2026 Theses Doctoral

Advanced Atomic-Scale Degradation Diagnostics for Cell-Level Impact in Li Batteries

Kwon, Yongbeom

Lithium-based batteries are central to portable electronics, electrified transportation, and grid-scale energy storage because they offer high energy density, high cell voltage, and practical rate capability. However, battery performance and failure do not arise from any single material property. Instead, they emerge from coupled processes that span atomic-scale redox chemistry, interfacial reactions, electrode microstructure, and full-cell transport. As a result, the cell-level observables most commonly used to evaluate battery health, such as capacity fade, impedance growth, and voltage hysteresis, are not mechanistically unique. Similar electrochemical behavior can originate from very different combinations of structural, chemical, and morphological degradation. Advancing Li battery performance therefore requires diagnostic tools that can identify the local origins of degradation and connect them explicitly to cell-level consequences.

In this dissertation, we develop and apply a multi-scale diagnostic framework for understanding degradation in Li batteries. This approach combines solution and solid-state nuclear magnetic resonance (NMR) spectroscopy, X-ray photoelectron spectroscopy (XPS), X-ray diffraction (XRD), Raman spectroscopy, electron paramagnetic resonance (EPR), microscopy, and electrochemistry to resolve how local structure, interfacial chemistry, and morphology evolve during operation. Across the studies presented here, the central goal is to move from electrochemical observation to physical mechanism by determining which local changes initiate degradation, how they propagate through interfaces and electrode architecture, and how they ultimately appear as measurable cell-level performance loss.

In the first chapter, we introduce the origins of battery performance and failure across multiple length scales and establish why cell-level observables alone are insufficient for mechanistic diagnosis. We then discuss how advanced characterization methods provide complementary forms of information that electrochemistry cannot supply in isolation. Particular emphasis is placed on NMR as a local probe of Li environments, speciation, and dynamics, XPS as a surface-sensitive probe of interphase chemistry, and XRD as a bulk structural probe of crystallographic evolution and disorder. These methods, together with Raman spectroscopy, EPR, and microscopy, define the diagnostic framework used throughout the remainder of the dissertation.

The second chapter examines how cathode identity controls the reversibility of Li deposition and stripping in anode-free batteries. By comparing LiNi0.8Mn0.1Co0.1O2 (NMC811)/Cu and LiFePO4 (LFP)/Cu cells, we show that the cathode does more than determine energy density and operating voltage. It also shapes electrolyte decomposition pathways, interphase chemistry, and the morphology and reversibility of deposited Li. Operando 7Li NMR, three-electrode electrochemistry, XPS, in situ solution NMR, and microscopy reveal that NMC811 promotes tortuous Li deposits, acidic electrolyte decomposition products, and the accumulation of electrochemically inactive dead Li0, whereas LFP produces denser Li deposits but greater Li loss to the solid electrolyte interphase and corrosion. This work establishes that battery reversibility must be understood as a full-cell problem in which cathode-driven interfacial chemistry directly influences anode-side morphology and irreversible Li loss.

The third chapter focuses on the methodological challenge of making operando NMR more representative of practical batteries. Although NMR is exceptionally powerful for probing local Li environments and interfacial processes, most prior operando experiments have relied on plastic cells or heavily modified geometries that differ substantially from commercial battery formats. Here, we develop and evaluate operando NMR methodology for realistic coin and pouch cells, with particular emphasis on radiofrequency skin-depth limitations, bulk magnetic susceptibility effects, component placement, and signal optimization in metallic cell hardware. By systematically studying how casing materials and internal components alter the observed 7Li spectra, we establish practical design rules for collecting interpretable NMR data from intact metallic cells. With these optimized configurations, operando NMR is used to visualize Li plating and stripping and identify the accumulation of inactive Li in a realistic coin cell. In multilayer pouch cells, the same approach enables direct observation of electrolyte wetting during early cycling, including the effect of thermal treatment on accelerating the wetting process. This chapter demonstrates that operando NMR can be extended to commercially relevant cell formats and used as a noninvasive probe of degradation under realistic operating conditions.

The fourth chapter applies the multi-scale diagnostic approach to synthetic graphite anodes, where electrochemical behavior depends sensitively on graphitization history and structural disorder. Using XRD as the principal structural probe, supported by Raman spectroscopy, EPR, XPS, microscopy, and electrochemistry, we investigate how annealing temperature changes interlayer spacing, stacking order, crystallite coherence, and defect populations, and how those structural variables map onto first-cycle efficiency, diffusional behavior, and exfoliation during cycling. These studies show that synthetic graphite evolves from defect-mediated adsorption and strongly disordered Li storage at low annealing temperatures toward staged intercalation and more three-dimensionally ordered graphitic domains at higher temperatures. At the same time, increasing graphitic order can also increase susceptibility to solvent co-intercalation and structural damage, demonstrating that graphite performance is not governed by crystallinity alone. Rather, it is controlled by the balance among interlayer spacing, stacking order, defect density, and surface reactivity. This work provides a framework for relating graphitization pathway to electrochemical behavior and for understanding why local structural changes in carbon can govern practical cell failure.

In the fifth chapter, the studies in this dissertation are taken together to show that degradation in Li batteries must be diagnosed as a coupled, multi-scale phenomenon. Local changes in bonding, defect chemistry, interphase composition, and morphology do not remain confined to the scale at which they originate, but instead propagate through the full cell to produce the macroscopic performance losses measured electrochemically. Advanced characterization therefore does not replace electrochemistry; it makes electrochemical behavior mechanistically interpretable. By connecting atomic- and interfacial-scale measurements to cell-level impact, this dissertation provides both new insight into degradation pathways in Li batteries and a practical framework for diagnosing and designing next-generation electrochemical energy storage systems.