We study electronic and atomistic properties of matter in condensed and molecular phases. Understanding the microscopic origin of the physical phenomena that underlie experimental observations, and developing novel theoretical and algorithmic tools to model the structure and dynamics of matter at the nanoscale are among our goals.

Clockwise from upper left corner: atomic wires, path integral representaton of a water molecule, ice lattice, carbon nanotube.

Our research makes extensive use of computer simulations based on methods that are rooted in condensed matter theory, classical and quantum mechanics, statistical mechanics. We use density functional theory and many-body theoretical approaches to model the electronic properties of materials and molecules. Electronic density functional theory in combination with classical molecular dynamics is used to simulate nuclear motion in liquids and solids. In this approach, called ab-initio molecular dynamics, the potential energy surface for nuclear dynamics is derived on the fly from the instantaneous ground-state of the electrons, capturing the interplay between molecular dynamics and chemical bond evolution. Ab-initio molecular dynamics was introduced by Car and Parrinello in1985 and, since then, has been applied to a number of problems in condensed matter and chemical physics, materials science and biochemistry, becoming a standard tool of molecular simulation. The ab-initio molecular dynamics scheme in the context of Feynman's path integral formulation of statistical mechanics allows us to account without approximation for quantum nuclear effects in equilibrium statistical properties. These effects are non-negligible, even at room temperature in systems containing light atoms such as hydrogen.

In spite of the important progress made in the last decades, many challenges remain. The general problem of dynamics of a quantum many-body system, as opposed to sampling equilibrium configurations, remains one of the most difficult and so far unsolved problems in the field. Nuclear tunneling events and non-adiabatic processes are examples of phenomena that require a satisfactory solution to the quantum dynamics problem. Another difficulty that plagues computer simulations, and especially ab-initio simulations, is associated to rare events, i.e. processes that can be exceedingly slow on the time scale of the simulations. Examples are chemical reactions and nucleation phenomena. Developing simulation techniques to sample rare events is an active area of current research. Finally, electron dynamics underlies processes like optical excitations in molecular and condensed phases, and electron transport in molecular devices. A proper treatment of electron excitations requires adopting either time-dependent density functional schemes or many-body theory methods. Transport at the nanoscale requires modeling an open quantum mechanical system and raises fundamental issues concerning the interplay of ballistic and dissipative processes. Our current research is actively involved in all the above themes.