Spin Injection
The magnetic dipole moment that is associated with a particle’s spin state is a fundamental property with manifestations in many branches of physics. In Condensed Matter Physics, electron (and hole) spin offers a rich phenomenology and this volume gives a
- PDF / 533,425 Bytes
- 29 Pages / 439.37 x 666.142 pts Page_size
- 63 Downloads / 223 Views
10.1 Introduction The magnetic dipole moment that is associated with a particle’s spin state is a fundamental property with manifestations in many branches of physics. In Condensed Matter Physics, electron (and hole) spin offers a rich phenomenology and this volume gives an appreciation of many of these topics. This chapter is concerned with transport phenomena that involve carrier spin as well as charge. Since electric currents in solids typically utilize carriers within a thermal energy range of the Fermi level EF , our focus is on the spin state of conduction electrons near EF and we examine how spin states can affect charge flow in the form of current and voltage distributions. A unique feature of this study is that interface effects are important, as can be understood with a simple description of basic concepts. An electric current in a ferromagnetic metal (F ) is spin polarized, a fact known since the early part of the last century [1]. An electric current in a nonmagnetic metal (N ) is not spin polarized. When a ferromagnet is in interfacial contact with a nonmagnetic metal, the current crossing the F /N interface is spin polarized. This phenomenon is broadly known as spin injection [2]. It results in a nonequilibrium population of spin polarized electrons in N, a spin accumulation. These nonequilibrium spins spread by diffusion, and this can result in a small current of oriented spins in N and/or F . The phenomenology of spin injection, accumulation and detection was developed for metals. Research on a variety of topics in the late 1990s caused high interest in the plausibility of spin injection in semiconductors. Recent results [3] have demonstrated the effect, and have confirmed that the models of spin injection, accumulation and detection developed for metals are also valid for semiconductors. 10.1.1 History Our knowledge of spin dependent transport in the solid state has derived from several key experiments and theoretical insights. Tedrow and Meservey [4, 5] fabricated
280
M. Johnson
planar F /I /S tunnel junctions, where S was superconducting aluminum, F was a transition metal ferromagnet, and I was an aluminum oxide tunnel barrier. After applying a field of order 1 tesla in the film plane, tunnel conductance spectroscopy was used to demonstrate that the current tunneling into the quasiparticle states of the aluminum was spin polarized. These experiments gave the first empirical estimate of the fractional polarization, P, of such currents. Shortly thereafter, Julliere [6] extended the work of Meservey et al. to form a structure that has become important for applications. For his Ph.D. thesis, he made a tunneling structure F 1/I /F 2 in which he substituted a second ferromagnetic film for the aluminum electrode, thereby inventing the magnetic tunnel junction. He succeeded in measuring a tunnel magnetoresistance only at low temperature, but 25 years later his invention would become a technological success. The prevailing opinion in the 1970s was that any spin polarized current that crossed a F /N interfac
Data Loading...
