Spin-injection into III-V Semiconductors
To realize semiconductor based spintronic devices, the basics of spin injection, transport, manipulation and detection must be understood. Our group focuses in MBE-grown epitaxial ferromagnetic metal/semiconductor heterostructures for spin injection and detection. We investigate and manipulate the interfacial properties through growth and structural characterization. The interface bonding and the electronic band structure as well as the doping profile of the semiconductor immediately beneath the metallization have been found to be important for controlling the interfacial spin transport. Both spin-LEDs and lateral 3 and 4-terminal spin transport measurements are done in collaboration with Professor Paul Crowell’s research group at the University of Minnesota. These structures are used to investigate spin-transport and the spin dynamics in the presence of spin-orbit and nuclear hyperfine interactions. Our interconnected MBE growth and surface characterization chambers allow for the growth of FM/SC heterostructures with high quality interfaces in addition to LEED, RHEED, STM and XPS studies of the growth process and also to study the fundamental effects of interfaces on spin injection. So far we have investigated epitaxial Fe and epitaxial Heusler alloys (Co2MnSi, Co2MnGe, Ni2MnIn) as spin contacts to III-V semiconductors. We are also interested in working on ferromagnetic materials with high spin polarization. Collaborations with Professor Chris Leighton (University of Minnesota) involve developing highly spin polarized CoS2 contacts to GaAs.
This work is primarily supported as part of the NSF-MRSEC at the University of Minnesota.
Magnetic Tunnel Junctions for Spin Torque Transfer Memories
Magnetic tunnel junctions (MTJs) are structures containing two ferromagnetic layers (one with fixed magnetization, the other free) with a thin (1-2 nm) insulator layer in between (typically MgO). Electrons tunneling from one ferromagnetic electrode through the tunnel barrier into the other experience a resistance determined by the magnetizations of the ferromagnetic electrodes. This is known as the spin valve effect. If the magnetization of the two layers are parallel, the electrons will experience a “low resistance”, or if the magnetization of the two layers are anti parallel, the electrons will experience a “high resistance”. The magnitude of this effect is paritally dependent on the spin polarization of the current. This ratio between the high and low resistance cases is called a tunneling magnetoresitance ratio (TMR). TMR ratios are important for use in the reading of magnetic storage devices, such as hard drives or magnetic random access memory (MRAM).
MRAM devices also employ the use of the spin torque effect. When an electron of a certain spin orientation travels through a ferromagnet, the ferromagnetic domains experience a torque that aligns their magnetic moments with the traveling spin. This effect can be used to write data bits in a spin-torque transfer memory (STTM) device. STTM devices could possibly have long data retention times, and low switching currents, which would dramatically reduce power consumption.
The Palmstrom group explores the use of half-metallic Heusler compounds as ferromagnetic electrodes in these structures. Materials that are predicted to be highly spin polarized (Co2MnSi, Co2MnGe, Co2FeSi) are being studied. The interconnected UHV system of the group allows for the growth of high quality interfaces between various layers of the heterostructure, in addition to studying the effects of doping on the reduction of minority in-gap states. Our work includes the study of crystalline quality, composition, and interfaces through in-situ STM and EDX, and X-ray diffraction. Vibrating sample magnetometry (VSM), magneto-optic Kerr effect (MOKE) and superconducting quantum interference devices (SQUID) allow for the study of the magnetic properties of the ferromagnetic Heusler compoungs. Spin-polarization can be measured through TMR ratio measurements, in situ point-contact Andreev reflections (PCAR), and through spin-tunneling spectroscopy. Jim Allen from the Department of Physics at UC-Santa Barbara and Alexander Kozhanov from the California Nanosystems Institute (CNSI) are collaborators on this project.
This work is primarily supported by the Semiconductor Research Corporation and Intel.