[molpro-user] CCSD-F12, density fitting problem for systems containing Helium

[Peterson, Kirk]

Dear Mark,

one workaround is to use the def2-QZVPP/JKFIT set of Weigend for He.  It’s fairly universal, so you can use it for DZ through QZ at least (this set is also good for the alkali and alkaline earth metals where JKFIT sets are also not available for cc basis sets).  In Weigend’s 2002 paper where the cc-pVnZ/JKFIT sets were reported, none of the rare gases were included.  In molpro we’ve generated JKFIT sets for Ne, Ar, and Kr by extrapolation of the JKFIT sets of the previous two elements.  It’s unfortunately not possible to do this for He.

Hope that helps,

-Kirk

On Oct 20, 2016, at 9:12 AM, Pfeifle, Mark Claudius Gerhard <mpfeifle at anl.gov<mailto:mpfeifle at anl.gov>> wrote:

Hello,

The following sample input, just as any other He containing system, will throw an error message at the beginning of the CCSD-F12 part with Molpro 2012 and 2015:

“***,He-He
memory,80,M
angstrom
geometry={
         He1
         He2, He1, dist
         }
dist=10.0 angstrom
basis=aug-cc-pVTZ

rhf

CCSD-F12”

This is the error message:

“Unable to assign basis for exchange fitting.Please specify DF_BASIS_EXCH
? Error
? Unknown df_basis_exch
? The problem occurs in f12_default_bases”

The same message is obtained for the cc-pVTZ-F12 basis. Explicitly specifying df_basis_exch or ri_basis within the CCSD-F12 directive does not help.
Using the def2-TZVPP basis instead helps though. Other noble gases work fine, too, with any of the above-mentioned bases. Looking at the basis set library, it seems like the JKFIT versions of cc-pVXZ are NOT present for the He atom.
Is there any way to work around this (except using the def bases obviously)?

Best regards,
Mark
_______________________________________________
Molpro-user mailing list
Molpro-user at molpro.net<mailto:Molpro-user at molpro.net>
https://urldefense.proofpoint.com/v2/url?u=http-3A__www.molpro.net_mailman_listinfo_molpro-2Duser&d=DQICAg&c=C3yme8gMkxg_ihJNXS06ZyWk4EJm8LdrrvxQb-Je7sw&r=mqaXWsZ0ghQR-pXvPLjY_g&m=4uBx2O5FTLdnavoLPlSG-l3tJDtj9_VVkWSD5056Y_s&s=mELHe_5EJ_KhM9t6yGdqAaJAwGoFIdwltwvXkXcPtsw&e=

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20161022/1a986dcb/attachment.html>