HP 5890 - FID Detector Problems

[Dank]

I am direct injecting (solvent flush) into an HP 5890 Series II FID and when it reaches around the 1.5 min mark of a run the detector signal drops to 0 and I have to re-light the detector. I have made sure that the detector is lit prior to the run. My signal is ~15 . Was ~19 prior to the first time this happened, signal has dropped since then when I re-light. I checked my flows and they are:

H2 carrier ~ 50 mL/min
Makeup (H2) ~ 10 mL/min
Air ~ 400 mL/min

The standard that I am shooting is a 50 ppm. Methanol, Ethanol, Ethyl Acetate in H20.



Thanks,

[Dank]

Forgot to mention: Oven Temp 40 C. Detector 280 C . Injector 255 C

When I turn the oven Temp to 80 C. This is no longer a problem. What is the likelihood of the water condensing and extinguishing the detector?

[Bigbear]

I feel the water is quenching your flame. Also your gas mix seems a little rich to me.
The make up flow seems a little low, 30-60 mls. is what I recall is suggested.

Can you run this procedure via headspace. You would get less water in the system which would also improve the chromatography .

[Dank]

Thank you bear,

This is what I feel as well. In order to confirm this suspicion further I turned the oven temp to 30 C and ran a propane std through a gas sampling valve. Detector did not turn off. When I ran the liquid solvent flush std....detector went out. Tried to confirm this with my boss.

He asked me "Do you really think 30 C - 80 C water would have any impact on a 280 C flame" .. I believe yes from what I have been seeing. But he thinks that I have a detached flame and is not lit in the correct spot. I have lit the flame repeatedly and have established a consistent signal.


So my question: Can water actually condense in the column or near detector to actually extinguish it" The amount of water that seems to "extinguish" it is .1-1 uL

[Consumer Products Guy]

Back in the packed column days, it wasn't uncommon for us to extinguish a flame when using water as the injection solvent. I see yours (50ml/min hydrogen) must also be packed column.

[Bigbear]

If you are shooting 1ul of water into your system it will flash to about 1200ul of vapor. That will cary "junk" outside your liner and most likely into the split vent line.
I doubt that your flame is detached flame ( do you have the correct jet installed?).

I again ask if you can change to headspace? If you are running a manual injection now it would only entail some sealed vials that you could incubate at a set temp. For your compounds you could use a water bath at 60-80C.
You could add some N-propanol as an internal standard and if you wanted could add something like t-butanol for a surrogate.

[Peter Apps]

Packed column or capillary ? If capillary you have a plug of condensed water that is pushed through the column and when it hits the hot detector it vapoprises and blows the flame out. This is the second time in a couple of months that this exact problem has been on the forum.

Peter

[aldehyde]

chance that it is the water: close to 100%

its possible it could be something else, but as soon as I read your question I started scanning your post for "..in water".