1 Introduction

The increasing markets of electrical vehicles, cell phones and new electronic devices make it urgently desirable to develop and optimize rechargeable batteries with high energy density [1,2,3,4]. Presently, rechargeable lithium ion batteries (LIB) hold the dominant share in the markets for portable electronic devices. Nevertheless, the energy and power density of LIBs still need significant improvements in order to support electric cars. In such a situation, developing higher level battery systems with higher energy density and safety through new configurations and electrode materials are of tremendous significance [5,6,7,8,9,10,11].

Sulfur has a set of advantages, viz., cheap, abundant and environment friendly [12]. Especially, sulfur with theoretical capacity of 1672 mAh g−1, almost ten times the capacity of traditional LIBs, is a charming cathode material [13,14,15]. Howbeit, the commercial progress of Li–S battery is impeded by its poor electrical conductivity, accompanied by low utilization of elemental sulfur. Another big issue for Li–S battery is the high dissolubility of the intermediate products called lithium polysulfides (Li2Sx, 4 ≤ x ≤ 8), which can reduce the cycle stability of battery [16,17,18,19]. The product (Li2S2 and Li2S) which deposited uncontrollably are acknowledged as another important reason for accelerated capacity deterioration and low Coulombic efficiency in Li–S batteries [20,21,22]. These deficiencies remarkably undermine energy density of Li–S battery.

To eliminate these shortcomings, substantial efforts have been made to design host materials such as carbon materials [5, 11, 23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38] and nanometal oxides [6]. These approaches have been considered to boost the electrical conductivity of sulfur materials and confine soluble polysulfides. For all that, the Li–S battery has not been available in practical application till today, because most of the reported electrodes have low areal mass loading, unavoidably decreasing the real capacity of the Li–S battery systems. Up till now, many efforts have been made to increase the areal mass loading, such as introducing 3D structure hosts [11, 26]. 3D graphene oxides with abundant porous structure have been applied as host materials to achieve high areal mass loading and excellent contact with sulfur for the sake of better electron transport. Particularly, Rehman et al. announced an interconnected and vertically aligned porous carbon nanosheets with good electrical conductivity, which show promising electrochemical properties as a sulfur host [11]. All in all, exploring efficient approaches to encapsulate sulfur into 3D carbon matrix has a bright prospect [24, 25].

Here, we introduce a simple method to synthesize three-dimensional porous graphene-wrapped porous carbon using PAA cation-exchange resin as raw material with nickel ions as catalysts. A melting diffusion approach was used to incorporate sulfur into 3D-G porous structure. The amount of sulfur mass can be strictly controlled by the impregnation approach. The key feature of 3D-G@S is that they have traits of both porous carbon and graphene: (a) 3D-G possesses high surface area of 1375.08 m2 g−1, which in its unique porous structure allows high sulfur loading for improving the energy density; (b) in addition, it can buffer the volume variation during the conversion processes between sulfur and its discharged products, preserve the integrity of the electrode structure, and boost the cycling performance of the batteries; (c) the graphene-like structure on the surface confines sulfur within the cathode and suppresses the free movement of the polysulfide species; (d) 3D-G helps to improve the electrical conductivity, decrease surface impedance and boost rate capability.

2 Materials and methods

Sodium hydroxide (AR), potassium hydroxide (AR), Nickel(II) acetate tetrahydrate (AR) and sulfur powders (AR) were bought from Aladdin. Polyacrylic acid cation-exchange resin was bought from Shanghai Hualing Company. Ethanol (AR) was bought from Tianjin Damao Company. All the chemicals were used with no further modification.

2.1 Material synthesis

2.1.1 Synthesis of 3D-G

Polyacrylic acid cation-exchange resin (500 g) was initially cleaned using deionized water, and immersed in 1 M HCl solution for more than 10 h. It was then washed with deionized water and dried at 70 °C in blast drying oven. The prepared hydrogen ion-exchange resin was immersed in 1 M NaOH solution for 10 h, which were washed and dried as above. 10 g obtained powders were added into 100 ml nickel acetate solution (0.05 mol L−1) to get nickel ion-exchange resin.

The KOH-ethanol solution was prepared by dissolving 40 g KOH in 400 ml anhydrous ethanol. The as-synthesized resin was added in the KOH-ethanol solution, then stirred and heated at 80 °C in water bath until complete vaporization of ethanol to form the 3D-G precursor. These precursor materials were kept at room temperature overnight, and then placed in the 70 °C blast drying oven for 6 h. After drying, the powder was heated for 2 h at 850 °C in Ar atmosphere with the heating-rate of 2 °C min−1 to produce the 3D-G material. These 3D-G materials were dissolved in 3 mol L−1 HCl in order to remove the residual KOH until neutral, and finally dried at 100 °C for 12 h under vacuum.

2.1.2 Synthesis of 3D-G @S

The as-prepared 3D-G and the elemental sulfur were mixed together in weight ratios of 1:4. Then sulfur/3D-G mixtures were put into a sealed vessel and heated at 155 °C for 12 h to let the melting sulfur absorbed into the pores of 3D-G via diffusion.

2.2 Material characterization

The characterization of the 3D-G and 3D-G@S composites were performed by thermogravimetric analyzer (TGA, TA Instrument SDT2960), Brunauer–Emmett–Teller (BET, ASAP 2020), X-ray diffraction (XRD, RIGAKUD/max-2500), and scanning electron microscopy (SEM, HITACHI S-3500 N).

2.3 Electrochemical measurements

The electrochemical performance of 3D-G@S was assessed by CR2016-type coin cells. The counter electrode is a lithium foil and the separator is a polypropylene microporous film (Celgard 2300). The cathode electrode was fabricated by blending 3D-G@S with PVDF binder and conducting additive (Super P) following the weight ratio of 8:1:1. In the cathode electrode, the mass loading of sulfur was about 3 mg cm−2. And the electrolyte was composed of 1.0 M LiTFSI (lithiumbis(trifluoromethanesulfonyl)imide) in 1,3-dioxolane(DOL) and 1,2-dimethoxyethane (DME) (1:1 in volume ratio, with 1.0 wt% LiNO3 as additive). 0.1 ml electrolyte was used in a coin cell. The cells were tested by Neware SBT7.5 in the voltage scope from 1.5 to 3.0 V at various current rates (1 C = 1675 mA g−1). Cyclic voltammetry (CV) measurements over the potential range of 1.5–3 V at 0.1 mV/s scanning rate and the electrochemical impedance spectrum (EIS) tests with a frequency from 100 kHz to 0.1 Hz were carried out on an electrochemical workstation (Solartron 1287/1260).

3 Results and Discussion

Figure 1 demonstrates the synthetic process of 3D-G. Firstly, polyacrylic acid cation-exchange resin was treated with HCl solution for more than 10 h to produce hydrogen ion-exchange resin. The prepared hydrogen ion-exchange resin was immersed in 1 M NaOH solution for 10 h to obtain sodium ion-exchange resin, which was then immersed in nickel acetate solution to get nickel ion-exchange resin. At high temperature nickel works as catalytic site to aid the growth of carbon. In the next step, direct carbonization of nickel ion-exchange resin with KOH as activator produces highly interconnected and porous 3D-G.

Fig. 1
figure 1

Schematics illustration on synthesis method for 3D-G

Full size image

The morphology of the pure resin carbon and 3D-G were characterized by FE-SEM. Figure 2a, b are the SEM images of pure resin carbon with different magnifications and exhibit that big pieces of granular carbon join together. Figure 2c, d show that the as-prepared 3D-G possesses white fungus-like structure with wrinkled graphene linked together.

Fig. 2
figure 2

The SEM images for pure resin carbon: a low resolution; b high resolution; SEM images for 3D-G: c low resolution; d high resolution

Full size image

The formation of the special 3D-G structure can be described by the cooperation effect of Ni atoms and KOH. It has been proclaimed that Ni and Cu perform as the most promising graphene catalytic metals via a dissolution–precipitation mechanism [39]. Blakely and coauthors have operated extensive studies on the formation of carbon films. They found that a single layer graphite forms first through carbon segregation, followed by precipitation of more carbon to grow graphite [39, 40].

Firstly potassium carbonate forms (at ≤ 600 °C), which is then decomposed to K2O and CO2 (at ≤ 700 °C). Finally, at T ≤ 800 °C, the produced CO2 reacts with carbon to produce porosity at high temperature [41]. Thanks to Ni catalysis and KOH activation, 3D-G has surface area as high as 1375.08 m2 g−1, which can offer enough space for high loading amount of sulfur to improve the energy density of lithium–sulfur battery. Unlike the conventional two-dimensional porous carbon sheets, 3D-G is the assembly of inner porous carbon and outer interconnected carbon nanosheets. This unique structure serves as a cage for trapping polysulfide anions and is highly important for alleviating the dissolution of polysulfides during battery cycling.

For further confirmation, X-ray diffraction was conducted. XRD patterns for pure S, 3D-G, 3D-G@S, and resin carbon are shown in Fig. 3a. Pure resin carbon has obvious sharp peaks at 26° and 43°, which correspond to the graphite phase (002) and (110) crystal plane, indicating that carbon generated from pure resin has good graphitic carbon structure. However, after dealing with the nickel catalytic reaction and KOH activation, the sharp peak at 2θ = 26° shifts to the left, demonstrating that the 3D-G spacing has become broader, which is mainly due to the catalytic function of nickel ions. In the forming process of 3D-G, nickel atom reacts with carbon to form nickel carbide at low temperature. When the temperature rises to 500 °C, nickel carbide is further decomposed to nickel and a stable crystal lattice of carbon, which accelerate the rearrangement and expansion of carbon layer. In addition, peak for crystalline sulfur is not observed in 3D-G@S, verifying that sulfur is completely distributed within the porous carbon structure and graphene sheets.

Fig. 3
figure 3

a XRD patterns for 3D-G, 3D-G@S and resin carbon; b nitrogen adsorption–desorption for 3D-G and 3D-G@S; c the distribution of pore size for 3D-G; d TGA profile image of pure S and 3D-G@S

Full size image

Nitrogen BET adsorption were used to measure the specific surface area of 3D-G and 3D-G@S composites. The nitrogen adsorption–desorption isotherms of 3D-G and 3D-G@S, are displayed in Fig. 3b, respectively. The shape of 3D-G isotherm is described as type I and indicates the presence of pore sizes varying from micropores to macropores. The quick increase at a comparatively low pressure of P/P0 (0 ≤ P/P0 ≤ 0.1, where P and P0, correspond, respectively, to the equilibrium and saturation pressure for N2 at 77 K), indicates the existence of micropores, which is based on the behaviors of nitrogen filling in micropores during the adsorption process. In the area P/P0 ≥ 0.1, the gas adsorption slowly increases, which reveals that the material has less mesoporous structure. As the pressure reaches to 1.0, a final increasing tail means the existence of macropores. It can be seen in Fig. 3c that 3D-G was primarily composed of micropores with a narrow pore-size distribution of about 0.7 nm. Compared with 3D-G, the gas adsorption curve of the 3D-G@S isotherm slowly increases, indicating that the micropores of the 3D-G material are almost filled with sulfur. As the relative pressure is close to 1.0, the slope of the curve is nearly not going up, which means that the macropores have also been stuffed with sulfur. It can be seen from Table 1 that the specific surface area of the materials decreased from 1375.08 to 128.41 m2 g−1, and the overall pore volume decreased from 0.82 to 0.04 cm2 g−1, which supports that the encapsulated sulfur is well dispersed into the pores of 3D-G.

Table 1 Comparison between 3D-G and 3D-G@S
Full size table

In Fig. 3d, TGA curve shows that pure sulfur begins to vaporize from ≈ 180 °C and stops at ≈ 300 °C, while 3D-G@S begins to vaporize at ≈ 250 °C and completes at ≈ 360 °C. The contents of sulfur in the composites are calculated to be 75.02%. Pure S has a lower sulfur vaporizing temperature than 3D-G@S because the interaction between sulfur and carbon-based matrix makes the evaporation of sulfur more difficult.

After confirmation from TGA that sulfur is well dispersed into the pores of 3D-G, now it is important to emphasize the morphology of materials. Figure 4a displays the morphology of large bulk of pure sulfur. Figure 4b, c compare the SEM images of 3D-G and 3D-G@S. It can be seen that graphene-like lamellar structure on the original 3D-G has disappeared after sulfur loading. It indicates the uniform encapsulation of sulfur within the graphene lamellar or graphene lamellar stacked holes, which agrees with the XRD diffraction analysis. The element mapping in Fig. 4d shows that 3D-G@S contains some oxygen element in addition to sulfur and carbon element. Therefore, 3D-G@S also consists of small amount of oxygen-containing functional groups, which could help to offer chemical adsorption to polysulfide ions.

Fig. 4
figure 4

The SEM images of a pure S; b 3D-G; c 3D-G@S; d the element mapping of 3D-G@S

Full size image

CV curves of 3D-G@S and pure sulfur were compared in Fig. 5a. During the first negative scan of 3D-G@S, it displays two main reductive peaks at around 2.25 V and 2.04 V. The first reduction peak at about 2.28 V corresponds to the conversion of S8 to long chain lithium polysulfides (Li2Sn, 4 ≤ n ≤ 8) [16]. The second peak at 2.04 V originates from the reduction of lithium polysulfide to Li2S2 and finally to Li2S. At 2.46 V, an anodic peak can be assigned to the oxidation of Li2S2/Li2S to form sulfur. During the second negative scan of 3D-G@S, the reductive peaks were noticed to positively shift from 2.25 to 2.30 V, indicating a reduced polarization in the process. As comparison, in the first cycle of pure sulfur, an oxidation peak was observed at 2.65 V, which was negatively shifted to 2.5 V in the second circle. The polarization and irreversible capacity loss of pure sulfur are obviously larger than 3D-G@S composite. Furthermore, the CV curve of the 3D-G@S composite in the first two circles is better coincident than pure S, which indicates that the material has a better reversible performance.

Fig. 5
figure 5

a The first and second CV curves of pure S and 3D-G@S; b the charge/discharge curves from the first cycle to fifth cycle at 0.1 C for pure S; c the first 200 cycles charge/discharge curves of pure S at 0.1 C; d the charge/discharge curves of 3D-G@S from the first cycle to fifth cycle at 0.1 C; e the first 200 cycles charge/discharge curves of 3D-G@S at 0.1 C

Full size image

Figure 5b, c show the discharge/charge profiles of pure S in the first 5 cycles and the first 200 cycles at 0.1 C rate, respectively. Pure S exhibits an initial discharge capacity 934.2 mAh g−1 and a capacity of 291.9 mAh g−1 in the 200th cycle at 0.1 C. In addition, the large potential difference between the discharge plateau and charge plateau suggested that the polarization of pure S electrode is large, which is consistent with Fig. 5a. Figure 5d, e are the discharge/charge profiles of 3D-G@S in the first 5 cycles and 200 cycles at 0.1 C rate, respectively. It shows that 3D-G@S composite material exhibits an initial discharge capacity as 1066.4 mAh g−1 at 0.1 C. As the cycle went on, capacity decayed slowly. 3D-G@S showed a specific capacity of 701.3 mAh g−1 with capacity retention of 55.6% after 200 cycles. In addition, 3D-G@S composite material shows an obvious activation process. It is found that capacity of 3D-G@S rose in the initial several cycles. The possible reason for this phenomenon is that in the initial several cycles, the wettability between the electrolyte and sulfur is not good as the channels in 3D-G carbon are blocked by sulfur materials. In this case, some sulfur does not react with the electrolyte at the beginning. After several cycles, the electrolyte contacts with sulfur materials thoroughly due to formation of polysulfides and the dredged inner pathway.

Table 2 is the redox potential of the first two CV cycles for pure S and 3D-G@S. 3D-G@S has lower oxidation peak potential than pure S, which means lower polarization in 3D-G@S.

Table 2 The redox potential of CV curves for pure S and 3D-G@S
Full size table

Figure 6a compares the cycling stability between 3D-G@S and pure S at 0.1 C. 3D-G@S has initial discharge capacity of 1066.4 mAh g−1, almost 98.7% coulombic efficiency and 63.7% utilization rate of the active materials. Capacity retention ratio was 54.7% after 200 cycles.

Fig. 6
figure 6

a The cycle performance of 3D-G@S and Pure S at 0.1 C; b the cycle performance of 3D-G@S at 0.2 C; c rate performance at various current rates (1 C = 1675 mA g−1); d the AC impedance diagram of pure S and 3D-G@S before cycles. e The model of equivalent circuit before cycles; Inset in d shows amplified plots from the high to medium frequency range

Full size image

It is clearly shown that the cyclic performance and coulomb efficiency has been boosted after 3D-G was served as a sulfur host. Compared to pure S, the utilization rate of the 3D-G@S composite material was increased by 8.0% and the capacity retention ratio was increased by 13.5%. Thanks to the unique 3D-G structure with high surface area as well as large pore volume, it can be served as effective host for loading sulfur and enhancing their high utilization. Furthermore, graphene-like nanosheets can effectively constrain the polysulfide shuttle effect to achieve stable cyclic life. Figure 6b shows the cycling stability of the 3D-G@S composite cathodes at 0.2 C rate, the initial discharge capacity is 1182.3 mAh g−1 with 70.4% utilization rate of active materials. The 3D-G@S material retains the specific capacity of 539.4 mAh g−1 with 45.6% capacity retention ratio after 200 cycles. Figure 6c demonstrates that the specific capacities at different C-rates of pure S and 3D-G@S, respectively. The cell made from 3D-G@S shows an average discharge capacity of 1250 mAh g−1, 950 mAh g−1, 720 mAh g−1, 680 mAh g−1, at 0.1, 0.2, 0.5, 1 C, respectively. As the current density is turned back to 0.1 from 1 C, a reversible capacity of 1000 mAh g−1 is achieved, indicating good rate performance of the battery. The 3D-G@S composite electrode has higher specific capacity than pure S electrode at different C-rates. Since the insulating pure S has bad electric conductivity, the unique interconnected graphene-like structures of 3D-G improve electric conductivity, and shorten the transport distance for electrons, therefore enhancing the rate capability.

Besides the cycling performance, to evaluate the effects of 3D-G structure, it is also necessary to conduct electrochemical impedance spectroscopy (EIS) to understand the reaction mechanisms during charge and discharge. Figure 6d shows the Nyquist plots for pure S and 3D-G@S, which consist of one depressed semicircle and one straight line in the whole frequency region. Notably, the charge transfer resistance at the interface of 3D-G@S is much smaller than pure S, indicating that 3D-G facilitates the charge transfer process. The equivalent circuits were pictured in Fig. 6e to clarify the interface charge-transfer resistance. The R1 represents a solution resistance and R2 is assigned to be the charge transfer resistance. Meanwhile, the CPE1 is the impedance of double layer capacitor. Diffusion generates an impedance called Warburg impedance. Figure 6e shows that R1 and R2 of pure S electrode are 5.16 Ω and 95.56 Ω, respectively, compared to 1.65 Ω and 11.87 Ω of 3D-G@S electrode. It is obvious that graphene-like structures on 3D-G@S composite materials have smaller charge transfer resistance to decrease the surface impedance and boost the rate capability of 3D-G@S.

4 Conclusion

To summarize, we report a simple, scalable fabrication process for preparing a novel graphene-wrapped porous carbon, by using low cost PAA cation-exchange resin as carbon precursor which introduces nickel ions to work as a catalyst during the carbonization process. It shows that the as prepared carbon materials possess white fungus-like structure. We consider that such a structure could endow sulfur particles with electrically conducting property, and restrict the dissolution of polysulfides. The 3D-G@S composite, as cathode materials for Li–S batteries, showed high specific capacity and good cycling stability. Owing to their unique structure, this 3D-G can be serving as new polysulfides reservoir with improved performance of Li–S batteries for future applications in renewable energy storage.