A neural network approach to classify inversion regions of high mobility ultralong channel single walled carbon nanotube field-effect transistors for sensing applications
Hari Krishna, S. V.
Date of Issue2013
International Nanoelectronics Conference (5th : 2013 : Singapore)
School of Mechanical and Aerospace Engineering
Millimetre long individual single walled carbon nanotubes (SWCNTs) were consistently grown and fabricated into carbon nanotube field effect transistors (CNTFETs). In this work, we extracted the effective mobilities in the strong inversion region, near-threshold region and subthreshold region respectively for these long-channel CNTFETs. Using the mobility data as an input parameter, an artificial neural network (ANN) employing multi-layer perceptron (MLP) architecture was used to classify the different inversion regions of the mobility curves with an accuracy of 90%.