Mass Spectrometry for Peptide Identity: Charge-State Envelopes, Monoisotopic vs. Average Mass, and the Limits of a Purity Number

A certificate of analysis usually carries two numbers: a purity figure from chromatography and a mass figure from mass spectrometry. They answer different questions, and the mass figure is the one most often reduced to a single line that hides how the measurement was made. This guide covers what an electrospray spectrum of a peptide actually contains, how a multiply charged ion series becomes an identity claim, why monoisotopic and average mass are different numbers for the same molecule, and where identity evidence and purity evidence each stop. It is written for someone holding a document and deciding how much weight it can carry.

The instrument measures mass-to-charge, not mass

A mass spectrometer separates ions by mass-to-charge ratio, m/z. Neutral mass is not observed directly; it is inferred. For a peptide ionized in positive mode by proton attachment, an ion carrying n protons appears at

m/z = (M + n × 1.00728) / n

where M is the neutral monoisotopic mass and 1.00728 Da is the mass of a proton. Every step from raw spectrum to reported mass runs through that relation, so an error in assigning n propagates directly into the reported mass.

The charge-state envelope is the identity evidence

Electrospray ionization does not produce one ion per molecule. In the paper that established the technique for large biomolecules, Fenn and colleagues described the distinguishing feature of electrospray spectra as coherent sequences of peaks whose component ions are multiply charged, each peak differing from its neighbors by one charge ^[1]. That sequence is the point. A single peak at a plausible m/z is a coincidence waiting to happen; a series of peaks that all reduce to the same neutral mass is a far tighter constraint, though it is only as strong as the assumption that the observed peaks are adjacent charge states.

Work the arithmetic for a hypothetical peptide of neutral monoisotopic mass 3000.0000 Da:

• 2+: (3000.0000 + 2 × 1.00728) / 2 = 3002.0146 / 2 = 1501.0073
• 3+: (3000.0000 + 3 × 1.00728) / 3 = 3003.0218 / 3 = 1001.0073
• 4+: (3000.0000 + 4 × 1.00728) / 4 = 3004.0291 / 4 = 751.0073
• 5+: (3000.0000 + 5 × 1.00728) / 5 = 3005.0364 / 5 = 601.0073

Four peaks, one molecule. Running it backward is the check that matters. Given two adjacent peaks at m/z values m₁ (the lower, higher-charge peak) and m₂, the charge on m₂ is (m₁ − 1.00728) / (m₂ − m₁). Using the 3+ and 2+ peaks above: (1001.0073 − 1.00728) / (1501.0073 − 1001.0073) = 1000.0000 / 500.0000 = 2. The charge is determined by the spacing, not assumed. Deconvolution software does this across the whole envelope and returns a single neutral mass. How many charge states agreed on that mass is what the number is worth, and a mass quoted on its own does not tell you.

At sufficient resolution there is a second, independent charge readout. Adjacent isotope peaks within one charge state are separated in neutral-mass terms by the difference between ¹³C and ¹²C, which is 1.00335 Da, so on the m/z axis they are separated by 1.00335 / n. A 3+ ion shows isotope peaks 1.00335 / 3 = 0.334 apart; a 2+ ion shows 1.00335 / 2 = 0.502. If a document reports a mass but shows no spectrum, you cannot tell whether either of these checks was performed.

What moves the envelope around

The distribution of charge states is not a fixed property of the peptide. Iavarone, Jurchen and Williams found that for cytochrome c and myoglobin, changing the 50% solvent fraction of a 47%/50%/3% water/solvent/acetic acid solution from water to methanol to acetonitrile to isopropanol shifted the charge-state distributions to lower charge and higher m/z — an effect they described as relatively small for these solvents — while adding less than 0.4% diethylamine produced dramatic shifts to lower charge, presumably due to preferential proton transfer from the higher charge state ions to diethylamine ^[2]. Their measurements were on those two proteins; the general lesson is that the envelope reports on the spray conditions as much as on the analyte.

Counterions do the same thing. Mirza and Chait showed that certain anions in the spray solution cause a marked decrease in the net average charge of peptide and protein ions, with the propensity following the order trichloroacetate > trifluoroacetate > acetate, approximately equal to chloride, and that the effect depended only on the anion and not on whether it arrived as the acid or the salt ^[3]. They proposed an ion-pairing mechanism and concluded that any attempt to correlate observed gas-phase charge state distributions with the net charge on the protein in solution must take this anion effect into account.

This matters practically because trifluoroacetic acid is the routine mobile-phase modifier in peptide reversed-phase chromatography. Kuhlmann and colleagues presented evidence that TFA suppresses signal for basic compounds through strong ion pairing between the TFA anion and the protonated analyte cation, effectively masking the cation from the electrospray fields, while weakly basic molecules were not suppressed by this process ^[4]. They reported that postcolumn addition of a propionic acid and isopropanol solution gave signal enhancement typically 10–50 times for peptides and other small basic molecules. So a weak or shifted spectrum can mean the peptide is scarce, or it can mean the mobile phase was working against the source. Those are not distinguishable from a number on a certificate.

Monoisotopic versus average mass

Two different masses can be calculated from the same formula, and they are not interchangeable.

Monoisotopic mass sums the mass of the lightest stable isotope of each element: ¹²C at exactly 12, ¹H at 1.00783, ¹⁴N at 14.00307, ¹⁶O at 15.99491. It corresponds to a specific, physically real molecule — the one containing no heavy isotopes.

Average mass sums standard atomic weights, which are isotope-weighted: carbon 12.011, hydrogen 1.008, nitrogen 14.007, oxygen 15.999. It corresponds to no individual molecule at all; it is the centroid of the whole isotope distribution.

The gap is dominated by carbon, and it scales with size. Per atom, the average-minus-monoisotopic difference is 0.011 for carbon, 0.00393 for nitrogen, 0.00409 for oxygen and 0.00017 for hydrogen. For a peptide with formula C₁₃₀H₂₀₀N₄₀O₄₀, the arithmetic is 130 × 0.011 = 1.430, plus 200 × 0.00017 = 0.034, plus 40 × 0.00393 = 0.157, plus 40 × 0.00409 = 0.164, for a total of about 1.785 Da. That is far larger than the tolerance of any high-resolution measurement. A reported mass that is silent about which convention it uses is ambiguous by about 1.8 Da at this size, which is the same order as several real modifications.

Which one you can actually measure depends on resolution and on size. As mass rises the monoisotopic peak shrinks relative to the rest of the envelope. Lermyte and colleagues, working on top-down proteomics, noted that the average mass is usually used in that context but is known to fluctuate significantly from both natural and technical causes, and that the monoisotopic precursor mass, which would be preferable, falls below the detection limit for all but the smallest proteins ^[5]. Their solution was to predict the monoisotopic mass from the exact mass of the most-abundant aggregated isotope peak using linear models, typically accurate to a few parts per million, with an associated confidence measure to handle off-by-one-dalton errors. Research peptides sit well below the size range where that problem becomes acute, but the direction of the effect is the same, and the off-by-one-dalton failure mode is worth knowing about because it lands close to the size of a deamidation: replacing an amide NH₂ with OH changes the monoisotopic mass by 15.99491 + 1.00783 − (14.00307 + 2 × 1.00783) = 0.98401 Da, against the 1.00335 Da spacing of a carbon-13 peak, so the two are indistinguishable at unit resolution and are told apart only by a measurement able to see the 0.01934 Da between them.

What mass accuracy can and cannot settle

Higher accuracy narrows the set of formulas consistent with an observation, and at a few parts per million it starts to separate residues that are nominally the same. Takach and colleagues demonstrated MALDI-TOF measurements with delayed extraction at errors below 5 ppm across a mixture of twelve peptides spanning roughly 900 to 3700 Da, and reported that it is possible to distinguish lysine from glutamine in peptides on that basis ^[6]. The two residues differ by CH₄ versus O: 12.000 + 4 × 1.00783 − 15.99491 = 0.0364 Da, a separation of about 28 ppm at 1300 Da, which a measurement accurate to 5 ppm separates comfortably and a unit-resolution measurement cannot approach.

Some things accuracy cannot settle at any resolution. Leucine and isoleucine have identical elemental composition and identical mass. Any two peptides built from the same residues in a different order have the same mass. The glossary entry on mass spectrometry makes the same point about sequence permutations, and it is the reason that a matching intact mass is consistent with the intended sequence rather than proof of it — fragmentation data, not a single mass, is what constrains order.

There is also a subtler collision. Ramjit and colleagues, resolving a 19-component dodecapeptide library by electrospray FT-ICR at an average resolution of 275,000, found isobaric multiplets in which the monoisotopic peak of one component overlapped the ¹³C-containing peak of another within the 1380–1385 Da and 1395–1400 Da windows, and reported that at that resolution these were mass resolved, accurately measured and identified from their computed molecular formulas ^[7]. In the same study they described chromatographic separation as of limited utility for that library, useful mainly for monitoring the overall extent of reaction and apparent product distribution.

What a purity assay alone cannot tell you

A chromatographic purity percentage is an area ratio at one detection wavelength under one gradient. It says how much of the detected material eluted in the main peak. It does not say what the main peak is, and it does not say what the other peaks are.

The scale of what hides in the remainder is worth seeing concretely. Li and colleagues at the BIPM, characterizing a synthetic human C-peptide material by LC coupled to high-resolution MS, identified, confirmed and quantified more than 65 structurally related peptide impurities, and estimated their total mass fraction at 83.3 mg/g with an expanded uncertainty of 3.0 mg/g (k = 2) ^[8]. Their framing is the important part: every purity assessment approach they list — mass balance, amino acid analysis, qNMR, nitrogen determination — requires a correction for structurally related peptide impurities in order to avoid bias. A purity number without impurity identification is an uncorrected number.

Lian and colleagues' review of LC-MS characterization of synthetic peptide materials makes the same structural point: despite their small size relative to proteins, synthetic peptides are subject to many complex modifications arising from starting materials, the manufacturing process and storage conditions, and their discussion of impurity identification by LC-MS focuses on structural isomers and stereoisomers, that is, peptide epimers ^[9].

The modifications that mass alone will not resolve

Two failure modes deserve naming because both survive an intact-mass check.

Epimerization. Zhang and colleagues, developing a chiral stationary phase HPLC method for aspartic acid epimerization in a model active-site peptide, state plainly that this modification causes no change in m/z and may therefore be overlooked during analysis ^[10]. A D-residue in place of an L-residue is a different molecule with the same formula. Mass spectrometry, on its own, is blind to it; separation on a chiral phase is what makes it visible.

Deamidation and isoaspartate formation. Deamidation of asparagine adds about one dalton, which is detectable, but it produces aspartate and isoaspartate, which are isomers of each other and therefore isobaric. Erckes and colleagues report that conventional chromatographic methods and standard mass spectrometric analyses often fail to distinguish structurally similar peptides with nearly identical physicochemical properties and masses, and evaluated CID and ETD tandem MS as a way to differentiate the resulting species without relying on chromatographic separation, with ETD further enabling semi-quantitative detection ^[11]. In their experiments, exposure to acidic conditions, and particularly to TFA as commonly used during HPLC purification, produced substantial direct deamidation by hydrolysis without detectable isoaspartate formation, and the degradation was site-dependent, with C-terminal amides markedly more susceptible in their study.

That last finding points at something uncomfortable: the analysis can create the modification it then reports. Liu and colleagues make the same argument for sample preparation, noting that deamidation artifact is easily introduced during proteolysis, that the neutral to mildly basic conditions of common digestion protocols favor it, and that performing Glu-C digestion at pH 4.5 in ammonium acetate effectively eliminated the artifact while preserving nearly identical sequence specificity compared with pH 8.0 ^[12]. A deamidation reported on a certificate may be in the material or may be in the method, and the document rarely lets you tell which.

Reading the document you actually have

Most certificates report a mass and a purity percentage and nothing else. The following are the questions a spectrum would answer and a number does not, and it is honest to treat them as unanswered rather than assume favorably:

• Is the reported mass monoisotopic or average? For the formula worked through above the two differ by about 1.8 Da, and the gap grows with the carbon count.
• How many charge states were observed, and did they agree? One peak is weaker evidence than an envelope.
• Was the isotope pattern resolved, or was the charge assumed?
• What was the mass error, in ppm or in daltons, and against which theoretical value?
• Were the minor chromatographic peaks identified, or only integrated?
• Was any fragmentation performed, or is sequence order inferred entirely from the intact mass?

What is measured and what is convention

Much of the surrounding practice is convention with thin primary literature behind it. That the intact mass is reported as monoisotopic below some size and average above it, that purity is quoted at a low-UV wavelength where the peptide bond absorbs, that a single number is considered sufficient without an accompanying spectrum — these are habits of the trade, not measured thresholds, and they vary between laboratories. The parts that are measured, and cited above, are narrower: how solvent and counterion shift a charge-state envelope, how large the monoisotopic-to-average gap is for a given formula, how many impurities a careful characterization of one synthetic peptide material found, and which modifications survive an intact-mass check.

What is genuinely not known, from a document alone, is whether any of it was done. A mass consistent with the intended sequence is real evidence and worth having. It is also the weakest form of identity evidence that gets called identity evidence, and knowing exactly how weak it is remains the useful thing to carry to the bench.

Echo Peptides supplies research peptides for laboratory and in-vitro research use only. Nothing here is guidance for use in humans or animals.