Publication date: 15th December 2025
Compositionally complex energy materials display structural and electronic behavior that develops across multiple length scales. In this talk, I will show how advanced atomistic modeling, ranging from first-principles theory to machine-learned interatomic potentials, can help clarify this global and local complexity. I first discuss the global free-energy landscape of FAPbI₃, where large-scale molecular dynamics simulations with a machine-learned potential resolve a previously unknown low-temperature structure and reveal kinetically trapped formamidinium configurations [1]. Extending to mixed-cation systems, I will show how FA/MA compositional tuning creates a morphotropic phase boundary characterized by competing octahedral tilt patterns, nanoscale structural fluctuations, and enhanced electron-phonon coupling [2]. Moving from global structural effects to local electronic ones, I will discuss self-trapped excitons in BiVO₄, where hybrid density functional theory calculations identify competing localization modes that influence charge separation and transport [3]. Together, these examples illustrate how complexity arises at different levels in energy materials, and how a holistic modeling framework is needed for linking global phase behavior to local excited-state phenomena.
