Two Certificates, One Batch, Two Different Numbers

Same batch, two laboratories, two different purity figures. No analyst blundered. The spread opened up because a purity percentage is the output of a procedure, and those were not the same procedure.

Where the missing points went

Detector setting moves the number furthest, so begin there. At 214 nm the detector answers to the amide backbone, which means essentially every peptide-related species in the injection registers. At 280 nm only aromatic residues answer. Should a sequence carry neither tryptophan nor tyrosine, it barely appears at all, and impurities without aromatics drop out of the trace entirely while sitting plainly visible at 214 nm. What follows is arithmetic. Fewer impurity peaks in the denominator, larger percentage on the page. The longer wavelength flatters.

Separation power is the second lever. Species eluting close to the target are resolved by a shallow gradient and swallowed by a steep one, and the usual near neighbor is a deletion sequence, a peptide short one residue. Whether such a species arrives as a peak of its own or merely as a bulge on the flank of the main one is settled by gradient slope, column chemistry and run length. Between those two outcomes, nothing whatever about the powder has changed.

Integration is the third lever and the least discussed. Somebody decides where the baseline runs, and somebody decides how small a feature has to be before software files it as noise. Both are judgment calls, encoded in a threshold or in an analyst habit, and each of them nudges the total.

A reading, not a property

None of this should surprise anyone once the definition is in view. Area of the main peak over total integrated area, under one stated set of conditions: that is the entire quantity. It records what a particular instrument did on a particular afternoon. The molecule does not carry it around.

The question sitting underneath purity

Homogeneity and correctness are separate claims, and the second is the weightier one. Establishing that the dominant peak really is the sequence ordered takes a mass spectrum agreeing with the expected molecular weight. Absent that, a high percentage reports that the contents are uniform without reporting what they are uniformly made of. How the two techniques divide the work between them is set out in our note on HPLC and mass spectrometry verification.

Compare method statements, not magnitudes

Ranking suppliers by the size of their percentages is therefore close to meaningless. Rank them by disclosure instead. A report naming column chemistry, gradient, wavelength and run time carries more information than a bigger number offered with no conditions attached, for the simple reason that only the first can be argued with. Better again is a reproduced chromatogram, where baseline steadiness and any shoulder riding on the main peak say things a summary figure never will. Background on working through these documents sits in how to read a certificate of analysis.

How our own paperwork is handled

Nothing printed goes in the box. The third-party reports we have obtained are published on our certificates of analysis page, and each one states which methods, HPLC or mass spectrometry or both, were run against the batch in question. Vials carry no batch numbering, so the link between a vial and its report is the cap and crimp coloring, checked against the photographs reproduced on the report itself.

ExoLabz supplies compounds for laboratory research use only. Nothing on this page is medical advice or a suggestion of human or veterinary use.

Related listings. Compounds leaving the column early, down in the polar region, are where all of this bites hardest; Selank 5mg is the plainest case in our range. Reports we have obtained appear on the certificates of analysis page.