How to read a peptide certificate of analysis

What a peptide certificate of analysis is actually claiming

A certificate is a record of measurement on one batch at one moment. It is not a statement about the seller, the compound in general, or the vial you were sent unless the lot numbers match. That match is the first thing to check and the most commonly missing.

A peptide certificate of analysis answers four questions, and most answer only three. Identity: is this the molecule named? Purity: how much of what the detector saw was that molecule? Content: how much peptide is actually in the vial? Contaminants: what else came along from the synthesis and the fill? The fourth is the one routinely absent, and its absence is not a fault so long as the document does not imply a test that was never run.

Everything else on the page describes how the measurement was made. Read the page in that order: identity, then purity, then content, then provenance. A peptide certificate of analysis becomes quick to judge once it is read that way.

It is worth being blunt about what the document is for. A certificate is a statement about material, not about use. It records what a named laboratory measured in a sample drawn from one batch, on one date, with named instruments. Nothing on it authorises anything. A research-use certificate in particular is not evidence of suitability for any purpose beyond the bench, and a seller that presents it as though it were has told you something about the seller. That limitation is exactly why the analytical fields matter: they are all a certificate can honestly carry, so they should all be there.

How to read a peptide certificate of analysis: every field, and what its absence means

A complete peptide certificate of analysis is short: usually one page, occasionally two once the chromatogram is attached. If any of the following is missing, the gap is the finding.

The sequence line is the field most often skimmed and the easiest to check. It is written in one-letter amino acid code, and the modifications matter as much as the letters: an Ac- prefix marks an acetylated N-terminus, a trailing -NH2 marks an amidated C-terminus, and non-standard residues appear by name, as in ipamorelin's Aib-His-D-2-Nal-D-Phe-Lys-NH2 or tesamorelin's trans-3-hexenoyl-GRF(1-44)-NH2. Two peptides with the same letters and different terminal chemistry are different molecules with different masses. If the sequence on the certificate does not match the sequence on the product page character for character, nothing further down the page is worth reading.

Read the table below by its last column first. For most of these fields, the consequence of an omission is larger than the value of the field when it is present.

HPLC purity on a peptide certificate of analysis, and what area percent leaves out

High-performance liquid chromatography separates the sample and reports each component as a percentage of the total area under the trace. When a certificate says 99%, it almost always means 99% area percent: of everything the detector saw, 99% was the target peptide.

That figure excludes what the detector does not see. Residual water, salts and counter-ions such as trifluoroacetate are real mass in the vial and are not part of an area-percent number. Net peptide content, how much of the vial's mass is peptide, is a different measurement, and it belongs to peptide purity testing rather than to the chromatogram. A peptide certificate of analysis that reports both figures is a more confident document than one that reports only the larger.

The chromatogram beneath the figure is worth a moment of its own, because the number was derived from it. Time runs along the horizontal axis and detector response up the vertical; the target peptide appears as one dominant peak at a characteristic retention time, and the integration table beside the trace lists every peak the software counted with its own area share. What you are looking for is a peak that returns cleanly to the baseline on both sides, a stated retention time rather than an unlabelled bump, and a small number of minor peaks that are actually resolved from the main one. A shoulder on the leading or trailing edge is co-elution: two components the column did not separate, counted by the software as one. That is how an area-percent figure can be technically correct and still flattering.

Do the arithmetic once and the distinction stops being abstract. The percentage is a ratio between peaks the detector saw; the content figure is a mass in milligrams. Neither can be worked out from the other, they do not average, and the larger one does not stand in for the smaller.

Certificates differ in how they express the second figure. Some give it as a percentage of vial mass. The laboratory that tests most of our lots reports it instead as a measured mass per vial, which is more directly checkable and occasionally surprising. Lot 2FRBRG of BPC-157 was reported at 99.51% by HPLC with 10.68 mg of peptide in a vial labelled 10 mg, because vials are deliberately overfilled to survive the fill tolerance. A content figure above the label is normal, and one a few per cent below it sits within ordinary fill tolerance. A figure well below the label, with no explanation beside it, is worth a question.

Identity: mass spectrometry, molecular weight and the name on the label

Purity says how much of one thing there is. Identity says whether it is the right thing. Mass spectrometry answers that by measuring the molecular mass and comparing it against the theoretical mass for the sequence.

A certificate showing observed and theoretical mass within a small tolerance has demonstrated identity. One showing purity alone has told you the sample was clean, without saying what it was clean of. Where the two masses are printed side by side, convention for electrospray ionisation on peptides puts acceptable agreement within roughly 0.1 to 0.5 daltons, and a certificate that prints both numbers has shown its working.

Not every laboratory reports it that way, and the alternative is not a weaker test. Freedom Diagnostics, which issues most of the certificates in our library, confirms identity by LC-MS against a certified reference standard and prints a confirmation rather than a pair of masses. That is a genuine identity determination made against a known material; it is simply expressed as a conclusion rather than as the two numbers behind it. What matters is that identity was determined by an instrument at all.

Three constants pin a molecule down where a trade name will not: molecular weight, molecular formula and CAS number. They belong on a certificate and they are trivially checkable against a reference. TB-500 is the clearest case of why. The name is used in this market for two different molecules: the acetylated seven-residue fragment Ac-LKKTETQ at 889.0 g/mol, which is what TB-500 is here, and the full-length 43-residue thymosin β4 protein, several times heavier at around 4,963 g/mol. Both are legitimately supplied. They are not the same material, and the field that separates them at a glance is the molecular weight. The question what is TB-500 is worth settling before comparing two suppliers' purity figures at all.

Where the impurities in the other one per cent come from

Peptides at this scale are built by solid-phase peptide synthesis: the chain is assembled one residue at a time on a resin, each coupling driven close to completion but never quite to it. The shortfall at each step is where the impurity profile comes from, and it has a shape worth knowing.

A coupling that fails and is then capped leaves a deletion sequence: the full chain, one residue short. A chain that stops early leaves a truncated fragment. Side reactions on protecting groups leave adducts. All three are structurally close relatives of the target, which is precisely why they are difficult: a deletion sequence differing by one amino acid can elute very close to the target on a C18 column, and on a certificate reporting purity alone it shows up as nothing more alarming than a slightly smaller number. Mass spectrometry is what separates clean from correct.

The rest of the non-peptide mass in the vial is not a synthesis error at all. Reversed-phase purification runs in trifluoroacetic acid, so the peptide leaves the column as a TFA salt unless it is deliberately exchanged to acetate or hydrochloride. The counter-ion is real mass, it is invisible to an area-percent figure, and the salt form is a named field on a well-made certificate rather than something to be inferred. Lyophilised peptides are also hygroscopic, and residual moisture is counted in the vial's mass the moment the seal is broken in a humid room.

Between them, salt form and residual moisture account for most of the distance between a 99% purity figure and the mass of peptide a vial actually holds, which is why a fill is set above the label strength. Neither is a defect. Both are reasons the two numbers were never going to be the same number.

What a standard peptide certificate of analysis does not test

Purity, identity and content are the three measurements almost every research-use certificate carries. Contamination is a fourth category, and it is commissioned separately, charged separately and reported separately. A certificate that says nothing about endotoxin has not failed an endotoxin test. It has not had one, and the distinction is the whole point.

The LAL assay (Limulus amebocyte lysate) detects bacterial endotoxin, the heat-stable cell-wall fragments left behind by Gram-negative bacteria, which survive processes that destroy the bacteria themselves. Results come back either as a figure in endotoxin units per millilitre or as a pass against a stated limit. The lots in our library that carry this test report it as a pass in duplicate against a limit of 0.05 EU/mL or lower, and two cite USP General Chapter <85> as the method.

Sterility is a different question again, and a slower one. The compendial method in USP General Chapter <71> incubates a sample in growth media for 14 days and looks for growth. Fourteen days is why it is rarely run on a research lot released in a week, and why a claim of sterility unsupported by a sterility test is worth nothing on either side of the argument.

Heavy metals are screened by ICP-MS (inductively coupled plasma mass spectrometry), which reports lead, cadmium, arsenic and mercury at parts per billion. Residual solvents are a gas-chromatography question: acetonitrile and DMF from the synthesis and purification, measured in parts per million against the limits set out in ICH Q3C. Neither appears on a standard certificate for this class of material.

It is worth stating plainly what no certificate of analysis establishes, however complete it looks. It does not establish sterility unless a sterility test is printed on it. It does not establish that the impurities it counted are of any particular character: area percent counts them, it does not name them. It says nothing about what happened to the vial between the laboratory bench and yours, which is why how to store peptides is a separate subject. And it is not a manufacturing record: it reports on a sample, not on the controls under which the batch was made.

Purity bands, and what a band does not settle

The market sorts HPLC figures into bands, and the bands are genuinely informative about one thing: the size of the impurity fraction. Moving from 95% to 99% is a fivefold cut in everything that is not the target peptide. Moving from 99% to 99.5% halves it again.

What a band cannot do is rank two certificates from different laboratories, because area percent depends on the method: the gradient, the column, the detection wavelength. An impurity that does not absorb where the detector is looking is not in the denominator. Two 99% figures are comparable only when the methods behind them are. That is one more reason the method line matters more than the percentage it qualifies.

Above roughly 99%, the honest answer is that the differences stop being the interesting question. Who measured this lot, whether the report names them, and whether the number belongs to the vial in front of you are all larger variables than the third decimal place.

Who ran the test named on a peptide certificate of analysis, and under what accreditation

A certificate produced in-house by the seller is an assertion. One produced by an independent laboratory is evidence, and it carries that laboratory's name, the date and the method.

Accreditation is a readable field, and the three abbreviations that appear on these documents do not mean the same thing. ISO/IEC 17025 is the international standard for the competence of testing laboratories: it accredits methods and the quality system behind them, scope by scope, and it is audited by an external body. GLP (Good Laboratory Practice) governs how studies are conducted and recorded rather than the accuracy of any single assay. cGMP applies to manufacture, not to testing at all, so a testing laboratory's credentials say nothing about the conditions the material was made under. An accredited laboratory named on a certificate is a stronger signal than an unaccredited one; an unnamed laboratory is not a signal at all.

Names are worth knowing because names can be checked. The certificates in our library are issued by Freedom Diagnostics and, for two lots, by pH Solutions. Elsewhere in this market you will see Janoshik Analytical, Eurofins BioPharma Product Testing and SGS Life Sciences on documents. An unfamiliar laboratory is not a red flag by itself. An anonymous one is.

A research peptide supplier should publish its certificates before purchase rather than produce them on request. Ours are: every lot's report is on the site, matched to the number printed on the vial, and can be looked up directly.

Verifying a certificate against the laboratory's own record

A PDF can be edited. That is the honest starting point, and it is why the strongest check does not involve reading the document more carefully. It involves finding a copy of the same information held by someone other than the seller.

Most contract laboratories assign each sample an accession number when it is booked in, and many print a search code alongside it that resolves in the laboratory's own lookup. Every Freedom Diagnostics certificate in our library carries both. Lot PSNN1N of GLP-3 RT, for instance, shows accession 2607030208 and search code Auth2607030208; lot GHK100-052026-4 of GHK-Cu shows 2605120139 and Whol2605120139. The prefixes vary across our reports according to how the sample was submitted, which is the kind of internal consistency a fabricated document tends not to have.

Two further checks cost nothing. Read the test date: the working convention in this market treats a certificate more than about twelve months old as stale, not because the paper expires but because a lot that has outlived its documentation usually means the seller has stopped testing. And read the lot number off the vial itself, then find that number on the seller's site rather than on the PDF the seller emailed you.

Certificates that should not be trusted

The commonest problem is not forgery but vagueness. A report with no lot number attaches to nothing. A report with no laboratory named attaches to nobody. A chromatogram with no axis labels, no scale and no retention times is a picture of a peak rather than a measurement of one.

Then there are the reused documents: one certificate cited across several products, or a date that never moves while lots keep shipping. Synthesis is a per-batch process, and a failure in one batch tells you nothing about the next, which is exactly why a peptide certificate of analysis has to be per batch too.

One widely repeated tell deserves qualifying, because it is repeated more confidently than it deserves. Purity reported as a flat 99% is often said to be suspicious on the grounds that real instruments produce decimals. It is a fair heuristic and it is not a rule. Two lots in our own library, GLP-3 RT lots Y4AWYQ and 2NPVBM, are reported at a flat 99% simply because that is how the issuing laboratory, pH Solutions, prints the figure. The rest of the library, measured by Freedom Diagnostics, runs from 99.247% to 99.984% with two or three decimals. A round number is a reason to read the method and the laboratory name, not a verdict on either.

The check that resolves nearly all of this takes ten seconds: read the lot number off the vial, and find that lot number on the seller's site.

Fifty-one certificates, and what a real library looks like

Reading about a peptide certificate of analysis is less useful than reading one. The library on this site holds 51 signed reports covering 11 compounds, each tied to a released lot number, and the figures in it behave the way measurements behave rather than the way marketing copy does.

The purity figures are not all 99.9%. Across the Freedom Diagnostics reports they run from 99.247% on a BPC-157 lot released in April to 99.984% on an ipamorelin lot, which is the ordinary spread of a batch process rather than a quality problem. The measured content moves around more than the purity does, because fill overage is set per run: a 50 mg GHK-Cu lot came back at 65.80 mg of peptide, and a 100 mg lot of the same compound at 94.85 mg.

Blends are where an itemised certificate earns the most. A single combined purity figure on a four-component vial answers almost nothing; what matters is the measured mass of each component. KLOW lot XX1EZC reports GHK-Cu at 58.65 mg, KPV at 8.80 mg, TB-500 at 10.16 mg and BPC-157 at 9.42 mg against an 80 mg label, and that itemisation is the only way to know what the ratio in a given lot actually is. GLOW vs KLOW is a question their certificates answer more precisely than their names do.

Getting a certificate when a seller will not publish one

Half the questions people ask around this subject are about obtaining a certificate rather than reading one, which is a fair reflection of how much of this market still emails documents on request.

Ask first, and ask for the lot. A request that names the number printed on the vial is much harder to answer with a generic document than a request for "the COA". If what comes back carries a different lot number, or no lot number, that is an answer in itself.

If a seller cannot or will not produce one, material can be tested independently. The analytical laboratories serving this market accept samples from anyone: a few milligrams in the original sealed vial, a submission form naming the compound and the assays wanted, and a report back in days rather than weeks. Pricing is quoted per sample and per assay. HPLC purity and mass-spectrometry identity are the cheapest pair on the list, and each contaminant assay is charged on top of them. Published rates move often enough that a figure quoted in an article is a figure that will be wrong by the time it is read, so ask the laboratory directly rather than trusting a number in a guide.

What commissioning your own test does not do is retrofit provenance. It tells you what is in the vial you sent, which is worth knowing, and nothing about the next vial from the same seller. A seller that publishes per lot has already answered the question for every buyer at once, and whether they do is one of the first things worth checking about a research peptide supplier.