CCL: Ab initio on bridged rings

 Sent to CCL by: "Wayne Steinmetz" [WES04747:+:pomona.edu]
 You are having problems with convergence so your system shows
 sensitivity to structural details.  I recommend that you first generate
 a structure using molecular mechanics and a good Class II force field,
 e.g. the Merck Molecular Force Field or Allinger's MM3.  Sometimes,
 intuition in initially drawing the structure leads one astray.  For
 example, when I drew the structure of calixarene, I started with a
 structure quite far in energy from the global minimum.  Hence, after you
 have minimized your initial structure, you might also want to search
 conformational space.  Spartan uses a Monte Carlo algorithm which
 usually works.  In difficult cases such as calixarene, I have used
 distance geometry to get out of a trap.
 You can rely on the results of molecular mechanics if you use a Class II
 force field.  Once you have located the global minimum, then proceed to
 Hartree Fock and then a more advanced method such as DFT with Becke's
 B3LYP functional or Truhlar's M06 functional.  If the convergence
 falters, you might tweak a portion of this structure.  This trick often
 helps getting out of a rut.
 Wayne E. Steinmetz
 Professor Emeritus of Chemistry
 USFS Volunteer
 Chemistry Department
 Pomona College
 645 North College Avenue
 Claremont, California 91711-6338
 USA
 phone: 1-909-621-8447
 FAX: 1-909-607-7726
 Email: wsteinmetz- -pomona.edu
 WWW: pages.pomona.edu/~wsteinmetz
 -----Original Message-----
 > From: owner-chemistry- -ccl.net [mailto:owner-chemistry-
 -ccl.net]
 Sent: Tuesday, August 05, 2008 12:30 AM
 To: Wayne Steinmetz
 Subject: CCL: Ab initio on bridged rings
 Sent to CCL by: "Simon  Halstead" [shalstead:hit.edu.cn]
 Dear all,
 I am trying to optimize some bridged (1-4) cyclohexane rings. I
 initially optimized them using sempiempirical methods. These
 calculations converged very rapidly. However, I then tried to optimize
 the resulting structures at a higher level of theory. These high level
 calculations have failed to converge, despite being run for a large
 number of steps. I have tried HF, MP2 and B3LYP with a variety of basis
 sets but nothing appears to converge. Visualizing the outputs of the
 failed calculations shows no problems with the structure.
 Does anyone have any experience with structures like this or can anyone
 give me a suggestion of how best to optimize bridged rings?
 I am using GAMESS if this makes any differnce.
 Thanks for your help,
 Simon Halstead
 -= This is automatically added to each message by the mailing script =-http://www.ccl.net/cgi-bin/ccl/send_ccl_messageSubscribe/Unsubscribe:
 Job: http://www.ccl.net/jobs http://www.ccl.net/spammers.txt-------------------------------------------------------------
 This message has been scanned by Postini anti-virus software.