Abstract:Developing high-entropy ceramics (HECs) with ultra-high melting points (Tm) is crucial for their applications in ultra-high-temperature environments. However, related research has seldom been reported. Here, taking high-entropy diborides (HEBs) as an example, we develop a data-driven method to efficiently explore HEBs with ultra-high Tm via transferable machine-learning-potential-based molecular dynamics (MD). Specifically, a moment tensor potential (MTP) for HEBs with nine transition metal elements of group IVB, VB, and VIB is first constructed based on unary and binary diborides. Further studies on the performance of our constructed MTP have confirmed its remarkable accuracy, transferability, and reliability across both equimolar and non-equimolar HEB systems. Tm of HEBs are then accurately simulated through MD simulations based on the constructed MTP, and 24 features are simultaneously collected to enable reliable machine learning training. Five descriptors with the gradient boosting regression model are derived as the optimal combination for accurate Tm predictions in HEBs with genetic algorithms. Based on our established model, Tm of 32563 HEBs are eventually determined, achieving the maximum Tm of 3688 K in (Ti0.1Zr0.1Hf0.6Ta0.2)B2. The work presents a feasible approach to develop HECs with ultra-high Tm.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to develop high - entropy ceramics (HECs) with ultra - high melting points ($T_m$), especially high - entropy diborides (HEBs). Specifically, the research aims to efficiently explore and predict HEBs with ultra - high melting points through machine - learning - potential - driven molecular dynamics simulations. The following is a detailed description of this problem: 1. **Background and Challenges**: - High - entropy ceramics (HECs) are of great value in extreme - environment applications, such as under high - temperature, high - pressure, strong - erosion and radiation conditions. - Among them, the ultra - high melting point ($T_m$) is crucial for high - temperature applications. - However, due to the huge compositional space, developing HECs with ultra - high melting points faces great challenges. 2. **Limitations of Existing Methods**: - Traditional experimental methods are time - consuming and costly, and it is difficult to obtain sufficient melting - point data. - Molecular dynamics (MD) simulation is an ideal method, but the lack of a reasonable inter - atomic potential function suitable for the HECs system is a major obstacle. 3. **Solutions**: - This paper proposes a molecular dynamics simulation method based on transferable machine - learning potentials (MLP) to efficiently predict the melting points of HEBs. - A moment - tensor potential (MTP) suitable for diborides of nine transition - metal elements (from groups IVB, VB and VIB) was constructed, and a training data set was generated through density - functional - theory (DFT) calculations. - Using genetic algorithms (GAs) combined with a gradient - boosting - regression (GBR) model, five optimal descriptor combinations were determined for accurate melting - point prediction. 4. **Objectives**: - Through the above methods, the melting points of 32,563 HEBs were finally predicted, among which the highest melting point reached 3,688 K, which occurred in the non - equimolar ratio (Ti$_{0.1}$Zr$_{0.1}$Hf$_{0.6}$Ta$_{0.2}$)B$_2$. - A feasible method for developing HECs with ultra - high melting points is provided, and the discovery of such materials is accelerated. In summary, this paper aims to solve the difficult problem of developing high - entropy ceramics with ultra - high melting points through machine - learning and molecular - dynamics simulations, thereby promoting their applications in extreme environments.

Machine learning potential-driven prediction of high-entropy ceramics with ultra-high melting points

Exploring mechanical and thermal properties of high-entropy ceramics via general machine learning potentials

Data-driven discovery of a formation prediction rule on high-entropy ceramics

Design high-entropy carbide ceramics from machine learning

Unrevealing hardening and strengthening mechanisms in high-entropy ceramics from lattice distortion

Discovering superhard high‐entropy diborides ceramics via the hybrid data‐driven and knowledge‐enabled model

Design of superhard high-entropy diborides v i a high-throughput DFT and thermodynamics calculations

Interpretable machine learning for simultaneous designing martensitic transformation temperature and thermal hysteresis of high-entropy shape memory alloys

Using Machine Learning and Feature Engineering to Characterize Limited Material Datasets of High-Entropy Alloys

Designing high-entropy ceramics via incorporation of the bond-mechanical behavior correlation with the machine-learning methodology

Temperature Dependent Thermal and Elastic Properties of High Entropy (Ti0.2Zr0.2Hf0.2Nb0.2Ta0.2)B2: Molecular Dynamics Simulation by Deep Learning Potential

Electronic structures and strengthening mechanisms of superhard high-entropy diborides

Rational design of high-entropy ceramics based on machine learning – A critical review

Machine learning-assisted mechanical property prediction and descriptor-property correlation analysis of high-entropy ceramics

Prediction of Ti-Zr-Nb-Ta high-entropy alloys with desirable hardness by combining machine learning and experimental data

A deep neural network potential model for transition metal diborides

Mechanical behavior of high-entropy intermetallic compounds and high-entropy ceramics

High-entropy ceramics: Present status, challenges, and a look forward

Phase stability, mechanical properties and melting points of high-entropy quaternary metal carbides from first-principles

Formation ability descriptors for high-entropy diborides established through high-throughput experiments and machine learning

Dense nine elemental high entropy diboride ceramics with unique high temperature mechanical and physical properties