Si GW corrections

[Laura Caputo]

Hello,

I've been doing some calculations to calculate the band gap of Si (following the tutorial:http://www.yambo-code.org/wiki/index.php?title=Silicon) with PPA.
On the website, it seems that the k-point mesh should be converged at circa 500 k-points and 50 bands. Doing the calculation, with a 14x14x14 k point grid, I could obtain an indirect band gap of 1.11eV which should be correct for Silicon. However, when I tried to go to denser k grids like 16x16x16, I noticed that this gap becomes smaller (1.09eV) which shouldn't be correct for Si.

In attachment the log.

[Daniele Varsano]

Dear Laura,

20meV I would say that is in the range of the precision of the GW method. Probably, in order to be totally confident of your results you should push also other parameters (bands, dielectric matrix size..).

Best,
Daniele