npsm 새물리 New Physics : Sae Mulli

pISSN 0374-4914 eISSN 2289-0041


Research Paper

New Phys.: Sae Mulli 2024; 74: 432-439

Published online May 31, 2024

Copyright © New Physics: Sae Mulli.

Silicon Carbon Nanotube Composite for Supercapacitor Applications

Mangesh Subedi1, Hae Kyung Jeong1,2*

1Department of Physics, Daegu University, Gyeongsan 38453, Korea
2Department of Energy System Engineering, Daegu University, Gyeongsan 38453, Korea

Correspondence to:*

Received: December 23, 2023; Revised: February 20, 2024; Accepted: March 27, 2024

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License( which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

Supercapacitor finds appealing uses as an electrical energy storage device in consumer electronics and as an alternative power source due to its quick discharge/charge times and long-term operation stability. This study examines silicon and CNT (Si-CNT) composite, synthesized by a simple and easy method, for supercapacitor applications. The supercapacitor electrodes active material was a composite of Si and CNT. It was made by magnetic stirring, mixing, ultrasonication, vacuum filtration, and drying process of the powdered Si and CNT with a binder. The final composite shows the specific capacitance of 121.3 F/g, energy density of 10.86 Wh/kg, and power density of 403.2 W/kg at 1 A/g. It demonstrates that the composite material could boost the supercapacitor efficiency.

Keywords: Carbon nanotubes, Supercapacitors, Silicon, Composite

In recognition of their swift charge/discharge rate, high power density, and exceptional cycle performance, supercapacitors-also known as electrochemical capacitors, are starting to gain popularity in the energy storage sector[1-3]. The primary functions of a supercapacitor remain the same: to store energy between a solid electrode and an electrolyte. Supercapacitors have a higher power density than batteries and fuel cells, but their energy storage capacity is lower. However, their energy density must be increased to comply with the constantly rising demand for electronic devices, such as electric cars and portable electronics[1, 4, 5]. Supercapacitors are categorized as electric double-layer capacitors (EDLCs) or faradaic pseudo-capacitors based on their charge storage mechanism. EDLCs store energy through ion adsorption and desorption at the electrode-electrolyte interface[6, 7]. However, pseudo capacitors store electric charges through a rapid redox reaction at the electrode surface[2, 6].

Silicon exhibits great potential as an electrode material for supercapacitors and batteries owing to its high estimated capacity of 4200 mAh/g, which is tenfold greater than that of graphite-based anode materials[1, 4, 7]. Silicon is an affordable and non-toxic electrode material that can be utilized in energy storage devices. Numerous research have been conducted to broaden and expand the application of Si-based materials and their derivatives, such as silicon-carbon nanofiber, silicon- graphite, silicon- carbon, silicon nanosphere-CNT, and silicon nanowire-carbon, as electrode materials in portable devices to enhance the electrochemical performance of supercapacitor[1, 6]. The exceptional electrical conductivity and advantageous surface area of many Si-based materials and their derivatives have drawn scientific attention.

The superiority of carbon nanotubes (CNTs) over graphite as a matrix material for supercapacitor has been demonstrated, owing to their electrical conductivity, greater mechanical strength, large surface area, and structural flexibility[1, 8, 9]. The higher reversible capacity of CNTs than graphite shows the potential replacement of graphite-based supercapacitor. CNT has a higher surface area, a lower mass density, excellent electrical conductivity, more sophisticated synthesis techniques, and a lower cost than other materials, making it capable of meeting all supercapacitor electrode requirements[2]. The interface between Si and CNTs determines the electrochemical properties of the composite. Recent studies have demonstrated that the intimate contact between Si nanoparticles and CNTs facilitates efficient electron transfer and increases the stability of the electrode during cycling. Additionally, the presence of CNTs can alleviate the volume expansion of Si during lithiation, mitigating electrode degradation and improving cycling stability[1, 2, 6]. Polystyrene sulfonate (PSS) is an excellent binder for Si particles on CNT surfaces due to its long-term stability, electrode compatibility, and flexibility, which allow for a strong bond between Si and CNT particles, maximizing active material utilization[10].

In the present work, Si and CNT composites are made with commercially available silicon powder and CNTs synthesized by ultrasonication, stirring, drying, and vacuum filtration by physical mixing with PSS. Si-CNTs composite material shows a high specific capacity of 121.3 F/g, 10.86 Wh/kg energy density, and power density of 403.2 W/kg at 1 A/g. Si-CNT has 90.7% capacitive retention, and 84.8% coulombic efficiency is obtained after 3,000 cycles. It demonstrates that the composite material could boost the supercapacitor efficiency, rechargeability, stability and high reversibility’s of electrochemical performance of supercapacitors.

1. Reagent and materials

Silicon (Si, powder, 325 mesh, 99%), carbon nanotubes (CNT, powder ≥ 98%), polystyrene sulfonate (PSS, powder), and sodium hydroxide (KOH) were purchased from Merck. Samchun Pure Chemical Co., Ltd. supplied the ethanol (Seoul, Republic of Korea). Reagents have been used without being purified further. The aqueous solutions employed in the experimental procedures were prepared utilizing deionized (DI) water.

2. Preparation of Si-CNT composite

1 g of silicon powder and 0.125 g of CNT were mixed and distributed in 30 ml of ethyl alcohol under magnetic stirring for 1 h. Subsequently, the mixture was ultrasonically suspended for 2 h to achieve a well dispersed Si-CNT mixture. On the other hand, to obtain a PSS solution, 0.125 g of PSS and 30 ml of deionized water were combined and stirred magnetically for 1 h. Following that, a highly stable Si-CNT dispersion was then achieved by adding a PSS solution to the mixture and magnetically stirring it for an hour. Afterward, the Si-CNT composite was then formed by vacuum filtration and dehydrated in a vacuum oven at 60 C for 24 h. The Si-CNT composite was applied to bath sonication for 1 h, preceded by vacuum filtration. The mass ratio of the Samples Si: CNT: PSS = 8:1:1, was used to prepare the composites to boost the electrochemical properties of the finalized materials.

3. Physicochemical characterization

Field emission scanning electron microscopy (FESEM, S-4300, Hitachi, Japan) images were utilized for the morphological studies of the samples, while energy dispersive X-ray spectroscopy (EDS, S-4300, JEOL) was applied for elemental composition and mapping. For the EDS mapping, the acceleration voltage was 20 kV. The materials' phase and vibration modes were evaluated using Raman spectroscopy (NANOBASE 100X XperRAM C, Korea, with a laser excitation of 532 nm).

4. Electrochemical characterization

Using Bio-Logic SP-150 (France), the electrochemical response of the materials was examined at room temperature. Ag/AgCl (3 M KCl saturated) served as the reference electrode, and platinum wire served as the counter electrode in 6 M KOH electrolyte[11]. The fabrication of the working electrodes involved the following steps: 5 mg each sample (Si, CNT, Si-CNT) was individually dispersed in 2 ml of isopropyl alcohol, and subjected to 5 μL of sonicated homogenous suspension was drop placed on the GCE (glassy carbon electrode, OD: 6 mm, ID: 3 mm) for the preparation of the working electrodes. The loaded active mass was 2.5 mg/m2 in the working electrodes. As the electrolyte, an aqueous 6 M KOH solution was used.

Galvanostatic charge-discharge (GCD), chrono coulometry (CC), cyclic voltammetry (CV), and potential electrochemical impedance spectroscopy (PEIS) were used to assess the samples' electrochemical characteristics[11-15]. To examine the reversibility and the capacitive response of the sample’s electrochemical reactions, CV measurements are performed at sampling rates of 200, 100, 50, 25, and 10 mV/s using a potential window of 0–0.8 V. CC measurements were used to determine the samples' electrochemically active surface areas[16, 17], as previously mentioned. To investigate the electrochemical impedance characteristics, PEIS was conducted in the frequency between 100 mHz and 500 kHz at an amplitude of 5 mV at open-circuit voltage. Supercapacitors were tested using GCD at current densities varying from 0.25–5 A/g and potential ranges ranging from 0–0.8 V to figure out their performance.

Figure 1 exhibits the synthesis process of Si-CNT. The morphological and microstructural characterization of Si, CNT, and Si-CNT composites was studied by SEM. Figure 2(a–c) illustrates the SEM image of silicon powder with a wide size variation and irregular forms for different magnifications. The surface of the CNT in Fig. 2(d–f) was highly smooth, tubular, and straight, but after the composition, Si nano particles are decorated on the surface of CNT in the Fig. 2(g–i). Furthermore, it is evident that there are isolated silicon particles in some areas that are either linked to or incorporated in the CNT. In the Si-CNT composite, the diameter of the CNT is both larger and broader than in the precursor, resulting in a good distribution of the silicon particles between the CNTs or on the surface. Figure 3 shows the composition and presence of carbon (C), oxygen (O), and silicon (Si) in the Si-CNT composite, confirming that the silicon is dispersed uniformly across the whole surface of the CNT.

Figure 1. (Color online) Schematic representation of the Si-CNT composite.

Figure 2. SEM images of the samples: Si (a–c), CNT (d–f), and Si-CNT (g–i).

Figure 3. (Color online) EDS analysis of Si-CNT with its composition elements and elemental spectrum.

The Raman spectra results of Si, CNT, and Si-CNT appear in Fig. 4. Raman spectroscopy is a more efficient method of analyzing carbon-based materials. Due to the crystalline silicon phonon scatterings, Pure Si has two peaks around 511 cm-1 and 945 cm-1[6]. Furthermore, G (graphitic carbon) and D (disordered carbon) bands were observed in the Si-CNT composite’s Raman spectra. Due to the strong scattering of Si-CNTs, a small peak around 511 cm-1 appears in the entire Raman spectrum of the composite. The most prominent CNT’s Raman features are the G (graphite) and D (disordered) bands, which are clearly defined in the spectrum range of 1200–1800 cm-1, with the G band (graphitic carbon) centered about 1571 cm-1, while the D band (disordered carbon) concentrated at 1350 cm-1[5, 18-20]. There is a decrease in peak intensities of Si peak with D, G, and 2D in Si-CNT composite than in Si and CNT, suggesting a uniform distribution of the Si, which supports the SEM results.

Figure 4. (Color online) Raman Spectra of Si, CNT, and the Si-CNT composite.

The experiment used a half-cell in a 6 M KOH electrolyte to assess the electrochemical properties Comparing the cyclic voltammetry (CV) between Si (Fig. 5(a)), CNT (Fig. 5(b)), Si-CNT (Fig. 5(c)) and PSS (Fig. 5(d)), Si-CNT shows the highest current density and increased current-carrying capability (Fig. 5(c)). At 50 mV/s, the Si-CNT exhibits a greater cyclic voltammetry (CV) current compared to other conditions, while the Si, CNT, and PSS lead to a lower integrated area under the CV curves with the potential range from 0–0.8 V. There is no notable peak in the curve for CNT, and the rectangular CV curve represents the ideal double-layer capacitor characteristic with a charge/discharge process. Furthermore, there is a quasi-rectangular shape in the Si-CNT, indicating the existence of an electric double-layer capacitance of CNT with silicon, suggesting excellent non-Faradaic capacitive behavior[21]. Figure 5(e) compares the CV results of all the sample at a scan rate of 100 mV/s. Figure 5(f) displays all the sample’s CV results at the different scan rates, which range from 200 mV/s to 10 mV/s. The appropriate specific capacitance (Cs) from the CV data could be approximated in F/g, as described elsewhere[13, 20, 22]. Figure 5(f) shows the specific capacitance that results as a proportion of the scan rate, with the Si-CNT sample significantly exhibiting the highest capacitance across all scan rates. As the scan rate increases, the Cs gradually decrease. At 50 mV/s, the Si-CNT sample has the highest capacitance of 122.5 F/g, which is nearly two times the capacitance of Si (66.38 F/g). The capacitance of CNT was also found to be 81.25 F/g. Because of high specific capacitance, the specified Si-CNT composite will boost the electrode surface area and enhance the excellent properties for supercapacitor applications. The specific capacitance of PSS is the lowest; its value is 31.25 F/g.

Figure 5. (Color online) CV results of (a) Si, (b) CNT, (c) Si-CNT, (d) PSS, with (e) comparing CV of PSS, Si, CNT, and Si-CNT at 100 mV/s, and (f) the corresponding capacitance at different scan rate.

Electrochemically active surface areas were calculated using CC values[23]. Figure 6(a) shows how the slope relates to the surface area. The electrochemically active surface areas were 3.09 ×10-5 cm2, 3.3 ×10-5 cm2, and 4.5 ×10-5 cm2 for the Si, CNT, and Si-CNT samples, orderly. Based on these observations, the Si-CNT sample's surface area is roughly 1.5 times larger than the Si samples. Along with the electrodes' conductivity characteristics, PEIS was used to investigate the rate of charge transfer along the electrode-electrolyte interface. Figure 6(b) shows a Nyquist diagram consisting of a semicircle and line along with an equivalent circuit (inset) that includes the Warburg component (W), double layer capacitance (CDL), internal resistance (Rs), and charge transfer resistance (RCT). The charge transfer resistance (RCT) of Si-CNT electrode is represented by the diameter of the depressed semicircle. Using Z-tuning software on the modified Randal equivalent circuit, the RCT was 44.32 Ω for Si, 29.61 Ω for CNT, and 23.4 Ω for Si-CNT. The Si-CNT has a significantly lower charge transfer resistance owing to enhanced conductivity following CNT insertion in Si, which agrees with the CV and CC results.

Figure 6. (Color online) (a) PEIS results and (b) CC results.

The capacitive overall performance of the samples was, in addition, further investigated by galvanostatic charge/discharge (GCD) curves. Figure 7(a–c) displays the GCD curves of Si, CNT, and Si-CNT, respectively at distinctive current densities that vary from 5 A/g to 0.25 A/g. When the suggested potential (0.8 V) was reached, a discharge current was passed through the electrodes until the potential was zero. The analysis of discharge curves can be used to estimate the sample’s specific capacitance, yielding specific capacitances of 66.25, 81.25, and 121.25 F/g obtained for Si, CNT, and Si-CNT, respectively at 1 A/g. As expected, Si-CNT exhibits the maximum capacitance at the identical current density. The calculated values of Cs obtained from Fig. 7(d) provides a summary of the GCD curves for various current densities. In Fig. 7(b), during the GCD test spanning from 5–0.25 A/g, the discharge segment displayed nearly symmetrical linearity, indicative of minimal internal resistance (IR drop) and high-speed charge/discharge characteristics, thereby highlighting excellent reversibility in electrochemical performance. It’s important to highlight that Si-CNT not only has a lower drop resistance but also has a longer discharge time than Si and CNT sample in a three-electrode system. At 1 A/g, the Si-CNT device’s specific capacitance can be determined using the GCD curves, and it measures 121.25 F/g. Capacitance performance as a component of current density has also been shown in Fig. 7(d). The energy density compared to power density of the Si-CNT supercapacitor device can be represented in the Ragone plot shown in Fig. 7(e)[24]. An energy density of 10.86 Wh/kg and a corresponding power density of 403.2 W/kg were computed at a current density of 1 A/g. The rise in current density leads to the reduction in energy density due to a drop in capacitance, but the energy density increases with discharge time. At a peak current density of 5 A/g, the achieved energy density was 10.08 Wh/kg, with a power density of 2016 W/kg. The comparison of Si-CNT supercapacitor electrode material with other published reports is summarized in Table 1. The capacitance of the Si-CNT device decreased very slowly and reduced to 90.7% of the initial capacitance and coulombic efficiency of 84.8% after 3,000 cycles, demonstrating the very stable and reliable performance of the Si-CNT supercapacitor device. The interior of Fig. 7(f) shows the PEIS measurements of the Si-CNT device prior to and after 3,000 cycles, demonstrating that the device is predictable and stable. It is determined that the Si-CNT device is well suited for supercapacitor applications because it not only has high capacitance but also excellent performance and rate capability.

Figure 7. (Color online) GCPL results of (a) Si, (b) CNT, and (c) Si-CNT with (d) the corresponding specific capacitance at different current densities, (e) Ragone plot, and (f) capacitance retention with Coulombic efficiency at 1 Ag-1 up to 3,000 cycles.

Comparison of Si- CNT supercapacitor electrode material with other published reports.

Electrode MaterialElectrolyteCapacitance (F g-1)Energy Density (Wh kg-1)Power density (Wkg-1)Ref
MoS2-CNTppy1M H2SO4275@1A g-113.1@1mA g-1625@1mA g-15
Si-CNF1M H2SO4206@1A g-121.6@1A g-1401@1A g-16
SiCNWs@Ni(OH)21M Na2SO41412@100A g-159.4@100 A g-127.5K@100 A g-17
HPC/SiC1M H2SO4234.2@1A g-125.20@1A g-1181@1A g-118
Si-CNT6M KOH121.3@1A g-1101.86@1A g-1403.2@1A g-1This work

Supercapacitors electrode material has been successfully synthesized and investigated using silicon-carbon nanotube electrodes (Si-CNT). The performance of the obtained composite material is attributed to its improved capacitance, broadened electroactive surface area, and dropped charge transfer resistance as compared to silicon and CNT themselves. The composite electrode had a 121.3 F/g specific capacitance in the three-electrode system, along with a high energy density of 10.86 Wh/kg, and the associated power density of 403.2 W/kg was noted. Effective cycling stability and reliability have improved, achieving 90.7% capacity retention and 84.8% coulombic efficiency after 3,000 cycles. This work could provide an efficient synthesis method for fabricating a supercapacitor device, which holds promise for upcoming generation energy storage devices. Si-CNT composite electrodes have exceptional electrochemical performance due to the synergistic effects of Si and CNTs, as well as the impact of composite morphology. Continued research into the underlying mechanisms and composite design parameters will improve the performance, scalability, and durability of these electrode materials for a wide range of energy storage applications.

  1. A. Vanpariya, et al., Int. J. Nanotechnol. 18, 483 (2021).
  2. M. Zhong, M. Zhang and X. Li, Carbon Energy 4, 950 (2022).
  3. K. M. Joseph, H. J. Kasparian and V. Shanov, Energies 15, 6506 (2022).
  4. H. Zhao, et al., Scr. Mater. 192, 49 (2021).
  5. S. M. S. Pandiyarajan, et al., ACS Appl. Energy Mater. 4, 2218 (2021).
  6. K. Pandey and H. K. Jeong, Chem. Phys. Lett. 834, 140972 (2024).
  7. L. Gu, et al., J. Power Sources 273, 479 (2015).
  8. G. Gyawali and H. K. Jeong, J. Energy Storage 40, 102806 (2021).
  9. H. Tang, et al., Coatings 12, 1515 (2022).
  10. R. D. Olmo, T. C. Mendes, M. Forsyth and N. Casado, J. Mater. Chem. A 10, 19777 (2022).
  11. K. Pandey and H. K. Jeong, J. Phys. Chem. Solids 181, 111520 (2023).
  12. Y. Li, et al., Ceram. Int. 45, 16195 (2019).
  13. K. Pandey and H. K. Jeong, Mater. Sci. Eng. B 295, 116601 (2023).
  14. J. Fu, et al., Front. Chem. 6, 00624 (2018).
  15. S. S. Rao and J. Energy, Storage 28, 101199 (2020).
  16. G. Gyawali and H. K. Jeong, New Phys.: Sae Mulli 73, 395 (2023).
  17. K. Pandey and H. K. Jeong, New Phys.: Sae Mulli 73, 403 (2023).
  18. Q. Tang, X. Chen, D. Zhou and C. Liu, Colloids Surf. A 620, 126567 (2021).
  19. E. Biserni, et al., Nanotechnology 27, 245401 (2016).
  20. A. L. M. Reddy and S. Ramaprabhu, J. Phys. Chem. C 111, 7727 (2007).
  21. Z. Wu, et al., J. Nanomater. 2021, 6650131 (2021).
  22. S. Chen, L. Qiu and H. M. Cheng, Chem. Rev. 120, 2811 (2020).
  23. K. Pandey and H. K. Jeong, New Phys.: Sae Mulli 72, 274 (2022).
  24. R. Vicentini, et al., Batter. Supercaps 4, 1291 (2021).

Stats or Metrics

Share this article on :

Related articles in NPSM