How to get the profile of dielectric constant

[wqtangecust]

Hi,

I want to know by setting which keyword I can get the distribution of dielectric constant.

Thank you in advance.

[TTsangSC]

Hi, it seems that lines 401–27 in solvation.F deals with the cavity, and the statement to write epsilon to the EPS file is usually disabled with an #ifdef debugsol clause. Hence I think you can either:

• Rebuild VASP + VASPsol with -Ddebugsol passed to the preprocessor along with the other -D flags, hence activating the debug switch. That requires minimal change to the code itself (only to the CPP_OPTIONS in the makefile.include), but also has the side effect of always writing said EPS file, along with the adjacent KAPPA2 file. Depending on your disk space limit this may or may not be acceptable since these are all rather chonky volumetric files.

• Patch the solvation.F source code to add your own customized logical flag, say LEPS, to the INCAR specifying whether to write the EPS file; then again rebuild VASP + VASPsol. See lines 3068–9 for an example – the RDATAB() subroutine defined in the VASP source is used to read in LRHOB flag defined by VASPsol. Since the flag only occurs and is parsed here, and doesn't otherwise interfere with other parts of VASP, I think you can safely adapt the snippet to define, read in, and manipulate arbitrary INCAR flags, provided that they don't clash with existing ones.

Unfortunately I don't think there's a way around a rebuild, which of course will require access to the VASP source code (plus some elbow grease).