by
lmh » Wed Nov 07, 2018 2:46 pm
The original assumption might just as well say "it can be assumed that the impurities have the same chromophore as the product", which is obviously a terribly bad assumption.
If the absorption spectrum for the impurity is the same as that of the product, then if you measure the product at a slightly suboptimal wavelength, the measurement of impurity will be equally suboptimal, and the calculation of percentage purity will be unaffected.
If the absorption spectrum of the impurity is different to that of the product then it's pretty much certain that its molar extinction coefficient will be different too (except, by purest accident, at some particular and unrecognisable wavelength), and in this case your calculations are, in any case, meaningless in an absolute sense; they can only used in a relative sense, to establish whether this batch is better than that batch. If working in a relative sense, again, it does not matter what wavelength you use for each.
There's no point in worrying too much about data based on dodgy assumptions; worry can't make the data better. It's more productive just to understand, and accept, the data's limitations. Not that the authorities will necessarily agree.