Graphene chemical and biological sensors : modeling, systems, and applications

Abstract

Graphene exhibits a unique combination of properties making it particularly promising for sensing applications. This thesis builds new graphene chemical and biological sensing technologies from the ground up by developing device models, systems, and applications. On the modeling side, this thesis develops a DC model for graphene electrolyte-gated field-effect transistors (EGFETs). It also presents a novel frequency-dependent (AC) small-signal model for graphene EGFETs and demonstrates the ability of these devices to operate as functional amplifiers for the first time. Graphene sensors are transitioned to the system level by developing a new sensor array architecture in conjunction with a compact and easy-to-use custom data acquisition system. The system allows for simultaneous characterization of hundreds of sensors and provides insight into graphene EGFET performance variations. The system is adapted to develop solution-phase ionized calcium sensors using a graphene EGFET array that has been functionalized using a polyvinyl chloride (PVC) membrane containing a neutral calcium ionophore. Sensors are shown to accurately quantify ionized calcium over several orders of magnitude while exhibiting excellent selectivity, reversibility, response time, and a virtually ideal Nernstian response of 30.1 mV/decade. A new variation-insensitive distribution matching technique is also developed to enable faster readout. Finally, the sensor system is employed to develop gas-phase chemiresistive ammonia sensors that have been functionalized using cobalt porphyrin. Sensors provide enhanced sensitivity over pristine graphene while providing selectivity over interfering compounds such as water and common organic solvents. Sensor responses exhibit high correlation coefficients indicating consistent sensor response and reproducibility of the cobalt porphyrin functionalization. Variations in sensitivity follow a Gaussian distribution and are shown to stem from variations in the underlying sensor source-drain currents. A detailed kinetic model is developed describing sensor response profiles that incorporates two ammonia adsorption mechanisms--one reversible and one irreversible.