Ghost peaks

[GregK]

Hello all. I am running the USP <467> (procedure A) residual solvents method on our new Agilent G1888 headspace sampler and our Agilent 6890. I'm just in the preliminary stages of getting it to run. To avoid using up our USP residual solvent standards, I made some pseudo USP standards with high purity (GC and/or LC grade) reagents. About two weeks ago my chromatography looked great: nice smooth baseline with the solvent peaks where they are supposed to be. Then we got a new gas manifold system installed; with new moisture traps and O2 traps . By plumbers . They used some sort of solder to connect the the smaller diameter tubing to the larger (1/8" to 1/4"), not, you know, connectors.

Anyway. Now every headspace I shoot (water blank, water with DMSO, water with DMF, air, just pushing start) has six peaks that come out during the 40°C - 240°C temperature ramp (around 180°C-200°C). I tried baking the whole system overnight with the inlet at 250°C, oven at 260°C, detector at 300°C, headspace sampler around 175°C. Peaks are still there.

After reading the "Analytes Clinging to COlumn" thread, I thought maybe there was something in my inlet. So, tonight I took it apart, changed the liner and septum, trimmed a bunch of my column and started it up again. Peaks are still there. They didn't even have the decency to go down in size.

Anyone have any suggestions? I'd post chromatograms, but we have a web filter that won't let me onto image hosting sites, sorry.

Sorry for the rambling, and thanks for any help.

-Greg

[Peter Apps]

Hi Greg

Sorry to ruin your weekend, but it sounds as if the plumbers have contaminated all your gas lines and the internal workings of the headspacer and GC with soldering flux

I am presuming that they used ordinary plumbing solder - not the high temp eutectic that uses no flux ?

The only thing to try is charcoal filters immediately upstream of the instrument in the hope that the internal plumbing is not too badly contaminated yet.

Peter

[GregK]

I'm not sure what they used. I just know that people said the plumbers were using torches within a few feet of our flammable solvent cabinet, and the hydrogen tank. There's a bunch of what looks like charcoal dust around all the unions. They used some kind of Teflon tape at some of the regular fittings, and some white pasty stuff at others (looks like melted Teflon tape).

I was not here when they were installing anything, as I work 2nd shift. They were gone before I came in. I'm not sure that I would have stopped them if I was here, but I sure wish that I'd at least had the opportunity. We have a whole drawer full of brass fittings they could have used to plumb it correctly...

The whole mess kinda pisses me off. Did I mention I'm supposed to get the residual solvent testing completed by the end of the month?

Time to go home and take out some aggression with my xBox 360...

[Peter Apps]

Hi Greg

You have good reason to be peed off, this is a potential disaster for every instrument in the lab - you now need to go into damage limitation. Turn off all the gas supplies to the instruments so that no more flux vapour gets into them, then go and explain carefully to whoever authorized the plumbers to set to work that the best case scenario is to replace every bit of the gas reticulation system, and the worse case is to replace all the pneumatics in the instruments as well.

They will not be happy, and the passing of bucks and covering of backs will start.

Peter

[krickos]

Hi Greg

Sorry to hear what seems to be a disaster of magnitude. Peter offer some really good advice here.

Never experianced something as serious as this, but we have had smaller incidents around new installations and repairs. In those cases it has been enough with disconnectiong the GC/columns etc and heat treat GC tubing with small toarch to vent off residual contaminats from the tubing. In fact I think we nowdays as standard always heat treat new tubing between supply and GC before installation even if it is good stuff just to be sure.

Here unfoutunately it seems as Peter says that a complete shutdown is necessary to do damage control and redo the whole thing with right tubing and right professionals.

[AICMM]

GregK,

Go grab a cylinder (cylinders) and clamp it to the table in front of the GC and plumb it in yourself with tubing. Leak check everything and start troubleshooting all over again. I would further advise dividing the system in 1/2. Take out the headspace sampler and run the pneumatics straight to the injection port. You may still be hosed but you will quickly identify if the problem will require new pneumatics or simply new gas manifold.

Best regards.

[Ron]

You have my sympathy, I was in a similar situation several years ago. We had a new system installed by a company that did medical gas systems for hospitals and doctor's offices, and the system was so contaminated that it never could be used for the instruments. There are companies around that specialize in gas distribution systems that do good work, they aren't cheap, but how much of a bargain is it to buy a system that never will work.

I wouldn't even try to use the manifold system, use gas cylinders close to the instruments. The safety people may not be too happy, but at least you will be able to do the analysis and ship product.

[larkl]

The goop that looks like "melted teflon" may be pipe dope. No telling what's in it, probably not good.

I know of a lab that had to replace EPC modules in 10-20 GC's at considerable expense. All because no one bothered to rinse out the tubing for their new carrier gas line before it was installed. A little bit of dirt or gravel goes a long way.

[GregK]

Thanks for the replies and the sympathy. A few hours of Mass Effect 2 last night has me calmed down a bit. And at least I have a concrete reason the project won't be done at month's end that is not MY fault.

I mentioned to my boss your suggestions. She told me that she gave "the powers that be" the numbers of 3 different companies that specialize in GC gas set-ups. They decided to go the cheaper route and contract some plumbers.

I had already taken the HSS out of the equation. Direct from new manifold to GC, peaks are still present. That's what made me really think (hope) it was in the inlet. I also thought about direct hook up of the He tank to the GC, but wanted management approval. She gave it, now its hooked up, waiting to equilibrate to do an injection. I'm not sure whether to hope for a clean chromatogram or to still have the peaks. Clean means I know what the problem is; dirty means its not in the new manifold.

About the pnuematics in the GC: how hard are those to replace? Is it something Agilent will have to come out and do, or is it something I can manage (not knowing my skill level hampers you ability to answer that, I know; I do lots of repairs, but never inside of the 6890)?

larkl- The melted teflon really looks like melted teflon tape. Some parts have normal teflon tape that blends into gummy, semi-transparent whitish gunk. Its like they heated the tubing at the connection after using the tape?

I'll know shortly if hooking the tank straight to the GC fixes it, keep your fingers crossed.

[Peter Apps]

Hi Greg

If you still see the peaks with a new He tank plumbed direct ot the GC it could mean that the GC plumbing is contaminated. Let if run for a while and see if the peaks get smaller.

Changing pneumatics is likely to be a bit of a fiddle, but if a service tech can do it I'm sure that a lab person can as well. I'm not sure if you would have to reset some software to recognise the new EPC modules ?

Peter

[AICMM]

GregK,

The "melted teflon" could be Swack from Swagelok. I hate the stuff, only because it gets all over fingers, clothing, instruments, etc....

Regarding replacing EPC's, it is relatively easy on he 6890, some torx screws and some patience and you should be fine. I would guess you might have to change your pressures a bit to get the retention times to reproduce (or change your rt table) but other than that, not a big deal.

Best regards.

[CE Instruments]

Short cuts

The goop that looks like "melted teflon" may be pipe dope. No telling what's in it, probably not good.

My money is on liquid PTFE used as a sealant it is ptfe in a solvent , will remain forever in the lines

Best suggestion is write off all the plumbing up to the GC and pray that running clean gas through the GC and headspace that this clears the solvent out. It should work as anything that made it that far will be volatile enough to go.

[GregK]

Okay, guys (and gals?), the direct He hook-up to the system removed all the ghosts peaks. Well, removed is a bit over-stated. Practically removed is more like it. Three peaks went from about 25-30pA to about 3pA in height the other three disappeared altogether. This was after about 2 hours of flush with the clean He. I ran a bunch of injections through the weekend, and the peaks didn't move or change height, so I'm assuming they are there to stay.

Now I have to get the USP residual solvents method running well enough to quantify my samples, then I can deal with the manifold. I was thinking about just replacing the tubing (and "fittings") from the moisture and O2 filters to the GC. Shouldn't take too long, or a lot of investment. But, that is for another day (week, month...).

Well, all I can say is its a good thing they decided to cut that corner and not get a real GC gas company in here. I wouldn't have spent a whole week trying to figure out what killed my runs otherwise. And I'm sure they saved more money than they paid me for a whole week of wasted time. Maybe next time something like this happens, I'll speak up BEFORE the work is done.

[Peter Apps]

Hi Greg

Well, it could have been worse. Change what you can easily and cheaply (like tubing and connectors) and hope that the contamination is sufficiently reduced.

Are there any other instruments that use the same gas showing the same problems ?

Peter