Les émulsions eau/eau (W/W) ont suscité un grand intérêt en raison de leur potentiel d'application dans différentes industries telles que l'agroalimentaire, les produits pharmaceutiques, les cosmétiques et les soins personnels. Le caractère particulier des émulsions W/W est leur stabilisation par ajout de particules. L’objectif de ce travail de thèse est de comprendre cet aspect en étudiant une émulsion modèle W/W à base de dextran et du poly(oxyde d'éthylène) stabilisée par des particules à base de protéines du lactosérum. Dans un premier temps, nous avons étudié l'effet de la morphologie des particules protéiques et leur partitionnement sur la stabilité des émulsions W/W. En particulier, la stabilité s’est révélée dépendre de la structure des particules quand ses derniers étaient sous forme de microgels, d’agrégats fractals ou de fibrilles. Il a été montré que la stabilité s'améliorait lorsque les particules se localiser préférentiellement dans la phase continue. Deuxièmement, nous avons étudié la gélification, des microgels et des agrégats fractals, induite en réduisant le pH entre 6,5 et 3,5 ou en ajoutant 0,3 M NaCl à pH 7,0 aussi bien quand l’excès des particules se situe dans la phase continue ou dispersée. Dans le premier cas, un réseau se formé dans la phase continue de dextran, permettant d’inhiber le crémage des gouttelettes de PEO, les agrégats fractals étant plus efficaces que les microgels. Dans le second cas, des particules protéiques denses pourraient être formées par gélification des gouttelettes de dextran dispersées. Finalement, nous avons exploré l'adsorption des protéines natives sur les particules de latex et leur capacité à stabiliser les émulsions. / Water/water (W/W) emulsions have attracted great interest recently due to their high potential for applications in different industries such as food and beverages, pharmaceutical, cosmetics and personal care. An important issue is the stabilization of W/W emulsions by adding particles. The aim of the research for this thesis was to shed light on this issue by studying a model W/W emulsion formed by mixing dextran and poly(ethylene oxide) with particles based on whey proteins. Firstly, we studied the effect of the morphology of protein particles and their partitioning on the stability of W/W emulsions. The stability was different when microgels, fractal aggregates or fibrils were added. We showed that stability improved when the particles partitioned to the continuous phase. Secondly, we investigated gelation of the fractal aggregates and microgels induced by reducing the pH between 6.5 and 3.5 or by adding 0.3M NaCl at pH 7.0 with excess particles either in the continuous or he dispersed phase. In the first case, a network was formed in the continuous dextran phase, making it possible to arrest creaming of PEO droplets, fractal aggregates being more effective than microgels. In the second case, dense protein particles could be formed by gelation of the dispersed dextran droplets. Thirdly, we explored the effect of adsorbing native proteins unto latex particles on their capacity to stabilize W/W emulsions.

Parmi les procédés de fabrication continue, l’extrusion par fusion à chaud est une technique dont l’intérêt dans le domaine pharmaceutique est grandissant. Ce procédé permet la formation des dispersions solides des substances actives au sein des matrices polymériques ou lipidiques. En fonction de l’excipient et de la substance active, cela peut être largement utilisé pour la conception des systèmes: (i) pour une libération immédiate, (ii) pour une libération modifiée et (iii) pour le masquage de goût. Les systèmes à libération modifiée sont des dispositifs intéressants qui permettent d’améliorer la biodisponibilité de la substance active, son efficacité ainsi que l’observance des patients. En fonction de la nature de l’excipient, différents systèmes avec des mécanismes de libération variés peuvent être produit, notamment des matrices inerte, érodable ou gonflante. Le poly éthylène oxide est un polymère semi- cristallin et hydrophile qui peut être utilisé pour la libération contrôlée. Son point de fusion compris entre 63 et 67 °C le rend adapté pour l’extrusion. Surtout, ses capacités de gonflement permettent d’administrer la substance active de façon contrôlée en fonction du poids moléculaire du poly éthylène oxide. Les objectifs de ce travail sont (i) d’étudier l’impact des paramètres critiques du procédé (température d’extrusion et vitesse des vis d’extrudeuse) sur le profil de libération de la substance active, (ii) de déterminer l’impact des paramètres de formulations (poids moléculaire du poly éthylène oxide, charge et type de la substance active) sur le profil de libération de la substance active et (iii) d’évaluer des formes galéniques solides conçues par le procédé d’extrusion à celui de la compression directe. Il a été montré que la variation de la température d’extrusion et de la vitesse des vis altérait l’apparence de l’extrudat et ainsi la distribution de la substance active au sein de l’extrudat. Il s’est avéré dans notre étude que la libération de la substance active n’était pas particulièrement affectée par ces changements de température et vitesse de vis de l’extrudeuse. De plus, cette étude a permis de fixer les paramètres pour les projets suivants: température 100 °C ; vitesse des vis 30 rpm ; longueur de la forme galénique 1 cm. Des extrudats de poly ethylène oxide contenant 10 % de théophylline et du poly éthylène oxide de 100 à 7000 kDa ont été utilisés dans ce travail. Il a été observé que lorsque le poids moléculaire du poly ethylène oxide augmente de 100 à 600 kDa, la libération en substance active diminue de façon importante alors qu’une augmentation jusqu’à 7000 kDa ne diminue que légèrement la libération. Des études du gonflement ont montré que ce phénomène corrélait aux variations de volume de la partie opaque de l’extrudat (gel non transparent et cœur solide). / Among continuous manufacturing processes, hot melt extrusion is a technique with growing interest in the pharmaceutical field. This process enables the formation of solid dispersions of many drugs within a polymeric or lipidic carrier. Hot melt extrusion can be widely used for different issues using the appropriate carrier and drug. Here are the mostly used concepts in pharmaceutical solid dosage forms: (i) immediate release, (ii) modified release and (iii) taste masking. Modified release systems have been taken into account to be very interesting devices for the improvement of drug- bioavailability, drug- efficacy as well as the patient compliance. Various systems with different release mechanisms can be manufactured, depending on the nature of the carrier (inert, erodible, and swelling matrices). Poly ethylene oxide is a semi crystalline and hydrophilic polymer which can be used to control drug delivery. The poly ethylene oxide melting point ranging from 63 to 67 °C makes it suitable for hot melt extrusion. Importantly, the swelling capacities of the hydrophilic poly ethylene oxide matrices are able to deliver drug in a time controlled manner, in respect of the poly ethylene oxide molecular weights. The purposes of this work were (i) to study the impact of critical process parameters (extrusion temperature and screw speed) on the drug release behavior, (ii) to determine the impact of formulation parameters (poly ethylene oxide molecular weight, nature of drug and drug loading) on drug release kinetics, and (iii) to evaluate solid dosage forms prepared by hot melt extrusion versus direct compression. Interestingly, the variation of the extrusion temperature and the screw speed leads to the altering of the extrudate appearance and thus the distribution of drug into the extrudate. However, this changing has not influenced the drug release remarkably. Thus, this study was useful to set the parameters for the following projects (temperature 100 °C; screw speed 30 rpm; dosage form size 1 cm). Poly ethylene oxide hot melt extrudates containing 10 % theophylline and based on 100 - 7,000 kDa poly ethylene oxide are used for this thesis. Importantly, the drug release decreased substantially with the increase of the poly ethylene oxide molecular weight from 100 to 600 kDa. However, further increasing of the molecular weights leads to only a slight decrease in the release rate. Swelling studies have shown that this phenomenon correlated with the change in volume of the opaque part of the extrudates (non-transparent gel and solid core).

No description available.

Morphological development during crystallization of thin films of poly(ethylene oxide) (PEO) / poly(methyl methacrylate) (PMMA) blends has been reported. Studies have focused on the effects of the blend composition, PMMA molecular weight, film thickness, and crystallization temperature on the observed crystal morphology. As the blend composition was varied from 90 to 30 wt% PEO, the crystal morphology varied from spherulites to needles and dendrites. Variation of the crystallization temperature and PMMA molecular weight resulted in similar changes in morphology. A morphological map demonstrating the roles of the experimental controls on the observed crystal morphology has been developed. This map was used as a tool for more detailed studies of the observed morphologies and morphological transitions. The dendritic region of the map (~ 30 = 40 wt% PEO) was studied in detail. Changes in the diffusion length were achieved through variation of the PMMA molecular weight, and were shown to influence the secondary sidebranch spacing. Sidebranch spacing measurements revealed that coarsening of the dendritic microstructure occurred well after the competition between diffusion fields of neighboring dendrite arms vanished, indicating the existence of another coarsening mechanism. These studies of dendritic sidebranching indicate that polymer dendrites develop by mechanisms similar to those in small molecules and metals. A number of in-situ observations of morphological transitions have also been reported, including a dense-branched morphology (DBM)/dendrite transition, a DBM/stacked-needle/needle transition, and a transition from dendrites with 90o sidebranching to dendrites with 45o branching or a dense-branched morphology, both of which grow at 45o to the original dendrite trunk. The DBM/dendrite transition occurred over a range of crystallization temperatures, indicating that the transition is not sharp. Crystal growth rate measurements verified this result. The DBM/stacked-needle/needle transitions demonstrated distinct jumps in the crystal growth rate, indicating a change in the growth mechanism or direction. For the transition involving a change in the growth direction, the effective level of noise (fluctuation) was found to be important in morphological selection. The results of this work have helped to define new directions for the study of crystal morphologies, especially in the areas of spherulite formation and dendritic growth. / Ph. D.

Crystalline polymer/salt complexes can conduct, in contrast to the view held for 30 years. The alpha-phase of the crystalline poly(ethylene oxide)₆:LiPF₆ is composed of tunnels formed from pairs of (CH₂-CH₂-O)ₓ chains, within which the Li⁺ ions reside and along which the latter migrate.¹ When a polydispersed polymer is used, the tunnels are composed of 2 strands, each built from a string of PEO chains of varying length. It has been suggested that the number and the arrangement of the chain ends within the tunnels affects the ionic conductivity.² Using polymers with uniform chain length is important if we are to understand the conduction mechanism since monodispersity results in the chain ends occurring at regular distances along the tunnels and imposes a coincidence of the chain ends between the two strands.² Since each Li⁺ is coordinated by 6 ether oxygens (3 oxygens from each of the two polymeric strands forming a tunnel), monodispersed PEOs with the number of ether oxygen being a multiple of 3 (NO = 3n) can form either “all-ideal” or “all-broken” coordination environments at the end of each tunnel, while for both NO = 3n-1 and NO = 3n+1 complexes, both “ideal” and “broken” coordinations must occur throughout the structure. A synthetic procedure has been developed and a series of 6 consecutive (increment of EO unit) monodispersed molecular weight PEOs have been synthesised. The synthesis involves one end protection of a high purity glycol, functionalisation of the other end, ether coupling reaction (Williamson's type ether synthesis³), deprotection and reiteration of ether coupling. The parameters of the process and purification methods have been strictly controlled to ensure unprecedented level of monodispersity for all synthesised samples. Thus obtained high purity polymers have been used to study the influence of the individual chain length on the structure and conductivity of the crystalline complexes with LiPF₆. The results support the previously suggested model of the chain-ends arrangement in the crystalline complexes prepared with monodispersed PEO² over a range of consecutive chain lengths. The synthesised complexes constitute a series of test samples for establishing detailed mechanism of ionic conductivity. Such series of monodispersed crystalline complexes have been studied and characterised here (PXRD, DSC, AC impedance) for the first time. References: 1. G. S. MacGlashan, Y. G. Andreev, P. G. Bruce, Structure of the polymer electrolyte poly(ethylene oxide)₆:LiAsF₆. Nature, 1999, 398(6730): p. 792-794. 2. E. Staunton, Y. G. Andreev, P. G. Bruce, Factors influencing the conductivity of crystalline polymer electrolytes. Faraday Discussions, 2007, 134: p. 143-156. 3. A. Williamson, Theory of Aetherification. Philosophical Magazine, 1850, 37: p. 350-356.

Ce travail est consacré à l'étude d'origine et signatures de la dépendance en concentration $\phi$ du paramètre de Flory $\chi _(eff)$ pour les polymères neutres solubles dans l'eau. Les trois versions du modèle à deux états qui ont été proposés pour expliquer le comportement de phase de Poly(oxyéthylène) dans l'eau mènent à $\chi _(eff)(\phi) $. Cette dépendance peut servir à tester la qualité des modèles: elle permet de distinguer entre les différents modèles et juger de la pertinence des paramètres impliqués. L'apparition de $\chi _(eff)(\phi)$ à partir des modèles alternatifs à un seul état est aussi discutée. Les signatures macroscopiques de $\chi _(eff)(\phi)$ sont illustrées avec l'exemple du gonflement des chaînes isolées et des brosses planes. La dépendance de $\chi _(eff)$ avec $\phi$ dans les systèmes à concentration uniforme a deux conséquences principales: entraînement du déplacement du croisement entre les régimes Gaussien et auto-évitant, et la possibilité d'une transition de phase du premier ordre pour les chaînes flexibles isolées. $\chi _(eff)(\phi)$ peut provoquer une séparation verticale de phase dans des brosses polymères qui implique la coexistence de deux phases de concentration finie. Cette approche est appliquée à l'interprétation d'un ensemble des résultats expérimentaux qui suggère que cet effet pourrait exister avec les brosses de Poly(N-isopropylacrylamide).

Nous reportons ici le comportement rhéologique de suspensions concentrées de particules non browniennes de fécule de maïs ou de polystyrène sphériques et monodisperses. Le système modèle de particules de polystyrène est utilisé afin de contrôler les propriétés physicochimiques de la suspension. Ce travail expérimental s'articule autour de deux aspects rencontrés chez les fluides complexes : le seuil d'écoulement et le seuil de rhéoépaississement. <br />Dans un premier temps, nous avons étudié l'origine du seuil d'écoulement. Nous avons, à partir de techniques de rhéologie classique et locale par IRM, montré que le seuil d'écoulement des suspensions modèles n'est observé que par l'introduction d'un contraste de densité entre les particules et le fluide suspendant. Ce contraste de densité introduit alors une consolidation à l'échelle des particules. Le seuil d'écoulement apparaît alors sans aucune sédimentation macroscopique de l'échantillon. Dans un second temps, nous avons étudié un phénomène de structuration dynamique, sous écoulement : le rhéoépaississement. Ainsi, en couplant des mesures macroscopiques originales (comme l'étude de l'influence de l'entrefer de la géométrie de mesure sur le comportement rhéoépaississant, et des mesures de dilatation volumique en fonction du taux de cisaillement) et des mesures locales des propriétés d'écoulement par IRM, nous montrons que le rhéoépaississement de la suspension de fécule de maïs est une conséquence directe de la dilatance de Reynolds. Par ailleurs, le système modèle, contrairement à la suspension de fécule de maïs, présente un rhéoépaississement qui est accompagné d'un régime de localisation qui devient responsable d'un blocage partiel de l'écoulement dans l'entrefer de la géométrie de Couette.

A more quantitative understanding of peptide loading and release from polyethylene oxide (PEO) brush layers will provide direction for development of new strategies for drug storage and delivery. The antimicrobial peptide nisin shows potent activity against Gram-positive bacteria including the most prevalent implant-associated pathogens, its mechanism of action minimizes the opportunity for the rise of resistant bacteria and it does not appear to be toxic to humans, suggesting good potential for its use in antibacterial coatings for selected medical devices. In this work, optical waveguide lightmode spectroscopy was used to record changes in adsorbed mass during cyclic adsorption-elution experiments with nisin, at uncoated and PEO-coated surfaces. PEO layers were prepared by radiolytic grafting of Pluronic® surfactant F108 or F68 to silanized silica surfaces, producing long- or short-chain PEO layers, respectively. Kinetic patterns were interpreted with reference to a model accounting for history-dependent adsorption, in order to evaluate rate constants for nisin adsorption and desorption, as well as the effect of pendant PEO on the lateral clustering behavior of nisin. Lateral rearrangement and clustering of adsorbed nisin was apparent on uncoated and F68-coated surfaces, but not on F108-coated surfaces. In addition, nisin showed greater resistance to elution by peptide-free buffer from uncoated and F68-coated surfaces. These results are consistent with shorter PEO chains allowing for peptide adsorption to the base substrate in the case of F68-coated surfaces, while adsorption to the F108-coated surfaces is apparently governed by the presence of a hydrophobic core within the brush layer itself. Further, these results suggest that while peptide location within the hydrophobic core provides stability against lateral rearrangement, the pendant PEO chains themselves provide no steric barrier to nisin rearrangement within the brush layer. / Graduation date: 2012

The marked increase in surface-to-volume ratio associated with microscale devices for hemodialysis leads to problems with hemocompatibility and blood flow distribution that are more challenging to manage than those encountered at the conventional scale. In this work, stable surface modifications with pendant polyethylene oxide (PEO) chains were produced on polycarbonate microchannel and polyacrylonitrile membrane materials used in construction of microchannel hemodialyzer test articles. These coatings were evaluated in relation to protein repulsion, impact on urea permeability through the membrane, and impact on bubble retention through single-channel test articles. PEO layers were prepared by radiolytic grafting of PEO-PBD-PEO (PBD = polybutadiene) triblock copolymers to microchannel and membrane materials. Protein adsorption was detected by measurement of surface-bound enzyme activity following contact of uncoated and PEO-coated surfaces with ��-galactosidase. Protein adsorption was decreased on PEO-coated polycarbonate and polydimethyl siloxane (PDMS) materials by 80% when compared to the level recorded on uncoated materials. Protein adsorption on membrane materials was not decreased with PEO-PBD-PEO treatment; a PEI (polyethylene imide) layer exists on the AN69 ST membrane which is intended to trap heparin during membrane pre-treatment. It is still unclear how this PEI layer interacts with PEO-PBD-PEO. Neither the PEO-PBD-PEO triblocks nor the irradiation process was observed to have any effect on polyacrylonitrile membrane permeability to urea, nor did the presence of additional fibrinogen and bovine serum albumin (BSA) in the urea filtrate. The PEO-PBD-PEO treatment was not able to visibly reduce bubble retention during flow through single-channel polycarbonate test articles, however, the rough surfaces of the laser-etched polycarbonate microchannels may be causing this bubble retention. This surface treatment holds promise as a means for imparting safe, efficacious coatings to blood processing equipment that ensure good hemocompatibility and blood flow distribution, with no adverse effects on mass transfer. / Graduation date: 2013

Nisin, an amphiphilic, antimicrobial peptide, has been shown to integrate into the hydrophobic inner region of poly(ethylene oxide) (PEO) brush layers; however, the presence of integrated nisin may compromise the protein repulsive character of the PEO layer. In particular, the introduction of fibrinogen to nisin-loaded brush layers has been observed to cause changes consistent with partial elution of nisin and/or location of fibrinogen at the interface. Questions surrounding the possibility of fibrinogen adsorption warrant further investigation, as the location of procoagulant proteins at a peptide-loaded PEO layer would significantly reduce the viability of a medical device coating based on such an approach. In this work, the preferential location of fibrinogen at PEO brush layers was investigated by: detection of FITC-labeled fibrinogen after sequential introduction of nisin and labeled fibrinogen; measurement of changes in the zeta potential of PEO coated and uncoated surfaces following nisin, fibrinogen, and/or buffer challenges; and evaluation of adsorption and elution kinetics in label-free, sequential adsorption experiments using optical waveguide lightmode spectroscopy (OWLS). PEO layers were constructed through radiolytic grafting of Pluronic�� F108 or F68 onto silanized silica surfaces producing long-chain or short-chain PEO layers, respectively. Adsorption results indicated that sequential introduction of nisin and fibrinogen to PEO brush layers, based on F108, does not result in fibrinogen adsorption beyond that expected for a nisin-free PEO layer. No evidence of nisin entrapment in fibrinogen-repellent F68 layers was recorded. Low-level fibrinogen adsorption observed at F68 layers following the introduction of nisin was determined to be a result of nisin adsorption at (uncoated) defect regions on the surface. In conclusion, retention of PEO layer capacity for protein repulsion after nisin entrapment is owing to a steric repulsive barrier provided by PEO segments extending beyond the level of entrapped nisin. It was then hypothesized that the immobilized, pendant PEO chains will inhibit exchange of entrapped nisin by competing proteins, and therefore prolong nisin activity retention. In order to evaluate nisin function following its entrapment, the antimicrobial activity of nisin-loaded, F108-coated silica surfaces was evaluated against the Gram-positive indicator strain, Pediococcus pentosaceous. The retained biological activity of these nisin-loaded layers was evaluated after incubation in the presence of bovine serum albumin (BSA), for contact periods up to one week. Surfaces were withdrawn at selected times and placed on plates inoculated with P. pentosaceous to measure kill zone radius in order to quantify nisin activity. In the presence of BSA, F108-coated surfaces retained more antimicrobial activity than the uncoated, hydrophobic surfaces. These results strongly suggest that PEO brush layers may serve as a viable drug storage platform due to the retained non-fouling character after bioactive peptide entrapment and the prolonged peptide activity in the presence of other proteins. / Graduation date: 2013