Sodium and potassium adducts in electrospray spectra: reading the satellite peaks

A mass spectrum of a synthetic peptide rarely consists of a single clean peak per charge state. Alongside the principal protonated ion there are frequently smaller signals at fixed offsets above it, and an unfamiliar reader may take these for impurities, degradation products, or a second species in the sample. In many cases they are none of these. They are adducts, ions in which a metal cation has taken the place of one or more of the protons that normally carry the charge. Recognizing them is a routine part of interpreting electrospray data, and the pattern they form says something useful about the sample’s history rather than about the peptide itself.

This post describes how adducts arise, how their mass offsets are calculated, how they are distinguished from true chemical variants, and what they can and cannot indicate on an analytical report. The discussion is limited to interpretation of analytical data.

How adduct ions form

Electrospray ionization produces gas-phase ions from charged droplets. In positive mode, a peptide is expected to acquire its charge by protonation at basic sites: the N-terminal amine and the side chains of arginine, lysine, and histidine. The protons are supplied by the solvent, which is usually water and acetonitrile acidified with formic acid or trifluoroacetic acid. If alkali metal cations are present in the droplet, they compete with protons for the same binding sites. Sodium and potassium are the usual culprits because they are ubiquitous at trace levels in glassware, water, reagents, and the peptide material itself.

An alkali cation attaches through coordination to carbonyl oxygens of the backbone and to carboxylate groups of aspartate, glutamate, and the C-terminus. Unlike a proton, a metal cation does not readily leave once the droplet has evaporated, and it is not removed by the collisional conditions normally used in the ion source. The result is an ion with the same peptide but a different charge-carrying cation. Because the peptide carries more than one charge in a typical spectrum, mixed species occur: an ion carrying two protons and one sodium appears as well as the all-proton form and the form with two sodium ions.

The competition is not even. Sodium and potassium bind more tightly to many peptide sites than a proton does under electrospray conditions, so a modest concentration of salt can produce a disproportionate adduct signal. Ion-suppression effects also play a role, since the salts reduce the efficiency with which droplets shrink and release ions, lowering the overall signal and degrading the signal-to-noise ratio for every species in the spectrum.

Calculating the offsets

The offset of an adduct from the protonated ion is determined by the mass difference between the metal ion and the proton it replaces. A sodium adduct is heavier than the protonated species by about 21.982 Da, the difference between sodium (22.990 Da) and hydrogen (1.008 Da). The corresponding potassium offset is about 37.956 Da. On the m/z axis these differences are divided by the charge state, which is the key to recognition: a singly charged ion shows satellites at +21.98 and +37.96, a doubly charged ion at roughly +10.99 and +18.98, and a triply charged ion at about +7.33 and +12.65.

Analysts therefore look for a regular pattern. If the satellites sit at consistent fractions of the same two mass differences across several charge states, and the spacing between the isotope peaks confirms the charge assigned to each cluster, the assignment is secure. Multiple adducts on the same ion give a progression, with each additional sodium adding another 21.98 Da to the total mass. Because the spacing is derived from exact masses, high-resolution instruments can distinguish a sodium adduct from other small offsets with overlapping nominal masses, such as an oxidation product (+15.995 Da) or a methyl ester (+14.016 Da), which differ in a way that is obvious at resolving powers commonly available.

A related, less frequently discussed feature is the co-presence of ammonium adducts, which appear about 17.03 Da above the protonated mass for a given charge. These are common when ammonium acetate or ammonium formate buffers are used and are generally less stable, losing ammonia on activation.

Distinguishing adducts from chemical variants

Telling an adduct from a covalent variant matters because the two carry different implications. A covalent impurity, such as a deletion product, an oxidized residue, or a deamidated form, is a distinct molecular species. It may separate chromatographically from the main peptide, so that it appears at a different retention time in an LC-MS run. An adduct, by contrast, is created in the ion source and therefore co-elutes exactly with the parent. This is the simplest and most reliable diagnostic: if a satellite peak has the same retention time and peak shape as the main ion across the whole chromatographic peak, it is almost certainly a gas-phase adduct.

Fragmentation provides a second line of evidence. When a sodiated ion is selected for MS/MS, the fragment spectrum often shows fragments that retain the metal, and the fragmentation pattern can differ from that of the protonated ion because the cation is less mobile than a proton. The cation may remain on the C-terminal side, producing sodiated y-type ions. A shift in the fragment masses by 21.98 Da in specific ions localizes the cation but does not change the underlying sequence information.

Finally, the intensity of adducts changes with sample handling. Diluting the sample, desalting it with a reversed-phase trap, or adding a small amount of acid typically reduces the sodium and potassium signals relative to the protonated peak. Covalent impurities do not respond to these steps in this way. This behavior is part of why direct-infusion analyses, which lack a chromatographic separation, are more prone to adduct-dominated spectra than LC-MS.

Sources of the salt and what the pattern indicates

Alkali contamination has several routes into the sample. Borosilicate glass releases sodium into aqueous solutions, particularly under acidic or basic conditions and on prolonged contact, which is why plastic or polymer-coated vials are often chosen for trace-level work. Reagent-grade water and solvents carry trace salts. Peptide material that has been through ion-exchange or neutralization steps during purification may retain residual sodium or potassium as a counterion for acidic side chains. Biological buffers, if the peptide was handled in phosphate or Tris-buffered saline at any stage, can introduce large amounts of salt that dominate the spectrum.

For this reason, the adduct pattern is informative about sample preparation rather than about identity. A spectrum with heavy sodium adduction suggests that the sample or vial contributed salt, and a spectrum with a strong potassium component often points to a different source than sodium, because the two ions come from different materials. A very clean spectrum with a single dominant protonated ion at each charge state suggests thorough desalting and a clean analytical path. None of this reflects the intrinsic quality of the peptide, and it should not be read as a purity statement.

Quantitation is affected as well. If the signal of a peptide is split among protonated, sodiated, and potassiated forms, the intensity of the protonated ion alone understates the total. Analysts who integrate extracted ion chromatograms for peptide quantitation generally sum the contributions of the major adducts or, preferably, control the salt level so that the protonated form dominates. When reporting relative abundance of impurities from MS data, failing to account for adducts can inflate the apparent number of species.

Limits of interpretation

Adduct identification has boundaries worth stating. First, the presence of adducts does not by itself show that a sample contains a given amount of salt; electrospray response to metal ions is nonlinear and depends on the peptide sequence, the solvent, and the instrument. A spectrum with strong sodium adducts could correspond to micromolar sodium in solution, whereas a peptide with many acidic residues might show adducts at lower salt levels than a basic one. Elemental or ion chromatographic methods, rather than mass spectrometry, are the appropriate tools for quantifying counterion content.

Second, adduct assignment depends on mass accuracy and on correct charge-state determination. At low resolution, an apparent sodium satellite can overlap with isotope peaks of the main ion, particularly for larger peptides where the isotope envelope is broad. Deconvolution software can mis-assign these overlaps unless the envelope is fitted explicitly.

Third, some satellites are neither adducts nor covalent variants. Dimer ions, in-source fragments, and contaminants such as polyethylene glycol, which produces a repeating series spaced by 44.026 Da, have their own signatures. Matching the observed offsets to a table of expected values, and then checking them against retention time behavior, is the practical way to separate these possibilities.

Synthesis

Salt adducts are a feature of the ionization process, not of the peptide’s covalent structure. Their offsets are predictable, their pattern across charge states is regular, and their co-elution with the parent ion distinguishes them from true impurities. Read correctly, they reveal something about the glassware, solvents, and sample preparation steps that preceded the measurement, and they flag when a spectrum needs to be repeated after desalting before any conclusion is drawn about composition. Read incorrectly, they can inflate an apparent impurity count or understate the peptide’s total signal. For anyone evaluating analytical documentation, the practical question is whether a report accounts for such species explicitly, specifies the integration approach used, and identifies the retention-time and mass-accuracy criteria used to assign each peak.