GC-ECD Baseline Noise Help

[noTech]

Hi Everyone,

I'm trying to get an Agilent 7890A GC with ECD detector up and running with very little GC experience (please forgive my lack of knowledge - it's trial by fire here!). The system has been sitting around for a few years. I recently changed the column and ran some samples, but am seeing a very noisy baseline. A picture of my blank solvent run (MTBE) is attached and my baseline is noisy with other solvents.
I've changed out the inlet liner and syringe and checked for leaks. Does anyone have an ideas of additional steps I should take?

Much appreciated!

https://postimg.org/image/i7kc0yuzf/

[tkubowicz]

Hello

I can't open attached pictures however if uECD hasn't been used for a while perhaps it needs good bake out.
Please check link below for procedure:

http://www.agilent.com/cs/library/suppo ... a15698.pdf

Good luck

Regards

Tomasz Kubowicz

[dblux_]

Hello

I can't open attached pictures however if uECD hasn't been used for a while perhaps it needs good bake out.
...

Then run a blank run (without injection) and show us the chromatogram.

[Peter Apps]

Welcome to the forum.

The source in an ECD has a limited lifetime whether you use it or not. Check how old yours is, and how long it can be expected to last. If it is past its use by date then you need a new/ refurbished detector.

Peter

[dblux_]

Welcome to the forum.

The source in an ECD has a limited lifetime whether you use it or not. Check how old yours is, and how long it can be expected to last. If it is past its use by date then you need a new/ refurbished detector.

Peter

Peter, this is Agilent 7890A GC introduced I guess in 2011.
And Ni 63 has a half-life ca. 100 years.

[Peter Apps]

It's not the half life that's the issue. It has to do with diffusion of the radioactive nickel into the foil as well. If I recall, ECD lifetimes are given at around 10 years (someone can check with Agilent) so after 5 years it should still be good.

[01310040231]

I contacted Agilent and they said the average life for a micro ECD is 4-10 years depending on its usage. Our oldest micro ECDs are on a 6890 and they are 17 years old. We have over 180k runs on these detectors and our background noise is almost the same as 17 years ago.

[James_Ball]

Hi Everyone,

I'm trying to get an Agilent 7890A GC with ECD detector up and running with very little GC experience (please forgive my lack of knowledge - it's trial by fire here!). The system has been sitting around for a few years. I recently changed the column and ran some samples, but am seeing a very noisy baseline. A picture of my blank solvent run (MTBE) is attached and my baseline is noisy with other solvents.
I've changed out the inlet liner and syringe and checked for leaks. Does anyone have an ideas of additional steps I should take?

Much appreciated!

https://postimg.org/image/i7kc0yuzf/

What is the temperature program you are running in that chromatogram?

Does it have the same baseline disturbance in the middle of the run if you use Hexane as solvent?

It looks like it could be some kind of flow disruption during the run, are they any timed events with the flows taking place?

[Peter Apps]

I contacted Agilent and they said the average life for a micro ECD is 4-10 years depending on its usage. Our oldest micro ECDs are on a 6890 and they are 17 years old. We have over 180k runs on these detectors and our background noise is almost the same as 17 years ago.

How does the signal per unit mass of a test compound compare then and now ?

Peter

[Peter Apps]

I can see the chromatogram now. The humps almost look like ghost peaks, or contamination accumulating on the column and then bleeding off as you programme.

If you have not done so already, put a new oxygen and moisture scrubber on the carrier gas line just upstream of the GC.

Peter

[Yama001]

Note that negative peaks (the plunge below baseline) often come from vials that have been contaminated with septum particles from repaeted injections.

The bakeout intructions that are linked above are worth following. It can be useful to use hydrogen as the make up gas during the bakeout.

[01310040231]

I contacted Agilent and they said the average life for a micro ECD is 4-10 years depending on its usage. Our oldest micro ECDs are on a 6890 and they are 17 years old. We have over 180k runs on these detectors and our background noise is almost the same as 17 years ago.

How does the signal per unit mass of a test compound compare then and now ?

Peter

Peter:

According to the Agilent manual if the background signal is below 200 then the detectors are in excellent shape. Our background was 130 & 140 when the system was installed 17 years ago and it is 180 & 190 today. We use the Agilent RMSH-2 trap for the gas line and it lasts about 20 months.

[noTech]

Thanks so much for all of your responses. I baked out the ECD as suggested and ran my method both without injection and with hexane:

No injection: https://postimg.org/image/adr90oxyb/
Hexane injection: https://postimg.org/image/cbcdyu0jf/

This seems much cleaner than before but still noisy. I recently changed out the syringe and liner and only use clean, new vials.

The method is as follows:
33 C held for 14 mins
5 C/min to 60 C, held for 5 mins
15 C/min to 275 C, held for 20 mins
No timed events

Any ideas as to what the huge dip in signal might be from 5-10 mins when I run hexane?
I have not yet put an oxygen/moisture filter on the carrier gas line as I'm waiting for a new one.

[tkubowicz]

Hello

Check below:

1.Graphite ferrule for uECD makeup adapter - replace it. It might be a leak...you have baseline disturbance when temperature in oven is going up.
2.Nitrogen for uECD must be pure. Make sure your N2 source is really clean.

Regards

Tomasz Kubowicz

[Bigbear]

Also check the ecd make up adapter for leaks. The weld where the N2 pipe enters is very weak.