In recent years, much attention has been paid to lithium-ion batteries (LIBs) and sodium-ion batteries (SIBs) to store clean energy [1-6]. LIBs have been widely applied in portable devices owing to their high energy density and long cycle life [7-9]. However, the shortage of lithium resources seriously hinders the development of LIBs in future applications [3, 10-12]. SIBs are promising alternative to LIBs because of the natural abundance and lower cost of sodium [13-15]. In spite of extensive work on electrode materials for both LIBs and SIBs, it still remains a challenge to develop the eco-friendly anode materials with high capacities and good rate capabilities.

Carbon materials are recognized as one of the most promising anode materials due to their high electrical conductivity, high chemical and thermal stability, large surface area, low cost and easy availability. Currently, a variety of carbon materials, including carbon spheres [16], carbon nanotubes [17], carbon nanofibers [18] and graphene [19] have been employed as anode materials for LIBs and SIBs. However, the synthesis processes of most carbon materials are complex, and expensive equipment are also needed. In recent years, it is revealed that biomass derived carbon materials are ideal anode materials for porous structure, which ensures high capacity and excellent cyclability, as well as the merits of cost-effective resources and eco-friendliness [20-22]. Various of biomass derived carbon have been proved to possess superior Li/Na storage performance, i.e., silicon/nitrogen-doped carbon hierarchically structured spheres from rice husk [23], nitrogendoped porous carbon nanosheets from eucalyptus tree leaves [24], porous hard carbon from Pomelo peels [25], hierarchically porous carbon from peanut skin [26], etc.

In this work, we chose bamboo leaves as the precursor to fabricate nanostructured carbon by carbonization and etching in HF. It is well known that bamboo distributes all over the world and needs short growth period, thus the raw materials are abundant and easily obtained [27]. Moreover, the connected three-dimensional microstructures in the bamboo leaves consisting of hydrate SiO₂ nanoparticles, lignin, semi-cellulose, and cellulose, provide a good template for the formation of hierarchically porous carbonaceous structures via carbonization and etching in HF [28, 29]. The obtained hierarchically porous carbon displays sheet-like nanostructure, high surface area, adjustablepore structure andmoderate nitrogen doping. Owing to the advantages, bamboo leaves derived hierarchically porous carbon materials exhibit remarkable electrochemical property. After optimizing the synthetic condition, the as-derived carbon could sustain a discharge capacity of 450 mAh/g after 500 cycles at the current density at 0.2 A/g for LIBs, and maintain a discharge capacity of 180 mAh/g after 300 cycles for SIBs at the current density of 0.1 A/g.

The dried bamboo leaves used in this work were collected in October 2018 from a bamboo garden in Hunan University. The bamboo leaves were firstly washed with deionized water and absolute ethanol, and dried at 100 ℃ overnight. Then, the leaves were soaked in 1 mol/L HCl solution at 100 ℃ for 3 h to remove alkali impurities and metal impurities. After being washed to neutral, the acid-treated bamboo leaves were carbonized in argon atmosphere at 600 ℃, 700 ℃ and 800 ℃ for 2 h, respectively. Finally, the samples were etched in HF solution to remove the biogenetic SiO₂ and washed several times to obtain N-doped porous carbon materials. Here, the as-received carbon materials were termed as C-600, C-700 and C-800 according to the corresponding annealing temperatures.

The morphology, structure and composition of the carbon materials were characterized by scanning electron microscopy (SEM) (HitachiS4800) with an accelerating voltage of 15 kV, transmission electron microscope (TEM) (JEOLJEM-3010) performed at 200 kV, X-ray diffractometer (XRD) (Bruker D8 Adv), X-ray photoelectron spectroscopy (XPS), Thermo Scientific Escalab 250Xi) and Raman scattering (LabRAM HR Evolution, HORIBA system, at an excitation wavelength of 532 nm). The surface area and pore diameter distribution were analyzed by BrunauerEmmett-Teller (BET) measurements using an Autosorb-iQ (Quantachrome Instruments).

To prepare electrodes, the carbon materials were mixed with acetylene black and carboxyl methyl cellulose at a weight ratio of 80:10:10 to form a slurry. Then, the homogeneous slurry was casted onto a piece of Cu foil current collectors, followed by drying in vacuum at 60 ℃ for 12 h. The average active material loading was controlled to be 1.0 mg/cm². For electrochemical test, two-electrode cells were made with the as-prepared electrode as working electrode and Li or Na metal as counter electrode for LIBs and SIBs, respectively. The polypropylene separator and LiPF6 in EC/DEC (1:1 by weight) for LIBs or NaPF₆ in EC/DEC (1:1 by weight) for SIBs were also used. The CR2032-type coin cells were fabricated in an argon-filled glove box. Galvanostatic chargedischarge testing was conducted on a Neware battery test system (CT-4008) between 0.01–3 V. Electrochemical impedance studies (EIS) were carried out using electrochemical workstation.

The synthesis process of the hierarchically porous carbon is schematically illustrated in Scheme 1. The alkali impurities and metal impurities were firstly removed from dried bamboo leaves via acid washing. Then, the bamboo leaves were carbonized in argon atmosphere at high temperatures to obtain intermediate product. Finally, the porous carbon materials were obtained through etching intermediate product.Figs. 1a, c and e reveal that all the as-prepared carbon obtained at different temperatures (C-600, C-700 and C-800) possess sheet-like microstructures. High-magnification SEM images of C-600, C-700 and C-800 (Figs. 1b, d and f) present numerous of pores on the surface of individual sheet, which might be caused by the etching of SiO₂, indicating the porous structure of the as-obtained samples. TEM images in Figs. 2a–f further verify the formation of hierarchically porous structure. The porous structure may provide favorable pathways for the diffusion and absorption of ions [30], and could also prevent and minimize the volume expansion of electrode materials during the charge-discharge process.

XRD was employed to confirm the phase of the as-prepared carbon materials. As shown in Fig. 3a, a broad peak at ~24° is detected, which can be indexed to the (002) peak of graphitic carbon [31]. The weak peak centered at 43° can be ascribed to the (100) diffraction peaks of graphitic carbon. The characteristic diffraction peaks of carbon materials indicate successful transformation of bamboo leaves to carbon [32]. Fig. 3b shows the Raman spectra of the as-derived carbon materials. All samples exhibit two peaks centered at 1330 cm^-1 (D band) and 1600 cm^-1 (G band). They are consistent with structural defects and disorder induced in the graphene layers of carbon and the first order stretching vibration of sp₂ carbon atoms in the two-dimensional hexagonal lattice, respectively [33]. We are known that the I_D/I_G manifest the level of structural disarrangement and defect intensity, high I_D/I_G manifest large defects of the structure. The I_D/I_G of C-600, C-700, and C-800 were measured to be 0.827, 0.830 and 0.835, respectively, indicating the improvement of graphitization degree of the carbon materials [34]. However, from the strong D-band peak of C-700 we can see that it had a low crystallinity, it can attribute to contained abundant disordered defects.

To further confirm the porosity of the carbon materials, N₂ adsorption-desorption isothermal analysis was used.Figs. 4a, c and e show the nitrogen adsorption-desorption isotherms of C-600, C-700, and C-800, respectively. They display the typical type-IV shape, indicating all carbon materials possess the substantial mesopores. On the basis of Brunauer-Emmett-Teller (BET) analysis, the specific surface areas are 526.9 m²/g, 475.9 m²/g, 393.8 m²/g for C-600, C-700 and C-800, respectively. The specific surface area of C-800 is slightly lower than these of C-600 and C-700, which was attributed to the partial pore collapse caused by higher carbonization temperature [35].Figs. 4b, d and f show their corresponding pore size distributions. It can be seen that the sizes of the mesopores in the carbon materials range from 3 nm to 5 nm. The mesopores of the carbon resulted from the removal of SiO₂. The porous structure provides efficient surfaces for the adsorption of Li/Na ions and short transmission pathway for the ion transfer and electrolyte pervasion, which may lead to excellent electrochemical performance, including higher specific capacity and excellent rate performance in both LIBs and SIBs [36].

XPS was used to investigate the elemental compositions and element binding state. The XPS survey spectra (Fig. 5a) present three peaks centered at 285, 400 and 531.7 eV, which can be attributed to C 1s, N 1s, and O 1s, respectively. The results confirm the formation of nitrogen doped porous carbons. Moreover, XPS analysis displays that the nitrogen content decreases as the pyrolysis temperature rises. The content of nitrogen for C-600, C-700 and C-800 are 3.83%, 2.54% and 1.59%, respectively. The loss of nitrogen may be caused by the breakage of the N—C bond at high temperature [37].

The N 1s spectra can be employed to distinguish the nitrogen configurations. As shown inFigs. 5b-d, the two peaks at around 398 and 400.2 eV can be attributed to pyridinic N (N atoms situated at the edges of graphene planes) and graphitic N (N atoms embed in the graphene layer), respectively [38]. Firstly, pyridinic nitrogen can improve the energy storage performance due to the free p-electrons in the outermost shell of the pyridinic nitrogen through absorbing ion and the N atoms can enhance the conductivity of carbon [14]. Secondly, the pyridinic N can produce substantial defects among the carbon materials, offering mass open channels to facilitate Li⁺ and Na⁺ storage and transport together [39].

Lithium ion storage behavior of the as-derived porous carbon materials was investigated by cyclic voltammetry (CV) and galvanostatic charge/discharge cycling.Figs. 6a-c exhibit the CV curves for C-600, C-700 and C-800 at a current density of 0.2 A/g in the voltage range of 0.01–3 V, respectively. A peak located at 0.5 V was observed in the first scan and it disappeared in the subsequent cycles. The peak can be attributed to the irreversible process of the formation of solid electrolyte interphase (SEI) layer [40, 41]. The SEI films or reduction of the electrolyte and inhibit the lithium insertion into the carbon material cause the large irreversible capacity decrease [42]. Figs. 6d-f exhibit the discharge/charge profile of the electrodes of the first five cycles at a current density of 0.2 A/g between 0.01–3.0 V for LIBs. During the first cycle, the carbon electrodes delivered the specific discharge capacities of 1128, 1300 and 963 mAh/g for C-600, C-700 and C-800, respectively. The initial Coulombic efficiency of the carbon electrodes are 43%, 48% and 47% for C-600, C-700 and C-800, respectively. After the 10th cycle, Coulombic efficiencies increase to over 95%, indicating that the carbon materials have excellent reversibility and structural stability. Fig. 6g shows the rate performance of C-600, C-700, C-800 between 0.1 A/g and 5 A/g for LIBs. C-700 exhibited the best rate performance among the three carbon materials. The discharge capacities of C-700 are 650, 500, 400, 350 and 250 mAh/g at current densities of 0.1, 0.2, 1, 2 and 5 A/g, respectively. When the current density came back to 0.1 A/g, the capacity could be recovered up to 650 mAh/g. This indicates that C-700 has the outstanding stability and reversibility. As shown in Fig. 6h, one can find that the reversible capacity of C-700 is higher than those of other materials, and maintains at 450 mAh/g after 500 cycles. The specific capacities of C-600 and C-800 electrodes are reduced to about 250 mAh/g. The C-700 electrode shows better electrochemical performance than the previously reported lignin-derived carbon fibrous [43], bisporus derived hierarchically porous carbon [44] and mangrove-charcoalderived carbon [45] in LIBs. Fig. 6i exhibits the EIS profile of fresh electrodes for LIBs in a frequency range of 0.01 Hz to 100 kHz. Nyquist plots exerts a semicircle in the high-frequency region and an inclined line in low-frequency region. The semicircle presents charge-transfer resistance ( R_ct), which is related to the electrical conductivity of the active materials. A more vertical line in the lowfrequency region indicates the electrode is closer to an ideal capacitor. By fitting with the Zview program in the Sai software set, the charge transfer resistance, R_ct, for C-600, C-700 and C-800 electrodes are 3520, 205.6 and 427.7 V, respectively, confirming the superior charge-transfer kinetics of C-700 [32-34].

Fig. 7 exhibits CV and discharge/charge profile for SIBs. From Figs. 7a-c, one can observe a pair of redox peaks in the lower potential region. That demonstrates the insertion/de-insertion of sodium ions in the graphitic micro crystallites [46], which is consistent with these of lithium ions in carbon materials. Notably, the 2^nd and 3^rd charge-discharge curves were found to almost overlap, manifesting that the as-derived carbon materials have excellent stability and reversibility.Figs. 7d-f exhibit the discharge/charge profile of the electrodes of the five cycles between 0.01–3.0 V at a current density of 0.1 A/g for SIBs. The plateau about 1.3 V for 1^st discharge cycle could confirm the forming of SEI, and some irreversible reactions on the surface of carbon materials and constrain sodium ions at graphitic interlayers or active sites [47]. This is consistent with the results from the CV observations. During the first discharge cycle, the carbon electrodes delivered the specific discharge capacities of 544, 515 and 563 mAh/g for C-600, C-700 and C-800, respectively. The initial Coulombic efficiency of the carbon electrodes are 31%, 34% and 35% for C-600, C-700 and C-800, respectively. After 10 cycles, Coulombic efficiencies can increase to over 95%, indicating that the carbon materials have excellent reversibility in the following cycles. The rate performance of C-600, C-700, C-800 for SIBs were tested at the current densities between 0.1 A/g and 5 A/g. The cycling performance of C-700 exhibits the best cycling performance, with a capacity of 180 mAh/g after 300 cycles. The electrochemical performance of C-700 is better than carbon materials derived from corncobs and porous carbon derived from pistachio shell in SIBs [48, 49]. In the EIS profiles in (Fig. 7i), C-700 exhibits a charge– discharge resistance of 295.6 V, which is much lower than C-600 and C-800 electrodes. Summarily, the excellent performance of C-700 can be attributed to the appropriate lager surface area, smaller charge–discharge resistance and nitrogen doping.

In this work, we report the bamboo leaves-derived nitrogen doped porous carbon as high-performance anode materials for both LIBs and SIBs. The optimized carbon delivers a high discharge capacity of 450 mAh/g after 500 cycles at the current density of 0.2 A/g for LIBs, and a discharge capacity of 180 mAh/g after 300 cycles at the current densityof0.1 A/gforSIBs.Thelager surface areas and heteroatom doping endow the carbon materials with excellent electrochemical performance. The excellent performance, low cost and environmental friendliness of as-derived carbon materials make them promising anode materials for LIBs and SIBs in the future.

The authors declare that there is no interest for this manuscript.

This work was supported by the National Natural Science Foundation of China (Nos.11675051, 51302079, 51702138), the Natural Science Foundation of Hunan Province (No. 2017JJ1008), and the Key Research and Development Program of Hunan Province of China (No. 2018GK2031).