High-throughput study and machine learning on MAX and MAB phases: new materials and fingerprints of superior lattice thermal conductivities

The ternary ceramic MAX and recently emerging MAB phases have exhibited a combination of properties of metals and ceramics, excellent mechanical properties along with high-damage tolerance. Thermal conductivity as one of the fundamental properties is closely associated with operating conditions, which could serve as an essential indicator of thermo-functional applications. Upon performing a high-throughput density functional theory calculation on M ₂AX (X = B, C, or N) and M ₂AB ₂ phases in a wide compositional space, a great number of new materials are stable. Several ultra-low/high lattice thermal conductors are identified. Combined with the machine-learning, the underlying origins of approaching superior lattice thermal conductivity are unraveled and verified to be self-consistent by reverse comparison. Towards the “truly” stable materials by including reaction enthalpies, about seventy materials are retained, in which Zr ₂SnC and Nb ₂SnB having ultralow κ ph (< 2 W (m K) ⁻¹ ) are successfully synthesized and further characterized. A new continent of ceramics with superior lattice thermal conductivities is thus reported, and we tend to lay foundations of the lattice thermal conductivity of such layered ceramics by machine learning methods and physical modeling. It is believed that this work would pave the way for rational design and high-throughput studies of materials.