Computational modeling and experimental verification of cathode catalyst layer on PEM fuel cells

Highlights

• The catalyst layer of fuel cells, which has a vital role for water management, should be optimized.

• Electrical conductivity of gas diffusion layer can be controlled by adjusting Catalyst:Nafion ratio.

• The current density can be estimated directly from the rotation rate and Pt coating by using the proposed mathematical model.

• Half-cell modeling makes performance values of the fuel cells more predictable.

Abstract

Fuel cell systems are environmentally friendly energy converters that directly transform the chemical energy of the fuel to electricity. The proton exchange membrane (PEM) fuel cells are one of the most common type of fuel cells since they deliver high power density and are lighter and smaller when compared to the other cells. However, commercialization of the PEM fuel cells is challenging due to the high cost of its components. In addition to high catalyst costs, the problem of poor water management is also a vital issue that needs to be overcome. While the gas diffusion layer of a fuel cell is essential for removing the by-product water, the Nafion solution contained in the catalyst layer has hydrophobic properties and is crucial for both preventing the water accumulation and increasing the effectiveness of the fuel cell. In this study, the effects of Carbon:Nafion ratio on the reduction potential was investigated. The cyclic voltammograms (CV) was produced for each ratio, and it was shown that the CVs exhibit characteristics of hydrogen adsorption/desorption peaks. All the linear sweep voltammogram (LSV) curves revealed well distinguished regions of kinetic, mixed and diffusion limited reaction rate. As a result, it was observed that the ratio of 1:5 resulted higher reduction potential compared to 1:3 and 1:7. Finally, a mathematical model was purposed, in which related the rotation rate and platinum coating with the current density, in order to gain insight about the responses of the fuel cell system. The constructed model is tested and validated experimentally for various parameters that are present in the system, and it may be utilized to determine oxygen reaction activities of the catalysts without performing any unnecessary electrochemical tests.

Introduction

The fact that the usage of fossil fuels are declining day by day and the environmental problems they cause during their use have led researchers to seek alternative energy sources. Utilization of renewable energy sources such as wind, solar, biomass and hydrogen has been seen as a solution to environmental problems and the use of these resources has increased [1]. At this point, the use of fuel cells that convert the chemical energy of the fuel into electricity has come into prominence. The use of proton exchange membrane (PEM) fuel cells is very common due to their wide range of applications, ease of design and high efficiency [2]. Hence, PEM fuel cell system is one of the most promising systems [3]. In PEM fuel cells, hydrogen gas is fed to the anode side of the cell, while oxygen gas or air is fed to the cathode side. Hydrogen is catalyzed on the anode side and separated into protons. The formed electrons and protons react with oxygen at the cathode side. As a result of this exothermic electrochemical reaction, water is formed and DC current is generated [4].

The main parts of a PEM fuel cell are membrane electrode assembly (MEA), bipolar plates and gasket. MEA consists a polymer membrane which provides proton conduction, gas diffusion layer (GDL) and catalyst layer (CL) [5]. One of the major obstacles in the commercialization of fuel cells is the high cost of platinum or platinum alloys used as electro catalysts due to their high activity for the oxygen reduction reaction (ORR) [6,7]. However, the net present value of fuel cells are comparable to the other energy sources when salvage value of platinum is considered [6]. In addition, the structure is supported by carbon in the form of nano-sized particles to reduce the cost [7]. The obstacle of cost is enforced by the problem of degradation in fuel cells. In order to overcome this challenge, numerous studies has been conducted. In a recent study, Dhimish et al. observed that the efficiency of PEM fuel cells drop around 7.2%–14.7% after 180 days of continuous working cycle [8]. However, numerous studies have been conducted with the aim of increasing the durability of the fuel cells. Wang et al. showed that the addition of CeO₂ nanoparticles to the microporous layer of the cell can significantly improve the cell durability [9]. Huang et al. reported a preparation and performance study of Pt–Co–ZrO₂ multi-walled carbon nanotubes, in which they showed that the cell durability of such PEM fuel cells has greatly surpassed that of Pt/C [10]. Furthermore, Jahromi & Heidary proposed a triple fuel cell stack configuration, which can prolong the lifetime of fuel cells by 18.93% and cut the stack cost around 14% [11]. In addition, nano-sized particles also increase the electrocatalyst durability and enhance the reactive surface sites [12]. Catalyst layer plays a vital role in removing water from the structure as a result of reactions in the fuel cell and also provides mechanical support to the MEA [13]. In addition to the catalytic materials catalyst layers include various components such as Nafion solution, pure water and isopropanol. Among them, Nafion solution has vital significance due to its direct effects on catalyst performance. Nafion solution is added into the catalyst mug, providing increased number 3-phase points and Pt utilization. Another important benefit of Nafion usage is the water management. Nafion is highly hydrophobic, therefore addition of Nafion into the catalyst layer prevents the water accumulation.

Another major challenge in wide spread utilization of fuel cells is the water accumulation at electrochemically active area [14,15]. Thus, implementation of effective water management mechanisms which sufficiently moisten the membrane while preventing the accumulation of water in the fuel cell components is essential [16]. Saturating the membrane with sufficient water is necessary to prevent a reduction in proton conductivity and fuel cell efficiency [17]. On the other hand, excessive use of Nafion may cause loss of electrical conductivity between catalyst particles and gas diffusion layer, preventing gases from reaching the catalytic site. This obstacle of electron transfer is observed as ohmic resistance in equivalent circuit and it may cause very high overpotential. Therefore, a precise Carbon:Nafion ratio needs to be determined. The determination of critical point for Catalyst:Nafion ratio via experiments is demanding and costly, since many other experimental parameters may affect the PEM fuel cell results. Moreover, the critical point may vary with respect to catalyst synthesis procedure and catalyst particle size. Thus, instead of conducting a large number of experiments, construction of a mathematical model based on specific experimental results is more convenient and significant. Such a model would be a useful tool for estimation of optimum ratio without conducting any experiment.

Mathematical models are tools that represent defined system behavior under possible conditions [18]. Various models have been developed for fuel cell systems to provide a deeper understanding of fuel cell processes [19]. Modeling is helpful in understanding the responses occurring under different parameters and operating conditions. Modeling on experimental data has a critical role for the optimization of fuel cell performance parameters. Fuel cell models can be examined in different categories, including time consideration, modelled process, model dimensions and level [1].

Electrochemical and thermal modeling studies on the PEM fuel cells are carried out using different dimensions, techniques and methods or software. Laribi et al., analyzed drying and flooding on PEM fuel cell and developed a zero-dimensional optimal impedance model at stationary state. With this study, the physical properties of the electrochemical impedance model can be predicted in the case of drying and flooding in PEM fuel cells [20]. Cheddie and Munroe developed a one-dimensional model by analyzing critical points of the performance. The main parameters investigated in this study were Pt weight fraction, GDL porosity, and catalyst activation. They concluded that increase in GDL porosity provided an increase in gas permeability and a decrease in the ionic and thermal conductivity [21]. Sun et al., analyzed effects of GDL porosity and GDL thickness on the cell performance by developing a two-dimensional model. It was observed that increases in GDL thickness causes decrease in the performance, and identifying and modeling other effects of GDL on the fuel cell performance is crucial for the advancement of fuel cell technologies [22]. Seinderberger et al., studied a model for predicting degradation and water transport in GDLs. A three-dimensional GDL model was constructed, but complexity prevents it from being used for process control [23]. The effect of catalyst layer thickness on PEM fuel cell performance and durability has also been investigated in the literature [24,25].

Fuel cells can be modelled by using some programs, such as Matlab/Simulink, Comsol Multiphysics, Ansys Fluent, Engineering Equation Solver etc. Han et al., used Matlab to optimize the fuel cell efficiency. The theoretical and experimental data obtained were compared and found to be similar to each other. This situation revealed that the proposed model will provide a significant increase in fuel cell efficiency [26]. Li and Sundén, utilized Ansys Fluent to construct a three-dimensional model to describe the effects of deformation of GDL at stationary state. The oxygen concentration, temperature, volumetric current density, and liquid water saturation of fuel cells were investigated [27]. Kwan et al., simulated a hybrid fuel cell system with the help of Comsol Multiphysics program. This three-dimensional model has revealed deep insights about the fuel cell-thermoelectric hybrid systems [28]. Schumann et al. developed and implemented an electrical equivalent circuit for electrically controllable fuel cells by using Matlab/Simulink [29]. Asensio et al., developed a theoretical, zero-dimensional model, which is used to control the system. The developed optimization strategy contributes to the reduction of costs related to the electricity generation [30].

Recent studies generally focused on optimizing parameters of the fuel cell. Kanchan et al., developed a model to investigate the effect of porosity configurations in GDL of a high temperature fuel cell [31]. Xiao et al., modified the structure of carbon paper and obtained a model from experimental data [32]. Li et al., constructed a two-dimensional model to examine transport properties of the fuel cell by changing membrane thicknesses [33]. However, as far as we know, the effect of Catalyst:Nafion ratio on the performance of the PEM fuel cells has not been thoroughly analyzed. Also, throughout our literature investigation, we have not found a study which directly relates the current density with rotation rate and Pt coating.

In this study, the main objective is to investigate the electrochemical properties of Catalyst:Nafion content in the electrochemically active field and determine how different ratios affect water transport. While the addition of Nafion prevents water accumulation, it also limits the electron transfer by obstructing gases from reaching the catalyst. Hence, it is crucial to determine optimum Carbon:Nafion ratio for production of efficient fuel cells. To this end, a series of experiments are conducted with the same catalyst, changing the Carbon:Nafion ratio. The effect of Nafion on catalyst performance towards the oxygen reduction reaction was investigated. The cyclic voltammogram (CV) of the electrodes were obtained for different Carbon:Nafion ratios. From these curves, electrochemically active surface areas (ECSA) were calculated for each Carbon:Nafion ratio. Linear sweep voltammetry (LSV) curves were also obtained for different Carbon:Nafion ratios and at different rotation rates to determine the number of electrons for each oxygen molecule (n value). Furthermore, by using the experimental data obtained, a mathematical model was developed to determine the current density directly from rotation rate and Pt coating. The obtained model is expected to be beneficial in terms of limiting costly and time-consuming PEM fuel cells experimental work for optimization. The researchers can use the proposed model to estimate current density and n value of the catalyst in our experimental range.