WRONG ANSWERS ARE CREATED BY HIGH ABSORBANCE VALUES
Frequently, users of QASoft begin by analyzing a test gas of known concentration, like 10 PPM of NO in nitrogen, or 20 PPM of CO in air. The answers they get sometimes turn out to be far from the correct answers. Then a search for an explanation begins. The user may well suspect that the absorbances of the reference spectra are wrong. Infrared Analysis, Inc. may well say we don’t think our absorbances could be that far off, and then we might ask back how the user handled his sample and what were the experimental conditions of recording the sample spectrum.
Many times it turns out that the sample spectrum was recorded with absorbances so high that they were well out of the linear absorbance region for the particular molecule being studied. High absorbance in the sample spectrum always leads to error. This is true not only of our quantitative analysis method, RIAS, but of all quantitative analysis methods for gases.
We therefore propose a cardinal rule for quantitative analysis of Gases:
NEVER INCLUDE IN THE ANALYSIS REGION ANY SPECTRAL FEATURES OF HIGH ABSORBANCE
The next question to be answered, then, is how low do the absorbances have to be. This depends on the type of molecule being measured, the total pressure of the sample, and the spectral resolution being used. With a relatively large molecule, like acetone, the spectral bands have no fine structure at atmospheric pressure, and the linear absorbance region can run from zero up to 0.5 absorbance units (base 10), or even higher. With such a molecule, spectral resolution does not matter much.
With a small molecule, like NO or CO, there are individual narrow lines in the spectrum, and total pressure and spectral resolution become important factors in determining the usable absorbance region. For this type of molecule, when working at half wavenumber resolution and one atmosphere total pressure, the linear absorbance region runs from zero to only about 0.1 absorbance units (base 10). Any absorbance values higher than 0.l should be considered OFF LIMITS.
This whole matter is rather complicated, but if you observe a rule that, when working at half wavenumber and one atmosphere total pressure, all absorbance values higher than 0.1 (base 10) are FORBIDDEN TERRITORY, you will be safe. In order to widen the range of concentrations for your quantitative analysis, you should make the noise level in your spectrum as low as possible. Consult our discussion of CONCENTRATION RANGE FOR QUANTITATIVE ANALYSIS.
Return to Table of Contents