Electrochimica Acta 55 (2010) 2319–2324
Contents lists available at ScienceDirect
Electrochimica Acta journal homepage: www.elsevier.com/locate/electacta
Using RTILs of EMIBF4 as “water” to prepare palladium nanoparticles onto MWCNTs by pyrolysis of PdCl2 Keqiang Ding ∗ , Guokai Yang College of Chemistry and Materials Science, Hebei Normal University, Yuhua East Road No: 113, Shijiazhuang 050016, PR China
a r t i c l e
i n f o
Article history: Received 30 July 2009 Received in revised form 29 November 2009 Accepted 30 November 2009 Available online 3 December 2009 Keywords: Pyrolysis Palladium nanoparticle RTILs MWCNTs Ethanol oxidation reaction (EOR)
a b s t r a c t For the first time, palladium nanoparticles supported on MWCNTs (multi-walled carbon nanotubes), denoted as Pd/MWCNTs, were prepared by a simple pyrolysis process of PdCl2 dissolved in room temperature ionic liquids (RTILs) of 1-ethyl-3-methylimidazolium tetrafluoroborate (EMIBF4 ) rather than water. X-ray diffraction (XRD), transmission electron microscopy (TEM) and energy dispersive spectroscopy (EDS) were used to characterize the structure of Pd/MWCNTs, and the results showed that Pd nanoparticles with highly crystalline structure and a diameter of around 4 nm were prepared, and more importantly, except for carbon and palladium no other elements were detected. The results obtained from a pyrolysis process only containing PdCl2 and EMIBF4 testified that in our developed pyrolysis process, EMIBF4 was used not only as ligands, to form a novel complex, but also as a reducing agent, to reduce Pd2+ . The electrocatalytic performance of Pd/MWCNTs-modified glassy carbon electrode towards ethanol oxidation reaction (EOR) was also probed by cycle voltammetry (CV), demonstrating that it was possible to utilize the obtained Pd/MWCNTs as anode materials in fuel cell. Initiating the application of RTILs in the pyrolysis process and finding that EMIBF4 could be employed as ligands and reducing agents are the main contributions of this preliminary work. © 2009 Elsevier Ltd. All rights reserved.
1. Introduction Recently, room temperature ionic liquids (RTILs) have attracted much more attentions due to its excellent features, for example, low-volatility, non-toxicity, non-flame, higher conductivity compared to common organic solvent, and higher solubility for organic substances when comparing with aqueous solutions [1–5]. Summarily, RTILs are mainly applied in the following fields of chemistry (1) being used as solvents in organic synthesis [6]. (2) Being employed as electrolytes in electrochemistry [7]. However, to the best of our knowledge, the application of RTILs in pyrolysis process was not reported so far, though numerous papers concerning RTILs were published every year. Meanwhile, carbon nanotubes (CNTs) has become an important research field especially since the work achieved by Iijima [8] was published. Many advantages of CNTs have been reported [9]. Of them, being employed as an ideal substrate to modify the electrode surface was thought as the main contribution of CNTs when used in electrochemistry [10,11]. Therefore, immobilizing metal nanoparticles onto CNTs has turned into a main task due to the key role of CNTs and metal nanoparticles in electrocatalysis, biosensor and so on [12,13].
∗ Corresponding author. Tel.: +86 311 86268311; fax: +86 311 86269217. E-mail address:
[email protected] (K. Ding). 0013-4686/$ – see front matter © 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.electacta.2009.11.094
Due to the advantages of CNTs, immobilizing metal nanoparticles (especially, platinum, gold and other transition metals) onto CNTs, has become an important task for electrochemistry researchers. Palladium nanoparticles, due to its potential applications in fuel cells, were also attached onto the surface of CNTs by many developed methods. So far, there are three typical methods to generate Pd nanoparticles onto a CNTs surface. (a) Chemical reduction. For example, Xing and co-workers [14] prepared carbon black-supported Pd nanoparticles by a conventional impregnation synthesis method, in which PdCl2 and NaBH4 were employed as the precursor of palladium and the reducing agent, respectively. Lin et al. immobilized Pd nanoparticles onto CNTs by the reduction of Pd(hfa)2 ·xH2 O in a supercritical carbon dioxide, in which hydrogen gas was used as reducing agents [15]. (b) Thermal decomposition. For instance, Chen and co-workers synthesized the carbon nanotubes-supported Pd nanoparticles by a solid-state reaction, in which hydrogen gas was used as a reductive reagent and also PdCl2 the precursors of palladium [16]. (c) Electrochemical reduction. For instance, Cui et al. [17] modified Pd nanoparticles onto the surface of CNTs by means of galvanostat method, in which a constant current of −5 mA was applied on a graphite electrode for 4 h in a 0.05 M H2 SO4 solution having 2 mM PdCl2 . In this work, we modified Pd nanoparticles onto the surface of MWCNTs through a very simple method of hydrolysis process in which no other reducing agents were introduced except for multiwalled carbon nanotubes (MWCNTs) and RTILs of EMIBF4. XRD,
2320
K. Ding, G. Yang / Electrochimica Acta 55 (2010) 2319–2324
TEM and EDS were employed to characterize the obtained samples, revealing that palladium nanoparticles were formed on MWCNTs. The electrocatalytic activities of Pd/MWCNTs towards EOR were examined by cyclic voltammograms (CVs) in a 1 M KOH solution, indicating that EOR could occur on this developed Pd-coated MWCNTs. 2. Experimental
Electrochemical experiments were conducted on a model CHI660B electrochemical workstation (Shanghai Chenhua Apparatus, China). A conventional three-electrode system was employed, in which a Pd/MWCNTs-modified GC electrode and platinum wire was used as the working electrode and counter electrode, respectively. It should be noted that the reference electrode is a saturated calomel electrode (SCE). All potentials in this paper were reported with respect to SCE. Experiments were carried out at room temperature.
2.1. Chemicals and materials 3. Results and discussion Multi-walled carbon nanotubes (MWCNTs, outer diameter: 10–20 nm, length: 0.5–500 m) were purchased from Shenzhen Nanotech Port Co., Ltd. (China). RTILs of EMIBF4 with a purity of more than 99% were obtained from Hangzhou Chemer Chemical Co., Ltd. (China). A glassy carbon (Ø = 2 mm) electrode purchased from Tianjin Aida Co., Ltd. (China) was used as the working electrode. 2.2. Pyrolysis of PdCl2 to form Pd nanoparticles onto MWCNTs in EMIBF4 MWCNTs supported Pd catalysts were synthesized by a simple pyrolysis process. Firstly, 2 ml EMIBF4 containing 5 × 10−3 M PdCl2 and 10 mg MWCNTs were mixed together to yield a suspension solution, and then this resultant solution was ultrasonicated for 30 min. Secondly, the obtained suspension solution was placed in a home-made autoclave at room temperature, and the well sealed autoclave was transferred to a box-type furnace. Lastly, the temperature of the box-type furnace was increased to 230 ◦ C within 20 min, and then it was kept for 3 h to fulfill the pyrolysis process. The resultant solution was filtered, and the obtained samples were washed with redistilled water thoroughly, and dried at ambient to generate the Pd-coated MWCNTs (denoted as Pd/MWCNTs). 2.3. Preparation of Pd/MWCNTs-coated glassy carbon electrode Prior to each experiment, the working electrode of a glassy carbon (GC) electrode with a diameter of 2 mm was successively polished with 1 and 0.06 m alumina powder on a microcloth wetted with doubly distilled water, leading to a electrode with a mirror like surface. For the preparation of Pd/MWCNTs-coated electrode, 4.6 mg of the resultant Pd/MWCNTs materials was added into a 1 ml aqueous solution of sodium lauryl sulfate (SDS) (the content of SDS is 1.5 mg/ml), and then the mixture was treated for 30 min with ultrasonication to form a uniform suspension. 5 L of this mixture was dropped onto the surface of a well-treated GC electrode. Finally, the resultant Pd/MWCNTs-modified GC electrode was dried with hot air prior to the following electrochemical experiments.
3.1. XRD analysis The typical XRD patterns of the obtained samples are shown in Fig. 1. For the pattern (a) in Fig. 1, the diffraction peaks at 2 of 26.1◦ and 43.2◦ are indexed to (0 0 2) and (1 0 1) planes of carbon nanotubes (CNTs), respectively, based on the data of JCPDS card 26-1077, according with the previous report very well [18]. After the pyrolysis process, the diffraction peaks corresponding to CNTs are still clearly exhibited, suggesting that the pyrolysis process did not destroy the crystal structure of MWCNTs. Also, some novel peaks were exhibited, and they can be assigned to (1 1 1), (2 0 0), (2 2 0) and (3 1 1) planes of face-centered cubic (fcc) palladium (JCPDS card, 89-4897), consistent with the former report [19] very well, and no impurity phases can be detected, indicating the formation of pure Pd and they are highly crystalline. Since no other diffraction peaks were found, Fig. 1 strongly demonstrates that only Pd particles are formed on the surface of MWCNTs. The particle average size was estimated using the Debye–Scherer formula, t = 0.89/(ˇ cos B ), where is the X-ray wavelength (1.5406 Å), B is the Bragg diffraction angle, and ˇ is the peak width at halfmaximum. The average size was 4.5 nm and 4.8 nm as calculated from the Debye–Scherer formula on the (1 1 1) and (2 0 0) peak, respectively [20]. Also, the calculated particles size is very close to that measured by TEM studied, as shown below. Fig. 2a is the TEM images of obtained samples. It can be seen that after the pyrolysis process, as shown by image (b), some black nanoparticles were exhibited on the surface of MWCNTs, which is
2.4. Characterization X-ray diffraction (XRD) analysis of the catalyst was carried out on a Bruker D8 ADVANCE X-ray diffractometer equipped with a Cu K␣ source ( = 0.154 nm) at 40 kV and 30 mA. The 2 angular region between 10◦ and 90◦ was explored at a scan rate of 1◦ /step. The obtained samples were characterized using transmission electron microscopy (TEM, HITACHI, H-7650, Japan). EDS spectrum analysis was carried out on a X-ray energy instrument (EDAX, PV9900, USA). UV–vis spectra were obtained on a spectrophotometer V-500 (JASCO, Japan). Pyrolysis was implemented in a SRJX-8-13 box-type furnace equipped with KSY 12-16 furnace temperature controller.
Fig. 1. XRD patterns for MWCNTs (a) and Pd/MWCNTs (b).
K. Ding, G. Yang / Electrochimica Acta 55 (2010) 2319–2324
2321
Fig. 2. (a) TEM images of MWCNTs (a) and Pd/MWCNTs (b). (b) EDS spectra for Pd/MWCNTs.
very similar to the former report [21]. The diameter of Pd nanoparticles is about 4–5 nm based on the TEM images, according with the results evaluated from XRD patterns very well. Moreover, as shown in Fig. 2b, carbon and palladium element peaks were clearly exhibited in the EDS spectra for the Pd/MWCNTs samples, and the weight contents of palladium and carbon were analyzed to be about 1.96 and 98.04 wt%, respectively, implying that no impurity was involved in the obtained samples. Numerous attempts told us that the palladium nanoparticles could not be prepared in the absence of RTILs of EMIBF4 in above pyrolysis process, that is to say, Pd nanoparticles could not be fabricated by above pyrolysis process having MWCNTs and PdCl2 aqueous solution. Thus, an unavoidable question came into being. What is the reducing agent in above pyrolysis process? Theoretically, there must be some substances to release their electrons to Pd2+ to yield Pd atoms. Therefore, the following pyrolysis process containing only PdCl2 and EMIBF4 was carried out. 3.2. Preparation of Pd nanoparticles by pyrolysis process containing PdCl2 and EMIBF4 To probe the role of EMIBF4 in above pyrolysis process, the pyrolysis process having only PdCl2 and EMIBF4 was conducted. As
shown in Fig. 3a, the transparent solution of EMIBF4 was changed into an orange red one as PdCl2 was introduced. After the pyrolysis process, a yellow solution was exhibited as shown by image (c) in Fig. 3a. The color variation indicated that an interaction between EMIBF4 and PdCl2 took place, probably, some complexes having Pd2+ and BF+ were formed based on the previous paper 4 reporting the complex of Pd2+ [22]. UV–vis spectra are shown in Fig. 3b, in which dotted line corresponds the pure EMIBF4, and the solid is the case when PdCl2 was dissolved in EMIBF4. Interestingly, after the pyrolysis process, as shown by the dashed line, the absorbance value at the wavelength of maximum absorbance peak (max ) was dramatically attenuated, indicating that some complexes were consumed in the pyrolysis process. Meanwhile, the phenomenon of red-shift was observed for the maximum absorbance peak, for example, for the pure EMIBF4, the value of max is about 304 nm though its absorbance value is only 0.15, while as PdCl2 was introduced, the value of max was altered to be around 336 nm with a absorbance value of 4.47, at least demonstrating that a novel complex was formed. Interestingly, along with the color variation and the decreased absorbance value, after the pyrolysis process, the value of max restored to be 304 nm. Therefore, Fig. 3b indicates that a complex was formed by dissolving PdCl2 into EMIBF4, moreover, the resultant complex was
2322
K. Ding, G. Yang / Electrochimica Acta 55 (2010) 2319–2324
Fig. 3. (a) Digital photos of samples: (a) pure EMIBF4; (b) 5 × 10−3 M PdCl2 dissolved in pure EMIBF4; (c) after pyrolysis process. (b) UV–vis spectra for samples in (a). Dotted line: for pure EMIBF4; solid line: for the 5 × 10−3 M PdCl2 dissolved in pure EMIBF4; dashed line: samples after pyrolysis process.
destroyed by this pyrolysis process. It should be mentioned that in the absence of PdCl2 , no change was found for the pure EMIBF4 after such a pyrolysis process, thus, it was confirmed that Pd2+ was reduced by EMIBF4, though we cannot present a satisfied interpretation on the reduction mechanism of Pd2+ with our present techniques. Fig. 4a is the SEM images for the Pd nanoparticles prepared by above pyrolysis process containing only PdCl2 and EMIBF4. Fig. 4b is the EDS spectra for the resultant samples of Fig. 4a, in which except for palladium element no other peaks was exhibited, suggesting that the obtained particles were pure Pd nanoparticles. Although, many methods were developed to fabricate Pd nanoparticles, Pd nanoparticles with such good cubic crystal structures as shown in Fig. 4a were seldom reported [23]. To our knowledge, this is the first time to report the preparation of Pd nanoparticles having a diameter about 200 nm by a pyrolysis process in which TRILs were employed as ligands, and reducing agents as well. It seems that a self-contradictory result was obtained, i.e., the diameter of Pd nanoparticles on MWCNTs is only about 4–5 nm, while, in the absence of MWCNTs, Pd particles with a diameter about 200 nm were exhibited. Unfortunately, we can only show our results in this preliminary work, and cannot provide a satisfied explanation at present research stage.
3.3. Electrocatalysis of Pd/MWCNTs towards EOR As shown by the dotted line in Fig. 5, no oxidation peaks was found on the MWCNTs-coated GC electrode in 1 M KOH having 1 M ethanol. While on the Pd/MWCNTs-modified electrode, two oxidation peaks were exhibited, as shown by the solid line, implying that the pyrolysis-prepared Pd nanoparticles have catalysis towards EOR, according with the previous report very well [24]. In Fig. 5, the anomalous oxidation peak appearing in the negative-direction potential scan is still clearly observed, implying that the oxidation mechanism of ethanol in alkaline solution was not greatly varied on our developed Pd/MWCNTs-coated GC electrode. But one can see that the ratio of I2 to I1 is highly larger than unit, rather different from that the result in Ref. [24] in which the ratio of I2 to I1 is close to unit. This abnormal ratio of I2 /I1 suggests that EOR on this developed Pd/MWCNTs-modified GC electrode undergoes a modified path. Another interesting result was shown in Fig. 6. It can be seen that both I1 and I2 were gradually increased at the scan rate of 50 mV/s in continuous 10 cycles, suggesting that Pd nanoparticles obtained by our developed method have catalysis towards EOR even after several potential cycles. To our knowledge, this is the first time to report this interesting result so far. Theoretically, the oxidation
K. Ding, G. Yang / Electrochimica Acta 55 (2010) 2319–2324
2323
Fig. 5. Cyclic voltammograms (CVs) for 1 M KOH having 1 M ethanol. Dotted line: obtained on a MWCNTs-coated GC electrode; solid line: obtained on a Pd/MWCNTscoated GC electrode.
dered the oxidation of ethanol. And in the negative-going potential sweep, it was proposed that the following reaction occurred [26]: Pd–O + H2 O + 2e− → Pd + 2OH− .
(4)
Hence, EOR could still occur on the released Pd active sites in the negative-going potential sweep. It should be noted that our employed EMIBF4 is free of water, or in other words, in our developed pyrolysis process, little water was introduced, therefore, the content of OH− is very low, i.e., palladium oxides were seldom formed in this pyrolysis process, that is to say, Pd nanoparticles prepared by this pyrolysis process have more larger fresh surface or more active sites when compared to Pd nanoparticles prepared
Fig. 4. (a) SEM images for the Pd particles prepared by the pyrolysis process containing EMIBF4 and PdCl2 . (b) EDS spectra for Pd particles shown in (a).
peak current should decrease with the potential sweep cycles since EOR is an irreversible oxidation reaction. How do we understand this phenomenon? Probably, the self-catalysis of Pd nanoparticles enhanced its catalysis towards EOR, further investigations are really required in the further work. Although the mechanism of EOR remains unclear, it has been widely accepted that OH− ions are first chemisorbed in the initial stage of the oxide formation and then they are transformed into higher valence oxides at higher potentials, as described by the following formulas [25]: Pd + OH− → Pd–OHads + e− −
Pd–OHads + OH → Pd–O + H2 O + e
(1) −
Pd–OHads + Pd–OHads → Pd–O + H2 O
(2) (3)
That is to say, in the positive direction potential scan, Pd–O was formed, which covered on some active sites of Pd surface and hin-
Fig. 6. CVs obtained on a Pd/MWCNTs-coated GC electrode in 1 M KOH having 1 M ethanol at 50 mV/s for 10 cycles.
2324
K. Ding, G. Yang / Electrochimica Acta 55 (2010) 2319–2324
by other methods, thus, probably the surface of Pd nanoparticles we prepared differed from that commonly purchased Pd substrate. Hence, a larger ratio of I2 to I1 was exhibited. Based on the CVs in Figs. 5 and 6, it is reasonable to think that for our developed Pd nanoparticles, they are less poisoned by the chemisorbed OH− ions when compared to a Pd electrode [24]. 4. Conclusions Pd nanoparticles coated MWCNTs were prepared by a pyrolysis process of PdCl2 dissolved in EMIBF4. XRD, TEM and EDS as well, all strongly demonstrated that Pd/MWCNTs were generated by our proposed novel method, in which no other reducing agents were introduced. Further studies revealed that EMIBF4 played an important role in above pyrolysis process, in which EMIBF4 was used as ligands, to form a complex containing Pd2+ , and reducing agents as well, to reduce Pd2+ to yield Pd atoms. The obtained CVs indicated that our developed Pd nanoparticles have more catalysis towards EOR when compared to the commonly used Pd substrate. Acknowledgements This work was financially supported by the Doctor Fund of Hebei Normal University, Key Project of Hebei Province Education Bureau (ZH2007106), Key Project Fund of Hebei Normal University (L2008Z08) and Special Assist Project of Hebei Province Personnel Bureau (106115).
References [1] T. Welton, Chem. Rev. 99 (1999) 2071. [2] H.S. Park, Y.S. Choi, Y.M. Jung, W.H. Hong, J. Am. Chem. Soc. 130 (2008) 845. [3] M.P. Jensen, J.A. Dzielawa, P. Rickret, M.L. Dietz, J. Am. Chem. Soc. 124 (2002) 10664. [4] J.S. Lee, N.D. Quan, J.M. Hwang, J.Y. Bae, H. Kim, B.W. Cho, H.S. Kim, H. Lee, Electrochem. Commun. 8 (2006) 460. [5] V.W.-W. Yam, J.K.-W. Lee, C.-C. Ko, N. Zhu, J. Am. Chem. Soc. 131 (2009) 912. [6] J.S. Wilkes, J. Mol. Catal. A: Chem. 214 (2004) 11. [7] Y. Zhang, J. Zheng, Electrochem. Commun. 10 (2008) 1400. [8] S. Iijima, Nature 354 (1991) 56. [9] C.E. Banks, T.J. Davies, G.G. Wildgoose, R.G. Compton, Chem. Commun. (2005) 829. [10] K.-Q. Ding, T. Okajima, T. Ohsaka, Electrochemistry 75 (2007) 35. [11] J. Yan, H. Zhou, P. Yu, L. Su, L. Mao, Electrochem. Commun. 10 (2008) 761. [12] J. Shen, Y. Hu, C. Li, C. Qin, M. Ye, Electrochim. Acta 53 (2008) 7276. [13] K.-Q. Ding, M. Cao, Russ. J. Electrochem. 44 (2008) 977. [14] Y. Huang, X. Zhou, J. Liao, C. Liu, T. Lu, W. Xing, Electrochem. Commun. 10 (2008) 1155. [15] Y. Lin, X. Cui, X. Ye, Electrochem. Commun. 7 (2005) 267. [16] B. Xue, P. Chen, Q. Hong, J. Lin, K.L. Tan, J. Mater. Chem. 11 (2001) 2378. [17] C. Cui, X. Quan, H. Yu, Y. Han, Appl. Catal. B: Environ. 80 (2008) 122. [18] J. Xu, K. Hua, G. Sun, C. Wang, X. Lv, Y. Wang, Electrochem. Commun. 8 (2006) 982. [19] W. Wang, Q. Huang, J. Liu, Z. Zou, Z. Li, H. Yang, Electrochem. Commun. 10 (2008) 1396. [20] V. Radmilovic, H.A. Gasteiger, P.N. Ross, J. Catal. 154 (1995) 98. [21] Z.-P. Sun, X.-G. Zhang, Y.-Y. Liang, H.-L. Li, Electrochem. Commun. 11 (2009) 557. ´ [22] A. Drelinkiewicz, M. Hasik, M. Choczynski, Mater. Res. Bull. 33 (1998) 739. [23] L. Zhang, K. Lee, J. Zhang, Electrochim. Acta 52 (2007) 3088. [24] S. Yang, X. Zhang, H. Mi, X. Ye, J. Power Sources 175 (2008) 26. [25] M. Grden, A. Czerwinski, J. Solid State Electrochem. 12 (2008) 375. [26] M. Grden, J. Kotowski, A. Czerwinski, J. Solid State Electrochem. 4 (2000) 273.