I am trying to develop a LC-MS/MS method for a compound. The calibration curve flattens at the higher concentration points when measuring in the range of 2.5-5000 ng/mL. It is lineair when measuring in the range of 1-500 ng/mL.
I have tried to measure with different transitions (naturally occuring isotopes +1+0, +1+1, +2+0, +2+1, +2+2 for the molecular ion and the product ion, respectively.) but the calibration curve still flattens out at the higher concentration points. Since the absolute area (cps) of the peaks do decrease as expected, I think there is saturation in the ion source.
If there is indeed an ion-source saturation, how can I solve this problem? I prefer not to dilute the samples because in the MRM method, I am analysing multiple analytes and I cannot lose more sensitivity for the other analytes. Also, I am curious if there are fundamental approaches to prevent this problem just for my knowledge.
With much less abundant isotopologues you still got saturation and so you did have ionization saturation. Certain degree of ionization saturation can be compensated to certain extent by using stable isotope labeled internal standard.
Ionization saturation means that increased analyte concentration does not result in more (possibly even less sometimes) target ions generated near the peak apex, making it impossible to perform quantification based on peak area or height. For severe ionization saturation, there is no GOOD option other than diluting the sample or reducing the injection volume, still there is a possibly acceptable option depending on your requirement, Width Based Quantitation: Article Width Based Quantitation of Chromatographic Peaks. Principle...
Hi.
Dilute the samples and/or reduce dramatically the injection volume. And recheck the calibration ranges. Linearity between 1 - 500 ppb?! How did you verify this?
Cheers!
Thank you all for the comments!
So, there is certainly an ionization saturation because the calibration curves of all of the less abundant isotopologues flattens at the higher concentration points.
Since I am already using a stable isotope as internal standard (3H2 or 3x deuterium). Does this mean that I can continue with the current method and just use a quadratic fit for the calibration curve? That gives a r2=0.9998 which is really good. H. Liu . I dont want to dilute the samples because of the other analytes that need the highest possible sensitivity.
John R Carney : Changing the weighing factor does not improve the fit in my case.
Mário Silva : The same happens for the 1 µL injections (UHPLC column with an internal diameter of 2.1 mm). Also, our lab works with protocols and validation methods according to the FDA rules, and the calibration curve is lineair when tested between 1 ng/mL - 500 ng/mL.
Hello,
You need to create more ions. This could be possible by changing ion source parameters (temperature, gas flow, ESI voltage) or small adjustment of the mobile phase composition. it is quite empirical work.
Asal Asali I understand this. We all do the same. However, in my experience the best conditions for sensitivity are not always the best for a quantitative work. I don't know how you did optimization (infusion or else) but maybe you were already saturating the spray? Sometimes it's worth to come back on initial optimization to do a check.
Dear all,
I have solved this matter by (dramatically) increasing my internal standard concentration. There is ofcourse still saturation at the source. I can see that in the gradually decreasing internal standard peak area's, because of supression of the analyte. However, the calibration curve is perfectly linear now.
- Badges
- Science topic
- Similar topics
- Mathematics
- Algebra
- Matrix