Two suppliers, one compound, two certificates. One reports 99.1 per cent purity, the other 98.2 per cent. The first looks better and may well be worse, and the only way to know is to compare the documents line by line rather than headline by headline.
This works through that comparison: which lines are directly comparable, which are not comparable at all, and how to reach a decision when the documents disagree about what to measure. The underlying read of a single document is covered in reading a certificate of analysis line by line; this is the next step.
Start with what each document chose to measure
Before comparing any value, list the tests each certificate reports. This is usually the most informative thing in the whole exercise and it takes a minute.
A certificate reporting identity, purity, related substances, net peptide content, water content and residual solvents is a different class of document from one reporting purity and appearance. The second is not a worse result; it is a smaller claim. A supplier who measures six things and reports all six has given you more to object to than one who measures two, which is the opposite of how it tends to feel.
The practical rule is that absence is information. A missing net peptide content line means the real cost per milligram cannot be calculated, and price per milligram explains why that single omission can hide a 20 per cent price difference. A missing water content line means residual moisture is unknown, and moisture drives slow hydrolysis in a lyophilised cake.
Why two purity figures rarely compare
Purity by chromatography is a percentage of peak area under a specific set of conditions, and changing the conditions changes the number. Gradient, column, detection wavelength and run length all affect how well impurities separate from the main peak and whether they are seen at all.
A short gradient on a short column resolves less than a long one. Impurities that co-elute with the main peak are counted as the main peak. So a method that separates poorly reports a higher purity for the same material — the number goes up precisely because the method is seeing less.
This is why the method matters more than the figure, and why a certificate that gives a percentage without stating the conditions is not comparable to anything. HPLC purity explained covers what the measurement represents and where it stops.
| Line | Supplier A | Supplier B | Comparable? |
|---|---|---|---|
| Purity | 99.1% | 98.2% | Only if the methods match |
| Method stated | No | Gradient, column, wavelength given | B is assessable, A is not |
| Related substances | Not reported | Largest single 0.6%, total 1.8% | No — one is absent |
| Identity | Mass spectrometry | Mass spectrometry | Yes |
| Net peptide content | Not reported | 87.4% | No — one is absent |
| Water content | Not reported | 4.1% by Karl Fischer titration | No — one is absent |
| Laboratory | Not named | Named, accredited | B is verifiable |
Two certificates for the same compound, compared line by line
On the headline figure A wins by 0.9 points. On everything else B is the only document that can be checked at all. A's 99.1 per cent is unfalsifiable — there is no stated method against which to judge it and no related-substances breakdown to reconcile it with. B's 98.2 per cent comes with a 1.8 per cent impurity total that accounts for the difference, which is internally consistent and therefore credible.
The arithmetic that catches an inconsistent document
Where a certificate reports both purity and related substances, the two should roughly reconcile. If purity is 98.2 per cent and total related substances are 1.8 per cent, the document adds up. If purity is 99.1 per cent and total related substances are 3 per cent, something is wrong — either the numbers come from different methods, or from different lots, or one of them is not what it claims to be.
This check costs ten seconds and finds a surprising number of problems. Peptide impurities explained covers the categories that should appear in the breakdown, and peptide purity grades covers what the headline percentage conventionally means.
Lines that are genuinely comparable
Three, reliably. Identity method — mass spectrometry against mass spectrometry is a fair comparison, and a certificate with no identity method at all has not established what is in the vial. Water content, where both report it by a named method such as Karl Fischer titration. And net peptide content, which is the one number that makes price comparable across suppliers.
Salt form is comparable in the sense that it must be known, though the forms themselves are not better or worse. TFA and acetate salt forms covers what each implies for mass.
Lines that are not comparable, and why
Appearance descriptions are prose and vary by house style. Specification ranges are the supplier's own and tell you what they committed to, not what the material is. And any figure without a method is a claim rather than a measurement.
A particular trap is comparing a specification against a result. A line reading 'not less than 98 per cent' is a limit; a line reading '98.2 per cent' is a measurement. A certificate that gives only limits has told you what the material was allowed to be, which is a different document from one telling you what it was.
Dates, and what they tell you
Three dates appear on certificates and they answer different questions, which is why comparing the wrong pair produces confusion.
The manufacture or synthesis date says when the material was made. The release or issue date says when the analysis was signed off. The re-test or expiry date says how long the supplier stands behind the result. A long gap between the first two means the lot waited for analysis; a long gap between the second and today means the lot has been in inventory.
Neither gap is a fault on its own. A lyophilised peptide in sealed glass under cold storage is not deteriorating quickly, and material held for a year may be indistinguishable from material released last week. What the dates give you is a basis for asking about storage in the interval, which is the question that actually matters. Retest dates and expiry dates covers what each commits the supplier to.
Where two certificates are otherwise equivalent, the fresher lot is the better purchase at the same price — not because the older one is suspect, but because you inherit more of the re-test window.
When one document is a scan and the other is not
Format is not quality, but it carries information. A certificate issued as a searchable document from a named laboratory is easier to verify than a photograph of a printed page, and a scanned image with no issuing header is the hardest case of all.
The reason this matters is traceability rather than aesthetics. A document you can read is a document whose issuer you can contact, and how to verify a certificate of analysis covers doing exactly that. A scan with the laboratory name cropped out cannot be verified at all, which collapses several of the comparison rows to unknown at once.
The practical response is to ask for the original. A supplier who has it will send it; a supplier who cannot produce anything beyond an image has answered a different question than the one asked, and that answer belongs in the comparison.
Reaching a decision
Rank on completeness first, verifiability second, and reported values last. A thorough document from a named accredited laboratory with a slightly lower purity figure is a better purchase than a bare percentage from an unnamed source, because you can act on the first and can only hope about the second.
Where both documents are thorough and genuinely comparable, then the numbers decide, and net peptide content usually matters more to real cost than the last half-point of purity. Where one document is thin, the useful move is to ask for the missing lines rather than to discount the supplier — the response is informative either way. Comparing suppliers on documentation rather than price covers reading the pattern across a catalogue, and how to verify a certificate of analysis covers confirming a document with the laboratory that issued it.