Improving headspace GC method; problem with separation

[MestizoJoe]

Hi,

I'm trying to improve my headspace GC method for ETO.

I have made some small changes that improved sensitivity and peak shape. Now I have one peak co-eluting with my ETO early in the method.

My solvent is water. I am doing a splitless injection with purge flow to vent at 0.01 min. I believe that if I can inject the headspace gas faster onto the colum I might be able to get these peaks resolved.

My question is how can I inject the headspace sample faster onto GC?

My vial parameters are

Fill mode: Flow to pressure
Fill pressure: 15
Fill flow: 50
Loop fill mode: default
Loop fill ramp rate: 20
Loop final pressure: 10.05
Loop equilibration: 0.05

On a 7697A

Can I just increase my loop fill ramp rate?

Thanks,

[chromatographer1]

If you are injecting splitless, then the headspace sample enters the injection inlet at the column flow rate and gas enters your column at the same rate.

You would have to increase the column flow rate to get the sample to onto the column faster.

Or am I missing something?

Rod

[Don_Hilton]

What kid of column are you using for the analysis - and what column conditions? Depending on the type of column, trapping analytes at the head of the column to focus the peak comes to mind, if possible.

[chromatographer1]

He has not mentioned having any subambient techniques available to him which would allow him to refocus the sample plug, but he seems to want a faster loading of sample onto the column.

Is that right, Joe?

Rod

[Peter Apps]

The only thing that you achieve with the very short splitless time is to cause pressure and flow fluctuations in the inlet. You would probably do better with a simple split injection as long as all the gas that goes into the column or out through the split is supplied to the inlet via the headspacer. If this is the setup that you describe in your other threads then you may be mixing the gas that comes from the headspacer with carrier gas coming via the inlet feed line. This will reduce the flow through the loop and transfer line, giving a wider (in time) plug of headspace sample diluted with carrier gas going onto the column, which is bad for both separation and sensitivity. So this is all about gas control - what goes where at what flow rates and pressures - study the operator's manual.

I am assuming of course that the two peaks you are trying to separate actually can be separated on the column that you are using. Do they separate if you do a simple liquid split injection ? - if not you need a different column.

Peter

[amazhad]

You will want to evaluate your splitless headspace method carefully. There is a reason the 7697A headspace Advanced Operation Guide says "With any inlet type, the HS supports only split inlet modes. Splitless inlet modes are not supported."

Not supported is not the same as saying it doesn't work, it will if the conditions are right (correct hardware and method setup). The basic problem with headspace splitless methods is in a splitless setup it takes a long time for the sample to get from the headspace to the inlet, causing peak broadening. I'm assuming this is why you ended up with a 0.01 minute purge time.

What you are running is not really splitless, with a purge time of 0.01 minutes you are basically doing a split injection 0.5 seconds after the sample loop switches to inject the sample. In a 7697A EPC flow control setup, how fast your sample travels from the headspace to the inlet in a split injection is a combination of the septum purge, purge/split vent, and column flows. In a splitless method how fast the sample travels is a combination of the septum purge and column flow.

I would expect if you take your current purge vent flow rate, and use that to run a true split method where the split vent flow is the same as your splitless purge vent flow, you will see similar results.

Before you try adjusting anything else, stick either with a true splitless or true split method. Don't use one inlet method to emulate the other. This means for splitless, the purge time has to be set long enough for the sample to reach the inlet and enter the column.

[Don_Hilton]

The comment on trying to focus the peak was that the objective seems to be better resolution - and sharper peaks can be easier to resolve than broader peaks. Large volumes passing through a column can give band spreading - fast or slow introduction to the column. The reason I asked about the column was to see if there was a chance of being able to reduce band spreading from injecting the desired sample volume all ont the column - by oven temperature or by some other means.

Splitting the the sample to inject whith higher flow through the inlet has some strong advantages - at the loss of some sample out the split vent. If there is enough sample, the split injection would be preferable.

Whe I hear splitless, I tend to assume (bad move) that we are going after sensitivity. If that is the case, the problem is to cram a lot of sample volume into a small space - and make it all small by the time the chroamtographic run starts...

[Peter Apps]

Questions that I forgot to ask earlier - how wide are the peaks that do not separate ? and how far apart are they ? Better yet, post a chromatogram.

Peter

[MestizoJoe]

The column is a DB-wax starting at 60C. I tried starting at 40C, but this ruined the peak shape because ETO is very volatile and Water (the solvent) is not.

I do want faster loading onto the column, but if the only way to do that is increased flow, I think I would still have the problem of coelution of a peak with my ETO.

Peter, I do get resolution using this system with another method, but that method gave me high RSD. I will check the manual and try to understand how the gas flows through the system, but yes, the two peaks are resolved using the other method.

I see what amazhad is saying. Yes, I did decrease purge time to try to sharpen peak shape and I see how that is just emulating a split.

Don, I thought if I could focus the peak I would acheive the resolution that I saw when I used a split method (with high RSD but good resolution).

The wideness of the peak is 0.16 min at 2.19 min RT. With the high rsd method the width was 0.09 near the same RT.

Here is an overlay. Thank you guys.

[Peter Apps]

If this is with the 75 m column in the other thread then you must have a howling gale of carrier gas going down it to get those retention times. Let's save a lot of back and forward- post all the pressures, flows, inlet liner type, column type and dimensions etc etc etc. then we know what we are trying to fix.

What is the huge peak eluting after the two small ones ?

Peter