Nine red flags on a peptide Certificate of Analysis

This is the first check and the most decisive. A COA is a statement about one specific lot. If the number on the document and the number on the vial differ — or the vial carries no lot number at all — then whatever the document says, it is not evidence about the material in your hand.

The related pattern is a supplier who publishes one COA for a product and serves it for every subsequent order. See why batch numbers matter.

An unattributed report cannot be checked by anyone. A credible COA names the laboratory that performed the work and carries a report reference. Where the analysis was performed in-house that is not automatically a problem — but it should say so, rather than implying independence it does not have.

A typed percentage discards all the evidence behind it: baseline quality, peak shape, shoulders, integration limits, late-eluting material. There is no way to audit a number presented alone.

Publishing the trace costs a supplier nothing beyond the willingness to be looked at closely, which is precisely why its absence is informative.

Purity is method-dependent. Without column, gradient, run time and detection wavelength, the figure cannot be compared against anything — not another supplier, not another batch, not a published method. As covered in what HPLC purity actually means, a short steep gradient can hide co-eluting deletion sequences that a longer one resolves.

A reported purity of exactly 100.00% is not a measurement; every chromatogram has a baseline and every integration has a threshold. Similarly, figures quoted to three decimal places imply a precision the technique does not have.

Neither proves dishonesty. Both indicate a document written by someone who is not thinking about measurement uncertainty.

A spectrum image with no axis labels, no annotated peak and no stated theoretical mass conveys no information. The reader cannot tell whether the observed mass matches the sequence, or even what charge state is being displayed. See what mass spectrometry tells you.

Both dates carry information. Analysis at release describes the material as produced. A large unexplained gap between manufacture and analysis, or a missing date entirely, leaves the storage history — and therefore the degradation state — entirely unknown. Peptides degrade predictably in ways covered in storage and stability.

For a peptide, the amino-acid sequence is the identity. A COA that names only a trade name or an abbreviation has not stated what was tested with enough specificity to check the theoretical mass. Abbreviations are also not standardised across suppliers — the same short name is used for different molecules by different vendors.

A COA is a measurement record. Superlatives, therapeutic framing, or claims about what the compound does are not analytical results and do not belong on one. Their presence usually indicates a document produced by a marketing function rather than a laboratory — and in the UK, claims of that kind about unlicensed compounds carry their own regulatory problems entirely separate from the analytical question.

Three things are frequently treated as suspicious but are ordinary:

• Purity below 99%. 98% with a full method and a visible trace is a better document than 99.9% without.
• A TFA counterion. This is the normal consequence of reversed-phase purification, not a contaminant that snuck in.
• Small differences between laboratories. A point or two of variation between competent labs reflects method and integration differences, not disagreement about the material.