What “98% by HPLC” Is a Percentage Of

Purity on a peptide certificate is almost always an HPLC area percentage. It is a precise number describing something narrower than most readers assume, and the gap between what it measures and what it is taken to mean is where most misreadings start.

It is a percentage of area, not of mass

The calculation is simple: integrate every peak in the chromatogram, add the areas, divide the main peak’s area by that total, multiply by a hundred.

Nothing in that arithmetic involves mass. Peak area is proportional to the number of molecules passing the detector multiplied by how strongly each one absorbs, so the figure describes a ratio of detector responses. It becomes a mass ratio only if every species responds identically per unit mass, which they do not.

What is excluded by construction

Only things that elute and absorb are in the denominator. That leaves four categories outside the number entirely:

• Water. A lyophilised peptide typically carries several percent residual moisture. It has no chromophore and does not appear.
• Counter-ion. Trifluoroacetate or acetate can be a substantial fraction of weighed mass. Not in the chromatogram.
• Salts and non-absorbing residues. Buffer components, residual reagents and anything else without a chromophore.
• Anything that does not elute. Material retained irreversibly on the column, or aggregates too large to pass, are absent from the total.

So a vial can be 99 percent pure by area and contain considerably less than 99 percent peptide by mass. The two questions are answered by different measurements: area percentage by HPLC, mass fraction by amino acid analysis or nitrogen determination, water by Karl Fischer titration. A document reporting one has not reported the others.

The denominator depends on the method

Two laboratories analysing the same material can compute different denominators without either making an error.

A shallower gradient resolves species that a steeper one merges, adding peaks to the total. A different detection wavelength changes which species contribute and by how much. A longer run captures late-eluting material a shorter one never sees. A different column selectivity separates a different subset of the impurities.

Each of those changes the sum in the denominator, and therefore the percentage, from the same vial.

Where the baseline goes

Integration is a judgement as well as a calculation. Where the baseline is drawn under a peak, and whether a shoulder is integrated separately or absorbed into the main peak, both move the number.

The effect is largest exactly where it matters most: on a broad or shouldered main peak, of the kind proline-rich sequences produce. Two competent analysts can differ by several tenths of a percent on the same trace. That is the practical floor on the precision of the figure, and it is discussed in peak integration and purity.

A single peak is not a single compound

Area percentage assumes each peak is one species. Anything co-eluting with the main peak is counted as main peak, which inflates the figure rather than reducing it.

This is the failure mode that matters most, because it moves the number in the flattering direction and leaves no visible trace on the chromatogram. Detecting it requires a second, orthogonal method rather than a closer look at the same trace. The mechanism is covered in co-elution and peptide purity.

Response factors make it non-comparable across species

At 214 nm, response scales roughly with the number of backbone amide bonds. A short truncated impurity therefore contributes less area per mole than the intact peptide, and is systematically under-represented.

At 280 nm, response depends on aromatic residues, so an impurity that lost the only tryptophan can be nearly invisible while remaining fully present. Neither wavelength gives a proportional picture of the mixture, and the distortion runs in different directions depending on which impurities are present. This is set out in why two peptides at 98 percent are not equally pure.

What the figure is genuinely good for

None of this makes area percentage a bad measurement. It is sensitive, reproducible within a method, cheap enough to run on every lot, and it detects the impurity classes that chemical synthesis actually produces: deletions, truncations, oxidations, incomplete deprotections.

Its proper use is comparison of like with like — one lot of a compound against another lot of the same compound by the same method. That comparison is exactly what it was designed to support, and within it the number is reliable.

What makes a reported figure interpretable

A wavelength, a gradient, a column and a run length. With those, a reader knows what was in the denominator. Without them, the number is a percentage of an unspecified total, and comparing it to a percentage from another laboratory compares two different calculations. Related reading: reading an HPLC chromatogram and net peptide content explained.