Why purity is the cheapest number to flatter

What follows is deliberately mechanical. We describe what the instrument does, in order, and identify at each stage the choice available to the analyst and its direction of effect on the reported figure. The purpose is not to suggest that anybody is manipulating the number. The purpose is to establish that two competent laboratories, both acting entirely properly, can return figures more than two percentage points apart on the same material — and that in a market where 99 is a selling point and 97 is a return, two points is the whole transaction.

Why this measurement and not another

Purity won on economics. A generic reversed-phase gradient occupies an instrument for between twelve and forty minutes, consumes a milligram of sample and a few millilitres of solvent, and produces a figure the same week. A peptide content determination by nitrogen analysis or quantitative amino-acid analysis costs several times as much and takes weeks. An aggregate determination requires a second technique nobody offers. Endotoxin requires a different laboratory. Sterility requires a fortnight and destroys the container.

Given that spread, a market with no regulator and no agreed release specification will settle on the cheapest comparable number, and it did. The trouble is what happens next. Once a market competes on a single metric, effort flows towards the metric. There are entirely legitimate ways to raise a reported purity figure that involve no change whatever to the material: run a shorter gradient, raise the integration threshold, widen the solvent-front exclusion, choose a detection wavelength less sensitive to the impurities present, inject a smaller load.

None of those is fraud. Each is a defensible analytical choice with a published rationale. Collectively they mean that the difference between a 99.4 and a 97.6 on two certificates may be entirely a difference of method, and that a buyer comparing them is comparing procedures rather than powders without knowing it.¹

Gradient slope, expressed properly

A gradient should be quoted as a rate, not as a duration. Twenty-five to forty-five per cent acetonitrile over forty minutes is half a percentage point of organic per minute. The same range in twelve minutes is about one and two-thirds points per minute. That threefold difference in slope is the difference between resolving a deamidated relative from its parent and delivering both as one peak.

The underlying relationship is well established in peptide chromatography: resolution of closely related species improves as gradient slope decreases, up to the point where peak broadening from extended run times starts to give the gain back. Peak capacity — the number of peaks a method can theoretically resolve across its run — rises with shallower gradients and with more efficient columns, and it is the honest single-number summary of what a separation can do. It is never quoted in this trade.

The Journal’s standing request is simply that the gradient be printed. It is three numbers: starting composition, ending composition, time. Nobody regards it as commercially sensitive, every laboratory has it in the method file, and its presence converts a purity figure from an assertion into something comparable with the next certificate. Its absence is the reason two figures from two suppliers cannot be placed side by side, and that absence is a documentary decision rather than a technical constraint.²

Injection load, linearity and the flattened peak

A detector responds linearly to concentration over a defined range and then stops. Overload the column or saturate the detector with too much sample and the main peak flattens at the top, its apex broadens, and its integrated area no longer represents the quantity present. Since the main peak is the numerator and dominates the denominator, distorting it distorts the purity figure — usually downwards, because the flattened peak loses area relative to a properly loaded one.

There is a competing pressure, and it is the reason overloading happens. Small impurities near the reporting threshold need adequate signal-to-noise to be integrated at all, and the way to raise their signal is to inject more sample. A laboratory hunting for 0.05 per cent impurities is tempted towards a load that compromises the main peak. The correct answer in regulated practice is two injections: a small load for the main peak and a larger one for the related-substances profile, with the results combined.

Column overload is a separate phenomenon from detector saturation and produces a characteristic asymmetric fronting peak. Both are visible on the chromatogram to anybody who is shown it, which is one of several reasons the Journal asks for the trace rather than the number. A purity figure calculated from a distorted main peak is arithmetically correct and analytically meaningless, and the only way to know is to look.³

Five things a purity figure cannot tell you

First, how much peptide is in the vial. Counter-ions, residual water, inorganic salts and non-absorbing excipients contribute mass and no chromatographic signal, which is how a preparation can be 99 per cent pure and substantially less than 99 per cent peptide. Purity and content are different quantities and the second is the one that enters any calculation involving a mass.

Second, whether anything is aggregated. Reversed-phase conditions dissociate most non-covalent aggregates before detection, so the monomer is what arrives at the detector. Only a size-based separation reports high molecular weight species.

Third, whether the sequence is correct. Retention-time agreement is consistency; molecular mass is composition; only fragmentation approaches sequence. Fourth, whether an isomeric degradation product is present, since isoaspartate and racemised residues change nothing about mass and may or may not resolve depending on the method. Fifth, anything at all about microbiological quality — bioburden, sterility, endotoxin — which is a separate discipline in a separate laboratory.

Stated as a list it reads like an indictment of the technique, and it is not. Reversed-phase chromatography answers its own question superbly. The list is an indictment of a market that asks it five questions and prints one answer.

The four services, and what they disclose

We put the same set of questions to Janoshik, Medutest, PeptideMeter and VendorInvestigate: which gradient does a standard peptide purity run use, at what wavelength, with what integration threshold, against what reference material, under what system suitability criteria, and is the test within an accredited scope. The purpose was not to rank them but to establish what a buyer can find out by asking.

A good deal, is the answer. Every service that responded was willing to describe its method when asked directly, and none treated any of it as confidential. That finding matters more than the individual answers, because it means the information gap between a certificate and a method disclosure is not protected by commercial sensitivity — it is a matter of report design. What appears on the document is a formatting decision, and formatting decisions are cheap to revise.

Where the services differ is in what reaches the report without being asked. Reports from these laboratories are generally better documented than certificates issued by suppliers, which is one reason a third-party report carries more weight in this market than a supplier’s own. The Journal’s view is that the four services are also the constituency best placed to change the convention: if all four printed gradient, wavelength and threshold as standard, supplier certificates would follow within a year, because buyers would start noticing the difference. The practice literature on contract analytical work makes the same argument from the laboratory’s side: a method transferred without its parameters is a method nobody downstream can reproduce.⁴

What we submitted, and how we designed it

The Journal buys material and has it tested, and the design of those exercises deserves the same disclosure we ask of others. For this piece we bought eight vials from a single lot from one supplier, held them together at two to eight degrees, and submitted them in pairs to four laboratories, asking each for a purity determination and for the chromatogram and method parameters alongside the figure. We did not disclose that the vials were from one lot and we did not disclose that the same material had gone elsewhere.

Separately, we commissioned a single laboratory to run one sample under four deliberately varied conditions: a twelve-minute generic gradient and a forty-minute shallow gradient, each integrated at a threshold of 0.1 per cent and 0.05 per cent. That produced four purity figures from one physical sample and one instrument on one afternoon, which isolates the method effect from every other source of variation.

The limitations are ours to state. One lot from one supplier is not a survey of the market. Single injections carry the variability of single injections. And a deliberately varied method study demonstrates the size of the method effect rather than the practice of any laboratory, since all four conditions were chosen by us. What it establishes is a floor on how much of the spread between two certificates can be method rather than material, and the floor is high.⁵

Documentation practice, named

Across the twenty companies the Journal tracks, every certificate states a purity figure. Rather fewer state the method in any form. A minority name the detection wavelength. Almost none states the gradient as a rate or a programme, and we have seen an integration threshold on a supplier-issued certificate twice.

Practice that deserves naming: CPC states the wavelength and the run time on its standard certificate. SSA lists the three largest impurities with relative retention times, which is the single most informative addition we have seen anybody make. QST and BCH supplied full method parameters on request within a working day. WXT and SWB attach the third-party laboratory report rather than transcribing a figure, which removes a transcription step and with it a class of error. QYB, MKM, HJ, KP, SGN, FGP, ERP and JEEP follow the standard convention of a figure without a method, as do WWB, QSC, GGPeps, GL Biochem, Homopeptide and TFC.

The criticism is of the convention, which the whole market adopted collectively and any member of it could leave unilaterally. No company named here has been shown to us to have misstated a result, and where we have queried a figure against a chromatogram the responses have generally been prompt and technical. What we are describing is a document format that omits the four values needed to compare one number with another, and that omission is not in anybody’s interest, including the sellers’.

Four values that would make purity figures comparable

The Journal’s ask on this subject is four values, all known to whoever generated the number, none of them commercially sensitive, none requiring any additional analysis.

The gradient, as a rate or a programme: starting and ending organic composition and the time between them. The detection wavelength. The integration threshold, or the reporting limit if the laboratory prefers that framing. And the solvent-front exclusion window, since it defines the denominator. Add a fifth if the laboratory is willing: the relative retention times and areas of the three largest impurities, which converts a scalar into a description.

With those values, two certificates become comparable, a buyer can tell whether a difference between suppliers is material or method, and a supplier that has invested in a genuinely better product can demonstrate it — which is the argument we would make to a seller rather than to a buyer. Under the current convention, a company running a forty-minute shallow gradient and reporting 98.2 per cent looks worse than a competitor running twelve minutes and reporting 99.4, and there is no mechanism by which the first can show a buyer why. The absence of method disclosure penalises the more rigorous laboratory, and that, more than anything else in this article, is the reason to fix it.

A note on method and sourcing

The compendial and regulatory material in this piece is taken from the current general chapters on chromatography and on validation of compendial procedures, from the European Pharmacopoeia chapters on liquid chromatography and on chromatographic separation techniques, and from the harmonised guidelines on analytical validation, on impurities and on specifications for biotechnological products, all read in the original. The separation science is drawn from the chromatography literature, with the peptide-specific behaviour cited where it differs from small-molecule practice.

Where the Journal reports a number it obtained, it states the number of vials, the number of laboratories, whether the vials came from one lot, whether the laboratories knew, and what method parameters were disclosed to us. Where we quote a figure from a certificate we state whether the method was disclosed on it. Where a laboratory or a company answered our questions we distinguish an answer from a refusal and a refusal from a non-response.

Nothing in this department is a recommendation to buy, use or avoid anything. The compounds referred to are sold for research use only and are not approved for human use in any jurisdiction. Corrections and disputes go to standards@compoundjournal.com. Readers with certificates or chromatograms they would like read should write to letters@compoundjournal.com; we do not identify the source of anything sent to us, and we do not publish a reader’s name without permission.

Efficiency and selectivity, the inseparable trade-off

Particle size determines efficiency but not selectivity. A column packed with 1.7-micron particles will produce sharper peaks and narrower bandwidths than a 5-micron equivalent, which means better resolution of closely spaced peaks, but both columns separate according to hydrophobicity and both will fail to resolve species that do not differ sufficiently in that property. Resolution—the separation of two peaks, measured by their distance relative to their width—is the joint product of efficiency and selectivity, and improvement in one does not compensate for inadequacy in the other.⁶

A peptide method development sequence therefore cannot stop at efficiency. Running a smaller particle after a failed separation is a rational experiment, but it is not the only experiment, and it is frequently not the right one. A shallower gradient, a different pH, a temperature shift or a stationary-phase change addresses selectivity directly, and a successful method development programme tests each before concluding that only a smaller particle will serve. The practical consequence is that method robustness and lifetime depend on whether the selectivity separation is known and defended, or whether the method relies on brute-force efficiency to hide a hidden selectivity problem.

The full spectrum, recorded and discarded

A diode-array detector records ultraviolet absorbance across a wavelength range—commonly 200 to 400 nanometres—at every point in the chromatogram. From that data, two things of immediate value can be extracted. First, the spectrum can be examined at each peak to confirm spectral homogeneity—a pure peak has a spectrum that does not drift across its width, while a co-eluted pair shows a spectrum that changes. Second, the full spectrum can be plotted to reveal absorbers that were not visible at the single monitoring wavelength.⁷ A 254-nanometre impurity in a sample monitored at 280 nanometres is invisible in the 214-nanometre chromatogram and is therefore invisible in the purity report, until a diode-array scan reveals it.

Almost no research-peptide certificate prints this information. The data exists—it is generated automatically—and its absence from the report is a formatting decision, not a technical limitation. The usefulness of seeing the full spectrum would be apparent on the first certificate where it resolves an otherwise inexplicable discrepancy between the purity and the mass-balance calculation. It is one of the few changes to the standard format that costs nothing and adds information the reader genuinely needs.

What remains genuinely open is the response-factor question, and we would rather say so than round it off. Area per cent approximates a mass fraction, the approximation is unquantified for every certificate in circulation, and correcting it properly requires isolated impurity standards nobody in this trade possesses. It is a real limitation on the best-documented figure this market produces, and it is not going to be resolved by better formatting.