What is measured, by whom, by which method, and what each figure can and cannot establish. The full method reference sits on its own page; this is the summary.
Independent laboratory testing
Testing is commissioned from a third-party laboratory with no interest in the result. The laboratory is named on the record it produced, because an unnamed "independent lab" is an assertion that cannot be checked. Internal quality control also happens; it is not what we ask anyone to rely on.
Identity verification
Identity is established by mass spectrometry: the observed molecular mass is compared against the mass calculated from the known structure. An identity failure ends the lot. It is not re-run, because a mass that does not match is not a measurement problem — the material is not what it was labelled as.
HPLC purity analysis
Purity is determined by reverse-phase HPLC and reported as area percent: the main peak's integrated area against the total integrated area of the run. That figure is a proportion of what the detector saw, which is why it is published alongside the method and the trace rather than on its own.
Mass spectrometry and LC-MS
Where a question needs both answers at once — not only how large an impurity peak is, but what it is — liquid chromatography and mass spectrometry are run coupled, so each separated component is identified rather than merely counted.
Chromatogram interpretation
The full chromatogram is published as issued, including baseline noise and minor peaks. Traces are not redrawn as summaries and are not cropped to move an inconvenient peak out of frame. Retention time, peak area, baseline behaviour and resolution are all readable from the trace as published.
Lot traceability
A lot code is printed on the vial and it is the key to everything else. It resolves to one permanent record carrying the manufacture and test dates, the laboratory, the methods, the results, the chromatogram and the certificate. Testing is reported per lot, never once per product line.
Certificate documentation
The certificate of analysis is published in full on the lot record. A purity figure is never published without the document behind it, a figure from an in-house laboratory is never presented as independent, and no result is rounded upward.
Release standards
A lot moves through a fixed sequence of states and a person advances each one. An out-of-specification result has two honest outcomes: a documented, assignable laboratory error with both the original and the repeat retained, or no assignable error, in which case the result stands and the lot is not released. Assays are not repeated until a number improves.