Key takeaways
- Reversed-phase HPLC separates a peptide from related impurities by hydrophobicity; the UV detector reads the peptide bonds at about 214 nm.
- Area-% purity assumes every impurity absorbs UV like the main peptide. Tryptophan, for one, absorbs about 30 times more than a peptide bond at 214 nm.
- The method sets the number: column, solvents, gradient and integration can merge or split peaks.
- A pure-looking peak can still be the wrong molecule. Identity needs mass spectrometry.
01What reversed-phase HPLC does
High-performance liquid chromatography pushes a dissolved sample at high pressure through a narrow column. Different molecules travel through it at different speeds, so they leave the column one after another.
In the reversed-phase mode used for peptides, the column is packed with particles carrying a water-repelling surface, commonly C18 chains. The liquid flowing through starts mostly as water with a little acid, such as trifluoroacetic acid (TFA) or formic acid, and the share of an organic solvent such as acetonitrile rises over the run. This rising mix is the gradient. Molecules that avoid water cling to the column longer and come off later.
Peptides that differ by a single amino acid, a missing residue or an oxidised side chain usually differ slightly in hydrophobicity. That is why the technique can pull a target peptide apart from most of the by-products of its synthesis.
02What the UV trace shows
As molecules leave the column they pass a UV detector, usually set between 210 and 220 nm. At these wavelengths the peptide bond itself absorbs light, so almost every peptide is visible. The result is the chromatogram: time on the horizontal axis, absorbance on the vertical axis, one peak per separated component.
The peaks are not equal reporters. A 2007 study measured absorbance at 214 nm for each amino acid and for the peptide bond. Tryptophan absorbed about 30 times more than a peptide bond; phenylalanine, tyrosine and histidine about six times more; most other residues much less. A by-product that lost a tryptophan, or gained one, will draw a peak that is smaller or larger than its true share.
Retention time
When a peak leaves the column. Useful for comparing runs on the same method, not proof of identity.
Peak area
The area under each peak. Purity is calculated from these areas, not from peak height.
Baseline
The line the peaks rise from. Where the software draws it changes every area.
03How area-% purity is calculated
The software integrates each peak above a reporting threshold, adds the areas, and divides the main peak by the total.
Purity (area %) = main peak area ÷ sum of all integrated peak areas × 100. Illustration only, not from any real sample: a main peak of 980 area units with side peaks of 12 and 8 gives 980 ÷ 1,000 × 100 = 98.0%.
Three assumptions sit inside that number. Every impurity absorbs UV about as strongly as the main peptide. Every impurity separated from the main peak instead of hiding under it. And everything in the vial both absorbed at that wavelength and came off the column during the run. Water and inorganic salts fail the last test, which is why purity says nothing about how much of the powder is peptide. That is net peptide content.
04Why the method and gradient matter
The same sample can give different purity figures on different methods. In a 2018 Waters application note screening columns for synthetic-peptide impurities, the separation of the peptide bivalirudin changed with the column and with the acid or buffer in the mobile phase. A steep, short gradient can push close impurities into the main peak and raise the number.
Some impurities are hard to separate at all. A 2014 review of related impurities in peptide medicines lists deletion and insertion sequences, diastereomers from racemisation during synthesis, leftover protecting groups, oxidation products, dimers, and degradation products such as pyroglutamate and succinimide forms. A diastereomer has the same mass and nearly the same hydrophobicity as the target. In a second Waters note, impurities co-eluting with the main peptide peak were only found by mass spectrometry.
The validation guideline used by regulators, ICH Q2(R2), says specificity can be shown by absence of interference or by comparing with an orthogonal procedure, and that where one procedure does not discriminate enough, two or more should be combined. For a peptide, HPLC with mass spectrometry is that combination.
- A credible report names the column, the mobile phases, the gradient, the flow, the wavelength and the run time.
- A purity figure with no method and no trace cannot be checked or repeated.
05What HPLC can't tell you
| Question | Does area-% HPLC answer it? | What does |
|---|---|---|
| Is it the right molecule? | No. A wrong peptide can be pure. | Mass spectrometry |
| How many mg of peptide are in the vial? | No. Area % is a ratio. | Net peptide content, or a quantitative assay against a characterised reference standard |
| How much TFA or acetate is present? | No | Counter-ion analysis, such as ion chromatography |
| How much water is in the powder? | No | Water content, such as Karl Fischer titration |
| Are there aggregates? | Not reliably | Size-exclusion chromatography |
| Is there bacterial endotoxin, or anything alive? | No | Endotoxin and sterility tests |
FDA made the same point in its July 2026 advisory-committee briefing on BPC-157. One certificate submitted to it listed impurity totals without identifying the impurities, so FDA called the impurity profile unclear; another reported purity only; and a literature search found that most BPC-157 certificates carry only a purity result.
06Reading a chromatogram, step by step
Batch number, compound and test date agree with the vial and the rest of the certificate.
Column, mobile phases, gradient, wavelength. No method, no comparison.
One tall peak with its retention time and area listed in the peak table.
A shoulder, a split top or a long tail can mean something is hiding underneath.
Every visible peak is listed, the areas add up to the stated purity, and the baseline is drawn under the peaks rather than through them.
The trace covers the full gradient, so late-eluting impurities had a chance to appear.
A mass spectrometry result for the same lot confirms the main peak is the intended peptide.
Plain-English summary of analytical chemistry sources and regulator documents, checked 17 November 2026. Not legal or medical advice.