Designing New Organic Molecules as Electrolyte Solvents for Realistic Lithium Metal Batteries

Here we rationally designed and synthesized new liquid molecules without CAS number as the single electrolyte solvent. Paired with low-concentration single salt, the electrolyte enabled realistic lithium metal batteries and even industrial anode-free pouch cells.

Like Comment
Read the paper

Organic chemistry has contributed significantly to the development of batteries for decades, and particularly, the synthesis of carbonate molecules (e.g. ethylene carbonate, dimethyl carbonate, fluoroethylene carbonate, etc.) results in optimal electrolytes for modern lithium-ion batteries, which was the subject of the 2019 Nobel Prize in Chemistry. In addition to the success of lithium-ion batteries, both industry and academia are working on reviving the lithium metal battery chemistry, trying to vastly boost the battery energy density. However, the commercial lithium-ion battery electrolytes do not work well with the lithium metal electrode. The electrolyte thus becomes the Achilles' Heel of lithium metal batteries.

To tackle this problem, researchers around the world are developing new electrolytes that are compatible with the lithium metal anode as well as high-voltage cathodes. Some of these novel electrolyte systems, such as high concentration electrolytes (Nat. Commun. 6, 6362 (2015)), local high concentration electrolytes (Nat. Energy 4, 796–805 (2019)), and electrolytes with special solvent combinations (Nat. Energy 4, 882–890 (2019) and Nat. Nanotechnol. 13, 715–722 (2018)), achieved superior performance even under strict testing conditions.

Despite these achievements, there is still a need for further understanding of the rational design of electrolytes. Therefore, we get back to the old-fashioned tool, organic synthesis, trying to create solvent molecules that were seldom explored or never identified (i.e. even without CAS number) for lithium metal batteries.

Our target is to develop a single-salt, single-solvent electrolyte with standard salt concentration, so that we not only obtain the knowledge of a clear structure-property relationship but also achieve practical lithium metal batteries. We selected the ether-based molecule 1,2-dimethoxylethane (DME, Figure a) as the starting structure since it has been reported to be fairly stable towards lithium metal but oxidized at high voltage (Joule 3, 1662–1676 (2019) and Nat. Commun. 6, 6362 (2015)). We first lengthened the backbone of DME to obtain 1,4-dimethoxylbutane (DMB, Figure b) and then chose −F as the functional group to further enhance both lithium metal stability and high-voltage tolerance, yielding fluorinated 1,4-dimethoxylbutane (FDMB, Figure c). It is worth noting that we only functionalized selected positions to −CF2− (orange part in Figure c) while leaving CH3−O−CH2− groups (wathet part in Figure c) intact. This design makes FDMB distinguished from typical hydrofluoroethers that cannot dissolve lithium salts.


Figure. Design scheme and chemical structures of three solvent molecules studied in our work: DME (a), DMB (b), and FDMB (c). 


With this designed molecule as the solvent and one molar (1 M) LiFSI solely as the salt, the 1 M LiFSI/FDMB electrolyte enabled realistic lithium metal batteries under harsh testing conditions. Limited-excess Li|NMC full cells retained 90% capacity after 420 cycles with an average Coulombic efficiency of 99.98%. Industrial anode-free Cu|NMC811 pouch cells achieved ~325 Wh kg-1 specific energy while Cu|NMC532 ones showed record-high 80% capacity retention after 100 cycles.

If you are interested in detailed insight into this electrolyte (bizarre color, intriguing Li−F coordination, and Li+ solvation structure in the 1 M LiFSI/FDMB electrolyte), please check out our paper here: “Molecular design for electrolyte solvents enabling energy-dense and long-cycling lithium metal batteries”.

Last but not the least, please do not hesitate to synthesize new electrolyte solvents, salts, or additives for improving batteries!

Zhiao Yu

PhD Candidate, Stanford University