Ab-initio simulation of excited-state potential energy surfaces with transferable deep quantum Monte Carlo

These authors contributed equally to this work. Abstract The accurate quantum chemical calculation of excited states is a challenging task, often requiring computationally demanding methods. When entire ground and excited potential energy surfaces (PESs) are desired, e.g., to predict the interaction of light excitation and structural changes, one is often forced to use cheaper computational methods at the cost of reduced accuracy. Here we introduce a novel method for the geometrically transferable optimization of neural network wave functions that leverages weight sharing and dynamical ordering of electronic states. Our method enables the efficient prediction of ground and excited-state PESs and their intersections at the highest accuracy, demonstrating up to two orders of magnitude cost reduction compared to single-point calculations. We validate our approach on three challenging excited-state PESs, including ethylene, the carbon dimer, and the methylenimmonium cation, indicating that transferable deep-learning QMC can pave the way towards highly accurate simulation of excited-state dynamics. Light-driven phenomena are also key to technological advancements, ranging from material design and chemical processing [4, 5] to biomedical technologies such as molecular motors and photo-controlled drug delivery [6, 7]. Despite the critical importance of these processes, their theoretical study is hindered by the need for accurate ab-initio descriptions of electronic excited states. Most quantum chemistry methods have been developed for the calculation of electronic ground states and their extensions to excited states are either limited or highly expensive and often require expert knowledge [8, 9].