Peptide purity testing: the methods, the laboratories and what they cost

The methods peptide purity testing usually involves

Each test below answers a different question, and none substitutes for another. A certificate of analysis is essentially a list of which peptide purity testing methods were run, so the first thing to read on a report is not the headline percentage but the list of methods above it.

Four of them make up almost every research-grade panel: chromatography for purity, mass spectrometry for identity, amino acid analysis for how much peptide is actually in the vial, and Karl Fischer analysis for how much water is left in the cake. Everything else in the table is a second tier, run on a first production lot or when a particular application calls for it.

The HPLC method line behind a peptide purity testing result

A purity percentage on its own is not a measurement. It is the output of a peptide purity testing method, and the method is normally printed on the certificate in one dense line that most readers skip. It is worth two minutes.

A standard analytical peptide method uses a C18 reverse-phase column with 0.05 to 0.1 per cent trifluoroacetic acid in both mobile phases as an ion-pairing agent, a gradient running from roughly 5 to 65 per cent acetonitrile over about twenty minutes, at a flow rate near 1 mL/min. Detection is by ultraviolet absorbance, usually at 210 to 220 nm, where the amide bonds of the backbone absorb, so every peptide is visible whatever its sequence. Some methods detect at 280 nm instead, which sees only tryptophan and tyrosine and is therefore blind to a peptide containing neither.

Each of those parameters moves the number. A shallower gradient resolves impurities that a steep one buries under the main peak; a different column chemistry changes which species co-elute; 280 nm detection can miss an impurity entirely. Where the operator sets the integration baseline and how small peaks are counted moves it again, usually by a tenth or two.

The practical consequence is that two competent laboratories can report figures one to three per cent apart on identical material and both be right. That is not a scandal, and it is not a reason to distrust chromatography. It is the reason a purity figure quoted without its method is a claim rather than a result, and why comparing two listings on their headline percentages alone is close to meaningless.

What mass spectrometry adds, and the signatures it sees

Chromatography tells you how much. Mass spectrometry tells you what. The instrument ionises the sample and measures mass-to-charge ratio, and the result is compared against the mass calculated from the sequence and formula.

On a nominal-mass instrument, an observed mass within about 1 Da of theoretical, or within 0.1 per cent for a long chain, is treated as a match. On a high-resolution instrument such as an Orbitrap or a Q-TOF the tolerance falls below 5 parts per million, and MS/MS fragmentation can walk the sequence residue by residue rather than only weighing the whole molecule. That distinction matters for a long peptide, where several different sequences can share the same nominal mass.

The more useful thing mass spectrometry does is see impurities the UV detector cannot. Oxidation adds sixteen daltons at methionine, tryptophan or cysteine, and the oxidised species often elutes so close to the parent that the chromatogram counts it inside the main peak. A certificate reporting 99.5 per cent by HPLC with no mass-spectrometry data has not ruled that out.

The impurities a purity percentage is counting

Solid-phase synthesis builds a chain one residue at a time, and every step has a yield below one hundred per cent. What survives purification is a population, not a single molecule, and the impurity profile on a certificate is the shape of that population.

Deletion sequences are chains missing one residue, from a coupling step that did not go to completion. Truncated peptides are chains that stopped early. Both are structurally close to the target and can be hard to resolve chromatographically, which is why an impurity profile that names related substances is worth more than one that reports a single number.

Residual protecting groups are fragments of the synthesis chemistry, tert-butyl, trityl, Pbf and their relatives, left on a side chain when the final cleavage was incomplete. Dimers and higher aggregates form where a free cysteine oxidises to a disulfide or where a hydrophobic sequence associates with itself; these show as peaks at multiples of the parent mass.

Then there is the material that is not peptide at all. Trifluoroacetate from the purification mobile phase stays with the product as a counter-ion and is invisible to a UV trace. Residual acetonitrile and other solvents remain at trace levels. Water absorbed into the lyophilised cake is mass in the vial that is not peptide. None of that appears in an area-percent purity figure, which is exactly why net peptide content is measured separately.

Degradation is the fourth category, and the only one that happens after the certificate was signed. Hydrolysis, oxidation and deamidation all proceed faster in the presence of moisture and heat, so a lot that measured 99.6 per cent on the day it was tested is not making a claim about what it measures a year later in a warm cupboard.

Purity bands: what 95, 98 and 99 per cent mean

There is no regulator setting grades for research-grade peptides, so the bands below are conventions rather than standards. They are useful anyway, because the market has converged on them and because the difference between them is a real difference in what is left in the vial.

The arithmetic is worth saying out loud: a 98 per cent result means two per cent of the detected material was something else. On a 10mg vial that is 200 micrograms of related substances, and the certificate rarely says what they were. Whether that matters depends entirely on the work. For a binding assay with a wide window it is irrelevant, and for anything quantitative it is not.

Purity is not content: a worked example from released lots

The two numbers are routinely confused, and the fastest way to separate them is to read a real certificate. Peptide purity testing reports both, and they are not the same quantity. HPLC purity is a ratio between peaks the detector saw. Net peptide content is a mass: how many milligrams of peptide the vial actually holds.

Take lot 2FRBRG of BPC-157. Freedom Diagnostics reported it on 29 May 2026 at 99.51 per cent by RP-HPLC, identity confirmed by LC-MS, with net peptide content of 10.68mg in a 10mg vial, under accession 2605270306. Both figures are on the same page of the same document, and neither one could be worked out from the other.

Those two numbers move independently. Lot PQ7MWD of TB-500 reported the same day at 99.58 per cent, a shade purer, with 12.16mg of peptide in a 10mg vial. A higher purity percentage with a higher content figure is an overfill, not a contradiction; purity describes the material, content describes the fill.

Blends are where the distinction stops being academic. A multi-component product cannot honestly carry one purity figure and one content figure, because there are several analytes in the cake. The certificate for GLOW lot GLOW-062026-4 reports content per component, GHK-Cu 59.92mg, Thymosin β4 11.71mg and BPC-157 10.86mg, and a listing that flattens that into a single number has thrown away the only part a reader needs.

Tests that appear on some certificates and not others

Endotoxin testing, usually by LAL assay and specified in the United States Pharmacopeia as chapter <85>, measures bacterial endotoxin in endotoxin units. It is a microbiological measurement rather than a chemical one, and it is commissioned where a downstream application is sensitive to it: cell culture work, for example, where endotoxin confounds results at very low concentrations. Where a certificate reports it, the useful detail is the limit tested against and whether both replicates passed.

Two neighbouring chapters turn up on fuller panels. USP <61> is bioburden: a plate count of viable organisms recovered from the sample, reported in colony-forming units. USP <71> is sterility, a fourteen-day incubation in two growth media reported as growth or no growth. Both describe what was recovered from the material tested. Neither is a purity measurement, and neither is run on routine research-grade lots.

Counter-ion content quantifies trifluoroacetate or acetate directly rather than leaving it implied, by ion chromatography or fluorine NMR. Residual solvent testing by headspace gas chromatography looks for what is left from synthesis and purification. Elemental impurity screening by ICP-MS reports metals in parts per million or billion. All three are characterisation tests, more common on a first production lot than on a routine release.

Appearance and solubility are recorded observationally: colour and form of the cake, and whether it goes cleanly into the stated solvent. Trivial to perform, and genuinely useful, because an off-colour cake or a solution that will not clear is a signal ahead of any instrument.

Independent laboratories that run peptide purity testing

This is the part of the subject most articles leave abstract, and it does not need to be. A working set of contract laboratories that run peptide purity testing publish their prices and their turnaround openly, and the table below is what those pages said when they were read on 19 September 2026. The laboratories are described by type rather than named, because a rate card can be rewritten overnight and a company name welded to a stale figure is worse than no name at all. The shape of the market is the durable part: the tiers, the spread between them, and what each one buys.

Two structural points before the table. First, price scales with scope rather than with quality: a $195 identity-and-purity result from a competent laboratory is the same measurement as the same test inside a $750 panel. Second, a laboratory that publishes its certificates in a searchable public archive has done something the others have not. It has made its own work checkable by people who are not its customers.

Serviam's own lots run through two laboratories, both named on the certificates themselves. Forty-nine of the fifty-one archived certificates were issued by Freedom Diagnostics, and 48 of the 51 carry an accession number and the laboratory's own search code. The remaining two, both GLP-3 RT, came from pH Solutions, which is ISO/IEC 17025 accredited and uses a different report format entirely: fill-weight conformance vial by vial, plus microbials, mycotoxins, residual solvents and elemental impurities that the Freedom certificates do not cover.

The spread in that table is a spread of scope rather than of rigour. The same RP-HPLC injection sits inside the cheapest tier and the dearest one; what the extra spend buys is more tests beside it and a longer calendar to run them in. Read that way, the question worth asking of a seller is never which laboratory is cheapest, but which panel a given lot needed and whether the report for it was published against the lot number.

What peptide purity testing costs, tier by tier

Published entry prices at independent United States laboratories sit between $195 and $250 for identity plus purity on a single sample. Adding a content measurement takes it to roughly $295. A full quality-control panel, meaning chromatography, identity, content, endotoxin, elemental impurities and microbials, runs $600 to $750. Above that the published figures thin out and widen: one laboratory lists a $1,195 tier that adds sterility with its fourteen-day incubation, and the broadest published per-sample range runs from $150 to $1,500 depending on how much scope is bought. The top of that range, not the $1,195 tier, is the highest single-sample figure on any page read.

Volume moves those figures. One laboratory publishes discounts of five per cent at fifteen to twenty-nine samples, ten per cent at thirty to forty-nine and twenty per cent above fifty, which is the economics that makes per-lot release testing realistic for a seller rather than aspirational. Method development, where no method exists for a compound, is charged separately and is the item most likely to surprise someone testing something unusual.

Set against those numbers, a missing certificate is not a cost problem. Routine release testing against an established method is a two-figure to low-three-figure charge spread across a whole lot of vials, so a seller who does not publish one has made a decision about disclosure rather than run into a budget.

Sending a sample for peptide purity testing yourself

Contract analytical laboratories accept samples from businesses and from individuals, and the process is more prosaic than people expect. What varies is the sample requirement, and it is larger than most guides suggest.

One laboratory's published requirement is a 10mg minimum per sample and three vials, rising to four vials where the fill is under 10mg, with lyophilised powder preferred and reconstituted solution accepted. The reason for multiple vials is that several of the tests are destructive and some are run in replicate. Sending a single vial usually means the panel has to be cut down.

On the submission form, the compound name and the sequence are the two fields that matter, because the sequence is what a theoretical mass is calculated from. Without it the laboratory can weigh your molecule but cannot tell you whether the weight is right. Turnaround is quoted from receipt, not from posting: the day the parcel lands is Day 0, and published standard turnarounds run from one to three business days at the fastest to two weeks at the slowest.

Results come back as a PDF, increasingly through a portal with a lookup code on the certificate. Free programmes exist, and at least one laboratory tests free of charge for individuals, but the trade is confidentiality: the result is published in a public dataset with the vendor named. That is a good bargain if the point of the exercise is to inform the market, and a poor one if you were hoping to keep a disappointing figure to yourself.

ISO/IEC 17025, and what accreditation does and does not cover

ISO/IEC 17025 is the international standard for the competence of testing and calibration laboratories. An accredited laboratory has demonstrated to an accreditation body that it runs particular methods correctly: that its instruments are calibrated and traceable, that its staff are qualified for the work, that its results are reproducible, and that it participates in proficiency testing against other laboratories.

The word doing the work in that paragraph is particular. Accreditation is granted against a scope, which is a published document listing the specific methods and matrices the laboratory is accredited for. A laboratory accredited for elemental analysis by ICP-MS is not thereby accredited for peptide chromatography. Asking whether a laboratory is accredited is the easy half of the question; asking whether the test on your certificate falls inside its accredited scope is the half that matters.

Two things accreditation is not. It is not GMP, which governs manufacturing rather than testing, and it is not a guarantee about any individual result. It is a statement about the system that produced it. Nor is it universal in this category: plenty of competent peptide purity testing is done by laboratories that are not accredited, and the honest position is that accreditation raises the floor rather than certifying the ceiling.

Two of Serviam's fifty-one archived certificates come from an ISO/IEC 17025 accredited laboratory, pH Solutions, and their format shows the difference: fill-weight conformance reported vial by vial across the whole run, alongside microbials, mycotoxins, residual solvents and elemental impurities. Purity on those reports is stated as conformance to a 99 per cent specification rather than to two decimal places, which is the document being quoted accurately rather than a weaker lot.

One vial is not a lot, and a certificate has a date

Testing is per lot, not per vial. A lot is one synthesis and one fill, so every vial from it shares the same measurement, and the lot number on the label is what ties a particular vial to a particular report. What does not carry across is one lot's result to the next. Synthesis is a per-batch process and failures are per-batch events, so a new lot needs a new test and no amount of testing on the previous one substitutes.

Even within a lot, a certificate describes the vials that were opened. Fill variation is real, which is why the accredited-format certificates report conformance across every vial in the run rather than a single figure, and why a content result of 10.68mg in a 10mg vial is a measurement of that vial rather than a promise about all of them.

The date on the certificate is the other half of this. Purity is measured on a day, and the material carries on ageing afterwards. Lyophilised powder held at −20°C or colder is stable for a long time; the same powder at room temperature is not. Once material is in solution the clock speeds up sharply. We do not publish a shelf-life figure for reconstituted peptide, because we have not measured one: the 28 days often quoted is USP General Chapter <797>'s limit on a punctured multiple-dose container of preserved diluent, which governs the water rather than the peptide dissolved in it. Repeated freeze-thaw cycles accelerate degradation independently of temperature. None of that is visible on a document signed six months ago.

The practical rule is to read the report date rather than the release date, and to treat a certificate whose date never moves while new stock keeps shipping as the clearest signal on the page. How to store peptides between the laboratory and the bench is as much a part of the purity question as the synthesis was.

Making a result checkable by someone who is not you

A test that only the seller can see is not evidence. The whole value of third-party peptide purity testing rests on someone outside the transaction being able to confirm the document is real, and there are only a few mechanisms that actually do that.

The strongest is an accession number: the laboratory's own internal identifier for the sample, printed on the certificate and searchable in the laboratory's records. Forty-eight of the 51 certificates in our archive carry one. Lot 2FRBRG is accession 2605270306, searchable as Auth2605270306, which means the document can be checked at the laboratory rather than only against the seller's own website. QR codes on certificates do the same job with less typing, and public certificate databases and vendor leaderboards do it at the level of the whole market.

The weakest mechanism, and the commonest, is a PDF hosted by the seller with no identifier on it that points anywhere. It may be perfectly genuine. It is simply not checkable, and a document nobody can check is doing the work of a logo.

What that means in practice is a short list of questions for any certificate: does the lot number match the vial in front of you, is the laboratory named, is there an identifier that resolves at the laboratory, and is the chromatogram a real trace with labelled axes and retention times rather than a picture of a peak. Reading a peptide certificate of analysis line by line takes about a minute, and checking a lot number against the published archive takes less.