Effectiveness of an electronic nose based on metal oxide semiconductor sensors for evaluating coffee quality.
Coffee quality, which is essentially reflected in the beverage’s aroma, depends on various factors that require multiple approaches to verify, entailing significant costs and lengthy analysis times. In this context, the aim of a recent study conducted by a group of Brazilian researchers (Mutz et al., 2025) was to develop and calibrate an electronic nose in combination with chemometrics for the analysis of the qualitative characteristics of this product. Specifically, twelve different metal oxide sensors were used in the development of the instrument.
The objectives of the analysis are as follows: (i) distinguishing between the species Coffea arabica and Coffea canephora, (ii) differentiating between roasting profiles (light, medium, and dark), and (iii) separating expired samples from non-expired ones. Principal component analysis (PCA) indicates good clustering of the tested samples based on their characteristics, highlighting the potential of volatile compounds in sample classification.
Furthermore, the use of independent soft modeling by class analogy (SIMCA), partial least squares discriminant analysis (PLS-DA), and least squares support vector machine (LS-SVM) yields excellent results with an accuracy exceeding 90% for each required objective. In conclusion, the combination of the electronic nose with various chemometric models can be effectively used for multifunctional classification tasks by coffee producers, representing a more economical, rapid, and effective alternative to traditional methods.
Use of Smart Electrochemical Sensors for Coffee Quality Assessment.
Quality control is mandatory in the food industry, and chemical sensors play a crucial role in this field. Coffee is one of the most widely consumed and marketed food products in the world, and its quality characteristics are therefore of fundamental importance to the industry. The objective of a recent study, conducted by a group of Italian researchers (Grasso et al., 2025), was to evaluate the ability of a smart electrochemical sensor to distinguish between different beverages prepared with coffee beans having different moisture contents (0, 2, and > 4%) and ground to three different particle sizes (fine, medium, and coarse).
These parameters, in fact, reflect real-world scenarios in which this product is processed and its quality is affected. A specific experimental setup was designed for the tests, and the data were analyzed using machine learning techniques. The results obtained from principal component analysis (PCA) and partial least squares discriminant analysis (PLS-DA) demonstrate the sensor’s ability to distinguish between samples of different quality, with a correct classification rate of 86.6%.
The ability to optimize the quality characteristics of coffee in real time by adjusting parameters such as moisture content and particle size paves the way for the development of intelligent machines capable of delivering significant savings in time and resources. However, further research is needed to further improve the classification capability of the proposed sensor.
References: Y.S. Mutz et al., Chemosensors, 2025, 13, 1–16;Grasso et al., Chemosensors, 2025, 13, 1–18.
