For 2 full days I have been trying to track down the source of a couple of ghost peaks on my newish 6890N w/ECD detector capillary column run in split mode. The part that mystifies me most is when I make a blank run oven ramping only no injections... the same two peaks at the same retention times... too strange! I've changed septum/liner/o-ring twice. The peaks are pretty well defined and do not decrease over successive oven ramps. What's happening??

Contamination is entering the column inlet in the injection port.

Where is it coming from?

Contaminated carrier gas, perhaps.

Backflashed material from previous injections can be another source.

More information about the peak area (is it the same or changing as time between oven ramps varies) and the elution temperature and capillary column used might be helpful in diagnosing your problem.

Good luck.

Additionally,

1. has ANYTHING changed recently? e.g. the carrier gas or the cylinder.

2. see if the peak size changes with altered split ratio

3. try blank runs with the injector cold

Regards,

Ralph

The elution time was 1 minute later after extending the hold time 1 minute on my blank oven ramp. The column is DB5 15 meter x 0.25 x 0.25 uF.
Split ratio of less yields smaller mystery peaks. Other oven ramps starting at 250 ramping to 300 don't show the peaks. My test oven ramp starts at 120 and ramps to 280.

I think it is time Rick changed his purifier cartridges on his carrier gas. Do you have these Rick? You are getting heavy not volatile contamination (oils?).

I bet you got a cylinder that was contaminated (maybe quite a while ago) and the garbage is just now getting through your traps and purifiers.

Clean all stainless steel lines, pressure regulators, valves, etc and make sure your carrier gas is clean.

The peaks will go away when you clean their source.

Good luck.

The idea of running a blank oven run with the injector cool was an interesting one and in fact the unknown high boiling peaks cleared up. This would seem to tell me that the column onward through detector is not the source of my ills. I believe that I had cleaned the split line and injector port with various solvents and a brush... so either my cleaning was somehow ineffective or I have bad gas recontaminating the inlet. I also seriously wonder about the 6890N and the Pneumatic pressure lines and hardware and how they could be responsible.

I originally suspected the carrier but, to review,

1.an oil contamination would give more than two peaks

2. They show when starting at a lower oven temp, so pointing to them condensing on the column between runs.

3. none with the cold injector suggests the injector itself - if it was contaminated carrier it would still be getting onto the column to produce peaks BUT..

4. A lower split ratio reduces the peak size - puzzling if the source is the injector - I can't work that one out

Do you have another GC and if so, is it behaving normally?

Ralph

I originally suspected the carrier but, to review,

1.an oil contamination would give more than two peaks

2. They show when starting at a lower oven temp, so pointing to them condensing on the column between runs.

3. none with the cold injector suggests the injector itself - if it was contaminated carrier it would still be getting onto the column to produce peaks BUT..

4. A lower split ratio reduces the peak size - puzzling if the source is the injector - I can't work that one out

Do you have another GC and if so, is it behaving normally?

Ralph

A contaminated gas carrier line from backflashed samples would be consistent with the evidence presented. A cool injector would not vaporize this material onto the column. The carrier gas would be clean until it passes by the 'heated' portion of the contaminated 'backflashed' injector port line.

Disconnect the carrier line going into the injector port after the EPC or pressure regulator and replace or clean it thoroughly.

If the other injection port does not show this problem the source is not the carrier gas.

Avoid injecting too quickly or too large a liquid sample solution into an injection port as the flash of the vaporization can deposit and degrade your sample just upstream of the injection port where it can slowly thermally decompose from heat of the injection port and cause the peaks you are seeing.

Rodney George
Senior Research and Development Scientist
Gas Separations Research
Supelco
595 North Harrison Road
Bellefonte, PA 16823

814-359-5737 voice
814-359-5459 fax

Have you checked your pressure regulator? Is it new? This must have a clean stainless steel diaphragm. If you have a heated purifier downstream of your regulator, this will crack most impurities from a (shudder) neoprene diaphragm into some fairly volatile compounds. We have also been shipped tanks with greased threads. These will give many peaks for a long time without a filter, but I have no experience with a heated filter in this case.

Have you checked your pressure regulator? Is it new? This must have a clean stainless steel diaphragm. If you have a heated purifier downstream of your regulator, this will crack most impurities from a (shudder) neoprene diaphragm into some fairly volatile compounds. We have also been shipped tanks with greased threads. These will give many peaks for a long time without a filter, but I have no experience with a heated filter in this case.

A long time ago I had a problem with a large ghost peak during each injection (ECD). I finally found this was caused by the liner vendor changing their cleaning and silanizing methods. I found if I disconnected the column, and connected a little "pigtail" of column to the injection port and allowed the carrier gas to go through the inlet at normal operating temperature overnight, the contamination would go away. I also found that the contamination from the liner would contaminate the whole injection port -- you could switch to a good liner, and still have the contamination.

This may not be your problem -- but that's what happened to me.