Rapid and non-destructive determination of vitamin C and antioxidant activity of intact red chilies using visible near-infrared spectroscopy and machine learning tools

Abstract

Red chili peppers are extensively used in the culinary industry due to their high vitamin C, antioxidant content, spice, and natural colorant properties. Traditional methods for determining these parameters are time-consuming and unstable. This study investigates the viability of visible near-infrared spectroscopy as a rapid and nondestructive method for determining vitamin C content and antioxidant activity in intact red chilies. Visible near-infrared spectroscopy measures the sample's light absorption, proportional to its chemical composition. Collecting intact red chili samples from various sources and determining their vitamin C content and antioxidant activity using conventional methods were required for the study. The visible near-infrared spectra of the pieces were collected using a spectrometer, and chemometric models were devised to correlate the spectra with the vitamin C content and antioxidant activity. Traditional methods were compared to visible near-infrared spectroscopy regarding their performance of prediction. In conjunction with chemometric models and machine learning algorithms, visible near-infrared spectroscopy could accurately predict red chilies' vitamin C content and antioxidant activity. The developed models exhibited high calibration and prediction coefficients, low root mean square errors, and high prediction-to-deviation ratios. The identified absorption peaks in the visible near-infrared spectra were related to the samples' color pigments and water content. This study demonstrates the feasibility of visible near-infrared spectroscopy as a rapid and nondestructive method for determining the vitamin C content and antioxidant activity of intact red chilies. The findings may have significant ramifications for the food industry, providing a more efficient and secure quality control and nutritional labeling method. It is recommended that additional research and validation be conducted to ensure the applicability of the developed models to various red chili peppers varieties and growing conditions.

Keywords

• Chemometrics tools
• Machine learning
• Red chilies
• Spectroscopy