The aim of the project is to investigate how a more viscous, gel like formulation can be made of BioGaia´s Prodentis Drops, which is a probiotic product for oral use. The two different strains of Lactobacillus reuteri that are used in the product, together called L. reuteri Prodentis, have been clinically proven to be effective in treatment of gingivitis and caries formation. The existing product is a highly liquid oil suspension that has been described as too runny and difficult to get into tooth pockets and between teeth. Therefore a gel formulation would be preferred. Pre-trials were excecuted to see what combinations and quantities of ingredients could work. Three different formulations using 3 %, 5 % and 7 % beeswax as thickening agent proceeded to another round of trials, as well as one formulation where the original oil was exchanged for hydrogenated rapeseed oil. In the beeswax formulations fumed silicon dioxide was being used as well. Three different analyses were executed, considering bacterial survival, viscosity and phase separation of gels. The bacterial survival proved to be acceptable in all samples even after 14 days of incubation in 37oC. The formulation with 7 % beeswax was the most viscous one, followed by 5 % beeswax, 3 % beeswax and the formulation with hydrogenated oil, respectively. Phase separation could be seen in the hydrogenated oil formulation already after seven days and even more so after 14 days. There were also signs of separation in the formulation with 3 % beeswax after 14 days. It was concluded that in further development of the Prodentis Drops it is recommendable to proceed with the 5 % beeswax formulation.

Cette thèse concerne l'étude de la dynamique d'impacts de gouttes, dans des situations de friction réduite entre le substrat solide et la goutte liquide. Cette diminution de friction s'est faite au moyen d'un film d'air inséré entre le liquide et le solide. Il existe plusieurs stratégies permettant l'existence de ce film d'air : la première est d'utiliser le phénomène de caléfaction, ou effet Leidenfrost : un liquide approché d'une surface chauffée au delà d'une température critique s'évapore suffisamment rapidement pour pouvoir léviter sur sa propre vapeur, et ainsi être isolé de la surface solide. Dans certaines conditions, les surfaces super-hydrophobes micro-texturées permettent au liquide de rester dans un état "fakir", c'est à dire de n'être en contact qu'avec le sommet de micro-piliers, le reste du liquide demeurant au dessus d'un coussin d'air. Enfin, il a également été constaté que l'écoulement d'air engendré par le mouvement d'une surface solide peut induire une force de portance sur une goutte, et ainsi lui permettre de léviter au dessus de cette surface / In this thesis we study the dynamic of drop impacts, in situations of low friction between the liquid and the solid surface. This low friction can be obtained thanks to an air cushion trapped between the liquid and the solid, which can be achieved by several ways. The first one is the Leidenfrost effect : when a liquid is moved close to a hot surface, the evaparation rate can be sufficient make liquid levitate on its own vapour. In certain conditions, onto micro-patterned super-hydrophobic surfaces, a drop can be in a "fakir" state, which means that the contact is limited to the top of micro-pillars, the rest of the liquid is at the top of an air cushion. Finally, we also observed that the air flow due to a moving surface can generate a lift force which can permit the levitation of the drop

Cette thèse traite des instabilités hydrodynamiques survenant dans les ligaments liquides en présence d’écoulement axial, afin de comprendre la formation des gouttes lors des processus d’atomisation. Il a été montré par des études précédentes que tout processus d’atomisation passe par l’étape de la formation des ligaments, qui se déstabilisent en gouttelettes. Le processus de fragmentation des ligaments est régi par deux instabilités : l’instabilité de Rayleigh-Plateau lorsque le ligament est infini, le phénomène du « end pinching », lorsque le ligament est semi-infini. Dans les deux cas le mécanisme responsable du pincement du ligament est la capillarité. Dans cette thèse, nous avons montré que, sous certaines conditions, le pincement du ligament peut être retardé ou évité : c’est le phénomène d’évitement de l’étranglement. Une exploration détaillée de la zone de constriction du ligament (là où le rayon est minimal), révèle des effets non linéaires liés à la viscosité du fluide : la couche visqueuse se développe, s’enroule puis décolle de l’interface sous forme de jet, en aval du cou avec formation d’anneau tourbillonnaire. L’écoulement dans cette zone du ligament est soumis à une accélération axiale correspondant à une baisse de la pression : c’est l’effet Venturi. Ce type d’écoulement a été étudié à travers l’instabilité du pont capillaire entre deux tubes, soumis à un écoulement axial. Tout au long de ces travaux, deux approches ont été utilisées : les simulations numériques et des observations expérimentales. L’essentiel des résultats présentés a fait l’objet de publication ou d’une soumission d’article. / This thesis deals with the hydrodynamic instabilities which occur in a liquid ligament in presence of axial flow. We investigate the formation of drops during the atomization process. Previous studies highlighted a common step to all types of atomization processes: the initial formation of the ligament which later evolves into droplets. The ligaments fragmentation process is governed by two possible instabilities: the “Rayleigh-Plateau” instability which is characteristic of an infinite ligament and the “end-pinching” phenomenon, which occurs in semi-infinite ligaments. In both cases, capillarity is the driving mechanism underlying the ligament segmentation. In this thesis we show that, under certain conditions, the liquid ligament can surprisingly escape from pinch-off through creation of a vortex ring (“escape phenomenon”). A detailed analysis of the constriction zone (neck of the ligament) during the “escape phenomenon” suggested that nonlinear effects associated to fluid viscosity might play an important role in the escape process. Both our experimental observations and numerical results confirmed the occurrence of the detachment of the viscous layer into a jet downstream of the neck through creation of a vortex ring, when fluid viscosity exceeds a threshold. Accordingly, the fluid in the constriction zone undergoes an axial acceleration associated to a decrease in the pressure: this is the so-called “Venturi effect”. This type of flow is characteristic of the instability which emerges at the capillary bridge between two tubes subjected to axial flow. Part of the results obtained were the subject of a publication or article submission.

Les écoulements de trains périodiques de gouttes monodisperses confinées dans des microcanaux sont largement utilisés pour diverses applications haut-débits en microfluidique digitale. Pour cela, il est nécessaire de réaliser différentes opérations sur ces gouttes comme les fragmenter, les fusionner ou les trier. Dans ce manuscrit, nous discutons de trois de ces opérations expérimentalement et théoriquement. La première concerne l’encapsulation d’un train de gouttelettes dans des gouttes. Nous étudions la dynamique d’encapsulation et nous présentons une nouvelle méthode d’encapsulation. Par suite, nous investiguons deux modes de fragmentation de gouttes, tous deux influencés par des interactions hydrodynamiques entre gouttes consécutives dans un train. Enfin, nous cherchons à comprendre la sélection du chemin suivi par des gouttes à des bifurcations successives. / Flows of periodic trains of monodisperse drops confined in microchannels are widely used for numerous high-throughput applications in digital microfluidics. The development of such applications requires performing and combining various operations on these drops like breakup, fusion or sorting. In this manuscript, we study experimentally and theoretically three of these operations. We first discuss the encapsulation of a train of droplets inside drops, focusing on the encapsulation dynamics. Also, we present a new way to encapsulate drops to produce double emulsions. We then investigate two ways to break drops against micro-obstacles, both being influenced by hydrodynamics interactions between two consecutives drops in a train. Lastly, we report the investigation of the path selection of drops at successive bifurcations.

Um dos grandes desafios no desenvolvimento de fórmulas farmacêuticas e cosméticas é a adequação de seus sistemas conservantes. No presente trabalho, empregou-se método de otimização destes para suspensão oftálmica de dexametasona e sulfato de polimixina B. O experimento foi conduzido utilizando-se planejamento estatístico do tipo simplex-lattice. A matriz de ensaio contemplou 17 fórmulas sendo que as variáveis independentes foram as concentrações de conservantes álcool feniletílico (X1) e digluconato de clorhexidina (X2) e EDTA (X3). A variável dependente ou resposta foi o valor D obtido do desafio das fórmulas com Pseudomonas aeruginosa, Pseudomonas cepacia, Staphy/ococcus aureus, Candida albicans e Aspergillus niger, empregando-se nas determinações, o Método de Regressão Linear. A análise estatística dos resultados, para cada microrganismo desafiante, conduziu à equações que expressaram matematicamente o fenômeno observado. Como conseqüência, foi possível o cálculo teórico do valor D em função das concentrações das variáveis independentes. Além das equações, obtiveram-se as superfícies de resposta e gráficos de contorno correspondentes aos ensaios, para cada germe. Para a seleção da fórmula ou região que atendesse o critério adotado e que conciliasse as condições ideais e factíveis (D ≤ 4 horas, para bactérias e D ≤ 28 horas, para fungos) utilizaram-se duas estratégias: gráfica e numérica. A primeira consistiu na sobreposição dos gráficos de contorno resultando em região da qual selecionou-se, aleatoriamente, a fórmula empregando X1 = 0,10; X2 = 0,80; X3 = 0,10. A segunda, baseou-se na aplicação da função \"desejo\", tendo sido a fórmula ideal resultante de cálculos desenvolvidos, a partir de premissas, sendo: X1 = 0,25; X2= 0,75; X3= 0,0. Ambas as fórmulas, derivadas das estratégias empregadas, foram submetidas à avaliação de seus sistemas conservantes 232 pelo método de Regressão Linear. Os valores D obtidos foram semelhantes àqueles calculados por meio das equações anteriormente mencionadas, exceto para Staphy/ococcus aureus. A aplicação de métodos de otimização estatística permitiu, de forma racional, atingir o objetivo fixado. Através dessas técnicas, tornou-se possível a definição de regiões experimentais em melhores condições, demonstrando vantagens no uso do álcool feniletílico, bem como um melhor conhecimento dos mecanismos envolvidos. / One of the greatest challenges in pharmaceutical and cosmetic formulations consist of the development of adequate preservative systems. Whose optimized development was preposed for ophthalmic suspension of dexametasone and polymyxin B. This was performed threugh simplex-lattice method. The matrix of essay completed 17 formulas, from which independent variables were the preservative concentrations: phenylethanol (X1), chlorhexidine digluconate (X2) and EDTA (X3). The dependent variable was the D-value obtained from the challenge of those formulas with Pseudomonas aeruginosa, Pseudomonas cepacia, Staphhylococcus aureus, Candida albicans and Aspergillus niger, ali result of the Linear Regression Method. The statistical analysis employment led to polynomial expressions which expressed mathematically the observed behavior in each organismo As a consequence the theoretical calculation of the D-value considering ali the independent variables was possible. Besides these expressions, the response surfaces and the contour graphics corresponding to the essays, for each microorganism, were obtained. To the formula or region selection which better complied to the adopted criteria and conciliated ideal and realistic conditions (D ≤ 4 hours, for bacteria and D ≤ 28 hours, for fungi) both graphic and numeric strategies were used. The former consisted of the superposition of contour graphics of each micreorganism, what resulted in a region from which the formula was randomly selected using X1 = 0,10; X2 = 0,80; X3 = 0,10. The latter was based on the application of \"wish\" function, and the ideal formula obtained from calculations developed from these premisses: X1 = 0,25, X2 = 0,75; X3 = 0,0. Both formulas, derived from the employed strategies were submitted to evaluation of their preservative systems by Linear Regression Method. The D-values obtained were similar to those calculated through the previously mentioned expressions, except for Staphylococcus aureus. The application of optimized statistical methods enabled us, in a rational way, to achieve the fixed goal. Through these techniques, both experimental regions in better conditions, demonstrating advantages of using of the involved mechanisms were attained.

Um dos grandes desafios no desenvolvimento de fórmulas farmacêuticas e cosméticas é a adequação de seus sistemas conservantes. No presente trabalho, empregou-se método de otimização destes para suspensão oftálmica de dexametasona e sulfato de polimixina B. O experimento foi conduzido utilizando-se planejamento estatístico do tipo simplex-lattice. A matriz de ensaio contemplou 17 fórmulas sendo que as variáveis independentes foram as concentrações de conservantes álcool feniletílico (X1) e digluconato de clorhexidina (X2) e EDTA (X3). A variável dependente ou resposta foi o valor D obtido do desafio das fórmulas com Pseudomonas aeruginosa, Pseudomonas cepacia, Staphy/ococcus aureus, Candida albicans e Aspergillus niger, empregando-se nas determinações, o Método de Regressão Linear. A análise estatística dos resultados, para cada microrganismo desafiante, conduziu à equações que expressaram matematicamente o fenômeno observado. Como conseqüência, foi possível o cálculo teórico do valor D em função das concentrações das variáveis independentes. Além das equações, obtiveram-se as superfícies de resposta e gráficos de contorno correspondentes aos ensaios, para cada germe. Para a seleção da fórmula ou região que atendesse o critério adotado e que conciliasse as condições ideais e factíveis (D ≤ 4 horas, para bactérias e D ≤ 28 horas, para fungos) utilizaram-se duas estratégias: gráfica e numérica. A primeira consistiu na sobreposição dos gráficos de contorno resultando em região da qual selecionou-se, aleatoriamente, a fórmula empregando X1 = 0,10; X2 = 0,80; X3 = 0,10. A segunda, baseou-se na aplicação da função \"desejo\", tendo sido a fórmula ideal resultante de cálculos desenvolvidos, a partir de premissas, sendo: X1 = 0,25; X2= 0,75; X3= 0,0. Ambas as fórmulas, derivadas das estratégias empregadas, foram submetidas à avaliação de seus sistemas conservantes 232 pelo método de Regressão Linear. Os valores D obtidos foram semelhantes àqueles calculados por meio das equações anteriormente mencionadas, exceto para Staphy/ococcus aureus. A aplicação de métodos de otimização estatística permitiu, de forma racional, atingir o objetivo fixado. Através dessas técnicas, tornou-se possível a definição de regiões experimentais em melhores condições, demonstrando vantagens no uso do álcool feniletílico, bem como um melhor conhecimento dos mecanismos envolvidos. / One of the greatest challenges in pharmaceutical and cosmetic formulations consist of the development of adequate preservative systems. Whose optimized development was preposed for ophthalmic suspension of dexametasone and polymyxin B. This was performed threugh simplex-lattice method. The matrix of essay completed 17 formulas, from which independent variables were the preservative concentrations: phenylethanol (X1), chlorhexidine digluconate (X2) and EDTA (X3). The dependent variable was the D-value obtained from the challenge of those formulas with Pseudomonas aeruginosa, Pseudomonas cepacia, Staphhylococcus aureus, Candida albicans and Aspergillus niger, ali result of the Linear Regression Method. The statistical analysis employment led to polynomial expressions which expressed mathematically the observed behavior in each organismo As a consequence the theoretical calculation of the D-value considering ali the independent variables was possible. Besides these expressions, the response surfaces and the contour graphics corresponding to the essays, for each microorganism, were obtained. To the formula or region selection which better complied to the adopted criteria and conciliated ideal and realistic conditions (D ≤ 4 hours, for bacteria and D ≤ 28 hours, for fungi) both graphic and numeric strategies were used. The former consisted of the superposition of contour graphics of each micreorganism, what resulted in a region from which the formula was randomly selected using X1 = 0,10; X2 = 0,80; X3 = 0,10. The latter was based on the application of \"wish\" function, and the ideal formula obtained from calculations developed from these premisses: X1 = 0,25, X2 = 0,75; X3 = 0,0. Both formulas, derived from the employed strategies were submitted to evaluation of their preservative systems by Linear Regression Method. The D-values obtained were similar to those calculated through the previously mentioned expressions, except for Staphylococcus aureus. The application of optimized statistical methods enabled us, in a rational way, to achieve the fixed goal. Through these techniques, both experimental regions in better conditions, demonstrating advantages of using of the involved mechanisms were attained.

L'objectif de cette thèse consiste à déterminer l'influence d'une pression ambiante comprise entre 100 à 600 KPa sur l'évaporation d'une goutte, puis sur l'évaporation d'un spray soumis aux mêmes conditions. La première étape consiste à étudier l'influence de la pression ambiante sur l'évolution de la forme, du diamètre, de la vitesse et du débit d'évaporation d'une goutte de liquide en chute libre. Pour cela, une technique de mesure optique a été utilisée et une méthodologie a été développée pour calibrer cette technique de mesure et ainsi minimiser les erreurs de mesures sur la taille des particules. En parallèle, un modèle analytique d'évaporation de gouttes en chute libre a été développé : une attention particulière a été portée sur la détermination d'une corrélation adaptée au calcul du coefficient de traînée afin de tenir compte de l'évolution de la forme des gouttes au cours de leur chute. Ce modèle d'évaporation de gouttes est comparé à un modèle d'évaporation de spray (prenant en compte l'entrainement d'air, la concentration de vapeur au loin de la goutte et l'influence de la pression ambiante) pour montrer les limites du modèle d'évaporation de gouttes lors de son application à l'évaporation d'un spray. La deuxième étape de l'étude a consisté à appliquer les techniques de mesure et d'analyse mises au point précédemment à l'étude de la granulométrie d'un spray pour caractériser l'influence de trois paramètres : la pression ambiante, la pression d'injection du liquide et la nature du liquide. L'analyse des résultats a permis de développer un modèle statistique pour déterminer la granulométrie de ces sprays.

During this investigation, vibration induced droplet generation from a liquid layer was examined as a means for achieving high heat flux evaporative cooling. Experiments were performed in which droplets were generated from a liquid layer using a submerged vibrating piezoelectric driver. Parameters determined during this investigation of droplet generation were droplet mass flow rate, droplet size, driver frequency, driver voltage, and liquid layer thickness. The results showed that as the liquid layer thickness was increased, the frequencies and frequency ranges at which droplet generation occurred decreased. Droplet mass flow rates were varied by adjustment of the liquid layer thickness, driver frequency, and driver voltage. The dependence of the drivers displacement, velocity, and acceleration on frequency and voltage was determined, and the drivers frequency response was related to the occurrence of droplet generation. As a result, a frequency-dependent dimensionless parameter was proposed as a method for predicting droplet generation from the surface of the liquid layer. The dimensionless parameter is a combination of the Froude number and the dimensionless driver acceleration. The measurements have shown that droplet generation occurs when the parameter is between distinct upper and lower bounds. An analytical heat transfer model of a droplet cooling heat transfer cell was developed to simulate the performance of such a cell for thermal management applications. Using droplet flow rates determined as functions of driver voltage, driver frequency, liquid layer thickness, and interception distance, the heat transfer rate of a droplet cooling heat transfer cell was predicted for varied heat source temperatures and cell conditions. The heat transfer model was formulated in such a way as to accommodate a number of parameter variations that can be used for the design of a simple heat transfer cell. The model was used to determine the effect of droplet cooling on the heat transfer rate from a heated surface, but it can also be used to determine the influence of any of the other embodied parameters that may be of interest for thermal management applications.

The focus of this Ph.D. thesis research is a new piezoelectrically driven micromachined ultrasonic atomizer concept that utilizes fluid cavity resonances in the 15 MHz range along with acoustic wave focusing to generate the pressure gradient required for droplet or jet ejection. This ejection technique exhibits low-power operation while addressing the key challenges associated with other atomization technologies including production of sub-5 um diameter droplets, low-temperature operation, the capacity to scale throughput up or down, and simple, low-cost fabrication. This thesis research includes device development and fabrication as well as experimental characterization and theoretical modeling of the acoustics and fluid mechanics underlying device operation. The main goal is to gain an understanding of the fundamental physics of these processes in order to achieve optimal design and controlled operation of the atomizer. Simulations of the acoustic response of the system for various device geometries and different ejection fluid properties predict the resonant frequencies of the device and confirm that pressure field focusing occurs. High-spatial-resolution stroboscopic visualization of fluid ejection under various operating conditions is used to investigate whether the proposed atomizer is capable of operating in either the discrete-droplet or continuous-jet mode. The results of the visualization experiments combined with a scaling analysis provide a basic understanding of the physics governing the ejection process and allow for the establishment of simple scaling laws that prescribe the mode (e.g., discrete-droplet vs. continuous-jet) of ejection. In parallel, a detailed computational fluid dynamics (CFD) analysis of the fluid interface evolution and droplet formation and transport during the ejection process provides in-depth insight into the physics of the ejection process and determines the limits of validity of the scaling laws. These characterization efforts performed in concert with device development lead to the optimal device design. The unique advantages enabled by the developed micromachined ultrasonic atomizer are illustrated for challenging fluid atomization examples from a variety of applications ranging from fuel processing on small scales to ultra-soft electrospray ionization of biomolecules for bioanalytical mass spectrometry.

Distributed hydrogen generation from liquid hydrocarbon fuels to supply portable fuel cells presents an attractive, high energy density alternative to current battery technology. Traditional unit operation reactor design for hydrogen generation becomes inadequate with decrease in scale because of the unique challenges of size and weight minimization. To address the challenge of reactor scale-down, the concept of multifunctional reactors has emerged, in which synergistic combination of different unit operations is explored to achieve improved performance. The direct droplet impingement reactor (DDIR) studied here is based on this approach in which the liquid feed is atomized using a regularly spaced array of droplet generators with unparalleled control over droplet characteristics, followed by vaporization and reaction directly on the catalyst surface. Considering each droplet generator in the array as a unit cell, a comprehensive, first-principles model of the DDIR has been developed by considering the intimately coupled processes of 1) droplet transport, heating, evaporation, and impingement on the catalyst surface, 2) liquid reagent film formation, capillary penetration, and vaporization within the catalyst layer, and 3) gas phase heat and mass transfer and catalytic reactions. Simulations are performed to investigate the effect of reactor operating parameters on performance. Experimental validation of the model is carried out by visualizing droplet impingement and liquid film accumulation while simultaneously monitoring reaction product composition over a range of operating conditions. Results suggest an optimal unit cell shape for reaction selectivity based on a balance between reagent back diffusion and catalyst bed thermal resistance. Further, achieving a target throughput is best accomplished by adding together a larger number unit cells with optimized geometry and lower throughput (per unit cell) to more effectively spread heat and avoid hotspots at the catalyst interface. At the same time, conditions must be satisfied for ensuring droplet impingement on the catalyst surface, which become more stringent as unit cell throughput is decreased.