I was hoping that someone could help me with some problems I am having with a validated HPLC impurity method. We have a peak eluting in our solvent front which was traditionaly attributed to a placebo peak. However this peak seems to be increasing in size in recently generated chromatography.Also this peak is not always present and in somes samples seems to decrease over time. Any ideas?

Just to add to this. Is the peak size varying because of where it is eluting. Also could be due to Ethanol can Ethanol be detected using RP HPLC?

Avoid analysing the void peak. It is really not reliable anyway. You need to do a second and preferable ortogonal separation compared to your initial separation technique.

Are these the same samples that have been run in the past (e.g. stability samples at a later time-point or control samples)? Can the change in that peak be correlated with change in active peak? You could be experiencing degradation with the degradate eluting in the solvent front, or possibly some form of conjugation of molecules (we see lactose adducts with some of our compounds).

It could still be a "placebo" peak - you could just be getting variance in the amount dissolving in your samples solutions.

Have you changed supplier of any of your actives/excipients or the solvents/reagents you use to make up your samples? Change in any of these could affect your chromatography, if the change was not fully evaluated.

As indicated by Einar, changes in the solvent front aren't reliable - you need to get your HPLC development people onto this one to see if you can separate it out. If that's not possible, you could possibly try running the sample through a Diode Array detector and see if you can determine anything from the spectrum of the solvent front.

Thanks for your suggestions. The size of this peak has not changed in the active. What would PDA tell me about the peak? I have limited knowledge of PDA and my understanding was that PDA determines peak purity but as peak is eluting in the solvent front the likelyhood is that it is spectrally impure.

What about collecting the peak, degassing and reinjecting on a systm where it is chromatographed? What about MS?

hi - if anon#2 = anon#1 then think about the ethanol. Ethanol is certainly detected in RP-HPLC at lower wavelengths. What is your detection wavelength?
Try to decrease %organic to get some more retention on the unknown peak. With a PDA you can get the UV spectrum. However, the UV spectrum will be of little help without some experience or reference libraries.
NO - the PDA can NOT determine peak purity. It might give a value called peak purity. Confusing -isnt it? The PDA measures UV spectra and some smart algorithms compare spectra from beginning, at the apex and at the end of the peak. Imagine a impurity coeluting with the active and having the same UV spectrum. Relying on the peak purity value you will not notice anything.

Alex

Just to clarify a few points.

PDA can give you useful purity information if the following conditions are met:

1: Impurity has some UV spectra
2: Impurity is partially resolved (PDA can not detect a perfect co-elution)
3: Impurity is there at a detectable level.
4: Spectra between co-eluters are different

What this really means is:

If a PDA detects an impurity, it is there. If a PDA says no impurities are present, there still may be an impurity peak.

When compounds elute in the void, there is usually no resolution between eluting peaks, therefore PDA can't really help with those void peaks.

Just a note to Alex, the better PDA purity algorithms compare every spectra across a peak to the apex, not just the leading and trailing edge.