From owner-chemistry@ccl.net Tue Aug 29 08:47:00 2006 From: "saikat Mukhopadhyay saikatrel=-=gmail.com" To: CCL Subject: CCL: DFT inputs Message-Id: <-32446-060829082948-9312-7wWqb9xkktPyXWqf4AqiQw^server.ccl.net> X-Original-From: "saikat Mukhopadhyay" Date: Tue, 29 Aug 2006 08:29:46 -0400 Sent to CCL by: "saikat Mukhopadhyay" [saikatrel^gmail.com] 576 Dear users, First of all I would like to inform that I am a beginner.I want to know ' how one can run a DFT calculation of a dimer( water dimer -say) with the input MO-Vectors achieved from the previous runs with the monomers(water)?' actually I want to have a UDFT run where the input MOVECS will be those achieved from the monomer runs and get back the MO-VECS for the dimer system. Nowhere in the manual it is described what would be the pattern of the input. Can anybody help me out here? Thanks in advance - Sincerely, Saikat Mukhopadhyay Pune India From owner-chemistry@ccl.net Tue Aug 29 09:58:01 2006 From: "Xiao xiaolzhao .. gmail.com" To: CCL Subject: CCL:G: Gaussian formchk err Message-Id: <-32448-060829051158-19440-PtGqq847BwM8mvu7ADtchQ[a]server.ccl.net> X-Original-From: Xiao Content-Type: multipart/alternative; boundary="----=_Part_4515_4850615.1156840345568" Date: Tue, 29 Aug 2006 16:32:25 +0800 MIME-Version: 1.0 Sent to CCL by: Xiao [xiaolzhao[a]gmail.com] ------=_Part_4515_4850615.1156840345568 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Content-Disposition: inline Hellow, everyone, I met a problem when I convert a gaussian checkpoint file to formatted form use "formchk". In the "Cartesian Gradient" section appears "-nan". ---------------------------- " 1.67613539E-09 1.53165212E-04 1.35253230E-04 Cartesian Gradient R N= 27 -nan -1.80112017E-03 5.10172364E-07 -3.91095638E-04 - 4.07779150E-04 -7.37741807E-08 -1.06821156E-03 -2.25327180E-04 1.18551596E-04 - 1.06800804E-03 " ----------------------------- the gaussian job was terminated normally. How to resolve this problem? Thank you for your kind help! ------=_Part_4515_4850615.1156840345568 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Hellow, everyone,
    I met a problem when I convert a gaussian checkpoint file to formatted form use "formchk". In the "Cartesian Gradient" section appears  "-nan".
----------------------------
"
  1.67613539E-09  1.53165212E-04  1.35253230E-04
Cartesian Gradient                         R   N=          27
            -nan -1.80112017E-03  5.10172364E-07 -3.91095638E-04 -4.07779150E-04
 -7.37741807E-08 -1.06821156E-03 -2.25327180E-04  1.18551596E-04 -1.06800804E-03
"
-----------------------------
 the gaussian job was terminated normally.

How to resolve this problem? Thank you for your kind help!
------=_Part_4515_4850615.1156840345568-- From owner-chemistry@ccl.net Tue Aug 29 16:06:00 2006 From: "Flavio Contreras ffcontreras_yyz^^^yahoo.es" To: CCL Subject: CCL: DIIS and gaps Homo-Lumo Message-Id: <-32449-060829130212-14243-S1sQQce0T8Zp8BRk/PXGFg++server.ccl.net> X-Original-From: "Flavio Contreras" Date: Tue, 29 Aug 2006 13:02:12 -0400 Sent to CCL by: "Flavio Contreras" [ffcontreras_yyz++yahoo.es] Dear CCLers I would like to know why I can not use the direct inversion in the iterative subspace algorithm, when the gaps Homo-Lumo is less than 0.5 eV. If I used it, taking in mind huge systems, the eigenvalues would be incorrect or not? Im going to aprecciate every reference. Thanks in advance. Flavio Contreras. From owner-chemistry@ccl.net Tue Aug 29 17:26:01 2006 From: "Jan Labanowski janl[A]speakeasy.net" To: CCL Subject: CCL: Check MMCC Results Newsletter in CCL Archives Message-Id: <-32450-060829172048-30595-RQBPSkbaOlh75Erif4aYhg ~~ server.ccl.net> X-Original-From: "Jan Labanowski" Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="iso-8859-1" Date: Tue, 29 Aug 2006 21:20:46 +0000 MIME-Version: 1.0 Sent to CCL by: "Jan Labanowski" [janl%a%speakeasy.net] I would like to thank Dr. Nageswar for a valuable contribution to CCL archives, namely: Molecular Modeling & Computational Chemistry Results (MMCC Results) Newsletter "MMCC Results" contains scientific literature reviews of 30 major journals by an extensive network of international contributors. The latest issue and archival issues starting from 2005 are now available on CCL in: http://server.ccl.net/cca/documents/MMCC_Results/ The newsletter is in the PDF format and contains short abstracts of important papers gathered from 30 leading journals. Sections cover papers dealing with molecular modeling and computational chemistry applications, namely: drug design, biopolymers, polymers, materials, surface, solvation, methodology, etc. Thank you again, Dr. Nageswar for sharing this useful contribution with CCL community. Jan Labanowski CCL Maintainer PLUG: P.S. Support CCL (please...). Did you upload useful stuff to CCL? Announced a new job opening in your group? An interesting conference? Placed a logo of your organization on CCL page ("Adopted the page" {:-)}) bought a DVD with CCL Archives (we had a tornado warning here yesterday, so be afraid... be very afraid...), or posted a nice message on the CCL? Just go to CCL Home page ( http://www.ccl.net ) and learn how you can help. Vacations is over... Get busy... From owner-chemistry@ccl.net Tue Aug 29 23:36:01 2006 From: "Konstantin Kudin konstantin_kudin]=[yahoo.com" To: CCL Subject: CCL: DIIS and gaps Homo-Lumo Message-Id: <-32451-060829223807-28566-xMAc073BLlt/JhQDcnp/Gw~~server.ccl.net> X-Original-From: Konstantin Kudin Content-Transfer-Encoding: 8bit Content-Type: text/plain; charset=iso-8859-1 Date: Tue, 29 Aug 2006 18:37:58 -0700 (PDT) MIME-Version: 1.0 Sent to CCL by: Konstantin Kudin [konstantin_kudin##yahoo.com] Dear Flavio, Your question as posed lacks some important qualifiers. Basically, do you want to use the DIIS (direct inversion of the iterative subspace) for the SCF (self-consistent field), or some other type of convergence? If it is the SCF convergence you are talking about, certainly there is no technical reason why you cannot use the DIIS for the cases you have mentioned. You certainly should be able to. An entirely different question if such an approach would be effective for such systems. The answer to this question is no. For systems with small gaps you either get SCF convergence that is very slow, or, in spatially large systems you might get charge waves from cycle to cycle during the SCF leading to divergence. Note that the plane-wave codes actually have to deal with small gaps a lot more often, so you may want to check the approaches used by those codes for the SCF convergence. For localized basis sets, you could check out this paper: http://dx.doi.org/10.1063/1.1470195 In any event, if you want a better answer, you will need a more pointed question. Regards, Kostya --- "Flavio Contreras ffcontreras_yyz^^^yahoo.es" wrote: > Sent to CCL by: "Flavio Contreras" [ffcontreras_yyz++yahoo.es] > Dear CCLers > > I would like to know why I can not use the direct inversion in the > iterative subspace algorithm, when the gaps Homo-Lumo is less than > 0.5 eV. If I used it, taking in mind huge systems, the eigenvalues > would be incorrect or not? > > Im going to aprecciate every reference. > > Thanks in advance. > > Flavio Contreras. > > > > -= This is automatically added to each message by the mailing script > =- > To recover the email address of the author of the message, please > change> Conferences: > http://server.ccl.net/chemistry/announcements/conferences/ > > Search Messages: http://www.ccl.net/htdig (login: ccl, Password: > search)> > > > __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com