nonstoichiometry

In school chemistry, compounds have tidy whole-number formulas: water is H2O, salt is NaCl, iron oxide is FeO — exactly one iron per oxygen. Nonstoichiometry is the surprising fact that many real solid compounds refuse to obey their neat formula precisely. Their true composition drifts away from the ideal ratio, and it does so continuously — you can have a range of compositions, all the same crystal, none of them a whole-number formula. Such a compound is called nonstoichiometric (an old name is a berthollide, as opposed to a well-behaved daltonide).

The textbook case is wustite, nominally FeO. Analyse real wustite and you never find exactly one iron per oxygen; instead the formula is Fe(1-x)O with x running from about 0.05 to 0.12 — there is always a DEFICIT of iron. How can a crystal simply be short of iron? Through point defects. The oxygen lattice stays full, but some of the iron sites are vacant. Charge must still balance, though: each missing Fe2+ (a vacancy of charge minus two) is compensated by converting two nearby Fe2+ ions into Fe3+ ions, each carrying one extra positive charge. So nonstoichiometry is not sloppy chemistry — it is a crystal quietly rearranging its defect population and its ion charges to accommodate an off-ideal composition.

Nonstoichiometry is common and important precisely in the compounds we most rely on for electronic and energy applications: transition-metal oxides and sulphides whose metal can adopt more than one charge state. That mixed-valence flexibility is exactly what makes them semiconduct, catalyse, store lithium, and change colour. Titanium oxide, uranium dioxide, the cobalt and manganese oxides in battery cathodes — all live and work as nonstoichiometric phases, and their useful electronic properties come directly from the defects that the off-ideal composition demands.