糖心TV

Single atoms are the smallest possible magnets, of interest for fundamental physics and of great promise for technological applications. When assembled on a surface, their properties can be probed and manipulated via scanning tunneling microscopy (STM) and inelastic scanning tunneling spectroscopy (ISTS). However, the theoretical description is challenging, due to the interplay between the reduced dimensionality, the interactions driving the magnetism, and the coupling to the surface. Furthermore, an accurate description of the low-energy physics due to the spin-orbit interaction and external magnetic fields is essential.

In J眉lich, we have recently developed a hierarchical theoretical approach to the static and dynamics properties of surface-supported magnetic nanostructures. The materials-specific information is supplied by density functional theory (DFT), which gives access to ground-state magnetic properties and magnetic interactions. The dynamics of the magnetic moments follows from time-dependent DFT (TDDFT), accounting for the impact of the surface electrons on the spin dynamics. The final level of theory addresses many-body effects, either through many-body perturbation theory (MBPT) or by solving a multi-orbital Anderson impurity model with quantum Monte Carlo, yielding realistic inelastic transport spectra.

In this talk I will present an overview of these theoretical methods while focusing on concrete physical systems: magnetic adatoms and small clusters on metallic surfaces, such as Cu(111) or Pt(111), in close connection with the available experimental information. Different magnetic adatoms on the same surface have very different static and dynamic properties [1-3], which influence their magnetic stability via zero-point spin fluctuations [4-6], while their tunneling spectra are only adequately described in MBPT [7-9]. The interactions between magnetic adatoms depend not only on their separation but also on their arrangement with respect to the surface, leading to widely different properties of apparently similar clusters [10-12]. I conclude by discussing the possible shortcomings and future research directions.