Reading a certificate of analysis

1. The lot number

Start here, not at the purity figure. A certificate that cannot be tied to the vial in your hand is decoration. You want a three-way match: the lot on the vial, the lot on the certificate, and the lot on the invoice. All three. Members have found mismatches, and a mismatch is a serious finding worth posting in #vendor-verification with photographs of both.

2. Purity, and what “purity” means here

Almost always HPLC area percent at a stated wavelength: the area of the main peak as a fraction of the total integrated area. Three things follow.

• It is a proportion, not a quantity. It says nothing about how much is in the vial.
• It depends on the detection wavelength. Peptides are usually read near 214nm, where the amide backbone absorbs and so nearly everything shows up. At 280nm you only see aromatic residues, and an impurity without one is invisible.
• It depends on the gradient. A shallow gradient can bury a closely-related impurity under the main peak and produce a better number than a steeper, more honest one.

A certificate that says “>98%” rather than a specific number is telling you nobody integrated carefully.

3. Assay / content

This is the milligrams. It is the test that catches the most common failure in this market, which is an underfilled vial rather than impure peptide, and it is the line most certificates omit. If it is absent, you can approximate it yourself: weigh the sealed vial, subtract a tared empty of the same type, and compare against the label — but only if you also know the water and counterion content, which is the next two sections.

4. Water content (Karl Fischer)

Lyophilised peptide is hygroscopic and holds water. Water content of a few percent is normal and entirely honest, and it is the reason a perfectly good 10mg vial can assay at 9.1mg of peptide. If a certificate reports it, that supplier is telling you something most do not. If it does not, ask — about half will send it.

5. Counterion content

Synthetic peptides come out of purification as a salt, usually trifluoroacetate or acetate. The counterion has mass. So net peptide content = gross powder mass − water − counterion, which is why an honest 10mg vial can contain 11.4mg of powder. TFA is heavier than acetate for the same peptide. If the certificate does not state the salt form, you cannot do the mass balance and any weighing exercise is guesswork.

6. The related-substances table

The interesting part. It lists what else was seen, usually by relative retention time. A table populated with real numbers means somebody looked. A column of “not detected” means either a clean synthesis or a search that was not conducted, and you cannot tell which from the document.

Common entries worth recognising: +1 Da is deamidation, +16 Da is oxidation, a mass deficit matching one or more residues is a truncation or deletion, and something eluting late at roughly double the mass is a dimer.

7. What the certificate is not telling you

• Identity. Retention time is a hint, not a fingerprint — two different peptides can co-elute. A stated molecular weight from a mass spectrometer is much stronger. Actual proof of identity is MS/MS sequencing, and you will almost never see it.
• Sterility and endotoxin. Separate tests, almost never run on research material. A purity certificate says nothing about either.
• Whether this lot is your vial. Suppliers test a lot, not a vial.

8. Spotting a forged or recycled certificate

1. The same chromatogram on two different lots. The single biggest tell there is. Keep a dated folder and compare.
2. A certificate date that precedes the lot date. Sometimes sloppiness, sometimes not.
3. Mixed fonts, inconsistent kerning, a number that sits slightly off the baseline. A retyped figure in an otherwise consistent document.
4. Retention times that do not move at all between runs months apart. Real columns age.
5. A laboratory name with no method, no instrument, no column and no operator. Real reports name their conditions.

The short version

Independent testing referenced throughout the archive comes from Janoshik, Medutest and PeptideMeter, with supplier writeups cross-checked against VendorInvestigate. Two laboratories reporting a point or two apart on the same lot is normal and constantly mistaken for one of them being wrong.