You say you've worked with gas samples for years? Then you probably have had it up to here with the behavior of some gases in some sample bags. Long ago, I was making a series of measurements using a 40 liter Tedlar (tm) sample bag containing 50 PPM nitric oxide (NO). Hour by hour, the instrument response kept falling off. I blamed the "faulty" sensor -- until I drew a fresh sample of nitric oxide from the cylinder. The signal returned to its initial value! Further tests showed that I was actually measuring the oxidation of NO by atmospheric oxygen, which was permeating not-so-slowly through the bag material.
Virtually every organic polymer is permeable to organic solvent vapors, as well as to tiny gas molecules like helium or hydrogen. Even with high-quality materials like Tedlar (tm), vapors of benzene, carbon tetrachloride, hydrogen sulfide, and ethyl mercaptan all decay rapidly through outward leakage. Air does not permeate very quickly, but there is so much of it outside the bag that its inward leakage can affect other gases (such as nitric oxide).
Our supplier now offers bags made from a proprietary material consisting of layers of polyethylene metallized with aluminum, so that even helium can diffuse through only very slowly. We offer the 12" x 12" and 24" x 24" sizes, often from stock.
The numbers below will serve as a clue as to what you can expect with the laminated material. The experiment began with trace concentrations of four sulfur compounds in the same bag, which were analyzed four times during the following week. At the end of seven days the concentrations of these four, normally unstable, gases were unchanged.
("Tedlar" and "Teflon" are trademarks of I.E. Dupont de Nemours Co.)
Table I. Initial and final concentrations of four sulfur gases over one week of storage in laminated material. Data from intermediate time periods have been omitted.