GC and soap leak detectors

[chromy]

Hi all,
I'm curious about the advice not to use soap-based leak detectors on GC systems. What sort of contamination does it cause? Does it affect only the column or the whole system? Gory details are welcome!
Thanks

[DR]

I've never been able to attribute any peaks to Snoop, but I only use it on fittings that are supposed to be under pressure (tank valve, regulator fittings, inlet to GC fittings) - never on an internal or on a column fitting.

[Bruce Hamilton]

As I noted elsewhere, I routinely used soap and a thermal conductivity leak detector. The important aspects are:-

Aqueous soap solutions are viscous, if the pressure is not sufficient, capillary attraction may fill the leak with soap solution - so the leak goes away until the solvent evaporates or is driven out by a pressure pulse.

When you find a leak, the general guidance is to reduce the pressure, and loosen, then retighten, the fitting. The soap can enter the system, and even deposit on the internal joint surfaces, which then creates a corrosion/deposit patch that leaks sometime in the future.

That's why it's really important to wash off all the soap with distilled water before reducing the system pressure.

The main immediate problem with soap solution, is that it can compromise other leak detectors. I had one TCD filament killed just after we bought the detector because a borrower ignored the instruction to rinse and dry all soap solution residues before using the detector.

I put a small cotton plug in the suction line, and never lost another cell ( but had to replace wet cotton many times ). It's amazing how many people can get degrees up to PhD, and yet fail to understand simple instructions.

The other disadvantage of soap solutions is that they often promote corrosion and/or freezing of threaded joints, especially in ovens and detectors ( less important these days, but old detector threaded fittings ( eg on Pye !04, Varian 1200 ) were notorius for freezing after a few months of thermal cycing, especially if soap had been used and not rinsed off, as it greatly accelerated metal corrosion. SS and brass compression fittings tend to tightly bind if soap residues dry out.

I've never seen a effect on the chromatography or baseline because of soap, but I've seen a few broken glass columns because the detector fitting froze after the use of soap.

So, soap is a good and cheap leak detctor, and don't be afraid to use it, but make sure you thoroughly rinse it off, and avoid using it on low pressure systems.

Please keep having fun,

Bruce Hamilton

[gcguy]

I have used Snoop to help seal push fit column connectors. Worked quite well and it appeared to "glue" the column ends into the connector. I didn't see any extra peaks when using it as a leak finder. Watch out if you spray it onto something hot which is above where your hand is....yes I have scalded my hand when it dripped onto it.
I now use an electronic sniffer.

GCguy

[chromy]

Ahhh...I see...*reaching for the DI water bottle!*
Thanks for the heads-up. I know we have used it in the past, and have never rinsed it off. Good to know!

[Peter Apps]

I am a fervent advocate of leak seekers rather than bubble liquids for finding leaks in GC systems, mainly because the consequences of contaminating a GC gas line far outweight the cost of a leak seeker. At the worst you could be looking at replacing all the plumbing on that line -fancy EPC controllers and all. Contamination anywhere except at the inlet would probably go unrecognised as peaks because all you would see is an increase in the baseline, and probably more baseline drift that would not go away when you did all the usual trouble-shooting.

Surprisingly, liquids can creep into leaking fittinge even while gas is flowing out through the leak - the gas velocity over the surface is near zero and a liquid film can creep by capillary attraction in the opposite direction to the gas flow. This is similar to the way that air gets into leaky systems.

The only time that I use a liquid is on air lines (and there I prefer to pressurize temporarily with helium and then use a leak seeker) or if I have a lot of potentially leaky connections in a confined space so that I cannot loctae the leak with a leak seeker. Then I only ise usopropanol, which is non-volatile enough to make bubbles, but voaltile enough to disappear quickly.

Peter

[chromy]

Thanks Peter, I agree that those are chances we don't need to be taking with our GC. My lab is going to start looking around for an electronic detector, and I just hope that our PhD can learn to use it, haha.
Thanks again to all!

[Bruce Hamilton]

I would discourage using isopropanol as a leak detector on or near any operating GCs.

It's a designated flammable liquid ( flash point 12C ), and whilst the autoignition temperature is relative high ( 425C ), the flammability range starts at 2%. Some companies prohibit flammables near GCs, other than those in sample and rinse vials.

If you want to use a liquid on very large leaks, and doen't want to use deterents, I suppose you you could use a nonflammable water/IPA mix.

I would only substitute a gas more sensitive to electronic leak detectors during commissioning or repair of an installation when known leaks couldn't be found, as normally a 24-48 hour pressure test requirement would catch most leaks. Any purifiers in the system will become saturated with the different gas, and breaking into the system downstream creates a possible new leak.

Once a system is installed, the only maintenance I do is annual drying/replacement of moisture purifiers. In my opinion, the risk of rinse water entering a pressurised system is trivial, provided recommended operational pressures are maintained.

I use the electronic detector when suspected leaks are difficult to access - as He and H2 can be trapped in an inverted small funnel held above the area. However, in my experience, one failing of electronic detectors is that they sometimes miss leaks on larger fittings inside instruments because the leak gases rise and disperse quickly. Using long tweezers to put a paper towel below the items and swabbing small amounts of detergent solution, followed by a water rinse works for me. YMMV.

However, your peace of mind in important, so choose what works for you.

[HW Mueller]

Well, I would think that it requires an extraordinary skill to produce a problem with alcohols in this connection. (Of course, one doesn´t use a gallon/leak).