A recyclable catalyst that precipitates at the end of the reaction.

Homogeneous catalysts--which exist in the same (usually liquid) phase as reactants and products--are usually more selective than heterogeneous catalysts and far less affected by limitations due to slow transport of reactants and products, but their separation from reaction products can be costly and inefficient. This has stimulated the development of strategies that facilitate the recycling of homogeneous catalysts. Some of these methods exploit the preference of a catalyst for one of two solvents with thermoregulated miscibility; others exploit a dramatic decrease in catalyst solubility as one reagent is consumed or temperature changed after completion of the reaction. Here we describe a tungsten catalyst for the solvent-free hydrosilylation of ketones that retains its activity until essentially all of the liquid substrate is converted to liquid products, which we can then simply decant to separate the catalyst that precipitates from the products of the reaction. We attribute the ability of the catalyst to retain its solubility and hence activity until completion of the reaction to the transient formation of liquid clathrate that contains a few molecules of the substrate per molecule of the otherwise solid catalyst. Insights into the fundamental processes controlling the formation of this liquid clathrate might help to tailor other catalysts and substrates, so as to develop efficient and solvent-free schemes for reactions of practical interest.

Synthesis and reactivity of silyl ruthenium complexes: the importance of trans effects in C-H activation, Si-C bond formation, and dehydrogenative coupling of silanes.

A series of octahedral ruthenium silyl hydride complexes, cis-(PMe(3))(4)Ru(SiR(3))H (SiR(3) = SiMe(3), 1a; SiMe(2)CH(2)SiMe(3), 1b; SiEt(3), 1c; SiMe(2)H, 1d), has been synthesized by the reaction of hydrosilanes with (PMe(3))(3)Ru(eta(2)-CH(2)PMe(2))H (5), cis-(PMe(3))(4)RuMe(2) (6), or (PMe(3))(4)RuH(2) (9). Reaction with 6 proceeds via an intermediate product, cis-(PMe(3))(4)Ru(SiR(3))Me (SiR(3) = SiMe(3), 7a; SiMe(2)CH(2)SiMe(3), 7b). Alternatively, 1 and 7 have been synthesized via a fast hydrosilane exchange with another cis-(PMe(3))(4)Ru(SiR(3))H or cis-(PMe(3))(4)Ru(SiR(3))Me, which occurs at a rate approaching the NMR time scale. Compounds 1a, 1b, 1d, and 7a adopt octahedral geometries in solution and the solid state with mutually cis silyl and hydride (or silyl and methyl) ligands. The longest Ru-P distance within a complex is always trans to Si, reflecting the strong trans influence of silicon. The aptitude of phosphine dissociation in these complexes has been probed in reactions of 1a, 1c, and 7a with PMe(3)-d(9) and CO. The dissociation is regioselective in the position trans to a silyl ligand (trans effect of Si), and the rate approaches the NMR time scale. A slower secondary process introduces PMe(3)-d(9) and CO in the other octahedral positions, most likely via nondissociative isomerization. The trans effect and trans influence in 7a are so strong that an equilibrium concentration of dissociated phosphine is detectable (approximately 5%) in solution of pure 7a. Compounds 1a-c also react with dihydrogen via regioselective dissociation of phosphine from the site trans to Si, but the final product, fac-(PMe(3))(3)Ru(SiR(3))H(3) (SiR(3) = SiMe(3), 4a; SiMe(2)CH(2)SiMe(3), 4b; SiEt(3), 4c), features hydrides cis to Si. Alternatively, 4a-c have been synthesized by photolysis of (PMe(3))(4)RuH(2) in the presence of a hydrosilane or by exchange of fac-(PMe(3))(3)Ru(SiR(3))H(3) with another HSiR(3). The reverse manifold - HH elimination from 4a and trapping with PMe(3) or PMe(3)-d(9) - is also regioselective (1a-d(9)() is predominantly produced with PMe(3)-d(9) trans to Si), but is very unfavorable. At 70 degrees C, a slower but irreversible SiH elimination also occurs and furnishes (PMe(3))(4)RuH(2). The structure of 4a exhibits a tetrahedral P(3)Si environment around the metal with the three hydrides adjacent to silicon and capping the P(2)Si faces. Although strong Si...HRu interactions are not indicated in the structure or by IR, the HSi distances (2.13-2.23(5) A) suggest some degree of nonclassical SiH bonding in the H(3)SiR(3) fragment. Thermolysis of 1a in C(6)D(6) at 45-55 degrees C leads to an intermolecular CD activation of C(6)D(6). Extensive H/D exchange into the hydride, SiMe(3), and PMe(3) ligands is observed, followed by much slower formation of cis-(PMe(3))(4)Ru(D)(Ph-d(5)). In an even slower intramolecular CH activation process, (PMe(3))(3)Ru(eta(2)-CH(2)PMe(2))H (5) is also produced. The structure of intermediates, mechanisms, and aptitudes for PMe(3) dissociation and addition/elimination of H-H, Si-H, C-Si, and C-H bonds in these systems are discussed with a special emphasis on the trans effect and trans influence of silicon and ramifications for SiC coupling catalysis.

Structure and reactivity of bis(silyl) dihydride complexes (PMe(3))(3)Ru(SiR(3))(2)(H)(2): model compounds and real intermediates in a dehydrogenative C-Si bond forming reaction.

A series of stable complexes, (PMe(3))(3)Ru(SiR(3))(2)(H)(2) ((SiR(3))(2) = (SiH(2)Ph)(2), 3a; (SiHPh(2))(2), 3b; (SiMe(2)CH(2)CH(2)SiMe(2)), 3c), has been synthesized by the reaction of hydridosilanes with (PMe(3))(3)Ru(SiMe(3))H(3) or (PMe(3))(4)Ru(SiMe(3))H. Compounds 3a and 3c adopt overall pentagonal bipyramidal geometries in solution and the solid state, with phosphine and silyl ligands defining trigonal bipyramids and ruthenium hydrides arranged in the equatorial plane. Compound 3a exhibits meridional phosphines, with both silyl ligands equatorial, whereas the constraints of the chelate in 3c result in both axial and equatorial silyl environments and facial phosphines. Although there is no evidence for agostic Si-H interactions in 3a and 3b, the equatorial silyl group in 3c is in close contact with one hydride (1.81(4) A) and is moderately close to the other hydride (2.15(3) A) in the solid state and solution (nu(Ru.H.Si) = 1740 cm(-)(1) and nu(RuH) = 1940 cm(-)(1)). The analogous bis(silyl) dihydride, (PMe(3))(3)Ru(SiMe(3))(2)(H)(2) (3d), is not stable at room temperature, but can be generated in situ at low temperature from the 16e(-) complex (PMe(3))(3)Ru(SiMe(3))H (1) and HSiMe(3). Complexes 3b and 3d have been characterized by multinuclear, variable temperature NMR and appear to be isostructural with 3a. All four complexes exhibit dynamic NMR spectra, but the slow exchange limit could not be observed for 3c. Treatment of 1 with HSiMe(3) at room temperature leads to formation of (PMe(3))(3)Ru(SiMe(2)CH(2)SiMe(3))H(3) (4b) via a CH functionalization process critical to catalytic dehydrocoupling of HSiMe(3) at higher temperatures. Closer inspection of this reaction between -110 and -10 degrees C by NMR reveals a plethora of silyl hydride phosphine complexes formed by ligand redistribution prior to CH activation. Above ca. 0 degrees C this mixture converts cleanly via silane dehydrogenation to the very stable tris(phosphine) trihydride carbosilyl complex 4b. The structure of 4b was determined crystallographically and exhibits a tetrahedral P(3)Si environment around the metal with the three hydrides adjacent to silicon and capping the P(2)Si faces. Although strong Si.HRu interactions are not indicated in the structure or by IR, the HSi distances (2.00(4) - 2.09(4) A) and average coupling constant (J(SiH) = 25 Hz) suggest some degree of nonclassical SiH bonding in the RuH(3)Si moiety. The least hindered complex, 3a, reacts with carbon monoxide principally via an H(2) elimination pathway to yield mer-(PMe(3))(3)(CO)Ru(SiH(2)Ph)(2), with SiH elimination as a minor process. However, only SiH elimination and formation of (PMe(3))(3)(CO)Ru(SiR(3))H is observed for 3b-d. The most hindered bis(silyl) complex, 3d, is extremely labile and even in the absence of CO undergoes SiH reductive elimination to generate the 16e(-) species 1 (DeltaH(SiH)(-)(elim) = 11.0 +/- 0.6 kcal x mol(-)(1) and DeltaS(SiH)(-)(elim) = 40 +/- 2 cal x mol(-)(1) x K(-)(1); Delta = 9.2 +/- 0.8 kcal x mol(-)(1) and Delta = 9 +/- 3 cal x mol(-)(1).K(-)(1)). The minimum barrier for the H(2) reductive elimination can be estimated, and is higher than that for silane elimination at temperatures above ca. -50 degrees C. The thermodynamic preferences for oxidative additions to 1 are dominated by entropy contributions and steric effects. Addition of H(2) is by far most favorable, whereas the relative aptitudes for intramolecular silyl CH activation and intermolecular SiH addition are strongly dependent on temperature (DeltaH(SiH)(-)(add) = -11.0 +/- 0.6 kcal x mol(-)(1) and DeltaS(SiH)(-)(add) = -40 +/- 2 cal.mol(-)(1) x K(-)(1); DeltaH(beta)(-CH)(-)(add) = -2.7 +/- 0.3 kcal x mol(-)(1) and DeltaS(beta)(-CH)(-)(add) = -6 +/- 1 cal x mol(-)(1) x K(-)(1)). Kinetic preferences for oxidative additions to 1 - intermolecular SiH and intramolecular CH - have been also quantified: Delta = -1.8 +/- 0.8 kcal x mol(-)(1) and Delta = -31 +/- 3 cal x mol(-)(1).K(-)(1); Delta = 16.4 +/- 0.6 kcal x mol(-)(1) and Delta = -13 +/- 6 cal x mol(-)(1).K(-)(1). The relative enthalpies of activation (-)(1) x K(-)(1)). Kinetic preferences for oxidative additions to 1 - intermolecular SiH and intramolecular CH - have been also quantified: Delta (H)SiH(add) = 1.8 +/- 0.8 kcal x mol(-)(1) and Delta S((SiH-add) =31+/- 3 cal x mol(-)(1) x K(-)(1); Delta S (SiH -add) = 16.4 +/- 0.6 kcal x mol(-)(1) and =Delta S (SiH -CH -add) =13+/- 6 cal x mol(-)(1) x K(-)(1). The relative enthalpies of activation are interpreted in terms of strong SiH sigma-complex formation - and much weaker CH coordination - in the transition state for oxidative addition.