band gap calculation
62 views
Skip to first unread message
박기명(일반대학원 물리학과)
Jul 23, 2025, 8:19:54 AMJul 23
to BerkeleyGW Help
Dear all,
I performed GW band gap calculations for silicon as a test case, and the results showed good agreement with experimental values (1.1–1.2 eV). Using the same procedure, I calculated the band gap of perovskite material including spin–orbit coupling (SOC). However, the computed band gap was significantly underestimated by about 0.5–0.6 eV compared to reported experimental electronic band gaps (not optical gaps).
In this case, I first performed SOC-DFT calculations, followed by epsilon and sigma steps. I used a dielectric cutoff energy of 20–30 Ry and included 3000–4000 bands, which should be sufficient for convergence.
Are there any methodological issues or additional considerations I should be aware of when including SOC in GW calculations?
Or are there any possible causes that might explain this underestimation when SOC is included in GW calculations?
Best,
GiMyung Park
I performed GW band gap calculations for silicon as a test case, and the results showed good agreement with experimental values (1.1–1.2 eV). Using the same procedure, I calculated the band gap of perovskite material including spin–orbit coupling (SOC). However, the computed band gap was significantly underestimated by about 0.5–0.6 eV compared to reported experimental electronic band gaps (not optical gaps).
In this case, I first performed SOC-DFT calculations, followed by epsilon and sigma steps. I used a dielectric cutoff energy of 20–30 Ry and included 3000–4000 bands, which should be sufficient for convergence.
Are there any methodological issues or additional considerations I should be aware of when including SOC in GW calculations?
Or are there any possible causes that might explain this underestimation when SOC is included in GW calculations?
Best,
GiMyung Park
Reply all
Reply to author
Forward
0 new messages