Innovative 96-Well Plate Imaging for Quantifying Hydrogen Peroxide in Cow’s Milk: A Practical Teaching Tool for Analytical Chemistry

Abstract

Initially, students performed a qualitative analysis to detect the presence of oxidizing compounds in cow’s milk samples. This was achieved by observing the oxidation of potassium iodide (KI) by hydrogen peroxide (H₂O₂). Subsequently, the concentrations of H₂O₂ in the milk samples were quantified using images of 96-well plates captured with a flatbed scanner. This method is straightforward, efficient, and ideal for high-throughput analysis. The RGB values from the 96-well plates were automatically extracted using the ImageJ plugin, ReadPlate. For the quantitative analysis, students explored various figures of merit, including the limit of detection (LOD), limit of quantification (LOQ), linearity, sensitivity, and recovery. Milk samples were spiked with H₂O₂ at three concentrations (0.03%, 0.06%, and 0.09%), and the measured concentrations in these spiked samples were compared to evaluate interclass repeatability using one-way ANOVA. Post hoc tests, including Games-Howell and Tukey, were used to identify significant differences between concentrations across the classes. Before conducting the one-way ANOVA, students assessed data normality using Q-Q plots, the Shapiro–Wilk test, and the Anderson–Darling test. They also evaluated the homogeneity of variance with Levene’s and Bartlett’s tests. Additionally, a two-way ANOVA was employed to analyze the effects of spiking concentrations and laboratory classes on the recovery. This analysis revealed a significant interaction between spiking concentrations and laboratory classes on recovery. All statistical tests were conducted using accessible and user-friendly freeware, providing students with practical experience in both data analysis and statistical interpretation.