B. Thorpe, S. Schirmer, K. Kalna, F. C. Langbein. Monte Carlo Simulations of Spin Transport in an InGaAs Field Effect Transistor. 34th International Conference on the Physics of Semiconductors (ICPS2018), Poster P3_176, 29th July to 3rd August 2018.
Electron spin in semiconductor devices can enable novel devices with a wide variety of potential applications like spin field effect transistors (SpinFETs) considered as a future candidate for high-performance computing and memory applications with ultra-low power consumption.
Here, we apply finite-element quantum-corrected ensemble Monte Carlo self-consistent device simulations with electron spin to a nanoscale III-V field effect transistor to investigate a spin transport. The simulations include spin as a separate degree of freedom via a spin density matrix. The spin-orbit interaction assumes the Dakov-Perrel mechanism with two terms: the Dresselhaus Hamiltonian which accounts for spin-orbit coupling, and the Rashba Hamiltonian which accounts for spin-orbit coupling due to structural inversion asymmetry.
We then investigate the spin dynamics across the channel of a 25nm gate length In0.3Ga0.7As MOSFET in which the source electrode was assumed to be ferromagnetic. This was used as means to inject polarized spin current into the device channel. We then varied the initial injection direction, the drain and gate biases and applied mechanical strain. The simulation results are interesting because they suggest that the polarisation of the electrons initially decays as the electrons traverse the channel, as expected. From a length of 0.8 at the source to a minimum of 0.24 at approximately 10nm past the gate. However the magnetisation then partially recovers to a a final length of 0.44 as the electrons approach the drain. Since the drain electrode is non-magnetic, the recovery of the magnetization cannot be attributed to existing polarized carriers inside the drain but must be assumed to be due to partial re-phasing of electron spins resulting in a net magnetization.
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.