Abstract
Increasing concerns with non-renewable energy sources drive research and development of sustainable energy technology. Fuel cells have become a central part in solving challenges associated with energy conversion. This review summarizes recent development of catalysts used for fuel cells over the past 15 years. It is focused on polymer electrolyte membrane fuel cells as an environmentally benign and feasible energy source. Graphene is used as a promising support material for Pt catalysts. It ensures high catalyst loading, good electrocatalysis and stability. Attention has been drawn to structural sensitivity of the catalysts, as well as polymetallic and nanostructured catalysts in order to improve the oxygen reduction reaction. Characterization methods including electrochemical, microscopic and spectroscopic techniques are summarized with an overview of the latest technological advances in the field. Future perspective is given in a form of Pt-free catalysts, such as microbial fuel cells for long-term development.
1. Introduction
Due to depleting fossil fuel resources accompanied with drastic climate changes, looking for renewable and sustainable energy is imminent. Fuel cell technology is among many approaches recently developed aiming at solving the global energy challenges. Fuel cells (FCs) are electrochemical devices that convert chemical energy stored in fuel molecules into electric energy via electrochemical reactions [1]. Basic elements in a fuel cell are cathode, anode, electrolyte and fuels such as hydrogen, methanol, ethanol and formic acid [2]. Critical catalyzed reactions occur at both electrodes. As an example, reactions for a direct methanol fuel cell are [3]:
Fig. 1 (Color online) Schematic overview of different types of fuel cells
Since bulk Pt atoms do not participate in reactions, Pt utilization is improved by increasing the surface area. Reducing the size of the particles (i.e., nanoparticles) increases the surface area. Another approach to limiting bulk Pt is core-shell nanoparticles where the core is made from less expensive metal than Pt and serves as a support for a thin layer or even a monolayer of Pt in the shell. Pt poisoning can be overcome by alloying with materials less prone to CO and CO2 poisoning, making the catalyst more resistant to poisoning [12]. Catalysts must be immobilized on a suitable supporting material for cathode or anode. The most common supports are carbon black, binary carbon catalyst support (Vulcan XC-72 and Black pearl 2000), conducting polymers (eliminating the need for Nafion® layers or impregnation), carbon nanohorn support and carbon nanotube hybrids [8]. Recently, graphene has been utilized as a catalyst support in FCs [11].
This review focuses on the application of graphene as a support for the catalytic material in the FCs. It includes an introduction to graphene as a catalyst support, with the progress on Pt in FCs, an overview of graphene–Pt hybrids in FC applications, a presentation of methodology used for the characterization of these FC systems and a perspective on Pt-free catalysts for future FCs.
2. Graphene
Graphene is a two-dimensional (2D) single sheet of sp2 carbon in a hexagononal arrangement [13]. Graphene possesses unique properties, such as high charge-carrier mobility (up to 105 cm2 V⁻¹ s⁻¹), super conductivity, ambipolar electric field effect, quantum Hall effects at room temperature, high mechanical strength (130 GPa) and high surface area (2,600 m2 g⁻¹) [14]. Graphene can be chemically synthesized from graphite and constructed in a form of three-dimensional (3D) foam or paper. These properties make graphene ideal as electrode support for catalysts in electrochemical energy systems [15]. Three-dimensional porous graphene has exhibited remarkable properties for electrochemical power systems. Such materials are classified as macroporous ([50 nm), mesoporous (2–50 nm) and microporous (\2 nm) according to their pore size [16]. Synthesizing 3D graphene with precise control of shape and pore size is required. Nanoparticles interact strongly with functional groups on graphene, e.g., hydroxyls, epoxides, carbonyls and carboxylates making exact engineering of graphene functionalization vital [17]. However, this compromises the high electronic conduc- tivity of graphene. Usually, graphene production is divided into ‘‘wet’’ and ‘‘dry’’ approaches [18, 19]. The latter is based on building graphene sheets from simple carbon- containing molecules, such as methane and ethanol [20]. Wet approach is chemical synthesis of graphene in solution from graphite via formation of graphene oxide (GO) [18]. A summary of reducing agents in reported procedures and corresponding properties is given in Table 2.
For fuel cell applications, porous 3D graphene structures are created in order to fully utilize its properties [15]. Porous graphene networks are in rapid development and most important are graphene nanomesh and graphene foam (Fig. 2). Graphene nanomesh is graphene structure with high-density nanoscale pores situated on top of conjugated carbon surfaces [15]. Its synthesis includes polymer building blocks, photo-, electron- and plasma- etching, template methods and chemical-etching methods [34]. Graphene foam (GF) is made by stacking graphene nanosheets into a connected network [15]. GFs are mechanically strong and compressible with multidimensional electron pathways which ensure excellent electrochemical performance. GFs are usually produced using hydrothermal methods, chemical reduction methods or template-directed chemical vapor deposition (CVD) methods. The advantage of chemical methods is that crosslinking sites are formed, making a chemically bonded GF framework. Worsley et al. [35] crosslinked GO in aqueous solution with resorcinol–formaldehyde by reducing it in a sol–gel procedure using NaCO3 as a catalyst, followed by supercritical CO2 drying and pyrolysis (1,050 °C) under N2 atmosphere. The obtained product showed high specific surface area (1,200 m2 g⁻¹). In template methods, graphene is deposited on a template by chemical vapor deposition.
Table 1 Overview of electrochemical reactions, electrolyte, electrodes catalysts and their properties in different types of fuel cells
Table 2 Summary of GO-reducing agents and reacting conditions in accordance with well-supported and proposed mechanisms
Fig. 2 Scanning electron microscopy (SEM) image of a pristine Ni foam and b doped 3D graphene. c, d Transmission electron microscopy (TEM) images of doped 3D graphene. Adapted with permission from Ref. [33]. Copyright 2013, American Chemical Society
Representative work was done using CH4 deposited at 1,000 °C on porous Ni foam as a template [36].
Current research focuses on enhancing mechanical strength of 3D graphene for practical application. The diameter of GF pores is in range of a few macrometer to a few hundred micrometers. Networks of nanometer-sized pores are needed to develop optimized fuel cell system.
3. Platinum in fuel cells
Platinum (Pt) was discovered as a valuable material in 1780s and is one of the most important catalysts for chemical reactions. The main drawback is the high price of Pt-based catalysts. Pt is a scarce element in the Earth’s crust, with a relative abundance of approximately 0.01 ppm (1 ppm = 1 mg L⁻¹) [37], and only a few hundred tons are produced annually. It is mostly used for catalytic converters (35 %–40 %), jewelry (up to 35 %) and petroleum and glass production (18 %) [38]. For these reasons, extensive studies have been undertaken to synthesize specific Pt nanoparticles (PtNPs) to maximize and optimize the catalytic surface area, while minimizing material consumption. Efforts have been put on optimizing shape, size and facet distribution of the NPs, leading to improved activity and selectivity. Syntheses through wet chemical reduction (WCR) enable nanoscale design by controlling synthetic parameters to tune shape and size of the NPs [39]. Formation of PtNPs in solution, includes reduction, nucleation, growth and ligand capping, processes that all affect shape and size [40]. Non-spherical PtNPs with high edge and corner atoms are presented in Fig. 3. Representative recipes for syntheses of PtNPs are summarized in Table 3. As an example, TEM pictures of PtNPs and loading on carbon support are shown in Fig. 4. Pt is primarily utilized as a catalyst for oxygen reduction reaction (ORR), one of key reactions in fuel cell systems [47]. ORR is a multiple electron reaction, which includes a number of elementary steps with different reaction inter- mediates [48]. Wroblowa et al. [49] proposed a reaction scheme elucidating the complexity of O2 reduction at the metal surface. O2 can be electrochemically reduced directly to H2O (‘‘direct’’, 4e⁻ reduction) or through intermediate formation of adsorbed H2O2 (‘‘series’’, 2e⁻ reduction). Adsorbed H2O2,ad can be electrochemically reduced into H2O (‘‘series’’, 2e⁻ pathway), decomposed at the metal surface or desorbed into the solution. Markovic´ et al. [48] have experimentally established an interpretation of the Pt-catalyzed ORR through a ‘‘series’’ pathway via H2O2,ad intermediate, concluding that negligible O–O bond splitting occurs prior to peroxide formation. H2O2,ad may or may not be reduced to water, but in both cases, the addition of the first electron to O2 is the rate-determining step in the ORR.
Fig. 3 TEM images of as-synthesized starch-capped PtNPs a and Pt–CB catalyst for PEMFC (b, c). For detailed information, see Ref. [41]
Table 3 Recipes for the syntheses of PtNPs
The development of Pt-based catalysts requires fundamental understanding of the Pt–electrolyte interfacial reactions, as well as optimization of the Pt surfaces. Promising catalyst must include both large surface area and high specific activity. Changing the local bonding environment, the distribution of active sites and intrinsic electronic properties allows for tailor-made catalysts with enhanced properties and ultra-low Pt loadings.
Fig. 4 High-resolution TEM images of PtNPs obtained with molar ratios between NaNO3 and H2PtCl6 of (a, e) 5.5 and (b, f) 11.0, respectively. The insets show fast Fourier transform (FFT) patterns used to determine the crystallographic directions marked in the insets. c, d, g, h Models showing that the growth of PtNPs was substantially enhanced at ridges and corners to form both octapods and tetrapods. Adapted with permission from Ref. [46]. Copyright 2004, American Chemical Society
The relationship between surface area and electrochemical reactivity has been termed as ‘‘structural sensitivity’’ describing a relationship between the kinetics of electrochemical reactions and crystal surface structures [50]. Correlations between the kinetics of the ORR and the surface coverage of chemisorbed oxygen-containing species have been reported on polycrystalline and single- crystal electrodes [51]. Chemisorbed oxygen-containing species are present on the Pt surfaces in a reversible form (denoted as OHad) and an irreversible form (denoted as ‘‘oxide’’). Experiments demonstrated that the order of activity of Pt(hkl) in 0.1 mol L⁻¹ KOH, under combined kinetic–diffusion control (potential range of O2 reduction E [ 0.75 V), was (100) \ (110) \ (111) [52]. However, reversible adsorption of hydroxyl ions (OHrv) on Pt(hkl) suppressed the kinetics of the ORR, but did not change the pathway of the reaction. The ‘‘oxide’’ form changed the pathway of the reaction, O2, was not fully oxidized, and peroxide was detected on the ring electrode [52]. Furthermore, trace amounts of chloride can significantly change the activity and the reaction pathway of ORR on Pt catalysts. In PEM fuel cells, chloride impurities occur at the ppm level arising from the membrane electrode assembly (MEA) preparation process or contamination of humidified feed streams. Even a 4-ppm chloride impurity can result in a voltage loss of 50 mV and equally affect the open-circuit cell voltage [48]. Therefore, MEA preparation and humidified feed streams require high purity.
4. Graphene–Pt hybrid materials
Hybrid graphene–platinum (G–Pt) material can be traced back to 1999, and it has been developing fast for the past decade [53]. G–Pt syntheses can be broadly divided into four strategies: (1) ‘‘One pot synthesis’’ where both precursors, GO and Pt salt, are reduced simultaneously [54], (2) GO reduced in the presence of PtNPs [55], (3) PtNP precursor reduced in the presence of reduced GO (graphene) [56] and (4) ‘‘separate reduction’’ where graphene is mixed with already prepared PtNPs which are then incorporated in the graphene network [57] (Fig. 5). G–Pt with specific features including nanotubes [58], nanospheres [59], nanofibers [60], nanocages [61] have been synthesized using different reduction methods, such as chemical reduction, electrochemical reduction, thermally or microwave-assisted methods and combined reduction methods. Density-functional calculations were carried out to study the interfaces of graphene with icosahedral sub-nanosize clusters of Pt13 and Au13 [62]. Introducing a carbon vacancy into graphene network increased adsorption energy of the clusters and graphene flakes containing five or seven rings increased the adsorption energy of the clusters more than a flat, defect-free graphene sheet. CO and H chemisorption energies become smaller on the metal clusters on graphene than on the clusters without carbon support. Introduction of graphene increased catalyst poison tolerance [62]. Graphene nanosheets decorated by Pt clusters with a range of 0.5–1.5 nm were reported [63].
Fig. 5 Routes of syntheses to anchor PtNPs on graphene. Graphene and GO are marked by black and orange, respectively
The obtained PtNPs exhibited enhanced specific surface area in comparison with Pt on carbon black (Vulcan XC-72R) and improved activity for the methanol oxidation reaction (MOR). A general approach to the preparation of graphene–metal nanoparticles in water–ethylene–glycol system was proposed using graphene oxide as a precursor and metal nanoparticles (Au, Pt and Pd) as building blocks [64]. In this method, metal NPs were adsorbed on the surface of GO, which was not only beneficial for the sub- sequent reduction of GO by ethylene glycol, but also prevented restacking of reduced GO sheets, resulting in the formation of stable G–NP composites. Kou et al. [65] synthesized a G–Pt hybrid material with enhanced durability. The properties of both G–Pt and commercial catalyst were investigated with cyclic voltammetry (CV) over 5000 cycles in N2-saturated 0.5 mol L⁻¹ H2SO4. The initial electrochemical surface areas (ECSAs) of G–Pt and commercial catalyst (E-TEK) were 108 and 75 m2 g⁻¹, respectively. The ECSA for both G–Pt and E-TEK decreased after the 5000 cycle degradation to 62.4 % (67.6 m2 g⁻¹) and 40 %, respectively. ORR activities decreased to 49.8 % for the G–Pt, while the commercial catalyst kept only 33.6 % of its original activity. In another study, G–Pt was prepared using chemical reduction of GO and H₂PtCl₆ by NaBH4₄ [66]. The peak current density of Pt at the potential of 0.652 V (vs. Ag/AgCl) showed that the composite had superior catalytic performance toward methanol oxidation, almost twice that of the commercial Pt–Vulcan catalyst (199.6 and 101.2 mA mg⁻¹ Pt, respectively). The PtNPs in this setup were in the range of 5–6 nm. The catalytic stabilities of the G–Pt and Pt–Vulcan were ex- amined by chronoamperometry in 0.5 CH3OH and 0.5 mol L⁻¹ H2SO4 at a fixed potential of 0.60 V. The catalysts initially showed the same tendency, where potentiostatic currents decreased rapidly. Formation of intermediate species like COads and CHOads, during methanol oxidation, might have been the cause of such an effect. Gradually, the current decay slowed down and a pseudo- steady state was achieved. It appeared that the current density of the G–Pt was higher than that of Pt–Vulcan catalyst electrodes during the whole testing duration, which indicated superior electrocatalytic stability of the G–Pt catalyst relative to the Pt–Vulcan catalyst. An example of graphene nanofoam (GNF) and application of its hybrids with PtNPs in methanol reduction are presented in Fig. 6. Pt loadings up to 80 wt% were achieved on surface-functionalized graphene nanosheets, with particle diameters less than 3 nm, resulting in current densities of methanol electrooxidation at least twice that of conventional Pt on carbon support [67]. Furthermore, these catalysts main- tained high Pt mass activities with increased Pt loading on the working electrode ranging from 0.2 to 2.0 mg cm⁻².
Highly controllable deposition of PtNPs has been established with uniform PtNPs and high dispersion assembled on graphene functionalized with poly(diallyldimethylammonium chloride) via a NaHB4 reduction process [68]. Loadings ranging from 18 wt% to 78 wt% of approximately 4.6 nm PtNPs were achieved exposing predominantly Pt(111) facets. An overview of representative electrocatalysts employed in PEMFC systems and relevant properties is given in Table S2 (online).
Bimetallic nanoparticles enhance the catalytic properties and resistance to catalyst poisoning [69]. G–Pt–Ru catalyst exerted higher electrocatalytic activity for both methanol and ethanol oxidation than monometallic G–Pt catalyst [70]. Assuming the tolerance toward catalytic poisoning is the ratio of the forward peak current density (If) to the reverse anodic peak current density (Ir), G–Pt hybrid provided a ratio of 6.52 which was much higher than those for Pt–C and Pt–graphite (1.39 and 1.03, respectively). Similarly, graphene-supported Pt, Pt₃Co, Pt₃Cr and G–Pt– Pd alloy nanoparticles are reported [71, 72]. These alloy electrocatalysts generally show enhanced properties toward ORR (or MOR), high electrocatalytic activities and low overpotentials, due to the properties of the individual metals and synergetic effects. The geometry and architecture of such NP assemblies (dendrites, core-shell, cage, etc.) also have significant effect. The suppression of the formation of OHad species on Pt surfaces can be achieved by alloying Pt with Co and Cr [73]. The specific activity of the Pt–Cr alloy catalysts for the ORR in the absence of methanol is higher than that of the pure Pt catalyst and increases with increasing Cr content. The higher overpotential for ORR is due to absorption of methanol species on the catalyst surface. The alloy catalysts can catalyze the oxygen reduction but effectively limit the oxidation of methanol [73]. The bimetallic Pt–Pd hybrids exhibited 3 to 9.5 higher electrocatalytic activities for MOR than their single-component counterparts [71]. The observed dendritic growth of Pt branches was attributed to the high rate of reduction and (111) facet epitaxial growth of Pt mediated by Pd seeds with (111) facets. This resulted in lowering of the potential at a given oxidation current density for the hybrid material when compared to E-TEK or platinum black (PB), commercial catalysts. It means that synthesized hybrid Pt–Pd on graphene exhibited better performance for methanol electrooxidation at applied potential (from 0.3 to 0.7 V) [71]. Jiao’s group [74] synthesized high-quality exfoliated graphene decorated with Pt nanocrystals (3 nm) by simple, low-cost and environmentally benign process of thermal expansion and liquid exfoliation, followed by solvothermal reaction (TELESR process). The hybrid material exhibited larger If/Ib ratio and higher peak current densities than commercial Pt–C catalyst used for MOR. The method can be universally applied to functionalize graphene nanosheets with Pt–M (M = Pd, Co) alloys, which demonstrate larger ECSA, improved efficiency for MOR and long-term stability.
Trinary alloy systems have been synthesized and studied due to even higher catalytic activity and stability than binary systems.
Fig. 6 a SEM images of the pristine GNF. b Mass and real area double-normalized CV profiles of Pt nanoclusters on GNFs for the methanol oxidation in 0.5 mol L⁻¹ H2SO4 ? 1.0 mol L⁻¹ CH3OH. Adapted with permission from ref. [4]. Copyright 2010, American Chemical Society
The PtPdAu/G catalyst (8.5 nm) has been obtained by a simple one-step ethylene glycol–water reduction process from GO and Pt, Pd and Au precursor salts [75]. The onset potential of MOR on PtPdAu/G shifted to the more negative values, indicating effective decrease in overpotential for the electrooxidation reaction. Moreover, the current of methanol oxidation on the PtPdAu/G electrode was high, and the current density on the forward scan of PtPdAu/G was 1.5, 2.3 and 2.8 times as high as those of PtPd/G, PtAu/G and Pt–G, respectively. However, trinary alloy systems consist of expensive, noble metals and still exert limited stability to intermediate species and anode fuel crossover. Reports on bi- and trimetallic Pt-based catalysts are summarized in Table 4 and S3 (online).
Gram-scale synthesis of defective GF decorated with PtNPs and Pt-decorated 3D graphene structures has been reported [85]. Defective GF as a support for PtNPs pro-vides a large surface area and small pore sizes. Loading of PtNPs can be increased from 20 wt% (for GF, carbon black, commercially available graphene) up to 33 wt% on defective GF [85]. Relative durability was tested by recording the retention of ECSA over 60,000 cycles via accelerated start/stop tests. GF–Pt lost all of the activity after only 30,000 cycles, due to facile oxidation of GF support [85]. In membrane assemblies, GF showed better performance than carbon black in the activation region, but significantly worse performance in the mass diffusion region [85]. Huang et al. [86] recently presented promising data for methanol oxidation catalysis using Pt on functionalized graphene 3D structures.
Amino species bring functionalization potentials to reduced GO. Pyridine-type nitrogen can attract PtNPs from colloidal solutions. Nitrogen-doped graphene materials are excellent supports for Pt nanoparticles. These hybrids have exhibited outstanding activity toward electrochemical methanol oxidation reaction due to the high dispersion state of the PtNP and superior conductivity induced by the incorporated nitrogen [87].
Table 4 Bimetallic Pt catalysts for fuel cell application
The methanol oxidation current of Pt catalyst on nitrogen-doped graphene (Pt–NG-800, 135 mA mg-1) was three times higher than those of the other composites and the commercial Pt–CB (27 mA mg-1).
In addition, the higher temperature treatment greatly reduced the number of defect sites in the hybrid, enhancing its conductivity. Pt-decorated 3D architectures were synthesized from graphitic carbon and nitride nanosheets by means of a co-assembly approach [86]. Interconnecting 3D structures with much higher nitrogen content than previously reported (29.4 at% compared to less than 10 at%) enhanced the PtNP adhesion, effectively avoiding their agglomeration [86]. As-synthesized NPs were closely packed and uniformly dispersed on graphene support, with an average particle size of 3.4 nm and prevailingly (111) facet surface. Moreover, a strong metal–support link (induced by N) resulted in reduced accumulation of COads on Pt, increasing the catalyst poison tolerance. The stability of the catalyst for methanol oxidation reaction (MOR) was investigated by chronoamperometry at 0.5 V for 2,000 s [86]. The prepared catalyst showed the lowest current decay and retained the highest oxidation current over time. Cyclic voltammetry confirmed these results. After 100 cycles, the initial forward peak current density decreased by 38.9 %, compared to 71.2 % for Pt–C catalyst [86]. SEM and TEM images were taken after the durability measurements, showing clearly integrated, well-preserved and highly dispersed Pt nanoparticles on the surface of the functionalized graphene structure.
5. Electrocatalysis, electrochemical investigation
Electrocatalysis is the change of rate and selectivity of electrochemical reactions on the electrode interface [88] and is studied using electrochemical techniques including cyclic voltammetry (CV), electrochemical impedance spectroscopy (EIS), chronoamperometry (CA), rotating disk electrode (RDE) and rotating ring disk electrode (RRDE) methods [89]. These methods provide information on the thermodynamics of redox processes, adsorption processes, kinetics of electron transfer reactions, reaction mechanisms, dielectric properties of a medium over a range of frequencies, capacitive properties of the material and quantitative determination of electroactive species adsorbed on the electrodes. RRDE provides information about specific products formed during catalytic reactions at the disk electrode based on the current measured at the surrounding ring electrode. Electrochemical oxygen reduction is one of the key reactions for the conversion of chemical energy into electrical energy [90]. Voltammetric experiments preformed under RRDE hydrodynamic conditions are the basis for the evaluation of the ORR electron transfer pathways as well as formation of intermediates. Although valuable information can be obtained by employing these methods, they were developed for flat surfaces and they do not fit to porous electrodes [91]. In this type of electrode, the ORR products do not follow the electrolyte flow perpendicular to the films generated by the electrode rotation, but stay in pores and channels inside the catalyst. Prolongation of the time products spend in contact with the catalyst layer may lead to secondary reactions, and the number of transferred electrons can be strongly influenced. Research in fuel cell catalytic material development is primarily focused on RDE and RRDE methods for catalyst characterization, even when working with porous structures. In order to characterize catalytic materials efficiently, in situ methods i.e., electrochemical methods combined with other techniques, are important.
Scanning electrochemical microscopy (SECM) is a microscopy-based electrochemical approach providing laterally resolved characterization of electrode surfaces [92]. The SECM tip is positioned in close proximity to the sample surface acting as a local sensor for the detection of reaction products generated at the surface of the catalyst. The spatial distribution of selective domains on electro- catalyst surfaces can be obtained with submicrometer resolution [93].
The selectivity of electrocatalysts can be identified with a combination of electrochemical and chromatographic techniques, such as gas chromatography (GC), ion chromatography (IC) and high-performance liquid chromatography (HPLC) [94]. Santasalo-Aarnio et al. [95] investigated the selectivity of Pt and Pd during oxidation of methanol and ethanol, respectively, in alkaline solutions. It was demonstrated that the electrochemical oxidation of methanol on Pt leads to the formation of two side products, formaldehyde and formate. In the case of Pd, only formate was obtained. Similarly, electrochemical oxidation of ethanol produced acetaldehyde and acetate on Pt, while it was more selectively oxidized to acetate on Pd [95].
To investigate the selectivity of Pt-based catalysts during methanol, ethanol and formic acid electrooxidation, differential electrochemical mass spectroscopy (DEMS) can be applied [96]. It combines electrochemical measurements with mass spectroscopy (MS) in order to detect reaction products. Scanning differential electrochemical mass spectroscopy (SDEMS) was employed to assess the ‘‘integral’’ selectivity of electrocatalyst and spatially map the catalyst selectivity [97]. An array of electrodeposited Pt and Pt–Ru catalysts was tested for their selectivity in methanol electrooxidation. Pure Pt exhibited extremely high conversion efficiency of methanol to CO2, reaching up to 90 %. The addition of 6 wt% Ru led to the highest current densities in case of both bimetallic Pt–Ru and pure Pt catalysts. Increased Ru content led to a decrease in the amount of formic acid produced during methanol electrooxidation. At Ru contents above 50 wt%, formic acid was non-detectable and formaldehyde was the only side product [97]. DEMS is suitable for volatile reaction products capable of passing through the separating membrane between the electrochemical cell and MS detector.
Detection of products of electrocatalytic reactions can be achieved by electrospray ionization mass spectrometry (ESI-MS) [98]. Limited fragmentation occurs during ionization which minimizes interference among different mass fragments. ESI-MS was used to simultaneously detect volatile and non-volatile side products of methanol electrooxidation on Pt-based catalysts [99].
A number of spectroscopic techniques have been combined with electrochemical techniques. Infrared absorption/ reflection spectroscopy (IRS) [100], surface-enhanced infrared absorption spectroscopy (ATR-SEIRAS) [101], electron plasmon resonance (EPR) [102], fast Fourier transform infrared spectroscopy (FTIRS) [103], attenuated total reflection–Fourier transform infrared spectroscopy (ATR-FTIRS) [104] and Raman spectroscopy [105] are spectroscopic techniques used for in situ probing of electrochemical systems. FTIRS and Raman are combined with electrochemical methods to determine the selectivity of electrocatalysts. In situ FTIRS elucidates mechanistic and kinetic information with respect to organic fuels used in fuel cells (methanol, ethanol, formic acid). Using the ATR- SEIRAS method, a triple-path mechanism in electrooxidation of formic acid on Pt catalyst in acidic media was discovered [101]. The adsorption–desorption equilibrium of adsorbed formate is quickly established, and a part of the adsorbed formate is decomposed to CO2 and H⁺ ions [101].
Durability of catalysts has to be tested at the developmental stage in order to assess long-term performance. Potentiostatic and galvanostatic polarization techniques are the most used electrochemical techniques to assess catalyst stability. Electrochemical impedance spectroscopy (EIS) is used to monitor degradation of electrocatalysts at the electrode–electrolyte interface [106]. CV provides simple means to get information about degradation at certain potential values. Electrochemical quartz crystal microbalance (EQCM) monitors degradation of metallic catalyst by weight [107]. An inductively coupled mass spectrometer (ICP-MS) equipped with electrochemical scanning flow cell offers in-depth investigation of degradation [108]. ICP- MS enables online downstream elemental surveillance of electrocatalyst degradation products at different electrode potentials.
Ex situ methods include microscopic and spectroscopic techniques. Morphology, distribution and size of the particles are imaged prior to and after the durability tests. Transmission electron microscopy (TEM), high-resolution TEM (HR-TEM), identical location TEM (IL-TEM) [109], identical location scanning electron microscopy (IL-SEM), atomic force microscopy (AFM) and electron energy loss spectroscopy (EELS) are advanced techniques that monitor ex situ degradation of nanocatalysts.
Spectroscopic techniques provide information on the degradation phenomena based on crystallographic structures, elemental states and distribution. Primarily used techniques include X-ray photoelectron spectroscopy (XPS) to analyze elemental composition and chemical states, powder X-ray diffraction (XRD) to analyze durability of particles, size and crystal orientations, X-ray adsorption spectroscopy (XAS) to investigate atomic local structure and electronic states and Raman spectroscopy which identifies molecules by vibrational and rotational frequencies of the system [79]. IR spectroscopy is used to determine functional groups in molecules. Applied technique depends on the type of the information which is sought. More detailed descriptions of techniques used to characterize different components of the fuel cell systems are given in Table S4 (online).
6. Future perspectives: Pt-free catalysts
Recently, Pt-free catalysts have been developed, such as non-precious metals (Fe, Co, Mn), metal oxides (Fe2O3, Fe3O4, Co3O4), nitrogen-doped graphene and biomaterials (bacteria and enzymes) [110, 111]. Metal and metal-oxide catalysts often suffer from dissolution, sintering and agglomeration under working conditions of a FC system. This results in catalyst degradation and strongly diminishing electrical conductivity. Electron transfer is then hampered during ORR process. Heteroatom doping of (N, B and S) graphene nanomaterials resulting primarily in high ECSA gave good electrochemical response and reactivity and are regarded as promising candidates in replacing noble elements as electrocatalysts toward ORR [18]. Nitrogen-doped graphene (N-G) exhibited high electrocatalytic activity with long durability as well as CO tolerance compared to conventional Pt catalysts [112]. The N-G film showed good performance for oxygen reduction reaction associated with alkaline fuel cells. The steady-state catalytic current at the N-G electrode was found to be almost three times higher than with Pt–C over a large potential range. The long-term stability, anti-poison effect and tolerance to crossover are better than Pt–C for oxygen reduction in basic electrolyte. It is notable that the N-doped graphene film possesses remarkable electrocatalytic properties for ORR similar to that of nitrogen-containing vertically aligned carbon nanotubes. Biofuel cells operate using biological systems as catalysts for the redox reactions. The biological systems are either enzymes or whole microorganisms, such as bacteria. Biofuel cells (BFC) are promising green energy sources that harvest electricity from various organic materials. The key advantages of these biological systems are sustainability and renewability. Most critical issues are short lifetime and low power densities. These features are in direct relation to enzyme/bacteria stability, electron transfer rate and catalyst loading [113]. Microbial catalysts are whole living organisms that are immobilized on electrodes and catalyze reactions in fuel cells (Fig. 7). Generally, these are robust systems which operate on a variety of fuels and are usually capable of oxidizing the substrate completely to CO2 and water. In a representative MFC, direct electrochemical communication of Shewanella putrefaciens with an anode is mediated by membrane cytochromes [113]:
Fig. 7 (Color online) Schematic principle of a microbial fuel cell (MFC). A microorganism serves as a catalyst
The most intriguing achievement made so far has been the fabrication of bioanodes and biocathodes by self-assembly of bacteria with graphene, in a single-step process [119]. This effective and sustainable approach needs optimization of operational and constructional features in order to develop cutting edge, high-performance fuel cells.
- Indo Tambangraya Megah (ITMG)
- Bukit Asam (PTBA)
- Baramulti Sukses Sarana (BSSR)
- Harum Energy (HRUM)
- Mitrabara Adiperdana (MBAP)
- Adaro Energy (ADRO)
- Bumi Resouces (BUMI)
- Samindo Resources (MYOH)
- United Tractors (UNTR)
- Berau Coal
