Sudden tailing of non-solvent peak

[Tremenda]

Dear all,

I will try to explain what goes beyond my understanding...

Everything was fine until I decided to run some capacity tests and I injected
5 ul of ethyl acetate in an injector using the splitless mode in an attempt to get more sample into the column.

After that I run a sample also in splitless mode and the large peak (not the solvent) eluting before the analyte started to tail to such a extent that a third of the analyte peak is now below such tail.

I have though of contamination, so I changed septum and glass liner.
I tried swapping the ends of the column and I get the same tailing.
I have run a blank in the same conditions and I see nothing...but the peak keeps tailing...

Could anybody please give some insight into what is going on?

Regards

[chromatographer1]

First: what you did that was improper.

By putting 5µL into a splitless injector you overwhelmed the volume of the injector and deposited non-volatiles upstream, downstream, and all around the town through the vaporization of the ethyl acetate.

LVI requires special hardware and injecting technique. You may wish to investigate that in the future.

Now, with the non-inertness of the pneumatics due to said deposits acting as active sites, this can change the shape of the analyte plug entering the column after the vent is opened, ie tailing.

The other possibility is you damaged the surface of the column or the phase itself with the overload of the EtoAc solvent, producing active sites.

The trouble shooting should start with a new column. Did the problem go away? Then the column was damaged.

No effect ?

Then clean the entire injector, and the upstream pneumatics, included flow controllers, and see if that solves the problem.

You won't get more sample onto a column in splitless mode by injecting more liquid into an injector. If you volatilize 2µL into 500 µL and you are only flowing 1cc/min through the column then it takes 30 seconds for the entire sample to enter the column, and that does not take into account the volume must be less flowing in due to fact that the headpressure is higher than ambient pressure. One cc going in becomes 2cc going out at 1 atmosphere pressure.

Don't forget to clean the septum vent lines.

best wishes,

Rod

[Tremenda]

I have to say that I first looked at the HP flow calculator to see if a 5 ul injection of EtoAc would saturate the 900 ul capacity of my injector. According to it, a 93%. My attemp of getting more sample into the column via a larger sample volume was combined with an increase in the headpressure and an increase in the splitless time. I was expecting to optimise these three variables to maximise the sample entering the column. I hope this redeems my sin to some extent.

Now I just wonder how complicated it would be to clean the pneumatics as well as the septum purge...do I have to dismantle the whole thing?

I will try to change the column and cross my fingers.

Thanks for you advice.
I will post my progress on this issue.

[Bigbear]

Whoah , do not chean any of the flow controller parts! What you can do is go to the Agilent website and download a paper " Cleaning injection ports to remove active sites". Do that , and if it's a 6890 remove the copper split vent line and solvent rinse it ( or better yet replace it if you have a spare). Also replace the trap at the end of the split vent line. Put in a new gold seal and liner and you should be good to go.

[chromatographer1]

You are correct, those parts should not be dismantled. I meant to type 'excluding' and typed 'included' instead.

Thanks for catching this error.

Rod

[Tremenda]

Hi again,

Is is not the column unfortunately, and after baking out the injector at 300 ºC overnight, the tailing has not changed at all...

Therefore I will have to go the hard way...I am working with a Varian 3400 GC. Does anybody know of something specific for this machine or is the agilent article generic enough to be applied everywhere?

Thanks for your help

[chromatographer1]

You did replace the column or put the column into another GC to test it?

The other issue with the tailing, is an improper installation of the column, at either end, or possibly inadequate makeup gas at the detector.

Just a few more things for you to verify.

You did test the column in another GC, right?

Rod

[Peter Apps]

Before you start disembowelling the pneumatics let's get some more details on the hardware and procedures.

What kind of liner do you have in the inlet ? , what temperature is the inlet set to, what is the splitless time ? NB that Varian liners go in with the narrow bore at the top, the exact opposite to Agilents. Have you checked that there is no dirt in the bottom of the inlet that would not be removed when you changed the liner ?

What are the dimensions and stationary phase of the column, what carrier gas are you using, and what is the flow rate ?

You specifically mention that it is a large peak (not the solvent) just before the analyte that tails, does this mean that the solvent peak is not tailing ? If it is only the one peak, what compound is it ? Do you see any tailing on any other peaks ?

Peter

[chromatographer1]

Ah ha !

You are working with a Varian 3400. A good workhorse of a GC.

The injector has a different design than the supposed but not professed Agilent GC.

Be certain the injection liner is installed properly, mounted flush into the heating block with no dead space or graphite exposed.

You may wish to place the gas lines instead of cleaning. I have found that to be a cheaper and better solution than trying to clean them. And a whole heck of a lot quicker.

Be sure you use both acid and base, both non-polar and polar solvents, to clean the lines. and use a final rinse of methanol with a drying of inert gas before you reinstall the lines. I have found glacial acetic acid to be a great cleaner for contaminated metal. Sodium hydroxide in methanol is also a good cleanser (use it before the acid).

Oh, and by the way,

You do have a clean and unblocked vent trap?

best wishes,

Rod

[Tremenda]

Dear Rod,

I do not have access to another GC, however I have a shorter column of exactly the same stationary phase. instead of 50 m it is a 15 m one. To be honest, it is hard to tell whether the same tailing takes place as everything is so clogged.
I could buy a new column, but I thought it would be faster and cheaper to get a few brushes and follow the relatively simple procedure for cleaning the injector and if that does not work, then get a new column of the same or even larger length. What are your thoughts about this?

Thank you.

Dear Peter,

I use 4 mm ID glass frit liners. The inlet temperature is 200 ºC and the splitless time 0.8 min.
I have not opened the bottom nut that would give access to the bottom of my 1077 split/splitless injector, yet. So I think that if I do it, I could take the opportunity to clean the whole thing with the brushes and the three solvents.

http://www.chem.agilent.com/Library/Sup ... a16022.pdf

I am using a 100% dimethylpolysiloxane phase colum 50m x 0.32 mm x 5um. I use nitrogen and the headpressure is 20 psi. It is a bit cumbersome and time-consuming to measure the column flow in my system, so I adjust the headpressure to find a compromise solution between resolution and analysis time.

The solvent used to tail, but I sorted that out by increasing the column flow to its maximum. When the splitless time is over, a massive flow of gas cleans out completely the injector chamber and no tailing occurs.

The tailing peak has to be a derivatisation by-product with MBDSTFA, such as TFA. I am not sure yet. Is it important to find out what it is?

Everything was fine until I did a large injection of EtoAc. It is the only peak tailing, the others seem to be fine, actually hidden under the tail of this peak.

Regards

[Tremenda]

Dear Rod,

I do not have access to another GC, however I have a shorter column of exactly the same stationary phase. instead of 50 m it is a 15 m one. To be honest, it is hard to tell whether the same tailing takes place as everything is so clogged.
I could buy a new column, but I thought it would be faster and cheaper to get a few brushes and follow the relatively simple procedure for cleaning the injector and if that does not work, then get a new column of the same or even larger length. What are your thoughts about this?

Thank you.

Dear Peter,

I use 4 mm ID glass frit liners. The inlet temperature is 200 ºC and the splitless time 0.8 min.
I have not opened the bottom nut that would give access to the bottom of my 1077 split/splitless injector, yet. So I think that if I do it, I could take the opportunity to clean the whole thing with the brushes and the three solvents.

http://www.chem.agilent.com/Library/Sup ... a16022.pdf

I am using a 100% dimethylpolysiloxane phase colum 50m x 0.32 mm x 5um. I use nitrogen and the headpressure is 20 psi. It is a bit cumbersome and time-consuming to measure the column flow in my system, so I adjust the headpressure to find a compromise solution between resolution and analysis time.

The solvent used to tail, but I sorted that out by increasing the column flow to its maximum. When the splitless time is over, a massive flow of gas cleans out completely the injector chamber and no tailing occurs.

The tailing peak has to be a derivatisation by-product with MBDSTFA, such as TFA. I am not sure yet. Is it important to find out what it is?

Everything was fine until I did a large injection of EtoAc. It is the only peak tailing, the others seem to be fine, actually hidden under the tail of this peak.

Regards

[Peter Apps]

If TFA is trifluoroacetic acid then the tailing is due to adsorptive activity - the other peaks are not affected because they do not have polar groups. This argues against depostits of non-specific gunk.

Your carrier pressure seems very high, even for a 50 m column, you could get a better trade off of speed and resolution by using a shorter column. A quick way to measure flow rate is to inject gas from a cigarette lighter and measure the retention time of the peak.

The inlet temperature is quite low, and this may be why you need a "massive" flow of gas to flush the inlet at the end of the splitless period. How big is massive in ml/min ?

The stationary phase in the column is thicker than I would have expected for derivatized compounds- which tend to have high MW. What is the temperature programme ?

When you put in the short column did you lower the gas pressure proportionately ? If not everythng woul dhave come racing through with very little chance of diagnosing peak shape.

Can you post chromatograms ? Instructions are in a sticky at the top of the page.

Peter

[Tremenda]

Before:


After:


The sample is the same, but in different days, so some sample degradation occurs, which does not affect the analysis of the problem.

I have just realised that I got tailing too in the sample before the 5 ul injection of ethyl acetate, which made me go through the logbook and see that I had had a problem with the septum purge only three injections before. However, I only saw something anomalous in the fourth, right before the solvent injection.

The septum purge in my Varian 3400 is a metal coil ended in nut that tightens a valve, but it does not look like a needle valve. I am saying this because in the manual there is no reference whatsoever to this important part of the GC, and therefore I do not know what I am dealing with at the moment.

The nut loosened after increasing the headpressure to 20 psi, and I was losing sample via the septum purge. Therefore, I tightened it, but I cannot control the flow precisely. If it is tight, only a tiny flow comes out (less than 0.1 ml/min). If it is too loose it leaks.

I have tried to run a sample with the septum purge valve wide open, and the tailing still occurs.

Could the septum purge be the source of the problem?
I wonder whether I could post a pic of the valve regulating this purge...