orbital overlap
/ OR-bih-tal OH-ver-lap /
Two people can only shake hands if they reach into the same patch of space. Orbitals are the same: two atomic orbitals can only combine into a molecular orbital where their clouds actually share territory. Orbital overlap is the amount of that shared region — how much the two electron clouds occupy the same place at the same time.
More precisely, overlap is measured by an overlap integral, a number that adds up where the two atomic orbital functions are both large and have the same sign. When the lobes meet with matching sign, the overlap is positive and electron density builds up between the nuclei: bonding. When they meet with opposite sign, the overlap is negative and density is cancelled there: antibonding. When the lobes are arranged so that the positive and negative parts exactly cancel — for instance an s orbital approaching a p orbital sideways — the net overlap is zero, and the two orbitals simply ignore each other no matter how close they get.
Overlap is one of the three conditions for orbitals to combine, alongside being close in energy and matching in symmetry. The bigger the net overlap, the larger the energy split between the resulting bonding and antibonding orbitals, and so the stronger the bond. This is why sigma bonds (head-on overlap) are usually stronger than pi bonds (sideways overlap), and why bonds weaken as atoms get larger and their diffuse orbitals overlap less effectively.
An s orbital approaching a p orbital end-on overlaps strongly and bonds well. But if the same s orbital approaches the p orbital sideways, its single sign meets the p orbital's plus lobe and minus lobe equally — the two contributions cancel, the net overlap is zero, and no bond forms.
Geometry decides overlap: end-on overlap bonds, but a mismatched approach can cancel to zero.
Big overlap is necessary but not sufficient — two orbitals that overlap a lot still cannot bond unless they are also close in energy and matched in symmetry.