Electronic nose detects multiple odors
The human nose consists of about ten million olfactory cells, with about 400 different olfactory receptors. These receptors detect the odors and generate a specific signal pattern. The brain assigns the signal pattern to a certain smell. “We have taken the biological nose as a model,” says Dr. Martin Sommer, who is in charge of the smelldect project at the KIT’s Institute of Microstructure Technology. “In our electronic nose, nanofibers react to complex gas mixtures – i.e. odors – and also form signal patterns that are used by the sensor to detect them. The goal of the project is to develop a low-cost odor sensor suitable for mass production and everyday use.
The electronic nose is only a few centimetres in size. It contains all the operating electronics, including the technology for evaluating the gases. The “nose” consists of a sensor chip on which nanowires of tin dioxide are attached to many individual sensors. The chip calculates specific signal patterns through the resistance changes of the individual sensors. These depend on the molecules in the ambient air, are different for different odors – and therefore characteristic and recognizable. If this pattern was previously taught into the chip, the odor sensor can detect it within seconds.
To get the process going, the researchers use a light emitting diode integrated into the sensor housing, which irradiates the nanowires with UV light. As a result, the originally very high electrical resistance of tin dioxide decreases to such an extent that changes in it – caused by the molecules responsible for the smell and deposited on the tin dioxide surface – can only be detected in the first place. “If the sensor detects a smell, the resistance drops even further. When the odor disappears, the original conditions are restored with a correspondingly high electrical resistance, so that the “nose” is ready for further odor measurements,” says Sommer.
The sensor chip can learn a wide variety of different odors and can therefore be used in many different ways: whether in the home to control the indoor air or as a fire detector, when shopping, to detect how fresh fish or meat is, in the final quality control of honey, for example, or as a nose for a robot. “The difficulty is that smell is not the same as smell. A rose, for example, smells differently in sunshine than in the rain,” says the physicist. “That’s why we are currently training the electronic nose for specific applications, which are, however, universally selectable.”
The scientists intend to develop a sensor that is as inexpensive as possible in order to make it suitable for mass production. “In the future, for example, the electronic nose could be installed in all electrical appliances to prevent cable fires. Or we equip smartphones with it. Everyone would then have their own highly sensitive electronic nose when shopping,” says Sommer.
Project partners JVI-Elektronik and FireEater support the KIT in industrial production and sales. Together with KIT, they already developed an intelligent fire detector based on an electronic nose in the EU project “SmokeSense” in 2015. It detects gases generated in smoldering and combustion processes and offers an analysis of which burning materials are involved.
Related articles:
Cyborg insects as wireless biorobotic sensing machines
Microsoft smart mat blurs digital, physical worlds