Machine learning interatomic potential can infer electrical response
This work extends machine learning interatomic potentials to predict electrical response without additional training data, enabling scalable modeling of electric-field-driven processes in diverse material systems, though it is incremental as it builds on existing frameworks.
The researchers tackled the challenge of modeling material and chemical systems' responses to electric fields by showing that polarization and Born effective charge tensors can be extracted from long-range machine learning interatomic potentials using the Latent Ewald Summation framework, solely from energy and force data, enabling predictions of infrared spectra, ionic conductivities, and phase transitions in systems like bulk water, superionic ice, and ferroelectric PbTiO3.
Modeling the response of material and chemical systems to electric fields remains a longstanding challenge. Machine learning interatomic potentials (MLIPs) offer an efficient and scalable alternative to quantum mechanical methods but do not by themselves incorporate electrical response. Here, we show that polarization and Born effective charge (BEC) tensors can be directly extracted from long-range MLIPs within the Latent Ewald Summation (LES) framework, solely by learning from energy and force data. Using this approach, we predict the infrared spectra of bulk water under zero or finite external electric fields, ionic conductivities of high-pressure superionic ice, and the phase transition and hysteresis in ferroelectric PbTiO$_3$ perovskite. This work thus extends the capability of MLIPs to predict electrical response--without training on charges or polarization or BECs--and enables accurate modeling of electric-field-driven processes in diverse systems at scale.