What peptide purity percentages do and do not mean

When a peptide is described as being of a given purity, the figure almost always comes from high-performance liquid chromatography. The sample is separated into components, a detector records a signal as each component passes, and the area of the target peak is expressed as a percentage of the total area of all peaks detected. The phrase for this is area normalisation, and every word of it constrains what the number means.

What the number measures

It measures the proportion of the detected signal attributable to the target peak, under the conditions of that particular method, using that particular detector. Three limits follow immediately.

1. Only what is detected is counted. Ultraviolet detection at a wavelength chosen for peptide bonds responds poorly or not at all to species that lack the relevant absorbance. Salts, water and many residual solvents contribute little or nothing to the trace and are therefore not in the denominator.
2. Only what is separated is counted separately. An impurity that co-elutes with the target sits inside the target peak and is counted as target. Resolving it requires a different method, which is why a single chromatogram is weaker evidence than two obtained by different mechanisms.
3. Response is not uniform. Area percentage treats every peak as though the detector responds to it identically. It does not, so area percentage is an approximation of composition rather than a measurement of it.

Purity presumes identity

A chromatogram shows that one component dominates. It does not show that this component is the intended peptide. That is a question of identity, and it is answered by mass spectrometry, by sequencing, or by comparison against a characterised reference standard. A pure sample of the wrong molecule produces an excellent purity figure.

Identity and purity are therefore complementary rather than alternatives, and a certificate reporting one without the other has answered half the question.

Where impurities come from

Peptide impurities are frequently close relatives of the target rather than unrelated contaminants. Chains missing a residue, chains carrying a residue twice, incompletely removed protecting groups, oxidised methionine and deamidated asparagine all arise from ordinary synthesis chemistry. Because these species resemble the target closely, they are the hardest to separate and the most likely to be counted as target by a method that was not designed to resolve them.

Reading a purity claim in practice

• Ask which method produced the figure, and at what detection wavelength.
• Ask whether identity was confirmed separately, and by what technique.
• Ask whether net peptide content was determined, and treat it as a different number from purity.
• Ask whether the figure belongs to the batch in front of you or to a representative batch. Those are not the same claim.
• Treat a purity figure quoted without a method, a batch and a date as an unsupported assertion.

None of this makes purity figures useless. It makes them specific. A specific number, correctly understood, is more useful than a general impression of quality.