NIR spectroscopy is a method developed not long ago for the rapid analysis of agricultural products by the USDA labs. Its main advantage lies in the fact that signals are much weaker that in classical IR spectroscopy. This makes the presentation of samples much easier by eliminating the need for laborous preparations. On the other hand these overtones and combination bands are not only much more diffuse and overlapping but also influenced by physical effects such as particle size and light scattering within the sample.
For the evaluation and analysis therefore first spectra pretreatments are necessary to eliminate these influences. Further the significant information lies within spectra consisting of a very high number (sometimes several thousand) of data points. Extraction mathods – called chemometrics are necessary which have been developed under the influence of NIR spectroscopy.
NIR analysis is only useful for certain applications, with practical measurement ranges depending on the sample material. We will provide advice by processing just a limited number of spectra – independent of the supplier of the instruments – which analytical accuracy can be expected when modelling is done correctly.