Synthesis of Novel Rare-Earth MXenes Using Density Functional Theory and Optimal Experiment Design

The emergent family of 2D materials exhibiting room temperature ferromagnetism hold massive potential for data storage, sensing, and quantum computing applications. This domain is currently limited to a few materials, but rapidly growing, and in need of scalable synthesis routes. MXenes, a family of highly tunable 2D transition metal carbides and nitrides, can potentially expand this set of atomically thin room temperature magnets. Specifically, the incorporation of rare-earth elements (Nd, Gd, Tb) into MXenes can potentially introduce magnetic properties, with potential further applications in energy storage and electromagnetic interference shielding. The domain of rare earth (RE)-MXenes has not yet been evaluated for magnetic properties. In this work, we predict stable rare-earth containing MXenes, along with their synthesis routes and magnetic properties, using optimal experiment design. We investigate four MAX systems (Mo-Nb, Mo-Ti, Cr-Ti, and Ti)-Al-C, with the incorporation of Nd, Gd, and Tb via doping into the different MAX arrangements, such as ordered and solid solutions. Doping of Mn is also studied, as Mn has been predicted to induce magnetism in MAX phases. We construct convex hulls for each RE-MAX system and predict decomposition of these phases, using DFT+U with linear response method to estimate Hubbard U parameters. From these MAX phase predictions, we implement an active learning loop using semi-supervised learning to predict synthesis of these RE-MAX phases and their etched RE-MXenes.