Ask what proves a research chemical is what its label says, and nine times out of ten the answer is a certificate of analysis — the most cited quality document in this market, and the most routinely misread. Everything here concerns material sold for research use only, not for human or animal consumption of any kind; the question is document literacy, not personal use. A certificate is genuinely informative about what it reports. Trouble starts when readers treat it as a blanket warranty instead of a bounded analytical record. This article walks the line between the two: what the document establishes, what it can't tell you, and what separates a rigorous certificate from a thin one.
What a Certificate of Analysis Is — and Who Writes It
In one sentence: a certificate of analysis is a lot-specific record of which tests a supplier ran on a particular batch and what numbers came back. Nothing more.
Two words there do a lot of work. Lot-specific means the document attaches to a batch, not a product line — a different lot of the same compound is different material with a different chromatogram. Supplier means the testing was almost always run by the manufacturer's or seller's own quality-control laboratory. That doesn't make the results wrong; in-house QC is how most analytical chemistry gets done. It does make the certificate a first-party document.
Regulators approach it exactly that way. FDA's current good manufacturing practice guidance frames a supplier's report of analysis as something to be qualified rather than simply accepted — a starting point for evaluating a supplier, not a substitute for looking at the material yourself. For the row-by-row tour of a typical document, we cover how to read a certificate of analysis field by field separately. Here we're concerned with scope.
The Identity Section: What the Spectrum Actually Proves
The identity section answers one question — is this the compound named on the label? — and the answer rests on mass, not on retention time.
Mass spectrometry carries that burden. It confirms the molecular mass of the material, and fragmentation experiments let analysts verify the amino acid sequence of a peptide rather than just its total mass. Reference-standard programs built for regulatory work follow the same logic: identity verified by mass spectrometry, with infrared, NMR, and ultraviolet methods brought in for selected compounds.
What identity does not rest on is where a peak shows up in the chromatogram. Retention time belongs to the method — the column, the gradient, the temperature — not to the molecule. Two different compounds can elute together on one method and separate cleanly on another.
Mass has its own blind spot. Leucine and isoleucine are isobaric: same formula, same mass, different structures. No mass measurement tells them apart. That's why amino acid analysis sits alongside mass spectrometry on serious certificates, confirming that composition ratios match the theoretical values for the stated sequence.
One compliance note worth stating plainly. Several research compounds share names with approved pharmaceuticals — semaglutide, tirzepatide, BPC-157. A certificate on research-grade material of that name records what the laboratory measured in that lot; it does not make the material equivalent to the FDA-approved pharmaceutical product of the same name.
The Purity Section: Why One Number Isn't the Whole Number
The headline purity figure on most certificates is chromatographic purity, and it's a narrower claim than most readers assume.
Chromatographic purity is an area percentage. Run the sample, integrate the peaks, report the target compound's share of total detected peak area. Target accounts for 98% of detected area, purity reads 98%. What that figure describes is how peptide-related species are distributed among themselves — the share of the detectable compound family that is the molecule you wanted.
What it doesn't describe is how much of the powder in the vial is that compound. Two different quantities, and the gap between them isn't small. We unpack the distinction at length in net peptide content and HPLC purity are two different numbers, but the short version: a lyophilized solid can be 98% pure by area percent and still be a meaningful fraction water and salt by mass.
A second, quieter assumption sits inside area percent — that every impurity responds to the detector the way the main component does. For UV detection that depends on each species carrying a comparable chromophore, which is an approximation rather than a fact. Methods that sidestep detector response entirely, such as quantitative NMR, where the reference standard is independent of the analyte's chemical structure, exist precisely because that approximation has limits.