Photochemical Metallonitrene Formation for N Atom Transfers. Metal- versus Azide-Centered Excited States and the Role of Their Energetic Order

Square-planar d⁸-configured transition metal complexes, [M^II(dppe)(N₃)₂] ([<b>1</b>_M]) with M = Ni, Pd, Pt and dppe = 1,2-bis(diphenylphosphino)ethane, are used as photolabile precursors for metallonitrenes, an important class of subvalent species that serve as reactive intermediates in the direct amination of CH bonds. We monitored the relaxation dynamics of [<b>1</b>_M], following their ultrafast excitation at 266 nm, with femtosecond-to-microsecond mid-infrared (MIR) spectroscopy in the antisymmetric N₃-stretching region. Regardless of M, the metallonitrene, [M^II(dppe)(N₃)(N)] ([<b>2</b>_M]), could be identified as the predominant photoproduct through the N₃-stretching absorption of its remaining intact azido ligand. Accompanying quantum-chemical calculations demonstrate that the N_α-N_β bond cleavage required for metallonitrene formation is facilitated by an intriguing triplet azide-ππ* state, which is electronically preconfigured for the adiabatic release of a neutral N₂ fragment. This state extrapolates smoothly via a small barrier to the electronic ground state of [<b>2</b>_M], which is the N-centered triplet diradical. No evidence was found for a closed-shell singlet nitrene intermediate, which was previously invoked in the photoinduced formation of organic nitrenes. The photochemical conversion yield was found to increase systematically in the order Ni > Pd > Pt, which was found to reflect an energetic reversal of the lowest metal-centered triplet d-d states and the productive, purely ligand-centered, azide-ππ* state. Our findings thus advocate a simple design guideline for efficient CH-amination catalysts: use strong-field ancillary ligands in combination with 4d- and 5d-transition metals to shift the lowest metal-centered state above the azide-ππ* state, thus eliminating detrimental deactivation pathways.