Surface Reaction Simulations for Battery Materials through Sample-Based Quantum Diagonalization and Local Embedding

Abstract: Accurate quantum chemistry calculations are essential for understanding electronic structures, yet exactly solving the many-body Schr\"odinger equation remains impractical due to electron correlation effects. While Density Functional Theory (DFT) is widely used, its inherent approximations limit its ability to capture strong correlation effects, particularly in surface reactions. Quantum computing offers a promising alternative, especially for localized electron interactions. In this work, we apply a quantum embedding approach to study the Oxygen Reduction Reaction at Lithium battery electrode surfaces. We employ an active space selection strategy, using Density Difference Analysis to identify key orbitals, which are subsequently refined with coupled-cluster natural orbitals. These are treated using Sample-Based Quantum Diagonalization (SQD) and its extended version (Ext-SQD), leveraging the Local Unitary Cluster Jastrow (LUCJ) ansatz for efficient state preparation on quantum hardware. Ext-SQD significantly improves upon standard SQD by incorporating excitation operators into the quantum-computed electron configurations used in classical Quantum Selected Configuration Interaction. Our approach demonstrates improved accuracy over Coupled Cluster Singles and Doubles (CCSD) calculations over ground state energies. The results, validated against Complete Active Space Configuration Interaction (CASCI), whenever computationally feasible, and Heat-Bath Configuration Interaction (HCI), provide new insights into $\mathrm{Li}$-$\mathrm{O_2}$ surface reactions, underscoring the potential of quantum computing in materials discovery and reaction modeling.