search

hackingmaterials / atomate

atomate is a powerful software for computational materials science and contains pre-built workflows.
https://hackingmaterials.github.io/atomate
Other
245 stars 175 forks source link

R2SCAN Workflow has inconsistent vdW settings #774

Closed Andrew-S-Rosen closed 1 year ago

Andrew-S-Rosen commented 1 year ago

https://github.com/hackingmaterials/atomate/blob/387938da7c578481c41aa4cfeb65a166778e6372/atomate/vasp/fireworks/core.py#L236-L246

In this code, BPARAM is set to 15.7 if the user specified vdw in their input params. However, this is the parameter to get SCAN+rVV10. If you want r2SCAN+rVV10, BPARAM should be set to 11.95. So, this value should vary depending on the functional. See https://www.vasp.at/wiki/index.php/Nonlocal_vdW-DF_functionals

CC @janosh, @rkingsbury

janosh commented 1 year ago

@arosen93 Good catch! As a VASP novice, IIUC b being too large causes underbinding?

Increasing C or b generally results in smaller vdW correction.

Andrew-S-Rosen commented 1 year ago

It just leads to the wrong functional. Each functional is paramaterized against test benchmarks, and the ideal vdW paramaters for SCAN and r2SCAN are different.

rkingsbury commented 1 year ago

Great catch @arosen93 ! This is a relic from when I originally wrote the workflow, before r2 SCAN had been released. Should definitely be updated now with logic to pick the correct parameters based on functional.