3D multi-physics modelling and validation of the model of a Polymer Electrolyte Membrane Fuel

This Ph.D. thesis focuses on testing and modelling PEMFC single cell systems, to better understand internal phenomena and to find out operative solutions able to increase the overall cell performance. Thus, the first step is based on the creation and validation of a wide spectrum of multi-physics models of PEMFC fed with hydrogen or methanol, by using Comsol® Multi-physics platform coupled with Matlab®. All models are able to work under different operating conditions and with materials of different characteristics (membranes and catalysts). Moreover, the efforts were also focused on the creation of models of systems similar to fuel cell, as the gas diffusion electrode (GDE). The GDE is usually employed to analyze the electrochemical properties of the catalytic layer. Each single model was validated against a huge set of experimental data (partly obtained at POLITO, partly provided by the partners of two research projects: DURAMET and NAMEDPEM). After the model validation, these models were used to investigate the internal phenomena, and how materials, geometry and operative conditions affect the cell performance. Furthermore, particular problems affecting the entire FC system such as water flooding, methanol crossover, flow patterns design and current density distribution were deeply investigated to provide reasonable solutions. In general, the 3D multi-physics, multi-component, multi-phase and not-isothermal models developed in this Ph.D. include Maxwell-Stefan, Navier-Stokes-Brinckman, and extended two-phase Darcy-law to solve velocity, pressure, and mass transfer equations, and modified Butler-Volmer and Tafel equations to describe the electrochemical kinetics. All the equations are coupled to each other to simulate the performance of a single cell PEMFC (or GDE), reproducing the electrochemical, fluid-dynamics, and thermal phenomena. Each model was validated by comparing the simulated results, in terms of electric performance (polarization curves and power density curves), with experimental data obtained by changing several parameters: -Type of membranes: Nafion® (N112, N115, N117), Fumapem® (F1850) for the DMFC, Nafion® (N-HP and NR-212) for the hydrogen-fed PEMFC. -Dimensions of active area of the single cell: 5cm2 and 25cm2. -Catalyst: eight different catalysts for the DMFC, four for the hydrogen-fed PEMFC, three for the GDE (commercial Pt/C, PtRu/C and lab-made FeNC-based catalysts). -Operative conditions: pressure, methanol inlet concentration, air or oxygen at the cathode, cell and flow temperatures, anode and cathode flow rates, humidification and stoichiometric ratio (for the hydrogen-fed PEMFC). -Flow field designs: unique serpentine, four parallel serpentines and four inlet serpentine. After the validation, the models were used to reproduce and study the multi-dimensional trends of particular phenomena which produce system losses and/or affect the performance. In first istance, the multi-physics analysis was used to improve the way to deposit the catalyst, thus the catalytic layer distribution was investigated in order to have a better uniformity in the current density distribution at the membrane/anodic catalyst interface. The proposed solutions, the 3-Layers MEA, was modelled in Comsol® and tested in the lab. It should avoid hot-spots on the membrane as a consequence of the better uniformity in the current density distribution, with a consequent increase in the life-time of the MEA (chapter I) The second step was the analysis of the influence of water flooding and catalyst materials. The extended two-phase Darcy-law was used into the model to describe the mass transport inside the micro-porous structure of the noble/non-noble metal cathode catalyst, produced in our labs. The multi-physics analysis displays a direct relationship between the water saturation, the oxygen diffusion flow, and the oxygen consumption. Thus, water condensation inside the micro-pores may produce the flooding of micro and meso-porous, showing a consequent link between condensation and decreasing of cell performance (chapter II). The third step of the multi-physics analysis was the study of the influence of FF design (unique serpentine, four parallel serpentines, four inlet serpentines.) and types of membrane on system performance. A large amount of lab tests and simulation were performed for each FF, by changing the temperatures, the inlet flow rates, the inlet methanol concentrations and the type of cathode flow. Pulse Field Gradient (PFG) NMR spectroscopy was used to get a direct measurement of the diffusion coefficients of water and methanol through the membranes. Thus, the model was used as a tool to investigate anodic overpotentials, water and methanol crossover flow rates, current density distribution along the membrane, understanding the relationship between the shape of the FF and cell performance (i.e. pressure and methanol consumption).(chapter III) To improve the research on the area of catalyst properties, the study focused on the cathode using the gas diffusion electrode (GDE), trying to find out key parameters which influence the performance of catalysts for the oxygen reduction reaction. A commercial Pt-based catalysts and the non-noble metal Fe–N–C catalyst prepared in-house were tested and modelled, to carry out a sensitive analysis by varying the inlet oxygen flow rate, showing the influence of oxygen diffusive flow on the catalyst performance. The multi-physics analysis provides the way to increase the performance of the non-noble metal catalyst by changing some system properties as tortuosity, porosity and hydrophobicity (chapter IV) After the large analysis developed for DMFC, the PhD work continued with the modelling of hydrogen-fed PEMFC, to complete, in such way, the general sensitivity analysis on PEMFC systems. Obviously, some innovations and changings were introduced to adapt the model with the new inlet fuel, as new equations and parameters for the electrochemical behaviours, initial and boundary conditions, H2 crossover and controlled parameters, i.e., relative humidity. The multi-physics analysis and the lab tests show how the relative humidity influence the performance, in relation to the variation of pressure and temperature (chapter V) After the experimental and modelling studies, the PhD work was focused on several aspects, in order to improve the performance of the single cell with Fe-N-C catalyst on the cathode, related to material improvement and the single cell system: the active surface area, the percentage of micro-pores present in the catalytic layer, the membrane type, the procedure of catalyst deposition, the back pressure and the closing force. Thus, the modelling and lab work performed outcome in an excellent result: the system performance increases four times than the initial computed value, from about 10 mW cm–2 to 40.6 mW cm–2.