Chromatography Forum

Hello, we have seen a progressive loss in linearity for a UPLC-APCI-MS assay we routinely perform. Gradient is acetonitrile/water, MS is LTQ. Standard curves are calculated over 3 orders of magnitude using analyte/IS response ratio ploted against analyte concentration. IS is not isotopic analog but related compound with different RT.

Std curve has gradually gone from linear with r2=0.999 to a curve with high concentrations tailing off over last 6 months. Now a quadratic or log-log fit has to be used. Non-linearity also suddenly became a lot worse after a recent manufacturer engineer visit and MS service.

We are sure this is not due to std preparation, UPLC hardware or columns (we routinely change these). Peak shapes and RT are exactly as they have been when the assay was linear. I suspect the MS. Unfortunately, our analyte is non-UV absorbant, so we can't check this with a UV or PDA detector inline between the UPLC and MS.

I am guessing this is a MS problem, perhaps from dirty ion optics? APCI, ion transfer tube, and first couple of lenses are cleaned/exhanged regularly but hasn't fixed the problem. Has anyone else seen this problem? Or suggest procedures to diagnose the cuase of the non-linearity?

thanks
Tony

Did your peak areas go up after the service call?
Is it possible that your instrument is more sensitive post service, and that you are now saturating your detector?

No - in fact, the peak areas went down 30-fold after the service! After extensive re-tuning and cleaning the ion optics, we managed to get the sensitivity up to 10-fold worse than it was before the service. Unfortunately, the instrument passed all the standard tests and checks done as part of the PM service, so engineers don't have anything obvious to go on.

Have since also perfomed loop injections of several dilution series to rule out the UPLC - it is definitely the MS. The calibration curve even "curves" now (i.e. is non-linear) at the lowest few concentrations, so I am at a loss....

thanks
Tony

Hello,

I'm having the same problem with my LC-ESI-MS-MS. I lost the linearty sudenly. After clean the source, lens, Q0, new mobile phases, testing injector, etc. I can't fix the problem.

The problem is more accented in molecules with higher mol. weight.
And some times the IS response drop in a 50% between samples.

I hope someone can help us.

Thanks

Sounds like a detector blinding issue - Especially with a 3 order curve. Check your electron multiplier values against what you had pre-PM, and contact the manufacturer if this hasn't changed.

If its not that, (and this is a rare one) - Are you aquiring the entire chromatographic window. I have seen instances of peak variations based on concentrations, and you may be losing analyte in on solvent front or "bull horn peaks". The ratio of loss from quantifying peak increases as concentration increases and results in bent curves.

Sorry but I don`t understand very much the propose of Rappoldbr...

I'm having a simmilar problem with an analysis in plasma samples by HPLC-MS/MS...I'm analysing two products in the same samples (with Internal Standard), but I'm only loosing linearity in one of them...
The validation process didn't had this problems, but since we began to analyse samples, the loss of linearity is increasing from day to day...I have clean the MS but I still have the problem...I have check the standard solutions, I have changed the chromatographic column, I have changed the position of the ESI probe (to "decrease the response in case it is a problem of too high response in the higher part of the column) but nothing has changed....

Any more ideas???? Thank you!!!!

It sounds like dirty ion optics, but if you have already cleaned that... Do a good mass and resolution calibration. Check your multiplier. Maybe your multiplier needs replacement?

Hi,
I'm interested to hear if you found a solution to this problem?

I'm maybe experiencing a similar problem with an analysis running in ESI neg mode. My highest calibration point started (some weeks ago) to deviate from its expected value (downwards) and after an annual service the situation got worse: also the next highest point deviates now, and some of the lower points deviate upwards.

Any ideas are more than welcomed!