GC-MS problem (methanol is converted to formaldehyd?)

[Chacer]

With a mass spectrometer - you will always see some oxygen. There is always a small leak through the seals of the mass spectrometer. An additional source of air entry into at GC/MS system can be the split/splitless inlet. If your head pressure is low enough, air will diffuse into the inlet if the split vent is open. This can be decreased by increasing the split flow. And you can have air enter through the septum purge and through a cored septum.

You should see some air background on the other GC system. How does it compare when you run the leak check on that instrument?

The reason I suggested injecting methanol was to eliminate reaction of anything but methanol from the picture. If the contaminant peak is still there, it has nothing to do with analytes or other contaminants. If the strange peak is still there, the next step is to push the run button on the GC as if making a manual injection. This will eliminate anything coming in through a syringe or in the solvent.

If you look in methanol for formaldehyde, you will find some. If you check the specifications on your methanol, you will most likely find that it is certified to contain less than some level of formaldehyde. And that specification is good on the date when the methanol was analyzed. If your methanol has been sitting in a stockroom for a couple of years, the level of formaldehyde can increase. Also expect to see some formic acid and methyl formate.

If the methanol has been stored in a clear glass bottle in the lab, look for these to grow even more.

The oxygen level on the other GC is 2.6%
Alle bottles are HPLC gade and less than 6 months old and they are stored in brown bottles...
Ill check the specs for formaldehyd ect.

[Don_Hilton]

The oxygen levels you see between GCs is fairly similar.

So the next question is what do you see when you shoot a methanol blank and then a standard? The point about formaldehyde in methanol is that you will find some, so no need to go looking for it. If your standard is made up in methanol - whatever is going on should go on in the standard if it involves formaldehyde, your compund of interst and the inlet. And the question, right now seems to be what is going on in the inlet.

But, if you have changed all the inlet parts and the column, everything you have in the flow path within the GC should be the same or better than in the the other GC. So the question is what is there outside the GC that gets into the GC that may differ. This could be the syringe or some thing in the folow path coming into the GC.

YOu might try the same solutin on both instruments, using the same syringe and wash solutions for the syringe (dirt in a wash solutin vial can become a peak in a chromatographic run). This would work toward evaluting that external source of the extraneous peak.

Are both GC's plumbed to the same gas bottle? Has anythign been done wiht the pumbing shortly before or about the time the problem started? Has anything been spilled across the fittings that lead into the GC in question? Is there a carbon trap that may have filled and has now broken through.

If you push the run button on the GC first thing in the morning wiht no injection, does that peak show up?

[Chacer]

The oxygen levels you see between GCs is fairly similar.

So the next question is what do you see when you shoot a methanol blank and then a standard? The point about formaldehyde in methanol is that you will find some, so no need to go looking for it. If your standard is made up in methanol - whatever is going on should go on in the standard if it involves formaldehyde, your compund of interst and the inlet. And the question, right now seems to be what is going on in the inlet.

But, if you have changed all the inlet parts and the column, everything you have in the flow path within the GC should be the same or better than in the the other GC. So the question is what is there outside the GC that gets into the GC that may differ. This could be the syringe or some thing in the folow path coming into the GC.

YOu might try the same solutin on both instruments, using the same syringe and wash solutions for the syringe (dirt in a wash solutin vial can become a peak in a chromatographic run). This would work toward evaluting that external source of the extraneous peak.

Are both GC's plumbed to the same gas bottle? Has anythign been done wiht the pumbing shortly before or about the time the problem started? Has anything been spilled across the fittings that lead into the GC in question? Is there a carbon trap that may have filled and has now broken through.

If you push the run button on the GC first thing in the morning wiht no injection, does that peak show up?

We have tried the same solution on both instruments and the peak only shows at 1 of.
Ill try new wash solutions ect.
Ille see if i can use the syringe on the good GC.

Both GCs has the same gas bottle. Nothing has been done to the GC or plumbing. It happend suddenly at got worse over the course of 2 weeks. It has been this way now for months. We have split almost everything apart and cleaned or replaced it (we still need to look into a few things in the inlet). Dont know about the carbon trap, ill look into that.
The peak only shows in out test solution and when a sample with amphetamine or methamphetamine is analysed.
It seems now that both these drugs acting very wierd.

Thanks for the great feedback. Ill bring updates as they come.

[Chacer]

We tried diluting our test solution but this didnt help.
We are running out of ideas.
We have talked to the specialists from Waters and are waiting for their reply.
Deriviation is an option, however we wish to avoid this.

Any ideas are welcome

[Chacer]

A system report indicates 1.6% water.
Is this too much?
I know that too much water in the system can cause adverse chromatographs.

[Peter Apps]

Hi Chacer

The identity of the mystery peak will help a lot. The spectra and chromatograms no longer show (there is a frog in an ice cube instead !). For lack of anything else, what does the MS library search throw up ?

Peter

[Don_Hilton]

The water leve you see on the system report is the water in the mass spectrometer. This results from the small amount of air that leaks into the mass spectrometer, water coming in as part of the air in other leaks and water that is slowly desorbing from surfaces from the last time the inside of the spectrometer was exposed to air.

Use the m/z 28 and 32 to evaluate the system for leaks. As long as the instrument is leak tight, then look at 18 to see that the spectrometer has pumped down long enough to sufficiently remove water.

Your water background in your instrument looks to be OK - from what you posted before, it is lower than m/z 28. This is good.

Water background in the mass spectrometer will not affect the chromatography.

If you see a chromatographic peak for water, this is a different matter - and a matter of sample preparation. And small quantities of water showing up in a chromatogram can be OK. But, I would not set the mass range low enough to pick up water unless you need signals in that range for other analytes.

[Chacer]

Latest spektra

[Chacer]

Everything you see, should be one big peak.
Im off now. More will properbly come monday.

[Peter Apps]

Latest spektra

I see a frog - with domain unregistered underneath it.

[Don_Hilton]

Looking at the spectrum with a base peak of 135 and noticible M+2. This is not typical of compunds containing only C, H, N, and O. While not a great match, there is enough similarity to benzisothiazole or benzothiazole, that I woner if we have a poor spectrum of one of these or if we have a related compund. Display of the ion traces for 135, 136, 138, and 139 may help to confirm that 138 and 139 are signals from the same compund (same retetion time and profile) as m/z 135.

The peaks between methanphetamine and the mystery peak are probably of some interst if they only show up in this particular sample. They may offer clues as to what is going on.

[Chacer]

The peaks between methanphetamine and the mystery peak, have some carectaristics of methamphetamine.
But some are completely off. At some points, there is evidence of formaldehyd reacting with the amine, and adding 1 C (+12) atom in the process.
Ill look at the ion traces and get back...

[Chacer]

Ion traces 135 , 136 , 138 , 139 +/- 0,3

trying a new image host

[Chacer]

The solution/sample, which we called Standard 1 contains: methamphetamine, amphetamine, MDMA, paracetamol, cocaine, heroin and flunitazepam and is dissovled in methanol.

[Chacer]

Ion traces 135 , 136 , 138 , 139 +/- 0,3

trying a new image host

I see signals at 135 + 136 and 137 (slightly) but none 138 and 139