An electronic nose using a single graphene FET and machine learning for water, methanol, and ethanol

Abstract

The poor gas selectivity problem has been a long-standing issue for miniaturized chemical-resistor gas sensors. The electronic nose (e-nose) was proposed in the 1980s to tackle the selectivity issue, but it required top-down chemical functionalization processes to deposit multiple functional materials. Here, we report a novel gas-sensing scheme using a single graphene field-effect transistor (GFET) and machine learning to realize gas selectivity under particular conditions by combining the unique properties of the GFET and e-nose concept. Instead of using multiple functional materials, the gas-sensing conductivity profiles of a GFET are recorded and decoupled into four distinctive physical properties and projected onto a feature space as 4D output vectors and classified to differentiated target gases by using machine-learning analyses. Our single-GFET approach coupled with trained pattern recognition algorithms was able to classify water, methanol, and ethanol vapors with high accuracy quantitatively when they were tested individually. Furthermore, the gas-sensing patterns of methanol were qualitatively distinguished from those of water vapor in a binary mixture condition, suggesting that the proposed scheme is capable of differentiating a gas from the realistic scenario of an ambient environment with background humidity. As such, this work offers a new class of gas-sensing schemes using a single GFET without multiple functional materials toward miniaturized e-noses.

Chemical Gas Sensors: Recent Developments, Challenges, and the Potential of Machine Learning—A Review

“…Therefore, their trained SVM model was successfully able to distinguish caffeine with an accuracy rate of 93.4% [139] (see Figure 17Bb-3). Most recently, Hayasaka et al [140] fabricated a highly selective sensor using pristine graphene and ALD-RuO 2 -based GFET devices with machine learning. In their proposed scheme, the measured V-shaped conductivity profiles were decoupled into four distinctive physical properties combined with other parameters.…”

“…In their proposed scheme, the measured V-shaped conductivity profiles were decoupled into four distinctive physical properties combined with other parameters. These four parameters were used as input feature vectors to classify different gases including electron mobility (µ e ), hole mobility (µ h ), ratio of the electron and hole concentration (n e/h ), the ratio of the residual carrier, and charge impurity concentration (n*/n imp ), represented in Equations ( 9)- (12), respectively [140].…”

“…The electron mobility and hole mobility turned out to be the most important features, with electron mobility having much of the information of the data. The confusion matrix displayed 100% correct values for ALD-RuO 2 -GFET, while pure graphene GFET confused some values with others [140] (see Figure 18). A multilayer perceptron classifier with a feed-forward neural network architecture was implemented and trained using the 4D input feature vector of the two GFETs (pure graphene and ALD-RuO 2 ).…”

“…The cross-validation results indicated that the pristine GFET device had a mean accuracy of 95.4% and a standard deviation of 2.5%, whereas the ALD-RuO 2 -GFET device had a mean accuracy of 99.6% and a standard deviation of 0.8%. [140]. A comparison of FET devices with extracted features and output accuracies is summarized in Table 2.…”

Machine-Perception Nanosensor Platform to Detect Cancer Biomarkers

“…21 "Optical" noses have been developed as well. 22,23 However, these developments are limited in their ability to detect molecules such as proteins and in physiological conditions and complex biofluids. To overcome limitations associated with one-to-one recognition elements, we are investigating the development of a perception-based methodology that uses ML processes coupled with a sensor array, where each element exhibits moderate selectivity for a wide range of molecules.…”