Espectroscopia de infravermelho próximo (NIR) em análises bromatológicas: uma revisão bibliográfica

Abstract

Spectroscopies are among the most widely used analytical methods in the world, serving as a reference for bromatological analyses, i.e., those aimed at determining the chemical composition of food, nutritional value, calories, physical, chemical, toxicological properties, among others. The objective of this work is to conduct a literature review on the use of near-infrared (NIR) spectroscopy in these analyses, highlighting its advantages and disadvantages. Based on the results found on Google Scholar and Capes Journals platforms, it is observed that NIR can be applied to forages, seeds, grains, pods, and silages, where its results indicate whether the sample is fit for consumption or not, in addition to determining parameters such as crude protein content, nitrogen content, dry fiber, moisture content, total ash, crude fiber, and crude fat. The use of chemometrics is essential for quantification via NIR, as NIR spectra lack selectivity, thus multivariate calibration methods such as principal component regression (PCR) and partial least squares regression (PLS) are crucial for the quantification of substances and parameters. To perform multivariate calibration, it is also necessary that reference analyses be used to correlate the spectra (NIR) with the property values found by the reference method. Considering the practicality, cost, and high analytical frequency of the technique, the importance of NIR spectroscopy in bromatological analyses is evident.