Thermal C-H Bond Activation of Benzene with Cationic [Pt(CX3(L)]+ Complexes in the Gas Phase: A Combined Experimental/Theoretical Study (X = H, D; L=1,10-Phenanthroline, 2,2′-Bipyrimidine, 2,2′-Bipyridine, and (o,o′-Cl2C6H3)N=C(CH3)-C(CH3)=N(o,o′-Cl2C6H3))

The collision-induced fragmentation behavior of cationic [Pt(CX3)(L)](+) complexes (X = H, D; L = 1,10-phenanthroline (phen), 2,2'-bipyrimidine (bipyrm), 2,2'-bipyridine (bipy) and (o,o'-Cl2C6H3)N=C-(CH3)-C(CH3)=N(o,o'-Cl2C6H3) ((Me)DAB(DCP))) and their thermal reactions with C6H6, [1,2,3]-C6H3D3, [1,3,5]-C6H3D3, and C6D6 were probed in gas-phase experiments. Collision-induced dissociations of the phen and bipyrm complexes result in preferential eliminations of neutral PtCH2 to afford the protonated heterocyclic ligands LH+ as ionic products; the bipy complex favors loss of methane due to a rollover cyclometalation process. Fragmentation of the (Me)DAB(DCP)complex leads to the elimination of up to four HCl molecules from the ligand; neither PtCH2 nor methane is liberated. In their bimolecular reactions with benzene, all platinum complexes lose methane from the initially formed adducts. The complexes with L = phen, bipyrm, and bipy react much more efficiently with benzene than the (Me)DAB(DCP) complex, and only in the latter case is significant adduct formation observed. For the formation of methane in the ion/molecule reactions of the complexes with deuterated benzenes, a microkinetic model was applied for fitting of the observed isotope patterns, which reveals the following: (i) the C-H bond activation of benzene prior to methane elimination is reversible, (ii) while for the (Me)DAB(DCP) complex the rate of H/D exchange is lower than that for the liberation of methane, an opposite behavior holds true for the other complexes; (iii) the overall kinetic isotope effects associated with the H/D exchange amount to ca. 1.5 for the phen bipyrin, and bipy complexes. DFTcalculations demonstrate that the popular B3LYP method rails to explain the observed H/D exchange but that the functionals mPW 1k, M05-2X. and mPW1PW91 can account for the experimental Findings. Nevertheless, serious discrepancies seem to exist that prevent an unambiguous distinction of mechanistic scenarios, i.e.. oxidative addition/reductive elimination versus sigma-CAM process.