Detection wavelength and response

A UV detector reports absorbance, and absorbance depends on what is absorbing. For peptides two chromophores matter: the amide bond of the backbone, absorbing strongly below about 230 nm, and the aromatic side chains of tryptophan, tyrosine and phenylalanine, absorbing around 275–280 nm.

Why the index prefers 214 nm

At 214 nm the molar response is approximately proportional to the number of amide bonds, so a deletion impurity missing one residue of fifteen responds about 7 % less strongly than the parent per mole. That is a small and predictable error. At 280 nm the same impurity, if it has lost the only tryptophan, responds essentially not at all — and an impurity that does not respond does not appear in the total, which inflates purity.

The consequence is that a purity figure measured at 280 nm is systematically higher than the same sample at 214 nm, and the gap depends on the aromatic content of the impurities rather than on anything about the material. The index records the wavelength on every report and does not pool 280 nm purity figures with 214 nm ones.

The dual-wavelength trick

Running both channels simultaneously gives a free piece of structural information: the 280/214 absorbance ratio at peak apex.

Main peak       A280 / A214 = 0.082
Impurity at −0.34 min   A280 / A214 = 0.081     → same aromatic content
Impurity at −0.66 min   A280 / A214 = 0.019     → aromatic residue lost or modified
Impurity at +0.48 min   A280 / A214 = 0.084     → same aromatic content

Reading: the −0.66 min species has lost or oxidised a Trp/Tyr.
For a compound whose expected impurity set includes both a des-Tyr
truncate and an oxidised-Trp species, this ratio distinguishes them
without a mass spectrum.

Peak purity by PDA

A photodiode-array detector acquires a full spectrum across the peak. If the spectrum at the leading edge, the apex and the trailing edge are identical, the peak is spectrally homogeneous — which is evidence against a co-eluting species with a different chromophore, and no evidence at all against a co-eluting diastereomer, whose spectrum is identical by definition.