From CHAMANKHAH@sask.usask.ca Tue Aug 17 17:02:12 1993 Date: Tue, 17 Aug 1993 23:02:12 -0600 (CST) From: CHAMANKHAH@sask.usask.ca Subject: Re: Computational Chemistry Text To: chemistry@ccl.net Message-Id: <01H1VANUBWW88WW655@SKYCAT.USask.CA> Dear Friends I also need an introductory text in computational chemistry which is useful specially for those who don't have good background in the field and didn't take any course before. Yours all Mahmood From dave@carbon.chem.csiro.au Wed Aug 18 10:22:02 1993 Message-Id: <199308180521.AA23816@shark.mel.dit.csiro.au> Date: Wed, 18 Aug 93 15:22:02 EST From: (Dr.) Dave Winkler Subject: Cost of 'mini Cray' To: chemistry@ccl.net I received some information from Cray about their new desktop (or deskside) Cray. It starts at $150,000 (US$ I guess) but they did not give any performance figures in the speil. Cheers, Dave __________________________________________________________________________ Dr. David A. Winkler Voice: 61-3-542-2244 Principal Research Scientist Fax: 61-3-543-8160 CSIRO Division of Chemicals and Polymers Private Bag 10 Clayton, Australia. "Life is what happens to you while you're making other plans" From schw0531@compszrz.zrz.tu-berlin.de Wed Aug 18 12:08:28 1993 Date: Wed, 18 Aug 93 10:08:28 +0200 From: Prof. Dr. Helmut Schwarz Message-Id: <9308180808.AA27552@compszrz.zrz.tu-berlin.de> To: chemistry@ccl.net, ramon@ce.ifisicam.unam.mx Subject: Re: Seeking words of wisdom Dear netters, I would be intersted to know about people working with the GAMESS program from Mike Schmidt on HP workstations. The instalation seems to be more tricky than for IBM. Finally I get the program to run on a HP 750, but in comparisson to uor instalations on IBM/6000 and CONVEX the program is 2-4 times slower. The code optimization during compilation led to an executable giving in few cases wrong results. Jan Hrusak schw0531@compszrz.zrz.tu-berlin.de From rhn900@anusf.anu.edu.au Wed Aug 18 14:31:36 1993 From: Ross Nobes Message-Id: <199308180931.AA01803@anusf.anu.edu.au> Subject: Disks for RS/6000 To: chemistry@ccl.net Date: Wed, 18 Aug 93 19:31:36 EST Dear Readers, The Research School of Chemistry at this University has recently bought an IBM RS/6000 Model 355 to run ab initio calculations on (using mainly Gaussian 92). The machine is configured with 64 Mbytes of memory and is running AIX 3.2. Would anyone have any performance data and/or recommendations on SCSI disks for this machine? For example, how does a Wren 9 compare with an Elite 3? Is disk striping an attractive option under AIX? If so, how would the performance of a striped pair of Wren 9 disks compare with a single Elite 3, for example? Thanks in advance, Ross Nobes -- * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * Ross Nobes * * Academic Consultant * * Supercomputer Facility Phone: + 61 6 249 4154 / 4161 * * Australian National University Fax: + 61 6 247 3425 * * Canberra, ACT 0200 * * Australia E-mail: R.Nobes@anu.edu.au * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * From omar@crs4.it Wed Aug 18 13:29:45 1993 Message-Id: <9308180929.AA18328@malena.crs4.it> To: chemistry@ccl.net (POST MSGS) Subject: Re: Seeking words of wisdom Date: Wed, 18 Aug 1993 11:29:45 +0200 From: "Omar G. Stradella" Your message dated: Tue, 17 Aug 1993 15:44:15 CDT > >Dear Netters: > >In our department we have, out of a sudden, an important amont of $$ to >expend in computer gadgets. > >I am wondering what a good investment could be. That is, we have words >that a mini-CRAY is in the market, and that HP-750 workstations can >give about half the speed of a real CRAY ( that is when one works on >the programming ). I am requesting opinions as to which one of this can >be more suitable for computational chemistry work, which includes >GAUSSIAN 92, MOPAC, Molecular Dynamics and other goodies. I would like >to hear also about their price/performance data (if available) and the cost >of maintainance as well. > >Obviously, if the mini-CRAY is very expensive we would go for several >HP's. > You should also consider the new line of symmetric multiprocessors from Silicon Graphics, the Power Challenge Series. Here you'll find some information (the prices are indicative only): Processor: MIPS SSR (300 Mflops peak) Cache: 32 KB (instruction), 16 KB (data), 2 or 4 MB (secondary) Number of processors: 2-18; System architecture: 64-bit Physical memory capacity: 16 GB; Disk capacity: 3.5 TB Disk performance: 450 MB/s; Standard configuration: 64 MB memory, no disk Power Challenge L: 2 procs. (0.6 Gflops peak): $119,900 4 procs. (1.2 Gflops peak): $209,900 6 procs. (1.8 Gflops peak): $229,900 Power Challenge XL: 2 procs. (0.6 Gflops peak): $169,900 8 procs. (2.4 Gflops peak): $439,900 18 procs. (5.4 Gflops peak): $889,900 Compare to: CRAY Y-MP/832 (1 proc., 6 ns): 0.333 Gflops peak CRAY Y-MP C90 (1 proc., 4.2 ns): 0.952 Gflops peak Most computational chemistry programs (such as those that you mention and more) are supported on SGI's machines. Omar. -- **************************************************************************** * Dr. Omar G. Stradella - Omar.Stradella@crs4.it * * CRS4 - Centro di Ricerca, Sviluppo e Studi Superiori in Sardegna * * - Centre for Advanced Studies, Research and Development in Sardinia * * -------------------------+----------------------------+----------------- * * Casella Postale 488 | Via Nazario Sauro, 10 | 39 13' 19" North * * 09100 Cagliari CA | 09123 Cagliari CA | 9 6' 17" East * * Italy | Italy | * * -------------------------+----------------------------+----------------- * * Phone: +39 70 27 96 415 FAX: +39 70 27 96 400 or +39 70 24 10 18 * **************************************************************************** From ipcakc@vigyan.ernet.in Sun Aug 18 10:35:44 1993 To: CHEMISTRY@ccl.net Subject: HNO L-J parameter Date: 18 Aug 93 15:35:44 EST (Wed) From: ipcakc@vigyan.iisc.ernet.in Message-Id: <9308181535.AA07892@vigyan.iisc.ernet.in> Can anybody give me the Lennard-Jones parameters for H-N-O molecule? I mean epcilon/k and sigma. If it is not available , what will be the nearest approx. molecule for which L-J parameters are available. D.Sengupta IISc Bangalore India. ipcakc@ipc.iisc.ernet.in From J_BROWN@uvmvax.uvm.edu Wed Aug 18 03:48:00 1993 Date: Wed, 18 Aug 1993 08:48 EST From: J_BROWN@uvmvax.uvm.edu Subject: response to question on which computer gadgets to purchase To: chemistry@ccl.net Message-Id: <01H1VV7Y9IIO000LDS@uvmvax.uvm.edu> Ramon Garduno writes: > Dear Netters: > In our department we have, out of a sudden, an important amont of > $$ to expend in computer gadgets. and then goes on to ask about "mini-crays" and "HP-750 workstations" We have several "muscle machines" at the university on a time share network. Due to the nature of the time share system, there are situations where a PC 486 DX is a faster solution (esp. around finals week were computer projects are suddenly "due" for grading). We have downloaded a large amount of mainframe software and compiled it with our WATCOM FORTRAN 77 PC compiler and discovered (much to our joy) that we can actually get some work done. I am not suggesting that you avoid these "towers of power" since some of our work (esp. ab initio calculations with a large number of atoms) take too long even on the fastest PC; hovever, for many applications the PC seems to be both an economical and time saving solution. Take a good look at your needs, you may discover that 20 PC's is worth a "RISC in the bush". - Jay From omar@crs4.it Wed Aug 18 17:24:21 1993 Message-Id: <9308181324.AA23237@malena.crs4.it> To: jle@world.std.com (Joe M Leonard) Subject: Re: Help need synchronizing I/O to standard out on RS/6000 Date: Wed, 18 Aug 1993 15:24:21 +0200 From: "Omar G. Stradella" Your message dated: Tue, 17 Aug 1993 19:36:13 EDT >Not knowing who to contact at IBM anymore, I'll ask the net for advice... > >I'm working with a collection of codes that combine C and F77 routines. >When the code's run from the terminal, the I/O comes out in the correct >order. When the code's redirected into an output file, the F77 output >preceeds the C output - suggesting there's two buffers involved and that >I have a sync issue. There are flush commands to dump buffers on non-IBM >machines, but I need to get the AIX magic... Joe, Yes, as far as I know, the FORTRAN does its own buffering independently from C, so it's not a good idea to mix them. Is it impossible for you to let FORTRAN only (or C only) to do the I/O ?. Anyway, if you have the latest version of xlf (2.3.*.*), there is a new routine to flush the FORTRAN buffers: INTEGER*4 LU CALL FLUSH_(LU) Note the underscore after the name of the rutine. If you want to be sure that it is there, do the following: $ ar -w /usr/lib/libxlf.a | grep flush_ .flush_ shr.o flush_ shr.o Omar. -- **************************************************************************** * Dr. Omar G. Stradella - Omar.Stradella@crs4.it * * CRS4 - Centro di Ricerca, Sviluppo e Studi Superiori in Sardegna * * - Centre for Advanced Studies, Research and Development in Sardinia * * -------------------------+----------------------------+----------------- * * Casella Postale 488 | Via Nazario Sauro, 10 | 39 13' 19" North * * 09100 Cagliari CA | 09123 Cagliari CA | 9 6' 17" East * * Italy | Italy | * * -------------------------+----------------------------+----------------- * * Phone: +39 70 27 96 415 FAX: +39 70 27 96 400 or +39 70 24 10 18 * **************************************************************************** From gerardo@houston-gate.cray.com Wed Aug 18 04:41:01 1993 From: gerardo@houston-gate.cray.com (Gerardo Cisneros) Message-Id: <9308181441.AA20613@houston-gate.cray.com> Subject: Re: Cost of 'mini Cray' To: chemistry@ccl.net Date: Wed, 18 Aug 1993 09:41:01 -0500 (CDT) David Winkler writes > I received some information from Cray about their new desktop (or deskside) > Cray. It starts at $150,000 (US$ I guess) but they did not give any > performance figures in the speil. The EL92 can have up to two processors each rated at 133 MFlops peak. They are binary-compatible with the Y-MP and C90 series systems. The following is an excerpt of the official press release: > The CRAY EL92 is available with one or two CPUs, and 256 to > 512 megabytes (32 to 64 megawords) of real main memory, he > said. The air-cooled system operates on standard 50 Hz and 60 > Hz power, and can easily be installed by customers in office > environments. It is offered with a fast HIPPI interface for > network connectivity applications such as Parallel Virtual > Machine (PVM), or to speed transfers of large datasets to CRAY > 90 and CRAY T3D systems [...] > > [Derek] Robb said the new system runs more than 600 applications > currently operating on larger Cray systems and includes the > company's industry-leading UNICOS operating system, which is > UNIX-based and POSIX-compliant. "The CRAY EL92 system > includes the same operating system, programming > environments, compilers and networking capabilities as more > powerful Cray Research Systems." In addition, he said, the > system can also be ordered with the CRAY T3D Emulator, a > software tool that helps programmers using the CF77 > programming environment to develop and test applications for > the CRAY T3D massively parallel system. I can e-mail the full press release, or post it if that is OK with the list's administrator. Saludos, Gerardo -- Dr. Gerardo Cisneros |Cray Research de M'exico, S.A. de C.V. | This Applications Scientist |Tuxpan 10-403, Col. Roma Sur | space gerardo@cray.com |06760 M'exico, D.F. | for (+52+5)622-8584 |MEXICO | rent From elewars@alchemy.chem.utoronto.ca Wed Aug 18 09:06:46 1993 Date: Wed, 18 Aug 93 13:06:46 -0400 From: elewars@alchemy.chem.utoronto.ca (E. Lewars) Message-Id: <9308181706.AA29939@alchemy.chem.utoronto.ca> To: chemistry@ccl.net Subject: Suggestions for Computational Chem Textbboks Dear netters: This is a response to the two requests for suggestions for introductory computational chemistry textbooks. So far as I know, basically there aren't any, but you might look at: 1. "Exploring Chemistry with Electronic Structure Methods: A Guide to Using Gaussian", J. B. Foresman, A. Frisch; Gaussian Inc., Carnegie Office Park, Building 6, Pittsburgh, PA 15106 (1993); (412) 279-6700. ca. $40. Although this book is oriented toward Gaussian 92, it has information that would be useful to beginners using any kind of quantum mechanical molecular structure program. MMX and semiempirical methods are mentioned only in passing, and Z-matrices are used throughout, although this kind of input is getting a bit oldfashioned. 2. "A Handbook of Computational Chemistry", Tim Clark; Wiley (1985). A nice feature of this book is that it deals with MMX, semiemp. and ab initio. As expected for a book 8 years old, there is much discussion on how to write Z-matrices and no mention of using an interactive model building program to create input. 3. "Computational Chemistry Using the PC", D. W. Rogers; VCH Publishers, Inc., 220 East 23rd St, Suite 909, New York, NY 10010 (1990). Exercises illustrating various principles relevant to computational chemistry. 4. Chapters 15 and, especially, 16 and 17 of "Quantum Chemistry" Ira N. Levine; Prentice Hall (1991). These give a nice, brief overview of computational chemistry today. The bulk of the book is, of course, a course in quantum mechanics. 5. "Ab Initio Molecular Orbital Theory", W. J. Hehre, L. Radom, P. v.R. Schleyer, J. A. Pople; Wiley (1986). Still the bible of ab initio computations, but probably not a textbook for an introductory course. You might also contact the company Wavefunction (18401 Von Karman, Suite 210, CA 92715; FAX (714)955-2118); they are interested in promoting the teaching of computational chemistry, and have produced a tutorial book and some "laboratory exercises" to be used with their SPARTAN program, which is a nice integrated MMX-semiempirical-ab initio package. Another company interested in comp. chem. in education is Autodesk, Inc., FAX(415)331-8093, Tel(415)332-2344, which makes Hyperchem. My (albeit yet brief) experience in teaching computational chemistry has convinced me that: 1. Hands-on practice on the computer is very important, preferably interactively with good graphical input and output, rather than the traditional batch submit-it-and-wait-a-week method. 2. Ab initio and semiempirical methods are best taught by first teaching the simple Hueckel method (which as Hess and Schaad have shown -- J Chem Educ 1974 51 640; Pure Appl Chem 1980 52 1471) -- is NOT obsolete). This can be used to present most of the basic ideas, even the SCF principle: see the books by Roberts and by Zimmerman ("Notes on Molecular Orbital Calculations"; Benjamin (1961). "Quantum Mechanics for Organic Chemists"; Academic Press (1975)). After this basic stuff, really important in my view (no book seems to agree) is to explain how matrices are used in computational chemistry. Book after book mentions matrices and talks about matrix equations and matrix elements, but one looks in vain for a matrix, a rectangular array of elements. In my opinion the only good way to teach their relevance to comp. chem. is to show them in all their glory, and explain their basic properties. Not much of a background in linear algebra is needed to grasp the idea of diagonalization (by an iterative method based ultimately on the old Jacobi method) of the Fock matrix to give a premultiplying (eigenvectors) and diagonal (eigenvalues) matrix. Using matrices properly--writing them out as arrays of rectangular elements--also simplifies teaching how population analysis is done and how vibrational frequencies are calculated (does any book bother to show what a hessian really looks like?). Semiempirical should follow ab initio, as a semiemp. program is an ab initio program that looks up the integrals (to oversimplify). I hope some of this is useful. Errol =============================================================================== elewars@trentu.ca E. Lewars FAX (705)748-1625 Chemistry Dept Trent University Peterborough, Ontario Canada K9J 7B8 =============================================================================== From virtual@quantum.larc.nasa.gov Wed Aug 18 09:33:45 1993 Message-Id: <9308181733.AA00465@quantum.larc.nasa.gov> Date: Wed, 18 Aug 93 13:33:45 -0400 From: Don H. Phillips To: CHEMISTRY@ccl.net Subject: Re: Seeking words of wisdom <199308172144.AA08056@oscsunb.ccl.net> In response to: >I am wondering what a good investment could be. That is, we have words >that a mini-CRAY is in the market, and that HP-750 workstations can >give about half the speed of a real CRAY Dr. Omar G. Stradella responds: >>You should also consider the new line of symmetric multiprocessors from >>Silicon Graphics, the Power Challenge Series. >> ... >>... (deleted specs and peak speed info for sgi and cray boxes) There are a lot of wonderful new machines out there but info on advertised peak speeds is nearly worthless in distinguishing between them. Information on sustained performance on a typical problem using Gaussian92, GAMESS, etc. is helpful to this audience. Actually, a few typical cases which differ in the vectorization of the code used (SCF, gradient, CI, etc) is even better since that shows the effects of differing cache designs. Is any info on the Power Challenge Series, Cray desk(top/side), etc compared to garden variety risc boxes available out there? Don Phillips From JSMCM@jazz.ucc.uno.edu Sun Aug 18 07:24:10 1993 Date: 18 Aug 1993 13:24:10 -0600 (CST) From: "JORGE M. SEMINARIO" Subject: Flushing on VMS To: omar@crs4.it Message-Id: <01H1W4KUAO7U8WW06T@jazz.ucc.uno.edu> Dear Friends, Regarding this very helpful hint by Dr. Omar Stradella: >Anyway, if you have the latest version of xlf (2.3.*.*), there is a new >routine to flush the FORTRAN buffers: > >INTEGER*4 LU > >CALL FLUSH_(LU) > >Note the underscore after the name of the rutine. If you want to be >sure that it is there, do the following: > >$ ar -w /usr/lib/libxlf.a | grep flush_ >.flush_ shr.o >flush_ shr.o I wonder if there is an equivalent way to flush the fortran buffers under VMS (VAX)? Your help is greatly appreciate. Jorge From rc_davis@srs.gov Wed Aug 18 10:37:55 1993 Date: Wed, 18 Aug 1993 14:37:55 -0400 (EDT) From: "Ricardo C. Davis" Subject: Re: Seeking words of wisdom To: "Omar G. Stradella" Message-Id: On Wed, 18 Aug 1993, Omar G. Stradella wrote: > Your message dated: Tue, 17 Aug 1993 15:44:15 CDT > > > >Dear Netters: > > > >In our department we have, out of a sudden, an important amont of $$ to > >expend in computer gadgets... > > You should also consider the new line of symmetric multiprocessors from > Silicon Graphics, the Power Challenge Series. Here you'll find some > information (the prices are indicative only): > > Processor: MIPS SSR (300 Mflops peak) > Cache: 32 KB (instruction), 16 KB (data), 2 or 4 MB (secondary) > Number of processors: 2-18; System architecture: 64-bit > Physical memory capacity: 16 GB; Disk capacity: 3.5 TB > Disk performance: 450 MB/s; Standard configuration: 64 MB memory, no disk > > Power Challenge L: > > 2 procs. (0.6 Gflops peak): $119,900 > 4 procs. (1.2 Gflops peak): $209,900 > 6 procs. (1.8 Gflops peak): $229,900 > > Power Challenge XL: > > 2 procs. (0.6 Gflops peak): $169,900 > 8 procs. (2.4 Gflops peak): $439,900 > 18 procs. (5.4 Gflops peak): $889,900 > > Compare to: > > CRAY Y-MP/832 (1 proc., 6 ns): 0.333 Gflops peak > CRAY Y-MP C90 (1 proc., 4.2 ns): 0.952 Gflops peak > > Most computational chemistry programs (such as those that you mention and > more) are supported on SGI's machines. Two points to beware: 1) Gaussian, Inc. does support G92 on SGI platforms running IRIX (I assume at least v. 4.0), but does that include the Power Challenge machines? Some codes that have been optimized for a particular platform sometimes don't perform well on new hardware. 2) Comparison of peak performance numbers usually means nothing when dealing with benchmarking and application comparison on high-performance computing platforms. (Of course, there is an obvious performance difference in running G92 on a Cray C90 vs. a 486 PC :-). What you really need to do is give some specific performance numbers running similar G92 data input, keeping outside performace factors at a minimum. Ricardo C. Davis Internet: rc_davis@srs.gov Building 773-42A, SRTC/SCS Phone: (803)725-5172 Westinghouse Savannah River Co. 725-8829(FAX) P.O. Box 616, Aiken, SC 29802 USA * Freedom is not the right to choose, freedom is the result of wise choices. From 100012.1163@CompuServe.COM Sun Aug 18 10:42:53 1993 Date: 18 Aug 93 14:42:53 EDT From: "100012.1163@compuserve.com" <100012.1163@compuserve.com> To: Subject: Re: Computat. Chemistry Text Message-Id: <930818184252_100012.1163_BHB60-1@CompuServe.COM> To: > internet:chemistry@ccl.net Dear Netters, it somewhat depends on your lecture. The following book is certainly suitable for teaching students with a background in organic chemistry. I attach Roald Hoffmann's forewaord and the table of contents for your information. Sincerely Dr. Rainer Stumpe Chemistry Editorial Springer-Verlag Tiergartenstr. 17 D-69121 Heidelberg Phone: +49-(0)6221-48 73 10 Fax: +49-(0)6221-41 39 82 INTERNET:stumpe@spint.compuserve.com ------------------------------------------------------------ V.I. Minkin, B.Ya. Simkin, R.M. Minyaev, Rostov University, USSR Quantum Chemistry of Organic Compounds - Mechanisms of Reactions 1990. Approx. 290 pp. 66 figs. 37 tabs. 262 schemes. Hardcover DM 148,- ISBN 3-540-52530-0 This textbook on the application of ab initio and semiempirical techniques for the analysis of organic reaction mechanisms is designed for chemistry undergraduates. The material is presented according to the mechanistic types: nucleophilic and electrophilic substitution, addition reactions, radical, pericyclic, proton and electron transfer reactions. Orbital and electrostatic models are used for structural correlations of interacting molecular systems along the reaction paths. Particular attention is focussed on the characteristics of the transition state. The text combines phenomenology with the basic theoretical principles needed to understand and predict chemical reactivity. This textbook is designed for students of physico-organic chemistry. Quantum-chemical models for the most important reaction mechanisms are presented and help the reader understand and predict the reactivity of molecules. Advanced Textbook ------------------------------------------------------------- Foreword Chemistry is the science of substances (today we would say molecules) and their transformations. Central to this science is the complexity of shape and function of its typical representatives. There lies, no longer dependent on its vitalistic antecedents, the rich realm of molecular possibility called organic chemistry. In this century we have learned how to determine the three- dimensional structure of molecules. Now chemistry as whole, and organic chemistry in particular, is poised to move to the exploration of its dynamic dimension, the busy business of transformations or reactions. Oh, it has been done all along, for what else is synthesis? What 1 mean is that the theoretical framework accom panying organic chemistry, long and fruitfully laboring on a quantum chemical understanding of structure, is now making the first tentative motions toward building an organic theory of reactivity. The Minkin, Simkin, Minyaev book takes us in that direction. lt incorporates the lessons of frontier orbital theory and of Hartree Fock SCF calculations; what chemical physicists have learned about trajectory calculations of selected reactions, and a simplified treatment of all-important solvent effects. lt is written by professional, accomplished organic chemists for other organic chemists; it is consistently even-toned in its presentation of contending approaches. And very much up to date. That this contemporary work should emerge from a regional university in a country in which science has been highly centralized and organic chemistry not very modern, invites reflection. lt is testimony to the openness of the chemical literature, good people, and the irrepressible streaming of the human mind toward wisdom. Cornell University Roald Hoffmann ------------------------------------------------------------- Table of Contents Chapter 1 Potential Energy Surfaces of Chemical Reactions 1 1.1 Introduction. Mechanism of Chemical Reaction and Quantum Chemistry . 1 1.2 Choice of a Coordinate System and the Representation of a PES 2 1.3 Topography of the PES and Properties of a Reacting System 6 1.3.1 Critical Points 6 1.3.2 The Regions of the Minima on the PES . 8 1.3.2.1 Vibrational Spectrum of Molecules 9 1.3.2.2 Calculation of Thermodynamic Functions of Molecules 11 1.3.2.3 Topological Definition of Molecular Structure 13 1.3.2.4 Structural Diagrams 18 1.3.3 Saddle Points on the PES. Transition States 20 1.3.3.1 Localization of the Transition States on the PES 21 1.3.3.2 Symmetry Selection Rules for Transition State Structures 23 1.3.3.3 Calculation of Activation Parameters of Reactions and of Kinetic Isotopic Effects . 27 1.3.4 Pathway of a Chemical Reaction 30 1.3.4.1 Ambiguity of the Definition 30 1.3.4.2 A More Accurate Definition of the MERP and the Reaction Coordinate . 31 1.3.4.3 Symmetry Demands on the Reaction Path 34 1.3.4.4 Chiral and Achiral Pathways of Degenerate Reactions. 36 1.3.5 Empirical Correlations of the Reaction Pathways 39 1.3.5.1 Molecular Vibrations and the Reaction Coordinate. 39 1.3.5.2 The Principle of Least-Motion . 41 1.3.5.3 Structural Correlations of the Pathways of Chemical Reactions 42 1.4 Dynamic Approach 5 1.5 Tunnelling Effects in Chemical Reactions 49 1.6 Description of Nonadiabatic Reactions 53 References 58 Chapter 2 Quantum Chemical Methods for Calculating Potential Energy Surfaces 61 2.1 General Requirements upon the Methods for Calculating Potential Energy Surfaces 61 2.2 Nonempirical (ab initio) Methode. The Hartree Fock Method 62 2.2.1 Closed Electron Shells 62 2.2.2 Open Electron Shells 64 2.2.3 Basis Sets of Atomic Orbitals 66 2.2.4 Electron Correlation 69 2.2.5 The Problem of Stability of Hartree Fock Solutions 75 2.3 Semiempirical Methods 76 2.3.1 The Extended Hckel Method 78 2.3.2 Semiempirical Selfconsistent Field Methods 78 2.3.2.1 The CNDO/2 Method 78 2.3.2.2 The MINDO/3 Method 80 2.3.2.3 The MNDO Method 82 2.3.2.4 The AM l Method 85 References 86 Chapter 3 Effects of the Medium 88 3.1 A General Scheme for Calculating the Solvation Effect", 89 3.2 Macroscopic Approximation 90 3.2.1 General Theory 90 3.2.2 Model Hamiltonians in the Macroscopic Approximation 92 3.2.2.1 Model Hamiltonian in the Kirkwood Approximation 93 3.2.2.2 A Model Hamiltonian Based on the Born Formula. Scheme of Solvatons 93 3.2.2.3 The Scheme of Virtual Charges 95 3.2.2.4 The Theory of Selfconsistent Reactive Field 96 3.3 Discrete Representation of Solvent Molecules. Model Hamiltonians in the Microscopic Approximation 97 3.4 Specific Features of the Supermolecular Approach in Studies of Solvation Effects 100 3.5 Statistical Methods fo)r Studying Solutions 102 References 104 Chapter 4 Orbital Interactions and the Pathway of a Chemical Reaction 106 4.1 The Role of Frontier Orbitals 106 4.2 Theory of Orbital Interactions 108 4.3 Components of the Interaction Energy of a Reacting System in a Transition State . . 4.4 Isolobal Analogy 113 References 115 Chapter 5 Substitution Reaction 116 5.1 Nucleophilic Substitution at a Tetrahedral Carbon Atom. . .117 5.1.1 The SN2 Reactions 118 5.1.1.1 Stereochemistry of the Reactions 118 5.1.1.2 Reaction Coordinate and the Structure of the Transition State 119 5.1.1.3 Energetics and Stochiometric Mechanism of the Gas-Phase SN2 Reactions 121 5.1.1.4 Effect of the Solvent 125 5.1.1.5 Reactions with Retention of Configuration of the Carbon Atom 128 5.1.2 The SNI Reactions 130 5.2 Electrophilic Substitution at the Tetrahedral Carbon Atom 132 5.3 Nucleophilic Substitution at the Carbon Atom of the Carbonyl Group 134 5.3.1 The Stochiometric Mechanism 134 5.3.2 Homogeneous Catalysis 138 5.3.3 Stereochemistry of the Reaction 142 5.3.3.1 The Direction of Nucleophilic Attack and Orbital Steering 142 5.3.3.2 Stereochemical Control of the Breakdown of the Tetrahedral Adduct 144 5.4 Aromatic Electrophilic Substitution Reactions 147 5.5 Nucleophilic Substitution at the Nitrogen, Phosphorus, and Sulfur Centers . . 5.5.1 Substitution at the Nitrogen Atom of Nitroso- and .154 Nitro- Groups 154 5.5.2 Substitution at the Dicoordinate Sulfur Atom 156 5.5.3 Substitution at Tricoordinate Sulfur and Phosphorus Centers 158 5.5.4 Substitution at Tetracoordinate Phosphorus 159 5.5.5 Substitution at Pentacoordinate Phosphorus 161 5.5.6 Inclusion of the Polytopal Rearrangements of intermediates 164 in the Overall Reaction Scheme 165 References Chapter 6 .170 Addition Reactions 6.1 Electrophilic Additions to Multiple Bonds 170 6.2 Nucleophilic Addition to Alkenes 174 6.3 Nucleophilic Addition to a Triple Bond 177 References 179 Chapter 7 Low-Energy Barrier Reactions- Structural Modelling 181 7,1 The Principle of Correspondence Between Structures of the Initial and the Transition State of Reaction 181 7.2 Nucleophilic Rearrangements and Tautomerizations 182 7.3 Cyclization Reactions 186 7.4 Topochemical Reactions 188 References 189 Chapter 8 Radical Reactions 190 8.1 specific Features of the Theoretical Analysis of Radical Reactions 190 8.2 Free-Radical Reactions 191 8.2.1 Bond-Cleavage and Addition Reactions 192 8.2.2 Radical Substitution Reactions at the Tetrahedral Carbon Atom 195 8.3 Reactions with Formation of Biradicals 196 8.4 The Reactions of Carbenes 200 8.4.1 Addition to the 1:)Double Carbon Carbon Bond 201 8.4.2 Insertion into a-Bonds 204 References 207 Chapter 9 Electron and Proton Transfer Reactions 9.1 Electron Transfer Reactions 210 9.1.1 Single Electron Transfer Reactions in organic chemistry 210 9.1.2 Elementary Act of Electron Transfer 211 9.1.3 Theoretical Studies of the Mechanism of SN1 Reactions 214 9.2 Proton Transfer Reactions 217 9.2.1 Potential Energy Curves and Activation Barriers 219 9.2.2 Stereochemistry 221 9.2.3 Proton Transfer in Systems with the Intramolecular Hydrogen Bonding 222 9.2.4 The Tunnelling Mechanism in Proton Transfer Reactions 227 9.2.5 Double Proton Migrations 231 References Chapter 10 Pericyclic Reactions 238 10.1 Reactions of Cycloaddition 239 10.1.1 [2+2]-Cycloaddition 239 10.1.2 [4+2]-Cycloaddition 242 10.2 Electrocyclic Reactions 246 10.3 Sigmatropic Rearrangements 251 10.4 Haptotropic Rearrangements 254 10.5 lon-Radical Pericyclic Reactions 257 References 261 List of Abbreviations 265 Subject Index 267 From jle@world.std.com Wed Aug 18 11:41:48 1993 Date: Wed, 18 Aug 1993 15:41:48 -0400 From: jle@world.std.com (Joe M Leonard) Message-Id: <199308181941.AA11066@world.std.com> To: chemistry@ccl.net Subject: Answer for my synchronization problem on the RS 6000 Well, it seems that the solution is to call a streams/IO function in the standard library - setlinebuf(stdout) at the start of the C parent routine. The quickie test says that it works - hope nothing untoward happens! Thanks to Pedro Vazquez (vazquez@iqm.unicamp.br) for the help!! Joe Leonard jle@world.std.com From kmoore@ncsc.org Wed Aug 18 11:37:29 1993 Date: Wed, 18 Aug 93 15:37:29 EDT From: Kevin Moore Message-Id: <9308181937.AA00964@duck.ncsc.org> To: chemistry@ccl.net Subject: response to question on which computer gadgets to purchase Date: Wed, 18 Aug 1993 08:48 EST From: J_BROWN@uvmvax.uvm.edu X-Vms-To: IN%"chemistry@ccl.net" Sender: chemistry-request@ccl.net Precedence: bulk Ramon Garduno writes: > Dear Netters: > In our department we have, out of a sudden, an important amont of > $$ to expend in computer gadgets. and then goes on to ask about "mini-crays" and "HP-750 workstations" We have several "muscle machines" at the university on a time share network. Due to the nature of the time share system, there are situations where a PC 486 DX is a faster solution (esp. around finals week were computer projects are suddenly "due" for grading). We have downloaded a large amount of mainframe software and compiled it with our WATCOM FORTRAN 77 PC compiler and discovered (much to our joy) that we can actually get some work done. I am not suggesting that you avoid these "towers of power" since some of our work (esp. ab initio calculations with a large number of atoms) take too long even on the fastest PC; hovever, for many applications the PC seems to be both an economical and time saving solution. Take a good look at your needs, you may discover that 20 PC's is worth a "RISC in the bush". - Jay Take a look at the DEC-Alpha. You can get one of them for less than 20K and you won't find 20 486's being worth one of them. They scream and for the price are hard to beat. There desktop models start at 6K and leave the 486 and Pentium in the dust (not to mention the Indigo which costs significantly more). The only thing they need is software. BTW, Gaussian has been ported along with most of the MSI software (including Quanta, Charmm and Xplor I believe). Things like Mopac will take a recompile. It is worth a long hard look! I'm not a DEC salesman and own no stock in the company, so no flames please. ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ || Kevin Moore North Carolina Supercomputing Center || || Scientific Support Analyst 3021 Cornwallis Rd. || || (919) 248-1179 Research Triangle Park, NC 27709 || ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ From d3f012@pellucidar.pnl.gov Wed Aug 18 07:27:28 1993 Date: Wed, 18 Aug 93 14:27:28 -0700 From: d3f012@pellucidar.pnl.gov Subject: Water Simulations To: dibug@comp.bioz.unibas.ch, chemistry@ccl.net Message-Id: <9308182127.AA02619@pellucidar.pnl.gov> Some time ago, I posted responses to an inquiry I made regarding the potential used in Discover (from Biosym) for simulating water. Just as a reminder, Biosym uses: 1) flexible water. 2) SPC LJ parameters for Oxygen. 3) TIP3P internal geometry. 4) LJ "A" param for hydrogen of 10^-8 (kcal Ang^12)/mol. Note, #4 is necessary to prevent a "coulomb catastrophe" that flexible water would allow if there were only LJ params on the oxygen, and one used Aij = sqrt(Aii*Ajj) as Discover does. In other words, without #4, the hydrogens are ripped away from their respective oxygens by the neighboring waters' oxygens (unless one uses rigid water). I have run some test simulations on SPC and Biosym water (hereafter referred to as BW) to compare the radial distribution functions between SPC and BW. I did this because, Biosym mentioned that the O-O rdf shows the second peak being off by about 1 angstrom and I wanted to see if I could reproduce this and understand why this second peak was off. The reference calculations I'll refer to are given in Jorgensen, et. al. J. Chem. Phys. vol 79, (1983) p. 926. Original reference to SPC is Berendsen, et. al. in "Intermolecular Forces" ed. B. Pullman (Reidel), (1981) p. 331. Also I'll refer to "Structural and energetic effects of truncating long ranged interactions in ionic and polar fluids" Brooks C. L., Pettitt and Karplus J. Chem. Phys. vol 83, (1985), p. 5897. The two water calculations I ran were: A). SPC (SHAKE), NVT, 1.0 fs. time step. 298 K, periodic boundary conditions (18.6 ang. box), 218 waters, minimum image convention. Equilibrate for 50 ps. simulate for 50 ps. Group cutoffs at 9 angstroms, rebuild NB list every 10 steps, Temp. coupling constant of 0.4 ps. Save configuration every 10 fs. B) Biosym Water, same as A) except used 0.5 fs. steps (1.0 fs steps shows poor energy conservation for NVE; consequence of flexible water), 30 ps. simulation. My rdf's for SPC water for O-O, O-H, and H-H are visually indistinguishable to those reported by Jorgensen. As reported by Biosym, the first peaks of the O-O, O-H, and H-H rdfs using BW were very close in position and height to the published SPC values as well. However, the O-O rdf for BW is very flat after the first peak which is what one expects from TIP3P. The results reported from Biosym indicated that there was a second peak in the BW O-O rdf that was off by ~1.0 Ang. I don't see a discernable second peak. I suspect that the discrepancy in the O-O rdf may be due to Biosym's use of switching functions. I have not confirmed this yet with a calculation, but I suspect it's true, based on the Brooks paper cited above. Brooks et. al. observed that switching and shifting functions can noticeably alter the O-O rdf functions beyond the first peak. Another recent reference on the effects of switching distances is Ding, Karasawa and Goddard Chem. Phys. Letts, vol 193, (1992), p. 197. I have not investigated the effect of using 1.0 fs time steps with flexible water, mainly, because the energy conservation should be poor if run in the NVE ensemble. Overall, based soley on the rdf functions, I would say that Biosym water is not better, but perhaps not a lot worse than SPC or TIP3P. Caveats to my observations: 1) I have not calculated other properties, and can't say how well BW would do. 2) These results are for small molecules. There can be other interpretations to the effects of truncation based on simulating macromolecules, such as discussed in "The Effects of Truncating Long-Range Forces on Protein Dynamics", Loncharich and Bernard Brooks, Proteins, vol 6, (1989), p. 32. 3) The comparison of hard cutoffs, switching functions, and shifting functions, will be affected by the actual cutoff distance as well. A more important question for me is what is the motivation for Biosym doing water in this fashion in the first place? Has anyone explored the effects of switching functions on rdf's for Biosym water? I'd be interested in hearing others experiences and results. Can anyone tell me what is the functional form of the Biosym switching function? Mark Thompson ************************************************************************** Mark A. Thompson Sr. Research Scientist email: d3f012@pnlg.pnl.gov Molecular Science Research Center FAX : 509-375-6631 Pacific Northwest Laboratory voice: 509-375-6734 PO Box 999, Mail Stop K1-90 Richland, WA. 99352 Disclaimer: The views expressed in this message are solely my own and do not represent Battelle Memorial Institute, Pacific Northwest Laboratory, or any of its clients. ************************************************************************** From b_duke@lacebark.ntu.edu.au Thu Aug 19 03:11:37 1993 From: b_duke@lacebark.ntu.edu.au Message-Id: <9308191411.AA08864@lacebark.ntu.edu.au> Subject: Re: Comp Chemistry texts. To: CHEMISTRY@ccl.net (chemistry) Date: Thu, 19 Aug 1993 09:11:37 -0600 (CDT) Recent posts have emhasised a number of books that, IMHO, are better suited for graduate work. There really is'nt much for undergraduates. I do quite a lot of computational quantum chemistry with our final year undergraduates. We use Huckel, EHM, and ab initio (GAUSSIAN). It is important to have well written programs for things like Huckel and EHM. I give people fairly extensive "laboratory" manuals for the work. One of these exercises was described by us in J Chem Ed last year - 69, 529, 1992. Students can refer to a number of texts but they are all introductory bonding theory texts such as "Coulson", "Murrell, Kettle and Tedder" etc. At a higher level, best used by graduate students, are:- D M Hirst "A Computational Approach to Chemistry", Blackwell Scientific Pub, 1990. A Hinchliffe "Computational Quantum Chemistry" Wiley, Chichester, 1988. Both authors are British, so they may be less well known in the US. They are both good books, Cheers, Brian. -- Associate Professor Brian Salter-Duke (Brian Duke) School of Chemistry and Earth Sciences, Northern Territory University, Box 40146, Casuarina, NT 0811, Australia. Phone 089-466702 e-mail: b_duke@lacebark.ntu.edu.au or b_duke@darwin.ntu.edu.au From JSMCM@jazz.ucc.uno.edu Sun Aug 18 15:46:20 1993 Date: 18 Aug 1993 21:46:20 -0600 (CST) From: "JORGE M. SEMINARIO" Subject: Re: Water Simulations To: d3f012@pellucidar.pnl.gov Message-Id: <01H1WLJRT39M8WW274@jazz.ucc.uno.edu> Dear Friends, Regarding the following post by Mark Thompson: >Just as a reminder, Biosym uses: > >1) flexible water. >2) SPC LJ parameters for Oxygen. >3) TIP3P internal geometry. >4) LJ "A" param for hydrogen of 10^-8 (kcal Ang^12)/mol. I recall that sometime ago I took a look to the force field available in DISCOVER and found out, just by looking at the manuals, that the force field for water contains also nondiagonal terms like: ... + SUM SUM F[b,theta]*(b-bo)(theta-thetao) b theta + SUM SUM F[b,b']*(b-bo)*(b'-bo') + ... b b' I even found the parameters on a Table explicitely stated for for water to be F[b,b']= -14.5 F[b,theta]= 31.3 Are those terms considered in your calculation? Can someone comment on those nondiagonal terms. Thanks, Jorge