Breakthrough in MAX phases: Machine learning unlocks new materials

Materials scientists have long been fascinated by the exploration of MAX phases, aiming to uncover new materials with enhanced properties for diverse applications. The unique combination of ceramic and metal-like characteristics in MAX phases makes them highly promising. However, the vast number of potential elemental combinations, reaching up to 4347 when considering a specific range of elements and structure limitations, poses a significant challenge. Manually screening through these combinations to identify stable and experimentally viable MAX phases is not only time-consuming but also extremely inefficient when relying on traditional first-principles calculations.

Recently, a group of researchers from the Harbin Institute of Technology made a remarkable breakthrough. They developed a machine-learning-based stability model for MAX phase, leveraging elemental features to revolutionize the screening process. This model can rapidly assess the stability of MAX phases using only basic elemental parameters.

The team published their work in Journal of Advanced Ceramics on February 21, 2025.

"We constructed this model by training it on a comprehensive dataset of 1,804 MAX phase combinations, sourced from existing literature," said Prof. Yuelei Bai, the corresponding author of the study and a leading expert in the field of ceramics at the Harbin Institute of Technology. "This allowed us to screen out 150 previously unsynthesized MAX phases that met the stability criteria and even guided the first-time experimental synthesis of Ti₂SnN."

Ti₂SnN, a newly-synthesized MAX phase, exhibits remarkable properties. It has a low elastic modulus, high damage tolerance, and a self-extrusion characteristic. "The experimental findings of Ti₂SnN are exciting and undoubtedly demonstrate the accuracy of the model. However, it is worth noting that it is prepared by the Lewis acid replacement method, the method of sintering the elemental powder without pressure has failed in the attempt, and we may need to try more preparation methods for the remaining predicted stable phases." said Zhiyao Lu (Ph.D candidate), one of the lead researchers on the project. Professor Bai also added, "Compared to discovering a specific MAX phase compound, this method, which predicts stability based solely on elemental composition and offers scalability, is more important."

Through their research, the team also discovered that average valence electron number and valence electron difference play a crucial role in determining the stability of MAX phases. "Understanding these factors gives us a deeper insight into the fundamental mechanisms of MAX phase formation. It's like unlocking a key part of the material-design puzzle," Prof. Bai added.

The research not only significantly improved the screening efficiency of MAX phases but also opened up new avenues for future material design. However, the researchers’ acknowledge that there is still more work to be done. "We plan to further explore the properties of the newly discovered MAX phases and extend model to the high-entropy MAX phases. Our long-term goal is to build a comprehensive database of stable MAX phases and their properties and to better serve their applications in thermal barrier coatings," Prof. Bai said.

Other contributors to this research include Yun Fan, Zhaoxu Sun, Xiaodong He, Chuchu Yang, Hang Yin, Jinze Zhang, Guangping Song, Yongting Zheng from the Harbin Institute of Technology.

This work was supported by the National Natural Science Foundation of China (Grant No. 51972080) and the science foundation of the National Key Laboratory of Science and Technology on Advanced Composites in Special Environments.

About Author

Yuelei Bai is a professor in mechanics at the Center for Composite Materials and Structures, School of Astronautics, Harbin Institute of Technology. In 2013, he received funding from the first batch of the Postdoctoral International Exchange Program (sponsored by the National Postdoctoral Management Committee under the Ministry of Human Resources and Social Security), and then conducted two years of postdoctoral research at Imperial College London and Nanyang Technological University in Singapore.

Currently, his research primarily focuses on both basic and applied research in the fields of ternary layered ceramics, multi-scale simulations of material behavior in extreme environments, ceramic-matrix composites, and thermal protection materials and systems. He is dedicated to addressing scientific and technological challenges in aerospace, aviation, transportation, and other sectors.

In recent years, he has published 56 academic papers in top-tier international journals, such as Acta Materialia and the Journal of the American Ceramic Society. These papers have been cited 1,958 times, and his H-index is 26.  He also serves as an Associate Editor for the Journal of the American Ceramic Society and is a member of the editorial board of Journal of Advanced Ceramics.