In electronics, a drift current is one where the motion of charge carriers is dictated by an applied electric field. In contrast, a diffusive current is one where their motion is due to diffusion associated with the slope of the carrier concentration in the absence of an electric field. The field of spintronics utilises not only the charge degree of freedom of carriers, but also their spin degree of freedom. It is therefore interesting and informative to explore how electrical signals generated from spin-polarised carries vary between the drift and diffusive current regimes. This is the focus of this week’s Pick of the Week, “Efficient conversion of light to charge and spin in Hall bar microdevices” by Nadvornik and Haigh et al. in Phys Rev B.
Seeing the Hall picture
Over the past 18 years there have been several papers that have studied Hall voltages associated with the ordinary Hall effect (OHE) in drift and diffusive regimes. In particular, some of these studies considered the geometry of the Hall cross device as well as the spatial variation of the applied magnetic field, which was achieved by using a magnetic force microscopy (MFM) tip. The work of Nadvornik and Haigh is an opto-spintronic equivalent of such OHE/MFM experiments, as spin currents are generated optically via the absorption of circularly polarised light, and Hall voltages are generated via the inverse spin-Hall effect (ISHE), a spin-orbit coupling phenomenon.
It has previously been demonstrated that the ISHE can be used to produce a voltage that depends on the degree of circular polarisation of incident light. This principle allows for electrical polarimeters that are scaleable, work at room temperature, and do not need mechanical components in their operation. The main result of the work of Nadvornik and Haigh is to show that the electrical signal of such polarimeters can be greatly enhanced when accessed in the drift regime.
Shedding light on the situation
Measurements are made on micron-sized Hall cross style devices of a GaAs/AlGaAs sample wafer. Electrical current flows largely through the GaAs layer as a result of localised doping of the AlGaAs layer and electron migration. Using a fairly simple optical set up, the intensity of an incident light beam on the Hall cross device is varied at a frequency f1, while its circular polarisation is varied at a frequency f2. By making electrical voltage detection measurements at lock-in frequencies f1 and f2, the resultant Hall voltage across the semiconductor device can be separated into components related to the intensity and degree of circular polarisation of the incident light. The device geometry also allows for a bias to be applied orthogonally in-plane to the Hall detection probes, to allow the drift regime to be accessed.
The electrical current that arises when shining the laser onto the device can be considered to be made up from four different contributions – one relating to charge diffusion, one relating to charge drift, one relating to the ISHE (which is appropriate for a spin-current when a charged diffusive current is absent), and one relating to the anomalous Hall effect for a polarised charge diffusive current. If the incident light is of a wavelength the same as the GaAs band gap, spin-polarised electron-hole pairs are generated, and so the Hall cross device hosts both non-equilibrium charge concentrations and spin polarisations.
Do you catch my drift?
Measurements are first presented for the diffusive regime. The voltage sensed at the transverse Hall cross probes at the f2 lock-in frequency is associated with the ISHE, and in the diffusive regime this has an expected variation with the spatial variation of the laser spot (and hence spin current). An electrical bias is then applied across the longitudinal channel of the Hall cross device to access the drift regime, and it is found that the Hall voltage increases by more than two orders of magnitude. This shows that the Hall voltage can be tuned by tailoring the type of regime of the charge current flow. Other ways of controlling the balance of the drift/diffusive regimes include varying the laser power, the size of the applied bias, and the relative size of the laser spot to the Hall cross channel. Indeed, for a laser spot larger than the channel width in the drift regime, the Hall cross acts as a point detector. One interesting feature that the spatial resolution of the experiment reveals is that the transverse voltage of the device goes to zero in the diffusive regime when the spot is centrally located between the probes, whereas it remains finite for this situation in the drift regime. This could be a ‘sweet spot’ so to speak if this experiment were to form the basis of an actual spintronic device.
It is not often that anything optical gets featured in Pick of the Week, but as this paper demonstrates, optical techniques can be a useful tool to integrate into spintronic experiments. For example, the OHE/MFM experiments mentioned before suffer from additional signals that arise from electrostatic coupling of the MFM tips to the semiconductor channel, and these issues are not present in the Nadvornik and Haigh experiment. The importance of charge-based applications using semiconductor materials (e.g. the transistor) goes without saying, as is true of optical- based applications (e.g. LEDs). Experiments such as that of Nadvornik and Haigh that add spin into the charge and optical semiconductor toolkit are vital to building a greater knowledge of the fundamental link between these semiconductor techniques, and thus may aid in the design of potential future opto-spintronic devices.
Nádvorník, L., Haigh, J., Olejník, K., Irvine, A., Novák, V., Jungwirth, T., & Wunderlich, J. (2015). Efficient conversion of light to charge and spin in Hall-bar microdevices Physical Review B, 91 (12) DOI: 10.1103/PhysRevB.91.125205