1.Diabatic potential energy matrices1
We use diabatic potential energy matrices, a generalization of potential energy surfaces, to study nonadiabatic chemistry. I spent times in Prof. Yarkony's group which is known for making probably the most dedicated potential energy matrices from derivative coupling diabatization. My recent focus is on the application of artificial neural networks and the extension to dipoles and spin-orbit couplings.
2.Spin-orbit coupling induced dynamics2
We have successfully constructed the spin-orbit coupling matrix elements from relativistic electronic structure theory. The spin-orbit coupling forms the spin-orbit conical intersection (SOCI). However, due to the fine-splitting of near degenerate states, the SOCI are not exactly a replica of conical intersection. Instead they present unique topography for spin-polarization simulations.
3.Reduced density matrix theory
Coulson's challenge discussed the possibility of using reduced density matrix to replace the traditional wavefunction methods. For molecular systems the energy is a linear function of two-particle reduced density matrix. The major challenge is the N-representability, which constrains the set of reduced density matrix to represent a realistic N-particle system.
4.Quantum computers for quantum chemistry
Quantum computers presents unique opportunities for chemical simulations. From electronic structure to machine learning potential to quantum dynamics, we explore the possibility of "quantum advantage". Special link between quantum state tomography and reduced density matrix is the focus at this point.
1.Wang, Y.; Xie, C.; Guo, H.; Yarkony, D. R. A quasi-diabatic representation of the 1,21A states of methylamine, J. Phys. Chem. A 2019, 123, 5231−5241
2.Wang, Y.; Guan, Y; Guo, H.; Yarkony, D. R. Enabling complete multichannel nonadiabatic dynamics: A global representation of the two-channel coupled, 1, 21A and 13A states of NH3 using neural networks, J. Chem. Phys. 2022, 154, 094121