From JSMCM@jazz.ucc.uno.edu Wed May 17 01:37:13 1995 Received: from jazz.ucc.uno.edu for JSMCM@jazz.ucc.uno.edu by www.ccl.net (8.6.10/930601.1506) id BAA20952; Wed, 17 May 1995 01:30:36 -0400 Received: from jazz.ucc.uno.edu by jazz.ucc.uno.edu (PMDF #2820 ) id <01HQL931M44G8WXVBP@jazz.ucc.uno.edu>; Wed, 17 May 1995 00:29:51 CST Date: 17 May 1995 00:29:51 -0600 (CST) From: JORGE Subject: Is DFT ab initio? To: CHEMISTRY@ccl.net Message-id: <01HQL931M44I8WXVBP@jazz.ucc.uno.edu> X-VMS-To: IN%"CHEMISTRY@ccl.net" X-VMS-Cc: JSMCM MIME-version: 1.0 Content-type: TEXT/PLAIN; CHARSET=US-ASCII Content-transfer-encoding: 7BIT On The Question: Is DFT an ab initio method? From my dictionary: Theory = System of rules and principles Empirical = Relying on experimental information Empirical = Relying on experience [what about theoretical experience?] Empirical = Not Relying on theory Approximation = very near, nearly correct My understanding of a pure ab initio method is of any method that solves (exactly or approximately) the Schrodinger equation, this means: using only the principles of quantum mechanics and not using empirical or other information extrapolated, interpolated, or transferred from other system(s). Whether the method is a very rough approximation, is not going to eliminate its ab initio character. In this sense, we have a spectrum of methods from the very precise to the very inaccurate ones, forcing us to a trade-off because the cost of the method is usually a direct function of its precision. In general any practical computational method will be part ab initio and part empirical in character, and it is not much subjective to determine if a given method can be classified as ab initio, empirical or semiempirical. The selection and use of a basis set is always based on experience; whether, theoretical or experimental, one can always create or improve a basis set based on how it behaved in earlier calculations. The basis set is telling your method the spectral region to look for the solutions but it does not constitute the solution itself. A fully 100% analytical ab initio method would be limited to the solution of the hydrogen-like atoms. Another pure 100% ab initio, but nevertheless approximated method, would be any of those variational-perturbational methods that solves the helium-like atoms with about 15 figures!! of precision, and where the basis set has been chosen with the only criteria of the lowest possible energy. Then come our standard ab initio methods, based on the central field approximation, or HF-based methods (HF, MP, CC, CI, MRHF,...) with several degrees of accuracy and several degrees of small empirical (due to experiment or experience from earlier calculations) corrections and truncations. We call all of them ab initio, and there is an implicit understanding that this is where our definition of ab initio is settled. Small corrections to those methods, like 10% correction to the ZPE do not change our attitude toward the ab initio character of the methods. There exists the accepted Gaussian2 theory where corrections, not necessarily small, are made to the energies based on precise experimental results. Neither our ability to improve the level of theory would exclude the ab initio character of a method. Now to the question if DFT methods are ab initio: First of all, you CAN derive DFT directly from the Schrodinger equation even without recurring, as in HF methods, to the central field approximation. The fact that a well defined but nowadays unsolvable functional shows up can not exclude the ab initio character of DFT. As in any of the ab initio methods above, approximations are made. A classification could be done whether the functional was derived using entirely first principles information or if it was obtained by using experimental or other transferred information Only on those cases where the functional is empirical, the particular DFT method can be catalogued as semiempirical. Functionals like the local approximation, gradient approximation are exact functionals (up to their order) and are obtained from pure first principles arguments. The GGA of Perdew and Wang for exchange-correlation, PW91, is strictly ab initio and there are evidence that it performs as well as QCI and MP4 methods. A predecessor to the PW91, the PW86, has been used with systems containing up to transition metals with great success. DFT methods can also be developed using fully numerical techniques, like in Becke's Numol, eliminating the use of basis sets. Using this method together with one of the ab initio functionals yields a 100% ab initio procedure, i.e., more ab initio than any of the standard ones. VERDICT: We find DFT ab initio Jorge Seminario jsmcm@jazz.ucc.uno.edu From ahirasim@agr.kyushu-u.ac.jp Wed May 17 04:52:18 1995 Received: from nogakubu.agr.kyushu-u.ac.jp for ahirasim@agr.kyushu-u.ac.jp by www.ccl.net (8.6.10/930601.1506) id EAA23082; Wed, 17 May 1995 04:47:29 -0400 Received: from [133.5.200.49] (pestic3.ach.agr.kyushu-u.ac.jp [133.5.200.49]) by nogakubu.agr.kyushu-u.ac.jp (8.6.9+2.4W/3.4Wbeta3) with SMTP id RAA25814 for ; Wed, 17 May 1995 17:48:08 +0900 Message-Id: <199505170848.RAA25814@nogakubu.agr.kyushu-u.ac.jp> Date: Wed, 17 May 1995 17:50:47 +0900 To: chemistry@ccl.net From: ahirasim@agr.kyushu-u.ac.jp (Akinori Hirashima) X-Sender: ahirasim@nogakubu.agr.kyushu-u.ac.jp Subject: CCL: web for chemistry: addition to summary Return-Receipt-To: ahirasim@agr.kyushu-u.ac.jp MIME-Version: 1.0 Content-Type: text/plain; charset=iso-2022-jp X-Mailer: Eudora-J(1.3.8.5-J13) Dear Netters: Recently I have posted a question concerning web for chemistry and its summary. I would like to express my sincere gratitude to the netters whose responses I received. Some of the replies were very helpful. I woul like to add non-USA sites to USA sites reported in my previous summary: http://www.shef.ac.uk/uni/academic/A-C/chem/chemistry-www-sites.html http://www.uia.ac.be/u/thys/chem.html http://www.ch.cam.ac.uk/ChemSites.html http://www.uia.ac.be/u/thys/chem.html http://hackberry.chem.niu.edu:70/0/webpage.html http://www.chem.vt.edu/Yip/www.html http://boris.qub.ac.uk/edward/index.html http://wwwhost.cc.utexas.edu/world/instruction/index.html http://www.rpi.edu/dept/chem/cheminfo/chemres.html http://www.yahoo.com/Science/Chemistry/ http://www.latrobe.edu.au/chejs/chem.html http://www.ch.ic.ac.uk/GIC/ http://www.oxmol.co.uk http://www.csc.fi/lul/chem/graphics.html http://www.rpi.edu/dept/chem/cheminfo/chemres.html http://galaxy.einet.net/galaxy/Science/Chemistry.html http://www.eecs.wsu.edu/IRL/ANIM/Anim.html#anims http://strange.engr.washington.edu/index.html http://www.geom.umn.edu/ http://csdvx2.ccdc.cam.ac.uk/ http://www.lib.utexas.edu/Libs/Chem/first/mallet.html http://www.polymers.com/phoenix.html http://pctc.chemie.uni-erlangen.de/fakultaet/chemie.html http://www.pc.chemie.th-darmstadt.de/ http://www.chemie.fu-berlin.de/index.html http://tci002.uibk.ac.at:80/noy/ http://www.chem.csiro.au/ http://www.uia.ac.be/u/thys/index.html http://www.netlib.org/index.html http://wucmd.wustl.edu/index.html http://www.jungle.com/ http://CHEDR.IDBSU.EDU/ http://atlas.nlm.nih.gov:5700/Entrez/version2/medline.html http://www.dcl.co.il/conf_clust.html http://www.ccc.uni-erlangen.de/info/OML/about-OML.html http://btm2xd.mat.uni-bayreuth.de/match/ http://www-kun.sci.kun.nl:80/sigma/ http://www.ucg.ie/chem/chemdept.htm http://http1.brunel.ac.uk:8080/depts/chem/ http://wavefun.com/ http://www.i2020.net/edusoft/hint.html http://www.che.chalmers.se/welcome.html http://www.biochem.mpg.de/ http://derioc1.organik.uni-erlangen.de/ccc-welcome.html http://scsg9.unige.ch/eng/toc.html http://rchs1.uni-regensburg.de:80/preprint.html ahirasim@agr.kyushu-u.ac.jp (Akinori Hirashima) From kochw@argon.chem.TU-Berlin.DE Wed May 17 05:08:32 1995 Received: from mailgzrz.TU-Berlin.DE for kochw@argon.chem.TU-Berlin.DE by www.ccl.net (8.6.10/930601.1506) id FAA23295; Wed, 17 May 1995 05:07:09 -0400 Received: from argon.chem.TU-Berlin.DE by mailgzrz.TU-Berlin.DE with SMTP (PP); Wed, 17 May 1995 11:06:45 +0200 Received: by argon.chem.TU-Berlin.DE (AIX 3.2/UCB 5.64/4.03) id AA28372; Wed, 17 May 1995 11:06:16 +0200 From: kochw@argon.chem.TU-Berlin.DE (Wolfram Koch) Message-Id: <9505170906.AA28372@argon.chem.TU-Berlin.DE> Subject: Q-Chem (fwd) To: chemistry@ccl.net Date: Wed, 17 May 1995 11:06:15 +0200 (DFT) X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 3678 Hi, since my original question with regard to QCHEM was of interest not only to me but to other subscribers as well, I append Benny Johnson's reply (thanks...) to this posting. Wolfram Forwarded message: > From JOHNSONB@B.PSC.EDU Tue May 16 18:46:59 1995 > Date: Tue, 16 May 1995 12:45:06 -0400 (EDT) > From: JOHNSONB@B.PSC.EDU > To: kochw@argon.chem.TU-Berlin.DE > Cc: JOHNSONB@B.PSC.EDU > Message-Id: <950516124506.20601b84@B.PSC.EDU> > Subject: Q-Chem > > Dear Prof. Dr. Koch, > > I am sorry not to reply to you sooner -- I have been out of the office. > > >we heard a lot of rumor about the Gaussian off-spin program called > >QCHEM. Can anyone - prefereably those in charge of QCHEM - > >provide some information, such as: what is it good for, > >where is it better than G9x, how much does it cost (very important...), > >where to get it from, any positive or negative experiences, for which > >platformns is it available, any GUI available, etc.. > > Here is some information on Q-Chem. I hope it is helpful in answering > your questions. > > Q-Chem was founded in 1993 to build a new quantum chemistry package from > the ground up for modern computer architectures. The principals of Q-Chem > are myself, Peter Gill (Massey University) and Martin Head-Gordon > (UC Berkeley). Q-Chem is not connected with Gaussian, Inc. in any way. > However, the three of us used to collaborate with Gaussian and have implemented > a number of recent features within the Gaussian program (e.g. DFT, 2-e > integrals and derivatives, direct SCF and MP2, QCI). > > Q-Chem incorporates many conventional ab initio (HF, MP, CIS, etc.) > as well as density functional methods. The main strength of Q-Chem over > currently available programs is that it is designed from the beginning with > new algorithms for the efficient treatment of very large molecular systems. > Additionally, the algorithms have been structured to take advantage of > massively parallel computing platforms from the outset, rather than attempting > to adapt pre-existing code to parallel architectures. In these ways Q-Chem > will significantly expand the range of tractable chemical problems. > > An example of this is Continuous Fast Multipole Method (CFMM), which is a > technique for evaluating the Coulomb matrix (the rate-limiting step in > large-scale density functional calculations) in an amount of work which is > linear in system size, instead of quadratic as required by other programs. > > Q-Chem is under development and is expected to be released publicly in > early 1996 on a variety of Unix workstations and supercomputers. Currently > we run on IBM, SGI, DEC, HP, Cray, Fujitsu and Convex platforms. Although > we have not yet fixed a definite price, Q-Chem will be competitive with other > commercial programs and will carry a substantial academic discount. > A GUI will definitely be available. Compatibility with a number of > established GUIs is planned. > > If you have additional questions about the program or the company, I would be > happy to answer them. > > Sincerely yours, > Benny Johnson, Ph.D. > President > Q-Chem, Inc. > 317 Whipple St. > Pittsburgh, PA 15218 > USA > -- _________________________________________________________________________ Prof. Dr. Wolfram Koch Institut fuer Organische Chemie, Sekr.C3 Technische Universitaet Berlin Strasse des 17.Juni 135, D-10623 Berlin Fon: +49 30 314 27870, Fax: +49 30 314 21102 e-mail: kochw@argon.chem.TU-Berlin.DE _________________________________________________________________________ From oscar@MPA-Garching.MPG.DE Wed May 17 05:37:17 1995 Received: from ibm-1.MPA-Garching.MPG.DE for oscar@MPA-Garching.MPG.DE by www.ccl.net (8.6.10/930601.1506) id FAA23660; Wed, 17 May 1995 05:31:18 -0400 Received: (oscar@localhost) by ibm-1.MPA-Garching.MPG.DE (8.6.12/8.6) id LAA40680; Wed, 17 May 1995 11:31:17 +0200 Date: Wed, 17 May 1995 11:31:16 +0200 (MDT) From: "Oscar N. Ventura" To: chemistry@ccl.net Subject: CCL:is DFT ab initio? Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Dear netters, Reading the messages about whether DFT is ab initio or not, I got a little bit puzzled. Why is it such a reluctance in admiting DFT as a truly separated category of theoretical methods? In my opinion, it is clear that semi-empirical and DFT methods share a key characteristic: there is no systematic way of improving the results. If for a system you got a bad result (be it far from experiment, inconsistent or just crazy) using the "best" semi-empirical or DFT methods available, then there is no room for systematic improvement. Of course, one can try to derive new parametrizations (PM3 vs AM1) or somehow mix calculations and shake a bit (Becke3LYP vs BLYP), but that is not truly systematic, is it? On the other side, DFT and ab initio share the characteristic that they are not resorting to experiment to determine the parameters they are using (in potentials and in basis sets for instance). In this respect they are clearly different from semi-empirical. On a second thought, it is also true that semi-empirical and ab initio do share some ground in what concerns approximations. What about adding -3.07 millihartrees for each bound paired valence electron in Gaussian 1 theory? And is it systematic really the passage from Gaussian 1 to Gaussian 2? So then, are Gaussian 1 and Gaussian 2 ab initio theories or semi-empirical ones? In my opinion, to treat Semi-empirical, Ab initio and DFT methods as different approximate models to attempt a solution of the Schroedinger equation is the most fair approach. Unless, obviously, that someone may think there is more prestige involved in studying DFT if it is catalogued as an ab initio theory rather than as a semi-empirical one. But surely that is not the case, is it? O. ------------------------------------------------------------------------------- Dr. Oscar N. Ventura Tel. +(5982)941860 MTC-Lab Fax +(5982)941906 Gral.Flores 2124, C. C. 1157 E-mail oscar@bilbo.edu.uy Montevideo 11800, URUGUAY MTC-Lab http://bilbo.edu.uy/MTC-Lab.html ------------------------------------------------------------------------------- From aiba@ir.phys.chem.ethz.ch Wed May 17 06:07:18 1995 Received: from bernina.ethz.ch for aiba@ir.phys.chem.ethz.ch by www.ccl.net (8.6.10/930601.1506) id GAA23978; Wed, 17 May 1995 06:05:34 -0400 Received: from volta (actually volta.ethz.ch) by bernina.ethz.ch with SMTP inbound; Wed, 17 May 1995 12:04:47 +0200 Date: Wed, 17 May 1995 12:04:54 +0200 Message-Id: <95051712045442@ir.phys.chem.ethz.ch> From: aiba@ir.phys.chem.ethz.ch (Ayaz Bakasov, Phys. Chem., ETH Zurich) To: chemistry@ccl.net Subject: CI on top of UHF ??? X-VMS-To: smtp%"chemistry@ccl.net" X-VMS-Cc: AIBA Dear Netters, The question I am posing is for those who are willing to get into some theoretical consideration. However, in case of a definitive response a pointer to (possibly) existing method, or code, or an option in a code will be very welcome. I have to calculate expectation values of an operator which is nonzero if: 1) Neither alpha- nor beta- spin component of the wave function is CONSTANT (including zero), i.e. both spin components are non-trivial functions of coordinates; 2) The alpha- and beta- spin components of the wave function are not equal to each other (i.e. are different functions of coordinates). Obviously, I am getting non-zero expectation values for a UHF wave function. It is highly desirable to have better wave function, such as CI. However, the CI available, for instance, from G92/94, is based on the RHF reference determinant. It means that whatever CI I will produce, this will be a combination of determinants with zero (or equal) alpha or beta spin component for a given spatial molecular orbital. If UHF gives already non-zero, are there CI (or analogous methods) available which are based on the UHF reference determinant ??? Sincerely, Ayaz Bakasov. From qsar!qsar.chem.msu.su!baskin@gamma.srcc.msu.su Wed May 17 07:37:19 1995 Received: from gw.srcc.msu.su for qsar!qsar.chem.msu.su!baskin@gamma.srcc.msu.su by www.ccl.net (8.6.10/930601.1506) id HAA24944; Wed, 17 May 1995 07:27:46 -0400 Received: from srcc.UUCP (uusrcc@localhost) by gw.srcc.msu.su (8.6.12/8.6.12) with UUCP id PAA03909 for CHEMISTRY@ccl.net; Wed, 17 May 1995 15:27:39 +0400 Received: by gamma.srcc.msu.su; Wed, 17 May 1995 15:06:22 +0400 Received: by qsar.chem.msu.su (dMail for DOS v1.23, 15Jun94); Wed, 17 May 1995 10:09:01 +0400 To: CHEMISTRY@ccl.net Message-Id: Organization: Lab. of Org.Synth., MSU Date: Wed, 17 May 1995 10:09:01 +0400 (MSD) From: "Igor Baskin" X-Mailer: dMail [Demos Mail for DOS v1.23] Subject: Partial derivatives of internal coordinates Lines: 18 Dear CCLers, I look for mathematical expressions (or a source code) for calculating the partial derivatives of internal molecular coordinates, such as valence and torsion angles, or their cosines, with respect to Cartesian coordinates of atoms. Such relations are widely used in MM and MD programs. Thank you in advance. -------------------------------------------------------------------- Dr. Igor I. Baskin ! e-mail: baskin@qsar.chem.msu.su Institute of Organic Chemistry & ! phone : 7-095-939-3557 Chemistry Department, ! FAX : 7-095-939-0290 Moscow State University, ! Moscow, Russia ! -------------------------------------------------------------------- From toukie@zui.unizh.ch Wed May 17 08:22:21 1995 Received: from rzusuntk.unizh.ch for toukie@zui.unizh.ch by www.ccl.net (8.6.10/930601.1506) id IAA25364; Wed, 17 May 1995 08:21:47 -0400 Received: by rzusuntk.unizh.ch (4.1/SMI-4.1.9) id AA14474; Wed, 17 May 95 14:21:37 +0200 X-Nupop-Charset: Swiss Date: Wed, 17 May 1995 14:21:27 +0100 (MET) From: "Hr Dr. S. Shapiro" Sender: toukie@zui.unizh.ch Reply-To: toukie@zui.unizh.ch Message-Id: <51687.toukie@zui.unizh.ch> To: chemistry@ccl.net Subject: DGEOM_95 ported to DOS/Windows Dear Colleagues; If anyone has successfully ported the programme DGEOM_95 for execution on a DOS/Windows PC, kindly contact me. I am interested in possibly using this programme on a Pentium/90 MHz PC fitted with MS-DOS 6.2/Windows for Workgroups 3.11, 16 MB RAM, and a 1 GB HD. Thanks in advance to all responders. Sincerely, (Dr.) S. Shapiro Inst. f. orale Mikrobiol. u. allg. Immunol. Zent. f. Zahn-, Mund- und Kieferheilkd. der Univ. ZH Plattenstr. 11 Postfach CH-8028 Zuerich 7 Switzerland Internet: toukie@zui.unizh.ch From siroiss@CERCA.UMontreal.CA Wed May 17 09:22:30 1995 Received: from condor.CC.UMontreal.CA for siroiss@CERCA.UMontreal.CA by www.ccl.net (8.6.10/930601.1506) id JAA26274; Wed, 17 May 1995 09:08:52 -0400 From: Received: from balzac.CERCA.UMontreal.CA by condor.CC.UMontreal.CA with SMTP id AA13877 (5.65c/IDA-1.4.4 for CHEMISTRY@ccl.net); Wed, 17 May 1995 09:08:18 -0400 Received: from pellan.CERCA.UMontreal.CA by balzac.CERCA.UMontreal.CA (950215.SGI.8.6.10/5.17) id JAA14363; Wed, 17 May 1995 09:09:14 -0400 Received: by pellan.CERCA.UMontreal.CA (8.6.10/5.17) id JAA28035; Wed, 17 May 1995 09:09:13 -0400 Message-Id: <199505171309.JAA28035@pellan.CERCA.UMontreal.CA> Subject: is DFT ab initio To: CHEMISTRY@ccl.net Date: Wed, 17 May 1995 09:09:12 -0400 (EDT) Cc: siroiss@CERCA.UMontreal.CA (Suzane Sirois) X-Mailer: ELM [version 2.4 PL23] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 975 Depuis que tout le monde boit du champagne on doit se demander si ce que l'on boit est vraiment du champagne! Obviously a question like "is DFT ab inito?" will not be asked by a connaisseur! Or may be we need to revised the definition of ab initio and semi-empirical ..., based on how the functionals in DFT are parametrized as well as how the basis set in HF are contracted. I have also a question: Is the Kohn-Sham methodology has come to a limit? because we see more and more that one functional alone can not treat at the same time with the same degree of accuracy both metallic and biological systems I will be delighted to hear one answer coming from a gourou who's working on solving directly the DFT equation and from one who's working with the SCF-KS method "Hope of a deliverance from the darkness that surrounds us" Paul McCartney Suzanne Sirois Centre de Recherche en Calcul Applique (CERCA) siroiss@cerca.umontreal.ca From m10!mjf!frisch@uunet.uu.net Wed May 17 10:07:21 1995 Received: from relay3.UU.NET for m10!mjf!frisch@uunet.uu.net by www.ccl.net (8.6.10/930601.1506) id JAA27389; Wed, 17 May 1995 09:52:28 -0400 Received: from uucp3.UU.NET by relay3.UU.NET with SMTP id QQyqct13904; Wed, 17 May 1995 09:52:28 -0400 Received: from m10.UUCP by uucp3.UU.NET with UUCP/RMAIL ; Wed, 17 May 1995 09:52:27 -0400 Received: from mjf by m10.gaussian.com; Wed, 17 May 1995 09:31:04 -0400 Received: by mjf (AIX 3.2/UCB 5.64/4.03) id AA22871; Wed, 17 May 1995 09:34:03 -0400 From: m10!mjf!frisch@uunet.uu.net (Mike Frisch) Message-Id: <9505171334.AA22871@mjf> Subject: UHF-CISD To: chemistry@ccl.net Date: Wed, 17 May 1995 09:34:03 -0400 (EDT) X-Mailer: ELM [version 2.4 PL24] Content-Type: text Content-Length: 2755 Ayaz Bakasov, Phys. Chem., ETH Zurich writes: I have to calculate expectation values of an operator which is nonzero if: 1) Neither alpha- nor beta- spin component of the wave function is CONSTANT (including zero), i.e. both spin components are non-trivial functions of coordinates; 2) The alpha- and beta- spin components of the wave function are not equal to each other (i.e. are different functions of coordinates). It is highly desirable to have better wave function, such as CI. However, the CI available, for instance, from G92/94, is based on the RHF reference determinant. It means that whatever CI I will produce, this will be a combination of determinants with zero (or equal) alpha or beta spin component for a given spatial molecular orbital. If UHF gives already non-zero, are there CI (or analogous methods) available which are based on the UHF reference determinant ??? Actually, the open-shell CISD code is Gaussian IS based on a UHF rather than an RHF reference determinant, and you can get the appropriate alpha and beta generalized density matrices if you wish to computer properties. So, I think this is just what you want. While we're on the subject, I'd like to clear up a related misconception which I've seen in the literature. The UCISD wavefunction in Gaussian is NOT an exact spin eigenfunction. It consists of all determinants produced by replacing one or two spin-orbitals which are occupied in the reference determinant with virtuals. For an open-shell system, this would not lead to a spin-eigenfunction even if the reference determinant had the same alpha and beta orbitals, because some determinants which are a double excitation over spacial orbitals but a triple excitation over spin-orbitals (because the spins are flipped) are needed to form proper spin-eigenfunctions. To show a simpler example with singles vs. doubles, if the reference determinant is 1a 1b 2a, then the determinant 1a 2a 3b is considered a single excitation over spin-orbitals but 1a 2b 3a, which should be mixed with 1a 2a 3b to form a spin-eigenfunction, is a double excitation of spin-orbitals from the reference. Of course, both of these would be included in a CISD wavefunction, but the same issue occurs with double and triple excitations. All of these problems are specific to the open-shell case and the closed-shell CISD wavefunction in Gaussian is a proper spin-eigenfunction. Also, note that for most properties of either closed or open-shell systems, QCISD is considerably more accurate than CISD with a modest difference in cost and is to be preferred. Mike Frisch Lorentzian, Inc. frisch@lorentzian.com From J.M.Jansen@farm.RUG.NL Wed May 17 10:52:21 1995 Received: from hearnvax.nic.surfnet.nl for J.M.Jansen@farm.RUG.NL by www.ccl.net (8.6.10/930601.1506) id KAA28363; Wed, 17 May 1995 10:40:24 -0400 Received: from mailhost.rug.nl by HEARNVAX.nic.SURFnet.nl (PMDF V4.2-12 #3330) id <01HQM787VPC0008WTW@HEARNVAX.nic.SURFnet.nl>; Wed, 17 May 1995 16:37:47 +0200 (MET-DST) Received: from dep.farm.rug.nl by mailhost.rug.nl with SMTP (PP); Wed, 17 May 1995 16:37:42 +0200 Received: from novell_farm1.farm.rug.nl by dep.farm.rug.nl (QAA20994); Wed, 17 May 1995 16:33:32 +0200 Received: from FARM1/MAIL by novell_farm1.farm.rug.nl (Mercury 1.20); 17 May 95 17:02:31 MET Received: from MAIL by FARM1 (Mercury 1.20); 17 May 95 17:02:25 MET Date: Wed, 17 May 1995 17:02:23 +0100 (CET) From: Hanneke Jansen Subject: Three Dimensions in Drug Design To: CHEMISTRY@ccl.net Message-id: <56A3FF00A56@novell_farm1.farm.rug.nl> Organization: Pharmacy Dept Groningen University X-Envelope-to: CHEMISTRY@ccl.net X-Mailer: Pegasus Mail/Windows (v1.22) Content-transfer-encoding: 7BIT Priority: normal THREE DIMENSIONS IN DRUG DESIGN a minisymposium University Centre for Pharmacy, Groningen, the Netherlands June 2nd, 1995 As soon as people realized that in order for a drug to elicit a response this compound has to complement the structure of a macromolecular target, attempts have been made to visualize this interaction. Nowadays we are well aware of the fact that the complementarity has to include all three dimensions in space. Synthetic and computational chemistry have evolved to a level at which the third dimension can be handled well. Overviews of the capabilities of the techniques involved, examples of their application and aspects of new methodology will be presented in the symposium "Three Dimensions in Drug Design" by the speakers listed below. Prof. Dr. U. Hacksell: Synthetic Aspects of Medicinal Chemistry Dr. J.T.F. Keurentjes: Membrane Technology for the Separation of Racemic Mixtures. Prof. Dr. J.P. Tollenaere: The Role of Computational Chemistry in Pharmaceutical Research Coffee-break Dr. J.P. Snyder: Taxol: A Complicated Molecule Made Simple Prof. Dr. H.C.J. Ottenheijm: Chirality: Food for Medicinal Chemists The last item on the program will present thoughts about various aspects of the third dimension and about the question which additional dimensions have to be added to really understand the drug-target interaction and its effect. All participants are invited to join the discussion following a short introduction. The time-frame of the program will probably not allow us to conclude this exchange of ideas, but a reception is planned after the closing of the symposium during which this discussion can be continued. The symposium will take place at the University of Groningen (the Netherlands), location Antonius Deusinglaan 1, ground floor, room BG04. The registration desk will be opened, and coffee and sandwiches will be available, from 13:30 h. onwards. The symposium will be opened at 14:00 h. and after the closing at 17:00 h. a reception will be offered to all participants. The official language of the symposium will be English. Before June 2nd, the symposium secretariat can be reached at Antonius Deusinglaan 2, 9713 AW Groningen, the Netherlands, phone: +31-50- 633296, fax: +31-50-636908, e-mail: sfc@farm.rug.nl; during symposium hours the phone-number is +31-50-638000. Registration for the symposium is free of charge and can be effected by sending your name, address and affiliation to the symposium secretariat at the above mentioned address. This symposium has been sponsored by Astra Pharmaceutica B.V., Glaxo B.V., the Groningen Centre for Drug Research, the Groningen Utrecht Institute for Drug Exploration, the Janssen Research Foundation, Nefarma, N.V. Organon, Solvay Duphar B.V. and Unilever. Organizing committee: Hanneke Jansen, Cor Grol, Willem Jan Drijfhout, Janita Zwinderman ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ J.M. Jansen e-mail: hanjan@farm.rug.nl Univ. Centre for Pharmacy tel: +31-50633305 Dept. Medicinal Chemistry fax: +31-50636908 A. Deusinglaan 2 NL-9713 AW Groningen the Netherlands ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ From acp37@rs1.rrz.Uni-Koeln.DE Wed May 17 11:22:29 1995 Received: from rs1.rrz.Uni-Koeln.DE for acp37@rs1.rrz.Uni-Koeln.DE by www.ccl.net (8.6.10/930601.1506) id LAA29418; Wed, 17 May 1995 11:17:41 -0400 Received: from axp1.rrz.Uni-Koeln.DE by rs1.rrz.Uni-Koeln.DE with SMTP id AA92484 (5.67b/IDA-1.5 for ); Wed, 17 May 1995 17:17:21 +0200 Received: by axp1.rrz.Uni-Koeln.DE; (5.65/1.1.8.2/23Apr94-0300PM) id AA06935; Wed, 17 May 1995 17:17:21 +0200 Date: Wed, 17 May 1995 17:17:21 +0200 (MET DST) From: Thorsten Koch To: CCL Subject: How to write symmetric z-matrices Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Dear netters, does anybody know references about how to write z-matrices which impose molecular symmetry and have the correct number of degrees of freedom? Any hints will be appreciated. T. Koch /-----------------------------------------------------------------\ | Thorsten Koch | | Institut fuer physikalische Chemie II der Universitaet zu Koeln | | acp37@rrz.uni-koeln.de | | Tel. +49 [0]221 470 4816 | \-----------------------------------------------------------------/ From thomas_w@btm2x2.mat.uni-bayreuth.de Wed May 17 13:07:23 1995 Received: from btr0x1.hrz.uni-bayreuth.de for thomas_w@btm2x2.mat.uni-bayreuth.de by www.ccl.net (8.6.10/930601.1506) id MAA02045; Wed, 17 May 1995 12:55:34 -0400 Received: from btm2x2.mat.uni-bayreuth.de by btr0x1.hrz.uni-bayreuth.de (4.1/btr0x1 (UBTGW/btr0x1-2.4.7)) id AA19184; Wed, 17 May 95 10:24:35 +0200 Received: by btm2x2.mat.uni-bayreuth.de (4.1/SMI-4.1) id AA16903; Wed, 17 May 95 10:23:51 +0200 Date: Wed, 17 May 95 10:23:51 +0200 From: thomas_w@btm2x2.mat.uni-bayreuth.de (Thomas Wieland) Message-Id: <9505170823.AA16903@btm2x2.mat.uni-bayreuth.de> To: chemistry@ccl.net Subject: 6th Int Conf Math Chem Dear Netters, I will attend the 6th International Conference on Mathematical Chemistry, 10-14 Jul 1995, in Pitlochry, Scotland, U.K., and I am looking for people who will also go there. In a poster session on Thursday afternoon, I will hold a computer demonstra- tion of our program system MOLGEN. The organizer of the conference wrote me that I should better bring my own notebook with me since the configuration of the hardware might cause some difficulties. I would now like to get in touch with other people who are also planning to make a computer demonstration there. Perhaps we can coordinate our talks in order to avoid everybody carrying one more computer. My program can run under MS-Windows 3.1, IBM OS/2 2.1 or 3.0 and Sun Solaris 2.3 (with Motif 1.2.3 installed). I'm looking forward to hearing from you. Best regards, Thomas Wieland +---------------+ Dipl. Math. |+---- +----+| Lehrstuhl II f. Mathematik |\ \ | || Universitaet Bayreuth | \ \ | || | \ \ | || 95440 Bayreuth | \ \\ || Germany | \ \\ || Tel. +49 (921) 553386 | \ \\ || Fax +49 (921) 553385 | \-------|| +---------------+ From SATYAM@vms.cis.pitt.edu Wed May 17 14:22:26 1995 Received: from myriad.cis.pitt.edu for SATYAM@vms.cis.pitt.edu by www.ccl.net (8.6.10/930601.1506) id OAA04360; Wed, 17 May 1995 14:18:29 -0400 From: Received: from vms.cis.pitt.edu by vms.cis.pitt.edu (PMDF V4.3-10 #10002) id <01HQM279VTTC9N41LO@vms.cis.pitt.edu>; Wed, 17 May 1995 14:18:23 -0400 (EDT) Date: Wed, 17 May 1995 14:18:23 -0400 (EDT) Subject: Need Address of Book shops/dealers/ for some classics To: chemistry@ccl.net Message-id: <01HQM279WWEA9N41LO@vms.cis.pitt.edu> X-Envelope-to: chemistry@ccl.net X-VMS-To: IN%"chemistry@ccl.net" MIME-version: 1.0 Content-type: TEXT/PLAIN; CHARSET=US-ASCII Content-transfer-encoding: 7BIT Dear Netters, Are there any book shops / second book shops / dealers / persons who sell some of the classic books on Quantum Chemistry e.g book by salem , streitwiesser, greenwood, strauss etc. Thanks for the help Satyam [ADVANCED TECHNOLOGY IS THE BUILDING..OLD CLASSICS ARE FOUNDATION ] From elewars@alchemy.chem.utoronto.ca Wed May 17 17:22:26 1995 Received: from alchemy.chem.utoronto.ca for elewars@alchemy.chem.utoronto.ca by www.ccl.net (8.6.10/930601.1506) id RAA08219; Wed, 17 May 1995 17:16:35 -0400 Received: (from elewars@localhost) by alchemy.chem.utoronto.ca (8.6.10/8.6.10) id RAA07822 for chemistry@ccl.net; Wed, 17 May 1995 17:16:32 -0400 Date: Wed, 17 May 1995 17:16:32 -0400 From: "E. Lewars" Message-Id: <199505172116.RAA07822@alchemy.chem.utoronto.ca> To: chemistry@ccl.net Subject: COMP CHEM BOOKS There was a question about (1) How to get out-of-print "classics" and (2) the new book on computational chemistry from Oxford University Press. (1) Hard-to-find books on science (and anything else) may sometimes be bought from Dover publishers. They reprint (all in paperback, I think) otherwise out-of-print books. Their catalog has many books on mathematics, physics and chemistry. Ask your library or bookstore for their address and order the Mathematics and Science Catalog. (2) The Oxford U Press book is *Computational Chemistry*, by Guy H. Grant & W. Graham Richards, cloth $37.50, paper $13.50. I think it's in a series of introductory texts. It was to have appeared in January, but my copy still hasn't arrived. ===== Errol Lewars ===== From !rgab@trpntech.com Wed May 17 17:37:28 1995 Received: from netcomsv.netcom.com for !rgab@trpntech.com by www.ccl.net (8.6.10/930601.1506) id RAA08222; Wed, 17 May 1995 17:16:41 -0400 Received: from .UUCP by netcomsv.netcom.com with UUCP (8.6.12/SMI-4.1) id OAA16053; Wed, 17 May 1995 14:07:37 -0700 Message-Id: <199505172107.OAA16053@netcomsv.netcom.com> Date: 17 May 1995 13:29:22 -0800 From: "Richard Bone" Subject: Topological Indices - progr To: "CCL" Subject: Topological Indices - programs? Hello, Computational Chemists: We are looking for programs to calculate molecular "topological indices". The larger the number of indices that are handled, the better. A stand-alone program is most desirable, as is one which reads MDL mol/RD/SD/-files. So far, we know of "MOLCONN-X" and Oxford Molecular's "Tsar", though the latter is not really "stand-alone". As usual, please reply to me and I'll summarise if there is sufficient interest/response level. Richard Bone __________________________________________________________ Richard G. A. Bone, PhD. Computational Chemist Terrapin Technologies, Inc. 750-H Gateway Blvd. South San Francisco CA 94080-7020 USA Tel. +1 (415) 244 9303 FAX +1 (415) 244 9388 E-mail rgab@trpntech.com From boyd@chem.iupui.edu Wed May 17 17:52:26 1995 Received: from indyvax.iupui.edu for boyd@chem.iupui.edu by www.ccl.net (8.6.10/930601.1506) id RAA08651; Wed, 17 May 1995 17:46:38 -0400 Return-receipt-to: Boyd Received: from macgw.chem.iupui.edu by INDYVAX.IUPUI.EDU (PMDF V4.3-13 #5862) id <01HQM7JC1OAO002OKH@INDYVAX.IUPUI.EDU>; Wed, 17 May 1995 16:46:34 -0500 Date: Wed, 17 May 1995 16:44:46 -0500 From: Boyd Subject: similarity To: OSC CCL Message-id: <01HQM7JC2H8I002OKH@INDYVAX.IUPUI.EDU> Content-transfer-encoding: 7BIT To the discussion on molecular similarity initiated by Shep Smithline, I would like add that "Reviews in Computational Chemistry" will have two relevant chapters: "Similarity Searching in Databases of Chemical Structures" by Geoffrey M. Downs and Peter Willett "Three-Dimensional Structure Database Searches" by Andrew C. Good and Jonathan S. Mason Volume 7 will be published in August and released at the American Chemical Society National Meeting in Chicago. Don Donald B. Boyd, Ph.D. Research Professor of Chemistry Co-editor, REVIEWS IN COMPUTATIONAL CHEMISTRY Department of Chemistry Indiana University-Purdue University at Indianapolis 402 North Blackford Street Indianapolis, Indiana 46202-3274 U.S.A. Telephone 317-274-6891 Facsimile 317-274-4701 Internet boyd@chem.iupui.edu World Wide Web URL http://chem.iupui.edu/ From boyd@chem.iupui.edu Wed May 17 17:55:05 1995 Received: from indyvax.iupui.edu for boyd@chem.iupui.edu by www.ccl.net (8.6.10/930601.1506) id RAA08473; Wed, 17 May 1995 17:32:31 -0400 Return-receipt-to: Boyd Received: from macgw.chem.iupui.edu by INDYVAX.IUPUI.EDU (PMDF V4.3-13 #5862) id <01HQM71USWOG004GLU@INDYVAX.IUPUI.EDU>; Wed, 17 May 1995 16:32:28 -0500 Date: Wed, 17 May 1995 16:31:07 -0500 From: Boyd Subject: DFT or not to DFT To: OSC CCL Message-id: <01HQM71UXZXU004GLU@INDYVAX.IUPUI.EDU> Content-transfer-encoding: 7BIT The recent kind remarks by Dr. Gregor Overney (University of New Mexico) about the "Reviews in Computational Chemistry" books are appreciated. > Recently I have received four volumes edited by K.B. Lipkowitz and >D.B. Boyd (Reviews in Computational Chemistry, VCH Publishers, Inc. >New York, 1990-1993). I strongly recommend those books for people >working in this field and for those who want to get started. > However, researchers doing approximations based on DFT will not like >volume four in which DFT seems not to be taken seriously. I hope this >deficiency will be eliminated in a later volume. I would like to assure Dr. Overney and the CCL subscribers that "Reviews in Computational Chemistry" does indeed want to have thorough coverage. Volume 7, which will be released this August at the American Chemical Society National Meeting in Chicago, will have two pertinent chapters: "An Introduction to Density Functional Theory" by Liberto J. Bartolotti and Ken Flurchick "Density Functional Methods in Biomolecular Modeling" by Alain St-Amant Don Donald B. Boyd, Ph.D. Research Professor of Chemistry Co-editor, REVIEWS IN COMPUTATIONAL CHEMISTRY Department of Chemistry Indiana University-Purdue University at Indianapolis 402 North Blackford Street Indianapolis, Indiana 46202-3274 U.S.A. Telephone 317-274-6891 Facsimile 317-274-4701 Internet boyd@chem.iupui.edu World Wide Web URL http://chem.iupui.edu/ From kmay@trans.net Wed May 17 20:07:29 1995 Received: from ns.trans.net for kmay@trans.net by www.ccl.net (8.6.10/930601.1506) id TAA10644; Wed, 17 May 1995 19:57:52 -0400 Received: from by ns.trans.net (4.1/SMI-4.1) id AB11864; Thu, 18 May 95 01:58:00 BST X-Sender: kmay@mailhost Message-Id: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Thu, 18 May 1995 01:58:21 +0200 To: chemistry@ccl.net From: kmay@trans.net (Klaus J.W.May) Subject: LIST: Software over Internet Hi netters! I reently asked for websites, where one can download computational chemistry sw. Akinori Hirashima asked a similar question about chemistry related websites. He got obviously a whole bunch of replies ( on the second run). My harvest was much lesser, but i am sure that at many of those websites one can download sw. Since I got more replies from colleagues who want the compilation than actually entries, I put it on the list : Tadaa, here it is ( pretty poor, isn't it?) B-) ----------------------------------------------------------- http://www.intsim.com/~isigen Program SCULPT from Interactive Simulations. Newest release is on air since one week. A new paradigm in interacting with biomolecules.A "physical model" in the computer. Realtime minimization of torsions, vdw and electrostatics from small organics to largest proteins with ligands, while applying forces ("pulls") to atoms,residues, whole substructures, or molecules. DEMOVERSION FREE Academic price US$ 600/Year. Commercial 2.250/year. Mark Surles, intsim.sales@intsim.com ----------------------------------------------------------- Hypercube has a web page set up now: http://www.hyper.com . It has a variety of things, including access to a working demo version of our HyperChem software (and slide-show demos of ChemPlus and HyperNMR). The demos are also available by ftp from ftp.hyper.com, /pub/demo/* Joel Polowin, Ph.D. Manager, Scientific Support Email to: polowin@hyper.com Hypercube Inc, 419 Phillip St, Waterloo, Ont, Canada N2L 3X2 (519)725-4040 ----------------------------------------------------------- Recently, I announced version 1.0 of my program MolEdit on this forum. Due to the fact that some people had problems to download the files from my FTP-server I uploaded them to ftp.ccl.net. This is also a newer version of MolEdit (1.0.b) that allows to create HPGL-files of the graphical display, so that the pictures can be used in other programs. Additional information is available at the WWW-page: http://pctc.chemie.uni-erlangen.de/~gedeck/moledit.html To download the program either access my FTP-server (in Europe) at pctc.chemie.uni-erlangen.de (131.188.120.2) in directory /pub/moled or use the CCL archives: via anonymous ftp www.ccl.net in /pub/chemistry/software/MS-DOS/MolEdit via gopher gopher gopher.ccl.net 73 software --> MS-DOS --> MolEdit via WWW: http://www.ccl.net/chemistry.html C.C.Archive --> Software --> software archives --> MS-DOS --> MolEdit Peter Gedeck Inst. f. Physikalische Chemie I Egerlandstrasse 3 91058 Erlangen ----------------------------------------------------------- You might want to include the PMD home page in your compilation. It is an experimental parallel molecular dynamics package that is freely available for academic purposes. It features the fast multipole algorithm for the efficient calculation of long range interactions. URL: "http://tincan.bioc.columbia.edu/pmd/" Andreas Windemuth |Columbia University, Department of Biochemistry and Biophysics |630 West 168th St. BB-221 | tel: (212)-305-6884, fax: 6926, NeXTmail |New York, NY 10032 | email: windemut@cumbnd.bioc.columbia.edu ----------------------------------------------------------- LOOK from MAG (molecular Applications Group) has an ftp site from which you can pick up the latest version and then you unlock the software for a trial run with a password obtained via email. Not freely available, but easy enough. The main problem to me is that it takes a LONG time to download 18 MB across the Atlantic. Leif Norskov Novo Nordisk A/S Copenhagen Denmark lnl@novo.dk ----------------------------------------------------------- please check http://www.mdli.com for the availability of ISIS/Draw. This is not a demo version: it is the full application and is FREE for academic and personal home use. Osman F. Guner, PhD, -- Senior Scientist MDL Information Systems, Inc. (510) 895-1313 Osman@mdli.com ----------------------------------------------------------- Klaus J.W. May e-mail: kmay@trans.net May & Partner SciTech Consult Phone/Fax: xx49-(89) 333 569 Mandlstr. 15 Mobile national: 0172-620 38 38 international: xx49-172-620 30 38 D - 80802 Munich GERMANY From shepard@dirac.tcg.anl.gov Wed May 17 20:22:28 1995 Received: from dirac.tcg.anl.gov for shepard@dirac.tcg.anl.gov by www.ccl.net (8.6.10/930601.1506) id UAA10898; Wed, 17 May 1995 20:17:32 -0400 Received: by dirac.tcg.anl.gov (4.1/SMI-4.1) id AA08097; Wed, 17 May 95 19:17:30 CDT Date: Wed, 17 May 95 19:17:30 CDT From: shepard@dirac.tcg.anl.gov (Ron Shepard) Message-Id: <9505180017.AA08097@dirac.tcg.anl.gov> To: CHEMISTRY@ccl.net Subject: CCL:UHF-CISD Cc: shepard@tcg.anl.gov In a previous posting, Mike Frisch wrote: >While we're on the subject, I'd like to clear up a related >misconception which I've seen in the literature. The UCISD >wavefunction in Gaussian is NOT an exact spin eigenfunction. It >consists of all determinants produced by replacing one or two >spin-orbitals which are occupied in the reference determinant with >virtuals. For an open-shell system, this would not lead to a >spin-eigenfunction even if the reference determinant had the same >alpha and beta orbitals, because some determinants which are a double >excitation over spacial orbitals but a triple excitation over >spin-orbitals (because the spins are flipped) are needed to form >proper spin-eigenfunctions. > >To show a simpler example with singles vs. doubles, if the reference >determinant is 1a 1b 2a, then the determinant 1a 2a 3b is considered a >single excitation over spin-orbitals but 1a 2b 3a, which should be >mixed with 1a 2a 3b to form a spin-eigenfunction, is a double >excitation of spin-orbitals from the reference. Of course, both of >these would be included in a CISD wavefunction, but the same issue >occurs with double and triple excitations. Whether a CI wave function ends up being a spin eigenfunction depends on the hamiltonian operator (or approximations thereof), the expansion space, and possibly the iterative procedure used to converge the eigenvalue equation. Mike Frisch was discussing the expansion space above, but perhaps was not clear on a couple of points. First, if a reference determinant is |1a1b2a> with the same spatial orbitals used for different spins (i.e. not UHF), then this reference determinant is a pure doublet (it is a spin eigenfunction of both Sz and S^2 operators). The usual hamiltonian operator (nonrelativistic, no spin-orbit, etc.) commutes with both Sz and S^2, so H|ref> will be also a pure doublet, H^2|ref> will be a pure doublet, and so on for the entire krylov sequence H^n|ref>. This applies also if the vectors are truncated to within some given expansion space. So if the converged CI wave function is computed as a linear combination of such krylov vectors, each of which is a pure doublet, then the final result will also be a pure doublet. This argument holds for any sequence of expansion vectors, truncated or not, provided the operators (and/or methods) used to generate them commute with Sz and S^2. And of course this holds also for any |ref> that is a spin eigenfunction, not just doublets. In the above example, the spin-adapted excitation operator E(3,1) acting on |1a1b2a> will give the combination |1b2a3a>-|1a2a3b>. The other determinant with the same spatial orbital occupations, |1a2b3a>, is a double spinorbital excitation, as noted above. It is not, by itself a pure doublet, but must be combined with the (+) combination of the other two determinants above to become one. The correct combination is 2*|1a2b3a> - |1b2a3a> - |1a2a3b> Some experimentation will show that this spin function cannot be generated from the reference function |1a1b2a> with any kind of spin-adapted SINGLE excitation, but it can be generated with a spin-adapted DOUBLE excitation operator (e.g. the product E(3,2)E(2,1) results in a linear combination of the above two spin functions). As Mike said in his comment, the same thing happens for certain double spatial orbital excitations that end up being triple spin-orbital excitations; these do not interact through the hamiltonian with the original reference function since the hamiltonian is only a spin-adapted double-excitation operator. It is possible to compute CI wave functions that consist only of interacting expansion terms. It does not matter if these wave functions are computed in terms of spin-adapted CSFs, or directly in terms of the determinants. There are, of course, both advantages and disadvantages to this approach in practice. For example, in the COLUMBUS MR-CI code, we use spin-adapted expansion terms (CSFs), but we can easily discard from the expansion space only those noninteracting CSFs that involve excitations from orbitals that are doubly occupied in all reference CSFs. The other possible non-interacting determinants involving partially occupied active orbitals are mixed together with the interacting ones. By the way, the thought might have occurred to someone that there are only two independent linear combinations of the three determinants discussed above. What happened to the other linearly independent combination, namely |1a2b3a> + |1b2a3a> + |1a2a3b>? The answer is that this is also a spin eigenfunction, but it is not a doublet. It is the Ms=1/2 component of the 3-electron quartet state, the high-spin component of which looks like |1a2a3a>. This finally brings up the question of whether you want a wave function that is formed only from the interacting space, or if you want all possible spatial-orbital excitations of a given excitation level included in the expansion space. The answer depends on what you want to do with the wave function. If you want to compare results of various spatial and spin symmetries in the most unbiased way possible, or to compute transition properties between such states, then it is often the case that the best way to achieve this is to ignore the interacting space simplifications. And I won't even try to discuss whether this is ab initio or semiempirical. ;-) -Ron Shepard