From owner-chemistry@ccl.net Tue Nov 20 01:08:00 2007 From: "Arvydas Tamulis tamulis(0)mserv.itpa.lt" To: CCL Subject: CCL: 6-31+G(d) in GAMESS-US Message-Id: <-35643-071120010611-30461-6wh0IGZ9or7AyzimESmmTg^server.ccl.net> X-Original-From: Arvydas Tamulis Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed Date: Tue, 20 Nov 2007 08:05:55 +0200 (EET) MIME-Version: 1.0 Sent to CCL by: Arvydas Tamulis [tamulis*o*mserv.itpa.lt] Dear Cplleagues, How to write in the input file of GAMESS-US request for 6-31+G(d) basis set? I am having now the request only for 6-31G: $CONTRL SCFTYP=RHF RUNTYP=energy MAXIT=200 NOSYM=1 $END $CONTRL ICHARG=0 MULT=1 $END $BASIS GBASIS=n31 ngauss=6 $END $DFT DFTTYP=b3lyp $END Cheers, Arvydas From owner-chemistry@ccl.net Tue Nov 20 02:18:01 2007 From: "Cheng, Tiejun ccl**mail.sioc.ac.cn" To: CCL Subject: CCL: Looking for a structure to name conversion software Message-Id: <-35644-071119212602-10539-7dTr0ypoI29Ksm/XxS3uXg_._server.ccl.net> X-Original-From: "Cheng, Tiejun" Content-Type: multipart/alternative; boundary="----=_Part_14156_28427136.1195522211471" Date: Tue, 20 Nov 2007 09:30:11 +0800 MIME-Version: 1.0 Sent to CCL by: "Cheng, Tiejun" [ccl{=}mail.sioc.ac.cn] ------=_Part_14156_28427136.1195522211471 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Hi, Saravanan I don't whether there is a free one, but ChemDraw can absolutely do this job. On 11/16/07, drmsaravanan=aim.com wrote: > > > Hi All, > > Is there any tool to convert 1000 structures to name at a stretch? it > would be better if it is available freely. > > regards, > > M. Saravanan > ------------------------------ > *Check Out the new free AIM(R) Mail*-- Unlimited storage and industry-leading spam and email virus protection. > > ------=_Part_14156_28427136.1195522211471 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline
Hi, Saravanan
 
I don't whether there is a free one, but ChemDraw can absolutely do this job.

 
On 11/16/07, drmsaravanan=aim.com <owner-chemistry a ccl.net> wrote:

Hi All,

Is there any tool  to convert 1000 structures to name at a stretch? it would be better if it is available freely.

regards,

M. Saravanan
Check Out the new free AIM(R) Mail -- Unlimited storage and industry-leading spam and email virus protection.
 

------=_Part_14156_28427136.1195522211471-- From owner-chemistry@ccl.net Tue Nov 20 03:37:01 2007 From: "Sebastian Rohrer sebastian.rohrer**mail.uni-wuerzburg.de" To: CCL Subject: CCL: Looking for a structure to name conversion software Message-Id: <-35645-071120033523-1152-nBkAMdlhYIDR4+1e5eSIDQ a server.ccl.net> X-Original-From: Sebastian Rohrer Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=ISO-8859-1; format=flowed Date: Tue, 20 Nov 2007 09:05:17 +0100 MIME-Version: 1.0 Sent to CCL by: Sebastian Rohrer [sebastian.rohrer-x-mail.uni-wuerzburg.de] Hi, OpenEye Lexichem (http://www.eyesopen.com/products/toolkits/lexichem.html) does the job. It's free to academics. Sebastian Cheng, Tiejun ccl**mail.sioc.ac.cn wrote: > Hi, Saravanan > > I don't whether there is a free one, but ChemDraw can absolutely do > this job. > > > On 11/16/07, *drmsaravanan=aim.com * > > wrote: > > > Hi All, > > Is there any tool to convert 1000 structures to name at a > stretch? it would be better if it is available freely. > > regards, > > M. Saravanan > ------------------------------------------------------------------------ > *Check Out the new free AIM(R) Mail* > > -- Unlimited storage and industry-leading spam and email virus > protection. > > > From owner-chemistry@ccl.net Tue Nov 20 04:12:01 2007 From: "James Robinson James.Robinson:prosonix.co.uk" To: CCL Subject: CCL:G: PCM thermodynamics Message-Id: <-35646-071119143502-4975-qBz0ys9oMKOkHPBYessrUg]-[server.ccl.net> X-Original-From: "James Robinson" Content-class: urn:content-classes:message Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="us-ascii" Date: Mon, 19 Nov 2007 19:01:09 -0000 MIME-Version: 1.0 Sent to CCL by: "James Robinson" [James.Robinson!=!prosonix.co.uk] Dear Tom, What is it you want to know? What information do you want from your calculation? If indeed, you want to compare free energies in solution you should reoptimise with PCM solvation and perform a frequency calculation in solution. I also recommend using tight, or indeed, verytight convergence for SCF cycles when using solvation.=20 Dr James J Robinson Senior Scientist Prosonix Ltd Magdalen Centre Robert Robinson Avenue Oxford Science Park Oxford, Oxon. OX4 4GA, UK. +44(0) 1865 784243 http://www.prosonix.co.uk =20 P Please consider the environment - do you really need to print this email? =20 =20 -----Original Message----- > From: owner-chemistry .. ccl.net [mailto:owner-chemistry .. ccl.net]=20 Sent: 19 November 2007 08:35 To: James Robinson Subject: CCL:G: PCM thermodynamics Sent to CCL by: "Tom de Greef" [t.f.a.d.greef+*+tue.nl] Dear all, I have read several post how to use PCM calculations to correct for solvent effects. I have the following problem: I have optimised two conformations of the same structure in the gas-phase and subsequently done a frequency calculation on both conformations. This has revealed to me the difference of deltaG in the gas phase. Now I want to check the difference in deltaG in the solvent (chloroform). The way I have done this is doing a single point calculation on the gas phase optimised structures using the pcm model in Gaussian03 supplied with the SCFVAC command.=20 From owner-chemistry@ccl.net Tue Nov 20 04:52:00 2007 From: "Daniel Bonniot dbonniot[-]chemaxon.com" To: CCL Subject: CCL: Looking for a structure to name conversion software Message-Id: <-35647-071120044908-10786-MnBHrtrGKrx3ioJ2o5B+3Q|-|server.ccl.net> X-Original-From: "Daniel Bonniot" Date: Tue, 20 Nov 2007 04:49:04 -0500 Sent to CCL by: "Daniel Bonniot" [dbonniot|,|chemaxon.com] > Is there any tool to convert 1000 structures to name at a stretch? it would be better if it is available freely. You can use ChemAxon's IUPAC name generator: http://www.chemaxon.com/marvin/doc/user/iupacnaming.html For naming 1000 structures or more, you will want to look at the batch tools in Marvin Beans (cxcalc, molconvert): http://www.chemaxon.com/marvin/index.html Or at Instant JChem: http://www.chemaxon.com/instantjchem/ Batch naming is free for academic users and non-commercial web sites. Follow this link if you would like to subscribe for the free usage or need more details: http://www.chemaxon.com/licensing.html Best regards, Daniel Bonniot From owner-chemistry@ccl.net Tue Nov 20 05:42:01 2007 From: "Rudolf Herrmann rudolf.herrmann-$-physik.uni-augsburg.de" To: CCL Subject: CCL: 6-31+G(d) in GAMESS-US Message-Id: <-35648-071120043243-31988-ydYI9r1F1OvkpDM+uEj6Qg[*]server.ccl.net> X-Original-From: "Rudolf Herrmann" Content-Transfer-Encoding: 8bit Content-Type: text/plain;charset=iso-8859-1 Date: Tue, 20 Nov 2007 09:57:32 +0100 (CET) MIME-Version: 1.0 Sent to CCL by: "Rudolf Herrmann" [rudolf.herrmann|,|physik.uni-augsburg.de] Hello, DIFFSP=.true. adds diffuse basis function to heavy elements NDFUNC=1 adds a polarisation function to heavy elements (you may add two or three with NDFUNC=2 or NDFUNC=3). It's all in the INPUT.DOC file of the documentation, and in more detail in the REFS.DOC. Ciao Rudolf. > > Sent to CCL by: Arvydas Tamulis [tamulis*o*mserv.itpa.lt] > Dear Cplleagues, > > How to write in the input file of GAMESS-US request for > 6-31+G(d) basis set? > > I am having now the request only for 6-31G: > $CONTRL SCFTYP=RHF RUNTYP=energy MAXIT=200 NOSYM=1 $END > $CONTRL ICHARG=0 MULT=1 $END > $BASIS GBASIS=n31 ngauss=6 $END > $DFT DFTTYP=b3lyp $END > > Cheers, > Arvydas > From owner-chemistry@ccl.net Tue Nov 20 06:01:00 2007 From: "akef afaneh akef_afnh!^!yahoo.com" To: CCL Subject: CCL: 6-31+G(d) in GAMESS-US Message-Id: <-35649-071120054544-23757-lgz7wN/Ir5jVnLIpSYDduw[]server.ccl.net> X-Original-From: akef afaneh Content-Transfer-Encoding: 8bit Content-Type: multipart/alternative; boundary="0-1071979424-1195555513=:63553" Date: Tue, 20 Nov 2007 02:45:13 -0800 (PST) MIME-Version: 1.0 Sent to CCL by: akef afaneh [akef_afnh:yahoo.com] --0-1071979424-1195555513=:63553 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit You can use the following link: http://classic.chem.msu.su/gran/gamess/ then manuals :-) Good Luck "Arvydas Tamulis tamulis(0)mserv.itpa.lt" wrote: Sent to CCL by: Arvydas Tamulis [tamulis*o*mserv.itpa.lt] Dear Cplleagues, How to write in the input file of GAMESS-US request for 6-31+G(d) basis set? I am having now the request only for 6-31G: $CONTRL SCFTYP=RHF RUNTYP=energy MAXIT=200 NOSYM=1 $END $CONTRL ICHARG=0 MULT=1 $END $BASIS GBASIS=n31 ngauss=6 $END $DFT DFTTYP=b3lyp $END Cheers, Arvydashttp://www.ccl.net/cgi-bin/ccl/send_ccl_messagehttp://www.ccl.net/chemistry/sub_unsub.shtmlhttp://www.ccl.net/spammers.txt--------------------------------- Be a better pen pal. Text or chat with friends inside Yahoo! Mail. See how. --0-1071979424-1195555513=:63553 Content-Type: text/html; charset=iso-8859-1 Content-Transfer-Encoding: 8bit
You can use the following link:
 
then manuals :-)
Good Luck

"Arvydas Tamulis tamulis(0)mserv.itpa.lt" <owner-chemistry++ccl.net> wrote:

Sent to CCL by: Arvydas Tamulis [tamulis*o*mserv.itpa.lt]
Dear Cplleagues,

How to write in the input file of GAMESS-US request for
6-31+G(d) basis set?

I am having now the request only for 6-31G:
$CONTRL SCFTYP=RHF RUNTYP=energy MAXIT=200 NOSYM=1 $END
$CONTRL ICHARG=0 MULT=1 $END
$BASIS GBASIS=n31 ngauss=6 $END
$DFT DFTTYP=b3lyp $END

Cheers,
Arvydas


http://www.ccl.net/cgi-bin/ccl/send_ccl_message
http://www.ccl.net/cgi-bin/ccl/send_ccl_message
http://www.ccl.net/chemistry/sub_unsub.shtml
http://www.ccl.net/spammers.txt




Be a better pen pal. Text or chat with friends inside Yahoo! Mail. See how. --0-1071979424-1195555513=:63553-- From owner-chemistry@ccl.net Tue Nov 20 06:37:00 2007 From: "Andreas Klamt klamt-#-cosmologic.de" To: CCL Subject: CCL:G: PCM thermodynamics Message-Id: <-35650-071120055859-32111-A/Qzct8qYTgZRl1mZ4HgKA*|*server.ccl.net> X-Original-From: Andreas Klamt Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=ISO-8859-1; format=flowed Date: Tue, 20 Nov 2007 11:58:30 +0100 MIME-Version: 1.0 Sent to CCL by: Andreas Klamt [klamt_+_cosmologic.de] Dear Tom, dear James, I strongly disagree: As I have already often communicated in the CCL, it does not make sense to do the frequency calculations in solution for getting the free energies in solution. Reasons are: 1) The solvation change in the translational/vibrational/rotational part of the free energy is already essentially included in the solvation model contribution (e.g. PCM, COSMO, or COSMO-RS), most like as part of the non-dielectric contributions, because these models are parameterized on the total free energy change of solvation, while not doing separate frequency calculations in both states. It is almost impossible to separate this contribution out of these models. Hence, if you add the contribution explicitely, you will double count it!!! 2) The connection between free energy and frequencies is by far not as simple in solution as it is in the gasphase. The 2nd drivatives of the total energy cannot be direcly interpreted as frequencies, because the include the (larger) contribution from the slow, reorientational polarization degrees of freedom. They will hardly be able to follow the vibrations of the molecule. Here a non-equilibrium treatment (only available in a few codes) would be required. Furthermore, in reality the vibrations will be coupled vibrations including the motion and inertia of the next neighbors. The inertia of the solvent is not at all included in models like PCM. Hence stay out of it. Do the best free energy calculation of the species in the gasphase, add the solvation (free) energy from the solvation model (including geometry optimization in the solvent, if this was done for the parametrization of the solvation model) and thus get the best estimate of the free energy in solution. Obviously, for a solvent as chloroform can hardly be realistically described by just a dielectric continuum solvation model. COSMO-RS can include non-dielectric effects in much more details. It would even be able to take into account a finte concentration of your solute in the solvent, or mixtures, and temperature changes, because it is based on a fluid pase thermodynamics treatment. Bst regards Andreas -- ----------------------------------------------------------------------------- Dr. habil. Andreas Klamt COSMOlogic GmbH&CoKG Burscheider Str. 515 51381 Leverkusen, Germany Tel.: +49-2171-73168-1 Fax: +49-2171-73168-9 e-mail: klamt()cosmologic.de web: www.cosmologic.de ----------------------------------------------------------------------------- COSMOlogic Your Competent Partner for Computational Chemistry and Fluid Thermodynamics ----------------------------------------------------------------------------- James Robinson James.Robinson:prosonix.co.uk schrieb: > Sent to CCL by: "James Robinson" [James.Robinson!=rosonix.co.uk] > > Dear Tom, > > What is it you want to know? What information do you want from your > calculation? If indeed, you want to compare free energies in solution > you should reoptimise with PCM solvation and perform a frequency > calculation in solution. I also recommend using tight, or indeed, > verytight convergence for SCF cycles when using solvation. > > Dr James J Robinson > > Senior Scientist > > Prosonix Ltd > > Magdalen Centre > > Robert Robinson Avenue > > Oxford Science Park > > Oxford, Oxon. > > OX4 4GA, UK. > > +44(0) 1865 784243 > > http://www.prosonix.co.uk > > > > P Please consider the environment - do you really need to print this > email? > > > > > > > -----Original Message----- > >> From: owner-chemistry]~[ccl.net [mailto:owner-chemistry]~[ccl.net] >> > Sent: 19 November 2007 08:35 > To: James Robinson > Subject: CCL:G: PCM thermodynamics > > > Sent to CCL by: "Tom de Greef" [t.f.a.d.greef+*+tue.nl] > Dear all, > I have read several post how to use PCM calculations to correct for > solvent effects. I have the following problem: I have optimised two > conformations of the same structure in the gas-phase and subsequently > done a frequency calculation on both conformations. This has revealed to > me the difference of deltaG in the gas phase. Now I want to check the > difference in deltaG in the solvent (chloroform). The way I have done > this is doing a single point calculation on the gas phase optimised > structures using the pcm model in Gaussian03 supplied with the SCFVAC > command. > > > > -=his is automatically added to each message by the mailing script =-> > > > > > From owner-chemistry@ccl.net Tue Nov 20 08:08:00 2007 From: "James Robinson James.Robinson#prosonix.co.uk" To: CCL Subject: CCL:G: PCM thermodynamics Message-Id: <-35651-071120080327-4166-5AAl8DuOMU70RvwHYiu19g=-=server.ccl.net> X-Original-From: "James Robinson" Content-class: urn:content-classes:message Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="us-ascii" Date: Tue, 20 Nov 2007 13:02:04 -0000 MIME-Version: 1.0 Sent to CCL by: "James Robinson" [James.Robinson++prosonix.co.uk] I suggested doing the frequency calculation in solution, as in the case of performing pKa calculations one needs to performing gas phase and solution phase calculations. It depends on the information one wants. I do not disagree with the email from Andreas Klamt.=20 Dr James J Robinson =20 =20 -----Original Message----- > From: owner-chemistry:-:ccl.net [mailto:owner-chemistry:-:ccl.net]=20 Sent: 20 November 2007 10:59 To: James Robinson Subject: CCL:G: PCM thermodynamics Sent to CCL by: Andreas Klamt [klamt_+_cosmologic.de] Dear Tom, dear James, I strongly disagree: As I have already often communicated in the CCL, it does not make sense to do the frequency calculations in solution for=20 getting the free energies in solution. Reasons are: 1) The solvation change in the translational/vibrational/rotational part of the free energy is already essentially included in the solvation=20 model contribution (e.g. PCM, COSMO, or COSMO-RS), most like as part of=20 the non-dielectric contributions, because these models are parameterized on the total free energy change of solvation, while not doing separate=20 frequency calculations in both states. It is almost impossible to=20 separate this contribution out of these models. Hence, if you add the=20 contribution explicitely, you will double count it!!! 2) The connection between free energy and frequencies is by far not as=20 simple in solution as it is in the gasphase. The 2nd drivatives of the=20 total energy cannot be direcly interpreted as frequencies, because the=20 include the (larger) contribution from the slow, reorientational=20 polarization degrees of freedom. They will hardly be able to follow the=20 vibrations of the molecule. Here a non-equilibrium treatment (only=20 available in a few codes) would be required. Furthermore, in reality the vibrations will be coupled vibrations including the motion and inertia=20 of the next neighbors. The inertia of the solvent is not at all included in models like PCM. Hence stay out of it. Do the best free energy calculation of the species in the gasphase, add the solvation (free) energy from the solvation=20 model (including geometry optimization in the solvent, if this was done=20 for the parametrization of the solvation model) and thus get the best=20 estimate of the free energy in solution. Obviously, for a solvent as chloroform can hardly be realistically=20 described by just a dielectric continuum solvation model. COSMO-RS can=20 include non-dielectric effects in much more details. It would even be=20 able to take into account a finte concentration of your solute in the=20 solvent, or mixtures, and temperature changes, because it is based on a=20 fluid pase thermodynamics treatment. Bst regards Andreas --=20 ------------------------------------------------------------------------ ----- Dr. habil. Andreas Klamt COSMOlogic GmbH&CoKG Burscheider Str. 515 51381 Leverkusen, Germany Tel.: +49-2171-73168-1 Fax: +49-2171-73168-9 e-mail: klamt=3D-=3Dcosmologic.de web: www.cosmologic.de ------------------------------------------------------------------------ ----- COSMOlogic Your Competent Partner for Computational Chemistry and Fluid Thermodynamics ------------------------------------------------------------------------ ----- James Robinson James.Robinson:prosonix.co.uk schrieb: > Sent to CCL by: "James Robinson" [James.Robinson!=3Drosonix.co.uk] > > Dear Tom, > > What is it you want to know? What information do you want from your > calculation? If indeed, you want to compare free energies in solution > you should reoptimise with PCM solvation and perform a frequency > calculation in solution. I also recommend using tight, or indeed, > verytight convergence for SCF cycles when using solvation.=20 > > Dr James J Robinson > > Senior Scientist > > Prosonix Ltd > > Magdalen Centre > > Robert Robinson Avenue > > Oxford Science Park > > Oxford, Oxon. > > OX4 4GA, UK. > > +44(0) 1865 784243 > > http://www.prosonix.co.uk > > =20 > > P Please consider the environment - do you really need to print this > email? > > =20 > > =20 > > > -----Original Message----- > =20 >> From: owner-chemistry]~[ccl.net [mailto:owner-chemistry]~[ccl.net]=20 >> =20 > Sent: 19 November 2007 08:35 > To: James Robinson > Subject: CCL:G: PCM thermodynamics > > > Sent to CCL by: "Tom de Greef" [t.f.a.d.greef+*+tue.nl] > Dear all, > I have read several post how to use PCM calculations to correct for > solvent effects. I have the following problem: I have optimised two > conformations of the same structure in the gas-phase and subsequently > done a frequency calculation on both conformations. This has revealed to > me the difference of deltaG in the gas phase. Now I want to check the > difference in deltaG in the solvent (chloroform). The way I have done > this is doing a single point calculation on the gas phase optimised > structures using the pcm model in Gaussian03 supplied with the SCFVAC > command.=20 > > > > -=3Dhis is automatically added to each message by the mailing script = =3D-> > > > > > -=3D This is automatically added to each message by the mailing script = =3D-http://www.ccl.net/cgi-bin/ccl/send_ccl_messageSubscribe/Unsubscribe:=20Job: http://www.ccl.net/jobs=20http://www.ccl.net/spammers.txt From owner-chemistry@ccl.net Tue Nov 20 08:42:00 2007 From: "Pierre Archirel pierre.archirel * lcp.u-psud.fr" To: CCL Subject: CCL: PCM: any reference? Message-Id: <-35652-071120081437-7435-zk9/zljzZKd9C+XUq/XRUA : server.ccl.net> X-Original-From: Pierre Archirel Content-Transfer-Encoding: 8bit Content-Type: text/plain; charset=ISO-8859-1; format=flowed Date: Tue, 20 Nov 2007 14:28:49 +0100 MIME-Version: 1.0 Sent to CCL by: Pierre Archirel [pierre.archirel*_*lcp.u-psud.fr] This is a question to Andreas Klamt: Your communication to CCL about PCM frequencies is very interesting. Could you give us a reference? Pierre Archirel __________________________________________________________ Pierre Archirel Groupe de Chimie Théorique Laboratoire de Chimie Physique Tel: 01 69 15 63 86 Bat 349 Fax: 01 69 15 61 88 91405 Orsay Cedex France pierre.archirel/./lcp.u-psud.fr __________________________________________________________ From owner-chemistry@ccl.net Tue Nov 20 09:17:01 2007 From: "Dirk Blunk d.blunk+*+uni-koeln.de" To: CCL Subject: CCL:G: Comparability of energies Message-Id: <-35653-071120052324-10688-HQIjm3o2MBQCySLsBh1tdA-.-server.ccl.net> X-Original-From: "Dirk Blunk" Date: Tue, 20 Nov 2007 05:23:20 -0500 Sent to CCL by: "Dirk Blunk" [d.blunk-x-uni-koeln.de] Dear all, in how far are the absolute energies calculated for a hydrogen bridged dimer comparable to that for a one dimensional chain of the same molecules, calculated under periodic boundary conditions (PBC) in Gaussian03? E.g.: dimer (A,V representing the two molecules) A V Chain: A A A A ... V V V ... The repeating unit here is again the "dimer". The chain is calculated to be energetically so much more favored that I am somewhat unsettled in view of possible effects of the number of basis functions or BSSE or so. According to the log-file the number of basis functions and primitive gaussians are the same in both calculations. Energy differences for dimerisation A+V -> AV(dimer) -70 kJ/mol chain-formation A+V -> AV(chain) -110 kJ/mol Furthermore it seems that frequency calculations are technically not possible with PBC. (Can you comment on that?) Would a counterpoise correction work with PBC in the usual way (either with keyword counterpoise or by setting ghost atoms in the old fashion)? Thank you very much in advance, with best regards Dirk From owner-chemistry@ccl.net Tue Nov 20 09:52:00 2007 From: "Andreas Klamt klamt%x%cosmologic.de" To: CCL Subject: CCL: PCM thermodynamics Message-Id: <-35654-071120084814-4725-R/vEVnk/eM/Gsob6F+Dz1Q{=}server.ccl.net> X-Original-From: Andreas Klamt Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=ISO-8859-1; format=flowed Date: Tue, 20 Nov 2007 14:48:01 +0100 MIME-Version: 1.0 Sent to CCL by: Andreas Klamt [klamt() cosmologic.de] I even doubt that the frequency calculations in solution make sense for pKa calculations. At least we are getting excellent results if we neglect them, assuming that this part (i.e. the vibrational free energy change due to the deprotonation) essentially cancels out if you consider a closed cycle, wher ethe proton is just added to another species (to H2O). Andreas James Robinson James.Robinson#prosonix.co.uk schrieb: > Sent to CCL by: "James Robinson" [James.Robinson++prosonix.co.uk] > I suggested doing the frequency calculation in solution, as in the case > of performing pKa calculations one needs to performing gas phase and > solution phase calculations. It depends on the information one wants. I > do not disagree with the email from Andreas Klamt. > > Dr James J Robinson > > > > > > -----Original Message----- > >> From: owner-chemistry]*[ccl.net [mailto:owner-chemistry]*[ccl.net] >> > Sent: 20 November 2007 10:59 > To: James Robinson > Subject: CCL:G: PCM thermodynamics > > > Sent to CCL by: Andreas Klamt [klamt_+_cosmologic.de] > Dear Tom, dear James, > > I strongly disagree: As I have already often communicated in the CCL, it > > does not make sense to do the frequency calculations in solution for > getting the free energies in solution. Reasons are: > 1) The solvation change in the translational/vibrational/rotational part > > of the free energy is already essentially included in the solvation > model contribution (e.g. PCM, COSMO, or COSMO-RS), most like as part of > the non-dielectric contributions, because these models are parameterized > > on the total free energy change of solvation, while not doing separate > frequency calculations in both states. It is almost impossible to > separate this contribution out of these models. Hence, if you add the > contribution explicitely, you will double count it!!! > 2) The connection between free energy and frequencies is by far not as > simple in solution as it is in the gasphase. The 2nd drivatives of the > total energy cannot be direcly interpreted as frequencies, because the > include the (larger) contribution from the slow, reorientational > polarization degrees of freedom. They will hardly be able to follow the > vibrations of the molecule. Here a non-equilibrium treatment (only > available in a few codes) would be required. Furthermore, in reality the > > vibrations will be coupled vibrations including the motion and inertia > of the next neighbors. The inertia of the solvent is not at all included > > in models like PCM. > > Hence stay out of it. Do the best free energy calculation of the species > > in the gasphase, add the solvation (free) energy from the solvation > model (including geometry optimization in the solvent, if this was done > for the parametrization of the solvation model) and thus get the best > estimate of the free energy in solution. > > Obviously, for a solvent as chloroform can hardly be realistically > described by just a dielectric continuum solvation model. COSMO-RS can > include non-dielectric effects in much more details. It would even be > able to take into account a finte concentration of your solute in the > solvent, or mixtures, and temperature changes, because it is based on a > fluid pase thermodynamics treatment. > > Bst regards > > Andreas > > -- ----------------------------------------------------------------------------- Dr. habil. Andreas Klamt COSMOlogic GmbH&CoKG Burscheider Str. 515 51381 Leverkusen, Germany Tel.: +49-2171-73168-1 Fax: +49-2171-73168-9 e-mail: klamt,cosmologic.de web: www.cosmologic.de ----------------------------------------------------------------------------- COSMOlogic Your Competent Partner for Computational Chemistry and Fluid Thermodynamics ----------------------------------------------------------------------------- From owner-chemistry@ccl.net Tue Nov 20 10:49:01 2007 From: "Andreas Klamt klamt^cosmologic.de" To: CCL Subject: CCL: PCM: any reference? Message-Id: <-35655-071120101714-5319-tQ+YArHvL76JMLXOyokjUg,,server.ccl.net> X-Original-From: Andreas Klamt Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset=ISO-8859-1; format=flowed Date: Tue, 20 Nov 2007 16:17:00 +0100 MIME-Version: 1.0 Sent to CCL by: Andreas Klamt [klamt*_*cosmologic.de] Pierre, I am afraid that I do not know a reference for my arguments on the=20 meaning and importance of vibrational contributions in solution. I hope=20 that they are quite obvious and true without a reference. Best regards Andreas Pierre Archirel pierre.archirel * lcp.u-psud.fr schrieb: > > Sent to CCL by: Pierre Archirel [pierre.archirel*_*lcp.u-psud.fr] > This is a question to Andreas Klamt: > Your communication to CCL about PCM frequencies is very interesting. > Could you give us a reference? > Pierre Archirel > > __________________________________________________________ > > Pierre Archirel > Groupe de Chimie Th=E9orique > Laboratoire de Chimie Physique Tel: 01 69 15 63 86 > Bat 349 Fax: 01 69 15 61 88 > 91405 Orsay Cedex > France pierre.archirel^lcp.u-psud.fr > __________________________________________________________ > > > > -=3D This is automatically added to each message by the mailing script = =3D- > To recover the email address of the author of the message, please chang= ehttp://www.ccl.net/chemistry/sub_unsub.shtm= lConferences:=20 > http://server.ccl.net/chemistry/announcements/conferences/ > > Search Messages: http://www.ccl.net/htdig (login: ccl, Password: searc= h)> > -+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-= + > > > > > --=20 -------------------------------------------------------------------------= ---- Dr. habil. Andreas Klamt COSMOlogic GmbH&CoKG Burscheider Str. 515 51381 Leverkusen, Germany Tel.: +49-2171-73168-1 Fax: +49-2171-73168-9 e-mail: klamt+/-cosmologic.de web: www.cosmologic.de -------------------------------------------------------------------------= ---- COSMOlogic Your Competent Partner for Computational Chemistry and Fluid Thermodynamics -------------------------------------------------------------------------= ---- From owner-chemistry@ccl.net Tue Nov 20 13:42:00 2007 From: "Barabara Jagoda-Cwiklik barbara . fh.huji.ac.il" To: CCL Subject: CCL:G: PCM thermodynamics Message-Id: <-35656-071120113612-13155-7jmyIM4OC2ZIyu3UGitcZg a server.ccl.net> X-Original-From: "Barabara Jagoda-Cwiklik" Content-Type: multipart/alternative; boundary="----=_Part_19828_16636219.1195570285456" Date: Tue, 20 Nov 2007 16:51:25 +0200 MIME-Version: 1.0 Sent to CCL by: "Barabara Jagoda-Cwiklik" [barbara**fh.huji.ac.il] ------=_Part_19828_16636219.1195570285456 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline On Nov 19, 2007 10:35 AM, Tom de Greef t.f.a.d.greef:tue.nl < owner-chemistry##ccl.net> wrote: > > Sent to CCL by: "Tom de Greef" [t.f.a.d.greef+*+tue.nl] > Dear all, > I have read several post how to use PCM calculations to correct for > solvent effects. I have the following problem: I have optimised two > conformations of the same structure in the gas-phase and subsequently done a > frequency calculation on both conformations. This has revealed to me the > difference of deltaG in the gas phase. Now I want to check the difference in > deltaG in the solvent (chloroform). The way I have done this is doing a > single point calculation on the gas phase optimised structures using the pcm > model in Gaussian03 supplied with the SCFVAC command. This has given me the > following results: > > Conformer 1 > > # freq b3lyp/6-311+g(d,p) > > Zero-point correction= 0.311308 > (Hartree/Particle) > Thermal correction to Energy= 0.335385 > Thermal correction to Enthalpy= 0.336330 > Thermal correction to Gibbs Free Energy= 0.254360 > Sum of electronic and zero-point Energies= -1327.753254 > Sum of electronic and thermal Energies= - 1327.729177 > Sum of electronic and thermal Enthalpies= -1327.728233 > Sum of electronic and thermal Free Energies= -1327.810202 > > > > b3lyp/6-311+g(d,p) scrf=(solvent=chloroform) geom=connectivity > volume=tight > -------------------------------------------------------------------- > Variational PCM results > ======================= > (a.u.) = -1328.064544 > ( a.u.) = -1328.080065 > (a.u.) = -1328.081509 > (a.u.) = -1328.062912 > ( a.u.) = -1328.081527 > Total free energy in solution: > with all non electrostatic terms (a.u.) = -1328.063559 > -------------------------------------------------------------------- > (Unpolarized solute)-Solvent (kcal/mol) = - 9.74 > (Polarized solute)-Solvent (kcal/mol) = -11.68 > Solute polarization (kcal/mol) = 1.02 > Total electrostatic (kcal/mol) = -10.66 > -------------------------------------------------------------------- > Cavitation energy (kcal/mol) = 33.71 > Dispersion energy (kcal/mol) = -23.68 > Repulsion energy (kcal/mol) = 1.25 > Total non electrostatic (kcal/mol) = 11.27 > DeltaG (solv) (kcal/mol) = 0.62 > -------------------------------------------------------------------- > > Conformer 2 > > # freq b3lyp/6-311+g(d,p) > > Zero-point correction= 0.311268(Hartree/Particle) > Thermal correction to Energy= 0.335317 > Thermal correction to Enthalpy= 0.336261 > Thermal correction to Gibbs Free Energy= 0.254121 > Sum of electronic and zero-point Energies= -1327.744984 > Sum of electronic and thermal Energies= -1327.720934 > Sum of electronic and thermal Enthalpies= -1327.719990 > Sum of electronic and thermal Free Energies= - 1327.802131 > > > b3lyp/6-311+g(d,p) scrf=(solvent=chloroform) geom=connectivity > volume=tight > -------------------------------------------------------------------- > Variational PCM results > ======================= > (a.u.) = -1328.056233 > (a.u.) = -1328.073249 > (a.u.) = - 1328.075019 > (a.u.) = -1328.054191 > (a.u.) = -1328.075039 > Total free energy in solution: > with all non electrostatic terms ( a.u.) = -1328.056641 > -------------------------------------------------------------------- > (Unpolarized solute)-Solvent (kcal/mol) = -10.68 > (Polarized solute)-Solvent (kcal/mol) = - 13.08 > Solute polarization (kcal/mol) = 1.28 > Total electrostatic (kcal/mol) = -11.80 > -------------------------------------------------------------------- > Cavitation energy (kcal/mol) = 33.47 > Dispersion energy (kcal/mol) = -23.16 > Repulsion energy (kcal/mol) = 1.23 > Total non electrostatic (kcal/mol) = 11.54 > DeltaG (solv) (kcal/mol) = - 0.26 > -------------------------------------------------------------------- > > Using the results of the gas phase calculations I have calculated a > difference in deltaG between cvonformer 1 and 2 of 5.1 kcal/mol. The > difference in Total free energy in solution( with all non electrostatic > terms) between conformer 1 and 2 is 4.35 kca/mol. However, if I correct > the difference in free energy in the gas-phase with the difference in de > DeltaG(solv) term (wich is 0.88 kcalmol) I get a value of 4.18 kcal/mol. > First question: wich of these two numbers is correct? > Second question: Is this the usual approach taken by people in the > computational field or do they really re-optimize the gas-phase minimized > structures using the PCM approach? Then the next question is: How do they > calculate the frequencies? Do you also have to calculate the frequencies > using the PCM option? > > Hope anyone can help me. > > With best regards, > > Tom> Hi Tom, Question No.2: the way you calculated ddG doesn't include the change in the geometry of both conformers in the solvent (because in principle this energy is different for each conformer). More "natural" is to take into consideration the "relaxed" solute geometry (for example, after optimization within PCM or e.g. from MD simulations what I was actually doing). Best regards, Basia -- Barbara Jagoda-Cwiklik, Ph.D The Fritz Haber Research Center for Molecular Dynamics Institute of Chemistry The Hebrew University of Jerusalem Jerusalem, ISRAEL e-mail: barbara##fh.huji.ac.il ------=_Part_19828_16636219.1195570285456 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline On Nov 19, 2007 10:35 AM, Tom de Greef t.f.a.d.greef:tue.nl <owner-chemistry##ccl.net> wrote:

Sent to CCL by: "Tom  de Greef" [t.f.a.d.greef+*+tue.nl]
Dear all,
I have read several post how to use PCM calculations to correct for solvent effects. I have the following problem: I have optimised two conformations of the same structure in the gas-phase and subsequently done a frequency calculation on both conformations. This has revealed to me the difference of deltaG in the gas phase. Now I want to check the difference in deltaG in the solvent (chloroform). The way I have done this is doing a single point calculation on the gas phase optimised structures using the pcm model in Gaussian03 supplied with the SCFVAC command. This has given me the following results:

                       Conformer 1

# freq b3lyp/6-311+g(d,p)

 Zero-point correction=                           0.311308    (Hartree/Particle)
 Thermal correction to Energy=                    0.335385
 Thermal correction to Enthalpy=                  0.336330
 Thermal correction to Gibbs Free Energy=         0.254360
 Sum of electronic and zero-point Energies=          -1327.753254
 Sum of electronic and thermal Energies=             - 1327.729177
 Sum of electronic and thermal Enthalpies=           -1327.728233
 Sum of electronic and thermal Free Energies=        -1327.810202



b3lyp/6-311+g(d,p) scrf=(solvent=chloroform) geom=connectivity volume=tight
--------------------------------------------------------------------
 Variational PCM results
 =======================
 <psi(0)|   H    |psi(0)>                     (a.u.) =   -1328.064544
 <psi(0)|H+V(0)/2|psi(0)>                     ( a.u.) =   -1328.080065
 <psi(0)|H+V(f)/2|psi(0)>                     (a.u.) =   -1328.081509
 <psi(f)|   H    |psi(f)>                     (a.u.) =   -1328.062912
 <psi(f)|H+V(f)/2|psi(f)>                     ( a.u.) =   -1328.081527
 Total free energy in solution:
 with all non electrostatic terms            (a.u.) =   -1328.063559
 --------------------------------------------------------------------
 (Unpolarized solute)-Solvent             (kcal/mol) =      - 9.74
 (Polarized solute)-Solvent               (kcal/mol) =     -11.68
 Solute polarization                      (kcal/mol) =       1.02
 Total electrostatic                      (kcal/mol) =     -10.66
 --------------------------------------------------------------------
 Cavitation energy                        (kcal/mol) =      33.71
 Dispersion energy                        (kcal/mol) =     -23.68
 Repulsion energy                         (kcal/mol) =       1.25
 Total non electrostatic                  (kcal/mol) =       11.27
 DeltaG (solv)                            (kcal/mol) =       0.62
 --------------------------------------------------------------------

                            Conformer 2

# freq b3lyp/6-311+g(d,p)

 Zero-point correction=                           0.311268 (Hartree/Particle)
 Thermal correction to Energy=                    0.335317
 Thermal correction to Enthalpy=                  0.336261
 Thermal correction to Gibbs Free Energy=         0.254121
 Sum of electronic and zero-point Energies=          -1327.744984
 Sum of electronic and thermal Energies=             -1327.720934
 Sum of electronic and thermal Enthalpies=           -1327.719990
 Sum of electronic and thermal Free Energies=        - 1327.802131


 b3lyp/6-311+g(d,p) scrf=(solvent=chloroform) geom=connectivity volume=tight
--------------------------------------------------------------------
 Variational PCM results
 =======================
 <psi(0)|   H    |psi(0)>                     (a.u.) =   -1328.056233
 <psi(0)|H+V(0)/2|psi(0)>                     (a.u.) =   -1328.073249
 <psi(0)|H+V(f)/2|psi(0)>                     (a.u.) =   - 1328.075019
 <psi(f)|   H    |psi(f)>                     (a.u.) =   -1328.054191
 <psi(f)|H+V(f)/2|psi(f)>                     (a.u.) =   -1328.075039
 Total free energy in solution:
 with all non electrostatic terms            ( a.u.) =   -1328.056641
 --------------------------------------------------------------------
 (Unpolarized solute)-Solvent             (kcal/mol) =     -10.68
 (Polarized solute)-Solvent               (kcal/mol) =     - 13.08
 Solute polarization                      (kcal/mol) =       1.28
 Total electrostatic                      (kcal/mol) =     -11.80
 --------------------------------------------------------------------
 Cavitation energy                        (kcal/mol) =       33.47
 Dispersion energy                        (kcal/mol) =     -23.16
 Repulsion energy                         (kcal/mol) =       1.23
 Total non electrostatic                  (kcal/mol) =      11.54
 DeltaG (solv)                            (kcal/mol) =      - 0.26
 --------------------------------------------------------------------

Using the results of the gas phase calculations I have calculated a difference in deltaG between cvonformer 1 and 2 of 5.1 kcal/mol. The difference in  Total free energy in solution( with all non electrostatic terms) between conformer 1 and 2 is 4.35 kca/mol. However, if I correct the difference in free energy in the gas-phase with the difference in de DeltaG(solv) term (wich is 0.88 kcalmol) I get a value of 4.18 kcal/mol.
First question: wich of these two numbers is correct?
Second question: Is this the usual approach taken by people in the computational field or do they really re-optimize the gas-phase minimized structures using the PCM approach? Then the next question is: How do they calculate the frequencies? Do you also have to calculate the frequencies using the PCM option?

Hope anyone can help me.

With best regards,

Tom



E-mail to subscribers: CHEMISTRY##ccl.net or use:
     http://www.ccl.net/cgi-bin/ccl/send_ccl_message

E-mail to administrators: CHEMISTRY-REQUEST##ccl.net or use
     http://www.ccl.net/cgi-bin/ccl/send_ccl_message

Subscribe/Unsubscribe:
     

Before posting, check wait time at: http://www.ccl.net

Job: http://www.ccl.net/jobs
Conferences: http://server.ccl.net/chemistry/announcements/conferences/

Search Messages: http://www.ccl.net/htdig  (login: ccl, Password: search)
     http://www.ccl.net/spammers.txt

RTFI: http://www.ccl.net/chemistry/aboutccl/instructions/


Hi Tom,

Question No.2: the way you calculated ddG doesn't include the change in the geometry of both conformers in the solvent (because in principle this energy is different for each conformer).
More "natural" is to take into consideration the "relaxed" solute geometry (for example, after optimization within PCM or e.g. from MD simulations what I was actually doing).
 
Best regards,
Basia

--
Barbara Jagoda-Cwiklik, Ph.D
The Fritz Haber Research Center for Molecular Dynamics
Institute of Chemistry
The Hebrew University of Jerusalem
Jerusalem, ISRAEL
e-mail: barbara##fh.huji.ac.il ------=_Part_19828_16636219.1195570285456-- From owner-chemistry@ccl.net Tue Nov 20 14:17:01 2007 From: "Alexander Hoepker achoepker-.-gmail.com" To: CCL Subject: CCL:G: IRC calculation Message-Id: <-35657-071120111926-5584-2AEzyXEopGJV+9qSif+eTA=-=server.ccl.net> X-Original-From: "Alexander Hoepker" Content-Disposition: inline Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=ISO-8859-1 Date: Tue, 20 Nov 2007 10:16:05 -0500 MIME-Version: 1.0 Sent to CCL by: "Alexander Hoepker" [achoepker.:.gmail.com] I'm performing a forward IRC calculation starting with a transition state that has been optimized in Gaussian 03 in B3LYP (6/31G(d)) level of theory. The transition state was verified by one imaginary frequency for this particular deprotonation. my keywords are: #T B3LYP/6-31G(d) IRC=(CalcFC,Forward,MaxPoints=40) After only a couple of points the IRC appears to search backwards instead of forwards. IRC calculations are black magic for me and to be honest it is not surprising to me that the program confused the backward with the forward reaction coordinate. Could someone give me a hint as what I might be doing wrong or how I could tip the calculation in the right direction. Thanks a lot in advance. Alex From owner-chemistry@ccl.net Tue Nov 20 14:52:00 2007 From: "Mateusz Witkowski big_mateo**inbox.com" To: CCL Subject: CCL:G: Intel Quad-core vs AMD Opteron Message-Id: <-35658-071120101041-2073-oi5U55ZMI6MpbsYRzXcN3w[A]server.ccl.net> X-Original-From: Mateusz Witkowski Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset=US-ASCII Date: Tue, 20 Nov 2007 06:00:00 -0800 Mime-Version: 1.0 Sent to CCL by: Mateusz Witkowski [big_mateo[-]inbox.com] Dear CCL-ers, I wonder what is your opinion about these two types of CPUs. We are going = to build a 64CPUs cluster but still haven't decided about its architecture. The cluster will be mainly used for the purpose of QM calculations = (Gaussian, NWchem, GAMMES, ...). My major questions are: 1) Which CPU is better/more efficient? Are they comparable? 2) Does the number of cores really improve calculations? Assuming that = each server will have 2CPUs (2.4GHz) on board - with Intel there will be 8 = cores while with AMD only 4 cores. 3) Is it true that Intel quad-core is also a 64bit one? If so, why Itanium = is more expensive? 4) What is the real advantage of using 64bit instead of 32bit for QM? - = the only case I'm aware of is the problem with >2GB file size (eg. rwf = files in Gaussian) Any other suggestions are welcomed. TIA Mateo=20 From owner-chemistry@ccl.net Tue Nov 20 15:55:00 2007 From: "Jerome Kieffer jerome.Kieffer=-=terre-adelie.org" To: CCL Subject: CCL:G: Intel Quad-core vs AMD Opteron Message-Id: <-35659-071120155206-22075-X/9897qgeKek7yug8wTr6A^^^server.ccl.net> X-Original-From: Jerome Kieffer Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset=ISO-8859-15 Date: Tue, 20 Nov 2007 21:53:52 +0100 Mime-Version: 1.0 Sent to CCL by: Jerome Kieffer [jerome.Kieffer{:}terre-adelie.org] On Tue, 20 Nov 2007 06:00:00 -0800 "Mateusz Witkowski big_mateo**inbox.com" wrote: > Sent to CCL by: Mateusz Witkowski [big_mateo[-]inbox.com] > Dear CCL-ers, > I wonder what is your opinion about these two types of CPUs. We are going= to build a 64CPUs cluster but still haven't decided about its architecture. > The cluster will be mainly used for the purpose of QM calculations (Gauss= ian, NWchem, GAMMES, ...). > My major questions are: > 1) Which CPU is better/more efficient? Are they comparable? I think there is little difference between the two with the same frequency/number of core, memory...=20 Opteron are better to allocate large amount of memory, Xeon have larger cache.=20 > 2) Does the number of cores really improve calculations? Assuming that ea= ch server will have 2CPUs (2.4GHz) on board - with Intel there will be 8 co= res while with AMD only 4 cores. This make the biggest difference. > 3) Is it true that Intel quad-core is also a 64bit one?=20 yes > If so, why Itanium is more expensive? another architecure, not desktop oriented.=20 > 4) What is the real advantage of using 64bit instead of 32bit for QM? - t= he only case I'm aware of is the problem with >2GB file size (eg. rwf files= in Gaussian) being able to have 1 calculation running on 8core with 32GB is nowadays common and helps to keep all 2e integrals in-memory what would not be possible with only 2GB of memory (depending on the size of the system). This speeds up the calculation a lot. Hope this helps --=20 J=E9r=F4me KIEFFER From owner-chemistry@ccl.net Tue Nov 20 16:54:00 2007 From: "Tanusree Sarkar sarkar.tanusree{}gmail.com" To: CCL Subject: CCL: MD Simulation of nanotube Message-Id: <-35660-071120160317-25676-kHFpkl7C2Iv+RIxPaBmfiw]~[server.ccl.net> X-Original-From: "Tanusree Sarkar" Content-Type: multipart/alternative; boundary="----=_Part_1353_10168622.1195586305041" Date: Tue, 20 Nov 2007 20:18:25 +0100 MIME-Version: 1.0 Sent to CCL by: "Tanusree Sarkar" [sarkar.tanusree_-_gmail.com] ------=_Part_1353_10168622.1195586305041 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Dear CCL Users, I want to simulate growth of carbon nanotube using MD simulation. Can you please refer some MD software for this? Any kind of help is greatly appreciated Thanking you With best regards Tanusree ------=_Part_1353_10168622.1195586305041 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline
Dear CCL Users,
                        I want to simulate growth of carbon nanotube using MD simulation. Can you please refer
some MD software for this?
Any kind of help is greatly appreciated
Thanking you
With best regards
Tanusree
------=_Part_1353_10168622.1195586305041-- From owner-chemistry@ccl.net Tue Nov 20 17:29:01 2007 From: "Shina Kamerlin skamerli+*+usc.edu" To: CCL Subject: CCL:G: IRC calculation Message-Id: <-35661-071120162329-3634-4AyahxnjTYweg7qP/3y44w_._server.ccl.net> X-Original-From: Shina Kamerlin Content-disposition: inline Content-language: en Content-transfer-encoding: 7BIT Content-type: text/plain; charset=us-ascii Date: Tue, 20 Nov 2007 12:38:08 -0800 MIME-version: 1.0 Sent to CCL by: Shina Kamerlin [skamerli++usc.edu] Hi! I don't know how helpful it would be, but usually when it gets confused in any sort of optimisation, I switch from CalcFC to CalcAll - by calculating the force constants at every step, you are helping it get less confused and "guiding" it in the right direction. This usually solves my problems, even if it is much more computationally demanding. Of course, your problems could come from something completely different....you could also try mapping out around the transition state to verify that it really is a transition state (i.e. do a small more o'ferrall jencks plot) before trying to do the IRC. Also, did you do the frequency calculation in gas phase or solution? Hope this helps, Lynn ----- Original Message ----- > From: "Alexander Hoepker achoepker-.-gmail.com" Date: Tuesday, November 20, 2007 11:59 am Subject: CCL:G: IRC calculation To: "Kamerlin, Shina " > > Sent to CCL by: "Alexander Hoepker" [achoepker.:.gmail.com] > I'm performing a forward IRC calculation starting with a transition > state that has been optimized in Gaussian 03 in B3LYP (6/31G(d)) level > of theory. The transition state was verified by one imaginary > frequency for this particular deprotonation. > > my keywords are: #T B3LYP/6-31G(d) IRC=(CalcFC,Forward,MaxPoints=40) > > After only a couple of points the IRC appears to search backwards > instead of forwards. IRC calculations are black magic for me and to be > honest it is not surprising to me that the program confused the > backward with the forward reaction coordinate. Could someone give > me a > hint as what I might be doing wrong or how I could tip the calculation > in the right direction. Thanks a lot in advance. > > Alex > > > > -= This is automatically added to each message by the mailing > script =- > To recover the email address of the author of the message, please > changethe strange characters on the top line to the -.- sign. You can > also> Conferences: > http://server.ccl.net/chemistry/announcements/conferences/ > Search Messages: http://www.ccl.net/htdig (login: ccl, Password: > search)> > -+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+- > +-+-+ > > > > From owner-chemistry@ccl.net Tue Nov 20 18:03:00 2007 From: "Daniel Jana dfjana*gmail.com" To: CCL Subject: CCL:G: Intel Quad-core vs AMD Opteron Message-Id: <-35662-071120172812-2676-QB42uEgMSeBpRNGmPa3poQ() server.ccl.net> X-Original-From: Daniel Jana Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=ISO-8859-1; format=flowed Date: Tue, 20 Nov 2007 21:23:33 +0000 MIME-Version: 1.0 Sent to CCL by: Daniel Jana [dfjana-,-gmail.com] Hello! Mateusz Witkowski big_mateo**inbox.com wrote: > My major questions are: > 1) Which CPU is better/more efficient? Are they comparable? > While I can't give you real data, I know of a cluster built in Germany (with 100+ computers) in the end of last year/beginning of this year that was built with AMD because, after benchmarking, it was an overall better option than with Intel. Based on some back of the envelope calculations we got to the same result, mainly because floating point operations where better in a cluster with AMD when price was considered (and not only the number of processors). Your mileage may vary though and this may be reasonable if you get a precompiled program that was built for a slightly different architecture (or was built more generally to accommodate several architectures). > 2) Does the number of cores really improve calculations? Assuming that each server will have 2CPUs (2.4GHz) on board - with Intel there will be 8 cores while with AMD only 4 cores. > Although I cannot (once again!) give real information on this, thinking in a purely abstract way, there comes a point where too many things are being written to the disk at the same time. If you have 8 cores and only one disk in the computer used for placing stuff (Gaussian scratch files, output results, swap, ...) you will definitely be nearer the limit of the connection between disk and CPU. Though I am not sure if it is by any way near this value, it did work as a disadvantage for the Intel when we where weighing pros and cons (but, as I said, with no real data on the subject). You could obviously consider using several disks, with RAID for splitting but that then increases the costs and may lower the number of computers you buy with the same money. On a side note, the number of cores does increase the number of calculations you can run. If you use programs that don't write to the disk except occasionally and do everything on the RAM then with 8 cores you will be doing twice the work you are doing with 4 so it does improve calculations. But, as above, your mileage may vary (mainly due to the disk access which will be the bottleneck). > 3) Is it true that Intel quad-core is also a 64bit one? If so, why Itanium is more expensive? > Yes! Check: http://en.wikipedia.org/wiki/Intel_Core_2 (although they are quad-core they belong to the Core 2 family). Also check this: http://en.wikipedia.org/wiki/Itanium Basically Itanium processors are a different architecture. They may be better (hence the price) but you then get into the problem of getting software for those (especially software that may not be accessible in the form of source code). There is a sentence in the Itanium page of the Wikipedia that reveals the state of things: "As of 2007, Itanium is the fourth-most deployed microprocessor architecture for enterprise-class systems, behind x86-64, IBM POWER, and SPARC." With x86-64 being the Intel Core 2 and AMD current architectures. > 4) What is the real advantage of using 64bit instead of 32bit for QM? - the only case I'm aware of is the problem with >2GB file size (eg. rwf files in Gaussian) > That I'd leave to the specialists... Daniel Jana From owner-chemistry@ccl.net Tue Nov 20 21:26:01 2007 From: "Hod Greeley hod-ccl===greeley.org" To: CCL Subject: CCL:G: Intel Quad-core vs AMD Opteron Message-Id: <-35663-071120212342-11894-9Sgg2X4IGwvJqWm9f7XmDA{}server.ccl.net> X-Original-From: "Hod Greeley" Content-Transfer-Encoding: 8bit Content-Type: text/plain;charset=iso-8859-1 Date: Tue, 20 Nov 2007 18:14:20 -0800 (PST) MIME-Version: 1.0 Sent to CCL by: "Hod Greeley" [hod-ccl||greeley.org] Hello, Regarding #2, there could be a few reasons why you would not expect the processing power to scale linearly with the number of cores. There is a lot more to a system than just the cpu. Programs like Jaguar and Gaussian do their calculations in double precision, that is, 64-bit floating point. A true 64-bit cpu will have a significant speed advantage with these programs. This has nothing to do with the addressable memory limit. Overall, what you are asking is not as simple a question as it might seem. For instance, you may not get twice the performance with an 8 processor board over a 4 processor one, but if you have to have a whole other machine (including motherboard, power supply, disk and all) to have two 4-core systems, your price/performance may not actually come out ahead. Hopefully someone can point you toward some benchmarks with QM applications to help you decide. I haven't looked through these in some time, but you might look at the benchmarks posted at http://www.spec.org/. They even include a QM code in their test. Be careful, though. You need to understand what the benchmarks are really saying, which is not always easy to decipher. Best, Hod > > Sent to CCL by: Mateusz Witkowski [big_mateo[-]inbox.com] > Dear CCL-ers, > I wonder what is your opinion about these two types of CPUs. We are going to build a 64CPUs cluster but still haven't decided about its > architecture. > The cluster will be mainly used for the purpose of QM calculations (Gaussian, NWchem, GAMMES, ...). > My major questions are: > 1) Which CPU is better/more efficient? Are they comparable? > 2) Does the number of cores really improve calculations? Assuming that each server will have 2CPUs (2.4GHz) on board - with Intel there will be 8 cores while with AMD only 4 cores. > 3) Is it true that Intel quad-core is also a 64bit one? If so, why Itanium is more expensive? > 4) What is the real advantage of using 64bit instead of 32bit for QM? - the only case I'm aware of is the problem with >2GB file size (eg. rwf files in Gaussian) > > Any other suggestions are welcomed. > TIA > Mateo> > > >