same compound 2 peaks doing liquids split

[MSCHemist]

I've been seeing an interesting phenomena. When I do liquids injections I see the same compound in two places. A small peak followed by the main peak about 1-2 min later. It only happens with some high abundant compounds when I inject split (50 to 500 to 1) using Agilent glass wool liner 5181-3316. I've seen it on different compounds on different samples. I've never seen it doing SPME where I use a spit liner with no glass wool and a 2:1 split.

What causes it?

[Peter Apps]

With split ratios that high it is unlikely to be a solvent effect, and 1 - 2 minutes is a long time for anything else like temporary pulses of gas flow into split lines. How do you know that it is the same compound ? - can you rule out the small peak being an impurity of the large one ?

Peter

[MSCHemist]

This is with GC/MS so I get identical spectra and high library hits and I've ruled out isomers. In some cases it is with pure compounds.

[Bigbear]

When was the last injector maintenance?
Sounds like you are washng something off your injection port. Do you see it with a solvent only injection?
Are you over loading your liner? How long do you have your split solenoid off durring injection?

[MSCHemist]

I wash my injector out pretty frequently and replaced the split trap a few months ago. I also switch liners frequently between splitless for 3-MCPD, SPME (Restek 20913), and liquids split with the Agilent glass wool liner.

I also checked the sept purge and it is blowing out ~2ml/min according to my flowmeter.

Most of my samples are concentrated materials or flavors in an oily solvent such as propylene glycol or glycerol tricaprylate (oleoresins). Most of those I just inject 0.2ul with a 500:1 split.

[MSCHemist]

One thing I do when doing a split liquids injection is to set the autosampler to leave the syringe in the injector for 5 seconds. I did some work with an MPS-2 where I saw a significant improvement injection precission by setting any delay of 5 sec or more. I am using an Agilent 7683 now but I assume the same principle applies.

[Peter Apps]

This is with GC/MS so I get identical spectra and high library hits and I've ruled out isomers. In some cases it is with pure compounds.

It could still be an impurity if the main compound is e.g. an alkane or a long-chain ester that produces a string of fragment peaks with no obvious molecular ion. What is the area ratio of the main to the "split" peak more or less, and is it consistent for a given material from injection to injection ?

You solvents have much higher boiling points than the run of the mill GC solvents like hexane or dichloromethane. What happens if you dilute say 10:1 in something low boiling and then split 50:1 instead of injecting neat and splitting 500:1 ?. The peak sizes should stay the same, and if the "split" peak is an impurity you will still see it. If it goes away you have some effect of the high boiling solvents that your samples come in - and a solution to the problem.

Peter

[Yama001]

Does the gas saver turn on at 1-2 minutes after injection? It may be worth setting this to a later time and see if the split peak follows it.