GC (MS) problem with tailing peaks

[soccerdad007]

Hi all,

I am in an educational establishment.
We are using a 6890GC coupled with a 5973 MS. The column is a HP5-MS.
We are demonstrating the technique of GCMS for students by running C8-C12 alkanes (roughly 100 ppm of each) with split of 20:1. The liner is a split liner (pressure drop type I think)

We used to be able to run this solution and get good symmetrical peaks of roughly equal heights/area but now we get a lot of tailing with the tailing increasing as we go from C9 to C12. C8 is symmetrical. We also find that the peak height (and area) decreases drastically as we go from C8 to C12.

I mention this application as this is the most frequent one that I use but other applications are also giving us non-optimal results

Any suggestions
Thanks
Kevin

[Don_Hilton]

When is the last time you changed the liner and septum and checked connections in the GC for leaks? Has anyone changed the method parameters, like flow rate or inlet temperature?

Also how to the retention times compare with chromatograms when all was well?

Let me also suggest that you take a look at the column vendor or GC manfactureer web sites. Some of these folks have nice troubleshooting guides and posters. In an educational setting, a poster showing common problems and solutions could be very helpful to students.

[soccerdad007]

Hi Don
Thanks for getting back to me so fast.
To answer your questions

I changed the septum and liner today (at separate times). The chromatography seemed to improve a little for the first run but then seemed to be as bad as ever.

I checked the obvious connections in the GC for leaks. The MS did not pick up any water or air leaks in the system when I checked its performance
The inlet temperature is the same as normal (150C) and the flow rate is the same (1ml/min - required for the MS)

The retention times are close to what they normally are.

I liked your idea about checking with the manufacturers for troubleshooting guides - we think alike as that was the first place I looked (and tell my students to look!!). I have several of those guides (and even some displayed in the lab). I couldn't find anything that dealt with tailing increasing with retention time - the opposite was frequently mentioned.

Am I right in thinking that it is inlet related as some wise person once told me that >80% of all GC problems are inlet related. Do you have any idea of things that I might try

Thanks
Kevin

[bhuvfe]

Kevin,

Maybe you can give a look at the tip of your column, and if necessary cut few cm.

Does someone else use this system and for what applications?

Last question. Do you have a split/splitless inlet or a PTV inlet?
150 C seems too low to me for a split/splitless (I would go certainly above 200, generally 250-280 C).

Regards,
bhuvfe

[Consumer Products Guy]

Kevin,

Maybe you can give a look at the tip of your column, and if necessary cut few cm.

Does someone else use this system and for what applications?

Last question. Do you have a split/splitless inlet or a PTV inlet?
150 C seems too low to me for a split/splitless (I would go certainly above 200, generally 250-280 C).

Regards,
bhuvfe

I'd start with what bhuvfe stated: inlet temperature, try 200 C and 250C, see if things are: better, worse, or the same. Only make one change at a time so you can learn what isn't right. Post i.d. of your column as well; 5973 should be able to handle 1 ml/min, especially if it's a turbo instead of diffusion pump MSD. Your issue is most likely a GC issue, not MSD.

[Don_Hilton]

Kevin,

What you have heard - that the bulk of GC problems is at the inlet is true. I've fixed mass spec sensitivivity poblems by replacing the inlet septum - more than once.

The incrased tailing with retention time suggests that it is a temperature issue. Inlet temperature could be too cold or the column could be ramping too slowly. (And, as I walked down the hall after posting this - if the retention times are right, the temperature ramp should be good.) Gas flow could be part of the issue as well - thus checking to see if carrier gas is escaping around the inlet ferrule or other inlet parts. Leaks can change with oven temperature - as the gas viscosity changes. And, an increasing leak with oven temperature would correspond to smaller areas at higher temeratures.

On other question that comes to mind: What was done to the instrument before it went bad? Did someone cut the end of the collumn - in which case the column may be installed incorrectly? Are you sure that you are using the same kind of liner as before. (Althought for a hydrocarbon mix, about any liner shoud do - if operated hot enough. I've not operated a split spltless inlet below 250 degrees that I can remember - well, at least intentionally.) Speaking of unintentional. Do you have a copy of the method from back when everything was working? Inadvertant changes are never logged...



Don

[AICMM]

As noted by others. According to Wiki, dodecane has a bp of 216C so you inlet is not hot enough to evaporate it. 250 to 275 C at a minimum.

Also, what is the solvent (or is it a neat mix) since you might want to alter oven profile to re-focus.

Best regards.

[Ron]

Dodecane will vaporize at room temperature, so even though 150C is low for the inlet it is unlikely to be the problem. You are more than likely dealing with a leak, probably at the column connection to the mass spectrometer. I would vent the mass spectrometer, replace the ferrule at the transferline interface, tighten a little past finger tight, flush the column with carrier gas for at least 10 min, heat the column to the maximum temp for about 5 min, cool to close to room temperature, then retighten the column nut. The graphite vespel ferrules don't seal well when cold, and tend to shrink after heating. If you heat, cool, then seal you will get a better seal.

[Don_Hilton]

Before you vent the MS, I'd take a leak sniffer to the inlet end. The tailing of peaks suggests a problem while the sample is still separating. Leaks at the MS end kill sensitivity (and this can vary with a leak varying with temperature), but are not as likely to change peak shape. It might save you a pump-down cycle ... or not. But we can always hope for the easy fix. (I also ocasionally buy a lottery ticket.)

[soccerdad007]

Hi all
Thanks for all the great suggestions...

To summarise what I tried today.

1. Increased the inlet temperature from 150C to 250C

Result no improvement.

2. Did a leak check at 40C (our starting oven temperature) using by spraying canned air (difluoroethane) at the various connections and seeing if the MS picked up any of difluoroethane.

Result no sign of a leak

3. Repeated step 2 but at an oven temperature of 200C

Result no sign of a leak

Is there something better to use to leak check the GCMS than this stuff?

4. Changed the liner to a splitless liner that was on hand

Result first injection looked very much improved, second injection and subsequent injections the problem had returned

5. Vented the MS - will trim the column tomorrow.

6. If that doesn't work I will try the column connection to the MS

Any other ideas
Kevin

[Don_Hilton]

If you have a leak detector (that detects helium coming from an instrument), I'd give it a try around the inlet. Normally a leak around the inlet will let some air into the column because diffusion runs both ways. But it he leak is strong enough, you may not catch it that way.

Be sure that column ends are square and you have the right length of column beyond the end of the transfer line. I've correctly installed a column into the inlet, only to tighten it enough to somehow break off the column at the end of the ferrule... It does not help sensitivity. So it is worth taking a flashlight and looking into the inlet to see that the end of the column is still standing. And, it is a good check for septum crumbs and such on the bottom of the liner.

[soccerdad007]

I just checked out a hunch - the peak areas of the C8-C12 aren't decreasing significantly, it was the increased tailing that was leading to poor integration of the peak area by the data system. Even visually the areas looked very different.

So to summarize
C8 - symmetrical and good peak height/area
C9-C12 increasing amount of tailing though peak area seems to be similar (as long as the integration is done manually on each peak)

Sorry for the confusion
Yours
Kevin

[Peter Apps]

Hi Kevin

What is your column temperature programme ?

Peter

[Don_Hilton]

Chec to be sure that the column dimensions are correctly entered into the MSD software (I assume that you are using ChemStation?). A wrong dimension can result in the calcuation of a wrong pressure profile for the run if you are using constant folow. --- but if that is the only change to the method, retention times should be off...

[Peter Apps]

Having checked back through the thread my money is now on a bit of silicone sitting on top of the gold seal at the bottom of the inlet - this will have escaped the liner change.

Peter