Aplicação da espectroscopia de fluorescência por análise direta na determinação de contaminantes em alimentos

Abstract

Food contamination poses serious health risks and remains a global challenge due to the presence of pathogens, toxins, and chemical residues. Current detection methods are often costly, time consuming, and require complex instrumentation, limiting their practicality for routine monitoring. Direct fluorescence spectroscopy presents a rapid, non-destructive, and cost effective alternative, enabling fast detection through intrinsic or modified fluorescence signals. This technique simplifies analysis and holds strong potential for real-time, food safety monitoring. In this study, three analytical methods were developed using spectrofluorometry to evaluate regulated chemical contaminants in various food matrices. A method for the determination of benzo[a]pyrene (B[a]P) in Pilsner malt was established and validated using liquid-liquid extraction and synchronous fluorescence. The optimized method showed a linear detection range of 0.5 - 3.0 μg kg-1, with limits of detection (LOD) and quantification (LOQ) of 0.24 μg kg 1 and 0.81 μg kg-1, respectively, and the experimental LOQ established as 0.50 μg kg-1. Precision and trueness assessments yielded relative standard deviations (RSD) below 10%, and recovery rates between 97.9% and 114.6% at B[a]P concentrations of 0.5, 1.5, and 2.5 μg kg-1. A second method was developed and validated for the detection of ochratoxin A (OTA) in red wine, combining liquid-liquid extraction with synchronous fluorescence enhanced by β-cyclodextrin inclusion complexation. The method exhibited a linear detection range of 0.5 - 3.0 μg L-1, with LOD and LOQ of 0.20 μg L-1 and 0.60 μg L-1, respectively, and the experimental LOQ established as 0.50 μg L-1. RSD values were below 30%, and recovery rates ranged from 84.9% to 95.1% at OTA concentrations of 0.5, 2.0, and 3.0 μg L-1. Additionally, an exploratory chemometric model based on fluorescence excitation-emission matrices and parallel factor analysis (PARAFAC) was developed to identify 5-hydroxymethylfurfural (HMF) in mead. Preliminary results demonstrated the technique's potential to discriminate fluorescent components associated with HMF, with a significant correlation observed between PARAFAC-extracted scores and reference concentrations. These findings support the feasibility of incorporating such modeling approaches into future multivariate calibration strategies, benefiting from second-order data advantages. The integration of multidimensional fluorescence spectroscopy, supramolecular modifications, and multivariate analysis provided high selectivity and sensitivity, highlighting direct fluorescence as a simple, cost-effective, and viable alternative to traditional analytical methods.