skip to content

Quantum materials and devices group

 

Conventional superconductivity is mediated by singlet Cooper pairs of electrons which have antiparallel spins. The proximity effect between a superconductor (S) and a ferromagnet (F) tends to strongly suppress superconductivity because the internal exchange field of the latter acts to align electron spins and so destroys pairs. Consequently the critical temperature (Tc) of a thin superconductor is strongly suppressed by contact with a ferromagnet, and in an F/S/F trilayer (the equivalent of the spintronic spin valve) the Tc suppression can be controlled by summing or subtracting the exchange fields of the magnetic layers via the relative alignment of their moments. 

Similarly, the Josephson critical current (Ic) between superconductors through a ferromagnetic barrier is much lower than through the equivalent thickness non-magnetic (N) barrier. Nevertheless, in the thickness range in which it is still finite, Ic can be shown to oscillate in sign owing to a distance-dependent phase-shift between the electrons in the pair. This can lead to a ground state Pi-phase difference across an S/F/S junction which has direct application as a phase battery in quantum circuits. By creating a barrier with two independent magnetic layers it is possible to magnetically-switch between 0 and Pi ground states.

Although the pair condensate in a singlet superconductor has zero net spin, this is not necessarily true for the unpaired electrons which form the quasiparticle excitations. Indeed, there are circumstances in which the quasiparticle spin-decay length in the superconducting state is much longer than in the normal state and a very large effective spin polarisation can be induced even by unpolarised current injection, and quasiparticle spin currents can be detected via the inverse spin Hall effect. However, in the superconducting state any quasiparticle spin currents must be diffusive and independent of the (zero-spin) charge supercurrent meaning that many of the familiar concepts of conventional spintronics such as giant magnetoresistance do not have a direct quasiparticle spin equivalent. 

The final, and potentially most important, ingredient is that an inhomogeneous magnetic interface can result in an unconventional S/F proximity effect mediated by triplet rather than singlet pairs. Consequently a triplet Josephson current can propagate through much thicker ferromagnet barriers than the equivalent singlet current and also carries spin, meaning that control of the charge supercurrent necessarily enables control of spin currents in the superconducting state.  

Spintronics emerged from the discovery of giant magnetoresistance (GMR) in the 1980s. It has rapidly achieved enormous technological success in the data storage field: initially as a means of improving magnetic field sensors for reading data from hard discs and more recently as the storage node in magnetic random access memory (MRAM). More broadly, spintronics has been promoted as an eventual low-power replacement for charge-based semiconductor (CMOS) logic in which information is carried by spin currents and controlled and sensed by magnetic elements within a circuit.

Combining superconductivity with spintronics brings in phenomena which do not exist in the normal state, such as quantum coherence and spin polarised supercurrents, thus enabling much lower energy spin-transfer and magnetic switching.The core vision of this proposal is to create a paradigm shift in spintronics by combining some or all of the S/F ingredients outlined above to enable the generation of spin currents with minimal ohmic losses, their control via a combination of magnetic state and superconducting coherence, and for the spin transfer to drive information storage and flow in a circuit. 

We are conducting experimental investigations into exploiting the coupling of superconductivity and magnetism to create novel device functions. This will progressively focus towards elements of future technologies. The eventual ambition is to create simple superconducting spintronic demonstrator devices which combine addressable memory and logic functions and hence ignite a technology field.