Direct methanol fuel cells (DMFCs) have been attracting extensive attention for many years as alternative energy devices. DMFCs are promising for use in portable electronic devices and field equipment, as well as light electric vehicles, because of their high fuel energy density, simplicity, rapid start-up, easy fuel recharging and low pollution. California-based Oorja Protonics launched its “ultra-powerful” DMFC technology in 2008. This company has also unveiled a variety of commercial products, such as Oorja Model 3 for the materials handling market and Oorja Model T-1 for stationary applications. Danish company IRD also has developed 500-W and 800-W DMFCs for use in lighthouses and telecommunication towers.
The maximum thermodynamic voltage of a DMFC at 25 °C is 1.18 V, as defined by its electrocatalytic redox reactions [1]:
Anode reaction:
Cathode reaction:
Overall reaction:
However, because of the different types of voltage loss caused by the complexity in DMFCs, the actual cell voltage is lower than the theoretical value. In particular, methanol crossover (MCO) from the anode to the cathode causes a mixed potential in the cathode, making the overall cell voltage much lower than that of hydrogen fuel cells. Although considerable improvement has been realized,DMFCs are still far from widespread commercial application. At present, electrocatalytic materials are a major obstacle limiting DMFC technology. Conventional Pt catalysts suffer from several issues including catalyst deactivation, low catalytic efficiency and high cost. Thus, current research focuses on the development of inexpensive catalysts that display high stability, durability, and performance.
Besides the catalysts,DMFCs also face issues related with the microstructure and electrochemical characteristics of the membrane electrode assembly (MEA), where the chemical energy is converted to electrical energy by electrocatalytic redox reactions. MEAs are typically composed of an anode diffusion layer (DL), an anode catalyst layer (CL), a proton exchange membrane (PEM), a cathode CL, and a cathode DL. MEAs provide a complex conductive environment for electron/proton conduction, reactant/product transportation and interfacial redox reactions. Normally, methanol diffuses into the anode DL from the anode flow field, and reaches the anode CL, where it takes part in the methanol oxidation reaction (MOR). Excess methanol will be transported to the cathode by osmotic pressure and electromigration across the PEM, and then reacts with oxygen (O2) as soon as it arrives at the cathode. Although the exact kinetics of this process are the subject of debat, the end result is a decrease of the cell voltage. At the cathode,O2 from the surrounding air is transferred through the cathode DL to the cathode CL, where it is reduced to water. At the same time, free electrons from the anode reactions flow through the external circuit to the cathode, which allows the reaction at the cathode to continue. Before the electrons reach the current collector, they must be able to flow through the cathode CL and DL. By optimizing every layer in a MEA, limitations such as transport and management issues and MCO can be overcome, and conduction and distribution can be enhanced [2].
The catalytic efficiency can be enhanced by optimizing the structure or composition of electrocatalytic materials. For instance, numerous studies have shown that constructing an appropriate porous network or ordered nanoparticle array in the CL can lead to enhanced catalytic performance and improved Pt utilization because of an increase of electrochemically active surface area (EASA) [3]. Regarding composition modification, many researchers have focused on introducing other metals or metal oxides into Pt catalysts to modify their interface electronic state [4, 5]. During such modification, the ligand effect can influence the poison tolerance of Pt by weakening the interaction between Pt and some adsorbates such as protons (H+), hydroxide ions (OH-) or adsorbed carbon monoxide (COads), and thus improve catalytic efficiency. For example, it is well known that Pt-Ru bimetallic catalysts can activate H2O at lower potential than Pt, the adsorbed OH- species on Ru can oxidize the COads on the surface of the Pt atoms to CO2 and thus increase the anti-poisoning ability of Pt; this process is usually called the bifunctional mechanism [6].
The three-dimensional (3D) network structure is considered suitable for constructing high-efficiency CLs because it facilitates internal transport by providing continuous passages and decreasing diffusion distances. The porous structure composed interspaces between separated branches can also provide sufficiently accessible active sites at interfaces to allow electrocatalytic reactions. The gaps between nanoparticles have been proved effective at preventing grain rearrangement and growth, improving the working life and durability of fuel cells [7]. Among various catalysts,Pt-based catalysts with advanced nanostructures are of particular interest for use in DMFCs.
As an example of a CL with a 3D network structure, Yamauchi et al. [8] used a soft template method with lyotropic liquid crystals (LLCs) to form highly ordered mesoporous materials with lamellar, two-dimensional hexagonal, 3D cubic, and other nanostructures. They also developed a very simple one-step method to prepare Pt nanodendrites at room temperature within 5 min using block copolymer as a structure-directing agent [9]. Transmission electron microscopy (TEM) images of these Pt nanodendrites are shown in Fig. 1. The block polymer with amphiphilic nature mediated the growth of branched nanostructures through physical and chemical interactions with Pt precursors. The Pt nanodendrites displayed not only a large active Pt surface area, but also favorable electrochemical performance. The mass-normalized current densities for Pt nanodendrites and commercially available Pt black were 0.52 and 0.11 A/mg, respectively, in the positive direction sweep. The current density of the Pt nanodendrites was 4.7 times higher than that of Pt black.
Recently,Fan et al. [10] reported a one-pot hydrothermal synthesis of porous Pt-Cu alloy networks using the block copolymer Pluronic-F127 as a structure-directing agent. A scanning electron microscopy (SEM) image of a Pt53Cu47 network is provided in Fig. 2. The as-prepared Pt-Cu alloy exhibited a unique 3D hierarchical porous network structure consisting of numerous interconnected nanodendrites. Remarkably, this network structure mostly presented in the form of monodisperse metal nanoparticles. In this synthesis, ethylene glycol served as a weak reducing agent to allow slow growth of Pt-Cu alloy nanostructures. The block copolymer was used as amphiphilic molecules to adjust the structure formation, making it important in the formation of porous Pt-Cu alloy nanostructures. In addition, the reduction kinetics substantially affected the nucleation, growth and resulting structure of the Pt-Cu alloys. Because of its porous network structure composed of unique nanodendritic building blocks, the Pt53Cu47 alloy catalyst exhibited superior electrocatalytic performance toward the MOR compared with that of Pt73Cu27 and Pt black (Fig. 3).
Similarly, various nanostructured Pt-Pd catalysts have been studied to obtain higher surface area and lower mass transfer resistance [11]. Among such catalysts, those with hollow porous structures have attracted special interest. Recently,Qiu and colleagues [12] developed a general electrostatic attraction-directed layer-by-layer assembly approach to construct novel Pt-Pd alloy hollow porous nanospheres (HPNSs) using SiO2 nanospheres as templates (Fig. 4). They found that the strong electrostatic attraction between charged species during the synthetic process played important roles in the homogeneous, compact deposition of alloy nanoparticles on the surface of the SiO2 templates and subsequent formation of alloy HPNSs. Because of their unique compositional and structural features, the optimized Pt-Pd HPNSs displayed markedly enhanced electrocatalytic performance in the MOR in terms of catalytic activity and durability compared with that of both commercial Pt black and Pd black catalysts. The peak polarization current of Pt-Pd HPNSs was 198.7 mA/mg, which is much higher than those of Pd black (45.7 mA/mg) and Pt black (120.3 mA/mg). The promising electrocatalytic performance of the Pt-Pd HPNSs makes them a suitable anodic catalyst for DMFCs.
Besides these template techniques, dealloying is another attractive method to prepare nanostructured catalysts. In dealloying, an additional metal is selectively dissolved from an alloy precursor to give a bimetallic catalyst with a self-supporting porous nanostructure [13]. Sun et al. [14] reported a novel ultrafine nanoporous Pt-Cu alloy with a Pt:Cu stoichiometric ratio of 3:1 (denoted np-Pt3Cu1), which was prepared by mechanical alloying and subsequent two-step chemical dealloying (Fig. 5). The specific activity of the np-Pt3Cu alloy was much higher than that of a reference Pt/C catalyst. The peak current density of the np-Pt3Cu alloy in the positive scan in cyclic voltammetry (CV) measurements in 0.5 mol/L sulfuric acid and 0.5 mol/L methanol solution at 25 °C was 1.38 mA/cm2, which was about 3.5 times that of Pt/C (0.39 mA/cm2).
These Pt-based catalysts with 3D network nanostructure generally exhibited improved reaction kinetics and catalytic efficiency compared with those of reference materials. This is because the 3D networks provided plentiful active sites and led to reinforced structural stability and extended working durability originating from their enhanced resistance to Ostwald ripening, dissolution and aggregation. Complex components seem to be promising to increase the catalytic activity of materials, and the use of nanostructures is becoming a new trend to overcome the major limitations of DMFCs [15].
Numerous researchers have used carbon supporting materials in their studies. The results have generally indicated that the introduction of nanostructured carbon materials into CLs can lower the reaction charge transfer resistance, improve mass transfer efficiency and decrease the usage of noble metal catalysts [16, 17]. Recently, hollow carbon materials with a core-shell structure have attracted extensive attention for use in CLs [18, 19]. For example, a periodically ordered bimodal porous carbon (POBPC) framework with a hollow core-mesoporous shell structure designed to support a Pt-Ru catalyst exhibited much higher specific activity in the MOR than the counterpart with commercial E-TEK [20]. This result was attributed to the high EASA and well-developed interconnected macroscale and mesoscale bimodal porosity of the POBPC framework, which allowed facile diffusion of reactants and products to and from the CL. Related reports had widely assumed that in MEAs, the mesopores and micropores act as mass transfer passages and the main reaction space for the MOR and oxygen reduction reaction, and the other macropores act as a pathway into the interior reaction sites.
To further improve the catalytic efficiency of nanostructured carbon-supported catalysts, many studies have focused on introducing heteroatoms to overcome the adverse effects caused by the relatively inert surface state of carbon [21, 22]. Among these advanced materials, various N-doped carbon supports have received extensive attention. Many researchers attributed the enhancement in catalytic efficiency caused by N doping to the resulting change of surface electron donor/acceptor characteristics, and thus suggested that N-doped nanostructured carbon materials could be ideal catalyst-supporting materials for DMFCs [23, 24]. Zhang et al. [25] successfully prepared hierarchical N-doped porous hollow carbon spheres (HNPHCSs) through in situ oxidation polymerization using polyaniline (PANI) as the precursor (Fig. 6). The physical and chemical characterization results indicated that the HNPHCSs have an interconnected bimodal pore system with a thin mesoporous shell and hollow macroporous core, which is potentially important for use as a supporting material.
Conductive polymer supporting materials can also help to decrease the content of COads on the surface of Pt particles and improve catalytic efficiency. Conductive polymers such as PANI have excellent electrochemical properties in the MOR, so many researchers have paid them great attention for a long time [26, 27]. Hybrid structures containing PANI and its derivatives are currently attracting increasing interest because of their superior performance to PANI alone [28]. For example,Shah and co-workers [29] coated a PANI/poly-ortho-aminophenol (POAP) bilayer structure on Pt electrodes for use in the MOR in DMFCs by electropolymerization. CV measurements revealed that the maximum anodic current of PANI/POAP deposited on a Pt electrode was 1.3 times higher than that of a reference POAP-Pt electrode. Furthermore, electrochemical impedance spectroscopy (EIS) showed that with increasing upper potential, the conductivity and capacitance of the PANI/POAP-Pt electrodes increased because of the desorption and oxidation of COads to CO2. Overall, conductive polymers with hybrid structures containing PANI and its derivatives can be considered as good supporting materials for CLs in DMFCs.
Pt-based catalysts are nearly irreplaceable for methanol electrocatalytic oxidation at normal temperature in an air atmosphere. However, a self-poisoning process on the Pt surface caused by reaction intermediates such as COads, which are formed during stepwise dehydrogenation of methanol, the catalytic properties of Pt are unstable. Metal oxides can change the surface electronic state of Pt particles and provide reactive oxygen species to react with COads, suppressing catalytic poisoning of Pt. Research has also indicated that the coexistence of multiple valence states in metal oxides is beneficial to electron transport during redox reactions [30].
As an example of how a metal oxide can affect the properties of Pt, addition of CeO2 can enhance the catalytic activity and stability of Pt-based catalysts by increasing the oxygen content on the catalyst surface [31]. Recently,Wang et al. [32] proposed a 3D Pt-rod-shaped CeO2/graphene sheet (Pt-NRCeO2/GN) catalyst to improve the electrocatalytic properties of Pt in the MOR. They synthesized rod-shaped CeO2 on graphene sheets using an in situ growth method. TEM images of Pt-NRCeO2/GN are shown in Fig. 7. The rod-shaped CeO2, which acted as a spacer to prevent graphene stacking, exposed a large amount of Pt particles in the effective reaction region to realize performance improvement. In electrochemical tests, the Pt-NRCeO2/GN catalyst showed a higher ESA value (72.6 m2/g) and larger catalytic current density (498 mA/mg) for methanol oxidation than those of commercial catalysts.
SnO2 is also of particular interest to prevent Pt catalyst poisoning because of its highly efficient COads oxidation caused by facile oxygen species transmission [33]. Pt-SnO2 catalysts on different nanostructured carbon supports exhibited superior electrochemical performance compared with that of commercial Pt/C catalyst [34, 35]. Recently, the catalytic activity of Pt-SnO2 catalysts doped with a third element in the MOR has been investigated. The indium-doped Pt-SnO2 catalysts developed by Feng et al. [36] exhibited almost twice the current density of undoped samples under the same experiment conditions. The same group also prepared Mo-doped SnO2-Pt catalysts for use in the MOR. Compared with the binding energies of the Pt/C counterpart, all the binding energies of Pt in the reported (MoO3)mSnO2-Pt/C catalysts were shifted to higher values. Some researchers proposed that these enhancements originated from the downshift of the d-band center of Pt, which led to a weakened interaction between Pt and the adsorbed reaction intermediates [37].
Among various metal oxides,TiO2 has received the most interest because of its large catalysis promoting effect and high working durability in DMFC environments. As an effective method to improve the electronic conductivity and EASA of Pt catalysts, a hybrid supporting material for Pt-based catalysts consisting of TiO2 and nanostructured carbon was developed [38, 39]. Characterization of the supported catalyst revealed that the support could efficiently improve the electrocatalytic performance of Pt in the MOR.
To obtain Pt/TiO2 catalysts with specific morphology and advanced performance, various methods to prepare Pt/TiO2 materials, such as electrospinning, electric deposition, coating pyrolysis, physical or chemical vapor deposition, and magnetron sputtering, have been investigated [40, 41]. For instance, a Pt/TiO2 nanofiber (TiO2NF) catalyst prepared by electrospinning was found to display highly effective catalysis under DMFC conditions [42]. This is because the surface and pore structure of TiO2NF led to facile transportation of reactants and products. In addition, the deactivation caused by COads was generally well suppressed by the Pt/TiO2NF catalyst. The assurance of sufficient oxygen supply from the TiO2 support has been attributed as the cause of the improvement of catalytic activity. However, to avoid the high electronic resistance and superfluous coverage of OH- caused by excess TiO2, the Pt/TiO2 ratio in Pt/TiO2NF catalysts should be carefully controlled. Zheng et al. [43] found that the maximum current densities in scans in the positive direction of the MOR in CV measurements decreased in the order 21% Pt/TiO2NFs ＞ 27% Pt/TiO2NFs ＞ 16% Pt/TiO2NFs ＞ 10% Pt/TiO2NFs. Therefore, the 21% Pt/TiO2NFs displayed the highest activity for the MOR with a peak current density of 1.781 mA/cm2, which was about 1.6 times that of commercial Pt/C (1.082 mA/cm2).
To further improve catalytic performance, researchers combined TiO2 substrates with carbon materials to enhance the electrical conductivity of Pt-based catalysts. These Pt/TiO2-C catalysts synthesized by chemical vapor deposition, a microwave-assisted polyol process and photodeposition showed excellent electrochemical properties when tested under DMFC conditions [44, 45]. For example,Fan's group [46] prepared Pt/TiO2-C catalysts by reducing H2PtCl6 with NaBH4 on the TiO2-C particles. This Pt/TiO2-C catalyst exhibited higher catalytic efficiency than commercial Pt/C catalyst in the MOR. Meanwhile,Yue et al. [47] proposed a Pt-TiO2/C electrode containing TiO2 nanotubes (TiO2NTs), which were prepared by pyrolysis, and investigated its electrochemical performance in a DMFC. During the pyrolysis process,H2PtCl6 and glucose decomposed simultaneously to achieve an ideal dispersion effect and particle configuration. SEM images of the as-prepared Pt/TiO2NT-C electrode are shown in Fig. 8. This electrode exhibited superior catalytic efficiency and working stability compared with a Pt-TiO2NT electrode. The observed enhancement in the catalytic performance of the Pt/TiO2NT-C electrode was attributed to the smaller particle size, more homogeneous distribution, better dispersion and stronger interaction of interface electrons between Pt nanoparticles and the TiO2NT-C support than in the Pt-TiO2NT electrode.
For several years, researchers have focused on achieving higher activity of Pt-TiO2 catalysts by doping with other metallic or non-metallic elements [48, 49]. Recently,Li et al. [50] reported that Pt/Ti0.9Sn0.1O2-C showed enhanced electrocatalytic activity and durability compared with those of Pt/TiO2-C and commercial Pt/C, as illustrated in Fig. 9. The improved performance was attributed to the presence of Ti0.9Sn0.1O2, in which the doped Sn increased the amount of adsorbed oxygen groups on the oxide surface, which aided the removal of COads intermediates from the active sites. In addition, the unique self-assembled Ti0.9Sn0.1O2-C structure provided junction deposition sites, allowing better dispersion of Pt nanoparticles.
A summary of the composition, structure, specific surface area, mass-normalized current density in the MOR and synthesis method of electrocatalytic materials developed for use in DMFCs is presented in Table 1. The catalysts characterized under alkaline conditions such as In0.1SnO2-Pt and (MoO3)0.2SnO2-Pt/C showed the best performance, but the acidic environment caused by the Nafion ionomer restricts their practicability in DMFCs. These catalysts are therefore not suitable for use in commercial fuel cells, unless researchers can make develop alkaline anion exchange membranes. Considering the rest of the reviewed catalysts, the Pt nanodendrites and Pt nanostars synthesized by the LLC template method achieved high electrocatalytic activities. However, it is difficult to expand the production scale of the LLC method to meet the requirements of mass production. In addition, the stability of the redox potential and fuel atmosphere also affect the choice of catalyst. Therefore,Pt-based catalysts supported by transition metal oxides and carbon materials and synthesized by a relatively simple method, such as Pt-NRCeO2/GN and Pt/Ti0.9Sn0.1O2-C, are potential candidates for DMFC application. Besides the development of novel structures and compositions, a thorough understanding of the reaction process in DMFCs is of utmost importance to achieve further advances. Further research efforts to improve the catalytic materials in DMFCs may focus on increasing catalyst durability.
Several issues in DMFC research such as anode CO2 removal, cathode water flooding and MCO are still awaiting more effective resolutions. Notably, these issues are all related to the structure parameters of the porous electrode in the MEA. Specifically, at both sides of the structural interface between the DL and CL, parameters such as porosity, topography and hydrophilicity are all different, and such differences can lead to large changes of transport properties over the entire MEA [51].
As described above, the performance of DMFCs is strongly influenced by the mass transfer process in MEAs. Therefore, many researchers have attempted to develop mathematical models to help optimize the electrode design of DMFCs by comprehensively evaluating the effects of different structural parameters on DMFC performance. For instance,Yuan et al. [52] developed a multiphysics model coupled with mass and momentum transport that defined the methanol concentration,CO2 volume fraction and polarization curve. As shown in Fig. 10, their simulation results indicated that the surface wettability of the anode DL determined the two-phase convection between CO2 and the methanol solution, and the optimal contact angle was 130°. Similarly,Sun and co-workers [53] developed a multiphase DMFC model to study the effect of microporous layers (MPLs) with different hydrophilicity properties on electrode performance. They found that as the contact angle increased along the flow direction of methanol in the anode DL, the liquid always needed to flow from a more hydrophilic region to a more hydrophobic one, leading to less MCO. The simulation results in Fig. 11 indicate that a higher anode DL contact angle generally led to more severe MCO, although it did increase the flow resistance at the inlet. In addition, the mass transport resistance at the DL/MPL interface was decreased, and the capillary-driven flow was enhanced inside the DL. However, because the MPL is very thin compared with the DL, it could act as a mass transfer barrier, mainly because of the contact angle differences across the MPL/DL and MPL/CL interfaces, which was the factor dominating the MPL function.
Weng et al. [54] experimentally investigated the MPL with a hydrophobic gradient at the cathode. Their experimental results revealed that the gradient MPL efficiently removed water from the electrode at high relative humidity. Liu and colleagues [55] compared several MEAs with different anode diffusion media, and concluded that the water transport coefficient through the membrane with a hydrophobic anode MPL was several times smaller than those of structures with a hydrophilic MPL or without an anode MPL.
Despite the advances achieved by this research, conventional MPLs made of carbon powder still suffer from the problem of cracks, which influence the permeability and capillary pressure of the electrodes. To tackle this issue,Yan et al. [56] proposed a MPL prepared with fluorinated carbon nanotubes (Fig. 12). The fluorinated nanotube MPL had no cracks and its contact angle was as high as 153°. Inclusion of the crack-free superhydrophobic MPL in a DMFC enabled a substantial increase in the water recovery flux, facilitated the MOR and thereby enhanced the cell performance. As a result, the maximum power density of the DMFC with the fluorinated nanotube MPL reached 25.6 mW/cm2, an improvement of more than 30% compared with that of the DMFC containing carbon powder (19.4 mW/cm2). Furthermore, the group proposed that the low surface energy of the material was the reason for its superhydrophobicity, and fluorine was effective to lower its surface free energy.
In a working DMFC, if the products cannot be removed in time, they may occupy the effective reaction region and block the influx of reactants, leading to performance degradation. At present, especially when a DMFC is working under high current conditions, the efficiency of mass transportation in the MEA is still systemically inefficient. Therefore,MCO and how to effectively remove the products from the CL to the flow field are two urgent problems waiting for solution.
To solve the issue of MCO, many researchers have focused on the introduction of novel porous structures such as porous carbon plates or polymer films as a methanol control layer to adjust the total water flux and thus mitigate MCO [57, 58]. Similarly, to improve product transportation, novel functional structures have been widely introduced into MEAs [59, 60]. Structural modification of CLs had received widespread attention. In particular,CLs with a hydrophobic gradient distribution have been shown to enhance oxygen diffusion and water removal [61]. Meanwhile, 3D CLs with designed pore distributions that were prepared by depositing the catalyst on polypyrrole-treated polystyrene spheres have also exhibited facile CO2 removal [62]. Some CLs with functional multilayer structures have been developed. Liu et al. [63] prepared a bilayer anode CL with a relatively dense inner layer to mitigate MCO, and a porous outer layer to meet the demands of the MOR. SEM images of the bilayer anode CL are shown in Fig. 13. The bilayer CL was prepared by an ultrasonic spray process that combined a catalyst-coated membrane transfer printing method with the direct spray technique. The performance improved from 116.8 mW/cm2 for the traditional MEA to 202.6 mW/cm2 for the optimized MEA at 80 °C. Another bilayer CL with a hydrophilic PtRu black inner layer and PtRu/C outer layer was developed by Suo's group [64]. The advanced cell performance was attributed to the gradients of catalyst concentration and porosity in the catalyst, which led to higher conductivity of charged particles and more facile reactant/product transmission. Moreover,Zhao et al. [65] prepared a multilayer electrode by the sputtering method (Fig. 14). This electrode displayed favorable cell performance and mass activity because of the good contact at the interface between the catalyst and PEM and low mass transfer resistance.
Many investigations have indicated that a comparatively high porosity is beneficial to improve cell performance [66, 67]. Introducing an appropriate amount of pore-forming additive during the preparation of the CL or MPL can improve cell performance [68]. Ammonium compounds that decompose at low temperature and acid-soluble oxide particles of appropriate sizes are both suitable for use as pore-forming agents in MEA modification because they can be entirely removed through heat treatment, hot pressing or acid washing [69, 70]. However, these pore-forming agents always induced aggregation of catalyst particles during their removal, so the formed pores tended to be disordered and relatively independent, which could lower their correlative beneficial effects.
To further optimize the performance of DMFCs,Yuan et al. [71] presented a method for anode optimization using a gradient porous medium (Fig. 15) to realize more effective control of the anode mass transportation. This functional medium was composed of a self-developed metal fiber sintered felt based on multi-tooth cutting and high-temperature sintering. The assembly with a low-porosity layer installed inside performed better than that with a reverse setup. Their results highlight the advantage of using a horn-like structure to promote fast removal of the CO2 produced on the anode side, because it induced the formation of a pressure gradient that can help to drive CO2 more effectively out of the electrode. Similarly,Chai et al. [72] successfully fabricated a nano-network structure (NNS) within the anode CL and MPL of a MEA using zinc oxide (ZnO) as effective sacrificial template (Fig. 16). The maximum power density increased from 29.0 mW/cm2 for the conventional MEA to 40.2 mW/cm2 for the optimized one. The improved performance of the MEA with the NNS was attributed to its higher catalyst utilization, higher mass transfer efficiency, and lower charge-transfer resistance compared with those of the conventional MEA. The connection of two network structures within different layers lowered the mass transfer resistance for methanol and water. However, the high content of water in the cathode strongly affected the reaction at the cathode and prevented the performance of the MEA from further improvement. This clearly indicates that the mass transfer on an anode could be improved considerably by the fabrication of a network structure. According to the results of EIS analysis (Fig. 17), the decrease in charge transfer resistance was one of the factors that could be responsible for the improvement in DMFC performance when a NNS was formed in the anode.
The electrocatalytic properties and mass transfer conditions of porous electrodes used in MEAs directly determine the power performance, stability and longevity of DMFCs. This review illustrates a variety of materials and structural modification methods for DMFCs reported recently. With the gradual expansion of related studies, the efficiency of electrocatalysis in DMFCs has been improved considerably. However, more work should be performed to clarify related mechanisms such as the reaction kinetics of multimetal catalysts, the functional principles of nanostructured supports, the adsorption mass transfer process at the MPL/CL interface and a practical MEA structure to allow inner water circulation. Accompanied with the accumulation of a complete theoretical system of DMFCs, additional breakthroughs to obtain important materials and continuous improvement of operating life, the prospects of DMFCs are bright.
2016,
Vol. 37 Issue (7): 1037-1048 
PDF (1477 KB)
2016,
Vol. 37 Issue (7): 1037-1048 PDF (1477 KB)