Optical spectroscopy covers wavelengths from 200 nm to 25 µm (see Figure 1) and is divided into three important spectral ranges:

• Ultraviolet/visible: 200 – 760 nm
• Near Infrared : 760 – 2500 nm
• Mid-Infrared : 2500 nm – 25 µm

Photons in the Ultraviolet/Visible spectral range have enough energy to excite or ionize materials by raising the energy level of bound electrons. NIR radiation has less energy/photon but does excite molecular vibrations.

The interaction of the electromagnetic radiation with the sample can be described by the following effects:

• Absorbance/Reflection (Figure 2a)
  The absorbance spectra are used for the qualitative and quantitative  determination of composition.
• Surface effects (Figure 2b)
  The evaluation of surface effects determine physical characteristics such as  roughness and grain size.
• Interface effects (Figure 2c)
  Interference patterns in returned spectra can be used to determine thickness of thin layers.
   
  

Chemometrics is the application of statistical methods (e.g., principal components analysis or partial least squares) to extract information from chemical or spectroscopic data. There are quali tative and quantitative methods for conducting multi-component analysis. Quantification, for example linking sensor signal with substance concentration, is possible even without specific interactions.

The wavelength of light employed in NIR spectroscopy excites vibrations of covalent mole cular bonds. Therefore, it is suitable for the determination of water content in food and agrochemical products (renewable raw materials, feedstuff, corn, milk). Organic and pharmaceutical products can be ex amined with respect to their protein (N-H bonds) or fat content (C-H bonds). In addition, various molecular structures and groups can be detected in polymers. An excellent example is given by the carboxylic groups (COOH). For these reasons, the method is commonly used in the chemical, pharmaceutical, and food industry for quality assurance and process control.