- Email: [email protected]
Evaluation of impregnated nanocomposite membranes for aqueous methanol electrochemical reforming
Solid State Ionics 283 (2015) 16–20
Contents lists available at ScienceDirect
Solid State Ionics journal homepage: www.elsevier.com/locate/ssi
Evaluation of impregnated nanocomposite membranes for aqueous methanol electrochemical reforming Sundar Pethaiah Sethu a,c,⁎, Mangalaraja Ramalinga Viswanathan b, Ulaganathan Mani c, Siew Hwa Chan c a b c
Technische Universität München — Campus for Research Excellence and Technological Enterprise, 1 CREATE Way, Singapore 138602, Singapore Advanced Ceramics and Nanotechnology Laboratory, Department of Materials Engineering, University of Concepcion, Concepcion 407-0409, Chile Energy Research Institute @NTU, Nanyang Technological University, 50 Nanyang Drive, Singapore 637553, Singapore
a r t i c l e
i n f o
Article history: Received 2 June 2015 Received in revised form 30 September 2015 Accepted 5 November 2015 Available online 18 November 2015 Keywords: Fuel cell Hydrogen Methanol Electrolysis Methanol crossover
a b s t r a c t In this work, we present the methanol electrolysis performance of the bimetallic Pt–Pd impregnated Nafion based nanocomposite polymer electrolyte membrane prepared by non-equilibrium impregnation–reduction (NEIR) method. The different atomic concentrations of Pt–Pd impregnated nanocomposite membranes were subjected to X-ray diffraction for their structural characteristics. The effect of various parameters such as methanol concentration, cell voltage and temperature on the cell performance was investigated. Further, the bimetallic concentration and catalyst loading were also optimized for methanol electrolysis. The results indicated that the bimetallic Pt–Pd of a ratio 70:30 impregnated Nafion based nano-composite membrane showed the best performance on the methanol electrolysis. © 2015 Elsevier B.V. All rights reserved.
1. Introduction The increase in the need of energy consumption to the futuristic world has led the scientists to search for a new and clean energy source. Hydrogen has high energy density and is considered as a most promising energy carrier for providing a clean and sustainable energy system to prevent from the climate changes related to the use of other fuels, especially the use of fossil fuels [1,2]. Electrolysis is a best option for producing hydrogen very quickly and conveniently. Particularly, electrolysis of aqueous methanol is a promising method for onsite hydrogen production, and it has been reported that very high purity hydrogen can be generated by electrolysis of methanol–water mixture at a very low operating voltage, compared to water electrolysis [2,3]. In a methanol electrolysis cell, methanol is fed to the anodic compartment where it can be completely oxidized in the presence of water, producing carbon dioxide and protons, i.e.,: At Anode : CH3 OH þ H2 O→CO2 þ 6Hþ þ 6e− E0a ¼ −0:016V vs SHE: ð1Þ
⁎ Corresponding author at: Technische Universität München — Campus for Research Excellence and Technological Enterprise, 1 CREATE Way, Singapore 138602, Singapore. Tel.: +65 65923013; fax: +65 68969950. E-mail address: [email protected] (S.P. Sethu).
http://dx.doi.org/10.1016/j.ssi.2015.11.006 0167-2738/© 2015 Elsevier B.V. All rights reserved.
Carbon dioxide evolves in the form of gaseous phase, whereas the electrons circulate in the external circuit and protons crossover the polymer electrolyte membrane, reaching the cathodic compartment where they are reduced by the electrons coming from the external circuit, thus producing hydrogen as follows: At Cathode : 6Hþ þ 6e− →3H2 E0c ¼ 0V vs SHE:
ð2Þ
This corresponds to the overall decomposition of methanol into hydrogen and carbon dioxide: Overall : CH3 OH þ H2 O→3H2 þ CO2 E0 ¼ −0:016V vs SHE:
ð3Þ
Hydrogen production by aqueous methanol electrolysis is simple in principle [4], in the patent by Narayanan et al. [5] who first described hydrogen production by electrolysis of aqueous organic solutions. There are several studies on aqueous methanol electrolysis have been reported and most of the works are focused on catalyst development and parameter optimizations [2–6]. However, it is well known that Nafion perfluorosulfonic acid polymers are commonly used as electrolyte membranes due to their excellent thermal and chemical stability, and their disadvantage is that they are quite permeable to methanol [7–9]. Methanol crossover may reduce the cell performance [10], purity of hydrogen and conversion loss in lost fuel. In addition cathode catalyst layer is poisoned by CO due to methanol oxidation intermediate. In order to mitigate the effect of methanol crossover in direct methanol fuel cells, a number of techniques have been reported [10–13] and the modification of Nafion without decreasing its proton conductivity is
S.P. Sethu et al. / Solid State Ionics 283 (2015) 16–20
one of the important techniques, such as hybridizing the Nafion membrane with sulfonated organic silica, zeolites, conducting polymers and polymer composites, metallic nanoparticles, heteropolyacids and composite organic/inorganic thin films onto the Nafion membrane [14]. Among the others, the surface modification of Nafion membrane with incorporation of palladium (Pd) layers and Pd nanoclusters is an interest of recent research [15]. It has been widely reported that the Pt catalyzed membrane is suitable for electrolyzer applications because of the use of a Pt-catalyzed membrane as an anode for methanol oxidation offers advantages over conventional cells [16–19]. Furthermore, the addition of Pd suppresses the methanol permeation and enhances proton conductivity which makes the bimetallic Pt–Pd/Nafion nanocomposite as the promising membrane for methanol electrolysis. To the best of our knowledge, so far no one has studied the Pt–Pd impregnated Nafion nanocomposite membrane for methanol electrolysis. On the other hand, the catalyzed membrane preparation is an attractive as it involves only simple steps and also for application processes. Furthermore, the development of a nanocatalyzed membrane by the non-equilibrium impregnation– reduction (NEIR) method is one of the best methods for electrolyzer applications [17,20,21]. Hence, in this study, bimetallic Pt–Pd nanophase was impregnated to modify the Nafion membrane. The Pt–Pd impregnated Nafion based nanocomposite polymer electrolyte membranes fabricated by NEIR method were investigated for methanol electrolyzer applications. 2. Experimental procedure The methodology used for membrane pre-treatment and Pt–Pd impregnated nanocomposite membrane preparation method has been reported already in our previous work [22]. The reproducibility of the Pt–Pd impregnated nanocomposite membranes varies with the fabrication procedure. However, the reproducibility of the Pt–Pd impregnated nanocomposite membranes extents to keep fabrication procedure identical [21]. Hence we keep the fabrication procedure identical to facilitate the reproducibility. Accordingly, Pt–Pd was deposited on one side of Nafion membrane and was used as an anode (Fig. 1). On the other hand, 40% Pt/C (Duralyst, USA) was used as a cathode catalyst with 1 mg cm−2 loading. The electrocatalyst slurry was prepared by dispersing a 40% Pt/C in water and followed by sonication at room temperature for 30 min. A 5 wt.% of Nafion ionomer, and isopropyl alcohol were added to the mixture and again sonicated at room temperature, and then the obtained slurry was coated onto the membrane using decal transfer method. In decal transfer method the above prepared catalyst ink was applied on a polyimide film and dried at 100 °C. The catalyst
Fig. 1. Pt–Pd impregnated nanocomposite membrane.
17
coated polyimide film was placed on pretreated membrane and was hot pressed to 1000 psi at 130 °C for 3 min. Later, the polyimide film was peeled off, leaving a uniform layer of the catalyst on the membrane. Finally again MEAs were hot pressed with microporous layer coated gas diffusion layer (GDL) with uniform holes on anode side and without holes on cathode side. The pressure was maintained to 1000 psi for 3 min at 130 °C (Fig. 2). For comparative evaluation, a normal MEA with both side 40% Pt/C catalysts was prepared by decal transfer method. X-ray diffraction (XRD) patterns of Pt–Pd/Nafion nanocomposite membranes were acquired at room temperature with an X'pert PRO PANalytical diffractometer using Cu-kα radiation as source and operated at 40 kV. The samples were scanned in the 2θ ranging from 10 to 80° for 2 s in the step scan mode. Single cell experiments were performed by placing the Pt–Pd impregnated Nafion nanocomposite MEA between two gaskets of thickness 0.25 mm and inserted between two graphite plates with straight parallel grooves and 10 cm2 active areas. The above items were clamped together using nuts and bolts, by applying uniform torque, to assemble the single cell. The cell had provision for heating, temperature control and suitable ports for feeding the reactants and removal of products. The electrolysis was conducted to generate hydrogen at the cathode, by applying current across the two terminals of the electrolyzer using a programmable DC power supply unit having constant current and voltage mode provisions. Hydrogen production rate was calculated from the cell current and was cross examined by gas volume measurement. The purity of hydrogen gas was checked by using a calibrated hydrogen purity analyzer (NOVA Gas Analyzer). 3. Results and discussions 3.1. Physical characterizations 3.1.1. XRD XRD patterns of different ratios of Pt–Pd layered Nafion 117 membranes and pure Nafion 117 membrane are shown in Fig. 3. The samples showed the crystalline characteristic peaks around at 2θ = 40, 47 and 68° which represented (1 1 1), (2 0 0), and (2 2 0) lattice planes, respectively, of face-centered cubic (fcc) structure of Pt and Pd [23]. A broad hump observed at 17° revealed the amorphous nature of Nafion membrane [24]. It was confirmed that the face centered cubic structure
Fig. 2. MEA with gasket.
18
S.P. Sethu et al. / Solid State Ionics 283 (2015) 16–20
Fig. 4. Cell voltage–current density characteristics of Pt–Pd composite MEA with different ratios and normal MEA at 80 °C.
Fig. 3. X-ray diffraction analysis of prepared membranes.
(fcc) of Pt and Pd was overlapped in all the samples. However, there was no characteristic peak of metallic or oxide form of Pd observed, but their presence could not be discarded because Pd might be presented in Pt lattice and formed alloy or even in an amorphous form. Similar observations were obtained for PtRu based catalyst layered systems by Naoko Fujiwara et al. [19,25]. From the XRD measurements, the crystallite size of the Pt–Pd catalyst was calculated from the broadening of the (1 1 1) diffraction peak using Debye–Scherer's relation. The calculated crystallite size, atomic and weight ratios are given in Table 1. It was observed that the crystallite was size increased with Pd content. Surface image and cross sectional view of the Pt–Pd (70:30) bimetallic layered Nafion® 117 membranes have been elaborately discussed in our previous work [22]. 3.2. Electrochemical characterizations The Nafion membrane was made of hydrophobic back-bone structure and a hydrophilic side chain. In a well-hydrated membrane, there will be approximately 20 water molecules for each SO3 side chain [16, 26]. Therefore, the impregnation solutions of Pt and Pd occupy the hydrated regions and precipitated in the hydrophilic region after adding the reducing agent NaBH4 then reduce the concentration of water molecules, which then decreases the proton conductivity and methanol crossover. Hence, it is important to determine the optimum Pt–Pd composition and catalyst loading. To optimize the Pt–Pd ratios and catalyst loading, the single cell performance was evaluated. Figs. 4 and 5 show the voltage–current density characteristics of the single cell for to optimize the Pt–Pd ratios and catalyst loading, respectively at 80 °C with 4 M of methanol as a fuel. It was noted from Fig. 4 that the various Pt:Pd atomic ratios in the Pt–Pd composite membranes had a significant effect on the cell performances. For a quick overview, the performance of the single cell containing Pt:Pd (70:30) nanocomposite membrane gave better performance than the other ratios and normal MEA under identical test conditions that might be due to the reduced methanol
crossover and better methanol oxidation catalytic activity of Pt–Pd bimetallic catalyst [27–29]. The Pt:Pd (50:50) nanocomposite membrane showed lower performance than the other ratios and that might be due to higher nanoparticle crystalline size [30]. However, detailed studies of the catalyst analysis and origin of the performance enhancement of Pt–Pd (70:30) will be reported in the future. Fig. 5 shows current density vs voltage curves of above better performed Pt:Pd (70:30)/Nafion nanocomposite membrane. The loading amount of Pt–Pd was 1, 2 and 3 mg cm−2 corresponding to one, two and three repetitions of NEIR method, respectively. From the plot, it was noted that the performance of the single cell increased with Pt–Pd loading up to 2 mg cm − 2 , above which the cell exhibited lower performance. Among the various loadings, 2 mg cm− 2 loaded membrane delivered superior performance than the other samples due to an increase in methanol oxidation rate by increasing the catalyst loading and the decreasing level of methanol crossover. However, further increase in catalyst loading causes Pt–Pd particle isolation and leads to a lack of three-phase zone which reduced the cell performance. Moreover, mentioned earlier increased Pt–Pd depositions occupy the hydrated regions and reduced the proton conductivity, which also lowered the cell performances. The effect of catalyst loading on mass activity is shown in Fig. 6. An increase of mass activity up to −2 1.48 A mg−1 catalyst loading, compared Pt:Pd was observed for 1 mg cm −1 with 0.8 and 0.44 A mgPt:Pd for 2 and 3 1 mg cm− 2 respectively (at 0.9 V). The mass activity increased when the catalyst loading decreased which might be due to high electrochemical surface area and high catalyst utilization [31]. However, due to ohmic and mass transport losses within the electrode matrix, higher electrochemical surface area did not increase the electrolyzer performance in our case [32].
Table 1 XRD and EDX characteristics of the prepared membranes. Metal compositions in solution (at. %)
XRD data
EDX data
Crystallite size (nm)
Metal compositions (at.%)
Metal compositions (wt.%)
Pt–Pd (90:10) Pt–Pd (70:30) Pt–Pd (50:50)
5.4 5.7 6.7
94.61:5.39 79.09:20.91 62.55:37.45
96.98:3.02 87.40:12.60 75.39:24.61
Fig. 5. Effect of Pt–Pd loading on the performance of electrolyzer at 80 °C with 4 M methanol.
S.P. Sethu et al. / Solid State Ionics 283 (2015) 16–20
19
Fig. 6. Effect of Pt–Pd loading on mass activity of MEAs at 80 °C with 4 M methanol. Fig. 8. Effect of operating temperature with Pt–Pd (70:30) nanocomposite membrane at 80 °C with 4 M methanol.
3.3. Operating parameter optimizations Fig. 7 shows the comparative study of Pt and Pt–Pd (70:30) catalysts with various methanol concentrations. Electrolysis of aqueous methanol–water solution was carried out at a cell voltage of 1 V, in a PEM electrolyzer cell of active area 10 cm2, using Nafion-117 membrane at 80 °C. Hydrogen and carbon dioxide gases were generated at cathode and anode, respectively. The concentration of methanol was varied from 2 to 10 M to study the effect of methanol concentration on current density, which was directly related to the rate of hydrogen production. From the results it was noted an increase in current density with methanol concentration, however, a sharp increase in current density was observed up to about 6 M methanol concentration and from 6 to 10 M, the incremental gain was low due to the reason that the variation of current density with methanol concentration might be correlated to the change of membrane conductivity with methanol concentration. Nafion membrane was described as a series of clusters interconnected by narrow pores. In each cluster, the fixed membrane charges extended inward the center of each sphere. Due to the lower polarity of methanol compared to water, as the aqueous methanol concentration was increased, the cluster size was increased and hence the ionic conductivity of the membrane decreased, resulting in lower cell performance at high methanol concentrations [2,33,34]. However, Pt–Pd composite membrane was performed better than normal membrane which is mainly due to the reduction in methanol crossover. Moreover, the Pt–Pd alloy catalyst with a suitable ratio has also been exhibit very good resistance against CO poisoning effect [35]. The synergistic effect between Pt and Pd enhances the electrocatalytic activity towards methanol oxidation reaction. The presence of Palladium dilutes the Pt sites and prevents
the presence of three adjacent sites that are necessary for the adsorption of poison species on the electrode surface. This leads to the overall enhancement of methanol oxidations [28,29,36] Fig. 8 shows the effect of the reaction temperature on the current density vs voltage by feeding 4 M of methanol to the anode side. The current density was found to increase with temperature. The observed improvement in cell performance with temperature may be attributed to the enhanced kinetics of the electrochemical reaction and the decreased ohmic resistance. For a quick overview Pt–Pd (70:30)/Nafion nanocomposite membrane anode in the electrolyzer delivered a maximum current density value of 1.9 A cm−2 at 1 V with the operation at 80 °C and 4 M of methanol concentration. The optimized single cell performance with hydrogen production rate is shown in Fig. 9. The maximum volumetric hydrogen production rate is directly proportional to the current density and it can be explained in Eq. (4) below,
Fig. 7. Effect of methanol concentration on current density of methanol electrolyzer at 1.0 V and at 80 °C.
Fig. 9. Optimized electrolyzer performance and hydrogen production rate for Pt–Pd (70:30)/Nafion nanocomposite MEA at 80 °C with 4 M methanol.
Number of hydrogen moles ¼
ð4Þ
ðcurrent in amperesÞðtime in secondsÞ : ðunsigned numeric charge on the ionÞðFaradayÞ
Fig. 10 compares the Faradaic hydrogen production rate and the real selectivity of hydrogen from the experiment (methanol conversion) for Pt–Pd (70:30)/Nafion nanocomposite membrane. The results confirmed that the Faradaic current efficiencies for hydrogen production were near to ≈ 97–99%. The purity of the hydrogen gas was then checked using a Nova hydrogen analyzer (Nova Analytical Systems,
20
S.P. Sethu et al. / Solid State Ionics 283 (2015) 16–20
Fig. 10. Experimental and Faradaic hydrogen production rate comparison.
USA) that indicated 99% hydrogen purity which is highly compatible for fuel cell applications. This simplifies the fuel cell with fuel processor concept for transport applications. 4. Conclusions In the present work, the bimetallic Pt–Pd impregnated Nafion nanocomposite polymer electrolyte membrane was successfully prepared and its performance on methanol electrolysis was evaluated. From the results of the single cell test, the Pt–Pd bimetallic/Nafion nanocomposite membrane performed better than normal one due to the combined effect of reduced methanol crossover and better catalytic activity. Pt– Pd (70:30)/Nafion nanocomposite membrane anode in the electrolyzer delivered a maximum current density value of 1.9 A cm−2 at 1 V with 80 °C and 4 M methanol concentration. References [1] A. Caravaca, F.M. Sapountzi, A. de Lucas-Consuegra, C. Molina-Mora, F. Dorado, J.L. Valverde, Int. J. Hydrog. Energy 37 (2012) 9504–9513. [2] G. Sasikumar, A. Muthumeenal, S. Sundar Pethaiah, N. Nachiappan, R. Balaji, Int. J. Hydrog. Energy 33 (2008) 5905–5910.
[3] J. Nakamura Barbara, NASA Tech. Brief 26 (6) (2009). [4] Zhiyi Hu, Mei Wu, Zidong Wei, Shuqin Song, Pei Kang Shen, J. Power Sources 166 (2007) 458–461. [5] S.R. Narayaman, W. Chun, B.J. Nakamura, S. Marino, T.I. Valdez (2001) Hydrogen generation by electrolysis of aqueous organic solutions, U.S. Patent 6799744. [6] Tetsuo Take, Kazuhiko Tsurutani, Minoru Umeda, J. Power Sources 164 (2007) 9–16. [7] H. Sun, G. Sun, S. Wang, J. X Liu, X. Zhao, G. Wang, H. Xu, S. Hou, Q. Xin, J. Membr. Sci. 259 (2005) 27–33. [8] M.F. Hogarth, G.A. Hards, Platin. Met. Rev. 40 (1996) 150–159. [9] K. Scott, W. Taama, J. Cruickshank, J. Power Sources 65 (1997) 159–171. [10] Fangjian Shang, Lei Li, Yongming Zhang, Hong Li, J. Mater. Sci. 44 (2009) 4384–4388. [11] A.M. Zainoodin, S.K. Kamarudin, M.S. Masdar, W.R.W. Daud, A.B. Mohamad, J. Sahari, Appl. Energ. 113 (2014) 946–954. [12] Rong Zeng, Yiwang Chen, Weihua Zhou, Shuqin Xiao, Jichun Xiao, Caisheng Song, J. Mater. Sci. 45 (2010) 1610–1616. [13] Joon-Hee Kim, Min-Jee Yang, Jun-Young Park, Appl. Energ. 115 (2014) 95–102. [14] Tiago P. Almeida, Celina M. Miyazaki, Valdecir A. Paganin, Marystela Ferreira, Margarida J. Saeki, Joelma Perez, Antonio Riul Jr., Appl. Energ. 323 (2014) 7–12. [15] L. Brandao, J. Rodrigues, L.M. Madeira, A. Mendes, Int. J. Hydrog. Energy 35 (2010) 11561–11567. [16] E.H. Jung, U. H Jung, T.H. Yang, D.H. Peak, D.H. Jung, S.H. Kim, Int. J. Hydrog. Energy 32 (2007) 903–907. [17] R. Liu, P. Fedkiw PS, J. Electrochem. Soc. 139 (1992) 3514–3523. [18] A. Katayama-Aramata, R. Ohnishi, J. Am. Chem. Soc. 105 (1983) 658–659. [19] N. Fujiwara, K. Yasuda, T. Ioroi, Z. Siroma, Y. Miyazaki, Electrochim. Acta 47 (2002) 4079–4084. [20] S. Sundar Pethaiah, G. Sasikumar, S.H. Chan, U. Stimming, J. Power Sources 254 (2014) 61–167. [21] Raymond Liu, Wei-Hwa Her, Peter S. Fedkiw, J. Electrochem. Soc. 139 (1992) 15–23. [22] S. Sundar pethaiah, M. Ulaganathan, R.V. Mangalaraja, S.H. Chan, RSC Adv. 5 (2015) 981–987. [23] S.R. Yoon, G.H. Hwang, W.I. Cho, I.H. Oh, S.A. Hong, H.Y. Ha, J. Power Sources 106 (2002) 215–223. [24] M. Ludvigsson, J. Lindgren, J. Tegenfeldt, J. Electrochem. Soc. 147 (2000) 1303–1305. [25] H. Qiao, T. Kasajima, M. Kunimatsu, N. Fujiwara, T. Okada, J. Electrochem. Soc. 153 (2006) A42–A47. [26] S.R. Sammas, S. Wasmus, R.F. Savinell, J. Electrochem. Soc. 143 (1996) 1498–1504. [27] T. Hejze, B.R. Gollas, R.K. Sauerbrey, M. Schmied, F. Hofer, J.O. Besenhard, J. Power Sources 140 (2005) 21–27. [28] Jin Kim, Won Choon Choi, Won Choon Choi, Seong Ihl Woo, Won Hi Hong, Electrochim. Acta 49 (2004) 3227–3234. [29] T. Lopes, E. Antolini, E.R. Gonzalez, Int. J. Hydrog. Energy 33 (2008) 5563–5570. [30] Sarawalee Thanasilp, Mali Hunsom, Renew. Energy 36 (2011) 1795–1801. [31] Akos Kriston, Tianyuan Xie, David Gamliel, Prabhu Ganesan, Branko N. Popov, J. Power Sources 243 (2013) 958–963. [32] N. Rajalakshmi, H. Ryua, K.S. Dhathathreyan, Chem. Eng. J. 102 (2004) 241–247. [33] Anh Tuan Pham, Toshio Shudo, Int. J. Automot. Eng. 3 (2012) 125–130. [34] S. Song, G. Wang, W. Zhou, X. Zhao, G. Sun, Q. Xin, S. Kontou, P. Tsiakaras, J. Power Sources 140 (2005) 103–110. [35] D. Keqiang, W. Yahui, Y. Hongwei, Z. Chunbao, Yanli Cao, W. Huige, W. Yiran, G. Zhanhu, Electrochim. Acta 100 (2013) 147–156. [36] T. Arikan, A.M. Kannan, F. Kadirgan, Int. J. Hydrog. Energy 38 (2013) 2900–2907.
Contents lists available at ScienceDirect
Solid State Ionics journal homepage: www.elsevier.com/locate/ssi
Evaluation of impregnated nanocomposite membranes for aqueous methanol electrochemical reforming Sundar Pethaiah Sethu a,c,⁎, Mangalaraja Ramalinga Viswanathan b, Ulaganathan Mani c, Siew Hwa Chan c a b c
Technische Universität München — Campus for Research Excellence and Technological Enterprise, 1 CREATE Way, Singapore 138602, Singapore Advanced Ceramics and Nanotechnology Laboratory, Department of Materials Engineering, University of Concepcion, Concepcion 407-0409, Chile Energy Research Institute @NTU, Nanyang Technological University, 50 Nanyang Drive, Singapore 637553, Singapore
a r t i c l e
i n f o
Article history: Received 2 June 2015 Received in revised form 30 September 2015 Accepted 5 November 2015 Available online 18 November 2015 Keywords: Fuel cell Hydrogen Methanol Electrolysis Methanol crossover
a b s t r a c t In this work, we present the methanol electrolysis performance of the bimetallic Pt–Pd impregnated Nafion based nanocomposite polymer electrolyte membrane prepared by non-equilibrium impregnation–reduction (NEIR) method. The different atomic concentrations of Pt–Pd impregnated nanocomposite membranes were subjected to X-ray diffraction for their structural characteristics. The effect of various parameters such as methanol concentration, cell voltage and temperature on the cell performance was investigated. Further, the bimetallic concentration and catalyst loading were also optimized for methanol electrolysis. The results indicated that the bimetallic Pt–Pd of a ratio 70:30 impregnated Nafion based nano-composite membrane showed the best performance on the methanol electrolysis. © 2015 Elsevier B.V. All rights reserved.
1. Introduction The increase in the need of energy consumption to the futuristic world has led the scientists to search for a new and clean energy source. Hydrogen has high energy density and is considered as a most promising energy carrier for providing a clean and sustainable energy system to prevent from the climate changes related to the use of other fuels, especially the use of fossil fuels [1,2]. Electrolysis is a best option for producing hydrogen very quickly and conveniently. Particularly, electrolysis of aqueous methanol is a promising method for onsite hydrogen production, and it has been reported that very high purity hydrogen can be generated by electrolysis of methanol–water mixture at a very low operating voltage, compared to water electrolysis [2,3]. In a methanol electrolysis cell, methanol is fed to the anodic compartment where it can be completely oxidized in the presence of water, producing carbon dioxide and protons, i.e.,: At Anode : CH3 OH þ H2 O→CO2 þ 6Hþ þ 6e− E0a ¼ −0:016V vs SHE: ð1Þ
⁎ Corresponding author at: Technische Universität München — Campus for Research Excellence and Technological Enterprise, 1 CREATE Way, Singapore 138602, Singapore. Tel.: +65 65923013; fax: +65 68969950. E-mail address: [email protected] (S.P. Sethu).
http://dx.doi.org/10.1016/j.ssi.2015.11.006 0167-2738/© 2015 Elsevier B.V. All rights reserved.
Carbon dioxide evolves in the form of gaseous phase, whereas the electrons circulate in the external circuit and protons crossover the polymer electrolyte membrane, reaching the cathodic compartment where they are reduced by the electrons coming from the external circuit, thus producing hydrogen as follows: At Cathode : 6Hþ þ 6e− →3H2 E0c ¼ 0V vs SHE:
ð2Þ
This corresponds to the overall decomposition of methanol into hydrogen and carbon dioxide: Overall : CH3 OH þ H2 O→3H2 þ CO2 E0 ¼ −0:016V vs SHE:
ð3Þ
Hydrogen production by aqueous methanol electrolysis is simple in principle [4], in the patent by Narayanan et al. [5] who first described hydrogen production by electrolysis of aqueous organic solutions. There are several studies on aqueous methanol electrolysis have been reported and most of the works are focused on catalyst development and parameter optimizations [2–6]. However, it is well known that Nafion perfluorosulfonic acid polymers are commonly used as electrolyte membranes due to their excellent thermal and chemical stability, and their disadvantage is that they are quite permeable to methanol [7–9]. Methanol crossover may reduce the cell performance [10], purity of hydrogen and conversion loss in lost fuel. In addition cathode catalyst layer is poisoned by CO due to methanol oxidation intermediate. In order to mitigate the effect of methanol crossover in direct methanol fuel cells, a number of techniques have been reported [10–13] and the modification of Nafion without decreasing its proton conductivity is
S.P. Sethu et al. / Solid State Ionics 283 (2015) 16–20
one of the important techniques, such as hybridizing the Nafion membrane with sulfonated organic silica, zeolites, conducting polymers and polymer composites, metallic nanoparticles, heteropolyacids and composite organic/inorganic thin films onto the Nafion membrane [14]. Among the others, the surface modification of Nafion membrane with incorporation of palladium (Pd) layers and Pd nanoclusters is an interest of recent research [15]. It has been widely reported that the Pt catalyzed membrane is suitable for electrolyzer applications because of the use of a Pt-catalyzed membrane as an anode for methanol oxidation offers advantages over conventional cells [16–19]. Furthermore, the addition of Pd suppresses the methanol permeation and enhances proton conductivity which makes the bimetallic Pt–Pd/Nafion nanocomposite as the promising membrane for methanol electrolysis. To the best of our knowledge, so far no one has studied the Pt–Pd impregnated Nafion nanocomposite membrane for methanol electrolysis. On the other hand, the catalyzed membrane preparation is an attractive as it involves only simple steps and also for application processes. Furthermore, the development of a nanocatalyzed membrane by the non-equilibrium impregnation– reduction (NEIR) method is one of the best methods for electrolyzer applications [17,20,21]. Hence, in this study, bimetallic Pt–Pd nanophase was impregnated to modify the Nafion membrane. The Pt–Pd impregnated Nafion based nanocomposite polymer electrolyte membranes fabricated by NEIR method were investigated for methanol electrolyzer applications. 2. Experimental procedure The methodology used for membrane pre-treatment and Pt–Pd impregnated nanocomposite membrane preparation method has been reported already in our previous work [22]. The reproducibility of the Pt–Pd impregnated nanocomposite membranes varies with the fabrication procedure. However, the reproducibility of the Pt–Pd impregnated nanocomposite membranes extents to keep fabrication procedure identical [21]. Hence we keep the fabrication procedure identical to facilitate the reproducibility. Accordingly, Pt–Pd was deposited on one side of Nafion membrane and was used as an anode (Fig. 1). On the other hand, 40% Pt/C (Duralyst, USA) was used as a cathode catalyst with 1 mg cm−2 loading. The electrocatalyst slurry was prepared by dispersing a 40% Pt/C in water and followed by sonication at room temperature for 30 min. A 5 wt.% of Nafion ionomer, and isopropyl alcohol were added to the mixture and again sonicated at room temperature, and then the obtained slurry was coated onto the membrane using decal transfer method. In decal transfer method the above prepared catalyst ink was applied on a polyimide film and dried at 100 °C. The catalyst
Fig. 1. Pt–Pd impregnated nanocomposite membrane.
17
coated polyimide film was placed on pretreated membrane and was hot pressed to 1000 psi at 130 °C for 3 min. Later, the polyimide film was peeled off, leaving a uniform layer of the catalyst on the membrane. Finally again MEAs were hot pressed with microporous layer coated gas diffusion layer (GDL) with uniform holes on anode side and without holes on cathode side. The pressure was maintained to 1000 psi for 3 min at 130 °C (Fig. 2). For comparative evaluation, a normal MEA with both side 40% Pt/C catalysts was prepared by decal transfer method. X-ray diffraction (XRD) patterns of Pt–Pd/Nafion nanocomposite membranes were acquired at room temperature with an X'pert PRO PANalytical diffractometer using Cu-kα radiation as source and operated at 40 kV. The samples were scanned in the 2θ ranging from 10 to 80° for 2 s in the step scan mode. Single cell experiments were performed by placing the Pt–Pd impregnated Nafion nanocomposite MEA between two gaskets of thickness 0.25 mm and inserted between two graphite plates with straight parallel grooves and 10 cm2 active areas. The above items were clamped together using nuts and bolts, by applying uniform torque, to assemble the single cell. The cell had provision for heating, temperature control and suitable ports for feeding the reactants and removal of products. The electrolysis was conducted to generate hydrogen at the cathode, by applying current across the two terminals of the electrolyzer using a programmable DC power supply unit having constant current and voltage mode provisions. Hydrogen production rate was calculated from the cell current and was cross examined by gas volume measurement. The purity of hydrogen gas was checked by using a calibrated hydrogen purity analyzer (NOVA Gas Analyzer). 3. Results and discussions 3.1. Physical characterizations 3.1.1. XRD XRD patterns of different ratios of Pt–Pd layered Nafion 117 membranes and pure Nafion 117 membrane are shown in Fig. 3. The samples showed the crystalline characteristic peaks around at 2θ = 40, 47 and 68° which represented (1 1 1), (2 0 0), and (2 2 0) lattice planes, respectively, of face-centered cubic (fcc) structure of Pt and Pd [23]. A broad hump observed at 17° revealed the amorphous nature of Nafion membrane [24]. It was confirmed that the face centered cubic structure
Fig. 2. MEA with gasket.
18
S.P. Sethu et al. / Solid State Ionics 283 (2015) 16–20
Fig. 4. Cell voltage–current density characteristics of Pt–Pd composite MEA with different ratios and normal MEA at 80 °C.
Fig. 3. X-ray diffraction analysis of prepared membranes.
(fcc) of Pt and Pd was overlapped in all the samples. However, there was no characteristic peak of metallic or oxide form of Pd observed, but their presence could not be discarded because Pd might be presented in Pt lattice and formed alloy or even in an amorphous form. Similar observations were obtained for PtRu based catalyst layered systems by Naoko Fujiwara et al. [19,25]. From the XRD measurements, the crystallite size of the Pt–Pd catalyst was calculated from the broadening of the (1 1 1) diffraction peak using Debye–Scherer's relation. The calculated crystallite size, atomic and weight ratios are given in Table 1. It was observed that the crystallite was size increased with Pd content. Surface image and cross sectional view of the Pt–Pd (70:30) bimetallic layered Nafion® 117 membranes have been elaborately discussed in our previous work [22]. 3.2. Electrochemical characterizations The Nafion membrane was made of hydrophobic back-bone structure and a hydrophilic side chain. In a well-hydrated membrane, there will be approximately 20 water molecules for each SO3 side chain [16, 26]. Therefore, the impregnation solutions of Pt and Pd occupy the hydrated regions and precipitated in the hydrophilic region after adding the reducing agent NaBH4 then reduce the concentration of water molecules, which then decreases the proton conductivity and methanol crossover. Hence, it is important to determine the optimum Pt–Pd composition and catalyst loading. To optimize the Pt–Pd ratios and catalyst loading, the single cell performance was evaluated. Figs. 4 and 5 show the voltage–current density characteristics of the single cell for to optimize the Pt–Pd ratios and catalyst loading, respectively at 80 °C with 4 M of methanol as a fuel. It was noted from Fig. 4 that the various Pt:Pd atomic ratios in the Pt–Pd composite membranes had a significant effect on the cell performances. For a quick overview, the performance of the single cell containing Pt:Pd (70:30) nanocomposite membrane gave better performance than the other ratios and normal MEA under identical test conditions that might be due to the reduced methanol
crossover and better methanol oxidation catalytic activity of Pt–Pd bimetallic catalyst [27–29]. The Pt:Pd (50:50) nanocomposite membrane showed lower performance than the other ratios and that might be due to higher nanoparticle crystalline size [30]. However, detailed studies of the catalyst analysis and origin of the performance enhancement of Pt–Pd (70:30) will be reported in the future. Fig. 5 shows current density vs voltage curves of above better performed Pt:Pd (70:30)/Nafion nanocomposite membrane. The loading amount of Pt–Pd was 1, 2 and 3 mg cm−2 corresponding to one, two and three repetitions of NEIR method, respectively. From the plot, it was noted that the performance of the single cell increased with Pt–Pd loading up to 2 mg cm − 2 , above which the cell exhibited lower performance. Among the various loadings, 2 mg cm− 2 loaded membrane delivered superior performance than the other samples due to an increase in methanol oxidation rate by increasing the catalyst loading and the decreasing level of methanol crossover. However, further increase in catalyst loading causes Pt–Pd particle isolation and leads to a lack of three-phase zone which reduced the cell performance. Moreover, mentioned earlier increased Pt–Pd depositions occupy the hydrated regions and reduced the proton conductivity, which also lowered the cell performances. The effect of catalyst loading on mass activity is shown in Fig. 6. An increase of mass activity up to −2 1.48 A mg−1 catalyst loading, compared Pt:Pd was observed for 1 mg cm −1 with 0.8 and 0.44 A mgPt:Pd for 2 and 3 1 mg cm− 2 respectively (at 0.9 V). The mass activity increased when the catalyst loading decreased which might be due to high electrochemical surface area and high catalyst utilization [31]. However, due to ohmic and mass transport losses within the electrode matrix, higher electrochemical surface area did not increase the electrolyzer performance in our case [32].
Table 1 XRD and EDX characteristics of the prepared membranes. Metal compositions in solution (at. %)
XRD data
EDX data
Crystallite size (nm)
Metal compositions (at.%)
Metal compositions (wt.%)
Pt–Pd (90:10) Pt–Pd (70:30) Pt–Pd (50:50)
5.4 5.7 6.7
94.61:5.39 79.09:20.91 62.55:37.45
96.98:3.02 87.40:12.60 75.39:24.61
Fig. 5. Effect of Pt–Pd loading on the performance of electrolyzer at 80 °C with 4 M methanol.
S.P. Sethu et al. / Solid State Ionics 283 (2015) 16–20
19
Fig. 6. Effect of Pt–Pd loading on mass activity of MEAs at 80 °C with 4 M methanol. Fig. 8. Effect of operating temperature with Pt–Pd (70:30) nanocomposite membrane at 80 °C with 4 M methanol.
3.3. Operating parameter optimizations Fig. 7 shows the comparative study of Pt and Pt–Pd (70:30) catalysts with various methanol concentrations. Electrolysis of aqueous methanol–water solution was carried out at a cell voltage of 1 V, in a PEM electrolyzer cell of active area 10 cm2, using Nafion-117 membrane at 80 °C. Hydrogen and carbon dioxide gases were generated at cathode and anode, respectively. The concentration of methanol was varied from 2 to 10 M to study the effect of methanol concentration on current density, which was directly related to the rate of hydrogen production. From the results it was noted an increase in current density with methanol concentration, however, a sharp increase in current density was observed up to about 6 M methanol concentration and from 6 to 10 M, the incremental gain was low due to the reason that the variation of current density with methanol concentration might be correlated to the change of membrane conductivity with methanol concentration. Nafion membrane was described as a series of clusters interconnected by narrow pores. In each cluster, the fixed membrane charges extended inward the center of each sphere. Due to the lower polarity of methanol compared to water, as the aqueous methanol concentration was increased, the cluster size was increased and hence the ionic conductivity of the membrane decreased, resulting in lower cell performance at high methanol concentrations [2,33,34]. However, Pt–Pd composite membrane was performed better than normal membrane which is mainly due to the reduction in methanol crossover. Moreover, the Pt–Pd alloy catalyst with a suitable ratio has also been exhibit very good resistance against CO poisoning effect [35]. The synergistic effect between Pt and Pd enhances the electrocatalytic activity towards methanol oxidation reaction. The presence of Palladium dilutes the Pt sites and prevents
the presence of three adjacent sites that are necessary for the adsorption of poison species on the electrode surface. This leads to the overall enhancement of methanol oxidations [28,29,36] Fig. 8 shows the effect of the reaction temperature on the current density vs voltage by feeding 4 M of methanol to the anode side. The current density was found to increase with temperature. The observed improvement in cell performance with temperature may be attributed to the enhanced kinetics of the electrochemical reaction and the decreased ohmic resistance. For a quick overview Pt–Pd (70:30)/Nafion nanocomposite membrane anode in the electrolyzer delivered a maximum current density value of 1.9 A cm−2 at 1 V with the operation at 80 °C and 4 M of methanol concentration. The optimized single cell performance with hydrogen production rate is shown in Fig. 9. The maximum volumetric hydrogen production rate is directly proportional to the current density and it can be explained in Eq. (4) below,
Fig. 7. Effect of methanol concentration on current density of methanol electrolyzer at 1.0 V and at 80 °C.
Fig. 9. Optimized electrolyzer performance and hydrogen production rate for Pt–Pd (70:30)/Nafion nanocomposite MEA at 80 °C with 4 M methanol.
Number of hydrogen moles ¼
ð4Þ
ðcurrent in amperesÞðtime in secondsÞ : ðunsigned numeric charge on the ionÞðFaradayÞ
Fig. 10 compares the Faradaic hydrogen production rate and the real selectivity of hydrogen from the experiment (methanol conversion) for Pt–Pd (70:30)/Nafion nanocomposite membrane. The results confirmed that the Faradaic current efficiencies for hydrogen production were near to ≈ 97–99%. The purity of the hydrogen gas was then checked using a Nova hydrogen analyzer (Nova Analytical Systems,
20
S.P. Sethu et al. / Solid State Ionics 283 (2015) 16–20
Fig. 10. Experimental and Faradaic hydrogen production rate comparison.
USA) that indicated 99% hydrogen purity which is highly compatible for fuel cell applications. This simplifies the fuel cell with fuel processor concept for transport applications. 4. Conclusions In the present work, the bimetallic Pt–Pd impregnated Nafion nanocomposite polymer electrolyte membrane was successfully prepared and its performance on methanol electrolysis was evaluated. From the results of the single cell test, the Pt–Pd bimetallic/Nafion nanocomposite membrane performed better than normal one due to the combined effect of reduced methanol crossover and better catalytic activity. Pt– Pd (70:30)/Nafion nanocomposite membrane anode in the electrolyzer delivered a maximum current density value of 1.9 A cm−2 at 1 V with 80 °C and 4 M methanol concentration. References [1] A. Caravaca, F.M. Sapountzi, A. de Lucas-Consuegra, C. Molina-Mora, F. Dorado, J.L. Valverde, Int. J. Hydrog. Energy 37 (2012) 9504–9513. [2] G. Sasikumar, A. Muthumeenal, S. Sundar Pethaiah, N. Nachiappan, R. Balaji, Int. J. Hydrog. Energy 33 (2008) 5905–5910.
[3] J. Nakamura Barbara, NASA Tech. Brief 26 (6) (2009). [4] Zhiyi Hu, Mei Wu, Zidong Wei, Shuqin Song, Pei Kang Shen, J. Power Sources 166 (2007) 458–461. [5] S.R. Narayaman, W. Chun, B.J. Nakamura, S. Marino, T.I. Valdez (2001) Hydrogen generation by electrolysis of aqueous organic solutions, U.S. Patent 6799744. [6] Tetsuo Take, Kazuhiko Tsurutani, Minoru Umeda, J. Power Sources 164 (2007) 9–16. [7] H. Sun, G. Sun, S. Wang, J. X Liu, X. Zhao, G. Wang, H. Xu, S. Hou, Q. Xin, J. Membr. Sci. 259 (2005) 27–33. [8] M.F. Hogarth, G.A. Hards, Platin. Met. Rev. 40 (1996) 150–159. [9] K. Scott, W. Taama, J. Cruickshank, J. Power Sources 65 (1997) 159–171. [10] Fangjian Shang, Lei Li, Yongming Zhang, Hong Li, J. Mater. Sci. 44 (2009) 4384–4388. [11] A.M. Zainoodin, S.K. Kamarudin, M.S. Masdar, W.R.W. Daud, A.B. Mohamad, J. Sahari, Appl. Energ. 113 (2014) 946–954. [12] Rong Zeng, Yiwang Chen, Weihua Zhou, Shuqin Xiao, Jichun Xiao, Caisheng Song, J. Mater. Sci. 45 (2010) 1610–1616. [13] Joon-Hee Kim, Min-Jee Yang, Jun-Young Park, Appl. Energ. 115 (2014) 95–102. [14] Tiago P. Almeida, Celina M. Miyazaki, Valdecir A. Paganin, Marystela Ferreira, Margarida J. Saeki, Joelma Perez, Antonio Riul Jr., Appl. Energ. 323 (2014) 7–12. [15] L. Brandao, J. Rodrigues, L.M. Madeira, A. Mendes, Int. J. Hydrog. Energy 35 (2010) 11561–11567. [16] E.H. Jung, U. H Jung, T.H. Yang, D.H. Peak, D.H. Jung, S.H. Kim, Int. J. Hydrog. Energy 32 (2007) 903–907. [17] R. Liu, P. Fedkiw PS, J. Electrochem. Soc. 139 (1992) 3514–3523. [18] A. Katayama-Aramata, R. Ohnishi, J. Am. Chem. Soc. 105 (1983) 658–659. [19] N. Fujiwara, K. Yasuda, T. Ioroi, Z. Siroma, Y. Miyazaki, Electrochim. Acta 47 (2002) 4079–4084. [20] S. Sundar Pethaiah, G. Sasikumar, S.H. Chan, U. Stimming, J. Power Sources 254 (2014) 61–167. [21] Raymond Liu, Wei-Hwa Her, Peter S. Fedkiw, J. Electrochem. Soc. 139 (1992) 15–23. [22] S. Sundar pethaiah, M. Ulaganathan, R.V. Mangalaraja, S.H. Chan, RSC Adv. 5 (2015) 981–987. [23] S.R. Yoon, G.H. Hwang, W.I. Cho, I.H. Oh, S.A. Hong, H.Y. Ha, J. Power Sources 106 (2002) 215–223. [24] M. Ludvigsson, J. Lindgren, J. Tegenfeldt, J. Electrochem. Soc. 147 (2000) 1303–1305. [25] H. Qiao, T. Kasajima, M. Kunimatsu, N. Fujiwara, T. Okada, J. Electrochem. Soc. 153 (2006) A42–A47. [26] S.R. Sammas, S. Wasmus, R.F. Savinell, J. Electrochem. Soc. 143 (1996) 1498–1504. [27] T. Hejze, B.R. Gollas, R.K. Sauerbrey, M. Schmied, F. Hofer, J.O. Besenhard, J. Power Sources 140 (2005) 21–27. [28] Jin Kim, Won Choon Choi, Won Choon Choi, Seong Ihl Woo, Won Hi Hong, Electrochim. Acta 49 (2004) 3227–3234. [29] T. Lopes, E. Antolini, E.R. Gonzalez, Int. J. Hydrog. Energy 33 (2008) 5563–5570. [30] Sarawalee Thanasilp, Mali Hunsom, Renew. Energy 36 (2011) 1795–1801. [31] Akos Kriston, Tianyuan Xie, David Gamliel, Prabhu Ganesan, Branko N. Popov, J. Power Sources 243 (2013) 958–963. [32] N. Rajalakshmi, H. Ryua, K.S. Dhathathreyan, Chem. Eng. J. 102 (2004) 241–247. [33] Anh Tuan Pham, Toshio Shudo, Int. J. Automot. Eng. 3 (2012) 125–130. [34] S. Song, G. Wang, W. Zhou, X. Zhao, G. Sun, Q. Xin, S. Kontou, P. Tsiakaras, J. Power Sources 140 (2005) 103–110. [35] D. Keqiang, W. Yahui, Y. Hongwei, Z. Chunbao, Yanli Cao, W. Huige, W. Yiran, G. Zhanhu, Electrochim. Acta 100 (2013) 147–156. [36] T. Arikan, A.M. Kannan, F. Kadirgan, Int. J. Hydrog. Energy 38 (2013) 2900–2907.