heteronuclear diatomic MO diagram

Now take two different atoms — carbon and oxygen, nitrogen and oxygen, hydrogen and fluorine. The molecular orbital picture still works, but the diagram is no longer symmetric. Because the two atoms differ in electronegativity, their atomic orbitals start at different energies, and the more electronegative atom's orbitals sit lower down the page. The whole diagram becomes lopsided, and that lopsidedness is the molecular orbital theory's way of drawing polarity.

Here is what the asymmetry does to the orbitals. A bonding molecular orbital is built mostly from the lower-energy atomic orbitals, so it leans toward the more electronegative atom — its electrons spend more time there, which is exactly a polar covalent bond. The matching antibonding orbital leans the other way, toward the less electronegative atom. The orbitals are no longer shared 50-50; the coefficients in the LCAO are unequal, mirroring the unequal pull of the two nuclei. The bigger the electronegativity gap, the more lopsided the orbitals, and at the extreme the bonding orbital is almost entirely on one atom — which is just the ionic limit drawn in MO language.

These diagrams explain real molecules a Lewis picture handles only awkwardly. For carbon monoxide, CO, MO theory shows that the highest occupied orbital is a carbon-rich, weakly antibonding lone pair — which is precisely the orbital CO uses to bond to metals through its carbon end, the foundation of metal-carbonyl chemistry. For nitric oxide, NO, the diagram puts one electron in an antibonding orbital, giving a bond order of 2.5 and one unpaired electron, correctly predicting that NO is a stable, paramagnetic radical. Lopsided diagrams turn electronegativity into something you can read off a page.