We are having a problem in recovery in one of our accuracy experiments. It is a dissolution profile. We are checking recovery at 5 conc. 10, 20, 40, 80, 100, 120 % of the working conc. Recovery is 100% at 100% conc. but increases to 110% at 10 % conc. solution and decrease a bit at 120% conc solution. What could be the reasons. Can a more experienced person suggest something?

Dear Anonymous,
Without knowing the concentrations your are working at, have you checked the linearity of the method yet? Maybe you are operating just outside the linear range for this compound. Also, have you run a placebo to make sure there is nothing uder the analyte peak? Just a couple of quick thoughts.

Regards,
Mark

Maybe you have solubility problem.
How did you prepare the samples?

Okay. the working conc. is 100 microgram /ml and the solutions were prepared at above mentioned concentrations. Linearity was checked before at the same test concentrations , correlation coefficient was 0.9999. If I calculate the recovery at the above conc from solutions prepared from linearity, I get the same recovery as I got in accuracy. There is nothing under the analyte peak. It is checked already. Doesnot seem to be the solubility problem. Samples are prepared in water. What other possibilities can we explore?

Carry over... See if the situation changes if you randomize the injection sequence of the standards.

How are you taking the samples?

Are the samples put in the disso tester and withdrawn with an automatic sampler?
I have seen carryover because the use of stainless steel components in the automatic sampler.

Regards

ther were no carry overs as we had done the % disso tests,it was never observed before. But the problem was solved by lower injection volume , % recovery at 110% level is aroud 102% now.