VTZ basis set Rb2

[Gileadc]

Dear all molpro users and developers

I want to calculate the excited state - PES of Rb2. I originally thought that I should use a cc-pVTZ for the basis set and close (or freeze) all core orbitals (I have done the same thing with K2 and Na2 and it worked alright). Unfortunately it seems that there is no VTZ basis for Rb, only cc-pVTZ-DK and PP. Hence I am asking what exactly is the difference between these sets? From my crude understanding DK is for relativistic effects, can it be used in a system where no relativistic effects are taking place? 

Thanks in advance

Gilead

[Peterson, Kirk]

Dear Gilead,

 

Rb is definitely heavy enough that you should be worried about scalar relativistic effects.  For all-electron work, the cc-pVTZ-DK is what you want (just remember to set dkho=2 after your basis set block).  If you want to use something based on a relativistic ECP, than cc-pVTZ-PP is the choice.  Just setting basis=cc-pVTZ-PP will also pick up the correct ECP parameters automatically.

 

best regards,

 

-Kirk

 

PS - also, for alkali dimers you probably should be correlating the outer-core electrons (4s and 4p for Rb) and then one should use cc-pwCVTZ-DK or cc-pwCVTZ-PP.  I would also test the effect of additional diffuse functions by prefixing "aug-" in front of either the valence or CV sets.

--
You received this message because you are subscribed to the Google Groups "molpro-user" group.
To unsubscribe from this group and stop receiving emails from it, send an email to molpro-user...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/molpro-user/6728099d-44d7-496d-8541-85f3a0d1121fn%40googlegroups.com.

[Gileadc]

Dear Kirk, Thank you for the fast reply!

For the moment being I will try using an all-electron approach. So as you said, since in Rb I should be worried about relativistic effects, apart from the basis set consideration, do I need to take relativity into account in any other way? can you please abbreviate further on setting "dkho=2" after my basis set block (what is the purpose of it and so on)?

as for your postscript, thanks for the advice, I will try using cc-pwCVTZ-DK. As for diffuse functions, I am intentionally omitting them so higher excited states would not have avoided crossing with Rydberg states near the ion curve.

Best regards

Gilead