Feasibility of Predicting Ease of Peeling of Tomato Fruits by Using a Handheld Infrared Spectrometer

Abstract

The potential of infrared (IR) handheld spectrometers, paired with chemometrics, to predict peelability of processing tomatoes was examined. A pilot plant scale study was carried out where tomato fruits underwent lye treatment and peeled with rotating rubber disks. The degree of peeling was regressed on the infrared spectral data collected from various regions on the fruits using partial least squares regression procedure. Peelability was best explained by spectra collected from areas adjacent to the blossom and stem ends. The regression model exhibited residuals with nonhomogeneous variance and was accurate only in predicting “very hard to peel” values, using IR frequencies associated with phenolics in the cuticular domain. The use of a handheld infrared spectrometer paired with chemometrics showed promise for predicting ease of peel of processing tomatoes. There is a need to improve spectral quality through better sample contact with the attenuated total reflectance accessory for predicting intermediate-peeling tomatoes. Practical Applications The tomato industry incurs in significant losses during the peeling stage, accounting for up to 28% of the waste generated in commercial canneries. Current technology used to assess tomato maturity (color, soluble solids, titratable acidity) has not correlated to tomato-peeling losses. The use of a handheld IR spectrometer paired with chemometrics has shown potential as a tool for predicting peelability of processing tomatoes. We have shown that infrared spectra reflect the peel's chemical state, which in turn could be related to the fruit's developmental stage.