Lithium sulfur batteries (LSB) have been attracting extensive interest as a promising next-generation high energy storage technology, due to the high theoretical specific capacity, low costs and environmental friendliness of the electrode materials [1][2][3][4] . Sulfur has been known to undergo multi-electron reactions with Li ions and exhibit a high theoretical specific capacity of 1672 mAh/g 4 . Ideally, the cathode materials for LSB should include a high surface area and large pore volume to accommodate a high loading of sulfur particles, strong polar absorption for soluble reactive intermediates, and highly conductive network for rapid transport of ions and electrons 5 . However, the performance of LSB has been limited by several challenging obstacles, such as fast capacity decay, low coulombic efficiency and poor rate capability, which greatly hinder the practical applications 3,4 . These issues are mainly ascribed to the low electrical conductivity of the active materials (e.g., sulfur, Li2S, and Li2S2), diffusion (and loss) of soluble polysulfide intermediates, and large volumetric changes of the cathode materials during the charge-discharge process.
These issues may be mitigated by the development of new, effective sulfur hosts 6,7 , modification of membrane surfaces [8][9][10] , and/or addition of electrolyte additives 11,12 . In a number of studies, conductive matrices, such as carbon materials and conductive polymers, have been employed to encapsulate sulfur, improve electrical conductivity of the cathode as well as minimize the loss of lithium polysulfides 13,14 . In particular, carbon materials with a high specific surface area and large pore volume have been used rather extensively, such as meso/microporous carbons 15 , graphene 6,[16][17][18] , hollow carbon nanofibers 19 , hollow carbon nanospheres 20 , and carbon nanotubes 21,22 . In addition, conductive polymers, such as polypyrrole (PPy) 23 , polyaniline (PANI) 24 and poly(3,4-ethylenedioxythiophne) (PEDOT) 7,25 , have also been used to host sulfur particles. The resulting sulfur-encapsulated nanocomposites typically exhibit enhanced specific capacity and good cycling performance during the initial cycles. But the coulombic efficiency in general remains low, and rapid capacity loss occurs during long-term cycling, as the non-polar carbon/polymer hosts cannot efficiently entrap the polar lithium polysulfide species because of weak interactions with sulfur.
Polar host materials, such as metal oxides of TiO2 26 and MnO2 27 , metal hydroxides of Ni(OH)2 28 , and metal-organic frameworks (MOFs) 29,30 , have been found to form strong chemical bonds with lithium polysulfides, which can significantly improve the long-term cycling performance of LSB. Of these, MnO2-based nanocomposites with a uniform structure and large surface area have been attracting particular attention 31 . For instance, Nazar and coworkers dispersed sulfur onto the surface of MnO2 nanosheets 32 and then covered the sulfur with a MnO2 shell 27 to improve the electrochemical performance. In another study 33 , Chen's group decorated hollow sulfur nanospheres with MnO2 nanosheets. Diao and coworkers synthetized unique sulfur/-MnO2 coreshell nanocomposites 34 . However, the electrical conductivity of these metal-oxide materials is typically low, in comparison with carbon and conductive polymers, which compromises the rate capability and specific capacity of LSB. Consequently, conductive additives are generally added to the cathode materials. This inevitably reduces the mass loading of active sulfur.
Therefore, it can be envisaged that nanocomposites based on the combination of conductive matrices and polar metal oxides may serve as effective host materials of sulfur. For instance, Lou and coworkers fabricated hollow carbon nanofibers filled with MnO2 nanosheets to host sulfur nanoparticles 35 . Kong's group used hollow MnO2 nanospheres with a PPy shell to encapsulate sulfur, which exhibited an excellent cycling performance 36 . Yu's group also synthesized PPy-MnO2 nanotubes as a sulfur host for high-performance lithium sulfur batteries 37 .
In this work, we prepared PPy-modified MnO2 nanotubes for effective encapsulation of sulfur nanoparticles. The MnO2 nanotubes were synthesized through a facile hydrothermal method and the PPy layer was formed in situ by using the MnO2 as the oxidant. Sulfur nanoparticles were then melted and diffused into the nanotubes. The resulting ternary structure exhibited at least two advantages. First, the hollow interior of the MnO2 nanotubes provided a large space for the loading of sulfur particles, and the strong chemical interactions with polysulfides intermediates helped minimize the loss of the active species. Second, the PPy shells efficiently enhanced the electrical conductivity of the cathode materials. These led to a remarkable performance as a LSB cathode material.
Pyrrole was used after purification by distillation. Concentrated sulfuric acid (H2SO4) and hydrochloric acid (HCl) were purchased from Baiyin Liangyou Chemical Reagents Co., Ltd. Potassium permanganate (KMnO4) were purchased from Guangfu Chemical Reagents Co. Sublimed sulfur (99.95%) was obtained from Aladdin Industrial Corporation.
As shown in Scheme 1, MnO2 nanotubes were first prepared by a facile hydrothermal method 38,39 . In brief, 0.658 g of KMnO4 was dissolved in 75 mL of deionized water. Then 1.5 mL of concentrated HCl was added into the solution in a dropwise fashion under magnetic stirring for 15 min at ambient temperature. The solution was then transferred to a 100 mL Teflon-lined stainless autoclave, and heated at 150 °C for 12 h. After being cooled down to room temperature, brown precipitates (MnO2 nanotubes) were filtered and washed with deionized water and ethanol, and then dried at 60 °C in an oven.
MnO2-PPy nanotubes were then prepared by using the obtained MnO2 nanotubes as reaction templates. Experimentally, 0.2 g of the as-prepared MnO2 nanotubes was dispersed in 1 M HCl solution (50 mL) under sonication. After magnetic stirring for 30 min in an ice bath, 980 μL of pyrrole was added to the suspension, and the polymerization was carried out in the ice bath for 12 h. Dark blue precipitates were obtained by centrifugation, washed with deionized water and ethanol several times, and then dried at 60 °C, affording MnO2-PPy nanotubes.
The obtained MnO2-PPy nanotubes were then homogeneously blended with sulfur as a mass ratio of 3:7, and the mixture was heated at 155 °C for 24 h in a nitrogen atmosphere, such that sulfur was melted and infiltrate the hollow interiors of the MnO2-PPy nanotubes. To remove sulfur on the outside surface of the MnO2-PPy nanotubes, the sample was heated at 200 °C for 2 h. The resulting sample was referred to as MnO2-PPy-S.
The surface morphology of the as-prepared nanocomposites was examined with a scanning electron microscope (SEM, Hitachi S-4800, Japan) equipped with energy dispersive X-ray spectroscopy (EDX), and a high-resolution transmission electron microscope (HR-TEM, JEOL TEM-2010). The sample crystallinity was characterized by using an X-ray diffractometer (XRD, Shimadzu Corp., Kyoto, Japan) with Cu Kradiation. Thermogravimetric analysis (TGA) (TA Instruments, New Castle, DE) was carried out under a N2 atmosphere at the heating rate of 10 °C/min.
To prepare working cathodes, the active material obtained above was blended with acetylene black as a conductive agent and polyvinylidenediuoride (PVDF) as binders, at the mass ratio of 7:2:1, in N-methyl-2-pyrrolidone (NMP) to form a uniform slurry.
The slurry was cast onto an Al foil current collector and dried at 40 °C for 12 h in a vacuum oven. CR3032 half coin cells were assembled in a glove box filled with argon.
Lithium foils were employed as both the counter and reference electrodes, the active material as the cathode and a Celgard 2400 membrane as the separator. The liquid electrolyte was composed of 1 M bis(trifluoromethane) sulfonimide lithium salt (LiTFSI) dissolved in a mixture of 1.3-dioxolane (DOL) and dimethoxymethane (DME)
(1:1 v:v) with 1% LiNO3 additive. Electrochemical performance was tested at various current densities within the voltage range of 1.7 to 2.8 V versus Li + /Li using a CT2001A battery testing system (LAND Electronic Co.). The electrodes were cycled with a CHI 660E electrochemical workstation in the potential window of 1.7 to 2.8 V versus Li + /Li at the scan rate of 0.1 mV/s.
The structure of the nanotube nanocomposites was first characterized by SEM and TEM measurements. From the SEM images in Figure S1, it can be seen that the MnO2 nanotubes exhibited a smooth surface morphology with an outer diameter of about 100 nm (Figure S1a), and the hollow tubular interior can be clearly identified in TEM measurements, with an inner diameter of about 70 nm (Figure 1a,b). After the coating of a PPy layer, the deposition of PPy nanoparticles rendered the MnO2 nanotube surfaces drastically roughened, as shown in Figure S1b,c and Figure 1c,d. The formation of this rather compact PPy layer was likely due to the MnO2 nanotubes that served both as a supporting scaffold and an oxidizing agent for pyrrole polymerization 37 . One can see that in the MnO2-PPy samples, the hollow nanotube structure was retained, which may be exploited for the loading of sulfur. This can be clearly seen in TEM studies (Figure 1e, f), whereas no obvious sulfur particles were found on the exterior of the MnO2-PPy nanotubes (Figure S1d-f), suggesting efficient confinement of sulfur within the MnO2 nanotubes. Indeed, EDS mapping analysis (Figure 2) shows that the elements of carbon, nitrogen, oxygen, sulfur and manganese were uniformly distributed throughout the sample, indicating the successful and homogeneous loading of PPy and sulfur into the MnO2 nanotubes. The crystalline structures of the samples were then examined by XRD measurements.
As shown in Figure S2a, MnO2 nanotubes exhibited a series of well-defined diffraction peaks at 12.6°, 18.1°, 28.8°, 37.6°, 41.9°, 49.9°, 56.2°, 60.2°, 65.1º and 69.7º, which can be ascribed, respectively, to the (110), ( 200), (310), ( 211), (301), (411), (521), ( 002) and (541) crystal planes of tetragonal-like α-MnO2 (JCPDS NO. 44-0141); whereas MnO2-PPy shows only a featureless profile except for a broad peak at ca. 24.4°, 23 suggesting an amorphous structure of a PPy outer layer. Interestingly, the diffraction patterns of the MnO2-PPy-S composite were dominated by those of sulfur, likely because of the high loading of sulfur.
Consistent results were obtained in FT-IR measurements. From Figure S2b, it is obvious that MnO2-PPy exhibited a spectral profile consistent with that of PPy, indicating that the MnO2 nanotubes were well coated with PPy layers. The characteristic bands at 1550 cm -1 and 1458 cm -1 can be ascribed to the fundamental vibrations of the polypyrrole ring, the bands at 1290 cm -1 and 1045 cm -1 are due to the C-H in-plane vibrations, and the band at 1180 cm -1 arises from the C-N stretching vibration of the polypyrrole chain [40][41][42] . Interestingly, after sulfur loading, these vibrational features became less well-defined for the MnO2-PPy-S sample 36 . The loading of sulfur in the MnO2-PPy-S composite was then quantitatively evaluated by TGA measurements. From Figure 3, one can see that the weight loss of the MnO2-PPy-S sample commenced at ca. 180 C, and the sample weight remained virtually unchanged at temperatures over ca. 310 C. This profile is very similar to that of pure sulfur, whereas PPy was rather stable within this temperature range. The total weight loss for MnO2-PPy-S was estimated to be 71%. That is, sulfur accounts for about surface area of the MnO2-PPy was calculated to be 111.51 m 2 /g, which diminished markedly to 21.46 m 2 /g for MnO2-PPy-S as sulfur impregnated the MnO2 hollow tubes.
The mesoporous size distributions of the samples are shown in Figure 4b. The MnO2-PPy showed a pore volume of 0.44 cm 2 /g with an average pore size of 13.87 nm, while after sulfur loading, the MnO2-PPy-S sample displayed a substantial decrease of the pore volume to 0.13 cm 2 /g, whereas the average pore size increased to 19.16 nm, likely because smaller pores were easier to fill up with sulfur impregnation. The performance of the MnO2-PPy-S composite as a cathode material for LSB was then evaluated electrochemically. Figure 5a shows the charging-discharging cycling performance of the sample at different current densities. The electrode was first cycled at a low current density of 0.05 C for activation and then charged and discharged at the current density of 0.2 C and 1 C, respectively. After activation for three cycles, the cathode delivered a specific capacity of 973.8 mAh/g at 0.2 C and 770.4 mAh/g at 1 C, respectively; and after 100 cycles, the capacity remained promising at 734.6 and 572.8 mAh/g.
To evaluate the rate capability of the MnO2-PPy-S composites, the electrode was charged and discharged from 0.2 C to 1 C, 2 C, 3 C, 5 C and finally back to 0.2 C at the voltage range of 0.01 V-3 V, as shown in Figure 5b. The initial specific discharge capacity was 803.3 mAh/g at 0.2 C, and then decreased slowly to 708.0 mAh/g at 1 C, 615.3 mAh/g at 2 C, 542.0 mAh/g at 3 C, and 470.0 mAh/g at 5 C. More importantly, the electrode was able to deliver a specific capacity of 726.6 mAh/g when the current density was re-increased to 0.2 C, more than 90% retention as compared to the initial specific capacity. This suggests high reversibility of the operation.
The durability of the MnO2-PPy-S electrode was further examined by charging and discharging at the current density of 0.2 C for 500 cycles. From Figure 6a (left y axis), one can see that during the initial activation at 0.05 C, the electrode delivered a specific capacity of 1469.2 mAh/g in the first cycle. Then as the current density increased to 0.2 C, the specific discharge capacity diminished to 973.8 mAh/g in the 4th cycle. In the following cycles, the discharge capacity declined much more slowly to 734.6 mAh/g in the 100 th cycle, 694.8 mAh/g in the 200 th , 671.9 mAh/g in the 300 th , and 632.1 mAh/g in the 400 th cycle and remained almost invariant at around 586 mAh/g after the 500 th cycle. This means that on average there was only 0.07 % capacity decay per cycle during this discharge-charge process (Figure 6b). Consistent behaviors can be observed with the corresponding coulombic efficiency (Figure 8a, right y axis), where the MnO2-PPy-S electrode can be seen to demonstrate an outstanding coulombic efficiency of 95.7 % on average. To evaluate the electrochemical reaction mechanism, the MnO2-PPy-S cathode was tested by cyclic voltammetric measurements at the scan rate of 0.1 mV/s from 1.7
V to 2.8 V for 5 cycles. From Figure 7a, the electrode was swept from open circuit voltage (OCV) to 1.7 V, where element sulfur was reduced to Li2S2/Li2S. Notably, the
Li2S2/Li2S species were not oxidized back to element sulfur during the charging process 43,44 . Two well-defined cathodic peaks appeared at ca. 2.3 V (peak i) and 2.1 V (peak ii), which might be ascribed to the reduction of high-order lithium polysulfides (e.g., Li2S8) to the low-order species (Li2Sx. 4 ≤ x ≤ 8), and the transformation of soluble lithium polysulfides to solid Li2S2/Li2S, respectively 23,36 . In the corresponding anodic scan, two adjacent peaks can be identified at 2.3 V (peak iii) and 2.4 V (peak iv), likely due to the conversion of the Li2S2/Li2S to low-order lithium polysulfides and then to high-order polysulfides, respectively 34 . In the following four cycles, the voltammograms overlapped with each other, demonstrating good cycling stability of the electrode.
Electrochemical impedance measurements of the MnO2-PPy-S electrode were then performed to examine the reaction dynamics for lithium insertion and extraction during the cycling tests. The Nyquist plots are depicted in Figure 7b. It can be seen that the sample exhibited two depressed semicircles in the high and middle frequency domains and a short inclined line in the low frequency domain. The semicircle in the high frequency region can be ascribed to the interfacial charge transfer while the semicircle in the middle frequency region is likely caused by mass transport for the formation of solid polysulfides (Li2S and Li2S2), which disappeared in the subsequent cycles as the Li2S2/Li2S were not converted back to element sulfur, consistent with results from the CV measurements (Figure 7a) 34,43,44 . Meanwhile, the typical Nyquist plots after 100 cycles exhibited a depressed semicircle in the high frequency region and an inclined line in the low frequency region, which likely reflected the charge-transfer resistance of the interface between the electrolyte and sulfur electrode and the lithium ion semi-infinite diffusion, respectively.
In addition, the resulting MnO2-PPy-S cathodes demonstrated a remarkable long cycling stability (586 mAh/g after 500 cycles), rate capability (470 mAh/g at 5C) and coulombic efficiency (average 95.7%) due to the fine structural combination of metal oxides (MnO2) and conducting polymer (PPy) which accommodate the volumetric changes and confine the soluble polysulfides. The electrode performance was higher than leading results reported in recent literature (Table 1).
In this study, a functional nanocomposite was prepared where polypyrrole modified MnO2 nanotubes were used as a host scaffold for the impregnation of sulfur.
The resulting composites showed a high-performance as the cathode material for lithium sulfur batteries, featuring high specific capacity, excellent cycling stability and good rate capabilities. This was ascribed to the hollow interior of the MnO2 nanotubes that accommodated the high loading and large volumetric expansion of sulfur particles, and the polypyrrole layer that facilitated charge transfer during the chargingdischarging processes.