From chemistry-request@ccl.net Sat Apr 18 06:20:21 1992 Date: Sat, 18 Apr 92 18:00:42 +1000 From: Brian Yates Subject: ampac/mopac geometry optimisation problems To: chemistry@ccl.net Status: R In MOPAC another way around the problem is to use the keyword XYZ. This instructs the optimization to be carried out in cartesian coordinates. brian yates. From chemistry-request@ccl.net Sat Apr 18 18:21:38 1992 Date: 18 Apr 92 15:25 LCL From: PA13808%UTKVM1.BITNET@OHSTVMA.ACS.OHIO-STATE.EDU Subject: BITNET mail follows To: CHEMISTRY@ccl.net Status: R MOPAC GEOM OPT A QUICK SOLUTION Whenever I ran into a problem with large floppy cyclic molecules using Mopac i did the same calc using the AM1 or MNDO option on the latest set of Gaussian prog. The math tecniques are entirely differe nt and probably more precise since the tolerences for convergence are much greater.(ANDY H. correct me if I am wrong). Anyway whatever the reason it worked. In one or two cases especially when the semiemp. predicted a geometry which I thought was unreasonable the G. prog gave me a more reasonable result. The difference in computer time was ma maybe a factor of two. (Since 2*0=0 I decided this was more efficient with respect to my time than fiddling with the Z matrix. However, I believe there is a recent paper on how to obtain a Z matrix for cyclic molecules probably in J comp. Chem 1991 (John E. Bloor, pa13808 at UTKVM1) From chemistry-request@ccl.net Sat Apr 18 21:20:32 1992 Date: Sat, 18 Apr 1992 16:40 EST From: IRIKURA@ENH.NIST.GOV Subject: Density functional theory: molecular software? To: chemistry@ccl.net Status: R I'm interested in obtaining a DFT package to run on an RS/6000-320 workstation or possibly on a Cray. What commercial and non-commercial software is available? The hope is to use these programs for molecules that are too large for the traditional approach (e.g., Gaussian, GAMESS). Thanks for any help! Karl Irikura irikura@enh.nist.gov (301) 975-2510 From chemistry-request@ccl.net Sat Apr 18 22:26:37 1992 Date: Sat, 18 Apr 1992 18:56 EST From: "DOUGLAS A. SMITH" Subject: AM1 and Z-matrix To: chemistry@ccl.net Status: R One problem which must always be avoided in the definition of a Z matrix is defining your rings simply as a ring. In other words, if you have a six membered ring with atoms numbered 1-6 (clockwise) so that the ring closure is defined by 6-1, you should not define your Z matrix such that the atomic distance column includes 1-2, 2-3, 3-4, 4-5 and 5-6, nor should the bond or dihedral angles be defined using such a scheme (i.e. 3-2-1, 4-3-2, or 4-3-2-1, etc.). This can lead to small changes at the beginning of the sequence causing large distance or angle changes at the end of the sequence. (Anybody remember playing crack the whip as a child?) I am not sure how PCModel defines rings, but if it does what I described above, it might lead to one of the errors originally posted here. We now use the AVS Chemistry Viewer to set up many of our jobs, and do not get poorly defined Z matrices. Doug Smith Assistant Professor of Chemistry The University of Toledo Toledo, OH 43606-3390 voice 419-537-2116 fax 419-537-4033 email fax0236@uoft02.utoledo.edu From chemistry-request@ccl.net Sun Apr 19 00:21:35 1992 Date: Sat, 18 Apr 1992 21:05 CST From: Andy Holder Subject: Semiempirical optimizations To: CHEMISTRY@ccl.net Status: R In reply to John Bloor, The criteria are not more stringent in GAUSSIAN than in AMPAC. What is different is the optimization algorithm. I believe that the default optimizer in GAUSSIAN is what is known as EF (eigenvector following) on the semiempirical side. (Someone please correct me if I am wrong. This is only an impression not a fact.) If this is the case, John McKelvey pointed this out before as a solution. I am looking at his other fix and trying to figure out just what overall effect it will have. I think that Henry Rzepa's implementation of EF will find its way into the next version of AMPAC. Andy Holder =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= DR. ANDREW HOLDER Assistant Professor of Computational/Organic Chemistry Department of Chemistry || BITNET Addr: AHOLDER@UMKCVAX1 University of Missouri - Kansas City || Internet Addr: aholder@vax1.umkc.edu Spencer Chemistry, Room 502 || Phone Number: (816) 235-2293 Kansas City, Missouri 64110 || FAX Number: (816) 235-1717 =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=