Title:Descriptor-Enabled Rational Design of High-Entropy Materials Over Vast Chemical Spaces

Abstract:The practically unlimited high-dimensional composition space of high-entropy materials (HEMs) has emerged as an exciting platform for functional materials design and discovery. However, the identification of stable and synthesizable HEMs and robust design rules remains a daunting challenge due to the difficulty in determining composition/structure-specific enthalpy and entropy contributions to the stability and formation of HEMs. In this work, using first-principles calculations, we find that (i) the stability and miscibility of HEMs strongly depend on the formation enthalpy of the HEM relative to the most stable completing phase rather than on the conventionally-viewed enthalpies of mixing over all possible competing phases, and (ii) the entropy forming ability of a HEM can be measured from the defect formation energy spectrum of tens of substitutional defects in ordered binary compounds, involving no sampling over numerous alloy configurations. Based on these findings, we propose a highly predictive Mixed Enthalpy-Entropy Descriptor (MEED), which enables the rational high-throughput first-principles design and screening of new HEMs over large chemical spaces. Applying the MEED to two structurally distinct material systems (i.e., 3D rocksalt carbides and 2D layered sulfides), not only all experimentally reported HEMs within each system are successfully identified, but a universal cutoff criterion for assessing their relative synthesizability also revealed. In addition, tens of new high entropy carbides and 2D high-entropy sulfides are also predicted. They have the potential for a wide variety of applications such as coating in aerospace devices, energy conversion and storage, and flexible electronics.