DETERMINATION OF APRICOT FRUIT QUALITY USING FOURIER TRANSFORM INFRARED SPECTROSCOPY

Abstract

This work reports the rapid determination of apricot fruit quality traits using Fourier Transform near (NIR) and mid (MIR) infrared reflectance spectroscopy. The FT-NIR (800-2500 nm) spectra were acquired on intact fruits using an integrating sphere (diffuse reflectance). The FT-MIR spectra were acquired on fruit slurries using a horizontal ATR cell (Attenuated Total Reflectance). Relationships between spectral data and quality attributes obtained by traditional methods are evaluated by the application of chemometric techniques based on partial least squares (PLS) on fruit set divided randomly into two groups: 598 fruits for calibration and 279 for validation. The non-destructive method using NIR spectroscopy allows the prediction of soluble solids content and titratable acidity (correlation coefficient of 0.92 and 0.89, respectively, and root mean square error of prediction of 0.98% Brix and 3.62 meq.100g -1 FW, respectively). The ATR-FTIR performed on apricot slurries is well-adapted for the prediction of the individual compounds: sucrose, glucose, malic and citric acids (correlation coefficients greater than or equal to 0.92 and errors smaller than 16%). Prediction model for fructose, the minor compound in apricot, showed the lowest performance with r = 0.86 and error of validation of 18%. Moreover, the technique that consists in acquiring spectra and performing PCA on spectra allows the discrimination of apricot sets according to their ripening stages and their genetic characteristics.