Introduction

It is possible to understand, explain and predict physical phenomena related to macroscopic systems by modelling and analyzing the behavior of their microscopic constituent elements (atoms). In the last century, a wide variety of atomistic modelling methods, primarily founded in quantum physics, were developed. Theoretical frameworks, computational methods and the software/hardware infrastructure related to atomistic modelling have been rapidly advancing, resulting in time and length scales available to modern simulations being comparable to experiments.

Research Interests

Our group develops and applies first-principles and machine-learning-based atomistic simulations — including density functional theory, ab initio molecular dynamics, many-body/quasiparticle methods, and machine-learned interatomic potentials — to understand, predict, and design the properties of materials.

Within this shared methodological foundation, individual members pursue a variety of particular research directions, including:

• Two-dimensional and van der Waals materials — electronic, magnetic, and vibrational properties of graphene, transition-metal dichalcogenides, and layered heterostructures.
• Surfaces and catalysis — adsorption, femtosecond-laser-induced dynamics, and photoinduced desorption of molecules on metal and oxide surfaces.
• Magnetism and multiferroics — magnetic ordering, proximity effects, and magnetoelectric coupling in layered and molecular materials.
• Molecular crystals and mechanochemistry — polymorphism, thermosalient effects, and mechanochemical reaction pathways.
• Machine learning for materials — interatomic potentials, crystal structure prediction, and data-driven property screening.

For a full account of our work, see our Publications, or visit People to learn more about individual members’ research.