Abstract:Accurate phase diagram prediction is crucial for understanding alloy thermodynamics and advancing materials design. While traditional CALPHAD methods are robust, they are resource-intensive and limited by experimentally assessed data. This work explores the use of machine learning interatomic potentials (MLIPs) such as M3GNet, CHGNet, MACE, SevenNet, and ORB to significantly accelerate phase diagram calculations by using the Alloy Theoretic Automated Toolkit (ATAT) to map calculations of the energies and free energies of atomistic systems to CALPHAD-compatible thermodynamic descriptions. Using case studies including Cr-Mo, Cu-Au, and Pt-W, we demonstrate that MLIPs, particularly ORB, achieve computational speedups exceeding three orders of magnitude compared to DFT while maintaining phase stability predictions within acceptable accuracy. Extending this approach to liquid phases and ternary systems like Cr-Mo-V highlights its versatility for high-entropy alloys and complex chemical spaces. This work demonstrates that MLIPs, integrated with tools like ATAT within a CALPHAD framework, provide an efficient and accurate framework for high-throughput thermodynamic modeling, enabling rapid exploration of novel alloy systems. While many challenges remain to be addressed, the accuracy of some of these MLIPs (ORB in particular) are on the verge of paving the way toward high-throughput generation of CALPHAD thermodynamic descriptions of multi-component, multi-phase alloy systems.

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is how to use machine - learning interatomic potentials (MLIPs) to accelerate the phase - diagram prediction of complex alloys. Specifically, the paper explores how to significantly improve the speed of phase - diagram calculation while maintaining sufficient accuracy by combining MLIPs with the CALPHAD method. The following are the specific problems that this paper attempts to solve: 1. **Limitations of the traditional CALPHAD method**: - Although the traditional CALPHAD method is robust, it is resource - intensive and depends on experimentally evaluated data, which limits its application in multi - component, multi - phase alloy systems. - For complex chemical spaces such as high - entropy alloys (HEA), the CALPHAD method can only predict the stability of phases that have been experimentally observed and partially evaluated, which is an important limitation when exploring new alloy systems. 2. **Requirement for efficient calculation**: - With the development of materials science, researchers need more efficient tools to quickly explore and design new alloy systems, especially those complex chemical spaces that have not been fully studied. - Although traditional calculation methods based on density functional theory (DFT) are accurate, their computational cost is extremely high, making it difficult to conduct high - throughput screening. 3. **Application potential of MLIPs**: - MLIPs such as M3GNet, CHGNet, MACE, SevenNet and ORB models can significantly accelerate phase - diagram calculation while maintaining a certain degree of accuracy, achieving a calculation speed more than three orders of magnitude faster than DFT. - By combining MLIPs with the ATAT tool, thermodynamic descriptions can be efficiently generated within the CALPHAD framework, thereby enabling high - throughput exploration of alloy systems. 4. **Challenges and solutions**: - Although MLIPs offer significant computational advantages, in some cases, their prediction accuracy may not be as good as DFT, especially in systems involving multiple competing phases. - The paper presents the performance of MLIPs (especially ORB) in different alloy systems through multiple case studies (such as Cr - Mo, Cu - Au and Pt - W alloy systems) and proposes thresholds to ensure the accuracy of MLIPs. In summary, this paper aims to overcome the limitations of the traditional CALPHAD method by introducing MLIPs, providing an efficient and accurate high - throughput thermodynamic modeling framework to accelerate the phase - diagram prediction of complex alloy systems and the design of new materials.

Accelerating CALPHAD-based Phase Diagram Predictions in Complex Alloys Using Universal Machine Learning Potentials: Opportunities and Challenges

Quasi-Phase Equilibrium Prediction of Multi-Element Alloys Based on Machine Learning and Deep Learning

Phase diagrams of alloys and their hydrides via on-lattice graph neural networks and limited training data

Accelerated design of high entropy alloys by integrating high throughput calculation and machine learning

Unraveling phase prediction in high-entropy alloys: A synergy of machine learning, deep learning, and ThermoCalc, validation by experimental analysis

Temperature-Dependent Design of BCC High Entropy Alloys: Integrating Deep Learning, CALPHAD, and Experimental Validation

Design and Selection of High Entropy Alloys for Hardmetal Matrix Applications using a Coupled Machine Learning and CALPHAD Methodology

The CALPHAD approach for HEAs: Challenges and opportunities

The High-dimensional Phase Diagram and the Large CALPHAD Model

Towards High-Throughput Microstructure Simulation in Compositionally Complex Alloys Via Machine Learning

Accelerating Phase Prediction of Refractory High Entropy Alloys Via Machine Learning

Deep Learning Accelerated Phase Prediction of Refractory Multi-Principal Element Alloys

Phase prediction in high entropy alloys with a rational selection of materials descriptors and machine learning models

Towards Predicting Metal Extraction from Oxides with Machine-Learning Augmented First-Principles Calculations

Integrating CALPHAD into Phase Field Simulations for Practical Applications

Third-Generation CALPHAD Modeling of Elemental Nb and Zr and Partial Re-Assessment of Their Binary Phase Diagram

Phase Prediction of High-Entropy Alloys Based on Machine Learning and an Improved Information Fusion Approach

Machine-Learning Surrogate Model for Accelerating the Search of Stable Ternary Alloys

Machine Learning-Based Classification, Interpretation, and Prediction of High-Entropy-Alloy Intermetallic Phases

Predicting Ti-Al Binary Phase Diagram with an Artificial Neural Network Potential