Dresden 2026 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
MM: Fachverband Metall- und Materialphysik
MM 17: Data-driven Materials Science: Big Data and Workflows II
MM 17.6: Talk
Tuesday, March 10, 2026, 15:15–15:30, SCH/A251
Many-body perturbation theory vs. density functional theory: A systematic benchmark for band gaps of solids — •Marc Thieme1,2, Max Großmann1, Malte Grunert1, and Erich Runge1 — 1Institute of Physics and Institute of Micro- and Nanotechnologies, Technische Universität Ilmenau, 98693 Ilmenau, Germany — 2Institute of Applied Physics, Friedrich Schiller Universität, 07743 Jena, Germany
The band gap is one of the most important material properties for optoelectronic applications. However, predicting band gaps remains a challenging task in materials science. Here, we benchmark many-body perturbation theory against density functional theory, the workhorse of computational materials science, for predicting the band gaps of solids. We systematically compared four GW variants—G0W0 using the plasmon-pole approximation (G0W0-PPA), full-frequency quasiparticle G0W0 (QPG0W0), full-frequency quasiparticle self-consistent GW (QSGW), and QSGW augmented with vertex corrections in W (QSGŴ)—against the currently best-performing and popular density functionals. Our results show that the QSGŴ produces band gaps so accurate that they can even flag questionable experimental measurements, albeit at an extremely high computational cost. To balance accuracy and efficiency, we identify lower-cost alternatives, such as the QPG0W0 and a rescaled version of the QSGW, which achieve nearly the same accuracy as the QSGŴ while being significantly more efficient, making them promising candidates for generating high-fidelity datasets in machine-learning-driven materials discovery.
Keywords: Band Gaps; DFT; Many-Body Perturbation Theory; GW; Benchmark