Palladium nanoparticles (4 nm, fcc) were prepared through decomposition of [Pd2(dba)3] by H2 in the presence of a chiral xylofuranoside diphosphite. These particles catalyze the allylic alkylation of rac-3-acetoxy-1,3-diphenyl-1-propene with dimethyl malonate leading to an almost total conversion of the (R) enantiomer and almost no reaction with the (S). This gives rise to 97% ee for the alkylation product and a kinetic resolution of the substrate recovered with ca. 90% ee. This behavior was compared to that of a molecular catalyst at various dilutions, and the differences between the two systems are discussed. This is the first colloidal system shown to display such a high enantioselectivity besides the well-known Pt/cinchonidine system.

No abstract

Metal nanoparticles have been used for a long time to catalyze chemical reactions in both heterogeneous and homogeneous phases.[1] The analysis of traditional heterogeneous and homogeneous catalysis requires very different techniques that are difficult to combine for the study of metal nanoparticles, in which distinguishing between colloidal and molecular catalysis is difficult.[2] Thus, many questions concerning the reactivity of metal nanoparticles are still open, particularly the nature of intermediate surface species, knowledge of which is important for the development of new nanocatalysts and new catalytic transformations. Some of us have used solid-state NMR spectroscopy for this purpose recently, [3] and herein we report the combination of this method with desorption techniques for investigating the reactivity of ruthenium nanoparticles.The synthesis of metal nanoparticles by hydrogenation of organometallic precursors in the presence of organic ancillary ligands, such as amines, thiols, or carboxylic acids as stabilizers, has been investigated for over fifteen years by some of us.[4] In particular, essentially monodisperse, very small ruthenium nanoparticles, which display a remarkable surface coordination chemistry, can be obtained using [Ru-(cod)(cot)] as a precursor (cod = 1,5-cyclooctadiene; cot = 1,3,5-cyclooctatriene). This system, and similar ones involving Pd, Pt, or Rh nanoparticles, catalyzes a number of chemical reactions such as olefin hydrogenation, CÀC coupling, and hydrogenation of aromatic hydrocarbons.[5] Some of us have shown independently that palladium nanoparticles stabilized by asymmetric phosphite groups are good enantioselective alkylation catalysts.[6] This result provides strong evidence for the direct coordination of ligands, in this case phosphite groups, to the palladium surface.The coordination of ligands such as CO, [7] amines, [8] and organosilanes, [9] has previously been established by NMR spectroscopy studies in solution or in the solid state. The coordination of hydrogen to metal nanoparticles, however, is especially important. Hydrogen binding to clean metal surfaces has been well established by surface science, and it is generally accepted that one hydrogen atom is adsorbed per surface metal atom. [10] We have recently demonstrated the presence of mobile hydrides, which are in slow exchange with gaseous dihydrogen, on the surface of amine-protected ruthenium nanoparticles using a combination of gas-phase 1 H NMR and solid-state 2 H NMR spectroscopy. [3] Furthermore, other species, such as alkenes or arenes, may adsorb on the surface during a catalytic process or give rise to new reactive intermediates, including alkyl groups and carbenes. The important question which then arises is whether these groups are stable and can be detected spectroscopically, as in organometallic complexes.Herein we describe: 1) the synthesis of a new class of phosphine-protected ruthenium nanoparticles, 2) the characterization of phosphine coordination by NMR spectroscopy techniques, 3) the p...