Sequential injection ATR-FTIR spectroscopy for taste analysis in tomato

Abstract The potential of attenuated total reflectance Fourier transform infrared spectroscopy (ATR-FTIR) for food analysis has been demonstrated in many publications. Most applications, however, deal with the major disadvantage of extensive sample manipulation and cleaning of the ATR sampling compartment, which greatly reduces the sampling rate. To address this problem, in this paper, the potential of sequential injection analysis (SIA) in combination with ATR-FTIR has been studied for sugar and organic acid analysis of Belgian tomato samples. SIA allows automation of the sampling and measurement protocol. The SIA-ATR-FTIR system is optimized for measurement performance accuracy using a central composite design. The four important operational parameters studied are spectral resolution, number of co-added scans, flow rate, and sample temperature. A clear optimum is found for the number of scans (256 co-added scans) and resolution (16 cm −1 ) while sample temperature and flow rate are found not to significantly influence the measurement performance on the most important taste components in fruit. Hence, a stopped flow analysis – to save sample – at room temperature with 256 co-added scans at a resolution of 16 cm −1 is selected as optimal for tomato juice measurements. This setup results in a sample throughput of 24 samples/h. Different tomato cultivars were discriminated based on their acid and sugar content, which are the most important chemical components determining tomato taste. Prediction models for d -glucose, d -fructose, citric acid and l -glutamic acid concentrations in tomato samples were successful for the measured tomato cultivars with RMSEP-values of 1.81 g L −1 , 1.63 g L −1 , 0.41 g L −1 and 0.35 g L −1 respectively. However, the prediction model for l -malic acid, which is present in very small concentrations in tomato fruit, was not successful.