Orbital Wars: When VSEPR Theory Fails (And Why It’s Okay)

🧪 SM-EDUCATE CHEMISTRY June 07, 2026

Orbital Wars:
When VSEPR Theory Fails Spectacularly

⚛️ Electron repulsion · Expanded octets · d‑orbital participation · Inorganic deep dive

Inorganic Chemistry VSEPR Theory Molecular Geometry Chemical Bonding Expanded Octets Visual Learning

Valence Shell Electron Pair Repulsion (VSEPR) is a reliable workhorse: predict the shape of a molecule by counting electron domains and assuming they repel each other. But some molecules laugh at VSEPR. They twist, distort, or adopt geometries that defy simple electron‑pair repulsion. Here’s where the orbital wars begin — and where VSEPR loses.

VSEPR successfully predicts octahedral (SF₆) and trigonal bipyramidal (PCl₅) geometries for many main‑group compounds. But when the central atom has lone pairs or d‑electrons, the rules start to fray.

1. The Classic VSEPR Assumptions

VSEPR treats electron pairs as negative clouds that repel. It works beautifully for molecules with only s and p valence orbitals. But for transition metals and heavier main‑group elements, d‑orbitals enter the stage. Electron pairs can be stereochemically active (affecting shape) or inactive (not affecting shape). VSEPR doesn't distinguish.

_xy d_xz d_z² When d‑orbitals are partially filled, electron repulsion is not isotropic. VSEPR fails to predict bond angles.

The d‑orbitals have directional shapes. When they contain electrons, repulsion patterns become anisotropic — VSEPR assumes uniform repulsion, so it fails.

2. Notable Failures of VSEPR

🔴 XeF₆: The Octahedral That Isn’t

Xenon hexafluoride (XeF₆) has six bonding pairs and one lone pair (if you count the Xe 4d electrons?). VSEPR predicts a distorted octahedron (like a pentagonal pyramid or capped octahedron). But experimental structures show a complex, dynamic geometry — not a single rigid shape. The lone pair is "stereochemically active" but the molecule is fluxional. VSEPR cannot decide between geometries.

❌ VSEPR prediction: Distorted octahedral (one lone pair) → square pyramid? Actually, for 7 electron domains (6 bonds + 1 lone pair) VSEPR gives pentagonal bipyramid, but that doesn't match XeF₆. The molecule adopts a distorted octahedron with the lone pair occupying a face — something VSEPR doesn't describe well.

🔴 Ni(CO)₄: Tetrahedral but … d‑Orbitals Intervene

Nickel tetracarbonyl is tetrahedral — VSEPR gets that right. But why is it tetrahedral, not square planar? VSEPR only counts electron pairs. The real reason lies in crystal field theory and d‑orbital splitting. The tetrahedral geometry minimises ligand‑ligand repulsion and gives favourable d‑orbital splitting for Ni(0) (d¹⁰). VSEPR doesn't explain why some d⁸ complexes are square planar (e.g., PtCl₄²⁻) while others are tetrahedral. It just counts pairs.

🔴 SF₄: The See‑Saw Works, but the Lone Pair Is Active

Sulfur tetrafluoride has 5 electron domains (4 bonds + 1 lone pair). VSEPR correctly gives a see‑saw shape. But the lone pair’s location and the bond angles (F–S–F ~100°) are not exactly predicted by simple repulsion — the equatorial lone pair compresses the axial bonds. It works qualitatively, but fails quantitatively. For larger atoms like Te, the lone pair can be stereochemically inactive — a total VSEPR breakdown.

🔴 Transition Metal Complexes: VSEPR Is Useless

VSEPR assumes the central atom uses s and p orbitals. Transition metals use d orbitals for bonding. The shapes (octahedral, square planar, tetrahedral) are determined by crystal field or ligand field effects, not electron‑pair repulsion. For example, [Co(CN)₆]³⁻ is low‑spin octahedral, while [CoF₆]³⁻ is high‑spin — both octahedral, but VSEPR has no way to distinguish spin states or Jahn–Teller distortions. In fact, the Jahn–Teller theorem predicts elongated octahedra for certain d‑electron counts (e.g., Cu²⁺). VSEPR misses this entirely.

Jahn–Teller distortion in Cu²⁺ complexes: two axial bonds elongate. VSEPR predicts a regular octahedron, but the actual geometry is distorted due to electronic effects in the d‑orbitals.

3. Why Does VSEPR Fail?

VSEPR treats electron pairs as classical point charges. It ignores:

• d‑orbital participation: When d‑orbitals are involved in bonding or contain lone pairs, repulsion is not spherically symmetric.
• Ligand sterics: Bulky ligands can distort geometries beyond what simple repulsion predicts.
• Second‑order Jahn–Teller effects: Mixing of orbitals can stabilise non‑VSEPR shapes (e.g., bent vs. linear for some d¹⁰ complexes).
• Fluxionality: Some molecules (like XeF₆ or IF₇) rapidly interconvert between geometries; VSEPR predicts a static shape.

🔬 Takeaway: VSEPR is a great starting point for main‑group molecules without d‑electrons. For transition metals, heavier p‑block elements, or any system with significant d‑orbital involvement, you need molecular orbital theory or ligand field theory to predict shapes.

4. The Orbital War: A Truce?

Modern chemists don't discard VSEPR. Instead, they know its limits. When a molecule violates VSEPR, that's a signal that d‑orbitals, spin states, or relativistic effects are at play. The real "war" is between simple models and quantum reality. And in that war, the molecules always win.

The peace agreement: VSEPR for simple main‑group chemistry, molecular orbital theory for the rest.

📚 Further Reading

• 1. Gillespie, R. J., & Hargittai, I. (2012). The VSEPR Model of Molecular Geometry. Dover.
• 2. Cotton, F. A., & Wilkinson, G. (1988). Advanced Inorganic Chemistry (5th ed.). Wiley. (Chapters on molecular shapes and VSEPR failures)
• 3. Pearson, R. G. (1976). “The VSEPR model and the crystal field theory.” Journal of Chemical Education, 53(8), 475.
• 4. Gillespie, R. J., & Silvi, B. (2002). “The VSEPR model revisited.” Coordination Chemistry Reviews, 233‑234, 53-62.

SM-EDUCATE CHEMISTRY — Beyond simple models: when VSEPR surrenders