- Email: [email protected]
Proton conducting graft copolymers with tunable length and density of phosphonated side chains for fuel cell membranes
Journal of Membrane Science 450 (2014) 362–368
Contents lists available at ScienceDirect
Journal of Membrane Science journal homepage: www.elsevier.com/locate/memsci
Proton conducting graft copolymers with tunable length and density of phosphonated side chains for fuel cell membranes Ivaylo Dimitrov a,1, Shogo Takamuku b, Katja Jankova a, Patric Jannasch b, Søren Hvilsted a,n a b
Department of Chemical and Biochemical Engineering, Danish Polymer Centre, Technical University of Denmark, DTU, DK-2800 Kgs. Lyngby, Denmark Department of Chemistry, Polymer and Materials Chemistry, Lund University, P.O. Box124, SE-22 100 Lund, Sweden
art ic l e i nf o
a b s t r a c t
Article history: Received 29 July 2013 Received in revised form 7 September 2013 Accepted 10 September 2013 Available online 20 September 2013
Polysulfones functionalized with highly phosphonated poly(pentafluorostyrene) side chains of different lengths were synthesized applying controlled polymerization and modification methods. The graft copolymers' thermal properties were evaluated by differential scanning calorimetry and thermal gravimetrical analyses. The proton conductivity of membrane prepared from the graft copolymer with the shortest phosphonated side chains was 134 mS cm 1 at 100 1C under fully immersed conditions. The graft copolymer TEM image shows a nanophase separation of ion-rich segments within the polysulfone matrix. Increasing the ionic groups content in the graft copolymers led to extensive membrane swelling. To improve the dimensional stability the graft copolymers were blended with pyridine-modified polysulfone. The blend membranes were transparent with formation of nano-phase domains as revealed from TEM images. The acid–base blend membranes exhibited a slightly higher thermal stability but lower proton conductivity compared to the membranes formed from pure graft copolymers. & 2013 Elsevier B.V. All rights reserved.
Keywords: Graft copolymer Acid–base blend Polymer electrolyte membrane Proton conductivity Thermal stability
1. Introduction Proton conducting polymer electrolyte membranes (PEMs) are one of the key components that determine the cost and performance of polymer electrolyte membrane fuel cells (PEMFCs). In recent years, there has been increased interest in the design and development of novel proton conducting PEMs able to operate at high temperatures (4100 1C) under low humidification conditions [1,2]. Higher operation temperatures result in faster electrode kinetics, reduced risk of catalyst poisoning, smaller heat exchanger and easier water/thermal management [3,4]. The most widely used commercially available PEMs are based on perfluorosulfonic acid (PFSA) ionomers, such as Nafions, produced by DuPont. Despite their good thermal and mechanical properties, the PFSA-based ionomers have several disadvantages including the high cost, decreased performance and dehydration at temperatures above 90 1C [5]. The evaluation of three model compounds containing different protogenic functionalities: sulfonic acid, phosphonic acid and imidazole, for high temperature and low humidity operation has shown that only the phosphonic acid-based system exhibits high conductivity in the temperature range 100–200 1C [6]. The significant strength of the carbon–phosphorus bond makes n
Corresponding author. Tel.: þ 45 45252965; fax: þ 45 4588 2161. E-mail address: [email protected] (S. Hvilsted). 1 Permanent address: Institute of Polymers, Bulgarian Academy of Sciences, 1113-Sofia, Bulgaria. 0376-7388/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.memsci.2013.09.016
the polymers bearing covalently attached phosphonic acid groups attractive for high temperature operation. Moreover, due to the amphoteric nature of phosphonic acid, the membranes are able to conduct protons by diffusion at low humidity, while the proton conductivity is facilitated by the dynamics of the water under immersed conditions. In terms of the copolymer architecture, the ion-containing block and graft copolymers are of significant interest due to their ability to self-assemble in ordered nanostructures that facilitate proton/water transport over a wide range of conditions [7,8]. A number of block and graft copolymers containing phosphonic acid groups have been reported and some of them have also been evaluated for potential fuel cell applications [9–13]. Generally, the proton conductivity and the water retention capacity of the membranes are enhanced when the content of ionic groups is increased [14]. However, upon reaching a certain level, the membranes become highly swollen and even soluble in water. A way to overcome this problem is to cross-link the polymer [15]. The process of cross-linking might be chemical [16–18] or physical [19–21]. The influence of tuning the ionic content of membranes by blending, as opposed to varying the degree of sulfonation, has been evaluated by blending highly sulfonated graft copolymers with fluorous homopolymers [22]. Most often the physical cross-linking is achieved via acid–base interactions between the proton conducting acidic polymer and a suitable basic polymer [23]. Very recently, we introduced a multistep synthetic strategy for the preparation of novel graft copolymers with a polysulfone (PSU)
I. Dimitrov et al. / Journal of Membrane Science 450 (2014) 362–368
backbone and highly phosphonated poly(pentafluorostyrene) (PFS) side chains [24]. The controlled polymerization and modification steps involved allow straightforward tuning of graft copolymer architecture in terms of length and density of the phosphonated side chains. The initial evaluation of copolymer membrane proton conductivity and thermal stability showed promising results. In the present work, we have extended the variety of phosphonated graft copolymers in the search for an optimal balance between high proton conductivity and membrane dimensional stability. We tuned the copolymer architecture to obtain a twice as high grafting degree as that achieved previously. The length of the side chains was also varied in a controlled manner. Furthermore, to improve the membrane dimensional stability, acid–base polymer blends were prepared from the phosphonated graft copolymers and a polysulfone bearing pyridine side groups. The membranes formed from the pure graft copolymers and from the acid–base blends were evaluated for their thermal properties, water uptake and proton conductivity.
2. Experimental 2.1. Materials Graft copolymers with PSU backbone and highly phosphonated PFS side chains of different lengths (PSU-g-PhPFS) were synthesized as previously described [24]. Briefly, PSU (Udels P-3500 LCD MB8, Mn ¼ 40,000 g mol 1, PDI: 1.95, Solvay Advanced Polymes) was lithiated with n-butyllithium in tetrahydrofuran (THF), followed by the addition of (chloromethyl)benzoyl chloride. The targeted degree of functionalization (DS ¼ 0.14, i.e., 14 substitutions per 100 PSU repeating units) was achieved and confirmed by 1H NMR spectroscopy. The benzyl chloride-containing PSU was converted into a multiazide-functionalized polymer (PSU-N) through reaction with sodium azide. Alkyne end-functionalized PFSs (PFSA) of three different number-average molar masses (Mn ¼3700, 7600 and 10,500 Da) were synthesized by atom transfer radical polymerization (ATRP) and “clicked” onto the PSU-backbone via copper catalyzed azide-alkyne 1,3-cycloaddition (CuAAC). Finally, the grafts were phosphonated through reaction with tris(trimethylsilyl) phosphite as described by Atanasov and Kerres [25]. PSU grafted with 40 mol% of basic pyridine side groups (PSU-Py) was synthesized as previously described [26]. 2.2. Membrane preparation The pure graft copolymers (1–2 wt%) were suspended in Nmethylpyrrolidone (NMP) at 80 1C. The temperature was increased to 150 1C and the mixtures were stirred overnight until the polymers dissolved. The membranes were cast at 80 1C during 4 days under nitrogen onto glass plates. In the case of the acid– base blend membranes, a solution of PSU-Py in NMP (2 wt%) was prepared separately at 80 1C. Then, it was slowly added to the copolymer solutions. The resulting mixtures were spread onto glass plates and the membranes were cast in the oven at 150 1C during 8 h. All membranes were detached from the glass plates after immersion in deionized water, followed by repeated rinsing and washing with deionized water. 2.3. Characterization 1
H NMR spectra were recorded in CDCl3 or DMSO-d6 on a Bruker Avance 300 MHz spectrometer. Size-exclusion chromatography (SEC) was carried out with a Viscotek GPCmax VE-2001 equipped with Viscotek TriSEC Model 302 triple detector array (refractive index detector, viscometer detector, laser light scattering detector with the
363
light wavelength of 670 nm and measuring angles of 901 and 71), and a Knauer K-2501 UV detector using two PLgel mixed-D columns from Polymer Laboratories. The samples were run in THF at 30 1C (1 mL min 1). The molar-mass characteristics were calculated using polystyrene standards. Thermal analyses were carried out on a differential scanning calorimeter DSC Q1000 (TA Instruments) in a temperature range of 25–400 1C at a heating and cooling rate of 10 1C min 1 under nitrogen. The glass transition temperatures (Tg) were determined during the second heating cycle at the inflection point of the thermal transition. Thermogravimetric analysis (TGA) was performed on a TGA Q500 instrument from 50 to 600 1C with a heating rate of 20 1C min 1 under nitrogen flow. The ion exchange capacity (IEC, meq. g 1) was calculated assuming 100% efficiency of the “click” grafting, 90% phosphonation of the grafts and considering both protons of each phosphonic acid group. The water uptake (WU) was measured from the weight increase of membranes after soaking them in deionized water at temperatures between 20 and 100 1C by increments of 20 1C to investigate the temperature dependence. It was calculated as: WU (wt%)¼(Wwet– Wdry) Wdry 1 100, where Wwet and Wdry are the membrane weights under hydrated and dry condition respectively. The number of water molecules per H þ in the phosphonic acid group (λ) was calculated as: λ ¼ 1000 (Wwet–Wdry) Wdry 1 (IEC 18) 1. The proton conductivity (S cm 1) of fully immersed membranes was measured in a sealed cell between 20 and 120 1C, using a Novocontrol high resolution dielectric analyzer V 1.01S in the frequency range 10–107 Hz at 50 mV voltage amplitude in a twoprobe method. The humidity dependence of the proton conductivity from 30% to 90% relative humidity (RH) was investigated at 80 1C by a four-probe method with a Gamry potentiostat/galvanostat/ZRA in the frequency range 10 1–105 Hz and a Fumatech MK3 conductivity cell, where the humidity was equilibrated by deionized water in a closed system. The samples for transmission electron microscopy (TEM) were stained with lead acetate and microtomed as previously described [22]. Electron micrographs were obtained with a Tecnai T20 G2 instrument operated at an accelerating voltage of 200 kV.
3. Results and discussion 3.1. Synthesis and characterization of the graft copolymers The synthetic sequence of efficient and controlled polymerization and modification steps allows us to finely tune the copolymer composition in the search for an optimal balance between proton conductivity and mechanical properties of PEMs. We decided to explore the influence of the grafting density as well as of the length of the phosphonated side chains on the copolymer membranes properties. The grafting sites on the PSU backbone were doubled as compared to those obtained in our first work and were set to 14 mol%. Separately, three alkyne end functionalized PFSs (PFS-A) of different lengths were synthesized and grafted onto the PSU backbone through copper catalyzed azide–alkyne 1,3-cycloaddition. The final reaction step was the phosphonation of the PFS side chains. The intermediate and final products were analyzed and characterized as previously described [24]. The precursor and copolymer characteristics are summarized in Table 1. The previously obtained copolymer compositions comprising the PSU backbone grafted with two different lengths of phosphonated PFS side chains are also added for comparison. By doubling the grafting sites, and further varying the length of the phosphonated side chains, our aim was to adjust the optimal composition for that copolymer architecture in order to achieve maximum proton conductivity without scarifying the polymer membrane mechanical properties.
364
I. Dimitrov et al. / Journal of Membrane Science 450 (2014) 362–368
3.2. Thermal properties of the graft copolymers The thermal decomposition and glass transition data of the graft copolymers are listed in Table 1. A clear two-step decomposition for the graft copolymers is observed (Fig. 1A). The initial weight loss up to 360 1C might be attributed to the reversible
desorption of water, produced by the condensation of phosphonic acid, as previously reported [27]. Starting at around 360–370 1C (depending on the size and/or the density of the grafts) the C–P bonds are cleaved. The second major weight loss above 400 1C is due to the backbone and the PFS decomposition. All graft copolymers (except one – P14-g-Ph39) revealed Td10% higher than that
Table 1 Compositions, molar-mass characteristics and thermal properties of precursors and graft copolymers. Precursors m
a
PFS-A1 PFS-A2d PFS-A3 PFS-A4
19 31 39 54
PSU-N1d PSU-N2 Nafions NRE212
90 90
Graft copolymers b
x
0.07 0.14
c
Mn (Da)
Mw/Mn
3700 6000 7600 10,500
1.5 1.3 1.3 1.2
43,100 –
1.9 –
c
PSU-g-PhPFS
Tg (1C)
Td
P14-g-Ph19 P7-g-Ph31d P14-g-Ph39 P7-g-Ph54d P14-g-Ph54
204 194 210 365 204
388 398 359 406 390
120
378
10%
(1C)
a
Degree of polymerization. Grafting density. Determined by SEC (vs. polystyrene standards). d Previously synthesized and added for comparison [24]. b c
Fig. 1. (A) TGA curves for phosphonated graft copolymers with different grafting degree and the same length of the side chains (data for native PSU and Nafions NRE212 were added as references); (B) DSC curves for the graft copolymers with 14 mol% grafting degree and different length of the side chains.
Fig. 2. Characteristics of the graft copolymer membranes as a function of temperature: (A) water uptake; and (B) proton conductivity, measured under immersed conditions (data for Nafions NRE212 were added as a reference).
I. Dimitrov et al. / Journal of Membrane Science 450 (2014) 362–368
of the reference Nafions NRE212. From the two copolymers having the same length of the side chains, the more densely grafted copolymer is slightly less thermally stable (Fig. 1A). The DSC measurements of the graft copolymers are presented in Fig. 1B and show that all graft copolymers have glass transition temperatures (Tg) in a close interval above 200 1C (204–210 1C), much higher than that of the Nafions reference sample. The Tgs are most probably due to the PSU main chain [12].
3.3. Graft copolymers membrane properties As noticed in our previous work, the graft copolymers show poor solubility in aprotic solvents at room temperature. Therefore, they were stirred in NMP at 150 1C overnight to dissolve and the membranes were cast at 80 1C. Out of three copolymer samples grafted with 14 mol% phosphonated side chains of different lengths only the one with the shortest grafts and lowest IEC (P14-g-Ph19) was able to form a tough and stable membrane in water. The rest of the membranes with IEC in the 5.4–5.8 meq. g 1 range disintegrated in water. The water uptake of the stable membrane, expressed as λ ¼ [H2O]/[H þ ] was measured under fully immersed conditions vs. temperature and was compared to those of the previously obtained membranes from the phosphonated graft copolymers with half the degree of grafting (Fig. 2A). Due to the high phosphonic acid content in P14-g-Ph19, it shows a steep, increase of the water uptake with increasing the temperature, close to that of Nafion reference sample. That is a clear indication of excessive swelling. Table 2 Graft copolymer membrane properties. Membranes
IEC (meq. g 1)
WUa(wt%)
λb
sc(mS cm 1)
P7-g-Ph31d P7-g-Ph54d P14-g-Ph19 P14-g-Ph39 P14-g-Ph54 Nafions NRE212
3.8e 4.8e 4.3e 5.4e 5.8e 0.9
7.5 188 183 – – 43
11 22 24 – – 26
48 82 134 – – 178
a
Determined at 100 1C under immersed conditions. Number of absorbed water molecules per H þ in the phosphonic acid group at 100 1C. c Measured under immersed conditions at 100 1C. d Previously obtained and added for comparison [24]. e Calculated values assuming complete grafting and 90% phosphonation of the grafts. b
365
The proton conductivity measurements of P14-g-Ph19 membrane were performed under fully immersed conditions in water between 20 and 120 1C (Fig. 2B). At subzero temperatures the graft copolymer shows higher conductivity (less pronounced effect of melting water) than the Nafions sample. At 100 1C the graft copolymer membrane conductivity reaches 134 mS cm 1. That is almost three times as high as that of copolymer P7-g-Ph31 with similar IEC, but less densely grafted PSU backbone (48 mS cm 1). Moreover, the copolymer with 7 mol% of long grafted phosphonated side chains (P7-g-Ph54) with much higher IEC was not able to reach such high conductivity. The membrane properties are summarized in Table 2.
3.4. Preparation and characterization of acid–base polymer blend membranes Due to the high content of phosphonic acid groups in the two graft copolymers with the highest IEC values, and the inability to form stable membranes of these copolymers, we decided to explore the possibility of the preparation of acid–base crosslinked polymer blend membranes. This strategy has previously been pioneered by Kerres and co-workers for the preparation of sulfonated membranes [28]. The graft copolymers were used as acidic blend components and PSU modified with approx. 40 mol% of pyridine side groups (PSU-Py) was used as a basic component. PSU was modified with pendant pyridine groups through lithiation chemistry, and the degree of functionalization was estimated by 1 H NMR spectroscopy. The copolymer blending was performed by mixing two separate polymer solutions in a PSU-g-PhPFS:PSUPy¼80:20 (wt/wt) ratio according to Scheme 1. The blend membranes were cast at 150 1C for 8 h from the combined solutions. The thermal properties of blend membranes are presented in Fig. 3. The thermal stability of the blend membranes is higher than that of the corresponding pure graft copolymers (Fig. 3A). The membranes start to decompose at around 400 1C with Td10% between 393 and 405 1C. All of them are more thermally stable than the reference Nafions sample (Td10% ¼378 1C). Unlike the degradation profile of the pure graft copolymers, the degradation steps are not so well defined. The PSU modified with pyridine side groups (PSU-Py) shows a two-step degradation with Td10% ¼413 1C (Fig. 3A). It also has a Tg at 191 1C, which is close to that of the neat PSU (188 1C). The blend membranes exhibit Tgs that are 4–71 lower than those of the pure graft copolymers (Fig. 3B). Similarly to the pure graft copolymer P14-g-Ph19, its blend with PSU-Py also has an additional Tg at 220 1C.
Scheme 1. Formation of the acid–base polymer blend membranes.
366
I. Dimitrov et al. / Journal of Membrane Science 450 (2014) 362–368
Fig. 3. (A) TGA; and (B) DSC traces of acid–base blend membranes and PSU with 40 mol% pendant pyridine groups – PSU-Py (TGA trace of the graft copolymer P14-g-Ph19 was added for comparison).
Fig. 4. Characteristics of acid-–base blend membranes: (A) water uptake as a function of temperature; (B) proton conductivity measured under immersed conditions as a function of temperature; (C) proton conductivity as a function of relative humidity at 80 1C (data for Nafions NRE212 were added as a reference).
The acid–base blend membranes obtained were stable in water under fully immersed conditions at room temperature and up to 60 1C. Above that temperature they became unstable which is expected since it is known that physically cross-linked membranes might lose their integrity at temperatures above 70 1C [15]. Therefore the water uptake of the blend membranes, expressed as λ, was measured as a function of temperature up to 60 or 80 1C and
presented in Fig. 4A. The water uptake increases moderately with the temperature and IEC, but it is smaller than that of Nafions NRE212. Increased IEC makes the blend membranes P14-g-Ph39-B and P14-g-Ph54-B unstable above 60 1C. The proton conductivities of the blend membranes were measured under fully immersed conditions in water in the temperature interval 20 up to 120 1C and shown in Fig. 4B. It
I. Dimitrov et al. / Journal of Membrane Science 450 (2014) 362–368
367
Table 3 Thermal and membrane properties of the acid–base cross-linked graft copolymers. PSU-g-PhPFS-B
P14-g-Ph19-B P14-g-Ph39-B P14-g-Ph54-B NafionsNRE212 PSU-Py
Copolymer blends
Membranes
Tg (1C)
Td
200 203 201 120 191
393 405 402 378 413
10%
(1C)
IEC (meq. g 1)
WUa(wt%)
λb
sc(mS cm 1)
3.3d 4.2d 4.5d 0.9 0.8d
71 96 111 36 –
12 13 14 22 –
54 51 70 125 –
a
Determined at 60 1C under immersed conditions. Number of absorbed water molecules per H þ in the phosphonic acid group at 60 1C. Measured under immersed conditions at 60 1C. d Calculated values. b c
Fig. 5. TEM images of: (A) graft copolymer P14-g-Ph19; (B) acid–base blend P14-g-Ph19-B (the bright domains are PSU-rich and dark domains are ion-rich).
can be seen that at higher temperatures the conductivity of the blend membranes varies less with the temperature than the Nafions NRE212 membrane. It should be noted that during the measurements some of the membranes became unstable at the highest temperatures. In accordance with the reduced IEC as a result of blending the conductivity of membrane P14-g-Ph19-B is much lower (68 mS cm 1 at 100 1C) compared to that of the pure graft copolymer P14-g-Ph19 (134 mS cm 1 at 100 1C). However, we were able to measure the conductivity of P14-g-Ph39-B and P14-g-Ph54-B blend membranes (Fig. 4B), which was not possible to do with the corresponding pure graft copolymers. All the three blend membranes show similar conductivity dependence on RH at 80 1C (Fig. 4C). The membranes proton conductivity increases with RH. Although, the values are lower than those of the reference sample NRE 212 at that temperature, the blend membranes show improved dimensional stability unlike the pure graft copolymers for which the measurement could not be performed due to their brittleness. The acid–base blends' thermal and membrane properties are summarized in Table 3. The membranes' characteristics under fully immersed conditions, including water uptake, λ and conductivity, are given at 60 1C. At that temperature all the samples were stable during the measurements. TEM images of the pure graft copolymer P14-g-Ph19 and the corresponding acid–base blend P14-g-Ph19B are shown in Fig. 5. A nanophase-separated morphology of ion-rich domains is clearly visible for both samples. At high magnification it seems that the sizes of ion-rich domains are not significantly different for both samples, while the hydrophobic ones are bigger for the blended polymers (Fig. 5B).
4. Conclusions A multistep synthetic strategy for controlled synthesis of graft copolymers comprising PSU backbone and phosphonated PFS side chains with tunable length and density was applied to find an optimal balance between proton conductivity and mechanical stability of the copolymer membranes for potential fuel cells applications. It was found that the graft copolymer with 14 mol% phosphonated short side chains shows high thermal stability and conductivity of 134 mS cm 1 at 100 1C under fully immersed conditions. Compared to other graft copolymers with similar or even higher IEC but with less densely grafted side chains, the above-mentioned copolymer shows superior conductivity. Further increase of the side-chains length leads to membranes with poor mechanical stability. To overcome this problem, acid–base membranes were prepared by blending the phosphonated graft copolymers with pyridine-modified PSU. The cross-linked structure of the membranes imparted higher dimensional stability without sacrifying their thermal stability. However, the membrane from pure graft copolymer performed better than its electrostatically cross-linked analog.
Acknowledgment The financial support from the Danish Council for Strategic Research through contract no. 09-065198 is gratefully acknowledged. We thank Mr. Lars Schulte for obtaining TEM images of the graft copolymers and their acid–base blends.
368
I. Dimitrov et al. / Journal of Membrane Science 450 (2014) 362–368
References [1] B. Lafitte, P. Jannasch, On the prospects for phosphonated polymers as proton exchange fuel cell membranes, in: T.S. Zhao, K.D. Kreuer, T. Van Nguyen (Eds.), Advances in Fuel Cells, vol. 1, 2007, Elsevier, Oxford, 2007, pp. 119–185. [2] S. Bose, T. Kuila, T. Nguyen, N. Kim, K. Lau, J. Lee, Polymer membranes for high temperature proton exchange membrane fuel cell: recent advances and challenges, Prog. Polym. Sci. 36 (2011) 813–843. [3] M. Rikukawa, K. Sanui, Proton-conducting polymer electrolyte membranes based on hydrocarbon polymers, Prog. Polym. Sci. 25 (2000) 1463–1502. [4] Q. Li, R. He, J. Jensen, N. Bjerrum, Approaches and recent development of polymer electrolyte membranes for fuel cells operating above 100 1C, Chem. Mater. 15 (2003) 4896–4915. [5] M. Hickner, H. Ghassemi, Y. Kim, B. Einsla, J. McGrath, Alternative polymer systems for proton exchange membranes (PEMs), Chem. Rev. 104 (2004) 4587–4612. [6] M. Schuster, T. Rager, A. Noda, K. Kreuer, J. Maier, About the choice of the protogenic group in PEM separator materials for intermediate temperature, low humidity operation: a critical comparison of sulphonic acid, phosphonic acid and imidazole functionalized model compounds, Fuel Cells 5 (2005) 355–365. [7] Y. Yang, S. Holdcroft, Synthetic strategies for controlling the morphology of proton conducting polymer membranes, Fuel Cells 5 (2005) 171–186. [8] Y. Elabd, M. Hickner, Block copolymers for fuel cells, Macromolecules 44 (2011) 1–11. [9] S. Subianto, N. Choudhury, N. Dutta, Palladium-catalyzed phosphonation of SEBS block copolymer, J. Polym. Sci. Polym. Chem. 46 (2008) 5431–5441. [10] D. Markova, A. Kumar, M. Klapper, K. Müllen, Phosphonic acid-containing homo-, AB and BAB block copolymers via ATRP designed for fuel cell applications, Polymer 50 (2009) 3411–3421. [11] R. Perrin, M. Elomaa, P. Jannasch, Nanostructured proton conducting polystyrene-poly(vinylphosphonic acid) block copolymers prepared via sequential anionic polymerizations, Macromolecules 42 (2009) 5146–5154. [12] J. Parvole, P. Jannasch, Polysulfones grafted with poly(vinylphosphonic acid) for highly proton conducting fuel cell membranes in the hydrated and nominally dry state, Macromolecules 41 (2008) 3893–3903. [13] M. Ingratta, M. Elomaa, P. Jannasch, Grafting poly(phenylene oxide) with poly (vinylphosphonic acid) for fuel cell membranes, Polym. Chem. 1 (2010) 739–746.
[14] B. Tripathy, V. Shahi, Organic–inorganic nanocomposite polymer electrolyte membranes for fuel cell applications, Prog. Polym. Sci. 36 (2011) 945–979. [15] J. Kerres, Blended and cross-linked ionomer membranes for application in membrane fuel cells, Fuel Cells 5 (2005) 230–247. [16] R. Nolte, K. Ledjeff, M. Bauer, R. Mülhaupt, Partially sulfonated poly(arylene ether sulfone) – a versatile proton conducting membrane material for modern energy conversion technologies, J. Membr. Sci. 83 (1993) 211–220. [17] S. Mikhailenko, K. Wang, S. Kaliaguine, P. Xing, G. Robertson, M. Guiver, Proton conducting membranes based on cross-linked sulfonated poly(ether ether ketone) (SPEEK), J. Membr. Sci. 233 (2004) 93–99. [18] N. Fukuzaki, K. Nakabayashi, S. Nakazawa, S. Murata, T. Higashihara, M. Ueda, Highly phosphonated poly(N-phenylacrylamide) for proton exchange membranes, J. Polym. Sci. Polym. Chem. 49 (2011) 93–100. [19] J. Kerres, A. Ullrich, F. Meier, T. Häring, Synthesis and characterization of novel acid–base polymer blends for application in membrane fuel cells, Solid State Ion. 125 (1999) 243–249. [20] P. Gode, A. Hult, P. Jannasch, M. Johansson, L. Karlsson, G. Lindbergh, E. Malmström, D. Sandquist, A novel sulfonated dendritic polymer as the acidic component in proton conducting membranes, Solid State Ion. 177 (2006) 787–794. [21] W. Li, Y. Fu, A. Manthiram, M. Guiver, Blend membranes consisting of sulfonated poly(ether ether ketone) and polysulfone bearing 4-nitrobenzimidazole for direct methanol fuel cells, J. Electrochem. Soc. 156 (2009) B258–B263. [22] M. Nielsen, A. Yang, K. Jankova, S. Hvilsted, S. Holdcroft, Enhancing the phase segregation and connectivity of hydrophilic channels by blending highly sulfonated graft copolymers with fluorous homopolymers, J. Mater. Chem. A 1 (2013) 8118–8126. [23] P. Jannasch, Fuel cell membrane materials by chemical grafting of aromatic main-chain polymers, Fuel Cells 5 (2005) 248–260. [24] I. Dimitrov, S. Takamuku, K. Jankova, P. Jannasch, S. Hvilsted, Polysulfone functionalized with phosphonated poly(pentafluorostyrene) grafts for potential fuel cell applications, Macromol. Rapid Commun. 33 (2012) 1368–1374. [25] V. Atanasov, J. Kerres, Highly phosphonated polypentafluorostyrene, Macromolecules 44 (2011) 6416–6423. [26] J. Kerres, A. Ullrich, M. Hei, Preparation and characterization of novel basic polysulfone polymers, J. Polym. Sci. Pol. Chem. 39 (2001) 2874–2888. [27] B. Lafitte, P. Jannasch, Phosphonation of polysulfones via lithiation and reaction with chlorophosphonic acid esters, J. Polym. Sci. Polym. Chem. 43 (2005) 273–286. [28] W. Cui, J. Kerres, G. Eigenberger, Development and characterization of ionexchange polymer blend membranes, Sep. Purif. Technol. 14 (1998) 145–154.
Contents lists available at ScienceDirect
Journal of Membrane Science journal homepage: www.elsevier.com/locate/memsci
Proton conducting graft copolymers with tunable length and density of phosphonated side chains for fuel cell membranes Ivaylo Dimitrov a,1, Shogo Takamuku b, Katja Jankova a, Patric Jannasch b, Søren Hvilsted a,n a b
Department of Chemical and Biochemical Engineering, Danish Polymer Centre, Technical University of Denmark, DTU, DK-2800 Kgs. Lyngby, Denmark Department of Chemistry, Polymer and Materials Chemistry, Lund University, P.O. Box124, SE-22 100 Lund, Sweden
art ic l e i nf o
a b s t r a c t
Article history: Received 29 July 2013 Received in revised form 7 September 2013 Accepted 10 September 2013 Available online 20 September 2013
Polysulfones functionalized with highly phosphonated poly(pentafluorostyrene) side chains of different lengths were synthesized applying controlled polymerization and modification methods. The graft copolymers' thermal properties were evaluated by differential scanning calorimetry and thermal gravimetrical analyses. The proton conductivity of membrane prepared from the graft copolymer with the shortest phosphonated side chains was 134 mS cm 1 at 100 1C under fully immersed conditions. The graft copolymer TEM image shows a nanophase separation of ion-rich segments within the polysulfone matrix. Increasing the ionic groups content in the graft copolymers led to extensive membrane swelling. To improve the dimensional stability the graft copolymers were blended with pyridine-modified polysulfone. The blend membranes were transparent with formation of nano-phase domains as revealed from TEM images. The acid–base blend membranes exhibited a slightly higher thermal stability but lower proton conductivity compared to the membranes formed from pure graft copolymers. & 2013 Elsevier B.V. All rights reserved.
Keywords: Graft copolymer Acid–base blend Polymer electrolyte membrane Proton conductivity Thermal stability
1. Introduction Proton conducting polymer electrolyte membranes (PEMs) are one of the key components that determine the cost and performance of polymer electrolyte membrane fuel cells (PEMFCs). In recent years, there has been increased interest in the design and development of novel proton conducting PEMs able to operate at high temperatures (4100 1C) under low humidification conditions [1,2]. Higher operation temperatures result in faster electrode kinetics, reduced risk of catalyst poisoning, smaller heat exchanger and easier water/thermal management [3,4]. The most widely used commercially available PEMs are based on perfluorosulfonic acid (PFSA) ionomers, such as Nafions, produced by DuPont. Despite their good thermal and mechanical properties, the PFSA-based ionomers have several disadvantages including the high cost, decreased performance and dehydration at temperatures above 90 1C [5]. The evaluation of three model compounds containing different protogenic functionalities: sulfonic acid, phosphonic acid and imidazole, for high temperature and low humidity operation has shown that only the phosphonic acid-based system exhibits high conductivity in the temperature range 100–200 1C [6]. The significant strength of the carbon–phosphorus bond makes n
Corresponding author. Tel.: þ 45 45252965; fax: þ 45 4588 2161. E-mail address: [email protected] (S. Hvilsted). 1 Permanent address: Institute of Polymers, Bulgarian Academy of Sciences, 1113-Sofia, Bulgaria. 0376-7388/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.memsci.2013.09.016
the polymers bearing covalently attached phosphonic acid groups attractive for high temperature operation. Moreover, due to the amphoteric nature of phosphonic acid, the membranes are able to conduct protons by diffusion at low humidity, while the proton conductivity is facilitated by the dynamics of the water under immersed conditions. In terms of the copolymer architecture, the ion-containing block and graft copolymers are of significant interest due to their ability to self-assemble in ordered nanostructures that facilitate proton/water transport over a wide range of conditions [7,8]. A number of block and graft copolymers containing phosphonic acid groups have been reported and some of them have also been evaluated for potential fuel cell applications [9–13]. Generally, the proton conductivity and the water retention capacity of the membranes are enhanced when the content of ionic groups is increased [14]. However, upon reaching a certain level, the membranes become highly swollen and even soluble in water. A way to overcome this problem is to cross-link the polymer [15]. The process of cross-linking might be chemical [16–18] or physical [19–21]. The influence of tuning the ionic content of membranes by blending, as opposed to varying the degree of sulfonation, has been evaluated by blending highly sulfonated graft copolymers with fluorous homopolymers [22]. Most often the physical cross-linking is achieved via acid–base interactions between the proton conducting acidic polymer and a suitable basic polymer [23]. Very recently, we introduced a multistep synthetic strategy for the preparation of novel graft copolymers with a polysulfone (PSU)
I. Dimitrov et al. / Journal of Membrane Science 450 (2014) 362–368
backbone and highly phosphonated poly(pentafluorostyrene) (PFS) side chains [24]. The controlled polymerization and modification steps involved allow straightforward tuning of graft copolymer architecture in terms of length and density of the phosphonated side chains. The initial evaluation of copolymer membrane proton conductivity and thermal stability showed promising results. In the present work, we have extended the variety of phosphonated graft copolymers in the search for an optimal balance between high proton conductivity and membrane dimensional stability. We tuned the copolymer architecture to obtain a twice as high grafting degree as that achieved previously. The length of the side chains was also varied in a controlled manner. Furthermore, to improve the membrane dimensional stability, acid–base polymer blends were prepared from the phosphonated graft copolymers and a polysulfone bearing pyridine side groups. The membranes formed from the pure graft copolymers and from the acid–base blends were evaluated for their thermal properties, water uptake and proton conductivity.
2. Experimental 2.1. Materials Graft copolymers with PSU backbone and highly phosphonated PFS side chains of different lengths (PSU-g-PhPFS) were synthesized as previously described [24]. Briefly, PSU (Udels P-3500 LCD MB8, Mn ¼ 40,000 g mol 1, PDI: 1.95, Solvay Advanced Polymes) was lithiated with n-butyllithium in tetrahydrofuran (THF), followed by the addition of (chloromethyl)benzoyl chloride. The targeted degree of functionalization (DS ¼ 0.14, i.e., 14 substitutions per 100 PSU repeating units) was achieved and confirmed by 1H NMR spectroscopy. The benzyl chloride-containing PSU was converted into a multiazide-functionalized polymer (PSU-N) through reaction with sodium azide. Alkyne end-functionalized PFSs (PFSA) of three different number-average molar masses (Mn ¼3700, 7600 and 10,500 Da) were synthesized by atom transfer radical polymerization (ATRP) and “clicked” onto the PSU-backbone via copper catalyzed azide-alkyne 1,3-cycloaddition (CuAAC). Finally, the grafts were phosphonated through reaction with tris(trimethylsilyl) phosphite as described by Atanasov and Kerres [25]. PSU grafted with 40 mol% of basic pyridine side groups (PSU-Py) was synthesized as previously described [26]. 2.2. Membrane preparation The pure graft copolymers (1–2 wt%) were suspended in Nmethylpyrrolidone (NMP) at 80 1C. The temperature was increased to 150 1C and the mixtures were stirred overnight until the polymers dissolved. The membranes were cast at 80 1C during 4 days under nitrogen onto glass plates. In the case of the acid– base blend membranes, a solution of PSU-Py in NMP (2 wt%) was prepared separately at 80 1C. Then, it was slowly added to the copolymer solutions. The resulting mixtures were spread onto glass plates and the membranes were cast in the oven at 150 1C during 8 h. All membranes were detached from the glass plates after immersion in deionized water, followed by repeated rinsing and washing with deionized water. 2.3. Characterization 1
H NMR spectra were recorded in CDCl3 or DMSO-d6 on a Bruker Avance 300 MHz spectrometer. Size-exclusion chromatography (SEC) was carried out with a Viscotek GPCmax VE-2001 equipped with Viscotek TriSEC Model 302 triple detector array (refractive index detector, viscometer detector, laser light scattering detector with the
363
light wavelength of 670 nm and measuring angles of 901 and 71), and a Knauer K-2501 UV detector using two PLgel mixed-D columns from Polymer Laboratories. The samples were run in THF at 30 1C (1 mL min 1). The molar-mass characteristics were calculated using polystyrene standards. Thermal analyses were carried out on a differential scanning calorimeter DSC Q1000 (TA Instruments) in a temperature range of 25–400 1C at a heating and cooling rate of 10 1C min 1 under nitrogen. The glass transition temperatures (Tg) were determined during the second heating cycle at the inflection point of the thermal transition. Thermogravimetric analysis (TGA) was performed on a TGA Q500 instrument from 50 to 600 1C with a heating rate of 20 1C min 1 under nitrogen flow. The ion exchange capacity (IEC, meq. g 1) was calculated assuming 100% efficiency of the “click” grafting, 90% phosphonation of the grafts and considering both protons of each phosphonic acid group. The water uptake (WU) was measured from the weight increase of membranes after soaking them in deionized water at temperatures between 20 and 100 1C by increments of 20 1C to investigate the temperature dependence. It was calculated as: WU (wt%)¼(Wwet– Wdry) Wdry 1 100, where Wwet and Wdry are the membrane weights under hydrated and dry condition respectively. The number of water molecules per H þ in the phosphonic acid group (λ) was calculated as: λ ¼ 1000 (Wwet–Wdry) Wdry 1 (IEC 18) 1. The proton conductivity (S cm 1) of fully immersed membranes was measured in a sealed cell between 20 and 120 1C, using a Novocontrol high resolution dielectric analyzer V 1.01S in the frequency range 10–107 Hz at 50 mV voltage amplitude in a twoprobe method. The humidity dependence of the proton conductivity from 30% to 90% relative humidity (RH) was investigated at 80 1C by a four-probe method with a Gamry potentiostat/galvanostat/ZRA in the frequency range 10 1–105 Hz and a Fumatech MK3 conductivity cell, where the humidity was equilibrated by deionized water in a closed system. The samples for transmission electron microscopy (TEM) were stained with lead acetate and microtomed as previously described [22]. Electron micrographs were obtained with a Tecnai T20 G2 instrument operated at an accelerating voltage of 200 kV.
3. Results and discussion 3.1. Synthesis and characterization of the graft copolymers The synthetic sequence of efficient and controlled polymerization and modification steps allows us to finely tune the copolymer composition in the search for an optimal balance between proton conductivity and mechanical properties of PEMs. We decided to explore the influence of the grafting density as well as of the length of the phosphonated side chains on the copolymer membranes properties. The grafting sites on the PSU backbone were doubled as compared to those obtained in our first work and were set to 14 mol%. Separately, three alkyne end functionalized PFSs (PFS-A) of different lengths were synthesized and grafted onto the PSU backbone through copper catalyzed azide–alkyne 1,3-cycloaddition. The final reaction step was the phosphonation of the PFS side chains. The intermediate and final products were analyzed and characterized as previously described [24]. The precursor and copolymer characteristics are summarized in Table 1. The previously obtained copolymer compositions comprising the PSU backbone grafted with two different lengths of phosphonated PFS side chains are also added for comparison. By doubling the grafting sites, and further varying the length of the phosphonated side chains, our aim was to adjust the optimal composition for that copolymer architecture in order to achieve maximum proton conductivity without scarifying the polymer membrane mechanical properties.
364
I. Dimitrov et al. / Journal of Membrane Science 450 (2014) 362–368
3.2. Thermal properties of the graft copolymers The thermal decomposition and glass transition data of the graft copolymers are listed in Table 1. A clear two-step decomposition for the graft copolymers is observed (Fig. 1A). The initial weight loss up to 360 1C might be attributed to the reversible
desorption of water, produced by the condensation of phosphonic acid, as previously reported [27]. Starting at around 360–370 1C (depending on the size and/or the density of the grafts) the C–P bonds are cleaved. The second major weight loss above 400 1C is due to the backbone and the PFS decomposition. All graft copolymers (except one – P14-g-Ph39) revealed Td10% higher than that
Table 1 Compositions, molar-mass characteristics and thermal properties of precursors and graft copolymers. Precursors m
a
PFS-A1 PFS-A2d PFS-A3 PFS-A4
19 31 39 54
PSU-N1d PSU-N2 Nafions NRE212
90 90
Graft copolymers b
x
0.07 0.14
c
Mn (Da)
Mw/Mn
3700 6000 7600 10,500
1.5 1.3 1.3 1.2
43,100 –
1.9 –
c
PSU-g-PhPFS
Tg (1C)
Td
P14-g-Ph19 P7-g-Ph31d P14-g-Ph39 P7-g-Ph54d P14-g-Ph54
204 194 210 365 204
388 398 359 406 390
120
378
10%
(1C)
a
Degree of polymerization. Grafting density. Determined by SEC (vs. polystyrene standards). d Previously synthesized and added for comparison [24]. b c
Fig. 1. (A) TGA curves for phosphonated graft copolymers with different grafting degree and the same length of the side chains (data for native PSU and Nafions NRE212 were added as references); (B) DSC curves for the graft copolymers with 14 mol% grafting degree and different length of the side chains.
Fig. 2. Characteristics of the graft copolymer membranes as a function of temperature: (A) water uptake; and (B) proton conductivity, measured under immersed conditions (data for Nafions NRE212 were added as a reference).
I. Dimitrov et al. / Journal of Membrane Science 450 (2014) 362–368
of the reference Nafions NRE212. From the two copolymers having the same length of the side chains, the more densely grafted copolymer is slightly less thermally stable (Fig. 1A). The DSC measurements of the graft copolymers are presented in Fig. 1B and show that all graft copolymers have glass transition temperatures (Tg) in a close interval above 200 1C (204–210 1C), much higher than that of the Nafions reference sample. The Tgs are most probably due to the PSU main chain [12].
3.3. Graft copolymers membrane properties As noticed in our previous work, the graft copolymers show poor solubility in aprotic solvents at room temperature. Therefore, they were stirred in NMP at 150 1C overnight to dissolve and the membranes were cast at 80 1C. Out of three copolymer samples grafted with 14 mol% phosphonated side chains of different lengths only the one with the shortest grafts and lowest IEC (P14-g-Ph19) was able to form a tough and stable membrane in water. The rest of the membranes with IEC in the 5.4–5.8 meq. g 1 range disintegrated in water. The water uptake of the stable membrane, expressed as λ ¼ [H2O]/[H þ ] was measured under fully immersed conditions vs. temperature and was compared to those of the previously obtained membranes from the phosphonated graft copolymers with half the degree of grafting (Fig. 2A). Due to the high phosphonic acid content in P14-g-Ph19, it shows a steep, increase of the water uptake with increasing the temperature, close to that of Nafion reference sample. That is a clear indication of excessive swelling. Table 2 Graft copolymer membrane properties. Membranes
IEC (meq. g 1)
WUa(wt%)
λb
sc(mS cm 1)
P7-g-Ph31d P7-g-Ph54d P14-g-Ph19 P14-g-Ph39 P14-g-Ph54 Nafions NRE212
3.8e 4.8e 4.3e 5.4e 5.8e 0.9
7.5 188 183 – – 43
11 22 24 – – 26
48 82 134 – – 178
a
Determined at 100 1C under immersed conditions. Number of absorbed water molecules per H þ in the phosphonic acid group at 100 1C. c Measured under immersed conditions at 100 1C. d Previously obtained and added for comparison [24]. e Calculated values assuming complete grafting and 90% phosphonation of the grafts. b
365
The proton conductivity measurements of P14-g-Ph19 membrane were performed under fully immersed conditions in water between 20 and 120 1C (Fig. 2B). At subzero temperatures the graft copolymer shows higher conductivity (less pronounced effect of melting water) than the Nafions sample. At 100 1C the graft copolymer membrane conductivity reaches 134 mS cm 1. That is almost three times as high as that of copolymer P7-g-Ph31 with similar IEC, but less densely grafted PSU backbone (48 mS cm 1). Moreover, the copolymer with 7 mol% of long grafted phosphonated side chains (P7-g-Ph54) with much higher IEC was not able to reach such high conductivity. The membrane properties are summarized in Table 2.
3.4. Preparation and characterization of acid–base polymer blend membranes Due to the high content of phosphonic acid groups in the two graft copolymers with the highest IEC values, and the inability to form stable membranes of these copolymers, we decided to explore the possibility of the preparation of acid–base crosslinked polymer blend membranes. This strategy has previously been pioneered by Kerres and co-workers for the preparation of sulfonated membranes [28]. The graft copolymers were used as acidic blend components and PSU modified with approx. 40 mol% of pyridine side groups (PSU-Py) was used as a basic component. PSU was modified with pendant pyridine groups through lithiation chemistry, and the degree of functionalization was estimated by 1 H NMR spectroscopy. The copolymer blending was performed by mixing two separate polymer solutions in a PSU-g-PhPFS:PSUPy¼80:20 (wt/wt) ratio according to Scheme 1. The blend membranes were cast at 150 1C for 8 h from the combined solutions. The thermal properties of blend membranes are presented in Fig. 3. The thermal stability of the blend membranes is higher than that of the corresponding pure graft copolymers (Fig. 3A). The membranes start to decompose at around 400 1C with Td10% between 393 and 405 1C. All of them are more thermally stable than the reference Nafions sample (Td10% ¼378 1C). Unlike the degradation profile of the pure graft copolymers, the degradation steps are not so well defined. The PSU modified with pyridine side groups (PSU-Py) shows a two-step degradation with Td10% ¼413 1C (Fig. 3A). It also has a Tg at 191 1C, which is close to that of the neat PSU (188 1C). The blend membranes exhibit Tgs that are 4–71 lower than those of the pure graft copolymers (Fig. 3B). Similarly to the pure graft copolymer P14-g-Ph19, its blend with PSU-Py also has an additional Tg at 220 1C.
Scheme 1. Formation of the acid–base polymer blend membranes.
366
I. Dimitrov et al. / Journal of Membrane Science 450 (2014) 362–368
Fig. 3. (A) TGA; and (B) DSC traces of acid–base blend membranes and PSU with 40 mol% pendant pyridine groups – PSU-Py (TGA trace of the graft copolymer P14-g-Ph19 was added for comparison).
Fig. 4. Characteristics of acid-–base blend membranes: (A) water uptake as a function of temperature; (B) proton conductivity measured under immersed conditions as a function of temperature; (C) proton conductivity as a function of relative humidity at 80 1C (data for Nafions NRE212 were added as a reference).
The acid–base blend membranes obtained were stable in water under fully immersed conditions at room temperature and up to 60 1C. Above that temperature they became unstable which is expected since it is known that physically cross-linked membranes might lose their integrity at temperatures above 70 1C [15]. Therefore the water uptake of the blend membranes, expressed as λ, was measured as a function of temperature up to 60 or 80 1C and
presented in Fig. 4A. The water uptake increases moderately with the temperature and IEC, but it is smaller than that of Nafions NRE212. Increased IEC makes the blend membranes P14-g-Ph39-B and P14-g-Ph54-B unstable above 60 1C. The proton conductivities of the blend membranes were measured under fully immersed conditions in water in the temperature interval 20 up to 120 1C and shown in Fig. 4B. It
I. Dimitrov et al. / Journal of Membrane Science 450 (2014) 362–368
367
Table 3 Thermal and membrane properties of the acid–base cross-linked graft copolymers. PSU-g-PhPFS-B
P14-g-Ph19-B P14-g-Ph39-B P14-g-Ph54-B NafionsNRE212 PSU-Py
Copolymer blends
Membranes
Tg (1C)
Td
200 203 201 120 191
393 405 402 378 413
10%
(1C)
IEC (meq. g 1)
WUa(wt%)
λb
sc(mS cm 1)
3.3d 4.2d 4.5d 0.9 0.8d
71 96 111 36 –
12 13 14 22 –
54 51 70 125 –
a
Determined at 60 1C under immersed conditions. Number of absorbed water molecules per H þ in the phosphonic acid group at 60 1C. Measured under immersed conditions at 60 1C. d Calculated values. b c
Fig. 5. TEM images of: (A) graft copolymer P14-g-Ph19; (B) acid–base blend P14-g-Ph19-B (the bright domains are PSU-rich and dark domains are ion-rich).
can be seen that at higher temperatures the conductivity of the blend membranes varies less with the temperature than the Nafions NRE212 membrane. It should be noted that during the measurements some of the membranes became unstable at the highest temperatures. In accordance with the reduced IEC as a result of blending the conductivity of membrane P14-g-Ph19-B is much lower (68 mS cm 1 at 100 1C) compared to that of the pure graft copolymer P14-g-Ph19 (134 mS cm 1 at 100 1C). However, we were able to measure the conductivity of P14-g-Ph39-B and P14-g-Ph54-B blend membranes (Fig. 4B), which was not possible to do with the corresponding pure graft copolymers. All the three blend membranes show similar conductivity dependence on RH at 80 1C (Fig. 4C). The membranes proton conductivity increases with RH. Although, the values are lower than those of the reference sample NRE 212 at that temperature, the blend membranes show improved dimensional stability unlike the pure graft copolymers for which the measurement could not be performed due to their brittleness. The acid–base blends' thermal and membrane properties are summarized in Table 3. The membranes' characteristics under fully immersed conditions, including water uptake, λ and conductivity, are given at 60 1C. At that temperature all the samples were stable during the measurements. TEM images of the pure graft copolymer P14-g-Ph19 and the corresponding acid–base blend P14-g-Ph19B are shown in Fig. 5. A nanophase-separated morphology of ion-rich domains is clearly visible for both samples. At high magnification it seems that the sizes of ion-rich domains are not significantly different for both samples, while the hydrophobic ones are bigger for the blended polymers (Fig. 5B).
4. Conclusions A multistep synthetic strategy for controlled synthesis of graft copolymers comprising PSU backbone and phosphonated PFS side chains with tunable length and density was applied to find an optimal balance between proton conductivity and mechanical stability of the copolymer membranes for potential fuel cells applications. It was found that the graft copolymer with 14 mol% phosphonated short side chains shows high thermal stability and conductivity of 134 mS cm 1 at 100 1C under fully immersed conditions. Compared to other graft copolymers with similar or even higher IEC but with less densely grafted side chains, the above-mentioned copolymer shows superior conductivity. Further increase of the side-chains length leads to membranes with poor mechanical stability. To overcome this problem, acid–base membranes were prepared by blending the phosphonated graft copolymers with pyridine-modified PSU. The cross-linked structure of the membranes imparted higher dimensional stability without sacrifying their thermal stability. However, the membrane from pure graft copolymer performed better than its electrostatically cross-linked analog.
Acknowledgment The financial support from the Danish Council for Strategic Research through contract no. 09-065198 is gratefully acknowledged. We thank Mr. Lars Schulte for obtaining TEM images of the graft copolymers and their acid–base blends.
368
I. Dimitrov et al. / Journal of Membrane Science 450 (2014) 362–368
References [1] B. Lafitte, P. Jannasch, On the prospects for phosphonated polymers as proton exchange fuel cell membranes, in: T.S. Zhao, K.D. Kreuer, T. Van Nguyen (Eds.), Advances in Fuel Cells, vol. 1, 2007, Elsevier, Oxford, 2007, pp. 119–185. [2] S. Bose, T. Kuila, T. Nguyen, N. Kim, K. Lau, J. Lee, Polymer membranes for high temperature proton exchange membrane fuel cell: recent advances and challenges, Prog. Polym. Sci. 36 (2011) 813–843. [3] M. Rikukawa, K. Sanui, Proton-conducting polymer electrolyte membranes based on hydrocarbon polymers, Prog. Polym. Sci. 25 (2000) 1463–1502. [4] Q. Li, R. He, J. Jensen, N. Bjerrum, Approaches and recent development of polymer electrolyte membranes for fuel cells operating above 100 1C, Chem. Mater. 15 (2003) 4896–4915. [5] M. Hickner, H. Ghassemi, Y. Kim, B. Einsla, J. McGrath, Alternative polymer systems for proton exchange membranes (PEMs), Chem. Rev. 104 (2004) 4587–4612. [6] M. Schuster, T. Rager, A. Noda, K. Kreuer, J. Maier, About the choice of the protogenic group in PEM separator materials for intermediate temperature, low humidity operation: a critical comparison of sulphonic acid, phosphonic acid and imidazole functionalized model compounds, Fuel Cells 5 (2005) 355–365. [7] Y. Yang, S. Holdcroft, Synthetic strategies for controlling the morphology of proton conducting polymer membranes, Fuel Cells 5 (2005) 171–186. [8] Y. Elabd, M. Hickner, Block copolymers for fuel cells, Macromolecules 44 (2011) 1–11. [9] S. Subianto, N. Choudhury, N. Dutta, Palladium-catalyzed phosphonation of SEBS block copolymer, J. Polym. Sci. Polym. Chem. 46 (2008) 5431–5441. [10] D. Markova, A. Kumar, M. Klapper, K. Müllen, Phosphonic acid-containing homo-, AB and BAB block copolymers via ATRP designed for fuel cell applications, Polymer 50 (2009) 3411–3421. [11] R. Perrin, M. Elomaa, P. Jannasch, Nanostructured proton conducting polystyrene-poly(vinylphosphonic acid) block copolymers prepared via sequential anionic polymerizations, Macromolecules 42 (2009) 5146–5154. [12] J. Parvole, P. Jannasch, Polysulfones grafted with poly(vinylphosphonic acid) for highly proton conducting fuel cell membranes in the hydrated and nominally dry state, Macromolecules 41 (2008) 3893–3903. [13] M. Ingratta, M. Elomaa, P. Jannasch, Grafting poly(phenylene oxide) with poly (vinylphosphonic acid) for fuel cell membranes, Polym. Chem. 1 (2010) 739–746.
[14] B. Tripathy, V. Shahi, Organic–inorganic nanocomposite polymer electrolyte membranes for fuel cell applications, Prog. Polym. Sci. 36 (2011) 945–979. [15] J. Kerres, Blended and cross-linked ionomer membranes for application in membrane fuel cells, Fuel Cells 5 (2005) 230–247. [16] R. Nolte, K. Ledjeff, M. Bauer, R. Mülhaupt, Partially sulfonated poly(arylene ether sulfone) – a versatile proton conducting membrane material for modern energy conversion technologies, J. Membr. Sci. 83 (1993) 211–220. [17] S. Mikhailenko, K. Wang, S. Kaliaguine, P. Xing, G. Robertson, M. Guiver, Proton conducting membranes based on cross-linked sulfonated poly(ether ether ketone) (SPEEK), J. Membr. Sci. 233 (2004) 93–99. [18] N. Fukuzaki, K. Nakabayashi, S. Nakazawa, S. Murata, T. Higashihara, M. Ueda, Highly phosphonated poly(N-phenylacrylamide) for proton exchange membranes, J. Polym. Sci. Polym. Chem. 49 (2011) 93–100. [19] J. Kerres, A. Ullrich, F. Meier, T. Häring, Synthesis and characterization of novel acid–base polymer blends for application in membrane fuel cells, Solid State Ion. 125 (1999) 243–249. [20] P. Gode, A. Hult, P. Jannasch, M. Johansson, L. Karlsson, G. Lindbergh, E. Malmström, D. Sandquist, A novel sulfonated dendritic polymer as the acidic component in proton conducting membranes, Solid State Ion. 177 (2006) 787–794. [21] W. Li, Y. Fu, A. Manthiram, M. Guiver, Blend membranes consisting of sulfonated poly(ether ether ketone) and polysulfone bearing 4-nitrobenzimidazole for direct methanol fuel cells, J. Electrochem. Soc. 156 (2009) B258–B263. [22] M. Nielsen, A. Yang, K. Jankova, S. Hvilsted, S. Holdcroft, Enhancing the phase segregation and connectivity of hydrophilic channels by blending highly sulfonated graft copolymers with fluorous homopolymers, J. Mater. Chem. A 1 (2013) 8118–8126. [23] P. Jannasch, Fuel cell membrane materials by chemical grafting of aromatic main-chain polymers, Fuel Cells 5 (2005) 248–260. [24] I. Dimitrov, S. Takamuku, K. Jankova, P. Jannasch, S. Hvilsted, Polysulfone functionalized with phosphonated poly(pentafluorostyrene) grafts for potential fuel cell applications, Macromol. Rapid Commun. 33 (2012) 1368–1374. [25] V. Atanasov, J. Kerres, Highly phosphonated polypentafluorostyrene, Macromolecules 44 (2011) 6416–6423. [26] J. Kerres, A. Ullrich, M. Hei, Preparation and characterization of novel basic polysulfone polymers, J. Polym. Sci. Pol. Chem. 39 (2001) 2874–2888. [27] B. Lafitte, P. Jannasch, Phosphonation of polysulfones via lithiation and reaction with chlorophosphonic acid esters, J. Polym. Sci. Polym. Chem. 43 (2005) 273–286. [28] W. Cui, J. Kerres, G. Eigenberger, Development and characterization of ionexchange polymer blend membranes, Sep. Purif. Technol. 14 (1998) 145–154.