Valence Bond Theory and Orbital Overlap

Valence bond overlap reasoning remains the everyday mental shorthand chemists use to predict and rationalize structure before reaching for computation. In organometallic and coordination chemistry, the directional overlap of metal d orbitals with ligand donor orbitals underlies the qualitative prediction of which geometries (square planar, octahedral, tetrahedral) give the strongest metal–ligand bonding for a given d-electron count, feeding directly into crystal field and ligand field reasoning used to rationalize catalyst design. In main-group and materials chemistry, overlap arguments explain why heavier p-block elements (Si, P, S versus C, N, O) show weaker, longer bonds and a documented reluctance to form strong π bonds — the "double bond rule" breakdown for heavier elements traces directly to poor sideways overlap between large, diffuse 3p (or higher) orbitals, a fact exploited deliberately in the synthesis of stable disilenes and diphosphenes using bulky substituents to protect the intrinsically weak π overlap. In drug design and enzyme mechanism, the geometric logic of maximum overlap underlies why catalytic transition states demand precise bond angles and orbital alignments — enzymes achieve rate acceleration partly by pre-organizing substrates into the orbital geometry that maximizes productive overlap along the reaction coordinate, a concept traceable to the same overlap-integral reasoning used for H₂. In spectroscopy and bonding analysis software, modern computational packages (natural bond orbital, or NBO, analysis) explicitly decompose delocalized molecular orbitals back into localized VB-like bonding pairs with associated overlap and hybridization percentages, making Pauling-era VB language directly recoverable from state-of-the-art quantum chemical calculations — a striking continuity between a 1927 hand calculation and a modern density-functional-theory output file.

Ask the tutor: How does adding hybridization change the overlap integral compared to using pure atomic orbitals, and why does that let sp³ carbon achieve stronger bonds than unhybridized p orbitals could?

Start with two isolated hydrogen atoms, each in its 1s ground state, infinitely far apart. Each has one electron. As the atoms approach, their 1s orbitals begin to overlap — the electron clouds interpenetrate in the region between the nuclei. What happens to the total energy depends critically on the relative spin orientation of the two electrons. If the electrons have opposite spins (a singlet state), quantum mechanics allows both electrons to occupy the same spatial overlap region with high probability. This buildup of electron density between the nuclei electrostatically screens the two positively charged nuclei from each other while simultaneously attracting both nuclei toward the shared density. The net effect, worked out rigorously by Heitler and London in 1927, is a potential energy curve with a genuine minimum: the energy drops below that of the separated atoms as the nuclei approach, reaches a minimum at the equilibrium bond length $r_e$, and rises steeply at shorter distances due to nuclear-nuclear and core-core repulsion. If the electrons have parallel spins (a triplet state), the Pauli exclusion principle forbids them from occupying the same spatial region with appreciable probability — this creates a "Fermi hole" that keeps the electrons apart, actually reducing density between the nuclei. The resulting curve is repulsive at all distances: no bond forms. This spin dependence is the deep reason the Lewis picture insists on paired electrons in a bond — pairing is not just bookkeeping, it is the condition under which the spatial wavefunction is symmetric and overlap-enhancing, given that the total (space × spin) wavefunction must be antisymmetric under electron exchange for two fermions. The key quantity controlling all of this is the overlap integral,

which measures the degree to which the two atomic orbitals $\psi_A$ and $\psi_B$ occupy the same region of space with matching phase. When $S$ is large and positive, constructive overlap builds density between the nuclei and stabilizes the bonded state; when the orbitals are oriented so that a positive lobe of one meets a negative lobe of the other, $S$ becomes negative and the interaction is destabilizing (antibonding-like in character, even within the VB framework). When the orbitals point away from each other or belong to symmetry-mismatched sets, $S \to 0$ and no bond forms regardless of energy matching. This is why atomic orbital shape dictates bond direction. An s orbital is spherically symmetric, so it can overlap effectively with an orbital pointing toward it from any direction — s–s bonds have no directional preference of their own. A p orbital, by contrast, has two lobes of opposite phase along a single axis; it overlaps well only with orbitals approaching along that axis, and the overlap is zero (by symmetry) with an orbital approaching exactly perpendicular to it, since the perpendicular approach samples equal amounts of the positive and negative lobe and they cancel. This single geometric fact is the reason p-orbital-based bonding in molecules like H₂S and PH₃ gives bond angles close to 90° rather than the tetrahedral 109.5° that a naive "carbon-like" guess might suggest — the three 2p (or 3p) orbitals of S or P point along mutually perpendicular axes, and each bonds independently to a hydrogen 1s orbital along its own axis.

The overlap integral also explains bond-strength trends down a group. Compare H–F, H–Cl, H–Br, H–I. Fluorine's bonding orbital is a compact 2p orbital; chlorine's is a more diffuse 3p orbital; bromine's is 4p, iodine's 5p. As the halogen orbital becomes larger and more diffuse, it still overlaps with the small, tight hydrogen 1s orbital, but the quality of overlap — how much of the diffuse orbital's density is concentrated in the internuclear region relative to its total extent — degrades, and the bond length increases while the bond dissociation energy falls (H–F 565 kJ/mol at 92 pm; H–I 299 kJ/mol at 161 pm). The mismatch in orbital size between H(1s) and the heavier halogen p orbital is compounded by a growing mismatch in orbital energy, since ionization energies (a proxy for orbital energy) fall down the halogen group; both effects push in the same direction, toward weaker bonding. Energy matching is the second pillar of VB overlap, alongside symmetry matching. Two orbitals will mix effectively only if their energies are comparable — if one orbital is much lower in energy (more tightly bound, more electronegative environment) than the other, the bonding combination remains dominated by the lower-energy orbital and the "bond" becomes strongly polarized, shading toward an ionic description. This is qualitatively why H–F is more polar than H–I: the 2p orbital on F sits much lower in energy (F is far more electronegative) than H 1s, giving a highly polarized, largely ionic-character bond, whereas the energy match between H 1s and I 5p is closer, giving a bond with more covalent, more symmetric density.

Finally, VB overlap draws the fundamental distinction between σ bonds (electron density concentrated symmetrically along the internuclear axis, formed by head-on overlap of s, p$_z$, or hybrid orbitals) and the beginnings of π bonds (sideways overlap of parallel p orbitals, with density above and below the axis and a node in the plane containing the nuclei) — a distinction developed fully once hybridization is introduced, but whose origin is entirely a matter of geometric overlap, exactly as described here.

The total electronic energy of the VB-treated two-electron system, in the Heitler–London approach, separates into pieces with transparent physical meaning:

where:

• $E_+$ is the energy of the singlet (paired-spin, bonding) state and $E_-$ the triplet (parallel-spin, antibonding) state,
• $E_{\text{Coulomb}}$ is the classical electrostatic (Coulomb) integral — attraction of each electron to both nuclei plus electron–electron and nuclear–nuclear repulsion, evaluated as if the electrons were classical charge distributions,
• $E_{\text{exchange}}$ is the purely quantum mechanical exchange integral, arising from the indistinguishability of the two electrons; it has no classical analogue and is the dominant term responsible for bonding,
• $S$ is the overlap integral defined above, which appears in the normalization denominator because the two-electron wavefunction built from non-orthogonal atomic orbitals is not automatically normalized.

Reading: the singlet combination ($+$ sign) lowers the energy relative to the separated atoms because $E_{\text{exchange}}$ is large and negative; the triplet ($-$ sign) raises it. This is the mathematical statement of "opposite spins bond, parallel spins don't," and it shows that overlap $S$ enters not just linearly but through a self-consistency (normalization) factor that becomes important as $S$ grows large at short internuclear distance. The second governing relationship is the qualitative maximum overlap principle, which for a pair of orbitals approaching along an axis making some overlap integral $S(r,\theta)$ a function of internuclear distance $r$ and relative orientation $\theta$, states that

where:

• $r_e$ is the equilibrium bond length (found by minimizing total energy, balancing overlap-driven attraction against nuclear repulsion at short range),
• $\theta$ is the angle of approach relative to the orbital's directional axis (for a p orbital, $\theta = 0°$ means head-on approach along the lobe axis; $\theta = 90°$ means perpendicular approach).

Reading: bonds form preferentially along the directions that maximize spatial overlap. For directional p or d orbitals this immediately predicts bond angles set by the orbitals' own geometry — mutually perpendicular p orbitals predict ~90° angles — before any hybridization correction is applied.

Example 1 — Foundational: locating the H₂ bond length from a simplified overlap-energy balance. Treat H₂ with a deliberately simplified (but physically correct in trend) model: the attractive energy contribution scales with the overlap integral, which for two 1s orbitals of Bohr radius $a_0 = 52.9$ pm falls off as

(the standard 1s–1s overlap integral in atomic-orbital form), while the repulsive nuclear–nuclear and core contributions grow steeply at short range. Rather than solving the full transcendental minimization here, use the known experimental bond length $r_e = 74.1$ pm to evaluate the overlap actually achieved at equilibrium and check that it is neither trivially small nor absurdly close to unity — i.e., that "substantial but incomplete" overlap is what nature selects.

Step 1. Compute $r_e / a_0$:

Step 2. Substitute into the overlap formula:

Step 3. Evaluate $e^{-1.401} = 0.2464$, so

Step 4 — Interpretation. An overlap integral of about 0.75 (on a scale where $S=0$ is no interaction and $S=1$ would mean the two orbitals are perfectly coincident, which is unphysical for two nuclei at finite separation) confirms that H₂'s equilibrium geometry sits at a point of large but not complete overlap — exactly the regime the maximum overlap principle describes: the system moves inward as long as increasing $S$ (via $E_{\text{exchange}}$) lowers energy faster than nuclear repulsion raises it, and stops once that balance reverses. This is a genuine, checkable number, not a qualitative gesture: it shows the bonding pair is substantially delocalized between the two protons at the real, measured bond length.

Example 2 — Applied: ranking H–X bond strengths in the hydrogen halides using overlap/size arguments. Given the bond dissociation energies (kJ/mol) and bond lengths (pm):

Step 1 — Identify the orbital pair for each bond. Each bond uses H(1s) and a halogen valence p orbital: F(2p), Cl(3p), Br(4p), I(5p). Orbital size increases sharply down the group because each successive orbital has a higher principal quantum number and more radial nodes/shielding.

Step 2 — Apply the overlap argument. A larger, more diffuse halogen p orbital has its electron density spread over a bigger volume; a fixed-size hydrogen 1s orbital (radius always set by $a_0 = 52.9$ pm, since H's electron always sees the same nuclear charge) captures a shrinking fraction of that diffuse density as the halogen orbital grows. Quantitatively, this is why the equilibrium bond length grows monotonically (92 → 127 → 141 → 161 pm) — the diffuse orbital's "sweet spot" of maximum overlap-per-unit-energy-cost sits farther out — while the achievable peak overlap density in the internuclear region falls.

Step 3 — Cross-check against bond energy. Bond dissociation energy falls monotonically in step with the falling overlap quality: 565 → 431 → 366 → 299 kJ/mol. The percentage drop from H–F to H–I is

nearly a factor of two weakening, tracking the near-doubling of bond length (92 pm → 161 pm, a 75% increase) — consistent with, though not numerically identical to, the general rule that overlap-driven bond strength degrades faster than linearly with increasing bond length, because $S(r)$ itself falls off exponentially with $r$.

Step 4 — Limiting-case sanity check. As the halogen's valence p orbital becomes arbitrarily diffuse and low-lying in energy relative to H(1s) (an extrapolation past iodine, hypothetically toward astatine or beyond), the overlap-based argument predicts continued weakening and lengthening — which is exactly the observed trend continuing to H–At (weaker still, though radioactive and hard to measure precisely). Conversely, in the opposite limit of a hypothetical "orbital exactly matched to H(1s) in both size and energy," the model predicts the strongest, shortest possible bond — realized, in fact, by H₂ itself (H–H: 74 pm, 436 kJ/mol), which is stronger and shorter than any H–halogen bond, precisely because there the two overlapping orbitals are identical in size and energy by definition. This limiting case is a genuine check on the logic, not a coincidence: identical orbitals overlap best.

Example 3 — Exam/competition level: predicting relative bond angles in AH₂ hydrides of Group 16 using pure p-orbital overlap, then correcting for lone-pair repulsion. A problem: gas-phase bond angles are H₂O 104.5°, H₂S 92.1°, H₂Se 91.0°, H₂Te 90.0°. Using only the maximum-overlap-of-unhybridized-p-orbitals model, explain the trend and estimate how much "extra opening" beyond the 90° baseline is attributable to s–p mixing (partial hybridization) in each case, given that s–p energy separation for the central atom is the controlling variable.

Step 1 — Baseline prediction. If each A–H bond used a pure, unhybridized p orbital on the central atom (A = O, S, Se, Te), the maximum-overlap principle (Step 2 of the derivation above) predicts exactly 90° for every member of the series, since all np orbitals are mutually perpendicular regardless of $n$.

Step 2 — Identify the deviation for each hydride.

Step 3 — Correlate with s–p energy gap. The central-atom valence $ns$–$np$ energy gap widens sharply going up the group (O: 2s/2p gap large because the small, tightly held O 2s orbital is strongly stabilized by poor shielding from the small 2p subshell; by S, Se, Te the ns orbital is more shielded and the ns–np gap narrows, and additionally the larger np orbitals are themselves more diffuse and less "pure" p in character). A small s–p gap favors mixing of some s character into the bonding orbitals (a preview of hybridization), which pulls the bond angle away from 90° toward the s-orbital's isotropic (angle-independent) character — and pure sp³ mixing in the tetrahedral limit would give 109.5°.

Step 4 — Interpret the trend as an "incomplete hybridization" ladder. Oxygen shows the largest deviation ($\Delta\theta = 14.5°$) because its small 2s–2p gap allows substantial s character to mix into the bonding hybrids, pulling the angle most of the way toward 109.5°. Sulfur, selenium, and tellurium show progressively smaller deviations (2.1°, 1.0°, 0.0°) because their larger, more shielded ns orbitals sit at increasingly different energies from their np orbitals (or, in a fully rigorous MO-based view, s–p mixing becomes less favorable as the orbitals delocalize and their overlap with the "wrong" symmetry combination weakens), so the bonding orbitals stay closer to pure, unhybridized p character, and the angle relaxes toward the 90° baseline set by orbital geometry alone.

Step 5 — Limiting case. In the limit of an idealized central atom with zero s–p mixing at all (an infinite s–p gap), the model predicts exactly 90° with no exceptions — the H₂Te value of 90.0° is the empirical realization of this limit, serving as the calibration point that validates the whole ladder: it shows the "pure p-orbital" baseline is not just a theoretical idealization but an experimentally attained limit.

Valence bond theory's quantitative foundation was laid by Walter Heitler and Fritz London, whose 1927 quantum mechanical treatment of the hydrogen molecule was the first genuinely successful ab initio explanation of a covalent bond — appearing barely a year after Schrödinger's wave equation itself. Their calculation, using exactly the symmetric/antisymmetric two-electron wavefunctions built from atomic orbitals as constructed in the derivation above, reproduced a bonding energy curve with a real minimum for paired spins and a purely repulsive curve for parallel spins, settling — with actual numbers — the question of why Lewis's shared electron pairs (proposed on pure chemical intuition in 1916, well before quantum mechanics) worked as well as they did. The theory was extended into a full working framework for polyatomic molecules chiefly by Linus Pauling, who introduced the maximum overlap principle explicitly as a guiding rule for bond directionality and combined it with the hybridization concept (built from the same orbital-overlap logic used here) to explain the tetrahedral carbon, the geometry of unsaturated systems, and — through resonance — the equalized bond lengths of benzene and the carboxylate/carbonate ions. Pauling's 1931 papers and his landmark book The Nature of the Chemical Bond (1939) turned Heitler and London's two-electron calculation into the qualitative, structure-predicting toolkit that remains standard in undergraduate chemistry.