Mass accuracy, resolving power, and ppm error: what 'matches the theoretical mass' actually establishes

Analytical reports for synthetic peptides routinely include a line stating that the observed mass matches the theoretical value. The statement is compact and reassuring, but it conceals a good deal of variation. A match to within 0.5 Da on a low-resolution quadrupole and a match to within 2 ppm on a high-resolution orbital trap are not equivalent claims, even though both are written as “conforms.” The strength of the statement depends on two instrument properties, mass accuracy and resolving power, and on how the measurement was calibrated.

This post defines those terms, works through what a given tolerance means for peptides of different sizes, and describes what a mass match can and cannot establish about identity. The discussion concerns interpretation of analytical data only.

Defining accuracy, precision, and resolving power

Mass accuracy describes how close a measured mass is to the true value. It is usually expressed as a relative error in parts per million: the difference between the measured and theoretical mass, divided by the theoretical mass, multiplied by one million. Expressing error relative to mass is convenient because many instrument errors scale with the m/z value being measured. Mass precision, by contrast, describes the scatter between repeated measurements. An instrument can be precise but inaccurate if it is consistently offset by a poor calibration, and a report that quotes only replicate agreement says nothing about whether the central value is correct.

Resolving power is a separate quantity. It is commonly defined as the m/z of a peak divided by its width at half of its maximum height. A higher value means narrower peaks and a greater ability to separate ions with nearly identical m/z. Resolving power and accuracy are related in practice, since a narrow peak is easier to centroid precisely, but they are not interchangeable. An instrument with modest resolving power can still deliver good accuracy for an isolated, well-calibrated peak, whereas an instrument with high resolving power can report a poor value if its calibration has drifted.

Different analyzer types occupy different regions of this space. Quadrupole and ion trap instruments used for routine confirmation often operate at unit resolution, adequate to assign a nominal mass but not to resolve fine structure. Time-of-flight analyzers commonly reach resolving powers in the tens of thousands, and orbital trap and ion cyclotron resonance instruments can exceed one hundred thousand. For orbital traps the resolving power also falls as m/z increases, so a specification quoted at one reference m/z does not apply unchanged across the scan range.

What a tolerance means at different masses

A tolerance in ppm translates into an absolute window that grows with mass. For a peptide with a monoisotopic mass of 1,000 Da, 5 ppm corresponds to 0.005 Da. For a 4,000 Da peptide, the same 5 ppm corresponds to 0.020 Da. A quoted agreement of 0.3 Da, typical of unit-resolution work, corresponds to roughly 300 ppm at 1,000 Da and about 75 ppm at 4,000 Da. The same absolute difference therefore carries different meaning depending on the size of the molecule, which is why reports that give only a difference in daltons are harder to compare across compounds than reports that give ppm.

The monoisotopic mass is the appropriate reference for high-resolution work, because the resolved lowest-mass isotopologue peak can be measured directly. Unit-resolution instruments, which cannot resolve the isotope envelope for larger species, report an average or centroid mass, and the corresponding theoretical value must be calculated on the same basis. Comparing a measured monoisotopic mass with an average theoretical mass, or the reverse, produces an apparent discrepancy of a fraction of a dalton to more than a dalton for peptides in the multi-kilodalton range. This basis mismatch is a common source of apparent failures that are really bookkeeping errors, and the distinction is treated in more detail in a separate post on monoisotopic and average mass.

Multiply charged ions add a further step. The instrument measures m/z, and the neutral mass is obtained by multiplying by the charge and subtracting the mass of the charge-carrying protons. An error in assigning the charge state, or the use of the mass of a hydrogen atom rather than a proton, introduces a systematic offset of about 0.0005 Da per charge. That is small in absolute terms but not negligible at the level of a few ppm on a small peptide.

Calibration and the sources of error

Most of the accuracy a high-resolution instrument delivers comes from calibration rather than from the analyzer alone. External calibration uses a reference mixture of known compounds acquired before or between samples. It is simple, but it assumes that instrument conditions do not change between the calibrant and the analyte, and temperature variation, electronic drift, and space-charge effects can each degrade it over hours. Internal calibration, in which a reference compound of known mass is introduced into the same spectrum as the analyte, corrects for these changes scan by scan and generally yields lower errors. Lock-mass approaches are a variant in which a background ion or continuously infused reference supplies the correction.

Space-charge effects deserve particular mention. When too many ions occupy the analyzer or trap at once, their mutual repulsion shifts the apparent m/z. A concentrated sample or an intense co-eluting contaminant can therefore push a measurement outside the instrument’s nominal accuracy specification. Detector saturation distorts peak shape and centroiding in a related way. A reported mass error that was obtained from a saturated or overloaded signal is less reliable than one from a signal well within the linear range.

Peak shape and overlap also matter. If an adduct, an isotope peak of a neighboring species, or a background ion overlaps with the signal of interest at the instrument’s resolving power, the centroid is pulled toward the interferent and the measured mass is biased. This is one practical reason that spectral interpretation should be paired with an inspection of the underlying data, not only the summary line.

What mass alone cannot distinguish

A mass match narrows the candidate structures but does not single one out. Isomeric species have identical elemental compositions and therefore identical exact masses, however accurate the measurement. A peptide and its variant with two adjacent residues transposed, or a form carrying a D-amino acid in place of the L-form, cannot be distinguished by mass at all. Distinguishing them requires chromatographic separation, fragmentation, or chiral analysis, topics treated elsewhere in this journal.

Even for species of different composition, the number of formulas that fit a given mass within a tolerance increases quickly with molecular size. For a small peptide, a tolerance of a few ppm may leave only one chemically reasonable formula. For a peptide of several kilodaltons, many compositions can fall inside the same window, and constraining the answer relies on knowing the expected formula in advance. In identity testing this is the case: the expected sequence is known, and the question is whether the observation is consistent with it. Consistency is a weaker statement than uniqueness, and it is the former that a mass match supplies.

Small mass differences between plausible variants illustrate the resolving power requirement. Deamidation of an asparagine or glutamine residue raises the monoisotopic mass by 0.984 Da, while the first carbon-13 isotope peak of the unmodified peptide lies 1.003 Da above its monoisotopic peak. The two differ by only about 0.019 Da. At a singly charged m/z of 4,000, separating them at half-height requires a resolving power in the region of 200,000, and at lower resolving power the deamidated species is simply unresolved and appears as a distortion of the isotope envelope. For multiply charged ions the required resolving power is lower at the same neutral mass, because the m/z value is smaller, but the principle holds: the ability to see a small modification depends on the instrument as much as on the sample.

Reading a mass statement critically

A useful report states the instrument type, the ionization mode, the charge states observed, the mass basis used for comparison, the calibration approach, and the error in ppm or in daltons. Where these details are absent, the statement can still be accurate, but its strength cannot be judged. A tolerance appropriate to the instrument is also informative in itself. A low-resolution acquisition reported with a tolerance of a few ppm would be implausible, and a high-resolution acquisition accepted at several hundred ppm suggests that the specification has been set to the capability of an older method.

The mass line on a certificate is best understood as one piece of an identity assessment that also draws on retention behavior, fragmentation, and, where relevant, orthogonal methods. Taken alone it establishes consistency with a composition, not structural proof. Taken with the instrument context and error value that accompany it, it becomes a quantified statement that can be compared between laboratories and between lots, which is the practical value of reporting accuracy in ppm in the first place.