Optical Microscopy Reveals Benefits of Strain for Spintronics

Daniel S. Burgess

In experiments at Los Alamos National Laboratory in New Mexico, Scott A. Crooker and Darryl L.G. Smith have demonstrated that strain can be used to control the spin of electrons flowing in bulk GaAs. They monitored the effects of applied electrical, magnetic and strain fields with a scanning Kerr microscope and determined that the spatial spin coherence was particularly robust for spin manipulation induced by strain.

Scanning Kerr microscopy reveals the lateral drift and diffusion of spin-polarized electrons in GaAs optically injected at the point indicated with the red spot (top). Applied strain enables control of electron spins with a high spatial coherence (bottom). Courtesy of Scott A. Crooker.

The researchers produced two-dimensional images of spin-polarized electrons in 1-μm-thick GaAs that they optically pumped with a circularly polarized, 10- to 25-μW diode laser. The experiments were performed in vacuum at 4 K. The microscope setup featured a tunable CW Ti:sapphire laser whose lin-early polarized 50- to 100-μW probe beam was raster-scanned across the sample with 4-μm resolution. The rotation of the polarization measured in the reflected beam indicated the electrons’ spin component normal to the plane of the semiconductor crystal.

The results of measurements of GaAs subjected to stress from cryogenically applied pressure demonstrated that the spins could be made to precess independent of applied electrical bias in a manner that was controllable, reversible and uniform across the sample.

The spatial coherence of the precession was more robust than that produced with applied magnetic fields, which tended to decohere the spins at distances greater than one precession period.

The findings have potential implications for spintronics, in which devices employ electron spin as well as charge. With current and near-term applications in data storage, spintronics also promises to enable the development of semiconductor-based quantum computers.

Physical Review Letters, June 15, 2005, 236601.