Spontaneous spin polarization in GaAs/AlGaAs split-gate heterostructures
Introduction
Spintronics is a new branch of electronics which involves the active control and manipulation of spin degrees of freedom in solid-state devices. Spin transport differs from charge transport as spin is a non-conserved quantity in solids due to spin–orbit and hyperfine coupling. With rapid scaling of CMOS technology the physical gate lengths of transistors have reached atomic length scales where short channel effects degrade the device performance drastically. The alternative technologies, like spintronics, can complement the existing scaled CMOS technology to extend Moore's law even further to dimensions as small as 8 nm. Extensive research has been going on in the spintronics field to overcome a variety of challenges posed in the form of efficient injection, transport and detection of spin polarized carriers.
Various conductance measurements have been performed on quantum point contacts (QPC) formed by the lateral confinement of a high mobility two-dimensional electron gas (2DEG) in split-gate GaAs/AlGaAs heterostructures. Conductance quantization [1] is observed in these mesoscopic devices, and in addition to the integer multiples of 2e2/h, some experimental groups have observed an extra feature at G∼0.7 (2e2/h) [2]. This feature at 0.7 G0 has been referred to as ‘0.7 conduction anomaly’ or simply ‘0.7 structure’. This feature has also been shown to vary between ∼0.5 and 0.7 depending on various parameters like the surface gate geometry, length of the channel and also electron density [3].
The above experimental observations initiated a number of theoretical efforts to understand and explain the 0.7 anomaly [4], [5]. The most accepted idea is to associate this anomaly with the electron–electron interaction inducing an onset of spontaneous spin polarization in the QPC. Many modeling attempts carried out along these lines have considered simple analytical model potentials to include the spin density functional formalism of Kohn and Sham in the local density approximation (LDA) [6]. The local exchange potential induces spontaneous local magnetization and a spin splitting of the sub-bands. The exchange interaction is, hence, seen to be the dominant mechanism towards the spin splitting.
In this paper, we try to extend the same modeling approach to real potentials obtained from a self-consistent 3D Poisson–1D Schrödinger problem. The LDA approximation in the Kohn–Sham density functional formalism is used to calculate the total effective potentials for the spin species. The device modeled is a GaAs/Al0.24Ga0.76As modulation doped heterostructure having a 35 nm GaAs quantum well. A schematic view of the device structure simulated in this work is shown in Fig. 1 where Nd1 and Nd2 are the two delta doped layers.
Section snippets
Theoretical modeling
The simulation procedure, used to obtain the self-consistent potential, is comprised of two parts. The first part involves the calculation of the self-consistent Hartree potential by using the in-house 3D Poisson–1D Schrödinger solver. The confinement in the growth direction (y direction) is found to be much stronger than the confinement in the z direction and we can assume, with no approximations involved, that only the first sub-band related to the y direction is occupied for low sheet
Simulation results
The important parameters required for the self-consistent calculation of the total effective potential—Schottky barrier potential and donor ionization energy—are validated with experimental data by performing the simulations on a Hall bar structure (see Fig. 2).
The simulated sheet densities are in good accordance with the experiment falling within the 7% error margins in the experiment. The spontaneous spin polarization in the QPC, given by p(x, z), for a voltage of −4 V on the split gates is
Conclusions
In summary, we have found evidence of spontaneous spin polarization in realistic QPC devices by developing a self-consistent simulation scheme using Kohn–Sham spin density functional formalism. The exchange potential is the dominant driver behind the polarization. These calculations have encouraged us to further investigate the dependence of the 0.7 anomaly on electron density, length of the channel region and surface gate geometry. In a future effort, we would like to build a Monte Carlo based
References (11)
- et al.
J. Phys. C
(1988) - et al.
Phys. Rev. Lett.
((1996))The 0.7 conductance feature: a mesoscopic mystery with important implications
(2003) - et al.
Phys. Rev. B
(2000) - et al.
Phys. Rev. B
(1996)et al.Phys. Rev. B
(1998) - et al.
Phys. Rev. B
(2001)
Cited by (1)
Computational electronics: Semiclassical and quantum device modeling and simulation
2017, Computational Electronics: Semiclassical and Quantum Device Modeling and Simulation