Changes in the phenolic acids composition during pancake preparation: Whole and refined grain flour and processed food classification by UV and NIR spectral fingerprinting method-Proof of concept

We investigated the changes in the levels of phenolic acids during pancake preparation (flour to batter-pancake) from two refined and whole-wheat wheat varieties. In addition, we evaluated if multivariate analysis of ultraviolet (UV) and near infrared (NIR) spectral fingerprinting data can be used for classification of samples based on preparation stage, wheat varieties, and flour types (whole and refined). Results indicated that total phenolic acids did not significantly change (< 10%) during preparation of pancake from the refined and whole wheat flours. Most (> 90%) of the phenolic acids existed in insoluble bound form and ferulic acid (82-92%) was the most abundant phenolic acid. Correlation between simple UV spectral scan and HPLC analysis for the assay of phenolic acids ranged from 0.771 to 1.00. Principle component analysis (PCA) of NIR spectral data from 4900 to 5900 cm−1 showed clear separation between flour, batter, and pancake. Additionally, the PCA of UV spectral data between 250 and 350 nm separated two into clusters well (refined and whole-wheat samples). The results presented in this manuscript with limited number of samples illustrate the proof of concept that spectral fingerprinting techniques show promising potential for whole and refined grains sample classification and their products.