Extraordinary Ultrahigh‐Capacity and Long Cycle Life Lithium‐Ion Batteries Enabled by Graphitic Carbon Nitride‐Perylene Polyimide Composites

Graphitic carbon nitride (g–C3N4) is widely used in organic metal‐ion batteries owing to its high porosity, facile synthesis, stability, and high‐rate performance. However, pristine g–C3N4 nanosheets exhibit poor electrical conductivity, irreversible metal‐ion storage capacity, and short‐term cycling owing to their high concentration of graphitic–N species. Herein, a series of 3,4:9,10‐perylenetetracarboxylic diimide‐coupled g–C3N4 composite anode materials, CN–PIx (x = 0.2, 0.5, 0.75, and 1), was investigated, which exhibited an unusually high surface nitrogen content (23.19–39.92 at.%) and the highest pyridinic–N, pyrrolic–N, and graphitic–N contents reported to date. The CN–PI1 anode delivers an unprecedented and continuously increasing ultrahigh discharging capacity of exceeding 8400 mAh g−1 (1.96 mWh cm−2) at 100 mA g−1 with high specific energy density (Esp) of ∼7700 Wh kg−1 and the volumetric energy density (Ev) of ∼14956 Wh L−1 and an excellent long‐term stability (414 mAh g−1 or 0.579 mWh cm−2 at 1 A g−1). Furthermore, the activation of the CN–PIx electrodes contributes to their superior electrochemical performance, resulting from the fact that the Li+ is not only stored in the CN–PIx composites but also CN–PIx activated the Li0 adlayer on the CN–PI1–Cu heterojunction as an SEI layer to avoid the direct contact of Li0 phase and the electrolyte. The CN–PI1 full cell with LiCoO2 as the cathode delivers a discharge capacity of ∼587 mAh g−1 at a 1 A g−1 after 250 cycles with a Coulombic efficiency nearly 99%. This study provides a strategy to develop N‐doped g–C3N4‐based anode materials for realizing long‐lasting energy storage devices.

Graphitic carbon nitride (g-C 3 N 4 ) is widely used in organic metal-ion batteries owing to its high porosity, facile synthesis, stability, and high-rate performance.However, pristine g-C 3 N 4 nanosheets exhibit poor electrical conductivity, irreversible metal-ion storage capacity, and short-term cycling owing to their high concentration of graphitic-N species.Herein, a series of 3,4:9,10-perylenetetracarboxylic diimide-coupled g-C 3 N 4 composite anode materials, CN-PI x (x = 0.2, 0.5, 0.75, and 1), was investigated, which exhibited an unusually high surface nitrogen content (23.19-39.92at.%) and the highest pyridinic-N, pyrrolic-N, and graphitic-N contents reported to date.The CN-PI 1 anode delivers an unprecedented and continuously increasing ultrahigh discharging capacity of exceeding 8400 mAh g −1 (1.96 mWh cm −2 ) at 100 mA g −1 with high specific energy density (E sp ) of ∼7700 Wh kg −1 and the volumetric energy density (E v ) of ∼14956 Wh L −1 and an excellent long-term stability (414 mAh g −1 or 0.579 mWh cm −2 at 1 A g −1 ).Furthermore, the activation of the CN-PI x electrodes contributes to their superior electrochemical performance, resulting from the fact that the Li + is not only stored in the CN-PI x composites but also CN-PI x activated the Li 0 adlayer on the CN-PI 1 -Cu heterojunction as an SEI layer to avoid the direct contact of Li 0 phase and the electrolyte.The CN-PI 1 full cell with LiCoO 2 as the cathode delivers a discharge capacity of ∼587 mAh g −1 at a 1 A g −1 after 250 cycles with a Coulombic efficiency nearly 99%.This study provides a strategy to develop N-doped g-C 3 N 4 -based anode materials for realizing long-lasting energy storage devices.
density metal-ion storage, the sp 2 -hybridized pyridinic-N and pyrrolic-N groups should primarily occur at the edge sites of the carbon lattice, which are believed to be electrochemically more favorable active centers owing to their electron-accepting capability, local charge distribution of the different bonding states of nitrogen, and local density of the states.Consequently, the most competitive metal-ion storage metrics of >1200 mAh g −1 have been reported for high nitrogendoping levels in the range of 5-33.7%. [14,15]A first-principles study conducted by Ma et al., [16] which included the impact of the different bonding states of nitrogen on Li + storage in N-doped graphene nanosheets, showed that pyridinic-N and pyrrolic-N groups are one of the most suitable dopant active centers for the adsorption of multiple Li + (high lithium uptake), resulting in an extremely high reversible lithium storage capacity of >1260 mAh g −1 , while the graphitic-N group was postulated to bind irreversibly with Li + and degrade the cyclability performance.Through combined theoretical simulations and experiments, Wang et al. [17] showed that the N-doped graphene has an abundance of pyrrolic-N "hole" defects located at the edges and surfaces, which exhibited an ultrahigh initial capacity of 1284 mAh g −1 and a reversible capacity of 432 mAh g −1 .Significant research efforts have also proven that a high level of pyridinic-N and pyrrolic-N located at the edge sites or inner surface of the carbon materials plays a crucial role in achieving a superior Li + storage performance. [11,18,19]Therefore, it is pivotal to enhance the pyridinic-N content with the highest pyrrolic-N configuration in nitrogen-doped carbon materials applied in high-performance energy storage devices.
Graphitic carbon nitride (g-C 3 N 4 ) is composed of a twodimensional layered structure analogous to graphene/graphite with a high nitrogen content (∼57 at.%).Moreover, g-C 3 N 4 possesses a unique structure with large triangular pores linked by six pyrrolic-N atoms (heptazine units), an easily scalable heating method using low-cost nitrogen-rich precursors, and a potentially higher rate capability than that of graphene/graphite. [20]Theoretical predictions in the pioneering studies have indicated that g-C 3 N 4 is capable of reacting with Li (or Na and K) atoms to generate Li 2 C 3 N 4 (or Na 2 C 3 N 4 and K 2 C 3 N 4 ) with a theoretical storage capacity of up to 524 mAh g −1 and average adsorption energy of 2.4 eV per Li atom, which are almost twice that observed for graphite (LiC 6 , 372 mAh g −1 ). [21]However, g-C 3 N 4 exhibits poor electronic conductivity, low reversible metal-ion storage capacities (serious irreversible capacity loss), and insufficient cycling lifespan owing to its high content of graphitic-N species within the g-C 3 N 4 nanosheets, which leads to the deterioration of crystallinity. [22,23][26][27][28][29][30][31][32] Nevertheless, nitrogen-doping strategies and composite technology utilizing the coupling of g-C 3 N 4 with inorganic semiconductors suffer from many detrimental deficiencies, including the formation of aggregation/carbonaceous species at high carbonization/pyrolysis condition (>720 °C), tedious post-etching process, volume change and pulverization of the nanoparticles during cycling, which restrict their practical applications for energy storage devices. [33][36][37][38][39][40] Among them, perylenetetracarboxylic dianhydride (PTCDA) possesses a large aromatic π-conjugated condensed structure and high electron mobility. [41]Benefiting from the PTCDA motifs and high-defect density of the g-C 3 N 4 nanosheets, we are particularly interested in constructing highperformance energy storage materials via the surface hybridization reaction between the abundant NH 2 groups located at the edge of the g-C 3 N 4 nanosheets and PTCDA motifs to establish an O=C-N-C=O covalent bond.
Herein, a series of PTCDI-coupled g-C 3 N 4 composites were developed, denoted as CN-PI x (where x = 0.2, 0.5, 0.75, and 1, Scheme 1), which consist of an unusually high surface nitrogen content with the highest pyridinic-N and pyrrolic-N contents by varying the PTCDA unit (0.2, 0.5, 0.75, and 1 wt.%), to understand how these composite structures act as anode materials in LIBs.CN-PI x composites exhibiting a range of excellent features (i.e., high specific surface areas of 49.42-64.38m 2 g −1 ) and abundant porosity (pore volume in the range of 0.184-0.225cm 3 g −1 ) were synthesized using a one-step thermal polycondensation (imidization reaction) between PTCDA and high defect-density g-C 3 N 4 nanosheets with a high surface nitrogen content (∼57 at.%) in a zinc acetate catalyst/molten imidazole system.When used as anodes in LIBs, the high content of PTCDI coupled in CN-PI 1 anode delivers an unprecedented and continuously increasing discharge capacity exceeding 8400 mAh g −1 (the areal capacity of 1.96 mWh cm −2 ) at a current density of 100 mA g −1 and a highly reversible lithium storage capacity of 1847 mAh g −1 (∼4.303 mWh cm −2 ) after the 210th cycle, while also achieving excellent longterm cycling stability after 1000 cycles (414 mAh g −1 or ∼0.579 mWh cm −2 at 1 A g −1 ) with a Coulombic efficiency (CE) of 96%.A CN-PI 1 full cell with LiCoO 2 cathode realized the attractive long-duration cyclic property (∼587 mAh g −1 at a 1 A g −1 after 250 cycles with a CE nearly 99%).

Structural and Compositional Characterization of g-C 3 N 4 and the CN-PI x Composites
The 3,4:9,10-perylenetetracarboxylic diimide (PTCDI unit, which is attached with melon motifs of g-C 3 N 4 ) was assembled with g-C 3 N 4 nanosheets (i.e., CN-PI x ; where x represents the weight ratio of 3,4:9,10-perylenetetracarboxylic dianhydride (PTCDA), x = 0.2, 0.5, 0.75, and 1 g; named as CN-PI 0.2 (CN 1 -PI 0.2 ), CN-PI 0.5 (CN 1 -PI 0.5 ), CN 1 -PI 0.75 (CN 1 -PI 0.75 ) and CN-PI 1 (CN 1 -PI 1 )), in which PTCDA was chosen to react with the as-prepared nitrogen-rich g-C 3 N 4 nanosheets (standard weight of CN is taken as 1 g) via a thermal polycondensation reaction in 95.7% yield based on a previously reported synthetic procedure by Wang et al. [37,38] The structural evolution of the as-synthesized g-C 3 N 4 nanosheets and CN-PI x composite samples was confirmed using FT-IR spectroscopy (Figure S1a,b, Supporting Information).The FT-IR spectrum obtained for the g-C 3 N 4 nanosheets exhibit the typical IR absorptions of melon-type g-C 3 N 4 . [42]The vibrational characteristic peaks for the g-C 3 N 4 nanosheets and CN-PI x composites appeared in the region of 1010-1679 cm −1 , which can be attributed to the characteristic aromatic C-N and C=N heterocyclic Energy Environ.Mater.2023, 6, e12553 2 of 28 stretching vibrations of the skeletal tri-s-triazine ring (heptazine units). [43]The sharp peaks located at ∼806 cm −1 originate from the breathing vibration mode of the tri-s-triazine ring as well as the symmetric in-plane bending of the imide C=O bonds originating from the PTCDI units present in the CN-PI x composites, which manifests the complete conversion of the anhydride groups of the PTCDA into the imide ring (PTCDI unit) in CN-PI x composites.The broad peaks appearing between 3000 and 3400 cm −1 were attributed to the N-H 2 stretching vibration modes of the uncondensed primary and secondary amino groups, [44,45] demonstrating that the g-C 3 N 4 nanosheets contain abundant -NH 2 groups at the edge and defect sites.In general, the FT-IR spectra obtained for the CN-PI x composites were almost identical to that of the g-C 3 N 4 nanosheets.Moreover, the FT-IR spectrum of PTCDA displays two sharp vibrational bands at 1766 and 1727 cm −1 , which were attributed to the asymmetric and symmetric stretching vibrational frequencies of carbonyl groups (C=O bonds), respectively.These vibrational bands completely disappear in all of the CN-PI x composite samples.The strong vibrational bands located at 1298 and 937 cm −1 correspond to the anhydride groups (C-O-C bonds) present in the PTCDA, which were absent in the CN-PI x composites.These results also demonstrate that the PTCDI unit was coupled with the g-C 3 N 4 nanosheets in the molecular structure of the CN-PI x composites (i.e., PTCDI unit was covalently incorporated within the g-C 3 N 4 framework).Fluorescence properties of CN-PI x (x = 0.2, 0.5, 0.75, and 1) samples were also studied to speculate the assembly of PTCDI units on the framework of g-C 3 N 4 nanosheets (Figure S1c, Supporting Information).Initially, the fluorescence spectrum of pristine g-C 3 N 4 nanosheets did not show any distinct emission peak excited at 500 nm.Whereas the pristine PTCDA showed two emission peaks at 529 and 565 nm, which are attributed to 0-0 and 0-1 singlet exciton transitions of PTCDA molecules.The fluorescence spectrum of CN-PI x (x = 0.2, 0.5, 0.75) samples showed a new emission peak with high intensity at 572 nm, corresponding to 0-1 singlet exciton transitions of PTCDI units.This emission peak diminished and became broad at 565 nm for CN-PI 1 , indicating the higher content of PTCDI units assembled with g-C 3 N 4 frameworks of CN-PI 1 .This result further supports the assembly of PTCDI units on the surface of CN-PI x samples.The structure of g-C 3 N 4 nanosheets and CN-PI x composites were further confirmed using elemental (CHNS) analysis (see Section 4 and Table S1, Supporting Information).In order to prove that the PTCDI units were incorporated and assembled on the surface NH 2 groups of the g-C 3 N 4 framework in the CN-PI x samples, the composition of the C, N, and O elements was determined using elemental analysis (Table S1, Supporting Information).It was found that the elemental ratio of C to O increases from g-C 3 N 4 nanosheets, CN-PI 0.2 to CN-PI 1 , while the ratio of N significantly decreases from 60.73% for the g-C 3 N 4 nanosheets to 44.16% for CN-PI 1 .
High-resolution X-ray photoelectron spectroscopy (HR-XPS) was conducted to investigate the surface elemental chemical states and composition of bulk g-C 3 N 4 , the g-C 3 N 4 nanosheets, and the CN-PI x composite samples (Figure 1a,b; Figures S2-S5, Supporting Information).Only carbon, nitrogen, and oxygen species were detected in the XPS survey spectra (Figure S2a, Supporting Information).As depicted in the HR-XPS spectra, the C1s spectra of bulk g-C 3 N 4 and the g-C 3 N 4 nanosheets exhibit four peaks at 284. 48, 287.68, 288.18, and  292.98 eV (Figure S2b, Supporting Information), corresponding to the carbon (C) atoms from the sp 3 /sp 2 -hybridized C=C/C-C bonds of the adventitious carbon species, C atoms in the O-C=O and C-NH 2 groups at the edge sites of the g-C 3 N 4 nanosheets, sp 2 -hybridized C atoms in the N-C=N/C−(N) 3 bonds of the aromatic rings, and π-excitations (and σ* c-c ) of the heptazine ring system, respectively. [40][36][37][38][39][40] Therefore, it is obvious that the chemical states of both the carbon and nitrogen atoms in the g-C 3 N 4 nanosheets were almost identical to those of bulk g-C 3 N 4 , and no obvious binding energy shift was observed in the C 1s and N 1s core electronic spectra.The deconvolution of the O 1s spectra obtained for both bulk g-C 3 N 4 and the g-C 3 N 4 nanosheets show two peaks at 531.58 and 532.98 eV (Figure S2d, Supporting Information), which can be attributed to the surface hydroxyl groups (i.e., N-C-O bond or N-C-OH groups) and adsorbed water molecules on the surface of the g-C 3 N 4 samples, respectively. [46]The peak located at 531.58 eV attributed to the N-C-O functional group corresponds to the deconvoluted O-C=O species observed in the C 1s spectra.According to the percentage of C to N elements determined using HR-XPS, the C content increases from 41.49% for bulk g-C 3 N 4 to 41.90% for g-C 3 N 4 nanosheets, whereas the N content decreases from 57.26% to 56.59% (Table S2, Supporting Information), which is consistent with the theoretical composition of reported in previous studies for g-C 3 N 4 (43% C, 56-57% N, and 1% O). [27,28] This result strongly confirmed the high surface nitrogen content of the g-C 3 N 4 nanosheets obtained via the thermal condensation reaction of a nitrogen-rich urea precursor.The HR-XPS survey spectra of the g-C 3 N 4 nanosheets and as-synthesized CN-PI x (x = 0.2, 0.5, 0.75, and 1) composite samples exhibit the presence of C (285 eV), N (399 eV), and O (532 eV) in the XPS survey spectra (Figure S3a, Supporting Information).According to quantitative XPS analysis (Table S2, Supporting Information), the N/C ratio seemingly decreases from 1.35 for the g-C 3 N 4 nanosheets to 0.75, 0.46, 0.354, and 0.350 for the CN-PI 0.2 , CN-PI 0.5 , CN-PI 0.75 , and CN-PI 1 samples, respectively.A similar decreasing trend was also observed in the elemental (CHNS) analysis (Table S1, Supporting Information).The high-resolution C 1s spectra obtained for the g-C 3 N 4 nanosheets and CN-PI x composites exhibit three typical peaks at 284.48, 287.68, and 288.18 eV (Figure S3b-f, Supporting Information), corresponding to the C=C/C-C bonds, O-C=O/C-NH 2 groups, and N-C=N/C−(N) 3 bonds, respectively.A new peak located at 285.98 eV attributed to the C=O groups originating from the PTCDI units was observed for all of the CN-PI x samples and its peak intensity was obviously intensified upon increasing the PTCDI unit from CN-PI 0.2 to CN-PI 1 .In addition, it was obvious that the binding energy peaks pertaining to the N-C=N/C−(N) 3 moieties were progressively attenuated, whereas the signals from the C=C/C-C species belonging to the as-synthesized CN-PI x samples were clearly more intense when compared to those of the g-C 3 N 4 nanosheets.Moreover, the more pronounced N-C=N/ C−(N) 3 peak observed for the g-C 3 N 4 nanosheets compared to those of the CN-PI x samples was consistent with their higher total surface nitrogen content (56.59 at.%) (Table S2, Supporting Information).These results strongly confirm the destruction of the g-C 3 N 4 framework and the formation of covalent bonds between the PTCDI units and g-C 3 N 4 framework in the CN-PI x samples.S4, Supporting Information present the N 1s HR-XPS spectra obtained for the g-C 3 N 4 nanosheets and CN-PI x (x = 0.2, 0.5, 0.75, and 1) composite samples, in which the N 1s region of all samples can be deconvoluted into four major peaks at 398.18, 399.58, 400.58, and 403.8 eV, which correspond to the pyridinic-N, pyrrolic-N, graphitic-N, and pyridinic−N + -O x − species, respectively.A new and weak peak observed at 399.84 eV was assigned to the imide nitrogen atoms (imide−N) originating from the O=C-N-C=O bonds and its intensity remained almost unchanged upon increasing the PTCDI unit from CN-PI 0.2 to CN-PI 1 .This result verified that the imide−N bond was formed by the imidization reaction between the PTCDA units and NH 2 groups in the g-C 3 N 4 nanosheets, indicating that the PTCDI units were covalently coupled with the g-C 3 N 4 framework. [27,40]The surface nitrogen content in the samples was comprehensively compared, as shown in Figure 1c and Table S3, Supporting Information.It can be seen that the three types of nitrogen atoms decrease significantly upon increasing the PTCDI unit from CN-PI 0.2 to CN-PI 1 when compared to the g-C 3 N 4 nanosheets.In addition, the g-C 3 N 4 nanosheets and CN-PI 0.2 achieve the largest surface nitrogen content of 56.59 and 39.92 at.%, respectively, yet the total percentage of pyridinic-N and pyrrolic-N were the highest (∼65.4-70.91%)for the other CN-PI x (x = 0.5, 0.75, and 1) samples.Consequently, the pyridinic-N and pyrrolic-N contents were observed to be higher than the graphitic-N content in the g-C 3 N 4 nanosheets and all of the CN-PI x (x = 0.2, 0.5, 0.75, and 1) samples, while the imide−N content was almost unchanged in the CN-PI x samples.These results suggest their potential application with superior Li + storage performance upon obtaining a large surface capacity contribution via the higher total surface N-doping level with the highest pyrrolic-N content.In addition, the fractional percentage of pyrrolic-N species in the CN-PI x samples increases in the order of CN-PI 0.2 (16.79%) < CN-PI 0.5 (21.60%) < CN-PI 1 (21.14%)< CN-PI 0.75 (24.67%), while the total surface nitrogen content decreases in the order of CN-PI 0.2 (39.92 at.%) > CN-PI 0.5 (28.47 at.%) > CN-PI 0.75 (23.19 at.%) > CN-PI 1 (23.22 at.%).These results strongly verified that not only the higher total surface nitrogen content but also the three major types of nitrogen bonding configurations can be tuned for the as-synthesized CN-PI x composite materials by establishing a surface hybridization reaction between the different ratios of PTCDA content and NH 2 groups in the g-C 3 N 4 nanosheets.It is worth to note that the overall C/N ratio was increased from CN-PI 0.2 , CN-PI 0.5 , CN-PI 0.75 , and CN-PI 1 composites regardless of the distribution of total surface nitrogen contents (i.e., pyridinic-N, pyrrolic-N, imide-N, graphitic-N, and pyridinic-N ) on the surface of the composite samples by both elemental analysis and XPS result (Tables S1 and S2, Supporting Information).According to Table S3, Supporting Information, it is also important to note that the distribution of pyridinic-N/pyrrolic-N ratio was relatively increased to 2.35 for CN-PI 1 sample regardless of the higher content of imide-N when addition of more PTCDA with g-C 3 N 4 .This could probably be reason for significant amount of imide-N was detected in CN-PI 1 when addition of more PTCDA for polycondensation reaction (Figure 1c; Table S3, Supporting Information).Moreover, it should be noted that CN-PI 1 has higher a total fractional percentage of pyridinic-N and pyrrolic-N content about 70.91% than those of CN-PI 0.75 (65.4%), which could be believed to be electrochemically more favorable active centers owing to their electron-accepting capability, local charge distribution of the different bonding states of nitrogen, and local density of the states.Even though the total surface nitrogen content of CN-PI 1 (23.22% and C/N ratio ∼2.83) is almost identical to that of CN-PI 0.75 (23.19% and C/N ratio ∼2.86), CN-PI 1 has higher fractional percentage of pyridinic-N content (49.77%) combined with pyrrolic-N content (21.14%) and significantly lowest graphitic-N content (21.07%) as compared to those of CN-PI 0.75 (pyridinic-N: 40.73%; pyrrolic-N: 24.67% and graphitic-N: 25.83%).These highest combined pyridinic-N and pyrrolic-N functionalities with significantly lowest graphitic-N content of CN-PI 1 are most favorable binding/active centers for an ultrahigh lithium uptake during (de)insertion process.These unique nitrogen configuration characteristics could be beneficial for the formation of very stable and thick Mosaic-like SEI layer on the surface of CN-PI 1 electrode during charging/discharging cycling, thereby resulting in larger Li + storage capability of CN-PI 1 anode than CN-PI 0.75 and other composite anodes.The HR-XPS O1s spectra obtained for the g-C 3 N 4 nanosheets can be resolved into two energy peaks at 531.78 and 533.38 eV (Figure S5, Supporting Information), which can be attributed to the N-C=O bonds and surface hydroxyl groups (N-C-OH) on the surface of the g-C 3 N 4 nanosheets, respectively. [47]When compared to the g-C 3 N 4 nanosheets, the HR-XPS O1s spectra of the CN-PI x samples show two novel deconvoluted peaks at 531.08 and 532.68 eV, corresponding to the oxygen species in the O=C-N-C=O or N-C=O bonds in the CN-PI x heterojunction and C=O groups, respectively. [40]The O1s peak intensities of the O=C-N-C=O bonds observed at 531.08 eV were similar and the peak intensity at 532.68 eV attributed to the carbonyl (C=O) groups obviously increases from CN-PI 0.2 to CN-PI 1 , which mainly originates from the C=O groups in the PTCDI units in the CN-PI x samples and N-C=O groups in the g-C 3 N 4 frameworks.A very weak peak observed at 536.28 eV was assigned to the adsorbed oxygen group present in the g-C 3 N 4 framework of the CN-PI x samples. [40]These results strongly verify that the CN-PI x (x = 0.2, 0.5, 0.75, and 1) heterojunctions were covalently constructed via the surface hybridization of the PTCDI unit and g-C 3 N 4 nanosheets.

Microstructural Characterization of the g-C 3 N 4 Nanosheets and CN-PI x Composites
X-ray diffraction (XRD) was used to monitor the in-plane structures of the bulk g-C 3 N 4 , g-C 3 N 4 nanosheets, and CN-PI x composites (Figure 1d; Figure S6a-c, Supporting Information).The g-C 3 N 4 nanosheets show two consistent diffraction peaks similar to those of bulk g-C 3 N 4 at 12.7°and 27.7°(Figure S6a, Supporting Information), which were assigned to the (100) and (002) planes of the in-planar structural packing motif of the tri-s-triazine unit and characteristic inter-planar stacking between the conjugated aromatic systems of the g-C 3 N 4 nanosheets, respectively. [13,36]This result suggests that the g-C 3 N 4 nanosheets have an identical crystal structure to that of the parent bulk g-C 3 N 4 .The reflection peak observed at 2.89°(100 planes) in bulk g-C 3 N 4 originates from the lattice planes parallel to the c-axis, which was less pronounced in the g-C 3 N 4 nanosheets.When compared to the bulk g-C 3 N 4 material, the peak originating from the periodic interlayer stacking of the g-C 3 N 4 nanosheets was red-shifted from 27.71°to 27.81°, demonstrating the decreased interplanar distance between the basic sheets in the g-C 3 N 4 nanosheets (Figure S6b, Supporting Information).After the incorporation of the PTCDI units on the -NH 2 groups (edge groups) of the g-C 3 N 4 nanosheets, the XRD spectra CN-PI x composites show the only in-plane structure of the g-C 3 N 4 nanosheets, while there were no diffraction peaks related to the PTCDA crystal phase (i.e., PTCDI units are hardly found in the prepared CN-PI x (x = 0.2, 0.5, Energy Environ.Mater.2023, 6, e12553 0.75, and 1) composites at lower PTCDA-to-g-C 3 N 4 weight ratio), as shown in Figure 1d and Figure S6c, Supporting Information.These results demonstrate that the PTCDI molecules are randomly assembled/ distributed layer by layer on the surface of the CN-PI x composite frameworks (melon sheets of g-C 3 N 4 ) rather than existing as solidstate π-π stacking structure arrangements.Therefore, PTCDI units have been randomly assembled layer by layer into the g-C 3 N 4 framework to form few-layer to multi-layered nanoplatelet-like structures (thick plates) with increasing the PTCDI unit from CN-PI 0.2 to CN-PI 1 samples.These few-layers to multi-layered nanoplatelet-like structures can be further justified from TEM images with corresponding selected area electron diffraction (SAED) patterns for CN-PI 0.5 to CN-PI 1 samples (Figure 2e1,e2; Figure S10, Supporting Information) as well as atomic force microscopy (AFM) images for thick nanoplatelet-like morphology with a thickness of about 300-440 nm for CN-PI 1 sample (Figure 2c; Figure S9c, Supporting Information).Due to these thick nanoplateletlike morphology, the XRD spectra of CN-PI x composites show the only in-plane structure of multi-layered nanoplatelet-like g-C 3 N 4 frameworks rather than the π-π stacking structure of randomly assembled PTCDI molecules.Therefore, the randomly distributed PTCDI molecules do not affect the XRD patterns of CN-PI x composite samples, even at the higher content of PTCDI units assembled in the CN-PI 1 sample.These observations are in good accordance with the low density of -NH 2 groups distributed over the edge sites of the g-C 3 N 4 nanosheets in the CN-PI x composites and the XRD results by earlier reports. [37,38,40]Moreover, the (100) and (002) crystal plane peaks observed for the CN-PI x composites were significantly weakened and shifted to lower angles upon increasing the PTCDI content due to the amorphous structures of the CN-PI x composites compared to the more crystalline g-C 3 N 4 nanosheets (i.e., increasing the face-to-face stacking of melon-PTCDI units) (Figure S6c, Supporting Information).Moreover, the decrease in the interplanar distance (d-spacing) in the CN-PI x composites upon increasing the PTCDI content was attributed to the formation of covalent bonds (O=C-N-C=O bonds) (Figure 1d).
Figure 1e shows that the Raman spectra obtained for the CN-PI x composites exhibit two diagnostic peaks located at 1362 and 1567 cm −1 corresponding to disordered amorphous carbon (D band) and graphite (G band), respectively, which demonstrate that the asprepared CN-PI x composites were more graphitic carbon-nitride-type materials when compared to g-C 3 N 4 nanosheets (Figure S6d, Supporting Information). [30]The broadband spanning the range of 2350-3300 cm −1 was assigned to the second-order Raman bands.An additional peak centered at 1291 cm −1 was observed, which corresponds to the C-H bending vibrational mode of the PTCDI molecules present in the CN-PI x composites. [48]he specific surface areas and pore size distributions of the g-C 3 N 4 nanosheets and CN-PI x composite samples were measured using nitrogen adsorption-desorption isotherms (Figure 1f).It was found that the CN-PI x samples exhibit an IUPAC type IV isotherm with a remarkable H4-type hysteresis loop at a high relative pressure (P/P 0 = 0.0-1.0) between the adsorption and desorption branches, indicating the existence of abundant mesopores in the g-C 3 N 4 nanosheets and CN-PI x composite samples.According to the Brunauer-Emmett-Teller (BET) analysis, the high specific surface area of the nitrogen-rich g-C 3 N 4 nanosheets was calculated to be 79.35 m 2 g −1 .This value was 197% higher than that previously reported for urea-driven g-C 3 N 4 powder (40-45.1 m 2 g −1 ) produced separately without a controlled NH 3 (g) atmosphere in the air, suggesting a large volume of Li + storage active sites was generated under the controlled NH 3 (g) atmosphere. [49]In contrast, the CN-PI x composites exhibit a decrease in the BET surface area from 64.38 m 2 g −1 for CN-PI 0.2 to 56.80 m 2 g −1 for CN-PI 1 upon increasing the PTCDI content (Table S4, Supporting Information).This result indicates that the PTCDI units are randomly distributed on the surface of the g-C 3 N 4 nanosheets and exist as micro-and mesopores upon increasing the PTCDI unit, as further confirmed by the SAED patterns shown in Figure 2e3 and Figure S10a4-c4, Supporting Information.Furthermore, the CN-PI x composites exhibit a larger pore size distribution ranging from 17.58 to 18.94 nm when compared to the g-C 3 N 4 nanosheets (147.7 Å), indicating that the as-prepared CN-PI x composites have a large number of mesopores due to the abundant redox-active sites with a range of different chemical environments available for realizing fast reversible Li + insertion/extraction kinetics. [27]3.Morphological Characterization of the g-C 3 N 4 Nanosheets and CN-PI x Composites The structural morphologies of the g-C 3 N 4 nanosheets and CN-PI x composite samples were characterized by scanning electron microscopy (SEM).Figure 2a shows the SEM images of the g-C 3 N 4 nanosheets, which has a graphene-like layered structure or stacked porous structures with relatively smooth surfaces when compared with their parent bulk g-C 3 N 4 material (Figure S7, Supporting Information).However, after the thermal polycondensation reaction using different amounts of PTCDA, the resulting CN-PI x composite samples were composed of layered graphitic nanoplatelet-like porous structures with thicknesses in the range of 100-200 nm, confirming that the ultra-thin g-C 3 N 4 nanosheets were self-assembled layer by layer to form a thick nanoplatelet-like structure morphology upon increasing the PTCDI unit from CN-PI 0.2 to CN-PI 1 , except for CN-PI 0.2 (Figure 2b; Figure S8, Supporting Information).Consequently, the overall layered nanoplatelet-like surface morphology of the g-C 3 N 4 nanosheets was almost retained after surface modification by the PTCDI units for CN-PI 0.2 to CN-PI 1 samples, which was further confirmed by the TEM images shown in Figure 2e1,e2 and Figure S10, Supporting Information.As can be seen in Figure 2b and Figure S9a,b, Supporting Information, the SEM images of CN-PI 1 show a thick nanoplatelet-like surface morphology as compared to that of other CN-PI x (x = 0.2, 0.5, and 0.75) samples (Figure S8, Supporting Information), indicating more PTCDI units covalently coupled with g-C 3 N 4 frameworks of the CN-PI 1 composite.To gain further insight into the morphology and thickness of CN-PI 1 sample, the tapping-mode AFM images were recorded (Figure 2c; Figure S9c, Supporting Information).The AFM topographic image of CN-PI 1 sample shows a very thick nanoplateletlike morphology (i.e., multi-layered nanosheets) deposited by spincoating on the FTO substrate.The corresponding cross-sectional height profile image shows the thickness of multi-layered CN-PI 1 nanosheets ranging from 300 to 440 nm (Figure S9c, Supporting Information).This result matches well with SEM images of CN-PI 1 samples (Figure 2b; Figure S9a,b, Supporting Information).
Figure 2d1 clearly shows the TEM images of g-C 3 N 4 nanosheets, which are constructed of many graphitic ultra-thin nanosheets with irregular, thick layered structures and transparency.Moreover, the high-resolution TEM (HR-TEM) image of g-C 3 N 4 shows the 2D-like graphitic structures were composed of many g-C 3 N 4 nanosheets with a high degree of porosity produced by clustering (Figure 2d2).The SAED patterns of the g-C 3 N 4 nanosheets present a weak outer diffraction ring and inner strong diffraction ring, which can be indexed as the Energy Environ.Mater.2023, 6, e12553 reflections of the (002) and (100) planes of the g-C 3 N 4 nanosheets, respectively, indicating that the g-C 3 N 4 nanosheets consist of significantly more crystallites with random orientations (Figure 2d3).This result was in good agreement with the XRD data (Figure 1d).After the thermal polycondensation reaction with different amounts of PTCDA, the CN-PI x (x = 0.5, 0.75, and 1) composite samples with higher PTCDI units were composed of an ultra-thin nanosheet morphology and amorphous character upon increasing the PTCDI unit from CN-PI 0.5 to CN-PI 1 (Figure 2e1,e2; Figure S10, Supporting Information).Especially, TEM images of CN-PI 1 show a very thick nanoplatelet-like morphology, indicating that more PTCDI unit covalently coupled with g-C 3 N 4 frameworks of CN-PI 1 (Figure S10c1,c2, Supporting Information), as strongly further justified by increased C/N ratio from the elemental analysis (Table S1, Supporting Information).Figure 2e2 and Figure S10a3-c3, Supporting Information show the HR-TEM images reveal the formation of multilayer nanosheets and thick nanosheets with increasing amorphous structure characteristics from CN-PI 0.5 to CN-PI 1 upon increasing the PTCDI unit, as strongly confirmed by the SAED patterns shown in Figure 2e3 and Figure S10a4-c4, Supporting Information.The HR-TEM image of CN-PI 1 shows more amorphous characteristics as compared to that of CN-PI 0.5 and CN-PI 0.75 samples, indicating that more PTCDI unit covalently coupled with the g-C 3 N 4 frameworks of CN-PI 1 (Figure S10c3, Supporting Information), as strongly evident from the SAED patterns for amorphous structure shown in Figure S10c4, Supporting Information.It is worth to note that as-synthesized CN-PI x (x = 0.5, 0.75, and 1) composites are composed of porous graphitic carbon nitride polymeric frameworks and PTCDI units.In CN-PI x composite samples, PTCDI units have randomly assembled in layer by layer into the g-C 3 N 4 framework to form few to multiple nanoplatelet (thick plates).Hence, TEM images showed the few-layer to multi-layered nanoplatelet-like structures for CN-PI x (x = 0.5, 0.75, and 1) samples (Figure 2e1,e2; Figure S10, Supporting Information).This inference was further strongly evident from AFM images of CN-PI 1 sample for thick nanoplatelet-like morphology with a thickness of about 300-440 nm (Figure 2c; Figure S9c, Supporting Information), in which very thick nanoplatelet-like morphology (i.e., multi-layered nanosheets) was observed for CN-PI 1 .These unique micro-morphologies and aggregated CN-PI x composites show good consistency with previous reports. [37,40]These observation suggested that thick graphene-like nanosheets with increased amorphous structures and abundant porosity at the edge and defect sites were produced in the CN-PI x composites.These features may provide abundant Li + diffusion channels and shorten the Li + diffusion pathways during Li + insertion/extraction kinetics. [50]Figure 2d4,e4 and Figure S10a5,b5, Supporting Information illustrate the STEM-EDX elemental mapping of the g-C 3 N 4 nanosheets and as-synthesized CN-PI x composite samples, demonstrating that the PTCDI units were incorporated into the framework of the g-C 3 N 4 nanosheets.These results were further supported by the weight and atomic percentage of the g-C 3 N 4 nanosheets and CN-PI x composite samples (Table S5, Supporting Information).

Electrochemical Properties of g-C 3 N 4 Nanosheets and CN-PI x Composite Anodes
Various electrochemical measurements were performed to investigate the electrochemical process and evaluate the Li + storage properties of the g-C 3 N 4 nanosheets, pristine PTCDA, and CN-PI x (x = 0.2, 0.5, 0.75, and 1) composite anodes (Figure 3; Figures S11 and S12, Supporting Information).Firstly, the cyclic voltammograms (CV) curve of pristine PTCDA displays five pairs of redox peaks at 2.4/2.4,2.07/2.3,1.29/1.014,0.96/0.88,and 0.5/0.01V in the first scan (Figure S11a, Supporting Information), corresponding to the Li + (de)-insertion reaction with C=O, anhydride rings, and C=C functional groups in the PTCDA unit and its interlayer spacing.These redox peaks area decreases with the increasing scanning from 2nd to 6th cycles, suggesting the high irreversible capacity with high polarization phenomenon and poor cycling reversibility of pristine PTCDA anode.Figure 3a shows the initial six CV curves obtained for the g-C 3 N 4 nanosheets anode within the potential window of 0.01-3.0V at a scan rate of 0.1 mV s −1 .The g-C 3 N 4 electrode shows typical CV curves for the initial six cycles, which shows good consistency with high-level pyridinic−N doped g-C 3 N 4 nanosheets and graphitic nitrogen-deficient g-C 3 N 4 nanosheets-based anode materials. [15,27]The strong reduction peaks located at 0.6 and 1.1 V during the first discharge cycle can be attributed to the irreversible formation of the solid-electrolyte interphase (SEI) film (i.e., formation and continuous growth of stable SEI film) on the surface of the g-C 3 N 4 electrode during the first full cycle, which causes the irreversible capacity loss in the initial cycle.This phenomenon was due to the occurrence of some side reactions on the electrode, leading to Li + trapping in the porous g-C 3 N 4 electrode induced by the poor extraction of Li + .These reduction peaks disappear in the consecutive cycles and almost overlap the CV curves from the second cycle onwards, indicating the highly stable and superior reversible electrochemical reactions occurring in the g-C 3 N 4 electrode after the first cycle.Another reduction peak occurs close to 0.01 V (R2) in the initial CV curves, which was attributed to the insertion of Li + into the interlayers and nanovoids of the g-C 3 N 4 nanosheets.This indicates the high density of Li + stored on the surface of the g-C 3 N 4 electrode. [26,28]During the charging process, the two broad anodic peaks located at 0.17 (O1) and 1.03 V (O2) can be assigned to the extraction of Li + (de-lithiation process) from the interlayers of the g-C 3 N 4 nanosheets and a large number of pores/surface defects, respectively. [29]Similar tendencies have also been observed in all of the CV curves obtained for the CN-PI x composites in terms of the irreversible formation of the SEI film, obvious reduction peaks (R1, R2, and R3) in the discharge cycles (insertion of Li + ), and oxidation peaks (O1 and O2) in the charge cycles (extraction of Li + ) (Figure 3b; Figure S12a-c, Supporting Information).It is worth to note that the strong reduction peak appeared at 1.02 V during the first cycle can be ascribed to the irreversible formation of the SEI film on the surface of the CN-PI 1 electrode, which causes the irreversible capacity loss at the first cycle.This is a reason for observing the initial irreversible capacity loss in the galvanostatic discharge/charge profile of CN-PI 1 (Figure 5f).In addition, small reduction peaks (R1) emerged in the initial CV curves at 1.77 V (CN-PI 0.2 ), 2.17 V (CN-PI 0.5 ), 2.38 V (CN-PI 0.75 ), and 2.11 V (CN-PI 1 ), which were attributed to the insertion of Li + into the carbonyl groups (C=O) of the PTCDI units and the interlayers of melon-PTCDI coupled in the CN-PI x composites.In other words, the carbonyl groups (C=O) were reduced with Li + to form C-O-Li bonds via an enolation reaction in the melon-PTCDI segment.Figure 3b and Figure S12a-c, Supporting Information show the aforementioned cathodic peaks (reduction peaks) in the CN-PI x composites almost overlapped or disappeared in the consecutive scans and then shifted to the lower potential region at 1.38 V (CN-PI 0.2 ), 1.84 V (CN-PI 0.5 ), 1.7 V (CN-PI 0.75 ), and 1.89 V (CN-PI 1 ), demonstrating the higher electrochemical stability and superior electrochemical reversible reaction in the CN-PI x composite electrodes.It is obvious that the CV curves of pristine PTCDA anode are quite different from the CV profiles of CN-PI x anodes except for a reduction peak located at 2.1 V, representing the fast reaction kinetics in CN-PI x anodes due to synergistic effect of g-C 3 N 4 frameworks in CN-PI x anodes (Figures S11 and S12, Supporting Information).Furthermore, the CV curves were obtained for the g-C 3 N 4 and CN-PI x electrodes at different scan rates to investigate the charge storage mechanism and electrochemical reaction kinetics (diffusion coefficient of Li + ) for all of the anodes (Figure 3c,d; Figure S12d-f, Supporting Information), as expressed by the modified Randles-Sevcik Equation (1). [47] where i p is the anodic current peak, n is the number of electrons transferred during the redox reaction (assuming n = 1 for the formation of Li-N=C-R 2 and C-O-Li species), A is the electrochemically active surface area (cm 2 ), D is the diffusion coefficient (cm 2 s −1 ), C is the concentration of redox species in 1 M LiPF 6 (mol cm −3 ), and v is the scanning rate of CV (mV s −1 ).The lithium diffusion coefficients (D o ) for all of the redox peaks were determined for the g-C 3 N 4 nanosheets and CN-PI x composite anodes, and the overall results are presented in Table S6, Supporting Information.CN-PI 0.75 shows the highest D o values at the highest intensity observed for the redox peaks when compared to those of the g-C 3 N 4 electrode and other CN-PI x (x = 0.2, 0.5, and 1) composite anodes.In particular, the D o value of CN-PI 0.75 was 7.9128 × 10 −5 cm 2 s −1 based on the maximum intensity observed in the reduction peak centered at 0.01 V (R3) and 1.006 × 10 −4 cm 2 s −1 at the highest intensity observed in the oxidation peak located at 0.17 V (O1 composite samples decreases upon increasing the PTCDA content from CN-PI 0.2 to CN-PI 1 , which can be ascribed to the increasing nature of the amorphous structures from crystalline CN-PI 0.2 to CN-PI 1 .These CV profiles of g-C 3 N 4 electrode and CN-PI x anodes were further confirmed by CV measurements at different high scan rates ranging from 0.5 to 10 mV s −1 , as shown in Figure S13, Supporting Information.All redox peaks of g-C 3 N 4 nanosheets and CN-PI x anodes were increased with increasing scan rates from 0.5 to 10 mV s −1 .It can be seen that CV curves were differentiated from CV profiles performed for all anodes at the low scanning rates, indicating excellent electrochemical properties at high scan rates.Furthermore, the sweep-rate-dependent CVs observed for the g-C 3 N 4 nanosheets and CN-PI x composite electrodes were investigated in terms of their charge storage kinetics.The ideal CV responses of the g-C 3 N 4 nanosheets and CN-PI x composite electrodes at various sweep rates (v) can be estimated using Equation (2): [51] i where i is the redox peak current (mA cm −2 ), v is the corresponding scan rate (mV s −1 ), and a and b are constants, which are important indicators of the charge storage kinetics.The b-value can be determined from the slope of the log (i) versus log (v) plot (Figure 3e,f; Figure S12g-i, Supporting Information).The b-value is an indicator of surface capacitive process that is free of diffusion limitations (if b = 1) and diffusioncontrolled process (if b = 0.5).In particular, the capacitive process free of diffusion limitations indicates the electrochemical double layer (EDL) and pseudocapacitive storage processes.The calculated b values at the cathodic peak regimes for all the anodes were estimated to be <0.5 (≈0.029-0.368),as shown in Figure 3e,f and Figure S12g-i, Supporting Information.These results indicate that the charge-discharge process in the g-C 3 N 4 nanosheets and CN-PI x anodes have predominantly pseudo-capacitive behavior or surface capacitive process as well as a combination of both diffusion capacitive and surface capacitive process.Generally, the range for b value is 0.5-1 corresponding to slow diffusion to highly capacitive nature of the electrode material.In our case, it is observed that b values for all anode showed lowest values than 0.5 when performing CV measurements at low scanning rates (0.1-0.5 mV s −1 ).The b values for all anodes were estimated to be <0.5 (≈0.029-0.368),which can be associated with thickness of active materials causing slow Li + diffusion rate with highly surface-capacitive effect and low current density during cycling at low current rates.Therefore, the CV measurement was performed at different high scan rates ranging from 0.5 to 10 mV s −1 for all anodes.The estimated the b values for g-C 3 N 4 and CN-PI x electrodes were found to be b = ∼0.611-1.0 for the reduction peaks and b = ∼0.684-0.877for the oxidation peaks (Figure S14, Supporting Information).The b values of g-C 3 N 4 (b = 0.611 for the reduction peaks) and CN-PI x electrodes (b = ∼0.640-1)strongly confirmed that pseudo-capacitive behavior or surface-capacitive effect is predominantly dominant in all anodes.This similar phenomena and tendencies were also observed for low b values obtained at the low scanning rates (Figure 3e,f; Figure S12g-i, Supporting Information).These results illustrated that a surface-capacitive lithium storage mechanism contributed a large portion of the total capacity for g-C 3 N 4 and CN-PI x anodes, especially at high-rate long-life cycling.Therefore, the surface-charge storage mechanism of g-C 3 N 4 and CN-PI x anodes can be expected to the superior rate performance of CN-PI x anodes.
To further elucidate the charge/discharge mechanism of g-C 3 N 4 and the CN-PI x anodes, the charge contribution ratios can be analyzed using the following two-part equation, that is, the surface-controlled capacitive process (k 1 v) and diffusion-controlled faradaic process (k 2 v 1/2 ): where k 1 v and k 2 v 1/2 are the capacitive-controlled and diffusioncontrolled charge contributions, respectively and v represents the scan rate.The testing current (i) at a certain potential (V) can be quantitatively analyzed using the values of k 1 (the slope of each line) and k 2 (the intercept of each line).The total charge contributed ratio was estimated for the g-C 3 N 4 and CN-PI x anodes under different scan rates (0.1-1.0 mV s −1 ) using Equation (3) from the shadow region observed in the CV profiles of the total stored charges for the capacitive-controlled and diffusion-controlled reaction, as shown in Figure 4a-d.Figure 4a shows the diffusion-controlled Faradaic process was indicated by the orange region for the g-C 3 N 4 anode, which accounted for 71.5% of the total stored charge at a scan rate of 0.1 mV s −1 .The blue region indicates the contribution of the surface capacitive-controlled process, which was 28.5%.The CN-PI 1 anode has capacitive-and diffusion-controlled contributions of 26.4% and 73.6%, respectively, at a scan rate of 0.1 mV s −1 (Figure 4c,f).The fraction of the surface capacitive contribution ratios gradually increases with the increasing the scan rates and then reaches a maximum value of 90.1% for the g-C 3 N 4 anode and 83.4% for the CN-PI 1 anode at a high scan rate of 1.0 mV s −1 , as indicated by the diffusion-and capacitive-controlled contribution curves shown in Figure 4b,d.Similar tendencies were also observed in the total contribution ratios of the CN-PI 0.2 and CN-PI 0.5 anodes in terms of increasing the capacitive contribution ratio in the range of 26.7-84.5% (CN-PI 0.2 ) and 25.8-81.6%(CN-PI 0.5 ) upon increasing the scan rate from 0.1 to 1.0 mV s −1 (Figure S15a,b, Supporting Information).It should be noted that the charge contribution ratios of 25.8-28.5% and 81.6-90.1% for the g-C 3 N 4 anode and CN-PI x (x = 0.2, 0.5, and 1) anodes mainly originate from the characteristic pseudo-capacitive contribution or surface-controlled capacitive process at 0.1 and 1.0 mV s −1 , respectively.This similar trend of dominant characteristic of surface-controlled charge contribution ratio was also observed and showed a liner relationship with increasing high scanning rates (0.5-10 mV s −1 ) for only g-C 3 N 4 and CN-PI 1 anodes (Figure S16, Supporting Information).These results strongly indicate that the unusual continuously increasing capacity of the g-C 3 N 4 nanosheets and CN-PI x (x = 0.2, 0.5, 0.75 and 1) anodes in LIBs mainly originates from the high contribution ratio of the surface capacitive-controlled process (Figure 4e,f; Figures S15 and S16, Supporting Information).This high fraction of surface capacitive contribution may be associated with the shorter ion-diffusion length and efficient electron transfer at high current densities, which can be further confirmed by the excellent rate performance.This result demonstrates that the CN-PI x (x = 0.2, 0.5, 0.75 and 1) composite anodes show fast redox reactions and rate-independent behavior during the charge-discharge process.Although the CN-PI 0.75 anode shows a small capacitive contribution ratio of 16.1% at a scan rate of 0.1 mV s −1 , it has an almost equal charge contribution ratio in the capacitive-and diffusion-controlled effects of 50.8% and 49.2%, respectively at a high scan rate of 1.0 mV s −1 (Figure S15c, Supporting Information).This result indicates that CN-PI 0.75 has dominant pyrrolic-N "hole" defects (high content of pyrrolic−N of ∼24.67% in Figure 1c) located at the edges and surfaces, which are responsible for the diffusion-controlled Faradaic behavior at high scan rates when compared to the other anodes.Moreover, the characteristic of a diffusion-controlled contribution effect significantly diminishes upon increasing the PTCDI content from CN-PI 0.2 to CN-PI 1 , demonstrating that the surface modification of the g-C 3 N 4 nanosheets with an increasing amount of PTCDI unit boosts the electrochemical Li + storage properties of the CN-PI x composites.Furthermore, we evaluated the lithium intercalation/deintercalation diffusion coefficient by applying the GITT technique for CN-PI x (x = 0.2, 0.5, 0.75, and 1) anode materials (Figure S17a-d, Supporting Information).The lithium diffusion coefficient (D Li þ ) of CN-PI x anode materials was determined using the Fick's second law, which can be expressed by Equation ( 5): where Δt, m B , M B , V M , and A represent the pulse time (s), mass, molar mass, molar volume of the electrode material (cm 3 mol −1 ), and electrode surface area (cm 2 ), respectively.for lithiation; and 5.78 × 10 −8 -1.58 × 10 −11 cm 2 s −1 for delithiation).This result indicated that CN-PI 1 anode after activation has a large Li + diffusion coefficient and superior charge transfer kinetics than that of other CN-PI x (x = 0.2, 0.5, and 0.75) anode materials during the (de)lithiation process owing to the existence of multilayered nanoplatelets (thick plates) and different range of abundant chemical environments in CN-PI 1 .

Electrochemical Performance of the g-C 3 N 4 Nanosheets and CN-PI x Composite Anodes
The electrochemical Li + storage performances of the g-C 3 N 4 nanosheets, pristine PTCDA, and CN-PI x anodes were investigated using galvanostatic discharge/charge (GDC) experiments in the potential window of 0.01-3.0V at a current density of 100 mA g −1 (Figure 5; Figure S18, Supporting Information).Figure S18a, Supporting Information presents the GDC voltage profiles of the g-C 3 N 4 electrode, in which the discharge and charge curves have two apparent voltage plateaus at 0.35 and 1.27 V.The discharge curve has a steeper region (sloping region) in the potential range of 3.0-0.35V versus Li + /Li, corresponding to the Li + insertion reaction and adsorption on the pyridinic-N/pyrrolic-N "hole" defects of the g-C 3 N 4 layers, whereas the shallower region (plateau region) between 0.3 and 0.01 V can be attributed to the large volume of Li + filling in the many closed pores and Li + binding with surface heteroatoms (i.e., N and O atoms) of the g-C 3 N 4 anode. [20,52]These results are in accordance with their corresponding CV results (Figure 3a).The g-C 3 N 4 electrode delivers an initial specific capacity of ∼72 mAh g −1 (the areal capacity of ∼0.233 mWh cm −2 ) and continuously increases up to ∼123 mAh g −1 (∼0.387 mWh cm −2 ) over 500 cycles (Figure S18b, Supporting Information).These GDC results are consistent with previous reports of super-lithiation in g-C 3 N 4 -derived materials. [27,28]In addition, this result suggests that a longstanding activation process is required to boost the specific capacity of the g-C 3 N 4 anode, and the time period for g-C 3 N 4 anode activation suffers from poor Li + insertion/extraction kinetics.Therefore, the g-C 3 N 4 electrode capacity was retested after the 100th cycle and exhibited a continuous increase in the specific capacity over 500 cycles.Many studies determined that about 10-20% of the initial capacity is generally consumed in irreversible formation of SEI layer (i.e., formation and continuous growth of stable SEI film) on the surface of g-C 3 N 4 anode during the first full cycle, which cause the irreversible capacity loss in the initial cycle. [29,53]Such a consuming 10-20% of the initial capacity could then provide the excellent cycling performance of g-C 3 N 4 anode by gradual kinetic activation of anode during the extended cycling. [27]This phenomenon is strongly evident from the extended deep cycling of g-C 3 N 4 nanosheets after 100 cycles (Figure S18a,b, Supporting Information).This similar tendency is also revealed in CN-PI x (x = 0.2, 0.5, 0. This result strongly evident from their CV results of CN-PI x anodes (Figure 3a,b; Figure S12a-c, Supporting Information), in which the cathodic peak is present at 1.0 V in the first cycle, yet absent in subsequent cycles.From the second cycles, the CE is greatly improved at nearly 100% with insufficient capacity fading during extended cycles (Figure 5e,f).Furthermore, the discharge capacities of CN-PI x (x = 0.2, 0.5, 0.75 and 1) anodes gradually increased to exhibit an extraordinary ultrahigh capacities (≈8400 mAh g −1 for CN-PI 1 ) with a CE nearly 100%.This phenomenon should be explained by the slow penetration of electrolyte in the porous CN-PI x anodes and then gradual kinetic enhancement/activation of the anodes via the gradual use of accessible defect/edge sites (nitrogen triangular "holes" defect sites) as well as the redox-active aromatic C=C bonds and C=O groups of PTCDI unit on the large surface area during deep-discharge cycles over 500 times.This impressive enhanced electrochemical Li + storage performance can be accompanied by a drastic decrease in the charge transfer resistance, which was strongly supported by impedance spectroscopy (Figure 7c).This demonstrates that cycling of the g-C 3 N 4 anode leads to the activation of the porous g-C 3 N 4 electrode with lower charge transfer resistance values, and thus, it shows a linear increase in the specific capacity of the g-C 3 N 4 anode over 500 cycles (Figure S18b, Supporting Information).These similar tendencies were also observed in the CN-PI x composite anodes in terms of the two apparent discharge/charge voltage plateau observed at 0. (∼0.985 mWh cm −2 , and 1150 mAh g −1 or 1.74 mWh cm −2 at the 380th cycle) for CN-PI 0.75 at a current density of 100 mA g −1 over 500 cycles (Figure 5e).Interestingly, CN-PI 1 delivers an unprecedented and highly stable discharge capacity exceeding 8400 mAh g −1 (∼1.96 mWh cm −2 ) even after 225 cycles and then shows saturated capacity retention (Figure 5f).These result demonstrated that the lithium storage in CN-PI x (x = 0.2, 0.5, 0.75, and 1) composite anodes follow an "intercalation/defect adsorption-closed pore filling" mechanism.That is, Li + intercalation reaction and defect adsorption mainly contributes to the sloping region in the discharge curves, whereas the long discharge plateau region is mainly attributed to closed pore filling of Li + in the g-C 3 N 4 frameworks of the CN-PI x composite anodes. [52]As a result, CN-PI 1 has long contribution of discharge plateau region than that of the other composite anodes (Figure 5a-d), which should be responsible for the exceptional lithium storage capacity exceeding 8400 mAh g −1 owing to large volume of Li + filling in the many closed pores in the g-C 3 N 4 frameworks of the CN-PI 1 during the long-term charging/discharging process.This phenomenon of closed pore filling of lithium storage mechanism contribution to the long discharge plateau region and large lithium storage capacities was well corroborated with as shown in the earlier investigation by Cheng et al. [52] Moreover, the capacities observed for CN-PI 0.5 , CN-PI 0.75 , and CN-PI 1 were about 1.02-12 times higher than the theoretical storage capacity of g-C 3 N 4 (524 mAh g −1 ), demonstrating the outstanding electrochemical performance of the CN-PI x composite anodes when compared to the g-C 3 N 4 anode.In contrast to g-C 3 N 4 anode, the battery performance of CN-PI 1 electrode was impressive.Hence, the electrochemical performance of PTCDA was investigated to better demonstrate the synergistic effect of g-C 3 N 4 and PTCDA content in CN-PI x (x = 0.2, 0.5, 0.75, and 1) composite anodes (Figure S18c,d S11a, Supporting Information).In order to provide reliable mechanism for electrochemically highly stable after inserting more lithium and capturing more electrons through multi-step reversible lithiation process at every carbon sites of PTCDA, the first deepdischarge cycle of pristine PTCDA was taken to account for the multi-step reversible lithium storage mechanism as shown in Figure S11b, Supporting Information.As Figure S11b, Supporting Information shows, the possible electrochemical process for Li-ion incorporation reaction involved in the two major steps with five distinct discharge voltage plateaus, so as forming plausibly a total of 28 Li-ions insertion (PTCDAÁ28Li) with the carbonyl groups, anhydride rings, and at every free carbon sites of PTCDA molecule in the deep-discharge voltage of 0.002 V (Figure S11b,c, Supporting Information and more detailed explanation in the Appendix S1, Supporting Information).As a result, the outstanding electrochemical performance of the CN-PI x (x = 0.2, 0.5, 0.75, and 1) composite anodes was attributed to the improved intrinsic electronic conductivity via the π-π stacking interactions of the aromatic rings of the melon-PTCDI segments, the excessive existence of graphene-like g-C 3 N 4 nanosheets with mesoporous structures and high specific surface area.To the best of our knowledge, the unprecedented and highly stable discharge capacity exceeding 8400 mAh g −1 (∼1.96 mWh cm −2 ) observed for the CN-PI 1 anode (cycling period >420 days) was much larger than the values obtained using graphitic carbon nitride, the theoretical value of graphene (744 mAh g −1 ), graphitic carbon-based anode materials, and N-doped g-C 3 N 4 -based carbon materials (Table S7, Supporting Information).Therefore, CN-PI 1 can be used as a low-cost anode material with a low working potential (<0.5 V vs Li + /Li), which makes it a highly promising anode candidate with both high specific energy density (E sp ) and volumetric energy density (E v ).The calculated E sp of ∼7700 Wh kg −1 and E v of ∼14 956 Wh L −1 were realized for CN-PI 1 anode, which shows a large improvement compared to those of CN-PI 0.2 (E sp : ∼150.6 Wh kg −1 and E v : 287 Wh L −1 ), CN-PI 0.5 (E sp : ∼507 Wh kg −1 and E v : 1232 Wh L −1 ), and CN-PI 0.75 (E sp : ∼726 Wh kg −1 and E v : 6342 Wh L −1 ), and also the commercial graphite (E sp : ∼200 Wh kg −1 and E v : ∼570 Wh L −1 ).Moreover, the CN-PI 0.2 and CN-PI 0.5 anodes delivered the increasing capacity with cycle numbers as compared to that of CN-PI 0.75 anode (i.e., the capacity of CN-PI 0.75 anode increases in the initial 350 cycles and fades in the following 150 cycles), indicating that CN-PI 0.2 and CN-PI 0.5 anodes have high structural integrity with high homogeneity of the composite during extended cycling when compared to CN-PI 0.75 anode.These phenomena can further be justified by the ex situ SEM images for the cycled CN-PI x (x = 0.2, 0.5, 0.75, and 1) anodes (Figure S19, Supporting Information).The ex situ SEM images of CN-PI 0.2 and CN-PI 0.5 anodes showed a high structural integrity with less crater morphology as well as very less delaminated electrode surface (Figure S19a,b, Supporting Information).This should be accounted by observing increased capacity of both CN-PI 0.2 and CN-PI 0.5 anodes with increasing cycle numbers.However, the ex situ SEM image of CN-PI 0.75 anode displayed the more unfavorable crater morphology with large size of holes and abundant cracks (Figure S19c, Supporting Information), whereas the ex situ SEM image of CN-PI 1 anode displayed almost homogeneity of the composite film with fewer holes and a very thick SEI layer over the electrode surface (Figure S19d, Supporting Information).The inset optical image of CN-PI 0.75 anode has more delaminated electrode surface morphology as compared to that of CN-PI 1 anode.Such a more delaminated electrode surface of CN-PI 0.75 anode should probably be a reason for observing the more volcano-shaped charge-discharge cycling after 380 cycles for CN-PI 0.75 anode (Figure 5e) and less capacity degradation after 225 cycles for CN-PI 1 anode (Figure 5f).Moreover, the CEs of all of the g-C 3 N 4 , PTCDA, and CN-PI x composite anodes show a rapid stabilization at 94-96% during the initial 20 cycles and then approached ∼100% (Figure 5e,f; Figure S18b,d, Supporting Information).Relatively low initial CEs were observed in the CN-PI x composite samples due to the side reaction with oxygen-containing functional groups or other structural defect sites (C 2 N species).As revealed by the differences in the Li + storage performance shown in Figure 5a-d, this was further supported by the different capacity versus voltage (dQ dV −1 vs V) plots obtained for the 1st and 500th charging cycles during the Li + extraction process at 100 mA g −1 (Figure S20a, b, Supporting Information).The dQ dV −1 versus V plots can be described by the unprecedented and highly stable increasing specific capacity extracted from the working anodes (g-C 3 N 4 and CN-PI x composites) over a given potential range of 0.01-3.0V.The dQ dV −1 versus V plots for the g-C 3 N 4 nanosheets and CN-PI x composite anodes show an unusually high-capacity contribution throughout the entire potential range of 0.01-3.0V, which was attributed to the high density of Li + extraction from the following sites: 1) carbon layers/unsaturated C 6 aromatic rings (i.e., the heptazine units in the g-C 3 N 4 and the perylene cores in the CN-PI x anodes) at 0.01-0.8V, [21,36] 2) the closed pores and defects sites in the voltage range of 1.0-1.4 or 0.9-1.4V, 3) the N-C-O-Li sites formed by the lithium enolate species between the N-C=O groups and Li + at 1.5-2.1 V, 4) high-density Li + binding with the combined pyridinic-N/ pyrrolic-N atoms present in the void spaces of the anode surface in the voltage range of 2.2-3.0V. [30] During the 1st charge cycle, the g-C 3 N 4 and CN-PI x anodes show a higher efficiency of Li + extraction at lower potential from the carbon layers/unsaturated C 6 aromatic rings of PTCDI units and closed pores/defects rather than extraction from the Li + bound to the N atoms (Figure S20a, Supporting Information).This trend was decreased upon increasing the PTCDI unit from CN-PI 0.2 to CN-PI 0.75 , which then increases the efficiency of Li + extraction in CN-PI 1 anode at lower potential with highly reversible and stable Li + extraction from the Li-N surface binding sites in the voltage range of 2.2-3.0V.At the 500th charge cycle, the g-C 3 N 4 and CN-PI x anodes show higher efficiency for Li + extraction from the Li-N surface binding sites rather than from the carbon layers/unsaturated C 6 aromatic rings of the PTCDI units and closed pores/defects (Figure S20b, Supporting Information), which enables unusually excellent capacity retention over 500 cycles.These results indicate that an unprecedented and highly stable incremental capacity appears from the CN-PI 0.2 to CN-PI 1 anodes due to the activation of the many closed pores and defects, an increase in the large amount of C=O/C-O species in PTCDI units, and the associated increase in the longstanding kinetic activity of the Li-N surface binding sites, as shown by the sharp peaks that emerge in the range of 2.5-3.0V (Figure S20b, Supporting Information).These observations were further correlated with the proposed electrochemical redox reaction mechanism of the CN-PI x composite anodes during the charge-discharge process shown in Figure 6a.The overall electrochemical Li + insertion/extraction process of the CN-PI x composite anodes involved the following distinct steps: 1) Step 1 involves that each PTCDI unit in the CN-PI x composites can reversibly exchange eight electrons through the direct formation of the octa-anions (CN-PI x 8− Á8Li + ) compounds via an enolization reaction between the opposite C=O/ C-O species and insertion of 8 mol of Li + ions in the voltage range of 3.0-2.0V.This electrochemical redox mechanism in the PTCDI units of the CN-PI x composites could be explained in three subsequent steps by fundamental electrochemistry of PTCDI units. [54]In the initial stage of discharge, one electron was initially captured from the C=O group of each PTCDI unit in the CN-PI Á8Li + compounds with insertion of 8 mol of Li + ions in the voltage range of 3.0-2.0V, while C=O groups were rebuilded and returned to the original state (accompanied by delithiation) during the charge process (oxidation process); 2) Step 2 involves a total of 16 electron transfer redox steps occur in the CN-PI x composite to generate totally CN-PI x

16−
Á16Li + compounds with another 8 mol of Li + insertion in the voltage range of 2.0-1.05 or 2.0-1.15V.In principle, the aromatic imides of each PTCDI unit in the CN-PI x composites have four C=O groups, which could reversibly be exchanged four electrons for electrochemical reversible accommodation of 4 Li + ions, so as forming in a total of 16 mol of Li + ions in the four PTCDI units of the CN-PI x composites to obtain electrogenerated CN-PI x 16− Á16Li + compounds, as shown in Step 2 of Figure 6a.Therefore, these electrochemical redox reactions fundamentally take place in the imide groups of PTCDI units through sequential CN-PI x composites structural rearrangements via an electrochemical enolization reaction between their 16 C=O groups and associating/disassociating of Li + ions from the oxygen atoms of the imide functionality.During the charging process in Step 2, the g-C 3 N 4 or CN-PI x composite losses electrons at the N atoms to generate positively charged (C) 3 -N + moieties at both the triazine ring of the heptazine units and imide position of the perylene core (=N + -(C) 2 ), at which PF 6 − will be absorbed on (C)  (C) 2 ) groups for compensating its charge.During the discharge process, PF 6 − will be released from CN-PI x composites, and the (C) 3 -N + and (=N + -(C) 2 ) bonds will be reversibly reduced into their original state of single (C) 3 -N and (-N-(C) 2 ) bonds in the voltage range of 2.0-1.05 or 2.0-1.15V. [54] The electrochemical PF 6 − storage of CN-PI x composites is mainly correlated to the reversible bond, forming/breaking of (C) 3 -N + and (=N + -(C) 2 ) to (C) 3 -N and (-N-(C) 2 ) bonds during the charge-discharge process.To support for Step 2 process, the variation of chargetransfer resistance was identified at different voltage ranges of 0.01-3 V (Figure S21, Supporting Information).The g-C 3 N 4 anode shows that the interfacial charge-transfer resistance (R ct ) with lithium diffusion resistances (R Li ) became smaller and decreased the interfacial resistance (R e ) between the electrolyte and electrode when voltage decreases from 3 to 0.01 V during discharging (Figure S21a, Table S8 Á172Li + species with a high density of 156 mol of Li + insertion at every free carbon sites of CN-PI x structure during deepdischarging in the voltage range of 1.05-0.01V.According to density functional theory (DFT) studies by Hankel et al., [55] the lithium mainly interacting with the pyridinic-N/pyrrolic-N functional groups (so-called "nitrogen triangular holes" defects sites) present in the g-C 3 N 4 nanosheets, which enabled an ultrahigh lithium uptake during (de)insertion process.As can be seen Figure 6a of Step 3, when deep-discharging in the voltage range of 1.05-0.01V, the high density of Li + ions were inserted into the g-C 3 N 4 frameworks of the CN-PI x anodes, in which one Li + was initially adsorbed/binding by three pyridinic-N atoms (i.e., one Li + adsorbed into one nitrogen triangular hole), followed by the adsorption of two, three, and up to six Li + insertion to each pyridinic-N atom in the void space to form the Li-N binding sites.In addition to that, the higher content of covalently coupled melon-PTCDI molecules also provided abundant redox-active sites for much higher Li + insertion with C=O groups, anhydride rings, and gradual incorporation of large Li + ions into the unsaturated carbon atoms of the C 6 aromatic rings of the composite structure (lithiation at every free carbon sites of PTCDI units and g-C 3 N 4 frameworks), as strongly evident from the first deep-discharge cycle for pristine PTCDA anode (Figure S11b,c, Supporting Information), in which forming plausibly a total of 28 Li + ions insertion (PTCDAÁ28Li) with the carbonyl groups, anhydride rings, and at every free carbon sites of PTCDA molecule in the deep-discharge voltage of 0.002 V (as already discussed above).Therefore, the electrochemical participation of these closed pores and defect sites of g-C 3 N 4 frameworks and organic functional groups of PTCDI with lithium contributed to have the superior lithium storage capacities for as-synthesized CN-PI x composite anodes.The calculated theoretical storage capacities (C theo , mAh g −1 ) for the electrogenerated species in the CN-PI x anodes are given in each step.The calculated theoretical capacities were relatively in good agreement with the discharge capacity observed in the CN-PI x anodes in the given voltage range (Figure 5).
Until now, it is noticed that as-synthesized CN-PI x composite exhibited an unprecedented and continuously increasing ultrahigh capacities up to 8400 mAh g −1 for CN-PI 1 anode.In order to support these ultrahigh lithium storage capacities of CN-PI 1 , we have proposed a possible lithium storage mechanism in which the theoretical storage capacities are up to 1777 mAh g −1 as per of total 172 electron transfer process through multi-step lithiation with the carbonyl groups, anhydride rings of PTCDI units and at every free carbon sites of composite structure as shown in Step 3 of Figure 6a.To further support the proposed lithium storage mechanism, we have also performed a theoretical studies for a single part (single segment; CN-PI) of CN-PI x composite structures (as shown in Figure 6b).To evaluate the energetic preference of formations between Li bulk versus Li-ion adsorbed on CN-PI structure (g-C 3 N 4 -PI), about 24 Li atoms were placed at the proposed active sites.With the computed Li adsorbed CN-PI configurations shown in Figure 6c, we calculated the energetic preferences by following Equation ( 6).
where E LiþCNÀPI is the energy of 24 Li + ions on the CN-PI; E CN-PI is the energy of the isolated CN-PI; E Li bulk is the energy of bulk Li.To be consistent with the number of Li + ion in bulk Li metal and Li + on CN-PI, the bulk system size was expanded to 1 × 1 × 6 with Gamma KPOINTS mesh, 2 × 2 × 1.From the Equation ( 6), a positive (negative) value indicates adsorbed Li + ions on CN-PI is more (less) favorable than the formation of bulk Li.The simulation provided a negative E pref (−0.93 eV) suggesting that the configuration of 24 Li + ions adsorbed on CN-PI structure is energetically less favorable than Li bulk formation.In other words, the bulk Li metal layer/deposition favorably forms during the electrochemical kinetic reactions in the long-term charging periods.This behavior could trigger the experimentally observed extraordinary ultrahigh capacities exceeding 8400 mAh g −1 .Consequently, theoretical calculation predicted that a total of 24 Li + ions (i.e., ∼17 Li + ions in a PTCDI unit and 7 Li + ions in a melem unit of g-C 3 N 4 ) can be adsorbed/stored in a CN-PI single structure.According to previous DFT studies by Hankel et al. [55,56] (i.e., the pores of g-C 3 N 4 could be preferred more adsorption sites for more Li binding) and our theoretical prediction, it can be assumed that a total of 115 Li atoms could be plausibly adsorbed/stored in a complete composite structures (i.e., 24  Li + ×4 single segment, 7 Li + ions in one melem unit and 6 Li + × 2 in each pore of heptazine structure).By considering a total adsorption of 115 Li + ions on the as-proposed complete CN-PI x composite structure, the theoretical storage capacities could be plausibly found to be ∼1188 mAh g −1 as per a total of 115 electron transfer process.Therefore, it is justified that the theoretical prediction is relatively matching in line with our proposed lithium storage mechanism and theoretical storage capacities (Figure 6a).The promising Li + storage properties of g-C 3 N 4 nanosheets, pristine PTCDA, and CN-PI x composite anodes were further emphasized by their superior high-rate performance.The rate performances of all the anodes were determined using GDC over 20 consecutive cycles ranging from 0.1 to 20 A g −1 (Figure 7a).
The CN-PI 1 anode exhibits a significantly superior rate capacity than that of g-C 3 N 4 , PTCDA, and the other CN-PI x (x = 0.2, 0.5, and 0.75) anodes (Figure S22a, Supporting Information).Specifically, initial discharge capacities of 934, 912, 737, 575, 426, 224, 75, and 33 mAh g −1 (∼0.077-2.176mWh cm −2 ) were realized for the CN-PI 1 anode at current densities of 0.1, 0.2, 0.5, 1, 2, 5, 10, and 20 A g −1 , respectively, delivering a highly reversible discharge capacity of 1847 mAh g −1 (∼4.303 mWh cm −2 ) after 210 Energy Environ.Mater.2023, 6, e12553 cycles at 0.1 A g −1 , which were much higher than g-C 3 N 4 -based anode materials and other optimized N-doped g-C 3 N 4 -based carbon materials summarized in Table S7, Supporting Information.On the other hand, the reversible discharge capacity of other CN-PI x anodes gradually increases and returns to ∼57 mAh g −1 (∼0.206 mWh cm −2 ) for g-C 3 N 4 ns and ∼163 mAh g −1 (∼0.171 mWh cm −2 ) for CN-PI 0.2 , ∼298 mAh g −1 (∼0.278 mWh cm −2 ) for CN-PI 0.5 , and ∼309 mAh g −1 (∼0.99 mWh cm −2 ) for CN-PI 0.75 over 210 cycles, which demonstrates the outstanding cycling stability and ultra-fast rate capability of the CN-PI x composites at various current densities.In contrast, the corresponding specific capacities of the g-C 3 N 4 anode were 88, 57, 31, 20, 13, 6, 2, and 2.8 mAh g −1 (∼0.011-0.318mWh cm −2 ), which are far below those of the CN-PI x composite anodes.In addition, the g-C 3 N 4 anode suffers from capacity degradation as the current density reverts to 0.1 A g −1 after 160 cycles due to the high nitrogen content (∼56.59%) in g-C 3 N 4 , leading to a degradation in the cyclability during the ultra-fast charging rate (Figure 7a).Moreover, the PTCDA battery exhibits a better rate capability with a reversible capacity of 82 mAh g −1 (∼0.32 mWh cm −2 ) at 0.1 A g −1 over 50 cycles.The CN-PI 0.5 anode showed low-rate performance at the current densities (∼0.1-5A g −1 ) compared with that of CN-PI 0.2 (Figure S22a, Supporting Information).However, CN-PI 0.5 has a high reversible capacity after being reversed back to 0.1 A g −1 .This was probably due to the longstanding activation and superlithiation process induced by the gradual use of accessible a highest pyridinic-N/pyrrolic-N configuration with a reduced graphitic-N ratio, the excessive porous g-C 3 N 4 framework, and higher PTCDI units in the CN-PI 0.5 .All g-C 3 N 4 nanosheets, pristine PTCDA, and CN-PI x composite anodes showed poor rate performance at an ultrahigh current density of 20 A g −1 (Figure 7a; Figure S22a, Supporting Information), implying that the rate capability for all anode materials was restricted by the insufficient wetting of the electrode.Therefore, this feature leads to poor Li + transports into or through the framework of g-C 3 N 4 nanosheets, PTCDA, and CN-PI x composites.
The long-term cycling stability and CE of the as-synthesized g-C 3 N 4 , PTCDA, and CN-PI x composite anodes were estimated using GDC measurements at a high charging rate of 1 A g −1 over 1000 cycles (Figure 7b; Figure S22b, Supporting Information).The specific capacity of the CN-PI x anodes continuously increases and unprecedented discharge capacities from a negligible contribution of active materials during the 10th cycle and good durability when compared to the g-C 3 N 4 and PTCDA anodes.In particular, CN-PI 1 shows an unprecedented discharge capacity of ∼414 mAh g −1 (∼0.579 mWh cm −2 ) over 1000 cycles, which is larger than that of the g-C 3 N 4 (∼57 mAh g −1 ; ∼0.22 mWh cm −2 ), PTCDA (∼32 mAh g −1 ; ∼0.13 mWh cm −2 ), CN-PI 0.2 (∼157 mAh g −1 ; ∼0.366 mWh cm −2 ), CN-PI 0.5 (∼351 mAh g −1 ; ∼0.981 mWh cm −2 ), and CN-PI 0.75 (∼403 mAh g −1 ; ∼0.751 mWh cm −2 ) anodes.Moreover, the pristine PTCDA anode exhibits a discharge capacity of 54 mAh g −1 (∼0.219 mWh cm −2 ) at 1 A g −1 over 2000 cycles (Figure S23a, Supporting Information), indicating outstanding cyclic stability of pristine PTCDA anode due to kinetic enhancement and stable SEI layer formation during the long-standing activation process.Another cell of pristine PTCDA anode shows an initial discharge capacity of 11 mAh g −1 (based on the high mass loading of electrode) and then continuously increased up to ∼24 mAh g −1 (∼0.097 mWh cm −2 ) over 8500 cycles at 1 A g −1 (Figure S23b, Supporting Information).These results strongly demonstrated that PTCDA molecule has more Li + ions binding active sites, in which about a total of 28 Li + ions could be lithiated with carbonyl groups, anhydride rings, and at every carbon sites of PTCDA molecule at the voltage below 1.06 V as strongly evident from Figure S11b,c, Supporting Information.These kinetic phenomena and synergistic effect of PTCDA should be the decisive factor for higher specific capacity and excellent electrochemical performance of the as-prepared CN-PI x composites.As a result, the excellent cycling stability of the CN-PI x anodes was comparable to the g-C 3 N 4 nanosheets and pristine PTCDA anode.It can be seen that the capacity of CN-PI 1 anode gradually increased from 106 to 220 mAh g −1 with increasing cycling over 330 times and then the capacity increased suddenly after 331 cycles and reached about 414 mAh g −1 over 1000 cycles (Figure 7b).These electrochemical phenomena can be explained by involving long-standing kinetic enhancement/activation process in the porous CN-PI 1 anode.Generally, a long-standing activation process is required to boost the overall capacity of graphite/graphitic carbon nitride anodes.However, the long time period for electrode activation significantly suffers from poor (de)lithiation kinetic reaction.In order to check this phenomenon, we retested the capacity of the electrode after 100th cycle for g-C 3 N 4 and 330th cycle for CN-PI 1 (Figure 7b; Figure S18b, Supporting Information), which exhibited a continuously increasing ultrahigh capacity of 122 mAh g −1 (g-C 3 N 4 ) and 414 mAh g −1 (CN-PI 1 ) during extended cycles over 500 and 1000 times, respectively.These similar trend was also observed in the CN-PI 0.2 , CN-PI 0.5 , and CN-PI 0.75 after 50 cycles (Figure 5e).This result indicated that a long-standing activation of CN-PI 1 and followed by super-lithiation process can be induced by the gradual use of accessible a highest pyridinic-N/pyrrolic-N "holes" defect sites, the excessive porous g-C 3 N 4 framework of CN-PI x anodes, and abundant C=C/ C=O groups of PTCDI units in the CN-PI x composite anodes during extended cycling.The electrochemical participation of these functional groups with lithium can be further justified from dQ dV −1 versus V plots (Figure S20, Supporting Information).Moreover, it is worth to note that CN-PI 1 anode exhibited an unprecedented and continuously increasing higher specific capacity at different current densities as compared to g-C 3 N 4 , pristine PTCDA, and other CN-PI x (x = 0.2, 0.5 and 0.75) anodes, which should be attributed to the following reasons: 1) CN-PI 1 has high specific surface area (BET surface area of 56.80 m 2 g −1 and pore size distribution of 18.14 nm), which can shorten the Li + diffusion pathways by providing a large electrode/electrolyte contact area to facilitate the extremely high density of Li + accession, 2) higher content of covalently coupled melon-PTCDI molecules on the surface of CN-PI 1 , which provide abundant redox-active sites for much higher Li + insertion, large Li + filling on the closed pores and gradual incorporation of large Li + into the unsaturated carbon atoms of the C 6 aromatic rings of the PTCDI molecules, 3) the formation of a reduced N content in the CN-PI 1 anode with dangling bonds around the pyridinic-N/pyrrolic-N dopants and an appropriate amount of graphitic-N located at the edges/defect sites (C 2 N species) in the g-C 3 N 4 frameworks of CN-PI 1 can provide more electrochemical redoxactive sites for high-density Li + accession, which facilitate a large amount of Li + migration within the framework and electron transport over the entire surface of the CN-PI 1 anode.Consequently, these characteristic features pertaining to the CN-PI 1 anode are enabled to have extraordinary ultrahigh specific capacities at different current densities as compared to g-C 3 N 4 and pristine PTCDA anodes (Figures 5 and 7) and other CN-PI x (x = 0.2, 0.5, and 0.75) anodes (Figure S22, Supporting Information).Moreover, it is well-known that the multilayered graphitic nanoplatelet-like structure, abundant porosity, unsaturated carbon sites with abundant carbonyl groups, and nature of the electrode surface are the main factors for greatly affecting the specific capacity of electrode materials.In comparison with CN-PI 0.75 anode, CN-PI 1 anode exhibited an exceptional ultrahigh capacities up to 8400 mAh g −1 .This is due to the fact that CN-PI 1 has the multilayered graphitic nanoplatelet-like structure morphology (thick nanoplatelet-like structures) as compared to that of CN-PI 0.75 sample.Moreover, CN-PI 1 anode has abundant porous surface of the electrode, large amount of redox-active aromatic C=C bonds and C=O groups present in the PTCDI units (high degree of redox activity), and the existence of a high level of pyridinic-N/pyrrolic-N (C 2 N species) with a reduced ratio of graphitic-N located at the edge/defects sites, very thin electrode surface morphology, and the formation of very thick SEI film with well-defined Mosaic-like nanostructures during longstanding kinetic enhancement/activation process, which have greatly enabled to exhibit an overwhelmingly superior Li + storage capability of CN-PI 1 than that of other composite anodes.In addition to that, the coexistence of more Li metal deposition at the Cu-CN-PI 1 heterojunction as well as at the bottom of the Cu foil have also greatly contributed to continuously increase the ultrahigh capacities up to 8400 mAh g −1 for CN-PI 1 anode as compared to CN-PI 0.75 and other anodes (see Figure 9 for more detailed explanation).
To figure out the capacity contribution of CN-PI x (x = 0.2, 0.5, 0.75, and 1) anodes, the cyclic voltammetry (CV) analysis and cyclic performance of CN-PI x anodes were further performed within different cut-off potential in the range of 0.005-1.5 and 0.1-3.0V, respectively (Figure S24, Supporting Information).The CV curves of CN-PI x (x = 0.2, 0.5, 0.75, and 1) anodes showed the decreasing redox characteristic peaks with the increasing PTCDI unit in both potential ranges (Figure S24a,c, Supporting Information), which are almost similar to the CV profiles of CN-PI x (x = 0.2, 0.5, 0.75, and 1) anodes obtained at the voltage window between 0.01 and 3.0 V (Figure 3b; Figure S12a-c, Supporting Information).Further, the cyclic performance of CN-PI x (x = 0.2, 0.5, 0.75, and 1) anodes at 1 A g −1 showed lower battery performance as compared to that of higher performance metrics obtained at the voltage window of 0.01-3.0V (Figure 7b; Figure S22b, Supporting Information).The capacity contribution ratio at the voltage range of 0.1-3.0V is about 2.8-3 times higher as compared to those obtained at the voltage ranges of 0.005-1.5V (Figure S24b,d, Supporting Information), whereas the capacity contribution ratio was further increased about 2-3.7 times in the potential range between 0.01 and 3.0 V (Figure 7b; Figure S22b, Supporting Information).This result demonstrated that the capacity contribution ratio is significantly larger at the voltage range of 0.01-3.0V, caused by the gradual kinetic activation/enhancement of multi-layered nanoplatelets-like porous structures, abundant redox-active sites for much higher Li + insertion process and enhanced electrolyte decompositions products to form highly stable SEI layer from the voltage step at 0.01 V versus Li + /Li during the lithiation process.
To better understand the superior Li + storage properties of the CN-PI x composite anodes, the Li + diffusion behavior/charge transfer resistance was investigated using electrochemical impedance spectroscopy (EIS) for the fresh and cycled cells over 1000 cycles, as shown in Fig- ure 7c,d and Figure S25, Supporting Information.The Nyquist plots obtained for the g-C 3 N 4 , PTCDA, and CN-PI x (x = 0.2, 0.5, 0.75, and 1) cells were composed of one semicircle in the high-and medium-frequency zones, followed by a straight line in the lowfrequency area in the fresh state.By fitting the EIS results based on the application of an equivalent circuit to analyze the impedance spectra (Table S9, Supporting Information), g-C 3 N 4 and CN-PI 0.2 show a much larger semicircle diameter when compared to other CN-PI x cells before cycling due to the inactivated g-C 3 N 4 nanosheets with a high Energy Environ.Mater.2023, 6, e12553 graphitic-N content available for Li + storage, surface roughness, and non-homogeneous electrode surface morphology of CN-PI 0.2 , which leads to higher charge-transfer resistance (R ct ) value observed for the fresh cell of g-C 3 N 4 (486 Ω), PTCDA (238 Ω) and CN-PI 0.2 (15 100 Ω).These R ct values were drastically reduced and found to be 13 Ω for g-C 3 N 4 , 179 Ω for PTCDA, and 437 Ω for CN-PI 0.2 after 1000 cycles.The g-C 3 N 4 showed the R e value of 21.7 Ω at open circuit voltage (OCV; Table S9, Supporting Information), yet it showed increase in R e of 51 Ω during the applied voltage of 3.0 V (Figure S21b, Table S8, Supporting Information).This similar trend was also observed in CN-PI 1 (R e of 58 Ω at OCV and 118 Ω at 3.0 V), indicating the increase of conductivity of the CN-PI 1 as compared to that of g-C 3 N 4 anode.The CN-PI 0.5 , CN-PI 0.75 , and CN-PI 1 anodes exhibit favorable R ct values of 275, 185, and 205 Ω, respectively, before cycling, which are drastically reduced to 27, 107, and 120 Ω over 1000 cycles, indicating the longstanding activity of the CN-PI x (x = 0.5, 0.75, and 1) anodes for the adsorption/diffusion of large amounts of Li + and their preferential storage properties, resulting in the increase of the charge transfer kinetics.These results are reflected in the unprecedented, highly stable incremental discharge capacities observed for the CN-PI 0.5 , CN-PI 0.75 , and CN-PI 1 anodes.Moreover, the high intrinsic electronic conductivity with noticeably reduced R ct values observed for the CN-PI x anodes can facilitate a large volume of Li + migration and electron transport over the entire CN-PI x anode surface.Consequently, it can suppress the volume change, pulverization, and aggregation of the g-C 3 N 4 nanosheets in the CN-PI x anodes.These results exhibit a very good correlation with the improved amorphous structures observed when increasing the PTCDI units, which maintains a high degree of graphitization, as revealed in the Raman spectra (Figure 1e), and high structural integrity by SEM analysis (Figure 9; Figure S27, Supporting Information).The electrical resistivity of g-C 3 N 4 nanosheets and CN-PI x composite anodes were also determined based on its electrical resistance values and found to be ρ = 6.681 × 10 −4 , 2.151 × 10 −5 , 1.181 × 10 −3 , 1.755 × 10 −3 and 1.584 × 10 −3 Ω m −1 for g-C 3 N 4 nanosheets and CN-PI 0.2 , CN-PI 0.5 , CN-PI 0.75 and CN-PI 1 anodes, respectively.The electrical resistivity of CN-PI x composites was seemingly increased with increasing PTCDI units in CN-PI x composites as compared to g-C 3 N 4 nanosheets, indicating PTCDIs assembled on the surface of CN-PI x composites.

Evaluation of the Cycled g-C 3 N 4 and CN-PI x Composite Anodes
The chemical changes in the g-C 3 N 4 and CN-PI 1 anodes observed during the electrochemical redox process were monitored using ex situ FT-IR spectra (Figure S26a-d, Supporting Information).The spectra for all of the cycled electrodes were captured at five different states to identify the Li + storage sites during the discharging/charging process.The pristine g-C 3 N 4 electrode showed prominent vibrational bands in the region of 1110-1450 cm −1 (especially at 1203, 1228, and 1311 cm −1 and possibly 1392 cm −1 ), corresponding to the stretching vibration characteristic of either trigonally connected C-N(-C)-C (full condensation motif) or C-NH-C (partial condensation motif) bonds. [15,40,57]These vibrational bands were clearly weakened and exhibit three broad peaks at 1163, 1232, and 1392 cm −1 when the g-C 3 N 4 electrodes were fully discharged to 1.28 and 0.35, and then deep-discharged to 0.002 V, respectively (Figure S26a, Supporting Information).The strong vibrational bands located at 1452, 1533, 1568, and 1628 cm −1 can be attributed to the stretching vibration characteristics of the heptazine-derived repeating units in the pristine g-C 3 N 4 electrode.These vibrational bands were weakened to a broadened band at 1622 cm −1 during discharging to 1.28, 0.35, and 0.002 V.These results strongly indicate that a large amount of Li + adsorption occurs in the pyridinic-N/pyrrolic-N "hole" defects/edge sites and the interlayer of the g-C 3 N 4 anode. [25]Similar trends were also observed when the g-C 3 N 4 electrode was fully charged to 2.08 and 2.8 V.The band at 889 cm −1 was attributed to the presence of C-H out-of-plane bonds in the aromatic domains of the polymeric g-C 3 N 4 nanosheets, which disappeared entirely in the discharged and charged states.The sharp band located at ∼806 cm −1 was assigned to the ring-sextant out-of-plane bending vibration mode of the heptazine ring system.This sharp band was red-shifted (shifted to a higher wavenumber) to become a broadened peak at 835 cm −1 when the g-C 3 N 4 electrode was fully discharged to 1.28, 0.35, and 0.002 V, and then fully recharged voltage of 2.08 and 2.8 V.The broadbands observed in the region of 2800-3350 cm −1 correspond to the N-H stretching vibration characteristics of secondary and primary amines present in the pristine g-C 3 N 4 electrode (Figure S26b, Supporting Information). [58]These broadbands were obviously weakened and become a relatively broadened band positioned at 2928 cm −1 , indicating that a large amount of Li + adsorption occurred or preferential Li + intercalation occurred at the pyridinic-N/pyrrolic-N "hole" defects (C 2 N species) on the g-C 3 N 4 nanosheets and the edge of g-C 3 N 4 nanosheets and g-C 3 N 4 interlayers during discharging at 1.28, 0.35, and 0.002 V. [25] Considering the highly reversible redox behavior of the polymeric g-C 3 N 4 nanostructures, even after being recharged to 2.08 and 2.8 V, the g-C 3 N 4 electrodes still exhibit broadbands at 1622, 1392, 1232, 1163, and 835 cm −1 (i.e., the broadbands were similar to their discharged states).These features can be attributed to the formation of an irreversible SEI film on the surface of the g-C 3 N 4 electrodes. [25,59]These results strongly indicate that the heptazine ring systems (i.e., C-N(-C)-C and/or C-NH-C connected motifs) and secondary/primary amine functionalities (N-H and/or NH 2 groups) of the polymeric g-C 3 N 4 nanosheets act as redox-active sites and participate in the highly stable reversible electrochemical redox reaction with Li + insertion.Similar results were found in a previous report by Li et al. [25] The ex situ FT-IR spectra of the pristine CN-PI 1 electrode and different states of the CN-PI 1 electrodes also display similar tendencies to those of the g-C 3 N 4 electrodes (Figure S26c,d, Supporting Information).In addition to that, the two distinct broad peaks detected in the pristine CN-PI 1 electrode were located at 1676 and 1630 cm −1 , corresponding to the asymmetric and symmetric stretching frequencies of the C=O groups present in the melon-PTCDI segments, respectively, which were gradually more intense when compared with the g-C 3 N 4 electrode and red-shifted to become a strong broadened band at 1632 cm −1 during discharge to 1.28, 0.35, and 0.002 V.This strong broadened band was obviously weakened at 1619 cm −1 during recharging to 2.08 and 2.8 V.These results were attributed to the asymmetric and symmetric stretching vibration characteristics of the C-O-Li bonds due to the formation of lithium enolate species with lithium during the discharge process.A new broadband observed at 733 cm −1 was ascribed to the symmetric in-plane bending vibration mode of the imide C=O bonds of the melon-PTCDI segments in the pristine CN-PI 1 electrode, which was blue-shifted (shifted to lower region) to become a sharp vibrational peak at ∼772 cm −1 during discharge to 1.28 and 0.35 V.When the deep-discharge voltage was Energy Environ.Mater.2023, 6, e12553 decreased to 0.002 V, this sharp vibrational peak gradually intensifies and becomes a strong broadened band along with the ring-sextant outof-plane bending vibration characteristic of heptazine ring systems (located at 835 cm −1 ).This result indicates the formation of new bonds between the imide C=O groups and lithium to form imide C-O-Li bonds.However, the broadened bands of the C-O-Li bonds located at 1619 and 772 cm −1 do not diminish when fully recharging the CN-PI 1 electrode to 2.08 and 2.8 V (Figure S26c, Supporting Information).Moreover, the strong vibrational peaks centered at 1560 and 1313 cm −1 were associated to the stretching vibration modes of the perylene C=C bonds and imide C-N bonds in the pristine CN-PI 1 electrode, respectively, which were gradually intensified and blue-shifted to become a broadened bands at 1517 and 1306 cm −1 after being discharged to 1.28 and 0.35 V, yet these peaks still shift to become a strong broadened bands at 1520 (perylene C=C stretching) and 1304 cm −1 (imide C-N stretching) upon deep-discharge to 0.002 V. Furthermore, the relative reversible recoveries of the perylene C=C bonds and imide C-N bonds were detected in the CN-PI 1 electrodes when recharged to 2.08 and 2.8 V, implying that the unsaturated C 6 aromatic rings of the PTCDI units (perylene core) and imide rings (C-N bonds of the imide functionality) of the melon-PTCDI molecules in the CN-PI 1 electrode also act as redox-active sites and participate in the reversible electrochemical redox processes in the potential range of 3.0-0.01V. [37] These phenomena were strongly justified for the unprecedented and continuously increased discharge capacity of the CN-PI 1 anode, as shown in Figure 5f.
Ex situ HR-XPS studies on the g-C 3 N 4 and CN-PI 1 electrodes were conducted to verify the evidence for the participation of the organic redox-active sites and the occurrence of a high-density Li + adsorption process into the pores/defects (nitrogen triangular holes/C 2 N species), edge sites, and interlayers of the g-C 3 N 4 nanosheets during the galvanostatic discharge-charge process.The typical HR-XPS spectra of C 1s, N 1s, and O 1s signals for the pristine g-C 3 N 4 electrode are similar to those of the graphitic-like g-C 3 N 4 nanosheets (Figure 8a-f; Figure S26e,f, Supporting Information).[62] These characteristic peaks were significantly decreased and shifted to higher binding energies (redshifted) when fully discharged to 0.002 V when compared to that when fully charged voltage of 2.8 V.In particular, the peaks observed at 284.48, 286.88, and 288.48 eV were attributed to the C=C, C=O/ HO-C=O, and N-C=N/C-(N) 3 bonds, respectively, which were remarkably diminished and red-shifted to become a broad peak during the discharge step to 0.002 V and then this peak was intensified when fully charged voltage of 2.8 V, demonstrating the incorporation of Li + into the basal plane of the g-C 3 N 4 nanosheets to produce different forms of intercalated compounds (i.e., LiC 6 and LiC y N x species) [17] and the formation of C-O-Li bonds during the discharge process, and then followed by the recovery of the original state of these peak features during being charged to 2.8 V. Figure 8b shows the HR-XPS N1s spectrum obtained for the pristine g-C 3 N 4 electrode can be resolved into four dominant peaks at 398.48, 399.58, 400.78, and 404.18 eV, corresponding to the aromatic pyridinic-N, pyrrolic-N, graphitic-N and C-NH 2 /pyridinic−N + -O x − groups, respectively. [28,60]In addition, the co-existence of a high C=N-C/C-N-H content and an appropriate amount of N-(C) 3 is capable of generating larger numbers of Li + storage sites to achieve a high density of Li + adsorption in the g-C 3 N 4 electrode during the electrochemical redox process, as described beforehand.[65][66] When fully discharged to 0.002 V, the peak corresponding to the pyridinic-N bonds diminished and red-shifted to 399.58 eV, while the graphitic-N peak was not distinguished from the pyrrolic-N peak, which was remarkably reduced and red-shifted to become a broad peak at 401.58 eV.The under-stoichiometric value of the pyridinic-N/graphitic-N peak (i.e., pyridinic-N/(graphitic-N + pyrrolic-N peak)) clearly confirms the existence of the pyrrolic-N peak combined with the graphitic-N peak to become a broadened peak of N-(C) 3 + C-N-H located at 401.58 eV. [67]These results demonstrate that the higher density of Li + adsorbed on the center of the pyridinic-N and pyrrolic-N hole defects to form a high density of -N-Li bonds and the nitrogen triangular holes in the g-C 3 N 4 nanosheets. [17,60]These peak features were significantly intensified to regain their original state and shift to lower binding energies by 0.7-1 eV during being charged to 2.8 V when compared to discharge voltage of 0.002 V.The HR-XPS O 1s spectra showed the very weak O1s peaks observed at 531.48 and 532.98 eV can be attributed to the surface hydroxyl groups (i.e., N-C-O bond or N-C-OH groups) and adsorbed H 2 O or CO 2 on the surface of g-C 3 N 4 samples, respectively (Figure S26e, Supporting Information). [46,68]The O1s peak observed at ∼531.48 eV for the N-C-O bond corresponds to the deconvoluted C=O/HO-C=O species observed in the C 1s spectra of the pristine g-C 3 N 4 electrode (Figure 8a). [63]When the voltage was fully discharged to 0.002 V, the strong peak detected at 531.78 eV was attributed to the formation of N-C-O-Li bonds and a weaker peak at 533.38 eV could also be assigned to the formation of N-C-O-Li bonds (Figure S26e, Supporting Information).These peaks originate from the reduction reaction of their respective N-C-O bonds and H 2 O or CO 2 molecules with lithium.The ratio of these peaks seemingly increases and shifts to lower binding energies (0.1 eV) during the discharge voltage of 0.002 V when compared to that fully charged at 2.8 V.This was probably due to the fact that the N-C-O/N-C-OH groups gain electrons from lithium to form their respective lithium enolate species during the discharge process, which strongly indicates that the N-C-O/N-C-OH groups (surface C-O species) and CO 2 molecules on the surface of the g-C 3 N 4 electrode also contribute to the reversible electrochemical redox processes.Furthermore, the Li 1s binding-energy peaks observed at 55.68 and 56.78 eV correspond to a range of Li + environments in the N-C-O-Li/LiC y N x bonds and LiC 6 species, respectively when discharged to 0.002 V (Figure 8c), demonstrating the coordination bond between the lithium and six nitrogen lone-pair electrons to form a range of Li + environments (i.e., -N-Li binding sites) within the g-C 3 N 4 nanosheets. [23,47]The ratio of the above peaks seemingly decreases and becomes a broadened peak upon charging to 2.8 V.A very weak peak corresponding to Li 2 O at 58.08 eV was observed and mainly originates from the SEI layer. [17]igure 8d-f and Figure S26f, Supporting Information illustrated that the C1s, N1s, O1s, and Li1s binding-energy peaks observed for the pristine and different states of the CN-PI 1 electrodes show very similar tendencies to those of the g-C 3 N 4 electrode in terms of the following aspects: 1) the XPS C1s spectrum of pristine CN-PI 1 shows four dominant peaks at 284.28, 285.58, 287.98, and 290.28 eV, corresponding to the pure graphitic carbon C=C and C-C/C-H bonds, C=O bonds, N-C=N/C-(N) 3 bonds, and π-excitations from the melon-PTCDI type repeating system, respectively (Figure 8d).These peak features were significantly diminished and become broad peaks during discharged to 0.002 when compared to fully charge voltage of 2.8 V, indicating that the Li + were gradually incorporated into the basal plane of the g-C 3 N 4 framework of the CN-PI 1 , the unsaturated C 6 aromatic rings of the perylene core and formation of imide C-O-Li bonds; [48] 2) the XPS N1s spectrum of pristine CN-PI 1 has four dominant peaks at 398.38, 399.28, 399.80, and 400.58 eV, corresponding to pyridinic-N, pyrrolic-N, imide-N atoms in the imide bonds (O=C-N-C=O bonds), and graphitic-N bonds, respectively (Figure 8e).The ratio of the pyridinic-N and graphitic-N + pyrrolic-N peaks (including imide-N peak) was remarkably decreased and blue-shifted to become two broad peaks at 398.28 and 399.68 eV, respectively during the discharged to 0.002 V, yet these peaks were significantly redshifted to 399.38 and 401.38 eV after being fully charged to 2.8 V, implying that the larger volume of Li + was adsorbed energetically on the center of the pyridinic-N/pyrrolic-N hole defects of the g-C 3 N 4 frameworks in the CN-PI 1 during the discharge process; 3) the O1s energy peaks observed at 531.28 and 532.98 eV can be attributed to the O=C-N-C=O bonds present in the melon-PTCDI molecules and surface C-O bonds in the pristine CN-PI 1 electrode, respectively (Figure S26f, Supporting Information).These peak intensities intensified and blue-shifted to become strong broadbands at 531.08 and 532.78 eV during discharge to 0.002 V when compared to charge voltage of 2.8 V, demonstrating the formation of C-O-Li bonds; 4) the Li 1s XPS spectrum obtained for pristine CN-PI 1 anode did not show any Li + species, yet the discharged CN-PI 1 displays two strong energy peaks at 54.58 and 55.38 eV, which were associated with a range of Li + environments containing C-O-Li and LiC y N x bonds and LiC 6 species, respectively (Figure 8f).Therefore, the overall HR-XPS characteristic features of the CN-PI 1 electrode strongly demonstrate that a range of different edge site environments, defects (C 2 N species), and organic redox-active sites (imide C=O/C-N groups and unsaturated C 6 aromatic ring of the perylene core) participated in the reversible electrochemical redox reaction with lithium.These results were further supported by the overwhelmingly superior Li + storage properties of the CN-PI 1 electrode, including its unprecedented, highly stable discharge capacity, ultrahigh-rate capability, and long cycle life (Figures 5 and 7; Figure S22, Supporting Information).

Morphologies of the Cycled g-C 3 N 4 and CN-PI x Composite Anodes
To elucidate the impact of SEI layer formation on all cycled electrodes and the origin of their outstanding Li + storage performance, the g-C 3 N 4 and CN-PI x composite anodes were disassembled after 1000 cycles at 1 A g −1 and the surface morphologies of anodes analyzed using FE-SEM, as shown in Figure 9 and Figure S27, Supporting Information.The pristine g-C 3 N 4 electrode shows that the active material was relatively well distributed with irregularly shaped aggregated nanosheets and particles (Figure 9a).In contrast, the CN-PI x (x = 0.2, 0.5, 0.75, and 1) composite electrodes display abundant nanoparticles and pronounced graphitic-like nanosheets with π-stacked aggregated structures when growing content of PTCDI unit in the g-C 3 N 4 nanosheets, which was well-distributed to form PTCDI-assembled Energy Environ.Mater.2023, 6, e12553 nanosheet g-C 3 N 4 surface morphologies (Figure 9c; Figure S27a,c,e, Supporting Information).The SEM image for the cycled g-C 3 N 4 electrode shows a relatively homogeneous distribution of graphitic-like nanosheets/particles covered with very poor SEI film on the surface of the g-C 3 N 4 electrode (Figure 9b and illustration in Mosaic SEI layer model for g-C 3 N 4 ns).The SEM image of the cycled CN-PI 0.2 electrode has a very weak SEI film consisting of inorganic and organic grains on the electrode surface (Figure S27b, Supporting Information and Mosaic SEI layer model for CN-PI 0.2 ).Whereas, the cycled CN-PI x (x = 0.5, 0.75, and 1) electrodes have pronounced Mosaic-like nanostructures with different aggregate sizes, high porosity, and the formation of a dense/robust Mosaic SEI layer on their corresponding electrode surfaces when compared to that of the cycled CN-PI 0.2 electrode (Figure 9d; Figure S27d,f, Supporting Information).[71][72] Therefore, the formation of a dense/robust SEI layer with more stable inorganic species near the surface of the CN-PI x (x = 0.5, 0.75, and 1) electrodes was attributed to the in situ formation of inorganic decomposition products (i.e., Li 2 O, Li 2 CO 3 , and LiF), which were highly thermodynamically stable against preferential Li + conduction/diffusion.These SEI species have already been verified using XPS studies of the cycled CN-PI x electrodes (Figures 8 and 10).This "inorganic SEI layer" or "lithium passivating layer" may be covered by organic decomposition products (i.e., lithium alkyl carbonates, ROCO 2 Li, and lithium carbonates; Li 2 CO 3 ) during the initial charging/ discharging cycles and then kinetically suppresses continuous electrolyte decomposition by preventing direct contact (electronic contact) between the surface of the CN-PI x electrodes and the liquid electrolyte, as well as solvent co-intercalation.Therefore, these lithium-passivation phenomena should be mainly responsible for the continuously increasing ultrahigh specific capacities and highly reversible Li + storage properties of the CN-PI x electrodes, high-rate capability, exceptional longterm cyclic stability, and impedance electrode-electrolyte interface resistance.These results suggest that the CN-PI x electrodes maintain their structural integrity and a high degree of crystallinity with compact Mosaic-like nanostructures and different aggregate sizes during the fast reversible redox reactions, as illustrated in the Mosaic SEI layer model for the corresponding electrodes.In particular, the outstanding performance of CN-PI 1 was attributed to the homogeneous electrode morphology and very thick SEI film with well-defined Mosaic-like nanostructures with different dimensions (favorable structural topology) and the high degree of crystallinity when compared to that observed for the cycled g-C 3 N 4 and other CN-PI x (x = 0.2, 0.5, and 0.75) electrodes, as clearly identified in the SEM images and Mosaic SEI layer models (Figure 9d).Therefore, these unique characteristic features allow for much shorter Li + diffusion pathways that give rise to a short electron/charge hopping distance between a large amount of redoxactive aromatic C=C bonds and C=O groups present in the PTCDI units (high degree of redox activity) and the existence of a high level of pyridinic-N/pyrrolic-N (C 2 N species) with a reduced ratio of graphitic-N located at the edge sites and defects in the framework of the CN-PI 1 anode. [28,40,68]Consequently, these characteristic features allow CN-PI 1 to have an overwhelmingly superior Li + storage capability after cycling due to the greater amount of CN-PI 1 active material accessed after the longstanding anode activation process through the increased electrolytes penetration amount into the framework of the CN-PI 1 anodes. [27,70]In contrast, the poor performance of the g-C 3 N 4 electrode mainly originates from a relatively unfavorable electrode morphology with very poor Mosaic SEI layer (Figure 9b), low crystallinity, and poor electronic conductivity due to the absence of the multiple aromatic rings of the PTCDI units within the g-C 3 N 4 anode compared with the highperformance CN-PI x anodes.These results show good agreement with the seemingly low Li + storage performance of g-C 3 N 4 nanosheets in comparison with the CN-PI x anodes, as shown in the GDC curves.
To further support and convince the rationalization of the ultrahigh capacity of CN-PI 1 (∼8400 mAh g −1 ) exceeding the theoretical limitation (Figure 6a), we have strongly justified with previous report by Chen et al. [59] on the high reversible capacity up to 5100 mAh g −1 versus Na/Na + for very thin g-C 3 N 4 film and further our ex situ analyses of the ultrahigh-capacity CN-PI 1 electrode after long-term cycling periods of >420 days (Figure 10; Figures S28 and S29, Supporting Information).The initial stage of research for the deposition (underpotential) and dissolution of lithium salts in aprotic solvents has been focused upon surface films generated on the copper (Cu) foil.During the electrochemical process, there are two types of phenomenon, the potentiostatic process by polarizing the electrode to low potentials (surface film formation in the 1.5-0.0V range and Li deposition below 0.0 V vs Li/Li + ) or galvanostatic process (Li metal deposition at different current densities).Upon the polarization to low potentials, the electrodes were covered with porous and rough surface films on their solution side, in which Li metal deposition in LiPF 6 solutions was found to be more uniform.This study concluded that the Li deposition on copper was much more homogeneous and uniform surface films with major LiF surface species. [73]Recently, Chen et al. [59] showed that the effect and electrochemical performance of ultrathin 2D g-C 3 N 4 films with the thickness of 10 and 150 nm deposited on a Cu metal electrode by a chemical vapor deposition method, which was evaluated directly as a flexible electrode for sodium-ion batteries.Among them, the 10 nm thick g-C 3 N 4 film exhibited a high reversible and an extraordinarily high capacity of ∼51 Ah g −1 (5100 mAh g −1 ) at an areal current density of 0.013 mA cm −2 , which was 40 times higher than metallic sodium (1.16 Ah g −1 ). [74]This result demonstrated that the g-C 3 N 4 -Cu heterojunctions effects were responsible for an unusually high capacity to store sodium metal at an underpotential of 1.2 V or lower potentials.The proposed mechanism of Na + storage in g-C 3 N 4 film demonstrates that the Na + is reduced and deposited onto the Cu metal by underpotential sodium deposition, thus forming ∼500 nm thick metallic Na layer below the g-C 3 N 4 film.In the charging process, g-C 3 N 4 films also activated the Na metal for reversible Na deposition on the Cu surface after penetrating from g-C 3 N 4 films by establishing electroneutrality.After the charging process, the g-C 3 N 4 film was moved up as an SEI layer on top of the growing Na 0 metallic layer.Thus, the moved g-C 3 N 4 film with the SEI layer could avoid the direct contact of the Na 0 metallic phase with the liquid electrolyte. [59]hese similar tendencies and charge storage mechanisms were also observed in CN-PI 1 anode coated on a Cu foil, which exhibited an extraordinary ultrahigh capacity to store lithium metal (Li 0 deposition) at an underpotential of 1.2 V or lower potential.The ex situ analyses of ultrahigh-capacity CN-PI 1 cell (after long-term charged CN-PI 1 electrode) were strongly supported the proposed lithium storage mechanism as follows: 1) The CN-PI 1 layer formed in semiconductor-metal heterojunctions at HOMO position or work function of +1.5 V (Figure 9e), which is comparable to Cu with +0.34 V.This is related to the very first contact zone (Schottky zone/layer) for an electron depletion in Cu of about 1.2 V (1.16 V). [59,75] This phenomenon is further supported by both ex situ FT-IR (Figure S28a,b, Supporting Information) and Cu 2p XPS results (Figure 10e).2) The work function of ∼1.2 V could thus be responsible for the underpotential lithium deposition on Cu foil, in which Li is preferentially inserted in the Schottky zone/layer to give back its electrons to Cu.In another word, Li 0 deposition (Li 0 adlayers) can form at lower potentials in the Schottky zone/ layer, while the zone/layer of the Cu gained the energy by the electrons back-transfer process (as illustrated in Figure 9f). [59,73]In the charging process, Li 0 is deposited on the top of the Cu surface after Li + permeation through the CN-PI 1 film by establishing electroneutrality phenomena, thereby the CN-PI 1 film is moved up as an SEI layer on top of the growing Li 0 adlayers (which acts as Li 0 anode).In this stage, the thickness of Li 0 adlayers is initially found to be a significant amount in Energy Environ.Mater.2023, 6, e12553 the controlled area, which could be responsible for realizing such an extraordinary charge uptake initially and then continuously increasing high capacity in CN-PI 1 anode.During long-term charging, the thickness of Li 0 adlayers or Cu-Li compounds increased for the thicker film even up to about 3 μm, as strongly evident from the cross-sectional ex situ SEM images (Figure 10i), which indicated that the effects of volume of the CN-PI 1 film on the Cu foil and the electric resistance of the CN-PI 1 film (Figure S28e,f, Supporting Information) are crucial factors for underpotential Li 0 deposition in CN-PI 1 anode.At the same time, the Mosaic-like SEI layer is also generated on the top of the CN-PI 1 film during the long-term charge-discharge process (Figure 10g), as already discussed above.3) After the charging state, the electron back-transfer process saturated significantly and became less and less with increasing Li 0 adlayer thickness, in which the effect of semiconductor-metal (CN-PI 1 -Cu) heterojunction has disappeared, as illustrated in Figure 9g.In this stage, Li 0 adlayers act as anode material, and the moved CN-PI 1 film works as SEI layer to avoid the direct contact of the Li 0 metallic phase with the liquid electrolyte.Meanwhile, the significant amount of Cu-Li compounds and inorganic/organic lithium salts is also deposited in a very controlled area at the bottom of the Cu foil with a thickness of 1-2 μm under the formation of SEI layer, which is attributed to the penetration of Li 0 adlayers or Cu-Li compounds from the top of the growing Li 0 deposition of Cu surface and then formed as Cu-Li compounds and inorganic/organic lithium salts at the bottom of Cu foil, as strongly evident from SEM images for the backside of the Cu foil after charged CN-PI 1 (Figure 10h) and XPS survey spectrum (Figure S29c, Supporting Information).Moreover, this result also concluded that Li 0 is obviously deposited in a very controlled area of the CN-PI 1 -Cu heterojunctions in a very special electronic zone in the interface layer, which furthermore generates Cu-Li compounds and inorganic/organic lithium salts locally in a very controlled manner at the bottom of the Cu surface of the CN-PI 1 electrode under the SEI layer during longterm charging-discharging process.This result indicates the permeation of Li 0 adlayer or Cu-Li compounds across the Cu foil and deposition of Li (Li salts) on the Cu backside of the charged CN-PI 1 electrode, as strongly justified by the cross-sectional SEM images (Figure 10i) and XPS results (Figure S29c, Supporting Information).This phenomenon can exclude the formation of Li-dendrites or uncontrolled Li 0 metal deposition for thermodynamic phenomena.Therefore, these Li 0 adlayers or Cu-Li compounds and inorganic/organic lithium salts of the Cu surface should also participate in an extraordinary electron/charge uptake for the overall chemical capacity of CN-PI 1 and thereby resulting in continuously increasing ultrahigh capacity of CN-PI 1 cell with long-term cycling periods of >420 days.Furthermore, these results demonstrate that the formation of Li 0 adlayers or Cu-Li compounds and inorganic/organic lithium salts thickness depends on the low thickness of the CN-PI 1 film coated on the Cu foil (loaded active mass of 0.12 mg cm −2 for CN-PI 1 ), which reflected the generation of maximal number electron/charge and physical transfer of electrons into the CN-PI 1 film due to abundant surface area with porosity and high-defect density of CN-PI 1 structures.It is also worth noting that the coulomb/ extra charge in CN-PI 1 is probably compensated by a Li + species in the pores (CN-PI 1 channels) by establishing electroneutrality via one electron-transfer/charge-transfer per 2 heptazine units of CN-PI 1 (Figure 9f), as strongly evident from XPS Li 1s spectrum (Figure 10f). [75]hese ex situ analyses for the charged CN-PI 1 electrode were also strongly supported and show consistency with the proposed mechanism as previously reported by Chen et al. [59] From FT-IR and XRD results, the remarkably observed difference between before and after charged CN-PI 1 electrodes justified the electron transfer from Cu to CN-PI 1 film, while the characteristic stretching frequencies of C=N and C-N heterocycles from 1200 to 1600 cm −1 and the characteristic breathing vibration mode of the tri-s-triazine ring as well as the symmetric in-plane bending of the imide C=O bonds from the PTCDI units of CN-PI 1 electrode at ∼809 cm −1 for the charged CN-PI 1 electrode.This result indicated that the presence of pyridinic-N/pyrrolic-N "hole" defects and different chemical environments have participated in an energetically more favorable Li + adsorption process and effective electrochemical redox process in CN-PI 1 anode, as strongly revealed in ex situ FT-IR results (Figure S28a,b, Supporting Information).The ex situ XRD result revealed the high structural integrity of CN-PI 1 electrode with significantly reduced d-spacing after long-term cycled CN-PI 1 electrode even >420 days (Figure S28c,d, Supporting Information).
Furthermore, the state of the chemical species for both Li 0 adlayer or Cu-Li compounds and inorganic/organic lithium salts were also evaluated by ex situ XPS analysis (Figure 10a-f), which further gives strong evidence to the proposed mechanism of lithium storage in the local scale for the extraordinary ultrahigh-capacity CN-PI 1 anode.Firstly, the XPS spectra of C 1s and N 1s signals for CN-PI 1 film on Cu foil in the before and after charging are strongly confirmed the generation of the CN-PI 1 -Cu heterojunction. [59,75]The XPS spectrum of C1s signal for CN-PI 1 film before charging can be resolved using three major peaks with binding energies of 288.0 eV (N-C=N/C-(N) 3 ), 285.7 eV (C=O), and 284.3 eV (C=C and C-C/C-H), which are typical peaks of CN-PI 1 electrode.After charging, CN-PI 1 showed that the high energy peak (N-C=N/C-(N) 3 ) located at 288.0 eV was obviously decreased to 287.7 eV with reduced peak intensities (Figure 10a).This result indicated that the electrons transferred from Cu surface to N-C=N/C-(N) 3 groups of CN-PI 1 film in the CN-PI 1 -Cu heterojunction.The N 1s spectrum of CN-PI 1 before charging has five peaks at 398.5 (Pyridinic-N), 399.5 (pyrrolic-N), 399.7 (imide-N), 400.6 (graphitic-N), and 404.2 eV (C-NH 2 /pyridinic−N + -O x − ), which are typically five major nitrogen configuration peaks of CN-PI 1 electrode.After charging, these five major nitrogen configuration peaks were obviously decreased with a reduction in their peak intensities and then shifted at 398.2, 397.2, 398.3, 399.3, and 403.5 eV, respectively (Figure 10b).This result also quantifies the electron transfer from Cu foil to the most N-dopant active centers of CN-PI 1 film, [71] which is also obviously seen in Cu + /Cu species spectrum in Cu 2p XPS spectra (Figure 10e).The O 1s spectrum of CN-PI 1 before charging has only two peaks at 531.3 eV (O=C-N-C=O) and 533 eV (C-O groups).After charging, these peaks are intensified and shifted to 531.08 and 532.6 eV, respectively (Figure 10c), indicating the gaining of electrons from Cu foil and transfer to the C=O groups of CN-PI 1 and then associated with Li + ions via enolation to form N-C-O-Li bonds.These phenomena strongly evidenced that CN-PI 1 layer moved up with the growing Li 0 deposition (Li 0 adlayers) and the effect of CN-PI 1 -Cu heterojunction has disappeared, as strongly evident from SEM images (Figure 10i).F1s spectrum of pristine CN-PI 1 electrode has two peaks at 687.3 and 689.4 eV, which corresponded to Li x PO y F z and LiPF 6 , respectively (Figure 10d).After charging, these peak intensities were diminished and shifted at 686.3 and 688.8 eV, respectively.The new peak for LiF appeared at 684.0 eV.The LiF and Li x PO y F z peaks originated from the decomposition of the LiPF 6 to form more stable LiF species. [73]This result indicates the generation of inorganic lithium salts/ organic lithium salts and Li 0 deposition on the surface of the CN-PI 1 film and the backside of Cu surface under the SEI layer.The XPS spectrum of pristine Cu foil (before loading) is similar to a Cu 2+ species (Figure 10e), which is attributed to the partial surface oxidation and it has to reflect the 1.5 V heterojunction to the CN-PI 1 film. [59,75]After loading CN-PI 1 material on Cu foil, the Cu is more like a Cu + species with electron-poor.In this case, the electron-poor Cu stabilized and balanced the missing charges, which will be taken up by the interface electrons by the incorporation of Li + within the CN-PI 1 film (Figure 10e).There is the absence of Cu + and Cu 2+ species for pristine CN-PI 1 electrode before charging, indicating the appearance of the effective CN-Energy Environ.Mater.2023, 6, e12553 PI 1 -Cu heterojunction.Moreover, the existence of coulomb/extra charge in CN-PI 1 is probably compensated by a Li + species in the pores of CN-PI 1 channels by establishing electroneutrality via one electrontransfer (charge-transfer) per 2 heptazine units (or N-dopant active centers) of CN-PI 1 , as strongly indicated from XPS Li 1s spectrum (Figure 10f).A similar tendency of XPS result for the charged CN-PI 1 electrode is also observed in the backside of the Cu surface for charged CN-PI 1 electrode (Figure S29c, Supporting Information).The higher content of the average atomic percentage of organic lithium salts and a trace amount of Cu-Li compounds were formed on the backside of the Cu surface for the charged CN-PI 1 electrode through the permeating Li 0 adlayer or Cu-Li compounds across Cu surface (as strongly evident from Figure 10i), thus forming inorganic/organic lithium salts with a thickness of about 1-2 μm under the formation of SEI layer (Figure S29a-c, Supporting Information).The ex situ TEM images also confirmed that the surface of the long-term cycled CN-PI 1 anode covered with distinct nature of Mosaic SEI layer (Figure S30a, Supporting Information).Moreover, long-term cycled CN-PI 1 electrode illustrated that the graphitic-like layer domains were disorderedly stacked together and naturally crooked to form both amorphous and turbostratic CN-PI matrix shown in Figure S30a, Supporting Information, which led to the formation of large number of closed pores in the CN-PI 1 electrode that should be probably reason for obeying Li + storage by "intercalation/defect adsorption-closed pore filling" mechanism. [52]This result was further strongly justified by amorphous characteristics with reduced diffraction (002) peak (reduced d-spacing of d (002) = 0.2932 Å) for long-term cycled CN-PI 1 electrode shown in Figure S28c,d, Supporting Information.Both amorphous and turbostratic characteristics with reduced d-spacing of CN-PI 1 matrix were well corroborated with previous investigation by Cheng et al. [52] The SAED pattern indicated that the inorganic compounds, such as Li 2 O(111), LiF (002), and Li 2 CO 3 (311), turbostratic carbon nitride(002), and amorphous CN-PI matrix were identified from the surface structure of cycled CN-PI 1 (Figure S30b, Supporting Information). [71]Furthermore, the STEM-EDX elemental mapping for the CN-PI 1 electrode confirmed the presence of inorganic and organic lithium salts (bottom images in Figure S30, Supporting Information), which was further justified with ex situ XPS results.Therefore, the formation of these Li 0 adlayers or Cu-Li compounds and inorganic/organic lithium salts of the Cu surface could also participate in an extraordinary electron/charge uptake for the overall chemical capacity of CN-PI 1 and thereby resulting in an extraordinary ultrahigh capacity of CN-PI 1 cell with longterm cycling periods of >420 days.This similar capacity tendency of the continuously increasing unusual high capacity was also observed for the earlier report by Seferos et al. on a ladder-type and high crystalline perylene diimide-based microporous polymer (∼783 mAh g −1 after 1000 cycles). [70]Therefore, the strong justification with previous report by Chen et al. [59] and our ex situ analyses for the long-term cycled CN-PI 1 electrode demonstrated that the continuously increasing the capacity of CN-PI 1 electrode has not only originated from its intrinsic property of CN-PI 1 electrode but also the coexistence of the bulk Li metal layer deposition at the Cu-CN-PI 1 heterojunction as well as at the bottom of the Cu foil.In other words, the Li-ions are stored in the form of a lump in between Cu-CN-PI 1 heterojunction as well as at the bottom of the Cu foil through Li permeation via the porosity of CN-PI 1 film during the long-term charging periods of >420 days.Hence, these phenomena and the ex situ analyses strongly evidenced that the Li is not only stored in the CN-PI 1 electrode, but also CN-PI 1 activated to Li 0 metal permeation through controlled porosity of CN-PI 1 anode for reversible Li 0 deposition (Li 0 adlayers) at the both inside of Cu-CN-PI 1 heterojunction and surface of Cu foil, which have greatly contributed to exhibit continuously increasing the overall chemical capacities exceeding 8400 mAh g −1 for CN-PI 1 anode.Furthermore, the full cells were fabricated using the extraordinary capacity-growing CN-PI 1 as the anode with commercial LiCoO 2 as the cathode in 1 M LiPF 6 electrolyte.The CV curve of CN-PI 1 in full cell shows three pairs of redox peaks at 1.83/2.2,1.09/1.75,and 0.45/0.81V in the initial cycle (Figure S31a, Supporting Information), which can be attributed to the Li + insertion/de-insertion process in a large number of pores/defects and interlayers of the g-C 3 N 4 frameworks of CN-PI 1 as well as C=C/ C=O functional groups of PTCDI units in CN-PI 1 .These redox peaks area increases as scan cycle increases and became broadened redox peaks at 2.55/2.1 and 0.88/0.74V, suggesting the superior cycling reversibility of CN-PI 1 in full cell.It can be seen that the CV curves of CN-PI 1 full cell are quite different from the CV profiles of CN-PI 1 halfcell (Figure 3b), which represented the fast reaction kinetics in CN-PI 1 anodes due to synergistic effect of g-C 3 N 4 frameworks and PTCDI units present in CN-PI 1 anode.The CN-PI 1 full cell achieves an initial charge/discharge capacity of 1065/469 mAh g −1 and then decreases up to 420/377 mAh g −1 over 25 cycles at a current density of 100 mA g −1 (Figure S31b, Supporting Information).For the longterm cyclic performance at 1 A g −1 , the CN-PI 1 full cell delivers an initial discharge capacity of 235 mAh g −1 and gradually increases up to 1037 mAh g −1 over 50 cycles (Figure S31c, Supporting Information).After 50 cycles, the discharge capacity decreases up to ∼587 mAh g −1 after 250 cycles with a CE nearly 99%, indicating the attractive longduration cyclic property of CN-PI 1 full LIBs.Such an increase in the capacity of CN-PI 1 full cell for the initial 50 cycles can be explained by the initial kinetic activation process of CN-PI 1 full cell during deep cycling and also the pseudocapacitive effect contribution for the overall capacity of CN-PI 1 full LIBs.Furthermore, the Nyquist plots of CN-PI 1 full cell shows that the charge-transfer resistance rapidly diminishes for cycled CN-PI 1 full cell (R ct ≈ 74 Ω) as compared to that of fresh cell (R ct ≈ 123 Ω), as shown in Figure S31d, Supporting Information.These results strongly evidenced that the LiCoO 2 //CN-PI 1 full cell exhibited the superior performance in terms of reversible capacity, long-duration cyclic property with high CE, demonstrating the potential application of the CN-PI 1 as an anode material for LIBs.

Conclusions
In summary, we successfully developed a series of CN-PI x (x = 0.2, 0.5, 0.75, and 1) composite materials with an unusually high surface nitrogen content (23.19-39.92at.% and 44.16-52.94wt.%) via controllable low-temperature engineering from the high-defect density of nitrogen-rich g-C 3 N 4 nanosheets using a single-step and cost-effective thermal polycondensation reaction with varying PTCDA amounts of 0.2, 0.5, 0.75, and 1 wt.%.The as-prepared CN-PI x (x = 0.2, 0.5, 0.75, and 1) materials as anode for LIBs have the cooperation of the two redox-active components that generate a synergistically enhanced high-rate energy storage performance with superior rate capability and remarkable cycling durability in Li + storage, which was much better than their counterparts.Most importantly, the higher PTCDA content in the CN-P1 1 composite anode delivers an unprecedented and continuously increasing ultrahigh discharging capacity exceeding 8400 mAh g −1 (∼1.96 mWh cm −2 ) even after 225 cycles at a current density of 100 mA g −1 (cycling period >420 days) with a Energy Environ.Mater.2023, 6, e12553 continuously stabilized CE of 96%.The CN-PI 1 anode delivered high specific energy density (E sp ) of ∼7700 Wh kg −1 and the volumetric energy density (E v ) of ∼14 956 Wh L −1 compared to those of other CN-PI x (x = 0.2, 0.5, 0.75, and 1) anode materials and also commercial graphite.The CN-PI 1 anode also displayed promising high-rate long-term cycling stability.A full cell, consisting of CN-PI 1 anode and LiCoO 2 as cathode, realized the attractive long-duration cyclic property (∼587 mAh g −1 at a 1 A g −1 after 250 cycles with a CE nearly 99%).As a result, the extensive range of high surface nitrogen content, particularly the highest pyridinic-N/pyrrolic-N contents, extraordinary porosity with high specific surface area, improved intrinsic conductivity channels, and the existence of pyridinic-N/pyrrolic-N "hole" defects and different chemical environments present in the melon-PTCDI molecules of CN-PI x composite materials were significantly contributed to the overwhelmingly superior lithium storage features by "intercalation/defect adsorption-closed pore filling" mechanism.Furthermore, the formation of Li 0 deposition layer (Li 0 adlayers) on the Cu surface between CN-PI 1 -Cu heterojunction at underpotential deposition and the formation of higher content of inorganic/organic lithium salts on the backside of the Cu surface have contributed to overall the chemical capacity of CN-PI x anodes.Hence, these findings concluded that both Li 0 adlayers or Cu-Li compounds and inorganic/organic lithium salts also participated in the overall chemical capacity of CN-PI 1 electrode by taking up the interface electrons, which enabled to have an extraordinary ultrahigh capacity of CN-PI 1 exceeding 8400 mAh g −1 .Therefore, our work has revealed a ground-breaking method for the nitrogen-deficient design of g-C 3 N 4 -based composite materials to realize outstanding and long-lasting electrochemical energy storage devices.

Experimental Section
Preparation of nitrogen-rich graphitic carbon nitride (g-C 3 N 4 ): The g-C 3 N 4 sample was synthesized via the thermal condensation of a nitrogen-rich urea precursor at 600 °C for 9 h at a heating rate of 1 °C min −1 in a muffle furnace.An alumina crucible loaded with urea granules (20 g, Alfa Aesar, >99%) was placed in the center of the muffle furnace.The thermal condensation reaction was performed in air, while the crucible was covered with aluminum foil to maintain the ambient pressure.Pale yellow polymeric bulk g-C 3 N 4 materials were isolated.We repeated the same thermal condensation reaction for the bulk g-C 3 N 4 materials and then isolated 0.6 g of the ultra-thin g-C 3 N 4 nanosheets.Anal.calcd for g-C 3 N 4 nanosheets: The Calcd for C x N y (C 30 H 8 N 44 ) found: C 35.58, H 1.47, N 60.73, O 2.07 (surface oxygen functional groups detected in g-C 3 N 4 ).
Preparation of g-C 3 N 4 -perylene polyimide composites (CN-PI x ): To modify the surface of nitrogen-rich graphitic carbon nitride (g-C 3 N 4 ) with PTCDA content (CN-PI x , where x represents the weight ratio of PTCDA, x = 0.2, 0.5, 0.75, and 1 g), PTCDA was chosen to react with the as-prepared high-defect density, nitrogen-rich/-NH 2 edge groups of g-C 3 N 4 nanosheets via a polycondensation reaction, as illustrated in Scheme 1.Briefly, PTCDA unit (0.14 g) was chosen to react with prepared nitrogen-rich g-C 3 N 4 nanosheets (0.7 g) yielding 0.72 g of CN-PI 0.2 composite sample.In a typical synthetic procedure, PTCDA (0.14 g), g-C 3 N 4 nanosheets (0.7 g), and imidazole (10 g) were placed in a 100 mL threeneck round-bottom (RB) flask equipped with a reflux condenser.This mixture was slowly heated to 140-150 °C under vigorous stirring and maintained at this temperature for 12 h under a nitrogen atmosphere.When the reaction mixture was cooled to 50 °C, 50 mL of ethanol was added, and the reaction mixture was stirred for another 12 h.The solution was then transferred to a 250 mL roundbottom flask containing 150 mL of 2 M HCl.After stirring at room temperature for 12 h, the precipitate was filtered under vacuum and washed thoroughly with methanol and then deionized water several times.The resulting reddish-pink solid was transferred to a 100 mL RB flask containing 50 mL of 10% potassium carbonate solution (aqueous K 2 CO 3 ), which was refluxed in an oil bath at 100 °C for 1 h.After cooling to 50 °C, the resulting precipitate was filtered under vacuum, washed three times with 300 mL of 10% K 2 CO 3 solution and 50 mL of 2 M HCl solution.The pink precipitate obtained was washed thoroughly with methanol and deionized water until the pH of the filtrate was neutral.The collected solids were dried under vacuum at 80 °C for 12 h to afford 0.72 g of a pink powder, which was marked as CN-PI 0.2 .Similarly, the different compositions of PTCDA (0.5, 0.75, and 1 g) were chosen to react with the nitrogen-rich g-C 3 N 4 nanosheets (weight ratio of CN is taken as one; 1 g) to prepare the CN-PI 0.5 , CN-PI 0.75 , and CN-PI 1 composites, respectively.Therefore, "x" (PTCDA content) has been optimized in CN-PI x by taking the weight ratio of PTCDA/CN (g/g): 0.2/1, 0.5/1, .75/1and 1/1 for CN-PI 0.2 , CN-PI 0.5 , CN-PI 0.75 , and CN-PI 1 composites, respectively, in which the weight ratio of CN is taken as 1 g.As-prepared g-C 3 N 4 nanosheets possess a high density of -NH 2 on the edges due to the higher content of nitrogen wt.% (Tables S1 and S2, Supporting Information).Consequently, PTCDIs were assembled on the edges of g-C 3 N 4 through the reaction between PTCDA and -NH 2 edge groups on the g-C 3 N 4 .Hence, it should be noticed that the availability of edge groups on the g-C 3 N 4 , the number of PTCDIs assembling on the edge of g-C 3 N 4 , multi-step purification process with batch-tobatch variation and solubility of starting materials would greatly affect the overall yield of composites.Hence, these parameters could be probably reason for observing the variation in the yield of composites in the range of 0. Characterization: All reagents, chemicals, organic solvents, and electrolytes were purchased from Sigma-Aldrich, Alfa Aesar, and Tokyo Chemical Industry Co., Ltd. and were used as received.Fourier transform infrared (FT-IR) spectra of the prepared composites were recorded on a Smart iTR NICOLET iS10 FT-IR spectrometer (Thermo Scientific).Thermogravimetric analysis (TGA) was performed using a TGA Q50 thermal analyzer at a heating rate of 10 °C min −1 from 0 to 900 °C under a nitrogen atmosphere.Elemental analysis of all the prepared g-C 3 N 4 nanosheets and CN-PI x composites was performed using a Thermo Scientific Flash 2000 Elemental (CHNS/O) Analyzer.XRD analysis was used to investigate the crystal structures of the bulk and electrode samples using Cu Kα radiation (operating voltage: 40 kV, current: 30 mA, λ = 1.5418Å).The photoluminescence (PL) spectrum of g-C 3 N 4 nanosheets and CN-PI x composite samples were recorded using Jasco FP-6500 spectrometer in ethanol-dispersed solutions.X-ray photoelectron spectroscopy (XPS) measurements were recorded using K-Alpha (Thermo Scientific) as an X-ray source to analyze the elemental surface compositions of the samples and electrodes.The morphologies of the samples were examined using field-emission scanning electron microscopy (FE-SEM) on an S-4800 (Hitachi) instrument.For the HR-TEM images, high-angle annular darkfield scanning TEM and EDS elemental mapping were acquired on a STEM-HAADF, FEI Tecnai G2 F20 transmission electron microscope equipped with an energy dispersive detector.The g-C 3 N 4 anode and CN-PI x composite materials were evaluated using a micro-Raman spectrophotometer (XploRA; Horiba) using an Ar + laser (532 nm).To identify the specific surface area, pore volume, and average pore size of the prepared samples, the nitrogen adsorption, and desorption isotherms were analyzed using a physisorption ion analyzer (BET, 3-flex; Micromeritics Instruments Corp.) located at the Core Research Support Center for Natural Products and Medical Materials, Yeungnam University.AFM measurements were made with a PSI AFM (XE-100).All the topography images were realized in noncontact mode using a PPP-NCHR (PointProbe ® Plus Non-Contact High-Resolution Frequency-Reflex Coating) silicon probe with tip radius of <10 nm (NanosensorsÔ).System control and data acquisition were performed by XEP software (Park Systems Corp.), and data analysis was done with XEI software (Park Systems Corp.).The galvanostatic intermittent titration technique (GITT) was measured at 50 mA g −1 with a pulse time (Δt) of 15 min and a relaxation time of 60 min.
Energy Environ.Mater.2023, 6, e12553 Computation methods: All plane-wave DFT calculations were performed using the projector augmented wave pseudopotentials [76] provided in the Vienna ab initio simulation package (VASP). [77]The Perdew-Burke-Ernzerhof exchange correlation [78] was used with a plane-wave expansion cutoff of 400 eV for all systems.To model CN-PI x composite structure, we took a part of CN-PI x composite as shown in Figure 6a, because the identical parts are repeated to form complete CN-PI x .The simulation cell was set to 40 Å × 40 Å × 30 Å with single part (single segment) of composite structures to avoid the self-interactions between periodic images.The configuration shown in Figure 6b was relaxed until the forces reached at least 0.03 eV Å−1 with Gamma KPOINTS mesh, 2 × 2 × 4. In terms of Li bulk metal, the same simulation settings were applied but the different the Gamma KPOINTS mesh, 2 × 2 × 2, were used.The relaxed bulk lattice constants of Li were predicted 4.31 Å.
Electrochemical performance measurements: The CN-PI x composite anodes were prepared by mixing the active materials (such as g-C 3 N 4 anode, PTCDA, CN-PI x composite), conductive carbon black (Super P; Sigma-Aldrich), and polymer binder (PVDF) at a weight ratio of 60:30:10 to form a well-dispersed slurry in N-methyl pyrrolidone (NMP) with stirring for 12 h.The slurry was uniformly spread over copper foil as a current collector using the doctor blade technique at a thickness of 20 μm.Subsequently, the coated copper foil was immediately dried in a blast oven at 100 °C for 12 h to produce the electrode films.A twin roller machine was used to press the coated electrodes to increase the contact between the active materials and copper foil.Finally, the electrodes were cut into small circular discs with a diameter of 12 mm.The mass loadings of the anode materials in all of the cells were maintained at 1.7 AE 0.28 mg cm −2 .After the electrodes were weighed, they were placed in an argon-filled glove box with moisture and oxygen/water levels <1 ppm and fabricated into LIBs using a 2032 coin cell assembly.Pure lithium foil was used as the counter electrode, 1 M LiPF 6 in a mixture of ethylene carbonate (EC)/dimethyl carbonate (DMC) (EC/DMC; 1:1 molar ratio) was used as the electrolyte, and a Celgard 2500 membrane was used as the separator.Full cells were also fabricated using capacity-growing CN-PI 1 anode (coated on Cu foil) and commercial LiCoO 2 as the cathode (coated on Al foil) without counter electrode (lithium foil) in the 1 M LiPF 6 electrolyte.The galvanostatic charge-discharge cycling of the fabricated cells was carried out using a Neware multi-channel battery tester (BTS-7.6.0,Shenzhen) in the voltage range of 0.01-3.0V (vs Li + /Li).Cyclic voltammetry (CV) was performed using a WonA-Tech WBCS3000 battery test unit.Electrochemical impedance spectroscopy (EIS) was conducted to estimate the charge transfer resistance of both the fresh and cycled battery cells using a Solartron impedance/gain phase analyzer (model SI 1255) coupled with a potentiostat (SI 1268).The frequency was varied from 100 kHz to 0.1 Hz with an alternating current signal amplitude of 10 mV.

Figure 1 .
Figure 1.High-resolution N 1s XPS spectra obtained for the a) g-C 3 N 4 nanosheets and b) CN-PI 1 samples.c) The relative fractions of the nitrogen functionalities (pyridinic-N, pyrrolic-N, imide-N, graphitic-N, and pyridinic-N + -O x − species) in the g-C 3 N 4 nanosheets and CN-PI x (x = 0.2, 0.5, 0.75, and 1) samples.d) XRD patterns of the g-C 3 N 4 nanosheets and CN-PI x samples.e) Raman spectra of the CN-PI x samples.f) Nitrogen adsorption-desorption isotherms (BET isotherm curves) obtained for the g-C 3 N 4 nanosheets and CN-PI x samples and their corresponding pore-size distributions (inset).

Figure
Figure 1a,b and FigureS4, Supporting Information present the N 1s HR-XPS spectra obtained for the g-C 3 N 4 nanosheets and CN-PI x (x = 0.2, 0.5, 0.75, and 1) composite samples, in which the N 1s region of all samples can be deconvoluted into four major peaks at 398.18, 399.58, 400.58, and 403.8 eV, which correspond to the pyridinic-N, pyrrolic-N, graphitic-N, and pyridinic−N + -O x − species,

Figure 2 .
Figure 2. Scanning electron microscopy images of the a) g-C 3 N 4 nanosheets and b) CN-PI 1 sample.c) Tapping-mode atomic force microscopy image of CN-PI 1 deposited on the FTO substrate with the image size of 3 μm × 3 μm.TEM and HR-TEM images of the g-C 3 N 4 nanosheets and CN-PI 1 sample at different magnifications; d1, d2) g-C 3 N 4 nanosheets and e1, e2) CN-PI 1 samples.d3, e3) The corresponding electron diffraction patterns of the selected areas.d4, e4) STEM-EDX elemental mapping of the g-C 3 N 4 nanosheets and CN-PI 1 sample, respectively.

Figure 3 .
Figure 3. Electrochemical properties of the g-C 3 N 4 nanosheets and CN-PI 1 electrodes in the potential range of 0.01-3.0V for Li + /Li.CV curves obtained for the a) g-C 3 N 4 nanosheets and b) CN-PI 1 electrodes at a scan rate of 0.1 mV s −1 .The CV curves obtained for the c) g-C 3 N 4 nanosheets and d) CN-PI 1 electrodes at different scan rates.The relationship between the square root of the scan rate (v 1/2 ) and peak current (i p ) of the e) g-C 3 N 4 nanosheets and f) CN-PI 1 electrodes.

Figure 4 .
Figure 4.The capacitive-and diffusion-controlled charge storage contribution of the a) g-C 3 N 4 nanosheets at a scan rate of 0.1 mV s −1 , b) g-C 3 N 4 nanosheets at 1.0 mV s −1 , c) CN-PI 1 electrode at 0.1 mV s −1 , and d) CN-PI 1 electrode at 1.0 mV s −1 .The normalized charge contribution ratio of the capacitive and diffusion-controlled capacities observed at different scan rates for the e) g-C 3 N 4 nanosheets and f) CN-PI 1 electrodes.

x composites ( 4 × 4 − 4 − 8 −
PTCDI unit) to form four radical anions ([CN-PI x ] 4Á− ) as intermediate compounds and were then accompanied by the insertion of 4 mol of Li + ions to form electrogenerated CN-PI x Á4Li + compounds.In the subsequent discharge process, another four electrons were captured from the oppositely located C=O groups of each PTCDI unit to form another four radical anions ([CN-PI x ] 4Á− ) species and were then associated with another 4 mol of Li + ions to form CN-PI x Á4Li + compounds.Hence, the total of eight radical anions ([CN-PI x ] 8Á− ) were initially generated and were then accompanied by the totally insertion of 8 mol Li + ions to generate the electro-generated CN-PI x Á8Li + compounds, as shown in the Step 1 of Figure6a.Therefore, a total of eight electron transfer steps occurred in the subsequent step to generate CN-PI x 8−

Figure 6 . 28 ©
Figure 6.a) A schematic representation of the graphitic carbon nitride-perylene polyimide composite anode materials indicating the proposed electrochemical redox reaction mechanism during the lithium-ion insertion/extraction process.b) Top and side views of pristine CN-PI x (single segment) and c) 24 Li-ion adsorbed in a CN-PI x .Color scheme: carbon (gray), hydrogen (white), nitrogen (blue), oxygen (red), and lithium (light purple).

Figure 7 .
Figure 7. a) Rate performance of the g-C 3 N 4 , PTCDA, and CN-PI 1 anodes at different current densities ranging from 0.1 to 20 A g −1 .b) Long-term cycling stability of the g-C 3 N 4 , PTCDA and CN-PI 1 anodes at a current density of 1 A g −1 over 1000 cycles.Nyquist plots obtained for the c) g-C 3 N 4 and d) CN-PI 1 anodes before and after being cycled over 1000 times.The inset shows the corresponding equivalent circuits for the c) g-C 3 N 4 nanosheets and d) CN-PI 1 anode.

Figure 8 .
Figure 8. High-resolution XPS a) C 1s, b) N 1s, c) Li 1s spectra obtained for the different states of the g-C 3 N 4 electrodes.The high-resolution XPS d) C 1s, e) N 1s, and f) Li 1s spectra of the different states of the CN-PI 1 electrodes.

Figure 9 .
Figure 9. Scanning electron microscopy images of a) pristine g-C 3 N 4 and b) the cycled g-C 3 N 4 electrode; c) pristine CN-PI 1 and d) the cycled CN-PI 1 electrode.The schematic illustration of initial Mosaic SEI layer formed on the surface of g-C 3 N 4 anode and CN-PI 1 anodes by inorganic and organic lithium salts (as shown in right side).Schematic illustration of the mechanism of the lithium storage in thin CN-PI 1 film-activated Cu foil: e) CN-PI 1 deposited on Cu foil (current collector) constitutes a semiconductor-metal (CN-PI 1 -Cu) heterojunction, in which electrons will transfer from Cu surface to CN-PI 1 f) In the charging process, Li + penetrate/permeate through thin CN-PI 1 films and deposit onto the Cu metal surface by underpotential lithium deposition, thus forming a Li 0 adlayer or Cu-Li compounds as Li 0 metallic anode, in which the CN-PI 1 film is moved up as an SEI layer on top of the growing Li 0 adlayer or Cu-Li compounds.g) After long-term charging state, Li 0 adlayer or Cu-Li compounds act as the anode material and thin CN-PI 1 film works as stable SEI layer to avoid the direct contact of the Li 0 metallic phase and liquid electrolyte.

Figure 10 . 28 ©
Figure 10.High-resolution XPS spectra of a) C 1s, b) N 1s, c) O 1s, d) F 1s, e) Cu 2p and f) Li 1s signal obtained for the front side of CN-PI 1 electrode before and after cycles.Scanning electron microscopy (SEM) images of g) front side and h) back side of the Cu surface for the charged CN-PI 1 electrode with different magnification.The cross-sectional SEM images of i) the back side of Cu foil for the charged CN-PI 1 electrode.