R-200.21.182 [2016-03] Protein Content in Malt – NIR

Reflectance spectroscopy in the near infrared spectral range (NIR) is employed as a physical method for the determination of substances found in grain. This method is based on the fact that different substances, e.g., proteins, in wavelengths ranging from approximately 800 to 2500 nm possess characteristic absorbance and reflection spectra.

Light of a defined wavelength is directed towards the (finely ground) sample in a measurement chamber and is reflected diffusely back to a detector, which measures the intensity of the reflected light. On the basis of the absorbance measured, an integrated computer calculates the protein content of the sample. In order to be able to perform these calculations, the NIR spectrometer must be specifically calibrated for each product (e.g., barley, wheat, barley malt, wheat malt) as well as for each constituent (in this case, protein). The calibration is carried out with data gathered from the absorbance of spectra from reference substances. The protein content of individual samples is determined chemically using a reference method, for example, with the Kjeldahl method and are then correlated with the respective spectral data. Through comparison of the spectra from the analysis sample with the reference sample used in the calibration, the protein content is determined. Since seasonal differences in barley kernels from year to year make it necessary to modify the spectral data so that they match the chemically determined data, calibration must be monitored and adjusted accordingly. Normally, the manufacturer supplies a standard calibration reference with the device.

The primary advantages offered by this method are the speed with which samples are analyzed (< 1 min per sample) and the fact that no chemicals are necessary. Grinding the sample at a defined setting prior to the analysis may be required; however, it is not necessary to weigh the sample. With the latest generation of NIR devices, samples no longer need to be ground. The accuracy of the measured values is dependent upon the quality of the calibration. If analysis is performed carefully, standard deviations of 0.1 % are attainable.