Geometrical magnetoresistance effect and mobility in graphene field-effect transistors

TL;DR

This paper introduces a geometrical magnetoresistance method to accurately measure charge carrier mobility in graphene FETs, overcoming limitations of traditional approaches and revealing insights into scattering mechanisms.

Contribution

The study presents a novel, assumption-free technique using gMR to evaluate mobility in GFETs, improving accuracy over conventional methods.

Findings

gMR effect dominates up to ~0.55 T magnetic field

gMR mobility is 2-3 times higher than drain resistance model estimates

Mobility variations linked to Coulomb and phonon scattering mechanisms

Abstract

Further development of the graphene field-effect transistors (GFETs) for high-frequency electronics requires accurate evaluation and study of the mobility of charge carriers in a specific device. Here, we demonstrate that the mobility in the GFETs can be directly characterized and studied using the geometrical magnetoresistance (gMR) effect. The method is free from the limitations of other approaches since it does not require an assumption of the constant mobility and the knowledge of the gate capacitance. Studies of a few sets of GFETs in the wide range of transverse magnetic fields indicate that the gMR effect dominates up to approximately 0.55 T. In higher fields, the physical magnetoresistance effect starts to contribute. The advantages of the gMR approach allowed us to interpret the measured dependencies of mobility on the gate voltage, i.e., carrier concentration, and identify the corresponding scattering mechanisms. In particular, the range of the fairly constant mobility is associated with the dominating Coulomb scattering. The decrease in mobility at higher carrier concentrations is associated with the contribution of the phonon scattering. Analysis shows that the gMR mobility is typically 2-3 times higher than that found via the commonly used drain resistance model. The latter underestimates the mobility since it does not take the interfacial capacitance into account.