How to Read a Peptide COA

How to Read a Peptide Certificate of Analysis

Most Certificates of Analysis in the US research peptide market get read for one number. A buyer scans for the purity figure, sees 99-point-something, and moves on. That number is the least informative thing on a well-built COA, and on a badly built one it is actively misleading.

Stated as bluntly as the analytical chemistry allows: an HPLC purity percentage is a statistic about a chromatogram, not a statement about the contents of a vial. A material can be 99% pure by peak area and still be substantially something other than the peptide by mass.

That is not hypothetical. It has been documented in the peer-reviewed literature, on commercially sourced material, by a team including authors from the United States Pharmacopeial Convention.

The finding that should change how you read the purity line

Choules and colleagues, publishing in the Journal of Pharmaceutical and Biomedical Analysis in 2020 (PMID 31671336), applied quantitative proton NMR to two custom synthetic peptides purchased commercially. They found undeclared mannitol at 20% and 43% w/w.

Two features of that result matter. First, mannitol is UV-transparent and highly polar — it produces no signal in the detection mode that generates a standard HPLC-UV purity number, and it does not retain on a reversed-phase column the way the peptide does. A UV-based purity assay is, in the authors’ word, “blind” to it. Second, quantitative NMR detected and quantified it readily, even on a 60 MHz benchtop instrument.

So the purity percentage on those materials was not false. It was answering a narrower question than the reader assumed. “What fraction of the UV-absorbing, column-retained material eluting in this run is the target peak?” is not the same question as “what fraction of the powder in this vial is the peptide?”

Everything below follows from keeping those two questions separate.

The header block

Start at the top, and start with the lot.

A COA is a record of tests performed on a specific batch of material. If the document does not carry a lot or batch number that matches the vial that shipped, it is a marketing asset with a laboratory letterhead. Generic COAs — one PDF served for every unit of a product line regardless of production run — are common in this market and should be treated as the absence of a COA rather than a weak version of one.

The header should also carry the compound name and sequence or CAS identifier, the manufacture or test date, the testing laboratory, and the methods used. Test date is quietly important: it fixes when the material was characterized, and everything afterwards — storage, handling, transit — is outside the document’s scope.

Purity: read the method, not just the number

Under the purity line there should be a method. In practice that is usually reversed-phase HPLC with UV detection, typically at 214 or 220 nm, where the peptide bond itself absorbs.

Three questions are worth asking of any reported purity figure:

What is the detection mode? UV at 214 nm sees the amide backbone, which is why it is used. It does not see excipients, sugars, salts, or most inorganic material. A number generated this way describes peptide-related composition, not total composition.

Is the chromatogram shown? A purity value without the trace is an assertion. The trace shows peak shape, baseline, whether the main peak is resolved from its nearest neighbors, and whether the run was long enough for late-eluting material to come off the column. General expectations for system suitability and chromatographic performance are set out in USP General Chapter <621>.

Purity by area, or content by mass? These are separate line items on a complete COA and they are not interchangeable.

Net peptide content is the number most COAs omit

Lyophilized peptide powder is peptide plus counterion plus bound water. It is not unusual for the peptide fraction of the total mass to sit well below 100%, which means the mass on the label and the mass of peptide in the vial are different quantities.

Net peptide content — usually determined by amino acid analysis or by quantitative NMR against a certified standard — is what converts a labeled milligram figure into an actual molar quantity. A study that calculates concentration from the label rather than from net content is running at a different concentration than its methods section states. For any work where concentration is the independent variable, this is the line item that matters most, and it is the one most often missing.

Counterion content is not a footnote

Peptides purified by reversed-phase HPLC generally arrive as trifluoroacetate salts, because trifluoroacetic acid is the standard ion-pairing modifier in the separation. Residual TFA is therefore expected. It is also biologically active in ways that contaminate results.

Cornish and colleagues demonstrated this in American Journal of Physiology in 1999 (PMID 10567002). TFA at 10⁻⁸ to 10⁻⁷ M reduced cell numbers and thymidine incorporation in fetal rat osteoblast cultures within 24 hours, with the same effect in articular chondrocytes and neonatal mouse calvariae — so it was not specific to one cell type or species. When TFA and hydrochloride salts of amylin, amylin-(1-8) and calcitonin were compared in osteoblasts, proliferation was consistently lower with the TFA salts, to the point of producing failure to detect a real proliferative effect and, in some comparisons, wrongly attributing an antiproliferative one. The authors concluded that this is likely relevant to any purified peptide used above 10⁻⁹ M in any cell or tissue type.

A COA that reports residual TFA lets a research group decide whether salt exchange is necessary before the assay. A COA that omits it leaves an unmeasured variable in every cell-based experiment run on that lot.

Identity: purity without identity means nothing

A 99% pure sample of the wrong molecule is 99% pure.

Identity is normally established by mass spectrometry — ESI-MS or MALDI-TOF — reported as observed mass against theoretical mass, ideally with the charge states or isotope pattern shown. Sequence-level confirmation by tandem MS is stronger still and is increasingly available on research-grade material.

The reason to insist on identity data is that the most common synthesis-derived impurities are structurally near-identical to the target. Wu and colleagues, writing in Rapid Communications in Mass Spectrometry in 2026 (PMID 42312586), compared impurity profiles in teriparatide reference materials from chemical synthesis and from recombinant DNA origin using LC-Orbitrap high-resolution MS. One isomer and three oxidation impurities were common to all materials, but six amino acid deletion impurities were found specifically in the chemically synthesized material, and eight impurities were newly characterized beyond previous work. Deletion sequences and oxidation products can co-elute closely with the parent peak. Mass, not retention time, is what separates them.

The impurity table

A strong COA names its impurities rather than reporting only a total. Named, quantified, individually reported impurities tell a research group what is present. A single “total impurities: 0.6%” line tells them only that the supplier’s integration software summed something.

Thresholds deserve a note of caution. The ICH Q3A and Q3B impurity frameworks were built around small-molecule drug substances and products, and synthetic peptides sit awkwardly against them. FDA addresses the peptide case separately in its guidance ANDAs for Certain Highly Purified Synthetic Peptide Drug Products That Refer to Listed Drugs of rDNA Origin, which is the appropriate reference point for how peptide-related impurities are expected to be identified, characterized and justified. Research-grade material is not regulated to that standard — but the standard is the right yardstick for judging how complete a COA is.

Water, solvents, and elemental content

Three further line items round out a thorough document:

Water content, by Karl Fischer titration (USP <921>). Feeds directly into the mass accounting above, and is a stability variable in its own right.
Residual solvents, typically by gas chromatography with headspace sampling, classified per ICH Q3C.
Elemental impurities, by ICP-MS, per ICH Q3D. Relevant wherever metal catalysis or metal-contact equipment features in the synthesis or workup.

Microbiological attributes

Endotoxin, bioburden and sterility are separate tests answering separate questions, and research-grade lyophilized peptide is generally not sterile. That is a design constraint to plan around, not a defect.

Where endotoxin is tested, the compendial method is USP <85>. Thota and colleagues, in the PDA Journal of Pharmaceutical Science and Technology in 2026 (PMID 41844320), compared traditional Limulus amebocyte lysate against recombinant cascade reagent across several categories of pharmaceutical sample and found the two equivalent, with no interference observed in certain samples where rCR was used. The US, European and Japanese pharmacopeias have assessed this equivalency, and rCR has been available as an animal-free alternative since 2024. A COA specifying either method reports against a recognized standard.

What a COA cannot tell you

It is a release document. It describes the material at the moment it was tested, at the facility that tested it. It says nothing about the thermal history of the shipment, the number of freeze-thaw cycles the vial has seen, or how long it sat on a loading dock.

This is the practical argument for domestic sourcing rather than a marketing one. Material crossing an international border sits in an uncontrolled thermal environment for an unpredictable interval, and the COA issued at release cannot account for any of it. Maple Research Labs ships same-day from our Santa Barbara, California facility on orders placed before cutoff, which compresses that window to a known and short one.

Every lot we stock carries a batch-specific Certificate of Analysis with third-party HPLC verification through an independent laboratory — the BPC-157 10MG batch COA, for example, is issued against a named report and lot rather than served generically across the product line. If a supplier cannot produce the COA for the specific lot that will ship, that is the answer to the question.

A reading checklist

Lot number on the COA matches the vial. No match, no document.
Purity method named, chromatogram shown, detection wavelength stated.
Net peptide content reported separately from purity by area.
Counterion and residual TFA quantified.
Identity confirmed by mass spectrometry, observed against theoretical.
Impurities named and individually quantified, not summed.
Water content by Karl Fischer.
Residual solvents and elemental impurities where the route warrants them.
Testing laboratory identified, and independent of the seller.
Test date present, and recent relative to the shipment.

Compliance notice. All compounds and analytical material referenced on this page are supplied for laboratory research use only. All applications discussed are in vitro systems and animal models. Nothing on this page is guidance for administration to people. These materials are not for human use, not for veterinary use, and not for diagnostic use.

References

Choules MP, Bisson J, Simmler C, McAlpine JB, Giancaspro G, Bzhelyansky A, Niemitz M, Pauli GF. NMR reveals an undeclared constituent in custom synthetic peptides. J Pharm Biomed Anal. 2020;178:112915. PMID 31671336.
Cornish J, Callon KE, Lin CQ, Xiao CL, Mulvey TB, Cooper GJ, Reid IR. Trifluoroacetate, a contaminant in purified proteins, inhibits proliferation of osteoblasts and chondrocytes. Am J Physiol. 1999;277(5):E779-83. PMID 10567002.
Wu P, Cheng Y, Li D, Tan H, Li J, Zhang D. Comparison of structurally related impurity profiles in teriparatide from synthetic and recombinant DNA origin using liquid chromatography-high resolution mass spectrometry. Rapid Commun Mass Spectrom. 2026;40(17):e70125. PMID 42312586.
Thota P, Kumar P, Teotia AK, et al. Study of LAL and recombinant cascade reagents methods for bacterial endotoxin testing in pharmaceutical products. PDA J Pharm Sci Technol. 2026;80(3):288-297. PMID 41844320.
US Food and Drug Administration. ANDAs for Certain Highly Purified Synthetic Peptide Drug Products That Refer to Listed Drugs of rDNA Origin — Guidance for Industry.
United States Pharmacopeia. General Chapters <621> Chromatography, <85> Bacterial Endotoxins Test, <71> Sterility Tests, <921> Water Determination, <1225> Validation of Compendial Procedures.
International Council for Harmonisation. Q2(R2) Validation of Analytical Procedures; Q3C Residual Solvents; Q3D Elemental Impurities; Q6A Specifications.