Introduction

To face the challenge of the energy depletion and global warming, there is an urgent need for the development of energy storage devices with high power density and high energy density simultaneously for advanced electronic equipment [1, 2]. Supercapacitor, as a promising candidate for energy storage system, has attracted tremendous attention owing to their high power density, long cycle life, and good reversibility [3,4,5]. It is well known that the electrode materials are the key to determine the electrochemical performance of supercapacitor, so the development of high-performance electrode materials is essential for the practical use of supercapacitors [6]. Carbon materials have long been recognized as ideal electrode materials for supercapacitors. In recent years, various carbon-based electrode materials have been designed and constructed from the metal-organic frameworks (MOFs) [7,8,9,10,11], polymer spheres [12], polypyrrole [13], and polybenzoxazine-based polymer [14] used as functional material precursors. Unfortunately, the preparation of abovementioned precursor materials more or less unsatisfactory, such as involving tedious procedures, using organic toxic chemicals, time-consuming and expensive, as well as not suitable for scale-up production and practical applications [15,16,17]. Therefore, selecting low-cost carbon precursors and appropriate preparation methods are great significance for prepared advanced high-performance porous carbons with adjustable pore structure parameters and economic advantages in the field of renewable energy research.

Recently, bio-derived activated carbon materials have been widely investigated because of their raw materials abundant and renewable in nature, high electrical conductivity, and excellent chemical stability [18,19,20]. Some natural biomass materials, such as rice husk [21], catkins [22], bagasse [23], ginkgo leaves [24], pomelo peel [25], waste celtuce leaves [26], and pomegranate husk [27], have been used in production of activated carbons. Among them, three-dimensional (3D) porous biomass-derived carbon materials are currently considered to be the ideal electrode materials [28]. The hierarchical porosity of these carbon materials provides both a conducting pathway for electrons and a fast ion-transport channel [29]. In addition, in order to enhance their electrochemical performance, the various activators are usually used to prepare biomass carbon materials with developed porosity [30, 31]. However, the common acid/base activators, such as KOH and H3PO4, have significant corrosive properties. Besides, the heavy metal salt activator CuCl2 possesses strong toxicity. Furthermore, the carbon materials activated by these activators usually display a large proportion of narrow microporous characteristics, which limits the transport of large-sized electrolyte ions and weakens the electrochemical performances of carbon-based electrode materials especially at high current densities [32]. Therefore, it is very encouraging to develop mild, low-cost, large-scale, and implementable preparation methods to produce 3D porous carbon with high electrochemical performances [33].

Herein, a novel tamarisk root-based honeycomb-like porous carbon (TRHPC) is successfully prepared by used biomass tamarisk roots as carbon precursor and employed NaCl and ZnCl2 as hybrid activators. The TRHPC with unique interconnected honeycomb-like morphology and abundant mesopore and micropore structure exhibits the high specific capacitance of 293 F g−1 at 0.5 A g−1. In addition, the assembled symmetric supercapacitor based on TRHPC electrodes shows high energy density of 16 Wh kg−1 at a high power density of 160 W kg−1 and good cycling stability with the capacitance retention rate reached 92% after 10,000 charging/discharging cycles. Therefore, the TRHPC with excellent electrochemical performances has a great potential as promising electrode for supercapacitor applications.

Experimental

Materials

Tamarisk roots were obtained from the Minqin County of China. Zinc chloride (ZnCl2) and sodium chloride (NaCl) were purchased from Tianjin Kaitong Chemical Reagent Co., Ltd. All chemical reagents were of analytical grade.

Preparation of TRHPC

Before the experiments, the obtained tamarisk roots were firstly soaked in dilute HCl solution (0.1 M) for 6 h, and then washed with distilled water and freeze-dried for 36 h. After that, the pretreated tamarisk roots (3.0 g) were firstly precarbonized at 550 °C for 2 h at a heating rate of 5 °C min−1 under an atmosphere of N2, and the resulting precarbonized tamarisk root product is named as PC-TR.

The TRHPC materials were synthesized as follows: the PC-TR (1.0 g) and ZnCl2 (2.0 g) were added into the 30 mL NaCl aqueous solutions with different concentrations (1 M, 2 M, and 3 M), followed by magnetic stir for at least 12 h at room temperature. And then, the suspension was further dried in an air-circulating oven at 60 °C to form solid precursor mixtures. Subsequently, the obtained precursor mixtures were activated and carbonized at 800 °C for 2 h at a heating rate of 5 °C min−1 under an atmosphere of N2. After the sample cooling to room temperature naturally, the resulting products were thoroughly washed with HCl (2 M) to remove inorganic salt, and then, the sample was repeatedly washed by distilled water until neutral pH was reached and dried at 60 °C for 24 h. The obtained tamarisk root-based honeycomb-like porous carbon materials were named TRHPC-1, TRHPC-2, and TRHPC-3 based on the used concentrations of NaCl solutions (1 M, 2 M, and 3 M), respectively.

To further explore the influence of NaCl and ZnCl2 for TRHPC, only 30 mL of 2 M NaCl solution or 2 g ZnCl2 dissolve in 30 mL distilled water mixed with PC-TR as comparison experiments were also conducted, respectively. The other preparation process is the same as the TRHPC. The obtained samples were named TRHPC-s (only NaCl treatment) and TRHPC-z (only ZnCl treatment), respectively. In addition, the control sample was also prepared by direct carbonation of PC-TR at 800 °C for 2 h at a heating rate of 5 °C min−1 under an atmosphere of N2, and the as-obtained sample was named TRHPC-c.

Characterization

The surface morphology of the materials was performed by field emission scanning electron microscopy (FE-SEM, Zeiss). The phase structure was characterized by using an X-ray diffractometer (XRD, D8 ADVANCE) equipped with graphite monochromatized Cu Kα radiation. Raman spectra were collected using an inVia Raman spectrometer (Rainie Salt Public Co. Ltd., Britain) with a laser wavelength of 514 nm. The composite was examined by Brunauer–Emmett–Teller (BET) measurements using the surface area analyzer (77 K, autosorb iQ2).

Electrochemical measurement

In a three-electrode system, the electrochemical measurements were performed at room temperature using a computer-controlled CHI660D electrochemical work station with the Hg/HgO (1 M KOH) as reference electrode, a carbon rod as the counter electrode, and platinum carbon electrode as the working electrode in a 2 M KOH solution, respectively. The working electrode was prepared by mixing the as-prepared materials with commercial conductive carbon black and polymer binder (polyvinylidene fluoride) (80:10:10, mass ratio) in N-methyl-2-pyrrolidone (NMP) solution until it forms a homogeneous slurry. The slurry was coated on nickel foam with a working area of 1.0 cm2, and the electrodes were dried at 60 °C for 12 h and then weighted and pressed into sheets under 15 MPa. The total mass of electroactive material was between 3 and 5 mg of each electrode. The electrochemical properties of the electrodes were measured by the cyclic voltammetry (CV), galvanostatic charge–discharge (GCD), and the electrochemical impedance spectroscopy (EIS) tests. The specific capacitance (Cm) value was defined and calculated according to Eq. (1):

$$ {C}_m=\frac{I\times \Delta t}{m\times \Delta v} $$
(1)

where I is the discharge current, ∆t is the discharge time, ∆v represents to the range of charge–discharge voltage, and m is the mass of the active material of the electrode.

For a two-electrode system, the working electrodes were prepared same to the three-electrode system. The two electrodes with identical or very close weight were selected and immersed in 0.5 M Na2SO4 electrolytes for 6 h before being assembled into the symmetric supercapacitor configuration. Then, those electrodes fitted with the separator (thin polypropylene film) and 0.5 M Na2SO4 electrolyte solution were symmetrically assembled into sandwich-type cell construction (electrode/separator/electrode). The energy density and power density of the symmetric supercapacitor were calculated using Eqs. (2) and (3):

$$ E=\frac{1}{7.2}\times {C}_m\times {\Delta v}^2 $$
(2)
$$ P=\frac{E}{\Delta t}\times 3600 $$
(3)

where Cm is specific capacitance and ∆v is the range of charge–discharge voltage.

Results and discussion

Figure 1 exhibits the preparation process of TRHPC. Firstly, tamarisk roots were performed a simple pretreatment and precarbonized at 550 °C under N2 atmosphere. Then, the resulting precarbonized products were mixed with hybrid inorganic salt activators (ZnCl2 and NaCl) and further one-step activation and high-temperature carbonization to obtain the final TRHPC materials. It is reported that the mixed NaCl solution in carbon precursor can provide high-energy Cl−1 ions to etch the carbon structures and form well-developed micro- and mesoporous structures during high-temperature carbonization process [34]. In addition, the high specific surface area and rich porosity generated in carbon materials can be obtained by the ZnCl2 activation approach. Therefore, the tamarisk roots mixed with hybrid activators of the ZnCl2 and NaCl are expected to exploit the novel porous carbon materials with unique morphology and developed porous structure to provide convenient electron/ion transport and diffusion channels for supercapacitor electrode, resulting in excellent electrochemical performance.

Fig. 1
figure 1

The schematic illustration of the synthesis of TRHPC

Full size image

The surface morphologies of as-prepared samples were investigated by FE-SEM. The photographs and FE-SEM images of original tamarisk roots and the corresponding carbon materials are shown in Fig. S1. The results show that the microstructure of the tamarisk roots is not destroyed and maintains good structural stability after high-temperature carbonization. Figure 2a–d gives the FE-SEM images of TRHPC-c, TRHPC-z, TRHPC-s, and TRHPC-2, respectively. It is shown that the TRHPC-c material (Fig. 2a) prepared through direct carbonation of PC-TR presents naturally rich porous and a typical 3D honeycomb-like structure consisting of plenty interconnected networks with a diameter from hundreds of nanometers to several microns. The TRHPC-z (Fig. 2b) prepared only by ZnCl2 treatment displays 3D honeycomb-like structure with some wrinkles in hole wall. The TRHPC-s prepared only by the NaCl treatment shows porous honeycomb-like structure (Fig. 2c), but some hole wall appears to be fractured significantly. As exhibited in Fig. 2d, TRHPC-2 displays the integrated porous honeycomb-like structure with abundant wrinkles in hole wall. The FE-SEM images of other samples (TRHPC-1 and TRHPC-3) prepared similar to TRHPC-2 but different concentrations of NaCl treatment are presented in Fig. S2. Similarly, TRHPC-1 and TRHPC-3 display the integrated porous honeycomb-like structure with some wrinkles in hole wall. The pore size of TRHPC-1 is similar to that of TRHPC-3, while the hole wall of TRHPC-3 is rougher than that of TRHPC-1. Therefore, it can be concluded that the TRHPC with highly crumpled structure in hole wall maintains the interconnected porous honeycomb structure caused by the coactivation of ZnCl2 and NaCl.

Fig. 2
figure 2

FE-SEM images of a TRHPC-c, b TRHPC-z, c TRHPC-s, and d TRHPC-2

Full size image

The XRD pattern of the TRHPC-2 in Fig. 3a exhibits two broad peaks around 23.3° and 42.8° corresponding to the (002) and (100) planes of graphite, respectively, demonstrating the formation of typical graphitic carbon structure [35]. By comparison, the weak and broadening peaks are observed in TRHPC-c, revealing the amorphous nature and low graphitization degree. Figure 3b shows the Raman spectrum of TRHPC-c and TRHPC-2. The two characteristic peaks around 1340 cm−1 and 1596 cm−1 correspond to the D and G bands of carbon. The D band is a typical characteristic of disordered structure carbon, while the G band is a result of a sp2 electronic configuration in the graphitic carbon. Therefore, the relative intensity ratio of the D peak to the G peak (ID/IG) is commonly used to measure the graphitization degree of carbon materials [36, 37]. One can calculate that the ID/IG ratio of TRHPC-c and TRHPC-2 were 1.03 and 0.96, respectively. The result indicated that the TRHPC-2 possesses a higher graphitization degree and a fewer defective nature compared with TRHPC-c, resulting in high electrical conductivity of TRHPC-2 material. Therefore, it can be concluded that the graphitization degree of carbon materials can be effectively improved by the combination of ZnCl2 and NaCl activating treatment.

Fig. 3
figure 3

a XRD pattern of TRHPC-c and TRHPC-2. b Raman spectrum of TRHPC-c and TRHPC-2. c Nitrogen adsorption isotherm. d Pore size distribution curves of TRHPC-c and TRHPC-2

Full size image

It is well accepted that specific surface area and pore structure are two important factors to determine the electrochemical properties of carbon materials. To further study the porosity of TRHPC, the nitrogen adsorption–desorption method was used to determine the BET specific surface area and pore structure of all TRHPC materials. As shown in Fig. 3c, TRHPC-2 and TRHPC-c exhibit typical type IV with a H3 hysteresis loop, suggesting the existence of mesopore structure of those materials [38, 39]. The amount of adsorbed N2 at relative high pressures P/P0 (> 0.90) increase dramatically in the TRHPC-2, demonstrating the presence of large cavities and/or macropore structure [40]. Observably, the adsorbed volume of TRHPC-2 is larger than that of TRHPC-c, meaning large specific surface area and developed pore structure in TRHPC-2. Therefore, it can be concluded that carbonized tamarisk roots integrated by ZnCl2and NaCl activation can improve porosity and specific surface area of carbon materials. Specifically, the BET specific surface area of TRHPC-2 is 484 m2 g−1, which slightly larger than that of TRHPC-c (437 m2 g−1). Besides, the corresponding pore-size distribution curves of TRHPC-2 and TRHPC-c calculated by the Barrett–Joyner–Halenda (BJH) method are shown in Fig. 3d. The TRHPC-2 exhibited high adsorption pore volume and the pore size mainly multilevel distributed in the range of 2–10 nm, while the pore size mainly centered in 12 nm in TRHPC-c. The pore volume of TRHPC-2 is 0.4313 cm3 g−1, which also larger than that of TRHPC-c (0.2866 cm3 g−1). The existence of large pore volume and abundant small mesopore structure in TRHPC-2 can not only guarantee a large electrode/electrolyte interface for electrostatic charge accumulation but also facilitate ion transport by shortening diffusion pathway, which was beneficial to the performance of the supercapacitor [41].

Electrochemical performances of as-prepared materials were firstly investigated using cyclic voltammetry (CV), galvanostatic charge–discharge (GCD), and electrochemical impedance spectroscopy (EIS) analysis in 2 M KOH electrolyte by a three-electrode system. The CV curve of TRHPC-c, TRHPC-z, TRHPC-s, and TRHPC-2 electrodes were performed at scan rate of 50 mV s−1, as shown in Fig. 4a. Observably, the curve of TRHPC-2 exhibits the rectangular-like shape and larger CV curve area than others, suggesting fast electrochemical response and large specific capacitance of TRHPC-2, which may result from the integrated porous honeycomb structure with abundant wrinkles in hole wall. In contrast, the other materials exhibit small irregular rectangular curve and even change to triangular tendency at high potentials, corresponding to low capacitances due to its pore dimensions are probably too small to allow ions to diffuse easily. Remarkably, the CV profiles of TRHPC-2 remain quasi-rectangular shape even when the scan rate increases from 10 to 100 mV s−1 (Fig. 4b), indicating the ideal capacitance behavior. The GCD tests of TRHPC-2 were carried out at various current densities ranging from 0.5 to 5 A g−1, as shown in Fig. 4c. The nearly symmetric and linear GCD curves confirm the reversible charging–discharging processes, which are consistent with the CV results. Besides, the comparison of GCD curves of different samples is given in Fig. S3 at a current density of 0.5 A g−1. As shown in Fig. S3, GCD curves of TRHPC-1, TRHPC-2, TRHPC-3, TRHPC-c, TRHPC-s, and TRHPC-z are clearly observed at different current densities of 0.5 to 5 A g−1. Additionally, the specific capacitances of electrode materials calculated from GCD curves at different current densities for all electrodes are shown in Fig. 4d. It can be seen that the specific capacitances of these electrode materials decreased gradually when the current densities increase. The specific capacitance of TRHPC-2 can be reached up to 293 F g−1 at 0.5 A g−1, which value is larger than that of TRHPC-c (126 F g−1), TRHPC-z (165 F g−1) and TRHPC-s (142 F g−1) under same current density. It was surprising that the specific capacitance of TRHPC-2 remains 195 F g−1 even at a super high current density of 30 A g−1 (about 66.6% of the capacitance retention), demonstrating remarkable rate performance. The specific capacitances of TRHPC-1, TRHPC-2, and TRHPC-3 are shown in Fig. S5. Furthermore, TRHPC-2 displayed capacitance retention of 90.2% after 10,000 cycling operation at current density of 5 A g−1 (Fig. S6).

Fig. 4
figure 4

a CV curves of TRHPC-2, TRHPC-s, TRHPC-z, and TRHPC-c at scan rate of 50 mV s−1. b CV curves of TRHPC-2 at different scan rates. c GCD curves of TRHPC-2 at different current densities. d Specific capacitance of different electrodes at different current densities

Full size image

To investigate the practical application of electrode materials, the two-electrode symmetric supercapacitor (SSC) device based on TRHPC-2 electrode was also assembled, and its electrochemical performance was further investigated. It was reported that the neutral aqueous electrolyte (such as Na2SO4 and Li2SO4) can achieve a higher working voltage than acid or alkali electrolytes, because the neutral electrolyte has a low concentration of hydrogen and hydroxide ions [42]. Therefore, the symmetric supercapacitor based on TRHPC-2 material was fabricated and measured in 0.5 M Na2SO4 electrolyte. As shown in Fig. 5a, the CV curves of as-assembled TRHPC-2//TRHPC-2 SSC were performed at operating voltage of 1.6 V at different scan rates (10~50 mV s−1). Moreover, the CV curves of the symmetric supercapacitor can still keep a stable shape even at high scan rate of 50 mV s−1, indicating fast ion transportation and good rate capability. Figure 5b displays the GCD curves of TRHPC-2//TRHPC-2 SSC at different current densities. All GCD curves show approximately the shape of an isosceles triangle, indicating that the device possesses excellent electrochemical reversibility.

Fig. 5
figure 5

Electrochemical performance of symmetric supercapacitor measured in a two-electrode system in 0.5 M Na2SO4 electrolyte. a CV curves of symmetric supercapacitor at different scan rates. b GCD curves of symmetric supercapacitor at various current densities. c Ragone plot of the symmetric supercapacitor. d Cycling performance of symmetric supercapacitor at a current density of 5 A g−1

Full size image

The Ragone plot (a plot of energy density vs. power density) of SSC calculated from GCD curves at different current densities is presented in Fig. 5c. It can be found that the energy density of TRHPC-2//TRHPC-2 SSC can be achieved to 16 Wh kg−1 at power density of 160 W kg−1 and remained as 2.92 Wh kg−1 at 2430 W kg−1, which value is much higher than those of previously reported carbon-based aqueous symmetric supercapacitor [43,44,45,46,47]. Figure S7 shows the Nyquist plots and equivalent circuit diagram of the symmetric supercapacitor. The device shows a small x-axis intercept and vertical line in the high-frequency and low-frequency regions, respectively, indicating a small diffusion resistance and an ideal capacitive behavior. The EIS data were fitted with ZSimp software to the equivalent circuit model (inset in Fig. S7). Rs, Rct, Cdl, ZW, Q, and CL represent the internal series resistance, charge transference resistance, electrochemical double-layer capacitance, Warburg impedance, constant phase element, and inductance capacitance, respectively. In terms of the EIS results, the TRHPC-2//TRHPC-2SSC has the low internal resistance (Rs) of 1.65 Ω and low charge transfer resistance (Rct) of 1.39 Ω. Moreover, the TRHPC-2//TRHPC-2 SSC exhibits an outstanding long-term electrochemical stability with 92% of the maximum capacitance remains after 10,000 cycles (Fig. 5d). The remarkable cycling stability can be attributed to the unique 3D honeycomb-like porous structure, which can accelerate diffusion/transport and alleviate the volume changes during the charge–discharge process to guarantee good stability.

Conclusion

In summary, we demonstrated a simple and effective route by using tamarisk roots as carbon precursor and mixed NaCl and ZnCl2 as hybrid activators to achieve the tamarisk root-based honeycomb-like porous carbon (TRHPC). The TRHPC displays the integrated porous honeycomb-like structure with abundant wrinkles in hole wall. Because of its unique structure, the TRHPC used as electrode for supercapacitors shows a high specific capacitance of 293 F g−1 at 0.5 Ag−1 and excellent rate performance. Furthermore, a symmetric supercapacitor is assembled based on TRHPC-2 electrodes, delivering a maximum energy density of 16 Whkg−1 at the power density of 160 W kg−1 and remarkable cycling stability with 92% of the initial capacitance retention after 10,000 cycles. Besides, the impressive electrochemical performances are highly comparable or superior to that of recently reported advanced porous carbons and commercially available activated carbons. Most importantly, the present synthesis strategy can avoid the use of corrosive and/or heavy metal salt activated reagents. The proposed novel method is not only cost-effective and environment-friendly but also suitable for large-scale production of porous carbon materials.