You are here

PATI employs FTIR both in the lab and in the field. The lab uses FTIR for TDT Air Scan® analyses, off gas studies, and other special projects. Typical applications for the field (extractive FTIR) are stack emission compliance tests, relative accuracy test audits (RATAs), and engineering studies. We also routinely perform EPA Methods 320, 321, and 318.

Fourier Transform Infrared (FT-IR) spectrometry is an instrumental method that permits simultaneous measurement of infrared absorption frequencies of chemical species in real time. The spectrometer functions by imparting a different modulation frequency to each individual infrared frequency or wavelength. The amplitude of each frequency detected is used to generate the infrared spectrum. Reduction in the amplitude at any specific frequency is the absorption due to a molecular species.

EPA Methods 320 and ASTM D6348 require that FTIR gas analysis data collected for compliance applications be reviewed for accuracy following collection. The review is performed to look for unknown or unexpected gases present in the data that may act as interference and cause the quantitative results to be incorrect. The validation also is required to make sure the automated analysis software is correctly calculating each concentration in the sample and that no biases are observed.

Prism has developed a very detailed FTIR data validation process. We perform this service as a third party verification for source testers using FTIR. Our FTIR data validation process validates the spectral results, that the data was collected correctly, and followed the proper method protocols. Contact Prism about how we can offer this low cost service to your FTIR business. Sample Validation

Most commercial FTIRs work by taking an infrared beam, splitting it into two equal halves by passing it through a "beamsplitter" that passes half the beam and reflects the other half. Very similar to looking through a window at night, you can see your reflection but you can also see outside. The two halves then travel away from the beamsplitter for certain distance and strike reflecting surfaces (mirrors) and then return to the beamsplitter. If the two beams travel exactly the same distance, all the wavelengths are in phase with their counterpart from the other beam and you get constructive interference (all wavelengths in phase also known as the "Centerburst"). If the beams travel different distances some of the waves will be in phase and some out of phase and you will get both constructive and destructive interference to that and having those two halves interfere with each other. To get a fully computed IR spectrum, one or both of the mirrors move a certain distance to provide the resolution required. The IR detectors measures the change in signal as these mirrors move and this is called an "Interferogram". The interferogram has some correction done on it including an Apodization function multiplied by it and then a Fourier Transform is performed to convert the Interferogram into a Single-Beam spectrum. Two single-beam spectra (background and sample) are then divided by each other and a Log(10) is performed to generate the absorbance spectrum that will be used to measure the gas concentrations.

In FTIR gas analysis, many very light compounds demonstrate significant fine structure (that is very sharp absorption bands), one in particular is H2O that has absorption bands throughout the IR spectrum and can be a main interference in FTIR gas analysis. If higher resolution FTIR instruments are employed (1 cm-1 resolution), the area between the water absorption lines can be viewed for other species present (such as HF, VOCs, NO, NO2, SO2, etc), thus allowing the software to pseudo ignore the large water absorption features. So, higher resolution allows for a better analysis of compounds when interference gases are present that have this fine structure.


Using higher resolution comes with a cost however and that is a loss in signal-to-noise or potentially higher detection limits. As higher resolution spectra are used the measured noise goes up as the resolution goes up. It also requires the user to incorporate smaller IR detectors that will have lower signal-to-noise causing the signal-to-noise to be even worse. So, the detection limits when using higher resolution can be higher than when using low resolution. If however there is an interference like water present when trying to measure a gas (HF, NO, NO2, etc), the noise created by the water interference can be much greater than the noise of the high resolution spectrum. The detection limit can then actually be better with high resolution because the interference can be reduced or eliminated.

If significant interference compounds are present in a spectral data set, higher resolution generally produce better results. However, if the compound you are trying to measure has no spectral interference then lower resolution will normally produce better results.

Most commercial FTIRs have a laser (either HeNe or diode laser) and it is used for a couple of purposes. First, it is used to track the movement of the mirror(s) position. In some systems it is used to tell the software when to collect data. Lastly, it is used as a frequency reference. The laser emission frequency is normally very stable and very narrow, this constant frequency is then used as a reference to calculate all the IR frequencies from the IR interferogram.

The laser is not the IR beam that is measuring the compounds, it is strictly for system operation only. If the laser fails or dims the instrument will not operate correctly or at all.

Since nearly every gas has an infrared absorption spectrum, it is best to answer what gases FTIR cannot measure. A special class of molecules, called homonuclear diatomics do not have an infrared active spectrum. Among these are, hydrogen (H2), nitrogen (N2), oxygen (O2), chlorine (Cl2), bromine (Br2) and Iodine (I2). In addition, monoatomic gases such as Argon (Ar), Krypton (Kr), Helium(He), and Xenon (Xe) do not have an infrared absorption.

Two other special gases, hydrogen iodide (HI) and hydrogen sulfide(H2S) have infrared spectra, but they are so weak that it would require 100’s of ppm of the gas to be able to detect them.