Hybridization

Plain (unhybridized) valence-bond overlap correctly explains simple diatomics like H2, F2 and HF, but it cannot explain several well-known observations: how beryllium forms TWO covalent bonds, how boron forms THREE, or how carbon forms FOUR, when their ground-state electron configurations (Be: $1s^22s^2$; B: $1s^22s^22p^1$; C: $1s^22s^22p^2$) show far fewer unpaired electrons than that; nor can it explain why, experimentally, all four C–H bonds of methane turn out completely identical in length, strength and angle, even though a naive picture would build them from differently-shaped, non-equivalent s and p orbitals. Hybridization resolves both problems: it is the mixing of valence orbitals belonging to the SAME atom, recast into an equal number of brand-new, mutually EQUIVALENT hybrid orbitals. The process happens in two steps. First, formation of the excited state: a paired electron in a lower-energy orbital (e.g. 2s) is uncoupled and promoted into an empty orbital of only slightly higher energy (e.g. 2p), so the atom now has as many singly-occupied (half-filled) orbitals as its known valency requires — e.g. Be, whose valency in BeF2 is two, promotes one 2s electron into an empty 2p orbital to give two half-filled orbitals. Second, mixing and recasting: these now-comparable-energy 's' and 'p' orbitals combine and redistribute their electron density an …