Why Some Atoms Bond and Others Don’t
Plain-language explainer for doi:10.5281/zenodo.21612627 (#710)
The mystery hiding in plain sight
Every chemistry student learns the HOMO-LUMO gap. HOMO is the Highest Occupied Molecular Orbital — the last seat filled on the electron bus. LUMO is the Lowest Unoccupied Molecular Orbital — the first empty seat. The gap between them controls whether a molecule will react: a big gap means unreactive, a small gap means chemically active.
What chemistry students are not told is why the gap varies so wildly across the periodic table. Noble gases have enormous gaps. Carbon has a moderate gap with a strong directional character. Transition metals like iron and cobalt have gaps so small that electrons can hop between oxidation states with ease. There is no unified account of this variation from elementary quantum numbers alone.
This paper provides one.
Winding numbers: a one-line summary of every orbital
Every electron orbital is labelled by two quantum numbers: n (the shell, 1, 2, 3, …) and ℓ (the angular momentum subshell: 0 for s, 1 for p, 2 for d, 3 for f).
This paper defines a single quantity:
w(n, ℓ) = n + ℓ + 1
This is called the winding number of the orbital. It counts how the orbital wraps around the Fock sphere — the hidden S³ symmetry of hydrogen that underlies the entire periodic table. The s-subshell of shell n has winding number n+1. The p-subshell has n+2. And so on.
Orbitals with the same winding number form a winding shell — a natural grouping that cuts across the conventional n-shells. The first winding shell (w=2) contains only 1s. The second (w=3) contains 2s and 2p. The third (w=4) contains 3s, 3p, and 4s. The fourth (w=5) contains 3d, 4p, and 5s. The fifth (w=6) contains 4d, 5p, 6s, and 4f — the same shell that contains all the transition metals and the entire lanthanide series.
The Madelung filling rule — the empirical recipe students memorise for filling orbitals in order — is simply electrons filling in order of increasing w.
The frontier winding gap
For any atom with atomic number Z, you can read off the winding numbers of its HOMO and LUMO from a table. The paper defines:
ΔwF = w(LUMO) − w(HOMO)
This one number — the frontier winding gap — stratifies all of chemistry into three universal regimes.
Regime 1: ΔwF ≥ 2 — noble gas inertness
For helium, neon, argon, krypton, xenon, and radon, the HOMO and LUMO sit in completely different winding shells, separated by two or more levels. There is no mechanism by which the two shells can mix. You cannot coax a noble gas into bonding because the topological gap is too large — the shells simply cannot reach each other.
This is not an energy argument. It is a structural one. The shells are on different rungs of the winding ladder.
Regime 2: ΔwF = 1 — directional covalent bonding
For the main-group elements — carbon, nitrogen, oxygen, the halogens — HOMO and LUMO are in adjacent winding shells. They can mix, but only in one direction, and only by a fixed amount. This produces directional bonds: the sp³ tetrahedral geometry of methane, the sp² planar geometry of ethylene, the sp linear geometry of acetylene. Valence is fixed. Carbon is always 4-valent. Nitrogen is always 3-valent (or 4, with a formal charge). The ladder has a definite step size.
This regime is standard organic chemistry. All of the molecules of life — proteins, sugars, lipids, DNA — are built from ΔwF = 1 atoms. The winding step enforces the specific, repeatable, programmable bonding geometry that makes molecular biology possible.
Regime 3: ΔwF = 0 — variable valence and catalysis
For the transition metals — iron, cobalt, copper, molybdenum, manganese, nickel — HOMO and LUMO live in the same winding shell. The frontier electrons can mix freely within the shell. There is no topological barrier to rearranging them. This is the origin of variable oxidation states: iron can be Fe(II) or Fe(III) or Fe(IV) because the d-shell electrons can be added or removed without crossing a winding-shell boundary. The same logic explains why transition metals form coordination complexes with variable coordination numbers, why they can bind substrates and release products, and why they are catalysts.
The ΔwF = 0 regime is where biology lives. Every enzyme that carries out chemistry at a metal centre — every cytochrome, every nitrogenase, every photosystem — uses a transition metal precisely because ΔwF = 0 is the topological prerequisite for catalysis. Iron, cobalt, copper, molybdenum, manganese, and nickel are not in biology by evolutionary accident. They are there because they are the only atoms that satisfy the winding condition for variable valence. The periodic table selects them uniquely.
The anomalies: ΔwF = −1
A handful of elements — chromium, copper, gold, platinum — have ground-state configurations that depart from the naive filling order. Chromium is [Ar] 3d⁵ 4s¹ instead of [Ar] 3d⁴ 4s². Copper is [Ar] 3d¹⁰ 4s¹ instead of [Ar] 3d⁹ 4s². Gold and platinum show similar departures driven by relativistic effects.
In the frontier winding gap picture, all of these become ΔwF = −1: the HOMO has been pulled into a lower winding shell than expected, by exchange energy (for Cr, Cu) or by relativistic mass increase (for Au, Pt). The unified signal for every d-block anomaly is a single integer.
Two famous rules derived for free
Two empirical rules of chemistry follow immediately from the winding-shell structure.
The 18-electron rule. Stable transition metal complexes — the kind that serve as industrial catalysts — almost always have exactly 18 electrons in their valence shell. The rule is memorised as a counting trick, like the octet rule but for metals. The paper derives it: the w=6 winding shell contains nd, (n+1)s, and (n+1)p subshells, with capacity 10 + 2 + 6 = 18. A filled winding shell is stable. The 18-electron rule is the transition-metal analogue of the octet rule, and both are winding-shell completeness conditions.
Hückel’s 4n+2 rule. Aromatic molecules — benzene, naphthalene, pyridine — are stable when their π-electron count is 2, 6, 10, 14, … (that is, 4n+2 for integer n ≥ 0). The paper derives this as a winding-shell completeness theorem for electrons on a ring (S¹). A complete winding level on a circle holds exactly 4n+2 electrons. Aromaticity is topological stability, not resonance energy.
A reactivity map without DFT
The paper introduces the (ΔwF, ΔC) plane — where ΔC is the difference in SO(4) Casimir values between LUMO and HOMO. Together, these two numbers form a quantitative reactivity map that can be computed in seconds from nothing but a quantum number table. No density functional theory, no basis sets, no self-consistent field calculation. Just two integers read off the periodic table.
Elements cluster into distinct regions of this plane: noble gases in one corner, main-group elements in a band, transition metals in another region, and the anomalous configurations as a labelled outlier set. The plane makes visible a structure that was always latent in Mendeleev’s table but had never been made explicit.
What this changes
The HOMO-LUMO gap has been studied for a century. The idea that it might be organised by a single topological integer — one that derives from the Fock-sphere symmetry of hydrogen and not from any approximation or fitting — is new.
The practical consequence is a new way to screen for catalytic activity. If ΔwF = 0, the atom is a candidate catalyst. If ΔwF ≥ 2, it is not. If ΔwF = 1, it is a structural atom, not a functional one. This two-second calculation replaces a first-pass that would otherwise require quantum chemical software.
The deeper consequence is conceptual. Reactivity is not a continuous property smoothly varying with electron count. It is discretised by topology into three regimes, separated by integer jumps in a quantity that is computable from first principles. Chemistry has more structure than the continuous models suggest, and that structure is visible as soon as you look at the winding numbers.
See also:
- Winding-Shell Branching Rules (#711) — SO(4) → SO(3) → G branching; Walsh diagrams, Jahn-Teller, and active space selection unified by winding shells
- No Shell Game (#712) — the minimal active space AS(w₀) = {w₀−1, w₀, w₀+1} derived from winding numbers; 20/20 transition-metal benchmark
- Knotted Orbitals: Kelvin-Tait Vindicated (#657) — orbital topology as torus knots on S³; Madelung rule as torus-knot index ordering
For the full technical treatment, see doi:10.5281/zenodo.21612627