Work

Exploratory assessment of rancidity intensity in blended olive oils using an electronic nose and its potential for quality grading

Year: 2025

Type: article

Abstract: An electronic nose (e-nose) prototype equipped with nine commercial metal oxide semiconductor (MOS) sensors was used as a rapid, non-invasive tool to evaluate olive oil quality. To assess its performance, extra virgin olive oils (EVOO), with medium or intense fruitiness, were blended with rancid oil at different proportions (0%, 2%, 4%, 8%, 16%, 32%, 100%). According to physicochemical parameters (free acidity, peroxide value, UV extinction), oils containing up to 32% rancid oil were classified as EVOO or virgin olive oil (VOO), while rancid oil alone was categorized as lampante (LOO). Sensory evaluation confirmed progressive quality deterioration, assigning blends as EVOO (0% and 2%), VOO (4% and 8%), and LOO (16% and 32%) based on rancid intensity perception by trained panelists. When coupled with linear discriminant analysis, the e-nose accurately distinguished oils by commercial category, achieving average sensitivities of 92% and 95% (repeated K-fold cross-validation), respectively, for rancid blends with medium or intense defect-free oils. Additionally, e-nose signal profiles allowed establishing simple linear regressions (Signal = slope × Rancid + intercept), enabling quantitative prediction of rancid defect intensity on a 0–10 scale. For medium or intense fruitiness levels, the regression models showed satisfactory fitting (R 2 : 0.9851-0.9948), low error (RMSE: 0.23-0.43), high predictive potential (RPD: 3.5-16.6) and low detection and quantification limits (LOD: 0.5-1.0; LOQ: 1.5-2.9). These results demonstrate that the prototype accurately classifies oil grades and quantifies rancidity, offering a practical, user-friendly tool for olive oil quality screening, though further validation with more samples and standard reference methods is needed.

Source: Food Control

DOI: 10.1016/j.foodcont.2025.111839

Authors: Fedi Landoulsi, António M. Peres, Ana C.A. Veloso, Maissa Khemakhem, José Alberto Pereira, Nuno Rodrigues

Institutions: Tunisia Private University, Universidade Politécnica de Bragança, Polytechnic Institute of Coimbra

Language: English

Topic: Advanced Chemical Sensor Technologies

Subfield: Biomedical Engineering

Field: Engineering

Domain: Physical Sciences

Open Access status: hybrid

Estimated APC paid: $4,340

Funders: Ministério da Ciência, Tecnologia e Ensino Superior, Fundação para a Ciência e a Tecnologia