Understanding Purity Testing

High-performance liquid chromatography separates the contents of a sample into peaks as they elute from a column. The area of the main peak versus everything else gives you the purity percentage.

Mass spectrometry weighs molecules. By comparing the measured molecular weight against the theoretical weight for your peptide sequence, you confirm the molecule on the chromatogram is the molecule on the label.

HPLC alone tells you something is pure — not what it is. Mass spec alone tells you the right molecule is present — not whether contaminants are too. You need both for a defensible result.

The impurity profile in a peptide batch has a short list of sources, and knowing them makes a chromatogram readable. Truncation sequences — chains missing one or more terminal residues — arise from incomplete coupling during synthesis and are the most common related substances. Protecting-group residues and scavenger carryover come from the synthesis and cleavage steps. Oxidised variants (for sequences with redox-sensitive residues) and deamidated forms appear with handling and age.

Diastereomers are the hardest class to see on a standard HPLC trace: they have the same mass and nearly the same retention behaviour, which is why identity confirmation relies on mass spectrometry plus, for critical work, orthogonal methods. None of these sources is exotic — they are the ordinary chemistry of peptide manufacture, which is exactly why every batch is tested rather than sampled.

A single purity threshold can't serve every use. Reference-standard work and assay calibration need the highest available main-peak value because the compound itself is the measurement. Cell-culture exposure work adds endotoxin constraints on top of purity. Formulation R&D has different tolerances again, because the downstream matrix dominates the final product's quality.

This is why a serious supplier publishes per-compound specifications rather than one blanket percentage — the release bar for a long, aggregation-prone sequence may legitimately differ from that for a short, robust one. What should never vary is the honesty of the reporting: the measured value, the method, and the batch, every time.