UC Berkeley

Solid electrolytes have been a long-standing goal of the battery industry since they have been considered safer than flammable liquid electrolytes and are capable of producing batteries with higher energy densities. The latter can be achieved by using a lithium metal anode, which is unstable against liquid electrolytes. Past attempts at polymer electrolytes for lithium-anode batteries have failed due to the formation of lithium dendrites after repeated cycling. To overcome this problem, we have proposed the use of microphase separated block copolymers. High ionic conductivity is obtained in soft polymers such as poly(ethylene oxide) (PEO) where rapid segmental motion, which is needed for ion transport, necessarily results in a decrease in the rigidity of the polymer. Block copolymers have the ability to decouple the requirements of high modulus, needed to prevent dendrite growth, and high ionic conductivity. Furthermore, the use of block copolymers may enable the creation of well-defined, optimized pathways for ion transport.

This dissertation presents studies of a poly(styrene-block-ethylene oxide) (SEO) copolymer blended with the lithium salt LiTFSI for use as a polymer electrolyte. In this case, the PEO is the ionically conducting block whereas the PS provides mechanical rigidity. The polymers used for this study were synthesized via anionic polymerization to obtain copolymers with low polydispersity. The introduction of a nonconducting microphase undoubtedly decreases the overall conductivity of the block copolymer relative to that of the ionically conducting homopolymer. Furthermore, the addition of salts into the block copolymer can be viewed as adding a selective solvent to the system. This invariably changes the energetic interactions in the systems. It is our goal to determine the correlation between the salt concentration and polymer phase behavior, and determine the effects of phase behavior on the ionic conductivity.

The polymer electrolyte system is designated as SEO (a-b)/LiTFSI where a = molecular weight of the PS block (kg/mol) and b = molecular weight of the PEO block (kg/mol). By varying the salt concentration, r = [Li]/[EO], and by varying a and b, several different morphologies such as alternating lamellae, hexagonally packed cylinders, and a cocontinuous network phase are obtained. Characterization of the electrolyte systems includes a combination of small-angle Xray scattering, optical birefringence measurements, and alternating current impedance spectroscopy.

The phase behavior and thermodynamics of the block copolymers as a function of LiTFSI concentration are also explored. It is assumed that the LiTFSI resides mainly in the PEO phase, the polymer with the higher dielectric constant, which is known for solvating lithium salts very effectively. Upon addition of LiTFSI salts to SEO systems, we obtain a disorder-to-order transition at a particular salt concentration. Further increases in the salt concentration have been shown to lead to other phase transitions such as lamellar to gyroid, or gyroid to cylinders. Changes in morphology cannot be attributed to increases in volume fraction of the PEO/LiTFSI phase alone. It is hypothesized that the presence of salts increases the effective Flory Huggins chi parameter. Using six different SEO/LiTFSI mixtures with accessible order-to-disorder transitions, we can develop a relationship to estimate the change in the effective chi parameter with salt concentration. It was established that this relationship is a linear function, in good agreement with theoretical predictions. This relationship was also obtained for a mixture of SEO polymers with the ionic liquid imidizolium TFSI (ImTFSI). The effective chi parameter relationships were approximately the same, indicating that the large anion drives the thermodynamics of the polymer/salt systems. The slope of the effective chi vs. r line, m, is compared to theoretical calculations. The theoretically determined values were consistently higher than experimentally determined ones.

In this study, ionic conductivity measurements through order-order and order-disorder phase transitions (OOTs and ODTs) in mixtures of SEO with LiTFSI were performed to determine the effect of morphology on conductivity. The molecular weight of the blocks and the salt concentration were adjusted to obtain OOTs and ODTs within the available experimental window. The normalized conductivity (normalized by the ionic conductivity of a 20 kg/mol homopolymer PEO sample at the salt concentration and temperature of interest), was also calculated to elucidate the effect of morphology. For samples with a major phase PEO block (e.g. volume fraction of PEO in SEO is greater than 0.5), no dramatic changes in conductivity were seen when transitioning through different morphologies. The well-known Vogel-Tamman- Fulcher (VTF) equation provides an excellent fit for the temperature dependence of the conductivities regardless of morphology. However, for samples with minor phase PEO block, the conductivity/structure relationship is more complex. Through in-situ conductivity/SAXS experiments, these samples show changes in conductivity with temperature, which are dependent upon the thermal history. The reason for these changes has not been established.