The Iraqis had an interesting way of doing poison gas. They used the binary method, but never developed artillery shells with reliable breakaway diaphrams to mix upon firing. They could never prevent slow leakage from shells, either. Also, since many nerve poisons have limited shelf-life, they did not want to prepare the stuff and seal it into shells. What they did was simple and workable. They manufactured (or purchased) the ingredients for binary nerve poisons. They kept the shells empty until just before firing. Then a single guy in a gas protective suit poured the two chemicals into the shell and screwed a steel stopper into place. The stuff was nearly ready to use by the time he had closed the stopper. It was moved to the artillery piece and fired immediately, so slow leakage was not a problem. Rotation during flight mixed the chemicals thoroughly. Also, shelf life was not a problem, since the precursors are much more stable than the final poison.

When we find the artillery shells with the supplies of binary agents nearby, the idiots will claim that these shells were "not ready for use". True. It would take perhaps 5 minutes to get them ready, however.