From owner-chemistry@ccl.net Wed Feb 20 03:17:01 2008 From: "Aleksey Kuznetsov AlexKuznetsov2007=-=yandex.ru" To: CCL Subject: CCL:G: Strange behaviour of SCF energies using TURBOMOLE for single-point calculations of Pd atom and dimer on TiO2(110) surface Message-Id: <-36327-080220031512-4497-oDhtiHHFVFS5E4ehgFog5Q[]server.ccl.net> X-Original-From: Aleksey Kuznetsov Content-Transfer-Encoding: 8bit Content-Type: text/plain; charset=KOI8-R Date: Wed, 20 Feb 2008 11:14:51 +0300 MIME-Version: 1.0 Sent to CCL by: Aleksey Kuznetsov [AlexKuznetsov2007 ~~ yandex.ru] Dear CCL Subscribers, I'd like to ask you once again for a piece of advice on the system containing Pd atom/dimer on TiO2(110) surface; this time I try to do single-point calculation using TURBOMOLE with the following conditions: Ti17O42H16 model, only Pd atoms in dimer were allowed to relax, for Pd atom I did single-point calculation, def-SV(P) basis set for all atoms, def-ecp ECP for Pd, ecp-10-mdf ECP for Ti, and ecp-2-sdf ECP for O; b3-lyp_Gaussian functional; closed shell shift 0.6; damp options are as follows: start=1.000 step=0.050 min=0.050. Here I encounter with something strange: after some iteration the SCF energy begins to fluctuate wildly, ranging from, say, -1437.374 till 1408.485 A.U. Do I suppose correctly that there is some incompatibility among used ECPs here which can cause such a strange behavior? Thank you very much, With kind regards, Aleksey Kuznetsov. ------------------------------------------- Dr. Aleksey Kuznetsov Cherry L. Emerson Center for Scientific Computation Emory University 1515 Dickey Drive Atlanta, GA 30322 USA Phone: (404)727-2381 Email: akuznets{=}euch4e.chem.emory.edu ------------------------------------------- -------- Завершение пересылаемого сообщения -------- -- ------------------------------------------- Dr. Aleksey Kuznetsov Cherry L. Emerson Center for Scientific Computation Emory University 1515 Dickey Drive Atlanta, GA 30322 USA Phone: (404)727-2381 Email: akuznets{=}euch4e.chem.emory.edu ------------------------------------------- From owner-chemistry@ccl.net Wed Feb 20 09:00:01 2008 From: "Nicolas Staelens nicolas.staelens]^[fundp.ac.be" To: CCL Subject: CCL: Ag and Ag+ params for Amber Message-Id: <-36328-080220085735-777-Ph8lMWaIPUMF4XZgN1PYMA---server.ccl.net> X-Original-From: "Nicolas Staelens" Date: Wed, 20 Feb 2008 08:57:31 -0500 Sent to CCL by: "Nicolas Staelens" [nicolas.staelens{:}fundp.ac.be] Dear CCL subscribers, I am interested in MD simulations of DNA in presence of Ag and Ag+. I would like to use the Amber force field to do so, but I could not find parameters for Ag or Ag+ for this force field anywhere. Could someone send me a link to a reference mentioning it? Thanks in advance Nicolas Staelens nicolas.staelens*fundp.ac.be Ph. D. Student PCI Laboratory University of Namur From owner-chemistry@ccl.net Wed Feb 20 09:35:00 2008 From: "Erdem YILDIRIM ekyildirim.^-^.gmail.com" To: CCL Subject: CCL: Gaussian03 error Message-Id: <-36329-080220042231-987-8gKjdetHRhF7uLUNniHBXA^-^server.ccl.net> X-Original-From: "Erdem YILDIRIM" Date: Wed, 20 Feb 2008 04:22:26 -0500 Sent to CCL by: "Erdem YILDIRIM" [ekyildirim*gmail.com] Hello Can anyone help me wtih the "Gradient out of range" error message in Gausian03. I am using B3LYP/LANL2DZ basis set and trying to do optimization and frequency calculatios. What shoul I do get rid of from this error. Thank you for your helps. Erdem From owner-chemistry@ccl.net Wed Feb 20 11:10:00 2008 From: "Uwe Huniar uwe.huniar^cosmologic.de" To: CCL Subject: CCL:G: Strange behaviour of SCF energies using TURBOMOLE for single-point calculations of Pd atom and dimer on TiO2(110) surface Message-Id: <-36330-080220110512-26164-ofpkhNDYzxiXdESuqv+NBA{=}server.ccl.net> X-Original-From: Uwe Huniar Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=KOI8-R; format=flowed Date: Wed, 20 Feb 2008 16:05:27 +0100 MIME-Version: 1.0 Sent to CCL by: Uwe Huniar [uwe.huniar__cosmologic.de] Hi, the default settings for damping and orbital shifts are well suited for systems with a gap that is not too small. In your case the settings should be changed, try: $scfdamp start=6.000 step=0.050 min=0.2 $scforbitalshift automatic 0.5 Just start dscf and redirect its output in a file like: dscf > dscf.out & during the calculation you can use 'cgnce dscf.out' if gnuplot is installed on your system. You will get a graphics with the SCF convergence. If the energy starts to oscillate, stop the calculation, raise DIIS damping ($scfdamp start value) and start from scratch, i.e. do not use the new molecular orbitals. I always save the original input, reuse it is such cases and set $scfiterlimit to a number that is lower than the one where oscillation started. There are several possible reasons for energy oscillations: 1. too high orbital shift together with low DIIS damping 2. too high DIIS damping (difference between several subsequent SCF steps is too small to get a numerical reasonable guess for the next iteration) 3. oscillation in occupation - if two orbitals alternate during the calculation. In that case I would recommend to do an unrestricted calculation and use $fermi (see manual) to get a reasonable occupation automatically. If DIIS is the reason, it often helps to raise damping, limit $scfiterlimit to a number that is smaller than the one where energy oscillations occurs, run dscf, remove the DIIS files afterwards (errvec, oldfock, fock) and restart dscf. For extremely difficult cases that has to be repeated a few times until convergence is reached. Turbomole is very often used for metal clusters, and playing with $scfdamp, $scforbitalshift, $fermi, and $scfiterlimit will solve your problem. However, at least for clusters, there is no ideal combination of damping and shifting that will always work for all systems... Regards, Uwe > I'd like to ask you once again for a piece of advice on the system containing > Pd atom/dimer on TiO2(110) surface; this time I try to do single-point > calculation using TURBOMOLE with the following conditions: Ti17O42H16 model, > only Pd atoms in dimer were allowed to relax, for Pd atom I did single-point > calculation, def-SV(P) basis set for all atoms, def-ecp ECP for Pd, > ecp-10-mdf ECP for Ti, and ecp-2-sdf ECP for O; b3-lyp_Gaussian functional; > closed shell shift 0.6; damp options are as follows: start=1.000 step=0.050 > min=0.050. Here I encounter with something strange: after some iteration the > SCF energy begins to fluctuate wildly, ranging from, say, -1437.374 till > 1408.485 A.U. Do I suppose correctly that there is some incompatibility among > used ECPs here which can cause such a strange behavior? From owner-chemistry@ccl.net Wed Feb 20 11:44:01 2008 From: "andrea spitaleri spitaleri.andrea[-]hsr.it" To: CCL Subject: CCL: Ag and Ag+ params for Amber Message-Id: <-36331-080220111310-27824-EKT7UR9pzV3Aqt1aNgAwMg===server.ccl.net> X-Original-From: andrea spitaleri Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=ISO-8859-1 Date: Wed, 20 Feb 2008 16:04:31 +0100 MIME-Version: 1.0 Sent to CCL by: andrea spitaleri [spitaleri.andrea.:.hsr.it] Hi, have look to Aqvist and/or Babu papers: J.phys.chem. A, 2006, 110, 691-9 & B 103, 1999,7958 & B 111, 2007, 2219-24 J. phys. chem. 1990, 94, 8021-24 in these papers, the authors have developed a way to calculate the parameters for ions. However, in amber you may have look to http://pharmacy.man.ac.uk/amber/ I hope this helps, Regards andrea Nicolas Staelens nicolas.staelens]^[fundp.ac.be wrote: > Sent to CCL by: "Nicolas Staelens" [nicolas.staelens{:}fundp.ac.be] > Dear CCL subscribers, > > I am interested in MD simulations of DNA in presence of Ag and Ag+. I would like to use the Amber force field to do so, but I could not find parameters for Ag or Ag+ for this force field anywhere. > Could someone send me a link to a reference mentioning it? > > Thanks in advance > > Nicolas Staelens > nicolas.staelens*fundp.ac.be > Ph. D. Student > PCI Laboratory > University of Namur> > > -- ------------------------------- Andrea Spitaleri PhD Dulbecco Telethon Institute c/o DIBIT Scientific Institute Biomolecular NMR, 1B4 Via Olgettina 58 20132 Milano (Italy) http://biomolecularnmr.ihsr.dom/ Tel: 0039-0226434348/5622/3497/4922 Fax: 0039-0226434153 ------------------------------- From owner-chemistry@ccl.net Wed Feb 20 13:05:00 2008 From: "Roger Kevin Robinson r.robinson{=}imperial.ac.uk" To: CCL Subject: CCL:G: What QM-Code/Basis-Set for looking at reactions on a metal surface Message-Id: <-36332-080220130215-4615-FJWZI1aTdbEjVGPzaO0pJQ . server.ccl.net> X-Original-From: Roger Kevin Robinson Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=KOI8-R; format=flowed Date: Wed, 20 Feb 2008 17:18:46 +0000 MIME-Version: 1.0 Sent to CCL by: Roger Kevin Robinson [r.robinson]*[imperial.ac.uk] Dear All, Previously I've been using Gaussian to look at reaction paths for organic molecules in the gas phase. A colleague of mine is interested in Activation Energies for the rearrangements of organic molecules whist bound to a metal surface. I did some quick calculations for him for the respective reactions in the gas phase, with out considering the metal, to get a basic starting point. They now need some more accurate numbers. I've never done any calculations on a metal surface before so I'm a bit lost. First of all I tried the same attaching the organic molecule to a single platinum atom. The only basis set I could find for which the Atomic number of platinum was not out of range were lanl2mb and lanl2dz. I also tried building a platinum cluster and minimising with b3lyp/lanl2mb and b3lyp/lanl2dz this is taking a long time. Obliviously UFF MM can also be used. I then decided to search the literature, most of the QM work on metal surface seems to be using different QM codes from Gaussian such as Castep. Although a lot of the work seems to be using DFT functionals, such as B3LYP, which I am familiar with, the Basis sets are unknown to me. Is it feasible to study reactions on a surface using Gaussian ? Should I start to look at another QM code ? Or should I be looking for other basis sets to use in Gaussian ? This is not my main work so I dont want to be spending too much time on it. Also if I am working with a 9 atom platinum cluster approximately how long would I be looking at for calculations times ? Any other help or advice appreciated. Thanks Roger From owner-chemistry@ccl.net Wed Feb 20 13:40:01 2008 From: "Wendy Warr wendy() warr.com" To: CCL Subject: CCL: ACS report Message-Id: <-36333-080220130803-6592-mDOhVA78g9bRXxIRDFDMQw|*|server.ccl.net> X-Original-From: "Wendy Warr" Date: Wed, 20 Feb 2008 13:07:59 -0500 Sent to CCL by: "Wendy Warr" [wendy{=}warr.com] Chemical Information and Computation 2007, Number Two Report on the 234th ACS National Meeting and Exposition, Boston, MA, August 19-23, 2007 The contents list is on the Web at http://www.warr.com/morepubs.html. The report includes: news from more than 70 companies (including some not exhibiting) condensed press releases for the last six months of 2007 accurate transcripts of 29 presentations, complete with color figures and literature references, with the following themes: o drug reprofiling (10 papers) o Herman Skolnik Award Symposium, honoring Robert Pearlman (9 papers) o Chemical Abstracts Service 100th Anniversary (5 papers) o cheminformatics techniques in bioinformatics-related applications (2 papers) o open meetings of the ACS committees on ACS Publications and CAS o tautomer generation (poster) o NLP and substructure search for literature mining (poster). Contact me by email for further details. Wendy Dr. Wendy A. Warr Wendy Warr & Associates 6 Berwick Court, Holmes Chapel Cheshire, CW4 7HZ, England Tel./fax +44 (0)1477 533837 wendy() warr.com http://www.warr.com