Multi-template synthesis of hierarchically porous carbon spheres with potential application in supercapacitors

Abstract

A new and simple multi-template approach towards hierarchical porous carbon (HPC) materials was reported. HPC spheres were prepared by using hierarchical silica capsules (HSCs) as the hard template and triblock copolymer Pluronic P123 as the soft template. Three types of pores were tunably constructed in the HPC spheres in a wide size range of 3.0 to 100 nm. Since the HSCs were in situ constructed by silica nanoparticles, which were formed from the sol–gel system of tetraethylorthosilicate and (3-aminopropyl)triethoxysilane (APTES), the porous structures of HPCs were simply controlled by changing the size of silica nanoparticles, i.e. by varying the APTES content. Thanks to their high surface areas and interconnected pores, the HPCs exhibited good electrochemical performance with specific capacitances of up to 170 F g⁻¹ and outstanding cycling stability without capacitance loss after 5000 cycles.

---

Introduction

Porous carbon materials have attracted significant research interest in recent years, owing to their great application potential in many areas, such as energy/gas storage and conversion, water purification, sensors, and solid catalyst carriers.^1–6 In particular, porous carbons have proven to be one of the most promising electrode materials for supercapacitors due to their high surface area, excellent physicochemical stability, good conductivity and relatively low cost.¹ So far, much effort has been devoted to develop efficient strategies for the fabrication of porous carbons with controllable pore structures, which may result in tuneable capacitive performance of the porous carbon materials.^3,5,7–9 Currently, the most popular approaches include hard-template,^10–19 soft-template,^20–24 and multi-template methods.^25–29 For example, Zhao and co-workers reported a facile fabrication of mesoporous carbon spheres with <3 nm nanopores using a Pluronic amphiphilic block copolymer as the soft template and resol as the carbon source.²⁴ However, pore structures of the porous carbons fabricated via hard-template or soft-template methods are generally monotonous and limited. In contrast, multi-template approach can produce porous carbons with a great variety of pore structures, including hierarchically porous carbons (HPCs).^5,27–29 For instance, Wen et al. have prepared three-dimensional HPCs by using colloidal crystal SiO₂ as the hard template and a triblock copolymer as the soft template.²⁷ HPCs have received considerable attention as electrode materials of energy storage devices including supercapacitors, since they afford synergistic effects of different pore structures. For instance, macropores or large mesopores may act as reservoir for electrolytes and thus shorten the ion diffusion distances from the exterior to the interior surfaces of electrode materials, while the small mesopores and micropores can enhance ion transport and charge storage,^30,31 leading to high capacitive performance of supercapacitors.^3,9 Although a few studies have heretofore devoted to the development of multi-template techniques for the preparation of HPC materials,^5,27–29 facile multi-template methods towards HPCs with tuneable porosity and pore size are still highly desirable.

Herein, we report a new and simple multi-template method for the fabrication of HPC spheres with tuneable pore sizes at mesoscale (Fig. 1), in which hierarchical silica capsules (HSCs) as the hard template play a critical role. HSCs are well constructed by silica nanoparticles, which were prepared in one step by using CaCO₃-rod stabilised water-in-oil (w/o) emulsions as the soft template. The primary size of silica nanoparticles can be simply controlled by varying the ratio of the precursors, e.g. tetraethylorthosilicate (TEOS) to (3-aminopropyl)-triethoxysilane (APTES), and it eventually makes the pore sizes of HPCs tuneable. In addition to HSCs, triblock copolymer Pluronic P123 (EO₂₀PO₇₀EO₂₀) is employed as the soft template to create ordered mesopores, which naturally makes the large pores interconnected in HPC spheres. After the removal of the templates and carbonization by thermal treatment at 900 °C, HPC spheres with three types of pores are successfully achieved, including ordered mesopores I (∼3.0 nm) templated by P123, mesopores II (5–20 nm) templated by the silica nanoparticles, and meso–macro pores (20–100 nm) templated by the aggregates of the silica nanoparticles. The HPC spheres possess high surface areas of up to ∼760 m² g⁻¹. Acting as electrode materials of supercapacitors, the HPC materials with hierarchical pores and high surface areas exhibit good electrochemical performance with specific capacitance of up to 170 F g⁻¹ and excellent cycling stability without capacitance loss after 5000 cycles.

Fig. 1 Illustration of the fabrication of HPCs. Firstly, hierarchical silica spheres are prepared by using CaCO₃-rod stabilised w/o emulsion as a template, in which TEOS and APTES are dissolved in paraffin oil as silica precursors. After aging at 60 °C for 20 h (step 1), the oil phase and CaCO₃ rods are washed out from the newly formed hierarchical silica (step 2) before the impregnation of Pluronic P123 (EO₂₀PO₇₀EO₂₀) and resol. HPCs have been successfully fabricated with the following solidification, carbonization and removal of P123 and silica templates.

Results and discussion

Hierarchical silica capsules were prepared by using a previously reported emulsion-template method.³² By using CaCO₃ rods (Fig. S1, ESI†) as stabilisers, the w/o emulsions were prepared by manually shaking of paraffin oil containing modified CaCO₃ rods, TEOS and APTES, with 2.5 wt% PVP aqueous solution. The optical micrographs of emulsions (Fig. S2a1–f1, ESI†) show that there is little difference between the emulsions despite different APTES contents, and those emulsions are polydisperse with droplet sizes in the region of 40–150 μm. Similar to the surface morphologies of CaCO₃ rod-stabilised o/w emulsions³² and air bubbles,³³ the droplets are fully covered with CaCO₃ rods, which self-assemble at oil–water interfaces and form 2D polydomains structures. Owing to the robust armoured structures, all the emulsions changed little after aging at 60 °C for 20 h (Fig. S2a2–f2, ESI†). However, when oil phase was washed out, it is found that the silica/CaCO₃ composite capsules behave differently with the increase of APTES content (Fig. S2a3–f3, ESI†). The composite capsules prepared from the system with 0.04 mL of APTES are well kept as spherical with robust shells (Fig. S2a3, ESI†), while there are a few cracks in the shells of composite capsules at 0.10 mL of APTES (Fig. S2b3, ESI†). As the APTES content increases to 0.20 mL, around one third composite capsules are broken owing to cracks in the shells (Fig. S2c3, ESI†). However, the composite capsules get better at 0.40 mL of APTES, and the cracks in the shells reduce dramatically (Fig. S2d3, ESI†). When the APTES content increases to 0.80 mL and 1.20 mL, the composite capsules become more robust with few cracks (Fig. S2e3 and f3, ESI†). It is interesting to know what will happen to the silica capsules after the CaCO₃ rods were removed.

Fig. 2 clearly shows the microstructure evolution of silica capsules as the APTES content increases from 0.04 mL to 1.20 mL. When there is 0.04 mL in the oil phase, hollow silica capsules have been obtained with very few silica nanoparticles in the capsules (Fig. 2a). The hollow silica capsules are similar with those we reported previously,³² but aging for a shorter time. That is why the silica shell is a bit thinner with about 30 nm in thickness. With the increase of APTES content, there are more and more silica nanoparticles formed in the silica capsules (Fig. 2b–f), while the silica shells become thinner and thinner. As the APTES content reaches 1.20 mL, the silica capsules are fully packed with silica nanoparticles, which are closely connecting with one another (Fig. 2f) to construct such hierarchical structures. It is important to note that the primary particle size of silica nanoparticles decreases from about 90 nm to 10 nm or smaller, as the APTES content increases from 0.10 mL to 1.20 mL. Undoubtedly, APTES plays a critical role in creating the hierarchical silica capsules. TEOS–APTES is a well-known self-catalysed system for the formation of silica,^32,34,35 but here we find something very interesting. APTES is water soluble due to the existence of aminopropyl groups, and moreover the amino groups as nucleophilic groups could easily get protonated in water. In a typical sol–gel process, APTES migrates from the oil phase to the oil–water interface first, and then dissolves in water and become protonated. Meanwhile, the dissolution and protonation of APTES will induce a pH increase, and consequently facilitate the hydrolysis and condensation of TEOS and APTES. When the APTES content is 0.04 mL, almost all protonated APTES molecules are located at oil–water interfaces and rod–water interfaces, owing its amphiphilic properties.³⁵ Therefore, hollow silica capsule is the most likely microstructure. As the APTES content increases, more and more APTES gets dissolved and protonated in water. It is expected that at the high APTES content, all the protonated molecules cannot be adsorbed at the oil–water and rod–water interfaces. The “excess” thus most likely diffuses in bulk water, and facilitate the hydrolysis and condensation of TEOS and APTES, as well as the formation of silica nanoparticles in the inner water phase. With the increase of APTES content, both the “excess” protonated APTES molecules and the pH of water phase increase, and as a result, there will be more and more nucleation sites of silica while the primary particle size of silica will decrease. In other words, as the APTES content increases, silica capsules evolve from hollow capsules to hierarchical silica capsules having nanoparticles (network) in their bulk, which in turns allow controlling capsule porosity.

Fig. 2 Microstructure evolution of hierarchical silica capsules with increasing the APTES content in the oil phase, imaged by Cryo-SEM. (a) 0.04 mL APTES. (b) 0.10 mL APTES. (c) 0.20 mL APTES. (d) 0.40 mL APTES. (e) 0.80 mL APTES. (f) 1.20 mL APTES. Scale bars: 50 μm for a1, b1, c1, d1, e1 and f1; 5 μm for a2, b2, c2, d2, e2 and f2; 0.5 μm for a3, b3, c3, d3, e3 and f3.

Besides the Cryo-SEM images of the cross section of silica capsules, the optical micrographs (Fig. S3, ESI†) and SEM images (Fig. 3) also demonstrate the microstructure evolution of silica capsules. After the removal of CaCO₃ rods, all the six types of silica capsules remain spherical or quasi-spherical in water dispersions despite the different APTES contents (Fig. S3, wet samples, ESI†). However, during the drying process, the silica capsules behave differently. At the low APTES contents (0.04 mL and 0.10 mL), the hollow silica capsules will get broken after drying (Fig. 3a and b and S3a and b, ESI†), since the shells are too thin to withstand the high osmotic pressure during the drying. On the other hand, at high APTES contents (0.80 mL and 1.20 mL), the hierarchical silica capsules shrink a little during the drying process (Fig S3e and f, ESI†), for silica nanoparticles get more closely packed with the evaporation of water. Their spherical shapes are well kept. Moreover, the SEM images (Fig. 3e and f) indicate that silica nanoparticles are well encapsulated by very thin silica shells, and without the shells there will be probably no robust hierarchical silica capsules. As it is expected, the transition morphologies and microstructures (Fig. 3c and d and S3c and d, ESI†) are observed from the silica capsules prepared at the medium APTES contents (0.20 mL and 0.4 mL). In particular, the capsules prepared from 0.20 mL APTES system behave like the hollow capsules (Fig. S3c, ESI†), collapsing rather than shrinking during the drying process, but there are lots of silica nanoparticles inside the capsules (Fig. 3c).

Fig. 3 Surface morphology variation of dried hierarchical silica capsules with increasing the APTES content in the oil phase, imaged by SEM. (a) 0.04 mL APTES. (b) 0.10 mL APTES. (c) 0.20 mL APTES. (d) 0.40 mL APTES. (e) 0.80 mL APTES. (f) 1.20 mL APTES. Scale bars, 200 μm for a1, b1, c1, d1, e1 and f1; 50 μm for a2, b2, c2, d2, e2 and f2; 5 μm for a3, b3, c3, d3, e3 and f3; 0.5 μm for a4, b4, c4, d4, e4 and f4.

Nitrogen adsorption/desorption analyses reveal the change of the BET surface areas of silica capsules as the APTES content varies (Table S1†). As the APTES content increases from 0.04 mL to 0.20 mL, the BET surface area of silica capsules decreases from 287 m² g⁻¹ to 129 m² g⁻¹, while it increases from 129 m² g⁻¹ to 422 m² g⁻¹ with increasing the APTES content from 0.20 mL to 1.20 mL. Since there are no ordered porous structures in those silica capsules which is proved by small angle X-ray diffraction (Fig. S4, ESI†), the BET surface areas are mainly contributed by the external surfaces of silica shells and nanoparticles. When looking at the hollow capsules prepared at 0.04 mL of APTES (Fig. 2a), we could find that the thickness of silica shells is around 30 nm or even thinner. Moreover, there are a few micropores randomly formed during the condensation of hydrolysed TEOS and APTES,³² so the BET surface area could reach 287 m² g⁻¹. Since there are a few big silica nanoparticles with diameters around 90 nm (Fig. 1b), the BET surface area of hollow capsules prepared at 0.10 mL of APTES decreases to 250 m² g⁻¹. As to the silica capsules prepared at 0.20 mL of APTES and above (Fig. 2c–f), silica nanoparticles will contribute the large percentage of their BET surface areas rather than silica shells. When the APTES content increases from 0.20 mL to 1.20 mL, the particle size of silica nanoparticles decreases from 50 nm to 10 nm or even smaller, and consequently their BET surface areas increase from 129 m² g⁻¹ to 422 m² g⁻¹. It is worth mentioning that there is a dramatic decrease of BET surface area when the APTES increases from 0.10 mL to 0.20 mL, because of the microstructure transition from hollow silica capsules to hierarchical silica capsules constructed by nanoparticles.

It was reported earlier that PVP could be used to facilitate the growth of silica on the surface of colloidal particles.³⁶ However, from the microstructures of silica capsules, here we assume that PVP plays a minor role on the formation of silica, e.g. silica layer at the oil–water interfaces and silica nanoparticles within the droplets. Therefore, a series of control samples have been produced by using the same process and similar recipes, but replacing 2.5 wt% PVP aqueous solution with deionized water. The optical micrographs (Fig. S5, ESI†) indicate that the morphology evolution of silica/CaCO₃ composite capsules is similar with that of the composite capsules prepared with PVP. As expected, the similar microstructure evolution of silica capsules is observed by optical microscope (Fig. S6, ESI†) and Cryo-SEM (Fig. S7, ESI†). On the other hand, it further proves that APTES plays a critical role in both the formation of silica shells and nanoparticles from another angle. It will provide a simple approach to scientist for the creation of various silica nanostructures.

By employing the hierarchical silica capsules prepared at 0.40 mL, 0.80 mL and 1.20 mL as hard templates, hierarchical porous carbons (HPCs) have been successfully prepared using a reported method.^14,37 In addition to silica capsules, Pluronic P123 is used as soft template to create ordered mesoporous structures, and resol is used as the precursor of carbon. From the SEM images of these hierarchical porous carbons (Fig. S8, ESI†), we can observe most of HPC spheres are immersed in the carbon matrix for the sample prepared from the original system with 0.40 mL APTES (the sample is denoted as HPC1, see Fig. S8a, ESI†), while we can also see individual HPC spheres in the samples prepared from the original systems with 0.80 mL and 1.20 mL APTES (the samples were named as HPC2 and HPC3, see Fig. S8b and c, ESI†). It may be mainly because of the microstructures of the silica capsules. As to the precursor capsules of HPC1, there are less silica nanoparticles in the capsules (Fig. 2d). During the impregnation and solidification process, the silica capsules will shrink and deform a bit more than the other two samples (Fig. 2e and f). Moreover, the solid content of silica capsules dispersion also increases with the increase of APTES content. As a result, the resulting HPC1 spheres are nearly immersed in the carbon matrix.

The ordered mesoporous structures could be clearly observed from the TEM images of HPC spheres (Fig. 4a–c). Moreover, with the increase of APTES content, the ordered mesopores in the resulting HPC spheres turn better in terms of both quality and quantity. It is further proved by nitrogen adsorption/desorption measurements (Fig. 4d and e). The BET surface area of HPC spheres increases from 408 m² g⁻¹ to 759 m² g⁻¹ when the APTES content increasing from 0.40 mL to 1.20 mL, as well as their total pore volume (Table S2†). The XRD patterns indicate HPC2 and HPC3 have an intense diffraction peak at a 2θ of around 0.8° (Fig. S9b and c, ESI†), while there is no peak in the XRD pattern of HPC1 (Fig. S9a, ESI†) since the ordered mesoporous structures are only a small portion of HPC1. To explain this interesting phenomenon, we need to understand the formation process of ordered mesoporous polymer. Here the ordered mesoporous polymer composites (OMPC) are prepared for the respective HPC spheres, by a solvent evaporation induced self-assembly method.^14,38 Fig. S10† shows the digital photos of the dried polymer composites. OMPC1 looks like semi-transparent gel or plastics, while OMPC3 is like a newly baked cake, and OMPC2 displays the transition appearance. It is mainly because that OMPC3 dries much faster than OMPC1, and OMPC2 is in the middle. The evaporation rate largely depends on the microstructures of silica capsules. The capsules prepared at 1.20 mL APTES behave like rigid porous spheres (Fig. 2f and S3f, ESI†), so ethanol could easily escape from the capsules and the assembled structures of the capsules. However, the capsules prepared at 0.4 mL APTES behave like soft porous beads (Fig. 2d and S3d, ESI†), which will interact with one another to form a network (e.g. semi-transparent gel) during the evaporation process, and consequently slow down the evaporation of ethanol. It is believed that the evaporation rate of ethanol will not only impact the appearances of the composites, but their microstructures. Since ethanol is evaporated out slowly, Pluronic P123 and low MW resol may migrate from the bulk phase to the surface. It is proved by the digital photos of OMPC1 and OMPC2 (Fig. S10b and c, ESI†). Therefore, it will induce the change to the self-assembly structures of Pluronic P123 in the polymer composites, and eventually result in less ordered mesopores in HPC1. The SEM images of cross-sections of HPC spheres (Fig. 5) also indicate their microstructure evolution. It could be found that the HPC spheres are nearly perfect replicas of their respective hierarchical silica capsules. With the increase of APTES content, there are more and more pores in the HPC spheres, which are mainly templated from primary silica nanoparticles or their aggregates.

Fig. 4 Microstructure characterization of the hierarchical porous carbons. (a) TEM images of HPC1 prepared from the hierarchical silica based on the system with 0.40 mL APTES. (b) TEM images of HPC2 prepared from the hierarchical silica based on the system with 0.80 mL APTES. (c) TEM images of HPC3 prepared from the hierarchical silica based on the system with 1.20 mL APTES. (d) Nitrogen adsorption–desorption isotherms of HPCs. (e) Pore size distribution of HPCs determined by Barrett–Joyner–Halenda (BJH) analysis. Scale bars, 200 nm for a1, b1 and c1; 50 nm for a2, b2 and c2.

Fig. 5 Microstructure evolution of the hierarchical porous carbon (HPC) spheres. The samples were cut by a scalpel blade before observed by SEM. (a) HPC1 prepared from the hierarchical silica based on the system with 0.40 mL APTES. (b) HPC2 prepared from the hierarchical silica based on the system with 0.80 mL APTES. (c) HPC3 prepared from the hierarchical silica based on the system with 1.20 mL APTES. Scale bars, 50 μm for a1, b1 and c1; 5 μm for a2, b2 and c2; 0.5 μm for a3, b3 and c3.

From Fig. 4d, we could also observe a steep increase in the adsorbed volume at very low relative pressure. It indicates there are abundant micropores in the HPCs, which are similar with many porous carbon materials reported earlier.^13,22,39,40 The unordered micropores are mainly generated from the PEO segments embedded in the phenolic resin walls during carbonization,²² since PEO blocks have good affinity with the precursors. Besides the micropores, Fig. 4e displays the broader pore size distribution of HPC spheres from 2 nm to 100 nm. The ordered mesopores of around 2.8 nm are templated from the self-assembly structure of Pluronic P123, which was decomposed at around 350 °C during calcination.²⁰ The ordered mesopores can be found in all the three kinds of HPC spheres. In accordance with the evolution of the primary size of silica nanoparticles, we can find that in the pore size distribution curves (Fig. 4e), there is a peak centred at around 18.0 nm for HPC1, a peak centred at around 10.0 nm for HPC2 and a peak centred at around 5.0 nm for HPC3 respectively. These pore sizes seem a little bit smaller than their respective silica nanoparticles, owing to the shrinkage of the matrix materials during the solidification and carbonization processes. The pores at around 18 nm in HPC1 can be clearly observed (Fig. 4a). However, it is a bit difficult to find single pores in HPC2 and HPC3 (Fig. 4b and c), since most of the pores are interconnected with one another, which are actually templated from the aggregates of silica nanoparticles. As a result, the pore size distributions of HPC2 and HPC3 are very broad. Especially for HPC3, there is a peak centred at around 50 nm (Fig. 4e), and the pores can be clearly observed in the TEM images (Fig. 4c). It is concluded that the pores at around 50 nm are largely determined by the most probable size of the aggregates formed by silica nanoparticles. By now, hierarchically porous carbon spheres have been successfully fabricated with integrated micro–meso–macro pores. It is worth noting that the pore sizes can be tuned as required by simply changing the APTES content.

The Raman spectra of the HPC samples show two strong peaks at 1345 and 1585 cm⁻¹, which can be assigned to the D and G bands of the disordered and graphitized carbons, respectively (Fig. S11† presents the spectrum of HPC2 as an example, ESI†). The similar intensity of the G band to that of the D band (intensity ratio of G to D ∼1) suggests a high degree of graphitization for the HPC samples, which would render the HPCs good electrical conductivity that is required for electrode materials of energy devices.^41,42

The hierarchically porous carbon spheres with large specific surface areas are expected to have potential applications as electrode materials of supercapacitors. The electrochemical performance of HPCs was evaluated in a three-electrode system with 5 M KOH electrolyte. Cyclic voltammetry (CV) curves measured at different scan rates exhibit a typical double-layer capacitive behavior with high capacitance (Fig. S12, ESI†). Moreover, the galvanostatic charge/discharge measurements gave nearly triangular charge/discharge curves at different current densities (Fig. 6a), indicative of an efficient ion transport throughout the HPC electrodes, confirming the typical double-layer capacitive behavior. Based on the discharge curve, the specific capacitance (C_s) of HPC2 was estimated to be ∼170 F g⁻¹ at 0.2 A g⁻¹ (Fig. 6b), according to the equation: C_s = IΔt/ΔVm, where I denotes the current; Δt is the discharge time; ΔV represents the voltage change during the discharge; and m expresses the total mass of the active materials in electrode.⁴³ The specific capacitance of HPC2 is slightly higher than that of HPC3 due to the similar specific surface areas, while it is much larger than that of HPC1 owing to the greater surface area. Moreover, the capacitance of HPC2 is also superior to those of many reported porous carbon materials, such as Ni-doped microporous carbon (154 F g⁻¹),⁴⁴ partially graphitized porous carbon (148 F g⁻¹),⁴⁵ hierarchical porous carbon aerogel derived from bagasse (142 F g⁻¹),⁴⁶ microwave exfoliated graphite oxide (154 F g⁻¹),⁴⁷ among others.

Fig. 6 Electrochemical performance of HPC materials as electrodes for supercapacitors. (a) Galvanostatic charge/discharge curves of HPC2 at different current densities in 5 M KOH electrolyte. (b) Capacitance rate performance of different HPC samples. (c) Cycling stability evaluated at a current density of 0.2 A g⁻¹. (d) Nyquist plots of the HPC samples.

The rate capacity of HPCs was also evaluated (Fig. 6b). The specific capacitances of the samples reduced quickly to <60 F g⁻¹ when the current density increased to 1.0 A g⁻¹. The capacitance reduction at increased current densities is owing to the ion transport limitation, which often occurs in electrode materials at high current densities.³⁰ It is well-known that the rate capability of electrode materials is mainly determined by the kinetics of ion diffusion and electronic conductivity. At lower current densities, the ions can sufficiently diffuse into almost all holes of the electrode. While the effective contact between the ions and the electrode is greatly reduced at higher current densities, leading to a lower capacitance. The rate capacity curves suggest that the HPC materials are suitable for the capacitive applications at low current densities. Finally, the cycling stability of the HPCs, e.g. HPC2, was evaluated by galvanostatic charge/discharge measurements at 0.2 A g⁻¹. Notably, almost no capacitance loss was observed after 5000 cycles, indicating an excellent cycling stability of HPC2 as electrode materials of supercapacitors (Fig. 6c). In the cycling curve, a part of slight enhanced capacitance may be attributed to the improvement of ion accessibility in the carbon frameworks during the cycling process, which leads to an increased accommodation behavior for charges.³¹

To further understand the satisfactory capacitive performance of HPC2, electrochemical impedance spectra (ESI†) of the HPCs were recorded (Fig. 6d). These spectra show obvious semicircle regions, which arise from charge-transfer resistance of electrode materials.^42,43 The spectra indicate that a smaller resistance (∼0.4 Ω) of HPC2 than those of HPC1 (∼1.1 Ω) and HPC3 (∼0.7 Ω). The lower resistance plays a crucial role in enhancing the electrical conductivity and power output of supercapacitors.^42,43

The good electrochemical performance of HPC2 is considered to arise from its hierarchically porous structure, which provides a synergistic effect of macro- and meso-/micropores.^31,41 First, the macropores and large mesopores may buffer ions and thus shorten the diffusion distances from the external electrolyte to the interior surfaces. Second, the micropores in the carbon walls can enhance ion transport and charge storage. Third, the conductive interconnected carbon walls which construct 3D frameworks may serve as multidimensional pathways to facilitate the transport of electrons in the bulk electrode. On the other hand, HPC2 has a larger specific surface area than that of HPC1, while in comparison to HP3 with mesopores of ∼5 nm mean size, HPC2 possesses a similar surface area and larger mesopores with an average diameter of ∼10 nm (Fig. 4e), which facilitate smooth mass transport and entirely access by electrolytes in the charge/discharge process such that the exposed surface in the larger mesopores can be efficiently used for charge storage.⁴⁸ Moreover, HPC2 shows better electrical conductivity than those of the other two samples. These factors contribute to the better capacitive performance of HP2 as electrode materials for supercapacitors.

Conclusions

In summary, here we demonstrate a new approach for the fabrication of hierarchical porous carbon spheres with multi-level porous structures, by using hierarchical silica capsules as templates. The pores are tuneable and well interconnected, ranging from 3 nm to 100 nm. It is very critical to create the hard template with tuneable microstructures, e.g. HSCs constructed by silica nanoparticles. As a result, the porous structures of HPC spheres can be simply tuned by changing the size of silica nanoparticles, that is, the ratio of TEOS and APTES. The HPC spheres have high specific surface areas of up to 760 m² g⁻¹. Serving as electrode materials of supercapacitors, the HPC spheres exhibit good electrochemical performance with specific capacitance of up to 170 F g⁻¹ and excellent cycling stability with no capacitance loss after 5000 cycles.

Experimental

Materials

Pristine CaCO₃ rods were supplied by Maruo Calcium Co., Ltd (Japan), with diameters of 0.5–1.0 μm and lengths in the region of 10–30 μm (Fig. S1, ESI†). Modified CaCO₃ rods were prepared according to the method reported previously.^32,33 Poly(propylene oxide)-block-poly(ethylene oxide)-block-poly(propylene oxide) triblock copolymer Pluronic P123 (EO₂₀PO₇₀EO₂₀), APTES and polyvinylpyrrolidone (PVP) with average molecular weight of 40 000 were purchased from Sigma-Aldrich. Paraffin oil, TEOS, ethanol, tetrahydrofuran (THF), phenol, formalin solution (37 wt%), sodium hydroxide, oleic acid, hydrochloric acid (HCl) and hydrofluoric acid (HF) were purchased from Sinopharm Chemical Reagent Co., Ltd (China). The resol precursor was prepared by using the literature method.²⁰ Deionized water was produced by using Millipore. All reagents were used as received without further purification.

Synthetic procedures

Preparation of hierarchical silica capsules. Paraffin oil (16.0 mL), TEOS (4.0 mL) and a certain amount of APTES (ranging from 0.04 mL to 1.20 mL) were mixed together to form oil phase in a 50 mL centrifuge tube. Modified CaCO₃ rods (0.6 g) were added to the mixed oil phase, and shaken by hand for 10 s to obtain a homogeneous dispersion. CaCO₃-rod stabilised w/o emulsion was prepared by mixing the oil dispersion with 2.5 wt% PVP aqueous solution (10.0 mL) and shaking for 30 s by hand. Since PVP can facilitate the growth of silica on the surfaces of colloidal particles,³⁶ we assume that PVP could also help with the formation of silica at the oil–water interface. The emulsion was aged at 60 °C for 20 h, and then filtrated and washed with THF (4 × 50 mL). After the removal of CaCO₃ rods with 0.5 M HCl solution, hierarchical silica capsules have been successfully prepared from the systems with different APTES contents. The capsules were washed by deionized water (5 × 100 mL) and ethanol (3 × 50 mL) in turn to obtain concentrated ethanol dispersions of silica capsules (15 mL) for further use. A series of control samples of silica capsules have also been prepared by simply using deionized water (10.0 mL) to replace 2.5 wt% PVP aqueous solution (10.0 mL) as the inner phase of w/o emulsions.

Preparation of hierarchical porous carbons. HPC spheres were prepared by a reported method with slight modifications.^20,37 For a typical procedure, concentrated ethanol dispersions of silica capsules (2.0 mL) were dispersed into an solution containing ethanol (10 mL), resol (1.0 g) and P123 (0.5 g). The resulting dispersions were placed in a 25 mL beaker, and the depth of the dispersions was around 15 mm. After totally evaporating out ethanol at ambient (20 °C) for about 15 h, the resulting cakes were further solidified at 100 °C for 24 h. With the presence of nitrogen, the silica/resol/P123 composites were calcined at 350 °C for 2 h at a heating rate of 1 °C min⁻¹ to remove P123 template, and further carbonized at 900 °C for 2 h at a heating rate of 5 °C min⁻¹. After the removal of silica by the mixed solvent of 5% HF (15 mL) and ethanol (5 mL), the hierarchical porous carbons were washed with deionized water (5 × 100 mL) and dried at 100 °C for 24 h.

Characterization

Hitachi Model S-4800 field emission scanning electron microscope (Japan) was used to collect scanning electron microscopy (SEM) images. Cryo-SEM images of silica capsules were also collected by Hitachi Model S-4800 equipped with Gatan Alto 2500. A small amount of silica aqueous dispersion was placed onto a copper holder, and rapidly immersed into liquid nitrogen for 15 s. Right after the freezing process, frozen samples were transferred into the attached cryo preparation chamber, fractured using a chilled scalpel blade, and then etched at −90 °C for 5 min. The specimen was coated with a sputtered film of platinum before characterization. Transmission electron microscopy (TEM) images were collected by a Tecnai G2 spirit BioTWIN electron microscope at 120 kV (FEI Company, USA). TEM samples were ultrasectioned directly from HPCs using a Leica UC6 Ultramicrotome (Leica, Wetzlar, Germany). The sections with the thickness of 80 nm were placed on a carbon-coated TEM grid for TEM analysis. Optical photographs were obtained using a Leica optical microscope (DM LB 2, Germany). Micromeritics ASAP 2010C system (USA) was used to measure nitrogen sorption isotherms at 77 K. All the samples were degassed under vacuum at 200 °C for 4 h before analysis. The Brunauer–Emmett–Teller (BET) method was utilized to calculate the specific surface areas (P/P₀ ranging from 0.05 to 0.25). The total pore volume (V_t) was obtained from the adsorbed amount at a relative pressure P/P₀ of 0.995. The pore-size distribution was calculated from the adsorption branch by the BJH model. X-ray diffraction (XRD) patterns were measured by a Bruker APLX-DUO single crystal diffractometer equipped with Cu Kα radiation. The patterns were recorded from 0.5° to 5° with 0.3° steps at 10 s per point.

Electrochemical measurements. Electrochemical performance of the HPC spheres as electrode materials of supercapacitor was evaluated on an EG & potentiostat/galvanostat model 2273 advanced electrochemical system. CV and charge–discharge galvanostatic measurements were performed in a three-electrode cell system. Working electrodes were prepared by mixing 80 wt% powder active materials (2 mg), 10 wt% carbon black (Mitsubishi Chemicals, Inc.), and 10 wt% polytetrafluoroethylene (PTFE) binder. These compositions were weighed by a highly accurate balance (Mettler Toledo XS105). Nickel foam was applied as a counter electrode with an Ag/AgCl electrode as a reference electrode. The experiments were carried out in a 5 M KOH solution. The potential range was between −1 to 0 V (vs. Ag/AgCl) at different scan rates and different current densities at ambient temperature. Nyquist plots of the samples were recorded by applying a sine wave with amplitude of 5.0 mV over a frequency range of 100 kHz to 0.01 Hz.

Acknowledgements

This work was sponsored by the National Basic Research Program of China (2015CB931801), National Natural Science Foundation of China (21374062, 21304057 and 51573091), China Scholarship Council (CSC) (201506235066), and the Program for Eastern Scholar in Shanghai. S. D. S. acknowledges financial support from European Cooperation in Science and Technology Actions MP1305 and MP1106.

Notes and references

1. L. L. Zhang and X. S. Zhao, Chem. Soc. Rev., 2009, 38, 2520.
2. M. Sevilla and R. Mokaya, Energy Environ. Sci., 2014, 7, 1250.
3. S. Dutta, A. Bhaumik and K. C. W. Wu, Energy Environ. Sci., 2014, 7, 3574.
4. W. Z. Shen and W. B. Fan, J. Mater. Chem. A, 2013, 1, 999.
5. H. Jiang, P. S. Lee and C. Z. Li, Energy Environ. Sci., 2013, 6, 41.
6. R. Ryoo, S. H. Joo, M. Kruk and M. Jaroniec, Adv. Mater., 2001, 13, 677.
7. L. Qie, W. M. Chen, H. H. Xu, X. Q. Xiong, Y. Jiang, F. Zou, X. L. Hu, Y. Xin, Z. L. Zhang and Y. H. Huang, Energy Environ. Sci., 2013, 6, 2497.
8. Y. F. Zhao, W. Ran, J. He, Y. F. Song, C. M. Zhang, D. B. Xiong, F. M. Gao, J. S. Wu and Y. Y. Xia, ACS Appl. Mater. Interfaces, 2015, 7, 1132.
9. Z. Chen, J. Wen, C. Z. Yan, L. Rice, H. Sohn, M. Q. Shen, M. Cai, B. Dunn and Y. F. Lu, Adv. Energy Mater., 2011, 1, 551.
10. D. W. Wang, F. Li, M. Liu, G. Q. Lu and H. M. Cheng, Angew. Chem., Int. Ed., 2008, 47, 373.
11. J. E. Hampsey, Q. Y. Hu, L. R. Rice, J. B. Pang, Z. W. Wu and Y. F. Lu, Chem. Commun., 2005, 25, 3606.
12. A. Taguchi, J.-H. Smått and M. Lindén, Adv. Mater., 2003, 15, 1209.
13. H. Yamada, H. Nakamura, F. Nakahara, I. Moriguchi and T. Kudo, J. Phys. Chem. C, 2007, 111, 227.
14. W. Xia, B. Qiu, D. G. Xia and R. Q. Zou, Sci. Rep., 2013, 3, 1935.
15. R. Ruiz-Rosas, M. J. Valero-Romero, D. Salinas-Torres, J. Rodriguez-Mirasol, T. Cordero, E. Morallon and D. Cazorla-Amoros, ChemSusChem, 2014, 7, 1458.
16. P. Strubel, S. Thieme, T. Biemelt, A. Helmer, M. Oschatz, J. Brückner, H. Althues and S. Kaskel, Adv. Funct. Mater., 2015, 25, 287.
17. S. Chaudhari, S. Y. Kwon and J. S. Yu, RSC Adv., 2014, 4, 38931.
18. L. F. Chen, X. D. Zhang, H. W. Liang, M. G. Kong, Q. F. Guan, P. Chen, Z. Y. Wu and S. H. Yu, ACS Nano, 2012, 8, 7092.
19. Z. Y. Wang, F. Li, N. S. Ergang and A. Stein, Chem. Mater., 2006, 18, 5543.
20. Y. Meng, D. Gu, F. Q. Zhang, Y. F. Shi, H. F. Yang, Z. Li, C. Z. Yu, B. Tu and D. Y. Zhao, Angew. Chem., Int. Ed., 2005, 44, 7053.
21. F. F. Wang, R. R. Song, H. H. Song, X. H. Chen, J. S. Zhou, Z. K. Ma, M. C. Li and Q. Lei, Carbon, 2015, 81, 314.
22. Y. Huang, H. Q. Cai, D. Feng, D. Gu, Y. H. Deng, B. Tu, H. T. Wang, P. A. Webley and D. Y. Zhao, Chem. Commun., 2008, 23, 2641.
23. X. D. Huang, B. Sun, D. W. Su, D. Y. Zhao and G. X. Wang, J. Mater. Chem. A, 2014, 2, 7973.
24. Y. Fang, D. Gu, Y. Zou, Z. Wu, F. Li, R. Che, Y. Deng, B. Tu and D. Y. Zhao, Angew. Chem., Int. Ed., 2010, 49, 7987.
25. N. Brun, S. R. S. Prabaharan, C. Surcin, M. Morcrette, H. Deleuze, M. Birot, O. Babot, M. F. Achard and R. Backov, J. Phys. Chem. C, 2012, 116, 1408.
26. D. D. Zhou, H. J. Liu, Y. G. Wang, C. X. Wang and Y. Y. Xia, J. Mater. Chem., 2012, 22, 1937.
27. X. R. Wen, D. S. Zhang, L. Y. Shi, T. T. Yan, H. Wang and J. P. Zhang, J. Mater. Chem., 2012, 22, 23835.
28. X. R. Wen, D. S. Zhang, T. T. Yan, J. P. Zhang and L. Y. Shi, J. Mater. Chem. A, 2013, 1, 12334.
29. L. Estevez, R. Dua, N. Bhandari, A. Ramanujapuram, P. Wang and E. P. Giannelis, Energy Environ. Sci., 2013, 6, 1785.
30. Z. Wu, Y. Sun, Y. Tan, S. Yang, X. Feng and K. Müllen, J. Am. Chem. Soc., 2012, 134, 19532.
31. X. Yang, X. Zhuang, Y. Huang, J. Jiang, H. Tian, D. Wu, F. Zhang, Y. Mai and X. Feng, Polym. Chem., 2015, 6, 1088.
32. W. Z. Zhou, G. S. Tong, D. L. Wang, B. S. Zhu, Y. Ren, M. Butler, E. Pelan, D. Y. Yan, X. Y. Zhu and S. D. Stoyanov, Small, 2016, 12, 1797.
33. W. Z. Zhou, J. Cao, W. C. Liu and S. D. Stoyanov, Angew. Chem., Int. Ed., 2009, 48, 378.
34. C. E. Fowler, D. Khushalani and S. Mann, J. Mater. Chem., 2001, 11, 1968.
35. X. L. Yang, N. Zhao, Q. Z. Zhou, Z. Wang, C. T. Duan, C. Cai, X. L. Zhang and J. Xu, J. Mater. Chem., 2012, 22, 18010.
36. C. Graf, D. L. J. Vossen, A. Imhof and A. van Blaaderen, Langmuir, 2003, 19, 6693.
37. H. J. Liu, X. M. Wang, W. J. Cui, Y. Q. Dou, D. Y. Zhao and Y. Y. Xia, J. Mater. Chem., 2010, 20, 4223.
38. Y. F. Lu, R. Ganguli, C. A. Drewien, M. T. Anderson, C. J. Brinker, W. L. Gong, Y. X. Guo, H. Soyez, B. Dunn, M. H. Huang and J. I. Zink, Nature, 1997, 389, 364.
39. W. Lu, M. Liu, L. Miao, D. Zhu, X. Wang, H. Duan, Z. Wang, L. Li, Z. Xu, L. Gan and L. Chen, Electrochim. Acta, 2016, 205, 132.
40. Y. Zhao, M. Liu, X. Deng, L. Miao, P. K. Tripathi, X. Ma, D. Zhu, Z. Xu, Z. Hao and L. Gan, Electrochim. Acta, 2015, 153, 448.
41. M. Liu, X. Wang, D. Zhu, L. Li, H. Duan, Z. Xu, Z. Wang and L. Gan, Chem. Eng. J., 2017, 308, 240.
42. M. Liu, X. Ma, L. Gan, Z. Xu, D. Zhu and L. Chen, J. Mater. Chem. A, 2014, 2, 17107.
43. Z. Lin, H. Tian, F. Xu, X. Yang, Y. Mai and X. Feng, Polym. Chem., 2016, 7, 2092.
44. Q. Lei, H. Song, D. Zhou, S. Zhang and X. Chen, RSC Adv., 2015, 5, 78526.
45. L. Ye, Q. Liang, Z. H. Huang, Y. Lei, C. Zhan, Y. Bai, H. Li, F. Y. Kang and Q. H. Yang, J. Mater. Chem. A, 2015, 3, 18860.
46. P. Hao, Z. Zhao, J. Tian, H. Li, Y. Sang, G. Yu, H. Cai, H. Liu, C. P. Wong and A. Umar, Nanoscale, 2014, 6, 12120.
47. M. Ghaffari, Y. Zhou, H. Xu, M. Lin, T. Y. Kim, R. S. Ruoff and Q. Zhang, Adv. Mater., 2013, 25, 4879.
48. H. Tian, Z. Lin, F. Xu, J. Zheng, X. Zhuang, Y. Mai and X. Feng, Small, 2016, 12, 3155.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra24524b

---