anion-conductive multiblock aromatic copolymer membranes...

Anion-Conductive Multiblock Aromatic Copolymer Membranes:Structure−Property RelationshipsDoh-Yeon Park,† Paul A. Kohl,‡ and Haskell W. Beckham*,†

†School of Materials Science and Engineering and ‡School of Chemical and Biomolecular Engineering, Georgia Institute ofTechnology, Atlanta, Georgia 30332, United States

ABSTRACT: Anion-conductive multiblock copoly(arylene ether sulfone)s(mPES) were synthesized with different block lengths and ion-exchangecapacities (IEC) to maximize ion conductivity and explore the relationshipbetween chemical structure and morphology in anion-exchange membranes(AEM). Nuclear magnetic resonance (NMR) relaxometry was used to probewater mobility and domain size. The multiblock copolymers were synthesized bypolycondensation of separately prepared hydroxy-terminated oligomers withfluoro-terminated oligomers. The polymers were made ion-conductive throughselective chloromethylation of one of the two block types, followed byquaternization and hydroxide ion exchange. The resulting block structure, inwhich one type is hydrophilic and one type is hydrophobic, was designed toensure a nanophase-separated morphology. The multiblock copolymers exhibitedhigher anion conductivity than their random copolymer counterparts at the sameIEC. The multiblock copolymer that exhibited the highest anion conductivity was not the one with the highest IEC, but it wasthe one with the shortest NMR relaxation times for water, indicating a greater fraction of water interacts with the ionic polymersegments. Higher IEC values led to increased water absorption, an effective dilution of the ionic groups, and increased NMRrelaxation times for water.

1. INTRODUCTION

Fuel cells directly convert chemical energy to electricity andcould potentially reduce pollution in transportation and otherareas.1 The proton-exchange membrane fuel cell (PEMFC) hasbeen extensively investigated as a power source in electricvehicles and stationary applications.2,3 In PEMFCs, Nafion,developed by DuPont in 1962, is the preferred polymerelectrolyte.4 Nafion is a perfluorosulfonated copolymer thatexhibits high ionic conductivity and dimensional stability.Nevertheless, commercialization of the PEMFC has beenhampered by intrinsic drawbacks such as the use of expensivemetal catalysts, limited lifetime due to degradation, complexwater management schemes, the use of poison-vulnerablecatalysts, and the slow rate of oxygen reduction under acidicconditions. These problems all originate from the use of protonsas conductive species.Anion-exchange membrane fuel cells (AEMFCs) have been

investigated as an alternative to PEMFCs because manydeficiencies can be addressed:5−9 (i) high pH provides kineticallyfavored oxidation of fuels, (ii) AEMFCs can operate withplatinum-free catalysts, (iii) new routes are available foraddressing water management, and (iv) the electro-osmoticdrag of water from the air cathode to the fuel anode lowers fuelcrossover. Despite these benefits, the AEMFC suffers from thelow diffusion coefficient of hydroxide ions.8 A membrane withhigh ionic conductivity may be obtained by the formation ofnanochannels to facilitate ion transport. Numerous studies haveattributed Nafion’s high conductivity to nanochannels created

through nanophase separation between a hydrophobic matrixand hydrophilic side groups.4,10

Attempts have been made to create proton conductivity inmaterials with Nafion-like morphology by incorporatingmonomers with perfluoroacid groups.11,12 Example alternativematerials also include sulfonated poly(ether sulfone),13 hydro-philic−hydrophobic multiblock copolymers based on poly-(arylene ether sulfone),14−16 and aromatic comb-shapedcopolymers with highly sulfonated side chains.17 Attempts havealso been made to create materials with hydroxide conductivityby use of sequential hydrophobic and hydrophilic blocks.Watanabe and co-workers18 prepared anion-conductive multi-block copoly(arylene ether)s that had higher hydroxide ionconductivity than their corresponding random copolymers. Theyattributed the enhanced conductivity of the multiblockcopolymers to their well-developed phase-separated morphol-ogy. Although the conductivity of AEM multiblock copolymershas been reported, the systematic study of water mobility anddomain size in these types of polymer membranes has not beenstudied. Understanding the relationships between polymerstructure, morphology, and ion conductivity is important tofind predictive pathways for improving conductivity.In this study, a series of anion-conductive multiblock

copoly(arylene ether sulfone)s (mPES) containing quaternaryammonium groups19 were synthesized with different lengths of

Received: December 5, 2012Revised: June 17, 2013Published: June 19, 2013

Article

pubs.acs.org/JPCC

© 2013 American Chemical Society 15468 dx.doi.org/10.1021/jp311987v | J. Phys. Chem. C 2013, 117, 15468−15477

pubs.acs.org/JPCC

hydrophobic and hydrophilic blocks to enable nanophase-separated morphologies. The corresponding random copoly-mers were also synthesized for comparison of the properties. Theprecise chemical structures of the copolymers were characterizedby one- and two-dimensional (1D and 2D) NMR. The phase-separated morphologies of the copolymers were probed bymeasuringNMR spin−lattice (T1) and spin−spin (T2) relaxationtimes of water absorbed into the anion-conductive polymericmatrices. Compared to atomic force (AFM) and transmissionelectron microscopy (TEM), NMR relaxation time measure-ments can be used to obtain quantitative information on watermobilities and domain sizes without laborious sample prepara-tion. The water mobility and domain size information extractedfrom the NMR relaxation times of water in the membranes wasthen correlated with ionic conductivity.

2. EXPERIMENTAL DETAILS2.1. Materials. 4,4′-Difluorodiphenyl sulfone (FPS), bi-

sphenol A (BPA), 1,1,2,2-tetrachloro ethane (TCE), trimethylamine aqueous solution (50 wt %), and calcium hydride wereobtained from TCI Co., Ltd. N,N-dimethylacetamide (DMAc),4,4′-(hexafluoroisopropylidene) diphenol (HFBPA), and tin-(IV) chloride were obtained from Alfa Aesar. Chloromethylmethyl ether (CMME), potassium carbonate, and dimethylsulfoxide-d6 (DMSO-d6) were obtained from Aldrich ChemicalCo. Chloroform-d was obtained from Cambridge Isotopes Inc.for the NMR studies. All chemicals were used as received, unlessotherwise specified.2.2. Polymerization. A typical synthetic procedure for the

hydroxy-terminated oligomer (for number of repeat units x =6.7) is as follows. DMAc was dried by distillation in vacuo at 130°C over CaH2 and stored with activated 3Amolecular sieves. FPS(1.63 g, 6.40 mmol), HFBPA (2.37 g, 7.04 mmol), K2CO3 (2.21g, 16.00 mmol), and DMAc (20 mL) were mixed under drynitrogen in a dry 100-mL two-neck round-bottomed flaskequipped with a condenser at room temperature for 10 min. Theresulting mixture was heated to 120 °C by use of an oil bath for3.5 h. HFBPA (0.79 g) was added to the mixture and allowed toreact for 1 h at 120 °C to ensure that the ends were hydroxyl-terminated. The slightly viscous mixture was poured into hotwater to precipitate the product. The reddish powder was washedwith water and methanol several times. The product wascollected, isolated by filtration, and dried at 80 °C for 24 h. Thenumber of repeat units (x = 6.7) and molecular weight (MW =4040 g/mol) were determined by 1H NMR.A similar synthesis, purification, drying, and 1H NMR

characterization procedure was followed to prepare a fluoro-terminated oligomer (number of repeat units y = 7.7). FPS (1.79g, 7.04 mmol), BPA (1.46 g, 6.40 mmol), and K2CO3 (1.86 g,12.80mmol) were dissolved in 20mL of dry DMAc and stirred at110 °C for 7 h. After addition of 0.27 g of FPS, the reaction wasallowed to continue at 110 °C for 1 h, which finally yielded awhite powder (y = 7.7, MW = 3660 g/mol).Copolymerization of the two oligomers was carried out to

yield the multiblock copolymer mPES-X6.7Y7.7. The hydroxy-terminated oligomer (0.70 g, 0.17 mmol) was stirred with thefluoro-terminated oligomer (0.63 g, 0.17 mmol), K2CO3 (0.24 g,1.70 mmol), and DMAc (7.8 mL) at 130 °C for 2 h undernitrogen. This yielded a white product with molecular weight(Mn ∼ 22.8 kg/mol) and degree of polymerization (N ∼ 3.9) asmeasured by gel permeation chromatography (GPC). We alsosynthesized the random copolymer (rPES-X0.5NY0.5N) throughthe polymerization of FPS (1.11 g, 4.38 mmol), HFBPA (0.74 g,

2.19 mmol), and BPA (0.50 g, 2.19 mmol) in the presence ofDMAc (10 mL) and K2CO3 (1.51 g, 10.9 mmol) at 130 °C for 4h. The post-reaction workup and GPC characterization were thesame as those described for the multiblock copolymerization.

2.3. Chloromethylation, Quaternization, and Mem-brane Casting. TCE was dried and stored over activated 3Amolecular sieves. Dry TCE (5 mL) was added to mPES-X6.7Y7.7(0.50 g). CMME (0.15 mL) and SnCl4 (0.01 mL) were theninjected into the solution. (Caution: chloromethyl methyl etheris carcinogenic and potentially harmful to human health.) Thereaction was allowed to continue for 4 h at 55 °C. The crudechloromethylated polymer was recovered by pouring thereaction mixture into methanol and washing the precipitatewith methanol several times.The chloromethylated copolymers (CmPES or CrPES) (0.50

g) were quaternized by reacting with 50 wt % trimethylamineaqueous solution (10 mL) at room temperature for 24 h. The dryquaternized copolymer (QPES) was recovered by evaporatingthe residual trimethylamine in a Petri dish. The QPES (0.50 g)was dissolved inN,N-dimethylformamide (DMF; 3 mL) and theresulting solution was filtered through a 0.45 μm poly-(tetrafluoroethylene) (PTFE) membrane filter. A film was castby pouring this solution into an aluminum dish, followed bydrying in a vacuum oven at 80 °C for 24 h. The free-standingpolymer films were about 100 μm thick and 3.5 × 3.5 cm. Thechloride ions in the film were exchanged for hydroxide ions bysoaking in 0.1 N KOH under nitrogen for about 12 h. After beingwashed several times with water, the QPES membranes inhydroxide form were stored in distilled water in a closed vial.

2.4. Measurements. The chemical structures of thesynthesized polymers were analyzed by a variety of NMRtechniques: one-dimensional 1H and 13C NMR, distortionlessenhancement by polarization transfer (DEPT), and two-dimensional heteronuclear multiple bond correlation(HMBC). A Bruker Avance III 400 spectrometer with 5-mmsample tubes was used. Chloroform-d or DMSO-d6 was used asthe NMR solvent. 1H spectra (16 scans) were collected at 400.13MHz with a 7.5 s recycle delay. The ion-exchange capacities(IEC) and degrees of chloromethylation (DC) were determinedfrom the respective 1H NMR spectra. The 13C and DEPT NMRspectra were collected at 100.61 MHz. The HMBC analysisemployed 32 scans, 128 increments along t1, 1024 data pointsalong t2, 160 Hz as a one-bond coupling constant, and 10 Hz as along-range coupling constant.To ensure full hydration, the QPES membranes in hydroxide

form were stored in distilled water for at least 1 week prior toNMR studies. A 3 × 10 mmmembrane strip was loaded into a 7-mm solid-state NMR rotor, and 1H NMR spin−lattice (T1) andspin−spin (T2) relaxation times were measured at 25 °C on aBruker DSX-300. The inversion recovery sequence 180°-τ-90°was used for the T1 measurements, and the Carr−Purcell−Meiboom−Gill pulse sequence (CPMG) was used for the T2measurements with an interpulse delay of 0.2 ms.The molecular weights of the polymers were determined by

GPC on a Waters 2690 separations module with a 2410differential refractive index detector, which was connected toWaters Styragel columns (HP 1, HP 3, HP 4). Tetrahydrofuran(THF) was used as the eluent and the solvent. The molecularweights were computed by a calibration curve based onpolystyrene standards.The water uptake of the membranes was evaluated according

to eq 1:

The Journal of Physical Chemistry C Article

dx.doi.org/10.1021/jp311987v | J. Phys. Chem. C 2013, 117, 15468−1547715469

=−

×W W

Wwater uptake 100w d

d (1)

where Wd is the dry mass of the membranes determined afterdrying in a desiccator andWw is the wet mass of the membraneswithout excess surface water after soaking for 24 h. The ionicconductivity measurements were performed in a four-probeelectrochemical impedance spectrometer by use of a PAR 2273potentiostat. Themembrane strips (1× 3 cm)weremounted in aconductivity cell and stabilized at a specific temperature (i.e., 25or 60 °C) under nitrogen. The frequency region from 1 Hz to 2MHz was scanned, where the impedance had a constant value.The ionic conductivity was calculated from eq 2:

σ =′L

Z A( ) (2)

where L is the length between sense electrodes (0.425 cm), Z′ isthe real component of the impedance response at high frequency,and A is membrane area available for hydroxide conduction.

3. RESULTS AND DISCUSSION3.1. Synthesis and Characterization of Multiblock and

Random Copolymers. A series of mPES multiblockcopolymers were synthesized by polycondensation of separatelyprepared HO-terminated and F-terminated oligomers, as shownin Scheme 1. The length of each oligomer was carefullycontrolled by optimizing the polymerization time (i.e., 3.5−7

Scheme 1. Synthesis of Multiblock Copoly(arylene ether sulfone), mPES-X6.7Y7.7a

ax = number of repeat units in hydrophobic block, y = number of repeat units in block that ultimately becomes hydrophilic, and N = degree ofpolymerization in the final polymer.



Figure 1. 1H NMR spectra of (a) HO-terminated oligomer (x = 6.7), (b) F-terminated oligomer (y = 7.7), and (c) resulting multiblock copolymermPES-X6.7Y7.7.



h). The integral ratio of 1H NMR peaks b′ and b allowscalculation of the degree of polymerization for the hydrophobicHO-terminated oligomer, which is x = 6.7 for the block shown inFigure 1a. In the same manner, the integral ratio of 1H NMRpeaks e′ and e allows calculation of the degree of polymerizationfor the F-terminated oligomer, which is y = 7.7 for the blockshown in Figure 1b. The F-terminated oligomer will become thecharge-carrying hydrophilic block after chloromethylation andquanternization, as described below. A summary of the oligomerlengths, which are assumed to correspond to block lengths afterincorporation into the multiblock copolymers, is shown in Table1, along with molecular weights and degrees of chloromethyla-tion for the prepared copolymers.

The GPC elution curves for the mPES copolymers wereunimodal and shifted to shorter elution times from those of theoligomers (cf. Figure 2), showing that the polymerization wassuccessful. The target molecular weight (Mn) of the mPES wasca. 30 kg/mol. This allowed solvent casting of a free-standingfilm.Random copoly(ether sulfone)s (rPES) were prepared in a

one-pot polycondensation reaction of FPS, HFBPA, and BPA inthe presence of K2CO3. The feed ratio of each monomer wascontrolled to give rPES copolymers with the same overall ratio ofrepeat units as a given mPES copolymer. DEPT-135 13C NMRspectra for rPES and mPES are shown in Figure 3. A majority ofthe peaks for the multiblock copolymer, seen in Figure 3a, aresharp singlets. In contrast, most of the peaks for the randomcopolymer, seen in Figure 3b, are multiplets. The randomlydistributed repeat units in the rPES cause the peak splitting.These results clearly show that the multiblock copolymers havehighly ordered structures compared to the random copoly-mers.20,21

3.2. Synthesis and Characterization of Chloromethy-lated and Quaternized Multiblock Copolymers. Multi-

block copolymers were chloromethylated by the Friedel−Craftsreaction with CMME and SnCl4 in TCE solution (cf. Scheme 1).The resulting chemical structures (CmPES) were characterizedand the degree of chloromethylation was evaluated by 1H NMR.A new peak at 4.53 ppm in the 1H NMR spectrum of CmPES-X6.7Y7.7, shown in Figure 4, was assigned to the methyleneprotons of the chloromethyl groups. New peaks at 6.85, 7.15, and7.35 ppm indicated that substitution of the -CH2Cl groupschanged the chemical shift of the BPA aromatic protons. Thepresence of these new peaks in the NMR spectra of the CmPESmaterials shows that the chloromethylation reaction wassuccessful. The degree of chloromethylation (DC) in Table 1was obtained by integrating the 1H NMR peak areas andcomparing the ratio of the -CH2Cl methylene protons to thedimethyl BPA protons. Previous studies reported the undesirablegelation of polymers caused by cross-linking during chlorome-thylation.22 GPC evaluation of the CmPES materials wasconducted to examine whether the molecular weight changeddue to chloromethylation. As shown in Table 1 and Figure 2, theCmPES copolymers do not exhibit significantly differentmolecular weight characteristics from their mPES precursorsand therefore did not undergo excessive branching or cross-linking during chloromethylation.It was expected that Friedel−Crafts alkylation of the

chloromethyl groups should selectively occur at the BPA moiety,which is more electron-rich than hexafluoroisopropylidenedi-phenyl and diphenylsulfone groups. The disappearance of thepeak at 6.9 ppm in Figure 1, peak h, after chloromethylationproves that chloromethyl groups are attached at the orthoposition to oxygen of the BPA. The two-dimensional HMBCspectrum was also obtained to verify the location of thechloromethyl group. In two-dimensional HMBC spectra, cross-peaks signify the coupling of proton and carbon nuclei separatedby two or three chemical bonds (cf. Figure 5).23 Unlike HSQCspectra that only reveal one-bond carbon−proton connectivity,HMBC spectra can be used to observe quaternary carbons andthereby provide valuable insight into substituents on aromaticrings.Cross-peak 1 in Figure 5 is assigned to correlate the methyl

protons (Hi) with the aromatic quaternary carbon (Ck) of theBPA. Cross-peak 2 correlates the aromatic quaternary carbon Ckwith the aromatic proton Hh′. Cross-peak 3 shows that thearomatic proton Hh′ is coupled to another aromatic quaternarycarbon (Ch) to which the chloromethyl group is attached. Theposition of the -CH2Cl groups are shown by the position of cross-

Table 1. Structural Characteristics of Multiblock and RandomCopoly(arylene ether sulfone)s before and afterChloromethylation

PESa CPESb

oligomer repeatunits (x/y)c Mn

d PDIe Mnd PDIe DCf

mPES-X6.7Y3.4

6.7/3.4 34.5K 2.9 47.2K 3.2 2.1

mPES-X9.2Y3.4

9.2/3.4 24.9K 2.1 26.5K 3.9 2.8

mPES-X11.3Y3.4

11.3/3.4 25.3K 2.3 25.9K 2.5 2.4

mPES-X6.7Y2.5

6.7/2.5 18.4K 2.2 21.4K 2.4 2.0

mPES-X6.7Y7.7

6.7/7.7 22.8K 2.1 27.5K 2.5 2.1

mPES-X6.7Y11.1

6.7/11.1 22.2K 2.2 28.5K 3.7 2.2

rPES-X0.5NY0.5N

55.3K 2.4 82.9K 4.7 2.0

rPES-X0.67NY0.33N

18.8K 2.0 19.2K 2.6 2.1

aPES = copoly(arylene ether sulfone)s. bCPES = chloromethylatedcopoly(arylene ether sulfone)s. cx = number of repeat units inhydrophobic block; y = number of repeat units in block that ultimatelybecomes hydrophilic. dMn = number-average molecular weight; K =kg/mol. ePDI = polydispersity index. fDegree of chloromethylation =number of chloromethyl groups per Y-type repeat unit.

Figure 2. Gel-permeation chromatograms of reactant oligomers, theresulting multiblock copolymer mPES-X6.7Y7.7, and the correspondingchloromethylated product CmPES-X6.7Y7.7.



peak 4. This peak shows the correlation between Ch and themethylene protons (Hj) of the chloromethyl groups.The quaternary ammonium salt of CmPES was formed by

immersing the dried powder in an aqueous trimethylaminesolution. The quaternized copolymer (QmPES) had differentsolubility characteristics from CmPES. The QmPES was solublein aprotic polar solvents such as DMSO, DMF, and DMAc. 1HNMR was used to investigate the chemical structure andcompletion of the chemical reaction for QmPES polymers, asshown in Figure 4b. The position of the amino methyleneprotons (4.63 ppm) shifted slightly downfield from that of thechloromethylene protons (4.53 ppm). The ratio of the integratedpeaks between amino methylene protons and dimethyl protonsof the BPA was consistent during the quaternization reaction.This comparison confirmed completion of the reaction. Theintegrated peak value for the trimethyl protons also supportsquantitative quaternization.The ion-exchange capacity (IEC) of the QPES copolymers

was computed from eq 3:

=+

×y

xM yMIEC

(DC)w, w,

1000phob phil (3)

where DC is the degree of chloromethylation, Mw,phob is therepeat-unit molecular weight of the hydrophobic block, andMw,phil is the repeat-unit molecular weight of the hydrophilicblock. The IEC values for the copolymers are summarized inTable 2.3.3. Membrane Formation, Water Uptake, and

Hydroxide Ion Conductivity of QmPES and QrPES. Free-standing QPES membranes were obtained by solution castingfrom DMF. The chloride ion was exchanged for hydroxide in the

membranes immediately before measuring the water uptake andionic conductivity to minimize the effect of converting thehydroxide into carbonates through contact with CO2 in air.Water uptake is particularly important for ion-exchangepolymers used as ionomers in fuel cell electrodes. Typically,ionic conductivity is proportional to water uptake. However,excessive water uptake can swell the polymer and result in poorreactant transport within the electrodes. Thus, the optimumionomer has moderate water uptake and high conductivity.24

Table 2 gives the water uptake at room temperature andhydroxide ion conductivity at 25 and 60 °C.Since the DC of the membranes is approximately the same (cf.

Table 1), the IEC is mostly determined by the ratio of the lengthsof the charge-carrying hydrophilic to hydrophobic blocks (y/x).As shown in Table 2 and Figure 8c, water uptake generallyincreases with increasing IEC values for the multiblockcopolymer membranes. However, the ionic conductivity doesnot exhibit a clear dependence on IEC (Figure 8a). Moreover, itshould be noted that QmPES-X6.7Y3.4 showed the highest anionicconductivity in comparison to the other AEMs in spite of itsmoderate IEC and water uptake. A comparison betweenmultiblock and random copolymers with similar IEC and y/xvalues, for example, mPES-X6.7Y3.4 versus rPES-X0.67NY0.33N andmPES-X6.7Y7.7 versus rPES-X0.5NY0.5N, clearly reveals that themultiblock copolymers exhibit higher ionic conductivity. Thehigh conductivities of multiblock copolymer AEMs are notsimply a result of the IEC and water uptake. Rather, thenanophase morphology is important; the specific nanophasestructure of the multiblock material facilitates ionic transport.

3.4. Morphology and Water Transport in QPESMembranes. Hydroxide conductivity in these AEMs is directlyproportional to ion concentration and ion mobility.16 Ion

Figure 3. Selected aromatic region of DEPT-135 13C NMR spectra of (a) multiblock copoly(arylene ether sulfone), mPES-X6.7Y7.7 and (b) randomcopoly(arylene ether sulfone), rPES- X0.5NY0.5N.



concentration is relatively easy to evaluate from the ion-exchangecapacity (IEC). On the other hand, effective ion mobility is lessstraightforward to assess because it is affected by polymermorphology such as the degree of phase separation and the shapeand size of the domains (i.e., channels).Several attempts to relate ion mobility to underlying polymer

morphology have been reported for proton conductivity inPEMs.16,25,26 Systematic studies of the relationship betweenmorphology and conductivity have employed block copolymerswith controlled segment lengths and water contents.27−29

However, such systematic studies of hydroxide ion mobilityand conductivity in AEMs have not appeared. Watanabe and co-workers18 used scanning transmission electron microscopy(STEM) to examine the morphology of multiblock AEMs, butthey did not correlate ion mobility with block lengths or ionconcentrations.While imaging experiments provide a visualization of the

morphology,18,30 the information does not directly show ion/

water transport or provide a quantitative metric for evaluatingdifferent materials. In addition, imaging techniques suffer fromcomplicated and sometimes inconsistent sample preparationprocedures. In contrast, NMR measurements enable directcharacterization of local molecular dynamics and ion/watermovement through domains inside membranes. Specifically,NMR spin−lattice (T1) and spin−spin (T2) relaxation timesprovide direct indications of the restricted motions of water inconfined environments or near surfaces. The spin−lattice NMRrelaxation time (T1) has been used to examine water confine-ment and physical interactions of water molecules with thepolymer in Nafion.31 Sierra-Martin et al.32 measured T2 to detectmotional confinement of water molecules in a poly(NIPAM)microgel. They found that T2 could be correlated with changes inwater mobility and degree of confinement. Morphologicaldifferences between random and multiblock copolymers inPEMs were revealed by analyzing bound versus free waterthrough T2 values (water associated with sulfonic acid groupsrelaxed faster).14

The nature of the ion/water channels in the copolymerssynthesized here were investigated by examining NMR relaxationtimes by use of the inversion recovery pulse sequence for T1measurements and the CPMG pulse sequence for T2 measure-ments. The relaxation times, summarized in Table 2, correspondto water in the membranes rather than the polymer protons. Thisis confirmed by the absence of peaks at 7−8 ppm belonging tothe polymer, as shown in Figure 6. In addition, the water contentin the membranes was high enough (8.4−54.7%) so that theNMR peaks due to water dominate the NMR spectrum.As shown in Table 2, the sample that exhibits the highest ion

conductivity (15.7 mS/cm at 25 °C and 37.7 mS/cm at 60 °C) isthe multiblock copolymer mPES-X6.7Y3.4. The random copoly-mer with the same IEC (1.38 meq/g), rPES-X0.67NY0.33N,exhibited much lower conductivity (5.7 mS/cm at 25 °C and13.6 mS/cm at 60 °C), but it also absorbed less water (16%versus 24.5%). However, despite the higher water content, themultiblock exhibits a T2 relaxation time that is significantlyshorter (20.2 ms) than T2 for the random copolymer (50.2 ms)(cf. Figure 7). The same relationships are also observed for

Figure 4. 1H NMR spectra of the multiblock copoly(arylene ethersulfone), mPES-X6.7Y7.7, after (a) chloromethylation to give CmPES-X6.7Y7.7 and (b) quaternization to give QmPES-X6.7Y7.7.

Figure 5. HMBC NMR spectrum of the chloromethylated multiblockcopoly(arylene ether sulfone), CmPES-X6.7Y7.7. See text for descriptionof assignments for cross-peaks 1−4.



another pair of multiblock and random copolymers withapproximately the same IEC, mPES-X6.7Y7.7 (IEC = 2.28 meq/g) and rPES-X0.5NY0.5N (IEC = 2.17 meq/g): the multiblockcopolymer exhibits higher water uptake and conductivity and alower T2 value. For this to occur, the water molecules that arepresent in the multiblock must be more closely interacting withthe solid polymer than the water in the random copolymer. Thisimplies that the water in the random copolymer pools in afashion that is inefficient for ion conductivity. This observation is

consistent with previously reported studies18 in which theauthors claimed that multiblock copolymer AEMs with well-developed phase-separated morphologies utilize water moleculesin hydrophilic blocks more efficiently than random copolymersfor hydroxide ion transport.The sample that exhibits the highest ion conductivity, mPES-

X6.7Y3.4, is not the sample with the greatest IEC or water uptake,in contrast to many, but not all,26,33 published reports on PEMpolymers. Conductivity versus IEC is shown in Figure 8a for themultiblock copolymers. The samples with IEC values greaterthan 1.38 meq/g exhibit lower conductivities than the mPES-X6.7Y3.4 multiblock copolymer. This trend is similar to the oneshown for a plot of conductivity versus water uptake in Figure 8b.The sample with the highest conductivity does not absorb themost water. Such trends have been reported in the literature byPeckham et al.26 for an α,α,β-trifluorostyrene-based PEM:maximum conductivity was reached at IEC ≈ 2 meq/g afterwhich the conductivity decreased and finally leveled off withincreasing IEC. Tsang et al.33 also observed the same trend forproton-conductive fluorous ionic graft copolymers. Bothpublications attributed this nonlinear relationship to a dilutionof the available sulfonic acid groups caused by a significantincrease in water. As shown in Figure 8c, water uptake increasesmonotonically with IEC, as is often observed and expected.Figure 8d shows that T2 also generally increases with IEC,especially for the two multiblock copolymers with IEC > 1.38meq/g, indicating that the fraction of water that interacts with theionic polymer segments decreases for the higher water contents.The observed relationships are indeed interesting and warrantfurther study.For the following group of multiblock copolymers that do not

contain the most conductive sample (i.e., mPES-X6.7Y2.5, mPES-X6.7Y7.7, and mPES-X6.7Y11.1), T1 and T2 increase as the length ofthe hydrophilic segment (y) increases for the same hydrophobicsegment length (x = 6.7). This trend is consistent with theconcept that longer hydrophilic blocks generate morphologieswith larger aqueous domains, and is consistent with the wateruptake values (8.4%, 47.5%, and 57.5%). No significant effect onthe NMR relaxation times is observed for the multiblockcopolymers that have the same hydrophilic segment length (y =3.4) but different hydrophobic segment lengths of x = 11.3 and9.2.A series of multiblock copolymers was also synthesized in

which the hydrophobicity of the hydrophobic segments wasincreased beyond those described here. The idea was to createmore strongly phase-separated multiblock copolymers toexamine whether the channel structure could be made moreefficient, and therefore more conductive, for a given IEC andwater content. The same trend was observed for conductivity

Table 2. Water Uptake, Ion Conductivity, and Water NMR Relaxation Times in QPES Membranes

conductivity (mS/cm)

y/xa IEC (meq/g) water uptake (%) at 25 °C at 60 °C T1 (ms) T2 (ms)

mPES-X11.3Y3.4 0.30 1.06 13.5 2.4 5.0 309 28.0mPES-X9.2Y3.4 0.37 1.45 18.0 4.9 12.7 304 33.3mPES-X6.7Y3.4 0.51 1.38 24.5 15.7 37.7 224 20.2mPES-X6.7Y2.5 0.37 1.02 8.4 3.2 7.2 229 22.2mPES-X6.7Y7.7 1.15 2.28 47.5 13.9 27.9 376 48.0mPES-X6.7Y11.1 1.66 2.82 57.5 14.2 29.0 483 95.6rPES-X0.5NY0.5N ∼1.0 2.17 31.1 11.5 25.0 324 57.2rPES-X0.67NY0.33N ∼0.5 1.38 16.0 5.7 13.6 212 50.2

aRatio of hydrophilic/hydrophobic repeat units.

Figure 6. 1HNMR stacked plot of CPMG spectra for QmPES-X6.7Y3.4 asa function of the number of cycles (n).

Figure 7. Signal intensity for CPMG spectra of QmPES-X6.7Y3.4 andQrPES-X0.67NY0.33N as a function of the number of cycles. Thesecopolymers are characterized by the same IEC (1.38 meq/g). The waterin the multiblock exhibits a shorter T2 relaxation time, despite a greaterwater content than the random copolymer (24.5% versus 16.0%; cf.Table 2).



versus IEC; that is, the most conductive sample was not the onewith the greatest IEC. These results and the usage of these newlysynthesized multiblock copolymers as membranes and/orionomers in hybrid polymer electrolyte fuel cells34 and alkalinedirect methanol fuel cells will be described in a future publication.

4. CONCLUSIONS

A systematic study of the effect of block lengths and the degree ofrandomness on the ionic conductivity in a series of multiblockand random aromatic copolymers containing quaternaryammonium groups was undertaken. The polymers weresynthesized by a polycondensation reaction. The reactionconditions were optimized and the length of the blocks, overallmolecular weight of the polymers, and degree of chloromethy-lation were controlled. NMR techniques, including DEPT-135and two-dimensional HMBC, were used to investigate the degreeof randomness and the position of chloromethyl groups withinthe aromatic copolymer backbone. Quaternary ammonium andanionic conductive membranes were prepared with reasonablyhigh conductivity despite the low water uptake: 37.7 mS/cm at60 °C in mPES-X6.7Y3.4 with 24.5% water. NMR relaxation times(T1 and T2) of water in the membranes were used to explain thehigh conductivity obtained with mPES-X6.7Y3.4 compared tomultiblock copolymers with higher IEC values, the higherconductivity of multiblock copolymer membranes compared totheir random copolymer counterparts, and the observednonlinear dependency of ionic conductivity on IEC and watercontent. Higher performance resulted from an optimizedcombination of IEC, chain structure, and morphology.

■ AUTHOR INFORMATION

NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTS

We thank Dr. Anselm Griffin and Dr. Stefan France for valuablediscussions and gratefully acknowledge the financial support ofthe Army Research Laboratory.

■ REFERENCES(1) Larminie, J. Fuel Cell Systems Explained; J. Wiley: Chichester, U.K.,2003.(2) Rikukawa, M.; Sanui, K. Proton-conducting polymer electrolytemembranes based on hydrocarbon polymers. Prog. Polym. Sci. 2000, 25,1463−1502.(3) Thounthong, P.; Rael, S.; Davat, B. Control strategy of fuel cell/supercapacitors hybrid power sources for electric vehicle. J. PowerSources 2006, 158, 806−814.(4) Mauritz, K. A.; Moore, R. B. State of understanding of Nafion.Chem. Rev. 2004, 104, 4535−4585.(5) Agel, E.; Bouet, J.; Fauvarque, J. F. Characterization and use ofanionic membranes for alkaline fuel cells. J. Power Sources 2001, 101,267−274.(6) Asazawa, K.; Yamada, K.; Tanaka, H.; Oka, A.; Taniguchi, M.;Kobayashi, T. A platinum-free zero-carbon-emission easy fuelling directhydrazine fuel cell for vehicles. Angew. Chem., Int. Ed. 2007, 46, 8024−8027.(7) Pan, J.; Lu, S. F.; Li, Y.; Huang, A. B.; Zhuang, L.; Lu, J. T. High-performance alkaline polymer electrolyte for fuel cell applications. Adv.Funct. Mater. 2010, 20, 312−319.(8) Merle, G.; Wessling, M.; Nijmeijer, K. Anion exchange membranesfor alkaline fuel cells: A review. J. Membr. Sci. 2011, 377, 1−35.(9) Couture, G.; Alaaeddine, A.; Boschet, F.; Ameduri, B. Polymericmaterials as anion-exchange membranes for alkaline fuel cells. Prog.Polym. Sci. 2011, 36, 1521−1557.(10) Diat, O.; Gebel, G. Proton channels. Nat. Mater. 2008, 7, 13−14.(11) Miyatake, K.; Shimura, T.; Mikami, T.; Watanabe, M. Aromaticionomers with superacid groups. Chem. Commun. 2009, 6403−6405.(12) Yoshimura, K.; Iwasaki, K. Aromatic polymer with pendantperfluoroalkyl sulfonic acid for fuel cell applications. Macromolecules2009, 42, 9302−9306.(13) Matsumoto, K.; Higashihara, T.; Ueda, M. Locally and denselysulfonated poly(ether sulfone)s as proton exchange membrane.Macromolecules 2009, 42, 1161−1166.(14) Roy, A.; Lee, H. S.; McGrath, J. E. Hydrophilic-hydrophobicmultiblock copolymers based on poly(arylene ether sulfone)s as novelproton exchange membranes - Part B. Polymer 2008, 49, 5037−5044.(15) Bae, B.; Yoda, T.; Miyatake, K.; Uchida, H.;Watanabe, M. Proton-conductive aromatic ionomers containing highly sulfonated blocks forhigh-temperature-operable fuel cells. Angew. Chem., Int. Ed. 2010, 49,317−320.

Figure 8. (a) Conductivity at 60 °C as a function of IEC, (b) conductivity at 60 °C as a function of water uptake, (c) water uptake as a function of IEC,and (d) T2 as a function of IEC for the series of multiblock copolymers with properties summarized in Tables 1 and 2.



(16) Peckham, T. J.; Holdcroft, S. Structure-morphology-propertyrelationships of non-perfluorinated proton-conducting membranes.Adv. Mater. 2010, 22, 4667−4690.(17) Li, N.; Wang, C.; Lee, S. Y.; Park, C. H.; Lee, Y. M.; Guiver, M. D.Enhancement of proton transport by nanochannels in comb-shapedcopoly(arylene ether sulfone)s. Angew. Chem., Int. Ed. 2011, 50, 9158−9161.(18) Tanaka, M.; Fukasawa, K.; Nishino, E.; Yamaguchi, S.; Yamada,K.; Tanaka, H.; Bae, B.; Miyatake, K.; Watanabe, M. Anion conductiveblock poly(arylene ether)s: synthesis, properties, and application inalkaline fuel cells. J. Am. Chem. Soc. 2011, 133, 10646−10654.(19) Zhou, J.; Unlu, M.; Vega, J. A.; Kohl, P. A. Anionic polysulfoneionomers and membranes containing fluorenyl groups for anionic fuelcells. J. Power Sources 2009, 190, 285−292.(20) Zhou, X. M.; Jiang, Z. H. Sequence analysis of poly(ether sulfone)copolymers by C-13 NMR. J. Polym. Sci., Part B: Polym. Phys. 2005, 43,1624−1630.(21) Lee, H.-S.; Roy, A.; Lane, O.; Dunn, S.; McGrath, J. E.Hydrophilic-hydrophobic multiblock copolymers based on poly(aryleneether sulfone) via low-temperature coupling reactions for protonexchange membrane fuel cells. Polymer 2008, 49, 715−723.(22) Wang, G.; Weng, Y.; Chu, D.; Chen, R.; Xie, D. Developing apolysulfone-based alkaline anion exchange membrane for improvedionic conductivity. J. Membr. Sci. 2009, 332, 63−68.(23) Khandelwal, D.; Hooda, S.; Brar, A. S.; Shankar, R. 1D and 2DNMR studies of isobornyl acrylate - Methyl methacrylate copolymers. J.Mol. Struct. 2011, 1004, 121−130.(24) Zhou, J.; Uenlue, M.; Anestis-Richard, I.; Kim, H.; Kohl, P. A.Solvent processible, high-performance partially fluorinated copoly-(arylene ether) alkaline ionomers for alkaline electrodes. J. Power Sources2011, 196, 7924−7930.(25) Kreuer, K. D. On the development of proton conducting polymermembranes for hydrogen and methanol fuel cells. J. Membr. Sci. 2001,185, 29−39.(26) Peckham, T. J.; Schmeisser, J.; Rodgers, M.; Holdcroft, S. Main-chain, statistically sulfonated proton exchange membranes: therelationships of acid concentration and proton mobility to watercontent and their effect upon proton conductivity. J. Mater. Chem. 2007,17, 3269−3269.(27) Bae, B.; Miyatake, K.; Watanabe, M. Sulfonated poly(aryleneether sulfone ketone) multiblock copolymers with highly sulfonatedblock: Synthesis and properties. Macromolecules 2010, 43, 2684−2691.(28) Zhao, C. J.; Lin, H. D.; Li, X. F.; Na, H. Morphologicalinvestigations of block sulfonated poly(arylene ether ketone)copolymers as potential proton exchange membranes. Polym. Advan.Technol. 2011, 22, 2173−2181.(29) Na, T. Y.; Shao, K.; Zhu, J.; Sun, H. C.; Liu, Z. G.; Zhao, C. J.;Zhang, Z. Q.; Lew, C.M.; Zhang, G. Block sulfonated poly(arylene etherketone) containing flexible side-chain groups for direct methanol fuelcells usage. J. Membr. Sci. 2012, 417, 61−68.(30) Zhang, X.; Hu, Z.; Luo, L.; Chen, S.; Liu, J.; Chen, S.; Wang, L.Graft-crosslinked copolymers based on poly(arylene ether ketone)-gc-sulfonated poly(arylene ether sulfone) for PEMFC applications.Macromol. Rapid Commun. 2011, 32, 1108−1113.(31) Nicotera, I.; Coppola, L.; Rossi, C. O.; Youssry, M.; Ranieri, G. A.NMR investigation of the dynamics of confined water in Nafion-basedelectrolyte membranes at subfreezing temperatures. J. Phys. Chem. B2009, 113, 13935−13941.(32) Sierra-Martin, B.; Choi, Y.; Romero-Cano, M. S.; Cosgrove, T.;Vincent, B.; Fernandez-Barbero, A. Microscopic signature of a microgelvolume phase transition. Macromolecules 2005, 38, 10782−10787.(33) Tsang, E.M.W.; Zhang, Z. B.; Yang, A. C. C.; Shi, Z. Q.; Peckham,T. J.; Narimani, R.; Frisken, B. J.; Holdcroft, S. Nanostructure,morphology, and properties of fluorous copolymers bearing ionic grafts.Macromolecules 2009, 42, 9467−9480.(34) Unlu, M.; Zhou, J. F.; Kohl, P. A. Hybrid polymer electrolyte fuelcells: Alkaline electrodes with proton conducting membrane. Angew.Chem., Int. Ed. 2010, 49, 1299−1301.



anion-conductive multiblock aromatic copolymer membranes...

Documents