New model brings spintronic transistor a step closer

University of Groningen PhD student Siddhartha Omar has analyzed the noise in spin transport in graphene. He used a theoretical model based on a circuit that he fed with experimental data. He concludes that impurities in graphene are the main hindrance to spin transport. His findings, which were published in the journal Physical Reviews B, pave the way to improving spin transport. This is necessary for building a transistor based on the principle.

Spintronics is an attractive alternative to realizing low-power electronic devices. It is based not on a charge current but on a current of electron spins. The spin is a quantum mechanical property of electrons, which can best be imagined as small spheres spinning around their own axis, causing them to behave like small compass needles. This spin can assume two values: ‘up’ or ‘down’, which makes it suitable, in principle, for storing or transporting information.

Wobble

Graphene, the two-dimensional form of carbon, is an excellent conductor of spins. But it isn’t living up to theoretical predictions, says Omar. He wanted to find out why, so he used a theoretical model to analyze the noise in spin signals. ‘We were the first to measure this noise, earlier this year. I wanted to know what causes it.’ The noise, which could be seen as a ‘wobble’ in the axis of the spinning electron, could flip the spin and thereby dissipate the spin signal.

Omar used a circuit model, originally proposed by 2007 Nobel laureate in physics Albert Fert and further simplified by one of Omar’s co-authors, Ivan Vera-Marun. ‘I adapted this model for my particular purpose.’ The circuit is based on resistors, and allows for simulations with relatively simple equations. ‘The model was fed with data from our experiments, and we could then estimate the contribution of different sources of noise.’

Impurities

His work suggests that impurities in the graphene are the main source of noise. ‘The good news is that we have now gained a better understanding of why the spin current dissipates in graphene’, Omar explains. The bad news is that impurities are an intrinsic property of the graphene Omar and his colleagues in the lab of University of Groningen spintronics researcher Bart van Wees use most of the time. ‘The source of our graphene is natural graphite, so we need to switch to a different source.’

One source they are now investigating is ultraclean synthetic graphene, made in the lab rather than harvested from natural graphite. This should increase the spin signal even further. Van Wees’s lab had already found several ways to improve this signal. One recent discovery was that by injecting spins through a double layer of boron nitride, the signal increased a hundredfold.

‘We are steadily improving our understanding of spin transport, so we can improve it ever further’, Omar continues. The ultimate aim is to use all this knowledge to build a real spin transistor. ‘That would be the first step towards realizing logic switches based on spin transport. This study brings us one step closer to that aim.’

Reference: S. Omar, B.J. van Wees and I.J. Vera-Marun: A two-channel model for spin-relaxation noise. Physics Review B, 2017, DOI 10.1103/PhysRevB.96.235439