Electronic nose is a technology that makes it possible to identify volatile chemicals in gas mixtures. The eponymous device operates on the same principle as the human nose, analyzing the smell as a whole. In this, it differs from combination gas chromatograph/mass spectrometers, which progressively check every component in a gas mixture. While highly accurate, gas chromatograph/mass spectrometers are bulky, which means they can only be used in laboratory conditions. This is why a more compact (book-sized) and mobile electronic nose can be used in field conditions, including for environmental monitoring. However, the device must be pre-configured in order to easily recognize the target substances.
The scientists from Skoltech and Kazakhstan’s Eurasian National University have proposed using an electronic nose to identify oil components. The authors took nine oil samples from three different oil fields in Kazakhstan, containing both light and heavy hydrocarbon fractions. The authors simulated an oil spill by adding samples to soil and keeping them in a gas analysis unit for half an hour, after which the air flow with volatile substances was examined by an electronic nose.
The electronic nose was trained with the use of artificial intelligence (AI): an algorithm analyzed the consequences of decision-making and produced a result at the output. With the help of AI, the scientists managed to adjust the processing of data coming from the device’s eight sensors, thanks to which the electronic nose started to identify the source of oil with high accuracy, regardless of its volatility. This adjustment turned out to be very effective: the device could detect the presence of oil in soil even 12 hours after the end of the experiment, when the oil had already partially evaporated.
“Our approach can be used at oil fields to determine how productive a particular formation is. The electronic nose will establish its chemical characteristics by the smell of the volatile fraction of oil in the air. Presumably, the electronic nose will also make it possible to identify the locations of new oil fields,” Valery Zaitsev, Skoltech postgraduate student, is quoted as saying by the Russian Science Foundation.
