The Testing Standard
Purity is the number this industry leads with, and it answers almost nothing alone. It says how much of the vial is one substance — never which substance, never how many milligrams you actually received, never what the rest is. Our standard is built around that question and one rule of order: the specification comes first. Every number a lot must meet is published before the lot exists. A standard written after the test is not a standard. It is a description.
Fourteen tests. Five gates. One record.
A lot is one production run: material synthesized together, filled into vials together, tested together. What the tests find is true of every vial in it — point 10 makes that claim honest. A specification is the list of numbers a lot has to meet. We call the system the 14-Point Quality Framework: fourteen tests, five release gates, one public record.
Eleven of the fourteen points run on every lot, without exception. The other three — aggregation, potency, and secondary structure — are qualification tests: the work done before a product first goes on sale, redone whenever the synthesis route, the container, or the supplier changes. Process-level checks re-run on a defined cycle and on any change.
Between synthesis and sale, every lot stops at five release gates and waits for evidence. No evidence, no movement.
Every test is run by an independent laboratory with no commercial stake in the result, and the certificate lives on that laboratory’s system, not ours — a file we host is a file we could have edited.
One public record holds every outcome. Passes go to the lot ledger. A failure means the same thing every time: the lot does not release, the material is destroyed, and the record is published on the failed lots page with the number it missed.
The fourteen points
Plain names first, instruments second. The last line of every card is identical, because the consequence is.
1 Identity — is it the molecule on the label?
Every lot.
High-resolution LC-MS/MS measures the molecule’s exact mass and its fragments; the first lot from each manufacturing site also gets full sequence confirmation. Purity is how much; identity is which — and nothing else on the certificate matters until this passes.
Failure: lot destroyed, record published.
2 Purity — how much of the vial is one substance?
Every lot.
RP-HPLC separates everything in the sample; full-spectrum PDA detection reads the whole light spectrum, so related impurities cannot hide under the main peak. Purity is real and necessary — just not the finish line: it counts one substance and cannot name it.
Failure: lot destroyed, record published.
3 Net peptide content — how many milligrams are actually there?
Every lot.
qNMR — quantitative NMR — measures the true mass of peptide in the vial, cross-checked by amino acid analysis. Label weight includes water and counterions; peptide is only part of it. This is the most commercially consequential number on any certificate: two vials with identical labels and purity can hold different amounts of peptide. The label is not this number.
Failure: lot destroyed, record published.
4 Water content — how much of the weight is water?
Every lot.
Karl Fischer titration, a method built to measure water and nothing else. Freeze-dried peptides absorb moisture without looking any different, and water is label weight that is not peptide — and a stability risk the buyer inherits unseen.
Failure: lot destroyed, record published.
5 Counterion content — what salt form makes up the mass?
Every lot.
Ion chromatography measures the counterion the peptide carries — TFA left over from synthesis, or acetate — and how much of the mass it accounts for. Two suppliers selling “the same” peptide can be selling different salt forms with different real content; labels rarely say which.
Failure: lot destroyed, record published.
6 Structural integrity — are the bonds in the right places?
Every lot, where the molecule has the feature.
Peptides with disulfide bridges get peptide mapping to confirm each bridge connects the right pair of cysteines; free thiol is measured by Ellman’s assay. Right mass, high purity, and a bridge in the wrong place is still the wrong molecule.
Failure: lot destroyed, record published.
7 Residual solvents — what synthesis chemicals remain?
Every lot.
GC-MS headspace analysis measures the solvents left behind by synthesis and purification. Solvents are a normal part of manufacturing; how much remains is measured against a published limit, not assumed.
Failure: lot destroyed, record published.
8 Elemental impurities — what metals came along?
Every lot.
ICP-MS measures metal residues, including palladium, a catalyst some synthesis routes rely on. Metals are invisible to an HPLC purity run — a vial can read 99% pure and still carry them.
Failure: lot destroyed, record published.
9 Endotoxin and bioburden — what did bacteria leave behind?
Every lot, where the format calls for it.
Bioburden (USP <61>/<62>) counts living microorganisms. Endotoxin (USP <85>) measures what dead ones leave behind — fragments of bacterial cell walls that survive sterilization. Contamination does not announce itself.
Failure: lot destroyed, record published.
10 Content uniformity — is every vial the same?
Every lot.
USP <905>: vials pulled from across the fill run are measured to confirm the first vial matches the last. A certificate describes a whole lot; uniformity is what makes one certificate honest for every vial in it — including yours.
Failure: lot destroyed, record published.
11 Aggregation state — has the peptide clumped?
Qualification tier.
SEC-MALS — size-exclusion chromatography with light-scattering detection — finds molecules stuck together into aggregates. Aggregated peptide can still look clean on a purity run; clumping is a different failure with its own instrument.
Failure: lot destroyed, record published.
12 Relative potency — does it perform against a reference?
Qualification tier.
SPR or a functional assay measures the lot against a reference standard: does the molecule bind and behave in the instrument the way the standard does. Identity says which molecule; potency asks whether this lot performs against a benchmark.
Failure: lot destroyed, record published.
13 Secondary structure — does it fold into the right shape?
Qualification tier, where the molecule has a defined shape.
CD and FTIR spectroscopy read the fold of the chain. A correct sequence folded wrong is not the material the specification describes; peptides with no stable fold carry this point as Not applicable, with the reason stated.
Failure: lot destroyed, record published.
14 Container closure integrity — does the sealed vial stay sealed?
Every lot.
Vacuum-decay and headspace testing confirm the stopper and crimp actually keep air and moisture out. Every point above describes the material on testing day; this one decides whether it is still that material after shipping and storage.
Failure: lot destroyed, record published.
Applicable. Not applicable. Triggered.
Every certificate lists all fourteen points, every time — none is quietly dropped. Each carries one of three marks.
- Applicable: the test ran, and the certificate shows the measured result beside the published limit.
- Not applicable: the attribute does not exist for this molecule or this format, and the certificate states why. A peptide with no cysteines has no disulfide bridges to map. That is not an exception to the every-lot rule — an attribute a molecule does not have cannot be tested, only acknowledged.
- Triggered: the test runs when a defined event calls for it — the first lot from a new site, or any change of route, container, or supplier.
Why the marks earn their space: endotoxin limits were written for sterile vial preparations, where bacterial residue in the sealed contents is the hazard being controlled. A capsule is a different presentation with different risks. It is still tested for identity, purity, content, and bioburden — but judging it against a vial’s endotoxin limit would produce a reassuring row in a table and no information. A certificate that prints a passing number for a test that cannot apply is telling you its table was assembled to look complete. “Not applicable,” written with its reason, is evidence that someone examined the actual product. That is the difference between a certificate and a certificate-shaped document.
The standard is half the skill
Fourteen tests, five gates, one record — none of it asks for your trust. It asks for your reading. Certificate reading is a learnable skill, and it transfers: read ours line by line and you can interrogate any supplier’s. Start with the guide, then resolve any lot number against the laboratory’s own records yourself.