ws cutoff CUTwsGvec

[sunxl]

Dear all，

I am studying a one-dimensional system placed in an orthorhombic cell, where the periodic direction is along z, and 20 Å of vacuum is added in the non-periodic directions. When applying the Coulomb cutoff, I use the ws Z scheme.

1、In Yambo, when using the ws Z cutoff, the parameter CUTwsGvec has a default value of 0.7 Ry. I want to examine whether different values of CUTwsGvec affect the results. Therefore, I set it to 0.7 Ry and 1.0 Ry, respectively. The resulting G₀W₀ calculations are identical, but the G₀W₀-BSE results show significant differences, leading to completely different exciton distributions. Should I gradually increase CUTwsGvec to achieve convergence? What criteria should be used to determine convergence?

2、I would like to perform GW corrections on top of a PBE0 calculation. Is the parameter CUTwsGvec equivalent to ecutvcut in QE, and should they be kept consistent during calculations?

Best，
sunxi

[Daniele Varsano]

Dear Sunxi,
CUTwsGvec indicates the number of G components of truncated potential are considered: yes it is a convergence parameter that should converge quite fast. CUTwsGvec (Yambo) is exactly the same of ecutvcut (QE). For consistency in a GW calculation it is safe to keep them consistent. For the BSE you can also use a different vlaue.

Best,

Daniele