Second-Order Calibration for Quantification of Food Colorants in Complex Samples: An Universal Calibration Approach

The utilization of colorants in food products is imperative to achieve visually appealing and natural shades that appeal to consumers. However, due to cost and stability considerations, natural colorants are frequently substituted with synthetic alternatives, which may present potential health risks. A variety of analytical methods have been proposed for the detection of food colorants [1, 2], with separation techniques being the most commonly employed. Nevertheless, the establishment of a universal calibration for accurate quantification across diverse real and complex samples remains a challenging endeavor. In order to address this issue, contemporary analytical instruments, when combined with multi-way multivariate calibration methodologies, offer effective solutions for quantification problems, particularly in the presence of unknown interferents. Second-order calibration methods demonstrate promise in developing universal calibration models for determining colorants in food products. This study introduces a higher-order multivariate chemometric approach to construct a universal calibration for quantifying Sunset Yellow (SY) as a model compound in various food products, even in the presence of non-calibrated interferences. By utilizing second-order pHspectrophotometric data obtained through hyper-spectral imaging, the MCR-ALS second-order calibration strategy achieves efficient quantitation of SY in different food products, providing a practical and effective approach for food analysis and quality control. The data pertaining to the standard samples and the real sample containing SY at 16 different pH were simultaneously measured using a hyperspectral imaging (HSI). This design enhances the analysis speed by leveraging the capabilities of the HSI device and creates a second order data to achieve the advantage of analyzing samples with non- calibrated components. The proposed method successfully analyzed real samples of colored drinks, jelly, and candy. The proposed method is applicable for the quantitation of all colorants that exhibit sensitivity to pH and function as weak acids or bases.