arXiv:2609.28369
Simulation of a Battery Cell on Quantum Computers: Reactions & Transport
Albert J. Pool, Michael Schelling, Birger Horstmann
Abstract
Simulations of electrochemical materials and systems accelerate technological progress, but are still limited by computational power. In particular, quantum computing offers prospects for higher resolutions, due to the exponential amount of data that can be stored in a quantum state. As current quantum computers are still noisy, we consider a hybrid quantum-classical algorithm, that divides the problem into smaller computational tasks. We describe how to implement such an algorithm for non-linear partial differential equations, the Feynman-Kitaev Hamiltonian, in a scalable way for an electrochemical system. We show how it can evaluate general electrochemical models and present a quantum simulation of the Single Particle Model with electrolyte (SPMe) as the first quantum simulation of a battery cell.
AI Summary
The paper asks how a coupled electrochemical battery model can be encoded in a variational quantum PDE solver. It treats the Single Particle Model with electrolyte (SPMe), combining spherical diffusion in representative electrode particles with a linearised Butler–Volmer reaction law and an algebraic fixed-current constraint. Electrolyte contributions to cell voltage are calculated analytically. The construction combines a Feynman–Kitaev spacetime formulation with quantum nonlinear processing units. It introduces circuits for spherical diffusion, nonperiodic Neumann boundaries, and concentration–potential coupling. Surface reactions are distributed over a thin boundary region rather than applied at a boundary point. Parameterised circuits encode the spacetime solution, with classical optimisation minimising the residual cost. Noiseless demonstrations use three spatial and three temporal qubits, representing an $8\times8$ grid, for constant-current charging at approximately $4.13$C and a current pulse followed by relaxation. Reported concentration-state infidelities against numerical reference solutions range from $3.0\times10^{-7}$ to $2.1\times10^{-5}$; potentials and derived cell voltages reproduce the reference trends. Optimisation uses statevectors and classically evaluated costs, reporting the best of 20 initialisations. A separate $4\times4$-grid sampling test uses an IBM fake backend with 2.5 million shots and M3 mitigation; reduced-connectivity circuits show less noise. The stated limitations include small grids, concentration-profile deviations, simplified reaction kinetics, and optimisation as a practical bottleneck of unclear complexity. Linear circuit-depth scaling does not establish end-to-end computational advantage. Higher spatial resolution is needed to better resolve relaxation dynamics, while reducing circuit complexity for noisy hardware can impair trainability.
Central claim may not hold: The claimed scalable implementation is supported by encoding and circuit-size arguments, but end-to-end solution complexity remains unestablished because optimization and measurement costs are not controlled. The demonstrated simplified SPMe does not establish the abstract's broader claim of evaluating general electrochemical models. The principal battery solutions use classically evaluated statevector optimization, while the main-text noisy experiment samples already optimized circuits rather than executing the full hybrid solver. The listed anode parameters imply a dimensionless diffusion coefficient of 0.9 rather than 1, and the GITT caption labels a dimensionless duration as seconds. The stated diffusion equation and flux boundary condition predict anode depletion and cathode accumulation for the listed currents, opposite to the reported charging behavior.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 9 pages, 5 figures