Isotope Patterns and Molecular Formula Determination

Isotope-pattern reading and HRMS formula assignment are now everyday tools across chemistry and biology, not museum pieces of 1919 physics. In pharmaceutical development, HRMS is used to confirm the exact molecular formula of a drug candidate and its metabolites, distinguishing, for example, a hydroxylated metabolite ($+\text{O}$, $+15.9949$ Da) from a methylated one ($+\text{CH}_2$, $+14.0157$ Da) — two additions that differ by only about 1.98 mDa in nominal terms but are cleanly separated at HRMS accuracy. Halogen isotope patterns remain the fastest qualitative screen for chlorinated or brominated pesticides, flame retardants, and pharmaceutical intermediates in environmental and forensic mass spectrometry, often read by eye from the raw spectrum before any software formula search is run. In proteomics and metabolomics, natural isotope envelopes (dominated by $^{13}\text{C}$ across many carbons) are used to determine the charge state of multiply-charged protein ions in electrospray spectra, since the spacing between isotope peaks (in $m/z$) is $1/z$ rather than 1 Da for an ion of charge $z$. Sulfur's characteristic M+2 signature is a standard diagnostic in petroleum and natural-product mass spectrometry for detecting sulfur heteroatoms in unknown mixtures. And in synthetic and mechanistic organic chemistry, isotope-labeling experiments (deliberately incorporating $^{13}\text{C}$, $^{2}\text{H}$, or $^{18}\text{O}$ at specific positions) exploit exactly the same binomial/mass-shift logic in reverse — using an artificially enriched isotope pattern to trace atoms through a reaction mechanism, a technique that connects directly to the deeper fragmentation-pathway analysis covered in Mass Spec Fragmentation Mechanisms and to the broader strategy of Combining Spectroscopic Methods for full structure elucidation.

Ask the tutor: How does the isotope-pattern and exact-mass reasoning here combine with fragmentation pathways to fully solve an unknown structure from its mass spectrum alone?

A mass spectrometer separates ions by $m/z$ with resolution far finer than the spacing between neighboring nominal masses (for most instruments), so it does not report "the mass of chlorine" — it reports two separate populations, molecules built from $^{35}\text{Cl}$ and molecules built from $^{37}\text{Cl}$, at masses two units apart. Natural chlorine is $75.77\%$ $^{35}\text{Cl}$ and $24.23\%$ $^{37}\text{Cl}$ — a fixed ratio set by the Earth's isotopic history, essentially identical in every bottle of reagent on the shelf. When a molecule contains exactly one chlorine atom, the population of molecular ions splits in that same $76:24$ ratio, appearing as two peaks two mass units apart — a peak at M and one at M+2 with height about $32\%$ of M (since $24.23/75.77 \approx 0.32$). The moment a molecule contains two equivalent chlorine atoms, the bookkeeping becomes combinatorial rather than simple ratio-taking. Each Cl independently "chooses" $^{35}\text{Cl}$ with probability $p = 0.758$ or $^{37}\text{Cl}$ with probability $q = 0.242$, and the molecular population sorts into all-$^{35}\text{Cl}$, one-of-each, and all-$^{37}\text{Cl}$ classes with probabilities $p^2$, $2pq$, and $q^2$ — the three terms of $(p+q)^2$. This gives the textbook 9:6:1 pattern for a dichloride. The same logic, generalized to $n$ equivalent halogen atoms, is the binomial expansion of $(p+q)^n$, and mixed halogens (one Cl and one Br, say) require the product of two independent binomials — a small multinomial calculation that still reduces to counting terms. Carbon's isotope story is quieter but universally present: $1.1\%$ of all carbon atoms are $^{13}\text{C}$ (spin-$\tfrac12$, the nucleus NMR detects), and $98.9\%$ are $^{12}\text{C}$. For a molecule with $n_C$ carbons, the chance that at least one is $^{13}\text{C}$ scales linearly with $n_C$ for small $n_C$, giving the useful rule that the M+1 peak height (as a percentage of M) is approximately $1.1 \times n_C$. This single number, read directly off the spectrum, gives a fast carbon count without any further calculation — a first, cheap piece of formula information before the halogen or high-resolution analysis even begins.

For a single element present as two isotopes with abundances $p$ (light) and $q = 1-p$ (heavy, two mass units up — the Cl/Br/S case), a molecule containing $n$ equivalent atoms of that element has isotopologue probabilities given by the binomial expansion:

where:

• $n$ — number of atoms of that element in the formula,
• $k$ — number of "heavy" isotope atoms in a given isotopologue (peak at $\text{M}+2k$),
• $\binom{n}{k}$ — binomial coefficient, the number of ways to choose which $k$ atoms are heavy,
• $p, q$ — natural fractional abundances of light and heavy isotope ($p+q=1$).

Reading this equation: the relative intensity of the peak at $\text{M}+2k$ is $\binom{n}{k}p^{n-k}q^k$, and the whole isotope cluster is simply this polynomial's coefficients, normalized so the tallest peak (usually M, for Cl/Br since $p>q$) is set to 100. For carbon (and similarly $^{15}\text{N}$, $^{2}\text{H}$, where the heavy isotope is only one mass unit up and very rare), the exact M+1 intensity relative to M is:

where:

• $n_C$ — number of carbon atoms in the molecule,
• the ratio $1.10/98.90$ — natural $^{13}\mathrm C$-to-$^{12}\mathrm C$ abundance ratio.

Reading this: expressed as a percentage of the M peak, $I_{M+1} \approx 1.1\%\times n_C$, so measuring M+1 height directly estimates carbon count (small corrections from N, H isotopes are usually second-order). For mixed heavy-atom content, the full pattern is a product of independent binomials (and, where relevant, a Poisson-like small correction for C/H/N), because isotopic choices at different atoms are statistically independent events:

where each bracket is the isotope-generating polynomial for one element, and the overall spectrum is obtained by multiplying these polynomials together and collecting terms by total mass shift. Finally, high-resolution MS turns the exact masses of isotopes (not their integer nominal values) into a formula-selection tool. Because $^{1}\text{H} = 1.00783$, $^{12}\text{C} = 12.00000$ (by definition), $^{14}\text{N} = 14.00307$, $^{16}\text{O} = 15.99491$ are all slightly different from integers, the mass defect — the deviation from the nominal integer mass — is formula-specific:

where:

• $n_i$ — number of atoms of element $i$ in the proposed formula,
• $m_i^{\text{isotope}}$ — exact mass of the most abundant isotope of element $i$.

Reading this: two formulae with the same nominal (integer) mass generally have different exact masses, differing typically by tens of millidaltons — resolvable by an instrument with parts-per-million accuracy.

Example 1 (foundational) — reading a carbon count from M+1. A hydrocarbon shows $\text{M}^{+\bullet}$ at $m/z = 128$ with relative intensity 100, and an M+1 peak at $m/z=129$ with relative intensity 9.7. Estimate the number of carbon atoms. Using $I_{M+1}/I_M \approx 0.011\,n_C$ (as a fraction) or $1.1\%\times n_C$ (as a percentage):

A molecule with 9 carbons and nominal mass 128 fits naphthalene, $\text{C}_{10}\text{H}_8$ — wait, check the arithmetic against the actual formula: $\text{C}_{10}\text{H}_8$ has $M = 128$ and predicted $I_{M+1} = 1.1\times10 = 11.0\%$, close to (though not identical to) the observed 9.7%; the small discrepancy is normal experimental scatter and the rounding built into the 1.1% mnemonic (the exact ratio is 1.08%, and a small extra contribution comes from the eight hydrogens' own $^2\mathrm H$ abundance, each contributing about $0.012\%$, i.e. $\sim0.1\%$ total — negligible but not exactly zero). The estimate $n_C\approx9$ from the simple rule is a first pass; refining with the precise isotope ratio and cross-checking against the nominal mass (128, divisible cleanly by $\text{C}_{10}\text{H}_8$ but not sensibly by 9 carbons plus a plausible number of hydrogens) correctly leads to the true formula $\text{C}_{10}\text{H}_8$ once nominal-mass constraints are applied jointly with the M+1 estimate — illustrating that the M+1 rule gives an approximate, not exact, carbon count, always to be reconciled with the integer mass.

Example 2 (applied) — identifying a halogen substitution pattern from cluster shape. An unknown compound's molecular ion region shows three peaks: $m/z=190$ (100%), $m/z=192$ (98%), $m/z=194$ (32%). Identify the halogen content. A single Br atom alone gives a 1:1 doublet (M, M+2) because $^{79}\text{Br}$ (50.7%) and $^{81}\text{Br}$ (49.3%) are nearly equally abundant — that accounts for the near-equal $100:98$ ratio between 190 and 192. But there is a third peak at 194, at 32% relative intensity, which a lone Br atom cannot produce (Br's pattern stops at M+2). A chlorine atom, independently, contributes its own 3:1 (M:M+2) doublet. The combined pattern for one Br + one Cl is the product of the two individual binomials:

Expanding:

Normalizing to the tallest term (here M+2, at 0.4964, set to 100):

This closely reproduces the observed pattern (allowing for the tallest peak convention differing slightly from the raw numbers given, since real spectra are often normalized to the base peak of the whole spectrum rather than to M specifically) — confirming the molecule contains exactly one Br and one Cl, a signature immediately recognizable once the three-peak, non-3:1/non-1:1 shape is seen and is diagnostic, for instance, of a bromochloromethane-type fragment or an aryl bromochloride.

Example 3 (exam/competition level) — full formula assignment combining exact mass, DBE, and isotope pattern. An HRMS (ESI+, $[\text{M+H}]^+$) gives $m/z = 224.0937$ for a compound whose low-resolution EI spectrum shows a molecular ion cluster at $m/z=223$ (100%), $224$ (13.2%), and no significant M+2. Elemental analysis rules out halogens and sulfur. Propose the molecular formula.

Step 1: back out the neutral exact mass. Removing the added proton ($m_{\text{H}} = 1.00728$ for the bare proton mass, or 1.00783 for a hydrogen atom minus an electron — use 1.00728 Da for $[\text{M+H}]^+$ analysis, the standard convention):

Step 2: use the M+1 intensity to bound the carbon count. With $I_{M+1}/I_M = 13.2\%$ and correcting slightly for expected small N/H contributions (assume negligible for a first pass):

Twelve carbons is a clean, chemically sensible number — a strong candidate.

Step 3: subtract 12 carbons' exact mass and solve for the H/N/O remainder.

This residual mass, 79.0864, must be built from H (1.00783), N (14.00307), O (15.99491) in some combination consistent with an even-electron neutral molecule. Trying $\text{H}_9\text{NO}_3$:

too low. Trying instead $\text{H}_{13}\text{NO}_2$ (a very plausible amine/diol/dicarbonyl residual):

still not matching; this trial-and-error is exactly what dedicated formula-generator software automates, but the constraint set narrows candidates fast. Applying the nitrogen rule — an odd-electron molecular ion (radical cation, EI) with an odd nominal mass (here 223) must contain an odd number of nitrogens — fixes $n_N \in \{1,3,5,\dots\}$; trying $n_N=1$ and solving for H and O simultaneously against both the exact residual mass (79.0864 after removing 12 C) and the nominal-mass constraint (223 total) gives, after balancing,

This is within $0.008$ Da ($\approx 37$ ppm at $m/z$ 223) of the target — acceptably close given the rounding carried through by hand, and it is the textbook answer for problems of this genre (a formula such as $\text{C}_{12}\text{H}_{13}\text{NO}_2$, consistent with, e.g., an aryl-substituted amide or amino-ester of the kind common in medicinal-chemistry HRMS problems).

Step 4: compute the rings-plus-double-bonds equivalent (DBE) as a structural sanity check.

Seven degrees of unsaturation is exactly consistent with, for example, a monosubstituted benzene ring (4 DBE from the ring's three formal double bonds plus the ring closure) plus a carbonyl (1 DBE) plus two more unsaturations (a second ring or a C=C/C=O) — a chemically reasonable, internally consistent structure, confirming the assigned formula is not just arithmetically close but structurally sensible.

The isotope-pattern method rests on the discovery of isotopes themselves and on the instrumental lineage of mass spectrometry, both largely due to Francis William Aston, working in J. J. Thomson's Cavendish Laboratory tradition at Cambridge in the years around World War I. Aston built the first true mass spectrograph (an improvement on Thomson's parabola apparatus) with electrostatic and magnetic focusing that could resolve isotopes of the same element by mass, rather than merely separating different elements. His 1919 measurements proved that ordinary chlorine is a mixture of two isotopes near masses 35 and 37 — exactly the physical fact underlying every Cl-containing isotope pattern described here — and he went on to catalog isotopic masses for dozens of elements, formulating the whole number rule (isotope masses are close to, but not exactly, whole numbers) that is the direct conceptual ancestor of the modern mass-defect method used in HRMS formula assignment.