Quantum Simulation of Strongly-Correlated Molecules with Rydberg Atom Arrays

One of the promising applications for near-term programmable quantum simulators is studying quantum chemistry and materials problems. In this project, we consider a hybrid quantum-classical research pipeline for studying low-energy properties of certain molecules with strong spin correlations. The procedure starts with an effective model Hamiltonian, such as a Heisenberg model, governing the low-energy spin behavior, computed with traditional quantum chemistry techniques. Then, a programmable quantum simulator can be used to probe the spectrum and eigenstates of the model. To efficiently simulate the dynamics, we develop a hardware-efficient encoding of higher-spin Heisenberg models for Rydberg atom arrays, utilizing multi-qubit gates, atom reconfiguration, and Floquet engineering. Then, we demonstrate how to extract both the spectrum and eigenstate properties of the model via snapshot measurements and ancilla-assisted control, which accesses exponentially many two-time correlation functions. This information can be used to compute important chemical information, such as relative energy splittings between low-energy spin states, and magnetic susceptibilities. Finally, we propose proof-of-concept experiments studying organo-metallic catalysts and single-molecular magnets for demonstrating these techniques. Our results provide a framework and roadmap for how quantum simulation can be used in collaboration with computational chemistry to study frontier problems in quantum chemistry and materials science.