The Mass Spectrum and the m/z Axis

Every commercial spectral database — NIST/EPA/NIH mass spectral library, Wiley Registry, METLIN for metabolomics — stores compound identities exactly as this article has framed them: a list of (m/z, relative abundance) pairs, base peak normalized to 100%, molecular ion flagged separately when observed. Automated library-search algorithms used in forensic toxicology screening, doping-control laboratories, and environmental contaminant analysis compute similarity scores by comparing an unknown's normalized stick spectrum against thousands of reference spectra stored in precisely this format. In pharmaceutical quality control, LC-MS/MS methods used to confirm drug identity and purity report both the molecular ion (often as a protonated adduct [M+H]⁺ in soft ionization) and characteristic fragment ions, with regulatory guidance (such as that from the FDA and EMA) explicitly requiring the analyst to state which peak is the base peak and which corresponds to the intact analyte ion, avoiding exactly the confusion this article warns against. In proteomics and intact-protein mass spectrometry, multiply-charged ions generated by electrospray ionization populate an entire m/z axis with a "charge-state envelope" — a series of peaks at different m/z all corresponding to the same protein molecule, differing only in how many protons it happens to be carrying. Deconvolution software untangles this envelope back into a single neutral molecular mass, a direct and more elaborate descendant of the m/z-versus-mass distinction introduced here. Portable and benchtop GC-MS instruments used in field forensics, arson investigation, and airport security screening rely on exactly the normalized-stick-spectrum display convention described above, because it is the only representation robust enough to survive comparison against a stored library regardless of the specific instrument's sensitivity or the amount of sample recovered from a crime scene or checked bag.

Ask the tutor: In my spectrum, the tallest peak isn't at the highest m/z value I can find — which one should I call the molecular ion, and how do I justify that choice?

Picture the ion source of an electron-ionization mass spectrometer, described mechanically in The Mass Spectrometer: Components and Workflow. A beam of ~70 eV electrons strikes gas-phase molecules of the analyte and knocks out one electron, producing a radical cation:

This species, called the molecular ion and given the symbol M⁺• (the dot marks the unpaired electron, the plus marks the missing electron's charge), carries essentially the full mass of the neutral molecule, now short one electron — an amount so small (0.00055 u) that it is ignored in ordinary integer-mass bookkeeping. The 70 eV impact deposits far more energy than a single C–C bond requires to break (roughly 3–4 eV), so a large fraction of molecular ions do not survive intact. They fragment, in ways that reflect the molecule's connectivity, into smaller ions and neutral pieces. Every ion produced — whether the surviving molecular ion or one of its fragments — is then accelerated, sorted by m/z in the mass analyzer, and struck against a detector that outputs a current or pulse count proportional to how many ions of that particular m/z arrived. The raw output is an ion current recorded as a function of m/z: a curve or, at typical unit-mass resolution, a set of discrete counts at integer m/z values. Two independent numbers describe each peak. First, its position on the x-axis — the mass-to-charge ratio itself, dimensionless by IUPAC convention (mass in atomic mass units divided by an integer charge number, itself dimensionless, though most working chemists still say "m/z = 91" as if it carried units of mass). Second, its height on the y-axis, which by long convention is rescaled relative to the single most abundant ion in the entire spectrum. That most-abundant ion, whatever its mass, defines the base peak and is assigned exactly 100% relative abundance; every other peak's height is reported as a percentage of that reference. This normalization is what allows spectra from different instruments, different ion currents, and different sample amounts to be compared directly — a spectrum plotted at 10⁻¹² A full-scale and one plotted at 10⁻⁹ A full-scale look identical once both are rescaled to base peak = 100%. It also means relative abundance carries no information at all about how much sample was present; a barely-detectable trace and a saturating overload give visually similar bar-chart shapes, though real-world dynamic range and detector saturation do impose practical limits.

Look at the figure's arithmetic carefully: the base peak at m/z 43 is set to 100% by definition, whatever its actual ion count was. The molecular ion at m/z 86 stands at only 22% relative abundance — smaller than the base peak, smaller even than the fragment at m/z 58 (30%) — yet it is still the molecular ion, because "molecular ion" is a statement about chemical identity (the intact molecule, radical-cation form) and mass, not about height. This is the single most common point of confusion for a first-time reader of a spectrum, and it is addressed head-on in the worked examples.

The mass-to-charge ratio is formally defined as a ratio of two integers (mass number in atomic mass units, charge number in units of elementary charge):

where: m is the ion's mass expressed in unified atomic mass units (u, numerically ≈ its nominal or monoisotopic mass), and z is the (positive integer, almost always 1 in routine EI spectra) number of charges the ion carries.

Reading: because z = 1 for the vast majority of ions formed by electron ionization, m/z is numerically equal to the ion's mass in daltons in nearly every spectrum an undergraduate will meet; the distinction becomes essential only for multiply-charged ions (electrospray of proteins, discussed elsewhere) where a single physical mass produces several different m/z peaks. Relative abundance is defined by a normalization against the tallest peak in the spectrum:

where: $I_i$ is the ion current (or integrated ion count) measured at a given m/z value $i$, and $I_{\text{base peak}}$ is the ion current at the single most abundant m/z in the spectrum — by construction the largest $I$ value in the whole data set, so the base peak's own relative abundance is always exactly 100%.

Reading: every other bar in the spectrum is a fraction of the tallest bar; this is why relative-abundance spectra look identical whether the instrument's raw ion current was measured in picoamps or nanoamps — the normalization divides that instrument-dependent scale factor clean away. The nominal mass of the molecular ion, when a molecule of known elemental formula $\text{C}_c\text{H}_h\text{N}_n\text{O}_o\ldots$ loses one electron, is essentially the sum of the most abundant isotope masses:

where: each coefficient is the exact mass of the most abundant naturally occurring isotope of that element (¹²C, ¹H, ¹⁴N, ¹⁶O, …), and the sum is taken over all atoms in the molecular formula, with the mass of the missing electron (0.00055 u) conventionally neglected at this level of precision.

Reading: this equation is the bridge between a structural formula on paper and a single number on the m/z axis — it is why a chemist can predict, before ever running the sample, roughly where the molecular ion peak should appear.

Example 1 (foundational) — building the normalized spectrum from a raw data table. A low-resolution EI mass spectrum of methyl butanoate (CH₃CH₂CH₂CO₂CH₃, MW 102) is recorded, and the electron multiplier reports the following peak-area currents (arbitrary units, a.u.), reading directly off the instrument's data system:

Step 1. Scan for the maximum raw current: 6,900 a.u. at m/z 71. This is the base peak, corresponding structurally to loss of •OCH₃ from the molecular ion (an acylium ion, CH₃CH₂CH₂C≡O⁺). Step 2. Normalize every entry against 6,900:

Step 3. Identify the molecular ion. Its m/z must equal the compound's molecular weight — here 102, matching the given formula's nominal mass ($5\times12 + 10\times1.00783 + 2\times15.99491 = 60 + 10.078 + 31.990 = 102.07$ u, nominal 102). So the peak at m/z 102, at 22.0% relative abundance, is M⁺•. Step 4. State the answer with correct terminology: the base peak is at m/z 71 (100%), the molecular ion is at m/z 102 (22.0%), and these are two different peaks — a result that is completely typical of esters, which fragment efficiently at 70 eV and rarely give a molecular ion taller than several of their fragments.

Example 2 (applied) — using base-peak identity as a first clue in unknown identification. A quality-control chemist runs GC-MS on a suspected batch of acetophenone (C₆H₅COCH₃, MW 120) contaminated with an unknown side product. The total ion chromatogram shows two resolved peaks; the mass spectrum under the second chromatographic peak shows a base peak at m/z 105 (100%), a moderate peak at m/z 77 (65%), a small peak at m/z 51 (18%), and a peak at m/z 120 (38%). Step 1. Interpret m/z 120 first, since it is a candidate molecular ion: it matches acetophenone's nominal mass exactly ($8\times12 + 8\times1.00783 + 15.99491 = 96+8.063+15.995=120.06$ u, nominal 120), so tentatively M⁺• = 120. Step 2. Interpret m/z 105 as loss of a 15-mass-unit fragment from the molecular ion:

producing the resonance-stabilized benzoyl cation, C₆H₅C≡O⁺, at m/z 105 — chemically reasonable as the dominant fragmentation channel, which is exactly why it is the base peak rather than the molecular ion. Step 3. Interpret m/z 77 as loss of carbon monoxide from the benzoyl cation:

giving the phenyl cation, C₆H₅⁺, at m/z 77 — a classic, textbook-recognizable fragment. Step 4. Interpret m/z 51 as loss of acetylene (C₂H�2, 26 u) from the phenyl cation:

Step 5. Conclude: every peak is self-consistently explained as a stepwise fragmentation of acetophenone (120 → 105 → 77 → 51), so the second chromatographic peak is authentic acetophenone, not a contaminant, and the base peak (m/z 105, 100%) is the acylium fragment, not the molecular ion (m/z 120, only 38%) — precisely the terminology distinction the article opened with, now applied to solve a real identification problem.

Example 3 (exam/competition level) — reconstructing an incomplete normalized spectrum and testing the ceiling constraint. A partially transcribed spectrum lists four peaks with the following relative abundances: m/z 27 (38.5%), m/z 55 (100%, "base peak" labeled by the original analyst), m/z 70 (61.9%), and m/z 98 ($x$%, illegible in the notebook). The chemist recalls that the raw ion currents, before normalization, were 640 a.u., 1,664 a.u., 1,030 a.u., and 292 a.u. respectively, for m/z 27, 55, 70, and 98, in that order. Determine $x$, and verify the "base peak" label was assigned correctly. Step 1 — verify the base peak claim independently of the label. Compare the four raw currents directly: 640, 1664, 1030, 292. The maximum is 1664 a.u., occurring at m/z 55 — so the base-peak assignment is correct; m/z 55 does have the largest raw ion current among the four listed peaks. Step 2 — recompute the two given percentages as an internal consistency check on the normalization constant.

Both match the notebook's transcribed values to the precision given, confirming 1,664 a.u. is indeed the correct normalization denominator throughout this data set. Step 3 — solve for the missing percentage at m/z 98.

Step 4 — sanity-check the answer against the ceiling rule. Since 292 a.u. < 1664 a.u., $x$ must lie strictly between 0% and 100%; 17.55% satisfies this comfortably, with plenty of margin, so no transcription error is indicated. Step 5 — identify m/z 98 chemically, as a bonus check. If this spectrum belongs to cyclohexanone (C₆H₁₀O, nominal mass $6\times12+10\times1.00783+15.99491 = 72+10.078+15.995=98.07$, nominal 98), then m/z 98 is the molecular ion itself, appearing here at 17.55% relative abundance — smaller than three separate fragment peaks (m/z 27, 55, and 70, the ring-opening/CO-loss fragments characteristic of cyclic ketones), once again illustrating that the molecular ion is frequently not the tallest peak in the spectrum, exactly as flagged in the very first figure of this article.

The vocabulary and conventions used throughout this article — m/z, the normalized base-peak scale, the very idea of plotting a "spectrum" of masses — trace back to the parallel invention of mass spectrography by Francis William Aston and, earlier, the foundational work of J. J. Thomson on positive rays. But the person most responsible for turning mass spectrometry from a physicist's tool for measuring isotope masses into a chemist's routine structural technique was Fred W. McLafferty, whose systematic cataloguing of fragmentation behavior (and the rearrangement that bears his name) established the interpretive logic — reading peak-to-peak mass losses as chemical bond cleavages — that Example 2 and Example 3 above rely on directly.