Establishment of a quantitative GC-MS method for acrylamide detection and in situ kinetic study of acrylamide formation in fried potato slices.

This study established a reliable quantification technique (acrylamide derivatization and GC-MS) for acrylamide detection due to the high accuracy, reproducibility, sensitivity, and good stability. Specifically, the pre-treatment conditions of extraction solvents and defatting methods were investigated and optimized. The acrylamide formation and quality indicators in fried potato slices at different temperatures (120-220 °C) for varying durations (0-40 min) were further measured. The significant correlations obtained between color parameters, moisture content, and acrylamide contents were observed. Moreover, in situ kinetic model coupled the formation, mass transfer, and elimination process of acrylamide in fried potato slices was constructed. Finally, three machine learning algorithms (BP Neural Network, Support Vector Machines (SVM), and Random Forest Models) were employed to predict 7 quality indicators (L*, △E, chroma, hue angle, water activity, moisture content, and weight loss rate) based on frying parameters with Random Forest demonstrating superior performance in most output variables (R2 > 0.85).