gas hold-up time?

[trigger]

hello,

I am just starting to use the GC-FID that has been in use in the lab for some time and has some problems described as follow by my colleagues:

The gas hold-up time changes depending if we switch from outdoor sample to calibration mixture. (and therefore all retention times are shifted, but apparently adjusted retention time are the same)

For both analysis all other parameters are the same.The only change is to disconnect from outdoor sampling and connect to the calibrator output.

I am just wondering where that could come from? very likely there is a difference in pressure: sample (atmospheric pressure) and calibrator output (??).
but there is also a GC-PID that uses the exact same path and is not affected.

Can we still trust the results and work like this? I personnaly think not and want to solve the problem.

Any help would be useful.
thanks

[chromatographer1]

Are you taking your injection from the same location, the same valve?

If not, then any change in the length of flow path and the amount of gas contained in even the same path but at different pressures will cause a timing change.

Can you describe your sampling arrangement in more detail?

Is this a new problem? What changed if so?

best wishes,

Rod

[AICMM]

Trigger,

How/where are you doing your injection? Syringe, gas sampling valve, some other way?....

Best regards.

[trigger]

Hi,

Thanks for your replies.

The sample goes through water removing -> particle filter -> to the GC.
The calibration mixture goes through calibrator -> water removing -> particle filter -> to the GC.

a small membrane pump in the instrument input the sample to a sample loop.
35ml sample gas are preconcentrated on a Tenax trap. This step is repeated 10 times and then the trap is purged by carrier gas (to remove water).

The sample is desorbed by quickly heating the preconcentration tube while flushing it with the carrier gas.

all of these steps are exactly the same for both calibration and sampling.

The problem has been here from the very beginning. And they have observed a shift of hold-up time of 1 to 2mins between calibration and samples.
This instrument is used for on line sampling of outdoor ambient air, and the hold up time changes from site to site. Apparently the shift is more important with higher relative humidity.
But for the calibration the hold up time is very stable in time and has not changed.

I dont't understand, because the injection depends on the desorption of the preconcentration trap, so how the former steps can affect this?

could it be that because RH is higher then purge takes longer... but then what?

Thanks for your help!

[MikeD]

For a better diagnosis you may need to give even more detail eg. temperature of Tenax trap in preconcentration step, carrier gas split ratio, what range of ambient air components and if the system is homemade in-house or commercial. If commercial a name would be useful. Could you confirm that your calibration gas is also added in 10 steps totalling 350 ml and that it is presumably dry ?

Based on your information that it is humidity dependent I would have to guess that water is not completely removed from the sample and at the moment of high temperature desorption the effective split ratio (if there is a split) is different for sample and calibration gas. Therefore the real column head pressure is different. If this is the case you are right to be concerned about the validity of the calibration.

Going back a few decades as I can, calibration in single-stage thermal desorption for ambient air measurements was OK for packed columns when everything was splitless. Calibration with capillary columns worked better with 2-stage thermal desorption.

[Ron]

The are tables on the SIS web site to show the breakthrough volumes for analytes on various materials. I agree that the most likely cause of your problem is inadequate dry purge time or flow. You can use the tables to calculate the approximate volume of gas needed to remove water without losing analytes. This can be difficult depending on the volatility of the analytes.

You are using a very large sampling volume. Is it necessary to use that large a volume? If so, you might want to consider a different material for the trap. You may be approaching the breakthrough volume for some low molecular weight analytes.

[trigger]

The GC is a commercial instrument and has been mainy manipulated by engineers from the company it was purchased until now.

And it seems that many parameters are not appropriate.

I think you're right the sampling volume is too important.

I now have several things to check and try, I probably will come back with more questions later.

Thank you very much that was very useful!!

[MikeD]

Another thing to consider is the maximum temperature of the Tenax trap you can get away with and not lose the volatile fraction of your analytes. In other words if you are using liquid nitrogen cooling you are trapping all the water, but at -10 degC not very much.

The SIS tables are quite useful for comparing the effect of temperature and the differences between sorbents. However I would allow x 0.5 to x 2 latitude on their absolute values because they use a special definition of breakthrough volume and their test trap dimensions are not necessarily the same as yours.
http://www.sisweb.com/index/referenc/breakthrough.htm