Mass Accuracy, Parts Per Million, and What a Mass Match Establishes

A certificate prints a measured mass beside a calculated one and the two agree to a couple of decimal places. Is that a good result? The honest answer is that the question cannot be settled from those two figures alone. Agreement is only interpretable against what the instrument was capable of resolving, and that capability gets stated in parts per million rather than in daltons.

A tolerance that scales with the molecule

Instrument error grows roughly in proportion to the mass under measurement. Quote it as a fixed number of daltons and it misleads at one end of the range or the other. Parts per million express it as a fraction instead: 5 ppm means five millionths of whatever value was measured.

One specification therefore implies quite different absolute tolerances depending on the molecule in front of it.

The first two sit comfortably inside the one-dalton differences that distinguish amidation or deamidation — assuming the instrument actually holds that specification. The third does not come close. A lower-resolution instrument quoting accuracy in the hundreds of ppm confirms that the compound is approximately the right size and stops there; it cannot tell you which terminal form you have.

The number belongs to a calibration, not to a machine

No instrument possesses mass accuracy intrinsically. What it possesses is accuracy while recently calibrated against standards of known mass, and that calibration drifts with temperature and with use.

Two approaches exist. External calibration is performed earlier in the day and hopes the drift since then is small. Internal calibration puts a reference compound into the same acquisition as the sample, correcting for drift at the instant of measurement, which makes it the more reliable of the two. Customer-facing reports seldom say which was used. The broader principle is set out in reference standards and traceability.

Isomers a mass measurement cannot separate

Suppose the tolerance is tight and the agreement is real. What has been established is that the measured mass is consistent with the molecular formula calculated for the claimed sequence. That is a meaningful constraint. It is not an identification.

Anything that rearranges the molecule while conserving its formula slips straight through. A sequence isomer carrying the same residues in a different order does. So does an isoaspartate rearrangement, a disulfide isomer with different connectivity, and a peptide containing a D-residue. Mass alone cannot separate any of them; racemization and chiral purity and disulfide bonds and scrambling treat two of those cases directly.

Separating them calls for fragmentation — breaking the molecule apart and measuring the pieces, which reports sequence instead of composition — or for an orthogonal separation. Routine identity packages include neither.

The figure worth looking for

When both masses appear, the informative quantity is the difference between them expressed in ppm. A report that states that figure is telling you something; a report listing two bare numbers is leaving you to guess. One prerequisite comes first, and it is covered in monoisotopic and average mass: both figures have to be the same kind of number before any comparison between them means anything.

All material is supplied for laboratory research use only. It is not a drug, not a supplement, and not for use in humans or animals.