The electronic rendition of the Hanle effect, which is interpreted as the ensemble dephasing of a spin accumulation in the semiconductor under a perpendicular magnetic field, has been one of the most widely utilized and effective methods of measuring spin lifetime, spin accumulation, and spin transport in semiconductors. However, the origin of the Hanle magnetoresistance in the three-terminal (3T) setup has been intensively questioned both theoretically and experimentally; this is in contrast to the nonlocal four-terminal (NL-4T) measurement, which is accepted as reflecting spin accumulation and its spatial decay in metals and semiconductors alike. Here, we present results from 3T and NL-4T Hanle measurements on the same spin injection and detection devices with an ${\mathrm{Al}}_{0.3}{\mathrm{Ga}}_{0.7}\mathrm{As}:\mathrm{Si}$ semiconducting channel. The use of ${\mathrm{Al}}_{0.3}{\mathrm{Ga}}_{0.7}\mathrm{As}:\mathrm{Si}$, a persistent photoconductor, enables examination of the evolution of both types of Hanle signals with varying carrier density in the channel on one and the same device via in situ photodoping. We observe that the 3T and NL-4T Hanle signals exhibit similar Lorentzian line shapes, and thus yield similar spin lifetimes at all carrier densities. Moreover, the amplitudes of both types of Hanle signals are found to be consistent with each other, showing a similar exponential decrease with carrier density and in agreement with the Valet-Fert theory, in contrast to devices with artificial oxide barriers. These observations provide compelling evidence that in devices in which the spin injectors and detectors are engineered to minimize the presence of localized states, the 3T Hanle measurements provide a reliable probe of the spin accumulation and its dynamics in the semiconductor channel.
Read full abstract