DMREF: Machine Learning Accelerated Design and Discovery of Rare-earth Phosphates as Next Generation Environmental Barrier Coatings

Environmental barrier coatings (EBCs) are key components that can greatly enhance the performance/longevity of structural materials such as ceramic-matrix composites against active oxidation in high speed hot gas streams and corrosion in reactive engine environments. Multi-generation EBCs have evolved, mainly based on silicate-based systems, but they suffer from the volatility of silicon due to water vapor attack and corrosion of molten glass attack.

Innovative design and discovery of EBCs with transformative performance are needed to meet even harsher environments of high temperature, high thermal flux and severe oxidation and corrosion for future aerospace and space systems. This Designing Materials to Revolutionize and Engineer our Future (DMREF) project will explore an innovative concept of using multiple component rare-earth phosphates as advanced EBCs, and develop a science-based paradigm guided by machine learning (ML) for accelerated materials design and discovery.

Both graduate and undergraduate students will be trained as the next-generation workforce in this data-driven materials research. K-12 students and underrepresented groups will be engaged through multiple outreach activities such as the Engineering Summer Exploration program at Rensselaer and the New Visions: Math, Engineering, Technology & Science program.

Materials data and computational tools developed will be contributed to the MPContribs Portal for public access on the Materials Project platform to facilitate data-driven material design.

Material design and discovery for advanced environmental barrier coatings (EBCs) have been greatly hindered by our limited understanding of how composition and microstructure affect materials properties and performance. This project will accelerate fundamental understanding of the influence of composition and microstructure on the phase stability and properties of multicomponent rare-earth phosphates, and use this understanding to optimize performance of next generation EBCs for ceramic matrix composites (CMCs) in reactive engine environments.

A multipronged data-driven machine learning (ML) approach will be developed to inform materials design and guide materials performance evaluation to discover new rare-earth phosphates that have unique attributes of EBCs for CMCs, compared to current state-of-the-art disilicates without the issue of silicon evaporation. An element-based ML will be trained on high throughput density functional theory calculations and will be used to guide the design and optimization of configurationally-disordered rare-earth phosphates with key characteristics of EBCs.

A microstructure-based ML will be trained on high-throughput finite element method calculations and will be used to predict the optimal microstructure and performance of EBCs against molten glass corrosion at elevated temperatures. The integration of multiscale computations, machine learning, and experimental demonstration and validation will provide a pathway for success in accelerating the design and discovery of rare-earth phosphates as next generation EBCs for CMCs.

This project is jointly funded by NSF’s Mathematical and Physical Sciences (MPS) Division of Materials Research (DMR) Designing Materials to Revolutionize and Engineer our Future (DMREF) program, and the Division of Civil, Mechanical, and Manufacturing Innovation (CMMI) in the Directorate for Engineering (ENG).

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.