A certificate is only evidence if it can be tied to a laboratory, a batch and a method that someone other than the seller can check. The document itself is easy to produce; anyone with a template can generate one. What is hard to fake is consistency — between the certificate and the vial, between the certificate and the testing laboratory, between one lot's figures and the next. Judging whether a certificate is genuine is a matter of testing those consistencies, and most of the tests take a minute.
The checks, in order of cost
| Check | What a genuine certificate shows | What a doubtful one shows |
|---|---|---|
| Lot match | The exact string printed on the vial | A different lot, no lot, or a product code instead |
| Laboratory | A named testing laboratory, distinct from the seller | No laboratory, or the seller's own name |
| Method detail | Column, gradient, wavelength, instrument | "HPLC", "MS", nothing more |
| Raw data | A chromatogram, a mass spectrum | Numbers only |
| Figures | 98.7%, 4,113.6 Da | 99%, 100%, round numbers |
| Verification route | A code, a database, a laboratory contact | An image file and nothing else |
| Date and signature | Present, plausible, accountable | Missing, or the same date on every lot |
Why is the lot number the first check?
Because it is the only field that connects the document to the material, and it is the one most often wrong. A certificate for the compound is not a certificate for your vial; a certificate for last year's lot is not a certificate for this one. If the string on the label and the string on the document do not match exactly, everything else on the document is about something you do not have. On this site the check is done by entering the label's number on the lot verification page, which returns the document for that lot or nothing. The label is where the string comes from.
What does an independent laboratory prove?
That the seller is not grading its own work. A certificate issued by a named third-party laboratory carries that laboratory's accountability, and the laboratory can be contacted to confirm it issued the document. A certificate with the seller's own name in the laboratory field may be entirely honest, but it cannot be checked against anyone, which is a different level of evidence. Some laboratories provide a verification code or a searchable database of the certificates they have issued; where one exists, using it is the single strongest check available.
What the numbers themselves reveal
Real instruments produce untidy figures. A purity of 98.73% and an observed mass of 4,113.6 daltons are what a chromatogram and a mass spectrometer return. A purity of 99% and a mass of 4,114 have passed through a person who rounded them, which is not dishonest but is a step removed from the measurement. A purity of 100% is a claim area normalisation cannot produce, and a certificate carrying it has either misunderstood its own method or invented the figure.
The same applies across lots. Two certificates for the same compound showing identical figures to the decimal are suspicious; batch synthesis does not produce identical batches. Figures that vary within a plausible band across lots look like a process; figures that never vary look like a template.
Can a genuine certificate still be misleading?
Yes, and this is the more common problem than outright fabrication. A genuine certificate for a real batch can be presented for a different batch. A genuine purity figure can be quoted without the content figure that would change its meaning. A genuine chromatogram can be from a run stopped before the late-eluting impurities appeared. None of these involves a false document; all of them involve a true document doing more work than it can bear. Reading each panel for what it proves is the defence, and it is why the lot match matters more than the document's authenticity.
What about certificates supplied as images?
An image is a picture of a document, and it is what most certificates are in practice — a scanned PDF. That is not a problem in itself. It becomes one when the image is the only form the certificate exists in: no laboratory to contact, no database to check it against, no lot in a supplier's own records to cross-reference. An image that can be checked against something is a certificate; an image that cannot is a claim. The question to ask is not "is this a real file" but "what, other than this file, confirms it".
What should I do if a certificate fails a check?
Ask, before concluding. A lot mismatch is sometimes a supplier's filing error, resolved by a request for the right document. A missing laboratory name is sometimes a formatting omission. The response to the question is itself the information: a correct certificate for the right lot within a day is one answer, and a specimen document, a delay, or a purity figure typed into an email is another. The same test that vets a supplier applies to a certificate, because the certificate is where a supplier's practice becomes visible.
Is a certificate ever unnecessary?
For qualitative work where a fifth either way changes nothing, the certificate's numbers matter little. Its existence still does, because it is the record that a batch was characterised at all, and because the work that seems qualitative today is the work that gets compared to something next month. Keeping the certificate with the lot record costs nothing at the time and cannot be done afterwards. It is the one document that turns a vial in a freezer into a material with a history.
All products referenced here are supplied for laboratory and research use only. They are not drugs, foods, supplements or cosmetics, and are not for human or veterinary use.
