Prediction of Single‐Phase High‐Entropy Nitrides from First‐Principles Calculations

Abstract

Entropy‐forming ability (EFA) is used as a descriptor to identify possible single‐phase high‐entropy nitrides (HENs) from eight cation candidates Al, Si, Ti, V, Cr, Zr, Nb, and Mo. Around 56 five‐cation HEN compositions using density functional theory calculations of 2744 ten‐atom supercells are evaluated. Ten HEN compositions with high EFA values are identified, and among them AlCrNbTiVN5 has the highest EFA value. Cation‐nitrogen bond length statistics shows that HENs with small lattice distortion tend to have large EFA values. The elemental dependence analysis shows that silicon introduces largest lattice distortion and hence small EFA values, which provides useful guidance for element selection of single‐phase HENs.

Machine Learning-Based Prediction of Stability in High-Entropy Nitride Ceramics

“…Such distortion affects the material's propensity to form a stable single-phase structure, as a uniform distribution of atoms is generally favored for higher EFA values. Our findings align with those of an existing study [32], which also reported that the incorporation of Si led to a compacting of the crystal lattice and consequently, a reduction in the EFA.…”

“…For the initial prediction phase, we compiled a dataset of 56 samples, each with its corresponding EFA value as the target property. These EFA values were sourced from the theoretical computational data based on bulk HEN ceramic materials featuring facecentered cubic structures [32]. The EFA values are shown in Table A1.…”

“…Our research adopted a data-driven approach by constructing a SISSO model, which was informed by an extensive dataset of HENs [32]. The performance of our SISSO model was meticulously validated against theoretical benchmarks documented in the literature and benchmarked against existing ML methodologies.…”