- Email: [email protected]
Electrocatalytic activity and operational stability of electrodeposited Pd–Co films towards ethanol oxidation in alkaline electrolytes
Journal of Power Sources 293 (2015) 815e822
Contents lists available at ScienceDirect
Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour
Electrocatalytic activity and operational stability of electrodeposited PdeCo films towards ethanol oxidation in alkaline electrolytes Lok-kun Tsui a, Claudio Zafferoni b, Alessandro Lavacchi c, Massimo Innocenti b, Francesco Vizza c, Giovanni Zangari a, * a b c
Department of Materials Science and Engineering and CESE, University of Virginia, 395 McCormick Rd, Charlottesville, VA 22904, USA Department of Chemistry, University of Florence, Via della Lastruccia 3-13, 50019 Florence, Italy ICCOM-CNR, Via Madonna del Piano 10, 50019 Florence, Italy
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
PdeCo catalysts were electrodeposited with compositions between 20 and 80 at% Co. A transformation of alloyed Co to CoOOH/Co3O4 flakes was observed after cycling. The Co in the films supplies eOH to EtOH oxidation, reducing onset potential. A max peak current density is found at an optimum composition of 77 at% Co. A 50 at% CoePd film exhibited the highest stable currents among PdeCo alloys.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 9 March 2015 Received in revised form 7 May 2015 Accepted 30 May 2015 Available online 11 June 2015
Direct alkaline ethanol fuel cells (DEFCs) are usually run with Pd anodic catalysts, but their performance can be improved by utilizing alloys of Pd and Co. The oxyphilic Co serves to supply ample eOH to the ethanol oxidation reaction, accelerating the rate limiting step at low overpotential under alkaline conditions. PdeCo films with compositions between 20 and 80 at% Co can be prepared by electrodeposition from a NH3 complexing electrolyte. Cyclic voltammetry studies show that the ethanol oxidation peak exhibits increasing current density with increasing Co content, reaching a maximum at 77% Co. In contrast, potentiostatic measurements under conditions closer to fuel cell operating conditions show that a 50 at% Co alloy has the highest performance. Importantly, the CoePd film is also found to undergo phase and morphological transformations during ethanol oxidation, resulting in a change from a compact film to high surface area flake-like structures containing Co3O4 and CoOOH; such a transformation instead is not observed when operating at a constant potential of 0.7 VRHE. © 2015 Elsevier B.V. All rights reserved.
Keywords: Direct ethanol fuel cell Ethanol oxidation Anode catalyst Palladiumecobalt alloy Cobalt oxide Electrodeposition
1. Introduction
* Corresponding author. E-mail address: [email protected] (G. Zangari). http://dx.doi.org/10.1016/j.jpowsour.2015.05.121 0378-7753/© 2015 Elsevier B.V. All rights reserved.
Direct Ethanol Fuel Cells (DEFCs) present various attractive features as power sources for portable and, more recently, transportation applications, since they exhibit energy densities
816
L.-k. Tsui et al. / Journal of Power Sources 293 (2015) 815e822
comparable to gasoline and, in contrast to hydrogen, benefit from a simpler and mostly existing storage and transport infrastructure [1,2]. The use of ethanol, as well as other high molecular weight alcohols such as ethylene glycol or glycerol, is advantageous with respect to methanol or formic acid feeds, due to their low toxicity [3,4], and the possibility to produce these alcohols from biomass (sugar- or cellulose-based) using renewable, CO2-neutral methods [5,6]. The main drawback of these devices however derives from the sluggish kinetics of ethanol oxidation, much slower than that of lighter alcohols or H2, which results in lower efficiency and inferior performance with respect to the corresponding fuel cells. The development of efficient and selective anodic electrocatalysts is necessary to overcome this hurdle. These anodic electrocatalysts could also be successfully exploited in the electrochemical reforming of bio-alcohols providing a sustainable route for the cogeneration of hydrogen and important raw chemicals [7e9]. Platinum catalysts are commonly used in acidic electrolytes because of their excellent ability to enhance the adsorption and dissociation of small organics molecules (SOMs); however, strongly adsorbed organic species such as carbonyl and CO, tend to poison the catalyst [1]. Furthermore, the high cost of Pt contributes significantly (~50%) to the total fuel cell stack cost. In contrast, in alkaline media the kinetics of ethanol oxidation is faster, and the adsorption strength of organic fragments is weaker. Importantly, less expensive electrocatalysts such as Pd can be used due to their stability in alkaline environments [10]. In an alkaline-type DEFC the reaction at the anode is targeted to be the complete oxidation of ethanol to CO2 (1), accompanied by the reduction of oxygen at the cathode (2), leading to the overall reaction of ethanol “combustion” (3) [6].
CH3 CH2 OH þ 16OH / 2CO3 2 þ 11H2 O þ 12e
(1)
3O2 þ 6H2 O þ 12e / 12OH
(2)
CH3 CH2 OH þ 3O2 þ 4OH / 5H2 O þ 2CO3 2
(3)
Reaction (1) consists of a series of intermediate steps, whereby adsorbed ethanol is oxidized via hydrogen abstraction to acetaldehyde; the latter may subsequently be oxidized further to CO2, if the CeC bond is broken, or transformed into a terminal acetate product (Fig. 1) [11,12]. The rate determining step in this reaction mechanism is the adsorption of hydroxyl (OHads) at low overpotential, or the removal of the adsorbed ethoxy (CH3COads) by adsorbed hydroxyl at higher potentials [13]. The overall reaction could be accelerated by addition of a more oxyphilic metal to Pt or Pd, which would enhance water dissociation and the formation of oxygenated species that become available for further oxidation to acetic acid. Most studied in this context are Ru and Sn, with the latter providing higher oxidation currents [14]; Co on the other
hand could provide the same function at a lower cost with respect to Ru, while possibly enhancing CeC bond breaking, due to the homologous electronic structure to Rh and Ir, both known to enhance selectivity towards complete oxidation [14,15]. The overall selectivity for ethanol complete oxidation is found to be generally low, with the exception of Rao et al. [16], who reported a 55% CO2 conversion efficiency in alkaline solutions and only 2% in acidic ones. CH3COO- is in most cases by far the major product of ethanol oxidation in alkaline environment, decreasing the faradaic efficiency by 1/3 as compared to carbonate formation. Despite this intrinsic loss of faradaic efficiency, the fast kinetics of acetate production on Pd results in relatively low anodic overpotential, providing DEFCs with power densities up to 335 mW cm2 [17,18]. Pd-Ni [19e22] and PdeCo [23e25] have been already investigated as anodic catalysts for alkaline DEFCs, each showing better performance than Pd alone when normalized per unit area. The alkaline environment allows complete elimination of Pt from these fuel cells. Indeed in alkaline environment iron and cobalt macrocycles are active for the oxygen reduction with similar performance to Pt based ORR catalysts, these materials being also inactive toward alcohol oxidation This is particularly relevant as it results in negligible voltage drop for alcohol crossover [26]. The studies referred to above speculate that the facilitated formation of eOH on the 3d transition metal in alkaline solutions is the main mechanism by which ethanol oxidation was enhanced. Liu et al. identified the equiatomic composition to be optimum for formic acid and methanol oxidation at PdeCo nanoparticles supported on carbon fibers [25]. Xu et al. mixed Pd nanoparticles 6e10 nm with oxides of Ce, Ni, Co, and Mn, and found that a Co3O4ePd composite with an oxide content of 0.05 mg cm2 yields the highest peak oxidation current for ethanol oxidation, while noting low stability for this oxide-Pd pairing [24]. So far however, the optimum features of the catalyst structure and composition for ethanol oxidation have not been assessed; in addition, the stability of the catalyst surface has received little to no attention. On a related issue, a rigorous protocol for the evaluation of the electrocatalytic activity in ethanol oxidation catalysts has been missing and current methods to compare catalyst activity are not standardized, hindering an accurate ranking of the relative activity of various catalysts. In this work, we report on the electrocatalytic activity and longterm stability of PdeCo alloy films in view of their potential application as anodic catalysts for alkaline DEFCs. We produce a set of thin film electrodes with a broad range of compositions, using electroplating from an ammonia-based electrolyte, and we evaluate different methods to rank activity towards ethanol oxidation. In the process, we identify an optimum alloy composition at a Co fraction around 50e60%, yielding the highest current densities for ethanol oxidation and also find that the reaction onset occurs at more negative potentials with increasing Co fraction. Accelerated oxidation of Co in the alloy to form Co (hydro)oxides is observed during the potential cycling of PdeCo in alkaline, ethanolcontaining electrolytes, which is accompanied by a morphological transformation to a high surface area flake-like structure, never explicitly reported before. 2. Experimental 2.1. Electrodeposition of PdeCo
Fig. 1. Simplified reaction mechanism for ethanol oxidation; the reaction can either lead to CO2 if the CeC bond is broken or end at acetate as a terminal product [11,12].
An ammonia-based complexing electrolyte was utilized in order to shift the redox potentials of the two metals closer together and thus facilitate the formation of alloy solid solutions [27]. The electrolyte was prepared as follows: first, two beakers were prepared, one containing a 20 mM PdCl2, 0.4 M (NH4)2SO4 and 0.5 M NH3
L.-k. Tsui et al. / Journal of Power Sources 293 (2015) 815e822
817
aqueous solution, and the other containing a 50 mM CoSO4, 0.4 M (NH4)2SO4 and 0.5 M NH3 solution. The latter solution was stirred until it turned black, indicating that the oxidation of Co(II) to Co(III) had reached completion. Oxidation of Co(II) to Co(III) prevents the spontaneous oxidation of Co(II) in the presence of Pd(II), which would reduce the Pd out of solution and hinder compositional control. Upon completion of the Co oxidation, the two solutions were mixed together to form the plating electrolyte. PdeCo films were deposited under current control on Au/Si substrates with a constant current density between 0.25 and 2.0 mA cm2 to a fixed charge density of 1.0 C cm2 (corresponding to 275e400 nm of PdeCo films as measured by cross sectional SEM). Pure Pd films were prepared in a solution of 10 mM PdCl2, 0.3 M (NH4)2SO4, and 0.5 M C7H6O6S as complexing agents at a current density of 1 mA cm2 to a fixed charge density of 1 C cm2 (460 nm). Pure Co films were instead obtained by potentiostatic deposition at 1.4 VSCE for 3 min in a solution of 0.3 M CoSO4 þ 0.25 M H2SO4 yielding a Co film of approximately 700 nm. 2.2. Electrocatalytic activity Electrochemical measurements were carried out in solutions containing 2 M KOH or 2 M KOH with 10 wt% ethanol (Alfa Aesar, 95þ%). A three electrode cell configuration was used which included a Pt mesh counter electrode, an Hg/HgO reference electrode (CH Instruments), and the working electrode. The onset potential for ethanol oxidation was determined by two distinct methods: (i) a linear intercept was fitted to the left side of the oxidation peak with the background current; alternately, (ii) the potential at which the current density exceeded 0.1 or 1 mA cm2 was recorded. 2.3. Materials characterization The morphology and composition of the PalladiumeCobalt films were characterized by scanning electron microscopy (SEM) and energy dispersive x-ray spectroscopy (EDS) in an FEI Quanta 650 SEM, respectively. Phase identification was carried out using Raman spectroscopy (Renishaw InVIA Raman Microscope) using a 488 nm laser with a 10s acquisition time. X-Ray Diffraction (XRD) was performed with Cu Ka radiation (l ¼ 0.154 nm) using a Panalytical X'PERT MPD x-ray diffractometer. Reference patterns were obtained from the ICDD PDF DDView þ database [28]. 3. Results and discussion 3.1. Composition of PdeCo films Fig. S1 shows the composition as a function of current density for PdeCo films grown on Au substrates by galvanostatic electrodeposition. Varying the applied current density between 0.25 and 1.5 mA cm2, the composition varies from 20% up to 80% Co. Beyond 1.5 mA cm2, the Co content saturates at 80%. Such a broad composition window allows us to explore in detail the effect of Co content on the ethanol oxidation activity of electrodeposited PdeCo films. The experimental phase diagram for bulk PdeCo shows a continuous series of face-centered cubic solid solutions, i.e. true alloys, with a possible transition to Co-rich hexagonal phases below 422 C and less than ~20 at% Pd [29]. In contrast, the FactSage database shows a miscibility gap below ~600 C and between 0 and 50 at% Pd [30,31], and only ordered structures of the L12 and L10 types are reported for Co > 25 at% [32,33]. Fig. 2 shows the XRD patterns obtained on the alloy films and the pure Pd and Co films for comparison. Besides the substrate peaks, all PdeCo films that we obtained show a peak at 41, located between the FCC Pd (111)
Fig. 2. XRD Patterns of electrodeposited PdeCo films along with pure Pd and pure Co films show, with increasing Pd content, an FCC alloy peak around 41 followed by the formation of a second phase above 46% Co, identified as HCP Co.
and HCP Co (100) peak. The position of this peak does not shift with composition; assuming a cubic lattice, the corresponding lattice constant is 0.380 nm ± 0.002 nm. As expected, this value lies between the lattice constants for FCC Pd (0.389 nm, PDF 00-0461043) and FCC Co (0.354 nm, PDF 00-015-0806), and corresponds approximately to a solid solution containing ~ 35 at% Co [34]. At and above 46 at% Co, new peaks emerge at 2q ¼ 41.6 and 44.5 , which are assigned to hexagonal Co. At high overvoltage, the probability of nucleation and growth of Co seeds in parallel with the solid solution is possible, resulting in a kinetically induced formation of a pure Co phase, and in two-phase growth; interestingly, the alloy peak shifts to lower angles when the pure Co phase forms, suggesting a Pd enrichment in the cubic phase. 3.2. Electrocatalytic activity of PdeCo alloy films Fig. 3(a) shows the cyclic voltammetry (CV) response over six successive cycles at a 50 at% Co, PdeCo film in 2 M KOH; this behavior is representative of all the alloy compositions investigated. Besides characterizing the electrochemical response, these data also enable the determination of the electrochemically active ski [35]. Fig. 3(b) surface area (EASA) of Pd as reported by Czerwin displays the CV response of PdeCo films in the full range of compositions between 0 and 100% Co in 2 M KOH and 10 wt% ethanol. In the KOH electrolyte (Fig. 3(a)), scanning towards positive potentials, a peak corresponding to the oxidation of cobalt is observed at about 1.0 VRHE, that is not present in pure Pd films; this is followed at higher potentials by Pd oxidation. On the return scan, two peaks corresponding to the sequential reduction of a monolayer of Pd oxide to Pd are observed, near 1.0 and 0.7 VRHE, respectively, followed by hydrogen adsorption and then hydrogen evolution at more negative potentials [36]. With repeated cycles, the peaks corresponding to PdO reduction and Co oxidation increase in intensity, suggesting a corresponding increase in surface area during cycling. When 10 wt% ethanol is added to the electrolyte (Fig. 3(b)), a new, much stronger asymmetric peak corresponding to the oxidation of ethanol appears centered at 0.8 to 1 VRHE; a corresponding smaller anodic peak on the return scan is associated with the oxidation of intermediate products still adsorbed on the alloy
818
L.-k. Tsui et al. / Journal of Power Sources 293 (2015) 815e822
Fig. 4(a) shows the three potentials thus calculated for PdeCo films as a function of composition. Using any of these methods, with more Co in the films, the reference potential for ethanol oxidation shifts towards more negative values. The linear intercept method gives a decrease of 100 mV with higher Co content across the various CoePd alloys, but taking the reference potential to be the value at which 0.1 mA cm2 is reached gives a decrease of 200 mV, while taking a measurement at 1.0 mA cm2 gives a decrease of 300 mV. No onset potential could be extracted from the pure Co film because no ethanol oxidation is detected. These results suggest that an increase of the Co fraction in the alloy enhances the oxidation peak intensity and reduces the overpotential needed to activate the ethanol oxidation current, favoring an improved performance. On the other hand, the various benchmarks used to rank the electrocatalytic activity of CoePd alloys result in the same ranking, but differ in the estimation of the relative activity. Since actual fuel cell operating conditions are closer to points at the onset of the oxidation wave rather than near the peak, it is important to monitor potentiostatic current transients under these
Fig. 3. (a) Cyclic voltammetry at 50 at% Co electrodeposited films over several cycles in ethanol free 2 M KOH and (b) 5th cycle of CV at PdeCo films of various compositions in 2 M KOH þ 10 wt% EtOH.
electrode [37]. The peak reaches a maximum and sharply decreases to zero when palladium oxidizes completely, deactivating the ethanol oxidation reaction [38]. Between 0 and 80% Co, the peak current density increases monotonously up to above 20 mA cm2, which should be compared to the 2 mA cm2 peak observed on pure Pd. There is also a shift in the position of this peak towards more positive potentials with increasing Co content, suggesting that the passivation of Pd is delayed in the presence of Co, which oxidizes preferentially. A pure Co film did not exhibit any significant current for ethanol oxidation, but a Co oxidation peak at 1.0 VRHE was detected. In the literature on ethanol oxidation, the maximum of the oxidation peak is often assumed to be representative of the electrocatalytic activity of the electrode [39e41]. However, experiments performed on working DEFC devices with Pt or PteSn catalysts show open cell voltages of 0.9 V, while peak power density occurs at a cell voltage of 0.3e0.4 V [42,43], suggesting that the ethanol oxidation in practice may be run at about 0.6e0.8 VRHE. That is close to the onset of the ethanol oxidation wave. For this reason, it may be useful to introduce an alternative benchmark for the catalytic activity towards ethanol oxidation, which could be for example the onset potential for the ethanol oxidation wave, or the potential where the oxidation current reaches, say 0.1 or 1 mA cm2, respectively representative of a significant oxidation current and of a fair performance of the fuel cell.
Fig. 4. (a) Representative potential for ethanol oxidation as a function of catalyst composition calculated by taking a linear extrapolation to the left side of the EtOH peak and the potential where 0.1 mA cm2 and 1 mA cm2 are reached, respectively. (b) Potentiostatic measurements at PdeCo films in ethanolic solutions show that the current density over the long term is optimized at a composition of PdeCo close to 50%.
L.-k. Tsui et al. / Journal of Power Sources 293 (2015) 815e822
conditions. Potentiostatic measurements on PdeCo films taken at 0.7 VRHE are shown in Fig. 4(b). Pure Pd shows the weakest currents, likely due to its susceptibility to passivation by the conversion of Pd-OHads to PdO at potentials close to the onset of ethanol oxidation [17]. In comparison, the higher currents observed at the PdeCo alloys show that an increase in Co fraction is beneficial to the alloy operation as a DEFC anode. In contrast to the CV data, the potentiostatic measurements suggest that under practical fuel cell operating conditions, the best performing PdeCo films should have a composition close to 50% Co. In order to fully understand the difference between the potentiodynamic and potentiostatic measurements, which respectively predict optimal performance for high Co concentration and around 50% Co, it is instructive to study the morphological and phase evolution during cycling and potentiostatic measurements. 3.3. Morphological and phase changes during electrochemical testing Fig. 5 shows the morphology of 55 at% Co, PdeCo film after undergoing potential cycling in KOH or KOH þ EtOH for 5 to 20 cycles and between 0.1 and 1.3 VRHE. The surface appears to become rougher in KOH solutions, consistent with the increase in the intensity of the PdO reduction peak seen in Fig. 3(a). For the EtOHcontaining solution in contrast, the formation of flake-like structures initiates as early as 5 cycles, growing in size and completely covering the surface of the sample after 15 cycles. EDS was performed to measure the compositional changes as a function of cycling. A 2% decrease in Co content relative to Pd was recorded after 20 cycles in a solution containing KOH only, and a 4% decrease in Co fraction was recorded in KOH with ethanol; these values are consistent with an average loss of 3% Co to soluble Co(OH)2 species as reported by Erts et al. for a concentrated KOH solution [44]. The
819
limited compositional variation indicates that no significant dealloying of the films occurs. EDS compositional mapping and semiquantitative determination of oxygen content (Fig. 6(a)) in the flake-like structures reveals that the presence of EtOH accelerates significantly the formation of an oxide, resulting in a stoichiometry close to Co3O4 in as little as 5 cycles. The mapping of Fig. 6(b) indicates that the flake-like structures are composed of Co and O, with little or no Pd incorporated into the flakes. Raman spectra (Fig. 7) were acquired to identify the phases of the cobalt oxides present after cycling. Raman spectra for Co3O4 are located at 197, 485, 620, and 691 cm1, while a characteristic peak at 505 cm1 is associated with CoOOH [45]. Weak, broad peaks are found after cycling in EtOH-free alkaline solution, whereas strong peaks indicating well crystallized oxide phases are found when cycling in KOH þ EtOH solutions. Co oxidation in the EtOH-containing solution generates a current that adds to that of ethanol oxidation, and its extent should be estimated in order to correct the intensity of the ethanol oxidation peak. The fraction of current responsible for Co oxidation can be calculated from the amount of Co being oxidized; this was estimated on a 75 at% Co alloy sample after cycling for 4 cycles in KOH þ EtOH. Fig. S2 compares the cross sectional SEM images of this sample before and after cycling, from which a change in thickness of the compact film layer from 410 to 350 nm was determined. This corresponds to an affected portion of 60 nm of the original film which was converted into flake-like Co oxides. Assuming as a first approximation that all the Co had been completely oxidized to Co3O4 with an average valence state of þ2.67 for each Co atom, the total charge required to carry out this oxidation is 0.16 C cm2. Since there may also be some partially oxidized Co within these flakes, this estimate should be treated as an upper bound. Negligible cathodic current is observed during the reverse scans outside of the area where PdO is reduced back to Pd,
Fig. 5. Morphology changes in a 55% Co film after cycling between 5 and 20 cycles in (a) KOH and (b) KOH þ EtOH.
820
L.-k. Tsui et al. / Journal of Power Sources 293 (2015) 815e822
Fig. 6. (a) Oxygen content as a function of the number of potential cycles in KOH and EtOH, showing rapid oxidation with EtOH. (b) EDS elemental mapping shows segregation of Co and O to flake like structures.
indicating that little of the Co-oxide is reduced back to Co. Given that the total accumulated charge is 0.99 C cm2 during the first four cycles, we estimate that 16% of this charge is used for Co oxidation. Since the anodic peak in the reverse scan is known to correspond to the oxidation of intermediates remaining on the surface, if we only count the forward scans towards the accumulated charge, 0.77 C cm2 of charge passed through the cell and 21% of this would be used for Co oxidation. We conclude that this correction is significant and, considering that the extent of Co oxidation may vary with alloy composition, it may change the previous ranking of electrocatalytic activity. On the other hand, the morphological evolution (Fig. 8) of PdeCo films tested under potentiostatic conditions is quite limited. These surfaces closely resemble the rough structures seen after cycling in EtOH-free electrolytes, indicating that no large scale reorganization of the surface occurs during potentiostatic polarization. The oxygen content of these samples after testing is high, suggesting that significant oxidation has occurred, but this fraction decreases with increasing Co content (Fig. S3), resulting in 30 at% O (O/(Co þ O)) above 35 at% Co, significantly below the Pd-rich samples. For comparison, the oxygen content in the potentiostatically biased samples is half the value observed during cycling in the EtOH electrolyte. These results show that assessing ethanol oxidation behavior using cyclic voltammetry methods is not recommended. Indeed, the PdeCo surface evolves into a flake-like structure at high anodic potential. Instead, since little change is recorded in the morphology, the potentiostatic method yields more representative results, and therefore, we can state based on these results that the optimum PdeCo content lies close to 50%.
Fig. 7. Raman spectra of PdeCo films (a) show the slow formation of Co3O4 after repeated potential cycling in KOH, and (b) the rapid formation of a mixed Co3O4 þ CoOOH phase structure after cycling in KOH þ EtOH.
In general, it is possible to surmise that PdeCo electrodes for ethanol oxidation perform better than pure Pd films and are relatively stable in alkaline solutions, forming oxides but otherwise showing limited degradation over time; this is valid however only under potentiostatic control; during transients in the cell it may well be possible that large shifts in the electrode potential may lead to a rearrangement on the anode material, resulting in performance changes over time.
4. Conclusion PdeCo films in a wide compositional range between 20 and 80% Co were synthesized by electrodeposition and tested for their catalytic activity towards ethanol oxidation and their operational stability over time. Although testing of the films by cyclic voltammetry seems to indicate that more Co produces higher ethanol oxidation current, potentiostatic measurements operating under conditions close to practical fuel cell operation evidence that a PdeCo film containing 50 at% Co has the highest performance, showing a sustained current density of 0.6 mA cm2 over 30 min. The onset potential at CoePd alloys is significantly lower than the
L.-k. Tsui et al. / Journal of Power Sources 293 (2015) 815e822
821
Fig. 8. Morphology of CoePd films of varying composition after a 30 min potentiostatic measurement at 0.7 VRHE. Flake like particles only form at high Co content and are smaller than those found during cycling, indicating greater stability.
onset potential of bare Pd, resulting in a lower overpotential needed to drive high currents. The shift towards more negative potentials for the onset and the higher sustained current density of the PdeCo alloys are attributed to the presence of Co partly transforming into oxides, thus supplying eOH to the ethanol oxidation reaction to accelerate acetate formation. These findings could be likely exploited in the design of highly performant anodic catalysts for the direct ethanol fuel cell technology. Acknowledgments We acknowledge the ARCS Foundation and NSF through the grant CMMI 1131571 for their financial support. Raman spectroscopy studies were supported through National Science Foundation Grant no. CMMI-1229603. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.jpowsour.2015.05.121. References [1] X. Teng, Anodic catalyst design for the ethanol oxidation fuel cell reactions, in: A. Mendez-Vilas (Ed.), Mater. Process. Energy Commun. Curr. Res. Technol. Dev., Formatex, 2013, pp. 473e484. ger, Recent [2] C. Lamy, A. Lima, V. LeRhun, F. Delime, C. Coutanceau, J.-M. Le advances in the development of direct alcohol fuel cells (DAFC), J. Power Sources 105 (2002) 283e296, http://dx.doi.org/10.1016/S0378-7753(01) 00954-5. [3] C. Lamy, E. Belgsir, J. Leger, Electrocatalytic oxidation of aliphatic alcohols: application to the direct alcohol fuel cell (DAFC), J. Appl. Electrochem. 31 (2001) 799e809, http://dx.doi.org/10.1023/A:1017587310150. [4] E. Peled, T. Duvdevani, A. Aharon, A. Melman, New fuels as alternatives to methanol for direct oxidation fuel cells, Electrochem. Solid-State Lett. 4 (2001) A38, http://dx.doi.org/10.1149/1.1355036. ger, Recent progress in [5] C. Lamy, S. Rousseau, E. Belgsir, C. Coutanceau, J.-M. Le the direct ethanol fuel cell: development of new platinumetin electrocatalysts, Electrochim. Acta 49 (2004) 3901e3908, http://dx.doi.org/10.1016/ j.electacta.2004.01.078. [6] E.H. Yu, U. Krewer, K. Scott, Principles and materials aspects of direct alkaline alcohol fuel cells, Energies 3 (2010) 1499e1528, http://dx.doi.org/10.3390/ en3081499. [7] Y.X. Chen, A. Lavacchi, H.A. Miller, M. Bevilacqua, J. Filippi, M. Innocenti, et al., Nanotechnology makes biomass electrolysis more energy efficient than water electrolysis, Nat. Commun. 5 (2014) 4036, http://dx.doi.org/10.1038/ ncomms5036. [8] V. Bambagioni, M. Bevilacqua, C. Bianchini, J. Filippi, A. Lavacchi, A. Marchionni, et al., Self-sustainable production of hydrogen, chemicals, and energy from renewable alcohols by electrocatalysis, ChemSusChem. 3 (2010) 851e855, http://dx.doi.org/10.1002/cssc.201000103. [9] Y.-X. Chen, A. Lavacchi, S.-P. Chen, F. di Benedetto, M. Bevilacqua, C. Bianchini, et al., Electrochemical milling and faceting: size reduction and catalytic activation of palladium nanoparticles, Angew. Chem. (Int. Ed. Eng.) 124 (2012) 8500e8504, http://dx.doi.org/10.1002/anie.201203589. [10] C. Bianchini, P.K. Shen, Palladium-based electrocatalysts for alcohol oxidation in half cells and in direct alcohol fuel cells, Chem. Rev. 109 (2009) 4183e4206,
http://dx.doi.org/10.1021/cr9000995. [11] J. Friedl, U. Stimming, Model catalyst studies on hydrogen and ethanol oxidation for fuel cells, Electrochim. Acta 101 (2013) 41e58, http://dx.doi.org/ 10.1016/j.electacta.2012.12.130. [12] Z.-Y. Zhou, Q. Wang, J.-L. Lin, N. Tian, S.-G. Sun, In situ FTIR spectroscopic studies of electrooxidation of ethanol on Pd electrode in alkaline media, Electrochim. Acta 55 (2010) 7995e7999, http://dx.doi.org/10.1016/ j.electacta.2010.02.071. [13] Z.X. Liang, T.S. Zhao, J.B. Xu, L.D. Zhu, Mechanism study of the ethanol oxidation reaction on palladium in alkaline media, Electrochim. Acta 54 (2009) 2203e2208, http://dx.doi.org/10.1016/j.electacta.2008.10.034. [14] A. Kowal, M. Li, M. Shao, K. Sasaki, M.B. Vukmirovic, J. Zhang, et al., Ternary Pt/ Rh/SnO2 electrocatalysts for oxidizing ethanol to CO2, Nat. Mater. 8 (2009) 325e330, http://dx.doi.org/10.1038/nmat2359. [15] M. Li, D. a Cullen, K. Sasaki, N.S. Marinkovic, K. More, R.R. Adzic, Ternary electrocatalysts for oxidizing ethanol to carbon dioxide: making ir capable of splitting CeC bond, J. Am. Chem. Soc. 135 (2013) 132e141, http://dx.doi.org/ 10.1021/ja306384x. [16] V. Rao, C. Cremers, U. Stimming, Investigation of the ethanol electro-oxidation in alkaline membrane electrode assembly by differential electrochemical mass spectrometry, Fuel Cells 7 (2007) 417e423, http://dx.doi.org/10.1002/ fuce.200700026. [17] V. Bambagioni, C. Bianchini, Y. Chen, J. Filippi, P. Fornasiero, M. Innocenti, et al., Energy efficiency enhancement of ethanol electrooxidation on Pd-CeO(2)/ C in passive and active polymer electrolyte-membrane fuel cells, ChemSusChem. 5 (2012) 1266e1273, http://dx.doi.org/10.1002/cssc.201100738. [18] Y. Chen, M. Bellini, M. Bevilacqua, P. Fornasiero, A. Lavacchi, H.A. Miller, et al., Direct alcohol fuel cells: toward the power densities of hydrogen-fed proton exchange membrane fuel cells, ChemSusChem. 8 (2015) 524e533, http:// dx.doi.org/10.1002/cssc.201402999. [19] C.-M. Chang, M.-T. Hsieh, W.-C. Kang, T.-J. Whang, Study of the Coelectrodeposited Pd-Ni alloy thin film and its performance on ethanol electro-oxidation, J. Electrochem. Soc. 161 (2014) D552eD557, http://dx.doi.org/ 10.1149/2.0911410jes. [20] K.S. Kumar, P. Haridoss, S.K. Seshadri, Synthesis and characterization of electrodeposited NiePd alloy electrodes for methanol oxidation, Surf. Coat. Technol. 202 (2008) 1764e1770, http://dx.doi.org/10.1016/ j.surfcoat.2007.07.035. [21] T. Maiyalagan, K. Scott, Performance of carbon nanofiber supported PdeNi catalysts for electro-oxidation of ethanol in alkaline medium, J. Power Sources 195 (2010) 5246e5251, http://dx.doi.org/10.1016/j.jpowsour.2010.03.022. [22] Z. Qi, H. Geng, X. Wang, C. Zhao, H. Ji, C. Zhang, et al., Novel nanocrystalline PdNi alloy catalyst for methanol and ethanol electro-oxidation in alkaline media, J. Power Sources 196 (2011) 5823e5828, http://dx.doi.org/10.1016/ j.jpowsour.2011.02.083. [23] X. Wang, G. Ma, F. Zhu, N. Lin, B. Tang, Z. Zhang, Preparation and characterization of micro-arc-induced Pd/TM(TM ¼ Ni, Co and Ti) catalysts and comparison of their electrocatalytic activities toward ethanol oxidation, Electrochim. Acta 114 (2013) 500e508, http://dx.doi.org/10.1016/ j.electacta.2013.10.049. [24] C. Xu, Z. Tian, P. Shen, S.P. Jiang, Oxide (CeO2, NiO, Co3O4 and Mn3O4)-promoted Pd/C electrocatalysts for alcohol electrooxidation in alkaline media, Electrochim. Acta 53 (2008) 2610e2618, http://dx.doi.org/10.1016/ j.electacta.2007.10.036. [25] D. Liu, Q. Guo, H. Hou, O. Niwa, T. You, Pd x Co y nanoparticle/carbon nanofiber composites with enhanced electrocatalytic properties, ACS Catal. 4 (2014) 1825e1829, http://dx.doi.org/10.1021/cs5000153. n, A. Heinzel, Estimation of the membrane methanol diffusion co[26] V. Barraga efficient from open circuit voltage measurements in a direct methanol fuel cell, J. Power Sources 104 (2002) 66e72, http://dx.doi.org/10.1016/S03787753(01)00896-5. _ ski, R. Kowalik, K. Fitzner, Analysis of CoePd alloys deposi[27] K. Mech, P. Zabi n tion from electrolytes based on [Co(NH3)6]3þ and [Pd(NH3)4]2þ complexes, Electrochim. Acta 104 (2013) 468e473, http://dx.doi.org/10.1016/
822
L.-k. Tsui et al. / Journal of Power Sources 293 (2015) 815e822
j.electacta.2012.12.006. [28] International Centre for Diffraction Data, ICDD DDViewþ, Version 4.8.0.5, Newtown Square, PA, USA, 2009. [29] K. Ishida, T. Nishizawa, The Co-Pd (Cobalt-Palladium) System, J. Phase Equilib. 12 (1991) 83e87, http://dx.doi.org/10.1007/BF02663680. lisle, P. Chartrand, S.A. Decterov, G. Eriksson, K. Hack, et al., [30] C.W. Bale, E. Be FactSage thermochemical software and databases - recent developments, Calphad Comput. Coupling Phase Diagrams Thermochem. 33 (2009) 295e311, http://dx.doi.org/10.1016/j.calphad.2008.09.009. [31] Co-Pd: Data from FSnobl e Fact Sage Noble Metal Databases, Cent. Res. Comput. Thermochem. http://www.crct.polymtl.ca/fact/phase_diagram.php? file¼Co-Pd.jpg&dir¼FSnobl (accessed 09.03.15) [32] Y. Matsuo, Ordered alloys in the cobaltepalladium system, J. Phys. Soc. Jpn. 32 (1972) 972e978. [33] S. Tominaka, T. Osaka, Nanoporous PdCo catalyst for microfuel cells: electrodeposition and dealloying, Adv. Phys. Chem. 2011 (2011) 1e13, http:// dx.doi.org/10.1155/2011/821916. [34] G. Zangari, B. Bozzini, P.L. Cavallotti, G. Fontana, P.G. Maisto, E. Terrenzio, Coercivity and microstructure in Co-Pt and Co-Pd thin films, J. Magn. Magn. Mater. 133 (1994) 511e515, http://dx.doi.org/10.1016/0304-8853(94)906106. ski, The adsorption of carbon oxides on a palladium electrode from [35] A. Czerwin acidic solution,, J. Electroanal. Chem. 379 (1994) 487e493, http://dx.doi.org/ 10.1016/0022-0728(94)87173-6. , M. Łukaszewski, G. Jerkiewicz, A. Czerwin ski, Electrochemical [36] M. Grden behaviour of palladium electrode: oxidation, electrodissolution and ionic adsorption, Electrochim. Acta 53 (2008) 7583e7598, http://dx.doi.org/ 10.1016/j.electacta.2008.05.046. [37] J. Liu, J. Ye, C. Xu, S.P. Jiang, Y. Tong, Kinetics of ethanol electrooxidation at Pd electrodeposited on Ti, Electrochem. Comm. 9 (2007) 2334e2339, http://
dx.doi.org/10.1016/j.elecom.2007.06.036. [38] Z.X. Liang, T.S. Zhao, J.B. Xu, L.D. Zhu, Mechanism study of the ethanol oxidation reaction on palladium in alkaline media, Electrochim. Acta 54 (2009) 2203e2208, http://dx.doi.org/10.1016/j.electacta.2008.10.034. [39] Z. Zhang, C. Zhang, J. Sun, T. Kou, Q. Bai, Y. Wang, et al., Ultrafine nanoporous PdFe/Fe3O4 catalysts with doubly enhanced activities towards electrooxidation of methanol and ethanol in alkaline media, J. Mater. Chem. A 1 (2013) 3620, http://dx.doi.org/10.1039/c3ta01464a. [40] Q. Yi, L. Sun, X. Liu, H. Nie, Palladiumenickel nanoparticles loaded on multiwalled carbon nanotubes modified with b-cyclodextrin for electrooxidation of alcohols, Fuel 111 (2013) 88e95, http://dx.doi.org/10.1016/ j.fuel.2013.04.051. [41] C. Qiu, R. Shang, Y. Xie, Y. Bu, C. Li, H. Ma, Electrocatalytic activity of bimetallic PdeNi thin films towards the oxidation of methanol and ethanol, Mater. Chem. Phys. 120 (2010) 323e330, http://dx.doi.org/10.1016/ j.matchemphys.2009.11.014. ger, Direct ethanol fuel cell [42] S. Rousseau, C. Coutanceau, C. Lamy, J.-M. Le (DEFC): electrical performances and reaction products distribution under operating conditions with different platinum-based anodes, J. Power Sources 158 (2006) 18e24, http://dx.doi.org/10.1016/j.jpowsour.2005.08.027. [43] J. Mann, N. Yao, A.B. Bocarsly, Characterization and analysis of new catalysts for a direct ethanol fuel cell, Langmuir 22 (2006) 10432e10436, http:// dx.doi.org/10.1021/la061200c. €rnsson, H. Olin, J. Prikulis, Studies of the initial [44] D. Erts, E. Ahlberg, J. Asbjo oxidation of cobalt in alkaline solutions using scanning electrochemical microscope, App. Phys. A 66 (1998) 477e480, http://dx.doi.org/10.1007/ s003390051186. [45] B.S. Yeo, A.T. Bell, Enhanced activity of gold-supported cobalt oxide for the electrochemical evolution of oxygen, J. Amer. Chem. Soc. 133 (2011) 5587e5593, http://dx.doi.org/10.1021/ja200559j.
Contents lists available at ScienceDirect
Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour
Electrocatalytic activity and operational stability of electrodeposited PdeCo films towards ethanol oxidation in alkaline electrolytes Lok-kun Tsui a, Claudio Zafferoni b, Alessandro Lavacchi c, Massimo Innocenti b, Francesco Vizza c, Giovanni Zangari a, * a b c
Department of Materials Science and Engineering and CESE, University of Virginia, 395 McCormick Rd, Charlottesville, VA 22904, USA Department of Chemistry, University of Florence, Via della Lastruccia 3-13, 50019 Florence, Italy ICCOM-CNR, Via Madonna del Piano 10, 50019 Florence, Italy
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
PdeCo catalysts were electrodeposited with compositions between 20 and 80 at% Co. A transformation of alloyed Co to CoOOH/Co3O4 flakes was observed after cycling. The Co in the films supplies eOH to EtOH oxidation, reducing onset potential. A max peak current density is found at an optimum composition of 77 at% Co. A 50 at% CoePd film exhibited the highest stable currents among PdeCo alloys.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 9 March 2015 Received in revised form 7 May 2015 Accepted 30 May 2015 Available online 11 June 2015
Direct alkaline ethanol fuel cells (DEFCs) are usually run with Pd anodic catalysts, but their performance can be improved by utilizing alloys of Pd and Co. The oxyphilic Co serves to supply ample eOH to the ethanol oxidation reaction, accelerating the rate limiting step at low overpotential under alkaline conditions. PdeCo films with compositions between 20 and 80 at% Co can be prepared by electrodeposition from a NH3 complexing electrolyte. Cyclic voltammetry studies show that the ethanol oxidation peak exhibits increasing current density with increasing Co content, reaching a maximum at 77% Co. In contrast, potentiostatic measurements under conditions closer to fuel cell operating conditions show that a 50 at% Co alloy has the highest performance. Importantly, the CoePd film is also found to undergo phase and morphological transformations during ethanol oxidation, resulting in a change from a compact film to high surface area flake-like structures containing Co3O4 and CoOOH; such a transformation instead is not observed when operating at a constant potential of 0.7 VRHE. © 2015 Elsevier B.V. All rights reserved.
Keywords: Direct ethanol fuel cell Ethanol oxidation Anode catalyst Palladiumecobalt alloy Cobalt oxide Electrodeposition
1. Introduction
* Corresponding author. E-mail address: [email protected] (G. Zangari). http://dx.doi.org/10.1016/j.jpowsour.2015.05.121 0378-7753/© 2015 Elsevier B.V. All rights reserved.
Direct Ethanol Fuel Cells (DEFCs) present various attractive features as power sources for portable and, more recently, transportation applications, since they exhibit energy densities
816
L.-k. Tsui et al. / Journal of Power Sources 293 (2015) 815e822
comparable to gasoline and, in contrast to hydrogen, benefit from a simpler and mostly existing storage and transport infrastructure [1,2]. The use of ethanol, as well as other high molecular weight alcohols such as ethylene glycol or glycerol, is advantageous with respect to methanol or formic acid feeds, due to their low toxicity [3,4], and the possibility to produce these alcohols from biomass (sugar- or cellulose-based) using renewable, CO2-neutral methods [5,6]. The main drawback of these devices however derives from the sluggish kinetics of ethanol oxidation, much slower than that of lighter alcohols or H2, which results in lower efficiency and inferior performance with respect to the corresponding fuel cells. The development of efficient and selective anodic electrocatalysts is necessary to overcome this hurdle. These anodic electrocatalysts could also be successfully exploited in the electrochemical reforming of bio-alcohols providing a sustainable route for the cogeneration of hydrogen and important raw chemicals [7e9]. Platinum catalysts are commonly used in acidic electrolytes because of their excellent ability to enhance the adsorption and dissociation of small organics molecules (SOMs); however, strongly adsorbed organic species such as carbonyl and CO, tend to poison the catalyst [1]. Furthermore, the high cost of Pt contributes significantly (~50%) to the total fuel cell stack cost. In contrast, in alkaline media the kinetics of ethanol oxidation is faster, and the adsorption strength of organic fragments is weaker. Importantly, less expensive electrocatalysts such as Pd can be used due to their stability in alkaline environments [10]. In an alkaline-type DEFC the reaction at the anode is targeted to be the complete oxidation of ethanol to CO2 (1), accompanied by the reduction of oxygen at the cathode (2), leading to the overall reaction of ethanol “combustion” (3) [6].
CH3 CH2 OH þ 16OH / 2CO3 2 þ 11H2 O þ 12e
(1)
3O2 þ 6H2 O þ 12e / 12OH
(2)
CH3 CH2 OH þ 3O2 þ 4OH / 5H2 O þ 2CO3 2
(3)
Reaction (1) consists of a series of intermediate steps, whereby adsorbed ethanol is oxidized via hydrogen abstraction to acetaldehyde; the latter may subsequently be oxidized further to CO2, if the CeC bond is broken, or transformed into a terminal acetate product (Fig. 1) [11,12]. The rate determining step in this reaction mechanism is the adsorption of hydroxyl (OHads) at low overpotential, or the removal of the adsorbed ethoxy (CH3COads) by adsorbed hydroxyl at higher potentials [13]. The overall reaction could be accelerated by addition of a more oxyphilic metal to Pt or Pd, which would enhance water dissociation and the formation of oxygenated species that become available for further oxidation to acetic acid. Most studied in this context are Ru and Sn, with the latter providing higher oxidation currents [14]; Co on the other
hand could provide the same function at a lower cost with respect to Ru, while possibly enhancing CeC bond breaking, due to the homologous electronic structure to Rh and Ir, both known to enhance selectivity towards complete oxidation [14,15]. The overall selectivity for ethanol complete oxidation is found to be generally low, with the exception of Rao et al. [16], who reported a 55% CO2 conversion efficiency in alkaline solutions and only 2% in acidic ones. CH3COO- is in most cases by far the major product of ethanol oxidation in alkaline environment, decreasing the faradaic efficiency by 1/3 as compared to carbonate formation. Despite this intrinsic loss of faradaic efficiency, the fast kinetics of acetate production on Pd results in relatively low anodic overpotential, providing DEFCs with power densities up to 335 mW cm2 [17,18]. Pd-Ni [19e22] and PdeCo [23e25] have been already investigated as anodic catalysts for alkaline DEFCs, each showing better performance than Pd alone when normalized per unit area. The alkaline environment allows complete elimination of Pt from these fuel cells. Indeed in alkaline environment iron and cobalt macrocycles are active for the oxygen reduction with similar performance to Pt based ORR catalysts, these materials being also inactive toward alcohol oxidation This is particularly relevant as it results in negligible voltage drop for alcohol crossover [26]. The studies referred to above speculate that the facilitated formation of eOH on the 3d transition metal in alkaline solutions is the main mechanism by which ethanol oxidation was enhanced. Liu et al. identified the equiatomic composition to be optimum for formic acid and methanol oxidation at PdeCo nanoparticles supported on carbon fibers [25]. Xu et al. mixed Pd nanoparticles 6e10 nm with oxides of Ce, Ni, Co, and Mn, and found that a Co3O4ePd composite with an oxide content of 0.05 mg cm2 yields the highest peak oxidation current for ethanol oxidation, while noting low stability for this oxide-Pd pairing [24]. So far however, the optimum features of the catalyst structure and composition for ethanol oxidation have not been assessed; in addition, the stability of the catalyst surface has received little to no attention. On a related issue, a rigorous protocol for the evaluation of the electrocatalytic activity in ethanol oxidation catalysts has been missing and current methods to compare catalyst activity are not standardized, hindering an accurate ranking of the relative activity of various catalysts. In this work, we report on the electrocatalytic activity and longterm stability of PdeCo alloy films in view of their potential application as anodic catalysts for alkaline DEFCs. We produce a set of thin film electrodes with a broad range of compositions, using electroplating from an ammonia-based electrolyte, and we evaluate different methods to rank activity towards ethanol oxidation. In the process, we identify an optimum alloy composition at a Co fraction around 50e60%, yielding the highest current densities for ethanol oxidation and also find that the reaction onset occurs at more negative potentials with increasing Co fraction. Accelerated oxidation of Co in the alloy to form Co (hydro)oxides is observed during the potential cycling of PdeCo in alkaline, ethanolcontaining electrolytes, which is accompanied by a morphological transformation to a high surface area flake-like structure, never explicitly reported before. 2. Experimental 2.1. Electrodeposition of PdeCo
Fig. 1. Simplified reaction mechanism for ethanol oxidation; the reaction can either lead to CO2 if the CeC bond is broken or end at acetate as a terminal product [11,12].
An ammonia-based complexing electrolyte was utilized in order to shift the redox potentials of the two metals closer together and thus facilitate the formation of alloy solid solutions [27]. The electrolyte was prepared as follows: first, two beakers were prepared, one containing a 20 mM PdCl2, 0.4 M (NH4)2SO4 and 0.5 M NH3
L.-k. Tsui et al. / Journal of Power Sources 293 (2015) 815e822
817
aqueous solution, and the other containing a 50 mM CoSO4, 0.4 M (NH4)2SO4 and 0.5 M NH3 solution. The latter solution was stirred until it turned black, indicating that the oxidation of Co(II) to Co(III) had reached completion. Oxidation of Co(II) to Co(III) prevents the spontaneous oxidation of Co(II) in the presence of Pd(II), which would reduce the Pd out of solution and hinder compositional control. Upon completion of the Co oxidation, the two solutions were mixed together to form the plating electrolyte. PdeCo films were deposited under current control on Au/Si substrates with a constant current density between 0.25 and 2.0 mA cm2 to a fixed charge density of 1.0 C cm2 (corresponding to 275e400 nm of PdeCo films as measured by cross sectional SEM). Pure Pd films were prepared in a solution of 10 mM PdCl2, 0.3 M (NH4)2SO4, and 0.5 M C7H6O6S as complexing agents at a current density of 1 mA cm2 to a fixed charge density of 1 C cm2 (460 nm). Pure Co films were instead obtained by potentiostatic deposition at 1.4 VSCE for 3 min in a solution of 0.3 M CoSO4 þ 0.25 M H2SO4 yielding a Co film of approximately 700 nm. 2.2. Electrocatalytic activity Electrochemical measurements were carried out in solutions containing 2 M KOH or 2 M KOH with 10 wt% ethanol (Alfa Aesar, 95þ%). A three electrode cell configuration was used which included a Pt mesh counter electrode, an Hg/HgO reference electrode (CH Instruments), and the working electrode. The onset potential for ethanol oxidation was determined by two distinct methods: (i) a linear intercept was fitted to the left side of the oxidation peak with the background current; alternately, (ii) the potential at which the current density exceeded 0.1 or 1 mA cm2 was recorded. 2.3. Materials characterization The morphology and composition of the PalladiumeCobalt films were characterized by scanning electron microscopy (SEM) and energy dispersive x-ray spectroscopy (EDS) in an FEI Quanta 650 SEM, respectively. Phase identification was carried out using Raman spectroscopy (Renishaw InVIA Raman Microscope) using a 488 nm laser with a 10s acquisition time. X-Ray Diffraction (XRD) was performed with Cu Ka radiation (l ¼ 0.154 nm) using a Panalytical X'PERT MPD x-ray diffractometer. Reference patterns were obtained from the ICDD PDF DDView þ database [28]. 3. Results and discussion 3.1. Composition of PdeCo films Fig. S1 shows the composition as a function of current density for PdeCo films grown on Au substrates by galvanostatic electrodeposition. Varying the applied current density between 0.25 and 1.5 mA cm2, the composition varies from 20% up to 80% Co. Beyond 1.5 mA cm2, the Co content saturates at 80%. Such a broad composition window allows us to explore in detail the effect of Co content on the ethanol oxidation activity of electrodeposited PdeCo films. The experimental phase diagram for bulk PdeCo shows a continuous series of face-centered cubic solid solutions, i.e. true alloys, with a possible transition to Co-rich hexagonal phases below 422 C and less than ~20 at% Pd [29]. In contrast, the FactSage database shows a miscibility gap below ~600 C and between 0 and 50 at% Pd [30,31], and only ordered structures of the L12 and L10 types are reported for Co > 25 at% [32,33]. Fig. 2 shows the XRD patterns obtained on the alloy films and the pure Pd and Co films for comparison. Besides the substrate peaks, all PdeCo films that we obtained show a peak at 41, located between the FCC Pd (111)
Fig. 2. XRD Patterns of electrodeposited PdeCo films along with pure Pd and pure Co films show, with increasing Pd content, an FCC alloy peak around 41 followed by the formation of a second phase above 46% Co, identified as HCP Co.
and HCP Co (100) peak. The position of this peak does not shift with composition; assuming a cubic lattice, the corresponding lattice constant is 0.380 nm ± 0.002 nm. As expected, this value lies between the lattice constants for FCC Pd (0.389 nm, PDF 00-0461043) and FCC Co (0.354 nm, PDF 00-015-0806), and corresponds approximately to a solid solution containing ~ 35 at% Co [34]. At and above 46 at% Co, new peaks emerge at 2q ¼ 41.6 and 44.5 , which are assigned to hexagonal Co. At high overvoltage, the probability of nucleation and growth of Co seeds in parallel with the solid solution is possible, resulting in a kinetically induced formation of a pure Co phase, and in two-phase growth; interestingly, the alloy peak shifts to lower angles when the pure Co phase forms, suggesting a Pd enrichment in the cubic phase. 3.2. Electrocatalytic activity of PdeCo alloy films Fig. 3(a) shows the cyclic voltammetry (CV) response over six successive cycles at a 50 at% Co, PdeCo film in 2 M KOH; this behavior is representative of all the alloy compositions investigated. Besides characterizing the electrochemical response, these data also enable the determination of the electrochemically active ski [35]. Fig. 3(b) surface area (EASA) of Pd as reported by Czerwin displays the CV response of PdeCo films in the full range of compositions between 0 and 100% Co in 2 M KOH and 10 wt% ethanol. In the KOH electrolyte (Fig. 3(a)), scanning towards positive potentials, a peak corresponding to the oxidation of cobalt is observed at about 1.0 VRHE, that is not present in pure Pd films; this is followed at higher potentials by Pd oxidation. On the return scan, two peaks corresponding to the sequential reduction of a monolayer of Pd oxide to Pd are observed, near 1.0 and 0.7 VRHE, respectively, followed by hydrogen adsorption and then hydrogen evolution at more negative potentials [36]. With repeated cycles, the peaks corresponding to PdO reduction and Co oxidation increase in intensity, suggesting a corresponding increase in surface area during cycling. When 10 wt% ethanol is added to the electrolyte (Fig. 3(b)), a new, much stronger asymmetric peak corresponding to the oxidation of ethanol appears centered at 0.8 to 1 VRHE; a corresponding smaller anodic peak on the return scan is associated with the oxidation of intermediate products still adsorbed on the alloy
818
L.-k. Tsui et al. / Journal of Power Sources 293 (2015) 815e822
Fig. 4(a) shows the three potentials thus calculated for PdeCo films as a function of composition. Using any of these methods, with more Co in the films, the reference potential for ethanol oxidation shifts towards more negative values. The linear intercept method gives a decrease of 100 mV with higher Co content across the various CoePd alloys, but taking the reference potential to be the value at which 0.1 mA cm2 is reached gives a decrease of 200 mV, while taking a measurement at 1.0 mA cm2 gives a decrease of 300 mV. No onset potential could be extracted from the pure Co film because no ethanol oxidation is detected. These results suggest that an increase of the Co fraction in the alloy enhances the oxidation peak intensity and reduces the overpotential needed to activate the ethanol oxidation current, favoring an improved performance. On the other hand, the various benchmarks used to rank the electrocatalytic activity of CoePd alloys result in the same ranking, but differ in the estimation of the relative activity. Since actual fuel cell operating conditions are closer to points at the onset of the oxidation wave rather than near the peak, it is important to monitor potentiostatic current transients under these
Fig. 3. (a) Cyclic voltammetry at 50 at% Co electrodeposited films over several cycles in ethanol free 2 M KOH and (b) 5th cycle of CV at PdeCo films of various compositions in 2 M KOH þ 10 wt% EtOH.
electrode [37]. The peak reaches a maximum and sharply decreases to zero when palladium oxidizes completely, deactivating the ethanol oxidation reaction [38]. Between 0 and 80% Co, the peak current density increases monotonously up to above 20 mA cm2, which should be compared to the 2 mA cm2 peak observed on pure Pd. There is also a shift in the position of this peak towards more positive potentials with increasing Co content, suggesting that the passivation of Pd is delayed in the presence of Co, which oxidizes preferentially. A pure Co film did not exhibit any significant current for ethanol oxidation, but a Co oxidation peak at 1.0 VRHE was detected. In the literature on ethanol oxidation, the maximum of the oxidation peak is often assumed to be representative of the electrocatalytic activity of the electrode [39e41]. However, experiments performed on working DEFC devices with Pt or PteSn catalysts show open cell voltages of 0.9 V, while peak power density occurs at a cell voltage of 0.3e0.4 V [42,43], suggesting that the ethanol oxidation in practice may be run at about 0.6e0.8 VRHE. That is close to the onset of the ethanol oxidation wave. For this reason, it may be useful to introduce an alternative benchmark for the catalytic activity towards ethanol oxidation, which could be for example the onset potential for the ethanol oxidation wave, or the potential where the oxidation current reaches, say 0.1 or 1 mA cm2, respectively representative of a significant oxidation current and of a fair performance of the fuel cell.
Fig. 4. (a) Representative potential for ethanol oxidation as a function of catalyst composition calculated by taking a linear extrapolation to the left side of the EtOH peak and the potential where 0.1 mA cm2 and 1 mA cm2 are reached, respectively. (b) Potentiostatic measurements at PdeCo films in ethanolic solutions show that the current density over the long term is optimized at a composition of PdeCo close to 50%.
L.-k. Tsui et al. / Journal of Power Sources 293 (2015) 815e822
conditions. Potentiostatic measurements on PdeCo films taken at 0.7 VRHE are shown in Fig. 4(b). Pure Pd shows the weakest currents, likely due to its susceptibility to passivation by the conversion of Pd-OHads to PdO at potentials close to the onset of ethanol oxidation [17]. In comparison, the higher currents observed at the PdeCo alloys show that an increase in Co fraction is beneficial to the alloy operation as a DEFC anode. In contrast to the CV data, the potentiostatic measurements suggest that under practical fuel cell operating conditions, the best performing PdeCo films should have a composition close to 50% Co. In order to fully understand the difference between the potentiodynamic and potentiostatic measurements, which respectively predict optimal performance for high Co concentration and around 50% Co, it is instructive to study the morphological and phase evolution during cycling and potentiostatic measurements. 3.3. Morphological and phase changes during electrochemical testing Fig. 5 shows the morphology of 55 at% Co, PdeCo film after undergoing potential cycling in KOH or KOH þ EtOH for 5 to 20 cycles and between 0.1 and 1.3 VRHE. The surface appears to become rougher in KOH solutions, consistent with the increase in the intensity of the PdO reduction peak seen in Fig. 3(a). For the EtOHcontaining solution in contrast, the formation of flake-like structures initiates as early as 5 cycles, growing in size and completely covering the surface of the sample after 15 cycles. EDS was performed to measure the compositional changes as a function of cycling. A 2% decrease in Co content relative to Pd was recorded after 20 cycles in a solution containing KOH only, and a 4% decrease in Co fraction was recorded in KOH with ethanol; these values are consistent with an average loss of 3% Co to soluble Co(OH)2 species as reported by Erts et al. for a concentrated KOH solution [44]. The
819
limited compositional variation indicates that no significant dealloying of the films occurs. EDS compositional mapping and semiquantitative determination of oxygen content (Fig. 6(a)) in the flake-like structures reveals that the presence of EtOH accelerates significantly the formation of an oxide, resulting in a stoichiometry close to Co3O4 in as little as 5 cycles. The mapping of Fig. 6(b) indicates that the flake-like structures are composed of Co and O, with little or no Pd incorporated into the flakes. Raman spectra (Fig. 7) were acquired to identify the phases of the cobalt oxides present after cycling. Raman spectra for Co3O4 are located at 197, 485, 620, and 691 cm1, while a characteristic peak at 505 cm1 is associated with CoOOH [45]. Weak, broad peaks are found after cycling in EtOH-free alkaline solution, whereas strong peaks indicating well crystallized oxide phases are found when cycling in KOH þ EtOH solutions. Co oxidation in the EtOH-containing solution generates a current that adds to that of ethanol oxidation, and its extent should be estimated in order to correct the intensity of the ethanol oxidation peak. The fraction of current responsible for Co oxidation can be calculated from the amount of Co being oxidized; this was estimated on a 75 at% Co alloy sample after cycling for 4 cycles in KOH þ EtOH. Fig. S2 compares the cross sectional SEM images of this sample before and after cycling, from which a change in thickness of the compact film layer from 410 to 350 nm was determined. This corresponds to an affected portion of 60 nm of the original film which was converted into flake-like Co oxides. Assuming as a first approximation that all the Co had been completely oxidized to Co3O4 with an average valence state of þ2.67 for each Co atom, the total charge required to carry out this oxidation is 0.16 C cm2. Since there may also be some partially oxidized Co within these flakes, this estimate should be treated as an upper bound. Negligible cathodic current is observed during the reverse scans outside of the area where PdO is reduced back to Pd,
Fig. 5. Morphology changes in a 55% Co film after cycling between 5 and 20 cycles in (a) KOH and (b) KOH þ EtOH.
820
L.-k. Tsui et al. / Journal of Power Sources 293 (2015) 815e822
Fig. 6. (a) Oxygen content as a function of the number of potential cycles in KOH and EtOH, showing rapid oxidation with EtOH. (b) EDS elemental mapping shows segregation of Co and O to flake like structures.
indicating that little of the Co-oxide is reduced back to Co. Given that the total accumulated charge is 0.99 C cm2 during the first four cycles, we estimate that 16% of this charge is used for Co oxidation. Since the anodic peak in the reverse scan is known to correspond to the oxidation of intermediates remaining on the surface, if we only count the forward scans towards the accumulated charge, 0.77 C cm2 of charge passed through the cell and 21% of this would be used for Co oxidation. We conclude that this correction is significant and, considering that the extent of Co oxidation may vary with alloy composition, it may change the previous ranking of electrocatalytic activity. On the other hand, the morphological evolution (Fig. 8) of PdeCo films tested under potentiostatic conditions is quite limited. These surfaces closely resemble the rough structures seen after cycling in EtOH-free electrolytes, indicating that no large scale reorganization of the surface occurs during potentiostatic polarization. The oxygen content of these samples after testing is high, suggesting that significant oxidation has occurred, but this fraction decreases with increasing Co content (Fig. S3), resulting in 30 at% O (O/(Co þ O)) above 35 at% Co, significantly below the Pd-rich samples. For comparison, the oxygen content in the potentiostatically biased samples is half the value observed during cycling in the EtOH electrolyte. These results show that assessing ethanol oxidation behavior using cyclic voltammetry methods is not recommended. Indeed, the PdeCo surface evolves into a flake-like structure at high anodic potential. Instead, since little change is recorded in the morphology, the potentiostatic method yields more representative results, and therefore, we can state based on these results that the optimum PdeCo content lies close to 50%.
Fig. 7. Raman spectra of PdeCo films (a) show the slow formation of Co3O4 after repeated potential cycling in KOH, and (b) the rapid formation of a mixed Co3O4 þ CoOOH phase structure after cycling in KOH þ EtOH.
In general, it is possible to surmise that PdeCo electrodes for ethanol oxidation perform better than pure Pd films and are relatively stable in alkaline solutions, forming oxides but otherwise showing limited degradation over time; this is valid however only under potentiostatic control; during transients in the cell it may well be possible that large shifts in the electrode potential may lead to a rearrangement on the anode material, resulting in performance changes over time.
4. Conclusion PdeCo films in a wide compositional range between 20 and 80% Co were synthesized by electrodeposition and tested for their catalytic activity towards ethanol oxidation and their operational stability over time. Although testing of the films by cyclic voltammetry seems to indicate that more Co produces higher ethanol oxidation current, potentiostatic measurements operating under conditions close to practical fuel cell operation evidence that a PdeCo film containing 50 at% Co has the highest performance, showing a sustained current density of 0.6 mA cm2 over 30 min. The onset potential at CoePd alloys is significantly lower than the
L.-k. Tsui et al. / Journal of Power Sources 293 (2015) 815e822
821
Fig. 8. Morphology of CoePd films of varying composition after a 30 min potentiostatic measurement at 0.7 VRHE. Flake like particles only form at high Co content and are smaller than those found during cycling, indicating greater stability.
onset potential of bare Pd, resulting in a lower overpotential needed to drive high currents. The shift towards more negative potentials for the onset and the higher sustained current density of the PdeCo alloys are attributed to the presence of Co partly transforming into oxides, thus supplying eOH to the ethanol oxidation reaction to accelerate acetate formation. These findings could be likely exploited in the design of highly performant anodic catalysts for the direct ethanol fuel cell technology. Acknowledgments We acknowledge the ARCS Foundation and NSF through the grant CMMI 1131571 for their financial support. Raman spectroscopy studies were supported through National Science Foundation Grant no. CMMI-1229603. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.jpowsour.2015.05.121. References [1] X. Teng, Anodic catalyst design for the ethanol oxidation fuel cell reactions, in: A. Mendez-Vilas (Ed.), Mater. Process. Energy Commun. Curr. Res. Technol. Dev., Formatex, 2013, pp. 473e484. ger, Recent [2] C. Lamy, A. Lima, V. LeRhun, F. Delime, C. Coutanceau, J.-M. Le advances in the development of direct alcohol fuel cells (DAFC), J. Power Sources 105 (2002) 283e296, http://dx.doi.org/10.1016/S0378-7753(01) 00954-5. [3] C. Lamy, E. Belgsir, J. Leger, Electrocatalytic oxidation of aliphatic alcohols: application to the direct alcohol fuel cell (DAFC), J. Appl. Electrochem. 31 (2001) 799e809, http://dx.doi.org/10.1023/A:1017587310150. [4] E. Peled, T. Duvdevani, A. Aharon, A. Melman, New fuels as alternatives to methanol for direct oxidation fuel cells, Electrochem. Solid-State Lett. 4 (2001) A38, http://dx.doi.org/10.1149/1.1355036. ger, Recent progress in [5] C. Lamy, S. Rousseau, E. Belgsir, C. Coutanceau, J.-M. Le the direct ethanol fuel cell: development of new platinumetin electrocatalysts, Electrochim. Acta 49 (2004) 3901e3908, http://dx.doi.org/10.1016/ j.electacta.2004.01.078. [6] E.H. Yu, U. Krewer, K. Scott, Principles and materials aspects of direct alkaline alcohol fuel cells, Energies 3 (2010) 1499e1528, http://dx.doi.org/10.3390/ en3081499. [7] Y.X. Chen, A. Lavacchi, H.A. Miller, M. Bevilacqua, J. Filippi, M. Innocenti, et al., Nanotechnology makes biomass electrolysis more energy efficient than water electrolysis, Nat. Commun. 5 (2014) 4036, http://dx.doi.org/10.1038/ ncomms5036. [8] V. Bambagioni, M. Bevilacqua, C. Bianchini, J. Filippi, A. Lavacchi, A. Marchionni, et al., Self-sustainable production of hydrogen, chemicals, and energy from renewable alcohols by electrocatalysis, ChemSusChem. 3 (2010) 851e855, http://dx.doi.org/10.1002/cssc.201000103. [9] Y.-X. Chen, A. Lavacchi, S.-P. Chen, F. di Benedetto, M. Bevilacqua, C. Bianchini, et al., Electrochemical milling and faceting: size reduction and catalytic activation of palladium nanoparticles, Angew. Chem. (Int. Ed. Eng.) 124 (2012) 8500e8504, http://dx.doi.org/10.1002/anie.201203589. [10] C. Bianchini, P.K. Shen, Palladium-based electrocatalysts for alcohol oxidation in half cells and in direct alcohol fuel cells, Chem. Rev. 109 (2009) 4183e4206,
http://dx.doi.org/10.1021/cr9000995. [11] J. Friedl, U. Stimming, Model catalyst studies on hydrogen and ethanol oxidation for fuel cells, Electrochim. Acta 101 (2013) 41e58, http://dx.doi.org/ 10.1016/j.electacta.2012.12.130. [12] Z.-Y. Zhou, Q. Wang, J.-L. Lin, N. Tian, S.-G. Sun, In situ FTIR spectroscopic studies of electrooxidation of ethanol on Pd electrode in alkaline media, Electrochim. Acta 55 (2010) 7995e7999, http://dx.doi.org/10.1016/ j.electacta.2010.02.071. [13] Z.X. Liang, T.S. Zhao, J.B. Xu, L.D. Zhu, Mechanism study of the ethanol oxidation reaction on palladium in alkaline media, Electrochim. Acta 54 (2009) 2203e2208, http://dx.doi.org/10.1016/j.electacta.2008.10.034. [14] A. Kowal, M. Li, M. Shao, K. Sasaki, M.B. Vukmirovic, J. Zhang, et al., Ternary Pt/ Rh/SnO2 electrocatalysts for oxidizing ethanol to CO2, Nat. Mater. 8 (2009) 325e330, http://dx.doi.org/10.1038/nmat2359. [15] M. Li, D. a Cullen, K. Sasaki, N.S. Marinkovic, K. More, R.R. Adzic, Ternary electrocatalysts for oxidizing ethanol to carbon dioxide: making ir capable of splitting CeC bond, J. Am. Chem. Soc. 135 (2013) 132e141, http://dx.doi.org/ 10.1021/ja306384x. [16] V. Rao, C. Cremers, U. Stimming, Investigation of the ethanol electro-oxidation in alkaline membrane electrode assembly by differential electrochemical mass spectrometry, Fuel Cells 7 (2007) 417e423, http://dx.doi.org/10.1002/ fuce.200700026. [17] V. Bambagioni, C. Bianchini, Y. Chen, J. Filippi, P. Fornasiero, M. Innocenti, et al., Energy efficiency enhancement of ethanol electrooxidation on Pd-CeO(2)/ C in passive and active polymer electrolyte-membrane fuel cells, ChemSusChem. 5 (2012) 1266e1273, http://dx.doi.org/10.1002/cssc.201100738. [18] Y. Chen, M. Bellini, M. Bevilacqua, P. Fornasiero, A. Lavacchi, H.A. Miller, et al., Direct alcohol fuel cells: toward the power densities of hydrogen-fed proton exchange membrane fuel cells, ChemSusChem. 8 (2015) 524e533, http:// dx.doi.org/10.1002/cssc.201402999. [19] C.-M. Chang, M.-T. Hsieh, W.-C. Kang, T.-J. Whang, Study of the Coelectrodeposited Pd-Ni alloy thin film and its performance on ethanol electro-oxidation, J. Electrochem. Soc. 161 (2014) D552eD557, http://dx.doi.org/ 10.1149/2.0911410jes. [20] K.S. Kumar, P. Haridoss, S.K. Seshadri, Synthesis and characterization of electrodeposited NiePd alloy electrodes for methanol oxidation, Surf. Coat. Technol. 202 (2008) 1764e1770, http://dx.doi.org/10.1016/ j.surfcoat.2007.07.035. [21] T. Maiyalagan, K. Scott, Performance of carbon nanofiber supported PdeNi catalysts for electro-oxidation of ethanol in alkaline medium, J. Power Sources 195 (2010) 5246e5251, http://dx.doi.org/10.1016/j.jpowsour.2010.03.022. [22] Z. Qi, H. Geng, X. Wang, C. Zhao, H. Ji, C. Zhang, et al., Novel nanocrystalline PdNi alloy catalyst for methanol and ethanol electro-oxidation in alkaline media, J. Power Sources 196 (2011) 5823e5828, http://dx.doi.org/10.1016/ j.jpowsour.2011.02.083. [23] X. Wang, G. Ma, F. Zhu, N. Lin, B. Tang, Z. Zhang, Preparation and characterization of micro-arc-induced Pd/TM(TM ¼ Ni, Co and Ti) catalysts and comparison of their electrocatalytic activities toward ethanol oxidation, Electrochim. Acta 114 (2013) 500e508, http://dx.doi.org/10.1016/ j.electacta.2013.10.049. [24] C. Xu, Z. Tian, P. Shen, S.P. Jiang, Oxide (CeO2, NiO, Co3O4 and Mn3O4)-promoted Pd/C electrocatalysts for alcohol electrooxidation in alkaline media, Electrochim. Acta 53 (2008) 2610e2618, http://dx.doi.org/10.1016/ j.electacta.2007.10.036. [25] D. Liu, Q. Guo, H. Hou, O. Niwa, T. You, Pd x Co y nanoparticle/carbon nanofiber composites with enhanced electrocatalytic properties, ACS Catal. 4 (2014) 1825e1829, http://dx.doi.org/10.1021/cs5000153. n, A. Heinzel, Estimation of the membrane methanol diffusion co[26] V. Barraga efficient from open circuit voltage measurements in a direct methanol fuel cell, J. Power Sources 104 (2002) 66e72, http://dx.doi.org/10.1016/S03787753(01)00896-5. _ ski, R. Kowalik, K. Fitzner, Analysis of CoePd alloys deposi[27] K. Mech, P. Zabi n tion from electrolytes based on [Co(NH3)6]3þ and [Pd(NH3)4]2þ complexes, Electrochim. Acta 104 (2013) 468e473, http://dx.doi.org/10.1016/
822
L.-k. Tsui et al. / Journal of Power Sources 293 (2015) 815e822
j.electacta.2012.12.006. [28] International Centre for Diffraction Data, ICDD DDViewþ, Version 4.8.0.5, Newtown Square, PA, USA, 2009. [29] K. Ishida, T. Nishizawa, The Co-Pd (Cobalt-Palladium) System, J. Phase Equilib. 12 (1991) 83e87, http://dx.doi.org/10.1007/BF02663680. lisle, P. Chartrand, S.A. Decterov, G. Eriksson, K. Hack, et al., [30] C.W. Bale, E. Be FactSage thermochemical software and databases - recent developments, Calphad Comput. Coupling Phase Diagrams Thermochem. 33 (2009) 295e311, http://dx.doi.org/10.1016/j.calphad.2008.09.009. [31] Co-Pd: Data from FSnobl e Fact Sage Noble Metal Databases, Cent. Res. Comput. Thermochem. http://www.crct.polymtl.ca/fact/phase_diagram.php? file¼Co-Pd.jpg&dir¼FSnobl (accessed 09.03.15) [32] Y. Matsuo, Ordered alloys in the cobaltepalladium system, J. Phys. Soc. Jpn. 32 (1972) 972e978. [33] S. Tominaka, T. Osaka, Nanoporous PdCo catalyst for microfuel cells: electrodeposition and dealloying, Adv. Phys. Chem. 2011 (2011) 1e13, http:// dx.doi.org/10.1155/2011/821916. [34] G. Zangari, B. Bozzini, P.L. Cavallotti, G. Fontana, P.G. Maisto, E. Terrenzio, Coercivity and microstructure in Co-Pt and Co-Pd thin films, J. Magn. Magn. Mater. 133 (1994) 511e515, http://dx.doi.org/10.1016/0304-8853(94)906106. ski, The adsorption of carbon oxides on a palladium electrode from [35] A. Czerwin acidic solution,, J. Electroanal. Chem. 379 (1994) 487e493, http://dx.doi.org/ 10.1016/0022-0728(94)87173-6. , M. Łukaszewski, G. Jerkiewicz, A. Czerwin ski, Electrochemical [36] M. Grden behaviour of palladium electrode: oxidation, electrodissolution and ionic adsorption, Electrochim. Acta 53 (2008) 7583e7598, http://dx.doi.org/ 10.1016/j.electacta.2008.05.046. [37] J. Liu, J. Ye, C. Xu, S.P. Jiang, Y. Tong, Kinetics of ethanol electrooxidation at Pd electrodeposited on Ti, Electrochem. Comm. 9 (2007) 2334e2339, http://
dx.doi.org/10.1016/j.elecom.2007.06.036. [38] Z.X. Liang, T.S. Zhao, J.B. Xu, L.D. Zhu, Mechanism study of the ethanol oxidation reaction on palladium in alkaline media, Electrochim. Acta 54 (2009) 2203e2208, http://dx.doi.org/10.1016/j.electacta.2008.10.034. [39] Z. Zhang, C. Zhang, J. Sun, T. Kou, Q. Bai, Y. Wang, et al., Ultrafine nanoporous PdFe/Fe3O4 catalysts with doubly enhanced activities towards electrooxidation of methanol and ethanol in alkaline media, J. Mater. Chem. A 1 (2013) 3620, http://dx.doi.org/10.1039/c3ta01464a. [40] Q. Yi, L. Sun, X. Liu, H. Nie, Palladiumenickel nanoparticles loaded on multiwalled carbon nanotubes modified with b-cyclodextrin for electrooxidation of alcohols, Fuel 111 (2013) 88e95, http://dx.doi.org/10.1016/ j.fuel.2013.04.051. [41] C. Qiu, R. Shang, Y. Xie, Y. Bu, C. Li, H. Ma, Electrocatalytic activity of bimetallic PdeNi thin films towards the oxidation of methanol and ethanol, Mater. Chem. Phys. 120 (2010) 323e330, http://dx.doi.org/10.1016/ j.matchemphys.2009.11.014. ger, Direct ethanol fuel cell [42] S. Rousseau, C. Coutanceau, C. Lamy, J.-M. Le (DEFC): electrical performances and reaction products distribution under operating conditions with different platinum-based anodes, J. Power Sources 158 (2006) 18e24, http://dx.doi.org/10.1016/j.jpowsour.2005.08.027. [43] J. Mann, N. Yao, A.B. Bocarsly, Characterization and analysis of new catalysts for a direct ethanol fuel cell, Langmuir 22 (2006) 10432e10436, http:// dx.doi.org/10.1021/la061200c. €rnsson, H. Olin, J. Prikulis, Studies of the initial [44] D. Erts, E. Ahlberg, J. Asbjo oxidation of cobalt in alkaline solutions using scanning electrochemical microscope, App. Phys. A 66 (1998) 477e480, http://dx.doi.org/10.1007/ s003390051186. [45] B.S. Yeo, A.T. Bell, Enhanced activity of gold-supported cobalt oxide for the electrochemical evolution of oxygen, J. Amer. Chem. Soc. 133 (2011) 5587e5593, http://dx.doi.org/10.1021/ja200559j.