“The benefits of NIR over traditional testing procedures lie in the speed of analysis, the simplicity, the ease of operation and the ability to measure several components simultaneously. NIR technology lends itself to in line measurements due to the ruggedness of the hardware and the fact that analysers can operate unattended for long periods with little or no need for maintenance.”
Near Infrared Spectroscopy:
In the mid 1960’s, an agricultural engineer, Karl Norris, USDA, Beltsville, Maryland, USA, was given the challenge to develop a rapid means of measuring protein, oil and moisture in soybeans. Mr. Norris is credited as being the father of NIR because he was the first person to use multiple linear regression analysis to develop mathematical models to relate the Near Infrared Reflectance spectra ground soybeans to the protein, moisture and oil concentrations. In the last 50 odd years, Near Infrared Spectroscopy has evolved to become a mainstay measurement tool for many food and agricultural products. Almost all of the world grain is traded based on the use of NIR analysers to rapidly measure protein and moisture in wheat, corn, barley, rice, pulses, oats, rye, sorghum and protein, oil and moisture in oil seeds such as canola, soybeans, sunflower, safflower, linseeds and others.
In the flour milling industry it is important to be able to measure moisture in tempered wheat before it is milled. Generally wheat is soaked in water for 8 to 10 hours to bring the moisture content up to between 15 and 17%. In line measurement and control of moisture in tempered wheat is difficult. The issues include the uneven distribution of water over the grain and the flow of grain through a sampling device. This article discusses the use of the CropScan 3000S In Line Grain Analyser to measure moisture in tempered wheat in a flour mill in Tamworth, NSW, Australia.
Figure 1. shows a schematic of the basic optics that is used in NIR analysers to generate spectra of flour and whole wheat grains. Light from a tungsten halogen lamp passes through a sample of either flour or whole grains. Energy is absorbed at specific frequencies by the chemical bonds in protein (N-H), water (O-H), oil (C-H) and carbohydrates (C-O-H). An optical element called a Diffraction Grating separates the light into individual frequencies from 720-1100nm. The dispersed light is detected by a silicon photodiode array detector, similar to that used in a flat bed scanner or a laser printer. The amount of light absorbed by the sample at the resonant frequencies that correspond to protein, moisture, oil and carbohydrate is proportional to the concentration of each component in the sample. Figure 2. shows the NIT spectra of wheat, durum, flour and semolina.
Applications for NIR Spectroscopy in the Flour and Wheat Milling Industry:
The starting point for the production of flour and semolina is wheat and durum. Typically NIR analysers measure the protein and moisture in whole grains of wheat and durum as basic trading parameters, however for millers the measurement of gluten, falling number, hardness, sedimentation and Zeleny are also important. Gluten is the major group of protein found in wheat. Gluten consists mostly of two protein, i.e., Glutenins and Gliadin. The question is whether the NIR spectra contains specific information that allows the analyser to measure gluten as part of the total protein compliment. The answer is no. The spectra of wheat grains are made up of a few broad absorption bands which are related to N-H bonds that are common to all proteins. The correlation between gluten and total protein is generally high and as such, the measurement of gluten using NIR is simply a proportional relationship to the total protein. Nonetheless, calibrations for gluten are commonly found in NIR analysers supplied to the flour milling industry. Likewise the physical parameters such as falling number, hardness, sedimentation and Zeleny are not directly measured from the NIR spectral but moreover indirect correlations to one or several of the chemical parameters, i.e., protein, moisture, oil and carbohydrates. The accuracy of NIR to measure these components in whole wheat grains is often poor. Grinding the grains into a powder and collecting the Near Infrared Reflectance spectra from 1000 to 2500nm, will improve the accuracy. Particle size of the powder is commonly correlated with measurements such as hardness, sedimentation and Zeleny and thereby explains why NIR Reflectance analysers are better for performing these measurements over NIR Transmission analysers.
In Line NIR Measurements in the Flour Milling Industry
NIR technology lends itself to in line measurements due to the ruggedness of the hardware and the fact that analysers can operate unattended for long periods with little or no need for maintenance. There are many locations within a flour mill to place NIR In Line analysers, i.e., incoming grain, tempering, milling streams and final products. Typically protein and moisture are the major parameters measured in line, however other quality and physical parameters can also be measured.
Milling of wheat to make flour requires the grain to be tempered, i.e. soaked in water and allowed to absorb the water for several hours before milling. Tempering can be achieved by several means. A flour mill in Tamworth, NSW, Australia employs a rotary drum mixer whereby the grain is sprayed with water. Following the mixer, the grain is conveyed up belt and dropped into a tank where it is held for several hours.
Figure 2 shows the installation of a special sampling head for the CropScan 3000S In Line Analyser as mounted on a rotary mixer used for tempering wheat prior to milling. The objective is to provide a feedback control of the amount of water sprayed on the wheat in order to get constant 15% moisture content.
The use of in line NIR analysers in the flour milling industry is becoming more popular. The benefits of NIR over traditional testing procedures lie in the speed of analysis, the simplicity, the ease of operation and the ability to measure several components simultaneously. The benefits of in line NIR lie in the ability to make measurements in real time and then to make decision before it is too late to do anything about it.