Webinar: A practical guide to GW-BSE and Sigma-functionals in the Amsterdam Modeling Suite

In this third webinar of the AMS 2024 webinar series, Dr. Arno Förster, Assistant Professor at the Vrije Universiteit Amsterdam, is presenting how GW-BSE and new Sigma-functionals in the Amsterdam Modeling Suite enable very fast and very accurate optical excitations, particularly for charge-transfer states, as well as highly accurate energetics at low computational cost.

Abstract

Hedin’s GW approximation[1] to the electronic self-energy gives access to accurate charged excitations and spectral functions. In combination with the Bethe-Salpeter equation (BSE@GW) it also gives access to accurate optical excitations.[2] Especially for charge-transfer states, BSE@GW is highly accurate and often challenges the performance of advanced wave-fucntion based methods at a fraction of the computational cost.[3]

Starting in 2020, BSE@GW has been implemented in AMS, and since then, the algorithms are being continuously refined and improved.[4–6] Practical BSE@GW calculations are complicated by two factors: a dependence of the final results on the DFT starting point, as well as a very slow convergence to the complete single-particle basis set limit.

In this talk, I will first give a short introduction to the BSE@GW method. I will then show how the starting-point dependence problem can be navigated, either through self-consistency in the GW approximation, or through a judiciously chosen starting point. I will then show how the strong basis-set dependence of GW can be overcome using data-driven techniques, which allow for a smooth extrapolation to the complete basis set limit.[7]

Finally, the GW method also gives access to relatively accurate total energies, through the random phase approximation (RPA). The RPA excels in the description of barrier heights and non-covalent interactions and I will show how, at practically no additional cost, the accuracy of RPA calculations can be greatly improved through sigma-functionals.[8] I will introduce the implementation of these functionals in AMS and show that Sigma-functionals should be used for highly accurate energetics at low computational cost.[9]

Webinar Details:
Date: Tuesday 26.11.2024
Time: 17.00 (CET) / 11.00 (ET)

Save Your Spot:

References
[1] L. Hedin, Physical Review 1965, 139, A796.
[2] X. Blase, I. Duchemin, D. Jacquemin, Chem Soc Rev 2018, 47, 1022–1043.
[3] A. Förster, L. Visscher, J Chem Theory Comput 2022, 18, 6779–6793.
[4] A. Förster, L. Visscher, J Chem Theory Comput 2020, 16, 7381–7399.
[5] A. Förster, L. Visscher, J Chem Theory Comput 2021, 17, 5080–5097.
[6] A. Förster, L. Visscher, Front Chem 2021, 9, 736591.
[7] F. Bruneval, I. Maliyov, C. Lapointe, M.-C. Marinica, J Chem Theory Comput 2020, 16, 4399–4407.
[8] S. Fauser, E. Trushin, C. Neiss, A. Görling, Journal of Chemical Physics 2021, 155, 134111.
[9] S. Fauser, A. Förster, L. Redeker, C. Neiss, J. Erhard, E. Trushin, A. Görling, J Chem Theory Comput 2024, 20, 2404–2422.