Chromatography Forum

hello!

I am rather new in chromatography and that's why I have been practicing with different kinds of injection techniques (I mean splitless, with the valve opening after 0.5, 1, 2 min or spilt). I had also tried spiltess with the valve closed all the time.

Is it acceptable to have the purge valve closed during the gc run or am I going to have problems in the future (after injecting a nuber of samples)?

Georgia[/list]

When you inject a sample in capillary GC, the solute and the solvent vaporise in the glass liner of the injector. The volume of sample vapor will tend to mix such that it occupies the whole volume of the liner-maybe 1-2mls. If the split valve is closed, the only way the sample can escape is through the column. If the column is running at 1 ml/min it will take a very long time to clear the sample completely from the injector. Because the sample continually diffuses and mixes it will take much longer than 2 minutes to clear an injector of volume 2mls. The result is that you get a long trailing solvent peak as the solvent is gradually eluted from the column. The solute is usually trapped on the cold front of the column while this is occurring. Early peaks on your chromatogram may be obscured by the solvent peak. If you open the split valve, the residual solvent and some sample is flushed out by the rush of gas through the split (maybe 50-100mls/min). But MOST of the sample is introduced.

Why not try this? Close the split valve, inject the solvent and see what happens. When you get bored watching the solvent tail, open the valve and you will see the peak crash back to the baseline.

Victor's advice is good.

But to answer your question, it all depends upon your sample.

Certainly it is not difficult to open your purge vent after a run, before you inject another sample, and then reclose it if you wish.

Will it cause problems not to do so?

Maybe, maybe not, but it could. Why take a chance?

What would you gain in doing this? You are injecting into a closed system with flow paths that available to the sample but would never get flushed clean. Aren't you worried about reactions depositing material, or sample to sample cross contamination?

Depending upon the volatility of your sample composition you may not be able to refocus the sample 'plug' and your peaks will be broad and overlapping. Why good is that?

Reread what Victor has written. Take his advice, open your splitter vent and purge your injector system between samples.

best wishes,

Rod

But if you inject a solvent that expands a lot (eg: CHCl3) with the valve closed, you may get "flashback". This is what happens when the pressure in the injector builds quickly and the column isn't the only place for the sample to go. It goes up the carrier gas line and slowly desorbs over time, but not smoothly and gradually. It desorbs in chunks, usually whenever you've convinced yourself that you've got your new column's bleed down to an absolute minimum, and you inject your only aliquot of some important trace analysis sample... wham! Lots of poorly resolved peaks foul your chromatogram.

Flashback is not good.

If you have a considerable volume or a solvent that's prone to flashback, read up on "cold trapping". This technique coupled with slow injections can yield good focusing in otherwise marginal conditions.

DR-what you say is true and is a useful cautionary note, but it does not directly address the original question. Splitless injection is a technique where by definition, the split valve is closed during the injection period. The liner should be selected so that it has sufficient volume to contain the original injection, or the injection volume should be reduced until this condition is met. Most splitless liners are designed with this in mind. Splitless injection as recommended by Grob IS peformed with the column below the solvent boiling point-so "cold trapping" of solvent is inevitably performed. This effect will condense vapor from the vaporising chamber more quickly than expected from the simple argument I gave below, reducing the problem somewhat.

I am not sure why you selected chloroform as your example for generating flashback. Chloroform has a higher MW than hexane and will thus generate less vapor than hexane. The real problems start with very low MW solvents- like water for example.

Yes, you can make a splitless injection with the purge valve closed during the entire run by using a specially designed liner "Drilled Uniliner", which has a hole on the upper or lower part of the liner and a tappered opening at the bottom where you insert the column. The column end won't go through the narrow opening beyond 10-12 mm. We've tried it with a short (5m) capillary column for a highly reactive compound at ppm levels and it works without any noticeable 'peak tailing' or 'band broadening' for the compound or the solvent (Methanol). The liner ensures a tighter seal and the procedure is called a 'direct injection', since 100% of the sample is getting into the column, unlike a splitless injection (95% gets into the column) which requires opening the purge valve after 0.25 or 0.50 min. However, after several months an extra column 'band broadening' and a slight 'peak tailing' are observed indicating that it's time to change the liner. I'm not sure whether Dr. Grob has ever mentioned it in his monographs.

AZK

One problem with leaving the split valve closed during the entire run is the possibility of damaging your column. Early HP 6890s would close the split vent when the instrument came ready. This reduced flow to the sum of column flow and septum purge flow, and the low flow rate could allow air to infitrate the column. I ran a sequence starting Friday afternoon, and the instrument sat most of the weekend at 150C with the purge valve closed. When I tried to run Monday the column was trashed and had to be replaced. Opening the split vent increases flow through the inlet and not only cleans the inlet but helps protect the column.