Machine Learning for Carbon Nanotube Optical Sensors

Abstract

Perception-based machine learning platforms, modeled after the complex olfactory system, can isolate individual signals through an array of relatively nonspecific receptors. Each receptor captures certain features, and the overall ensemble response is analyzed by the neural network in our brain, resulting in perception. Previous attempts to develop perception-based sensing platforms include “electronic nose” technologies for gas sensing, based on conducting polymers, DNA-decorated field-effect transistors, and other technologies based on protein recognition using simple data analytic techniques. We developed single-walled carbon nanotube (SWCNT)-based sensor arrays wherein the optical responses of the nanotubes were used to train machine learning models. Due to the existence of several mechanisms of optical modulation of SWCNTs, as well as multiple carbon nanotube species with differing bandgaps, and the increasing number of moieties that can be used for colloidal suspension and functionalization, new methods in data analytics can be used to facilitate selective recognition of bioanalytes to detect and process multi-parametric spectroscopic fingerprints. These methods can improve the detection of biological analytes and to diagnose diseases.