2024/09/24 by Nerea Núñez, Javier Saurina, Óscar Núñez · 1 voice
Medicine · Agricultural and Biological Sciences · #Coffee research and impacts #Tea Polyphenols and Effects #Food Chemistry and Fat Analysis
paper · doi:10.1016/j.microc.2024.111770
openalex publication_date 2024/09/24 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
• Targeted LC-HRMS characterization of 26 polyphenolic compounds in coffee samples. • Polyphenolic profiles resulted in excellent sample chemical descriptors for coffee authentication. • Good PLS-DA coffee classification based on variety, production region and roasting degree. • The most discriminant polyphenols on each coffee class were tentatively annotated. • Quantitation of adulteration levels in coffee by PLS regression with overall prediction errors below 7.8%. The importance of monitoring the presence of bioactive compounds as food attributes for sample classification and characterization is increasing. In this study, targeted Liquid Chromatography coupled with High-Resolution Mass Spectrometry (LC-HRMS) was employed to analyze the chemical profile of polyphenolic compounds as the source of information for the characterization and classification of 306 commercial coffee samples. Coffee holds a distinguished position as one of the most widely popular beverages globally but also one of the most easily adulterated. Regrettably, in recent times, instances of coffee adulteration have been on the rise. Consequently, implementing rigorous quality control measures for coffee becomes imperative to guarantee its quality. The results obtained in this work confirm that the proposed chemical profiles serve as excellent descriptors for sample characterization and classification through the implementation of principal component analysis (PCA) and partial least squares-discriminant analysis (PLS-DA), achieving classification rates higher than 83.3% in PLS-DA validation. Moreover, the proposed LC-HRMS polyphenolic approach was employed to identify and measure adulteration levels in coffee samples using partial least squares (PLS) regression with prediction errors below 7.8%.