does anyone use liquid injection to calibrate gas injection?

[zimanli]

Greetings.

We are trying to quantify gas injections by using standard liquid solution injections to calibrate. We understand that it might not be fair to calibrate sample with standard in different phase, but we do not know the different phase effect. Does anyone have any experience with it?

Thanks,
ziman

[Peter Apps]

Hi Ziman

At best this would be difficult to validate (to put it mildly) and the validity would depend on exactly which compounds you are analysing, in what matrix, by what method, using what standards dissolved in what solvent etc etc etc.

In other words we need to know the details please.

Peter

[MikeD]

I have some experience with this, but not a good one. If your inlet system was genuinely splitless it ^might^ be valid, but don't bank on it.

If there is a column inlet split and there was some way of separating the substance of interest from 99.9% of the solvent well before the split point then it ^might^ be valid. A miniature sorbent trap might do this, but don't bank on it.

Otherwise you are wasting your time. Rapid volatilization of the solvent changes the split ratio relative to pure gaseous split.

[jimmyhu]

Hi Ziman,

I have a very very approximate way of doing this analysis when a "semi-quantification" is required here, and cost and time are in priority.

Things need to be considered, such as that the expansion in the injector may not be reproducible, some will be lost via the split (splitless must be in use) and also via the septum purge. Also the actual temperatures and pressure of expansion may be different inside the injector.

For example, a liquid of toluene standard in CS2 is prepared, and 1 uL of aliquot is injected in a spitless injector. This liquid (1 uL) will be expanded into 322 uL in gas in injection port using HP Pressure/Flow Calculator. A calibration curve can be obtained for toluene with the knowm amount of toluene and the integrated area obtained from GC run. Then 500 uL of gas sample from an airbag in which the toluene is present is injected.

If the quantification of the gas sample based on the above calibration curve is given to be 100ug, for example, so a factor of 0.644 from 322uL/500uL is used to correct the raw data of 100ug. The final result should be 64.4 ug (100 x 0.644) of toluen in the sample bag. Does it sound OK?

Cheers
Jimmy