From chemistry-request@server.ccl.net Mon Jul 9 08:05:29 2001 Received: from web13601.mail.yahoo.com ([216.136.175.112]) by server.ccl.net (8.11.0/8.11.0) with SMTP id f69C5T513737 for ; Mon, 9 Jul 2001 08:05:29 -0400 Message-ID: <20010709120529.54005.qmail@web13601.mail.yahoo.com> Received: from [129.215.223.79] by web13601.mail.yahoo.com via HTTP; Mon, 09 Jul 2001 13:05:29 BST Date: Mon, 9 Jul 2001 13:05:29 +0100 (BST) From: =?iso-8859-1?q?tj=20su?= Subject: calculation using Gaussian 94 To: chemistry@ccl.net MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Dear CCLers, We had difficulties on calculation of the bond orders of a molecule using the Gaussian 94 package. Hopefully, someone can help us to solve this problem. Thanks a lot in advance. The input and output are as follows: iuput: ------------------------------------------------------------ #n b3lyp/6-311g* scf=tight density=current aim=bo scfcon=5 ------------------------------------------------------------ output: ------------------------------------------------------------------------------- 3 attractors - 2 bond point(s) + 0 ring point(s) - 0 cage point(s) = 1 WARNING: RMS ERROR HAS INCREASED WARNING: RMS ERROR HAS INCREASED NEWTON STEP FAILED FOR SURFACE SHEET 2 Error termination via Lnk1e in /usr/local/package1/g98/l609.exe. ------------------------------------------------------------------------------- May May Lin ____________________________________________________________ Do You Yahoo!? Get your free @yahoo.co.uk address at http://mail.yahoo.co.uk or your free @yahoo.ie address at http://mail.yahoo.ie From chemistry-request@server.ccl.net Sun Jul 8 21:38:22 2001 Received: from lacebark.ntu.edu.au ([138.80.63.12]) by server.ccl.net (8.11.0/8.11.0) with SMTP id f691cK525518 for ; Sun, 8 Jul 2001 21:38:21 -0400 Received: (qmail 21563 invoked by uid 200); 9 Jul 2001 10:47:32 +0930 Date: Mon, 9 Jul 2001 10:47:32 +0930 From: Brian Salter-Duke To: chemistry@ccl.net Subject: Re: CCL:compile gaussian98 for Redhat 7.1 with pgf77 Message-ID: <20010709104732.B20594@lacebark.ntu.edu.au> Mail-Followup-To: chemistry@ccl.net Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.3.10i There have been several requests for information on compiling G98 on Redhat 7.1 linux. I think all the answers have already been posted, but here is a summary, followed by some questions:- Add:- #include in g98/bsd/mdutil.c after #ifdef i386 #include #include Alter these sections of i386.make, under the heading:- # Exceptional link commands: l402.exe: $(EXTRAS) l402.a $(RUNFSP) -e m$* $(BSDDIR)/main.F $(RUNMAKE) $(MAKELOC) OPTFLAG='$(OPTFLAG)' $(EXTRAS) $*.a m$*.o cp l402.a l402_save.a ar x l402.a bd0402.o ar d l402.a bd0402.o $(RUNF77) $(FFLAGS) -o $*.exe m$*.o $(EXTOBJ) $*.a bd0402.o \ $(EXTRAS) $(GAULIB) $(LIBS) rm -f m$*.* bd0402.o mv l402_save.a l402.a l609.exe: $(EXTRAS) l609.a $(RUNFSP) -e m$* $(BSDDIR)/main.F $(RUNMAKE) $(MAKELOC) OPTFLAG='$(OPTFLAG)' $(EXTRAS) $*.a m$*.o cp l609.a l609_save.a ar x l609.a bd0609.o ar d l609.a bd0609.o $(RUNF77) $(FFLAGS) -o $*.exe m$*.o $(EXTOBJ) $*.a bd0609.o \ $(EXTRAS) $(GAULIB) $(LIBS) rm -f m$*.* bd0609.o mv l609_save.a l609.a I have a few questions. The following flags are used:- UNROLL = -Munroll VECTOR = -Mvect=cachesize:524288 MACH = -tp p6 giving:- pgf77 -O2 -tp p6 -Munroll -Mvect=cachesize:524288 '-tp p6' should be unnecessary as the compiler is supposed to select the correct flag for the machine you are running on. Does -Munroll improve performance? What does -Mvect=cachesize:524288 do? I used it for compiling Gamess(US) and I got several errors including a total failure in a AM1 test. Would altering the number help? Would any other flags improve performance? For Gamess(US), I tried '-fast' or its equivalent '-O2 -Munroll -Munframe'. The results are a disaster - wrong energies, no SCF convergence, etc. '-Munroll' is the problem. Cheers, Brian. -- Associate Professor Brian Salter-Duke (Brian Duke) b_duke@lacebark.ntu.edu.au Chemistry, School of BECS, SITE, NT University, Darwin, NT 0909, Australia. Phone 08-89466702. Fax 08-89466847. http://www.smps.ntu.edu.au/ From chemistry-request@server.ccl.net Mon Jul 9 12:07:45 2001 Received: from c001.snv.cp.net ([209.228.32.132]) by server.ccl.net (8.11.0/8.11.0) with SMTP id f69G7i518511 for ; Mon, 9 Jul 2001 12:07:44 -0400 Received: (cpmta 28413 invoked from network); 9 Jul 2001 09:07:38 -0700 Date: 9 Jul 2001 09:07:38 -0700 Message-ID: <20010709160738.28412.cpmta@c001.snv.cp.net> X-Sent: 9 Jul 2001 16:07:38 GMT Received: from [199.98.21.13] by mail.roberttopper.com with HTTP; 09 Jul 2001 09:07:38 PDT Content-Type: text/plain Content-Disposition: inline Mime-Version: 1.0 To: CHEMISTRY@ccl.net From: Robert Topper Cc: roberttopper@roberttopper.com X-Mailer: Web Mail 3.9.3.3 Subject: SPARTAN 5.0 vs 5.1: coordinate driving X-Sent-From: roberttopper@roberttopper.com Hi all, a SPARTAN question. I've searched the archives but to no avail...and I also have asked the folks at Wavefunction, but they seemed to not be able to respond to this question. There are some important differences between versions 5.0 and 5.1 of SPARTAN. Specifically, the interfaces for coordinate driving and conformer searching are quite different. In 5.0, the conformer search module can be used to generate single-point energies as a function of a stiffly rotated dihedral angle (without doing, say, a constrained optimization at each angle). However, in 5.1 this does not seem to be possible, at least not with the dialogs and keywords we have tried. The coordinate driving module seems to insist on doing a constrained optimization, whereas all I want is a series of single-point energies as a function of the dihedral angle. Has anyone figured out how to do this in SPARTAN 5.1? Thanks, and cheers- Robert Q. Topper Dept of Chemistry Cooper Union New York, NY 10003 From chemistry-request@server.ccl.net Mon Jul 9 13:22:39 2001 Received: from deimos.email.Arizona.EDU (IDENT:root@[128.196.133.166]) by server.ccl.net (8.11.0/8.11.0) with ESMTP id f69HMd520220 for ; Mon, 9 Jul 2001 13:22:39 -0400 Received: from demetrio.u.arizona.edu (128.196.86.129) by deimos.email.Arizona.EDU (5.1.056) id 3B479AE900010C63 for chemistry@ccl.net; Mon, 9 Jul 2001 10:22:37 -0700 Message-Id: <5.1.0.14.0.20010709101047.00aabbf0@deimos.email.arizona.edu> X-Sender: dasf@deimos.email.arizona.edu X-Mailer: QUALCOMM Windows Eudora Version 5.1 Date: Mon, 09 Jul 2001 10:21:40 -0700 To: chemistry@ccl.net From: Demetrio Antonio da Silva Filho Subject: BSSE Summary Mime-Version: 1.0 Content-Type: multipart/alternative; boundary="=====================_155367634==_.ALT" --=====================_155367634==_.ALT Content-Type: text/plain; charset="iso-8859-1"; format=flowed Content-Transfer-Encoding: quoted-printable Dear CCL'ers, I would like to say thanks to Donald Boyd, Hong Jiang, Jaime Combariza, Douglas Fox, Sergio Galembeck, Tanja van Mourik and Robert Topper. The solution to my problem is perfectly discribed by Dr. Tanja Mourik= e-mail. Thanks a lot! Demetrio Filho Here is the summary. ++++++++++++++++++++++++++++++++++++++++++++++++++ Hi Demetrio, Take a look at the chapter by N. R. Kestner and J. E. Combariza, in Reviews= in Computational Chemistry, K. B. Lipkowitz and D. B. Boyd, Eds., Wiley-VCH,= New York, 1999, Vol. 13, pp. 99-132. Basis Set Superposition Errors: Theory and Practice. It has some examples. Good luck, Don Donald B. Boyd, Ph.D. ++++++++++++++++++++++++++++++++++++++++++++++++++ Hi, Demetrio Filho, Counterpoise correction (CC) can be done in Gaussian by using the key word,= =20 "massage". There is an article about BSSE in "Review in computational=20 chemistry" (the volume in 2000?), and at the end of the paper there is an=20 example of using CC in Gaussian. I hope that will be helpful for you. Good luck! Hong ++++++++++++++++++++++++++++++++++++++++++++++++++ The keyword is MASSAGE. I was not sure about the proper input MP2/6-31++g Massage test ...... input blank line 1 Nuc 0.0 2 Nuc 0.0 3 Nuc 0.0 Assuming you want to get rid of the first three atoms. Massage is used also to add functions to a particualr center --- Jaime E. Combariza Ph.D. Research Support Group ++++++++++++++++++++++++++++++++++++++++++++++++++ See the MASSAGE keyword in the G98 User's Reference for an example, page 118. Douglas J. Fox Director of Technical Support help@gaussian.com ++++++++++++++++++++++++++++++++++++++++++++++++++ Demetrio, You can use BSSE for dimers composed of equal molecules without any problem. Abra=E7os, S=E9rgio ++++++++++++++++++++++++++++++++++++++++++++++++++ If you calculate a binding energy De as: De =3D E{AB} -E{A} - E{B} Using counterpoise, you calculate the monomer energies E{A} and E{B} also in the complete dimer basis set. In Gaussian, you do this by setting the nuclear charge of the second dimer (B for E{A} and A for E{B}) to zero. This is done by the Massage keyword; see the examples in the Gaussian manual (www.gaussian.com/00000452.htm). (but don't believe what the manual says about counterpoise only providing "a crude estimate" - see papers below). Depending on how much your monomer geometries change upon complexation, you may also have to take into account the monomer deformation energies (the energy it takes to bring an anthracene molecule to the geometry it has in the complex). Some references: - F.B. van Duijneveldt, J.G.C.M. van Duijneveldt-van de Rijdt, and J.H. van Lenthe, Chem. Rev. 94, 1873, 1994 (on counterpoise theory) - S. Simon, M. Duran, J.J. Dannenberg, J. Chem. Phys. 105, 1996 (on geometry optimizations on BSSE-corrected potential energy surfaces). - F.B. van Duijneveldt, "Basis Superposition Error" in "Molecular Interactions" S. Scheiner (ed.), Wiley (1997) - T. van Mourik, A.K. Wilson, K.A. Peterson, D.E. Woon, T.H. Dunning, Jr., Adv. Quant. Chem. 31, 105, 1999 (showing BSSE effects on interaction energies, distances, and frequencies). You may also want to search the ccl archives; there have been a few= disussions on counterpoise the last few years. Hope that helps, Tanja --=20 ++++++++++++++++++++++++++++++++++++++++++++++++++ Dear Demetrio, There is an article in Reviews in Computational Chemistry which is quite good. It addresses your question directly. Neil R. Kestner and Jaime E. Combariza, "Basis Set Superposition Errors: Theory and Practice," Reviews in Computational Chemistry Vol. 13, pp. 99-132 (Wiley-VCH, New York, 1999). Hope this helps. Best regards, Robert Topper --=====================_155367634==_.ALT Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Dear CCL'ers,

I would like to say thanks to Donald Boyd, Hong Jiang, Jaime Combariza,
Douglas Fox, Sergio Galembeck, Tanja van Mourik and Robert Topper.

The solution to my problem is perfectly discribed by Dr. Tanja Mourik e-mail.


Thanks a lot!

Demetrio Filho




Here is the summary.

++++++++++++++++++++++++++++++++++++++++++++++++++

Hi Demetrio,
Take a look at the chapter by N. R. Kestner and J. E. Combariza, in Reviews in
Computational Chemistry, K. B. Lipkowitz and D. B. Boyd, Eds., Wiley-VCH, New
York, 1999, Vol. 13, pp. 99-132. Basis Set Superposition Errors: Theory and
Practice. It has some examples.
Good luck, Don
Donald B. Boyd, Ph.D.

++++++++++++++++++++++++++++++++++++++++++++++++++

Hi, Demetrio Filho,
Counterpoise correction (CC) can be done in Gaussian by using the key word, "massage". There is an article about BSSE in "Review in computational chemistry" (the volume in 2000?), and at the end of the paper there is an example of using CC in Gaussian.
I hope that will be helpful for you.
Good luck!
Hong

++++++++++++++++++++++++++++++++++++++++++++++++++


The keyword is MASSAGE.
I was not sure about the proper input

MP2/6-31++g Massage test

...... input



blank line
1 Nuc 0.0
2 Nuc 0.0
3 Nuc 0.0


Assuming you want to get rid of the first three atoms.
Massage is used also to add functions to a particualr center

---
Jaime E. Combariza Ph.D.
Research Support Group

++++++++++++++++++++++++++++++++++++++++++++++++++

See the MASSAGE keyword in the G98 User's Reference for
an example, page 118.


Douglas J. Fox
Director of Technical Support
help@gaussian.com

++++++++++++++++++++++++++++++++++++++++++++++++++

Demetrio,

You can use BSSE for dimers composed of equal molecules
without any problem.

Abra=E7os,
S=E9rgio

++++++++++++++++++++++++++++++++++++++++++++++++++

If you calculate a binding energy De as:
De =3D E{AB} -E{A} - E{B}
Using counterpoise, you calculate the monomer energies E{A} and E{B}
also in the complete dimer basis set. In Gaussian, you do this by
setting the nuclear charge of the second dimer (B for E{A} and
A for E{B}) to zero. This is done by the Massage keyword; see the
examples in the Gaussian manual (www.gaussian.com/00000452.htm).
(but don't believe what the manual says about counterpoise only
providing "a crude estimate" - see papers below).
Depending on how much your monomer geometries change upon complexation,
you may also have to take into account the monomer deformation energies
(the energy it takes to bring an anthracene molecule to the geometry it
has in the complex).
Some references:
- F.B. van Duijneveldt, J.G.C.M. van Duijneveldt-van de Rijdt, and
J.H. van Lenthe, Chem. Rev. 94, 1873, 1994 (on counterpoise theory)
- S. Simon, M. Duran, J.J. Dannenberg, J. Chem. Phys. 105, 1996
(on geometry optimizations on BSSE-corrected potential energy surfaces).
- F.B. van Duijneveldt, "Basis Superposition Error"
in "Molecular Interactions" S. Scheiner (ed.), Wiley (1997)
- T. van Mourik, A.K. Wilson, K.A. Peterson, D.E. Woon, T.H. Dunning, Jr.,
Adv. Quant. Chem. 31, 105, 1999 (showing BSSE effects on interaction
energies, distances, and frequencies).
You may also want to search the ccl archives; there have been a few disussions
on counterpoise the last few years.
Hope that helps,
Tanja
--
++++++++++++++++++++++++++++++++++++++++++++++++++

Dear Demetrio,
There is an article in Reviews in Computational Chemistry which
is quite good. It addresses your question directly.
Neil R. Kestner and Jaime E. Combariza, "Basis Set Superposition Errors:
Theory and Practice," Reviews in Computational Chemistry Vol. 13,
pp. 99-132 (Wiley-VCH, New York, 1999).
Hope this helps.
Best regards,
Robert Topper
 --=====================_155367634==_.ALT-- From chemistry-request@server.ccl.net Mon Jul 9 12:46:08 2001 Received: from c001.snv.cp.net ([209.228.32.130]) by server.ccl.net (8.11.0/8.11.0) with SMTP id f69Gk8519377 for ; Mon, 9 Jul 2001 12:46:08 -0400 Received: (cpmta 5598 invoked from network); 9 Jul 2001 09:46:03 -0700 Date: 9 Jul 2001 09:46:02 -0700 Message-ID: <20010709164602.5597.cpmta@c001.snv.cp.net> X-Sent: 9 Jul 2001 16:46:02 GMT Received: from [199.98.21.13] by mail.roberttopper.com with HTTP; 09 Jul 2001 09:46:02 PDT Content-Type: text/plain Content-Disposition: inline Mime-Version: 1.0 To: dasf@u.arizona.edu From: Robert Topper Cc: CHEMISTRY@ccl.net X-Mailer: Web Mail 3.9.3.3 Subject: Re: CCL:BSSE on Dimers X-Sent-From: roberttopper@roberttopper.com Dear Demetrio, There is an article in Reviews in Computational Chemistry which is quite good. It addresses your question directly. Neil R. Kestner and Jaime E. Combariza, "Basis Set Superposition Errors: Theory and Practice," Reviews in Computational Chemistry Vol. 13, pp. 99-132 (Wiley-VCH, New York, 1999). Hope this helps. Best regards, Robert Topper On Fri, 06 July 2001, Demetrio Antonio da Silva Filho wrote: > > Dear CCL'ers, > > I am doing some MP2/6-311G(2d,2p) canculation on > anthracene dimers (using Gaussian'98) and I want to > know how I can eliminate the BSSE (basis set > superposition error) using the counterpoise method > for that system of two equal molecules. > > I wonder if someone can point me to articles or > a scheme in order to do that using Gaussian'98. > > Thanks in advance, > > Demetrio Filho > From chemistry-request@server.ccl.net Mon Jul 9 20:48:34 2001 Received: from morita.chem.sunysb.edu ([129.49.49.135]) by server.ccl.net (8.11.0/8.11.0) with ESMTP id f6A0mY527557 for ; Mon, 9 Jul 2001 20:48:34 -0400 Received: from morita.chem.sunysb.edu (080-043.dialup.sunysb.edu [129.49.80.43]) by morita.chem.sunysb.edu (980427.SGI.8.8.8/980728.SGI.AUTOCF) via ESMTP id UAA15668 for ; Mon, 9 Jul 2001 20:52:32 -0400 (EDT) Sender: cuigl@morita.chem.sunysb.edu Message-ID: <3B4A5140.BAC91FCF@morita.chem.sunysb.edu> Date: Mon, 09 Jul 2001 20:50:08 -0400 From: Guanglei Cui Organization: SUNY at Stony Brook X-Mailer: Mozilla 4.76 [en] (X11; U; Linux 2.4.2-2 i686) X-Accept-Language: en MIME-Version: 1.0 To: chemistry@ccl.net Subject: summary: compile g98 in redhat 7.1 with pgf77 Content-Type: text/plain; charset=gb2312 Content-Transfer-Encoding: 7bit Dear all, I'd like to thank those who helped me with compiling Gaussian98 in RedHat 7.1, mostly Prof. Brain Salter -Duke and Dr. Doug. Most of the details have been described in Prof. Salter's previous email, including modifying mdutil.c, i386.make, etc. I also need to add one line in bsd/bldg98 right after mkdir temputil to copy ntran.inc to temputil. With all these changes, I am able to compile Gaussian98 with PGI compiler under RedHat 7.1 with BLAS library provided with Gaussian98 package (not optimized). The performance on Athlon 1.0GHz with 512MB memory for a 2-hour long HF/6-31G* ESP (electrostatic potential) calculation (on SGI Origin 2000) is 10 min faster and gives identical results. I tried other BLAS libraries from Greg Henry's websites. The recommended 1.2f gives some reference error while compiling, the latest 1.3e for PIII passed the compilation, however failed at run time (terminated at l401.exe). What BLAS library should I use to boost the performance most? Best regards, Guanglei Cui