Epitaxial growth of Co on GaAs͑001͒ and its in-plane magnetic anisotropy are studied using reflection high-energy electron diffraction, a high-resolution transmission electron microscope, and the magneto-optical Kerr effect. In the initial and final stages of growth, Co exists in single-crystalline body-centered-cubic ͑bcc͒ and hexagonal-closed-packed ͑hcp͒ phases, respectively, while in the middle stage the coexistence of the bcc and hcp structures is observed. For the bcc Co thin films on GaAs͑001͒, a fourfold in-plane magnetic anisotropy with easy axes along the ͗100͘ directions is realized and discussed. ͓S0163-1829͑98͒04915-7͔The 3d transition metals exist in a variety of crystallographic and magnetic phases. Thin-film growth of these materials on crystalline substrates allows the forces present at the interface to drive the film into specific crystalline structures. These structures may be in a thermodynamically stable phase, a known high-pressure or high-temperature phase, or even a phase not previously observed. They greatly increase the variety of magnetic materials by essentially making ''new'' materials from ''old'' elements. 1The epitaxial growth of Co films serves as a good example. It is known that the hexagonal-close-packed ͑hcp͒ and face-centered-cubic ͑fcc͒ structures are, respectively, stable and metastable phases of Co. The body-centered-cubic ͑bcc͒ structure, which does not occur in nature, was realized by Prinz with epitaxial growth on a GaAs͑110͒ substrate. However, it was later pointed out by Liu and Singh that bcc Co is not a true metastable phase but a force-induced phase. 3The in-plane magnetic anisotropy of such a bcc Co thin film on GaAs͑110͒ was further determined and a negative value for the cubic anisotropy constant K 1 was proposed.2 If this were true, a fourfold in-plane magnetic anisotropy with easy axes along the ͗110͘ direction would then be expected in the bcc Co films on GaAs͑001͒ substrates. In fact, a fourfold in-plane magnetic anisotropy with the easy axes along the ͗100͘ rather than the ͗110͘ direction was observed by Blundell et al. 4 Interestingly, it was later argued by Gu et al. that Co films grown on GaAs͑001͒ were actually not bodycentered cubic but two-domain hexagonal close packed by which the fourfold magnetic anisotropy along the ͗100͘ direction could be explained by such a microstructure.5 Obviously, the epitaxial structure of Co on GaAs͑001͒ and its magnetic anisotropy are still very controversial. In this work, we present a clear picture of the epitaxial growth of Co on GaAs͑001͒, which clears up the previous controversy about the structure of Co thin films on GaAs͑001͒. With the help of this clear picture, we prove that the bcc Co films on GaAs͑001͒ show a fourfold in-plane magnetic anisotropy.Co films were grown in a molecular-beam epitaxy ͑MBE͒ growth chamber connected with the VG-ESCALAB-5 electron spectrometer system. The Te-doped GaAs͑001͒ singlecrystal wafers were polished and treated by ordinary device cleaning process. The final substrate cleaning w...

The body-centered-cubic (bcc) phase of Ni, which does not exist in nature, has been achieved as a thin film on GaAs(001) at 170 K via molecular beam epitaxy. The bcc Ni is ferromagnetic with a Curie temperature of 456 K and possesses a magnetic moment of 0.52+/-0.08 micro(B)/atom. The cubic magnetocrystalline anisotropy of bcc Ni is determined to be +4.0x10(5) ergs x cm(-3), as opposed to -5.7x10(4) ergs x cm(-3) for the naturally occurring face-centered-cubic (fcc) Ni. This sharp contrast in the magnetic anisotropy is attributed to the different electronic band structures between bcc Ni and fcc Ni, which are determined using angle-resolved photoemission with synchrotron radiation.

Permalloy with a body-centered-cubic structure has been grown on GaAs(001) by molecular beam epitaxy. Its magnetism, Curie temperature, and magnetic anisotropy are determined experimentally and compared to those of conventional face-centered-cubic Permalloy. Unexpectedly the vanishing magnetic cubic anisotropy in Permalloy is found to be independent of its atomic structure but depends only upon the stoichiometry of Fe and Ni in the FexNi1-x alloy. This observation is further investigated and confirmed by first-principles electronic band calculations, which help to understand the long-standing issue of why Permalloy should be a soft magnet.

Single crystalline Fe x Cu 1−x alloys over the entire composition range have been prepared successfully on GaAs͑001͒ via molecular beam epitaxy. The films are body centered cubic (bcc) at high Fe concentration ͑x Ͼ 0.75͒, body centered tetragonal (bct) or face centered cubic (fcc) at the lower Fe concentration depending on the film thickness. Long-range ferromagnetic order is observed when x Ͼ 0.33 where the thickness is 6 nm, and the effective magnetic moment per Fe atom decreases as the Fe concentration increases. Fe-Cu alloy films with four-fold magnetocrystalline anisotropy were obtained.

No abstract