Decoratypes: An Extensible Crystal Taxonomy for Machine Learning-Guided Materials Discovery
Kyle D. Miller, Michele Campbell, Danilo Puggioni, James M. Rondinelli
TL;DR
This work addresses the need for scalable, structure-aware classification to guide materials discovery. It introduces decoratypes, an extensible taxonomy that couples prototype-level descriptions with site-specific decorations, and defines polaritype and related subclasses to capture structure–property mappings. A data-driven discovery workflow combines polaritype screening, active learning, and high-fidelity DFT to identify six novel ferroelectric or hyperferroelectric candidates in anti-Ruddlesden-Popper derivatives, including strain-activated cases. The results demonstrate robust generalization across chemical space and reveal clear structure–property trends, enabling targeted design of strain-tunable ferroic materials with potential for advanced functional applications.
Abstract
We introduce decoratypes as a structure taxonomy that classifies compounds based on site decorations of specific structural prototypes. Building on this foundation, a ferroelectric materials discovery framework is developed, integrating decoratypes with an active learning approach to accelerate exploration. In addition, six novel ferroelectric candidates are predicted, including three strain-activated ferroelectrics and three strain-activated hyperferroelectrics. These findings highlight the potential of the decoratype taxonomy to enhance our understanding of structure-driven material properties and facilitate the discovery of promising yet underexplored regions of chemical space.