Langmuir 2004, 20, 5403-5411
5403
Highly Ion Conductive Poly(ethylene oxide)-Based Solid Polymer Electrolytes from Hydrogen Bonding Layer-by-Layer Assembly Dean M. DeLongchamp† and Paula T. Hammond* Department of Chemical Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139 Received January 23, 2004. In Final Form: March 22, 2004 We report the development of a solid polymer electrolyte film from hydrogen bonding layer-by-layer (LBL) assembly that outperforms previously reported LBL assembled films and approaches battery integration capability. Films were fabricated by alternating deposition of poly(ethylene oxide) (PEO) and poly(acrylic acid) (PAA) layers from aqueous solutions. Film quality benefits from increasing PEO molecular weight even into the 106 range due to the intrinsically low PEO/PAA cross-link density. Assembly is disrupted at pH near the PAA ionization onset, and a potential mechanism for modulating PEO:PAA ratio within assembled films by manipulating pH is discussed. Ionic conductivity of 5 × 10-5 S/cm is achievable after short exposure to 100% relative humidity (RH) for plasticization. Adding free ions by exposing PEO/ PAA films to lithium salt solutions enhanced conductivity to greater than 10-5 S/cm at only 52% RH and tentatively greater than 10-4 S/cm at 100% RH. The excellent stability of PEO/PAA films even when exposed to 1.0 M salt solutions led to an exploration of LBL assembly with added electrolyte present in the adsorption step. Fortuitously, the modulation of PEO/PAA assembly by ionic strength is analogous to that of electrostatic LBL assembly and can be attributed to electrolyte interactions with PEO and PAA. Dry ionic conductivity was enhanced in films assembled in the presence of salt as compared to films that were merely exposed to salt after assembly, implying different morphologies. These results reveal clear directions for the evolution of these promising solid polymer electrolytes into elements appropriate for electrochemical power storage and generation applications.
Introduction In recent years, layer-by-layer (LBL) assembly has emerged as a promising functional thin film fabrication technique. Best described as a type of assisted selfassembly, the LBL method allows one to construct a film atop a substrate of almost any composition or topology by alternating its exposure to solutions containing species of opposite multivalent attractive affinities. The flexibility, ease, and accessibility of this approach have resulted in an explosion of focused research efforts.1-4 LBL assembly is most often achieved by exploiting electrostatic forces: an anionic substrate is exposed to polycation solution and rinsed, then exposed to polyanion solution and rinsed.5 Each exposure results in the reproducible and conformal deposition of an overcompensating layer of polyion, so that surface charge reversal is assured. This cycle can be repeated for a nominal number of layer pairs in a scheme that provides unprecedented control over film thickness and compositional variation normal to the substrate. Each layer pair thus created is inherently a two-component composite that can contain almost any conceivable combination of materials or functionalities.1-4 Solid electrolytes are functional films of increasing relevance to a wide variety of applications. In different forms, these ion-conductive media are employed in lithium batteries, fuel cells, and many photovoltaic panels. * To whom correspondence may be addressed. † Current address: National Institute of Standards and Technology, Polymers Division, Gaithersburg, MD. (1) Decher, G. Science 1997, 277, 1232. (2) Hammond, P. T. Curr. Opin. Colloid Interface Sci. 2000, 4, 430. (3) Schonhoff, M. Curr. Opin. Colloid Interface Sci. 2003, 8, 86. (4) Multilayer thin films; Decher, G., Schlenoff, J. B., Eds.; Wiley: Weinheim, 2003. (5) Decher, G.; Hong, J. D.; Schmitt, J. Thin Solid Films 1992, 210, 831.
Development of these power sources is critical to enabling the portability of current and next-generation electronic devices. Polymer-based solid electrolytes are the most heavily researched type because they can be processed as flexible and conformal films, are generally low cost, and often conduct well at room temperature.6-8 Solid polymer electrolytes must provide a mechanically robust matrix within which ion dissociation and migration can occur to compensate electrode electrochemistry. These requirements necessitate selecting a polymer (or a localized portion: block or graft) with high dielectric constant and generous room-temperature chain mobility or alternatively employing a polymer with an affinity for a polar liquid plasticizer. Because of these necessary qualities, most solid electrolyte polymers can be dissolved or dispersed in water, and their molecular structures frequently include ionic or other affinity groups. Common solid electrolyte polymers are thus predisposed to facile LBL assembly, and the LBL assembly process may be applied to form composite solid electrolyte films and tailor the architecture of those films at the molecular level.9 As an additional advantage, the excellent uniformity that is a hallmark of LBL assembled films may allow reduction of solid electrolyte thickness as a power cell design parameter, lowering electrolyte resistance and yielding higher power density devices. To harness these potential benefits, we previously pursued a solid electrolyte design strategy by electrostatic LBL assembly of the polycation (6) Armand, M. B. In Fast Ion Transport in Solids; VanGool, W., Ed.; North-Holland: Amsterdam, 1973. (7) MacCallum, J. R.; Vincent, C. A., Eds. Polymer electrolyte reviews; Elsevier: London, 1987; Vol. 1. (8) Gray, F. M. Polymer Electrolytes; The Royal Society of Chemistry: Cambridge, U.K., 1997. (9) DeLongchamp, D. M.; Hammond, P. T. Chem. Mater. 2003, 15, 1165.
10.1021/la049777m CCC: $27.50 © 2004 American Chemical Society Published on Web 05/18/2004
5404
Langmuir, Vol. 20, No. 13, 2004
linear poly(ethylene imine) (LPEI) with several polyacids;9 this work resulted in ionic conductivity levels greater than 10-5 S/cm, far exceeding levels reported in LBL assembled films thus far,10 and appropriate for photovoltaic applications.11 An important conclusion from our previous work was that high ionic cross-link density slowed free ion mobility, presumably due to a hindrance of polymer chain motion and because ion pair sites can behave in a manner similar to ion-exchange resins,12 creating “Coulomb traps” or associative sites for mobile ions. A critical further research direction was that LBL assembly exploiting nonelectrostatic interactions such as hydrogen bonding might circumvent these limiting effects. Our objective in this present work is to explore the LBL assembly of a film by hydrogen bonding forces and then develop that film as a solid polymer electrolyte. Harnessing hydrogen bonding to drive LBL assembly was introduced by Stockton and Rubner, who demonstrated assembly based on poly(aniline).13 Robust and electronically conductive films were assembled in a LBL fashion by pairing poly(aniline) with several nonionic polymers. The behavior of similar films was explored by Wang and co-workers in a series of reports addressing hydrogen bonding pyridinecontaining polymers.14-17 More recently, Sukhishvili and Granick have described hydrogen bonded LBL assembled films formed from poly(acrylic acid) (PAA) and poly(methacrylic acid) (PMAA) paired with nonionic polymers such as poly(vinylpyrrolidone) and poly(ethylene oxide) (PEO). These films dissolve at pH sufficiently basic to ionize the polyacid, leading to potential applications for these films as controllably dissolvable drug delivery vehicles.18,19 The solubility behavior of similar poly(acrylamide)-based systems has since been exploited by Yang and Rubner to attain patterned films by inkjet printing water of controlled pH,20 while analogous stimuliinduced dissolution of an electrostatically LBL assembled film has been demonstrated by Dubas and Schlenoff, involving a salt-mediated dissolution mechanism.21 The breadth of application of hydrogen-bonded LBL assembled films is expanded here by the development of one such system as a solid polymer electrolyte. Assembly by hydrogen bonding allows us the opportunity to employ PEO as a film constituent. PEO is the most commonly employed polymer in lithium batteries due to its ability to solvate low lattice energy lithium salts. We combine PEO in this work with PAA, and first elaborate the effects of assembly pH and PEO molecular weight on film growth. Optimal pH and molecular weight are identified, and evaluations of ionic conductivity by impedance spectroscopy are then presented. We employ water as a model plasticizer because its concentration is controllable by exploiting its well-characterized vapor-liquid equilibrium. We demonstrate that the free ion population within PEO/ PAA films can be greatly enhanced by exposing already(10) Durstock, M. F.; Rubner, M. F. Langmuir 2001, 17, 7865. (11) Tokuhisa, H.; Hammond, P. T. Adv. Funct. Mater. 2003, 13, 831. (12) Farhat, T. R.; Schlenoff, J. B. Langmuir 2001, 17, 1184. (13) Stockton, W. B.; Rubner, M. F. Macromolecules 1997, 30, 2717. (14) Wang, L. Y.; Wang, Z. Q.; Zhang, X.; Shen, J. C.; Chi, L. F.; Fuchs, H. Macromol. Rapid Commun. 1997, 18, 509. (15) Wang, L. Y.; Fu, Y.; Wang, Z. Q.; Wang, Y.; Sun, C. Q.; Fan, Y. G.; Zhang, X. Macromol. Chem. Phys. 1999, 200, 1523. (16) Wang, L. Y.; Fu, Y.; Wang, Z. Q.; Fan, Y. G.; Zhang, X. Langmuir 1999, 15, 1360. (17) Wang, L. Y.; Cui, S. X.; Wang, Z. Q.; Zhang, X.; Jiang, M.; Chi, L. F.; Fuchs, H. Langmuir 2000, 16, 10490. (18) Sukhishvili, S. A.; Granick, S. J. Am. Chem. Soc. 2000, 122, 9550. (19) Sukhishvili, S. A.; Granick, S. Macromolecules 2002, 35, 301. (20) Yang, S. Y.; Rubner, M. F. J. Am. Chem. Soc. 2002, 124, 2100. (21) Dubas, S. T.; Schlenoff, J. B. Macromolecules 2001, 34, 3736.
DeLongchamp and Hammond
assembled PEO/PAA films to lithium salt solutions. Finally, variations in the thickness and roughness of PEO/ PAA films were observed when deposition solutions contained additional electrolyte. These variations, which are explained by the physical interactions of PEO and PAA with the electrolyte, indicate that the composition and morphology of hydrogen bonded LBL assembled films might be modulated by the presence of ions during deposition in a manner similar to that of electrostatically LBL assembled films. We aim here to provide a preliminary exploration; we are currently developing targeted studies to optimize ionic conductivity, increase compatibility with common power storage and generation applications, and evaluate alternative uses for this promising new class of LBL assembled films. Experimental Section Materials. Nonionic PEO (1500, 20 000, and 4 000 000 Mw, Polysciences) and polyanions PAA (90 000 Mw, Polysciences) and poly(methacrylic acid) (PMAA, 100 000 Mw, Polysciences) were used as received. These polymers were weighed and dissolved in Millipore MilliQ filtered water (deionized to 18.2 MΩ‚cm and filtered through a 0.22 µm membrane), and the resultant solutions were then pH adjusted with dilute aqueous HCl. PEO solution was heated to 60 °C to ensure complete dissolution. All polymer solutions were 0.020 M (polymer concentrations with respect to repeat unit). Some assembly conditions required the dissolution of LiCF3SO3 or MgCl2 salts (Aldrich), which were added directly to aqueous polymer solutions after polymer dissolution and pH adjustment. After salt addition, pH was remeasured; in cases of very high ionic strength, further HCl solution addition was required to attain pH 2.5 (the only pH employed in salt-addition studies). Deposition substrates in every case were 1 in. × 2 in. indium-doped tin oxide (ITO) coated glass purchased from Donelly Applied Films and patterned by DCI, Inc., to form multiple 3 mm ITO stripes. ITO film resistance was measured to be 28 Ω/square after patterning. The ITO substrates were cleaned by ultrasonication (Bransonic) in a series of solvents for 15 min each in the following order: dichloromethane, acetone, methanol, and MilliQ-filtered water. Following the water cleansing, the substrates were dried under a jet of desiccated nitrogen gas. Immediately before use, the ITO glass substrates were cleaned in a Harrick PCD 32G plasma cleaner with oxygen bleed for 5 min. Following plasma cleaning, substrates were immediately immersed in a pH 4 solution of LPEI (25 000 Mw, Polysciences, Inc.) at 0.020 M with respect to the repeat unit molecular weight for 10 min to prepare the ITO surface for PEO/PAA deposition. Assembly. Films were constructed using a modified Carl Ziess DS50 programmable slide stainer. Substrates were exposed first to PEO solution for 10 min, followed by 4 min of rinsing in three MilliQ-filtered water baths that had also been pH-adjusted, and then exposed to polyanion solution for 10 min and rinsed, and then the cycle was repeated for the required number of layer pairs. After assembly, all films were dried without further rinsing and then stored in a desiccator. Some films were invested with salt ions by soaking the films after storage in aqueous solutions of LiCF3SO3 or MgCl2 that had been adjusted to pH 2.5. Soaking was of 24 h duration, and films were dried without further rinsing under desiccated nitrogen immediately after. Soaking was performed before test bed fabrication. Thickness and roughness measurements were performed on the constructed films with a Tencor P10 profilometer using a 2 µm stylus tip and 5 mg stylus force; profiling was performed by scoring a portion of the film down to the substrate and then profiling the score. All thickness measurements were made atop the ITO-coated portion of the deposition substrates, though thickness measurements on the glass portions were found to be of similar magnitude. Film thickness was measured by a Gaertner single-wavelength ellipsometer atop silicon substrates for films of thickness less than 1000 Å, using a fixed incident angle of 70° and a fixed refractive index of 1.51. Test Bed Fabrication. After assembly, films for ionic conductivity evaluation were dried at 110 °C for 24 h, which has been shown to effectively remove water from LBL assembled
Ion Conductive Polymer Electrolytes films.22 The drying was followed by thermal evaporation through a custom-designed shadow mask 2 mm wide, with 1000 Å thick gold electrodes perpendicular to the 3 mm wide patterned ITO stripes. This processing creates two-electrode test beds of 6 mm2 area in which the LBL assembled film is sandwiched between ITO and gold electrodes. Electrode contacts are offset from each other and the measured film cell. Dimensions allowed eight such cells per substrate. The cells were profiled by Tencor P10 profilometer to verify the absence of significant gold penetration into the LBL assembled film by comparing the height of the gold layer atop the film to the expected height from coevaporation thickness monitoring. Testing. The cells were exposed to a variety of controlled humidity environments. First the cells were exposed to a chamber that contained anhydrous CaSO4 (Drierite) the solid-vapor equilibrium of which controls humidity to approximately 17% relative humidity (RH), as measured by a VWR pen thermometer/ hygrometer (all RH measurements (2%) at a room temperature of 25 °C. The CaSO4-equilibrated atmosphere is described as “dry” or 17% RH throughout this report. The chamber was approximately 0.05 m3 and contained a fan recirculating at 0.15 m3/min; equilibrium RH at any humidity level was reached in the chamber interior within approximately 5 min with this configuration. The cells were exposed to the relatively dry CaSO4equilibrated environment for 7 days. After equilibration, ionic conductivity was evaluated within the chamber by impedance spectroscopy. Substrates were accessed by means of electrodes built into the chamber wall. Impedance spectroscopy was performed using a Solartron 1260 impedance analyzer scanning from 1 MHz to 1 Hz. Due to noise at low frequency and high impedance, the lowest frequency included in analysis was variable and typically greater than 1 Hz; the lowest frequency was chosen so that the measurement would be within the
