The sodium dodecylbenzene sulfonate (SDBS, 98%) and TAPB (99.67%) were purchased from Tianjin Damao Chemical Reagent Factory, while BTMB (99.98%) and TP (97%) were obtained from Jilin Province Yanshen Technology Co, Ltd. BPY (97%) and glacial acetic acid (99.5%) were procured from Shanghai Macklin Biochemical Co, Ltd. Electrolytes (1 M LiPF₆) in ethylene carbonate (EC)/dimethyl carbonate (DMC)/ethyl methyl carbonate (EMC) (volume ratio of 1:1:1) and CR2032 coin type cells were purchased from DoDochem.

The SDBS (1 mg mL⁻¹ in chloroform solvent, 20 μL) was dripped into the deionized water (50 mL) in crystal dish, to be dispersed on the liquid surface for 40 min. The TAPB monomer (1 mg mL⁻¹ in 0.12 M HCl solution) was added into the liquid phase under SDBS and maintained for 1 h. And then another BPY monomer (1 mg mL⁻¹ in 0.12 M HCl solution) and acetic acid catalyst were added to complete the polymerization reaction of TAPB and BPY monomers. After 7 days, the COF_TAPB-BPY nanofilm was grown between the gas/liquid interface [24,25,26].

Similarly, the COF_BTMB-TP film was synthesized by the same method, except for the TP (2 mg mL⁻¹ in 0.12 M HCl solution) and BTMB (2 mg mL⁻¹ in 0.12 M HCl solution) monomers, without additional acetic catalyst.

The chemical structures of COF_TAPB-BPY and COF_BTMB-TP nanofilms were investigated by Fourier transform infrared spectroscopy (FTIR, Thermo Scientific Nicolet iS10 spectrometer), Raman (Renishaw qontor), X-ray photoelectron spectroscopy (XPS, Thermo Scientific K-Alpha spectrometer) test. The morphologies of COF_TAPB-BPY and COF_BTMB-TP nanofilms were observed by the optical microscope (OM, LEICA DM750M), scanning electron microscope (SEM, JEOL), transmission electron microscopy (TEM, D/MAX-2500, Rigaku) and atomic force microscope (AFM, Dimension Icon, Bruker). The surface morphologies of COF_TAPB-BPY and COF_BTMB-TP nanofilms were monitored by the scanning electrochemical microscopy (SECM, VersaSCAN, Advance Measurement Technology, InC) though a three-electrode system including COF nanofilms coated Si wafer, Ag/AgCl reference and probe in K₃Fe(CN)₆/K₄Fe(CN)₆ electrolyte (0.1 mmol L⁻¹). The lipophilicity of COF_BTMB-TP was tested by ⁷Li solid-state NMR (Agilent 600M) analysis, after soaking COF_BTMB-TP nanofilm in LiPF₆ solution.

The prepared COF_BTMB-TP anode nanofilm was transferred to a copper foil (ϕ = 12 mm) to assemble into the half-cell device, with polypropylene film and Li metal as the separator and reference electrode, respectively, using the 1 M LiPF₆ in EC/DME/EMC (volume ratio, 1:1:1) (40 μL) as the electrolyte, in the CR2032 coin cell. All the assembly process was carried out in the argon-filled glove box (Etelux, water and oxygen content of less than 0.1 ppm).

Then, the COF_BTMB-TP half-cell was treated by charge/discharge cycles at 0.1 C in the voltage range of 0‒2 V versus Li/Li⁺ on LAND battery testing system to achieve pre-lithiation of COF_BTMB-TP nanofilm anode. After pre-lithiation, the COF_BTMB-TP half-cell was transferred into the glove box and remove the negative tab. The obtained pre-lithiated COF_BTMB-TP anode nanofilm was assembled into the COF_TAPB-BPY//COF_BTMB-TP LIC, with COF_TAPB-BPY nanofilms as the cathode, in the similar process with above full-cell assembly.

Electrochemical tests contain the cyclic voltammetry (CV), galvanostatic charge/discharge (GCD), electrochemical impedance spectroscopy (EIS), which can be performed in the Princeton Applied Research electrochemical workstation (VersaSTAT 3), except GCD in the LAND battery testing system.

CV measurement of COF_TAPB-BPY//COF_BTMB-TP LIC was performed in the potential range of 0‒2 V at various scan rates from 10 to 100 mV, and GCD test with the voltage window of 0‒1.5 V at different current densities from 0.01 to 0.1 mV cm⁻². In these CV curves, the current can be divided into surface capacitance (k₁v) and diffusion-control reaction (k₂v^1/2), which can be further quantified as Eq. (1) [27]:

In order to assess the charge storage capacity of these COF_TAPB-BPY//COF_BTMB-TP LIC devices, the important index such as areal specific capacitance (C_A, mF cm⁻²), volumetric specific capacitance (C_v, mF cm⁻³), energy density (E, Wh cm⁻³) and power density (P, W cm⁻³) can be calculated based on the GCD curves, which can be derived from Eqs. (2‒5) [28,29]:

where i (A) is the discharge current, Δt (s) is the discharge time, Δv (V) is the potential windows, A (cm²) and h (cm) are the area and thickness of active material.

EIS plots of COF_TAPB-BPY//COF_BTMB-TP LIC device was tested in the frequency range of 10 mHz to 100 kHz with an applied potential amplitude of 5 mV. In addition, the EIS plots of anode COF_BTMB-TP nanofilm and bulk COF_BTMB-TP were also measured at different voltages ranging from 0.01 to 2 V versus Li/Li⁺, to determine the Warburg impedance (W_s), which is closely related with the Li⁺ diffusion coefficient (D'_Li⁺, cm³ s⁻¹), as shown in Eq. (6) [30]:

where n is the number of the electrons involved in the electrode reaction, σ_ω value is the slope of Z ~ ω^1/2 plots, C (mol cm⁻³) is the concentration of lithium-ions in the lattice (roughly estimated as electrolyte concentration).

Scheme 1 illustrates the synthesis of COF_TAPB-BPY and COF_BTMB-TP nanofilms, which are then assembled into COF_TAPB-BPY//COF_BTMB-TP nanofilm LIC. The COF nanofilms were prepared through the reversible Schiff base polymerization of BTMB and TP monomers (or TAPB and BPY) using a soft template derived the surfactant SDBS. The COF_BTMB-TP nanofilm was generated by undergoing the irreversible enol‒to‒keto tautomerization after above reversible Schiff base reaction, which only involves the chemical bond conversion, without destroying any crystallinity [31]. Considering the possible rapid energy storage kinetic process of COF_BTMB-TP nanofilm with strong electronegativity-CF₃ groups, the COF_BTMB-TP nanofilm acted as the LIC anode, to match with the capacitance-type cathode. The COF_TAPB-BPY nanofilm was applied as the LIC cathode, whose thickness was optimized by controlling the concentration of adding monomers, to obtain the high-performance COF_TAPB-BPY//COF_BTMB-TP LIC device.

The chemical structures of COF_TAPB-BPY and COF_BTMB-TP nanofilms were demonstrated by FTIR, Raman and XPS techniques. As shown in Fig. S1a, the C=N stretching and bending vibration peaks newly appear in 1623 and 1252 cm⁻¹, respectively. Additionally, the N-H stretching vibration peak at 3218 cm⁻¹ in TAPB and the-CHO stretching vibration peak at 1630 cm⁻¹ in BPY disappear. These changes can be attributed to the Schiff base reaction between the TAPB and BPY, resulting in the formation of the C=N bond in the COF_TAPB-BPY film [32]. However, the COF_BTMB-TP (Fig. S1b) contains newly generated C=O at 1631 cm⁻¹ and C-N at 1188 cm⁻¹ stretching vibration peaks, confirming the occurrence of enol-to-keto tautomerism and the accurate synthesis of COF_BTMB-TP film [33], where the C-F bonds at 1049 and 1009 cm⁻¹ are not involved in the reaction. Meanwhile, the C=N bond at 1602 cm⁻¹, aromatic C-H/C=C vibration bond severally at 1181 and 1542 cm⁻¹, deformation vibration of benzene ring at 1460 cm⁻¹ [34], can be detected in the COF_TAPB-BPY Raman spectrum (inset of Fig. 1a). This provides an initial indication of the successful polymerization between TAPB and BPY. On the other hand, in the Raman spectrum of COF_BTMB-TP (inset of Fig. 1b) without the presence of-NH₂,-CHO, or-OH groups, different vibrational bonds are detected: C-N at 846 cm⁻¹, C-H at 965 cm⁻¹, C-F at 1317 cm⁻¹, C=O at 1594 cm⁻¹, and deformation vibration of the benzene ring at 1455 cm⁻¹. These observations further confirm the Schiff base reaction between the BTMB and TP monomers. It is worth noting that in the case of COF_BTMB-TP, there is also evidence of enol‒to‒keto tautomerization, leading to the formation of C=C-N bonds.

Figure S1c shows the P-XRD patterns of bulk COF_BTMB-TP and COF_TAPB-BPY, which can be used to reflect the COF_BTMB-TP and COF_TAPB-BPY nanofilm. The diffraction peak of (100) crystal plane appear at 2θ = 2.9° for COF_BTMB-TP and 2θ = 2.2° for COF_TAPB-BPY, which can be further calculate the pore size of COF_BTMB-TP about 2.9 nm and COF_TAPB-BPY about 3.9 nm. Moreover, the structural result of COF_BTMB-TP and COF_TAPB-BPY can be also confirmed by simulation in Fig. S1d. The COF_BTMB-TP by A-A stacking arrangement exhibits the pore size of 2.9 nm, and layer spacing of 0.44 nm. The COF_TAPB-BPY by A-A stacking arrangement exhibits the pore size of 3.9 nm, and layer spacing of 0.35 nm. XPS analysis was performed to further research the chemical elements/bonds of COF_TAPB-BPY and COF_BTMB-TP. The C 1s core-level spectrum (Fig. S2a) of COF_TAPB-BPY can be divided into C−C/C=C at 284.2 eV, C=N in imine group at 284.5 eV [35], while C−F at 292 eV, C=O at 287.9 eV, C−N at 284.5 eV, C−C/C=C at 283.9 eV for COF_BTMB-TP [36], further proving the polymerization of TAPB-BPY, coupling of enol‒to‒keto tautomerization of BTMB-TP monomer pair. The N 1s spectrum (Fig. S2b) contains the C−N at 399.6 eV and C−N at 398.4 eV for COF_TAPB-BPY, while only C−N peak at 399.5 eV for COF_BTMB-TP. In addition, the unique O and F elements for COF_BTMB-TP were also analyzed as follows: C=O bond at 533.5 eV (Fig. S2c) in the O 1s spectrum, the C−F bond at 688.25 eV in F 1s spectrum, which show that the C−F bond exists in the form of semi-ionic bond in COF_BTMB-TP nanofilm (Fig. S2d).

The OM, SEM, TEM and AFM images of COF nanofilms were collected to observe their morphology. As shown in Fig. 1a, b, the OM photographs show that the smooth and uniform nanofilms can be distinguished for COF_TAPB-BPY and COF_BTMB-TP nanofilms, without obvious cracks and defects, even if the SEM (Fig. 1c, d) and TEM (Fig. 1e, f) images at a high magnification. The recognized slight folds and edges are the typical of large-area 2D nanofilm. Notably, these COF nanofilms can be transferred into different substrates without destroying the integrity of nanofilms, indicating the high mechanical strength. The thickness of COF_TAPB-BPY and COF_BTMB-TP nanofilms is measured to be about 8.3 and 4.1 nm, respectively, based on AFM test (Fig. 1g, h). In the 2D and 3D SCEM images (Fig. 1i-l), the observed maximum current signal difference resulting from the redox reaction within the 0.5 × 0.5 mm² scanning region is about 2.5 nA. It is indicated the presence of a flat and intact film across a large surface area for the prepared COF_TAPB-BPY and COF_BTMB-TP nanofilms, as determined by monitoring the distance between the nanofilm surface and probe.

In view of the absorption of special COF on Li⁺, the lithium storage performance of COF_BTMB-TP nanofilm was evaluated in a half-cell with lithium metal as the counter electrode. In Fig. S4, a pair of distinct oxidation/reduction peaks separately at 1.3 and 0.73 V can be detected in the CV curves of 0‒2 V versus Li/Li⁺ at scan rate of 0.2 mV s⁻¹, corresponding to the fact as follows: (i) the F^‒ and Li⁺ can be bonded by the electrostatic attraction, in virtue of the C-F transformation from semi-ionic into ionic bond (charge). (ii) Afterward, the F semi-ionic bond to nearby carbocation will be reformed, to attach the Li⁺ de-intercalation (discharge). As reported, the highly reversible conversion of C-F ionic/semi-ionic bond exists in the charge/discharge process of C-F contained 2D conductive materials [37], which can provide enormous lithium storage active sites and improve the cycle stability. In addition, there is no obvious variation for CV integral area in the first three cycles, proving that the high reversibility of Li⁺ intercalation/de-intercalation.

The similar-shaped CV curves at different scan rates (Fig. 2a) and the GCD curves at different current densities (Fig. 2b) are also the signal of stable electrochemical performances of COF_BTMB-TP electrode. Based on the discharge time in GCD curves, the high areal specific capacity of COF_BTMB-TP electrode is 5.15, 3.90, 3.25, 2.86 and 2.50 μAh cm⁻² at current densities of 0.01, 0.03, 0.05, 0.07 and 0.1 mA cm⁻², respectively. The COF_BTMB-TP electrode with only the thickness of 4.1 nm exhibits the extremely high volumetric specific capacity, contrasting significantly with currently reported commercial LiCoO₂ (LCO)/Li₄Ti₅O₁₂ electrode with the thickness of 1300 μm and an area specific capacity of 28.6 mAh cm⁻² [38]. This proves the excellent lithium storage capacity of COF_BTMB-TP nanofilm. The high specific capacity of COF_BTMB-TP electrode benefits from the unique highly ordered 1D channel, which enables full exposure of Li⁺ to active sites, along with the presence of abundant active sites such as C-F, C=O, C-N. Moreover, because of the thickness of COF_BTMB-TP film at nanometer level, it can overcome the problem of covering the active site by layer accumulation of powder COF material. So that all lithium storage sites can be fully utilized. In Fig. 2c, it is observed that the specific capacity retention maintains about 50% even with an increased current density from 0.01 to 0.1 mA cm⁻², and it still approaches to 90% when the current density cycles back to 0.01 mA cm⁻². The good rate capability of COF_BTMB-TP is an unattainable target especially for organic compounds. The COF_BTMB-TP electrode shows acceptable cycle performance after 1000 cycles with a specific capacity of 2.0 μAh cm⁻² at 0.01 mA cm⁻², and coulombic efficiency with no obvious deformation (Fig. 2d). This performance can be explained by the similar morphology in SEM images obtained after 300th with 1st cycle (inset of Fig. 2d). Compared to the original nanofilm, the obvious solid electrolyte interface (SEI) layer can be found in the cross section of nanofilm after 1st charge/discharge cycle, which exhibit no obvious variation after the following 300 cycles. This good rate capability and structural stability can be attributed to the enhanced structure and SEI stability, which result from the covalent-bonded rigid skeleton structure and C-F bonds.

In the ex situ XPS spectra of COF_BTMB-TP nanofilm in LiClO₄ (1 M, in EC:PC = 1:1 Vol%) alternative electrolyte (Fig. 2e), all semi-ionic C‒F bonds exist in the original state, gradually transitioning into ionic C‒F bonds on account of Li⁺ addition during charging, and further recover to semi-ionic C‒F bonds due to the Li⁺ removal during the discharge process. This realizes the highly reversible conversion of semi-ionic to ionic C‒F bonds throughout the whole lithium storage process. Similarly, in Fig. S4b, c, the lithium storage mechanism of C=O and C‒N bonds was also confirmed in the charge/discharge process. In the ⁷Li solid-state NMR spectra, the chemical shift of Li⁺ transforms from −1.622 to −1.242 ppm after adding COF_BTMB-TP nanofilm (Fig. 2f). This shift is attributed to the decreasing local electron cloud density caused by the presence of strong electron-withdrawing-CF₃ and C=O groups, confirming the interaction between COF_BTMB-TP and LiPF₆. In order to verify these Li⁺ storage active sites at the theoretical level, the corresponding lithium absorption energy values were calculated using DMol3 module based on DFT. In Fig. 2g, the Li⁺ absorption energy values for oxygen (‒1.37 eV), nitrogen (‒1.13 eV) and fluorine (‒1.31 eV) atoms in COF_BTMB-TP are higher than those of PF₆⁻ (‒0.82 eV), EC (‒0.84 eV), DMC (‒1.04 eV) and EMC (‒1.11 eV) in electrolyte. This observation indicates the strong interaction between COF_BTMB-TP and Li⁺, making it more favorable for Li⁺ to bind with COF after separation from electrolyte. In fact, it is very difficult for the Li⁺ absorption on C atoms, but C-F semi-ionic bond can enhance the Li⁺ adsorption capacity of C atoms, enabling higher capacity for energy storage [39]. The strong interaction between the functional groups on COF_BTMB-TP skeleton and Li⁺ is supported by the distribution of Mulliken bonds. The exceedingly low lattice energy of C=O (‒0.454 eV), C-N (‒0.555 eV) and C-F (‒0.301 eV) bonds make it easier for them to compete with PF₆⁻ and obtain more Li⁺. In addition, the calculated electrostatic potential distribution of COF_BTMB-TP (inset in Fig. 2g) shows that the electrons in COF_BTMB-TP skeleton can be transferred into C-F (red area), leading to high electron cloud density in C-F region. Therefore, the Li⁺ migration process from electrolyte components to COF_BTMB-TP is drawn in Fig. 2h as follows: the lithium-ions undergo the desolvation from electrolyte component due to the strong electronegative C-F, subsequently absorbing onto the COF_BTMB-TP, simultaneously affected by repulsive force of C-F on electrolyte anions.

In order to figure out the internal kinetics of COF_BTMB-TP nanofilm in energy storage process, the surface capacitance and diffusion-controlled contribution for COF_BTMB-TP electrode were analyzed at a series of scan rates. In Fig. 3a, the green-highlighted surface capacitance contribution constitutes a significant portion of CV integral area, amounting to 84% at a scan rate of 1 mV s⁻¹ and this contribution increases with the scan rates escalate. The proportion of surface capacitance contribution is high, reaching 70.1%, 76.8%, 80.2%, 82.4% and 84.0% at scan rate of 0.2, 0.4, 0.6, 0.8 and 1.0 mV s⁻¹, respectively (Fig. 3b). This phenomenon explains the excellent charge storage kinetics behavior of COF_BTMB-TP electrode, which perfectly match the capacitive kinetic process of the LIC cathode. This kinetic process can be also effectively evaluated by ion diffusion coefficient from EIS measurement. In Fig. 3c, dynamic EIS spectra of COF_BTMB-TP nanofilm at different discharge potentials exhibits the semicircular arc and straight line at high and low frequent region, respectively; these features correspond to the charge transfer resistance and ion diffusion Warburg resistance. At the ion diffusion region, these plots of Z’ versus ω^−1/2 (Fig. 3d) were summarized to calculate the slope values at different potentials (Fig. S5e), which were subsequently employed to calculate the Li⁺ diffusion coefficient based on Eq. (6). Similarly, as a comparison, the calculated Li⁺ diffusion coefficient of bulk COF_BTMB-TP can reach 5.6 × 10⁻¹¹ cm² s⁻¹ (Fig. S5f), surpassing many other electrode materials such as LiNi_0.5Mn_0.5O₂ (3.7 × 10⁻¹³ cm² s⁻¹) [40] and LFP/CZIF-8 (1.17 × 10⁻¹³ cm² s⁻¹) [41]. In view of the nano-level thickness of film, the volumetric Li⁺ diffusion coefficient at different potentials (Fig. 3e) show that COF_BTMB-TP nanofilm exhibits the significantly higher volumetric Li⁺ diffusion coefficient of 1.15 × 10⁻⁶ cm³ s⁻¹, compared with that of bulk COF_BTMB-TP (1.86 × 10⁻⁸ cm³ s⁻¹). Moreover, with the increasing potential, the attenuation of Li⁺ diffusion coefficient appears at 1.2 V versus Li/Li⁺ for bulk COF_BTMB-TP; there is no change for COF_BTMB-TP nanofilm. This suggests the wider operating voltage range for highly ordered COF_BTMB-TP nanofilm. To bridge the enormous gap of cathode/anode in the output capacity arising from different energy storage mechanisms in LICs, the thickness values of cathodic COF_TAPB-BPY are adjusted about 4.8, 6.3, 8.3, 9.0, and 9.6 nm (Fig. 3f), by varying the concentration of adding TAPB at 0.2, 0.3, 0.4, 0.5, 0.6 mmol L⁻¹ and corresponding BPY monomers (molar ratio of BPY to TAPB = 1.5) to ensure a balanced charge of Q₊  = Q₋. It can be found that the thickness of cathodic COF_TAPB-BPY is not increased significantly after adding 0.3 mmol L⁻¹ TAPB and 0.45 mmol/L BPY monomers. Correspondingly, the energy density of assembled LIC devices incorporated these cathodic COF_TAPB-BPY nanofilms shows the similar trend with the thickness of cathodic nanofilms. Therefore, the COF_TAPB-BPY nanofilm with a thickness of 8.3 nm was selected to serve as cathode in the LIC, matched with the COF_BTMB-TP nanofilm anode to ensure capacity balance. In the inset of Fig. 3f, the coulomb efficiency of the assembled LIC device decreases to a certain extent at the current density of 0.01 mA cm⁻², which can be explained by the impact of internal resistance from incomplete electrode activation at the low current density.

Since above kinetics and capacity matching, the pre-lithiated COF_BTMB-TP and optimal COF_TAPB-BPY nanofilm with the thickness of 8.3 nm were integrated into a COF_TAPB-BPY//COF_BTMB-TP nanofilm LIC device as the cathode/anode, respectively, as shown in the internal structure diagram of button-type device (inset in Fig. 4a). The COF_TAPB-BPY//COF_BTMB-TP nanofilm LIC device demonstrates the excellent capacitive performance, exhibited by the standard shuttle-like shaped CV curve (Fig. 4a) and quasi-straight GCD curves without obvious platforms (Fig. 4b). In details, both the areal and volumetric specific capacitances of LIC devices increase continuously with the thickness of cathodic COF_TAPB-BPY nanofilm rises from 4.8 to 8.3 nm. And then they remain constant even with the use of an 9.0 or 9.6 nm COF_TAPB-BPY nanofilm cathode and COF_BTMB-TP anode (Figs. 4c and 3g). All these involved data have been also evaluated by reproducibility test including 6 identical devices (inset in Fig. 4c), exhibiting the reasonable error margin, which can prove the robust feasibility of COF_TAPB-BPY//COF_BTMB-TP nanofilm LIC device. Figure 4d collects the Ragone plots of energy and power density for this COF_TAPB-BPY//COF_BTMB-TP nanofilm LIC device (318 mWh cm⁻³ at 6 W cm⁻³), together with reported film-type LICs and supercapacitors as the contrast. The COF_TAPB-BPY//COF_BTMB-TP nanofilm LIC device outperforms reported LICs in highest energy/power density, including PAF-5-LIMC (71.1 mWh cm⁻³ at 1.9 W cm⁻³) [42], LTO//AG-LIMCs-80 (53.5 mWh cm⁻³) [43], as well as supercapacitors such as Co-COF_TAPB-DHPA (230.4 mWh cm⁻³ at 5.9 W cm⁻³) [19]. Notably, this COF_TAPB-BPY//COF_BTMB-TP nanofilm LIC device achieves a comparable energy density to that of the commercial Panasonic (17,500) lithium-ion battery, which is about 340 mWh cm⁻³. This similarity is particularly pronounced at higher power densities. The Nyquist plots (Fig. 4e) of COF_TAPB-BPY//COF_BTMB-TP nanofilm LIC shows the negligible semicircular at high frequency region, indicating the low charge transfer resistance and interface impedance. At high frequency region, the linear portion with a high slope can be fitted by the equivalent circuit to obtain the low Li⁺ diffusion resistance of about 2.1 Ω. In addition, this COF_TAPB-BPY//COF_BTMB-TP nanofilm LIC device exhibits acceptable long-cycle stability, with 77% capacitance retention after 5000 cycles at 0.05 mA/cm² (Fig. 4f) and coulombic efficiency of about 100%. The common self-discharge phenomenon of COF_TAPB-BPY//COF_BTMB-TP nanofilm LIC device was tested by collecting the potential signal after 10 charge/discharge cycles and standing for 1 h at 1 V. In Fig. 4g, the COF_TAPB-BPY//COF_BTMB-TP nanofilm LIC shows slow self-discharge with up to 80 h self-discharge time in most parallel experiments. This phenomenon is primarily attributed to the presence of an energy barrier of 0.04 eV within the anode COF_BTMB-TP structure, which controls the Li⁺ diffusion rate (Fig. 4h), as reported in literature[44]. Upon comparison, the surprising electrochemical performances of COF_TAPB-BPY//COF_BTMB-TP nanofilm LIC device can be evaluated in the current level, such as the extremely high energy and power density, good cycle stability, slow self-discharge, and low resistance (Fig. 4i).