Background: To facilitate faster phenotyping of onions (Allium cepa L.), Fourier-transform mid infrared (FT-MIR) spectroscopy with partial least squares (PLS) regression modelling was evaluated for the determination of pungency (pyruvate), sweetness (free sugars) and fructan in juice samples (n = 605) expressed from bulbs from breeding populations.
Results: Fourier-transform infrared (FTIR) spectra (range 1700-900 cm-1 ) were obtained from droplets (30 μL) of unprocessed juice. Goodness-of-fit (r2 ) and prediction errors (standard error of cross validation) for optimal PLS models were: soluble solids (0.997, 0.1 °Brix), pyruvate [0.825, 0.8 μmol g-1 fresh weight (FW)], fructan (0.98, 1.9 mg g-1 FW), glucose (0.941, 1.1 mg g-1 FW), fructose (0.967, 1.0 mg g-1 FW) and sucrose (0.919, 1.7 mg g-1 FW). FTIR models for industry sweetness indices based on glucose or sucrose equivalents were also developed. Because of its very low concentration (0.8-12 μmol g-1 FW) relative to other compounds, pyruvate was the weakest model developed. Fructan could be determined spectroscopically without the need for enzymatic digestion.
Conclusions: All of the chemometric models developed are acceptable for screening purposes. Those for soluble solids, fructan and fructose are also suitable for routine analysis. FT-MIR can therefore be utilised for the simultaneous determination of pungency, sweetness and fructan in this crop. © 2018 Society of Chemical Industry.
Keywords: Allium cepa; Fourier-transform infrared; composition; fructo-oligosaccharide; multivariate analysis; sweetness index.
© 2018 Society of Chemical Industry.