No abstract

No abstract

The aim of this work was to compare two near infrared spectrometers (960-1650 nm) to assess coffee matrices for real-time characterization during processing. A benchtop spectrophotometer built for process measurement, equipped with a rotating acquisition system suitable for in-line analyses (device 1), and a portable spectrophotometer for at-line analyses (device 2) were tested. The experimentation was conducted on green, roasted bean, and ground coffee, relating to three coffee blends, employing samples collected at different steps from the process. A total of 399 (247 green, 76 roasted, and 76 ground coffee samples) and 229 (142 green, 43 roasted and 44 ground coffee samples) samples were analyzed using device 1 and device 2, respectively. Principal component analysis was applied to the spectra of the coffee samples to evaluate the feasibility of realtime characterization for each matrix type, during the different steps of the process. Partial least square regression analysis was applied to develop calibrations to predict moisture content, tap density, and powder granulometry on ground coffee and to estimate moisture on roasted bean with a view of real-time measurements. The Passing and Bablok regression method and Bland-Altman analysis were performed to compare results obtained from the tested devices. Considering the models built on different datasets were based on different blends, quality parameters, and matrices, the results obtained from partial least square models, in cross-validation, gave R 2 values between 0.21 and 0.98 and between 0.18 and 0.84 for devices 1 and 2, respectively. However, considering the comparison of model results derived from the two devices, no significant differences are noticeable for the most part of the dataset analyzed. Hence, based on the strategy to be undertaken by coffee industry operators, a future real-scale application could be envisaged directly in-line using device 1 or at-line using device 2.

The application of optical techniques may help MDC buyers to monitor the quality of postharvest products, leading to an effective optimization of the entire supply chain. © 2017 Society of Chemical Industry.

No abstract