A certificate of analysis is a lab report that says what a specific batch of material was tested for and what the results were. It is batch-specific by definition. If a supplier sends you the same PDF regardless of what you ordered or when, you have a marketing document, not a COA.

Here is what should be on the page and how to read each part.

The header, which is more important than it looks

Four things need to be present and consistent: the peptide name and sequence, the batch or lot number, the date of analysis, and the name of the testing laboratory. If the lot number on the COA does not match the number printed on the vial you received, the document is not about your material. That single check eliminates a large share of what is in circulation.

Look also at whether the testing lab is the manufacturer or an independent third party. In-house testing is not automatically dishonest, plenty of legitimate manufacturers run their own analytics, but a third party report from a lab you can look up carries more weight, and a third party report you obtained yourself carries the most.

HPLC purity

High performance liquid chromatography separates the components of your sample by how strongly they stick to a column. The output is a chromatogram: a baseline with peaks, where each peak is a distinct compound and the area under the peak is proportional to how much of it there is. Purity is reported as the main peak area as a percentage of total peak area, usually with UV detection at 214 nm because that is where the peptide bond absorbs.

What good looks like: 98 percent or higher for a research grade peptide, and you should see the actual chromatogram image, not just a number in a table. Study the picture. A single sharp symmetrical main peak with a flat baseline is what you want. Warning signs include a broad or split main peak, a cluster of small peaks eluting just before or after the main one (these are usually deletion sequences or oxidised variants from a sloppy synthesis), and any peak large enough to have its own labelled area sitting at over 1 percent.

The number alone can be gamed. Integration limits can be set to exclude inconvenient regions, and a chromatogram run over too short a gradient will not resolve close-eluting impurities. Check that the run time and gradient are stated. A ten minute run reporting 99.8 percent purity is not telling you much.

Mass spectrometry

HPLC tells you the sample is pure. It does not tell you the pure thing is the right thing. Mass spec answers identity by measuring molecular weight. The report should give you a theoretical or expected molecular weight from the sequence and an observed value, and they should agree to within about one dalton for a peptide of a few thousand daltons on a standard ESI or MALDI instrument.

Two practical notes. First, on ESI you will often see multiply charged species, so the raw spectrum may show peaks at M+2H over 2 and M+3H over 3 rather than the plain molecular ion. A deconvoluted mass should be reported. Second, a mass that is 16 higher than expected usually means oxidation, typically at a methionine, and a mass low by 18 suggests dehydration. These are the kinds of details a real analytical report will annotate and a fabricated one will not.

Endotoxin

Bacterial endotoxin (lipopolysaccharide from Gram negative cell walls) survives the sterilisation processes that kill the bacteria themselves. It is pyrogenic, meaning it causes fever, chills and inflammatory response, and it is the single most likely cause of a bad reaction to an otherwise correctly manufactured peptide. Testing is by LAL assay, reported in endotoxin units per milligram or per vial.

For anything intended for parenteral use the standard reference point is under 0.25 EU per mg, and pharmacopoeial limits for injectables work out around 5 EU per kg of body weight per hour. Most COAs in the grey market omit endotoxin entirely. That omission tells you the material was not made under conditions where anyone was thinking about it.

Water content and counterion

Two entries people skip past, both of which change your dose.

  • Water content by Karl Fischer titration. Lyophilised peptides are hygroscopic and typically carry 3 to 10 percent residual water. If your vial is labelled 5 mg and the material is 8 percent water, you have about 4.6 mg of peptide plus water.
  • Counterion content, usually trifluoroacetate from the purification step, sometimes acetate. TFA can account for 5 to 20 percent of the mass of a basic peptide. A COA that reports net peptide content (sometimes called peptide content by AAA or by nitrogen analysis) is doing you a real service, because gross weight and net peptide can differ by a quarter.

Put those two together and a vial sold as 10 mg gross might be 7.5 mg of peptide. If you are reconstituting to a calculated concentration, that is the difference between the dose you think you took and the dose you took.

Fast checks for a worthless COA

  1. No lot number, or a lot number that does not match the vial
  2. No date, or a date years old on freshly shipped material
  3. A purity figure with no chromatogram attached
  4. Chromatogram axes with no units, no run conditions, or an image that has clearly been cropped from another document
  5. No mass spec section at all
  6. The same file served for multiple different products (check the PDF metadata and file name)
  7. Perfectly round numbers everywhere, such as exactly 99.00 percent purity and exactly 0.00 percent impurity

The strongest position is not trusting a supplier's paperwork at all. Independent third party testing, where you send a sample of what arrived at your door to a lab yourself, costs a few hundred rand to a few thousand depending on the panel and settles the question completely.