Atomistic insights explain voltage hysteresis in silicon battery anodes

Background

Silicon anodes promise much higher lithium storage than graphite, but their practical use is limited by voltage hysteresis, long relaxation times, and complex Li–Si phase changes. These effects make it difficult to define a reliable open-circuit potential, estimate state of charge, and design battery management strategies for high-energy lithium-ion cells.

New insights

In a recent study, Papadopoulos and co-workers from Porsche, Volkswagen, Leipniz Institute INM, Saarland University, RWTH Aachen University, and Saarene combine electrochemical experiments with atomistic modeling to describe silicon lithiation in the Li-Si system. Their work uses a physically grounded, multi-species and multi-reaction framework in which silicon is not treated as a single empirical voltage curve, but as a cascade of phase-specific redox reactions. By assigning each SiLiₓ phase its own equilibrium potential and kinetic parameters, the model captures asymmetric lithiation and delithiation pathways. The study highlights how ReaxFF-based atomistic modeling can support battery research across both academic and industrial settings.

ReaxFF molecular dynamics in the Amsterdam Modeling Suite was used as part of an atomistic workflow to generate amorphous Li–Si configurations, while optimized crystalline and amorphous structures supplied phase energetics for the electrochemical model. This atomistic input helped connect microscopic Li–Si energetics to macroscopic voltage behavior, providing a route beyond purely phenomenological hysteresis models.

The resulting framework captures key features that conventional Doyle–Fuller–Newman models miss. It reproduces voltage hysteresis, phase-fraction evolution, and relaxation behavior in silicon half-cells, with reported root-mean-square errors from 5.4 to 36.9 mV across constant-current and pulse-relaxation protocols. The simulations also reveal that silicon particles do not simply stop evolving during rest. Lithium continues to redistribute between Li–Si phases until phase-specific equilibrium potentials and kinetics drive the system toward a new thermodynamic balance.

For battery R&D, this is a useful cross-scale connection: atomistic calculations help parameterize phase-resolved electrode models, while electrochemical experiments validate which microscopic assumptions matter at cell scale. Such models can support better interpretation of silicon voltage curves, improved state estimation, and more robust design strategies for next-generation high-capacity anodes.