Sulfonated poly(ether ether ketone)-based hybrid membranes containing graphene oxide with acid-base pairs for direct methanol fuel cells

Accepted Manuscript Title: Sulfonated poly(ether ether ketone)-based hybrid membranes containing graphene oxide with acid-base pairs for direct methanol fuel cells Author: Yongheng Yin Haiyan Wang Li Cao Zhen Li Zongyu Li Mingyue Gang Chongbin Wang Hong Wu Zhongyi Jiang Peng Zhang PII: DOI: Reference:

S0013-4686(16)30839-8 http://dx.doi.org/doi:10.1016/j.electacta.2016.04.040 EA 27075

To appear in:

Electrochimica Acta

Received date: Revised date: Accepted date:

17-12-2015 8-3-2016 9-4-2016

Please cite this article as: Yongheng Yin, Haiyan Wang, Li Cao, Zhen Li, Zongyu Li, Mingyue Gang, Chongbin Wang, Hong Wu, Zhongyi Jiang, Peng Zhang, Sulfonated poly(ether ether ketone)-based hybrid membranes containing graphene oxide with acid-base pairs for direct methanol fuel cells, Electrochimica Acta http://dx.doi.org/10.1016/j.electacta.2016.04.040 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Sulfonated poly(ether ether ketone)-based hybrid membranes containing graphene oxide with acid-base pairs for direct methanol fuel cells

Yongheng Yina,b, Haiyan Wanga,b, Li Caoa,b, Zhen Lia,b, Zongyu Lia,b, Mingyue Ganga,b, Chongbin Wanga,b, Hong Wua,b,c *, Zhongyi Jianga,b, Peng Zhangd

a

Key Laboratory for Green Chemical Technology of Ministry of Education, School of

Chemical Engineering and Technology, Tianjin University, Tianjin 300072, China b

Collaborative Innovation Center of Chemical Science and Engineering (Tianjin),

Tianjin 300072, China c

State Key Laboratory of Separation Membranes and Membrane Processes, School of

Materials Science and Engineering, Tianjin Polytechnic University, Tianjin 300387, China d

Key Laboratory of Nuclear Analysis Techniques, Institute of High Energy Physics,

Chinese Academy of Sciences, Beijing 100049, China

*Corresponding

author: School of Chemical Engineering and Technology, Tianjin University, Tianjin 300072, P R China. Tel: 86-22-23500086. Fax: 86-22-23500086. E-mail: [email protected] (H. Wu) 1

Graphical abstract

H+

MeOH

Abstract The graphene oxide (GO) sheets are functionalized by histidine molecules and incorporated into sulfonated poly (ether ether ketone) (SPEEK) matrix to fabricate hybrid polymer electrolyte membranes for direct methanol fuel cells (DMFCs). The loading of functionalized GO is varied to investigate its influence on cross-sectional morphology, crystalline structure, polymer chain stiffness, thermal stability and fractional free volume of membrane, etc. The acidic –SO3H groups (proton donors) in SPEEK and basic imidazole groups (proton acceptors) in histidine molecules form acid-base pairs and transport protons synergistically, thus yielding efficient proton 2

channels inside the hybrid membranes. The maximum proton conductivity (at 100% RH) of hybrid membranes is elevated by 30.2% compared with the plain SPEEK membrane at room temperature. The functionalized GO flakes also confer the hybrid membranes low methanol permeability in the range of 1.32 – 3.91 × 10-7 cm2 s-1. At the filler content of 4 wt%, the hybrid membrane shows a superior selectivity of 5.14 × 105 S s cm-3 and its maximum power density of single DMFC cell (43.0 mW cm-2) is 80.7% higher than that of plain SPEEK membrane. Keywords: Graphene oxide; Hybrid membrane; Acid-base pair; Proton conductivity; Methanol permeability

1. Introduction Direct methanol fuel cell (DMFC) has received tremendous attention in recent years as methanol is one of the most promising fuels besides hydrogen due to its abundant source, high energy content, easy storage and transportation, etc. [1] As the performance-limiting component of DMFC, polymer electrolyte membrane (PEM) needs to meet two basic requirements, high proton conductivity and low/zero methanol permeability [2, 3]. Unfortunately, like the commonly found trade-off effect between separation factor and permeability for polymeric membranes used in separation processes [4, 5], a similar trade-off limit has also been found in DMFC 3

membranes, i.e. increasing the methanol resistance property (reducing methanol permeability) usually compromises the proton conductivity and vice versa [6]. This is because that the water molecules in membrane will promote the proton conduction via Vehicular mechanism (mainly in the form of H3O+) and the diffusion of methanol molecules simultaneously. Developing organic-inorganic hybrid membranes seems to be an effective approach to overcome the trade-off effect between proton conductivity and methanol permeability [7, 8]. The recent discoveries and researches of 2D carbonaceous materials represented by graphene oxide (GO) have aroused particular interest in the area of new and high-performance materials [9-11]. It has been demonstrated that the protons can be transferred by the carboxylic acid groups along edges of GO and epoxy groups on the surface of GO [12, 13], consequently, the proton conductivity of single-layer GO can reach as high as 10-2 S cm-1 [11]. Additionally, the GO nanosheets also possess an outstanding blocking effect on methanol molecules [14].These unique characteristics make GO a promising candidate as inorganic component in the fabrication of hybrid proton conductive membranes. Currently, graphene oxide sheets are usually sulfonated by either direct oxidation [15, 16], physical adsorption [17] or chemical grafting [18], and then blended with polymer to make hybrid PEMs. Nature offers various inspirations for fabricating new membrane materials [19-21]. Ion channels, which are ubiquitous in cell membrane, play an important role in intercellular mass transport such as proton transfer [22, 23]. Usually, the proton 4

transfer occurs along the amino acid-lined channels with high efficiency, for example, 105 protons could be conducted per second through one channel in adenosine triphosphate (ATP) synthase systems [24]. It has been demonstrated that for acid-base pairs, the acidic groups act as proton donors and the basic groups act as proton acceptors, resulting in efficient proton conduction [25]. Based on these findings, acid-base pairs have been introduced into polymer to fabricate PEMs [26-28]. Tributsch et al. incorporated silica nanoparticles attached with lysine into the pores of polyethylene terephthalate [24], the power output of the obtained membrane with acid-base pairs approached the commercial Nafion® membrane. Wang et al. used dopamine-modified silica nanoparticles to prepare sulfonated poly (ether ether ketone)-based hybrid membranes [29], the hybrid membrane achieved a high conductivity of 4.52 × 10-3 S cm-1 at 120 οC under anhydrous condition. We have previously introduced acid-base pair functionalized titania-silica into chitosan membrane and found that the highest selectivity of the as-prepared hybrid membrane was 4.85 × 104 S s cm-3, which was ~ 3 times higher than that of pure chitosan membrane [30]. The aim of this study is developing hybrid membranes using functionalized GO as additives in order to pursue both high proton conductivity and strong methanol resistance for DMFC application. Herein, GO is functionalized by histidine molecules and then incorporated into sulfonated poly (ether ether ketone) (SPEEK) to prepare organic-inorganic hybrid membranes with acid-base pairs. The fabricated hybrid membranes are characterized by SEM, FTIR, DSC, PALS, etc, and the membrane is 5

made into membrane electrode assembly for single DMFC performance test. The reason why as-prepared hybrid membranes could solve the trade-off effect between proton conductivity and methanol permeability is discussed. 2. Experimental 2.1 Materials and chemicals Poly (ether ether ketone) (PEEK) was purchased from Victrex High-performance Materials Co., Ltd (Shanghai, China). Histidine (purity 98%), oxalic acid dihydrate (analytical grade), calcium hydride (analytical grade) and thionyl chloride (analytical grade) were obtained from Aladdin. Sulfuric acid (H2SO4), N, N-dimethylformamide (DMF), tetrahydrofuran (THF), methanol and other reagents were purchased from Guangfu Fine Chemical Research Institute (Tianjin, China). DMF was dehydrated by stirring with calcium hydride and distilling under reduced pressure before use. Deionized water was used in all the experiments. 2.2 Preparation and functionalization of graphene oxide Pristine graphene oxide (GO) was synthesized using a modified Hummers method [31], then pristine GO was esterified to produce more carboxylic acid groups [32]. Typically, the prepared GO was suspended in 100 mL of H2O (5 mg mL-1), 6 g of oxalic acid dehydrate was added into the mixture and stirred for 24 h. Subsequently, the suspension was washed with H2O for 3 times and finally dispersed in 20 mL of DMF. The GO was functionalized by histidine through an amidation reaction. Firstly, the above suspension was added into a three-necked flask containing 100 mL of thionyl 6

chloride under N2 atmosphere. The flask was equipped with a water-cooled reflux condenser and heated at 65 °C for 24 h. The GO with acid chloride groups (denoted as GO-COCl) was washed with THF until the pH of filtrate reached neutral. Secondly, GO-COCl was mixed with aqueous solution of histidine (5 mg mL-1), and the suspension was heated at 100 °C for 40 h. Finally, the histidine-functionalized GO (denoted as GO-his) was filtered, washed with H2O, and dried under vacuum at 40 °C for 48 h. 2.3 Preparation of sulfonated poly (ether ether ketone) (SPEEK) SPEEK was prepared by a post-sulfonation method. Typically, 18 g of dried PEEK was dissolved gradually in 200 mL of concentrated sulfuric acid (98%) for 4 h at room temperature, then the mixture was stirred vigorously at 45 οC for 8 h. After reaction, the solution was slowly precipitated in water under mechanical agitation. The obtained SPEEK precipitate was filtered, washed several times with water until the pH reached neutral. Finally, the SPEEK polymer was dried at room temperature for 24 h and at 60 οC for another 24 h. The sulfonation degree of as-prepared SPEEK was measured to be 66% by acid-base titration. 2.4 Preparation of hybrid membranes 0.6 g of SPEEK was dissolved in 4 g of DMF under stirring at 25 oC. A premeasured amount of GO-his was dispersed in 4 g of DMF under ultrasonication, then the GO-his suspension was added into the SPEEK solution, followed by vigorously stirring for 12 h to ensure a fine dispersion. The mixture was degasified and cast onto a glass plate, followed by successively dried at 60 οC for 12 h and 7

annealing at 80 οC for another 12 h. The prepared hybrid membranes were designated as SPEEK/GO-his-X, where X was the weight ratio percentage of inorganic fillers to SPEEK. As a comparison, plain SPEEK membrane and hybrid membranes incorporated with pristine GO (SPEEK/GO-X) were also fabricated. The thickness of all the membranes was in the range of 60 – 65 μm. 2.5 Characterizations Transmission electron microscopy (TEM) observation was performed by a Tecnai G2 F20. Fourier transform infrared spectra (FTIR) of GO and membranes were recorded on a Nicolet MAGNA-IR 560 spectrometer in the wavenumber range of 4000 – 400 cm-1. The elemental composition of GO was investigated by X-ray photoelectron spectroscopy (XPS, PHI-1600). The cross-sectional morphologies of the membranes were observed using magnified field emission scanning electron microscope (FESEM, Hitachi S-4800) operated at 3 kV and FESEM (Nanosem 430) operated at 10 kV. The crystalline structure was characterized with X-ray diffractometer (XRD, RigakuD/max2500v/Pc) using Cu Kα radiation, the scanning angle ranged from 2° to 55° with a scanning rate of 2° min-1. Glass transition temperature was measured by differential scanning calorimetry (DSC, 204 F1 NETZSCH), samples were preheated under N2 atmosphere from 25 to 150 °C at 10 °C min-1, then cooled to 90 °C and reheated to 260 °C. Thermal stability was characterized by thermogravimetric analyzer (TGA, Perkin-Elmer Pyris), each sample was heated under N2 flow over the temperature range of 20 – 800 oC at a heating rate of 10 oC min-1. The mechanical properties of the membranes were measured using an 8

electronic tensile machine (WDW-2, Yangzhou Zhongke Measuring Apparatus Co., China) with an elongation rate of 10 mm min-1 at room temperature. The water uptake and swelling degree of membranes were measured as described in literature [33] and calculated by Eqs. (1) and (2), respectively:

Water uptake (%) 

Ww  Wd  100 Wd

Swelling degree(%) 

(1)

Aw  Ad  100 Ad

(2)

where Wd and Ad are the weight and cross-sectional area (length × width) of the dry membrane, Ww and Aw are the weight and cross-sectional area of the wet membrane after complete hydration at 25 °C. The ion-exchange capacity (IEC) of membranes was measured by a classical acid-base titration method as described in our previous work [30]. Positron annihilation lifetime spectroscopy (PALS) experiment was performed with an EG&G ORTEC fast-fast coincidence system. The

22

Na source was sandwiched

between two pieces of membrane samples (thickness 1.0 mm). Assuming that the location of o-Ps occurs in a sphere potential well surrounded by an electron layer of a constant thickness Δr (0.1656 nm), the radius of free volume cavity (r3) was calculated from the pick-off annihilation lifetime of o-Ps (τ3) by the following semi-empirical Eq.: r 1  1   2r3   3  1  3    sin  2  r3  r  2   r3  r 

1

(3)

The volume of equivalent sphere (Vf) and the fractional free volume (FFV) could be calculated by Eqs. (4) and (5), respectively, 9

4 3 r3 3

(4)

FFV  Vf I 3

(5)

Vf 

where I3 is the intensity of o-Ps. 2.6 Measurement of proton conductivity, methanol permeability and selectivity The membrane impedance was measured using a frequency response analyzer (PARSTAT 2273, Princeton) over the frequency range of 1 Hz – 1 MHz. For the measurement of proton conductivity at 100% relative humidity (RH), the temperature was controlled from 25 to 75 οC. For the test of proton conductivity at low RH, the saturated steam of KCl, NaNO3, NaBr and MgCl2 was used to keep the RH at 78.9%, 65.4%, 51.4% and 26.1%, respectively. The proton conductivity (σ, S cm-1) was calculated by:



l0 AI

(6)

in which l0, A and I denote the distance between the electrodes, the effective membrane area and the membrane impedance, respectively. The methanol permeability of membranes was measured as described in our previous work [34] using a glass diffusion cell, which includes a methanol compartment and a water compartment. The methanol permeability (P, cm2 s-1) was determined by the following Eq.: P  S'

VBl AC A0

(7)

where VB is the volume of water compartment, CA0 is the feed concentration of methanol in methanol compartment (1, 2, 5 or 10 M), l and A are the thickness and 10

effective area of membrane respectively, S’ can be derived from the slope of the straight line of methanol concentration versus time. The comprehensive performance of membranes was evaluated by selectivity (β, S s cm-3), which is defined as:





(8)

P

2.7 Single DMFC Performance The electrocatalysts used in anode and cathode were PtRu/C (60 wt%) and Pt/C (60 wt%), respectively. The catalyst ink was prepared by mixing catalyst with 5% Nafion in isopropanol water solution, then required amount of catalyst (1.0 mg cm-2) was brushed onto carbon paper. The membrane was sandwiched between the two gas-diffusion electrodes, and hot-pressed into a membrane electrode assembly (MEA) at 4 MPa and 160 °C for 5 min. For the direct methanol single cell system, the test was operated at 60 °C with 2 M methanol aqueous solution as the fuel, the flow rate of hydrated O2 was ~ 100 mL min-1 for the cathode side. 3. Results and Discussion 3.1 Characterization of the GO The GO-his was synthesized as illustrated in Scheme 1. Pristine GO was first esterified to produce more carboxylic acid groups, and then activated by acyl chlorination to obtain GO-COCl. Finally, histidine was covalently bonded to the surface of the above pretreated GO through amidation reaction. TEM images in Fig. 1 show the morphology of GO sheets before and after functionalization. The pristine GO (Fig. 1a) is a single layered sheet with a flat 11

structure. The GO-COCl sheet (Fig. 1b) appears also as an individual sheet but with some wrinkles. After functionalization with histidine, the GO-his sheet (Fig. 1c) displays a more obviously wrinkled surface because of the mutual interaction between functional groups attached to GO. In general, the GO sheets remain the exfoliated layered structure without evident aggregation during the whole functionalization process. For chemical structure analysis, the FTIR spectra of GO-COCl and GO-his are compared with that of pristine GO (Fig. 1d). The pristine GO exhibits three characteristic bands. The strong band at 1726 cm-1 is attributed to the C=O stretching of COOH groups [35], the band at 1330 cm-1 is attributed to the O–H deformation of C–OH groups and the band at 1210 cm-1 corresponds to the asymmetric stretching of C–O–C groups [36]. The results clearly demonstrate that the graphite has been oxidized to GO successfully. After the acyl chlorination of GO, two new bands at ~ 1230 and 1380 cm-1 appear, indicating the formation of –COCl groups [37]. In the case of GO-his, the new bands at 1660 and 3100 cm-1 are assigned to the stretching vibration of C=O and N–H in amide groups, the C=N and C–N stretching vibration in imidazole groups can also be observed at 1420 and 1570 cm-1, respectively [38]. The FTIR spectra confirm the successful functionalization of GO by histidine. The chemical status and composition of elements in pristine GO and GO-his are investigated by XPS. According to the wide range XPS patterns shown in Fig. 1e, a new band can be found for GO-his at the binding energy of 400.0 eV, which is attributed to the N1s component in histidine. Based on the atomic concentration of 12

N1s, the weight fraction of histidine molecules on GO-his is calculated to be ~ 7.8%. The deconvoluted XPS spectra of O1s for pristine GO (Fig. 1f) shows that the O1s spectrum consists of three main peaks at ~ 533.0, 531.7 and 530.5 eV, which can be attributed to the oxygen species of C–O–C, C–OH and C=O groups, respectively [39]. Similarly, three peaks centered at ~ 402.2, 400.6 and 397.0 eV can be used to fit the N1s spectrum of GO-his (Fig. 1g). These three peaks are assigned to the protonated N atom (C–N–C), unprotonated N atom (C=N–C) in imidazole ring and the N atom in amide bonds, respectively. 3.2 Characterization of the membranes The cross-sectional morphologies of plain SPEEK membrane and hybrid membranes are shown in Fig. 2. It can be observed that the plain SPEEK membrane possesses a smooth surface (Fig. 2a), while after the introduction of GO, flake-like fillers are visible in the SPEEK/GO-4 hybrid membrane and a very rugged cross-section occurs (Fig. 2b). The rough morphology of SPEEK/GO-4 is most likely caused by the poor compatibility between pristine GO and SPEEK. Compared with the SPEEK/GO-4 hybrid membrane, the SPEEK/GO-his-4 hybrid membrane displays a relatively smooth morphology (Fig. 2c). The magnified FESEM image of SPEEK/GO-his-4 (Fig. 2d) shows that the GO-his sheets are dispersed homogeneously in polymer matrix. This is because the GO-his contribute to a stronger interfacial interaction with acidic groups in SPEEK compared with the pristine GO, which weakens the surface tension of GO and avoids the micro-phase separation. When the GO-his content is increased up to 6 wt%, a more wrinkled 13

structure can be found due to the stacking and agglomeration of GO-his flakes (Fig. 2e). As shown in Fig. 2f, the SPEEK/GO-his-6 hybrid membrane is formed with some cracks and pinholes along the edges of GO-his sheets. The FTIR spectra shown in Fig. 3a illustrate the interaction between inorganic filler and polymer matrix. All the SPEEK-based membranes show characteristic bands at ~ 1221, 1080 and 1020 cm-1, which are assigned to the symmetric and asymmetric vibrations of –SO3H groups [40]. Due to the electrostatic force generated between GO and SPEEK, the peak intensity of –SO3H groups in hybrid membranes decreases apparently compared with plain SPEEK membrane. The SPEEK/GO-his hybrid membranes exhibit bigger decrease in peak intensity than SPEEK/GO hybrid membranes at the same filler content, indicating that there exist strong acid-base interactions between –SO3H and imidazole groups in the SPEEK/GO-his hybrid membranes [41]. As displayed in Fig. 3b, the plain SPEEK membrane possesses crystalline peaks at 2θ of 17° and 23°, which correspond to the reflections from (1 1 0) and (2 0 0) planes, respectively [42]. After the introduction of inorganic fillers, the intensity of these two characteristic peaks is obviously weakened. When the GO-his loading is over 4 wt%, the crystalline peaks almost disappear. The reduction in intensity is caused by the hydrogen bonds generated between the polar groups in GO sheets (–OH, –O–, C=O) and the sulfonic acid groups in SPEEK, which significantly disturb the originally ordered packing of polymer chains. Fig. 3c illustrates the polymer chain mobility of the membranes investigated by 14

DSC. The glass transition temperature (Tg) value of the plain SPEEK membrane is 142.6 °C, the incorporation of GO increases the Tg of membranes, suggesting that the flexibility of polymer chains is restrained by the inorganic filler. The hybrid membranes with the GO-his content of 1, 2, 4, 6 wt% show gradually enhanced Tg values of 147.0, 151.5, 158.9 and 163.4 °C, respectively. The results show that as the GO-his filler content increases, the polymer chains become stiffer and a higher activation energy is required for the identical motion state. 3.3 Thermal stability of the membranes The thermal stability of the as-obtained membranes, which is important for the membrane lifetime, is determined by TGA and displayed in Fig. 4. Both the hybrid membranes and the plain SPEEK membrane show similar decomposition processes with a three-step weight loss behavior. The first weight loss below 240 °C is due to the evaporation of free water, bound water and residual solvent within the membrane. The second sharp weight loss in the range of 280 – 380 °C is related to the pyrolysis of –SO3H in SPEEK chains. It can be seen that the second degradation stage is inhibited after the incorporation of GO. In addition, the SPEEK/GO-his-2 and SPEEK/GO-his-4 hybrid membranes show a retarded decomposition process compared

with

SPEEK/GO-2

and

SPEEK/GO-4

hybrid

membranes.

The

experimental results mean that the electrostatic interactions generated between GO-his and SPEEK in the form of –S–O-···+H–HN– and –S–O-···+H–N= slow down the pyrolysis of –SO3H groups in SPEEK, leading to an improvement in thermal stability. The third weight loss starting from 400 °C is associated with the 15

decomposition of polymer main chains. It is worth noting that all the hybrid membranes are sufficiently stable below 250 °C, which is adequate for the practical application in DMFCs [43]. 3.4 Mechanical stability of membranes The mechanical properties of as-prepared membranes are evaluated from the stress-strain curves shown in Fig. 5, and the results are displayed in Table 1. The tensile strength of the SPEEK/GO-his hybrid membranes is similar to that of the plain SPEEK and Nafion® 117 membranes, indicating a good mechanical stability. The Young's modulus of plain SPEEK membrane is measured to be 731.1 Mpa, after the introduction of GO-his fillers, the Young’s modulus is increased to 742.7 − 823.5 Mpa. The elongation at break of SPEEK/GO-his hybrid membranes is lower than that of plain SPEEK membrane and decreases with the increase of filler content. The enhanced Young’s modulus and decreased elongation at break are reasonably ascribed to the fact that the incorporation of rigid inorganic fillers restricts the movement of polymer chains. Nevertheless, the elongations of SPEEK/GO-his hybrid membranes with filler contents lower than 6 wt% are still higher than 40%, flexible enough to be used as PEMs in DMFCs [44]. 3.5 Water uptake and swelling degree Water uptake reflects the hydrophilicity and free volume property of the membranes, while excess swelling caused by too much water uptake will lead to the deterioration of the interface between membrane and catalyst. The water uptake and swelling degree values of all the membranes at 25 and 55 °C are shown in Table 2. 16

Generally, both the water uptake and swelling degree are reduced after the GO is filled into the polymer. As the GO-his content increases from 0 wt% to 6 wt%, the water uptake of SPEEK/GO-his hybrid membranes decreases from 12.72% to 8.84% at 25 °C and from 62.83% to 42.28% at 55 °C. The decrease is caused by the fact that the hydrophilicity of functional groups in GO filler (e.g. –COOH, –OH, C–O–C, imidazole groups) is relatively lower compared with the –SO3H groups in SPEEK matrix [45]. The swelling degree is also suppressed when the GO sheets are introduced into SPEEK and the value decreases with the increase of filler content. At the same filler content, the SPEEK/GO-his hybrid membranes show stronger dimensional stability in water compared with SPEEK/GO hybrid membranes. This is because the acidic and basic groups in SPEEK/GO-his hybrid membranes endow it strong electrostatic interaction, decreasing the interchain distance and dimensional size of hydrophilic clusters. At room temperature, both the water uptake and swelling degree of SPEEK/GO-his hybrid membranes are lower than that of the Nafion® 117 membrane. At 55 °C, the water uptake is higher while the swelling degree is lower compared with Nafion® 117 membrane, suggesting an acceptable stability of as-prepared hybrid membranes in water. 3.6 Ion-exchange capacity The ion-exchange capacity (IEC) indicates the density of H+ exchangeable functional groups in the sample. The IEC values of as-prepared membranes are listed in Table 2, the plain SPEEK membrane displays an IEC of ~1.85 mmol g-1, corresponding to a sulfonation degree of 66%. The IEC values of SPEEK/GO and 17

SPEEK/GO-his hybrid systems vary from 1.84 to 1.58 mmol g-1. The IEC drop can be attributed to the fact that the carboxylic acid and imidazole groups in GO and GO-his exhibit lower H+ dissociation ability than sulfonic acid groups. For SPEEK/GO-his system, the acid-base interaction between heterocycles and sulfonic acid groups also restricts the H+ dissociation from the acidic groups, so the exchangeable H+ per unit mass decreases as the GO-his content increases. 3.7 Proton conduction of the membranes The proton conductivities (σ) of all the membranes at 100% RH are presented in Fig. 6. Obviously, the proton conductivities of all the hybrid membranes are higher than the plain SPEEK membrane. The notable increase in proton conductivity demonstrates that the proton-transfer ability of the hybrid membranes are intensified by the functional groups in GO-his and the proton conduction process is thermally activated. Varying the loading amount of GO-his from 1 to 4 wt% results in a positive effect on the proton conductivity of hybrid membranes. The SPEEK/GO-his-4 membrane exhibits proton conductivity values of 6.94 × 10-2 S cm-1 at 25 °C and 0.29 S cm-1 at 75 °C, which are 30% and 35% higher than the plain SPEEK membrane under the same conditions. These proton conductivity values are comparable to the Nafion® 117 membrane. However, when the filler loading is higher than 4 wt%, aggregation effect of GO-his emerges, which blocks the proton channels within hybrid membranes and reduces the proton conductivity. For comparison, the proton conductivity values of SPEEK/GO hybrid membranes are also measured. The SPEEK/GO-2 membrane shows a lower proton conductivity than plain SPEEK 18

membrane because the proton mobility in pristine GO is relatively much lower than that in SPEEK matrix. At the GO content of 4 wt%, an increase in proton conductivity can be found compared with SPEEK/GO-2, which is probably due to that the C–O–C and –COOH groups on GO nanosheets transfer protons more efficiently. The activation energy values (Ea) for proton conduction can be estimated from the Arrhenius equation,

   0e



Ea RT

(9)

where σ0 is the pre-exponential factor, R is the gas constant and T is the Kelvin temperature. The Ea values of all the prepared membranes are in the range of 26 – 31 kJ mol-1, demonstrating that both Grotthuss mechanism and Vehicular mechanism exist in these membranes, whereas the former is predominant [34]. The possible proton-transfer mechanism inside the hybrid membranes is depicted in Fig. 7 and interpreted as follows. (i) After GO-his sheets are introduced into the SPEEK matrix, the acidic –SO3H groups and basic imidazole groups are linked by intermediate water bridges to form loose acid-base complexes (the number of water molecules forming a water bridge is usually below three [46]). The proton first interacts with acidic –SO3H group through hydrogen bonds, transfers along a cluster of hydrogen-bonded H2O molecules to arrive at the basic imidazole group. Then the proton triggers the protonation of imidazole and finally moves out of the imidazolium through hydrogen bonds with another H2O molecule. In this manner, the protons will be transferred continuously by acid-base pairs via Grotthuss mechanism, in which the hydrogen-bonding network is formed and broken alternately. (ii) Some protons 19

interact with the free water molecules in the membrane to generate hydronium ions which further facilitate the diffusion of protons through continuous nanochannels via Vehicular mechanism. To evaluate the potential performance in fuel cell application, the proton conductivities of membranes as a function of humidity at 65 °C are also measured as shown in Fig. 8. For plain SPEEK control membrane, the proton conductivity declines sharply from 0.183 (RH = 100%) to 1.05 × 10-4 (RH = 26%) S cm-1 with a reduction of ~ 99.94%. This remarkable decrease is mainly caused by the serious water loss in SPEEK membrane, which reduces the proton transfer via Vehicular mechanism. Nevertheless, the SPEEK/GO-his hybrid membranes display higher proton conductivity values compared with the plain SPEEK membrane under the same RH and the enhancement is more evident when the RH is decreased. At the RH of 26%, SPEEK/GO-his-2, SPEEK/GO-his-4, and SPEEK/GO-his-6 exhibit the proton conductivity of 1.82 × 10-3, 4.71 × 10-3 and 5.57 × 10-4 S cm-1, which is 17.3, 44.9 and 5.3 times higher than that of the plain SPEEK, respectively. The significant improvement in proton conductivity can be interpreted by the fabrication of acid-base proton channels within the hybrid membranes, which enhance the Grotthuss-type proton conduction. 3.8 Methanol permeability and selectivity In DMFCs, methanol permeation through the PEMs would cause the fuel loss, electro catalyst poisoning and reduced open-circuit potential, so methanol crossover is one of the most important problems that need to be solved. Table 3 shows the proton 20

conductivity (σ, at 100% RH), methanol permeability (P, in 2M methanol solution) and selectivity (β) of the as-prepared membranes at 25 °C. Compared with the plain SPEEK membrane (5.07 × 10-7 cm2 s-1), the methanol permeability values are significantly reduced after the introduction of GO sheets (1.32 – 3.91 × 10-7 cm2 s-1). The methanol permeability of SPEEK/GO-his membranes is ~ an order of magnitude lower than that of the Nafion® 117 membrane. Fig. 9 shows the methanol concentration dependence of methanol permeability for all the membranes at 25 °C. With an increase in methanol concentration from 1 M up to 10 M, the methanol permeability for plain SPEEK membrane increases from 4.86 × 10−7 up to 5.53 × 10−7 cm2 s-1, indicating that the methanol permeability in such a PEM is controlled by methanol concentration. While for hybrid membranes, the methanol permeability keeps constant or even decreases slightly with increasing methanol concentration, this result is consistent with the trend found in literatures [47, 48]. All the results demonstrate that the fabricated hybrid membrane is a good methanol barrier. The free volume characteristics of membranes refer to the static voids generated by chain packing and the transient gaps created by thermally induced chain rearrangement [49, 50]. Since the methanol passes through the membrane via the solution-diffusion mechanism, the amount of free volume in membranes determines the diffusion of methanol molecules. The free volume parameters of as-prepared membranes are measured by PALS and shown in Table 4. The fractional free volume of hybrid membranes decreases from 0.42 to 0.32% as the GO-his content increases from 0 to 6 wt%. The decrease of fractional free volume is fairly beneficial to 21

suppress the methanol crossover through the membranes. Besides the reduction of hydrophilic clusters, it has been demonstrated in literatures that the physical property and defects of GO sheets play major roles in determining the mass transport of membranes. At a constant oxidation level, methanol permeability decreases linearly with increasing the GO mean flake size [51]. The bond length between carbon atoms (dC–C) in graphene is 0.142 nm, implying that the pore size of each hexagonal lattice is 0.246 nm considering the nuclei alone, this size is much smaller than the dynamic diameter of methanol molecule (0.38 nm), so the existence of GO sheets blocks the diffusion way of methanol molecules to some extent. In conclusion, the synergistic effect of acid-base pairs formed between SPEEK and GO-his enhances the proton conductivity, meanwhile, the block effect and channel size reduction after the incorporation of GO-his sheets are in favor of methanol resistance, thus resulting in a high selectivity. Among all the hybrid membranes, SPEEK/GO-his-4 shows the highest selectivity of 5.14 × 105 S s cm-3, which is ~ 5 times higher than that of the plain SPEEK membrane (1.05 × 105 S s cm-3). Additionally, the selectivities of as-prepared hybrid membranes are much higher than that of the Nafion® 117 membrane. 3.9 Single DMFC performance test Considering its relatively good comprehensive performance, SPEEK/GO-his-4 is utilized to prepare MEA with the SPEEK control membrane as comparison. The single cell performance is tested with methanol/O2 at 60 °C and the polarization and power density curves are shown in Fig. 10. The open-cell voltage (OCV) of the cell 22

made with SPEEK/GO-his-4 hybrid membrane (0.69 V) is slightly higher than that of the cell with the plain SPEEK membrane (0.65 V), the lower OCV is probably caused by higher methanol crossover. The peak power density of the cell with the SPEEK/GO-his-4 membrane is 43.0 mW cm-2, which is 80.7% higher than the cell with plain SPEEK membrane. Additionally, the maximum current density of SPEEK/GO-his-4 membrane is also 52.6% higher than that of SPEEK membrane. The test results suggest that the incorporation of histidine functionalized GO nanosheets can substantially improve the fuel cell performance.

4. Conclusions Based on the discovery that acid-base pairs can construct efficient proton transfer sites, we synthesized the histidine-functionalized GO sheets and introduced them into SPEEK polymer matrix to design new polymer electrolyte membranes with continuous pathways for proton conduction. The acid-base pairs formed by imidazole groups in histidine and –SO3H groups in SPEEK endow the membranes high proton conductivity. The reduction of fractional free volume caused by acid-base interaction and the blocking effect of GO-his flakes offer the membrane a strong resistance to methanol crossover. As a consequence, the SPEEK/GO-his-4 hybrid membrane shows a highest selectivity of 5.14×105 S s cm-3. The power density of the cell with the SPEEK/GO-his-4 membrane reaches 43.0 mW cm-2, which is 80.7% higher than the cell with plain SPEEK membrane. This study provides a new and facile method towards achieving both high proton conductivity and low methanol permeability for 23

DMFC membranes.

Acknowledgement The authors gratefully acknowledge financial support from Program for from National Natural Science Foundation of China (21576189), State Key Laboratory of Separation Membranes and Membrane Processes (Tianjin Polytechnic University) (M2-201504) and the Programme of Introducing Talents of Discipline to Universities (No. B06006).

References [1]

C. Koenigsmann, S.S. Wong, One-dimensional noble metal electrocatalysts: a promising structural paradigm for direct methanol fuel cells, Energy Environ. Sci. 4 (2011) 1161.

[2]

J. Wang, X. Yue, Z. Zhang, Z. Yang, Y. Li, H. Zhang, X. Yang, H. Wu, Z. jiang, Enhancement of proton conduction at low humidity by incorporating imidazole microcapsules into polymer electrolyte membranes, Adv. Funct. Mater. 22 (2012) 4539.

[3]

H. Zhang, P.K. Shen, Advances in the high performance polymer electrolyte membranes for fuel cells, Chem. Soc. Rev. 2012, 41 (2012) 2382.

[4]

L.M. Robeson, The upper bound revisited, J. Membr. Sci. 320 (2008) 390. 24

[5]

S. Jessie Lue, S.H. Peng, Polyurethane (PU) membrane preparation with and without hydroxypropyl-β-cyclodextrin and theirpervaporation characteristics, J. Membr. Sci. 222 (2003) 203.

[6]

C.H. Lee, H.B. Park, Y.S. Chung, Y.M. Lee, B.D. Freeman, Water Sorption, proton conduction, and methanol permeation properties of sulfonated polyimide

membranes

cross-linked

with

N,N-Bis(2-hydroxyethyl)-2-aminoethanesulfonic acid (BES), Macromolecules 39 (2006) 755. [7]

C. Laberty-Robert, K. Valle, F. Pereira, C. Sanchez, Design and properties of functional hybrid organic–inorganic membranes for fuel cells, Chem. Soc. Rev. 40 (2011) 961.

[8]

D. Kim, H. Park, J. Rhim, Y. Lee, Proton conductivity and methanol transport behavior of cross-linked PVA/PAA/silica hybrid membranes, Solid State Ion. 176 (2005) 117.

[9]

C.Y. Tseng, Y.S. Ye, M.Y. Cheng, K.Y. Kao, W.C. Shen, J. Rick, J.C. Chen, B.J. Hwang, Sulfonated polyimide proton exchange membranes with graphene oxide show improved proton conductivity, methanol crossover impedance, and mechanical properties, Adv. Energy Mater. 1 (2011) 1220.

[10]

H. Tateishi, K. Hatakeyama, C. Ogata, K. Gezuhara, J. Kuroda, A. Funatsu, M. Koinuma, T. Taniguchi, S. Hayami, Y. Matsumoto, Graphene oxide fuel cell, J. Electrochem. Soc. 160 (2013) F1175.

[11]

M.R. Karim, K. Hatakeyama, T. Matsui, H. Takehira, T. Taniguchi, M. 25

Koinuma, Y. Matsumoto, T. Akutagawa, T. Nakamura, S. Noro, T. Yamada, H. Kitagawa, S. Hayami, Graphene oxide nanosheet with high proton conductivity, J. Am. Chem. Soc. 135 (2013) 8097. [12]

W. Gao, G. Wu, M.T. Janicke, D.A. Cullen, R. Mukundan, J.K. Baldwin, E.L. Brosha, C. Galande, P.M. Ajayan, K.L. More, A.M. Dattelbaum, P. Zelenay, Ozonated graphene oxide film as a proton-exchange membrane, Angew. Chem. Int. Ed. 53 (2014) 3588.

[13]

K. Hatakeyama, M.R. Karim, C. Ogata, H. Tateishi, A. Funatsu, T. Taniguchi, M. Koinuma, S. Hayami, Y. Matsumoto, Proton conductivities of graphene oxide nanosheets: single, multilayer, and modified nanosheets, Angew. Chem. Int. Ed. 53 (2014) 6997.

[14]

T. Yuan, L. Pu, Q. Huang, H. Zhang, X. Li, H. Yang, An effective methanol-blocking membrane modified with graphene oxide nanosheets for passive direct methanol fuel cells, Electrochim. Acta 117 (2014) 393.

[15]

B.G. Choi, J. Hong, Y.C. Park, D.H. Jung, W.H. Hong, P.T. Hammond, H. Park, Innovative polymer nanocomposite electrolytes: nanoscale manipulation of ion channels by functionalized graphenes, ACS nano 5 (2011) 5167.

[16]

S. Gahlot, P.P. Sharma, V. Kulshrestha, P.K. Jha, SGO/SPES-based highly conducting polymer electrolyte membranes for fuel cell application, ACS Appl. Mater. Inter. 6 (2014) 5595.

[17]

Z. Jiang, X. Zhao, Y. Fu, A. Manthiram, Composite membranes based on sulfonated poly(ether ether ketone) and SDBS-adsorbed graphene oxide for 26

direct methanol fuel cells, J. Mater. Chem. 22 (2012) 24862. [18]

H. Zarrin, D. Higgins, Y. Jun, Z. Chen, M. Fowler, Functionalized graphene oxide nanocomposite membrane for low humidity and high temperature proton exchange membrane fuel cells, J. Phys. Chem. C 115 (2011) 20774.

[19]

J. Zhao, X. Zhao, Z. Jiang, Z. Li, X. Fan, J. Zhu, H. Wu, Y. Su, D. Yang, F. Pan, J. Shi, Biomimetic

and

bioinspired

membranes:

preparation and

application, Prog. Polym. Sci. 39 (2014) 1668. [20]

U.G.K. Wegst, H. Bai, E. Saiz, A.P. Tomsia, R. Ritchie, Bioinspired structural materials, Nat. Mater. 14 (2015) 23.

[21]

X. Hou, L. Jiang, Learning from nature: building bio-inspired smart nanochannels, ACS nano 3 (2009) 3339.

[22]

E.R. Brown, S. Piscopo, Ion channels in key marine invertebrates; their diversity and potential for applications in biotechnology, Biotechnol. Adv. 29 (2011) 457.

[23]

S. Majd, E.C. Yusko, Y.N. Billeh, M.X. Macrae, J. Yang, M. Mayer, Applications

of

biological

pores

in

nanomedicine,

sensing,

and

nanoelectronics, Curr. Opin. Biotech. 21 (2010) 439. [24]

H.J. Leem, I. Dorbandt, J. Rojas-Chapana, S. Fiechter, H. Tributsch, Bio-analogue amino acid-based proton-conduction wires for fuel cell membranes, J. Phys. Chem. C 112 (2008) 2756.

[25]

M. Yamada, I. Honma, A biopolymer composite material as an anhydrous proton-conducting membrane, Angew. Chem. Int. Ed. 43 (2004) 3688. 27

[26]

S.S. Nagarkar, S.M. Unni, A. Sharma, S. Kurungot, S.K. Ghosh, Two-in-one: inherent anhydrous and water-assisted high proton conduction in a 3d metal– organic framework, Angew. Chem. Int. Ed. 53 (2014) 2638.

[27]

Y. Liu, J. Wang, H. Zhang, C. Ma, J. Liu, S. Cao, X. Zhang, Enhancement of proton conductivity of chitosan membrane enabled by sulfonated graphene oxide under both hydrated and anhydrous conditions, J. Power Sources 269 (2014) 898.

[28]

M.M. Hasani-Sadrabadi, E. Dashtimoghadam, F.S. Majedi, H. Moaddel, A. Bertsch, P. Renaud, Superacid-doped polybenzimidazole-decorated carbon nanotubes: a novel high-performance proton exchange nanocomposite membrane, Nanoscale 5 (2013) 11710.

[29]

J. Wang, H. Bai, H. Zhang, L. Zhao, H. Chen, Y. Li, Anhydrous proton exchange membrane of sulfonated poly(ether ether ketone) enabled by polydopamine-modified silica nanoparticles, Electrochim. Acta 152 (2015) 443.

[30]

Y. Yin, T, Xu, X. Shen, H. Wu, Z. Jiang, Fabrication of chitosan/zwitterion functionalized titania–silica hybrid membranes with improved proton conductivity, J. Membr. Sci. 469 (2014) 355.

[31]

W. Lv, D.M. Tang, Y.B. He, C.H. You, Z.Q. Shi, X.C. Chen, C.M. Chen, P.X. Hou,

C.

Liu,

Q.H.

Yang,

Low-temperature

exfoliated

graphenes:

vacuum-promoted exfoliation and electrochemical energy storage, ACS nano 3 (2009) 3730. 28

[32]

Y. Liu, R. Deng, Z. Wang, H. Liu, Carboxyl-functionalized graphene oxide– polyaniline composite as a promising supercapacitor material, J. Mater. Chem. 22 (2012) 13619.

[33]

Y. Yin, T. Xu, G. He, Z. Jiang, H. Wu, Fabrication of sulfonated poly(ether ether ketone)-based hybrid proton-conducting membranes containing carboxyl or amino acid-functionalized titania by in situ sol–gel process, J. Power Sources 276 (2015) 271.

[34]

H. Wu, X. Shen, T. Xu, W. Hou, Z. Jiang, Sulfonated poly(ether ether ketone)/amino-acid

functionalized

titania

hybrid

proton

conductive

membranes, J. Power Sources 213 (2012) 83. [35]

C.W. Lin, Y.S. Lu, Highly ordered graphene oxide paper laminated with a Nafion membrane for direct methanol fuel cells, J. Power Sources 237 (2013) 187-194.

[36]

A. Enotiadis, K. Angjeli, N. Baldino, I. Nicotera, D. Gournis, Graphene-based nafion nanocomposite membranes: enhanced proton transport and water retention by novel organo-functionalized graphene oxide nanosheets, Small 8 (2012) 3338.

[37]

J. Chu, X. Li, P. Xu, Fluorescent features of CdTe nanorods grafted to graphene oxide through an amidation process, J. Mater. Chem. 21 (2011) 11283.

[38]

M.S. Boroglu, S.U. Celik, A. Bozkurt, I. Boz, The synthesis and characterization of anhydrous proton conducting membranes based on 29

sulfonated poly(vinyl alcohol) and imidazole, J. Membr. Sci. 375 (2011) 157-164. [39]

L. Zhang, L. Ji, P.A. Glans, Y. Zhang, J. Zhu, J. Guo, Electronic structure and chemical bonding of a graphene oxide–sulfur nanocomposite for use in superior performance lithium–sulfur cells, Phys. Chem. Chem. Phys. 14 (2012) 13670.

[40]

H. Wu, X. Shen, Y. Cao, Z. Li, Z. Jiang, Composite proton conductive membranes

composed

of

sulfonated

poly(ether

ether

ketone)

and

phosphotungstic acid-loaded imidazole microcapsules as acid reservoirs, J. Membr. Sci. 451 (2014) 74. [41]

Y. Yin, J. Wang, S. Jiang, X. Yang, X. Zhang, Y. Cao, L. Cao, H. Wu, Novel composite membranes based on sulfonated poly(ether ether ketone) and adenosine triphosphate for enhanced proton conduction, Rsc Adv. 5 (2015) 75434-75441.

[42]

X. Jin, M. T. Bishop, T.S. Ellis, F. E. Karasz, A sulphonated poly(aryl ether ketone), Brit. Polym. J. 1985, 17, 4.

[43]

S. Feng, Y. Shang, G. Liu, W. Dong, X. Xie, J. Xu, V.K. Mathur, Novel modification method to prepare crosslinked sulfonated poly(ether ether ketone)/silica hybrid membranes for fuel cells, J. Power Sources 195 (2010) 6450-6458.

[44]

H. Wu, Y. Cao, Z. Li, G. He, Z. Jiang, Novel sulfonated poly (ether ether ketone)/phosphonic acid-functionalized titania nanohybrid membrane by an in 30

situ method for direct methanol fuel cells, J. Power Sources 273 (2015) 544-553. [45]

S.J. Paddison, K.D. Kreuer, J. Maier, About the choice of the protogenic group in polymer electrolyte membranes: Ab initio modelling of sulfonic acid, phosphonic acid, and imidazole functionalized alkanes, Phys. Chem. Chem. Phys. 8 (2006) 4530.

[46]

L. Wu, C. Huang, J.J. Woo, D. Wu, S.H. Yun, S. J. Seo, T. Xu, S.H. Moon, Hydrogen bonding: a channel for protons to transfer through acid-base pairs, J. Phys. Chem. B 113 (2009) 12265.

[47]

M.H. Yildirim, A. Schwarz, D.F. Stamatialis, M. Wessling, Impregnated membranes for direct methanol fuel cells at high methanol concentrations, J. Membr. Sci. 328 (2009) 127-133.

[48]

K.-i. Okamoto, Y. Yin, O. Yamada, M.N. Islam, T. Honda, T. Mishima, Y. Suto, K. Tanaka, H. Kita, Methanol permeability and proton conductivity of sulfonated co-polyimide membranes, J. Membr. Sci. 258 (2005) 115-122.

[49]

T.C. Merkel, B.D. Freeman, R.J. Spontak, Z. He, I. Pinnau, P. Meakin, A.J. Hill, Ultrapermeable, reverse-selective nanocomposite membranes, Science 296 (2002) 519-522.

[50]

V.P. Shantarovich, I.B. Kevdina, Y.P. Yampolskii, A.Y. Alentiev, Positron annihilation lifetime study of high and low free volume glassy polymers: Effects of free volume sizes on the permeability and permselectivity, Macromolecules 33 (2000) 7453-7466. 31

[51]

A. Paneri, S. Moghaddam, Impact of synthesis conditions on physicochemical and transport characteristics of graphene oxide laminates, Carbon 86 (2015) 245.

32

e)

d)

Atomic Conc (%) N 1s 1.9

pristine GO

C 1s

Intensity (a.u.)

Transmittance

O 1s

GO-COCl

N 1s

GO-his GO O 1s

C 1s

GO-his

3500

3000

2500

2000

1500

1000

500

800

Wavenumber (cm-1)

700

600

500

400

300

200

Binding energy (eV)

f)

g)

pristine GO

GO-his N 1s

O 1s C=N-C C-O-C

Intensity (a.u.)

Intensity (a.u.)

C-OH

C=O

538

536

534

532

530

528

C-N-C CO-NH

406

Binding energy (eV)

404

402

400

398

396

Binding energy (eV)

Fig. 1. TEM images of a) pristine GO, b) GO-COCl and c) GO-his. d) FTIR spectra of pristine GO, GO-COCl and GO-his. e) Wide range XPS patterns of pristine GO and GO-his. f) Deconvoluted XPS spectra in the O1s region for pristine GO. g) Deconvoluted XPS spectra in the N1s region for GO-his.

33

Fig. 2. FESEM images of the cross-section of a) plain SPEEK, b) SPEEK/GO-4, c), d) SPEEK/GO-his-4 and e), f) SPEEK/GO-his-6 hybrid membranes.

34

a)

SPEEK SPEEK/GO-2

Transmittance

SPEEK/GO-4

SPEEK/GO-his-1 SPEEK/GO-his-2 SPEEK/GO-his-4 SPEEK/GO-his-6

1221 2000

1800

1600

1400

1080 1200

1020 1000

800

Wavenumber (cm-1)

b)

Intensity

SPEEK SPEEK/GO-2 SPEEK/GO-4 SPEEK/GO-his-1 SPEEK/GO-his-2 SPEEK/GO-his-4 SPEEK/GO-his-6 10

20

30

40

2θ (degrees)

50

c) 142.6 SPEEK

151.0

Heat flow

SPEEK/GO-2

156.9

SPEEK/GO-4 147.0 SPEEK/GO-his-1 151.5 SPEEK/GO-his-2

158.9

SPEEK/GO-his-4

163.4

SPEEK/GO-his-6 100

120

140

160

180

Temperature (°C)

Fig. 3. a) FTIR, b) XRD and c) DSC of plain SPEEK, SPEEK/GO and SPEEK/GO-his hybrid membranes.

35

SPEEK SPEEK/GO-2 SPEEK/GO-4 SPEEK/GO-his-1 SPEEK/GO-his-2 SPEEK/GO-his-4 SPEEK/GO-his-6

100

90

90

SPEEK/GO-his-4

70

60

Weight (%)

Weight (%)

80

85

SPEEK/GO-4 SPEEK/GO-his-6 SPEEK/GO-his-1

80

SPEEK/GO-his-2

50 SPEEK/GO-2

75

SPEEK

40

300

320

340

360

380

400

420

Temperature (°C)

100

200

300

400

500

600

700

800

Temperature (°C)

Fig. 4. TGA curves of plain SPEEK membrane, SPEEK/GO hybrid membranes and SPEEK/GO-his hybrid membranes.

36

50

Stress (Mpa)

40

30

a

e 20

a---SPEEK b---SPEEK/GO-4 c---SPEEK/GO-his-2 d---SPEEK/GO-his-4 e---SPEEK/GO-his-6

b 10

0 0

10

c

d

20

30

40

50

60

Strain (%)

Fig. 5. Stress−strain curves for plain SPEEK membrane, SPEEK/GO hybrid membranes and SPEEK/GO-his hybrid membranes at room temperature.

37

0.30

SPEEK SPEEK/GO-2 SPEEK/GO-4 SPEEK/GO-his-1 SPEEK/GO-his-2 SPEEK/GO-his-4 SPEEK/GO-his-6

Proton Conductivity (S cm-1)

0.25

0.20

0.15

0.10

0.05 20

30

40

50

60

70

80

Temperature (°C)

Fig. 6. Proton conductivity of plain SPEEK membrane, SPEEK/GO hybrid membranes and SPEEK/GO-his hybrid membranes at 100% RH as a function of temperature.

38

Fig. 7. Proposed mechanism for proton conduction in the SPEEK/GO-his hybrid membranes.

39

Proton Conductivity (S cm-1)

1

SPEEK SPEEK/GO-4 SPEEK/GO-his-2 SPEEK/GO-his-4 SPEEK/GO-his-6

0.1

0.01

1E-3

1E-4 20

30

40

50

60

70

80

90

100

Relative Humidity (%)

Fig. 8. Proton conductivity of plain SPEEK membrane, SPEEK/GO hybrid membranes and SPEEK/GO-his hybrid membranes as a function of humidity at 65 °C.

40

SPEEK Methanol permeability (×10-7 cm2 s-1)

5

4

3

SPEEK/GO-his-1

SPEEK/GO-2 SPEEK/GO-his-2

2

SPEEK/GO-4 SPEEK/GO-his-4

1

SPEEK/GO-his-6 2

4

6

8

10

Methanol concentration (M)

Fig. 9. Methanol permeability of plain SPEEK membrane, SPEEK/GO hybrid membranes and SPEEK/GO-his hybrid membranes as a function of methanol concentration at 25 °C.

41

50 0.7

SPEEK SPEEK/GO-his-4 40

0.6

Voltage (V)

30 0.4 20 0.3

0.2

Power Density (mW cm-2)

0.5

10

0.1 0 0.00

0.05

0.10

0.15

0.20

0.25

-2

Current Density (A cm )

Fig. 10. Single DMFC performance of the plain SPEEK and SPEEK/GO-his-4 membranes operated at 60 °C and fed with 2 M methanol.

42

Scheme 1. Schematic representation of the functionalization process of graphene oxide.

43

Table 1. Mechanical properties of as-prepared membranes and Nafion® 117 as a comparison. Tensile strength (MPa)

Young's modulus (MPa)

Elongation at break (%)

SPEEK

41.5

731.1

52.4

SPEEK/GO-4

33.9

865.5

5.4

SPEEK/GO-his-2

37.2

742.7

45.2

SPEEK/GO-his-4

38.8

792.0

42.2

SPEEK/GO-his-6

44.4

823.5

24.6

Nafion® 117

37.4

185.0

297.6

Membrane

Table 2. Water uptake, swelling degree and ion-exchange capacity (IEC) of as-prepared membranes and Nafion® 117 as a comparison. Water uptake (%) Swelling degree (%) IEC Membrane (mmol g-1) 25 °C 55 °C 25 °C 55 °C SPEEK

12.72

62.83

5.81

44.73

1.85

SPEEK/GO-2

12.02

54.55

5.32

38.74

1.69

SPEEK/GO-4

10.13

50.89

4.06

34.99

1.61

SPEEK/GO-his-1

15.29

55.38

7.03

36.79

1.84

SPEEK/GO-his-2

9.70

48.31

3.72

34.90

1.76

SPEEK/GO-his-4

9.47

46.90

2.73

31.43

1.62

SPEEK/GO-his-6

8.84

42.28

3.02

28.30

1.58

Nafion® 117

19.2

31.9

13.1

47.4

0.90

44

Table 3. Proton conductivity (σ), methanol permeability (P) and selectivity (β) of as-prepared membranes and Nafion® 117 as a comparison at 25 °C. σ (×10-2 S cm-1) P (×10-7 cm2 s-1) β Membrane 5 100% RH (×10 S s cm-3) 2 M methanol SPEEK

5.33

5.07

1.05

SPEEK/GO-2

3.52

2.74

1.28

SPEEK/GO-4

5.37

1.92

2.80

SPEEK/GO-his-1

6.34

3.91

1.62

SPEEK/GO-his-2

6.54

2.34

2.79

SPEEK/GO-his-4

6.94

1.35

5.14

SPEEK/GO-his-6

5.37

1.32

4.07

Nafion® 117

5.70

29.40

0.19

Table 4. Free volume parameters of plain SPEEK and hybrid membranes. Membrane

τ3 (ns)

I3 (%)

r3 (nm)

Vf (nm3)

FFV (%)

SPEEK

2.026

4.220

0.2872

0.0992

0.4185

SPEEK/GO-4

2.043

4.150

0.2888

0.1008

0.4185

SPEEK/GO-his-2

2.113

3.140

0.2950

0.1075

0.3375

SPEEK/GO-his-4

2.144

3.221

0.2977

0.1105

0.3558

SPEEK/GO-his-6

2.157

2.880

0.2988

0.1117

0.3217

45