How does NIR work?
NIR instruments can be divided into two groups: filtered-based and scan-based. Although both groups are based on the same theory, they differ in frequency and bandwidth. Before I discuss the differences, let’s take a look at the inside of the instruments. First you need an illumination source and a sensor. In most instruments, a halogen-tungsten bulb is used for illumination because the range is visible and reaches into the NIR region. The sensor is the most important part of the instrument. It detects the full desired range, but must be sensitive enough to detect subtle differences in the amount of reflected or transmitted radiation. The latter is used to relate the concentration of a compound, so we can determine the amount present (e.g. 14% protein).
At this point we need a filter to sort the various wavelengths. A filter is placed in the path of the illuminating energy, which restricts the range of the radiation to just a few wavelengths. This range falls on the sensor, which generates an electrical current proportional to the amount of energy it received. Early instruments had as few as six filters or ranges, while current models have up to 20.
The only difference between filter-based instruments and scan-based instruments is that scan-based use a dispersing material, like a prism, to produce much larger, narrow sets of ranges. The most popular dispersing material is called a grating. The way the grating is implemented determines the resolution of the instrument. There are other methods to disperse energy, but these are the most common.
The reference is another important part that you may or may not see, but is in all instruments. It is a highly reflective material that is used to set the maximum energy levels that the sensors will see. This material must be equally reflective over the full range of the sensor, as it becomes the base to which all other measurements are compared. The most common reference is Spectralon, which is about 97% reflective over the entire NIR range.
THE FINAL RESULT. Let’s put this all together by following the path of the illuminate. The radiated energy is directed to illuminate the sample. The energy interacts with the sample and is reflected (or transmitted for NIT instruments). Various optics control the path of the illumination and reflected energy. It is up to the designer how this is accomplished, but usually mirrors or fiber optics move the reflected light through the filters or to the dispersing material. For a filter-based system, the filters are rotated through the energy path and the sensor takes a reading in synchrony. For the scanning instruments the grating is usually rotated. You could rotate the sensor; it also takes readings in synchrony. The process is the same whether a sample or reference is used, but at some point the reference must be scanned.
The sample is illuminated and the sensor determines how much energy is reflected after passing through the various filters. It produces a valued based on the relationship of the amount of energy and the wavelengths. In the Spring 2006 issue of the AIB Quarterly, I’ll explain this relationship in more detail, as it that allows NIR instruments to calculate a meaningful value. AIB
The author is Director, Product and Technology Development, AIB.