Effect of Confinement on Proton Transport Mechanisms in Block Copolymer/Ionic Liquid Membranes

Nanostructured membranes containing structural and proton-conducting domains are of great interest for a wide range of applications requiring high conductivity coupled to high thermal stability. Understanding the effect of nanodomain confinement on protonconducting properties in such materials is essential for designing new, improved membranes. This relationship has been investigated for a lamellae-forming mixture of poly(styrene-b-2-vinyl pyridine) (PS-b-P2VP) with ionic liquid composed of imidazole and bis(trifluoromethylsulfonyl)imide, where the ionic liquid selectively resides in the P2VP domains of the block copolymer. Quasi-elastic neutron scattering and NMR diffusion measurements reveal increased prevalence of a fast proton hopping transport mechanism, which we hypothesize is due to changes in the hydrogen bond structure of the ionic liquid under confinement. This, in combination with unique ion aggregation behavior, leads to a lower activation energy for macroscopic ion transport compared with that in a mixture of ionic liquid with P2VP homopolymer. The proton transference number in both samples is significantly higher than that in the neat ionic liquid, which could be taken advantage of for applications such as proton exchange membrane fuel cells and actuators. These results portend the rational design of nanostructured membranes having improved mechanical properties and conductivity. ■ INTRODUCTION Nanostructured membranes containing structural and ionconducting phases are of great interest for a wide variety of applications requiring high ionic conductivity coupled to mechanical durability. Such membranes may be designed to exhibit continuous ion-conducting channels that lead to enhanced conductivity compared with nonordered materials. For instance, Nafion (Dupont) is currently the industry standard proton-conducting polymer membrane for many applications because of its exceptional conductivity and mechanical properties that have been attributed to its nanoscale phase separation into conducting, water-filled domains and structural, hydrophobic domains. Furthermore, materials designed to have well-ordered, continuous nanostructures have significantly higher conductivities than comparable nonordered materials. Understanding the effect of morphology on conductivity has become increasingly important in recent years as a tool for designing new, improved membranes.4−9 Of the many factors contributing to improved conductivity in nanostructured membranes (including conducting phase continuity and increased contrast between conducting and structural phases), we expect differences in the local charge carrier environment near the “walls” of conducting phases to have a significant impact on the mechanism of conductivity. It has been shown, for example, that hydrogen bonding is significantly different at the walls of confined water domains10−13 and that such effects have significant implications for proton transport in hydrated Nafion.14−18 A few studies of nanostructured materials with ion carriers other than water have also identified wall effects on the mechanism of conductivity. These effects were deduced largely based on bulk conductivity measurements. In one case, NMR experiments also provided clues as to how the local dynamics change, but determining exactly how confinement affects the molecular mechanisms of conductivity remains difficult. Mixing a block copolymer with an ionic liquid is one route to obtaining nanostructured, ion-conducting membranes having high ionic conductivity coupled to favorable mechanical durability. Ionic liquids have been selectively incorporated Received: December 19, 2011 Revised: March 13, 2012 Published: March 23, 2012 Article pubs.acs.org/Macromolecules © 2012 American Chemical Society 3112 dx.doi.org/10.1021/ma202741g | Macromolecules 2012, 45, 3112−3120 into one phase of a diblock copolymer, where the second phase imparts mechanical durability to the membrane.21−26 Such membranes self-assemble into well-defined nanostructures, making them ideal materials for studying the relationship between structure and conductivity. In this work, the effect of confinement on the mechanism of conductivity has been investigated for mixtures of the diblock copolymer poly(styrene-b-2-vinyl pyridine) (PS-b-P2VP) with the protonc o n d u c t i n g i o n i c l i q u i d , i m i d a z o l i u m : b i s (trifluoromethylsulfonyl)imide ([Im][TFSI]) having an excess of imidazole, in which the ionic liquid is selective for the P2VP phase of the block copolymer. Nonstoichiometric [Im][TFSI] containing excess imidazole conducts protons via two mechanisms: a vehicle mechanism whereby protons are carried by imidazolium cations and a proton-hopping mechanism whereby protons are transferred between hydrogen bonded imidazole molecules. Excess imidazole molecules act as proton “acceptors” that facilitate proton hopping, increasing proton conduction vis-a-̀vis conduction by vehicle diffusion alone. Because the proton-hopping mechanism is strongly related to the hydrogen bond network of imidazole and hydrogen bonding is known to be affected by confinement, we surmise that proton hopping in imidazole will also be affected by confinement (i.e., to block copolymer nanodomains). Previous studies of block copolymer/ionic liquid membranes have focused on bulk conductivity measurements. These measurements have provided insight into the relationship between structure and conductivity but afford limited information regarding the mechanism of ionic liquid conductivity. Combined conductivity and NMR studies of mixtures of ionic liquids with homopolymers and random copolymers have shown that specific interactions between the ionic liquid and polymers greatly affect the prevalence of ion aggregation and the number of effective charge carriers, yet there have been no similar studies on mixtures containing protic ionic liquids or having well-defined nanostructures. In this work, we employ a combination of experimental techniques that together probe the dynamics of a variety of molecular and atomic species on length scales ranging from tenths of nanometers to micrometers to millimeters to determine the effect of confinement on the mechanisms of proton transport in a selfassembled mixture of PS-b-P2VP block copolymer with nonstoichiometric [Im][TFSI] ionic liquid, where the ionic liquid is confined to P2VP domains. A mixture of nonstoichiometric [Im][TFSI] with P2VP homopolymer has been studied in addition to the block copolymer mixture to discriminate effects of confinement from effects of mixing with P2VP. We show that combining the ionic liquid with both P2VP and PS-b-P2VP has significant effects on ion aggregation and the proton transference number and, most strikingly, that the amount of proton hopping compared with vehicle diffusion is greatly increased in the nanostructured membrane. These results portend the rational design of nanostructured membranes having improved mechanical properties and conductivity. ■ EXPERIMENTAL SECTION Polymer Synthesis and Characterization. Hydrogenated and deuterated poly(2-vinyl pyridine) (P2VP and dP2VP) and poly(styrene-b-2-vinyl pyridine) (PS-b-P2VP and dPS-b-dP2VP) were synthesized via anionic polymerization as previously described. d8Styrene monomer (Polymer Source) and d7-2-vinyl pyridine monomer (Isotec) were used to synthesize deuterated polymers. The molecular weights of the polystyrene blocks (PS or dPS) were determined using gel permeation chromatography (GPC), and the total molecular weights of the block copolymers were determined using 1H NMR (Bruker AVB 300 MHz). The molecular weight of P2VP was determined using 1H NMR end-group analysis, and the molecular weight of dP2VP was determined using GPC. GPC was used to assess the polydispersity of each polymer. The degree of polymerization of the P2VP or dP2VP block, NP2VP, the volume fraction of PS or dPS, f PS, and the polydispersity index (PDI) of each polymer are given in Table 1. Ionic Liquid Purification and Preparation. Imidazole (≥95%), d4-imidazole (≥98% deuteration), and bis(trifluoromethylsulfonyl)imide (HTFSI, ≥95%) were purchased from Sigma Aldrich and purified by sublimation under vacuum. The final purity of each starting material was assessed using differential scanning calorimetry (DSC) and 1H NMR. Purified imidazole or d4-imidazole and HTFSI were combined in a 4:1 molar ratio and heated to 100 °C for 2 to 3 h. The compositions of the resulting nonstoichiometric [Im][TFSI] and [dIm][TFSI] ionic liquids were confirmed by comparing the measured melting points and 1H NMR profiles to literature. The structures of the ionic liquid molecules are shown in Figure 1. Because of their hygroscopic nature, the ionic liquids and their starting materials were handled in an argon atmosphere glovebox and sealed sample holders at all times. Preparation of Polymer/Ionic Liquid Mixtures. Dichloromethane and tetrahydrofuran were degassed using three freeze, pump, thaw cycles, dried by stirring over CaH2 overnight, and stored on molecular sieves in an argon atmosphere glovebox. All further sample preparation was performed within the glovebox. Predetermined masses of ionic liquids and polymers were added to glass vials. Ca. 5 wt % solutions were prepared by dissolving in dichloromethane (block copolymer mixtures) or tetrahydrofuran (homopolymer mixtures), and the solutions were stirred overnight. Samples were cast one drop at a time into sample holders for DSC, small-angle X-ray scattering (SAXS), quasi-elastic neutron scattering (QENS), and AC impedance spectroscopy. Samples were cast onto pieces of Kapton for NMR. Solvent was removed carefully so as not to remove simultaneously excess imidazole. Block copolymer samples were heated to 65 °C for precisely 30 min. Homopolymer samples were dried at 35 °C for 3 days. Complete solvent removal and negligible imidazole loss were confirmed by 1H NMR. After solvent removal, DSC, SAXS, QENS, and conductivity sample holders were sealed shut, whereas NMR samples were scraped off of the Kapton and into sealable NMR tubes (for diffusion measurements) or 4 mm (outer diameter) magic-angle spinning (MAS) rotors. Samples were sealed in jars containing desiccant for transportation to experimental apparatuses. The exclusion of water from samples was confirmed by 1H NMR. Table 1. Degree of Polymerization, Volume Fraction, and PDI of Polymers Studied polymer NP2VP f PS PDI