Have you checked that there are no leak in your GC inlet (septumn and liner o-ring seal). Did you overload your column. Check the load capacity of your column.

Thanks for the information and I'll check out the liner o-ring seal and septum today on a couple of our GCs, we currently are running 10 GCs. I am certain we did not overload the column due to we have been using the same type of columns, same prepping method and same GC program for awhile now and have never had this problem occur but I'll check to make sure our GC program has not changed. We are analyzing FAMEs and injecting 2uL of hexane layer.

Although I've not run GC for FAMEs however I have experienced issues from running larger batches than usual.  If it is ONLY happening towards the end of long runs, I first thought of the septum as well.  However, besides RT shifts, septum bleed also usually causes tailing peaks and/or noisy baseline.  What about your inlet liner, could it be part-way clogged, due to high sample load, and therefore restricting the flow?  If they are changed between runs, you may not have that problem, i.e. the liner may never get clogged enough to affect resolution.  That may be reaching though, good luck!  Let us know what it was when you figure it out please!

John R Carney

We are using the 2560 columns for trans analysis. All 10 GCs are using the same method and all 10 GCs developed the same issue around the same time with the retention time becoming shorter and shorter as the runs go on. Yes there seems to be a trend which is with each injection the retention time is becoming shorter.

We do change our septum and liner regular at every 250 injections.

We do not currently have any flowmeters to measure the air flow at the GC. We use gas cylinders for nitrogen gas which throughout the day does not change. For hydrogen gas we have Hydrogen generators which the electronic readout is stable. For the air we have a generator for that with a constant read out.

We had to monitor the oven temperature a month or two ago so we placed a thermometer in a few GCs ovens to monitor the temperature. With the readout from the GC and the thermometers the GCs were going through each temperature ramp correctly.

Would it be possible that we got a bad batch of columns and the film is degrading so it is not retaining the FAMEs as well with each injections?

change the septum

What flow setting for Gas Saver function you are using? Disable it and set split flow >50mL/min after the injection cycle. Try to inject a batch of samples to see if it helps.

You have a decrease in retention time, and so you can safely ignore the suggestions about septa or clogged liners, which are based on the possibility of leak or a constriction in gas flow, which will always lead to increased retention. You also (if I interpret correctly) have a trend for the increase to get larger as sample batch size gets larger. So you can eliminate anything that will cause random fluctuations. 

The flow of gasses to the detector cannot affect retention. What is you carrier gas ?

Shorter retentions need faster carrier gas flow or higher oven temperatures. Faster carrier gas flow could be due to a partially blocked split vent, which is not likely to affect 10 GCs all at the same time and to the same extent. A higher oven temperature is an electronics problem that is vanishingly unlikely to affect 10 GCs all at once.

If you start a new batch, do the retentions return to their expected value, or are they the same as they were at the end of the previous batch ? If they are the same as it the  end of the previous batch you probably have progressive column deterioration. 

Are all the peaks similarly affected - if the IS elutes 1 s later, do all the other peaks do the same ? If the shifts are different for different peaks is there a pattern such as earlier or later peaks being more or less retarded. Do you notice any shifts in relative retention or resolution ?

Do your GC run time long enough to elute all peaks during run.

Ah yes, I didn't think about the gas saver.  I've had issues with one myself.  And good point Peter regarding the fact again that it's affecting all 10 GCs, which (unless it is a chemical problem, i.e. column overload) makes me think carrier gas could be the issue.

Are your N2 bottles outside, or indoors?  Is daytime heating of the cylinders affecting this, or does it happen no matter what the time of day/night the GCs are running?  What about internal gas regulation on the 2010 (I'm more familiar with Agilent)?  The suggestion far up above (in John Carney's answer) regarding the non-heating split vent line and it's poss. issues running heavier condensates, warrants further investigation as well, imo.  Again, good luck!

Peter Apps: Our carrier gas is H2. I just checked out batches from 2 different days from the same GC, there was about a 8hour period were no samples ran. From the last samples that ran from day 1 to the first sample that ran on day 2 there is an additional decrease in retention time. It seems the hexane peak is unaffected and shows up at the same time point. However, any peak that is after the hexane peak is showing the trend of having a lower decrease retention time as more samples are going thru the column. Also, from comparing a few chromatograms on the same graph it seems the FAMEs that come off more towards the end are showing a larger decrease in retention times per sample then the early eluting FAMES. For the most part I would say our resolution is good except we have been seeing a merger of C24:1n9 (nervonic) and C23:3n3 (our IS) at a much lower injection count then we have before for our columns. For example, we recently replaced a column with only 400 injections due to this problem where as before we got 2000+ injections before coming across any problems with the column.

Cheung Raymond: We do not use a gas saver function on our GC programs, at least not for our main programs we run.

Joel Berzas: All our N2 bottles are indoors so they are always at a constant temperature.

I cannot think of anything utility-wise that we haven't already ruled out.  Therefore it sounds to me like the issue is either sample and/or column related and if all GCs are using columns from the same lot, it certainly could be a bad batch of columns (as you mentioned previously).  If you have one, maybe try putting an old column back on one of the GCs and see if the issue replicates.  I'd also use guard columns in the future (if you're not using them already).

Please check one point. After you notice a shift to lower retention time, set oven temperature to your column allow maximun,say 320 degree C, for 20 minutes. Then reinject the last sample with the original GC method. This step is to check any accumulation matrix in your sample. 

Hi Jeremy

The finer points of the symptoms - increased retention from the end of one batch to the beginning of another, and all of the GCs similarly affected rule out pretty much everything except a bad batch of columns, which I agree with John Carney is very unlikely, and a carrier gas problem. With hydrogen from a generator there is a possibility of a progressive increase in water content as membranes deteriorate (depending on what generator you have) and this will certainly affect retention on a very polar column (it will also kill the columns much more quickly). What maintenance do you do on your hydrogen generator, and do you have indicating moisture traps between it and the GCs ?

Peter has, as usual, hit the nail squarely on the head. Your retention times are shifting more for heavier compounds than for lighter; you are seeing phase degradation. Oxygen and moisture are especially hard on polar phases, and this sounds suspiciously like what happens when a polar phase begins to degrade.

Thanks for the information.  Been a busy week at work with people being on vacation and haven't had time to get back to you all.  We did have 1 'old' column that was still good to be used from before we started having this issue.   The 'old' column was in a GC but was taken out and ends caped because we needed to use a different type of column for a research project.  We just recently put this 'old' column back in 1 of our GC and will run samples on it over the weekend.  If it shows the same trend then I am guessing what John Carney is saying about H2 generators could be the source of the problem and causing our columns to experience phase degradation.  However, if the column does not show the decrease in retention times then would it be safe to conclude the batch of columns was just 'bad'?