From cmartin@rainbow.uchicago.edu Fri Sep 30 00:11:14 1994 Received: from midway.uchicago.edu for cmartin@rainbow.uchicago.edu by www.ccl.net (8.6.9/930601.1506) id XAA01795; Thu, 29 Sep 1994 23:41:46 -0400 Received: from rainbow.uchicago.edu by midway.uchicago.edu for chemistry@ccl.net Thu, 29 Sep 94 22:41:46 CDT Received: by rainbow.uchicago.edu (920330.SGI/920502.SGI) for @midway.uchicago.edu:chemistry@ccl.net id AA26345; Thu, 29 Sep 94 23:04:32 -0500 From: cmartin@rainbow.uchicago.edu (Charles Martin) Message-Id: <9409300404.AA26345@rainbow.uchicago.edu> Subject: charges and pertrurbation theories To: chemistry@ccl.net Date: Thu, 29 Sep 1994 23:04:31 -0600 (CDT) X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 275 Netters: I agree with Wendy. I would like to hear more about these perturbation theory methods. Additionally, I would like to hear what some people think about the limitations of the current models for electrostatics in general. Chuck Martin The University of Chicago From qftramos@usc.es Fri Sep 30 06:11:20 1994 Received: from uscmail.usc.es for qftramos@usc.es by www.ccl.net (8.6.9/930601.1506) id GAA04301; Fri, 30 Sep 1994 06:02:48 -0400 Received: by uscmail.usc.es (4.1/1.34) id AA17212; Fri, 30 Sep 94 11:03:52 +0100 Date: Fri, 30 Sep 94 11:03:52 +0100 From: qftramos@usc.es (Antonio Fernandez Ramos) Message-Id: <9409301003.AA17212@uscmail.usc.es> To: chemistry@ccl.net Subject: semiclassical tunneling Dear netters: I'm interested in the semiclassical tunneling models. Particularly, the models developed by N. Makri and W. H. Miller are a powerful tool to calculate the splitting in symmetric systems where there are protonic transference. Do programs exist to make these calculations (using basis set methods, TDSCF or/and classical trajectory simulations), if I have output data of ab-initio calculations (minima, TS, frequencies, IRC, LRP, etc ..)? Thanks A. F. Ramos E-mail: qftramos@usc.es From frits@rulglj.LeidenUniv.nl Fri Sep 30 06:28:34 1994 Received: from rulway.LeidenUniv.nl for frits@rulglj.LeidenUniv.nl by www.ccl.net (8.6.9/930601.1506) id FAA04154; Fri, 30 Sep 1994 05:18:57 -0400 Received: from rulglj.LeidenUniv.nl by rulway.LeidenUniv.nl with SMTP id AA18363 (5.65c+/IDA-1.4.4 for <@RULWAY.LEIDENUNIV.NL:chemistry@ccl.net>); Fri, 30 Sep 1994 11:18:50 +0200 Received: by rulglj.LeidenUniv.nl (920330.SGI/920502.SGI) for @RULWAY.LEIDENUNIV.NL:chemistry@ccl.net id AA17736; Fri, 30 Sep 94 10:19:27 GMT Date: Fri, 30 Sep 94 10:19:27 GMT From: frits@rulglj.LeidenUniv.nl (Frits Daalmans) Message-Id: <9409301019.AA17736@rulglj.LeidenUniv.nl> To: chemistry@ccl.net Subject: large 2-electron files on IBM RS6000; small suggestion (mail to criscuol@gensia.edu bounced, so I send this to the list) To: criscuol@gensia.edu Subject: Re: CCL:Gaussian92 run time problem under AIX 3.2.5 Well.. I know of a "dirty trick" that I used on a NAS9000 mainframe running under CMS: the integral file in the version of GAMESS I used then was called mt2 instead of ed2, which already suggests what kind of file it was: I think (but I could be wrong) that integral files are sequentially written and read, so you could possibly use a tapestreamer instead of a disk to write the integrals to. This assumes that you have unlimited CPU time, of course, since reading/writing from tape is MUCH slower than disk access. But the integrals should only be needed in constructing the Fock matrix so, depending on your calculation, they are only made a couple of times. You could give it a try to do a setenv ed2 /dev/rmt1.0 (assumed /dev/rmt1.0 is the device name of a self-rewinding tape drive, and assumed the integrals are read one-by-one sequentially) Please mail me if this works, or why it cannot work :-) Greetings, Frits Frits Daalmans OIO Conformational Analysis Gorlaeus Laboratoria Leiden, The Netherlands E-mail: frits@rulglj.leidenuniv.nl Tel: [+31] (0)71-274505 Frits Daalmans OIO Conformational Analysis Gorlaeus Laboratoria Leiden, The Netherlands E-mail: frits@rulglj.leidenuniv.nl Tel: [+31] (0)71-274505 From m10!frisch@uunet.uu.net Fri Sep 30 10:11:22 1994 Received: from relay3.UU.NET for m10!frisch@uunet.uu.net by www.ccl.net (8.6.9/930601.1506) id JAA06959; Fri, 30 Sep 1994 09:33:07 -0400 Received: from uucp2.UU.NET by relay3.UU.NET with SMTP id QQxjpe01633; Fri, 30 Sep 1994 09:33:04 -0400 Message-Id: Received: from m10.UUCP by uucp2.UU.NET with UUCP/RMAIL ; Fri, 30 Sep 1994 09:33:04 -0400 Received: by m10.gaussian.com; Fri, 30 Sep 94 09:20:47 EDT Date: Fri, 30 Sep 94 09:20:47 EDT From: m10!frisch@uunet.uu.net (Michael Frisch) Subject: Re: CCL:large 2-electron files on IBM RS6000; small suggestion To: uunet!ccl.net!chemistry@uunet.uu.net In-Reply-To: uunet!rulglj.LeidenUniv.nl!frits (Frits Daalmans), Fri, 30 Sep 94 10:19:27 GMT (mail to criscuol@gensia.edu bounced, so I send this to the list) To: criscuol@gensia.edu Subject: Re: CCL:Gaussian92 run time problem under AIX 3.2.5 My email directly to criscuo also bounced. The main point in reply to this problem is that Gaussian 92 with SCF=Direct should use NO disk space at all for the integral file and should be compute bound. The problems described are only posible if the user is not in fact doing direct SCF and will go away once this is used. Worrying about how to store big integral files is beside the point; even jobs which produce less than 2 GBytes of integrals will use less CPU as well as less elapsed time with SCF=Direct. In fact, since direct SCF is automatically converted to in-core if the integrals fit, there really isn't any size of calculation with should be run with conventional SCF on an RS/6000. Mike Frisch ------- From smb@smb.chem.niu.edu Fri Sep 30 11:11:24 1994 Received: from mp.cs.niu.edu for smb@smb.chem.niu.edu by www.ccl.net (8.6.9/930601.1506) id KAA08773; Fri, 30 Sep 1994 10:41:23 -0400 Received: from cz2.chem.niu.edu by mp.cs.niu.edu with SMTP id AA15777 (5.67b/IDA-1.5 for <@mp.cs.niu.edu.chem.niu.edu:CHEMISTRY@ccl.net>); Fri, 30 Sep 1994 09:41:26 -0500 Received: from smb.chem.niu.edu by cz2.chem.niu.edu via SMTP (920330.SGI/890607.SGI) (for @mp.cs.niu.edu.chem.niu.edu:CHEMISTRY@ccl.net) id AA07136; Fri, 30 Sep 94 09:21:17 -0500 Received: by smb.chem.niu.edu (920330.SGI/890607.SGI) (for @cz2.chem.niu.edu:CHEMISTRY@ccl.net) id AA14924; Fri, 30 Sep 94 09:21:16 -0500 Date: Fri, 30 Sep 94 09:21:16 -0500 From: smb@smb.chem.niu.edu (Steven Bachrach) Message-Id: <9409301421.AA14924@smb.chem.niu.edu> To: CHEMISTRY@ccl.net Subject: Re: large 2-electron files Mike Frisch recently commented: >In fact, >there really isn't any size of calculation with should be run with >conventional SCF on an RS/6000. Is this really true? If one has a small basis set, say less than 100 functions is the cost of re-evaluating integrals still less than I/O time? Is the same thing true for other platforms (in particular SGI)? Thanks, Steve Steven Bachrach Department of Chemistry Northern Illinois University DeKalb, Il 60115 Phone: (815)753-6863 smb@smb.chem.niu.edu Fax: (815)753-4802 From fh@qc.ag-berlin.mpg.de Fri Sep 30 11:13:21 1994 Received: from oberon.qc.ag-berlin.mpg.de for fh@qc.ag-berlin.mpg.de by www.ccl.net (8.6.9/930601.1506) id LAA09372; Fri, 30 Sep 1994 11:09:46 -0400 Received: by oberon.qc.ag-berlin.mpg.de (AIX 3.2/UCB 5.64/4.03) id AA21537; Fri, 30 Sep 1994 16:08:52 +0100 From: fh@qc.ag-berlin.mpg.de (Frank Haase) Message-Id: <9409301508.AA21537@oberon.qc.ag-berlin.mpg.de> Subject: large 2-electron files To: chemistry@ccl.net (ccl) Date: Fri, 30 Sep 1994 16:08:49 +0100 (NFT) X-Mailer: ELM [version 2.4 PL23] Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 947 Of course, doing conventional SCF is meaningless on fast workstations. But I do not think that the user had his 2GB file problem with an SCF calculation rather with a correlated calculation. For correlated calculations G92 has a serious disadvantage, in my opinion. ALL integral types and perhaps other data will be written into ONE file ( a DA file I guess ). So one hits very fast the 2GB limit (which falls finally in AIX 4.2) and has no chance to manage the job. In TURBOMOLE, for each integral type one file can be specified, and moreover for the largest files like the halftransformed ints a multiple file specification is possible. This allows a more flexible file management and of course the use of several GB. The additional bookholding for this file splitting is very easy to implement. So I do not know the reason for doing all I/O to a single rwf-file in correlated runs in G92. frank haase fh@oberon.qc.ag-berlin.mpg.de From ryszard@msi.com Fri Sep 30 13:16:05 1994 Received: from msi.com for ryszard@msi.com by www.ccl.net (8.6.9/930601.1506) id MAA00934; Fri, 30 Sep 1994 12:54:48 -0400 Received: from aga (aga.msi.com) by msi.com (4.1/SMI-4.1) id AA02085; Fri, 30 Sep 94 12:09:42 EDT Received: by aga; (5.65/1.1.8.2/29Apr94-0859AM) id AA07545; Fri, 30 Sep 1994 12:10:58 -0400 Date: Fri, 30 Sep 1994 12:10:58 -0400 From: Ryszard Czerminski X 217 Message-Id: <9409301610.AA07545@aga> To: chemistry@ccl.net, cornell@cgl.ucsf.edu Subject: Re: CCL:charges First many thanks to Wendy D. Cornell for an excelent summary. The treatment of INTERmolecular interactions by perturbation theory has quite a long history (a lot of reference articles and books). I do not have references at hand but just to give you few names to look for (in no particular order): W. Kolos, B. Jeziorski, L. Piela, P. Claverie, A.D. Buckingham, A. Stone, P. Hobza, M.M. Szczesniak ... Molecular multipole expansion (as opposed to multicenter multipole expansion - atom centered charges are the simplest model of this kind) works properly only in cases when interacting molecules are smaller (and preferably much smaller) then distance between them. For two interacting nucleic acid bases it is a disaster. For calculating electrostatic interactions I have personally used approach developed by Claverie and co-workers -> J. Langlet, P. Claverie et al. Int. J. Quantum Chem., 19 (1981) 299 as well as by Sokalski -> W.A.Sokalski et al. J. Phys. Chem., 87 (1983) 2803 The main idea of this approach (as it was already mentioned earlier) is to go beyond charges (and atoms in Claverie's case) in Mulliken scheme. This is definitely better description of electrostatics then any atom based point charges could provide. The main problem with applying it to large molecular systems (definition of what is large depends which generation of computers are we using and/or how many digits of accuracy are we hoping to achieve) is greater computational complexity (if we include dipoles, quadrupoles etc...) and/or larger number of point charges not longer centered on atoms - which results in much higher CPU costs. Conformational dependence problems would be probably similar as with point charges derived e.g. from electrostatic potential fitting - although I am not aware of any detail study on this subject. From bio- simulations point of view (~ tousands of atoms) we are stack with atomic charges models for a while - probably. One interesting avenue could be to see if there is a way of deriving point charges equivalent to ESP/RESP fitted charges directly from wave function i.e. basically generalizing Thole & van Duijnen approach to higher moments (- Thole & van Duijnen, "A general population analysis preserving the dipole moment", Theor. Chim. Acta 63, 209 (1983).) This is probably not possible exactly but maybe some further progress is possible in this direction. Sincerely, Ryszard Czerminski +------------------------------------------------------------------+ | Molecular Simulations Incorporated | phone : (617)229-8875 x 217 | | 16 New England Executive Park, | fax : (617)229-9899 | | Burlington, MA 01803-5297 | e-mail: ryszard@msi.com | +------------------------------------------------------------------+ From 21NBCJ@NPD.UFPE.BR Fri Sep 30 14:16:05 1994 Received: from npd1.npd.ufpe.br for 21NBCJ@NPD.UFPE.BR by www.ccl.net (8.6.9/930601.1506) id NAA02266; Fri, 30 Sep 1994 13:59:14 -0400 Received: from NPD.UFPE.BR by NPD.UFPE.BR (PMDF V4.2-11 #6339) id <01HHQ6ZOXODC8WX7CT@NPD.UFPE.BR>; Fri, 30 Sep 1994 15:01:43 -0300 Date: Fri, 30 Sep 1994 15:01:43 -0300 From: Nivan Bezerra <21NBCJ@NPD.UFPE.BR> Subject: CI code for atoms using STOs wanted To: chemistry@ccl.net Message-id: <01HHQ6ZOY7O28WX7CT@NPD.UFPE.BR> X-VMS-To: IN%"chemistry@ccl.net" MIME-version: 1.0 Content-type: TEXT/PLAIN; CHARSET=US-ASCII Content-transfer-encoding: 7BIT Dear Netters, I would greatly appreciate receiving information of an availbale CI code for atoms using STOs. Thank you very much, Nivan ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Nivan Bezerra da Costa Junior Fone: (081) 271-8440 Universidade Federal de Pernambuco Fax: (081) 271-8442 Centro de Ciencias Exatas e da natureza Departamento de Quimica Fundamental E-mail: Cidade Universitaria 50739 - Recife - PE INTERNET: nivan@vaxdqf.ufpe.br 21nbcj@npd.ufpe.br ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ From hinsenk@ERE.UMontreal.CA Fri Sep 30 14:17:31 1994 Received: from condor.CC.UMontreal.CA for hinsenk@ERE.UMontreal.CA by www.ccl.net (8.6.9/930601.1506) id NAA01677; Fri, 30 Sep 1994 13:28:46 -0400 Received: from eole.ERE.UMontreal.CA by condor.CC.UMontreal.CA with SMTP id AA15621 (5.65c/IDA-1.4.4 for chemistry@ccl.net); Fri, 30 Sep 1994 13:27:21 -0400 Received: from cyclone.ERE.UMontreal.CA by eole.ERE.UMontreal.CA (940406.SGI/5.17) id AA21319; Fri, 30 Sep 94 13:27:20 -0400 Received: by cyclone.ERE.UMontreal.CA (940406.SGI/5.17) id AA06533; Fri, 30 Sep 94 13:27:19 -0400 Date: Fri, 30 Sep 94 13:27:19 -0400 From: hinsenk@ERE.UMontreal.CA (Hinsen Konrad) Message-Id: <9409301727.AA06533@cyclone.ERE.UMontreal.CA> To: cornell@cgl.ucsf.edu Cc: chemistry@ccl.net In-Reply-To: <199409292323.QAA04058@socrates.ucsf.EDU> (cornell@cgl.ucsf.edu) Subject: Re: CCL:charges Wendy Cornell writes: I would be very interested to see a summary of this approach posted here. It sounds like this is a molecule-based rather than an atom-based description. How many parameters ("true unimolecular properties such as multipoles and polarizabilties") would describe a molecule such as alanine? Can you even treat a molecule that is that big? Isn't this description also subject to conformational dependence? Has this model been tested for its ability to reproduce properties other than gas phase interaction energies, such as conformational energies or condensed phase properties? Although I am rather new to the field of simulating big molecules, I do have a lot of experience with multipole expansions, and I would like to warn about a problem to watch out for: A multipole expansion is an expansion of the potential generated by a spatially localized charge distribution for long distances. Its convergence is guaranteed only outside a sphere containing all the charges. To see if you can use multipole expansions for molecular systems, draw a sphere around every molecule. If any two spheres overlap, you are in trouble. In practice this means that multipole expansions are useful only for approximately spherical molecules, or for molecules in a gas phase. Even if you don't care about the convergence of the whole series ("A dipole is good enough for me."), you can still get some problems. For example, if you have a system of induced dipoles (i.e. molecules with a dipole polarizability) and place them too close to each other in some external field, then the strength of the induced dipoles will diverge. The reason is that such a system is unphysical: you can't have arbitrarily polarizable materials, so to increase the polarizability of a real object, you have to increase its size. But that puts a lower limit on the distance between two such objects. You can actually calculate the limit in this way: take a sphere of radius r made out of the most polarizable material (a perfect conductor) and calculate its dipole polarizability (which is r^3 if I remember correctly). That's the highest polarizability allowed for a distance of 2r. Finally, if you find that you can work with a multipole expansion, don't assume that a dipole is sufficient. It might be, but there is no way to know without trying with higher multipole orders. So make sure your code can handle them. And take care to pick a suitable representation. Most people use multipole moments defined in terms of spherical harmonics, because that's what you find in textbooks on electromagnetic theory. There are however good reasons to choose other representations for numerical work. If you can live with redundant elements, the Cartesian multipole expansion (essentially a Taylor series) is very simple to use. If you can't, have a look at irreducible Cartesian tensors, which are equivalent to spherical harmonics, but can be calculated without having to deal with complex numbers and sines and cosines. ------------------------------------------------------------------------------- Konrad Hinsen | E-Mail: hinsenk@ere.umontreal.ca Departement de Chimie | Tel.: +1-514-343-6111 ext. 3953 Universite de Montreal | Fax: +1-514-343-7586 C.P. 6128, succ. A | Deutsch/Esperanto/English/Nederlands/ Montreal (QC) H3C 3J7 | Francais (phase experimentale) ------------------------------------------------------------------------------- From becky@msi.com Fri Sep 30 15:16:08 1994 Received: from msi.com for becky@msi.com by www.ccl.net (8.6.9/930601.1506) id OAA03079; Fri, 30 Sep 1994 14:38:17 -0400 Received: from franz.MSI.COM by msi.com (4.1/SMI-4.1) id AA02548; Fri, 30 Sep 94 14:38:50 EDT Received: by franz.MSI.COM (AIX 3.2/UCB 5.64/4.03) id AA21582; Fri, 30 Sep 1994 14:38:47 -0400 Date: Fri, 30 Sep 1994 14:38:47 -0400 From: becky@msi.com (Becky Rone X 276) Message-Id: <9409301838.AA21582@franz.MSI.COM> To: chemistry@ccl.net Subject: CCL:charges I would like to respond to this question: > Is there any reference on charge assignment method for CHARMM, > and/or CHARMm? > Xinjun J. Hou hou@agouron.com The CHARMm references are: F.A. Momany, R. Rone, and L. Schafer, "Geometry Optimization, Energetics, and Solvation Studies on Six-membered Cyclic Peptides, using QUANTA3.3/CHARMm22", Peptides, Proceedings of the Thirteenth American Peptide Symposium, (1993). F.A. Momany, R. Rone, H. Kunz, R. F. Frey, S. Q. Newton, and L. Schafer, "Geometry Optimization, Energetics, and Solvation Studies on Four and Five Membered Cyclic and Disulfide Bridged Peptides, using the Programs QUANTA3.3/CHARMm22", J. Mol. Struct. (Theochem), 286, 1-18 (1993). F.A. Momany and R. Rone, "Validation of the General Purpose QUANTA 3.2/CHARMm Force Field," J. Comp. Chem. 13 (7), 888-900 (1992). F.A. Momany, R. Rone, R.F. Frey, and L. Schafer, "On the Use of Correlation Ab Initio Studies in the Development of the CHARMm Empirical Molecular Force Field," Chemical Design Automation News (7), 1, 38-41(July, 1992). To those of you who have QUANTA please note that the last reference is reprinted as the QUANTA Parameter Handbook which is shipped out with your tapes. Reprints for these papers are available from the support hotline at MSI (617) 229-9800 for all MSI customers. For other scientists, if you need reprints contact my secretary Susan Rostoff at the same number. Cheers, Rebecca Rone Senior Scientist Molecular Simulations, Inc. From sling@euclid.chem.washington.edu Fri Sep 30 16:16:07 1994 Received: from euclid.chem.washington.edu for sling@euclid.chem.washington.edu by www.ccl.net (8.6.9/930601.1506) id PAA03882; Fri, 30 Sep 1994 15:17:43 -0400 Received: by euclid.chem.washington.edu (AIX 3.2/UCB 5.64/4.03) id AA17577; Fri, 30 Sep 1994 12:16:43 -0700 Date: Fri, 30 Sep 1994 12:16:43 -0700 From: sling@euclid.chem.washington.edu (Song Ling) Message-Id: <9409301916.AA17577@euclid.chem.washington.edu> To: CHEMISTRY@ccl.net Subject: semiclassical tunneling Sorry but I lost the address on the subject, a more recent paper along the lines of Miller, Makri, and so on is as follows: QIN Y; THOMPSON DL. SEMICLASSICAL TREATMENT OF TUNNELING EFFECTS IN HONO CIS-TRANS ISOMERIZATION. JOURNAL OF CHEMICAL PHYSICS, 1994 MAY 1, V100 N9:6445-6457. Miller's address: miller@neon.cchem.berkeley.edu From ross@cgl.ucsf.EDU Fri Sep 30 16:20:20 1994 Received: from socrates.ucsf.EDU for ross@cgl.ucsf.EDU by www.ccl.net (8.6.9/930601.1506) id PAA04742; Fri, 30 Sep 1994 15:53:08 -0400 Received: (from ross@localhost) by socrates.ucsf.EDU (8.6.9/GSC4.24) id MAA04450 for chemistry@ccl.net; Fri, 30 Sep 1994 12:53:06 -0700 Date: Fri, 30 Sep 1994 12:53:06 -0700 Message-Id: <199409301953.MAA04450@socrates.ucsf.EDU> From: ross@cgl.ucsf.edu (Bill Ross ) To: chemistry@ccl.net Subject: multipole expansion From: hinsenk@ERE.UMontreal.CA (Hinsen Konrad) ... A multipole expansion is an expansion of the potential generated by a spatially localized charge distribution for long distances. Its convergence is guaranteed only outside a sphere containing all the charges. To see if you can use multipole expansions for molecular systems, draw a sphere around every molecule. If any two spheres overlap, you are in trouble. In practice this means that multipole expansions are useful only for approximately spherical molecules, or for molecules in a gas phase. ... Another use of multipole expansions is for long-range forces exerted by collections of point charges (typically in a subcube of the simulated volume) as a means to faster calculations in vacuum or solvated systems. There is an excellent brief summary of these methods by Leslie Greengard in a recent _Science_ issue featuring computers in science. Bill Ross From hinsenk@ERE.UMontreal.CA Fri Sep 30 18:16:08 1994 Received: from condor.CC.UMontreal.CA for hinsenk@ERE.UMontreal.CA by www.ccl.net (8.6.9/930601.1506) id RAA06443; Fri, 30 Sep 1994 17:34:53 -0400 Received: from eole.ERE.UMontreal.CA by condor.CC.UMontreal.CA with SMTP id AA23259 (5.65c/IDA-1.4.4 for chemistry@ccl.net); Fri, 30 Sep 1994 17:09:47 -0400 Received: from cyclone.ERE.UMontreal.CA by eole.ERE.UMontreal.CA (940406.SGI/5.17) id AA27325; Fri, 30 Sep 94 17:09:46 -0400 Received: by cyclone.ERE.UMontreal.CA (940406.SGI/5.17) id AA10449; Fri, 30 Sep 94 17:09:46 -0400 Date: Fri, 30 Sep 94 17:09:46 -0400 From: hinsenk@ERE.UMontreal.CA (Hinsen Konrad) Message-Id: <9409302109.AA10449@cyclone.ERE.UMontreal.CA> To: ryszard@msi.com Cc: chemistry@ccl.net In-Reply-To: <9409301610.AA07545@aga> (message from Ryszard Czerminski X 217 on Fri, 30 Sep 1994 12:10:58 -0400) Subject: Re: CCL:charges The treatment of INTERmolecular interactions by perturbation theory Actually I wouldn't call multipole expansions perturbation theory. There is nowhere any assumption about some parameter being small. It is simply an expansion in terms of the relation between the radius of the charged system and the distance to the point where the potential is to be calculated. Molecular multipole expansion (as opposed to multicenter multipole expansion - atom centered charges are the simplest model of this kind) works properly only in cases when interacting molecules are smaller (and preferably much smaller) then distance between them. This is not exactly true - what you have mentioned is the condition for fast convergence. You can even treat two touching spheres with a multipole expansion, but you might need fifty terms for that. Whether or not that is feasible depends on the application. ------------------------------------------------------------------------------- Konrad Hinsen | E-Mail: hinsenk@ere.umontreal.ca Departement de Chimie | Tel.: +1-514-343-6111 ext. 3953 Universite de Montreal | Fax: +1-514-343-7586 C.P. 6128, succ. A | Deutsch/Esperanto/English/Nederlands/ Montreal (QC) H3C 3J7 | Francais (phase experimentale) ------------------------------------------------------------------------------- From leboeuf@CHIMCN.UMontreal.CA Fri Sep 30 19:16:10 1994 Received: from condor.CC.UMontreal.CA for leboeuf@CHIMCN.UMontreal.CA by www.ccl.net (8.6.9/930601.1506) id SAA07855; Fri, 30 Sep 1994 18:30:49 -0400 Received: from chims1.CHIMCN.UMontreal.CA by condor.CC.UMontreal.CA with SMTP id AA25724 (5.65c/IDA-1.4.4 for chemistry@ccl.net); Fri, 30 Sep 1994 18:29:46 -0400 Received: by chims1.CHIMCN.UMontreal.CA (930416.SGI/5.17) id AA28442; Fri, 30 Sep 94 18:28:33 -0400 From: leboeuf@CHIMCN.UMontreal.CA (Leboeuf Martin) Message-Id: <9409302228.AA28442@chims1.CHIMCN.UMontreal.CA> Subject: Re: multipole expansion To: chemistry@ccl.net Date: Fri, 30 Sep 1994 18:28:33 -0400 (EDT) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 1456 > From: hinsenk@ERE.UMontreal.CA (Hinsen Konrad) > > ... > A multipole expansion is an expansion of the potential generated > by a spatially localized charge distribution for long distances. > Its convergence is guaranteed only outside a sphere containing > all the charges. To see if you can use multipole expansions > for molecular systems, draw a sphere around every molecule. > If any two spheres overlap, you are in trouble. In practice > this means that multipole expansions are useful only for > approximately spherical molecules, or for molecules in a > gas phase. > ... In other words, the multipole expansion (ME) is the solution of Laplace equation, which is why you have to be outside of the sphere containing the charge distribution. But recently, Koester et al. (JCP, 99 p1224 (1993)) developped a model for the solution of the Poisson equation. Since you are now solving Poisson's equation, there is no longer the restriction of being outside of your charge distribution. And in fact, we have implemented this scheme in our DFT code (deMon) and the molecular electrostatic potential calculated with this model has the right behavior both at long AND short distances from the molecules. And the computational effort is very similar to a ME calculation. Martin Leboeuf Departement de Chimie Universite de Montreal Montreal, Canada email: leboeuf@cerca.umontreal.ca From mf12101@sc.msc.edu Fri Sep 30 21:16:11 1994 Received: from noc.msc.edu for mf12101@sc.msc.edu by www.ccl.net (8.6.9/930601.1506) id UAA08752; Fri, 30 Sep 1994 20:26:16 -0400 From: Received: from sc.msc.edu by noc.msc.edu (5.65/MSC/v3.0.1(920324)) id AA01859; Fri, 30 Sep 94 19:24:57 -0500 Received: by sc.msc.edu (5.61/MSC/v3.0s(901129)) id AA412488; Fri, 30 Sep 94 19:24:55 -0500 Message-Id: <9410010024.AA412488@sc.msc.edu> Subject: partial charges To: chemistry@ccl.net (Comp. Chem. list) Date: Fri, 30 Sep 94 19:24:54 CDT Cc: cramer@chemsun.chem.umn.edu (Christopher J. Cramer), giesen@chemsun.chem.umn.edu (Dave Giesen), mf12141@sc.msc.edu (Greg Hawkins), eeaston@prometheus.chem.umn.edu (Evan Easton), chambers@t1.chem.umn.edu (Candee C. Chambers), guxx0004@gold.tc.umn.edu (Michael Gu) X-Mailer: ELM [version 2.3 PL11] I would like to contribute some thoughts on the subject of partial charges of atoms in molecules, which has received some attention on the CCL this week. In particular I note that Ryszard Czerminski wrote: "One interesting avenue could be to see if there is a way of deriving point charges equivalent to ESP/RESP fitted charges directly from a wave function.... This is probably not possible exactly but maybe some further progress is possible in this direction." In the following I will comment directly on recent progress in this area. Partial charges are an important theoretical subject because they are so useful in molecular modeling, and they are a fascinating subject because, since they are not physical observables, their definition is--to some extent-- at our disposal. In recent work [J. W. Storer, D. J. Giesen, C. J. Cramer, and D. G. Truhlar, "Class IV Charge Models: A New Semiempirical Approach in Quantum Chemistry," Journal of Computer-Aided Molecular Design, in press], we have distinguished four classes of models of partial charges. Class I charges are defined by simple models that make no reference to the quantum mechanical character of electronic structure. An example would be obtaining the partial charges of a heteronuclear diatomic molecule by dividing the dipole moment by the bond length. Class II charges are obtained by some prescription for partitioning the density distribution corresponding to some approximate electronic wave function (or the exact wave function if we had it) of a molecule into components associated with individual atoms. Examples include Mulliken or Lowdin population analysis or Richard Bader's method of partitioning the charge distribution. Class III partial charges are based on a more utilitarian approach. As stated above, partial charges of atoms in molecules are interesting first of all because they are useful for molecular modeling. Now there are several approaches to modeling the electrostatic properties of molecules based on placing a set of point multipoles at each of several sites in a molecule. One example encountered in practice is to place partial charges both at the nuclei and in the lone pair regions; this approach is often used for water. Another example is placing a partial charge and a point dipole at each nuclear center. The latter is an example of a distributed multipole representation. By far the most common example of this approach is the nuclear-centered distributed-monopole representation in which we ignore the higher multipoles (dipole, quadrupole, octupole, hexadecapole, ...) and simply place a partial charge at each nuclear center. The simplicity of this approach makes it useful for inclusion in force fields designed for conformational analysis, interaction potentials, solvation modeling, and molecular dynamics. Class III partial charges are an attempt to find the best set of nuclear-centered partial charges for such modeling efforts. Thus class III charges are defined such that physical observables calculated from such partial charges agree as well as possible (a subjective elements creeps in here) with the same physical observable calculated using the continuous psi squared charge density of an electronic wave function. The most frequently used physical observable is the electrostatic potential (ESP) at selected points around a molecule; a special case of this would be fitting the dipole moment, which is equivalent to fitting the electrostatic potential of a polar molecule on a hypersphere of very large radius. Examples of partial charge methods based on ESP-fitting are the ChelpG method of Michelle Francl, Ken Wiberg, and coworkers and the similar fitting procedure of Kenny Merz and Peter Kollman. Class III partial charges, like any other modeling tool, have some deficiencies. The first type of deficiency is numerical. Francl herself has pointed out that the equations one obtains in ESP fitting are often ill- conditioned. In a similar vein, Bill Jorgensen has pointed out that the charges on buried atoms may be especially poorly determined by ESP fitting. A second problem with class III charges is that, while they make up (as well as possible) for deficiencies in the replacement of a continuous electron density function (corresponding to some electronic structure level X and basis Y) by a set of nuclear-centered partial charges, they do not make up for the deviation of psi squared at level X/Y from the exact psi squared. Such deficiencies can in fact be quite serious even for popular levels X/Y that are considered to be high levels. [For example, for MeSO_3H, the HF/6-31G* dipole moment is 3.24 D, whereas the more accurate MP2/cc-pVDZ dipole moment is 2.33 D.] Despite these deficiencies, ESP fitting is a powerful technique, and it is often very useful, but Class IV charges represent an attempt to make up for both sets of deficiencies. Class IV charges are obtained by starting with class II charges and mapping them to a new set of charges (the class IV charges) with mapping parameters determined semiempirically such that the new charges reproduce experimental observables as well as possible. Although ESPs are in principle observable, dipole moments are more widely available and have been used for developing class IV charge models so far. We have so far (in the preprint mentioned above) parameterized two class IV charge models, which we call CM1A and CM1P. The former begins the map with AM1 Mulliken charges, and the latter begins with PM3 Mulliken charges. The parameters in our mappings were based on 204 neutral compounds containing a wide variety of functional groups. Mapping parameters are available for the following atom types: H, C, N, O. F, Si, S, Cl, Br, and I. (Note the conspicuous absence of P. We believe that a good map for P should start with an ab initio wave function, or at least with something better than AM1 or PM3. Work is "in progress" on this.) For 23 compounds we tested the partial charges against those obtained by ChelpG analysis of MP2/6-3/G* wave functions. This chart gives the RMS errors in various calculated dipole moments for these 23 compounds. The first row is based on a continuous charge distribution; the others are based on partial charges: dipole calculated from RMS error (D) ________________________ ____________ MP2/6-31G*, psi squared 0.21 HF/6-31G* ChelpG charges 0.33 HF/6-31G* Mulliken charges 0.93 AM1 Mulliken charges 0.89 PM3 Mulliken charges 1.00 AM1-CM1A class IV charges 0.27 PM3-CM1P class IV charges 0.20 The cost of the mapping is totally negligible; thus with PM3-CM1P class IV charges one obtains MP2/6.31G* accuracy with NDDO cost. If you would like a preprint of our paper, send e-mail to truhlar@t1.chem.umn.edu with your full mailing address and request UMSI Research Report 94/144 by Storer et al. CM1A and CM1P partial charges may be calculated for gas-phase molecules using AMSOL-version 4.5 [Reference: C. J. Cramer, G. D. Hawkins, G. C. Lynch, D. J. Giesen, D. G. Truhlar, and D. A. Liotard, QCPE Bull. 14, 55- 57 (1994)]. This program is available from QCPE at Indiana University--it is program 606 in their catalog. Don Truhlar Dept. of Chemistry University of Minnesota