Magnetoelectric Gates for Graphene Spin Transistors

(See also: http://www.advancedsciencenews.com/magnetoelectric-oxide-gates-enable-graphene-spintronics/)

As an all carbon material, graphene has long been a target for spintronics applications where maximizing spin diffusion lengths is desired. There are many challenges associated with graphene-based spintronics [1] (phonons, contact effects, etc.). One important issue is that, for versatile spintronic applications, one needs not only to transport spin efficiently but also to manipulate it electronically. For this purpose some form of magnetic gate is required.

Years ago, our NC State colleagues led by Prof. Kim proposed [2] a graphene spin field effect transistor (spin-FET) that would use a magnetic insulator for a gate. For this purpose, the magnetization in the gate should also be switchable with an applied voltage. Thus the number of viable materials for a graphene spin FET gate are severely restricted to those that are "magnetoelectric" [3].

The classic magnetoelectric material is Cr2O3, the most stable chromium oxide under many conditions. Our group set out to grow Cr2O3 films on graphitic substrates to determine the materials science challenges associated with this spin-FET strategy. It is not trivial to make a uniform film on graphene or other graphitic carbon. Since these materials are quite inert, it can be favorable for growing films to "ball up" on the substrate and make rough discontinuous films not suitable for device applications.

Our approach was to use the very high instantaneous deposition rates achievable with pulsed laser deposition to create uniform Cr2O3 films on graphene and graphite [4]. Importantly, we used magnetoelectric annealing combined with electric and magnetic force microscopy (Figure 1) to show that the films maintained the crucial continuous electrically controlled magnetization needed for spin-FET gates (notably at room temperature).

Figure 1. A) AFM image of a region that had been written with a programmed potential gradient using a conducting AFM tip; B) EFM image of the same region; C) MFM of the same region; D) Averaged line profiles showing the continuous electric and magnetic gradient in the written region.

Importantly, the electrical and magnetic signals seen in Figure 1 are seen under zero applied field conditions. This shows that the magnetoelectric effect at the surface is different than in the bulk (where magnetization would vanish in zero applied field). Our experiments thus provide further evidence supporting a growing body of work on the unique properties of the Cr2O3 surface [5,6] that may make it better for spin FET gates than might be predicted based on its known bulk properties.

References

[1] Han et al., Nat. Nanotechnology 9, 794 (2014).

[2] Semenov et al., Appl. Phys. Lett. 91, 153105 (2007).

[3] Fiebig, J. Phys. D: Appl. Phys. 38, R123 (2005).

[4] Stuart et al., Phys. Status Solidi RRL, 10, 242 (2016).

[5] Belaschenko Phys. Rev. Lett. 105, 147204 (2010).

[6] He et al., Nature Mater. 9, 579 (2010).