Page 1 of 1

Accuracy (Recovery) Issue – sample diluent with citrate

Posted: Thu Mar 19, 2015 4:53 pm
by Rob Burgess
Hi there,

We have experienced issues recently with an accuracy spiking method validation exercise, spiking drug into a liquid nebulizer solution that contains 10 mM of a citrate buffer. The concentration of drug is prepared accurately at about 5 mg/mL. The drug itself is a zwitterionic .HCL salt. The sample preparation for LC injection involves a 50 fold dilution with 25/75 AcN/water to 0.1 mg/mL concentration. So the final sample diluent contains 2% v/v of 10 mM citrate buffer.

The assay standards are prepared slightly differently, by dissolving directly into a 25/75 AcN/water v/v diluent. So there is no citrate in these standard solutions.

Assaying the accuracy (spiked) validation samples against the standards we consistently got between 93-94% recoveries (outside of our validation SOP criteria). We then investigated adding different levels of citrate to the samples and obtained the following results.


Citrate Addition % v/v Drug Recovery %
0 99.9
1 97.1
2 95.6
4 94.7
6 94
8 94.2
10 94.1

So there appears to be some kind of leveling off effect with the citrate concentration in the sample solution.What could be the cause of this missing 6-7% when we run mismatched samples and standards? Is it some kind of ion-pairing/chelating effect with citrate that is causing a decrease in our assay recovery by LC? It is the first time I have come across something like this of this big a significance.

Consequently when we added 2% v/v citrate to our standard diluent preparation we achieve ca. 100% recoveries for our samples. I also compared UV stand-alone spectra of standards (no citrate) with standards with 2 and 10% citrate (diluted from the same stock standard so all at the same concetration) and there was perfectly overlay - i.e. no shifts in response or wavelength.

Re: Accuracy (Recovery) Issue – sample diluent with citrate

Posted: Thu Mar 19, 2015 6:27 pm
by tom jupille
Maybe something along the lines of the "strong diluent effect" where the presence of the citrate alters the equilibria a bit until the injected sample gets diluted by mobile phase (the same kind of thing you see if you inject a sample dissolved in 100% organic)? In that situation, the peak shapes with/without citrate might be subtly different, enough to integrate slightly differently (yes, I'm hand-waving here!).

Regardless, matching the sample and standard diluents as you have done seems to me to be the most appropriate fix.

Re: Accuracy (Recovery) Issue – sample diluent with citrate

Posted: Thu Mar 19, 2015 7:11 pm
by Consumer Products Guy
Regardless, matching the sample and standard diluents as you have done seems to me to be the most appropriate fix.

That's also what we would do.

Re: Accuracy (Recovery) Issue – sample diluent with citrate

Posted: Tue Mar 24, 2015 12:59 pm
by lmh
playing devil's advocate, if you genuinely have some strange mechanism by which citrate is actually removing analyte from the system (e.g. some weird ion-pair effect by which it causes a small amount of active ingredient to bind strongly enough to plastic tubes that you lose it...), then by matching your standards to the samples, you achieve a situation where all of your results are uniformly wrong. Neither the samples, nor the standards, contain what you expect, but since they agree about what they contain, everyone is happy...