Not so! For example, water has the inverse recombination reaction of \

H₃O^- -> 2H₂O with a diffusion controlled reaction rate constant of 1.3 x 10¹¹ M⁻¹ s⁻¹ at room temperature. This, of course, assumes pure water and ignores such things as isotope effects (deuterium oxide, D₂O, self-ionizes less than "regular" water, for example).

Thus even state change (water can change to steam and back again, for instance) isn't final.