Synthesis and performance evaluation of a polymer mesh supported proton exchange membrane for fuel cell applications

Journal of Membrane Science 350 (2010) 417–426

Contents lists available at ScienceDirect

Journal of Membrane Science journal homepage: www.elsevier.com/locate/memsci

Synthesis and performance evaluation of a polymer mesh supported proton exchange membrane for fuel cell applications Susanta K. Das a,∗ , K.J. Berry a , Hedrick Jamie b , Ali R. Zand b , Lars G. Beholz c a b c

Center for Fuel Cell Systems and Powertrain Integrations, Kettering University, 1700 West Third Avenue, Flint, MI 48504, United States Department of Chemistry, Kettering University, 1700 West Third Avenue, Flint, MI 48504, United States Beholztech, Inc., 132 West First Street, Flint, MI 48502, United States

a r t i c l e

i n f o

Article history: Received 20 July 2009 Received in revised form 6 January 2010 Accepted 9 January 2010 Available online 18 January 2010 Keywords: Proton exchange membrane Fuel cells Fabrication of PEM Proton transfer Performance evaluation

a b s t r a c t A novel approach to the design and fabrication of proton exchange membrane (PEM) has been developed whereby a non-structural polymer fabricated with high proton exchange capacity was bound to an inert polymer matrix. The patented fabrication techniques used here allow greater flexibility in PEM design. Results related to proton exchange performance of these novel PEMs are presented here. The proton exchange material described herein is a ter-polymer composed of various ratios of monomers. These materials were bound to an inert ethylene–tetrafluoroethylene (ETFE) copolymer mesh that had been rendered adhesive using patented hydroxylation technique in a two-step water-borne process. The basic characteristics of the new membranes were compared to those of Nafion® 212. An aqueous two-cell testing unit is utilized by which the rate of protons transferred from one cell through the membrane into the other cell was determined by monitoring the change in pH of the cells. Results indicated that the new membrane could transfer protons approximately 10 times faster per unit area compared to Nafion® 212 under the test conditions utilized at 80 ◦ C. In addition to improvement in induction time and reduced resistance, the new membrane conducts protons at reduced membrane water content compared to Nafion® 212. © 2010 Elsevier B.V. All rights reserved.

1. Introduction The essential requirements of PEMs for fuel cell applications include the following: (i) high proton conductivity, (ii) minimal thickness in order to minimize resistance resulting in fuel cell ohmic drop, (iii) high thermal stability, (iv) excellent mechanical properties (strength, flexibility, and processability), (v) excellent chemical stability, (vi) low water drag, (vii) rapid adjustment of fast kinetics for electrode reactions, (viii) low or minimal gas permeability, and finally (ix) low cost and high availability. Currently, a benchmark commercially produced and widely used proton exchange membrane (PEM) for fuel cell applications is the Nafion® [1–8]. Nafion® is a sulfonated fluoropolymer [1]. Nafion® has a number of limitations such as an operating temperature range of 50–90 ◦ C [2–8], undesirable gas permeability—on the order of 10−6 cm2 /s [6–11] which results in decreased fuel cell performance, limited operational hydration range [1–4,8–10],

∗ Corresponding author at: Department of Mechanical Engineering, Center for Fuel Cell Systems and Powertrain Integrations, Kettering University, 1700 West Third Avenue, Flint, MI 48504, United States. Tel.: +1 810 762 9916; fax: +1 810 762 7860. E-mail address: [email protected] (S.K. Das). 0376-7388/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.memsci.2010.01.019

and high cost, US $800/m2 and US $160/kW [1–3]. To overcome these limitations, a variety of alternative materials for PEM fabrication has been explored. For example, “self-humidifying” poly-tetrafluoroethylene (PTFE) polymers have been developed [2–8]. Furthermore, strong acids such as phosphoric acid, sulfuric acid, and alpha-F sulfuric acid [2–11] have been added to a variety of materials to improve proton conductivity. Among these, the most common proton transfer functional group has been the sulfonate group [1–3]. The proton conductivity of acid-doped membranes can be several orders of magnitude greater than that of Nafion® but unbound acid species are rapidly leached from the membrane in tens to hundreds of hours [2–10]. A common polymer backbone used in fuel cell membrane preparation is polystyrene (PS) [9]. This polymer has been doped with sulfate functionality [3,4] or sulfonated directly [1–4]. Direct sulfonation greatly increases the stability of the sulfonate groups within the PEM and thus greatly increases membrane life expectancy. Hence, this type of membrane has contributed greatly to the use of sulfonated polymers as fuel cell electrolytes [4–6]. These sulfonated PS based PEMs have disadvantages of their own including lower C–H bond (chemical) stability compared to perfluoronated PEMs [1–6]. Sulfonated PS based polymer membranes further suffer from a short lifetime relative to their perfluoronated PEM counterparts. This is due to the sensitivity to oxidation of the

418

S.K. Das et al. / Journal of Membrane Science 350 (2010) 417–426

tertiary C–H bonds in the styrene chains by oxygen and hydrogen peroxide under fuel cell operating conditions. Chemically, the bond strength for C–F is about 485 kJ/mol, much higher than that of C–H bonds (typically 350–435 kJ/mol) or C–C bonds (typically 350–410 kJ/mol) [10]. Polymers containing C–F bonds have also been shown to have high thermal and chemical stability [1,10–12]. For example, PTFEs, consisting of repeated unit of –[CF2 –CF2 ]–, have an excellent chemical stability and a high melting temperature around 370 ◦ C relative to the sulfonated PS based PEMs [10]. In attempts to prepare a more structurally robust PEM, additives such as silica have been used [2–5]. The addition of other additives has sometimes been for the purpose of providing an anchor for the PEM [3–5]. Ethylene–tetrafluoroethylene (ETFE) film has similar properties to the current fluorinated polymers used to prepare PEMs and thus has been used to provide a robust matrix from which sulfonate based PEMs were prepared. These copolymer films are currently prepared through radiation grafting [11,12]. For example, toluene and styrene have been grafted onto ETFE film, with and without divinylbenzene cross-linker [12]. These grafted films were then sulfonated using chlorosulfonic acid to provide proton exchange capability [10–12]. Although in many aspects, these membranes performed at least as good as Nafion® , their manufacture involves the use of both radiation and chlorosulfonic acid and thus will not likely be industrially feasible on a large scale. Although perfluorosulfate based membranes have steadily advanced in fuel cell performance, one major drawback of these membranes is their low conductivity and thus poor cell performance under low humidity and elevated temperatures (above 90 ◦ C) due to water loss [3,4]. Many additional polymer types have been explored for the preparation of PEMs. For example, polysiloxane polymers have been examined [13–16]. Furthermore, aromatic polymers containing a phenylene backbone have been explored [17]. Other fluorinated polymer membranes have also been actively investigated, for example, poly-tetrafluoroethylene-hexafluoropropylene (PTFEP) films [18–20]. The PTFEP film is irradiated first, and then styrene groups are grafted on it with divinylbenzene (DVB) as a cross-linker [19,20]. In this material, the proton conductivity is introduced by sulfonating the aryl groups. A recent work using this type of membrane has reported a fuel cell lifetime over 5000 h at 85 ◦ C [20]. Poly-vinylidene fluoride (PVDF) based polymer membranes have also been developed through grafting and then sulfonating the styrene groups [21–24]. In these PVDF based PEMs, the PVDF provides good physical stability and chemical resistance while the sulfonated polystyrene provides good conductive properties and high water uptake [25–27]. However, high conductivity is only obtained at high levels of sulfonation which also results in high swelling and thus poor mechanical properties, especially at higher operating temperatures [28]. Great efforts have been made to reduce the swelling by ionic cross-linking [29–32]. Another method used to maintain high conductivity while preventing swelling has been to prepare inorganic–organic composites [33–37]. Inorganic–organic composite membranes have been prepared by (i) casting a bulk mixture of powder or colloidal inorganic components with a polymer solution, or (ii) in situ formation of inorganic components in a polymer membrane or in a polymer solution. The bulk polymer solution enables the nano-sized particles to be dispersed in the formed membranes [38]. Some of these composite membranes exhibit promising conductivities at temperatures above 100 ◦ C [17,39]. However, most of these composite membranes have not been tested in fuel cell applications. Acid–base polymers are another class of proton exchange membranes which provide good performance at high temperatures [17,40]. One of these, ionically cross-linked phosphoric acid-doped poly[2,2 -(m-phenylene)-5,5 -benzimidazole] (PBI) has recently

received much attention [30,41]. In these acid–base polymer membranes, basic polymers can be doped with an amphoteric acid, which acts as both a proton donor and an acceptor and therefore allows for proton migration. It was found that the base protonation and strong hydrogen bridging in acid–base blend membranes markedly reduce polymer swelling [29]. The resulting acid–base blends constitute a new class of proton conducting membranes with high conductivity, thermal stability, and mechanical flexibility and strengths [41]. In general, polymers bearing basic sites such as ether, alcohol, imine, amide or imide groups react with strong acids such as phosphoric acid or sulfuric acid. The basic components of these polymers enable the establishment of hydrogen bonds with the acid. Because of their unique proton conduction mechanism by self-ionization and self-dehydration, H3 PO4 and H2 SO4 exhibit effective proton conductivity even in an anhydrous form. When basic components are present, the interaction between these and the acids increases the extent of the acid dissociation relative to that of the anhydrous acids [42]. Most of the studied acid/polymer systems are not entirely anhydrous, as water is present as a necessary plasticizer for improving conductivity and mechanical properties [42]. When doped with acids as well as strong bases, the acid–base polymeric electrolyte has been proposed for fuel cell membrane electrolytes at temperatures above 100 ◦ C [43,44]. Acid–base membranes have been cast from solutions of different concentrations of organic solvents, for example, 2.5–3.0% solution in a mixture of NaOH and ethanol [45]. The cast membrane is then doped with the acid in order to obtain sufficient conductivity [42]. Sulfonated polysulfone/PBI (SPSF/PBI) membranes doped with phosphoric acid have been investigated and shown to exhibit excellent chemical and thermal stability as well as good proton conductivity [46]. Using these materials, good fuel cell performance has been achieved at a temperature of 110 ◦ C and pressure of 1.5 atm [47,48]. The objective of the present study is to develop a novel, efficient and inexpensive proton exchange membrane (PEM) for fuel cell applications that comprises a proton exchange medium cast onto a robust polymer mesh to form the hybrid PEM. In this manner, many of the membrane’s stability requirements (i.e. mechanical and chemical) can be met with the polymer mesh support while the proton exchange media may be tailored to most efficiently exchange protons. The polymer mesh used here is a surface modified ethylene–tetrafluoroethylene (ETFE) copolymer, which has similar properties to fluorinated polymers [1–5]. In this study, a novel water-borne chemical procedure was used to functionalize ETFE mesh onto which were cast varying compositions of random ter-polymers comprising acrylic acid, styrene and vinylsulfonic acid. To evaluate the performance of these hybrid PEMs, an easy and inexpensive test method was developed. Details of materials and methods are discussed in Section 2. Experimental details for performance evaluation of the novel membranes are presented in Section 3. Results and discussions are provided in Section 4 and finally conclusions, based on the results obtained for the new membrane compared to Nafion® 212 membrane, are drawn in Section 5. 2. Materials and methods 2.1. Materials Ethylene–tetrafluoroethylene (ETFE) mesh (70 ␮m nominal aperture, 66.7 threads/in.2 , 70 ␮m monofilament diameter, 21% open area) was purchased from Goodfellow Corporation, Oakdale, PA, U.S.A. 15% sodium hypochlorite and 85% phosphoric acid were purchased from PVS Nolwood, Detroit, MI, U.S.A. All other chemicals and polymers were purchased from Sigma–Aldrich Corporation, St. Louis, MO, U.S.A. unless otherwise indicated.

S.K. Das et al. / Journal of Membrane Science 350 (2010) 417–426

419

2.2. Manufacturing procedures The manufacturing process consists of three steps: (i) surface modification of ETFE mesh by chemical treatment, (ii) preparation of proton exchange polymer and finally (iii) casting the proton exchange polymer onto the surface modified ETFE mesh to develop the hybrid PEM. Firstly, the ETFE mesh was chlorinated by immersion in 15% sodium hypochlorite into which phosphoric acid was carefully added until constant light bubbling was achieved. The solution was then stirred overnight. After chlorination, the mesh was rinsed with water and subsequently hydroxylated by placing in an aqueous 1 M sodium hydroxide (NaOH) solution overnight. This method resulted in the preparation of a hydroxylated surface on the ETFE mesh rendering it adhesive or optional chemically reactive as shown in Fig. 1. Secondly, the proton exchange polymer was prepared by adding the desired proportions of monomers—acrylic acid, styrene, and vinylsulfonic acid to a vial followed by a small amount of benzoyl peroxide to act as a radical initiator. For example (exact ratios are not mentioned here for proprietary reasons), 5 ml of acrylic acid, 8 ml of styrene, 0.1 ml of vinylsulfonic acid and 11.0 mg of benzoyl peroxide would form a typical membrane cocktail. Enough ethanol was then added to achieve a homogeneous solution. This solution was then very slowly heated in a sand bath to a temperature between 100 ◦ C and 110 ◦ C. The reaction time, monomer concentration and percentage of conversion

Fig. 2. (a) General structure of proton exchange polymer cast onto the ETFE mesh and (b) the final product of SAS proton exchange membrane.

during polymerization are given in Table 1. After polymerization was complete, the solution was removed from the sand bath and allowed to cool at ambient temperature (25 ◦ C). After cooling, the polymer was isolated and then re-dissolved in ethanol to prepare the solution cocktail into which the hydroxylated ETFE mesh was placed. The ETFE mesh now coated with the proton exchange polymer was then spun dried in a centrifuge. The new styrene–acrylic acid–vinylsulfonate (SAS) proton exchange polymer will subsequently be referred to as “SAS” polymer. This casting procedure was repeated with drying in between each casting step. The new SAS PEM has the general structure indicated in Fig. 2a wherein the mechanically fragile SAS polymer matrix that exchanges protons, denoted by light green, was cast onto the mechanically stable hydroxylated ETFE mesh to form the final SAS PEM shown in Fig. 2b. Since only the surface structure of the ETFE was modified through patented chemical processing techniques, it was anticipated that the ETFE mesh would provide the structural support of the new SAS PEMs whereas the SAS proton exchange polymer would act as an efficient proton exchange medium. Initial tests demonstrated [49] that increased sulfonate functionality (increased vinylsulfonic acid) to the polymer backbone provides increased proton conductivity by allowing the formation of proton migration channels throughout the polymer membrane. 2.3. Casting methods In this study, various methods were examined to enhance the homogeneity of the proton exchanging polymer cast onto the ETFE mesh. Different casting methods were explored. These methods are described below. Table 1 Reaction time, monomer concentration and percentage of conversion during polymerization. Name of monomer

Fig. 1. Schematic of ETFE functionalization processes.

Acrylic acid Styrene Vinylsulfonic acid

Reaction time

Concentration

% of conversion

24.3 h

40% 100% 70%

27.8 26.4 45.8

420

S.K. Das et al. / Journal of Membrane Science 350 (2010) 417–426

2.3.1. Dip casting In this method, the proton exchange polymer was initially dissolved in a solvent, usually ethanol, and the treated ETFE mesh was then dipped into this solution. The solution was then heated to 75 ◦ C for 10 min after which the mesh was removed and allowed to air dry. Further drying was achieved through microwave irradiation. The newly cast membrane was examined under a microscope relative to the untreated ETFE mesh to ensure the presence of the proton exchanging polymer. 2.3.2. Centrifuge casting In this procedure, the treated ETFE mesh was suspended inside a centrifuge tube containing proton exchanging polymer in ethanol solution. The solution level alternately covered only half of the total mesh surface area (called partially filled centrifuge method) or the solution level covered the entire mesh surface area (called filled centrifuge method). This was centrifuged at 1000 rpm for 5 min. The cast mesh (PEM) was removed and allowed to air dry. The PEM was further dried by additional centrifugation in an empty tube. The PEM was examined under a microscope as previously described. 3. Experimental measurement of membrane performance 3.1. Measurement of proton transfer rate and membrane resistance In this study, an aqueous two-cell testing unit (schematic is shown in Fig. 3) was used wherein the rate of proton transfer from one cell through the membrane into the other cell was determined by monitoring the change in pH of the cells. The proton exchange membrane (PEM) was placed into the PEM holder between the two cells. To ensure that the transfer of protons from one cell to the other did not occur without passing through the PEM, silicone grease was applied around the outer edge of the PEM holder. First, an equal amount of de-ionized water was added to each cell. Then the cells were allowed to equilibrate so that the water on each side of the membrane achieved the same depth. A pH meter was fitted into each cell to accurately record the pH readings. After the pH meters reached equilibrium, simultaneously 20 drops of deionized water were added to the left cell and 20 drops of 20% HCl solution were added to the right cell using the same size pipette. This procedure was used to ensure the volume of liquid in each cell remained the same so that there would be no liquid forced through the membrane by pressure differences (i.e. avoided pressure driven flow). A half-cell voltage circuit with 1.3583 V for Cl− [41,42] was established between the water and acid cells in order to maintain

Fig. 4. Membrane resistance test apparatus: two-cell method. Acid cell, water cell and test membrane sections are labeled and indicated by arrow sign. pH meter is attached to each of the cells and pH is recorded at regular time intervals through the monitor.

electrical neutrality. The half-cell circuit is commonly used in electroplating. By establishing half-cell reaction circuit between the cells we assured high proton concentrations in the acid cell. The chloride ion in the acid cell becomes chlorine gas and the protons transferred in the water cell become hydrogen gas with the help of half-cell voltage circuit. The half-cell voltage circuit electron activity is shown below: 2HCl ↔ 2H+ + 2Cl− 2Cl− → 2e− + Cl2

Acid cell :

Water cell : H+ + 2e− → H2 Fig. 4 illustrates the complete experimental set-up to measure change in pH of the cells. The implementation of the half-cell circuit is clearly demonstrated in Fig. 4. The initial pH and temperature of both cells were recorded immediately. Since there will be a negative gradient in the concentrations of protons between two cells, protons should be moved from the acid cell through the PEM to the water cell. pH readings of both acid cell and water cell were then taken at regular intervals, usually every 2–5 min, until constant pH values were achieved in both cells. All measurements were carried out without stirring under quiescent conditions. After collecting sufficient readings, the experiment was terminated. The experiments were repeated with the same set-up and initial conditions to test the performance of each of the membrane types reported here. The total proton concentration was calculated using the theoretical model based on the rate of transfer of protons in the water cell through the membrane [49] given as:



Cw (t) = Cf 1 − 0.3679

Fig. 3. Schematic of water-borne two-cell unit with PEM holder used to test proton exchange capacity. PEM indicated in purple color at the center. (For interpretation of the references to color in this figure caption, the reader is referred to the web version of the article.)

 C − C  0 f Cf

(1)

where C0 and Cf are the values of proton concentration at time t = 0 and at time t = ∞, respectively, and Cw (t) is the proton concentration in the water cell (see Fig. 4). The values of C0 and Cf were determined from the experiment shown in Fig. 4. Finally, the exact resistance of the membrane was calculated by subtracting the solution’s resistance from the total resistance [49]

S.K. Das et al. / Journal of Membrane Science 350 (2010) 417–426

as given by: R=



r−

1 am

−

1 mw



= 2t = T

421

(2)

where R is the membrane resistance, r is the total resistance,  aw and  mw are the interfacial/solution resistances, t is the instant of time and T is the total time. 3.2. pH measuring process and calculation of proton concentration (a) At the beginning of the experiment, the pH of both the acid cell and water cell was measured. (b) Once the experiment was started, the measurement of pH was done at regular intervals of time. (c) The experiment was terminated after the pH of both cells had reached equilibrium. (d) The final value of pH in each cell (acid and water) was recorded. (e) The concentration of protons was then obtained using the relation: [H+ ] = 10−pH . 4. Results and discussions The patented technology [50,51], used in this study as described in Fig. 5, resulted in the attachment of hydroxyl groups onto the polyethylene portions of the ETFE copolymer onto which the proton exchange matrices were attached. As shown in Fig. 5, according to Refs. [50,51], through BT process-A (details are not discussed due to proprietary issues) side chains of the polymer surface can be replaced by atoms or other functional groups just as an anchor. On the other hand, following the BT process-B side chains of the polymer surface can be modified in order to attach other species (i.e. desired chemical species) onto the polymer backbone. The properties of new SAS PEM were compared with commercially manufactured Nafion® 212. In this study, we compared properties of two SAS membranes, SAS type I and SAS type II, with those of Nafion® 212. The differences between the SAS type I and SAS type II are the compositions of monomer ratios (not discussed here due to proprietary issues). The final thickness of both of the SAS membranes reported herein is 50.4 ␮m which is comparable to the thickness of Nafion® 212 of 50.8 ␮m. The change of pH in the water cell as a function of time without placing membrane between the two cells in order to test diffusion rate of protons (conduction of protons) is discussed in Ref. [49]. A complete profile of change of pH in the water cell [49] clearly

Fig. 5. A brief schematic of the patented technology [50,51] used in this study to manufacture the SAS proton exchange membrane.

Fig. 6. Experimental results for the change of pH in water cell as a function of time with different membranes at 80 ◦ C. Linear regression equations at each of the three distinct phases, induction phase, transfer phase and equilibrium phase, are shown in the onset. The slopes of the curves indicate the rate of change of pH in each phase.

displays three distinct phases—an induction phase, a proton transfer phase and an equilibrium phase. Fig. 6 represents the change in pH in the water cell at 80 ◦ C for different membranes tested here as a function of time. Since the same initial concentration of protons was used for each of the membrane trials, the initial pH was the same for each trial as can be seen from Fig. 6. Comparing the results of proton concentration without placing the membrane [49] and Fig. 6 revealed the variations in the profiles of proton concentration in the water cell due to the presence of the membrane which leads to quantify proton transfer rate through the membrane. The profiles of rate of pH change in the water cell (see Fig. 6) provided a curve consisting of three segments [49]: an initial slightly negative slope representing an induction phase, a greater negative slope denoting a transfer phase and then a final slightly negative slope denoting an equilibrium phase. The entire profile of the rate of pH change curve was then separated into three separate curves, one for each region of different slopes, and a linear regression line fitted to each curve segment with corresponding linear equations. The induction phase is used to determine the time required for each of the membrane to start the transfer of proton to the water cell. The proton transfer phase was used to determine the concentration of protons (i.e. total amount of protons) transferred into the water cell per minute and the time from initiation of proton transfer into the initially water-only cell till attainment of the equilibrium pH between the two cells. The slope of the middle, strongly negative curve, proton transfer phase curve, represents the maximum rate of proton transfer per minute and can be calculated as the “change in negative logarithm of the hydrogen ion (proton) concentration per minute” or log[H+ ]/min. The two intersections of the three curves, shown in Fig. 6, were then obtained to provide the onset of protons crossing the membrane and the attainment of pseudo-equilibrium [49]. The slope was then converted to moles of protons to obtain the “change in concentration of protons per minute” and the final proton concentration is multiplied by 6.02 × 1023 (Avogadro number) to obtain the actual number of protons transferred through the membrane in the water cell per minute. The proton concentration thus obtained was then inverted to obtain the time required per mole of protons and the time required per proton to pass through the membrane, respectively.

422

S.K. Das et al. / Journal of Membrane Science 350 (2010) 417–426

Fig. 7. Slopes (rate of change) of pH profiles in water cell as a function of time with different membranes at 80 ◦ C. SAS membranes show very sharp decrease of slopes in the transfer phase compared to Nafion® 212 membrane.

Fig. 8. Concentration profiles of proton flow in the water cell at 80 ◦ C as a function of time for different membranes. Symbols represent experimental results. Solid, dashed and dotted lines represent theoretical model predictions for different membranes.

Using the regression line equations, given in Ref. [49], y1 = m1 t + d1 ,

(3)

y2 = m2 t + d2 ,

(4)

Among the profiles of proton flow for different membranes, SAS type I PEM has the highest peak and Nafion® 212 has the lowest peak. This indicates that the SAS type I PEM is able to transfer more protons per unit time than the Nafion® 212 membrane at 80 ◦ C.

and y3 = m3 t + d3 ,

(5)

where m1 , m2 , and m3 represent the rates of pH change in the water cell at each of the three distinct phases; induction phase, transfer phase and equilibrium phase, respectively. The rate of change (slope) of pH profiles for each of the tested membrane at 80 ◦ C was then calculated. The rate of change of pH profiles in the water cell, obtained from Fig. 6, is presented in Fig. 7. In Fig. 7, it can be seen that the induction phase provided a steady constant slope while significant slope variations were obtained in the proton transfer phase. Finally, a steady constant slope was again obtained in the equilibrium phase and it signaled the attainment of pseudo-equilibrium of proton concentration between the acid cell and the water cell. The slope characteristics of the pH profiles in the water cell described how the proton conduction is taking place between the two cells. From Fig. 7, we can see the sharp decrease in slopes throughout the transfer phase for both of the SAS membranes compared to Nafion® 212 membrane. Using the rate of pH change in the water cell in Fig. 8 and the relation given in the equation below [49]: Concentration of protons : [H+ ] = 10−pH ,

(6)

we calculated moles of protons transferred through the membrane and obtained the concentration of protons per minute for each of the membranes examined. Fig. 8 represents the concentration profiles of proton flow as a function of time in the water cell for various membranes. This was obtained experimentally by placing the membrane between the two test cells (shown in Fig. 4) and the corresponding theoretical results were obtained using the theoretical model, given in Eq. (1). In this calculation, we used the values of C0 and Cf obtained experimentally through the rate of pH change as shown in Fig. 7. In Fig. 8, both the experimental and theoretical results are presented. As can be seen, there is an excellent agreement between the experimental and theoretical profiles of proton flow (see Fig. 8) in all the three phases. The peak of the profiles of proton flow represents the maximum rate of proton transfer.

4.1. Membrane conductivity at different temperatures To determine the stability of the SAS membranes in hydrolytic conditions, i.e. the fuel cell operating conditions, the membranes were immersed into water at different temperatures for 24 h and then their proton transfer capacity was tested. The proton transfer capacity for each of the membranes tested here was determined in the range of 25–90 ◦ C in order to judge the membrane conductivity at different temperatures. The experimental procedures discussed in Section 3 were followed to examine the proton transfer capacity for each of the membranes at different temperatures. This experimental test was done by placing the membrane sandwiched between the conductive media as shown in Fig. 4. The reported values are the mean of at least five experiments. The standard deviation from the mean values of proton transfer capacity was ±2%. Fig. 9 shows the maximum proton transfer capacity for each of the membranes examined. Maximum proton transfer capacity is determined using the highest peak slope shown in Fig. 7 at different temperatures examined in this study. From Fig. 9, it can be seen that both SAS type membranes were able to transfer higher numbers of protons in the temperature range 25–90 ◦ C than Nafion® 212 membrane. The maximum proton transfer rate increases as the temperature increases for all the membranes. Proton transfer rate through the membranes was almost constant between 25 ◦ C and 60 ◦ C (Fig. 9). The SAS type membranes started transfer of protons rapidly after 60 ◦ C compared to Nafion® 212 membrane. At 80 ◦ C, the SAS type I membrane provided the highest proton transfer rate, approximately 18 mol/min, compared to the Nafion® 212 membrane, approximately 1.8 mol/min. This indicates that SAS type I membrane had 10 times faster proton transfer capability than Nafion® 212 membrane at low temperature PEM fuel cell operating conditions, i.e. at 80 ◦ C. Fig. 9 also shows a slow decrease in proton concentration profiles after 80 ◦ C for SAS type I membrane as temperature increases. Since the protons present in the water cell are in the form of H3 O+ and not simply H+ , it is not known

S.K. Das et al. / Journal of Membrane Science 350 (2010) 417–426

Fig. 9. Maximum proton transfer capacity among membranes at different temperatures. Dashed, dotted and solid lines represent experimental results and symbols represent theoretical results [49].

immediately what the significance of this shifted trend would be after 80 ◦ C especially when considering a hydrogen fuel cell. Thus, further investigation is needed to understand the trend. Fig. 10 represents the minimum time required for the protons to pass through the membrane at different temperatures. The minimum time required for each membrane was determined by the difference in time it just started to transfer protons into the water cell and the time it takes to transfer protons at its highest capacity, determined by the peak in the concentration profiles at different temperatures as shown in Fig. 8 at 80 ◦ C. Comparing the results given in Fig. 10 suggests that SAS type I membrane took 71% less time at 25 ◦ C and 87% less time at 80 ◦ C compared to the Nafion® 212 membrane to transfer a mole of proton. Both the experimental and theoretical results are presented in Fig. 10, where dotted, dashed and solid lines denote experimental results and symbols represent theoretical results [49]. Both experimental and theoreti-

Fig. 10. Minimum time required for protons to pass through the membrane at different temperatures. Dotted, dashed and solid lines represent experimental results and symbols represent theoretical results [49].

423

Fig. 11. Induction time required for protons to pass through the membrane at different temperatures. Dotted, dashed and solid lines represent results for SAS type I, SAS type II and Nafion® 212 membranes, respectively.

cal results are in good agreement. Fig. 11 shows the induction time required for protons to begin passing through the membrane at different temperatures and was calculated using Ref. [49]: t1 =

d2 − d1 . m1 − m2

(7)

From Fig. 11, it can be determined that at low temperature (i.e. 25 ◦ C), Nafion® 212 membrane has a higher induction time compared to SAS type membranes. The induction time decreases with increasing temperature. At 80 ◦ C, the induction time of Nafion® 212 is 2 times longer than that of the SAS type I membrane (Fig. 11). This indicates that the SAS type I membrane is able to start transferring protons 2 times earlier compared to Nafion® 212 at 80 ◦ C. 4.2. Relative resistance of membranes Fig. 12 provides the minimum relative resistance between the membranes at different temperatures. The theoretical relative resistance was calculated according to the theoretical model [49]. According to the theoretical model [49], the membrane’s relative resistance is directly proportional to the total time taken by the membrane to allow a specific amount of protons to pass through it. Using the concentration profiles of protons in the water cell, as shown in Fig. 8 at 80 ◦ C, we evaluated the relative membrane resistance at different temperatures. The minimum membrane resistance (Rmin ) was calculated from the time it took each membrane to transfer a mole of proton at its peak transfer rate at different temperatures. The results in Fig. 12 suggest that the relative resistance for Nafion® 212 is 72% higher than the SAS membrane at 25 ◦ C and 87% more at 80 ◦ C. Membrane resistance has a great impact on the performance of low temperature proton exchange membrane fuel cell since high membrane resistance causes a drop in fuel cell’s overall ohmic voltage [52]. Among the membranes reported in this study, the SAS type I membrane has the lowest resistance and Nafion® 212 has the highest resistance at all temperatures (Fig. 12). Fig. 13 represents the 4-electrode conductivity measurement at different relative humidity (RH) levels at 30 ◦ C, 80 ◦ C and 120 ◦ C using industry-standard test protocol performed by BekkTech, a commercial third party membrane conductivity measurement service provider. From Fig. 13, we can see that our SAS PEM

424

S.K. Das et al. / Journal of Membrane Science 350 (2010) 417–426

Fig. 12. Membrane’s minimum relative resistance at different temperatures. Dashed, dotted and solid lines represent experimental results and symbols represent theoretical results [49].

(sample—SAS type I) is able to conduct proton at different RH levels especially at 80 ◦ C but required further improvement in terms of proton conductivity at a reduced humidity condition, i.e. 50% RH. From Fig. 13, we see that our SAS membrane performance is well comparable with that of the commercial membrane Nafion® 212 at 80–90% RH cycling at 80 ◦ C but fell far below our laboratory test results presented in Fig. 9. We refined our SAS membrane further according to the procedures described above and tested the performance using industry-standard test protocol. The results closely validated the theoretical results and laboratorybased results reported here in Figs. 9–12. Due to proprietary issues and patenting restrictions, further details of membrane performance using industry-standard test protocol are not documented here. Details of SAS membrane’s structural characterization results will be published shortly elsewhere. 4.3. Water uptake content The swelling characteristics of the SAS membranes were determined by water uptake measurements. The membrane samples

Fig. 13. Direct 4-electrode conductivity measurement results of SAS type membrane (sample—SAS type I) with different relative humidity (RH) levels at 30 ◦ C, 80 ◦ C and 120 ◦ C using industry-standard test protocol.

Fig. 14. Membrane’s water uptake content at different temperatures. Dotted, dashed and solid lines represent SAS type I, SAS type II and Nafion® 212 membranes, respectively.

were dried, weighed and soaked in de-ionized water for 24 h at temperatures ranging from 25 ◦ C to 90 ◦ C. The membranes were then blotted dry and further air dried after which they were re-weighed. This procedure was repeated at least 5 times until a satisfactory reproducibility was obtained. The water uptake content was calculated using the following relationship: Water uptake content, w (%) =

wwet − wdry wdry

× 100,

(8)

where wdry and wwet are the weights of the dry and wet membrane samples, respectively. Fig. 14 shows the water uptake content of membranes at different temperatures. Dotted, dashed and solid lines represent the SAS type I, SAS type II and Nafion® 212 membranes, respectively. From Fig. 14, we see that the water uptake content for Nafion® 212 membrane is very high and increased almost linearly with increasing temperature compared to SAS type membranes. This implies that to conduct proton transfer efficiently, Nafion® 212 requires more water (i.e. higher humidity level) than the SAS type membranes. It appears that SAS type membranes are capable of transferring protons efficiently at low water content than Nafion® 212. The dependence of the liquid water uptake could have significant implications for the use of a membrane in PEMFCs [52]. For example, in one common mode of fabrication of membraneelectrode assemblies (MEAs), the membrane and electrodes are hot-pressed together at higher temperature (e.g. at 120 ◦ C). During this process, all water is lost from the membrane and the operational temperature to which the membrane is exposed (e.g. 80 ◦ C) could result in incomplete rehydration. If less water is taken up by the membrane, a decrease in the maximum attainable conductivity would occur since the conductivity depends strongly on membrane water content and hence on the membrane’s relative humidity. In general, it is assumed that the membrane will regain the required hydration level once soaked into the water prior to use in the fuel cell environment. A rigorous relative humidity (RH) cycle measurement will be studied in the near future to verify the performance of these membranes in terms of conductivity, resistance, temperature, relative humidity and other parametric conditions.

S.K. Das et al. / Journal of Membrane Science 350 (2010) 417–426

5. Conclusions The purpose of the research reported here is the development of a complete manufacturing procedure to fabricate a novel PEM, for fuel cell applications, that has the potential to be more efficient, robust and less expensive than Nafion® . The objectives of improved proton conductivity through a proton exchanging polymer matrix cast onto a very mechanically and chemically stable and robust ETFE mesh were achieved by separating different PEM requirements and distributing them among different PEM polymer matrices. In the present study, many of the structural and mechanical requirements were met by the ETFE mesh while the proton exchange polymer medium was designed to optimize proton transfer capacity. Although the materials used to prepare these hybrid membranes were purchased from specialty chemical and polymer companies, the cost of ETFE used in the current study was approximately $353 m−2 and that of the chemicals was under $20.00 m−2 . Manufacturing costs are also relatively low and the preparation and casting of the membrane cocktail onto the mesh did not require any expensive special equipment. It is thus anticipated that large scale manufacture of SAS type PEMs will be significantly less expensive than the manufacture of Nafion® -type membranes. The results of the current study also indicate that the hybrid SAS type proton exchange polymer matrix may be tailored to meet a range of low temperature fuel cell requirements. In this study, we have shown that the basic requirements such as proton exchange capacities, relative resistance, temperature variation effect, induction time and water uptake content are significantly improved in SAS type membranes, especially SAS type I membrane, compared to Nafion® 212 membrane. A detailed experimental set-up to validate the performance with the theoretical model has also been developed and described. The results also show that the theoretical model predictions are in excellent agreement with the experimental observations. These initial laboratory-based results reported here for the new SAS type PEMs are promising and further characterization of new SAS type PEM using industrystandard commercial equipments is presently underway and will be reported in future work. Acknowledgement This work is accomplished under the funding support provided by the U.S. Department of Energy (DOE) grant award number GO86056. References [1] Nafion® , E.I. DuPont de Nemours and Company, U.S.A., 2004. [2] H.L. Yeager, A. Steck, Cation and water diffusion in Nafion ion exchange membranes influence of polymer structure, J. Electrochem. Soc. 128 (9) (1981) 1880–1884. [3] C. Yang, S. Srinivasan, A.B. Bocarsly, S. Tulyani, J.B. Benziger, A comparison of physical properties and fuel cell performance of Nafion and zirconium phosphate/Nafion composite membranes, J. Membr. Sci. 237 (2004) 145–161. [4] M. Doyle, L. Wang, Z. Yang, S.K. Choi, Polymer electrolytes based on ionomer copolymers of ethylene with fluorosulfonate functionalized monomers, J. Electrochem. Soc. 150 (2003) D185. [5] K.D. Kreuer, On the development of proton conducting polymer membranes for hydrogen and methanol fuel cells, J. Membr. Sci. 185 (2001) 29. [6] N.H. Jalani, R. Datta, The effect of equivalent weight, temperature, cationic forms, sorbates, and nanoinorganic additives on the sorption behavior of Nafion, J. Membr. Sci. 264 (2005) 167–175. [7] R. Lawson, C. Wang, J. Hong, J. Ma, B. Fang, D. Chu, Nafion–bimevox composite membrane for fuel cell applications, J. Electrochem. Soc. 154 (2007) B48–B52. [8] T.A. Zawodzinski, J. Davy, J. Valerio, S. Gottesfeld, The water content dependence of electro-osmotic drag in proton-conducting polymer electrolytes, Electrochim. Acta 40 (1995) 297. [9] L.J.M.J. Blomen, M.N. Mugerwa (Eds.), Fuel Cell Systems, Plenum Press, New York, 1993, p. 614. [10] S. Srinivasan, B.B. Dave, K.A. Murugesamoorthi, A. Parthasarathy, A.J. Appleby, Overview of fuel cell technology, in: L.J.M.J. Blomen, M.N. Mugerwa (Eds.), Fuel Cell Systems, Plenum Press, New York, 1993, pp. 37–72.

425

[11] M. Doyle, M.E. Lewittes, M.G. Roelofs, S.A. Perusich, Ionic conductivity of nonaqueous solvent-swollen ionomer membranes based on fluorosulfonate, fluorocarboxylate, and sulfonate fixed ion groups, J. Phys. Chem. B 105 (2001) 9387. [12] S. Chen, L. Krishnan, S. Srinivasan, J. Benziger, A.B. Bocarsly, Ion exchange resin/polystyrene sulfonate composite membranes for PEM fuel cells, J. Membr. Sci. 243 (2004) 327–333. [13] G. Akovali, A. Oözkan, Notes on modification of polystyrene by sulphonation: some properties of poly(styrenesulphonic acid), Polymer 27 (1986) 1277–1280. [14] N. Carretta, V. Tricoli, F. Picchioni, Ionomeric membranes based on partially sulfonated poly(styrene): synthesis, proton conduction and methanol permeation, J. Membr. Sci. 166 (2000) 189–197. [15] A.G.G. Garcia, P.N. Pintauro, M.W. Verbrugge, E.W. Schneider, Analysis of radiation-grafted membranes for fuel cell electrolytes, J. Appl. Electrochem. 22 (1992) 204–214. [16] L. Hong, N. Chen, Proton-conducting polymer membrane based on sulfonated polystyrene microspheres and an amphiphilic polymer blend, J. Polym. Sci. B 38 (2000) 1530–1538. [17] D.J. Jones, J. Rozière, Recent advances in the functionalisation of polybenzimidazole and polyetherketone for fuel cell applications, J. Membr. Sci. 185 (2001) 41–58. [18] B. Gupta, G.G. Scherer, Proton exchange membranes by radiation grafting of styrene onto FEP films. I. Thermal characteristics of copolymer membranes, J. Appl. Polym. Sci. 50 (1993) 2129–2134. [19] B. Gupta, F.N. Buechi, G.G. Scherer, A. Chapiro, Development of radiationgrafted FEP-g-polystyrene membranes: some property–structure correlations, J. Polym. Adv. Technol. 5 (1994) 493–498. [20] H. Kuhn, L. Gubler, T.J. Schmidt, G.G. Scherer, H.P. Brack, K. Simbek, T. Rager, F. Geiger, Development of polymer electrolyte membranes, in: Proceedings of the 2nd European PEFC Forum, June 30–July 4, Lucerne, Switzerland, 2003, p. 69. [21] S. Hietala, M. Koel, E. Skou, M. Elomaa, F. Sundholm, Thermal stability of styrene grafted and sulfonated proton conducting membranes based on poly(vinylidene fluoride), J. Mater. Chem. 8 (1998) 1127–1132. [22] S. Holmberg, J.H. Nasman, F. Sundholm, Synthesis and properties of sulfonated and crosslinked poly[(vinylidene fluoride)-graft-styrene] membranes, Polym. Adv. Technol. 9 (1998) 121–127. [23] S. Hietala, S.L. Maunu, F. Sundholm, T. Lehtinen, Water sorption and diffusion coefficients of protons and water in PVDF-g-PSSA polymer electrolyte membranes, J. Polym. Sci. B 37 (1999) 2893–2900. [24] T. Lehtinen, G. Sundholm, F. Sundholm, P. Bjornbom, M. Bursell, Electrochemical characterization of PVDF-based proton conducting membranes for fuel cells, Electrochim. Acta 43 (1998) 1881–1890. [25] D.I. Ostrovskii, L.M. Torell, M. Paronen, S. Hietala, F. Sundholm, Water sorption properties of and the state of water in PVDF-based proton conducting membranes, Solid State Ionics 97 (1997) 315–321. [26] S. Hietala, S.L. Maunu, F. Sundholm, Sorption and diffusion of methanol and water in PVDF-g-PSSA and Nafion® 117 polymer electrolyte membranes, J. Polym. Sci. B 38 (2000) 3277–3284. [27] M. Torkkeli, R. Serimaa, V. Etelaniemi, M. Toivola, K. Jokela, M. Paronen, F. Sundholm, ASAXS study of styrene-grafted sulfonated poly(vinylidene fluoride) membranes, J. Polym. Sci. B 38 (2000) 1734. [28] C. Bailly, D.J. Williams, F.E. Karasz, W. MacKnight, The sodium salts of sulphonated poly(aryl-ether-ether-ketone) (PEEK): preparation and characterization, Polymers 28 (1987) 1009–1016. [29] J.A. Kerres, Development of ionomer membranes for fuel cells, J. Membr. Sci. 185 (2001) 3–27. [30] R. Nolte, K. Ledjeff, M. Bauer, R. Mulhaupt, Partially sulfonated poly(arylene ether sulfone)—a versatile proton conducting membrane material for modern energy conversion technologies, J. Membr. Sci. 83 (1993) 211–220. [31] J. Kerres, W. Zhang, W. Cui, New sulfonated engineering polymers via the metalation route. II. Sulfinated/sulfonated poly(ether sulfone) PSU Udel and its crosslinking, J. Polym. Sci. A 36 (1998) 1441–1448. [32] L. Kerres, W. Cui, M. Junginger, Development and characterization of crosslinked ionomer membranes based upon sulfinated and sulfonated PSU crosslinked PSU blend membranes by alkylation of sulfinate groups with dihalogenoalkanes, J. Membr. Sci. 139 (1998) 227–241. [33] H. Nakajima, S. Nomura, T. Sugimoto, S. Nishikawa, I. Honma, High temperature proton conducting organic/inorganic nanohybrids for polymer electrolyte membrane, J. Electrochem. Soc. 149 (2002) A953–A959. [34] P. Staiti, M. Minutoli, S. Hocevar, Membranes based on phosphotungstic acid and polybenzimidazole for fuel cell application, J. Power Sources 90 (2000) 231–235. [35] J.M. Amarilla, R.M. Rojas, J.M. Rojo, M.J. Cubillo, A. Linares, J.L. Acosta, Antimonic acid and sulfonated polystyrene proton-conducting polymeric composites, Solid State Ionics 127 (2000) 133–139. [36] C. Poinsignon, I. Amodio, D. Foscallo, J.Y. Sanchez, Low cost filled thermostable ionomer membrane for PEMFC, Mater. Res. Soc. Symp. Proc. 548 (1999) 307–312. [37] S.P. Nunes, B. Ruffmann, E. Rikowski, S. Vetter, K. Richau, Inorganic modification of proton conductive polymer membranes for direct methanol fuel cells, J. Membr. Sci. 203 (2002) 215–225. [38] B. Bonnet, D.J. Jones, J. Roziere, L. Tchicaya, G. Alberti, M. Casciola, L. Massinelli, B. Baner, A. Peraio, E. Ramunni, Hybrid organic–inorganic membranes for a medium temperature fuel cell, J. New Mater. Electrochem. Syst. 3 (2000) 87–92.

426

S.K. Das et al. / Journal of Membrane Science 350 (2010) 417–426

[39] G. Alberti, M. Casciola, R. Palombari, Inorgano-organic proton conducting membranes for fuel cells and sensors at medium temperatures, J. Membr. Sci. 172 (2000) 233. [40] M. Rikukawa, K. Sanui, Proton-conducting polymer electrolyte membranes based on hydro-carbon polymers, Prog. Polym. Sci. 25 (2000) 1463–1502. [41] J.S. Wainright, M.H. Litt, R.F. Savinell, in: W. Vielstichm, A. Lamm, H.A. Gasteiger (Eds.), Handbook of Fuel Cells, vol. 3, John Wiley & Sons Ltd., New York, 2003, p. 436. [42] R.J. Gillespie, E.A. Robinson, in: T.C. Waddington (Ed.), Nonaqueous Solvent Systems, Academic Press, New York, 1965, pp. 117–210. [43] B. Xin, O. Savadogo, Hydrogen/oxygen polymer electrolyte membrane fuel cells (PEMFCs) based on alkaline-doped polybenzimidazole (PBI), Electrochem. Commun. 2 (2000) 697–702. [44] K.T. Adjemian, S. Srinivasan, J. Benziger, A.B. Bocarsly, Investigation of PEMFC operation above 100 ◦ C employing perfluorosulfonic acid silicon oxide composite membranes, J. Power Sources 109 (2002) 356–364. [45] M. Litt, R. Ameri, Y. Wang, R. Savinell, J.S. Wainwright, High-temperature polybenzimidazole fuel cell membranes, Mater. Res. Soc. Symp. Proc. 548 (1999) 313–323.

[46] C. Hasiotis, Q. Li, V. Deimede, J.K. Kallitsis, C.G. Kontoyannis, N.J. Bjerrum, Development and characterization of acid-doped polybenzimidazole/sulfonated polysulfone blend polymer electrolytes for fuel cells, J. Electrochem. Soc. 148 (2001) A513. [47] L. Jorissen, V. Gogel, J. Kerres, J. Garche, New membranes for direct methanol fuel cells, J. Power Sources 105 (2002) 267–273. [48] P. Choi, N.H. Jalani, T.M. Thampan, R. Datta, Consideration of thermodynamic, transport, and mechanical properties in the design of polymer electrolyte membranes for higher temperature fuel cell operation, J. Polym. Sci. B: Polym. Phys. 44 (2006) 2183–2200. [49] S.K. Das, K.J. Berry, Two-cell theory to measure membrane resistance based on proton flow: theory development and experimental validation, J. Power Sources 173 (2007) 909–916. [50] L.G. Beholz, Process for Producing Painted Polymeric Articles, US Patent No. 6,100,343 (2000). [51] L.G. Beholz, Apparatus for Treating Polymeric Material to Improve Surface Adhesion, US Patent No. 7,022,291 (2006). [52] J. Larminie, A. Dicks, Fuel Cell Systems Explained, John Wiley & Sons Inc., New York, 2000.