1 Introduction

Over the past years, supercapacitors hold the promise of energy storage in various applications due to their long life span and high power density [1]. These properties suggest a range of applications of supercapacitors which include high-power delivery system, backup power sources in electric vehicles, electronic devices, few to mention [2]. However, their applications are hampered by their low energy densities, and thus, researchers are trying to improve the energy density by increasing specific capacitances.

Energy storage in supercapacitors can be characterized as electric double-layer energy storage mechanism or pseudocapacitance mechanism. The electric double-layer energy storage mechanism requires high surface area and well-structured pores for high specific energy storage. However, developing high surface area and highly porous carbon remains a challenge. To defeat the aforesaid challenge, efforts have been made by synthesizing porous carbon from biomasses, especially agriculture wastes such as bamboo shoot shells, biogas slurry, orange peels, concorb and others [3,4,5,6,7,8,9]. The interest in biomass-based supercapacitor electrodes is catalyzed by the exciting properties (long life cycle, low cost, deformability and high specific capacitance) [10]. The focus has been synthesizing porous materials with high surface area, interconnected pores and pore sizes that will allow the formation of carbon/electrolyte double-layer interface for increasing specific capacitance.

Biomass-derived porous carbon materials have been reported to be synthesized through hydrothermal carbonization [11,12,13], direct carbonization followed by activation [14,15,16,17,18], soft and hard template carbonization [19,20,21] and molten salt carbonization [22,23,24]. Of the aforementioned routes, direct carbonization followed by activation has gained more attention because it increases the surface area of the porous carbon and easily adjusts the mesopores to micropores which plays an important role in energy storage.

Vitex mombassae plants are well grown in woodland, especially in Tanzania Zimbabwe and South Africa. Vitex mombassae produces fruits which are eaten as wild fruits [25], and they are used to prepare jam; however, their seeds are discarded as wastes. This study utilized the vitex mombassae seed waste as carbon precursor for synthesizing high surface area nanomaterials. The high surface area, microstructured and highly pore textured porous materials are obtained through varying activation temperature from 600 to 800 °C. The fabricated electrodes demonstrate a high gravimetric capacitance of about 437 F g−1 at a scan rate of 5 mV s−1 in 6 M KOH electrolyte.

2 Experimental sections

2.1 Materials

Vitex mombassae (VM) fruits wastes were collected from Iringa Tanzania. Potassium hydroxide and hydrochloric acids were purchased from Sigma-Aldrich. Polyvinylidene difluoride (PVDF), conducting carbon and N-methyl pyrrolidone were purchased from MTI Corporation. All chemicals were used without further purification.

2.2 Synthesis of activated carbon

The VM seed wastes were collected in large amount, washed and dried in air for 1 week. Dried VM seed wastes were then crushed into small particles of around 160 µm. The crushed VM seed waste particles were put into alumina boat crucible and carbonized in horizontal tube furnace (CTF12/65/550) from Carbolite, UK, at 450 °C for 1 h in nitrogen atmosphere at a heating rate of 5 °C/min. The carbonized VM seeds were further activated with KOH at carbon-to-KOH mass ratio of 1:3 which has been reported to be optimal mixing ratio for many biomass carbon precursors [4, 26]. The mixtures were activated at 600, 700 and 800 °C for 2 h in nitrogen atmosphere at the heating rate of 5 °C min−1. The tube furnace was left to cool naturally to room temperature, and then, activated carbon was taken out. The final products were then washed with 1 M HCl and sufficient distilled water in order to remove other organics. The samples were designated as VM-T with T representing the activation temperature, whereas the unactivated sample was designated as VM.

2.3 Characterizations of activated carbon

FLASH 2000 Organic Elemental Analyzer, Thermo Scientific, was used to study the bulk elemental compositions of the samples. Sorption studies of the samples were performed using Quantachrome Autosorb iQ instrument using nitrogen gas sorption at 77.35 K. All samples were degassing at 200 °C for 5 h prior to the adsorption experiments in order to eliminate the surface gaseous contaminants. The BET surface area was calculated using the Brunauer–Emmet–Teller (BET) equation. The external surface area, micropore surface area and pore volume were determined through t plot method. The pore size distributions were calculated from nitrogen adsorption branch using nonlocal density functional theory (NLDFT) adsorption branch model. Scanning electron microscope (FEI Nova NanoSEM 450) with an accelerating voltage of 15 kV having a beam current of 2.0 mA, magnification capability of 100,000 and working distance of 14,800 μm was used to observe the microscopic surface morphology of the samples. The X-ray spectrophotometer spectra (XPS) were studied through PHI model Quantum 2000 spectrometer using Al Kα radiation as the radiation source with wavelength (λ = 1486.6 eV). The microstructure of the materials was obtained using high-resolution transition electron microscope (Talos F200X microscope) operated at voltage of 200 kV. The diffraction pattern of the materials was identified through powder X-ray diffractometer (PXRD) performed on Rigaku Ultima IV diffractometer (Cu Kα radiation, λ = 1.5406 Å) with acceleration voltage and current of 40 kV and 44 mA, respectively. The diffraction patterns were obtained in the step scanning of 0.02 (2θ°) in the range of 5 < (2θ°) < 75.

2.4 Electrode preparation and electrochemical measurements

The electrochemical measurements were performed using PGSTAT204 Autolab workstation (Metrohm). The working electrodes were prepared by mixing VM porous carbon, conducting carbon and polyvinylidene fluoride (PVDF) (in N-methylpyrrolidone) in the weight ratio of 80:10:10. The working electrodes were pressed in a nickel foam and dried at 100 °C for 10 h. Ag/AgCl and platinum wire were used as reference and counter electrodes, respectively. Cyclic voltammetry (CV) measurements were conducted in the voltage range of − 1.0 to 0.0 V at a scan rate of 5, to 200 mV s−1. Electrochemical impedance spectroscopy (EIS) measurements were taken in a frequency range of 1 mHz to 1 MHz. The specific capacitance was calculated from CV using Eq. (1):

$$C = \frac{1}{{2mv(E_{2} - E_{1} )}}\int\limits_{{E_{1} }}^{{E_{2} }} {i(E){\text{d}}E}$$
(1)

where (E2 − E1) is the applied potential window, m (g) is the weight of the active material, v (mV s−1) is the scan rate and \(\int_{V1}^{V2} {i(V){\text{d}}V}\) is the total voltammetric charge obtained by integration of negative and positive sweeps in cyclic voltammograms.

3 Results and discussion

3.1 Elemental composition

In order to qualify VM for supercapacitors application, bulk elemental compositions of the raw and carbonized samples were studied using CHNOs elemental analyzer. It can be seen in Table 1 that the quantities of carbon are 45%, 75%, 87% 84% and 67% for raw (uncarbonized), VM, VM600, VM700 and VM800, respectively. The carbon content in the samples decreased with activation temperature from 87% at 600 °C to 67% at 800 °C. The percentage of carbon in the samples qualifies VM for supercapacitor application.

Table 1 Bulk elemental compositions of the samples
Full size table

Sorption studies of the prepared samples as done through N2 adsorption and desorption are shown in Fig. 1. The unactivated sample clearly shows type IV isotherms which demonstrate the presence of mesopores (Fig. 1a). As the activation temperature increased, the isotherms of VM600, VM700 and VM800 demonstrated overwhelmingly microporous textures with the coexistence of mesopores with hysteresis loop type IV (Fig. 1b–d). Figure 2 shows the pore size distribution of the synthesized samples. From Fig. 2a, b, it is observed that all samples (except VM) exhibited a combination of mesopores and micropores. As shown in Fig. 2b, the micropores are centered at 1.5 and 2 nm while the mesopores are centered between 2 and 4.5 as well as between 30 and 40 nm as evidenced in Fig. 2a, b, respectively. The combination of micropores and mesopores in the synthesized porous materials is beneficial to energy storage as it is accepted that the mesoporous structure can improve the charge transfer while micropores have high surface area which in turn offer more active sites for charge adsorption and hence increase specific capacitance.

Fig. 1
figure 1

Nitrogen adsorption–desorption curves of a VM, b VM600, c VM700, d VM800

Full size image
Fig. 2
figure 2

Pore size distribution: a pore width 0–80 nm, b pore width 0–7 nm

Full size image

The influence of activation temperature on BET, micropore and mesopore surface areas, pore size and total pore volume is depicted in Table 2. The specific surface area as calculated by Brauner–Emmet–Teller (BET) method was found to be 14, 1253, 2044 and 1622 m2 g−1 for VM, VM600, VM700 and VM800, respectively. It can be noted that increasing activation temperature from 600 to 700 °C, the BET surface area was increased by 790 m2 g−1. When the activation temperature was further altered to 800 °C, the BET surface area was decreased by 422 m2 g−1. It can be stretched that as the activation temperature was increased from 600 to 700, the reaction rate between C and KOH increased which in turn increased carbon burn-off. Similarly, the devolatilization process of the carbon chars increased with activation temperature to produce more pores from the carbon chars. Carbon burn-off and evolvement of volatiles increase BET surface area. However, at high activation temperature (800 °C), the BET surface area decreased because of excessive carbon burn-off which resulted into pores widening and collapse of the pores and loss of walls, hence limiting the BET surface area [27]. The micropore surface area (Smicro) increased with activating temperature and reached its optimal value at 700 °C and then decreased at higher temperature 800 °C. On the other hand, mesopore surface area (Smeso) decreased with increasing activation temperature and increased at higher temperature as depicted in Table 2. The contribution of micropore surface area in percentage (Smicro/BET) was 17, 78, 95 and 72% for VM, VM600, VM700 and VM800, respectively. VM700 exhibited high Smicro 1945.8 m2 g−1 and up to 95% micropore surface area and 89% micropore volume. It can be seen further that the total pore volume increased with activation temperature. At 600 °C activation temperature, the mesopore and micropore volume was 50% each. The ratio between micropore surface area and mesopore surface area (Smicro/Smeso) corresponds well with the ratio between micropore to mesopore volumes (Vmicro/Vmeso) which is the highest at activating temperature of 700 °C.

Table 2 Textural properties of synthesized samples
Full size table

The morphologies of the synthesized porous materials derived from VMs are shown in Fig. 3. It can be seen that the morphologies are dependent on the activation temperature. At 700 °C and 800 °C, the morphologies of the samples acquired the cylindrical structures of the pores. The developed pore structures in Fig. 3b–d may have resulted due to introduction of KOH which break down the cellulosic and lignin structure of the VM seeds into microstructures by burning more carbon to produce pores. For VM600–VM700, the pores developed are vividly shown to be qualifying the materials for energy storage usage. The formation of pores was further confirmed in TEM image (Fig. 4a–b). The existence of these pores is important for charge transfer in energy storage devices.

Fig. 3
figure 3

SEM micrographs of a VM, b VM600, c VM700 and d VM800

Full size image
Fig. 4
figure 4

TEM images of a VM and b VM 700

Full size image

Figure 5a shows XRD patterns for VM and VM700 as representative samples. It can be seen that all samples show two broad peaks at around 2θ = 24° and 42° indexed to (200) and (100) plane of reflection, respectively. The peak at 2θ = 24° is more pronounced in VM than VM700 while at 2θ = 42, VM700 had more pronounced peak than VM. The presence of strong background at low angle for the synthesized samples signifies that the rich micropores were embedded in the carbon framework which in turn disturbed the lattice periodicity, thus reducing the peak of the plane indexed to (200) [28, 29]. The structure of the VM700 as representative sample was further investigated through Raman analysis. Figure 5b depicts two carbon materials prominent peaks located at 1348 and 1602 cm−1 which correspond to D and G bands. It is worth to note that D peak is due to disordered carbon structure while G band represents the sp2 hybridization graphitic carbon [30]. The ratio of intensity of the D band to that of G band (ID/IG) of VM700 was about 0.84 which indicates the graphitization of the synthesized sample.

Fig. 5
figure 5

a XRD patterns of VM and VM700, b Raman spectrum of VM700

Full size image

As shown in Fig. 6, X-ray photoelectron spectroscopy was used to investigate the surface chemistry of the synthesized materials. XPS survey spectra of VM and VM700 as representative samples are shown in Fig. 6a. It can be pointed out the samples contained C1s, O1s CKLL and O KLL core level peaks with main peaks at 284, 400 and 532 eV which correspond to multiple coexistence of C O and N, respectively [31]. One can see that carbon is the main component in the materials followed by oxygen with small content of N in VM; however, the content of N was negligible for VM700, indicating that the KOH activation influences the N content. The presence of oxygen groups and nitrogen groups in the synthesized samples confirms that the porous materials were self-doped. Furthermore, these results are consistent well with CHNOs results. C1s spectra for VM and VM700 were deconvoluted into two main peaks at 284.8 and 286.8 eV for epoxy and alkoxy, respectively. The O1s was further deconvoluted into two oxygen-containing functional groups. These groups include C=O (531 eV) and C–O (532.1 eV) for VM sample while VM700 was deconvoluted into 533 eV, 535.4 eV. N1s spectra of VM were deconvoluted into only one type (a pyridinic) which corresponded to 399.1 eV. It is assumed that pyridinic nitrogen species contribute to pseudocapacitance [14, 32].

Fig. 6
figure 6

XPS results. a Survey spectra for VM and VM700, b C1s deconvolution of VM, c C1s deconvolution of VM700, d O1s deconvolution of VM, e O1s deconvolution of VM700 and N1s deconvolution of VM

Full size image

3.2 Electrochemical performances

In order to assess the effectiveness of the synthesized materials for supercapacitor electrodes cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS) were carried out. Figure 7 shows the voltammograms of the fabricated electrodes VM, VM600, VM700 and VM800. It can be seen that all voltammograms acquired the quasirectangular shape in the voltage window of − 1 to 0 V. This is an indication that the fabricated electrodes had EDL characteristics. The area under the CV curves increased with activation temperature Fig. 7b, c and then dropped at high temperature (Fig. 7d). There were some small eversible bumps in the CV curves (Fig. 7a–d), and these may have been attributed by the presence of heteroatoms functional groups in the synthesized porous materials as confirmed by XPS results.

Fig. 7
figure 7

Cyclic voltammograms of a VM, b VM600, c VM700, d VM800

Full size image

Figure 8a shows the specific capacitance of the electrode as calculated from CV curves. It is shown the activation temperature leads to dramatic increase in specific capacitance from 118 to 402, 437 and 377 F g−1 for VM, VM600, VM700 and VM800, respectively, at a scan rate of 5 mV s−1. VM700 attained the highest specific capacitance which could have been attributed to its high BET surface area as well as high micropore content. The higher surface area creates large sites to be accessed by ions. The specific capacitance obtained in this study (437 F g−1) was higher than that for mesoporous carbon derived from fullerene crystals (286 F g−1) [33], Metaplexis (256.5 F g−1) [34], biogas slurry (289 F g−1) [4] and coconut shell (228 F g−1) [35] at a scan rate of 5 mV s−1. The results presented in Fig. 8b suggest that there is close relationship between the BET surface area, (Smicro/Smeso), (Vmicro/Vmeso) and specific capacitance exhibited by the electrodes. One can see in Fig. 8b that VM had the highest (Smicro/Smeso), (Vmicro/Vmeso), BET surface area and so was the specific capacitance implying that pore textures influence the electrochemical performances of electrodes.

Fig. 8
figure 8

a Specific capacitances, b relationship between pore textures and specific capacitance, c Nyquist plots, d cyclic stability

Full size image

Figure 8c shows the electrochemical impedance spectroscopy (EIS) of fabricated electrodes. It is revealed from Nyquist plot that all the synthesized carbon materials exhibited high ionic conductivity and the intercepts with horizontal axis of VM 600, VM700 and VM800 electrodes were between 2.5 and 3.5 Ω (see the insert of Fig. 8c). The internal resistances were 0.29, 0.31 and 0.32 Ω for VM600, VM700 and VM800, respectively. The smaller semicircle with vertical line at the lower frequency existed for all samples which implied that the electrodes had low charge transfer resistance at the electrode/electrolyte interface; thus, there was fast electrolyte ion diffusion into mesopores. Additionally, the supercapacitor electrode (VM700) demonstrates high cyclic stability with capacitance retention of over 90% after 50,000 cycles which qualifies the materials for supercapacitor application.

4 Conclusion

In summary, highly porous carbon nanomaterials were synthesized by carbonizing Vitex mombassae (VM) seeds followed by subsequent activation. The effect of activation temperature (600–800 C) on the textural microstructure and specific capacitances was studied. Highly porous carbon materials were synthesized with high specific surface area. The specific surface area increased with activation temperature and then dropped at higher temperature. On the other hand, the total pore volume increased with activation temperature. The resulting VM seeds possess unique porous surface structure with pore volume increasing from 0.8 to 1.1 cm3 g−1. Moreover, the optimized VM700 have a high BET specific area (2043.8 m2 g−1), high micropore surface area (95%) and micropore pore volume (89%) which facilitates the electrochemical double-layer capacitance. Due to these advantages, the VM700 exhibit a high specific capacitance of 437 F g−1 at a scan rate of 5 mV s−1 in 6 M KOH electrolyte. More importantly, the electrodes exhibited over 90% capacity retention after 50,000 cycles, indicating that VM-based porous materials are promising electrode materials for supercapacitors.