Large-Scale spin-orbit coupled GW-Bethe-Salpeter equation calculations in AMS
The GW method is a state-of-the-art approach for the calculation of the electronic structure of molecules and can be used to calculate ionization potentials (IP) and electron affinities (EA) with high accuracy. The presence of heavy elements then often necessitates to explicitly consider spin-orbit effects which in can change IPs by up to 0.5 eV compared to scalar relativistic calculations.
Accounting of spin-orbit effects is typically associated with high computational cost. Building on the efficient GW-Bethe-Salpeter equation (GW-BSE) functionality in AMS(1–3), a new implementation of this method in the ADF and BAND modules of AMS now enables the application of fully relativistic GW to molecules of unprecedented size. In a recent paper(4), this new implementation has been used to benchmark a variety of different GW methods against accurate experimental reference data for a large set of molecules containing heavy elements. Already the relatively cheap G0W0 method based on a functional with a relatively high fraction of exact exchange can deliver IPs with mean absolute deviations to experiment below 0.2 eV when spin-orbit effects are taken into account as in the new implementation. Even higher accuracy can be achieved using computationally more demanding partially self-consistent GW methods.
The new implementation also allows for the calculation of excited states including spin-orbit effects and is fully available in AMS 2023.1.
1. A. Förster, L. Visscher, Low-Order Scaling G0W0 by Pair Atomic Density Fitting. J Chem Theory Comput. 16, 7381–7399 (2020).
2. A. Förster, L. Visscher, Low-Order Scaling Quasiparticle Self-Consistent GW for Molecules. Front Chem. 9, 736591 (2021).
3. A. Förster, L. Visscher, Quasiparticle Self-Consistent GW-Bethe-Salpeter equation calculations for large chromophoric systems. J Chem Theory Comput. 18, 6779–6793 (2022).
4. A. Förster, E. van Lenthe, E. Spadetto, L. Visscher, Two-component GW calculations: Cubic scaling implementation and comparison of partially self-consistent variants. J Chem Theory Comput (2023)
