search

MolSSI-BSE / basis_set_exchange

A repository for quantum chemistry basis sets
https://molssi-bse.github.io/basis_set_exchange/
BSD 3-Clause "New" or "Revised" License
153 stars 44 forks source link

mTZVPP basis for r2SCAN-3c #152

Open susilehtola opened 3 years ago

susilehtola commented 3 years ago

This is a modification of the mTZVP basis (#151) for the r2SCAN-3c method, as described in J. Chem. Phys. 154, 064103 (2021)

susilehtola commented 3 years ago

@awvwgk do you have this as well?

awvwgk commented 3 years ago

Here are the collected version of the mTZVPP basis: def2-mTZVPP.basis.txt def2-mTZVPP.basis.txt def2-mTZVPP.jbas.txt

susilehtola commented 3 years ago

Looks like the def2-mTZVPP basis file is missing a final *

susilehtola commented 3 years ago

Looks like this set doesn't have elements 87-89 i.e. Fr, Ra, and Ac, though.

awvwgk commented 3 years ago

It is probably safe to alias everything beyond 86 with the def2-TZVP basis set. r²SCAN-3c itself is not available beyond Rn anyway since the DFT-D4 model is only available up to Rn.

william-dawson commented 2 years ago

I would also like to have access to mTZVPP (as well as mTZVP) available in the json format. It looks like you have everything you need here, but is there any way I can help finish this?

susilehtola commented 2 years ago

@william-dawson It's been a while since there was activity on this, but looking at #197 it looks like the issue was that the files that @awvwgk sent run all the way to Z=103 but the BSE only has the ECPs up to Z=86...

awvwgk commented 2 years ago

Sorry for the confusion, I just pulled everything out of Turbomole's basis set library which was labeled as def2-mTZVP and def2-mTZVPP. The def2-mTZVP and def2-mTZVPP are aliases for the def2-TZVP basis set for the heavier elements, it seems that the def2-TZVP is mislabeled for everything higher than Z>86. The clarification in https://github.com/MolSSI-BSE/basis_set_exchange/issues/195#issuecomment-806231938 did settle the issue for me, therefore I didn't follow up here.

Since r²SCAN-3c is in any case only defined up to Z=86, due to the D4 and gCP components, the discussion over the basis set beyond this point is not really meaningful.

Is there anything left blocking the addition of the mTZVP and mTZVPP basis set which needs resolution?

william-dawson commented 2 years ago

Sounds good to me, I think this (mTZVPP) and #151 (mTZVP) should be good to go!