CCL: all-electron basis sets for Au; ORCA, TURBOMOLE, MOLPRO



 Sent to CCL by: Gerald Knizia [knizia^_^theochem.uni-stuttgart.de]
 On Monday 05 January 2009 09:20, Berger Raphael berger,chem.helsinki.fi wrote:
 >   3. Molpro-related:
 >                    Orca prints also the respective RI-auxilliary basis set:
 >                    Type Au : 25s9p9d5f4g contracted to 25s9p9d5f4g pattern
 >
 > {1111111111111111111111111/111111111/111111111/11111/1111} So I thought of
 > using DF- methods and finally maybe even explicitly correlated methods
 > (F12), but these require besides the DF_BASIS also an RI_BASIS. What could
 > be a sensible choice for that?
 1) At this moment F12 methods in MOLPRO will not work in conjunction with ECPs
 or DKH Hamiltonians (currently, dipole fields are the only supported core
 Hamiltonian patches. In F12 methods this issue is more complicated than it
 looks). This will likely change in the future, but not within the next few
 weeks.
 2) Basis sets for RI: MP2FIT basis sets from the Ahlrichs group are available
 for pretty much all elements, and these would probably do the trick if used
 as RI bases.
 We typically used Weigend's JKFIT basis sets for RI purposes, but if I'm not
 mistaken, these are only available for 1st- and 2nd-row main group elements.
 Kirk A. Peterson recently developed some new Optimized RI basis sets, but
 these are also only available for this limited set.
 --
 Gerald Knizia