UC Berkeley

The widespread use of variable and intermittent renewable energy sources will require robust energy storage systems. Rechargeable batteries have the potential to meet this need. The lithium-ion battery is the one of the most popular rechargeable batteries, consisting of a graphite anode, a liquid organic electrolyte, and a metal oxide cathode. Unfortunately, the energy density in commercial lithium-ion batteries is approaching the theoretical limit for lithium-ion technology, and improvements are necessary to enable large scale grid energy storage applications. Safety is also an important criterion for energy storage applications, necessitating a fundamental improvement in the highly flammable lithium-ion batteries. Replacing the graphite anode with pure lithium metal can enable higher energy densities and replacing the flammable organic liquid solvent with a nonvolatile polymer electrolyte can improve safety. We are interested in polymer electrolytes because of their improved safety conditions and their compatibility with lithium metal anodes. However, ion transport and thermodynamics in polymer electrolytes is fundamentally different than in traditional liquid electrolytes. This thesis focuses on understanding ion transport and thermodynamics in polymer electrolytes. Ion transport of polymer electrolyte systems is characterized under the framework of Newman’s concentrated solution theory. We explicitly define the cation transference number, which describes the fraction of current carried by the cation. Two experimental approaches (Hittorf and Bruce-Vincent) for measuring the cation transference number are described. We illustrate the importance of the cation transference number and its dependence on reference frames, and compare values for commonly reported transference numbers in PEO electrolyte systems. We also provide a comprehensive description of the electrochemical properties of poly(ethylene oxide) (PEO) electrolytes that accounts for error propagation and uncertainty. Data from 64 independent PEO electrolyte samples, with PEO molecular weights ranging from 5 kg mol-1 to 275 kg mol-1, are compiled and analyzed to demonstrate that ion transport properties remain constant above the entanglement threshold. A preliminary investigation into the possibility of a salt diffusion coefficient dependence on molecular weight in PEO electrolytes and a discussion of its implications is presented. PEO has been the most extensively studied polymer for lithium battery applications, yet its use as an electrolyte is limited by its poor ionic conductivity and low cation transference number with respect to solvent velocity. Two strategies for developing superior ion transport properties are presented. (1) Moving beyond homopolymer electrolytes (i.e. single polymer and salt) to block copolymer systems. The ion transport and thermodynamics of the hybrid inorganic-organic block copolymer poly(ethylene oxide)-b-polyhedral oligomeric silsesquioxane (PEO-POSS) are reported. We find that PEO-POSS electrolytes self-assembles into a variety of morphologies ranging from lamellar to disorder, and these morphologies affect ion transport. Decreasing the length of the alkyl chain substituent on the POSS monomer is found to improve ion transport. (2) Mixing two chemically distinct polymers with lithium salt to create miscible polymer blend electrolytes. We use small angle neutron scattering (SANS) as an unambiguous method for determining the phase behavior of polymer blend systems with and without salt. We demonstrate that PEO and poly(1,3,5-trioxocane) (P(2EO-MO)) are miscible in the presence of lithium salt. Miscibility, or homogeneity, is governed by the thermodynamics of the polymer blend system. We apply the Random Phase Approximation to homogeneous PEO/P(2EO-MO) electrolyte systems to calculate the Flory-Huggins parameter, which describes the enthalpic interaction between two different polymers. We also show that PEO and poly(1,3,-dioxolane) (P(EO-MO)) follow the same miscibility trend as PEO/P(2EO-MO) blends, but that P(2EO-MO)/P(EO-MO) blends are only miscible in the salt-free state. The Flory-Huggins interaction parameter is also relevant beyond battery applications to topics such as separation where uptake in polymer membranes is dictated by thermodynamics, e.g. Donnan equilibrium. We have developed a set of equations describing the thermodynamics of univalent and multivalent charged polymer networks in electrolytic solution. In effect, we have modified the classical Donnan equilibrium model to account for both enthalpic interactions and the elasticity of the gel. We model experimental data obtained in a block copolymer membrane soaked in a traditional liquid electrolyte. We also compare experimental data from literature with our model results. This thesis provides a thorough description for the measurement and analysis of ion transport in polymer electrolytes. It is a good starting point for further studies into thermodynamics of polymer blend electrolytes and polymer membranes.