Guillermo Medina-González

Ensuring food authenticity, traceability, and quality has become a critical challenge in increasingly complex and globalized food supply chains. Conventional post-harvest analytical approaches, while powerful, often operate in isolation and fail to fully capture the influence of pre-harvest conditions on food composition. In parallel, the emergence of smart farming technologies has enabled the collection of high-resolution environmental and agronomic data, offering new opportunities to establish baseline signatures linked to geographical origin and production practices. This review explores the integration of pre-harvest data from precision agriculture with advanced post-harvest analytical techniques, focusing on spectroscopic and isotopic methods for food authentication. Recent advances in vibrational spectroscopy, including near- and mid-infrared, Fourier-transform infrared, and Raman techniques, alongside complementary methods such as nuclear magnetic resonance and fluorescence spectroscopy, have enabled rapid and non-destructive food fingerprinting. In parallel, isotope ratio mass spectrometry and compound-specific isotope analysis provide robust markers of origin, climate conditions, and agricultural inputs through the analysis of stable isotopes of carbon, hydrogen, oxygen, nitrogen, and sulfur. The combination of these analytical approaches with chemometric and machine learning tools facilitates the extraction of meaningful patterns from complex datasets. A central focus of this review is the development of integrated farm-to-fork frameworks that use multi-source data, including field sensor technologies, spectral fingerprints, and isotopic signatures, to enhance traceability and authentication. Applications across a wide range of food systems, including edible oils, beverages, plant-based products, and animal-derived foods, are critically evaluated to highlight the strengths and limitations of current methodologies. Key challenges related to data standardization, system interoperability, cost, portability, miniaturization and regulatory acceptance are discussed, alongside emerging solutions such as artificial intelligence-driven models, digital twins, and blockchain-enabled traceability systems. The review underscores a paradigm shift from reactive testing toward predictive and real-time food authentication systems, driven by the convergence of smart agriculture and advanced analytical chemistry. This integrated approach has the potential to significantly enhance transparency, trust, and sustainability in the global food system.

The accuracy of determining arsenic and lead using the optical technique Slurry–Total Reflection X-ray Fluorescence (Slurry-TXRF) was significantly enhanced through the application of a machine learning method, aimed at improving the ecological risk assessment of agricultural soils. The overlapping of the arsenic Kα signal at 10.55 keV with the lead Lα signal at 10.54 keV due to the relatively low resolution of TXRF could compromise the determination of lead. However, by applying a Partial Least Squares (PLS) machine learning algorithm, we mitigated interference variations, resulting in improved selectivity and accuracy. Specifically, the average percentage error was reduced from 15.6% to 9.4% for arsenic (RMSEP improved from 5.6 mg kg⁻¹ to 3.3 mg kg⁻¹) and from 18.9% to 6.8% for lead (RMSEP improved from 12.3 mg kg⁻¹ to 5.03 mg kg⁻¹) compared to the previous univariable model. This enhanced predictive accuracy, within the set of samples concentration range, is attributable to the efficiency of the multivariate calibration first-order advantage in quantifying the presence of interferents. The evaluation of X-ray fluorescence emission signals for 26 different synthetic calibration mixtures confirmed these improvements, overcoming spectral interferences. Additionally, the application of these models enabled the quantification of arsenic and lead in soils from a viticultural subregion of Chile, facilitating the estimation of ecological risk indices in a fast and reliable manner. The results indicate that the contamination level of these soils with arsenic and lead ranges from moderate to considerable.

A machine learning method was applied to improve the accuracy of the determination of Arsenic and Lead by Slurry - Total Reflection X-ray fluorescence (Slurry-TXRF) with the idea of being applied to the ecological assessment of agricultural soils. Due to TXRF's relatively low resolution, a particular and well-known overlapping of arsenic signal Kα at 10.55 keV with Lαsignal at 10.54 keV of the lead can compromise its determination. Applying a multivariate calibration method based on a machine learning algorithm, for example, Partial Least Squares, could reduce variations due to interference and, consequently, improve the selectivity and accuracy in arsenic and lead determination. In this work the X-Ray fluorescence emission signals was evaluated for a set of 26 different synthetic calibration mixtures and a significant accuracy improvement for arsenic and lead determination was observed, overcoming the problems associated with spectral interferences. Furthermore, with these models, arsenic and lead were quantified from soils of a viticultural subregion of Chile, allowing the estimation of ecological indices in a fast and reliable way. The results report that the level of contamination of these soils concerning arsenic and lead is moderate to considerable.

The study aims to evaluate the applicability of the slurry-TXRF method for estimating background contents and ecological indices in a rapid and convenient way. For this reason, the agricultural soils of the Itata Valley were used as a case study, where 48 soil samples were collected and analyzed. This rapid, minimally sample-intensive, and simultaneous multi-element quantification technique presented high accuracy but lower precision (approx. 20% RSD) compared to the classic total reflection X-ray fluorescence and flame/graphite furnace atomic absorption spectrometry methods, which require sample digestion. Due to the analytical characteristics of Slurry-TXRF, it can be concluded that the lower precision is likely compensated for, and this method represents a valuable alternative for the rapid and efficient assessment of trace element contamination in agricultural soils. The regional median concentrations of Cr, Ni, Cu, Zn, and Cd in the Itata Valley surface soils were found to be 63.7, 9.57, 31.0, 41.1, and 0.56 mg kg⁻¹, respectively, with corresponding upper limits of 47.6, 6.82, 17.0, 30.7, and 0.284 mg kg⁻¹. The ecological indices, including the geoaccumulation index, contamination factor, enrichment factor, and degree of contamination, suggest moderate levels of contamination in the region.

A multivariate calibration method was applied to improve the accuracy of the determination of cadmium in the presence of potassium by total reflection X‐ray fluorescence (TXRF). Due to TXRF's relatively low resolution, a particular interference caused by potassium at line Kα 3.31 keV as interference can compromise the cadmium determination at line Lα 3.13 keV, respectively. The method is based on the hypothesis that the application of a multivariate calibration method, for example, partial least squares, could reduce variations due to interference and, consequently, improve the selectivity and accuracy in Cd determination. In this work, this strategy was evaluated on the X‐ray fluorescence emission signals between 2.50 and 3.90 keV for a set of 26 different synthetic calibration mixtures and eight different mixtures for external calibration. Based on a certified reference material (Clay 2—CRM051) with high potassium levels, a significant improvement of accuracy for cadmium determination was observed, overcoming the problems associated with spectral interferences by potassium.