Ab inito/DFT calculations on non-unity multiplicity metal complexes

["Sheldon M. Williams" <smwilliams ( ( at ) ) mail.cm.utexas.edu>]

 Hello,
 I have been having problems with SCF convergence and running out of memory
 trying to perform ab inito geometry optimization calculations on Cu(II)
 and Fe(III) ion complexes with crown ethers. I run MMFF conformational
 searches with the ion constrained near the center of the crown ether ring
 to get close to the global minimum, as I get poor configurations with low
 level calculations if I do not contrain these metal ions. I then do a PM3
 geometry
 optimization, if possible. After this I have attempted both ROHF/ and
 UOHF/3-21G*, 6-31G*, and LACVP* geometry optimizations with SCF cycles set
 to up to 5000, but the calculation always runs out of memory. I'm
 performing the calculations with Spartan PC 2002 on a 2.0GHz P4 with 1.0GB
 of SDRAM. Trying to perform a DFT calculation directly from the MMFF or
 semi-emperical geometry has worse SCF convergence and memory problems.
 It appears to be a problem of the multiplicities of these complexes as
 Ag(I) and Cu(I) work. Almost all papers on ab initio or DFT calculations
 on transition metal-ligand complexes use the oxidation state with
 singlet multiplicity. I haven't gotten any useful info from support at
 wavefunction. I would be grateful if you could suggest what the
 problem might be or direct me somewhere I could find help?
 Thank you for your time and attention,
 Sheldon M. Williams
 Ph.D. Candidate, Analytical Chemistry
 Jennifer Brodbelt Lab
 Chem & Biochem Department
 1 University Station A5300
 University of Texas at Austin
 Austin, TX 78712
 Phone:(512)471-0041
 FAX: (512)471-8696
 E-mail: smwilliams ( ( at ) ) mail.cm.utexas.edu