Silicon-containing lithium-ion batteries can exhibit capacity gain early in life, which makes forecasting future cell behavior difficult. We have observed these anomalous trends even in conditions where known mechanisms, such as overhang equalization and excessive electrolyte oxidation, are unlikely to be significant. Here, we combine simulations and experiments to analyze four cases that can produce increased capacity in Si cells. Three of these pathways relate to break-in processes, where improved mass and charge transport can lead to increased access to active electrode domains and decreased cell impedance. The fourth case occurs at high levels of prelithiation, when the positive electrode (PE) is completely replenished with Li+ at the end of cell discharge. We show that the commonality among these mechanisms is that the underlying transformations change the potentials experienced by electrodes at the end of half-cycles, increasing the Li+ inventory available to the cell. A quantitative framework to describe these effects is presented, enabling these ideas to be extended to other battery systems.

ABSTRACT Electrochemical activation is a critical step for optimal functioning of Li‐ and Mn‐rich (LMR) cathodes, yet the underlying mechanism for such activation remains elusive. Here, by using scanning/transmission electron microscopy (S/TEM) combined with the associated energy‐dispersive x‐ray spectroscopy (EDS) and electron energy‐loss spectroscopy (EELS), we decipher the origin of the activation enhanced electrochemical properties. We reveal that activation induces the formation of a spinel‐like phase within the C2/m domains of the LMR cathode, where the transition‐metal ions partially occupy both the tetrahedral (8a) and octahedral (16c) sites of the Fd m spinel lattice, distinguishing the spinel‐like phase from the conventional high‐voltage spinel. Systematic varying the cycling voltage reveals a critical activation voltage above which this spinel‐like phase forms, while lower voltages preserve the layered bulk structure. As the spinel‐like phase is a stable structure for electrochemical cycling, the present findings provide direct mechanistic insight into the voltage‐dependent activation process and explain how the C2/m to spinel‐like transformation upon activation contributes to the electrochemical performance of LMR cathodes, providing guidance for the rational design of Li‐rich cathodes with enhanced cycling durability.

Electrodes containing SiOx/graphite (Gr) materials are attractive as anodes for high-energy lithium-ion batteries. However, their mechanical deformation, electrochemical response, and impedance evolution during long-term cycling are strongly coupled, complicating accurate diagnosis of performance fade mechanisms. In this work, the behavior of electrochemically prelithiated SiOx/Gr anodes paired with NMC811 cathodes is systematically investigated using techniques that include in-situ dilatometry, three-electrode electrochemistry, and multiscale post-cycling microscopy. The SiOx/Gr electrode exhibits a maximum expansion of 49% upon lithiation to 10 mV vs Li+/Li, with 84% of the expansion and 91% of the capacity being reversible. In full cells, relatively stable cycling with only 12% capacity fade over 500 cycles is observed. Three-electrode experiments reveal cell-level impedance growth, which is dominated by the NMC811 cathode: the SiOx/Gr anode exhibits minimal net impedance rise and an initial impedance decrease at low potentials. Despite this apparent electrochemical stability, cross-sectional SEM, PFIB tomography, and cryo-STEM reveal irreversible anode thickening caused by the accumulation of an inorganic-rich solid electrolyte interphase (SEI) permeating the anode bulk. Electrode potential-shift analysis further demonstrates that Li+ ions released from lithium reservoirs in the prelithiated anode mask true lithium inventory loss during aging. These results demonstrate that low-expansion SiOx/Gr anodes can simultaneously exhibit favorable cycling and impedance metrics while undergoing substantial, hidden degradation, underscoring the importance of electrode-resolved diagnostics for evaluating prelithiated silicon-based anodes.

Li- and Mn-rich (LMR) layered oxides are known to exhibit a thin surface reconstruction layer, which grows during electrochemical cycling in a manner that depends on exposed crystallographic facets, cycling conditions, and electrolyte chemistry. Direct characterization of this layer has traditionally relied on high-resolution electron microscopy, which is inherently limited to small fields of view. Here, we employ four-dimensional scanning transmission electron microscopy (4D-STEM) combined with unsupervised machine-learning clustering to quantitatively map phase distributions over large areas and track their evolution in LMR cathodes during electrochemical aging. Our results show that the surface reconstruction layer consists predominantly of a rocksalt phase, whose thickness varies across different facets following activation cycling and becomes substantially thicker and more uniform during calendar aging. In contrast, a spinel-like phase is observed within the particle bulk. Large-area phase mapping and correlative high-resolution imaging reveal that this spinel-like phase preferentially nucleates at bulk crystallographic defects, including boundaries between 60 degrees-rotated layered domains and associated mixed-phase regions, rather than exclusively at the particle surface. Our findings establish a mechanistic distinction between surface-driven rocksalt formation and bulk-defect-mediated spinel nucleation while demonstrating the unique capability of 4D-STEM to provide statistically robust, mesoscale insight into complex phase-evolution processes in LMR cathodes.

Voltage-hold (V-hold) protocols have shown promise toward calendar lifetime analysis of cells with graphite (Gr) and silicon (Si) anodes. In this work, repeat V-holds are performed on Gr and Si cells paired with lithium iron phosphate cathodes to delineate their beneficial role in formation and conditioning. We find that V-hold at the top of charge supplements constant current cycling in conditioning the cell to higher capacities for both Gr and Si cells after the first V-hold. A reduced order model provides the irreversible capacity proportions of each V-hold. With each repeat V-hold, parasitic loss of lithium to the solid electrolyte interphase (SEI) decreases on both Gr and Si cells. Gr cells show the square-root-of-time capacity loss behavior within 200 h of V-hold, indicative of its fast relaxation and low impact of reference performance test cycles on the SEI growth. Lifetime estimates from repeat V-holds on Gr can reach years. Si exhibits longer transition times from kinetic to diffusion-limited SEI growth, evidenced by the 400 h and 200 h holds showing square-root-of-time and linear behavior, respectively. Lifetime predictions from repeat V-holds on Si only reach 1–2 months, highlighting its limitations. Recommended duration of V-holds for Si cells should be ≥400 h.

Understanding parasitic reactions is critical for improving the lifetime of lithium- and manganese-rich layered oxide (LMR)/graphite (Gr) lithium-ion cells. Here, calendar aging of LMR/Gr cells is investigated using a voltage-hold protocol designed to isolate electrolyte oxidation and reduction processes. All cells exhibit progressive capacity loss with increasing hold potential, accompanied by minor impedance growth and voltage fade in the LMR cathode. During voltage-holds, however, significant capacity gains are observed, which originate from electrolyte oxidation reactions that inject electrons into the cathode while Li+ ions enter the LMR lattice to maintain charge neutrality. These cathode-side electrolyte oxidation reactions strongly influence the capacity evolution and partially mask lithium inventory loss caused by anode-side reduction reactions. Our results reveal how competing oxidation and reduction pathways govern calendar aging in LMR/Gr cells and underscore the need for electrolytes with improved oxidative and reductive stability.

The volumetric changes of silicon electrodes, along with the strong adhesive properties of certain binders, can lead to plastic deformation of the current collector and create damage in the electrode coating. Here, we report a detailed study of silicon coatings on a high-tensile alloy (HTA) foil of copper with strength over twice that of conventional copper foils. The HTA current collectors with high mechanical strength can mitigate plastic deformation upon continuous cycling. At moderate areal capacities (2.5-3 mAh cm-2), conventional copper foils show significant wrinkling after only a few electrochemical cycles, whereas the HTA foils remain intact. We demonstrate viability of the HTA foils in large format xx6395 pouch cells, in which the HTA current collectors remain intact even at an areal capacity of 4.5 mAh cm-2; in contrast, wrinkles form in conventional copper current collectors increasing the likelihood of lithium plating. Computational studies show that stresses generated during cycling of silicon electrodes are very high in the current collector and at the current collector-coating interface, explaining the wrinkling of conventional Cu foils. Our studies highlight importance of current collector to solve the electrochemical and chemo-mechanical performance challenges associated with high-loading silicon electrodes.

The technology of silicon anodes appears to be reaching maturity, with high-energy Si cells already in pilot-scale production. However, the performance of these systems can be difficult to replicate in academic settings, making it challenging to translate research findings into solutions that can be implemented by the battery industry. Part of this difficulty arises from the lack of access to engineered Si particles and anodes, as electrode formulations and the materials themselves have become valuable intellectual property for emerging companies. Here, we summarize the efforts by Argonne’s Cell Analysis, Modeling, and Prototyping (CAMP) Facility in developing Si-based prototypes made entirely from commercially available materials. We describe the many challenges we encountered when testing high-loading electrodes (>5 mAh cm −2 ) and discuss strategies to mitigate them. With the right electrode and electrolyte design, we show that our pouch cells containing ≥ 70 wt% SiO x can achieve 600–1,000 cycles at C/3 and meet projected energy targets of 700 Wh L −1 and 350 Wh kg −1 . These results provide a practical reference for research teams seeking to advance silicon-anode development using accessible materials.

Accurate and efficient parameter estimation is essential for battery diagnostics and aging analysis. In this study, we compare two optimization-based approaches-gradient descent and Bayesian optimization-for extracting parameters from differential voltage analysis in lithium-ion batteries. While these techniques are widely used, their relative strengths and limitations for this application are not well understood. The study evaluates the trade-offs between these methods in terms of result quality, computational cost, and reliability within this specific application. The diagnostic results from our battery data suggest adopting gradient descent as an initial method for rapid and efficient analysis, while employing more stable optimization techniques, such as Bayesian optimization, as a verification step to mitigate potential instability. Comparing the two methods provides information on algorithmic choice, while inspiring further discussions on selecting appropriate techniques for specific research tasks.

The novelty of this study includes correlating stress with the complex structural transformation and volume changes in sodium-ion battery cathodes.

The widespread use of silicon (Si)-rich anodes in lithium-ion batteries (LIBs) is impeded by an unstable solid electrolyte interphase (SEI) incurring insufficient cell life. Fluoroethylene carbonate (FEC) additive in the electrolyte significantly improves cycle life. However, the gains on calendar life remain unclear; the SEI structure still undergoes detrimental alterations at rest. Thus, elucidating the SEI dynamics during calendar aging is critical to mitigating time-dependent capacity degradation. ATR-FTIR, XPS, and ToF-SIMS are used herein to investigate the SEI structure before and after calendar aging. Si cycled without FEC exhibits no notable SEI chemistry changes Pre- and Post-aging, leaving poor passivation as the main failure pathway. Conversely, the FEC-SEI starts as short oligomeric species from FEC/EC electroreduction prior to aging; after calendar aging, polymerized carbonates become consistently more prominent. Unexpectedly, the deposition of self-polymerized FEC species results from time exposure to the delithiated Si specifically as opposed to the lithiated surface. This unexpected finding is supported by another recent Si calendar-aging research, which albeit not investigating FEC, finds global failure of the SEI upon delithiation resulting in ~247 fold more reactive surface compared to the lithiated. Figure 1

Li- and Mn-rich layered oxides (LMRs), a class of earth-abundant materials for rechargeable Li-ion battery cathodes, crystallize into layered structures of two different symmetries: C2/m represented by Li2MnO3 and R3̅m represented by LiMn0.5Ni0.5O2. Fundamental questions about how the C2/m and R3̅m domains spatially correlate within the same oxide grain and how the C2/m stacking faults arrange themselves when this happens still remain. Here, by using integrated differential phase contrast imaging in scanning transmission electron microscopy (STEM-iDPC), we probe the structural and compositional details of a prototypical, cobalt-free LMR material, 0.3Li2MnO3·0.7LiMn0.5Ni0.5O2 (Li1.13Mn0.57Ni0.3O2). The connection between the C2/m and R3̅m domains is found to be abrupt, facilitated by the small lattice mismatch between the two structures. Stacking faults in the C2/m domains feature atomic plane shifting that accommodates stacking sequence changes, which explains why the stacking faults form in a random manner. Furthermore, a local disordering mechanism was identified to correlate with the C2/m stacking faults. Chemically, it is found that Ni coexists with Mn at the transition metal sites within the nominal Li2MnO3 domain. This study demonstrates that STEM-iDPC is a very useful tool for capturing all the elements in a single image, revealing atomic details on domain connections and stacking faults in the LMRs.

A characterization technique was developed to generate 3D chemo-mechanical reconstructions of battery microstructures. The technique was used to observe degradation mechanisms caused by aging over time in Si anode-based lithium-ion batteries.

LiNi0.5Mn1.5O4 (LNMO) is a high-capacity spinel-structured material with an average lithiation/de-lithiation potential at ca. 4.6-4.7 V vs Li+/Li, far exceeding the stability limits of electrolytes. An efficient way to enable LNMO in lithium-ion batteries is to reformulate an electrolyte composition that stabilizes both graphitic (Gr) negative electrode with solid-electrolyte-interphase and LNMO with cathode-electrolyte-interphase. In this study, we select and test a diverse collection of 28 single and dual additives for the Gr||LNMO battery system. Subsequently, we train machine learning models on this dataset and employ the trained models to suggest 6 binary compositions out of 125, based on predicted final area-specific-impedance, impedance rise, and final specific-capacity. Such machine learning-generated new additives outperform the initial dataset. This finding not only underscores the efficacy of machine learning in identifying materials in a highly complicated application space but also showcases an accelerated material discovery workflow that directly integrates data-driven methods with battery testing experiments.

Spinel-structured LiNix Mn2-x O4 (LNMO), with low-cost earth-abundant constituents, is a promising high-voltage cathode material for lithium-ion batteries. Even though extensive electrochemical investigations have been conducted on these materials, few studies have explored correlations between their loss in performance and associated changes in microstructure. Here, down to the atomic scale, the structural evolution of these materials is investigated upon the progressive cycling of lithium-ion cells. Transgranular cracking is revealed to be a key feature during cycling; this cracking is initiated at the particle surface and leads to the penetration of electrolytes along the crack path, thereby increasing particle exposure to the electrolyte. The lattice structure on the crack surface shows spatial variances, featuring a top layer of rock-salt, a sublayer of a Mn3 O4 -like arrangement, and then a mixed-cation region adjacent to the bulk lattice. The transgranular cracking, along with the emergence of local lattice distortion, becomes more evident with extended cycling. Further, phase transformation at primary particle surfaces and void formation through vacancy condensation is found in the cycled samples. All these features collectively contribute to the performance degradation of the battery cells during electrochemical cycling.

Silicon-based lithium-ion batteries exhibit severe time-based degradation resulting in poor calendar lives. This has been identified as the major impediment towards commercialization with cycle life considered a solved issue through nanosizing and protective coatings allowing over 1000 cycles of life to be achieved. In this work, rapid screening of sixteen electrolytes for calendar life extension of Si-rich systems (70 wt% Si) is performed using the voltage hold (V-hold) protocol. V-hold significantly shortens the testing duration over the traditional open circuit voltage reference performance test allowing us to screen electrolytes within a span of two months. We find a novel ethylene carbonate (EC) free electrolyte formulation containing lithium hexafluorophosphate (LiPF6) salt, and binary solvent mix of fluoroethylene carbonate (FEC), ethyl methyl carbonate (EMC) that extends calendar life of Si cells as compared to conventional EC based electrolyte. Our coupled experimental-theoretical analysis framework provides a decoupling of the parasitic currents during V-hold, allowing us to extrapolate the capacity loss to predict semiquantitative calendar lifetimes. Subsequently, cycle aging and oxidative stability tests of the EC free system also show enhanced performance over baseline electrolyte.

In this presentation we will introduce dilatometry and profilometry using Operando Energy Dispersive X-Ray Diffraction conducted at the Advanced Photon Source, Argonne National Laboratory. This approach is applied to a Li metal battery with a high-voltage NMC811 cathode but the method is suitable for any and all materials and electrodes. In the technique, micrometer-wide collimated X-rays traverse the electrode assembly, and the Bragg peaks from different crystalline phases are used to track the hard edges of the solid materials as they move due to volume changes in the cell. We visualize, with submicron resolution, the movement of the oxide cathode and microporous separator: the latter is observed through the diffraction of X-rays off the lamellae in the polymer matrix. One can map movement of the cathode and also compression in the porous separator – a task that is beyond the capabilities of mechanical dilatometers. We use our approach to show periodic expansion and contraction of metallic lithium during delithiation and relithiation of the cathode and to quantify formation of mossy lithium on the metal surface. The method is universal, easy to implement in-operando, and is selective in distinguishing between different phases in materials. Another dividend of this method is that it allows the profiling of lithiation gradients in the layered-oxide cathode. In our NMC811/Li cells, the cathode gradients are seen even with weak currents, they occur only during relithiation of the cathode, and they persist as the currents abate during constant-voltage discharge. Other particulars of our experimental approach and observations will also be discussed during the talk. Acknowledgement: This document has been created by UChicago Argonne, LLC, Operator of Argonne National Laboratory (“Argonne”). Argonne, a U.S. Department of Energy Office of Science laboratory, is operated under Contract No. DE-AC02-06CH11357.

As the need for research and development efforts to enable alternative, sustainable lithium-ion cathode material is increasing, fine-tuning of the cathode composition and local structure has been focused on the use of dopants/ coatings and varying synthesis conditions such as annealing temperature and time. While changing these parameters has been shown to have significant effect on materials performance, the relation of synthesis parameters with local structure and composition and electrochemical activity is an ongoing focus area1. Solid state NMR is an important characterization technique for LIB systems as 6Li and 7Li solid state NMR can directly “see and follow” the Li cations within the structure, providing crucial information on domain structures, lithium local order changes and metal segregation. A multinuclear NMR approach is also crucial to track the dopants and study how/ if the dopants are diffusing into the bulk or remain on the surface as unintended coatings. This work covers an overview of how multinuclear solid-state NMR can be utilized to understand cathode local structure changes with synthesis, and electrochemical cycling, how dopants play a role in transition metal segregation and cathode interphase, and they affect materials performance. The NMR characterization studies focus on lithium-ion cathodes systems including lithium and manganese rich transitional metal oxide cathodes, lithium nickel oxide and nickel manganese cobalt oxides and are assisted by density functional theory calculations in providing a better understanding of the experimental characterization data and therefore local environments, domain structures and dopant effects. References: Brandon R., Long, Jason R. Croy, Fulya Dogan, Matthew R. Suchomel, Baris Key, Jianguo Wen, Dean J. Miller, Michael M. Thackeray, Mahalingam Balasubramanian, Chem.Mater.2014, 26, 3565−3572 Dogan F., Long B. R., Croy J. R., Gallagher K. G., Iddir H., Russell J. T., Balasubramanian M., Key B., Journal of the American Chemical Society (2015), 137(6), 2328-2335 Figure 1