Where to start troubleshooting

[Charles A. Burger]

I have a combipal injector on a Shimadzu 2014
Running a packed column on a split injection system. (I know already that I shouldn't be doing this but shimadzu has an adapter that allows for it and I'm stuck with what I have.) - split and purge vents are off.

Very simple run. Compounds are Cetyl and Stearyl Alcohol in 200 proof ethanol. 5 injections of a standard only used for resolution requirements, 2 injections of 1 sample prep. Calc is area percent.

Couldn't be happier with my standard precision, both peaks area RSD's under 1.2%.

Problem rears its head with the sample injections. The first one has noticable drop of in Area and Height. Tailing increases, retention time decreases from established 5 rep. Next sample injection, everything back to normal and right in line with the 5 rep. Final injection was of the standard, again, right in line with the 5 rep. So we seem to have one problem injection.

Since I'm new to the combipal scene I wandered what would happen if I injected different amounts of a standard. 0.5, 1.0, and 1.25µL used. Both compounds generated 0.9996 correlation coef. Also, RT's right on top of each other and the tailing was spot on. My thought was what if I had an air bubble that might have reduced my injectable amount. Doesn't seem to be the case.

What direction should I go now?

[JTM]

dilute the sample a little and run it again. (10x)

[Don_Hilton]

First question - are your samples in 200 prro ethanol also? If not, it may be a solvent issue.

[chromatographer1]

That first injection may have had a leak in the septum due to unseen causes, but that corrected itself with later injections.

When in doubt repeat the series.

Unless the first sample injection had a base in its solution, that is.

Rod

[Charles A. Burger]

First question - are your samples in 200 prro ethanol also? If not, it may be a solvent issue.

Yes they are.

[Peter Apps]

A problem is not a problem if it only happens once; so follow Rod's advice and re-run the sequence.

Peter

[Charles A. Burger]

FYI the shimadzu system has a thick plug for a septa. I would love to know why, other than trying to get more bang for the buck. But having said that, my guess is that a spetum leak would be difficult with this style of septa. I'm not discounting the possibility, but I thought that should be known.

What I ended up doing last night was prepare 3 vials with the same standard prep. Each vial was injected 12 times. The data that I observed today was interesting. Of the 36 injections only 2 were substantialy lower in area. Inconjuction, both of those injections had poorer symmetry numbers. Also, both poor injections were the first ones out of their respective vials

Of the 36 injections 6 of them had noticably different elution times (faster), however only one of those matched the 2 from above. There was no discernable pattern observed.

[Peter Apps]

These intermittent problems are the very worse kind ! Try a sequence of twelve vials, one injection out of each. Watch the autosampler while it works.

What are you washing the syringe with between injections ? what kind of caps do you have on the vials, and is the instrument lab at the same temperature as the lab where the vials are filled - if it is cooler you mihgt be getting a partial vacuum in the vial that equalises once the septum is punctured.

Peter

[chromatographer1]

First of a vial injection problems is usually associated with the needle of the syringe. The tip may not curve over the opening (ignore if you are using a side port needle) properly. A bit of septum is being cored and the uptake and expelling of the sample is being affected.

Other issues may be the warming or cooling of the sample vial, increasing or decreasing the vapor pressure inside the vial which can affect the uptake and the amount of sample injected.

Ain't science neat?

Rod

[ingochrist]

I like to add my 5 cents to the topic. When doing a validation on a CTC PAL, we require using a brand new 10uL 26 AS gauge (it states in the IQ/OQ document) Hamilton syringe -- best would be part number 203205. Make sure not to use any bevel tip syringe, I am sure any other brand is fine, too.

If you want to make sure the PAL is not the source of the problem, fill 10 vials with a known amount of H2O and remove a certain amount and weigh it again. I would use a 100uL syringe for that test (it fits in the 10uL syringe holder), since it would otherwise be hard to determine if the error is from the balance, fingerprints or who knows what.

Feel free to contact us, if you believe the PAL is the source of your reproducibility problem.

Best Regards,
Ingo at autosamplerguys dot com

[AICMM]

Mr. Burger,

Of the two injections that were bad, it was probably the second and third vial and not the first? This would suggest Rod's idea of pressure build up in the vial from ethanol evaporating.

Other than cooling the tray, I am not yet sure how I would attack this problem....

Best regards,

AICMM