How to Review a Peptide Certificate of Analysis (COA)

Who issued the document, for which lot, and when

Every certificate is a statement made by a specific party about a specific quantity of material at a specific moment. The masthead establishes all three, and it is the part of the document most often skimmed.

The issuing party matters because certificates come from three sources that carry different weight. A manufacturer's own quality control group tests material it produced, which is normal practice and not a weakness in itself, but the figures are self-reported. An independent contract laboratory tests material submitted to it, which introduces an outside party and also a question about what was submitted. A distributor may reissue or summarize either of those on its own letterhead, at which point the reader is two steps from the instrument. A complete document names the facility that performed the testing, with an address, rather than leaving the reader to infer it from a logo.

Sample provenance is the field most often absent. Wording such as "results apply to the sample as received" is standard contract-laboratory language and it is honest, but it means the laboratory characterized what arrived in its sample vial rather than the lot in your freezer. What links those two things is the lot identifier and the chain of custody, not the certificate itself.

The lot or batch identifier converts a general product description into a record about material you can hold. It must be a specific alphanumeric string, and the same string must appear on the vial label. A document with no lot identifier is not a certificate of analysis; it is a product datasheet describing what the compound is supposed to be, and it supports no claim about any particular container.

Dates come in at least three flavors and a good document separates them. The manufacture or fill date is when the lot came into existence, the test date is when the reported measurements were taken, and the report date is when somebody signed off. Those can be months apart, and the gaps are informative. The signature block should name an individual and a role rather than carry an unattributed company stamp, because that line says who is accountable for the figures above it.

Masthead fields and what their absence leaves unresolved

Together these fields answer one question: does this paper describe the object in front of me, and can I tell who says so. If the lot string is missing or the signature is an unattributed logo, nothing further down the page can be attributed to your vial, however strong the numbers look.

The identity block: name, sequence, formula and theoretical mass

Below the masthead sits the block defining what the material is supposed to be. A competent reader can check this section independently, because three of its fields are arithmetically related to one another.

The compound name should be the established name for the sequence, not a category. A certificate naming a mechanism class rather than a molecule has described a family, and families contain compounds with very different sequences and masses. Where a compound is widely known by a code number or trade shorthand, both should appear with one designated as primary. Synonyms are not decoration; they are how a receiving lab reconciles a purchase order, a label and a certificate that each use a different name for the same thing.

The sequence should be printed in full, in one-letter or three-letter code, with terminal and side-chain modifications shown explicitly. Acetylation at the N-terminus, amidation at the C-terminus, a disulfide bridge or a fatty-acid conjugation each change the molecular formula and therefore the theoretical mass. A certificate that names a modified compound but prints the parent sequence has printed a formula that will not reconcile.

The molecular formula and theoretical mass are where the cross-check happens. Given a printed sequence the formula is determined, and given the formula so is the mass, which means a reader can recalculate both. That is the most useful property of the identity block, and it explains why a missing sequence matters more than it first appears: without it the theoretical mass cannot be verified, and the identity result later in the document is compared against a number nobody outside the issuing lab can reproduce.

One distinction is quietly important. The formula and mass for the free peptide and for the salt form are different, and most certificates report the free peptide while shipping the salt. That is conventional and correct, but the document should say which basis it uses, because the difference propagates into every mass calculation downstream.

Identity fields, correct form, and the defect each usually shows

A useful habit is to recalculate the theoretical mass from the printed sequence before reading any result. It takes a minute with a sequence calculator, it catches transcription errors in the identity block, and it means the observed mass further down is compared against a figure you derived rather than one you were handed.

Specification, result and method belong in the same row

The results table is the part everybody reads and the part most often misconstrued, because a row can be printed several ways and only one of them carries full information. A complete row has four parts: the attribute tested, the method used, the specification it is judged against, and the measured result.

Drop the method and the result becomes uncomparable. Purity by reversed-phase HPLC at one wavelength and gradient is not the same measurement as purity by another, and two lots cannot be compared unless both rows name the conditions. Drop the result and the row collapses into a restatement of the product definition: a line reading "purity: not less than 98 percent" tells you what the supplier intends to sell, not what this lot measured. That is a specification masquerading as a finding, and it is the most common weakness in circulating certificates.

Conforms and complies are the same problem in different clothing. They record that somebody made a pass-or-fail judgment, which is worth something, but they discard the value. Two lots both reading conforms against a 98 percent specification might have measured 98.1 and 99.6, materially different materials with identical paperwork.

Not detected needs a limit attached. A negative result is only as strong as the sensitivity behind it, so a residual-solvent or impurity row reading none detected with no limit of detection or quantitation stated is not a result. The same applies to identity rows reading conforms to reference spectrum without printing an observed mass: something was compared to something, and the reader is asked to accept the conclusion without the data.

The counterintuitive case is a row reporting a result outside its specification, annotated and signed. That is a stronger document than one with no out-of-specification results anywhere, because it shows results are transcribed from instruments rather than assembled to match the specification column.

How a results row is printed and what it is worth

When comparing two certificates for the same compound, compare row structure before comparing numbers. A document printing methods, specifications and results for six attributes is a different class of record from one printing two percentages and a signature, and that gap is larger than any difference between the percentages.

What a purity figure leaves out of the mass balance

Purity is the number everyone quotes and the one most often misread, because at least three quantities travel under that word and they answer different questions about the same vial.

Chromatographic purity is an area ratio within the ultraviolet-absorbing material that came off the column. If the figure is 98 percent, then 98 percent of the integrated peak area belongs to the target and the rest belongs to other peptide-related species: truncations, deletions, oxidized forms, diastereomers. That calculation is confined to the peptide fraction. Anything in the vial that does not absorb meaningfully at the detection wavelength never enters the arithmetic.

Net peptide content, sometimes called peptide content or peptide assay, is a mass ratio. It reports what fraction of the weighed powder is peptide rather than water, counter-ion and residual salt, and it is normally determined by amino acid analysis or a nitrogen determination rather than by chromatography. For a lyophilized research peptide it is routinely in the seventies or eighties, which is expected rather than alarming.

Two things fill the gap between those figures. The first is the counter-ion. Peptides purified by reversed-phase chromatography with trifluoroacetic acid as the ion-pairing agent are typically isolated as TFA salts, and TFA associates with basic residues, so a sequence carrying several arginines or lysines takes on proportionally more. The second is water: lyophilized peptide is hygroscopic and retains bound water, which is why a Karl Fischer or loss-on-drying figure appears on thorough certificates.

The practical consequence is that a vial labeled with a nominal amount of powder does not contain that amount of peptide unless the fill basis says so. A competent certificate or label states whether the figure is gross powder mass or net peptide mass. Where measurement arithmetic for preparing laboratory aliquots starts from the label figure, that convention is the difference between the concentration you calculate and the one you have.

Illustrative mass balance on a 100 mg powder basis; arithmetic only, not a specification

Those numbers are illustrative and should not be used as a correction factor for a real lot, because counter-ion load scales with the number of basic residues and water content depends on lyophilization and handling. The point is the shape of the arithmetic: a high chromatographic purity figure and a much lower peptide-by-mass figure are entirely consistent, and a certificate reporting the first without the second has left the mass question unanswered.

Appearance, water, counter-ion and the storage statement

The fields around the edges of a certificate get treated as boilerplate, and they are the ones most often carrying recoverable information.

Appearance is the cheapest test on the document and the only one a receiving lab can repeat without instruments. A stated appearance such as a white to off-white lyophilized powder or cake gives a technician something to compare against, and a discrepancy is a real finding: a collapsed or discolored cake in a vial whose certificate describes a white cake is evidence of something that happened between test and receipt. An appearance row reading conforms or as described forfeits that check.

Water content matters twice. It is mass, so it displaces peptide in the vial, and it is a reactant, so residual moisture drives hydrolysis and other degradation during storage. A water figure with no method is weak, since loss on drying and Karl Fischer titration do not measure quite the same thing, and a certificate with no water figure has left one of the two largest non-peptide fractions unquantified.

Residual solvents are carryover from cleavage, purification, precipitation and drying. A row saying the lot meets requirements without naming which analytes were screened is uninformative; the analytes actually looked for are the content of that test.

The counter-ion or salt form should be named. It changes the mass per vial, and TFA in particular is documented as affecting some cell-based assays, so a lab running that work needs to know whether it holds a TFA salt or an acetate salt.

The storage statement is where certificates get vague. A usable statement gives a temperature, says whether the container should be protected from light and kept desiccated, and separates conditions for the sealed lyophilized powder from conditions after reconstitution. Store in a cool, dry place is not a storage statement. Any retest or re-evaluation interval only means something read against the test date, because elapsed months since testing come out of that interval before the material reaches you.

Secondary fields, what each constrains, and wording too weak to rely on

The last two rows interact in a way worth stating plainly. A certificate with a generous storage claim and a test date well in the past describes the material as it was on the test date, not as it is on your bench. That is not a defect, but the reported purity is a historical measurement, and the further back that date sits, the more of the claim rests on storage and shipping rather than on the assay.

Filing a certificate so it still means something later

A certificate that cannot be retrieved and tied back to a specific container is functionally the same as no certificate, and that is usually where the process fails: not at review, but months later when somebody asks which lot produced a given experiment.

The cross-check at receipt runs across four artifacts, not one. The vial label, the packing list or invoice, the certificate, and the receiving log entry carry overlapping fields, and the useful exercise is confirming they agree rather than reading any one closely. Compound name, lot string and stated amount per vial should be identical across all four. Where a shipment contains vials from more than one lot, each lot needs its own certificate and its own log line; a single certificate filed against a mixed shipment quietly breaks the link for every vial it does not describe.

This is also where the difference between a certificate and a datasheet gets enforced. A generic datasheet carries a compound name, a sequence, a specification range and often a nominal purity figure, and it looks convincing enough to get filed as though it were lot-specific evidence. The test is mechanical: no lot string and no test date means no lot-specific claim. Filing one in place of the other is the hardest documentation failure to detect afterward, because everything else about the file looks right.

Filing itself should be boring and consistent. Name the stored file with the compound and the lot string so it sorts predictably. Record in the receiving log the file name, who reviewed it, on what date, and any discrepancy found and how it was resolved. Retain it at least as long as data derived from that lot remains in use.

Where a certificate is requested before purchase, it should be for the lot that will actually ship, and it is reasonable to ask for exactly that. A certificate for a previous lot shows the supplier tests material, which is genuinely useful, but says nothing about the container that will arrive.

Cross-checking a certificate against the receiving record

None of this requires software. A dated spreadsheet row per received lot, with a file name in one column and reviewer initials in another, closes the loop well enough that a question raised months later is answered from the record rather than from memory.

Questions this walkthrough gets asked

Where to read next

• Lot traceability: matching label, COA and records the receiving-log side of the cross-check in more detail
• Reading an HPLC chromatogram: purity by area where the chromatographic purity figure comes from
• Mass spectrometry for peptide identity confirmation how the observed mass is produced and judged
• Peptide purity versus peptide identity why the two figures cannot substitute for each other
• Third-party lab testing and certificates