Regular and inverse secondary kinetic enthalpy effects (KHE) for the rate of inversion of thioether and 1,1'-biisoquinoline complexes of ruthenium and osmium.

Abstract

Thioether complexes with the formula Delta/Lambda-chloro(thioether)bis(2,2'-bipyridine)metal(II) (M = Ru, Os; thioether = dimethyl sulfide (3a(+)), diethyl sulfide (3b(+)), and tetrahydrothiophene (3c(+))) have been synthesized. The rates of inversion at the sulfur atom of the thioether ligands have been measured by spin-inversion transfer and line-shape NMR methods. In every case, the ruthenium derivative exhibits a faster inversion frequency at a given temperature than the corresponding osmium derivative. In contrast, similar complexes with the formula chloro(delta/lambda-1,1'-biisoquinoline)(2,2':6',2"-terpyridine)metal(II), 4(M=Ru,Os)(+), undergo atropisomerization of the misdirected 1,1'-biisoquinoline (1,1'-biiq) ligand with rates that are faster for osmium than ruthenium. As a result of the lanthanide contraction effect and the similar metric parameters associated with the structures of second-row and third-row transition metal derivatives, steric factors associated with the isomerizations are presumably similar for the Ru and Os derivatives of these compounds. Since third-row transition metal complexes tend to have larger bond dissociation enthalpies (BDE) than their second-row congeners, we conclude the difference in reactivities of 3(M=Ru)(+) versus 3(M=Os)(+) and 4(M=Ru)(+) versus 4(M=Os)(+) are attributed to electronic effects. For 3, the S3p lone pair of the thioether, the principal sigma donor orbital, is orthogonal to the metal sigma acceptor orbital in the transition state of inversion at sulfur and the S 3s orbital is an ineffective sigma donor. Thus, a regular relationship between the kinetic and thermodynamic stabilities of 3(M=Ru)(+) and 3(M=Os)(+) is observed for the directed <==> [misdirected] <==> directed (DMD) isomerization (the more thermodynamically stable bond is less reactive). In contrast, atropisomerization of 4(+) involves redirecting (strengthening) the M-N bonds of the misdirected 1,1'-biiq ligand in the transition state. Therefore, an inverse relationship between the kinetic and thermodynamic stabilities of 4(M=Ru)(+) and 4(M=Os)(+) is observed for the misdirected <==> [directed] <==> misdirected (MDM) isomerization (the more thermodynamically stable bond is more reactive). The rates obtained for 4(+) are consistent with the rates of atropisomerization of Delta/Lambda-(delta/lambda-1,1'-biisoquinoline)bis(2,2'-bipyridine)metal(II), 1(M=Ru,Os)(2+), and (eta(6)-benzene) Delta/Lambda-(delta/lambda-1,1'-biisoquinoline)halometal(II), 2(M=Ru,Os;halo=Cl,I)(+), that we reported previously. We term the relative rates of reaction of second-row versus third-row transition metal derivatives kinetic element effects (KEE = k(second)/k(third)). While the KEE appears to be generally useful when comparing reactions of isostructural species (e.g. the relative rates of 1(M=Ru)(2+), 1(M=Os)(2+), and 1(M=Ir)(3+)), different temperature dependencies of reactions prevent the comparison of related reactions between species that have different structures (e.g., the 1,1'-biiq atropisomerization reactions of 1(M=Ru,Os)(2+) versus 2(M=Ru,Os;halo=Cl,I)(+) versus 4(M=Ru,Os)(+)). This problem is overcome by comparing entropies of activation and kinetic enthalpy effects (KHE = DeltaH(third)/DeltaH(second)). For a given class of 1,1'-biiq complexes, we observe a structure/reactivity relationship between DeltaH and the torsional twist of the 1,1'-biiq ligands that are measured in the solid state.