The development of improved water management strategies for the polymer electrolyte membrane fuel cell (PEMFC) could stand to benefit from an improved understanding of the surface and internal structure of the gas diffusion layer (GDL). The GDL is a fibrous porous material enabling mass transport between the PEMFC catalyst layer and flow fields. Fluorescence-based visualizations of liquid water droplet evaporation on GDL surfaces were performed to investigate water droplet pinning behaviours. The heterogeneous in-plane and through-plane porosity distributions of untreated GDLs were studied using computed tomography visualizations. The through-plane porosity distributions were utilized to calculate heterogeneous local tortuosity, relative diffusivity, and permeability distributions. Finally, the heterogeneous through-plane porosity distributions of GDLs treated for increased hydrophobicity were investigated. This work provides new insight into GDL material properties to better inform future PEMFC models.

The anisotropic mass transport issues inside a fuel cell membrane have been studied in this thesis using computer modelling. The polymer electrolyte membrane (PEM) conductivity of a PEM fuel cell (PEMFC) depends on the hydration state of the hydrophilic charged sites distributed in the pores of the membrane. Water humidification of these charged sites is crucial for sustaining the membrane conductivity and reducing concerning voltage losses of the cell. During the operation of a PEMFC, the transport of humidified inlet gases (fuel/oxidant) is influenced by external design factors such as flow field plate geometry of the gas circulating channels. As a result, there arises a distribution in the mass transport of water inside the membrane electrode assembly. A two-dimensional, cross-the-channel, fuel cell membrane layer mass transport model, developed in this work, helps the study of the impact of factors causing the distribution in the membrane ionic conductivity on ohmic losses.<p>The governing equations of the membrane mathematical model stem from the multicomponent framework of concentrated solution theory. All mass transport driving forces within the vapour and/or liquid equilibrated phases have been accounted in this research. A computational model, based on the finite control volume method, has been implemented using a line-by-line approach for solving the dependent variables of the mass transport equations in the two-dimensional membrane domain. The required boundary conditions for performing the anisotropic mass transport analysis have been obtained from a detailed agglomerate model of the cathode catalyst layer available in the literature.<p>The results obtained using boundary conditions with various flow field plate channel-land configurations revealed that the anisotropic water transport in the cathode half-cell severely affects the ohmic losses within the membrane. A partially humidified vapour equilibrated membrane simulation results show that a smaller channel-land ratio (1:1) sustains a better membrane performance compared to that with a larger one (2:1 or 4:1). Resistance calculations using the computer model revealed that ohmic losses across the membrane also depend on its physical parameters such as thickness. It was observed that the resistance offered by a thinner membrane towards vapour phase mass transport is comparatively lower than that offered by a thicker membrane. A further analysis accounting the practical aspects such as membrane swelling constraints, imposed by design limitations of a fuel cell, revealed that the membrane water content and ionic conductivity are altered with an increase in the compression constraint effects acting upon a free swelling membrane.

The anisotropic mass transport issues inside a fuel cell membrane have been studied in this thesis using computer modelling. The polymer electrolyte membrane (PEM) conductivity of a PEM fuel cell (PEMFC) depends on the hydration state of the hydrophilic charged sites distributed in the pores of the membrane. Water humidification of these charged sites is crucial for sustaining the membrane conductivity and reducing concerning voltage losses of the cell. During the operation of a PEMFC, the transport of humidified inlet gases (fuel/oxidant) is influenced by external design factors such as flow field plate geometry of the gas circulating channels. As a result, there arises a distribution in the mass transport of water inside the membrane electrode assembly. A two-dimensional, cross-the-channel, fuel cell membrane layer mass transport model, developed in this work, helps the study of the impact of factors causing the distribution in the membrane ionic conductivity on ohmic losses.<p>The governing equations of the membrane mathematical model stem from the multicomponent framework of concentrated solution theory. All mass transport driving forces within the vapour and/or liquid equilibrated phases have been accounted in this research. A computational model, based on the finite control volume method, has been implemented using a line-by-line approach for solving the dependent variables of the mass transport equations in the two-dimensional membrane domain. The required boundary conditions for performing the anisotropic mass transport analysis have been obtained from a detailed agglomerate model of the cathode catalyst layer available in the literature.<p>The results obtained using boundary conditions with various flow field plate channel-land configurations revealed that the anisotropic water transport in the cathode half-cell severely affects the ohmic losses within the membrane. A partially humidified vapour equilibrated membrane simulation results show that a smaller channel-land ratio (1:1) sustains a better membrane performance compared to that with a larger one (2:1 or 4:1). Resistance calculations using the computer model revealed that ohmic losses across the membrane also depend on its physical parameters such as thickness. It was observed that the resistance offered by a thinner membrane towards vapour phase mass transport is comparatively lower than that offered by a thicker membrane. A further analysis accounting the practical aspects such as membrane swelling constraints, imposed by design limitations of a fuel cell, revealed that the membrane water content and ionic conductivity are altered with an increase in the compression constraint effects acting upon a free swelling membrane.

The effects of actively pulsing reactant flow rates into solid polymer electrolyte fuel cells were investigated in this thesis. First, work was conducted to determine the magnitude of voltage response to pulsed reactant flow on a direct hydrogen proton exchange membrane (PEM) cell. The effects of pulsed reactant flow into a direct methanol fuel cell (DMFC) were then considered. The PEM work showed substantially greater response to pulsed air flow than to pulsed fuel flow. It was found that several parameters affect the magnitude of cell response to active flow control (AFC). Increasing current load, increasing the magnitude of flow oscillation, decreasing the frequency of oscillation, and decreasing the average level of excess reactant supplied were found to maximize both the level of voltage oscillations and the decrease in cell power from steady state performance. Greater response to pulsed oxidant flow is believed to have been observed due to effects brought about by changes in membrane humidity. In contrast, pulsed fuel flow showed the greatest response in the study of DMFC technology. In this case, time averaged cell voltage was found to increase as the time averaged fuel flow rate was reduced. The increase in average cell power is the result of a reduction in methanol crossover; sustainable increases of up to 6% in power output were measured. The parameters found to effect the increase in cell power observed include the frequency of oscillation and the time-averaged NOSfuel. Pulsed air flow on the DMFC did not show any such rise in voltage, supporting the hypothesis that a reduction in methanol crossover is the phenomenon which brings about enhanced performance.

Kyoto University (京都大学) / 0048 / 新制・課程博士 / 博士(エネルギー科学) / 甲第19090号 / エネ博第314号 / 新制||エネ||64 / 32041 / 京都大学大学院エネルギー科学研究科エネルギー基礎科学専攻 / (主査)教授 萩原 理加, 教授 佐川 尚, 教授 野平 俊之 / 学位規則第4条第1項該当

This dissertation focuses on the preparation of new Nafion®/ ex-situ silica nanocomposites membranes and the impact of particle size of spherical silica particles on the nanocomposites' properties. To achieve acceptable power production, fuel cell polymer membranes are required with good proton conductivity, water retention, thermal and mechanical stability. However, to avoid poisoning of fuel cell electrocatalysts with CO or other fuel contaminants, they must be operated at temperatures (>100 °C). At these temperatures, fuel cell membranes dehydrate resulting in dramatic decreases in proton conductivity or complete failure as membranes crack due to volumetric stress from water loss. Even if fuel cell is kept in a humidified chamber, increasing temperature will eventually shut the cell down as Nafion®'s bicontinuous structure "dissolves" into a single poorly conducting phase at temperatures above the polymer's Tg.This research provides systematic studies of effects of silica particle size on properties of silica-Nafion® nanocomposites. Results of this study include new insights into requirements for reproducible particle syntheses, practical methods for avoiding silica particle floatation during Nafion® nanocomposite membranes preparation, and a summary of the influence of particle size and functionalization on Nafion® membrane properties. Stober particle syntheses showed high sensitive to ammonia concentration and we discovered that literature procedures' variability is likely due to researchers failure to actually measure ammonia concentration in their aqueous base (which can be 50% or more off). Homogeneous nanocomposite membranes, as determined by AFM and SEM, were successfully prepared using more viscous dispersions. It was observed that nanocomposites membranes with small particles (<50 nm) showed significant increases in proton conductivity at temperatures above 80 °C. Surface modification of the silica particles improved the proton conductivity at 80 °C. Enhancement on proton conductivity was more pronounced with small modified particles at temperatures < 80 °C but unmodified particles were better than modified particles at temperatures >80 °C. Small, unmodified particles led to enhanced thermal stability of the Nafion® ionic domain, however, surface modification did not result in any thermal stability enhancement. Contrary to the expected, mechanical properties of the Nafion® were degraded by adding the silica particles, especially with smaller particles (<50nm).

Pressing demands for high power and high energy densities in novel electrical energy storage units have caused reconsiderations regarding both the choice of battery chemistry and design. Practical concerns originating in the conventional use of flammable liquid electrolytes have renewed the interests of using solvent-free polymer electrolytes (SPEs) as solid ionic conductors for safer batteries. In this thesis work, SPEs developed from two polymer host structures, polyethers and polycarbonates, have been investigated for all-solid-state Li- and Li-ion battery applications. In the first part, functional polyether-based polymer electrolytes, such as poly(propylene glycol) triamine based oligomer and poly(propylene oxide)-based acrylates, were investigated for 3D-microbattery applications. The amine end-groups were favorable for forming conformal electrolyte coatings onto 3D electrodes via self-assembly. In-situ polymerization methods such as UV-initiated and electro-initiated polymerization techniques also showed potential to deposit uniform and conformal polymer coatings with thicknesses down to nano-dimensions. Moreover, poly(trimethylene carbonate) (PTMC), an alternative to the commonly investigated polyether host materials, was synthesized for SPE applications and showed promising functionality as battery electrolyte. High-molecular-weight PTMC was first applied in LiFePO4-based batteries. By incorporating an oligomeric PTMC as an interfacial mediator, enhanced surface contacts at the electrode/SPE interfaces and obvious improvements in initial capacities were realized. In addition, room-temperature functionality of PTMC-based SPEs was explored through copolymerization of ε-caprolactone (CL) with TMC. Stable cycling performance at ambient temperatures was confirmed in P(TMC/CL)-based LiFePO4 half cells (e.g., around 80 and 150 mAh g-1 at 22 °C and 40 °C under C/20 rate, respectively). Through functionalization, hydroxyl-capped PTMC demonstrated good surface adhesion to metal oxides and was applied on non-planar electrodes. Ionic transport behavior in polycarbonate-SPEs was examined by both experimental and computational approaches. A coupling of Li ion transport with the polymer chain motions was demonstrated. The final part of this work has been focused on exploring the key characteristics of the electrode/SPE interfacial chemistry using PEO and PTMC host materials, respectively. X-ray photoelectron spectroscopy (XPS) was used to get insights on the compositions of the interphase layers in both graphite and LiFePO4 half cells.

Foram realizadas pesquisas em uma série de condutores iônicos que apresentam aplicações na área dos dispositivos eletroquímicos de estado sólido, utilizando-se basicamente a técnica de Ressonância Magnética Nuclear (RMN). A primeira parte deste trabalho é dedicada aos compostos de intercalação baseados na matriz de dissulfeto de molibdênio (MoS2) onde as espécies intercalantes (íons de lítio e moléculas de aminas) são inseridas num espaço de dimensionalidade reduzida gerado pela matriz. Já a segunda parte envolve os condutores iônicos poliméricos do tipo compósitos, nos quais foram utilizadas nanopartículas de carbono (Carbon Black) e titânio (TiO2) no eletrólito formado pelo poli(óxido de etileno) (POE) e o perclorato de lítio (LiClO4). Todos estes sistemas apresentam, em geral, uma considerável complexidade estrutural, o que significa que os movimentos moleculares e de difusão iônica se produzem num meio semicristalino (caso dos compósitos) ou num meio de dimensionalidade reduzida (caso dos intercalados). A espectroscopia de RMN dos núcleos de 7Li e 1H é uma técnica conveniente para o estudo destes materiais, pois é possível avaliar, através dos resultados obtidos das medidas dos tempos de relaxação e formas de linha, os efeitos provocados pela baixa dimensionalidade dos movimentos em estruturas laminares (caso dos intercalados), bem como identificar e aferir as interações e os mecanismos de relaxação resultantes dos diferentes graus de liberdade dos movimentos (iônicos e moleculares), fornecer parâmetros estruturais (distâncias interatômicas) que auxiliam na proposta de possíveis modelos estruturais e caracterizar completamente a escala temporal dos movimentos iônicos e moleculares. / Nuclear Magnetic Resonance (RMN) techniques were used to study a series of ionic conductor materials, which present applications in the area of the solid-state electrochemical devices. The first part of this work is dedicated to the study of intercalation compounds based on the molybdenum disulfide matrix (MoS2), where the intercalated species (lithium ion and amine molecule) are inserted in the low-dimensionality space generated by the matrix. The second part involves the study of a composite polymer electrolyte, employing fillers like Carbon Black and titanium dioxide (Tio2) nano particles in the electrolyte formed by the poly(ethy1ene oxide) and a lithium salt (LiClO4). In general, these systems present a considerable structural complexity, meaning that the molecular movements and ionic diffusion are produced in a semicrystalline environment (case of the composites) or in an environment of reduced dimensionality (case of intercalates). The 7Li and 1H NMR spectroscopy is a convenient technique for the study of these materials. Relaxation time and line shape measurements may provide a tool to investigate the effects provoked by the low-dimensionality of the movements in laminate structures (case of intercalate), to identify the interactions and relaxation mechanisms of the ionic and molecular motions, to supply structural parameters (interatomic distances) that would help the proposal of possible structural models, and finally, to characterize the time scale of the ionic and molecular movements completely.

Kyoto University (京都大学) / 0048 / 新制・課程博士 / 博士(工学) / 甲第15584号 / 工博第3296号 / 新制||工||1497(附属図書館) / 28105 / 京都大学大学院工学研究科材料化学専攻 / (主査)教授 瀧川 敏算, 教授 木村 俊作, 教授 長谷川 博一 / 学位規則第4条第1項該当

Philosophiae Doctor - PhD / High Temperature Polymer Electrolyte Membrane Fuel Cells (HT-PEMFC) have received renewed interest in recent years due to its inherent advantages associated with the limitations faced by Low Temperature Polymer Electrolyte Membrane Fuel Cells (LT-PEMFC). The high Pt loadings required for PEMFCs have significantly hindered its commercialisation. Electrophoretic Deposition (EPD) is a promising route to reduce the noble metal loading. EPD is a method in which charged colloidal particles are deposited onto a target substrate under the force of an externally applied electric field. To effectively study the EPD method, the methodology of this study was divided into two parts: (i) the EPD method was studied via known empirical methods to fabricate, test and characterise MEAs suitable for HT-PEMFCs. The feasibility of the EPD method was determined by comparing the performance of the fabricated EPD MEAs to MEAs fabricated via spraying methods, and (ii) due to the promising results obtained in part (i) of the methodology, a theoretical model was developed to obtain a deep understanding about nature of the interactions between the Pt/C particles in a colloidal suspension. The theoretical model will serve as a foundation for future studies. In part (i) of the methodology, the Pt/C particles were studied in organic solutions (i.e. Isopropyl Alcohol, IPA) via the Zetasizer Nano ZS instrument under various salt (NaCl) concentrations and pH conditions while introducing polymeric surfactants, i.e. Nafion® ionomer and Polytetrafluoroethylene (PTFE) to the suspension. The optimum catalyst suspensions were selected to fabricate GDEs via the EPD method. Physical characterisations revealed that the EPD GDEs exhibited cracked morphology with high porosity. Electrochemical characterisations revealed that the EPD MEA showed significantly better performance (i.e. 73% higher peak power) compared to the hand vi sprayed MEA due to lower charge transfer and mass transport resistance at high current densities. Compared to the ultrasonically sprayed MEA, the EPD MEA exhibited a peak power increase of ~12% at a slightly lower Pt loading (i.e. ~4 wt%). A comparative study between the Nafion® ionomer and PTFE in the CLs of two EPD MEAs revealed superior performance for the EPD MEA with the PTFE in the CLs. Part (ii) of the methodology deals with the electrical interfacial properties of the aqueous Pt/C suspension. The study consists of two sets of measurements (i.e. electrophoretic and coagulation dynamic studies) conducted for different electrolyte compositions. A theoretical background on determining the interfacial potential and charge from electrophoretic and coagulation dynamic measurements are provided. Detailed statements of the Standard Electrokinetic and Derjaguin, Landau, Vervey and Overbeek Models are given in the forms that are capable of addressing electrophoresis and the interaction of particles for an arbitrary ratio of the particle to Debye radius, interfacial potential and electrolyte composition. The obtained experimental data were processed by using numerical algorithms based on the formulated models for obtaining the interfacial potential and charge. While analysing the dependencies of interfacial potential and charge on the electrolyte compositions charge, conclusions were made regarding the mechanisms of charge formation. It was established that the behaviour of system stability is in qualitative agreement with the results computed from the electrophoretic data. The verification of quantitative applicability of the employed models was conducted by calculating the Hamaker constant from the experimental data. It was proposed how to explain the observed variations of the predicted Hamaker constant and its unusually high value.