Reliable spectroscopic screening protocols for the eco-design of analytical workflows in laboratories ensuring EVOO quality categorisation

Abstract

The increasing demand for food safety/quality and environmental sustainability requires the development of rapid, reliable and green analytical screening methods for food quality assessment. Although precise, traditional reference methods are often time-consuming, resource-intensive and environmentally taxing. To address this challenge, we developed fast, reliable and sustainable analytical screening methods for assuring the extra virgin olive oil (EVOO) quality category, a strictly regulated sector. The potential of two analytical techniques by using miniaturised instrumentation was explored: portable spatially offset Raman spectrometry (SORS) and benchtop low-field nuclear magnetic resonance (LF-NMR) spectrometry. 387 olive oil samples of different commercial categories were analysed. Screening methods were developed to ensure the compliance to EVOO category against three quality characteristics of olive oils: acidity, peroxide value and fatty acid ethyl esters. The core of the methodology involved building multivariate classification models based on machine learning. Conservative screening thresholds were established to ensure the highest probability of correctly classifying compliant samples to the EVOO category. The performance of the models was evaluated, achieving 100 % precision in almost all cases, outstanding particularly in screening olive oils based on acidity. Furthermore, whiteness evaluation yielded scores above 80 %, revealing the efficiency and sustainability of the proposed methods, and the high savings indices showed their potential to reduce the number of samples requiring conventional reference analyses, which are environmentally unfriendly. Results demonstrate that both SORS and LF-NMR, coupled with a multivariate approach, offer a viable, green and cost-effective alternative to support routine quality control of olive oils, significantly optimising analytical resources.