Poor accuracy (clomiphene citrate tablet)

[chicay]

I am validating analytical method for Clomiphene citrate tablet dissolution using HPLC (there are 2 peaks). The problem is, although in the specificity test there is no peak at the clomiphene citrate retention time, the method has a poor accuracy. I prepared my accuracy sample by spiking a quantity of the standard solution to the filtered placebo solution. I was using 4 level of concentration. And the result:
Level Recovery
30% 107%
60% 104%
100% 100%
120% 100%.
Can anyone help me? Thanx before.

[Kostas Petritis]

Can you please give more information on column, mobile phase, gradient and type of detector used?

[chicay]

Column : CN 250 x 4.6 mm
Mobile phase : Methanol : Phosphate Buffer contains 0.6% TEA pH 2.5 (55:45)
Detector: UV- Vis

[JGK]

Are you using a unit response factor (single point) to calculate concentration or a multi level calibration?

If you're using the single point method, your accuracy values are not unsurprising. they will tend to deviate higher or lower the further away from the calibration level you get.

[chicay]

I used a multi level calibration (3 level concentration).

[mbicking]

What procedures did you use for spiking? What concentrations, what volumes, and how did you measure these volumes (pipet, pipetter,etc.)? It looks like the only problem is at 30%.

[Dan]

I will second Merlin's request for your spiking procedure. The are several problems that can be encountered in the spiking experiment for a dissolution method.

The order of addition of analyte and placebo powders and dissolution media can greatly influence the results. In some instances, the spiking should be done with a solution of your analyte rather than directly spiking the material in powder form. See more about it at the forum for the Dissolution Discussion Group: http://www.dissolution.com/vbulletin/sh ... php?t=2296

So, please give a little more info on your spiking procedure and we can better troubleshoot.

Regards,
Dan

[chicay]

For sample preparation, we transfer a certain volume of our analyte solution to volumetric flask, then we add filtered placebo solution to the mark.