Description
Two innovative mild solar drying processes on slices of white onion and summer melon were carried out. During the whole process, NIR spectra were collected in reflectance mode, between 900 and 1700 nm, using the MicroNIR OnSite-W (VIAVI Solutions S.r.l., Italy) portable spectrometer. Spectra were acquired continuously throughout the process by placing the thermally insulated NIR probe over a sample slice. Samplings were performed every 90 minutes for the determination of: weight; moisture content (MC); water activity (aw). The drying process was stopped when the samples reached a constant weight. During the process, data for the external and internal air temperature and relative humidity were recorded as well. MC and aw of processed onion varied from 90.3 to 16.3% and from 0.9975 to 0.4552, respectively. Summer melon aw varied from 0.9994 to 0.4666. Partial least square regressions showed good prediction ability for the determined parameters. Onion MC: R2CV = 0.94; RMSECV=6.4%; onion aw: R2CV = 0.95; RMSECV=0.05; summer melon aw: R2CV = 0.97; RMSECV=0.04. The PLS models were then used to predict the parameters profiles along the process. Generally, the predicted parameters resulting from the NIR spectra provided a good estimation of the data measured offline, describing the trend of these parameters during the whole process. These preliminary results were useful for monitoring the progress of the process. Furthermore, real-time prediction of aw provided a useful tool for defining the end of the dehydration process and suggesting the potential safety of the final product. Given the RMSECV, the MC model should be improved by adding samples with MC <20%. The MC calibrations for the summer melon samples are in progress.
