How to read a certificate of analysis

A certificate of analysis is a laboratory's record of what it found in one batch of material, on one date, using named methods. It is a statement about that batch and nothing else. Two vials with different batch numbers have two certificates, and a certificate for one tells you nothing about the other.

This is why a figure is worth asking for by batch number rather than by product name, and why the page on reading a certificate starts from the number printed on the label.

Purity by HPLC

High-performance liquid chromatography separates a sample into its components and measures the area under each peak. The purity figure is the area of the main peak as a percentage of all the peak area detected. A figure of 99% means the target peptide accounts for 99% of what the column separated.

The impurity peaks matter as much as the headline. A batch reported at one purity figure with a single large impurity is a different material from one reported at the same figure with four small ones, and a certificate that gives only the headline has not shown you which you have. A report worth reading lists every impurity peak.

The chromatogram

The purity figure is a summary. The chromatogram is the record it was calculated from, and it is only readable with axes: retention time along one, detector signal along the other. A picture without scales is decoration.

Two places are worth looking at. The baseline should sit flat before and after the main peak, and the integration limits should be visible, because where the integration starts and stops decides the number. And the region immediately after injection: whatever appears there travelled with the void volume and was not retained by the column at all. It is usually salt or the solvent front rather than peptide, and it does not belong in the purity calculation.

The case that matters is a shoulder on the main peak. Deletion sequences, missing a single residue, elute close to the target, and on a short gradient they disappear under its flank. A report showing a high figure, one narrow peak and a very short run time has shown less than it appears to.

Mass by mass spectrometry

Purity says how much of one thing is in the vial. Mass says which thing it is. A certificate states the measured mass beside the theoretical mass calculated from the sequence, and the two should agree to within the resolution of the instrument.

This is the figure to check first, because it is the one that catches a substitution. Two compounds sold under the same name can differ by a factor of five: TB-500 is used for both full-length thymosin beta-4, around 4963 Da, and for a short fragment of around 889 Da. CJC-1295 exists with and without a DAC complex, about 3647 g/mol against about 3368. A listing that gives a name and no mass has not told you which one it is selling.

The public record shows how easily that goes wrong, and it is worth working through once. UniProt P62328 lists thymosin beta-4 as 44 residues at 5053 Da. That is the translated chain, initiator methionine included. The same entry records that the methionine is removed and the new N-terminus acetylated, and the arithmetic of that - 5053 less the methionine residue, plus the acetyl group - lands near 4964, which is where the figure around 4963 Da comes from.

Both numbers are defensible and they describe different things: one a database chain, the other the material in a vial. Neither is recoverable from a product name, and a certificate reporting a measured mass is what tells you which one you have.

Peptide content

The share of the whole vial contents that is peptide at all. It is typically 75-85% for a lyophilised peptide, and the remainder is not contamination - it is residual water and the counter-ions left behind by synthesis and freeze-drying, usually acetate or trifluoroacetate. Trifluoroacetic acid does both jobs on the usual route, cleaving the peptide from the resin and then pairing ions in the reversed-phase purification, so freeze-drying leaves a trifluoroacetate salt unless it is deliberately exchanged. The review Pharmaceuticals 2025, on TFA as a counterion in synthetic peptides sets that out.

Purity and peptide content answer different questions and a seller quoting only the first has answered the easier one. There is a separate guide on why.

Water content

Measured by Karl Fischer titration. A lyophilisate holds some residual moisture; a figure in the low single digits is ordinary. It is part of the reason peptide content is not 100%.

The method, the date and the laboratory

A result with no method is not a result. A report should name the method - RP-HPLC, ESI-MS - the date of the test, the laboratory, and that laboratory's own report number, which is what lets the finding be checked somewhere the seller does not control.

A specification says what a material is supposed to meet: purity at least this, water at most that. A certificate of analysis says what was measured in one particular batch. The first is an intention and the second is a finding, and the two are routinely offered under the same heading.

The difference is visible in the document. A specification carries limits and words like minimum and not more than. A certificate carries single values, a batch number, a test date and the laboratory's own report number. Where a page headed Certificate of Analysis contains only limits, it is a specification with the wrong title, and it says nothing about the vial in front of you.

The same test applies to the specification table on a product page, including the ones on this site. What those state is what is expected of a material of that class. The measured value for one batch is in the certificate issued by the laboratory that tested it.

It cannot tell you what the material does. A certificate is an analytical result, not a finding about biology, and any seller who moves from one to the other is no longer quoting a laboratory.

It cannot tell you that the vial in your hand is the vial that was tested. That link is the batch number on the label, and it is only worth anything if the number can be looked up independently.

And a certificate that exists only as a PDF on the seller's own site proves very little. It is a document, and documents are editable. Ask whether the laboratory publishes the same result under its own report number.

Two limits remain even when a report contains everything it should. They are properties of the methods rather than oversights by the laboratory.

The first is stereochemistry. A peptide with one residue in the D form instead of the L form has the same molecular formula and the same mass as the correct one. Mass spectrometry cannot tell them apart because there is nothing there to tell apart: they weigh the same. Chromatographically they are diastereomers, separable in principle on an achiral column but not thereby separated in practice. For sequences that contain D residues on purpose this is not a marginal question: ipamorelin carries two.

The second is coelution. Anything leaving the column at the same time is integrated as one peak and counted as one substance. That is why a second method run under different conditions says more than a higher number from the same one, and why 99% purity with no column, gradient or run time stated is a weaker claim than the figure suggests.

The check that separates most offers costs three lines. A contract laboratory will confirm its own report on request, and has no reason not to.

Three outcomes are possible and all three are answers. The laboratory confirms, and the report is genuine. The laboratory does not, and the question is settled. Or the seller names neither a laboratory nor a report number, so there is nobody to send the enquiry to.

A manufacturer's certificate and a report from an independent contract laboratory are not the same document, even when they carry the same lines. The manufacturer is testing what it produced. The contract laboratory is testing a sample that was sent to it, and has no commercial interest in the outcome.

Either can be sound, and a manufacturer's certificate is not worthless: in chemistry it is the normal case. The difference is what you are relying on, the manufacturer's own quality system in one case and a second party with nothing to sell in the other.

An offer that does not say which one it is has not answered the question. And where a contract laboratory is named, the second question is who drew the sample: one selected and sent by the seller says less about the rest of the batch than the document implies.

A certificate carries a test date, and the date is not a formality. It says when the batch was examined. It does not say how long the result holds.

Two intervals are worth comparing. Between manufacture and testing: a test long after manufacture examined material that had already been stored. And between the test date and today: a lyophilised peptide kept dry and cold is comparatively stable, but a report from years ago describes the state it was in then.

A report substantially older than the goods on offer usually describes an earlier batch. The batch number is where that shows up, and it is the only place it shows up.

1. Is there a certificate for this batch number, or one certificate reused across the product?
2. Does it state the measured mass, or only the purity?
3. Are the impurity peaks listed, or only the headline figure?
4. Can the result be checked with the laboratory rather than only with the seller?
5. Does the document carry single measured values, a batch number and a report number, or only limits?
6. Are the column, the gradient and the run time stated, so the purity figure can be placed?

The same questions are worth putting to this site. What Peptorium publishes is the chemistry that identifies a material and what a complete certificate contains; a measured figure belongs with the report and the laboratory that produced it. That is the standard the six questions above hold anyone to, this site included.