CNT Supported Pt and PtRu Nanocatalysts for Direct Methanol

Apr 29, 2009 - Pt/MnO2/carbon nanotube (CNT) and PtRu/MnO2/CNT nanocomposites were synthesized by successively loading hydrous MnO2 and Pt (or ...
2 downloads 0 Views 4MB Size
pubs.acs.org/Langmuir © 2009 American Chemical Society

MnO2/CNT Supported Pt and PtRu Nanocatalysts for Direct Methanol Fuel Cells Chunmei Zhou, Hongjuan Wang, Feng Peng,* Jiahua Liang, Hao Yu, and Jian Yang The School of Chemistry and Chemical Engineering, South China University of Technology, Guangzhou 510640, China Received January 20, 2009. Revised Manuscript Received March 21, 2009 Pt/MnO2/carbon nanotube (CNT) and PtRu/MnO2/CNT nanocomposites were synthesized by successively loading hydrous MnO2 and Pt (or PtRu alloy) nanoparticles on CNTs and were used as anodic catalysts for direct methanol fuel cells (DMFCs). The existence of MnO2 on the surface of CNTs effectively increases the proton conductivity of the catalyst, which then could remarkably improve the performance of the catalyst in methanol electro-oxidation. As a result, Pt/MnO2/CNTs show higher electrochemical active surface area and better methanol electro-oxidation activity, compared with Pt/CNTs. As PtRu alloy nanoparticles were deposited on the surface of MnO2/CNTs instead of Pt, the PtRu/MnO2/CNT catalyst shows not only excellent electro-oxidation activity to methanol with forward anodic peak current density of 901 A/gPt but also good CO oxidation ability with lower preadsorbed CO oxidation onset potential (0.33 V vs Ag/AgCl) and peak potential (0.49 V vs Ag/AgCl) at room temperature.

1. Introduction

*Corresponding author. Fax: +86 20 87114916 E-mail: cefpeng@ scut.edu.cn.

with smaller amounts of noble metal and higher electrocatalytic performance. Recently, TiO210 and SnO211,12 have been found to have the similar enhancing effect as hydrous RuO2. And in MartinezHuerta’s work,13 MoOx was found to have the role of increasing the CO tolerance and reducing the amount of precious metals Pt and Ru when it was used in PtRu/MoOx catalyst. With many appealing electrochemical properties, such as excellent proton conductivity, various structures and oxidation states in electrolyte, manganese dioxide (MnO2) has become widely used to store and convert charges in electrochemical supercapacitors.14-16 Recently, MnO2 was also used in a DMFC catalyst to improve the methanol oxidation activity by some researchers. A MnOx/Ru catalyst was prepared to enhance the electrochemical activities for DMFCs.17 A Pd/β-MnO2 composite nanotube was proved to have an excellent electrochemical performance for methanol oxidation in an alkaline solution.18 A composite electrode composed of Pt nanoparticles immobilized on MnO2 nanowire was also found to show 110 mV decreased overpotential and 2.1-fold enhanced voltammetric forward peak current.19 However, the single MnO2 has poor electric conductivity, and the preparation of MnO2 nanotubes or nanowires is a relatively complex and costly course. Then the MnO2/ carbon nanotube (CNT) composite, with the synergistic effect, has been employed to improve the poor electric conduction and deficient charge transfer channel of single MnO2.20,21

(1) Liu, H. S.; Song, C. J.; Zhang, L.; Zhang, J. J.; Wang, H. J.; Wilkinson, D. P. J. Power Sources 2006, 155, 95–110. (2) Rolison, D. R.; Hagans, P. L.; Swider, K. E.; Long, J. W. Langmuir 1999, 15, 774–779. (3) Chen, Z. G.; Qiu, X. P.; Lu, B.; Zhang, S. C.; Zhu, W. T.; Chen, L. Q. Electrochem. Commun. 2005, 7, 593–596. (4) Scheiba, F.; Scholz, M.; Cao, L.; Schafranek, R.; Roth, C.; Cremers, C.; Qiu, X.; Stimming, U.; Fuess, H. Fuel Cells 2006, 6, 439–446. (5) Cao, L.; Scheiba, F.; Roth, C.; Schweiger, F.; Cremers, C.; Stimming, U.; Fuess, H.; Chen, L. Q.; Zhu, W. T.; Qiu, X. P. Angew. Chem., Int. Ed. 2006, 45, 5315–5319. (6) Villullas, H. M.; Mattos-Costa, F. I.; Bulhoes, L. O. S. J. Phys. Chem. B 2004, 108, 12898–12903. (7) Profeti, L. P. R.; Simoes, F. C.; Olivi, P.; Kokoh, K. B.; Coutanceau, C.; Leger, J. M.; Lamy, C. J. Power Sources 2006, 158, 1195–1201. (8) Jeon, M. K.; Won, J. Y.; Woo, S. I. Electrochem. Solid-State Lett. 2007, 10, B23–B25. (9) Huang, S. Y.; Chang, C. M.; Wang, K. W.; Yeh, C. T. ChemPhysChem 2007, 8, 1774–1777.

(10) Song, H. Q.; Qiu, X. P.; Guo, D. J.; Li, F. S. J. Power Sources 2008, 178, 97–102. (11) Ke, K.; Waki, K. J. Electrochem. Soc. 2007, 154, A207–a212. (12) Mecheri, B.; D’Epifanio, A.; Traversa, E.; Licoccia, S. J. Power Sources 2008, 178, 554–560. (13) Martinez-Huerta, M. V.; Rodriguez, J. L.; Tsiouvaras, N.; Pena, M. A.; Fierro, J. L. G.; Pastor, E. Chem. Mater. 2008, 20, 4249–4259. (14) Ghaemi, M.; Ataherian, F.; Zolfaghari, A.; Jafari, S. M. Electrochim. Acta 2008, 53, 4607–4614. (15) Devaraj, S.; Munichandraiah, N. J. Phys. Chem. C 2008, 112, 4406–4417. (16) Liu, X. M.; Zhang, X. G. J. Inorg. Mater. 2003, 18, 1022–1026. (17) Rebello, J. S.; Samant, P. V.; Figueiredo, J. L.; Fernandes, J. B. J. Power Sources 2006, 153, 36–40. (18) Xu, M. W.; Gao, G. Y.; Zhou, W. J.; Zhang, K. F.; Li, H. L. J. Power Sources 2008, 175, 217–225. (19) Zhao, G. Y.; Li, H. L. Appl. Surf. Sci. 2008, 254, 3232–3235. (20) Xie, X. F.; Gao, L. Carbon 2007, 45, 2365–2373. (21) Ma, S. B.; Nam, K. W.; Yoon, W. S.; Yang, X. Q.; Ahn, K. Y.; Oh, K. H.; Kim, K. B. J. Power Sources 2008, 178, 483–489.

Direct methanol fuel cells (DMFCs) as a “clean” power source for portable devices have attracted more and more attention due to their advantages of relatively compact system design and higher energy densities as compared with existing technologies.1 Pt is the most active anodic electrocatalyst for methanol oxidation. However, the effective utilization factor of precious metal Pt and the inhibiting effects of intermediate species, mainly as CO, are two vital problems that exigently need to be solved. One or more other metals combining with Pt as cocatalysts have been proposed for the oxidation of methanol. Among them, bimetallic PtRu catalyst is considered to be the most promising DMFC anodic catalyst.2 In the work of Qiu and his co-workers, hydrous ruthenium oxide had been used as a novel support material for fuel cell electrocatalyst to enhance the intrinsic proton conductivity because of the high electron and proton conductivity.3-5 And many other researches around RuO2 in the field of DMFCs are being developed quickly.6-9 However, Ru is also a noble metal and not very stable in acidic media. So more durable and lower costing anode catalysts are preferable for DMFCs. Many researches in this field involve development of new materials

Langmuir 2009, 25(13), 7711–7717

Published on Web 04/29/2009

DOI: 10.1021/la900250w

7711

Article

In this work, novel composites of Pt/MnO2/CNTs and PtRu/ MnO2/CNTs for DMFC anodic catalyst were fabricated for the first time. Their methanol and CO electro-oxidation activity were investigated. The role of MnO2 in the electrocatalytic oxidation of methanol is discussed.

2. Experimental Section 2.1. Catalyst Preparation and Characterization. CNTs were treated by a well-known acid oxidation method to introduce oxygenous groups that increase the surface activity of CNTs. For Pt/CNTs, the functionalized CNTs were added to ethylene glycol under constant stirring for 10 min. Then, the appropriate amounts of H2PtCl6 and KOH were added with constant stirring. The slurry was refluxed at 140 °C for 2 h to ensure a complete reduction of Pt(IV). The solid product was filtered and then rinsed repeatedly with deionized water. And the solid product was dried in vacuum at 70 °C for 10 h. The PtRu/CNTs catalyst was also prepared with a similar method. Functionalized CNTs were added to ethylene glycol with constant stirring for 10 min. Then RuCl3, H2PtCl6, and KOH were added with constant stirring. The resultant slurry was refluxed at 140 °C for 2 h to form PtRu alloy on CNTs. As to the deposition of hydrous manganese dioxide on CNTs, 400 mg of functionalized CNTs and 20 mL of KMnO4 (0.05 M) aqueous solution were dispersed in 100 mL of deionized water with 15 min ultrasonic, and then 20 mL of citric acid (0.05 M) was added dropwise at room temperature under vigorous stirring, until the purple of the suspension was found faded completely. Then, the suspension was refluxed at 80 °C for 7 h. The obtained catalyst is denoted as MnO2/CNTs. Pt or PtRu alloy was deposited on MnO2/CNTs with the aforementioned method to obtain Pt/MnO2/CNTs and PtRu/MnO2/CNTs. In order to prove the role of hydrous MnO2 in PtRu/MnO2/CNTs, we treated MnO2/ CNTs under 10% H2/90% Ar atmosphere at 600 °C (removing the combining structural water and reducing MnO2) to prepare MnO2-x/CNTs. Using the MnO2-x/CNTs as support, PtRu/ MnO2-x/CNTs were prepared according to the same method of preparation for PtRu/MnO2/CNTs. Atomic adsorption spectroscopy (Shimadzu, AA-6800) was used to analyze the content of Mn, Pt, and Ru in the filtrate, which was collected during catalyst preparation to determine the exact loading amount of Pt, Ru, and Mn. The mass contents of samples are as follows: MnO2/CNTs with 10.5% Mn; Pt/MnO2/CNTs with 15% Pt and 6.8% Mn; PtRu/MnO2/CNTs with 15% Pt, 8.0% Ru, and 6.8% Mn; and PtRu/CNTs with 15% Pt and 9.5% Ru. The morphology of the catalyst was characterized by transmission electron microscopy (TEM, JEOL, JEM2010) with the instrument operating at 200 kV. Structure characterization of the catalyst was carried out by X-ray diffraction (XRD, D/maxIIIA spectrometer), thermal gravimetric analysis (TGA, Netzsch STA449C), and TEM (JEOL JEM-2010). The chemical valence of Pt, Ru, and Mn in catalyst prepared was analyzed by X-ray photoelectron spectroscopy (XPS, VG Scientific, ESCALAB MK II) using Al KR radiation (1486.71 eV). Spectra correction was conducted by using a C 1s binding energy of 284.6 eV. The Brunauer-Emmett-Teller (BET) specific surface areas were measured by N2 adsorption at 77 K using a TriStar 3000 instrument (Micromeritics) after the sample was degassed in a vacuum at 120 °C overnight. The temperature programmed reduction (TPR) profiles of the catalysts were measured in an AutoChem 2920 (Micromeritics) chemisorption analyzer. 2.2. Electrochemical Performance Measurement. The fabrication of a glassy carbon (GC) electrode modified with CNT-based electrocatalysts was described in our previous paper in detail.22 Electrochemical measurements were carried out at room temperature in a three-electrode cell connected to (22) Wang, H. J.; Yu, H.; Peng, F.; Lv, P. Electrochem. Commun. 2006, 8, 499–504.

7712 DOI: 10.1021/la900250w

Zhou et al. an electrochemical analyzer (Eco Chemie B. V., Autolab PGSTAT30). A glassy carbon coated with 0.01 mg of catalyst was used as the working electrode. An Ag/AgCl electrode and a Pt electrode were used as the reference and counter electrodes, respectively. The cyclic voltammogram curves were not recorded in 1 M HClO4 with a scan rate of 0.1 V/s until a stable response was obtained. Methanol oxidation activity was measured in 1 M CH3OH with 1 M HClO4 as electrolyte at 25 °C with the same scan rate of 0.1 V/s. For the electrochemical active surface area (EAS) measurement, 10 consecutive cyclic voltammogram sweeps were performed in 1 M HClO4 solution (sweep rate 100 mV/s) to obtain a stable curve. The average Coulombic charge of hydrogen adsorption and hydrogen desorption was used to calculate the electrochemical active platinum surface area of the electrodes. For CO stripping measurements, pure CO was first bubbled into the electrolyte for CO adsorption on the electrode, and then nitrogen was used to purge CO reversibly adsorbed on the surface. Three CVs were recorded from -220 to 900 mV versus Ag/AgCl. The first anodic sweep was performed to electro-oxidize the irreversibly adsorbed CO, and the subsequent sweeps were carried out in order to verify the completeness of the CO oxidation.23

3. Results and Discussion The morphology of MnO2/CNTs was observed by TEM, as shown in Figure 1. Most of MnO2 particles were floc and uniformly covered the surface of the CNTs, as shown in Figure 1a. Only some crystalline MnO2 nanoparticles were found from high-resolution (HR)-TEM micrographs, and the distance between two adjacent lattice planes is approximately 0.245 nm (as in Figure 1b), which is agreement with the XRD analysis (inset in Figure 4a) and corresponds to R-MnO2(400) (JCPDS #44-0141). The XPS spectra of Mn 2p in MnO2/CNTs (Figure 2) show two main peaks located at 642.2 and 653.5 eV, which could be assigned to Mn 2p3/2 and Mn 2p1/2 of Mn(IV) in MnO2 respectively.20 Meanwhile, Mn(III) and Mn(V) species are also present in the XPS spectra. It is reported that the relative intensity of the Mn 2p core level components is extremely sensitive to small changes in structures of reduced Mn 3d hybridization. The XPS intensity of Mn 2p is influenced by 3d electrons due to “final state effect”.24 Thus we cannot calculate the composition of Mn 3+, 4+, and 5+ simply by assuming their 2p signal intensity, as their response factors of each state are different. The content of structural water in MnO2 can be determined by thermal gravimetric analysis. The structural water of MnO2 3 xH2O mainly loses in the temperature range of 200-600 °C. To exclude the influence of functional groups on the surface of CNTs, the weight loss of water in MnO2 3 xH2O was obtained by subtracting the weight loss of CNTs from the measured weight loss of MnO2 3 xH2O/CNTs in the temperature range of 200-600 °C (in Figure 3). The x value calculated by TGA is approximately 0.5. For the XRD patterns of MnO2/CNTs (shown in Figure 4a), only a few broad peaks were found attributed to MnO2. The peaks around 2θ = 18.1° and 37.0° are similar to those in the literature,20,25 and they are attributed to MnO2(200) and MnO2(400) (JCPDS #44-0141), respectively. The XRD results agree well with the TEM results. And after the subsequent loading of Pt or PtRu to the MnO2/ CNTs, as shown in Figure 4b and c, the MnO2(400) peak at 2θ = 37.0° is covered by that of Pt(111); however, the MnO2 (200) peak at 2θ = 18.1° can still be observed. It means that the properties of the Mn oxide in the samples of MnO2/CNTs, Pt/MnO2/CNTs (23) Pozio, A.; De Francesco, M.; Cemmi, A.; Cardellini, F.; Giorgi, L. J. Power Sources 2002, 105, 13–19. (24) Schieffer, P.; Krembel, C.; Hanf, M. C.; Gewinner, G. J. Electron Spectrosc. Relat. Phenom. 1999, 104, 127–134. (25) Xu, C. J.; Li, B. H.; Du, H. D.; Kang, F. Y.; Zeng, Y. Q. J. Power Sources 2008, 180, 664–670.

Langmuir 2009, 25(13), 7711–7717

Zhou et al.

Figure 1. TEM (a) and HR-TEM (b) images of MnO2/CNTs.

Figure 2. XPS spectra of Mn 2p for MnO2/CNTs.

Figure 3. Thermal gravimetric analyses of CNTs and MnO2/ CNTs.

Figure 4. XRD patterns of MnO2/CNTs (a), Pt/MnO2/CNTs (b), and PtRu/MnO2/CNTs (c). Langmuir 2009, 25(13), 7711–7717

Article

and PtRu/MnO2/CNTs are similar. As for PtRu/MnO2/CNTs (in Figure 3c), only diffractions from face-centered-cubic (fcc) Pt metal are observed and those from hexagonal-closed-packed (hcp) Ru metal are absent. This is a typical characteristic of PtRu alloy containing g50 atom % Pt.26 Compared Pt/MnO2/CNTs with PtRu/MnO2/CNTs, the position of the Pt(111) peak remains unchanged at ∼39.7°, which implies that the crystalline phases for both samples are Pt metal. The peaks of Pt fcc lattices for PtRu/ MnO2/CNTs were found to be much broader than those of Pt/ MnO2/CNTs, which may caused by the alloying of Pt and Ru.27 The electrochemical property of the MnO2/CNTs in the electrolytes was characterized by cyclic voltammogram (CV) method, as shown in Figure 5. The background current of MnO2/CNTs in 1 M HClO4 solution is 3 times larger than that of CNTs. It indicates that the capacity and proton conductivity of MnO2/CNTs are much higher than those of CNTs,2 which is in agreement with the references.20,21 And the addition of CH3OH to HClO4 electrolyte does not change the CV curves of MnO2/ CNTs. Table 1 shows that the transported Coulombic charge (Q) of MnO2/CNTs between -0.1 and 0.8 V retains well in 1 M CH3OH + 1 M HClO4 solution, with about 94% of its original Q reserved after the 1000 cycles CV scan, which indicates that the proton conductivity stability of MnO2/CNTs is very good in the electrolyte of 1 M CH3OH + 1 M HClO4 solution. The BET specific surface areas of the functionalized CNTs (105.09 m2/g) and the MnO2/CNTs (101.23 m2/g) were found to be similar. It indicates that MnO2 enhances the intrinsic proton conductivity of the catalysts, rather than increases the specific surface area. It is well-known that the electrocatalytic oxidation of methanol is essentially a process of deprotonation and electron formation; thus, the proton conductivity of the catalyst is very important for the electrocatalytic oxidation of methanol, which will influence the electrocatalytic activity of the catalyst. With different proton conductivity, the methanol electrocatalytic oxidation activity of Pt/CNTs and Pt/MnO2/CNTs should be different. The cyclic voltammogram curves of methanol electrocatalytic oxidation for Pt/CNTs and Pt/MnO2/CNTs are shown in Figure 6. The methanol oxidation activity can be denoted by the forward anodic peak current (If) in the cyclic voltammogram of methanol oxidation. The responding current of methanol oxidation by Pt/MnO2/CNTs is 2.3 times as large as that of Pt/ CNTs, which suggests that Pt/MnO2/CNT catalyst has higher methanol electrocatalytic activity. The promoting effect may further attribute to the improvement of its CO oxidation activity (capacity) or its utilization ratio of Pt. The CO oxidation ability of the Pt/CNTs and Pt/MnO2/CNTs catalysts were evaluated by conducting a CO-stripping experiment, as shown in Figure 7. It can be seen that the onset potential of CO electro-oxidation with Pt/MnO2/CNTs (0.53 V) is similar to that of Pt/CNTs (0.56 V), and both of their peak potentials are about 0.68 V, which demonstrates that the addition of MnO2 does not enhance the CO oxidation ability (the onset oxidation potential). The utilization ratio of Pt is closely interrelated with its dispersion situation and the electrochemical active surface area (EAS). The EAS of Pt/CNT and Pt/MnO2/CNT catalysts can be measured based on H2 adsorption and desorption and preadsorbed CO electro-oxidation, as shown in Figure 7. The CV curves of the second cycle of preadsorbed CO oxidation (not shown for concision) are almost overlapped with the stable cycles (26) Chakraborty, D.; Chorkendoff, I.; Johannessen, T. J. Power Sources 2007, 173, 110–120. (27) Gu, Y. J.; Wong, W. T. Langmuir 2006, 22, 11447–11452.

DOI: 10.1021/la900250w

7713

Article

Zhou et al.

Figure 5. Cyclic voltammograms in 1 M HClO4 solution for CNTs (a) and MnO2/CNTs (b); 5th cycle (c) and 1000th cycle (d) for MnO2/CNTs in 1 M CH3OH + 1 M HClO4 solution. Table 1. Electrochemical Properties of MnO2/CNTs in the Electrolytes samples CNTs

SBET (m2/g)a

electrolyte

105.09

cycle no.

Figure 8. Cyclic voltammograms in 1 M HClO4 solution for PtRu/MnO2/CNTs (a) and PtRu/CNTs (b). Dotted lines represent the stable cycles before CO preadsorption, and the solid lines represent the first cycle of preadsorded CO oxidation.

Q (mC)b

1 M HClO4 5th 5th 1 M HClO4 101.23 1 M CH3OH + 5th MnO2/CNTs (10.5 wt % Mn) 1 M HClO4 1 M CH3OH + 1000th 1 M HClO4 a BET specific surface areas. b Transported Coulombic between -0.1 and 0.8 V in Figure 5.

0.37 1.12 1.13 1.06 charge

Figure 9. Cyclic voltammograms of PtRu/MnO2/CNTs (a) and PtRu/CNTs (b) in 1 M CH3OH with 1 M HClO4.

Figure 6. Cyclic voltammograms of Pt/MnO2/CNTs (a) and Pt/ CNTs (b) in 1 M CH3OH with 1 M HClO4.

Figure 7. Cyclic voltammograms in 1 M HClO4 solution for Pt/ MnO2/CNTs (a) and Pt/CNTs (b). Dotted lines represent the stable cycles before CO preadsorption, and the solid lines represent the first cycle of preadsorded CO oxidation. The filled areas in (a) represent the exchanged charge of H2 adsorption (QH2) and desorption (QH1) and CO oxidation (QCO) on Pt sites. 7714 DOI: 10.1021/la900250w

before CO preadsorption (dotted lines in Figure 7), which suggests the CO can be completely oxidized in the first cycle of preadsorbed CO oxidation. In the stable cycles before CO preadsorption, the peaks representing H2 adsorption and desorption are clear in the potential range of -0.23 to +0.2 V. These peaks disappear in the first cycle of preadsorbed CO oxidation and recover in the second cycle after the CO is completely removed by oxidation. It indicates that the H2 and CO adsorption are in the same Pt active sites. The electrochemical active surface area based on the area of the H2 adsoprtion and desorption peak (EAS-H) can be calculated according to the equation of EAS-H = QH/210 (μC/cm2)/w.23 The results are 28 m2/gPt for Pt/CNTs and 49 m2/gPt for Pt/MnO2/CNTs. In the same way, the EAS bases on the area of the CO oxidation peak (EAS-CO) according to the equation of EAS-CO = QCO/484 (μC/cm2)/w23 are 21 and 40 m2/gPt for Pt/CNTs and Pt/MnO2/CNTs, respectively. The calculated EAS-H and EAS-CO for the two catalysts are comparable, and the Pt/MnO2/CNTs in both cases exhibit much higher EAS in comparison with Pt/CNTs. Now, it can be concluded that the MnO2 in the Pt/MnO2/ CNTs catalyst can increase the EAS of Pt but cannot enhance its CO oxidation ability. For an excellent methanol anodic catalyst, it should have not only large EAS but also good CO oxidation ability. Then, on the basis of the Pt/MnO2/ CNTs, we should further improve its CO oxidation ability. For promoting the CO oxidation ability, the introduction of Ru has been proved to be successful as mentioned in the Introduction. So we also introduce Ru into Pt/MnO2/CNTs to obtain PtRu/ MnO2/CNTs catalyst. The EDS (energy dispersive X-ray Langmuir 2009, 25(13), 7711–7717

Zhou et al.

Article

Figure 10. TEM images and particle size distribution histograms of Pt/MnO2/CNTs (a, b, and e) and PtRu/MnO2/CNTs (c, d, and f).

spectroscopy) result proved the Pt, Ru, and Mn composition existed in the PtRu/MnO2/CNTs (see Supporting Information Figure S1). For this catalyst, PtRu alloy is effective for CO oxidation and the MnO2 can increase the proton conductivity and EAS. So PtRu/MnO2/CNTs are expected to have both higher methanol electro-oxidation activity and better CO oxidation ability. From Figure 8, we can see that the PtRu/CNTs have a lower onset potential of CO electro-oxidation (0.33 V) than Pt/CNTs (0.56 V). The introduction of Ru to Pt/MnO2/CNTs shows the Langmuir 2009, 25(13), 7711–7717

same effect; the onset potential of CO electro-oxidation for PtRu/MnO2/CNTs also recedes to 0.33 V, and the peak potential recedes to 0.49 V. Hence, the CO oxidation ability of PtRu/ MnO2/CNTs is much better than that of Pt/MnO2/CNTs. In Figure 8, we can also see that the QH and QCO values of PtRu/MnO2/CNTs are also higher than those of PtRu/ CNTs. It proves again that the MnO2 can improve the EAS. And the forward anodic peak current of methanol oxidation for PtRu/MnO2/CNTs is 1.5 times higher than that of PtRu/CNTs (shown in Figure 9) and also higher than that of Pt/MnO2/CNTs. DOI: 10.1021/la900250w

7715

Article

Zhou et al. Table 2. Performances of Electrocatalytic Oxidation and Utilization of Pt for Different Catalysts EASact (m2/g Pt)d

ECO (V)a samples

If (A/g Pt)

onset

peak

d (nm)b

Sgeom (m2/g Pt)c

EASact-H

Sact-CO

ΔSact/Sgeome

Pt/CNTs 187 0.56 0.68 3.6 78 28 21 0.31 431 0.52 0.68 3.5 80 49 40 0.57 Pt/MnO2/CNTs PtRu/CNTs 362 0.33 0.65 3.2 87 81 63 0.83 906 0.33 0.49 2.5 112 112 103 0.95 PtRu/MnO2/CNTs 0 0 0 MnO2/CNTs 357 0.40 0.57 3.3 86 78 65 0.83 PtRu/MnOx/CNTs a Onset potential and peak potential (vs Ag/AgCl) of electro-oxidation of preadsorbed CO. b Average diameter of Pt nanoparticles based on TEM. c Geometrical surface area (Sgeom) of Pt based on TEM. d Electrochemical active surface, Sact-H and Sact-CO, obtained by Coulombic charge of hydrogen absorption (QH) and preadsorbed CO oxidation (QCO), respectively. e Electrochemical active surface (ΔSact) calculated with ΔSact = (Sact-H + Sact-CO)/2.

So, the PtRu/MnO2/CNTs, as expected, are an excellent methanol oxidation anodic catalyst with not only a larger EAS but also higher CO oxidation ability. However, the EAS alone cannot perfectly and exactly represent the utilization ratio of Pt that is also influenced by the dispersion situation of Pt. So, the dispersion of Pt should also be considered in the analysis. From the XRD pattern of Figure 4, the widths of Pt(111) peaks for Pt/MnO2/CNTs and PtRu/MnO2/CNTs are different. The average sizes of the Pt nanoparticles estimated by Scherrer’s equation are 3.5 and 2.5 nm for Pt/MnO2/CNTs and PtRu/MnO2/CNTs, respectively, which indicates that the Pt dispersion situations are different in the two electrocatalysts. Different dispersion situations provide different geometrical specific surface areas (Sgeom) of Pt nanoparticles. A comparison of EAS with Sgeom can tell how many surface atoms are electrochemically active, that is, contributing to the electrochemical reactions. Thus, the ratio of EAS/Sgeom is considered here as an effective evaluation of Pt utilization ratio for a catalyst. The larger the ratio is, the higher electroactivity the catalyst has. For a catalyst with an activity of 1.0, it means that every surface atom on the Pt nanoparticles is electrochemically active.28 TEM was used to observe the morphology and dispersion situation of the catalysts and obtain the information of the Sgeom of Pt. As shown in Figure 10a-c, the metal nanopaticles are uniformly dispersed on the surfaces of the CNTs for both Pt/MnO2/CNTs and PtRu/MnO2/CNTs. As for the Pt/MnO2/ CNTs, according to the high-resolution TEM image of nanoparticles (Figure 10b inset), the distances between two adjacent lattice planes for most nanoparticles are approximately 0.228 nm, which are characteristic of Pt(111) and are consistent with the results obtained from X-ray diffraction (in Figure 4b). For the PtRu/ MnO2/CNTs, many crystal spacings of 0.220-0.226 nm assigned to the Pt(111) plane (in Figure 10d) were observed. The crystal spacings shrank distinctly as compared with the 0.228 nm spacing of the pure Pt metal. It suggests that some Ru has alloyed with Pt during the catalyst preparation process.29,30 Some nanoparticles with the distance of 0.245 nm between two adjacent lattice planes can also be seen, which is characteristic of MnO2(400). The particle size distribution histograms of these two catalysts are determined on the basis of analysis of more than 500 particles, and only some of them are included in Figure 3 for concision. The average particle sizes determined by TEM are about 3.5 nm for Pt/MnO2/CNTs and 2.5 for PtRu/MnO2/CNTs. And as to our previous work,31 the average particle sizes determined (28) Xing, Y. C. J. Phys. Chem. B 2004, 108, 19255–19259. (29) Guo, J. S.; Sun, G.; Sun, S. G.; Yan, S. Y.; Yang, W. Q.; Qi, J.; Yan, Y. S.; Xin, Q. J. Power Sources 2007, 168, 299–306. (30) Guo, J. S.; Sun, G. Q.; Wu, Z. M.; Sun, S. G.; Yan, S. Y.; Cao, L.; Yan, Y. S.; Su, D. S.; Xin, Q. J. Power Sources 2007, 172, 666–675. (31) Peng, F.; Zhou, C. M.; Wang, H. J.; Yu, H.; Liang, J. H.; Yang, J. Catal. Commun. 2009, 10, 533–537.

7716 DOI: 10.1021/la900250w

by TEM for Pt/CNTs and PtRu/CNTs are 3.6 and 3.2 nm, respectively. The valences of Pt and Ru for the PtRu/CNTs and PtRu/ MnO2/CNTs were also determined by XPS (see Supporting Information Figure S2). The results showed the PtRu/CNTs and PtRu/MnO2/CNTs were mainly composed of metallic Pt, Ru, and RuO2. It was found that the Ru 3d5/2 peaks of PtRu/ MnO2/CNTs for the metallic Ru (281.0 eV) and RuO2 (282.2 eV) were 0.5 eV higher than that of PtRu/CNTs for the metallic Ru (280.5 eV) and RuO2 (281.7 eV), which indicated the strong interaction of Ru and MnO2. The interactions of Pt, RuO2, and MnO2 were also demonstrated by the TPR (temperature programmed reduction) results of catalysts (see Supporting Information Figure S3). The geometrical specific surface areas of Pt nanoparticles can be calculated by Sgeom = 6/(Fd),28 where F is the density of Pt (21.09 g/cm3) and d is the mean diameter of the Pt nanoparticles in the catalyst. For Pt/CNTs, the EAS is much lower than the Sgeom, and the utilization ratio of Pt (EAS/Sgeom) is only about 0.31. When introducing MnO2 (for Pt/MnO2/CNTs), the utilization ratio of Pt increases to near 0.57. The same promotion is also found in PtRu/CNTs (0.83) and PtRu/MnO2/CNTs (0.95). So, it is clear that the role of the MnO2 is to increase the EAS and, as a result, to increase utilization ratio of Pt. From Table 2, we can clearly see that MnO2 in Pt/MnO2/CNTs cannot influence the Pt particle size but influence the EAS. Generally speaking, the EAS is closely related with the Pt particle size. The smaller the particle size of Pt is, the larger the EAS is. However, there is difference in our reaction. So, why are the EAS/ Sgeom values of Pt/MnO2/CNTs and PtRu/MnO2/CNTs larger than those of the corresponding Pt/CNTs and PtRu/CNTs? Because the similar problem has been encountered by many researchers,28 three possible reasons are proposed to address the above problem. First, the hydrous MnO2, as well as hydrous RuO2, provides fast and barrier-free channels for electron and proton transfer, which is favorable to hydrogen adsorption or desorption and CO electro-oxidation. Second, proton and CO probably spill over from the Pt site to MnO2 during the cyclic voltammogram measurements (see Supporting Information Table 1S). A similar effect was also observed by Tseung and Chen in Pt/WO3 systems where the protons produced on the platinum surface during electro-oxidation of methanol transferred to WO3 and then generated hydrogen tungsten bronze species (HxWO3).32 Third, it has been reported that methanol is able to diffuse into the hydrous regions of RuO2.4 Because hydrous MnO2 has similar aqueous regions as RuO2, protons and CO as smaller species might also diffuse into the hydrous regions of MnO2. All these three possible explanations are responsible for the larger EAS with the same particle size, and they are also (32) Tseung, A. C. C.; Chen, K. Y. Catal. Today 1997, 38, 439–443.

Langmuir 2009, 25(13), 7711–7717

Zhou et al.

responsible for the higher methanol electro-oxidation activity. Three explanations also suggest that the crystalline water connected to MnO2 in Pt/MnO2/CNTs and PtRu/MnO2/CNTs is very important to increase its EAS and the methanol electrooxidation activity. In order to prove the effect of hydrous MnO2 in PtRu/MnO2/ CNTs on catalytic activity, the electrochemical property of the MnO2-x/CNTs and PtRu/MnO2-x/CNTs in the electrolytes was measured by cyclic voltammogram method (see Supporting Information Figures 4S-6S). The results showed that the QH and QCO values of PtRu/MnO2-x/CNTs were far lower than those of PtRu/MnO2/CNTs; the responding current of methanol oxidation by PtRu/MnO2/CNTs was 2.5 times larger than that of PtRu/MnO2-x/CNTs. In Table 2, it is shown that the electrooxidation activity of PtRu/MnO2-x/CNTs is similar to that of PtRu/CNTs. In a word, the PtRu/MnO2-x/CNTs could not improve the performances of PtRu/CNTs, which inversely proved the hydrous MnO2 in PtRu/MnO2/CNTs was very important to increase its EAS and the methanol electro-oxidation activity.

4. Conclusions In summary, MnO2 3 xH2O/CNTs were synthesized and used to support Pt and PtRu as DMFC catalysts. MnO2 3 xH2O in the

Langmuir 2009, 25(13), 7711–7717

Article

catalysts is capable of increasing the proton conductivity and Pt utilization ratio for the catalysts and then improving the performance for direct methanol electro-oxidation. In the presence of MnO2 3 xH2O, Pt/MnO2/CNTs present high electrochemical active surface area and superior methanol electro-oxidation activity. Using PtRu nanoparticles rather than pure Pt, the as-obtained PtRu/MnO2/CNT catalyst has even higher electrooxidation activity to methanol and better oxidation ability to CO. In view of the outstanding performance of the PtRu/MnO2/CNT catalyst, it has great promising potential in DMFCs. Acknowledgment. We gratefully acknowledge the support of the Guangdong Provincial Science and Technology Project (No. 2006A10903002) and the Guangzhou Civic Science and Technology Project (No. 2007Z3-D2101) for this work. Supporting Information Available: EDS pattern of PtRu/ MnO2/CNTs, XPS analyses of PtRu/CNTs and PtRu/ MnO2/CNTs, and TPR profiles of the catalysts are shown in Figure S1-S3. Detailed analyses of the effect of the hydrous MnO2 on PtRu/MnO2/CNTs are given; the curves of CV are shown in Figures S4-S6. This material is available free of charge via the Internet at http://pubs.acs.org.

DOI: 10.1021/la900250w

7717