Near-infrared reflectance spectroscopy model predictive of cadmium concentration in peanut kernels

• Published in: Journal of food measurement & characterization 2023-12, Vol.17 (6), p.5730-5735
• Main Authors: Gao, Wei Wei, Yang, Zhen, Jiang, Chun Jiao, Sun, Hao Jie, Wang, Zhi Wei, Zhang, Ming Jun, Wei, Ying, Wang, Chuan Tang
• Format: Article
• Language: English
• Subjects: Cadmium; Chemistry; Chemistry and Materials Science; Chemistry/Food Science; Engineering; Food Science; Inductively coupled plasma mass spectrometry; Infrared reflection; Infrared spectra; Infrared spectroscopy; Kernels; Mass spectrometry; Mass spectroscopy; Mean square errors; Near infrared radiation; Original Paper; Peanuts; Prediction models; Reflectance; Root-mean-square errors
• ISSN: 2193-4126; 2193-4134
• DOI: 10.1007/s11694-023-02064-7

Peanut is a major food crop prone to enrichment in toxic cadmium (Cd). To control Cd levels in peanut, this study aimed to develop a near infrared spectroscopy (NIRS) model for determining Cd concentration in peanut kernels as an alternative to traditional wet chemistry techniques, which can be expensive, time-consuming, and destructive. Near-infrared (NIR) diffuse reflectance spectra of 110 bulk peanut kernel samples were collected, and Cd concentration of the kernel samples was determined by inductively coupled plasma mass spectrometry. A robust quantitative NIRS prediction model for Cd concentration in peanut kernels was developed for the first time. In the calibration set, the best model had a high coefficient of determination (R cal 2  = 0.9194) and a low root mean square error of cross-validation (RMSECV = 0.0388). In the prediction set of 105 additional peanut kernel samples not involved in the model development, the coefficient of determination (R p 2 ) was as high as 0.9539 and the root mean square error of prediction was as low as 0.0341. This study provides a rapid and low-cost screening tool for low-Cd breeding and Cd management in peanut.