Longitudinal unzipped carbon nanotubes with high specific surface area and trimodal pore structure

Abstract

This study reports unzipped carbon nanotubes (CNTs) with a trimodal (micro–meso–macro) pore structure using KOH as the activating agent. It is possible to unzip CNTs under severe conditions (in our study, CNT (C)/KOH = 1 : 10 (w/w) at 1000 °C) in contrast to the surface activation of CNTs under general conditions (in our study, C/KOH = 1 : 4 (w/w) at 900 °C). After severe alkali activation, various pores were initially formed on the surface. Subsequently, a longitudinally unzipped structure was obtained as the individual pores connected. In contrast with other methods used to prepare unzipped and porous CNTs, this method is economical and scalable because it enables a one-step synthesis of unzipped and porous CNTs. As per the non-localized density functional theory, the distribution of micro–meso pores provides evidence of unzipping because the peak for pore sizes <1 nm, measured from the partially opened tips of the pristine CNTs, was broadened. The perfectly opened tips observed after activation indicate that the micropores on the unzipped structure increased. In addition, the results indicated that the unzipped porous CNTs exhibited a trimodal pore structure. This structure resulted in increased specific surface area as well as energy storage and adsorption capacities. The maximum energy density of the unzipped porous CNTs in ultracapacitors based on an organic electrolyte was 50 W h kg⁻¹. Thus, the method is suitable for fabrication of unzipped porous CNTs, which demonstrate potential as electrode materials for ultracapacitors.

---

Introduction

Carbon nanotubes (CNTs) have been studied because of their unique properties including their cylindrical graphitic structure, excellent conductivity, and higher mechanical strength than steel since being discovered by Iijima in 1991.¹ Thus, CNTs have been applied in various fields including hydrogen storage, field-emission materials, and electrode materials for ultracapacitors.^2–7 Among these applications, ultracapacitors have been widely studied because of their satisfactory performance including their high power density (>10 kW kg⁻¹) and long cycle stability (>10⁵ cycles) in energy-storage devices.^8–12 These devices can store electrical energy by forming an electrical double layer at the interface between the electrode and electrolyte. CNTs are suitable materials for ultracapacitor electrodes for the following reasons: (1) conductive pathways enable fast transport of the electrolyte ions and (2) the active electrode can easily adjust to the volumetric changes during charge and discharge because of their inherent resiliency, which affects the high cycling performance.¹³

CNTs can be classified according to their structure: single-walled CNTs (SWCNTs), double-walled CNTs (DWCNTs), and multi-walled CNTs (MWCNTs). SWCNTs are composed of graphene rolled into a single cylindrical wall and exhibit the best performance (electrical conductivity of ∼10⁷ S m⁻¹, specific surface area (SSA) of 1320 m² g⁻¹).^14,15 DWCNTs exhibit a combination of properties: the flexibility of SWCNTs and electrical and thermal stability of MWCNTs. In MWCNTs, because graphene is rolled into many layers, the internal wall or surface of MWCNTs cannot be properly utilized. However, MWCNTs are useful for commercialization because they are the least expensive and easiest to handle of the three types of CNTs. Therefore, to utilize MWCNTs, it is necessary to overcome their major disadvantage: their insufficient surface area for use in some applications such as energy-storage and adsorption devices.^16,17 In other words, the commercial value of MWCNTs would significantly improve if their surface areas were similar to those of SWCNTs.

Strategies to increase the surface area of CNTs include partial etching of the CNT surface and unzipping the CNTs using wet chemical routes such as alkali or acid treatments.^18–24 The wet chemical route is scalable and effective for modifying CNTs. Alkali treatment, for example, is a facile method in which carbon is consumed and pores are formed by pyrolysis when an activating agent (e.g., ZnCl₂, KOH, or NaOH) is used under an inert atmosphere. Pyrolysis occurs at low temperatures (approximately 400 °C). At high temperatures (approximately 800–1000 °C), metallic gases are generated. Consequently, a large internal surface area is obtained by the formation of accessible pore structures.²⁵ Zhang et al. recently reported that the porous structures of CNTs could be adjusted under different alkali activation conditions, and these modified CNTs could be used as CO₂ capture materials.²¹ In addition, there are many other studies on CNT alkali activation for ultracapacitors.^19,26–28 However, in these studies, the CNT surface areas could not be adequately increased. An unzipped structure was not produced because the severity of the activation conditions was insufficient, and the CNT surfaces were only partially etched after activation.

During acid treatment, significant defects are formed on the CNT surfaces, and subsequently, the CNTs are unzipped. A representative method for unzipping CNTs is the oxidization reported by Tour et al.²² In this method, new pores are formed on the surface by the redox reaction of an acid (e.g., KMnO₄) with the CNTs, and the unzipped structure is formed as the surface is continually etched.²² After the treatment, the rolled-up graphene becomes an unrolled graphene sheet because the CNTs unfold, which are called graphene nano-ribbons (GNRs) or a graphene nano-sheet (GNS). GNRs can be generally defined as a 1D sp² hybridized carbon crystal with boundaries that possesses a large aspect ratio.²⁹ GNRs contain abundant edges on their unzipped structure, which provides an advantage as an ion diffusion pathway for energy-storage materials.²³ However, if only the acid treatment is used, the surface area of CNTs is increased only slightly (∼40 to 80 m² g⁻¹), although the specific capacitance, which is associated with the edge sites, is high.²⁷ In contrast, using a three-step process, Zheng et al. (2014) reported unzipped porous CNTs with abundant edge sites. The CNTs were prepared using the Brodie method with chemical activation, which resulted in a large surface area and high energy density.²³ The first step involved unzipping the CNTs by acid treatment, followed by alkali treatment of the porous CNTs and subsequent reduction to remove oxygen functional groups.

Therefore, an economical and one-step method with severe activation conditions is necessary to obtain CNTs with a large surface area. In this paper, we present unzipped porous MWCNTs with increased surface area prepared by a one-step alkali activation procedure under severe conditions, which is more economical than acid treatment. Furthermore, we examine the unzipping mechanism and change in the MWCNT pore structure as activation proceeds under various conditions. Finally, we evaluate the change in the electrochemical performance of the MWCNTs after activation.

Experimental

Preparation of unzipped MWCNTs by alkali activation

Commercial MWCNTs with an average diameter of ∼10 nm (CNT MR99, Carbon Nanomaterial Technology Co. Ltd., Korea) were mixed well with KOH flakes (Sigma-Aldrich Co. LLC., Korea) at C/KOH weight ratios of 1 : 4, 1 : 8, and 1 : 10 using an electric blender. We used 3 g pristine MWCNTs and 12, 24, and 30 g KOH flakes, respectively, according to C/KOH ratio. Then, the mixed powder was placed in a nickel furnace and treated at 900 °C (1 : 4, 1 : 8 w/w) or 1000 °C (1 : 10 w/w) for 1 h under Ar flow (the heating rate was 2 °C min⁻¹, and the Ar gas flow rate was 300 mL min⁻¹). The powder was then cooled to ambient temperature. The resulting powder was first washed with a 0.1 M HCl solution and then with distilled water until a pH of 7 was achieved. Finally, the resulting product was dried at 80 °C in an oven for 24 h. After the activation, 1.6, 1.5, and 1.0 g MWCNTs remained.

Preparation of oxidized MWCNTs by acid treatment

Next, 3.5 g potassium permanganate (99%, Sigma-Aldrich) was added into a mixture with 1 g pristine MWCNTs (CNT MR99, Carbon Nanomaterial Technology Co. Ltd., Korea) and 50 mL of sulphuric acid solution (95%, Samchun Pure Chemical Co., Ltd.). The mixture was oxidized at 35 °C for 5 h. Then, 150 mL of distilled water and 50 mL of hydrogen peroxide (35%, Junsei Chemical Co., Ltd.) were added dropwise into the mixture. The black product was then repeatedly washed with 0.3 M hydrochloric acid (35–37%, Samchun Pure Chemical Co., Ltd.) and distilled water. Finally, the obtained black powder was freeze-dried.

Materials characterization

The unzipped structures and morphologies of the samples were examined using high-resolution transmission electron microscopy (HR-TEM, JEOL, JEM-2000EX, Japan) and field-emission scanning electron microscopy (FE-SEM, JEOL, JSM-6700F, Japan), respectively. The various pore structures and pore distributions were inspected using a gas analyser (Belsorp-Mini II, Japan) and studied using N₂ adsorption–desorption isotherms and non-localized density functional theory (NLDFT). NLDFT methods have already been commercialized for calculation of pore size distributions from adsorption isotherms. The NLDFT method has been widely applied and was featured in a recent ISO standard (ISO-15901-3). The SSA was evaluated using the Brunauer–Emmett–Teller (BET) method. The crystal structures of the samples were examined using X-ray diffraction (XRD, Rigaku D/Max 2500/PC, Japan) with Cu Kα radiation (λ = 1.54 Å) operating at 40 kV and 200 mA and Raman spectrometry (Horiba Jobin Yvon, LabRam Aramis, Japan) with an Ar-ion laser beam. The surface components of the samples were confirmed using X-ray photoelectron spectroscopy (XPS, PHI 5000 VersaProbe, ULVAC-PHI, Japan). The sheet resistance of the materials was measured using a four-probe tester (Loresta, MCP-T610).

Electrochemical characterization

Coating type electrodes were prepared from pristine and unzipped MWCNTs as active materials along with polyacrylamide (PAA, Sigma-Aldrich Co. LLC., Korea) as the binder (the weight ratio of active materials/PAA was 9 : 1). The obtained electrodes were used as the working electrodes, and 1.0 M lithium hexafluorophosphate (LiPF₆) in ethylene carbonate/dimethyl carbonate (EC/DMC) = 1/1 (v/v) was used as the electrolyte in CR2032 coin cells. To investigate the electrochemical characteristics of the samples, a galvanostatic charge–discharge (CD) test and cyclic voltammetry (CV) were performed using a potentiostat (VMP3, Biologic, France). To confirm the rate capability, various current densities ranging from 0.1 to 10 A g⁻¹ and various scan rates ranging from 0.1 to 1000 mV s⁻¹ were applied. Furthermore, the cyclic performance was confirmed using a CD test after 1500 cycles at 5 A g⁻¹. To calculate the specific capacitance, we used the following equation:
Here, C_SP is the specific capacitance, I is the discharge current (A), t is the discharge time (s), ΔV is the potential window (V) in discharge, and m is the mass of active materials on one electrode. The energy and power density were derived from the following equations:

Here, E is the energy density (W h kg⁻¹), C_SP is the specific capacitance, P is the power density (W kg⁻¹). To calculate the electrical conductivity, the following equation was used:
Here, σ is the conductivity (S m⁻¹), R_s is the sheet resistance (Ω sq⁻¹), and t is the thickness of the film (μm).

Results and discussion

Unzipped CNTs are more effective than pristine CNTs for energy-storage materials because of their numerous active sites. Moreover, unzipped CNTs have a larger SSA than pristine CNTs. One of the methods used to unzip CNTs is a harsh acid treatment by oxidation;^22–24 however, a two- or three-step process is required to obtain unzipped, porous CNTs, which is uneconomical. When oxidation is used, the method is expected to sever chemical bonds by only oxidizing the CNT surface, creating severe defects. In the intermediate step of Hummers' method, KMnO₄ is transformed into Mn₂O₇, which then oxidizes carbon.³⁰ However, for KOH activation, K₂CO₃ decomposes into K₂O and CO₂. The CO₂ partially consumes carbon; however, metallic K gas is formed at temperatures above 800 °C, which creates new pores during the chemical reaction.²⁵ Relatively few carbon defects are produced, which is the main difference between the oxidation and activation methods. In our work, MWCNTs with an unzipped and porous structure were prepared via KOH activation under severe conditions. KOH activation is a representative method of preparing porous carbons because it effectively provides a large SSA. However, when KOH activation is used as the CNT modification method and the activation condition is mild, the CNT SSA increases but the CNTs do not unzip^19,26–28 because the driving energy is insufficient. Therefore, we introduce an economical one-step method to obtain unzipped porous CNTs with low damage. Fig. 1(a) presents the scheme of the sequential unzipping phenomenon as the reaction proceeds by one-step alkali activation. To explain the mechanism of unzipping in MWCNTs, we examined the KOH activation process through the following reactions:²⁵

6KOH + 2C → 2K + 3H₂ + 2K₂CO₃ (1)

4KOH + CH_x → K₂CO₃ + K₂O + (2 + x/2)H₂ (2)

K₂CO₃ → K₂O + CO₂ (3)

K₂CO₃ + 2C → 2K + 3CO (4)

K₂O + C → 2K + CO. (5)

Fig. 1 Schematic illustration of unzipping process of MWCNTs via KOH activation.

At approximately 400 °C, potassium carbonate (2K₂CO₃) forms by pyrolysis (reactions (1) and (2)). At approximately 700 °C, 2K₂CO₃ decomposes into potassium oxide (K₂O) and carbon dioxide (CO₂), and subsequently, carbonization is performed (reaction (3)). In this step, pores are partially formed on the walls of the MWCNTs because of the surface etching. Then, CO₂ is reduced to CO, and K₂CO₃ is transformed into metallic K gas at temperatures above 800 °C (reactions (4) and (5)),²⁵ and as a result, the pores develop because carbon is consumed and metallic K gas etched on the surface. Subsequently, the individual pores connect, and the MWCNTs are longitudinally unzipped as the reaction gradually proceeds under severe activation conditions.

As demonstrated in Fig. 2(a) and S1(a),† pristine MWCNTs have a graphitic wall; however, the structure changes after KOH activation. Mild activation conditions (in our study, MWCNT (C)/KOH = 1 : 4 (w/w) at 900 °C) result in mostly opened-tip and porous MWCNTs without unzipped structures, as observed in Fig. 2(b) and S1(b);† the surface pore diameter is <4 nm.

Fig. 2 HRTEM image of the (a) pristine MWCNTs with closed tips and graphitic walls, (b) activated MWCNTs with pores formed under mild activation conditions (C/KOH weight ratio = 1 : 4), and (c) and (d) unzipped MWCNTs formed under severe activation conditions (C/KOH weight ratio = 1 : 8 and 1 : 10).

However, as demonstrated in Fig. 2(c) and (d), severe conditions (in our study, C/KOH (w/w) = 1 : 8 and 1 : 10 at 900 and 1000 °C, respectively) result in pore formation and unzipping of MWCNTs, as the C/KOH ratio and reaction temperature are increased. The pore diameter is <10 nm for the unzipped MWCNTs. The sample obtained from C/KOH = 1 : 10 (w/w) at 1000 °C is denoted as “unzipped MWCNTs”. Under mild conditions, the driving energy to unzip CNTs is insufficient; however, under severe conditions, the appropriate amount of energy allows the CNTs to unzip.

Fig. 3(a) and (b) present SEM images of the pristine and unzipped MWCNTs, respectively, with the insets showing the associated TEM images. The entangled MWCNTs are converted into the aggregated MWCNTs because of pore formation on the surface and unzipping of the MWCNTs after severe activation. In other words, the large empty spaces among entangled MWCNTs (Fig. 3(a) and S1(c)†) are reduced by connection of the unzipped walls formed under high activating agent content and temperature (Fig. 3(b) and S1(d)†). This phenomenon results in a 3D network structure that could facilitate ion migration and improve electrochemical performance for energy-storage materials.³¹

Fig. 3 High-resolution SEM image of the (a) entangled MWCNTs before activation and (b) unzipped and aggregated MWCNTs after activation (C/KOH weight ratio = 1 : 10) (low-resolution TEM images are shown in the insets), (c) nitrogen adsorption–desorption isotherm, and (d) NLDFT analysis result for the pristine MWCNTs and those modified under various activation conditions.

To confirm that the MWCNT pore structure changes as the C/KOH ratio is increased, we investigated N₂ adsorption–desorption isotherms and the micro–meso pore size distribution using NLDFT, and the results are presented in Fig. 3(c) and (d), respectively. As observed in Table 1, the BET SSA of the pristine MWCNTs was 214 m² g⁻¹. After activation, the samples corresponding to C/KOH (w/w) = 1 : 4, 1 : 8 and unzipped MWCNTs exhibited BET SSA values of 628, 662, and 1123 m² g⁻¹, respectively. Mild conditions result in surface-etched CNTs; however, the severe conditions used here resulted in porous and simultaneously unzipped CNTs, which contributed to the larger SSA values. Acid treatment can unzip CNTs but cannot make the CNT surfaces porous, which is why we selected severe alkali activation. Table 1 lists the total pore volumes calculated at the highest relative pressure (P/P₀ = 0.99); micropore volumes calculated at the lowest relative pressure (P/P₀ = 0.1); mesopore volumes measured using the Barrett–Joyner–Halenda (BJH) method; and macropore volumes of pristine, activated, and unzipped MWCNTs. The unzipped CNTs exhibited a trimodal (micro–meso–macro) pore structure. The total pore volume increased from 1.28 to 2.38 cm³ g⁻¹, which indicates that the mesopore structure increased after activation.³² Moreover, the micropore volume increased from 0.08 to 0.40 cm³ g⁻¹ because of activation. The modified pore structure and formation of new pores indicate that the SSA and total pore and micropore volumes dramatically increased as the C/KOH weight ratio, temperature, and thus the driving energy increased. As observed in Fig. 2(c), at the lowest relative pressure, the N₂ adsorption–desorption isotherms gradually grow as the C/KOH weight ratio increases, which indicates that the number of MWCNT micropores increases. In addition, the isotherm is a type II hysteresis plot, as designated by IUPAC, which indicates that the number of mesopores also increases.³³ However, the isotherm of the pristine MWCNTs is type I, which indicates that the material is not porous. Thus, the number of micropores and mesopores increased simultaneously after the activation. These pore structures improve the energy-storage capability and ion pathways in the electrolyte of energy-storage devices. In the micropore and mesopore distributions obtained by NLDFT in Fig. 2(d), the peak corresponding to MWCNT pore widths <1 nm broadens after the perfect unzipping of the MWCNTs, which indicates that micropores are formed as the number of MWCNT partially opened tips increases. In fact, the large number of edge sites and micropores created on the surface by unzipping after severe activation indicates perfect CNT unzipping, which yields GNRs with abundant edges on their unzipped structure and with an advantage as an ion diffusion pathway for energy-storage materials.²³

Table 1 Pore structure characteristics of pristine, unzipped, and oxidized MWCNTs^a

Sample	C/KOH ratio	Temperature (°C)	Yield (%)	SBET (m^2 g^−1)	Vtotal (cm^3 g^−1)	Vmicro (cm^3 g^−1)	Vmeso (cm^3 g^−1)	Vmacro (cm^3 g^−1)
a SBET: BET specific surface area; Vtotal: total pore volume; Vmicro: micropore volume; Vmeso: mesopore volume; Vmacro: macropore volume. Total pore volume measured by BET from P/P0 = 0.99; micropore volume measured by BET from P/P0 = 0.1; mesopore volume measured by BJH; macropore volume (total pore volume − micropore volume − mesopore volume). Yield: the mass change of the activated or unzipped MWCNTs after the KOH activation; 3 g pristine MWCNTs and 12, 24, and 30 g KOH flakes according to the C/KOH ratio were used. After activation, 1.6, 1.5, and 1.0 g MWCNTs remained.
Pristine MWCNTs	—	—	—	214.3	1.28	0.08 (6.3%)	0.47 (36.8%)	0.73 (56.9%)
Activated MWCNTs #1	1 : 4	900	53%	627.8	1.36	0.23 (16.9%)	0.91 (66.9%)	0.22 (16.2%)
Activated MWCNTs #2	1 : 8	900	51%	661.5	1.58	0.23 (14.6%)	1.06 (67.1%)	0.29 (18.3%)
Activated MWCNTs #3 (unzipped MWCNTs)	1 : 10	1000	34%	1123.2	2.38	0.40 (16.8%)	1.54 (64.7%)	0.44 (18.5%)

---

The samples corresponding to C/KOH (w/w) = 1 : 4 and 1 : 8 are more porous and partially unzipped; however, the peak for pore widths <1 nm is sharp, which indicates that few micropores form in the samples because the samples are not perfectly unzipped. In addition, large macropores with diameters >300 nm formed because of the empty space among entangled MWCNTs. However, after activation, micropores and mesopores with diameters of <1 nm and <4 nm, respectively, were created because of etching of the surface, as confirmed by the TEM images. Mesopores with diameters <22 nm were generated because of the reduction of space among MWCNTs because of the connection of individual pores on the surface and aggregation of MWCNTs, as observed in the SEM image of the unzipped MWCNTs. This finding indicates that after activation, micropores and different-sized mesopores are formed, which is consistent with the inferences drawn from the isotherms. These changes in the pore structure could improve the characteristics of materials such as the adsorption and energy-storage capacity.²¹

To analyse the structure of the unzipped MWCNTs, XRD patterns and Raman spectra were obtained. As observed in Fig. 4(a), the peaks corresponding to the (002) plane at 26° and the (100) plane at 43°, indicating a graphitic structure, are broader in the unzipped MWCNTs because the graphitic walls of MWCNTs partially collapse because of KOH etching after the activation, as indicated by the TEM images.

Fig. 4 (a) XRD patterns and (b), (c) Raman spectra of the pristine and unzipped MWCNTs, (c) magnified Raman spectra from 1500 to 1700 cm⁻¹.

The Raman spectra in Fig. 4(b) and (c) reveal that the intensity at 1580 cm⁻¹ (G band), corresponding to the crystalline graphitic layer intensity, is relatively lower than that at 1348 cm⁻¹ (D band), which indicates that disordered carbon increased after activation. The calculated I_D/I_G ratio, representing the degree of disorder in crystalline structures, increased from 1.20 to 1.47 after unzipping. In addition, the peak at 2692 cm⁻¹ (2D band) is lower than that before activation. This finding indicates that defects on the MWCNTs develop because of the formation of many pores on the MWCNT walls after activation. In addition, Fig. 4(c) presents the magnified spectra: the G and D′ peaks are broader and right-shifted from 1581 to 1588 cm⁻¹ and from 1614 to 1619 cm⁻¹, respectively, which also indicates that the graphitic structure is transformed to the amorphous structure via activation.³⁴

Thus, the MWCNT SSA increases as the graphitic walls collapse, and the number of defects increases because of unzipping and the creation of new pores in the MWCNTs after severe activation corresponding to previous analyses.

To confirm the surface components of the samples, XPS spectra were obtained. Fig. 5(a) and (c) present the XPS survey spectra with the inset showing the magnified O 1s spectra (as shown in Fig. S2†), and Fig. 5(d) and (e) present the C 1s spectra. After severe activation, the percentage of C 1s decreases from 99.7% to 94.5%, whereas the percentage of O 1s increases from 0.3% to 5.5%; however, the value is lower than that of oxidized MWCNTs, as observed in Fig. S3† and Table 2. The oxygen content of the oxidized MWCNTs is 27.8% and offers many oxygen functional groups, which suggests that the damage of unzipped MWCNTs formed by alkali treatment is less than that of unzipped MWCNTs formed by acid treatment.

Fig. 5 (a) and (b) XPS survey spectra (inset is O 1s spectra) and (c) and (d) C 1s spectra of the (a) and (c) pristine MWCNTs and (b) and (d) unzipped MWCNTs, respectively.

Table 2 Relative percentage of components on surfaces of pristine, unzipped, and oxidized MWCNTs^a

Sample	C/KOH ratio	C (%)		sp^2 C=C		sp^3 C–C		C–O		C=O		–COO
		C 1s	O 1s	BE (eV)	C (%)	BE (eV)	C (%)	BE (eV)	C (%)	BE (eV)	C (%)	BE (eV)	C (%)
a C: percentage of four components in the samples; BE: binding energy.
Pristine MWCNTs	—	99.7	0.3	284.5	70.7	285.1	20.3	286.3	9.0	—	—	—	—
Unzipped MWCNTs	1 : 10	94.5	5.5	284.5	65.6	285.1	16.1	286.0	11.6	287.5	6.5	—	—
Oxidized MWCNTs	—	72.2	27.8	—	—	285.0	53.0	286.3	10.0	287.3	21.1	288.9	15.9

---

As observed in Fig. 5(c) and (d), the XPS C 1s spectra for both samples show the graphitic structure (sp², C=C) at 285.4 eV, and the bonding percentage decreases after activation. In addition, peaks corresponding to the sp³, C–C bond at 285.1 eV and the C–O bond at 286.3 eV exist for both samples. Actually unzipped CNTs possess oxygen functional groups such as carbonyls (C=O), carboxyls (–COOH) and hydroxyls (C–OH) at their edges and surface.²² The C–C bond is predominant in oxidized MWCNTs, which indicates that there are many oxygen functional groups on the sample surface because of defects formed by severe acid treatment. The peak corresponding to the C=O bond at 287.5 eV is observed for unzipped MWCNTs, which corresponds to the deconvolution of the O 1s spectra in Fig. S2.† This finding indicates that defects and oxygen functional groups are created on the unzipped MWCNTs after severe activation. Table 2 lists the percentages of the four components on the sample surfaces.³⁵

To confirm the electrochemical performance of the unzipped MWCNTs, CV at various scan rates was conducted (Fig. 6).

Fig. 6 (a) CV and (b) CV curve at 50 mV s⁻¹ and (c) rate capability at various scan rates; (d) Ragone plot of the samples.

For unzipped MWCNTs, the specific capacitance is four times higher than that of pristine MWCNTs. This result could arise because the 3D structure forms by the connection between pores, and abundant edge sites are created on unzipped structures after severe activation. The 3D structure, abundant edge sites, and trimodal (micro–meso–macro) pores provide a rapid pathway for ion diffusion. Moreover, the edge sites contain plentiful functional groups; the sites are unstable and contribute to increased energy-storage and adsorption capacity.²³ The maximum specific capacitance is 160 F g⁻¹ at 0.1 mV s⁻¹. The rate capabilities at various scan rates exhibit slight differences, although many defects are generated after activation (the rate capability slightly decreases from 42% to 39%), as observed in Fig. 6(c). This result occurs because the 3D structure offers rapid ion-diffusion pathways, and the electrical conductivity of the pristine MWCNTs changes because of the defects formed on the MWCNT walls. In addition, the galvanostatic test shows that the specific capacitance remains 99% at 5 A g⁻¹ during 1500 cycles, as shown in Fig. S4.†

The electrical conductivity decreases from 1760 to 628 S m⁻¹, as observed in Fig. S5,† which is 35% of the electrical conductivity of the pristine MWCNTs but 80 times higher than that of commercial activated carbon (7.74 S m⁻¹). Thus, the unzipped MWNCTs exhibit high SSA and relatively high electrical conductivity.

Conclusions

In summary, we prepared unzipped and porous MWCNTs via a one-step KOH activation method under severe conditions (C/KOH = 1 : 10 (w/w) at 1000 °C). At low temperature, various pores were partially formed on the MWCNT walls because of etching of the surface. Moreover, the individual pores were connected and the MWCNTs were longitudinally unzipped as the reaction gradually proceeded.

The alkali activation is a more economical and simple process than acid treatment because it makes the MWCNT surface widely porous and simultaneously unzips the MWCNTs. Although the acid treatment can unzip the MWCNTs, the MWCNTs are only made slightly porous. Moreover, the treatment requires two or three steps to remove oxygen functional groups.

The unzipped CNTs exhibited trimodal (micro–meso–macro) pore structure. This pore structure provided a large SSA and 3D network for facilitating ion migration and improved the electrochemical performance for energy storage. In addition, the electrical conductivity of the unzipped MWCNTs was 628 S m⁻¹, which is a relatively high value that also contributed to the improved electrochemical performance. Although the electrical conductivity was lower than that of the raw materials, good electrochemical performance was maintained because of the modified pore structure obtained after activation. In brief, we have provided a facile and low defect method to prepare unzipped and porous CNTs that show potential as electrode materials for ultracapacitors.

Acknowledgements

This work was supported by the Energy Efficiency & Resources Core Technology Program of the Korea Institute of Energy Technology Evaluation and Planning (KETEP) granted financial resource from the Ministry of Trade, Industry & Energy, Republic of Korea (No. 20142020104160, No. 20122010100140) and this work was supported by the Technology Innovation Program (10049587, Development of hard carbon-based activated carbon and electrode material for supercapacitor) funded By the Ministry of Trade, industry & Energy (MI, Korea). This work was also supported by the industrial promotion program of economic cooperation area of MOTIE/KIAT (R004005, development of EDLC for high temperature (140 °C) application). And this work was supported by a grant from the Fundamental R&D program and funded by the Korea Institute of Ceramic Engineering and Technology (KICET) and Ministry of Trade, Industry and Energy (MOTIE), Republic of Korea. This work was also supported by the Energy Efficiency & Resources program of the Korea Institute of Energy Technology Evaluation Planning (KETEP), and was granted financial resources from the Ministry of Trade, Industry & Energy, Republic of Korea (No. 20152020105770).

Notes and references

1. S. Iijima, Nature, 1991, 354, 56–58.
2. Q. Wang, A. Setlur, J. Lauerhaas, J. Dai, E. Seelig and R. Chang, Appl. Phys. Lett., 1998, 72, 2912–2913.
3. Q. Jiang, M.-Z. Qu, B.-L. Zhang and Z.-L. Yu, Carbon, 2002, 40, 2743–2745.
4. M. Cinke, J. Li, C. W. Bauschlicher Jr, A. Ricca and M. Meyyappan, Chem. Phys. Lett., 2003, 376, 761–766.
5. R. Bacsa, C. Laurent, R. Morishima, H. Suzuki and M. Le Lay, J. Phys. Chem. B, 2004, 108, 12718–12723.
6. S. Arepalli, H. Fireman, C. Huffman, P. Moloney, P. Nikolaev, L. Yowell, K. Kim, P. A. Kohl, C. D. Higgins, S. P. Turano and W. J. Ready, JOM, 2005, 57, 26–31.
7. F. Yu, Y. Wu, X. Li and J. Ma, J. Agric. Food Chem., 2012, 60, 12245–12253.
8. P. L. Taberna, S. Mitra, P. Poizot, P. Simon and J. M. Tarascon, Nat. Mater., 2006, 5, 567–573.
9. X. Ji, K. T. Lee and L. F. Nazar, Nat. Mater., 2009, 8, 500–506.
10. S. Zhang, Y. Li and N. Pan, J. Power Sources, 2012, 206, 476–482.
11. H. Pan, J. Li and Y. P. Feng, Nanoscale Res. Lett., 2010, 5, 654–668.
12. F. Liu and D. Xue, Chem.–Eur. J., 2013, 19, 10716–10722.
13. H. Pan, et al., Nanoscale Res. Lett., 2010, 5, 654–668.
14. W. S. Bao, S. A. Meguid, Z. H. Zhu and G. J. Weng, J. Appl. Phys., 2012, 111, 093726.
15. F. H. Gojny, M. H. G. Wichmann, B. Fiedler, W. Bauhofer and K. Schulte, Composites, Part A, 2005, 36, 1525–1535.
16. C. Niu, E. K. Sichel, R. Hoch, D. Moy and H. Tennent, Appl. Phys. Lett., 1997, 70, 1480–1482.
17. D. Mhamane, A. Suryawanshi, S. M. Unni, C. Rode, S. Kurungot and S. Ogale, Small, 2013, 9, 2801–2809.
18. Q. Jiang, M.-Z. Qu, B.-L. Zhang and Z.-L. Yu, Carbon, 2002, 40, 2743–2745.
19. C. Li, D. Wang, T. Liang, X. Wang and L. Ji, Mater. Lett., 2004, 58, 3774–3777.
20. J. J. Niu and J. N. Wang, Solid State Sci., 2008, 10, 1189–1193.
21. X. Zhang and W. Chen, Microporous Mesoporous Mater., 2015, 206, 194–201.
22. D. V. Kosynkin, A. L. Higginbotham, A. Sinitskii, J. R. Lomeda, A. Dimiev, B. K. Price and J. M. Tour, Nature, 2009, 458, 872–876.
23. C. Zheng, X. F. Zhou, H. L. Cao, G. H. Wang and Z. P. Liu, J. Mater. Chem. A, 2014, 20, 7484–7490.
24. H. Wang, Y. Wang, Z. Hu and X. Wang, ACS Appl. Mater. Interfaces, 2012, 4, 6827–6834.
25. J. Romanos, M. Beckner, T. Rash, L. Firlej, B. Kuchta, P. Yu, G. Suppes, C. Wexler and P. Pfeifer, Nanotechnology, 2012, 23, 015401.
26. B. Xu, F. Wu, Y. Su, G. Cao, S. Chen, Z. Zhou and Y. Yang, Electrochim. Acta, 2008, 53, 7730–7735.
27. Q. Jiang, M. Z. Qu, G. M. Zhou, B. L. Zhang and Z. L. Yu, Mater. Lett., 2002, 57, 988–991.
28. Y.-M. Dai, W.-J. Liu, T.-C. Pan and J.-M. Jehng, Appl. Surf. Sci., 2012, 258, 3027–3032.
29. M. Terrones and A. R. Botello–Méndez, et al., Nano Today, 2010, 5, 351–372.
30. J. Zhao, L. Liu and F. Li, Graphene Oxide: Physics and Applications, Springer, 2015.
31. D.-W. Wang, F. Li, M. Liu, G. Q. Lu and H.-M. Cheng, Angew. Chem., 2008, 120, 379–382.
32. T. Kim, G. Jung, S. Yoo, K. S. Suh and R. S. Ruoff, ACS Nano, 2013, 7, 6899–6905.
33. P. He, L. Liu, W. Song, G. Xiong, T. S. Fisher and T. Chen, RSC Adv., 2015, 5, 31837–31844.
34. K. N. Kudin, B. Ozbas, H. C. Schniepp, R. K. Prud'homme, I. A. Aksay and R. Car, Nano Lett., 2008, 8, 36–41.
35. C.-C. Wang, Y.-C. Hsueh, C.-Y. Su, C.-C. Kei and T.-P. Perng, Nanotechnology, 2015, 26, 254002.

---

Footnote

† Electronic supplementary information (ESI) available: TEM images of MWCNTs before and after activation; XPS spectra of unzipped and oxidized MWCNTs; galvanostatic discharge curve and capacitance retention of MWCNTs before and after activation; SEM cross-section images and electrical conductivity of MWCNTs before and after activation; synthesis of oxidized MWCNTs. See DOI: 10.1039/c5ra22527b

---