Identifying the mechanisms of p-to-n conversion in unipolar graphene field-effect transistors

The mechanisms of p-to-n conversion and vice versa in unipolar graphene field-effect transistors (GFETs) were systematically studied using Raman spectroscopy. Unipolar p-type GFETs are achieved by decorating the graphene surface with a thin layer of titanium (Ti) film, resulting in a Raman D peak. The D peak is observed to recover by annealing the GFET in nitrogen ambient followed by silicon nitride (Si3N4) deposition, suggesting that the Ti adatoms are being partially removed. Furthermore, unipolar n-type GFETs are obtained after the passivation on p-type GFETs. The threshold voltage of the n-type GFET is dependent on the thickness of the Si3N4 layer, which increases as the thickness decreases. A comparison between the Si3N4 and SiO2 passivation layers shows that SiO2 passivation does not convert the GFET into n-type graphene, which identifies the significance of ammonia (NH3) for the formation of the n-type GFETs. This study provides an insight into the mechanism of controlling the conduction behavior of unipolar GFETs. (paper)