Charge transfer is essential for all electrochemical processes, such as in batteries where it is facilitated through the incorporation of ion–electron pairs into solid crystals. The low solubility of lithium (Li) in some of these host lattices cause phase changes, which for example happens in FePO₄. This results in the growth of interfacial patterns at the mesoscale between a Li-poor and Li-rich phase, FePO₄ and LiFePO₄ respectively. Conventionally, the effect of charge transfer on the evolution of these phases is usually modelled using the Butler–Volmer equation. However, the exponentially increasing current–overpotential relation in this formalism becomes problematic for battery systems operating under high currents. In this study, we implement a phase-field model to investigate two electrochemical reaction models: the Butler–Volmer and the Marcus–Hush–Chidsey formulation. We assess their effect on the spatial and temporal evolution of the FePO₄ and LiFePO₄ phases. Both reaction models demonstrate similar microstructural patterns in equilibrium. Nevertheless, a significant increase in current density is caused by using the Butler–Volmer expression, leading to an accelerated reaction rate at high overpotentials and an exaggerated delithiation. Furthermore, we show that including anisotropic elastic strain fields in the phase-field model accelerates the delithiation process, reaching the bulk mass transport limitation faster. These elastic effects, when included in the overpotential, can cause the current density to exceed its limits, a problem inherently mitigated by the Marcus–Hush–Chidsey model.

Energy is stored in a LiFePO₄ battery electrode through the intercalation of Li. As Li incorporate into the crystal lattice of Fe⁡(III)⁢PO₄, electrons reduce Fe(III) into Fe(II). The interactions of Li and its vacant site (Va) with these localized electrons (holes), so-called polarons, cause phase separation during battery operation. These fundamental interactions are however difficult to quantify using standard electronic structure calculations. In this paper, we utilize DFT+𝑈 with occupation matrix control to compute interaction energies at varying Li-Fe(II) and Va-Fe(III) pair separations. The increased energy with separation warrants the use of an electrostatic description. The DFT+𝑈 data are fitted to a Coulombic potential with two-body corrections and used in a Monte Carlo scheme. The coordination of the species determines their short-range ordering, showing that the Li and Va create chains bridged by their associated polarons which dissociate into dimers at higher temperatures. This dissociation happens at higher temperatures for Va than for Li, indicating a more pronounced clustering behavior during the formation of FePO₄. Notably, a significant amount of uncoordinated Li exists at elevated temperatures, challenging the simplified picture of complete Li-Fe(II) pairing.