Spectroscopy of Rashba spin splitting in InSb quantum wells

Abstract

We use electron spin resonance in the far infrared to probe the Landau-level spin splitting in symmetric and asymmetric InSb quantum wells. The asymmetric wells exhibit a strong deviation in behavior from the symmetric wells at low magnetic fields with apparent g factors far in excess of the bulk g factor of InSb. These asymmetry-induced shifts in the spin resonance depend on Landau-level The study of electron spin phenomena in semiconductor heterostructures has intensified dramatically in recent years. The field, "spintronics" has emerged with the purpose of developing devices combining the charge and spin degrees of freedom. In particular, spin splitting caused by bulk inversion asymmetry and structural inversion asymmetry (SIAoften called Rashba splitting) has attracted much attention. Understanding the spin-orbit interaction caused by SIA, which is predicted to lead to spin splitting even without an applied magnetic field, 1,2 is important for developing spinbased devices. To date, experimental evidence for Rashba splitting in two-dimensional electron systems (2DESs) has been chiefly confined to the observed beating of Shubnikov-de Haas (SdH) oscillations 3,4 that is presumed to arise from different populations of electron spin orientations. This beating has been interpreted in a number of heterostructures [3][4][5][6][7][8][9][10][11][12][13][14][15][16] to deduce the Rashba parameter ␣, which is expected to depend on materials parameters and the electric field. The zero-field spin splitting has also been inferred from Raman-scattering spectra 17 in GaAs and weak anti-localization measurements in InGaAs quantum wells 18 . In this paper, we report on electron spin resonance (ESR) experiments on 2DESs in InSb quantum wells. By observing transitions between states with the same Landau level index but opposite spin orientations, we directly measure the spin splitting as a function of applied magnetic field B. In addition to the B = 0 spin splitting, the Rashba model 1 predicts considerable modification of the Landau-level energy structure at nonzero B for asymmetric heterostructures, which has not been previously confirmed. Because of the large effective g factor and large predicted Rashba effect, 6 the spin splitting in InSb-based 2DESs should occur at far infrared frequencies for B Ͻ 10 T. Our experiments demonstrate that the Rashba effect is relatively strong in InSb quantum wells and that ESR is a powerful technique for studying the Rashba effect.Our samples are InSb single quantum wells of widths 30 and 20 nm with Al 0.09 In 0.91 Sb barrier layers that are ␦ doped with Si. The Si ␦ layers are located either singly on one side of the quantum well (asymmetric samples) or equidistant on both sides of the quantum well (symmetric samples). The ␦-doped layers within the barrier layers are typically located 70 nm from the well center. The barrier layers on the substrate and surface sides of the wells are 3 m and 160 nm thick, respectively. Silicon dopants are placed near the surface to limit surface depletion. ...

Zero-field spin splitting and spin-dependent broadening in high-mobilityInSb/In1−xAlxSbasymmetric quantum well heterostruct

“…Such inhomogeneities may result from spatial variations in well width and/or interface roughness in the sample 20 and perhaps reflects the difficulty in the growth of high-quality InSb heterostructures on highly mismatched GaAs substrates. However, it is interesting to note that estimating ⌫ from data taken from a similar InSb QW sample 17 grown from a different MBE source appears to show similar levels of broadening to those found here. We point out that this conjecture is clearly not universal for remote doped heterostructures, e.g., a narrow broadening of ϳ0.6 meV was found in a similar narrow gap InGaSb/AlInSb structure with a large 50 nm spacer layer in Ref.…”

“…Dedigama et al recently reported the first observations of WAL in a ͑low mobility͒ InSb QW supporting the presence of large SO coupling. 16 In an alternative approach, Khodaparast et al 17 extrapolated a spin splitting to zero field in an asymmetric 30 nm InSb/InAlSb QW from electron-spinresonance measurements. Assuming the Rashba interaction to be dominant, ␣ R was determined to be as high as 1.3 ϫ 10 −11 eV m. This is larger than recent theoretical calculations in InSb/InAlSb QWs estimating ␣ R to be in the range of 2-7ϫ 10 −12 eV m. 18 In this work it was also demonstrated that the bulk inversion asymmetry ͑BIA͒ contribution to spin splitting can be of significant and comparable value to that of the Rashba depending on the specific details of the heterostructure.…”

Experimental determination of the Rashba coefficient in InSb/InAlSb quantum wells at zero magnetic field and elevated temperatures

“…The common techniques for measuring it require either extrapolation from high magnetic fields [11,12] or the fitting of quantum interference corrections to the magnetoconductivity at low fields [13]. Extrapolation from high magnetic fields is non-trivial, not-only because the Zeeman effect obscures the Rashba contribution to the splitting, but also because the latter is rapidly quenched by the field [14].…”

Interplay of spin and orbital magnetogyrotropic photogalvanic effects in InSb/(Al,In)Sb quantum well structures

“…4 This novel material is the subject of numerous experimental studies of transport, optical, magneto-optical, and spin-related phenomena. [1][2][3][4][5][6][7][8][9][10][11][12][13][14] The characteristics driving the interest in this novel narrow gap material are the high carrier mobility, small effective masses, large Landé g * factor, possibility of the mesoscopic spindependent ballistic transport, and a strong spin-orbit coupling. The latter gives rise to a number of optoelectronic effects such as, e.g., terahertz photoconductivity 15 and the circular photogalvanic effect [16][17][18][19][20][21][22] recently observed in InSb QWs.…”

“…Indium-antimonide-based quantum wells (QWs) have attracted growing attention for high-speed transistors, 1 quantum computing, 2,3 and infrared lasers. 4 This novel material is the subject of numerous experimental studies of transport, optical, magneto-optical, and spin-related phenomena.…”