Reconsidering the Need for Empirical Alignment and Wavelength Calibration Steps in the Building of a Dispersive NIR Spectrometer with an Application for Ethanol Quantification Using a Polymer Filament 3D Printer.

The construction of a dispersive optical spectrometer using three-dimensional (3D) design software and printing, without applying any optical adjustments, its validation, and application to quantification of ethanol in multiproduct liquids, is the objective of this work. A 3D design software was used to design a near-infrared (NIR) spectrometer in the region from 800 to 1600 nm from the dimensions of commercially available optical components. The project was printed on a polymer filament 3D printer, and the components were fitted to the printed part. Software calculations using the model design parameters were applied to attribute the wavelength values to the abscissa axis in the spectra and estimate errors due to 3D printing limitations. The alignment of the spectrum was proven using the chloroform absorbance spectrum, which presented a maximum mispositioning of 4.1 nm concerning the literature data and effective bandwidths equivalent to commercial instruments. The 3D-printed instrument was applied to quantify ethanol in samples of cachaça, rum, beer, brandy, whiskey, vodka, mouth wash, alcohol gel, and commercial alcohol solutions. Partial least-squares regression models were built for the 3D-printed instrument and two commercial NIR instruments, the MPA II (Bruker) and the NIR DLP NIRscan (Texas Instruments), using a group of 180 standards. The three instruments reached excellent predictive capability with similar root mean square error of cross-validation (2.36-2.68) and prediction (2.31-2.87). The correlation coefficient of cross-validation and prediction for all models were between 0.97 and 0.98. The results show the feasibility of building a 3D-printed dispersive spectrometer ready for application with the simple docking of the optics, presenting acceptable accuracy to the project design concerning the printing limitations.

FT-NIR and linear discriminant analysis to classify chickpea seeds produced with harvest aid chemicals.

Spectroscopy and machine learning (ML) algorithms have provided significant advances to the modern food industry. Instruments focusing on near-infrared spectroscopy allow obtaining information about seed and grain chemical composition, which can be related to changes caused by field pesticides. We investigated the potential of FT-NIR spectroscopy combined with Linear Discriminant Analysis (LDA) to discriminate chickpea seeds produced using different desiccant herbicides at harvest anticipation. Five herbicides applied at three moments of the plant reproductive stage were utilized. The NIR spectra obtained from individual seeds were used to build ML models based on LDA algorithm. The models developed to identify the herbicide and the plant phenological stage at which it was applied reached 94% in the independent validation set. Thus, the LDA models developed using near-infrared spectral data provided to be efficient, quick, non-destructive, and accurate to identify differences between seeds due to pre-harvest herbicides application.