English

• With the soaring consumer market of electric vehicles and wearable electronics, energy storage systems with higher energy density, higher power density and better safety are in high demand. Lithium metal batteries (LMBs) are attractive energy storage devices with the advantages of low redox potential (−3.04 V vs. SHE) and high specific capacity (3860 mAh/g) of lithium (Li) metal anodes [1]. However, the uncontrolled lithium dendrite growth of Li anodes tends to incur short circuits and cause safety issues. Building artificial SEI [2], creating three-dimensional structured anode [3], adding electrolyte additive [4], or fabricating gel/solid electrolyte [5,6], are effective strategies to protect the Li anode. Recently, the role separators paly in regulating ion transport and deposition behavior has been gaining a lot attention [7]. Functional separators possessing high porosity and good wettability towards electrolyte can increase electrolyte uptake and retention rate, conductive to rapid electrolyte permeation to establish a fast Li⁺ transport channel, thus regulate the transport behavior of Li⁺ with uniform deposition [8]. Electrospun nonwovens composed of randomly arranged fibers, with materials including cellulose [9], poly(acrylonitrile) [10,11], polyimide [12], polyvinylidene fluoride [13], poly(ether ketone) [14], etc., have high porosity and specific surface area, in addition to reduced ion diffusion pathways and enhanced ion conductivity [15,16], making them appealing materials of functional separator for protecting Li anode.

  As a component connecting the cathode and anode in a battery, advanced separators can also manipulate the cathode interface via the special chemical composition and high specific area [17]. Specifically, some positive electrode active materials based on transition metal compounds experience dissolution of transition metal ions into the electrolyte solution, which is especially pronounced in spinel structured materials such as LiMn₂O₄ (LMO), damaging the durability of batteries [18–20]. Introducing a separator with the capability to adsorb dissolved active materials will improve the cycle performance. For example, composite separators with nitrogen functionalities can trap Mn cations resulting in reduced Mn amounts on negative electrodes and improved capacity retention in both LiMn₂O₄/Li (LMO/Li) and LMO/graphite cells [21].

  Protein materials have been proven capable of anchoring anions and accelerating Li⁺ ions transport [22,23], in addition to trapping soluble active materials to block their diffusion from cathode to anode side [24–26], thus are ideal candidates for modifying or building multifunctional separator to manipulate both the cathode and anode interfaces. Gelatin protein with rich polar groups shows excellent wettability with electrolyte. Furthermore, electrospun gelatin nonwovens with benign porous structure and high specific area can guarantee fast ion transport and provide sufficient trapping sites [27,28]. However, the gelatin nonwoven fabric is delicate and fragile, which can be easily dissolved in moisture or melted when touched by sweaty hands, leading to inconvenience for handling and storage.

  Herein, we proposed an in-situ crosslinked gelatin nonwoven (CGN) as separator for LMBs, which obtains enhanced mechanical properties via the crosslinking reaction between gelatin and glutaraldehyde (GTA). The CGN separator can stabilize the Li⁺ deposition to relieve dendrite growth and anchor the active material dissolved in electrolyte to reinforce the cyclic stability, thus promoting the electrochemical performance of LMO/Li cells by stabilizing the cathode and anode interfaces. Satisfyingly, the corresponding Li/Li cells hold extremely stable overpotential of 20 mV for 1900 h at 0.5 mA/cm², and the LMO/Li cells achieve high reversible capacity of 103 mAh/g after 100 cycles at 0.3 C with high capacity-retention ratio of 83.7%, manifesting the dual positive effect and application advantages of CGN separator.

  To prepare the CGN, a gelatin nonwoven (GN) was fabricated by electrospinning, which was then placed in a sealed container with saturated GTA vapor inside to induce in-situ crosslinking reaction via the Schiff-base reaction between the amino group of gelatin and the aldehyde group of GTA [29]. The obtained crosslinked gelatin nonwoven (CGN) shows enhanced tensile property as shown in Fig. 1a. The longer the reaction time, the higher the tensile strength. However, the CGN becomes brittle when reacted for too long, reflected by the decreased elongation after 5 days' reaction. The strength, elongation and the calculated fracture energy of CGN with different crosslinking time can be seen in Fig. S1 (Supporting information). The morphology of gelatin nonwoven as-electrospun and after crosslinking is shown in Figs. 1b and c. The GN exhibits highly porous structure constructed by gelatin fibers overlapping loosely while the CGN presents slightly fiber merging to obtain a firmer and more compact nonwoven membrane with uniform pores distribution. Further elongating the reaction time, gelatin fibers merge severely and pores from fiber overlapping start disappearing leading to a flat membrane with low porosity (Fig. S2 in Supporting information). For a separator with balanced strength and toughness, in addition to nice porous structure, the sample crosslinked for 2 days is chosen as CGN in the following, if not specifically indicated.

  Figure 1

  

  Figure 1.  Physical and chemical properties of CGN separator. (a) The stress-strain curves of gelatin nonwoven with different crosslinking time. The SEM images of (b) GN and (c) CGN. (d) The FT-IR spectra of GN and CGN. (e) The contact angle between separators and liquid electrolyte. (f) Nyquist plots of stainless steel/separator/stainless steel cells. (g) Porosity and ionic conductivity of the CGN and Celgard separator. Current response and interface impedance spectra before and after polarization of (h) Li/Celgard/Li cell and (i) Li/CGN/Li cell.

  DownLoad: Full-Size Img PowerPoint

  The crosslinking reaction was also confirmed via FTIR spectra (Fig. 1d). The two peaks of GN at 1630 and 1523 cm^-1 correspond to C=O stretching (amide Ⅰ) and N—H bending (amide Ⅱ), respectively. A noticeable red shift and peak intensity reduction of amide Ⅱ groups in CGN indicate the crosslinking between GTA and gelatin and some hydrogen bonds formation. Moreover, the enhanced broad band at ~3300 cm^-1 in CGN ascribed to the overlapped stretching bands of O—H and the intra- and inter-molecular hydrogen bonds [30]. The DSC results indicate a more stable chemical property of CGN, confirming the occurrence of crosslinking reactions (Fig. S3 in Supporting information). The dimensional stability of CGN under heating treatment is also outstanding (Fig. S4 in Supporting information). With rich polar groups, the CGN shows superior wettability than the Celgard separator as indicated in Fig. 1e. The contact angle between CGN and electrolyte is 34.8° at the first contact, which gradually decreases to 21.7° after one second and eventually the electrolyte is all absorbed by the CGN and contact angle become 0° after 10 s. Conversely, the contact angle of Celgard is 68.2° at the beginning and soon stabilizes at 67.0°. The excellent wettability can lead to easy electrolyte absorption and high ionic conductivity [31]. The Nyquist plot of the stainless steel(ss)/separator/ss cells with Celgard and CGN can be found in Fig. 1f, and the corresponding ionic conductivity are calculated and present in Fig. 1g CGN separator has an ionic conductivity of 3.75 mS/cm which is more than five times higher than that of Celgard (0.57 mS/cm). The porosity of CGN and commercial separator Celgard is also tested and the CGN exhibits a high porosity of 78.7%, while that for Celgard separator is only 50.4%. Especially, CGN separator displays a high Li-ion transference number (t_Li⁺) of 0.70 characterized via AC impedance and DC polarization while that of Celgard is only 0.31 (Figs. 1h and i). The equivalent circuit models used to fit the impedance spectra can be seen in Fig. S5 (Supporting information) and the resistance and current values used to calculate the t_Li⁺ are exhibited in Table S1 (Supporting information). The positively charged amino acid residues on the side chain of gelatin can attract anions, while the backbone oxygens in gelatin polypeptide chains with good affinity to Li-ions can enable fast Li⁺ hopping [27,32], leading to the high t_Li⁺ and ionic conductivity [33].

  With the above advantages, CGN plays an excellent role in stabilizing Li metal (Fig. 2a). The symmetric cells with CGN separator achieve stable Li plating/stripping cycling for ~1900 h with an overpotential of 20 mV at a current density of 0.5 mA/cm² and capacity of 0.5 mAh/cm². On contrast, the Celgard separator shows a gradually increasing overpotential to ~100 mV during the first 900 h and then the overpotential suddenly surges which implies the failure of the cells. The Nyquist plots of Li/Li symmetric cells after different cycles were collected and shown in Figs. 2b and c to reveal the interfacial resistance of different separators. The diameter of the semicircle on the real axis at high frequency corresponds to the interfacial resistance between electrolyte and anode surface (R_SEI). Apparently, the fresh cells with CGN separator have a significantly smaller R_SEI value than that with Celgard separator. The R_SEI values of both cells decrease with cycled deposition/stripping and the CGN cells keep advantage of smaller interfacial resistances all over different cycles, indicating CGN endows uniform and stable SEI film which benefits homogeneous deposition of Li ions. The morphology of Li metal disassembled from cycle Li/Li cells also reflect the superiority of CGN separator in inhibiting dendrite. The Li metals from CGN cell are clean, shiny and flawless and that from Celgard cell have lots of tousy “dead” lithium which losses electric contact with Li metal and form the dark regions (Fig. S6 in Supporting information). The SEM images show that the surfaces of Li metal from CGN cell in both deposition and stripping state are flat and smooth. On contrast, moss-like dendrites are formed on the deposited Li metal from Celgard cell and cracks can be seen on the stripped Li metal with a rough surface (Figs. 2d and e, Figs. S7 and S8 in Supporting information). It should be mentioned that the CGN maintains fibrous structure after cycling in Li/Li symmetric cell, with unchanged morphology (Fig. S9 in Supporting information), indicating the superior stability of crosslinked gelatin in electrolyte. The fibrous CGN with 3D network and rich lithophilic sites can homogenize the Li⁺ flux distribution, decrease the Li⁺ concentration gradient thus relieve the formation and growth of Li dendrite [34].

  Figure 2

  

  Figure 2.  Li stripping/plating behaviors of CGN separator compared with Celgard. (a) Voltage profiles of Li/Li symmetric cells at a current density of 0.5 mA/cm² and an areal capacity of 0.5 mAh/cm². Nyquist plots f Li/Li symmetric cells with (b) Celgard and (c) CGN separator after different cycles. The solid lines indicate the fitting results with the equivalent circuit model shown in inset. Morphology of Li metal in deposition state from cycled Li/Li cells with (d) Celgard and (e) CGN separator.

  DownLoad: Full-Size Img PowerPoint

  The feasibility of CGN separator working in batteries was examined based on LMO/Li cells. Figs. 3a and b show the charge/discharge voltage profiles of LMO/Li cells with Celgard or CGN separator at different cycles at 0.3 C. The CGN cell exhibits small polarization reflected by the narrow gap between the platforms of discharge and charge curves. On contrast, the Celgard cell presents larger and increasing polarization with cycles. Meanwhile, the capacity of CGN cell maintains 103 mAh/g after 100 cycles at 0.3 C, with a retention rate of 83.7% and the Celgard cell suffers rapid capacity decay with a remaining capacity of 83 mAh/g and retention rate of 70.3% (Fig. 3c). Moreover, the EIS spectra of the LMO-Li cells were collected (Figs. 3d and e). The semicircle diameter at the high-to-medium frequency region represents the charge transfer resistance (R_ct). Benefited from the highly porous structure and benign wettability with electrolyte, the CGN cell at fresh state shows a smaller R_ct value (128 Ω) than that of Celgard cell (212 Ω). After 5 cycles, the CGN cell still delivers lower charge transfer resistance than Celgard cell.

  Figure 3

  

  Figure 3.  Electrochemical performance of different separators in LiMn₂O₄/Li (LMO/Li) cells (the loading of LMO is ~3.5 mg/cm²). Voltage profiles of (a) Celgard and (b) CGN separator at different cycles. (c) Cycle performance comparation at 0.3 C. Nyquist plots of the LMO/Li cells with (d) Celgard and (e) CGN separator at fresh state or after 5 cycles. The solid lines indicate the fitting results with the equivalent circuit model shown in inset.

  DownLoad: Full-Size Img PowerPoint

  To verify the interactions between CGN and Mn ions dissolved in the electrolyte of LMO/Li cell, the CGN membrane was immersed in an electrolyte with Mn(TFSI)₂ added. The XPS spectra of CGN before and after soaking were compared in Fig. 4a. The O 1s spectra can be deconvoluted into two peaks at 532.9 and 534.2 eV ascribing to C=O and C—O bonds, respectively. After interacting with Mn²⁺ ions, the two peaks both shifted to higher value of binding energy, indicating an apparent interaction between oxygen atoms of gelatin and Mn²⁺. This phenomenon suggests that the oxygen in gelatin imparted partial charges to Mn²⁺ during the process of adsorption, leading to the decreased electron cloud density and increased binding energy [35]. Fig. 4b shows the cycled CGN from LMO/Li cell and some Mn-contained species can be seen on the fiber surface, illustrating strong trapping capability of gelatin protein to dissolve active materials. As for the anode side, the Li metal protecting function of CGN separator can be seen in the SEM image of cycled Li anode in LMO/Li cell as compared in Fig. 4c. The Li metal with Celgard separator shows large holes caused from uneven striping of Li and corrosion by the Mn ions, while the surface of Li anode from CGN cell is flat and film-like. The chemical composition of the SEI from cycled Li anode is analyzed via XPS and the deconvoluted F 1s and P 2p spectra are shown in Figs. 4d and e. With CGN separator, the SEI layer contains a more inorganic Li-F component which is a superior electronic insulator and allows the transport of Li⁺ ions, benefiting to form uniform SEI layer [36]. Besides, the Li_xPO_y and Li_xPF_y that from the decomposition of LiPF₆ present different distribution. The SEI with CGN separator has more Li_xPO_y which helps reduce the interfacial desolvation energy barriers thus contributing to enhanced transport kinetics of Li⁺ ions [5]. The feasibility of CGN separator in LiFePO₄/Li cells is also examined and the CGN separator greatly improved the cycle performance (Fig. S10 in Supporting information). In addition, sulfur cathode also suffers severe shuttle effects due to the soluble intermediate polysulfides and the CGN separator can improve the capacity of lithium-sulfur (Li-S) batteries by trapping the dissolved polysulfides (Fig. S11 in Supporting information), indicating that CGN is a promising functional separator in LMBs.

  Figure 4

  

  Figure 4.  The functions of CGN at trapping active materials and protecting Li anode. (a) Deconvoluted O 1s XPS spectra of CGN before and after immersion in electrolyte containing Mn(TFSI)₂. (b) SEM images of CGN separator disassembled from cycled LMO/Li cells. (c) Morphology of Li anode surface from cycled LMO/Li cells with Celgard or CGN separator. High-resolution (d) F 1s and (e) P 2p spectra of the SEI layer on the surface of lithium metal anode in LMO/Li cells with different separators.

  DownLoad: Full-Size Img PowerPoint

  In summary, the CGN separator fabricated with gelatin protein via electrospinning and in-situ crosslinking can manipulate both the cathode and anode interfaces in LMBs to boost its electrochemical performance. On the anode interface, the CGN separator can fix the anions by the positively charged amino acid residues on the side chain of gelatin to provide a homogenous electric field, in addition to promoting Li ions transport, leading to uniform Li ions deposition and a dendrite-free anode. The CGN separator with a high ionic conductivity (3.75 mS/cm) and a high Li⁺ transference number (0.70) enables the Li/Li symmetric cell to have a long cycle life of ~1900 h. On the cathode interface, the CGN separator can anchor Mn ions and block their diffusion to anode side with the rich oxygen-containing polar groups in backbone gelatin in together with the high specific area of nonwoven structure providing sufficient trapping sites, contributing to stable cycle performance and enhanced capacity. The LMO/Li cell with CGN separator remains a specific capacity of 103 mAh/g after 100 cycles at 0.3 C. The proposed CGN fabricated with wide-spread and green biomaterial provides a new perspective for designing dual functional separators for advanced metal batteries.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  CRediT authorship contribution statement

  Weijie Cai: Investigation, Formal analysis, Data curation. Xinxin Han: Validation, Methodology, Investigation. Min Chen: Writing – original draft, Methodology, Funding acquisition, Conceptualization. Haoyuan Chen: Methodology, Investigation. Hao Wang: Methodology, Investigation. Zhixiang Chen: Visualization, Formal analysis. Mengmeng Shao: Resources, Methodology. Ke Zheng: Validation, Conceptualization. Wenlong Wang: Validation, Resources. Rui Hong: Writing – review & editing, Supervision, Project administration, Data curation. Xiaodong Shi: Writing – review & editing, Writing – original draft, Supervision, Project administration, Funding acquisition.

  Acknowledgments

  This work was supported by National Natural Science Foundation of China (No. 22309029), Guangdong Basic and Applied Basic Research Foundation (No. 2024A1515140011), Dongguan Social Development Technology Foundation (No. 20231800907933), and Collaborative Innovation Center of Marine Science and Technology of Hainan University (No. XTCX2022HYC14). Additionally, the authors acknowledge the supports of comprehensive characterizations by Dongguan University of Technology Analytical and Testing Center, Pico Election Microscopy Center of Hainan University and Shiyanjia lab (http://www.shiyanjia.com).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2025.111809.

  
  
• 1. [1]

     G. Yang, Z. Hao, C. Fang, et al., Chin. Chem. Lett. 36 (2025) 111185. doi: 10.1016/j.cclet.2025.111185

  2. [2]

     Y. Ye, Y. Zhao, T. Zhao, et al., Adv. Mater. 33 (2021) 2105029. doi: 10.1002/adma.202105029

  3. [3]

     T. Lyu, F. Luo, Z. Wang, et al., Chem. Eng. J. 466 (2023) 143357. doi: 10.1016/j.cej.2023.143357

  4. [4]

     C. Wang, X. Fu, C. Ying, et al., Chem. Eng. J. 437 (2022) 135283. doi: 10.1016/j.cej.2022.135283

  5. [5]

     P. Wang, H. Li, N. Li, et al., Adv. Funct. Mater. 35 (2025) 2413544. doi: 10.1002/adfm.202413544

  6. [6]

     Y. Zheng, Y. Shen, J. Guo, et al., Nano Res. Energy 3 (2024) e9120118. doi: 10.26599/nre.2024.9120118

  7. [7]

     A. Ye, Z. Zhu, Z. Ji, et al., Adv. Funct. Mater. 34 (2024) 2310593. doi: 10.1002/adfm.202310593

  8. [8]

     Z. Fan, X. Chen, J. Shi, et al., Nano-Micro Lett. 17 (2025) 128. doi: 10.1007/s40820-024-01596-x

  9. [9]

     H. Kim, U. Mattinen, V. Guccini, et al., ACS Appl. Mat. Interfaces 12 (2020) 41211–41222. doi: 10.1021/acsami.0c08820

  10. [10]

      T. Dong, W.U. Arifeen, J. Choi, et al., Chem. Eng. J. 398 (2020) 125646. doi: 10.1016/j.cej.2020.125646

  11. [11]

      M. Guo, H. Zhu, P. Wan, et al., Adv. Fiber Mater. 4 (2022) 1511–1524. doi: 10.1007/s42765-022-00190-3

  12. [12]

      L. Kong, X. Fu, X. Fan, et al., Nanoscale 11 (2019) 18090. doi: 10.1039/c9nr04854e

  13. [13]

      X. l. Shen, Z. j. Li, N.P. Deng, et al., Chem. Eng. J. 387 (2020) 124058. doi: 10.1016/j.cej.2020.124058

  14. [14]

      D. Li, D. Shi, K. Feng, et al., J. Membr. Sci. 530 (2017) 125–131. doi: 10.1016/j.memsci.2017.02.027

  15. [15]

      D. Ji, Y. Lin, X. Guo, et al., Nat. Rev. Methods Primers 4 (2024) 1. doi: 10.1038/s43586-023-00278-z

  16. [16]

      Y. Huang, A. Jiang, D. Wang, et al., J. Energy Storage 107 (2025) 114854. doi: 10.1016/j.est.2024.114854

  17. [17]

      F. Yang, D. Wang, Z. Zhang, et al., Energy Stor. Mater. 65 (2024) 103134.

  18. [18]

      A. Banerjee, Y. Shilina, B. Ziv, et al., J. Am. Chem. Soc. 139 (2017) 1738–1741. doi: 10.1021/jacs.6b10781

  19. [19]

      Y. Zhang, A. Hu, D. Xia, et al., Nat. Nanotechnol. 18 (2023) 790–797. doi: 10.1038/s41565-023-01367-6

  20. [20]

      F.A. Susai, H. Sclar, Y. Shilina, et al., Adv. Mater. 30 (2018) 1801348. doi: 10.1002/adma.201801348

  21. [21]

      A. Banerjee, B. Ziv, S. Luski, et al., J. Power Sources 341 (2017) 457–465. doi: 10.1016/j.jpowsour.2016.12.036

  22. [22]

      X. Fu, R. Odstrcil, M. Qiu, et al., Energy Storage Mater. 42 (2021) 22–33. doi: 10.1016/j.ensm.2021.07.010

  23. [23]

      C. Wang, X. Fu, S. Lin, et al., J. Energy Chem. 64 (2022) 485–495. doi: 10.1016/j.jechem.2021.05.014

  24. [24]

      M. Chen, X. Fu, N.D. Taylor, et al., ACS Sustain. Chem. Eng. 7 (2019) 15267–15277. doi: 10.1021/acssuschemeng.9b02383

  25. [25]

      M. Chen, C. Li, X. Fu, et al., Adv. Energy Mater. 10 (2020) 1903642. doi: 10.1002/aenm.201903642

  26. [26]

      Y.X. Tang, J.Y. Deng, W.L. Li, et al., Adv. Mater. 29 (2017) 1701828. doi: 10.1002/adma.201701828

  27. [27]

      M. Chen, Z. Chen, X. Fu, W.H. Zhong, J. Mater. Chem. A 8 (2020) 7377–7389. doi: 10.1039/d0ta01989e

  28. [28]

      M. Chen, X. Fu, J. Liu, et al., ChemSusChem 14 (2021) 2226–2236. doi: 10.1002/cssc.202100568

  29. [29]

      J.H. Bowes, C.W. Cater, Biochim. Biophys. Acta (BBA) - Protein Struct. 168 (1968) 341–352. doi: 10.1016/0005-2795(68)90156-6

  30. [30]

      S. Das, G. Dalei, Sci. Total Environ. 875 (2023) 162660. doi: 10.1016/j.scitotenv.2023.162660

  31. [31]

      C. Ma, S. Zou, Y. Wu, et al., Angew. Chem. Int. Ed. 63 (2024) e202402910. doi: 10.1002/anie.202402910

  32. [32]

      X. Fu, L. Scudiero, W.H. Zhong, J. Mater. Chem. A 7 (2019) 1835–1848. doi: 10.1039/c8ta11384j

  33. [33]

      Z. Ju, X. Tao, Y. Wang, et al., Energy Environ. Sci. 17 (2024) 4703–4713. doi: 10.1039/d4ee01359j

  34. [34]

      Z. Ju, J. Nai, Y. Wang, et al., Nat. Commun. 11 (2020) 488. doi: 10.1038/s41467-020-14358-1

  35. [35]

      J.L. Yang, H.H. Liu, X.X. Zhao, et al., J. Am. Chem. Soc. 146 (2024) 6628–6637. doi: 10.1021/jacs.3c12638

  36. [36]

      M. Chen, X. Li, W. Cai, et al., Chin. Chem. Lett. 36 (2025) 110712. doi: 10.1016/j.cclet.2024.110712

• 

  Figure 1  Physical and chemical properties of CGN separator. (a) The stress-strain curves of gelatin nonwoven with different crosslinking time. The SEM images of (b) GN and (c) CGN. (d) The FT-IR spectra of GN and CGN. (e) The contact angle between separators and liquid electrolyte. (f) Nyquist plots of stainless steel/separator/stainless steel cells. (g) Porosity and ionic conductivity of the CGN and Celgard separator. Current response and interface impedance spectra before and after polarization of (h) Li/Celgard/Li cell and (i) Li/CGN/Li cell.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Li stripping/plating behaviors of CGN separator compared with Celgard. (a) Voltage profiles of Li/Li symmetric cells at a current density of 0.5 mA/cm² and an areal capacity of 0.5 mAh/cm². Nyquist plots f Li/Li symmetric cells with (b) Celgard and (c) CGN separator after different cycles. The solid lines indicate the fitting results with the equivalent circuit model shown in inset. Morphology of Li metal in deposition state from cycled Li/Li cells with (d) Celgard and (e) CGN separator.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  Electrochemical performance of different separators in LiMn₂O₄/Li (LMO/Li) cells (the loading of LMO is ~3.5 mg/cm²). Voltage profiles of (a) Celgard and (b) CGN separator at different cycles. (c) Cycle performance comparation at 0.3 C. Nyquist plots of the LMO/Li cells with (d) Celgard and (e) CGN separator at fresh state or after 5 cycles. The solid lines indicate the fitting results with the equivalent circuit model shown in inset.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  The functions of CGN at trapping active materials and protecting Li anode. (a) Deconvoluted O 1s XPS spectra of CGN before and after immersion in electrolyte containing Mn(TFSI)₂. (b) SEM images of CGN separator disassembled from cycled LMO/Li cells. (c) Morphology of Li anode surface from cycled LMO/Li cells with Celgard or CGN separator. High-resolution (d) F 1s and (e) P 2p spectra of the SEI layer on the surface of lithium metal anode in LMO/Li cells with different separators.

  下载: 全尺寸图片 幻灯片
•