Graphene Transistors

The recent discovery of graphene (Novoselov et al., 2004), a single atomic sheet of graphite, has ignited intense research activities to explore the electronic properties of this novel twodimensional (2D) electronic system. Charge transport in graphene differs from that in conventional 2D electronic systems as a consequence of the conical energy dispersion relation near the charge neutrality (Dirac) point in the electronic band structure (Zhang et al., 2005). Field-effect mobility as high as 15 000 cm2/V.s and a Fermi velocity of ~108 cm/s have been demonstrated at room temperature (Geim & Novoselov 2007). These properties make graphene a possible candidate for electronic devices in the future. The major benefit of graphene frequently quoted is superior electron/hole mobility compared to other semiconductors, but as Fig. 1 shows, this is not experimentally the case yet when compared to narrow bandgap III-V semiconductors for comparable carrier densities. In addition, the lack of a bandgap limits the usage of two dimensional graphene for digital switching, where high on/off ratios are necessary. However, several potential advantages may be listed: the perfect 2D confinement of carriers, electron/hole symmetry originating from a conical bandstructure, and the possibility of opening bandgaps lithographically by fabricating graphene nanoribbons (GNRs). In Section 2, high field characteristics of 2D exfoliated graphene are reported on both short-channel and longchannel back gated field effect transistors (FETs). We will elaborate on the problem of metal contact formation and high field transport of 2D graphene. The comparison of fabricated devices based on exfoliated, epitaxial and chemical vapour deposited (CVD) graphene will follow. The large area graphene opens the possibility to make 2D graphene devices for RF/analog amplification (Lin et al., 2010). Superior frequency performance of such devices makes it a promising application of 2D graphene FETs. The real advantage of graphene, however, can be highlighted from a novel proposed device architecture that is yet to be demonstrated (Zhang et al., 2008). If GNRs are made by lithographic patterning, and are either chemically or electrostatically doped into GNR p-n junctions, then the planar form yields to a tunnelling field-effect transistor (TFET). Towards the realization of the GNR TFET we designed a single but tuneable junction structure to demonstrate the operational principle. We discuss these GNR p-n junction transistors in Section 3 along with the developed analytical and device models used to explain the measured transistor characteristics.