Dielectric Surface Charge Engineering for Electrostatic Doping of Graphene

Controlling the doping level in graphene during integration into silicon CMOS compatible devices is an open challenge. In general, the doping level in graphene is influenced via substrate interactions, metal contacts, and encapsulation layers. Here, we demonstrate a method to control the Fermi level in graphene through transfer onto ionic-doped oxide surfaces. The substrates were prepared to this end by diffusion of ammonia and aluminum on the oxide surface, which induces positive (NSiO+) and negative (AlSiO-) charges on the oxide layer. Van der Pauw measurements show that the charge neutrality or Dirac voltage in graphene can be shifted from about -60 V (n = -8.62 x 10(12) cm(-2)) on standard SiO2 to about 13 V (n = 2.17 x 10(12) cm(-2)) on negatively doped SiO2 layers by manipulating the surface charge. Hall measurements show that the electron mobility in graphene transferred on an as-grown oxide surface is higher than for graphene on a doped oxide because of additional scattering centers. Transfer line method measurements show that the contact resistance between graphene and nickel electrodes varies in average from 683.3 Omega.mu m on SiO2 to 1046.6 Omega.mu m on negatively doped SiO2 and that it depends on both the substrate surface charge and on graphene sheet resistance. Ionic-doped oxide surfaces are generally temperature-stable with respect to front- and back-end-of-the-line semiconductor manufacturing. The method presented here allows adjustments of the surface charge density of the substrate, and thus in graphene, which cannot be realized by organic or metallic functionalization. Therefore, the method may be suitable for engineering graphene-based devices and circuits, in particular, for applications that require complementary devices or a specific position of the Fermi level in graphene, for example, to adjust contact resistivity, sheet resistance, or sensor sensitivity.