The two theories used to predict molecular geometry are Valence Shell Electron Pair Repulsion (VSEPR) theory and Valence Bond Theory (VB theory). Both approaches start from the idea that electrons around a central atom determine a molecule’s three-dimensional shape, but they get there by different routes. VSEPR focuses on how electron pairs push each other apart, while Valence Bond Theory explains shape through the mixing, or “hybridization,” of atomic orbitals.
VSEPR Theory: Shape From Repulsion
VSEPR theory predicts molecular geometry based on a simple principle: electron pairs in the outer shell of a central atom repel each other and arrange themselves as far apart as possible. That arrangement determines the molecule’s shape. You count up all the electron groups around the central atom, both bonding pairs (shared between atoms) and lone pairs (not shared), and the geometry follows from how many groups there are.
Eight basic postulates guide the theory:
- Electron pairs repel one another and spread out to minimize that repulsion.
- The outer electron shell is treated as spherical.
- Double and triple bonds count as a single electron group.
- Lone pairs take up more space than bonding pairs because they’re held by only one nucleus instead of two.
- Repulsion strength follows a hierarchy: lone pair/lone pair repulsion is strongest, lone pair/bonding pair is moderate, and bonding pair/bonding pair is weakest.
That hierarchy of repulsion is what makes VSEPR so useful for explaining why real molecules deviate from “perfect” angles. A tetrahedral arrangement of four electron groups gives an ideal angle of 109.5°. But in water, two of those groups are lone pairs, which squeeze the two O-H bonds closer together. The measured H-O-H angle is 104.5°, not 109.5°. Ammonia has one lone pair and three bonding pairs, so its H-N-H angle lands at 106.6°, compressed but not as much as water’s.
Common VSEPR Shapes and Bond Angles
Chemists use a shorthand called AXE notation, where A is the central atom, X represents bonded atoms, and E represents lone pairs. The number of X’s and E’s together determines the electron pair geometry, while the arrangement of just the atoms (ignoring lone pairs) gives you the molecular geometry.
- AX2 (2 bonds, 0 lone pairs): Linear, 180°
- AX3 (3 bonds, 0 lone pairs): Trigonal planar, 120°
- AX2E (2 bonds, 1 lone pair): Bent, slightly less than 120°
- AX4 (4 bonds, 0 lone pairs): Tetrahedral, 109.5°
- AX3E (3 bonds, 1 lone pair): Trigonal pyramidal, slightly less than 109.5°
- AX2E2 (2 bonds, 2 lone pairs): Bent, around 104.5°
For atoms that can hold more than eight electrons in their outer shell, the shapes expand further. Five electron groups (sp3d hybridization territory) produce a trigonal bipyramidal starting point, which can become see-saw, T-shaped, or linear as lone pairs replace bonds. Six electron groups (sp3d2) give an octahedral base that can become square pyramidal or square planar.
Valence Bond Theory: Shape From Orbital Mixing
Valence Bond Theory takes a quantum mechanical approach. It explains bonding as the overlap of atomic orbitals on neighboring atoms. The problem is that the standard s, p, d, and f orbitals don’t point in the right directions to explain the shapes molecules actually have. Carbon’s four bonds in methane, for instance, point toward the corners of a tetrahedron, but carbon’s native orbitals don’t naturally aim that way.
The solution is hybridization: atomic orbitals mathematically combine to form new, equivalent “hybrid” orbitals that do point in the right directions. The type of hybridization directly determines the geometry:
- sp hybridization: One s orbital combines with one p orbital, producing two equivalent hybrid orbitals oriented 180° apart. Result: linear geometry. Beryllium hydride (BeH2) is a classic example.
- sp2 hybridization: One s orbital combines with two p orbitals, producing three equivalent hybrids oriented 120° apart in a plane. Result: trigonal planar geometry.
- sp3 hybridization: One s orbital combines with three p orbitals, producing four equivalent hybrids at 109.5° to each other. Result: tetrahedral geometry. Methane (CH4) is the textbook case.
- sp3d hybridization: Five hybrid orbitals in a trigonal bipyramidal arrangement.
- sp3d2 hybridization: Six hybrid orbitals in an octahedral arrangement.
Valence Bond Theory gives you a reason for the geometry rooted in quantum mechanics, not just electron repulsion. It explains why bonds form where they do and why certain bond angles are favored energetically.
How the Two Theories Differ
VSEPR is qualitative. You draw a Lewis structure, count electron groups, and read off the predicted shape. It requires no math and no knowledge of orbital theory. This makes it the go-to method for quickly predicting geometry in introductory chemistry.
Valence Bond Theory is a quantum mechanical model. It describes specific bonds as the overlap of specific orbitals and uses hybridization to explain why atoms adopt particular geometries. It goes deeper than VSEPR by providing a physical mechanism for the shape, not just a prediction rule. For simple molecules, both theories give the same answer. VB theory becomes more useful when you need to understand the nature of the bonding itself, not just the shape.
In practice, chemists often use the two theories together. VSEPR quickly identifies the geometry from a Lewis structure, and hybridization from VB theory then explains what the orbitals are doing to produce that geometry. If VSEPR tells you a molecule is tetrahedral, VB theory tells you the central atom is sp3 hybridized.
Where These Theories Fall Short
Both theories work well for small molecules built around main-group elements (the s- and p-block of the periodic table), but they have blind spots. VSEPR struggles with transition metal compounds, where d electrons behave in ways the simple repulsion model doesn’t capture. It also fails for some molecules with identical electron counts that should have the same shape but don’t. For example, VSEPR predicts that both IF7 and TeF7⁻ (which have 56 valence electrons each) should be pentagonal bipyramidal, but experimental data shows the fluorine atoms in TeF7⁻ don’t sit in the same plane as predicted.
Another known problem involves the “inert pair effect,” where a lone pair on a heavy atom doesn’t influence the geometry the way VSEPR expects. Molecules like SeCl6²⁻ and TeCl6²⁻ should have seven electron pair positions according to VSEPR, but they turn out to be regular octahedra because one electron pair is stereochemically inactive.
For high-accuracy geometry predictions, especially for larger or more complex molecules, computational methods have largely taken over. Density functional theory (DFT) calculates molecular geometry by optimizing the energy of the entire electron system, producing results that closely match experimental measurements. But VSEPR and Valence Bond Theory remain the standard teaching tools and the fastest way to reason about molecular shape with pencil and paper.