Peaks on my ECD with a blank run (No injection)

[Lrouxph]

Hi,

The method i am using is as follows

Split/Splitless inlet at 290 C (I run it splitless for 0.8 min and at 50 ml/min split thereafter)
RTX 5 column (30m x 0.32mm x 25um) - Oven starts at 90C/1min - up to 300 C @ 20 C/min
ECD detector set at 350 C

My problem is this, I am getting peaks eluting even when i inject nothing. The peaks are the same for every run.

I have replaced the liner, replaced the column, changed carrier gas supply, changed carrier gas type.

ECD blanked off without column is clean.

The first peaks elute when the oven hits 148 C.

What else can I Check?

[queenann]

I would clean the purge line and swap out the purge cartridge next.

[hellyesidid]

Sharp peaks would generally be from the injector side of the instrument. However, I have seen sharp peaks from cracks in the makeup gas adapter (the cracks open up and allow air in as the oven temperature increases). Did you run the full oven program with the column off and the detector capped - and did you see peaks? If so, then the makeup gas adapter needs to be replaced.

[AICMM]

Lrouxph,

Is this a 5890 or 6890? As queenann notes, I would look at the split vent line first. If a 5890 I would solvent rinse and if a 6890 I would replace. Second thing I would make sure you have septum purge flow since this could easily build up phthalates that would respond nicely (and chromatograph) on an ECD equipped instrument.

Best regards,

AICMM

[Peter Apps]

Do you have scrubbers on your carrier gas line, and are they due for replacement ?

Peter

[Bigbear]

If you have been running this method for a while you may want to consider either "gun brushing" your injector or replacing it.

[eayres]

i found that certain vendors have changed the manufacturing process of their injection port o-rings. after this change, i started seeing a "ghost" peak on my ECDs. Something was baking off of the o-rings. It would actually collect on the head of the column and sometimes took a few injections to get down to just a single peak. when this happened, it took months for me to isolate it to the o-rings. in the mean time, i determined the elution temperature of the peak and started keeping all my ECDs' GC ovens above that temperature while idle. I mentioned my problem to the vendor, and got "we haven't heard of that from anyone" and a lot of trial o-rings. Luckily, it only affected my ECDs and I was able to use up the o-rings on my FIDs.

[thohry]

I would clean the purge line and swap out the purge cartridge next.

Ten years ago, I got this exact problem with a 6890. Finally we found the peak came from the purge line .

[aldehyde]

If inlet maintenance doesn't get it (I second the recommendation to change split vent line and check the inlet and the split vent line port for gunk) try removing the column from detector and using a blank nut to seal just the detector. rerun your method and see if you still get the contamination peak/baseline change. If you do its either the ECD liner, the make up gas adapter, or the ECD cell itself.

[Herbert West]

Triphenylphosphine oxide (TPPO) is a common contaminant that bleeds out of Viton O-rings, but it should go down in response as the O-ring gets conditioned. Unfortunately, by then it may be time to change the liner (and thus, the O-ring) and then you get it again. However, TPPO usually elutes relatively late in the chromatogram. Another suspect is bisphenol AF (a fluorinated bisphenol).

I'm presuming you have an Agilent split/splitless inlet, so you might try graphite O-rings instead. These are usually used for higher temperature work, but you can get away from background contamination with polymer O-rings by going to graphite. Just make sure that you get the proper ID (there are 6.35 and 6.5mm varieties depending on your liner OD). Be sure and leak check after installation.

You could consider the Merlin MicroSeal septum, too, which will eliminate any background issues from septum bleed, including eliminating septum coring and subsequent transfer of septum pieces to the GC inlet liner.

http://blog.restek.com/?p=3590

http://www.rtvanderbilt.com/documents/M ... it006a.pdf