How to read a peptide certificate of analysis

What a certificate of analysis is for

The certificate of analysis (COA) summarizes the tests performed on a specific batch. For a laboratory or a research group it is the first quality filter: peptide-related impurities can influence functional studies and lead to erroneous conclusions². The EMA guideline, in effect since 1 June 2026, lists the expected specification, from identification and purity to counter-ion, residual TFA, water, residual solvents, bacterial endotoxins and microbiological purity⁵. That list is a good template for reading any certificate.

HPLC purity: what "≥98%" means

The most visible number usually comes from reversed-phase liquid chromatography with an ultraviolet detector, the routine quality-control method for peptides³. The calculation is area normalization: the area of the main peak against the area of all detected peaks. In a quality-control study from Ghent University, detection was at 210 and 215 nm, with a reporting threshold of 0.1% of the main-peak area¹; detection at 220 nm has been described for oxytocin and acetates⁸. So "≥98%" means that at least 98% of the integrated ultraviolet signal belongs to the main peak.

That definition has limits worth knowing. Water and counter-ions are measured by separate tests and are not part of that percentage⁵⁷. An impurity that elutes with the main peak is hidden inside it: the EMA asks manufacturers to consider the risk of co-eluting impurities and, when they appear as one peak, to apply the 1.0% qualification threshold⁵. A good certificate states the method and attaches the chromatogram with the integrated peaks; FDA rules require laboratory records to keep those data¹⁵.

What is in the remaining percentage? Mostly impurities typical of solid-phase synthesis: sequences missing an amino acid (deletions) or carrying an extra one (insertions), diastereomers from racemization, protecting-group adducts, oxidations and dimers². Regulators set precise thresholds. Under the European Pharmacopoeia, peptide-related impurities are reported above 0.1%, identified above 0.5% and qualified above 1.0%⁵. The FDA, in its guidance on synthetic generics of glucagon, liraglutide, nesiritide, teriparatide and teduglutide (summarized in an official 2022 presentation), asks that any new impurity between 0.10% and 0.5% be identified and justified, and states that a new impurity "> 0.5% is not acceptable"⁶.

Identity: mass spectrometry

Purity tells how much of the material is a single component; it does not say which component. That is what mass spectrometry is for. Electrospray ionization (ESI) produces multiply charged ions: the spectrum of a large molecule appears as a series of peaks, each differing from its neighbor by one charge⁴. That is why a peptide of mass M shows up at mass-to-charge ratios such as (M + 2H)/2 or (M + 3H)/3, and the instrument software reconstructs the mass from that series. The EMA asks for tables of theoretical and observed mass values⁵. When reading a certificate, check that the observed mass matches the theoretical mass of the stated sequence, that it is clear whether monoisotopic or average mass is reported, and that the spectrum is attached.

Mass has a blind spot: two diastereomers weigh exactly the same, and racemization is a known synthesis impurity². The EMA recommends at least two orthogonal identification methods (for example, mass together with relative retention time, peptide mapping, amino acid analysis or NMR) and notes that controlling diastereomers may require specific methods⁵. In the Ghent study, the main compound of one sample had a different structure from the one ordered¹, an error that purity alone would not have revealed.

Purity is not content: counter-ions, salts and water

This is the point that causes the most confusion. A lyophilized powder with 99% HPLC purity is not 99% peptide by weight: it also contains counter-ions, water and sometimes solvents. The best-documented case is a candidate certified reference material of angiotensin II from the National Research Council Canada: the trifluoroacetate counter-ion accounted for nearly 25% of the mass and the final assigned content was 691 ± 9 mg/g⁷. In other words, about 69% of the powder was peptide. That value is the net peptide content; the EMA states that a chromatographic assay is expressed on the anhydrous, counter-ion-free substance⁵.

Why are there counter-ions? Peptides are cleaved from the resin and purified with trifluoroacetic acid, so cationic peptides are obtained as trifluoroacetate salts; most approved peptides, by contrast, are marketed as acetates⁸. The difference is not only about weight. In osteoblast cultures, TFA at 10⁻⁸–10⁻⁷ M reduced cell numbers, and the TFA salts of amylin and calcitonin masked a proliferative effect or suggested a nonexistent antiproliferative one; the authors recommended converting peptides to the hydrochloride before assessing their biological effects⁹. The counter-ion can be quantified by ion chromatography, and lyophilization with hydrochloric or acetic acid reduces TFA¹⁰. Salt forms also matter for GLP-1s, as explained in Compounded GLP-1s: salts, impurities and antibodies.

Water is the third component. Residual moisture is a critical quality attribute of lyophilized products, and Karl Fischer titration is the usual method for measuring it, with near-infrared spectroscopy as a nondestructive alternative¹¹; the EMA includes it in the specification and expects it in stability studies when the powder is hygroscopic⁵. The practical consequence for the researcher is knowing whether the milligrams on the label refer to powder or to net peptide: for example, 10 mg of powder with 70% peptide content provide 7 mg of peptide. How to store that powder afterwards is covered in Peptide stability.

Endotoxins, sterility and residual solvents

The endotoxin test and the sterility test measure different things. Endotoxins are determined with the Limulus amebocyte lysate (LAL) assay, which the FDA has accepted for more than 30 years in place of the rabbit pyrogen test and which is described in USP chapter <85>¹². The result should be reported with a number and units, for example in EU/mg, alongside its limit. A Hungarian study of semaglutide sold online shows why both tests matter: the three lyophilized vials purchased had no viable microorganisms at the time of testing, yet all contained endotoxins, between 2.16 and 8.95 EU/mg¹³.

Residual solvents are governed by the ICH Q3C guideline, which places them in three classes: solvents to be avoided (class 1), solvents to be limited (class 2) and solvents with low toxic potential (class 3)¹⁴. For class 2 it sets specific limits: 410 ppm for acetonitrile, 600 ppm for dichloromethane and 880 ppm for N,N-dimethylformamide; for class 3, up to 0.5% is considered acceptable without further justification¹⁴.

Batch, dates and laboratory

A certificate is valid only for the batch it describes. FDA good manufacturing practice rules require laboratory records to identify the sample with its lot number, the date it was taken and received, the method used, all data obtained (including instrument graphs, charts and spectra identified with the lot), the calculations, the initials or signature of whoever performed each test with the date, and review by a second person¹⁵. The EMA expects material traceability from synthesis through to the final active substance⁵.

In-house or third-party testing? The Ghent group found that although all 98 peptides came with a certificate stating ≥95% purity, only 43 met it, and recommended that researchers maintain their own quality control¹. For a laboratory, the strongest option is to combine the supplier's certificate with independent verification, at least of identity by mass, for the key batches of each research protocol.

Red flags and the limits of a certificate

• No mass spectrum or observed mass: purity alone does not confirm identity¹⁵.
• Identical chromatograms (same retention times, same areas, same baseline) across different batches: each batch generates its own data¹⁵.
• No lot number, no test dates or no signatures¹⁵.
• Purity without peptide content, counter-ion or water when the material is weighed to prepare solutions⁵⁷.
• Endotoxins reported as "conforms" with no number or units¹²¹³.
• A 99.9% figure with no integrated chromatogram or method, or a laboratory that cannot be identified¹¹⁵.

Even a flawless certificate has limits. It is a snapshot of the batch on the test date; that is why the EMA requires separate stability studies⁵. Market studies also show how far the declared can be from the measured: in three vials of semaglutide labeled at 99%, measured purity was 7.7–14.4%¹³, a finding reviewed in What is really in the vial?. A well-read certificate does not remove those risks; it lets you detect them in time.