- Email: [email protected]
Mechanically activated Pt–Ni and Pt–Co alloys as electrocatalysts in the oxygen reduction reaction
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e9
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/he
Mechanically activated PteNi and PteCo alloys as electrocatalysts in the oxygen reduction reaction Claudia A. Corte´s-Escobedo a,*, Rosa de G. Gonza´lez-Huerta b, Ana M. Boları´n-Miro´ c, Fe´lix Sa´nchez de Jesu´s c, Q. Zhu d, S.E. Canton d, Karina Suarez-Alcantara d,1, M. Tufin˜o-Velazquez e a
Centro de Investigacio´n e Innovacio´n Tecnolo´gica del Instituto Polite´cnico Nacional, Cda. Cecati s/n, Col. Sta. Catarina, CP 02250 Azcapotzalco, D.F., Mexico b ESIQIE-Instituto Polite´cnico Nacional, Laboratorio de Foto-Electrocata´lisis, UPALM, CP 07738 DF, Mexico c ´ Area Acade´mica de Ciencias de la Tierra y Materiales, Universidad Auto´noma del Estado de Hidalgo, CU, Carr. Pachuca-Tulancingo Km. 4.5, Mineral de la Reforma, CP 42184 Hidalgo, Mexico d Department of Synchrotron Radiation Instrumentation, Lund University, Ole Ro¨mers va¨g 1, SE-22363 Lund, Sweden e ESFM-Instituto Polite´cnico Nacional, Laboratorio de Fı´sica Avanzada, UPALM, CP 07738 DF, Mexico
article info
abstract
Article history:
Mixtures of powders of platinum with nickel or cobalt to obtain Ni0.75Pt0.25 or Co0.75Pt0.25
Received 27 November 2013
were mechanical alloyed by high energy ball milling. The results of crystal structure,
Received in revised form
morphology and electrocatalytic performance are presented for mechanically activated
24 February 2014
powders after 3 and 9 h of ball milling. Total solid solutions of Ni and Co with platinum
Accepted 5 March 2014
were analyzed by X-ray diffraction after 3 h of ball milling. After 9 h of ball milling, in both
Available online xxx
cases, the total solid solution was accompanied by the appearance of NiO or CoO and ZrO associated with a redox reaction with the milling media. The presence of zirconium
Keywords:
monoxide was confirmed by energy dispersive spectroscopy analysis. In both cases, an
NiPt
amorphization was detected. X ray absorption spectroscopy measurements showed
CoPt
changes in atomic and electronic environment of platinum, a reduction of the distance to
ORR
the first coordination sphere and increased d-band vacancy vs pure Pt and Pt nanoparticles
Electrocatalysts
were observed for both studied systems. The electrocatalytic activity was determined using
Mechanical alloying
cyclic and linear voltammetry. The Co0.75Pt0.25 alloy milled for 9 h showed a higher elec-
Anchor effect
trochemical activity for the oxygen reduction reaction (ORR) compared with the other samples, including Pt-Etek. The degree of the ORR electrochemical activity was correlated with the presence of ZrO, which could affect the oxygen adsorption and improve the catalytic activity for the oxygen reduction reaction. Copyright ª 2014, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
* Corresponding author. Tel.: þ52 55 57296000x68311, fax: þ52 5555617536. E-mail addresses: [email protected], [email protected] (C.A. Corte´s-Escobedo). 1
Present address: UNAM-IIM Morelia, Antigua carretera a Pa´tzcuaro 8710, Col. Ex-hacienda de San Jose´ de la Huerta, Morelia, Michoaca´n, 58190, Mexico. http://dx.doi.org/10.1016/j.ijhydene.2014.03.025 0360-3199/Copyright ª 2014, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Corte´s-Escobedo CA, et al., Mechanically activated PteNi and PteCo alloys as electrocatalysts in the oxygen reduction reaction, International Journal of Hydrogen Energy (2014), http://dx.doi.org/10.1016/ j.ijhydene.2014.03.025
2
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e9
Introduction Intense efforts have been made to develop new materials for electrocatalysis in low temperature polymer electrolyte fuel cells (PEMFCs) [1,2]. Both anode and cathode electrocatalysts require low platinum content in order to reduce costs. At the same time, it is intended to have high activity towards the oxygen reduction reaction (ORR), i.e. fewer kinetic limitations; to achieve low cathodic overpotential and to reduce the CO poisoning effect. Anode and cathode electrodes made of nanosized materials have shown the best activity towards the ORR. The electrocatalytic activity is strongly influenced by its particle size, the nature of the support and the preparation method [12]. To achieve an effective reduction in platinum content and poisoning effect, some transition metals have been used [3e16]. However, preparation of transition metal particles with a diameter size of only a few nanometers i.e. large surface areas is a challenge. Many researchers have employed different methods of synthesis to obtain novel nano-electrocatalyst of transition metals. Among these methods of synthesis, the mechanical milling (also known as ball milling) has been recently used [17e20]. With this technique, it is possible to obtain alloys, metastable phases, solid solutions, amorphous phases, nanocrystalline structures or composites. The mechanical milling can be used to prepare advanced materials that are difficult to produce by conventional fusion techniques. This work presents the mechanical synthesis of nano-sized Pt alloys (PteNi and PteCo) and evaluates their physical properties and electrocatalytic activity towards the ORR in an acid medium.
Materials and methods Elemental powders of Pt (SigmaeAldrich, >99.9%), Co (SigmaeAldrich, >99.9%) and Ni (SigmaeAldrich, >99.9%) were used as precursors. The raw materials Co and Ni were mixed in the appropriate weight ratio to obtain two different alloys: Co0.75Pt0.25 and Ni0.75Pt0.25. A total amount of 1 g of powder mixture together with 6 zirconia (ZrO2) balls, 11 mm in diameter, were loaded into a zirconia vial. The ball-to-powder weight ratio was 25:1. The mechanical alloying process was carried out at room temperature in air atmosphere using a shaker mixer/mill machine. To prevent excessive overheating of the vials, all experiments were carried out using cycles of 60 min of milling and 15 min of resting. The milling time spanned from 0 to 9 h to evaluate the effect of this parameter on the particle size distribution and the synthesis of the metallic alloy. X-Ray diffraction (XRD) patterns of the milled powders were used to study the phase transformations as a function of the composition and milling time, using a Bruker D2 Phaser. Diffraction parameters were the diffracting angle, A) was 2q, ranging from 20 to 90 . Cu Ka radiation (l ¼ 1.5418 used in all experiments. The morphology and composition of the milled powders were analyzed using a FEI Quanta 3D FEG scanning electron microscope under high vacuum. X-ray absorption spectroscopy (XAS) spectra were taken at the I811 beamline of MAX-Lab synchrotron facility, Sweden [21,22]. XAS at Pt L2,3-edge, Ni K-edge and Co K-edge of the 9-h
ball milled samples were collected. 1 mg of the sample was distributed over a 1.5 cm2 area of the glue-side of Kapton tape. The samples over Kapton were placed inside a He-filled chamber. The sample-side was oriented at 45 with respect to the incident beam, then XAS spectra were taken in fluorescence mode. Pt L3-edge spectra were taken between 100 and þ900 eV vs E0 ¼ 11564 eV energy, with a step ¼ 0.4 eV. Pt L2-edge spectra were taken between 100 and þ450 eV vs E0 ¼ 13273 eV energy, with a step ¼ 0.4 eV. A Zn filter was used before the Lytle detector during Pt L2,3-edge spectra collection, 5 scans were averaged. Ni K-edge spectra were taken between 50 and þ600 eV vs E0 ¼ 8333 eV, with a step ¼ 0.4 eV. A Co filter was used before the Lytle detector during Ni K-edge spectra collection, 5 scans were averaged. Co K-edge spectra were taken between 100 and þ800 eV vs E0 ¼ 7709 eV, with a step ¼ 0.4 eV. Fe filter was used before the Lytle detector during the Co K-edge spectra collection, 5 scans were averaged. Data were processed with the Athena program [23]: proper calibration with Pt, Ni and Co foil respectively was performed, background removal and normalization was performed before whiteline area integration. Electrochemical evaluation was carried out using a conventional three-electrode electrochemical cell. The electrodes were: 1) glassy carbon covered with a thin electrocatalyst film as working electrode (0.196 cm2 geometric area), 2) Hg/Hg2SO4, 0.5 M H2SO4 (E ¼ 0.680 V/NHE) as reference electrode and 3) platinum mesh as counter electrode (with more than 10 cm2 of surface area). A 0.5 M H2SO4 (Merck, p.a.) solution was used as electrolyte. The working electrodes were prepared according to the methodology reported elsewhere [17]. Briefly, the as milled electrocatalysts were supported in carbon Vulcan to a load of 10 wt% Pt/C: 5 mg of electrocatalyst, 24 mg of Vulcan and 200 mL of ethanol were sonicated for 20 min. Then the powders were dried at 40 C and placed in a closed glass container until their utilization. A catalytic “ink” of each electrocatalyst was prepared by mixing 60 mL of ethanol, 6 mL of Nafion (5 wt%, Du Pont 1000 EW) and 1 mg of electrocatalyst supported on carbon Vulcan (10 wt% Pt/C). 8 mL of the Ni0.75Pt0.25 or Co0.75Pt0.25 catalytic “ink” was uniformly dispersed over the surface of a glassy carbon electrode. After drying a thin film of electrocatalyst was formed. The rotating disk electrode (RDE) electrochemical characterization technique was performed by means of a 263 A potentiostat/galvanostat (EG&G PAR). The rotating rates were fixed at 100, 200, 400, 900, 1600 or 2500 rpm; the scan velocity was set at 5 mV s1. Previous (and necessarily) electrochemical activation was carried out by 20 successive potential sweep cycles at 50 mV s1 in a potential window of 0e1.2 V/ NHE to eliminate impurities. Potentials are reported with respect to the normal hydrogen electrode unless otherwise specified. Pt mass-specific activity (Im) was calculated at E ¼ 0.90 V/ NHE for an ORR polarization curve measured at a scan rate of 5 mV s1. The mass activities are estimated via calculation of jk (kinetic current) and normalization to LPt (the working electrode Pt loading). Catalyst electrocatalytic activity towards the ORR is quantified at E ¼ 0.90 V because interferences from mass-transport losses cannot be completely excluded at higher current densities observed below E ¼ 0.90 V [24].
Please cite this article in press as: Corte´s-Escobedo CA, et al., Mechanically activated PteNi and PteCo alloys as electrocatalysts in the oxygen reduction reaction, International Journal of Hydrogen Energy (2014), http://dx.doi.org/10.1016/ j.ijhydene.2014.03.025
3
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e9
Results and discussion In Fig. 1, the X-ray diffraction patterns of Ni0.75Pt0.25 electrocatalyst after 3 and 9 h of ball milling are presented. Ni powder without milling also is presented as visual comparison for crystal deformations. The Ni powder reflection peaks are consistent with the cubic close-packed structure of pure nickel (ICSD-41508). Fig. 1 reveals that 3 h of ball milling of Ni and Pt is enough to produce a total solid solution, where the Pt atoms are introduced into the Ni crystal lattice. A slight displacement of the nickel reflection peaks toward the left was observed from 44.4 to 43.74 in 2q degree indicating an expansion of the Ni crystal lattice caused by the big size of Pt lattice parameter a of nickel from 3.52 to 3.61 A. The Ni0.75Pt0.25 diffraction reflection peaks become wider as ball milling time increases due to the particle and grain size diminution. After 9 h of ball milling, the unexpected formation of NiO (ICSD61324) and ZrO (ICSD-72956) were evident. The formation of these oxides can be attributed to a redox reaction expressed in the following partial reactions due to Ni reactivity, Pt catalyst and the high energy process:
1
Ni þ /2 O2 / NiO þ 2e
(1)
ZrO2 þ 2e / ZrO þ 1/2O2
(2)
Ni (311)
Ni (220)
Ni (111)
Ni (200)
Results of Rietveld refinement of the XRD pattern indicated a reduction in crystal size from 573 to 166 A, an increment in cell parameter a from 3.52 to 3.61 A and an increment in rms microstrain from 0.0001 to 0.0031, however, because of the crystallite size, diffraction peaks are near to the level of noise, and the c2 is too high. Fig. 2 shows the XRD patterns of Co0.75Pt0.25 milled for 3 h and Co0.75Pt0.25 milled for 9 h. XRD of Co powder without milling is presented as visual reference. The Co powder is composed of 2 phases; the cubic Co-a (Co Fm-3m, ICSD 41507) and the hexagonal Co-b (ICSD-050727, P63/mmc, hcp). After 3 h
Intensity (a.u.)
Ni (222)
Pt
Ni
Ni
Pt
- 3h
Ni
Pt
- 9h ZrO NiO
20
30
40
ZrO NiO
50
60
NiO
70
80
NiO
90
100
2θ (degree) Fig. 1 e X-Ray diffraction of Ni and Ni0.75Pt0.25 mixtures milled for 3 and 9 h.
Fig. 2 e X-Ray diffraction patterns for Co and Co0.75Pt mixtures after high energy ball milling.
0.25
of milling, the Co0.75Pt0.25 powder presents only a wide peak at 43.9 in 2q, this can be associated to a deformed hexagonal structure, i.e. the amorphization of the Co-b structure and loss of Co-a structure. Pt diffraction peaks are not evident, confirming the incorporation of Pt into the deformed Co structure. The increase in the milling time promotes further deformation of the Co lattice, due to the mechanical energy induced by the process. After 9 h of ball milling cubic Co-a are observed. Again, two reflection peaks at about 2q ¼ 51.7 and 44 corresponding respectively to (200) and (111) are evident. The reversible transformation between Co-b (hcp) and Co-a (fcc) was already known, this depend upon the milling parameters and the formation of solid solutions [16]. Interesting, the formation of CoO (ICSD-9865) and ZrO (ICSD-72956) was observed. As in the Ni0.75Pt0.25 and nearly at the same time of ball milling, the oxide formation was present also in Co0.75Pt0.25. This cannot be associated with ball or vial fractures. We attribute the formation of ZrO and CoO to the following redox pair reactions:
Co þ 1/2 O2 / CoO þ 2e
(3)
ZrO2 þ 2e / ZrO þ 1/2 O2
(4)
Results of Rietveld refinement showed a reduction in crystallite size from 200 to 78 A after 3 h of milling, but then, after ZrO appearance, in 30 w% crystallite size increases up to 287 A and a cell parameter the beta structure of cobalt increases from 2.511 to 3.372 A after 3 h of milling but after 9 h, decrease to 2.841 A. While c parameter of the same structure diminishes from 4.083 to 3.258 A after 3 h and change to 3.885 A after 9 h of milling. Rms microstrain in the first step of milling increases from 0.0004 to 0.0739 and then decrease up to 0.0312 which can indicate a release of stresses promoted by the PteZrO presence. Again, diffraction peaks are near to the level of noise, and the c2 is too high.
Please cite this article in press as: Corte´s-Escobedo CA, et al., Mechanically activated PteNi and PteCo alloys as electrocatalysts in the oxygen reduction reaction, International Journal of Hydrogen Energy (2014), http://dx.doi.org/10.1016/ j.ijhydene.2014.03.025
4
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e9
Fig. 3 e SEM images of the Ni0.75Pt0.25 and Co0.75Pt
Fig. 3 presents the micrographs of Ni0.75Pt0.25 and Co0.75Pt0.25 electrocatalysts milled for 3 and 9 h. For both alloys, after 3 h of milling the presence of large aggregates of large particles are manifest. After 9 h of milling, the dispersion of small particles with grain sizes of approximately 60 nm were found, thus the available surface area is notably larger for samples milled for 9 h of ball milling. The Co0.75Pt0.25 surface also landscapes porosity. Table 1 collects the results of the elemental composition obtained by EDS for obtained alloys. Important is to note variations in atomic compositions due to increase in zirconium monoxide and cobaltum oxide formation, discussed before. Fig. 4(a) presents the Pt L3-edge XAS spectra for Ni0.75Pt0.25 and Co0.75Pt0.25. The plot also presents the XAS spectra of Pt foil and Pt commercial nanoparticles (ETEK) as reference for nanoparticle and alloy formation. The whole XAS spectra of Ni0.75Pt0.25 and Co0.75Pt0.25 show similar oscillations features; meanwhile different compared with the Pt foil, the Pt commercial nanoparticles. This reflect the changes in the Pt
Table 1 e Results of elemental composition by EDS for systems Ni0.75Pt0.25 and Co0.75Pt0.25 with milling. Sample
Ni Ni25Pt 3 h Ni25Pt 9 h Co Co25Pt 3 h Co25Pt 9 h
EDS composition O at%
Co at%
Ni at%
Pt at%
Zr at%
0 9.8 23.2 0 21.7 36.2
0 0 0 100 64.7 53.2
100 78.5 59.3 0 0 0
0 7.3 6.4 0 8.3 4.6
0 4.4 11.1 0 5.3 6.0
0.25
systems ball milled for 3 and 9 h.
surroundings according the synthetic procedure and the alloy composition. These changes in the geometrical environments are more evident at the EXAFS Fourier transformation (inset at ˚ 1, k3-weighted, the right in Fig. 4(a), k-range from 1.5 to 15.7 A phase corrected). The Pt foil and the Pt ETEK show the PtePt bond distance at 2.8 A while the Ni0.75Pt0.25 and Co0.75Pt0.25 show Pt-M (M ¼ Ni and Co) distances at 2.5 A in agreement with the alloy formation; and also in agreement with results obtained by X-ray diffraction. The inset at the left presents a zoom view of the Pt L3-edge whiteline and some oscillations above the edge. Evidently, changes in the electronic environment are induced after alloying. The Pt L3-edge whiteline results from the electronic transition 2p3/2 to 5d5/2 and its intensity is related to the Pt 5d vacancies [25]. However a complete analysis of the 5d-band vacancies is only completed by a simultaneous analysis of Pt L2-edge. Fig. 4(b) presents the near edge features of Pt L2-edge of the studied materials. Again, the oscillations above the Pt L2-edge of Ni0.75Pt0.25 and Co0.75Pt0.25 show similarities, meanwhile different from Pt foil and Pt ETEK nanoparticles. The whiteline at the Pt L2-edge is due to 2p1/2 to 5d5/2 transition, and is not so intense as the whiteline of Pt L3-edge: a phenomenon already described as non-statistical distribution of the densities of states of Pt 5d5/2 and 5d3/2 [26]. Murkerjee et al. [27] quantified the fractional change in d-band vacancies (fd) relative to a reference material (i.e. the Pt foil) as: fd ¼
DA3 þ 1:11DA2 ; ðA3 þ 1:11A2 Þr
(5)
where DA3 ¼ A3s A3r; DA2 ¼ A2s A2r; A3 and A2 are the areas under L3 and L2 edges respectively (10 to þ13 relative to the
Please cite this article in press as: Corte´s-Escobedo CA, et al., Mechanically activated PteNi and PteCo alloys as electrocatalysts in the oxygen reduction reaction, International Journal of Hydrogen Energy (2014), http://dx.doi.org/10.1016/ j.ijhydene.2014.03.025
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e9
Fig. 4 e (a). Pt L3-edge XAS spectra for Ni0.75Pt0.25 and Co0.75Pt0.25. (b). Near edge features of Pt L2-edge of the Ni0.75Pt0.25 and Co0.75Pt 0.25 systems.
edge position); s and r refer to the sample and reference respectively. The d-band vacancies of Pt samples are calculated as: hj
total sample
¼ 1 þ fd hj total
reference
(6)
Where the total number of unoccupied d-states, (hj)total reference, for pure Pt have been evaluated as 0.3 [27,28]. Thus the Pt d-orbital vacancy per atom, (hj)total sample, is 0.327 for Pt ETEK nanoparticles; 0.336 for Co0.75Pt0.25 and 0.334 for Ni0.75Pt0.25. Nanoparticle formation increases the Pt d-band vacancy. The alloy PteCo has an increased d-band vacancy compared to PteNi alloy. The d-band vacancy can play an important role for the activity towards the oxygen reduction reaction. Complementary information was obtained by XAS at Ni and Co K-edge. The K-edge XAS spectra of 3d transition metals describe the 1s to 3d transition if symmetry broken (pre-edge, no present here) and the 1s to 4p transition. Fig. 5(a) presents the XAS spectra at K-edge of Ni0.75Pt0.25 and Ni foil. The K-edge spectrum of Ni0.75Pt0.25 is strongly dominated by the NiO contribution [29,30]. The Ni at of Ni0.75Pt0.25 can be described as a mixture of metallic Ni (minor component)), NiO (mayor component) and NiePt alloy. Particles of NiO have been used to reduce the CO poisoning of Pt particles at fuel cells [31]. The inset of Fig. 5(a) presents the EXAFS Fourier transformation (k-
5
Fig. 5 e (a). XAS spectra at K-edge - EXAFS Fourier ˚ L1, k3-weighted, transformation (k-range from 2 to 10.5 A phase corrected) e of Ni0.75Pt0.25 and Ni foil. (b). XAS spectra at K-edge - EXAFS Fourier transformation (k-range from 2 ˚ L1, k3-weighted, phase corrected) e of Co0.75Pt0.25. to 10.5 A
˚ 1, k3-weighted, phase corrected). The Ni range from 2 to 10.5 A foil present the NieNi bond distance at 2.44 A (Ni ICSD-41508); A that is the NieO bond NiO the Ni0.75Pt0.25 has peaks at 2.08 (ICSD-61324), and at 2.66 A that is the NiePt bond (ICSD646297, PtNi alloy). Fig. 5(b) presents the Co K-edge spectrum of Co0.75Pt0.25, it is dominated by the features of CoO i.e. the shift of the edge position and the intense whiteline peak [30]. The inset of Fig. 5(b) displays the EXAFS Fourier transformation (k-range ˚ 1, k3-weighted, phase corrected). The peak at from 2 to 9.5 A 2.44 A agrees with the CoeCo bond distance of cubic Co. The peak at 2.08 A correlates with the CoeO bond distance (ICSD9865) and the 2.61 A is the CoePt bond distance (ICSD-102620). Thus, the ball milled Co0.75Pt0.25 can be described as a mixture of metallic Co (minor component), CoO (mayor component) and CoPt alloy. Cyclic voltammetry has become a technique for initial electrochemical studies of new systems and has proved being very useful in obtaining information about stability in reaction media. Fig. 6 shows cyclic voltammetry (CV) curves of the Ni0.75Pt0.25/C 3 and 9 h ball milled and Co0.75Pt0.25/C 3 and 9 h
Please cite this article in press as: Corte´s-Escobedo CA, et al., Mechanically activated PteNi and PteCo alloys as electrocatalysts in the oxygen reduction reaction, International Journal of Hydrogen Energy (2014), http://dx.doi.org/10.1016/ j.ijhydene.2014.03.025
6
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e9
Fig. 6 e Cyclic voltammograms synthesized by mechanical milling of the a) Ni0.75Pt0.25/C and b) Co0.75Pt0.25/C electrocatalysts. Data recorded in outgassed 0.5 M H2SO4 at 25 C. Scan rate of 50 mV sL1.
ball milled samples, in the de-aerated 0.5 M H2SO4 at 50 mV s1 in the range from 0.0 to 1.20 V at 25 C. The Ni0.75Pt0.25/C 3 h catalyst in Fig. 6(a) shows a striking low current density magnitude for all potential scans. This low electrochemically active area at both the hydrogen oxidation and the oxygen reduction regions and the low double layer capacitance, are presumably due to large aggregates and large particles. No evident influence of the large porosity of electrocatalyst particles was observed. The CV curve of Ni0.75Pt0.25/C 9 h (Fig. 6(a)) shows a significant improvement in the electrochemical activity. Typical characteristics of Pt nanoparticles were observed: the hydrogen adsorption and desorption peaks at the hydrogen region between 0.0 and 0.36 V, Pt oxidation starting at 0.80 V, and a reduction of Pt oxide film centered at 0.75 V. Fig. 6(b) shows the cyclic voltammetry (CV) curves of the Co0.75Pt0.25/C 9 h and 3 h catalysts. The Co0.75Pt0.25/C 3 h catalyst shows the same behavior as Ni0.75Pt0.25/C 3 h, almost none definition of Pt electrochemical features for oxygen reduction and hydrogen absorption/oxidation regions. The catalytic activity is low due to the large aggregates and large particles observed in Fig. 3. Co0.75Pt0.25/C 9 h presents a better
electrochemical response. Even more, the typical characteristics of Pt are better defined in the case of the Co0.75Pt0.25/C 9 h catalyst than in the case of Ni0.75Pt0.25/C 9 h interestingly, the Co0.75Pt0.25/C 9 h shows additional peaks at the hydrogen absorption/desorption region. These peaks, between 0.0 and 0.36 V, must be related to the presence of the second metal, Co and ZrO. Another effect of Co is the shift of the oxygen reduction potential from approximately 0.05 V to a more positive value of 0.71 V, which is showed by Ni0.75Pt0.25/C 9 h. The Co0.75Pt0.25/C catalysts shows high stability in the reaction medium, the CVs responses were equal after tenth cycle, which is showed in Fig. 6(b). The stability can be attributed to the presence of ZrO, as has been reported previously [32,33], causing an anchor effect of the platinum atoms, which inhibits agglomeration. The last effect is found only for the Co-based system. Due to the low solubility of oxygen in acid media, the oxygen reduction reaction (ORR) depends strongly on the hydrodynamic conditions. The ORR characteristics of polarization curves from RDE measurements, at different rotation rates, of Ni0.75Pt0.25/C 9 h and Co0.75Pt0.25/C 9 h are summarized in Fig. 7(a and b). Comparing the two electrocatalyst, two characteristics
Fig. 7 e Currentepotential curves for oxygen reduction for a) Ni0.75Pt0.25/C 9 h and b) Co0.75Pt0.25/C 9 h, and in 0.5 M H2SO4 saturated with oxygen at different rotation rates. Currents recorded at 5 mV sL1. Please cite this article in press as: Corte´s-Escobedo CA, et al., Mechanically activated PteNi and PteCo alloys as electrocatalysts in the oxygen reduction reaction, International Journal of Hydrogen Energy (2014), http://dx.doi.org/10.1016/ j.ijhydene.2014.03.025
7
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e9
Table 2 e Kinetic parameter obtained from the oxygen reduction reaction (ORR). Electrocatalyst
Eca V/ENH
-b V dec1
a
jo (mA cm2)
j/mA cm2 at V ¼ 0.85 V
LPt mgPt
Im mA 2 mg1 Pt cm
Ni0.75Pt0.25/C-9 h Co0.75Pt0.25/C-9 h Pt-Etek
0.95 0.06 1.05 0.05 0.98 0.03
0.100 0.008 0.083 0.005 0.080 0.002
0.59 0.03 0.70 0.02 0.71 0.01
1.49 1040.00004 1.38 1040.00003 1.25 1040.00002
0.60 0.08 3.33 0.05 1.36 0.03
7.7 5.5 12
0.031 0.129 0.080
are revealed: 1) The Co0.75Pt0.25/C 9 h electrocatalyst shows the higher open circuit potential, i.e. an overpotential reduction of about 50 mV compared to Ni0.75Pt0.25/C 9 h (see Table 2). Also there is an improvement in the total current density observed at the same hydrodynamic conditions. During the electrochemical reactions, either the charge or the mass transfer define the rate determining step. Both processes are distinguished at the polarization curves. The RDE curves of both materials present well defined regions: 1) at low overpotential (0.98e0.9 V/NHE), the charge transfer dominates and the current density is independent of the electrode rotation speed. 2) The mixed kinetic-diffusion control region (0.9e0.7 V/NHE), and 3) the diffusion limiting current region (0.7e0.2 V/NHE), where the increase of the rotation rate results in an increase of current density. Thus the kinetic parameters exchange current density and Tafel slope were estimated. RDE experiments were performed from high to low rotating rates, Co0.75Pt0.25/C catalysts presents a well-defined kinetic region, Fig. 7(b), while Ni0.75Pt0.25/C catalysts has low stability in the reaction medium, Fig. 7(a) shows a slight shift to lower potential, between each experiment, in the same region. Fig. 8 shows the mass-transfer-corrected Tafel plots of Ni0.75Pt0.25/C (3 h and 9 h) and Co0.75Pt0.25/C (3 h and 9 h) prepared by mechanical milling. Fig. 8 also shows the kinetic current density of Pt-Etek only for comparison purposes. The kinetic current densities were normalized to the geometrical area of the electrode, the mathematical procedure for kinetic parameters estimation is described elsewhere [34]. The values of open circuit potential, Eoc, current density exchange jo and Tafel slope are collected in Table 2 The open circuit potential,
Eoc, is monitored each experiment, in ORR it indicates material stability and is directly related to catalytic activity, in platinum compound its value should be from 0.9 to 1 V, Co0.75Pt0.25/C 9 h shows higher Eoc than Pt-Etek, Table 2. The comparison of current densities between the Ni0.75Pt0.25/C and Co0.75Pt0.25/C electrocatalysts at 0.85 V shows that the highest current density is obtained for Co0.75Pt0.25/C 9 h. The Tafel slope, b, of about 0.080 V dec1 was deduced for both the Co0.75Pt0.25/C-9 h and the P-Etek electrocatalysts, as expected for a first electron transfer as the rate determining step [19]. Values of the exchange current densities are consistent with those reported for nanosized catalysts containing Pt [12]; indicating that the ball milling as synthetic procedure is appropriate for electrocatalyst preparation. It is well known that the correlation between the d-band vacancies, the Pt-Metal bond distances, absorption of oxygenated species [27] and particle sizes have a combined effect on the electrocatalytic activity towards the ORR. Electrochemical results show that the incorporation of Co into the Pt structure improves its catalytic activity towards the ORR. Xray diffraction and X-ray absorption confirm a reduced PtMetal bond distance (vs PtePt). This reduced Pt-Metal bond distance have been identified as an active site suitable for absorption and further cleavage of oxygen. Both electrocatalyst show similar bond distances, thus another parameter is the main responsible for the enhanced electrochemical response. The Pt L3-edge XAS reveals that the Co0.75Pt0.25 has increased Pt d-band vacancy and also have improved activity towards the ORR compared with Co0.75Pt0.25, Pt nanoparticles and Pt foil. Additionally SEM of 9 h ball milled Co0.75Pt0.25 demonstrate the nanosized nature of this electrocatalyst, feature not shared with the Ni0.75Pt0.25. Finally the oxide formation ZrO/CoO or ZrO/NiO can act as co-catalyst towards the ORR [33]. The numerical values of the Pt mass-specific activity (Im) at E 0.90 V versus NHE derived from the ORR polarization curves measured at 5 mV s1, shown in Fig. 7, are detailed in Table 2. 2 A Pt-Etek has a Im of 0.080 mA mg1 Pt cm . The Im of the Co0.75Pt0.25/C-9 h is w60% higher than that of the Pt-Etek electrode and. Ni0.75Pt0.25/C-9 h has the lowest Im, 0.031 mA 2 mg1 Pt cm . The Im values reported in Table 2 demonstrate the importance of having a good electrode when evaluating the activity of a catalyst.
Conclusions
Fig. 8 e Mass transfer-corrected Tafel plots deduced from RDE analysis of oxygen-saturated 0.5 M H2SO4 at 25 C.
Ni0.75Pt0.25 and Co0.75Pt0.25 alloys, obtained by mechanical milling showed activity towards ORR. These samples are composed of a solid solution of platinum atoms inserted into Ni or Co structures after 3 h of milling. After 9 h of milling, the unexpected presence of ZrO/CoO or ZrO/NiO was observed,
Please cite this article in press as: Corte´s-Escobedo CA, et al., Mechanically activated PteNi and PteCo alloys as electrocatalysts in the oxygen reduction reaction, International Journal of Hydrogen Energy (2014), http://dx.doi.org/10.1016/ j.ijhydene.2014.03.025
8
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e9
consequence of a redox reaction between the milling media and the Ni or Co. SEM micrographs show that both materials are porous; particularly the 9 h ball milled Co0.75Pt0.25 presents large available surface area for ORR. The Pt L3-edge XAS reveals that the Co0.75Pt0.25 increased Pt d-band vacancy compared with Pt nanoparticles and Pt foil. Ni and Co K-edges indicate dominance of NiO or CoO respectively. The combination ZrO/CoO seems to play an important role in electrocatalytic activity due to the anchor effect. The electrocatalytic activity of Co0.75Pt0.25/C-9 h in the ORR is higher than Ni0.75Pt0.25/C-9 h. Co0.75Pt0.25/C-9 h also showed better stability in acid medium than Ni0.75Pt0.25/C-9 h. The Im of the Co0.75Pt0.25/C-9 h system is w60% higher than that of the PtEtek electrode.
Acknowledgments Authors thank the Center of Nanosciences and Micro and Nanotechnology of the National Polytechnic Institute, in particular Dr. Jose´ Alberto Andraca Adame and Dr. Hugo Martı´nez Gutie´rrez for structural and morphological characterization. Thanks for financial support: SECITI agreement ICYTDF/127/2012, CONACYT under projects CB-157925, CB130413 and SIP-IPN under project MULTI-1540. The authors are grateful to MAX-Lab synchrotron. Authors acknowledge the advice of Dr. Stefan Carlson and Dr. Katarina Nore´n during XAS data collection. KSA, QS and SEC were supported by the Swedish Research Council and the Crafoord Foundation.
references
[1] Barbir F. PEM Fuel cells, theory and practice. California: Academic Press, Sustainable world series; 2005. [2] Norskov JK, Rossmeisl J, Logadottir A, Lindqvist L, Kitchin JR, Bligaard T, et al. Origin of the overpotential for oxygen reduction at a fuel-cell cathode. J Phys Chem B 2004;108:17886e92. [3] Zignani SC, Antolini E, Gonzalez ER. Stability of PteNi/C (1:1) and Pt/C electrocatalysts as cathode materials for polymer electrolyte fuel cells: effect of ageing tests. J Power Sources 2009;191:344e50. [4] Zhao J, Jarvis K, Ferreira P, Manthiram A. Performance and stability of PdePteNi nanoalloy electrocatalysts in proton exchange membrane fuel cells. J Power Sources 2011;196:4515e23. [5] Zhao J, Manthiram A. Preleached PdePteNi and binary PdePt electrocatalysts for oxygen reduction reaction in proton exchange membrane fuel cells. App Cat B: Enviro 2011;101:660e8. [6] Su F, Poh CK, Zeng J, Zhong Z, Liu Z, Lin J. Pt nanoparticles supported on mesoporous carbon nanocomposites incorporated with Ni or Co nanoparticles for fuel cells. J Power Sources 2012;205:136e44. [7] Koh S, Toney MF, Strasser P. Activityestability relationships of ordered and disordered alloy phases of Pt3Co electrocatalysts for the oxygen reduction reaction (ORR). Electrochem Acta 2007;52:2765e74. [8] Di Noto V, Negro E. Development of nano-electrocatalysts based on carbon nitride supports for the ORR processes in PEM fuel cells. Electrochem Acta 2010;55:7564e74.
[9] Neyerlin KC, Srivastava R, Yu C, Strasser P. Electrochemical activity and stability of dealloyed PteCu and PteCueCo electrocatalysts for the oxygen reduction reaction (ORR). J Power Sources 2009;186:261e7. [10] Lee MH, Do JS. Kinetics of oxygen reduction reaction on Corich coreePtrich shell/C electrocatalysts. J Power Sources 2009;188:353e8. [11] Jayasayee K, Van Veen JAR, Manivasagam TG, Celebi S, Hensen EJM, Bruijn FA. Oxygen reduction reaction (ORR) activity and durability of carbon supported PtM (Co, Ni, Cu) alloys: Influence of particle size and non-noble metals. Appl Cat B: Enviro 2012;111:515e26. [12] Godı´nez-Salomo´n F, Hallen-Lo´pez M, Solorza-Feria O. Enhanced electroactivity for the oxygen reduction on Ni@Pt core-shell nanocatalysts. Int J Hyd Energy 2012;37:14902e10. [13] Garcı´a-Contreras MA, Ferna´ndez-Valverde SM, Vargas´ ngelesGarcı´a JR, Corte´s-Ja´come MA, Toledo-Antonio JA, A Chavez C. PtCo and PtNi electrocatalysts prepared by mechanical alloying for the oxygen reduction reaction in 0.5 M H2SO4. Int J Hyd Energy 2008;33:6672. [14] Antolini E, Salgado JRC, Gonza´lez ER. The methanol oxidation reaction on platinum alloys with the first row transition metals The case of PteCo and eNi alloy electrocatalysts for DMFCs: a short review. Appl Cat B: Enviro 2006;63:137e49. [15] Antolini E, Salgado JRC, Giz MJ, Gonza´lez ER. Effects of geometric and electronic factors on ORR activity of carbon supported PteCo electrocatalysts in PEM fuel cells. Int J Hyd Energy 2005;30:1213e20. [16] Betancourt-Cantera JA, Sa´nchez-De Jesu´s F, TorresVillasen˜or G, Boları´n-Miro´ AM, Corte´s-Escobedo CA. Extended solid solubility of a CoeCr system by mechanical alloying. J All Comp 2012;529:58e62. [17] Ezeta A, Arce EM, Solorza O, Gonzalez Huerta RG, Dorantes H. Effect of the leaching of RueSeeFe and RueMoeFe obtained by mechanical alloying on electrocatalytical behavior for the oxygen reduction reaction. J All Comp 2009;483:429e31. [18] Collins-Martı´nez V, Gonza´lez-Huerta RG, Lo´pez-Ortiz A, Delgado-Vigil D, Solorza-Feria O. Oxygen reduction kinetics of ball-milling synthesis Pd electrocatalysts in acid medium. J New Mat Electrochem Systems 2009;12:63e7. [19] Gonzalez-Huerta RG, Rodriguez Pierna A, Solorza-Feria O. Kinetic study of oxygen reduction reaction on an amorphous Ni59Nb40Pt0.6Ru0.4 catalyst in acid media. J New Mat Electrochem Systems 2008;11:63e7. [20] Sotelo-Mazo´n P, Gonza´lez-Huerta RG, Caban˜as-Moreno JG, Solorza-Feria O. Mechanically milled RuxFey electrocatalyst for oxygen reduction in acid media. Int J Electrochem Sci 2007;2:523e33. [21] Grehk TM, Nilsson PO. The design of the material science beamline, I811, at MAX-II. Nucl Instr and Meth in Phys Res A 2001;467-468:635e8. [22] Carlson S, Clausen M, Gridneva L, Sommarin B, Svensson C. XAFS experiments at beamline I811, MAX-lab synchrotron source. Sweden. J Synchrotron Rad 2006;13:359e69. [23] Ravel B, Newville M. ATHENA, ARTEMIS, HEPHAESTUS: data analysis for X-ray absorption spectroscopy using IFEFFIT. J Synchrotron Radiat 2005;12. 537e341. [24] Garsany Y. Experimental methods for quantifying the activity of platinum electrocatalysts for the oxygen reduction reaction. Anal Chem 2010;82:6321e8. [25] Merte LR, Behafarid F, Miller DJ, Friebel D, Cho S, Mbuga F, et al. Electrochemical oxidation of size-selected Pt nanoparticles studied using in situ high-energy-resolution Xray absorption spectroscopy. ACS Catal 2012;2:2371e6. [26] Sham TK, Ward MJ, Muephy MW, Lui LJ, Han WQ. Pt L3,2edge whiteline anomaly and its implications for the chemical behavious of Pt 5d5/2 and 5d3/2 electronic states e
Please cite this article in press as: Corte´s-Escobedo CA, et al., Mechanically activated PteNi and PteCo alloys as electrocatalysts in the oxygen reduction reaction, International Journal of Hydrogen Energy (2014), http://dx.doi.org/10.1016/ j.ijhydene.2014.03.025
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e9
[27]
[28] [29]
[30]
a study of PteAu nanowires and nanoparticles. J Phys Conf Ser 2013;430:012018e24. Mukerjee S, Srinivasan S, Soriaga M, McBreen J. Role of structural and electronic properties of Pt and Pt alloys on electrocatalysis of oxygen reduction. J Electrochem Soc 1995;142:1409e22. Brown M, Peierls RE, Stern EA. White lines in X-ray absorption. Phys Rev B 1997;15:738e44. Calandra M, Rueff JP, Gougoussis C, Ceolin D, Gorgoi M, Benedetti S, et al. K-edge X-ray absorption spectra in transition-metal oxides beyond the single-particle approximation: Shake-up many-body effects. Phys Rev B 2012;86:165102. Cantane DA, Oliveira FER, Santos SF, Lima FHB. Syntesis of Pt-based hollow nanoparticles using carbon-supported Co@Pt and Ni@Pt core-shell structures as templates: Electrocatalytic activity for the oxygen reduction reaction. App Catal B: Enviro 2013;136-137:351e60.
9
[31] Park B, Cairns EJ. Electrochemical performance of TiO2 and NiO as fuel cell electrode additives. Electrochem Comm 2011;13:75e7. [32] Liu G, Zhang H, Zhong H, Hu J, Xu D, Shao Z. A novel sintering resistant and corrosion resistant Pt4ZrO2/C catalyst for high temperature PEMFCs. Electrochim Acta 2006;51:5710e4. [33] He Q, Mukerjee S, Zeis R, Parres-Esclapez S, Illa´nGo´mez MJ, Bueno-Lo´pez A. Enhanced Pt stability in MO2 (M ¼ Ce, Zr or Ce0.9Zr0.1)-promoted Pt/C electrocatalysts for oxygen reduction reaction in PAFCs. App Catal A 2010;381:54e65. [34] Subramanian NP, Kumaraguru SP, Colon-Mercado H, Kim H, Popov BN, Black T, et al. Studies on Co-based catalysts supported on modified carbon substrates for PEMFC cathodes. J Power Sources 2006;157:56e63.
Please cite this article in press as: Corte´s-Escobedo CA, et al., Mechanically activated PteNi and PteCo alloys as electrocatalysts in the oxygen reduction reaction, International Journal of Hydrogen Energy (2014), http://dx.doi.org/10.1016/ j.ijhydene.2014.03.025
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/he
Mechanically activated PteNi and PteCo alloys as electrocatalysts in the oxygen reduction reaction Claudia A. Corte´s-Escobedo a,*, Rosa de G. Gonza´lez-Huerta b, Ana M. Boları´n-Miro´ c, Fe´lix Sa´nchez de Jesu´s c, Q. Zhu d, S.E. Canton d, Karina Suarez-Alcantara d,1, M. Tufin˜o-Velazquez e a
Centro de Investigacio´n e Innovacio´n Tecnolo´gica del Instituto Polite´cnico Nacional, Cda. Cecati s/n, Col. Sta. Catarina, CP 02250 Azcapotzalco, D.F., Mexico b ESIQIE-Instituto Polite´cnico Nacional, Laboratorio de Foto-Electrocata´lisis, UPALM, CP 07738 DF, Mexico c ´ Area Acade´mica de Ciencias de la Tierra y Materiales, Universidad Auto´noma del Estado de Hidalgo, CU, Carr. Pachuca-Tulancingo Km. 4.5, Mineral de la Reforma, CP 42184 Hidalgo, Mexico d Department of Synchrotron Radiation Instrumentation, Lund University, Ole Ro¨mers va¨g 1, SE-22363 Lund, Sweden e ESFM-Instituto Polite´cnico Nacional, Laboratorio de Fı´sica Avanzada, UPALM, CP 07738 DF, Mexico
article info
abstract
Article history:
Mixtures of powders of platinum with nickel or cobalt to obtain Ni0.75Pt0.25 or Co0.75Pt0.25
Received 27 November 2013
were mechanical alloyed by high energy ball milling. The results of crystal structure,
Received in revised form
morphology and electrocatalytic performance are presented for mechanically activated
24 February 2014
powders after 3 and 9 h of ball milling. Total solid solutions of Ni and Co with platinum
Accepted 5 March 2014
were analyzed by X-ray diffraction after 3 h of ball milling. After 9 h of ball milling, in both
Available online xxx
cases, the total solid solution was accompanied by the appearance of NiO or CoO and ZrO associated with a redox reaction with the milling media. The presence of zirconium
Keywords:
monoxide was confirmed by energy dispersive spectroscopy analysis. In both cases, an
NiPt
amorphization was detected. X ray absorption spectroscopy measurements showed
CoPt
changes in atomic and electronic environment of platinum, a reduction of the distance to
ORR
the first coordination sphere and increased d-band vacancy vs pure Pt and Pt nanoparticles
Electrocatalysts
were observed for both studied systems. The electrocatalytic activity was determined using
Mechanical alloying
cyclic and linear voltammetry. The Co0.75Pt0.25 alloy milled for 9 h showed a higher elec-
Anchor effect
trochemical activity for the oxygen reduction reaction (ORR) compared with the other samples, including Pt-Etek. The degree of the ORR electrochemical activity was correlated with the presence of ZrO, which could affect the oxygen adsorption and improve the catalytic activity for the oxygen reduction reaction. Copyright ª 2014, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
* Corresponding author. Tel.: þ52 55 57296000x68311, fax: þ52 5555617536. E-mail addresses: [email protected], [email protected] (C.A. Corte´s-Escobedo). 1
Present address: UNAM-IIM Morelia, Antigua carretera a Pa´tzcuaro 8710, Col. Ex-hacienda de San Jose´ de la Huerta, Morelia, Michoaca´n, 58190, Mexico. http://dx.doi.org/10.1016/j.ijhydene.2014.03.025 0360-3199/Copyright ª 2014, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Corte´s-Escobedo CA, et al., Mechanically activated PteNi and PteCo alloys as electrocatalysts in the oxygen reduction reaction, International Journal of Hydrogen Energy (2014), http://dx.doi.org/10.1016/ j.ijhydene.2014.03.025
2
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e9
Introduction Intense efforts have been made to develop new materials for electrocatalysis in low temperature polymer electrolyte fuel cells (PEMFCs) [1,2]. Both anode and cathode electrocatalysts require low platinum content in order to reduce costs. At the same time, it is intended to have high activity towards the oxygen reduction reaction (ORR), i.e. fewer kinetic limitations; to achieve low cathodic overpotential and to reduce the CO poisoning effect. Anode and cathode electrodes made of nanosized materials have shown the best activity towards the ORR. The electrocatalytic activity is strongly influenced by its particle size, the nature of the support and the preparation method [12]. To achieve an effective reduction in platinum content and poisoning effect, some transition metals have been used [3e16]. However, preparation of transition metal particles with a diameter size of only a few nanometers i.e. large surface areas is a challenge. Many researchers have employed different methods of synthesis to obtain novel nano-electrocatalyst of transition metals. Among these methods of synthesis, the mechanical milling (also known as ball milling) has been recently used [17e20]. With this technique, it is possible to obtain alloys, metastable phases, solid solutions, amorphous phases, nanocrystalline structures or composites. The mechanical milling can be used to prepare advanced materials that are difficult to produce by conventional fusion techniques. This work presents the mechanical synthesis of nano-sized Pt alloys (PteNi and PteCo) and evaluates their physical properties and electrocatalytic activity towards the ORR in an acid medium.
Materials and methods Elemental powders of Pt (SigmaeAldrich, >99.9%), Co (SigmaeAldrich, >99.9%) and Ni (SigmaeAldrich, >99.9%) were used as precursors. The raw materials Co and Ni were mixed in the appropriate weight ratio to obtain two different alloys: Co0.75Pt0.25 and Ni0.75Pt0.25. A total amount of 1 g of powder mixture together with 6 zirconia (ZrO2) balls, 11 mm in diameter, were loaded into a zirconia vial. The ball-to-powder weight ratio was 25:1. The mechanical alloying process was carried out at room temperature in air atmosphere using a shaker mixer/mill machine. To prevent excessive overheating of the vials, all experiments were carried out using cycles of 60 min of milling and 15 min of resting. The milling time spanned from 0 to 9 h to evaluate the effect of this parameter on the particle size distribution and the synthesis of the metallic alloy. X-Ray diffraction (XRD) patterns of the milled powders were used to study the phase transformations as a function of the composition and milling time, using a Bruker D2 Phaser. Diffraction parameters were the diffracting angle, A) was 2q, ranging from 20 to 90 . Cu Ka radiation (l ¼ 1.5418 used in all experiments. The morphology and composition of the milled powders were analyzed using a FEI Quanta 3D FEG scanning electron microscope under high vacuum. X-ray absorption spectroscopy (XAS) spectra were taken at the I811 beamline of MAX-Lab synchrotron facility, Sweden [21,22]. XAS at Pt L2,3-edge, Ni K-edge and Co K-edge of the 9-h
ball milled samples were collected. 1 mg of the sample was distributed over a 1.5 cm2 area of the glue-side of Kapton tape. The samples over Kapton were placed inside a He-filled chamber. The sample-side was oriented at 45 with respect to the incident beam, then XAS spectra were taken in fluorescence mode. Pt L3-edge spectra were taken between 100 and þ900 eV vs E0 ¼ 11564 eV energy, with a step ¼ 0.4 eV. Pt L2-edge spectra were taken between 100 and þ450 eV vs E0 ¼ 13273 eV energy, with a step ¼ 0.4 eV. A Zn filter was used before the Lytle detector during Pt L2,3-edge spectra collection, 5 scans were averaged. Ni K-edge spectra were taken between 50 and þ600 eV vs E0 ¼ 8333 eV, with a step ¼ 0.4 eV. A Co filter was used before the Lytle detector during Ni K-edge spectra collection, 5 scans were averaged. Co K-edge spectra were taken between 100 and þ800 eV vs E0 ¼ 7709 eV, with a step ¼ 0.4 eV. Fe filter was used before the Lytle detector during the Co K-edge spectra collection, 5 scans were averaged. Data were processed with the Athena program [23]: proper calibration with Pt, Ni and Co foil respectively was performed, background removal and normalization was performed before whiteline area integration. Electrochemical evaluation was carried out using a conventional three-electrode electrochemical cell. The electrodes were: 1) glassy carbon covered with a thin electrocatalyst film as working electrode (0.196 cm2 geometric area), 2) Hg/Hg2SO4, 0.5 M H2SO4 (E ¼ 0.680 V/NHE) as reference electrode and 3) platinum mesh as counter electrode (with more than 10 cm2 of surface area). A 0.5 M H2SO4 (Merck, p.a.) solution was used as electrolyte. The working electrodes were prepared according to the methodology reported elsewhere [17]. Briefly, the as milled electrocatalysts were supported in carbon Vulcan to a load of 10 wt% Pt/C: 5 mg of electrocatalyst, 24 mg of Vulcan and 200 mL of ethanol were sonicated for 20 min. Then the powders were dried at 40 C and placed in a closed glass container until their utilization. A catalytic “ink” of each electrocatalyst was prepared by mixing 60 mL of ethanol, 6 mL of Nafion (5 wt%, Du Pont 1000 EW) and 1 mg of electrocatalyst supported on carbon Vulcan (10 wt% Pt/C). 8 mL of the Ni0.75Pt0.25 or Co0.75Pt0.25 catalytic “ink” was uniformly dispersed over the surface of a glassy carbon electrode. After drying a thin film of electrocatalyst was formed. The rotating disk electrode (RDE) electrochemical characterization technique was performed by means of a 263 A potentiostat/galvanostat (EG&G PAR). The rotating rates were fixed at 100, 200, 400, 900, 1600 or 2500 rpm; the scan velocity was set at 5 mV s1. Previous (and necessarily) electrochemical activation was carried out by 20 successive potential sweep cycles at 50 mV s1 in a potential window of 0e1.2 V/ NHE to eliminate impurities. Potentials are reported with respect to the normal hydrogen electrode unless otherwise specified. Pt mass-specific activity (Im) was calculated at E ¼ 0.90 V/ NHE for an ORR polarization curve measured at a scan rate of 5 mV s1. The mass activities are estimated via calculation of jk (kinetic current) and normalization to LPt (the working electrode Pt loading). Catalyst electrocatalytic activity towards the ORR is quantified at E ¼ 0.90 V because interferences from mass-transport losses cannot be completely excluded at higher current densities observed below E ¼ 0.90 V [24].
Please cite this article in press as: Corte´s-Escobedo CA, et al., Mechanically activated PteNi and PteCo alloys as electrocatalysts in the oxygen reduction reaction, International Journal of Hydrogen Energy (2014), http://dx.doi.org/10.1016/ j.ijhydene.2014.03.025
3
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e9
Results and discussion In Fig. 1, the X-ray diffraction patterns of Ni0.75Pt0.25 electrocatalyst after 3 and 9 h of ball milling are presented. Ni powder without milling also is presented as visual comparison for crystal deformations. The Ni powder reflection peaks are consistent with the cubic close-packed structure of pure nickel (ICSD-41508). Fig. 1 reveals that 3 h of ball milling of Ni and Pt is enough to produce a total solid solution, where the Pt atoms are introduced into the Ni crystal lattice. A slight displacement of the nickel reflection peaks toward the left was observed from 44.4 to 43.74 in 2q degree indicating an expansion of the Ni crystal lattice caused by the big size of Pt lattice parameter a of nickel from 3.52 to 3.61 A. The Ni0.75Pt0.25 diffraction reflection peaks become wider as ball milling time increases due to the particle and grain size diminution. After 9 h of ball milling, the unexpected formation of NiO (ICSD61324) and ZrO (ICSD-72956) were evident. The formation of these oxides can be attributed to a redox reaction expressed in the following partial reactions due to Ni reactivity, Pt catalyst and the high energy process:
1
Ni þ /2 O2 / NiO þ 2e
(1)
ZrO2 þ 2e / ZrO þ 1/2O2
(2)
Ni (311)
Ni (220)
Ni (111)
Ni (200)
Results of Rietveld refinement of the XRD pattern indicated a reduction in crystal size from 573 to 166 A, an increment in cell parameter a from 3.52 to 3.61 A and an increment in rms microstrain from 0.0001 to 0.0031, however, because of the crystallite size, diffraction peaks are near to the level of noise, and the c2 is too high. Fig. 2 shows the XRD patterns of Co0.75Pt0.25 milled for 3 h and Co0.75Pt0.25 milled for 9 h. XRD of Co powder without milling is presented as visual reference. The Co powder is composed of 2 phases; the cubic Co-a (Co Fm-3m, ICSD 41507) and the hexagonal Co-b (ICSD-050727, P63/mmc, hcp). After 3 h
Intensity (a.u.)
Ni (222)
Pt
Ni
Ni
Pt
- 3h
Ni
Pt
- 9h ZrO NiO
20
30
40
ZrO NiO
50
60
NiO
70
80
NiO
90
100
2θ (degree) Fig. 1 e X-Ray diffraction of Ni and Ni0.75Pt0.25 mixtures milled for 3 and 9 h.
Fig. 2 e X-Ray diffraction patterns for Co and Co0.75Pt mixtures after high energy ball milling.
0.25
of milling, the Co0.75Pt0.25 powder presents only a wide peak at 43.9 in 2q, this can be associated to a deformed hexagonal structure, i.e. the amorphization of the Co-b structure and loss of Co-a structure. Pt diffraction peaks are not evident, confirming the incorporation of Pt into the deformed Co structure. The increase in the milling time promotes further deformation of the Co lattice, due to the mechanical energy induced by the process. After 9 h of ball milling cubic Co-a are observed. Again, two reflection peaks at about 2q ¼ 51.7 and 44 corresponding respectively to (200) and (111) are evident. The reversible transformation between Co-b (hcp) and Co-a (fcc) was already known, this depend upon the milling parameters and the formation of solid solutions [16]. Interesting, the formation of CoO (ICSD-9865) and ZrO (ICSD-72956) was observed. As in the Ni0.75Pt0.25 and nearly at the same time of ball milling, the oxide formation was present also in Co0.75Pt0.25. This cannot be associated with ball or vial fractures. We attribute the formation of ZrO and CoO to the following redox pair reactions:
Co þ 1/2 O2 / CoO þ 2e
(3)
ZrO2 þ 2e / ZrO þ 1/2 O2
(4)
Results of Rietveld refinement showed a reduction in crystallite size from 200 to 78 A after 3 h of milling, but then, after ZrO appearance, in 30 w% crystallite size increases up to 287 A and a cell parameter the beta structure of cobalt increases from 2.511 to 3.372 A after 3 h of milling but after 9 h, decrease to 2.841 A. While c parameter of the same structure diminishes from 4.083 to 3.258 A after 3 h and change to 3.885 A after 9 h of milling. Rms microstrain in the first step of milling increases from 0.0004 to 0.0739 and then decrease up to 0.0312 which can indicate a release of stresses promoted by the PteZrO presence. Again, diffraction peaks are near to the level of noise, and the c2 is too high.
Please cite this article in press as: Corte´s-Escobedo CA, et al., Mechanically activated PteNi and PteCo alloys as electrocatalysts in the oxygen reduction reaction, International Journal of Hydrogen Energy (2014), http://dx.doi.org/10.1016/ j.ijhydene.2014.03.025
4
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e9
Fig. 3 e SEM images of the Ni0.75Pt0.25 and Co0.75Pt
Fig. 3 presents the micrographs of Ni0.75Pt0.25 and Co0.75Pt0.25 electrocatalysts milled for 3 and 9 h. For both alloys, after 3 h of milling the presence of large aggregates of large particles are manifest. After 9 h of milling, the dispersion of small particles with grain sizes of approximately 60 nm were found, thus the available surface area is notably larger for samples milled for 9 h of ball milling. The Co0.75Pt0.25 surface also landscapes porosity. Table 1 collects the results of the elemental composition obtained by EDS for obtained alloys. Important is to note variations in atomic compositions due to increase in zirconium monoxide and cobaltum oxide formation, discussed before. Fig. 4(a) presents the Pt L3-edge XAS spectra for Ni0.75Pt0.25 and Co0.75Pt0.25. The plot also presents the XAS spectra of Pt foil and Pt commercial nanoparticles (ETEK) as reference for nanoparticle and alloy formation. The whole XAS spectra of Ni0.75Pt0.25 and Co0.75Pt0.25 show similar oscillations features; meanwhile different compared with the Pt foil, the Pt commercial nanoparticles. This reflect the changes in the Pt
Table 1 e Results of elemental composition by EDS for systems Ni0.75Pt0.25 and Co0.75Pt0.25 with milling. Sample
Ni Ni25Pt 3 h Ni25Pt 9 h Co Co25Pt 3 h Co25Pt 9 h
EDS composition O at%
Co at%
Ni at%
Pt at%
Zr at%
0 9.8 23.2 0 21.7 36.2
0 0 0 100 64.7 53.2
100 78.5 59.3 0 0 0
0 7.3 6.4 0 8.3 4.6
0 4.4 11.1 0 5.3 6.0
0.25
systems ball milled for 3 and 9 h.
surroundings according the synthetic procedure and the alloy composition. These changes in the geometrical environments are more evident at the EXAFS Fourier transformation (inset at ˚ 1, k3-weighted, the right in Fig. 4(a), k-range from 1.5 to 15.7 A phase corrected). The Pt foil and the Pt ETEK show the PtePt bond distance at 2.8 A while the Ni0.75Pt0.25 and Co0.75Pt0.25 show Pt-M (M ¼ Ni and Co) distances at 2.5 A in agreement with the alloy formation; and also in agreement with results obtained by X-ray diffraction. The inset at the left presents a zoom view of the Pt L3-edge whiteline and some oscillations above the edge. Evidently, changes in the electronic environment are induced after alloying. The Pt L3-edge whiteline results from the electronic transition 2p3/2 to 5d5/2 and its intensity is related to the Pt 5d vacancies [25]. However a complete analysis of the 5d-band vacancies is only completed by a simultaneous analysis of Pt L2-edge. Fig. 4(b) presents the near edge features of Pt L2-edge of the studied materials. Again, the oscillations above the Pt L2-edge of Ni0.75Pt0.25 and Co0.75Pt0.25 show similarities, meanwhile different from Pt foil and Pt ETEK nanoparticles. The whiteline at the Pt L2-edge is due to 2p1/2 to 5d5/2 transition, and is not so intense as the whiteline of Pt L3-edge: a phenomenon already described as non-statistical distribution of the densities of states of Pt 5d5/2 and 5d3/2 [26]. Murkerjee et al. [27] quantified the fractional change in d-band vacancies (fd) relative to a reference material (i.e. the Pt foil) as: fd ¼
DA3 þ 1:11DA2 ; ðA3 þ 1:11A2 Þr
(5)
where DA3 ¼ A3s A3r; DA2 ¼ A2s A2r; A3 and A2 are the areas under L3 and L2 edges respectively (10 to þ13 relative to the
Please cite this article in press as: Corte´s-Escobedo CA, et al., Mechanically activated PteNi and PteCo alloys as electrocatalysts in the oxygen reduction reaction, International Journal of Hydrogen Energy (2014), http://dx.doi.org/10.1016/ j.ijhydene.2014.03.025
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e9
Fig. 4 e (a). Pt L3-edge XAS spectra for Ni0.75Pt0.25 and Co0.75Pt0.25. (b). Near edge features of Pt L2-edge of the Ni0.75Pt0.25 and Co0.75Pt 0.25 systems.
edge position); s and r refer to the sample and reference respectively. The d-band vacancies of Pt samples are calculated as: hj
total sample
¼ 1 þ fd hj total
reference
(6)
Where the total number of unoccupied d-states, (hj)total reference, for pure Pt have been evaluated as 0.3 [27,28]. Thus the Pt d-orbital vacancy per atom, (hj)total sample, is 0.327 for Pt ETEK nanoparticles; 0.336 for Co0.75Pt0.25 and 0.334 for Ni0.75Pt0.25. Nanoparticle formation increases the Pt d-band vacancy. The alloy PteCo has an increased d-band vacancy compared to PteNi alloy. The d-band vacancy can play an important role for the activity towards the oxygen reduction reaction. Complementary information was obtained by XAS at Ni and Co K-edge. The K-edge XAS spectra of 3d transition metals describe the 1s to 3d transition if symmetry broken (pre-edge, no present here) and the 1s to 4p transition. Fig. 5(a) presents the XAS spectra at K-edge of Ni0.75Pt0.25 and Ni foil. The K-edge spectrum of Ni0.75Pt0.25 is strongly dominated by the NiO contribution [29,30]. The Ni at of Ni0.75Pt0.25 can be described as a mixture of metallic Ni (minor component)), NiO (mayor component) and NiePt alloy. Particles of NiO have been used to reduce the CO poisoning of Pt particles at fuel cells [31]. The inset of Fig. 5(a) presents the EXAFS Fourier transformation (k-
5
Fig. 5 e (a). XAS spectra at K-edge - EXAFS Fourier ˚ L1, k3-weighted, transformation (k-range from 2 to 10.5 A phase corrected) e of Ni0.75Pt0.25 and Ni foil. (b). XAS spectra at K-edge - EXAFS Fourier transformation (k-range from 2 ˚ L1, k3-weighted, phase corrected) e of Co0.75Pt0.25. to 10.5 A
˚ 1, k3-weighted, phase corrected). The Ni range from 2 to 10.5 A foil present the NieNi bond distance at 2.44 A (Ni ICSD-41508); A that is the NieO bond NiO the Ni0.75Pt0.25 has peaks at 2.08 (ICSD-61324), and at 2.66 A that is the NiePt bond (ICSD646297, PtNi alloy). Fig. 5(b) presents the Co K-edge spectrum of Co0.75Pt0.25, it is dominated by the features of CoO i.e. the shift of the edge position and the intense whiteline peak [30]. The inset of Fig. 5(b) displays the EXAFS Fourier transformation (k-range ˚ 1, k3-weighted, phase corrected). The peak at from 2 to 9.5 A 2.44 A agrees with the CoeCo bond distance of cubic Co. The peak at 2.08 A correlates with the CoeO bond distance (ICSD9865) and the 2.61 A is the CoePt bond distance (ICSD-102620). Thus, the ball milled Co0.75Pt0.25 can be described as a mixture of metallic Co (minor component), CoO (mayor component) and CoPt alloy. Cyclic voltammetry has become a technique for initial electrochemical studies of new systems and has proved being very useful in obtaining information about stability in reaction media. Fig. 6 shows cyclic voltammetry (CV) curves of the Ni0.75Pt0.25/C 3 and 9 h ball milled and Co0.75Pt0.25/C 3 and 9 h
Please cite this article in press as: Corte´s-Escobedo CA, et al., Mechanically activated PteNi and PteCo alloys as electrocatalysts in the oxygen reduction reaction, International Journal of Hydrogen Energy (2014), http://dx.doi.org/10.1016/ j.ijhydene.2014.03.025
6
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e9
Fig. 6 e Cyclic voltammograms synthesized by mechanical milling of the a) Ni0.75Pt0.25/C and b) Co0.75Pt0.25/C electrocatalysts. Data recorded in outgassed 0.5 M H2SO4 at 25 C. Scan rate of 50 mV sL1.
ball milled samples, in the de-aerated 0.5 M H2SO4 at 50 mV s1 in the range from 0.0 to 1.20 V at 25 C. The Ni0.75Pt0.25/C 3 h catalyst in Fig. 6(a) shows a striking low current density magnitude for all potential scans. This low electrochemically active area at both the hydrogen oxidation and the oxygen reduction regions and the low double layer capacitance, are presumably due to large aggregates and large particles. No evident influence of the large porosity of electrocatalyst particles was observed. The CV curve of Ni0.75Pt0.25/C 9 h (Fig. 6(a)) shows a significant improvement in the electrochemical activity. Typical characteristics of Pt nanoparticles were observed: the hydrogen adsorption and desorption peaks at the hydrogen region between 0.0 and 0.36 V, Pt oxidation starting at 0.80 V, and a reduction of Pt oxide film centered at 0.75 V. Fig. 6(b) shows the cyclic voltammetry (CV) curves of the Co0.75Pt0.25/C 9 h and 3 h catalysts. The Co0.75Pt0.25/C 3 h catalyst shows the same behavior as Ni0.75Pt0.25/C 3 h, almost none definition of Pt electrochemical features for oxygen reduction and hydrogen absorption/oxidation regions. The catalytic activity is low due to the large aggregates and large particles observed in Fig. 3. Co0.75Pt0.25/C 9 h presents a better
electrochemical response. Even more, the typical characteristics of Pt are better defined in the case of the Co0.75Pt0.25/C 9 h catalyst than in the case of Ni0.75Pt0.25/C 9 h interestingly, the Co0.75Pt0.25/C 9 h shows additional peaks at the hydrogen absorption/desorption region. These peaks, between 0.0 and 0.36 V, must be related to the presence of the second metal, Co and ZrO. Another effect of Co is the shift of the oxygen reduction potential from approximately 0.05 V to a more positive value of 0.71 V, which is showed by Ni0.75Pt0.25/C 9 h. The Co0.75Pt0.25/C catalysts shows high stability in the reaction medium, the CVs responses were equal after tenth cycle, which is showed in Fig. 6(b). The stability can be attributed to the presence of ZrO, as has been reported previously [32,33], causing an anchor effect of the platinum atoms, which inhibits agglomeration. The last effect is found only for the Co-based system. Due to the low solubility of oxygen in acid media, the oxygen reduction reaction (ORR) depends strongly on the hydrodynamic conditions. The ORR characteristics of polarization curves from RDE measurements, at different rotation rates, of Ni0.75Pt0.25/C 9 h and Co0.75Pt0.25/C 9 h are summarized in Fig. 7(a and b). Comparing the two electrocatalyst, two characteristics
Fig. 7 e Currentepotential curves for oxygen reduction for a) Ni0.75Pt0.25/C 9 h and b) Co0.75Pt0.25/C 9 h, and in 0.5 M H2SO4 saturated with oxygen at different rotation rates. Currents recorded at 5 mV sL1. Please cite this article in press as: Corte´s-Escobedo CA, et al., Mechanically activated PteNi and PteCo alloys as electrocatalysts in the oxygen reduction reaction, International Journal of Hydrogen Energy (2014), http://dx.doi.org/10.1016/ j.ijhydene.2014.03.025
7
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e9
Table 2 e Kinetic parameter obtained from the oxygen reduction reaction (ORR). Electrocatalyst
Eca V/ENH
-b V dec1
a
jo (mA cm2)
j/mA cm2 at V ¼ 0.85 V
LPt mgPt
Im mA 2 mg1 Pt cm
Ni0.75Pt0.25/C-9 h Co0.75Pt0.25/C-9 h Pt-Etek
0.95 0.06 1.05 0.05 0.98 0.03
0.100 0.008 0.083 0.005 0.080 0.002
0.59 0.03 0.70 0.02 0.71 0.01
1.49 1040.00004 1.38 1040.00003 1.25 1040.00002
0.60 0.08 3.33 0.05 1.36 0.03
7.7 5.5 12
0.031 0.129 0.080
are revealed: 1) The Co0.75Pt0.25/C 9 h electrocatalyst shows the higher open circuit potential, i.e. an overpotential reduction of about 50 mV compared to Ni0.75Pt0.25/C 9 h (see Table 2). Also there is an improvement in the total current density observed at the same hydrodynamic conditions. During the electrochemical reactions, either the charge or the mass transfer define the rate determining step. Both processes are distinguished at the polarization curves. The RDE curves of both materials present well defined regions: 1) at low overpotential (0.98e0.9 V/NHE), the charge transfer dominates and the current density is independent of the electrode rotation speed. 2) The mixed kinetic-diffusion control region (0.9e0.7 V/NHE), and 3) the diffusion limiting current region (0.7e0.2 V/NHE), where the increase of the rotation rate results in an increase of current density. Thus the kinetic parameters exchange current density and Tafel slope were estimated. RDE experiments were performed from high to low rotating rates, Co0.75Pt0.25/C catalysts presents a well-defined kinetic region, Fig. 7(b), while Ni0.75Pt0.25/C catalysts has low stability in the reaction medium, Fig. 7(a) shows a slight shift to lower potential, between each experiment, in the same region. Fig. 8 shows the mass-transfer-corrected Tafel plots of Ni0.75Pt0.25/C (3 h and 9 h) and Co0.75Pt0.25/C (3 h and 9 h) prepared by mechanical milling. Fig. 8 also shows the kinetic current density of Pt-Etek only for comparison purposes. The kinetic current densities were normalized to the geometrical area of the electrode, the mathematical procedure for kinetic parameters estimation is described elsewhere [34]. The values of open circuit potential, Eoc, current density exchange jo and Tafel slope are collected in Table 2 The open circuit potential,
Eoc, is monitored each experiment, in ORR it indicates material stability and is directly related to catalytic activity, in platinum compound its value should be from 0.9 to 1 V, Co0.75Pt0.25/C 9 h shows higher Eoc than Pt-Etek, Table 2. The comparison of current densities between the Ni0.75Pt0.25/C and Co0.75Pt0.25/C electrocatalysts at 0.85 V shows that the highest current density is obtained for Co0.75Pt0.25/C 9 h. The Tafel slope, b, of about 0.080 V dec1 was deduced for both the Co0.75Pt0.25/C-9 h and the P-Etek electrocatalysts, as expected for a first electron transfer as the rate determining step [19]. Values of the exchange current densities are consistent with those reported for nanosized catalysts containing Pt [12]; indicating that the ball milling as synthetic procedure is appropriate for electrocatalyst preparation. It is well known that the correlation between the d-band vacancies, the Pt-Metal bond distances, absorption of oxygenated species [27] and particle sizes have a combined effect on the electrocatalytic activity towards the ORR. Electrochemical results show that the incorporation of Co into the Pt structure improves its catalytic activity towards the ORR. Xray diffraction and X-ray absorption confirm a reduced PtMetal bond distance (vs PtePt). This reduced Pt-Metal bond distance have been identified as an active site suitable for absorption and further cleavage of oxygen. Both electrocatalyst show similar bond distances, thus another parameter is the main responsible for the enhanced electrochemical response. The Pt L3-edge XAS reveals that the Co0.75Pt0.25 has increased Pt d-band vacancy and also have improved activity towards the ORR compared with Co0.75Pt0.25, Pt nanoparticles and Pt foil. Additionally SEM of 9 h ball milled Co0.75Pt0.25 demonstrate the nanosized nature of this electrocatalyst, feature not shared with the Ni0.75Pt0.25. Finally the oxide formation ZrO/CoO or ZrO/NiO can act as co-catalyst towards the ORR [33]. The numerical values of the Pt mass-specific activity (Im) at E 0.90 V versus NHE derived from the ORR polarization curves measured at 5 mV s1, shown in Fig. 7, are detailed in Table 2. 2 A Pt-Etek has a Im of 0.080 mA mg1 Pt cm . The Im of the Co0.75Pt0.25/C-9 h is w60% higher than that of the Pt-Etek electrode and. Ni0.75Pt0.25/C-9 h has the lowest Im, 0.031 mA 2 mg1 Pt cm . The Im values reported in Table 2 demonstrate the importance of having a good electrode when evaluating the activity of a catalyst.
Conclusions
Fig. 8 e Mass transfer-corrected Tafel plots deduced from RDE analysis of oxygen-saturated 0.5 M H2SO4 at 25 C.
Ni0.75Pt0.25 and Co0.75Pt0.25 alloys, obtained by mechanical milling showed activity towards ORR. These samples are composed of a solid solution of platinum atoms inserted into Ni or Co structures after 3 h of milling. After 9 h of milling, the unexpected presence of ZrO/CoO or ZrO/NiO was observed,
Please cite this article in press as: Corte´s-Escobedo CA, et al., Mechanically activated PteNi and PteCo alloys as electrocatalysts in the oxygen reduction reaction, International Journal of Hydrogen Energy (2014), http://dx.doi.org/10.1016/ j.ijhydene.2014.03.025
8
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e9
consequence of a redox reaction between the milling media and the Ni or Co. SEM micrographs show that both materials are porous; particularly the 9 h ball milled Co0.75Pt0.25 presents large available surface area for ORR. The Pt L3-edge XAS reveals that the Co0.75Pt0.25 increased Pt d-band vacancy compared with Pt nanoparticles and Pt foil. Ni and Co K-edges indicate dominance of NiO or CoO respectively. The combination ZrO/CoO seems to play an important role in electrocatalytic activity due to the anchor effect. The electrocatalytic activity of Co0.75Pt0.25/C-9 h in the ORR is higher than Ni0.75Pt0.25/C-9 h. Co0.75Pt0.25/C-9 h also showed better stability in acid medium than Ni0.75Pt0.25/C-9 h. The Im of the Co0.75Pt0.25/C-9 h system is w60% higher than that of the PtEtek electrode.
Acknowledgments Authors thank the Center of Nanosciences and Micro and Nanotechnology of the National Polytechnic Institute, in particular Dr. Jose´ Alberto Andraca Adame and Dr. Hugo Martı´nez Gutie´rrez for structural and morphological characterization. Thanks for financial support: SECITI agreement ICYTDF/127/2012, CONACYT under projects CB-157925, CB130413 and SIP-IPN under project MULTI-1540. The authors are grateful to MAX-Lab synchrotron. Authors acknowledge the advice of Dr. Stefan Carlson and Dr. Katarina Nore´n during XAS data collection. KSA, QS and SEC were supported by the Swedish Research Council and the Crafoord Foundation.
references
[1] Barbir F. PEM Fuel cells, theory and practice. California: Academic Press, Sustainable world series; 2005. [2] Norskov JK, Rossmeisl J, Logadottir A, Lindqvist L, Kitchin JR, Bligaard T, et al. Origin of the overpotential for oxygen reduction at a fuel-cell cathode. J Phys Chem B 2004;108:17886e92. [3] Zignani SC, Antolini E, Gonzalez ER. Stability of PteNi/C (1:1) and Pt/C electrocatalysts as cathode materials for polymer electrolyte fuel cells: effect of ageing tests. J Power Sources 2009;191:344e50. [4] Zhao J, Jarvis K, Ferreira P, Manthiram A. Performance and stability of PdePteNi nanoalloy electrocatalysts in proton exchange membrane fuel cells. J Power Sources 2011;196:4515e23. [5] Zhao J, Manthiram A. Preleached PdePteNi and binary PdePt electrocatalysts for oxygen reduction reaction in proton exchange membrane fuel cells. App Cat B: Enviro 2011;101:660e8. [6] Su F, Poh CK, Zeng J, Zhong Z, Liu Z, Lin J. Pt nanoparticles supported on mesoporous carbon nanocomposites incorporated with Ni or Co nanoparticles for fuel cells. J Power Sources 2012;205:136e44. [7] Koh S, Toney MF, Strasser P. Activityestability relationships of ordered and disordered alloy phases of Pt3Co electrocatalysts for the oxygen reduction reaction (ORR). Electrochem Acta 2007;52:2765e74. [8] Di Noto V, Negro E. Development of nano-electrocatalysts based on carbon nitride supports for the ORR processes in PEM fuel cells. Electrochem Acta 2010;55:7564e74.
[9] Neyerlin KC, Srivastava R, Yu C, Strasser P. Electrochemical activity and stability of dealloyed PteCu and PteCueCo electrocatalysts for the oxygen reduction reaction (ORR). J Power Sources 2009;186:261e7. [10] Lee MH, Do JS. Kinetics of oxygen reduction reaction on Corich coreePtrich shell/C electrocatalysts. J Power Sources 2009;188:353e8. [11] Jayasayee K, Van Veen JAR, Manivasagam TG, Celebi S, Hensen EJM, Bruijn FA. Oxygen reduction reaction (ORR) activity and durability of carbon supported PtM (Co, Ni, Cu) alloys: Influence of particle size and non-noble metals. Appl Cat B: Enviro 2012;111:515e26. [12] Godı´nez-Salomo´n F, Hallen-Lo´pez M, Solorza-Feria O. Enhanced electroactivity for the oxygen reduction on Ni@Pt core-shell nanocatalysts. Int J Hyd Energy 2012;37:14902e10. [13] Garcı´a-Contreras MA, Ferna´ndez-Valverde SM, Vargas´ ngelesGarcı´a JR, Corte´s-Ja´come MA, Toledo-Antonio JA, A Chavez C. PtCo and PtNi electrocatalysts prepared by mechanical alloying for the oxygen reduction reaction in 0.5 M H2SO4. Int J Hyd Energy 2008;33:6672. [14] Antolini E, Salgado JRC, Gonza´lez ER. The methanol oxidation reaction on platinum alloys with the first row transition metals The case of PteCo and eNi alloy electrocatalysts for DMFCs: a short review. Appl Cat B: Enviro 2006;63:137e49. [15] Antolini E, Salgado JRC, Giz MJ, Gonza´lez ER. Effects of geometric and electronic factors on ORR activity of carbon supported PteCo electrocatalysts in PEM fuel cells. Int J Hyd Energy 2005;30:1213e20. [16] Betancourt-Cantera JA, Sa´nchez-De Jesu´s F, TorresVillasen˜or G, Boları´n-Miro´ AM, Corte´s-Escobedo CA. Extended solid solubility of a CoeCr system by mechanical alloying. J All Comp 2012;529:58e62. [17] Ezeta A, Arce EM, Solorza O, Gonzalez Huerta RG, Dorantes H. Effect of the leaching of RueSeeFe and RueMoeFe obtained by mechanical alloying on electrocatalytical behavior for the oxygen reduction reaction. J All Comp 2009;483:429e31. [18] Collins-Martı´nez V, Gonza´lez-Huerta RG, Lo´pez-Ortiz A, Delgado-Vigil D, Solorza-Feria O. Oxygen reduction kinetics of ball-milling synthesis Pd electrocatalysts in acid medium. J New Mat Electrochem Systems 2009;12:63e7. [19] Gonzalez-Huerta RG, Rodriguez Pierna A, Solorza-Feria O. Kinetic study of oxygen reduction reaction on an amorphous Ni59Nb40Pt0.6Ru0.4 catalyst in acid media. J New Mat Electrochem Systems 2008;11:63e7. [20] Sotelo-Mazo´n P, Gonza´lez-Huerta RG, Caban˜as-Moreno JG, Solorza-Feria O. Mechanically milled RuxFey electrocatalyst for oxygen reduction in acid media. Int J Electrochem Sci 2007;2:523e33. [21] Grehk TM, Nilsson PO. The design of the material science beamline, I811, at MAX-II. Nucl Instr and Meth in Phys Res A 2001;467-468:635e8. [22] Carlson S, Clausen M, Gridneva L, Sommarin B, Svensson C. XAFS experiments at beamline I811, MAX-lab synchrotron source. Sweden. J Synchrotron Rad 2006;13:359e69. [23] Ravel B, Newville M. ATHENA, ARTEMIS, HEPHAESTUS: data analysis for X-ray absorption spectroscopy using IFEFFIT. J Synchrotron Radiat 2005;12. 537e341. [24] Garsany Y. Experimental methods for quantifying the activity of platinum electrocatalysts for the oxygen reduction reaction. Anal Chem 2010;82:6321e8. [25] Merte LR, Behafarid F, Miller DJ, Friebel D, Cho S, Mbuga F, et al. Electrochemical oxidation of size-selected Pt nanoparticles studied using in situ high-energy-resolution Xray absorption spectroscopy. ACS Catal 2012;2:2371e6. [26] Sham TK, Ward MJ, Muephy MW, Lui LJ, Han WQ. Pt L3,2edge whiteline anomaly and its implications for the chemical behavious of Pt 5d5/2 and 5d3/2 electronic states e
Please cite this article in press as: Corte´s-Escobedo CA, et al., Mechanically activated PteNi and PteCo alloys as electrocatalysts in the oxygen reduction reaction, International Journal of Hydrogen Energy (2014), http://dx.doi.org/10.1016/ j.ijhydene.2014.03.025
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e9
[27]
[28] [29]
[30]
a study of PteAu nanowires and nanoparticles. J Phys Conf Ser 2013;430:012018e24. Mukerjee S, Srinivasan S, Soriaga M, McBreen J. Role of structural and electronic properties of Pt and Pt alloys on electrocatalysis of oxygen reduction. J Electrochem Soc 1995;142:1409e22. Brown M, Peierls RE, Stern EA. White lines in X-ray absorption. Phys Rev B 1997;15:738e44. Calandra M, Rueff JP, Gougoussis C, Ceolin D, Gorgoi M, Benedetti S, et al. K-edge X-ray absorption spectra in transition-metal oxides beyond the single-particle approximation: Shake-up many-body effects. Phys Rev B 2012;86:165102. Cantane DA, Oliveira FER, Santos SF, Lima FHB. Syntesis of Pt-based hollow nanoparticles using carbon-supported Co@Pt and Ni@Pt core-shell structures as templates: Electrocatalytic activity for the oxygen reduction reaction. App Catal B: Enviro 2013;136-137:351e60.
9
[31] Park B, Cairns EJ. Electrochemical performance of TiO2 and NiO as fuel cell electrode additives. Electrochem Comm 2011;13:75e7. [32] Liu G, Zhang H, Zhong H, Hu J, Xu D, Shao Z. A novel sintering resistant and corrosion resistant Pt4ZrO2/C catalyst for high temperature PEMFCs. Electrochim Acta 2006;51:5710e4. [33] He Q, Mukerjee S, Zeis R, Parres-Esclapez S, Illa´nGo´mez MJ, Bueno-Lo´pez A. Enhanced Pt stability in MO2 (M ¼ Ce, Zr or Ce0.9Zr0.1)-promoted Pt/C electrocatalysts for oxygen reduction reaction in PAFCs. App Catal A 2010;381:54e65. [34] Subramanian NP, Kumaraguru SP, Colon-Mercado H, Kim H, Popov BN, Black T, et al. Studies on Co-based catalysts supported on modified carbon substrates for PEMFC cathodes. J Power Sources 2006;157:56e63.
Please cite this article in press as: Corte´s-Escobedo CA, et al., Mechanically activated PteNi and PteCo alloys as electrocatalysts in the oxygen reduction reaction, International Journal of Hydrogen Energy (2014), http://dx.doi.org/10.1016/ j.ijhydene.2014.03.025