 |
|
| History Inline NIRS spectroscopie |
||||||
|
In 1800 the Englishman William Herschel discovered the infra red region of the light, by accident. He tried to measure, what each spectral colour of the sun light contributes in heating a body exposed to sun light. He used a glass prism to produce a spectra, while he used a therometer to measure the temperature for each colour. The sum of all temperature increase however was much lower than the temperature caused by the entire sun light. By accident the thermometer moved outside the spectra, just next to the red band. Suddenly he saw a dramatically increase in temperature. He postulated, that there exists an invisible band of light beyond red. He called this band beyond red = latin infra red. This band starts at 700nm. The band up to 2500nm is called the near infra red band = NIR band. Until the end of the 1960s all IR measurements have been made in the band above 2800nm, because the fundamental bands of the organic compounds occurred at 2800nm to 50.000nm. The NIR band consist only of overtone bands and combination bands. End of the 1960s, however the the potential of NIR measurements for agriculture products was discovered. NIR instruments In general a NIR instrument works as follows: - The product is illuminated with a white lamp Until a few years ago the main technologies for NIR instruments were filter instruments or spectrometers where the spectra of the received light is produced by a moving diffraction grating. A filter instrument needs for each constituent to be measured an optical filter to be installed in the instrument. This should be an optical filter narrow banded around the spectral line of the constituent. To measure several constituents it is also necessary to have several filters in the instrument. While measuring, this filters have to be moved into the received light beam. This usually is done by a filter wheel. This is a relatively complex mechanism, as well as it needs maintenance. At spectrometers with a moving grating the light is divided into a spectra. The number of spectral lines depends on the design (resolution) of diffraction grating. This method has the advantage, that it is possible to measure all constituents of the product, which contribute to the measured spectra. Diode array spectrometers With upcoming semiconductor light detectors and the rapid miniaturisation of these detectors it was possible to manufacture detector arrays with up to 1024 independent light detectors at on chip. Now it is possible to use a fixed diffraction grating and measure each spectral line with an own detector. This has two main advantages: - No moving parts. Until now this instruments are mainly used in laboratory or scientific applications, as the instruments are not rugged enough for industrial use and they also are relatively expensive. Simplified basics of spectral analysis The goal of spectral analysis is to calculate, out of the measured spectra, the amount (concentration) of the unknown constituents of interest in a sample. As with all quantitative analysis it is assumed, that the measuring values (the spectra) are somehow related to the concentration of the constituents of interest in the sample. The job is to create a calibration equation, which when applied to unknown data, can predict the concentration of the constituents of interest. This calibration equation is also called the calibration model. One measures a set of spectra with known concentrations of the constituents of interest. The calibration model is applied to these spectra’s and delivers (hopefully) a set of constants which allows the prediction of the constituents of interest in an unknown spectra. With other words: The calibration procedure delivers a set of constants for the calibration equation, which when applied to an unknown spectra allows the calculation of the concentration of the constituents of interest. The foundation for calibrating a spectrometer is the Beer-Lambert-Law: A(l) = E(l) * D * C A(l): Absorption at wave length l This law shows a simple linear relationship between the absorption and the concentration of a constituent at a given wave length. In other words: When moving a sample into the beam of a spectrometer, there is a direct and linear relationship between the concentration of the constituents and the absorbed light. The Beer-Lambert-Law also is additive: A(l) = E1(l)*D*C1 + E2(l)*D*C2 + …. + En(l)*D*C2 I.e. The absorption at a certain wave length is linear related to the concentration of all constituents of a sample. Therefore it is theoretically possible to calculate the concentration of all constituents in a sample. One only has to solve an equation system of n-equations and n-unknowns, what with today computing power is no problem. The Dosco spectrometer Dosco currently let develop a spectrometer for industrial on-line applications. The goal is to develop a rugged instrument for a reasonable price. This first instrument will operate in a wave length band of 600nm to 1100mn. Until end of 2006 the wave length range will be extended to 1700nm and 2500nm. The instrument is designed to measure simultaneously up to four constituents of interest. For each constituent is a separate 0/4-20mA output available.
Application: Instrument: Sensors:
|
|
|||||