Ghost peak in SEC

[bostonian737]

I am seeing a ghost peak in a validated assay that did not show up in validation. It coelutes with the aggregate of the samples we are testing. The samples are injected in triplicate (from the same vial) yet the ghost peak will show up in 1 or 2 or 3 of the injections, its always different. The rentention time does not shift. It will show up in different types of samples including the mobile phase blanks. We have tried new mobile phases, reagents used to make the mobile phases, new resevior bottles, different LC system, replaced the lamp, pump head, needle & seat, injector valve, new guard & analytical columns, and operator.

Thanks,
Any suggestions are welcome, just frustrated

[unmgvar]

have you changed the vials and septas?
we had a cases where the purchase dept. decided to do a major save of nickles and dimes and got cheaper stuff
after a very annoying investigation we found that our problem in our given case was the quality of the septa.
it was not a SEC application, but it looks like you tried everything but that

[danko]

Let me guess: You don't see the peek initially when you install a new column. Right?
I thik you have a build up, due to unspecific retention of the aggregate molecule.
Then once in awhile when there is too much of it, the equilibrium shifts an some of the stuff elutes.

Best Regards

[bostonian737]

Thanks for the responses,
unmgvar- we use glass vials & ran test injections without any caps to rule out the septa.

danko- we actually see it on new columns & guard columns. we've tried different lots from the manufacturer.

[danko]

Then I think you’re dealing with some sort of pollution. Might be the injector.
What happens if you just inject several (6 – 7) injections of the blank (mobile phase)? Will the peak appear and disappear randomly during the sequence, or will it just “fade away”?

Best Regards

[bostonian737]

I'm also leaning towards a pollutant possibly in the sample that's hanging up somewhere in the system.

We ran 10 mobile phase blanks, the first 4 had decreasing peaks but the last 6 injections had a constant ghost peak.

I'm going to sit down with all the data again today. Looking at the specific runs after LC parts were replaced & using a different system.

Thanks again

[jgarey]

PTFE-Silicon-PTFE are only good for one injection if THF or Acetonitrile are involved. The Si can create an impurity peak. I've had issues with ghost peaks being created because the sample diluent and mobile phase didn't match. SEC/GPC functions better if you can match the mobile phase and diluent as closely as possible. I saw this using a Corona CAD though, so it may be much less of a problem with UV/Vis. I pre-mixed my mobile phase (instead of online mixing) and used the same mix for sample prep and mobile phase. That solved my issue. I would have expected the ghost peak to come out at the Total Exclusion time, but it actually came out later and created interference for me.

[danko]

Hi jgarey,

But the bostonian737 has injected 10 (at least I believe) blanks i.e. mobile phase.

Best Regards