Quantum Machine Learning of Molecular Energies with Hybrid Quantum-Neural Wavefunction

Abstract

Quantum computational chemistry holds great promise for simulating molecular systems more efficiently than classical methods by leveraging quantum bits to represent molecular wavefunctions. However, current implementations face significant limitations in accuracy due to hardware noise and algorithmic constraints. To overcome these challenges, we introduce a hybrid framework that learns molecular wavefunction using a combination of an efficient quantum circuit and a deep neural network. This approach enhances computational efficiency and accuracy, surpassing traditional quantum computational chemistry methods. Numerical benchmarking on molecular systems shows that our hybrid quantum-neural wavefunction approach achieves near-chemical accuracy, comparable to advanced quantum and classical techniques. Experimental validation on a superconducting quantum computer, using the isomerization reaction of cyclobutadiene, further demonstrates its practical applicability, with high accuracy in energy estimation and significant resilience to noise.