Papers

AVS1996 Session MI-ThA: Growth and Structure of Magnetic Surfaces

Thursday, October 17, 1996 1:30 PM in Room 106A/B

Thursday Afternoon

Start	Invited?	Item
1:30 PM		MI-ThA-1 Structure, Morphology, and Giant Magnetoresistance of Spin Valves W. Egelhoff, Jr. (National Institute of Standards & Technology) The three basic types of GMR spin valves (top, bottom, and symmetric) have been investigated and optimized. The GMR values achieved are the largest values ever reported for such structures (top 10.5%, bottom 17.0%, and symmetric 23.4%). The key ingredients in this achievement stem from applying basic lessons of surface science to the optimization of these systems. Some elements in optimizing these systems are choice of substrate, preparation of substrate, suppression of interdiffusion at interfaces, the use STM to monitor roughness, low-temperature film growth, and the use of surfactants. The author wishes to ackknowledge his collaborators in this work: P.J. Chen, M.D. Stiles, C.J. Powell, R.D. McMichael, H.D. Chopra, J.H. Judy, K. Takano, and A.I. Berkowitz
2:10 PM		MI-ThA-3 Schottky Barrier Formation vs Interfacial Structure for Fe on Reconstructed Surfaces of GaAs(001) B. Jonker, E. Kneedler, P. Thibado, O. Glembocki, L. Whitman, B. Bennett (Naval Research Laboratory) Electrical contacts to a semiconductor formed using a magnetic material are the basis for a spin injection device. We have measured the Schottky barrier height for Fe on GaAs(001) as a function of the GaAs surface reconstruction, and correlated these measurements with the atomic scale structure of the Fe/GaAs interface as derived from photoelectron diffraction (PED) and STM. We discuss these results in the framework of several models for Schottky barrier formation. The samples were grown by MBE in a multi-chamber uhv facility which includes chambers for PED and STM. After coating with 50 \Ao\ of Au, the samples were removed from the system for photoreflectance measurements to determine the Schottky barrier height. On the As-rich 2x4 and c(4x4) surfaces, we find nearly identical barrier heights of 0.8 eV, and a very low density of interface states in the Fe/2x4 samples. From the STM and PED data, we conclude that the Fe/GaAs interface is formed by Fe bonding with an As monolayer after dissolution of the surface As dimers and some Ga into the Fe film.\super 1\ Our measured barrier heights and interface model are consistent with the effective work function model of the Schottky barrier, which predicts a value of 0.8 eV (Fe-As bonding), and inconsistent with the pure Schottky model which predicts a value of 0.43 eV. We compare these results with similar data for growth on the Ga-rich 4x6 reconstruction and on ZnSe(001) surfaces. \super 1\ E. Kneedler, et al. JVST-B, July/August (1996)
2:30 PM		MI-ThA-4 In-situ Ellipsometric Control of Metallic Multilayer Thin Films X. Gao, S. Heckens, D. Glenn, J. Woollam (University of Nebraska, Lincoln) Thin metallic multilayers with the layer thickness in the range of less than one nanometer to several tens of nanometers are of special interest as magneto-optical recording and giant magnetoresistance media. Precise control of layer thicknesses is very important for studying magneto-optical Kerr rotation, interlayer coupling and other magnetic properties of these multilayers. In-situ spectroscopic ellipsometry (In-situ SE) has been used to determine the growth rates of different metal films deposited in a sputtering chamber in order to establish growth conditions for the multilayers. Spectroscopic ellipsometry data were also taken during growth of multilayer thin films. By analyzing the SE data, we were able to determine the thicknesses of the individual layers in the superlattice as well as their spectrally dependent optical constants as thin films. The in-situ SE data and ellipsometric analysis of [Co/Au] and [Co/Ni] multilayer on thick gold coated silicon are presented in this work. Good consistency was obtained by applying the in-situ SE model to the ex-situ ellipsometric data for the [Co/Au] multilayer with thick (6nm) gold films. Oxidation effects were found in the [Co/Au] structures with thin gold films (2nm) and [Co/Ni] structures. X-ray diffraction (XRD) studies were also used to determine the periodic superlattic thicknesses. Less than a few percent difference was found by comparing the XRD results to the in-situ SE results. /super */ Research supported by NSF grants OSR-9255225 and DMR-9222976.
2:50 PM		MI-ThA-5 Structural and Magnetic Properties Calculated using Density Functional Theory H. Jansen (Oregon State University) The strength of calculations performed in the framework of Density Functional Theory is the ability to calculate the relation between total energy and structure, even for transition metal compounds. These results can be used to help interpret experiments in which transition metals are grown in a variety of crystal structures on appropriate surfaces. Besides the crystal structure one can also investigate the magnetic structure. A very interseting subject in this context is magnetic anisotropy, where energy differences are very small. Although in principle our calculations will yield the correct results, in practice we encounter limitations due to the approximations used to treat exchange and correlation. In this talk I present a general overview and some specific examples of recent calculations relevant to surface science. This work is supported by ONR, grant N00014-9410326.
3:30 PM		MI-ThA-7 The Structure of the Ferromagnetic MnNi Surface Alloy S. Banerjee, S. Ravy, J. Denlinger, X. Chen, D. Saldin, B. Tonner (University of Wisconsin, Milwaukee) The bulk MnNi alloy is an antiferromagnet. We haved studied the surface alloy of this composition, both by growing ordered MnNi as a c(2x2) structure on Ni(100) substrates, and by co-deposition to form MnNi on Cu(100). In both cases, the alloy that is formed is ferromagnetic, with the Mn spins aligned parallel to those of the Ni. An earlier study of c(2x2) MnNi using LEED [1] concluded that the surface alloy was a bulk termination structure, with planes of Mn atoms arranged perpendicular to the surface. This structure does not agree with the measured ferromagnetism of the surface alloy. In order to learn more about the structure of this new, ferromagnetic, phase of MnNi, we have completed a detailed structure determination using x-ray photoelectron diffraction. We will present results analyzed by quantitative modelling, and by direct inversion (photoelectron holography). We propose that ferromagnetic Mn results from a structure in which Mn has only Ni atom near-neighbors. [1]M. Wuttig, T. Flores, and C.C. Knight, Phys. Rev. B 48, 12082 (1993).
3:50 PM		MI-ThA-8 Growth, Morphology and Magnetic Properties of Monolayer Epitaxial Co Films on Cu(100) S. Coyle, G. Hembree, M. Scheinfein (Arizona State University) Monolayer thick Co films grown epitaxially on single-crystal (bulk) Cu(100) were investigated using in-situ nanometer resolution ultrahigh-vacuum scanning electron microscopy, Auger electron spectroscopy (AES), and the surface magneto-optic Kerr effect (SMOKE). Growth rates were determined by AES peak height ratios. Growth was monitored from the initial nucleation of islands through thicknesses of several monolayers (ML) at room temperature and 60 C. Microstructural evolution during growth indicated that steps (bands) initiated exchange diffusion resulting in irregular, yet crystallographic faceting even at low coverages. Large facets which grow deep into step bands produce pits from which Cu is believed to migrate. AES measurements showed evidence for a Cu capping layer for coverages less than ~2 ML. In-situ SMOKE measurements indicate that these films become ferromagnetic at room temperature at ~1.7 ML. A second Co phase was detected in many films with out-of-plane remnence and a coercivity 5-10 times the in-plane value of ~10 Oe. Magnetic properties will be correlated with outstanding features in film morphology.This work is supported by the Office of Naval Research under grant No. N00014-93-1-0099.
4:10 PM		MI-ThA-9 Growth, Structure, and Magnetic Properties of Fe\sub 3\O\sub 4\ and \gamma\-Fe\sub 2\O\sub 3\ Epitaxial Thin Films Grown by Oxygen-Plasma-Assisted Molecular Beam Epitaxy Y. Gao (Pacific Northwest National Laboratory); P. Lubitz (Naval Research Laboratory); S. Chambers (Pacific Northwest National Laboratory) Single-crystal epitaxial thin films of Fe\sub 3\O\sub 4\ (001) and \gamma\-Fe\sub 2\O\sub 3\ (001) have been grown on MgO (001) using oxygen- plasma-assisted molecular beam epitaxy. The structure and magnetic properties of these films have been examined with a variety of techniques, including reflection high-energy electron diffraction (RHEED), low-energy electron diffraction (LEED), x-ray photoelectron spectroscopy and diffraction (XPS/XPD), and ferromagnetic resonance. Real-time RHEED reveals that film growth occurs in a layer-by-layer fashion. The \gamma\-Fe\sub 2\O\sub 3\ (001) film surface is unreconstructed whereas the Fe\sub 3\O\sub 4\ (001) surface exhibits a (\sr\2x\sr\2)R45\super o\ reconstruction. Application of a simple electron counting rule reveals that the bulk-terminated \gamma\-Fe\sub 2\O\sub 3\ (001) surface is autocompensated, but that of Fe\sub 3\O\sub 4\ (001) is not. The latter reconstructs by forming an ordered array of tetrahedral Fe vacancies and autocompensates in the process. Significant Mg outdiffusion accompanies film growth of both Fe\sub 3\O\sub 4\ and \gamma\-Fe\sub 2\O\sub 3\ grown at >= 400 \super o\C. XPD of Fe 2p and Mg KLL emission reveals that Mg substitutionally incorporates in the Fe\sub 3\O\sub 4\ and \gamma\-Fe\sub 2\O\sub 3\ lattices, forming solid solutions. Mg outdiffusion is kinetically hindered at a growth temperature of 250 \super o\C. The measured magnetic moments are 2300 G and 4500 G for the \gamma\-Fe\sub 2\O\sub 3\ films grown at 250\super o\ C and 450\super o\ C, respectively. The high value at 450 \super o\C may be a result of 10 at. % outdiffused Mg, which replaces Fe on minority sites.
4:30 PM		MI-ThA-10 Mg Outdiffusion on Ferrite Films Grown on MgO (001) K. Shaw, E. Lochner, D. Lind (MARTECH and the Florida State University); J. Anderson, M. Kuhn, U. Diebold (Tulane University) STM studies of 1-\mu\m thick films of single crystalline Fe\sub 3\O\sub 4\ grown on MgO(001) indicate that repeated annealing of the sample in UHV cause Mg diffusion through the Fe\sub 3\O\sub 4\ film. The onset of this effect was clearly seen by STM at room temperature for samples raised above 475 degrees C. It appears that the annealing process causes the migration of Mg from the substrate entirely through the Fe\sub 3\O\sub 4\ lattice, and that the migration tends to fill the surface layer first, with lower layers filling as anneal time is increased. Upon detection of this effect, several complementary sample analysis techniques were employed to determine the extent of the changes observed. XRD studies indicate shifts in the lattice constant from the cubic constant of magnetite, Fe\sub 3\O\sub 4\, (8.396\Ao\ ) toward the cubic lattice constant of magnesioferrite, MgFe\sub 2\O\sub 4\ (8.375\Ao\ ). SQUID magnetometry studies reveal a significant change in the magnetic behavior of the film - large decreases in the saturation moment, remanence, and coercive field. The Verwey transition is greatly altered in these films after the annealing sequence. XPS studies of the films confirm the presence of magnesium in the uppermost layers of the film, and indicate a concentration gradient, with the highest concentrations of magnesium in the surface layer. X-ray fluorescence in SEM qualitatively indicate the presence of magnesium throughout the film, consistent with migration of the magnesium from the substrate. These results are compared with those on an unannealed Fe\sub 3\O\sub 4\ film of the same thickness and growth parameters, which shows no magnesium migration into the film during growth up to substrate temperatures of 300 C. Transport studies on these films to determine the differences in conductivity changes due to the alterations in stoichiometry and the concentration gradient of the magnesium in the film will also be discussed. Research funded by grants from NSF.