AMSOLRATE 8.6/P8.5.1-A6.6 Page i AMSOLRATE Manual Program Version: 8.6/P8.5.1-A6.6 Program Version Date: January 4, 2000 Date of this manual update: January 4, 2000 Copyright 2000 Yao-Yuan Chuang, Yi-Ping Liu, and Donald G. Truhlar Department of Chemistry and Supercomputer Institute University of Minnesota, Minneapolis, MN 55455-0431 AMSOLRATE provides an interface between two other programs: POLYRATE-version 8.5.1 by J. C. Corchado, Y.-Y. Chuang, P. L. Fast, J. Villa, W.-P. Hu, Y.-P. Liu, G. C. Lynch, K. A. Nguyen, C. F. Jackels, V. S. Melissas, B. J. Lynch, I. Rossi, E. L. Coitino, Antonio Fernandez-Ramos , R. Steckler, B. C. Garrett, A. D. Isaacson, and D. G. Truhlar Department of Chemistry and Supercomputer Institute University of Minnesota, Minneapolis, MN 55455-0431 and AMSOL-version 6.6 by Gregory D. Hawkins, David J. Giesen, Gillian C. Lynch, Candee C. Chambers, Ivan Rossi, Joey W. Storer, Jiabo Li, Tianhai Zhu, Paul Winget, Daniel Rinaldi, Daniel A. Liotard, Christopher J. Cramer, and Donald G. Truhlar. Department of Chemistry and Supercomputer Institute University of Minnesota, Minneapolis, MN 55455-0431 Abstract: AMSOLRATE is a set of FORTRAN subprograms and Unix scripts for interfacing the POLYRATE dynamics program and the AMSOL electronic structure program for the purpose of direct dynamics calculations of chemical reaction rates of polyatomic species. Reaction rates may be calculated for reactions in the gas phase or in liquid-phase solutions. The interface is based on the POLYRATE hooks protocol. The dynamical methods used are variational or conventional transition state theory and multidimensional semiclassical approximations for tunneling and nonclassical reflection. Gas-phase rate constants may be calculated by any of the dynamical methods available in the POLYRATE-version 8.5.1 program for canonical or microcanonical ensembles or for specific vibrational states of selected modes with translational, rotational, and other vibrational modes treated thermally. Reaction rates in liquid-solutions may be calculated by either the separable equilibrium solvation (SES), equilibrium solvation path (ESP), or nonequilibrium solvation (NES) approximation. One may, if desired, optimize the orientation of the generalized transition-state dividing surface, as well as its location along the reaction path. Bimolecular and unimolecular reactions are included. Both single-level and dual-level dynamical methods are available. In the single-level mode, potential energies and gradients can be calculated by any of the semiempirical molecular orbital methods in the AMSOL version 6.6 package, with or without reaction fields and atomic surface tensions to represent the solvent. In the dual-level mode, these same options are available for the lower-level part of the calculation, and higher-level data are read from an external file. AMSOLRATE 8.6/P8.5.1-A6.6 Page ii CONTENTS ---- TITLE PAGE .......................................................... i CONTENTS ............................................................ ii REFERENCES FOR AMSOLRATE PROGRAM .................................... iv USER AGREEMENT ...................................................... iv DISTRIBUTION AND CONTACT INFORMATION ................................. v EXPLANATION OF VERSION NUMBERS ...................................... v VERSION HISTORY ..................................................... vi AMSOLRATE VERSIONS AND AUTHORS ...................................... vii 1. INTRODUCTION .................................................... 1-1 2. ORGANIZATION OF THE MANUAL ...................................... 2-1 3. INSTALLATION .................................................... 3-1 3.A. Description of files ....................................... 3-2 3.A.1. List of files ..................................... 3-2 3.B. Compiling, linking, and running the test suite ............ 3-3 4. POTENTIAL ENERGY SURFACE FROM MOLECULAR ORBITAL THEORY............ 4-1 4.A. Interface ................................................. 4-1 4.B. Reactions in solution ...................................... 4-2 5. DESCRIPTION OF PROGRAM STRUCTURE AND SUBPROGRAMS ................ 5-1 6. INPUT/OUTPUT FILE USAGE ......................................... 6-1 7. SELECTED DETAILS OF THE CALCULATION ............................. 7-2 7.A. Restart option ............................................ 7-2 8. DESCRIPTION OF INPUT DATA ....................................... 8-1 8.A. Description for esp.fu71 through esp.fu78 files ........... 8-1 8.B. Description of other input and unavailable units ........... 8.4 9. DESCRIPTION OF SAMPLE TEST RUNS ................................. 9-1 9.A. GAS-PHASE REACTIONS 9.A.1. Gas test run 1 .................................... 9-3 9.A.2. Gas test run 2 .................................... 9-4 9.A.3. Gas test run 3 .................................... 9-5 9.A.4. Gas test run 4 .................................... 9-6 9.A.5. Gas test run 5 .................................... 9-7 9.A.6. Gas test run 6 .................................... 9-8 9.A.7. Gas test run 7 .................................... 9-9 9.A.8. Gas test run 8 .................................... 9-10 9.A.9. Gas test run 9 .................................... 9-11 9.A.10.Gas test run 10 .................................... 9-12 9.A.11.Gas test run 11 .................................... 9-13 9.A.12.Gas test run 12 .................................... 9-14 9.A.13.Gas test run 13 .................................... 9-15 9.A.14.Gas test run 14 .................................... 9-16 9.A.15.Gas test run 15 .................................... 9-17 AMSOLRATE 8.6/P8.5.1-A6.6 Page iii 9.B. LIQUID-PHASE REACTIONS 9.B.1. Solution test run 1 ............................... 9-18 9.B.2. Solution test run 2 ............................... 9-19 9.B.3. Solution test run 3 ............................... 9-20 9.B.4. Solution test run 4 ............................... 9-21 9.B.5. Solution test run 5 ............................... 9-22 9.B.6. Solution test run 6 ............................... 9-23 9.B.7. Solution test run 7 ............................... 9-24 9.B.8. Solution test run 8 ............................... 9-25 9.B.9. Solution test run 9 ............................... 9-26 9.B.10.Solution test run 10 ............................... 9-27 10. COMPUTER ON WHICH AMSOLRATE HAS BEEN TESTED ..................... 10-1 11. TEST RUN TIMINGS ................................................ 11-1 12. BIBLIOGRAPHY ..................................................... 12-1 13. NUMSTEP AND SCFCRT ............................................... 13-1 14. ACKNOWLEDGMENTS .................................................. 14-1 AMSOLRATE 8.6/P8.5.1-A6.6 Page iv REFERENCES FOR AMSOLRATE PROGRAM: Persons using AMSOLRATE should cite the AMSOLRATE package with the following reference: AMSOLRATE-version 8.6/P8.5.1-A6.6 by Y.-Y. Chuang, Y.-P. Liu, and D.G. Truhlar, University of Minnesota, Minneapolis, 1999 based on POLYRATE- version 8.5.1 by J. C. Corchado, Y.-Y. Chuang, P. L. Fast, J. Villa, W.-P. Hu, Y.-P. Liu, G. C. Lynch, K. A. Nguyen, F. Jackels, V. S. Melissas, B. J. Lynch, I. Rossi, E. L. Coitino, A. Fernandez-Ramos, R. Steckler, B. C. Garrett, A. D. Isaacson, and D. G. Truhlar, University of Minnesota, Minneapolis, 2000 and AMSOL-version 6.6 by G. D. Hawkins, D. J. Giesen, G. C. Lynch, C. C. Chambers, I. Rossi, J. W. Storer, J. Li, T. Zhu, D. Rinaldi, D. A. Liotard, C. J. Cramer, and D. G. Truhlar, University of Minnesota, Minneapolis, 2000. USER AGREEMENT AMSOLRATE is a licensed program, and use of this program implies acceptance of the terms of the license, which are repeated here for convenience: 1. No user or site will redistribute the source code or executable code to a third party in original or modified form without written permission of the principal investigator (Donald G. Truhlar). A license does not entitle the licensee to relicense the code or distribute it in original or modified form to parties not covered by the license. The licensee has no ownership rights in the AMSOLRATE software or in any copyrights for the AMSOLRATE software or documentation through this license. A user license covers the work of a single research group and the code may be shared and disseminated within a group without requiring permission. Site-license inquires should be directed to the principal investigator (D.G.T.). 2. Publications resulting from using this package or the POLYRATE or AMSOL subsystems used by this package will cite the corresponding program. The recommended references are given in the documentation (see above for the recommended reference for AMSOLRATE). 3. No guarantee is made that this program is bug-free or suitable for specific applications, and no liability is accepted for any limitations in the mathematical methods and algorithms used within the program. 4. No consulting or maintenance services are guaranteed or implied. 5. The POLYRATE-8.5.1 and AMSOL-6.6 codes required for use of AMSOLRATE are covered by separate licenses. AMSOLRATE 8.6/P8.5.1-A6.6 Page v DISTRIBUTION AND CONTACT INFORMATION Licenses and distribution procedures for the AMSOLRATE and POLYRATE packages are available on the world wide web at http://comp.chem.umn.edu. The AMSOL package is available without support to non-profit users at this same URL. AMSOL licenses with user support are available to both commercial and academic users from Oxford Molecular Group, Ltd. and by Makolab and will soon be available from Semichem; contact information for all authorized AMSOL distributors is available at http://comp.chem.umn.edu. Prospective users of AMSOLRATE need to obtain licensed copies of AMSOL and POLYRATE as well as a copy of AMSOLRATE. For more information or questions or problems with AMSOLRATE contact the principal investigator: Donald G. Truhlar Department of Chemistry and Supercomputer Institute University of Minnesota 207 Pleasant Street S.E. Minneapolis, MN 55455-0431, U.S.A. http://comp.chem.umn.edu fax: (612) 626-9390 AMSOLRATE VERSIONS Explanation of version numbers: In the AMSOLRATE version number, the first number (i.e., the part before the virgule) refers to the version of the AMSOLRATE package and is unique. That is, this number changes if the interface changes or if the POLYRATE and/or the AMSOL part changes. The second and third numbers (which are optional parts of the version designation) give extra information by referring to the versions of POLYRATE and AMSOL, respectively, upon which the given version of AMSOLRATE is built. In particular, in the version number 8.6/P8.5.1-A6.6, the first 8.6 is the version number of the whole package, including the interface, P8.5.1 specifies that the version of POLYRATE is 8.5.1, and A6.6 specifies that the version of AMSOL is 6.6. Thus it is equally acceptable to say AMSOLRATE-version 8.6 or to say AMSOLRATE-version 8.6/P8.5.1-A6.6. Note: we sometimes update one or more of the manuals without updating the version number. The manual version is determined by the date of its most recent change and is given on its first page. Any changes other than the manual(s) always involve a change in version number of the code. AMSOLRATE 8.6/P8.5.1-A6.6 Page vi VERSION HISTORY Version 7.9.1/P7.9.1-A6.5.1 This was the first distributed version of AMSOLRATE. Version 8.0/P8.0-A6.5.2 (Nov. 1998) 1. This version is based on new versions of POLYRATE and AMSOL. 2. A new gas-phase test run testr15 is add as an example of VTST-ISPE and VRP calculation. Version 8.0.1/P8.0-A6.5.3 (Jan. 1999) 1. This version is based on a new version of AMSOL. Version 8.1.1/P8.1.1-A6.5.3 (July 1999) 1. This version is based on a new version of POLYRATE. Version 8.1.2/P8.1.2-A6.5.3 (August 1999) 1. This version is based on a new version of POLYRATE. Version 8.2/P8.2-A6.5.3 (August 1999) 1. This version is based on a new version of POLYRATE. 2. A new solution-phase test run testr10 is add as an example of nonequilibrium solvation calculation. 3. The compilation script for SGI workstations amratecl.sgi is set to 32-bit. Version 8.4/P8.4-A6.5.3 (December 1999) 1. This version is based on a new version of POLYRATE. Version 8.5/P8.4-A6.6 (January 2000) 1. This version is based on a new version of POLYRATE. Version 8.6/P8.5.1-A6.6 (November 2000) 1. This version is based on a new version of POLYRATE. AMSOLRATE 8.6/P8.5.1-A6.6 Page vii AMSOLRATE VERSIONS AND AUTHORS In the list below, if we list several versions in succession and then a set of authors, it means the authors are the same for all those versions. AMSOLRATE-version 7.9.1-8.6 Y.-Y. Chuang, Y.-P. Liu, and D. G. Truhlar AMSOLRATE 8.6/P8.5.1-A6.6 Page 1-1 1. INTRODUCTION AMSOLRATE is a program for the analysis of reactants, products, and transition states of chemical reactions and for direct dynamics calculations of variational transition state theory (VTST) rate constants and multi- dimensional semiclassical tunneling probabilities using molecular orbital theory to represent the potential energy of interaction and its gradient wherever they are needed. This program version, 8.6/P8.5.1-A6.6, combines POLYRATE-version 8.5.1, which is a program for dynamical rate calculations, with the semiempirical electronic structure program AMSOL-version 6.6. In the current version of the program, both unimolecular and bimolecular gas-phase reactions can be treated, and the maximum number of atoms can be reset by changing the parameters NATOMS, MAXHEV, and MAXLIT in the appropriate include files. The analysis of reactants and products can be performed for up to two reactants and two products. The reaction path (RP) can be calculated by the Euler single-step (ES), Euler stabilization (ES1 or ES1*), or Page-McIver (PM), or variational reaction path (VRP) methods. The first three of these choices are numerical methods for finding the steepest- descent path, which is the minimum-energy path (MEP). The RP is always defined in mass-scaled coordinates, and the MEP in mass-scaled coordinates is identical to what some researchers call the IRC. Various anharmonicity options are available for the vibrational modes of the reactants, products, and/or generalized transition state species. The VTST rate constants can be computed using canonical variational theory (CVT), improved canonical variational theory (ICVT), or microcanonical variational theory (muVT). Conventional transition state theory (TST) may also be employed. The tunneling probabilities can be computed using the zero-curvature tunneling (ZCT) method, the centrifugal-dominant small-curvature tunneling (SCT) method, the large-curvature tunneling (LCT) method (which is the same as LCG3 and which includes tunneling into excited states), and the microcanonical optimized multidimensional tunneling (muOMT) method. In single-level mode, the AMSOLRATE program computes the rate constants either from energies and gradients calculated by semiempirical molecular orbital (MO) theory, or from a restart file which uses the reaction-path information from a previous calculation. In either case, the semiempirical MO methods that can be used to represent the surface are the same as those available in the AMSOL package, version 6.6, namely, MNDO, AM1, and PM3 for gas-phase calculations and SMx models for liquid-phase calculations. The package also supports gas-phase configuration interaction calculations based on semiempirical Hamiltonians. In dual-level mode, AMSOLRATE computes the potential energy at the lower level by one of the options available for single-level calculations, and it reads the data needed for the high-level from a file assigned to FORTRAN unit fu50. For reactions in solution, the SES, ESP, and NES approximations are supported in the current version of AMSOLRATE. More details of these methods are given in Section 4.B of this manual. AMSOLRATE 8.6/P8.5.1-A6.6 Page 2-1 2. ORGANIZATION OF THE MANUAL This AMSOLRATE manual assumes that the user also possesses manuals for POLYRATE-version 8.5.1, AMSOL-version 6.6 and that the user is familiar with the theory behind both POLYRATE and AMSOL. Throughout this manual, we capitalize AMSOLRATE, POLYRATE, AMSOL, subprogram names, FORTRAN variables (except FORTRAN unit designators like fu6), AMSOL keywords, POLYRATE keywords, and acronyms in which every letter stands for a new word. File names (except some include files) and Unix commands are in lower case. In this manual, POLYRATE keywords (for example, EZERO) are in capital letters, whereas the values that these keywords take (for example, "calculate") are in double quotation marks. Section 3 discusses how to install AMSOLRATE. Some aspects of the interface of POLYRATE with the AMSOL semiempirical molecular orbital package are described in Section 4. Section 5 describes the program structure and the subprograms that make up the AMSOLRATE program, and it lists the files that comprise this version of AMSOLRATE. Section 6 describes the FORTRAN unit numbers used for the input and output. Section 7 contains selected details of some of the options available in this version of AMSOLRATE. Section 8 contains a description of the input required by the AMSOLRATE program. Section 9 describes the test runs that make up the test suite for this version of the program. Section 11 contains the timings for the test suite for this version of the code on all the machines on which it has been tested. Section 12 contains a bibliography. Section 13 compares results obtained with various programs and numerical parameters. Section 14 has acknowledgments. AMSOLRATE 8.6/P8.5.1-A6.6 Page 3-1 3. INSTALLATION Notice that the licenses for POLYRATE and AMSOL may be obtained via the internet, in particular from the URL: http://comp.chem.umn.edu A user should install POLYRATE 8.5.1 and AMSOL 6.6 before installing the AMSOLRATE program. The AMSOLRATE 8.5 package contains several compilation scripts (/amsolrate8.6/src/amratecl.*) for various platforms. The AMSOLRATE scripts will look for the file .poly_path8.5.1 for the location of the POLYRATE8.5.1 package and will ask the user to input the path of the AMSOL6.6 source codes (e.g. /usr/people/chuang/amsol6.6). One can carry out the installation by following the instructions later in this section. The program is distributed with the short output files (in particular the fu15 file for all test runs and the fu14 file for the test runs in which FORTRAN unit fu14 is used) for all the test runs in the test suite. The program is always distributed in the Unix tarred format, and upon issuing the command tar -xvf , the files that make up the distribution package (see Subsection 3.A) will be partitioned into the following structure: amsolrate8.5 | --------------------------------------------------------- | | | | | | | | doc src gas solution testo data check_test.jc | check_all.jc --------- | | gas solution amsolrate8.5 this directory contains all the files doc this directory contains the present AMSOLRATE manual src this directory contains all the source code and all the job control files needed to compile and link the source codes gas this directory contains the sample test run files for gas-phase reactions solution this directory contains the sample test run files for solution-phase reactions testo/gas this directory contains all the short output files for the /gas test runs from Cray C90 testo/solution this directory contains all the short output files for the /solution test runs from Cray C90 AMSOLRATE 8.6/P8.5.1-A6.6 Page 3-2 data this directory contains files with parameters for solvation models check_test.jc a job control file that checks a specific test run check_all.jc a job control file that checks all the test runs 3.A. DESCRIPTION OF FILES Version 8.6/P8.5.1-A6.6 of the AMSOLRATE code consists of several files which may be classified according to the following categories: 1) FORTRAN source files 2) C source files 3) Include files 4) Manual 5) Job control files 6) Data files 7) Input for test runs 8) Output for test runs 3.A.1 List of files FORTRAN source files: amdrv.f amrate.f amsol.f fromblas.f hooks.f amsol_mach.sgi amsol_mach.ibm amsol_mach.cray C source files: dateclock.c dateclock1.c second1.c Include files: param1.inc param2.inc param3.inc param4.inc Manual: amsolrate.doc Job control files (C shell scripts for compiling, linking, and running test runs): amratecl.cray amratecl.ibm amratecl.sgi concat run_all.jc check_test.jc check_all.jc in /gas run_all.jc testr1.jc testr2.jc testr3.jc testr4.jc testr5.jc testr6.jc testr7.jc testr8.jc testr9.jc testr10.jc testr11.jc testr12.jc testr13.jc testr14.jc testr15.jc AMSOLRATE 8.6/P8.5.1-A6.6 Page 3-3 in /solution run_all.jc testr1.jc testr2.jc testr3.jc testr4.jc testr5.jc testr6.jc testr7.jc testr8.jc testr9.jc testr10.jc Data files: solv.prp solv.txt solvR.prp Input for testruns in /gas testr1.70 testr12.75 testr2.71 testr6.73 testr1.71 testr13.70 testr2.72 testr6.74 testr1.73 testr13.71 testr2.73 testr6.75 testr1.75 testr13.72 testr2.74 testr7.70 testr10.70 testr13.73 testr2.75 testr7.71 testr10.71 testr13.74 testr3.70 testr7.72 testr10.72 testr13.75 testr3.71 testr7.73 testr10.73 testr14.70 testr3.72 testr7.74 testr10.74 testr14.71 testr3.73 testr7.75 testr10.75 testr14.72 testr3.74 testr8.70 testr11.70 testr14.73 testr3.75 testr8.71 testr11.71 testr14.75 testr4.70 testr8.72 testr11.72 testr15.70 testr4.71 testr8.73 testr11.73 testr15.71 testr4.72 testr8.74 testr11.74 testr15.72 testr4.73 testr8.75 testr11.75 testr15.73 testr4.74 testr9.70 testr12.70 testr15.74 testr4.75 testr9.71 testr12.71 testr15.75 testr5.70 testr9.72 testr12.72 testr15.77 testr6.70 testr9.73 testr12.73 testr15.78 testr6.71 testr9.74 testr12.74 testr2.70 testr6.72 testr9.75 testr1.dat testr13.dat testr3.dat testr7.dat testr10.dat testr14.dat testr4.dat testr8.dat testr11.dat testr15.dat testr5.dat testr9.dat testr12.dat testr2.dat testr6.dat testr9.50 testr10.50 testr12.50 testr15.51 testr11.50 testr13.50 testr3.50 in /solution testr1.70 testr2.74 testr5.70 testr6.74 testr1.71 testr2.75 testr5.71 testr6.75 testr1.72 testr3.70 testr5.72 testr7.70 testr1.73 testr3.71 testr5.73 testr7.71 testr1.74 testr3.73 testr5.74 testr7.72 testr1.75 testr3.75 testr5.75 testr7.73 testr2.70 testr4.70 testr6.70 testr7.74 testr2.71 testr4.71 testr6.71 testr7.75 testr2.72 testr4.73 testr6.72 testr8.71 testr2.73 testr4.75 testr6.73 testr8.72 testr8.73 testr8.74 testr8.75 testr9.70 testr9.71 testr9.72 testr9.73 testr9.74 AMSOLRATE 8.6/P8.5.1-A6.6 Page 3-4 testr9.75 testr10.70 testr10.71 testr10.72 testr10.73 testr10.74 testr10.75 testr1.dat testr2.dat testr3.dat testr4.dat testr5.dat testr6.dat testr7.dat testr8.dat testr9.dat testr10.dat Output files: in /testo/gas testr1.fu15 testr2.fu15 testr3.fu15 testr4.fu15 testr5.fu15 testr6.fu15 testr7.fu15 testr8.fu15 testr9.fu15 testr10.fu15 testr11.fu15 testr12.fu15 testr13.fu15 testr14.fu15 testr15.fu15 in /testo/solution testr1.fu15 testr2.fu15 testr3.fu15 testr4.fu15 testr5.fu15 testr6.fu15 testr7.fu15 testr8.fu15 testr9.fu15 testr10.fu15 3.B. SPECIFYING THE SIZE OF THE AMSOLRATE EXECUTABLE FILE A user can easily adjust the size of the executable file to be created by modifying the include files needed by the source code. A full description of how to carry out these modifications is given in subsection 3.C. Dimensions in the AMSOLRATE program, like those in POLYRATE, are governed by an include file called param.inc; and the dimensions for the AMSOL part are governed by the include file called SIZES.i, which is distributed in the AMSOL package. Note that the file SIZES.i and all references to it are case- sensitive when this program is used in a Unix or other case-sensitive environment. As explained in Subsection 5.A.2 of the POLYRATE manual, any changes made to include files must be done consistently, and the code must be recompiled for the changes to become effective. The include file param.inc may be modified as described in Section 5 of the POLYRATE manual. However, one must be wary of the value used for the maximum number of atoms, NATOMS, in this include file as it must be less than or equal to the sum of MAXHEV and MAXLIT in the include file SIZES.i. This sum sets the maximum number of atoms allowed in the AMSOL portion of the code. In the include file SIZES.i, the values of MAXHEV and MAXLIT may be modified as described in the AMSOL manual. Note that, MAXHEV + MAXLIT must be greater than or equal to the POLYRATE variable NATOMS. In the subprogram SETUP, a check is carried out to ensure that NATOMS is less than or equal to MAXHEV + MAXLIT before starting the full calculation. AMSOLRATE 8.6/P8.5.1-A6.6 Page 3-6 3.C. COMPILING, LINKING, AND RUNNING THE TEST SUITE AMSOLRATE-version 8.5 is distributed with four sample param.inc files. The sample param.inc files are param1.inc, param2.inc, param3.inc, and param4.inc. Compiling and linking the source code ------------------------------------- This subsection describes the use of the scripts provided for compiling and linking the AMSOLRATE-version 8.5 source code. Four scripts are included in the amsolrate8.6/src directory: amratecl.ibm ---- for IBM workstations amratecl.cray ---- for Cray supercomputers amratecl.sgi ---- for SGI workstations Change the working directory to amsolrate8.6/src, and choose the appropriate script from the list above for the type of machine you are working on, and then do the following: Execute the script: For example, if your machine is an IBM RS/6000 and you want to use param1.inc as the AMSOLRATE parameter include file and the AMSOL6.6 source codes are located in the /usr/people/chuang/amsol6.6 directory, then you type: amratecl.ibm 1 /usr/people/chuang/amsol6.6 This script will copy, compile, and link all the necessary files to generate an executable called amslrate.exe in /src. Use a different script name if a different type of machine is being used. Note: Each script requires two arguments: (1) the number to indicate which param#.inc is used. (2) the full path of the directory that contains the AMSOL6.6 source codes. AMSOLRATE 8.6/P8.5.1-A6.6 Page 4-1 4. POTENTIAL ENERGY SURFACE FROM MOLECULAR ORBITAL THEORY 4.A. INTERFACE The interface between POLYRATE and AMSOL consists of four subroutines: AMSET, AMPEG, AMOPT, AMHES. These subroutines are in the file amdrv.f. The subroutine AMSET sets up some common blocks needed by AMSOL, checks that certain AMSOL and POLYRATE arrays are dimensioned in a compatible way, and parses the fu70 input file. The primary subroutine for the interface between POLYRATE and AMSOL is AMOPT. This subroutine, which is a modification of the main program of AMSOL, sets up all the necessary parameters for the semiempirical molecular orbital calculation and reads the initial geometry. AMOPT also calls the AMSOL optimization subroutines, such as FLEPO and EF, to optimize the geometry of a stationary point, if a geometry optimization is requested. In AMSOLRATE, the geometry optimizations for the reactants, products, and saddle point species are carried out by AMSOL routines only (not by POLYRATE routines), and AMOPT is called once for each molecular chemical species,i.e., once for each reactant and product and once for the saddle point. The subroutine AMPEG evaluates the potential and its gradient as functions of the Cartesian coordinates of the atoms. It serves as a link between the POLYRATE and AMSOL programs, and it functions in the same way as in the case of an analytic potential energy surface used in POLYRATE. The subroutine AMPEG takes a set of Cartesians from POLYRATE, converts them to angstroms, and passes this information to the AMSOL subroutine COMPFG which calculates the energy and gradient by the self-consistent-field (SCF) method. The subroutine AMPEG then makes the necessary unit conversions and returns the energy and gradient to POLYRATE in atomic units. If HHOOK is specified for the HESSCAL keyword in section *SECOND of the POLYRATE fu5 input file, the subroutine AMHES returns to POLYRATE the Hessian matrix of the potential at a given point calculated using the AMSOL Hessian evaluation subroutine FMAT. However, if the GHOOK option is used for the HESSCAL keyword in the *SECOND section, the subroutine AMPEG is used for the gradients of the potential at a given point with AMSOL gradient routine COMPFG. Then the Hessian is evaluated by numerical differences in POLYRATE. AMSOLRATE 8.6/P8.5.1-A6.6 Page 4-2 4.B. REACTIONS IN SOLUTION This version of AMSOLRATE supports three approximations for calculating reaction rates in solution. First is the Separable Equilibrium Solvation (SES) approximation, second is the Equilibrium Solvation path (ESP) approximation, and the third is the Nonequilibrium Solvation (NES) approximation. See the references in Section 12 of this manual for theoretical background. In the SES approximation, a gas-phase reaction path is used, and the solvation free energy is added along this path with a solvation model where only the energy along the path is changed from the gas phase, not the path itself or the Hessian. In the ESP approximation, the free energy of solvation is included during the geometry optimization, gradient calculation, and Hessian evaluations. As a consequence, the reaction-path-following algorithm yields a liquid-phase reaction path. This solvated reaction path is called the equilibrium solvation path. In the NES approximation, a collective solvent mode is added to estimate the nonequilibrium solvation (friction) effects. To carry out this type of calculation, the ESP keyword is required in the fu70 input file. An important difference from the gas-phase-only MORATE package is that AMSOLRATE requires a fu70 file. In this file, there are three major options: GAS, ESP, and SES. To calculate reaction rates in the gas phase, it is required to input the keyword GAS in the fu70 file and to input AMSOL gas-phase keywords in the fu71-fu78 files. Similarly, for the ESP approximation, one needs to place the keyword ESP in the fu70 file and have the AMSOL keywords for the solvation model in the fu71-fu78 files. For the SES approximation, it is required to include an SES list keyword in the fu70 file. The SES keyword has the following options. SMODEL The SMODEL keyword specifies the parameters of the solvation model. The keywords for the currently supported models are: SM5.4A, SM5.4P, SM5.4U, SM5.2R, and SM5.42R. The following methods are available: Method SMODEL Keyword in fu71-fu78 -------- -------- ---------------------- SM5.4/A SM5.A AM1 SM5.4/P SM5.4P PM3 AM1-SM5.4/U SM5.4U AM1 PM3-SM5.4/U SM5.4U PM3 SM5.2R/AM1 SM5.2R AM1 SM5.2R/PM3 SM5.2R PM3 SM5.2R/MNDO SM5.2R MNDO SM5.42R/AM1 SM5.42R AM1 SM5.42R/PM3 SM5.42R PM3 SM5.42/AM1 SM5.42R AM1 SM5.42/PM3 SM5.42R PM3 AMSOLRATE 8.6/P8.5.1-A6.6 Page 4-3 Note that SM5.4/A may also be called SM5.4/AM1, and SM5.4/P may also be called SM5.4/PM3. The models are parameterized only for the temperature of 298 K. SOLVNT The SOLVNT keyword specifies the type of solvent. The options are: WATER, GENORG, CHCL3, BENZENE, and TOLUENE. (Notice that CHCL3, BENZENE, and TOLUENE are valid choices only for the SM5.4/A and SM5.4/P solvation methods and notice that water is the only valid choice for SM5.4/U methods.) The default is WATER. IOFR The IOFR keyword specifies the index of refraction of the desired solvent. The default is 0. ALPHA The ALPHA keyword specifies Abraham's descriptor for the ability of solvent to donate a proton to a hydrogen bond. The default is 0. BETA The BETA keyword specifies Abraham's descriptor for the ability of solvent to accept a hydrogen bond. The default is 0. GAMMA The GAMMA keyword specifies the macroscopic surface tension of the solvent in SI units of mN/m. The default is 0. DIELEC The DIELEC keyword specifies the solvent dielectric constant, that is, the relative permittivity. The default is 0. FACARB The FACARB keyword specifies the fraction of non-hydrogenic atoms in the solvent that are aromatic carbons. The default is 0. FEHALO The FEHALO keyword specifies the fraction of non-hydrogenic atoms in the solvent that are electronegative halogen atoms, i.e., F, Cl, or Br. The default is 0. For detailed information about these keywords, please refer to the AMSOL manual. AMSOLRATE 8.6/P8.5.1-A6.6 Page 4-4 To apply the SES approximation, one just needs to include the appropriate gas- phase keywords for the AMSOL input files (i.e. fu71-fu78) and to add the SES section into fu70. For example, to calculate a reaction rate in methanol with the SES approximation, one should input the following lines to fu70: SES SMODEL SM5.4P SOLVNT GENORG IOFR 1.3288 ALPHA 0.43 BETA 0.47 GAMMA 22.1 DIELEC 32.63 FACARB 0.00 # optional, this is the default value. FEHALO 0.00 # optional, this is the default value. END In the current version of AMSOLRATE, if one wants to make a calculation in the SES approximation for a reaction that involves a single atom (anion or cation) as one of the reactants or products, the solvation model specified on fu70 is not applied to that particular species. It is required to have the solvation model keywords in the corresponding fu71-fu78 file instead of gas phase keyword for atomic species (see, e.g., the following test run: solution/testr4). Notation: note that ESP calculations are generally denoted SMx/y whereas SES calculations are denoted SMx/y//y, where y is AM1 of PM3. In principle one could have SMx/y//z, but such calculations are not supported by this version of AMSOL. AMSOLRATE 8.6/P8.5.1-A6.6 Page 5-1 5. DESCRIPTION OF PROGRAM STRUCTURE AND SUBPROGRAMS This section describes the program structure and the subprograms for the AMSOLRATE code. As mentioned in the introduction, the AMSOLRATE program has two major options for obtaining the reaction-path information necessary for rate calculations. The first option is to calculate this information directly using semiempirical molecular orbital theory to represent the potential energy of interaction. The second option is to read it from a restart file which was saved from a previous AMSOLRATE run. These options are implemented in the AMSOLRATE program in precisely the same way as in the POLYRATE program. The next five paragraphs describe the use of these options either for a single-level calculation or for the lower level of a dual-level calculation. First, the program reads the keyword input from different input files. Then the input options are analyzed for possible conflicts (i.e., unsupported combinations of options) in subroutine OPTION. If it is a restart run, the program reads the restart information in RESTOR, and the rate calculation proceeds as described in the POLYRATE manual. If the calculation is not a restart run, the geometry and the MEP-following information in the unit fu5 input file is used for the next stages of calculations. Depending on the options chosen, the properties of the reactants, products and generalized transition state are calculated next. To calculate the stationary points, AMSOLRATE requires to optimize the geometries in the AMSOL portion, then pass the final geometry back to the POLYRATE portion for further calculations. This means that the keyword INITGEO can only has the value of HOOKS. The initial guess geometries will be required in the files fu71-fu78 instead of fu5. For the reactants and/or the products, the equilibrium geometries and properties are optionally obtained in the subprogram DOREPR. It should be noted that in this version of AMSOLRATE the force constant matrix can either be obtained directly from AMSOL, or it can be calculated in POLYRATE routines from the numerical derivatives of the gradients obtained in AMSOL. Next, the program optionally determines the saddle point geometry and properties. This is done in SADENG. The user must supply an approximate geometry of the saddle point in an AMSOL input data file, esp.fu75. After determining the geometries and the properties for the reactants, products, and/or the saddle point, the program proceeds in the same way as described in the POLYRATE manual. The precise sequence of calculations depends on the combination of options chosen. Any further calculations of the energy and gradients needed by the POLYRATE subprograms are accomplished through calls to the interface subroutines EHOOK and GHOOK. These routines call the AMSOL-specific interface routine AMPEG. This routine converts the geometry from atomic units to angstroms, and it calls the AMSOL subroutine COMPFG to determine the energy and the gradients. In the present version the subroutine COMPFG is called such that the energy calculations are done with a full SCF, and the gradients are only calculated as required. The subroutine AMPEG then converts the energy and the gradients to atomic units and passes them back to EHOOK and GHOOK. AMSOLRATE 8.6/P8.5.1-A6.6 Page 5-2 In dual-level mode, AMSOLRATE then reads the required data for the higher level from an external file assigned to FORTRAN unit fu50. 5.B. DESCRIPTION OF AMSOLRATE SUBPROGRAMS This section discusses the subprograms used in AMSOLRATE. The subprograms consists of the new routines created for the interface between POLYRATE and AMSOL. Descriptions of subroutines in the POLYRATE and AMSOL portions are already given in their individual manuals; thus only routines that were changed or included specifically for AMSOLRATE are discussed here. The amdrv.f file is made up of four subroutines, as discussed in Section 4.A. The subroutine AMPEG calls the AMSOL subroutines to calculate the energy and the gradients. The subroutine AMOPT is used to read the AMSOL input file and to optimize the geometry. The subroutine AMHES calculates the force constant matrix from the gradients by the central difference method with an user- specified step size (the default is 0.0001 bohr). The subroutine AMSET checks the compatibility of system sizes allowed in POLYRATE and AMSOLRATE; it also performs some initialization and parses the fu70 input file. AMSOLRATE 8.6/P8.5.1-A6.6 Page 6-1 6. INPUT/OUTPUT FILE USAGE The program uses several files for input data, for storing restart information, and for output. The program explicitly opens and closes all files required for a given run, and the files are assumed to have the file names poly.fu# and esp.fu7#, where # denotes an integer. The file poly.fu# is assigned to FORTRAN I/O unit fu# which is defined in the include file param.inc (see the POLYRATE on-line manual). The files esp.fu7# are linked to the FORTRAN unit numbers fu71 through fu78. These unit numbers are also defined in the include file param.inc. Because of these I/O specifications, at the beginning of any run all the AMSOLRATE input data files that have logical file names differing from poly.fu# or esp.fu7# must be linked or copied to the corresponding poly.fu# or esp.fu7# file. Unit Usage fu1 Reaction-path information for restarting the program fu2 Additional reaction-path information to be merged with that from unit=fu1 fu7 Output for merged reaction path information fu5 General POLYRATE input data fu6 Output (This is the full output file. Selected subsets of this output are also written to files poly.fu14 and poly.fu15.) fu14 Output table of dynamical bottleneck properties fu15 Output table of selected forward rate constants fu22 Detailed output for the LCT calculations fu25 Output table of energetic information along the MEP fu26 Output table of GTS frequency information along the MEP fu27 GTS geometry information in the XYZ input format of the XMOL program fu50 Additional input data for VTST-IOC calculations (also called dual- level mode or /// dynamics) fu51 Additional input data for VTST-ISPE calculations fu61 Information about the stationary points from POLYRATE calculation for the second restart option AMSOLRATE 8.6/P8.5.1-A6.6 Page 6-2 fu70 AMSOLRATE input data for phase information fu71 AMSOL input data for reactant 1 fu72 AMSOL input data for reactant 2, if it exists fu73 AMSOL input data for product 1 fu74 AMSOL input data for product 2, if it exists fu75 AMSOL input data for the saddle point or starting geometry fy76 reserved for use by electronic structure package interfaces in general, but not used by AMSOLRATE fu77 AMSOL input data for the wellr, if it exists fu78 AMSOL input data for the wellp, if it exists Running the test runs --------------------- To run the test runs, type the name of the job control file at the prompt: For example: %testr1.jc This means type testr1.jc after the prompt, which is %. Note: This script creates a temporary subdirectory called testr1 in the current working directory (e.g., /amsolrate8.6/gas/testr1). The calculation is carried out in this subdirectory. If this subdirectory already exists when the script is run, the script will move the existing testr1 subdirectory to /testr1_old, write an error message in the long output file testr1.fu6, and continue. No error checking is carried out for the existence of the subdirectory testr1_old. To run the whole test suite, type run_all.jc. Note: to run the whole test suite, make sure that the amslrate.exe is compiled and linked with param1.inc. The dimensions in the other supplied paramx.inc files are not large enough to run the entire test run. To check the result of a test run, type check_test.jc #, where # is the number of the test run. For example: %check_test.jc gas 1 The differences will be listed in the file gas_1.chk. Then, to check the results for the whole test suite, type check_all.jc. AMSOLRATE 8.6/P8.5.1-A6.6 Page 7-1 7. SELECTED DETAILS OF THE CALCULATION The methods used by this program are described in detail in the POLYRATE and the AMSOL manuals and in the references within those manuals. However, the restart option requires special attention here. 7.A. RESTART OPTION The restart option is controlled by the RESTART keyword in the GENERAL section of the unit fu5 input file. This option is designed to allow calculations based on information stored on the save grid (the grid controlled by DELSV) in the restart file. One of the simplest calculations which can be carried out with the restart flag is to calculate rate constants at temperatures different from those specified in the original run. If the restart calculation does not involve LCG3 tunneling calculations, then the only input file, other than the restart file, required for the calculation is the FORTRAN unit fu5 file. If the restart calculation does LCG3 tunneling, then the esp.fu75 file is also required; the esp.fu71 through esp.fu74 files are never required for a restart calculation. This option can also be used to limit searches for variational transition states to localized ranges of s by using SLMG > SLM and/or SLPG < SLP, but SLM and SLP cannot be different from the values used in the run that created the restart file. Perhaps the most powerful use of the restart option, however, is that a restart file used for LCG3 ground-state tunneling calculations can be used for LCG3 excited-state calculations. This means that the contribution to the LCG3 tunneling probabilities from excited states can be checked without calculating the MEP if the restart file is saved during the LCG3 ground-state calculation; this leads to significant savings in computer time. A few additional points should be made with respect to the POLYRATE/AMSOLRATE restart option. First, with regard to tunneling, if the restart run is required to do SCT tunneling calculations, then the run that created the restart file must also have done SCT calculations, and similarly for LCT calculations. Second, unlike a full calculation in which the MEP information is determined, there is no error-checking for a restart calculation. Therefore, if the restart file being used does not contain all the information needed for the calculation specified, no error messages will be printed, but the calculation will abort. Also, a VTST-IOC calculation can only be restarted with a restart file that was generated in a VTST-IOC calculation, since the unit fu50 file can only be read in a non-restart run. AMSOLRATE 8.6/P8.5.1-A6.6 Page 8-1 8. DESCRIPTION OF INPUT DATA The unit fu5 input file for AMSOLRATE is very similar to the unit fu5 input file for POLYRATE, except that OHOOK must be specified in fule fu5, and the initial geometries for the optimization are taken from input files fu71-fu78. The fu71-fu78 files are in the standard AMSOL format. For information on the esp.fu70 file, see section 4.C. 8.A. DESCRIPTION FOR esp.fu71 THROUGH esp.fu78 FILES The data files esp.fu71 through esp.fu78 are AMSOL-type data files whose content and form are specified in the AMSOL manual (except esp.fu76 which is not used in this version). The first issue which must be considered when creating these files is that the order of the specified atoms in unit fu75 must be consistent with the order of the atoms input in the ATOMS keyword in the GENERAL section of the poly.fu5 file, and the order of the atoms in units fu71-fu78 must be consistent with the index specifications in the GEOM keyword in the REACT1, REACT2, PROD1, PROD2, START, WELLR, and WELLP sections of the unit poly.fu5 input file. Second, the method chosen for the semiempirical molecular orbital calculations for a given reaction must be the same in all the esp.fu71 through esp.fu78 files representing the species for that reaction. Also, the specification of the 1SCF keyword in files fu71-fu78 is optional. When this keyword is not included, an optimization will be carried out. When the geometry is already optimized, 1SCF may be specified to prevent an attempt at re-optimization. In general, though, this saves negligible time since for an optimized geometry, the gradient norm will be small and no optimization will be carried out even if 1SCF is not specified. However, the calculation proceeds by a different path through AMSOL routines when 1SCF is or is not specified, so the results will differ slightly due to the finite precision of the calculations. Generally speaking, it is not recommended to use the 1SCF keyword with AMSOLRATE, especially in the fu75 file when there is a saddle point. The esp.fu71 and esp.fu72 files specify information for the calculations on reactants, including geometry optimizations if desired. The esp.fu73 and esp.fu74 files specify analogous information for the products. The esp.fu75 file is used for the saddle point or starting geometry of the MEP as well as for all subsequent calls to AMSOL subprograms, in particular, for calculations at generalized transition states and at points in the nonadiabatic region of LCG3 calculations. AMSOLRATE 8.6/P8.5.1-A6.6 Page 8.4 The file esp.fu75 requires further explanation. When there is a saddle point, the initial geometry in the esp.fu75 file should be the initial guess for the saddle point geometry, and any keywords related to geometry optimization will apply to this step. Other keywords, like SCFCRT, also apply to the subsequent calls to AMSOL subprograms. If there is a saddle point but it is already optimized, then the optimization with AMSOL is still used to obtain the geometry from esp.fu75 file. When there is no saddle point, the initial geometry in the esp.fu75 file should be the starting geometry for the MEP calculation, and the 1SCF keyword should be specified. In this case, too, the relevant keywords also apply to the calculations at further points along the reaction path. The files esp.fu77 and esp.fu78 are used to specify information for the wells. 8.B. DESCRIPTION OF OTHER INPUT AND UNAVAILABLE UNITS Units fu29 and fu40 of POLYRATE are not supported in the AMSOLRATE package, even though they can be used in POLYRATE when it is not linked to AMSOL. This means that neither the IVTST-0 nor the IVTST-1 method is allowed in AMSOLRATE via fu29. (However, the IVTST-0 interpolation method can be employed for selected vibrational frequencies by using the LOWFREQ keyword in the PATH section of the POLYRATE fu5 input). Furthermore, the external electronic structure input file fu40 cannot be used for single-level or dual-level calculations in AMSOLRATE. AMSOLRATE 8.6/P8.5.1-A6.6 Page 9-1 9. DESCRIPTION OF SAMPLE TEST RUNS The input and short output data files for the twenty-two test runs are included in the AMSOLRATE distribution package. In a few cases, some of the input data files (such as those that are used as esp.fu71 through esp.fu78 files) are the same for two or more test runs. In such cases the file is included more than once, with different names, e.g., testr2.71 and testr7.71, for the convenience of new users and to ensure that each test run package is complete. The results obtained on different Cray computers or on workstations are not always identical to the full number of digits printed (due to round-off, SCF convergence, etc.), especially if there are any low-frequency modes. In all cases, the distributed versions of the fu14 and fu15 files for this version of AMSOLRATE are from runs on the Cray C90 computer. The reactions involved in the examples are: In /gas TR1: CH CH / \\ // \ H2C=CH CH -> CH CH=CH2 (symmetric 1,5-hydrogen shift | | reaction; CH3 CH3 TST rate constants only) TR2: CD3 + CH4 -> CD3H + CH3 (TST rate constants only) TR3: Cl- + CH3Cl -> CH3Cl + Cl- (symmetric SN2 reaction; AM1-IOC calculation) TR4: Cl- + CH3BR -> CH3Cl + Br- (exoergic SN2 reaction; forward rates calculation at 2 temperatures) TR5: Cl- + CH3BR -> CH3Cl + Br- (exoergic SN2 reaction; forward rates calculation at 2 temperatures, restart) TR6: CD3 + CH4 -> CD3H + CH3 (forward and reverse rates with LCG3 tunneling) TR7: Br + CH4 -> HBr + CH3 (forward and reverse rates with LCG3 tunneling) TR8: Br + CH4 -> HBr + CH3 (forward and reverse rates with LCG3 tunneling, including excited final states, from a restart calculation) TR9: CH4 + OH -> CH3 + H2O (VTST-IOC calculation including LCG3 tunneling) TR10: CH4 + OH -> CH3 + H2O (VTST-IOC calculation including LCG3 tunneling) AMSOLRATE 8.6/P8.5.1-A6.6 Page 9-3 TR11: CH3Cl + F-(H2O) -> CH3F + Cl-(H2O) (TST and TST-IOC calculations) TR12: CF3 + CD3H -> CF3H + CD3 (VTST-IOC calculations including LCG3 tunneling) TR13: N2H2 + H -> N2H + H2 (ICL-ECKART calculation with AM1 surface) TR14: HBr + C2H2 -> CHBrCH2 (Redundant curvilinear internal coordinates with AM1 surface) TR15: Cl- + CH3Cl -> CH3Cl + Cl- (symmetric SN2 reaction with wells, AM1 VTST-ISPE calculation with VRPE method for reaction path following) In /solution TR1: CH3Cl + Br- -> CH3Br + Cl- (TST rate of SN2 reaction with SM5.4/AM1 in methanol) TR2: CH3Cl + Cl- -> CH3Cl + Cl- (SM5.4/AM1 calculation of aqueous SN2 reaction) TR3: NH4+ + NH3 -> NH3 + NH4+ (SM5.4/PM3 calculation of aqueous proton transfer reaction) TR4: NH4+ + NH3 -> NH3 + NH4+ (SM5.2R/PM3//PM3 calculation of aqueous proton transfer reaction) TR5: CH3Cl + NH3 -> Cl- + CH3NH3+ (SM5.4/PM3 calculation of aqueous Menshutkin reaction) TR6: CH3Cl + NH3 -> Cl- + CH3NH3+ (SM5.4/PM3//PM3 calculation of aqueous Menshutkin reaction) TR7: HCOOH + HCOOH' -> HCOOH'+ HCOOH (SM5.2R/PM3//PM3 calculation of aqueous double proton transfer reaction) TR8: NH4+ + NH3 -> NH3 + NH4+ (SM5.2R/PM3//PM3 calculation of aqueous proton transfer reaction using VTST-ISPE) AMSOLRATE 8.6/P8.5.1-A6.6 Page 9-3 9.A. GAS-PHASE REACTIONS 9.A.1. Gas test run 1 ---------------------- H3C-CH=CH-CH=CH2 -> H2C=CH-CH=CH-CH3 TST rate constants only Potential : PM3 Vibrations : harmonic RODS : no Dynamics : TST Tunneling : no Dual Level : no Solvation : no This sample input is for a conventional transition state theory calculation for the 1,5-hydrogen shift in the cis-1,3-pentadiene molecule using a PM3 potential energy surface. The harmonic approximation is assumed for all normal modes and generalized normal modes. Neither the MEP nor any transmission coefficients are calculated. Conventional transition state theory (TST) rate constants are calculated at seven temperatures. The activation energy is calculated for the interval 460-480K. AMSOLRATE I/O Files AMSOLRATE Filenames testr1.70 esp.fu70 input data for phase testr1.71 esp.fu71 input data for reactant 1 testr1.73 esp.fu73 input data for product 1 testr1.75 esp.fu75 input data for the saddle point testr1.dat poly.fu5 input data for AMSOLRATE testr1.fu6 poly.fu6 long output file testr1.fu15 poly.fu15 summary output file AMSOLRATE 8.6/P8.5.1-A6.6 Page 9-4 9.A.2. Gas test run 2 ---------------------- CD3 + CH4 -> CD3H + CH3 TST rate constants only Potential : MNDO Vibrations : harmonic RODS : no Dynamics : TST Tunneling : no Dual Level : no Solvation : no This sample input is for an MNDO direct dynamics calculation for the CD3 + CH4 -> CD3H + CH3 reaction. Only conventional transition state theory rate constants are computed. AMSOLRATE I/O Files AMSOLRATE Filenames testr2.70 esp.fu70 input data for phase testr2.71 esp.fu71 input data for reactant 1 testr2.72 esp.fu72 input data for reactant 2 testr2.73 esp.fu73 input data for product 1 testr2.74 esp.fu74 input data for product 2 testr2.75 esp.fu75 input data for the saddle point testr2.dat poly.fu5 input data for AMSOLRATE testr2.fu6 poly.fu6 long output file testr2.fu15 poly.fu15 summary output file AMSOLRATE 8.6/P8.5.1-A6.6 Page 9-5 9.A.3. Gas test run 3 ---------------------- Cl- + CH3Cl -> CH3Cl + Cl- AM1 harmonic TST and CVT rate constants Potential : AM1 (low-level) Vibrations : harmonic rectilinear RODS : no Dynamics : TST, CVT, ICVT Tunneling : SCT Dual Level : ICA for vibrations and ECKART method for VMEP Solvation : no This sample input is for an AM1 and an AM1-IOC direct dynamics calculation on the Cl- + CH3Cl -> CH3Cl + Cl- SN2 reaction. All normal modes and generalized normal modes are treated using the harmonic approximation. The Page-McIver integration method is used with a gradient step size of 1.E-2 bohr to follow the MEP, and a save step size of 2.0E-2 bohr is used between generalized normal mode analyses. Both MEPSAG and CD-SCSAG tunneling corrections are computed, and both CVT and ICVT rate constants are calculated at three temperatures. Derivatives of the gradient with respect to the reaction coordinate, as required for the CD-SCSAG calculation, are found by a quadratic fit. Higher-level data for the VTST-IOC calculations are obtained from the MP2 calculations of S. C. Tucker and D. G. Truhlar, J. Phys. Chem. 93, 8138 (1989). Note: In the poly.fu5 input data file the masses of the species in this reaction are arranged in the following order: C, Cl, Cl, H, H, H. The indices used for the reactant and product specifications in the poly.fu5 file correspond to the order of the masses; i.e. reactant 1 is Cl-, which is index 2, reactant 2 is CH3Cl, represented by indices 1, 3, 4, 5, and 6, and similarly for the products. The esp.fu71 through esp.fu75 files must be consistent with the representation used for the species in the poly.fu5 file. For example, esp.fu71, which represents reactant 1, must be a AMSOL-type input data file for Cl-. AMSOLRATE I/O Files AMSOLRATE Filenames testr3.70 esp.fu70 input data for phase testr3.71 esp.fu71 input data for reactant 1 testr3.72 esp.fu72 input data for reactant 2 testr3.73 esp.fu73 input data for product 1 testr3.74 esp.fu74 input data for product 2 testr3.75 esp.fu75 input data for the saddle point testr3.dat poly.fu5 input data for AMSOLRATE testr3.fu6 poly.fu6 long output file testr3.fu15 poly.fu15 summary output file testr3.50 poly.fu50 additional input data for the VTST-IOC calculation AMSOLRATE 8.6/P8.5.1-A6.6 Page 9-6 9.A.4. Gas test run 4 ---------------------- Cl- + CH3Br -> CH3Cl + Br- AM1 harmonic TST and CVT rate constants Potential : AM1 Vibrations : harmonic redundant curvilinear RODS : no Dynamics : TST, CVT Tunneling : no Dual Level : no Solution : no This sample input is for an AM1 direct dynamics calculation on the Cl- + CH3Br -> CH3Cl + Br- exoergic SN2 reaction. All normal modes and generalized normal modes are treated using the harmonic approximation with redundant curvilinear coordinates. The Page-McIver integration method with cubic starting algorithm is used. A gradient step size of 5.E-3 bohr is used to follow the MEP, and a save step size of 1.0E-2 bohr is used between generalized normal mode analyses. A restart file containing reaction-path information is written to unit fu1 for use in test run 5. No tunneling corrections are computed, and only forward CVT reaction rate constants are calculated at two temperatures. AMSOLRATE I/O Files AMSOLRATE Filenames testr4.70 esp.fu70 input data for phase testr4.71 esp.fu71 input data for reactant 1 testr4.72 esp.fu72 input data for reactant 2 testr4.73 esp.fu73 input data for product 1 testr4.74 esp.fu74 input data for product 2 testr4.75 esp.fu75 input data for the saddle point testr4.dat poly.fu5 input data for AMSOLRATE testr4.fu6 poly.fu6 long output file testr4.fu1 poly.fu1 reaction path information testr4.fu15 poly.fu15 summary output file AMSOLRATE 8.6/P8.5.1-A6.6 Page 9-7 9.A.5. Gas test run 5 ---------------------- Cl- + CH3Br -> CH3Cl + Br- AM1 restart calculation Potential : AM1 Vibrations : harmonic redundant curvilinear RODS : no Dynamics : TST, CVT Tunneling : no Dual Level : no Solvation : no This sample input is for an AM1 direct dynamics calculation on the Cl- + CH3Br -> CH3Cl + Br- exoergic SN2 reaction. The reaction-path information is read from a restart file on unit fu1. The restart file must be created by running test run 4 prior to running test run 6. The reaction rate constants are calculated for the forward and the reverse directions, and the rate constants are calculated at six temperatures. All other parameters are the same as those in the previous test run. AMSOLRATE I/O Files AMSOLRATE Filenames testr5.70 esp.fu70 input data for phase testr5.dat poly.fu5 input data for AMSOLRATE testr4.fu1 poly.fu1 reaction path information (input data for a restart calculation) testr5.fu6 poly.fu6 long output file testr5.fu15 poly.fu15 summary output file AMSOLRATE 8.6/P8.5.1-A6.6 Page 9-8 9.A.6. Gas test run 6 ---------------------- CD3 + CH4 -> CD3H + CH3 LCG3 ground-state-to-ground-state Potential : MNDO Vibrations : harmonic redundant curvilinear RODS : no Dynamics : TST, CVT Tunneling : ZCT, SCT, LCT, muOMT Dual Level : no Solvation : no This sample input is for an MNDO direct dynamics calculation for the CD3 + CH4 -> CD3H + CH3 reaction. All normal modes and generalized normal modes are treated using the harmonic approximation. The Euler integrator is used. A gradient step size of 4.E-3 bohr is used to follow the MEP, and a save step size of 1.6E-2 bohr is used between generalized normal mode analyses. Zero-(ZCT), small-(SCT), and large-(LCT) curvature tunneling contributions are computed, and both forward and reverse reaction rate constants are calculated. In the LCG3 calculations, only ground-state-to- ground-state tunneling processes are included. AMSOLRATE I/O Files AMSOLRATE Filenames testr6.70 esp.fu70 input data for phase testr6.71 esp.fu71 input data for reactant 1 testr6.72 esp.fu72 input data for reactant 2 testr6.73 esp.fu73 input data for product 1 testr6.74 esp.fu74 input data for product 2 testr6.75 esp.fu75 input data for the saddle point testr6.dat poly.fu5 input data for AMSOLRATE testr6.fu6 poly.fu6 long output file testr6.fu14 poly.fu14 summary output file testr6.fu15 poly.fu15 summary output file AMSOLRATE 8.6/P8.5.1-A6.6 Page 9-9 9.A.7. Gas test run 7 ----------------------- Br + CH4 -> HBr + CH3 LCG3 ground-state-to-ground-state Potential : MNDO Vibrations : harmonic redundant curvilinear RODS : no Dynamics : TST, CVT Tunneling : ZCT, SCT, LCT, muOMT Dual Level : no Solvation : no This sample input is for an MNDO direct dynamics calculation for the Br + CH4 -> BrH + CH3 reaction. All normal modes and generalized normal modes are treated using the harmonic approximation. The Euler stabilization- version 1 (ES1) method integrator is used. A gradient step size of 1.E-3 bohr is used to follow the MEP, and a save step size of 2.0E-2 bohr is used between generalized normal mode analyses. A generalized normal mode analysis is also performed at eighty-six special save points. Zero-(ZCT), small-(SCT), and large-(LCT) curvature ground-state tunneling contributions are computed, and both forward and reverse reaction rate constants are calculated. In the LCG3 calculations, only ground- state-to-ground-state processes are included. A restart file containing reaction-path information is written to unit fu1 for use in test run 11. AMSOLRATE I/O Files AMSOLRATE Filenames testr7.70 esp.fu70 input data for phase testr7.71 esp.fu71 input data for reactant 1 testr7.72 esp.fu72 input data for reactant 2 testr7.73 esp.fu73 input data for product 1 testr7.74 esp.fu74 input data for product 2 testr7.75 esp.fu75 input data for the saddle point testr7.dat poly.fu5 input data for AMSOLRATE testr7.fu6 poly.fu6 long output file testr7.fu1 poly.fu1 reaction path information testr7.fu14 poly.fu14 summary output file testr7.fu15 poly.fu15 summary output file AMSOLRATE 8.6/P8.5.1-A6.6 Page 9-10 9.A.8. Gas test run 8 ----------------------- Br + CH4 -> HBr + CH3 LCG3 with tunneling into excited states Potential : MNDO Vibrations : harmonic redundant curvilinear RODS : no Dynamics : TST, CVT Tunneling : ZCT, SCT, LCT, muOMT Dual Level : no Solvation : no This sample input is for an MNDO direct dynamics calculation for the Br + CH4 -> BrH + CH3 reaction. This calculation is a restart calculation which uses the reaction path information in the restart file generated by test run 10. The parameters are the same as for test run 8 except for the large-curvature tunneling option which is set to include all accessible product excited states. This directs the program to carry out tunneling calculations for ground-state transmission coefficients by the ZCT, SCT, and LCG3 methods and to include excited final states in the LCG3 calculation. For the LCG3 calculation, the program determines the highest excited state which is accessible in the LCG3 exit arrangement, and probabilities for tunneling into those excited states are included in the LCG3 calculation. There is approximately a 10% increase in the LCG3 kappa factors in this calculation versus those from test run 10. Note that the reaction as written is endoergic. Therefore the code treats HBr + CH3 as the LCG3 entrance arrangement, in which only the ground state is included in tunneling calculations, and it treats Br + CH4 as the LCG3 exit arrangement, in which excited states are considered as well. The forward transmission coefficient is set equal to the backward one because of microscopic reversibility. AMSOLRATE I/O Files AMSOLRATE Filenames testr8.70 esp.fu70 input data for phase testr8.75 esp.fu75 input data for the saddle point testr7.fu1 poly.fu1 reaction path information (input data for a restart calculation) testr8.dat poly.fu5 input data for AMSOLRATE testr8.fu6 poly.fu6 long output file testr8.fu14 poly.fu14 summary output file testr8.fu15 poly.fu15 summary output file AMSOLRATE 8.6/P8.5.1-A6.6 Page 9-11 9.A.9. Gas test run 9 ----------------------- CH4 + OH -> CH3 + H2O LCG3 tunneling Potential : PM3 (low-level) Vibrations : harmonic rectilinear with one hindered-rotor mode RODS : no Dynamics : TST, CVT Tunneling : ZCT, SCT, LCT, muOMT Dual Level : ICA for vibrations, original interpolated IC method for VMEP Solvation : no This sample input is for a PM3 and a PM3-IOC direct dynamics calculation for the CH4 + OH -> CH3 + H2O reaction. The lowest vibrational mode of the generalized transition state is treated as a hindered rotator. All other modes are treated using the harmonic approximation. A gradient step size of 2.E-3 bohr is used to follow the MEP, and a save step size of 1.E-2 bohr is used between generalized normal mode analyses. Zero-, small-, and large- curvature tunneling corrections are computed and the microcanonical SCT and LCT tunneling probabilities are used to compute the muOMT tunneling approximation. The forward reaction rate constants are calculated at 15 temperatures. The higher-level data for the PM3-IOC calculation are obtained from an adjusted correlation-consistent polarized triple-zeta basis set and Moller-Plesset second-order perturbation theory, scaling all correlation energy(SAC) of V. S. Melissas and D. G. Truhlar, J. Chem. Phys., 1993, 99, 1013. AMSOLRATE I/O Files AMSOLRATE Filenames testr9.70 esp.fu70 input data for phase testr9.71 esp.fu71 input data for reactant 1 testr9.72 esp.fu72 input data for reactant 2 testr9.73 esp.fu73 input data for product 1 testr9.74 esp.fu74 input data for product 2 testr9.75 esp.fu75 input data for the saddle point testr9.dat poly.fu5 input data for AMSOLRATE testr9.fu6 poly.fu6 long output file testr9.fu15 poly.fu15 summary output file testr9.50 poly.fu50 input data for the VTST-IOC calculation AMSOLRATE 8.6/P8.5.1-A6.6 Page 9-12 9.A.10. Gas test run 10 ------------------------ CH4 + OH -> CH3 + H2O LCG3 tunneling Potential : PM3 (low-level) Vibrations : harmonic redundant curvilinear with one hindered-rotor mode RODS : no Dynamics : TST, CVT Tunneling : ZCT, SCT, LCT, muOMT Dual Level : ICA for vibrations, SECKART method for energy Solvation : no This sample input is for a PM3 and a PM3-IOC direct dynamics calculation for the CH4 + OH -> CH3 + H2O reaction. The lowest vibrational mode of the generalized transition state is treated as a hindered rotator. All other modes are treated using the harmonic approximation. A gradient step size of 2.E-3 bohr is used to follow the MEP, and a save step size of 1.E-2 bohr is used between generalized normal mode analyses. Zero-, small-, and large- curvature tunneling corrections are computed. The forward reaction rate constants are calculated at 15 temperatures. The higher-level data for the PM3-IOC calculation are obtained from an adjusted correlation-consistent polarized triple-zeta basis set and Moller-Plesset second-order perturbation theory, scaling all correlation energy(SAC) of V. S. Melissas and D. G. Truhlar, J. Chem. Phys., 1993, 99, 1013. AMSOLRATE I/O Files AMSOLRATE Filenames testr10.70 esp.fu70 input data for phase testr10.71 esp.fu71 input data for reactant 1 testr10.72 esp.fu72 input data for reactant 2 testr10.73 esp.fu73 input data for product 1 testr10.74 esp.fu74 input data for product 2 testr10.75 esp.fu75 input data for the saddle point testr10.dat poly.fu5 input data for AMSOLRATE testr10.fu6 poly.fu6 long output file testr10.fu15 poly.fu15 summary output file testr10.50 poly.fu50 input data for the VTST-IOC calculation AMSOLRATE 8.6/P8.5.1-A6.6 Page 9-13 9.A.11. Gas test run 11 ------------------------ CH3Cl + F-(H2O) -> CH3F + Cl-(H2O) TST and TST-IOC calculations Potential : PM3 (low-level) Vibrations : harmonic RODS : no Dynamics : TST Tunneling : no Dual Level : ICA for vibrations and SECKART method for energy Solvation : no This sample input is for a PM3 and a PM3-IOC TST calculation for the CH3Cl + F-(H2O) -> CH3F + Cl-(H2O) reaction. The TST-IOC calculation is a special case of the VTST-IOC calculation, where all the reactant and saddle point properties are replaced by higher-level ab initio calculations (in this case, MP2/aug-cc-pVDZ calculations). All modes are treated using the harmonic approximation. The forward reaction rate constants are calculated at 3 temperatures. AMSOLRATE I/O Files AMSOLRATE Filenames testr11.70 esp.fu70 input data for phase testr11.71 esp.fu71 input data for reactant 1 testr11.72 esp.fu72 input data for reactant 2 testr11.73 esp.fu73 input data for product 1 testr11.74 esp.fu74 input data for product 2 testr11.75 esp.fu75 input data for the saddle point testr11.dat poly.fu5 input data for AMSOLRATE testr11.fu6 poly.fu6 long output file testr11.fu15 poly.fu15 summary output file testr11.50 poly.fu50 input data for the TST-IOC calculation AMSOLRATE 8.6/P8.5.1-A6.6 Page 9-14 9.A.12. Gas test run 12 ------------------------ CF3 + CD3H -> CF3H + CD3 LCG3 tunneling Potential : MNDO (low-level) Vibrations : rectilinear, harmonic with one hindered-rotor mode RODS : no Dynamics : TST, CVT Tunneling : ZCT, SCT, LCT, muOMT Dual Level : ICA for vibrations and SECKART method for energy Solvation : no This sample input is for a VTST-IOC direct dynamics calculations for the CF3 + CD3H -> CF3H + CD3 reaction. The higher-level surface is chosen to be the AM1-SRP-2 surface from Y.-P. Liu, D.-h. Lu, A. Gonzalez-Lafont, D. G. Truhlar, and B. C. Garrett, J. Am. Chem. Soc. 115, 7806 (1993). The lowest vibrational mode of the generalized transition state is treated as a hindered rotator. All other modes are treated using the harmonic approximation. A gradient step size of 5.E-3 bohr is used to follow the MEP from s = -3.00 bohr to s = 3.00 bohr, and a save step size of 3.E-2 bohr is used between generalized normal mode analyses. Zero-, small-, and large curvature tunneling (ground-state to ground-state) corrections are computed. The forward reaction rate constants are calculated at 11 temperatures. AMSOLRATE I/O Files AMSOLRATE Filenames testr12.70 esp.fu70 input data for phase testr12.71 esp.fu71 input data for reactant 1 testr12.72 esp.fu72 input data for reactant 2 testr12.73 esp.fu73 input data for product 1 testr12.74 esp.fu74 input data for product 2 testr12.75 esp.fu75 input data for the saddle point testr12.dat poly.fu5 input data for AMSOLRATE testr12.fu6 poly.fu6 long output file testr12.fu15 poly.fu15 summary output file testr12.50 poly.fu50 input data for the VTST-IOC calculation AMSOLRATE 8.6/P8.5.1-A6.6 Page 9-15 9.A.13. Gas test run 13 ------------------------ N2H2+ H -> N2H + H2 ICL-DECKART calculation in curvilinear internal coordinates Potential : AM1 (low-level) Vibrations : harmonic nonredundant curvilinear RODS : no Dynamics : TST, CVT Tunneling : ZCT, SCT, LCT, muOMT Dual Level : ICL for vibrations and DECKART method for energy Solvation : no This sample input is for an AM1 and an AM1-ICL-ECKART direct dynamics calculation for the N2H2 + H -> N2H + H2 reaction. The higher-level information is taken from D. P. Linder, X. Duan, and M. Page, J. Chem. Phys. 104, 6298 (1994). Curvilinear internal coordinates are used to obtain better low-frequency modes. Both gradient and save steps are 0.01 bohr with the MEP followed from -1.0 to 0.4 bohr. Zero-, small, and large curvatures tunneling corrections are computed. The forward reaction rate constants are calculated at 9 temperatures with symmetry number equals to 2. AMSOLRATE I/O Files AMSOLRATE Filenames testr13.70 esp.fu70 input data for phase testr13.71 esp.fu71 input data for reactant 1 testr13.72 esp.fu72 input data for reactant 2 testr13.73 esp.fu73 input data for product 1 testr13.74 esp.fu74 input data for product 2 testr13.75 esp.fu75 input data for the saddle point testr13.dat poly.fu5 input data for AMSOLRATE testr13.fu6 poly.fu6 long output file testr13.fu15 poly.fu15 summary output file testr13.50 poly.fu50 input data for the VTST-IOC calculation AMSOLRATE 8.6/P8.5.1-A6.6 Page 9-16 9.A.14. Gas test run 14 ------------------------ HBr+ C2H2 -> CHBrCH2 AM1 calculation in redundant internal coordinates Potential : AM1 Vibrations : harmonic redundant curvilinear RODS : no Dynamics : TST, CVT Tunneling : ZCT, SCT, LCT, muOMT Dual Level : no Solvation : no This sample input is for an AM1 direct dynamics calculation for the HBr + C2H2 -> CHBrCH2 reaction. The harmonic vibrational frequencies are evaluated with the redundant internal coordinates option (curv3). The MEP is followed from -3.3 to 1.5 bohr. The gradient step is 0.01 bohr and the save step is 0.02 bohr. Zero-, small-, and large-curvature tunneling contributions are computed. The forward rate constants are calculated at 9 temperatures. AMSOLRATE I/O Files AMSOLRATE Filenames testr14.70 esp.fu70 input data for phase testr14.71 esp.fu71 input data for reactant 1 testr14.72 esp.fu72 input data for reactant 2 testr14.73 esp.fu73 input data for product 1 testr14.75 esp.fu75 input data for the saddle point testr14.dat poly.fu5 input data for AMSOLRATE testr14.fu6 poly.fu6 long output file testr14.fu15 poly.fu15 summary output file AMSOLRATE 8.6/P8.5.1-A6.6 Page 9-17 9.A.15. Gas test run 15 ------------------------ Cl- + CH3Cl -> CH3Cl + Cl- AM1 harmonic TST and CVT rate constants Potential : AM1 (low-level) Vibrations : harmonic rectilinear RODS : no Dynamics : TST, CVT, ICVT Tunneling : SCT Dual Level : ISPE Solvation : no This sample input is for an AM1 and an AM1 VTST-ISPE direct dynamics calculation on the Cl- + CH3Cl -> CH3Cl + Cl- SN2 reaction. All normal modes and generalized normal modes are treated using the harmonic approximation. The VRPE integration method is used with a gradient step size of 1.E-2 bohr to follow the MEP, and a save step size of 2.0E-2 bohr is used between generalized normal mode analyses. Both MEPSAG and CD-SCSAG tunneling corrections are computed, and both CVT and ICVT rate constants are calculated at three temperatures. Derivatives of the gradient with respect to the reaction coordinate, as required for the CD-SCSAG calculation, are found by a quadratic fit. For the VTST-ISPE calculation, the energies of the two turning points at the representative tunneling energy of the SCT calculated at 300 K are evaluated at the MP2/6-31G** level. AMSOLRATE I/O Files AMSOLRATE Filenames testr15.70 esp.fu70 input data for phase testr15.71 esp.fu71 input data for reactant 1 testr15.72 esp.fu72 input data for reactant 2 testr15.73 esp.fu73 input data for product 1 testr15.74 esp.fu74 input data for product 2 testr15.75 esp.fu75 input data for the saddle point testr15.77 esp.fu77 input data for wellr testr15.78 esp.fu78 input data for wellp testr15.dat poly.fu5 input data for AMSOLRATE testr15.fu6 poly.fu6 long output file testr15.fu15 poly.fu15 summary output file testr15.51 poly.fu51 input data for the VTST-ISPE calculation AMSOLRATE 8.6/P8.5.1-A6.6 Page 9-18 9.B. SOLUTION-PHASE REACTIONS 9.B.1. Solution test run 1 ---------------------------- CH3Cl + Br- -> CH3Br + Cl- SM5.4/AM1//AM1 SES SN2 calculation TST rate only (in methanol) Potential : SM5.4/AM1//AM1 Vibrations : harmonic RODS : no Dynamics : TST Tunneling : no Dual Level : no Solvation : SES, methanol solvent This sample input is for an SES dynamics calculation for the reaction CH3Cl + Br- -> CH3Br + Cl- in methanol. The harmonic vibrational frequencies are evaluated using the AM1 Hamiltonian. The forward rate constant is calculated by conventional transition state theory without tunneling (except for the simple Wigner estimate) at 298 K. AMSOLRATE I/O Files AMSOLRATE Filenames testr1.70 esp.fu70 input data for phase testr1.71 esp.fu71 input data for reactant 1 testr1.72 esp.fu72 input data for reactant 2 testr1.73 esp.fu73 input data for product 1 testr1.74 esp.fu74 input data for product 2 testr1.75 esp.fu75 input data for the saddle point testr1.dat poly.fu5 input data for AMSOLRATE testr1.fu6 poly.fu6 long output file testr1.fu15 poly.fu15 summary output file AMSOLRATE 8.6/P8.5.1-A6.6 Page 9-19 9.B.2. Solution test run 2 ---------------------------- CH3Cl + Cl- -> CH3Cl + Cl- SM5.4/AM1 ESP aqueous SN2 reaction CVT and ICVT calculations with SCT Potential : SM5.4/AM1 Vibrations : harmonic redundant curvilinear RODS : no Dynamics : TST, CVT, ICVT Tunneling : SCT Dual Level : no Solvation : ESP, water solvent This sample input is for an ESP dynamics calculation for the reaction CH3Cl + Cl- -> CH3Cl + Cl- in water. The harmonic vibrational frequencies are evaluated using the SM5.4/AM1 model. The forward rate constant is calculated at 298 K. The scaling mass used to define the isoinertial coordinate system is 20.58 amu. The reaction path is followed with the Page-McIver algorithm with a step size of 0.01 bohr in the range of -0.5 bohr < s < 0.5 bohr. Both ICVT and CVT calculations are carried with SCT tunneling corrections. AMSOLRATE I/O Files AMSOLRATE Filenames testr2.70 esp.fu70 input data for phase testr2.71 esp.fu71 input data for reactant 1 testr2.72 esp.fu72 input data for reactant 2 testr2.73 esp.fu73 input data for product 1 testr2.74 esp.fu74 input data for product 2 testr2.75 esp.fu75 input data for the saddle point testr2.dat poly.fu5 input data for AMSOLRATE testr2.fu6 poly.fu6 long output file testr2.fu15 poly.fu15 summary output file AMSOLRATE 8.6/P8.5.1-A6.6 Page 9-20 9.B.3. Solution test run 3 ---------------------------- NH4+ + NH3 -> NH3 + NH4+ SM5.4/PM3 ESP aqueous reaction CVT calculation with muOMT Potential : SM5.4/PM3 Vibrations : harmonic redundant curvilinear RODS : no Dynamics : TST, CVT Tunneling : ZCT, SCT, LCT, muOMT Dual Level : no Solvation : ESP, water solvent This sample input is for an ESP dynamics calculation for the reaction NH4+ + NH3 -> NH3 + NH4+ in water. The harmonic vibrational frequencies are evaluated using the SM5.4/PM3 model. The forward rate constant is calculated at 298 K. ZCT, SCT, LCT, and muOMT tunneling corrections are calculated with CVT. The scaling mass used to define the isoinertial coordinate system is 1.0 amu. The reaction path is followed with the Page-McIver method with a step size of 0.015 a0 within the range -1.3 a0 < s < +1.3 a0. The lowest real frequency mode is fixed at 60 cm-1 along the path with the LOWFREQ keyword. AMSOLRATE I/O Files AMSOLRATE Filenames testr3.70 esp.fu70 input data for phase testr3.71 esp.fu71 input data for reactant 1 testr3.73 esp.fu74 input data for product 1 testr3.75 esp.fu75 input data for the saddle point testr3.dat poly.fu5 input data for AMSOLRATE testr3.fu6 poly.fu6 long output file testr3.fu15 poly.fu15 summary output file AMSOLRATE 8.6/P8.5.1-A6.6 Page 9-21 9.B.4. Solution test run 4 ---------------------------- NH4+ + NH3 -> NH3 + NH4+ SM5.2R/PM3//PM3 SES aqueous reaction CVT calculation with muOMT Potential : SM5.2R/PM3//PM3 Vibrations : harmonic redundant curvilinear RODS : no Dynamics : TST, CVT Tunneling : ZCT, SCT, LCT, muOMT Dual Level : no Solution : SES, water solvent This sample input is for an SES dynamics calculation for the reaction NH4+ + NH3 -> NH3 + NH4+ in water. The harmonic vibrational frequencies are evaluated using the PM3 hamiltonian. The forward rate constant is calculated at 298 K. ZCT, SCT, LCT, and muOMT tunneling corrections are calculated with CVT. The scaling mass used to define the isoinertial coordinate system is 1.0 amu. The reaction path is followed with the Page-McIver method with a step size of 0.015 a0 within the range -1.3 a0 < s < +1.3 a0. The lowest real frequency mode is fixed at 35 cm-1 along the entire reaction path with the LOWFREQ keyword. AMSOLRATE I/O Files AMSOLRATE Filenames testr4.70 esp.fu70 input data for phase testr4.71 esp.fu71 input data for reactant 1 testr4.73 esp.fu73 input data for product 1 testr4.75 esp.fu75 input data for the saddle point testr4.dat poly.fu5 input data for AMSOLRATE testr4.fu6 poly.fu6 long output file testr4.fu15 poly.fu15 summary output file AMSOLRATE 8.6/P8.5.1-A6.6 Page 9-22 9.B.5. Solution test run 5 ---------------------------- CH3Cl + NH3 -> CH3NH3+ + Cl- SM5.4/PM3 ESP aqueous reaction CVT calculation with muOMT Potential : SM5.4/PM3 Vibrations : harmonic redundant curvilinear RODS : no Dynamics : TST, CVT Tunneling : ZCT, SCT, LCT, muOMT Dual Level : no Solvation : ESP, water solvent This sample input is for an ESP dynamics calculation for the reaction CH3Cl + NH3 -> CH3NH3+ + Cl- in water. The harmonic vibrational frequencies are evaluated using the SM5.4/PM3 hamiltonian. The forward rate constant is calculated at 298 K. ZCT, SCT, LCT, and muOMT tunneling corrections are calculated with CVT. The scaling mass used to define the isoinertial coordinate system is 1.0 amu. The reaction path is followed with the Page-McIver method with a step size of 0.01 a0 within the range -2.0 a0 < s < +1.56 a0. AMSOLRATE I/O Files AMSOLRATE Filenames testr5.70 esp.fu70 input data for phase testr5.71 esp.fu71 input data for reactant 1 testr5.72 esp.fu72 input data for reactant 2 testr5.73 esp.fu73 input data for product 1 testr5.74 esp.fu74 input data for product 2 testr5.75 esp.fu75 input data for the saddle point testr5.dat poly.fu5 input data for AMSOLRATE testr5.fu6 poly.fu6 long output file testr5.fu15 poly.fu15 summary output file AMSOLRATE 8.6/P8.5.1-A6.6 Page 9-23 9.B.6. Solution test run 6 ---------------------------- CH3Cl + NH3 -> CH3NH3+ + Cl- SM5.4/PM3//PM3 SES aqueous reaction CVT calculation with muOMT Potential : SM5.4/PM3//PM3 Vibrations : harmonic redundant curvilinear RODS : no Dynamics : TST, CVT Tunneling : ZCT, SCT, LCT, muOMT Dual Level : no Solvation : SES, water solvent This sample input is for an SES dynamics calculation for the reaction CH3Cl + NH3 -> CH3NH3+ + Cl- in water. The harmonic vibrational frequencies are evaluated using the PM3 hamiltonian. The forward rate constant is calculated at 298 K. ZCT, SCT, LCT, and muOMT tunneling corrections are calculated with CVT. The scaling mass used to define the isoinertial coordinate system is 1.0 amu. The reaction path is followed with the Page-McIver method with a step size of 0.01 a0 within the range -4.0 a0 < s < 2.0 a0. AMSOLRATE I/O Files AMSOLRATE Filenames testr6.70 esp.fu70 input data for phase testr6.71 esp.fu71 input data for reactant 1 testr6.72 esp.fu72 input data for reactant 2 testr6.73 esp.fu73 input data for product 1 testr6.74 esp.fu74 input data for product 2 testr6.75 esp.fu75 input data for the saddle point testr6.dat poly.fu5 input data for AMSOLRATE testr6.fu6 poly.fu6 long output file testr6.fu15 poly.fu15 summary output file AMSOLRATE 8.6/P8.5.1-A6.6 Page 9-24 9.B.7. Solution test run 7 ---------------------------- HCOOH + HCOOH' -> HCOOH' + HCOOH SM5.2R/PM3//PM3 SES aqueous reaction CVT calculation with muOMT Potential : SM5.4/PM3//PM3 Vibrations : harmonic redundant curvilinear RODS : no Dynamics : TST, CVT, LCT, muOMT Tunneling : ZCT, SCT Dual Level : no Solvation : SES, water solvent This sample input is for an SES dynamics calculation for the double proton- transfer reaction HCOOH + HCOOH' -> HCOOH' + HCOOH in water. The harmonic vibrational frequencies are evaluated using the PM3 hamiltonian. The forward rate constant is calculated at 298 K. ZCT, SCT, LCT, and muOMT tunneling corrections are calculated with CVT. The scaling mass used to define the isoinertial coordinate system is 1.0 amu. The reaction path is followed with the Page-McIver method with a step size of 0.01 a0 within the range -1.5 a0 < s < 1.5 a0. AMSOLRATE I/O Files AMSOLRATE Filenames testr7.70 esp.fu70 input data for phase testr7.71 esp.fu71 input data for reactant 1 testr7.72 esp.fu72 input data for reactant 2 testr7.73 esp.fu73 input data for product 1 testr7.74 esp.fu74 input data for product 2 testr7.75 esp.fu75 input data for the saddle point testr7.dat poly.fu5 input data for AMSOLRATE testr7.fu6 poly.fu6 long output file testr7.fu15 poly.fu15 summary output file AMSOLRATE 8.6/P8.5.1-A6.6 Page 9-25 9.B.8. Solution test run 8 ---------------------------- NH4+ + NH3 -> NH3 + NH4+ SM5.2R/PM3//PM3 SES aqueous reaction CVT calculation using VTST-ISPE Potential : SM5.2R/PM3//PM3 Vibrations : harmonic redundant curvilinear RODS : no Dynamics : TST, CVT Tunneling : ZCT, SCT Dual Level : ISPE Solution : SES, water solvent This sample input is for an SES dynamics calculation for the reaction NH4+ + NH3 -> NH3 + NH4+ in water. The harmonic vibrational frequencies are evaluated using the PM3 model. The forward rate constant is calculated at 298 K. ZCT and SCT tunneling corrections are calculated with CVT. The scaling mass used to define the isoinertial coordinate system is 1.0 amu. The reaction path is followed with the Page-McIver method with a step size of 0.015 a0 within the range -1.3 a0 < s < 1.3 a0 The lowest real frequency mode is fixed at 35 cm-1 along the path with the LOWFREQ keyword. SES approximation is performed using VTST-ISPE approach. The result of this test run should be the same as test run 4. AMSOLRATE I/O Files AMSOLRATE Filenames testr8.70 esp.fu70 input data for phase testr8.71 esp.fu71 input data for reactant 1 testr8.72 esp.fu72 input data for reactant 2 testr8.73 esp.fu73 input data for product 1 testr8.74 esp.fu74 input data for product 2 testr8.75 esp.fu75 input data for the saddle point testr8.dat poly.fu5 input data for AMSOLRATE testr8.51 poly.fu50 input data for VTST-ISPE testr8.fu6 poly.fu6 long output file testr8.fu15 poly.fu15 summary output file AMSOLRATE 8.6/P8.5.1-A6.6 Page 9-26 9.B.9. Solution test run 9 ---------------------------- CH3Cl + Cl- -> CH3Cl + Cl- SM5.42R/AM1//AM1 ESP aqueous SN2 reaction CVT and ICVT calculations with SCT Potential : SM5.42R/AM1//AM1 Vibrations : harmonic redundant curvilinear RODS : no Dynamics : TST, CVT, ICVT Tunneling : SCT Dual Level : no Solvation : SES, water solvent This sample input is for an SES dynamics calculation for the reaction CH3Cl + Cl- -> CH3Cl + Cl- in water. The harmonic vibrational frequencies are evaluated using the SM5.42R/AM1 model. The forward rate constant is calculated at 298 K. The scaling mass used to define the isoinertial coordinate system is 20.58 amu. The reaction path is followed with the Page-McIver algorithm with a step size of 0.01 bohr in the range of -0.5 bohr < s < 0.5 bohr. Both ICVT and CVT calculations are carried with SCT tunneling corrections. AMSOLRATE I/O Files AMSOLRATE Filenames testr9.70 esp.fu70 input data for phase testr9.71 esp.fu71 input data for reactant 1 testr9.72 esp.fu72 input data for reactant 2 testr9.73 esp.fu73 input data for product 1 testr9.74 esp.fu74 input data for product 2 testr9.75 esp.fu75 input data for the saddle point testr9.dat poly.fu5 input data for AMSOLRATE testr9.fu6 poly.fu6 long output file AMSOLRATE 8.6/P8.5.1-A6.6 Page 9-28 9.B.10. Solution test run 10 ------------------------------ CH3Cl + Cl- -> CH3Cl + Cl- SM5.4/AM1 NES aqueous SN2 reaction CVT and ICVT calculations with SCT Potential : SM5.4/AM1 Vibrations : harmonic redundant curvilinear RODS : no Dynamics : TST, CVT, ICVT Tunneling : SCT Dual Level : no Solvation : NES, water solvent This sample input is for an NES dynamics calculation for the reaction CH3Cl + Cl- -> CH3Cl + Cl- in water. The method used for the NES calculation is the linear-response method of Ref. 4 of Section 12. The harmonic vibrational frequencies of the solute are evaluated using the SM5.4/AM1 model. The forward rate constant is calculated at 298 K. The scaling mass used to define the isoinertial coordinate system is 20.58 amu. The reaction path is followed with the Page-McIver algorithm with a step size of 0.01 bohr in the range of -0.5 bohr < s < 0.5 bohr. The solvent friction time is 10 fs, and the diffusion coefficients were determined using scheme A which is based on the solvent-accessible surface areas of the atoms; these are the same areas that used in the CDS surface tension calculation of the SM5.4 model. Both ICVT and CVT calculations are carried with SCT tunneling corrections. AMSOLRATE I/O Files AMSOLRATE Filenames testr10.70 esp.fu70 input data for phase testr10.71 esp.fu71 input data for reactant 1 testr10.72 esp.fu72 input data for reactant 2 testr10.73 esp.fu73 input data for product 1 testr10.74 esp.fu74 input data for product 2 testr10.75 esp.fu75 input data for the saddle point testr10.dat poly.fu5 input data for AMSOLRATE testr10.fu6 poly.fu6 long output file testr10.fu15 poly.fu15 summary output file AMSOLRATE 8.6/P8.5.1-A6.6 Page 10-1 10. COMPUTERS AND OPERATING SYSTEMS ON WHICH AMSOLRATE HAS BEEN TESTED Version 7.9.1 Cray C90 (UNICOS 8.0.3) IBM RS/6000 model 590 (AIX 4.1) Silicon Graphics Power Challenge R10000 (IRIX 6.2) Version 8.0 Cray C90 (UNICOS 8.0.3) IBM SP with Power604e (AIX 4.3.1) Silicon Graphics Power Challenge R10000 (IRIX 6.2) Version 8.1.1 IBM SP with Power604e (AIX 4.3.1) Silicon Graphics Power Challenge L with R10000 (IRIX 6.5) Version 8.2 IBM SP with Power604e (AIX 4.3.1) Silicon Graphics Power Challenge L with R10000 (IRIX 6.5) AMSOLRATE 8.6/P8.5.1-A6.6 Page 11-1 11. TEST RUN TIMINGS Here are the timings for the test runs as measured by the Unix time command. The following timings (in seconds) apply to version 8.4: IBM SGI SP/604e Power Challenge L /R10000 in /gas Test run 1 8.0 11.3 Test run 2 3.9 5.6 Test run 3 27.9 31.2 Test run 4 26.1 34.1 Test run 5 0.0 0.0 Test run 6 568.4 518.5 Test run 7 1182.5 836.2 Test run 8 1158.9 845.3 Test run 9 401.9 432.1 Test run 10 416.7 429.9 Test run 11 2.0 2.2 Test run 12 1208.5 1244.8 Test run 13 120.9 127.0 Test run 14 132.9 126.0 Test run 15 95.2 116.9 in /solution Test run 1 1.9 2.1 Test run 2 43.2 59.9 Test run 3 185.5 246.8 Test run 4 152.7 235.2 Test run 5 322.5 448.3 Test run 6 520.6 671.8 Test run 7 180.3 235.5 Test run 8 124.7 259.4 Test run 9 71.5 73.4 Test run 10 45.6 59.9 AMSOLRATE 8.6/P8.5.1-A6.6 Page 12-1 12. BIBLIOGRAPHY For references of solvation models please refer to the AMSOL manual. For references of dynamics calculations with variational transition state theory and multidimensional semiclassical tunneling approximations, please refer to the POLYRATE manual. For an explanation of the methods used for calculating reaction rates in solution, see: 1. "Tunneling in the Presence of a Bath: A Generalized Transition State Theory Approach," D. G. Truhlar, Y.-P. Liu, G. K. Schenter, and B. C. Garrett, Journal of Physical Chemistry 98, 8396-8405 (1994). 2. "The Interface of Electronic Structure and Dynamics for Reactions in Solution," Y.-Y. Chuang, C. J. Cramer, and D. G. Truhlar, International Journal of Quantum Chemistry, submitted for publication in International Journal of Quantum Chemistry 70, 887-896 (1998). (Sanibel issue) 3. "Direct dynamics for Free Radical Kinetics in Solution: Solvent Effect on the Rate Constant for the Reaction of Methanol with Atomic Hydrogen," Y.-Y. Chuang, M. L. Radhakrishnan, P. L. Fast, C. J. Cramer, and D. G. Truhlar, Journal of Physical Chemistry A 103, 4893-4909 (1999). 4. "Nonequilibrium Solvation Effects for a Polyatomic Reaction in Solution," Y.-Y. Chuang and D. G. Truhlar, Journal of American Chemical Society, in press. AMSOLRATE 8.6/P8.5.1-A6.6 Page 13-1 13. NUMSTEP AND SCFCRT The numerical methods in the packages for semi-empirical molecular orbital theory are usually inadequate to provide fully converged values for the frequencies of the low-frequency modes. Due to the lack of analytical gradients in some cases and the lack of analytical Hessians in others, different step sizes (called NUMSTEP in fu5 and called DLX in the code) for the numerical derivatives that are used to calculate Hessians from gradients can result in significantly different low frequencies in various runs. Here are three tables of the tests from various programs (not using POLYRATE). Example 1 is based on default values of convergence parameters, and example 2 and 3 show the effect of varying the SCFCRT convergence parameter of the SCF calculations. 1) The lowest three real frequencies (in cm-1) of the gas-phase complex NH4+...NH3 evaluated at the AMSOL6.5.1 optimized geometry. PM3 Mode 19 Mode 20 Mode 21 ----------------------------------------------------------------- AMSOL6.5.1 296 296 13 AMPAC5.4m1 296 296 6 MOPAC5.07mn 297 296 6 GAUSSIAN94 295 295 19i GAMESS 294 294 60 ------------------------------------------------------------------ MNDO Mode 19 Mode 20 Mode 21 ------------------------------------------------------------------ AMSOL6.5.1 94 94 23 AMPAC5.4m1 94 94 23 MOPAC5.07mn 94 94 20 GAUSSIAN94 93 71i 403i GAMESS 58 37 15 ------------------------------------------------------------------ MNDO/d Mode 19 Mode 20 Mode 21 ------------------------------------------------------------------ AMPAC5.4m1 94 94 6 ------------------------------------------------------------------ AMSOLRATE 8.6/P8.5.1-A6.6 Page 13-2 2) The lowest real frequency (in cm-1) of the optimized saddle point in the gas-phase reaction CH3 + CH4 -> CH4 + CH3. a b SCFCRT AMSOL MOPAC5.07mn MOPAC5.07mn with PRECISE ---------------------------------------------------------------------------- 1E-7 21 63 20 1E-8 24 17i 16 1E-9 24 24 19 1E-10 24 14 23 1E-11 f 15 21 1E-12 f 16 24 1E-13 f 16 22 1E-14 f 16 22 ---------------------------------------------------------------------------- a: This is the value of SCFCRT specified in the input file. b: Note that the PRECISE keyword in MOPAC does two things: 1. forces central differences rather than one-sided differences in the derivations of the gradients. 2. changes the SCF convergence criterion to a factor of 100 tighter than specified by the input value of SCFCRT. f: Failed 3) The lowest real frequency (in cm-1) of the saddle point (with geometry optimized using the MOPAC5.07mn program with PRECISE and SCFCRT=1E-14) in the gas-phase reaction CH3 + CH4 -> CH4 + CH3. a b SCFCRT AMSOL MOPAC5.07mn MOPAC5.07mn with PRECISE ---------------------------------------------------------------------------- 1E-7 24 67 26 1E-8 23 14i 18 1E-9 23 25 19 1E-10 23 14 22 1E-11 f 17 20 1E-12 f 16 22 1E-13 f 16 22 1E-14 f 16 22 ---------------------------------------------------------------------------- a: This is the value of SCFCRT specified in the input file. b: Note that the PRECISE keyword in MOPAC does two things: 1. forces central differences rather than one-sided differences in the derivations of the gradients. 2. changes the SCF convergence criterion to a factor of 100 tighter than specified by the input value of SCFCRT. f: Failed These tables are given simply as a warning that it is the user's responsibility to test stability of the results with respect to NUMSTEP and SCFCRT. The default values of SCFCRT and NUMSTEP are 1E-7 and 0.0001 a0 in AMSOLRATE. However, the recommended values are 1E-9 and 0.011 a0. If stable values for low frequencies cannot be obtained or if the semiempirical method is judged to be unreliable for low-frequency modes, then it may be advisable to use the LOWFREQ keyword as a way to make the runs reproducible and/or more accurate. AMSOLRATE 8.6/P8.5.1-A6.6 Page 14-1 14. ACKNOWLEDGMENTS The development of AMSOLRATE and AMSOL has been supported by grants from the National Science Foundation. The development of POLYRATE has been supported by grants from the U. S. Department of Energy, Office of Basic Energy Sciences.