Hi Christine.

I think you had made injection very high concentration in your mass spectrometer.

Try dilute more your solution if possible. Another alternative decrease the injection volume.

I don't know what is your mass spectrometer model, but you have very high signal. Probably you have saturated the mass spectrometer signal.

Regards.

Hi Christine.

I think you had made injection very high concentration in your mass spectrometer.

Try dilute more your solution if possible. Another alternative decrease the injection volume.

I don't know what is your mass spectrometer model, but you have very high signal. Probably you have saturated the mass spectrometer signal.

Regards.

Hi all, I agree with Marcos .. MS detector signal will not be linear for these high concentrations .. Christine, would you please try 1:100.000 and 1:200.000 dilutions and see if AUC will be proportional to concentrations?

i agree with both answers, but consider cleaning your MS first, sometimes high concentrations contaminate the system, and all later results will not be efficient .

Good advice above.  As a general rule in our lab, we start with very low concentrations and work up rather than starting with high concentrations and working down.  If you have a high-resolution instrument, and are saturating the detector,  you'll find (do the experiment!) that the mass accuracy gets better as the concentration gets lower.

Thank you very much for all the input! I have to say that originally, I started measuring concentrations in the range of 10-80μg/L. The results were very unusual. That is why I thought I try to measure higher concentrations of my stock, yet with similar results.

I took your thoughts into consideration and cleaned my MS again, measuring only 10-40-80μg/L today. I used a phenyl column with a flow of 0.200ml/min. My organic solvent was MeOH. This set up gave me the highest AUC and stable RT so far. Total runtime is 15min with ascending straight to 75% B at 7min, then going back down to 0%. I am working on Dionex Ultimate 3000 + Bruker microTOF Q-II.

I measured 10-40-80μg/L from stock dissolved in and diluted with H2O or MeOH. Standards prepared in H2O were not visible. Standards prepared in MeOH gave results similar  to as before: 10μg/L - AUC 154.000 – int. 10^4, 40μg/L - AUC 33.000 – int. 2000, 80μg/L - AUC 89.000 – int. 8000

Interestingly, I was not able to see a strong signal for m/Z 363.2 for cortisol within my peaks today. This makes me doubt my measurements even further.

I would be pleased to hear your opinions about these results!

 Do you monitor the pH of your solvent system?  I notice that you don't mention acetic or formic acid in your solvents.  Cortisol has a carboxylic acid function, and you're doing the analyses in positive mode in water or water/methanol, so there's a possibility that you're not getting much protonation (or are getting varying protonation, which could account for your lack of reproducibility).  You might try water/acetonitrile systems with around 0.1% formic or acetic acid in both solvents (I like acetic acid).  Negative mode might be worth a try.

Also, I've been a bit confused by your use of 'AUC', since that has a specific meaning in pharmacokinetics, and I'm still not sure if you're trying to measure that or by AUC you simply mean the integration of the m/z 363.2 peak in the extracted-ion or total-ion chromatograms.

Dear Dr. Wishnok,

thank you for your reply. All my solvents contain 0,1% formic acid. I tried using water/acetonitrile up until now but switched to water/methanol since results (though still unproportional) turned out better. I am operating in positive mode. When mentioning AUC I mean the integration of m/z 363.2 in the extracted-ion chromatogram.

This is puzzling indeed.  Incidentally, what are you using for an internal standard?   Also, how reproducible are several injections from a given vial?    Are your injection solutions the same ratio of water/methanol as the initial ratio of your gradients?

So far I have not used an internal standard. Unfortunately, deuterium labelled cortisol is out of my purchase range for this research. I know about the problems of verifying measurements without an internal standard. My alternative now will be using prednisolone for tests in the upcoming weeks.

I switched to negative mode now and tried the suggested repeated measurements of the same vial. With my previous set ups I only repeated a couple of measurements from dilutions mixed on different days. This gave me no further insight though into weather my results are truly reproducible or not. Since neither stock now dilutions made in water or methanol turned out to be better or worse, I will now work with stock and standards diluted in 50:50 water/methanol. I will update his thread when results are ready.

Update: Problems have been solved. It seems like a small leakage in the HPLC injection pump was part of my deceptive results. With MS set to negative mode, an adapted gradient up to 50% B as well as sample preparation involving SPE results are well reproducible. 

Whew!

:)

(As is often the case, the simple things can be the most important.)