Intelligent Sniffer: A Chip-Based Portable E-Nose for Accurate and Fast Essence Evaluation

Yiwei Jiang,Ziyi Wang,Yan Hu,Wenhao Lin,Longchao Yao,Wenfang Xiao,Jiangxue Hu,Weijie Liu,Chenghang Zheng,Linghong Chen,Xiang Gao
DOI: https://doi.org/10.1016/j.snb.2024.136989
IF: 9.221
2024-01-01
Sensors and Actuators B Chemical
Abstract:Electronic nose (e-nose) is capable of identifying gases of different odors and concentrations. However, accurate odor evaluation of complicated volatile compounds (e.g., from essence) at human perception level remains challenging, especially in a fast and cheap way. This study presents a portable e-nose designed to evaluate different essence samples and their mixtures that outperforms human olfaction. Given the complexity and similarity of ingredients among different essence types, six sensors were integrated on a miniaturized chip after the sensor array optimization from 18 gas-sensing materials with varying doping, film thicknesses, operated at different temperatures. A pattern recognition model based on Long-short term memory network and convolutional neural network was developed to automatically extract sensors’ dynamic response features to determine essence type and concentration. Our e-nose exhibited high sensitivity and repeatability to analyze low-concentration essence volatiles better than human olfactory tests. The classification accuracy for four similar textile essence and their mixtures could reach up to 96.9%. And the R2 and RMSE for relative concentration achieved 0.98 and 0.068%, respectively. Furthermore, only the first 60s signals are needed for the sample analysis which significantly reduced the sample testing time compared to the traditional base-peak-base test method (over 200s). This research provides a favorable opportunity for essence fragrance prediction in various applications such as food, textile, cosmetics, etc.
What problem does this paper attempt to address?