GC/MS baseline problems in SIM mode

[Karen]

Hi,

We have been having problems with our Agilent GC/MS in both NCI and EI, but only in SIM mode.

We first noticed this after injecting a 1 pg standard, but now that we look back, we have been having this problem in SIM mode for quite some time. The baseline sits at 0 with random fluctuations, which look as if they are just electronic noise. We are still getting peaks. This is more of a problem in some SIM windows than others. For example, a SIM window containing 6 ions (dwell time = 100 msec, so 1.6 cycles per second) looks terrible. Each SIM window is set up so that the cycles per second is more than 1.5.

Agilent thought the problem was with our sensitivity, however increasing the EMVolts has no effect on the baseline. And since we have had this problem so long and the fullscan baseline looks fine, we have ruled out problems with our consumables. We are also experiencing the problem with 2 different columns and different instrumental methods.

I have also been looking at some pine needle samples that have been run in SIM mode. These have a normal baseline, well at least a baseline that looks normal to my eyes. I'm continuing to look back at more of our runs from the last year.

We have also compared our 1 pg standard to a similar 1 pg standard run on another GC/MS in SIM mode - the other GC/MS has a normal looking baseline.

Any ideas??? I will try to work out how to post a chromatogram so I can show you guys

Thanks
Karen

[Karen]

[jh1]

I guess I'm a bit confused as to your problem, is it the tiny peak preceding the bigger peak or the way the noise fluctuates?

[thohry]

If scan mode is ok, you can do the extract ion for the ions you have in the sim mode to check where the problem come from. For certain ions, the peakshape is very bad but still used for quantitation.

[Don_Hilton]

In scan mode you are typically looking at more stuff on injection. SIM, particularly when peaks are small represents just a little stuff on column -- So, I have to ask. When was inlet maintanance last done. And do the two instruments have identical liners, columns and conditions? (I won't tell why I ask -- therefore I don't have to admit to anything!)

[chemstation]

Hi,
Its hard to read the baseline counts of the Simionich lab, but it appears ~45, where as the Hageman clearly show ~0.
What are the threshold values set to?. ( n.b. I did read that changing the EM voltage made no difference.)
Have you convert your Full scan method to the SIM method to check that the original method hasn't corrupted its self, and that it is using the same tune file.

Are the attached chromatograms of the same Standard mixture (concentration and chemicals)?

Alex

[JI2002]

The baseline looks bad compared to Simonich Lab because the scales in abundance are different:0-90 compared to 0-3000. There is nothing wrong with the baseline itself. The low sensitivity may be caused by problems in the front: split ratio too large? leak? low injector temperature?

The cycles per second is a little low. For a peak with PW of 0.1 min, you only collect about 10 data points across the peak. It's recommended to have 15 data points for a peak. So shorten the dwell time to make it to 2.5 cycles per second.

[Karen]

JH1 - We are mostly concerned with the baseline and how the noise fluctuates.

Don_Hilton - We are very regular with our inlet maintenance. I also changed the liner, septum and cleaned the inlet, and tried re-running with no luck (gold seal doesn't look too worn). As we seem to have had this problem for a while and consumables have been changed a lot, we thought we could rule that out. The two instruments do have different columns (in length that is) and conditions. I'm not completely sure about the liners, though I expect they would be using Agilent deactivated liners too. However, we have used a similar column and exactly the same instrumental method before and still had the problem then.

Chemstation - Yes, the baseline for Simonich is ca 45, while ours is 0. These are both run in SIM mode - threshold is a fullscan parameter, although that was one of our first thoughts (it seems as if the bottom of the baseline has been chopped off?!). We have run fullscan and SIM scan using the same method and still have the problem so don't think the method is corrupt. Also, we are having the same problem with different instrumental methods and different columns. The standard mixture is not exactly the same - they are both mixtures of organic contaminants (organochlorine pesticides, other pesticides, polychlorinated biphenyls) - the Simonich standard contains more analytes and slightly more variety of chemical classes than our standard (it was run a few years ago when my supervisor worked in the Simonich lab).

JI2002 - We are running pulsed splitless injection with inlet temperature of 250 deg C and for some methods 270 deg C, depending on what we are looking for. I'm running the standard again with cycles per second no lower than 2.5 so will see how it affects the chromatogram. Dwell time and cycles per second seems to be one of those iffy things. Agilent recommended 6 scans across a peak, while someone else recommended no less than 1.5 cycles per second.

I will try and post some more pictures of the baseline.

[Karen]

Here we go. You can see a few different SIM windows here. The middle one looks like it has good baseline to me. But the first and last look like they have been cut off.

[Don_Hilton]

What do the traces look like on a blank run? The baseline could be elevated because there is a peak tail running along under it in that middle section.

[Karen]

Good idea. I will try a blank run and see what happens.

I tried changing the dwell times so the cycles per second are always more than 2.5 - no luck. Interestingly, some of the sizes of the peaks have changed - some have gotten bigger, some have gotten smaller. But no change in the baseline.

[giacomo56]

Hi Karen,

First thing....are both labs running the same instrumentation? Which version of the GC/MS do you both have? ....be specific there's multiple version of each instrument.....and I know software for each GC/MSD is different.

[AICMM]

Karen,

May I suggest you post both tunes from each instrument? Also, it would be very useful to know exactly what liner and with that, what the column position is. If you are running pulsed splitless and trying to get 1 pg sensitivity, the liner you choose will make a huge difference.

Best regards.

[Karen]

Hi Guys,

Thought you might be interested in an update of this situation.

We are finding that the parameters from the autotune are not corresponding to the tune parameters that the instrumental method is using. We suspect this is due to a glitch in the software.

Karen

[moraespack]

Hi everybody, I´m also having this kind of trouble with sim mode.

First, i´d like to know if the concerned issue was resolved and how it was.

Second, my problem is that the baseline of my gc/ms is worse than Karen´s .

The interesting part is that when we do a maintenance in any part and turn the gc on, the baseline seems to be OK. But, after 1 hour, the baseline starts to be noisy.

I have changed gás filter, liner, septum, washes solvent, helium cylinder, column, o-ring, but it didn´t work.

I have also cleaned injector body, and ion source, but it didn´t work. Could it be some contamination from the vacum pump oil?

Thanks

Marcelo