TDDFT simulations of the energy loss of moving projectiles in solids and nanostructures

Abstract

We have recently developed a code to perform real time time-dependent density-functional theory simulations. Our method is based on the SIESTA code and uses a linear combination of atomic orbitals as a basis set. Previous versions of our code had been used to study the optical response of finite systems, i.e., electron dynamics was followed after the system was initially perturbed while nuclei were kept fixed in their initial positions. Our most recent version, however, allows performing coupled electron-nuclear dynamics within the Ehrenfest approximation and has been applied to study the problem of radiation damage in solids and nanostructures. Although radiation damage processes are of extraordinary fundamental and technological importance, ab initio simulations of these effects in solids are still very scarce to date. Most simulations for solids and condensed systems are based on semi-empirical approaches, like SRIM. The energy transferred to the solid goes both onto displacements of the target ions (nuclear stopping) and electronic excitations (electronic stopping). While at very low velocities nuclear stopping can we dominant, at moderate, intermediate and high energies the most efficient energy loss mechanism is the electronic stopping. The effect of electronic stopping is frequently incorporated in simulations through an ion and target dependent friction coefficient. Thus, the electronic stopping is assumed to depend linearly on velocity. This is generally true for simple metals, for which the friction coefficient can be estimated very efficiently using a jellium model plus scattering theory. However, it has been recently observed that there are significant deviations from linearity at low velocities in insulators and noble metals, both showing different kinds of threshold effects. Understanding of such effects demands an explicit treatment of the electronic stopping in the presence of the actual atoms and actual electronic structure of the host system. Our simulations using time-evolving TD-DFT could reproduces the anomalies in the stopping power observed experimentally for projectile velocities below 0.3 a.u., for insulators and noble metals. In addition, we could analyze the Barkas effect (difference in stopping between protons and antiprotons) in LiF, and the He/H anomaly in Au (the stopping is larger for He at all velocities, contrary to expectations based on free electron models). Our approach has quite general applicability and we plan to apply it to other radiation damage problems. As an example, we have recently studied the influence of the electrons being excited on the effective internuclear forces when an Al target is bombarded with protons. Finally, the dependence of the electron dynamics on the size and dimensionality is an important issue in many fields. For example, it determines the efficiency and time scale of the screening of interactions, the rate of many chemical reactions at surfaces and the optical response of nanoobjects. We plan to use our methods to investigation some of these issues. In particular, I will present some of our recent semi-classical results on the influence of the localized-hole screening on the energy losses during photoemission from metal clusters.