Bias Dependent Variability of Low Frequency Noise in Single Layer Graphene FETs

TL;DR

This study investigates the bias-dependent variability of low-frequency noise in single-layer graphene FETs, providing a physics-based model validated by experiments, revealing the physical mechanisms behind noise fluctuations.

Contribution

It introduces the first comprehensive statistical analysis and analytical model of LFN variability in GFETs, linking physical mechanisms to noise deviations under various bias conditions.

Findings

LFN variability is driven by carrier number fluctuations and mobility fluctuations.

LFN deviations follow an M-shape versus gate bias, with a minimum at the charge neutrality point.

Trap statistics in GFETs differ from classical Poisson distribution, likely due to electrolyte interface effects.

Abstract

Low-frequency noise (LFN) variability in graphene transistors (GFETs) is for the first time researched in this work. LFN from an adequate statistical sample of long-channel solution-gated single-layer GFETs is measured in a wide range of operating conditions while a physics-based analytical model is derived that accounts for the bias dependence of LFN variance with remarkable performance. It is theoretically proved and experimentally validated that LFN deviations in GFETs stem from physical mechanisms that generate LFN. Thus, carrier number DN due to trapping/detrapping process and mobility fluctuations Dm which are the main causes of LFN, define its variability likewise as its mean value. DN accounts for an M-shape of normalized LFN variance versus gate bias with a minimum at the charge neutrality point (CNP) as it was the case for normalized LFN mean value while Dm contributes only near the CNP for both variance and mean value. Trap statistical nature is experimentally shown to differ from classical Poisson distribution at silicon-oxide devices, and this is probably caused by electrolyte interface in GFETs under study. Overall, GFET technology development is still in a premature stage which might cause pivotal inconsistencies affecting the scaling laws in GFETs of the same process.