An explanation for gas line leaks

[jsupica]

I work as an instrument specialist with GCMS systems and a technician approached me with a question:

"We saw an air leak on a GC and fixed it by tightening a loose helium line between the tank and the GC. If the lines are pressurized, wouldn't helium just get out instead of air getting in?"

Right now my best guess is that since the loose fitting was a pinch point in the line the addition of air is because of the Venturi effect. I'm almost regretting not taking fluid dynamics in college but I have no response for the technician. Can anyone provide an answer as to exactly why this happens?

[Peter Apps]

A gas (including air) is able to diffuse upstream against a flow. While the bulk of the gas diffuses quite slowly there is a small portion of it with a very rapid diffusion rate that overcomes even rapid downstream flow. In addition the flow profile through an orifice or along a pipe is parabolic, with the gas near the wall moving very slowly downstream - this slow moving layer offers an easy path for upstream diffusion.

Peter

[GasMan]

jsupica wrote "If the lines are pressurized, wouldn't helium just get out instead of air getting in?"

This is a big misunderstanding and one of the reasons that GC suppliers do not recommend using plastic tubing for the gas lines. Besides what Peter mentioned, you have to take into account partial pressures of gases. If you have a line with ultra pure helium in it at high pressure, the partial pressure of nitrogen and oxygen is virtually zero. Outside of the the tubing you will have nitrogen at a partial pressure of approx 600 mm of Hg and oxygen at 150 mm Hg. The plastic tubing acts as a membrane, and with a pressure difference for the nitrogen and oxygen, they will diffuse INTO the helium line. The same can happen if a flexible 'O' ring is used anywhere in the gas system for sealing purposes.

Gasman

[Don_Hilton]

This is also the reason why a high air background in the MS sometimes goes away when a worn septum is replaced on a GC in a GC/MS system...

[Peter Apps]

jsupica wrote "If the lines are pressurized, wouldn't helium just get out instead of air getting in?"

This is a big misunderstanding and one of the reasons that GC suppliers do not recommend using plastic tubing for the gas lines. Besides what Peter mentioned, you have to take into account partial pressures of gases. If you have a line with ultra pure helium in it at high pressure, the partial pressure of nitrogen and oxygen is virtually zero. Outside of the the tubing you will have nitrogen at a partial pressure of approx 600 mm of Hg and oxygen at 150 mm Hg. The plastic tubing acts as a membrane, and with a pressure difference for the nitrogen and oxygen, they will diffuse INTO the helium line. The same can happen if a flexible 'O' ring is used anywhere in the gas system for sealing purposes.

Gasman

For a spectacular demonstration of how gasses diffuse through plastics, spray some soapy water onto the plastic-lined braided metal lines that some cylinder manifold gas systems have between cylinder and regulator - there is enough gas leaking through to generate a thick foam. Works best with helium of course.

Peter

[jsupica]

GasMan and Peter

Excellent information, thanks very much!

[MSCHemist]

Aren't there flexible o-rings all over the instrument including the inlet/liner, side plate, the KF connections, vent-valve, and where the transferline weldment meets the interface chamber (as well as several other places).

[jsupica]

MSChemist,

You're right, O-rings abound. Are you wondering if helium diffuses through O-rings, then why would they design a system with so many of them?

I'm guessing that the GCMS specific O-rings may be formulated to limit diffusion. They're also very small, maybe diffusion is only really an issue if you use off the shelf O-rings at a junction between the instrument and the tanks.

Total guesses, what do you think?