Highly Conductive Anion-Exchange Membranes Based on Cross-Linked Poly(norbornene): Vinyl Addition PolymerizationMrinmay Mandal, Garrett Huang, and Paul A. Kohl*

School of Chemical and Biomolecular Engineering Georgia Institute of Technology Atlanta, Georgia 30332-0100, United States

*S Supporting Information

ABSTRACT: Cross-linked (XL) anion-exchange membranes (AEMs) synthe-sized by vinyl addition polymerization of norbornene were prepared for use inanion-exchange membrane electrochemical devices, including fuel cells andelectrolyzers. Tetrablock copolymers composed of an all-hydrocarbonbackbone with a very high ion-exchange capacity (IEC), 3.46 mequiv/g,were synthesized. Light cross-linking was found to be adequate for providingcritical control over unwanted water uptake. This enabled use of very high IECmembranes. Without light cross-linking, the unwanted water uptake wouldcause swelling and softening of the membrane. The best performing membranehad no significant drop in ionic conductivity over 1000 h of aging in 1 MNaOH at 80 °C and a record high ionic conductivity of 198 mS/cm at 80 °C(for a chemically stable AEM). The number of bound and free water moleculesper ion pair is described along with ion mobility comparisons to previousmaterials. The membranes are suitable for electrochemical devices and wereused in AEM fuel cells.

KEYWORDS: anion-exchange membranes (AEMs), vinyl addition, poly(norbornene), cross-linking, fuel cells

■ INTRODUCTION

Fuel cells are a clean energy conversion technology with thepotential to reduce the use of fossil fuels.1 Fuel cells can beused in stationary power generation, portable electronics, andtransportation.2,3 In addition, fuel cells are environmentallyfriendly, can be easy to refuel, and can have high energyconversion efficiency.4 Solid polyelectrolyte membranes, suchas anion-exchange membranes (AEMs) and proton-exchangemembranes (PEMs), simplify the fabrication of electrodes witha three-phase boundary because liquid/gas pressures do nothave to be balanced like in liquid electrolyte devices. High pHAEMs have facile oxygen reaction kinetics compared to acidconducting PEMs and offer the opportunity to use nonpre-cious metal catalysts, and reduced fuel crossover.4−9 However,early membranes suffered from low ion conductivity, poorchemical stability at high pH, and high water uptake.10−13

More recently, higher conductivity (e.g., 100 mS/cm at 80 °C)and chemical stability (80 °C in 1 M NaOH) have beenachieved by a number of researchers, as reviewed by Arges.14

This notable progress has shown that certain structuralmoieties can be used to address pervious AEM deficiencies.The structure of the polymer backbone, the position of the

cations in the polymer architecture, and the nature of thecations determine the conductivity and long-term alkalinestability of AEMs. Polymer backbones containing polysulfone,polyketone, and poly(aryl ether) moieties are susceptible tohydroxide attack and polymer backbone degradation.15−22

Backbone degradation problems are mitigated by the use of apolymer with an all-hydrocarbon backbone and headgroup

tether.12 Cation degradation can be mitigated by positioningthe headgroup at the end of a long alkyl chain tether, typically4−6 carbons long.23 In addition, the alkyl tether can isolate thecation headgroup from the electron withdrawing inductiveeffect of aromatic groups in the tether or polymer backbone (ifaromatic groups are present). Hence, stable AEMs underrealistic operating conditions (e.g., 80 °C and 1 M KOH) canbe synthesized by combining an all-hydrocarbon backbonewith tethered cations on long alkyl chains.24−26

In addition to alkaline stability, electrochemical devicesrequire AEMs with high conductivity to achieve low ohmicresistance losses. Hydroxide conductivity is a function of theion mobility and ion-exchange capacity (IEC). The IEC of theAEM is often kept to a modest value in an effort to avoid highwater uptake which can result in swelling of the membrane andlow ion mobility. The mobility can be improved by theformation of efficient ion conducting channels (such as by theuse of block copolymers) and the prevention of excess wateruptake within the membrane.27 Thus, membranes face theconundrum of striving to achieve high IEC but suffering theconsequences that come from the water that the ions attract.12

Cross-linking can be used to address excess water uptake butoften at the expense of low ion mobility.Phase segregation within block copolymers aids in the

formation of hydrophobic and hydrophilic regions within the

Received: November 28, 2018Accepted: March 20, 2019Published: March 20, 2019

Article

www.acsaem.orgCite This: ACS Appl. Energy Mater. 2019, 2, 2447−2457

© 2019 American Chemical Society 2447 DOI: 10.1021/acsaem.8b02051ACS Appl. Energy Mater. 2019, 2, 2447−2457

Dow

nloa

ded

via

GE

OR

GIA

IN

ST O

F T

EC

HN

OL

OG

Y o

n A

pril

25, 2

019

at 1

8:20

:33

(UT

C).

Se

e ht

tps:

//pub

s.ac

s.or

g/sh

arin

ggui

delin

es f

or o

ptio

ns o

n ho

w to

legi

timat

ely

shar

e pu

blis

hed

artic

les.

polymer. The quaternary ammonium headgroups within thehydrophilic phase are where hydroxide transport occurs.11 It isnow evident that carbonate conductivity is very important inAEM (vs hydroxide conductivity) because of the uptake ofcarbon dioxide from the ambient air. In a fuel cell, the carbondioxide at the air cathode is readily absorbed and converts thehydroxide produced at the cathode to bicarbonate orcarbonate. Once the bicarbonate or carbonate is transportedto the hydrogen anode in a fuel cell, the evolved carbondioxide will build-up within the recycled hydrogen fuel alongwith the water produced at the anode. Both carbon dioxide andwater can diffuse back through the membrane continuing theprocess of hydroxide neutralization and carbonate migration.Fuel cell tests using fresh feed hydrogen avoid facing thiscritical issue of carbon dioxide build-up and carbonateconduction. Thus, it is imperative that the IEC and ionmobility be as high as possible for efficient carbonateconduction. Carbonate mobility is much lower than hydroxidemobility.One of the remaining challenges in the design of stable, high

conductivity AEMs is water uptake. Excessive water uptake canoccur at high IEC causing channel flooding and membraneswelling. This leads to mechanical distortion and softening ofthe membranes.12 Materials with high IEC have a tendency toadsorb large quantities of water. Some water is needed to formthe ion solvent shell as well as dilute the hydroxide salt withinthe membrane. The absorbed water must be adequate for ionsolvation; however, excess free water is not productive ordesired. Thus, the water content can be divided into boundwater (for forming the solvent shell) and free water.11,12

Hence, it is necessary to choose an IEC which balances theamount of free and bound water inside the membranes to yieldmaximum ion mobility (i.e., conductivity) while maintainingAEM mechanical properties. Although this study does not dealwith polymers with a different number of blocks, this too hasan impact on conductivity.28−33

A previous report of using vinyl addition poly(norbornene)in an AEM only produced a very low conductivity (4 mS/cm at80 °C) and showed a modest decline in conductivity aftersoaking in 6 M NaOH at room temperature.34 It should benoted that the ion-exchange capacity (1.83 mequiv/g) wassignificantly less than what is reported in this paper, thebackbone was not a block copolymer, and the headgrouptethers contained ether linkages which are known to besusceptible to hydroxide attack. In another report, a blockcopolymer form of poly(norbornene) was synthesized by vinyladdition polymerization for use as a pervaporization mem-brane.35 It was not ion conductive, did not have tetheredcation headgroups, and did not face the same challenges ofbalancing ion conductivity, water uptake, and chemical stabilityin base.A strategy to avoid the unwanted water swelling issue is to

utilize cross-linking to increase the mechanical stability.Previously, several research groups have reported the synthesisof cross-linked AEMs with improved alkaline stability,dimensional stability, and swelling resistance.36−43 Whilemany of these cases showed high hydroxide conductivity, thereported AEMs did not have adequate long-term alkalinestability. Zhang et al. synthesized cross-linked AEMs with ahydroxide conductivity of ∼200 mS/cm at 80 °C; however, thelong-term alkaline stability was not measured.44 Recently, Zhuet al. reported cross-linked AEMs with high ionic conductivity(200 mS/cm at 80 °C); however, there was 27% degradation

when the membrane was aged in 1 M NaOH solution at 80 °Cfor 500 h.41 In addition, recently, Wang et al. synthesizedhighly conductive (>200 mS/cm at 80 °C) AEMs with 6.2%conductivity loss after 500 h in 1 M hydroxide solution at 80°C.45

A further consideration is the need for exceptionally highconductivity in fuel cell membranes. Conductivity tests areoften performed with only hydroxide conductive ions.Although hydroxide may be produced at the alkaline fuel cellcathode, under steady-state operating conditions, it quicklyforms carbonate due to the accumulation of carbon dioxide inthe recycled hydrogen at the anode. The only means to removethe carbon dioxide is to vent unused hydrogen (a veryundesirable process) or let the carbon dioxide diffuse backacross the membrane, which converts hydroxide to carbonate.Thus, under steady-state conditions, carbonate conductivitywill dominate. Carbonate mobility is much less than hydroxidemobility. Thus, extra conductivity is needed.In this study, the benefits of light polymer cross-linking are

investigated as a means of implementing polymers with veryhigh IEC yet maintaining high hydroxide mobility (withproven stability) without excessive water uptake or membraneswelling. Poly(norbornene) is a favorable polymer backbonefor low cost AEMs because of its all-hydrocarbon backboneand inexpensive starting material, dicyclopentadiene. Thispaper shows that very high IEC polymers can be obtained(through the use of low molecular weight for norbornenemonomers) and used to achieve record hydroxide conductivityvia light cross-linking. Poly(norbornene)s can be polymerizedthrough several different synthetic routes. In this study, theresults of vinyl addition polymerization of norbornenes areexplored. The results from ring-opening metathesis polymer-ization (ROMP) of norbornenes are disclosed in a companionpublication of this study (Scheme S1).46 The methods ofsynthesis and resulting properties of ROMP and vinyl additionpolymer are different. Especially notable is the glass transitiontemperature (Tg). We note that several types of ion conductingpolymer can be useful in electrochemical devices (e.g., fuelcells, electrolyzers, flow batteries). The membrane separatingthe two electrodes should have the highest possible ionconductivity and low fuel crossover. The ionomer used tomake the electrodes should be compatible with catalyst andelectrode fabrication methods. There is also the need for ionconducting adhesion layers between the electrode andmembrane. Thus, there is interest in ion conducting polymerswith different properties. Vinyl addition polymerized norbor-nenes have a high glass transition temperature.47−50 Vinyladditional poly(norbornene) itself has a Tg of 390 °C whilepoly(hexylnorbornene) has a Tg of 265 °C. A variety ofnorbornene copolymers were shown to have a Tg between 340and 355 °C,47 203 and 331 °C,48 293 and 360 °C,49 and over380 °C for poly(methylnorbornene).50

It was shown previously that the vinyl addition polymer-ization route produces AEMs with high hydroxide conductivity(123 mS/cm) and excellent alkaline stability (<1% loss ofconductivity in 1200 h in 1 M NaOH at 80 °C).12 For AEMs,there is a compromise between the polydispersity of the blocks,overall molecular weight of the polymer, and product yield. Atetrablock copolymer was chosen for study because it has asufficiently high molecular weight and good yield. Futurepublications may investigate the trade-offs in block size andnumber. In this study, it was found that light cross-linking ofpostpolymerized norbornene extended the usable IEC range to

ACS Applied Energy Materials Article

DOI: 10.1021/acsaem.8b02051ACS Appl. Energy Mater. 2019, 2, 2447−2457

2448

very high values without unwanted water uptake or swelling.To the best of our knowledge, the hydroxide conductivityreported here (198 mS/cm at 80 °C) is the highest reportedfor a chemically stable polymer (at 80 °C in 1 M NaOH).

■ EXPERIMENTAL SECTIONMaterials. The chemicals used in this study are as follows: 1-

Hexene, 5-bromo-1-pentene, and dicyclopentadiene were obtainedfrom Alfa Aesar (used as-received); the butyl norbornene (BuNB)and bromopropyl norbornene (BPNB) monomers were prepared by aDiels−Alder reaction, following a previously published procedure.51

The monomers were purified by distillation over sodium in threefreeze−pump−thaw cycles. The synthesis reactions were carried outin a dry argon glovebox with care to limit the exposure to air andmoisture. The toluene was heating at reflux for 6 h over sodium andbenzophenone to remove water. Freshly distilled toluene was used inthe synthesis. Triisopropylphosphine and [(η3-allyl)Pd(Cl)]2 (Sigma-A l d r i c h ) we r e u s ed a s - r e c e i v ed . (A l l y l ) p a l l a d i um -(triisopropylphosphine)chloride ((η3-allyl)Pd(iPr3P)Cl) was used asthe catalyst and was prepared as described previously.52 Lithiumtetrakis(pentafluorophenyl)-borate·(2.5Et2O) (Li[FABA]) (BoulderScientific Co.) was used as-received. N,N,N′,N′-Tetramethyl-1,6-hexanediamine (TMHDA); α,α,α-trifluorotoluene (TFT), anhydrous,≥99%; and tetrahydrofuran (THF) (Sigma-Aldrich) were used as-received.Synthesis of Tetrablock Copolymer [Poly(BuNB-b-BPNB-b-

BuNB-b-BPNB)]. Tetrablock PNB-X34-Y66 (PNB = polynorbornene,X34 = mole percent of combined hydrophobic blocks, Y66 = molepercent of combined halogenated blocks) was prepared as describedpreviously.12 The catalyst was made by dissolving (η3-allyl)Pd(iPr3P)Cl (26 mg, 0.074 mmol) and Li[FABA] (65 mg, 0.074 mmol) in 0.5 gof TFT and 0.5 g of toluene. BuNB (0.28 g, 1.86 mmol) and toluene(6 mL) were then added and stirred. The catalyst was added undervigorous stirring. The BuNB polymerization reaction was complete in10 min. The product was checked by gel permeation chromatography(GPC). BPNB (1.6 g, 7.44 mmol) and toluene (32 mL) were addedto the catalyst-containing solution and stirred for 3 h to add theBPNB block to the BuNB block. After consumption of BPNB, theproduct was checked by GPC analysis. The third block was formed byadding BuNB (0.28 g, 1.86 mmol) and toluene (6 mL) and allowed toreact for 10 min. Finally, BPNB (1.6 g, 7.44 mmol) and toluene (32mL) were added and stirred for 3 h to form the fourth block on thepolymer. The reaction product was quenched by precipitation inmethanol. The polymer was purified over activated charcoal andfiltered to remove catalyst residue. The polymer product wasprecipitated twice in methanol and vacuum-dried at 60 °C.Nuclear Magnetic Resonance (NMR) and GPC. The polymers

were studied using 1H NMR (Bruker Avance 400 MHz instrument)with samples in CDCl3. The number-average molecular weight (Mn)and polydispersity index (Mw/Mn) of PNB-X34-Y66 were found byGPC (Shimadzu with LC-20 AD HPLC pump and a refractive indexdetector, RID-20 A, 120 V). The GPC sample was in THF with theeluent flow rate of 1.0 mL/min at 30 °C with polystyrene standard.Membrane Casting and Ion-Exchange. The tetrablock

copolymer, PNB-X34-Y66 (0.1 g), was taken up in 5 mL of chloroform.In situ cross-linking was performed by adding a cross-linking agent tothe polymer/solvent mixture when the membrane was cast, followedby reaction after casting. The cross-linking agent, TMHDA, wasadded to the solution at different mole ratios: 4 mol %, 5 mol %, 7mol %, 10 mol %, 20 mol %, and 50 mol %, with respect to the molesof brominated monomer in the polymer (i.e., those monomers whichwere capable of forming a quaternary ammonium headgroup). Thecross-linker concentration in this paper is given in terms of mol %TMHDA cross-linker added to the polymer. For example, 5 mol %TMHDA means that the up to 10% of the available headgroups areconsumed by TMHDA cross-linker. It is noted that even if all thecross-linker were to react, the fraction of intramolecular cross-linkingvs intermolecular cross-linking would be difficult to evaluate. Thesolution was filtered through a 0.45 μm poly(tetrafluoroethylene)

(PTFE) membrane syringe filter, and a film was cast and dried at 60°C for 24 h. The film was colorless, transparent, and flexible. Themembranes were aminated by immersion in 50 wt % aqueoustrimethylamine solution (48 h at room temperature). The quaternizedmembranes were washed with deionized (DI) water. The bromideions were converted to hydroxide ions by soaking the membranes in 1M NaOH solution under nitrogen for 24 h.

Hydroxide Conductivity. The membrane conductivity wasmeasured using four-point probe electrochemical impedance spec-trometry with a PAR 2273 potentiostat. The conductivity of themembranes was measured in HPLC-grade water in a nitrogenatmosphere. The membranes were allowed to sit for 30 min beforeeach measurement. The in-plane ionic conductivity was calculated, eq1.

σ = LWTR (1)

In eq 1, σ is the ionic conductivity in S/cm, and L is the lengthbetween sensing electrodes in cm. W and T are the width andthickness of the membrane in cm, respectively, and R is the resistancemeasured in ohms. The long-term (>1000 h) alkaline stability testingwas performed by immersing the membrane in 1 M NaOH solution at80 °C in a Teflon-lined Parr reactor. Prior to each measurement, themembranes were taken out of solution and thoroughly washed withDI water. After each measurement, the membranes were stored in thereactors with a freshly prepared NaOH solution. The change in ionicconductivity was used to evaluate the long-term alkaline stability.During measurement, each data point was measured in triplicate andthe average value was reported. The deviation in the measurements ofeach data point was <1%. In addition, the alkaline stability was furtheranalyzed by characterizing the chemical structure using a Nicolet 6700FT-IR spectrometer.

Ion-Exchange Capacity (IEC), Water Uptake (WU), Numberof Freezable (Nfree) and Bound Nonfreezable (Nbound) WaterMolecules, and Hydration Number (λ). 1H NMR was performedon the preaminated samples to determine the IEC of the membranes.Further, titration was used to show that the quaternization reactionwas quantitative. The titration involved converting the counteranionto chloride, followed by titration of the chloride in the membrane. Itwas previously found that IEC measurements obtained via 1H NMR(preaminated samples) and titration (postaminated samples) were thesame (within experimental error).12 For example, the IEC of PNB-X67-Y33 was found by titration and NMR, and results were 1.90 and1.92 mequiv/g, respectively.12 The fact that they match shows thateach bromoalkyl group was quantitatively converted into a quaternaryammonium headgroup. That is, each available bromoalkyl group wasreacted with trimethyl amine. 1H NMR was found to be the moredependable method and will be reported here for the materials. Thewater uptake of the membranes was calculated using eq 2.

=−

×M M

MWU (%) 100w d

d (2)

In eq 2, Md is the dry mass and Mw is the wet mass of the membraneafter removing surface water. The membranes were measured at roomtemperature in OH− form. The number of water molecules per ionicgroup (λ) was calculated using eq 3.

λ = ××

1000 WU%IEC 18 (3)

The numbers of freezable (Nfree) and bound (Nbound) water moleculesper ion pair in the membrane were found by differential scanningcalorimetry (DSC). DSC measurements were performed with aDiscovery DSC with autosampler (TA Instruments). The membraneswere hydrated, and excess water was removed from the surface. A 5−10 mg sample was sealed in a DSC pan. The sample was cooled to−50 °C and then heated to 30 °C at a rate of 5 °C/min under N2 (20mL/min). The amount of freezable and nonfreezable water wascalculated, eqs 4−6.53−55

ACS Applied Energy Materials Article

DOI: 10.1021/acsaem.8b02051ACS Appl. Energy Mater. 2019, 2, 2447−2457

2449

λ= ×NMMfree

free

tot (4)

Mfree is the mass of freezable water, and Mtot is the total mass of waterin the membrane. The weight fraction of freezable water wascalculated, eq 5.

=−

MM

H HM M M

/( )/

free

tot

f ice

w d w (5)

Mw is the wet membrane mass, and Md is the dry mass of themembrane. Hf is the enthalpy found by the integration of the DSCfreezing peak, and Hice is the enthalpy for fusion for water, correctedfor the subzero freezing point, eq 6.

= − Δ Δ°H H C Tice ice p f (6)

ΔCp is the difference between the specific heat capacity of liquid waterand ice. ΔTf is the freezing point depression.Small Angle X-ray Scattering (SAXS). SAXS was used to

analyze the phase segregation of block copolymer AEMs. Hydratedmembranes in bromide form were tested in air using the NSLS-IIbeamline at the Center for Functional Nanomaterials (BrookhavenNational Laboratory, Upton, NY). The wave vector (q) was calculatedusing eq 7, where 2θ is the scattering angle.

πθ

=ql

4sin 2 (7)

The characteristic separation length or interdomain spacing (d) (i.e.,the Bragg spacing) was calculated, eq 8.

π=dq

2(8)

Fuel Cell Testing. The XL5-PNB-X34-Y66 membrane was selectedfor electrochemical testing in an alkaline fuel cell (AEMFC). Theanode and cathode were fabricated using the slurry method asdescribed previously.12 A low molecular weight form of thepoly(BuNB-b-BPNB-b-BuNB polymer (20.5 kDa) was used as theionomer based on results from a previous investigation.56 Theionomer and 50% platinum on Vulcan XC-72 catalyst were groundtogether in isopropanol. The catalyst/ionomer slurry was thensonicated at room temperature to ensure uniform mixing. The slurrywas spray coated onto 1% water-proofed Toray TGPH-060 carbonpaper and allowed to dry at ambient temperature. The platinumloading was 2.1 mg/cm2, and the ionomer-to-carbon ratio was 40%.This metal loading was intentionally used to avoid kinetic losses in thenonoptimized electrode.The electrode−membrane assembly was soaked in 1 M NaOH for

1.5 h to exchange the bromide for hydroxide. A Fuel CellTechnologies test station with single-pass serpentine graphite platesand PTFE gaskets was used. The tests were performed on a Scribner850e Fuel Cell Test Station operated at 60 °C using humidified H2and O2 gases, each at 0.5 L/min. The dew points of the anode andcathode gas streams were adjusted during the experiments.

■ RESULTS AND DISCUSSIONSynthesis and Characterization of Tetrablock Co-

polymer, Gel Permeation Chromatography (GPC), andIon-Exchange Capacity (IEC). The catalyst, (η3-allyl)Pd-(iPr3P)Cl, and the two monomers (butyl norbornene (BuNB)and bromopropyl norbornene (BPNB)) were preparedfollowing a previous report.51,52 The catalyst was synthesizedusing (η3-allyl)Pd(iPr3P)Cl and Li[FABA] in a 1:1 ratio,sufficient for generating the cationic Pd for initiating thepolymerization. The BuNB-to-catalyst ratio was 25:1 ([M]0/[Pd] = 25). The reaction was allowed to proceed for 10 min.The second block of the tetrablock polymer was formed byadding BPNB at a monomer-to-catalyst ratio of 100-to-1([M]0/[Pd] = 100) and allowed to react for 3 h. The final two

blocks of the tetrablock polymer were synthesized by repeatingthe two steps described above.The number-average molecular weights (Mn’s) of the first,

second, third, and fourth blocks of PNB-X34-Y66 were 5.17,11.16, 5.34, and 10.68 kDa, respectively, based the GPCanalysis of the polymer samples extracted during synthesis,Figure 1. The Mn of the tetrablock copolymer was 32.35 kDawith a polydispersity index (PDI) of 2.04.

The IEC was evaluated via 1H NMR analysis by integrationof the terminal methyl protons (Ha) of the hydrophobic blockwhich resonate at 0.89 ppm and the methylene protons (Hb)adjacent to the bromine atom of the halogenated block whichappear at 3.42 ppm. 1H NMR is a more accurate means ofdetermining IEC because of the quantitative nature of solutionNMR. The integration ratio of Ha and Hb was used to calculatethe IEC (Figure 2). The Ha:Hb ratio was 1.55:2, as shown inFigure 2. This value was used to calculate the molar ratiobetween the hydrophobic and halogenated blocks byrecognition of the fact that the hydrophobic block has threemethyl protons and the halogenated block has two methyleneprotons. Using 1H NMR spectroscopy, the combined fractionof hydrophobic blocks and halogenated blocks was 34 and 66mol %, respectively. Next, the IEC was calculated from theratio of the masses of the two types of blocks. The IEC was3.46 mequiv/g for the block copolymer in hydroxide formwithout cross-linking, PNB-X34-Y66. The IEC is slightly lowerwith the addition of the cross-linking agent, as shown in Table1. The IEC values obtained by titration and NMR werepreviously found to be the same, which shows quantitativeconversion of the bromoalkyl endgroup to the quaternaryammonium form, as described above.12,34

The precursor polymer, PNB-X34-Y66, was cross-linked withN,N,N′,N′-tetramethyl-1,6-hexanediamine (TMHDA) at dif-ferent cross-linker concentrations (4−50 mol %) by adding aspecific mol % TMHDA with respect to the total moles of thehalogenated monomer in the tetrablock polymer (Scheme 1).The membranes without light cross-linking were soft and notable to be handled in membrane form without breaking. Thisresult shows one reason why the IEC of non-cross-linkedpolymers must be kept to modest values. High IEC values donot yield usable membranes. Attempts to form and handle

Figure 1. GPC trace of PNB-X34-Y66, showing the sequential growthof each block during the formation of tetrablock copolymer.

ACS Applied Energy Materials Article

DOI: 10.1021/acsaem.8b02051ACS Appl. Energy Mater. 2019, 2, 2447−2457

2450

membranes without any cross-linking failed because themembrane was too delicate. Thus, one immediate benefit oflight cross-linking was that usable high IEC membranes couldbe fabricated and tested. Table 1 includes a summary of themembrane properties.Glass Transition Temperature (Tg). Vinyl addition

poly(norbornene) copolymers are known to have a high Tg

(250 to 400 °C).47−50 The addition of flexible alkyl side-chainstends to lower the Tg of the polymer. DSC experiments wereperformed on the tetrablock copolymers used in this study.

The AEM samples were heated from 25 to 400 °C. However, aTg was not detected below the decomposition temperature ofthe polymer (<300 °C). The quaternary ammonium head-groups are known to break down at or below 250 °C.12 The Tg

of the polymer in other forms (e.g., nonquaternized form) is ofno interest because forming the ionic headgroups andabsorbing water will affect the Tg of the final polymer.

Hydroxide Conductivity (σ) and Ionic Area SpecificResistance (ASR). High hydroxide conductivity is needed formembranes used in electrochemical devices. Figure 3 shows

Figure 2. 1H NMR spectrum of tetrablock PNB-X34-Y66 in CDCl3.

Table 1. Properties of Cross-Linked Poly(butyl norbornene-b-bromopropyl norbornene-b-butyl norbornene-b-bromopropylnorbornene) (PNB-X34-Y66) Membranes in Hydroxide Form

OH−

conductivity(mS/cm)a

block copolymercross-linker conc

(mol %) 25 °C 80 °CIEC

(meq/g)b σ/IECcionic ASRd

(Ω cm2)water

uptakee (%)hydrationno. λ Nfree Nbound

d-spacing(nm)

XL50-PNB-X34-Y66 50 29.0 74.0 3.15 23.5 0.092 42.0 7.39 0.00 7.40 49.7XL20-PNB-X34-Y66 20 45.9 111 3.33 33.3 0.065 50.8 8.47 0.96 7.52 51.8XL10-PNB-X34-Y66 10 74.8 167 3.39 49.3 0.033 52.9 8.67 2.31 6.35 51.8XL7-PNB-X34-Y66 7 81.2 175 3.41 51.4 0.040 66.2 10.8 4.22 6.56 48.6XL5-PNB-X34-Y66 5 95.2 198 3.43 57.8 0.032 69.1 11.2 5.21 6.00 48.9f

XL4-PNB-X34-Y66 4 86.8 184 3.43 53.5 0.033 73.7 11.9 6.20 5.73 49.0aOH− conductivity was measured by four-probe conductivity cell. bIEC was determined by 1H NMR cIonic conductivity at 80 °C/IEC. dIonicASR was calculated using the following equation: ASR = L/σ where L = film thickness in cm and σ = ion conductivity in S/cm (at 80 °C). eWateruptake was measured at room temperature. XL = cross-linked; PNB = polynorbornene. fd-spacing was estimated via linear interpolation. X =hydrophobic block; Y = halogenated block. Numbers in the subscript indicate the molar ratio of each block.

ACS Applied Energy Materials Article

DOI: 10.1021/acsaem.8b02051ACS Appl. Energy Mater. 2019, 2, 2447−2457

2451

the increase in conductivity with temperature from 25 to 80°C. This phenomenon is due to greater thermal motion of theions at elevated temperature.57 The apparent activation energy(Ea) was estimated from the slope of ln(σ) vs 1/T and foundto be 11.7 to 14.9 kJ/mol (Figure S1). The Ea values arecomparable to previously reported high performance AEMsand PEMs, such as Nafion-117.58−60

Figure 4 shows the effect of cross-linker concentration onionic conductivity. A high degree of cross-linking can stabilizepolymer membranes and inhibit excess water swelling, butoften at the expense of ion mobility. It was observed that the

ionic conductivity increased slightly with light cross-linking, 5mol % cross-linker concentration, and then decreased at highercross-linking. For XL5-PNB-X34-Y66 (5 mol % cross-linkerconcentration), the hydroxide conductivity at 25 and 80 °Cwas 95.2 and 198 mS/cm, respectively. XL50-PNB-X34-Y66(50 mol % cross-linker concentration) had lower ionicconductivity: 29 and 74 mS/cm at 25 and 80 °C, respectively.Without cross-linking, the water uptake was so high that stablefilms could not be made because of excessive swelling. Thetrend can be seen in the slightly lower conductivity and higher

Scheme 1. Synthesis of Cross-Linked AEMs

Figure 3. Plot of ionic conductivity of XL AEMs at differenttemperatures. Figure 4. Variation of hydroxide ion conductivity with cross-linker

concentration.

ACS Applied Energy Materials Article

DOI: 10.1021/acsaem.8b02051ACS Appl. Energy Mater. 2019, 2, 2447−2457

2452

water uptake with the 4% cross-linker sample, although the 4%and 5% samples are quite close in water uptake values. At highcross-linker concentration, the membrane is too tightly cross-linked, which limits mechanical deformation and inhibits ionmobility.43

The hydroxide conductivity normalized by the IEC (σ/ΙΕC)is representative of the hydroxide ion mobility in themembrane. It measures the average effectiveness of the cationswithin the membrane to contribute to hydroxide conduction.Since the IEC value of each membrane is approximately thesame (the only mass change is due to the added cross-linker),the hydroxide mobility tracks with conductivity. The data inTable 1 shows that XL5-PNB-X34-Y66 had the highestefficiency whereas XL50-PNB-X34-Y66 displayed the lowestefficiency.Finally, ionic ASR is a key membrane metric. On the basis of

the polymer conductivity at 80 °C and membrane thickness,ionic ASR was calculated using the following equation: ASR =L/σ, where L is the film thickness and σ is the ion conductivity.The ASR of the XL5-PNB-X34-Y66 membrane is 0.032 ohmcm2 which meets the ARPA-E IONICS (Department ofEnergy, USA) target of ≤0.04 ohm cm2. The ionic ASR valuesfor the other membrane samples can be found in Table 1.Water Uptake (WU), Hydration Number (λ), and

Numbers of Freezable (Nfree) and Bound Nonfreezable(Nbound) Water Molecules. For each polymer there is anoptimum amount of water uptake needed for ion hydrationand efficient channel conduction. Excess water in the form offree water can lead to overswelling of the ion conductionchannels and poor performance due to membrane softeningand channel flooding. As shown in Table 1, the WU of themembranes had a power law relationship with cross-linkerconcentration. The best performing membrane, XL5-PNB-X34-Y66, had 69.1% WU with a conductivity of 198 mS/cm at80 °C. At a slightly lower cross-linker concentration, XL4-PNB-X34-Y66, the membrane had slightly higher WU (73.7%)and also lower conductivity (184 mS/cm at 80 °C). Themembrane with the highest cross-linker concentration, XL50-PNB-X34-Y66, had the lowest WU (42%) and also the lowestconductivity (74 mS/cm at 80 °C) due to poor ion mobility.

The number of water molecules per ion pair (headgroup andmobile counterion) and hydration number (λ) can be furtherparsed into bound or nonfreezable (Nbound) water andunbound or freezable (Nfree) water. The amount of each canbe determined in DSC freezing point measurements. As shownin Table 1, the hydration numbers for the samples testedincreased with decreasing cross-linker concentration, similar tothe WU. The bound water was calculated by subtracting thefree water from hydration number. The results of all themembranes are given in Table 1. All of the membrane samples,regardless of their conductivity, had 6−7 bound watermolecules per ion pair, while the number of free watermolecules ranged from 0 to 6.20 per ion pair. The membranewith highest conductivity (XL5-PNB-X34-Y66, 198 mS/cm at80 °C) had Nfree = 5.21 water molecules and Nbound = 6.00water molecules per ion pair. On the other hand, the worstperforming membrane (XL50-PNB-X34-Y66, 70.4 mS/cm at 80°C) had Nfree = 0.00 and Nbound = 7.40 water molecules per ionpair in the membrane. Representative DSC curves for sampleswith high and low cross-linking concentration are shown in theSupporting Information (Figures S2 and S3). Enthalpyintegrations of the peak in the cooling curves are annotatedin Figures S2 and S3. These peaks were used for calculating thenumber of freezable waters.As shown in a previous report for PNB-X54-Y46, water is able

to freely populate high IEC membranes without cross-linking,and the number of Nfree and Nbound water molecules can be aslarge as 10.6 and 17.9, respectively.12 At a high cross-linkdensity, it is increasingly difficult for water molecules,especially free water, to populate the membranes because ofthe lack of flexibility within the tightly cross-linked membrane.For un-cross-linked membranes that had less than 6 free watermolecules per ion pair, the conductivity was less than 70 mS/cm.12 The conductivity of XL50-PNB-X34-Y66 was low due tothe lack of free water. This shows that some free water isessential for channel hydration and high ion mobility. It is alsonoted that the number of free waters decreased with highercross-link density, which is likely due to the restrictedconnectivity between hydrophilic domains. It is also notedthat the domain distance was little changed for the sampleswith different cross-link density. Hence, an optimization of free

Figure 5. Alkaline stability of cross-linked AEMs in 1 M NaOH solution at 80 °C. Monitoring the drop in OH− conductivity over time (left). FT-IR spectra of XL10-PNB-X34-Y66 for characterization of the chemical structure (right).

ACS Applied Energy Materials Article

DOI: 10.1021/acsaem.8b02051ACS Appl. Energy Mater. 2019, 2, 2447−2457

2453

and bound water molecules in the membranes is necessary forobtaining maximum efficiency.11,12,61,62

Alkaline Stability. The long-term alkaline resistance ofAEMs is a concern, especially for electrochemical devicesoperating for thousands of hours at high pH and temperature.The stability measurements were performed by soaking themembranes in freshly prepared 1 M NaOH solution at 80 °Cfor more than 1000 h. The conductivity was measuredperiodically during the aging process. It was found that theAEMs lost between 1.22% and 1.40% over the >1000 h agingperiod, as shown in Figure 5 (left). Each data point in Figure 5is the average of three individual measurements. There was<1% deviation between the individual measurements of eachdata point. The three measurements varied only in the thirdsignificant figure. This value of conductivity loss is low andacceptable for ARPA-E IONICS targets. The conductivity lossfor PNB-X62-Y38 without cross-linking was 0.81% compared to1.22% loss here for XL10-PNB-X34-Y66 with 10 mol % cross-linker.12 Although the loss in conductivity is slightly higher forthe samples tested here compared to the values reportedpreviously for poly(norbornene) AEMs without cross-linking,the cross-linked AEMs in this study have significantly higherIEC and less water uptake per ion pair making the pH insidethe membranes higher than for un-cross-linked samples.12 Thatis, although all the samples were soaked in 1 M NaOH, thelocal hydroxide concentration within the membranes isdifferent. The hydroxide concentration inside the membraneis better described by λ−1, the number of hydroxide ions perwater molecule. This comparison shows that the hydroxide is68% more concentrated in the cross-linked AEMs than in theprevious un-cross-linked AEMs. A lower WU per ion paircreates a more alkaline environment within the membrane,which can accelerate the degradation. Additional attempts weremade to analyze the AEMs for hydroxide degradation. FT-IRanalysis of the chemical structure before and after the alkalineaging showed no new peaks, Figure 5 (right). The C−Nstretching frequencies at 838, 913, 968, and 1060 cm−1 showthat the chemical structure of the membrane is indeedintact.63,64 However, it is noted that FT-IR is only semi-quantitative, and structural changes at the 1% level are difficultto analyze by this method. It is also noted that quantitativesolution NMR analysis is not possible because the samples areno longer soluble due to the cross-linking.Morphological Characterization. SAXS was used to

investigate the phase segregation and microstructure of thecross-linked poly(norbornene) membranes. During casting, thepoly(norbornene) block copolymers phase segregate into ionconduction channels based on the thermodynamic dissim-ilarities between the halogenated and hydrophobic blocks.Once cross-linking begins to occur, the cross-linking agentfurther limits the self-assembly, thereby locking in themicrostructure of the membrane upon curing. The inter-domain spacing (d-spacing), or the average separation lengthbetween inhomogeneities in the membranes, was determinedfrom the Bragg spacing of the primary scattering peak in theSAXS spectra, as shown in Figure 6. The interdomain spacingvalues were determined by a Lorentz curve fitting function andare listed in Table 1. It was found that the domain distances forthe polymers with different cross-link densities were all similar,ranging only from 48.6 to 51.8 nm. This is unlike the domaindistance for the membranes without cross-linking, which areunhindered and can range from 37.2 to 86.4 nm.12

Fuel Cell Testing. The XL5-PNB-X34-Y66 was chosen fordemonstration is an alkaline fuel cell because it had high ionicconductivity and chemical stability. The membrane wasmechanically robust and easily assembled into the fuel cellhardware. The fuel cell tests were performed at 60 °C, which isa common operating temperature. The fuel cell was firstconditioned at a cell voltage of 0.5 V for 1 h followed by 1 h at0.2 V. After conditioning, the open circuit voltage (OCV) was1.042 V. A current−voltage voltammogram and impedancespectrum at 0.4 V were periodically recorded. The dew pointsof the anode and cathode feed gases were set at 52 °C (i.e.,74.8% RH) and 56 °C (i.e., 86.6% RH), respectively.Figure 7 shows the cell voltage and power as a function of

the current density. The peak power density was 510 mW/cm2

at 0.534 V and 954 mA/cm2. Figure 7 also shows the voltage asa function of current density after correction for the iR drop

Figure 6. SAXS spectra of cross-linked tetrablock copolymerpoly(BuNB-b-BPNB-b-BuNB-b-BPNB) membranes in bromide form.

Figure 7. Polarization data for XL5-PNB-X34-Y66 AEMFC (2.1 mgcm−2 Pt, 40% ionomer to carbon ratio) with and without iRcorrection. Cell temperature was 60 °C with anode and cathode dewpoints both set at 52 and 56 °C, respectively. Flow rates of humidifiedH2 and O2 were both 0.5 L/min.

ACS Applied Energy Materials Article

DOI: 10.1021/acsaem.8b02051ACS Appl. Energy Mater. 2019, 2, 2447−2457

2454

across the membrane. The membrane resistance was 95 mΩcm2, as obtained from the high frequency intercept of theimpedance plot. The iR corrected cell power was 573 mW/cm2. While the power output is modest compared to the 2 W/cm2 cell reported by Wang et al., the membrane electrodeassembly used here was not optimized for ionomer and catalystlayer content, the operating temperature was lower (60 °Chere vs 80 °C by Wang et al.), and the membrane was thickerthan necessary, 64 μm.45 These fuel cell results show that themembrane can be successfully integrated into a workingelectrochemical device. Improvements in the electrodeassembly and membrane thickness are underway and will bethe subject of future publications.

■ SUMMARYCross-linked anion conductive polymers synthesized via vinyladdition polymerization of norbornenes were studied. Themembranes had very high ionic conductivity, up to 198 mS/cmat 80 °C. It was found that only light cross-linking was neededto mitigate water swelling problems which have plagued otherhigh IEC AEMs. The cross-linking was light enough so as notto cause problems encountered by highly cross-linked AEMs.The dimensional stability was attributed to cross-link density,which can be tuned to balance the free and bound watercontent within the membrane. There were 5.21 free watermolecules and 6.00 bound water molecules per ion pair withinthe optimized membrane. Excellent alkaline stability in 1 MNaOH solution at 80 °C was demonstrated (<1.5%conductivity loss in >1000 h at 80 °C). The hydrogen/oxygenfuel cell tests at 60 °C had a peak power density of 510 mW/cm2 at 0.534 V and 954 mA/cm2. These tetrablock copolymerswith very high ionic conductivity and alkaline resiliency areexcellent candidates for high performance electrochemicaldevices.

■ ASSOCIATED CONTENT*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acsaem.8b02051.

Synthesis scheme of ROMP homopolymer and blockcopolymer, Arrhenius plot of ln(σ) vs inverse temper-ature, DSC curve with enthalpy integration from XL5-PNB-X34-Y66 and XL20-PNB-X34-Y66, and TEM micro-graph (PDF)

■ AUTHOR INFORMATIONCorresponding Author*E-mail: kohl@gatech.edu. Phone: 404-894-2893.ORCIDMrinmay Mandal: 0000-0002-3404-9588Paul A. Kohl: 0000-0001-6267-3647NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSThe authors gratefully acknowledge the financial support of theARPA-E IONICS (United States Department of Energy)program and the helpful discussions of Dr. Edmund Elce. Wewould also like to thank Dr. Sungmin Park for his invaluablehelp with SAXS measurements at Brookhaven NationalLaboratory.

■ REFERENCES(1) Mekhilef, S.; Saidur, R.; Safari, A. Comparative Study ofDifferent Fuel Cell Technologies. Renewable Sustainable Energy Rev.2012, 16, 981−989.(2) Carrette, L.; Friedrich, K. A.; Stimming, U. Fuel Cells-Fundamentals and Applications. Fuel Cells 2001, 1, 5−39.(3) Winter, M.; Brodd, R. J. What Are Batteries, Fuel Cells, andSupercapacitors? Chem. Rev. 2004, 104, 4245−4269.(4) Steele, B. C. H.; Heinzel, A. Materials for Fuel-CellTechnologies. Nature 2001, 414, 345−352.(5) Varcoe, J. R.; Atanassov, P.; Dekel, D. R.; Herring, A. M.;Hickner, M. A.; Kohl, P. A.; Kucernak, A. R.; Mustain, W. E.;Nijmeijer, K.; Scott, K.; Xu, T. W.; Zhuang, L. Anion-exchangemembranes in electrochemical energy systems. Energy Environ. Sci.2014, 7, 3135−3191.(6) Lu, S.; Pan, J.; Huang, A.; Zhuang, L.; Lu, J. Alkaline polymerelectrolyte fuel cells completely free from noble metal catalysts. Proc.Natl. Acad. Sci. U. S. A. 2008, 105, 20611−20614.(7) Yu, E. H.; Wang, X.; Krewer, U.; Li, L.; Scott, K. Direct oxidationalkaline fuel cells: from materials to systems. Energy Environ. Sci. 2012,5, 5668−5680.(8) Hickner, M. A.; Herring, A. M.; Coughlin, E. B. Anion exchangemembranes: Current status and moving forward. J. Polym. Sci., Part B:Polym. Phys. 2013, 51, 1727−1735.(9) Zhou, J.; Joseph, K.; Ahlfield, J. M.; Park, D. Y.; Kohl, P. A.Poly(arylene ether) Ionomers with Pendant Quinuclidium Groupsand Varying Molecular Weight for Alkaline Electrodes. J. Electrochem.Soc. 2013, 160, F573−F578.(10) Mohanty, A. D.; Bae, C. Mechanistic analysis of ammoniumcation stability for alkaline exchange membrane fuel cells. J. Mater.Chem. A 2014, 2, 17314−17320.(11) Liu, L.; Ahlfield, J.; Tricker, A.; Chu, D.; Kohl, P. A. Anionconducting multiblock copolymer membranes with partial fluorina-tion and long head-group tethers. J. Mater. Chem. A 2016, 4, 16233−16244.(12) Mandal, M.; Huang, G.; Kohl, P. A. Anionic MultiblockCopolymer Membrane Based on Vinyl Addition Polymerization ofNorbornenes: Application in Anion-Exchange Membrane Fuel Cells.J. Membr. Sci. 2019, 570−571, 394−402.(13) Mohanty, A. D.; Tignor, S. E.; Sturgeon, M. R.; Long, H.;Pivovar, B. S.; Bae, C. Thermochemical Stability Study of Alkyl-Tethered Quaternary Ammonium Cations for Anion ExchangeMembrane Fuel Cells. J. Electrochem. Soc. 2017, 164, F1279−F1285.(14) Arges, C. G.; Zhang, L. Anion Exchange Membranes’ EvolutionToward High Hydroxide Ion Conductivity and Alkaline Resiliency.ACS Appl. Energy Mater. 2018, 1, 2991−3012.(15) Zhang, X.; Shi, Q.; Chen, P.; Zhou, J.; Li, S.; Xu, H.; Chen, X.;An, Z. Block poly(arylene ether sulfone) copolymers tetheringaromatic side-chain quaternary ammonium as anion exchangemembranes. Polym. Chem. 2018, 9, 699−711.(16) Shi, Q.; Chen, P.; Zhang, X.; Weng, Q.; Chen, X.; An, Z.Synthesis and properties of poly(arylene ether sulfone) anionexchange membranes with pendant benzyl-quaternary ammoniumgroups. Polymer 2017, 121, 137−148.(17) Weiber, E. A.; Meis, D.; Jannasch, P. Anion conductingmultiblock poly(arylene ether sulfone)s containing hydrophilicsegments densely functionalized with quaternary ammonium groups.Polym. Chem. 2015, 6, 1986−1996.(18) Akiyama, R.; Yokota, N.; Nishino, E.; Asazawa, K.; Miyatake, K.Anion Conductive Aromatic Copolymers from Dimethylaminome-thylated Monomers: Synthesis, Properties, and Applications inAlkaline Fuel Cells. Macromolecules 2016, 49, 4480−4489.(19) Shimada, M.; Shimada, S.; Miyake, J.; Uchida, M.; Miyatake, K.Anion conductive aromatic polymers containing fluorenyl groups:Effect of the position and number of ammonium groups. J. Polym. Sci.,Part A: Polym. Chem. 2016, 54, 935−944.(20) Fujimoto, C.; Kim, D. S.; Hibbs, M.; Wrobleski, D.; Kim, Y. S.Backbone stability of quaternized polyaromatics for alkalinemembrane fuel cells. J. Membr. Sci. 2012, 423−424, 438−449.

ACS Applied Energy Materials Article

DOI: 10.1021/acsaem.8b02051ACS Appl. Energy Mater. 2019, 2, 2447−2457

2455

(21) Nunez, S. A.; Hickner, M. A. Quantitative 1H NMR Analysis ofChemical Stabilities in Anion-Exchange Membranes. ACS Macro Lett.2013, 2, 49−52.(22) Arges, C. G.; Ramani, V. Two-dimensional NMR spectroscopyreveals cation-triggered bacbone degradation in polysulfone-basedanion exchange membranes. Proc. Natl. Acad. Sci. U. S. A. 2013, 110,2490−2495.(23) Long, H.; Kim, K.; Pivovar, B. S. Hydroxide DegradationPathways for Substituted Trimethylammonium Cations: A DFTStudy. J. Phys. Chem. C 2012, 116, 9419−9426.(24) Lee, W. H.; Mohanty, A. D.; Bae, C. Fluorene-Based HydroxideIon Conducting Polymers for Chemically Stable Anion ExchangeMembrane Fuel Cells. ACS Macro Lett. 2015, 4, 453−457.(25) Ono, H.; Kimura, T.; Takano, A.; Asazawa, K.; Miyake, J.;Inukai, J.; Miyatake, K. Robust anion conductive polymers containingperfluoroalkylene and pendant ammonium groups for high perform-ance fuel cells. J. Mater. Chem. A 2017, 5, 24804−24812.(26) Lee, W. H.; Kim, Y. S.; Bae, C. Robust Hydroxide IonConducting Poly(biphenyl alylene)s for Alkaline Fuel Cell Mem-branes. ACS Macro Lett. 2015, 4, 814−818.(27) Park, D. Y.; Kohl, P. A.; Beckham, H. W. Anion-ConductiveMultiblock Aromatic Copolymer Membranes: Structure-PropertyRelationships. J. Phys. Chem. C 2013, 117, 15468−15477.(28) Chen, X. C.; Wong, D. T.; Yakovlev, S.; Beers, K. M.; Downing,K. H.; Balsara, N. P. Effect of Morphology of Nanoscale HydratedChannels on Proton Conductivity in Block Copolymer ElectrolyteMembranes. Nano Lett. 2014, 14, 4058−4064.(29) Inceoglu, S.; Rojas, A. A.; Devaux, D.; Chen, X. C.; Stone, G.M.; Balsara, N. P. Morphology−Conductivity Relationship of Single-Ion-Conducting Block Copolymer Electrolytes for Lithium Batteries.ACS Macro Lett. 2014, 3, 510−514.(30) Sun, J.; Liao, X.; Minor, A. M.; Balsara, N. P.; Zuckermann, R.N. Morphology-Conductivity Relationship in Crystalline andAmorphous Sequence-Defined Peptoid Block Copolymer Electro-lytes. J. Am. Chem. Soc. 2014, 136, 14990−14997.(31) Sun, J.; Jiang, X.; Siegmund, A.; Connolly, M. D.; Downing, K.H.; Balsara, N. P.; Zuckermann, R. N. Morphology and ProtonTransport in Humidified Phosphonated Peptoid Block Copolymers.Macromolecules 2016, 49, 3083−3090.(32) Price, S. C.; Ren, X.; Jackson, A. C.; Ye, Y.; Elabd, Y. A.; Beyer,F. L. Bicontinuous Alkaline Fuel Cell Membranes from Strongly Self-Segregating Block Copolymers. Macromolecules 2013, 46, 7332−7340.(33) Meek, K. M.; Sharick, S.; Ye, Y.; Winey, K. I.; Elabd, Y. A.Bromide and Hydroxide Conductivity−Morphology Relationships inPolymerized Ionic Liquid Block Copolymers. Macromolecules 2015,48, 4850−4862.(34) He, X.; Liu, J.; Zhu, H.; Zheng, Y.; Chen, D. Novel QuaternaryAmmonium Functional Addition-Type Norbornene Copolymer asHydroxide-Conductive and Durable Anion Exchange Membrane forDirect Methanol Fuel Cells. RSC Adv. 2015, 5, 63215−63225.(35) Kim, D.-G.; Takigawa, R.; Kashino, T.; Burtovyy, O.; Bell, A.;Register, R. A. Hydroxyhexafluoroisopropylnorbornene Block andRandom Copolymers via Vinyl Addition Polymerization and TheirApplication as Biobutanol Pervaporation Membranes. Chem. Mater.2015, 27, 6791−6801.(36) He, S. Q.; Liu, L.; Wang, X. C.; Zhang, S. F.; Guiver, M. D.; Li,N. W. Azide-Assisted Self-Crosslinking of Highly Ion ConductiveAnion Exchange Membranes. J. Membr. Sci. 2016, 509, 48−56.(37) Xu, W.; Zhao, Y.; Yuan, Z.; Li, X.; Zhang, H.; Vankelecom, I. F.J. Highly Stable Anion Exchange Membranes with Internal Cross-linking Networks. Adv. Funct. Mater. 2015, 25, 2583−2589.(38) Wang, J.; Wang, J. H.; Zhang, S. B. Synthesis andCharacterization of Crosslinked Poly(Arylene Ether Ketone)Containing Pendant Quaternary Ammonium Groups for Anion-Exchange Membranes. J. Membr. Sci. 2012, 415−416, 205−212.(39) Tibbits, A. C.; Mumper, L. E.; Kloxin, C. J.; Yan, Y. S. A Single-Step Monomeric Photo-Polymerization and Crosslinking via Thiol-Ene Reaction for Hydroxide Exchange Membrane Fabrication. J.Electrochem. Soc. 2015, 162, F1206−F1211.

(40) Gu, S.; Cai, R.; Yan, Y. S. Self-Crosslinking for DimensionallyStable and Solvent-Resistant Quaternary Phosphonium BasedHydroxide Exchange Membranes. Chem. Commun. 2011, 47, 2856−2858.(41) Zhu, L.; Zimudzi, T. J.; Li, N.; Pan, J.; Lin, B.; Hickner, M. A.Crosslinking of comb-shaped polymer anion exchange membranes viathiol-ene click chemistry. Polym. Chem. 2016, 7, 2464−2475.(42) Cheng, J.; Yang, G.; Zhang, K.; He, G.; Jia, J.; Yu, H.; Gai, F.;Li, L.; Hao, C.; Zhang, F. Guanidimidazole-quanternized and cross-linked alkaline polymer electrolyte membrane for fuel cell application.J. Membr. Sci. 2016, 501, 100−108.(43) Hu, E. N.; Lin, C. X.; Liu, F. H.; Yang, Q.; Li, L.; Zhang, Q. G.;Zhu, A. M.; Liu, Q. L. Cross-Linked Poly(vinylbenzyl chloride) AnionExchange Membranes with Long Flexible Multihead for Fuel Cells.ACS Appl. Energy Mater. 2018, 1, 3479−3487.(44) Zhang, W.; Liu, Y.; Liu, X.; Horan, J. L.; Jin, Y.; Ren, X.; Ertem,S. P.; Seifert, S.; Liberatore, M. W.; Herring, A. M.; Coughlin, E. B.Crosslinked anion exchange membranes with connected cations. J.Polym. Sci., Part A: Polym. Chem. 2018, 56, 618−625.(45) Wang, L.; Bellini, M.; Miller, H. A.; Varcoe, J. R. A highconductivity ultrathin anion-exchange membrane with 500+ h alkalistability for use in alkaline membrane fuel cells that can achieve 2 Wcm−2 at 80 °C. J. Mater. Chem. A 2018, 6, 15404−15412.(46) Chen, W.; Mandal, M.; Huang, G.; Wu, X.; He, G.; Kohl, P. A.Highly Conductive Anion-Exchange Membranes Based on Cross-linked Poly(norbornene): Ring Opening Metathesis Polymerization.ACS Appl. Energy Mater. 2019, DOI: 10.1021/acsaem.8b02052.(47) Grove, N. R.; Kohl, P. A.; Bidstrup Allen, S. A.; Jayaraman, S.;Shick, R. Functionalized Polynorbornene Dielectric Polymers:Adhesion and Mechanical Properties. J. Polym. Sci., Part B: Polym.Phys. 1999, 37, 3003−3010.(48) Shin, B.-G.; Cho, T.-Y.; Yoon, D. Y.; Liu, B. Structure andProperties of Polynorbornene Derivatives: Poly(norbornene dicarbox-ylic acid dialkyl ester)s and Poly(norbornene dimethyl dicarboxylate)-s. Macromol. Res. 2007, 15, 185−190.(49) Tetsuka, H.; Isobe, K.; Hagiwara, M. Synthesis and Propertiesof Addition-Type Poly(norbornene)s with Siloxane Substituents.Polym. J. 2009, 41, 643−649.(50) Dorkenoo, K. D.; Pfromm, P. H.; Rezac, M. E. Gas TransportProperties of a Series of High Tg Polynorbornenes with AliphaticPendant Groups. J. Polym. Sci., Part B: Polym. Phys. 1998, 36, 797−803.(51) Martínez-Arranz, S.; Albeniz, A. C.; Espinet, P. Versatile Routeto Functionalized Vinylic Addition Polynorbornenes. Macromolecules2010, 43, 7482−7487.(52) Lipian, J.; Mimna, R. A.; Fondran, J. C.; Yandulov, D.; Shick, R.A.; Goodall, B. L.; Rhodes, L. F.; Huffman, J. C. AdditionPolymerization of Norbornene-Type Monomers. High ActivityCationic Allyl Palladium Catalysts. Macromolecules 2002, 35, 8969−8977.(53) Lue, S. J.; Shieh, S. J. Water States in Perfluorosulfonic AcidMembranes Using Differential Scanning Calorimetry. J. Macromol.Sci., Part B: Phys. 2009, 48, 114−127.(54) Mecheri, B.; Felice, V.; Zhang, Z.; D’Epifanio, A.; Licoccia, S.;Tavares, A. C. DSC and DVS Investigation of Water Mobility inNafion/Zeolite Composite Membranes for Fuel Cell Applications. J.Phys. Chem. C 2012, 116, 20820−20829.(55) Moster, A. L.; Mitchell, B. S. Hydration and proton conductionin Nafion/ceramic nanocomposite membranes produced by solid-state processing of powders from mechanical attrition. J. Appl. Polym.Sci. 2009, 113, 243−250.(56) Ahlfield, J.; Huang, G.; Liu, L.; Kaburagi, Y.; Kim, Y.; Kohl, P.A. Anion Conducting Ionomers for Fuel Cells and Electrolyzers. J.Electrochem. Soc. 2017, 164, F1648−F1653.(57) Li, Q.; Liu, L.; Miao, Q.; Jin, B.; Bai, R. A novel poly(2,6-dimethyl-1,4-phenylene oxide) with trifunctional ammonium moietiesfor alkaline anion exchange membranes. Chem. Commun. 2014, 50,2791−2793.

ACS Applied Energy Materials Article

DOI: 10.1021/acsaem.8b02051ACS Appl. Energy Mater. 2019, 2, 2447−2457

2456

(58) Dang, H.-S.; Jannasch, P. Exploring Different Cationic AlkylSide Chain Designs for Enhanced Alkaline Stability and HydroxideIon Conductivity of Anion-Exchange Membranes. Macromolecules2015, 48, 5742−5751.(59) Pan, J.; Lu, S.; Li, Y.; Huang, A.; Zhuang, L.; Lu, J. High-Performance Alkaline Polymer Electrolyte for Fuel Cell Applications.Adv. Funct. Mater. 2010, 20, 312−319.(60) Lin, B.; Qiu, L.; Lu, J.; Yan, F. Cross-Linked Alkaline IonicLiquid-Based Polymer Electrolytes for Alkaline Fuel Cell Applications.Chem. Mater. 2010, 22, 6718−6725.(61) Liu, L.; Kohl, P. A. Anion Conducting Multiblock Copolymerswith Different Tethered Cations. J. Polym. Sci., Part A: Polym. Chem.2018, 56, 1395−1403.(62) Liu, L.; Huang, G.; Kohl, P. A. Anion conducting multiblockcopolymers with multiple headgroups. J. Mater. Chem. A 2018, 6,9000−9008.(63) Amel, A.; Yitzhack, N.; Beylin, A.; Pan, J.; Hickner, M. A.; Ein-Eli, Y. Chemical and Thermal Stability of Poly(phenylene oxide)-Based Anion Exchange Membranes Containing Alkyl Side Chains. J.Electrochem. Soc. 2018, 165, F1133−F1138.(64) Amel, A.; Smedley, S. B.; Dekel, D. R.; Hickner, M. A.; Ein-Eli,Y. Characterization and Chemical Stability of Anion ExchangeMembranes Cross-Linked with Polar Electron-Donating Linkers. J.Electrochem. Soc. 2015, 162, F1047−F1055.

ACS Applied Energy Materials Article

DOI: 10.1021/acsaem.8b02051ACS Appl. Energy Mater. 2019, 2, 2447−2457

2457