Recently,Prof. Zhenan Bao, Prof. Yi Cui and Prof. Jian Qin (corresponding author) of Stanford Universityet al. studied the performance of ether-based electrolyte solvents in lithium metal batteries by adjusting the degree of fluorination of their ether-based electrolyte solvents. In this study, three fluorinated DEE solvent molecules (F1F0, F1F1, and F1F2) with low fluorination were obtained by introducing a monofluorinated substituent into the 1,2-dioxy solvent to adjust the degree of fluorination.
These three solvents exhibit higher solvation strength and ionic conductivity than highly fluorinated electrolytes and maintain good oxidation stability. Through the whole cell test of the lithium metal anode and the nickel-rich cathode, it was found that the higher degree of fluorination is conducive to the cycling performance, and the cycle stability is in the order of F1F0 < F1F1 < F1F2. Specifically, F1F0 exhibits poor cyclic stability due to instability to the negative and positive electrodes. Although both F1F1 and F1F2 exhibit good stability to lithium metal anodes, their relative long-term oxidation stability has an impact on their performance, with F1F1 and F1F2 maintaining 80% capacity for about 20 and 80 cycles, respectively. Finally, the researchers demonstrated that in an actual lithium iron phosphate (LFP) bag battery, F1F2 is capable of achieving 90 cycles before reaching 80% capacity retention. This study demonstrates the importance of regulating the degree of fluorination of the electrolyte solvent and that the method is applicable to a wide range of cathode materials.
In the Li+Li redox reaction, the high number of transfer electrons per atomic mass and the low electrochemical potential make Li-metal an ideal anode material for high-energy-density batteries. Despite these advantages, lithium electrolysis produces dead lithium during the plating stripping cycle, as well as poor coulombic efficiency (CE) and circularity, which greatly hinder practical implementation. The key factor in achieving stable lithium deposition is the formation of a robust solid electrolyte interface (SEI) that allows for efficient Li+ transfer and uniform Li-deposition.
Brief introduction of the results
Figure 1Chemical structure of a fluorinated 1,2-dioxyethane solvent molecule.
Fluorinated ether molecules (i.e., F1F0, F1F1, and F1F2) are synthesized by SN2 reaction. Will 1Three single-solvent electrolytes were prepared by adding 2 mmol of difluoromethanesulfonimide (Lifsi) to 1 mL of solvent molecules. Among them, consistent with the authors' expectations, increasing the fluoridation degree of the ether solvent will reduce its solvation ability; However, the effect on chemical changes is not as pronounced as for highly fluorinated FDEE electrolytes, suggesting that the authors' new electrolyte improves salt solvation (or cation-anion dissociation).
Figure 2Characterization of electrolytes
The authors then used 7LI NMR to analyze the solvation structure of the salts in the electrolyte, using the LICL in D2O as the internal reference. With the increase of fluorination, the chemical shift of 7Li gradually shifted upward. This increased electronic shielding is due to a weaker cation-anion interaction due to reduced solvation capacity. Measurements from Raman showed that the F1F0 and F1F1 electrolytes were predominantly solvent-separated ion pairs (SSIPS, 720 cm-1). In F1F2, a wide shoulder peak indicates a large ratio of CIPS and AGGS, indicating a diminished solvation intensity. All three electrolytes exhibited relatively high ionic conductivity (005-0.17 ms cm), and they follow the trend f1f0 f1f1 f1f2, which is proportional to the solvation strength.
Figure 3Calculation of solvation structure
The authors further performed molecular dynamics (MD) simulations to gain insight into the solvation behavior. The O Coordination Probability as a function of the Li+ center distance, i.e., the radial distribution function (RDF), shows the tendency of the proportion of solvated O atoms from the FSI anion to F1F0 F1F1 F1F2. This trend is consistent with the order of solvation strength. Therefore, in the electrolyte system with a low degree of solvation, the overall positive and negative ion interactions are stronger.
Figure 4Lithium metal full battery performance
A representative set of battery cycle results showed a clear trend in cycle life: f1f0, f1f1, f1f2. The capacity of F1F0 rapidly decays to near zero over 30 cycles, and its CE decreases significantly despite the presence of excess lithium reservoirs, indicating that F1F0 is unstable for both positive and negative electrodes. In addition, the decay mode is the opposite of that of DEE electrolytes (fluorine-free electrolytes), which exhibit stable capacity for the first 18 cycles without significant reduction, and then rapidly decline to 80% capacity over the next 37 cycles. The authors further studied the charge-discharge curves under different cycles. The results show that the charge-discharge curves of F1F0 at the 1st, 10th and 20th cycles, in which the overpotential at the beginning of the charge-discharge is almost unchanged, while the capacity decays rapidly during the cycle, indicating that the volume and interface resistance are retained, but the counter electrode is unstable.
Figure 5Lithium metal half-cell cycle analysis
The authors then tested lithium metal by li||Stability of the Cu half-cell cycle. The modified Aurbach method was used to perform a typical evaluation of CE. The CE of F1F0, F1F1, and F1F2 were respectively. 3% and 987%。A closer look at the potential curve reveals that all three electrolytes have a "scalable" peak during the peel cycle, which is attributed to the transition in phase-to-phase dynamics. At the same time, the authors noted a significant capacity loss during the initial cleaning cycle of F1F0 and a slow increase in the peel overpotential in later cycles, indicating an increase in the cathode instability and impedance of F1F0. Notably, the peel cycle overpotential curve exhibits a spike characteristic, which may be related to the reconnection of "dead" lithium during stretching.
Therefore, the authors conclude that F1F0 has poor electrochemical stability and lithium deposition is unstable and uneven. In contrast, F1F1 and F1F2 exhibited higher CE at 96., respectively3% and 987% at 0At 5 mA cm2, the average overpotentials of F1F1 and F1F2 were 11 mV and 13 mV, respectively, and were lower than those of F5DEE (20 mV) due to their higher ionic conductivity. To further verify the stability of the lithium-ion cycle, the authors conducted a review of the li||The Cu half-cell underwent a long-term cycle. Interestingly, F1F0 exhibits an extremely low 50% CE in the first cycle, then increases and reaches a plateau value of 88% over the next 20 cycles. However, after 100 cycles, this value slowly drops to 50%. The results showed that the average CE value for the first 12 cycles was 75%, which was consistent with the modified measurements for 12 times of lithium plating stripping. Among them, F1F0 exhibits poor stability in lithium cycles and full cell cycles. On the other hand, both F1F1 and F1F2 exhibit stable lithium cycling over 200 cycles, with an average CE of 98., respectively5% and 988%。
Figure 6With li||XPS analysis of lithium deposited in Cu half-cells
Although the species are similar, a careful comparison of the relative abundances of the elements reveals that the content of O and F in F1F1 is relatively higher than that of F1F2 after the first cycle (271% to 108%，17.8% to 45%)。This difference may be due to the higher electrochemical sensitivity of monofluorinated substituents than difluorinated substituents. It is important to note that the SEI content observed in XPS is an evaluation of the results of a variety of factors: 1) the solvation of the electrolyte determines the composition of the SEI generated; 2) Dissolution of SEI in the corresponding electrolyte. A direct comparison of SEI levels between different electrolytes does not seem to make sense to elucidate their relative performance, as the observed levels are "stable" residues in a particular electrolyte.
Figure 7Long-term oxidation stability
Figure 7 shows the use of li||PT half-cell at 4Leakage current of the electrolyte over time at a constant voltage of 4 V. Under long-term high-voltage sustainment, F1F0 and F1F1 exhibit significant leakage currents, which are broken over time. In contrast, the F1F2 has minimal leakage current and remains stable over a period of more than 18 hours. Therefore, the superior oxidation stability of F1F2 compared to F1F1 is the reason why it is more stable for cycling when fully charged with the cell.
Figure 8Performance of a non-negative pouch battery
To evaluate the application of these new electrolytes in real-world batteries, the authors performed cyclic tests on commercial multilayer anode-free LFP pouch batteries with a relatively high area load of 21 mah/cm2。At 55 cycles, the volume retention of F1F1 electrolytes reached 80%. With improved oxidation stability, F1F2 can reach 90 cycles. In contrast, the state-of-the-art electrolyte F5DEE achieves a high cycle number of 110. Therefore, the F1F2 electrolyte, although it cannot surpass the F5DEE electrolyte, is suitable for study in actual batteries.