Composition-aware optimization of interatomic potentials for glasses toward experimental properties

The accuracy of molecular simulations is fundamentally limited by the interatomic potentials that govern atomic interactions. Traditional potential development, which relies heavily on ab initio calculations, frequently struggles to reproduce the experimentally observed properties that govern real material behavior. To address this challenge, we present a practical, composition-aware workflow that integrates established machine-learning surrogate modeling, active learning, and standard global-optimization algorithms to calibrate classical interatomic potentials directly against experimental properties. Our method, here showcased on soda-lime borosilicate glasses, targets both global (density) and local (boron coordination) structural features across a wide range of compositions. By combining a surrogate model with iterative active learning, the framework efficiently explores a five-dimensional parameter space using only 400 molecular dynamics simulations over 17 iterations, making it highly data-efficient and eliminating the need for extensive simulation campaigns. Two optimized parameter sets are identified, each demonstrating good agreement with experimental measurements, including glass density, fraction of four-fold boron, and X-ray structure factor. The framework effectively captures and manages inherent trade-offs between structural objectives and compositional regimes, providing insights into the coordination behavior of boron in complex glass networks. Rather than yielding a single universally optimal parameterization, the results reveal intrinsic trade-offs between density and B4 fraction predictions and composition-dependent performance. The resulting classical force fields are generalizable and do not require reparameterization for individual compositions. Altogether, this work demonstrates how the careful integration of established ML and optimization tools can deliver a scalable, experimentally grounded route to transferable interatomic potentials for complex multi-component glasses, an application area where data-efficient calibration against experimental targets has remained a persistent challenge.