Prof. Alex Khaetskii (Dept. of Physics, SUNY, Buffalo, NY, USA)
Title: " Intrinsic Mechanism of Spin Hall effect. Myth and Reality."
Time: Tuesday, January 17, 12:00
Place: Seminar room of Theoretical Physics Dept.
It is quite amazing that despite intensive theoretical study, the physics of the edge spin density accumulation (or spin Hall effect) for the intrinsic mechanism has not been understood until recently. We recall that the new boom about spin currents happened 10 years ago exactly for the reason that the intrinsic mechanism (i.e. related to the spin-orbit-related splitting of electron spectrum) became the central issue. The recent reviews give to a reader a totally wrong impression about the status of the topic in the case of semiconductors. The intrinsic mechanism is mostly pronounced in the limit of large mean free path, and two sources of edge spin accumulation are possible, the bulk one due to the spin current from the bulk, and generation of the spin density upon the boundary scattering itself. The relative role of those mechanisms has been studied in Ref.  with the conclusion which is exactly opposite to the one reached by the previous researchers. It was analytically proved that the main contribution to the edge spin density is due to the boundary scattering itself, and not due to the spin flux from the bulk. It has direct relation to experiments [2,3] with 2D holes. In addition, we have studied  the edge spin accumulation due to boundary scattering in a high mobility 2D electron gas formed in a symmetric well with two subbands. This study is strongly motivated by the recent experiment of Hernandez et al.  who demonstrated the spin accumulation near the edges of a bilayer symmetric GaAs structure in contrast to no effect in a single-layer configuration. The intrinsic mechanism of the spin-orbit interaction we consider arises from the coupling between two-subband states of opposite parities. We show that the presence of a gap in the system, i.e., the energy separation between the two subband bottoms, changes drastically the picture of the edge spin accumulation. As a result, a parametrically large magnitude of the edge spin density for a two-subband well as compared to the usual single-subband structure is found. By changing the gap from zero up to 1-2 K only, the magnitude of the effect changes by three orders of magnitude. This opens up the possibility for the design of new spintronic devices.
1. A. Khaetskii,, Phys. Rev. B 89, 195408 (2014).
2. J. Wunderlich, B. Kaestner, J. Sinova, and T. Jungwirth, Phys. Rev. Lett. 94, 047204 (2005).
3. K. Nomura, J. Wunderlich, J. Sinova et al., Phys. Rev. B72, 245330 (2005).
4. F. Hernandez, L.M. Nunes, G.M. Gusev, and A.K. Bakarov, Phys. Rev. B 88, 161305(R) (2013).
5. A. Khaetskii and J. Carlos Egues, ArXiv:1602.00026 (2016).