From chemistry-request@ccl.net Fri Jul 31 07:57:01 1992 To: rak@igor.caltech.edu (Robert A. Kumpf) Subject: Re: Phone Number for QCPE? Date: Fri, 31 Jul 92 10:39:28 +0100 From: Omar G. Stradella Status: R Your message dated: Fri, 05 Jun 92 14:37:00 PDT > Does anyone have the current phone number for QCPE? > The number I have ( (812) 335-4784 ) is no longer in > service. Phone number is 812-855-4784 Fax number is 812-855-5539 ---- **************************************************************************** * Dr. Omar G. Stradella - omar@crs4.it * * CRS4 - Centro di Ricerca, Sviluppo e Studi Superiori in Sardegna * * - Centre for Advanced Studies, Research and Development in Sardinia * * -------------------------+----------------------------+----------------- * * Casella Postale 488 | Via Nazario Sauro, 10 | 39 13' 19" North * * 09100 Cagliari CA | 09123 Cagliari CA | 9 6' 17" East * * Italy | Italy | * * -------------------------+----------------------------+----------------- * * Phone: +39 70 27 96 236 FAX: +39 70 27 96 220 * **************************************************************************** --- Administrivia: This message is automatically appended by the mail exploder. CHEMISTRY@ccl.net --- everybody; CHEMISTRY-REQUEST@ccl.net --- coordinator only OSCPOST@ccl.net : send something from chemistry; FTP: www.ccl.net --- From chemistry-request@ccl.net Fri Jul 31 14:07:47 1992 Date: Fri, 31 Jul 1992 11:23 EST From: J_BROWN@uvmvax.uvm.edu Subject: least squates fitting 1D nmr spectra software To: chemistry@ccl.net Status: R To: Jim Macmillan (macmillian@iscsvax.uni.edu): There is a program from QCPC for least squares optimization of 1D NMR simulations (DNMR5: QCPE # 365). It will "optimize" up to 16 parameters including chemical shifts, coupling constants, populations, T2 values, and rate constants. It is only fair to warn you that it is a very "complex" program to use due to the nature of the calculations involved. --- Administrivia: This message is automatically appended by the mail exploder. CHEMISTRY@ccl.net --- everybody; CHEMISTRY-REQUEST@ccl.net --- coordinator only OSCPOST@ccl.net : send something from chemistry; FTP: www.ccl.net --- From chemistry-request@ccl.net Fri Jul 31 15:11:52 1992 Date: Fri, 31 Jul 92 08:35:50 -0700 From: zheng@violet.berkeley.edu To: chemistry@ccl.net Subject: Thanks those who responded to my question on DFT and ab initio Status: R I would like to thank all those who responded to my question about DFT and ab initio calculations. It's very helpful. The book by Jan Labanowski and Jan Andzelm (Density Functional Methods in Chemistry, QD462.6 D45 L33 1991) is highly recommended for those who are interested in DFT theory and it's applications. Yajun --- Administrivia: This message is automatically appended by the mail exploder. CHEMISTRY@ccl.net --- everybody; CHEMISTRY-REQUEST@ccl.net --- coordinator only OSCPOST@ccl.net : send something from chemistry; FTP: www.ccl.net --- From chemistry-request@ccl.net Fri Jul 31 19:09:55 1992 Date: Fri, 31 Jul 92 16:30:15 EDT From: m10!frisch@uunet.UU.NET (Michael Frisch) Subject: Re: Natural Populations in G90 To: chemistry@ccl.net Status: R Steven Bachrach asked: Is there a keyword in GAUSSIAN-90 that will do natural population analyis? I know that it can be done by explicitly stating the non-standard route, but this is such a pain. You don't need a non-standard route, even though you want to run an extra link. Just add EXTRALINKS=L607 to your standard route to get the default NBO analysis. You have to set an option (using the IOP keyword) if you want the NBO code to read input. In G92, you just say POP=NPA, POP=NBO, or POP=NBORead for Natural population analysis, full NBO analysis, or reading input to the NBO code. Mike Frisch ------- --- Administrivia: This message is automatically appended by the mail exploder. CHEMISTRY@ccl.net --- everybody; CHEMISTRY-REQUEST@ccl.net --- coordinator only OSCPOST@ccl.net : send something from chemistry; FTP: www.ccl.net --- From chemistry-request@ccl.net Fri Jul 31 20:06:33 1992 Date: Fri, 31 Jul 92 15:39:42 EDT From: m10!frisch@uunet.UU.NET (Michael Frisch) Subject: frozen-core derivatives To: chemistry@ccl.net Status: R Fellow netters: I have recently had the chance to use Gaussian92 for an MP2 analytical hessian run on a structure that I had optimized using Gaussian88. Much to my surprise, I found that the default is to use the frozen core unstead of the full correlation option. It was my understanding (however wrong this may be) that the full correlation is needed when doing MP derivatives. I have not been able to track down a paper that discusses the use of the frozen core for the derivatives. If someone out there knows of the reference or understands the mathematics behind this, I would be most appreciative of any help. Also, any experience between the difference in structures and frequencies obtained by the two methods would be useful. Frozen-core gradients of CISD energies were published by Rice and Amos several years ago. Versions of Gaussian through G90 could only do gradients of correlated energies with the full set of orbitals. This led to inconsistencies such as MP single points default to frozen-core but optimizations defaulting to full. Gaussian 92 can do frozen core gradients for all post-SCF methods, and can also do second derivatives for MP2(FC). So the default is now frozen-core for all correlated methods all the time. In addition to being simpler and more consistent, there are substantial performance improvements >from using frozen-core. The techniques used to do this for gradients are a generalization of the previous gradient work, while the second derivatives involve some additional new ideas. The paper describing this in detail is still in preparation. The differences between FULL and FC structures are quite small as compared to the errors these methods make in predicting structures, however, frequency calculations should be done with the same set of orbitals as the preceeding optimization to ensure that modest differences in forces don't affect the vibrational frequencies. This means that if you have a structure done with G90, you have two choices: a) run a new frozen-core optimization and then frozen-core frequencies, or b) run the frequencies with FULL rather than frozen-core. In G92, the savings in using frozen-core in the MP2 frequencies is modest -- the main value is that once can use frozen-core in the optimization. Consequently, the choice between a) and b) depends on how many more structures you plan to optimize and whether you have the checkpoint files >from the MP2(FULL) optimizations you've already done. If you have the checkpoint files, then MP2 OPT=ReadFC Geom=Check should get you from the old (FULL) to the new (FC) structure in just a couple of steps. If you don't have force constants lying around, then the FC optimizations will be more costly and you may prefer to use FULL on your remaining structures rather than re-optimize the ones which you've already done. Mike Frisch ------- --- Administrivia: This message is automatically appended by the mail exploder. CHEMISTRY@ccl.net --- everybody; CHEMISTRY-REQUEST@ccl.net --- coordinator only OSCPOST@ccl.net : send something from chemistry; FTP: www.ccl.net ---