Theoretical approaches of magnetism of transition-metal thin films and nanostructures on semi-infinite substrate

Abstract

Tremendous progress has been made in the field of magnetic materials and technology over the past few years. Superior properties and novel scientific questions arise from our ability to either synthesize artificial structures or tailor microstructures at the appropriate length scale. The study of low-dimensional magnetism of thin films, surfaces and interfaces has revealed new and sometimes unexpected properties. In this review we focus on the determination of the local magnetic moments distribution at the surface of transition metals, for very thin transition-metal films and nanostructures on noble and transition-metal substrates at T = 0 K. Two different approaches are commonly used. Ab initio and semi-empirical tight-binding (TB) calculations appear to be complementary to determine the electronic structure. The first provides very precise results for periodically ordered systems, whereas the second is able to study complex systems with a large number of non-equivalent sites. We discuss these approaches and point out recent developments which combine the versatility of the ab initio method with the rapidity of the TB calculation. Special attention is devoted to the magnetic frustration which occurs when both ferromagnetic and antiferromagnetic couplings are present. The Fe Cr interface which has been, since 10 years, one of the most studied systems is a good example. Recent experimental and theoretical efforts have addressed the question of a possible onset of ferro- or antiferromagnetism in nanostructures, free-standing clusters, thin-film structures of metals that are non-magnetic in the bulk form. V, Pd, Rh and Ru have been singled out as potential candidates. Thus particular attention will be devoted to V free-standing films, as well as V monolayer on Ag and Fe. Rh and Ru have been found magnetic experimentally in the form of cluster (Rh) and for Ru monolayer on graphite. However, Rh monolayer on Ag is found experimentally non-magnetic whereas calculations show that magnetism is killed when surface alloy is considered. Pd is another important element to consider because, being non-magnetic in the atomic form as well as in bulk form it presents high susceptibility. Therefore, it can become magnetic if its lattice parameter is slightly increased. It is believed to be magnetic in the case of a bilayer on Ag(0 0 1). Induced polarization is obtained when Pd is in contact with a magnetic metal. Because of its exotic structural and magnetic properties, Mn is an interesting candidate for thin-film growth as it is expected to accept different local configurations. Experimentally, one may attempt to stabilize normally high-temperature phases of Mn by epitaxial growth on a suitable substrate. We will show that Mn takes on the c(2 × 2) magnetic configuration for the Mn monolayer on Fe(0 0 1) and Co(0 0 1). Calculations show that Ir and Pt, in the monolayer range, are magnetic when grow on Ag. However, no experiment has displayed any kind of magnetism yet and more realistic calculation has shown that spin-orbit effect kills the magnetic moment. Therefore, only V, Cr, Mn, Co and Ni for 3d, and Ru, Rh and Pd for 4d will be considered in this review.