Impurities assay OOS

3 posts / 0 new
Last post
mrm707
mrm707's picture
Impurities assay OOS

Hi,
 
I have an impurities assay (HPLC) test method on which I obtained an OOS (high) result  for an unknown impurity for Morphine Sulfate API.  This happened over 2 years ago.  Two injections are run on the same sample preparation. 

We ran a retest protocol on 5 new sample preparations.  This unknown impurity is still present (in all of these preps) but at a much lower conc (approx 1/3 that of the OOS conc.).  A different HPLC system, columns, and solutions were used.  The method is gradient with a UV detector at 284 nm, inj. volume 30 ul, and flow rate 1.0ml/min.  The run time is 72 min.  Column type is 4um C18 dim = 4.6 x 150 mm.  A USP reference standard of Morphine sulfate is used to calculate impurities conc (based on average area count of 6 injects). 

I was originally thinking that contamination of the orig sampe prep was the root cause.  But since this same impurity appears in the 5 resample preps, over 2 years later, I am thinking there is some other cause.

Do you have any ideas? 

Thanks!
mrm707

Dr. Analytical
Dr. Analytical's picture
sorry for the late response.

sorry for the late response.

It is possible that the same contamination exists in your laboratory even after two years, but I agree that this is not a common situation.  However, glassware contamination from cleaning solutions is a likely source of external contamination.  Other glassware, lab supplies, and even filters can contribute to contamination.

Have you completely checked a "reagent blank?"  Process a solvent sample through exactly the same procedure and look for the impurity.

Also program a "blank gradient" (no injection) to verify that the system is not contaminated.

If these experiments do not indicate a problem, then you should submit the sample for further testing to identify the impurity.  LCMS would be the recommended approach.

mrm707
mrm707's picture
Thanks for these strategies! 

Thanks for these strategies!  This is a great help.