DFT Computation of Relative Spin-State Energetics of Transition Metal Compounds

DFT is often used to predict the energetics of transition metal compounds. In particular, energy differences between states of different spin are of great interest. This review discusses the accuracy of such computations for spin-state splittings of monometallic complexes where the differences in spin-pairing occur mostly or exclusively on the metal itself. Large differences are found for computed splittings using different functionals, in particular between pure and hybrid functionals. An optimum exact exchange admixture of ~15% seems to yield accurate results in many cases, but experimental and especially theoretical uncertainties suggest that this should not be relied upon, and the development of new functionals is seen to be highly desirable.