Interplay Between Hydroxyl Density and Relaxations in Poly

Jan 9, 2018 - Those available include the contrasting works on a soft non-cross-linked membrane based on the commercial FAA-3 chemistry designed for a...
0 downloads 7 Views 2MB Size
Subscriber access provided by University of Florida | Smathers Libraries

Article

Interplay Between Hydroxyl Density and Relaxations in poly(vinylbenzyltrimethyl ammonium)-b-poly(methylbutylene) Membranes for Electrochemical Applications Keti Vezzu', Ashley M. Maes, Federico Bertasi, Andrew Motz, TsungHan Tsai, E. Bryan Coughlin, Andrew M Herring, and Vito Di Noto J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b10524 • Publication Date (Web): 09 Jan 2018 Downloaded from http://pubs.acs.org on January 9, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Journal of the American Chemical Society is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 14 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Interplay Between Hydroxyl Density and Relaxations in poly(vinylbenzyltrimethyl ammonium)-b-poly(methylbutylene) Membranes for Electrochemical Applications Keti Vezzù,a,b‡ Ashley M. Maes,c‡ Federico Bertasi,a,d Andrew R. Motz,c Tsung-Han Tsai,e E. Bryan Coughlin,e Andrew M. Herring,c* and Vito Di Noto.a,b,d,f* a

Section of Chemistry for Technology, Department of Industrial Engineering, University of Padova, in Department of Chemical Sciences, Via Marzolo 1, I-35131 Padova, Italy. b Consorzio Interuniversitario Nazionale per la Scienza e Tecnologia dei Materiali (INSTM), d Interdepartmental Centre Giorgio Levi Cases for Energy Economics and Technology, Via Marzolo 9, I-35131 Padova, Italy c Department of Chemical and Biological Engineering, Colorado School of Mines, Golden, CO 80401-1887 USA. e Department of Polymer Science and Engineering, University of Massachusetts Amherst, 120 Governors Drive, Amherst, MA 01003 USA. f

Material Science and Engineering Department, University of Carlo III , Av. De la Universidad, 30, 28911 Leganés, Madrid, Spain

ABSTRACT: Anion-exchange membranes (AEMs) consisting of poly (vinyl benzyl trimethyl ammonium)-b-poly (methylbutylene) of three different ion exchange capacities (IECs), 1.14, 1.64, and 2.03 meq g -1, are studied by High-Resolution Thermogravimetry, Modulated Differential Scanning Calorimetry, Dynamic Mechanical Analysis, and Broadband Electrical Spectroscopy in their OHform. The thermal stability and transitions are elucidated, showing a low temperature Tg and a higher temperature transition assigned to a disorder-order transition, T, which depends on the IEC of the material. The electric response is analyzed in detail, allowing the identification of three polarizations (only two of which contribute significantly to the overall conductivity, EP and IP,1) and two dielectric relaxation events (β1 and β2), one associated with the tolyl side groups (β1) and one with the cationic side chains (β2). The obtained results are integrated in a coherent picture and a conductivity mechanism is proposed, involving two distinct conduction pathways, EP and IP,1. Importantly, we observed a reordering of the ion pair dipoles which is responsible for the T at temperatures higher than 20°C, which results in a dramatic decrease of the ionic conductivity. Clustering is highly implicated in the higher IEC membrane in the hydroxyl form, which reduces the efficiency of the anionic transport.

1. INTRODUCTION Anion exchange membranes (AEMs) are well established as separators and membranes, but their use in electrochemical devices has been limited in the past by their low anionic conductivity and poor chemical and mechanical durability. From 2005, when Varcoe and Slade first proposed that purpose designed AEMs could be used in an AEM fuel cell, 1 interest has rapidly grown in the use of these materials in a variety of electrochemical energy conversion devices such as fuel cells, electrolyzers, or redox flow cells.2 For these devices to work at their highest power densities it is necessary to design materials in which anion transport is maximized and water transport optimized for the particular application. The catholyte and anolyte in a redox flow battery are aqueous and the membrane should maintain the concentration of these on both sides, in an electrolyzer water transport should be minimized, as removing water from the produced hydrogen is expensive, and in a fuel cell where water is a reactant on the cathode for system simplicity it should be provided by back diffusion to the cathode from the anode where it is a product. For these reasons, fundamental studies of anion and water transport and their relation to the chemistry and morphology of the polymer are extremely important so that polymer design for next generation AEMs can incorporate this learning. To date there are a limited number of detailed studies on anion and water transport in AEMs and very few have appeared where

hydroxide is considered. Those available include the contrasting works on a soft non-crosslinked membrane based on the commercial FAA-3 chemistry designed for a fundamental study,3 a rigid electron-beam-benzyl trimethyl ammonium grafted ETFE material designed to be a practical material with high anionic conductivity with little swelling,4 and a block copolymer polyphenyleneoxide-b-poly[vinylbenzyltrimethylammonium] (PPO-b-PVBTMA) that phase separates in to a biscontinuous phase.5 This research represents some of the first values reported for pure hydroxide conductivity in humidified CO2 free gases. To date the established conclusion is that hydroxide is fully dissociated with adequate hydration and that differing amounts of water are taken up by the film depending on the counter anion. This gives rise to different ionic conductivities at the same water activity.3 The hydroxide conductivity for the softer materials based on FAA-3 reaches a maximum of ca. 60 mS cm-1 on adequate hydration at 25 °C, and shows an ionomer peak in the SAXS that swells on a log scale to ½ indicating that the swelling is only in two dimensions.3 The hydroxide conductivity of the material based on ETFE reaches a maximum of 132 mS cm-1 at 80 °C and 95% RH and shows no peaks in the SAXS.4 The hydroxide conductivity of the PPO-bPVBTMA reaches 138 mS cm-1 at 80 °C and 95% RH similar to that of the radiation grafted ETFE material.5 These studies primarily use ionic conductivity, often in-plane, and diffusion

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 14

In our previous report on the Br- form of this material we were able to fully assign the electric response of the material and propose a mechanism for bromide conduction which involves two distinct ionic pathways.19 Importantly, we observed a reordering of the ion pair dipoles which is responsible for a disorder– order transition (T) at higher temperatures, observed for the first time in AEMs, which results in a dramatic decrease of the ionic conductivity. We wanted to extend these studies with the model polymer system to come up with some basic design rules for the synthesis of future AEMs. For this reason it was necessary a) to contrast ionic conductivity of bromide with that of a more dissociated hydroxide anion, b) understand how the interactions between side groups are correlated to the dielectric relaxations of backbone and side chains; change the IEC and so the density of the side chains to see the effect at a short and at a long distance of the interactions between polar side chains on the efficiency of ionic conductivity; c) confirm that the presence of T is not depending on the type of anion composing the ionomer, but is a general phenomenon; study the correlation between T and the density of polar side chains in order to detect if it is a potential parameter diagnostic of the thermal, chemical and electrochemical membrane stability. These objectives are easily achieved by studying the same model polymer system at three different IECs in the hydroxide form and contrasting the results obtained with the Br- form of the polymer studied previously.19 This report describes the results obtained from the detailed BES and thermoanalytical characterization of the [PVBTMA][X]-b[PMB] where X = OH- or Br-. In this study, the materials were studied at three differing degrees of functionalization, DFs in the hydroxide form and compared to the data of the same polymer in the bromide form, by means of high-resolution thermogravimetry (HR-TG), modulated differential scanning calorimetry (MDSC) and dynamic mechanical analysis (DMA). Broadband electrical spectroscopy (BES) is adopted to determine the electric response of the membranes in the fully hydrated state. Once the commonality of T in these materials was established, a similar polymer immersed in liquid water was investigated by SAXS as a function of temperature to show that cations in the polymer must be re-ordering as T was passed. The interplay between the thermoanalytical and electrical properties of the membranes is elucidated, and a mechanism for the long-range charge transport is proposed. These studies are extremely challenging as hydroxide is a strong nucleophile and tends to attack polar polymer backbones and many of the organic cations that are attached to them.21 In addition it is well known that the current level of 400 ppm of CO 2 found in the atmosphere quickly reacts with the hydroxide in the AEM, yielding carbonate and bicarbonate species in the membrane.22 The HCO3- and CO32- anions of the carbonated AEM are much larger and less mobile with respect to the OHanion.2 Consequently, carbonation has a remarkable effect on the properties of the AEM, especially in regard to the electric properties and the charge transfer mechanism. In practical applications, CO2 is reduced to low levels from the air fed to the AEMFC cathode with the specific purpose of minimizing carbonation and obtaining an optimal performance from the device. The AEMs studied in this work are manipulated in an inert, CO2-free environment; they are never exposed to the open atmosphere to prevent carbonation. All the results reported here are determined on AEMs in their OH- form and compared to a previous study in which the AEM was in the bromide form. 19

as measured by electrochemical impedance and tracer diffusion from pulse field gradient spin echo NMR to correlate transport with water uptake data and morphology inferred from SAXS. To complete the understanding of anion transport in polyelectrolytes it is extremely informative to study these materials over a very large frequency range by broadband electrical spectroscopy (BES). Because of this wide dynamic range in frequency, BES enables all of the relaxations both mechanical and ionic to be probed at the macromolecular level and interpreted in the polymer systems. Increasing the IEC of an AEM without structuring it or crosslinking does not increase anionic conductivity beyond a certain point as the ions cluster.6 The concept of using block polymers to create an advantageous nanostructured phase separation has been applied to proton conducting membranes successfully for sometime to increase proton conductivity.7 In the last 5 years the use of block polymers to also produce nano-structured morphologies for improved anion and water transport properties has seen a large increase in activity. A large variation of cationic blocks have already been utilized, with recent examples such as, poly vinylbenzyltrimethylamonium,8-10 dioxiphenylene blocks with multiple quaternary ammonium groups,11 poly 2vinyl n-methylpyridinium,12 polymerized ionic liquids,13-14 poly(2-acryloxy)ethyltributylphosphonium,15 poly(diallylpiperidinium hydroxide),16 or polymethylstyrene functionalized with poly tris(2,4,6-trimethoxyphenyl)phosphine17 paired with the usual hydrophobic blocks such as polystyrene, polybutylene, polyphenylene oxide,5 and polysulfone. These phase separated materials show enhanced anion conductivity with varying degrees of success, and so it is still necessary to further elucidate the mechanism of anion conduction in phase separated materials as a function of not just morphology, but also, the mechanical relaxations and chemistry of the material. In order to fully understand anionic and water transport in these materials we designed a model polymer system with well defined phase separated lamellar morphology, and easily variable degree of functionalization, so that the BES measurements could be related to the length scales as measured by SAXS. The AEMs studied in this report, consist of a poly(vinylbenzyltrimethylammonium)-b-poly(methylbutylene) diblock polymer with either hydroxide or bromide counter anions ([PVBTMA][X]-b-[PMB] where X = OH- or Br-).18-19 When this diblock polymer is cast from the correct solvent, i.e., chloroform, it phase separates in to a well-defined lamellar structure, which is designed to avoid formation of large clusters of anions. The anion-exchange block PVBTMA of the AEMs consists of benzyltrimethylammonium moieties. The polymer is given structural rigidity and flexibility by using the rubbery PMB as the hydrophobic block, allowing the membrane to maintain a lamellar structure while still being able to swell with water. Each membrane is distinguished by a different degree of functionalization (DF) with quaternary ammonium groups, defined as the percentage of side tolyl groups bearing an anionexchange trimethylammoinum moiety. The SAXS data for this polymer, with the block lengths described in this study, and an ion exchange capacity, IEC of 2.03 mmol g-1 has been reported confirming the lamellar morphology and giving a d spacing for the lamella from 38.9 to 42.8 nm for dry and wet conditions respectively.20 The modeling of this polymer by an entirely physical model predicts that at very high water concentrations the water in the channel may not be homogeneously distributed. 20

2 ACS Paragon Plus Environment

Page 3 of 14 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society where 𝜀∞ is the electronic contribution to the total permittivity. The first term is associated with the electrode (j=EP) and interdomain (j=IP,1 and IP,2) polarizations. j and j are the conductivity and the relaxation time associated with each polarization event. j accounts for the broadening of the jth peak. The second term accounts for the dielectric relaxations that are detected. k, k, k and βk correspond to the dielectric strength, relaxation time, symmetric and asymmetric shape parameter for each dielectric relaxation event. SAXS data was collected on an Anton Parr SAXSess instrument using a high resolution image plate. The x-ray source was a PANalytical PW3830 x-ray generator, using Cu K radition with an energy of 8.04 keV. Background scans were collected using water placed in a plastic capillary tube, which was sealed in a rotor cell. Samples of polymer were measured in liquid water in the same capillary. The temperature of the rotor cell was controlled using an Anton Parr TCU50 temperature control unit, from 20-80°C in 10°C increments. The equilibration time was 15 minutes at the desired temperature, and the exposure time was 10 minutes.

2. EXPERIMENTAL.

2.1 Materials. [PVBTMA][Br]-b-[PMB] is synthesized by the published method23 and obtained from the University of Massachusetts, Amherst at DF = 41.8%, 62.9% , and 80.7%, designated A, B, and C, respectively. 24 The final OH- exchanged A, B, and C membranes are obtained from the respective brominated precursors by immersion in 1 M NaOH for a total of 20 h; they were then rinsed with deionized water three times over the course of a further 24 h. The entire OH - exchange process was carried out inside a homemade CO2 free wet box. 2.2 Instruments and Methods. Thermogravimetric analyses are carried out with a high-resolution TGA 2950 (TA Instruments) thermobalance using a working N2 flux of 100 mL min-1. The TG profiles are collected in the temperature range between 25 and 800 °C. Ca. 5 mg of material is analyzed in an open platinum pan. Differential Scanning Calorimetry (DSC) measurements are carried out with a MDSC 2920 instrument (TA Instruments) equipped with a LNCA low-temperature attachment operating under a helium flux of 30 mL·min -1. Measurements are performed with a heating rate of 3°C·min -1 between -100 and 100°C on ca. 4 mg of sample in a hermitically-sealed aluminium pan. Dynamic mechanical analysis (DMA) data are collected with a TA Instruments DMA Q800 instrument using the film/fiber tension clamp. The instrument was placed inside a glove bag filled with N2(g) to protect the sample from the CO2 in the ambient air. The temperature spectra are determined by subjecting a rectangular dry film sample of ca. 25 (height) mm × 6 (width) mm × 0.2 (thickness) mm to an oscillatory sinusoidal tensile deformation with an amplitude of 4 m at 1 Hz and with a 0.05 N preload force. The temperature range of -100 to 150°C, is scanned at a rate of 4°C min-1. The elastic (storage) (E’) and viscous (loss) modulus (E’’) are at the basis of the analysis of the mechanical response of the materials. Tan  = E”/E’ is analyzed as a function of the temperature to evaluate the material damping features. The samples analyzed by HR-TG, DSC and DMA are dried for two days under vacuum before the measurements, and are manipulated under Ar. They were never exposed to the open atmosphere. BES are collected in the frequency and temperature range from 0.01 to 107 Hz and -105 to 150°C, respectively, using a Novocontrol Alpha-A analyzer. The temperature is increased in 10°C steps using a home-made cryostat operating with a N2 gas jet heating and cooling system. The temperature is measured with an accuracy higher than ±0.4°C. The geometrical cell constant is determined by measuring the electrode–electrolyte contact surface and the distance between the electrodes. No corrections for the thermal expansion of the cell are used. The complex impedance (Z*(ω) = Z’(ω) + iZ”(ω)) is converted into complex conductivity (σ*(ω) = σ’(ω) + iσ”(ω)) and complex permittivity (ε*(ω) = ε’(ω) - iε”(ω)) using the equations σ*(ω) = k[Z*(ω)]-1 and σ*(ω) = iωεoε*(ω), respectively, where k is the cell constant and ω = 2πf (f is the frequency in Hz). The wet samples are soaked in water for 2 h prior to measurement and 100 μL of doubly-distilled water is added to the measurement cell. The spectra are analysed, as elsewhere reported,19 by fitting simultaneously ε*(ω), σ*(ω) and tan δ(ω) profiles of the electric response of the materials using the equation: 𝜀 ∗ (𝜔) = ∑3𝑗=1

𝛾 𝜎𝑗 (𝑖𝜔𝜏𝑗 ) 𝑗 𝛾 𝑖𝜔[1+(𝑖𝜔𝜏𝑗 ) 𝑗 ]

Δ𝜀

+ ∑4𝑘=1 [1+(𝑖𝜔𝜏𝑘 )𝜈𝑘 ] + 𝜀∞ 𝑘

3. RESULTS AND DISCUSSION

3.1. Composition and Thermal Stability. In this report the electric response and thermal and mechanical relaxations of three films, which are based on the same Poly(vinylbenzyltrimethylammonium hydroxide)-b-poly(methylbuthylene) block copolymer with the chemical structure shown in Figure 1 are studied. A [PVBTMA][X]-b-PMB was synthesized with a 1:3 PVBTMA to PMB block ratio, divided into three batches and each brominated to a differing degree resulting in the same polymer backbone with three different degrees of functionalization (DFs) or ion exchange capacity (IECs). The films investigated have an increasing DF or IEC and are labeled A, B, and C with DF = 41.8%, 62.9%, and 80.7% and IEC = 1.14, 1.64, and 2.03 meq g-1, respectively. DF is the percentage of the polar vinylbenzyltrimethylammonium hydroxide ([VBTMA][OH]) side chains in the hydrophilic A block, shown in Figure 1 and determined using Eq. 2: DF% = n[VBTMA][OH]⁄(n[VBTMA][OH] + n[4MS] ) ∙ 100

(2)

where n[VBTMA][OH] and n[4MS] are the moles of [VBTMA][OH] and tolyl (or 4-methyl styrene, 4MS) repeat units, respectively. The composition of A, B and C and their basic properties are summarized in Table 1. The number of repeat units in the hydrophobic (A) and hydrophilic (B) domains are m = 989 and n = 318, respectively. This indicates that the thickness of the hydrophilic domain is on average ca. 1/5 of the entire thickness of the block copolymer lamella. This information reveals that this large hydrophobic domain, which as expected exhibits a low permittivity, acts to: a) mechanically and thermally stabilize the ionomer; b) provide well nanostructured percolation pathways for long range efficient hydroxyl migration processes; and c) reduce the water uptake (WU), which for the A, B, and C samples show values increasing with DF and IEC, respectively, in the order AB>C, is strongly dependent on the DF. This indicates that, T, which is correlated to the electrostatic interactions occurring between the polar cationic side chains in the hydrophilic domains, is a diagnostic parameter to study the thermal stability and interactions of the AEMs. DMA was utilized to study the evolution of the mechanical properties of the membrane C (highest DF) in the completely dry state (Figure 4). Unfortunately the hydroxide exchanged A and B membranes where too brittle to analyze by DMA in the dry state. The results depicted in Figure 4 show that the mechanical properties of the membrane C are closely dependent on the T. In particular, the storage modulus E’ drops from 100 to 6 MPa at T ~ Tg ~ -40°C. At T > Tg, for membrane C, in the dry state, E’ is mostly related to the effect of steric hindrance in the polar domains, triggered by the dipole-dipole interactions between the side groups (see the upper inset in Figure 3). The increase observed in E’ at T > 15°C is attributed to the T. This transition occurs due to the relative translational dynamics of the different backbone chains ordering the dipoles as shown in the inset to Figure 4. Stronger interactions between side groups occur above the T, leading to the observed increase in E’. This same phenomenon has also been observed before in a polysulfone

3.3 Broadband electrical spectroscopy studies. The electric response of the A, B, and C membranes is investigated by Broadband Electric Spectroscopy (BES) in order to detect the influence of the DF on: a) the relaxation phenomena characterizing the [PVBTMA][X]-b-PMB membranes in terms of polarization events and dielectric relaxations; and b) the correlation existing between the thermal, mechanical, and dielectric relaxations and the conductivity mechanisms of the materials investigated. Only the data for the wet membranes is considered here. Attempts to dry the membranes resulted in brittle films. The three-dimensional permittivity ”() surfaces of the [PVBTMA][OH]-b-PMB membranes in their fully hydrated state are shown below in Figure 5. The other components of the dielectric response are shown in Figure S1 and Figure S2 in the Supplementary Material to further aid the reader in the identification and illustration of the observed transitions. We discuss the results in comparison to the brominated [PVBTMA][Br]-b-PMB membrane with IEC = 2.2 mmol g-1 studied previously.19 For the sake of clarity, the brominated [PVBTMA][Br]-b-PMB membrane is henceforth designated the label “Br”. The qualitative analysis of the 3D surfaces, shown in Figure 5, and of the 2D-single profiles of both ”() and ‘() components of *() (See Figure S3 of Supplementary Information) reveals three polarization events, one associated with the electrode polarization and two with the inter domain polarizations (EP, IP,1, and IP,2) and two dielectric relaxation phenomena (1 and 2) for the [PVBTMA][X]-b-PMB membranes (X = Br- and OH-) in their fully hydrated state.

5 ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 14

Figure 5. Three-dimensional ε”() surfaces of fully hydrated [PVBTMA][Br]-b-PMB and [PVBTMA][OH]-b-PMB membranes with IEC of 1.14 (A), 1.64 (B), 2.03 (C), 2.2 (Br) mmol g -1.

membranes no -relaxations were observed either corresponding to the hydrophobic PMB blocks or of the hydrophilic backbone of the [PVBTA] domains as they are too weak in comparison to the observed polarization events.

The polarizations events are easily distinguished from the dielectric phenomena as they are order of magnitudes stronger, and EP is distinguished from IP,1 and IP,2 as it is affected by the membrane thickness whereas the inter domain polarizations are not. EP is due to the accumulation of charges at the interfaces between the electrode and the membrane, and IP,i is due to the accumulation of charge at the interfaces between nano domains in the membrane with different permittivities. The assignment of the events detected here is in accordance with the results reported for the [PVBTMA][Br]-b-PMB,19 which are included here only for the sake of comparison. 1 and 2 are the relaxation modes attributed to the local fluctuation of tolyl and benzyl trimethyl ammonium (-BTMA+) dipole moments of the side groups in the hydrophilic domains, respectively (Figure 6). These modes peak at very low temperature and at high frequencies, as the water acts to plasticize the membrane facilitating the movement of the side chains in the hydrophilic domains. Furthermore, as the temperature is increased above 0 °C where water melts, the ”() profiles show only a modest step-increase of permittivity in comparison to a typical perfluorinated membranes,26 indicating that the size of the water clusters embedded in the hydrophilic domains is significantly smaller. In the wet

Figure 6. Dielectric relaxations in the [PVBTMA][X] hydrophilic domains.  1- and 2-mode are associated to the fluctuation of the dipole moment of the tolyl and benzyl trimethyl ammonium side groups, respectively.

6 ACS Paragon Plus Environment

Page 7 of 14 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

In addition, a drop of permittivity is revealed in correspondence to T the disorder-order transition, also detected by the DSC and confirmed by the DMA measurements as discussed above and shown in Figures 3 and 4. We postulate that T appears to be a general phenomenon associated with anion exchange membranes of this class. To confirm this assignment we performed SAXS in liquid water as a function of temperature for [PVBTMA][Br]-b-PMB n = 29.6 and m = 71.5 kg/mol, DF 89.4% and IEC of 1.94 meq g-1. The SAXS profile with intensity normalized to the primary peak is shown in Figure 7, which clearly shows the expected peak of a material giving long-range lamellar order with swollen lamella spacing’s of 64 nm. Note the similarity of this pattern, Figure 7a, to that of the Br material elsewhere reported, the only difference being the width of the lamella.19 Intriguingly we can see that the higher order peaks show several isoscattering points as the temperature is increased from ambient to 80°C, while the position of the primary peak only varies slightly in position, Figure 7a. This is interpreted as a further indication that the electron density is heterogeneously20 distributed across` the channel and is changing with temperature. Unfortunately, the first peak is too close to the beam stop to be fully analyzed but we can background correct and decompose the higher order peaks, see Figure S7-8, in order to analyze their position and intensity, Figure 7b and c and Figure S8. Examination of the peak position, d x and intensity Ai, consistently shows a break at ca., 50°C with first a rapid rising of the property to the transition temperature and then a shallower change in the property at higher temperatures, Figure 7b and c. This is entirely consistent with T, and further proof of the disorder-order transition where as T increases the disordered dipoles are pulled together by the dipole attractive forces and then the ordered clusters of dipoles expand more slowly with increasing temperatures. The conductivity, ’() profiles of the [PVBTMA][X]-b-PMB membranes in the fully hydrated state are reported in Figure 8. In the ’() profiles, three plateaus (EP, IP,1, and IP,2) are distinguished which are followed by a subsequent decrease of ’() values toward the low frequency wing of the spectra, these are the polarization events. The AEMs in the hydroxyl form show the frequencies of the polarization relaxations shift to higher values compared to the Br membrane, thus confirming that the mobility at 25 °C of the anion in the membrane rises in the expected order: Br- (78.1 -1cm2mol-1) 0°C, a well-defined high-frequency plateau of EP is revealed, which is mostly responsible for the overall conductivity (0 ≈ EP + IP,1).30 0 is higher at a DF < 80% (membrane A and membrane B). It is also observed that the conductivity of the membranes drops at T > T; T decreases as the DF is raised in the [PVBTMA][OH]-b-PMB membranes, from a value of 80°C (membrane A) to 40°C (membrane C). For these reasons, it is expected that this event corresponds to the same phenomenon observed by DSC and DMA (see Figure 3 and 4), and is associated to a disorder-order transition involving the side groups of the [PVBTMA][X] hydrophilic blocks (see Figure 8e). T likely depends on a number of factors, including the type of anion neutralizing the –TMA+ groups and the size and the shape of the water domains impregnating the [PVBTMA][X] blocks. The plots of the conductivity values associated with the polarization events determined for the fully hydrated [PVBTMA][X]b-PMB membranes as a function of 1/T are reported in Figure 9 below. Figure 9 shows four conductivity regions (I, II, III, and IV) delimited by the thermal transitions detected by DSC and DMA measurements. These are, I below Tg, II between Tg and 0°C (the melting point of ice), III between 0°C and T, and IV, above T. Clearly IP,2 has an insignificant contribution to the overall conductivity and is not further discussed, it probably arises from a distribution of polymer blocks that does not promote significantly ionic conductivity. Linear behavior in these data sets may be interpreted by use of the Arrhenius equation and is independent of the polymer motions. When the data is

𝜎𝐸𝑃 = 𝐴𝜎 𝑇 1/2 exp (−

𝐸𝑎 𝑅(𝑇−𝑇0 )

)

(3)

where A is proportional to the density of the carrier ions, R is the gas constant, Ea is the activation energy for conductivity, and T0 is the thermodynamic ideal glass transition temperature. This type of  vs. 1/T profiles are usually observed in ion conducing materials where the dynamics of polymer backbone chains play a crucial role in promoting the long range migration mechanisms. A careful analysis, within each region, of the i behavior on 1/T allow us to identify that in regions I and II (at T < 0°C), at DF > 42% (B and C membranes) 0  EP + IP,1 with IP,1 > EP. This result indicates that at high IEC, the conductivity predominantly occurs by the exchange of anions between cations of the side chains. In region III (0°C ≤T≤ Tδ) it is revealed that:  for the A and B membranes with low and medium IEC, EP > IP,1, IP,1 cannot be detected, and 0  EP, thus showing that when the –TMA+ sites at the interfaces between hydrophobic and hydrophilic domains becomes less frequent, long range OH- migration processes occur predominantly by the exchange events of anions between water-impregnated hydrophilic domains. In addition, the dependence of EP on 1/T shows the typical VTF trend30 as the conductivity is modulated by the segmental motion of the polymer matrix.  for the C sample, with the highest IEC, it is observed that 0  IP,1 which shows that in this case long range OH- migration phenomena take place predominately by hopping processes of the X- through the –TMA+ sites.

9 ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 14

Figure 10. Dependence on 1/T of the frequencies of the dielectric relaxations of the [PVBTMA][X]-b-PMB AEMs with X = Br- and OH. Br is the membrane in Br- form with an IEC of 2.2 mmol g-1. A, B, and C are membranes in OH- form with an IEC of 1.14, 1.64, and 2.03 mmol g-1, respectively.



for Br both EP and IP,1 are of the same order of magnitude and both conduction pathways contribute as Br is much less dissociated than hydroxide.

The analysis of the dielectric strength, r,i on 1/T (see Figure S4 in the Supplementary Information) of both β1 and β2 relaxation modes supports this evidence showing that as expected: a) β1 < β2; and b) β2 values of [PVBTMA][Br] are higher than those of the [PVBTMA][OH] blocks. This later assignment is confirmed by the intensity of the dipole moment of polar side groups in both Br- and OH- form, which are decreasing concurrently from 11.45 to 6.29 D, respectively for Br-19 and OH- AEMs. These dipole moments are determined as reported previously19 by DFT calculations.32-33 To elucidate the correlation that exists between the charge migration phenomena in the bulk membranes and the reorientational relaxations which characterize the polymer matrix, the activation energies of both i and βi modes are determined. Figure 11 compares in the delineated conductivity regions the activation energies (Ea) of the i and βi modes. Ea values are obtained by fitting the data of Figures 9 and 10 with an Arrheniuslike or Vogel-Tamann-Fulcher – like equation as reported elsewhere.29 In accordance with other studies19, 28 the correspondence of the activation energy of conductivity with that of dielectric relaxation is a strong evidence of coupling effects existent between the dielectric relaxation and the long-range charge migration phenomena.19 It is observed that in regions I and II, at T < 0°C, for A, B, and C membranes:  Ea,σIP,1 and Ea,β2 are of the same order of magnitude and concurrently depend on the DF. This indicates that the long-range OH- transfer processes of IP,1 are coupled with the dipole relaxations of the –BTMA+ side groups;  Ea,σEP > Ea,β2 and decreases in the order I > II. These results demonstrate that EP is completely decoupled from the dynamics of β2, i.e., the relaxation of the host polymer matrix is more facile and have a negligible effect on EP, which likely depends on the morphology and charge exchange processes occurring in the H2O domains impregnating the hydrophilic components of the membranes.

In region IV, above the T, where 0  EP for all the AEMs, as expected, 0 increases as the DF of the materials rises. Taking all of this information together, when the conductivity is dominated by EP, and in agreement with other studies,19 the long range charge migration occurs by the exchange of X - between different delocalization bodies (DBs).28 As elsewhere defined,28 a DB is the volume of hydrophilic polymer matrix impregnated by H2O. In DBs X- is dissociated from the –TMA+ cations and solvated by H2O molecules. Thus, within DBs, the X- is exchanged between coordinating H2O molecules with relaxation times that are at least 3 orders of magnitude faster than that of the overall conductivity phenomena. To complete the BES analysis, Figure 10 reports the frequencies of the dielectric relaxations of the various AEMs as a function of 1/T. Results reported in Figure 10 show that the trends of the relaxation frequencies vs. 1/T are always of the Arrhenius-type. This confirms the previous assignment that the detected dielectric relaxation modes correspond to the local fluctuation of dipole moments of the side chains in the hydrophilic domains. 28 The 𝑓𝛽1 and 𝑓𝛽2 modes correspond to the dipole moment fluctuation phenomena of tolyl and –BTMA+ side groups, respectively.19, 28 In the [PVBTMA][OH]-b-PMB membranes, 𝑓𝛽1 decreases as DF is raised, until at the highest DF (i.e., in membrane C) 𝑓𝛽1 is not detected. Indeed, the 𝑓𝛽1 frequency is consistent with that observed for the β-relaxation of polystyrene,31 lending further support to the proposed assignment of the dielectric events discussed above. For the membranes in OH- form, the 𝑓𝛽1 mode shifts to slightly higher frequencies as the IEC rises. This result suggests that as the IEC rises the rate of the reorientational processes of the polar side groups is facilitated.

10 ACS Paragon Plus Environment

Page 11 of 14 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Figure 12. Correlation of 𝑫𝝈𝒊 (i =EP and IP,1) with 𝒇𝜷𝟐 relaxation mode for wet [PVBTMA][X]-b-PMB membrane with X= Br- or OH-. 𝑫𝝈𝒊 values are evaluated as reported elsewhere34 by using the Nernst-Einstein equation: 𝐃𝐢 = 𝛔𝐧𝐢⋅𝐑∙𝐓 , where 𝐧𝐗 = ∙𝐅𝟐 𝐗

𝝋 ⋅ 𝒅, R the gas constant, T the absolute temperature, 𝛔𝐢 are the EP or IP,1 conductivity values, F the Faraday constant, 𝝋 the total anion exchange capacity. d values are reported in Table 1. Full lines show the Einstein-Smoluchowski-like fits obFigure 11. Activation energies of the i and  i modes in region I, II, and III for the [PVBTMA][X]-b-PMB AEMs with X = Brand OH-. Br is the membrane in Br- form with an IEC of 2.2 mmol g-1. A, B, and C are the membranes in OH- form with an IEC of 1.14, 1.64, and 2.03 mmol g -1, respectively.

tained using the following Equation: 𝐃𝐢 = 𝐟𝜷𝟐 = 𝐃𝟎,𝐢 𝐟𝜷𝟐 .

𝟐 𝟔𝛕

=

𝛑∙𝟐 𝟑



where i is either EP or IP,1, 𝐷𝜎𝑖 is the diffusion coefficient determined from the conductivity values via the Nernst-Einstein equation34 (Figure 12), 𝐷𝜎0𝑖 is the diffusion coefficient at 1 Hz, i.e., the purely ionic conductivity decoupled from the movement of the polymer,  is the slope of the correlation, and 𝑓𝛽2 is the frequency of the movement of the cationic side chain. It is observed that for the higher IEC C membrane the linear dependence of Di vs. 𝑓𝛽2 : a) is shifted toward lower values; and b) exhibits a higher slope. This indicates that at high IEC, the anion migration process is less effective and more strongly dependent on the dynamics of –BTMA+ side groups than for the lower IEC hydroxyl form and the bromide form of the membrane. For all the membranes other than the higher IEC OH- C membrane, the fitting of the data gives  = 1.12  0.04 and log𝐷𝜎0𝑖 = 15.4  0.2. For these membranes as D and 𝑓𝛽2 are correlated,  is ca. 1, we can interpret the data on the basis of the Einstein-Smoluchowski equation, which describes pure percolative conduction mechanisms in a rigid medium, rewritten in the form:

Furthermore, in region I we can observe that for the Br material, Ea,σEP  Ea,σIP,1  Ea,β2. This indicates that in this case both EP and IP,1 are significantly influenced by the dynamics of the – BTMA+ side groups. In Region III, at T > 0°C, 𝐸𝑎,𝜎𝐸𝑃 increases with the DF. 𝐸𝑎,𝜎𝐸𝑃 values are consistent with the energy of hydrogen bonding networks (i.e., ~10 kJ·mol-1) and of the 2 mode. Therefore, it is hypothesized that in III, the relaxation of polar side chains are coupled with the dynamics of the H2O domains, which are responsible of the long-range charge transfer processes in the membranes. Further confirmation of the influence of the β2 dielectric relaxation of the cationic side chains, on the conductivity pathways detected in the membranes is determined by analyzing the correlation map shown in Figure 12. Here in Figure 12, on a log log-plot we show the dependence on fβ2 of the diffusion coefficients determined by using the Nernst-Einstein equation34 and the conductivity values of Figure 9. It is observed that both EP and IP,1 pathways are clearly correlated to the β2 mode. The data reported in Figure 12 can be fitted with the empirical equation: 𝑙𝑜𝑔𝐷𝜎𝑖 = 𝑙𝑜𝑔𝐷𝜎0𝑖 + 𝛼 ∙ 𝑙𝑜𝑔𝑓𝛽2

Di = D0,if where D0,i = /32i the diffusive coefficient at a frequency of 1 Hz and i the elementary migration step length. Using the fitted value of D0 (Figure 12) in the Einstein-Smoluchowski model it is possible to obtain the elementary migration distance

11 ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

at β2 = 1 Hz which is of 1.95 Å. This latter value which corresponds to the average distance between neighboring –BTMA+ groups along the hydrophilic polymer chains of the membranes. In the case of the higher IEC membrane C, where the EinsteinSmoluchowski does not hold, it is strongly implicated that the cationic groups are not distributed and must be clustered, leading to lower values of the diffusion coefficient. Taking all of above evidence together, we can conclude that in accordance with our other studies on AEMs19, 35 two types of mechanisms are contributing to the overall conductivity of the investigated membranes. The first mechanism is associated with EP and occurs when X- is exchanged between different delocalization bodies (DBs). In this case an X- migration distance of the order of 50 nm (see Figure S6 of Supporting Information) is revealed. This is consistent with the sizes of the lamella reveled by the SAXS measurement. These results indicate that this conductivity pathway in AEMs is modulated significantly by local dipole fluctuations of the –TMA+ side groups and to a lesser extent by relaxation phenomena of the host polymer matrix (segmental motions).

Page 12 of 14

independent of IEC and ascribed to the hydrophobic backbone of the polymer; and (b) an disorder-order transition associated to a structural rearrangement of the membrane at 30 ≤ T ≤ 85°C, the temperature varying with IEC and ascribed to the dipolar cationic side chains of the polymer. The analysis of the BES results obtained on the [PVBTMA][X]-b-[PMB] membranes in their fully hydrated state shows four conductivity regions (indicated as I, II, III, and IV in Figures 9-10), delimited by the thermal transitions of the materials (i.e., Tg, 0°C, T). Two polarization events, EP and IP,1, are observed at T > 0°C and dominate the electric response of the AEMs. Two dielectric relaxations, β1 and β2, tolyl and benzyl trimethyl ammonium side groups, respectively, are observed at T < 0°C. The overall conductivity of the materials (0) is equal to the sum of EP and IP,, which are correlated with β2. In all the investigated materials, it is demonstrated that two types of migration events, which are correlated to the dynamics of –TMA+ sites are revealed which play a crucial role in the modulation of the overall conductivity of proposed membranes. Indeed, the same two long-range charge migration pathways identified elsewhere19 also predominate in 0=EP+IP,1 of [PVBTMA][OH]-b-[PMB] membranes. In detail, IP,1 involves hopping processes of the anions along the interfaces between the hydrophobic and hydrophilic domains of the [PVBTMA][X]-b-[PMB] membranes, assisted by H2O solvation and 2 relaxations; EP consists of long-range charge transfer events occurring owing to the exchange of anions between different DBs in hydrophilic nanodomains. The contributions of each EP and IP,1 conductivity pathways to the overall conductivity depends on the IEC of membranes, water uptake and temperature. For the membrane with the highest IEC, 2.03 mmol g1 , cationic side clustering is observed to inhibit the anionic transport. In conclusion we showed that: a) the more dissociated hydroxide anion was much more mobile than the bromide ion as evidenced by the polarizations shifting to higher frequencies. By contrasting the bromide form of the materials with the hydroxide we showed that main contribution of the detected conductivity pathways to 0 (EP + IP = 0) depends on the type of anion neutralizing the cationic side chains of the membranes. 0 is equally modulated byEP and IP in all conductivity regions for Br- while in the OH- membranes where OH- is more disassociated from the cationic side chains it depends on the conductivity region being probed. b) The density of the polar side chains determine the type of dipole-dipole and dipole-ion interactions responsible for the nano-morphology of the sample and of the types of conductivity pathways contributing to the overall conductivity of the materials. It was detected that in the membranes with the more dissociated OH- cations at 0°C ≤ T ≤ T as the IEC was increased the overall conductivity switches from long range charge migration events based on exchange processes of OH- groups between delocalization bodies to that which occurs between neighboring coordination sites. This indicates that as the IEC in AEMs rises the density of the R-N(CH)3+X- … X-(CH3)N+-R dipole-dipole interactions increases which act to localize the anions and to modify the nano-domains of the materials.

The second mechanism (IP,2), which occurs with an average migration distances on the order of 10 Å (see Figure S6 of supplementary materials), consists of the shorter-range anion migration processes which occur owing to elementary X- by “hopping” events between the –TMA+ cations. These phenomena, which take place along the hydrophobic/hydrophilic interfaces of AEMs are assisted by the β2 relaxation and depend also on the type and shape of H2O solvation “shell” covering the polar –BTMA+ side groups. In conclusion, both these two conductivity pathways (EP and IP,1)35 a) contribute to the overall conductivity of the materials (0  EP + EP) and b) are modulated mostly by the β2 relaxation. Depending on the IEC and on the density of X -, one of these two mechanisms will predominate in the overall conductivity in each of the four conductivity regions previously described. This is easily identified by analyzing the φi parameters (φi=i/0 with i= EP, IP,1 and IP,2 and 𝜎0 = ∑𝑛𝑖=1 𝜎𝑖 ) shown in Figure S5 of Supporting Information. φi values show that: as the DF rises from A to C, the temperature range where IP,1 is dominating 0 becomes larger extending from I up to III regions for the higher IEC C sample. In a complementary way, as the DF decreases from membrane C to A, the contribution of the EP to 0 dominates from IV up to the region II for the A sample. In conclusion, when less polar side groups are present in the hydrophilic domains, larger clusters of water are required to interconnect the side groups (migration between DBs). In contrast when the density of polar side chains is larger and the polar side groups are closer each other, the conductivity is modulated the exchange of X- between neighboring side chains (hopping processes). In this latter case, the migration between DBs is less significant. 4. CONCLUSIONS The thermomechanical properties and the electrical response of [PVBTMA][OH]-b-[PMB] membranes are elucidated by HRTG, MDSC, DMA, and BES and compared with a membrane previously studied with Br- as the counter anion. The membranes are thermally stable up to ca. 140°C, and show two main thermal transitions: (a) a glass transition at Tg = -51°C which is

12 ACS Paragon Plus Environment

Page 13 of 14 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society 13. Meek, K. M.; Elabd, Y. A., J. Mater. Chem. A 2015, 3 (48), 24187-24194. 14. Nakabayashi, K.; Umeda, A.; Sato, Y.; Mori, H., Polymer 2016, 96, 81-93. 15. Cotanda, P.; Sudre, G.; Modestino, M. A.; Chen, X. C.; Balsara, N. P., Macromolecules 2014, 47 (21), 7540-7547. 16. Strasser, D. J.; Graziano, B. J.; Knauss, D. M., J. Mater. Chem. A 2017, 5 (20), 9627-9640. 17. Zhang, W.; Liu, Y.; Jackson, A. C.; Savage, A. M.; Ertem, S. P.; Tsai, T.-H.; Seifert, S.; Beyer, F. L.; Liberatore, M. W.; Herring, A. M.; Coughlin, E. B., Macromolecules 2016, 49 (13), 4714-4722. 18. Tsai, T.-H. Ionic Copolymers for Alkaline Anion Exchange Membrane Fuel Cells (AAEMFCs). University of Massachusetts, Amherst, 2013. 19. Noto, V. D.; Giffin, G. A.; Vezzu, K.; Nawn, G.; Bertasi, F.; Tsai, T.-h.; Maes, A. M.; Seifert, S.; Coughlin, E. B.; Herring, A. M., Physical Chemistry Chemical Physics 2015, 17 (46), 31125-31139. 20. Herbst, D. C.; Witten, T. A.; Tsai, T.-H.; Coughlin, E. B.; Maes, A. M.; Herring, A. M., The Journal of Chemical Physics 2015, 142 (11), 114906. 21. Couture, G.; Alaaeddine, A.; Boschet, F.; Ameduri, B., Progress in Polymer Science 2011, 36 (11), 1521-1557. 22. Merle, G.; Wessling, M.; Nijmeijer, K., J Membrane Sci 2011, 377 (1-2), 1-35. 23. Tsai, T.-H. Ionic Copolymers for Alkaline Anion Exchange Membrane Fuel Cells (AAEMFCs). University of Massachusetts, Amherst, Amherst, MA, 2013. 24. Tsai, T.-H. Ionic Copolymers for Alkaline Anion Exchange Membrane Fuel Cells. University of Massachusetts - Amherst, Amherst, MA, 2014. 25. Narducci, R.; Chailan, J. F.; Fahs, A.; Pasquini, L.; Di Vona, M. L.; Knauth, P., Journal of Polymer Science Part B: Polymer Physics 2016, 54 (12), 1180-1187. 26. Giffin, G. A.; Haugen, G. M.; Hamrock, S. J.; Di Noto, V., Journal of the American Chemical Society 2013, 135 (2), 822834. 27. Atkins, P.; Julio de Paula, J., Atkins' Physical Chemistry. 10th ed.; Oxford University Press: 2014. 28 Di Noto, V.; Piga, M.; Giffin, G. A.; Vezzù, K.; Zawodzinski, T. A., Journal of the American Chemical Society 2012, 134 (46), 19099-19107. 29. Di Noto, V.; Giffin, G. A.; Vezzù, K.; Piga, M.; Lavina, S., Broadband Dielectric Spectroscopy: A Powerful Tool for the Determination of Charge Transfer Mechanisms in Ion Conductors In Solid State Proton Conductors, Knauth, P.; Di Vona, M. L., Eds. John Wiley & Sons: Chichester, 2012; pp 109-183. 30. Di Noto, V., The Journal of Physical Chemistry B 2002, 106 (43), 11139-11154. 31. Mc Grum, N. G.; Read, B. E.; Williams, G., Anelastic and Dielectric Effects in Polymeric Solids. Dover: New York, 1967. 32. Delley, B., J Chem Phys 1990, 92 (1), 508-517. 33. Delley, B., FJ Chem Phys 2000, 113 (18), 7756-7764. 34. Di Noto, V.; Vittadello, M.; Yoshida, K.; Lavina, S.; Negro, E.; Furukawa, T., Electrochimica Acta 2011, 57, 192-200. 35. Giffin, G. A.; Lavina, S.; Pace, G.; Di Noto, V., The Journal of Physical Chemistry C 2012, 116 (45), 23965-23973.

c) At T the conductivity is observed to drop by several orders of magnitude due to the clustering events of the dipolar cationic groups. We demonstrated that this is a disorder-order event and is a general diagnostic phenomena for studying AEMs of this type.

ASSOCIATED CONTENT Supporting Information. Additional BES and SAXS results are available free of charge via the Internet at http://pubs.acs.org.

AUTHOR INFORMATION Corresponding Author * Authors to whom correspondence should be addressed: [email protected]; [email protected]

Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. ‡These authors contributed equally.

ACKNOWLEDGMENT This work was funded by the European office of the US Army Research Office under grant W911NF-13-1-0400 and the US Army Research Office by a MURI grant W911NF-10-1-0520. The authors wish to acknowledge use of the CSM SAXS facility.

REFERENCES 1. Varcoe, J. R.; Slade, R. C. T., Fuel Cells 2005, 5 (2), 187200. 2. 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.; Zhuang, L., Energy Environ. Sci. 2014, 7 (10), 3135-3191. 3. Marino, M. G.; Melchior, J. P.; Wohlfarth, A.; Kreuer, K. D., J Membrane Sci 2014, 464, 61-71. 4. Pandey, T. P.; Maes, A. M.; Sarode, H. N.; Peters, B. D.; Lavina, S.; Vezzù, K.; Yang, Y.; Poynton, S. D.; Varcoe, J. R.; Seifert, S.; Liberatore, M. W.; Di Noto, V.; Herring, A. M., Physical Chemistry Chemical Physics 2015, 17 (6), 4367-4378. 5. Pandey, T. P.; Sarode, H. N.; Yang, Y.; Yang, Y.; Vezzù, K.; Noto, V. D.; Seifert, S.; Knauss, D. M.; Liberatore, M. W.; Herring, A. M., Journal of The Electrochemical Society 2016, 163 (7), H513-H520. 6. Tsai, T.-H.; Ertem, S. P.; Maes, A. M.; Seifert, S.; Herring, A. M.; Coughlin, E. B., Macromolecules 2015, 48 (3), 655-662. 7. Elabd, Y. A.; Hickner, M. A., Macromolecules 2011, 44 (1), 1-11. 8. Li, Y.; Jackson, A. C.; Beyer, F. L.; Knauss, D. M., Macromolecules 2014, 47 (19), 6757-6767. 9. Yang, Y.; Knauss, D. M., Macromolecules 2015, 48 (13), 4471-4480. 10. Akiyama, R.; Yokota, N.; Nishino, E.; Asazawa, K.; Miyatake, K., Macromolecules 2016, 49 (12), 4480-4489. 11. Weiber, E. A.; Meis, D.; Jannasch, P., Polymer Chemistry 2015, 6 (11), 1986-1996. 12. Arges, C. G.; Kambe, Y.; Suh, H. S.; Ocola, L. E.; Nealey, P. F., P Chemistry of Materials 2016, 28 (5), 1377-1389.

13 ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

SYNOPSIS TOC

14 ACS Paragon Plus Environment

Page 14 of 14