We study the pattern evolution of pre-rippled Au͑001͒ during sputtering by an ion beam that is incident perpendicular to the initial ripple in azimuth at a grazing angle. Prepatterned ripples decay exponentially with time and new ripples develop only after extended flat areas form along the crossing-ion beams. Hence, the superposition of the initial and new ripple patterns does not occur. The kinetic behaviors of new ripples growing on pre-rippled Au͑001͒ by the crossing-ion beams are distinct from those on initially flat Au͑001͒. When comparing the pre-rippled surface to the initially flat surface, the morphological evolution is substantially influenced by enhanced nonlinear effects such as redeposition.

Au(001) surfaces that have been prepatterned into a rippled morphology develop one-dimensional nanodot arrays (nanobeads) selectively along the ripples when bombarded with energetic ions at an angle that is normal to the average surface orientation. By quantifying the shape and morphology of these arrays, we show experimentally and by numerical simulations of an extended Kuramoto-Sivashinsky equation that the degree of one-dimensional order of the nanobeads can be optimized by considering initial rippled surfaces with various wavelength and roughness values. Our simulations employ physical units and use the experimental topographies as initial conditions. Such nonideal shapes are key to elucidating the influence of nonlinear effects (like conformal interface motion and local redeposition) since the early stages of the dynamics for these prepatterned systems. In spite of the fact that the evolution of the surface morphology becomes far from trivial under these circumstances, our continuum model is able to reproduce the experimental results quantitatively, in contrast to relevant alternative models in the context of surface nanopatterning by ion-beam bombardment.

We grow Fe film on (4 × 2)-GaAs(100) at low temperature, (∼ 130 K) and study their chemical structure by photoelectron spectroscopy using synchrotron radiation. We observe the effective suppression of As segregation and remarkable reduction of alloy formation near the interface between Fe and substrate. Hence, this should be a way to grow virtually pristine Fe film on GaAs(100).Further, the Fe film is found stable against As segregation even after warmed up to room temperature. There only forms very thin, ∼ 8Å thick interface alloy. It is speculated that the interface alloy forms via surface diffusion mediated by interface defects formed during the low temperature growth of the Fe film. Further out-diffusion of both Ga and As are suppressed because it should then proceed via inefficient bulk diffusion.1 Fe film on GaAs(100) has been extensively studied as a representative system for ferromagnetic metal-semiconductor heterostructure.[1] Due to the small lattice mismatch (∼ 1.3%) between the double of the lattice constant of Fe and that of GaAs, epitaxial growth of Fe film on GaAs(100) is achieved. However, alloy formation near the interface and serious outdiffusion of both Ga and As from the bulk have been notorious problems.[2, 3, 4] There have been various attempts to solve those problems such as S-passivation of GaAs surface[5] and insertion of Er layer between Fe and GaAs [6]. For both cases, the interfacial reaction is reduced to some extent, but the segregated As is still observed. Chye et al. [7] grow Fe film on GaAs (100) at 120 K and also insert Al interlayer. They find improved squareness in magnetic hysteresis and reduced interface states in photo-luminescence spectra. However, no direct investigation on atomic and chemical structure of the Fe film is made.In the present work, we grow Fe film on GaAs(100) around 130 K, and examine the possibility of kinetic stabilization of the Fe film by photoelectron spectroscopy. Here, we report direct evidence for the effective suppression of the outdiffusion of both Ga and As and the minimal formation of the interface alloy during the growth of the Fe film. When the film is warmed up to room temperature, there forms very thin alloy limited near to the interface. Further outdiffusion of both Ga and As is, however, still suppressed.All the experiments are performed at 2B1 beam line of Pohang light source in Korea. It is equipped with an electron energy analyzer and low energy electron diffraction (LEED) optics. The base pressure of the chamber is below 5 × 10 −10 Torr. For the substrate, we use Si-doped, n-type GaAs(100). Repeated sputtering and annealing produce clean and wellordered GaAs substrate; sputtering is made by Ar ion beam of 0.5 K eV, with its incidence angle 45 o from surface normal to minimize surface damage. Annealing is made at 840 K for 30 minutes. As-prepared surface shows well-defined 4 × 2 LEED pattern.We use an e-beam evaporator to deposit Fe film whose thickness is determined by a quartz microbalance that is calibrated by in situ surfac...

No abstract