Hybridization — where the directions come from

VSEPR told you the shape; this is its cause. A carbon's bare orbitals — one round 2s and three perpendicular 2p — point the wrong way to make four equal bonds at 109.5°. So they mix: one s plus three p recombine into four identical sp³ hybrids aimed at the corners of a tetrahedron. Use fewer p's and you get fewer, wider hybrids — sp² at 120°, sp at 180° — each leaving spare p orbitals behind that become π bonds. This is the engine under the last two rungs: water's oxygen is sp³ (the 104.5°), CO₂'s carbon is sp (the linear, plus two π's).

Bridge-Burners LLC · Fiddler · s + p → hybrids · #hybrids = σ+lone domains · leftover p = π · anchor: AKASHA

State

hybridizationsp³
recipes + 3p
# hybrids4
angle / geometry109.5° tetrahedral
leftover p (π)0
s-character25%

Closes the loop

sp³O in H₂O → the 104.5° bend; C in CH₄
sp²C=C ethene, BF₃ → 120°, 1 π
spC in CO₂, C≡C → 180°, 2 π

Status discipline

Literal#hybrids = σ-bond + lone-pair domains; sp³/sp²/sp give 109.5°/120°/180° and 25/33/50% s-character; leftover p orbitals (0/1/2) are the π-bond capacity. Predicts the observed geometries.
BridgeHybridization as the origin of VSEPR's directions; the spare p's as the source of double and triple bonds.
SpeculativeIt's a model, not an event — no "promotion then mixing" in time. Methane's photoelectron spectrum shows two ionization energies, not four identical sp³ orbitals; the delocalized-MO picture is equally valid. Useful for geometry, often overstated.