A universal strategy for producing 2D functional carbon-rich materials from 2D porous organic polymers for dual-carbon lithium-ion capacitors

XIN Xiao-yu, ZHAO Bin, YUE Jin-shu, KONG De-bin, ZHOU Shan-ke,HUANG Xiao-xiong, WANG Bin, ZHI Lin-jie,,*, XIAO Zhi-chang,*

（1. Department of Chemistry, College of Science, Hebei Agricultural University, Baoding 071001, China;2. College of New Energy, China University of Petroleum (East China), Qingdao 266580, China;3. CAS Key Laboratory of Nanosystem and Hierarchical Fabrication, National Center for Nanoscience and Technology, Beijing 100190, China）

Abstract： Two-dimensional (2D) carbon materials have attracted enormous attention, but the complicated synthesis methods, inhomogeneous structure and uncontrollable properties still limit their use. Here we report a universal protocol for fabricating a series of heteroatom-doped 2D porous polymers, including pyrrole and indole as nitrogen-dopant sources, and 3,4-ethoxylene dioxy thiophene as a sulfur-dopant source by a simple chemical crosslinking reaction. This bottom-up strategy allows for the large-scale synthesis of functionalized ultrathin carbon nanosheets with a high heteroatom doping content and abundant porosity. Consequently,the obtained N-doped carbon-rich nanosheets (NCNs) sample has a specific capacity of 573.4 mAh g-1 at 5 A g-1 as an anode for lithium-ion capacitors (LICs), and the optimized sample has a specific capacitance of 100.0 F g-1 at 5 A g-1 when used as a cathode for a LIC. A dual-carbon LIC device was also developed that had an energy density of 168.4 Wh kg-1 at 400 W kg-1, while maintaining outstanding cycling stability with a retention rate of 86.3% after 10 000 cycles. This approach has the potential to establish a way for the precise synthesis of substantial amounts of 2D functionalized carbon nanosheets with the desired structure and properties.

Key words: Functional carbon nanosheets；Porous organic polymer；Formation mechanism；Universal synthetic strategy；Dual-carbon lithium-ion capacitor

1 Introduction

The intriguing properties of two-dimensional(2D) carbon materials, including exceptional mechanical strength, high electronic and thermal conductivity,have garnered significant attention in the scientific community. Researchers are particularly captivated by the vast diversities offered by these materials, ranging from their synthetic strategies and porosity to their chemical structures and functional properties[1]. Generally speaking, 2D carbon materials can be achieved by 2 methods. One is through the top-down method[2]like mechanical cleavage, liquid-phase exfoliation and electrochemical exfoliation. However, the yield of the target nanosheets is often quite low, and the 2D materials may exhibit inhomogeneous thickness and small lateral size; meanwhile, uncontrollable defects can be inevitable due to harsh conditions such as high-shear mixing, sliding frictional forces, and high friction between 2D flakes[3]. Another is the bottom-up method[4]starting from the small molecules. For instance, a versatile active-salt-templating strategy has been proposed as a means of synthesizing 2D porous carbon materials derived from layered organic-inorganic hybrids[5]. Unfortunately, the conventional approach often necessitates the use of environmentally hazardous reagents to remove the template, resulting in complicated processes. CVD as another bottom-up method can avoid this problem; however, the yield through this method is very low[6]. Therefore, it remains an ongoing challenge to fabricate 2D materials with precise structures and high yield through controllable chemical reactions.

Porous organic polymers (POPs)[7]exhibit remarkable properties, characterized by their abundant porosity and high specific surface area (SSA), have sparked considerable interest in a wide range of scientific applications. These versatile materials are being actively investigated for their potential in diverse fields, encompassing gas separation/storage, catalysis,energy storage/conversion, etc. Their unique features make them highly attractive for addressing various challenges and driving advancements in these areas.Meanwhile, the unique chemical structure of the building blocks endows these polymers with great potential to synthesis materials with desirable function and morphology[8]. On one hand, a variety of techniques have been employed to synthesize porous polymers, including templating strategies, self-assembly of block copolymers, as well as wet-chemical methods[9].On the other hand, a lot of building blocks with different geometries can be employed for the development of POPs[10]. Consequently, it provides us with a great opportunity to design polymers with different functional groups and morphology by elaborately choosing the suitable synthetic methods and building blocks. However, the 2D materials derived from POPs are still scarce, which is mainly due to the lack of efficient synthetic strategies. Although a controlled and versatile approach to pattern 2D free-standing surface has been reported, it necessitates the assistance of monomicelle close-packing assembly of block copolymers. This is primarily due to the limited porosity exhibited by certain polymers, such as polypyrrole and polyaniline. These materials may lack the desired pore structure required for efficient patterning without the aid of block copolymer assembly[11]. Additionally, a class of layered solvent knitting hyper-cross-linked microporous polymers (SHCP) have been prepared with thickness varying from 2 to 50 nm; yet considering the polymers were always block particles, further rigorous treatment was required to obtain thinner SHCP nanosheets[12]. Therefore, deeply understanding the formation mechanism of these 2D polymers,combined with an efficient strategy for their controllable chemical functionalization, has remained a formidable challenge.

The sheet-like structure of 2D carbon materials holds significant promise for facilitating efficient mass transfer in LICs, which offer a distinct combination of the benefits observed in high-energy lithiumion batteries (LIBs) as well as high-power supercapacitors (SCs). The utilization of 2D carbon materials in LICs facilitates improved mass transport, thereby enhancing the overall performance and bridging the gap between LIBs and SCs[13]. Nonetheless, it is important to note that the working mechanisms differ significantly between the anode, resembling a battery, and the cathode, resembling a capacitor. In the former, energy storage primarily relies on Faraday reactions,yielding a high energy density of 150-300 Wh kg-1.However, this attribute is accompanied by a relatively constrained power density, typically exhibiting values lower than 1 kW kg-1. On the other hand, the latter operates on the basis of physical adsorption/desorption of ions, driving the electrochemical process and enabling a high power density of up to 10 kW kg-1.However, this comes with the trade-off of a limited energy density ranging from 5 to 10 Wh kg-1. Consequently, it becomes essential to strategically engineer electrode materials that address the disparity in reaction kinetics between the cathodic and anodic processes in LICs, aiming for the simultaneous attainment of high energy density and power density.

In this work, the aforementioned scientific challenges for fabricating 2D polymers with well-defined structure are successfully solved based on heterocyclic aromatic hydrocarbons and the external crosslinker via a facile knitting reaction. The obtained Ndoped polymer nanosheets (NPNs) featuring a thickness of only ～70 nm and high N content of 10.4%(atomic percent), could be converted into N-doped carbon-rich nanosheets (NCNs) by a chemical crosslinking reaction. Impressively, the formation mechanism of the 2D morphology is experimentally clarified,and this chemical reaction is demonstrated to be a universal tool for producing a series of 2D porous organic polymers with variable heteroatom dopants,namely, pyrrole and indole as N source, and EDOT as S source. Benefitting from the abundant porosity, ultrathin thickness of ～25 nm, as well as a high N-doping content (6.11%), the NCNs could serve as both battery-type anodes and capacitor-type cathodes. As a result, by employing these 2D carbon-rich nanosheets,a dual-carbon LIC is successfully assembled. The electrode material exhibits an extraordinary specific capacitance of 75.7 F g-1at 0.2 A g-1, coupled with an excellent energy density of 168.4 Wh kg-1at 400 W kg-1. Moreover, the LIC also exhibits exceptional cycling stability, retaining 86.3% capacity after 10 000 cycles.

2 Experimental

2.1 Sample preparation

Synthesis of NPNs: First, 1.3 g (0.01 mol) of AlCl3was dissolved in 20 mL of dichloroethane (DCE) with the aid of 0.04 mol of FDA (3 600 μL). Subsequently,0.01 mol (695 μL) of pyrrole was gradually introduced into the solution. To ensure uniform blending of the monomers, the resulting mixture was stirred in an ice bath. Subsequently, the mixture was subjected to stirring at 45 °C for 5 h, leading to the formation of the initial network. To achieve complete conversion,the reaction was continued at 80 °C for 19 h. The resulting precipitate was thoroughly washed with hydrochloric acid, ethanol, and deionized water until the solution became colorless, ensuring the removal of any residual impurities. It was then thoroughly washed with ethanol in a Soxhlet apparatus for 24 h,followed by overnight drying in an oven at 60 °C.

Synthesis of indole- and EDOT-NPNs: The method was followed a similar procedure as that of NPNs, with the only difference being the amount of FDA used. For indole-NPNs, 0.08 mol (7 200 μL) of FDA was employed, while for EDOT-NPNs, 0.02 mol(1 800 μL) of FDA was used. This adjustment in the amount of FDA ensures the desired composition and properties of the synthesized indole- and EDOTNPNs.

Synthesis of NCNs, indole- and EDOT-NCNs:NCNs, indole-NCNs, and EDOT-NCNs were synthesized by subjecting NPNs, indole-NPNs, and EDOTNPNs to pyrolysis under an argon flow at 800 °C for 1 h. The pyrolysis process was performed using a heating rate of 3 °C min-1to ensure precise and effective transformation of the precursor materials into the intended carbon-based nanosheets, ensuring controlled and efficient conversion.

Synthesis of ANCNs: Typically, a mixture of solid KOH and NPNs was dissolved in 25 mL of deionized water, with varying mass ratios of 2∶1, 3∶1 and 4∶1 respectively. After the KOH was fully dissolved, the solution was transferred into a nickle crucible. After the solution was dried, the crucible containing the solid was carefully transferred to the furnace for pyrolysis. The solid underwent thermal treatment at 600 °C for 1 h under a controlled argon flow,with a heating rate of 2 °C min-1. Following the pyrolysis process, the resulting solid was thoroughly washed with hydrochloric acid, ethanol, and deionized water until the solution reached a neutral pH level. The resulting black powders, designated as ANCNs-2, ANCNs-3 and ANCNs-4, respectively, were obtained after drying the solution overnight in an oven.

2.2 Characterization

The chemical structures were analyzed using various spectroscopic techniques. Fourier transform infrared spectrometer (FTIR, Perkin, Spectrum One)was employed to investigate the functional groups present. The13C CP/MAS NMR spectra were recorded using a WB 400 MHz Bruker Avance III spectrometer, providing valuable information about the carbon bonding environment. X-ray photoelectron spectroscopy (XPS) was performed using an ESCALAB250Xi system equipped with a monochromatic AlKα X-ray source, allowing for the characterization of the elemental composition and oxidation states. The graphitization structure was characterized using X-ray diffraction (XRD) analysis (Rigaku D/max-2500B2+/PCX system). This technique provided information about the degree of graphitization. The Renishaw invia plus instrument was utilized to acquire Raman spectra, allowing for the analysis of the defective structures. The morphology was characterized using multiple techniques. Scanning electron microscopy (SEM) analysis was performed using a Hitachi S4800 instrument, providing high-resolution imaging of the surface morphology. Field-emission transmission electron microscopy (FE-TEM) analysis was carried out using a Tecnai G2 F20 U-TWIN instrument, enabling detailed examination of the sample’s internal structure and morphology at high magnification. Additionally, atomic force microscopy(AFM) imaging (Veeco Dimension 3100 Atomic Force Microscope) allowed for the visualization of surface topography and lamellar thickness. To prepare the samples for AFM analysis of the 2D porous carbon-rich materials, the powders were dispersed into ethanol at a concentration of approximately 0.01 mg mL-1. The dispersion was subjected to sonication for 10 min using an Ultrasonic Cell Disruptor to ensure proper dispersion and uniformity. Subsequently,the suspension was applied onto a freshly cleaved silicon wafer surface and subjected to vacuum drying at a temperature of 60 °C. This procedure allowed for the formation of a thin, homogeneous layer of the sample on the substrate, suitable for AFM imaging and analysis. The Micrometritics ASAP 2020 analyzer was employed to analyze the nitrogen adsorption and desorption isotherms, facilitating the determination of the Brunauer-Emmett-Teller (BET) SSA. Additionally, a density functional theory (DFT) method was employed to calculate the pore size distribution(PSD), providing a comprehensive characterization of the porous structure.

2.3 Electrochemical measurements

Fabrication of the half cells: The composition of the working electrodes consisted of NCNs, acetylene black, and polyvinylidene fluoride (PVDF) binder in a weight ratio of 80∶10∶10. These components were thoroughly mixed in N-methyl-2-pyrrolidinone (NMP)to form a homogenous slurry. Subsequently, the slurry was carefully applied onto a copper foil substrate using a slurry-casting technique, ensuring precise and controlled deposition of the NCNs anode material.The electrolyte employed in this study consisted of a 1 mol L-1LiPF6solution dissolved in a blend of ethylene carbonate (EC) and dimethyl carbonate (DMC) in a 1∶1 volume ratio. The cell assembly process was carried out inside an argon-filled glove box to maintain an inert atmosphere. A Celgard 2400 membrane was employed as the separator to facilitate ion transport while preventing direct contact between the anode and cathode.

Fabrication of the supercapacitors: The weight ratio of ANCNs, acetylene black, and polytetrafluoroethylene (PTFE) binder in the working electrodes was 85∶10∶5, respectively. To achieve a homogeneous mixture, these components were carefully dispersed in ethanol, serving as a dispersant. The mixture was thoroughly blended into a paste using a steel pestle. The electrolyte used was identical to that used in the half cells. The assembly of the device was also carried out inside an argon-filled glove box, with a glass fiber membrane serving as the separator.

Fabrication of the LICs: Coin-type cells(CR2032) were used to assemble LICs. The pre-lithiated NCNs were utilized as the anode, achieved by direct contact with Li metal and a few drops of electrolyte in between, followed by a half-hour waiting period. The cathode was prepared with a mass ratio of 8∶1∶1 for the 3 components, representing ANCNs,acetylene black and PVDF, respectively. The optimized mass ratio of ANCNs to NCNs in this study was found to be 4.0. The ANCNs electrode utilized aluminum foil as the current collector, while Celgard 2400 membrane was chosen as the separator. For the electrolyte, a 1 mol L-1LiPF6solution in a 1∶1 volume ratio of EC to DMC was employed.

Electrochemical measurements: The Land battery test system was employed to evaluate the galvanostatic charge-discharge (GCD) performance of the half cells. Furthermore, the Biologic VSP electrochemical workstation was utilized to conduct cyclic voltammogram (CV) measurements and electrochemical impedance spectroscopy (EIS). To calculate the specific capacitance (C, F g-1) of the ANCNs samples in the symmetric supercapacitor, the following formula was applied:

whereIrepresents the applied constant current (A),mdenotes the total mass of the active material in both electrodes (g), and dV/dtrepresents the slope of the discharge curve (V s-1).

The specific capacitance (C, F g-1), gravimetric energy density (E, Wh kg-1) and gravimetric power density (P, W kg-1) of the LICs were determined using the following formulas:

whereIrepresents the applied constant current (A),mdenotes the net mass of active material in both electrodes (g), Δtis the discharge time (s),VmaxandVminare the highest and lowest voltage values (V), andt1andt2represent the starting and ending times of the discharge process, respectively (s).

3 Results and discussion

3.1 Exploration for the formation mechanism of the 2D morphology

The synthesis of pyrrole-based porous organic polymers is quite facile (as detailed in the Experimental section) and was achieved by the knitting reaction between pyrrole and the external crosslinker(formaldehyde dimethyl acetal, FDA). Surprisingly,the product exhibited homogeneous 2D morphology(Fig. S1), which triggered us to further investigate the formation mechanism of the sheet-like structure.

Fig. 1 Exploration for the formation mechanism of the NPNs: (a) Illustration for synthesis of heteroatom-doped 2D porous polymer and porous carbon-rich materials; (b)-(c) SEM images of CNPNs at different magnifications; (d)-(f) element mapping images of aluminum and oxygen; (g) element content analysis of CNPNs

As reported previously, uniform leaf-like boehmite nanosheets with high anisotropy could be achievedviaa simple hydrothermal method[14], with the similar size (a lateral size of (4.5±0.5 μm)×(9.0±1.0 μm) and a thickness of 60-90 nm) of NPNs in this work. Consequently, it is reasonable to hypothesize that the formation mechanism of NPNs is inducedviathe chemical crosslinking of the pyrrole monomers, which are adsorbed on the AlCl3layered crystals solubilized by FDA. To disclose the existence of the 2D AlCl3layered crystals, the product without removal of the catalyst is subjected to a direct carbonization under air to thoroughly remove the polymer composition (denoted as CNPNs). Indeed,the CNPNs reveal a similar sheet-like structure(Fig. 1b-c) to the 2D polymer. Furthermore, the components of the CNPNs are investigated by energy-dispersive X-ray (EDX) analysis. Obviously, CNPNs are mainly comprised of aluminum and oxygen, which are derived from AlCl3compound oxidized by air. The disappearance of element carbon and N confirms that the polymer on the CNPNs is removed thoroughly from the products (Fig. 1d-f). Additionally, the EDX analysis (Fig. 1g) discloses that CNPNs mainly contain aluminum (45.3%, mass fraction) and oxygen(41.8%, mass fraction), which agree well with the element mapping results. Based on these results, a selfassembled temporary template-induced mechanism is confirmed for the formation of the 2D porous polymer nanosheets (Fig. 1a): AlCl3is first solubilized in the solvent by FDA, and form the temporary sheetlike self-assembled template. Afterwards, the pyrrole monomers are oriented and crosslinked along the 2D template with FDA to form the temporary product. Finally, the 2D N-doped polymer nanosheets (NPNs)are achieved after the removal of the temporary template. In addition, a further thermochemical crosslinking reaction could convert NPNs into N-doped carbon-rich nanosheets (NCNs) with similar morphology yet enhanced electrical conductivity, which will be further discussed in the following section.

3.2 Universality for constructing other 2D porous organic polymers

The bottom-up strategy, employing a controllable chemical approach, is further applied to synthesize other heteroatom-doped 2D porous organic polymers. Here we take indole and EDOT as examples.All the FTIR (Fig. S2), solid-state13C NMR spectroscopy (Fig. S3) and XPS measurements (Fig. S4)demonstrate that the indole and EDOT monomers can be successfully knitted into organic polymers by external crosslinkerviathe controllable knitting reaction (further information can be found in the Supporting information for more details). SEM, TEM and AFM are employed to investigate the morphology and microstructure of indole-based N-doped polymer nanosheets (denoted as indole-NPNs), EDOT-based sulfur-doped polymer nanosheets (denoted as EDOTSPNs), and their corresponding carbon nanosheets(denoted as indole-NCNs and EDOT-SCNs, respectively). The sheet-like morphology of indole-NPNs and EDOT-SPNs is depicted in Fig. 2a-b and Fig. 2fg, respectively, enabling the estimation of their thicknesses to be approximately 20 nm (Fig. S5). Notably,after heat treatment at 800 °C, all the samples could retain the similar 2D morphology (Fig. 2c-d and Fig. 2h-i), but ultrathin thicknesses of ～8 and ～10 nm (Fig. S6), respectively, are observed, which is due to the partial degradation of the polymer precursors.Typical AFM and thickness analyses (Fig. 2e and Fig. 2j) further disclose the similar 2D morphology with a uniform thickness of ～7 and ～6.5 nm, respectively, which is in good agreement with the SEM results. These results provide strong evidence for the effectiveness of our bottom-up approach in fabricating 2D porous polymers with diverse heteroatom dopants. Additionally, the weight loss of the three 2D porous organic polymers (i.e., NPNs, indole-NPNs and EDOT-SPNs) upon heating to 800 °C is 33.97%,25.84% and 58.90% (Fig. S7), indicating these 2D porous organic polymers are competent in producing carbon-rich materials[15].

3.3 Chemical composition and microstructure characterization of NPNs and NCNs

Fig. 2 Universality for constructing 2D porous organic polymers: (a, b) SEM images of indole-NPNs under different magnification settings; (c, d) SEM images of indole-NCNs at different magnifications; (e) AFM image of indole-NCNs; (f, g) SEM images of EDOT-SPNs under different magnification settings;(h, i) SEM images of EDOT-SCNs at different magnifications; (j) AFM image of EDOT-SCNs

The chemical composition of the NPNs and NCNs are first investigated by FTIR, solid-state13C NMR and XPS measurements. The FTIR analysis reveals characteristic peaks in the spectrum. Specifically, vibrations corresponding to the pyrrole ring skeleton are observed at approximately 1 477, 1 375 and 1 101 cm-1, while the presence of methylene groups is confirmed by peaks around 2 918 and 2 847 cm-1(Fig. 3a). The solid-state13C NMR spectra of NPNs exhibit a prominent resonance peak at around 130 ppm, indicating the presence of substituted aromatic carbon. Another distinct resonance peak is observed at approximately 1.27×10-4, corresponding to non-substituted aromatic carbon. Moreover, the presence of a resonance peak around 3.1×10-5can be assigned to carbon atoms within the methylene linker in the skeleton of NPNs[16](Fig. 3b). XPS analysis is employed to provide valuable insights into the elemental composition and bonding states of NPNs and NCNs(Fig. 3c-d). As expected, the NPN precursor shows a single peak at 399.6 eV, corresponding to N atoms within the pentagonal ring of pyrrole[17]. Interestingly,after undergoing heat treatment at 800 °C, the N1s spectrum reveals 4 distinct peaks: pyridinic N (398.2 eV), pyrrolic N (399.6 eV), quaternary N (401.0 eV),and N-oxide (403.3 eV)[18]. This observation suggests that a portion of the N atoms within the pyrrole ring undergoes conversion into other types of N. The changes in the N1s spectra may be originated from the enhanced thermal stability of graphitic nitrogen and local alterations in the carbon structure. At the elevated temperature of 800 °C, pyrrolic-N becomes less stable, leading to the transformation of pyrrolic-N species into pyridinic-N and graphitic-N species through condensation reactions. As the carbon structure undergoes transformation under high-temperature conditions, part of N heteroatoms tend to escape from the graphite lattice[19]. Notably, a high N content of 6.11 at% within the conductive carbon skeleton could be well maintained even after the high-temperature treatment, which shows promise for electrochemical applications[20].

Fig. 3 Chemical structure of pyrrole-based 2D materials: (a) FTIR spectra; (b) 13C solid state NMR spectra;High-resolution N1s spectra of (c) NPNs and (d) NCNs

The graphitic structure is further examined by XRD and Raman spectroscopy measurements. The XRD pattern reveals that the NPNs are consisted of amorphous polymeric backbone[21]. Yet for NCNs, the(002) peak at ～26° becomes narrower, indicating an increased graphitic degree after heat treatment(Fig. S8a)[22]. Meanwhile, the Raman spectrum of NPNs presents polymeric features[23], while that of NCNs reveals two prominent peaks (Fig. S8b),namely, theDandGbands, located around 1 355 cm-1and 1 585 cm-1. The observed peaks in the Raman spectra can be attributed to specific vibrational modes of graphite, namely the defect-induced mode and theE2gmode[24]. These findings provide strong evidence for a higher degree of graphitic structure within the NCNs skeleton, aligning with the results obtained from XRD analysis. Hence, the combined analysis of XRD and Raman spectra implies that the heat treatment process effectively enhances the graphitic stacking in the NCNs material.

Fig. 4 Nitrogen adsorption/desorption isotherms at 77.3 K of (a) NPNs and NCNs, (b) ANCNs. Insets show the pore-size distribution results based on the DFT model; (c) HRTEM image; (d) Dark-field TEM and elemental mapping images of NPNs; (e) AFM image of NCNs

N adsorption/desorption measurements are conducted to analyze the porous characteristics of the 2D materials. The adsorption/desorption isotherms of NPNs and NCNs (Fig. 4a) exhibit a combined curve of type I and type IV, indicating a hierarchical porous structure[25]. The calculated SSA of NPNs and NCNs are 302.7 and 247.5 m2g-1, respectively. The PSD analysis reveals that the materials predominantly consist of micropores with a size of 1.5 nm and mesopores ranging from 10 to 50 nm, which can enhance mass transfer during the electrochemical process[26].To enhance the charge storage capacity, the development of high-SSA carbons is crucial. In this study,KOH is selected as the activating reagent due to its lower activation temperature, higher yields, and welldefined pore-forming capability[27]. To maintain the 2D structure of the derived carbon materials, an activation process was implemented by using different mass ratios (2, 3 and 4) of KOH to NPNs (Details can be found in Experimental section). The obtained activated N-doped carbon nanosheets (ANCNs) are denoted as ANCNs-2, ANCNs-3 and ANCNs-4, respectively. As shown in Fig. 4b, the adsorption/desorption isotherms of ANCNs-2 and ANCNs-3 show a joint curve of type I and type IV, indicating a hybrid of micro- and mesoporous structure. Yet ANCNs-4 demonstrates type I isotherm, which reflects a typical microporous character. The calculated SSA of ANCNs-2,ANCNs-3 and ANCNs-4 are as high as 1 018.5,1 166.0 and 1 407.6 m2g-1, respectively, which are promising as electrode materials for supercapacitors.The PSD analysis indicates that the porosity of both ANCNs-2 and ANCNs-3 is primarily constituted by micropores with dimensions less than 1 nm, in addition to the presence of mesopores around 4-5 nm.However, with an increase in the mass ratio of KOH,the mesopores in ANCNs-4 are observed to diminish significantly. XRD and Raman spectroscopy techniques are employed to analyze the degree of graphitization and the presence of defects in the ANCNs samples. The XRD patterns (Fig. S9a) show that ANCNs-4 have a weaker (002) peak absorption compared to ANCNs-2 and ANCNs-3, indicating a more disordered structure and a disrupted graphitic carbon skeleton with increasing amounts of KOH[28]. Additionally, Raman spectroscopy analysis reveals that ANCNs-2 exhibits the lowestID/IGvalue of 0.84 (Fig.S9b), while theID/IGratio increases with the increasing KOH content, suggesting a decrease in graphitic degree and an increase in defects, which is consistent with the XRD measurements. A high specific surface area, appropriate defect structures, and abundant N doping will synergistically improve the specific capacitance, rate capability and long-term cycling stability in supercapacitors[29-30].

The morphology and microstructure of NPNs,NCNs and ANCNs is further investigated. As shown in Fig. S10a-b, substantial homogeneous free-standing sheets of NPNs precursors are observed, exhibiting a size range extending from 500 nm to several micrometers. Remarkably, the resulting sheets possess a low thickness of ～70 nm. Meanwhile, the sheet-like morphology as well as the porous nature can be further demonstrated by TEM and high-resolution TEM images (HRTEM, Fig. S11 and 4c). The element mapping images (Fig. 4d) further reveal the distribution of nitrogen atoms on the carbon skeleton, indicating a homogeneous composition of the nanosheets. Remarkably, even after heat treatment at 800 °C, the 2D structure can still be preserved for NCNs, whereas the thickness is reduced to ～30 nm (Fig. S12), which is due to the partial degradation of the NPNs precursors.HRTEM image of NCNs (Fig. S13) indicates the essence of amorphous structure, whilst partial graphitic character could be distinguished among the disordered skeleton, which could be helpful to increase the Li+storage capability. By conducting AFM and thickness analyses (Fig. 4e), it is evident that the nanosheets maintain a consistent 2D morphology,similar to the observations from SEM and TEM. The thickness measurements reveal a uniformity of approximately 25 nm throughout the samples. In addition, the morphology of ANCNs is also characterized by SEM (Fig. S14). As expected, the ANCNs-2 and ANCNs-3 can retain the 2D sheet-like structure with similar thickness. However, the ANCNs-4 possesses irregular bulk particles, which is probably due to the excessive KOH that induces pronounced swelling and destroys the polymer skeleton during the activating process[27].

3.4 Electrochemical performance of NCNs and ANCNs

The electrochemical performances of NCNs are initially evaluated by its lithium storage capability in CR2032 coin half-cells. Fig. S15a illustrates the initial discharge process in the cyclic voltammetry of NCNs, displaying 2 distinct cathodic peaks. These peaks, located at approximately 0.8 V and 0.01-0.5 V,correspond to specific electrochemical processes. The peak at 0.8 V indicates the formation of the solid electrolyte interphase (SEI) layer, whereas the peaks within the range of 0.01-0.5 V signify the reversible insertion of lithium into the graphite layers and nanocavities[31]. The following 4 cycles overlap on each other,implying that the NCN electrode is quite stable and undergo a reversible electrochemical process. The GCD curves of the initial 3 cycles at 0.2 A g-1are shown in Fig. S15b. Significantly, during the first cycle, the GCD curve demonstrates a plateau in the range of 0.7-0.9 V, which corresponds to the formation of the SEI layer and other side reactions, consistent with the CV results. To assess cycling stability,the NCN electrode is subjected to high current density cycling at 5 A g-1. Remarkably, the NCN electrode maintains a consistent reversible capacity of 519.7 mAh g-1even after 1 000 cycles (Fig. S15c),showcasing its exceptional capacity and prolonged cycling life. Furthermore, the NCN electrode displays impressive rate capability, delivering high specific capacities of 1 068.8, 986.0, 855.0, 760.9, 700.8 and 573.4 mAh g-1at 0.1 0.2, 0.5, 1, 2 and 5 A g-1, respectively (Fig. S15d). Notably, the electrode exhibits remarkable stability, maintaining a capacity of 570.6 mAh g-1even under a high current density of 5 A g-1,and it can restore up to a capacity of 1 032.1 mAh g-1when the current density is reduced back to 0.1 A g-1.These exceptional properties demonstrate the excellent lithium-ion storage capability of NCNs.

The EDLC performance of the 2D materials is further examined using a symmetrical supercapacitor configuration, where two identical ANCNs electrodes are assembled with a LiPF6-based organic electrolyte.Fig. 5a shows the CV curves of ANCN electrodes,employing a scan rate of 100 mV s-1and covering a voltage range of 0-2.7 V. Notably, no significant peaks are observed in these electrodes, and the rectangular shape of the CV curves indicates an ideal double-layer capacitive behavior[32]. Obviously, ANCN-3 exhibits the highest current density among the three counterparts. The GCD curves at 0.5 A g-1are presented in Fig. 5b. Accordingly, ANCN-3 shows the highest specific capacitance of 151.5 F g-1, yet ANCN-2 and ANCN-4 exhibit only 110.3 and 128.5 F g-1, respectively. To further understand the difference in capacitive behavior of the three samples, EIS was carried out to investigate the electrolyte ion transport capability (Fig. 5c). In the low-frequency range,ANCN exhibited nearly vertical curves, signifying their optimal capacitive behavior[33]. Interestingly, the charge transfer resistance (Rct) value of 18.4 Ω for ANCN-3 is much lower than that of 26.0 and 33.9 Ω for ANCN-2 and ANCN-4, which implies a superior mass transfer process within ANCN-3. Consequently,ANCN-3 is demonstrated to be the best electrode among the ANCNs for EDLC, and a closer observation of its EDLC performance was deemed worthy.The CV curves of ANCN-3 exhibit a consistent rectangular shape from 2 to 100 mV s-1, indicating their excellent electric double layer capacitive properties(Fig. 5d). Fig. 5e illustrates the GCD curves of the supercapacitor utilizing ANCNs-3 at different current densities. The specific capacitance of ANCNs-3 in the Li+electrolyte is quantified as 167.0, 151.5, 139.0,123.9 and 100.0 F g-1corresponding to current densities of 0.2, 0.5, 1, 2 and 5 A g-1, respectively. Furthermore, the electrochemical stability of ANCNs-3 is assessed through the GCD technique at 2 A g-1, as depicted in Fig. 5f. Impressively, the results reveal a mere 5% decay in capacitance even after undergoing 10 000 cycles, emphasizing the excellent long-term durability of ANCN-3. Collectively, ANCN-3 demonstrates excellent EDLC performance, including high specific capacitance as well as good rate capability,indicating its potential application in supercapacitors.

Fig. 5 EDLC performance of ANCNs: (a) CV curves at a scan rate of 100 mV s-1 ; (b) GCD profiles at 0.5 A g-1; (c) EIS curves. Supercapacitor performance based on ANCNs-3: (d) CV curves at scan rates from 2 to 100 mV s-1; (e) GCD profiles from 0.2 to 5 A g-1;(f) long-term cycling performance for 10 000 cycles at 2 A g-1

Till date, the kinetic mismatch and rapid electrochemical process still impede the full realization of high-performance LIC[34]. To further investigate the potential electrochemical application of the 2D materials in this work, NCNs//ANCNs dual-carbon LIC was assembled by choosing NCNs as anode and ANCNs as cathode in LiPF6-containing organic electrolyte.Before the assmebling of LIC, the NCNs were prelithiated (further information can be found in the Experimental section) to achieve a high operation voltage in the cell[35]. To prevent the electrolytes from undergoing oxidative decomposition, the electrochemical tests were conducted within the voltage range of 0-4.0 V[36]. Notably, the CV curves of NCNs//ANCNs-3 LIC exhibit a gradual deviation from the ideal rectangular shape as the scan rate increases (Fig. 6b),which is due to the different energy storage mechanisms between the 2 electrodes[37]. Likewise, the GCD curves depicted in Fig. 6c also deviate from the typical linear slopes. The specific capacitance values at various current densities, namely 0.2, 0.5, 1, 2 and 5 A g-1(based on the total mass of both cathode and anode), are measured to be 75.7, 68.2, 61.0, 53.8 and 44.0 F g-1, respectively. Notably, the overall capacitance retention of this system is determined to be 58.1%. In contrast, only 39.6% capacitance can be retained for NCN//ANCN-4 LIC (Fig. S16), confirming that 2D nanosheets could indeed facilitate the fast mass transportation. While the NCN//ANCN-2 LIC exhibits an appreciable capacitance retention of 52.0%benefitted from the 2D carbon skeleton, the specific capacitance is much lower (51 F g-1at 0.2 A g-1). This is probably due to the inferior SSA of ANCN-2,which provides insufficient sites for the charge adsorption/ desorption during the electrochemical process. Thus, it is reasonable to conclude that both SSA and the 2D skeleton play synergetic effects on the performance of NCN//ANCN LIC. Furthermore, the practical application of NCN//ANCN-3 LIC was examined and it could deliver both a high-energy density of 168.4 Wh kg-1(at the power density of 400 W kg-1) and a high-power density of 10.0 kW kg-1(with a high energy density of 98.0 Wh kg-1), which outperforms NCNs//ANCNs-2 LIC, NCNs//ANCNs-4 LIC and other previously reported dual-carbon LIC(Fig. 6d), such as NPCS-1//NPCS-1 LIC[35], PDAGN//PG LIC[38], TTF-COF/CNT//AC HLIC[39], YP-50F//PLPCC-700 LIC[40], CTAB-Sn(IV)@Ti3C2//AC LIC[41], G1P4//AC LIC[42], NbN@C//AC LIC[43], GNS-13//AC LIC[44]. Remarkably, the NCN//ANCN-3 LIC also exhibits an excellent cycle stability with the capacity retention of 86.3% after 10 000 cycles at 2 A g-1(Fig. 6e). It is worth mentioning, an LED panel composed of 16 blue LEDs (Fig. 6f), can be successfully powered by the NCNs//ANCNs-3 LICs, demonstrating its promising practical potential. The remarkable performance of NCN//ANCN LIC can be attributed to the synergistic effects of several key factors, as depicted in Fig. 6a. (1) The 2D nanosheets’ high aspect ratio and excellent conductivity facilitate smooth electron and ion transport across the electrodes, resulting in high reversible capacities, rapid rate capability, and exceptional cycling stability. (2) The high SSA and abundant porosity which are interconnected in the carbon skeleton, result in a large ion-accessible surface area within the unique hierarchical porous nanosheets.(3) The incorporation of N-doping in the 2D porous carbon not only enhances the Li+storage capacity but also improves the rate capability of the electrode. It is the combined effect of these factors, rather than any individual factor, that results in the superior performance of ANCN-based LICs, despite their relatively modest surface areas and N-species contents.

Fig. 6 (a) Schematic illustration, (b) CV curves at different scan rates and (c) GCD profiles of the NCNs//ANCNs-3 LIC; (d) Comparative analysis of Ragone plots: NCNs//ANCNs dual-carbon LIC versus previously reported dual-carbon LICs; (e) long-term cycling stability at 2 A g-1 (insets are the GCD profiles of the first three and last three cycles); (f) the digital photograph of an LED panel powered by the NCNs// ANCNs-3 LIC

4 Conclusions

In conclusion, we have developed a universal protocol for synthesizing a series of heteroatom-doped 2D porous polymers, employing pyrrole, indole, and EDOT as N and S dopant sourcesviaa controllable chemical crosslinking reaction. This adaptable bottom-up strategy facilitates large-scale synthesis of 2D porous carbon materials characterized by abundant heteroatom doping content and porosity, along with ultralow thickness achieved by direct pyrolysis. As a result, the derived 2D porous carbon-rich materials demonstrate outstanding electrochemical performance as anodes for high Li+storage (with a specific capacity of 519.7 mAh g-1at 5 A g-1for 1 000 cycles)and as electrodes for organic system EDLCs (exhibiting only 5% capacitance decay after 10 000 cycles at 2 A g-1), while also showcasing potential for high-performance dual-carbon LICs (with excellent cycling stability, 86.3% capacity retention after 10 000 cycles,and a high energy density of 168.4 Wh kg-1). This bottom-up approach, capitalizing on a knitting reaction with a variety of aromatic molecules, may pave the way for the synthesis of heteroatom-doped 2D porous polymers and carbon materials with tunable properties and broad applicability in energy-related fields.

Data availability statement

The data that support the findings of this study are openly available in Science Data Bank at https://www.doi.org/10.57760/sciencedb.j00125.00050 or https://resolve.pid21.cn/31253.11.sciencedb.j00125.00 050.

Acknowledgements

The authors acknowledge the financial support from the National Natural Science Foundation of China (22005084 and U20A20131), Science and Technology Project of Hebei Education Department(BJK2023021), the Natural Science Foundation of Hebei Province (E2019204131), and Talents Introduction Plan of Hebei Agricultural University(YJ201819).