pubs.acs.org/Langmuir © 2010 American Chemical Society
Enhancement in Electro-Oxidation of Methanol over PtRu Black Catalyst through Strong Interaction with Iron Oxide Nanocluster† Min Ku Jeon, Ki Rak Lee, and Seong Ihl Woo* Department of Chemical and Biomolecular Engineering (BK21 Graduate Program) & Center for Ultramicrochemical Process Systems, Korea Advanced Institute of Science and Technology, Yuseong-gu, Daejeon 305-701, Republic of Korea Received April 6, 2010. Revised Manuscript Received May 10, 2010 One of the major issues in direct methanol fuel cell research is to develop a new catalyst for methanol electro-oxidation reaction (MOR) with high activity and low cost. In this study, a new, simple, and economic way was introduced to improve the catalytic activity of commercial PtRu black catalyst for the MOR. A nanocomposite electrode was fabricated by mixing the PtRu catalyst with Fe2O3 nanoclusters. When 10 wt % of the PtRu catalyst was replaced by the Fe2O3 nanoclusters, mass activity (A/gPt) increased by 80% compared to that of the pure PtRu catalyst. Specific activity of the mixed catalyst was 100% higher than that of the pure PtRu catalyst. The nanocomposite catalysts were also applied to single cells. Although the amount of the PtRu catalyst was reduced by 10 wt %, 10% higher potential was observed in the nanocomposite catalysts at a current density of 100 mA/cm2.
1. Introduction Methanol electro-oxidation reaction (MOR) catalysts are under intensive research for application in direct methanol fuel cells (DMFCs). DMFCs use methanol and oxygen as anode and cathode reactants, respectively, to produce electricity, CO2, and water. Reduction of the required volume was achieved by using liquid fuel, methanol, while other fuel cells use gas fuel, hydrogen, as their anode reactant. However, the use of methanol reduces the catalytic activity of anode catalysts, which results in poor performance of DMFCs.1-3 Pt was first introduced as a MOR catalyst, but rapid decrease of activity was observed due to CO poisoning. Incorporation of Ru could dramatically improve CO electro-oxidation activity of Pt catalysts, leading to high MOR activity.4-6 The improved CO electro-oxidation activity can be explained by two mechanisms: mainly bifunctional mechanism7,8 and minor electronic effect.9-12 The bifunctional mechanism works by discharging water on the Ru surface to produce Ru-OH, which reacts with Pt-CO. The electronic effect works by changing electronic state of Pt; incorporation of Ru or transition metals lowers the electron density of Pt and weakens Pt-CO bonding. †
Part of the Molecular Surface Chemistry and Its Applications special issue. *To whom correspondence should be addressed. Telephone: þ 82-42-3503918. Fax: þ 82-42-350-8890. E-mail:
[email protected]. (1) Arico, A. S.; Srinivasan, S.; Antonucci, V. Fuel Cells 2001, 1, 133–161. (2) Thomas, S. C.; Ren, X.; Gottesfeld, S.; Zelenay, P. Electrochim. Acta 2002, 47, 3741–3748. (3) Petrii, O. A. J. Solid State Electrochem. 2008, 12, 609–642. (4) Watanabe, M.; Motoo, S. J. Electroanal. Chem. 1975, 60, 267–273. (5) Markovic, N. M.; Gasteiger, H. A.; Ross, P. N., Jr. Electrochim. Acta 1995, 40, 91–98. (6) Chrzanowski, W.; Wieckowski, A. Langmuir 1998, 14, 1967–1970. (7) Gasteiger, H. A.; Markovic, N.; Ross, P. N.; Cairns, E. J. J. Phys. Chem. 1993, 97, 12020–12029. (8) Yajima, T.; Uchida, H.; Watanabe, M. J. Phys. Chem. B 2004, 108, 2654– 2659. (9) Rodriguez, J. A.; Goodman, D. W. Science 1992, 257, 897–903. (10) Kitchin, J. R.; Nørskov, J. K.; Barteau, M. A.; Chen, J. G. Phys. Rev. Lett. 2004, 93, 156801. (11) Lu, C.; Rice, C.; Masel, R. I.; Babu, P. K.; Waszczuk, P.; Kim, H. S.; Oldfield, E.; Wieckowski, A. J. Phys. Chem. B 2002, 106, 9581–9589. (12) Rigsby, M. A.; Zhou, W.-P.; Lewera, A.; Duong, H. T.; Bagus, P. S.; Jaegermann, W.; Hunger, R.; Wieckowski, A. J. Phys. Chem. C 2008, 112, 15595– 15601.
Langmuir 2010, 26(21), 16529–16533
Previously, we reported improved CO and methanol electrooxidation activity of the PtRuFe/C catalyst.13 In this study, it was revealed that Fe existed as its oxide form, indicating that the PtRuFe/C catalyst is a nanoscale mixture of PtRu and Fe2O3. An increase in binding energy showed that incorporation of Fe2O3 lowered the electron density of Pt, which resulted in weak Pt-CO bonding. This result suggested that alloying is not essential to improve the MOR activity when metal oxide is incorporated. And naturally, a question has risen: Can we improve the MOR activity just by mixing metal oxides and conventional catalysts? There have been reports that incorporation of CeO2 into Pt14,15 or PtRu16 catalyst could achieve higher MOR activity without alloy formation. In addition, Wang et al.17 reported that they could observe improved CO electro-oxidation and MOR activity by mixing carbon nanotube supported Pt with CeO2 nanoparticles. In the present study, we fabricated composite electrodes by mixing commercial PtRu black catalyst with iron oxide (Fe2O3) nanoclusters to improve CO and methanol electro-oxidation activity by modifying the electronic structure of Pt.
2. Experimental Section 2.1. Synthesis and Characterization of Fe2O3 Nanoclusters. The Fe2O3 nanoclusters were synthesized by reducing iron precursor with NaBH4. Ammonium iron sulfate, (NH4)2Fe(SO4)2, was dissolved in deionized (DI) water and then stirred for 1 h at 80 °C. NaBH4 (0.2 M) solution was added to the precursor solution as a reducer followed by stirring for 3 h to complete the reduction reaction. The final mixture was filtered and washed with DI water. The resulting powder was dried in an oven at 100 °C overnight. Structural characterization of the Fe2O3 nanoclusters was performed (13) Jeon, M. K.; Won, J. Y.; Lee, K. R.; Woo, S. I. Electrochem. Commun. 2007, 9, 2163–2166. (14) Campos, C. L.; Roldan, C.; Aponte, M.; Ishikawa, Y.; Cabrera, C. R. J. Electroanal. Chem. 2005, 581, 206–215. (15) Takahashi, M.; Mori, T.; Vinu, A.; Kobayashi, H.; Drennan, J.; Ou, D.-R. J. Mater. Res. 2006, 21, 2314–2322. (16) Guo, J. W.; Zhao, T. S.; Prabhuram, J.; Chen, R.; Wong, C. W. J. Power Sources 2006, 156, 345–354. (17) Wang, J.; Deng, X.; Xi, J.; Chen, L.; Zhu, W.; Qiu, X. J. Power Sources 2007, 170, 297–302.
Published on Web 05/26/2010
DOI: 10.1021/la101355d
16529
Article
Jeon et al.
via X-ray diffraction (XRD). The XRD measurement was performed in a continuous scan mode over a 2θ range between 20 and 80° at a scan rate of 3 o/min. Transmission electron microscopy (TEM) measurement was also performed.
2.2. Electrochemical Characterization of PtRu-Fe2O3 Nanocomposite Catalysts. A thin film method was used to prepare half-cell working electrodes for the electrochemical experiments.18 Certain amounts of the Fe2O3 nanoclusters and the commercial PtRu catalyst (HiSPEC6000, Johnson-Matthey) were put into a vial with various PtRu/Fe2O3 mass ratios of 100:0, 90:10, 80:20, 70:30, and 50:50. The mixture was dispersed in DI water via sonication to obtain homogeneous mixing, and then 20 μL of the dispersion was dripped on a glassy carbon working electrode (3 mm dia., BAS Co., Ltd., MF-2012). Catalyst loading was 1.13 mg of (PtRu þ Fe2O3)/cm2. After drying in air, 20 μL of 5 wt % Nafion ionomer solution was dripped on the catalyst layer to provide mechanical strength in the catalyst layer. For the electrochemical experiments, a beaker-type three electrode cell was employed. Pt wire and Ag/AgCl electrode (BAS Co., Ltd., MF-2052 RE-5B) were used as the counter and reference electrodes, respectively. For CO stripping tests, CO was adsorbed on the working electrode by bubbling CO through the cell for 1 h, and then dissolved CO was removed by purging the electrolyte with N2. During the adsorption and purging processes, the working electrode was kept at 0.1 V (vs reversible hydrogen electrode, RHE). Electro-oxidation of CO was measured by increasing the potential from 0.1 to 1.2 V (vs RHE) at a scan rate of 15 mV/s, and then the potential was cycled between 0 and 1.2 V (vs RHE). HClO4 (1 M) solution was used for the CO stripping experiments. MOR activity was measured by potential cycling between 0 and 0.8 V (vs RHE) at a scan rate of 15 mV/s. Nitrogen purged 1 M H2SO4 þ 1 M methanol solution was used as the electrolyte. All potentials in this paper were scaled versus RHE.
2.3. Fabrication and Characterization of Membrane Electrode Assembly (MEA). Two different anodes were prepared by
using (1) the pure PtRu black catalyst and (2) the 10 wt % Fe2O3 þ 90 wt % PtRu mixed catalyst, respectively. Pt black catalyst (HiSPEC1000, Johnson-Matthey) was employed for the cathode electrodes. Catalyst inks were prepared by mixing the catalyst, DI water, isopropyl alcohol, and Nafion ionomer solution. MEAs were fabricated by a decal transfer method. The inks were put through sonication for homogeneous mixing and then sprayed on poly(tetrafluoroethylene) (PTFE) sheets until total metal loading of 2 mg/cm2 was obtained. The PTFE sheets were hot pressed with Nafion115 membrane (DuPont) at 1000 psi and 120 °C for 3 min. After the hot pressing, the PTFE sheets were removed, leaving the catalyst layers on both sides of the Nafion membrane. The active area of the MEAs was 4 cm2. The MEAs were operated at 70 °C. In the anode, 1 M methanol solution was fed at a flow rate of 0.5 mL/min, and oxygen was supplied into the cathode at a flow rate of 100 cc/min. During electrochemical impedance spectroscopy (EIS) measurement, hydrogen was fed into the cathode at a flow rate of 30 cc/min to work as both the counter and reference electrodes.19 The EIS measurement was performed by changing the frequency from 105 to 0.01 Hz using a Solartron 1255B frequency response analyzer.
3. Results and Discussion The XRD and TEM results of the NaBH4 reduced Fe2O3 nanoclusters are shown in Figure 1a and b, respectively. The XRD results clearly showed that the γ-Fe2O3 phase was formed. The crystallite size determined by Debye-Scherrer equation was 13.0 nm. TEM results showed that a spherical shape was obtained in the synthesis (18) Schmidt, T. J.; Gasteiger, H. A.; St€ab, G. D.; Urban, P. M.; Kolb, D. M.; Behm, R. J. J. Electrochem. Soc. 1998, 145, 2354–2358. (19) Jeon, M. K.; Won, J. Y.; Oh, K. S.; Lee, K. R.; Woo, S. I. Electrochim. Acta 2007, 53, 447–452.
16530 DOI: 10.1021/la101355d
Figure 1. (a) XRD pattern and (b) TEM images of synthesized γ-Fe2O3 nanoclusters.
procedure, and the particle size was not much different from the XRD results. Figure 2a shows the CO stripping results of the PtRu-Fe2O3 composite catalysts, where the mass ratio of Fe2O3 was changed from 0 to 10, 20, 30, and 50 wt % of total catalyst mass. Enlarged images near CO electro-oxidation onset potential are displayed in Figure 2b, where lowered onset potentials for the CO electrooxidation are clearly shown in the nanocomposite catalysts. The onset potential for the CO electro-oxidation was 0.4 V (vs reversible hydrogen electrode, RHE) in the pure PtRu catalyst, which was reduced to 0.3 V (vs RHE) in the nanocomposite catalysts. This result clarifies that the CO electro-oxidation activity was significantly improved by mixing the Fe2O3 nanoclusters with the commercial PtRu catalyst. The electrochemically active surface area (EAS) was calculated from the CO electro-oxidation area based on 420 μC/cm2 for monolayer Pt, and the results are listed in Table 1. A more rapid decrease of EAS than increase of the mass ratio of Fe2O3 might come from blockage of active sites of the PtRu catalysts by the Fe2O3 nanoclusters and isolation of the PtRu particles surrounded by Fe2O3 nanoclusters. The MOR activity measurement results are shown in Figure 3. An especially high current density of 18 mA/cm2 was observed in the 10 wt % Fe2O3 mixed catalyst at 0.5 V. Further increase in Fe2O3 content resulted in the reduction of current density. The mass activity of the 10 wt % Fe2O3 catalyst was 18 A/gPtRu which was 88% higher than 9.6 A/gPtRu of the pure PtRu catalyst. The 20 and 30 wt % Fe2O3 mixed catalysts also showed higher mass activity than that of the pure PtRu catalyst. Specific activities can be calculated by dividing the mass activities by EAS, which are listed in Table 1. Interestingly, 10 wt % Fe2O3 catalyst exhibited 100% higher specific activity of 380 mA/m2 than 190 mA/m2 of the pure PtRu catalyst, indicating that mixing the PtRu catalyst with the Fe2O3 nanoclusters could dramatically improve the MOR activity of the PtRu catalyst. In the case of the 20 and 30 wt % Fe2O3 mixed catalysts, the same specific activity of 320 mA/m2 Langmuir 2010, 26(21), 16529–16533
Jeon et al.
Article
Figure 2. (a) CO stripping results of (1) pure PtRu and Fe2O3 mixed nanocomposite catalysts which contain (2) 10, (3) 20, (4) 30, and (5) 50 wt % Fe2O3. (b) Enlarged images of (a) near onset potentials. Table 1. Electrochemical Experiment Results of the Pure and Fe2O3 Mixed PtRu Catalysts Fe2O3 content
EAS (m2/gcat) I (mA/cm2)a
mass activity specific activity (A/gPtRu)a (mA/m2)a
0 wt % 51 11 9.6 10 wt % 42 18 18 20 wt % 41 15 16 30 wt % 26 9.5 12 50 wt % 18 1.7 3.0 a Based on current density at 0.5 V (vs RHE).
190 380 320 320 83
was observed, which is still 68% higher than that of the pure PtRu catalyst. When mass ratio of the Fe2O3 nanoclusters was 50%, the lowest mass and specific activities of 3.0 A/gPtRu and 83 mA/m2 were observed, respectively. This rapid decrease of activities can be explained by a percolation effect; Fe2O3 is a semiconductor material, and thus, isolation of the PtRu catalyst by the Fe2O3 nanoparticles can cause significant drop of performance as the isolated PtRu catalyst particles have no paths for electron transportation. We applied the 10 wt % Fe2O3 catalyst into a single cell by fabricating the MEAs, and I-V test results are shown in Figure 4a. The effect of the Fe2O3 nanoclusters was compared by fabricating two MEAs which have different anode catalysts of (1) the pure Langmuir 2010, 26(21), 16529–16533
Figure 3. (a) Methanol electro-oxidation activity measurement results of (1) pure PtRu and Fe2O3 mixed nanocomposite catalysts which contain (2) 10, (3) 20, (4) 30, and (5) 50 wt % Fe2O3. (b) Enlarged images of (a) near 0.5 V.
PtRu (MEA-P) and (2) mixture of 10 wt % Fe2O3 and 90 wt % PtRu (MEA-M). In the MEA-M, the potential of the single cell at 100 mA/cm2 was 0.45 V, which was 10% higher than 0.41 V in the MEA-P. In Figure 4a, the difference of potentials between the two MEAs is larger at the low potential region of 0-100 mA/cm2, where catalytic activity is the most dominant factor among the three major parameters (activation loss from low catalytic activity, IR loss from resistance of components and interfaces, and diffusion limitation by slow mass transport) which determines the performance of DMFCs.20 On the other hand, a higher rate of performance drop in the MEA-M than in the MEA-P was observed when the current density was higher than 100 mA/cm2, which means that the MEA-M has higher IR resistance than the MEA-P. Here, quantitative analysis of the MEAs was performed by using the EIS analysis technique. Figure 4b shows the EIS measurement results of the MEAs and fitting results. To remove the effect of the cathode, hydrogen was fed into the cathode so that the cathode could work as both counter and reference electrodes. An equivalent circuit used for the fitting of the experimental results is shown in Figure 4c, and the fitting results are listed in Table 2. In the table, R1 indicates IR resistance and (20) O’Hayre, R.; Cha, S.-W.; Colella, W.; Prinz, F. B. Fuel Cell Fundamentals; John Wiley & Sons: Hoboken, New Jersey, 2006; pp 207-208.
DOI: 10.1021/la101355d
16531
Article
Jeon et al. Table 2. Fitting results of EIS data for the MEAs consisted of two different anode catalysts: 1) the pure PtRu and 2) 10 wt % Fe2O3 þ 90 wt % PtRu catalyst parameters
pure PtRu
10 wt % Fe2O3 þ 90 wt % PtRu
R1 (Ω) R2 (Ω) CPE1-T (F) CPE1-P R3 (Ω) L1 (H)
0.208 0.325 0.923 0.881 0.124 0.112
0.223 0.315 0.853 0.885 0.115 0.0949
Figure 5. Pt 4f7/2 XPS results of the MEAs which consists of (1) pure PtRu and (2) 10 wt % Fe2O3 mixed PtRu catalysts as their anode electrode.
Figure 4. (a) I-V test results of the two MEAs of which the anode was made with pure PtRu and 10 wt % Fe2O3 mixed PtRu catalyst. (b) EIS measurement results of the MEAs (dots) and fitting results (solid lines). (c) Equivalent circuit used for fitting of EIS measurement results.
R2 means reaction resistance for the MOR. R3 indicates CO electro-oxidation reaction resistance. CPE1 is a capacitive component including a diffusion effect which is composed of CPE-T and CPE-P. CPE-T is a capacitance and CPE-P is a nonhomogeneous constant. L1 is an inductive element which implies phase delay by slow relaxation of CO adsorbed on Pt.21 In the EIS results, IR resistance increased from 0.208 to 0.223 Ω when the Fe2O3 nanoclusters were mixed, which is in good agreement with the I-V measurement results. This 7% increase in the IR resistance presumably came from the semiconductor behavior of the Fe2O3 nanoclusters. In the case of the reaction resistances, R2 and R3, a decrease of the resistances was observed when the Fe2O3 nanoclusters were mixed. Only 3% decrease was observed in R2, while 8% decrease was observed in R3, indicating that the higher (21) M€uller, J. T.; Urban, P. M.; H€olderich, W. F. J. Power Sources 1999, 84, 157–160.
16532 DOI: 10.1021/la101355d
performance of the MEA-M was originated from the improved CO oxidation activity. This result is in good agreement with the CO stripping results. The high performance of the MEA-M means that the PtRu-Fe2O3 nanocomposite catalyst is a promising anode electrode to achieve high performance with reduced amounts of expensive PtRu by a conventional MEA fabrication technique. Interaction between Pt and the Fe2O3 nanoclusters was investigated by using X-ray photoelectron spectroscopy (XPS), and the results are shown in Figure 5. It should be noted that the d-band center is considered as a key parameter which determines Pt-CO bonding strength.22 However, the XPS peak position of Pt can be employed as an indirect indicator of Pt-CO bonding strength.23,24 The binding energy of Pt 4f7/2 was moved to higher energies in the 10wt % Fe2O3 mixed PtRu catalyst, which is the same phenomenon observed in the PtFe23 and the PtRuFe13 catalysts. The shift of binding energy of Pt to higher energy indicates lowered electron density of Pt which results in weak Pt-CO bonding.24 Here, we could conclude that the electronic state of Pt could be modified by mixing with the Fe2O3 nanoclusters. This observation is supported by recent studies on Pt skin or core-shell type catalyst which showed that the Pt skin layer was affected by the underlying Pt-metal alloy core.24-27 Zhai et al.28 (22) Hammer, B.; Morikawa, Y.; Nørskov, J. K. Phys. Rev. Lett. 1996, 76, 2141– 2144. (23) Watanabe, M.; Igarashi, H.; Fujino, T. Electrochemistry 1999, 67, 1194– 1196. (24) Watanabe, M.; Zhu, Y.; Uchida, H. J. Phys. Chem. B 2000, 104, 1762–1768. (25) Toda, T.; Igarashi, H.; Uchida, H.; Watanabe, M. J. Electrochem. Soc. 1999, 146, 3750–3756. (26) Adzic, R. R.; Zhang, J.; Sasaki, K.; Vukmirovic, M. B.; Shao, M.; Wang, J. X.; Nilekar, A. U.; Mavrikakis, M.; Valerio, J. A.; Uribe, F. Top. Catal. 2007, 46, 249–262. (27) Stamenkovic, V. R.; Fowler, B.; Mun, B. S.; Wang, G.; Ross, P. N.; Lucas, C. A.; Markovic, N. M. Science 2007, 315, 493–497. (28) Zhai, J.; Huang, M.; Dong, S. Electroanalysis 2007, 19, 506–509.
Langmuir 2010, 26(21), 16529–16533
Jeon et al.
prepared Au/Pt core-shell catalysts which have various numbers of Pt layers, and showed that the performance of the core-shell catalysts resembles that of pure Pt with an increasing number of Pt layers. Dissolution of Fe2O3 should be mentioned, because the dissolution of Fe2O3 can cause a problem in the long term stability of DMFCs by reducing the activity of the anode catalyst.29 However, Virtanen et al.30 reported that the dissolution rate of Fe2O3 is very low in a sulfuric acid environment. It is still not clear how fast the Fe2O3 nanoclusters will be dissolved under a DMFC anode environment, and more study is required. (29) Chen, W.; Xin, Q.; Sun, G.; Yang, S.; Zhou, Z.; Mao, Q.; Sun, P. Electrochim. Acta 2007, 52, 7115–7120. (30) Virtanen, S.; Schmuki, P.; Davenport, A. J.; Vitus, C. M. J. Electrochem. Soc. 1997, 144, 198–204.
Langmuir 2010, 26(21), 16529–16533
Article
4. Conclusion The MOR activity of the commercial PtRu black catalyst was dramatically improved by simply mixing with Fe2O3 nanoclusters. In the 10 wt % Fe2O3 mixed catalyst, the mass and specific activities increased by 88 and 100%, respectively, compared with those of the pure PtRu catalyst. The improved MOR activity was also observed in the single cell tests. The reason of higher performance was verified by XPS results, which showed that the electronic state of Pt was modified by the mixing with the Fe2O3 nanoclusters. This nanocomposite catalyst provides an easy way to fabricate high performance MEAs with a reduced amount of the expensive noble metal catalyst. Acknowledgment. This work was supported by a National Research Foundation of Korea (NRF) grant funded by the Korea government (MEST) (No. 2009-0092783).
DOI: 10.1021/la101355d
16533
