PURGING THE SAMPLE COMPARTMENT IS NOT RECOMMENDED
When doing gas analysis with a long path cell, the sample compartment should not be purged. The spectrometer itself should be desiccated and sealed up, but the sample compartment may be left open to the atmosphere. Three disadvantages of purging are:
(1) Purging requires gas tanks, gas lines, gas regulators, and careful attention from the spectrometer operator.
(2) Purging wastes gas.
(3) Purging can upset the balance of the water absorption lines between the background single-beam spectrum and the sample single-beam spectrum.
If the purge improves during the interval between recording the background spectrum and recording the sample spectrum, a negative error will be introduced into any measurement of water vapor. If the purge slows down during that interval, a positive water error will be introduced.
If the sample compartment has room air in it, and only a short time is allowed between the recording of the background and sample single-beam spectra, there will be no water error in the resultant absorbance spectrum, because the humidity in a room does not change very fast.
Another point to be remembered is that when working through a long optical path, the spectrum of water in the cell usually is far stronger than the spectrum of water due to any water changes in the sample compartment. In most long path gas studies, the presence of water vapor in the sample compartment is therefore irrelevant.
The sample compartment water only becomes relevant when one is trying to measure traces of water as an impurity in some other gas. In that case, it is recommended that the sample compartment be sealed, but that there should not be any purge.