QuantumATK Forum
QuantumATK => General Questions and Answers => Topic started by: Ángel D. on April 10, 2022, 00:56
-
Hello, QATK community!
I hope I haven't posted on the wrong place. I was comparing some QATK capabilities on a MoS_2 bilayer system, trying to take into account van der Waals forces in order to predict some layer properties. Following many references, I wanted to try optB88. According to the reference given in QATK manual, its exchange part has one parameter, so I guess the corresponding exchange functional is OneparameterB88, right? What about the correlation part? If I'm right, it consists of a local LDA part and a nonlocal part (the most important for vdW in this case, I guess). I noticed there's an available correlation functional named Becke88DionVanDerWaals, but I don't think that's the good one, because it cannot reach convergence (9x9x1 k points and 55 H density-mesh cutoff). Which one would make the famous complete optB88?
Thanks in advance!
-
Nomenclature for XC functionals is a quagmire...! I hope I am right in this, but I believe optB88 is just an exchange functional and has itself no vdW in it. There is a version, or combination, referred to as optB88-vdW as used e.g. in https://www.mdpi.com/2073-4352/9/7/348/pdf (with PBE as correlation) but as far as I know, that requires that vdW is implemented separately in the functional. This is currently not available in QuantumATK, but I hope we can include it in the future.
-
Thank you very much for your answer, sir!