Chromatography Forum

Let's backtrack a bit:

1. It is definitely better to have the same %recovery for both the IS and the analyte (i.e. "relative recovery" of 100%).

2. However, if you carry your calibrator solutions through the same workup procedure as your samples, and if your relative recovery is in fact constant (e.g., always 90%), then you should get the same relative recovery in both calibrators and samples.

3. In that case, the relative recovery effects will cancel out, and you will get the correct values for your analyte in your samples.

The entire point of using the IS is so that you need not worry about the vagaries of extractions. If your ratio of IS area to Analyte (spike) area is constant over replicate extractions, you should be fine.

Thanks Tom
Does that mean that the working standard should be treated exactly as the real samples even by adding it to the same matrix?

Thank you.

As a minimum, I would do a set of replicates on a spiked matrix. If you can demonstrate that you get the same relative recovery from the spiked matrix and from plain calibrators carried through the same procedure, then there is no further need to use spiked matrix as calibrator.

If the relative recovery from the spiked matrix differs from the 90% that you obtained with plain calibrators, then you will have to use spiked matrix calibrators. At that point, the alternative of looking for a different IS becomes much more attractive!

Thanks DR and Tom

I repeated the experiment at lower concentrations of both the IS and the analyte and I was surprised by a %relative recovery of 96.6 +/- 0.31

Do you think it is a matter of saturating the upper layer at the high concentrations?

Actually, it doesn't matter. The fact that your relative recovery is inconsistent means that the use of that IS will probably make your results worse instead of better . If it's any consolation, better to find this out now rather than after you have run a number of real samples!