discrepancies in epsilon calculation results

[subrata]

Dear All,

I have a problem understanding the results from a paper and the results reported in other papers for the same method.
All, calculations are done in VASP

1) In the paper

J. Phys.: Condens. Matter 32 (2020) 015502

Table 2 reported beta = 0.28 for NiO

2) This is for DD-RSH-CAM, which is reported in

PHYSICAL REVIEW MATERIALS 2, 073803 (2018)

This is the method implemented in VASP.

3) Now, let's consider the same method and the same NiO results from

PHYSICAL REVIEW MATERIALS 2, 073803 (2018)

In TABLE IV of their paper, they reported \epsilon = 7.16 for NiO

which is according to \beta = 1/epsilon = 0.13966 for NiO

Already big discrepancies

4) Similarly, in the supporting information of the paper

PHYSICAL REVIEW MATERIALS 9, 053808 (2025)

Table S2, the reported value \beta = 1/epsilon = 0.13966 for NiO

5) This is the discrepancy not only for NiO, but also for CoO.

I did not check MnO

But seems there are code-by-code big discrepancies

6) Regarding \mu values, the results should not change much due to \mu variation, which is also my experience.

7) So, my question is, when we are using DD-RSH-CAM and self-consistently calculating \epsilon, then why are the results varying from code to code? At least all codes are using the same DFPT method.

Thanks
Subrata

[martin.schlipf]

Please contact the original authors of the papers for the input files of the calculations. Perhaps one or more of them are in this forum and can reply to this thread. We cannot provide support for papers that are not written by us.