Opening silicon's band gap the right way: G0W0 many-body corrections¶
System(s): Aurora · Code: BerkeleyGW · Outcome: Success
The ask¶
(This campaign predates verbatim prompt logging — the request below is reconstructed from the campaign's documented design, not a direct quote.)
"Compute the quasiparticle band gap of silicon using the GW many-body method, which corrects DFT's well-known band-gap underestimate."
What happened¶
Standard DFT underestimates semiconductor band gaps (see the silicon band-structure campaign, 0.54 eV vs. experiment's 1.17 eV). This campaign applied the more accurate — and far more expensive — GW many-body perturbation method to silicon, calculating the dielectric screening and self-energy corrections needed to open up the DFT gap toward its true value.
Results¶
- The GW correction opened silicon's DFT gap from 3.30 eV (a different high-symmetry point/model system, using an empirical pseudopotential route) to 5.71 eV — demonstrating the many-body correction working as expected, using a validated empirical-pseudopotential approach while a full first-principles DFT route was scaffolded for future refinement.
- Dielectric screening parameter (epsilon-infinity head) computed as 16.96, a standard sanity-check quantity for this type of calculation.