1. Introduction
Transition-metal oxides with perovskite-type structures have received great attention because of their excellent physical properties and potential applications. It is well known that perovskite materials have wide applications in catalysis because of their defective structures and excellent oxygen mobility
[1-
5]. However, the major drawback of using perovskites as electro catalysts is their low specific surface area owing to the high temperature required for perovskite phase crystallization
[5]. The synthetic method and conditions have been observed to have a strong influence on the catalytic activities of perovskites
[6,
7]. LaCoO
3 has many practical applications owing to its excellent physical and chemical properties. Synthesis of LaCoO
3 has been achieved with a number of methods, including a sol-gel method
[8,
9], a polymerizable complex method
[10], combustion synthesis
[11], molten chloride flux
[12], mechanochemical synthesis
[13], alkaline co-precipitation
[14], and electrochemical oxidation
[15]. The properties of the final materials obtained are highly dependent on the preparation method
[12-
16]. The reverse micro emulsion, which is also known as W/O (water in oil) micro emulsion method, is a recently developed technique, ideal for the preparation of inorganic nanoparticles
[17]. Water in oil micro emulsion proceeds by dispersing water phase (dispersion phase) in oil (continues phase). Oil phase goes to micelles formation on addition of surfactant while water phase containing reactants goes inside the micelles on mixing these two phases. Further mixing results in nucleating the nanoparticles via exchanging the counter ions at the interface between micelles and water. This exchange process is the rate determining step. The micelles act as nano reactors as growth of particle is being progressed inside them. The surfactant nature, water to surfactant ratio, solvent: Oil and reducing agent to surfactant ratio control the final particle size. When water ratio is lowered it result in rigidity, i.e., low mixing of reactants and hence slow growth. While, increasing water ratio leads to better solvation and good inter-micellar exchange rate which results in decrease in particle size. Hence, optimum water to surfactant ratio is used. Solvent affect particle growth rate and final product size. When the solvent having smaller molecular volumes were used, they penetrate into the tails of surfactant giving rigidity and slow growth rate. So, keeping this effect in mind, Isooctane which is somehow bulky and have larger molecular volume were used to control loose penetration at the tails of surfactant and increasing growth rate to control crystallites size
[18]. The main advantages of this approach are that the reactants are mixed at a more homogeneous level in this oil-based solution, and grains are effectively inhibited from coalescing during the synthetic process. The final materials have superior properties; including nanometric dimension grain sizes, uniform size distribution, and phase purity
[19].
Oxygen reduction and oxygen evolution are essential electrochemical reactions in rechargeable metal– air batteries, fuel cells, chloralkali cells, water electrolyzers, and CO
2 reduction. Rechargeable metal–air batteries generate electricity through a redox reaction between the metal and oxygen in the air. Alkaline solutions are commonly used as electrolytes in aqueous metal–air batteries. In aqueous electrolytes, metal–air batteries encounter large discharge-charge over potentials (i.e., large voltage gaps and low round-trip eciency), arising from slow rates of the oxygen reduction reaction (ORR) and the oxygen evolution reaction (OER) in the absence of catalysts. The presence of an oxygen electro catalyst can accelerate the ORR and OER processes
[20,
21]. Noble metals such as Pt (performs poorly in OER) and noble metal oxides such as IrO
2 or RuO
2 (performs poorly in ORR) are used as state-of-the-art electro catalysts for ORR and OER, respectively
[21,
22]. However, noble metals are expensive, and hence, research efforts are focused on developing low-cost transition-metal oxides as alternative electro catalysts. Metal oxide catalysts incorporated into the carbon electrode enhance both oxygen reduction and evolution kinetics, and increase the specific capacity of the cathode
[23]. Shao-Horn and co-workers have intensively studied perovskite oxides with the help of molecular orbital principles, and showed that LaCoO
3 is one of the promising candidates among the various perovskite-type oxides, possessing distinctive intrinsic activity for both ORR and OER
[24]. However, despite numerous studies, the exact mechanism of ORR on perovskites is still unclear. Both direct 4e
− [24] and series 2e
− + 2e
− pathways
[25] have been proposed for perovskite cathodes. The reactions at the alkaline electrolyte can be written, as shown in
Table 1.
Table 1.
Different pathways for reactions in alkaline electrolyte
However, mass-transport resistance and/or incomplete wetting of the electrode structure often hamper the evaluation of the intrinsic electro catalytic activity of catalysts. It is proposed that the use of rotating disk electrode (RDE) measurements, a fast screening tool with well-known mass-transport properties where the electro catalyst and Nafion solution are deposited onto a glassy carbon (GC) RDE, would help avoid or reduce these problems
[26]. When an electrode is rotated, the mass transfer of reactants and products occurs by a convective-diffusional mechanism. The theory of convective flow at the RDE is identical to the theory of flowing fluids (hydrodynamics)
[27], and is thus termed hydrodynamic voltammetry. At the RDE, a hydrodynamic flow pattern results from centrifugal forces that move the liquid horizontally out and away from the center of the disk while fresh solution continually replaces it with the flow normal to the electrode surface.
It has been demonstrated that perovskites alone show low electro catalytic activity, which may be increased by a few orders of magnitude when carbon is added to the catalytic layer. There is a synergistic chemical coupling between perovskite and carbon
[28]. In this study, a thin-film RDE technique was used, whereby a very thin layer of LaCoO
3 electro catalyst, synthesized by the micro emulsion method, was tested in order to reveal the influence of carbon super P on the catalytic activity of the LaCoO
3 electro catalyst.
3. Results and Discussion
Powder XRD data (
Fig. 1) shows the as-synthesized perovskite oxides LCO-ME and LCO-SP obtained from the micro emulsion and sol-gel methods, respectively. XRD patterns of both samples exhibited diffraction peaks consistent with those of LaCoO
3 (JCPDS No.-01-084-0848), irrespective of the preparation method. This showed that highly crystalline phases formed due to well-mixed cations in the hydrolysis and condensation steps of sol-gel synthesis and the nucleation and growth steps in the micro emulsion method. XRD patterns of both samples presented strong reflections at 32.9° (110) and 33.3° (104), corresponding to hexagonal crystal symmetry with an R3c space group. XRD results showed no trace of hexagonal La(OH)
3, corresponding to the 2θ values 27.3° (110), 28.0° (101), and 39.5° (201), which indicated the high degree of La and Co oxide incorporation into the perovskite structure achieved by both methods. The lattice parameters calculated from these XRD patterns are shown in
Table 2. The lattice parameters and cell volumes of the perovskite phase were almost similar for both samples and in agreement with the literature
[29]. The crystallite size was estimated using the Scherrer formula (D = 0.9λ/β cosθ, where d, λ, θ, and β are the crystallite size, X-ray wavelength (1.5418 Å), Bragg diffraction angle, and full width of the half maximum (FWHM) of the diffraction peak, respectively). Crystallite sizes of the samples prepared by the two different methods were calculated from the broadening of the (024) reflection of the LaCoO
3 phase at a 2θ angle of 47.5° and are tabulated in
Table 2. It was evident from the calculation that LaCoO
3 prepared by the micro emulsion method exhibited the lowest crystallite size (~23 nm) compared to the catalysts prepared by the sol-gel method and calcined at the same temperature (650℃). The larger crystallite size of the LaCoO
3 phase observed in the LCO-SP sample was indicative of the greater growth of its crystal domains. This result reveals that intended for smaller crystallite size mixings of cations by way of nucleation and growth steps is more appropriate over the following sequence such as hydrolysis, condensation, and then densification.
Fig. 1.
XRD patterns of two LaCoO3 samples prepared by sol-gel (LCO-SP) and micro emulsion (LCO-ME) methods.
Table 2.
Structure parameters of LaCoO3 synthesized by two different methods
Morphologies of LaCoO
3 synthesized with the sol-gel and micro emulsion methods are presented in
Figs. 2(a) and
(b), respectively. A comparison of these processes reveals that the micro emulsion method resulted in a more reduced particle size and homogeneity, while the powder prepared by the sol-gel method showed a relatively larger particle size and an irregular shape. The average particle size of LaCoO
3 synthesized by the micro emulsion method is about 100 nm with a spherical shape, whereas that from the sol-gel process has a broader particle size distribution. Thus, SEM results are consistent with the results of crystallite size estimation (see
Table 2). Particle size is different from crystallite size. A particle may be made up of several different crystallites. LaCoO
3 synthesized by the micro emulsion method (LCO-ME sample) showed a smaller particle and crystallite size, and hence a higher active surface area than that of the LCO-SP sample. In the sol-gel process, after hydrolysis and condensation the resultant gels have to be densified by thermal treatment. In this study, the metal-citrate precursor gel was vacuum dried at 393 K for densification. This drying stage is connected to shrinking; this step usually constrains the particle size of the final product
[30]. Thus, the nanocrystals possibly aggregated to form the large particle during the drying step, which is uncontrollable. Further heat treatment was applied for driving out the residual organic components and for perovskite phase formation. In contrast, the main strategy for the synthesis of nanoparticles in W/O micro emulsions consists of mixing two micro emulsions, one containing the metallic precursor and the other containing a precipitating agent. Upon mixing, both reactants contact each other due to droplet collisions and coalescence, and react to form precipitates of nanometric size that remain conned to the interior of micro emulsion droplets. The droplet size is governed by the ratio of water and surfactant
[31]. The particles can be made larger by increasing the droplet size. This makes the particles grow within the space hindered by the micelles, offering homogeneous nanoparticle formation
[31]. This resulted in smaller crystallites, smaller particle size, and large surface area, affording good electro catalytic properties. The sample prepared in this study by the micro emulsion method had much narrower particle size distribution (
Fig. 2(b)), owing to the isolated conditions used during nucleus formation and the subsequent grain growth for each particle.
Fig. 2.
SEM micrographs of two LaCoO3 samples prepared by (a) sol-gel and (b) microemulsion methods.
The N
2 adsorption desorption isotherm and the corresponding Barrett-Joyner-Halenda (BJH) pore size distributions are depicted in
Fig. 3 and corresponding average values regarding specific surface area, pore size and volume are presented in
Table 3. Both LCO-SP and LCO-ME samples present typical type-IV N
2 sorption isotherms with distinct H3 hysteresis loops that can linked to slit-shaped pores. The BET surface area of LCO-ME nanoparticles is 19.09 m
2 g
−1 which is much larger than that of the prepared LCO-SP nanoparticles (5.49 m
2 g
−1). The pore-size distribution curve shows porous LCO-ME nanoparticles with an average pore width ~12.39 nm. The single point total pore volume of LCO-SP and LCO-ME are reveled to be 0.016 and 0.061 cm
3 g
−1, respectively. Based on aforementioned analysis, LaCoO
3 perovskite prepared by micro emulsion method is believed to be a superior catalyst. Increasing number of pores and specific surface area in this sample is possibly able to provide more free channels for oxygen transportation and electrochemical active site for reaction, respectively, during ORR. The results from the above characterization of these materials revealed that it was possible to obtain a perovskite material with a smaller particle size as well as higher surface area and pore volume, with the micro emulsion technique than with sol-gel preparation method. Therefore, reverse micro emulsion method was a suitable route for the preparation of nanometer-sized perovskitetype solids with higher active surface areas.
Fig. 3.
Nitrogen adsorption-desorption isotherm and pore size distribution (inset) of two LaCoO3 samples prepared, by using (a) sol-gel (LCO-SP) and (b) micro emulsion (LCO-ME) method.
Table 3.
Surface properties of LaCoO3 synthesized by two different methods
The ORR and OER activities of nano-LaCoO
3 synthesized by using micro emulsion method (LCO-ME) were tested using RDE measurements. To assess the ORR catalytic activity, the samples were loaded onto GC electrodes for CV in O
2 versus Ar-saturated 0.1 M KOH solution at a scan rate of 5 mVs
−1 in the potential range −0.6 to 0.1 V. LCO-ME sample with or without carbon super P, were tested with the same mass loading. Featureless voltammetric responses were observed for both LaCoO
3 and LaCoO
3-Super P mixture-modified GC electrodes in Ar-saturated solutions within the studied potential range. In contrast, when the electrolyte solution was saturated with oxygen, the reduction current appeared as a well-defined cathodic peak around −0.47 V, suggesting pronounced electro catalytic activity of the LaCoO
3 particle-modified GC electrode in oxygen reduction (
Fig. 4). Both LaCoO
3 and the LaCoO
3-Super P mixture showed certain activity during O
2 reduction in alkaline solution, but the oxygen reduction peak of the latter was more positive. This was probably due to carbon participating in the catalysis of reduction in the latter case
[25].
Fig. 4.
Cyclic voltammetry curves of LaCoO3 and the LaCoO3-Super P mixture on glassy carbon electrodes in O2-saturated or Ar-saturated 0.1 M KOH solution. Catalyst loading is the same in both samples.
Fig. 5(a) shows linear sweep voltammograms representing the ORR in 0.1 M KOH using a thin film of LaCoO
3, Super P carbon, and the LaCoO
3-Super P mixture deposited on the rotating disk working electrode, rotated at 1600 rpm during the experiment. As a control experiment, the Ar-saturated solution displayed no noticeable reduction feature between −0.6 and 0.2 V, as expected due to the absence of oxygen. The GC electrode (the substrate onto which the film was deposited) in an O
2-saturated system led to an onset of reduction current at approximately −0.33 V, which increased gradually in the potential window studied (up to −0.6 V), as shown in
Fig. 5(a). The observed current, nevertheless, was minor when compared to the observed reduction currents for LaCoO
3, Super P, and the LaCoO
3-Super P mixture film-deposited GC electrode, indicating that the three modified electrodes acted as catalysts in ORR. A changing trend, similar to that observed in the CV curves, was also observed in the RDE curves, i.e., a relatively high ORR current appeared over the LCO-Super P mixture. Evidently, using LaCoO
3 without carbon led to a comparatively poor performance relative to even that of carbon Super P; this was probably associated with the nonhomogeneous electron transfer inside the film due to the poor electronic conductivity of perovskite oxides
[25]. For comparison, and to clearly display the onset potential, a portion of the graph was magnified and is presented in
Fig. 5(b). Remarkably, the LaCoO
3-Super P mixture showed a much more positive ORR onset potential (0.15 V) than both Super P carbon alone (−0.18 V) and LaCoO
3 (−0.15 V). In particular, the ORR current improved from LaCoO
3, to Super P, and again to the LaCoO
3-Super P mixture. Adding carbon to LaCoO
3, in turln, eads to a better performance than when carbon and LaCoO
3 were used alone. This suggested a synergy between LaCoO
3 and carbon. These synergistic effects are consistent with those reported in the literatures
[25,
28].
Fig. 5.
(a) Background (argon-saturated) and oxygen reduction reaction (ORR) polarization curves of thin films of glassy carbon, carbon Super P, LaCoO3, and the LaCoO3-Super P mixture at 1600 rpm. (b) A magnified portion of panel (a) to pinpoint the onset ORR potentials.
The ORR catalytic activity of LaCoO
3 (LCO-SP) samples which was synthesized by using sol-gel process is also observed by applying RDE measurement following the same recipe as described earlier for LaCoO
3 (LCO-ME)-Super P modified GC electrode to compare the different particle effect on the oxygen reduction activity of LaCoO
3 oxide material. For metallic catalysts such as Pt, the increase in electro-chemical surface area (ECSA) resulting from size reduction has typically been attributed as the reason for enhanced catalytic activity for the undesired reactions
[32]. However, for the oxide material enhanced ORR activity resulting from size reduction has attributed to increased conductivity and an increased number of active sites on the reaction surface
[33]. In our sample using micro emulsion method (LCO-ME), the particle size was decreased, but also the internal pore was well-developed than the sample using sol-gel method (LCO-SP). Therefore, the LCO-ME particles have comparatively higher ORR activity as shown in
Fig. 6, due to their higher electrochemical catalytic activity.
Fig. 6.
ORR polarization curves of LaCoO3 samples prepared by using sol-gel (LCO-SP) and micro emulsion (LCO-ME) method with super P with rotation rate at 1600 rpm.
Another important parameter characterizing the catalytic performance is the number of electrons (n) exchanged during ORR.
Fig. 7(a-c) shows typical ORR current densities as a function of rotation rate. All three sample polarization curves consisted of three ranges at “zero rotation”: the kinetics-controlled range (low current), the mixed kinetics and diffusion-controlled range (middle current), and the diffusion-controlled range (plateau current). However, the diffusion-controlled ORR current gradually diminished upon increasing the rotation rate from 200 to 1600 rpm, following the order Super P < LaCoO
3-Super P mixture < LaCoO
3 (
Fig. 5a-c). This behavior could be explained by a larger amount of O
2 being adsorbed in the thin film at lower rotation rates, leading to a more noticeable occurrence of an electro reduction current prior to the electro reduction current of O
2 in the bulk electrolyte
[25].
Fig. 7.
Oxygen reduction reaction (ORR) polarization curves of (a) LaCoO3, (b) Super P, and (c) the LaCoO3-Super P mixture under different rotating rates. (d) Koutecky-Levich plot based on ORR polarization curves at −0.52 V vs. the Ag|AgCl (3 M NaCl) reference electrode.
ORR reaction kinetics was examined using the Koutecky-Levich correlation:
where i corresponds to the measured current, n is the overall transferred electron number, F is the Faraday constant, C0 is the saturated concentration of oxygen in 0.1 M KOH solution, A is the geometric area of the electrode, ω is the rotating rate, DO2 is the diffusion coefcient of oxygen, ν is the kinetic viscosity of the solution, and k is the rate constant for oxygen reduction.
ORR can occur by two pathways in an alkaline electrolyte: the first is the direct reduction of O
2 to OH
− (
via a four-electron transfer), referred to as a “direct” 4e
− pathway, and the other is a “series” pathway, involving the reduction of O
2 to HO
2 (
via a two-electron transfer). The latter pathway may be followed either by a further 2e
−-reduction of peroxide, or the chemical disproportionation of peroxide
[28]. Koutecky-Levich plots are shown in
Fig. 7(d). Based on the average values calculated from different potentials, the overall electron transfer numbers of carbon Super P, LaCoO
3, and the LaCoO
3-Super P mixture were 2.4, 2.7, and 3.67, respectively. The overall electron transfer number of the LaCoO
3-carbon mixture was close to the theoretical value (4.0) for “direct” ORR in alkaline solution. On the other hand, the overall electron transfer numbers of carbon Super P and LaCoO
3 were close to 2 when used alone.
It seems that carbon in the LaCoO
3-Super P mixture plays a double role: firstly, by improving the ORR current of the catalytic layer, as shown in
Fig. 5(a), and secondly, by acting as a catalyst in the reduction of O
2 to HO
2− (perhaps further reduced to the LaCoO
3 oxide), explained by an assumed 2+2-electron transfer pathway. Thus, the ORR processes catalyzed by the LaCoO
3-Super P mixture were assumed to be dominated by a 2+2-electron transfer pathway rather than a 4-electron “direct” pathway to form the final product, OH
− .
The OER polarization curves were also examined in O
2-saturated 0.1 M KOH from 0 to 0.7 V by the RDE technique with the same electrolyte, working electrode, and scan rate as for ORR (
Fig. 8). In particular, the LaCoO
3-Super P mixture exhibited higher OER currents over the entire potential range compared to carbon Super P and the GC substrate. The OER current of carbon Super P was very low, showing that while carbon can act as a mediocre ORR catalyst, it is not at all effective for OER.
Fig. 8.
Oxygen evolution reaction polarization curves of glassy carbon, carbon Super P, and the LaCoO3-Super P mixture in O2-saturated aqueous electrolyte, 0.1 M KOH, with a sweep rate of 5 mVs−1.