On the Optimization of Graphene Liquid-Gate Transistors for Sensing Applications

Abstract

Introduction. Graphene, a carbon single-atom layer, is one of the most promising materials for future electronic devices. Inherit to the atomic thickness and the subsequential high surface-to-volume ratio, graphene presents high electrical and thermal mobilities, optimized electrostatic control of the channel, and sensitivity to changes in its surrounding. These properties make it optimum for biosensing applications. For biological recognition, electrochemical gated sensors employing electrolytes such as ionic liquids and aqueous solutions are widely reported [1,2]. Nonetheless, due to the high sensitivity of graphene and the complex nature of the electrolyte solutions, reproducibility among devices can be challenging [3]. In this work, we propose the use of double gate transistors which allows us both to understand the instability phenomena affecting the sensor performance (interface traps, defects, etc.) and to improve the reliability and variability of the sensors. Experimental Setup. The fabrication process is summarized in Figure 1. Cr/Au contacts are directly deposited on SiO2/Si substrates to perform four devices in a chip. Then, low-pressure chemical-vapor deposited graphene is transferred using the PMMA-based technique [4]. Graphene is annealed and etched to design the transistors and then, partially passivated using deposited Al2O3 in a lift-off photolithography process to avoid any potential drain/source shortcut. For the structural characterization, Raman, XPS, and atomic force microscopy (AFM) are used. For the electrical characterization, semiconductor analyzers and temperature- and pressure-controlled probe stations are employed. Results. Figure 2.a shows the operation of three different fabricated sensors when using PBS as liquid front-gate. The minimum of ID-VFG corresponds to the Dirac point, where variability among sensors is observed. To analyze the origin of this variability, Figure 2.b depicts the drain current vs. time characterization for a specific gate bias. The temporal signature of the current follows an initial decay, a plateau and finally a decay for the three sensors. To determine what is the contribution of the liquid gate in the inter-device variability, Figure 3 shows the same characterization using the back gate analysis. Devices also show the Dirac point shift when biased through the back gate. In this case, in the temporal characterization, a short current increment followed by the typical current decay (usually attributed to defects) is observed. Temporal signature comparison of both bias configurations attributes the current plateau to the liquid gate effect (mobile ions in the liquid or effect of the electric double-layer formation) while the decay periods observed in both cases can be associated with the interfaces in common (SiO2 and drain/source contacts). Conclusions. The objective of using double-gate graphene transistors as sensors is twofold: i) understanding the instability phenomena which affect graphene sensors in terms of reliability and inter-device variability and, ii) taking advantage of the inter-gate coupling between top and bottom interfaces to optimize a biosensing platform with reduced variability.[1] F. Chen et al., J. Am. Chem. Soc. 131, (2009).[2] N. Liu et al., Sensors, 19 (2019).[3] A. Pirkle et al., Appl. Phys. Lett. 99, 122108, (2011).[4] G. Borin, et al., Carbon N. Y. 84, 82–9, (2015). Figure 1