Guoying Rena,
Fengshi Cai*a,
Shoucheng Wanga,
Zhiqiang Luo*b and
Zhihao Yuanc
aTianjin Key Lab for Photoelectric Materials and Devices, School of Materials Science and Engineering, Tianjin University of Technology, Tianjin 300384, China. E-mail: caifs@tjut.edu.cn
bKey Laboratory of Display Materials and Photoelectric Devices (Tianjin University of Technology), Ministry of Education of China, Tianjin 300384, China. E-mail: zhqluo@email.tjut.edu.cn
cNational Demonstration Center for Experimental Function Materials Education, Tianjin University of Technology, Tianjin 300384, China
First published on 22nd June 2023
Covalent organic frameworks (COFs) are considered as promising candidate organic electrode materials for lithium-ion batteries (LIBs) because of their relatively high capacity, ordered nanopores, and limited solubility in electrolyte. However, the practical capacity of COF materials is mainly affected by their low electronic/ionic conductivity and the deep-buried active sites inside the COFs. Here, we synthesize an iodine doped β-ketoenamine-linked COF (2,6-diaminoanthraquinone and 1,3,5-triformylphloroglucinol, denoted as COF-I) by a facile one-pot solvothermal reaction. The introduction of iodine can make the COF more lithiophilic inside and exhibit high intrinsic ion/electron transport, ensuring more accessible active sites of the COFs. Consequently, when used as the cathode of LIBs, COF-I demonstrates a high initial discharge capacity of 140 mA h g−1 at 0.2 A g−1, and excellent cycling stability with 92% capacity retention after 1000 cycles. Furthermore, a reversible capacity of 95 mA h g−1 at 1.0 A g−1 is also achieved after 300 cycles. Our study provides a facile way to develop high-performance COF electrode materials for LIB applications.
Recently, several strategies are attempted to overcome these above-mentioned issues in COFs. For example, hybridization with conductive carbon materials such as carbon nanotube and graphene can effectively increase their conductivity and specific surface area, thus enhancing capacity and rate performance.10–13 Nevertheless, the introduction of a large amount of conductive materials reduces the amount of active material COFs in the electrode, generating practically low capacity and utilization.14 In addition, an exfoliation strategy has been developed to expose the buried redox-active sites in the pores of COF, and thus enhancing storage capacity of metal ions in rechargeable batteries.15–17 However, using the exfoliation technology to produce COF few-layer nanosheets also faces some problems such as low yield and restacking, which can hinder the further practical application of COF materials for energy-storage batteries.
Introducing active guest species is another effective method to improve the electrical conductivity and the storage performance of COF materials.18–22 For example, Dichtel et al. introduced poly(3,4-ethylenedioxythiophene) (PEDOT) into the pores of COF to improve the electrical conductivity by secondary polymerizations.23 Liu et al. significantly improved the electrical conductivity of TTF-based COFs via doping with iodine or tetracyanoquinodimethane (TCNQ).24 Besides, NiPc-CoTAA framework,25 2D ZnPc-pz COF,26 fluorene-based COF,27 and TANG-COF28 were also reported to enhance the electrical conductivity by absorbing iodine strategy. However, introducing a large amount of guest molecules into the COF pores can hinder ion transport pathway toward redox active sites of the COF, resulting in low capacity and poor rate performance. Therefore, it is still desirable to develop promising COF materials with high inherent ion/electron transport capability for rechargeable batteries.
The β-ketoenamine-linked COF (DAAQ-TFP COF), firstly reported for the energy storage in 2013, has showed the promise for electrical energy storage devices due to its inherent high surface areas and high density of redox-active sites.29,30 The reported results show that the low utilization of anthraquinone redox units is still the main factor limiting the application of bulk DAAQ-TFP COF materials. In a later study, it would be desirable to develop high-performance devices from bulk COF powders. Herein, iodine doped β-ketoenamine-linked COF (COF-I) was prepared through a facile one-pot solvothermal reaction between 1,3,5-triformylphloroglucinol (TFP) and 2,6-diaminoanthraquinone (DAAQ) by adding iodine into the synthetic process of COF. Iodine doping and rod-flower structure endow COF-I more accessible active sites and high intrinsic ion/electron transport capability. Consequently, COF-I cathode demonstrates a significantly increased discharge capacity of 140 mA h g−1 at 0.2 A g−1 and good rate performance in LIBs. This work provides a facile approach for the development of high electrochemical performance COF electrode materials without complicate fabrication and the need for the use of conductive substrates.
X-ray photoelectron spectroscopy (XPS) was conducted to characterize the changes of COF surface information before and after iodine doping. The full spectra of both samples (Fig. 1b) show that the elements C, N and O are the main surface chemical components, while the signal from iodine element can be detected in the COF-I sample. The high-resolution I 3d curve (Fig. 1c) presents a couple of fitting peaks at 618.7 eV and 632.0 eV for I 3d5/2 and I 3d3/2, respectively, which are ascribed to iodine anions chemical state.25,34 For pristine COF (Fig. S2†), the high-resolution C 1s XPS curve reveals three peaks at 284.7, 286.1, and 288.7 eV for CC, C–N, and CO bonds,35 respectively. The N 1s XPS indicates that the peak position at 399.5 and 400.1 eV are assigned to imine N and sp3 N in COF,36 respectively. In high-resolution O 1s XPS spectrum of COF, it can observe two peaks at 531.2 and 533 eV, corresponding to CO and C–O bonds,17 respectively. After iodine doping, these peak positions of C, N, O elements undergo slight shift, suggesting the charge-transfer between iodine and various C, N, O species in COF.37,38 Raman spectroscopy of the COF-I sample was measured and depicted in Fig. S3.† In the Raman spectrum, the COF-I exhibits two characteristic peaks at 113 and 167 cm−1, which can be assigned to the symmetric stretching vibration of I3− and I5−, respectively, which further indicates that the charge transfer occurred between the guest iodine and the electron-rich host network at COFs.39
Powder X-ray diffraction (PXRD) pattern of COF exhibits the obvious peaks at 3.5° and 26.9° (Fig. 1d), corresponding to the (100), and (001) crystal planes, respectively, indicating a highly crystalline honeycomb frame and the ordered porous structure. This result is in accordance with previous reports.5,31,40 Pawley refinement of COF using a P6/m unit cell (Fig. S4†) was performed, and the simulated PXRD pattern is in good agreement with the experimental data. It should be noted that the broad diffraction peak position at 26.5° in COF-I has a slight downshift compared to pristine COF (Fig. 1d), ascribed to the slightly enlarged distance of two COF layers along the c direction after iodine doping. Besides, a weaker diffraction peak intensity at 3.5° for COF-I is observed, implying the decreased crystallinity of COF structure after the induction of iodine species.41
The porosity of COF and COF-I was evaluated at 77 K by N2 adsorption–desorption measurements. Both COFs show reversible isotherms with a steep uptake at low pressures (P/P0 < 0.1) (Fig. 1e), indicating their microporous nature. The Brunauer–Emmett–Teller (BET) specific surface area of COF-I is 1056 m2 g−1, larger than that of pristine COF (517 m2 g−1). The pore size distributions of both COFs are analyzed based on nonlocal density functional theory (NLDFT). COF-I has an average pore size distribution of about 2.39 nm, while COF shows small pore size distribution of about 1.56–2.29 nm (Fig. 1e). The high specific surface area and large pore size distribution is probably originated from the aggregation state change induced by the iodine doping. In addition, Fig. 1f shows the electron paramagnetic resonance (EPR) spectra. A strong EPR signal at g = 2.0045 can be observed for COF-I, implying the presence of TFP radical cations from oxidation by iodine.42,43 The charged nanochannels can create a benign environment for capturing numerous Li+ ions from the electrolyte, which is conducive to Li+ moving toward redox-active sites in the channels of COF.
The morphology and microstructure of materials were analyzed by SEM and TEM. It can be seen from Fig. 2a and b that pristine COF has the flower-like morphology with nanorods (diameter of ∼400 nm and length of ∼3.5 μm). Comparatively, the COF-I (Fig. 2d and e) maintains the flower-like morphology, but the nanorods have smaller diameter (∼300 nm) and coarser surface. This helps to provide a larger contact area with the electrolyte and improve Li+/electron transfer channels on the nanorods. High-resolution TEM (HRTEM) images (Fig. 2c and f) show that pristine COF presents the distinct lattice fringes with interlayer distance of 2.1 nm, while COF-I exhibits weak lattice fringe and the slightly extended interlayer distance of 2.3 nm after the iodine introduction. This result is consistent with the previous PXRD analysis. Energy-dispersive spectroscopy (EDS) mapping images further exhibit that the COF-I framework has a uniform distribution of C, N, O, and I elements (Fig. 2g), and the content of iodine is about 2 wt%. (Fig. S5†). The elemental mapping images of pristine COF are also shown in Fig. S6.† To evaluate the reproducibility of iodine doped COF material synthesis, we prepared ten different batches of COF-I samples. The SEM images and the corresponding EDS for ten different batches of COF-I were measured and depicted in Fig. S7.† The morphology and the iodine content of each batch display the similar results.
The thermodynamic stability of both COF and COF-I was further investigated by thermogravimetric analysis (TGA) (Fig. S8†). The TGA curves of COF and COF-I exhibit a similar thermal decomposition process. Before 250 °C, the weight loss is attributed to the evaporation of water and solvent remaining in the nano-pore of samples. In the range of 250–520 °C, the weight loss can result from the breaking of chemical bonds and the decomposition of some oligomers. Obviously, both COF and COF-I show high thermal stability. In addition, the chemical stability of COF-I was also assessed in five solvents (1,3-dioxolane/dimethoxyethane, dichloromethane, tetrahydrofuran, water, hydrogen chloride). The digital images (Fig. S9†) show that solution color does not change after soaking the COF-I in different solvents for two months at room temperature. The FT-IR spectra (Fig. S10†) and UV-vis spectra (Fig. S11†) display that there are no dissolution of COF-I. This also further indicates that there is a strong interaction between the doped iodine and the COF skeleton, showing the high chemical stability of COF-I.
Before the electrochemistry test, Hall effect measurement was employed to investigate the electrical properties of COF-I in vacuum condition at 298 K. Fig. 3a shows that COF-I is a n-type semiconductor with carrier density of 1.3 × 1013 cm−3 and the direct-current Hall electron mobility of 11.05 cm2 V−1 s−1, which represents relatively high values compared to some previously reported 2D-COFs materials (Table S1†). The DC bulk conductivity was evaluated using the four-point probe measurement on a pressed powder pellet, and the corresponding electrical conductivity of COF-I is 0.034 S m−1 at 298 K (Fig. 3b). For comparison, the electrical conductivity of COF-I samples containing 1% and 5% of iodine was also measured (Fig. S12†). The results exhibit that COF containing 2% iodine (COF-I-2%) has the highest electrical conductivity. To further understand the enhanced electronic conductivity in COF-I, solid-state UV-vis spectroscopy (Fig. S13†) measurement was conducted and the results reveal that the absorption edge of COF-I has a red shift, which shows a narrower optical band gap of 1.9 eV compared with that of COF (2.0 eV) determined by the Kubelka–Munk (K–M) function, conducing to rapid electron transport in the COF. As a result, the utilization of active groups in the framework can improve with enhancing electron transport of the COF electrode.31,44
Fig. 3 (a) Hall effect measurement of COF-I; (b) relationship between pressure and resistivity of COF and COF-I. |
The electrochemical behavior of COF and COF-I as cathode materials for LIBs was evaluated by assembling the CR2032 coin cells. Fig. 4a exhibits the cyclic voltammetry (CV) profiles of COF and COF-I, where a pair of redox peaks can be observed, corresponding to the reversible lithiation/delithiation reaction on the CO active sites belonging to AQ.15 Compared with the redox peak current of COF, the COF-I shows dramatically enhanced peak current, indicating its high electrochemical activity.45 It should be noted that the redox peaks from iodine are not observed in CV profiles (2.5 to 3.2 V).46 The galvanostatic discharge/charge curves of COF-I cathode in Fig. 4b show a higher capacity of 140 mA h g−1 at 0.2 A g−1 compared with that of the COF (62 mA h g−1). The rate performances of COF and COF-I were evaluated at different current densities (Fig. 4c). COF-I has a reversible capacity of 164, 140, 121, 104, and 95 mA h g−1 at 0.1, 0.2, 0.4, 0.8, and 1.0 A g−1, respectively, significantly exceeding those of the pristine COF at the same current density. After the current density backing to 0.1 A g−1 successively, the discharge capacities well recover to 164 mA h g−1 with about 100% capacity restorations, demonstrating good rate capability of COF-I. To evaluate the cycling stability of COF and COF-I, long-term cycling tests were carried out (Fig. 4d). The COF-I delivers a high initial discharge capacity of 140 mA h g−1 at 0.2 A g−1 and retains the reversible capacity of 129 mA h g−1 after 1000 cycles. More impressively, the COF-I cathode yields a respectable discharge capacity of 95 mA h g−1 at 1.0 A g−1 (Fig. S14†). From the Ragone plot in Fig. 4e, the rate capability of COF-I is comparable with the some previously reported COF cathode materials, such as 2D-PAI,47 HATN-AQ-COF,48 DAPQ-COF,11 E-TP-COF,17 DAPH-TFP COF,32 DAAQ-TFP COF,32 PI-COF,49 Tp-DANT COF.4 Besides, the cycling performances of the six different batches of COF-I samples were measured over 300 cycles at 0.2 A g−1 and depicted in Fig. S15.† It indicates that the Li+-storage behavior of all batches of COF-I is similar, further demonstrating the good repeatability in preparing COF-I materials.
To further understand electrochemical kinetics of COF and COF-I, CV measurement was conducted at different scan rates. As shown in Fig. 5a, the shift of redox peaks (denoted as Peaks R and O) can be observed within the scan rate of 0.3–3.0 mV s−1. According to the equation: i = avb (peak current i, scan rate v and variable coefficients a and b), the b-value of 0.5 means a diffusion controlled process, whereas the b-value of 1.0 presents a capacitive-controlled behaviour.50 The b-value was determined by the slope of the linear fit of the log(i)–log(v) plot (Fig. 5b). COF-I shows the b-value of 0.719 and 0.768 for the oxidation process (bO) and the reduction process (bR), respectively, slightly lower than that of COF (0.77 and 0.841), indicating both capacitive and diffusion-limited redox process. Furthermore, the contribution ratio of diffusive (k2v1/2) and capacitive (k1v) component was calculated according to the equation: i = k1v + k2v1/2, where k1 and k2 are constants, v is scan rate.51 As the scan rate increases from 0.3 to 3.0 mV s−1, the capacitance proportion in COF-I progressively enhances from 53.47% to 91.68% (Fig. 5c), revealing the fast pseudocapacitive behavior at high rates, which stabilizes the cycling performance. The average lithium ion diffusion coefficients for COF and COF-I were calculated to be 2.92 × 10−12 and 3.42 × 10−10 cm2 s−1, respectively, by galvanostatic intermittent titration technique (GITT) testing (Fig. 5d, and Fig. S16†), indicating the faster Li+ diffusion kinetics in COF-I.52 Electrochemical impedance spectroscopy (EIS) was carried out to better understand the kinetic process. As shown in Fig. 5e, the Nyquist plots show a semicircle (high-medium frequency region) and an inclined line (low-frequency region), representing the system resistance (including the solution resistance, SEI resistance and charge transfer resistance) and the diffusion resistance of ions, respectively.6 The smaller semicircle and the higher the slope of COF-I reveal the lower charge transfer resistance and faster Li-ion diffusion compared with pristine COF, representing the faster redox reaction kinetics.53 Particularly, zeta potential test (Fig. 5f) shows that iodine-doped COF has negative zeta potential value (−36 mV), whereas zeta potential of pristine COF is close to 0 mV with neutral characteristics. This indicates the negative polarity on porous surfaces of the COF-I, which is beneficial to improve the lithium ion affinity of COF pore walls through electrostatic interaction.54 The abundant lithiophilic sites (anion I3− and I5− species) in the channels of COF-I can capture numerous Li+ ions from the electrolyte, which is conducive to elevating the Li+ concentration near the surface of COF channels.18 Meanwhile, the carbonyl oxygen atoms of COF could also serve as Li-anchoring sites effectively coordinating Li ions. Thus, Li+ could transport by site-to-site hopping along the 1D pore walls with the aid of cation–dipole interactions, which is conducive to Li+ moving toward redox-active sites in the channels of COF, as shown in Fig. S17.†
Footnote |
† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d3ra01414b |
This journal is © The Royal Society of Chemistry 2023 |