1.778996 0.604215 -5.0 0.5 9 1.5 5 6 1.779340 -0.778757 -0. K Nishimoto for pointing out several scientific problems with running this calculation.777138 2.1 0 0 0 0 0 0 0 0 -2.778348 -0..5 10 1. Here is some output from the ONIOM=SValue option: S-Values (between gridpoints) and energies: high med low 4 -39. Overview of geometry optimizations in Gaussian 03. which are the default. and it includes these subsections: • • • Options to the Opt keyword. CISD. Ways of generating initial force constants. S-Value Test. MP4(SDQ).479426 -153. The remainder of this quite lengthy section discusses various aspects of geometry optimizations. Opt This keyword requests that a geometry optimization be performed. . CCD. CASSCF. However.930320 -189.588420 3 -38. These are the s-values obtained with the absolute energies. and all DFT and semi-empirical methods.529].106289 8 -114.507971 model mid real The integers are the gridpoints.107344 2 -39.577651 6 -112.322207 7 -39.Note that the atom will also be frozen during non-ONIOM optimizations provided that they are performed in coordinates other than redundant internal coordinates.305712 9 -114. QCISD.801632 -193.15] (specified by the Redundant option).266459 1 -38. use the Opt=ModRedundant option to freeze atoms. be aware than when applying the S-value test. CCSD. if the field is set to a negative value other than -1. For the latter.341899 -150. The geometry will be adjusted until a stationary point on the potential surface is found. Horizontally between the grid points are the S-values.148. For ONIOM jobs only. the default algorithm for both minimizations (optimizations to a local minimum) and optimizations to transition states and higher-order saddle points is the Berny algorithm using redundant internal coordinates [149. For the Hartree-Fock. CIS.041481 -153. CID. it is treated as part of a rigid fragment during the optimization: all atoms with the same value (< -1) move only as a rigid block.160170 -192. MP3. relative energies and S-values need to be used (see reference [160]). Gradients will be used if available. The Berny algorithm using internal coordinates (Opt=Z-matrix) is also available [136. MP2. and under each one is the energy value.118688 5 -39. The default algorithm for all methods lacking analytic gradients is the eigenvalue-following algorithm (Opt=EF). product. No coordinate may be frozen during this type of calculation. Note that the atoms must be specified in the same order within the three structures.• • • • Optimizing to transition states and higher-order saddle points. GENERAL PROCEDURAL OPTIONS MaxCycle=N Sets the maximum number of optimization steps to N. See also Appendix B if you are interested in details about setting up Zmatrices for various types of molecules. including examples of Opt input and output and using the ModRedundant option. and initial TS structures as input. This option requires the reactant.01N Bohr or radians. The default value for N is 30. Summary of the Berny optimization algorithm. TS Requests optimization to a transition state rather than a local minimum. QST3 Search for a transition structure using the STQN method. Basic information as well as techniques and pitfalls related to geometry optimizations are discussed in detail in chapter 3 of Exploring Chemistry with Electronic Structure Methods [308]. This option requires the reactant and product structures as input. Examples for Opt=Z-matrix. QST2 Search for a transition structure using the STQN method. specified in three consecutive groups of title and molecule specification sections. Users should consult those subsection(s) that apply to their interests and needs. MaxStep=N Sets the maximum size for an optimization step (the initial trust radius) to 0. Path=M In combination with either the QST2 or the QST3 option. Note that the atoms must be specified in the same order in the two structures. TS should not be specified with QST2. The default is the maximum of 20 and twice the number of redundant internal coordinates in use (for the default procedure) or twice the number of variables to be optimized (for other procedures). specified in two consecutive groups of title and molecule specification sections. TS should not be specified with QST3. Saddle=N Requests optimization to a saddle point of order N. Notes on optimizing in redundant internal coordinates. . requests the simultaneous optimization of a transition state and an M-point reaction path in redundant internal coordinates [164]. At each step in the path relaxation. OptReactant Specifies that the input structure for the reactant in a simultaneous optimization calculation should be optimized to a local minimum. Instead.If QST2 is specified. The highest energy structure becomes the initial guess for the transition structure. Optproduct may not be combined with BiMolecular. The treatment of the input reactant and product structures is controlled by other options: OptReactant. the central point is optimized to the transition structure. BiMolecular. By default. The remaining M-2 points on the path are then generated by linear interpolation between the reactant and product input structures. This anchor point will not appear as one of the M points on the path. first between the reactant and transition structure and then between the transition structure and product. the highest point at each step is optimized toward the transition structure. In the output for a simultaneous optimization calculation. OptReactant may not be combined with BiMolecular. a third set of title and molecule specification sections must be included in the input as a guess for the transition state as usual. Note that the SCF wavefunction for structures in the reactant valley may be quite different from that of structures in the product valley. M must be an odd number so that the points on the path may be distributed evenly between the two sides of the transition structure. OptProduct. This is the default. BiMolecular Specifies that the reactants or products are bimolecular and that the input structure will be used as an anchor point. the predicted geometry for the optimized transition structure is followed by a list of all M converged reaction path structures. In this case. . If QST3 is specified. NoOptReactant retains the input structure as a point that is already on the reaction path (which generally means that it should have been previously optimized to a minimum). By default. it will be used instead to control how far the reactant side spreads out from the transition state. The remaining M-3 points on the path are generated by two successive linear interpolations. Guess=Always can be used to prevent the wavefunction of a reactant-like structure from being used as a guess for the wavefunction of a product-like structure. This is the default. NoOptProduct retains the input structure as a point that is already on the reaction path (which generally means that it should have been previously optimized to a minimum). OptProduct Specifies that the input structure for the product in a simultaneous optimization calculation should be optimized to a local minimum. the title and molecule specification sections for both reactant and product structures are required as input as usual. regardless of the ordering of the energies. this option is off. No other input is needed (see the examples). the default action is to add the specified coordinate. and +=value increments the coordinate by value. Lines in a ModRedundant input section use the following syntax: [Type] N1 [N2 [N3 [N4]]] [[+=]value] [A | F] [[min] max]] [Type] N1 [N2 [N3 [N4]]] [[+=]value] B [[min] max]] [Type] N1 [N2 [N3 [N4]]] K | R [[min] max]] [Type] N1 [N2 [N3 [N4]]] [[+=]value] D [[min] max]] [Type] N1 [N2 [N3 [N4]]] [[+=]value] H diag-elem [[min] max]] [Type] N1 [N2 [N3 [N4]]] [[+=]value] S nsteps stepsize [[min] max]] N1. This option requires a separate input section following the geometry specification. delete or modify redundant internal coordinate definitions (including scan and constraint information). See the discussion of the CASSCF keyword for details and examples. a ModRedundant input section must follow each geometry specification. Value specifies a new value for the specified coordinate. When used in conjunction with QST2 or QST3. Restart Restarts a geometry optimization from the checkpoint file.Conical Search for a conical intersection or avoided crossing using the state-averaged CASSCF method. N2. NoFreeze Activates (unfreezes) all variables (normally used with Geom=Check). Avoided is a synonym for Conical. and any dummy atoms are not counted. Note that CASSCF=SlaterDet is needed in order to locate a conical intersection between a singlet state and a triplet state. the entire route section will consist of the Opt keyword and the same options to it as specified for the original job (along with Restart). If no action code is included. N3 and N4 are atom numbers or wildcards (discussed below). In this case. The atom numbers and coordinate value are followed by a one-character code letter indicating the coordinate modification to be performed. Atom numbering begins at 1. AddRedundant is synonymous with ModRedundant. ModRedundant Add. the action code is sometimes followed by additional required parameters as indicated above. These are the available action codes: • A Activate the coordinate for optimization if it has been frozen. . D Calculate numerical second derivatives for the row and column of the initial Hessian for this coordinate. By default. K Remove the coordinate and kill all related coordinates containing this coordinate.• • • • • • • F Freeze the coordinate in the optimization. Min and max then define a range (or maximum value if min is not given) for coordinate specifications containing wildcards. value. B Bond length A Valence angle D Dihedral angle L Linear bend specified by three atoms (or if N4 is -1) or by four atoms. min and max are interpreted as the X. bond stretch for 2 atoms. . Here are some examples of wildcard use: • • • • • • • * All atoms specified by Cartesian coordinates ** All defined bonds 3* All defined bonds with atom 3 * * * All defined valence angles * 4 * All defined valence angles around atom 4 * * * * All defined dihedral angles * 3 4 * All defined dihedral angles around the bond connecting atoms 3 and 4 When the action codes K and B are used with one or two atoms. The action specified by the action code is taken only if the value of the coordinate is in the range. H Change the diagonal element for this coordinate in the initial Hessian to diag-elem. performing an optimization from each resulting starting geometry. S Perform a relaxed potential energy surface scan. the coordinate type is determined from the number of atoms specified: Cartesian coordinates for 1 atom. B Add the coordinate and build all related coordinates. not just those involving defined coordinates. where the fourth atom is used to determine the 2 orthogonal directions of the linear bend. Set the initial value of this coordinate to value (or its current value). In this case. An asterisk (*) in the place of an atom number indicates a wildcard. valence angle for 3 atoms and dihedral angle for 4 atoms. the meaning of a wildcard is extended to include all applicable atoms. and increment the coordinate by stepsize a total of nsteps times. Y and Z coordinates (respectively). Optionally. R Remove the coordinate from the definition list (but not the related coordinates). Type can be used to designate these and additional coordinate types: • • • • • X Cartesian coordinates. the mixed optimization format with all atoms specified via Cartesian lines in the Z-matrix can be used along with Opt=Z-matrix if these features are needed (see Appendix B for details and examples). value. that the variables including inactive variables are linearly independent and span the degrees of freedom allowed by the molecular symmetry). When a Z-matrix without any variables is used for the molecule specification. . CHarmonic is a synonym for this option. Cartesian Requests that the optimization be performed in Cartesian coordinates. ReadHarmonic=N Add harmonic constraints to a structure read in the input stream (in the input orientation). It also suppresses the frequency analysis at the end of optimizations which include second derivatives at every point (via the CalcAll option). O Out-of-plane bending coordinate for a center (N1) and three connected atoms. IHarmonic is a synonym for this option. the keyword FOpt rather than Opt requests that the program verify that a full optimization is being done (i.• In this case.e. COORDINATE SYSTEM SELECTION OPTIONS Redundant Perform the optimization using the Berny algorithm in redundant internal coordinates. with force constant N/1000 Hartree/Bohr2. In this case. This is the default for methods for which analytic gradients are available.and Opt=Z-matrix is specified. InitialHarmonic=N Add harmonic constraints to the initial structure with force constant N/1000 Hartree/Bohr2. specifying the two orthogonal bending components. then the optimization will actually be performed in Cartesian coordinates. Z-matrix Perform the optimization in internal coordinates.. min and max are each pairs of numbers. See the examples later in this section for illustrations of the use of this keyword. using the Berny algorithm. The POpt form requests a partial optimization in internal coordinates. Note that the initial structure may be input using any coordinate system. RHarmonic is a synonym for this option. ChkHarmonic=N Add harmonic constraints to the initial structure saved on the chk file with force constant N/1000 Hartree/Bohr2. No partial optimization or freezing of variables can be done with purely Cartesian optimizations. making previous internal coordinate force constants useless.8) The force constants are in lower triangular form-((F(J. FCCards Reads the Cartesian forces and force constants from the input stream after the molecule specifications. This has nothing to do with computation of vibrational frequencies.16) Cartesian forces (lines of format 6F12. In order to pass force constants estimated in this way to the MurtaughSargent program. where NAt3 is the number of Cartesian coordinates. These will typically be the final approximate force constants from an optimization at a lower level.ReadFC).J=1. it is necessary to do one run with Opt=StarOnly to produce the force constants.NAt3).I). This is the default. RCFC Specifies that the computed force constants in Cartesian coordinates from a frequency calculation are to be read from the checkpoint file. NewEstmFC Estimate the force constants using a valence force field. The format for this input is: • • • Energy (format D24. ReadFC Extract force constants from a checkpoint file.I). Only available for the Berny algorithm. . StarOnly Specifies that the specified force constants are to be estimated numerically but that no optimization is to be done. can be constructed using the ModRedundant option. This can be used to read force constants recovered from the Quantum Chemistry Archive using its internal FCList command. ReadCartesianFC is a synonym for RCFC.OldRedundant Use the Gaussian 94 redundant internal coordinate generator.8) Force constants (lines of format 6F12.I=1. such as distance matrix coordinates. EstmFC Estimate the force constants using the old diagonal guesses. and then run the actual optimization with Opt(MS. Note that a variety of other coordinate systems. or the force constants computed correctly by a lower-level frequency calculation (the latter are greatly preferable to the former). This is used when the definitions of variables are changed. CASSCF. Tight This option tightens the cutoffs on forces and step size that are used to determine convergence. and it is equivalent to CalcFC for DFT methods. CalcAll Specifies that the force constants are to be computed at every point using the current method (available for the HF. and semi-empirical methods only). transition state optimizations converge even if the test is not passed. and semi-empirical methods only). CalcHFFC is used with MP2 optimizations. DFT. MP2. VCD Calculate VCD intensities at each point of a Hartree-Fock Opt=CalcAll optimization. CONVERGENCE-RELATED OPTIONS These options are available for the Berny algorithm only. For DFT calculations.CalcHFFC Specifies that the analytic HF force constants are to be computed at the first point. Note that vibrational frequency analysis is automatically done at the converged structure and the results of the calculation are archived as a frequency job. VTight is a synonym for VeryTight. The test is on by default only for transition states in internal (Z?matrix) or Cartesian coordinates. DFT. for which it is recommended. An optimization with Opt=Tight will take several more steps than with the default cutoffs. NoRaman Specifies that Raman intensities are not to be calculated at each point of a Hartree-Fock Opt=CalcAll job (since it includes a frequency analysis using the results of the final point of the optimization). The Raman intensities add 10-20% to the cost of each intermediate second derivative point. but NoEigenTest is only recommended for those with large computing budgets. MP2. Occasionally. VeryTight Extremely tight optimization convergence criteria. CalcFC Specifies that the force constants be computed at the first point using the current method (available for the HF. For molecular systems with very small force constants (low frequency vibrational modes). this may be necessary to ensure adequate convergence and reliability of frequencies computed in a subsequent job step. Int=UltraFine should be specified as well. CASSCF. . EigenTest EigenTest requests and NoEigenTest suppresses testing the curvature in Berny optimizations. Int=UltraFine should be specified as well. This option can only be used with Berny optimizations. For DFT calculations. even for mechanical embedding. . CheckCoordinates Rebuild the connectivity matrix before each optimization step. and it does not work for electronic embedding. and may be appropriate for initial optimizations of large molecules using DFT methods which are intended to be followed by a full convergence optimization using the default (Fine) grid. quadratic macro step is used during ONIOM(MO:MM) geometry optimizations. The default.01 au and an RMS force of 0. NoExpert enforces the default limits and is the default. ALGORITHM-RELATED OPTIONS Micro Use microiterations in ONIOM(MO:MM) optimizations. These values are consistent with the Int(Grid=SG1) keyword. NoQuadMacro is the default. This option can lead to faster convergence but is quite dangerous.Expert Relaxes various limits on maximum and minimum force constants and step sizes enforced by the Berny program. Mic120 says to use microiterations in L120 for ONIOM(MO:MM). with selection of L120 or L103 for the microiterations depending on whether electronic embedding is on or off. It is the default. TrustUpdate TrustUpdate requests and NoTrustUpdate suppresses dynamic update of the trust radius in Berny optimizations. This is possible with mechanical embedding but not with electronic embedding. RFO Requests the Rational Function Optimization [530] step during Berny optimizations. It is not recommended for use by itself. NoMicro forbids microiterations during ONIOM(MO:MM) optimizations. Loose Sets the optimization convergence criteria to a maximum step size of 0. rebuild the redundant internal coordinate system. This is the default for electronic embedding. This option is off by default. QuadMacro Controls whether the coupled. It is used by experts in cases where the forces and force constants are very different from typical molecules and Z-matrices.0017 au. The default is to update for minima. Linear Linear requests and NoLinear suppresses the linear search in Berny optimizations. The default is to use the linear search whenever possible. It is the default for mechanical embedding. Mic103 says to perform microiterations in L103 for ONIOM(MO:MM). and sometimes in conjunction with Opt=CalcFC or Opt=CalcAll. If there is any change in it. tight optimizations and molecules with flat potential energy surfaces.536]. Bofill. This option can be turned off using Opt=Small. D2CorrBFGS. D2CMix and None. BFGS. ND2Corr. first. the defaults. then it is be scaled back. but this may not be as effective as the normal method involving eigenvector following. EF Requests an eigenvalue-following algorithm [530. . PDBFGS. CalcFC. UpdateMethod=keyword Specifies the Hessian update method. and EigenvalueFollow are all synonyms for EF. OD2Corr.535. EigFollow. This option is turned off by the RFO and Newton options. It is the default for semiempirical calculations. It may be useful when starting far from the minimum. Note that when analytic gradients are available and the lowest eigenvector is being followed. Large is a synonym for Big. with second. NoNRScale causes a minimization on the surface of the sphere of maximum step size [534]. or EnOnly. This is only compatible with Berny local minimum optimizations. Recommended for use with large systems.532. EigenFollow. QST2 and QST3 calculations are guided using an associated surface approximation. Newton Use the Newton-Raphson step rather than the RFO step during Berny optimizations. the eigenvector following methods (Opt=TS) cannot be used in conjunction with it. Steep Requests steepest descent instead of Newton-Raphson steps during Berny optimizations. Keyword is one of: Powell. Available for both minima and transition states.GDIIS Specifies the use of the modified GDIIS algorithm [531. NRScale NRScale requests that if the step size in the Newton-Raphson step in Berny optimizations exceeds the maximum. but is unlikely to reach full convergence. then the default Berny algorithm has all of the features of the eigenvalue-following algorithm. Scaling is the default for transition state optimizations and minimizing on the sphere is the default for minimizations. Big Requests the optimization to be done using the fast equation solving methods [537] for the coordinate transformations and the Newton-Raphson or RFO step. This option is default for semiempirical calculations. This method avoids the matrix diagonalizations.533]. or no analytic derivatives as indicated by CalcAll. Consequently. respectively). in format 2F10.HFError Assume that numerical errors in the energy and forces are those appropriate for HF and PSCF calculations (1.0D-06. SG1Error Assume that numerical errors in the energy and forces are those appropriate for DFT calculations using the SG-1 grid (1. Gaussian performs the optimization in redundant internal coordinates. By default. ReadError Read in the accuracy to assume for the energy and forces. Pulay has demonstrated [540. SEError is a synonym for this option. Schlegel and coworkers [149]. This is the default for optimizations using a DFT method and using the default grid (or specifying Int=FineGrid). Gaussian performs optimizations via the Berny algorithm in redundant internal coordinates. ease of use in specifying certain types of molecules). For example. This is the default for optimizations using those methods. OVERVIEW OF GEOMETRY OPTIMIZATIONS IN GAUSSIAN By default. respectively). FineGridError Assume that numerical errors in the energy and forces are those appropriate for DFT calculations using the default grid (1. since all structures are converted internally to redundant internal coordinates. irrelevant. however.6 (there is no terminating blank line for this input section since it is always a single line). and Baker reached a similar conclusion when he compared redundant internal coordinates to Cartesian coordinates [543].0D-07 and 1. quite literally.0D-07 and 1. that redundant internal coordinates are the best choice for optimizing polycyclic molecules.541.0D-07.0D-05. . There has been substantial controversy in recent years concerning the optimal coordinate system for optimizations. This is a change from previous versions of the program. Cartesian coordinates were shown to be preferable to internal coordinates (Z-matrices) for some cyclic molecules [538]. these procedures are also the work of H. This is the default for optimizations using a DFT method and Int(Grid=SG1Grid). and it has no effect on the way the optimization proceeds.0D-07 and 1.542]. This optimization procedure operates somewhat differently from those traditionally employed in electronic structure programs (including Gaussian 94 and earlier versions): • The choice of coordinate system for the starting molecular structure is. Similarly. respectively). as these values are also appropriate for semi-empirical calculations (for which it is also the default). All of the efficiency factors in the various coordinate systems are of no consequence. B. mixed internal and Cartesian coordinates were shown to have some advantages for some cases [539] (among them. is still available. but for some cases. there are several methods for providing improved force constants: • • • Use force constants from a lower-level calculation: The force constants can be read from the checkpoint file (Opt=ReadFC). Similarly.544]. via the Opt=Z-Matrix option. Extract Cartesian force constants from a checkpoint file: The Cartesian (as opposed to internal) force constants can be read from the checkpoint file. However. Note that Cartesian force constants are only available on the checkpoint file after a frequency calculation. and instead the command Opt=RCFC is used to read the Cartesian force constants and transform them to the current Z-matrix variables. such as Z-matrices with unusual arrangements of dummy atoms. In these cases. See the examples subsection for details. Normally it is preferable to pick up the force constants already converted to internal coordinates as described above. a frequency calculation occasionally reveals that a molecule needs to distort to lower symmetry. These will typically be the final approximate force constants from an optimization at a lower level or (much better) the force constants computed correctly at a lower level during a frequency calculation. In addition. some knowledge of the curvature around the saddle point is essential. This scheme usually works fine. You cannot use this option after an optimization dies . the initial guess may be so poor that the optimization fails to start off properly or spends many early steps improving the Hessian without nearing the optimized structure. WAYS OF GENERATING INITIAL FORCE CONSTANTS Unless you specify otherwise. Optimizations in internal coordinates. for optimizations to transition states (see also below).• All optimizations in redundant internal coordinates are full optimizations unless variables are explicitly frozen using the ModRedundant option. The approximate matrix is improved at each point using the computed first derivatives. In this case the computed force constants in terms of the old Z-matrix variables cannot be used. a Berny geometry optimization starts with an initial guess for the second derivative matrix-also known as the Hessian-which is determined using connectivity determined from atomic radii and a simple valence force field [149. Optimizations in redundant internal coordinates do make use of geometry constraint information and numerical differentiation specifications. the requirement that all variables in the Z-matrix be linearly independent does not apply to these optimizations. Usually this means that a new Z-matrix with fewer symmetry constraints must be specified to optimize to the lower energy structure. Including a separate constant variable section in the molecule specification does not result in any frozen variables. which was the default procedure in Gaussian 92. and the default approximate Hessian must always be improved. 0 Z-matrix R1 1.55 • • • The first line specifies that the angle formed by atoms 1. In this case each specified variable will be stepped in only one direction.5 A1 104.5 D 2 3 4 110. This is only possible for HF. they are updated at each point using the gradient information available from the points done in the optimization. and the second line sets the initial value of the bond between atoms 1 and 2 (the variable R in the Z-matrix) to 0.• • • • because of a wrong number of negative eigenvalues in the approximate second derivative matrix. or semi-empirical optimization. Needless to say. Note that this option is equivalent to CalcFC for DFT methods. This procedure is requested by a flag (D) on the variable definition lines: Redundant Internals 1 2 1. MP2.5 D A2 110. MP2. Calculate new force constants at every point: Normally after the initial force constants have been decided upon.0 H 0. this requires that the program do an extra gradient calculation for each specified variable.5.55 hartree/au2. this is very expensive. you can specify Opt=CalcAll. Calculate initial force constants at the current level of theory: You can request that the second derivatives of the method being used in the optimization be computed at the first point by specifying Opt=CalcFC. DFT. you may want to start from the most recent geometry and compute some derivatives numerically. DFT or post-SCF gradient optimizations. Note that the units here are Hartrees and Bohrs or radians. Of course. Input new guesses: The default approximate matrix can be used. 2 and 3 (the variable A in the Z-matrix) is to start at the value 104. This is specified in the ModRedundant input or on the variable definition lines in the Zmatrix. For a Hartree-Fock.0 D 2 3 1. and semi-empirical methods.0 R2 1.55 Z-matrix A 104. This is done by specifying Opt=CalcHFFC. Calculate initial force constants at the HF level: You can also request that the analytic Hartree-Fock second derivatives be calculated at the first point of the optimization.5 R 1.0 . not both up and down as would be required for an accurate determination of force constants.55 Angstroms. which requests that second derivatives be computed at every point in the optimization. In that case. This option is valid only with the Berny algorithm. but with new guesses read in for some or all of the diagonal elements of the Hessian. Compute some or all of the Hessian numerically: You can ask the optimization program to compute part of the second derivative matrix numerically. For example: Redundant Internals 1 2 3 104. The resulting second-derivatives are not as good as those determined by a frequency calculation but are fine for starting an optimization. This can be used with HF.5 1 2 3 104.5 1 2 1. The letter H on the second line indicates that a diagonal force constant is being specified for this coordinate and that its value is 0.0 H 0. as its input. uses a quadratic synchronous transit approach to get closer to the quadratic region of the transition state and then uses a quasi-Newton or eigenvector-following algorithm to complete the optimization.3)/A1 from the three points. This method. In contrast. a geometry with the bond between atoms 1 and 2 (R1) incremented slightly. and it then goes on to optimize that starting structure to a first-order saddle point automatically. implemented by H. The order of the atoms must be identical within all molecule specifications. Despite the superficial similarity. This option is only available with the Berny and EF algorithms. while QST3 requires three molecule specifications: the reactants. Traditional Transition State Optimizations Using the Berny Algorithm. The options to request this procedure are Opt=TS for a transition state (saddle point of order 1) or Opt(Saddle=N) for a saddle point which is a maximum in N directions. if it is coaxed along properly. QST2 and QST3 do locate proper transition states. Gaussian includes the STQN method for locating transition structures. QST2 requires two molecule specifications. it performs optimizations by default in redundant internal coordinates. and an initial structure for the transition state. Like the default algorithm for minimizations. This method will converge efficiently when provided with an empirical estimate of the Hessian and suitable starting structures. The Linear Synchronous Transit method merely locates a maximum along a path connecting two structures which may be used as a starting structure for a subsequent manually-initiated transition state optimization.• This input tells the program to do three points before taking the first optimization step: the usual first point. the Berny algorithm uses a combination of rational function optimization (RFO) and linear search steps to achieve speed and reliability (as described below). in that order. Schlegel and coworkers [149. The program will use the default diagonal force constants for the other two coordinates and will estimate all force constants (on and off diagonal) for bond(1. B.2. When searching for a local minimum. This method is requested with the QST2 and QST3 options. for the reactants and products. Opt=QST2 generates a guess for the transition structure that is midway between the reactants and products in terms of redundant internal coordinates. 2 and 3 (A1) incremented slightly.2)/R1 and angle(1.150]. See the examples for sample input for and output from this method. The Berny optimization program can also optimize to a saddle point using internal coordinates. LST does not locate a proper stationary point. this method is very different from the Linear Synchronous Transit method for locating transition structures requested with the nowdeprecated LST keyword. OPTIMIZING TO A TRANSITION STATE OR HIGHER-ORDER SADDLE POINT Transition State Optimizations Using Synchronous Transit-Guided Quasi-Newton (STQN) Methods. the products. and a geometry with the angle between atoms 1. This linear search step cannot be applied when searching for a . starting with Opt=CalcFC) and steps into a region of undesirable curvature.5 4 • The Nth mode in order of increasing Hessian eigenvalue can be requested by placing a 10 after the Nth variable definition line. If the number is not correct (1 for a transition state). transition state optimizations are much more sensitive to the curvature of the surface. A .. A transition state optimization should always be started using one of the options described above for specifying curvature information. For suggestions on locating transition structures. or gradients are not available (in which case Berny can't be used anyway). as in this input file: # Opt=(EF. An eigenvalue-following (mode walking) optimization method [146.147] can be requested by Opt=EF. refer to the literature [148]. EF is seldom preferable unless its ability to follow a particular mode is needed. the Berny optimization program checks the curvature (number of negative eigenvalues) of its approximate second derivative matrix at each step of a transition state optimization.1.CN H. the Opt=CalcAll option may be useful. HCN 60. the lowest mode is followed.1 N C.0 By default.TS) HCN --> HNC transition state search This job deliberately follows the wrong (second) mode! 0.20 10 Requests the second mode. so it is best to try to make the reaction coordinate (direction of negative curvature) correspond to one or two redundant internal coordinates or Z-matrix variables (see the examples below). This default can be overridden in two ways: • The mode having the largest magnitude component for a specific Z-matrix variable can be requested by placing a 4 on the variable definition line: Ang1 104.g.CH. This is correct when already in a region of correct curvature and when the softest mode is to be followed uphill.2. In the extreme case in which the optimization begins in a region known to have the correct curvature (e. Consequently. but since the RFO step [530] has now been incorporated into the Berny algorithm.HCN CN 1. This is quite expensive.1. This was sometimes superior to the Berny method in Gaussian 88. Without a full second derivative matrix the initial step is dependent on the choice of coordinate system. This algorithm has a dimensioning limit of 50 active variables. the job is aborted.transition state.3 CH 1. By default. but the full optimization procedure with correct second derivatives at every point will usually reach a stationary point of correct curvature if started in the desired region. and using a modification of the original Schlegel update procedure for optimizations in internal coordinates. and consequently it is appropriate to summarize the current status of the Berny algorithm here. Schlegel which implemented his published algorithm [136]. At each step of a Berny optimization the following actions are taken: • • • • • The Hessian is updated unless an analytic Hessian has been computed or it is the first step. a simple cubic is fit is done Any quintic or quartic step is considered acceptable if the latest point is the best so far but if the newest point is not the best. if it does not have a minimum in the acceptable range (see below) or if second derivatives are not available. Finally. Any components of the gradient vector corresponding to frozen variables are set to zero or projected out. If this fit fails or if the resulting step is unacceptable. but upon request either a unit matrix or a diagonal Hessian can also be generated as estimates. a constrained quartic fit is attempted. the linear search must return a point in between the most recent and the best step to be acceptable. although it may also indicate that the optimization has moved away from the desired minimum and is headed through a transition state and on to a different minimum. thereby ensuring that the polynomial itself has exactly one minimum. no . The trust radius (maximum allowed Newton-Raphson step) is updated if a minimum is sought. If a minimum is sought. a transition state optimization has less chance of success if the curvature is wrong at the current point.547]. in which case an estimate of the Hessian is made. Normally the update is done using an iterated BFGS for minima and an iterated Bofill for transition states in redundant internal coordinates. thereby eliminating their direct contribution to the next optimization step. This fits a quartic polynomial to the energy and first derivative (along the connecting line) at the two points with the constraint that the second derivative of the polynomial just reach zero at its minimum. using the method of Fletcher [545. B. perform a linear search between the latest point and the best previous point (the previous point having lowest energy).search for a minimum will often succeed in spite of bad real or approximate curvature.By default. The program has been considerably enhanced since this earlier version using techniques either taken from other algorithms or never published. On the other hand. Cubic steps are never accepted unless they are in between the two points or no larger than the previous step. If second derivatives are available at both points and a minimum is sought. because the steepest descent and RFO parts of the algorithm will keep the optimization moving downward. if all fits fail and the most recent step is the best so far. If NoEigenTest is used. it is best to MaxCycle to a small value (e.g. a quintic polynomial fit is attempted first. However. THE BERNY OPTIMIZATION ALGORITHM The Berny geometry optimization algorithm in Gaussian is based on an earlier program written by H.546. the test can be suppressed with the NoEigenTest option. 5) and check the structure after a few iterations. this is derived from a valence force field [544]. The RFO step behaves better than the Newton-Raphson method used in earlier versions of Gaussian when the curvature at the current point is not that desired. The old Newton-Raphson step is available as an option. maximum step component.3 of the closest pair have their distances added to the internal coordinates. If the quadratic step exceeds the trust radius and a minimum is sought. The step is the change between the most recent point and the next to be computed (the sum of the linear and quadratic steps). the quadratic step is taken from the point extrapolated using the linear search and uses forces at that point estimated by interpolating between the forces at the two points used in the linear search. the linear step is taken to the midpoint of the line connecting the most recent and the best previous points.536]. Finally. then the geometry is considered converged even if the displacement is larger than the cutoff value. As usual. . Any components of the step vector resulting from the quadratic step corresponding to frozen variables are set to zero or projected out. When the forces are two orders of magnitude smaller than the cutoff value (i.. In addition. CHANGE IN TRADITIONAL CONVERGENCE CRITERIA BEGINNING WITH GAUSSIAN 98 Gaussian 98 introduced one small but significant change in the criteria for determining when a geometry has converged. If a transition state is sought or if NRScale was requested. By default. then the related angles and dihedrals are added in order to ensure a complete coordinate system. the ModRedundant option can be used to add or remove any coordinates manually. a quadratic step is determined using the current (possibly approximate) second derivatives. then no angles or dihedrals involving both fragments are added. If at least 3 such pairs are found. convergence is tested against criteria for the maximum force component. 1/100th of the limiting value).147. as discussed by Jorgensen [534]. This test was introduced to facilitate optimizations of large molecules which may have a very flat potential energy surface around the minimum. root-mean square force. if only 1 or two pairs of atoms are close. this step uses the Rational Function Optimization (RFO) approach [146.e. We include Hydrogen bonds automatically. If the latest point is the best so far or if a transition state is sought. all pairs of atoms with one atom in each fragment having distance within a factor of 1. the quadratic step is simply scaled down to the trust radius. The generation of redundant internal coordinates for weakly bound complexes was also updated with Gaussian 98.• • • • linear step is taken. in connecting different fragments which are only weakly bound (hydrogen-bonded and otherwise). the step is reduced in length to the trust radius by searching for a minimum of the quadratic function on the sphere having the trust radius.530. If a linear search was done. and root-mean-square step. However. If all fits fail and the most recent step is not the best. CASSCF. For example. examples illustrating traditional.0 A=104. Basic Optimization Input. Scan. all DFT methods. In early versions of Gaussian. IRC. MP4(SDQ). Since the OH bond distance is specified using the same variable for both hydrogen atoms. .00 0. Force The examples in the subsection will focus on normal optimization procedures in Gaussian 03.97 0. and this holds for Gaussian 03 as well. MP3. CIS. Zmatrix-based optimizations using the Berny algorithm will also be given. CISD. the input file in the left column below could be used for such an optimization on water: # HF/6-31G(d) Opt Test Water opt 0 1 O1 H1 O1 R H2 O1 R H1 A Variables: R=1. Eigentest and EstmFC options are available for the Berny algorithm only. CCSD. MP2.00 1. at the end of the subsection.25 This Z-matrix specifies the starting configuration of the nuclei in the water molecule. They both will result in a Berny optimization in redundant internal coordinates. and all semi-empirical methods. giving identical final output. The Cartesian coordinate input in the right column is equivalent to the Z-matrix in the left column. Z-matrix input could be used for optimizations in either coordinate system. By contrast. The Tight. such input would lead to an optimization performed in Cartesian coordinates. Traditionally. by Gaussian 92.5 # HF/6-31G(d) Opt Test Water opt 0 O H H 1 0. CID. However. VeryTight. beginning with Gaussian 98 these two input files are exactly equivalent.00 0.00 0.00 0. geometry optimizations required a Z-matrix specifying both the starting geometry and the variables to be optimized. however. Expert.00 0. this Z-matrix also imposes (appropriate) symmetry constraints on the molecule.00 -0.Analytic gradients are available for the HF. QCISD. CCD. It also specifies that the optimization should determine the values of R and A which minimize the energy. 1.1). initialization pass. estimate D2E/DX2 ! ! A1 A(2. For example. all coordinates in use are displayed in the table (not merely those present in the molecule specification section): GradGradGradGradGradGradGradGradGradGradGradGradGradGradGradGradGrad Berny optimization.0001 ! ----------------------------------------------------------------------- The redundant internal coordinate definitions are given in the second column of the table. Each subsequent step of the optimization is delimited by lines like these: GradGradGradGradGradGradGradGradGradGradGradGradGradGradGradGradGrad Berny optimization. ---------------------------! Optimized Parameters ! ! (Angstroms and Degrees) ! --------------------------------------! Name Definition Value Derivative Info.9892 -DE/DX = 0. delimits the output from the Berny optimization procedures. ! ----------------------------------------------------------------------! R1 R(2. defined as R(2.1.5 estimate D2E/DX2 ! -------------------------------------------------------------------- The manner in which the initial second derivative are provided is indicated under the heading Derivative Info. the final structure is displayed: Optimization completed. ---------------------------! Initial Parameters ! ! (Angstroms and Degrees) ! ------------------------------------------! Name Definition Value Derivative Info.3) 104. The opt.. For optimizations in redundant internal coordinates.1) 0.0002 ! ! A1 A(2. -. algorithm is identified by the header format & this line. the variable R1.3) 100. ! ----------------------------------------------------------------------! R1 R(2. In this case the second derivatives will be estimated.Stationary point found. estimate D2E/DX2 ! ! R2 R(3.9892 -DE/DX = 0. On the first.004 -DE/DX = 0..0002 ! ! R2 R(3.Output from Optimization Jobs. Step number 4 out of a maximum of 20 Once the optimization completes. The string GradGradGrad. Search for a local minimum. . the program prints a table giving the initial values of the variables to be optimized.1) 1. The numbers in parentheses refer to the atoms within the molecule specification. specifies the bond length between atoms 1 and 2.1) 1. Initialization pass.1) 0. 000 -0. The energy for the optimized structure will be found in the output from the final optimization step.295 O7 1. it may be included if setting any of its options is desired.755 -1. It is also common to follow an optimization with a single point energy calculation at a higher level of theory.651 -0.000 O3 1. the Opt keyword may be combined with Freq in the route section of an input file. Compound Jobs. The following route section automatically performs an HF/631G(d.242 H8 1.505 -0.000 C2 0.898 1.p)//HF/6-31G(d.006 0.047 H4 -1. . and this combination will automatically generate a two-step job.000 0.484 1.1 C1 0.001 0. The following input file illustrates the method for specifying redundant internal coordinates within an input file: # HF/6-31G(d) Opt=ModRedun Test Opt job 0.000 0. More detailed information about the out put from geometry optimizations is provided in Chap. whenever possible.000 1. Note that this line adds only the bond between these two atoms. 3 of Exploring Chemistry with Electronic Structure Methods. The associated angles and dihedral angles would need to be added as well if they were desired.065 1.024 0. However. To facilitate this procedure.866 2.499 This structure is acetaldehyde with an OH substituted for one of the hydrogens in the methyl group.p) single point energy calculation # MP4/6-31G(d.364 -0.000 H5 -0. which precedes this table in the output file.938 3 2 8 1 3 0. Specifying Redundant Internal Coordinates. this output will be followed by an expression of the optimized structure in that format.p) optimization followed by an MP4/6-31G(d.When a Z-matrix was used for the initial molecule specification.735 H6 -0. the first input line for ModRedundant creates a hydrogen bond between that hydrogen atom and the oxygen atom in the carbonyl group. Optimizations are commonly followed by frequency calculations at the optimized structure.p) Test Note that the Opt keyword is not required in this case. 1 * * * K Define all bonds between atoms within 1. A distance matrix coordinate system can be activated using the following input: * * B * * * K Define all bonds between pairs of atoms Remove all other redundant internal coordinates The following input defines partial distance matrix coordinates to connect only the closest layers of atoms: * * B 1. Note that the lines containing the B action code will generate Cartesian coordinates for all of the coordinates involving the specified atom since only one atom number is specified: N1 B . The second input line for ModRedundant specifies the C-C=O bond angle.0 .0 Add 10.0 to the values to dihedrals involving N2-N3 bond Additional examples are found in the section on relaxed PES scans below. The following job illustrates the method for freezing variables during a redundant internal coordinate optimization: # HF/6-31G* Opt=ModRedundant Test Partial optimization 1 1 C H 1 R1 H 1 R1 2 A1 O 1 R2 2 A2 3 120. ensuring that its value will be displayed in the summary structure table for each optimization step..Displaying the Value of a Desired Coordinate. Performing Partial Optimizations. Nn B * F Generate Cartesian coordinates involving atom N1 Generate Cartesian coordinates involving atom Nn Freeze all Cartesian coordinates The following input defines special "spherical" internal coordinate appropriate for molecules like C60 [548] by removing all dihedral angles from the redundant internal coordinates: * * * * R Remove all dihedral angles The following input rotates the group about the N2-N3 bond by 10 degrees: * N2 N3 * +=10. Using Wildcards in Redundant Internal Coordinates.1 Å Remove all other redundant internal coordinates The following input sets up an optimization in redundant internal coordinates in which atoms N1 through Nn are frozen (such jobs may require the NoSymm keyword).. in this case the value of second modified redundant internal coordinate defaults to the value from the Z-matrix (180.0 -. This input fixes the O-H bond and the dihedral angle for the final hydrogen atom.) Once you have the structure in Cartesian coordinates.3 1.0 R=1.0 H6 1.1 C7 1.0 -.3 F F The structure is specified as a traditional Z-matrix.. Note that any value specified in this manner need not be the same as the one listed in the preceding Z-matrix (as is the case for the O-H bond length). and then adding three additional hydrogen atoms bonded to that carbon.2 2. Use the Cartesian coordinates version of the optimized structure as your starting point.9 H8 0.9 0.0 H10 C7 R H6 A C2 180. Additional atoms may be specified in either Cartesian or internal coordinates.1 4 5 5 4 3 2 1.0 .2 0. with its variables defined in a separate section. The latter could be given in internal coordinates: H6 1. The final input section gives the values for the ModRedundant option..0 H11 C7 R H8 A C2 -180.0 -. Modifying Optimized Structures (Why You Don't Need a Z-matrix). you can use it in a variety of ways: • • Add and/or remove atoms from it.0). For example. Modify it by substituting atoms or groups: For example. then you can use Geom=ModRedundant rather than this approach.0 -. If all you want to do is change the value or activate/frozen status of one or more variables.2 2.0 A=120.9 0.1 H7 1. .0 A1=120.2 0.3 1. R3=1. It can be generated by a route like this one: # Guess=Only Geom=Check (It can also be extracted from an archive entry. the structure is adjusted to enforce this constraint.5 The new structure on the right also uses an additional redundant internal coordinate (specifying Opt=ModRedundant on the final job) to alter the bond distance for the new carbon atom which is replacing the hydrogen (bonded to atom 2). you could change a hydrogen to a methyl group by editing the structure. replacing the desired hydrogen with a carbon atoms.0 H9 C7 R H5 A C2 180.9 H8 0.0 7 2 1. The constrained value is optional.H 4 R3 3 A3 2 180. The QST3 option allows you to specify a better initial structure for the transition state.Restart. this route section restarts a Berny optimization to a second-order saddle point: # RHF/6-31G(d) Opt=(Saddle=2.MaxCyc=50) Test Reading a Structure from the Checkpoint File. The QST2 option initiates a search for a transition structure connecting specific reactants and products. The input for this option has this general structure: # HF/6-31G(d) Opt=QST2 (Opt=QST2.ModRedun) First title section Molecule specification for the reactants for the reactants Second title section the reactants Molecule specification for the products for the products the products (optional) # HF/6-31G(d) First title section Molecule specification ModRedundant input for Second title section Molecule specification ModRedundant input for Note that each molecule specification is preceded by its own title section (and separating blank line). third title and molecule specification sections for the initial transition state geometry (along with the usual blank line separators). The read-in structure may be altered by specifying Geom=ModRedundant as well. Gaussian will automatically generate a starting structure for the transition structure midway between the reactant and product structures. If the ModRedundant option is specified. modifications have a form identical to the input for Opt=ModRedundant: [Type] N1 [N2 [N3 [N4]]] [[+=]Value] [Action [Params]] [[Min] Max]] Locating a Transition Structure with the STQN Method. Redundant internal coordinate structures may be retrieved from the checkpoint file with Geom=Checkpoint as usual. It requires the two title and molecule specification sections for the reactants and products as for QST2 and also additional. A failed optimization may be restarted from its checkpoint file by simply repeating the route section of the original job. For example. adding the Restart option to the Opt keyword. then each molecule specification is followed by any desired modifications to the redundant internal coordinates. as well as three corresponding modifications to the redundant internal coordinates if the ModRedundant . and then perform an optimization to a first-order saddle point.Restarting an Optimization. 4154 1. ! ! R2 R(3.3) 1. ! -------------------------------------------------------------------- In addition to listing the optimized values. this input adds a bond between atoms 2 and 3.4233 1. the table includes those for the reactants and products..3984 -DE/DX = 0. ! ! . ! ! R4 R(5. For example. Note that all other dihedrals around the bond should be removed: * N2 N3 * R N1 N2 N3 N4 S 20 2.05 Wildcards in the ModRedundant input may also be useful in setting up relaxed PES scans. Relaxed PES scans are available only for the Berny algorithm. If any scanning variable breaks symmetry during the calculation. The program will then locate the transition structure connecting the reactants and products closest to the specified initial geometry. a relaxed PES scan steps over a rectangular grid on the PES involving selected internal coordinates.4426 -DE/DX = -0.3989 1. Performing a Relaxed Potential Energy Surface Scan.083 1. It differs from the operation of the Scan keyword in that a constrained geometry optimization is performed at each point. The optimized structure found by QST2 or QST3 appears in the output in a format similar to that for other types of geometry optimizations: ---------------------------! Optimized Parameters ! ! (Angstroms and Degrees) ! ----------------------------------------! Name Definition Value Reactant Product Derivative Info.0985 1. Like the scan facility provided by previous versions of Gaussian.084 -DE/DX = 0. setting its initial value to 1. Redundant internal coordinates specified with the Opt=ModRedundant option may be scanned using the S code letter: N1 N2 [N3 [N4]] [[+=]value] S steps step-size.3989 1.3952 -DE/DX = -0.05 Å each: 2 3 1. and specifying three scan steps of 0. ! ! R3 R(4.option is specified. ! -------------------------------------------------------------------! R1 R(2.1) 1.1) 1.0 S 3 0. The Opt=Z-matrix and Opt=ModRedundant keywords may also be used to perform a relaxed potential energy surface (PES) scan.0 Remove all dihedrals involving the N2-N3 bond Specify a relaxed PES scan of 20 steps in 2º increments .4347 1.3) 1.1) 1.1009 1.0836 1. then you must include NoSymm in the route section of the job. ! ! R5 R(6.0 Å.0995 -DE/DX = 0. the following input is appropriate for a potential energy surface scan involving the N1-N2-N3-N4 dihedral angle. For example..4047 1. or it will fail with an error. 9 A=105. the Berny optimization algorithm makes a distinction between full and partial optimizations.1 A=105.4 By contrast. The one on the left has been constrained to C2v symmetry. Note that an optimization in redundant internal coordinates which begins from a C2v structure will retain that symmetry throughout the optimization. separated by the usual variables section by a blank line or a line containing a space in the first column and the string Constants:. the Z-matrix on the right is unconstrained since the two bond lengths are specified by different variables having different initial values. Note that the FOpt keyword form is used to request that the optimization variables be tested for linear independence prior to beginning the optimization.4 O H 1 R1 H 2 R2 2 A R1=0.4 Breaking Symmetry During an Optimization in Internal Coordinates.Full vs.9 R2=1. Those variables whose values should be held fixed are specified in a separate input section.4 degrees throughout the optimization: # HF/6-31G(d) Opt=Z-matrix Test Partial optimization for water 0 1 O H1 O R H2 O R H1 A Variables: R 1. Partial Optimizations. When it is performed in internal (Z-matrix) coordinates. but not the angle A. their values will always be the same: O H 1 R1 H 1 R1 2 A R1=0. which will be held fixed at 105. while partial optimizations optimize only a specified subset of the variables. For Opt=Z-matrix. For example. Below are two geometry specifications for water. Relaxed PES Scans. since the same variable is used for both bond lengths.0 Constants: A 105. a relaxed PES scan is requested simply by tagging the Z-matrix variables whose values are to be incremented with the S code letter . the following input file will optimize only the bond distance R. Full optimizations optimize all specified variables in order to find the lowest energy structure. 277.and the number of steps and the increment size.9 S 5 0. PSI is a synonym for WFN.275.4 S 2 1. on a separate line.274.273. WFN Write a PROAIMS wavefunction (.jpl.278]: • • • • • • Nuclear electric quadrupole constants: all jobs Rotational constants: Freq=(VibRot[. by 0.279] to the output file in the form of input for Pickett's program [280] (see spec. and the variable A2 to be incremented twice. The name for the created file is read from the input stream. For example.0 . This causes the variable R1 to be incremented five times.gov)..05 R2 1. Output The Output keyword requests output of Fortran unformatted files.05 Å each time.Anharmonic]) Quartic centrifugal distortion terms: Freq=(Anharmonic) Electronic spin rotation terms: NMR Nuclear spin rotation terms: NMR Dipolar hyperfine terms: all jobs ..212. the following input file requests a relaxed PES scan for the given molecule: # HF/6-31G(d) Opt=Z-matrix Test Relaxed PES scan 0 1 O H 1 R1 C 1 R2 2 A2 .213.wfn) file. Pickett Write g tensors and other tensors for hyperfine spectra [272..214. Variables: R1 0.276. resulting in a total of 18 geometry optimizations (the initial values for each variable also constitute a point within the scan).. Its options control the contents of the created file.nasa. by 1 degree each time.1 A2 115. The following tensors can be computed by Gaussian 03 [207. or with Tran=IJAB to save on disk space at the expense of CPU time. only ionization potentials which are < 20 eV are computed. By default. By default. For unrestricted calculations. separate ranges are specified for alpha and beta orbitals (on the same input line). but can be run Tran=Full to save CPU time at the expense of disk usage.249. eight "interesting" atoms are selected automatically by the program. all orbitals are used. in a separate.244.247. Single point energy calculations only.549]. In the latter case. By default. FC The frozen-core options for defining inner-shells to be excluded from the correlation calculation are valid with this keyword.245.246. OVGF calculations default to storage of <ia||bc> integrals. blank-terminated input section. ReadOrbitals Specify starting and ending orbitals to refine. Use ReadOrbitals option to specify the starting and ending orbitals to refine as input. and this input section is blank-terminated.248. electron affinities are not computed. . See the discussion here for details. Punch OVGF These method keywords request an Outer Valence Green's Function (propagator) calculation of correlated electron affinities and ionization potentials [243. Atoms numbers are specified in free format.• Fermi contact terms: all jobs ReadAtoms Read a list of the atoms to include in the input for Pickett's program (note that this program only accepts tensors for eight nuclei). NKPoint=N Do approximately N k-points.840 (PS) Converged 3rd order P3 pole: -0.854 (PS) The second output line gives the estimate of ionization potential/electron affinity for the specified orbital (which property is given depends on whether the orbital is occupied or not. and this keyword is used only to control how PBC calculations are performed. Note that orbitals are listed in the output in order of symmetry (and not necessarily in numerical order). PBC This keyword allows you to specify options for Periodic Boundary Conditions jobs.63722D+00 au -17. you do not have to include the keyword to perform a PBC calculation. NCellXC is a synonym for this option.For OVGF calculations. NCellMax=N Include at most N cells in any part of the calculation. respectively) .0 as the maximum value. GammaOnly Do just the Γ point (k=0) rather than full k-integration.340 eV 0. The pole strength is a measure of how easy it is to make this excitation. . NCellMin=N Include at least N cells. the results for each orbital appear as follows: Summary of results for alpha spin-orbital 6 P3: Koopmans theorem: -0.598 eV Converged second order pole: -0.61437D+00 au -16. If you do not need any of these options.718 eV 0. CellRange=N Go out N Bohr in each direction in setting up image cells. with 1. Note PBC is turned on simply by including translation vectors in the input structure.72022D+00 au -19. NCellDFT=N Include at least N cells in DFT XC quadrature. PM3 PM3MM The method keywords request a semi-empirical calculation using the PM3 Hamiltonian [55. if exact exchange is included. Note that energy differences computed from the values in semi-empirical calculations are in Hartrees and may be compared directly with energy differences computed from jobs using other methods. then this is twice the number of cells used for overlap-related quantities and XC quadrature. PM3MM specifies the PM3 model including the optional molecular mechanics correction for HCON linkages. By default.56].000000 . but this keyword may be combined in the same job with numerical . The parameter for Li has been updated as specified in [402].080731473251 NIter= 10. No basis set keyword should be specified with either of these keywords. Polar This method keyword requests that the dipole electric field polarizabilities (and hyperpolarizabilities. if possible) be computed. Dipole moment= . The PM3 energy appears in the output file as follows (followed by the x. Energies. y.NCellK=N Include at least N cells in exact exchange. No geometry change or derivatives are implied. "analytic" gradients.739540 The energy is as defined by the PM3 semi-empirical model. and z components of the dipole moment): Energy= -.000000 -. See the "Specifying Periodic Systems" subsection of the "Overview of Molecule Specifications" section. and numerical frequencies. The default for methods for which only analytic first derivatives are available. to produce hyperpolarizabilities. CCSD(T). Step=N Specifies the step size in the electric field to be 0. and so on). since analytic first derivatives will also be differentiated twice.552. This is possible for RHF and UHF and MP2 for which it is the default. Restart Restarts a numerical polarizability calculation from the checkpoint file.264. of course.269.0001N atomic units.268.305.553] may also be predicted via the OptRot option [223.271. However.266. QCISD(T).263. Cubic Numerically differentiate analytic polarizabilities to produce hyperpolarizabilities. BD. EnergyOnly. Note that Polar is done by default when second derivatives are computed analytically.270.221. a synonym for EnOnly. Freq and Polar may not be combined for methods lacking analytic gradients (MP4(SDTQ).262.267. if possible.224. Normally. not the expectation value in the case of MP2 or CI energies). frequency-dependent polarizabilities and hyperpolarizabilities [220. EnOnly Requests double numerical differentiation of energies to produce polarizabilities. Analytic Analytically compute the polarizability and.554]. Optical rotations [261. Numerical Computes the polarizability as a numerical derivative of the dipole moment (itself the analytic derivative of the energy. The polarizability is always computed during analytic frequency calculations.222. OptRot Perform optical rotation calculation. A failed Polar . DCSHG Do extra frequency-dependent CPHF for dc-SHG (direct current second harmonic generation) hyperpolarizabilities. polarizabilities and hyperpolarizabilities are computed using static frequencies.differentiation of forces by specifying both Freq and Polar in the route section.265.225] may be computed by including CPHF=RdFreq in the route section and specifying the desired frequency in the input file. the hyperpolarizability. is a misnomer.551. This option implies CPHF=RdFreq as well. when they are available.550. 100000: 1 2 3 1 0.112001D+02 3 0. CASSCF and DFT methods. here are the polarizability values for a frequency-dependent job (ω=0. MP3. Dipole Compute the dipole polarizabilities (this is the default). CCD. making EnOnly the default). CID. For example. Freq CPHF=RdFreq Frequency-Dependent Properties.000000: 1 2 3 1 0. Note that Polar is not available for any semi-empirical method. QCISD. Polarizabilities and hyperpolarizabilities will be automatically computed for HF and MP2. CCSD.calculation may be restarted from its checkpoint file by simply repeating the route section of the original job.491893D+01 2 0.165696D+02 Isotropic polarizability for W= 0.1 Molecule specification 0. adding the Restart option to the Polar keyword.482729D+01 2 0.1 Performing a frequency-dependent Polar calculation results in the results for the specified frequency following those for the static case within the output. The following job will frequency-dependent polarizabilities and hyperpolarizabilities using ω=0. and Polar=EnOnly will produce both polarizabilities and hyperpolarizabilities for CIS. MP2. No other input is required.000000D+00 0.115663D+02 .1 Hartree): SCF Polarizabilityfor W= 0. Polar will produce only polarizabilities for all other methods (for which no analytic derivatives are available.000000D+00 0. Polar will compute polarizabilities only.000000D+00 0.87 Bohr**3 SCF Polarizability for W= 0.000000 10.000000D+00 0. CISD.1 Hartrees: # Polar CPHF=RdFreq HF/6-31G(d) Frequency-dependent calculation: w=0. MP4(SDQ). In this case. Optical Rotation Beta= 2.0).2384 au. to avoid repeating any costly calculations. for which the default is Pop=Full (see below). Molar Mass = 74. G' tensor for W= 0. except for Guess=Only jobs. The default is to print just the total atomic charges and orbital energies.u. Optical Rotations. The static results are listed first in the output (ω=0. Similar output follows for hyperpolarizabilities and additional properties.74920313 9.6569 au.. The density that is used for the population analysis is controlled by the Density keyword.62301919 4.[Alpha]D = 643.4103 grams/mole.000000D+00 0.22 Bohr**3.52918305 10.u. Optical Rotation Beta= 1.10 deg. [Alpha] ( 5000. Molar Mass = 74.30 deg.000000 a. A static polarizability calculation would include only the first section.48555729 -7. we have performed a frequency-dependant calculation by including CPHF=RdFreq in the route section and specified a frequency of 500 nm: w= 0.3 0. Note that only one density and method of charge fitting can be used in a job step. If several combinations are of interest. The specific rotation value is highlighted in the output.64293589 28. Populations are done once for single-point calculations and at the first and last points of geometry optimizations. additional jobs steps can be added which specify Guess=Only Density=Check. Here is the key part of the output for optical rotations jobs (OptRot option).4103 grams/mole.000000D+00 0. Population analysis results are given in the standard orientation.100000 11..26760578 w= 0.091127: -27.171826D+02 Isotropic polarizability for W= 0.88112715 8. followed by those for the specified frequency.27183975 58. Population This properties keyword controls printing of molecular orbitals and several types of population analysis and atomic charge assignments.091127 a. Output controlled by the Pop keyword includes: • • • Molecular orbitals and orbital energies Atomic charge distribution Multipole moments: dipole through hexadecapole .50024234 -14.0 A) = 1917. NOAB Do separate natural orbital analyses for the α and β densities. AlphaNatural Do separate natural orbital analyses for the α and β densities. along with the density matrices and a full (orbital by orbital and atom by atom) Mulliken population analysis. Regular The five highest occupied and five lowest virtual orbitals are printed. and no population analysis is done. it can become quite substantial for larger molecules. The other options control how much is printed. but some other semi-empirical programs interpret these coefficients as referring to raw atomic orbitals. Gaussian prints molecular orbitals and performs population analyses regarding the MO coefficients from a semi-empirical calculation as coefficients of orthogonalized atomic orbitals (OAO's). Full Same as the Regular population analysis. This is a Mulliken population analysis in which only density terms involving pairs of basis functions on different centers are retained. There are important theoretical reasons for preferring this interpretation. Minimal Total atomic charges and orbital energies are printed. NO is a synonym for NaturalOrbitals. Since the size of the output depends on the square of the size of the molecule.wfn file (see Output=WFN). None No orbitals are printed. This is the default for all job types except Guess=Only. NaturalSpinOrbitals is a synonym for NOAB. NATURAL ORBITAL-RELATED OPTIONS NaturalOrbitals Do a natural orbital analysis of the total density. NOA is a synonym for AlphaNatural. . but store only the α densities for use in a . except that all orbitals are printed.By default. Use IOp(4/24=3) to compare orbitals from semi-empirical calculations to the results of such other programs. BONDING ANALYSIS OPTION Bonding Do a bonding population analysis in addition to the standard analysis. By default. terminated by a blank line. CHelp Produce charges fit to the electrostatic potential at points selected according to the CHelp scheme [218]. NOB is a synonym for BetaNatural. ESP and MerzKollman are synonyms for MK. SpinNatural Generate natural orbitals for the spin density (with α considered positive). but store only the β densities for use in a . constrain them to reproduce the dipole moment. These are read as pairs of atom number and radius.217]. CHelpG Produce charges fit to the electrostatic potential at points selected according to the CHelpG scheme [219].wfn file (see Output=WFN). MK Produce charges fit to the electrostatic potential at points selected according to the MerzSingh-Kollman scheme [216.Only.BetaNatural Do separate natural orbital analyses for the α and β densities. These are read as pairs of atomic symbol and radius. terminated by a blank line.Save. Dipole When fitting charges to the potential. natural orbitals are not included in the checkpoint file.NaturalOrbitals) Geom=AllCheck Run the formchk utility on the resulting checkpoint file to prepare the orbitals for visualization. ESPDipole is a synonym for Dipole. ReadRadii Read in alternative radii (in Angstroms) for each element for use in fitting potentials. AtomDipole When fitting charges to the potential. also fit a point dipole at each atomic center. ReadAtRadii Read in alternative radii (in Angstroms) for each atom for use in fitting potentials. Use a second job step of this form to place the natural orbitals into the checkpoint file: --Link1-%Chk=name # Guess=(Read. NBO-RELATED OPTIONS . NCSAll Requests an NCS analysis of all tensor components.C. Output=WFN . al.18. Weinhold and T. using NBO version 3 [12. NoNCS skips this analysis. NPA Requests just the Natural Population Analysis phase of NBO. Bohmann.17. J. Phys.19].16. which is based upon the NBO analysis method. Implies that NBO input will be read. Refer to the NBO documentation for details on this input.14. SaveNBOs Save natural bond orbitals in the checkpoint file (for later visualization). Chem. 107 (1997) 1173. Density. F. NBODel Requests NBO analysis of the effects of deletion of some interactions. with input controlling the analysis read from the input stream. By default. NBORead Requests a full NBO analysis.13. Farrar. refer to the NBO documentation for details. NCS Requests a partitioning of the NMR shielding tensors (computed using GIAOs) into magnetic contributions from bonds and lone pairs using the Natural Chemical Shielding Analysis of Bohmann et.NBO Requests a full Natural Bond Orbital analysis.15. an analysis of the isotropic shielding is performed. SaveNLMOs Save natural localized molecular orbitals in the checkpoint file (for later visualization).]. SaveMixed Save the NBOs for the occupied orbitals and the NLMOs for the unoccupied orbitals in the checkpoint file (for later visualization). Only possible with SCF methods. Note that NBO input starts in column 2 so that the UNIX shell does not interpret the initial $. Use this option to specify keywords for NBO. NCSDiag Requests an NCS analysis of the diagonal tensor elements.A. [ J. 000000 -0.237787 0.The following input file requests a bond order analysis using NBO 5: # B3LYP/6-31G(d. the potential.. The value should be specified as an option: # .919239 0.665676 1.000000 0.919278 0. and electric field gradient at each nucleus are computed.237787 -0.377.555].000000 0.5 The default is 1 atmosphere. .919239 0.237739 1. PROPERTY SELECTION OPTIONS EFG Specifies that potential.278.000000 $nbo bndidx $end Pressure Specifies the pressure to be used for thermochemistry analysis (in atmospheres). Pressure=1..237739 -1.000000 0. This is the default.665676 -1. field and field gradient are to be computed. Prop This properties keyword tells Gaussian to compute electrostatic properties [276. electric field.000000 0.p) Pop=NBORead Example of NBO bond orders C H C H H 0 1 0. By default.000000 0.000000 0.919278 -0. The density used for the electrostatic analysis is controlled by the Density keyword. Potential Specifies that the potential but not the field or field gradient are to be computed. but not the field gradient. EPR Compute the anisotropic hyperfine coupling constants (i. Opt Causes the program to read a list of centers as in Prop=Read. Grid Specifies that the potential is to be calculated at one or more grids of points and written to an external file (generally superseded by cubegen).Y1. a single line of input supplies all of the necessary information: . Dipole Constrain fitted charges to the dipole moment. Field Specifies that the potential and field. the order of the input sections is fixed points (Read). FitCharge Fit atomic charges to the electrostatic potential at the Van der Waals surface.Z2 Fortran unit for write. # grid column & horizontal step size. # grid rows & vertical step size.Z1 N2. Three additional input lines are required for a uniform grid: KTape.Y2.ZO N1. then optimized points (Opt).XO. coords. with one center per line. but then to locate the minimum in the electric potential closest to each specified point. in the standard orientation.X1. The points can be specified as a uniform rectangular grid. as an arbitrary collection read from an auxiliary file (both described below). of map's lower left corner. For points read from an auxiliary file.X2.YO.278. INPUT SOURCE-RELATED OPTIONS If both Read and Opt are specified. Read Causes the program to read a list of additional centers at which properties will be computed from the input stream. are to be computed. The Cartesian coordinates of each center in angstroms are read in free field.. This option requests mapping of the electric potential over a 2D grid of points. or via the input format used by Cube=Potential (see Appendix D). NoPotential suppresses computation of the electric potential and higher properties. spin-dipole EPR terms) [276.e.377]. KTape The coordinates of N points in Angstroms will be read from unit LTape. field. the following input indicates that 19. the Pseudo keyword applies to all layer of the ONIOM. CISD. CIS.LTape. Cards is a synonym for Read. LTape defaults to 52.NEFG. LP-31. MP2.12. Gaussian supports a new effective core potential (ECP) input format (similar to that used by ExtraBasis) which is described below. If used the ONIOM. Old Read pseudo-potential data using the old format (used by Gaussian 92 and earlier versions).696 points for the electrostatic potential (code 3) will be read from Fortran unit 10. cubegen Pseudo This keyword requests that a model potential be substituted for the core electrons. Read Read pseudo-potential data from the input stream. LANL1. CID. CCD. potential and field (NEFG=2). The potential (NEFG=3). The Cards option is by far its most-used mode. Density. LANL2. . SDD and SHC. When reading-in pseudopotentials. CCSD and QCISD. do not give them the same names as any internally-stored pseudopotentials: CEP. For example. If you want to read in ECPs only for one ONIOM layer.N. and field gradient (NEFG=1) will be computed and written to unit KTape. in format 3F20.10.3. with output written to Fortran unit 11: 19696. or potential. MP4(SDQ). all DFT methods. then use the GenECP keyword instead.11 HF. CHF. MP3. Input is described in the next subsection below. Gaussian radial functions and angular momentum projection operators. exponent) for each potential for each term in each angular momentum of the ECP. LANL1 Requests the LANL1 potentials. the input includes the general (d and higher) term. p.CHF Requests the Coreless Hartree-Fock potentials. LANL2 Requests the LANL2 potentials. and (2) the extra terms for each special angular momentum. what to add to the general term to make the s component) and the p-d term. and number of core electrons replaced by the potential. SHC Requests the SHC potentials.e.Max.. This option is normally used with the LP31G basis sets. Each block is introduced by a line containing the center numbers (from the molecule specification) and/or atomic symbols. If Name matches the name of a previous potential. For each component (I=1 to Max) of the current potential. 2 if there are special s and projections. and then includes a collection of triplets of: (coefficient. The pseudo-potential for those centers/atoms follows: Name. containing the following information: . Since only the first few angular momentum components have different terms. maximum angular momentum of the potential (i. 3 if there are s. All ECP input is free-format. Thus for an LP-31G potential. a group of terms is read. and d projections). FULL ECP INPUT FORMAT Effective Core Potential operators are sums of products of polynomial radial functions. ECP input therefore specifies which potential to use on each atomic center. typically d or f and higher projection.. that potential is reused and no further input other than the terminator line (see below) is required. specifying the atoms and/or atoms types to which it applies (just as for general basis set input-see the discussion of the Gen keyword).e. the s-d term (i. the potential is expressed as (1) terms for the general case. The list ends with a value of 0. which includes special s and p projected terms. power of R.ICore Name of the potential. NPower includes the R2 Jacobian factor. KEYWORDS FOR STUTTGART/DRESDEN ECP INPUT In Pseudo input. NPower.or twovalence electron atoms SDF is a good choice. The following table shows the availability of the various XY combinations.Expon. keywords for these ECP's are of the form XYn where n is the number of core electrons which are replaced by the pseudopotential and X denotes the reference system used for generating the pseudopotential (S for a single-valence-electron ion or M for a neutral atom). WB for Wood-Boring quasi-relativistic and DF for Dirac-Fock relativistic.Coef Power of R. For one.Title A description of the block. In this case. along with valid values for n. otherwise MWB or MDF is recommended (although for small atoms or for the consideration of relativistic effects. SIMPLIFIED ECP INPUT FORMAT Gaussian adds flexibility to ECP input by allowing it to include pre-defined basis sets names. the corresponding SHF and MHF pseudopotentials may be useful). Valid values of n for given values of X and Y . An example of an input file which includes a nonstandard ECP with its associated basis set is given below. The Defaults columns list the equivalencies for the SDD keyword (which selects an all electron basis set through Cl and ECPs thereafter) and when IOp(3/6) is set to 6 (which selects ECPs for all elements). exponent. the ECPs within the specified basis set corresponding to the specified atom type(s) will be used for that atom (see the examples). and coefficient for each of the NTerm terms. The Stuttgart/Dresden ECPs are not uniformly available across the periodic table. NTerm Number of terms in the block. and gradients through d functions only. Y specifies the theoretical level of the reference data: HF for Hartree-Fock. Energies through f functions only. An ECP definition may be replaced by a line containing the standard keyword for a pre-defined basis set. not otherwise used. Atom 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 He Li Be B C N O F Ne Na Mg Al Si P S Cl Ar K Ca Sc Ti V Cr Mn Fe Co Ni Cu Zn Ga Ge As Se Defaults IOp(3/6=6) MWB SDD keyword D95 D95 D95 D95 D95 D95 D95 D95 D95 D95 6-31G 6-31G D95 D95 D95 D95 D95 6-31G MWB10 MWB10 MDF10 MDF10 MDF10 MDF10 MDF10 MDF10 MDF10 MDF10 MDF10 MDF10 MWB28 MWB28 MWB28 MWB28 SDF SHF MDF MHF SDF2 SDF2 MWB2 MWB2 MWB2 MWB2 MWB2 MWB2 SDF10 SDF10 MWB10 MWB10 MWB10 MWB10 MWB10 MWB10 MWB10 MWB10 MDF10 MDF10 MDF10 MDF10 MDF10 MDF10 MDF10 MDF10 MDF10 MDF10 MWB28 MWB28 MWB28 MWB28 2 2 2 2 2 2 2 2 2 2 2 2 2 10 10 10 10 10 10 10 10 18 18 18 18 10 10 10 10 10 10 10 10 10 10 MDF10 MDF10 MDF10 MDF10 MDF10 MDF10 MDF10 MDF10 MDF10 MDF10 10 10 10 10 10 10 10 10 28 28 28 28 28 28 28 28 28 28 28 28 . 52 52.53 53.50.35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 Br Kr Rb Sr Y Zr Nb Mo Tc Ru Rh Pd Ag Cd In Sn Sb Te I Xe Cs Ba La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB46 MWB46 MWB46 MWB46 MWB46 MWB46 MWB46 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB46 MWB46 MWB46 MWB46 MWB46 MWB46 MWB46 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 MWB28 28 28 28 28 28 28 28 28 28 28 28 28 28 28 46 46 46 46 46 46 46 46 28.49 49.46.53 28.51.49.51 51.55 28.48 48.58.59 59 .58 28.56 28.55.57 28.47.47 47.52 28.57.55 55.56 56.51 28.50 50.54 54.52.59 28 28 36 36 36 36 28 28 28 28 28 28 28 28 28 28 46 46 46 46 46 46 54 54 54 46 46 46.57 57.49 28.47 28.56.53.48.54.48 28.50 28.58 58.59 28.54 28. 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 Lu Hf Ta W Re Os Ir Pt Au Hg Tl Pb Bi Po At Rn Ac Th Pa U Np Pu Am Cm Bk Cf Es Fm Md No Lr Rf MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB78 MWB78 MWB78 MWB78 MWB78 MWB78 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB78 MWB78 MWB78 MWB78 MWB78 MWB78 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 MWB60 28.78 78 78 78 78 78 78 60 60 60 60 60 60 60 60 60 60 60 60 60 60 60 78 60 60 60 60 60 60 60 60 60 60 60 60.78 78 78 78 78 78 78 60 60 60 60 60 60 60 60 60 60 60 60 60 60 60 92 Note: These ECPs are not available for elements 87 (Fr).60 60 60 60 60 60 60 60 60 60. 88 (Ra). and 105 and higher . 06623814 S-D projection angular momentum).40000000 2 1. ExtraBasis. Gen.97790770 2 30. applies to d & higher.ChkBasis. ECPs for ECP name=OLP.A3 H. 1 80.47 General basis set input **** O 0 the oxygen atoms. 2 2 52. in this case oxygen atom 1. Description for the Number of Corrections for projected terms (lowest Corrections for projected terms (highest Blank line indicates end of the ECP The basis set data follows the molecule specification section.49630500 block for oxygen. with the basis set and ECP data read from the input file: # HF/Gen Pseudo=Read Test Hydrogen peroxide 0.51654843 2 9.96 R3=1.2.3427019 27.1.3033500 17. D component general terms. OLP 2 2 replaces 2 electrons. 3 0 0. it is a .3.A3. The potential on this center is named OLP. 3 terms to follow.1.1 O H.6481971 2. This input file runs an RHF/LP-31G calculation on hydrogen peroxide.0000000 -1.0 R2=0. GenECP Specifying an ECP.0953760 -0.9212952 0.R2 O.7220233 -16.180.48 A3=109. The first line of the ECP data requests that a potential be read in (type 7) for atoms number 1 and 3 (the oxygen atoms) and that no potential is to be used for atoms 2 and 4 (the hydrogen atoms).60000000 1 30.04478500 P-D angular momentum).R3.R2. The second line of ECP data begins the input for the first center requiring a read-in potential.39552179 0 28..0000000 -0.2.1. . and each of the components.general term and applies to angular momentum 2 (D) and higher. it does not format the MO information in Gamess input format.) Options are used to specify what information should be output. Gamess) sends both the molecular orbital and Gamess input information to the file. Using Standard Basis Set Keywords to Specify ECPs. . the next potential. Finally. except that only one of MO and NaturalOrbitals can be requested. consists of a single line.7 under UNIX. Next comes a title for the general term. however. Punch(MO.0 0. C O 0 6-31G(d) **** Cr 0 LANL2DZ **** Cr 0 LANL2DZ ECP for chromium atom. lowest angular momentum first. that they are distinct and non-interacting. For example. The following input file illustrates the use of the simplified ECP input format: # Becke3LYP/Gen Pseudo=Read Opt Test HF/6-31G(d) Opt of Cr(CO)6 0 1 Cr 0. Note. All of these options can be combined. The output is disposed of in whatever manner is usual for Fortran alternate-unit output under the appropriate operating system (for example. It uses the same name as a previous potential (that of center 1) and so the information already read in is reused. Note that the maximum angular moment and number of core electrons must still be specified. unit 7 is sent to the file fort. the number of components of that term. send to a separate output file-useful information at various points in the calculation.0 0.0 molecule specification continues . even though they will generally be the same for all uses of a given potential. followed by the corrections for the projected terms. and the potential replaces two electrons. for center 3 in this case. Punch This output specification keyword allows the user to "punch"-in more modern parlance.. Use the ECP in this basis set. NaturalOrbitals Punches natural orbitals (for the density selected with the Density keyword). GAMESSInput Punches an input deck for GAMESS. All Punches everything except natural orbitals.Archive Requests that a summary of the important results of the calculation be punched. MO Punches the orbitals in a format suitable for Guess=Cards input. Note that this keyword requests only QCISD and does not include the triples correction [556.557] by default (see T below). Coord Punches the atomic numbers and Cartesian coordinates in a form which could be read back into Gaussian. HondoInput Punches an input deck for one version of Hondo. suitable for later use with Opt=FCCards. and second derivatives in format 6F12. Cartesian nuclear coordinate derivatives. Output QCISD This method keyword requests a Quadratic CI calculation [72]. Title Punches the title section. including single and double substitutions.8. This output is in the same format used by the Browse Quantum Chemistry Database System. which is probably easily modified to fit most others. . Derivatives Punches the energy. MaxCyc=n Specifies the maximum number of cycles.T Requests a Quadratic CI calculation including single and double substitutions with a triples contribution to the energy added [72].558]. J. CCSD The predicted energy from a QCISD calculation appears in the output in the final QCISD iteration: DE(CORR)= -. Note that Q1 is analogous to the T1 diagnostic for CCSD when it is computed using QCISD instead of the Coupled Cluster method. Conver=N Sets the convergence calculations to 10-N on the energy and 10-(N+2) on the wavefunction. Analytic energies and gradients for QCISD. T1Diag Computes the Q1 diagnostic of T. E4T Requests a Quadratic CI calculation including single and double substitutions with a triples contribution to the energy and also an evaluation of MP4 triples. The default is 50. See the discussion here for details. and numerical frequencies for all methods.7501966245D+02 . Must be specified with the T option. The default is N=7 for single points and N=8 for gradients.54999890D-01 E(CORR)= -. FC The frozen-core options for defining inner-shells to be excluded from the correlation calculation are valid with this keyword. TQ Requests a Quadratic CI calculation including single and double substitutions with an energy contribution from triples and quadruples [64] added. numerical gradients for QCISD(T). Lee and coworkers [423. 129.135]. For closed shell systems. Test SAC-CI The keyword selects the Symmetry Adapted Cluster/Configuration Interaction (SAC-CI) methods of Nakatsuji and coworkers [121. AnionDoublet.126. In this case.When QCISD(T) is specified. Quartet.134.122.132. you must either select an ROHF ground state wavefunction by including ROHF in the route section in addition to SAC-CI.124.125.jp/nakatsuji-lab.123.sbchem. Archive. Triplet.ac.127.kyoto-u. For open shell ground states.128. Other spin state selection options are CationDoublet (Doublet is a synonym). consult the SAC-CI documentation available at the following web site: www.75019725718D+02 ReArchive This calculation type keyword requests that the information on the checkpoint file be used to generate an archive entry. no new calculation is performed.130. Sextet and Septet. the default RHF wavefunction used by SAC-CI is appropriate. More than one spin state may be specified. the preceding output is followed by the energy including the non-iterative triples contribution: QCISD(T)= -. For detailed information on this method. See the examples for more information.131.133. SAC-CI jobs must specify a reference state for the subsequent excited states calculations. SPIN STATE OPTION Singlet=(suboptions) Specifies that singlet states are to be calculated. . or you must specify a closed shell state for the ground state calculation using the AddElectron or SubElectron option. Quintet. The parenthesized list of suboptions specifies the desired states and other calculation parameters. This is the default for such systems for CationDoublet. D2 for Td). and so on). 8 for D2h.e. and n is the solution number in the desired spin state (determined by a previous energy calculation). depending on the molecular symmetry (e. or for use with the Density keyword.g. OTHER COMMONLY-USED OPTIONS TargetState=(SpinState=s.. and between the lowest SAC-CI states and SAC-CI excited states for other spin states. Doublet.. This is the default for such systems for AnionDoublet. TransitionFrom=(SpinState=s. AddElectron Add one electron to the open shell reference SCF configuration. SpinState=(NoTransitionDensity) By default. Symmetry=m. S is the keyword indicating its spin multiplicity (i... Root=n) . Root=n) Specifies the target state for a geometry optimization or a gradient calculation. Doublet.g. SpinState=(Density) Calculate unrelaxed density matrices and perform Mulliken population analysis for all computed SAC-CI states of spin SpinState.. Singlet. etc. m is the irreducible representation number of its point group. SubElectron Subtract one electron from the open shell reference SCF configuration. The shorthand form NState=N specifies a value of N for each irreducible representation.SPIN STATE SUBOPTIONS SpinState=(NState=(i1.i2. See the examples for more information. Quartet and Sextet.. Symmetry=m.). Implies the FullActive option as well. SpinState=(SpinDensity) Calculate spin density matrices for all computed SAC-CI states of spin SpinState. NoTransitionDensity disables these calculations for the corresponding spin state. 4 for C2v. the transition density and oscillator strength are calculated between the SAC ground state and the SAC-CI singlet excited states when SpinState is Singlet. Degeneracies are handled by assuming the closest linear symmetry (e.)) Sets the number of states of the specified type to be calculated for the various irreducible representations of the molecule's point group. Up to eight values may be specified. You can use this option to determine the state number of a particular state in which you are interested (e. SpinState=(InCoreDiag) Force use of the in-core algorithm.e. all electrostatic properties and the diamagnetic terms (shielding and susceptibility).Specifies the initial state for for calculating transition density matrices. m is the irreducible representation number of its point group. ADDITIONAL OPTIONS FOR EXPERT USERS ADDITIONAL SPIN STATE SUBOPTIONS SpinState=(MaxR=N) Set the maximum excitation level to N. Note that this option only applies to the excited state portion of the calculation (the ground state calculation always uses a nonvariational procedure). This option applies to all spin states which specify the Density suboption. Variational proceeds by diagonalizing symmetrized matrices. Doublet. etc. and n is the solution number in the desired spin state (as for TargetState above).). SelectCISOnly Terminate the calculation after the CIS initial guess has been calculated. . NoProperty Don't calculate any molecular properties. AllProperties Calculate multipole moments through hexadecapole. Singlet. all Nth moment to the 4th moment. SpinState=(Iterative=item) Force the use of an iterative algorithm. and it is the default..g. for TargetState). Item specifies the initial guess type: SInitial for CIS and SDInitial for CISD. See the examples for an alternative method. SpinState=(NonVariational) Solve the SAC-CI equations for non-symmetric matrices. S is the keyword indicating its spin multiplicity (i.. SACOnly Performs only the calculation for the reference state and does not compute any excited states. PROCEDURAL OPTIONS FC The frozen-core options for defining inner-shells to be excluded from the correlation calculation are valid with this keyword. LevelOne Set the threshholds for selection of the double excitation operators to the lowest recommended level. For this reason. and it is the default. LevelThree is the most accurate level. DConvDiag=M Set the diagonalization energy convergence criteria to 10-M. This is the default. using a full orbital window is recommended. DConvSAC=M Set the energy convergence criteria to 10-M when solving the SAC equations. WithoutDegeneracy . The default value of N is 0. MaxItDiag=N Set the maximum number of diagonalization iterations. InCoreSAC For solution of the SAC equations using the in-core algorithm. General-R Perform the calculation including linked excitation operators through sextuples. MaxItSAC=N Set the maximum number of iterations for solving the SAC equations. LevelTwo is intermediate in accuracy between the other two levels. The available types are PM (Pipek-Mezey) and Boys. In general. the size of the active space greatly affects the accuracy of SAC-CI calculations. LMO=type Use the specified type of localized MO as reference orbitals. ACCURACY LEVEL OPTIONS SD-R Perform the calculation using singles and doubles linked excitation operators. Full is the default for geometry optimizations and gradient calculations. MacroIteration=N Requests the use of N macroiterations within an optimization step. FullRGeneration Generate all higher-order linked operators in the General-R scheme up to MaxR=4 and then perform perturbation selection as above. . FullUnlinked Include all types of unlinked terms. NoLinkedSelection Disables perturbation selection threshholds for linked operators (i. ionized and electron-attached states. all three of these preceding options are required.e. then at some or all subsequent points with CalcGSUM and then finally at all points with AfterGSUM.e. The actual results are provided by the final calculation. This procedure is only valid for singlet. all operators are included). resulting in reduced computational requirements. This is the default for gradient calculations and geometry optimizations. all operators are included). Use of this option is not recommended for production use. The Scan calculation must be performed three times: at the first point with BeforeGSUM. This option results in a tradeoff between decreased accuracy and computational requirements. LevelTwo and LevelThree options. perturbation selection is performed so that degeneracies are retained. In order to include all terms. This option suppresses this test. NoUnlinkedSelection Disables perturbation selection threshholds for unlinked operators (i. Perturbation selection threshholds are set via the LevelOne... EgOp Generate quadruple and higher-order linked operators in the General-R scheme via the exponential generation algorithm. The highest order excitation level is specified via the MaxR option (up to a maximum of 6). and it is not compatible with the General-R option. WithoutR2S2 Ignore R2S2 unlinked integrals. Forces the use of the in-core algorithm. GROUP SUM OPERATION OPTIONS These options are used to ensure consistency between all points in multipoint calculation types like potential energy surface scans. This is the default for single point energy calculations.By default. triplet. currently at a considerable performance penalty. 000. This will search for 8 singlet states. Use this option in a calculation at the first point.. The default is 100. Geometry optimizations default to using a full window. MEMORY USE OPTIONS These options can be used to increase the program default settings after a failed job has indicated that a resource shortfall was the problem. . Specifying a different frozen core option for an optimization will result in numerical gradient calculations and correspondingly poorer performance. Analytic energies and optimizations and numerical frequencies. The two lowest excited states will probably be among those found by the calculation. AfterGSUM Perform SAC-CI calculations at each point using the GSUM data collected previously with the CalcGSUM option.Singlet=(NState=8))/6-31G(d) NoSymm . MaxR2Op=N Set the maximum number of R2 operators after perturbation selection to N.. ignoring symmetry. you could use a route like the following: # SAC-CI=(Full. CalcGSUM Collect data and determine the threshholds and operator selections at specified points in order to form a consistent set which can then be used at every point.000. MaxEgOp=N Set the maximum number of operators in the General-R method to N.BeforeGSUM Initialize a series of linked calculations. Density If you want to locate the lowest two singlet excited states. The default is 5. TargetState=(SpinState=Singlet. use a route like this one: # SAC-CI=(Full.Density))/631G(d) .0))/6-31G(d) .Quartet=(NState=3)) . the following route will locate 4 singlet excited states of each symmetry type: # SAC-CI=(Full... Geometry Optimizations. and it computes three doublet and three quartet excited states for each irreducible representation. then you would omit the Density suboption to the Singlet option.2))/6-31G(d) . For example: # SAC-CI=(Full. use a route like this one: # SAC-CI=(Full. You could use a similar approach for the triplet ground state of methylene.1.. Computing Densities and Molecular Properties. To specify the desired number of singlet excited states for each irreducible representation for a molecule with C2v symmetry. use the TargetState option: # Opt SAC-CI=(Singlet=(Nstate=4)..Doublet=(NState=3)..Density). This job could be followed by a normal (LevelThree) calculation for the state(s) of interest..2.LevelOne)/6-31G(d) . This calculation will locate the lowest four singlet excited states for each irreducible representation. This specifies the use of an ROHF wavefunction for the ground state.. Locating States with an Inexpensive Initial Calculation... Calculations on Open Shell Systems. a neutral doublet radical. Singlet=(2. To predict excited states for vinyl radical.Singlet=(NState=4)...Triplet=(. To compute the unrelaxed density and population analysis for all predicted excited states.0.Root=2))/6-31G(d) . you could use the following route: # SAC-CI=(Full.Symmetry=1...Singlet=(. lower-accuracy calculation in order to locate a desired excited state at reduced computational cost.1.Alternatively. To optimize a specific excited state.Singlet=(1..Singlet=(NState=4))/6-31G(d) .. .. For example. You can use a preliminary.. If you wanted to compute the unrelaxed density and population analysis only for the triplet states.... you could use a route like the following: # ROHF/6-31G(d) SAC-CI=(Full. 6714 0.4502 0.0334 -0.0460 A1 2 18.2989 0..0 Excitations are from this state.7019 0.)) Density=Current .0000 A2 2 18.0000 0.6670 0.0000 0.3505 A1 4 18.0000 0.1159 0.5135 0.0000 0..0000 0.0581 0. The value should be specified as an option: # .0000 0.0000 -0.0000 0.95 The default is 1.0000 0.0000 0.. use a route like the following: # SAC-CI=(Full.0070 --------------------------------------------------------------------- Note that the various excited states are grouped by symmetry type—and not in order of increasing energy—in the output.0000 0. energy (eV) X Y Z strength --------------------------------------------------------------------A1 0 0.0077 0.8155 0.0000 0.0940 A1 3 18.2042 B1 3 22.3644 B2 4 23.0000 0.0000 1.7395 -0.2740 0.0000 0.To compute the relaxed density and population analysis for only one specified state.0000 0.. SAC-CI Output.0000 0.0000 -0.0000 0.0000 A2 1 7.0000 0.0000 0.0000 B1 1 1.1500 0. SAC-CI calculations produce a table like the following for each requested spin state (this example is for singlet states): --------------------------------------------------------------------Transition dipole moment of singlet state from SAC ground state --------------------------------------------------------------------Symmetry Sol Excitation Transition dipole moment (au) Osc.. .8904 0.6587 0.5055 0. Scale=0.8252 0.1099 0..7764 0.TargetState=(.0000 0.0023 B1 2 18.2639 B2 1 11.9280 0. A1 1 8.0000 0. Scale Specifies the frequency scale factor to be used for thermochemistry analysis.0422 0.7853 0.0000 0.1915 -0.4645 0.0000 -0.0000 0.0000 -0.0000 0.8696 B2 3 24.1671 B2 2 15.0122 B1 4 15.0000 0. Note that this job will be much more computationally expensive than the previous one as it requires a full gradient calculation.0 except for compound methods where the default specified by the method is used.0000 0.Singlet=(NState=4).5153 0. 0 A2 120.56) will be done for each combination of other variables.2. and 1.0 This input causes variable R1 to be stepped 3 times by 0. following the variable's initial value.Scan This calculation type keyword requests that a potential energy surface (PES) scan be done. 1. The number of steps and step size for each variable are specified on the variable definition lines. or the job will fail with an error.41. 1. Restart Restarts a PES scan calculation. A relaxed PES scan (with geometry optimization at each point) is requested with the Opt keyword.5 2 1. All in all.41 3 0. R1 values (1. Similarly. which consists of single point energy evaluations over a rectangular grid involving selected internal coordinates. For example: R1 1.05 A1 104. If any scanning variable breaks symmetry during the calculation. Thus four. . and A2 will be held fixed at 2. A rigid PES scan is performed. Any number of variables can be stepped. The units of the step-sizes are controlled by the Units keyword and default to Angstroms and degrees. then you must include NoSymm in the route section of the job.05. The molecular structure must be defined using Z-matrix coordinates. A failed Scan calculation may be restarted from its checkpoint file by simply repeating the route section of the original job. 3 values for A1 will be used.51.46. No other input is required. a total of 12 energy evaluations will be performed. adding the Restart option to the Scan keyword. Opt Output files from PES scans conclude with a table summarizing the results for the job: Scan completed. SCF This keyword controls the functioning of the SCF procedure.41306 104.5000 -38.0600 10 0.39412 107.Summary of the N R ---. Single point direct SCF calculations are run with modest convergence criteria automatically in the interest of speed. The default SCF procedure uses a combination of EDIIS [559] and CDIIS.9600 8 1.39041 104. with no damping or Fermi broadening.0100 6 1.41657 107.9600 2 1.5000 -38. and so on.42453 106.5000 -38.0100 9 1.5000 -38. Click here for more information on maximizing performance in the SCF for different problems. electrostatic potential derived charges. SCF=Tight requests full convergence for this case.5000 -38.----------104.0600 7 0.42564 107.5000 -38.1 kcal mole-1 accuracy in the SCF energy and 3 decimal places in the density matrix-sufficient for population analysis. sometimes it is useful to start off optimizations with less accurate integral. SCF and DFT single point energy calculations involving basis sets which include diffuse functions should always use the SCF=Tight keyword to request tight SCF convergence criteria. The Sleazy option reduces all of these cutoff values.39172 105.41547 106.39296 106. SCF. At the other extreme.41430 105. .0100 12 1.42336 105. The default for this case is sufficient for 0. and CPHF cutoffs and convergence criteria and then to enable the more accurate and expensive limits only when the geometry has stabilized.----------- Chapter 8 of Exploring Chemistry with Electronic Structure Methods [308] provides a detailed discussion of potential energy surface scans.5000 -38.--------- potential surface scan: A HF --------.9600 5 1.0600 4 0. It also turns off archiving. and the like.5000 -38. Options are used to specify the desired behavior.5000 -38. alternate algorithms.0100 3 1.0600 ---.9600 11 1.5000 -38.42668 --------.5000 -38.--------1 0.5000 -38. QC Calls for the use of a quadratically convergent SCF procedure [563]. This method is slower than regular SCF with DIIS extrapolation but is more reliable. See reference [560] for a discussion of SCF convergence and stability. XQC Add an extra SCF=QC step in case first-order SCF has not converged. Fermi Requests temperature broadening during early iterations [562]. MaxConventionalCycles=N Sets the limit on conventional SCF cycles during SCF=XQC to N. SCF=QC is not available for restricted open shell (RO) calculations. SSD Does scaled steepest descent SCF. NDamp=N Allow dynamic damping for up to N SCF iterations (the default is 10). Damp Turn on dynamic damping of early SCF iterations. combined with CDIIS and damping. . NoFermi suppresses Fermi broadening and is the default. ALGORITHM SELECTION OPTIONS DIIS DIIS calls for and NoDIIS prohibits use of Pulay's Direct Inversion in the Iterative Subspace extrapolation method [561]. NoDamp is the default. Fermi implies Damp as well by default. CDIIS Use only CDIIS. CDIIS implies Damp as well. SD Does steepest descent SCF. Note that damping and EDIIS do not work well together. damping is enabled if SCF=Fermi or SCF=CDIIS is requested. However.Single point energy calculations involving basis sets which include diffuse functions should always use the SCF=Tight keyword to request tight SCF convergence criteria. and also include level shifting. By default this involves linear searches when far from convergence and Newton-Raphson steps when close (unless the energy goes up). a full minimization is done only if the initial microiteration caused the energy to go up.e. the convergence criterion is tightened up as the rotation gradient is reduced. Note that with DIIS turned on. VeryTightLinEq Use even tighter convergence in the linear equation solutions (microiterations) . steepest descent is used. By default. Pass For in-core calculations. MaxRot=N Set the maximum rotation gradient for a Newton-Raphson step in SCF=QC to 10-N. By default. N defaults to 100. saves the integrals on disk as well. SD. NoVShift is equivalent to this setting. TightLinEq Use tight convergence in linear equation solution throughout SCF=QC. This is the default for direct SCF. Available only for RHF closed shell and UHF open shell calculations. nonincremental iteration after an SCF using DIIS or a direct SCF has converged. Only useful for frequency jobs in conjunction with SCF=InCore. scaled steepest descent is used. and it is the default for conventional SCF.. memory requirements increase with increasing maximum number of cycles. FinalIteration FinalIteration performs and NoFinalIteration prevents a final non-extrapolated. N millihartrees). NoPass forces integrals to be recomputed during each in-core phase. MaxCycle=N Changes the maximum number of SCF cycles permitted to N. VShift[=N] Shift orbital energies by N*0. Above this. NoIncFock prevents the use of incremental Fock matrix formation.DM Calls for use of the direct minimization SCF program [564]. N=-1 disables level shifting. above 100 times this.001 (i. The default is NoFinalIteration. This option disables automatic archiving. to avoid recomputing them in Link 1002. IncFock Forces use of incremental Fock matrix formation. the default is 64 (or 512 for SCF=DM and SCF=QC). It is usually inferior to SCF=QC and retained for backwards compatibility and as a last resort. FullLinear Specifies that L508 (SCF=QC. or SSD) should do full linear searches at each iteration. The default value for N is 2. and NoVarInt is a synonym for NoVarAcc. Synonymous with NoSinglePoint.throughout the QCSCF. Conver=N Sets the SCF convergence criterion to 10-N. The default for direct SCF. This option is sometimes needed for nearly linearly-dependant cases. This is possible for all available methods.Direct). Sleazy is a synonym for SinglePoint. VTL is a synonym for VeryTightLinEq. INTEGRAL STORAGE OPTIONS Direct Requests a direct SCF calculation. equivalent to SCF=(Conv=4. except for MCSCF second derivatives and anything using complex orbitals. Note that for single-point direct SCF calculations. a loose convergence criterion (10-4) is used in the interest of speed. respectively. Tight Use normal. VarInt is a synonym for VarAcc. NoSleazy and TightIntegrals. for which it is in terms of the orbital change and energy change. NoSP. SCF=InCore is available to force in-core storage or abort the job if not enough is available. This is the default SCF procedure in Gaussian. SinglePoint Requests the loose SCF convergence criteria appropriate for single points. This is a density-based convergence criterion except for GVB and CASSCF. InCore Insists that the SCF be performed storing the full integral list in memory. Can be abbreviated SP. NoInCore prohibits the use of the in-core procedure. The default for everything except CASSCF and direct SCF single points. in which the two-electron integrals are recomputed as needed. for both the SCF and CPHF. VerySleazy Reduce cutoffs even further. tight convergence in the SCF. NoDirect is a synonym for Conventional. uses Int=CoarseGrid and single-point integral accuracy .NoFinal. VarAcc Use modest integral accuracy early in direct SCF. This is done automatically in a direct SCF calculation if sufficient memory is available. The default for single point CASSCF or direct SCF.VarInt. switching to full accuracy later on. can be turned off via NoVarAcc. Conventional The two-electron integrals are stored on disk and read-in each SCF iteration. IntRep Calls for the SCF procedure to account for integral symmetry by replicating the integrals using the symmetry operations. NoIDSymm is the default. This is the default. FockSymm Calls for the SCF procedure to account for integral symmetry (use of the "petite" integral list) by symmetrizing the Fock matrices. DSymm Symmetrize the density matrix at every SCF iteration to match the symmetry of the molecule ("density symmetrize"). NoDSymm is the default. This is the default for direct SCF. ROHF and UHF) and L508 (SCF=QC). SCF=DM cannot be restarted. It is the default only for GVB calculations. It is synonymous with Guess=NoSymm and Symm=NoSCF. so the SCF can be restarted.during iterations. SYMMETRY-RELATED OPTIONS IDSymm Symmetrize the density matrix at the first iteration to match the symmetry of the molecule ("initial density symmetrize"). Restart Restart the SCF from the checkpoint file. Allows use of a short integral list even if the wavefunction does not have the full molecular symmetry. Available for L502 (the default for RHF. Use this option to retain a specific state of the wavefunction throughout the calculation. Symm Retain all symmetry constraints: make the number of occupied orbitals of each symmetry type (abelian irreducible representation) match that of the initial guess. SCRF . Not recommended for production quality calculations. NoSymm Requests that all orbital symmetry constraints be lifted. NoSave suppresses saving the wavefunction. followed by a single iteration with the usual single point grid (MediumGrid). DSymm implies IDSymm. FSymm is a synonym for FockSymm RESTART-RELATED OPTIONS Save Save the wavefunction on the checkpoint file every iteration. Polarizable Continuum (PCM) models in which the cavity is created via a series of overlapping spheres. the input consists of a line specifying the dielectric constant of the solvent and an optional isodensity value (the default for the latter is 0. Gaussian 03 can also carry out a PCM calculation using Klamt's form of the conductor reaction field (COSMO) [567] and generate the input data for the COSMO-RS solubility programs. OPTION FOR SPECIFYING THE SOLVENT Solvent=item Selects the solvent in which the calculation is to be performed.297.303] and Tomasi. the solute radius in Angstroms and the dielectric constant of the solvent are read as two free-format real numbers on one line from the input stream.287.284. see the discussion of the Volume keyword.300.296.289. REQUIRED INPUT: IPCM AND SCI-PCM MODELS For the IPCM and SCI-PCM models. Keywords within this section follow general Gaussian input rules.290. which places the solute in a spherical cavity within the solvent reaction field.de. using one of the following models: • • • • The Onsager model [281. A suitable solute radius is computed by a gas-phase molecular volume calculation (in a separate job step).298]. REQUIRED AND OPTIONAL INPUT: PCM MODELS Keywords and options specifying details for a PCM calculation (SCRF=PCM.291. www. CPCM or IEFPCM) may be specified in an additional blank-line terminated input section provided that the Read option is also specified. Note that the solvent may .292. See the discussion of the COSMORS keyword for details. The current implementation is the work of Barone and coworkers [285.0004).295].286.299.282. The available keywords are listed in a separate subsection following the examples.565.288. A Self-Consistent Isodensity PCM (SCI-PCM) model [307].302.This keyword requests that a calculation be performed in the presence of a solvent.287.566]. COSMO-RS is distributed as COSMOtherm by COSMOlogic GmbH. Mennucci and coworkers [293.cosmologic.301. initially devised by Tomasi and coworkers [285. REQUIRED INPUT: ONSAGER MODEL For the Onsager model (SCRF=Dipole).293. A static isodensity surface polarized continuum model (IPCM) [307].286.294.283. 39 Acetonitrile or CH3CN: ε=36.519 Xenon or Xe: ε=1. Item is a solvent name chosen from the following list: • • • • • • • • • • • • • • • • • • • • • • • • • Water or H2O: ε=78. the solvent defaults to water.293] (see below).247 Toluene or C6H5CH3: ε=2.335 DiChloroMethane or MethyleneChloride or CH2Cl2: ε=8.43 Quinoline: ε=9.36 CarbonTetrachloride or CCl4: ε=2. Thus.2 Heptane or C7H16: ε=1.7 TetraHydroFuran or THF: ε=7.92 CycloHexane or C6H12: ε=2.295].63 Ethanol or CH3CH2OH: ε=24. METHOD SELECTION OPTIONS PCM For quantum mechanical calculations. IEFPCM Perform a PCM calculation using the integral equation formalism model [288. but be aware it is only one of many internal parameters used to define solvents.294.also be specified in the input stream in various ways for the different SCRF methods. Also.43 Krypton or Kr: ε=1.379 ChloroBenzene or C6H4Cl: ε=5.293.7 Argon or Ar: ε=1. simply changing the ε value will not define a new solvent properly.023 Aniline or C5H5NH2: ε=6. This is the default. . some details of the formalism and the implementation have changed. Note that this option has changed in meaning with respect to Gaussian 98. If unspecified.228 Benzene or C6H6: ε=2.621 NitroMethane or CH3NO2: ε=38.03 Chloroform or CHCl3: ε=4.290.93 DiChloroEthane or CH2ClCH2Cl: ε=10.706 We list the ε values here for convenience.9 Ether or DiEthylEther or CH3CH2OCH2CH3: ε=4.64 DiMethylSulfoxide or DMSO: ε=46.7 Methanol or CH3OH: ε=32. The model of Chipman [568] is closely related to this earlier one [569].58 DiMethylSulfoxide or DMSO or CH3SOCH3: ε=46.89 Acetone or CH3COCH3: ε=20.55 Isoquinoline: ε=10. performs a reaction field calculation using the IEFPCM model [288. as described in [302]. Dipole Perform an Onsager model reaction field calculation. using the Read option (see below). . DIPOLE MODEL OPTIONS A0=val Sets the value for the solute radius in the route section (rather than reading it from the input stream). SCIPCM Perform an SCI-PCM model reaction field calculation: perform an SCRF calculation using a cavity determined self-consistently from an isodensity surface. CPCM Perform a PCM calculation using the CPCM polarizable conductor calculation model [292. after the geometry. Dielectric=val Sets the value for the dielectric constant of the solvent. Structures may be optimized with SCRF=COSMO prior to COSMORS single point calculations. If this option is included.Note that if IEF-PCM is used for an anisotropic or ionic solvent. Isodensity is a synonym for IPCM. This option overrides Solvent if both are specified. basis set and other data. COSMO Perform a PCM calculation using the CPCM model with Klamt's radii and iterative solution. COSMORS Requests a conductor PCM calculation (CPCM) using atomic radii and other parameters as suggested by Klamt for his models. IPCM Perform an IPCM model reaction field calculation.303]. PCM MODELS OPTION Read Indicates that a separate section of keywords and options providing calculation parameters should be read from the input stream (as described above). then Solvent or Dielectric must also be included. The name of the text file to write with input data for COSMO-RS is read from the input stream. then items in the PCM input section must be used to select the anisotropic and ionic dielectric models for these types of solvents. . MP4(SDQ). HF. CASSCF. CID and CISD energies and HF. MP2. performed in gas-phase and in solvent. Int=AM1 must be used in the route section if SCRF AM1 is specified. GradRho Use density basins for the numerical integration. UseMOs Force the use of MOs in evaluating the density. MP2. CIS. However. The solvent reaction field for PCM MP2 calculations is equilibrated to the solute electronic density obtained at the SCF level.Modify Pick up SCRF information from the checkpoint file. CCSD. TD. GasCavity Use the gas phase isodensity surface to define the cavity rather than solving for the surface self-consistently. The job may fail if non-nuclear attractors are present. CIS PCM [298] and TD PCM [300] calculations are by default non-equilibrium calculations with respect to the polarization process between the solvent reaction field and the charge density of the electronic state indicated in the input (where the ground state is the default). The PCM models are available for semi-empirical. SCI-PCM MODEL OPTIONS UseDensity Force the use of the density matrix in evaluating the density. DFT. IEFPCM and PCM may be used to compute frequencies for the methods listed for gradients. but must be obtained by comparing the results of two separate calculations. CCD. equilibrium CIS PCM calculations are the default for geometry optimizations. QCISD. DFT. IPCM MODEL OPTIONS GradVne Use Vne basins for the numerical integration. MP3. Note that ΔGsolvation=EPCM-MP2–EMP2 cannot be obtained using the PCM SCFVac option. CIS and CASSCF gradients. but also read modifications from the input stream. This is mainly a debugging option. solute electronic density polarization process. MP4(SDQ). CCSD. CID.569083211 A. MP3. and for HF and DFT optimizations and frequency calculations.573228 Total free energy in solution: . MP4(SDQ). CCD. CCSD. SCRF=SCIPCM calculations which fail during the SCF iterations should be restarted via the SCF=Restart keyword.4249D-05 -V/T = 2. Only single-point calculations are possible with COSMORS option. Calculation of non equilibrium solute-solvent interaction involving two different electronic states (e.u. The IPCM model is available for HF.568013 <psi(f)|H+V(f)/2|psi(f)> (a. The SCI-PCM model is available for HF and DFT energies and optimizations and numerical frequencies. CCD.) = -98. in two separate job steps (see the PCM input section below). followed by additional information about the calculation. Volume. These calculations will typically be done as single-point solvated calculations using SCRF=PCM optimized geometries.g. QCISD.u.) = -98. QCISD. after 5 cycles Convg = 0. SCF PCM Energy.0033 S**2 = 0. MP2. and CISD energies only.0000 -------------------------------------------------------------------Variational PCM results ======================= <psi(f)| H |psi(f)> (a.By default. MP3. Energy output from the SCRF models other than Onsager appears in the normal way within the output file. The Onsager model is available for HF. the initial and final states of a vertical transition) can be performed using the NonEq=type PCM keyword. DFT. here is the section of the output file containing the predicted energy from a PCM calculation: SCF Done: E(RHF) = -98. CID. and CISD energies. For example. DFT. MP2. The Opt Freq keyword combination may not be used in SCRF=Onsager calculations. SCRF=PCM and SCRF=IPCM jobs can be restarted from the read-write file by using the Restart keyword in the job's route section. CASSCF PCM [297] calculations corresponds to an equilibrium calculation with respect to the solvent reaction field.U. Here is a sample input file: # B3LYP/6-311+G(2d.95061789532 COSMO/RS Example.60 -------------------------------------------------------------------- Additional output lines may appear when various PCM options are included.r.08 Repulsion energy (kcal/mol) = 0.p) SCF=Tight SCRF=(PCM. Note that the PCM results also include the dipole moment in the gas phase and in solution (not shown here). They are placed in a separate input section.1. note that the energy to use is the one preceding the Convergence achieved message (i.96 a 104. as in this example: # HF/6-31++G(d. the one from the final iteration of the SCRF method). Onsager Energy.with all non electrostatic terms (a.e.1.cosmo.) = -98.34 Total non electrostatic (kcal/mol) = 2.569083 -------------------------------------------------------------------(Polarized solute)-Solvent (kcal/mol) = -3..cosmo This job will produce the data file water.a r .r h.5 water.Solvent=Cyclohexane) Test PCM SP calculation on hydrogen fluoride .Read.2. and the various components of the predicted SCRF energy. Additional Keywords for PCM Calculations Additional input keywords may be specified for PCM SCRF calculations.34 Dispersion energy (kcal/mol) = -3. The energy computed by an Onsager SCRF calculation appears in the output file as follows: Total energy (include solvent energy) = -74. For all iterative SCRF methods.2p) SCF=(Tight) SCRF=COSMORS Water generating COSMO-RS input 0 1 o h.u. The total energy in solution is the sum of the SCF energy and all of the non-electrostatic energy terms (both are highlighted in the output).27 -------------------------------------------------------------------Cavitation energy (kcal/mol) = 5. 21 and a value of 70 tesserae per sphere. and so you will probably want to set all of them appropriately. RSOLV=radius Solvent radius in Angstroms. NOREP Skip the calculation of repulsion solute-solvent interaction energy.21 TSNUM=70 This Gaussian job performs a PCM energy calculation on the molecule HF using the solvent cyclohexane. . The following keywords are available for controlling PCM calculations (arranged in groups of related items): SPECIFYING THE SOLVENT The solvent for the PCM calculation may be specified using the normal Solvent option to the SCRF keyword. The calculation is performed at a temperature of 300 K using a scaling factor for all atoms except acidic hydrogens of 1. DENSITY=val Density of the solvent EPSINF=val Optional value for the dielectric constant at infinite frequency.0 ALPHA=1.1 H F 1 R R=0. the others default to the values for water. Note that if any of these parameters are specified. Alternatively. the EPS and RSOLV keywords may be used in the PCM input section to define a solvent explicitly: EPS=e Dielectric constant of the solvent. The solvent name keyword or ID number may also be placed within the PCM input section. The final input section ends as usual with a blank line.9161 TABS=300. CALCULATION METHOD VARIATIONS NODIS Skip the calculation of dispersion solute-solvent interaction energy.0. This analysis involves a fitting of atomic charges to the molecular electrostatic potential in solution. the variation of the dipole moment in solution and so on. SCFVAC Performs the gas phase calculation before that in solution.e. and DCav may be used to include them for the rare cases where the non-electrostatic energy terms are known to affect the geometry. except for COSMO-RS. This is the default. DRep. at "fixed cavity"). but only by HF or DFT methods.NOCAV Skip the calculation of the cavitation energy. Such cases will require care during optimization. MXITER=N Specify the maximum number of iterations allowed to the iterative solution of the electrostatic problem.e. By default. FIXHSS Compute the electrostatic energy second derivatives neglecting the geometrical contributions (i. non-electrostatic energy contributions are computed and printed. FITPOT Performs analysis of the solute solvent interaction energy in terms of atomic or atomic groups additive contributions. but they are not added into the energy and its derivatives during geometry optimizations. NOSCFVAC is the default. ITERATIVE Solve the PCM electrostatic problem to calculate polarization charges through a linear scaling iterative method using a Jacobi-like scheme. It allows for the calculation of ΔGsolvation. Tight (10- . MobGrd is the default. This is the default for COSMO-RS. at "fixed cavity"). The recommended radii for this calculation type are the United Atom Topological Model applied on radii optimized for the HF/631G(d) level of theory (specified with RADII=UAHF). 200 is the default. INVERSION Solve the PCM electrostatic problem to calculate polarization charges through an inversion matrix algorithm. QCONV=type|N Set the convergence threshold for the iterative calculations of the PCM polarization charges to 10-N or to one of the following predefined types: VeryTight (10-12). and the optimization process may be trickier and more lengthy. FIXGRD Compute the electrostatic energy gradients neglecting the geometrical contributions (i. The keywords DDis. MobHss is the default. NODIIS Skip the DIIS algorithm for the iterative solution of the PCM problem when the Jacobi scheme is exploited. CG Set the iterative algorithm to a conjugate gradient. BoxLen=N Set the length in Angstroms of the FMM box. Default convergence values are QConv=Tight for PCM energy calculation and QCONV =VeryTight for PCM energy gradients calculations. ANISOTROPIC AND IONIC SOLVENTS ANISOTROPIC Performs a PCM calculation for anisotropic solvent according to the IEF-PCM formalism. 6. The 3-rank symmetric tensor representing the dielectric constant must be . PRECOND=N Set the preconditioner type for the PCM iterative solution. 6 is the default.0 is the default. FMM is the default. 1 corresponds to a simple Jacobi preconditioner. 0 means no preconditioner. The ICOMP keyword. and the algorithm defaults to Jacobi when it is used. CGS Set the iterative algorithm to a squared conjugate gradient.9 ) and Sleazy (10-6). is no longer needed and is deprecated. 2 is the default. LMax=N Set the degree of the polynomial for the electrostatic potential multipole expansion in the FMM. while 2 is a preconditioner based on the correlation considered only for charges located on the same sphere. formerly used to specify the charge compensation mode. This is the default for CPCM calculations. BiCGS Set the iterative algorithm to a stabilized biconjugate gradient The DIIS option is not allowed with this keyword. MxDIIS=N Number of vectors used in the DIIS extrapolation NoFMM Turn off the use of the Fast Multipole Method in the iterative solution. BONDI: Use the Bondi's atomic radii (explicit hydrogens). These are the recommended radii for for the calculation of ΔGsolvation via the SCFVAC PCM keyword. SPECIFYING THE MOLECULAR CAVITY By default. SPHEREONH=N When using the UA0 model. The ionic strength in mol/dm3 Å2 has to be specified as the value to the keyword DISM. RADII=model Indicates the topological model and/or the set of atomic radii used. places an individual sphere on the hydrogen at the Nth position in the atoms list. the program builds up the cavity using the United Atom (UA0) model. EPSX=value). changing sphere parameters and the general cavity topology.specified as the values for these six additional keywords: EPSX. Available models and sets are: UA0: Use the United Atom Topological Model applied on atomic radii of the UFF force field. adding extra spheres to the cavity built by default. UFF: Use radii from the UFF force field. UAHF: Use the United Atom Topological Model applied on radii optimized for the HF/6-31G(d) level of theory. Hydrogens have individual spheres (explicit hydrogens). The cavity can be extensively modified in the PCM input section: putting spheres around specified hydrogens. UAKS: Use the United Atom Topological Model applied on radii optimized for the PBE0/6-31G(d) level of theory. i.g. EPSZ. EPSY. and EUPSI (all of them take a parameter: e. by putting a sphere around each solute heavy atom: hydrogen atoms are enclosed in the sphere of the atom to which they are bonded. There are three UA models available (see below). This is the default when SCRF=CosmoRS is used.. IONIC Performs a PCM calculation for ionic solution according to the IEF-PCM formalism. and so on. EUTHE. PAULING: Use the Pauling (actually Merz-Kollman) atomic radii (explicit hydrogens). KLAMT: Use atomic radii from the COSMO method. . The whole molecular cavity can be also provided by the user in the input section.e. EUPHI. S. specified on lines of the following format: . Available options are: SES: Solvent Excluding Surface. O.Y. X. in the standard NSPH=N The cavity is built just from the N spheres provided by the user.SPHEREONACIDICH When using the UA0 model.2. SAS: Solvent Accessible Surface. VDW: Van der Waals surface. The radius of the solvent is added to the unscaled radii of atoms and/or atomic groups. put individual spheres on acidic hydrogens (those bonded to N. This is the default for electrostatic contribution. MODIFYSPH Alter parameters for one or more spheres. The surface is generated by the atomic or group spheres and by the spheres created automatically to smooth the surface ("added spheres"). Uses unscaled atomic radii and skip the generation of "added spheres" to smooth the surface.Z are the Cartesian coords. The modified spheres can be indicated in the PCM input using the following format: ModifySph atom_number radius [alpha] EXTRASPH=N Add N user-defined spheres to the cavity. ADDSPH is the default. Cl and F atoms). SURFACE=type Specify the type of molecular surface representing the solute-solvent boundary. ALPHA=scale Specify the electrostatic scaling factor by which the sphere radius is multiplied. NOADDSPH Avoid the generation of added spheres to smooth the cavity surface. Parameters of the spheres can be indicated using the following format: ExtraSph=N X Y Z radius [alpha] orientation. P. The default value is 1. X.89.2 is the default value).org) which can be used to visualize the molecular cavity. SymmCav is the default.4.Y. Increasing this value results in a smaller number of added spheres. Decreasing this index results in a smaller number of added spheres. Reducing this value results in a finer surface discretization. The default value is 0. This files contains input for the GeomView program (see geomview. Values suggested as the best compromise between accuracy and numerical stability range between 0.Z are the Cartesian coords. or even larger for molecular mechanics calculations.off describing the cavity. The default value is 0.2. SMALLTESSERA=value Threshold to discard small tesserae (the default is 10-4 Å2). TSARE=area Set the average area of the tesserae generated on each sphere in the cavity surface.atom_number radius [alpha] X Y Z radius [alpha] orientation. OFAC=value Specify the overlap index between two interlocking spheres [570]. in units of Å2 (area=0. PCMDOC Include the descriptions and values of all the internal PCM parameters in the Gaussian log file. SHORTEDGE=value Threshold to discard short edges in a tessera (the default is 5. in the standard NOSYMMCAV Do not impose the molecular symmetry to the cavity. RMIN=value Set the minimum radius in Angstroms for SES added spheres.2 and 0.0*10-7 Angstroms). OUTPUT OPTIONS GEOMVIEW Create the file tesserae. SP . Tight Sets the cutoff to 1 * 10-10. Loose Sets the cutoff to 5 * 10-5. Hartree-Fock and DFT methods (closed shell calculations).This calculation type keyword requests a single-point energy calculation. This keyword may also be used within method specifications for ONIOM layers. FMM . Medium Sets the cutoff to 5 * 10-7. Sparse Use sparse matrix storage for performance enhancement of large calculations (above around 400 atoms) [34]. Energies and gradients for AM1. All methods. This is the default for semi-empirical methods. N Sets the cutoff to 1 * 10-N. This is the default for DFT methods. It is useful for AM1 calculations of more than 200 atoms. See the discussion of the methods keywords for examples of their energy output formats. The keyword's option allows you to specify the cutoff value for considering matrix elements to be zero. It is the default when no calculation type keyword is specified. The code which checks for a complex stability (Link 902) is older and less reliable and should not be used unless complex orbitals are of interest. Gaussian has the ability to test the stability of a singledeterminant wavefunction with respect to relaxing various constraints [106. RRHF Constrain the wavefunction testing or reoptimization to be real. Synonymous with Singlet. By default. Note that analytic frequency calculations are only valid if the wavefunction has no internal instabilities. only real instabilities (i. an internal instability only affects the validity of the results if the pairs of orbitals mixed are of the same spatial symmetry.. spin-restricted. These include: • • • Allowing an RHF determinant to become UHF. In examining the results prior to a MøllerPlesset calculation. In examining the results prior to a frequency calculation. Int Test for internal instability (a lower determinant with the same constraints) only.Stable This calculation type method requests that the stability of the Hartree-Fock or DFT wavefunction be tested. The Stable keyword causes the program to compute a wavefunction as usual and then to determine if the resulting determinant is a local minimum with the specified degrees of freedom taken into consideration. . it suffices to see if any singlet instabilities exist for restricted wavefunctions or if any instabilities (singlet or triplet) exist for unrestricted wavefunctions. The default is to test for all instabilities but not to re-optimize the wavefunction.e. Reducing the symmetry of the orbitals. The validity of restricted Møller-Plesset energies based on wavefunctions which are unstable with respect to becoming UHF is also questionable [571]. not complex) are sought. Allowing orbitals to become complex.107] (see also [560]). If Stable=Opt is specified. and Møller-Plesset calculations are only valid if the wavefunction has no internal instabilities within the constrained symmetry. by default the wavefunction is allowed to be unrestricted if necessary. GENERAL OPTIONS RExt Test for external real instability as well as internal instability (the default). 1Opt Redo the SCF once if an instability is detected.e. reoptimize the wavefunction with the appropriate reduction in constraints. WAVEFUNCTION REOPTIMIZATION OPTIONS Opt If an instability is found. Synonymous with Triplet. CUHF Allow testing for real to complex instabilities in spin-unrestricted wavefunctions. in the MO basis). Restart Restarts the calculation off the checkpoint file. spin-unrestricted. AO Forces a calculation using the AO integrals (written to disk). NoOpt prevents reoptimization and is the default. Also implies SCF=Restart. It is the default when SCF=Conven is also specified..RUHF Constrain the wavefunction testing or reoptimization to be real. RepOpt is a synonym for Opt. avoiding an integral transformation. ICDiag Forces in-core full diagonalization of the matrix formed in memory from transformed integrals. ALGORITHM-RELATED OPTIONS Direct Forces a direct calculation (the default). CRHF Allow testing for real to complex instabilities in spin-restricted wavefunctions. . The AO basis is seldom an optimal choice. except for small molecules on systems having very limited disk. InCore Forces an in-core algorithm. MO Forces a stability calculation using transformed two-electron integrals (i. repeating stability tests and reoptimizations until a stable wavefunction is found. It implies the use of MO integrals. Loose Tells the program to use looser cutoffs in determining symmetry at the first point. SCF Symmetry This keyword specifies the uses of molecular symmetry within the calculation. Int Int enables and NoInt disables use of integral symmetry (use of the "petite list"). and input for properties and background charge distributions is required in the standard orientation. By default. symmetry is used wherever possible to reduce CPU.HF and DFT methods. Derivatives are then rotated back to the original (input) orientation. called the standard orientation. Grad NoGrad disables and Grad enables use of symmetry in integral derivative evaluation. If symmetry is in use. Symmetry use can be completely disabled by NoSymm. Synonymous with Int=[No]Symm. Tight says to use the regular criteria at the first point. which is used by default only for GVB calculations. The NoSymmetry keyword prevents the reorientation and causes all computations to be performed in the Z-matrix orientation. disk storage. It is designed for use with suboptimal input geometries. the molecule may be rotated to a different coordinate system. or modified by the Symm keyword and one or more options. and it is the default. SCF=NoSCF is equivalent to Guess=LowSym and combining all irreducible representations together. . and I/O requirements. Orbitals are printed in the standard orientation. SCF NoSCF disables and SCF enables use of N3 symmetry in SCF. before the calculation is performed. Follow Try to follow point group/orientation during optimization. so it should only be used if you know what you're doing! Int. 50-50 Solve for half triplet and half singlet states. SCF TD This method keyword requests an excited state energy calculation using the timedependent Hartree-Fock or DFT method [109. such as with massage.111].258. The default is the first excited state (N=1). Electronic circular dichroism (ECD) analysis is also performed during these calculations [255.110. NStates=M Solve for M states (the default is 3). for which it is the default.. On Turn on symmetry when it would otherwise be off. . Note that the normalization criteria used is <X+Y|X-Y>=1. Only effective for closed-shell systems. If 50-50 is requested.259.e. Triplets Solve only for triplet excited states. the default is 3 singlets and 3 triplets).260] Singlets Solve only for singlet excited states. Only effective for closed-shell systems.256. Axis=[X|Y|Z] Specify axis to help specify subgroup.257. NStates gives the number of each type of state for which to solve (i.PG=group Use no more symmetry than that found in the specified point group. Only effective for closed-shell systems. Root=N Specifies the state of interest. This can cause wrong answers. This is the default for TD Hartree-Fock. Note that. Optimizations are available using numerical gradients.0356 . Zero frequency is always done and need not be in the list. and it is the default for TD-DFT.68197 This state for optimization and/or second-order correction. Energies using Hartree-Fock or a DFT method. etc. all excited states are solved for.70318 6. NonEqSolv is the default. NoIVOGuess forces the use of canonical single excitations for guess.1280 eV 300. By default. Output Here is the key part of the output from a TD excited states calculation: Excitation energies and oscillator strengths: Excited State 1: Singlet-A2 4. SOS Do sum-over states polarizabilities. Copying the excited state density for this state as the 1-particle RhoCI density. Excited State 2: 8 -> 10 Singlet-B2 0.4912 eV 191.00 nm f=0. A list of frequencies at which to do the sums is read in. Read Reads initial guesses for the states off the checkpoint file. This option implies Read as well. unlike for SCF. CIS.Add=N Read converged states off the checkpoint file and solve for an additional N states. ZINDO. The HFIVOGuess option forces the use of Hartree-Fock IVOs for the guess. IVOGuess Force use of IVO guess. an initial guess for one basis set cannot be used for a different one. even for TDDFT. EqSolv Whether to perform equilibrium or non-equilibrium PCM solvation.35 nm f=0.0000 8 -> 9 0. which is the default.0007 -0. the excitation energy.0200 1. The value should be specified as an option: # . Rotatory Strengths (R) in cgs (10**-40 erg-esu-cm/Gauss) state X Y Z R(length) 1 -0. The ECD results appear in the output as follows: <0|del|b> * <b|rxdel|0> (Au). Archive.4378 eV 166.0007 5. and (on the second line for each state) the largest coefficients in the CI expansion.Excited State 3: 8 -> 11 Singlet-A1 0.0045 -0. Rearchive .3201 <0|r|b> * <b|rxdel|0> (Au).0007 -0. Its antonym is Archive.0043 1. Temperature=300 The default is 298.0017 -0.0068 3 0.0111 -0.3067 Temperature Specifies the temperature to be used for thermochemistry analysis (in Kelvin)... including the spin and spatial symmetry.0541 The results on each state are summarized.0001 5.7826 2 0.9442 3 -0.0193 0. Rotatory Strengths (R) in cgs (10**-40 erg-esu-cm/Gauss) state X Y Z R(velocity) 1 0.6444 2 -0. the oscillator strength.70219 7. Note that archive entries may be extracted from Gaussian log files after the fact using the pluck utility.15 K.0024 0. Test This keyword suppresses the automatic creation of an archive entry (formerly intended for the Browse Quantum Chemistry Database System).0018 -2.0300 0.0083 -3.0048 0.0004 0.0034 0.69 nm f=0.0040 -0. TestMO The cutoffs used in computing and storing integrals and the convergence criteria applied in SCF and CPHF calculations are appropriate for most molecules and basis sets. (Note that this corresponds to a coefficient of 1012 for the contribution of an AO integral to an MO integral involving four virtual orbitals. By default CPHF and post-SCF calculations are aborted if any MO coefficient is larger than 1000. INTEGRAL TRANSFORMATION ALGORITHM OPTIONS Direct Requests that the direct transformation routines be used. This is the default. TestMO is the default. It should be used only after careful thought. InCore Forces use of the in-core algorithm in Link 804 . if a nearly linearly dependent basis set is used. and semi-direct methods automatically. TrackIO This keyword requests routine-by-routine statistics of I/O and CPU usage. Link 804 will select between the in-core. fully direct. very large MO coefficients may occur and in combination with the finite accuracy of other terms lead to substantial numerical errors.) The NoTestMO keyword suppresses this check. Equivalent to L804. as well as the types of transformed integrals produced. However. #P Transformation This keyword controls the algorithm used for integral transformation. IJKA Produce <IJ||AB>. <IJ||KL>. Molecular Mechanics Methods . IAJB Produce <IJ||AB> and <IA||JB> integrals. IJAB Produce only <IJ||AB> integrals. <IA||JB>. <IJ||KA>.FullDirect Forces use of the fully direct (MO integrals in core) method in Link 804. including transformed integrals involving all virtuals). SemiDirect Forces use of the semi-direct algorithm in Link 804. New2PDM Causes the 2PDM to be generated. This option cannot be used for frozen-core calculations. INTEGRAL SELECTION OPTIONS Full Forces a transformation over all orbitals (i.. NoDirect is a synonym for Conventional. This is slow. <IJ||KL>. and <IJ||KA> integrals. and it must be used for frozen-core calculations. used. This is the default and fastest method. <IA||JB>. and <IA||BC> integrals.e. and discarded by L1111 in post-SCF gradient calculations. <IA||JB>. but it reduces memory requirements. Old2PDM Forces the old-fashioned process of the 2PDM in post-SCF gradients (sorted in L1111 and then processed in L702 and L703). Conventional Requests that the original transformation method based on externally stored integrals be used. This was the only choice in Gaussian 90 and earlier versions. IJKL Produce <IJ||AB>. and <IJ||KL> integrals. ABCD is a synonym for Full. IABC Produce <IJ||AB>. DREIDING: The DREIDING force field as described in [38]. Hence. with hard-wired parameters having priority over the read-in. when no relevant option is given. Soft parameters are ones specified by the user in the input stream for the current job (or a previous job when reading parameters from the checkpoint file). By default.. PARAMETER PRECEDENCE OPTIONS Terminology: Gaussian contains built-in parameter sets for the built-in force fields listed above. no charges are assigned to atoms by default when using any molecular mechanics force field. UnTyped Assign QEq charges only to those atoms for which the user did not specify a particular type in the input. HardFirst Read additional parameters from the input stream. these are referred to as hard-wired parameters. We use this current version from the AMBER web site (amber. read-in parameters are used only if there is no .dat) have been updated slightly since the publication of this paper. UFF: The UFF force field as described in [39].edu). Options are available to estimate charges at the initial point using the QEq algorithm under control of the following options for any of the mechanics keywords: QEq Assign charges to all atoms using the QEq method [40]. The following force fields are available: AMBER: The AMBER force field as described in [37]. No basis set keyword should be specified with these keywords.There are three molecular mechanics methods available in Gaussian. CHARGE ASSIGNMENT-RELATED OPTIONS Unless set in the molecule specification input. UnCharged Assign QEq charges for all atoms which have charge zero (i. the hard-wired parameters are the only ones used. but they are also available as independent methods.scripps.e. soft ones. They were implemented for use in ONIOM calculations. all atoms which were untyped or which were given a type but not a charge in the input). The actual parameters (parm96. use the last one found. . unless HardFirst is also specified. HANDLING MULTIPLE PARAMETER SPECIFICATION MATCHES Since parameters can be specified using wildcards. with soft (read-in) parameters having priority over the hard-wired values. Use SoftFirst if you want to override hard-wired parameter matches. Note that wildcards matches within the hardwared parameter set take precidence over soft parameters. Any non-standard (soft) parameters present in the checkpoint file are used with higher priority than corresponding hard-wired parameters. Modify Read modifications and additions to the parameter set (after it has been constructed from hard and/or soft parameters). FirstEquiv If there are equivalent matches for a required parameter. even when the latter contains an exact match for the same item. LastEquiv If there are equivalent matches for a required parameter. ChkParameters Read parameters from the checkpoint file. ignoring hard-wired parameters. INPUT CONVENTIONS AMBER calculations require that all atom types be explicitly specified using the usual notation within the normal molecule specification section: C-CT Specifies an SP3 aliphatic carbon atom. C-CT-0.32.32 Specifies an SP3 aliphatic carbon atom with a partial charge of 0. NewParameters Ignore any parameters in the checkpoint file. The default is to abort if there are any ambiguities in the force field. SoftOnly Read parameters from the input stream and use only them.corresponding hard-wired value. use the first one found. SoftFirst Read additional parameters from the input stream. The following options specify other ways of dealing with multiple matches. it is possible for more than one parameter specification to match a given structure. distances are in Angstroms. First. using a factor 0.0 for pairs that are separated by three bonds).O-O--0. gradients. we calculate the interactions between all pairs. angles are in degrees. and some value between 0. the program will attempt to determine atom types automatically. ONIOM. Since this involves only pairs that are close to each other based on the connectivity. we subtract out the contributions that should have been scaled. we can use computationally efficient (linear scaling) algorithms. and frequencies. Consult the AMBER paper [37] for definitions of atom types and their associated keywords. . R refers to distances and θ refers to angles. In MM force fields. In equations.5. but they are not required. the computer time for this step scales again linearly with the size of the system. Analytic energies.0 and 1. They are indicated by a 0 or an asterisk. In this step. However. Geom=Connect GENERAL MOLECULAR MECHANICS FORCE FIELD SPECIFICATIONS Unless otherwise indicated. There are a number of ways to implement the calculation of non-bonded interactions. interactions between pairs of atoms that are separated by three bonds or less are usually scaled down (in most force fields. but were included in the first step. We follow a two-step procedure. without taking the scaling into account. Wildcards may be used in any function definition.5 Specifies a carbonyl group oxygen atom with a partial charge of -0. Atom types and charges may also be provided for UFF and DREIDING calculations. Function equivalencies to those found in standard force fields are indicated in parentheses. the overall algorithm is the more efficient than the alternatives.0 for pairs separated by one or two bonds. Although at first sight it seems that too much work is done. energies are in Kcal/mol and charges are in atomic units. In the second step. the non-bonded (Vanderwaals and electrostatic) interactions are evaluated for every possible pair of atoms. For these methods. 0. NE Slater-Kirkwood effective number of valence electrons (dimensionless). used for NBDir and NBTerm (See MMFF94 below for MMFF94-type Vanderwaals parameters). This function will be expanded into pairs and a direct function (NBDir and NBTerm) before evaluation of the MM energy.In the soft force field input. However. you can specify just the non-bonded master function NonBon.0 for acceptor type. the NBDir function entry corresponds to the calculation of all the pairs. and the NBTerm entry is used for the subsequent subtraction of the individual pairs. Vanderwaals parameters. Scale2 Scale factor (dimensionless). which is automatically expanded into the actual functions NBDir and NBTerm during pre-processing. Scale1 Scale factor (Angstrom1/4). otherwise 0. NonBon V-Type C-Type. MMFF94 electrostatic buffering Buf94 Atom-type Value Non-bonded interaction master function. VDW94 Atomic-pol NE Scale1 Scale2 DFlag Atomic-pol Atomic polarizability (Angstrom3). 2. DFlag 1. V-Cutoff C-Cutoff VScale1 VScale2 VScale3 CScale1 CScale2 CScale3 V-Type is the Vanderwaals type: 0 No Vanderwaals 1 Arithmetic (as for Dreiding) 2 Geometric (as for UFF) 3 Arithmetic (as for Amber) 4 MMFF94-type Vanderwaals C-Type is the Coulomb type: 0 No Coulomb 1 1/R 2 1/R2 3 1/R buffered (MMFF94) . to make things easier.0 for donor type atom. VDW Bond-length Well-depth MMFF94 type Vanderwaals parameters (used for NBDir and NBTerm). 2 scaling is used (as for Amber). C-Type. C-Type. V-Scale. and C-Cutoff as above. If any scale factor < 0.V-Cutoff and C-Cutoff are the Vanderwaals and Coulomb cutoffs (respectively): 0 No cutoff >0 Hard cutoff <0 Soft cutoff VScale1-3 are Vanderwaals scale factors for 1 to 3 bond separated pairs. NBDir V-Type C-Type V-Cutoff C-Cutoff V-Type. CScale1-3 are Coulomb scale factors for 1 to 3 bond separated pairs. C-Cutoff. Coulomb and Vanderwaals direct (evaluated for all atom pairs). the 1/1. Atomic single bond radius AtRad Atom-type Radius Effective charge (UFF) EffChg Charge GMP Electronegativity (UFF) EleNeg Value Step down table Table Original-atom-type Stepping-down-type(s). Coulomb and Vanderwaals single term cutoffs NBTerm Atom-type1 Atom-type2 V-Type C-Type V-Cutoff C-Cutoff V-Scale C-Scale V-Type. V-Cutoff. and C-Scale as above.0. Harmonic stretch I (Amber [1]): ForceC*(R-Req)2 . V-Cutoff. Harmonic stretch III (UFF [1a]): k*(R-Rij)2 Equilibrium bond length Rij = (1 . Zi and Zj are the effective atomic charges defined with EffChg. Morse stretch I (Amber): DLim*(e-a(R-Req)-1)2 where a = Sqrt(ForceC/DLim) MrsStr1 Atom-type1 Atom-type2 ForceC Req DLim ForceC Force constant Req Equilibrium bond length DLim Dissociation limit Morse stretch II (Dreiding [5a]): DLim*exp[-a(Ri+Rj-Delta)]-1)2 where a = Sqrt(ForceC/DLim) MrsStr2 Atom-type1 Atom-type2 ForceC Delta DLim .PropC*lnBO)*(Ri + Rj) + Ren Force constant: k = 664. Xi and Xj are GMP electronegativity values defined with EleNeg.Sqrt(Xj)]2/(Xi*Ri + Xj*Rj) HrmStr3 Atom-type1 Atom-type2 BO PropC BO Bond order (if <0. it is determined on-the-fly) PropC Proportionality constant Ri and Rj are atomic bond radii defined with AtRad.HrmStr1 Atom-type1 Atom-type2 ForceC Req ForceC Force constant Req Equilibrium bond length Harmonic stretch II (Dreiding [4a]): ForceC*[R-(Ri+Rj-Delta)]2 HrmStr2 Atom-type1 Atom-type2 ForceC Delta ForceC Force constant Delta Delta Ri and Rj are atomic bond radii specified with AtRad.12*Zi*Zj/(Rij3) Electronegativity correction: Ri*Rj*[Sqrt(Xi) . Xi and Xj are GMP electronegativity values defined with EleNeg. Morse stretch III (UFF [1b]): A1*A3*(exp[-a(R-Rij)]-1)2 where a = Sqrt(k/[BO*PropC]) Equilibrium bond length Rij = (1 . Zi and Zj are the effective atomic charges defined with EffChg. it is determined on-the-fly) PropC Proportionality constant Ri and Rj are atomic bond radii defined with AtRad.12*Zi*Zj/Rij3 Electronegativity correction: Ren = Ri*Rj*(Sqrt(Xi) . Quartic stretch I (MMFF94 [2]): (Req/2)*(R-ForceC)2*[1+CStr*(R-ForceC+(7/12)*CStr2*(R-ForceC)2] QStr1 Atom-type1 Atom-type2 ForceC Req CStr ForceC Force constant (md-Angstrom-1) Req Equilibrium bond length (Angstrom) CStr Cubic stretch constant (Angstrom-1) Atomic torsional barrier for the oxygen column (UFF [16]) UFFVOx Barrier Atomic sp3 torsional barrier (UFF [16]) UFFVsp3 Barrier Atomic sp2 torsional barrier (UFF [17]) UFFVsp2 Barrier .Sqrt(Xj))2/(Xi*Ri + Xj*Rj) MrsStr3 Atom-type1 Atom-type2 BO PropC BO Bond order (if <0.PropC*lnBO)*(Ri + Rj) + Ren Force constant k = 664.ForceC Force constant Delta Delta DLim Dissociation limit Ri and Rj are atomic bond radii defined with AtRad. Xj and Xk are GMP electronegativity defined with EleNeg. Xi. C1 = -4*C2*cos(θeq) and C0=C2*(2*cos(θeq2)+1) Force constant: k = 664. it is determined on-the-fly) PropC Proportionality constant Ri. it is determined on-the-fly) BO23 Bond order for Atom-type2–Atom-type3 (when <0.12*Zi*Zk*(3*Rij*Rjk*(1-cos(θeq2))-cos(θeq)*Rik2)/Rik5 UFFBnd3 Atom-type1 Atom-type2 Atom-type3 θeq BO12 BO23 PropC θeq Equilibrium angle BO12 Bond order for Atom-type1–Atom-type2 (when <0. UFF 2-term bend (UFF [10]): [k/(Per2)]*[1-cos(Per*θ)] Force constant: k = 664.Harmonic bend (Amber [1]): ForceC*(T-θeq)2 HrmBnd1 Atom-type1 Atom-type2 Atom-type3 ForceC θeq ForceC Force constant (in kcal/(mol*rad2) θeq Equilibrium angle Harmonic Bend (Dreiding [10a]): [ForceC/sin(θeq2)]*(cos(θ)-cos(θeq))2 HrmBnd2 Atom-type1 Atom-type2 Atom-type3 ForceC θeq ForceC Force constant θeq Equilibrium angle Dreiding Linear Bend (Dreiding [10c]): AForceC*(1+cos(θ)) LinBnd1 Atom-type1 Atom-type2 Atom-type3 ForceC ForceC Force constant UFF 3-term bend (UFF [11]): k*(C0 + C1*cos(θ))+C2*cos(2θ) where C2=1/(4 * sin(θeq2)). Rj and Rk are atomic bond radii defined with AtRad.12*Zi*Zk*(3*Rij*Rjk*(1-cos(Per2))-cos(Per)*Rik2)/Rik5 . Zj and Zk are effective atomic charges defined with EffChg. Zi. 4 for square-planar. 3 for trigonal.. Zi. Zero bend term: used in rare cases where a bend is zero. Rj and Rk are atomic bond radii defined with AtRad. Zj and Zk are effective atomic charges defined with EffChg. Xi. it is determined on-the-fly) PropC Proportionality constant Ri. Dreiding torsion (Dreiding [13]): V*[1-cos(Period*(θ-PO))]/(2*NPaths) . Xj and Xk are GMP electronegativity defined with EleNeg.. determined on-the-fly).Mag4 V/2 magnitudes NPaths Number of paths (if < 0.4 (Magi*[1+cos(i*θ-I(i+4))])/NPaths AmbTrs Atom-type1 A-type2 A-type3 A-type4 PO1 PO2 PO3 PO4 Mag1 Mag2 Mag3 Mag4 NPaths PO1-PO4 Phase offsets Mag1.UFFBnd2 Atom-type1 Atom-type2 Atom-type3 Per BO12 BO23 PropC Per Periodicity: 2 for linear. ZeroBnd Atom-type1 Atom-type2 Atom-type3 Cubic bend I (MMFF94 [3]): (ForceC/2)*(1+CBend*(θ-θeq))*(θ-θeq)2 CubBnd1 Atom-type1 Atom-type2 Atom-type3 ForceC θeq CBend ForceC Force constant (in md*Angstrom/rad2) θeq Equilibrium angle CBend "Cubic Bend" constant (in deg-1) MMFF94 Linear Bend (MMFF94 [4]): ForceC*(1+cos(θ)) LinBnd2 Atom-type1 Atom-type2 Atom-type3 ForceC ForceC Force constant (md) Amber torsion (Amber [1]): Σi=1. BO12 Bond order for Atom-type1–Atom-type2 (when <0. it is determined on-the-fly) BO23 Bond order for Atom-type2–Atom-type3 (when <0. This term is needed for the program not to protest about undefined angles. determined on-the-fly. UFF torsion with constant barrier height (UFF [15]): [V/2]*[1cos(Period*PO)*cos(V*θ)]/NPaths UFFTorC Atom-type1 Atom-type2 Atom-type3 Atom-type4 Period PO V NPaths Period Periodicity PO Phase offset V Barrier height V NPaths Number of paths. UFF torsion with atom type-based barrier height (UFF [16]): [V/2]*[1-cos(Period*PO)* cos(Period*θ)]/NPaths where V=Sqrt(Vj*Vk) UFFTor1 Atom-type1 Atom-type2 Atom-type3 Atom-type4 Period PO NPaths Period Periodicity PO Phase offset NPaths Number of paths. . it is determined on-the-fly) NPaths Number of paths (when <0. determined on-the-fly).18*Log(BO12)] UFFTorB Atom-type1 Atom-type2 Atom-type3 Atom-type4 Period PO BO12 NPaths Period Periodicity PO Phase offset BO12 Bond order for Atom-type1–Atom-type2 (when <0. determined on-the-fly.DreiTrs Atom-type1 Atom-type2 Atom-type3 Atom-type4 V PO Period NPaths V Barrier height V PO Phase offset Period Periodicity NPaths Number of paths (if < 0. When zero or less. Vj and Vk are atomic constants defined with UFFVsp3. When zero or less. UFF torsion with bond order based barrier height (UFF [17]): [V/2]*[1-cos(Period*PO)* cos(Period*θ)]/NPaths where V = 5*Sqrt(Uj*Uk)*[1+4. it is determined on-the-fly) Uj and Uk are atomic constants defined with UFFVsp2. 0.0. PO=0. Period=6.0. C3 Coefficients Harmonic Wilson angle (MMFF94 [6]): (ForceC/2)*(θ2) summed over all three Wilson angles θ.0. with the following parameters: • • • If there are three atoms bonded to the third center and the fourth center is H. Vj and Vk are atomic constants from UFFVOx.0. OldTor Atom-type1 Atom-type2 Atom-type3 Atom-type4 Improper torsion (Amber [1]): Mag*[1+cos(Period*(θ-PO))] ImpTrs Atom-type1 Atom-type2 Atom-type3 Atom-type4 Mag PO Period Mag V/2 Magnitude PO Phase offset Period Periodicity Three term Wilson angle (Dreiding [28c]. but the fourth center is not H.0. Wilson Atom-type1 Atom-type2 Atom-type3 Atom-type4 ForceC C1 C2 C3 ForceC Force constant C1. PO=0. then these values are used: V=4. UFF [19]): ForceC*(C1 + C2*cos(θ) + C3*cos(2θ)) averaged over all three Wilson angles θ. these values are used: V=1.0. . and NPaths=-1. determined on-the-fly. Otherwise. When zero or less. During processing. it is removed. and at least one of them is H. Dreiding special torsion for compatibility with Gaussian 98 code.0. it is replaced with DreiTRS.UFF torsion with atom type based barrier height (UFF [16]) (differs from UFFTor1 in that the atomic parameter that is used): [V/2]*[1cos(Period*PO)*cos(Period*θ)]/NPAths where V=Sqrt(Vj*Vk) UFFTor2 Atom-type1 Atom-type2 Atom-type3 Atom-type4 Period PO NPaths Period Periodicity PO Phase offset NPaths Number of paths. If there are three atoms bonded to the third center. and NPaths=-1. Period=3. C2. 50 Double bond: 1...50 Triple central bond: bond order ≥ 2. Req23 Equilibrium bond lengths θeq Equilibrium angle USING SUBSTRUCTURES Substructures may be used to define different parameter values for a function for distinct ranges of some geometrical characteristic. For dihedral angles.00 ≤ bond order < 1. one or two substructures may be used (e. HrmStr-1. Substructure numbers are appended to the function name.00 ≤ bond order < 1. separated by a hyphen (e. HrmStr-2 and so on).50 Double central bond: 1.50 Triple bond: bond order ≥ 2.50 . Use a zero for the first substructure to specify only the second substructure.50 ≤bond order < 2.50 ≤ bond order < 2.HrmWil Atom-type1 Atom-type2 Atom-type3 Atom-type4 ForceC ForceC Force constant Stretch-bend I (MMFF94 [5]): (ForceC1*(R12-Req12)+ForceC2*(R32-Req23))*(θ-θeq) StrBnd1 Atom-type1 Atom-type2 Atom-type3 ForceC1 ForceC2 Req12 Req23 θeq ForceC1. The following substructures apply to functions related to bond stretches: • • • -1 -2 -3 Single bond: 0.g.50 The following substructures apply to functions for bond angles (values in degrees): First substructure: • • • -1 -2 -3 0 ≤ θ ≤ 45 45 < θ ≤ 135 135 < θ ≤ 180 Second substructure: • -i-n Number of atoms bonded to the central one. AmbTrs-1-2). First substructure: • • • • -0 -1 -2 -3 Skip this substructure (substructure "wildcard") Single central bond: 0. ForceC2 Force constants (in md/rad) Req12.g. 0 1. and their input is always interpreted in units of Angstroms and degrees. Cube and Massage keywords are not affected by the setting of the Units keyword.0 -1.0 120. RESTRICTIONS The Charge.0 -1.0 180. AU Distances are in atomic units (Bohrs). The defaults are Angstroms and degrees.40 C_2 * 50.70) Amide central bond (priority over resonance) None of the above Here is some simple MM force field definition input: HrmStr1 HrmStr1-1 HrmStr1-2 HrmBnd2 DreiTrs-1 DreiTrs-2 H_ C_2 C_2 * * * C_2 360. such as step-sizes in numerical differentiation.0 2. .0 Units The Units keyword controls the units used in the Z-matrix for distances and angles and related values.50 C_2 500.0 1.0 C_2 C_2 * 5.0 180. Ang Distances are in Angstroms (this is the default).0 1. Rad Angles are in radians.0 C_2 C_2 * 45. Deg Angles are in degrees (the default).08 C_2 350.0 2.Second substructure: • • • -i-1 -i-2 -i-3 Resonance central bond (1.30 ≤ bond order ≤ 1. CIS.95]. By default. all DFT methods. CCSD and QCISD.001 electrons/bohr3 density. Since other. CISD.Volume This keyword requests that the molecular volume be computed. the computed volume is only accurate to about two significant figures. MP3. selected with the W1U keyword.5Å larger than the radius corresponding to the computed volume) is printed in the output. but this is sufficient to estimate a radius for use with the Onsager solvent reaction field model. The first. This is the W1 method modified to use UCCSD instead of ROCCSD for open shell systems. MP2. The density to be used can be specified with the Density keyword. The recommended radius (which is 0. SCRF=Dipole W1U W1BD These method keywords request two variations of the W1 method of Martin [94. CCD. CID. defined as the volume inside a contour of 0. is the W1U method. Hartree-Fock. the volume is computed to an accuracy of about 10%. W1BD requests a related method which substitutes BD for coupled cluster [96]. . MP4(SDQ). This method is both more expensive and more accurate than CBS-QB3 and G3. Tight Requests an increased density of points for more accurate integration. Since a Monte-Carlo integration is done. more accurate solvent models are available in Gaussian 03. this keyword has applicability only in preparation for frequency calculations using SCRF=Dipole. Use of this option is recommended if the computed molecular volume is needed more quantitatively. 99916).458287 -76.479709 Temperature= E(ZPE)= W1 (0 K)= W1 Enthalpy= The predicted energy is given followed by values for the thermochemistry analysis. 1 atmosphere.g. arranged in the same order as they appeared in the molecule specification section.462067 -76. ReadIsotopes Specify alternate temperature. ZINDO . However. these values must be real numbers. and/or isotopes (the defaults are 298. Calculation Summary Output.150000 0... Here is the key part of the output from a W1U calculation: W1 Electronic Energy 298. isotope mass for atom n Must be real numbers.483031 Pressure= 1.15 K. and Gaussian uses the value 17. Gaussian prints a table of results for these methods. After all of the output for the component job steps. This information appears in a separate input section having the format: temp pressure [scale] isotope mass for atom 1 isotope mass for atom 2 . and an optional scale factor for frequency data when used for thermochemical analysis (the default value defined by the specified method is used if scale is omitted or set to 0.You should specify alternative isotopes for W1 jobs using the standard method.020965 -76. where temp.. pressure. the program will automatically use the corresponding actual exact mass (e. 18 specifies O18.000000 E(Thermal)= 0. pressure. and the most abundant isotopes). If integers are used to specify the atomic masses. the ReadIsotopes option is retained for rerunning completed calculations under different conditions. The remaining lines hold the isotope masses for the various atoms in the molecule.459231 W1 Free Energy= -76.0). pressure. Restart Restart an incomplete W1 calculation.023800 W1 Energy= -76. and scale are the desired temperature. 113. A value of zero indicates the first or last orbital. ." The default is the first excited state (N=1).This method keyword requests an excited state energy calculation using the ZINDO/S method [112. depending on where it is used.117. Window=(m[. then the highest |m| occupied and lowest |m| virtual orbitals are retained. Only effective for closed-shell systems.118. Root=N Specifies the "state of interest. By default.119. for which it is the default.e. NStates=M Solve for M states (the default is 3). then the highest n orbitals are frozen. NStates gives the number of each type of state for which to solve (i. the value for the last orbital is negative (-n). Use the Window option to define a different orbital set. If 50-50 is requested. If m is negative and n is omitted. n defaults to 0. Energies only. then the highest m orbitals are retained.114.115.120]. a ZINDO calculation is performed using the ten highest occupied orbitals and the ten lowest virtual orbitals. Only effective for closed-shell systems. Note that ZINDO calculations must not specify a basis set keyword.116..n]) The two values specify the starting and ending orbitals to be used. Add=N Read converged states off the checkpoint file and solve for an additional N states. The Density keyword is ignored for ZINDO calculations. If the value for the first orbital is negative (-m). 50-50 Solve for half triplet and half singlet states. Only effective for closed-shell systems. If m is positive and n is omitted. Triplets Solve only for triplet excited states. the default is 3 singlets and 3 triplets). Singlets Solve only for singlet excited states. spec may take on either of the forms used for the Read-Write file (described above). Link 0 commands may be up to 500 characters in length. .loc2.size2. MW or GW (with no intervening spaces) to indicate units. and a value of 0 indicates that an existing segment should retain its current size. which are optional and precede the route section if present. . For example. %KJob LN [M] Tells the program to stop the run after the Mth occurrence of Link N. KW. %Int=spec Locates and names the two-electron integral file(s). MB. the value may be optionally followed by KB. GB.size1. %Mem=N Sets the amount of dynamic memory used to N words (8N bytes). %KJob L502 2 will cause the run to terminate after Link 502 has been run for the second time. Note that directory specifications must include terminal slashes (on UNIX systems). Each location is followed by a maximum size for the file segment at that location. The locations may be either directory locations.. %Chk=file Locates and names the checkpoint file. %RWF=file Locates and names a single. %D2E=spec Locates and names the two-electron integral derivative file(s). it defaults to 1. The default is 6MW. N may be optionally followed by a units designation: KB. or full pathnames. The default units for each size is words. TD Gaussian 03 Online Manual Last update: 10 October 2003 This section lists all Link 0 commands. See this page for a more detailed discussion of the scratch file naming commands. %RWF=loc1. spec may take on either of the forms used for the Read-Write file (described above).. An alternate syntax is provided for splitting the Read-Write file among two or more disks (or file systems). KW. MB. A value of -1 for any size parameter indicates that any and all available space may be used. GB. M may be omitted.CIS. MB or GW. unified Read-Write file (old-style syntax). %NProcLinda=N Requests that the job use up to N processors for distributed memory parallel execution. then one processor is used. and the %NProcShared Link 0 command is used to override this local default (e.. the number of processors to use in production runs is usually set in the Default. including any files that were named explicitly following this directive. if a file is named before %NoSave is encountered. the file will be retained. On parallel machines. and no default is provided in the Default. to run debug jobs on a single processor even if the default is to use 4 processors).Route file.. these commands specify a name for the checkpoint file. to run debug jobs on a single processor even if the default is to use 4 processors). The directory specification should be in the usual . This capability is only available on some computer systems. %NoSave Causes Link 0 to delete scratch files at the end of a run. if the % directive naming the file appears after the %NoSave directive. For example %SUBST L913 /user/chem will cause /user/chem/l913.g. and Gaussian must have been built with parallel processing enabled. and an alternate name and directory location for the read-write file. In other words. If both %Save and %NoSave are specified. and Gaussian must have been built with parallel processing enabled. and cause only the checkpoint file to be saved at the conclusion of the Gaussian job: %RWF=/chem/scratch2/water %NoSave %Chk=water Files to be deleted go here.Route file. and no default is provided in the Default. However. %NProcShared=N Requests that the job use up to N processors for shared memory parallel execution on SMP multiprocessor computers.Route file. then the one appearing latest in the input file takes precedence. By default. %Save Causes Link 0 to save scratch files at the end of the run. and the %NProcLinda Link 0 command is used to override this local default (e. the number of processors to use in production runs is usually set in the Default. On parallel machines. Files to be saved go here. If %NProcShared is not used. all non-specified scratch files are deleted and all named scratch files are saved when the run completes successfully. This capability is only available on some computer systems.exe file) for a link from an alternate directory. then one processor is used. If %NProcLinda is not used.exe to be run instead of the default executable (in $g03root). it will not be saved. %Subst LN dir Tells Link 0 to take the executable (. For example. Note: the %NProc directive used in earlier program versions is obsolete.g.Route file. the desired options. giving the overlay number. An argument to the program chaining routine can override the jump. the file name must have the standard form of lnnnn. specifies a run through the links 702. at which point the line following the end of the loop is executed..Link. and 716 (in this order).format for the machine involved. This is given in parentheses at the end of the link list.3.. It indicates which overlay line is executed after completion of the current overlay... A further feature of the route specification is the jump number. on the other hand. If all options have their default value. the default value is +0. in execution order.3. This is used during geometry optimizations to loop over a sequence of overlay lines until the optimization has been completed. For example: 7/5=3.. If it is omitted. as in 7//2. then a complete sequence of overlays and links with associated options can be read in. This feature permits loops to be built into the route and is useful for optimization runs. The job-type input section begins with the line: # NonStd This is followed by one line for each desired overlay. and finally a semicolon: Ov/Opt=val.3. the line would be 7//2. the list of links to be executed./Link. Specifying Non-Standard Routes If a combination of options or links is required which is drastically different than a standard route. just before the semicolon.16.exe. . If the jump number is set to -4..Opt=val.16. with option 5 set equal to 3 and option 7 equal to 4 in each of the links.7=4/2.16(-4). where nnnn is the Link number.. indicating that the program will proceed to the next line in the list (skipping no lines). then execution will continue with the overlay specified four route lines back (not counting the current line). a slash. another slash. Only the directory can be specified. 703. and that Link 9999 (generating an archive entry) follows it.9=1/99. 4/7=1/1. 2/10=1. 99/5=1.Note that non-standard routes are not generally created from scratch but rather are built by printing out and modifying the sequence produced by the standard route most similar to that desired.11. 3/11=1.2. 912. This can be accomplished most easily with the testrt utility. an MP4 single point has integral transformation (links 801 and 802) and the MP calculation (links 901. 910.30=1/1.12=2/2. except that only Link 301 (set up of basis set) would be included from overlay 3 and that Link 402 (code excerpted from the MOPAC program) would replace Link 502. except that Link 1 (reading and interpreting the route section) precedes the actual calculation.8=2. 6/7=2. The resulting sequence of programs is illustrated below: The basic sequence of program execution is identical to that found in any ab initio program.25=14.19=1. 911. 5//2.28=1/1. The standard route: # RHF/STO-3G causes the following non-standard route to be generated: 1/29=10000/1. and 913) inserted . Similarly.3.10=2.9=2. 909. A Simple Route.14. An AM1 single-point would be similar. 1//3(-5).8=2.14.39=1/1. The resulting sequence of program execution is illustrated below: .25=14. 4/5=5. 99//99. 3/11=1.3.3.27=1.29=10000/1.25=14.28=1/1.16=2/1. 6/7=2.3. 4/7=1/1.10=2. 5//2. 1/10=7/3(1).28=1/1.after the population analysis and before Link 9999. 6/7=2.30=1.11. 5//2.3.9=2.7=1. A Route Involving Loops.29=1/1. 3/11=1.9=2.2. The standard route: # RHF/STO-3G Opt produces the following non-standard route: 1/10=7.3.10=2. 3/11=1.11.30=1/1. 2/10=1. 99/9=1/99.8=2. Link 9999 automatically terminates the job step when it completes.2. 7/25=1. 2//2. 7/27=1/1.14.2.12=2/2.16.16.2.3.30=1/1. recognize that the structure is optimized. since several actions must be performed only once. the program should calculate the gradients once. not at the relatively uninteresting intermediate geometries. Link 103) deciding whether another geometry was required or the structure has been optimized. These include reading the initial Z-matrix and generating the initial orbitals. If a converged geometry is supplied. There must be a loop over geometries. and quit. with the optimization program (in this case the Berny optimizer. .Several considerations complicate this route: • • • • The first point of the optimization must be handled separately from later steps. Population analysis and orbital printing should be done only at the first and last points. SCF. They are always executed after all standard links in that . causing the backward jump on line 15 to be executed. The route for Opt=Restart is basically just the main loop from the original optimization. Normally. The same twophase route structure is used for numerical differentiation to produce frequencies or polarizabilities. Link 103 (which does its own initialization). The forward jump on the eighth line has the effect that if Link 103 exits normally (without taking any special action) the following line (invoking Link 9999) is skipped. causing Link 9999 to be invoked by the following line and the job to complete. which prints the final multipole moments as well as the orbitals and population analysis if so requested. and integral derivatives in the route. with all but Link 301 omitted from the invocations of overlay 3. it exits with a flag which suppresses the jump. Finally. and has options set to tell Link 401 to generate an initial guess. Link 103 chooses a new geometry and exits normally. Link 9999 generates the archive entry and terminates the job step. it exits and suppresses the jump. guess. If Link 103 finds that the geometry has converged. If the geometry is still not converged. followed by the rest of the second energy+gradient sequence. The second sequence uses geometries produced in Link 103 in the course of the optimization. It concludes with Link 103. The next link to be executed will be Link 202. which processes the new Z-matrix. MP and CI optimizations have the transformation and correlation overlays (8 and 9) and the postSCF gradient overlays (11 and 10. with the special lines for the first step omitted. and has options set to tell Link 401 to retrieve the wavefunction from the previous geometry as the initial guess for the next. The second invocation of Link 103 is kept and does the actual restarting. Lines 10-15 form the main optimization loop. This evaluates the integrals. and the next line processed to be line 10. causing the concluding lines (16-18) to be processed. KEYWORDS RELATED TO NON-STANDARD ROUTES ExtraLinks Enables the inclusion of extra links in an otherwise standard route (the link names are specified as its options). in that order) inserted before overlay 7. If the second invocation of Link 103 finds that the geometry is converged. which constitutes the main optimization loop. beginning a new cycle. The concluding line generates the multipole integrals at the final geometry for use in Link 601.The first point has been dealt with by having two basic sequences of integrals. in this second invocation of Link 103 the initial gradient will be examined and a new structure chosen. Routes for AM1 optimizations are similar. The first sequence includes Link 101 (to read the initial geometry). wavefunction. and gradient for the second and subsequent points in the optimization. Link 402 replacing Link 501. and overlay 7 omitted (the MOPAC code in Link 402 computes the gradient information internally). since the user can provide new options to an additional link.occurrence of the overlay. Skip This keyword allows the user to skip past a certain number of overlay lines in a standard route generated by the parser. Link 99 line: 99//99. are described in the Gaussian 03 Programmer's Reference. This provides greater flexibility than the ExtraLinks keyword. but not recommended for production level calculations. Skip=M Skip the first M overlays. after any links in that overlay would be executed anyway. It can be invoked in two ways: Skip=Ovn Skip all overlays until the first occurrence of overlay n. Most of the options are for debugging. instead of just accepting those which happen to be already there for a given overlay. These are overlay lines as described above. they are described in the Gaussian 03 Programmer's Reference. For example. useful for developing new methods and other debugging purposes. • • • • • • • ExtraLinks ExtraOverlays IOp2 and its synonyms MDV and Core IOp33 Restart Skip Use . the program expects one or more lines of input after the blank line following the route section. The program will parse the standard route and add any extra overlay lines to the route just before the last overlay. ExtraLinks=(L901) specifies that Link 901 is to be included in every occurrence of overlay 9. Program Development-Related Keywords The following keywords. When specified. Use Allows the user to request an alternative algorithm for certain phases of the calculation. See also the discussion of the %KJob Link 0 command. ExtraOverlays Provides a mechanism for customizing a route which is somewhat intermediate between using ExtraLinks and reading in an entirely new non-standard route. generated in the standard route. A blank line is then used to separate the last extra overlay line from the title section. we strongly recommend converting to the up-to-date equivalents given in the table.The Gaussian 03 IOps Reference also documents all internal options (IOps).htm. While all of them are still supported by Gaussian 03. Obsolete Keywords The following table lists obsolete keywords used by previous versions of Gaussian. They are also documented at www. Obsolete Keyword Alter BD-T BeckeHalfandHalf Camp-King CCSD-T CubeDensity Cube=Divergence DIIS Direct GridDensity Guess=Restart MP2=Stingy and VeryStingy NoDIIS NoExtrap NoRaff OldConstants Opt=AddRedundant OptCyc=n OSS PlotDensity Prop=Grid QCID QCISD-T QCSCF Raff Save SCFCon=n Replacement Keyword & Option Guess=Alter BD(T) BHandH SCF=Camp-King CCSD(T) cubegen cubegen SCF=DIIS SCF=Direct cubegen SCF=Restart none (options are a no-op) SCF=NoDIIS SCF=NoExtrap Int=NoRaff Constants=1979 Opt=ModRedundant Opt(MaxCyc=n) GVB(OSS) cubegen cubegen CCD QCISD(T) SCF=QC Int=NoRaff none (Save is a no-op) SCF(Conver=n) .gaussian.com/iops. FChk (note the mixed case). EField: Write the electric field properties (in Cartesian coordinates). Its functionality is now handled by formchk and unfchk. the electrostatic potential. CCD+STCCD Specifies a coupled cluster calculation using double substitutions and evaluation of the contribution of single and triple excitations through fourth order using the CCD wavefunction. . OptCart: Write the intermediate structures from an optimization in Cartesian coords. and Laplacian of the density over a 3 dimensional grid (cube) of points.SCFCyc=n SCFDM SCFQC SCRF=Checkpoint VShift[=n] SCF(MaxCyc=n) SCF=DM SCF=QC Field=EChk SCF(VShift[=n]) Obsolete Utility The chkmove utility. the electron density. which is always preferable. FORMCHK OPTIONS All: Write everything to the formatted checkpoint file. This keyword is deprecated in favor of the formchk utility. density gradient. ForceCart: Write forces in Cartesian coordinates. ST4CCD is a synonym for CCD+STCCD. CPHF=DirInv Invert the A-matrix directly. ForceInt: Write forces in internal coordinates. This keyword cannot store transition densities or natural orbitals in the formatted checkpoint file. is no longer provided. Its use is deprecated in favor of the cubegen utility. OptInt: Write the intermediate structures from an optimization in internal coords. FormCheck Requests that a formatted version of the checkpoint file be written at the end of a successful run. the norm of the density gradient. The default is the iterative solution. It is superseded by CCSD(T). and it is placed into the default directory from which the job is run. which converted checkpoint files to and from binary and text formats for transfer between different computer architectures. Cube This properties keyword can be used to evaluate molecular orbitals. The formatted checkpoint file always has the name Test. the structure resulting from an LST calculation may be suitable as input for a subsequent Opt=TS. Note that an LST calculation does not actually locate a proper transition state. CurrentDensity: Write the generalized density for the current method. Counterpoise and other keywords. MO: Write the Molecular orbitals. etc. In general. CurrEx1PDM: Write the CI-Singles 1PDM for the current state. Cartesian coordinates can be included in molecule specifications without any special options being necessary. coefficients. See below for its full description. . It can be used to locate transition structures and higher saddle points. Charge. However. CurrTrans: Write the transition density between the ground and current state. LST and LSTCyc Requests that an initial guess for a transition structure be generated using Linear Synchronous Transit [575]. use the natural orbital representation (the default is the density lower triangle). GroundTrans: Write the transition densities between the ground and all excited states. Opt=FP Requests the Fletcher-Powell optimization algorithm [144]. The LST procedure locates a maximum along a path connecting two structures and thus provides a guess for the transition structure connecting them. densities involving either ground or current state. SCFDensity: Write the SCF density. This keyword is deprecated in favor of ExtraBasis. UseNO: If densities are requested. This option requires that the Hessian be read in via ReadFC or RCFC. Opt=EnOnly Requests an optimization using a pseudo-Newton-Raphson method with a fixed Hessian and numerical differentiation of energies to produce gradients. LST is not valid with AM1.). GroundCurrTrans: Write all trans.Basis: Write the basis set data (exponents. AllDensities: Write all available densities. Massage The Massage keyword requests that the molecule specification and basis set data be modified after it is generated. which does not require analytic gradients. AllTrans: Write all transition densities. the LST method has been superseded by Opt=QST2. AllEx1PDM: Write all CI-Singles 1PDMs. Spin: Write separate α and β components (default=total density). Geom=Coord Indicates that the geometry specification is in Cartesian coordinates. however. Opt=Grad Requests a gradient optimization. which corresponds to the Gaussian 98 SCRF=PCM option except for some minor implementation details [302]. It requires the dielectric constant of the solvent and the number of points per sphere as input. The radii of the spheres may optionally be specified for each atom type by including the ReadRadii option. This model is no longer recommended for general use. The default SCRF method is IEF-PCM. SCRF=Numer Force numerical SCRF rather than analytic. Output=PolyAtom This requests output of an integral file in one variant of the format originated for the PolyAtom integrals program. The Murtaugh-Sargent optimization method is an obsolete alternative. SCRF=DPCM Uses the polarizable dielectric model [285. Opt=UnitFC Requests that a unit matrix be used instead of the usual valence force field guess for the Hessian. No gradients are available for this option. if possible) force constants be computed and used to start the (presumably ab initio) optimization. This is only of interest to users of the Caltech programs. This option implies the use of spherical cavities.287]. This is the default whenever analytic gradients are available and is invalid otherwise. Opt=MS Specifies the Murtaugh-Sargent optimization algorithm [145]. and is retained in Gaussian 03 only for backwards compatibility. which are not recommended. This keyword is required for multiple orders beyond Dipole. Output=Trans Write an MO coefficient file in Caltech (Tran2P5) format. The format produced by default is that used by the Caltech MQM programs. .286. Alternate radii for each atom for use in fitting potentials may be input via the ReadAtRadii option. Opt=MNDOFC Requests that the MNDO (or AM1. but the code in Link 9999 is easily modified to produce other variations on the same theme. SCRF=OldPCM The PCM model present in Gaussian 94 may be accessed using this option to SCRF. using the default method unless another option is specified. In addition. evenly distributed over the rectangular grid generated by the program (which is not necessarily a cube). use Density=Current to evaluate the cube over the density from a correlated or CI-Singles wavefunction instead of the default Hartree-Fock density. Which density is used is controlled by the Density keyword. the electrostatic potential. so Cube does not require that the cube be specified by the user.SCR scratch file. the norm of the density gradient. the output filename must always be provided (see below). Quadrupole.Only) Density=Checkpoint in the route section of a subsequent job (or job step) in order to evaluate a different quantity without repeating any of the other steps of the calculation. Note that only one of the available quantities can be evaluated within any one job step.000 points (1003). However. All but Dipole require that the Numer option be specified as well. By default. Cube evaluates the electron density (corresponding to the Density option).SCRF=Dipole The options Dipole. Description of Cube The Cube properties keyword can be used to evaluate molecular orbitals. Stable=Symm Retain symmetry restrictions. Gaussian provides reasonable defaults for grids. and include Guess=(Read. Cube=100 specifies a grid of 1. immediately after the line specifying the dielectric constant and radius (three free-format reals). For example. Alternatively.nd Hexadecapole specify the order of multipole to use in the SCRF calculation. Octopole. NoSymm relaxes symmetry restrictions and is the default. and Laplacian of the density over a 3 dimensional grid (cube) of points. SCRF=Cards Begin the SCRF=Numer calculation with a previously computed reaction field read from the input stream. the electron density. Save the checkpoint file (using %Chk). density gradient. Its use is deprecated in favor of the cubegen utility. the input format used by earlier versions of Gaussian . %SCR Used to specify the location of the .000. Cube may be given a parameter specifying the number of points to use per "side" (the default is 80). If IFlag is less than 0.is still supported. Note that Pop=None will inhibit cube file creation. they are not orthogonalized. Y0. in freeformat. Z3 size in the Z-direction. Medium and Fine may also be specified as the parameter to Cube. an unformatted file will be written. The options Coarse. X1. according to the following syntax: Output-file-name IFlag. Z2 size in the Y-direction. INPUT FORMAT When the user elects to provide it. the cube filename (or cube filename and cube specification input) is immediately followed by a list of the orbitals to evaluate. they are interpreted as Angstroms (this keyword is not affected by the setting of the Units keyword). Y1.6). If the Orbitals option is selected. Cube=Cards indicates that a grid will be input. They correspond to densities of 3. then a formatted file will be produced. If N1<0 the input cube coordinates are assumed to be in Bohr. Y2. |N1| is used as the number of X-direction points in any case. the grid information is read from the input stream. so the grid need not be rectangular. Note that the three axes are used exactly as specified. X0. X2. must conform to format (I5. Subsequent lines. Output unit number and Number of points and stepNumber of points and stepNumber of points and step- IFlag is the output unit number. which are included only with Cube=Cards. These options are designed to facilitate uniform quality in grid sampling across the range of molecular sizes. Z0 initial point. In addition to numbers for the orbitals (with β orbitals numbered starting at N+1). otherwise. N2. N3. The first line-required for all Cube jobs-gives a file name for the cube file. otherwise. terminated by a blank line. X3. 6 and 12 points/Bohr. Y3. N1. Required in all Cube jobs. The files created by Cube can be manipulated using the cubman utility.3F12. It may be used to specify a grid of arbitrary size and shape. Z1 size in the X-direction. respectively. the following abbreviations can appear in the list: HOMO The highest occupied molecular orbital LUMO The lowest unoccupied molecular orbital . regardless of the input units. Y-Origin. unformatted files have one row per record (i. then there may be blank space in some lines. 3*N3. charge. but with two writes for each row (of lengths N3 and 3*N3). the output is unformatted.. X2. Density+gradient grids are similar.5). for a density cube.. Z-Origin N1. Z2 N3. N1*N2 records each of length N3). Y3. the output is formatted. Density + gradient + Laplacian grids have 3 writes per row (of lengths N3. X3. one value per point). and N3) For example. X1. Y2. then the IFlag parameter's sign determines the output file type. regardless of the input units. if N3 is not a multiple of six. X1. For density and potential grids. If IFlag>0. Z3 # of increments in the fastest running direction IA1. and coordinates of the first atom . In this case. each row is written out in format (6E13. and are written out in the same manner. Y1.e. .e. Y1. For formatted output. Using the default input to Cube produces an unformatted output file (you can use the cubman utility to convert it to a formatted version if you so desire). X-Origin. All values in the cube file are in atomic units. the output file looks like this: NAtoms. OUTPUT FILE FORMATS All values in the cube file are in atomic units.. If IFlag<0. Chg1. The norm of the density gradient and the Laplacian are also scalar (i. When the Cards option is specified.OCCA All occupied (α) orbitals OCCB All β occupied orbitals for UHF ALL All orbitals VALENCE All occupied non-core orbitals VIRTUALS All virtual orbitals See the examples section for sample input files. Z1 # of increments in the slowest running direction N2.. Z1 Atomic number. written for each pair of I1. and an additional record follows the data for the final atom (in format 10I5 if the file is formatted): NMO.N3) Read(n. I3=1. respectively.N2. . For molecular orbital output.'(6F13.'(6E13. a correct set of Fortran loops would be: Do 10 I1 = 1.I3=1. then each record is NMO*N3 long and has the values for all orbitals at each point together.NMO) Number of MOs and their numbers If NMO orbitals were evaluated. and the increment is (X1.N3.I3) has the coordinates: X-coordinate: X0+(I1-1)*X1+(I2-1)*X2+(I3-1)*X3 Y-coordinate: Y0+(I1-1)*Y1+(I2-1)*Y2+(I3-1)*Y3 Z-coordinate: Z0+(I1-1)*Z1+(I2-1)*Z2+(I3-1)*Z3 The output is similar if the gradient or gradient and Laplacian of the charge density are also requested.I2.N1).Z1). then point (I1.I3.I1).I3=1. N1 Do 10 I2 = 1.Z0).N2.I2. N2 Read(n.'(6F13.N3) 10 Continue where n is the unit number corresponding to the cube file. NAtoms will be less than zero. Zn atom (N1*N2) records. N1 Do 10 I2 = 1.I2.N1). READING CUBE FILES WITH FORTRAN PROGRAMS If one wishes to read the values of the density.5)') ((G(IXYZ.IXYZ=1. G(3. I2 values.IAn. if the density and gradient are to be read into arrays D(N3. each of length N3 Atomic number.N3) 10 Continue where again n is the unit number corresponding to the cube file.N1) code like the following Fortran loop may be used: Do 10 I1 = 1.I1). If the origin is (X0. Laplacian. RL(N3.I=1.Y0. and coordinates of the last Values of the density at each point in the grid Note that a separate write is used for each record. or potential back into an array dimensioned X(N3.Y1.3).N2. Chgn. charge. N2 Read(n.5)') (D(I3. Xn.N1).N2.I1).I2. (MO(I). Thus.5)') (X(I3. Yn. except that in these cases there are two or three records. For example. evenly distributed over the rectangular grid generated by the program (which is not necessarily a cube). Orbitals Compute the values of one or more molecular orbitals at each point.000 points (1003).GRID-RELATED OPTIONS N Number of points to use per "side" (the default is 80). Cube=100 specifies a grid of 1. Fine 12 points/Bohr. This is the default for the density. Full Evaluate the density including all electrons. Laplacian Compute the Laplacian of the density ∇2ρ). Potential Compute the electrostatic potential at each point. CUBE CONTENTS OPTIONS Density Compute just the density values.000. Cannot be combined with the Volume keyword or the Cube=Density option. Divergence is a synonym for Laplacian. and is not allowed for the potential. FC is a synonym for FrozenCore. Gradient Compute the density and gradient. . FrozenCore Remove the SCF core density. MO is a synonym for Orbitals. Medium 6 points/Bohr. Coarse 3 points/Bohr. NormGradient Compute the norm of the density gradient at each point. Cannot be combined with the Volume keyword or the Cube=Orbitals option. Density.47122063 orbitals.2.1.1. This is the default Alpha Use only the alpha spin density. #n rhf/6-31g* 5d scf=tight cube=(orbitals) test HOMO and LUMO in default cube 0. Cards Read grid specification from the input stream (as described above).A3 Variables: R2=0.R3. Arbitrary Read in a list of arbitrary points.96 R3=1.1 O H.42 A3=109.cube containing the HOMO and LUMO. Spin Use the spin density (difference between alpha and beta densities).cube homo lumo The following cube file illustrates the method for defining your own cube via Cube=Cards: # rhf/6-31g* 5d scf=tight cube=(density.R2 F. cubegen The following job will create a cube file named orbitals. Beta Use only the beta spin density.cards) test .Total Use the total density. 6-311G(3d1f). The standard basis functions are assigned to atoms before Massage alterations take place.1. cY.cube -51 -2.0 -1.R3.0 0.42 A3=109.96 R3=1. Calculations with massaged basis set data cannot generate archive entries.R2 F.0 0. Common polarization or diffuse functions can be added in this way to standard basis sets for which these functions are not internally defined.0 -2.1 O H.A3 Variables: R2=0. func. Counterpoise and other keywords.2.0 40 0.1. For example. The Massage keyword thus makes it possible to add additional uncontracted basis functions to a standard basis set. diffuse functions could be added to the 3-21G basis set to form 3-21+G.0 0. Charge. Similarly. This keyword is deprecated in favor of ExtraBasis. cZ ] . exp.0 20 0.Density cube with user-defined cube 0.0 0. Massage may also be used for counterpoise calculations and BSSE (see the examples). and do not take advantage of molecular symmetry. Point charges may also be specified for single point energy calculations using Charge.1 40 0.0 0. INPUT Massage requires one or more lines of input in the following format: center. Some of this functionality of Massage has been superceded by the ExtraBasis keyword. [cX. which is larger than the largest internally stored 6-311G-based basis set.1 0.1 Description of Massage The Massage keyword requests that the molecule specification and basis set data be modified after it is generated. while the number of electrons is computed from the atomic numbers after the modifications.47122063 density. polarization functions might be added to 6-311G to form a 6-311G(5d3f) basis. func is a code indicating the type of modification (see below). Note that this keyword is not affected by the setting of the Units keyword. and its input is always interpreted as Angstroms. exp is the exponent of Gaussian or new nuclear charge (a value of 0 says to add a ghost atom). and cX. The following input file adds point charges to a calculation on water using the Massage keyword. Charge.where center is the center number (numbering follows the ordering of the molecule specification section). 4 or S Add an S shell. ExtraBasis. -1 or Ch Add a point charge. A blank line terminates this input section. 1 or SP Add an SP shell. Gen. 2 or D Add a D shell. 5 or F Add an F shell.cZ are the coordinates of the point charge in Angstroms when func is -1 (see below). Counterpoise Adding Point Charges. func can take on these values: 0 or Nuc Change the nuclear charge. # RHF/6-31G(d) Massage Test Water with point charges . 3 or P Add a P shell.cY. Note: This is usually done with the Charge keyword and input. 464 0.0 -1.5 1.051401 0.0 Adding Basis Functions.143 0.000136 0.612551 1 D 0.015128 H-Bq 3.021888 -0. Note: The Counterpoise keyword is now used to perform this type of calculation. Note: This is usually done with the ExtraBasis keyword and input. H2O removed 0 1 H 0.0 0.0 1. The atoms to be removed are simply designated with the ghost atom suffix (Bq).0 1 O -0.001967 -0.177 H -0.0 0 ch 2.026973 0.441 -0. Note the the Massage keyword is not used.464 1.0 1.771917 O-Bq 2.184042 0.961882 A=104.789231 -0. The utilities are discussed in alphabetical order within this chapter. .0 Computing Counterpoise Corrections Manually.0 0.000050 -0.0 3 P 1.0 0 ch 2. The following input adds functions to the D95 basis set (in order to reproduce a calculation from the literature that used a non-standard basis set).137 H 0. # RQCISD(Full)/D95 Freq=Numer Massage Test H2O Frequencies at QCISD(Full)/DZP 0 1 O H 1 R H 1 R 2 A R=0.004924 0. # b3lyp/3-21G** nosymm scf=tight test HBr + H2O manual counterpoise calculation.85 2 P 1.536864 H-Bq 3. The following input file performs a counterpoise calculation.784986 -0.185282 Utility Programs This page discusses various utility programs included with Gaussian 03.0 1.685176 Br -0.0 1. temperature. be sure to consult the release notes accompanying the program for information pertaining to specific operating systems. pressure and scale factor can be specified for the thermochemistry analysis. . freqchk* freqmem gauopt ghelp mm newzmat* testrt* unfchk* GAUSS_MEMDEF Environment Variable The GAUSS_MEMDEF environment variable may be used to increase the memory available to utilities which do not offer such an option themselves.Most utilities are available for both UNIX and Windows versions of Gaussian. Prints frequency and thermochemistry data from a checkpoint file. On-line help for Gaussian. Standalone molecular mechanics program. Conversion between a variety of molecular geometry specification formats. and so on). Manipulates Gaussian-produced cubes of electron density and electrostatic potential (allowing them to be added. Converts a binary checkpoint file into an ASCII form suitable for use with visualization programs and for moving checkpoint files between different types of computer systems. Displays the route and title sections from a checkpoint file. Determines memory requirements for frequency calculations. This page presents a brief overview of traditional Z-matrix descriptions of molecular systems. Alternate isotopes. Route section syntax checker and non-standard route generation. Standalone cube generation utility. However.g. Performs optimizations of variables other than molecular coordinates. Its value should be set to the desired amount of memory in words. Convert a formatted checkpoint file back to its binary form (e.. The following lists the available utilities and their functions (starred items are included on the Gaussian 03W Utilities menu): c8603 chkchk* cubegen* cubman* formchk* Converts checkpoint files from previous program versions to Gaussian 03 format. after moving it from a different type of computer system). subtracted. and the dihedral (torsion) angle formed by the plane containing atom1. atom 2. As an initial example.9 O 2 1.format-code] Although these examples use commas to separate items within a line. as in an ONIOM calculation. For the syntax being described here. The most-used Z-matrix format uses the following syntax: Element-label. If the elemental symbol is used. It defines the other hydrogen as bonded to the second oxygen with an H-O distance of 0.Using Internal Coordinates Each line of a Z-matrix gives the internal coordinates for one of the atoms within the molecule.9 Angstroms. Element-label is a character string consisting of either the chemical symbol for the atom or its atomic number. it may be optionally followed by other alphanumeric characters to create an identifying label for that atom. The third line defines another oxygen with an O-O distance of 1. where the charge and spin multiplicity line is line 0. from atom 2.0 1 120.9 2 105. Atom1. The optional format-code parameter specifies the format of the Z-matrix input. the angle formed by this bond and the bond joining atom1 and atom2. atom1 and atom2. bond-angle. . an H-O-O angle of 105 degrees and a H-O-O-H dihedral angle of 120 degrees.e.4 Angstroms (i. Note that bond angles must be in the range 0º < angle < 180º. atom2. dihedral-angle [. The next line lists an oxygen atom and specifies the internuclear distance between it and the hydrogen as 0. Dihedral angles may take on any value.4 1 105. this code is always 0.0 The first line of the Z-matrix simply specifies a hydrogen. C2. Alternatively. A common practice is to follow the element name with a secondary identifying integer: C1. atom3 are the labels for previously-specified atoms and are used to define the current atoms' position. etc.0 H 3 0. The position of the current atom is then specified by giving the length of the bond joining it to atom1. A Z-matrix for this structure would be: H O 1 0. any valid separator may be used. the other oxygen) and having an O-O-H angle (with atoms 2 and 1) of 105 degrees..9 Angstroms. atom 3. consider hydrogen peroxide. This code is needed only when additional parameters follow the normal Z-matrix specification data. atom 1. The fourth and final line is the only one for which all three internal coordinates need be given. the other atoms' line numbers within the molecule specification section may be used for the values of variables. atom2 and atom3 with the plane containing the current atom. bond-length. Variables may be used to specify some or all of the values within the Z-matrix.39 -0.4 A 105.00 0. the values of the variables will be optimized to locate the lowest energy structure.0 Constants: D 120.92 It is also possible to use both internal and Cartesian coordinates within the same Zmatrix.00 0.00 1.0 Symmetry constraints on the molecule are reflected in the internal coordinates.88 0. as in this example: C C H H H H H H 0. Mixing Internal and Cartesian Coordinates Cartesian coordinates are actually a special case of the Z-matrix.92 1.51 0. variables in a second section (often labeled Constants:) are held fixed in value while those in the first section are optimized: Variables: R1 0. as are the two H-O-O bond angles.92 1.39 1.00 -0. Here is another version of the previous Z-matrix: H O 1 R1 O 2 R2 1 A H 3 R1 2 A 1 D Variables: R1 0.88 0.51 -1.88 0. The two H-O distances are specified by the same variable.00 -0.51 0. as in this example: O 0 xo C 0 0.88 0.0 See the examples in the discussion of the Opt keyword for more information about optimizations in internal coordinates. yc zo 0.00 1.52 -0.00 0. For a partial optimization (POpt). For a full optimization (FOpt).4 A 105.02 0. the variables are required to be linearly independent and include all degrees of freedom in the molecule.9 R2 1.0 D 120. 0.9 R2 1. When such a Z-matrix is used for a geometry optimization in internal coordinates (Opt=Z-matrix).39 -0.51 -0. .02 -0. 0 xn 0. This Z-matrix has several features worth noting: • • • The variable names for the Cartesian coordinates are given symbolically in the same manner as for internal coordinate variables. H 1 nh 2 hnx . Alternate Z-matrix Format An alternative Z-matrix format allows nuclear positions to be specified using two bond angles rather than a bond angle and a dihedral angle.C N H H H H H 0 0.0 C2 110. Cartesian coordinates can be related by a sign change just as dihedral angles can.08 r3 1.02 a1 125. yc 1. which indicates a dihedral angle as the third component): C4 O1 0. The following example illustrates the use of a dummy atom to fix the three-fold axis in C3v ammonia: N X 1 1. b1 90.4 1 C6 O1 R C2 A1 C3 A2 1 The first line uses a dihedral angle while the latter two use a second bond angle. 0.9 C2 120. Using Dummy Atoms This section will illustrate the use of dummy atoms within Z-matrices. b2 -90.4 C4 105. a2 125. -yc 0. The integer 0 after the atomic symbol indicates symbolic Cartesian coordinates to follow. which are represented by the pseudo atomic symbol X. d3 160. r1 1. 2 r1 3 a1 1 b1 2 r2 3 a2 1 b2 3 r1 2 a1 1 -b1 3 r2 2 a2 1 -b2 4 r3 2 a3 3 d3 Variables: xo -1.0 0 C5 O1 1.08 r2 1. This is indicated by a 1 in an additional field following the second angle (this field defaults to 0. zo 0.3 O2 180. xn 1. 0 hnx 70. 1 180.0 hcco 130.20 ch 1. The following examples illustrate the use of dummy atoms for specifying linear bonds. Second. First.0 The position of the dummy on the axis is irrelevant. X halfcc O 90. a dummy atom is placed at the center of the CC bond to help constrain the cco triangle to be isosceles.H 1 nh 2 hnx 3 120. 1 90. hnx is the angle between an NH bond and the threefold axis. and the distance 1. ox is then the perpendicular distance from O to the C-C bond.0 This example illustrates two points. C1 180. Geometry optimizations in internal coordinates are unable to handle bond angles of l80 degrees which occur in linear molecular fragments. H 2 ch 3 90. Difficulties may also be encountered in nearly linear situations such as ethynyl groups in unsymmetrical molecules. Here is a Z-matrix for oxirane: X C1 O C2 H1 H2 H3 H4 X halfcc X ox C1 90. and the angles oxc are held at 90 degrees.08 hcc 130. cn 1. some of the entries in the Z-matrix are represented by the negative of the dihedral angle variable hcco.0 C1 ch X hcc O hcco C1 ch X hcc O -hcco C2 ch X hcc O hcco C2 ch X hcc O -hcco halfcc 0.0 used could have been replaced by any other positive number.06 .0 H 1 nh 2 hnx 3 -120. such as acetylene or the C4 chain in butatriene. These situations can be avoided by introducing dummy atoms along the angle bisector and using the half-angle as the variable or constant: N C 1 cn X 2 1.0 ch 1.0 nh 1.75 ox 1. each of which defines an atom (by atomic symbol) and its connectivity. etc.Similarly.Geom This changes the orientations of the I-J and K-L bonds about the J-K bond. in this Z-matrix intended for a geometry optimization. Geom is either . Trig. or an atomic symbol (e. Then X will appear in row Y and Y will appear in row X.J. There are zero or more of each of the following lines. or Line. in ethylene is trigonal. IPr. The short formula matrix may be followed by optional lines modifying the generated structure.I.g.I. The AtomGeom line changes the value of the bonds at center I. Let I be the atom to the right of X in row Y and J be the atom to the right of Y in row X. The short formula matrix also implicitly defines the rotational geometry about each bond in the following manner. or one of the strings Tetr. half represents half of the NCO angle which is expected to be close to linear. Suppose atoms X and Y are explicitly specified. Note that a value of half less than 90 degrees corresponds to a cis arrangement: N C X O H 1 2 2 4 cn 1. a collection of lines.0 half 80. NBu. It is requested with the ModelA or ModelB options to the Geom keyword.). F) to which the current atom is connected by a terminal bond. NH2.. or a symbol for a terminal functional group which is bonded to the current atom. 1 half co 3 half 1 180.0 cn 1.0 oh 2 coh 3 0. Bent. which must be grouped together in the order given here: AtomGeom. NPr.K. Model Builder Geometry Specifications The model builder is another facility within Gaussian for quickly specifying certain sorts of molecular systems. IBu. and it requires additional input in a separate section within the job file.0 coh 105.20 co 1. The basic input to the model builder is called a short formula matrix.L. Then atoms I and J are put in the trans orientation about the X-Y bond. and TBu. Pyra.Geom Normally the local geometry about an atom is defined by the number and types of bond about the atom (e. Me. Et.g. BondRot. All bond angles at one center must be are equal.3 oh 1. H. carbon in methane is tetrahedral. Each of these can be either an integer. Geom may be the angle as a floating point number. The functional groups currently available are OH. which is the number of the line defining another explicitly specified atom to which the current atom is bonded. by up to six more entries. Pittsburgh.NewLen This sets the length of the I-J bond to NewLen (a floating point value). and J. A. -H). Rohlfing. and J. 1982). W. 1986). Robb. Gaussian 86 (Gaussian. P. Pittsburgh. A. the appropriate Model B bond length is used instead. Krishnan. M. 7 M. R. Kahn. R. Whiteside. PA. Binkley. R. C. J.I. Whiteside. C. Gaussian 76(Carnegie-Mellon University. J. H. C. R. Seeger. J. M. L. S. Defrees. J. which are specified by a minus sign before the atomic symbol (e. E. M. S. Kahn. D. R. Binkley. E. Frisch. J.J.) of a bond in assigning bond lengths.g. A. L. J. Gonzalez. Gaussian 88 (Gaussian. Raghavachari. Program No. R. Pople. S. Melius. and J. its extra valence can be "tied down" using dummy atoms. Seeger. Melius. D. while model B bond lengths depend only on the types of the atoms involved. Hehre. J. W. K. and J. M. J. The model builder can only build structures with atoms in their normal valencies. Hariharan. Frisch. double. M. M. P. Seeger. S. Trucks. 1 W. J. 1980). . etc. C. Binkley. Defrees. Only terminal atoms can be dummy atoms. Schlegel.. 237. J. J. Topiol. D. Newton. K. PA. B. H. Foresman. Gaussian 70. Gonzalez. W. Topiol. A. R. 1976). Gaup (≥+60). F. J. Inc. Stewart. D. J. J. B. B. 1988). Schlegel. A. Head-Gordon. S. Fox. Kahn. J. E. A. Schlegel. PA. H. R. Stewart. 3 J. A. A. PA. A. J. R. R. J. A. Gaussian 82 (Carnegie-Mellon Quantum Chemistry Publishing Unit. Gaussian 80 (Carnegie-Mellon Quantum Chemistry Publishing Unit. Raghavachari. Pople. Head-Gordon. R. 2 J. Martin. J. Binkley. and M. Raghavachari. If Model A is requested and an atom is used for which no Model A bond length is defined. Model B is available for all atoms from H to Cl except He and Ne.. R. triple. L. A. S. J. Whiteside. R. Pople. F. F. Fluder. P. Schlegel. D. Seeger. Frisch. D. Pittsburgh. D. Pople. Lathan. PA. Pople. 1970). M. Bobrowicz. J. Pittsburgh. J.. L. H. W. B. B. Schlegel. R.the dihedral angle or one of the strings Cis (≥0). Trans (≥180). Binkley. Pople. D. Ditchfield. C. Defrees. Binkley. 6 M. J. M. Whiteside. Defrees. Fox. R. A. Defrees. Hehre. Quantum Chemistry Program Exchange. S. D. Frisch. Fluder. Inc. Fox. Baker. The two available models (A and B) differ in that model A takes into account the type (single. J. S. If a radical is desired. or Gaum (≥-60). BondLen. C. K. R. R. Pittsburgh. Fluder. 4 J. A. B. G. J. 5 M. and J. Krishnan. Newton. M. D. Whiteside. H. C. J. Gaussian 94 (Gaussian. B. PA. A. Nanayakkara. H. K. Cammi. Chem. Chem. Raghavachari. Frisch. M. and J. Gomperts. E. R. 1992). K. Ö. A. J. and J. Baker. C. Gaussian 90 (Gaussian. W. D. L. Challacombe. Johnson. J. Head-Gordon. Scuseria. 7211 (1980). A. Ortiz. Weinhold. Am. Dapprich. and J. Gaussian 92/DFT (Gaussian. Pople. Dannenberg. W. Weinhold. 1987. 4066 (1983). E. K. G. W. Struct. Gaussian 92 (Gaussian. Gill. W. Pomelli. M. S. M. A. A. E. Millam. PA. K.. 1998). A. Topiol. Trucks. Pittsburgh. G. W. (Theochem) 169. J. Ayala. Komáromi. 1993). S. Gomperts. F. Gonzalez. M. Cioslowski. 102. Schlegel. Petersson. J. J. Keith. Head-Gordon. Raghavachari. Stewart. P. Gill. J. C. M. 1736 (1983). Weinhold. J. Chen. Pople. R. J. T. Replogle. I. Gill. Clifford. G. B. Cui. Al-Laham. Carpenter. Liu. D. B. J. Stewart. V. B. D. 41 (1988). Mennucci. J. 12 13 14 15 16 17 J. B. J. Defrees. W. B. M. J. Replogle. Ortiz. K. B. G. B. P. R. Baker. H. A. Montgomery Jr. W. Binkley. J. Fox. C. V. E. Chem. C. A. J. A. Frisch. Wong. J. J. M. M. Gill. J. P. L. J. B. J. Gomperts. G. S. Pople. Phys. Rabuck. Cossi. M. Head-Gordon. and J. Reed and F. G. C. S. P. Frisch. P.. R. E. M. Foster and F. C. Daniels. J. R. A. Replogle. Fox. Gomperts. 78. Andres. P. Piskorz. H. J. Trucks. Weinhold. S. Gonzalez. Replogle. J. Ochterski. Stewart. Baboul. S. K. 8 M. L. J. Stefanov. Carpenter and F. Trucks. Inc. A. E. M. Petersson. C. G. S. G. Cheeseman. A. P. Robb. J.. J. Burant. Pittsburgh. Farkas. K. G. J. and F. J. Kudin. A. Chen. . M. J. Mol. M. B. J. G. C. Zakrzewski. M. Y. Foresman. M. PA. W. Head-Gordon. Baker. R. B. A. W. J. E. Al-Laham. M. R. B. J. Martin. Robb. Cioslowski. A. Y. Baker. M. Reed. Keith. Reed and F. J. Morokuma. Phys. S. 83. Pittsburgh. Andres. 11 M. M. V. A. B. W. J. C. P. B. P. Stratmann. Defrees. Gonzalez. Johnson. D. L. E. E. Binkley. R. A. Trucks. L. Gaussian 98 (Gaussian. J. D. J. A. M. R. J. D. Schlegel. W. PhD thesis. Wong. Foresman. Binkley. Martin. A. N. D. H. Pittsburgh.. Schlegel. G. J. G. M. M.. Weinstock. J. J. P. Andres. Johnson. B. Liashenko. Ayala. Defrees. Raghavachari. D. J. P. Madison. E. P. Foresman. Y. 1995). Stefanov. A. Salvador. G. Seeger. Melius. J. Raghavachari. Pittsburgh. W. Malick. Peng. Chem. Robb. Q. Andres. and J. Cheeseman. R. Fox. 735 (1985). Wong. J. 9 M. Martin. L. S. Johnson. University of Wisconsin. C. Y. J. A. J. Fluder. 10 M. A. Pople. Pople. W. V. PA. J. WI. Phys. Frisch. Inc. G. 1990). R. Fox. E. D. B. W. J. Challacombe. V. J. R. B. Strain. A. B. Inc. Zakrzewski. J. L. Inc. E. Montgomery. J. J. J. Gonzalez. Martin. Schlegel. D. Inc. R. A. R. M. Kahn. Peng. R. J.. Stewart. A. M. Tomasi. S. Foresman. Adamo. J. T. Nanayakkara. L. P. Barone. Wong. L.Whiteside. PA. Soc. Weinhold. Robb. M. J. Pople. 31 J. Struct. Strain. Science 271. Bayly. Head-Gordon. C. 5569 (1997). Frisch. 166. Plenum 227 (1988). Phys. Faegri Jr. M. Chem. Scuseria. 122 (1994). Fox. D. J. 106. M. and M. Lett. Pople. W. G. Frisch. 258. C. and J. 89. Mayo. and F. and M. Int. 166. I. M. C. 107. Phys. Head-Gordon. Head-Gordon and J. (Theochem) 529. Spellmeyer. A. Head-Gordon and T. Daniels. Phys. K. 143 (1994). Almloff. M. Chem. D. Chem. 3140 (1983). A. Scuseria. D. Soc. and W. Lett. Scuseria. 521 (1987). Phys. A. Mol. and P. J. Reed. G. M. C. Am. M. Chem. 37 W. Phys. Phys. B. Olafson. Phys. 51 (1996). J. M. Pople. J. Gould. G. T. B. E. Chem.18 19 20 21 22 23 24 25 26 27 28 29 30 A. 425 (1997). J. Caldwell.. 31. Goddard. Phys. J. J. C. M. 94. D. Ferguson. P. C. 45 (1996). Chem. Burant. C. 503 (1988). Phys. A. Scuseria. 899 (1988). 275 (1990). 248. E. E. J. Chem. A. 8969 (1996). 33 34 35 36 M. 220. J. and K. Burant. 37 (2000). Weinhold. Chem. 88. Taylor. Strain. Lett. R. Curtiss. 32 J. 385 (1982). Burant. E. 3. Head-Gordon. E. 189 (1990). J. P. D. P. Lett. M. 141. Scuseria. Chem. Frisch. 8897 (1990). Pople. Dunlap. J. Chem. Gill. Rev. Head-Gordon. 153. Chem. J. Weinhold and J. Chem. B. 25. Chem. I. G. I. and J. J. Merz Jr. M. Strain. Pople. Carpenter. 117. Cornell. Lett. Frisch. F. L. M. E. 38 S. Lett. Cieplak. A. E. and G. J. A. Head-Gordon. J. Phys. Millam and G. A. and M. M. and M. Frisch. . Quant. Chem. 5179 (1995). J. Frisch. Chem. J. Phys. L. R. J. 43 (1996). Scuseria.. E. 281 (1990). Chem. J. Adv. J. Phys. M. J. W. J. Chem. M. Comp. 78. Dunlap. Kollman. A. M. Millam. Quant. J. Frisch. and M. Frisch. and J. Chem. Korsell. M. 5777 (1988). 195. C. J. I. K. Phys. G. J. P. Thiel. Phys.39 A. J. Rzepa. Am. J. Phys. 2026 (1967). 49. Healy. McWeeny and G. 1964 (1985). C. 10. Dewar. J. J. and J. . S. Pople. Chem. Am. Soc. 50 M. J. Soc. P. J. Chem. Dierksen. Am. M. Soc. J. J. 3902 (1985). 571 (1954). al. M. 54 55 56 57 58 59 M. Chem. L. Goddard III. C. J. Mod. 51 M. Casewit. S. 95. S. 52 M. W. Dewar. J. M. 97. Am. J. C. 44. S. 3607 (1978). 3289 (1966). et. Organometallics 4. J. 433 (1981). A. Roothan. J. J. F. 1285 (1975). and H. H. M. Dewar. 99. Phys. A.. Stewart.. Chem. Soc. Chem. Stewart. J. Am. Healy. F. Pople and R. J. J. Phys. P. Am. Comp. J. D. al. Rappé. J. J. Davis. L. C. 10024 (1992). 23. S. Chem. A. K. Dewar and E. M. Am. R. Rev. S. 4899 (1977). J. Chem. Soc. Phys. Pople. Chem. McKee. G. Skiff. E. Rzepa. 4. J. 100. McKee. J. 2. Chem. Colwell. 221 (1989). Chem. 99. Dewar and W. Dewar and H. J. Dewar and W. J. Comp. Chem. 4852 (1968). 209 (1989). Thiel. J. and E. J. 107. 542 (1983). S. K. Am. Bingham and M. 10. Nesbet. 106. Comp. S. Chem. Chem. 4. P. et. 2. Comp. J. Phys. S. Stewart. 49 M. J. Rappé and W. 40 41 42 43 44 45 46 47 48 A. K. Comp. L. Soc. Chem. S. J. Dewar and C. R. J. Dewar. L. 4499 (1977). M. and P. 100. Reynolds. Beveridge. 84 (1983). 53 M. G. and W. Chem. 3358 (1991). Chem. A. J. K. 69 (1951). Dewar and M. 6771 (1984). 140 (1986). Soc. Zoebisch. Grady. Goddard III. Chem. Dewar. 114. 47. S. Comp. Segal and J. 777 (1978). Soc. J. 22. Dobosh. Chem. J. Chem. J. J. S. Chem. 70 G. A. and K. 4: The Self-Consistent Field for Molecular and Solids (McGraw-Hill. G. S. 164. 618 (1934). A. Binkley. W. Chem. 98. F. J. Phys. Lett. A. Chem. Johnson. Rev. Chem. A. Krishnan. J. D. Moller and M. R. Chem. Kohn and L. 557 (1992). Pople. Scuseria and H. J. Phys. 72 J. Binkley. A1133 (1965). M. E. A. S. Replogle. J. and H. Lett. and R. XIV. Adv. 68 69 J. Gill. 89. Quantum Theory of Molecular and Solids. 76. H. Trucks. Bartlett. Pople. Scuseria. 83 (1989). A. R. 199. Purvis and R. 1974). Head-Gordon. G. A. 545 (1978). R. Janssen. 149 (1977). P. J. Pople. P. Head-Gordon. Phys. Quant. Phys. 466 (1977). Int. Chem. Hohenberg and W. Pople. 45. New York. Phys. 1910 (1982). M. Handy. Krishnan. C. J. 14. S. 87. Chem. 94. Krishnan. J. 5968 (1987). J. Phys. 77 J. 185 (1989). Plesset. Chem. Chem. 64 K. W. J. Raghavachari. 91 (1978). and J. 5648 (1993). Chem. . E. B864 (1964). W. Phys. Phys. J. 5579 (1990). E. Dykstra. A. R. 74 75 76 C. 90. Frisch. 71 G. Kohn. J. D. 65 66 S. Chem. Rev. Int. Krishnan and J. Chem. Cizek. 78 79 J. 140. Int. Pople. Becke. 3700 (1989). C. Schaefer III. 14. Phys. Lett. Chem. Almlof. J. C. Schaefer III. E. 1 (1976). M. and R. 7382 (1988). J. Phys. F. 72. Pople. A. Sham. J. Seeger. Seeger. 4244 (1980). Symp. B. A. Int. and B. J. 35 (1969). 154. Pople. J. K. Phys. M. Raghavachari. Head-Gordon. Quant. S. and J. Phys. Raghavachari. Pople. Phys. Saebo and J. Chem. Slater. Symp. 73 N. and M. Lett. Phys. Vol. 11. 10. and G. Rev . Quant. J. Phys. Schlegel. 136. Chem. Pople. Chem. J. L. J. Quant. 46.60 61 62 63 C. 67 J. . Chem. Curtiss. Redfern. Phys. Chem. Phys. 4703 (1999). W. C.. Parthiban and J. and K. Petersson and M. 81 L. G. 110. 7650 (1999). M. Martin. L. 2537 (1990). Fox. A. Nyden and G. G. 1843 (1999). A. M. Phys. A. Chem. J. S. Hegarty and M. M. Curtiss. A. C. A. Montgomery Jr. Raghavachari. Redfern. J. Frisch.80 J. Phys. A. 93 J. 6014 (2001). J. A. 109. Phys. A. Jones. . Phys. A. Curtiss. Phys. J. Phys. D. L. Phys. J. P. de Oliveira. G. Pople. W. Chem. J. 7221 (1991). and G. K. Petersson. Ochterski. C. Petersson. Pople. 38. Rassolov. J. W. P. Head-Gordon. Curtiss. J. V. and J. T. Petersson. Chem. J. Petersson. M. 87 88 M. M. J. 84 L. 5900 (1994).. Montgomery Jr. Chem. A. 93. Phys. Montgomery Jr. 1843 (1981). and L. J. 101. Raghavachari. 85 L. Chem Phys. A. and J. A. M. L. R. Phys. Curtiss.. J.. in prep (2003). G. W. A. A. A. Redfern. P. J. 112. J. J. 92 J. A. Rassolov. Phys. V. Martin. Chem. W. Pople. Petersson. 90 J. M. Raghavachari. 6532 (2000). Ochterski. K. 82 L. Raghavachari. Tensfeldt. 5622 (1989). Martin and G. L. Chem. Phys. 104. K. J. 90. Chem. Chem. A. A. A. Raghavachari. Mol. 6091 (1991). and J. 1795 (1979). and J. A. 94. 111. C. J. J. 83 L. K. 7764 (1998). Montgomery Jr. Raghavachari. A. Chem. Pople. 1293 (1993). and J. and G. Phys. 94 95 96 97 J. 110. Ochterski. Al-Laham. and G. 75. A. J. Trucks. 91 J. A. A. G. A. Frisch. A. 94. 94. 110. A. 98. J. Curtiss. G. K. A. and J. and J. J. J. K. Chem. Trucks. 6081 (1991). 89 G. Phys. Pople. Chem. Montgomery Jr. Robb. Pople. Chem. 114. 2598 (1996). Ochterski. and J. Montgomery Jr. Phys. J. Curtiss. Baboul. W. 86 A. Frisch. Petersson. D. Raghavachari. Chem. J. 2822 (1999). Inorganic Chem . Pople. J. Soc. Phys. C. 115 116 J. Chim. Chem. 21 (1979). Siegbahn. Chem. 106 107 R. 25. 183 (1988). 362 (1981). Chem. Kirchner. Phys. 9047 (1996). 108. Celani. McDouall. Ridley and M. and M. B. D. 105 F. C. Niazi. Mueller-Westerhoff.98 99 100 101 R. 223 (1976). Phys. J. 96. Lett. A. Casida. Eade and M. Goddard III. 2728 (1986). Klene. Zerner. M. H. Klene. E. Lett. A. Blancafort. Ed. Chem. Chem. Modern Theoretical Chemistry. 23 (1990). A. 148. H. Phys. Zerner. Robb. Robb. Chem. Ridley and M. Chim. C. Chem. Am. A. Chim. C. 454 (1996). J. K. Schaefer III. and M. M. Zerner. P. M. 589 (1980). J. Phys. C. Lett. Chem. Scuseria. 135 (1992). A. New York. Seeger and J. Anderson. 110 R. Bacon and M. J. J. Acta 42. O. Robb and U. Frisch. in prep (2003). J. Acta 53. Bobrowicz and W. and U. Stratmann. Ahlrichs. 66. 103 J. Reports in Molecular Theory 1. 104 M. Chem. 102. J. Chem. F. 109 R. Phys. A. E. in Methods of Electronic Structure Theory. G. J. D. 114 M. 109. A. Acta 32. 417 (1984). E. E. Vol. R. L. C. Olsen. 109. Bauernschmitt and R. Zerner. J. 89. E. Edwards. Phys. J. 4439 (1998). A. Lett. J. J. . and M. Robb. Phys. (1977). Frisch. A. 256. F. R. and M. Theo. P. and M. Head-Gordon. E. J. E. 8218 (1998). Jamorski. J. Zerner. and H. 111 (1973). W. J. Ahlrichs. Peasley. Foresman. Robb. 1977) 79-126. M. Lowe. Phys. Bauernschmitt and R. Jensen. 3 (Plenum. Chem. Pople. 5653 (2000). 104. Frisch. 113. 113 W. 108 J. 2185 (1988). T. 112 A. 83. M. B. R. Casida. Jørgensen. Chem. Salahub. Theo. M. K. C. J. Phys. H. and P. 111 M. G. and D. Roos. Frisch. Theo. 102 M. W. Phys. 4728 (1987). Hirao. Ishida. Nakatsuji. Phys. H. Phys . Soc. 131 M. Nakajima and H. Nakatsuji. M. 68. M. Chem. Phys. Thompson and M. Chem. Toyota. Ehara. Ehara. 134 K. in Computational Chemistry—Reviews of Current Trends. 280. M. 2 (VCH Publishing. Nakatsuji. H. 130 T. 53 (1996). Soc. Ed. M. 109. 1997) 62-124. Lipkowitz and D. B. P. Vol. 94. J. and H. Lett. Nakajima and H. Phys. Lett. 7179 (1993). Ehara. 362 (1978). 118 M. Ed. Zerner. Quant. Toyota. 251 (1982). Sci. Phys. Int. Ed. J. 79 (1997). C. 135 K. Phys. . Singapore. Lett. C. 2053 (1978). H. 1 (2002). Zerner. 98. Toyota. Chem. Vol. Prog. Nakatsuji. 121 122 123 124 125 126 127 128 H. M. in Reviews in Modern Quantum Chemistry. Nakatsuji. 329 (1979). Nakatsuji and K.. Toyota. T. Nakatsuji. Ehara. Am. Leszcynski. Fajer. K. Chem. 493 (2001). Ehara. C. Chem. Chem.117 M. J. 331 (1991). Lett. J. Phys. 67. Frisch. Boyd. Lett. J. Zerner. K. 1991) 313-366. K. Nakatsuji. Chem. 242. 129 H. Lett. 119 M. 351 (2001). Chem. Lett. 177. 59. Phys. J. New York. 350. and H. Chem. Phys. 177 (1999). K. and H. 2 (World Scientific. Lett. D. Surf. H. and M. Phys. M. Nakatsuji. in press (2003). H. Chem. Sen (World Scientific. Nakatsuji. Nakatsuji. Chem. Nakatsuji. I. 21. Nakatsuji. J. K. Ishida. Nakatsuji. Mayumi. 132 M. 356. 113. 133 M. Thompson. Singapore. and H. C. 120 L. A. Chem. A. M. Toyota. Nakatsuji and M. 8210 (1991). Phys. 2002) 293. Ishida. in Rev. H. Chem. B. Zerner. Chem. 347. Am. and M. Hanson. J. Ehara. K. Comp. Correa de Mello. Phys. 6716 (1991). Chem. Hehenberger. and H. Chem. J. H. B. J. 19. 17. Sieber. Morokuma. McDougall. Fletcher and M. Pople. 33. J. 140 N. 13. 153. Comp. E. W. Chem. Comp. H. Schlegel. Comp. J. 72. A. H. 490 (1988). Chem. 138 F. J. Phys. Maseras and K. Chem. Baker. Lett. Watts. Quant. Chem. 5523 (1990). Schlegel. 143 144 145 146 147 R. 16. Peng. Morokuma. Israel J. Phys. J. B. Phys. Humbel. J. 93. D. Schlegel. Symp. B. 147. E. R. 1101 (1996). B. 5031 (1984). 214 (1982). Schlegel. Phys. J. Phys. 105. and J. Lett. A. Schlegel. Symp. J. 139 J. Chem. 43 (1982). Robb. and K. F. 137 (1984). Chem. F. and R. C. A. J. W. 6. 1170 (1995). Schlegel and M. 163 (1963). B. 148 H. 150 151 152 153 154 155 C. Sosa. Sieber. S. 325 (1979). Peng and H. 449 (1994). and R. 90. Binkley. and H. 7. J. A. Int. H. 149 C. 385 (1986). J. J . B. Trucks. Soc. 1989)33 -53. Krishnan. A. Comp. Matsubara. 1959 (1996). Murtaugh and R. Far. Ed. Y. Chem. Sargent. Comp. J. J. Phys. The Netherlands. 8. 563 (1987). 94. Lett. Bertran (Kluwer Academic. 4654 (1980). Chem. M. J. B. Bernardi. Chem. P. Handy and H. Schaefer III. J. B. Pople. A. H. T. J. 2154 (1989). Ayala. J. Schlegel. C. D. Phys. and M. Chem. Chem . C. Robb. Morokuma. S. 49 (1996). 3. W. 81. Gonzalez and H. Bartlett. Quant. J. Schlegel. Gonzalez and H. Chem. Chem. S. W. and K. Int. J. J. . Frisch. Phys. A. 13. 60. Chem. Schlegel. C. Bottini. R. Comput. A. J. J. Krishnan. in New Theoretical Concepts for Understanding Organic Reactions. Salter. S.136 137 H. Bartlett. Chem. 185 (1970). 359 (1988). Powell. J. 142 G. Salter. Comput. Baker. Chem. B. 141 G. Chem. Schlegel. B. Trucks. B. and J. Matsubara. 107. Chem. . 45. J. 21. 5704 (1998). K. P. Morokuma. Vreven. Svensson. and G. Res. Schlegel. K. Petersson. 217 (1992). 197. Berry. Grev. Chem. 170 G. Humbel. Byun. T. Humbel. Irikura and D. 367 (2001). (1980). T. Chem. Morokuma. 158 S. Chem. K. Washington. Frurip. J.. Y. C. Farkas. D. Schlegel. A. Truhlar. Comp. J.. V. 124. Sieber. and K. D. Chem. 167 F. Frisch. Frisch. Chem. A. R. Vreven. K. A. J. 105. K. G.156 M. D. Phys. 321 (1996). H. Phys. ACS Symposium Series (Amer. S. R. Robb. J. I. G. J. Ragazos. J. C. Komáromi. B. B. J.. (2003). 102. and M. Morokuma. 3654 (1996). Byun. D. Bernardi. Chem. Bernardi. Farkas. 19357 (1996). Reviews 25. 84. G. Soc. Khoroshun. Dapprich. 164 P. in press (2003). and M. Vreven. 10074 (1998). D. Torrent. J. 169 B. Musaev. Svensson. Garrett. A. & Dev. Chem.. A. J. Musaev. 1998) 237. F. Mol. Bearpark. S. K. B. Marshall. 677 ed. T. J.. J. Chem. A. Morokuma. Schwartz. Schlegel. J. K. Petersson. and K. Morokuma. K. Montgomery Jr. 166 I. and H. Dapprich. 160 K. J. 168 D. 162 T. Phys. Robb. M. 100. Torrent. Malick. Morokuma. Ö. 161 M. S. Frisch. Lett. Morokuma. Froese. 223. and D. M. Ö. R. I. 165 M. Vreven and K. Phys. H. 163 T. Soc. and J. J. N. Morokuma. J. A. 375 (1997). B. Chem. A. and A. D. Chem. Lett. Olivucci. Schlegel. in prep. IBM J. 269 (1994). J. Struct. 1 (1999). Phys. K. Olivucci. and M. Phys. J. S. 192 (2002). 171 M. Petersson. Vreven. Comp. Am. Chem. Soc. G. 1419 (2000). S. Robb. S. and M. 108. Ayala and H. S. J. and M. Phys. (Theochem) 462. and H. in Computational Thermochemistry. Ehlers. Chem. 159 T. Montgomery Jr.C. Magnusson. Ed. M. 157 M. Basch. M. Phys. Komáromi. S. Thompson (World Scientific. Chem. Millam. Petersson. H. L. Phys. 189 H. Peslherbe. M. and M. A. FL. J. G. P. Schlegel. Chem. M. W. D. Bakken. Schlegel. B. Lett. Truhlar. Truhlar. J. Li. Kuppermann. J. J. E. 113. R. H. and M. 145 (1990). 10570 (1998). G. Soc. A. M. M. . Scuseria. Bakken. Phys. Li. A. Ed. Meth. S. C. Stamford. 77. A. Chem. 3. Clark. and H. and H. Hase. J. Wilson. Phys. Voth. Am. Singapore. in Modern Methods for Multidimensional Dynamics Computation in Chemistry. 1-3. M. V. D. 287 (1971). D. 111. Uggerud and T. J. W. Ed. 114. 1991). W. Frisch.172 G. A. Phys. S. Soc. 2041 (1970). T. Chem. Phys. Schlegel. Schleyer. Allinger. 228. D. Chem. Phys. A. J. 8694 (2002). Gasteiger. N. Thompson. Frisch. D. J. Phys. Ochterski. G. J. D. Chem. 4265 (1992). 10062 (2000). 93. Thompson. Helgaker. H. J. Kollman. Chem. 1985). Chem. Chem. G. L. Schlegel. R. J. 8773 (1999). Baer (CRC. Ed. 9758 (2001). 179 Advances in Classical Trajectory Methods. B. and B. Am. Uggerud. Boca Raton. 109. Millam. S. Voth. Phys. 5955 (1982). Chem. Chen. 107 (1935). Millam. Ed. Daniels. Garrett. Hase. Schlegel. and M. Hase. Malick. G. Phys. J. W. Phys. L. 114. W. V. Lett. Raff and D. E. and B. Schlegel. E. 1840 (1971). L. 173. Chen. 53. CT. J. Truhlar and A. M. B. A. Hase (JAI. J. W. F. 10. J. Iyengar. L. L. J. D. J. v. Chem. 183 K. 3800 (1999). 117. G. Chem. 111. 173 174 175 176 177 H. Phys. Skodje. Eyring. Frisch. J. W. Jensen. 178 L. M. Montgomery. Vol. Chichester. G. Scuseria. Chem. Comp. G. K. P. B. H. 181 182 T. 436 (1994). Bunker. in Theory of Chemical Reaction Dynamics. 188 H. E. Iyengar. and H. and B. J. J. 184 185 W. J. Millam. L. and G. 1998) 3506. J. L. 186 J. 1998). 180 D. X. Helgaker. in Encyclopedia of Computational Chemistry. T. and P. Bolton. L. Phys. 187 X. Schaefer III. Millam. Schreiner (Wiley. R. 89. 80. Phys. A. Yamaguchi. Binkley. and M. Chem. 123. T. Lett. Millam. A. J. J. 90. Chem. Chem. Lett. Chem. Chem. Frisch. Y. E. J. and H. W. and J. Frisch. E.. Stratmann. Frisch. Phys. 10175 (1997). Chem. Frisch. T. Handy. F. Phys. Scuseria. M. S. Lett. Schlegel. Phys. A. Bacskay. H. 922 (1988). Phys. A. J. W. G. 84. A. 115. 204 W. P. J. F. 207 208 209 V. H. Allen. 199 R. Yamaguchi. Schlegel. Barone. M. Chem. Meyer. 531 (1986). B. E. Phys. N. S. J. J. Pulay and W. Comp. Burant. J. Adams. Chem. Binkley. Phys. S. J. H. in prep. M. Gready. Chem. Miller. C. 196 197 198 R. 108. Iyengar. 84. in Potential Energy Surfaces and Dynamical Calculations. 57. Phys. Chem. Frisch. and M. G. Scuseria. 99 (1980). Gaw. Rev .190 S. Frisch. Johnson and M. F.Binkley. 216. J. J. J. B. Phys. New York. D. 191 192 R. J. 133 (1993). Phys. Hush. G. M. Phys. Johnson and M. and J. S. 194 M. 187 (1988). B. Pople. Chem. Chem. (2003). 373 (1996). 2150 (1979). L. S. Miller. Jaffe. Phys. A. Phys. Chem. D. B. Voth. Phys. . M. Clabo. J. and J. Chem. E. 100. Frisch. and H. Truhlar (Plenum. Quant. W. B. G. 88. K. J. Yamaguchi. 250. 167 (1981). Page and J. J. J. 55. Remington. H. 106. Chem. Phys. J. and N. 205 W. J. 1976 (1984). J. G. 1981) 265. Kormornicki and R. and J. Phys. H. Parrinello. Schaefer III. 7429 (1994). Amos. 10291 (2001). Chem. Miller. D. J. Chem. H. 467 (1978). Ruff. 6298 (1988). G. B. Raghavachari and J. 206 D. 2262 (1986). J. 195 N. Lett. G. McIver Jr. Ed. Chem. 193 Y. J. R. Car and M. Schaefer III. J. and Y. J. 185 (1984). Yamamoto. Schlegel. H. B. 72. 3337 (1972). 71. E. Vreven. Int. 200 201 202 203 A. Robb. 2471 (1985). Schaefer III. Y. C. Pople. 20. Chang. Phys. J. B. 129 (1984). Mol. and P. 10666 (1994). 789 (1974). Chem. E. 976 (1986). 93. L. L. Radium. Phys. Handy. 1990). R. J. 1028 (1962). 1419 (1980). Chem. Barone. 894 (1987). R. J. Chem. M. London. H. Chem. Chem. Comp. J. 226 227 228 229 230 F. J. R. Phys. 5634 (1990). 82. Chem. Chem. 3717 (1994). 112. 85. B. and J. 361 (1990). Dupuis. 365 (1995). Singh and P. Wolinski. H. Minichino and V. J. 8251 (1990). 11. Sanz. McWeeny. Doubleday. 93. J. J. C. . 101. 9 (Elsevier. and P. Barone and C. Chem. V. 41. Francl. Adamo. Comp. International Journal of Quantum Chemistry 43. A. 431 (1990). Comp. J. Barone. E. Barone. 94. Paris 8. J. Comp. J. Sadlej. Kollman. F. THEOCHEM 330. 5. Rev..210 C. Phys. Wiberg. J. Kollman. 8828 (1990). C. and V. E. Chem. J. Colwell. K. Phys. C. 487 (1991). 100. 223 224 S. J. Phys. U. 126. J. Merz Jr. J. 4959 (1991). Breneman and K. Rega. M. Phys. Cossi. Page. Chem. 8. Jørgensen. Ditchfield. Mulliken. R. N. Comp. Phys. N. M. in Theoretical Biochemistry: Processes and Properties of Biological Systems. 11. 225 J. 397 (1937). and A. Rice and N. Handy. C. Chem. Phys. Chem. Chirlian and M. 23. McWeeny. Am. Besler. Handy. Rice and N. Dodds. J. J. 222 J. Phys. S. McIver Jr. and J. 211 212 213 214 215 216 217 218 219 220 221 M. Phys. B. Mol. 91 (1992). D. C.. E. Sekino and R. K. Minichino. 27. 1833 (1955). 12. M. Amos. Theoretical and Computational Chemistry. Vol. J. M. 3235 (1985). Phys. V. R. Hilton. Olsen and P. J. A. Karna and M. Chem. W. Bartlett. Phys. Chem. J. Rice. S. J. P. C. C. New York. Soc. Pulay. Quant. Trucks. R. Chem. Bader. Handy. and G. Frisch. Walker. 113. Symp. 236 K. V. G. 89. Comp. 583 (1995). M. A 106. Gauss. Chem. J. 252. W. 194. Chem. Phys. Chem. J. Quant. Snyder. J. 3629 (1993). Ortiz. Phys. Lett. F. R. Lett. J. 223 (1993). and L. Cheeseman. 99. R. S. J. J. Phys. A. 1. 57 (1984). Chem. 23. Chem. Phys. 99. Cheeseman. T. T. and D. 53. and M. R. Scuseria. Helgaker. J. Peralta. Frisch. and N. A. Ortiz. Stephens. F. Contreras. E. J. Bak. Phys. Grafenstein. 3847 (1993). Symp. in prep. V. 246 247 248 249 250 V. and T. Phys. 321 (1989). V. M. 5497 (1996). F. Keith and R. Devlin. Gauss. M. J. J. H. J. Cremer. Chem. Cheeseman. (2003). Phys. B8. J. J. Keith. 6348 (1988). V. and P. J. Cederbaum. Ruud. K. M. Phys. 13 (1993). Quant. P. 5607 (2002). Cederbaum. 52.231 232 233 234 T. Frisch. E. Stephens. Phys. S. 1 (1992). H. 237 238 239 P. Zakrzewski and J. Zakrzewski and W. Phys. T. in prep. J. E. Ortiz. 245 W. Sychrovsky. Phys. R. Ortiz. 1001 (1995). Cheeseman. W. 99. Keith and R. 241 J. Chem. Schirmer. 14. . V. 431 (1988). Bader. J. Chem. Jørgensen. and H. J. Chem. Helgaker. V. Rep. G. Int. Contreras. 9402 (2000). Chem. J. 242 J. J. Phys. Phys. 243 244 J. R. 211 (1996). Frisch. Jensen. Peralta. 290 (1975). P. J. Int. von Niessen. 113. von Niessen. A. L. Chem. J. J. G. A. J. 1311 (1970). J. Int. J. PCCP Phys. J. 240 V. (2003). J. J. J. F. C. Barone. J. Chem. L. Phys. and J. 22. Lett. Chem. W. Devlin. Chem. Chem. Comp. Watson. J. Chem. 210. 3530 (2000). 235 J. 104. 256 K. M. 763 (1974). D. S. and J. D. Chirality 14. Stephens. M. J. J. Gordon. 4595 (2002). and E. J. and P. 27. K. R. 107 (1967). 1944) 342. C. and G. Phys. Chem.-K. L. Barron. 258 J. Acta 90. J. J. Theor. Helgaker and P. Phys. 76. 453 (1969). Lett. van Gisbergen. . Eyring. Chem. 441 (1995). 87. Olsen. Phys. J. Helgaker. J. Atkins and L. F. L. R. Mol. 98. K. W. 10768 (1992). Frisch. 421 (1995). L. 533 (2002). Stephens. Mol. Phys. 352. 257 K. R. Tomasi. Barron and A. Autschbach. and P. A. Chem. 268 P. Phys. Cheeseman. A. P. Bak. 12. R. 6930 (2002). 95. (2002). Mol. A. Lett. Chem. Lett. Bortolini. Devlin. J. Helgaker. Cimiraglia. A. 1111 (1971). J. Chem. 169 (1980). Jensen. 20. 8873 (1993). L. Hansen. Org. and C. Walter. Enantiomer 4. Phys. D. New York. Bak. Jørgensen. Bak. and H. Buckingham and G. K. Olsen. Chem. Bak. A. T. 259 J. Koseki. Chim. J. 263 H. E. F. Phys. D. Abegg and T. 455 (1999). Amos. Frisch. Quantum Chemistry (Wiley. 260 261 262 A. Devlin. 30. Mol. Cheeseman. Lett. D. Ruud and T. Kimball. T. J. 14. Helgaker. Rosini. Ha. 16. Phys. Ruud. Acta 90. O. 269 P. J. W. 4. Ziegler.251 252 253 254 255 P. S. T. Helgaker. 23 (1982). K. M. Ruud. and P. J. Chem. and M. 579 (1975). Buckingham. Ruud. Phys. Adv. T. P. 63 (1968). Schmidt. Mol. Jørgensen. E. M. 2595 (1991). Jørgensen. E. S. Chim. W. Theor. Chem. J. 96. T. K. Buckingham. Phys. Phys. Hansen and K. Besse. J. Abegg. L. Persico. D. J. 116. Longuet-Higgins. Phys. W. 264 265 266 267 P. Jørgensen. Baerends. Phys. J. Bak. 115. K. J. Ruud. J. Springer Series in Chemical Physics 40 (Springer-Verlag. J. J. Ruud. Kuchitsu (Blackwell. W. N. M. . Frisch. and J. Stephens. 4776 (1991). Frisch. J. Phys. Helgaker. Miertus and J. J. Ed. 287 M. Curl Jr. F. Chem. Wiberg. 11080 (2001). J. 275 E. Cvitaš. 571 (1994). Phys. M. 165 (1994). 1985) 184-193. Chem. Wong. Wiberg. Cheeseman. 117 (1981). 271 P. M. J. Tomasi. W. Phys.. 99. J. Wong. T. Tomasi. 8991 (1991). Boca Raton. L. Phys. 327 (1996). T. and M. Homann. 255. Brown. Chem. High-Resolution Spectroscopy of Transient Molecules. Chem. Hirota. Tomasi. 55. T. 201 (1996). F. 2nd ed. 272 R. 371 (1991). and M. 276 277 278 279 280 V. Chem. Rega. Chem. Tomasi. J. J. Barone. Hirota. 3032 (1997). K. B. Am. Mol. Soc. T. 282 283 M. 1645 (1992). Phys. H. J. Cossi. and J. K. Mennucci. Devlin. and J. Chem. Hirota. Am. J. 95. B. K. Shida. Cancès. B. Soc. J. 105. 273 E. 523 (1992). 9. and T. Mol. CRC Press. Olsen. 66. 288 M. Chem. 2804 (1996). Scrocco. Oxford. P. W. Chem. 148. J. 284 M. 105. R. Am. Phys. Lett. 285 286 S. Spec. Chem. and J. Frisch. S. dist. F. Miertus. Wiberg. Barone. Pickett. Frisch. Cammi.270 T. Pure & Appl. Berlin. 107.. and K. and V. Wong. J. Phys. R. 11060 (1996). Chem. 262. Soc. N. Barone. B. Units and Symbols in Physical Chemistry. J. Helgaker. Phys. A 105. Mills. Gauss. and M. Neese. K. and K. 274 Quantities. Lett. E. 281 M. J. and M. Phys. Chem. Hougen. V. B. 113. 1993). K. 65. Wong. Phys. Cossi. 585 (1965). Jørgensen. M. Chem. J. Wiberg. I. Faraday Discuss. J. K. Chem. J. 239 (1982). Frisch. 5356 (2001). M. Phys. Kallay. 114. 114. and N. W. and J. Barone. Tomasi. and V. Cappelli. Wiberg. B. Mol. Chem. 291 292 293 294 V. B. Frisch. M. and M. Phys. 253 (1998). Chem. Mennucci. B. J. Cammi. 286. Mennucci. A 102. J. T. A. and M. A 106. Cossi. and V. B. Cammi. V. Scalmani. and J. S. Phys. Chem. G. J. Frisch. G. Phys. Chem. Phys. PA. 306 J. R. Cossi. B. A 106. Chem. A. B. Phys. K. Gabriel. C. R. 3210 (1997). 5631 (2000). Barone. Phys. Inc. J. J. (Theochem) 464. Foresman and Æ. 6102 (2002). Chem. Corni. Tomasi. Phys. Chem. 1995 (1998). 4. 117. Tomasi. Tomasi. 12331 (2002).Tomasi. J. 308 J. Cossi and V. Cancès. M. Tomasi. Cancès. and J. J. G. 4708 (2001). Cossi. Phys. Barone. and J. Phys. Cheeseman. Stephens. Corni. Phys. Chem. M. Mennucci. Chem. Phys. Cammi. N. and S. Cammi. Chem. Lett. J. 10506 (1997). Barone. and V.289 V. Rega. and P. J. Barone and M. 43 (2002). B. Phys. Mennucci and J. Tomasi. R. J. Barone. 304 C. Scalmani. Chem. Tomasi.. Phys. Rega. R. Chem. N. Mennucci. Comp. Mennucci. R. 114. Chem. 1996). 5295 (1999). Cappelli. Chem. 5697 (2002). Tomasi. 2nd ed. Phys. A 104. Mennucci. Foresman. and J. Cossi. J. J. 107. Mennucci. Barone. and J. and E. and J. Rega. Pittsburgh. Chem. V. M. 9100 (1999). J. Chem. J. Phys. Comp. Cossi and V. Struct. . 19. 16098 (1996). 106. B. Devlin. Cossi. 290 M. 296 297 298 299 300 301 302 303 R. Cossi. J. Phys. B. Mennucci. M. J. 111. Tomasi. Barone. Frisch. J. Barone. N. F. in press (2003). A 103. B. 2427 (2000). 305 B. M. M. Chem. M. Chem. J. Phys. J. E. J. Cammi. Cossi. J. 100. 295 J. Tomasi. Barone. 5691 (2001). 404 (1998). Chem. V. J. M. B. S. J. 5151 (1997). (Gaussian. J. 211 (1999). Exploring Chemistry with Electronic Structure Methods. 307 J. Scalmani. 112. Robb. 115. Cossi. B 101. Keith. Snoonian. Hehre. Ratner. Chem. A.-P. 322 J. Am. 939 (1980). J. and J. W. M. Chem. 2657 (1969). M. M. Stewart. A. Curtiss. Francl. 1223 (1998). J. J. Pietro. Binkley. Chem. A. J. W. and J. Chem. Chem. Phys. J. 880 (1987). 2193 (1988). Windus. Curtiss. J. Chem. W. Dobbs and W. M. S. 724 (1971). Hariharan and J. Blaudeau. Mol. J. J. 8. 54. Pople. Phys. J. 11. 104. J. 359 (1986). A. Radom. D. J. Hehre. M. S. C. Schleyer. A. Binkley. 5142 (1976). Chem. Hehre. A. 27. Chem. Comp. 104. R. J. Pietro. Acta 28. Pople. 107. Comp. Pople. 163 (1980). W. J. Pople. J. J. G. A. L. S. M. Lett. Pople. 213 (1973). Petersson. J. Pople. 310 J. Hehre. Phys. J. Soc. 109. Dobbs and W. Shirley. and J. Ditchfield. Phys. B. 22. R. Chem. Phys. and J. A. Hehre. Al-Laham. Phys. M. Hehre. A. Defrees. Phys. J. Chem. Pople. Soc. Chem. and M. J. D. J. 56. Theo. A. J. J. 325 V. 209 (1974). A. 8. 64. J. Am. Dobbs and W. 5039 (1982). R. Chem. A. Chem. 311 J. 51. A. P. 324 R. Hariharan and J. D. DeFrees. L. Gordon. J. S. A. McGrath. Binkley. J. C. 327 G. J. K. Hehre. Pople. v. D. D. Rassolov. F. 976 (2001). J. Pople. J. Comp. K. and L. J. Binkley. S. S. Soc. Collins. P. 7. J. P. J. P. A. A. 89. 326 V. and J. Pople. 3654 (1982). S. 2257 (1972). W. 2797 (1982). Chem. J. J.309 W. Comp. Hehre. Pietro. J. J. Hehre. 861 (1987). A. A. P. and L. J. Phys. 323 M. A. Gordon. A. Chem. J. Ratner. Binkley. Comp. and J. Ditchfield. Curtiss. Phys. 102. J. R. A. Gordon. Pople. . and T. Chim. W. and W. W. Pople. M. S. 5016 (1997). Chem. Chem. 76. and L. C. 1206 (1990). 77. Am. J. Binning Jr. J. Tensfeldt. C. A. Rassolov. T. and W. A. 312 M. Francl. Bennett. 314 315 316 317 318 319 320 321 K. Phys. 313 W. Redfern. Mantzaris. Hehre. Phys. Chem. 343 P. J. L. M. Phys. Preuss. Preuss. Basch. Fuentealba. M. Phys. Stoll. Chem. Phys. McLean and G. 1976) 1-28. P. 1062 (1989). Stevens. L. S. 82. Chem. 93. and J. C. Jasien. Stevens. Schwerdtfeger. Seeger. Dunning Jr.-P. J. Phys. 418 (1989). Wadt. Trucks. Chem. Preuss. 330 331 332 A. Hay and W. 3 (Plenum. J. Can. Curtiss. 81. v. P. Pople. J. Ed. W. 81. P. Binning Jr. 5639 (1980). J. Lett. J. R. Szentpaly. J. Schwerdtfeger. H. Radom. Vol. Stoll. in Modern Theoretical Chemistry. v. and L. 334 M. Szentpaly. P. Phys. 66. Phys. 1033 (1970). 345 P. R. Phys. Basch. Chem. P. 336 337 A. 338 W. R. J. and H. v. 555 (1982). J. Smedly. and P. Phys. Phys. Rappé. Preuss. T. P. 70. Schaefer III. 329 R. 103. Hay and W. J. Raghavachari and G. P. Chem. 612 (1992). R. J. Goddard III. H. J. W. Chem.. Fuentealba. 511 (1991). J. J. J. J. H. Flad. Hay. Fuentealba. Phys. Davis. 6104 (1995). 346 H. J. New York. K. Szentpaly. J. J. Chem. Stevens. Krauss. Phys. G. v. Chem. Phys. P. H. 4377 (1977). N. D. F. 94. Stoll. Szentpaly. 1323 (1983). Cundari and W. A. J. Chem. 85. Phys. Phys. R. 270 (1985). Phys. 344 L. P. J. and P. Chem. Chem. A. S. J. J. Krishnan. H. E. H. Chem. and H. 82. K. 299 (1985). 1662 (1981). Stoll. Hay. H. 335 T. H. Chem. B 16. Chem. and J. Binkley. McGrath. 72. 82. Hay. 91. 2732 (1984). 52. J. . 650 (1980). 333 L. R. J. 98. A. 89. Wadt and P. 6026 (1984). Wadt. Blaudeau. and L. Chem. and H. Krauss. W. and W. J. 284 (1985). 339 340 341 342 T. H. 72. J. Lett. Phys. 5555 (1993). Radom. Wachters. McGrath and L. Chem. Chandler.328 A. Fuentealba. Haeussermann. H. Stoll. 1211 (1993). Preuss. P. 5852 (1993). Chim. 94. Preuss. U. G. M. Reidel. Stoll. H. 90. Dolg. D. Savin. Kuechle. Neumann. Phys. Chem. Stoll. and H. 94. 173 (1989). Preuss. 78. 356 W. M. 354 D. and H. H. Preuss. 1287 (1985). 97. Stoll. M. H. and H. H. H. H. Dolg. Chem. Kaupp. and H. Wedig. Phys. Theor. Preuss. Kuechle. 359 A. Schleyer. Stoll. 1360 (1991). Dolg. Mol. P. J. Preuss. Kuechle. Acta 75. M. Stoll. Stoll. J. 100. Phys. Wedig. Dolg. U. and H. 1052 (1996). Dolg. J. 1762 (1989). Preuss. Phys. and P. 65. Igel-Mann. and H. v. B 18. Preuss. Dolg. 353 M. Bergner.-J. and H. J. Phys. M. Bowmaker. Phys. H. J. Phys. Stoll. Chem. Preuss. Phys. and H. J. in Quantum Chemistry: The Challenge of Transition Metals and Coordination Chemistry. W. and H. 349 350 351 M. Preuss. Nicklass. J. Phys. Flad. H. Fuentealba. 8942 (1995). Leininger. L.-S. 360 361 362 363 364 M. H. G. Phys. A. Dolg. H. 365 T. 441 (1993). 355 M. A. Schwerdtfeger. Preuss. Dolg. . H. Stoll. U. Andrae. and H. Preuss. J. M. 91. Dolg. Schwerdtfeger. Chim. Schwarz. Stoll. Mol. Stoll. 7535 (1994). 1730 (1989). Preuss. H. Ed. and R. M. Chem. J. 86. 102. Stoll. 357 M. Chem. Chem. E. Veillard. Chem. M. Kuechle. 74. C. H. Dolg. M. Haeussermann. 3011 (1991). and Dordrecht (1986) 79. 352 P. Chim. Stoll. Phys. Phys.347 P. and H. M. Preuss. Preuss. Dolg. and H. Mol. and H. Stoll. 1245 (1991). Phys. Chem. 358 M. H. Acta 85. Boyd. M. and H. Stoll. 1431 (1993). Stoll. Dolg. 80. Dolg. M. Phys. J. Chem. Preuss. and H. Szentpaly. Pitzer. 1321 (1988). A. W. 1162 (1992). and H. H. 123 (1990). H. R. H. W. W. Phys. H. Stoll. Theor. J. Chem. Theor. Fulde. and P. W. v. Dolg. 97. Nicklass. Acta 77. 866 (1987). A. 348 U. A. Dolg. Stoll. Mol. 105. Dolg. H. D. V. V. Dolg. P. Mol. Dunning Jr. Jorge. Trsic. 1358 (1993). 377 V.. 90. Phys. (Theochem) 388. Lett. T. R. H. and A. 6796 (1992). Horn. D. de Castro. Barone. R. 108. 369 R. Struct. and A. J. J. 7410 (1994). J. 1996). 5829 (1994). J. A. Trsic. Cao and M. C. R. V. J. E. Phys. 100. Y. Chem. 86. B. 97. Phys. F. Jorge. 115. A. Truhlar. and J. Chem. 370 T. J. Costa. M. A. Chong (World Scientific Publ. Singapore. 380 F. 18. 2571 (1992). 139 (2002). 20. 384 J. 376 R. Phys. and M. Welch. D. Dunning Jr. 5043 (1987). da Silva. J. E. Woon and T. Dreizler. R. Phys. Chem. and T. Mol. Woon. Int. da Costa. E. 281 (1996). J. Chem. Phys. Co. Trsic. 216. H. and R. d. Phys. Peterson. Chem. Dunning Jr.. Struct. da Silva. 260. E. Phys. 339 (1997). M. Theor. Wilson. Trsic. F. 1987 (1987). F. Easton. 91 (1989). Part I. J. Mol. Quant. C. F. 98. R. H. Chem. 379 F. and D. de Castro and F. J. Mohallem. in Recent Advances in Density Functional Methods. Phys. 514 (1996). (Theochem) 581. Huber. H. R. J. Comp. Ahlrichs. G. M. Chem. Symp. Dunning Jr. H. Mohallem and M. A.. 371 K. 7348 (2001).. Giesen. Schaefer. Chem. Chem. 378 E. T. B. Y. 372 A. 317 (1997). 62. J. 381 382 383 A. H. and R. R.. de Castro. Chim. 373 374 375 E. H. 62. Acta 93. Chem. and M. B. X. 100. Chem.366 367 368 X. Cao and M. 5225 (1998). J.. Phys. and T. E. Davidson. Ahlrichs. Schaefer. A. and R. da Silva. Cramer. 1565 (1997). Ed. Dunning Jr. Phys. 45 (1986). E. R. E. Mohallem. D. Chem. F. Chem. J. J. J. Harrison. . H. J. Kendall. van Mourik. Phys. 96. Jorge. Phys. Mol. M. Dolg. 1007 (1989). E. J. Andzelm. C. and G. (Theochem) 187. Theor. C. Dewar and Y. 96. Macbeth. Chem. and E. J. Hutter. Schleyer. and J. in Modern Methods for Multidimensional Dynamics Computation in Chemistry. Mol. Acc . Jie.1606-1611). S.385 D. Jie. B. Chem. S. C. and W. Wilson. and M. Thompson (World Scientific. and P. 1318 (1991). G. E. 95. Andzelm. Nieuwpoort. 421 (1978). D. Comp. J. G. Cheng. Chem. Chem. Lippert. G. R. Can. Phys. Hase. Nieuwpoort. Pople. Chem. J. Freunscht. Dewar. M. and M. J. J. Salahub. M. K. J. and P. Organometallics 7. Comp. Quant. Dobbs. 477 (1997).40-107 (London. C. Godbout. B.iv. 522 (1988). Chem. 388 C. J. Mol. J. J. Dewar and A. Organometallics 7. Ed. Merz. 29. J. W. Spitznagel. 560 (1992). Schlegel and M. Int. J. and M. D. 83 (1995). 386 D. and D. Frisch. J. J. Klein. W. E. J. 57. Lippert. J. Frisch. 124 (1999). M. 70. 294 (1983). 3265 (1984). 3881 (1990). Kudin. H. S. K. Clark. Sosa. Parrinello. 513 (1988). 4. Frisch. J. 14. Elkin. Scuseria. M. Zoebisc. Singapore. Koch. Int. Ch em. 1998) 143. Dewar and K. 387 N. G. Chem. Holder. 103. S. Peshlherbe. 80. Bolton. Anders. A. in prep.-C. Lett. 14. E. Hutter. c. R. Organometallics 9. Trucks. Chem. S. 635 (1978). 403 K. W. C hem. 389 T. J. L. and G. Silver and W. J. M. L. Phys. H. M. G. R. (2003). Binkley. III. Martyna. M. Phys. 92. J. Shakespeare. Wimmer. v. Inorg. and E. Dixon. . S. D. Wimmer. C. J. L. Silver. M. 508 (1990). Parrinello. 1301 (1993). W. 1 (1989). J. Struct. G. A. Chandrasekhar. J. Yuan. Chem. 390 391 392 393 394 395 396 397 398 399 400 401 402 M. 6630 (1992). Dewar and C. Quant. Phys. Phys. M. 54. J. S. 255. I. 102. Soc. A. Phys. . B. 1966) 253. Wang. Phys. O. G. Robb. 414 415 416 417 418 S. S. Hamilton and P. A. 110. M. Löwdin (Academic Press. Bottoni. Soc. J. Bernardi. Soc. Am. M. Soc. Chem. 483 (1990). 99. Swamy. Garavelli. M. J. Peslherbe. Mezey. 671 (1996). Robb. Chem. Matsunaga. A. A. Am. 410 I. 4916 (1989). and H. M. F. A. Bearpark. M. 5081 (1992). Chem. K. 524 (1992). 189. Varandas. A. J. M.-H. F. Koseki. 119. and M. Robb. Rev. I. J. J. F. 419 S. Robb. Olivucci. 411 G. M. F. Frisch. and M. F. Tonachini. Robb. A. Gordon. Pulay. Chem. J. Chem. 10430 (1998). A. 112. N. 406 F. G. M. Phys. H. J. H. T. Foster and S. Komornicki. Program Exchange 16. J. Mod. 5993 (1988). M. 409 F. Robb. 412 T. Am. 32. Olivucci. Bottoni. N. A. P. 420 J. R. J. J. 12687 (1997). Schlegel. Pipek and P. Clifford. 90. Phys. Ph ys. Bernardi. Chem. Quantum Chem. Chem. Chem. in Quantum Theory of Atoms. Hase. and R. and M. Bernardi. Bottoni. M. Hillier. 3050 (1988). 12764 (1995). R. A. and H. New York. 4926 (1988). M. Pulay. 2121 (1994). Soc. and M. Ed. 408 F. Bernardi. Boys. Am. D. 413 S. Koseki. and A. 88. J. Chem. Schmidt. Chem. 112. 407 F. Org. F. and N. S. W. 2106 (1990). 57. J. M. Schmidt.404 W. Ragazos. Duchovic. S. Phys. J. and A. 90. 320 (1996). M. J. Robb. Phys. Olivucci. V. Palmer. Boys. B. Bernardi. Chem. J. Schlegel. Ragazos. F. Guest. Olivucci. 300 (1960). L. K. J. Robb. Field. M. Phys. M. I. A. 405 M. and G. J. W. Venturini. H. 116. S. Vreven. H. J. Bernardi. N. J. Schlegel. Lett. I. Tonachini. 110. A. Lu. Hu. P. Schlegel. M. Linde. Palmer. Soc. A. Am. Venturini. Molecules and the Solid State. and M. M. J. Lim. M. Gordon. Wolfe. Chem. Robb. H. Chem. Chem. Bofill and P. Am. Lett. B. 3637 (1989). B. Bernardi. X. A. Phys. 38. 21 (1994). Chem. McWeeny. 19. L. Int. 429 E. 219. Duran. J. 553 (1970). Stevens. 199 (1989). 246. Head-Gordon. M. Mills. Smith. Simon. C. F. L. Maurice. 32. Phys. Head-Gordon. 25. and J. Oumi. and A. and T. S. Symp. J. W. R. 114 (1995). Elmsford. R. Rev. 717 (1973). 11024 (1996). Chem. Quant. M. J. P. R. Phys. Phys. 1028 (1961). Rev. Dodds. Purvis. Oumi. G. Lipscomb. and J. Mod. Chem. Phys. Raynes. 128. Rico. 23. 63 (1986). 34. 1986). R. Int. 881 (1984). 14. J. R. Boys and F. T. Chim. Lee. Chem. G. 5043 (1983). Lett. J. Quant. Gerratt and I. Cohen and B. and M. Chem. Chem. Acta 69. 441 J. Theo. P. 1779 (1977). J. Chem. Sadlej. 431 432 433 434 435 436 437 438 439 Pure and Applied Chemistry 51. Phys. 105. Dannenberg. M. M. and W. R. 78. Phys. J. Boys. D. Lett. M. 1980). 296 (1960). McWeeny. Phys. McWeeny. Pitzer. Quant. Mol. D. J. Taylor. 430 R. . Weast. Rev. Phys. Riley. Hall and C. Chem. Phys. Phys. J. Lee and P. Pure and Applied Chemistry 2. 427 428 M. J. FL. 1 (1979). Mol. McWeeny. Int. G. CRC Handbook of Chemistry and Physics (Chemical Rubber Company. Pulay. Bartlett and G. M. R. Mol. 611 (1977). Mohr and Taylor. Mod. CODATA Bulletin (Pergamon. Smith and G. J. 426 M. 516 (1978). F. J. N. C. p. NY. Bernardi. M. R. The 1986 Adjustment of the Fundamental Physical Constants. R. S. Hall. J. N. Dodds. J. Phys.421 422 423 424 425 S. T. Boca Raton. R. 1719 (1968). 440 J. BG6 (August 2000). 32. Chem. 49. 550 (1963). 335 (1960). 33. Taylor. J. 96. G. D. J. C. Chem. A. G. Osamura. 1991). J. Perdew and Y. Parr and W. 197. T. Chem. J. J. W. 6671 (1992). Yamaguchi. and H. Jackson. Chem. Frisch. 97. J. G. Pederson. Phys. and H. Chem. Phys. Becke. 453 454 455 456 457 C. J. Chem. Gill.442 Y. D. 1253 (1981). Chabalowski. Perdew. Press. D. Wimmer. Wiberg. J. . B. Chem. Phys. Andzelm and E. 447 K. Chevary. Rev. Stephens. 499 (1992). R. Rev. 98. 444 445 446 C. Chem. J. Ed. E. Oxford. Pople. Chem. Ed. J. M. J. R. S. A. D. Chem. Y. M. B 46. 2919 (1981). 99. Phys. 75. C. J. Phys. Phys. 459 P. C. 448 The Challenge of d and f Electrons. Osamura. Schaefer III. M. J. Singh. Zerner (ACS. J. 7528 (1992). Faraday Discuss. Labanowski and J. Salahub and M. 1989). 671 (1992). Frisch. K. Lett. Breneman. Diercksen and Sadlej. B. A. J. 75. 96. Phys. Becke. 452 Density Functional Methods in Chemistry. C. 443 Y. 77. Ashvar. 2155 (1992). Fiolhais. 8004 (1993). F. P. LePage. 1989). C. Phys. New York. Phys. W. Roos. Andzelm (Springer-Verlag. Phys. and M. 103 (1994).C. H. Hadad. and Sadlej. Vosko. Lett. M. Schaefer III. Wang. 59. Lee. Frisch. Density-functional theory of atoms and molecules (Oxford Univ. 96. 29 (1981). 383 (1982). and C. 97. Yamaguchi. and M. 449 R. Diercksen. S. 9173 (1992). Scuseria. Yang. Y. Devlin. J. Chem. Chem. 450 J. Phys. K. Sosa and C. J. A. 458 P. Johnson. F. Phys. 109. F. A. 388 (1984). D. Chem. Dykstra and P. F. 1280 (1992). J.. Phys. and M. 451 J. G. E. Jasien. J. B 45. Washington. J. P. Phys. 13244 (1992). Wilk. C. Chabalowski. Chevary. 419 (1998). Phys. in Electronic Structure of Solids ‘ 91. Barone. J. J. 461 462 A. Perdew. B 33. Phys. 467 468 469 470 471 472 473 474 475 476 477 478 479 J. K. 1200 (1980). Handy and A. W. L.460 P. M. A 38. Berlin. Phys. J. Chem. A. Mol. Rev. and M. 116. Phys. 8822 (1986). . A. J. J. S. Dobson. Phys. K. P. 785 (1988). Stoll. J. P. F. and M. H. Perdew and A. Chem. Rev. Wang. M. Lee. S. Lett. J. Phys. J. P. 157. K. B 46. 9003 (1995). 16533 (1996). Lett. B 48. P. Wang. 5933 (2002). Phys. A. and C. J. S. Barone. R. Ed. Vosko. Perdew. K. Preuss. Chevary. A. Phys. Lett. Chem. Phys. Perdew. Adamo and V. Stephens. Phys. A. D. Nusair. Rev. Burke. and H. Perdew. M. B 37. 466 J. Adamo and V. Miehlich. Perdew. Phys. B 23. Barone. Vignale. Rev. Singh. Ernzerhof. Burke. 403 (2001). W. Ed. K. 98. and C. Becke. Perdew. and Y. and C. and M. D. and R. Mol. Burke. J. and M. P. Ziesche and H. C. Perdew. Yang. Perdew. Fiolhais. J. C. Rev. Fiolhais. 1991) 11. Pederson. F. Singh. N. H. 99. 99. Phys. G. Vosko. G. P. Eschrig (Akademie Verlag. F. Ricca and C. Phys . Bauschlicher Jr. J. H. H. Parr. J. Phys. Phys. 433 (1996).. J. M. J. Zunger. 664 (1998). Phys. Burke. B 54. W. P. Gill. J. Cohen. C. Rev. Rev. Rev. P. D. Comp. Phys. J. Jackson. Adamo and V. 464 J. Ernzerhof. Savin. 465 J. Jackson. 3865-3868 (1996). 77. (1993). Chem. 5048 (1981). 19. A. 89. Das (Plenum. P. (1992). C. J. and Y. 77. Vosko. Chem. B. 58. J. J. 1998). in Electronic Density Functional Theory: Recent Progress and New Directions. P. 108. Devlin. 463 K. R. 3098 (1988). P. Frisch. Can. P. 11623 (1994). Rev. Pederson. 3865 (1996). 200 (1989). Chem. Barone. N. 485 486 487 488 C. C. 9624 (1998). Phys. 483 A. Phys. Greengard. Phys. Papousek and M. Van Voorhis and G. Theor. Chem. 489 A. 9413 (1997). 274. Chem. J. C. C. G. N. New York. 114. L. and N. Lett. Adamo and V. Science 265. 498 D. Doltsinis. Durig (Elsevier. B. P. 400 (1998). 9233 (2001). Cambridge. J. 909 (1994). H. J. 107. . Chem. J. D. 1988). 108. 1982). C. J. Phys. Ayala and H. Frisch. Montgomery. J. Phys. J. Tozer. 7560 (1998). Phys. Schmider and A. 484 A. Greengard and V. D. Phys . Hernandez. J. 104. R. Y. see also the supp. Schlegel. Boese. J. 8554 (1997). A. Handy. Wilson. 491 L. 1976). Rokhlin. M. D. 108. D. J. Willets. D. J. Boese and N. J. 115. and D. Phys. Comput. 62 (1990). Chem. Bradley. 496 497 A. 482 F. 242 (1997). Petersson. and A. McQuarrie. T. B.480 481 A. Tozer. and M. Chem. Phys. Chem. Acc. Chem. Statistical Thermodynamics (Harper and Row. 1040 (1996). 109. Austin. Vibrational States (Wiley. J. Becke. Baboul and H. P. and G. 107. J. 325 (1987). J. E. 5497. 6264 (1998). N. The Rapid Evaluation of Potential Fields in Particle Systems (MIT Press. 1973). Chem. Jayatilaka. A. Scuseria. 107. Aliev. D. material: EPAPS Document No. 172. J. A. D. A. 490 L. New York. Ed. Phys. Chem. 494 D. S. Greengard. H. 495 W. Phys. 1670 (2000). J. in Molecular Vibrational Spectra. J. E (2001). Chem. Chem. 73. Sprik. L. Phys. Hamprecht. A. MA. Becke. R. Phys. Chem. T. Handy. 112. J. London. Miller. Becke. 180 (2002). Handy. Handy. R. Cohen. Califano. 492 493 L. J. 109. Chem. Lett. Schlegel. Russian Acad. 20. S. A 32. Phys. Phys. A. Hoffmann. Math. Douglas and N. L. R. 114. Gordon. Murphy. R. Phys. R. Conf. Inorg. A 33. 516 517 518 519 520 521 V. J. Zh. Phys. Pyykko and L. J. I. W. 1397 (1963). 40. A. 1968). 2474 (1964). J. Phys. Tetrahedron 22. Lebedev. A. Lohr. (NY) 82. J. 17. Rev. M. 587 (1992). J. Eqn. Phys. R. A. Fitzpatrick and G. 40. Phys. V. M. A. 1950 (1981). B. Chem. Ann. and D. 1 978. 521 (1966). J. Diff. W. P. Chem. Vychisl. Hoffmann. Chem. Ann. J. deJong. 361 (1963). Mat. Mat. Phys. .499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 S. 1770 (1985). Rev. Hoffmann. Lebedev. 89 (1974). 16. B 31. J. 139 (1986). Novosibivsk. Fiz. 48 (1975). Zh. Chem. Chem.. 4253 (1967). Mat. A 39. Hoffmann. 3742 (1986). Rev. M. L. 6016 (1989). Goddard III and L. N. A. Harding. Comp. B. Chem . Chem . A. 15. Phys. A. J. Hess. M. 45. Thermochemical Kinetics (Wiley and Sons. Am. Phys. 89. Dokl. Math. V. Rev. 293 (1976). Harrison. J. Phys. Sobolev (Nauka. Sci. Math. in Proc. 363 (1978). Chim. 88. R. D. Hess. Harris. I. Inorg. 3226 (1998). Hoffmann. Ed. Dixon. Laaksonen. Kroll. J. Mat. 39. 2480 (1964). Krack and A. Pople and M. Koster. G. B. Chem. Phys. 4892 (1984). 515 V. P. Lebedev. Jansen and B. S. Acta 111. Novosibivsk. Pyykko and L. McLaren. J. Hess. 1980) 110-114. 108. Numer. 48 (2001). Vychisl. Phys. I. Fiz. R. Chem. 2745 (1964). New York. Soc. 756 (1985). Chem. W. H. 40. I. Benson. Rev. Lebedev. Ö. Csaszar and P. Eötvös Loránd University and Hungarian Academy of Sciences. J. J. Chem. Jørgensen. 523 L. J. Comp. Fogarasi. 529 H. Ozment. G. and M. Bannerjee. Chem. M. Ed. 52 (1985). L. 533 534 535 536 537 538 539 Ö. F. G. J. 2745 (1983). J. D. H. 4. Simons. Int. Dyall. Budapest. THEOCHEM 573. E. and H. 524 525 526 527 528 E. 213 (1996). Phys. Phys . 111. and P. Phys. J. F. 2856 (1992). 111 (1976). Visscher and K. Chem. and J. Fogarasi. 10806 (1999). Phys. G. Schlegel. Boggs. Pang. . Mol. Trucks. 207 (1997). Yarkony (World Scientific Publishing. 65. Chem. Atomic Data and Nuclear Data Tables 67. King and M. Singapore. Dupuis. and J. J. 21. Adams. Chem. Chem. J. Chem. 12. H. J. Rys. 541 P. J. N. J. P. Chem Phys . Frisch. 530 J. B. Comp. J. 2800 (1981). and H. Sadlej. 87. Chem. King. 1995). 257. Miller. J. 532 Ö. H. G. J. J. 540 P. Scuseria. Chem. Jørgensen. F. J. F. R. Yeager. Lett. J. Phys. in Modern Electronic Structure Theory. 154 (1983). in prep. Barysz and A. J. Symp. Phys. Golab. J. 175 (1983). D. Taylor. 101. J. Farkas. 75. Soc. Phys. (Theochem) 114. R. 7100 (1998). 144 (1976). Phys. B. 531 P. J. Schlegel. 89. Simons. 2550 (1979). Farkas and H. Stratmann. Cheeseman. Chem. Quant. PhD (CsC) thesis. Schlegel. Rys. Schlegel. Struct. Frisch. T. B. Baker and W. 96. J. 31 (1984). 181 (2001). Cerjan and W. Shepard. 1995. 26.522 M. 243 (1992). B. 78. (2003). E. Comp. Am. J. 606 (1991). M. C. Pulay. Dupuis. J. A. Pulay. H. King. W. 109. and R. Chem. Pulay and G. Farkas and H. and M. Hehre. Dupuis. Chem Phys . J. Chem. Bofill and M. Chem. Chem. . J. Chem. E. 15. Pulay. N. 5463 (1990). 561 562 P. Hargitai. Phys. Schlegel. New York. Chem. G. Phys. Chem. P. J. Soc. Theor. 114. 90. Phys. P. G. P. Cancès. W. E. N. M. Jørgensen. Pople. 550 551 552 553 554 555 556 557 558 559 L. B. Chem. Soc. Comp. Chem. G. 1085 (1993). J. 1326 (1995). Johnson. J. Scuseria. Veress. Zakrzewski. 150. B. Fogarasi. Z. K. R. 1997) 465518. J. and R. Beratan. Am. J. 89. and P. 239 (1993). New York. 9. 1752 (1989). U. Kryachko (Kluwer Academic. and D. J. T. Lett. and J. Pulay. Gauss and C. M. Condon. Charney. 695 (1999). C. Rabuck and G. Acta. and E. Fletcher. and O. Chem. Calais and E. Lee. Wipf.-L. Ogretir and I. X. Csizmadia (Kluwer Academic. 1991) 167-185. A. 2204 (1998). Chem. 1979). Dolgounircheva. Kudin. 3. Chem. Phys. Salter. Rosenfeld. Physik 52. Rev. 432 (1937). A. Comp. 110. Comp. 543 544 545 546 547 548 J. Bartlett. E. H. Chem. Trucks. J. J. E. Ed. B. A. Chim. H. 206. J. 116. J. 3654 (1988). R. Phys. 14. 556 (1982). The Molecular Basis of Optical Activity (Wiley. Taylor. K. 66. P. Kondru. V. J. J. 549 J. 33 (1984). G. 120. Olsen. in Conceptual Perspectives in Quantum Chemistry. Phys. 94. V. and P. and Ö. Gill. Ortiz. R. Phys. 280 (1988). 8191 (1992). 560 H. in prep. Rendell. Chem. Zhou. 161 (1928). P. Taylor. G. in Computational Advances in Organic Chemistry. R. E. Farkas. The Netherlands. M. Cremer. Scuseria. Practical Methods of Optimization (Wiley. 8255 (2002). J. Am. (2003). Comp. 1 (1994).542 G. Phys. Jensen. 1980). Ed. G. Bofill. W. J. J. J. and J. A. Mod. McDouall. Lett. Comajuan. Baker. Schlegel and J. 16. J. A. Klamt. Lipscomb. G. Phys. Phys. Phys. Chim. B. AIChE J. Phys. Comp. Chem. S. NATO-ASI Series C 339 (Kluwer Academic. 225 (1977). 335 (1973). J. M. Soc.563 564 565 566 567 568 569 570 571 G. Eckert and A. Schlegel and M. A. Schlegel. P. J. J. Lett. C. Acta 80. 114. NATO-ASI Series C 406 (Kluwer Academic. B. 15. F. Seeger and J. Ed. 5558 (2000). L. Chem. Onsager. 574 575 R. Phys. J. G. The Netherlands. 48. M. Formosinho. Foresman and H. B. R. J. 572 H. A. 58. Pascual-Ahuir and E. Frisch. G. J. 61. Hubak. Mennucci. Hollas. Chem. 1127 (1994). Pople. J. J. in Theoretical and Computational Models for Organic Chemistry. Chem. Fausto and J. N. Carsky and E. J. L. J. Silla. Chem. 1486 (1936). 351 (1934). T. Phys. Chem. Bacskay. and L. D. 65. Csizmadia. Chem. R. . 265 (1976). Chem. in Recent experimental and computational advances in molecular spectroscopy. Halgren and W. Phys. I. 20. 407 (1991). Arnaut. 49. 385 (1981). 1991) 5-33. Ed. Chipman. Am. 112. 369 (2002). 2. Lett. 4744 (2001). Theo. Raffenetti. Kirkwood. 1993) 11-26. 573 J. B. E. The Netherlands. Cancès and B. Chem.