Quantum spin Hall effect spintronics

In this project we shall investigate the potential for spintronics of the quantum spin Hall (QSH) regime in hybrid nanostructures made by attaching ferromagnetic metal contacts to the edge states of two-dimensional topological insulators. These 2D materials will be formed from semiconducting InAs/GaSb coupled quantum wells. Being able to harness the spin-momentum-locked helical edge states in the QSH regime will have the potential for realising dramatic reductions in the power consumption of classical ICT hardware, and in the longer term offer the prospect of being useful for topological quantum computing.

To build such spintronic devices, we need to know the conditions under which current flows through their edge states. We need to know the spin polarisation of a current injected from a ferromagnet into the QSH edge state, and which ferromagnetic contact material provides the largest spin-polarisation. We need to know how efficiently spins can be injected and detected in these QSH edge channels using ferromagnetic metal contacts. We also need to know over what distance spin information can propagate in the QSH edge states, and in what circumstances this distance is the longest.

The project is a collaboration between the School of Physics and Astronomy, who have expertise in spintronics and the study of devices incorporating ferromagnetic materials, as well as topological materials, and the School of Electronic and Electrical Engineering, who are capable of growing ultra-high quality InAs/GaSb coupled quantum wells in their III-V semiconductor molecular beam epitaxy system. We will begin by constructing contacted InAs/GaSb mesas with top and bottom gates that allow them to be tuned into a charge-neutral and non-trivial regime, which are the correct conditions for current to flow only in the edge states. We will attach normal drain contacts on either side of a ferromagnetic source contact on a InAs/GaSb mesa and measure the drain currents from left- and right-flowing edge states in the non-trivial edge state regime; the spin-momentum locking in the QSH edge states will mean that these spatially separated currents directly correspond to the spin-resolved currents, allowing a direct measurement of the spin-polarisation of the current injected from the ferromagnet. We shall try different ferromagnetic metals to determine which one works best. We will then study the flow of a current in a QSH edge state between two closely-spaced ferromagnetic contacts, which is expected to be larger when the current flow direction is spin-momentum locked to the majority spin direction of the contacts; reversing the magnetisation direction in the contacts will invert this diode-like behaviour. The difference between forward and reverse currents will tell us the efficiency of the spin injection and detection. Moving the contacts apart will allow us to determine the length over which spins can flow coherently within the edge states by measuring the decline in difference between forward and reverse currents with spacing; we shall study this as a function of temperature in order to determine the physical mechanisms causing the loss of spin coherence.

The results we shall obtain will not only lead to high impact publications and conference presentations by shedding light on the possibilities offered by this novel combination of materials, but also develop valuable know-how in the field of quantum spin Hall spintronics for technological applications.

Our project is inspired by the increasingly urgent need to find new ways of processing information using vastly less energy than at present. It is now over a decade since the carbon footprint of the internet exceeded that of commercial air travel, and ICT consumes almost 2,000 TWh per year, using 10% of global electricity. These figures are expected to rise to 8,000 TWh and 21 % by 2030. Against this background of constantly rising demand for computing power, performance per watt in processors is beginning to tail off in conventional CMOS architectures and is expected to hit physical limits within the coming decades.

In order to circumvent these problems, entirely new device architectures are needed. Spintronics is already a commercial success story that has enabled the huge amounts of extremely cheap data storage needed to provide social media, such as Facebook, Instagram, and Youtube, free to users. Spintronics researchers are now turning their attention from data storage to data processing, where an important goal is to use spin-based phenomena to more efficiently process information.

One way to do so is to combine storage and memory, as is done in the recently announced MESO concept from Intel. There, the flow of currents is controlled by the state of bistable nanomagnets which store binary bits in a non-volatile manner. The structures we propose here share this feature, where bistable nanomagnets form contacts to the helical edge states of a quantum spin Hall material and control the current flow within it by their magnetic state. Furthermore, whether or not this happens can be controlled by electrical gates. This presents a new set of spintronic phenomena ready to be exploited for logic-in-memory architectures.

It is also possible to look beyond von Neumann machines executing Boolean logic towards quantum computing, motivated as much by new functionality as by energy efficiency. Topological quantum computing has been proposed as a method for overcoming the decoherence problem that plagues conventional qubits, but hardware implementations are as-yet non-existent. This approach requires the use of quasiparticles known as anyons that obey neither Bose-Einstein nor Fermi-Dirac quantum statistics. Majorana zero modes can act as anyons in many respects, and have been shown to exist in topological insulators in proximity contact with superconductors, including the helical edge states of InAs/GaSb coupled quantum wells. The use of the InAs/GaSb system is particularly noteworthy since it is a conventional III-V semiconductor heterostructure system, and so already looks familiar to the microelectronics industry. It does not present the special fabrication and processing challenges that arise due to the presence of exotic elements in many other topological materials. Adding spintronic functionality offers new dimensions to this form of topological qubit.

Our programme of work offers the possibility of developing a new family of spintronic devices and we will liaise directly with our project partner, QinetiQ, as well as through industrial partnerships developed through the Royce institute to explore pathways to impact for the advances that we shall make.