LVI woes: peak splitting

[tangaloomaflyer]

I am running a 6890/5973 with an Optic programmable inlet. I am injecting 25ul pet ether, with a typical LVI program (solvent vent split/splitless transfer/split - but i am not using a keeper solvent). My first internal standard has slowly become split. Recently, it became a lot worse and my first eluting compound is now splitting.

First I thought it was the liner. We recently tried changing the cleaning method to include a chromic acid wash. After replacing the liners i still had the problem. to confound things further i am getting what looks like a hydrocarbon footprint on my autotune and a CO2 peak at around 20% my target ion. This contamination happened on another Optic instrument (not sure if it is still affected as this instrument is not being used by me anymore)

I tried to clean the split lines, by bumping the split flow to 200ml/min and heating the inlet to 500degrees. When I got it going again, my retention times have shifted back and I can't resolve my first peak!

What causes peak splitting? Am i right in thinking it is excess solvent reaching the column?

[Peter Apps]

Excess solvent on the column can cause peak splitting, but then so can a host of other things. If you want advice on how to avoid it we need details of temperatures, flows, column phase and dimensions, and what the analytes are.

Peter

[little_brad]

use liquid Co2 to cool the head of the column which will focus the peaks or use a guard column

[danicrd]

Try to increase a little bit the vent flow. Normally would stop to split peak
Good luck

[Blossom]

If the peak splitting is worse for earlier eluting peaks and the column is being temperature programmed, then this is usually a sign that the column is overloaded with solvent. To solve this problem use a retention gap (empty piece of fused silica between inlet and column) with a polarity similar to your solvent or adjust the parameters you use to vent the solvent prior to injection, to reduce the amount of solvent entering the column. In my experience, the time required for venting would be the one to change, assuming of course that the temperature at which you vent the solvent has been optimised.
This problem sometimes gets worse with time if you are analysing dirty samples: residue builds up and reduces the rate at which the solvent evaporates.

[tangaloomaflyer]

Thanks,

I've tried playing around with the vent flow, but the advice I was given was to reduce it, but increasing it does make more sense.

The thing is though, i have a solvent monitor, and I still get splitting when all the solvent has passed through the split (but I guess if solvent is going on-column then it won't even get to the solvent detector)

Will report back soon

[tangaloomaflyer]

Increasing the vent flow reduced my responses but did'nt affect the peak splitting. I assume my analytes are being blown out the split.

I will try changing the vent time next.

If the column has been overloaded with solvent, would it need replacing? I am using a J&W DB5-MS

[Ron]

If the peaks looked good initially, but are splitting now something in the system has changed. The most likely sources of the problem are the liner becoming contaminated, or the head of the column building up nonvolatile residue. I would change the liner for a new one, clip off around 15 cm of the column at the injector end, and see if that helps the problem.

[tangaloomaflyer]

O.K.

We just installed a new inlet, same model, same method, still splitting. Therefore I agree probably a liner issue. The peak splitting happened soon after we changed our liner cleaning procedure, but I had ruled that out because I cleaned a batch using the old method. I will revisit the liners again.

But, just to complicate things I also have a problem with the new inlet, which may or may not be causing peak splitting. After injection i get a solvent peak (on the solvent monitor - not a GC peak) which starts to form normally but then rises sharply giving a fronted peak.

The sharp rise seems to occur when the syringe is withdrawn from the septum.

What could be the cause of this?

[tangaloomaflyer]

I've just pretty much rebuilt the method. I've dropped the inlet start temp to 35degrees (from 70) and raised the oven start temp to 70 degrees (from 50). Also increased the vent time a little. From the solvent monitor it seems I'm just catching a little bit of solvent before switching to splitless. I ran a low level spike and it seemed like I had got rid of the splitting, but when I ran a high level I realised the peak has been split into two resolved peaks, which have a good shape to them. i have also got a lot more stuff coming through that presumably was being boiled off in the inlet before.

Is it poor practice to quantitate using a fully resolved split peak? I guess that's a stupid question...

[Peter Apps]

Petroleum "ether" is a mixed hydrocarbon fraction with a specified boiling point range. If it contains two major components with different boiling points (say two adjacent linear alkanes) you will have all sorts of trouble doing large volume injections and solvent effect injections. Try replacing the "ether" with the alkane that has the closest boiling point.

Peter

[tangaloomaflyer]

It is one of our long-term goals to switch to hexane. I have been trying to get around to changing but we are a very busy lab and it can be hard to get time for validations and the like. The main reason we have'nt switched yet is it takes longer to concentrate using our current prep method, but i just got a new toy for concentrating samples so hopefully I'll get the chance to run some trials this week

[tima]

I found peak splitting is common when you inject a sample that has been placed in mixture of a couple of different solvents. Non polar vs polar say isoctane and ethanol. Or there is a mixture of solvents that have a big difference in polarity and boiling point. I had this problem with a derivatization method where the agent was in the mix in significant levels to cause a peak split. So I got around it by raising the initial temperature from 65 C (good for ethyl acetate focussing) to 90 C (better for pyridine focussing) and this was a good compromise.....no more peak split! Timothy

[Peter Apps]

If hexane takes longer to evaporate than the ether it most probably has a higher boiling point. What is the BP of the ether ?

Peter