Ghost peak

[Ryklys]

Hello,

I am doing chlorinated pesticides assays on Agilent 6890-ECD. Pair of chromatographers. Detectors (mikro ECD) are quite old ones (~ 6years). Probably they have been dead by now, but we do not have many samples. Gas carrier is N2. Problem is just one annoing ghost peak on boths colums (HP5, DB35MS). And most annoing of all — those ghost peaks acts as one of my sreened pesticides. And reacently it began to act as one of PCB (diferent assay, but same owen and column method), that newer happened before. Interesting thing is I do not see those ghost peaks while chronographing standart solution or blank solution. It only happens during sample sreening.
Things tried: baking of colums, twice, heating of ECDs (many times for another reasons), colums trimming (twice already), inlet cleaning (really — washing with solvent like petroleum ether, n-hexane, acetone). Have tried working surfaces cleanliness, working solvents, blank cleaning gels, cleanliness of equipment that is used. Didn't find the source.
Last thing. Our equipment have been used with splitless liners instead of spilt ones for ~5 years (in split mode). So could happen kind of bild up of fatty residues somewhere, but colums were trimmed and inlets washed, liners, septas, ferules, goldplated seals changed.
Any ideas?
Thanks.

[gcguy]

Are you taking multiple injections from the same vial? It could be the vial cap leaching something in that is being detected. The reason I am suggesting this is that we had a similar problem recently.

GCguy

[Rolandas Plausinaitis]

Labas/Hi

You did not mention sample preparation principle.

When You say "blank" - do You mean just solvent used for the standart/sample dissolution or was the "blank" processed trough all the sample preparation steps? Maybe you are catching something from the equipment used for the samples extraction?

[Ryklys]

No, Gcguy, I don't talk about multiple injections. Every vial is injected twice (into one and then the other GC, usually within 5-40 minutes interval), except standart solution for bracketing calibration. But there never are any problems with those ghost peaks in the standart solution chromatograms, but almost always are in samples. I have heard about vial lid septa caused problems. So did sort of experiment: left blank solvent in the closed vial for a day or two — no ghost peaks.

Labas (Hi), Rolandai, in this case I wrote about blank solvent — acetone:ciklohexane (50:50).
We use 3 different methods: high water content (wegetables), low water content (feed) and fatty products. First two are quite similar liquide extractions using dichlormethane and cleanup in packed column (silicagel or florisil). Fatts are extracted by petroleum ether and n-hexsane (80:20), cleaned up by florisil packed column. I did equipment and working surfaced tests (tried all used solvents and clean up gels as well): rinsed or cleaned with acetone, concentrated up to 1 ml and analysed on GC. Din't find anuthing warning.

[Rolandas Plausinaitis]

Some more info on the peak might help.

Since You mentioned PCB, I presume peak appears later in chromatogram (or it shifts retention time, some pattern = moving later, earlier since first appearance?).

Is it regular shape - are you sure it is not late eluting peak from previous injection? Big, small, fronting, tailing? Is retention time/area the same for both injections from the same vial - and for several vials in same batch? Does it appear in all samples of the batch? Only for some particular products? Completely randomly?

[Ryklys]

Peaks (by multiple I mean single one on both GCs) appear almost in the midle of the cycles. The ghost "concentrations" are in the range of 100-300 ppb. And they differ with time. Usually they are 2-7 areas of the inner standarts. In my assay they have Endryn retention time (or just recently PCB 118). I get just one nice solid peak. On each GC. Anybody who is not informed wouldn't doubt the result. The concentrations of those peaks ussually are quite simillar in bouth columns. I get those peaks with every fatty and vegetable sample and in most cases (but not always) with feed as well. Strange thing with feed is —ghost peaks are present in kind of fatty matrixes: fish powder and so, but absend or clearly identifyable as unknown residues in cereal matrix. Don't see any time drift. This problem lasts for a while already. I may say peaks areas decreased a little bit, but no clear retention time drift.
Quite sure it isn't peak from previous injection: never happens in blank solvent.

[Peter Apps]

If the peak does not appear in solvent blanks and standards you are right to be sure that it is not left over from previous injections - in which case it is not a ghost peak at all, it is a contaminant.

There are two possibilities; the peak is present in your samples and your clean up method is not removong it when it should. This is not likely since you find it in all types of samples. Or: the peak is being introduced during sample prep.

You first need to run a blank of the whole sample prep, in other words do the whole procedure with everything except the sample. Then either break your sample prep down into stages and try each one, or eliminate / change the source of every solvent, cartridge, reagent until the contaminant peak goes away.

Straightforward but tedious.

Peter

[Rolandas Plausinaitis]

Hi,

It is clear that it is not GC problem - it is completely unlikely two instruments would have same problem simultaneously. Regular peak shape indicates that "peak" enters trough injector - othervise could not have Gausian shape.

As mentioned above - this leaves two possibilities:
A - it is in the sample (including packaging and sampling equipment used)
B - it s coming from chemicals or equipment/glasware/plasticware used in sample preparation.

There is no other way - You have to go trough the method systematicaly and try to isolate the source as Peter suggested.

First try to identify the any change after which contamination peak appeared (probably you will not be succesful, but have to try)- samples started to arrive in different wessels, or new type of closure; new batch/supplier of some chemical, solvent; new glassware; equipment (evaporators) repaired; changed suplier of vials, septa; change in used labware cleaning procedure?

Decreasing peak area would indicate it could be coming from sample preparation equipment and gradualy "washing out" - some new plastic or rubber equipment/tubing coming in contact with the extract?

I presume you have separate sample prep procedures for fatty and non-fatty products. If it is so, try to think over which chemicals/equipment are used in both methods, then concentrate on those which are used only for fatty products.

If the peak is big enough and you have some friend/colleague working on GC/MS who could measure your sample for free as a favor- maybe you could identify what the compound is.

[Ryklys]

Thank you.

Somehow I did not realise myself to try my samples on diferent GC. I will try to locate the nearest awailable facility. Thank you again, for good idea.

[Rolandas Plausinaitis]

Good luck,
Inform us if you find the source - it is educational for everyone.

[Pinal]

DELETED BY MODERATOR

[dalarie]

I have had the same problem. I use DB-5 and DB-17 for organoclorinated pesticides and pcbs. I have noticed the peak at the same time as endrin in db-5 and pp-ddt in the db-17. Just today i only changed the liner (cause it was breaking down endrin) to a brand new liner and the peak appeared again. I had this problem for months, it was in the blank solvent and with no solvent at all.

It must be something in the liner or inlet because i didn't change anything else but the liner

[Ryklys]

Kind of progress report.

Dalarie: I use DB5 and DB 35 colums.

After few tests and studies of chromatograms, I have a feeling that it is equipment contamination. Could be inlet fouling with faty residues.
Chromatograms thorough examinations show slow decrease of the unwanted peak. It apears that I have very spontaneus shaped peak on almost every conditioning (blank cycle with no solwent or solution used) cycle. With decreasing tendence and decreasing peak if machines are used every day (that hapens rarely) up to no peak at all; increasing after GC left standing in ,,sleep'' regime (decreased gas flow: for column, make up gas, decreased detector temperature, switched of output) for few days or more.

I did not do wegetables blank yet, but feed and fatty blanks did not show any contamination. That is where I am confused. Couse my idea was: fatty matrix residues pulls ''endrin'' contaminants out of inlet compartment.

[Peter Apps]

If it is equipment contamination then the contaminant peak should appear in all the blanks, and letting the machine stand should decrease the contamination, not increase it. Do you have scrubbers on your carrier and mak-up gasses ?

Peter

[Ryklys]

/.../Do you have scrubbers on your carrier and mak-up gasses ?

Peter

Yes, we do.