TXRF spectral information enhanced by multivariate analysis: A new strategy for food fingerprint

The increased costumers' request of safe and high-quality food products makes food traceability a priority for frauds identification and quality certification. Elemental profiling is one of the strategies used for food trace-ability, and TXRF spectroscopy is widely used in food analysis even if its potentialities have not been fully investigated. In this work, a new method for food traceability using directly TXRF spectra coupled with multi-variate analyses, was tested. Twenty-four different beans' genotypes (Phaseolus vulgaris L.) grown onto two different sites have been studied. After the development of the method for beans' analysis, TXRF spectra were collected and processed with PCA combined with SNV and GLSW filter obtaining a perfect clustering of the seeds according to their geographical origin. Finally, using PLS-DA, beans were correctly classified demonstrating that TXRF spectra can be successfully used as fingerprint for food/seed traceability and that elemental quantification procedure is not necessary to this aim.

Automated summary

The increased demand for safe and high-quality food products from customers has made food traceability a priority for identifying frauds and certifying quality. Elemental profiling, particularly using TXRF spectroscopy, is a widely used strategy for food traceability, although its full potential has not been fully explored. In this study, a new method using directly TXRF spectra coupled with multivariate analyses was tested for food traceability. The study focused on 24 different genotypes of beans (Phaseolus vulgaris L.) grown in two different locations. After developing the method for bean analysis, the TXRF spectra were collected and processed using PCA combined with SNV and GLSW filters, resulting in a perfect clustering of the beans based on their geographical origin. Finally, using PLS-DA, the beans were correctly classified, demonstrating that TXRF spectra can be successfully used as a fingerprint for food and seed traceability and that elemental quantification is not necessary for this purpose.