- Email: [email protected]
sulfonated silica nanoparticles for proton exchange membranes
Journal of Membrane Science 332 (2009) 121–128
Contents lists available at ScienceDirect
Journal of Membrane Science journal homepage: www.elsevier.com/locate/memsci
Preparation and properties of nanocomposite membranes of polybenzimidazole/sulfonated silica nanoparticles for proton exchange membranes Suryani, Ying-Ling Liu ∗ Department of Chemical Engineering and R&D Center for Membrane Technology, #200, Chung-Pei Road, Chung Yuan Christian University, Chungli, Taoyuan 32023, Taiwan
a r t i c l e
i n f o
Article history: Received 19 October 2008 Received in revised form 10 January 2009 Accepted 26 January 2009 Available online 6 February 2009 Keywords: Polybenzimidazole Silica Nanocomposite membrane Fuel cell
a b s t r a c t Sulfonated silica nanoparticles (SA-SNP) are utilized as a functional additive to prepare polybenzimidazole (PBI)/SA-SNP nanocomposite membranes. The sulfonic acid groups of SA-SNP form ionic linkages with PBI chains, improve the compatibility between SA-SNP and PBI, and enhance the mechanical strength of the PBI/SA-SNP nanocomposite membranes. After acid doping with phosphoric acid, PBI/SA-SNP nanocomposite membranes exhibit depressions on methanol permeability and enhancements on proton conductivity comparing to the pristine PBI membrane. The proton conductivity of PBI/SA-SNP-10 (possessing 10 wt% SA-SNP) membrane is of about 3-folds of that observed with pristine PBI membrane. The selectivity (the ratio of proton conductivity to the methanol permeability) of PBI/SA-SNP-15 is of about 1.3-folds of that of the Nafion® 117 membrane. The above results indicate the PBI/SA-SNP nanocomposite membranes could be utilized as the proton exchange membranes for direct methanol fuel cells. © 2009 Elsevier B.V. All rights reserved.
1. Introduction Polybenzimidazole (PBI) is one of the most studied materials for applications in proton exchange membrane fuel cell (PEMFC). As PBIs are basic polymers, acid doping in PBIs results in an increase both in their proton conductivities and thermal stability [1,2]. The effects of acid-doping levels, temperatures, and relative humidity on the proton conductivities of PBI membranes were studied [3–10]. The proton conductivities of acid-doped PBI membranes were also dependent to the doped acids in the order of H2 SO4 > H3 PO4 > HClO4 > HNO3 > HCl [11]. However, the acids doped in PBI membranes might leak in fuel cell applications. One promising approach to overcome this drawback was to covalently bond the acid groups, especially sulfonic acid groups, to the PBI chains [12–19]. The sulfonated PBI membranes showed relatively high proton conductivities comparing to the corresponding pristine PBI membranes. Due to the complicated synthesis routes for preparation of sulfonated PBIs and their poor solubility, sulfuric acids were also introduced to PBI membranes through blending PBI with sulfonated polymers. Blending PBI with Nafion® enhanced the stability [20] and depressed the methanol crossover [21] of the resulting composite membranes. Similar effects were also reported
∗ Corresponding author. Tel.: +886 3 2654130; fax: +886 3 2654199. E-mail address: [email protected] (Y.-L. Liu). 0376-7388/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.memsci.2009.01.045
with the blended composite membranes composed PBI/sulfonated polyether [22] and PBI/polyphosphazene [23]. Another approach to improve the thermal and chemical stability and to depress the methanol crossover of PEMs was formation of PEM nanocomposite membranes [24–29]. Addition of inorganic proton conductors to PBI membranes was demonstrated to enhance the proton conductivities and mechanical strengths of the membranes [29]. The improvements on the mechanical strengths and methanol barrier property of PBI membranes were also reported with PBI/silica (from a sol–gel process) [30] and PBI/montmorillonite [31] nanocomposite membranes. Unfortunately, decreases in the proton conductivities accompanied with the PBI nanocomposite membranes since addition of inorganic nanomaterials diluted the concentration of proton-conducting groups. The decreases in the proton conductivities of nanocomposite PEMs could be compensated using sulfonated nanomaterials as reinforcements [26–28]. Some acidic inorganic fillers were also utilized for PBI membranes to enhance their proton conductivity [32,33]. In this work, we reported the attempts on preparation of high performance PBI nanocomposite membranes for fuel cell application using sulfonated silica nanoparticles (SA-SNP) as nano-reinforcements. PBI nanocomposites membranes possessing non-sulfonated silica nanoparticles were also prepared to study the effects of sulfuric acid groups on the properties of PBI nanocomposite membranes. The sulfuric acid groups of SA-SNP formed ionic linkages with PBI chains, improved the compatibility between SASNP and PBI, and enhanced the mechanical strength of the PBI/SA-
122
Suryani, Y.-L. Liu / Journal of Membrane Science 332 (2009) 121–128
SNP nanocomposite membranes. PBI/SA-SNP nanocomposite membranes exhibited depressions on methanol permeability and enhancements on proton conductivity comparing to the pristine PBI membrane, to promise their application potentials for fuels cells. 2. Experimental 2.1. Materials 3,3 -Diaminobenzidine (DAB, 99%, Aldrich), isophthalic acid (IPA, 99%, Alfa Aesar), and polyphosphoric acid (PPA, ca. 84%, Alfa Aesar) were used as received. Sodium bicarbonate (99.5%) and ortho-phosphoric acid (85%, HPLC grade) were purchased from Showa Chemical Co. (Japan) and Scharlau Chemie (Germany), respectively. Methanol and N,N-dimethylacetamide (DMAc) in HPLC grade were received from TEDIA. DMAc was dried over 4 Å molecular sieves (Acros) for at least 2 days before use. Silica nanoparticles (SNP) in diameters of 10–20 nm were received from Nissan Chemical Co., Japan. Sulfonated silica nanoparticles were prepared in the laboratory according to the reported method [34]. The content of sulfuric acid group of SA-SNP is 0.57 mmol/g measured with elemental analysis.
and the heated to 900 ◦ C. Tensile properties were measured with an instron (Instron model 5543) with an elongation rate of 0.5 mm/min at room temperature. Dynamic mechanical analysis (DMA) was carried out with a Thermal Analysis dynamic mechanical analyzer (TA Q-800 DMA). The analytical conditions were tension force of 0.05N, frequency of 1 Hz, amplitude of 20 m, temperature range of 25–500 ◦ C, and heating rate of 5 ◦ C/min. Scanning electron microscopy (SEM) was performed on a field-emission SEM of Hitachi S-4800. Elemental analysis and element mapping were conducted with a Hitachi S-3000N SEM equipped with an energy dispersive X-ray spectroscopy (EDX) of Horiba EDX-250. 2.5. Water uptake and acid-doping level Water uptake measurement was performed by dipping the dried membranes (dried at 120 ◦ C in vacuo for 5 h) in distilled water for 24 h at 25 or 80 ◦ C. The membranes were taken out, wiped with paper, and then weighted. The water uptake was calculated by Eq. (1): Water uptake (%) =
Ws − Wd × 100 Wd
(1)
2.2. Synthesis of PBI PBI was synthesized by polycondensation reaction according to the methods described elsewhere [35]. 15 g of DAB and 250 g of PPA were added into a 1-L round bottom flask equipped with magnetic stirrer. The solution was heated to 140 ◦ C under a nitrogen stream to dissolve DAB in PAA. 11.63 g of IPA was then added and the temperature was raised to 200 ◦ C. After 15 h, the viscous solution was poured into a large amount of cold water. The fibers formed were kept in water for at least 3 days. The excess acid was neutralized with a sodium bicarbonate aqueous solution (1 M). The polymer was then collected from filtration, washed thoroughly with water and methanol, and then dried under vacuum at 130 ◦ C overnight. Yield: 81%. 1 H NMR (300 MHz, ppm) 13.23 (2H), 9.13 (1H), 8.30 (2H), 7.97 (1H), 7.79 (3H), 7.61 (3H); Inherent viscosity (0.5 g/dL in DMAc at 25 ◦ C): 1.42 dL/g. 2.3. Preparation of nanocomposite membranes A PBI solution was prepared by dissolving 1 g PBI in 20 mL DMAc at 120 ◦ C. A certain amount of SNP or SA-SNP was added. The homogeneous solution was filtered to remove trace precipitate, and then poured into a Petri dish. Solvent was gradually evaporated at 80 ◦ C overnight. Membranes were detached from the Petri dish by water immersion, treated in boiled water for 3 h, and then dried at 190 ◦ C for 3 h to remove the residual DMAc. Nanocomposite membranes possessing 0, 5, 10, 15, 20 wt% of SNP (PBI/SNP-X, X denoting to the SNP content) or SA-SNP (PBI/SA-SNP-X, X denoting to the SA-SNP content) were prepared. The membrane thickness was in the range of 40–60 m. 2.4. Characterization The inherent viscosity of PBI polymer was measured with a Cannon-Fenske Routine Viscometer at 25 ◦ C using a PBI solution in DMAc (0.5 g/dL). 1 H NMR spectrum of PBI was recorded on a Brüker MSL 300 NMR (300 MHz) using deuterated dimethylsulfoxide (DMSO-d6 ) as a solvent. FTIR spectra of nanocomposite membranes were recorded on a Perkin-Elmer Spectrum One FTIR. Thermogravimetric analysis (TGA) was performed by a ThermalAnalysis (TA) TGA Q-500 at a heating rate of 10 ◦ C under nitrogen. All samples were preheated at 120 ◦ C for 20 min to remove absorbed water from membranes then equilibrated at ca. 35 ◦ C for 30 min,
where Ws and Wd are membrane weights after and before dipping, respectively. Acid-doping level was obtained by a similar manner. Membranes were dipped in an 85% H3 PO4 solution for 1 week at room temperature. To exclude the weight gain due to water uptake, doped membranes were dried at 110 ◦ C under vacuum until the membrane weights unchanged with time. The weight change in acid doping was measured and used for the calculation of aciddoping level (acid numbers per PBI repeating unit). 2.6. Methanol permeability Methanol permeability measurement was performed on a tworeservoir compartment in which the membrane was clamped between the two reservoirs. The membranes were hydrated in deionized water at room temperature for 24 h prior to measurement. Source cell (VA = 20 mL) was filled with a 3-M methanol aqueous solution, and receiving cell (VB = 18 mL) was filled with deionized water. The concentration of methanol in receiving cell was measured versus time by gas chromatography GC390B (CL Sciences Inc., Japan) with a gas flow rate of 30 mL/min. The methanol concentration of the source cell was considered much higher than that of the receiving cell during measurements. Methanol permeability was calculated by Eq. (2): CB (t) =
A DKCA (t − t0 ) VB L
(2)
where CB is the methanol concentration in receiving cell, CA the initial methanol concentration, A the cross-section area, L the membrane thickness, D the methanol diffusivity and K the partition coefficient. The product DK is the membrane permeability to methanol. Parameter t0 is the “lag time” corresponding to the time necessary for the methanol to pass through membrane. 2.7. Proton conductivity Proton conductivity of the membranes was measured with a Solartron 1255B frequency response analyzer equipped with a Solartron 1287 electrochemical interface with an oscillation amplitude of 10 mV and a frequency range of 0.1 Hz to 1 MHz. The measurements were taken at 20–80 ◦ C under a 95% relative humidity.
Suryani, Y.-L. Liu / Journal of Membrane Science 332 (2009) 121–128
123
Fig. 1. Chemical structure of sulfonated silica nanoparticles (SA-SNP) and PBI.
3. Results and discussion 3.1. Characterizations on PBI nanocomposite membranes PBI and SA-SNP (Fig. 1) were prepared in our laboratory. PBI and SA-SNP could form homogeneous solutions. The PBI/SA-SNP nanocomposites membranes were therefore obtained with a casting method. PBI/SNP membranes were also prepared with uses of non-sulfonated SNP to replace SA-SNP to examine the effect of sul-
fonic acid groups on the properties and performances of the PBI nanocomposites membranes. Fig. 2 shows the SEM micrographs of the PBI/silica nanocomposite membranes. Pristine PBI membranes exhibited dense and homogeneous morphology for both surface (Fig. 2a) and cross-section (Fig. 2b) observations. The silica particles in PBI/SNP-10 membrane could be seen in its surface SEM micrograph (Fig. 2c). Particles in sizes of about 20 nm aggregate together to form silica domains in micrometers, indicating the interactions between silica nanoparticles are stronger than that between
Fig. 2. SEM micrographs of PBI ((a) surface and (b) cross-section) and its nanocomposites of PBI/SNP-10 ((c) surface and (d) cross-section) and PBI/SA-SNP-10 ((e) surface and (f) cross-section).
124
Suryani, Y.-L. Liu / Journal of Membrane Science 332 (2009) 121–128
Fig. 3. Elemental mapping of sulfur (a) and silicon (b) on PBI/SA-SNP-20 membrane observed with a SEM-EDX.
PBI molecules and silica nanoparticles. The agglomerated silica domains were also observed in the cross-sectional SEM micrograph of PBI/SNP-10 (Fig. 2d). The SEM observations indicate that the compatibility between PBI and silica nanoparticles and the homogeneity of the PBI/SNP membranes are somewhat poor. However, the dispersion homogeneity of silica nanoparticles in PBI membranes could be significantly improved with using SA-SNP as the additives. As shown in Fig. 2e, isolated silica particles were observed in the surface SEM micrograph of PBI/SA-SNP nanocomposite membranes, demonstrating the high compatibility between PBI polymer and SA-SNP particles and the high dispersion ability of SA-SNPs in PBI matrix. Particle agglomeration was still not observed in the cross-sectional SEM micrograph of PBI/SA-SNP-10 (Fig. 2f). PBI has a pKa value of about 5.5. The basicity of PBI makes it readily reactive toward the sulfonic acid groups of SA-SNPs through ionic interaction. The ionic linkages between PBI and SA-SNP contribute to their high compatibility. Fig. 3 shows the SEM-EDX elemental mapping micrographs of PBI/SA-SNP-20. Silicon and sulfur elements were observed to demonstrate the presence of SA-SNP particles in PBI nanocomposite membranes. In addition, the distributions of silicon and sulfur in the micrographs are homogeneous, indicating that there was not particle agglomeration and micro-phase separation in the nanocomposites membranes. Therefore, SA-SNP particles could be considered homogeneously dispersing in the PBI polymer matrix.
Fig. 4. FTIR spectra of PBI membrane and its nanocomposite membranes of PBI/SNP10 and PBI/SA-SNP-10.
Fig. 4 shows the FTIR spectra of PBI and its nanocomposite membranes. Characteristic absorption of PBI appeared in the spectra for all membranes [35]. Absorption peaks at 3390 and 3150 cm−1 correspond to the non-hydrogen bonded N–H stretching and hydrogen bonded (self-associated) N–H groups, respectively. The absorption peaks at 1438 and 1532 cm−1 arise from the in-plane deformation of benzimidazole, and the peak at 1282 cm−1 is from the breathing mode of imidazole rings. The absorption peak at about 1100 cm−1 in relatively low intensity might correspond to the in-plane C–H deformation vibrations [2]. After addition of silica nanoparticles, the absorption intensities at about 1100 cm−1 increased due to the contribution of Si–O–Si absorption. The result indicates the presence of silica in both PBI/SNP-10 and PBI/SA-SNP-10 nanocomposite membranes. In addition, PBI/SA-SNP-10 exhibited some minor absorption peaks at 1370 and 1168 cm−1 corresponding to the absorptions from the sulfuric acid groups. A decrease in the relative intensity of absorption peak at 3390 cm−1 was observed with PBI/SA-SNP-10, indicating some of the non-hydrogen bonded N–H groups in PBI chains bond to the sulfonic acid groups of SA-SNP via H-bonding and ionic interaction. 3.2. Properties of PBI nanocomposite membranes The glass transition temperatures (Tg ) of PBI/SA-SNP nanocomposite membranes were read from the tan ı peaks in DMA thermograms (figures not shown). Pristine PBI showed a Tg of 395 ◦ C, which shifted to higher temperatures (450–455 ◦ C) while formation of nanocomposites with SA-SNP. The Tg increase indicates that the motion ability of PBI chains is restricted with the presence of SA-SNP. The chain motion restriction should be majorly contributed from the strong ionic linkages between PBI and SA-SNP. The ionic linkages bring a cross-linking effect to the PBI/SA-SNP nanocomposite membranes, so as to increase the Tg s of the PBI/SA-SNP nanocomposites. The contributions of –SO3 H/–NH ionic linkages to Tg enhancement was also demonstrated with the relatively low Tg of PBI/SNP-20 (423 ◦ C), as SNP does not possess sulfonic acid groups. On the other hand, the presence of SNP or SA-SNP did not significantly alter the thermal degradation behaviors of PBI polymer, as shown with their TGA thermograms (Fig. 5). The –SO3 H/–NH ionic linkages in the SA-SNP/PBI nanocomposites membranes broke down before the degradation of PBI chains. Due to the presence of the silica nanoparticles did not involve in the degradation reactions of polymer portion [36], all PBI/SA-SNP nanocomposites as well as PBI polymer exhibited similar degradation temperatures, which was due to the degradation of PBI chains. However, the presence of inorganic portions in the PBI nanocomposite membranes might reduce the weight loss rate of the PBI nanocomposites in high temperature region and increase the char yields of the nanocomposites. The relatively high char yield of PBI/SA-SNP-20, comparing to that of PBI/SNP-20, is attributed
Suryani, Y.-L. Liu / Journal of Membrane Science 332 (2009) 121–128
Fig. 5. TGA thermograms of PBI-based membranes. PBI/SNP-10 and PBI/SNP-20: nanocomposites membranes of PBI and silica nanoparticles (SNP) possessing 10 and 20 wt% of SNP, respectively; PBI/SA-SNP-10 and PBI/SA-SNP-20: nanocomposites membranes of PBI and silica nanoparticles (SNP) possessing 10 and 20 wt% of SA-SNP, respectively.
to the promoting effect of sulfonic acid groups on the dehydration/carbonization during PBI thermal degradation. The mechanical properties of the PBI-based membranes were measured with an Instron. The stress–strain curves are shown in Fig. 6. Pristine PBI membrane showed a Young’s modulus of 2.9 GPa and an elongation at break of 40%. Formation of nanocomposites with silica nanoparticles results in increases in tensile strengths and decreases in elongations at break for the PBI-based membranes. For PBI/SA-SNP nanocomposites membranes, the presence of ionic linkages between PBI and silica particles attributes to their high Young’s modulus, especially for high SA-SNP loaded samples of PBI/SA-SNP-15 and PBI/SA-SNP-20 (Young’s modulus above 3.9 MPa). Introduction of silica nanoparticles increased the brittleness of the PBI-based membranes with the decreases in their elongations at break. The brittleness of the membranes increased with increasing their silica contents. The effects of SNP and SA-SNP on the elongations at break of the nanocomposite membranes were
Fig. 6. Stress–strain curves of the nanocomposite membranes of (i) PBI and silica nanoparticles (SNP): PBI/SNP-X, X = 5, 10, 15, 20 which denoting to the SNP weight percents in the nanocomposites and (ii) PBI and sulfonated silica nanoparticles (SASNP): PBI/SA-SNP-X, X = 5, 10, 15, 20 which denoting to the SA-SNP weight percents in the nanocomposites.
125
Fig. 7. Water uptakes of the PBI nanocomposite membranes of (i) PBI and silica nanoparticles (SNP) nanocomposites (PBI/SNP) and (ii) PBI and sulfonated silica nanoparticles (SA-SNP) nanocomposites (PBI/SA-SNP) at 25 and 80 ◦ C.
similar. The decreases in the elongation at break and increases in the brittleness were widely observed with polymer nanocomposites and could be attributed to the presence of inorganic reinforcement. Moreover, the ionic linkages in PBI/SA-SNP as well as the hydrogen bonding in PBI/SNP also bring some cross-linking effect to the PBI chains so as to in decrease the elongation at break of the polymer. Therefore, the PBI/SA-SNP membranes could be superior to the PBI/SNP membranes in mechanical characteristics. Moreover, the high brittleness of PBI/SNP-20 and PBI/SA-SNP-20 (possessing 9% of elongation at break) might limit their applications for fuel cell membranes. Comparatively, PBI/SA-SNP-15, which showed an elongation at break of 16%, is more suitable for the membrane applications. PBI exhibit a water uptake of 16.6% at 25 ◦ C, corresponding to 2.74 molecules of water associating with one PBI repeating unit. The measured value of water uptake is reasonable comparing to the proposed values by Li et al. [5], who suggested that N atom and N–H groups in PBI repeating units formed intermolecular hydrogen bonds with 2–4 water molecules. Addition of silica nanoparticles to PBI membranes increased their water uptakes (Fig. 7) due to the hygroscopic nature of silica nanoparticles. The sulfonic acid groups in SA-SNP nanoparticles further increased the water uptakes to greater extents. PBI/SNP-20 and PBI/SA-SNP-20 showed a water uptake of 21.6 and 23.8 wt% at 25 ◦ C, respectively. Besides, the water uptakes of the PBI-based membranes measured at 25 ◦ C were higher than the values measured at 80 ◦ C. Sannigrahi et al. [37] reported that PBI polymer chain shrank in DMAc at elevated temperature, which was caused by disruption of interchain hydrogen bonding, and underwent conformational change from extended helical conformer to collapsed compact coil conformer. The low water uptakes observed with PBI membranes at high temperatures might be attributed to the PBI chain shrinkages in water. Similar results were also reported with Qing et al. [13,14] for sulfonated PBI polymers. The PBI-based membranes were doped with 85% H3 PO4 to prepare the acid-doped membranes for fuel cell applications. Fig. 8 shows the acid-doping levels of the PBI-based nanocomposite membranes. Acid-doping level (H3 PO4 molecules per PBI repeating unit) of pristine PBI membrane was 12.3, which decreased to 9.8 and 8.6 for PBI/SNP-20 and PBI/SA-SNP-20, respectively (Fig. 8). Some of the amino groups in PBI chains were blocked with the sulfonic acid groups of SA-SNP, to decrease the binding ability of PBI
126
Suryani, Y.-L. Liu / Journal of Membrane Science 332 (2009) 121–128
Fig. 8. Acid-doping levels of PBI nanocomposite membranes with H3 PO4 : (i) PBI and silica nanoparticles (SNP) nanocomposites (PBI/SNP) and (ii) PBI and sulfonated silica nanoparticles (SA-SNP) nanocomposites (PBI/SA-SNP).
chains to H3 PO4 molecules. Although high acid-doping levels are expected for fuel cell applications, these prepared PBI nanocomposites membranes still showed high proton conductivities, to be discussed below. 3.3. Methanol permeability and proton conductivity of PBI nanocomposite membranes The methanol permeability of PBI-based membranes was measured with a 3-M methanol aqueous solution at 25 ◦ C by a side-by-side method. Pure PBI membrane (without acid doping) exhibited a methanol permeability of 1.37 × 10−9 cm2 /s, which increased to 5.5 × 10−7 cm2 /s after being doped with H3 PO4 . The protonation of PBI chains with acid doping reduced the intermolecular interaction forces of PBI and increased the free volume of PBI, consequently to generate some additional spaces for methanol crossing the PBI membrane. The presence of silica nanoparticles in the PBI nanocomposite membranes played some counteraction to this effect. As shown in Fig. 9, the PBI nanocomposites exhibited relatively low methanol permeability. The increases in methanol
Fig. 9. Methanol permeability of PBI nanocomposite membranes: (i) PBI and silica nanoparticles (SNP) nanocomposites (PBI/SNP) and (ii) PBI and sulfonated silica nanoparticles (SA-SNP) nanocomposites (PBI/SA-SNP).
Fig. 10. Proton conductivities of PBI nanocomposite membranes at 95% relative humidity: (i) PBI and silica nanoparticles (SNP) nanocomposites (PBI/SNP) and (ii) PBI and sulfonated silica nanoparticles (SA-SNP) nanocomposites (PBI/SA-SNP). The proton conductivity of the Nafion® 117 membrane is also shown for comparison.
permeability of PBI/SNP-15 and PBI/SNP-20 membranes could be due to the poor compatibility between PBI and SNP particles and the agglomeration of SNPs in the PBI/SNP nanocomposites membranes. Since SA-SNP particles are much compatible to PBI, the continuing decreases in the methanol permeability were reasonably observed with PBI/SA-SNP-15 and PBI/SA-SNP-20. It is noteworthy that the methanol permeability of PBI/SA-SNP-20 membrane (3.3 × 10−7 cm2 /s) is of about one-tenth of the value read with Nafion® (3.7 × 10−6 cm2 /s) [24]. The proton conductivities of the phosphoric acid-doped PBI membranes at 95% relative humidity as a function of temperatures are shown in Fig. 10. The proton conductivities of all membranes increased with increasing the operation temperatures. Addition of the un-modified SNP showed a negative effect on the proton conductivities of the PBI/SNP membranes. Similar results were also reported to other nanocomposites membranes like Nafion/silica membranes [23,38] and PBI/silica membranes [30]. On the other hand, comparing to the pristine PBI and Nafion® 117 membranes, the relatively high proton conductivities of the PBI/SA-SNP nanocomposites membranes are noteworthy. Li et al. [24] used sulfonated organosilica to modify Nafion® membranes and they found the proton conductivities of the modified membranes were still lower than that of the neat Nafion® membrane. The reduction in the proton conductivity was attributed to the silica-induced changes in the tortuous paths and in the distribution of hydrophilic/hydrophobic domains. However, in this work an enhancing effect on the proton conductivity was observed with addition of sulfonated silica nanoparticles to PBI membranes. High proton conductivities were observed with the PBI/SA-SNP nanocomposites membranes, even the membranes possessed relatively low acid-doping levels comparing to the pristine PBI membrane. High doping levels have been recognized to result in high proton conductivities for H3 PO4 -doped PBI membranes, due to the excess H3 PO4 could form anion pathways for proton conduction [4]. However, the H3 PO4 doping levels of the PBI/SA-SNP membranes were still high enough for proton conduction [29]. The sulfonic acid groups of SA-SNP also contribute to the acid-doping effect to PBI polymer so as to enhance the proton conductivity of PBI/SA-SNP nanocomposites membranes [12–19]. Moreover, the presence of SA-SNP particles induced proton conductive path-
Suryani, Y.-L. Liu / Journal of Membrane Science 332 (2009) 121–128
127
The high proton conductivities and low methanol permeability of the PBI/SA-SNP nanocomposite membranes indicate their potentials of applications as the proton exchange membranes in direct methanol fuel cells (DMFC), which are evaluated with the membrane selectivity, i.e. the ratio of the proton conductivity to the methanol permeability [39]. The relative selectivity of the PBI-based membrane is higher than or comparable to that of the Nafion® 117 membrane as shown in Fig. 12. PBI/SA-SNP15 membrane shows the highest selectivity among the evaluated membranes, which is of about 1.3-folds of the selectivity of the Nafion® 117 membrane. Moreover, the PBI/SA-SNP membranes could be highly superior over the Nafion® 117 membrane at high temperatures and dry conditions, since under this operation condition the Nafion® 117 membrane shows a very low selectivity. 4. Conclusions
Fig. 11. Proton conductivities of PBI/SA-SNP nanocomposite membranes at high temperature region and under dry environment. The proton conductivity of the Nafion® 117 membrane is also shown for comparison.
ways attributing to the increases in the proton conductivities of PBI/SA-SNP membranes. The phenomena were similar to the results reported in our previous work [28]. The proton conductivities of the acid-doped PBI/SA-SNP nanocomposites membranes were further studied at high temperatures under an anhydrous environment (Fig. 11). All the PBI/SA-SNP nanocomposite membranes still exhibited higher proton conductivities that did the pristine acid-doped PBI membrane. Since the Grotthuss-type diffusion mechanism without assistance of the vehicle mechanism has been proposed for the proton conduction under anhydrous environments [3,10], the proton conductivity of the acid-doped PBI/SA-SNP membranes could be synergistically resulted from their SA-SNP contents and the acid-doping levels. Consequently, the highest proton conductivity was obtained with the acid-doped PBI/SA-SNP-10 membrane, which possess the moderate SA-SNP content and acid-doping level. It is noteworthy that Nafion® 117 shows relatively low proton conductivities under the anhydrous condition and almost loses its proton conductivities at high temperatures. Comparing to Nafion® 117 membrane, the high proton conductivities of PBI/SA-SNP-10 indicate that this membrane could be utilized as proton exchange membranes for fuel cells at high temperatures and dry conditions.
Fig. 12. The relative selectivity (over the Nafion® 117 membrane) of membranes of PBI and sulfonated silica nanoparticle (SA-SNP) nanocomposites (PBI/SA-SNP).
Sulfonated silica nanoparticles were effective additives for preparation of homogeneous PBI/SA-SNP nanocomposite membranes. The sulfonic acid groups of SA-SNP enhanced the compatibility between silica nanoparticles and PBI polymer matrix and improved the mechanical properties of the resulting PBI/SASNP nanocomposite membranes. Incorporation of SA-SNP to PBI membranes also depressed their methanol permeability and increased their proton conductivities, to result in their high selectivity comparing to pristine PBI and Nafion® membranes. The performances of the PBI/SA-SNP nanocomposite membranes are attractive and the membrane could be used as proton exchange membranes for direct methanol fuel cells and high temperature/low humidity proton exchange membrane fuel cells. Acknowledgements We thank The Ministry of Education, Taiwan for their financial support on this work under the Center-of-Excellence Program on Membrane Technology (2006–2010). Special thanks are given to Professor Steve Lien-Chung Hsu (National Cheng Kung University, Taiwan) for his kind help and discussion on the proton conductivity measurements. References [1] J.S. Wainright, J.-T. Wang, D. Weng, R.F. Savinell, M.H. Litt, Acid doped polybenzimidazoles: a new polymer electrolyte, J. Electrochem. Soc. 142 (1995) L121. [2] X. Glipa, B. Bonnet, B. Mula, D.J. Jones, J. Rozière, Investigation of the conduction properties of phosphoric and sulfuric acid doped polybenzimidazole, J. Mater. Chem. 9 (1999) 3045–3049. [3] Y.L. Ma, J.S. Wainright, M.H. Litt, R.F. Savinell, Conductivity of PBI membranes for high-temperature polymer electrolyte fuel cells, J. Electrochem. Soc. 151 (2004) A8–A16. [4] Q. Li, R. He, J.O. Jensen, N.J. Bjerrum, PBI-based polymer membranes for high temperature fuel cells-preparation, characterization, and fuel cell demonstration, Fuel Cells 4 (2004) 147–159. [5] Q. Li, R. He, R.W. Berg, H.A. Hjulers, N.J. Bjerrum, Water uptake and doping of polybenzimidazoles as electrolyte membranes for fuel cells, Solid State Ionics 168 (2004) 177–185. [6] J.A. Asensio, S. Borrós, P. Gómez-Romero, Polymer electrolyte fuel cells based on phosphoric-acid impregnated poly(2,5-benzimidazole) membranes, J. Electrochem. Soc. 151 (2004) A304–A310. [7] H. Pu, Q. Liu, L. Qiao, Z. Yang, Studies on proton conductivity of acid doped polybenzimidazole/polyimide and polybenzimidazole/polyvinylpyrrolidone blends, Polym. Eng. Sci. 45 (2005) 1395–1400. [8] R. He, Q. Li, A. Bach, J.O. Jensen, N.J. Bjerrum, Physicochemical properties of phosphoric acid doped polybenzimidazole membranes for fuel cells, J. Membr. Sci. 277 (2006) 38–45. ˜ [9] J. Lobato, P. Canizares, M.A. Rodrigo, J.J. Linares, J.A. Aguilar, Improved polybenzimidazole films for H3 PO4 -doped PBI-based high temperature PEMFC, J. Membr. Sci. 306 (2007) 47–55. [10] R. He, Q. Li, J.O. Jensen, N.J. Bjerrum, Doping phosphoric acid in polybenzimidazole membranes for high temperature proton exchange membrane fuel cells, J. Polym. Sci. Part A: Polym. Chem. 45 (2007) 2989–2997.
128
Suryani, Y.-L. Liu / Journal of Membrane Science 332 (2009) 121–128
[11] B.Z. Xing, O. Savadogo, The effect of acid doping on the conductivity of polybenzimidazole, J. New Mater. Electrochem. Syst. 2 (1999) 95–101. [12] J.A. Asensio, S. Borros, P. Gomez-Romero, Proton-conducting polymers based on benzimidazole and sulfonated benzimidazole, J. Polym. Sci. Part A: Polym. Chem. 40 (2002) 3703–3710. [13] S. Qing, W. Huang, D. Yan, Synthesis and characterization of thermally stable sulfonated polybenzimidazoles obtained from 3,3 -disulfonyl-4,4 dicarboxyldiphenylsulfone, J. Polym. Sci. Part A: Polym. Chem. 43 (2005) 4363–4372. [14] S. Qing, W. Huang, D. Yan, Synthesis and properties of soluble sulfonated polybenzimidazoles, React. Funct. Polym. 66 (2006) 219–227. [15] J. Jouanneau, R. Mercier, L. Gonon, G. Gebel, Synthesis of sulfonated polybenzimidazoles from functionalized monomers: preparation of ionic conducting membranes, Macromolecules 40 (2007) 983–990. [16] Z. Bai, G.E. Price, M. Yoonessi, S.B. Juhl, M.F. Durstock, T.D. Dang, Proton exchange membranes based on sulfonated polyarylenethioethersulfone and sulfonated polybenzimidazole for fuel cell applications, J. Membr. Sci. 305 (2007) 69–76. [17] N. Li, Z. Cui, S. Zhang, W. Xing, Sulfonated polyimide bearing benzimidazole groups for proton exchange membranes, Polymer (2007) 7255–7263. [18] H. Xu, K. Chen, X. Guo, J. Fang, J. Yin, Synthesis of novel sulfonated polybenzimidazole and preparation of cross-linked membranes for fuel cell application, Polymer 48 (2007) 5556–5564. [19] J. Peron, E. Ruiz, D.J. Jones, J. Roziere, Solution sulfonation of novel polybenzimidazole a proton electrolyte for fuel cell application, J. Membr. Sci. 314 (2008) 247–256. [20] Y. Zhai, H. Zhang, Y. Zhang, D. Xing, A novel H3 PO4 /Nafion-PBI composite membrane for enhanced durability of high temperature PEM cells, J. Power Sources 169 (2007) 259–264. [21] J.K. Lee, J. Kerres, Synthesis and characterization of sulfonated poly(arylene thioether)s and their blends with polybenzimidazole for proton exchange membranes, J. Membr. Sci. 294 (2007) 75–83. [22] R. Wycisk, J.K. Lee, P.N. Pintauro, Sulfonated polyphosphazene-polybenzimidazole membranes for DMFCs, J. Electrochem. Soc. 152 (2005) A892–A898. [23] N. Miyake, J.S. Wainright, R.F. Savinell, Evaluation of a sol–gel derived Nafion/silica hybrid membrane for proton electrolyte membrane fuel cell applications. I. Proton conductivity and water content, J. Electrochem. Soc. 148 (2001) A898–A904. [24] C. Li, G. Sun, S. Ren, J. Liu, Q. Wang, Z. Wu, H. Sun, W. Jin, Casting Nafionsulfonated organosilica nano-composite membranes used in direct methanol fuel cells, J. Membr. Sci. 272 (2006) 50–57. [25] Y.H. Su, Y.L. Liu, Y.M. Sun, J.Y. Lai, M.D. Guiver, Y. Gao, Using silica nanoparticles modifying sulfonated poly(phthalazinone ether ketone) membrane for direct
[26]
[27] [28]
[29]
[30]
[31]
[32] [33]
[34]
[35]
[36]
[37]
[38]
[39]
methanol fuel cell: a significant improvement on cell performance, J. Power Sources 155 (2006) 111–117. C.H. Rhee, H.K. Kim, H. Chang, J.S. Lee, Nafion® /sulfonated montmorillonite composite: a new concept electrolyte membrane for direct methanol fuel cells, Chem. Mater. 17 (2005) 1691–1697. P. Bébin, M. Caravanier, H. Galiano, Nafion® /clay–SO3 H membrane for proton exchange membrane fuel cell application, J. Membr. Sci. 278 (2006) 35–42. Y.H. Su, Y.L. Liu, Y.M. Sun, J.Y. Lai, D.M. Wang, Y. Gao, B. Liu, M.D. Guiver, Proton exchange membranes modified with sulfonated silica nanoparticles for direct methanol fuel cells, J. Membr. Sci. 296 (2007) 21–28. R. He, Q. Li, G. Xiao, N.J. Bjerrum, Proton conductivity of phosphoric acid doped polybenzimidazole and its composites with inorganic proton conductors, J. Membr. Sci. 226 (2003) 169–184. S.W. Chuang, S.L.C. Hsu, Y.H. Liu, Synthesis and properties of fluorine-containing polybenzimidazole/silica nanocomposite membranes for proton exchange membrane fuel cells, J. Membr. Sci. 305 (2007) 353–363. S.W. Chuang, S.L.C. Hsu, Y.H. Liu, Synthesis and properties of fluorine-containing polybenzimidazole/montmorillonite nanocomposite membranes for direct methanol fuel cell applications, J. Power Sources 168 (2007) 172–177. P. Staiti, Proton conductive membranes absorbed on silicotungstic acid/silica and polybenzimidazole, Mater. Lett. 47 (2001) 241–246. P. Staiti, M. Minutoli, S. Hocevar, Membranes based on phosphortungstic acid and polybenzimidazole for fuel cell application, J. Power Sources 90 (2000) 231–235. Y.L. Liu, C.Y. Hsu, Y.H. Su, J.Y. Lai, Chitosan–silica complex membranes from sulfuric acid functionalized silica nanoparticles for pervaporation dehydration of ethanol–water solutions, Biomacromolecules 6 (2005) 368–373. ˜ J. Lobato, P. Canizares, M.A. Rodrigo, J.J. Linares, G. Manjavacas, Synthesis and characterization of poly[2,2-(m-phenylene)-5,5-bibenzimidazole] as polymer electrolyte membrane for high temperature PEMFCs, J. Membr. Sci. 280 (2006) 351–362. Y.L. Liu, C.Y. Hsu, W.L. Wei, R.J. Jeng, Preparation and thermal properties of epoxy-silica nanocomposites from nanoscale colloidal silica, Polymer 44 (2003) 5159–5167. A. Sannigrahi, D. Arunbabu, R.M. Sankar, T. Jana, Aggregation behavior of polybenzimidazole in aprotic polar solvent, Macromolecules 40 (2007) 2844–2851. R. Jiang, H.R. Kunz, J.M. Fenton, Composite silica/Nafion® membranes prepared by tetraethylorthosilicate sol–gel reaction and solution casting for direct methanol fuel cells, J. Membr. Sci. 272 (2006) 116–124. D.K. Kim, G.P. Robertson, M.D. Guiver, Comb-shaped poly(arylene ether sulfone)s as proton exchange membranes, Macromolecules 41 (2008) 2126–2134.
Contents lists available at ScienceDirect
Journal of Membrane Science journal homepage: www.elsevier.com/locate/memsci
Preparation and properties of nanocomposite membranes of polybenzimidazole/sulfonated silica nanoparticles for proton exchange membranes Suryani, Ying-Ling Liu ∗ Department of Chemical Engineering and R&D Center for Membrane Technology, #200, Chung-Pei Road, Chung Yuan Christian University, Chungli, Taoyuan 32023, Taiwan
a r t i c l e
i n f o
Article history: Received 19 October 2008 Received in revised form 10 January 2009 Accepted 26 January 2009 Available online 6 February 2009 Keywords: Polybenzimidazole Silica Nanocomposite membrane Fuel cell
a b s t r a c t Sulfonated silica nanoparticles (SA-SNP) are utilized as a functional additive to prepare polybenzimidazole (PBI)/SA-SNP nanocomposite membranes. The sulfonic acid groups of SA-SNP form ionic linkages with PBI chains, improve the compatibility between SA-SNP and PBI, and enhance the mechanical strength of the PBI/SA-SNP nanocomposite membranes. After acid doping with phosphoric acid, PBI/SA-SNP nanocomposite membranes exhibit depressions on methanol permeability and enhancements on proton conductivity comparing to the pristine PBI membrane. The proton conductivity of PBI/SA-SNP-10 (possessing 10 wt% SA-SNP) membrane is of about 3-folds of that observed with pristine PBI membrane. The selectivity (the ratio of proton conductivity to the methanol permeability) of PBI/SA-SNP-15 is of about 1.3-folds of that of the Nafion® 117 membrane. The above results indicate the PBI/SA-SNP nanocomposite membranes could be utilized as the proton exchange membranes for direct methanol fuel cells. © 2009 Elsevier B.V. All rights reserved.
1. Introduction Polybenzimidazole (PBI) is one of the most studied materials for applications in proton exchange membrane fuel cell (PEMFC). As PBIs are basic polymers, acid doping in PBIs results in an increase both in their proton conductivities and thermal stability [1,2]. The effects of acid-doping levels, temperatures, and relative humidity on the proton conductivities of PBI membranes were studied [3–10]. The proton conductivities of acid-doped PBI membranes were also dependent to the doped acids in the order of H2 SO4 > H3 PO4 > HClO4 > HNO3 > HCl [11]. However, the acids doped in PBI membranes might leak in fuel cell applications. One promising approach to overcome this drawback was to covalently bond the acid groups, especially sulfonic acid groups, to the PBI chains [12–19]. The sulfonated PBI membranes showed relatively high proton conductivities comparing to the corresponding pristine PBI membranes. Due to the complicated synthesis routes for preparation of sulfonated PBIs and their poor solubility, sulfuric acids were also introduced to PBI membranes through blending PBI with sulfonated polymers. Blending PBI with Nafion® enhanced the stability [20] and depressed the methanol crossover [21] of the resulting composite membranes. Similar effects were also reported
∗ Corresponding author. Tel.: +886 3 2654130; fax: +886 3 2654199. E-mail address: [email protected] (Y.-L. Liu). 0376-7388/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.memsci.2009.01.045
with the blended composite membranes composed PBI/sulfonated polyether [22] and PBI/polyphosphazene [23]. Another approach to improve the thermal and chemical stability and to depress the methanol crossover of PEMs was formation of PEM nanocomposite membranes [24–29]. Addition of inorganic proton conductors to PBI membranes was demonstrated to enhance the proton conductivities and mechanical strengths of the membranes [29]. The improvements on the mechanical strengths and methanol barrier property of PBI membranes were also reported with PBI/silica (from a sol–gel process) [30] and PBI/montmorillonite [31] nanocomposite membranes. Unfortunately, decreases in the proton conductivities accompanied with the PBI nanocomposite membranes since addition of inorganic nanomaterials diluted the concentration of proton-conducting groups. The decreases in the proton conductivities of nanocomposite PEMs could be compensated using sulfonated nanomaterials as reinforcements [26–28]. Some acidic inorganic fillers were also utilized for PBI membranes to enhance their proton conductivity [32,33]. In this work, we reported the attempts on preparation of high performance PBI nanocomposite membranes for fuel cell application using sulfonated silica nanoparticles (SA-SNP) as nano-reinforcements. PBI nanocomposites membranes possessing non-sulfonated silica nanoparticles were also prepared to study the effects of sulfuric acid groups on the properties of PBI nanocomposite membranes. The sulfuric acid groups of SA-SNP formed ionic linkages with PBI chains, improved the compatibility between SASNP and PBI, and enhanced the mechanical strength of the PBI/SA-
122
Suryani, Y.-L. Liu / Journal of Membrane Science 332 (2009) 121–128
SNP nanocomposite membranes. PBI/SA-SNP nanocomposite membranes exhibited depressions on methanol permeability and enhancements on proton conductivity comparing to the pristine PBI membrane, to promise their application potentials for fuels cells. 2. Experimental 2.1. Materials 3,3 -Diaminobenzidine (DAB, 99%, Aldrich), isophthalic acid (IPA, 99%, Alfa Aesar), and polyphosphoric acid (PPA, ca. 84%, Alfa Aesar) were used as received. Sodium bicarbonate (99.5%) and ortho-phosphoric acid (85%, HPLC grade) were purchased from Showa Chemical Co. (Japan) and Scharlau Chemie (Germany), respectively. Methanol and N,N-dimethylacetamide (DMAc) in HPLC grade were received from TEDIA. DMAc was dried over 4 Å molecular sieves (Acros) for at least 2 days before use. Silica nanoparticles (SNP) in diameters of 10–20 nm were received from Nissan Chemical Co., Japan. Sulfonated silica nanoparticles were prepared in the laboratory according to the reported method [34]. The content of sulfuric acid group of SA-SNP is 0.57 mmol/g measured with elemental analysis.
and the heated to 900 ◦ C. Tensile properties were measured with an instron (Instron model 5543) with an elongation rate of 0.5 mm/min at room temperature. Dynamic mechanical analysis (DMA) was carried out with a Thermal Analysis dynamic mechanical analyzer (TA Q-800 DMA). The analytical conditions were tension force of 0.05N, frequency of 1 Hz, amplitude of 20 m, temperature range of 25–500 ◦ C, and heating rate of 5 ◦ C/min. Scanning electron microscopy (SEM) was performed on a field-emission SEM of Hitachi S-4800. Elemental analysis and element mapping were conducted with a Hitachi S-3000N SEM equipped with an energy dispersive X-ray spectroscopy (EDX) of Horiba EDX-250. 2.5. Water uptake and acid-doping level Water uptake measurement was performed by dipping the dried membranes (dried at 120 ◦ C in vacuo for 5 h) in distilled water for 24 h at 25 or 80 ◦ C. The membranes were taken out, wiped with paper, and then weighted. The water uptake was calculated by Eq. (1): Water uptake (%) =
Ws − Wd × 100 Wd
(1)
2.2. Synthesis of PBI PBI was synthesized by polycondensation reaction according to the methods described elsewhere [35]. 15 g of DAB and 250 g of PPA were added into a 1-L round bottom flask equipped with magnetic stirrer. The solution was heated to 140 ◦ C under a nitrogen stream to dissolve DAB in PAA. 11.63 g of IPA was then added and the temperature was raised to 200 ◦ C. After 15 h, the viscous solution was poured into a large amount of cold water. The fibers formed were kept in water for at least 3 days. The excess acid was neutralized with a sodium bicarbonate aqueous solution (1 M). The polymer was then collected from filtration, washed thoroughly with water and methanol, and then dried under vacuum at 130 ◦ C overnight. Yield: 81%. 1 H NMR (300 MHz, ppm) 13.23 (2H), 9.13 (1H), 8.30 (2H), 7.97 (1H), 7.79 (3H), 7.61 (3H); Inherent viscosity (0.5 g/dL in DMAc at 25 ◦ C): 1.42 dL/g. 2.3. Preparation of nanocomposite membranes A PBI solution was prepared by dissolving 1 g PBI in 20 mL DMAc at 120 ◦ C. A certain amount of SNP or SA-SNP was added. The homogeneous solution was filtered to remove trace precipitate, and then poured into a Petri dish. Solvent was gradually evaporated at 80 ◦ C overnight. Membranes were detached from the Petri dish by water immersion, treated in boiled water for 3 h, and then dried at 190 ◦ C for 3 h to remove the residual DMAc. Nanocomposite membranes possessing 0, 5, 10, 15, 20 wt% of SNP (PBI/SNP-X, X denoting to the SNP content) or SA-SNP (PBI/SA-SNP-X, X denoting to the SA-SNP content) were prepared. The membrane thickness was in the range of 40–60 m. 2.4. Characterization The inherent viscosity of PBI polymer was measured with a Cannon-Fenske Routine Viscometer at 25 ◦ C using a PBI solution in DMAc (0.5 g/dL). 1 H NMR spectrum of PBI was recorded on a Brüker MSL 300 NMR (300 MHz) using deuterated dimethylsulfoxide (DMSO-d6 ) as a solvent. FTIR spectra of nanocomposite membranes were recorded on a Perkin-Elmer Spectrum One FTIR. Thermogravimetric analysis (TGA) was performed by a ThermalAnalysis (TA) TGA Q-500 at a heating rate of 10 ◦ C under nitrogen. All samples were preheated at 120 ◦ C for 20 min to remove absorbed water from membranes then equilibrated at ca. 35 ◦ C for 30 min,
where Ws and Wd are membrane weights after and before dipping, respectively. Acid-doping level was obtained by a similar manner. Membranes were dipped in an 85% H3 PO4 solution for 1 week at room temperature. To exclude the weight gain due to water uptake, doped membranes were dried at 110 ◦ C under vacuum until the membrane weights unchanged with time. The weight change in acid doping was measured and used for the calculation of aciddoping level (acid numbers per PBI repeating unit). 2.6. Methanol permeability Methanol permeability measurement was performed on a tworeservoir compartment in which the membrane was clamped between the two reservoirs. The membranes were hydrated in deionized water at room temperature for 24 h prior to measurement. Source cell (VA = 20 mL) was filled with a 3-M methanol aqueous solution, and receiving cell (VB = 18 mL) was filled with deionized water. The concentration of methanol in receiving cell was measured versus time by gas chromatography GC390B (CL Sciences Inc., Japan) with a gas flow rate of 30 mL/min. The methanol concentration of the source cell was considered much higher than that of the receiving cell during measurements. Methanol permeability was calculated by Eq. (2): CB (t) =
A DKCA (t − t0 ) VB L
(2)
where CB is the methanol concentration in receiving cell, CA the initial methanol concentration, A the cross-section area, L the membrane thickness, D the methanol diffusivity and K the partition coefficient. The product DK is the membrane permeability to methanol. Parameter t0 is the “lag time” corresponding to the time necessary for the methanol to pass through membrane. 2.7. Proton conductivity Proton conductivity of the membranes was measured with a Solartron 1255B frequency response analyzer equipped with a Solartron 1287 electrochemical interface with an oscillation amplitude of 10 mV and a frequency range of 0.1 Hz to 1 MHz. The measurements were taken at 20–80 ◦ C under a 95% relative humidity.
Suryani, Y.-L. Liu / Journal of Membrane Science 332 (2009) 121–128
123
Fig. 1. Chemical structure of sulfonated silica nanoparticles (SA-SNP) and PBI.
3. Results and discussion 3.1. Characterizations on PBI nanocomposite membranes PBI and SA-SNP (Fig. 1) were prepared in our laboratory. PBI and SA-SNP could form homogeneous solutions. The PBI/SA-SNP nanocomposites membranes were therefore obtained with a casting method. PBI/SNP membranes were also prepared with uses of non-sulfonated SNP to replace SA-SNP to examine the effect of sul-
fonic acid groups on the properties and performances of the PBI nanocomposites membranes. Fig. 2 shows the SEM micrographs of the PBI/silica nanocomposite membranes. Pristine PBI membranes exhibited dense and homogeneous morphology for both surface (Fig. 2a) and cross-section (Fig. 2b) observations. The silica particles in PBI/SNP-10 membrane could be seen in its surface SEM micrograph (Fig. 2c). Particles in sizes of about 20 nm aggregate together to form silica domains in micrometers, indicating the interactions between silica nanoparticles are stronger than that between
Fig. 2. SEM micrographs of PBI ((a) surface and (b) cross-section) and its nanocomposites of PBI/SNP-10 ((c) surface and (d) cross-section) and PBI/SA-SNP-10 ((e) surface and (f) cross-section).
124
Suryani, Y.-L. Liu / Journal of Membrane Science 332 (2009) 121–128
Fig. 3. Elemental mapping of sulfur (a) and silicon (b) on PBI/SA-SNP-20 membrane observed with a SEM-EDX.
PBI molecules and silica nanoparticles. The agglomerated silica domains were also observed in the cross-sectional SEM micrograph of PBI/SNP-10 (Fig. 2d). The SEM observations indicate that the compatibility between PBI and silica nanoparticles and the homogeneity of the PBI/SNP membranes are somewhat poor. However, the dispersion homogeneity of silica nanoparticles in PBI membranes could be significantly improved with using SA-SNP as the additives. As shown in Fig. 2e, isolated silica particles were observed in the surface SEM micrograph of PBI/SA-SNP nanocomposite membranes, demonstrating the high compatibility between PBI polymer and SA-SNP particles and the high dispersion ability of SA-SNPs in PBI matrix. Particle agglomeration was still not observed in the cross-sectional SEM micrograph of PBI/SA-SNP-10 (Fig. 2f). PBI has a pKa value of about 5.5. The basicity of PBI makes it readily reactive toward the sulfonic acid groups of SA-SNPs through ionic interaction. The ionic linkages between PBI and SA-SNP contribute to their high compatibility. Fig. 3 shows the SEM-EDX elemental mapping micrographs of PBI/SA-SNP-20. Silicon and sulfur elements were observed to demonstrate the presence of SA-SNP particles in PBI nanocomposite membranes. In addition, the distributions of silicon and sulfur in the micrographs are homogeneous, indicating that there was not particle agglomeration and micro-phase separation in the nanocomposites membranes. Therefore, SA-SNP particles could be considered homogeneously dispersing in the PBI polymer matrix.
Fig. 4. FTIR spectra of PBI membrane and its nanocomposite membranes of PBI/SNP10 and PBI/SA-SNP-10.
Fig. 4 shows the FTIR spectra of PBI and its nanocomposite membranes. Characteristic absorption of PBI appeared in the spectra for all membranes [35]. Absorption peaks at 3390 and 3150 cm−1 correspond to the non-hydrogen bonded N–H stretching and hydrogen bonded (self-associated) N–H groups, respectively. The absorption peaks at 1438 and 1532 cm−1 arise from the in-plane deformation of benzimidazole, and the peak at 1282 cm−1 is from the breathing mode of imidazole rings. The absorption peak at about 1100 cm−1 in relatively low intensity might correspond to the in-plane C–H deformation vibrations [2]. After addition of silica nanoparticles, the absorption intensities at about 1100 cm−1 increased due to the contribution of Si–O–Si absorption. The result indicates the presence of silica in both PBI/SNP-10 and PBI/SA-SNP-10 nanocomposite membranes. In addition, PBI/SA-SNP-10 exhibited some minor absorption peaks at 1370 and 1168 cm−1 corresponding to the absorptions from the sulfuric acid groups. A decrease in the relative intensity of absorption peak at 3390 cm−1 was observed with PBI/SA-SNP-10, indicating some of the non-hydrogen bonded N–H groups in PBI chains bond to the sulfonic acid groups of SA-SNP via H-bonding and ionic interaction. 3.2. Properties of PBI nanocomposite membranes The glass transition temperatures (Tg ) of PBI/SA-SNP nanocomposite membranes were read from the tan ı peaks in DMA thermograms (figures not shown). Pristine PBI showed a Tg of 395 ◦ C, which shifted to higher temperatures (450–455 ◦ C) while formation of nanocomposites with SA-SNP. The Tg increase indicates that the motion ability of PBI chains is restricted with the presence of SA-SNP. The chain motion restriction should be majorly contributed from the strong ionic linkages between PBI and SA-SNP. The ionic linkages bring a cross-linking effect to the PBI/SA-SNP nanocomposite membranes, so as to increase the Tg s of the PBI/SA-SNP nanocomposites. The contributions of –SO3 H/–NH ionic linkages to Tg enhancement was also demonstrated with the relatively low Tg of PBI/SNP-20 (423 ◦ C), as SNP does not possess sulfonic acid groups. On the other hand, the presence of SNP or SA-SNP did not significantly alter the thermal degradation behaviors of PBI polymer, as shown with their TGA thermograms (Fig. 5). The –SO3 H/–NH ionic linkages in the SA-SNP/PBI nanocomposites membranes broke down before the degradation of PBI chains. Due to the presence of the silica nanoparticles did not involve in the degradation reactions of polymer portion [36], all PBI/SA-SNP nanocomposites as well as PBI polymer exhibited similar degradation temperatures, which was due to the degradation of PBI chains. However, the presence of inorganic portions in the PBI nanocomposite membranes might reduce the weight loss rate of the PBI nanocomposites in high temperature region and increase the char yields of the nanocomposites. The relatively high char yield of PBI/SA-SNP-20, comparing to that of PBI/SNP-20, is attributed
Suryani, Y.-L. Liu / Journal of Membrane Science 332 (2009) 121–128
Fig. 5. TGA thermograms of PBI-based membranes. PBI/SNP-10 and PBI/SNP-20: nanocomposites membranes of PBI and silica nanoparticles (SNP) possessing 10 and 20 wt% of SNP, respectively; PBI/SA-SNP-10 and PBI/SA-SNP-20: nanocomposites membranes of PBI and silica nanoparticles (SNP) possessing 10 and 20 wt% of SA-SNP, respectively.
to the promoting effect of sulfonic acid groups on the dehydration/carbonization during PBI thermal degradation. The mechanical properties of the PBI-based membranes were measured with an Instron. The stress–strain curves are shown in Fig. 6. Pristine PBI membrane showed a Young’s modulus of 2.9 GPa and an elongation at break of 40%. Formation of nanocomposites with silica nanoparticles results in increases in tensile strengths and decreases in elongations at break for the PBI-based membranes. For PBI/SA-SNP nanocomposites membranes, the presence of ionic linkages between PBI and silica particles attributes to their high Young’s modulus, especially for high SA-SNP loaded samples of PBI/SA-SNP-15 and PBI/SA-SNP-20 (Young’s modulus above 3.9 MPa). Introduction of silica nanoparticles increased the brittleness of the PBI-based membranes with the decreases in their elongations at break. The brittleness of the membranes increased with increasing their silica contents. The effects of SNP and SA-SNP on the elongations at break of the nanocomposite membranes were
Fig. 6. Stress–strain curves of the nanocomposite membranes of (i) PBI and silica nanoparticles (SNP): PBI/SNP-X, X = 5, 10, 15, 20 which denoting to the SNP weight percents in the nanocomposites and (ii) PBI and sulfonated silica nanoparticles (SASNP): PBI/SA-SNP-X, X = 5, 10, 15, 20 which denoting to the SA-SNP weight percents in the nanocomposites.
125
Fig. 7. Water uptakes of the PBI nanocomposite membranes of (i) PBI and silica nanoparticles (SNP) nanocomposites (PBI/SNP) and (ii) PBI and sulfonated silica nanoparticles (SA-SNP) nanocomposites (PBI/SA-SNP) at 25 and 80 ◦ C.
similar. The decreases in the elongation at break and increases in the brittleness were widely observed with polymer nanocomposites and could be attributed to the presence of inorganic reinforcement. Moreover, the ionic linkages in PBI/SA-SNP as well as the hydrogen bonding in PBI/SNP also bring some cross-linking effect to the PBI chains so as to in decrease the elongation at break of the polymer. Therefore, the PBI/SA-SNP membranes could be superior to the PBI/SNP membranes in mechanical characteristics. Moreover, the high brittleness of PBI/SNP-20 and PBI/SA-SNP-20 (possessing 9% of elongation at break) might limit their applications for fuel cell membranes. Comparatively, PBI/SA-SNP-15, which showed an elongation at break of 16%, is more suitable for the membrane applications. PBI exhibit a water uptake of 16.6% at 25 ◦ C, corresponding to 2.74 molecules of water associating with one PBI repeating unit. The measured value of water uptake is reasonable comparing to the proposed values by Li et al. [5], who suggested that N atom and N–H groups in PBI repeating units formed intermolecular hydrogen bonds with 2–4 water molecules. Addition of silica nanoparticles to PBI membranes increased their water uptakes (Fig. 7) due to the hygroscopic nature of silica nanoparticles. The sulfonic acid groups in SA-SNP nanoparticles further increased the water uptakes to greater extents. PBI/SNP-20 and PBI/SA-SNP-20 showed a water uptake of 21.6 and 23.8 wt% at 25 ◦ C, respectively. Besides, the water uptakes of the PBI-based membranes measured at 25 ◦ C were higher than the values measured at 80 ◦ C. Sannigrahi et al. [37] reported that PBI polymer chain shrank in DMAc at elevated temperature, which was caused by disruption of interchain hydrogen bonding, and underwent conformational change from extended helical conformer to collapsed compact coil conformer. The low water uptakes observed with PBI membranes at high temperatures might be attributed to the PBI chain shrinkages in water. Similar results were also reported with Qing et al. [13,14] for sulfonated PBI polymers. The PBI-based membranes were doped with 85% H3 PO4 to prepare the acid-doped membranes for fuel cell applications. Fig. 8 shows the acid-doping levels of the PBI-based nanocomposite membranes. Acid-doping level (H3 PO4 molecules per PBI repeating unit) of pristine PBI membrane was 12.3, which decreased to 9.8 and 8.6 for PBI/SNP-20 and PBI/SA-SNP-20, respectively (Fig. 8). Some of the amino groups in PBI chains were blocked with the sulfonic acid groups of SA-SNP, to decrease the binding ability of PBI
126
Suryani, Y.-L. Liu / Journal of Membrane Science 332 (2009) 121–128
Fig. 8. Acid-doping levels of PBI nanocomposite membranes with H3 PO4 : (i) PBI and silica nanoparticles (SNP) nanocomposites (PBI/SNP) and (ii) PBI and sulfonated silica nanoparticles (SA-SNP) nanocomposites (PBI/SA-SNP).
chains to H3 PO4 molecules. Although high acid-doping levels are expected for fuel cell applications, these prepared PBI nanocomposites membranes still showed high proton conductivities, to be discussed below. 3.3. Methanol permeability and proton conductivity of PBI nanocomposite membranes The methanol permeability of PBI-based membranes was measured with a 3-M methanol aqueous solution at 25 ◦ C by a side-by-side method. Pure PBI membrane (without acid doping) exhibited a methanol permeability of 1.37 × 10−9 cm2 /s, which increased to 5.5 × 10−7 cm2 /s after being doped with H3 PO4 . The protonation of PBI chains with acid doping reduced the intermolecular interaction forces of PBI and increased the free volume of PBI, consequently to generate some additional spaces for methanol crossing the PBI membrane. The presence of silica nanoparticles in the PBI nanocomposite membranes played some counteraction to this effect. As shown in Fig. 9, the PBI nanocomposites exhibited relatively low methanol permeability. The increases in methanol
Fig. 9. Methanol permeability of PBI nanocomposite membranes: (i) PBI and silica nanoparticles (SNP) nanocomposites (PBI/SNP) and (ii) PBI and sulfonated silica nanoparticles (SA-SNP) nanocomposites (PBI/SA-SNP).
Fig. 10. Proton conductivities of PBI nanocomposite membranes at 95% relative humidity: (i) PBI and silica nanoparticles (SNP) nanocomposites (PBI/SNP) and (ii) PBI and sulfonated silica nanoparticles (SA-SNP) nanocomposites (PBI/SA-SNP). The proton conductivity of the Nafion® 117 membrane is also shown for comparison.
permeability of PBI/SNP-15 and PBI/SNP-20 membranes could be due to the poor compatibility between PBI and SNP particles and the agglomeration of SNPs in the PBI/SNP nanocomposites membranes. Since SA-SNP particles are much compatible to PBI, the continuing decreases in the methanol permeability were reasonably observed with PBI/SA-SNP-15 and PBI/SA-SNP-20. It is noteworthy that the methanol permeability of PBI/SA-SNP-20 membrane (3.3 × 10−7 cm2 /s) is of about one-tenth of the value read with Nafion® (3.7 × 10−6 cm2 /s) [24]. The proton conductivities of the phosphoric acid-doped PBI membranes at 95% relative humidity as a function of temperatures are shown in Fig. 10. The proton conductivities of all membranes increased with increasing the operation temperatures. Addition of the un-modified SNP showed a negative effect on the proton conductivities of the PBI/SNP membranes. Similar results were also reported to other nanocomposites membranes like Nafion/silica membranes [23,38] and PBI/silica membranes [30]. On the other hand, comparing to the pristine PBI and Nafion® 117 membranes, the relatively high proton conductivities of the PBI/SA-SNP nanocomposites membranes are noteworthy. Li et al. [24] used sulfonated organosilica to modify Nafion® membranes and they found the proton conductivities of the modified membranes were still lower than that of the neat Nafion® membrane. The reduction in the proton conductivity was attributed to the silica-induced changes in the tortuous paths and in the distribution of hydrophilic/hydrophobic domains. However, in this work an enhancing effect on the proton conductivity was observed with addition of sulfonated silica nanoparticles to PBI membranes. High proton conductivities were observed with the PBI/SA-SNP nanocomposites membranes, even the membranes possessed relatively low acid-doping levels comparing to the pristine PBI membrane. High doping levels have been recognized to result in high proton conductivities for H3 PO4 -doped PBI membranes, due to the excess H3 PO4 could form anion pathways for proton conduction [4]. However, the H3 PO4 doping levels of the PBI/SA-SNP membranes were still high enough for proton conduction [29]. The sulfonic acid groups of SA-SNP also contribute to the acid-doping effect to PBI polymer so as to enhance the proton conductivity of PBI/SA-SNP nanocomposites membranes [12–19]. Moreover, the presence of SA-SNP particles induced proton conductive path-
Suryani, Y.-L. Liu / Journal of Membrane Science 332 (2009) 121–128
127
The high proton conductivities and low methanol permeability of the PBI/SA-SNP nanocomposite membranes indicate their potentials of applications as the proton exchange membranes in direct methanol fuel cells (DMFC), which are evaluated with the membrane selectivity, i.e. the ratio of the proton conductivity to the methanol permeability [39]. The relative selectivity of the PBI-based membrane is higher than or comparable to that of the Nafion® 117 membrane as shown in Fig. 12. PBI/SA-SNP15 membrane shows the highest selectivity among the evaluated membranes, which is of about 1.3-folds of the selectivity of the Nafion® 117 membrane. Moreover, the PBI/SA-SNP membranes could be highly superior over the Nafion® 117 membrane at high temperatures and dry conditions, since under this operation condition the Nafion® 117 membrane shows a very low selectivity. 4. Conclusions
Fig. 11. Proton conductivities of PBI/SA-SNP nanocomposite membranes at high temperature region and under dry environment. The proton conductivity of the Nafion® 117 membrane is also shown for comparison.
ways attributing to the increases in the proton conductivities of PBI/SA-SNP membranes. The phenomena were similar to the results reported in our previous work [28]. The proton conductivities of the acid-doped PBI/SA-SNP nanocomposites membranes were further studied at high temperatures under an anhydrous environment (Fig. 11). All the PBI/SA-SNP nanocomposite membranes still exhibited higher proton conductivities that did the pristine acid-doped PBI membrane. Since the Grotthuss-type diffusion mechanism without assistance of the vehicle mechanism has been proposed for the proton conduction under anhydrous environments [3,10], the proton conductivity of the acid-doped PBI/SA-SNP membranes could be synergistically resulted from their SA-SNP contents and the acid-doping levels. Consequently, the highest proton conductivity was obtained with the acid-doped PBI/SA-SNP-10 membrane, which possess the moderate SA-SNP content and acid-doping level. It is noteworthy that Nafion® 117 shows relatively low proton conductivities under the anhydrous condition and almost loses its proton conductivities at high temperatures. Comparing to Nafion® 117 membrane, the high proton conductivities of PBI/SA-SNP-10 indicate that this membrane could be utilized as proton exchange membranes for fuel cells at high temperatures and dry conditions.
Fig. 12. The relative selectivity (over the Nafion® 117 membrane) of membranes of PBI and sulfonated silica nanoparticle (SA-SNP) nanocomposites (PBI/SA-SNP).
Sulfonated silica nanoparticles were effective additives for preparation of homogeneous PBI/SA-SNP nanocomposite membranes. The sulfonic acid groups of SA-SNP enhanced the compatibility between silica nanoparticles and PBI polymer matrix and improved the mechanical properties of the resulting PBI/SASNP nanocomposite membranes. Incorporation of SA-SNP to PBI membranes also depressed their methanol permeability and increased their proton conductivities, to result in their high selectivity comparing to pristine PBI and Nafion® membranes. The performances of the PBI/SA-SNP nanocomposite membranes are attractive and the membrane could be used as proton exchange membranes for direct methanol fuel cells and high temperature/low humidity proton exchange membrane fuel cells. Acknowledgements We thank The Ministry of Education, Taiwan for their financial support on this work under the Center-of-Excellence Program on Membrane Technology (2006–2010). Special thanks are given to Professor Steve Lien-Chung Hsu (National Cheng Kung University, Taiwan) for his kind help and discussion on the proton conductivity measurements. References [1] J.S. Wainright, J.-T. Wang, D. Weng, R.F. Savinell, M.H. Litt, Acid doped polybenzimidazoles: a new polymer electrolyte, J. Electrochem. Soc. 142 (1995) L121. [2] X. Glipa, B. Bonnet, B. Mula, D.J. Jones, J. Rozière, Investigation of the conduction properties of phosphoric and sulfuric acid doped polybenzimidazole, J. Mater. Chem. 9 (1999) 3045–3049. [3] Y.L. Ma, J.S. Wainright, M.H. Litt, R.F. Savinell, Conductivity of PBI membranes for high-temperature polymer electrolyte fuel cells, J. Electrochem. Soc. 151 (2004) A8–A16. [4] Q. Li, R. He, J.O. Jensen, N.J. Bjerrum, PBI-based polymer membranes for high temperature fuel cells-preparation, characterization, and fuel cell demonstration, Fuel Cells 4 (2004) 147–159. [5] Q. Li, R. He, R.W. Berg, H.A. Hjulers, N.J. Bjerrum, Water uptake and doping of polybenzimidazoles as electrolyte membranes for fuel cells, Solid State Ionics 168 (2004) 177–185. [6] J.A. Asensio, S. Borrós, P. Gómez-Romero, Polymer electrolyte fuel cells based on phosphoric-acid impregnated poly(2,5-benzimidazole) membranes, J. Electrochem. Soc. 151 (2004) A304–A310. [7] H. Pu, Q. Liu, L. Qiao, Z. Yang, Studies on proton conductivity of acid doped polybenzimidazole/polyimide and polybenzimidazole/polyvinylpyrrolidone blends, Polym. Eng. Sci. 45 (2005) 1395–1400. [8] R. He, Q. Li, A. Bach, J.O. Jensen, N.J. Bjerrum, Physicochemical properties of phosphoric acid doped polybenzimidazole membranes for fuel cells, J. Membr. Sci. 277 (2006) 38–45. ˜ [9] J. Lobato, P. Canizares, M.A. Rodrigo, J.J. Linares, J.A. Aguilar, Improved polybenzimidazole films for H3 PO4 -doped PBI-based high temperature PEMFC, J. Membr. Sci. 306 (2007) 47–55. [10] R. He, Q. Li, J.O. Jensen, N.J. Bjerrum, Doping phosphoric acid in polybenzimidazole membranes for high temperature proton exchange membrane fuel cells, J. Polym. Sci. Part A: Polym. Chem. 45 (2007) 2989–2997.
128
Suryani, Y.-L. Liu / Journal of Membrane Science 332 (2009) 121–128
[11] B.Z. Xing, O. Savadogo, The effect of acid doping on the conductivity of polybenzimidazole, J. New Mater. Electrochem. Syst. 2 (1999) 95–101. [12] J.A. Asensio, S. Borros, P. Gomez-Romero, Proton-conducting polymers based on benzimidazole and sulfonated benzimidazole, J. Polym. Sci. Part A: Polym. Chem. 40 (2002) 3703–3710. [13] S. Qing, W. Huang, D. Yan, Synthesis and characterization of thermally stable sulfonated polybenzimidazoles obtained from 3,3 -disulfonyl-4,4 dicarboxyldiphenylsulfone, J. Polym. Sci. Part A: Polym. Chem. 43 (2005) 4363–4372. [14] S. Qing, W. Huang, D. Yan, Synthesis and properties of soluble sulfonated polybenzimidazoles, React. Funct. Polym. 66 (2006) 219–227. [15] J. Jouanneau, R. Mercier, L. Gonon, G. Gebel, Synthesis of sulfonated polybenzimidazoles from functionalized monomers: preparation of ionic conducting membranes, Macromolecules 40 (2007) 983–990. [16] Z. Bai, G.E. Price, M. Yoonessi, S.B. Juhl, M.F. Durstock, T.D. Dang, Proton exchange membranes based on sulfonated polyarylenethioethersulfone and sulfonated polybenzimidazole for fuel cell applications, J. Membr. Sci. 305 (2007) 69–76. [17] N. Li, Z. Cui, S. Zhang, W. Xing, Sulfonated polyimide bearing benzimidazole groups for proton exchange membranes, Polymer (2007) 7255–7263. [18] H. Xu, K. Chen, X. Guo, J. Fang, J. Yin, Synthesis of novel sulfonated polybenzimidazole and preparation of cross-linked membranes for fuel cell application, Polymer 48 (2007) 5556–5564. [19] J. Peron, E. Ruiz, D.J. Jones, J. Roziere, Solution sulfonation of novel polybenzimidazole a proton electrolyte for fuel cell application, J. Membr. Sci. 314 (2008) 247–256. [20] Y. Zhai, H. Zhang, Y. Zhang, D. Xing, A novel H3 PO4 /Nafion-PBI composite membrane for enhanced durability of high temperature PEM cells, J. Power Sources 169 (2007) 259–264. [21] J.K. Lee, J. Kerres, Synthesis and characterization of sulfonated poly(arylene thioether)s and their blends with polybenzimidazole for proton exchange membranes, J. Membr. Sci. 294 (2007) 75–83. [22] R. Wycisk, J.K. Lee, P.N. Pintauro, Sulfonated polyphosphazene-polybenzimidazole membranes for DMFCs, J. Electrochem. Soc. 152 (2005) A892–A898. [23] N. Miyake, J.S. Wainright, R.F. Savinell, Evaluation of a sol–gel derived Nafion/silica hybrid membrane for proton electrolyte membrane fuel cell applications. I. Proton conductivity and water content, J. Electrochem. Soc. 148 (2001) A898–A904. [24] C. Li, G. Sun, S. Ren, J. Liu, Q. Wang, Z. Wu, H. Sun, W. Jin, Casting Nafionsulfonated organosilica nano-composite membranes used in direct methanol fuel cells, J. Membr. Sci. 272 (2006) 50–57. [25] Y.H. Su, Y.L. Liu, Y.M. Sun, J.Y. Lai, M.D. Guiver, Y. Gao, Using silica nanoparticles modifying sulfonated poly(phthalazinone ether ketone) membrane for direct
[26]
[27] [28]
[29]
[30]
[31]
[32] [33]
[34]
[35]
[36]
[37]
[38]
[39]
methanol fuel cell: a significant improvement on cell performance, J. Power Sources 155 (2006) 111–117. C.H. Rhee, H.K. Kim, H. Chang, J.S. Lee, Nafion® /sulfonated montmorillonite composite: a new concept electrolyte membrane for direct methanol fuel cells, Chem. Mater. 17 (2005) 1691–1697. P. Bébin, M. Caravanier, H. Galiano, Nafion® /clay–SO3 H membrane for proton exchange membrane fuel cell application, J. Membr. Sci. 278 (2006) 35–42. Y.H. Su, Y.L. Liu, Y.M. Sun, J.Y. Lai, D.M. Wang, Y. Gao, B. Liu, M.D. Guiver, Proton exchange membranes modified with sulfonated silica nanoparticles for direct methanol fuel cells, J. Membr. Sci. 296 (2007) 21–28. R. He, Q. Li, G. Xiao, N.J. Bjerrum, Proton conductivity of phosphoric acid doped polybenzimidazole and its composites with inorganic proton conductors, J. Membr. Sci. 226 (2003) 169–184. S.W. Chuang, S.L.C. Hsu, Y.H. Liu, Synthesis and properties of fluorine-containing polybenzimidazole/silica nanocomposite membranes for proton exchange membrane fuel cells, J. Membr. Sci. 305 (2007) 353–363. S.W. Chuang, S.L.C. Hsu, Y.H. Liu, Synthesis and properties of fluorine-containing polybenzimidazole/montmorillonite nanocomposite membranes for direct methanol fuel cell applications, J. Power Sources 168 (2007) 172–177. P. Staiti, Proton conductive membranes absorbed on silicotungstic acid/silica and polybenzimidazole, Mater. Lett. 47 (2001) 241–246. P. Staiti, M. Minutoli, S. Hocevar, Membranes based on phosphortungstic acid and polybenzimidazole for fuel cell application, J. Power Sources 90 (2000) 231–235. Y.L. Liu, C.Y. Hsu, Y.H. Su, J.Y. Lai, Chitosan–silica complex membranes from sulfuric acid functionalized silica nanoparticles for pervaporation dehydration of ethanol–water solutions, Biomacromolecules 6 (2005) 368–373. ˜ J. Lobato, P. Canizares, M.A. Rodrigo, J.J. Linares, G. Manjavacas, Synthesis and characterization of poly[2,2-(m-phenylene)-5,5-bibenzimidazole] as polymer electrolyte membrane for high temperature PEMFCs, J. Membr. Sci. 280 (2006) 351–362. Y.L. Liu, C.Y. Hsu, W.L. Wei, R.J. Jeng, Preparation and thermal properties of epoxy-silica nanocomposites from nanoscale colloidal silica, Polymer 44 (2003) 5159–5167. A. Sannigrahi, D. Arunbabu, R.M. Sankar, T. Jana, Aggregation behavior of polybenzimidazole in aprotic polar solvent, Macromolecules 40 (2007) 2844–2851. R. Jiang, H.R. Kunz, J.M. Fenton, Composite silica/Nafion® membranes prepared by tetraethylorthosilicate sol–gel reaction and solution casting for direct methanol fuel cells, J. Membr. Sci. 272 (2006) 116–124. D.K. Kim, G.P. Robertson, M.D. Guiver, Comb-shaped poly(arylene ether sulfone)s as proton exchange membranes, Macromolecules 41 (2008) 2126–2134.