# diffusion of gases in both directions

I am a technical specialist that works on analytical equipment. Gas chromatographs use carrier gases like helium and hydrogen that flow through long, thin glass tubes coated on the inner surface with compounds that separate different compounds, usually by boiling point. These glass tubes are called columns and are usually run at elevated temperatures from 60-200 C depending on the application. If enough oxygen contaminates the carrier gas, the columns can be damaged or destroyed.

Sorry for the long intro. There is literature that states that if there is a small leak of helium into the atmosphere through a connection fitting, atmospheric oxygen can enter through this leak into the helium even under pressure. I have seen this actually happen several times over my career. I have also seen this happen in some of my other work.

My questions are these:

Is this diffusion? Most examples of diffusion I have read are the release of a volume of a certain substance into a larger volume of another substance and the attendant speed of mixing. That seems close, but not quite right.

Is this effusion? Most examples of effusion I have read start out with a divided chamber with a gas mixture on one side of the chamber such as 50% hydrogen and 50% carbon monoxide. There is a vacuum in the other chamber. The chambers are connected by a small hole. Because of the size of the hole and differences between mean paths, cross section, pressure difference and other factors of each of the compounds, the hydrogen molecules (H2) will move from the gas side to the vacuum side at as substantially higher rate than the Carbon Monoxide. This does not seem quite right either.

Is this a combination the two, or something completely different?

I am quite sure that the velocity inside the leaking column connections is not bringing in the atmosphere through Bernoulli or pitot phenomenon.

My only assumption is that the leak or hole has to be larger than oxygen to pass into the helium. Possibly the pathway has to be relatively straight. Shorter pathway is probably better a longer one.

Near as I can tell, I am seeing an exchange of gases in opposite directions. Atmospheric oxygen is about 21%. It takes 100 ppmv of oxygen or greater over time to damage a column. Helium pressure inside a column is usually from 60 psig at one end and atmospheric at the other end over a length of 30 meters. These are common numbers.

If you have read down to here, I would appreciate a response and hopefully an explanation of what I am seeing.