Electrical Current Characteristics of Simulated Carbon Nanotube Network Field Effect Transistors

\begin{figure}[htbp] \centerline{\includegraphics[width=0.32in,height=0.33in]{270920191.eps}} \label{fig1} \end{figure} \begin{figure}[htbp] \centerline{\includegraphics[width=0.25in,height=0.25in]{270920192.eps}} \label{fig2} \end{figure} \begin{figure}[htbp] \centerline{\includegraphics[width=0.32in,height=0.33in]{270920193.eps}} \label{fig3} \end{figure} \begin{figure}[htbp] \centerline{\includegraphics[width=0.25in,height=0.25in]{270920194.eps}} \label{fig4} \end{figure} \begin{figure}[htbp] \centerline{\includegraphics[width=0.32in,height=0.33in]{270920195.eps}} \label{fig5} \end{figure} \begin{figure}[htbp] \centerline{\includegraphics[width=0.25in,height=0.25in]{270920196.eps}} \label{fig6} \end{figure} \begin{figure}[htbp] \centerline{\includegraphics[width=0.32in,height=0.33in]{270920197.eps}} \label{fig7} \end{figure} \begin{figure}[htbp] \centerline{\includegraphics[width=0.25in,height=0.25in]{270920198.eps}} \label{fig8} \end{figure} Carbon nanotube (CNT) network field effect transistors (FETs) offer a promising method for creating biosensors. Sensing occurs due to electrostatic gating which impacts the metallic-semiconducting (m-s) junctions in particular, and is heavily influenced by the morphology of the CNT network. Using a simulated random stick network, we assigned sticks to be either metallic (m) or semiconducting (s) with ratios and densities similar to actual devices and simulated electrostatic gating at each m-s junction in the network. For biosensing applications CNT FETs should have optimized sensitivity. However, it is not fully understood how a network's morphological parameters impact its overall network sensitivity. Using our simulation, we mapped sensitivity as the impact of gating each m-s junction within the network to the overall change in network current. This process was done for multiple simulated networks of varying tube densities. Our results showed that not all m-s junctions influence the network the same, and allowed us to determine which m-s junctions act. In addition, we verified that m-s junctions most influence the network response when the networks are of low density affirming that sparse networks have higher sensitivity.