Researchers use clove, eucalyptus, lemon and rose scents to test their machine learning-enabled, graphene-based e-olfaction sensor.

Electronic sensing devices intended to detect odors such as human noses do have great potential in a wide range of modern-day applications. Researchers supported in part by the EU-funded CARBO-IMmap and SMELLODI projects have now developed e-olfaction sensors and introduced a method for assessing their olfactory performance towards volatile organic compounds (VOCs). Their study was published in the journal ‘Applied Physics Reviews’.

VOCs are chemicals used and produced in the manufacture of paints, pharmaceuticals and refrigerants. Exposure to VOC vapors can cause health conditions ranging from eye irritation and headaches to liver and kidney damage. Our ability to detect potentially harmful VOCs and identify their source using our sense of smell is a valuable tool for survival. In modern-day applications, electronic noses, or e-noses, aim to achieve the same thing by digitizing the sense of smell.

The e-olfaction sensing device has a single-channel nanosensor – unlike conventional e-nose systems that utilize sensor arrays – and operates at room temperature. “This unique design shows great potential for miniaturization and portability,” write the authors in their study.

Testing with four scents
To test their sensing device’s olfactory performance in terms of odor threshold performance, odor discrimination and odor identification, the team selected four VOC-based odors widely used to assess people’s sense of smell: eucalyptol (eucalyptus scent), 2-nonanone (lemon scent), eugenol (clove scent) and 2-phenylethanol (rose scent). The device was exposed to the rose scent at decreasing concentrations ranging from 19 to 4.4 parts per million, and was able to pick up the scent even at the lowest concentration. For the odor discrimination test, the sensor was exposed to the four odors and able to discriminate between them with a precision approaching 83.3 %. Furthermore, the use of supervised machine learning classifier algorithms, such as linear discriminant analysis, resulted in high odor identification accuracy (97.5 %).

In addition to the individual odors, the team also investigated the device’s response to mixtures of two odors, finding it capable of processing them efficiently. “The response to binary odor mixtures is shown to behave close to an individual odor while the response of the other odor is partially suppressed. This phenomenon is analogous to the overshadowing effect in human olfaction perception upon processing binary odor mixtures,” the authors report.

The research team used molecular dynamics simulations and density functional theory calculations to explain the adsorption interaction between odorant molecules and sensing materials. They also provided insight into how humidity affects the method and validated it by identifying other VOCs tested in gas sensing applications using humid air as a carrier gas. Their results showed that their e-olfaction sensor can sniff out VOC-based odors effectively.

The e-olfaction platform “leverages arrays of highly sensitive nanomaterials functionalized in diverse ways, thereby enabling the detection and discrimination of a much larger amount of target odor molecules and their complex mixtures. When coupled with mobile devices for data analysis, it holds great promise for assisting individuals with olfactory disorders in the near future. In addition, it has the potential to be applied in numerous emerging fields, such as environmental monitoring or public security.”

CARBO-IMmap (Immune Activity Mapping of Carbon Nanomaterials) ended in 2022. The SMELLODI (Smart Electronic Olfaction for Body Odor Diagnostics) project ends in March 2025.

For more information, please see:
CARBO-IMmap project website
SMELLODI project website

Keywords
CARBO-IMmap, SMELLODI, odor, olfaction, e-olfaction, sensor, scent, smell, volatile organic compound