Origin of the Magnetic Proximity Effect
- PDF / 155,994 Bytes
- 11 Pages / 612 x 792 pts (letter) Page_size
- 36 Downloads / 234 Views
Q5.2.1
ORIGIN OF THE MAGNETIC PROXIMITY EFFECT Miguel Kiwi Facultad de F´ısica, Pontificia Universidad Cat´olica de Chile, Casilla 306, Santiago, Chile 6904411 (Dated: January 16, 2003)
Abstract The magnetic proximity effect (MPE) has attracted the attention of theorists and experimentalists for at least three decades. Lately, the relevance of the effect for the development of nanodevices has revived interest on the subject. Here we review how the field has evolved, centering our attention on metal-metal and metal-insulator systems. We describe, and critically compare, the different theoretical approaches that have been put forward, as well as their limitations. An evaluation of the relationship between existing theories and available experimental results is also attempted.
Q5.2.2
1.
INTRODUCTION
Almost 30 years ago Zuckermann [1], in a pioneer work, showed theoretically that a system formed by a thin film of a weak itinerant ferromagnet (FM), in atomic contact with a thick film of an enhanced paramagnetic metal (PM), can exhibit a nonzero Curie temperature. Shortly thereafter this theory was extended to a system formed by a ferromagnet in contact with an antiferromagnet [2]. The technique that was used consisted in solving a simplified version of the integral equation derived from the Landau-Ginzburg theory of phase transitions. This equation reads as follows:
M(r) =
d3 r U(r) χT (r − r ; 0) M(r ) − U(r) M 3 (r)
ω
G0p (ω) ,
(1.1)
where M(r) is the local magnetization at point r, U(r) the Hubbard exchange constant, χT (r; 0) the static r-dependent magnetic susceptibility of the non-interacting conduction electrons and G0p (ω) the propagator for these momentum p conduction electrons. The above equation was analyzed using the procedure set forth by Werthamer [3], for the proximity effect between two superconductors. This way the integral equation 1.1 is transformed into a differential equation. The solution of this equation was obtained by matching the superconducting order parameter and its slope at the interface. However, this procedure is not reliable when applied to magnetic interfaces. First, with the possible exception of very weak magnets, the coherence length of the magnetic order parameter M(r) is smaller than a typical lattice parameter, which excludes a continuous differential equation as a valid description. In addition, there is no a priori reason for requiring that the slope of M(r) be continuous across an interface. For all these reasons a difference equation formulation seems to be appropriate. On the other hand, in 1978 Bergmann [4] published experimental results on very thin film magnetization of Ni, Co and Fe deposited on a PM metallic substrate (Pb3 Bi). He interpreted his measurements to imply that the magnetization of the first few FM layers, closest to the PM, was substantially different from the bulk value. Moreover, he concluded that Ni did not develop a magnetic moment in films thinner than three monolayers. This in itself is an interesting fact, since magnetic dead layers
Data Loading...
