Highly sensitive room temperature nanomaterial gas sensor array for smart electronic nose systems

Abstract

The electronic nose, an imitation of the mammal’s nose, could transform gas information into the electronic signal, thus it can meet the desire of “smell” the world. Typically, an electronic nose system is made of a gas sensor array, a readout circuit and a classification algorithm. Metal oxide semiconductor gas sensors have been applied in the electronic nose system for decades because of their stability, low cost, sensitivity and robustness. Moreover, the thrivingly developing nano-technology in both fabrication process and gas-sensitive materials has provided a new approach for metal oxide semiconductor based gas sensors to achieve room temperature gas detection ability with solid performance. The readout circuit could change the analogue signal from the sensor array to the digital output for the following classification algorithm. The pre-trained classification algorithm can accurately distinguish detected gases or even the gas mixture. In this thesis, we firstly demonstrate the unique U-shape response curve of a hierarchical ZnO gas sensor towards breath level acetone with temperature modulation. Through the temperature modulation, the unique U-shape response curve could be utilized as the fingerprint response pattern for acetone detection and discrimination. The proposed ZnO sensor presents the potential of gas sensors’ application in non-invasive health monitoring through breath. Besides the single gas sensor, we developed an array of sensors using hybrid materials for gas sensing. Through modulating the combination ratio of polyvinylpyrrolidone and tin oxide nanoparticle prepared through hydrothermal method, a sensor array that has the ability to identify drunk driving has been demonstrated. Also, by further adding noble metal nanoparticles as the catalyst and using polyvinylpyrrolidone and indium tin oxide as the sensing material, the micro SD card size electronic nose could classify six different solutions with the classification performance of 99.2%. In addition, we present a room temperature and highly sensitivity ultra-low power tin oxide gas sensor array which is based on the open-ended nanotube structure. Through conformal deposition of tin oxide nanoparticle on the porous alumina membrane template, the tin oxide nanotube array shows an excellently mechanical strength. In addition, four different conductive materials are deposited on the tin oxide nanotube array to form a monolithic gas sensor array. Benefiting from the Knudsen diffusion and platinum nanoparticle decoration, the sensor array has reached the state-of-art hydrogen and benzene detection capability. Moreover, because of the energy-hungry heater is removed, the electronic nose system (including the sensor array and readout circuit) could be powered by batteries and realize wireless sensing and transmitting. Combining with the algorithm, the electronic nose could identify four different gases. Further improvement on the algorithm demonstrates the electronic nose’s capability of identifying six organic solutions with different concentrations. Furthermore, we increase the number of the sensors in the array, achieving a sixteen-sensors array on one tin oxide nanotube membrane. By depositing eight different metals or metalloid materials as the electrodes, the sensor array could provide sixteen dimensions’ gas information for the classification algorithm. The combination of a k-fold cross validation and bagging decision tree algorithm have reached the classification performance of 96.9%, 96.8% and 80.2% for three kinds of binary gases mixture (hydrogen and ethanol, hydrogen and acetone, hydrogen and methanol), respectively. This work shed light on the future gases mixture detection and classification. The thesis explores five different gas sensor or electronic noses. Benefiting from the nano-technology, all the electronic noses could be operated at room temperature. Combining the readout circuit and classification algorithms, the electronic noses demonstrated in the thesis present excellent performance in gases, solution vapors or gas mixtures detection and classification.

Date of Award	2018
Original language	English
Awarding Institution	The Hong Kong University of Science and Technology