Eliminating split peaks on 6850?

[TJ]

Running GC's on diethyl diallyl malonate in methanol. I keep getting a splitting of the peak at the foot (ie a small peak, and then the regular expected peak). Varying inlet and detector temperature improves it, but it appears to return after a few runs. For instance, the first 10 or so samples are great, then about 20 are terrible, then 5 are great again, it appears to be quite random.

Any suggestions?

[Peter Apps]

The first idea I have is that you need to post the details of your analytical conditions (inlet temp, flow rates, inj volume, split ratio etc etc etc).

Peter

[Consumer Products Guy]

More details needed, such as whether split injection, exactly what type of liner is installed, injection volume, split ratio (if split), etc.

[TJ]

Sorry I didnt have the info earlier, posted question while away from the lab.

Agilent 6850 Series II

Oven set 180 oC for 5 minutes

Detector: heater (200oC), H2 Flow (45ml/min), airflow (450ml/min), makeup flow (N2 23ml/min).

Column: Constant flow, He Flow: 13.10 psi, 1.7ml/min, average velocity 36cm/s

Inlet: Split, (He gas), Heater 200oC, 13.10 psi, total flow 72.5ml/min
split ratio 40:1, split flow 67.9ml/min

Injection volume is 1ul

Liner: phenomenex tapered liner for split/splitless, (AGO-7515)

I should also stress that this is only happening for this particular substrate on this particular GC, and not others. As well, the samples have been cross checked with another GC in the facility, and the peaks were very clean with no doubling.

[Peter Apps]

What column are you using ? stationary phase and dimensions ?

Can you reproduce the eratic appearance and disappearance of the "split" peak by replicate injections from one vial ?

Do you have other peaks on the chromatogram, how do they look ?

Can you post a chromatogram - instructions in a sticky at the top of the LC page.

Peter

[TJ]

Repeat trials from the same sample show splitting as well, but the split peak to main peak ratios are not steady, and some are even negligeable. There is also THN internal standard in every sample, which is showing random splitting as well.



Column Info:

J&W Scientific
19091Z-413E
HP-1

Length: 30m, I.D. (mm) 0.32widebore, Film (um) 0.25

Temperature Limits -60 to 325 oC




TJ

[Peter Apps]

This is probably some kind of solvent effect - methanol as a sample solvent is not first choice for a non-polar methyl silicone column. Can you change the solvent ?

Also, the peaks are concentration overloaded (front tailed) so increasing the split ratio would also be useful.

Peter

[chromatographer1]

I would agree with Peter, but would also like to know the internal volume of the injection liner and for you to confirm the proper installation of the column within the liner (centered and at the proper depth).

I would try injecting the sample volume slowly instead of quickly (burst mode), injecting the 1µL over 1-3 seconds for example.

Lastly, I would suggest to try injecting smaller amounts to see if that affects the peak splitting.

best wishes,

Rod

[AICMM]

TJ,

A question and a comment. Is this the only 6850 running this sample or are other 6850's getting better results? Comment, I agree with Chrom.1 and Peter so try running 1/2 uL instead of 1 uL. Expansion volume will be less, should fit in the liner you are using and may resolve your problem. Should be possible with a 7673 if you use a 5 uL syringe instead of a 10 uL syringe.

Good luck.

[TJ]

Thanks for the help all.

The GC I cross checked with is a 6890.

I've switched to ether as solvent, and increased split ratio to 60:1. It seems to have improved greatly, even with same injection volume. I've only tested it with high concentration single vial setup. Tonight, I am trying it with my actual reaction, and 48 vials. Hopefully it will be fine.