Gas Diffusion Layers for PEM Fuel Cells

CHAPTER

GAS DIFFUSION LAYERS FOR PEM FUEL CELLS: EX- AND IN-SITU CHARACTERIZATION

2.28

Adnan Ozden1, Ibrahim E. Alaefour1, Samaneh Shahgaldi1, Xianguo Li1, C. Ozgur Colpan2, Feridun Hamdullahpur1 _ University of Waterloo, Waterloo, ON, Canada1; Dokuz Eylul University, Izmir, Turkey2

1. INTRODUCTION Sustainable and alternative methods of power generation have been under vehement development to meet ever-increasing energy and power requirements [1]. Fuel cells, specifically proton exchange membrane (PEM) fuel cells, which produce electric energy from the direct electrochemical reaction between a fuel (usually hydrogen) and an oxidant (usually oxygen), have become promising power sources for many practical applications ranging from stationary to transportation [2]. This popularity arises from their unique characteristics, such as high energy conversion efficiency, minimal to zero total emissions, relatively low operating temperatures, quick start-up and shut down, fast response to dynamic loads, quiet operation, and the ability to operate without moving parts [3e5]. Currently, PEM fuel cells have reached the early stage of commercialization with further technical barriers to be overcome for cost reduction, durability, and reliability; these have much to do with the gas diffusion layer, a key component of PEM fuel cells [6e8]. The gas diffusion layer (GDL), as shown in Fig. 1, is one of the critical components of a PEM fuel cell. It has crucial roles in the operation of a PEM fuel cell by (1) controlling the transport of the reactants and by-products to and from the catalyst layers (CLs); (2) providing pathways for electron conduction between the CL and the flow field (FF); (3) maintaining a delicate balance between membrane hydration and water removal, thus offering acceptable heat and water management; and (4) reinforcing the mechanically sensitive CL. The GDL, which is typically a carbon-based and wetproofed product used in a membrane-electrode assembly (MEA), is situated between the CL and cathode flow field (CFF) (Fig. 1A and B); its design strongly affects PEM fuel cell performance [9]. The liquid and gas species travel through the void region of the GDL, whereas electrons that are generated at the anode catalyst layer and transported to the cathode by means of an external circuit are carried through the solid region. Specifically, the void region is functionally responsible for the transport of the liquid and gas species, whereas the solid region basically acts as a bridge by providing a pathway for electrons and heat [10]. Although carbon cloth or paper has conventionally been Exergetic, Energetic and Environmental Dimensions. http://dx.doi.org/10.1016/B978-0-12-813734-5.00040-8 Copyright © 2018 Elsevier Inc. All rights reserved.

695

696

CHAPTER 2.28 GAS DIFFUSION LAYERS FOR PEM FUEL CELLS

FIGURE 1 (A) A schematic of a proton exchange membrane (PEM) fuel cell and its components, (B) a schematic of the working principle of a PEM fuel cell. Adapted from Wu H. Mathematical modeling of transient transport phenomena in PEM fuel cells [Ph.D. thesis]. University of Waterloo; 2009.

employed as porous materials to fulfill the aforementioned functions of the GDLs, carbon paper is generally preferred for both the anode and cathode sides. Hydrophobic treatment of a single-layer GDL has recently come into prominence in PEM fuel cell applications to control the wettability of the GDL (Fig. 2), that is, to effectively remove the accumulated water at the cathode. It has been demonstrated that deposition of a microporous layer (MPL), consisting of a carbon or graphite powder and a polymeric binder [e.g., polytetrafluoroethylene (PTFE), fluorinated ethylene propylene (PVDF),

1. INTRODUCTION

(A)

(B)

697

(C)

FIGURE 2 Scanning images of commonly used carbon papers as the gas diffusion layers in proton exchange membrane fuel cells: (A) face view of Avcarb EP40 with 0% polytetrafluoroethylene (PTFE), (B) face view of TORAY-TGPH120 with 40% PTFE, and (C) face view of microporous layerecoated Avcarb EP40 with 0% PTFE.

and fluorinated ethylene propylene (FEP)], to the face of a GDL in contact with the CL improves cell performance, particularly at high current densities, by promoting effective water removal [11]. In principle, an ideal MPL should enhance PEM fuel cell performance by (1) assisting in the distribution of oxygen to the cathode catalyst layer (CCL); (2) reducing liquid water saturation from the CCL to the CFF by increasing capillary pressure owing to its hydrophobicity and smaller pores (1e10 mm) compared with the GDL (10e100 mm), but larger pores compared with the CCL (0.001e5 mm); (3) decreasing the electrical interfacial resistance between the CCL and GDL; (4) enhancing the mechanical compatibility between the GDL and CCL; (5) providing mechanical support for the delicate CL; and (6) increasing catalyst utilization by preventing the precious metal catalyst from penetrating deeply into the GDL [13,14]. Up to now, because the design parameters in the manufacture of MPLs [e.g., wettability, porous structure, carbon loading (or layer thickness), polymeric binder content, placement configuration, type of carbon-based material] affect the architecture, physical and structural characteristics of GDLs, numerous studies have focused on these parameters [15e17]. With exponentially growing interest in the improvement of GDLs to achieve improved characteristics for PEM fuel cells, it has become essential to understand the state-of-the-art architectures of GDLs, as well as ex-situ and in-situ characterization techniques that have been intensively used in GDL characterization. Because GDLs serve as a bridge to electrochemical reactions in the CLs and eventual removal of reaction product water and heat through the reactant flow field and/or cooling flow, and the ability to transport reactant gas, product water, and heat through the GDL is critical to PEM fuel cell performance, durability, and reliability, this chapter focuses on a review of the structural and physical characteristics of GDLs (e.g., porosity, electrical and thermal conductivity, gas permeability, and wettability) that are directly related to the transport phenomena, and related experimental techniques used for their ex- and in-situ evaluation. Furthermore, critical design parameters for the MPLs (e.g., type and loading of carbon black powder, hydrophilic and hydrophobic treatment, and microstructure modification) are described, with a particular emphasis on their potential effect on PEM fuel cell performance.

698

CHAPTER 2.28 GAS DIFFUSION LAYERS FOR PEM FUEL CELLS

2. CHARACTERISTICS OF GAS DIFFUSION LAYER An ideal GDL, which is basically a carbon-based product (e.g., woven carbon cloth, carbon foam and nonwoven carbon paper) sandwiched between the CL and FF plate, should fulfill the following requirements [18e20]: 1. Have sufficient porosity to allow the flow of both the reactant gases and by-products that transport in the opposite direction. 2. Ensure balanced hydrophilicity (water retention) and hydrophobicity (water expulsion). 3. Provide acceptable electrical and thermal conductivities in both through- and in-plane directions. 4. Maintain a crack-free morphology during long-term operation. 5. Offer optimum mechanical integrity to support the delicate CL during the operation. 6. Allow acceptable thermal and chemical resistance along with sufficient durability. Because some of these requirements conflict with each other, they need to be carefully balanced to achieve performance improvement. This balance can only be achieved through an overall understanding of the main characteristics of GDLs. Therefore, this section provides a comprehensive review of the main characteristics of GDLs.

2.1 POROSITY AND PORE SIZE DISTRIBUTION A GDL is basically composed of an electrically conductive porous medium, primarily referred to as a macroporous substrate (MPS), and a relatively thin MPL (Fig. 2C). As a favorable MPS material, as presented in Fig. 2A, carbon fiberebased products, including woven cloth and nonwoven paper, have been extensively employed due to their high electrical conductivity and porosity [21]. On the other hand, the MPL, which is responsible for managing water transport, improving catalyst utilization, and reducing the electrical interfacial resistance between the CL and GDL, is basically composed of a carbon black powder and a hydrophobic agent (Fig. 2C). Owing to its unique structure (its small pore size), the MPL serves as a pressure valve with two functions: pushing water from the cathode to the anode (a process referred to as “back diffusion”) and allowing pressure to build up and expel the water across the less-hydrophobic pores of the GDL [22]. In the former case, the droplets are forced to pass through the membrane by high liquid water pressure arising mainly from the oxygen reduction reaction (ORR) and electro-osmotic drag, whereas in the latter case, the water droplets generated at the CCLemembrane interface are transported through the GDL and then expelled from the cell via cathode flow channels [16]. As has been emphasized, equilibrium between the former and latter cases is required for effective water removal and sufficient membrane hydration, and the porosity and pore size distribution play a key role in this context. As the critical structural characteristics of a GDL, porosity and pore size distribution are indeed of paramount importance, because they significantly influence several performance-effecting phenomena, including transport of reactants and by-products, ohmic resistance, and formation of liquid water saturation profiles [23]. The pore size in an MPS (10e100 mm) is distinguishably higher than that of an MPL (1e10 mm) [13]. The pores of the GDLs presented in Fig. 2A and B are therefore much more visible than those of the MPL-coated GDL displayed in Fig. 2C, even though all the images were obtained with the same measurement scale (200 mm). Small pore sizes within a GDL improve its water expulsion characteristics and make it hydrophobic.

2. CHARACTERISTICS OF GAS DIFFUSION LAYER

699

However, large pores are important for reactant transport. In general, pores within GDLs are classified based on their size [24]: 1. Pores smaller than 0.01 mm are classified as micropores. 2. Pores in the range of 0.01e5 mm are classified as mesopores. 3. Pores larger than 5 mm are classified as macropores. For micropores (<0.01 mm), a diffusion mechanism, primarily referred to as Knudsen diffusion, is dominant. However, diffusion through macropores is basically controlled by another diffusion mechanism generally referred to as bulk diffusion [24]. In macropores, therefore, gas molecules principally diffuse based on molecular collisions, which further confirms the importance of macropores in effective mass transport. These conflicting effects make optimization of these three pore types inside GDLs inevitable because optimized pore size distribution facilitates both reactant distribution and water removal [25]. Here, pores classified in the range of 0.01e5 mm represent a transition between micropores and macropores. Thus, mesopores are practically preferred because they balance the aforementioned effects.

2.2 ELECTRICAL CONDUCTIVITY In a well-designed PEM fuel cell, ohmic losses, which mainly result from the internal (bulk) resistances of each component [e.g., CL, MPL, MPS, anode flow field (AFF), and CFF], as well as from the interfacial resistances among these components, should be minimized. Internal resistances are primarily attributed to the characteristics of the CL (e.g., thickness, tortuosity, uniformity, porosity, surface roughness, and ionomer content), GDL [e.g., porosity, tortuosity, thickness, and hydrophobic agent (PTFE) content], MPL (e.g., type and loading of carbon black powder, hydrophobic agent content, and uniformity), and AFF and CFF plates (e.g., material properties, FF geometry, and thickness) [26,27]. Among these components, the GDL is a critical one that significantly contributes to ohmic losses, because it is mainly responsible for the transport of electrons from and to FF plates, and electrical contact between the GDL and FF plates is not continuous, rather periodic, due to the presence of the FFs. Therefore, electrons travel both in the in-plane and through-plane directions for different parts of the GDL. Depending on the morphological structures of the MPS and MPL, the transport characteristics of the GDL (e.g., electrical conductivity) demonstrate significant differences in both the through- and in-plane directions [15]. Nevertheless, in many developed computational fluid dynamics models, the transport characteristics of GDLs (e.g., electrical conductivity) have customarily been assumed to be isotropic [28e30]. This assumption is considered “questionable” [31] because GDLs are composed of carbon fibers, and these carbon fibers are normally oriented through the in-plane direction (Fig. 2A and B). Thus, the main transport characteristics, such as gas permeability and electrical and thermal conductivity, are reported to be completely different in the through- and in-plane directions. For instance, electrical conductivity in the through-plane direction is reported to be at least one order of magnitude lower than that in the in-plane direction [32,33]. In the operation of a PEM fuel cell, the GDL does not entirely contact the FF plates because of the geometry of the FF plates. Therefore, electrical conductivity in the in-plane direction becomes important in compensating for the ohmic losses arising from the places where through-plane electrical conduction is relatively poor (such as underneath the flow channels) [34]. In this context, determination of the electrical conductivity characteristics of GDLs in both the through- and in-plane directions is required for GDL optimization.

700

CHAPTER 2.28 GAS DIFFUSION LAYERS FOR PEM FUEL CELLS

2.3 THERMAL CONDUCTIVITY An operating PEM fuel cell generates a considerable amount of heat, which causes the formation of a temperature gradient as a result of continuous electrochemical reactions, transport of electrons and protons, and irreversibilities arising from the electrochemical reactions [15]. The heat generated inside the cell is transported by conduction through the components, together with relatively less dominant radiation and convection mechanisms. The control and effective removal of generated heat from the cell via implementation of efficient, reliable and cost-effective cooling strategies is of paramount importance for achieving short- and long-range efficiency, reliability and durability goals [35]. Hence, the transport of water and species, reaction kinetics at both the anode and cathode, and level of phase change are influenced by the temperature distribution and heat transfer mechanism inside the cell [36]. In parallel, performing thermal analysis for PEM fuel cells is essential to determine the thermal conductivity characteristics of each individual component. Toward this goal, significant research efforts have been made to understand the thermal conductivity characteristics of the main components, such as FF plates, membranes and GDLs [37]. Unlike with FF plates and membranes, determination of the thermal conductivities of GDLs in both the in- and through-plane directions is relatively challenging because of their anisotropic architecture (in which brittle micron-scale carbon fibers are positioned randomly) [15]. In addition, significant differences in thermal conductivity between the carbon fibers and water/air phases make the determination of thermal conductivity characteristics of GDLs even more challenging [36]. Nevertheless, for effective heat management inside PEM fuel cells, it is indeed important to determine the heat transfer characteristics of GDLs in both the through- and in-plane directions, because heat transfer occurs in both directions, particularly underneath the flow channels [38].

2.4 GAS PERMEABILITY In the operation of a PEM fuel cell, the generated or transported water at the cathode is absorbed into the ionomer embedded into the CL and then transported through the void regions of the CCL and GDL via convection with both in- and through-plane velocity components. However, such water transport occurs only in cases in which the negative-pressure gradient of the gas phase is overcome by the capillary pressure gradient, a mechanism mainly governed by the porosity and gas permeability of GDLs [17]. GDLs typically undergo numerous chemical, mechanical, and thermal processes as well as treatments, and their physical and transport properties, particularly gas permeability, electrical and thermal conductivity, and wettability, are widely affected by the specifications of these processes and treatments [39]. In addition to the manufacturing processes, GDLs are commonly treated with hydrophobic agents (Fig. 2B), which usually makes them relatively condenser and reduces their through-plane gas permeability, to facilitate the detachment of liquid water and mitigate mass transport limitations associated with water flooding [15]. However, gas permeability in both the through- and in-plane directions is an important characteristic of GDLs, because it influences the reactant transport, removal of by-products, and water saturation inside the cell. Thus, gas permeability in both the through- and in-plane directions, like many other transport characteristics of GDLs, should be accurately measured to help optimize the design and manufacturing parameters affecting it [11].

2. CHARACTERISTICS OF GAS DIFFUSION LAYER

701

2.5 WETTABILITY Particularly, at high current densities, the water production rate becomes greater, leading to increased liquid water formation; eventually it becomes problematic because it blocks electrochemically active sites within the CCL. It is well established that the hydrophobic (water-expelling) or hydrophilic (water-retaining) characteristics of GDLs are paramount in water management [40,41]. The hydrophobic or hydrophilic characteristics of GDLs depend on the roughness and energy of the surface, which are conventionally represented by the concept of a static contact angle [42]. In principle, this angle is described as the measured angle at the solid, liquid and gas interface. As illustrated in Fig. 3A, a static contact angle higher than 90 degrees is characterized as hydrophobic, whereas one lower than 90 degrees is characterized as hydrophilic [43]. The static contact angle on an ideally flat surface is theoretically defined by the Young’s equation given in Eq. (1), and basically indicates the surface free energies of materials (a GDL in this case) because it is solely dependent on the intrinsic properties of materials [42]:
 gsg  gsl 
cosðqc Þ ¼ (1) glg where gsg , gsl , and glg are the surface free energies of solidegas, solideliquid, and liquidegas interfaces, respectively, and qc represents the static contact angle. Herein, it is important to note that Eq. (1) is valid only for smooth surfaces, because surface roughness has a critical role in the wettability characteristics of materials [44].

FIGURE 3 Illustrations of (A) static contact angle ðqc Þ measurements on hydrophilic and hydrophobic surfaces, and (B) sliding contact angle ðqs Þ measurement. Adapted from Jiao K, Li X. Effect of surface dynamic wettability in proton exchange membrane fuel cells. International Journal of Hydrogen Energy 2010;35:9095e03. http://dx.doi.org/10.1016/j.ijhydene.2010.05.027.

702

CHAPTER 2.28 GAS DIFFUSION LAYERS FOR PEM FUEL CELLS

Although measurement of the static contact angle can give a general understanding of materials’ hydrophobicity and hydrophilicity, it is insufficient for evaluating the mobility of a liquid on the surface of the material [42]. Thus, as an alternative approach to the static contact angle, a method, referred to as the sliding contact angle, has been employed to evaluate the dynamic wettability characteristics of solid surfaces [45]. As seen from Fig. 3B, the sliding angle demonstrates the critical angle at which a droplet with a certain weight begins to slide down from an inclined surface. In this context, a lower sliding angle indicates a more hydrophobic surface characteristic, whereas a higher one indicates a more hydrophilic surface characteristic [42]. The correlation derived by Furmidge [46], as given in Eq. (2), describes the relationship between the sliding angle and the surface and liquid properties: sinðqs Þmg ¼ sRkðcosðqr Þ  cosðqa ÞÞ

(2)

where qs , qr , and qa represent the sliding angle, the receding contact angle, and the advancing contact angle, respectively (Fig. 3B), and m is the mass of liquid droplet, g is the gravity, s is the surface tension coefficient, R is the length scale, and k is the shape constant. Further information on the sliding contact angle measurement can be found in Jiao and Li [42] and Guo and Liu [47]. The approaches discussed in this section are solely applicable for determining the qualitative wettability characteristics of GDLs via measuring the external contact angles. However, specifically in cell operation, the specifications of chemical compositions (e.g., carbon powder loading and the content of the hydrophobic agent in the MPL) of GDLs become more dominant and affect performance differently. Consequentially, the external contact angle measurements have some restrictions in reference to providing quantitative information for the purposes of design or calculation [48].

3. EX-SITU CHARACTERIZATION OF GAS DIFFUSION LAYER The structural and physical characteristics of GDLs, such as porosity, electrical and thermal conductivity, gas permeability, and wettability, can be investigated by ex-situ characterization techniques. These characterizations can be performed for both pristine and post-mortem GDLs, which enables fuel cell researchers to obtain secondary information about the possible failure modes of GDLs [34]. Therefore, this section focuses on the recent and most common ex-situ characterization techniques employed to determine the structural and physical characteristics of GDLs.

3.1 POROSITY AND PORE SIZE DISTRIBUTION The microstructures and surface characteristics of GDLs (e.g., porosity and pore size distribution), which are useful for comparing different GDLs, have been successfully determined through several techniques, such as intrusion methods based on mercury [11,49], water [50], or kerosene [51], capillary flow porometry [34], and mercury standard porosimetry (MSP) [52]. Among these techniques, the method of mercury intrusion porosimetry (MIP), MSP, and capillary flow porometry have been extensively used owing to the wide range of spectrum of measurable pore radii (from 2 to 105 nm) [53], which is practically sufficient for the range of pore sizes commonly encountered in GDLs as the components of PEM fuel cells. However, all these techniques principally involve general assumptions, because the capillary structure of the sample is basically represented by a bunch of tubes with a certain

3. EX-SITU CHARACTERIZATION OF GAS DIFFUSION LAYER

703

range of radii; therefore, it is strongly suggested that the inferences about the internal structures of the measured samples should be made with particular consideration [34].

3.1.1 Mercury Intrusion Porosimetry MIP basically relies on using high pressure to force mercury into pore spaces within a porous medium so as to determine its porosity and pore size distribution [15]. For this technique, the GDL sample is situated inside a vacuumed glass tube to guarantee that the only liquid inside the tube is mercury. Assuming that the mercury will only penetrate the porous medium only as a result of externally applied pressure, the initial volume of mercury is measured after it is put inside the glass tube. External pressure is then applied to help the small amount of mercury permeate the largest pores of the porous medium (because the higher the pore size is, the less resistance to penetration the mercury will experience). The volume of the mercury that has penetrated is recorded. The pressure is then increased gradually until the smaller pores also fill with the mercury. The data on the volumes of mercury intrusion at specific pressure values are used to generate a capillary pressure versus saturation curve. Overall, with this method, it is possible to determine the hydrophobic and hydrophilic pore volume (total pore volume). However, accurate detection of total pore volume is not always possible, particularly for porous medium samples with smaller pore volume [54]. Thus, the same analysis may be performed with different working fluids, such as water and kerosene, to achieve more accurate results.

3.1.2 Method of Standard Porosimetry The main concept behind MSP is that changes in the weight of a working fluid in a GDL can be used to determine the capillary pressure and wettability characteristics of the test samples [55]. The type of working fluid is practically determined according to the test of interest, specifically to determine the pore size distribution (volume of hydrophobic and hydrophilic pores) in the GDL. Ethanol and octane are preferentially employed, whereas larger pores (hydrophilic pores) are more easily measured when water is used as the working fluid. This method employs three samples: two of them are standard (23 mm diameter) and the other is the tested GDL sample. These samples are stacked (the GDL under test is placed between the standard samples) (Fig. 4), and it is known that they are in capillary equilibrium. Thus, the capillary pressure of the standard samples is equal to that of the sample under test. Any change in their capillary pressure will affect both the standard and test samples equally; therefore, the capillary pressure of the GDL sample can be obtained from the capillary pressure curve generated for the standard samples [15]. To this end, the standard and GDL samples are placed inside a glass tube and vacuumed at a temperature of 180 C for 2 h to minimize the number of pores filled with air. Thereafter, the samples are immersed in the working fluid and air is penetrated to evaporate the working fluid within the samples. Finally, the samples are weighed at intervals and their weights are recorded while they are still stacked, to determine the amount of saturation of the working fluid. Because the capillary pressure changes with the saturation of the working fluid, it is possible to determine the capillary pressure of the GDL sample by comparison with the known capillary pressure curve of the standard sample. Generating the capillary pressure versus the saturation curve for the tested sample involves the steps, schematically presented in Fig. 4 and summarized as follows [15]: •

Step 1: Determination of the sample’s respective saturation levels from the data obtained for each individual sample from the intermittently measured weight of each sample, because their

704

CHAPTER 2.28 GAS DIFFUSION LAYERS FOR PEM FUEL CELLS

FIGURE 4 Schematic of the experimental apparatus, and the steps applied to generate capillary pressure curves for a gas diffusion layer (GDL) using a standard porosimetry method. Adapted from Zamel N. Transport properties of the gas diffusion layer of PEM fuel cells [PhD thesis]. University of Waterloo; 2011 and Zamel N, Li X. Effective transport properties for polymer electrolyte membrane fuel cellsdwith a focus on the gas diffusion layer. Progress in Energy and Combustion Science 2013;39:111e46. http://dx.doi.org/10.1016/j.pecs.2012.07.002.

•

•

capillary pressure can be found from the saturation of the working fluid (which is the knowledge obtained from the intermittent measurements) Step 2: Determination of the capillary pressure of the test sample by assuming both the test and standard samples are in capillary equilibrium (thus, the capillary pressure of the test sample is equal to that of the standard samples) Step 3: Repetition of the first two steps for each individual saturation measurement

3.1.3 Capillary Flow Porometry Capillary flow porometry is similar to intrusion methods based on different working fluids [56]. For the potential working fluids, instead of nonwetting fluids (e.g., mercury), wetting fluids, including octane and water, are commonly used. Thus, the method of capillary flow porometry is considered to be advantageous compared with MIP, because it is relatively faster and nondestructive [34]. In addition, because of the highly wetting properties of octane, this method enables the determination of both hydrophobic and hydrophilic pores (however, water is preferentially employed to determine hydrophilic pores). In this method, first, gas pressure and flow rates through the tested GDL sample are

3. EX-SITU CHARACTERIZATION OF GAS DIFFUSION LAYER

705

recorded, and then the wetting liquid is soaked into the pores of the sample and the bulk material porosity of the sample is measured [57]. Once the pores of the tested sample are completely filled with the wetting fluid, a pressurized and nonreacting gas is applied to the surface of the porous medium to remove the wetting liquid from the pores. The pressure required to remove the wetting liquid from the pores of the sample can be described, as in Eq. (3), as the function of pore diameter (D), the surface tension of the wetting liquid (g), and the contact angle of the wetting liquid (q) [34]: P¼

4g cosq D

(3)

Using the pressure difference between the wetting liquid and gas, the function of pore size distribution (f) can be described as in Eq. (4) [34]:   Fw D Fd f¼  100ð%Þ (4) DD where Fw and Fd are the flow rates of wet and dry gas, respectively. That is, in this technique, the pore size distribution throughout the porous medium can be determined from the difference between the flow rates of wet and dry gases using the small pressure difference for through-plane flow [58].

3.2 ELECTRICAL CONDUCTIVITY This section presents two extensively employed techniques to determine the electrical conductivity characteristics of GDLs in the through- and in-plane directions.

3.2.1 Through-Plane Electrical Conductivity In its simplest form, measurement of through-plane electrical conductivity, as discussed in detail in Ismail et al. [32] and Reum [59], involves (1) sandwiching a disk-shaped GDL sample (with a diameter of 10 mm) between two highly conductive steel disks (with a diameter and thickness of 10 and 1.5 mm, respectively), (2) placing the obtained stack between the highly conductive electrodes (usually goldplated electrodes), (3) situating the GDL-electrode assembly between two insulator plates, and (4) applying a gradually increasing compression force through the bolts of the setup together with DC current, as illustrated in Fig. 5. Furthermore, it includes measuring the resistance of the electrodes together with the metal disks, including the contributions from the interfacial resistances between the tested GDL and metal disks, as well as the bulk resistance of the GDL, by using the voltage drop. Eq. (5) shows the contributions to the measured resistance value [32]: Rtotal ¼ 2Rel þ 2Rst þ RGDL þ 2RGDLst þ 2Rstel

(5)

where Rel , Rst , and RGDL are the bulk resistances of the electrode, steel disk, and GDL, respectively. However, RGDLst and Rstel represent the contact resistances between the couples of the GDLesteel disk and steel diskeelectrode, respectively. For the resistance of the setup (in the case in which no GDL is situated between the steel disks), the correlation expressed in Eq. (6) can be used: Rsetup ¼ 2Rel þ Rst þ 2Rstel

(6)

706

CHAPTER 2.28 GAS DIFFUSION LAYERS FOR PEM FUEL CELLS

FIGURE 5 Schematic of the experimental apparatus for the through-plane electrical conductivity measurement of the gas diffusion layer (GDL). Adapted from Ismail MS, Damjanovic T, Ingham DB, Pourkashanian M, Westwood A. Effect of polytetrafluoroethylene-treatment and microporous layer-coating on the electrical conductivity of gas diffusion layers used in proton exchange membrane fuel cells. Journal of Power Sources 2010;195:2700e8. http://dx.doi.org/10.1016/j.jpowsour.2009.11.069.

where Rsetup is the total resistance of the setup schematically displayed in Fig. 5. Eq. (7) represents the contact resistance between the GDL and steel plate, found by subtracting Eq. (6) from Eq. (5): 1 RstGDL ¼ ðRtotal  Rsetup  Rst  RGDL Þ 2

(7)

The bulk resistance of the steel disk can be calculated by multiplying its resistivity and thickness, whereas that of the GDL can be obtained by following a procedure similar to that reported in Zamel and Li [15] and Arvay et al. [34]. Because of the porous structure of the GDL, the bulk resistivity of the GDL ðrGDL Þ can be calculated by considering the volume fraction weighted harmonic mean of the resistivity of the carbon fiber and air, as given in Eq. (8) [60]: rf ¼ ð1  εÞrGDL

(8) 5

where rf is the resistivity of the carbon fiber, estimated to be 4.02  10 U m, as reported in Zamel and Li [15] and Mishra et al. [60]. In light of this equation, the porosity of the GDL can be calculated,

3. EX-SITU CHARACTERIZATION OF GAS DIFFUSION LAYER

707

as reported in Barbir [61], by using the correlation in Eq. (9). In this equation, the bulk density of the GDL is suggested to be taken as 1.80 g/cm3 [62]: !  Areal weight ðkg=m2 ε¼1 (9) Thickness of the GDL ðmÞ  Bulk density of the GDLðkg=m3 Þ The calculated porosity of the GDL via Eq. (9) can also be used to determine the resistivity of the GDL through Eq. (8). The calculated resistivity of the GDL can then be converted into its bulk resistance by multiplying it with the thickness (d).

3.2.2 In-Plane Electrical Conductivity The electrical conductivity of GDLs is usually measured by a method, primarily referred to as the method of four-probe, described by Smits [63]. In this technique, the GDL sample, as seen in Fig. 6, is positioned between the copper electrodes and an insulating plate. DC is applied from each end of the GDL sample through the copper electrodes via a four-probe device, as illustrated in Fig. 6. The voltage difference between the two selected points (which are located at the middle of the GDL) is measured via two gold-plated probes and the four-probe device. Then, the resistance between these selected points can simply be calculated by Ohm’s Law. Here, the distance between the probes should be kept the same to achieve correct measurements. In his technique, Smits [63] also described a general mathematical formula that involves a term defined as a geometry-dependent correction factor and is both the function of the dimensions of the GDL sample and the space between the probes [32]. Here, the geometry-dependent correction term is the function of the ratio between the length and width of a GDL and the width and distance between the probes. These ratios were found by Ismail et al. [32] to be

FIGURE 6 Schematic of the experimental apparatus for in-plane electrical conductivity measurement of gas diffusion layers (GDLs). Adapted from Ismail MS, Damjanovic T, Ingham DB, Pourkashanian M, Westwood A. Effect of polytetrafluoroethylene-treatment and microporous layer-coating on the electrical conductivity of gas diffusion layers used in proton exchange membrane fuel cells. Journal of Power Sources 2010;195:2700e8. http://dx.doi.org/10.1016/j.jpowsour.2009.11.069.

708

CHAPTER 2.28 GAS DIFFUSION LAYERS FOR PEM FUEL CELLS

5 and 1, respectively. To calculate the electrical resistivity of the GDL sample ðrGDL Þ, the linear correlation given in Eq. (10) can be used [63]: rGDL ¼ CtR

(10)

where C is the geometry-dependent correction factor, t is the thickness of the GDL, and R is the measured electrical resistance between the probes. The calculated resistivity of the GDL ðrGDL Þ can also be used to determine the electrical conductivity of the GDL in the in-plane direction. Eq. (11) gives the correlation between the in-plane electrical conductivity ðsGDL Þ and the electrical resistivity of the GDL: sGDL ¼

1 rGDL

(11)

Here, the in-plane conductivity measurements should be conducted with multiple samples (at least six) that have been cut off from the GDL sheet in different directions, to keep the error range associated with the anisotropy of the GDL at a minimum.

3.3 THERMAL CONDUCTIVITY This section presents the recent and most extensively used characterization techniques for the determination of the through- and in-plane thermal conductivity characteristics of GDLs.

3.3.1 Through-Plane Thermal Conductivity The technique presented in this section, generally referred to as the guarded heat flux meter method, has been extensively used to determine through-plane thermal conductivity as well as thermal contact resistance between thin films [64,65]. As schematically presented in Fig. 7, in this method, the GDL sample is sandwiched between two highly conductive cylindrical bars. The top cylindrical bar is completely in contact with a hot plate and the bottom one is in contact with a cold plate, and thus there is continuous heat flux from top to bottom through the axis of the cylindrical bars. Hence, the experimental apparatus is designed to allow heat transfer in only the axial direction. The temperature distribution gradient on the bars is measured by thermocouples (which are, as seen in Fig. 7, located side by side, with the same intervals on the top and bottom bars). The compression load applied to the GDL sample is also controlled via a load device. After the temperature variation in each measurement point becomes less than 0.5 C in w30 min, the eventual temperature versus location plot is drawn. To calculate the thermal conductivity of the GDL sample, first, the total heat flux through the axis   of the cylindrical top bar is calculated using the temperature differences dT dx between the measurement points, the cross-sectional area ðAc Þ of the cylindrical bars, and the thermal conductivity ðkðTÞÞ of the bars as given in Eq. (12): dT (12) dx To determine the drop in temperature due to the presence of the GDL sample, the total resistance between the positions (where the upper and lower thermocouples are placed) and calculated heat flux can be used as given as in Eq. (13): Q ¼ kðTÞAc

DTsample ¼ Rtotal Q

(13)

3. EX-SITU CHARACTERIZATION OF GAS DIFFUSION LAYER

709

FIGURE 7 Schematic of the experimental apparatus for through-plane thermal conductivity measurement of gas diffusion layers (GDLs). DAQ, data acquisition; MEA, membrane-electrode assembly. Adapted from Khandelwal M, Mench MM. Direct measurement of through-plane thermal conductivity and contact resistance in fuel cell materials. Journal of Power Sources 2006;161:1106e15. http://dx.doi.org/10.1016/j.jpowsour.2006.06.092.

where DTsample , Rtotal , and Q represent the temperature drop due to the existence of the GDL sample, the total resistance between the upper and lower thermocouples, and the heat flux [as calculated using Eq. 12], respectively. To calculate the drop in temperature because of the GDL sample, the total resistance ðRtotal Þ between the upper and lower thermocouples can be calculated using Eq. (14): Rtotal ¼ RGDL þ 2RGDLSt

(14)

where RGDL and RGDLSt represent the thermal resistance of the GDL sample and the thermal contact resistance between the sample and cylindrical bar. RGDL and RGDLSt are reported to be the functions of thermal conductivity, surface roughness, and the micro-hardness of the GDL sample, respectively [66]. To generate a linear slope demonstrating the relation between the resistance obtained with a stack of GDL samples and the corresponding thickness of the GDL stack, the same steps can be repeated

710

CHAPTER 2.28 GAS DIFFUSION LAYERS FOR PEM FUEL CELLS

several times with double and triple GDL samples. Herein, the total resistance should be calculated considering the thermal contact resistances between the GDLs. Then, the effective thermal conductivity of the tested GDL ðkGDL Þ can be described as in Eq. (15): kGDL ¼ slope  AC

(15)

where slope represents the generated line’s slope showing the relationship between the resistances of the GDL stacks at corresponding thicknesses.

3.3.2 In-Plane Thermal Conductivity 3.3.2.1 Conventional In-Plane Thermal Conductivity Measurement Determining the in-plane thermal conductivity characteristics of GDLs is relatively more challenging than determining through-plane ones. Therefore, thus far, only a few research studies have targeted inplane thermal conductivity measurements of GDLs [36,67]. In one of these studies, a technique, similar to that discussed for through-plane thermal conductivity measurement, was introduced by Sadeghi et al. [36] for the in-plane thermal conductivity of GDLs. The schematic of the experimental apparatus employed in their study is presented in Fig. 8. The GDL samples are fixed by their ends to the sample holders. Both the design of the experimental apparatus and the experimental conditions are determined to allow heat transfer only through the samples. The heat flux, which is from the top-left to the bottom-right, is generated via an electrical heater located at the left-top of the apparatus. The temperature distribution profiles along the upper and lower flux meters are measured via thermo  couples attached to the flux meters. The differences dT dz in the measured temperatures obtained from different locations of the flux meter can then be used to calculate the heat flux ðQÞ, using Fourier’s Law of one-dimensional heat conduction, as given in Eq. (16): Q ¼ kfluxmeter A

dT dz

(16)

where kfluxmeter represents the thermal conductivity of the flux meter. Because the heat flux is constant from the top-left to bottom-right, the total resistance between the two sample holders, which is equal to that given in Eq. (17), can be determined using the temperature difference ðDTÞ between the two sample holders and the heat flux calculated using Eq. (16): Rtotal ¼

RGDL þ Rcontact1 þ Rcontact2 DT ¼ Q N

(17)

where RGDL represents the resistance of each GDL sample, and Rcontact1 and Rcontact2 represent the thermal contact resistances at the end points of each sample and are considered to be the same for each side. Thus, in the following equations, the contribution from these two thermal contact resistances is represented as Rcombined . In principle, to reduce the error range in the measurements, the GDLs are tested together as a stack (double, triple, or multiple GDLs together), where N represents the number of GDL samples in the stack. In Eq. (17), Rcontact1 and Rcontact2 (or Rcombined ), and RGDL are unknown. Therefore, all the steps should also be repeated with GDL samples of different lengths. Upon

3. EX-SITU CHARACTERIZATION OF GAS DIFFUSION LAYER

711

FIGURE 8 Schematic of the experimental apparatus for in-plane thermal conductivity measurement of gas diffusion layers (GDLs). Adapted from Sadeghi E, Djilali N, Bahrami M. A novel approach to determine the in-plane thermal conductivity of gas diffusion layers in proton exchange membrane fuel cells. Journal of Power Sources 2011;196:3565e71. http://dx.doi.org/10.1016/j. jpowsour.2010.11.151.

completion of all measurements, the in-plane thermal conductivity of the tested GDL ðkGDLin Þ can be found using Eq. (18): kGDLin ¼

NðDLÞ DRtotal AC

(18)

where DL represents the difference between the initial and final lengths of the GDL samples, and DRtotal represents the difference between the total resistances measured with the initial and final GDL samples. However, AC demonstrates the cross-sectional area of the GDL samples (which is perpendicular to the heat flux).

712

CHAPTER 2.28 GAS DIFFUSION LAYERS FOR PEM FUEL CELLS

3.3.2.2 Parallel Thermal Conductance Technique The technique presented in this section, which was originally introduced by Aaron et al. [68] and is primarily referred to as the parallel thermal conductance method, mainly relies on the determination of the in-plane thermal conductivity of GDLs through two consecutive measurements. The experimental apparatus employed for these experiments is schematically displayed in Fig. 9. The steps for the first measurement involve (1) situating two layers with a low thermal conductivity (usually glass layers) between the two ends of a vertical sample holder, (2) evacuating the air in the vacuum chamber together with applying heat from the top-left side of the apparatus, (3) arranging the voltage of the DC power supply to reach a steady-state source temperature, (4) keeping the temperature of the base plate at a constant temperature, and (5) recording the data for both the applied current and voltage to the heater and thermocouple readings once steady-state conditions are reached (typically requires 2e3 h). Upon completion of these steps, the thermal resistance ðR0 Þ between the two ends of the vertical sample holder can be calculated by using Eq. (19) [67]: R0 ¼

1 ðTh  Tc Þ IV

(19)

FIGURE 9 (A) Schematic of the experimental apparatus for in-plane thermal conductivity measurement of gas diffusion layers (GDLs). (B) Schematic of the thermal resistance network for in-plane thermal conductivity measurement of a GDL. Adapted from Teertstra P, Karimi G, Li X. Measurement of in-plane effective thermal conductivity in PEM fuel cell diffusion media. Electrochimica Acta 2011;56:1670e5. http://dx.doi.org/10.1016/j.electacta.2010.06.043.

3. EX-SITU CHARACTERIZATION OF GAS DIFFUSION LAYER

713

where I and V are the applied current and voltage values to the heater, and Th and Tc represent the temperatures at the top and bottom sides of the vertical sample holder, respectively. In this equation, the thermal resistance ðR0 Þ captures the contribution from the heat losses due to (1) conduction through the glass layers, thermocouples, and copper heater wires, and (2) radiation to the surroundings. Because the experimental apparatus is located inside an evacuated chamber, the natural convection effect is minimized, and any remaining would have also been captured in the thermal resistance ðR0 Þ. The second measurement involves all the steps given above for the first measurement. However, in the second measurement, the steps are repeated once the GDL sample with a known width and thickness is fastened in the horizontal sample holder with thermal conducting grease to minimize the contact resistance between the holder and the GDL sample (Fig. 9A). The distance between the two edges of the sample holder is measured to determine the length of the sample through which heat conduction occurs. For the second measurement, as displayed in Fig. 9B, the overall thermal resistance network can be defined as in Eq. (20) [67]: 1 1 1 ¼ þ R R0 Rs

(20)

where R is the overall resistance based on the second measurement, R0 is the resistance obtained from the first measurement, and Rs is the thermal resistance due to the presence of the GDL sample. The overall resistance ðRÞ for the second measurement can be calculated using the measured temperature differences between the two ends of the horizontal sample holder and current and voltage applied to the heater [as calculated through Eq. 19 for the first measurement]. The thermal conductivity of the GDL sample ðkÞ can be calculated by neglecting the heat losses associated with radiation via Eq. (21):   1 1 L  (21) k¼ R R0 tW where L is the length of the sample through which heat conduction occurs, t is the thickness of the GDL sample, and W is the width of the sample. The main advantage of this technique compared with the conventional one is its simplicity in measurement, because the in-plane thermal conductivity of the thin layers (GDL in this case) can be easily calculated through two simple measurements, which reduces the time required for the experiment. Unlike the conventional in-plane thermal conductivity measurement method, the experiments can be conducted with a single layer, which further minimizes the time required for sample preparation and experiments. The accuracy of the measurement could be another important advantage, because this technique enables the determination of in-plane thermal conductivity of thin layers directly, without any measurement or assumption for contact resistances.

3.4 GAS PERMEABILITY This section presents the techniques generally used to determine the gas permeability characteristics of GDLs in the through- and in-plane directions.

3.4.1 Through-Plane Gas Permeability The experimental apparatus displayed in Fig. 10 has been conventionally used by many research groups to explore the anisotropic permeability characteristics of GDLs under different compressions

714

CHAPTER 2.28 GAS DIFFUSION LAYERS FOR PEM FUEL CELLS

FIGURE 10 Schematic of the experimental apparatus for through-plane gas permeability measurement of gas diffusion layers. Adapted from Gostick JT, Fowler MW, Pritzker MD, Ioannidis MA, Behra LM. In-plane and through-plane gas permeability of carbon fiber electrode backing layers. Journal of Power Sources 2006;162:228e38. http://dx.doi.org/10.1016/j.jpowsour.2006. 06.096.

[69,70]. The experimental setup was specifically designed to replicate the pore structure of the GDL inside a single cell. As seen from Fig. 10, in this experimental apparatus, the GDL sample (generally in the shape of a disk with a diameter of 30 mm) is sandwiched between two metal plates that simulate the GDL situated between the FF plates, with variable spacing to control the compression level. The setup also allows for measurement of the pressure difference at the inlet and outlet of the cell via a differential pressure transducer, which is directly connected to a computer so that the pressure drop between the inlet and outlet of the cell can be precisely recorded. In addition, the flow rate of the air passing through the GDL sample is measured via a digital flow meter [70]. With an assumption that the velocity change of the air in the experimental setup is negligible (due to the superfine architecture of the GDL), the properties of air are assumed to be constant, the flow inside the apparatus is assumed to be single-phase, and the Reynolds number is assumed to be significantly low. With all these assumptions, Darcy’s Law given in Eq. (22) can be used to calculate the single-phase permeability of the GDL sample ðkGDL Þ [70]: kGDL ¼

vair mair Dx DPair

(22)

where vair , mair , and DPair represent the velocity, viscosity, and pressure drop of the air along the gas travel path, respectively.

3.4.2 In-Plane Gas Permeability To explore the gas permeability of GDLs in the in-plane direction, as seen from Fig. 11A and B and discussed in detail in Gostick et al. [69], an experimental apparatus is employed. In it, a rectangular GDL sample is sandwiched via feeler gauges between two plates with adjustable spacing. The sides of the experimental apparatus are sealed by clamping face plates (Fig. 11B), whereas the intersection of

3. EX-SITU CHARACTERIZATION OF GAS DIFFUSION LAYER

715

FIGURE 11 Schematic of the experimental apparatus for in-plane gas permeability measurement of gas diffusion layers. (A) Sectioned view for the internal components. (B) Exploded view. Adapted from Gostick JT, Fowler MW, Pritzker MD, Ioannidis MA, Behra LM. In-plane and through-plane gas permeability of carbon fiber electrode backing layers. Journal of Power Sources 2006;162:228e38. http://dx.doi.org/10.1016/j.jpowsour.2006. 06.096.

the face plates and the main body is sealed by means of a rubber gasket to ensure that the measured drop in the pressure of air along the travel path results only from the presence of the GDL sample. Dry air as the flowing fluid is supplied to the cell from the inlet at a certain pressure (the pressure of the supplied air is controlled via a pressure gauge) and allowed to flow only through the GDL. The pressure at the outlet of the cell is considered to be equal to atmospheric pressure. The experimental setup allows control of both the pressure and flow rate via a digital flow meter. The resistance to the flow, which is simply estimated by measuring the pressure difference between the inlet and outlet, can be used to characterize the gas permeability of the GDL sample [71].

3.5 WETTABILITY This section presents two of the most commonly used ex-situ characterization techniques to determine the wettability characteristics of GDLs.

3.5.1 Sessile Drop Technique The sessile drop technique is one of the most extensively used methods for measuring the external contact angles of GDLs. It primarily involves setting a liquid (generally deionized water) droplet on the surface of a GDL, fitting a tangent to the solid/liquid/gas phase point (which overlaps with the line where the liquid droplet touches the GDL surface), and measuring the external contact angle [72]. The experimental apparatus used for this technique is composed of a digital video camera with high resolution, software for calculating the external contact angle and surface energy, and a dispenser for generating droplets with appropriate radii. The apparatus is also connected to a sample holder (to hold the sample with different angles from 0 to 90 degrees), which makes it usable for both static and sliding contact angle measurements. During the measurements, the liquid droplet volume should be as small

716

CHAPTER 2.28 GAS DIFFUSION LAYERS FOR PEM FUEL CELLS

as possible to prevent errors arising from the weight of a droplet itself, and the data should be taken before any change in the volume and shape of the droplet due to evaporation [72].

3.5.2 Wilhelmy Plate Technique The Wilhelmy plate technique, which can be applied both statically and dynamically, is another common practice for the external contact angle measurements of GDLs. The static version mainly involves inserting a GDL sample, preferably a rectangular one, into a liquid to a known distance and measuring the force required for that insertion. However, for the dynamic version, the GDL sample is again vertically dipped into the liquid and then hoisted at a constant rate, and the force required for these steps is recorded. The Wilhelmy plate technique is reported to offer more accurate results for the external contact angle measurements of GDLs than its sessile drop counterpart, because of the larger number of measurements [34].

3.6 CLOSURE The GDL, which is a critical component of a PEM fuel cell, is anisotropic, and so the transport properties are directionally dependent. For successful simulation of the transport phenomena within a GDL, information regarding porosity, gas permeability, thermal and electrical conductivity, and wettability is needed. This chapter has therefore provided a comprehensive review of these transport properties, with a specific focus on ex-situ characterization techniques used to help understand the physical and structural characteristics of GDLs.

4. IN-SITU CHARACTERIZATION OF GAS DIFFUSION LAYERS (ELECTROCHEMICAL PERFORMANCE ASSESSMENT) Although the ex-situ characterization techniques introduced in section on the ex-situ characterization of GDLs are important for estimating the structural and physical characteristics of GDLs, in-situ characterization is essential for understanding the behavior of GDLs under actual operating conditions [34]. In principle, in-situ characterization techniques require the placement of the MEAs with GDLs that have certain characteristics inside the cell and enable a comprehensive investigation of performance under actual operating conditions. The following section is therefore focused on an assessment of the electrochemical performance of GDLs as the commonly used in-situ characterization technique. Numerous approaches have been developed to optimize the structural and physical characteristics of GDLs and hence mitigate performance deterioration associated with mass transfer limitations by tailoring their architectures. One practical approach to modifying the architectures of GDLs is to use a double-layer GDL (an MPL-coated GDL), which mainly consists of (1) a nonwoven carbon paper or carbon-fiber woven cloth and (2) a delicate MPL composed of a carbon or graphite black powder and a hydrophobic agent (e.g., PTFE, PVDF, and FEP). In principle, the manufacture of double-layer GDLs involves three steps: MPL ink preparation, coating and sintering. In the ink preparation and coating steps, carbon or graphite powder is mixed consecutively with PTFE-dispersed deionized water, organic solvents, and additives; then the resulting ink is sprayed onto the MPS. In the final step, the dual-layer GDL is first dried at w80 C for 1 or 2 h to evaporate the surfactants, and then it is sintered at w340 C for 30 min to distribute the PTFE homogeneously throughout the GDL [21]. Extensive research has

4. IN-SITU CHARACTERIZATION OF GAS DIFFUSION LAYERS

717

focused on investigating the effects of the design parameters of MPLs (e.g., type and loading of carbon black powder, hydrophilic and hydrophobic treatment, and microstructure modification) on the structural and physical characteristics of GDLs as well as on the performance of PEM fuel cells. Thus, this section reviews the effects of critical MPL design parameters on the characteristics of GDLs, with a particular interest in their potential influences on the performance of PEM fuel cells.

4.1 EFFECT OF TYPE AND LOADING OF CARBON BLACK POWDER Because of their desirable properties, including wide availability, high corrosion resistance, environmental acceptability, high thermal and electrical conductivity, and unique surface properties, carbonbased nanomaterials have been commonly used in the fuel cell industry [25]. For example, various types of carbon-based nanomaterials, including Vulcan XC-72 [21], Ketjenblack EC-300J [73], Black Pearls 2000 [74], Ketjenblack EC-600JD [75], and Acetylene Black [74], have been introduced into the MPL to modify the structural and physical characteristics of GDLs. This modification has been mainly targeted to alleviate the performance losses associated with mass transport limitations and water management through the optimization of the structural characteristics (e.g., porosity and pore size distribution). As the critical structural characteristic of a GDL, pore size distribution is indeed of paramount importance, because it affects the gas permeability and wettability characteristics simultaneously. For instance, small pore sizes within a GDL improve its water-expelling characteristics and make it hydrophobic. However, large pores are important for reactant transport. In an attempt to optimize the pore size distribution inside GDLs, Chen et al. [74] prepared cathode MPLs with different carbon powders (Acetylene Black, Black Pearls 2000, and composite carbon black based on Acetylene Black and Black Pearls 2000) and different loadings (0.50, 1.00, 1.50, and 2.00 mg/cm2) and conducted electrochemical performance studies for the MEAs based on these MPLs at different air relative humidity conditions in a single PEM fuel cell. Their results showed that it was possible to achieve improved cell performance, particularly at low relative humidity (<40%) by employing composite carbon black (containing 30 wt% Black Pearls 2000 and 70 wt% Acetylene Black) with an optimum loading (1.5 mg/cm2). Hence, Black Pearls 2000 was found to be more hydrophilic because it included the most micropores and macropores and the fewest mesopores, making it more prone to water flooding and deteriorating its mass transport properties. However, Acetylene Black was found to be comparatively hydrophobic because it included the most mesopores and the fewest hydrophilic micropores. The composite carbon black with the previously mentioned specifications was found to be promising in terms of cell performance because it included sufficient micro-, meso-, and macropores for both reactant distribution and water removal. Passalacqua et al. [76] prepared different MPLs with different carbon black powders, including Vulcan XC-72, Asbury Graphite 850, Shawinigan Acetylene Black, and Mogul L. The physical properties of these carbon powders as well as those of GDLs with these carbon powders in their MPLs are summarized in Table 1. The results obtained from the polarization behaviors of the cells with those MPLs further confirmed that the type and specifications of carbon powder affected performance considerably. The results obtained from this study were also in good agreement with the observations reported by Chen et al. [74]. In short, the physical properties of the carbon-based nanomaterials employed inside MPLs were found to influence the mass transport and water removal characteristics of PEM fuel cells. For MPL loading or thickness, the general consensus is that a very thick MPL causes high mass transfer limitations as well as electrical resistance, due to a lengthened diffusion path between the

718

CHAPTER 2.28 GAS DIFFUSION LAYERS FOR PEM FUEL CELLS

Table 1 Summary of Specifications for Both the Different Carbon Powders and the Double-Layer Gas Diffusion Layers With These Carbon Powders in Their Microporous Layers Carbon Powder Type

Surface Area (m2/g)

Pore Volume (cm3/g)

VP (cm3/g)

VS (cm3/g)

APR (mm)

APRP (mm)

Vulcan XC-72 Asbury 850 Acetylene Black Mogul L

250 13 70

0.489 0.346 0.594

0.319 0.212 0.368

0.170 0.134 0.226

1.80 3.50 1.70

0.24 0.29 0.27

4.90 8.60 4.30

140

0.276

0.157

0.119

6.00

0.20

13.60

APRS (mm)

APR, average pore volume; V, pore volume. Subscripts P and S represent the primary and secondary pores, respectively. Reorganized from Passalacqua E, Squadrito G, Lufrano F, Patti A, Giorgi L. Effects of the diffusion layer characteristics on the performance of polymer electrolyte fuel cell electrodes. Journal of Applied Electrochemistry 2001;31:449e54. http://dx.doi.org/10. 1023/A:1017547112282.

FF and CL, whereas a very thin MPL does not offer sufficient reactant distribution and interfacial characteristics, and thus both increases interfacial electrical and thermal resistances and mass transport limitations. In addition, with increasing MPL thickness, the water saturation level at the CCLeMPL interface was found to increase slightly, whereas it decreased in the pores of the macroporous structure [77]. These changes were considered to be beneficial for the transport of reactants. Consequently, an ideal carbon powder loading or MPL thickness should be (1) high enough to keep the water management and interfacial characteristics at an optimum level and (2) low enough to facilitate oxygen transport [16].

4.2 EFFECT OF HYDROPHOBIC AND HYDROPHILIC TREATMENT An appropriate hydrophobic agent content in a GDL, particularly on the cathode side, not only prevents the membrane from drying under low-humidity conditions but also offers effective water management under high-humidity conditions [78]. Hence, the content of the hydrophobic agent inside a GDL determines its hydrophobic and hydrophilic characteristics. These characteristics further influence the water transport mechanism inside the GDL. In principle, within the GDL, as the pressure of water vapor exceeds the saturation level, water condensation occurs. The condensed water droplets are then driven from the CL to the FF by the difference between the gas and liquid phases, primarily referred to as capillary pressure. The capillary pressure is reported to be positive and negative for hydrophilic and hydrophobic GDLs, respectively [79]. However, the force subjected to water droplets inside the GDLs demonstrates significant differences depending on their hydrophobicity and hydrophilicity. Namely, water removal becomes more effective in hydrophobic GDLs; thus, both MPS and MPL are generally treated with different hydrophobic agents (e.g., PTFE, PVDF, and FEP) with various contents. The optimal PTFE content loadings in MPSs and MPLs are reported to be in the range of 10e20 and 5e35 wt%, respectively, depending on the conditions of the operation, to achieve effective water management and reactant distribution [80,81].

5. SUMMARY

719

4.3 EFFECT OF MICROSTRUCTURE MODIFICATION The literature review presented in the previous sections indicates that the deterioration in performance associated with water management and mass transport limitations is mainly influenced by the architecture of GDLs. Furthermore, it is emphasized that a delicate balance between water management and reactant distribution can be constructed through an appropriate pore size distribution within the GDL. In this context, several research groups have modified the porous structure of double-layer GDLs by introducing various pore-forming agents, including Li2CO3 and NH4Cl, into the MPL. For example, Tang et al. [82] prepared MPLs with graded porosity by introducing an NH4Cl pore former into MPLs with different contents and assessed their performance in a single PEM fuel cell. Their results indicated that manufacturing MPLs with graded porosities facilitates the transport of liquid water while allowing effective reactant distribution, because of the presence of both small pores (which allow paths for reactant distribution) and large pores (which facilitate liquid water removal). Chun et al. [83] also prepared MPLs with graded porosities by introducing NH4Cl into MPL inks with different contents and assessed their performance in a single PEM fuel cell. Their results demonstrated that MPLs with the greatest macropore and least micropore volume performed better under high-humidity conditions, whereas MPLs with the least macropore and greatest micropore volumes exhibited more promising performance under dry conditions. Thus, it can be concluded that bimodal porosity distribution (the presence of large pores together with small pores) improves both water management and reactant distribution.

4.4 CLOSURE The design parameters of MPLs discussed in the sections on the effect of the type and loading of carbon black powder, the effect of hydrophobic and hydrophilic treatment, and the effect of microstructure modification influence the structural and physical characteristics of GDLs. These modifications may also affect the performance of PEM fuel cells, as summarized in Table 2. However, it is crucial to note that GDLs demonstrate different behaviors under different operating conditions, depending on their structural (e.g., porosity and pore size distribution) and physical (e.g., electrical and thermal conductivity, gas permeability, and wettability) characteristics. Consequently, a PEM fuel cell constructed with a more hydrophobic GDL may exhibit deterioration in performance due to membrane dehydration under low-humidity conditions, but it may demonstrate more promising performance under high-humidity conditions. However, a PEM fuel cell assembled with a more hydrophilic GDL may perform better under low-humidity conditions due to its ability to retain water, whereas it can show a deterioration in performance under high-humidity conditions because of water flooding [89]. Thus, specifically to improve PEM fuel cells’ viability for large-scale commercialization, the optimization of GDLs should be undertaken while simultaneously considering various operating conditions (e.g., low-, intermediate-, and high-humidity conditions), to meet ever-changing loading demands during operation [90].

5. SUMMARY GDLs are vital components for PEM fuel cells owing to their critical functions, such as (1) allowing the distribution of reactants over the CL, (2) removing by-products from the CL, (3) offering pathways for electron transport between the CL and FF plates, (4) maintaining a delicate balance between membrane hydration and water removal, (5) providing effective heat management, and (6) supporting

720

CHAPTER 2.28 GAS DIFFUSION LAYERS FOR PEM FUEL CELLS

Table 2 Summary of Main Characteristics of Gas Diffusion Layers (GDLs), Design Parameters Affecting GDL Characteristics, and the Potential Effects of GDL Characteristics on the Performance of Proton Exchange Membrane (PEM) Fuel Cells Characteristics of GDL Gas permeability

Design Parameters Affecting GDL Characteristics

Potential Effects on PEM Fuel Cell Operation

· Material properties of raw

· Affects the transportation of gas species for water · Critical management properties · Important for limiting

·

Thermal conductivity

Electrical conductivity

Porosity

· · · · · · ·

GDL (e.g., pore size, pore shape, thickness, and compressibility) Hydrophobic agent content in both raw GDL and MPL Type and loading of carbon black powder Thickness of GDL and MPL Material properties of raw GDL Geometry of raw GDL Type and loading of carbon powder Hydrophobic agent content in MPL Thickness of GDL and MPL

· Material properties of raw GDL properties of · Adhesion MPL · Type and loading of · · · · ·

carbon black powder in MPL Hydrophobic agent content in MPL Thickness of GDL and MPL Hydrophobic agent content in MPL and GDL Average pore size of carbon black powder in MPL MPL coating technique

References [70,84,85]

current density due to its influence on mass transfer limitations

for heat · Critical management properties for oxygen · Critical reduction reaction ohmic losses due · Impacts to membrane dehydration for water · Critical condensation and

[32,36,86]

evaporation

· Critical for design, · ·

efficiency, reliability, and durability of cell and longevity of components Critical for reducing ohmic losses Affects performance, specifically at intermediate current densities

· Critical for water · ·

management and reactant and by-product transportation Affects heat management Important for ohmic losses

[19,22,31]

[70,87,88]

REFERENCES

721

Table 2 Summary of Main Characteristics of Gas Diffusion Layers (GDLs), Design Parameters Affecting GDL Characteristics, and the Potential Effects of GDL Characteristics on the Performance of Proton Exchange Membrane (PEM) Fuel Cellsdcont’d Characteristics of GDL Wettability

Design Parameters Affecting GDL Characteristics

properties of raw · Material GDL agent · Hydrophobic content in GDL and MPL and loading of · Type carbon black powder in MPL

· Arrangement of MPL

(symmetric/only on cathode/only on anode)

Potential Effects on PEM Fuel Cell Operation

for water · Crucial management for reactant and · Critical by-product distribution start-up · Impacts performance · Critical for degradation of

References [21,70,88]

components of membrane electrode assembly (e.g., membrane degradation)

MPL, microporous layer.

the mechanically weak CL. The structural (e.g., porosity) and physical (e.g., electrical and thermal conductivity, gas permeability, and wettability) characteristics of GDLs, which are determined by the design parameters of MPLs (e.g., type and loading of carbon black powder, hydrophilic and hydrophobic treatment, and microstructure modification) are of great importance for fulfilling these functions. In this context, evaluation of GDLs through accurate and comprehensive ex- and in-situ characterization techniques is necessary for their optimization. Hence, this optimization is believed to facilitate the design and manufacture of high-performance, durable and reliable PEM fuel cells. Thus, we put forward a comprehensive review of the main structural and physical characteristics of GDLs, with a specific focus on the set of tools and techniques for their ex-situ evaluation. In addition, the critical design parameters of GDLs (e.g., type and loading of carbon black powder, hydrophilic and hydrophobic treatment, and microstructure modification) have been reviewed with particular emphasis on their potential effect on the performance of PEM fuel cells.

ACKNOWLEDGMENTS This work is financially supported by OntarioeChina Research and Innovation Fund (OCRIF Round 3) and the Natural Sciences and Engineering Research Council of Canada (NSERC) via a Discovery Grant. Also, the technical support of Mr. Mustafa Ercelik in the drawing of the schematics of the experimental apparatus and characterization techniques is gratefully acknowledged.

REFERENCES [1] Ozden A, Ercelik M, Ouellette D, Colpan CO, Ganjehsarabi H, Hamdullahpur F. Designing, modeling and performance investigation of bio-inspired flow field based DMFCs. International Journal of Hydrogen Energy 2017:1e13. http://dx.doi.org/10.1016/j.ijhydene.2017.01.007.

722

CHAPTER 2.28 GAS DIFFUSION LAYERS FOR PEM FUEL CELLS

[2] Chan C, Zamel N, Li X, Shen J. Experimental measurement of effective diffusion coefficient of gas diffusion layer/microporous layer in PEM fuel cells. Electrochimica Acta 2012;65:13e21. http://dx.doi.org/10.1016/ j.electacta.2011.12.110. [3] Wang FC, Guo YF. Robustness analyses of PEMFC systems on the production line. International Journal of Hydrogen Energy 2015;40:1959e66. http://dx.doi.org/10.1016/j.ijhydene.2014.11.146. [4] Neelakandan S, Rana D, Matsuura T, Muthumeenal A, Kanagaraj P, Nagendran A. Fabrication and electrochemical properties of surface modified sulfonated poly(vinylidenefluoride-co-hexafluoropropylene) membranes for DMFC application. Solid State Ionics 2014;268:35e41. http://dx.doi.org/10.1016/ j.ssi.2014.09.027. [5] Zhao D, Zheng Q, Gao F, Bouquain D, Dou M, Miraoui A. Disturbance decoupling control of an ultra-high speed centrifugal compressor for the air management of fuel cell systems. International Journal of Hydrogen Energy 2014;39:1788e98. http://dx.doi.org/10.1016/j.ijhydene.2013.11.057. [6] Lobato J, Zamora H, Plaza J, Can˜izares P, Rodrigo MA. Enhancement of high temperature PEMFC stability using catalysts based on Pt supported on SiC based materials. Applied Catalysis B: Environmental 2016;198: 516e24. http://dx.doi.org/10.1016/j.apcatb.2016.06.011. [7] Yang D, Yan Z, Li B, Higgins DC, Wang J, Lv H, et al. Highly active and durable Pt-Co nanowire networks catalyst for the oxygen reduction reaction in PEMFCs. International Journal of Hydrogen Energy 2016;41: 18592e601. http://dx.doi.org/10.1016/j.ijhydene.2016.08.159. [8] Sa´nchez M, Pierna AR, Lorenzo A, Del Val JJ. Effect of cocatalyst and composition on catalytic performance of amorphous alloys for ethanol electrooxidation and PEMFCs. International Journal of Hydrogen Energy 2016;41:19749e55. http://dx.doi.org/10.1016/j.ijhydene.2016.06.064. [9] Thomas YRJ, Benayad A, Schroder M, Morin A, Pauchet J. New method for super hydrophobic treatment of gas diffusion layers for proton exchange membrane fuel cells using electrochemical reduction of diazonium salts. ACS Applied Materials & Interfaces 2015;7:15068e77. http://dx.doi.org/10.1021/acsami.5b04428. [10] Zamel N. Transport properties of the gas diffusion layer of PEM fuel cells [PhD thesis]. University of Waterloo; 2011. [11] Park S, Popov BN. Effect of cathode GDL characteristics on mass transport in PEM fuel cells. Fuel 2009;88: 2068e73. http://dx.doi.org/10.1016/j.fuel.2009.06.020. [12] Wu H. Mathematical modeling of transient transport phenomena in PEM fuel cells [PhD thesis]. University of Waterloo; 2009. [13] Ferreira RB, Falca˜o DS, Oliveira VB, Pinto AMFR. Experimental study on the membrane electrode assembly of a proton exchange membrane fuel cell: effects of microporous layer, membrane thickness and gas diffusion layer hydrophobic treatment. Electrochima Acta 2017;224:337e45. http://dx.doi.org/10.1016/ j.electacta.2016.12.074. [14] Kim T, Lee S, Park H. A study of water transport as a function of the micro-porous layer arrangement in PEMFCs. International Journal of Hydrogen Energy 2010;35:8631e43. http://dx.doi.org/10.1016/ j.ijhydene.2010.05.123. [15] Zamel N, Li X. Effective transport properties for polymer electrolyte membrane fuel cells e with a focus on the gas diffusion layer. Progress in Energy and Combustion Science 2013;39:111e46. http://dx.doi.org/ 10.1016/j.pecs.2012.07.002. [16] Park S, Lee JW, Popov BN. A review of gas diffusion layer in PEM fuel cells: materials and designs. International Journal of Hydrogen Energy 2012;37:5850e65. http://dx.doi.org/10.1016/ j.ijhydene.2011.12.148. [17] Cindrella L, Kannan AM, Lin JF, Saminathan K, Ho Y, Lin CW, Wertz J. Gas diffusion layer for proton exchange membrane fuel cells-A review. Journal of Power Sources 2009;194:146e60. http://dx.doi.org/ 10.1016/j.jpowsour.2009.04.005.

REFERENCES

723

[18] He G, Yamazaki Y, Abudula A. A three-dimensional analysis of the effect of anisotropic gas diffusion layer(GDL) thermal conductivity on the heat transfer and two-phase behavior in a proton exchange membrane fuel cell(PEMFC). Journal of Power Sources 2010;195:1551e60. http://dx.doi.org/10.1016/ j.jpowsour.2009.09.059. [19] Vikram A, Chowdhury PR, Phillips RK, Hoorfar M. Measurement of effective bulk and contact resistance of gas diffusion layer under inhomogeneous compression - Part I: electrical conductivity. Journal of Power Sources 2016;320:274e85. http://dx.doi.org/10.1016/j.jpowsour.2016.04.110. [20] Ge N, Chevalier S, Lee J, Yip R, Banerjee R, George MG, Liu H, Lee C, Fazeli M, Antonacci P, Kotaka T, Tabuchi Y, Bazylak A. Non-isothermal two-phase transport in a polymer electrolyte membrane fuel cell with crack-free microporous layers. International Journal of Heat and Mass Transfer 2017;107:418e31. http:// dx.doi.org/10.1016/j.ijheatmasstransfer.2016.11.045. [21] Chun JH, Park KT, Jo DH, Kim SG, Kim SH. Numerical modeling and experimental study of the influence of GDL properties on performance in a PEMFC. International Journal of Hydrogen Energy 2011;36:1837e45. http://dx.doi.org/10.1016/j.ijhydene.2010.01.036. [22] Li H, Tang Y, Wang Z, Shi Z, Wu S, Song D, Zhang J, Fatih K, Zhang J, Wang H, Liu Z, Abouatallah R, Mazza A. A review of water flooding issues in the proton exchange membrane fuel cell. Journal of Power Sources 2008;178:103e17. http://dx.doi.org/10.1016/j.jpowsour.2007.12.068. [23] Pasaogullari U, Wang CY. Two-phase transport and the role of micro-porous layer in polymer electrolyte fuel cells. Electrochimica Acta 2004;49:4359e69. http://dx.doi.org/10.1016/j.electacta.2004.04.027. [24] Tseng C-J, Lo S-K. Effects of microstructure characteristics of gas diffusion layer and microporous layer on the performance of PEMFC. Energy Conversion and Management 2010;51:677e84. http://dx.doi.org/ 10.1016/j.enconman.2009.11.011. [25] Xu JB, Zhao TS. Mesoporous carbon with uniquely combined electrochemical and mass transport characteristics for polymer electrolyte membrane fuel cells. RSC Advances 2013;3:16. http://dx.doi.org/10.1039/ c2ra22279e. [26] Mason TJ, Millichamp J, Neville TP, El-Kharouf A, Pollet BG, Brett DJL. Effect of clamping pressure on ohmic resistance and compression of gas diffusion layers for polymer electrolyte fuel cells. Journal of Power Sources 2012;219:52e9. http://dx.doi.org/10.1016/j.jpowsour.2012.07.021. [27] Netwall CJ, Gould BD, Rodgers JA, Nasello NJ, Swider-Lyons KE. Decreasing contact resistance in protonexchange membrane fuel cells with metal bipolar plates. Journal of Power Sources 2013;227:137e44. http:// dx.doi.org/10.1016/j.jpowsour.2012.11.012. [28] Nam J, Chippar P, Kim W, Ju H. Numerical analysis of gas crossover effects in polymer electrolyte fuel cells (PEFCs). Applied Energy 2010;87:3699e709. http://dx.doi.org/10.1016/j.apenergy.2010.05.023. [29] Perng SW, Wu HW. Non-isothermal transport phenomenon and cell performance of a cathodic PEM fuel cell with a baffle plate in a tapered channel. Applied Energy 2011;88:52e67. http://dx.doi.org/10.1016/ j.apenergy.2010.07.006. [30] Perng SW, Wu HW. Effect of the prominent catalyst layer surface on reactant gas transport and cell performance at the cathodic side of a PEMFC. Applied Energy 2010;87:1386e99. http://dx.doi.org/10.1016/ j.apenergy.2009.08.006. [31] Ismail MS, Hughes KJ, Ingham DB, Ma L, Pourkashanian M. Effects of anisotropic permeability and electrical conductivity of gas diffusion layers on the performance of proton exchange membrane fuel cells. Applied Energy 2012;95:50e63. http://dx.doi.org/10.1016/j.apenergy.2012.02.003. [32] Ismail MS, Damjanovic T, Ingham DB, Pourkashanian M, Westwood A. Effect of polytetrafluoroethylenetreatment and microporous layer-coating on the electrical conductivity of gas diffusion layers used in proton exchange membrane fuel cells. Journal of Power Sources 2010;195:2700e8. http://dx.doi.org/10.1016/ j.jpowsour.2009.11.069.

724

CHAPTER 2.28 GAS DIFFUSION LAYERS FOR PEM FUEL CELLS

[33] Ismail MS, Borman D, Damjanovic T, Ingham DB, Pourkashanian M. On the through-plane permeability of microporous layer-coated gas diffusion layers used in proton exchange membrane fuel cells. International Journal of Hydrogen Energy 2011;36:10392e402. http://dx.doi.org/10.1016/j.ijhydene.2010.09.012. [34] Arvay A, Yli-Rantala E, Liu CH, Peng XH, Koski P, Cindrella L, Kauranen P, Wilde PM, Kannan AM. Characterization techniques for gas diffusion layers for proton exchange membrane fuel cells - a review. Journal of Power Sources 2012;213:317e37. http://dx.doi.org/10.1016/j.jpowsour.2012.04.026. [35] Sadeghi E, Djilali N, Bahrami M. Effective thermal conductivity and thermal contact resistance of gas diffusion layers in proton exchange membrane fuel cells. Part 1: effect of compressive load. Journal of Power Sources 2011;196:246e54. http://dx.doi.org/10.1016/j.jpowsour.2010.06.039. [36] Sadeghi E, Djilali N, Bahrami M. A novel approach to determine the in-plane thermal conductivity of gas diffusion layers in proton exchange membrane fuel cells. Journal of Power Sources 2011;196:3565e71. http://dx.doi.org/10.1016/j.jpowsour.2010.11.151. [37] Ramousse J, Didierjean S, Lottin O, Maillet D. Estimation of the effective thermal conductivity of carbon felts used as PEMFC gas diffusion layers. International Journal of Thermal Sciences 2008;47:1e6. http:// dx.doi.org/10.1016/j.ijthermalsci.2007.01.018. [38] Sui PC, Djilali N. Analysis of coupled electron and mass transport in the gas diffusion layer of a PEM fuel cell. Journal of Power Sources 2006;161:294e300. http://dx.doi.org/10.1016/j.jpowsour.2006.03.079. [39] Ahmed DH, Sung HJ, Bae J. Effect of GDL permeability on water and thermal management in PEMFCs-II. Clamping force. International Journal of Hydrogen Energy 2008;33:3786e800. http://dx.doi.org/10.1016/ j.ijhydene.2008.04.023. [40] Parry V, Appert E, Joud JC. Characterisation of wettability in gas diffusion layer in proton exchange membrane fuel cells. Applied Surface Science 2010;256:2474e8. http://dx.doi.org/10.1016/ j.apsusc.2009.10.091. [41] Antonacci P, Chevalier S, Lee J, Ge N, Hinebaugh J, Yip R, Tabuchi Y, Kotaka T, Bazylak A. Balancing mass transport resistance and membrane resistance when tailoring microporous layer thickness for polymer electrolyte membrane fuel cells operating at high current densities. Electrochimica Acta 2016;188:888e97. http://dx.doi.org/10.1016/j.electacta.2015.11.115. [42] Jiao K, Li X. Effect of surface dynamic wettability in proton exchange membrane fuel cells. Internation Journal of Hydrogen Energy 2010;35:9095e103. http://dx.doi.org/10.1016/j.ijhydene.2010.05.027. [43] Das PK, Li X, Liu Z-S. Analysis of liquid water transport in cathode catalyst layer of PEM fuel cells. International Journal of Hydrogen Energy 2010;35:2403e16. http://dx.doi.org/10.1016/ j.ijhydene.2009.12.160. [44] Marmur A, Bittoun E. When wenzel and cassie are right: reconciling local and global considerations. Langmuir 2009;25:1277e81. http://dx.doi.org/10.1021/la802667b. [45] Wang Y, Al Shakhshir S, Li X. Development and impact of sandwich wettability structure for gas distribution media on PEM fuel cell performance. Applied Energy 2011;88:2168e75. http://dx.doi.org/10.1016/ j.apenergy.2010.12.054. [46] Furmidge CGL. Studies at phase interfaces. I. The sliding of liquid drops on solid surfaces and a theory for spray retention. Journal of Colloid Science 1962;17:309e24. http://dx.doi.org/10.1016/0095-8522(62) 90011-9. [47] Guo ZG, Liu WM. Sticky superhydrophobic surface. Applied Physics Letters 2007;90. http://dx.doi.org/ 10.1063/1.2745251. [48] Gurau V, Bluemle MJ, De Castro ES, Tsou Y-M, Mann Jr JA, Zawodzinski Jr TA. Characterization of transport properties in gas diffusion layers for proton exchange membrane fuel cells: 1. Wettability (internal contact angle to water and surface energy of GDL fibers). Journal of Power Sources 2006;160:1156e62. http://dx.doi.org/10.1016/j.jpowsour.2006.03.016.

REFERENCES

725

[49] Gostick JT, Ioannidis MA, Fowler MW, Pritzker MD. Direct measurement of the capillary pressure characteristics of water-air-gas diffusion layer systems for PEM fuel cells. Electrochemistry Communications 2008;10:1520e3. http://dx.doi.org/10.1016/j.elecom.2008.08.008. [50] Harkness IR, Hussain N, Smith L, Sharman JDB. The use of a novel water porosimeter to predict the water handling behaviour of gas diffusion media used in polymer electrolyte fuel cells. Journal of Power Sources 2009;193:122e9. http://dx.doi.org/10.1016/j.jpowsour.2008.11.055. [51] Mathur RB, Maheshwari PH, Dhami TL, Tandon RP. Characteristics of the carbon paper heat-treated to different temperatures and its influence on the performance of PEM fuel cell. Electrochimica Acta 2007;52: 4809e17. http://dx.doi.org/10.1016/j.electacta.2007.01.041. [52] Volfkovich YM, Bagotzky VS. The method of standard porosimetry. 1. Principles and possibilities. Journal of Power Sources 1994;48:327e38. http://dx.doi.org/10.1016/0378-7753(94)80029-4. [53] Unsworth G. Measurement and characterization of heat and mass diffusion in PEMFC porous media [M.A.Sc. thesis]. 2012. [54] Senthil Velan V, Velayutham G, Rajalakshmi N, Dhathathreyan KS. Influence of compressive stress on the pore structure of carbon cloth based gas diffusion layer investigated by capillary flow porometry. International Journal of Hydrogen Energy 2014;39:1752e9. http://dx.doi.org/10.1016/j.ijhydene.2013.11.038. [55] Fairweather JD, Cheung P, St-Pierre J, Schwartz DT. A microfluidic approach for measuring capillary pressure in PEMFC gas diffusion layers. Electrochemistry Communications 2007;9:2340e5. http:// dx.doi.org/10.1016/j.elecom.2007.06.042. [56] Litster S, Sinton D, Djilali N. Ex situ visualization of liquid water transport in PEM fuel cell gas diffusion layers. Journal of Power Sources 2006;154:95e105. http://dx.doi.org/10.1016/j.jpowsour.2005.03.199. [57] Williams MV, Begg E, Bonville L, Kunz HR, Fenton JM. Characterization of gas diffusion layers for PEMFC. Journal of The Electrochemical Society 2004;151:A1173e80. http://dx.doi.org/10.1149/ 1.1764779. [58] Jena A, Gupta K. Advances in pore structure evaluation by porometry. Chemical Engineering & Technology 2010;33:1241e50. http://dx.doi.org/10.1002/ceat.201000119. [59] Reum M. Sub-millimeter Resolved measurement of current density and membrane resistance in polymer Electrolyte fuel cells (PEFC). 2008. Diss. No. 17979. [60] Mishra V, Yang F, Pitchumani R. Measurement and prediction of electrical contact resistance between gas diffusion layers and bipolar plate for applications to PEM fuel cells. Journal of Fuel Cell and Technology 2004;1:2. http://dx.doi.org/10.1115/1.1782917. [61] Barbir F. PEM fuel cells: theory and practice. 2nd ed. Waltham, MA: Elsevier; 2012. [62] Wilde PM, Ma¨ndle M, Murata M, Berg N. Structural and physical properties of GDL and GDL/BPP combinations and their influence on PEMFC performance. Fuel Cells 2004;4:180e4. http://dx.doi.org/ 10.1002/fuce.200400022. [63] Smits FM. Measurement of sheet resistivities with the four-point probe. Bell System Technology Journal 1958;37:711e8. http://dx.doi.org/10.1002/j.1538-7305.1958.tb03883.x. [64] Khandelwal M, Mench MM. Direct measurement of through-plane thermal conductivity and contact resistance in fuel cell materials. Journal of Power Sources 2006;161:1106e15. http://dx.doi.org/10.1016/ j.jpowsour.2006.06.092. [65] Nitta I, Himanen O, Mikkola M. Thermal conductivity and contact resistance of compressed gas diffusion layer of PEM fuel cell. Fuel Cells 2008;8:111e9. http://dx.doi.org/10.1002/fuce.200700054. [66] Bahrami M, Yovanovich MM, Culham JR. Thermal contact resistance: effect of elastic deformation. Semiconductor Thermal Measurement and Management Symposium 2005:44e52. http://dx.doi.org/ 10.1109/STHERM.2005.1412158. [67] Teertstra P, Karimi G, Li X. Measurement of in-plane effective thermal conductivity in PEM fuel cell diffusion media. Electrochimica Acta 2011;56:1670e5. http://dx.doi.org/10.1016/j.electacta.2010.06.043.

726

CHAPTER 2.28 GAS DIFFUSION LAYERS FOR PEM FUEL CELLS

[68] Aaron K, Tang X, Tritt TM. Measurement of the in-plane thermal conductivity of single crystals by the parallel thermal conductance technique. International Conference on Thermoelectrics ICT, Proc 2005;2005: 87e90. http://dx.doi.org/10.1109/ICT.2005.1519894. [69] Gostick JT, Fowler MW, Pritzker MD, Ioannidis MA, Behra LM. In-plane and through-plane gas permeability of carbon fiber electrode backing layers. Journal of Power Sources 2006;162:228e38. http:// dx.doi.org/10.1016/j.jpowsour.2006.06.096. [70] Tamayol A, McGregor F, Bahrami M. Single phase through-plane permeability of carbon paper gas diffusion layers. Journal of Power Sources 2012;204:94e9. http://dx.doi.org/10.1016/j.jpowsour.2011.11.084. [71] Nitta I, Hottinen T, Himanen O, Mikkola M. Inhomogeneous compression of PEMFC gas diffusion layer. Part I. Experimental. Journal of Power Sources 2007;171:26e36. http://dx.doi.org/10.1016/ j.jpowsour.2006.11.018. [72] Christoph R, Siegenthaler H. In situ. Situ 2010;33:433e46. http://dx.doi.org/10.1038/nn.2294. [73] Orogbemi OM, Ingham DB, Ismail MS, Hughes KJ, Ma L, Pourkashanian M. The effects of the composition of microporous layers on the permeability of gas diffusion layers used in polymer electrolyte fuel cells. International Journal of Hydrogen Energy 2016;41:21345e51. http://dx.doi.org/10.1016/ j.ijhydene.2016.09.160. [74] Chen HH, Chang MH. Effect of cathode microporous layer composition on proton exchange membrane fuel cell performance under different air inlet relative humidity. Journal of Power Sources 2013;232:306e9. http://dx.doi.org/10.1016/j.jpowsour.2013.01.079. [75] Oono Y, Sounai A, Hori M. Influence of the phosphoric acid-doping level in a polybenzimidazole membrane on the cell performance of high-temperature proton exchange membrane fuel cells. Journal of Power Sources 2009;189:943e9. http://dx.doi.org/10.1016/j.jpowsour.2008.12.115. [76] Passalacqua E, Squadrito G, Lufrano F, Patti A, Giorgi L. Effects of the diffusion layer characteristics on the performance of polymer electrolyte fuel cell electrodes. Journal of Applied Electrochemistry 2001;31: 449e54. http://dx.doi.org/10.1023/A:1017547112282. [77] Wu R, Zhu X, Liao Q, Wang H, Ding YD, Li J, Ye DD. A pore network study on water distribution in bi-layer gas diffusion media: effects of inlet boundary condition and micro-porous layer properties. International Journal of Hydrogen Energy 2010;35:9134e43. http://dx.doi.org/10.1016/j.ijhydene.2010.06.051. [78] Chang HM, Lin CW, Chang MH, Shiu HR, Chang WC, Tsau FH. Optimization of polytetrafluoroethylene content in cathode gas diffusion layer by the evaluation of compression effect on the performance of a proton exchange membrane fuel cell. Journal of Power Sources 2011;196:3773e80. http://dx.doi.org/10.1016/ j.jpowsour.2010.12.090. [79] Pasaogullari U, Wang CY. Liquid water transport in gas diffusion layer of polymer electrolyte fuel cells. Journal of The Electrochemical Society 2004;151:399e406. http://dx.doi.org/10.1149/1.1646148. [80] Su H. The effect of gas diffusion layer PTFE content on the performance of high temperature proton exchange membrane fuel cell. International Journal of Electrochemical Science 2016;11:2919e26. http:// dx.doi.org/10.20964/110402919. [81] Van Nguyen T, Ahosseini A, Wang X, Yarlagadda V, Kwong A, Weber AZ, Deevanhxay P, Tsushima S, Hirai S. Hydrophobic gas-diffusion media for polymer-electrolyte fuel cells by direct fluorination. Journal of The Electrochemical Society 2015;162:F1451e60. http://dx.doi.org/10.1149/2.0411514jes. [82] Tang H, Wang S, Pan M, Yuan R. Porosity-graded micro-porous layers for polymer electrolyte membrane fuel cells. Journal of Power Sources 2007;166:41e6. http://dx.doi.org/10.1016/j.jpowsour.2007.01.021. [83] Chun JH, Park KT, Jo DH, Lee JY, Kim SG, Lee ES, Jyoung J, Kim SH. Determination of the pore size distribution of micro porous layer in PEMFC using pore forming agents under various drying conditions. International Journal of Hydrogen Energy 2010;35:11148e53. http://dx.doi.org/10.1016/ j.ijhydene.2010.07.056.

REFERENCES

727

[84] Tamayol A, Bahrami M. In-plane gas permeability of proton exchange membrane fuel cell gas diffusion layers. Journal of Power Sources 2011;196:3559e64. http://dx.doi.org/10.1016/j.jpowsour.2010.11.109. [85] Morgan JM, Datta R. Understanding the gas diffusion layer in proton exchange membrane fuel cells. I. How its structural characteristics affect diffusion and performance. Journal of Power Sources 2014;251:269e78. http://dx.doi.org/10.1016/j.jpowsour.2013.09.090. [86] Alhazmi N, Ingham DB, Ismail MS, Hughes KJ, Ma L, Pourkashanian M. Effect of the anisotropic thermal conductivity of GDL on the performance of PEM fuel cells. International Journal of Hydrogen Energy 2013; 38:603e11. http://dx.doi.org/10.1016/j.ijhydene.2012.07.007. [87] Fishman Z, Bazylak A. Heterogeneous through-plane porosity distributions for treated PEMFC GDLs. ii. Effect of MPL cracks. Journal of The Electrochemical Society 2011;158:B846e51. http://dx.doi.org/ 10.1149/1.3594636. [88] Kandlikar SG, Garofalo ML, Lu Z. Water management in a PEMFC: water transport mechanism and material degradation in gas diffusion layers. Fuel Cells 2011;11:814e23. http://dx.doi.org/10.1002/ fuce.201000172. [89] Kitahara T, Konomi T, Nakajima H. Microporous layer coated gas diffusion layers for enhanced performance of polymer electrolyte fuel cells. Journal of Power Sources 2010;195:2202e11. http://dx.doi.org/10.1016/ j.jpowsour.2009.10.089. [90] Oh H, Park J, Min K, Lee E, Jyoung JY. Effects of pore size gradient in the substrate of a gas diffusion layer on the performance of a proton exchange membrane fuel cell. Applied Energy 2015;149:186e93. http:// dx.doi.org/10.1016/j.apenergy.2015.03.072.