From dyang@acs.ucalgary.ca Thu Aug 4 03:29:31 1994 Received: from acs7.acs.ucalgary.ca for dyang@acs.ucalgary.ca by www.ccl.net (8.6.9/930601.1506) id CAA05856; Thu, 4 Aug 1994 02:32:29 -0400 Received: from acs2.acs.ucalgary.ca by acs7.acs.ucalgary.ca (AIX 3.2/UCB 5.64/4.03) id AA66066; Thu, 4 Aug 1994 00:41:22 -0600 Received: by acs2.acs.ucalgary.ca (AIX 3.2/UCB 5.64/4.03) id AA17730; Thu, 4 Aug 1994 00:41:22 -0600 Message-Id: <9408040641.AA17730@acs2.acs.ucalgary.ca> Subject: any software for corrosion inhibitors To: chemistry@ccl.net Date: Thu, 4 Aug 94 0:41:21 MDT From: "Danya Yang" X-Mailer: ELM [version 2.3 PL11y] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 646 Dear Netters, I am planning to do some modeling work on corrosion inhibitors (organic compounds) by studying adsorption behavior of a molecular layer at the metal-solution interface. I am looking for a visualization software which can show how those organic molecules pack together (to form a molecular layer) on a metal surface, and how the molecular layer interacts with the metal surface, or a software which models adsorption phenomenon ( physical or chemical ). I appreciate any suggestions from you regarding to the softwares or published papers in this field. I'll summarize the replies if there is enough interest. Danya Yang From suter@physik.unizh.ch Thu Aug 4 03:40:01 1994 Received: from milliways.physik.unizh.ch for suter@physik.unizh.ch by www.ccl.net (8.6.9/930601.1506) id DAA05912; Thu, 4 Aug 1994 03:05:34 -0400 Received: by milliways.physik.unizh.ch (5.0cf2.19) id AA19694; Thu, 4 Aug 94 09:10:59 +0200 From: Hans-Ueli Suter Message-Id: <9408040710.AA19694@milliways.physik.unizh.ch> Subject: Answers to AO based DFT 600 Lines To: CHEMISTRY@ccl.net Date: Thu, 4 Aug 1994 09:10:57 +0200 (MET DST) Cc: suter@ccl.net (Hans-Ueli Suter) X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 23412 Dear Netters, two weeks ago (12.Juli 1994) I have asked a general question about the current use of AO based DFT methods. May aim was to get insigth in the feelings and experiences by the use of these programs, something which would naturally not be part of scientific literature. I thougth this was interesting, also because my experiences, coming from ab-initio, have not been to positive (all: speed, accuracy and stability of the programs). At that time I have used DMol. Not as many have answered the questions, so I am even more grateful for those who have done it. The list of programs include now DMol, G92/DFT and CADPAC, DeFT, DeMon, ADF and DGauss, which I have forgotten. There may be added the "historic" programs of Dunlap and Jones the one of Clementi (KGNMOL) and the newer of him, for which I have no information. Regarding that DFT has also been included in the Turbomol (C.v.Wuellen, Chem.Phys.Lett. 219 (1994) 8), the list will still not be complete. Replies have been received for DeFT, deMon, ADF, DGauss and DMol. All replies were positive. The advantage of DeMon may be the inclusion of NMR-Properties and the one of DGauss its speed, but figure out that on the following list. I also have been pointed to a very interesting review article of Ziegler (Chem.Rev. 91 (1991) 651), which was a pleasure to read. Thanks again to all who have taken time to answer my questions, Hans Ulrich Suter, Centro Svizzero di Calcolo Scientifico, CH-6928 Manno, Switzerland. For point1 : Two programs should been added to the list this is the DGauss-Code and the CADPAC program. ------------------------------------------------------------------------------ >From mckelvey@Kodak.COM Tue Jul 12 17:35 MET 1994 Received: from Kodak.COM by milliways.physik.unizh.ch (5.0cf2.19) id AA29047; Tue, 12 Jul 94 17:35:04 +0200 Received: from NTPROF.DECnet MAIL11D_V3 by Kodak.COM (5.61+/2.1-Eastman Kodak) id AA00299; Tue, 12 Jul 94 11:37:28 -0400 Date: Tue, 12 Jul 94 11:37:25 -0400 Reply-To: mckelvey@Kodak.COM Message-Id: <9407121537.AA00299@Kodak.COM> From: mckelvey@Kodak.COM To: "suter@physik.unizh.ch"@Kodak.COM Subject: RE: CCL:AO based DFT codes Content-Type: text Content-Length: 81 Status: RO Missing a VERY fast code, DGauss, for Cray and SGI. John McKelvey Eastman Kodak >From apisan@redvax1.dgsca.unam.mx Tue Jul 12 17:54 MET 1994 Received: from redvax1.dgsca.unam.mx by milliways.physik.unizh.ch (5.0cf2.19) id AA00633; Tue, 12 Jul 94 17:53:48 +0200 Received: from standard-input by redvax1.dgsca.unam.mx; Tue, 12 Jul 94 09:43:48 CST Sender: apisan@redvax1.dgsca.unam.mx Date: Tue, 12 Jul 1994 09:38:53 -0600 (CST) From: Pisanty Baruch Alejandro-FQ Subject: Re: CCL:AO based DFT codes To: Hans-Ueli Suter In-Reply-To: <9407121306.AA09273@milliways.physik.unizh.ch> Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Content-Length: 624 Status: RO Lieber Hans-Ueli, shouldn't you include DGauss too? I am interested in a summary - not that much on what is in the codes, more on how users now feel about them. Ihr Alejandro . . . . . . . . . . . . . . . . . . . . . . . . . . Dr. Alejandro Pisanty, Secretary of the Advisory Council on Computing, UNAM, and Head of the Graduate Division, Faculty of Chemistry, UNAM Universidad Nacional Autonoma de Mexico (UNAM) Ciudad Universitaria 04510 Mexico DF MEXICO Tel. (+52-5) 622 4181, 616 1649; Fax 550 0904, 616 2010 . . . . . . . . . . . . . . . . . . . . . . . . . . >From nelson@iqm.unicamp.br Tue Jul 12 18:52 MET 1994 Received: from styx.iqm.unicamp.br (iqm.unicamp.br) by milliways.physik.unizh.ch (5.0cf2.19) id AA05987; Tue, 12 Jul 94 18:50:42 +0200 Received: (nelson@localhost) by styx.iqm.unicamp.br (8.6.8/iqm-ansp-1.4) id NAA09392; Tue, 12 Jul 1994 13:36:57 -0300 From: Nelson Henrique Morgon Message-Id: <199407121636.NAA09392@styx.iqm.unicamp.br> Subject: DGauss To: suter@physik.unizh.ch Date: Tue, 12 Jul 1994 13:36:56 -0300 (EST) X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 1722 Status: RO To Hans-Ueli Suter suter@physik.unizh.ch You forgot DGauss (UNICHEM) I use DeFT, it is very simple to use. I don't experience with other programs. _________________________________________________________________ | | | NELSON HENRIQUE MORGON #=============# | | INSTITUTE OF CHEMISTRY #=|-----------# | | | DEPARTMENT OF PHYSICAL CHEMISTRY #=|-|---------# | | | | STATE UNIVERSITY OF CAMPINAS (UNICAMP) | | | o-----o | | | | | POSTAL BOX 6154 - ZIP CODE 13083-970 | | | | NHM | | | | | | CAMPINAS - SP - BRAZIL. | | | o-----\ | | | | | PHONE : (0192) 39-3805 (Ramal: 8640) | | #=============# | | FAX: 0055-192-393805 | #=============# | | #=============# | | Email: NELSON@IQM.UNICAMP.BR | | QUANTUM CHEMISTRY | |_________________________________________________________________| For the DeMon program: ------------------------------------------------------------------------- >From fournie@ned1.sims.nrc.ca Tue Jul 12 18:20 MET 1994 Received: from ned1.sims.nrc.ca by milliways.physik.unizh.ch (5.0cf2.19) id AA03056; Tue, 12 Jul 94 18:20:52 +0200 Message-Id: <9407121620.AA16364@ned1.sims.nrc.ca> From: fournie@ned1.sims.nrc.ca (Dr. Rene Fournier) Subject: DFT programs: add CADPAC & DGauss To: suter@physik.unizh.ch Date: Tue, 12 Jul 94 12:20:46 EDT X-Mailer: ELM [version 2.3 PL11] Content-Type: text Content-Length: 919 Status: RO Dear Hans; Maybe you can add CADPAC to the list. You may enquire about it by e-mail to Prof. N. Handy, but I'm not sure of the e-mail adress, I think it is: N. C. Handy : nch1@phx.cam.ac.uk You should certainly include the Cray software "DGauss" in that list. If I remember correctly, DGauss is part of the Unichem package but it can also be used as a stand-alone program. It can run on Silicon Graphics in addition to Cray computers. I am using deMon exclusively, so I can not compare with other programs. deMon is "reasonably" user-friendly and stable: these two aspects are not a concern in my day-to-day work. I can not comment on the speed of the program as I did not benchmarked it or compared to other programs in any way. Cheers, Rene Fournier fournie@ned1.sims.nrc.ca Steacie Institute for Molecular Sciences, NRC, Ottawa. (613) 990 0964 >From oles@kjemi.uio.no Tue Jul 12 21:25 MET 1994 Received: from sunic.sunet.se by milliways.physik.unizh.ch (5.0cf2.19) id AA20512; Tue, 12 Jul 94 21:25:30 +0200 Received: from pat.uio.no by sunic.sunet.se (8.6.8/2.03) id VAA16960; Tue, 12 Jul 1994 21:25:23 +0200 Received: from ulrik.uio.no by pat.uio.no with local-SMTP (PP) id <22349-0@pat.uio.no>; Tue, 12 Jul 1994 21:23:26 +0200 Received: from kelvin by ulrik.uio.no ; Tue, 12 Jul 1994 21:23:07 +0200 Received: by kelvin ; Tue, 12 Jul 1994 21:20:36 +0200 From: Date: Tue, 12 Jul 1994 21:20:36 +0200 Message-Id: <9407121920.AA02548@kelvin> To: suter@physik.unizh.ch In-Reply-To: Hans-Ueli Suter's message of Tue, 12 Jul 1994 15:05:58 +0200 (MET DST) <9407121306.AA09273@milliways.physik.unizh.ch> Subject: Re: CCL:AO based DFT codes Content-Type: text Content-Length: 1631 Status: RO Of the programs you mention, I have only been using ADF and deMon, so all I can do is to compare the two. Very crudely speaking, from my experience (only about one year) ADF has the most features and the best convergence, while deMon has the easiest input format. In more detail: ADF: More features. Most symmetry groups implemented, including non-Abelian groups. Well documented. Very reasonable rates for non-profit use, source code is included. Reasonably well tested but: Somewhat clumsy input for geometry optimizations. Unnecessary large output, with a lot of redundance and (in my opinion) unsufficient flexibility in user's choice of output. deMon: Very easy input. I like the output, too. but: Expensive, even for non-profit use, and no source code is included No symmetry implemented We've had convergence difficulties that the vendor could not sort out for us. We solved the problem by switching to ADF. Both programs have analytical gradients for optimization, and run on several different makes of hardware (ADF on more than deMon). Both come with basis set libraries, and both can handle relativistic effects in a scalar fashion. I have not conducted any detailed speed comparisons, but there are no big difference between the two programs. For systems with high symmetry, ADF is much faster because it can exploit the symmetry. Again, I must stress that these are my personal opinions only. A colleague says very nice things about Dmol, a program that I have never tried. Best regards, Ole Swang >From AZHARI@FRCU.EUN.EG Wed Jul 13 12:58 MET 1994 Received: from FRCU.EUN.EG by milliways.physik.unizh.ch (5.0cf2.19) id AA11062; Wed, 13 Jul 94 12:58:21 +0200 Received: from FRCU.EUN.EG by FRCU.EUN.EG (PMDF V4.2-11 #3805) id <01HENRWL771S000VGQ@FRCU.EUN.EG>; Wed, 13 Jul 1994 13:58:37 O Date: Wed, 13 Jul 1994 13:58:37 +0000 (O) From: AZHARI@FRCU.EUN.EG Subject: AO based DFT codes To: suter@physik.unizh.ch Message-Id: <01HENRWL9VHU000VGQ@FRCU.EUN.EG> X-Vms-To: IN%"suter@physik.unizh.ch" Mime-Version: 1.0 Content-Transfer-Encoding: 7BIT Content-Type: TEXT/PLAIN; CHARSET=US-ASCII Content-Length: 943 Status: RO Dear Suter: 1) I have the DeFT code but I did not use it yet. 2) From its manual and sample input and output it has many disadvantages. a) It dose not use the standard basis sets to make a comparison with the calculations at the SCF and MP2 levels. b) It dose not print the force fields and calculation of vibrational frequencies of isotopmers will require a separate job! which is definitaly a big disadvantage. c) It dose not allow to use the symmetry explicitly. 3) I think the code from cadpac would be much better. I think the paper of amos JPC, 1993 97, 4392 ( I hope I got the page number correct) will explain the advantages an the facilities of the DFT code implemented in the CADPAC program. We are tring to get the new version of the CADPAC program for this reason. Adel El-Azhary Cairo University >From CHERM@frcpn11.in2p3.fr Wed Jul 13 17:00 MET 1994 Received: from frcpn11.in2p3.fr by milliways.physik.unizh.ch (5.0cf2.19) id AA24145; Wed, 13 Jul 94 17:00:06 +0200 Received: from FRCPN11.IN2P3.FR by frcpn11.in2p3.fr (IBM VM SMTP V2R2) with BSMTP id 1170; Wed, 13 Jul 94 17:00:12 SET Received: from FRCPN11 (CHERM) by FRCPN11.IN2P3.FR (Mailer R2.08 R208004) with BSMTP id 0152; Wed, 13 Jul 94 17:00:12 SET Message-Id: <9407131500.AA24145@milliways.physik.unizh.ch> Date: Wed, 13 Jul 94 16:30:19 SET From: Henry Chermette Subject: Re: CCL:AO based DFT codes To: Hans-Ueli Suter In-Reply-To: Your message of Tue, 12 Jul 1994 15:05:58 +0200 (MET DST) X-Acknowledge-To: Content-Type: text Content-Length: 1631 Status: RO Hi, I am using presently deMon and ADF, and have collaborations with people using G92/DFT. as you know, DGauss is missing from your list, as well as CADPAC (Handy) and others codes coming from ab_initio programs, e.g. CRS4 (Clementi). Also is missing the LCGTO-MP-(N)LSD developped by Dunlap, Rosch, et al., which uses Hermite-gaussians, and which has common ancestors with DGauss, deMon and deFT. ADF is robust, available to tackle any atoms in the periodic table, it includes relativistic corrections, etc. it is fast, makes a lot of I/O, and can be applied to periodic systems 1D, 2D (slabs), 3D. deMon uses cartesian gaussians, as in DGauss, deFT, with auxiliary basis sets for fitting the density and the xc pot. it may contains many extensions for calc. of properties. ETC. (I have no time to summarize all diff. and simil. of these codes, this has published in part, e.g. in D.R.Salahub, in Metal-ligand interactions, from atoms to clusters, to surfaces, Kluwer, 1992, D.R. Salahub, N. Russo, ed. D.R. Salahub, in Relativistic and Electron correlation effects in molecules and solids, Plenum press, 1994, G.L. Malli, ed (unfortunatly the table comparing the programs has been de-tabulated) T. Ziegler, Chem. Review, 1991, 91, 651 * a comparison of DMol, DGauss, deMon applied to triazene systems will appear soon in J. Comp. Chem. (A. Schmiedekamp et al.) * other comp. have also been published, e.g. Dixon (I have not the ref. here) Henry Chermette Institut de Physique Nucleaire de Lyon 43 bd du 11 novembre 1918 F-69622 VILLEURBANNE Cedex Tel (33) 72 44 84 27 Fax (33) 72 44 80 04 e-mail CHERM@FRCPN11.in2p3.fr GENERAL >From proynov@CHIMCN.UMontreal.CA Tue Jul 12 21:36 MET 1994 Received: from condor.CC.UMontreal.CA by milliways.physik.unizh.ch (5.0cf2.19) id AA21590; Tue, 12 Jul 94 21:36:34 +0200 Received: from chims1.CHIMCN.UMontreal.CA by condor.CC.UMontreal.CA with SMTP id AA12703 (5.65c/IDA-1.4.4 for suter@physik.unizh.ch); Tue, 12 Jul 1994 15:13:28 -0400 Received: by chims1.CHIMCN.UMontreal.CA (930416.SGI/5.17) id AA23816; Tue, 12 Jul 94 15:12:45 -0400 From: proynov@CHIMCN.UMontreal.CA (Proynov Emil) Message-Id: <9407121912.AA23816@chims1.CHIMCN.UMontreal.CA> Subject: deMon news To: suter@physik.unizh.ch Date: Tue, 12 Jul 1994 15:12:45 -0400 (EDT) Mime-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=US-ASCII Content-Length: 2109 Status: RO Dear Dr. Suter, We understood from the network that you are interested in DFT programs for molecular calculations. Such programs are currently developed and used in our group under the supervision of prof. Dennis Salahub. At the moment we have several working versions of the so called package "deMon", based on the LCGTO-MO-DFT program written originaly by Dr. Alain St. Amant a couple of years ago. Let us now give gust a short gist of the options currently available in the code and if you are interested in more details you please do not hesitate to contact with us again. The code deMon basically performs a calculation of the electronic structure of molecules via SCF solution of the Kohn-Sham DFT equations. What one receives at the output is the total energy of the system (including the electron correlation) from which you can estimate the binding energy; the Kohn-Sham MO energies (which actually have the same chemical meaning as in the conventional MO-LCAO approach); the optimized molecular geometry (several different optimization procedures are currently included); optional detail vibrational analysis and evaluation of the IR spectral characteristics; complete Mullicken population analysis; evaluation of the dipol moments, polarizabilities and hyper-polarizabilities. You can try to maximize the quality of the calculations by testing several different options for the local and nonlocal exchange-correlation functionals, which are the soil of the method. Besides the well established already in the field options: LSD-Vosko-Wilk-Nossair, gradient-corrected schemes (GGA of Perdew and Becke), soon highly efficient new functionals will be available in the new release of deMon, developed in our group. Another interesting option you may use is the NMR appendix providing a detail calculation of NMR and EPR spectroscopic characteristics. Looking forward to hearing from you again, Best regards from Montreal, Mark Casida, Andreas Koster, Emil Proynov >From gene@jersey.cray.com Tue Jul 12 18:57 MET 1994 Received: from cray.com (timbuk.cray.com) by milliways.physik.unizh.ch (5.0cf2.19) id AA06608; Tue, 12 Jul 94 18:57:01 +0200 Received: from jersey.cray.com by cray.com (Bob mailer 1.3) id AA04977; Tue, 12 Jul 94 11:56:18 CDT Received: by jersey.cray.com id AA21741; 5.0/CRI-5.6; Tue, 12 Jul 94 12:54:58 EDT Date: Tue, 12 Jul 94 12:54:58 EDT From: gene@jersey.cray.com (Eugene Fleischmann) Message-Id: <9407121654.AA21741@jersey.cray.com> To: suter@physik.unizh.ch Cc: chemistry@ccl.net, gene@jersey.cray.com, george@gravity.cray.com, raeuchle@gravity.cray.com Subject: Re: CCL:AO based DFT codes Content-Type: text Content-Length: 3622 Status: RO > Greetings, Prof. Suter. With regard to your inquiry: > (i) Is this list complete, which program should be addeed? The only addition to your list is: DGauss, from Cray Research, Inc. > (ii) What program are you using, what is your experience > with it. how fast, how user-friendly and how stable > is your program? I use DGauss regularly, for obvious reasons. It is a VERY user-friendly DFT program as a stand-alone code, and even MORE user-friendly when used with the GUI for Silicon Graphics known as: UniChem. DGauss is VERY fast on Cray vector hardware, and it is also fast on SGI workstations. The program uses gaussian functions, in the spirit of standard "ab initio" programs today, to build the density and solve the Kohn-Sham equations. In its present implementation, DGauss uses an all-electron non-relativistic Hamiltonian. The major functions of DGauss include geometry optimizations for stationary points (minima and saddle points) on a potential energy hypersurface. These optimizations are carried out using analytical first derivatives. Until the next release, vibrational frequencies as well as infrared intensities are computed with finite difference techniques. For fixed or optimized geometries, DGauss evaluates molecular properties such as Mulliken charges, multipole moments, molecular orbitals, electron densities, spin densities, and electrostatic potentials. At present, SCF calculations can be carried out on the local spin density (LSD) functional level, or at the non-local (NLSD) level using Becke-Perdew or Becke-Stoll-Preuss-Pavlidou gradient corrections after the SCF or Becke-Perdew or Becke-Lee-Yang-Parr corrections self-consistently. DGauss calculates molecular orbitals (MOs) and the corresponding orbital energies. It is designed to take full advantage of the current knowledge of vector and parallel algorithms in the evaluation of integrals and other computationally demanding numerical procedures, such as the calculation of the exchange-correlation terms which are calculated numerically on a finite grid. For more information about DGauss and/or UniChem please send mail to any of the following individuals at Cray Research, Inc: Dr. George Fitzgerald george@gravity.cray.com Dr. Thomas Raeuchle raeuchle@gravity.cray.com Dr. Eugene Fleischmann gene@jersey.cray.com Best Wishes, Gene ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Eugene D. Fleischmann, Ph.D. Computational Chemist Cray Research, Inc. (609) 252-1250 121 Commons Way gene@jersey.cray.com Princeton, NJ 08540 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ >From inoue@greencross.co.jp Sat Jul 16 05:12 MET 1994 Received: from wincgw1.winc.ad.jp by milliways.physik.unizh.ch (5.0cf2.19) id AA23788; Sat, 16 Jul 94 05:12:29 +0200 Received: by wincgw1.winc.ad.jp (5.67+1.6W/4.18:3.7:winc:931216) id AA29421; Sat, 16 Jul 94 11:58:06 JST Received: by gccss2.greencross.co.jp (4.1/6.4J.6-agu4) id AA28302; Sat, 16 Jul 94 11:18:39 JST Date: Sat, 16 Jul 94 11:18:39 JST From: inoue@greencross.co.jp (Yoshihisa Inoue) Return-Path: Message-Id: <9407160218.AA28302@gccss2.greencross.co.jp> To: suter@physik.unizh.ch Subject: AO based DFT codes Content-Type: text Content-Length: 1219 Status: RO Dear Prof. Suter. (i) Is this list complete, which program should be addeed? DGauss, and I do not know about Prof. Ziegler's program. (ii) What program are you using, what is your experience with it. how fast, how userfriendly and how stable is your program? DMol. DMol is very good. If you want to use Gaussian type MP2, it will take much time, but DMol little. For example, I could not calculate C6NH9 system using MP2 optimization with 64Mbyte memory on IRIS/indigo2(with 10 days, I could not get even 1 cycle optimization result), but using DNP basis sets and FINE mesh and non-local system(they say almost same as MP2/6-31G*) I calculated these system within 2 days on IRIS/indigo. And the results were very good! Recently I tried C16N2H18 system and got optimized results within 5 days using DNP/FINE/non-local. I prefer DMol version 2.3 to version 2.3.5, but am waiting for the new version! Best regard, ____/ ___/ ___/ Yoshihisa INOUE (^_^) the Green Cross Corp. / / / 2-25-1 Shodai-Ohtani,Hirakata,Osaka 573 JAPAN / _ / / / tel: +81-720-56-9328 / / / / fax: +81-720-68-9597 _____/ ____/ ____/ E-mail: inoue@greencross.co.jp For DMol >From inoue@greencross.co.jp Sat Jul 16 05:12 MET 1994 Received: from wincgw1.winc.ad.jp by milliways.physik.unizh.ch (5.0cf2.19) id AA23788; Sat, 16 Jul 94 05:12:29 +0200 Received: by wincgw1.winc.ad.jp (5.67+1.6W/4.18:3.7:winc:931216) id AA29421; Sat, 16 Jul 94 11:58:06 JST Received: by gccss2.greencross.co.jp (4.1/6.4J.6-agu4) id AA28302; Sat, 16 Jul 94 11:18:39 JST Date: Sat, 16 Jul 94 11:18:39 JST From: inoue@greencross.co.jp (Yoshihisa Inoue) Return-Path: Message-Id: <9407160218.AA28302@gccss2.greencross.co.jp> To: suter@physik.unizh.ch Subject: AO based DFT codes Content-Type: text Content-Length: 1219 Status: RO Dear Prof. Suter. (i) Is this list complete, which program should be addeed? DGauss, and I do not know about Prof. Ziegler's program. (ii) What program are you using, what is your experience with it. how fast, how userfriendly and how stable is your program? DMol. DMol is very good. If you want to use Gaussian type MP2, it will take much time, but DMol little. For example, I could not calculate C6NH9 system using MP2 optimization with 64Mbyte memory on IRIS/indigo2(with 10 days, I could not get even 1 cycle optimization result), but using DNP basis sets and FINE mesh and non-local system(they say almost same as MP2/6-31G*) I calculated these system within 2 days on IRIS/indigo. And the results were very good! Recently I tried C16N2H18 system and got optimized results within 5 days using DNP/FINE/non-local. I prefer DMol version 2.3 to version 2.3.5, but am waiting for the new version! Best regard, ____/ ___/ ___/ Yoshihisa INOUE (^_^) the Green Cross Corp. / / / 2-25-1 Shodai-Ohtani,Hirakata,Osaka 573 JAPAN / _ / / / tel: +81-720-56-9328 / / / / fax: +81-720-68-9597 _____/ ____/ ____/ E-mail: inoue@greencross.co.jp From skw@crystal.uwa.edu.au Thu Aug 4 04:29:30 1994 Received: from uniwa.uwa.edu.au for skw@crystal.uwa.edu.au by www.ccl.net (8.6.9/930601.1506) id EAA06254; Thu, 4 Aug 1994 04:05:04 -0400 Received: from graf.crystal.uwa.edu.au (graf.crystal.uwa.edu.au [130.95.232.1]) by uniwa.uwa.edu.au (8.6.8/8.6.4) with SMTP id QAA03903 for ; Thu, 4 Aug 1994 16:10:31 +0800 Received: from pack.crystal.uwa.edu.au by graf.crystal.uwa.edu.au with SMTP (5.61+IDA+MU) id AA09506; Thu, 4 Aug 1994 17:22:31 +0800 From: skw@crystal.uwa.edu.au (Stephen Karl Wolff) Message-Id: <9408040758.AA01809@pack.crystal.uwa.edu.au> Subject: Hermitian diagonalizer wanted. To: chemistry@ccl.net Date: Thu, 4 Aug 94 15:58:40 "WST Mailer: Elm [revision: 70.85] Does anybody have a Hermitian diagonalizer routine, written in FORTRAN, that they could lend us. The matrices we wish to diagonalize are not sparse, and are mostly real or have small complex parts. email: skw@crystal.uwa.edu.au From HODGKINE@prince.mm.wyeth.com Thu Aug 4 04:32:01 1994 Received: from mail1 for HODGKINE@prince.mm.wyeth.com by www.ccl.net (8.6.9/930601.1506) id EAA06242; Thu, 4 Aug 1994 04:02:24 -0400 Received: from mr.wyeth.com by mail1.wyeth.com (PMDF V4.2-14 #5660) id <01HFHXPSFJCG000LGT@mail1.wyeth.com>; Thu, 4 Aug 1994 04:08:37 EST Received: with PMDF-MR; Thu, 4 Aug 1994 04:06:04 EST MR-Received: by mta PRIN1; Relayed; Thu, 04 Aug 1994 04:06:04 -0500 MR-Received: by mta MAIL1; Relayed; Thu, 04 Aug 1994 04:07:18 -0500 Disclose-recipients: prohibited Date: Thu, 04 Aug 1994 09:10:22 -0500 (EST) From: "Hodgkin, Ed" Subject: Summary-Transition States To: "chemistry@ccl.net" Message-id: <01HFHXPTKSP2000LGT@mr.wyeth.com> MIME-version: 1.0 Content-type: TEXT/PLAIN; CHARSET=US-ASCII Content-transfer-encoding: 7BIT UA-content-id: Summary-Transition States X400-MTS-identifier: [;40604040804991/175599@PRINCE] A1-type: MAIL Hop-count: 1 From: Hodgkin, Ed Date: Thu, Aug 4, 1994 9:10 AM Subject: Summary-Transition States To: chemistry@ccl.net Dear All Thanks for your replies to my question about modelling transition states and reaction pathways. Here is a summary. The original question was: > I am hoping to model a reaction mechanism using either MOPAC or Gaussian80 and > need some help with constructing transition states or following the reaction > coordinate. I understand it is possible to do this in both packages. > > My reaction shows unusual selectivity between two products. Each of the > products is formed by a concerted intramolecular mechanism, which I imagine is > the most straightforward case to model. Furthermore reactants and products are > conformationally constrained so there is no problem with flexibility. > > I would be very grateful for any hints you can offer. ============================================================ Message for Hodgkin, Ed From: DR. LIOTTA Date: Thu, Jul 28, 1994 3:40 pm Subject: Re: MOPAC and GAUSSIAN80 To: Hodgkin, Ed File(s): Attach.Txt1 Regarding Message ID : 40245182704991/414481@MAIL1 From: NAME: DR. LIOTTA To: NAME: hodgkine Date : 28-JUL-1994 15:40:00.00 I have recently being studying transition states using MOPAC, but have not been happy with the results so I'm also trying to study them using Gaussian92. I can give you a few pointers for MOPAC and would be interested in hearing what other people tell you about Gaussian80. Firstly you need to create a reaction surface by taking the product and systematically increasing the bond lengths of the bonds which form during your reaction. This should give you a rough estimate of the saddle point. Use the saddle point for further calculations, removing the constraints and using the keywords NNLSQ GRADIENTS CYCLES=100 (or 50-300) AM1 (or PM3 etc.). Repeat this until your gradient is <3 and then use the keyword SIGMA instead of NNLSQ. When a stationary point is found, test if it is a transition structure by using the keywords FORCE and LARGE instead of SIGMA and GRADIENTS. If one negative force constant is obtained, and the bonds which are forming or breaking are in the first vibration, you have a transition state. Hope this helps. I was not sure how much you have already done. V.Cornel (Emory University) ============================================================ Message for Hodgkin, Ed From: kcousins Date: Thu, Jul 28, 1994 12:38 pm Subject: Modeling TS's To: Hodgkin, Ed File(s): Attach.Txt1 Regarding Message ID : 22633182704991/414258@MAIL1 From: NAME: kcousins To: NAME: hodgkine Date : 28-JUL-1994 12:38:51.00 Edward, I have some experience modeling TS's using AMPAC and MOPAC, and will share my wealth with you. There are several papers and reviews on this subject, although I don't have the references. You might try searching the CCL archives as these kinds of questions come up frequently. You might also check out the five (or six) volumes of "Reviews in Computational Chemistry" edited by Lipkowitz and Boyd. I've modeled several 3,3 sigmatropic rearrangement TS's with eventual success. I had an idea what the TS might look like to start with, a somewhat dangerous position that also makes the job alot easier. I defined both the product and starting geometries (Z-matrices) so that all atoms corresponded in number and attachement (e.g. carbon atom 1 in the starting material was also carbon atom 1 in the product, and in both cases the carbon atom is bound to hydrogen atoms 10 and 11). This is not a trivial process, but is sometimes simplified by deriving both geometries in cartesian coordinates, then reordering the atoms so that both are in the same order, then running 1 SCF in Mopac to obtain the Z-Matrix. The correspondance of the two geometries must, however, still be varified. In addition, both Z-Matrices must have the same changing bond length, angle, or dihedral. This variable must be one that changes significantly during the reaction, and can be used to determine the reaction path. In my reaction, the 1 and 6 atoms in the 1,5 unsaturated system undergoing a 3,3-sigmatropic rearrangement are defined such that the sixth atom has a bond length relative to the first atom. This enables me to progressively shortent the atom1-atom6 bond distance to create an approximate reaction path. The reaction path calculation itself is built into MOPAC, and the input consists of the starting geometry with the reaction coordinate indicated with -1, the product geometry likewise, and step-wise values of the reaction coordinates. I vary the 1-6 bond length by .2 angstroms normally. Once the reaction path calculation is complete, the highest energy point is assumed to be an approximate TS and is optimized further using one of the routines to locate saddle points. When a stationary point is reached, a force calculation is performed to verify that that point is indeed a TS. There is also a routine in MOPAC called SADDLE that supposedly does alot of this for you--but you still have to define the two geometries consistantly. I've had more luck with SADDLE when I've used geometries close to the transition state, say two points obtained during the reaction path calculations. Finally, its sometimes difficult to locate a TS structure because the calculation get's "stuck" going around a peak. One way I handle this is to "tweek" one or two atoms (move them a little in the calculation geometry) in an effort to move away from the problem area. Hope these experiences are helpful. If you require further assistance, please ask. Best wishes! Kimberley Cousins Department of Chemistry California State University, San Bernardino 5500 University Parkway San Bernardino, CA 92407 (909)880-5391 kcousins@wiley.csusb.edu ============================================================ Message for Hodgkin, Ed From: m10!frisch Date: Fri, Jul 29, 1994 8:33 am Subject: Re: CCL:Modelling Transition States To: Hodgkin, Ed File(s): Attach.Txt1 Regarding Message ID : 01139092704991/415125@MAIL1 From: NAME: m10!frisch To: NAME: Hodgkin, Ed Date : 29-JUL-1994 08:33:46.00 From: Hodgkin, Ed Date: Thu, Jul 28, 1994 3:38 PM Subject: Modelling Transition States To: chemistry@ccl.net Dear All I am hoping to model a reaction mechanism using either MOPAC or Gaussian80 and need some help with constructing transition states or following the reaction coordinate. I understand it is possible to do this in both packages. If this is not a typo and you're really running Gaussian 80, you might want to consider Gaussian 92. It's only $600 for a Windows PC and most things should be faster in G92 on a 486 PC than in Gaussian 80 on any platform. Gaussian 92 does follow the intrisic reaction coordinate path down from a transition state to the reactants and products connected by that transition state. Gaussian 80 cannot do this, and neither can most versions of MOPAC (I don't think the latest ones can, either, but I'm not positive). There is advice on finding transition structures in our book "Exploring Chemistry Using Electronic Structure Theory", also available from Gaussian, Inc. While the examples are for Gaussian 92, the general principles are applicable with any package. If you'd like more information, you can send email to info@gaussian.com Mike Frisch frisch@gaussian.com ============================================================ Message for Hodgkin, Ed From: E. Lewars Date: Mon, Aug 1, 1994 4:19 pm Subject: modelling transition states To: Hodgkin, Ed File(s): Attach.Txt1 Regarding Message ID : 74127110804991/418217@MAIL1 From: NAME: E. Lewars To: NAME: hodgkine Date : 1-AUG-1994 16:19:16.00 You asked about finding TS's with MOPAC and G80. First, these are not the easiest programs to use, especially G80. Get hold of Spartan if you possibly can. It's made by Wavefunction, Inc, 18401 Von Karman, Suite 210, Irvine, CA 92715; (714)955-2120. If you must use MOPAC or G80, I hope you have a visualizing program like, e.g. PcModel (Serena Software, Box 3076, Bloomington, IN 47402-3076; (812)333-0823). Helpful books on your problem are A Spartan Tutorial (Wavefunction) Experiments in Computational Chemistry, Hehre, Burke, Shusterman, Pietro (Wavefunction) Practical Strategies for Electronic Structure Calculations Wavefunction Evaluation of Modern Electronic Structure Models (Wavefunction) (the above books are oriented to Spartan, but are likely to be useful with MOPAC & G80 too. Also: Exploring Chemistry with Electronic Structure Methods, J. B. Foresman and A. Frisch (Gaussian, Inc., 4415 Fifth Ave., Pittsburgh, PA 15213; (info@gaussian.com). Good Luck === By the way, the Gaussian TS algorithm is particularly a nuisance to use. It is not very robust, and frequently fails. if you must use it, try to get your TS first using a semiemp program like MOPAC, them use that TS as input for G80. Errol Lewars === ============================================================ From pettsj@visigoth.demon.co.uk Thu Aug 4 05:29:30 1994 Received: from post.demon.co.uk for pettsj@visigoth.demon.co.uk by www.ccl.net (8.6.9/930601.1506) id EAA06647; Thu, 4 Aug 1994 04:37:42 -0400 Received: from visigoth.demon.co.uk by post.demon.co.uk id aa15194; 4 Aug 94 9:36 GMT-60:00 X-Sender: (Unverified) Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Thu, 4 Aug 1994 09:38:31 +0000 To: chemistry@ccl.net From: James Petts Subject: Recommendation for Colour Printer? Message-ID: <9408040937.aa15194@post.demon.co.uk> Can anybody recommend a good quality dye-sublimation printer for attaching to a Personal Iris for producing screen dumps and spooled output? -- === James Petts === From oles@kjemi.uio.no Thu Aug 4 06:29:31 1994 Received: from pat.uio.no for oles@kjemi.uio.no by www.ccl.net (8.6.9/930601.1506) id FAA07197; Thu, 4 Aug 1994 05:56:24 -0400 From: Received: from ulrik.uio.no by pat.uio.no with local-SMTP (PP) id <21472-0@pat.uio.no>; Thu, 4 Aug 1994 12:01:44 +0200 Received: by kelvin ; Thu, 4 Aug 1994 12:01:42 +0200 Date: Thu, 4 Aug 1994 12:01:42 +0200 Message-Id: <9408041001.AA04659@kelvin> To: CHEMISTRY@ccl.net In-Reply-To: Hans-Ueli Suter's message of Thu, 4 Aug 1994 09:10:57 +0200 (MET DST) <9408040710.AA19694@milliways.physik.unizh.ch> Subject: Re: CCL:Answers to AO based DFT 600 Lines Thanks to Hans-Ueli Suter for the summary of responses concerning DFT packages. One of the responses was mine; I wrote about the ADF program: > Somewhat clumsy input for geometry optimizations. > > Unnecessary large output, with a lot of redundance and > (in my opinion) unsufficient flexibility in user's choice of output. I feel obliged to mention that in the meantime, a new version (1.1.0) of ADF has been launched, in which the abovementioned drawbacks have been remedied. Ole Swang From aiba@volta.vmsmail.ethz.ch Thu Aug 4 07:29:32 1994 Received: from bernina.ethz.ch for aiba@volta.vmsmail.ethz.ch by www.ccl.net (8.6.9/930601.1506) id GAA07432; Thu, 4 Aug 1994 06:44:52 -0400 Message-Id: <199408041044.GAA07432@www.ccl.net> Received: from VOLTA.vmsmail.ethz.ch by bernina.ethz.ch id <06550-0@bernina.ethz.ch>; Thu, 4 Aug 1994 12:50:04 +0200 X-Vms-To: ethz::"chemistry@ccl.net" To: chemistry@ccl.net From: aiba@volta.vmsmail.ethz.ch (Aiaz Bakassov, Phys. Chem., ETH Zurich) Subject: Reply to T. Daniel Crawford Date: Thu, 4 Aug 1994 12:50:04 +0200 Dear Daniel Crawford, > I have also derived this expression, > but obtained the same answer as S&O. Yes, their expression is the correct one. Who said it is incorrect ? What I was posting was absolutely about my own failure, and all wanted was to make sure that the expression in the book is correct and free from misprints for example. > I believe that if you derive the equation > using second-quantization that > you will obtain the correct result. That's exactly what I have done and got exactly their result. I have already responded to some people that I had got the S&O result using second quantization. The latter is a tool I am really used to, for I am a pure theoretical physicist and am in the quantum chemistry business for a year may be for two -- what an accident of life! :-) My problem was that I was failing to get their result, using that clumsy determinantal technique. Which merely means that I am still missing some point in the book. It is a purely working situation and those who wished to help me (without any other prejudice intended) directed their mails to me. Science must be free >from destructive emotions. Sincerely, Ayaz Bakasov. From hodgkina@vax.ox.ac.uk Thu Aug 4 08:29:34 1994 Received: from oxmail2.ox.ac.uk for hodgkina@vax.ox.ac.uk by www.ccl.net (8.6.9/930601.1506) id HAA07753; Thu, 4 Aug 1994 07:37:30 -0400 From: Received: from vax.ox.ac.uk by oxmail2.ox.ac.uk. with SMTP (PP) id <23497-0@oxmail2.ox.ac.uk.>; Thu, 4 Aug 1994 12:41:50 +0100 Received: by vax.ox.ac.uk (MX V4.1 VAX) id 4; Thu, 04 Aug 1994 12:42:23 +0100 Sender: hodgkina@vax.ox.ac.uk Date: Thu, 04 Aug 1994 12:42:22 +0100 To: CHEMISTRY@ccl.net Message-ID: <0098271E.BC9F35CC.4@vax.ox.ac.uk> Subject: Wanted testers for Mac NMR s/w Dear Netters If you are interested in beta-testing a program for NMR simulation on the Mac please read on: gNMR: (previously avaliable as a DOS package geNMR) is a versatile and comprehensive NMR simulation package which is also easy to use. Its functions include: (1) simulation of one-dimensional NMR spectra of single molecules or mixtures (2) iterative calculations so that you may fit calculated to experimental spectra (3) conversion of experimental spectra from various TforeignU file formats (Bruker, Lybrics, General Electric, Varian, ASCII .....) (4) facilities for chemical exchange calculations of the intra- and inter-molecular type (5) the option to input chemical formulae from chemistry drawing programs The program has a very thorough manual. We hope to find a limited number of volunteers (at most six) to add to our program of beta-testers. If you are keenly interested to test the package, please give me some indication of your Mac expertise and of your familiarity with NMR methods. The program is designed for the Chemist who needs NMR not necessarily the NMR guru. Cherwell Scientific is a specialist publisher of scientific software for the Windows and Mac platforms. We publish programs in chemistry, molecular biology, genetics, statistics, image processing, and other scientific disciplines. Our programs are commercially marketed, documented and suppported, with royalties paid to the developers. The prices are in the region of $200-750. Your comments will be extremely useful to us. If you are keen to beta-test and we are unableto include you in the testing cycle - we cannot make the testing open-ended as the cost of duplicating and shipping manuals is a factor for us - we will inform you of the packages release date etc in due course. Thank you for your help and consideration. Adam Hodgkin Cherwell Scientific The Magdalen Centre Oxford Science Park Oxford OX4 4GA Tel: 44-865-784800 Fax: 44-865-784801 adam.hodgkin@queens.oxford.ac.uk From beck@work2.ch-cip.Uni-Koeln.DE Thu Aug 4 09:29:35 1994 Received: from work2.ch-cip.Uni-Koeln.DE for beck@work2.ch-cip.Uni-Koeln.DE by www.ccl.net (8.6.9/930601.1506) id JAA09469; Thu, 4 Aug 1994 09:02:47 -0400 Received: by work2.ch-cip.Uni-Koeln.DE (5.65/CL/1.1) id AA12112; Thu, 4 Aug 1994 15:08:33 +0200 From: beck@work2.ch-cip.Uni-Koeln.DE (Michael Beck tel. 4816) Message-Id: <9408041308.AA12112@work2.ch-cip.Uni-Koeln.DE> Subject: excited states-MOPAC93 To: CHEMISTRY@ccl.net (mailing list) Date: Thu, 4 Aug 1994 15:08:32 +0200 (MET DST) Cc: reben@work2.ch-cip.Uni-Koeln.DE (Rupert Rebentisch Tel. 0221/470/4816) X-Mailer: ELM [version 2.4 PL13] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 395 Dear Netters! One colleaque of mine has the following trouble: She wants to optimize geometries for exicited states of molecules like stilbene and to calculate their normal coordinates. How shall she use the keywords CISD, C.I.= and ROOT= to get, say, 50 energy selected configurations. Only single and double excitations shall be taken into account. Thanx M.E.Beck for Barbara Tilmanns From ccl@probe.ac1.uni-duesseldorf.de Thu Aug 4 09:30:53 1994 Received: from probe.ac1.uni-duesseldorf.de for ccl@probe.ac1.uni-duesseldorf.de by www.ccl.net (8.6.9/930601.1506) id JAA09723; Thu, 4 Aug 1994 09:20:35 -0400 Received: by probe.ac1.uni-duesseldorf.de (5.61/1.34) id AA01057; Thu, 4 Aug 94 15:24:01 +0200 From: ccl@probe.ac1.uni-duesseldorf.de (Computational Chemistry List) Message-Id: <9408041324.AA01057@probe.ac1.uni-duesseldorf.de> Subject: stable/unstable conformations To: chemistry@ccl.net (Theochem-Liste) Date: Thu, 4 Aug 1994 14:24:00 +0100 (MET) X-Mailer: ELM [version 2.4 PL2] Content-Type: text Content-Length: 425 Hi Netters! For my investigations, I calculate the energy of molecules containing the ethane-rotor. To calculate the energy difference between the 3 canonical rotamers I start with 3 different structures and optimize them. Unfortunately they all give the same result (only one rotamer). Any suggestions how I can solve this problem. I will send the result of my question in this mailing-list. Thanks a lot Kay Kreidler From mccarthy@convx1.ccit.arizona.edu Thu Aug 4 12:33:32 1994 Received: from Arizona.EDU for mccarthy@convx1.ccit.arizona.edu by www.ccl.net (8.6.9/930601.1506) id MAA14163; Thu, 4 Aug 1994 12:24:04 -0400 Received: from convx1.ccit.arizona.edu by Arizona.EDU (PMDF V4.3-9 #2381) id <01HFI8WY5OAO95MVYZ@Arizona.EDU>; Thu, 04 Aug 1994 09:28:55 -0700 (MST) Received: from localhost by convx1.ccit.arizona.edu; Thu, 4 Aug 1994 09:25:25 -0700 Date: Thu, 04 Aug 1994 09:25:25 -0700 From: mccarthy@convx1.ccit.arizona.edu (William Mccarthy) Subject: Re: CCL:stable/unstable conformations To: ccl@probe.ac1.uni-duesseldorf.de, chemistry@ccl.net Message-id: <199408041625.JAA07884@convx1.ccit.arizona.edu> MIME-version: 1.0 Content-type: TEXT/PLAIN; CHARSET=US-ASCII Content-transfer-encoding: 7BIT Perhaps you've not taken into account the symmetry of the internal rotation (i.e. V=cos3x, x=angle or internal rotation), or maybe this is an artifact of the model potential energy surface upon which you're optimizing structure; in an investigation of the inversion of C3v symmetry anions, we've found other local minima by more rigorously describing electron correlation. Good luck, Will McCarthy From tgm@SSD.intel.com Thu Aug 4 14:30:58 1994 Received: from SSD.intel.com for tgm@SSD.intel.com by www.ccl.net (8.6.9/930601.1506) id NAA15652; Thu, 4 Aug 1994 13:48:20 -0400 Received: from idaho.SSD.intel.com by SSD.intel.com (4.1/SMI-4.1) id AA29258; Thu, 4 Aug 94 10:53:20 PDT Date: Thu, 4 Aug 94 10:53:20 PDT From: "Timothy G. Mattson" Message-Id: <9408041753.AA29258@SSD.intel.com> To: CHEMISTRY@ccl.net Subject: Intel Comp. Chem. mail group We have created an unmoderated mail group for people doing computational chemistry on Intel supercomputers. To join the group, send mail to: chemistry-request@ssd.intel.com To send mail to everyone on the list, send mail to: chemistry@ssd.intel.com If you have any questons about the group, send mail to me (Tim Mattson) at: tgm@ssd.intel.com From jdryan@mmm.com Thu Aug 4 22:29:42 1994 Received: from pigseye.mmm.com for jdryan@mmm.com by www.ccl.net (8.6.9/930601.1506) id VAA21082; Thu, 4 Aug 1994 21:58:41 -0400 Received: by pigseye.mmm.com (4.1/SMI-4.1) id AA00432; Thu, 4 Aug 94 21:02:11 CDT Date: Thu, 4 Aug 1994 21:03:08 -0500 Message-Id: <9408050203.AA29846@MOLE.mmm.com> From: jdryan@mmm.com (Jerry D. Ryan - Tech. Computations - 612 733-3595 - jdryan@mmm.com) To: chemistry@ccl.net Subject: Availability of DERMAL Program Hi all, One of our computational chemists has heard of a program DERMAL, but is unable to identify its source or availability. If you know of information relating to the availability of DERMAL, we would appreciate receiving this information. Thanks for your assistance! Jerry Ryan 3M Technical Computations jdryan@mmm.com