- Email: [email protected]
ionic-liquid-graphite-oxide composite membrane for high temperature polymer electrolyte membrane fuel cells
Journal of Power Sources 274 (2015) 922e927
Contents lists available at ScienceDirect
Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour
A polybenzimidazole/ionic-liquid-graphite-oxide composite membrane for high temperature polymer electrolyte membrane fuel cells Chenxi Xu a, b, 1, Xiaoteng Liu b, 1, Jigui Cheng a, *, Keith Scott b, ** a b
School of Materials Science and Engineering, Hefei University of Technology, Hefei, Anhui, China School of Chemical Engineering and Advanced Materials, Newcastle University, Newcastle NE1 7RU, United Kingdom
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
1-(3-Aminopropyl)-3-methylimidazolium groups (ionic liquid) was bonded to the graphite oxide. ILGO filler enhanced the proton conductivity of PBI membrane. ILGO/PBI membrane showed superior performances in HT-PEMFC.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 3 October 2014 Received in revised form 20 October 2014 Accepted 21 October 2014 Available online 28 October 2014
Graphite oxide is successfully functionalised by 3-aminopropyltriethoxysilane ionic liquid and used as a filler material in a polybenzimidazole (PBI) membrane for high temperature proton exchange membrane fuel cells. The ionic-liquid-graphite-oxide/polybenzimidazole (ILGO/PBI) composite membrane exhibits an appropriate level of proton conductivity when imbibed with phosphoric acid at low phosphoric acid loading, which promotes its use in fuel cells by avoiding acid leakage and materials corrosion. The ionic conductivities of the ILGO/PBI membranes at 175 C are 0.035 S cm1 and 0.025 S cm1 at per repeat units of 3.5 and 2.0, respectively. The fuel cell performance of ILGO/PBI membranes exhibits a maximum power density of 320 mW cm2 at 175 C, which is higher than that of a pristine PBI membrane. © 2014 Published by Elsevier B.V.
Keywords: Ionic liquid graphite oxide Polybenzimidazole Polymer electrolyte membrane fuel cells
1. Introduction In the last decades, considerable efforts have been made to develop high temperature (>100 C) proton exchange membrane fuel cells (PEMFCs) using polymer acid complexes (PACs) because it offers significant advantages in this temperature range, such as (1)
* Corresponding author. ** Corresponding author. E-mail addresses: [email protected] (C. Xu), [email protected] (J. Cheng). 1 These authors contributed equally to this work. http://dx.doi.org/10.1016/j.jpowsour.2014.10.134 0378-7753/© 2014 Published by Elsevier B.V.
improved CO tolerance, (2) enhanced efficiency, (3) avoidance of flooding by water, (4) opportunity to use non-noble metal catalysts, and (5) system simplification [1e5]. Solid-state electrolytes loaded with phosphoric acid provide less corrosion and more immobilisation compared to aqueous phosphoric acid fuel cells (PAFCs) [6e8]. The phosphoric-acid loaded polybenzimidazole (PBI) is the best-known example of a membrane; PBI has been used to produce reasonably successful membranes for fuel cells with excellent thermo-chemical stability and good conductivity [6,7]. However, in many cases, PBI/H3PO4 membranes exhibit high conductivity values only with high acid
C. Xu et al. / Journal of Power Sources 274 (2015) 922e927
loadings, which are usually at least higher than 5.0H3PO4 molecules per repeat unit (PRU) of PBI [8]. Such a high PA acid content creates problems of mechanical strength reduction and elution problems of electrolytes, as well as catalyst corrosion problems associated with the use of an excess of highly concentrated PA at high temperature [9,10]. One promising idea to reduce the loading level is to fill solid proton conductors into the PBI membrane, such as solid acids, ionic liquid and graphite oxide [10e12]. Graphite oxide (GO) consists of carbon, oxygen and hydrogen and is exfoliated into individual graphite oxide nanoplatelets in water because of the hydrophilic oxygen groups attached to the graphene sheets [13]. The presence of acidic groups of GO, such as carboxylic acid and epoxy oxygen, could facilitate hopping of protons [14e16]. According to our previous work, the sulfonic acid groups functionalised to GO could improve water retention ability and provide hydrogen bond for the conductivity increase [17]. So, the sulphonic GO with PBI exhibited improved performance at low PA loading level. Functionalised graphite oxide (FGO) is also easy to disperse in a solvent, resulting in a homogeneous distribution into polymer matrices; in addition, the functional groups could improve the proton conductivity. Hence, the functionalised GO is a suitable candidate filler material for use in membranes to improve the proton conductivity and avoid excessive content of PA. Recently, functionalised GO was used in composite membranes and has attracted much attention. Functionalised groups, such as 3mercaptopropyl trimethoxysilane (MPTMS) [15], added into graphite oxide were incorporated with the perfluorinated polymer Nafion®. The functionalised graphite controlled the state of water by means of nanoscale manipulation of the physical geometry and the chemical functionality of ionic channels. The confinement of bound water within the reorganised nanochannels of composite membranes enhanced the proton conductivity at high temperature via the low activation energy for ionic conduction [16]. Additionally, the increase of FGO nanofiller loading extended the number of available ion exchange sites per cluster, resulting in the increased proton mobility in the membrane at higher temperature and lower humidity [15]. Ionic liquids (ILs) are organic salts with high ionic conductivity and wide electrochemical potential windows. ILs could improve safety and enable a higher operating temperature range to be used [18,19]; therefore, they were considered as an additive in the proton exchange membrane. According to the oxygen group located in GO, the ILs could be readily bonded to the GO. The combination of GO and IL may enhance the PBI membrane conductivity because both the ILs and GO may provide hydrogen bond as the proton transport route [20]. In this work, the GO sheets were functionalised by an ionic liquid group 1-(3-aminopropyl)-3-methylimidazolium (ILGO) to combine with the PBI membrane loaded with phosphoric acid. The membrane proton conductivity and fuel cell performance were studied, and the results indicated that the performance of the PBI/ H3PO4 membrane was enhanced with the addition of ILGO. 2. Experimental methods A graphite oxide (GO) sheet was synthesised from graphite (Shandong Qingdao graphite company, China) by oxidation with KMnO4 in concentrated H2SO4 according to the Hummers method [21]. A graphite sheet, was placed in cold (0 C) concentrated H2SO4 (85 wt.%), and KMnO4 powder was added gradually by stirring and cooling to ensure a temperature below 20 C. This mixture was stirred at 35 C for 30 min, and then the temperature was increased to 98 C, and 150 mL of distilled water was later added slowly. This temperature was maintained for 15 min. The reaction was subsequently terminated by adding a large amount of distilled water and
923
30% H2O2 solution. Finally, the mixture was filtered and rinsed with 5% aqueous HCl solution until the sulphate could not be detected with BaCl2, and then the mixture was dried under a vacuum at 50 C [21]. 1-(3-aminopropyl)-3-methylimidazolium bromide (IL-NH2) was prepared through 1-methylimidazole (0.395 mL, 5 mmol) and 3-bromopropylamine hydrobromide (1.1000 g, 5 mmol) in 12.5 mL ethanol under nitrogen for 24 h. The resulting turbid mixture was re-crystallised by ethyl acetate as an anti-solvent. Finally, the resulting white powder was dried under vacuum at 60 C overnight. The ionic liquid graphite oxide (ILGO) was later synthesised by an epoxide ring-opening reaction between graphite oxide (GO) and IL-NH2 [22]. First, 0.05 g of GO was homogeneously dispersed in 50 mL of deionised water, and then 0.05 g (8.9 104 mol) of KOH was added into in the mixture, followed by 30 min of ultrasonication. Second, IL-NH2 (0.05 g) was then added into the mixture and ultrasonicated for another 1.0 h. The turbid mixture was then transferred into a three neck bottle and vigorously stirred at 80 C for 24 h. The resulting ILGO was subsequently washed with ethanol and water, and then air-dried [22]. ILGO was dispersed in a poly (2,20 -m-(phenylene)-5,50 -bibenzimidazole (PBI)/N,N-dimethylacetamide (DMAc)) solution with a mass ratio of 5 wt.%, and the membrane was loaded with phosphoric acid of 2 mol L1 and 4 mol L1 for three days each. The membrane in-plane conductivity was measured using a four-point probe with a frequency response analyser (Voltech TF2000, UK). The four-point probe method involves four equally spaced probes in contact with the membrane: two of the probes were used to source current whilst the other two were used to measure the voltage drop. The membranes were cut into 10 mm 50 mm strips and placed across four platinum foils with equal spacing of 5 mm. The AC impedance values were measured between frequencies of 1e20 kHz. To ensure the membrane reached a steady state, the membranes were held at the each desired conditions for 30 min before performing the measurements. The crystal structures of different membranes were analysed by X-ray diffraction (XRD, PANalytical X'Pert Pro Diffractometer), with a 2q range of 5e70 . The membrane morphologies were measured by a JSM-5300LV (Japan) Scanning Electron Microscope (SEM). Fourier transform infrared (FTIR) spectroscopy was performed on a Varian 800 FT-IR spectrometer system between 4000 cm1 and 400 cm1. Catalyst inks were prepared by blending carbon supported catalysts (50 wt.% Pt/C, Alfa Aesar) and polytetrafluoro-ethylene (PTFE, 60 wt.% Aldrich) in a watereethanol mixture under ultrasonic vibration for 10 min [23,24]. Gas diffusion electrodes (carbon paper) incorporated with wet proofed micro-porous layer (H2315 T10AC1) obtained from Freudenburg (FFCCT, Germany) were used as substrates to deposit the catalyst layer for both the anode and the cathode. The catalyst inks were sprayed onto carbon substrates at 100 C, and the electrodes were held at 150 C for 2 h to allow any liquid evaporation. The Pt loadings on the cathode and the anode were 0.9 mg cm2 and 0.5 mg cm2, respectively. A 4 mol L3 H3PO4 solution in a water/ethanol mixture was added onto the surface of the electrodes with a micropipette, and the electrodes were kept at 80 C to remove the residue water and ethanol [24]. The MEA was finally obtained by hot pressing the electrodes onto phosphoric acid loaded composite membranes at 150 C for 10 min with a load of 40 kg cm2. The MEA was fixed between two high-density graphite blocks (impregnated with phenolic resin) with parallel gas flow channels, and the active electrode area was 1 cm2. Electric cartridge heaters were mounted at the rear of the graphite blocks to maintain the desired temperature, which was monitored using imbedded
924
C. Xu et al. / Journal of Power Sources 274 (2015) 922e927
thermocouples and controlled with a temperature controller. H2 and O2/air were fed into the cell at flow rates of 100 cm3 min1 and 50 cm3 min1, respectively.
3. Results and discussion SEM observation was performed after loading the membrane with PA. Fig. 1 shows the cross-section of the membranes. The ILGO sheets remained exfoliated in the polymer matrix and were tightly held in the PBI matrix according to the strong interfacial interactions. Therefore, the conduction paths were established through the PBI matrix over the entire membrane, and these conduction paths provided a possible interpretation for the high conductivity of the composite membrane, as will be discussed later. The ILGO/PBI/PA composite membrane exhibited some slight wrinkling compared to the PBI/PA membrane. Due to the large particle size of the ionic liquid group bonded to GO, the dispersion of ILGO in the PBI matrix was not completely homogeneous, which may be the cause of the wrinkles. The X-ray diffraction pattern of graphite (Fig. 2) exhibited a typical graphite diffraction peak (002) at 2q ¼ 26.4 , and the interlay space was 0.34 nm, while the corresponding (002) reflection peak of GO was at 2q ¼ 11 with an interlayer spacing of 0.78 nm. The increase of the inter-planar distance of the GO comes from the existence of oxygen functional groups, which indicates that the oxidation of graphite was successful. The band of ILGO is at 2q ¼ 9.5 with an interlayer spacing of 0.93 nm, so the ILGO had a higher layer distance than GO as a result of the introduction of a larger size ionic liquid functional group onto the carbon sheets. FTIR studies confirmed the successful oxidation of graphite, as shown in Fig. 3. The FTIR spectrum of GO exhibited the different types of oxygen functionalities in graphite oxide at 3300 cm1 (OeH stretching vibrations) and at 1720 cm1 (stretching vibrations from C]O); the two peaks at 2854 cm1 and 2922 cm1 correspond to symmetric Vs (CH2) and asymmetric Vas (CH2) vibrations, while no significant peak was found in graphite. These results depicting OH and other functionalities, such as COOH groups, confirmed the successful oxidation of graphite. The characteristic band of the carboxylic group in ILGO appeared at 1725 cm1 (C]O stretching), the CeO vibrations of epoxy groups in GO appeared at 1153 cm1, and the CH3(N) stretching, CH2(N) stretching, and ring in-plane asymmetric stretching arising from the imidazolium ring appeared at 1571 cm1 and 1348 cm1, which confirmed the successful attachment of amine-terminated IL-NH2 to GO nanosheets [12].
Fig. 2. XRD patterns of graphite, GO and ILGO.
Fig. 3. Infrared spectra of graphite, GO, and ILGO.
The curve in Fig. 4 clearly showed that the ILGO/PBI had a higher conductivity than the pristine PBI membrane at different loading levels, and the higher H3PO4 content provided a higher conductivity. The ILGO/PBI of 2 PRU exhibited a similar conductivity performance with the PBI membrane of PRU 3.6. This result indicated that the ILGO as a filler used in the polymer matrix could be an ideal way to reduce the loading level of the membrane whilst achieving a
Fig. 1. SEM images of PBI/PA and ILGO/PBI/PA.
C. Xu et al. / Journal of Power Sources 274 (2015) 922e927
925
Fig. 4. Conductivities of PBI and ILGO/PBI composite membranes loaded with similar PRU of 1.9 and 3.6 under anhydrous conditions.
similar conductivity. The ionic liquid group could form the bond with H2PO 4 , which provided some pathways for proton transfer, and the rapid transfer through these sites should be quite fast. Therefore, the ILGO is a material that could enhance the proton conduction and reduce the required PA loading of the membranes. The ILGO enhanced the conductivity of the PBI with bonded acid, which indicates that the composite membrane is a suitable candidate for use in fuel cells. Fuel cell performance tests were performed with H2/O2 at atmospheric pressure. The polarisation and power density curves of the H2/O2 obtained at 175 C under anhydrous conditions for PBI and ILGO/PBI with different PA loadings are shown in Fig. 5. The open circuit voltages (OCVs) of all membranes were more than
0.9 V, which indicated that the membranes were pore-free and exhibited low gas crossover. The PBI/PA and ILGO/PBI/PA based fuel cells exhibited significant activation polarisation, with an over potential of some 0.25 V at 0.1 A cm2. The performance of the cells with the ILGO/PBI composite membranes was significantly better than that of the pristine PBI membrane at the same PA loading. The peak power densities of PBI and ILGO/PBI with oxygen were 0.26 W cm2 and 0.32 W cm2 at 2.0 PRU and 3.5 PRU, respectively. This improved performance was mainly attributed to the superior proton conductivity of the latter membrane and also the strong acid and water retention properties of the composite membrane at low acid loading. The membrane of higher loading level exhibited better performance indicating that H3PO4 played a major role in the
Fig. 5. Polarisation and power density curves of a fuel cell operated at 175 C with H2/O2 atmospheric pressure. Pt loading: cathode 0.9 mg cm2; anode 0.5 mg cm2; anhydrous conditions, H3PO4 PRU: 1.9 and 3.5.
926
C. Xu et al. / Journal of Power Sources 274 (2015) 922e927
improved conductivity. Hydrogen bonds in ILGO, which form acidic functional groups, such as carboxylic acid, epoxy oxygen and, especially the aminopropyl-methylimidazolium group, could provide more facile hopping of protons to enhance the conductivity. The internal resistance values were estimated from the voltage losses in the intermediate current density range. The voltage losses of the ILGO/PBI membrane based fuel cell were 0.141 V and 0.186 V (from 0.3 to 0.8 A cm2) at PRU values of 3.5 and 2.0, respectively, and resulted in cell conductivities of approximately 0.0136 S cm1 and 0.0093 S cm1 of ILGO/PBI at PRU values of 3.5 and 2.0, respectively. This conductivity was much less than that of the membranes alone (approximately 50% lower), which indicated a significant voltage loss in the electrode layers (and other cell components), i.e., the catalyst compositions in the MEA were not “optimal” for the fuel cell. Due to the low PA loading used in electrode, the electrode layer would have a relatively low ionic conductivity; thus, the catalyst utilisation in the electrode reactions was low. Essentially, only the Pt particles adjacent to the membrane were active, consequently explaining, in part, the high electrode activation lost. Fig. 6 shows IR corrected polarisation curves of the fuel cells operated with H2/O2 at 175 C. IR correction is based on the combined resistance of the membrane and the electrode, measured
from the slope of the cell voltage vs. current density at the high current density region (0.3e0.8 A cm2). In this region, the voltage loss associated with electrode polarisation is small compared to that of the internal resistance. From the plot of the IR corrected voltage vs. log (current density I), the slopes (Tafel type) of the line of PBI and QPBI are approximately 92 mV and 100 mV per decade at PRU of 3.5 and 2.0, respectively. The values are close to the literature reported values (100 mV dec1) on oxygen reduction reaction (ORR) electrodes for phosphoric-acid loaded PBI fuel cells [22,23]. Overall, although the ILGO-PBI membrane is a potential membrane for polymer electrolyte membrane fuel cells, more studies are required to investigate the electrode catalyst layer composition and to establish a suitable MEA preparation method. Ideally, this preparation method should incorporate the membrane materials in the catalyst layers, which is the subject of on-going research. 4. Conclusions Inorganiceorganic composite electrolytes, made from ILGO in PBI, were prepared for use in high temperature PEMFCs. The composite membranes, loaded with a low content of H3PO4, had similar conductivities to those of the higher phosphoric acid loading with pristine PBI membranes. The ILGO/PBI/PA composite membranes exhibited a proton conductivity of 0.035 S cm1 for a value of PRU of 3.5 at 175 C. In the fuel cell tests, the ILGO/PBI/PA composite membranes exhibited superior performance compared to that of a PBI/PA membrane. The peak power density of ILGO/PBI reached 320 mW cm2 and was higher than that of pristine PBI/PA at a value of PRU of 3.5 under H2/O2 conditions. The data indicated that the PBI composites with ILGO and loaded with phosphoric acid may be potential membranes for high temperature PEMFCs because of their high conductivity and fuel cell performance. Acknowledgements The author thanks the funding support by the 56th China Postdoctoral Science Foundation funded project, No. 2014M560506 and the Fundamental Research Funds for the Central Universities, No. 2014HGQC0004. This work is also supported by the EPSRC grant number is EP/C002601, Supergen, fuel cell consortium. References
Fig. 6. (a) IR corrected polarisation curves of a ILGO/PBI membrane at PRU values of 3.5 and 2.0; (b) Tafel plots obtained from the polarisation curves (I is the current density).
[1] T. Norby, The promise of protonics, Nature 410 (2001) 877e878. [2] C. Yang, P. Costamagna, S. Srinivasan, J. Benziger, A.B. Bocarsly, J. Power Sources 103 (2001) 1. [3] Q. Li, R. He, J.O. Jensen, N.J. Bjerrum, Fuel Cells 4 (3) (2004) 147e159. [4] Q. Li, R. He, J.O. Jensen, N.J. Bjerrum, Chem. Mat. 15 (2003) 4896e4915. [5] M. Li, Z. Shao, K. Scott, J. Power Sources 183 (1) (2008) 69e75. [6] Q. Li, J.O. Jensen, R.F. Savinell, N.J. Bjerrum, Prog. Polym. Sci. Oxf. 34 (5) (2009) 449e477. [7] D.A. Boysen, C.R.I. Uda, T. Chisholm, S.M. Haile, Science 303 (5654) (2004) 68e70. [8] M. Cappadonia, O. Niemzig, U. Stimming, Solid State Ionics 125 (1) (1999) 333e337. [9] P. Heo, N. Kajiyama, K. Kobayashi, M. Nagao, M. Sano, T. Hibino, Electrochem. Solid-State Lett. 11 (6) (2008) B91eB95. [10] S.Y. Oh, Toshihiro Yoshida, Go Kawamura, Hiroyuki Muto, Mototsugu Sakai, Atsunori Matsuda, J. Mater. Chem. 20 (30) (2010) 6359e6366. [11] X. Liu, E.H. Yu, K. Scott, Appl. Catal. B Environ. 162 (2015) 593e601. [12] G. Eda, M. Chhowalla, Adv. Mater. 22 (22) (2010) 2392e2415. [13] I. Jung, D. Dikin, S. Park, W. Cai, S.L. Mielke, R.S. Ruoff, J. Phys. Chem. C 112 (51) (2008) 20264e20268. [14] C. Xu, X. Wu, X. Wang, M. Mamlouk, K. Scott, J. Mater. Chem. 21 (2011) 6014. [15] H. Zarrin, D. Higgins, Y. Jun, Z. Chen, M. Fowler, J. Phys. Chem. C 115 (42) (2011) 20774e20781. [16] B.G. Choi, J. Hong, Y.C. Park, D.H. Jung, W.H. Hong, P.T. Hammond, H. Park, ACS Nano 5 (6) (2011) 5167e5517. [17] C. Xu, Y. Cao, R. Kumar, X. Wu, X. Wang, K. Scott, J. Mater. Chem. 21 (2011) 11359. [18] J. Le Bideau, L. Viau, A. Vioux, Chem. Soc. Rev. 40 (2) (2011) 907e925.
C. Xu et al. / Journal of Power Sources 274 (2015) 922e927 [19] H. Ye, J. Huang, J.J. Xu, N.K.A.C. Kodiweera, J.R.P. Jayakody, J. Power Sources 178 (2) (2008) 651e660. [20] J. Gao, J.G. Liu, W.M. Liu, B. Li, Y.C. Xin, Y. Yin, J. Gu, Z.G. Zou, Int. J. Electrochem. Sci. 6 (2011) 6115e6122. [21] W.S. Hummers Jr., R.E. Offeman, Chem. Soc. 80 (6) (1958) 1339.
927
[22] H. Yang, C. Shan, F. Li, D. Han, Q. Zhang, L. Niu, Chem. Commun. 26 (2009) 3880e3882. [23] R. Kannan, M. Parthasarathy, S.U. Maraveedu, S. Kurungot, V.K. Pillai, Langmuir 25 (14) (2009) 8299e8305. [24] Z. Liu, J.S. Wainright, R.F. Savinell, Chem. Eng. Sci. 59 (2004) 4833e4838.
Contents lists available at ScienceDirect
Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour
A polybenzimidazole/ionic-liquid-graphite-oxide composite membrane for high temperature polymer electrolyte membrane fuel cells Chenxi Xu a, b, 1, Xiaoteng Liu b, 1, Jigui Cheng a, *, Keith Scott b, ** a b
School of Materials Science and Engineering, Hefei University of Technology, Hefei, Anhui, China School of Chemical Engineering and Advanced Materials, Newcastle University, Newcastle NE1 7RU, United Kingdom
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
1-(3-Aminopropyl)-3-methylimidazolium groups (ionic liquid) was bonded to the graphite oxide. ILGO filler enhanced the proton conductivity of PBI membrane. ILGO/PBI membrane showed superior performances in HT-PEMFC.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 3 October 2014 Received in revised form 20 October 2014 Accepted 21 October 2014 Available online 28 October 2014
Graphite oxide is successfully functionalised by 3-aminopropyltriethoxysilane ionic liquid and used as a filler material in a polybenzimidazole (PBI) membrane for high temperature proton exchange membrane fuel cells. The ionic-liquid-graphite-oxide/polybenzimidazole (ILGO/PBI) composite membrane exhibits an appropriate level of proton conductivity when imbibed with phosphoric acid at low phosphoric acid loading, which promotes its use in fuel cells by avoiding acid leakage and materials corrosion. The ionic conductivities of the ILGO/PBI membranes at 175 C are 0.035 S cm1 and 0.025 S cm1 at per repeat units of 3.5 and 2.0, respectively. The fuel cell performance of ILGO/PBI membranes exhibits a maximum power density of 320 mW cm2 at 175 C, which is higher than that of a pristine PBI membrane. © 2014 Published by Elsevier B.V.
Keywords: Ionic liquid graphite oxide Polybenzimidazole Polymer electrolyte membrane fuel cells
1. Introduction In the last decades, considerable efforts have been made to develop high temperature (>100 C) proton exchange membrane fuel cells (PEMFCs) using polymer acid complexes (PACs) because it offers significant advantages in this temperature range, such as (1)
* Corresponding author. ** Corresponding author. E-mail addresses: [email protected] (C. Xu), [email protected] (J. Cheng). 1 These authors contributed equally to this work. http://dx.doi.org/10.1016/j.jpowsour.2014.10.134 0378-7753/© 2014 Published by Elsevier B.V.
improved CO tolerance, (2) enhanced efficiency, (3) avoidance of flooding by water, (4) opportunity to use non-noble metal catalysts, and (5) system simplification [1e5]. Solid-state electrolytes loaded with phosphoric acid provide less corrosion and more immobilisation compared to aqueous phosphoric acid fuel cells (PAFCs) [6e8]. The phosphoric-acid loaded polybenzimidazole (PBI) is the best-known example of a membrane; PBI has been used to produce reasonably successful membranes for fuel cells with excellent thermo-chemical stability and good conductivity [6,7]. However, in many cases, PBI/H3PO4 membranes exhibit high conductivity values only with high acid
C. Xu et al. / Journal of Power Sources 274 (2015) 922e927
loadings, which are usually at least higher than 5.0H3PO4 molecules per repeat unit (PRU) of PBI [8]. Such a high PA acid content creates problems of mechanical strength reduction and elution problems of electrolytes, as well as catalyst corrosion problems associated with the use of an excess of highly concentrated PA at high temperature [9,10]. One promising idea to reduce the loading level is to fill solid proton conductors into the PBI membrane, such as solid acids, ionic liquid and graphite oxide [10e12]. Graphite oxide (GO) consists of carbon, oxygen and hydrogen and is exfoliated into individual graphite oxide nanoplatelets in water because of the hydrophilic oxygen groups attached to the graphene sheets [13]. The presence of acidic groups of GO, such as carboxylic acid and epoxy oxygen, could facilitate hopping of protons [14e16]. According to our previous work, the sulfonic acid groups functionalised to GO could improve water retention ability and provide hydrogen bond for the conductivity increase [17]. So, the sulphonic GO with PBI exhibited improved performance at low PA loading level. Functionalised graphite oxide (FGO) is also easy to disperse in a solvent, resulting in a homogeneous distribution into polymer matrices; in addition, the functional groups could improve the proton conductivity. Hence, the functionalised GO is a suitable candidate filler material for use in membranes to improve the proton conductivity and avoid excessive content of PA. Recently, functionalised GO was used in composite membranes and has attracted much attention. Functionalised groups, such as 3mercaptopropyl trimethoxysilane (MPTMS) [15], added into graphite oxide were incorporated with the perfluorinated polymer Nafion®. The functionalised graphite controlled the state of water by means of nanoscale manipulation of the physical geometry and the chemical functionality of ionic channels. The confinement of bound water within the reorganised nanochannels of composite membranes enhanced the proton conductivity at high temperature via the low activation energy for ionic conduction [16]. Additionally, the increase of FGO nanofiller loading extended the number of available ion exchange sites per cluster, resulting in the increased proton mobility in the membrane at higher temperature and lower humidity [15]. Ionic liquids (ILs) are organic salts with high ionic conductivity and wide electrochemical potential windows. ILs could improve safety and enable a higher operating temperature range to be used [18,19]; therefore, they were considered as an additive in the proton exchange membrane. According to the oxygen group located in GO, the ILs could be readily bonded to the GO. The combination of GO and IL may enhance the PBI membrane conductivity because both the ILs and GO may provide hydrogen bond as the proton transport route [20]. In this work, the GO sheets were functionalised by an ionic liquid group 1-(3-aminopropyl)-3-methylimidazolium (ILGO) to combine with the PBI membrane loaded with phosphoric acid. The membrane proton conductivity and fuel cell performance were studied, and the results indicated that the performance of the PBI/ H3PO4 membrane was enhanced with the addition of ILGO. 2. Experimental methods A graphite oxide (GO) sheet was synthesised from graphite (Shandong Qingdao graphite company, China) by oxidation with KMnO4 in concentrated H2SO4 according to the Hummers method [21]. A graphite sheet, was placed in cold (0 C) concentrated H2SO4 (85 wt.%), and KMnO4 powder was added gradually by stirring and cooling to ensure a temperature below 20 C. This mixture was stirred at 35 C for 30 min, and then the temperature was increased to 98 C, and 150 mL of distilled water was later added slowly. This temperature was maintained for 15 min. The reaction was subsequently terminated by adding a large amount of distilled water and
923
30% H2O2 solution. Finally, the mixture was filtered and rinsed with 5% aqueous HCl solution until the sulphate could not be detected with BaCl2, and then the mixture was dried under a vacuum at 50 C [21]. 1-(3-aminopropyl)-3-methylimidazolium bromide (IL-NH2) was prepared through 1-methylimidazole (0.395 mL, 5 mmol) and 3-bromopropylamine hydrobromide (1.1000 g, 5 mmol) in 12.5 mL ethanol under nitrogen for 24 h. The resulting turbid mixture was re-crystallised by ethyl acetate as an anti-solvent. Finally, the resulting white powder was dried under vacuum at 60 C overnight. The ionic liquid graphite oxide (ILGO) was later synthesised by an epoxide ring-opening reaction between graphite oxide (GO) and IL-NH2 [22]. First, 0.05 g of GO was homogeneously dispersed in 50 mL of deionised water, and then 0.05 g (8.9 104 mol) of KOH was added into in the mixture, followed by 30 min of ultrasonication. Second, IL-NH2 (0.05 g) was then added into the mixture and ultrasonicated for another 1.0 h. The turbid mixture was then transferred into a three neck bottle and vigorously stirred at 80 C for 24 h. The resulting ILGO was subsequently washed with ethanol and water, and then air-dried [22]. ILGO was dispersed in a poly (2,20 -m-(phenylene)-5,50 -bibenzimidazole (PBI)/N,N-dimethylacetamide (DMAc)) solution with a mass ratio of 5 wt.%, and the membrane was loaded with phosphoric acid of 2 mol L1 and 4 mol L1 for three days each. The membrane in-plane conductivity was measured using a four-point probe with a frequency response analyser (Voltech TF2000, UK). The four-point probe method involves four equally spaced probes in contact with the membrane: two of the probes were used to source current whilst the other two were used to measure the voltage drop. The membranes were cut into 10 mm 50 mm strips and placed across four platinum foils with equal spacing of 5 mm. The AC impedance values were measured between frequencies of 1e20 kHz. To ensure the membrane reached a steady state, the membranes were held at the each desired conditions for 30 min before performing the measurements. The crystal structures of different membranes were analysed by X-ray diffraction (XRD, PANalytical X'Pert Pro Diffractometer), with a 2q range of 5e70 . The membrane morphologies were measured by a JSM-5300LV (Japan) Scanning Electron Microscope (SEM). Fourier transform infrared (FTIR) spectroscopy was performed on a Varian 800 FT-IR spectrometer system between 4000 cm1 and 400 cm1. Catalyst inks were prepared by blending carbon supported catalysts (50 wt.% Pt/C, Alfa Aesar) and polytetrafluoro-ethylene (PTFE, 60 wt.% Aldrich) in a watereethanol mixture under ultrasonic vibration for 10 min [23,24]. Gas diffusion electrodes (carbon paper) incorporated with wet proofed micro-porous layer (H2315 T10AC1) obtained from Freudenburg (FFCCT, Germany) were used as substrates to deposit the catalyst layer for both the anode and the cathode. The catalyst inks were sprayed onto carbon substrates at 100 C, and the electrodes were held at 150 C for 2 h to allow any liquid evaporation. The Pt loadings on the cathode and the anode were 0.9 mg cm2 and 0.5 mg cm2, respectively. A 4 mol L3 H3PO4 solution in a water/ethanol mixture was added onto the surface of the electrodes with a micropipette, and the electrodes were kept at 80 C to remove the residue water and ethanol [24]. The MEA was finally obtained by hot pressing the electrodes onto phosphoric acid loaded composite membranes at 150 C for 10 min with a load of 40 kg cm2. The MEA was fixed between two high-density graphite blocks (impregnated with phenolic resin) with parallel gas flow channels, and the active electrode area was 1 cm2. Electric cartridge heaters were mounted at the rear of the graphite blocks to maintain the desired temperature, which was monitored using imbedded
924
C. Xu et al. / Journal of Power Sources 274 (2015) 922e927
thermocouples and controlled with a temperature controller. H2 and O2/air were fed into the cell at flow rates of 100 cm3 min1 and 50 cm3 min1, respectively.
3. Results and discussion SEM observation was performed after loading the membrane with PA. Fig. 1 shows the cross-section of the membranes. The ILGO sheets remained exfoliated in the polymer matrix and were tightly held in the PBI matrix according to the strong interfacial interactions. Therefore, the conduction paths were established through the PBI matrix over the entire membrane, and these conduction paths provided a possible interpretation for the high conductivity of the composite membrane, as will be discussed later. The ILGO/PBI/PA composite membrane exhibited some slight wrinkling compared to the PBI/PA membrane. Due to the large particle size of the ionic liquid group bonded to GO, the dispersion of ILGO in the PBI matrix was not completely homogeneous, which may be the cause of the wrinkles. The X-ray diffraction pattern of graphite (Fig. 2) exhibited a typical graphite diffraction peak (002) at 2q ¼ 26.4 , and the interlay space was 0.34 nm, while the corresponding (002) reflection peak of GO was at 2q ¼ 11 with an interlayer spacing of 0.78 nm. The increase of the inter-planar distance of the GO comes from the existence of oxygen functional groups, which indicates that the oxidation of graphite was successful. The band of ILGO is at 2q ¼ 9.5 with an interlayer spacing of 0.93 nm, so the ILGO had a higher layer distance than GO as a result of the introduction of a larger size ionic liquid functional group onto the carbon sheets. FTIR studies confirmed the successful oxidation of graphite, as shown in Fig. 3. The FTIR spectrum of GO exhibited the different types of oxygen functionalities in graphite oxide at 3300 cm1 (OeH stretching vibrations) and at 1720 cm1 (stretching vibrations from C]O); the two peaks at 2854 cm1 and 2922 cm1 correspond to symmetric Vs (CH2) and asymmetric Vas (CH2) vibrations, while no significant peak was found in graphite. These results depicting OH and other functionalities, such as COOH groups, confirmed the successful oxidation of graphite. The characteristic band of the carboxylic group in ILGO appeared at 1725 cm1 (C]O stretching), the CeO vibrations of epoxy groups in GO appeared at 1153 cm1, and the CH3(N) stretching, CH2(N) stretching, and ring in-plane asymmetric stretching arising from the imidazolium ring appeared at 1571 cm1 and 1348 cm1, which confirmed the successful attachment of amine-terminated IL-NH2 to GO nanosheets [12].
Fig. 2. XRD patterns of graphite, GO and ILGO.
Fig. 3. Infrared spectra of graphite, GO, and ILGO.
The curve in Fig. 4 clearly showed that the ILGO/PBI had a higher conductivity than the pristine PBI membrane at different loading levels, and the higher H3PO4 content provided a higher conductivity. The ILGO/PBI of 2 PRU exhibited a similar conductivity performance with the PBI membrane of PRU 3.6. This result indicated that the ILGO as a filler used in the polymer matrix could be an ideal way to reduce the loading level of the membrane whilst achieving a
Fig. 1. SEM images of PBI/PA and ILGO/PBI/PA.
C. Xu et al. / Journal of Power Sources 274 (2015) 922e927
925
Fig. 4. Conductivities of PBI and ILGO/PBI composite membranes loaded with similar PRU of 1.9 and 3.6 under anhydrous conditions.
similar conductivity. The ionic liquid group could form the bond with H2PO 4 , which provided some pathways for proton transfer, and the rapid transfer through these sites should be quite fast. Therefore, the ILGO is a material that could enhance the proton conduction and reduce the required PA loading of the membranes. The ILGO enhanced the conductivity of the PBI with bonded acid, which indicates that the composite membrane is a suitable candidate for use in fuel cells. Fuel cell performance tests were performed with H2/O2 at atmospheric pressure. The polarisation and power density curves of the H2/O2 obtained at 175 C under anhydrous conditions for PBI and ILGO/PBI with different PA loadings are shown in Fig. 5. The open circuit voltages (OCVs) of all membranes were more than
0.9 V, which indicated that the membranes were pore-free and exhibited low gas crossover. The PBI/PA and ILGO/PBI/PA based fuel cells exhibited significant activation polarisation, with an over potential of some 0.25 V at 0.1 A cm2. The performance of the cells with the ILGO/PBI composite membranes was significantly better than that of the pristine PBI membrane at the same PA loading. The peak power densities of PBI and ILGO/PBI with oxygen were 0.26 W cm2 and 0.32 W cm2 at 2.0 PRU and 3.5 PRU, respectively. This improved performance was mainly attributed to the superior proton conductivity of the latter membrane and also the strong acid and water retention properties of the composite membrane at low acid loading. The membrane of higher loading level exhibited better performance indicating that H3PO4 played a major role in the
Fig. 5. Polarisation and power density curves of a fuel cell operated at 175 C with H2/O2 atmospheric pressure. Pt loading: cathode 0.9 mg cm2; anode 0.5 mg cm2; anhydrous conditions, H3PO4 PRU: 1.9 and 3.5.
926
C. Xu et al. / Journal of Power Sources 274 (2015) 922e927
improved conductivity. Hydrogen bonds in ILGO, which form acidic functional groups, such as carboxylic acid, epoxy oxygen and, especially the aminopropyl-methylimidazolium group, could provide more facile hopping of protons to enhance the conductivity. The internal resistance values were estimated from the voltage losses in the intermediate current density range. The voltage losses of the ILGO/PBI membrane based fuel cell were 0.141 V and 0.186 V (from 0.3 to 0.8 A cm2) at PRU values of 3.5 and 2.0, respectively, and resulted in cell conductivities of approximately 0.0136 S cm1 and 0.0093 S cm1 of ILGO/PBI at PRU values of 3.5 and 2.0, respectively. This conductivity was much less than that of the membranes alone (approximately 50% lower), which indicated a significant voltage loss in the electrode layers (and other cell components), i.e., the catalyst compositions in the MEA were not “optimal” for the fuel cell. Due to the low PA loading used in electrode, the electrode layer would have a relatively low ionic conductivity; thus, the catalyst utilisation in the electrode reactions was low. Essentially, only the Pt particles adjacent to the membrane were active, consequently explaining, in part, the high electrode activation lost. Fig. 6 shows IR corrected polarisation curves of the fuel cells operated with H2/O2 at 175 C. IR correction is based on the combined resistance of the membrane and the electrode, measured
from the slope of the cell voltage vs. current density at the high current density region (0.3e0.8 A cm2). In this region, the voltage loss associated with electrode polarisation is small compared to that of the internal resistance. From the plot of the IR corrected voltage vs. log (current density I), the slopes (Tafel type) of the line of PBI and QPBI are approximately 92 mV and 100 mV per decade at PRU of 3.5 and 2.0, respectively. The values are close to the literature reported values (100 mV dec1) on oxygen reduction reaction (ORR) electrodes for phosphoric-acid loaded PBI fuel cells [22,23]. Overall, although the ILGO-PBI membrane is a potential membrane for polymer electrolyte membrane fuel cells, more studies are required to investigate the electrode catalyst layer composition and to establish a suitable MEA preparation method. Ideally, this preparation method should incorporate the membrane materials in the catalyst layers, which is the subject of on-going research. 4. Conclusions Inorganiceorganic composite electrolytes, made from ILGO in PBI, were prepared for use in high temperature PEMFCs. The composite membranes, loaded with a low content of H3PO4, had similar conductivities to those of the higher phosphoric acid loading with pristine PBI membranes. The ILGO/PBI/PA composite membranes exhibited a proton conductivity of 0.035 S cm1 for a value of PRU of 3.5 at 175 C. In the fuel cell tests, the ILGO/PBI/PA composite membranes exhibited superior performance compared to that of a PBI/PA membrane. The peak power density of ILGO/PBI reached 320 mW cm2 and was higher than that of pristine PBI/PA at a value of PRU of 3.5 under H2/O2 conditions. The data indicated that the PBI composites with ILGO and loaded with phosphoric acid may be potential membranes for high temperature PEMFCs because of their high conductivity and fuel cell performance. Acknowledgements The author thanks the funding support by the 56th China Postdoctoral Science Foundation funded project, No. 2014M560506 and the Fundamental Research Funds for the Central Universities, No. 2014HGQC0004. This work is also supported by the EPSRC grant number is EP/C002601, Supergen, fuel cell consortium. References
Fig. 6. (a) IR corrected polarisation curves of a ILGO/PBI membrane at PRU values of 3.5 and 2.0; (b) Tafel plots obtained from the polarisation curves (I is the current density).
[1] T. Norby, The promise of protonics, Nature 410 (2001) 877e878. [2] C. Yang, P. Costamagna, S. Srinivasan, J. Benziger, A.B. Bocarsly, J. Power Sources 103 (2001) 1. [3] Q. Li, R. He, J.O. Jensen, N.J. Bjerrum, Fuel Cells 4 (3) (2004) 147e159. [4] Q. Li, R. He, J.O. Jensen, N.J. Bjerrum, Chem. Mat. 15 (2003) 4896e4915. [5] M. Li, Z. Shao, K. Scott, J. Power Sources 183 (1) (2008) 69e75. [6] Q. Li, J.O. Jensen, R.F. Savinell, N.J. Bjerrum, Prog. Polym. Sci. Oxf. 34 (5) (2009) 449e477. [7] D.A. Boysen, C.R.I. Uda, T. Chisholm, S.M. Haile, Science 303 (5654) (2004) 68e70. [8] M. Cappadonia, O. Niemzig, U. Stimming, Solid State Ionics 125 (1) (1999) 333e337. [9] P. Heo, N. Kajiyama, K. Kobayashi, M. Nagao, M. Sano, T. Hibino, Electrochem. Solid-State Lett. 11 (6) (2008) B91eB95. [10] S.Y. Oh, Toshihiro Yoshida, Go Kawamura, Hiroyuki Muto, Mototsugu Sakai, Atsunori Matsuda, J. Mater. Chem. 20 (30) (2010) 6359e6366. [11] X. Liu, E.H. Yu, K. Scott, Appl. Catal. B Environ. 162 (2015) 593e601. [12] G. Eda, M. Chhowalla, Adv. Mater. 22 (22) (2010) 2392e2415. [13] I. Jung, D. Dikin, S. Park, W. Cai, S.L. Mielke, R.S. Ruoff, J. Phys. Chem. C 112 (51) (2008) 20264e20268. [14] C. Xu, X. Wu, X. Wang, M. Mamlouk, K. Scott, J. Mater. Chem. 21 (2011) 6014. [15] H. Zarrin, D. Higgins, Y. Jun, Z. Chen, M. Fowler, J. Phys. Chem. C 115 (42) (2011) 20774e20781. [16] B.G. Choi, J. Hong, Y.C. Park, D.H. Jung, W.H. Hong, P.T. Hammond, H. Park, ACS Nano 5 (6) (2011) 5167e5517. [17] C. Xu, Y. Cao, R. Kumar, X. Wu, X. Wang, K. Scott, J. Mater. Chem. 21 (2011) 11359. [18] J. Le Bideau, L. Viau, A. Vioux, Chem. Soc. Rev. 40 (2) (2011) 907e925.
C. Xu et al. / Journal of Power Sources 274 (2015) 922e927 [19] H. Ye, J. Huang, J.J. Xu, N.K.A.C. Kodiweera, J.R.P. Jayakody, J. Power Sources 178 (2) (2008) 651e660. [20] J. Gao, J.G. Liu, W.M. Liu, B. Li, Y.C. Xin, Y. Yin, J. Gu, Z.G. Zou, Int. J. Electrochem. Sci. 6 (2011) 6115e6122. [21] W.S. Hummers Jr., R.E. Offeman, Chem. Soc. 80 (6) (1958) 1339.
927
[22] H. Yang, C. Shan, F. Li, D. Han, Q. Zhang, L. Niu, Chem. Commun. 26 (2009) 3880e3882. [23] R. Kannan, M. Parthasarathy, S.U. Maraveedu, S. Kurungot, V.K. Pillai, Langmuir 25 (14) (2009) 8299e8305. [24] Z. Liu, J.S. Wainright, R.F. Savinell, Chem. Eng. Sci. 59 (2004) 4833e4838.