Carry Over Peak Help Suggestions needed

[jwolff]

I have a carryover peak that I am having a hard time identifying the source of. This is an established method, from which I can not deviate from.

Here is the background:
-Isocratic, 200ul inj. non buffered MP
-Peak does not appear in blank
-Peak does not appear in 1st inj. of std, but does in 2nd inj (related to actives/analyte of interest).
-When run time is doubled, peak still appears in 2nd inj at same retention time.
-Peak elutes just prior to and interferes with the peak of interest
-Consitently appears on different instruments.
-Needlewash is functioning properly, needlewash solution has been routinely used in this analysis, increase organic conc made no difference, sample is soluble in organic, but is in an aqueous solution for this analysis.

Performed following inj. sequence:

-Blank, 2 Std, 2 Blank: Peak is present in 2 inj std and 1st of the following blank, 2nd blank inj is clean.

Blank, 2 Std, 2 zero volume inj, 2 Blank: Peak is present in 2 inj std and 1st of the zero vol, 2nd zero vol. inj is clean, last 2 blanks clean.

Blank, 1-Std (half inj vol), 2 Blank: Peak is present in 1st of the following blank at half the size, 2nd blank inj is clean.

Anyoneone have any ideas? It seems like it is geting hung up somewhere in the system and not releasing until thhe next injection.

Thanks
Julie

[HW Mueller]

This doesn´t make too much sense. What happens if you inject, lets say, 10 std consecutively? Do any peak areas change?

[Bruce Hamilton]

If the method doesn't work, then you should deviate from it. I assume the method used to work OK. If so, examine what has changed.

Is the method isocratic or gradient?. If isocratic, I'd try a couple of gradient runs to ensure the late material is not deposited in the flow path, waiting for the next 200ul of sample solvent ( unlikely, but.. )

200 ul is relative large injection volume, and I would dissolve the standard in the initial mobile phase at a higher concentration, and inject varying volumes to ascertain if there is an injector or solubility issue.

Please keep having fun,

Bruce Hamilton

[jwolff]

This doesn´t make too much sense. What happens if you inject, lets say, 10 std consecutively? Do any peak areas change?

If I inject 10 stds, it will be the same size in each injection. The only time the carryover peak changes in size is when the concentration of the solution changes....so lower conc std..next inj, smaller carry over. Higher conc std, higher area carry over in the next.

[Uwe Neue]

I very vaguely remember something similar from an older post. The source was sample volume that was parked somewhere in the injector loop and was injected when the new sample was loaded. Search for some older posts on similar problems.

[Kostas Petritis]

I was thinking something on the same lines like Uwe but you mentioned that you get the same problem in different instruments. What do you mean when you say different instruments (i.e. two different autosamplers of the same brand, different autosamplers etc)?

In order to narrow it down to the autosampler and to some kind of carry over you can try to inject your blank, 2 sample, 2 blank by using a manual injection valve.

[HW Mueller]

What bothers me is the rt difference between carryover and std. There must be some strange plumbing involved, or some strange chemistry?? What happens if you collect, concentrate, and reinject the carryover peak? Do you have other methods to characterize this carryover?
What happens if the std is in a different solvent (for instance mp as Bruce suggested)?

[unmgvar]

from what i read we are missing out on key information.
what is the equipment used?
what is the type of injection used- split loop or pulled loop?
what is the solvent used to dissolve the sample?
what is the mobile phase?
what is used to clean the needle and how much?
what is your sample nature.

for example last week i had to troubleshhot a case of carry over for a caffeine application.
two things stood out and help solve the case.
the sample nature was soft drinks (high conc. of sugar in them)
the needle wash was 50% AC 50% H2O.
simply moving to pure water and increasing the amount used for washing the needle solved it from 10% carry over to none

[HW Mueller]

unmgvar, did your carryover come at a different retention time than peaks originally seen in your sample?

[unmgvar]

no the RT was the same,
but this was just an example, no similarities to this case.

[jwolff]

We may have solved the problem. We are running on waters Alliance systems. Both instruments we tried were getting close to their next scheduled PM. We were suspecting that the "squeegee" that clean the needle between injection may be slightly worn and not cleaning as effectively, so we switched to a third LC that was new.

On this test run we did not see any carryover. We are going to start our system suitability, so I am keeping my figers crossed for now.

Thank you for the suggestions, I hope that this is the solution, however if not I will update with answers to the questions posted and try some of the suggestions.

Julie