August 4, 2026
Two new arXiv preprints (both under consideration as journal submissions) summarize our recent work on more accurate electronic structure methods for hybrid peroskites.
Wentao Zhang built on Rundong Zhao's earlier work in our group to advance fully relativistic DFT - the quasi-four-component (Q4C) approach - to extremely large systems:
A Large-scale Parallel Implementation of Quasi-Four-Component Relativistic Density Functional Theory with Numeric Atom-centered Orbitals
Like it or not, the most important halide perovskites today contain heavy elements like lead or bismuth. Their p-orbitals in particular are crucial for excitations and they are heavily affected by relativistic effects. The Q4C approach employs the Dirac-Coulomb Hamiltonian, not approximate scalar relativity or approximate spin-orbit coupling and it is now applicable to large enough system (above 3,000 atoms shown) to capture the structure sizes needed for our perovskite work.
Aaron Schankler pushed a separate development over the finish line, concerning dispersion interactions in halide perovskites:
Halide perovskites need an appropriate inclusion of dispersion interactions for adequate electronic structure theory. We have been quite successful in relying on the simple but effective Tkatchenko-Scheffler approach in our work, but unfortunately, the original parameterization overestimated the dispersion interactions associated with alkali metals. And indeed, in many halide perovskites, cesium plays an central role. Two revised versions, which we have used for several years now, address this issue, but for broader use, a clear benchmark of any new density functional (which this technically is) is essential. Here is that benchmark.