This thesis presents a compendium of work related to performance analyses of a proton electrolyte membrane (PEM) fuel cell with two novel design configurations. The finite element based numerical analysis has been carried out to solve the numerical transport models involved in a PEM fuel cell coupled with the flow in a porous medium, charge balance, electrochemical kinetics and membrane water content. The scope of this research work focuses on improving the performance of the PEM fuel cell by optimizing the thermo-fluid properties of the reactant species instead of analysing the complex electro-chemical interactions. Two new design configurations have been numerically analyzed; in the first design approach, a perforated-type gas flow distributor is used instead of a conventional gas flow distributor such as a serpentine, straight or spiral shape; the second design approach examines the effect of reactant flow pulsation on the PEM fuel cell performance. Results obtained from the numerical analyses were also compared with the experimental data and a good agreement was found. Performance of the PEM fuel cell with a perforated-type gas distributor was analyzed at different operating and geometric conditions to explore the merits of this new design configuration. Two-dimensional numerical analyses were carried out to analyze the effect of varying the different operating parameters; threedimensional numerical analyses were carried out to study the variation of different geometric parameters on overall performance of the new design configuration of the PEM fuel cell. The effects of the reactant flow pulsation on the performance of PEM fuel cell were analyzed using a two-dimensional numerical approach where both active and passive design configurations were numerically simulated to generate the pulsations in the reactant flow. The results showed a considerable increase in overall performance of the PEM fuel cell by introducing pulsations in the flow.

A l’aide de la Représentation Energétique Macroscopique (REM) comme outil de modélisation graphique, la modélisation et la gestion d’énergie d’une application stationnaire isolée à base d’un système PEMFC couplé à l’énergie solaire photovoltaïque comme source principale d’énergie sont développées. Afin d’assurer une autonomie du système en combustible, un électrolyseur PEM est intégré au dispositif. En outre, des packs de batteries et de supercondensateurs permettent un stockage d’énergie et de puissance.Grâce à la modularité de la REM, les modèles respectifs des différentes entités énergétiques du système ont été développés avant de les assembler pour reconstituer un modèle global. Une caractéristique propre de la REM étant la commande, une Structure Maximale de Commande (SMC) est déduite du modèle REM du système par application de règles d’inversion.Le phénomène d’effet échelle a permis de dimensionner le système grâce à un profil de consommation domestique d’énergie électrique. Une stratégie de gestion énergétique basée sur la méthode du bilan des flux de puissance et prenant en compte les dynamiques de chaque source a été développée. Différents modes de fonctionnement ont été étudiés. Grâce è un profil d’ensoleillement d’une journée, la pertinence du modèle a été évaluée. Il a été en outre introduit un couplage entre la méthode du bilan des flux de puissance et la logique floue afin que la stratégie de gestion redéfinisse les références des grandeurs électriques en tenant compte de l’état de charge des batteries et de celui des supercondensateurs. / A stand alone multi-source system based on the coupling of photovoltaic energy and both a PEM electrolyser and a PEMFC for stationary application is studied. The system gathers photovoltaic array as main energy source, ultracapacitors and batteries packs in order to smooth respectively fast and medium dynamic by supplying the load or by absorbing photovoltaic source overproduction. Because of the necessity of fuel availability, especially for islanding application like this one, a PEM electrolyser is integrated to the system for in situ hydrogen production.The main purpose being modeling and management of the power flows in order to meet the energy requirement without power cut, a graphical modeling tool namely Energetic Macroscopic Representation (EMR) is used because of its analysis and control strengths. Thanks to the modular feature of the EMR, the different models of each energetic entity of the system are performed before their assembling.By using scale effect, the energetic system sizing is performed according to a household power profile. Then, by the help of the multi-level representation, the maximal control structure (MCS) is deduced from the system EMR model. The electrical reference values of the MCS are generated by applying the power balancing method involving the own dynamic of each source into the energy management strategy. Different behavior modes are taken into account. By considering an irradiance profile for one day, the system is simulated highlighting its suitable behaviour. Moreover, the relevance of the introduced coupling between fuzzy logic controller and the power balancing method is pointed out.

In order to achieve a more stable and highly proton conducting membrane that is also cost effective, the perfluoroalkyl benzenesulfonylimides (PFSI) polymers are proposed as electrolyte for Proton Exchange Membrane Fuel Cells. 4-(trifluoromethyl)-2-diazonium perfluoro-3, 6-dioxa-4-methyl-7-octene benzenesulfonyl imide (I) is synthesized from Nafion monomer via a 5-step schematic reaction at optimal reaction conditions. This diazonium PFSI zwitterionic monomer can be further subjected to polymerization. The loss of the diazonium N2+ functional group in the monomer is believed to form the covalent bond between the PFSI polymer electrolyte and carbon electrodes support. All the intermediates and final products were characterized using 1H NMR, 19F NMR and IR spectrometry.

Fuel cells are predicted to be the power delivery devices of the future. They have many advantages such as the wide fuel selection, high energy density, high efficiency and an inherent safety which explains the immense interest in this power source. The need for advanced designs has been limited by the lack of understanding of the transport processes inside the fuel cell stack. The reactant gases undergo many processes in a fuel cell that cannot be observed. Some of these processes include convective and diffusional mass transport through various types of materials, phase change and chemical reaction. In order to optimize these variables, an accurate mathematical model can provide a valuable tool to gain insight into the processes that are occurring. The goal of this dissertation is to develop a mathematical model for polymer electrolyte-based fuel cells to help contribute to a better understanding of fuel cell mass, heat and charge transport phenomena, to ultimately design more efficient fuel cells. The model is a two-phase, transient mathematical model created with MATLAB. The model was created by using each fuel cell layer as a control volume. In addition, each fuel cell layer was further divided into the number of nodes that the user inputs into the model. Transient heat and mass transfer equations were created for each node. The catalyst layers were modeled using porous electrode equations and the Butler-Volmer equation. The membrane model used Fick's law of diffusion and a set of empirical relations for water uptake and conductivity. Additional work performed for this dissertation includes a mathematical model for predicting bolt torque, and the design and fabrication of four fuel cell stacks ranging in size from macro to micro scale for model validation. The work performed in this dissertation will help improve the designs of polymer electrolyte fuel cells, and other polymer membrane-based fuel cells (such as direct methanol fuel cells) in the future.

Le développement de nouveaux convertisseurs énergétiques non polluants et non dépendants des énergies fossiles est un enjeu environnemental, économique et politique crucial auquel les piles à combustible à membrane échangeuse de protons (PEM) pourraient répondre. La démocratisation de ces systèmes passe par l'amélioration de leurs performances, la réduction de leur coût lié essentiellement à l'utilisation de platine et par l'augmentation de leur durée de vie. Ce travail vise à améliorer la compréhension des mécanismes mis en jeu et à apporter une solution aux différents verrous technologiques en utilisant des matériaux modèles tels que les aérogels de carbone comme support de catalyseur dans les couches catalytiques cathodiques de piles à combustible PEM. Pour répondre à ces objectifs, l'impact de la texture du support carboné sur les performances électrochimiques en Assemblage Membrane Electrodes (AME) a été étudié. Puis nous avons analysé l'influence de la composition de la couche catalytique cathodique sur les performances, comparé différentes méthodes de dépôt de platine et enfin étudié le vieillissement des aérogels de carbone en AME sur banc monocellule. Ces études ont montré l'influence de l'architecture du support carboné sur les performances, l'impact positif de l'utilisation de PTFE dans la couche catalytique cathodique ainsi que l'intérêt de valider le protocole de dépôt de platine en fonction des performances au temps t=0 et au cours du vieillissement. Enfin, l'étude de la tenue au vieillissement des aérogels de carbone a montré que des progrès restaient à faire dans ce domaine.

Many modeling studies have been carried out to simulate the current distribution across the channel and shoulder direction in a proton exchange membrane (PEM) fuel cell. However the modeling results do not show agreement on the current density distribution. At the same time, no experimental measurement result of current density distribution across the channel and the shoulder direction is available to testify the modeling studies. Hence in this work, an experiment was conducted to separately measure the current densities under the channel and the shoulder in a PEM fuel cell by using the specially designed membrane electrode assemblies. The experimental results show that the current density under the channel is lower than that under the shoulder except when the fuel cell load is high. Afterwards two more experiments were carried out to find out the reason causing the higher current density under the shoulder. The effects of the electric resistance of gas diffusion layer (GDL) in the lateral and through-plane directions on the current density distribution were studied respectively. The experimental results show that it is the through-plane electric resistance that leads to the higher current density under the shoulder. Moreover, a three-dimensional fuel cell model is developed using FORTRAN. A new method of combining the thin-film model and homogeneous model is utilized to model the catalyst layer. The model is validated by the experimental data. The distribution of current density, oxygen concentration, membrane phase potential, solid phase potential and overpotential in a PEM fuel cell have been studied by the model. The modeling results show that the new modeling method provides better simulations to the actual transport processes and chemical reaction in the catalyst layer of a PEM fuel cell.

Vehicles powered by proton exchange membrane (PEM) fuelcells are approaching commercialisation. Being inherently cleanand efficient sources of power, fuel cells constitute asustainable alternative to internal combustion engines to meetfuture low-emission legislation. The PEM fuel cell may befuelled directly by hydrogen, but other alternatives appearmore attractive at present, due to problems related to theproduction, transportation and handling of hydrogen. Fuelling with an alcohol fuel, such as methanol, which isoxidised directly at the anode, offers certain advantages.However, the efficiency of the direct-methanol fuel cell (DMFC)is still significantly lower than that of the conventionalhydrogen-fuelled PEM fuel cell, due to some technical problemsremaining unsolved. Hence, indirect fuelling by a reformedliquid fuel may be the most feasible option in the early stagesof the introduction of fuel cell vehicles. The work presented in this thesis concerns the developmentof catalysts for production of hydrogen from methanol bypartial oxidation, steam reforming or a combination thereof.The work contributes to the understanding of how thepreparation route affects catalyst morphology and howphysicochemical properties determine catalytic behaviour andreaction pathways. The thesis is a summary of seven papers published inscientific periodicals. The first paper (Paper I) reviews thecurrent status of catalytic hydrogen generation from methanol,focusing on the fuel cell application. Paper II investigatesthe partial oxidation of methanol over Cu/ZnO catalystsprepared in microemulsion and by a conventionalco-precipitation technique. The activity for methanolconversion in the low-temperature regime is found to besignificantly higher over the former materials and the workcontinues by determining the nature of possible Cu-ZnOinteractions in the catalysts by studying their physicochemicalproperties more thoroughly (Paper III). In Paper IV, thepathways for methanol conversion via both partial oxidation andsteam reforming are elucidated. In Paper V, partial oxidation of methanol is studied overPd/ZnO catalysts prepared by microemulsion technique and againcompared to conventional materials. This investigationdemonstrates that although possessing high methanol conversionactivity, palladium-based catalysts are not suitable forreforming in fuel cell applications due to the considerableamounts of carbon monoxide formed. In Paper VI, methanol reforming is investigated over acommercial Cu/ZnO/Al2O3 catalyst. The mechanisms for carbonmonoxide formation and strategies for its suppression arediscussed, as well as reactor design aspects. The study alsoincludes some simple kinetic modelling. Finally, Paper VIIdescribes the optimisation of catalyst composition and processconditions to reach high hydrogen production efficiency at lowoperating temperatures and with minimum carbon monoxideformation. <b>Keywords:</b>PEM fuel cells, hydrogen, methanol, reforming,(partial) oxidation, reaction pathways, carbon monoxide,catalyst, microemulsion, Cu/ZnO, Pd/ZnO, copper, redoxproperties, oxidation state

Excessive water formation inside the polymer electrolyte membrane (PEM) fuel cell’s structures leads to the flooding of the cathode gas diffusion layer (GDL) and cathode gas flow channels. This results in a negative impact on water management and the overall cell performance. Liquid water generated in the cathode catalyst layer and the water moved from anode to cathode side due to electro-osmotic drag transport through the GDL to reach the gas flow field channels, where it is removed by air cathode gas stream. Due to high and uniform capillary force distribution effect of the pores through the GDL plane and surface tension between the water droplets and gas flow field channels surfaces, liquid water tends to block/fill the pores of the GDL and stick to the surface of the GDL and gas flow channels. Therefore, it is difficult to remove the trapped water in GDL structure which can lead to flood of the PEM fuel cell. The GDL surfaces are commonly treated uniformly with a hydrophobic material in order to overcome the flooding phenomena inside PEM fuel cell. Despite the importance impact of the surface wettability of both channel and GDL surface characteristics especially for the cathode side on the water management, few experimental studies have been conducted to investigate the effect of the two-phase flow in cathode gas flow channel and their crucial role. The work presented in this thesis covers contributions that provide insight, not only into the investigation of the effects of hydrophobic cathode GDL and cathode gas flow channels, on water removal, two phase flow inside the channel, and on PEM fuel cell performance, but also the superhydrophobic and superhydrophilic GDLs and gas flow channels effects. Further, the effects of a novel GDL designs with sandwich and gradient wettability with driving capillary force through GDL plane have been investigated. Two-phase flow especially in the cathode gas flow field channels of PEM fuel cell has a crucial role on water removal. Hence, in this research, ex-situ investigations of the effects of channels with different surface wettability; superhydrophobic, hydrophobic, slightly hydrophobic, and superhydrophilic on the two-phase flow characteristics have been tested and visualized at room temperature. Pressure drop measurements and two-phase flow visualization have been carried out using high speed camera. The effect of the various coating materials on graphite and GDL surface morphology, roughness, static contact angle (θ), and sliding contact angle (α) have been investigated using scanning electron microscopy (SEM), Profilometry, and sessile drop technique, respectively. It has been observed that the two-phase flow resistance is considerably affected by surface wettability of the channels. Further, the overall cell performance can be improved by superhydrophobic gas flow channels mainly at high current density over slightly hydrophobic and superhydrophilic cases tested. In addition, sandwich wettability GDL has been coated with a silica particle/ Polydimethylsiloxane (PDMS) composite. The porometric characteristics have been studied using, method of standard porosimetry (MSP). It has been found that sandwich wettability GDL has superhydrophobic surfaces with (θ = 162±2°), (α = 5±1°), and the internal pores are hydrophilic, while the mean pore radius is 7.1μm. This shows a low resistance to gas transport. On the other hand, performance testing indicates that (PEM) fuel cell equipped with sandwich wettability GDL results in the best performance compared to those with raw (non-coated) (slightly hydrophobic), PTFE coated (commercial with micro-porous layer (MPL)) (superhydrophobic), and silica coated (superhydrophilic) GDL. The wettability gradient has been introduced through plane of the one side hydrophobic GDL by coating one side of non-coated GDL with 15 wt. % of PTFE solution; however, the other side remains uncoated. The effects of wettability gradient on the water removal rate, droplet dynamics, and PEM fuel cell performance have been covered in this thesis. Water removal rate is determined using a 20 ml syringe barrel, wherein a 13 mm diameter GDL token is fixed on the barrel opening. The droplets penetrating through the GDL are visualized via a high speed camera to study the droplets’ dynamic characteristics. The GDL wettability gradient has a significant impact on water removal rate, droplets’ dynamic characteristics, and consequently enhances the overall PEM fuel cell performance.

The polymer electrolyte membrane fuel cell (PEMFC) is an electrochemical device used for the production of power, which is a key for the transition towards green and renewable power delivery devices for mobile, stationary and back-up power applications. PEMFCs consume hydrogen and oxygen to produce power, water and heat. The transient start-up from sub-zero freezing temperature conditions is a problem for the successful, undamaged and unhindered operation. The generation and presence of water in the PEMFC stack in such an environment leads to the formation of ice that hinders the flow of gases, causes morphological changes in the membrane electrode assembly (MEA) leading to reversible and irreversible degradation of stack performance. Start-up performance is highly dependent on start-up operational conditions and procedures. The previous state of the stack will influence the ability to perform upon the next start-up and operation. Water generated during normal operation is vital and improves performance when properly managed. Liquid water present at shut-down can form ice and cause unwanted start-up effects. This phase change may cause damage to the MEA and gas diffusion media due to volume expansion. Removal of high water content at shutdown decreases proton conductivity which can delay start-up times. The United States Department of Energy (DOE) has established a set of criteria that will make fuel cell technology viable when attained. As specified by DOE, an 80 kWe fuel cell will be required by 2015 to reach 50% power in 30 seconds from start-up at an ambient temperature of -20°C. This work investigates freeze start-up in a multi-kilowatt stack approaching both shut-down conditioning and start-up operations to improve performance, moderate fuel cell damage and determine the limits of current stack technology. The investigation involved a Hydrogenics Corporation 5 kW 506 series fuel cell stack. The investigation is completed through conditioning the fuel cell start-up performance at various temperatures ranging from -5°C to below -20°C. The control of system start-up temperature is achieved with an environmental chamber that maintains the desired set point during dwell time and start-up. The supply gases for the experiment are conditioned at ambient stack temperature to create a realistic environment that could be experienced in colder weather climates. Temperature controls aim to maintain steady ambient temperatures during progressive start-up in order to best simulate ambient conditions. The control and operation of the fuel cell is maintained by the use of a fuel cell automated test station (FCATS™). FCATS supplies gas feeds, coolant medium and can control temperature and reactant humidity in reactants according to a prescribed procedure for continuous operation. The iv collection of data occurs by the same system recording cell voltage, temperatures, pressures, flow rates and current densities. A procedural start-up and characterization are conducted in order improve start-of performance and examine reactant flows, coolant activation time, stack conditioning and the effects by freezing temperatures. The resulting degradation is investigated by polarization curves and various ex-situ measurements. In this work, it was found that freeze start-up of a fuel cell stack can be aided and managed by conditioning the stack at shut-down and applying a procedure to successfully start-up and mitigate the damage that freezing can cause.

Polymer electrolyte membrane (PEM) fuel cells have increasingly become promising green energy sources for automobile and stationary cogeneration applications but its success in commercialization depends on performance optimization and manufacturing cost. The activation losses, expensive platinum catalyst, and water flooding phenomenon are the key factors currently hindering commercialization of PEM fuel cells. These factors are associated with the cathode catalyst layer (CCL), which is about ten micrometers thick. Given the small scale of this layer, it is extremely difficult to study transport phenomena inside the catalyst layer experimentally, either intrusively or non-intrusively. Therefore, mathematical and numerical models become the only means to provide insight on the physical phenomena occurring inside the CCL and to optimize the CCL designs before building a prototype for engineering application. In this thesis research, a comprehensive two-phase mathematical model for the CCL has been derived from the fundamental conservation equations using a volume-averaging method. The model also considers several water transport and physical processes that are involved in the CCL. The processes are: (a) electro-osmotic transport from the membrane to the CCL, (b) back-diffusion of water from the CCL to the membrane, (c) condensation and evaporation of water, and (d) removal of liquid water to the gas flow channel through the gas diffusion layer (GDL). A simple analytical model for the activation overpotential in the CCL has also been developed and an optimization study has been carried out using the analytical activation overpotential formulation. Further, the mathematical model has been simplified for the CCL and an analytical approach has been provided for the liquid water transport in the catalyst layer. The volume-averaged mathematical model of the CCL is finally implemented numerically along with an investigation how the physical structure of a catalyst layer affects fuel cell performance. Since the numerical model requires various effective transport properties, a set of mathematical expressions has been developed for estimating the effective transport properties in the CCL and GDL of a PEM fuel cell. The two-dimensional (2D) numerical model has been compared with the analytical model to validate the numerical results. Subsequently, using this validated model, 2D numerical studies have been carried out to investigate the effect of various physical and wetting properties of CCL and GDL on the performance of a PEM fuel cell. It has been observed that the wetting properties of a CCL control the flooding behavior, and hydrophilic characteristics of the CCL play a significant role on the cell performance. To investigate the effect of concentration variation in the flow channel, a three-dimensional numerical simulation is also presented.