Abstract
Ring size effects on geometries and electronic structures were investigated for the (C (n) H (n) )M(C (m) H (m) ) (n = 4, 5, or 6; m = 8, 7, or 6; m + n = 12; M = Ti-Ni) systems using density functional theory. The lowest-energy C(12)H(12)M structures for the early transition metals titanium, vanadium, and chromium are the experimentally known singlet (η(5)-C(5)H(5))Ti(η(7)-C(7)H(7)), doublet (η(5)-C(5)H(5))V(η(7)-C(7)H(7)), and singlet (η(6)-C(6)H(6))(2)Cr, respectively. The likewise experimentally known singlet (η(6)-C(6)H(6))(2)Ti, doublet (η(6)-C(6)H(6))(2)V, and singlet (η(5)-C(5)H(5))Cr(η(7)-C(7)H(7)) are the second-lowest-energy structures with only a small energy difference between the two vanadium structures. For the later transition metals, dibenzenemetal complexes are the lowest-energy C(12)H(12)M species with two fully bonded hexahapto benzene rings in the lowest-energy manganese and iron derivatives and one hexahapto and one dihapto benzene ring in the lowest-energy cobalt and nickel derivatives. The lowest-energy (C(5)H(5))M(C(7)H(7)) structures for the later transition metals iron, cobalt, and nickel have partially bonded nonplanar C(7)H(7) rings with one or two uncomplexed C=C bonds. The (C(4)H(4))M(C(8)H(8)) (M = Ti-Ni) structures with the metal sandwiched between four- and eight-membered rings were found to be much higher in energy than their (C(5)H(5))M(C(7)H(7)) and (C(6)H(6))(2)M isomers.