ECD problem

[rtanco1]

I'm using ECD in a HP5890 II GC. The signal is in the range 20-30 so it is OK. If I open makeup gas valve and inject solvent (i.e. methanol) and I don't see the peak. If I close make up gas valve and inject I see the peak of methanol but after that the baseline starts raising slowly and then goes down slowly after 10 or 15 minutes.

Does anybody have an idea? Maybe the detector is dirty? try to bakeout detector?

Thank you very much.
Raul

[dblux_]

But the idea of ECD is that it should not be sensitive to methanol.

[rtanco1]

Yes, but if used as solvent I should get a peak because it contains oxygen. Anyway, I inject dichloromethane and no peak appears. Thank you.

[Steve Reimer]

If you don't get a very large peak with a dcm injection, either your ECD isn't working or your column isn't connected properly.
What is the normal baseline signal for that detector and how old is it?

[rtanco1]

The point is that if I close makeup and anode purge valves I get the peaks although with low sensibility. When makeup and anode purge valves are opened, baseline signal goes down but no peak appears.
Thanks. Raul

[dblux_]

The point is that if I close makeup and anode purge valves I get the peaks although with low sensibility. When makeup and anode purge valves are opened, baseline signal goes down but no peak appears.
Thanks. Raul

That's the way makeup gas works. It dilutes your analytes.

[rtanco1]

But, why Agilent recommends makeup gas flow of 60 ml/min ?
Thank you. Raul

[gcguy]

My understanding of the design of Agilent detectors is that the make up gas is to sweep the flow from the column rapidly to the point at which it is measured to reduce the chance of the peaks becoming too wide.
If you can make your ECD work without make-up gas and as long as it is reproducible and the peak shape is ok then don't use make up. I suspect that the baseline will also become noisy with time though.

If the make up gas dilutes your sample below the LOD you require then make your sample more concentrated or increase the injection volume or reduce the split flow.

GCguy

[dblux_]

If there is no methylene chloride peak than something went wrong. You have to find what.
Did it happen suddenly ?
Were you doing any maintenance works just before ?

Seems you completely lost ECD sensitivity.
You may find standard Agilent procedure to check ECD sensitivity and criteria to sensitivity assesment.

GCguy already explaind you the purpose of makeup gas in case of capillary columns.

[pierre514]

Hi, sorry for asking but, are you injecting in a capillary colomn, a wide bore colomn or packed column? Because if you are injecting in a packed colomn and or a wide bore you may have enough flow in the colomn so you don't need make up at the detector.
Next you can troubleshout for your ecd with a proper sample witch gave you good results before, good luck

[Steve Reimer]

If you don't see a peak with a dcm (methylene chloride) injection you won't see a regular sample. On the last ECD system I used (6890 with micro ECDs) a dcm injection would peg the detector for several minutes, if you aren't seeing a response something is broken or disconnected.

[rtanco1]

Thanks a lot for your answer. I'm using capillary column (0.25mm ID). When I close make up gas I can see peaks but they are wide and the sensitivity is low.When make up is opened i can`t see any peak, not even dcm peak.

May be ECD is contaminated? Thanks again

[AICMM]

rtanco1,

Let's start with the fundamentals. What are you trying to measure and at what levels? ECD may not be the detector of choice.

Then, let me toss this out. You are using a o.25 id column right? Flow is perhaps 1-3 mL/minute? Way, way too low for ECD w/o make-up, just for the flow through the detector perspective. Have to add make-up, either nitrogen or P-10. Likewise, if you are using just helium for the make-up, won't work because the ECD has to have a moderating gas to convert high energy electrons to thermal electrons.

An answer to the first part is very important.

Best regards,

AICMM