Cereal Chem 52:561 - 576.  |  VIEW ARTICLE
Application of Near Infrared Reflectance Spectroscopy to Analysis of Cereal Grains and Oilseeds.

P. C. Williams. Copyright 1975 by the American Association of Cereal Chemists, Inc. 

Infrared reflectance spectroscopy (IRS) was first introduced in 1971 for rapid on-the-spot analyses of soybeans for oil, protein, and moisture. The equipment utilizes the principle of infrared reflectance spectroscopy, together with a unique multiple-wavelength filter system, which facilitates scanning of the near infrared area of the spectrum. Calibration against known standards is necessary. IRS equipment has been calibrated for analyses of a variety of materials, including hard red spring wheat (HRS), hard and soft wheat flours, vital wheat gluten, barley, oats, rapeseed, and soybeans. Calibration depends upon statistical principles, using either a multiple regression or simultaneous equation-type program. In either case, computer facilities are necessary. Accuracy of analyses, as measured by the standard error of estimate, is about +/ 0.22% protein and +/-0.16% moisture in the case of red spring wheat, giving a coefficient of variability of 1.5% in each case. Accuracy of analyses of other cereals, oilseeds, and legumes for oil, protein, and moisture varies, but coefficients of variability are usually between 1-5%. Analytical precision as measured by check sample analyses is comparable with that of standard laboratory methods. Technical difficulties which may be encountered with calibration and operation of equipment are discussed. The most important factor influencing the accuracy and precision of analysis with IRS equipment is mean particle size of ground sample presented to the instrument, which is, in turn, influenced mainly by method of grinding. Other factors which are discussed include mixing of sample, type of grain, constitution of grain, growing environment of grain, including seasonal effect, and factors associated with instrument. Method of sample preparation is of fundamental importance to successful use of IRS instrumentation. Both accuracy and precision of analyses may be affected by method of sample preparation and also by method of presentation to instrument. Errors of over 2% protein have been encountered in analyses of HRS wheat when the instrument was used to analyze a sample which had been prepared by a system for which instrument had not been calibrated. Method of grinding a sample of grain affects both mean particle size of sample and type of particles produced. This, in turn, influences penetration and reflectance of light from sample presented to sensing unit. Extent of packaging inside presentation cell may also influence analytical results.