Need help troubleshooting Varian 450-GC

[JWF239]

I had been having some issues with my GC which I finally got fixed but it required a good bake out afterwards so I baked the column and PDHID detector over the weekend. When I returned I noticed that my scans were completely different than they used to be.



The red scan is one of my old scans and what I was expecting. The blue one was what I got when I returned from the bake out. I even tried again the day after to see if the temperatures had not stabilized and I got the same result.

You can see the hydrogen peak shifted a bit but kept mostly the same shape and size so that is not a huge deal. The issue is with everything else though. You can see how the argon and oxygen used to overlap but had enough separation to decipher the peaks from each other. Also nitrogen would elude at around 13 minutes. Fast forward to the blue scan and it is impossible to tell what is going on. The argon/oxygen peaks have either completely separated or completely joined together, and depending on which of those happened, the nitrogen either shifted by 8 minutes or disappeared completely. Not to mention the peak shapes and sizes are all different. I let it run for 30 minutes and nothing else eluded after the 5 minute mark.

I believe the bakeout either ruined the column or deposited something on the inlet liner. Problem is I don't even know where the inlet liner is located or how to get to it to see if if looks like it needs to be replaced.

Has anyone else experienced anything like this or have any idea or advice on how to proceed next?

[aidnai]

Well, the fact that you are separating argon and oxygen really narrows down the type of column you might possibly be using (molesieve 5a?), BUT -- you should still tell us what column. If you are worried that you damaged the column, why not include some details like the max rating of the column and the temperature you baked it at?

Good separation of argon and oxygen often requires cryo (unless you have a PLOT column perhaps), is that working properly?

Bakeouts are necessary for molesieve columns every once in a while, but the problem is they are normally run isothermal and the bakeout thermal cycle can loosen ferrules that were fine beforehand. If you have leaks present during bakeout, you can certainly introducing a bunch of atmospheric CO2 and H2O which would trash your chromatography. This hyposthesis is supported somewhat by the fact that it looks like your RTs are way down -- the nanopores in the zeolite are already stuffed with H2O and CO2, so your components of interest are unretained. This would hopefully be reversible with another leak-free bakeout.

I would forget about your idea about liner deposits for now. Although the fact you even make that comment makes me wonder what you are putting on your molesieve (assuming that is what you are using...)...

Aidan

[Peter Apps]

I'll echo aidan - the more you tell us the more we can help.

At the risk of stating the obvious; the inlet liner would be in the inlet if you had one, but with gas analyses it is usual to introduce the sample to the column using a six-port valve. If you do have an inlet, how are you introducing the samples ?

Check the flow through the column with a flow meter connected to the outlet of the detector.

Peter

[JWF239]

Thank you for the quick reply.

It is indeed a moleseive 5a. The column was already installed when I started here and I cannot find the box but everywhere online has the maximum temperature at 300. Another column has a maximum temperature of 240 C. I baked out at 150 C from Friday to Monday because I was advised that temperature should be plenty high enough to get rid of any contaminants. Since I cannot find the box I am also not sure if it is a PLOT column or not but from what I can tell all 5a columns are PLOT columns. I do not use any cryogenics with this system.

I was looking at some trouble shooting manuals and most of them listed liner issues when I looked at the issues I have been having so that is the only reason I brought it up. I thought baking for the weekend may have deposited something in it.

I am baking out at 200 C tonight and am going to try a blank run tomorrow to see if any discrete peaks are seen.

The issue you described explains why the argon and oxygen peaks eluded together and why my nitrogen peak is coming out so fast so I hope you are right and another bake out will clear it up.

[JWF239]

I'll echo aidan - the more you tell us the more we can help.

At the risk of stating the obvious; the inlet liner would be in the inlet if you had one, but with gas analyses it is usual to introduce the sample to the column using a six-port valve. If you do have an inlet, how are you introducing the samples ?

Check the flow through the column with a flow meter connected to the outlet of the detector.

Peter

Also thank you for the help!

I do have the 6 port valve for injecting, that is why I was having issues picturing where the inlet would actually be to check it. Like I said in the previous post, I only considered that because the trouble shooting documents I was looking at mentioned that for the issues I was seeing so that eliminates the inlet concern. I will check the flow through the column and see what I come up with.

I don't have a ton of GC background and this system was set up years ago by someone that does not work here anymore so my training and experience are all over the place. Sure that was pretty obvious!

[aidnai]

Hey, don't feel bad for asking questions. Troubleshooting guides pointed you towards liners, you asked about it. No shame in that.

The reason I told you to forget about liners in your case is that you are analyzing permanent gases. Even if you had tar in your inlet, it wouldn't cause "inlet discrimination" with your analytes the way it would with organics. And of course even if it could, where would the tar have come from?

Based on the temperature of bake out you mentioned, it seems unlikely that you damaged the column. If another bake out does the trick -- voila! now you at least know what it looks like when a molesieve needs to be regenerated.

BTW the options for molesieve columns are packed or PLOT (capillary like). The difference between them is very obvious from opening your oven door -- packed will be a metal column, usually 1/8" OD and a few meters long, whereas PLOT will be a very thin (e.g. 0.53mm) much longer column (15-60 meters). Based on your fairly lengthy retention time and Ar/O2 separation without cryo, I am confident that you indeed have a PLOT column.

Before you bake out tonight, do a thorough leak check using a leak sniffer like the Restek one that came up in a thread recently.

[JWF239]

Hey, don't feel bad for asking questions. Troubleshooting guides pointed you towards liners, you asked about it. No shame in that.

The reason I told you to forget about liners in your case is that you are analyzing permanent gases. Even if you had tar in your inlet, it wouldn't cause "inlet discrimination" with your analytes the way it would with organics. And of course even if it could, where would the tar have come from?

Based on the temperature of bake out you mentioned, it seems unlikely that you damaged the column. If another bake out does the trick -- voila! now you at least know what it looks like when a molesieve needs to be regenerated.

BTW the options for molesieve columns are packed or PLOT (capillary like). The difference between them is very obvious from opening your oven door -- packed will be a metal column, usually 1/8" OD and a few meters long, whereas PLOT will be a very thin (e.g. 0.53mm) much longer column (15-60 meters). Based on your fairly lengthy retention time and Ar/O2 separation without cryo, I am confident that you indeed have a PLOT column.

Before you bake out tonight, do a thorough leak check using a leak sniffer like the Restek one that came up in a thread recently.

Well thanks for being so understanding. I learned about the theory of GCs in school but for this job I was kind of just thrown into it and trouble shooting on an instrument that you did not set up is a nightmare.

The column is labeled as a 19' x 1/16" MS 5A 80/100 silco steel. It looks metal so to me it seems like I have a packed rather than a PLOT. We buy ultra high pure carrier grade gas as well as use a purifier for the system so that may be why we get separation of argon and oxygen.

Thank you again for all your help, it can be very frustrating trying to do this alone. Will update in the morning when I recheck the system.

[JWF239]

Based on the temperature of bake out you mentioned, it seems unlikely that you damaged the column. If another bake out does the trick -- voila! now you at least know what it looks like when a molesieve needs to be regenerated.

Ok so I baked out at a higher temp overnight, at 225 C. Here is what I got this morning.



So you can see that it was certainly useful, my peaks are much cleaner and closer to where I expected them, but I still don't have any resolution between the argon and the oxygen peak that I had before. Should I continue to bake out? Or any other suggestions how to get my separation back?

[GOM]

Hi

"A wise man asks questions"

As a background - forgive me if I am stating the obvious

The 80/100 refers to the particle size range of the packing based on sieving through an 80 and 100 mesh sieve. http://www.sigmaaldrich.com/chemistry/s ... rsion.html

Standard packed columns are usually 1/8" or 1/4" OD - outer diameter ( an historical hang up even though lengths are metric). 1/16" OD is 1mm ID -inner diameter

So you have a packed column. The Silcosteel is a deactivated metal Restek brand. The diameter suggests a "micro- packed" although the length at 19 feet is unusual.

The usual bake out temp for Mol Sieve 5A is 300°C

see
http://blog.restek.com/?p=10617

Regards

Ralph

[aidnai]

Hi Ralph, thanks for contributing that restek blog -- reading that about 6 years ago could have saved me a lot of time between then and now!

I have never heard of a micropacked column that long, and honestly I don't have much experience with micropacked of any length. Perhaps it is a custom column specifically for the argon/oxygen separation.

OP: follow the bakeout directions in the restek blog. I have done bakeouts with helium in the past (no one told me not to) but go ahead and try with nitrogen as they suggest. 3-4 hours is sufficient. Often the real reason for overnight bakeouts is that the GC is available then. If you have another column in the oven that can't take 300°C then remove it (and cap all of the connections).

[GOM]

Hi Aidnai

Like you I have no experience of micropacked columns -only standard columns in the analysis of exploding shampoo sachets in India and advertising claims for oxygenated toothpaste.

As you say it may well be a custom length.

I found that that baking out out at 300°C worked

Regards

Ralph

[JWF239]

Ok so I plan on following the instructions of the article but I wanted to ask a few things first. This is mostly a copy paste of an email I sent to someone that has seen our system so if you need any clarification just ask.

It says to bake at 300 C but I have a different column in the oven that cannot go above 240 (HSQ). I believe the HSQ is for the FID which seems to be working just fine so removing this column and plugging everything before baking the mol sieve at 300 should not cause any other issues correct?

Also, in the article it specifically mentions using nitrogen as the carrier gas multiple times. Do you know if this really matters? And if so, why? Currently there is no nitrogen going to the GC at all so the only way I could use nitrogen as the carrier for the reconditioning is to switch the nitrogen and helium tanks. Assuming the fittings will even allow for them to just be switched, will this hurt anything that you can think of? Helium is used for the PDHID but as long as I turn that off before switching the tanks there will be no flow going to the detector, correct?

Also, should I turn off the rest of the gases for this or just leave everything else running?

[GOM]

Yes stay with Helium. Yes remove your other column during bakeout

Regards

Ralph

[JWF239]

Ok guys,

so I baked out last night at 240 C because I did not have time to remove the other column before I had to leave. Here is what I got.



you can clearly see that the argon and oxygen peaks are beginning to separate but didn't quite get there. My next step is going to be to remove the rtx-1301 column and the Haye Sep Q column so that I can bake at 300 C.

The rtx-1301 column goes to my MS which is currently not working so removing it isn't a huge deal but I don't have a seal that fits where I removed the column. It connects to an inlet I don't even use since I have the 6 valve injector. If I remove the inlet, I have a rubber piece that is made to seal that during bakeouts.

I am going to lower the injector temperature so that it does not melt the rubber cover, remove the injector cap and replace with the rubber cover, and bake out like this with no cap on the oven side of the injector since the other side is capped.

I also have another column in my oven that is not labeled running to my FID. It is much shorter, not coiled at all, maybe 10-12 inches long. I am pretty sure this is my methanizer column but it is not labeled, I have no box for it, and I cannot find any images of a methanizer column online. Is the methanizer column actually located in the column oven like all the others, or is it somewhere else?

[JWF239]

So I have everything ready for a more complete bake out but it is telling me "front EFC pressure time out". I can go into the method and disable the front EFC, I just want to make sure this won't hurt anything?