Spinorbit interaction in the 2D electron system of InAs
Christopher Schierholz, Guido Meier, Toru Matsuyama
Introduction
Spinelectronic devices such as a spintransistor are based on control of electron
spin. The current is spinpolarized in the injector, spin modulated in the semiconductor
channel via spinorbit interaction originating from the Rashba effect, and detected
by a spinanalyzer. For such a device to function, spinmodulation in the semiconductor
must exist and the strength of it known. Control of the spinmodulation by an
applied voltage is wishful.
The common method to determine spinorbit interaction in a semiconductor is
by analysis of beating patterns occurring in Shubnikovde Haas (SdH) oscillations
[1,2]. The analysis of weak localisation and antilocalization provides an additional
method of measuring this interaction [3,4], and, as opposed to SdH analysis
it is performed in low external fields, in which the proposed spinelectronic
devices are to function.
Fig. 1: Image of a sample in Corbino geometry used in transport experiments.
Weak Localization and Antilocalization
Weak localization and antilocalization are caused by intrinsic interference
of transport electrons in semiconductors whilst being scattered in a closed
loop. Weak localization is a common phenomenon depending on inelastic and elastic
scattering, weak antilocalization however only occurs in the presence of spinorbit
interaction. Then the positive magnetoconductance caused by localization can
be turned into negative magnetoconductance for low magnetic fields.
External magnetic fields applied perpendicular to the plane of transport destroy
timereversal symmetry of closed paths in the semiconductor and thus also the
localization effects. As a consequence, when increasing the magneticfield strength
from zero one at first observes a decreasing conductance originating from the
suppression of antilocalization followed by an increasing conductance due to
the destruction of localization. This can be observed for the experimental data
in Fig. 2 (solid line).
Fig. 2: Experimental localization data and fit result.
By fitting the experimental data with the theoretical model of Lyanda, Geller,
and Pikus [5,6] we can determine inelastic and spinorbit characteristic fields
in the semiconductor. We use these to determine the spinorbit interaction strength,
i.e. the Rashba parameter [7] whereby we find good agreement with current band
structure calculations [8,9]. We aim at deeper understanding of the semiconductor
system and the tunability of spinorbit interaction. Furthermore, our goal is
to design semiconductor heterostructures that exhibit strong spinorbit interaction.
Literature
[1] T. Matsuyama et al., Phys. Rev. B 61, 15588 (2000)
[2] C.M. Hu et al., Phys. Rev. B 60, 7736 (1999)
[3] P.D. Dresselhaus et al. , Phys. Rev. Lett. 68, 106 (1992)
[4] G.L. Chen et al. , Phys. Rev. B 47, 4084 (1993)
[5] S.V. Iordanskii, Y.B. LyandaGeller, and G.E. Pikus, JETP Lett. 60,
206 (1994)
[6] W. Knap et al., Phys. Rev. B 53, 3219 (1996)
[7] Ch. Schierholz, T. Matsuyama, U. Merkt, and G. Meier, Phys. Rev. B 70, 233311 (2004)
[8] S. Lamari, Phys. Rev. B 64, 245340 (2001)
[9] R. Winkler, private communication
further reading: C. Schierholz: "Rashba SpinOrbit Interaction in Low and High Magnetic Fields", Cuvillier Verlag, ISBN: 3865374492
