Semiclassical interpretation of the spin interference effect observed in square loop arrays of In0.53Ga0.47As/In

Abstract

We present a semiclassical interpretation of the time-reversal spin interference (SI) observed in the square loop arrays made of In 0.53 Ga 0.47 As quantum wells [T. Koga et al., Phys. Rev. B 74, 041302 (2006)]. The simulated amplitude of SI as a function of the Rashba parameter α captured characteristic features in the experimental results if γ 8 eVÅ 3 is assumed for the bulk Dresselhaus spin-orbit constant γ . Our work proves the validity of the semiclassical approach to predict the effect of time-reversal quantum interference in mesoscopic systems and the values of the spin-orbit coefficients recently deduced from the weak localization/antilocalization experiment.

Direct extraction of electron parameters from magnetoconductance analysis in mesoscopic ring array structures

“…1(b), makes the trajectories inherently chaotic. This is an important improvement from our previous studies [11,12,16].…”

“…IV D) samples in the main text, whose sample names are KH3-1 and KH1-3, respectively. from 1.3 × 10 16 to 2.4 × 10 16 m −2 and from 1.1 to 2.2 μm, respectively, for gate voltages from 0.2 to 1.4 V [16,34,35].…”

“…The data in Fig. 2(b) (V g = 0.55 and 1.15 V) represent the cases where the primary h/2e oscillation is diminished substantially by carefully adjusting the SOI strength by the gate voltage [11,12,16]. This adjustment was made by setting the spin precessions per side of polygonal closed loops to some special angles [41].…”

“…In this study, we revisit our original idea that the electron properties including the Rashba SOI can be determined using the shape and size information of a nanostructure as a measurement reference [11]. We fabricated mesoscopic ring array (MRA) structures using the same pristine QW wafers as those used in the previous studies [12,14], but paying a particular attention to the MRA shape so that the resultant electron orbitals become chaotic [11,12,16]. The trigger of this work was really an unexpected observation that the amplitude of the Altshuler-Aronov-Spivak-(AAS-) type oscillations in our MRA was diminished when the temperature was lowered from ∼1 K to the base temperature of the dilution refrigerator (∼15 mK) in our preliminary experiment.…”

“…It is found that the semiclassical billiard ball simulations can reproduce the measured MC oscillations of the MRA very accurately, including their fine structure and background components. Our simulations account for the time-reversal interference on all return trajectories in the given system [16,23,24], assuming the spin precessions caused by the Rashba and residual SOIs [14,28,29]. The strengths of the SOIs are parametrized by the spin precession angles per unit length, which are twice the SOI-related wave numbers k α ≡ m * ᾱ h 2 (m * andh being the effective mass and Planck constant h divided by 2π , respectively) and k c for the Rashba and residual SOIs, respectively, where the explicit correspondence between the SOIs and actual motions of spins was maintained.…”