A trilayer stacked tunneling medium enhances spin injection into graphene

Spintronics is a field that adds an additional degree-of-freedom to conventional currents by utilizing the quantum spin state of electrons to store information. In this way, using spin currents that do not require moving electrons can improve conventional transistors. This can give rise to smaller, more sensitive devices that can operate at lower powers. However, work is still needed to efficiently inject, transport, and control the spin states through a sensing medium and realize the commercialization of spintronic transistors.

Leutenantsmeyer et al. observed promising results for enhancing the spin injection into the highly spin conductive material of graphene. Previous work showed that electron spin states can be efficiently transported through graphene, but that injection can be relatively difficult to perform efficiently due to the low spin polarization from the materials used for the electron contacts.

The authors used a pair cobalt electrodes, which typically yield a spin injection polarization of a few percent, but separated from the graphene with a trilayer of hexagonal boron nitride that served as a tunneling medium for the spin controlled electrons.

Applying a DC current through the graphene over a pair of these contacts allows tuning the DC spin polarization to be above 50 percent, which is a 30 percent improvement over previous work. This improvement directly enables larger spin injection efficiencies into graphene than observed before.

This research shows that the boron nitride trilayer is a significant improvement over previous attempts and represent a promising leap towards application in graphene spintronic devices. It also provides a deeper insight into the bias dependence of spin injection polarization.

Source: “Efficient spin injection into graphene through trilayer hBN tunnel barriers,” by Johannes Christian Leutenantsmeyer, Josep Ingla-Aynés, Mallikarjurna Gurram, and Bart J. van Wees, Journal of Applied Physics (2018). The article can be accessed at https://doi.org/10.1063/1.5050874 .