UC Berkeley

Rechargeable batteries are ubiquitous in our daily lives thanks to advancements in technology and low manufacturing costs for large scale production of lithium-ion batteries. State of the art, commercially available batteries are comprised of a graphite anode and liquid electrolyte. There is considerable interest in developing next-generation lithium batteries based on a lithium metal anode and solid-state electrolyte. While several commercial ventures have indicated that there is demand for such a product and significant capital to back them, lithium metal batteries have not been widely implemented due to high cost and poor cyclability compared to lithium-ion technology. We are motivated to study solid electrolytes which are compatible with lithium metal anodes. The majority of this thesis is focused on a nanostructured block copolymer electrolyte with phase-separated mechanically rigid and ion conducting domains: polystyrene-block-poly(ethylene oxide) (SEO) with lithium bis(trifluoromethanesulfonyl)imide salt (LiTFSI). Polystyrene (PS) provides mechanical rigidity which aides stable lithium deposition during battery charging and poly(ethylene oxide) (PEO) solvates the lithium salt and conducts ions between the electrodes. A variety of morphologies are accessible by tuning the volume fraction of one phase and the chain length. The viability of SEO/LiTFSI electrolytes in lithium metal batteries has been well established in the literature. However, there are many open questions about the mechanisms by which SEO facilitates the transportation of lithium ions between electrodes. Our goal is to shed light on the phenomena which govern the performance of block copolymer electrolytes via rigorous characterization of their ion transport properties and transient structure. The primary tools used in our analysis are electrochemical techniques involving a potentiostat to study the rate of ion transport under different conditions and small angle X-ray scattering (SAXS) to characterize block copolymer structure at equilibrium and while ions flow through the electrolyte. Unlike simple electronic conductors (e.g., a copper wire) which have only one charge carrier (electrons), a battery electrolyte has two charge carriers (cations and anions). When implementing an electrolyte into a lithium-ion or lithium metal battery, the electrodes are designed such that only lithium cations can cross the electrode/electrolyte interface. The unavoidable consequence is that passing ionic current through the electrolyte (i.e., charging or discharging the battery) results in accumulation of ions near one electrode and depletion of ions near the other. In other words, a salt concentration gradient develops along the axis perpendicular to the planar electrodes. Under galvanostatic conditions (constant current), thermodynamics requires that a larger driving force (over potential) be required to maintain the same current as the salt concentration gradient evolves. Under potentiostatic conditions (constant voltage), thermodynamics requires that a smaller current density be passed through the electrolyte as the salt concentration gradient evolves. Consequently, the nature of the salt concentration gradients which develop during battery operation are closely tied to material performance. In this dissertation, we characterize salt concentration gradients in symmetric cells composed of two lithium electrodes sandwiched around a polymer electrolyte. After introducing lithium-based batteries, polymer electrolytes, and SAXS in Chapter 1, we develop a framework for quantifying electrolyte performance in the presence of small salt concentration gradients. Measurement of the ionic conductivity, κ, by ac impedance spectroscopy characterizes ion transport in the absent of concentration gradients. In the presence of concentration gradients, electrolyte performance is quantified by the product of κ and the current fraction, ρ_+. The current fraction is the ratio of the steady state current, i_ss, to the initial current, i_0, in a potentiostatic experiment. We rank order a list of electrolytes reported on in the literature based on κρ_+ and find that there is a tradeoff between total ionic conductivity and selective cation transport. In Chapter 3, we explore the relationship between ρ_+ and the cation transference number, t_+, to show how the two parameters dictate ion transport at early and long times when an electrolyte of initially uniform salt concentration is subjected to ionic current. Negative transference numbers have been reported in the literature, however the result has not been definitively proven. Our analysis leads us to propose a cell based on anion exchange membranes which could be used to conclusively prove the result. In Chapter 4, we return our attention to SEO/LiTFSI block copolymer electrolytes and extend our discussion of electrochemical characterization beyond Chapter 2. By measuring κ, ρ_+, the salt diffusion coefficient, D, and the equilibrium potential of concentration cells, U, we fully characterize ion transport in a series of SEO/LiTFSI electrolytes with various morphologies and chain lengths to solve for the transference number with respect to the solvent velocity, t_+^0. We obtain negative values of t_+^0 over a wide range of salt concentrations and morphologies. Universal relationships are uncovered for ρ_+ and dU/(d ln⁡m ) which apply for any PEO/LiTFSI or SEO/LiTFSI electrolyte. The result allows us to predict salt concentration gradients for any SEO/LiTFSI electrolyte by simply measuring κ as a function of salt concentration. An important consideration when studying any electrochemical system is the interactions which take place at the electrode/electrolyte interface. For PEO-containing electrolytes in contact with a lithium electrode, we found that lithium metal is sparingly soluble in PEO. The primary evidence is obtained from nuclear magnetic resonance spectroscopy (NMR), which is discussed in Chapter 5. We show that this phenomenon is detectable by phase transitions in a low molecular weight SEO/LiTFSI electrolyte. We take this result into account when studying the phase behavior of SEO/LiTFSI electrolytes in lithium symmetric cells in the Chapters 6 and 7. In Chapter 6 and 7, we describe simultaneous polarization and SAXS experiments on SEO/LiTFSI electrolytes to study the transient structure in the presence of ionic current and salt concentration gradients. Our methodology allows us to study the nanostructure of the electrolyte as a function of time and distance from an electrode during polarization. In Chapter 6, we focus on a low molecular weight (3.1 kg mol-1) SEO/LiTFSI electrolyte which has accessible phase transitions over a wide salt concentration window. Under dc polarization, the initially lamellar morphology transforms into a disordered morphology near the negative electrode where salt concentration is lower, and pockets of the gyroid morphology form near the positive electrode where salt concentration is higher. The emergence of the gyroid phase is unexpected as it is only thermodynamically favored at significantly higher salt concentrations than should be accessible based on the initial salt concentration. We hypothesize the presence of the gyroid morphology is indicative of localized salt dense pockets which form at bottlenecks in the block copolymer structure. In Chapter 7, we study a higher molecular weight (39 kg mol-1) SEO/LiTFSI electrolyte which exhibits a lamellar morphology at all salt concentrations studied. The domain spacing, d, increases with salt concentration, and we study the change in d as a function of position in the cell. By measuring d at several current densities, we show that the spatial gradient in domain spacing aligns with model predictions at low to moderate current densities based on the measurements made in Chapter 4. The behavior at large current densities indicates that the rearrangement of the structure may limit the maximum sustainable current density which can be passed through the electrolyte. In our sample which consists of a distribution of lamellar grain orientations, we find those with PS/PEO interfaces aligned perpendicular to the current are distorted more than those with interfaces aligned parallel to the current. While grains with interfaces oriented perpendicular to the current do not provide pathways for current to pass between the electrodes, we show that the distortion of the structure is critical for facilitating the formation of a salt concentration gradient. In this dissertation, we discuss the physics that are responsible for the development of a salt concentration gradient when current flows through an electrolyte. Careful characterization of transport parameters provides microscopic insight into the mechanisms of ion transport and the magnitude of salt concentration gradients. Structural characterization of block copolymer electrolytes subjected to ionic current provides fresh insight into the mechanisms which limit the rate of ion transport between the electrodes. Our characterization of the dynamic relationship between salt concentration gradients, ion transport, and nanostructure in block copolymer electrolytes provides guiding principles for the design of novel electrolyte materials aimed at enabling the lithium metal anode.