La structuration de l'innovation dans les industries Science-based sous forme de réseaux est devenue depuis une dizaine d'années la nouvelle dynamique organisationnelle de l'innovation. Plus spécifiquement, dans l'industrie biopharmaceutique, cette nouvelle dynamique organisationnelle renvoie vers une logique d'efficience et de rationalisation des efforts de R&D. La nature de l'innovation et les interdépendances technologiques entre les différents acteurs sont autant des éléments qui poussent à cette organisation en réseaux d'innovation. Dans cette perspective, l'étude du processus d'émergence, structuration et évolution des clusters d'innovation de l'industrie biopharmaceutique nous permet de comprendre, en plus des avantages liés à la proximité spatiale, d'autres avantages liés à des formes de proximités organisationnelles et stratégiques. Cette thèse tente aussi de répondre à la question des facteurs indispensables à l'émergence et le développement de ces clusters d'innovation et ça à travers une étude de plusieurs cas d'émergence de clusters d'innovation. Dans notre étude empirique nous avons choisi d'analyser cette structuration en termes de clusters d'innovation pour deux régions en France. A travers l'approche des projets collaboratifs nous sommes arrivés à reconstruire les deux réseaux et identifier le rôle des grands acteurs dans la définition de la morphologie du réseau et la transmission et diffusion des connaissances au sein de ce dernier. Les résultats obtenus montrent aussi une structuration en réseaux à plusieurs échelles. Le travail réalisé contribue à la compréhension de la dynamique d'interaction au sein des réseaux de R&D dans les secteurs sciences-based / The structuring of innovation in science-based industries in the form of networks has become the new organizational dynamics of innovation over the past decades. More specifically, in the biopharmaceutical industry, this new organizational dynamic refers to the logic of efficiency and rationalization efforts in R&D. The nature of innovation and technological interdependencies between different actors are all elements that drive the organization in innovation networks. In this perspective, the study of the process of emergence, structuration and evolution of clusters of innovation in the biopharmaceutical industry enables us to understand the benefits of spatial proximity (knowledge spillovers, shared infrastructure...) and other benefits of organizational and strategic proximity. This thesis also attempts to answer the question of essential factors in the emergence and development of these innovation clusters and that through several case studies of the emergence of clusters of innovation in the biopharmaceutical industry in the United States and Europe.In our empirical study, we chose to analyze this structure in terms of innovation clusters for two regions in France, Ile-de-France and the Lyon region. Through the approach of collaborative projects, we are come to rebuild the two networks and identify the role of some major players in defining the morphology of the network and the transmission and dissemination of knowledge within it. The results also show a structure in networks at multiple scales. The work done contributes to the understanding of the dynamics of interaction within networks of R&D in science-based sectors.

No description available.

This thesis aims to better constrain the high-mass end of the globular cluster system (GCS) mass - Halo mass relation. This relation between the total mass contained in the globular clusters in a galaxy and the mass of its dark matter halo has been found to be nearly linear. The measured mass ration between the mass of the GCS and the halo mass, defined as eta, has been found to be constant over a large range of galaxy masses, however there is comparatively less data for galaxies with dark matter halo masses >10^13 solar masses. This research analyzes the GCSs of a sample of eleven high-mass brightest cluster galaxies (BCGs) through the use of the photometry program DOLPHOT on Hubble Space Telescope (HST) images in the F814W filter. The mass of the galaxies' GCSs were calculated from their GC radial distributions, and the dark matter halo masses were determined from the known relationship between total stellar mass and halo mass for BCGs. This research utilizes a new standardization technique to calculate the size of the GCS based on the galaxy's virial radius. These GC and halo masses were then compared to determine eta at this high-mass end, which was found to be eta=(6.84+/- 1.47)\times10^-5. When adding the sample BCGs to a catalogue of 303 lower-mass galaxies a total value for eta was found to be eta=(2.99 +/- 0.06)\times10^-5, which is within the literature range for of eta, and is both slightly higher and has a reduced uncertainty compared to the value for the catalogue galaxies alone which was found to be eta=(2.87 +/- 0.11)\times10^-5. This consistency between the eta-values of high-mass and lower-mass galaxies implies that in order for these BCGs to have such massive GCSs, a large proportion of these GCs must be accreted from galaxy mergers at late redshift. / Thesis / Master of Science (MSc)

Star cluster formation involves the conversion of molecular gas into stars inside giant molecular clouds (GMCs). Such a process involves many dynamical evolution mechanisms, including mergers between smaller star clusters (subclusters) on which we focus in this thesis. We take results of simulations performed by Howard et al. 2018 (H18) which found that young massive cluster (YMC) formation is heavily dependant on the process of subcluster mergers, and we simulate said mergers at higher resolution. Subclusters inside such GMC simulations are modelled using the sink particle prescription which does not resolve individual star particles or gas parcels inside the subcluster they represent. We employ a more controlled method in simulating subcluster mergers to better understand the response of the stellar and gas components of a subcluster from the merger process. To do this, we take the parameters of the sink particles created in H18 and set up spheres of stars and gas. We use the AMUSE framework to couple the N-body evolution of the stars to the smoothed particle hydrodynamics (SPH) evolution of the gas such that both components of a given cluster can realistically react to each other. We model 15 of these mergers and find that once the velocity at which the two clusters collide (collisional velocity) exceeds $\approx 10$kms$^{-1}$, the resultant cluster is not monolithic (i.e. it still contains two separate stellar components) while all other simulations merge into one monolithic stellar and gas component cluster. We also find that, regardless of the collisional velocity of masses of the component clusters, all resultant clusters lose a fraction of their stellar and gas mass. This fraction is directly proportional to the collisional velocity and is a discrepancy between the sink particle prescription (where all mass is contained inside a constant sink particle accretion radius) and real cluster mergers. A further discrepancy we find is that all simulations result in a cluster whose outermost regions are expanding and that the rate of this expansion is somewhat proportional to the collisional velocity of the merger. These results point to the inaccuracy of the sink particle prescription and allow us to develop tools to improve on it in future simulations. Next, we fit commonly used analytical density profiles to both the stellar and gas component of our resultant clusters and find that, while they do not provide particularly excellent fits, they provide constraints on what is an acceptable fit. Lastly, we analyze the amount by which gas with potentially star forming densities increase due to the merger and we find that all mergers increase their star forming gas mass fraction by roughly 50 per cent implying that mergers may be an effective tool for triggering star formation. / Thesis / Master of Science (MSc)

No description available.

1 volume

Yes / The paper examines the views, often associated with Porter, that clusters with deep collaborative networks and established local supply chains have good performance. The view that good cluster performance is not connected to the industrial sector is also assessed. Data from a Department of Trade and Industry (DTI) study on UK clusters are used to assess the impact on performance (employment growth and international competitiveness) of cluster depth, the stage of development of local supply chains, and industrial sector. The results of the analysis of the DTI data on clusters do not provide strong support for Porter-type views on cluster policy. Although established clusters are linked to employment growth, deep clusters are not associated with employment growth or international competitiveness, and clusters in the services, and media, computer-related and biotechnology sectors are more likely than manufacturing clusters to have good performance. Some of the major policy implications of the results are discussed in the light of the literature on the importance of regional, national, and international networks for the performance of clusters.

De nos jours, réussir à contrôler la microstructure de la phase active des catalyseurs hétérogènes reste encore un défi d’envergure. Le présent travail s’inscrit dans cette mouvance en proposant de valoriser l’utilisation de clusters moléculaires comme précurseurs métalliques. Ces entités organométalliques comprenant au moins trois atomes métalliques (identiques ou non) en leur cœur apparaissent comme des candidats de choix pour contrôler la taille et la composition chimique des particules métalliques supportées à l’échelle nanoscopique. L’objectif de la présente thèse est de déterminer si l’utilisation de clusters moléculaires en présence d’un support carboné approprié permet d’obtenir ce contrôle. Tout d’abord, des clusters carbonylés de palladium de la littérature ont été synthétisés et une nouvelle voie de synthèse en une étape de ces entités a été découverte. Ensuite, le support carboné de départ a été dérivatisé afin d’y introduire des sites d’ancrage pour les clusters (phosphines chélatantes). Dans la troisième partie du travail, les clusters synthétisés ont été incorporés sur le support fonctionnalisé et sur celui de départ. L’obtention de nanoparticules supportées bien dispersées a démontré l’intérêt d’utiliser des clusters moléculaires comme précurseurs en présence d’un support fonctionnalisé. Les mécanismes d’ancrage ont également été étudiés et il a été prouvé que les clusters se fixent au support de manière covalent par échange de ligands. Enfin, les clusters supportés ont été activés et les catalyseurs Pd/C formés ont été testés dans l’hydrogénation du nitrobenzène en aniline. Les catalyseurs se trouvent être actifs dans cette réaction mais aucune corrélation « structure-activité » n’a pu être mise en évidence.

Made available in DSpace on 2016-04-15T00:46:24Z (GMT). No. of bitstreams: 2 Mario Fernandes da Costa1.pdf: 2190710 bytes, checksum: 2ab31184ac3f5e136b1cca3fe610ca39 (MD5) Mario Fernandes da Costa2.pdf: 2712653 bytes, checksum: 418d75699b624bfa642648479839eea3 (MD5) Previous issue date: 2010-01-28 / Fundo Mackenzie de Pesquisa / The aim of this study is to identify the existence of clusters of logistics services providers in Brazil. This clustering of companies, and the competitive advantages deriving therein, are matters that have received increasing attention from those engaged in the study of areas such as industrial economy, regional economy, and geography and economics. The geographic and sector-related concentration of companies is capable of providing local savings to the participants, who benefit from the increase in competitiveness. In spite of the importance of logistics services nowadays, there is evident a dearth of studies focused upon localization and the relations between the participating companies which form part of the logistics services provider chains. Within this context, this study has been based on cluster literature, using geographic concentration indicators, such as the Herfindhal index, and that of regional specialization, such as the locating quotient, which is used to identify the location of the agglomerations of companies providing logistics services. The RAIS/MTE database was used in 2006 and 2007. The results of this study identify the existence of 929 clusters present in 617 Brazilian municipalities. Of this total, some 41% of the occurrences have taken place in the Southeastern Region, and 33% in the Southern Region of Brazil, where the majority of these clusters are concentrated. All together, the States of São Paulo, Rio Grande do Sul, Minas Gerais, Santa Catarina, and Paraná host some 66% of the total clusters. / O objetivo deste trabalho é identificar a ocorrência de aglomerações de prestadores de serviços de logística no Brasil. A aglomeração de empresas, e as vantagens competitivas dela advindas são questões que receberam atenção crescente dos estudiosos de áreas como economia industrial, economia regional e geografia econômica. A concentração geográfica e setorial das empresas é capaz de proporcionar economias externas locais aos participantes, que se beneficiam pelo aumento de competitividade. Apesar da importância dos serviços logísticos na atualidade, nota-se uma lacuna de estudos voltados para a localização e as relações entre as empresas participantes da cadeia de prestadores de serviços de logística. Nesse sentido, este trabalho se apoia na literatura de aglomerações, utilizando indicadores de concentração geográfica, como o índice Herfindahl, e de especialização regional, como o quociente locacional, para a identificação da localização de aglomerações de empresas prestadoras de serviços de logística. A bas de dados utilizada foi a RAIS/MTE nos anos de 2006 e 2007. Os resultados deste trabalho identificam a existência de 929 aglomerações presentes em 617 Municípios no Brasil. Deste total, cerca 41% das ocorrências são verificadas na Região Sudeste e 33% na Região Sul do Brasil onde estão concentradas a maioria das aglomerações. Os Estados de São Paulo, Rio Grande do Sul, Minas Gerais, Santa Catarina e Paraná somados detém cerca de 66% das aglomerações totais.

Reaction of the tetracobalt cluster Co4(CO)10(t 4-PPh)2 with 1,2-bis(dimethylphosphino)ethane (dmpe) affords the bis-substituted cluster Co4(CO)8(dmpe)(t 4-PPh)2. The bidentate dmpe ligand is shown to bind to the cluster in a chelating fashion by IR, NMR, and X-ray diffractions analyses. The fluxional nature of the ancillary carbonyl groups has been studied by variable temperature 13C NMR measurements which reveal two distinct carbonyl scrambling pathways. The stability of the phosphine-ligated cluster has been examined using in situ Cylindrical Internal Reflection (CIR) Spectroscopy. The effect of the dmpe ligand on the cluster polyhedron will be discussed with respect to the observed crystallographic and spectroscopic results