No abstract

No abstract

No abstract

Uniaxial anisotropy has been found in ultrathin cobalt films grown on a Cu(l 1 13) surface. Our studies using scanning electron microscopy with polarization analysis clearly show that the easy axis of magnetization is parallel to the direction of the step edges of the Cu(l 113) substrate. In spite of the different anisotropy behavior, the domain structures in Co/Cu(00l) and Co/Cu(l 1 13) are similar, which indicates that the domain pattern in ultrathin films is little affected by the anisotropy. PACS numbers: 75.30.Gw, 75.60.Ch, 75.70.Ak Recent scientific and technical advances in surface science and thin-film preparation methods have opened up a new class of research activities, i.e., the investigation of new artificially grown materials. One fascinating aspect of such studies is the ability to compare properties of the same material in different forms. Different crystal structures can be stabilized by the appropriate choice of substrates, phases that otherwise do not occur in nature such as bcc Co [1], fee Co [2], fee Fe [3], and bec Cu [4l.A further item of interest with these artificial materials concentrates on the exploitation of the transition from threeto two-dimensional crystals. For that purpose, the investigation of ultrathin films, i.e., films of a few monolayer thickness, has gained more and more importance in recent times. Apart from the general interest in studying dimensionality effects, there is a profound interest in understanding ultrathin films from the technological point of view, as novel devices continue to shrink in size.Investigations of magnetism in ultrathin films impressively demonstrate the variety of effects which can be found in ferromagnets of diminishing thickness. In many cases the effects manifest in the magnetic anisotropy of the films [5], For the interpretation and understanding of the magnetic properties it turns out that it is of great importance to distinguish the purely magnetic properties from those induced via magnetoelastic interactions by the film/substrate interface [6]. One approach to solve that problem is to study the ultrathin-film magnetism in a film system with a perfect and ideal substrate/film interface. The system Co/Cu(001) is well known from the literature to fulfill the above condition, and is well characterized concerning growth as well as magnetic properties [2,7-9]. The system exhibits layer-by-layer growth. The anisotropy behavior is determined by the film symmetry; no interface (i.e., magnetoelastic) effects altering these symmetry properties have been found for the cobalt films. Thus Co/Cu(001) is the ideal reference system for the investigations of substrate-induced magnetic film properties. These considerations let us use a slightly different template with a well-characterized and defined modification of the Cu(001) surface, i.e., the Cu(l 1 13). The main difference between Cu(00l) and Cu(l 1 13) is the reduced symmetry of the Cu(l 1 13) surface due to the existence of well-oriented steps. The influence of the symmetry on the magnetic properties o...

We studied the surface diffusion of hydrogen on Ni(lOO) by linearly diffracting a probe laser beam from a monolayer grating of adsorbed hydrogen. In the temperature range from 110 to 161 K, the Arrhenius plot shows a significant level off of the diffusion coefficient D{T) at around 144 K. Such a deviation from the behavior of a thermally activated diffusion suggests the onset of quantum tunneling diffusion.PACS numbers: 68.35.Fx, 66.10.Cb, 68.35.Ja One of the most interesting aspects of the transport of massive particles in condensed media is quantum tunneling motion. The process is most sensitive to various interactions and their interplay. The motion of the hydrogen atom and its isotopes in solids has attracted considerable attention in the past two decades [1,2]. Because the characteristic vibrational energies of hydrogen atoms in or on a solid are much closer to the substrate phonon energies, the screening effect is expected to be much stronger. The isotope dependence of the quantum motion of hydrogen should be the strongest of all stable elements because of the large mass ratios. As chargeless particles, hydrogen atoms may be one of the best systems for investigation of transport phenomena in the strong scattering limit, which will complement the investigation on electrons and electromagnetic radiation [3][4][5][6]. Theoretical studies and indirect experimental evidences indicate that the wave functions of hydrogen atoms on metals have sufficiently large spreads that the transport and vibrational properties may behave more like band electrons [7]. Experimentally, hydrogen atoms are relatively easily dissolved into many metals [1]. They also readily adbsorb onto the surfaces of many materials [8].However, direct experimental observation of the quantum motion of hydrogen atoms is rather scarce. The mobilities of hydrogen atoms in bulk solids have not been found other than those characteristics of thermally activated diffusions. The theories of quantum transports of hydrogen atoms such as the ''small-polaron'' theories and narrow-band particle models, all predicting power-law behaviors for diffusion coefficients at low temperatures, have yet to be tested against experimental observation [1,9,10].Recently, Gomer and co-workers measured the surface diffusion coefficients D(T) of hydrogen and its isotopes on W and Ni single crystals using a field-emission microscope (FEM) [11,12]. They found that the Arrhenius plots of D(T) level off at temperatures as high as 130 K and that D(T) exhibits large and yet reduced (from a rigid lattice model) isotope dependence, indicating the onset of quantum tunneling motion and a large mass renormalization effect. Even more interesting is that in a large temperature range from roughly 100 down to 30 K D(T) shows little temperature dependence. These results present a new case which is not yet explained satisfactorily by the existing theories [13]. On the experimental side. Comer's reports stand alone and the progress with the field-emission microscopy has been slow. This is mainl...