Better understanding of spin transfer to non-conducting magnets

Spintronics promises more efficient electronics by using the spin of electrons. But spins can also be carried by magnons, magnetic quasiparticles that can travel through electric insulators. The transfer of spin information from electrons to magnons is important to the future of spintronics. A new study by scientists from the Universities of Groningen and Manchester, which was published in Applied Physics Letters on 21 February 2019, shows how best to study this transfer.

Spin is a quantum-mechanical property of electrons that can best be imagined as electrons spinning around their own axis, causing them to behave like small magnets. This spin can have two directions – usually designated as ‘up’ and ‘down’ – which can be used to store or transmit information.

Spin is not limited to electrons. Magnons can be created in non-conducting magnetic materials. ‘These are a kind of spin-wave’, says Ivan Vera Marun, a lecturer at the University of Manchester. Magnons travel like a Mexican wave through a stadium: the electrons in the material don’t move, but their spin orientation does. Vera Marun published the Applied Physics Letters paper together with his former supervisor Bart van Wees from the University of Groningen, with whom he often still works. They also both supervised the experimental effort led by Kumar Das as part of his doctoral work and Fasil Dejene, who also obtained their PhDs at the Van Wees lab.

Transistor

‘There are several advantages to using magnons in spintronics’, explains Vera Marun. ‘They can travel through magnetic insulators, which means we can use different materials from those suited to electron spin transport. Furthermore, the spin of magnons can travel longer distances and be influenced by a magnetic field, which opens up the possibility of magnon-based logic.’ Scientists from the Van Wees lab recently reported on a magnon-transistor that they had built, and two other labs produced similar results.

However, to use magnons in spintronic devices, you have to convert spins carried by electrons into magnons – and vice versa. This requires interfaces of different materials, and their properties have to be optimized for the conversion process. In their new study, Vera Marun and his colleagues analysed this conversion interface in a new way.

‘In our device, we injected spins using aluminium as the transport channel’, says Vera Marun. This metal is not magnetic, and it has low spin-orbit coupling, which means the atoms barely interfere with the spin direction. ‘Previous studies have used heavy metals, which have high spin-orbit coupling, or ferromagnets to inject spins into the non-conducting magnet. However, these materials could affect the magnetic properties of the latter.’

Temperature

With their device, the scientists studied the temperature dependence of the spin conversion. ‘This had never been done before; other studies used just one temperature.’ They measured what is known as the effective spin-mixing conductance (which represents spin conversion efficiency) over a range of 4.2 Kelvin to room temperature (293 K), and this neatly confirmed their predictions, thus validating the theory. ‘We therefore dispelled some of the controversy surrounding measurements made when ferromagnets or heavy metals were used to inject the spins’, says Vera Marun.

A second representation of spin conversion is the ‘real spin-mixing conductance’. This is often measured in a different experiment, but it can also be calculated from the effective spin-mixing conductance. ‘We were able to both directly measure the real spin-mixing conductance and calculate it from the effective one, in one device. This allowed us to compare direct measurements and indirect calculations, and we found a significant disagreement. Our analysis suggests that it is sometimes more reliable to calculate the real spin-mixing conductance from the effective spin-mixing conductance.’

The new study provides valuable data on how to analyse the efficiency of the conversion of electron spins into magnons in different devices and predict its operation at other temperatures. ‘Our results provide information on how to interpret previous experiments’, says Vera Marun. ‘And they will assist the development of new, more efficient interfaces for the injection of spins into non-conducting magnets.’