Discriminative Detection of Different Cigarette Brands Using a Fast-Response Electronic Nose

Electronic noses are an artificial olfactory system that mimics the animal olfactory function. Currently, metal oxide semiconductor (MOS) gas sensors play a vital role in the development of high-performance electronic noses and have widespread applications across various fields. They are particularly valuable in ensuring food safety, monitoring air quality, and even detecting explosives for antiterrorism purposes. However, there is an increasing demand for electronic noses to exhibit faster response times in large-scale commercial applications. To address this challenge, we developed a novel MOS gas sensor with a porous ceramic substrate, specifically designed to facilitate rapid gas diffusion. The sensing performance of the sensor array was evaluated and the result showed that the T₉₀ time of porous ceramic-assisted MOS sensor was significantly (57%) shorter than sensors with a normal substrate. Moreover, the electronic nose system had demonstrated remarkable capability in accurately distinguishing between five distinct types of hazardous gases, including VOCs as well as ammonia. Furthermore, a low-cost electronic system was developed and applied to cigarette brand identification; 2490 groups of data were collected for each individual test at only a cost of 20 s. By employing a machine learning algorithm to analyze the data, an accuracy higher than 95% was achieved (96.29% for K nearest neighbor and 96.32% for random forest). We found that our system can resolve the onset time of electronic nose measurement with enough precision, and it was expected that this special approach by using porous ceramic as an insulating substrate can provide a simple and reliable method to manufacture a fast-response electronic nose.