Fast and Selective Microfluidic Chips for Electrochemical Antioxidant Sensing in Complex Samples

In this work, capillary electrophoresis chips with electrochemical detection have been assessed as creative and selective microfluidic platforms to integrate and simplify on a microscale the traditional methods for complex natural antioxidants determination. Depending on the acid−base properties of the analytes, two approaches (class-selective electrochemical index determination (CSEID) and individual antioxidant determination (IAD)) were investigated for the analysis of nine antioxidants ((+)-catechin, rutin, quercetin, chlorogenic, ferulic, caffeic, protocatechuic, vanillic, and gallic acids) in food samples. First, the novel concept of a class-selective electrochemical index is proposed allowing a fast and reliable determination of the main antioxidant classes (flavonoids and phenolic acids) in less than 100 s. In addition, an impressive separation of nine antioxidants is also offered in less than 260 s with the individual antioxidant determination approach. Qualitative and quantitative performances of both approaches were studied. The analytical figures of merit (i.e., electroosmotic flow (EOF) precision as relative standard deviation (RSD), resolution, signal precision as RSD, limit of detection, limit of quantification, and accuracy as recovery) of both approaches were <4%, ≈1, ≤5%, <8 μM, 30 μM, between 91% and 104%, and <4%, ≤2%, <9%, ≤6 μM, ≤20 μM (with the exception of protocatechuic acid, which shows values of 40 and 130 μM, respectively), between 80% and 107% for the CSEID and IAD concepts, respectively, which are excellent for food samples analysis. A set of representative samples was analyzed including apple and pear skins and pulps, red and white wines, and green tea tablets. High agreement was observed between the results of the sample analyses from the two microchip-based approaches, and good correlation was observed with results obtained from traditional methods. Although the prominent phenolic antioxidant classes and compounds were successfully determined, some nonprominent peaks were not detected in the samples when applying the IAD approach. Different integration strategies on microchip platforms were further explored, looking for a simplification of the overall analytical process without losing the excellent analytical characteristics obtained in both approaches. The results are promising and indicative of the progress of analytical microfluidics toward the "plateau of productivity" and the routine laboratory application.