College Park, Maryland      June 6 - 10 , 2004

ACNS Home Meeting Announcement Committees and Contacts Registration Abstract Submission Travel Reimbursement Dates to Remember

M2-A3 (11:15 AM): Magnetic behavior of epitaxial Europium thin films

S. Soriano, K. Dumesnil, C. Dufour, T. Gourieux (L.P.M. (UMR-CNRS 7556) Universite H. Poincare, BP 239, 54506 Vandoeuvre Cedex, France), A. Stunault (Institut Laue Langevin), M. Hennion (Laboratoire Léon Brillouin, C.E.A Saclay, 91191 Gif-sur-Yvette Cedex), J. Borchers (NIST Center for Neutron Research), Ph. Mangin (L.P.M. (UMR-CNRS 7556) Universite H. Poincare, BP 239, 54506 Vandoeuvre Cedex, France; NIST Center for Neutron Research)

We present the magnetic behavior of thin films of a very atypical rare earth, Europium, which has only recently been grown by molecular beam epitaxy along the [110] direction to form a bcc single-domain crystal[1]. In these (110) Eu films, the [001] axis lies in the growth plane, whereas the two others cubic axes ([100] and [010]) are out of the plane at 45° from the growth direction. Below a temperature T_cl, the sample is clamped to the (sapphire + niobium) substrate, which gives rise to strains eps(T). T_cl decreases and eps(T) lowers as the europium thickness is increased. The magnetic structures have been determined using Resonant x-ray Magnetic Scattering at ESRF and neutron scattering at LLB and at the NIST Center for Neutron Research. As in bulk europium the thin films exhibit a helical magnetic ordering at T_N = 90 K with magnetic propagation vectors along the cubic axes [100], [010] (out of the plane of the sample) and [001] (in the plane of the sample), which gives rise to three kinds of magnetic domains D1, D2 and D3 respectively. However, as the temperature is lowered two phenomena are observed:

i) The wave vectors of the D1 and D2 domains leave the cubic directions and move closer to the [110] growth direction. They rotate of an angle b(T).

ii) The D3 domains vanish at a temperature T₁ and are only restored (and even, as in the bulk, fill then the whole sample) if a critical field H_c(T) is applied along [001]. β (T), T1 and H_c(T) are shown to be correlated to the strain components eps(T), themselves related to the thicknesses of the samples. The role of the shear deformations is shown to be of prime importance. In addition, unlike in the bulk [2], it seems the magnetic transition is no longer of first order. To explain these evolutions, several mechanisms are discussed. They are based on surface dipolar effects, exchange anisotropy induced by the deformations and on magnetoelastic contributions.

[1] S. Soriano, K. Dumesnil, C. Dufour, D. Pierre (in press) Crystal Growth (2004)
[2] A.H. Milhouse, and K.A. McEwen, Solid State Com., 13, 339-345 (1973)

Back to the Program