Why one peptide gives several peaks in a mass spectrum
An electrospray mass spectrum of a pure peptide shows a series of peaks, none of them at the molecular mass. We explain charge states, isotopes and adducts.
- Published
- Theme
- Analytical methods
- Reading time
- 5 minutes
A certificate of analysis states a theoretical mass, say a little over 1400 Da. Attached to it is a mass spectrum. We look for a peak at 1400 and find nothing there. Instead there are tall signals near 710 and near 474, and a few small ones beside them. The peptide is fine. The spectrum is simply answering a different question from the one most readers expect.
A mass spectrometer does not measure mass. It measures mass divided by charge, written m/z. Once that is clear, the rest of the picture falls into place.
Charge states: the same molecule, counted several times
Most peptide spectra on certificates are recorded with electrospray ionisation. In the source, the peptide picks up protons, mainly on its basic sites: the N-terminal amine and the side chains of lysine, arginine and histidine. A peptide with several such sites can carry one, two, three or more protons, and each version appears at its own position on the axis.
The arithmetic is short. For a peptide of mass M carrying n protons, the signal appears at (M + n × 1.007) / n. The 1.007 is the mass of one proton, in daltons.
| Ion | Calculation for M = 1419.6 | Expected m/z |
|---|---|---|
| [M+H]+ | (1419.6 + 1.007) / 1 | 1420.6 |
| [M+2H]2+ | (1419.6 + 2.014) / 2 | 710.8 |
| [M+3H]3+ | (1419.6 + 3.021) / 3 | 474.2 |
The mass in the table is an invented example, chosen only to make the division visible. The pattern is what matters: several peaks, one molecule. Which charge state dominates depends on the number of basic residues, on the solvent and on the instrument settings. A short peptide with a single basic site may show almost only the singly charged ion. A longer one, rich in lysine and arginine, may hardly show it at all.
How we check that the peaks belong together
We take each candidate peak and work backwards: multiply the m/z by the assumed charge, then subtract the protons. If 710.8 is a doubly charged ion, the mass is 710.8 × 2 − 2.014 = 1419.6. If 474.2 is triply charged, 474.2 × 3 − 3.021 gives 1419.6 again. Two peaks, one answer. That agreement is the evidence. A single peak can be coincidence; two or three that converge on the same mass rarely are.
Most modern software does this automatically and prints a "deconvoluted" mass. We still do the sum by hand on at least one peak, because a deconvoluted figure with no raw spectrum behind it asks us to trust a step we cannot see.
Isotopes: why each peak is a small cluster
Zoom in on any one of those signals and it is not a single line. It is a cluster of lines, evenly spaced. The cause is carbon-13. About one carbon atom in a hundred is the heavier isotope, and a peptide contains dozens of carbons, so a fair share of the molecules carry one, two or three extra neutrons.
The spacing inside the cluster tells us the charge directly. Neighbouring isotope lines differ by one dalton in mass, so on the m/z axis they sit 1 apart for a singly charged ion, 0.5 apart for a doubly charged one and about 0.33 apart for a triply charged one. If the instrument resolves the cluster, we do not need to assume the charge. We can read it.
This is also where two different "theoretical masses" come from. The monoisotopic mass counts only the lightest isotope of each element and corresponds to the first line of the cluster. The average mass uses the natural isotope mixture and falls somewhere inside the cluster. For a peptide of this size the two differ by roughly one dalton. A certificate should say which one it quotes. When it does not, and the observed value looks about one unit off, this is the first thing we check before suspecting the product.
Adducts: small companions at fixed distances
To the right of a main peak there are often smaller ones at regular offsets. The usual explanation is that a sodium or potassium ion has taken the place of a proton. Sodium adds about 22 Da to a singly charged ion compared with the protonated form; potassium adds about 38. On a doubly charged ion the same swap moves the peak by half as much, about 11 and 19.
These adducts come from glassware, solvents and buffers, not from the synthesis. They tell us about sample handling at the analytical lab. They do not count as impurities of the peptide, and a spectrum with a modest sodium peak is entirely ordinary.
Signals that do deserve a question
Some extra peaks are not arithmetic artefacts. After deconvolution, a second mass that differs from the main one by the mass of one amino acid residue points to a deletion sequence, a chain that missed one coupling step. A difference of about +16 Da often indicates oxidation, most commonly on methionine. A difference matching a known protecting group suggests incomplete deprotection.
We read these with caution. Mass spectrometry confirms identity well and counts badly. Ionisation efficiency differs from one species to another, so the height of a peak in the spectrum says little about how much of that species is present. Quantity is the job of the chromatogram. The two documents answer different questions and we need both.
What we expect to find on the document
Our reading order for the mass spectrometry part of a certificate is fixed:
- The theoretical mass, with a note of whether it is monoisotopic or average.
- The raw spectrum, with the m/z axis legible and the main peaks labelled.
- The charge state assigned to each labelled peak.
- The observed mass and its difference from the theoretical one.
- The batch number on the spectrum itself, matching the one on the certificate.
The last point is easy to overlook. A spectrum without a batch number could belong to any batch. For a plain-language overview of what a buyer can verify on such documents, PEPTIDÁE has published an article on how to check peptide purity, in Romanian, and their product listing shows how research compounds are presented by a catalogue that sells them for laboratory use only.
When the numbers converge, the cluster spacing fits the assigned charge and the batch number matches, we write "identity confirmed" in our own records, with the date. When one of those is missing, we write what is missing and ask for it.