nagetive peak in USP method

[zhengmin]

Hello everyone:

We are determining related substance of the capecitabine API by HPLC which is USP method, now we found a nagetive peak just before at a related substance peak(RT=2.7min), it disturb the quatity for the impurity.the mobile phase is water(containing 0.1 glacial acetic acid):methanol:ACN=60:35:5, column is C18 4.6*250mm, 5u; diluted solution is water:methanol:ACN=60:35:5,we adjusted the mobile phase, change the source of water, methanol, ACN, acetic acid(they are all HPLC grade), but all cannot remove the nagetive peak. Finally, we change the diluted solution to water:(methanol:ACN=35:5)=9:1, the nagetive peak is almost disappeared. Because the method is USP method, so we don't know weather it is permitted by FDA. What should be done if we change the diluted solution? We should validat the whole method? And how can we eliminate the nagetive peak?

Wish your help.Thanks of all.

Zhengmin


2008.2.1

[AdrianF]

if you see no difference in the peak areas after using you alternative diluent I don't see their is a problem as longa s you document your change .

As a first choice I would try using the mobile phase as diluent ie use the acetic acid. I cannot see there could be any criticism for doing that unless there is a question opf stability in a weak acid solution.

In either case you will need to run comparison chromatograms and show that the only thing to change is the negative peak.

[lupetto]

hello Zhengmin,
sometimes you can easily get rid of negative peaks by changing the solvent you dissolved your sample in, especially if your target compound elutes that early

[Bruce Hamilton]

If you are reporting results as complying with the USP method, and reporting the product as complying with the USP, you would have to initiate a series of actions before you could change the sample solvent.

In general, quality people are more likely to regard such a change in a compendial procedure as a potentially serious event, and trigger an Out-of-specification process. That requires a thorough investigation and quality signoff.

The obvious issue is :- what is causing the negative peak?. By changing the solvent you may no longer be extracting something that should not be present, or your analytical system is misbehaving.

Any investigation should initially focus on the cause of the negative peak, and only then can you decide whether elimination of the problem by changing solvents is acceptable. You would certainly have to validate any procedural changes.

Bruce Hamilton

[zhengmin]

Thanks for all reply. Now we know how to do it.

[danko]

Assuming the flow rate is 1 mL/min (about the optimal for this column) the injection peak/disturbance should elute exactly at the time of this negative peak in question. Now, if the detection wavelength is relatively short, the acetic acid in the mobile phase would absorb a god deal of energy/light. So when an injection of a sample dissolved in a solvent less absorbing than the mobile phase (no acetic acid in the solvent) is performed, a negative peak at dead volume is inevitable.
The fact that an impurity/a part of the analyte elutes at that time indicates that the method developer either was an amateur or just a careless guy looking forward to his/her retirement.

Best Regards