Research

"We have to evolve means for obtaining energy from stores which are forever inexhaustible, to perfect methods which do not imply consumption and waste of any material whatever"

-Nikola Tesla-

Research Areas

Our research vision is to control and design battery electrode interphases & chemistries that enable energy-dense, safe, and durable rechargeable batteries for a variety of applications.

High-energy Cell Chemistries

Electrolyte Engineering

Battery Data Science

Battery AI

High-energy Cell Chemistries

Lithium (Li) and silicon (Si) are intrinsically the most energy-dense anode materials for lithium batteries, providing far greater gravimetric and volumetric capacities than conventional graphite anodes.

Achieving next-generation, high-energy-density lithium batteries demands replacing graphite anodes with Li and/or Si. Yet, legacy Li-ion battery electrolytes engender an inadequate interphase chemistry, producing an unstable solid-electrolyte interphase (SEI) that cannot retain its integrity over repeated cycles.

Inhomogeneous SEI formation

Local SEI defect formation

Inactive SEI & Li accumulation

Uncontrolled Li-electrolyte rxn

Non-uniform Li electrodeposition

Uneven Li stripping

Thick SEI formation

Mechanical SEI fracture

SEI accumulation

Large Si expansion

Substantial Si contraction

Si structural damage

Thus, spatiotemporally stabilizing Li/Si SEIs is a critical step in enabling high-energy lithium batteries. Such interphasial instability is significantly amplified when Li/Si anodes are paired with high-energy cathodes. The high-energy cathodes include (1) Layered Oxides - Lithium Nickel Manganese Cobalt Oxide (NMC) & Lithium Nickel Cobalt Aluminum Oxide (NCA) with Ni content >80 % for both NMC and NCA, (2) Olivine Phosphates - Lithium Iron Phosphate (LFP) & Lithium Manganese Iron Phosphate (LMFP), (3) Lithium-rich Layered Oxides, (4) Disordered Rocksalt Cathodes - Cobalt-free & lithium excess cathodes, and (5) Conversion-type Cathodes - Elemental sulfur (S8) & Air (O2). Each category triggers its own form of destructive crosstalk, such as SEI poisoning, transition metal dissolution, surface reconstruction, and shuttling effects that degrade cell cycling performance and cause cell safety issues. Ultimately, the intertwined challenges of realizing practical high-energy cell chemistries boil down to the interphase control. Thoughtfully engineered electrolytes are therefore pivotal - they govern interphase stability and, in turn, unlock both long-term cyclability and robust safety.

Our group will tailor cell chemistries to each application by strategically pairing high-energy anodes and cathodes to meet precise targets for energy density and performance. This balancing act brings us to the field’s central challenge: stabilizing the electrode interphases, which we tackle through advanced electrolyte engineering.

Electrolyte Engienering

To advance the technologies required for rigorous interphase control in high-energy lithium batteries, focused research in electrolyte engineering is indispensable. Because the electrolyte remains in constant contact with a highly reducing anode and a strongly oxidizing cathode, it ultimately defines the cell’s practical electrochemical stability window. High-energy chemistries exacerbate this constraint by pairing electrodes whose operating potentials surpass the intrinsic stability range of almost every known electrolyte. Inevitably, electrolyte molecules (precisely Li-ion solvation shells for liquid electrolytes) decompose during cycling, creating SEIs that afford metastable protection, temporarily extending the usable voltage range. Identifying electrolyte formulations - and, by extension, interphase chemistries - that can reliably sustain such energy-dense systems therefore remains a formidable and urgent challenge.

Dual-phase Electrolyte

Liquid

+

Solid

Our group is committed to developing practical, yet advanced electrolytes precisely tailored to each high-energy cell chemistry. Our design approach begins with re-imagining both liquid and solid electrolytes, recognizing that conventional single-phase formulations rarely deliver the full trifecta of high ionic conductivity, intimate interfacial contact, and long-term interphase stability. By engineering dual-phase (solid–liquid) electrolytes, we can merge the complementary strengths of each phase - rapid ion transport from the liquid component and mechanical/chemical robustness from the solid - to forge interphases that remain stable under demanding operating conditions.

Designing new Molecules & Blends

Liquids

Solids

Fine-tuning

Synergy

New Electrolyte Generations

A one-formulation-fits-all electrolyte for every high-energy battery chemistry is unrealistic, so we embrace a modular, application-driven design philosophy. We systematically fine-tune molecular architectures, solvent–salt-additive families, and phase compositions while balancing the unique trade-offs each electrode pair demands. This iterative approach will yield successive generations of both single- and dual-phase electrolytes, each precisely engineered for specific cell chemistries and end-use requirements.

Battery Data Science

Empirical
Exploration

Data Acquisition
& Processing

Data-Driven
Modeling

Database Management & Integration

Predictive Insights
& Decision Making

Empirical exploration of emerging battery chemistries, electrolytes, and interphases generates vast streams of electrochemical traces, spectroscopic maps, microscopy images, and multimodal datasets, each representing significant investments of time, effort, and resources. However, venturing into previously unexplored chemical spaces often means traditional analytical frameworks can interpret only a fraction of these signals, leaving substantial portions of valuable data underutilized. As the volume and complexity of our datasets grow, critical insights risk remaining undiscovered without sophisticated analysis.

To fully leverage every piece of experimental data, our lab employs a data-centered approach that emphasizes learning from observation, integrating advanced artificial intelligence (AI), statistical analyses, and predictive modeling techniques. This methodology enables us not only to identify hidden patterns and latent features within extensive datasets but also to reveal unexpected correlations that conventional analyses might overlook. By transforming raw experimental data into actionable insights, we significantly accelerate battery innovation and discovery.

In battery data science, rigorous, standardized, and repeatable measurement protocols form the essential foundation for establishing robust, predictive relationships between material properties, electrochemical behaviors, and battery performance. Given the unpredictable outcomes associated with new battery chemistries and evolving electrode-electrolyte interphases, our battery data science research first emphasizes generating accurate, consistent, and reproducible data across all scales - from individual materials to full-cell configurations.

Leveraging this high-quality data, we deploy advanced machine-learning frameworks capable of quantifying uncertainties, guiding hypothesis-driven research, and proactively suggesting experiments likely to yield the greatest knowledge gains. By fostering an interdisciplinary environment where engineers, electrochemists, chemists, data scientists, and materials scientists collaborate closely, our lab creates a robust platform to systematically address key battery performance bottlenecks. Ultimately, our comprehensive, data-driven workflow facilitates the rapid development of safer, more durable, and higher-energy rechargeable batteries.

Battery AI

Structured Battery Data Repository

Feature Selection and Prioritization

AI & ML

Integration

Predictive Discovery

of Battery Materials

Experimental Validation & Implementation

Battery Data Science involves systematically analyzing experimental datasets using statistical and machine-learning techniques to identify hidden correlations, quantify uncertainty, and enhance understanding of existing battery chemistries and behaviors. In contrast, Battery AI expands upon these insights, employing advanced computational models, including generative algorithms and reinforcement learning, to proactively predict novel electrolyte formulations, interphase chemistries, and nanostructures of SEIs. While Battery Data Science is about extracting knowledge from data, Battery AI is about strategically leveraging that knowledge to accelerate innovation and actively guide future experimental decisions.

Our group will bridge the frontiers of High-Energy Cell Chemistry, Electrolyte/Interphase Engineering, and Battery Data Science, transforming diverse experimental data into powerful predictive tools that redefine battery development. Our Battery AI strategies start by systematically analyzing extensive multimodal datasets derived from rigorous empirical investigations of electrolyte formulations, interphase chemistries, and a variety of cell chemistry behaviors. By applying state-of-the-art machine learning algorithms (including neural networks, reinforcement learning, and generative models), we rapidly pinpoint electrolyte compositions and interphase characteristics that enhance the stability, performance, and safety of energy-dense batteries.

Through the integration of AI-driven data analytics with electrolyte/interphase engineering efforts, we proactively predict electrolyte stability windows, identify promising molecular structures, and optimize dual-phase electrolyte systems. Our computational approach will significantly reduce the trial-and-error traditionally associated with high-energy cell chemistry development, which dramatically accelerates innovation cycles.

Note that Battery AI can be more than an analytical tool. It is a transformative approach enabling the discovery of new battery electrolyte/interphase chemistries and formulations. By merging experimental rigor with computational insight, our lab not only anticipates battery performance enhancements but also guides experiments toward the most promising pathways.