Accepted answer
Your reasoning breaks in three separate places, and each break is independently sufficient to kill the claim.
1. Identical-mass impostors are not unlikely, they are guaranteed
Solid-phase synthesis produces a specific family of by-products that have exactly the parent mass and cannot be seen by any mass measurement at any resolution:
- Iso-aspartate. The aspartimide intermediate that forms during piperidine deprotection can reopen at either carbonyl. Reopening the wrong way gives a beta-linked backbone through the Asp side chain. Same atoms, same mass, different molecule, different three-dimensional structure. Semaglutide has one Asp, tirzepatide has two.
- D-amino acid epimers. Racemisation at a stereocentre, most often at Asp or during activation of Cys and His. A D-residue is a stereoisomer: identical formula, identical mass, identical MS/MS fragment masses. Completely invisible to mass spectrometry.
- Sequence scrambling and transposition. If two residues swap positions, the composition is unchanged and so is the mass. MS/MS can sometimes catch this from the fragment ladder; intact MS never can.
- Diketopiperazine and cyclic species that happen to be isobaric with the linear parent.
These are not exotic. In a poorly controlled Asp-containing synthesis, iso-Asp plus D-Asp together can run to several percent, and an intact-mass scan will report the batch as a single clean species.
2. A mass spectrum is not a quantitative measurement
Peak intensity in ESI is the product of how much material is present and how well that particular molecule ionises. Ionisation efficiency varies between structurally similar peptides by factors of several, sometimes an order of magnitude, and it varies with the mobile phase, the concentration, the presence of the other components (ion suppression), and the day.
So even if you could see every impurity, you could not turn intensities into percentages. There is no calibration in a scan. The number "100%" has no experimental basis: to state a percentage you need a detector with a response you can relate to mass, which for peptides at 214 nm is roughly the count of peptide bonds — imperfect, but at least monotonic and correctable with a response factor.
3. "No impurity peaks detected" has no detection limit behind it
Ask the vendor what impurity level that experiment could have detected. There is no answer, because nobody spiked a known impurity at a known level to find out. In an infused intact-mass scan of a 4 kDa peptide, a 1% impurity of comparable ionisation efficiency is often lost in the isotope envelopes and noise. Something at 0.3% is essentially undetectable. Meanwhile an HPLC method with a validated disregard limit will report peaks down to 0.05% of the main peak area and can tell you it did so.
The claim is therefore not merely unsupported, it is unfalsifiable as written.
What each test actually establishes
| Question | Test that answers it | Test that does not |
|---|
| Is this molecule semaglutide? | ESI-MS intact mass | HPLC alone (retention time is weak evidence) |
| What fraction of the organic material is the target? | RP-HPLC area % at 214 nm | MS |
| How many mg of target are in the vial? | Quantitative HPLC against a reference standard | Both of the above |
| Is the sequence and the modification correct at every position? | LC-MS/MS peptide mapping | Intact MS |
| Are there stereochemical or isomeric impurities? | Chiral amino-acid analysis after hydrolysis, or an orthogonal separation | Any mass measurement |
Is there an MS experiment that supports a purity claim?
Sort of, with caveats. LC-MS with UV in line is the useful configuration: quantify off the UV trace, use the MS to identify each UV peak. That gives you a purity number with structural assignment for every impurity, which is far more informative than either alone. This is what a decent related-substances method looks like.
What LC-MS still cannot do is see isomers that co-elute and have identical fragments. For those you need a genuinely orthogonal dimension: a chiral separation, ion mobility, or electron-transfer dissociation to catch iso-Asp via the diagnostic c and z ion mass shifts.
So: MS is necessary and not sufficient. Treat "MS-verified purity" as a phrase that tells you something about the vendor's analytical literacy rather than about the vial. Independent reports from Janoshik, Medutest and PeptideMeter give you an HPLC purity figure and an MS identity confirmation as separate line items, precisely because they are separate questions.
edited 26 Feb 2025 by forty_two_c — tightened the wording; no substantive change
2The unfalsifiability point is the one to use in a vendor conversation. It ends the argument quickly. – sian_llewellyn 6 months ago Would add that a UV-MS split with the UV as the quantitative channel is also how the pharmacopoeial methods are set up. – assay_blank 4 months ago add a comment