Project: Accelerating Discovery and Characterization of Functional 2D Materials for Catalysis and Solar Energy Conversion Using Data Science Methods

Figure: Interplay between first-principles simula-tions, materials synthesis, materials characterization, and machine learning to predict and develop new 2D materi-als.
The discovery of graphene has sparked interest in 2D layered materials as a testbed for new physics and potentially transformative applications in flexible optoelectronics, solar energy conversion, catalysis and chemical sensing. Two decades after graphene’s discovery, the predicted stable 2D chemistries are numbered in the thousands and span the range from insulators, to semiconductors, topological materials, metals and even superconductors. Beyond the chemical structure, electronic and optical properties of 2D materials can be engineered by the presence of edges, defects, or by introducing guest species between layers.

As a result, the parameter space to explore is extremely large, and non-linear dimensional reduction and other unsupervised learning approaches present a promising approach to identifying the most promising 2D structures for specific applications.

In this work, we will combine first-principles electronic structure theory, multi-scale methods and data-driven machine and deep learning approaches with fabrication and experimental characterization of selected identified 2D structures (chalcogenides and MXenes) to validate and
provide feedback to the models that will be developed to predict relevant 2D material parameters: carrier density, mobility, life-time and sensitivity to disorder. Experimental materials characterization tools in the Titova lab are available as source for fast acquisition of data useable for training machine learning algorithms.

Researchers:   Dr. Titova, Physics &   Dr. Deskins, Chem. Eng. & Data Science faculty TBD.