Everything a buyer assumes has been checked, ranked by whether it has

It is important to be clear that these absences are not concealment. They are the shape of the form. Research-grade material is sold as a chemical, not as a preparation for administration, and a chemical supplier’s certificate reports the properties a chemist buying a reagent would want: identity, purity, appearance, sometimes water. Nobody removed the sterility line. It was never on the page, because the product is not being sold as a sterile preparation and its documentation does not pretend otherwise.

Water, counter-ion and the mass in the vial

A lyophilised peptide is not pure peptide even when it is chromatographically pure. It is a salt, usually of trifluoroacetic or acetic acid, containing residual water that a hygroscopic powder acquires readily, and sometimes residual solvent from purification. Three lines on a certificate address this and they are usually absent: water content, counter-ion identity and content, and residual solvent.

Water is determined by Karl Fischer titration or by loss on drying, and the pharmacopoeial methods for it are old, settled and inexpensive.¹ A peptide containing eight per cent water by mass contains eight per cent less peptide than its label implies, and the figure is not stable: it depends on how the vial was stoppered and how long it has been open. Counter-ion content is a larger contribution still for basic peptides purified in trifluoroacetic acid, where the counter-ion fraction can reach ten to twenty per cent of total mass.²

Put these together and the practical statement is the one this department repeats: the nominal mass on a research vial is an upper bound on the peptide it contains, not a value. A certificate that reports purity and is silent on water and counter-ion has told you the material is clean and nothing at all about how much of it there is.

What a batch number is supposed to mean

A batch, or lot, is a defined quantity of material produced in a single process run or a defined series of runs, homogeneous within itself, and identified by a unique code. That definition does real work: it is what makes it meaningful to test a sample and draw conclusions about the whole. If the material identified by one code is not homogeneous — if it was blended from separate syntheses, or filled across several sessions from stock stored differently — then a result on one sample generalises less well than the certificate implies.

In this trade batch codes range from the highly informative to the arbitrary. A code encoding the year, the month, the product and a sequence number tells a reader something and can be checked for internal consistency across documents. A four-digit code with no discernible structure cannot. Neither format is wrong; the difference is whether a reader can detect an anomaly.

The Journal’s dossier programme asks each company how batch codes are constructed and whether one code corresponds to one synthesis, one fill, or one shipment. The answers vary considerably and several companies have not previously been asked. We publish the answers without comment, because a code that identifies a fill session rather than a synthesis is a perfectly reasonable convention as long as a reader knows which convention is in use.

The certificate is a snapshot, and the vial has aged

Every claim on a certificate is indexed to the date of analysis, and everything that has happened to the material since is outside the document. For lyophilised peptides stored cold, dry and dark, the rate of change is slow but not zero: deamidation proceeds even in the solid state at a rate that depends on residual water, oxidation proceeds in the presence of air and light, and aggregation can occur after a temperature excursion that leaves no other trace.³

The practical significance depends on the interval. A certificate dated three weeks before shipment describes material that is, for most purposes, the material in the vial. A certificate dated fourteen months before shipment describes an earlier object. The Journal’s audit of certificates supplied through the dossier programme found a median interval between manufacture and analysis of eleven days, which is reassuring, and a median interval between analysis and the customer receiving the vial of somewhat over four months, which is the number nobody reports.

None of this argues for retesting every vial. It argues for reading the date, which takes two seconds, and for treating purity figures as historical rather than current. It also argues for taking the appearance line seriously, since a change in the cake is one of the few observations a buyer can make that bears on what has happened since the document was written.

The certificate as a file

A certificate that arrives as a PDF is a file with properties, and those properties are frequently more informative than the page. The document metadata usually records the producing application: chromatography data systems, laboratory information management systems, word processors and web-based form tools all leave distinguishable signatures. Creation and modification timestamps are often present. The original filename sometimes survives in the title field, and it occasionally names a different product or lot than the page does.

Structural properties help too. A page whose text can be selected and searched was generated as text; a page that is a single raster image was scanned or screenshotted, which destroys the metadata trail and is worth noticing when a supplier has the ability to send the original. Embedded font lists distinguish documents produced on different systems. A chromatogram present as vector graphics came from the instrument software; the same chromatogram as a low-resolution bitmap has been through a screenshot at some point.

None of this is proof of anything and it should not be presented as such. It is evidence about how a document came to exist, which the content of the page cannot supply. The Journal records these observations, asks the company about them, and publishes the exchange. Where a supplier responds by sending the original instrument-generated report, the question usually resolves immediately and in the supplier’s favour, which is worth saying because it happens more often than the alternative.

The systematic absences, and what each would cost

Peptide content is absent from almost every research certificate, and it is the absence with the largest practical consequence, because it is the number that determines how much peptide a nominal mass represents. Determination by elemental nitrogen analysis or quantitative amino-acid analysis is routine chemistry, and the compendial approach to amino-acid analysis for biotechnological articles is long established.⁴

Water content is absent nearly as often and is cheap to determine. Counter-ion identity and content are almost universally absent. Residual solvent appears occasionally. Bacterial endotoxin is absent, and its determination requires a different laboratory discipline and different reagents. Sterility is absent, requires fourteen days of incubation, and cannot be compressed. Container closure integrity is absent and is a packaging test rather than a chemical one.

The pattern is consistent: the tests that appear are the ones a chemical supplier’s laboratory already performs, and the ones that do not are the ones that would require a different laboratory. This is a rational commercial arrangement and it becomes a problem only when the resulting document is read as a general assurance of quality rather than as a chemical identity and purity statement, which is what it is and all it claims to be.

The ten-minute check

Minute one: find the batch number on the certificate and find it on the vial. Not the carton. If they do not match, or the vial has no number, stop and ask the supplier what the relationship is. Minutes two and three: find the date of manufacture and the date of analysis, and compute the interval. Then compute the interval between the date of analysis and today.

Minutes four and five: read the test table and count the columns. If the specification column is missing, the results cannot be assessed. If the method column is missing or says only HPLC, the purity figure cannot be compared with anybody else’s. Minute six: check the identity line for a theoretical mass, and check whether the convention — monoisotopic or average — is stated. Minute seven: read the signature block for a name and a role.

Minutes eight to ten: list what is not there. Content, water, counter-ion, residual solvent, endotoxin, sterility. Then decide whether any of those matter for what you are doing, which is a question only the reader can answer. The exercise does not establish that a certificate is right or wrong. It establishes whether the document can be checked at all, and in the Journal’s experience roughly a third of certificates in general circulation fail before minute five. Readers who work through this and find something they cannot interpret are welcome to write to standards@compoundjournal.com.

How this publication handles certificates

Four rules, arrived at over two years and revised twice. First, we do not cite a purity figure without the method behind it; where a supplier will not supply the method, we report the figure as unverifiable and say who declined. Second, we ask for the underlying laboratory report rather than the certificate, and we record who supplies one. Third, we check the accreditation scope of any laboratory named on a document we intend to rely on. Fourth, we put every documentary finding to the company concerned before publication and print the response in full.

The fourth rule is the one that has changed our coverage most. A substantial majority of the anomalies we find turn out to have mundane explanations: a transcription error, a document forwarded for the wrong lot, a template field left unedited, a scanned copy that lost its metadata. Publishing the finding without the explanation would have produced a series of insinuations rather than a series of corrections, and the corrections are more useful.

The rules also mean we publish less than we could. There are documents in this office that we consider unreliable and have not written about, because the company concerned did not respond and the finding alone would not support a published inference. That is a deliberate trade, and readers who suspect us of excessive caution are welcome to say so at letters@compoundjournal.com, where several already have.

A necessary distinction between a bad document and a bad product

Almost nothing in this article supports an inference about the contents of a vial. A certificate missing a specification column, an unsigned footer, a stale date of analysis and a batch number that appears only on the carton is a poor document. The material it accompanies may be excellent, and in the Journal’s experience frequently is: the analytical work behind these products is often better than the paperwork that reports it, because the paperwork is produced by a commercial function and the analysis by a laboratory.

The reverse also holds. A beautifully constructed certificate with four columns, two signatures and a named method is evidence of a functioning documentary process and is not evidence about the vial either, since a document cannot testify to material it does not accompany. This is why the Journal reports documents as documents and material as material, and declines to convert one into a claim about the other.

We labour the point because the alternative is a genre of coverage that treats documentary weakness as proof of dishonesty, and that genre is both unfair and useless. Unfair because most documentary weakness in this trade is inherited convention rather than intent. Useless because it gives a reader nothing to do. Reading the document properly gives a reader something to do, which is the entire purpose of this piece.

The Journal will keep asking the twenty companies in its dossier programme for the same seven lines each quarter, and will keep publishing who supplies them. Two years of doing this has produced measurable movement: the number of programme certificates carrying a specification column for every test has risen from four to nine, which is slow and is not nothing.