Size controlled ruthenium nanoparticles have been prepared by decomposition of the organometallic precursor Ru(COD)(COT) (COD = 1,5-cyclooctadiene, COT = 1,3,5-cyclooctatriene) under an H2 atmosphere in organic solutions. The stabilisation of these nanoparticles has been carried out in the presence of either a pure alcohol or an alcohol/THF mixture, or organic ligands as further stabilizers. An organisation of the particles has also been realised using mesoporous alumina membranes and silica materials. The characterisation of the particles has been realised by different techniques both from organometallic or solid state chemistry such as electronic microscopy (TEM, HRTEM), X-ray diffraction (XRD), wide-angle X-ray scattering (WAXS) and photoelectron spectroscopy (XPS). Finally, some catalytic experiments have been performed. When the particles are synthesised in pure alcohols, the colloidal solutions are stable for long periods of time, until more than one year for some of them. The TEM micrographs reveal the presence of sponge-like agglomerates formed from individual particles or of isolated and well dispersed monocrystalline particles depending upon the alcohol alkyl chain; long alkyl chains favouring the formation of monodisperse individual particles of small size. In all cases the size distributions are narrow. WAXS and XRD analyses indicate the exclusive presence of hcp ruthenium in these materials. The size of the agglomerates or of the particles can be controlled by adjusting the reaction temperature or the composition of the solvent mixture. In the case of MeOH/THF mixtures, a linear correlation was established between the solvent composition and the size of the particles in the range 2-85 nm. This variability was attributed to the formation of a microemulsion acting as a nanoreactor and resulting from the segregation of the cyclooctane formed in the reaction from the polar medium. In the case of heptanol, the heptyl chain allows the medium to be homogeneous and therefore the particles remain well separated. 13C and 1H NMR studies evidence the weak coordination of heptanol on the surface of the particles and the presence of a fast exchange between free and coordinated ligands. The most interesting aspect of this study is the observation of dihydrogen evolution in the NMR tube. It is hence probably due to the dihydrogen desorption resulting from the formation of surface hydrides in the synthesis conditions. This demonstrates the concomitant presence of both functional ligands and labile ligands at the surface of nanoparticles and therefore the possibility to develop a novel complex chemistry at the surface of metal nanoparticles using the traditional concepts of organometallic chemistry and monodisperse particles as "super-atoms". The particles can also be stabilized by long chain amines or thiols to give after appropriate work-up colloids which can be obtained as powders and handled like organometallic compounds. Whereas the amine ligands exchange rapidly at the surface of the particles with free amines, the thiols oxidatively add to ruthenium and excess thiol leads to the reductive elimination of disulfides which are released into the solution and do not exchange with the ligands present at the surface of the particles. This may be the reason why the colloids adopt so different structures and super-structures, even though the alkyl chains are the same and the functional groups similar. The alkyl chains of the sulphur ligands will encircle the particle and interpenetrate with the alkyl chains of other ligands either free or located at the surface of other particles hence leading to the formation of superstructures. A slow catalytic coupling of thiols into disulfides which may be removed from the colloid surface was observed. The presence of some organization of the amines in THF could allow the growth of ruthenium particles in the channel created in this way and hence explain the vermicular aspects of the colloids. Alternatively, the dynamics of the amines could also allow changes in the coordination sites at the surface of the particles and therefore, the preferred coordination of the amine along the growth axis of the hcp structure, perpendicular to the basal plane. In any case, the dynamics of the amine will favour the coalescence of initially formed spherical particles and therefore favour the formation of worm-like particles. This study demonstrates that the coupling of traditional techniques of characterization of nanomaterials (HREM, WAXS) with simple techniques of organometallic chemistry may provide useful information on the chemical reactivity of the surface of the nanoparticles and shed light on elementary surface reaction steps such as substitution, oxidative addition and reductive elimination. Such a study may also help designing nanoparticles of defined shape and organization at the mesoscopic level. The NMR studies evidence the processes of dynamic ligand exchange and catalytic coupling of thiols at the surface of metal nanoparticles, and the dynamics of coordinated ligands with the shape of the particles. As a whole, this experiment evidences a rare example of observation of a dynamic process at the surface of metal nanoparticles by solution NMR. The originality of the present system lies in the observation of both the spectra for coordinated ligands and those for the rapid exchange. Ruthenium nanoparticles were stabilized by chiral ligands, namely aminoalcohols and amino-oxazolines. The surface coordination of the ligands leads to an excellent stabilization of the metal particles in solution as well as to an influence on their catalytic performances. Although the catalytic activities are not very good, these nanocatalysts are active under mild conditions and, in some cases, with oxazolines, more efficient but less selective than the corresponding molecular catalysts. This activity and the very low enantiomeric excess obtained compared to those of the molecular systems both suggest a fluxional behaviour of the ligands at the surface of ruthenium as recently observed for amine/Ru systems. This is also confirmed by the decrease in activity and increase in selectivity observed upon addition of excess ligand to the catalytic system. It is interesting to note the very modest but significant and reproducible ee associated with an increase of the oxazoline concentration in solution. Ruthenium nanoparticles included in alumina membranes have been prepared either by dihydrogen decomposition of the complex Ru(COD)(COT) to give pre-prepared colloidal solutions which were further used to fill the pores of nanoporous alumina membranes by vacuum introduction, or by in situ decomposition. These hybrid systems were tested in the hydrogenation of 1,3- butadiene in the gas phase. The use of Ru loaded membranes resulted in considerably increased activities compared with grinded samples. This work appears as a convincing proof for the efficiency of such catalytic systems consisting of nanoparticles included into nanoporous membranes for gas phase catalytic applications. A particularly attractive field of applications would be the catalytic transformation of unwanted gases present in trace quantities either in the atmosphere or as a result of some chemical productions into harmless compounds. Ruthenium nanoparticles have been organised in mesoporous silica. We used different functionalised silica. The organization of the particles inside these material was more or less successful depending on the functional group of the silica. In the case of a phosphonated material we obtained nanowires of Ruthenium inside the channels. This work shows that the organometallic complex Ru(COD)(COT) is a very interesting precursor for the preparation of various nanoparticles whose size and surface state can be controlled. In addition, these particles are active in catalysis. At this date, not many works are available in the literature concerning ruthenium particles.