Preparation of High-Performance Enteromorpha Prolifera–Based Porous Carbons by Nitrogen Modification and Their Electrochemical Performance

This study presents a novel method to improve the electrochemical performance of porous carbons (PCs) by simply adjusting the elemental composition of their precursors with nitrogen-rich materials as additives. Nitrogen-modified enteromorpha prolifera–based (EP-based) PCs are prepared from EP and urea (or melamine). Overall, compared with the control PC without nitrogen modification, their pore structures and surface chemical properties present similar improvement. When used in supercapacitors, their specific capacitances increase by approximately 22% due to their significant development of mesopores at 2.5–7.0 nm, which increases the effective surface areas. With an appropriate amount of nitrogen-containing or oxygen-containing functional groups maintaining surface wettability, the notable increase of graphitized N improves their conductivity. Due to the higher graphitization degrees, their resistances are reduced. With more mesopores transporting ions, they exhibit excellent high-rate capacitive performance. Moreover, they show remarkable long cycle performance with the specific capacitance retention of larger than 92% after 10,000 cycles.


INTRODUCTION
Benefiting from the diversity of composition and structure of biomass, their derived PCs with developed pore structures have high potential as electrode materials for supercapacitors Yang et al., 2019;Zhao et al., 2019;Khan et al., 2020;Qian et al., 2020;Liang et al., 2021). Generally, to prepare high-performance PCs from selected biomass, some measures are taken to optimize their microstructures during the preparation process (Xu et al., 2017;He et al., 2019;Li et al., 2020b;Lin et al., 2020;Luo et al., 2020;Sun et al., 2020;Wu et al., 2020a;Xie et al., 2020a). Although those previous measures can improve the electrochemical properties, they may also show some adverse effects. For example, element doping is an effective method to improve surface property due to the generated surface functional groups, but those functional groups may also cause structural instability, pore blockage, conductivity decline, and other problems; templateassisted methods can be used to optimize pore structures, but they may also show some shortcomings, such as impurity residue, complex synthesis process, limited application scope, etc. Therefore, it is still necessary to further explore more advanced methods to fabricate high-quality PC products.
Nitrogen-rich materials are usually used as nitrogen sources for manufacturing N-doped PCs, but few studies are on their other potential applications (Chen et al., 2017;Zhang et al., 2019a;Li et al., 2020a;Song et al., 2020). By comparing the morphologies and pore structures of charcoals derived from alginic acid and ammonium alginate, our team found that nitrogen in ammonium alginate can promote a large number of thin-walled cavities in the carbonized sample, thus forming greatly developed pore structure in its derived PC. Although the mechanism of nitrogen regulation microstructure is still unclear, it is believed that nitrogen element in precursor can regulate the pore structure of its derived PC. In other words, it is possible to make the pore structure of the obtained PC better by adding nitrogen element into its precursor. Meaningfully, this provides a new application direction for nitrogen-rich substances in preparing PCs.
EP is a kind of natural wild green algae with strong natural reproduction ability. Although it is nontoxic, due to global climate change, water eutrophication, and other reasons, its large-scale growth destroys marine ecosystem (Smetacek and Zingone, 2013;Zhou et al., 2015;Sun et al., 2018;Zhang et al., 2019b;Zhang et al., 2021a). Typically, EP is used to produce lowcost agricultural fertilizers, and animal husbandry feeds, but this is not enough to achieve pollution control. For environmental protection and high-quality utilization of this biomass resource, researchers have tried to apply EP in pharmaceutical, chemical, environmental, and other industries Li et al., 2018;Shao et al., 2019;Li et al., 2020c;Zhu et al., 2020;Zhang et al., 2021b). In recent years, PCs derived from EP for energy storage have attracted considerable attention. However, it still needs to improve their properties, such as pore structure, conductivity, surface properties, etc.
This study presents a facile and effective nitrogen modification strategy (NMS): add nitrogen-rich substances to biomass to change the elemental composition of precursors, and then improve the electrochemical performance of those prepared PCs by optimizing their pore structures and surface chemical properties. Use EP as raw material, urea or melamine as nitrogenrich additive, and synthesize precursors by thoroughly mixing them. Then, PCs are prepared by activating carbonized samples, which are prepared by simply carbonizing precursors. Further, the morphologies, pore structures, and surface chemical properties of these samples are characterized, and their electrochemical properties are tested. The similarities and differences of the effects of the two different nitrogen-rich additives are analyzed by comparing the microstructures and electrochemical performance of the obtained PCs.

Preparation of Nitrogen Modification Porous Carbons
After thoroughly washing with deionized water to remove impurities, EP (Ningbo, Zhejiang, China) was put into a drying oven and heated at 120°C until dry. Then it was pulverized into powder and screened with an 80-mesh sieve. Dry EP powder and urea (or melamine) (Sinopharm Chemical Reagent Co., Ltd., China) were mixed to get homogeneous precursor samples. The mass ratio of EP to urea (or melamine) was 5:1. The precursor composed of EP and urea as EPU-5, the precursor composed of EP and melamine were marked as EPM-5. Besides, EP was selected as the control precursor.
30 g precursor was placed in a horizontal tube furnace and heated to 600°C in a 1-L min −1 nitrogen stream, and then kept at this temperature for 2 h and cooled down to 80°C to obtain a carbonized product. The carbonized product was pickled with 5 mol L −1 HCl (Sinopharm Chemical Reagent Co., Ltd., China) and rinsed thoroughly with deionized water until neutral. Then it was heated at 105°C until dry to obtain a carbonized sample.
Then, based on previous studies by (Wang and Kaskel, 2012;Zhou et al., 2014;Gao et al., 2015;Lee et al., 2016;Li et al., 2017;Lin et al., 2018;Su et al., 2018), KOH was selected as the activator to develop pore structures, and optimized activation conditions were implemented. With a mass ratio of KOH to the carbonized sample of 4:1 and at 105°C, the carbonized sample was immersed in saturated KOH (Tianjin Kermel Chemical Reagent Co., Ltd., China) solution for 12 h. The mixture was placed in a muffle furnace and heated to 800°C in a 0.6-L min −1 nitrogen flow, and then kept at this temperature for 2 h and cooled down to 80°C to obtain an activation product. The activation product was pickled with 5 mol L −1 HCl and then rinsed thoroughly with deionized water until neutral. Then it was heated at 105°C until dry to obtain a PC sample. Throughout the experiment, a fixed heating rate of 5°C min −1 was used. The PC samples derived from EP, EPU-5, and EPM-5 were marked as EP-PC, EPU-PC-5, and EPM-PC-5, respectively. Besides, EP-PC was selected as the control PC.

Material Characterization
The morphologies of samples were observed by a SUPRA55 scanning electron microscope (SEM, Carl Zeiss AG, Germany). After 5 h of vacuum degassing at 300°C, their pore structures were analyzed with a JW-BK132F instrument (Beijing JWGB Sci&Tech Ltd., China) by nitrogen adsorption and desorption at 77.4 K. Their specific surface areas (S BET s) were calculated by the Brunauer-Emmett-Teller (BET) method based on the nitrogen adsorption data within the P/P 0 range of 0.10 and 0.25. The amount of adsorbed nitrogen at P/P 0 0.993 was regarded as their total pore volumes (V total s). Their micropore surface areas (S micro s) and mesopore surface areas (S meso s) were estimated by the t-plot method based on carbon black model. Their micropore volumes (V micro s), average micropore sizes (D micro s), and micropore size distributions were examined by the Horvath-Kawazoe (HK) method based on nitrogen adsorption data and a slit pore model. Their mesopore size distributions and average mesopore sizes (D meso s) were obtained by the Barrett-Joyner-Halenda (BJH) method. Then their mesopore volumes (V meso s) were calculated by the differences between V total s and V micro s. The states of C, N, and O elements in PCs were surveyed by X-ray photoelectron spectroscopy (XPS) with an ESCALAB 250XI (Thermo Fisher Scientific Inc., United States) instrument. Their surface chemical

Electrochemical Characterization
Absolute ethyl alcohol (Sinopharm Chemical Reagent Co., Ltd., China) was used as the solvent, and then the homogeneous gluelike mixture of conducting graphite (TIMREX KS6), polytetrafluoroethylene (D210C, DAIKIN), and PC with a weight ratio of 1:1:8 was prepared. Then it was painted evenly on nickel foam pieces (300 g m −2 , about 1 mm thickness). Then these pieces were pressed at 12 MPa for 1 min, put into a vacuum, and kept at 80°C for 12 h to get dry electrode pieces. PCs' electrochemical performance was measured with an electrochemical workstation (CS310H, Wuhan CorrTest Instruments Corp., Ltd., China) and a two-electrode system. A piece of separator paper (NKK-MPF30AC-100, Nippon Kodoshi Corporation), a pair of cell shells (2016-type), two electrode pieces (radius about 8 mm, thickness about 0.1 mm) with PC loads of about 4 mg, and KOH solution (6 mol L −1 ) were assembled into a cell. These cell materials were purchased from Shanxi Cell Materials Co., Ltd., China.
Galvanostatic charge-discharge (GCD) tests were conducted with a current-density range of 0.1-50 A g −1 , and a potential window of 0-1 V. Cyclic voltammetry (CV) tests were conducted with a scan-rate range of 20-200 mV s −1 , and a potential window of 0-1 V. Electrochemical impedance spectroscopy (EIS) were measured with a frequency range of 10 5 -10 −3 Hz and an AC amplitude of 5 mV. Specific capacitance, discharge current, discharge time, the mass load of the PC in two electrode pieces, potential difference, energy density, and power density were marked as C, I, Δt, m, ΔV, E, and P, respectively. Then, the C of PC was calculated by Eq. 1 based on the data from the discharge branch of GCD curve. And the E and P of a cell were estimated by Eqs. 2, 3, respectively: (3)

RESULTS AND DISCUSSION
As shown in Figure 1, urea or melamine is added to EP to adjust the elemental composition of the precursor. During the carbonization process, EP and nitrogen-rich additive undergo complex physical and chemical reactions. Then due to the interaction of these reactions, the microstructures of the prepared carbonized sample are regulated. Afterward, its derived PC is improved accordingly in the microscopic morphologies, pore structures, and surface chemical properties and further shows excellent electrochemical properties. The carbonization reactions of urea and melamine are different, leading them to play different regulation effects on PCs. The details are as follows.
The SEM images of these prepared PCs are shown in Figure 2. EP-PC, EPU-PC-5, and EPM-PC-5 are, respectively, derived from EP, urea-modified EP, and melamine-modified EP, and they roughly present the structural characteristics of massive cracking, small carbon particle accumulation, and a large number of small slits, respectively. Then it can be concluded that NMS can effectively regulate the microstructure of the prepared PC, and the regulation effects of urea and melamine are different.
The nitrogen adsorption and desorption isotherm curves of EP-PC, EPU-PC-5, and EPM-PC-5 are shown in Figure 3A. When P/P 0 < 0.4, the trends of these three curves demonstrate that adsorption and filling occur in micropores (Yuan et al., 2021). And when P/P 0 > 0.4, the hysteresis loops on these curves demonstrate that capillary condensations occur in mesopores (Hamouda et al., 2021). Hence, all these PCs contain micropores and mesopores. However, the hysteresis loop corresponding to EP-PC is type H3, and those corresponding to EPU-PC-5 and EPM-PC-5 are type H4, indicating that NMS can change the pore architecture of PCs.
Based on the data from Table 1, calculations show that the S BET s of EPU-PC-5 and EPM-PC-5 are 0.98 and 1.01 times that of EP-PC, respectively, indicating that the S BET s of these three PCs are almost equal. The V total s of EPU-PC-5 and EPM-PC-5 are 1.19 and 1.15 times that of EP-PC, respectively, while the V micro s of these three are similar, suggesting that the increase of V total s is mainly caused by the increase of V meso s. Furthermore, these three PCs have equal D micro s, while the relationship of their D meso s is EPU-PC-5 > EPM-PC-5 > EP-PC. Therefore, it can be concluded that the effects of urea or melamine modification on pore structures are generally similar: they have little effects on micropore structures but can significantly promote the development of mesopore structures with almost no change in S BET s. Besides, in comparison, urea modification in promoting the development of mesopores is slightly better than that of melamine modification.
The micropore size distributions of EP-PC, EPU-PC-5, and EPM-PC-5 are shown in Figure 3B. The trends of these three curves are similar, and micropores are mainly distributed in the range of below 0.80 nm with almost the same positions of the extreme points on each curve. In the range of below 0.59 nm, the relationship of the extreme values corresponding to the same extreme point is roughly EPM-PC-5 > EP-PC > EPU-PC-5. However, in the range of 0.59-2.00 nm, the extreme values corresponding to the same extreme point are almost equal. Thus, the variation of these micropore size distributions can suggest that NMS has little effect on micropore size distributions, but only has a particular effect on the distributions of smaller pores (<0.59 nm). Their mesopore size distributions are shown in Figure 3C. On the whole, the trends of these curves are similar, and mesopores are mainly distributed in the range of 2.0-7.0 nm, but they differ in the range of 2.0-3.0 nm. Besides, it is observed that in the range of 2.5-7.0 nm, EPU-PC-5 has more mesopores than EPM-PC-5 and EPM-PC-5 has more mesopores than EP-PC, which also demonstrates that NMS can promote the development of mesopores, and those increased mesopores are primarily located in the range of 2.5-7.0 nm. The Raman spectra of EP-PC, EPU-PC-5, and EPM-PC-5 are shown in Figure 4. The D bands located around 1,349 cm −1 of them are considered as the disordered vibration peaks caused by structural defects, their G bands located around 1,590 cm −1 are caused by the vibration of sp 2 C, and the 2D bands located around 2,871 cm −1 mean that these PCs have good atomic order (Teng et al., 2020;Nguyen et al., 2021). The relative intensity ratios of G band to D band (I G /I D ) can indicate their degrees of graphitization. Accordingly, the relationship of graphitization degree of  them is EPU-PC-5 > EPM-PC-5 > EP-PC. However, their values of I G /I D are very close, suggesting that NMS can slightly improve graphitization degree, and the effect of urea is relatively better than that of melamine. The XPS spectra of EP-PC, EPU-PC-5, and EPM-PC-5 are shown in Figure 5. All these three PCs have C, N, and O elements ( Figures 5A-C). Furthermore, the data from Table 2 shows that the contents of each element in EPU-PC-5 and EPM-PC-5 are almost equal, and they have more C and less N and O than EP-PC, indicating that both of urea and melamine modification can increase the content of C and reduce the contents of N and O. More C may mean stronger conductivity, and less N and O may mean weaker surface wettability. Their C1s spectra can be deconvoluted into four separate peaks ( Figures 5D-F). The relative intensity contents of each peak for these three PCs are similar ( Table 2), indicating that NMS hardly change the existing state of C element. Besides, the relative intensity ratios of C C and C−C are larger than those of the others, implying that these three PCs have good conductivity. The N1s spectra can be deconvoluted into four separate peaks ( Figures 5G-I). EP-PC has more pyridine N and pyrrole N, the relative intensity ratios of these four peaks of EPU-PC-5 are all above 20%, and EPM-PC-5 has more pyrrole N and graphitized N ( Table 2). Generally, it is considered that pyridine N and pyrrole N are helpful to improve surface wettability, and graphitized N can enhance conductivity, and oxidized N is similar to nitrogen-oxygen functional groups (Lin et al., 2018;Wu et al., 2019). Thus, it can be concluded that NMS can significantly increase graphitized N and reduce pyridine N. Besides, it is noticed that the existing states of N element in EPU-PC-5 and EPM-PC-5 are different. The O1s spectra can be deconvoluted into four separate peaks (Figures 5J-L). All these PCs have more C O and C-OH and C-O-C than others ( Table 2), which helps to enhance surface wettability (Lin et al., 2021). Then, it is concluded that both urea and melamine modifications can increase the contents of C element, and the content of each element in these two PCs is similar, but their existing states differ a lot, especially N element.
Further, the electrochemical performance of these obtained PCs is tested on a two-electrode system. The CV curves of EP-PC, EPU-PC-5, and EPM-PC-5 are shown in Figures 6A-C. The rectangular CV curves corresponding to these three PCs indicate that the tested cells present good electric double-layer behaviors, implying that the pseudocapacitances can be ignored. And it is observed that when the scan rate changes from 20 -200 mV s −1 , the relationship of the rectangular shape distortions of these CV curves is EPU-PC-5 > EP-PC > EPM-PC-5, meaning that the relationship of their charge transfer capacities is EPU-PC-5 < EP-PC < EPM-PC-5. Compared with EP-PC, the charge transfer capacity of EPU-PC-5 decreases, which is because although the more mesopores of EPU-PC-5 are beneficial to improve its charge transfer capacity, the weakened effects caused by the decrease of N, O, pyridine N, and pyrrole N are stronger; and the charge transfer capacity of EPM-PC-5 increases, which is because limited by the pore structure, although EP-PC has more surface functional groups, they only have a small effect on improving the charge transfer capacity, while the more mesopores of EPM-PC-5 improve its charge transfer capacity greatly. Besides, it is noticed that although both of EPU-PC-5 and EPM-PC-5 have more mesopores, the effects of those mesopores on improving their charge transfer capacity are different, which may be caused by the different three-dimensional structures of their pores (Figure 2). These results indicate that NMS can affect charge transfer capacity through regulating pore structures and surface properties.
The GCD curves corresponding to these three PCs are typical performance curves of electric double-layer capacitors ( Figures  6D-F). According to Eq. (1), it is calculated that when the current density is 0.1, 0.2, 0.5, 1.0, and 2.0 A g −1 , the specific capacitances of EP-PC are 278.3, 264.9, 252.3, 244.7, and 236.2 F g −1 , respectively, and those of EPU-PC-5 are 344.6, 323.9, 308.0, 297.8, and 286.9 F g −1 , respectively, and those of EPM-PC-5 are 352.9, 334.3, 315.0, 305.8, and 296.5 F g −1 , respectively. Compared with those reported EP-based PCs, the specific capacitances of EPU-PC-5 and EPM-PC-5 are significantly larger (Gao et al., 2014a;Gao et al., 2014b;Gai et al., 2015;Tian et al., 2016;Ren et al., 2018;Xie et al., 2020b). Further calculations indicate that under the same current density, the specific capacitance of EPU-PC-5 is approximately 1.22 times that of EP-PC, and the specific capacitance of EPM-PC-5 is approximately 1.26 times that of EP-PC. It is noticed that the S BET s of EP-PC are similar to those of the other two PCs (Table 1), and it has more N and O ( Table 2), whose corresponding functional groups can provide pseudocapacitance. Still, its specific capacitance is significantly lower than that of the other two. For these tested cells, the surfaces in PCs that can provide capacitance are effective, and those surfaces which ions cannot access are ineffective. Thus, it can be inferred that the effective S BET s of EP-PC are markedly smaller than the other two. That is to say, the increase in specific capacitances of EPU-PC-5 and EPM-PC-5 is mainly due to the development of their mesopore structures which can increase their effective S BET s.
According to Eqs. 2, 3, the energy densities and power densities of these three PC-based cells are, respectively,  Figure 7A. EP-PC-based cell has an energy density of 9.7 Wh kg −1 with a power density of 49.9 W kg −1 , EPU-PC-5-based cell has an energy density of 11.9 Wh kg −1 with a power density of 49.7 W kg −1 , and EPM-PC-5-based cell has an energy density of 12.2 Wh kg −1 with a power density of 46.0 W kg −1 . Besides, it can be observed that under the same power density, the energy densities of EPU-PC-5-and EPM-PC-5-based cells are larger than that of EP-PC-based cell, indicating that NMS can improve the energy storage performance of the PC corresponding devices. The rate performance curves of EP-PC, EPU-PC-5, and EPM-PC-5 in Figure 7B present similar trends. Their specific capacitances decrease as current densities increase, which is mainly caused by the increase of electrode concentration polarization due to the slow diffusion of ions, which increases the internal resistance (Zhang et al., 2018;Mehare et al., 2020). It is observed that when less than 2 A g −1 , their specific capacitances show rapid decay, and when larger than 2 A g −1 , their specific capacitances show slow decay. Furthermore, when less than 30 A g −1 , the decay rates of these three curves are similar, and when larger than 30 A g −1 , the relationship of their decay rates is EP-PC > EPM-PC-5 > EPU-PC-5, indicating that the relationship of their rate performance is EP-PC < EPM-PC-5 < EPU-PC-5. The better rate performance of EPU-PC-5 and EPM-PC-5 should be mainly due to their more developed mesopore structures.
The Nyquist plots of EP-PC, EPU-PC-5, and EPM-PC-5 are shown in Figure 8A. In low-frequency range, these three curves are nearly vertical, indicating that their corresponding cells present idealized electric double-layer behaviors. In the highfrequency range, the diameter of the semicircular arc in the curve usually represents the charge-transfer resistance related to Faradic reactions between ions and oxygen-containing or nitrogen-containing functional groups on electrode material,  which corresponds to R ct in Figure 8B (Song et al., 2015). Therefore, as shown in Supplementary Table S1, the relationship of the charge-transfer resistances of cells corresponding to these three PCs is EPU-PC-5 (0.85 Ω) > EP-PC (0.32 Ω) > EPM-PC-5 (0.17 Ω). Evidently, large chargetransfer resistance means weak charge-transfer capacity. Therefore, the above result is consistent with the relationship of the charge-transfer capacity reflected by those CV curves ( Figures 6D-F). The intercepts on the Z' axis represent the equivalent-series resistances (ESRs) related to the intrinsic resistances of electrode materials and electrolyte, and the contact resistance between working and reference electrodes, which corresponds to R s in Figure 8B (Yu et al., 2017;Zhu et al., 2017). Thus, as shown in Supplementary Table S1, the relationship of the ESRs of the cells corresponding to PCs is EP-PC (0.45 Ω) > EPM-PC-5 (0.40 Ω) > EPU-PC-5 (0.38 Ω). Then it is concluded that NMS can reduce the ESRs of those cells, which is mainly attributed to the enhancement of the conductivities of those nitrogen modified PCs with the improvement of graphitization degrees. The cycle performance curves of EP-PC, EPU-PC-5, and EPM-PC-5 in Figure 9 obtained at 5 A g −1 for 10,000 cycles show similar three stages, and the comparison diagrams of GCD curves corresponding to the first and 10000th cycles are shown in Supplementary Figure S1. Each stage in each curve has almost equal retention of specific capacitance, suggesting that highly reversible double-layer behaviors occur at electrodes. As curves go from stage 1 to stage 2 and then to stage 3, the specific capacitance retentions suddenly increase and then suddenly decrease. These changes may be due to ion desolvation occurring, which allows ions to enter micropores to generate capacitance (Urita et al., 2014;Wu et al., 2020b), and then micropores are filled with ions and no capacitance is generated. Besides, all their specific capacitance retentions are larger than 92.0%, indicating that the cells corresponding to these nitrogen modified EP-based PCs can maintain excellent longcycle performance.

CONCLUSION
In summary, NMS can improve the electrochemical performance of PCs by optimizing their pore structures and surface chemical properties. Urea and melamine are used as nitrogen-rich additives to prepare nitrogen-modified EP-based PCs through carbonization and activation. On the one hand, the two nitrogen-  rich additives can promote the development of the mesoporous structures of the obtained PCs and can improve their graphitization degrees with almost unchanged S BET s. The effect of urea is slightly better than that of melamine. On the other hand, both of them can increase the contents of C element in PCs and can greatly increase the relative intensity contents of graphitized N. It is noticed that the content of each element in those two nitrogen modified EP-based PCs is almost equal, but the existing state of each element is different, especially N element. Benefiting from the optimized microstructures, the specific capacitances of nitrogen modified EP-based PCs are increased by approximately 22%, their high-rate performance is enhanced, the ESRs of their corresponding cells are reduced, and they show excellent long-cycle performance. Considering the simple process and remarkable effect of NMS, it offers great potential to apply in the preparation of PCs, and it can combine with other postprocessing methods to further improve the performance of PCs.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.

AUTHOR CONTRIBUTIONS
ML conducted the experiments and scientific discussions. KH contributed to research suggestions. ZT helped with writing. MW helped with characterization. YC and XL helped with experiments. All authors listed made a substantial and intellectual contribution to the work and approved it for publication.

FUNDING
This work was supported by the Natural Science Foundation of Shandong, China (ZR2017MEE010), and the Fundamental Research Funds of Shandong University (2016JC005).