From chemistry-request@www.ccl.net Mon Nov 9 02:38:44 1998 Received: from isis.u-strasbg.fr (isis.u-strasbg.fr [130.79.200.1]) by www.ccl.net (8.8.3/8.8.6/OSC/CCL 1.0) with ESMTP id CAA13494 Mon, 9 Nov 1998 02:38:39 -0500 (EST) Received: from amon.u-strasbg.fr (amon.u-strasbg.fr [130.79.200.3]) by isis.u-strasbg.fr (8.9.0/8.9.0) with ESMTP id IAA29216; Mon, 9 Nov 1998 08:38:30 +0100 (MET) Received: from gina (gina.u-strasbg.fr [130.79.79.207]) by amon.u-strasbg.fr (8.9.0/8.9.0) with SMTP id IAA07841; Mon, 9 Nov 1998 08:38:29 +0100 (MET) Received: from gina (localhost [127.0.0.1]) by gina (950413.SGI.8.6.12/950213.SGI.AUTOCF) via SMTP id IAA14812; Mon, 9 Nov 1998 08:42:39 +0100 Sender: wurtz@titus.u-strasbg.fr Message-ID: <36469CEF.1372@igbmc.u-strasbg.fr> Date: Mon, 09 Nov 1998 08:42:39 +0100 From: Jean-Marie WURTZ Organization: IGBMC du CNRS, Labo. de Biologie Structurale X-Mailer: Mozilla 3.01SGoldC-SGI (X11; I; IRIX 6.3 IP32) MIME-Version: 1.0 To: chemistry@www.ccl.net CC: wurtz@titus.u-strasbg.fr Subject: normal modes Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Dear CCLer, Could you indicate me a package able to calculate normal modes or to extract low frequencies of large systems (a few thousand heavy atoms). thanks for your help, -- WURTZ Jean-Marie ________________________________________________________________________ IGBMC E-mail : wurtz@igbmc.u-strasbg.fr Laboratoire de Biologie Structurale Phone : (33) 3 88 65 35 29 Parc d'Innovation Fax : (33) 3 88 65 32 01 1, rue Laurent Fries, BP 163 67404 ILLKIRCH, CU STRASBOURG France ________________________________________________________________________ From chemistry-request@www.ccl.net Mon Nov 9 09:17:29 1998 Received: from rs6.csrsrc.mi.cnr.it (rs6.csrsrc.mi.cnr.it [155.253.3.36]) by www.ccl.net (8.8.3/8.8.6/OSC/CCL 1.0) with ESMTP id JAA15405 Mon, 9 Nov 1998 09:15:27 -0500 (EST) Received: from rs5.csrsrc.mi.cnr.it (rs5.csrsrc.mi.cnr.it [155.253.3.35]) by rs6.csrsrc.mi.cnr.it (8.8.8/8.8.8) with ESMTP id PAA24776 for ; Mon, 9 Nov 1998 15:03:36 +0100 Received: (from dario@localhost) by rs5.csrsrc.mi.cnr.it (8.8.8/8.8.8) id OAA10186 for CHEMISTRY@www.ccl.net; Mon, 9 Nov 1998 14:54:53 +0100 Date: Mon, 9 Nov 1998 14:54:53 +0100 From: Dario Bressanini Message-Id: <199811091354.OAA10186@rs5.csrsrc.mi.cnr.it> To: CHEMISTRY@www.ccl.net Subject: 100 MUST HAVE BOOKS IN CHEMICAL PHYSICS ??? Hello, A few days ago i asked for suggestions about books on reaction dynamics. and received from Gert von Helden the following: R.D. Levine, R.B. Bernstein, "Molecular Reaction Dynamics and Chemical Reactivity", Oxford University Press. G. Scoles (ed.), Atomic and Molecular Beam Methods, Oxford University Press (Vol. 1 and Vol. 2) NOW: I am a researcher of a new Italian University (Universita' dell'Insubria) and I have been assigned the duty of building the 'core' of the new chemistry library. Send me your suggestions of "MUST HAVE", both classic and recents (like the one suggested above for reaction dynamics" I can start with, say, 100 books and for now i am interested in chemical physics/spectroscopy/computational chemistry/modelling/etc.... So, to summarize my request: "WHAT ARE THE 100 MUST HAVE BOOKS IN CHEMICAL PHYSICS ?" I will summarize the replies, and put the list on the Web. Dario Bressanini From chemistry-request@www.ccl.net Mon Nov 9 11:55:08 1998 Received: from lgsx01.lg.ehu.es (lgsx01.lg.ehu.es [158.227.2.34]) by www.ccl.net (8.8.3/8.8.6/OSC/CCL 1.0) with ESMTP id LAA18067 Mon, 9 Nov 1998 11:54:41 -0500 (EST) Received: from lgdx04.lg.ehu.es (lgdx04.lg.ehu.es [158.227.1.36]) by lgsx01.lg.ehu.es (8.8.5/8.8.5) with SMTP id RAA19091 for ; Mon, 9 Nov 1998 17:54:02 +0100 (MET) Date: Mon, 9 Nov 1998 17:54:06 +0100 (MET) From: Pablo Vitoria Garcia To: ccl Subject: AIMPAC problem with >1000 primitives Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Hi, Regarding the problem I found with AIMPAC opening a .wfn file of a big calculation (more than 1000 primitives), and whose origin I wrongly thought was the f functions, this may be of interest to other people. Some AIMPAC programs (EXT94b and cubev, at least) cannot read .wfn files if the calculation was done with more than a thousand primitives, bacause FORMAT line 102 tries to read a I3 integer (<999). It should be changed to: 102 FORMAT (4X,A4,11X,3(I4,16X)) and the problem disappears. Pablo -------------------------------------------------------------------------------- Pablo Vitoria Garcia Departamento de Quimica Inorganica, Facultad de Ciencias Universidad del Pais Vasco (UPV/EHU) Apartado 644, E-48080 Bilbao SPAIN e-mail: qibvigap@lg.ehu.es Phone: +34 4 4647700 Ext. 2450 -------------------------------------------------------------------------------- From chemistry-request@www.ccl.net Mon Nov 9 12:40:56 1998 Received: from arthur.INS.CWRU.Edu (root@arthur.INS.CWRU.Edu [129.22.8.215]) by www.ccl.net (8.8.3/8.8.6/OSC/CCL 1.0) with ESMTP id MAA18452 Mon, 9 Nov 1998 12:40:55 -0500 (EST) Received: from raman2.cwru.edu (chem51374.CHEM.CWRU.Edu [129.22.200.174]) by arthur.INS.CWRU.Edu with SMTP (8.8.8+cwru/CWRU-3.4) id MAA16635; Mon, 9 Nov 1998 12:40:56 -0500 (EST) (from hxt10@po.cwru.edu for ) Message-ID: <000b01be0c10$ecaca4e0$aec81681@raman2.cwru.edu> From: "Gavin Tsai" To: Subject: fitting problem Date: Mon, 9 Nov 1998 12:44:03 -0600 MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_0008_01BE0BDE.A1CF1160" X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 4.72.3110.5 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3110.3 This is a multi-part message in MIME format. ------=_NextPart_000_0008_01BE0BDE.A1CF1160 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Dear all, =20 I try to use Hessian biased singular value decomposition = (HBFF/SVD) method to optimize the potential form and Force Field constants to perform the vibration frequency calculations of some medium size molecules. This method is based on singular value decomposition = (SVD) of Jacobian which is the partial derivatives in force(F) and = Hessian(H) with respect to FF parameters (geometry (c) and force constants (k) = ). The Jacobian matrix has four main componets =20 [ dF/dc, dF/dk ] [ dH/dc, dH/dk], =20 with [dF/dk,dH/dk] being much smaller than [ dF/dc, dH/dc]. My question is as below: =20 Because of the different magnitudes of the components of the = Jacobian matrix, the step change for the carestian coordinates is acceptable, = but is very SMALL for the force contsnts. I multiplied the step change = of force constants by a scaling factor, but it is not very efficient because of their different magnitudes ( e.g. stretching force = constants > bending force constants > cross term force constants).=20 In brief, Is there any efficient method to fit different magnitudes = of parameters (like coordinates amd force constants)? Is there any suggestions on how = to nondimensitionize this Jacobian Matrix? =20 Any suggestions will be welcomed and appreciated. I will = summarize the responses. Thanks! =20 Gavin Tsai =20 ------=_NextPart_000_0008_01BE0BDE.A1CF1160 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable

    Dear all,
   =20
        I try to use Hessian = biased=20 singular value decomposition (HBFF/SVD)
    method to = optimize=20 the potential form and Force Field constants to
    = perform=20 the vibration frequency calculations of some medium = size
   =20 molecules.  This method is based on singular value decomposition=20 (SVD)
    of Jacobian which is the partial derivatives = in=20 force(F) and Hessian(H)
    with respect to FF = parameters=20 (geometry (c) and force constants (k) ).
    The = Jacobian=20 matrix has four main componets
    =
    [=20 dF/dc, dF/dk ]
    [ dH/dc, = dH/dk],
   =20
    with [dF/dk,dH/dk] being much smaller than [ = dF/dc,=20 dH/dc].

      My question is as=20 below:
   
       = Because of=20 the different magnitudes of the components of the = Jacobian
   =20 matrix, the step change for the carestian coordinates is acceptable,=20 but
    is very SMALL for the force contsnts.  I=20 multiplied the step change of
    force constants by a = scaling=20 factor, but it is not very efficient
    because of = their=20 different magnitudes ( e.g. stretching force = constants
   =20 > bending force constants > cross term force = constants). 
 
     In brief,=20 Is there any efficient method to fit different magnitudes  of=20 parameters
(like coordinates amd = force=20 constants)? Is there any suggestions on how to = nondimensitionize=20 this Jacobian Matrix?
   =20
       Any suggestions will be = welcomed and=20 appreciated.  I will summarize the responses. = Thanks!
   =20
    Gavin Tsai
   =20

------=_NextPart_000_0008_01BE0BDE.A1CF1160-- From chemistry-request@www.ccl.net Mon Nov 9 16:16:29 1998 Received: from dirac.cnrs-orleans.fr (dirac.cnrs-orleans.fr [163.9.6.67]) by www.ccl.net (8.8.3/8.8.6/OSC/CCL 1.0) with ESMTP id QAA24635 Mon, 9 Nov 1998 16:16:28 -0500 (EST) Received: (from hinsen@localhost) by dirac.cnrs-orleans.fr (AIX4.3/UCB 8.8.8/8.7) id WAA16180; Mon, 9 Nov 1998 22:13:08 +0100 Date: Mon, 9 Nov 1998 22:13:08 +0100 Message-Id: <199811092113.WAA16180@dirac.cnrs-orleans.fr> X-Authentication-Warning: dirac.cnrs-orleans.fr: hinsen set sender to hinsen@cnrs-orleans.fr using -f From: Konrad Hinsen To: wurtz@titus.u-strasbg.fr CC: chemistry@www.ccl.net In-reply-to: <36469CEF.1372@igbmc.u-strasbg.fr> (message from Jean-Marie WURTZ on Mon, 09 Nov 1998 08:42:39 +0100) Subject: Re: CCL:normal modes References: <36469CEF.1372@igbmc.u-strasbg.fr> > Could you indicate me a package able to calculate normal modes or to > extract low frequencies of large systems (a few thousand heavy atoms). > thanks for your help, Some very efficient methods for dealing with such problems are described in the article K. Hinsen Analysis of domain motions by approximate normal mode calculations Proteins 33(3), 417-429 (1998) They have been implemented in the program DomainFinder, available for free at http://dirac.cnrs-orleans.fr/DomainFinder/ together with analysis techniques which are specific to proteins. If you get the "developer release" of DomainFinder, you can also use these techniques individually for whatever purpose you like. -- ------------------------------------------------------------------------------- Konrad Hinsen | E-Mail: hinsen@cnrs-orleans.fr Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.55.69 Rue Charles Sadron | Fax: +33-2.38.63.15.17 45071 Orleans Cedex 2 | Deutsch/Esperanto/English/ France | Nederlands/Francais ------------------------------------------------------------------------------- From chemistry-request@www.ccl.net Mon Nov 9 23:04:11 1998 Received: from darkwing.uoregon.edu (genghis@darkwing.uoregon.edu [128.223.142.13]) by www.ccl.net (8.8.3/8.8.6/OSC/CCL 1.0) with ESMTP id XAA27689 Mon, 9 Nov 1998 23:04:10 -0500 (EST) Received: from localhost (genghis@localhost) by darkwing.uoregon.edu (8.8.8/8.8.8) with SMTP id UAA27394 for ; Mon, 9 Nov 1998 20:04:10 -0800 (PST) Date: Mon, 9 Nov 1998 20:04:10 -0800 (PST) From: "Dale A. Braden" To: cclpost Subject: wave function stability question Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Dear CCL, I'm doing DFT calculations on an open-shell transition metal compound using G94. I saw that 3 of the virtual orbitals had negative eigenvalues: ..... Alpha occ. eigenvalues -0.20751 -0.14471 Alpha virt. eigenvalues -0.03482 -0.01181 -0.01122 0.01779 0.03974 ..... which I thought warned that the SCF had converged to an excited state, so I performed a wave function stability calculation. This reported that the wave function was stable "under the perturbations considered", with an eigenvalue of 0.0119 for the "stability matrix". This was confusing, so I have three questions: 1) Am I correct that the stability calculation is NOT guaranteed to recognize an excited state? 2) Are negative eigenvalues for virtual orbitals a SURE sign that an excited state has been obtained? 3) In case I really still do have an excited state wave function, is the only way to direct it towards the ground state to guess which orbitals should be occupied, and then use guess=(read,alter) to fix their occupations? (I'd like to avoid hit-or-miss strategies!) Best wishes to all, Dale Braden Department of Chemistry University of Oregon Eugene, OR 97403-1253 genghis@darkwing.uoregon.edu