• lithium–sulfur batteries;
  • metal-organic frameworks;
  • microporous carbon materials;
  • porous materials

As a promising rechargeable battery system, lithium–sulfur (Li–S) batteries can deliver an exceptionally high theoretical specific capacity of 1672 mA h g−1 and an energy density of 2500 Wh kg−1 with the low-cost and environment-friendly sulfur as the cathode material.15 Although the potential use of sulfur as a cathode material has long been discovered, several severe drawbacks have hindered the realization of Li–S batteries.2, 3 One limitation is the insulating nature of sulfur with a very low conductivity of 5×10−30 S cm−1, which results in low utilization of sulfur. Another well-known problem is associated with the easy dissolution of polysulfides, the intermediate products formed during the electrochemical reaction, in organic electrolytes. The dissolved polysulfides “shuttle” between the electrodes, leading to the low Coulombic efficiency and deposition of a highly resistive layer on the surface of electrodes. These detrimental issues result in unsatisfactory electrochemical performance with rapid fading of capacity.

Several approaches have been proposed to overcome the above-mentioned challenges in Li–S batteries, such as developing novel electrolytes and electrode materials.413 Among these efforts, using sulfur-containing composites instead of pure sulfur as the cathode materials has been demonstrated as an effective way towards high-performance Li–S batteries.1422 Polymers and porous carbons are the common candidates to form composites with sulfur, which immobilize the loaded sulfur, and probably also the derived polysulfides via physical and/or chemical interactions. In addition, the electrical conductivity of composite materials is also better than that obtained with pristine sulfur. In particular, porous carbon materials have attracted intensive attention due to their good compatibility with sulfur, easy accessibility, and the abundance of candidates with diverse porosity and structures. Mesoporous carbon materials have been widely studied as the host materials to confine sulfur.45, 18, 23 For example, nanocomposites consisting of sulfur and ordered mesoporous carbon or mesoporous hollow carbon spheres have shown improved sulfur utilization and cycling stability.4, 8 Nonetheless, continuous capacity fading upon prolonged cycling is still commonly observed, and the use of optimized ether-based electrolytes seems to be indispensable. Recent reports on carbon materials with rich micropores have revealed distinct characteristics.24, 25 Sulfur embedded in microporous carbon shows a pronounced discharge plateau at a lower potential of about 1.8 V versus Li+/Li, which is different from the two plateaus of a typical sulfur cathode. More importantly, these microporous carbon/sulfur nanocomposites generally show outstanding capacity retention upon cycling and good compatibility with conventional carbonate-based electrolytes. However, the origins of the unusual characteristics of microporous carbon are not fully understood yet.

In recently years, syntheses of porous carbon materials from metal-organic frameworks (MOFs) or porous coordination polymers (PCPs) have attracted growing attention due to the facile preparation procedures, high carbon yield, and unique porous structures.2630 For example, carbonization of MOF-5 with furfuryl alcohol results in nanoporous carbon, which shows excellent supercapacitive performance.26 The carbon materials with fiber-like morphology prepared from Al-based PCPs exhibit remarkably high porosity.29 In particular, MOFs and PCPs are very attractive as both the template and the precursor for the fabrication of microporous carbon. Compared with many other highly porous carbon materials, such as those prepared by post-activation processes,31 the porous carbon derived from MOFs and PCPs exhibits highly uniform porosity, largely originating from the ordered crystalline structures of the MOFs and PCPs.27, 28 However, the interesting application of these carbon materials derived from MOFs and PCPs for Li–S batteries needs to be further explored.32

Herein, we report the facile synthesis of microporous carbon polyhedrons (MPCPs) using unique MOF polyhedrons as both the template and precursor, and their use as carbon host to incorporate sulfur for Li–S batteries. The as-prepared MPCPs with abundant and uniform micropores serve as an ideal model system for investigating the electrochemical behaviors of sulfur embedded in microporous carbon. Comparative investigations have been carried out to reveal the effects of several important parameters, such as the sulfur loading temperature, sulfur content, and the electrolyte, which offer the opportunity to improve the electrochemical performance of the nanocomposite carbon/sulfur electrodes. Under optimized conditions, the MPCP/sulfur composite can deliver stable cycling performance.

We first prepared a well-studied zeolitic imidazolate framework (ZIF) material, ZIF-8 [Zn(MeIM)2; MeIM=2-methylimidazole] with a unique polyhedral morphology as the MOF precursor.33 Field-emission scanning electron microscopy (FESEM) images given in Figure 1 A show particles with the uniform polyhedral shape obtained from the facile room-temperature synthesis. These polyhedrons are about 2 μm in size with smooth surface. Powder X-ray diffraction (XRD) analysis (see Figure S1 in the Supporting Information) confirms the product as phase-pure ZIF-8 with good crystallinity.33 Transmission electron microscopy (TEM) images (Figure 1 B, C) reveal the solid texture of the ZIF-8 polyhedrons, viewed from different projection directions. According to the FESEM and TEM observations, the geometric shape of the as-prepared ZIF-8 crystals is rhombic dodecahedron with 12 exposed {110} facets, consistent with previous reports.33 Additionally, some slightly truncated rhombic dodecahedrons are also observed in the product. A schematic illustration of the rhombic dodecahedron and truncated rhombic dodecahedron is given in Figure 1 D for better visualization of the polyhedral shape of the ZIF-8 crystals.

thumbnail image

Figure 1. A) FESEM and B, C) TEM images of ZIF-8 polyhedrons. D) Schematic illustration of a rhombic dodecahedron and a truncated rhombic dodecahedron. E, F) FESEM and G, H) TEM images of the MOF-derived microporous carbon polyhedrons (MPCPs).

Download figure to PowerPoint

The MPCPs were then obtained by heating the as-prepared ZIF-8 polyhedrons in an inert gas at 1000 °C. During the annealing process, the ligand of ZIF-8 (2-methylimidazole) is subjected to carbonization, while the Zn2+ species are reduced to metallic Zn that then vaporizes at high temperature.27, 28 FESEM images (Figure 1 E and F) show that the uniform morphology is perfectly preserved after the high-temperature carbonization process, whereas the surface becomes rougher and the size is slightly shrunk. The TEM images reveal the carbon particle exhibits a uniform texture without cracks or a large cavity, suggesting the good stability of the carbon framework. The magnified TEM image (Figure 1 G) on the edge of the polyhedron clearly shows the presence of abundant micropores, while the disordered graphitic layers suggest a relatively low degree of graphitization. The N2 adsorption–desorption isotherm of the MPCPs (Figure 2 A) belongs to typical Type I isotherms with the sharp uptake at the low relative pressure, indicating that the pores in the carbon material are mostly micropores.28 This is further confirmed by the close values of the micropore volume (0.30 cm3 g−1) and the total pore volume (0.34 cm3 g−1). The abundant micropores give rise to a high Brunauer–Emmett–Teller (BET) surface area of 849 m2 g−1. The narrow pore size distribution calculated using the DFT method (Figure 2 B) confirms the uniform size of the micropores in the as-prepared carbon material. The XRD pattern of the MPCPs (the lowest curve in Figure 2 C) displays two broad peaks at around 24 and 44° that are assigned to the (002) and (101) diffraction peaks of carbon, respectively. The broad diffraction peaks again confirm the relatively low graphitization. The absence of additional peaks from possible impurities (such as ZnO and Zn) suggests the complete vaporization of Zn during the carbonization process.

thumbnail image

Figure 2. A) N2 adsorption–desorption isotherm and B) the corresponding DFT pore size distribution of MPCPs. C) XRD patterns of MPCPs, MPCP-S-I, and MPCP-S-II. D) TGA curves of MPCP-S-I and MPCP-S-II in N2 flow.

Download figure to PowerPoint

The MOF-derived MPCPs exhibit several appealing features. First, the facile preparation method is suitable for large-scale synthesis. Second, the as-prepared carbon material exhibits high porosity with exclusively uniform micropores. Moreover, the microsized particles with high uniformity favor a high packing density, thus improving the volumetric energy density of the composite cathode material. In view of the promising electrochemical performance of microporous carbon/sulfur composites in recent reports,24, 25 we investigated the controllable sulfur loading into the MPCPs. A new two-step process in a sealed vessel was applied to embed sulfur into the micropores. The low-temperature step (155 °C) allowed the sulfur to melt and infuse into the MPCPs, followed by the high-temperature step (300 °C) to further promote the sulfur infusion into the central region and the small micropores in the MPCPs. Two MPCP/sulfur composites with different sulfur contents have been prepared, denoted as MPCP-S-I and MPCP-S-II. As shown in Figure 2 C, the MPCP-S-I sample only displays the broad peaks assigned to the porous carbon, suggesting that the sulfur is well dispersed in the carbon matrix and exists in an amorphous state. In contrast, sharp diffraction peaks from elemental sulfur are observed for MPCP-S-II, which has a higher sulfur content. Thermogravimetric analysis (TGA) was used to determine the percentage of sulfur in the MPCP/sulfur composites (Figure 2 D). Interestingly, the weight loss for MPCP-S-I due to the evaporation of sulfur occurs in a wide temperature range mainly between 200 and 480 °C with a total sulfur content of about 43 wt %. On the contrary, two steps in the sulfur evaporation process are clearly observed for the MPCP-S-II sample with a higher sulfur content of about 63 wt %. The sample firstly undergoes a rapid weight loss between 200 and 270 °C, followed by a second step, which displays a gradual loss until around 480 °C. Therefore, for the MPCP-S-I sample, we postulate that all of the sulfur is embedded into the micropores, where the stronger interaction between the carbon matrix and sulfur results in the higher vaporizing temperature. The sulfur content is also close to the theoretical value that would be expected assuming all of the micropores are filled up with sulfur (ca. 38 wt %). In the case of MPCP-S-II, due to the excess sulfur that cannot be fully accommodated in the micropores, sulfur exists also in the crystalline form, probably in some large cavity or on the surface of the MPCPs.

To verify the above hypotheses, we also examined the morphology of the MPCP/sulfur composites. As depicted in Figure 3 A, no sulfur particles were found in the MPCP-S-I sample. The TEM images (Figure 3 B) reveal that the MPCP-S-I particles resemble the MPCP particles before sulfur loading, with slightly increased contrast due to the filling of pores with sulfur. Elemental mapping on a single MPCP-S-I particle confirms the uniform presence of sulfur in the carbon matrix as given in Figure 3 C. Meanwhile, although no large sulfur aggregates are found in the MPCP-S-II sample (Figure 3 D), closer examination reveals some additional sulfur covering the surface and linking the polyhedral particles (Figure 3 E). Such additional sulfur outside the micropores corresponds to the first weight loss step in the TGA curve at low temperature. Based on the above results, it can be concluded that all of the sulfur is strongly embedded in the micropores for the MPCP-S-I sample, whereas for MPCP-S-II the sulfur exists both in the micropores and outside the microporous carbon particles. Notably, the high-temperature step during the sulfur loading is essential to embed sulfur into the micropores. With merely a prolonged heating process at 155 °C for 20 h, only a small amount of the sulfur infuses into the micropores of the MPCPs, as evidenced by the TGA curve of the sample (see Figure S2 in the Supporting Information). Hence for microporous carbon, both the sulfur content and loading temperature appear to be crucial to the final distribution of sulfur in the carbon host.

thumbnail image

Figure 3. A) FESEM, B) TEM images, and C) elemental mapping of MPCP-S-I. D, E) FESEM images of MPCP-S-II.

Download figure to PowerPoint

The different physicochemical environment of sulfur in the composites would largely determine the electrochemical performance of sulfur. Additionally, the electrolytes are also anticipated to have critical effects.34 Here we investigate three different electrolytes: 1.0 M LiPF6 in EC/DEC, 1.0 M LiTFSI in TEGDME, and 1.0 M LiTFSI in DOL/DME. The first carbonate-based electrolyte is commonly used for conventional lithium-ion batteries, whereas the two ether-based electrolytes are widely applied in Li–S batteries. Interestingly, the MPCP/sulfur composites with different sulfur contents show distinct charge–discharge voltage profiles. As an example, the charge–discharge curves of MPCP/sulfur composites using the DOL/DME electrolyte are given in Figure 4 A and B for comparison. For the MPCP-S-I sample, only one distinct plateau at about 1.8 V (vs. Li/Li+) is observed during the discharge process. In contrast, multiple discharge plateaus are identified in the profile of the MPCP-S-II sample. The first two plateaus at 2.3 and 2.1 V are in agreement with that of conventional carbon/sulfur composites, which are due to the conversion of sulfur molecules (S8) to soluble polysulfides, and the reduction of polysulfides to insoluble Li2S2/Li2S, respectively.4, 5 These two plateaus likely correspond to the sulfur located in the large pores or on the surface of the MPCPs, whereas the third plateau at 1.8 V, similar to the one of the MPCP-S-I sample, would be ascribed to the sulfur embedded in micropores.24, 25 The charge profiles of both samples are also in good agreement with the reversion of the corresponding discharge processes. The distinct charge–discharge curves clearly indicate different electrochemical characteristics of sulfur embedded in the micropores during charging/discharging, compared to that of bulk sulfur and typical carbon/sulfur nanocomposites.4, 5 The high initial discharge capacity, which is close to the theoretical value, and the plateau with reduced voltage might be ascribed to the intimate contact and strong binding between the microporous carbon and the confined sulfur. A recent study suggests that sulfur might exist as small molecules in the micropores of the carbon matrix, which leads to the unusual electrochemical performance.25 However, further investigation would be necessary to unambiguously reveal the electrochemical reactions involved.

thumbnail image

Figure 4. Charge–discharge profiles of A) MPCP-S-I and B) MPCP-S-II in the DOL/DME electrolyte. Cycling performance of C) MPCP-S-I and D) MPCP-S-II in different electrolytes. Current density is 100 mA g−1 based on the mass of composites.

Download figure to PowerPoint

Our study also demonstrates some unexpected results regarding the compatibility issues of electrolytes. As shown in Figure S3A in the Supporting Information, the MPCP-S-I sample in the EC/DEC electrolyte displays similar profiles to that with the DOL/DME electrolyte. However, unstable charge profiles and severe overcharging are found when the TEGDME electrolyte is used (see Figure S3B in the Supporting Information). The reason for such an unusual phenomenon might be related to the dissolution of intermediates and the difficulty of converting Li2S/Li2S2 back to elemental sulfur inside the micropores with the TEGDME electrolyte. In contrast, the MPCP-S-II sample is incompatible with EC/DEC, with immediate capacity fading to zero after initial discharge (see Figure S4A in the Supporting Information), which is similar to conventional carbon/sulfur composites due to the reactions between polysulfides and the carbonate-based electrolyte.34 Our result again confirms that the presence of weakly bound sulfur would cause fatal damage to the electrode in carbonate-based electrolytes. Indeed, the MPCP-S-II sample behaves quite normally in the TEGDME electrolyte (see Figure S4B in the Supporting Information), despite the presence of a certain amount of sulfur embedded in the micropores.

The cycling stability of the MPCP/sulfur composites in different electrolytes is shown in Figure 4 C and D. The MPCP-S-I sample exhibits similar performance in both DOL/DME and EC/DEC electrolytes, whereas the electrode with the EC/DEC electrolyte shows slightly higher capacity after prolonged cycling (Figure 4 C). After the rapid capacity decline in the first few cycles, relatively stable cycling performance is established with slow capacity fading. After 100 cycles, the MPCP-S-I sample delivers reversible capacities of about 420 and 490 mA h g−1 based on the mass of sulfur in DOL/DME and EC/DEC electrolytes, respectively, which correspond to specific capacities of about 180 and 210 mA h g−1, respectively, based on the mass of the MPCP-S-I composite. Except for the initial cycle, the very close values between the charge and discharge capacities of the same cycle indicate the high Coulombic efficiency (see Figure S5 A in the Supporting Information), and the absence of overcharging or other undesirable side reactions during the whole course of the testing. Meanwhile, the MPCP-S-II sample shows a less attractive performance in the two compatible electrolytes (Figure 4 D). The capacity quickly decays in the DOL/DME electrolyte and only remains at a value of around 250 mA h g−1 based on the mass of sulfur after 100 cycles. Although the capacity fading is less severe in the TEGDME electrolyte, the cycling stability is still worse than that of the MPCP-S-I sample. At the end of the cycling test, the discharge capacity remains about 340 mA h g−1 based on the mass of sulfur, which corresponds to around 210 mA h g−1 based on the mass of composite due to the relatively high sulfur content. However, a certain degree of overcharging is observed after around 20 cycles (see Figure S5B in the Supporting Information), which is a common drawback for sulfur cathodes. Besides, the MPCP/sulfur composites show a normal rate performance in proper electrolytes (see Figure S6 in the Supporting Information). The above comparison studies suggest that sulfur confined in micropores tends to have a stable capacity retention and better compatibility with conventional carbonate-based electrolytes, although some problems, such as the large initial capacity loss, reduced voltage plateau, and limited pore volume, are to be further addressed.

In summary, we have prepared microporous carbon polyhedrons (MPCPs) with abundant and uniform micropores from a unique MOF. We have used these MPCPs as a model microporous carbon support to prepare carbon/sulfur composites for Li–S batteries. The correct sulfur content and preparation procedure are critical to embed all of the sulfur into the micropores of the MPCPs and to achieve strong binding between sulfur and the carbon host. Systematic investigation on the electrochemical performance reveals the unusual characteristics of the MPCP/sulfur composites. The composite with sulfur exclusively embedded in micropores exhibits a stable cycling performance with high Coulombic efficiency in both DOL/DME and EC/DEC electrolytes, whereas the composite with additional sulfur outside the micropores shows inferior performance as well as different electrolyte compatibility. Although the overall performance of the MPCP/sulfur composites is still not satisfactory, we anticipate that further improvement could be achieved by optimizations, such as controlling the sulfur content, introducing meso-/macropores for electrolyte penetration, and modifying the electrolyte. The present work provides some guidance for further studies on the development of cathode materials for Li–S batteries.

Supporting Information

  1. Top of page
  2. Supporting Information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.


Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.