Magnetic properties of thin epitaxial films investigated by spin-polarized photoemission

Magnetic behavior of metastable Fe films grown on Ir(1 1 1)

We investigated the growth of ultra-thin Fe films on Ir(1 1 1) by means of in situ low energy electron diffraction and spin-resolved photoemission techniques. We observe a (1  ×  1) diffraction pattern, characteristic of the fcc substrate, below four monolayers (ML). Then, a complex superstructure starts to develop, compatible with the formation of bcc-like Fe domains aligned with the substrate according to the Kourdjumov-Sachs orientation relationships. The analysis of the diffraction patterns reveals a progressive evolution towards a fully relaxed bcc lattice, characteristic of bulk Fe. Both photoemission (filled states) and inverse photoemission (empty states) results show characteristic features related to the contribution of the Fe layer, evolving towards those observed on the Fe (1 1 0) bcc surface. Spin resolution allows to detect a spectral polarization above 4 ML, corresponding to the formation of bcc Fe, which gradually increases indicating the formation of an in-plane magnetized ferromagnetic layer in thick films. No in-plane net magnetization is detected in thinner films, independent of the sample temperature down to 30 K. Following recent investigations on the Fe/Ir(1 1 1) system with microscopy techniques, we link this observation to the stabilization of a non collinear spin structure yielding an overall nil magnetization.

X-Ray Photoelectron and Auger Electron Forward-Scattering Studies of the Epitaxial Growth of Fe on Ag(100)

A controversy has arisen in the past year over whether or not the growth of Fe on Ag(100) at room temperature occurs by a layer-by-layer mechanism. The present work attempts to address this controversy with an investigation of the issues, primarily by x-ray photoelectron (XPS) and Auger electron forward scattering, but with important supporting data from low-energy electron diffraction (LEED), and reflection high-energy electron diffraction (RHEED) oscillations. The results of this work suggest that the origin of the controversy lies in different substrate preparation techniques which produce different atomic step densities on the Ag(100) surface. The step sites are implicated as being the initiators of major departures from a layer-by-layer growth mode whenever most of the deposited Fe atoms have sufficient mobility to reach these steps. However, even when the Fe atoms cannot reach these steps it appears that atomic place-exchange occurs with ≥25% of the top-layer Ag atoms. Atomic place-exchange mechanisms, which could account for this intermixing, have been observed in recent molecular-dynamics simulations of epitakial growth. Thus it seems probable that under the conditions that appear to produce layer-by-layer growth, the growth begins as layer-by-layer growth of an FeAg alloy, and only becomes layer-by-layer in pure Fe as the segregating Ag atoms gradually get left behind in the growing Fe film.

The Curie temperature of thin ferromagnetic films

The thickness-dependent Curie temperature T(c)(d) of thin ferromagnetic films is calculated within the molecular field approximation ('Weiss mean field') of the Heisenberg model. Two higher mean field theories are applied to obtain a quantitative improvement of the results: the Oguchi cluster method and the 'constant coupling approximation' (CCA). Analytical expressions are derived from difference equations or eigenvalue problems with an unknown parameter which can be solved numerically. Explicit expressions for T(c)(d) can be given if the interaction is restricted to next neighbour monolayers only, for any value of the spin S within the Weiss mean field and for S = 1/2 within the CCA. Effects of an enhanced interaction within the surface layers are briefly investigated. Calculated values of T(c)(d) for EuO are presented within the three models.