- Email: [email protected]
Metalloporphyrin-modified perovskite-type oxide for the electroreduction of oxygen
Journal of Power Sources 293 (2015) 760e766
Contents lists available at ScienceDirect
Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour
Metalloporphyrin-modified perovskite-type oxide for the electroreduction of oxygen Tsukasa Nagai*, Shin-ichi Yamazaki, Masafumi Asahi, Zyun Siroma, Naoko Fujiwara, Tsutomu Ioroi Research Institute for Ubiquitous Energy Devices, National Institute of Advanced Industrial Science and Technology (AIST), 1-8-31, Midorigaoka, Ikeda, Osaka 563-8577, Japan
h i g h l i g h t s We prepared cobalt octaethylporphyrin (Co-OEP)-modified perovskite/carbon catalysts. ORR activity of perovskite/carbon was enhanced by Co-OEP-modification. RRDE measurements suggested that the 2 þ 2 electron reduction of O2 is promoted. The porphyrin plays a role as a two-electron O2 reduction catalyst to give HO 2. HO 2 is further reduced to OH by the perovskite-type oxide.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 8 March 2015 Received in revised form 25 May 2015 Accepted 1 June 2015 Available online xxx
Perovskite-type oxide-carbon (Vulcan XC72) mixture (La0.6Sr0.4Mn0.6Fe0.4O3/C) was modified by a metalloporphyrin (cobalt octaethylporphyrin: Co-OEP) having two-electron O2 reduction activity, and its electrochemical reduction activity for O2 (ORR) was investigated in an alkaline solution by rotating ring disk electrode (RRDE) voltammetry. The Co-OEP/La0.6Sr0.4Mn0.6Fe0.4O3/C catalyst showed improved ORR activity, with a positive shift of the onset potential. In addition, a decreased ring current compared to CoOEP/C suggested that the quasi-four-electron reduction of O2 was also enhanced. Further experiments showed that ORR activity was also enhanced by Co-OEP-modification of other types of carbon (Ketjenblack EC600JD, Denka Black) or perovskite-type oxide (La0.6Ca0.4Mn0.6Fe0.4O3, La0.8Sr0.2Co0.6Fe0.4O3). In the case of the addition of other porphyrin complexes (cobalt tetraphenylporphyrin (Co-TPP), iron octaethylporphyrin (Fe-OEP)) to a La0.6Sr0.4Mn0.6Fe0.4O3/C catalyst, the onset potential did not shift to the positive side due to the lower activity compared to Co-OEP. © 2015 Elsevier B.V. All rights reserved.
Keywords: Oxygen reduction reaction (ORR) Perovskite Porphyrin Alkaline Fuel cell
1. Introduction The electrocatalytic oxygen reduction reaction (ORR) is an important process in energy-conversion technologies such as fuel cells and metal-air batteries. Compared to acid-based proton exchange membrane fuel cells, alkaline fuel cells (AFCs) have attracted attention because of the possibility of faster kinetics and the potential use of less costly non-precious metal catalysts [1e5]. The mechanism of the ORR in alkaline media has been reported
* Corresponding author. E-mail addresses: [email protected] (T. Nagai), [email protected] (S.-i. Yamazaki), [email protected] (M. Asahi), [email protected] (Z. Siroma), [email protected] (N. Fujiwara), [email protected] (T. Ioroi). http://dx.doi.org/10.1016/j.jpowsour.2015.06.004 0378-7753/© 2015 Elsevier B.V. All rights reserved.
previously. In the ORR under alkaline conditions, there are two main pathways: a direct four-electron pathway (eq. (1)) and a twoby-two electron pathway (eq. (2) and (3)) which involves an alkaline-stabilized hydrogen peroxide intermediate (HO 2 ) [6e8]: O2 þ 2H2O þ 4e / 4OH
(1)
O2 þ H2 O þ 2e /HO 2 þ OH
(2)
HO 2 þ H2 O þ 2e /3OH
(3)
Additionally, hydrogen peroxide can undergo disproportionation according to eq. (4): H2O2 / O2 þ 2H2O
(4)
T. Nagai et al. / Journal of Power Sources 293 (2015) 760e766
To date, various kinds of non-Pt-based catalysts that offer the possibility of faster kinetics in alkaline solution have been proposed, such as carbon material [9,10], non-Pt-metal [3,11e13], metal oxide (ex., manganese oxide [14,15], perovskite-type oxide [16e18]), metal macrocycles [19,20]. Among them, perovskite-type oxides are considered to be an attractive candidate for use as alkaline-resistant electrocatalysts with high ORR activity. Furthermore, the enhancement of ORR activity and durability are also expected to result from optimization of the constituent elements and their composition, since the perovskite-type structure is composed of a 12-coordinate cation (mainly, rare-earth or alkaline earth) site (A-site), and a tetrahedrally-coordinated transition metal cation site (B-site). For example, in LaeMn-based oxides, it has been shown that partial substitutions at the La3þ site by alkaline-earth metals (ex. Sr2þ, Ca2þ) and at the Mn3þ site by transition metals (ex. Fe3þ) can improve both the ORR activity and durability [21,22]. To apply a perovskite-type oxide as an electrocatalyst, conductive additive carbon is often added to the perovskite electrode since perovskite possesses relatively low electrical conductivity. Carbon also plays a role as a two-electron O2 reduction catalyst to give a hydrogen peroxide intermediate (HO 2 ) in alkaline media (eq. (2)) [23,24]. Recently, the ORR activities of perovskite-type oxide thin film electrodes, for which can be neglected the contribution of carbon in ORR kinetics, have been studied, and it has been verified that perovskite thin films (LaeMn-based and LaeCo-based oxide) possess faster kinetics for 2-electron HO 2 reduction (eq. (3)) and/or disproportionation of H2O2 (eq. (4)) than for four-electron O2 reduction (eq. (1)) [25]. Poux et al. [26] also reported that the LaeMn based perovskite (La0.8Sr0.2MnO3) and LaeCo based perovskite (LaCoO3) in a carbon supported catalyst contribute to the reduction and/or disproportionation of H2O2. These studies suggest that O2 reduction reaction to HO 2 (eq. (2)) may be the ratelimiting process for a carbon-supported perovskite electrocatalyst. Therefore, it should be possible to promote the overall reaction on a carbon-supported perovskite catalyst by introducing a co-catalyst with high two-electron O2 reduction activity. As catalysts with two-electron reduction activity to generate HO 2 from O2, metal complexes such as cobalt (or iron) porphyrin and phthalocyanine are well known [27e30]. So far, catalysts that combine these complexes with manganese oxides, which cause the disproportionation of H2O2, have been reported [31]. This combined catalyst promotes the two-electron reduction reaction to give HO 2 from O2 (eq. (2)) and the subsequent chemical decomposition of H2O2 to O2 and H2O (eq. (4)) to achieve a pseudo four-electron reduction process. From this background, it should be possible to enhance the ORR activity of a perovskite-type oxide based catalyst through modification with a metallocomplex having two-electron O2 reduction activity. Up to now, catalysts using metalloporphyrin and perovskite-type oxide were reported; however, it was a heattreated-product (at 700 C) and the effects and roles of metalloporphyrins remain unclear [32]. In this study, we prepared a porphyrin complex-modified perovskite-type oxide/carbon catalyst. The ORR activity and electron number of the catalyst was investigated in alkaline media by the rotating ring disk electrode (RRDE) method. The influence of the composition of the catalysts on the ORR activity was also examined by using several types of carbon black, porphyrin complexes, and perovskite-type oxides.
761
oxide is considered to improve its catalytic activity. The polyvinyl pyrrolidone (PVP) method [33] was used to prepare nano-sized perovskite-type oxides. The perovskite-type oxides (La0.6Sr0.4Mn0.6Fe0.4O3, La0.6Ca0.4Mn0.6Fe0.4O3, and La0.8Sr0.2Co0.6Fe0.4O3) were synthesized using PVP as follows. Stoichiometric amounts of the corresponding metal nitrates (La(NO3)3$6H2O (Wako, 99.9%; 1.2990 g (for La0.6Sr0.4Mn0.6Fe0.4O3, La0.6Ca0.4Mn0.6Fe0.4O3) or 1.7320 g (for La0.8Sr0.2Co0.6Fe0.4O3)), Sr(NO3)2 (Aldrich, 99.95%; 0.4233 g (for La0.6Sr0.4Mn0.6Fe0.4O3) or 0.2117 g (for La0.8Sr0.2Co0.6Fe0.4O3)), Ca(NO3)2$4H2O (Aldrich, 99.9%; 0.4723 g), Mn(NO3)2$6H2O (Wako, 99.9%; 0.8611 g), Co(NO3)2$6H2O (Wako, 99.9%; 0.8731 g) and Fe(NO3)3$9H2O (Wako, 99.9%; 0.8080 g) and PVP (ca. 6.0 g) were dissolved in ultrapure water (100 ml). The aqueous solutions were heated to vaporize water at 100 C, and the obtained precursors were fired in air at 600 C for 6 h. For calcination, the temperature was increased at a constant rate of 20 C min1. The perovskiteecarbon mixture (perovskite/C) was prepared by mechanical milling [34]. Perovskite-type oxide (0.24 g) and carbon black (Vulcan XC72, Ketjenblack EC600JD, Denka Black; 0.08 g) were mixed in a weight ratio of 3:1 in an agate pot at a rotation speed of 400 rpm for 30 min using a planetary ball milling apparatus (FRITSCH GmbH, P-7). Porphyrin complexes were adsorbed on perovskite/C by an evaporation-to-dryness method [35]. We used three porphyrin complexes with different molecular structures (Fig. 1(a)): cobalt octaethylporphyrin (Co-OEP; Aldrich ([CoII(OEP)])), iron octaethylporphyrin (Fe-OEP; Aldrich ([FeII(OEP)(Cl)])), and cobalt tetraphenylporphyrin (Co-TPP; Aldrich ([CoII(TPP)])). The amount of porphyrin complex (0.6 mmol) on the perovskite/C support (30 mg) was constant at 20 mmol g1 support. Fig. 1(b) shows a schematic illustration of the porphyrin-modified perovskite/carbon catalyst. In the present catalyst, both a porphyrin complex and a perovskitetype oxide were supported on carbon. The porphyrin (and carbon) acts as a two-electron O2 reduction catalyst to give a hydrogen peroxide intermediate (HO 2 ), and the produced HO2 is then
2. Experimental 2.1. Sample preparation An increase in the specific surface area of a perovskite-type
Fig. 1. Schematic illustration of the porphyrin-modified perovskite/carbon catalyst (a), and structures of the porphyrin complexes which used in this study (b).
762
T. Nagai et al. / Journal of Power Sources 293 (2015) 760e766
reduced to OH (and/or decomposed to O2 and H2O) by the perovskite-type oxide. Therefore, the catalyst promotes the entire 2 þ 2 electron O2 reduction. 2.2. Sample characterization Sample characterization was performed by using X-ray powder diffraction (XRD) with Cu Ka radiation (RINT-Ultimaþ, Rigaku) and transmission electron microscopy (TEM; TITAN G2 60e300, FEI). The specific surface area was calculated by the BrunauereEmmetteTeller (BET) method (Bellsorp mini II, Bell Japan). Detailed characterization results (Scanning Transmission Electron Microscope-energy dispersive X-ray analysis (STEM-EDX) elemental mapping image, crystal diameter of the perovskitephase) of perovskite/C are given in Refs. [33,36]. 2.3. Rotating ring-disk electrode measurements The oxygen reduction activity of each catalyst was evaluated by hydrodynamic voltammetry using a rotating ring-disk electrode (a glassy carbon (GC) disk and a Pt ring). The area of GC disk electrode is 0.1256 cm2. A ring-disk electrode was polished to a mirror finish with 0.05 mm alumina powder. Aqueous suspensions (10 ml) containing ultrasonically dispersed perovskite-based catalyst (25 mg) and anion exchange resin (0.13 ml; AS-4, Tokuyama Corporation) were dropped onto the GC substrate, and the electrode was then dried in a vacuum at 60 C for 1 h. The catalyst loading amount on the GC substrate was 199 mg cm2. When we examined a catalyst that did not contain a perovskite-type oxide (e.g., carbon,
porphyrin-modified carbon) for comparison, the catalyst content on GC was fixed at 49.7 mg cm2 (i.e., quarter the amount of perovskite/C) to adjust the amount of carbon, since we mixed perovskite and carbon in a weight ratio of 3:1. In the case of a porphyrin-modified carbon (porphyrin/C) catalyst, the amount of porphyrin complex on carbon was 80 mmol g1 support to adjust the amount of porphyrin to that in the porphyrin/perovskite/C catalyst. Electrochemical measurements were performed using these electrodes in 0.1 M KOH saturated with O2 at 25 C. All measurements were conducted at 10 mV s1 using an electrochemical analyzer (Model 701E, ALS) equipped with a rotation apparatus (RRDE-3A, ALS). The rotation speed of the working electrode was 900 rpm. The data measured in an O2 atmosphere were background-corrected by the corresponding data in an Ar atmosphere. The ring electrode was polarized at 1.4 V (RHE), at which the hydrogen peroxide intermediate (HO 2 ) is completely oxidized. A reversible hydrogen electrode (RHE) and a platinum electrode were used as reference and counter electrodes, respectively. The onset potential was determined to be the potential at which the disk current value reached 0.005 mA. The fraction of hydrogen peroxide intermediate (HO 2 ) formation during oxygen reduction [37] was calculated from RRDE voltammograms. The collection efficiency (N ¼ Iring/Idisk) used in this study was determined to be 0.43 in 0.1 M KOH containing 1 mM K4[Fe(CN)6]. 3. Results and discussion Fig. 2(a) shows XRD patterns of the as-prepared perovskite-type
Fig. 2. XRD patterns of the as-prepared perovskite-type oxides (a) and perovskite/Vulcan XC72 (b), and a typical TEM image of La0.6Sr0.4Mn0.6Fe0.4O3/C (c).
T. Nagai et al. / Journal of Power Sources 293 (2015) 760e766
763
oxide (La0.6Sr0.4Mn0.6Fe0.4O3, La0.6Ca0.4Mn0.6Fe0.4O3, and La0.8Sr0.2Co0.6Fe0.4O3). All of the diffraction peaks of the as-prepared oxides can be assigned to a single phase of perovskite-type oxide, and do not indicate the presence of any crystalline impurities. The BET surface areas of prepared perovskite-type oxides were measured to be 31 m2 g1, 32 m2 g1, and 26 m2 g1 for La0.6Sr0.4Mn0.6Fe0.4O3, La0.6Ca0.4Mn0.6Fe0.4O3, and La0.8Sr0.2Co0.6Fe0.4O3, respectively [33]. We confirmed that the perovskite-type structure was maintained after ball milling treatment with carbon black (Fig. 2(b)). Fig. 2(c) shows a typical TEM image of carbon-supported La0.6Sr0.4Mn0.6Fe0.4O3 (La0.6Sr0.4Mn0.6Fe0.4O3/Vulcan XC72). The particle size of La0.6Sr0.4Mn0.6Fe0.4O3 was ca. 10e30 nm, and the perovskite-type oxide particles were uniformly dispersed on the carbon support. Fig. 3 shows the RRDE voltammograms for oxygen reduction with Co-OEP-modified La0.6Sr0.4Mn0.6Fe0.4O3/Vulcan XC72 (CoOEP/La0.6Sr0.4Mn0.6Fe0.4O3/C) in O2-saturated 0.1 M KOH solution obtained at 10 mV s1, 900 rpm, and 25 C. To benchmark the catalyst, the results for Co-OEP/Vulcan XC72, La0.6Sr0.4Mn0.6Fe0.4O3/ Vulcan XC72, carbon (Vulcan XC72) and commercially available Pt/ C (TEC10E50E (Tanaka Kikinzoku Kogyo), loading amount: 63.4 mg cm2) are also shown. The onset potential of Co-OEP/ La0.6Sr0.4Mn0.6Fe0.4O3/C (C) is shifted to the positive side and the ORR current at the disk electrode was drastically increased compared with that of the La0.6Sr0.4Mn0.6Fe0.4O3/C (-). Therefore, the ORR activity of the La0.6Sr0.4Mn0.6Fe0.4O3/C catalyst is considered to be improved by the addition of Co-OEP as a co-catalyst. The formation of a hydrogen peroxide intermediate (HO 2 ) was confirmed by the ring electrode current during oxygen reduction.
For Co-OEP/C with mainly two-electron O2 reduction activity (:), increased ring current which suggested the production of HO 2 was clearly observed from ca. 0.82 V (vs RHE), corresponding to oxygen reduction. The ring current of the Co-OEP/La0.6Sr0.4Mn0.6Fe0.4O3/C catalyst was significantly lower than that of Co-OEP/C. Fig. 3(c) presents the fraction of HO 2 production during oxygen reduction (XH2O2), as estimated from RRDE voltammograms [37]. The values for catalysts containing perovskite (La0.6Sr0.4Mn0.6Fe0.4O3/C, CoOEP/La0.6Sr0.4Mn0.6Fe0.4O3/C) were much lower than those for catalysts without perovskite (carbon, Co-OEP/C), indicating that O2 is reduced more completely to OH at the electrode. These RRDE results for the present catalysts suggest that the HO 2 generated by Co-OEP (and carbon) is efficiently reduced to OH by La0.6Sr0.4Mn0.6Fe0.4O3. In order to investigate the O2 reduction mechanism of the present Co-OEP/La0.6Sr0.4Mn0.6Fe0.4O3/C catalyst, RRDE measurements were also performed with various amounts of Co-OEP/ La0.6Sr0.4Mn0.6Fe0.4O3/C (Fig. 4). The catalyst contents on a GC disk electrode were 199 mg cm2 (C), 99.5 mg cm2 (-), and 49.7 mg cm2 (:). Co-OEP/La0.6Sr0.4Mn0.6Fe0.4O3/C showed a lower onset potential and increased ring current with a decrease in the loading amount on GC. Thus, HO 2 production during oxygen reduction was increased. Previous studies on Pt/C catalysts [38e40] have described a decrease in ORR activity (especially, increased H2O2 production) with a decrease in catalyst loading. We discuss here how the ORR on Pt/C catalysts occurs via a two-by-two electron pathway through a H2O2 intermediate as well as via a direct four-electron pathway. After two-electron O2 reduction on a Pt
Fig. 3. RRDE voltammograms in O2-saturated 0.1 M KOH at 25 C ((a) disk electrode, (b) ring electrode) and fractions of HO 2 production during oxygen reduction (c) for CoOEP/La0.6Sr0.4Mn0.6Fe0.4O3/Vulcan XC72 (C), La0.6Sr0.4Mn0.6Fe0.4O3/Vulcan XC72 (-), Co-OEP/Vulcan XC72 (:), and Vulcan XC72 (solid line). RRDE data for commercially available Pt/C catalyst (TEC10E50E, loading amount: 63.4 mg cm2) is also shown in (a) as a broken line. Scan rate and rotation speed were 10 mV s1 and 900 rpm, respectively.
Fig. 4. RRDE voltammograms of CoOEP/La0.6Sr0.4Mn0.6Fe0.4O3/Vulcan XC72 in O2saturated 0.1 M KOH at 25 C ((a) disk electrode, (b) ring electrode) and fractions of HO 2 production during oxygen reduction (c). Measurements were performed with various amounts of Co-OEP/La0.6Sr0.4Mn0.6Fe0.4O3/C catalyst (catalyst content on GC: 199 (C), 99.5 (-), and 49.7 (:) mg cm2). Scan rate and rotation speed were 10 mV s1 and 900 rpm, respectively.
764
T. Nagai et al. / Journal of Power Sources 293 (2015) 760e766
Table 1 BET specific surface area for La0.6Sr0.4Mn0.6Fe0.4O3/carbon (carbon; Vulcan XC72, Ketjenblack EC600JD, Denka Black). BET surface area/m2 g1 LSMF/Vulcan XC72 LSMF/Ketjenblack EC600JD LSMF/Denka Black Vulcan XC72 Ketjenblack EC600JD Denka Black
76 209 50 232 1282 63
surface, desorbed H2O2 is re-adsorbed and reduced on a nearby Pt particle, or diffuses away into the bulk solution. When Pt/C catalysts are highly agglomerated on GC (interparticle distance of Pt is short), the H2O2 molecules can easily reach nearby Pt particles, and an increased disk current and decreased ring current is observed in RRDE measurements. In these terms, the present Co-OEP/ La0.6Sr0.4Mn0.6Fe0.4O3/C catalyst is considered to show a consecutive two-by-two electron pathway for O2 reduction; two-electron O2 reduction to HO 2 occurs on Co-OEP (and carbon), and HO2 is captured and reduced on a nearby La0.6Sr0.4Mn0.6Fe0.4O3 particle. O2 is further reduced to OH due to the closeness between Co-OEP and La0.6Sr0.4Mn0.6Fe0.4O3, which enhances the likelihood of capturing HO 2. We also prepared the Co-OEP/La0.6Sr0.4Mn0.6Fe0.4O3/carbon catalyst using other types of conductive carbon (Ketjenblack EC600JD and Denka Black), and performed RRDE measurements. The BET specific surface area of La0.6Sr0.4Mn0.6Fe0.4O3/carbon is summarized in Table 1. The surface area of La0.6Sr0.4Mn0.6Fe0.4O3/C varied depending on the specific surface area of carbon. Fig. 5 shows hydrodynamic voltammograms of La0.6Sr0.4Mn0.6Fe0.4O3/C and Co-OEP/La0.6Sr0.4Mn0.6Fe0.4O3/C. The onset potential for each La0.6Sr0.4Mn0.6Fe0.4O3/C was approximately the same. Furthermore, under modification by Co-OEP, increased catalytic activity was confirmed regardless of the type of conductive carbon. The reduction currents for the Ketjenblack EC600JD-supported sample (La0.6Sr0.4Mn0.6Fe0.4O3/Ketjenblack, and Co-OEP/La0.6Sr0.4Mn0.6Fe0.4O3/Ketjenblack) were slightly lower than those for the samples with Vulcan XC72 and Denka Black. One of the possible causes is as follows. For the sample supported on Ketjenblack EC600JD, the catalyst loading with respect to the surface area of carbon becomes small due to the enormously high specific surface area of
Fig. 5. Hydrodynamic voltammograms of La0.6Sr0.4Mn0.6Fe0.4O3/carbon (LSMF/C, broken line) and Co-OEP/La0.6Sr0.4Mn0.6Fe0.4O3/carbon (Co-OEP/LSMF/C, solid line) in O2-saturated 0.1 M KOH solution obtained at 10 mV s1, 900 rpm, and 25 C (Vulcan XC72 (C), Denka Black (-), and Ketjenblack EC600JD (:)).
Ketjenblack. Thus, uncovered and inactive areas remain over the carbon surface, and it may be difficult for perovskite oxide to effectively capture the HO 2 that participates in the ORR reaction. To compare the ORR performances of porphyrin/perovskite/ carbon with different structures of porphyrin complex or compositions of perovskite oxide, we prepared catalysts with several types of porphyrin and perovskite. First, porphyrin complex with two different structures (iron octaethylporphyrin (Fe-OEP) and cobalt tetraphenylporphyrin (Co-TPP)) were added to the La0.6Sr0.4Mn0.6Fe0.4O3/Vulcan XC72 support, and their ORR activities were investigated (Fig. 6-I). All of the complexes added to La0.6Sr0.4Mn0.6Fe0.4O3/Vulcan XC72 showed a two-electron reduction activity of oxygen to HO 2 as shown in Fig. 6-II. The onset potentials for each catalyst are summarized in Table 2. As described in Fig. 3, a large potential shift was observed when the La0.6Sr0.4Mn0.6Fe0.4O3/Vulcan XC72 catalyst was modified by CoOEP (C). For the sample modified by an Fe-OEP complex (-), which has the same ligands as Co-OEP, the ORR activity was not enhanced. This can explain the low ORR activity of Fe-OEP/C (onset potential: ca. 0.70 V (vs RHE)) compared to Co-OEP/C (onset potential: ca. 0.83 V (vs RHE)). Along with the decreased ORR activity, the production of HO 2 by Fe-OEP/Vulcan XC72 was observed from significantly negative potential. Furthermore, the onset potential for Fe-OEP/C is lower than that for La0.6Sr0.4Mn0.6Fe0.4O3/C (ca. 0.85 V (vs RHE)), and therefore, the oxygen reduction on La0.6Sr0.4Mn0.6Fe0.4O3/C catalyst is less promoted by modification with an Fe-OEP complex. In the case of modification with Co-TPP complex, which shows relatively high activity on La0.6Sr0.4Mn0.6Fe0.4O3/Vulcan XC72 (:), the ORR activity was not increased as drastically as that of the CoOEP-modified catalyst. The reason remains unclear, however, the principal reason for this difference is considered to be the difference in the molecular structures of Co-TPP and Co-OEP. Co-TPP complex has a bulky structure with a phenyl group attached to the porphyrin ring, while Co-OEP has a relatively planar structure. It has been reported that the amount of Co-TPP adsorbed on Vulcan XC72 is significantly lower than that of Co-OEP under equilibrium adsorption conditions, almost due to the bulkiness and orientation of the substituents of Co-TPP [41]. The amount of the porphyrin adsorbed on carbon reflects the strength of a specific interaction between the porphyrin and carbon (i.e. p-p interaction between the porphyrin ring and carbon surface). Strong interaction between Co-OEP and carbon would be favorable for the electron transfer from an electrode to the molecule. The possible difference in the interaction might be related to the difference in the activity. For a La-based perovskiteetype oxide, enhanced catalytic activity has been achieved by doping the La3þ site with alkaline-earth metals (Sr, Ca), and by including Mn or Co at a transition metal cation site. To investigate the effect of modification with a porphyrin complex with different compositions of perovskite, we also prepared La0.6Ca0.4Mn0.6Fe0.4O3/Vulcan XC72 and La0.8Sr0.2Co0.6Fe0.4O3/Vulcan XC72, and modified these catalysts with CoOEP complex. Fig. 7(a) shows the hydrodynamic voltammograms for Co-OEP-modified perovskite/Vulcan XC72 catalysts. The disk currents of the La0.6Ca0.4Mn0.6Fe0.4O3/C and La0.8Sr0.2Co0.6Fe0.4O3/C without Co-OEP were comparable to that of La0.6Sr0.4Mn0.6Fe0.4O3/ C. For La0.6Ca0.4Mn0.6Fe0.4O3/C (-, solid line) and the La0.8Sr0.2Co0.6Fe0.4O3/C (:, solid line), the addition of Co-OEP produced a positive shift of the onset potential, as seen with Co-OEP/ La0.6Sr0.4Mn0.6Fe0.4O3/C (C, solid line). Although all of the prepared samples exhibited enhanced O2 reduction activity, the ring current (Fig. 7(b)) for Co-OEP/La0.8Sr0.2Co0.6Fe0.4O3/C was higher than those for LaeMn-based catalysts, and large fractions of HO 2 production (XH2O2) were observed (Fig. 7(c)). This is due to the large XH2O2 (and ring current) for the La0.8Sr0.2Co0.6Fe0.4O3/C, which indicates that
T. Nagai et al. / Journal of Power Sources 293 (2015) 760e766
765
Fig. 6. I: RRDE voltammograms for porphyrin-modified La0.6Sr0.4Mn0.6Fe0.4O3/Vulcan XC72 (porphyrin/LSMF/C). II: RRDE voltammograms for porphyrin-modified Vulcan XC72 (porphyrin/C). Each measurement was conducted in 0.1 M KOH saturated with O2 at 10 mV s1, 900 rpm, and 25 C ((a) disk electrode, (b) ring electrode). The fraction of HO 2 production during oxygen reduction (XH2O2) is also shown in (c).
HO by La0.8Sr0.2Co0.6Fe0.4O3 2 is not efficiently reduced to OH compared to LaeMn-based perovskite (La0.6Sr0.4Mn0.6Fe0.4O3 and La0.6Ca0.4Mn0.6Fe0.4O3). To obtain high four-electron O2 reduction activity on the Co-OEP-modified perovskite/C catalyst, it is preferable to apply perovskite with a higher HO 2 reduction activity (i.e., perovskite/C showing a small XH2O2 value). Based on these experimental results, the ORR activities of the perovskite-type oxides prepared in this study were enhanced by the addition of a Co-OEP complex. In addition, the HO 2 reduction activity of the perovskitetype oxide plays an important role in the enhancement of the quasi-four-electron reduction of O2.
4. Conclusion In this study, we prepared a cobalt octaethylporphyrin (CoOEP)-modified perovskite-type oxide/carbon catalysts (Co-OEP/ perovskite/C, perovskite: La0.6Sr0.4Mn0.6Fe0.4O3, La0.6Ca0.4Mn0.6Fe0.4O3, and La0.8Sr0.2Co0.6Fe0.4O3). The electrocatalytic reduction activities of oxygen (ORR) of the prepared catalysts were investigated in alkaline media by rotating ring disk electrode (RRDE)
Table 2 Onset potential for porphyrin-modified La0.6Sr0.4Mn0.6Fe0.4O3/Vulcan XC72 (porphyrin/LSMF/C) and porphyrin-modified carbon (porphyrin/C). The onset potential is the potential at which the disk current reaches 0.005 mA. Onset potential/V vs RHE Co-OEP/C Fe-OEP/C Co-TPP/C Co-OEP/LSMF/C Fe-OEP/LSMF/C Co-TPP/LSMF/C LSMF/C Vulcan XC72
0.829 0.702 0.824 0.906 0.879 0.868 0.854 0.677
Fig. 7. Hydrodynamic voltammograms for perovskite-type oxide/C and Co-OEPmodified perovskite-type oxide/C catalysts in O2-saturated 0.1 M KOH solution obtained at 10 mV s1, 900 rpm, and 25 C (C; La0.6Sr0.4Mn0.6Fe0.4O3/C (LSMF/C), :; La0.6Ca0.4Mn0.6Fe0.4O3/C (LCMF/C), -; La0.8Ca0.2Mn0.6Fe0.4O3/C (LSCF/C). The solid and broken lines represent the samples with and without Co-OEP, respectively.
766
T. Nagai et al. / Journal of Power Sources 293 (2015) 760e766
voltammetry. The onset potential of Co-OEP/perovskite/C was shifted to the positive side compared with that of perovskite/C at a disk electrode by the addition of Co-OEP complex with twoelectron O2 reduction activity to the perovskite/C. A decreased ring current and a small fraction of HO 2 formation relative to CoOEP/C were also confirmed. Further measurements with various amounts of catalyst were similar to the case with Pt/C catalyst; a lower onset potential and increased ring current with a decrease in the loading amount on GC. It is considered that quasi-four-electron O2 reduction is promoted on Co-OEP (O2 reduction to HO 2 ) and perovskite (HO 2 reduction). With regard to modification of the La0.6Sr0.4Mn0.6Fe0.4O3/C catalyst with other porphyrin complex (FeOEP and Co-TPP), the ORR activity of porphyrin/perovskite/C catalyst is varied with difference of their activity. Acknowledgment The authors are grateful to Tokuyama Corporation for providing the anion exchange resin (AS-4). We also thank Dr. Mitsunori Kitta (AIST) for his assistance with TEM. References [1] M. Savy, Electrochim. Acta. 3 (1968) 1359e1376. [2] H.A. Gasteiger, S.S. Kocha, B. Sompalli, F.T. Wagner, Appl. Catal. B 56 (2005) 9e35. [3] J.S. Spendelow, A. Wieckwski, Phys. Chem. Chem. Phys. 9 (2007) 2654e2675. [4] K. Asazawa, K. Yamada, H. Tanaka, A. Oka, M. Taniguchi, T. Kobayashi, Angew. Chem. Int. Ed. 46 (2007) 8024e8027. [5] D.A. Slanac, A. Lie, J.A. Paulson, K.J. Stevenson, K.P. Johnston, J. Phys. Chem. C 116 (2012) 11032e11039. [6] L.N. Nekrasov, L. Müller, D. Akad, Nauk. SSSR 149 (1963) 1107. [7] A. Damjanovic, M.A. Genshaw, J.O.'M. Bockris, J. Electrochem. Soc. 114 (1967) 1107e1112. [8] A. Damjanovic, M.A. Genshaw, J.O.'M. Bockris, J. Electroanal. Chem. Interfacial Electrochem. 15 (1967) 173e180. [9] I. Morcos, E. Yeager, Electrochim. Acta 15 (1970) 953e975. [10] P.S.D. Brito, C.A.C. Sequeira, J. Power Sources 52 (1994) 1e16. [11] V.S. Bagotzky, N.A. Shumilova, G.P. Samoilov, E.I. Khrushcheva, Electrochim. Acta 17 (1972) 1625e1635. [12] D. Sepa, M. Vojnovic, A. Damjanovic, Electrochim. Acta 15 (1970) 1355e1366. [13] B.B. Blizanac, P.N. Ross, N.M. Markovic, J. Phys. Chem. B 110 (2006) 4735e4741. [14] P. Zoltowski, D.M. Drazic, L. Vorkapic, J. Appl. Electrochem. 3 (1973) 271e283. [15] L. Mao, T. Sotomura, K. Nakatsu, N. Koshiba, D. Zhang, T. Ohsaka,
J. Electrochem. Soc. 149 (2002) A504eA507. [16] D.B. Meadowcroft, Nature 226 (1970) 847e848. [17] Y. Shimizu, K. Uemura, H. Matsuda, N. Miura, N. Yamazoe, J. Electrochem. Soc. 137 (1990) 3430e3433. [18] J. Suntivich, H.A. Gasteiger, N. Yabuuchi, H. Nakanishi, J.B. Goodenough, Y. Shao-Horn, Nat. Chem. 3 (2011) 546e550. [19] A. Bettelheim, B.A. White, R.W. Murray, J. Electroanal. Chem. Interfacial Electrochem. 217 (1987) 271e286. [20] J. Zagal, M. Paez, A.A. Tanaka, J.R. Dossantos, C.A. Linkous, J. Electroanal. Chem. 339 (1992) 13e30. [21] J. Tulloch, S.W. Donne, J. Power Sources 188 (2009) 359e366. [22] M. Yuasa, N. Yamazoe, K. Shimanoe, J. Electrochem, Soc. 158 (2011) A411eA416. [23] K. Kinoshita, Carbon-electrochemical and Physicochemical Properties, Wiley, New York, USA, 1988. [24] C. Song, J. Zhang, Electrocatalytic oxygen reduction reaction, in: J. Zhang (Ed.), PEM Fuel Cell Electrocatalysts and Catalyst Layers: Fundamentals and Applications, Springer, London, 2008, pp. 89e134. [25] Y. Miyahara, K. Miyazaki, T. Fukutsuka, T. Abe, J. Electrochem. Soc. 161 (2014) F694eF697. [26] T. Poux, F.S. Napolskiy, T. Dintzer, G. Kerangueven, S.Y. Istomin, G.A. Tsirlina, E.V. Antipov, E.R. Savinova, Catal. Today 189 (2012) 83e92. [27] F. Van Den Brink, W. Visscher, E. Barendrecht, J. Electroanal. Chem. Interfacial Electrochem. 157 (1983) 283e304. [28] F. Van Den Brink, W. Visscher, E. Barendrecht, J. Electroanal. Chem. Interfacial Electrochem. 157 (1983) 305e318. [29] J. Ouyang, K. Shigehara, A. Yamada, F.C. Anson, J. Electroanal. Chem. Interfacial Electrochem. 297 (1991) 489e497. [30] O. El Mouahid, C. Coutanceau, E.M. Belgsir, P. Crouigneau, J.M. Leger, C. Lamy, J. Electroanal. Chem. 426 (1997) 117e123. [31] L. Mao, K. Arihara, T. Sotomura, T. Ohsaka, Electrochim. Acta 49 (2004) 2515e2521. [32] Y. Kiros, Int. J. Electrochem. Sci. 2 (2007) 285e300. [33] T. Nagai, N. Fujiwara, M. Asahi, S. Yamazaki, Z. Siroma, T. Ioroi, J. Asian Ceram. Soc. 2 (2014) 329e332. [34] N. Fujiwara, Z. Siroma, S. Yamazaki, M. Asahi, T. Nagai, T. Ioroi, in: Proceedings of 80th Meeting of the Electrochemical Society of Japan 3A21, 2013. [35] S. Yamazaki, Y. Yamada, S. Takeda, M. Goto, T. Ioroi, Z. Siroma, K. Yasuda, Phys. Chem. Chem. Phys. 12 (2010) 8968e8976. [36] T. Nagai, N. Fujiwara, M. Kitta, M. Asahi, S. Yamazaki, Z. Siroma, T. Ioroi, J. Fuel Cell. Sci. Tech. 12 (2015), 021007-1e021007-5. [37] U.A. Paulus, T.J. Schmidt, H.A. Gasteiger, R.J. Behm, J. Electroanal. Chem. 495 (2001) 134e145. [38] M. Inaba, H. Yamada, J. Tokunaga, A. Tasaka, Electrochem. Solid-State Lett. 7 (2004) A474eA476. [39] K.J.J. Mayrhofer, D. Strmcnik, B.B. Blizanac, V. Stamenkovic, M. Arenz, N.M. Markovic, Electrochim. Acta 53 (2008) 3181e3188. [40] E. Fabbri, S. Taylor, A. Rabis, P. Levecque, O. Conrad, R. Kctz, T.J. Schmidt, Chem. Cat. Chem. 6 (2014) 1410e1418. [41] S. Yamazaki, Y. Yamada, T. Ioroi, N. Fujiwara, Z. Siroma, K. Yasuda, Y. Miyazaki, J. Electroanal. Chem. 576 (2005) 253e259.
Contents lists available at ScienceDirect
Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour
Metalloporphyrin-modified perovskite-type oxide for the electroreduction of oxygen Tsukasa Nagai*, Shin-ichi Yamazaki, Masafumi Asahi, Zyun Siroma, Naoko Fujiwara, Tsutomu Ioroi Research Institute for Ubiquitous Energy Devices, National Institute of Advanced Industrial Science and Technology (AIST), 1-8-31, Midorigaoka, Ikeda, Osaka 563-8577, Japan
h i g h l i g h t s We prepared cobalt octaethylporphyrin (Co-OEP)-modified perovskite/carbon catalysts. ORR activity of perovskite/carbon was enhanced by Co-OEP-modification. RRDE measurements suggested that the 2 þ 2 electron reduction of O2 is promoted. The porphyrin plays a role as a two-electron O2 reduction catalyst to give HO 2. HO 2 is further reduced to OH by the perovskite-type oxide.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 8 March 2015 Received in revised form 25 May 2015 Accepted 1 June 2015 Available online xxx
Perovskite-type oxide-carbon (Vulcan XC72) mixture (La0.6Sr0.4Mn0.6Fe0.4O3/C) was modified by a metalloporphyrin (cobalt octaethylporphyrin: Co-OEP) having two-electron O2 reduction activity, and its electrochemical reduction activity for O2 (ORR) was investigated in an alkaline solution by rotating ring disk electrode (RRDE) voltammetry. The Co-OEP/La0.6Sr0.4Mn0.6Fe0.4O3/C catalyst showed improved ORR activity, with a positive shift of the onset potential. In addition, a decreased ring current compared to CoOEP/C suggested that the quasi-four-electron reduction of O2 was also enhanced. Further experiments showed that ORR activity was also enhanced by Co-OEP-modification of other types of carbon (Ketjenblack EC600JD, Denka Black) or perovskite-type oxide (La0.6Ca0.4Mn0.6Fe0.4O3, La0.8Sr0.2Co0.6Fe0.4O3). In the case of the addition of other porphyrin complexes (cobalt tetraphenylporphyrin (Co-TPP), iron octaethylporphyrin (Fe-OEP)) to a La0.6Sr0.4Mn0.6Fe0.4O3/C catalyst, the onset potential did not shift to the positive side due to the lower activity compared to Co-OEP. © 2015 Elsevier B.V. All rights reserved.
Keywords: Oxygen reduction reaction (ORR) Perovskite Porphyrin Alkaline Fuel cell
1. Introduction The electrocatalytic oxygen reduction reaction (ORR) is an important process in energy-conversion technologies such as fuel cells and metal-air batteries. Compared to acid-based proton exchange membrane fuel cells, alkaline fuel cells (AFCs) have attracted attention because of the possibility of faster kinetics and the potential use of less costly non-precious metal catalysts [1e5]. The mechanism of the ORR in alkaline media has been reported
* Corresponding author. E-mail addresses: [email protected] (T. Nagai), [email protected] (S.-i. Yamazaki), [email protected] (M. Asahi), [email protected] (Z. Siroma), [email protected] (N. Fujiwara), [email protected] (T. Ioroi). http://dx.doi.org/10.1016/j.jpowsour.2015.06.004 0378-7753/© 2015 Elsevier B.V. All rights reserved.
previously. In the ORR under alkaline conditions, there are two main pathways: a direct four-electron pathway (eq. (1)) and a twoby-two electron pathway (eq. (2) and (3)) which involves an alkaline-stabilized hydrogen peroxide intermediate (HO 2 ) [6e8]: O2 þ 2H2O þ 4e / 4OH
(1)
O2 þ H2 O þ 2e /HO 2 þ OH
(2)
HO 2 þ H2 O þ 2e /3OH
(3)
Additionally, hydrogen peroxide can undergo disproportionation according to eq. (4): H2O2 / O2 þ 2H2O
(4)
T. Nagai et al. / Journal of Power Sources 293 (2015) 760e766
To date, various kinds of non-Pt-based catalysts that offer the possibility of faster kinetics in alkaline solution have been proposed, such as carbon material [9,10], non-Pt-metal [3,11e13], metal oxide (ex., manganese oxide [14,15], perovskite-type oxide [16e18]), metal macrocycles [19,20]. Among them, perovskite-type oxides are considered to be an attractive candidate for use as alkaline-resistant electrocatalysts with high ORR activity. Furthermore, the enhancement of ORR activity and durability are also expected to result from optimization of the constituent elements and their composition, since the perovskite-type structure is composed of a 12-coordinate cation (mainly, rare-earth or alkaline earth) site (A-site), and a tetrahedrally-coordinated transition metal cation site (B-site). For example, in LaeMn-based oxides, it has been shown that partial substitutions at the La3þ site by alkaline-earth metals (ex. Sr2þ, Ca2þ) and at the Mn3þ site by transition metals (ex. Fe3þ) can improve both the ORR activity and durability [21,22]. To apply a perovskite-type oxide as an electrocatalyst, conductive additive carbon is often added to the perovskite electrode since perovskite possesses relatively low electrical conductivity. Carbon also plays a role as a two-electron O2 reduction catalyst to give a hydrogen peroxide intermediate (HO 2 ) in alkaline media (eq. (2)) [23,24]. Recently, the ORR activities of perovskite-type oxide thin film electrodes, for which can be neglected the contribution of carbon in ORR kinetics, have been studied, and it has been verified that perovskite thin films (LaeMn-based and LaeCo-based oxide) possess faster kinetics for 2-electron HO 2 reduction (eq. (3)) and/or disproportionation of H2O2 (eq. (4)) than for four-electron O2 reduction (eq. (1)) [25]. Poux et al. [26] also reported that the LaeMn based perovskite (La0.8Sr0.2MnO3) and LaeCo based perovskite (LaCoO3) in a carbon supported catalyst contribute to the reduction and/or disproportionation of H2O2. These studies suggest that O2 reduction reaction to HO 2 (eq. (2)) may be the ratelimiting process for a carbon-supported perovskite electrocatalyst. Therefore, it should be possible to promote the overall reaction on a carbon-supported perovskite catalyst by introducing a co-catalyst with high two-electron O2 reduction activity. As catalysts with two-electron reduction activity to generate HO 2 from O2, metal complexes such as cobalt (or iron) porphyrin and phthalocyanine are well known [27e30]. So far, catalysts that combine these complexes with manganese oxides, which cause the disproportionation of H2O2, have been reported [31]. This combined catalyst promotes the two-electron reduction reaction to give HO 2 from O2 (eq. (2)) and the subsequent chemical decomposition of H2O2 to O2 and H2O (eq. (4)) to achieve a pseudo four-electron reduction process. From this background, it should be possible to enhance the ORR activity of a perovskite-type oxide based catalyst through modification with a metallocomplex having two-electron O2 reduction activity. Up to now, catalysts using metalloporphyrin and perovskite-type oxide were reported; however, it was a heattreated-product (at 700 C) and the effects and roles of metalloporphyrins remain unclear [32]. In this study, we prepared a porphyrin complex-modified perovskite-type oxide/carbon catalyst. The ORR activity and electron number of the catalyst was investigated in alkaline media by the rotating ring disk electrode (RRDE) method. The influence of the composition of the catalysts on the ORR activity was also examined by using several types of carbon black, porphyrin complexes, and perovskite-type oxides.
761
oxide is considered to improve its catalytic activity. The polyvinyl pyrrolidone (PVP) method [33] was used to prepare nano-sized perovskite-type oxides. The perovskite-type oxides (La0.6Sr0.4Mn0.6Fe0.4O3, La0.6Ca0.4Mn0.6Fe0.4O3, and La0.8Sr0.2Co0.6Fe0.4O3) were synthesized using PVP as follows. Stoichiometric amounts of the corresponding metal nitrates (La(NO3)3$6H2O (Wako, 99.9%; 1.2990 g (for La0.6Sr0.4Mn0.6Fe0.4O3, La0.6Ca0.4Mn0.6Fe0.4O3) or 1.7320 g (for La0.8Sr0.2Co0.6Fe0.4O3)), Sr(NO3)2 (Aldrich, 99.95%; 0.4233 g (for La0.6Sr0.4Mn0.6Fe0.4O3) or 0.2117 g (for La0.8Sr0.2Co0.6Fe0.4O3)), Ca(NO3)2$4H2O (Aldrich, 99.9%; 0.4723 g), Mn(NO3)2$6H2O (Wako, 99.9%; 0.8611 g), Co(NO3)2$6H2O (Wako, 99.9%; 0.8731 g) and Fe(NO3)3$9H2O (Wako, 99.9%; 0.8080 g) and PVP (ca. 6.0 g) were dissolved in ultrapure water (100 ml). The aqueous solutions were heated to vaporize water at 100 C, and the obtained precursors were fired in air at 600 C for 6 h. For calcination, the temperature was increased at a constant rate of 20 C min1. The perovskiteecarbon mixture (perovskite/C) was prepared by mechanical milling [34]. Perovskite-type oxide (0.24 g) and carbon black (Vulcan XC72, Ketjenblack EC600JD, Denka Black; 0.08 g) were mixed in a weight ratio of 3:1 in an agate pot at a rotation speed of 400 rpm for 30 min using a planetary ball milling apparatus (FRITSCH GmbH, P-7). Porphyrin complexes were adsorbed on perovskite/C by an evaporation-to-dryness method [35]. We used three porphyrin complexes with different molecular structures (Fig. 1(a)): cobalt octaethylporphyrin (Co-OEP; Aldrich ([CoII(OEP)])), iron octaethylporphyrin (Fe-OEP; Aldrich ([FeII(OEP)(Cl)])), and cobalt tetraphenylporphyrin (Co-TPP; Aldrich ([CoII(TPP)])). The amount of porphyrin complex (0.6 mmol) on the perovskite/C support (30 mg) was constant at 20 mmol g1 support. Fig. 1(b) shows a schematic illustration of the porphyrin-modified perovskite/carbon catalyst. In the present catalyst, both a porphyrin complex and a perovskitetype oxide were supported on carbon. The porphyrin (and carbon) acts as a two-electron O2 reduction catalyst to give a hydrogen peroxide intermediate (HO 2 ), and the produced HO2 is then
2. Experimental 2.1. Sample preparation An increase in the specific surface area of a perovskite-type
Fig. 1. Schematic illustration of the porphyrin-modified perovskite/carbon catalyst (a), and structures of the porphyrin complexes which used in this study (b).
762
T. Nagai et al. / Journal of Power Sources 293 (2015) 760e766
reduced to OH (and/or decomposed to O2 and H2O) by the perovskite-type oxide. Therefore, the catalyst promotes the entire 2 þ 2 electron O2 reduction. 2.2. Sample characterization Sample characterization was performed by using X-ray powder diffraction (XRD) with Cu Ka radiation (RINT-Ultimaþ, Rigaku) and transmission electron microscopy (TEM; TITAN G2 60e300, FEI). The specific surface area was calculated by the BrunauereEmmetteTeller (BET) method (Bellsorp mini II, Bell Japan). Detailed characterization results (Scanning Transmission Electron Microscope-energy dispersive X-ray analysis (STEM-EDX) elemental mapping image, crystal diameter of the perovskitephase) of perovskite/C are given in Refs. [33,36]. 2.3. Rotating ring-disk electrode measurements The oxygen reduction activity of each catalyst was evaluated by hydrodynamic voltammetry using a rotating ring-disk electrode (a glassy carbon (GC) disk and a Pt ring). The area of GC disk electrode is 0.1256 cm2. A ring-disk electrode was polished to a mirror finish with 0.05 mm alumina powder. Aqueous suspensions (10 ml) containing ultrasonically dispersed perovskite-based catalyst (25 mg) and anion exchange resin (0.13 ml; AS-4, Tokuyama Corporation) were dropped onto the GC substrate, and the electrode was then dried in a vacuum at 60 C for 1 h. The catalyst loading amount on the GC substrate was 199 mg cm2. When we examined a catalyst that did not contain a perovskite-type oxide (e.g., carbon,
porphyrin-modified carbon) for comparison, the catalyst content on GC was fixed at 49.7 mg cm2 (i.e., quarter the amount of perovskite/C) to adjust the amount of carbon, since we mixed perovskite and carbon in a weight ratio of 3:1. In the case of a porphyrin-modified carbon (porphyrin/C) catalyst, the amount of porphyrin complex on carbon was 80 mmol g1 support to adjust the amount of porphyrin to that in the porphyrin/perovskite/C catalyst. Electrochemical measurements were performed using these electrodes in 0.1 M KOH saturated with O2 at 25 C. All measurements were conducted at 10 mV s1 using an electrochemical analyzer (Model 701E, ALS) equipped with a rotation apparatus (RRDE-3A, ALS). The rotation speed of the working electrode was 900 rpm. The data measured in an O2 atmosphere were background-corrected by the corresponding data in an Ar atmosphere. The ring electrode was polarized at 1.4 V (RHE), at which the hydrogen peroxide intermediate (HO 2 ) is completely oxidized. A reversible hydrogen electrode (RHE) and a platinum electrode were used as reference and counter electrodes, respectively. The onset potential was determined to be the potential at which the disk current value reached 0.005 mA. The fraction of hydrogen peroxide intermediate (HO 2 ) formation during oxygen reduction [37] was calculated from RRDE voltammograms. The collection efficiency (N ¼ Iring/Idisk) used in this study was determined to be 0.43 in 0.1 M KOH containing 1 mM K4[Fe(CN)6]. 3. Results and discussion Fig. 2(a) shows XRD patterns of the as-prepared perovskite-type
Fig. 2. XRD patterns of the as-prepared perovskite-type oxides (a) and perovskite/Vulcan XC72 (b), and a typical TEM image of La0.6Sr0.4Mn0.6Fe0.4O3/C (c).
T. Nagai et al. / Journal of Power Sources 293 (2015) 760e766
763
oxide (La0.6Sr0.4Mn0.6Fe0.4O3, La0.6Ca0.4Mn0.6Fe0.4O3, and La0.8Sr0.2Co0.6Fe0.4O3). All of the diffraction peaks of the as-prepared oxides can be assigned to a single phase of perovskite-type oxide, and do not indicate the presence of any crystalline impurities. The BET surface areas of prepared perovskite-type oxides were measured to be 31 m2 g1, 32 m2 g1, and 26 m2 g1 for La0.6Sr0.4Mn0.6Fe0.4O3, La0.6Ca0.4Mn0.6Fe0.4O3, and La0.8Sr0.2Co0.6Fe0.4O3, respectively [33]. We confirmed that the perovskite-type structure was maintained after ball milling treatment with carbon black (Fig. 2(b)). Fig. 2(c) shows a typical TEM image of carbon-supported La0.6Sr0.4Mn0.6Fe0.4O3 (La0.6Sr0.4Mn0.6Fe0.4O3/Vulcan XC72). The particle size of La0.6Sr0.4Mn0.6Fe0.4O3 was ca. 10e30 nm, and the perovskite-type oxide particles were uniformly dispersed on the carbon support. Fig. 3 shows the RRDE voltammograms for oxygen reduction with Co-OEP-modified La0.6Sr0.4Mn0.6Fe0.4O3/Vulcan XC72 (CoOEP/La0.6Sr0.4Mn0.6Fe0.4O3/C) in O2-saturated 0.1 M KOH solution obtained at 10 mV s1, 900 rpm, and 25 C. To benchmark the catalyst, the results for Co-OEP/Vulcan XC72, La0.6Sr0.4Mn0.6Fe0.4O3/ Vulcan XC72, carbon (Vulcan XC72) and commercially available Pt/ C (TEC10E50E (Tanaka Kikinzoku Kogyo), loading amount: 63.4 mg cm2) are also shown. The onset potential of Co-OEP/ La0.6Sr0.4Mn0.6Fe0.4O3/C (C) is shifted to the positive side and the ORR current at the disk electrode was drastically increased compared with that of the La0.6Sr0.4Mn0.6Fe0.4O3/C (-). Therefore, the ORR activity of the La0.6Sr0.4Mn0.6Fe0.4O3/C catalyst is considered to be improved by the addition of Co-OEP as a co-catalyst. The formation of a hydrogen peroxide intermediate (HO 2 ) was confirmed by the ring electrode current during oxygen reduction.
For Co-OEP/C with mainly two-electron O2 reduction activity (:), increased ring current which suggested the production of HO 2 was clearly observed from ca. 0.82 V (vs RHE), corresponding to oxygen reduction. The ring current of the Co-OEP/La0.6Sr0.4Mn0.6Fe0.4O3/C catalyst was significantly lower than that of Co-OEP/C. Fig. 3(c) presents the fraction of HO 2 production during oxygen reduction (XH2O2), as estimated from RRDE voltammograms [37]. The values for catalysts containing perovskite (La0.6Sr0.4Mn0.6Fe0.4O3/C, CoOEP/La0.6Sr0.4Mn0.6Fe0.4O3/C) were much lower than those for catalysts without perovskite (carbon, Co-OEP/C), indicating that O2 is reduced more completely to OH at the electrode. These RRDE results for the present catalysts suggest that the HO 2 generated by Co-OEP (and carbon) is efficiently reduced to OH by La0.6Sr0.4Mn0.6Fe0.4O3. In order to investigate the O2 reduction mechanism of the present Co-OEP/La0.6Sr0.4Mn0.6Fe0.4O3/C catalyst, RRDE measurements were also performed with various amounts of Co-OEP/ La0.6Sr0.4Mn0.6Fe0.4O3/C (Fig. 4). The catalyst contents on a GC disk electrode were 199 mg cm2 (C), 99.5 mg cm2 (-), and 49.7 mg cm2 (:). Co-OEP/La0.6Sr0.4Mn0.6Fe0.4O3/C showed a lower onset potential and increased ring current with a decrease in the loading amount on GC. Thus, HO 2 production during oxygen reduction was increased. Previous studies on Pt/C catalysts [38e40] have described a decrease in ORR activity (especially, increased H2O2 production) with a decrease in catalyst loading. We discuss here how the ORR on Pt/C catalysts occurs via a two-by-two electron pathway through a H2O2 intermediate as well as via a direct four-electron pathway. After two-electron O2 reduction on a Pt
Fig. 3. RRDE voltammograms in O2-saturated 0.1 M KOH at 25 C ((a) disk electrode, (b) ring electrode) and fractions of HO 2 production during oxygen reduction (c) for CoOEP/La0.6Sr0.4Mn0.6Fe0.4O3/Vulcan XC72 (C), La0.6Sr0.4Mn0.6Fe0.4O3/Vulcan XC72 (-), Co-OEP/Vulcan XC72 (:), and Vulcan XC72 (solid line). RRDE data for commercially available Pt/C catalyst (TEC10E50E, loading amount: 63.4 mg cm2) is also shown in (a) as a broken line. Scan rate and rotation speed were 10 mV s1 and 900 rpm, respectively.
Fig. 4. RRDE voltammograms of CoOEP/La0.6Sr0.4Mn0.6Fe0.4O3/Vulcan XC72 in O2saturated 0.1 M KOH at 25 C ((a) disk electrode, (b) ring electrode) and fractions of HO 2 production during oxygen reduction (c). Measurements were performed with various amounts of Co-OEP/La0.6Sr0.4Mn0.6Fe0.4O3/C catalyst (catalyst content on GC: 199 (C), 99.5 (-), and 49.7 (:) mg cm2). Scan rate and rotation speed were 10 mV s1 and 900 rpm, respectively.
764
T. Nagai et al. / Journal of Power Sources 293 (2015) 760e766
Table 1 BET specific surface area for La0.6Sr0.4Mn0.6Fe0.4O3/carbon (carbon; Vulcan XC72, Ketjenblack EC600JD, Denka Black). BET surface area/m2 g1 LSMF/Vulcan XC72 LSMF/Ketjenblack EC600JD LSMF/Denka Black Vulcan XC72 Ketjenblack EC600JD Denka Black
76 209 50 232 1282 63
surface, desorbed H2O2 is re-adsorbed and reduced on a nearby Pt particle, or diffuses away into the bulk solution. When Pt/C catalysts are highly agglomerated on GC (interparticle distance of Pt is short), the H2O2 molecules can easily reach nearby Pt particles, and an increased disk current and decreased ring current is observed in RRDE measurements. In these terms, the present Co-OEP/ La0.6Sr0.4Mn0.6Fe0.4O3/C catalyst is considered to show a consecutive two-by-two electron pathway for O2 reduction; two-electron O2 reduction to HO 2 occurs on Co-OEP (and carbon), and HO2 is captured and reduced on a nearby La0.6Sr0.4Mn0.6Fe0.4O3 particle. O2 is further reduced to OH due to the closeness between Co-OEP and La0.6Sr0.4Mn0.6Fe0.4O3, which enhances the likelihood of capturing HO 2. We also prepared the Co-OEP/La0.6Sr0.4Mn0.6Fe0.4O3/carbon catalyst using other types of conductive carbon (Ketjenblack EC600JD and Denka Black), and performed RRDE measurements. The BET specific surface area of La0.6Sr0.4Mn0.6Fe0.4O3/carbon is summarized in Table 1. The surface area of La0.6Sr0.4Mn0.6Fe0.4O3/C varied depending on the specific surface area of carbon. Fig. 5 shows hydrodynamic voltammograms of La0.6Sr0.4Mn0.6Fe0.4O3/C and Co-OEP/La0.6Sr0.4Mn0.6Fe0.4O3/C. The onset potential for each La0.6Sr0.4Mn0.6Fe0.4O3/C was approximately the same. Furthermore, under modification by Co-OEP, increased catalytic activity was confirmed regardless of the type of conductive carbon. The reduction currents for the Ketjenblack EC600JD-supported sample (La0.6Sr0.4Mn0.6Fe0.4O3/Ketjenblack, and Co-OEP/La0.6Sr0.4Mn0.6Fe0.4O3/Ketjenblack) were slightly lower than those for the samples with Vulcan XC72 and Denka Black. One of the possible causes is as follows. For the sample supported on Ketjenblack EC600JD, the catalyst loading with respect to the surface area of carbon becomes small due to the enormously high specific surface area of
Fig. 5. Hydrodynamic voltammograms of La0.6Sr0.4Mn0.6Fe0.4O3/carbon (LSMF/C, broken line) and Co-OEP/La0.6Sr0.4Mn0.6Fe0.4O3/carbon (Co-OEP/LSMF/C, solid line) in O2-saturated 0.1 M KOH solution obtained at 10 mV s1, 900 rpm, and 25 C (Vulcan XC72 (C), Denka Black (-), and Ketjenblack EC600JD (:)).
Ketjenblack. Thus, uncovered and inactive areas remain over the carbon surface, and it may be difficult for perovskite oxide to effectively capture the HO 2 that participates in the ORR reaction. To compare the ORR performances of porphyrin/perovskite/ carbon with different structures of porphyrin complex or compositions of perovskite oxide, we prepared catalysts with several types of porphyrin and perovskite. First, porphyrin complex with two different structures (iron octaethylporphyrin (Fe-OEP) and cobalt tetraphenylporphyrin (Co-TPP)) were added to the La0.6Sr0.4Mn0.6Fe0.4O3/Vulcan XC72 support, and their ORR activities were investigated (Fig. 6-I). All of the complexes added to La0.6Sr0.4Mn0.6Fe0.4O3/Vulcan XC72 showed a two-electron reduction activity of oxygen to HO 2 as shown in Fig. 6-II. The onset potentials for each catalyst are summarized in Table 2. As described in Fig. 3, a large potential shift was observed when the La0.6Sr0.4Mn0.6Fe0.4O3/Vulcan XC72 catalyst was modified by CoOEP (C). For the sample modified by an Fe-OEP complex (-), which has the same ligands as Co-OEP, the ORR activity was not enhanced. This can explain the low ORR activity of Fe-OEP/C (onset potential: ca. 0.70 V (vs RHE)) compared to Co-OEP/C (onset potential: ca. 0.83 V (vs RHE)). Along with the decreased ORR activity, the production of HO 2 by Fe-OEP/Vulcan XC72 was observed from significantly negative potential. Furthermore, the onset potential for Fe-OEP/C is lower than that for La0.6Sr0.4Mn0.6Fe0.4O3/C (ca. 0.85 V (vs RHE)), and therefore, the oxygen reduction on La0.6Sr0.4Mn0.6Fe0.4O3/C catalyst is less promoted by modification with an Fe-OEP complex. In the case of modification with Co-TPP complex, which shows relatively high activity on La0.6Sr0.4Mn0.6Fe0.4O3/Vulcan XC72 (:), the ORR activity was not increased as drastically as that of the CoOEP-modified catalyst. The reason remains unclear, however, the principal reason for this difference is considered to be the difference in the molecular structures of Co-TPP and Co-OEP. Co-TPP complex has a bulky structure with a phenyl group attached to the porphyrin ring, while Co-OEP has a relatively planar structure. It has been reported that the amount of Co-TPP adsorbed on Vulcan XC72 is significantly lower than that of Co-OEP under equilibrium adsorption conditions, almost due to the bulkiness and orientation of the substituents of Co-TPP [41]. The amount of the porphyrin adsorbed on carbon reflects the strength of a specific interaction between the porphyrin and carbon (i.e. p-p interaction between the porphyrin ring and carbon surface). Strong interaction between Co-OEP and carbon would be favorable for the electron transfer from an electrode to the molecule. The possible difference in the interaction might be related to the difference in the activity. For a La-based perovskiteetype oxide, enhanced catalytic activity has been achieved by doping the La3þ site with alkaline-earth metals (Sr, Ca), and by including Mn or Co at a transition metal cation site. To investigate the effect of modification with a porphyrin complex with different compositions of perovskite, we also prepared La0.6Ca0.4Mn0.6Fe0.4O3/Vulcan XC72 and La0.8Sr0.2Co0.6Fe0.4O3/Vulcan XC72, and modified these catalysts with CoOEP complex. Fig. 7(a) shows the hydrodynamic voltammograms for Co-OEP-modified perovskite/Vulcan XC72 catalysts. The disk currents of the La0.6Ca0.4Mn0.6Fe0.4O3/C and La0.8Sr0.2Co0.6Fe0.4O3/C without Co-OEP were comparable to that of La0.6Sr0.4Mn0.6Fe0.4O3/ C. For La0.6Ca0.4Mn0.6Fe0.4O3/C (-, solid line) and the La0.8Sr0.2Co0.6Fe0.4O3/C (:, solid line), the addition of Co-OEP produced a positive shift of the onset potential, as seen with Co-OEP/ La0.6Sr0.4Mn0.6Fe0.4O3/C (C, solid line). Although all of the prepared samples exhibited enhanced O2 reduction activity, the ring current (Fig. 7(b)) for Co-OEP/La0.8Sr0.2Co0.6Fe0.4O3/C was higher than those for LaeMn-based catalysts, and large fractions of HO 2 production (XH2O2) were observed (Fig. 7(c)). This is due to the large XH2O2 (and ring current) for the La0.8Sr0.2Co0.6Fe0.4O3/C, which indicates that
T. Nagai et al. / Journal of Power Sources 293 (2015) 760e766
765
Fig. 6. I: RRDE voltammograms for porphyrin-modified La0.6Sr0.4Mn0.6Fe0.4O3/Vulcan XC72 (porphyrin/LSMF/C). II: RRDE voltammograms for porphyrin-modified Vulcan XC72 (porphyrin/C). Each measurement was conducted in 0.1 M KOH saturated with O2 at 10 mV s1, 900 rpm, and 25 C ((a) disk electrode, (b) ring electrode). The fraction of HO 2 production during oxygen reduction (XH2O2) is also shown in (c).
HO by La0.8Sr0.2Co0.6Fe0.4O3 2 is not efficiently reduced to OH compared to LaeMn-based perovskite (La0.6Sr0.4Mn0.6Fe0.4O3 and La0.6Ca0.4Mn0.6Fe0.4O3). To obtain high four-electron O2 reduction activity on the Co-OEP-modified perovskite/C catalyst, it is preferable to apply perovskite with a higher HO 2 reduction activity (i.e., perovskite/C showing a small XH2O2 value). Based on these experimental results, the ORR activities of the perovskite-type oxides prepared in this study were enhanced by the addition of a Co-OEP complex. In addition, the HO 2 reduction activity of the perovskitetype oxide plays an important role in the enhancement of the quasi-four-electron reduction of O2.
4. Conclusion In this study, we prepared a cobalt octaethylporphyrin (CoOEP)-modified perovskite-type oxide/carbon catalysts (Co-OEP/ perovskite/C, perovskite: La0.6Sr0.4Mn0.6Fe0.4O3, La0.6Ca0.4Mn0.6Fe0.4O3, and La0.8Sr0.2Co0.6Fe0.4O3). The electrocatalytic reduction activities of oxygen (ORR) of the prepared catalysts were investigated in alkaline media by rotating ring disk electrode (RRDE)
Table 2 Onset potential for porphyrin-modified La0.6Sr0.4Mn0.6Fe0.4O3/Vulcan XC72 (porphyrin/LSMF/C) and porphyrin-modified carbon (porphyrin/C). The onset potential is the potential at which the disk current reaches 0.005 mA. Onset potential/V vs RHE Co-OEP/C Fe-OEP/C Co-TPP/C Co-OEP/LSMF/C Fe-OEP/LSMF/C Co-TPP/LSMF/C LSMF/C Vulcan XC72
0.829 0.702 0.824 0.906 0.879 0.868 0.854 0.677
Fig. 7. Hydrodynamic voltammograms for perovskite-type oxide/C and Co-OEPmodified perovskite-type oxide/C catalysts in O2-saturated 0.1 M KOH solution obtained at 10 mV s1, 900 rpm, and 25 C (C; La0.6Sr0.4Mn0.6Fe0.4O3/C (LSMF/C), :; La0.6Ca0.4Mn0.6Fe0.4O3/C (LCMF/C), -; La0.8Ca0.2Mn0.6Fe0.4O3/C (LSCF/C). The solid and broken lines represent the samples with and without Co-OEP, respectively.
766
T. Nagai et al. / Journal of Power Sources 293 (2015) 760e766
voltammetry. The onset potential of Co-OEP/perovskite/C was shifted to the positive side compared with that of perovskite/C at a disk electrode by the addition of Co-OEP complex with twoelectron O2 reduction activity to the perovskite/C. A decreased ring current and a small fraction of HO 2 formation relative to CoOEP/C were also confirmed. Further measurements with various amounts of catalyst were similar to the case with Pt/C catalyst; a lower onset potential and increased ring current with a decrease in the loading amount on GC. It is considered that quasi-four-electron O2 reduction is promoted on Co-OEP (O2 reduction to HO 2 ) and perovskite (HO 2 reduction). With regard to modification of the La0.6Sr0.4Mn0.6Fe0.4O3/C catalyst with other porphyrin complex (FeOEP and Co-TPP), the ORR activity of porphyrin/perovskite/C catalyst is varied with difference of their activity. Acknowledgment The authors are grateful to Tokuyama Corporation for providing the anion exchange resin (AS-4). We also thank Dr. Mitsunori Kitta (AIST) for his assistance with TEM. References [1] M. Savy, Electrochim. Acta. 3 (1968) 1359e1376. [2] H.A. Gasteiger, S.S. Kocha, B. Sompalli, F.T. Wagner, Appl. Catal. B 56 (2005) 9e35. [3] J.S. Spendelow, A. Wieckwski, Phys. Chem. Chem. Phys. 9 (2007) 2654e2675. [4] K. Asazawa, K. Yamada, H. Tanaka, A. Oka, M. Taniguchi, T. Kobayashi, Angew. Chem. Int. Ed. 46 (2007) 8024e8027. [5] D.A. Slanac, A. Lie, J.A. Paulson, K.J. Stevenson, K.P. Johnston, J. Phys. Chem. C 116 (2012) 11032e11039. [6] L.N. Nekrasov, L. Müller, D. Akad, Nauk. SSSR 149 (1963) 1107. [7] A. Damjanovic, M.A. Genshaw, J.O.'M. Bockris, J. Electrochem. Soc. 114 (1967) 1107e1112. [8] A. Damjanovic, M.A. Genshaw, J.O.'M. Bockris, J. Electroanal. Chem. Interfacial Electrochem. 15 (1967) 173e180. [9] I. Morcos, E. Yeager, Electrochim. Acta 15 (1970) 953e975. [10] P.S.D. Brito, C.A.C. Sequeira, J. Power Sources 52 (1994) 1e16. [11] V.S. Bagotzky, N.A. Shumilova, G.P. Samoilov, E.I. Khrushcheva, Electrochim. Acta 17 (1972) 1625e1635. [12] D. Sepa, M. Vojnovic, A. Damjanovic, Electrochim. Acta 15 (1970) 1355e1366. [13] B.B. Blizanac, P.N. Ross, N.M. Markovic, J. Phys. Chem. B 110 (2006) 4735e4741. [14] P. Zoltowski, D.M. Drazic, L. Vorkapic, J. Appl. Electrochem. 3 (1973) 271e283. [15] L. Mao, T. Sotomura, K. Nakatsu, N. Koshiba, D. Zhang, T. Ohsaka,
J. Electrochem. Soc. 149 (2002) A504eA507. [16] D.B. Meadowcroft, Nature 226 (1970) 847e848. [17] Y. Shimizu, K. Uemura, H. Matsuda, N. Miura, N. Yamazoe, J. Electrochem. Soc. 137 (1990) 3430e3433. [18] J. Suntivich, H.A. Gasteiger, N. Yabuuchi, H. Nakanishi, J.B. Goodenough, Y. Shao-Horn, Nat. Chem. 3 (2011) 546e550. [19] A. Bettelheim, B.A. White, R.W. Murray, J. Electroanal. Chem. Interfacial Electrochem. 217 (1987) 271e286. [20] J. Zagal, M. Paez, A.A. Tanaka, J.R. Dossantos, C.A. Linkous, J. Electroanal. Chem. 339 (1992) 13e30. [21] J. Tulloch, S.W. Donne, J. Power Sources 188 (2009) 359e366. [22] M. Yuasa, N. Yamazoe, K. Shimanoe, J. Electrochem, Soc. 158 (2011) A411eA416. [23] K. Kinoshita, Carbon-electrochemical and Physicochemical Properties, Wiley, New York, USA, 1988. [24] C. Song, J. Zhang, Electrocatalytic oxygen reduction reaction, in: J. Zhang (Ed.), PEM Fuel Cell Electrocatalysts and Catalyst Layers: Fundamentals and Applications, Springer, London, 2008, pp. 89e134. [25] Y. Miyahara, K. Miyazaki, T. Fukutsuka, T. Abe, J. Electrochem. Soc. 161 (2014) F694eF697. [26] T. Poux, F.S. Napolskiy, T. Dintzer, G. Kerangueven, S.Y. Istomin, G.A. Tsirlina, E.V. Antipov, E.R. Savinova, Catal. Today 189 (2012) 83e92. [27] F. Van Den Brink, W. Visscher, E. Barendrecht, J. Electroanal. Chem. Interfacial Electrochem. 157 (1983) 283e304. [28] F. Van Den Brink, W. Visscher, E. Barendrecht, J. Electroanal. Chem. Interfacial Electrochem. 157 (1983) 305e318. [29] J. Ouyang, K. Shigehara, A. Yamada, F.C. Anson, J. Electroanal. Chem. Interfacial Electrochem. 297 (1991) 489e497. [30] O. El Mouahid, C. Coutanceau, E.M. Belgsir, P. Crouigneau, J.M. Leger, C. Lamy, J. Electroanal. Chem. 426 (1997) 117e123. [31] L. Mao, K. Arihara, T. Sotomura, T. Ohsaka, Electrochim. Acta 49 (2004) 2515e2521. [32] Y. Kiros, Int. J. Electrochem. Sci. 2 (2007) 285e300. [33] T. Nagai, N. Fujiwara, M. Asahi, S. Yamazaki, Z. Siroma, T. Ioroi, J. Asian Ceram. Soc. 2 (2014) 329e332. [34] N. Fujiwara, Z. Siroma, S. Yamazaki, M. Asahi, T. Nagai, T. Ioroi, in: Proceedings of 80th Meeting of the Electrochemical Society of Japan 3A21, 2013. [35] S. Yamazaki, Y. Yamada, S. Takeda, M. Goto, T. Ioroi, Z. Siroma, K. Yasuda, Phys. Chem. Chem. Phys. 12 (2010) 8968e8976. [36] T. Nagai, N. Fujiwara, M. Kitta, M. Asahi, S. Yamazaki, Z. Siroma, T. Ioroi, J. Fuel Cell. Sci. Tech. 12 (2015), 021007-1e021007-5. [37] U.A. Paulus, T.J. Schmidt, H.A. Gasteiger, R.J. Behm, J. Electroanal. Chem. 495 (2001) 134e145. [38] M. Inaba, H. Yamada, J. Tokunaga, A. Tasaka, Electrochem. Solid-State Lett. 7 (2004) A474eA476. [39] K.J.J. Mayrhofer, D. Strmcnik, B.B. Blizanac, V. Stamenkovic, M. Arenz, N.M. Markovic, Electrochim. Acta 53 (2008) 3181e3188. [40] E. Fabbri, S. Taylor, A. Rabis, P. Levecque, O. Conrad, R. Kctz, T.J. Schmidt, Chem. Cat. Chem. 6 (2014) 1410e1418. [41] S. Yamazaki, Y. Yamada, T. Ioroi, N. Fujiwara, Z. Siroma, K. Yasuda, Y. Miyazaki, J. Electroanal. Chem. 576 (2005) 253e259.