NIR Reflectance Spectroscopy Analyzes Grain
Brent D. Johnson
Advances in genetics and biotechnology have made it possible to manipulate the chemical constituents of poultry and swine feed to increase the value of a unit of grain. As a result, the traditional view of grain as a static commodity is giving way to that of a configurable product, and photonics is finding a place in the analysis of these products.
Over the last few years, scientists at the Iowa Grain Quality Initiative at Iowa State University in Ames have been investigating the use of near-infrared reflectance spectroscopy to study grain. Such a noncontact approach is crucial for this application because many of the samples are fatty or greasy. The scientists are working to make the technique more accurate for the measurement of compositional factors such as moisture, protein, oil, starch, hardness or texture and, in the case of corn, fermentability for the production of ethanol for oxygenated automotive fuel.
Researchers at Iowa State University are using near-infrared reflectance spectroscopy to measure the compositional factors of grain. Lab technician Leah Schultz examines a sample of corn with the Perten Instruments DA 7200 analysis system. Courtesy of Charles R. Hurburgh.
Charles R. Hurburgh, who is in charge of the program, has been using a DA 7200 analysis system from Perten Instruments AB of Huddinge, Sweden, since the last harvest. The dual-beam system allows him and his colleagues to look at moving samples of products such as feeds and grains.
The instrument has a spectral range of 950 to 1650 nm and a nominal resolution of 6 nm, which seem exceptional for those accustomed to Fourier transform techniques but which are common values for near-IR reflectance spectroscopy. It displays a shift in spectral measurement of less than 0.1 nm over two years. The spot size is 4 cm2, so the analyzer provides a viewing area of 180 cm2 over the entire field of rotation. And unlike monochromator-based systems in which the grating is on a pivot, the DA 7200 features a stationary grating so that the light spreads directly onto a circular, 256-pixel InGaAs array, eliminating misalignment.
The analyzer has a 30- to 100-ms integration time and takes 30 to 80 spectra per second. One rotation of the sample cup takes three seconds. To illustrate the speed of the technique, approximately 300 spectra must be analyzed to characterize a sample of corn.
The DA 7200 is performing well in the studies, Hurburgh said. It displays precise spectra and accurate calibration, reducing random error. Moreover, it measures whole grain as well as intermediate products such as mash and slurry, and it is configured so that multiple products can be measured. Other instruments may require that the material flow or that it be packed, which is not good for sticky powder or slurries.
He said that the biggest limitation of the technique is the need for the manufacturer to supply calibrations. A user typically does not want to be concerned with performing chemometrics and producing reference values. Hurburgh's current goal is to develop a calibration database for corn and soybeans. The researchers will perform field tests this fall, and they expect to have a complete database after two more growing seasons.
MORE FROM PHOTONICS MEDIA