Predicting Synthesizability and Mechanical Properties of High-Entropy Borides through First-Principles and Machine Learning Simulations

High-entropy borides (HEBs) with superior mechanical properties are promising candidates for extreme-environment applications. Here, we study the synthesizability of hexagonal five-metal HEBs (with period 4-6 transition metals) by computing their entropy formation ability (EFA) descriptors using density functional theory (DFT) calculations. We also train machine learning (ML) models using compositional features for additional data analysis and EFA target prediction. Additionally, we employ DFT calculations with special quasi-random structure (SQS) supercells to evaluate the HEB mechanical properties, including bulk and shear moduli, as well as hardness. A comparison of our simulations to experimental results will be discussed. Our study, combining first-principles and ML simulations, provides a framework for identifying HEBs with optimal synthesizability and superior mechanical properties, helping to accelerate the discovery of novel high-entropy materials for applications in extreme conditions.