Quantum Multiscale Modeling: A Hierarchy of Algorithms for Complex Chemical Systems

Computational chemistry faces a fundamental hurdle: physical phenomena—from the electronic interactions governing bond breaking to the macroscopic fluid dynamics in a reactor—span multiple orders of magnitude in time and length. Classical multiscale modeling bridges these gaps by passing information (like energy barriers or rate constants) between distinct algorithmic layers. However, this process is inherently lossy, as each scale transition requires approximating complex quantum mechanical data into simplified classical parameters.

This paper proposes a systematic framework for composing fault-tolerant quantum algorithms across these scales. The authors map specific quantum algorithms to each regime: Quantum Phase Estimation (QPE) for electronic structure, Gibbs state preparation and Hamiltonian simulation for atomistic dynamics, quantum random walks for mesoscopic kinetics, and Quantum Singular Value Transformation (QSVT) for continuum reactor physics. By mapping these, the authors illustrate how a coherent quantum pipeline could theoretically replace the classical hierarchy.

Crucially, the authors argue that individual quantum speedups at each scale are insufficient for end-to-end advantage. If a quantum algorithm at the electronic scale produces a high-precision result that is then measured and reduced to a classical rate constant, the quantum advantage is effectively discarded. The paper shifts the focus from isolated algorithm performance to the design of inter-scale quantum channels. By using a case study of CO oxidation on Pt(111), the authors demonstrate how information—such as adsorption energies—can be passed between scales as coherent quantum registers, potentially preserving quantum information throughout the entire simulation pipeline.