Batch-Matched vs Generic: The Difference Between a COA and a Brochure

A batch-matched certificate of analysis carries four things a brochure never does: the lot number printed on the container in front of you, test dates that fall before the material was packed and dispatched, an identifiable laboratory that performed the work, and numerical results reported against stated acceptance limits. A generic certificate — the same file attached to every order of a given item, often undated, often with the lot field blank, typed in by hand, or filled with a number that appears nowhere on the label — is a marketing document set in laboratory typography. It describes what some batch of that compound looked like once, somewhere, possibly years ago. It says nothing about the material you are holding.

This is not a vendor convention or a matter of taste. It is written into the documents that govern how analytical results may be reported. ISO/IEC 17025:2017, the international standard for the competence of testing laboratories, requires a test report to carry unambiguous identification of the item tested, the date the item was received, the date or dates on which the laboratory work was performed, the date of issue, identification of the method used, and an explicit statement that the results relate only to the items tested ^[1]. That last clause is the entire argument in one line: a report is evidence about the specific material that went into the instrument, and about nothing else.

What separates a batch-matched COA from a generic one

Work through these in order. The first failure ends the exercise — there is no point scrutinising a purity figure on a document that cannot be tied to the container it arrived with.

• Lot or batch identifier. The certificate names a lot, and that lot string matches the label character for character. A blank field, a handwritten addition, or a mismatch is disqualifying.
• Dates. There is a date of receipt of the sample, a date of analysis, and a date of issue — and all three precede the date the shipment left the supplier ^[1].
• Issuing laboratory. A named organisation with an address, not a logo. If the analysis was performed by a third party, ICH Q7 expects the certificate to name the laboratory that performed it and to reference the original manufacturer's batch certificate ^[2].
• Numerical results against limits. Each test is listed with its acceptance criterion and the number obtained, not a bare 'conforms' or 'pass' ^[2].
• Method identification. Each result names the analytical procedure that produced it — the technique, the column or instrument conditions where relevant, and the detection wavelength for chromatographic purity ^[1].

What a certificate of analysis is required to contain

Three independent bodies have published what belongs on such a document, and they agree closely. ICH Q7, adopted by the FDA as guidance for active pharmaceutical ingredients, states that authentic certificates should be issued for each batch; that the name of the material, its grade where relevant, the batch number and the release date appear on the certificate; that each test be listed with its acceptance limits and the numerical result obtained; and that the certificate be dated and signed by authorised personnel of the quality unit, showing the name, address and telephone number of the original manufacturer ^[2]. The World Health Organization publishes a model certificate of analysis with an equivalent field set, intended precisely so that a receiving party can read a certificate from an unfamiliar source and know what is missing from it ^[4]. ISO/IEC 17025 covers the same ground from the laboratory's side ^[1]; for what each individual entry on such a document is doing there, a field-by-field reading of a peptide certificate works through them one at a time.

Matching the lot on the paper to the lot on the container

The match must be exact and it must be visible on the physical label, not only in an order confirmation email. Lot strings are commonly formatted as a compound code plus a date plus a sequence — the specific scheme does not matter, but the string on the certificate and the string on the label are either identical or they are not. Transcription differences are the common failure: a hyphen dropped, a leading zero lost, an I read as a 1. Photograph the label at receipt and compare against the certificate on screen rather than from memory.

A second, subtler failure is the repackaged lot. Where material is subdivided or relabelled after the original manufacturer released it, the new lot number is not the tested lot number, and a certificate carrying the manufacturer's original batch number no longer describes the container in front of you without a documented chain linking the two. ICH Q7 addresses this directly: certificates issued by or on behalf of repackers, agents or brokers should name the laboratory that performed the analysis, reference the original manufacturer, and attach a copy of the original batch certificate ^[2]. If a supplier repackages and issues its own certificate with no such reference, the document has lost its connection to any analysis.

Why the test date must precede the ship date

A certificate dated after the material left the warehouse cannot have been produced from that material before release, which means the release decision was made on something other than the result you are reading. It may still be a real analysis of a real lot — but it is a retrospective document, and the sequence matters when you are reconstructing what was known and when. Certificates issued to close a gap after a customer asked for one are common, and they are usually detectable by exactly this ordering. ISO/IEC 17025 requires both the date of receipt of the item and the dates on which the laboratory activity was performed to appear on a report ^[1], and those two fields are what make the sequence auditable at all.

The corresponding weakness in an undated certificate is total. Without a date of analysis, no statement about stability, storage or age can be checked, and the document cannot be placed in time relative to anything. An undated certificate is not a weak certificate; it is a description with no evidentiary status.

Is the issuing laboratory identifiable and independent?

"Third-party tested" is a claim, not a credential. The verifiable version names the organisation, gives an address, and — where the laboratory holds accreditation — cites the accreditation body and the certificate number, which can be checked against that body's public register. Accreditation bodies themselves operate under ISO/IEC 17011, which sets requirements for how they assess and monitor the laboratories they accredit ^[6]; the laboratory is assessed against ISO/IEC 17025 ^[1]. The register lookup is the point of the whole arrangement, and it takes under a minute.

Two honest caveats. First, accreditation always has a defined scope: a laboratory accredited for one technique on one matrix is not accredited for everything it will accept payment to run, so the scope document is worth reading before treating the logo as decisive. Second, in this category most reported analysis is not performed by an accredited laboratory at all, and saying otherwise would be inaccurate. That does not make an unaccredited result worthless — a named laboratory with a stated method and a signature is far more accountable than an anonymous one — but it does mean the strongest available assurance is usually traceability and internal consistency rather than accreditation. Regulated supply chains handle exactly this residual uncertainty by not relying on the supplier's paperwork alone: under 21 CFR Part 211, a supplier's report of analysis may be accepted in place of full testing only if the receiving party conducts at least one specific identity test itself and validates the supplier's results at appropriate intervals ^[3].

What a generic or undated certificate does not tell you

It is worth being concrete about the gap, because a generic document still looks like evidence and files like evidence — and the published record of independent testing of grey-market vials is why the distinction is not an academic one. Here is what it leaves unanswered ^{[1] [2]}.

• Whether your lot was tested at all, as opposed to some earlier lot of the same item.
• Whether the sample tested came from the same synthesis, the same purification run, or the same supplier as your material.
• How old the result is, and therefore whether any storage or stability claim attached to it still applies.
• Which analytical method produced the purity figure, and at what detection wavelength.
• What the acceptance limit was, and therefore whether the reported result is close to it or comfortably inside it.
• Who is accountable if the number is wrong — an unsigned certificate names no one.

A purity number without a method is not a result

The single most over-read line on any certificate is a percentage next to the word purity. On peptide material, that figure is normally chromatographic purity by reversed-phase HPLC, reported as the area of the main peak as a percentage of total integrated peak area, most often with ultraviolet detection near 214 nm where the peptide bond absorbs. That is a genuine measurement, and it is not a measurement of how much of the powder is the compound. Lyophilised peptide material also contains residual solvent, water, and counterion from purification — trifluoroacetate is the usual one — so net peptide content by mass is routinely well below chromatographic purity. The two figures answer different questions, and a certificate that reports only one of them has answered only one.

Chromatographic purity is also blind to anything the detector does not see: a co-eluting impurity, a non-absorbing counterion, or a species that never left the column will not appear in the area percentage. This is why identity is confirmed by a separate orthogonal technique — mass spectrometry establishing the observed monoisotopic or average mass against the theoretical mass for the sequence — rather than being inferred from a purity number. Even that confirmation has a boundary: a matching mass does not distinguish between sequences of identical composition, which is the limit of what a mass match proves. ICH Q2(R2) sets out what validating an analytical procedure requires, including specificity, accuracy, precision and range ^[5]; a certificate that names its method lets you ask whether that work was done, while a certificate that reports a bare percentage does not.

A receiving check that takes five minutes

Run this at unpacking, before anything goes into storage, and record the outcome in whatever log you keep.

1. Photograph the container label so the lot string is legible.
2. Open the certificate and compare the lot string character by character. Stop here if it does not match.
3. Read the three dates — sample received, analysis performed, certificate issued — and confirm all three precede dispatch ^[1].
4. Note the issuing laboratory's name and address. If accreditation is claimed, look up the certificate number on the accreditation body's register ^[6].
5. For each test, check that a method, an acceptance limit and a numerical result are all present ^[2].
6. Confirm the certificate is signed or otherwise attributed to a named authorising person ^[2].
7. File the certificate against the lot, not against the product, so that the next shipment of the same item does not overwrite it.

The honest limit of all of this: even a flawless batch-matched certificate is a statement about a sample drawn from a lot, not a measurement of your specific container, and it is only as reliable as the laboratory that issued it. Document review narrows uncertainty; it does not eliminate it. That is precisely why the regulated version of this workflow pairs the supplier's paperwork with an identity test performed by the receiving party ^[3]. What a batch-matched certificate does give you is a traceable claim, attached to a named lot, dated, attributed to a named laboratory, and reproducible against a register — which is the difference between a record and an advertisement.