Huge signal in GC-ECD

[Juan Pedro]

Hi!

I´m using a GC-ECD system. When I inject 500ppb standards it shows huge peaks and saturation, as if the solution was very concentrated (the same standards gave a normal signal 3 months ago. I´ve checked the standars and set the contact potential. Any idea?

Thanks a lot!

[AICMM]

Juan Pedra,

Things to look at. Assuming you have always run splitless, double check injection volume, column position, wool or no wool. Look at the make-up gas flow rate (ECD's are concentration dependent.) Try an injection by hand just to make sure it is not the autosampler.

Has the zero significantly changed?

Best regards.

[Juan Pedro]

Thanks for your reply!

The zero has not significantly changed. I was surprised because even the solvent peak increased a lot.

The people from Varian have told me that the most probable reason is that the cell from ECD is runing out.They avised me to calibrate again with the new peaks and when they become too big I should change the ECD.

Best regards,

Juan Pedro

[haiedc]

Thanks for your reply!

The zero has not significantly changed. I was surprised because even the solvent peak increased a lot.

The people from Varian have told me that the most probable reason is that the cell from ECD is runing out.They avised me to calibrate again with the new peaks and when they become too big I should change the ECD.

Best regards,

Juan Pedro

It sounds strange to me that before an ECD fails, it gives stronger response. Anyway, you can test the detector to see if the response is stable, then it may be a better ECD (before it goes to a complete failure) and you can rebuild the calibration and analyze samples as normal.

[Peter Apps]

Hi Juan Pedro

Dieing detectors do not usually improve in signal strength. Your first step needs to be to check the detector's current performance against the manufacturer's specification. Assuming 1 ul splitless injection you are putting 0.5 ng of analyte through the detector, this is a lot for an ECD and it should give you a big peak.

My suspicion is that you have inadvertently fixed a problem that was previously causing a loss of response. This might not have anything to do with the detector at all.

Peter

[fsistere]

Hi Juan Pedro

I have a ECD from Varian. I think it's contamination. You could try a termal cleaning. See the instructions in the manual.

[Zdenek]

Hi Juan Pedro

check leaks on ECD, already small quantum of oxygene increases signal.

[haiedc]

Hi Juan Pedro

check leaks on ECD, already small quantum of oxygene increases signal.

I do not think it's a leak because the signal is big, not the noise (baseline).

[Zdenek]

I have old ECD on HP 5890 II. When I start the instrument, signal on one of two ECD is big (signal on instruments display after push the SIG1 botton). As ECD is heated, signal decrease. Colleague tell me that is possible any leak on nut in supply gas capillary. Termal dilatation can wafer the leak. It confirm also servise engineer. I will try it.

It is one of many reasons why is signal high...
Good luck with solution yours problem.

[Peter Apps]

Hi zdenek

The "signal" that you are reading when there is no sample component going through the detector is in fact the baseline (i.e. background) noise. It is well known that oxygen in an ECD increases the noise background, but (as far as I know) it does not increase the signal when an analyte passes through the detector.

Peter

[AICMM]

FYI, small quantities of added oxygen have been used in the past to enhance the sensitivity of certain compounds like chloroethene. (Detectors for Capillary Chromatography, Hill and McMinn)

Which is not likely to be the case here in any way shape or form. Usually it is an issue of cooking off the water and air to get the detector to it's normal operating conditions and carefully guarding for leaks and impurities.

Best regards.

[Peter Apps]

Thanks AICMM, presumably there are some drawbacks or the use of oxygen-doped nitrogen as makeup would be standard practise.

Peter

[AICMM]

Peter,

Good point. Response to other things tank, for example CCl4 goes from 10k to 1.9 relative molar response.

Thus, only applies to a very narrow range of components and not the ones usually thought of for ECD.

Best regards.