Metoda analizy widm mierzonych z wykorzystaniem spektrometrów OP-FTIR w monitorowaniu powietrza atmosferycznego oraz gazów w procesach przemysłowych
Artykuł w czasopiśmie
MNiSW
15
Lista A
| Status: | |
| Warianty tytułu: |
Open path FT-IR spectra analysis method for monitoring of environment and processes with varying conditions
|
| Autorzy: | Cięszczyk Sławomir, Komada Paweł, Akhmetova Ardak, Mussabekova Assel |
| Rok wydania: | 2016 |
| Wersja dokumentu: | Drukowana | Elektroniczna |
| Język: | polski |
| Numer czasopisma: | 2 |
| Wolumen/Tom: | 18 |
| Strony: | 218 - 234 |
| Web of Science® Times Cited: | 8 |
| Bazy: | Web of Science | Web of Science Core Collection |
| Efekt badań statutowych | NIE |
| Materiał konferencyjny: | NIE |
| Publikacja OA: | TAK |
| Licencja: | |
| Sposób udostępnienia: | Otwarte czasopismo |
| Wersja tekstu: | Ostateczna wersja opublikowana |
| Czas opublikowania: | W momencie opublikowania |
| Abstrakty: | angielski |
| Open-Path Fourier Transform Infrared (OP-FTIR) can be used for monitoring of atmospheric environment. The open path technique is based on the measurement of the absorption along the atmosphere path between radiation source and spectrometer. Measurement paths used in this method have a considerable length from tens of meters to several kilometers. The main advantage of OP-FTIR spectrometry is the possibility of continuously and simultaneously measuring concentrations of multiple compounds. Unfortunately, quantitative analysis of the spectra of such measurements is a difficult issue due to the changing atmospheric conditions and overlapping of the absorption spectra of various components. Numerous algorithms used for the interpretation of the measured spectra have been proposed. They can be classified into methods using classical chemometric calibration and iterative algorithms. Classical Least Square CLS and Partial Least Square PLS are the most commonly used methods of OP-FTIR spectrometry. Iterative methods are based on comparing measured data with synthetic spectra, that is computational models of investigated optical path transmission. For this purpose, databases such as HITRAN are used. Transmission model must take into account not only the spectral characteristics of gases, but also the measuring instrument influence on the measured spectrum. As an example of modeling the spectra of NH3 and HCl gas are used. Modeling of gas spectra with different resolution is shown. Classical methods of building a chemometric calibration model require appropriate reference samples. This is usually associated with considerable cost and time-consuming calibration process. In addition, correct calibration requires maintaining the same conditions during the calibration, as in practical measurements. This is possible only in the case of laboratory measurements. In particular, it is necessary to maintain a constant temperature and pressure of examined substances. It is connected with changes in width and intensity of gas rotational lines. In classical spectroscopy, changing environmental conditions require new calibration measurements. In the open path spectroscopy, changes in conditions occur naturally along with the changes in the examined environment (object, process). If the measurement conditions in the environment differ from those in calibration measurements, significant errors in determining the content of the ingredients may appear. Greater changes in conditions may occur in a variety of chemical or physical processes. Sometimes it is not possible to perform measurements in conditions similar to those occurring in a particular industrial facility. In such cases, synthetic spectra may be used in two ways: in an iterative process to compare with measured spectra or to form a chemometric calibration models. In the latter case, the problem of changing conditions can be solved in several ways. The simplest method is to build separate calibration models for all conditions that can occur during the measurement. However, in order to use this method, it is necessary to measure the existing conditions and choose an appropriate local model. Another method is to correct the measured spectra and to adapt them to the standard conditions. The third option is to build global models. The spectra of all the conditions that may occur during the measurement are then used for building a calibration model. Then, the effect of temperature on the determination of gas content for local calibration models is investigated. Finally, a global calibration model insensitive to temperature changes in 10-40 degrees C range is built. |
