Repetitive bending test of membrane electrode assembly for bendable polymer electrolyte membrane fuel cell

G Model JIEC 3207 No. of Pages 6

Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Journal of Industrial and Engineering Chemistry journal homepage: www.elsevier.com/locate/jiec

Repetitive bending test of membrane electrode assembly for bendable polymer electrolyte membrane fuel cell Yun Sik Kanga,1, Taehyun Parka,b,1, Segeun Jangb,c , Mansoo Choib,c , Sung Jong Yooa,d,* , Suk Won Chab,* a

Fuel Cell Research Center, Korea Institute of Science and Technology (KIST), 5 Hwarang-ro 14-gil, Seongbuk-gu, Seoul 02792, Republic of Korea Department of Mechanical and Aerospace Engineering, Seoul National University, 1 Gwanak-ro, Gwanak-gu, Seoul 08826, Republic of Korea Global Frontier Center for Multiscale Energy Systems, 1 Gwanak-ro, Gwanak-gu, Seoul 08826, Republic of Korea d Clean Energy and Chemical Engineering, Korea University of Science and Technology, 217 Gajeong-ro, Yuseong-gu, Daejeon 34113, Republic of Korea b c

A R T I C L E I N F O

Article history: Received 21 September 2016 Received in revised form 15 November 2016 Accepted 30 November 2016 Available online xxx Keywords: Polymer electrolyte membrane fuel cell (PEMFC) Membrane electrode assembly (MEA) Bending fatigue Carbon cloth Bendable fuel cell

A B S T R A C T

Membrane electrode assemblies (MEAs) with carbon paper and cloth for bendable polymer electrolyte membrane fuel cell were characterized as it is subject to repetitive bending. The performance of the MEA with carbon paper was decreased significantly while the MEA with carbon cloth remained constant after repetitive bending. Electrochemical impedance spectroscopy revealed ohmic and charge transfer resistances of the MEA with carbon paper were increased by repetitive bending. Such performance degradation is due to physically observed damages in carbon paper and its detachment from the MEA, which was not in the MEA with carbon cloth due to its intrinsic flexibility. © 2016 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.

Introduction Polymer electrolyte membrane fuel cells (PEMFCs) are considered as one of the portable power sources of the future on account of their potential high energy density as well as high efficiency and scalability [1–6]. Although, the current technologies for miniaturization of PEMFCs are not yet fully matured, the aforementioned characteristics enable them to be used as portable auxiliary power sources [7–10]. Even if it may not be easy to employ a PEMFC as a direct power source in portable applications, this technology is undoubtedly a strong candidate for achieving a higher energy density than the current market-leading lithium-ion batteries. With this variation in the power sources, ‘flexible’ or ‘bendable’ electronic devices are emerging as interesting future electronics. The development of such flexible electronics has been motivated by their use in a variety of applications such as wearable and epidermal electronics [11–15]. Consequently, research on flexible electronics has been actively pursued for decades. Given that fuel cells are capable of attaining higher energy density than lithium-

* Corresponding authors. E-mail addresses: [email protected], [email protected] (S.J. Yoo), [email protected] (S.W. Cha). 1 These authors contributed equally to this work.

ion batteries, a ‘flexible and bendable fuel cell’ can be a part of future portable power sources. That is why there have been several studies on bendable fuel cells [16–22]. Among these, bendable fuel cells based on polydimethylsiloxane (PDMS) coated with Ag nanowires even showed usable power (>1 W) [23–27]. This suggests that the technological advancements of bendable fuel cells are close to commercialization. Interestingly, a characteristic common to all the bendable fuel cells was the employment of normal membrane electrode assemblies (MEAs), which are generally used in normal PEMFCs, without any modifications. Although a MEA is composed of a Nafion1 membrane, Pt/C catalyst layers, and gas diffusion layers (GDLs), which are types of bendable materials, each component could be damaged by repetitive bending. Here, a bendable fuel cell will be exposed to repetitive bending in a real operation environment. Thus, if normal MEAs need to be used in any bendable fuel cell templates, performance variation of the MEA under repetitive bending should be investigated. Furthermore, the issue with previous flexible PEMFCs was their fabrication using brittle carbon papers as GDLs. In this case, it is possible that such GDLs will get damaged under real operating conditions of flexible electronics, such as repetitive bending. Therefore, it is necessary to investigate the bending durability of carbon papers. However, to the best of our knowledge, such research has not yet been reported.

http://dx.doi.org/10.1016/j.jiec.2016.11.048 1226-086X/© 2016 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.

Please cite this article in press as: Y.S. Kang, et al., Repetitive bending test of membrane electrode assembly for bendable polymer electrolyte membrane fuel cell, J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.11.048

G Model JIEC 3207 No. of Pages 6

2

Y.S. Kang et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx

Thus, in this study we have investigated the performance variation of normal MEAs as they were subjected to repetitive bending cycles. As mentioned earlier, because carbon paper has been widely used as a GDL material, a MEA with carbon paper was first examined. A carbon cloth, which is another representative GDL material, was also tested, because being a kind of fabric, it is not brittle unlike carbon paper. This implies that the carbon cloth will be bent easily without any damages. Our results showed that the carbon-paper-based MEA was fragile to repetitive bending, while the carbon-cloth-based MEA showed good bending-durability. The difference in the behavior of the two MEAs was analyzed via electrochemical measurements and by observing digital camera images of a physically damaged carbon-paper-based MEA. Interestingly, the repetitive bending increased both the ohmic and charge transfer resistances of the carbon-paper-based fuel cell. It was observed that the damages to carbon paper by repetitive bending resulted from its detachment from the MEA and extrinsic deterioration due to the generation of macro-scale cracks. However, no noticeable damages were detected for carbon cloth, due to its softness. This result signifies the need for further development of carbon cloths as GDLs in PEMFCs, which have not been in use due to the high performance of PEMFCs with carbon paper. Experimental Flow-field plates were fabricated using PDMS. The weight ratio between PDMS and a curing agent was 10:1. It was solidified at 70  C for 4 h. This gel-state PDMS solution was poured over a specially designed mold in which the structure of flow-channels was made to protrude inversely so that the channels could be etched on the solidified PDMS. The etched flow-channels were 1 mm in both width and height. The width of the rib was also 1 mm. The flow-channels were of mixed parallel-serpentine style: three channels comprised one main reactant stream from inlet to outlet. The resulting thickness of a single PDMS plate was 7 mm. The reactive area defined by the flow-channels was 3 cm  3 cm. The MEAs were prepared as follows: first, a Nafion1 212 membrane (Dupont Co., United States) was employed as an electrolyte without pre-treatment. It was mounted on a suction-type hotplate. The temperature of the hot-plate was 80  C. The catalyst ink for the anode and cathode catalyst layers was prepared by mixing water, 5 wt.% Nafion1 solution (Dupont Co., United States), and isopropyl alcohol (IPA) (Sigma-Aldrich Co., USA) with 40 wt.% Pt/C (Johnson Matthey Co., United Kingdom). The prepared catalyst ink was blended by ultrasonic treatment and sprayed onto the anode and cathode sides of the mounted Nafion1 212 membrane. Pt loading of the as-fabricated catalyst-coated membrane (CCM) of each electrode was 0.45 mg/cm2 and the active geometric area of the MEAs was 9.0 cm2. After this process, the carbon papers (10BC, SGL Carbon Co., Germany) or carbon cloths (A-type cloth, E-Tek Inc., United States) were hot-pressed on this CCM. The temperature and pressing time were 110  C and 10 min, respectively. A Ti gauze (280 mm thick, Alfa Aesar Co., USA) was used as the current collector in the fabricated MEAs. The width of the Ti mesh was 3 cm and it covered the entire surface of the GDLs. The Ti mesh was placed between the PDMS plates and MEA as illustrated in Fig. 1. The as-fabricated fuel cell was mounted on a vise to bend it. The fuel cell performance was measured for four shapes, namely, flat and bent with bending radii of 83, 58, and 47 mm. The volumetric rate of both the hydrogen and air streams for the fuel cell was 3.33 cm3/s. They were humidified by flowing through water at room temperature, which corresponds to the relative humidity and temperature of 94% and 25  C, respectively. The single cell performance and the corresponding electrochemical impedance spectroscopy (EIS) data were measured using Solartron 1260/1287

Fig. 1. Architectural schematic of the MEA-bending test setup (bendable PEMFC).

(Solartron Analytical Co., United Kingdom) at room temperature (25  C). Before measuring the performance, it was activated by measuring current–voltage (I–V) curves repetitively. The I–V curves were measured in the voltage range from the open circuit voltage (OCV) to 0.25 V. Potentiodynamic mode was used to measure it. The voltage scan rate was 15 mV/s. Right after measuring the I–V curves, the corresponding EIS were investigated at 0.5 V vs. RHE. A sinusoidal input with an amplitude of 30 mV was applied and the resulting current response was monitored. The frequency range was 105–0.1 Hz, from high to low frequency and the corresponding ohmic and charge transfer resistances were calculated. After measuring the performance, the assembled MEA was removed from the fuel cell. It was then repetitively bent more than 100 times. The bending radius was <47 mm in this deliberate bending process (hereafter, ‘before’ and ‘after’ in all the figures refer to the MEA before and after repetitive bending test, respectively.). This bending-treated MEA was characterized again by following the same process as described above. Results and discussion The results of the repetitive bending test performed for the MEAs in the bendable PEMFCs are shown in Fig. 2. According to Fig. 2(a) and (b), the OCVs of the PEMFC with carbon cloth (CC-FC) are 0.89 V while the PEMFC with carbon paper (CP-FC) are 0.99 V. It has been commonly reported that the OCVs are typically more or less than 0.9 V if the PEMFCs are operated with ambient air at room temperature [28–31]. Accordingly, it can be inferred that sealing of the bendable PEMFCs is not problematic. In addition, there was no mass transport loss in the high current density range equivalent to low voltages of <0.3 V. Because of the high flow rates of hydrogen and air (3.33 cm3/s), they are abundant stoichiometrically, which ensures the sealability of the fuel cell. The OCVs of the CP-FC of all shapes are higher than those of the corresponding CC-FC before the repetitive bending test (Fig. 2(c) and (d)). We believe that this difference between the OCVs of the two types of PEMFCs resulted from the unoptimized structure and lower electric conductivity of carbon cloth. Recently, carbon papers have become the chief materials for GDLs in the MEAs of PEMFCs instead of carbon cloths that were previously used, since their material characteristics, such as microstructures, relative diffusivity, and electric conductivity are superior to those of carbon cloths [32–34]. That is why in the development of GDLs, the focus has been on carbon papers. The aforementioned poor characteristics of carbon cloths might lead to insufficient supply of

Please cite this article in press as: Y.S. Kang, et al., Repetitive bending test of membrane electrode assembly for bendable polymer electrolyte membrane fuel cell, J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.11.048

G Model JIEC 3207 No. of Pages 6

Y.S. Kang et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx

3

Fig. 2. The I–V curves of bendable PEMFCs with the MEAs with (a) carbon cloth and (b) carbon paper GDLs before and after repetitive bending for >100 times. (c) and (d) show the variation of the OCVs and maximum power densities in (a) and (b), respectively. ‘R’, ‘Before’, and ‘After’ in symbol captions mean bending radius, before repetitive bending, and after repetitive bending of the MEA, respectively.

the reactant to the triple-phase boundaries or poor collection of the generated electrons from the fuel cell reaction, resulting in the lower OCVs of the CC-FCs. It is believed that the overall performances are also related to these features of carbon cloths. That is why not only the OCVs but also the performances of the CC-FC are lower than those of the CP-FC. However, this was the case only when the MEAs were not subject to the repetitive bending load. As shown in Fig. 2, both the OCVs and performances of the bendable PEMFCs changed after the application of the repetitive bending load. Interestingly, both the OCV and performance of the CP-FC become lower than those of the CC-FC as it was bent repetitively. The maximum power density of the CP-FC before the repetitive bending test (35–70 mW/cm2) was lowered significantly (18–40 mW/cm2) after the repetitive bending test. However, the CC-FC retained its power density range (20– 43 mW/cm2) even after the repetitive bending test. This clearly means that the MEA with carbon cloth has greater durability with regard to bending than the MEA with carbon paper. This will be further discussed later together with the results of the corresponding EIS. In addition, the OCVs of the MEA with carbon cloth increased after the repetitive bending test. Contrary to this, the MEA with carbon paper showed significant decrease in OCVs after the repetitive bending test, due to its lower bendability and corresponding physical damages of carbon papers. For the MEA with carbon cloth, however, it seemed no damage on carbon cloth after the repetitive bending test and the corresponding MEA performances were maintained. In this regard, we suggest

3 reasons for the increased OCVs of the MEA with carbon cloth after the repetitive bending test. First, due to the repetitive single cell testing, Pt catalyst for each electrode can be activated and the humidity of the MEAs can be increased from water molecules by cathode reaction. The repetitive single cell testing can clean the surface of Pt nanoparticles and generate excess water molecules, which can be absorbed in MEA and increase MEA performance and OCV, as a result. It can be the first reason for the higher OCV values. Second, the generation of cracks in catalyst layers during the repetitive bending test can increase the MEA performance and OCV. From the previous research about stretched catalyst-coated MEA, cracked Pt electrode was generated and exhibited improved single cell performance and higher OCV values than normal MEA due to the improvement in mass transport [35]. Lastly, it is commonly believed that the OCV values be easily affected by the assembly of cell components and corresponding gas sealing of each electrode. If each electrode is not separated well, the gas crossover can occur and the resulting mixed potential can deteriorate single cell performance [3]. In other words, the degree of cell assembly and the gas sealing of the MEA with carbon cloth could be improved after the repetitive bending test because there are no physical damages on carbon cloth, contrary to carbon paper. Another interesting behavior of the performance variation is that the overall power density increased with the increased bending in both CC-FC and CP-FC. It has been already proposed that the increasing tendency of the performance in a bendable fuel cell stems from the increased compressive stress on the MEA caused by the

Please cite this article in press as: Y.S. Kang, et al., Repetitive bending test of membrane electrode assembly for bendable polymer electrolyte membrane fuel cell, J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.11.048

G Model JIEC 3207 No. of Pages 6

4

Y.S. Kang et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx

(a)

(b) -9

Carbon cloth, 0.5 V vs. RHE, operated at RT

R= ∞, Before R= 83 mm, Before R= 58 mm, Before R= 47 mm, Before R= ∞, After R= 83 mm, After R= 58 mm, After R= 47 mm, After

2

ZIm (Ohm cm )

-7 -6

RΩ

CPE

-5

Rct

-4

Carbon paper, 0.5 V vs. RHE, operated at RT

-8

R= ∞, Before R= 83 mm, Before R= 58 mm, Before R= 47 mm, Before R= ∞, After R= 83 mm, After R= 58 mm, After R= 47 mm, After

-7 2

-8

ZIm (Ohm cm )

-9

-3 -2 -1

-6

CPE

RΩ

-5

Rct

-4 -3 -2 -1

0

0 0

1

2

3

4

5

6

7

8

9

10

0

1

2

2

3

4

5

6

7

8

9

10

2

ZRe (Ohm cm )

ZRe (Ohm cm )

Fig. 3. EIS of (a) CC-FC and (b) CP-FC corresponding to Fig. 2(a) and (b) at 0.5 V vs. RHE, respectively.

bending [23,24,26,27]. Although, the platform for the bendable fuel cells here is not exactly the same as in literature (Ti gauze vs. Ag nanowires percolation current collectors), according tothis research and as discussed elsewhere [23,24,26], the performance enhancement caused by the bending in bendable PEMFCs is related to the basic structure of the MEA and not linked to the current collectors. The poor bending durability of the MEA with carbon paper is further visualized clearly in the EIS in Fig. 3(a) and (b). Based on

(a)

5

Fig. 2(a), it is reasonable to estimate that there will be almost no difference between the EIS of the CC-FC before and after repetitive bending. In agreement with this estimation, it can be seen from Fig. 3(a) that the diameters of the half-circles in the EIS spectra before and after repetitive bending with a particular bending radius remain almost overlapped. On the contrary, it is apparent that not only the diameter but also the first (the left) intercepts to ZRE-axis of the half-circles of the CP-FC increased after repetitive

(b) Carbon cloth, 0.5 V vs. RHE

3

2

1

(c)

7

R= ∞

R= 83 30 mm

R= 58 mm

3

2

0

R= 47 mm

(d) Carbon cloth, 0.5 V vs. RHE

Before After

7

R= ∞

R= 83 mm

R= 58 mm

Carbon paper, 0.5 V vs. RHE

6

R= 47 mm

Before After

5

Rct (ohm cm2)

5

Rct (ohm cm2)

Before After

1

6

4 3 2 1 0

Carbon paper, 0.5 V vs. RHE

4

Rohm (ohm cm2)

Rohm (ohm cm2)

4

0

5

Before After

4 3 2 1

R= ∞

R= 83 mm

R= 58 mm

R= 47 mm

0

R= ∞

R= 83 mm

R= 58 mm

R= 47 mm

Fig. 4. Comparisons of the ohmic resistances of the (a) CC-FC and (b) CP-FC before and after repetitive bending test. (c, d) Same comparisons as (a) and (b) but of the charge transfer resistances.

Please cite this article in press as: Y.S. Kang, et al., Repetitive bending test of membrane electrode assembly for bendable polymer electrolyte membrane fuel cell, J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.11.048

G Model JIEC 3207 No. of Pages 6

Y.S. Kang et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx

bending test. In the above discussion, the first intercept to ZRE-axis and the diameter of the half-circle in EIS can be regarded as the ohmic and charge transfer resistances, respectively, if the EIS is matched to the equivalent circuit as shown in the insets of Fig. 3(a) and (b) [36,37]. If the variations of the ohmic and charge transfer resistances presented in Fig. 3(a) and (b) are quantitatively extracted from the EIS, the effect of repetitive bending can be visualized more clearly as shown in Fig. 4. According to Fig. 4(a) and (b), the ohmic resistance in the CC-FC remained almost constant, while it increased significantly in the case of the CP-FC. Even the variation tendency of the charge transfer resistances (see Fig. 4(c) and (d)) in the CC-FC and CP-FC is same as the variation of the ohmic resistances. This implies that the polarization characteristics shown in Fig. 2(b), in which the performance of the CP-FC decreased with repetitive bending, resulted from the above mentioned increase in both ohmic and charge transfer resistances. An intriguing behavior exhibited in Fig. 4(a) and (c) is that the ohmic and charge transfer resistances obtained with an infinite bending radius are somewhat different from the values derived from other shapes. The ohmic resistance is higher after the repetitive bending test, while the charge transfer resistance is higher before the repetitive bending test. This inconsistency in the variation of the resistances did not result in a big difference in the cell performances. It is speculated that such behavior was induced by the unstable compression caused by the flow-field plates. Although, an infinite bending radius of the PEMFC implies that it is flat, there is actually a residual compression by just the assembly of two PDMS flow-field plates because total thickness of the sealant applied is thinner than that of the MEA, GDLs, and Ti meshes altogether, thereby leading to the slightly bent shape of the flow-

5

field at each side, i.e. anode and cathode side. Accordingly, this would generate a not precisely controlled compression, leading to fluctuations in the performance and in the corresponding ohmic and charge transfer resistances. A similar phenomenon has been observed in the other types of bendable fuel cells described elsewhere [23–25]. In spite of the bending durability of the MEA with carbon paper, the quantitative values of the resistances obtained herein are still higher than those reported (<0.08 and <0.28 V cm2 vs. >0.9 and >1.2 V cm2 of ohmic and charge transfer resistances, respectively) in the literature [38], where they fabricated similar MEAs as used in this work, but with even lower Pt loading (0.12 mg/cm2). Considering the measured voltage (0.6 V vs. RHE) and the Pt loading, the ohmic and charge transfer resistances should be higher than here, however, it is not so. It can, therefore, be inferred that the conventional architecture of the MEAs is not optimized for a flexible and bendable system like the fuel cell template employed in this research. There are several parameters that can influence the ohmic and charge transfer resistances, namely, architecture of the flow-channels, assembly pressure (compressive force) on the MEA, and a resulting optimal pressure on the ribs in the flowchannels and well-defined triple-phase boundaries in the catalyst layer [1,3,39]. The aforementioned resistances (<0.08 and <0.28 V cm2) are supposed to be the results from the other optimized parameters listed above. Thus, for an absolute enhancement of the performance of a bendable PEMFC, it is necessary to re-design the MEA and other parts as well. Because the repetitive bending test is a type of mechanical durability and stability test, the mechanical, i.e. componential damages of the MEA would cause such increase in the ohmic and charge transfer resistances. According to Fig. 5(a) and (b),

Fig. 5. Digital camera images of the MEAs composed of (a) carbon cloths and (b) carbon papers before repetitive bending test. (c) and (d) are the images of (a) and (b) after repetitive bending test, respectively. Insets are the top-view images of the MEAs before and after repetitive bending test.

Please cite this article in press as: Y.S. Kang, et al., Repetitive bending test of membrane electrode assembly for bendable polymer electrolyte membrane fuel cell, J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.11.048

G Model JIEC 3207 No. of Pages 6

6

Y.S. Kang et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx

irrespective of the GDL material (carbon cloth and paper), the GDLs are attached well to Nafion1. While the MEA with carbon cloth maintained its structure, in CP-FC the carbon paper is detached from the MEA after the repetitive bending test (Fig. 5(c) and (d)). In addition, the top-view images of the MEAs showed that after the repetitive bending test, many cracks were generated on the surface of the carbon paper, but not on top of the carbon cloth. (Insets of Fig. 5(a)–(d)) Because ohmic resistance is a function of the electrical conductivity of the material itself and of the contact resistance between every component, both the phenomena, i.e. detachment of the carbon paper GDLs from the MEA and crackgeneration inside the carbon paper GDLs, are believed to increase the ohmic resistance in the CP-FC. In other words, such bending fatigue will not affect the carbon cloth negatively since they are intrinsically bendable and even foldable, while it will do so to the carbon papers due to the fold lines as well as cracks on the microporous layers and carbon paper layers. These factors would decrease the electrical conductivity and increase the ohmic resistance. Likewise, it is speculated that such fold lines and cracks would also damage the triple-phase boundaries and thereby decrease the reactive area that is related to the charge transfer resistance. This illustrates that the classical structure of the MEA with a carbon-paper-based GDL is susceptible to bending and is not stable or durable in bendable fuel cells, while a carbon-cloth-based GDL retains its mechanical durability and results in a stable performance. Although recent research has been focused on MEAs with carbon paper, it is necessary to develop carbon-cloth GDLbased MEAs for further development of bendable fuel cells in order to cope with the demands of future electronics. Conclusion MEAs with carbon-paper- and carbon-cloth-based GDLs for PEMFCs were fabricated and characterized as they were subjected to repetitive bending more than 100 times. The results showed that the MEA with carbon cloth exhibited lower but stable performance with good bending durability, while the MEA with carbon paper showed significant performance degradation after the repetitive bending test at all the bending radii tested. By investigating via EIS, it was found that both the ohmic and charge transfer resistances increased with repetitive bending of the CP-FC. However, these resistances did not change even after bending of the CC-FC, thereby showing good bending durability. The performance degradation of the carbon-paper-based MEA resulted from damaged components such as cracks and fold lines. Accordingly, it is suggested that further development of carbon cloth and the MEA thereof is required in order to realize high performance and highly durable, bendable PEMFCs. Acknowledgments This work was supported by the Global Frontier R&D Program on Center for Multiscale Energy System funded by NRF (2016M3A6A7945505, 2015M3A6A7065442) and the NRF grant funded by MSIP (2014R1A2A2A04003865). This work was also supported by the New & Renewable Energy Core Technology Program of KETEP, granted financial resource from MOTIE, Republic of Korea (20143030031340,20133030011320). The IAMD at Seoul National University and BK21 plus is also acknowledged for their partial support. References [1] F. Barbir, Pem Fuel Cells: Theory and Practice, 2nd ed., Academic Press, Massachusetts, 2012.

[2] F. Migliardini, O. Veneri, P. Corbo, J. Ind. Eng. Chem. 17 (3) (2011) 633, doi: http://dx.doi.org/10.1016/j.jiec.2011.05.011. [3] R. O’Hayre, S.-W. Cha, W. Colella, F.B. Prinz, Fuel Cell Fundamentals, 2nd ed., John Wiley & Sons, New York, 2009. [4] S. Kim, I. Hong, J. Ind. Eng. Chem. 14 (3) (2008) 357, doi:http://dx.doi.org/ 10.1016/j.jiec.2008.01.007. [5] S. Dutta, J. Ind. Eng. Chem. 20 (4) (2014) 1148, doi:http://dx.doi.org/10.1016/j. jiec.2013.07.037. [6] D.J. Kim, M.J. Jo, S.Y. Nam, J. Ind. Eng. Chem. 21 (2015) 36, doi:http://dx.doi.org/ 10.1016/j.jiec.2014.04.030. [7] T. Henriques, B. César, P.J.C. Branco, Appl. Energy 87 (4) (2010) 1400, doi:http:// dx.doi.org/10.1016/j.apenergy.2009.09.001. [8] J.H. Wee, J. Power Sources (2006) 1, doi:http://dx.doi.org/10.1016/j.jpowsour.2006.07.032. [9] C.K. Dyer, J. Power Sources 106 (1–2) (2002) 31, doi:http://dx.doi.org/10.1016/ S0378-7753(01)01069-2. [10] P. Agnolucci, Int. J. Hydrogen Energy 32 (17) (2007) 4319, doi:http://dx.doi.org/ 10.1016/j.ijhydene.2007.03.042. [11] C.-W. Sher, K.-J. Chen, C.-C. Lin, H.-V. Han, H.-Y. Lin, Z.-Y. Tu, H.-H. Tu, K. Honjo, H.-Y. Jiang, S.-L. Ou, R.-H. Horng, X. Li, C.-C. Fu, H.-C. Kuo, Opt. Express 23 (19) (2015) A1167, doi:http://dx.doi.org/10.1364/OE.23.0A1167. [12] S.-Y. Lee, K.-H. Choi, W.-S. Choi, Y.H. Kwon, H.-R. Jung, H.-C. Shin, J.Y. Kim, Energy Environ. Sci. 6 (8) (2013) 2414, doi:http://dx.doi.org/10.1039/ c3ee24260a. [13] S. Lam Po Tang, Trans. Inst. Meas. Control 29 (2007) 283, doi:http://dx.doi.org/ 10.1177/0142331207070389. [14] J.J. Boland, Nat. Mater. 9 (10) (2010) 790, doi:http://dx.doi.org/10.1038/ nmat2861. [15] A. Nathan, A. Ahnood, M.T. Cole, S. Lee, Y. Suzuki, P. Hiralal, F. Bonaccorso, T. Hasan, L. Garcia-Gancedo, A. Dyadyusha, S. Haque, P. Andrew, S. Hofmann, J. Moultrie, D. Chu, A.J. Flewitt, A.C. Ferrari, M.J. Kelly, J. Robertson, G.A.J. Amaratunga, W.I. Milne, Proc. IEEE 100 (2012) 1486. [16] C. Weinmueller, G. Tautschnig, N. Hotz, D. Poulikakos, J. Power Sources 195 (12) (2010) 3849, doi:http://dx.doi.org/10.1016/j.jpowsour.2009.12.092. [17] F.-K. Hsu, M.-S. Lee, C.-C. Lin, Y.-K. Lin, W.-T. Hsu, J. Power Sources 219 (2012) 180, doi:http://dx.doi.org/10.1016/j.jpowsour.2012.07.054. [18] J. Wheldon, W.-J. Lee, D.-H. Lim, A.B. Broste, M. Bollinger, W.H. Smyrl, Electrochem. Solid-State Lett. 12 (5) (2009) B86, doi:http://dx.doi.org/10.1149/ 1.3095682. [19] S. Tominaka, H. Nishizeko, S. Ohta, T. Osaka, Energy Environ. Sci. 2 (8) (2009) 849, doi:http://dx.doi.org/10.1039/b906216e. [20] S. Tominaka, H. Nishizeko, J. Mizuno, T. Osaka, Energy Environ. Sci. 2 (10) (2009) 1074, doi:http://dx.doi.org/10.1039/b915389f. [21] R. Hahn, S. Wagner, A. Schmitz, H. Reichl, J. Power Sources 131 (1–2) (2004) 73, doi:http://dx.doi.org/10.1016/j.jpowsour.2004.01.015. [22] I. Chang, M.H. Lee, J.H. Lee, Y.S. Kim, S.W. Cha, Int. J. Precis. Eng. Manuf. 14 (3) (2013) 501, doi:http://dx.doi.org/10.1007/s12541-013-0067-1. [23] I. Chang, T. Park, J. Lee, M.H. Lee, S.H. Ko, S.W. Cha, J. Mater. Chem. A 1 (30) (2013) 8541, doi:http://dx.doi.org/10.1039/c3ta11699a. [24] I. Chang, T. Park, J. Lee, H.B. Lee, S. Ji, M.H. Lee, S.H. Ko, S.W. Cha, Int. J. Hydrogen Energy 39 (14) (2014) 7422, doi:http://dx.doi.org/10.1016/j.ijhydene.2014.03.017. [25] T. Park, I. Chang, J. Lee, S.H. Ko, S.W. Cha, ECS Trans. 64 (3) (2014) 927, doi: http://dx.doi.org/10.1149/06403.0927ecst. [26] I. Chang, T. Park, J. Lee, H.B. Lee, S.H. Ko, S.W. Cha, Int. J. Hydrogen Energy 41 (14) (2016) 6013, doi:http://dx.doi.org/10.1016/j.ijhydene.2016.02.087. [27] T. Park, I. Chang, H.B. Lee, S.H. Ko, S.W. Cha, Int. J. Hydrogen Energy (2016), doi: http://dx.doi.org/10.1016/j.ijhydene.2016.08.022 in press. [28] S. Litster, G. McLean, J. Power Sources (2004) 61, doi:http://dx.doi.org/10.1016/ j.jpowsour.2003.12.055. [29] V. Mehta, J.S. Cooper, J. Power Sources (2003) 32, doi:http://dx.doi.org/ 10.1016/S0378-7753(02)00542-6. [30] S.A. Vilekar, R. Datta, J. Power Sources 195 (8) (2010) 2241, doi:http://dx.doi. org/10.1016/j.jpowsour.2009.10.023. [31] J. Zhang, Y. Tang, C. Song, J. Zhang, H. Wang, J. Power Sources 163 (1) (2006) 532, doi:http://dx.doi.org/10.1016/j.jpowsour.2006.09.026. [32] Y. Wang, C.Y. Wang, K.S. Chen, Electrochim. Acta 52 (12) (2007) 3965, doi: http://dx.doi.org/10.1016/j.electacta.2006.11.012. [33] S. Park, B.N. Popov, Fuel 90 (1) (2011) 436, doi:http://dx.doi.org/10.1016/j. fuel.2010.09.003. [34] V. Radhakrishnan, P. Haridoss, Mater. Des. 32 (2) (2011) 861, doi:http://dx.doi. org/10.1016/j.matdes.2010.07.009. [35] S.M. Kim, C.-Y. Ahn, Y.-H. Cho, S. Kim, W. Hwang, S. Jang, S. Shin, G. Lee, Y.-E. Sung, M. Choi, Sci. Rep. 6 (2016) 26503, doi:http://dx.doi.org/10.1038/ srep26503. [36] X. Yuan, H. Wang, J. Colin Sun, J. Zhang, Int. J. Hydrogen Energy (2007) 4365, doi:http://dx.doi.org/10.1016/j.ijhydene.2007.05.036. [37] C. Lai, J. Lin, K. Hsueh, C. Hwang, K. Tsay, L. Tsai, Y. Peng, Int. J. Hydrogen Energy 32 (17) (2007) 4381, doi:http://dx.doi.org/10.1016/j.ijhydene.2007.03.040. [38] S.M. Kim, Y.S. Kang, C. Ahn, S. Jang, M. Kim, Y.E. Sung, S.J. Yoo, M. Choi, J. Power Sources 317 (2016) 19, doi:http://dx.doi.org/10.1016/j.jpowsour.2016.03.083. [39] S. Kim, I. Hong, J. Ind. Eng. Chem. 13 (5) (2007) 864, doi:http://dx.doi.org/ 10.1243/0954406JMES654.

Please cite this article in press as: Y.S. Kang, et al., Repetitive bending test of membrane electrode assembly for bendable polymer electrolyte membrane fuel cell, J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.11.048