Understanding a Single-Li-Ion COF Conductor for Being Dendrite Free in a Li-Organic Battery

In addition to improving ion conductivity and the transference number, single-Li-ion conductors (SLCs) also enable the elimination of interfacial side reactions and concentration difference polarization. Therefore, the SLCs can achieve high performance in solid-state batteries with Li metal as anode and organic molecule as cathode. Covalent organic frameworks (COFs) are leading candidates for constructing SLCs because of the excellent 1D channels and accurate chemical-modification skeleton. Herein, various contents of lithium-sulfonated covalently anchored COFs (denoted as LiO3S-COF1 and LiO3S-COF2) are controllably synthesized as SLCs. Due to the directional ion channels, high Li contents, and single-ion frameworks, LiO3S-COF2 shows exceptional Li-ion conductivity of 5.47 × 10−5 S · cm−1, high transference number of 0.93, and low activation energy of 0.15 eV at room temperature. Such preeminent Li-ion-transported properties of LiO3S-COF2 permit stable Li+ plating/stripping in a symmetric lithium metal battery, effectively impeding the Li dendrite growth in a liquid cell. Moreover, the designed quasi-solid-state cell (organic anthraquinone (AQ) as cathode, Li metal as anode, and LiO3S-COF2 as electrolyte) shows high-capacity retention and rate behavior. Consequently, LiO3S-COF2 implies a potential value restraining the dissolution of small organic molecules and Li dendrite growth.


Introduction
Due to the high mechanical strength and inflammable features of solid-state electrolytes (SSEs), the solid-state Li metal batteries matched cathode materials with highly specific capacity and SSEs can achieve high-energy density and safety properties. Therefore, SSEs play a crucial role in realizing solid-state batteries [1][2][3][4][5]. Among various SSEs, single-Li + conductors (SLCs) show high ion conductivity and transference number. Moreover, the polarization of concentration difference can be addressed because there are no randomly removable anion and solvent in SLCs. Notably, the multiple species of freely mobile anion and solvent molecule inevitably trigger ununiform Li + deposition and dendrite formation on the Li metal surface [6][7][8][9]. Consequently, designing and constructing advanced SLCs without anion and solvent are urgently needed.
Although COF-based SLCs have been reported and studied such as TpPa-SO 3 Li with ion conductivity of 2:7 × 10 −5 S · cm −1 and transference number of 0.9 [32] as well as LiCON-3 with ion conductivity of 3:21 × 10 −5 S · cm −1 and transference number of 0.92 [33], the ion conductivity and transference number are still far below expectations. The main reason is that finite active site is anchored onto COFs, generating a lower Li + content [35,36]. Besides, these Li +conducting COFs are established by injecting Li salts or solvent molecules into the frameworks ( Figure S1a and 1b), resulting in anion-caused interfacial side reactions, which impedes the Li + movement [37][38][39]. Furthermore, the multicomponent of solvent and anion in the COF-based SSEs might generate ambiguously interfacial reaction mechanism [40][41][42]. Therefore, it is critically necessary to accurately modify and design a multiactive center for increasing ion conductivity and transference number.
Herein, we designed various COF-based SLCs by covalent modification of active-site SO 3 Li into the pores of COFs without additional padding Li salts ( Figure S1c). The synthetic route and chemical structure of single-ion COF conductors (LiO 3 S-COF1 and LiO 3 S-COF2) are provided in Figure 1. Particularly, the anionic framework could avoid the interfacial side-reaction and restrain dendrite growth. As a result, LiO 3 S-COF2 permits stable Li + plating/stripping in Li/Li symmetric battery. Furthermore, the Li + migration's dynamics and distance are improved. It leads to big ion   3 Li 2 Research conductivity and small activation energy. The Li + migration behaviors are fully deduced by DFT theoretical calculation. Moreover, the electrolyte is successfully implemented in solid-state organic Li battery. This strategy can hinder Li dendrite growth and avoid the organic cathode dissolution, leading to excellently cyclic and rate performances. We therefore anticipate that this study will lead to the development of a new technology that will enable COFbased single-ion conductors in solid-state batteries. Therefore, the characteristic peak of C-N is detected by the Fourier transform infrared (FT-IR) spectrum. Figure S2-S3 display obvious peaks for HO 3 S-COF1 and HO 3 S-COF2. It is ascribed to the stretching vibration of C-N [43]. The C-N bonds indicate successful condensation reaction between −CHO and −NH 2 groups and complete conversion into COFs. Moreover, the complete chemical conversion is further proved by the C=C bonds (1572 cm −1 for HO 3 S-COF1 and 1576 cm −1 for HO 3 S-COF2) [44,45]. The crystallinity of obtained HO 3 S-COF1 and HO 3 S-COF2 is investigated with powder X-ray diffraction (PXRD) measurements. For HO 3 S-COF1, Figure 2(a) shows three obvious peaks at 2θ of 4.76, 8.05, and 26.4°. They are attributed to the (100), (110), and (001) faces [43]. For further study structure, Pawley refinement is implemented. Low values of R wp = 2:67% and R p = 2:08% ( Figure 2(a), pink) suggest high in accordance with experiment ( Figure 2(a), green). In comparison with alternative AA and AB models, the HO 3 S-COF1 adopts the AA stacking ( Figure S4). No peaks of starting materials of Tp and Pa-1SO 3 H appear in HO 3 S-COF1 ( Figure S5), demonstrating complete conversion. Consequently, the synthesis of the targeted sample of LiO 3 S-COF1 is implemented by suspending the HO 3 S-COF1 and Li 2 CO 3 . The LiO 3 S-COF1 ( Figure S6) has similar PXRD with HO 3 S-COF1 after replacement of H atom by Li + . For the HO 3 S-COF2, two obvious peaks at   Figure 2(b). The two peaks are attributed to (100) and (001) faces [46,47]. Moreover, the Pawley refinement ( Figure 2(b), pink) with low values of R wp = 2:57% and R p = 1:98% indicates a good structure consistent with experimental data (Figure 2(b), green). HO 3 S-COF2 has AA stacking structure ( Figure S7). No peaks of starting materials of Tp and Pa-2SO 3 H exist in the HO 3 S-COF2 ( Figure S8). And the PXRD pattern of lithium-sulfonated LiO 3 S-COF2 has a similar profile to HO 3 S-COF2 ( Figure S9). Moreover, the solid-state 13  The microstructures of prepared COFs are studied by a scanning electron microscope (SEM) and transmission electron micrographs (TEM). Wirelike micrographs of HO 3 S-COF1 can be found in SEM and TEM images ( Figure S17a-d). After lithiation, the wirelike morphology is not changed (Figure S17b-d) and the high-resolution TEM of the LiO 3 S-COF1 image shows a smooth structure, indicating no Li metal nanoparticles or clusters on the LiO 3 S-COF1 surface. The EDS mappings of LiO 3 S-COF1 can also suggest that the elements are uniformly distributed ( Figure S17g-l). For the HO 3 S-COF2 and LiO 3 S-COF2, a similar morphology can be observed in SEM and TEM images (Figures 3(a)-3(f)). In addition, the C, N, O, and S elements are uniformly distributed (Figures 3(g)-3(l)). No 4 Research obvious micrograph changes of the prepared COF before and after lithiation suggest outstanding structure stability.

Electrochemical
Performances. According to the above result, the SLCs of LiO 3 S-COFn are successfully synthesized. In order to detect the chemical environment of Li + in LiO 3 S-COF2, the solid-state 7 Li NMR spectrum is used and illustrates a singlet at 1.03 ppm (Figure 4(a)). This result proves an equivalent chemical environment [33,34]. In addition, the single peak demonstrates Li + dissociation, fast diffusion, and migration [48]. The electrochemical performance for LiO 3 S-COF2 is studied. The self-standing pellet is obtained by the cold-pressing method. Electrochemical impedance spectroscopy (EIS) is employed to investigate the Li + transportation ability of LiO 3 S-COF2. Nyquist plots of LiO 3 S-COF2 are recorded under various temperatures (20 to 80°C) (Figure 4(b)). The EIS curve shows a semicircular profile made by plots of real component (Z) versus the  For the LiO 3 S-COF1, the EIS is displayed in Figure S18. It is obviously found that LiO 3 S-COF1 exhibits larger resistances than LiO 3 S-COF2. Consequently, the calculated Li + conductivity at 20°C is 2:67 × 10 −5 S · cm −1 according to impedance results as shown in Table S1. According to the Arrhenius plots (Figure 4(c)), LiO 3 S-COF2 and LiO 3 S-COF1 show low Ea value of 0.15 eV and 0.18 eV, respectively. The obtained small Ea is adjacent to other Li + conductors, certifying the directional Li-ion migration channel. The Li + transference number (t Li + ) is measured to be 0.93 for LiO 3 S-COF2 and 0.91 for LiO 3 S-COF1 ( Figure S19) at 20°C using potentiostatic polarization means, demonstrating Li + contribution to the ion conductivity [49]. Besides, the electrochemical window is studied by the linear sweep voltammetry (LSV). The LSV is recorded under a sweep rate of 10 mV · s -1 in a voltage range from 0.5 to 6.0 V (vs. Li/Li + ) at room temperature. As demonstrated in Figure S20, the current density keeps constant until the voltage is a higher than 4.3 V for LiO 3 S-COF2. Thus, LiO 3 S-COF2 exhibits an electrochemical window of 4.3 V than LiO 3 S-COF1 (4 V) within published work [33], indicating a wide range of working voltage. The Ea, Li + conductivity, and t Li + for LiO 3 S-COF1 and LiO 3 S-COF2 are compared and displayed in Table S1. The LiO 3 S-COF2 conductor shows better electrochemical behaviors than LiO 3 S-COF1, mainly caused by the more active-center content on the LiO 3 S-COF2. More notably, the comparison of Li + conductivity, Ea, and t Li + indicates distinctly better electrochemical performance for LiO 3 S-COF2 than other reported materials (Table S2). Compared with Li-ion conductors with propylene carbonate (PC), ethylene carbonate (EC) solvent, and Li salt (Table S2), LiO 3 S-COF2 presents remarkable Li + conductivity. This is because LiO 3 S-COF2 has numerous easily accessible sites and well-designed directional Li-ion channel. LiO 3 S-COF2 has Higher t Li + than others (Figure 4(d)), implying that it has outstanding single-Li +conducting behavior.
We further assess the application of LiO 3 S-COF2 in solid-state Li metal batteries. The prepared LiO 3 S-COF2 SSE film shows high flexibility (exhibited in Figure 4(e) inset). As shown in Figure 4(e), the Li metal electrode assembled by symmetric lithium metal battery configuration (inset of Figure 4(e)) is made to assess the practical use of the LiO 3 S-COF2 conductor. So, the galvanostatic Li plating/ stripping is studied under a current density of 50 μA·cm −2 for 1 h each cycle. Figure 4(e) demonstrates stable Li plating/ stripping behaviors more than 450 h without obvious fluctuation of potential. The prominent result is caused by the single-Li-ion-conducting behavior. The stability of the framework for LiO 3 S-COF2 after electrochemical measurement is investigated by PXRD and FT-IR. From the FT-IR ( Figure S21) and PXRD ( Figure S22) patterns before and after the test, no obvious changes can be observed. This result implies that LiO 3 S-COF2 has excellent skeleton stability.
The potential application of LiO 3 S-COF2 is further evaluated by constructing an organic battery. Universally, the organic carbonyl cathodes cause plentiful attentions in green electrode for rechargeable Li-ion batteries [50][51][52], as these compounds show eco-friendly and renewable great preponderance. However, the dissolution of this electrode material in liquid electrolyte will inevitably cause fast decay of specific capacity and hinder cyclability and rate behaviors. This issue could be resolved by fabricating solid-state batteries. Therefore, the performance of LiO 3 S-COF2 is further evaluated in the Li-organic battery. The organic battery is made of Li metal as anode and anthraquinone (AQ) as cathode. The AQ has ultrahigh capacity in traditional LIBs. Nevertheless, the capacity of AQ cathode decays distinctly and shows poor performances. This issue can be worked out in the solid-state LIBs [33]. Therefore, we assemble AQ | LiO 3 S-COF2 | Li solid cell and measure its electrochemical performances. Liquid electrolyte (10 μL LiPF 6 in EC/DEC v/v = 1 : 1) is added to the electrode surface to improve interface contact [30]. Figure S23 exhibits the galvanostatic charging-discharging curves at a current density of 500 mA·g −1 . The cathode AQ displays a discharging-specific capacity of 228 mAh·g −1 , lightly lower than the liquid cell AQ | LiTFSI | Li with 278 mAh·g −1 . Additionally, the AQ | LiO 3 S-COF2 | Li has better rate behaviors ( Figure S24) and cycling stability (Figure 4(f)) than AQ |LiTFSI| Li battery (Figure 4(f) and Figure S25). The excellent cycling stability of AQ|LiO 3 S-COF2|Li is caused by restraining the dissolution of AQ cathode in liquid electrolyte.
Li dendrite growth and dead Li on the Li metal electrode surface are the major threats and challenges of Li metal batteries. Therefore, the surface morphology of Li metal electrodes in a symmetric battery, quasi-state-solid battery, and liquid battery is investigated by SEM technique. As shown in Figure 5(a), for the Li metal surface of symmetric battery Li| LiO 3 S-COF2 |Li, no dead Li can be observed. The result suggests that Li-ion is uniformly deposited to the surface of metal Li in LiO 3 S-COF2. For the AQ jLiT F SIj Li liquid battery, the inhomogeneous Li deposition is obviously observed by SEM after cycling ( Figure 5(b)). That is because there is no functional material to guide the uniform Li-ion flux ( Figure S26a). However, due to the welldefined channels, COF materials can achieve uniform Li + deposition ( Figure S26b). Hence, no dead Li or dendrite after cycles for AQjLiO3S − COF2jLi is supported by SEM ( Figure 5(c)), implying homogeneous Li + deposition. The aforementioned phenomenon demonstrates that LiO 3 S-COF2 displays well-done single-Li-ion conductor behavior, which can avoid unwanted side reactions [37][38][39][40]. As a consequence, the single-Li-ion conductor LiO 3 S-COF2 has potential value in a solid-state Li metal battery.

Mechanism Studies.
The density functional theory (DFT) calculation is carried out to study the Li-ion 6 Research migration routes in LiO 3 S-COF2. LiO 3 S-COF2 is a 2Dextended material. Generally, two routes are parallel and perpendicular for Li + migration, which are divided into planar and axial approaches (Figures 6(a) and 6(b)), respectively. Different pathways for Li + migration are expounded by evaluation of migratory barriers at rate-determining steps for planar and axial. Figure S27 shows the optimized Li + geometries. The oxygen atom of keto-form groups in COF promotes for the migration of Li + via the cation-dipole interaction [53,54]. As shown in Figures 6(d) and 6(f), axial shows a lower migratory barrier of 0.245 eV than planar of 0.85 eV (Figures 6(c) and 6(e)). Some conditions are beneficial for Li + migration with axial. Firstly, the pore of LiO 3 S-COF2 results in an entered condition than interplanar distance. The pore size (1.3 nm) of LiO 3 S-COF2 is larger than the Li + radius (0.076 nm) than interplanar distance (0.48 nm), as shown in Figure S27. Therefore, Li-ion migrates in the pore. Secondly, the distance of the axial pathway is shorter than the planar pathway ( Figure S28). Beyond all question, the C=O groups provide electrostatic interaction and accelerate the dissociation of −SO 3   7 Research directionally transports across the stacked channel of LiO 3 S-COF2 under the reliable assistance of O atoms and the π-electronic system.

Conclusion
In conclusion, a novel lithium-sulfonated covalent organic framework is successfully prepared as single-Li + SSE. The active center of −SO 3 Li is covalently tethered into the 1D porous channel of COF to achieve single-ion behaviors. The unique framework structure and abundant active centers are beneficial for Li + migration. Thus, LiO 3 S-COF2 has high conductivity and transference number. Li-ion migration behaviors and routes are thoroughly studied and illustrated using DFT. Based on directional ion channel, the Li-ion conductor has exceptional electrochemical performance in quasi-solid-state batteries. This strategy can solve the crucial issue of the organic cathode dissolution in liquid electrolyte. Therefore, our work might accelerate the advancement of COF-based solid-state electrolyte.

Materials and Methods
4.1. Chemicals. The chemical reagents and characterized apparatus are described in the supporting information.

Electrochemical
Measurements. All the electrochemical measurements are provided in the supporting information.

Data Availability
The data used to support the findings of this study are included within the article and the supplementary information file.