Ferromagnetic Metal/Compound Semiconductor Heterostructures: Growth, Interfacial Reactions and Spin Injection

Christopher J. Palmstrom
University of Minnesota

Current semiconductor technology relies on the charge of an electron and the ability to control the transport of charges for the device performance. Spin has recently demonstrated potential as a storage medium for classical and quantum information within semiconductors, which has lead to the concept of 'spintronics' where the spin state of the carrier in the semiconductor is utilized. An ideal spintronic material would be one that enables the transport of only one spin carrier across the interface between it and an unpolarized material without any spin flip scattering. This concept has been the driving force for investigating epitaxial single crystal magnetic material/semiconductor heterostructures with minimal interfacial defects from, for example, magnetic dead layers formed by interfacial reactions, dislocations or other spin flip scattering mechanisms.

Molecular beam epitaxial (MBE) growth in combination with in-situ STM, XPS, Auger, RHEED and LEED and ex-situ XRD, RBS, TEM, magnetotransport and magnetic characterization was used to develop an understanding of issues of lattice matching and strain, bonding, minimization of interfacial reactions and growth temperature, and control of structural, electronic and magnetic properties during the growth of elemental and metallic compound/compound semiconductor heterostructures. The efficiency of the spin polarized current injected from the ferromagnetic contact was determined by measuring the electroluminescence polarization of the light emitted from a Al1-xGaxAs quantum well embedded within the depletion region of an n-p diode. The influence of interfacial reactions and the semiconductor band structure on the measured electroluminescence polarization will be discussed.