A Fluorinated, Ion-Conducting and Adaptive Supramolecular Polymer Protective Layer for Stabilizing Lithium Metal Anodes

Open AccessCCS ChemistryCOMMUNICATIONS17 Jan 2024A Fluorinated, Ion-Conducting and Adaptive Supramolecular Polymer Protective Layer for Stabilizing Lithium Metal Anodes Tao Chen†, Bo Qin†, Yuncong Liu, Zhekai Jin, Haiping Wu, Chao Wang and Xi Zhang Tao Chen† *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] Institute of Smart City and Intelligent Transportation, Southwest Jiaotong University, Chengdu 610032 Key Lab of Organic Optoelectronics and Molecular Engineering, Department of Chemistry, Tsinghua University, Beijing 100084 , Bo Qin† Key Lab of Organic Optoelectronics and Molecular Engineering, Department of Chemistry, Tsinghua University, Beijing 100084 , Yuncong Liu Key Lab of Organic Optoelectronics and Molecular Engineering, Department of Chemistry, Tsinghua University, Beijing 100084 , Zhekai Jin Key Lab of Organic Optoelectronics and Molecular Engineering, Department of Chemistry, Tsinghua University, Beijing 100084 , Haiping Wu Department of Battery Technology, BYD Shanghai Co., Ltd., Shanghai 201611 , Chao Wang *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] Key Lab of Organic Optoelectronics and Molecular Engineering, Department of Chemistry, Tsinghua University, Beijing 100084 and Xi Zhang Key Lab of Organic Optoelectronics and Molecular Engineering, Department of Chemistry, Tsinghua University, Beijing 100084 https://doi.org/10.31635/ccschem.023.202303380 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail Perfluoropolyether (PFPE) is a promising material for protective coatings on Li metal anodes due to its chemical inertness and minimal swelling in electrolytes. However, a conventional PFPE coating with poor ionic conductivity and mechanical stability is still not satisfactory for long-term cycling of Li anodes. Here, we design and synthesize an adaptive and high-conductivity supramolecular polymer (PFPE-EG-I). This polymer is constructed from PFPE chains, ethylene glycol (EG) segments, and hydrogen-bonding moieties derived from isophoronediisocyanate, serving as a multifaceted artificial solid electrolyte interphase (SEI). The incorporated EG segments enhance the Li+ conductivity of the SEI, and the hydrogen-bonding units introduce a dynamic self-adaptive behavior to the polymer matrix. A solution-processed PFPE-EG-I coating is demonstrated to promote uniform Li deposition and mitigate side reactions between Li and the electrolyte. Consequently, this leads to enhanced coulombic efficiency and prolonged cycle longevity in lithium metal batteries (LMBs). The innovative design of this multifunctional artificial SEI offers a promising avenue for the realization of dendrite-free Li anodes, paving the way for the advancement of high-performance LMBs. Download figure Download PowerPoint Introduction Lithium metal batteries (LMBs) are receiving significant attention as one of the next generations of high energy density systems.1–3 The theoretical specific capacity of the Li metal anode is 3860 mAh g−1, much higher than the graphite anode (372 mAh g−1) employed for lithium ion batteries. For the LMBs systems, achieving a safe and long-term stable Li metal anode would potentially be energy transformative. However, the implementation of Li metal anodes today still faces serious challenges due to interface instability.4,5 Li metal with low electrochemical potential (−3.04 V versus standard hydrogen electrode) is highly reactive. Thus, it easily reacts with electrolytes chemically and electrochemically once in contact. These side reaction products result in the formation of a heterogeneous solid electrolyte interphase (SEI) at the electrolyte/Li interface.6,7 In addition, Li metal as a conversion type hostless anode suffers infinite volume changes during the Li plating/stripping process. This leads to continuous crack-and-regeneration of the fragile SEI, which accumulates at the interface until the cell fails. Consequently, the inherently unstable SEI promotes nonuniform Li deposition, leading to dendrite formation. This process results in the continuous consumption of both active Li metal and electrolyte, thereby compromising the cycle longevity and safety of LMBs.8,9 Recently, a unique strategy of addressing Li anode-electrolyte interfacial instabilities is designing a polymer coating for the Li anode to regulate the Li deposition/dissolution behaviors.10–15 Perfluoropolyether (PFPE) with its [CF2CF2O]x(CF2O)y backbone exhibits low surface energy, chemical inertness, minimal swelling in the electrolyte, and the higher lowest unoccupied molecular orbital energy lever.16–18 Artificial SEIs constructed with it such as trifluoromethyl-capped PFPE oil drop or H-bonding reinforced PFPE derivatives have been used to improve the cycling stability of Li anodes.19,20 However, existing PFPE-based SEIs still have difficulties in achieving both high ionic conductivity and good mechanical stability, and the uniformity of lithium deposition needs to be further improved. In this study, we synthesize a series of polymers with the same PFPE backbones having varying graft segments which enables the systematic evaluation of the electrodeposition stability in relation to the segment structure. Meanwhile, the synthesized supramolecular polymer (PFPE-EG-I) allows the construction of an adaptive and high ionic conductivity artificial SEI for high-performance Li metal anodes. The introduction of the ethylene glycol (EG) chain can significantly increase the Li+ transport rate, and the hydrogen bond-containing segment enables the polymer to have excellent dynamic self-adaptability and adhesion. Compared to unprotected configurations (Scheme 1a), this supramolecular polymer coating has several advantageous properties: it greatly mitigates Li dendrite formation, minimizes the depletion of active Li and electrolyte, and enhances the safety and prolonged cycling stability of Li metal anodes (Scheme 1b). As a result, the PFPE-EG-I coated Li (PFPE-EG-I@Li) anode enables stable Li plating/stripping over 500 cycles in Li||Cu cells, five times longer than a bare Li anode. Furthermore, 90% capacity retention for over 300 cycles in a PFPE-EG-I@Li||LiFePO4 (LFP) full cell was achieved using ultrathin Li foils (50 μm thick) and high area loading LFP cathode (17.8 mg cm−2) sheets. The efficient artificial SEI protection design is promising for LMBs. Scheme 1 | (a) Nonuniform Li deposition and electrolyte consumption for unprotected Li metal anodes. (b) The covered supramolecular polymer (PFPE-EG-I) can suppress electrolyte decomposition and dendrite growth. Download figure Download PowerPoint Results and Discussion Characterization of PFPE supramolecular polymers Three PFPE-based polymers were synthesized (see Supporting Information Figure S1) and characterized. As shown in Figure 1a, new peaks in the range from 4.0 to 4.5 ppm corresponding to the protons near the formed urethane or ester groups were observed in the 1H NMR spectra of three polymers. Meanwhile, the peaks in the range from 3.2 to 3.4 ppm corresponding to the proton next to the isocyanate groups disappeared, indicating the formation of PFPE-based polyurethane. Additionally, the emergence of a new band at 1701 cm−1, attributed to the –C=O– functional groups within the urethane units, corroborating the synthesis of polyurethane (Figure 1b). For comparative purposes, the presence of a band at 1738 cm−1, representative of the –C=O– groups in ester units, signifies the successful formulation of a PFPE-based polyester. Figure 1 | (a) NMR, (b) Fourier transform infrared, and (c) DSC analysis of PFPE derivatives. (d–f) Rheology measurements of PFPE-I (d), PFPE-EG (e), and PFPE-EG-I (f). (g) Adhesion tests of three PFPE derivatives. (h) Ion conductivity tests of the polymers. (i) Current response of Li || Cu cells scanned at −0.3 ∼ 0.6 V. The scan rate is 1 mV s−1. Download figure Download PowerPoint Figure 1c shows the differential scanning calorimetry (DSC) results of PFPE-based polymers. As a control, the prepared PFPE-EG did not show a clear glass transition in the range of −70 to 150 °C, which was due to the lack of hydrogen bonds and the homogeneous structures between very soft PFPE segments and alkyl chains. In contrast, the Tgs of PFPE-I and PFPE-EG-I were recorded at −27.7 and −19.9 °C, respectively. This transition was ascribed to the hard segments formed by the hydrogen bonds in the polyurethane.21 Meanwhile, PFPE-EG-I displayed slightly higher Tg than PFPE-I, which was likely due to the existence of more hydrogen bonds between EG units and urethane groups. To further characterize the viscoelasticity of the polymers, rheology measurements at 1% strain over a frequency range of 0.05–100 Hz were carried out (Figure 1d–f). Both PFPE-I and PFPE-EG-I exhibit pronounced viscoelasticity, a feature attributable to the presence of hydrogen bonds. At frequencies exceeding approximately 10 Hz, their storage moduli (G′) consistently surpass their respective loss moduli (G″). Conversely, PFPE-EG does not display any elasticity within the tested frequency sweep range, adopting liquid-like characteristics, likely due to the lack of hydrogen bonding. Notably, the PFPE-EG-I polymer further demonstrates self-healing capabilities upon needle puncture (see Supporting Information Figure S2). In addition, the adhesive performance of PFPE-based polymers was characterized by the interfacial adhesion tests (Figure 1g). It is noteworthy that the adhesion energy of PFPE-I and PFPE-EG-I is much higher than that of PFPE-EG, which is ascribed to the abundant hydrogen bonds in the urethane motifs. Meanwhile, owing to the existence of more hydrogen bonds between EG units and urethane groups, PFPE-EG-I displays particularly high adhesion energy. In summary, the supramolecular design of the PFPE-EG-I material endows it with superior viscoelasticity and interfacial adhesion properties for the artificial SEI layer. Furthermore, the PFPE-EG-I also has excellent antiswelling in electrolyte, even after one week of immersion (see Supporting Information Figure S3). These characteristics enable the material to remain conformally adhered to the Li anode, accommodating volume fluctuations throughout the cycling process. The contact angles between electrolyte and the three PFPE-based SEI layers were in the range of 25°–35° (see Supporting Information Figure S4), indicating good interfacial compatibility. In terms of Li+ ion conductivity, electrochemical impedance spectroscopy measurements show low ionic conductivity for PFPE-I (1.4 ± 0.7 × 10−8 S cm−1) at 25 °C after mixing 10 wt % lithium bis(trifluoromethanesulfonyl)imide (LiTFSI) due to the lack of ion-conducting EG chain segments in the network. In contrast, considerable ion conductivities for PFPE-EG (4.9 ± 0.4 × 10−6 S cm−1) and PFPE-EG-I (1.1 ± 0.4 × 10−6 S cm−1) are presented (see Supporting Information Figure S5). On the other hand, PFPE-EG-I showed higher conductivity (4.8 ± 0.8 × 10−4 S cm−1) compared to PFPE-I and PFPE-EG after soaking in the electrolyte used for battery testing (1 M LiTFSI in 1,3-dioxolane/1,2-dimethoxyethane (v/v = 1:1) with 5 wt % LiNO3) due to the presence of ion-conducting EG chain and hydrogen-bond adhesion segments in the network (Figure 1h). The high Li+ ion conductivity of the SEI facilitates an increased critical Li nucleate size, allowing for uniform Li deposition and reducing parasitic reactions with Li metal. Cyclic voltammetry measurements of Li||Cu half cells are performed to investigate the electrochemical reversibility in the batteries with different PFPE-based artificial SEI protection. The current response in the cell with PFPE-EG-I artificial SEI increased significantly (Figure 1i), corresponding to the rapid Li+ transport and more reversible reaction kinetics. This is also consistent with the results of the ionic conductivity tests. The Li anode/electrolyte interface Figure 2 shows the morphology of the lithium on the control (bare Cu) and polymer-modified electrodes. Upon the deposition of 3 mAh cm−2 of lithium, the structures observed on the bare Cu electrode exhibited dendritic, needle-like formations and were not tightly adhered to the electrode. Notably, the areas demarcated in yellow indicate regions where metallic Cu remains exposed on the surface (Figure 2a and see Supporting Information Figure S6a). The nonuniform Li deposition would consume more Li and electrolyte in undesirable side reactions, which tends to accelerate cell failure. When a layer with high adhesion and low ionic conductivity was coated on the Cu (PFPE-I@Cu) electrode, dense but irregular Li deposition was presented (Figure 2b). As the ionic conductivity of the coating increased (PFPE-EG@Cu), the deposited Li showed large sphere-like structures with some porosity (Figure 2c). A further increase in ionic conductivity and adhesion of the coating (PFPE-EG-I@Cu) promoted more homogenous and dense lithium deposition (Figure 2d and see Supporting Information Figure S6b). These scanning electron microscopy (SEM) observations confirm that the PFPE-EG-I design improves the lithium growth morphology significantly. Figure 2 | SEM images of 3 mAh cm−2 of lithium deposited at 0.2 mA cm−2 on (a) bare Cu, (b) PFPE-I@Cu, (c) PFPE-EG@Cu, and (d) PFPE-EG-I@Cu electrodes. Download figure Download PowerPoint To investigate the chemical composition of SEI, Cu electrodes with and without polymer coating after Li stripping were detected by the X-ray photoelectron spectroscopy (XPS) (Figure 3). In O 1s spectra, the SEI layer on the bare Cu electrode showed three peaks: CH2-O-C O-CH2 (∼531.2 eV), R OCO2Li (∼530.8 eV), and metal-O (∼528.0 eV) corresponding to exposure of Cu. The appearance of the metal-O peak illustrated the irregular Li plating/stripping behaviors, which is also consistent with the previous SEM result. On the contrary, no metal-O peaks were detected on the other three polymer-coated Cu electrodes except for the addition of the -CF2- O- (∼534.7 eV) and CH2- O-CH2 (∼532.8 eV) peaks, supporting the improved uniformity of Li deposition. From C 1s spectra, the main species on the bare Cu electrode contained C–C/C–H (∼284.6 eV), C–O (∼285.9 eV), C=O (∼287.3 eV), and –CO3− (∼289.9 eV). The three other Cu electrodes with PFPE-based polymer coating had additional signals of –CF2– peaks (∼294.0 and ∼292.4 eV), and the ratios of –CO3− in the C-containing components for the polymer-coated Cu electrodes were significantly lower than that in the bare Cu electrode, indicating that the decomposition of the electrolyte was suppressed. In F 1s spectra, the SEI of polymer-coated Cu electrodes was dominated by PFPE chains (–CF2–, ∼687.7 eV and ∼688.3 eV), whereas that of bare Cu contained only the LiF (∼684.5 eV) component derived from the reactions between Li metal and TFSI− anion. Besides, in the SEI components of all electrodes, the sulfur-based species were derived only from the reductive decomposition of the LiTFSI salt. Thus, we can directly distinguish the degradation of electrolyte at different electrodes. From the S 2p spectra, the sulfur-based species on the bare Cu electrode were identified as -SO2-F (∼168.9 eV) and Li2S (∼160.5 eV). In contrast, only the SO2-F component, which is attributed to the incomplete decomposition of LiTFSI, was detected on the PFPE-I@Cu and PFPE-EG@Cu electrodes. Notably, this component was less pronounced on the PFPE-EG@Cu electrode. Intriguingly, no decomposed sulfur-based species were observed on the PFPE-EG-I@Cu electrode, suggesting effective inhibition of LiTFSI decomposition. We further analyzed the electrode in depth by XPS and still no decomposition of the sulfur-based species peaks was observed. However, the LiF component derived from the reactions between Li metal and PFPE-EG-I gradually increased with etching depth (see Supporting Information Figure S7), which also contributes to the improvement of the interfacial stability of the electrode. Figure 3 | O 1s, C 1s, F 1s, and S 2p XPS profiles of bare Cu and different artificial SEI-coated Cu electrodes when 3 mAh cm−2 areal capacity of predeposited Li was stripped at 1 mA cm−2. Download figure Download PowerPoint Overall, compared to the bare Cu electrode, the PFPE chains were the dominant components of the SEI on the three other polymer-coated Cu electrodes. These artificial SEIs could mitigate the electrolyte penetration and prevent continuous reactions between Li and electrolyte. Among them, the PFPE-EG-I polymer provided the best protection for Li metal because of its higher ionic conductivity and stronger adhesion, followed by the PFPE-EG and PFPE-I polymer. Electrochemical cycling performance As a representative of these artificial SEIs, the PFPE-EG-I coating on the Li electrode exhibits a smooth and uniform surface, which facilitates the homogeneous Li+ nucleation and distribution, and this coating layer was around 5 μm thick (see Supporting Information Figure S8). Meanwhile, the PFPE-EG-I@Li electrode also exhibits superior air stability over bare Li (see Supporting Information Figure S9), even at 49% humidity. To evaluate the electrochemical stability of the SEI on Li metal, we assessed the performance of batteries incorporating PFPE-I, PFPE-EG, and PFPE-EG-I polymers as artificial SEIs. This was done by analyzing the Coulombic efficiencies (CEs) of Li plating/stripping in Li||Cu half cells (Figure 4a and see Supporting Information Figure S10). At a current density of 1 mA cm−2 with an area capacity of 1 mAh cm−2, uncoated lithium metal exhibited a precipitous decline in its CE. Over the span of 110 cycles, from the 10th to the 110th cycle, it sustained an average CE of 97.8% before the cell underwent failure. In contrast, all the polymer-coated Li cells showed improvement in cycle life. The cells with PFPE-I@Li and PFPE-EG@Li electrodes could be stably cycled for 160 cycles and 200 cycles, with an average CE of 97.4% (cycle 10 to 160) and 97.6% (cycle 10 to 200), respectively. More significantly, the CE of the PFPE-EG-I-coated Li cell was maintained at ∼98.3% for over 500 cycles. These results showed that increasing the ionic conductivity of artificial SEI could improve the cell CE due to more uniform Li stripping/plating behaviors. Also, enhancing the interfacial adhesion of the artificial SEI could greatly improve the cycle life of the cell. Figure 4 | Electrochemical characterizations. (a) CE evolution of Li||Cu half Cells using the bare Li and polymer-coated Li electrodes at 1 mA cm−2 with a plating capacity of 1 mAh cm−2. (b) Cycling stability of Li||LFP full cells using the bare Li and polymer-coated Li electrodes at 0.5 C with Li thickness of 50 μm and LFP loading of 17.8 mg cm−2. (c) Rate performance of bare Li||LFP and PFPE-EG-I@Li||LFP full cells at various rates from 0.2 to 4 C. Corresponding the charge/discharge curves at different rates for bare Li (d) and PFPE-EG-I@Li electrode (e). Download figure Download PowerPoint Because of the artificial SEI's protection, the improved Li deposition stability can be further demonstrated by cycling a limited Li (50 μm thick) with a highly loaded LFP cathode (17.8 mg cm−2). Figure 4b and Supporting Information Figure S11 showed the cycling performance of the Li||LFP cells at 0.5 C. Although the cell with bare Li delivered a higher initial discharge capacity of 148.0 mAh g−1 due to relatively fast interfacial ion transport initially, it decayed rapidly to 115.4 mAh g−1 after 80 cycles, with a low-capacity retention of only 77.9%. For electrodes designated as PFPE-I@Li, PFPE-EG@Li, and PFPE-EG-I@Li, which exhibit progressively increasing ionic conductivity, the initial discharge capacities of the assembled cells were measured at 135.6, 138.7, and 142.2 mAh g−1, respectively. This indicated a positive correlation between the interfacial ion transport rate and the cell discharge capacity. Besides, the cells with PFPE-I@Li and PFPE-EG@Li electrodes could be cycled stably for 180 and 230 cycles, at which the discharge capacities were 131.6 and 134 mAh g−1, respectively. Remarkably, cells utilizing the PFPE-EG-I@Li electrode demonstrated stability over 300 cycles, maintaining an impressive capacity retention of 98.7%. These results suggest that utilizing an artificial SEI characterized by high ionic conductivity and enhancing the SEI's interfacial adhesion to mitigate delamination during cycling can substantially improve the long-term cycling stability of Li metal batteries. Furthermore, enhanced rate performance for the cell with PFPE-EG-I@Li electrode was observed with reversible capacities of 153.4, 142.8, 133.4, 118.5, 108.3, and 98.1 mAh g−1 under current densities of 0.2, 0.5, 1, 2, 3, and 4 C, respectively, superior to the performance of the cell with bare Li which only showed a discharge specific capacity of 60.4 mAh g−1 at 4 C (Figure 4c). The corresponding charge/discharge curves at different rates for the cells with bare Li and PFPE-EG-I@Li electrodes were shown in Figure 4d,e, respectively. In LFP full cells at elevated current densities, the PFPE-EG-I@Li electrode exhibited enhanced capacity utilization and reduced voltage hysteresis compared to the uncoated Li electrode. It indicates that the PFPE-EG-I polymer, as an artificial SEI layer, can effectively regulate the Li/electrolyte interface, ensuring a more efficient Li+ transport pathway. This strategy also demonstrated improved electrochemical performance in a conventional carbonate electrolyte system (1.0 M LiPF6 in ethylene carbonate/diethyl carbonate with 5 wt % fluoroethylene carbonate). The PFPE-EG-I@Li||LiNi0.8Co0.15Al0.05O2 (NCA) cell exhibited excellent cycling stability as well as lower polarization compared to the control Li||NCA cell (see Supporting Information Figure S12). The above suggests that the PFPE-EG-I artificial SEI has excellent compatibility in both ether and carbonate electrolytes, realizing effective protection of the lithium anode. Conclusion In summary, a multifunctional supramolecular polymer, PFPE-EG-I, has been designed for application as an artificial SEI in LMBs. Comprising PFPE chains, EG segments, and hydrogen-bonded units, this polymer simultaneously provides excellent resistance to electrolyte swelling, selective permeability to Li+, and dynamic adaptability of the SEI. The PFPE-EG-I coating effectively mitigates electrolyte infiltration, minimizes side reactions at the Li metal-electrolyte interface, and ensures homogeneous Li deposition. Remarkably, with the application of this coating, the cycle life of the PFPE-EG-I@Li anode in Li||Cu cells exceeded 500 cycles—a fivefold improvement over an uncoated Li anode. Moreover, PFPE-EG-I@Li||LiFePO4 full cells, employing ultrathin Li foils (50 μm) and high area loading LiFePO4 cathodes (17.8 mg cm−2), retained 90% of their capacity after 300 cycles. This innovative artificial SEI design offers a potent approach to addressing the challenges of interfacial instability in LMBs with ether- or carbonate-based electrolytes. Supporting Information Supporting Information is available and includes experimental procedures, synthesis process, Li anode morphology, interfacial analysis and electrochemical curves. Conflict of Interest There is no conflict of interest to report. Funding Information This research was made possible as a result of a generous grant from National Natural Science Foundation of China (grant no. 22075164) and Fundamental Research Funds for the Central Universities, Southwest Jiaotong University (grant no. 2682023CX005). Acknowledgments The authors wish to acknowledge and for their with the and the Liu Li Liu Anodes for and Wang The of Lithium Metal Liu Wang Liu Liu Tao of and in Lithium Metal Zhang Li Metal and Zhang Wang for Lithium Metal Battery and in Zhang on Wu Wang C. Zhang in on Lithium Metal Liu Wang for Lithium Metal Anodes in Li Li Li C. Layer for Lithium Metal Liu Liu Liu Liu Metal Anodes with an Adaptive Protective Wang C. Wang Liu and Supramolecular Polymer Protective Layer and Lithium Metal Anodes by on of Wu Li Zhang Liu Wang as for Lithium Metal Wang Li Wang Wang for Lithium Metal under and in V. for Lithium Zhang Zhang Li Li the of an Organic to Lithium Metal of Polymer at for Stabilizing Lithium Metal Li as a with to a Li Metal Wang with and Information solid electrolyte authors wish to acknowledge and for their with the and the times