Red Flags on a Certificate of Analysis

No Lot Number Printed on the Report

Every certificate of analysis should carry a lot number (sometimes called a batch number) that matches, character for character, the lot number printed on the vial label or shipping documentation. This is the single link between the paper and the physical material. Without it, the certificate could describe any batch the supplier has ever tested, including one that has nothing to do with what you received.

A lot number is not a serial number for the sale or the shipment. It identifies a specific manufacturing or fill run, meaning every vial from that run should share the same certificate and the same result set. If a supplier cannot tell you which lot number belongs to the vial in front of you, or reissues a generic report with no lot reference at all, the report has no evidentiary value regardless of how good the numbers look.

Some suppliers work around this by printing a lot number on the certificate that refers to their own internal purchase order rather than the manufacturer's actual production batch. That number will look correct at a glance and still tell you nothing about which physical fill run the vial came from, so it is worth asking directly whether the lot number on the page is the manufacturer's own batch identifier.

No Analytical Method Stated

A purity or identity figure without a stated method is an unverifiable claim. High performance liquid chromatography (HPLC), mass spectrometry, and Karl Fischer titration each measure different things and each has its own limitations. A legitimate report names the method for every result: HPLC purity by area percent at a stated wavelength, mass confirmation by LC-MS or MALDI-TOF, water content by Karl Fischer titration, and so on.

Method matters because it tells you what the number can and cannot rule out. An HPLC purity figure with no method detail could have been run on a column and gradient poorly suited to resolve closely related impurities, inflating the apparent purity. A report that just states "99% pure" with nothing else is not a result, it is an assertion.

The instrument model, column type, and detection wavelength do not need to appear on a routine commercial certificate, but the method category itself should always be named. If a report lists a purity figure next to nothing but a percent sign, ask which method produced it before relying on the number for anything.

A Purity Figure With No Chromatogram Attached

The number on a certificate is a summary of a chromatogram, not a replacement for it. A trustworthy report either attaches the actual HPLC trace or makes it available on request, showing the main peak, retention time, integration parameters, and any minor peaks with their relative area percentages. If a supplier will only give you the final percentage and refuses to produce the underlying trace, there is no way to check whether the integration was done honestly.

Small variations in how a peak is integrated, where the baseline is drawn, and whether shoulder peaks are folded into the main peak can move a reported purity figure by several percentage points without changing the underlying chemistry. A chromatogram lets a chemist judge that directly instead of taking the summary number on faith.

When you do get a chromatogram, check that the axis labels, the lot number, and the run date on the image actually match the certificate it supposedly supports. A trace with no lot reference printed on it can be swapped between reports just as easily as a bare number can.

A Report That Names No Laboratory

A certificate of analysis should identify the testing laboratory by name, and ideally by address and accreditation status. A report that only carries a supplier's own logo, with no laboratory name anywhere on it, tells you the supplier is vouching for its own material with no independent check at all.

This is different from an in-house quality control report, which has its place but should be labeled as such. The problem is a document formatted to look like third-party testing that on close reading never actually names who ran the instrument. If you cannot find a laboratory name, ask for it directly, and call the lab to confirm the report is real if the stakes justify it.

A named laboratory should also be searchable. A lab name with no findable address, phone number, or web presence anywhere is functionally the same as no laboratory named at all, even if a name does appear somewhere on the page.

Missing or Implausible Dates

A complete certificate carries a test date (or analysis date) and, where applicable, a report issue date. These should be close together and should make sense relative to the lot's manufacturing date. A report with no date anywhere cannot be checked against anything, and one where the test date predates the material's manufacture is a copy-paste error at best.

• No date at all on the document
• A test date years before the lot could plausibly have existed
• An issue date that is suspiciously identical across many different lots
• A date format that changes between one section of the report and another, suggesting the document was assembled from two different sources

A reissued certificate, one regenerated after an original was lost, should say so and should carry the date of reissue alongside the original test date. A reissued document presented as if it were the original, with only the newer date visible, hides how old the underlying analytical work actually is.

The Lot Number on the Report Does Not Match the Vial

This sounds obvious, but it is the single most common mismatch and the easiest to check. Read the lot number off the physical vial label and compare it character by character to the lot number on the certificate. A transposed digit, an extra letter, or a lot number that is close but not identical is not a clerical accident worth ignoring. It means the paperwork you were sent does not describe the material you were sent.

If a supplier explains the mismatch as "the same batch, just an old label" or "we update lot numbers internally," ask for a corrected certificate with the matching number before regarding the material as verified.

This check takes seconds and catches more genuine problems, both innocent shipping errors and deliberate substitution, than any other single item on this page. Make it the first thing you do with every new vial, not an occasional spot check.

Identical Figures Across Different Lots

Real analytical results carry natural variation. Two different lots of the same peptide, run independently, will show slightly different purity figures, slightly different retention times, and slightly different water content, even when both are well within specification. If you collect certificates for several lots from the same supplier and the purity figure reads 99.2% on every single one, down to the decimal point, that is not consistency. That is one report being reused with the lot number swapped.

The tell is usually in the numbers that should be hardest to fake by accident: retention time to the hundredth of a minute, or a water content figure that repeats exactly across supposedly independent runs. Request certificates for two or three different lots side by side and compare them directly.

A small amount of consistency is expected and even desirable, a well-controlled manufacturing process should produce similar purity figures from run to run. The distinction is between similar and identical. Numbers that vary in the tenths place but repeat exactly past that point, run after run, are the pattern worth questioning.

A Scanned Image With No Traceable Source

A certificate delivered as a flattened image (a photo of paper, or a screenshot) with no accompanying PDF, no letterhead detail that survives zooming in, and no way to trace it back to an issuing lab is much easier to alter than a native document. This does not mean every image-based certificate is fraudulent, plenty of labs still fax or scan reports, but it removes a layer of verification that a native PDF with embedded text and metadata provides.

Where possible, ask for the original PDF rather than a photo of a printout. A PDF retains selectable text, creation metadata, and often the name of the software that generated it, all of which give you more to check than a flattened image ever will.

If a PDF is available, try selecting and copying text out of it rather than just reading it on screen. A document built from a genuine export usually has clean, selectable text throughout; one built by pasting a scanned image into a template often has text that behaves inconsistently or will not select at all.

Results Reported to Impossible Precision

Analytical instruments have real, known limits of precision, and a certificate that reports figures beyond what the method can actually resolve is a sign the numbers were written rather than measured. HPLC purity by area percent is typically meaningful to one decimal place; a figure reported to four decimal places (99.8734%) implies a level of instrument precision that does not exist for that method.

The same logic applies to a mass confirmation reported with more decimal places than the mass spectrometer's resolution supports, or a water content figure reported to a level Karl Fischer titration on a small sample cannot actually achieve. When the precision on a report exceeds the precision of the instrument that supposedly produced it, the number was not measured, it was typed.

Oversized precision is also a useful tell because it is hard to fake convincingly by accident. A person building a fabricated report often reaches for extra decimal places to make a number look more scientific, when in fact a real result from a real instrument rounds off well before that point.