HP 5890 with regular interval negative peaks

[escott]

I have an HP 5890 GC with uECDs and we have recently started observing regular interval negative peaks appearing in our chromatography. They don't appear in every sample that is run but may appear several times during a sequence.

We are using a DB-1701P @ 30m long x 0.53mm ID x 1um film thickness.

Any help will be greatly appreciated.

[sassman]

I have seen this problem before with a dirty ECD. I'd suggest getting your detector cleaned.

[AICMM]

escott,

I agree with Sassman that this is usually an indicator of a dirty ECD although I have observed that the micro-ECD's generally stay much cleaner, much longer, than the older macro-ECD's. Does someone else have the same experience?

You can try thermal cleaning but probably just best to ship them to reputable cleaning facility.

Best regards.

[AKChemist]

I have a Varian 3800 dual ECDs. One ECD is new(only one year). Both ECDs have the same regular interval nagative peaks. I have baked both ECDs at 400 degree C for 4 hrs but the peaks are still there. I didn't run a lot of samples on this GC. Any other possiblity other than getting dirty?
By the way, who is the best company for the cleaning service?
Thanks

[fsistere]

I have a varian 3800 too, with a ECD. I analize THM's extracted with isoctane or pentane and it appears always 2 negative peaks at the same tr. I think that impurities of hydrocarbons in carrier gas make negative peaks in ECD. I run a blank and i substract the chromatogram obteined.

[AICMM]

fsistere, AKChemist,

First, Varian does it a little differently than HP in that they have a different function you control (junction voltage? I can't remember...) that may cause your spiking. I will ask someone I know who has 3800 ECD's.

Second, why do a blank subtraction if the negative peak does not occur near the components of interest? Curious?

Third, why pentane extraction with the ECD's, why not headspace. You might get a 5 or 10:1 concentration factor with solvent (ie 20 of 50 mL of water by 5 or 10 of pentane) but then you do a 1:1000 shot of that by shooting 1 or 2 uL?? Why not just do headspace and shoot 1 mL? Curious? Looking to do this application myself to try out non-radioactive ECD I am working on.

Best regards.

[LCbob]

Hi

Are you seeing the problem just in the chromatograms or do you see negative dips when the system is idle?

If the dips only occur during runs then they are probably a function of your samples or any changes during the run, i.e. temperature.

If the dips appear when the system in idle then as has already been quoted the ECD is dirty.

Out of interest has the general baseline noise changed since the problem appeared?

[AKChemist]

The baseline has changed from a few mV to 30mv.
When the machine is idle, no negative peaks show up.
When I run samples or even solvent blanks, the negative peaks show up about every 3 minutes during 30 minutes run time. Both ECDs have the same negative peak pattern.
Thanks

[sassman]

Are the retention times for the negative peaks reproducible or more random? If they always come out at the same time, then it is probably something in your sample.

When you say that there are no negative peaks when the machine is idle. I assume you are measuring the baseline with the detector on? What happens if you inject just air? What happens if you do no injection, but just run the temperature/flow program?

[AKChemist]

The negative peak RT is same, showing up at the same time. What I injected is only the pesticide grade hexane which I used before for a straight baseline blk.

When I injected only air, the negative peaks show up the same way as other injections. When no injection, only running oven temp ramp, the baseline is ok without nagative peaks.

Thanks

[AKChemist]

The negative peaks show up when I inject pesticide grade hexane.
Tha UHP grade carry gas and makeup gas are both through gas filters before into GC. The negative peaks appeared this year only. I have baked injector port and both ECDs according to Varian instructions but the problem is still there.

I am not really convinced that the peaks are caused by contamination.