Structural Characteristics and Electrochemical Performance of N,P-Codoped Porous Carbon as a Lithium-Ion Battery Anode Electrode.

Biomass-derived heteroatom-doped carbons have been considered to be excellent lithium ion battery (LIB) anode materials. Herein, ultrathin g-C3N4 nanosheets anchored on N,P-codoped biomass-derived carbon (N,P@C) were successfully fabricated by carbonization in an argon atmosphere. The structural characteristics of the resultant N,P@C were elucidated by SEM, TEM, FTIR, XRD, XPS, Raman, and BET surface area measurements. The results show that N,P@C has a high specific surface area (S BET = 675.4 cm3/g), a mesoporous-dominant pore (average pore size of 6.898 nm), and a high level of defects (I D/I G = 1.02). The hierarchical porous structural properties are responsible for the efficient electrochemical performance of N,P@C as an anode material, which exhibits an outstanding reversible specific capacity of 1264.3 mAh/g at 100 mA/g, an elegant rate capability of 261 mAh/g at 10 A, and a satisfactory cycling stability of 1463.8 mAh/g at 1 A after 500 cycles. Because of the special structure and synergistic contributions from N and P heteroatoms, the resultant N,P@C endows LIBs with electrochemical performance superior to those of most of carbon-based anode materials derived from biomass in the literature. The findings in this present work pave a novel avenue toward lignin volarization to produce anode material for use in high-performance LIBs.

Introduction

Lithium ions batteries (LIBs) have garnered a large amount of attention from the public due to their prominent features of eminent power density, good cycle life, decreased pollution, and no memory effects.[1] Commercial LIBs are very suitable for use in various portable electronic devices and hybrid electric vehicles.[2] Recently, to bridge the gaps in energy capacity and reliable operation using graphite as a LIB anode, biomass-derived porous carbonaceous materials have been used as alternative anode electrodes for LIBs, including carbon nanotubes,[3] carbon nanofibers,[4] carbon nanobeads,[5] hollow carbon nanospheres,[6] graphene,[7] graphitic carbon,[8] hierarchical porous carbon,[9] and their hybrids.[10] However, more attention should be paid to improving the electrochemical performance of these biomass-derived carbonaceous anode electrodes.

Modifying carbonaceous anode electrodes by incorporating heteroatoms such as N, P, S, O, B, Sn, and Ge may imporove the electrochemical characteristics of anode electrode, since incorporating heteroatoms in the carbon skeleton can provide more active sites and increase the ion-storage capacity compared to that of pristine carbon.[11−13] For instance, N,O-codoped hierarchical porous carbon derived from biomass could enhance the wettability between the electrolyte and the electrode to deliver an eminent reversible capacity and an excellent cycling stability in LIBs.[11] N,P-Codoping could result in increased active site exposure for better electrochemical performance than that of sole-atom-doped carbon.[14,15] As anodes for LIBs, vesicle-structured Sn4P3@P/N dual-doped carbon nanocomposites (Sn4P3@PNC) exhibited good rate capabilities and satisfactory cycling stabilities, which were ascribed to the intimate contact of Sn4P3 with PNC and the unique vesicle structure.[16] G,P-Codoped porous carbon nanoparticles (GPBN/C) were used as LIB anode materials with efficient electrochemical performance.[17] The higher the specific surface area and the pore volume of carbon materials, the higher the exposure of active sites and the energy density of batteries.[18] Therefore, adjusting the microstructure of the biomass-derived carbonaceous anode materials through heteroatom incorporation is of great importance to attain high electrochemical performance in LIBs systems.

Lignin, the largest aromatic biopolymer in nature, is a characteristic component of secondary walls of biomass, with a content of 15–30 wt %.[19] With the development of biomass biorefinery, the productivity of lignin has sharply increased in the last two decades. Lignin can be upgraded into biofuels, biochemicals, and biomaterials to facilitate the economic conversion of biomass toward a fossil-free society.[20] However, little information on lignin-derived porous carbon doped with heteroatoms as LIBs anode materials is avilable in the literature. A graphitic carbon (g-C3N4) nanosheet was always synthesized via a polycondensation reaction by annealing dicyandiamide, melamine, and urea at 550 °C in air for 4 h.[21,22] Researchers have demonstrated that g-C3N4 is a good interface enabler for depressing the formation of the dendritic lithium for solid-state lithium-metal batteries.[22−24] Additionally, g-C3N4 can also serve as a photocatalyst,[25] and g-C3N4-supported single-atom catalysts have been used for energy storage.[26] However, little information has been available on the use of an active carbon material with g-C3N4 as a soft template for LIB anodes so far.

In the present study, a simple carbonization approach is proposed to fabricate N,P-codoped lignin-derived hierarchical porous carbon (N,P@C) using ZnCl2 as a pore-forming agent and g-C3N4 as a soft template for use as efficient anode material in LIBs. The structure of the as-obtained N,P@C was comprehensively elucidated by Fourier transform infrared spectroscopy (FT-IR), X-ray diffraction (XRD), scanning electron microscopy (SEM), Raman spectroscopy (Raman), transmission electron microscopy (TEM), X-ray photoelectron spectroscopy (XPS), and N2 adsorption–desorption isotherm curves for Brunauer–Emmet–Teller (BET) surface area measurements. As counterparts, lignin-derived carbon (C) without a heteroatom dopant and N-doped lignin-derived carbon (N@C) and P-doped lignin-derived carbon (P@C) with a sole heteroatom dopant were also synthesized following a procedure similar to that for the synthesis of N,P@C . Serving as the anode electrode for LIBs, the as-obtained N,P@C exhibited a higher reversible specific capacity, a better rate capability, and a more satisfactory cycling stability than other carbonaceous counterparts derived from biomass.

Materials and Methods

Materials

Ligin (lot no. 04414PEV, CAS 8068-05-1) and urea (A.R) were both available from Sigma Co. Ltd. (Shanghai, China). Polyvinylidene fluoride (PVDF), polyvinylpyrrolidone (PVP), and N-methyl pyrrolidone (NMP) were purchased from Beijing Chemical Factory (Beijing, China). Analytic-grade zinc chloride (ZnCl2) and N,N-dimethylformamide (DMF) were purchased from Sinopharm Chemical Reagent Co. Ltd. (Shanghai, China). Lecithin (the content of elemental P was 2.7%) was gifted from Hebei Meiyesiwei Biological Technology Co. Ltd. (Hebei Province, China).

Fabrication of the N,P@C Material

The synthesis scheme for N,P-codoped lignin-derived carbon (N,P@C) is shown in Figure a. First, 10.5 g of g-C3N4 as the N source and a soft template was uniformly dispersed in 300 mL of deionized water, then 3.5 g of lignin and 14 mmol Zn2+ were dissolved in the water.[20,21] The mixture was stirred at a rotating speed of 800 rpm for 2 h and then aged at 25 °C for 12 h. Afterward, the mixture was centrifuged, and the residual solid was dried at 80 °C for 5 h to yield the g-C3N4 templated coated with lignin/Zn2+ coordination (lignin/Zn@g-C3N4). Subsequently, 100 mg of lecithin (the content of elemental P was 2.7%) was dissolved in 20 mL of ethanol under ultrasonic conditions for 30 min. Then, 3 g of lignin/Zn@g-C3N4 was dispersed in this solution, and the mixture was stirred at 800 rpm for 48 h. After being kept still overnight, the mixture was centrifuged at 5500 rpm for 15 min, and the residue solid was dried at 50 °C to yield the lignin/Zn/P@g-C3N4 composite.

Figure 1

Overview of the synthesis procedure and morphology of N,P@C, which was imaged by SEM and TEM methods. (a) Synthesis scheme of N,P@C. (b) SEM image, (c) TEM image, (d) high-resolution transmission electron microscopy (HR-TEM) image, (e) and high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) elemental mapping of N,P@C.

Furthermore, 3 g of the lignin/Zn/P@g-C3N4 composite as a precursor was uniformly placed in a porcelain boat. The precursor in boat was annealed at a temperature program in a tube furnace with flowing argon gas. The temperature program was as follows: the temperature was increased from 25 to 550 °C with the heating rate of 25 °C/min and kept at 550 °C for 2 h, then increased to 950 °C with the heating rate of 10 °C/min and kept h at 950 °C for 4, and finally naturally cooled to 30 °C. The resultant solid was washed using DMF as the solvent for 36 h, then washed with deionized water to remove impurities. The treated solid was dried at 80 °C under vacuum conditions to obtain N,P-codoped lignin-derived carbon (N,P@C) sample.

As counterparts, lignin-derived carbon (C) without a heteroatom dopant and N-doped lignin-derived carbon (N@C) and P-doped lignin-derived carbon (P@C) with a sole heteroatom dopant were synthesized using a procedure similar to the aforementioned N,P@C synthesis.

Structural Characteristics of the N,P@C Material

The morphology images of N,P@C were recorded using a JEOLJSM-7500F SEM (Japan) and a JEM-2100F TEM (Japan). The crystalline phase of N,P@C was measured by XRD (Germany). The defect degree of N,P@C was determined by Raman spectroscopy. The linkage bond of N,P@C was confirmed by Fourier transform infrared spectroscopy (FT-IR); spectra were recorded from 500 to 4000 cm–1 at a resolution of 2 cm–1. The specific surface areas and pore sizes of N,P@C were determined by nitrogen adsorption–desorption isotherm curves. The element compositions and species of N,P@C were analyzed by XPS.

Electrochemical Measurements

Electrochemical experiments for carbon materials include the preparation of a carbon-based working electrode and electrochemical activity measurements.[1,2] To prepare a carbon-based working electrode, a total weight of 18 mg of carbon material (14.4 mg, 80%), acetylene black (1.8 mg, 10%), and PVDF (1.8 mg, 10%) were evenly stirred and ground for 5 min. Afterward, the appropriate NMP was put into the mixture to form a diluted slurry. To prepare an anode plate, the slurry was blade-coated on the copper foil and then dried at 120 °C in a vacuum-drying oven for 12 h. The dried anode plate was then pressed into a wafer with a 1 cm diameter, which could be used as working electrode. Subsequently, CR2025 coin-type cells were prepared with the working electrode in an argon-filled glovebox in which the moisture and oxygen pressures were both less than 1.0 ppm. The reference electrode was lithium foil, and the electrolyte was 1.0 M LiPF6 in a mixture of 50% ethylene carbonate (EC) and 50% diethyl carbonate (DEC). The electrochemical activity of the carbon-based working electrode was measured as follows: Cyclic voltammetry (CV) was conducted on a CS350 electrochemical workstation from 0 to 3.0 V with a 5 mV amplitude at a scan rate of 0.05 mV/s. The galvanostatic charge–discharge curves were recorded at 100 mA/g on a CT3008 battery test system in a voltage windows of 0.1–3.0 V. The rate capacity was tested at different current densities of 0.1, 0.5, 1, 2, 5, and 10 A for 10 cycles. The cycling performance and its corresponding Coulombic efficiency (CE) were both analyzed at 0.5 and 1 A/g for 500 cycles and at 5 A/g for 1000 cycles.

Results and Discussion

The schematic for the fabrication of N,P@C and its morphology as imaged by SEM and TEM are shown in Figure . As shown in Figure a, a lignin/Zn2+ supermolecular composite was formed via the covalent coordination assembly of lignin and Zn2+ in an aqueous system at mild temperature, which was detailed in our previous work.[19] The lignin/Zn2+ supermolecular composite was coated on a g-C3N4 nanosheet to achieve lignin/Zn@g-C3N4 complexes. Afterward, P was impregnated into the lignin/Zn@g-C3N4 framework in an ethanol solution using lethicin as the P source. Subsequently, the precursor lignin/Zn/P@g-C3N4 was annealed in a tube furnace in an argon atmosphere to yield the N,P@C sample. At 950 °C the occurrence of phenomena of Zn2+ evaporation, g-C3N4 decomposition, and lignin graphitization could result in the hierarchical porous structure with defect carbon of N,P@C due to the rearrangement of carbon atoms.[21]

As shown by the SEM image in Figure b, the morphology of the resultant N,P@C is a stacked wrinkle nanosheet, similar to N@C (Figure S1) produced using ultrathin g-C3N4 as a soft template. In contrast, the morphologies of the lignin/Zn2+ assembly, C, and P@C imaged by SEM are smooth without the g-C3N4 template (Figure S1). The TEM image of N,P@C in Figure c shows the layered silk-like structure of the sample, indicating the ultrathin feature of the nanosheets. The HR-TEM image in Figure d confirms the presence of a large quantity of pores in the N,P@C material with a degree of low graphitization. It was demonstrated that these tiny pores could offer reservoirs for lithium ion storage, which improved the LIB capacity significantly.[1] HAADF-STEM elemental mapping (Figure e) points out that elements N and P are uniformly distributed on the surface of N,P@C. The codoping of N and P into the final porous carbonaceous material is further confirmed by the FT-IR and XPS spectra in Figure a and Figure , respectively.

Figure 2

Structural properties of N,P@C as detected by FT-IR, XRD, Raman, and BET techniques. (a) FT-IR spectrum, (b) XRD pattern, (c) Raman spectrum, and (d) BET isothermal curve of N,P@C.

Figure 3

Elemental compositions and species of N,P@C measured by XPS. (a) XPS survey spectrum, (b) high-resolution XPS C 1s spectrum, (c) high-resolution XPS N 1s spectrum, and (d) high-resolution XPS P2p spectrum of N,P@C.

Furthermore, XRD, Raman, FT-IR, and BET measurements were conducted to shed light on the detailed structural information on N,P@C. The results are shown in Figure .

The FT-IR spectrum in Figure a reveals bands at 1460 and 1238 cm–1 that are assigned to the C–N linkage bond and bands at 1115 and 829 cm–1 that are ascribed to the P–O linkage bond.[14] It suggests that heteroatoms N and P are codoped into the N,P@C material. This result agrees well with the HAADF-STEM image in Figure e. The XRD pattern in Figure b displays a relatively strong and broad graphitic carbon (002) peak at about 20° and a weak and narrow carbon (100) peak at 42°, revealing the formation of graphitic carbon during the pyrolysis treatment. Similar XRD patterns were also found in C, N@C, and P@C materials (Figure S2). This means that pyrolysis treatment of carbon is prone to causing graphitization.[1,2,19−21,27] Studies have revealed that the graphitic carbon framework is beneficial for Li+ ion intercalation and diffusion.[28,29]

Raman spectroscopy was further conducted to measure the graphitization degree of N,P@C. As shown in Figure c, two distinct signal peaks are present at 1342 (D band, ascribed to disorder and defective graphitic structure) and 1599 cm–1 (G band, assigned to ordered graphitic carbon), and the ID/IG intensity ration is 1.02. This indicates that N,P@C tends to be a disordered amorphous or defective carbon framework due to heteroatom doping and Zn2+ evaporation.[11,27,30] The graphitization degree of N,P@C (ID/IG = 1.02) is higher than that of its counterparts (Figure S3) C (ID/IG = 1.15), N@C (ID/IG = 1.04), and P@C (ID/IG = 1.11). The improved graphitization structure of N,P@C can expose more active sites for lithium ion intercalation and transportation.[31,32]

The BET specific surface area (SBET) and pore structure of the carbon material are two important indicators of lithium ion intercalation and deinsertion.[28] The N2 adsorption–desorption isotherm curve of N,P@C in Figure d exhibits the characteristic type IV isotherm with a clear H3 hysteresis loop, which indicates that N,P@C has hierarchical porous features of micro-, meso-, and macropores.[33] Many researchers have demonstrated that macropores can provide transportation channels for lithium ion diffusion, micropores can act as reservoirs to improve the capacity of lithium ion storage, and mesopores offer highways for lithium ion transportation.[9,11,33−36] The values for SBET, the average pore size, and the pore volume of N,P@C were calculated to 675.4 m2/g, 6.898 nm, and 2.383 cm3/g, respectively. In comparison with most of the biomass-derived anode materials in the literature, N,P@C shows higher SBET value, which is associated with the synthesis procedure of the carbon material.[1,32,37] Both the high SBET value and the hierarchical pore structure benefit the efficient electrochemical performance of lithium ion battery by exposing more active sites and facilitating lithium ion diffusion.[37,38]

To further explore the elemental compositions and species of N,P@C, XPS survey spectra were recorded, as shown in Figure . The existence of C, N, and P peaks in the XPS survey spectrum (Figure a) further confirms the multiheteroatom codoping of the final carbon skeleton with N and P elements. This phenomenon is also demonstrated by the HAADF-STEM (Figure e) and FTIR (Figure a) results. The presence of the O peak in Figure a reveals oxygen residue in the final carbon material from the lignin starting material (the content of elemental O ranges from 35 to 53 at %). The element composition of N,P@C are mainly consisted of carbon (87.65 at %), nitrogen (4,16 at %), phosphorus (0.65 at %), and oxygen (7.53 at %) (Figure a). As can be seen from the high-resolution XPS C 1s spectrum in Figure , the weak O—C=O peak at 287.9 eV is deconvoluted, suggesting that the source of the oxygen element was probably the oxygenated groups on lignin.[14] Two strong peaks are deconvoluted, namely, the C=C peak at 284.6 eV (ascribed to sp2 graphitic carbon) and the C—N/C=N peak at 285.6 eV (ascribed to C-bonded carbon), indicating the successful N-doping in the final carbonized sample.[39] As seen from the high-resolution XPS N 1s spectrum in Figure c, four peaks are deconvoluted: that for oxidized N at 403.8 eV, that for graphitic N at 401.6 eV, that for pyrrolic N at 400.2 eV, and that for pyridinic N at 398.5 eV.[14] In the high-resolution XPS P2p spectrum shown from Figure d, two peaks are fitted: that for the P–C bond at approximately 131.6 eV and that for the P–O bond at approximately 133.2 eV.[14] Therefore, the XPS survey spectra of N,P@C demonstrate the desired elemental composition (C, N, P, and O) without any other impurity contamination. A heteroatom dopant (like N, O, and P) can enhance the electrochemical performance by enhancing the reactive activity and offering lots of ion storage sites.[7,11,40]

After elucidating the structural properties, experiments on the electrochemical performance of N,P@C were further carried out, and the results are shown in Figure .

Figure 4

Electrochemical performances of the N,P@C anode material. (a) Cyclic voltammetry curves of three cycles in voltage windows from 0.1 to 3.0 V at 100 mA/g. (b) Galvanostatic charge–discharge curves at 100 mA/g. (c) Rate capability at different current densities of 01, 0.5, 1, 2, 5, and 10 A. (d) Cycling performance and Coulombic efficiency at 500 mA/g. (e) Cycling performance and Coulombic efficiency at 1 A/g. (f) Cycling performance and Coulombic efficiency at 5 A/g.

The lithium ion storage behavior of N,P@C can be understood from the CV measurement. As shown in from Figures a and S4, the CV curves of the as-obtained carbon materials, namely, N,P@C, C, N@C, and P@C, are similar in shape and reduction potential except for the reversibility of the first cycle. The broad irreversible negative peak at about 0.75 V of the first cycle is characteristic of CV curves of biomass-derived carbon as anode material. The probably reasonable explanation is the decomposition of the electrolyte and the formation of a solid electrolyte interface (SEI) film.[2,4,5,41,42] In the second and third cycles, this wide reduction peak disappears, and the CV curves overlap. The galvanostatic charge–discharge curve of N,P@C in Figure b shows that the initial charge specific capacity is 1106 mAh/g and the corresponding initial CE is approximately 37%, slightly higher than those of C (31%), N@C (34%) and P@C (32%) (Figure S5). It is universally acknowledged that a low CE is a big challenge in the practical application of biomass-derived carbons as LIBs anode materials.[1,2,41−45]

Appealingly, the N,P@C anode retains its high Li+ ion storage and excellent rate capability (Figure c). As shown in Figure c, the reversible capacities at 0.1, 0.5, 1, 2, 5, and 10 A/g for 10 cycles were calculated to be approximately 1260, 687, 584, 476, 342, and 261 mAh/g, respectively. Even at an extremely high current density of 10 A/g, the reversible capacity remained 261 mAh g–1, approximately 70% the theoretical capacity of graphite (372 mAh/g). It should be noticed that the discharge capacity of N,P@C can recover 1264.3 mAh/g when the current density is returned to 0.1 A/g. As an example, the rate capability curves of the N@C counterpart are shown in Figure S6a; its discharge capacity is greatly inferior to that of N,P@C when the current density is returned to 0.1 A/g (Figure S6b). The reasonable explanation is ascribed to the synergistic contributions of heteroatom doping.[4,7,11−13,32,39,40] These data indicate that the N,P@C anode shows a good rate capability and excellent reversibility.

Furthermore, the long cycling performace of N,P@C was tested to evaluate its durability at 0.5 and 1 A after 500 cycles and at 5 A after 1000 cycles. As shown in Figure d, the reversible specific capacity of N,P@C is 1867.4 mAh/g at 0.5 A after 500 cycles, while the reversible specific capacities of C, N@C, and P@C are only 241.8, 547, and 230 mAh/g, respectively (Figure S7). The highly efficient performance of the N,P@C anode is ascribed to the combined effects of both the mesoporous-dominant hierarchical structure with a large SBET and the synergistic contribution of heteroatom doping.[7,10−14,27,28,31,32] As shown by the TEM images in Figure S8a and b, the morphology of the used N,P@C anode remains intact at a relative low current density of 0.5 A after 500 cycles, confirming the satisfactory structural stability of the N,P@C material. However, the thermostability of the used N,P@C material is higher after 500 cycles than that of the pristine material (Figure S8c). This can probably be ascribed to some Li+ ion adsorption on the interlayer and surface of N,P@C after cycling. When the current density was increased to 1 A/g (Figure e), N,P@C showed a stable capacity of 1463.8 mAh/g after 500 cycles, a durability than that of N@C. When the current density was further enhanced to 5 A/g, the reversible specific capacity of N,P@C was about 625.8 mAh/g after 500 cycles and 198 mAh/g after 1000 cycles (Figure f). The decrease in capacity during cycling mainly arises from the structural destruction caused by large quantities of defects at a high current density of 5 A/g. As an example, the cycling performance and the corresponding CE of N@C were examined at 1, 2, and 5 A/g. It can be observed in Figures S9a and c that N@C has a much lower stable capacity at 1 (471 mAh/g after 500 cycles) and 5 A/g (115.5 mAh/g after 1000 cycles).

To better highlight the electrochemical performance of N,P@C, an impedance test was conducted in lithium batteries, and the result is shown in Figure .

Figure 5

Electrochemical impedance spectra of N,P@C in lithium batteries.

As shown in Figure , the Ohmic resistance at the high-frequency semicircle expresses the resistance of the electrolyte or the contact resistance,[46] while the Warburg impedance at the lower-frequency inclined line at approximately 45° is associated with the diffusion of Li+ ions in the carbon anode electrode.[47] As can be observed from the Nyquist plot in the inset of the figure, the charge transfers resistance (Rct) of the N,P@C anode electrode was found to be 75.56 Ω, significantly lower than those of biomass-derived carbon from ramie fiber (Rct = 239.7 Ω) and corncob (Rct = 129.2 Ω).[48] This indicates that the electronic conductivity of the N,P@C electrode is satisfactory in a LIB system.

Based on is outstanding cycling performance and satisfactory reversibility, lignin-derived N,P@C shows potential for promising practical applications as an anode material in LIBs. Table S1 comprehensively compares the electrochemical performance of lignin-derived N,P@C and those of biomass-derived carbonaceous materials in LIBs. Obviously, the specific capability of N,P@C is much better than those of most biomass-derived carbonaceous anode materials in the literature (Table S1). Although a lot of research on biomass-derived carbonaceous anode materials has been reported, more efforts should be made to improve the specific capacity of batteries and the initial CE in the coming practical applications.

Conclusions

In summary, N,P@C was successfully fabricated by coating a lignin/Zn+ composite on a g-C3N4 soft template, followed by an annealing procedure using lecithin as P source. The resultant N,P@C was characterized by SEM, TEM, FTIR, XRD, XPS, Raman spectroscopy, and BET surface area measurements. The combined effects of the hierarchical porous structure, the high SBET, and the synergistic contribution of multiheteroatom dopants are responsible for the outstanding reversible capacity, excellent rate capability, and satisfactory cycling performance of N,P@C as an anode material in LIBs. Compared to previous research on electrochemical performance, the specific capacity N,P@C is superior those of most biomass-derived carbonaceous anode materials in earlier works. The findings not only reveal that N,P@C is a promising potential anode material in practical LIB applications but also paves a novel avenue for lignin valorization toward battery anode materials.