Mobility Degradation and Series Resistance in Graphene Field-Effect Transistors

Abstract

Accurate device models and parameter extraction methods are of utmost importance for characterizing graphene field-effect transistors (GFETs) and for predicting their performance in circuit applications. For DC characterization, accurate extraction of mobility and series resistance is of particular concern. In this paper, we show how a first-order mobility degradation model can be used to separate information about mobility degradation and series resistance for a set of GFETs of different channel lengths. Data from a large set of top-gated GFETs based on chemical vapor deposited (CVD) graphene was analyzed to validate the proposed model and extraction procedures. For removing any uncertainties caused by observed device-to-device data variations due to the uneven quality of CVD graphene, the same methods were applied to a set of closely located bottom-gated GFETs found in literature. Those GFETs were designed for transfer length methods and fabricated on exfoliated graphene of homogenous quality. Similar mobility degradation behavior was observed for both sets of devices with the mobility being reduced to half for a voltage-induced charge carrier density of 10 13 cm -2 .

An Extraction Method for Mobility Degradation and Contact Resistance of Graphene Transistors

“…The drain current I D of GFETs given by a straightforward adaptation of a widely used I-V model [1] considering both an appropiate charge control description and an explicit contribution of the extrinsic mobility degradation coefficient θ [10], [21], [22] is expressed as 1 [5]…”

“…This model considers the impact of vertical fields and impurities on the carrier transport and has enabled to obtain relevant device parameters since its proposal, e.g., a low-field mobility [2]- [4]. Recently, this well-known model has been adapted in order to explicitly consider mobility degradation effects in the performance description of GFET technologies [5]. This novel approach has been motivated by experimental observations of a carrier density-dependent mobility in GFETs under different scenarios [6]- [8].…”

“…From a device characterization point of view, the fitting procedure relying on an initial guess of R c towards obtaining the other three model parameters in [5] might become an important constraint for the characterization of GFETs at a large scale due to the computational burden and challenges related to randomly finding adequate initial values. A straightforward parameter extraction methodology considering experimental data is required to expand and ease the use of this novel modeling-based characterization approach.…”

“…In this work, by following a methodical treatment of experimental data based on the concept of the well-known Y -function, a parameter extraction methodology for the mobility degradation-based transport model is proposed using individual device transfer characteristics of a given GFET technology with devices of different gated-channel lengths. In contrast to the original approach [5], the extraction methodology proposed here is more robust for device characterization purposes since it yields values of three out of four of the model parameters without needing any initial random assumption. The fourth parameter, related to charge-neutrality, is extracted by appropriately adjusting the model to experiments at very low transversal electric fields.…”

“…By considering the characterization of GFET technologies including sub-µm gate-lengths, the model-based characterization is proven here to be for general purposes and not only for long devices as shown in [5]. Further adaptations of the underlying model to include velocity saturation effects -relevant at high lateral fields-are out of the scope of this work, however, the extracted parameters here can be relevant for the extraction of velocity saturation related parameters.…”