How to select the UV wavelength for related substances?

[geneleo]

Would anyone like to share your experience about how to select the UV wavelength for related substance determination? It's quite sure the maximum UV wavelength is selected for assay determination of the main component itself. But for related substances, the UV absorption of them may be quite different with that of main peak. Mostly general wavelength such as 210 nm/220 nm or the maximum wavelength of the main peak will be selected. Could we select the maximum wavelength of the major impurities for detection?

[DR]

Provided you have a multichannel or diode array detector, sure you can. Scan the impurity of interest to determine an optimal wavelength, then use it (provided you have a relative response factor or standard of said impurity to quantitate against).

[Gerhard Kratz]

If you don't have a DAD, but a UV photometer just scan the UV spectra of your different compounds and overlay the spectra to find a wavelength susable for all your compounds.

[HPLCaddict]

First of all, if your'e in a regulated environment, don't even think of changing the detection wavelength without revalidating the whole method!

If you're indeed free to change the method or planning to revalidate it anyway, it might be a good idea to optimize the detection wavelength. The maximum absorption of the main compound is usually used, as unknown degradation products are supposed to be structurally similar and thus exhibit similar absorption properties - the best you can do if all degradants are unknown. If you actually know that the impurities have quite different absorption maxima, you may try to use a different wavelength in order to maximize the sensitivity of your method towards these impurities. Nevertheless, be sure that you're method still has enough sensitivity towards the main compound as unknowns are usually calculated against a reference standard of this main compound.
If you indeed change the detection wavelength, one of the first things I'd do is to check (at least quick 'n dirty ) if the LOD/LOQ of the adjusted method for the impurities AND the main compound still fits the specifications (regarding e.g. s/n ratio and peak area rsd%).

[HW Mueller]

The ideal method would be to switch to the optimum wave length of each compound at a retention window of this compound.

[Rob Burgess]

Following on from these discussions, I have a current instance where the UV profile of a starting material (potential impurity) is wildly different from the API. I haven't calculated the relative response factor for this impurity yet (@ the wavelength maxima of the API = 252 nm) but I hazard a guess its > 50. What would people do given this situation? I can't seem to find a common wavelength to suit both main peak API and ALL of the impurites (i.e. synthetic and degradation impurities).

PS. We are calculating impurity profile nased on % area/area results, so therefore we will see very diffferent % impurity profile results depending on what wavelength is selected.