The fabrication process of the g-C3N4@COF-3D network/polymer electrolyte is illustrated as shown in Fig. 1. First, g-C3N4 nanosheets were prepared by the thermal polymerization of urea. Then the g-C3N4@COF heterojunction filler could be obtained by in-situ growth of COF nanoparticles on g-C3N4 nanosheets. In order to further reinforce the composite polymer electrolyte, a robust 3D network with high mechanical strength was constructed by using g-C3N4@COF heterojunction filler as reinforcement phase. Finally, both of the g-C3N4@COF heterojunction filler and the 3D network were introduced into the polymer electrolyte to fabricate the g-C3N4@COF-3D network/polymer electrolyte.
As shown in Fig. S1, the g-C3N4 shows a multi-layered structure with thin nanosheets. The unique layered structure endows the g-C3N4 with large surface areas, which can provide large filler/polymer interfaces for fast ion transport. In addition, the nitrogen atoms in surface of the g-C3N4 can absorb TFSI− and thus promote the dissociation of LiTFSI for more mobile Li+. Figure 2a and 2b show SEM images of the COF material, which are nanoparticles with uniform size. Figure 2c and 2d show SEM images of the g-C3N4@COF heterojunction filler with the g-C3N4/COF ratio of 1:1, there is severe agglomeration of COF nanoparticles in g-C3N4@COF heterojunction filler. The high content of COF leaded to the uneven and uncontrollable COF growth.
when the g-C3N4/COF ratio increased to 1.5:1, COF nanoparticles are uniformly distributed on the nanosheets of g-C3N4, which can effectively avoid the aggregation of COF nanoparticles in the polymer electrolyte. Moreover, the uniformly distributed COF nanoparticles in g-C3N4@COF heterojunction filler can promote Li+ transport because of nanopores and ordered channels inside COF nanoparticles. Further increasing the g-C3N4/COF ratio to 2:1, the COF nanoparticles are unevenly distributed on the surface of g-C3N4 nanosheets. The low ratio of COF resulted in limited growth kinetics for COF formation, which leaded to the uneven distribution of COF nanoparticles on the surface of g-C3N4 nanosheets.
Figure 3a shows the XRD patterns of the g-C3N4, COF nanoparticles and g-C3N4@COF heterojunction fillers with different g-C3N4/COF ratios. The as-prepared g-C3N4 exhibits characteristic diffraction peaks at 12.9° and 27.6°, which correspond to (100) and (110) crystal planes of the typical g-C3N4 material, indicating the successful fabrication of the sheet-like g-C3N4. As for the COF nanoparticle, it shows amorphous nature. Therefore, the diffraction pattern of the g-C3N4@COF is similar to that of the g-C3N4, illustrating the amorphous nature of COF nanoparticles. In order to investigate the potential of g-C3N4@COF heterojunction fillers, composite polymer electrolytes based on g-C3N4@COF heterojunction fillers with different g-C3N4/COF ratios were constructed. SEM images of the composite polymer electrolytes based on g-C3N4@COF heterojunction fillers with different g-C3N4/COF ratios are shown in Fig. S2.
Figure 3b shows the XRD patterns of the composite solid electrolytes based on g-C3N4@COF heterojunction fillers with different g-C3N4/COF ratios. The PEO-LiTFSI electrolyte shows diffraction peaks at 19.0° and 23.3°. After introduction of the g-C3N4@COF heterojunction fillers, the peak intensity of the composite solid electrolytes decreases to some degree. Of note, the peak intensity of the g-C3N4@COF(1.5:1)-PEO-LiTFSI electrolyte is the lowest. The decreased peak intensity suggests that g-C3N4@COF heterojunction filler with g-C3N4/COF ratio of 1.5:1 can reduce the crystallinity of the polymer electrolyte, thereby promoting the transport of Li+. Moreover, Li+ transport of composite solid electrolytes based on g-C3N4@COF heterojunction fillers with different g-C3N4/COF ratios was investagated, and the testing results are shown in Fig. 3c. The g-C3N4@COF(1.5:1)-PEO-LiTFSI electrolyte possesses the smallest ionic impedance, indicating the fastest Li+ transport inside the g-C3N4@COF(1.5:1)-PEO-LiTFSI electrolyte. According to the EIS results, the high-temperature ionic conductivities of composite solid electrolytes based on g-C3N4@COF heterojunction fillers with different g-C3N4/COF ratios were calculated, respectively. The ionic conductivity of the g-C3N4@COF(1:1)-PEO-LiTFSI electrolyte is 2.66×10− 4 S cm− 1 at 60 ℃. And the ionic conductivity of the g-C3N4@COF(1.5:1)-PEO-LiTFSI electrolyte increases to 4.04×10− 4 S cm− 1 at 60 ℃. However, with the further decrease of the proportion of the COF to 2:1, the ionic conductivity of the g-C3N4@COF(2:1)-PEO-LiTFSI electrolyte reduces to 3.26×10− 4 S cm− 1 at 60 ℃. The EIS results suggest that the optimal g-C3N4/COF ratio in g-C3N4@COF heterojunction filler is of 1.5:1, which facilitates the fast Li+ transport. In addition, the electrochemical stability of composite solid electrolytes based on g-C3N4@COF heterojunction fillers with different g-C3N4/COF ratios was tested by linear sweep voltammetry (LSV). All three composite solid electrolytes display remarkable electrochemical stability, their electrochemical windows exceed 4.5 V.
Moreover, t (Li+) (Li+ transference number) of the composite polymer electrolytes based on g-C3N4@COF heterojunction fillers with different g-C3N4/COF ratios were tested and the results are shown in Fig. 4. The t (Li+) of the g-C3N4@COF(1:1)-PEO-LiTFSI electrolyte is 0.43 (Fig. 4a), and the t (Li+) of the g-C3N4@COF(1.5:1)-PEO-LiTFSI electrolyte increases to 0.47 (Fig. 4b). For the g-C3N4@COF(2:1)-PEO-LiTFSI electrolyte, the t (Li+) decreases to 0.44 (Fig. 4c). The t (Li+) testing results indicate that the optimal g-C3N4/COF ratio in g-C3N4@COF heterojunction filler is 1.5:1, which helps to realize selective Li+ transport.
Moreover, in order to further improve the room-temperature ionic conductivity of g-C3N4@COF-PEO-LiTFSI electrolytes, succinonitrile (SN) was introduced into the composite polymer electrolyte. The corresponding characterization and electrochemical tests are shown in Fig.S4-9. Although the introduction of SN can improve the ionic conductivity, SN simultaneously degrades the mechanical strength and electrochemical stability of the composite polymer electrolyte. Therefore, the robust 3D network was introduced into the composite polymer electrolyte to fabricate the g-C3N4@COF-3D network/polymer electrolyte.
As shown in Fig. 5a, the 3D network consists of nanofibers and shows a cross-linked structure, the large pores of the 3D network facilitate the infiltration of the polymer electrolyte. More importantly, the robust 3D network can provide mechanical support and reinforce the polymer electrolyte for suppressing lithium dendrite growth. The SEM image of the g-C3N4@COF-3D network/polymer electrolyte is shown in Fig. 5b, the polymer electrolyte fills up the 3D network and the 3D network is fully incorporated in the polymer electrolyte. The g-C3N4@COF-3D network/polymer electrolyte presents a dense and smooth surface without obvious pores and cracks, which can increase the stability and safety of batteries.
Besides, XRD patterns of the 3D network, the g-C3N4@COF/polymer electrolyte and the g-C3N4@COF-3D network/polymer electrolyte are shown in Fig. 5c. The 3D network exhibits a characteristic diffraction peak at 20.4°. For g-C3N4@COF/polymer electrolyte, there are two peaks between 18° to 25°, which reflects the presence of the polymer electrolyte. The diffraction pattern of the g-C3N4@COF-3D network/polymer electrolyte is similar to that of the g-C3N4@COF/polymer electrolyte, and there is a peak at 20.4°, reflecting the introduction of the 3D network. Notably, the peak intensity of the g-C3N4@COF-3D network/polymer electrolyte decreases slightly after introducing the 3D network. The decreased peak intensity of g-C3N4@COF-3D network/polymer electrolyte suggests that the introduction of the 3D network can reduce the crystallinity of the polymer electrolyte, which helps to promote the transport of Li+. Mechanical strength of polymer electrolytes is of great significance for the safety of cells. Figure 5d shows stress-strain curves of the g-C3N4@COF/polymer electrolyte and the g-C3N4@COF-3D network/polymer electrolyte. The tensile strength of the g-C3N4@COF/polymer electrolyte is 1.87 MPa. In contrast, after introduction of the 3D network, the tensile strength of g-C3N4@COF-3D network/polymer electrolyte increases to 8.613 MPa, which helps to inhibit the growth of lithium dendrites for long-life cycle.
The ionic conductivities of as-prepared composite electrolytes in this work are estimated by testing electrochemical impedance spectroscopy (EIS). Figure 6a compares the EIS curves of the g-C3N4@COF/polymer electrolyte and the g-C3N4@COF-3D network/polymer electrolyte at 30℃. According to the EIS results, the ionic conductivity of the g-C3N4@COF/polymer electrolyte is 1.39×10− 4 S cm− 1at 30℃. And the ionic conductivity of the g-C3N4@COF-3D network/polymer electrolyte slightly decreases to 1.25×10− 4 S cm− 1 at 30℃. Although the introduction of the 3D network sacrifices the ionic conductivity of the composite electrolyte to some degree, the 3D network reinforces the composite electrolyte, which can improve the mechanical strength and the electrochemical stability. Figure 6b shows LSV curves of the g-C3N4@COF/polymer electrolyte and the g-C3N4@COF-3D network/polymer electrolyte, respectively. The electrochemical stability window of the g-C3N4@COF/polymer electrolyte is 4.6 V. In contrast, electrochemical stability window of the g-C3N4@COF-3D network/polymer electrolyte reaches up to 5.0 V, indicating the increased electrochemical stability of the g-C3N4@COF-3D network/polymer electrolyte. The increased electrochemical stability can be attributed to the introduction of the 3D network, which can match high-voltage electrodes for higher energy density.
Moreover, Li-Li cells with different electrolytes were fabricated to investigate the t (Li+), and the corresponding results are as shown in Fig. 6c and 6d. The t (Li+) of the g-C3N4@COF/polymer electrolyte and the g-C3N4@COF-3D network/polymer electrolyte are 0.42 and 0.40, respectively. The enhanced t (Li+) helps to promote homogeneous Li deposition, which plays an important role on inhibition of the dendrite growth.
To demonstrate the practical possibility of the g-C3N4@COF-3D network/polymer electrolyte, LiFePO4 was selected as the cathode material and lithium metal was selected as the anode material to fabricate LiFePO4//Li cells. Figure 7a displays the charge-discharge curves of the first cycle of as-assembled LiFePO4//Li batteries with the g-C3N4@COF/polymer electrolyte and the g-C3N4@COF-3D network/polymer electrolyte. The initial discharge specific capacity of the LiFePO4//Li battery with the g-C3N4@COF/polymer electrolyte is 149.17 mA h g− 1, and the corresponding Coulombic efficiency is 93.38%. The charge-discharge curve is smooth without obvious fluctuation, indicating the reversible electrochemical process. And the LiFePO4//Li battery with the g-C3N4@COF-3D network/polymer electrolyte shows the initial discharge specific capacity of 126.36 mA h g− 1 with the charge-discharge efficiency of 84.18%. Although the LiFePO4//Li battery with the g-C3N4@COF/polymer electrolyte presents higher initial discharge specific capacity and Coulombic efficiency, the cycling stability needs to be further improved. For example, the LiFePO4//Li battery with the g-C3N4@COF/polymer electrolyte maintains a capacity retention of 84.48% after 21 cycles. In contrast, the LiFePO4//Li battery with the g-C3N4@COF-3D network/polymer electrolyte displays remarkable cycling stability with capacity retention of 99.71% after 600 cycles. The above test results demonstrate the considerably enhanced electrochemical stability of the g-C3N4@COF-3D network/polymer electrolyte for long-life Li metal batteries. The g-C3N4@COF-3D network/polymer electrolyte shows high mechanical strength, which can effectively suppress the dendrite growth. In addition, the higher electrochemical stability of the g-C3N4@COF-3D network/polymer electrolyte helps to suppress the side reactions between the solid electrolyte and the electrodes (both of cathode material and lithium metal).