Multiwfn forum

wham09

Member

Registered: 2025-09-19

Posts: 62

Advantages of (meta-)GGA functionals

Dear Prof. Lu,

If I understood your blog post correctly (http://sobereva.com/272), (meta-)GGA functionals are generally recommended when there is a risk of static correlation in ground state calculation, while not really recommended for anything else. (Please correct me if I'm wrong)

But I also know that nowadays, range-separated functionals such as wB97M-V are very sophisticated. And if I think about the concept of range-separation, I don't really see how a good-enough range-separated functional should be any worse than (meta-)GGA functionals even for moderate static correlation. And if static correlation was really severe, DFT will not be suitable anyways.

I understand the merits of B97-3c and r2SCAN-3c, but my question is: Does a traditional (meta-)GGA functional like PBE or TPSS have any reason to be utilized for non-periodic, non-TM-cluster calculations if I only care about accuracy?

sobereva

Tian Lu (Multiwfn developer)

From: Athens, Greece

Registered: 2017-09-11

Posts: 2,217

Website

Re: Advantages of (meta-)GGA functionals

Traditional pure functionals like PBE and TPSS nowdays have little value for calculating isolated systems, except for transition metal clusters. In addition, when PBE and TPSS are suitable for use, using r2SCAN instead is likely a better choice.