The unique structural and electronic properties of graphene have great potential for the development of opto-electronic applications, although significant further studies and optimization are required. In such studies, to minimize parasitic doping effects caused by uncontrolled material adsorption, graphene is often investigated under vacuum. However, we have shown that an entirely unexpected strong n-doping of graphene due to chemical species generated by common ion high-vacuum gauges occurs in such systems. The effect – reversible upon re- exposing the graphene to air – is significant, as doping rates can largely exceed 1012 /cm·hour, depending on pressure and the relative position of the gauge and the graphene device.
(a) Large-area CVD graphene sample contacted by Au/Ti electrodes evaporated through a shadow mask, with few PMMA residues from the transfer process visible as bright features near the edges. (b) Raman spectrum demonstrating the monolayer character of our CVD-grown samples. (c) AFM image of a sample transferred onto a Si/SiO2 substrate, with a roughness of ~6 Å. Small wrinkles a few nanometers high (white lines) and nanoparticle residues from the transfer process (white spots) are also visible.
Conductivity of the graphene device in an ultrahigh vacuum environment is strongly influenced by species emitted by the pressure-monitoring ion gauge, especially at higher pressures during chamber pump-down. Main panel: the thick solid lines—from right (black) to left (light blue)—show the conductivity σ versus gate voltage VG, with each curve measured after having exposed the graphene device to the ion gauge for an increasingly long period of time. The shift of the charge neutrality point into negative VG values indicates the occurrence of n-doping. The thin dashed lines (from left to right) illustrate the ‘recovery’ of the sample upon exposure to air, showing the reversibility of the effect. Inset: similar measurements performed on an exfoliated graphene device show qualitatively similar trends.