Magnetic nanostructures

This amazing property of quantum corrals can be used to tailor magnetic interaction between single adatoms  if those are placed in the focuses of the corral or even a part of a corral. Our calculations show that by carefully adjusting the size and shape of the corral one can enhance, quench or even reverse the coupling between individual spins at the surface (Fig. 2)^[2].

Yet it is not always necessary to use corrals, that are extremely hard to produce experimentally, to confine surface electrons. Natural or self-assembled surface  structures can serve the same purpose. So, for example, our calculations have proven, that quantum confinement of surface state electrons on hexagonal  copper islands can be used to tailor the interaction between single magnetic adatoms  adsorbed on top of them (Fig. 3)^[3] .

While confinement of surface state electrons by paramagnetic structures can aid in controlling the interaction of single magnetic impurities adsorbed within, the confinement itself can also be spin-dependent. So, f. e., our calculations have predicted that the spin-dependent confinement on triangular cobalt islands leads to a strong variation of the local magnetization within a single island/nanostructure ^[4]. Quite recently our predictions have been confirmed by our experimental colleagues (Fig. 4)^[5].

We also study other types of electron confinement. So, for example, electrons confined between a buried magnetic layer or cluster and the vacuum barrier at the surface of the crystal form quantum well states and can significantly affect the orientation of single spins at the surface (Fig. 5)^[6].

While it is paramount to understand the intrinsic magnetic and electronic properties of nanostructures to learn to control them, it is equally important to find means of externally affecting those properties to ultimately gain control over magnetism gown to the single spin level. We thus strive to find such means. A powerful and, even more importantly, local way to affect the magnetic coupling in surface nanostructures is by using the electric field. Contrary to its magnetic counterpart, electric field can be delivered locally by means of a scanning tunneling microscope tip of a nanojunction. Our calculations show that external electric fields can control magnetism in bi- or even multi-stable magnetic units on a metal surface (Fig. 6)^[7].

Another way of locally controlling the orientation of single spins is the use of a magnetic tip of a scanning tunneling (STM) or atomic force (AFM) microscope. We show that the spin orientation of single atoms and dimers on the surface and the conductance through those structures can be controlled by vertical or lateral movements of a spin-polarized STM tip (Fig. 9) or a finite bias applied to the tip-structure-substrate nanojunction ^{[9], [10]}.

An ultimately important aspect of surface science is the ability to construct nanoscale systems and control their dynamics. The former is a prerequisite to any fundamental experiment or industrial application and the latter is essential for assessment of the system's stability and evolution. We know that the dynamics of a system is intimately bound to the system's geometry and electronic structure. Understanding and controlling them means understanding and controlling the dynamics.

We use ab initio density functional theory methods to describe the underlying electronic effects and kinetic Monte Carlo simulations to reveal the dynamics of the system in real time. For example, precise description of the Shockley surface state ^[12] on a metal surface and its effect on the long ranged interaction of surface adsorbates allows us to accurately describe the dynamics of adatoms immersed in the 2D electron gas of the surface state. Our calculations predict that such dynamics may lead to the assembly of the adsorbates into a Kondo superlattice (Fig. 8) which lies in perfect agreement with experimental observations (Fig. 9)^{[13], [14]}.

On certain stepped surfaces the very same physics may lead, according to our calculations and experimental evidence of our colleagues^[15], to the formation of one-dimensional nanostructures, i.e. atomic strings (Fig. 12).

Our calculations allow us to predict the dynamics of adatoms in more complex systems. For example, in circular quantum corral, the diffusing adatoms would prefer to form concentric rings (Fig. 13), rearranging themselves in the semblance of a quantum onion^[16].

By careful choice of geometry one can use the electronic properties of the surface state to trap adatoms in otherwise open geometries. For example, the energy landscape for the diffusion of a single adatom formed by an open "quantum resonator" (Fig. 14) consisting of two parallel monatomic chains, causes the adatom adsorbed within to hop along linear channels and prevents it from leaving the resonator^[17].