jwlimgroup

Li Metal Batteries (LMBs)

Non-fluorinated diluents offer a sustainable choice for localized high-concentration electrolytes, but their vast chemical space makes purely experimental screening prohibitively time- and cost-intensive. To navigate this space efficiently, we establish physically meaningful computational descriptors that quantitatively capture critical molecular interactions and chemical properties. We then construct a comprehensive dataset through quantum chemical calculations and train an uncertainty-aware machine learning model based on molecular graphs. This model provides both target property predictions and calibrated confidence estimates to ensure high screening reliability. Utilizing this uncertainty-guided high-throughput framework, we successfully screen the extensive chemical space to identify highly promising non-fluorinated diluent candidates.

SEI layer analysis

Lithium metal batteries (LMBs) represent the ultimate frontier for next-generation energy storage, yet their commercial viability is severely hindered by complex nanoscale degradation at the interface. During repeated cycling, large volume changes of lithium trigger a cascade of interfacial failures, including SEI cracking, continuous electrolyte consumption, dead Li accumulation, and catastrophic dendrite growth that leads to internal short circuits. To overcome these critical barriers, unravelling the precise degradation mechanisms at the atomic and nano-scale is imperative. To this end, we employ scanning transmission X-ray microscopy (STXM) to quantitatively analyze the chemical composition of the solid electrolyte interphase (SEI) and map the evolving lithium morphology. By correlating these nanoscale chemical and structural insights, we systematically investigate the impact of tailored electrolytes and protective layers to guide the rational design of highly stable, long-lasting lithium metal interfaces.

Liquid electrolytes

Developing advanced liquid electrolytes for stable lithium metal batteries (LMBs) requires identifying optimal molecular and electrochemical parameters across a vast chemical space. To accelerate this process, we first employ DFT-based screening to efficiently predict key molecular properties and establish robust design parameters for screening candidates. Concurrently, we investigate non-fluorinated electrolytes to identify specific parameters that ensure both environmental sustainability and chemical stability against the lithium metal anode, allowing us to propose novel eco-friendly molecules. Furthermore, corrosion phenomena and their governing parameters are systematically analyzed by integrating targeted experimental methods with machine learning (ML) models. Ultimately, this unified framework combines computational, experimental, and data-driven approaches to extract critical parameters, accelerating the rational design of high-performance liquid electrolytes.

Polymer electrolytes

Polymer electrolytes are attractive candidates for next-generation battery systems owing to their excellent processability and intimate interfacial contact with electrodes compared with other solid electrolytes. However, their practical application remains limited by low ionic conductivity and insufficient mechanical robustness, which typically exhibit a trade-off relationship. Our research addresses these limitations by developing advanced polymer electrolyte systems for the stable operation of Li-metal batteries (LMBs). In particular, we focus on electrolyte design and interfacial analysis to promote the formation of stable SEI layers on Li-metal anodes, while simultaneously enhancing ionic transport, mechanical integrity, and interfacial stability. Through this approach, we seek to contribute to safer, longer-lasting, and high-energy-density battery technologies.