[molpro-user] fe basis set

[Grant Hill]

Dear Jen,

Part of the reasoning behind this is that F12 calculations on transition metal containing complexes are not particularly well tested, meaning selection of the various auxiliary fitting basis sets is left to the user.

You don't state which orbital basis you'd really like to use, but assuming something like aug-cc-pVTZ I would use def2-qzvpp/jkfit for the exchange fitting and aug-cc-pVTZ/mp2fit for the RI. I haven't tried this myself, so I've no idea how large the RI error would be in this case (or if the def2-qzvpp/jkfit set would be much different in the RI).

See http://www.molpro.net/info/2012.1/doc/manual/node432.html for more specifics on how to do this. Depending on the system you may want to define 'sets' in your basis block (http://www.molpro.net/info/2012.1/doc/manual/node133.html#sec:dfit).

Best regards,

Grant


On 17 Nov 2012, at 15:57, jen schmitt <jschmitt1531 at gmail.com> wrote:

> Hi I am new to molpro. 
> I have been trying to run a ccsd(t)-F12
> However my compound contains Fe and i get the following error
> 
> Cannot find default basis SMALL for atom Fe
> Type=FIT
> Context=JKFIT
> 
> I have tried numerous different basis sets that say they are for Fe and always obtain the same error either for basis set or for df_basis_exch
> How do i set it up?
> 
> -- 
> Jennifer
> _______________________________________________
> Molpro-user mailing list
> Molpro-user at molpro.net
> http://www.molpro.net/mailman/listinfo/molpro-user