How to read a certificate of analysis in two minutes

One page, three questions

A vial of white powder looks the same whether it holds the right peptide, the right peptide mixed with its near-misses, or something else entirely. Only analytical chemistry can tell them apart, and the certificate of analysis is where that chemistry is written down.

A good COA answers three questions. Is this the right molecule? How much of the vial is that molecule? And who measured it, when, and on which batch? A consortium of US proteomics labs convened by the National Cancer Institute put it simply: for purified peptides, the paperwork should include mass spectrometry data and the HPLC chromatogram itself, not just a number ^[3].

Field 1: product, lot number and dates

The lot (or batch) number ties every result on the page to one production run, and it should match the number on the vial label. Then check the dates: manufacture, testing, and retest or expiry. The test date has to come after manufacture.

Field 2: the testing lab and the signature

The COA should name the laboratory that ran the tests, with an address, a report number and the signature of the person who approved the results. An independent laboratory has no stake in whether a batch passes, which is why its report carries more weight than a seller's own sheet.

Many testing labs are accredited to ISO/IEC 17025, the international standard for testing laboratories, through national accreditation bodies linked by the International Laboratory Accreditation Cooperation (ILAC) ^[9]. A lab that exists, holds accreditation and will confirm a report number is a lab whose numbers can be checked.

Field 3: appearance

Usually one line: "white to off-white lyophilised powder". It is a crude check, but a yellowed, sticky or collapsed cake is worth a question.

Field 4: identity by mass spectrometry

A mass spectrometer weighs molecules by turning them into ions and measuring their mass-to-charge ratio (m/z). The COA should list a theoretical mass, calculated from the sequence, and an observed mass, measured on this lot.

Take BPC-157 as a worked example. Its formula is C62H98N16O22, which gives two "theoretical" masses ^[1]:

• Monoisotopic mass: 1,418.70 Da. The mass if every atom is its most common isotope (carbon-12, hydrogen-1 and so on).
• Average molecular weight: 1,419.5 g/mol. The mass averaged over the natural mix of isotopes, the number used for weighing out powder.

It is the difference between weighing one coin and the average coin in a jar. Neither is wrong; a COA just has to compare like with like.

Electrospray instruments add protons, so BPC-157 typically appears as [M+H]+ at about m/z 1,419.7 or as the doubly charged [M+2H]2+ at about m/z 710.4. "Theoretical 1,419.5, observed 1,419.7" is consistent. A gap of tens of daltons is not: a peak 57 Da light suggests a missing glycine, while one 22 Da heavy is usually a harmless sodium adduct. The NCI consortium showed exactly this kind of evidence in a lower-purity sample, where a peak matched the peptide minus its first residue ^[3].

Mass spectrometry has blind spots. Leucine and isoleucine weigh exactly the same, so telling them apart takes amino-acid analysis ^[4], and mirror-image (D-) amino acids have the same mass too. A correct mass also says nothing about how much of the powder is that molecule. That is HPLC's job.

Field 5: purity by HPLC

Reverse-phase HPLC pushes the sample through a water-repelling column with a rising gradient of acetonitrile in water and trifluoroacetic acid (TFA), and peptides leave in order of how strongly they stick ^[2]. A UV detector records them as peaks, usually at 210–220 nm, where the peptide bond itself absorbs light ^[2].

Purity is the area of the main peak divided by the total area of all peaks. The chromatogram should be attached, showing one dominant peak, a few small ones, the retention time and the detection wavelength.

That wavelength matters. Only aromatic side chains (phenylalanine, tyrosine, tryptophan) absorb at 250–290 nm ^[2]. BPC-157 has none of them, so a trace recorded at 280 nm would show almost nothing and could make a dirty sample look clean.

A chromatogram run with a steep gradient can hide impurities underneath the main peak. The NCI consortium recommended gradients of no more than 2% organic solvent per minute for exactly this reason ^[3].

The small peaks are usually the synthesis's near-misses: sequences missing a residue, truncated chains, leftover protecting groups or oxidised forms ^[7]. Our guide on how peptides are tested explains where each one comes from.

Purity is not the same as peptide content

Peptides are sold as salts: their charged groups pair up with counter-ions such as trifluoroacetate (from HPLC purification), acetate or chloride, and the powder also holds some water. Net peptide content is the percentage of the weighed powder that is actually peptide, excluding water and counter-ions ^[3]. It is normally measured by amino-acid analysis ^[3].

A worked example from the US Pharmacopeia shows the gap. A reference peptide measured 99.08% pure by HPLC, but acetic acid made up 5.58% of its weight and water another 1.5–2%, so each milligram of powder held about 0.91–0.92 mg of peptide ^[4]. With TFA salts the gap can be much larger: in one 2025 study, trifluoroacetate made up as much as 35% of the weight of the peptide salts tested ^[5].

The counter-ion is not always inert. In cell cultures, trifluoroacetate at 10⁻⁸ to 10⁻⁷ M reduced the growth of bone cells, enough to mask or invent an effect of the peptide under test ^[6], which is why some labs ask for acetate or hydrochloride salts.

Optional tests

Their absence is not a red flag by itself, but it limits what the material has been shown to be suitable for.

• Endotoxin. Bacterial cell-wall fragments that cause fever and inflammation, measured by the bacterial endotoxins test and reported in endotoxin units (EU) per mg or mL ^[8]. Most relevant for cell and animal work.
• Sterility. A growth test for live microbes, meaningful only for material filled under sterile conditions.
• Residual solvents and counter-ion content. Leftovers from synthesis and purification, each measured with its own method ^[4].
• Heavy metals or elemental impurities, and water content.

The two-minute table

Red flags

• No lot number. The COA then describes no vial in particular.
• A number with no data. "Purity 99%" without a chromatogram, or "MS conforms" without a spectrum, cannot be checked.
• A suspiciously round 99.99%. Real chromatograms produce numbers like 98.6% or 99.08% ^[4]. A clean 99.99% deserves a hard look at the chromatogram behind it.
• The same COA for every lot. Identical numbers, dates and chromatograms across batches mean the document is a template.
• A lab that cannot be verified. No address, no report number, or a lab that does not answer.
• A COA older than the product. A test date before the manufacture date, or a report for a lot that expired long ago.

At Synthos

Every Synthos lot comes with its certificate of analysis, and readers can ask to see the COA for a specific lot before ordering.

FAQ