Perovskite La0.6Sr0.4CoO3-δ as a new polysulfide immobilizer for high-energy lithium-sulfur batteries

Author’s Accepted Manuscript Perovskite La0.6Sr0.4CoO3-δ as a New Polysulfide Immobilizer for High-Energy Lithium-Sulfur Batteries Zhangxiang Hao, Rui Zeng, Lixia Yuan, Qiming Bing, Jingyao Liu, Jingwei Xiang, Yunhui Huang www.elsevier.com/locate/nanoenergy

PII: DOI: Reference:

S2211-2855(17)30516-5 http://dx.doi.org/10.1016/j.nanoen.2017.08.039 NANOEN2153

To appear in: Nano Energy Received date: 23 June 2017 Revised date: 29 July 2017 Accepted date: 21 August 2017 Cite this article as: Zhangxiang Hao, Rui Zeng, Lixia Yuan, Qiming Bing, Jingyao Liu, Jingwei Xiang and Yunhui Huang, Perovskite La 0.6Sr0.4CoO3-δ as a New Polysulfide Immobilizer for High-Energy Lithium-Sulfur Batteries, Nano Energy, http://dx.doi.org/10.1016/j.nanoen.2017.08.039 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Perovskite La0.6Sr0.4CoO3-δ as a New Polysulfide Immobilizer for High-Energy Lithium-Sulfur Batteries

Zhangxiang Hao a, Rui Zeng a, Lixia Yuan a, *, Qiming Bing b, Jingyao Liu b, Jingwei Xiang a, Yunhui Huang a, *

a

State Key Laboratory of Materials Processing and Die & Mould Technology, School of

Materials Science and Engineering, Huazhong University of Science and Technology, Wuhan, Hubei 430074, P. R. China. b

Laboratory of Theoretical and Computational Chemistry, Institute of Theoretical Chemistry,

Jilin University, Changchun, Jilin 130023, P. R. China.

Abstract To tackle the issue of low sulfur utilization and inferior cycle stability of sulfur cathode, we first report a new perovskite-type La0.6Sr0.4CoO3-δ (LSC) immobilizer to anchor the intermediate polysulfides via chemical interaction. The experimental results and theoretical calculations demonstrate that Sr doping results in valence variation in Co along with oxygen vacancy; The Co ions with mixed valence have strong adsorption to the polysulfide ions while the existence of oxygen vacancy enhances the binding strength between Li2S4 and LSC. Based on LSC, a dual coxial LSC/S@C nanocable is successfully designed and fabricated. With a sulfur loading of 2.1 mg cm2, the LSC/S@C cathodes demonstrate a high reversible capacity of 996 mAh g1 at 0.5 C and an outstanding cycle stability with only 0.039% capacity fade

1

per cycle over 400 cycles. Even with a high sulfur loading of 5.4 mg cm2, the LSC/S@C cathode can still deliver similar sulfur utilization and excellent cycling stability. The excellent cycle stability benefits from the chemical interaction between LSC and polysulfides, and the physical entrapment of the carbon shell. Moreover, the highly conductive LSC@C host and the porous interconnected fiber web-like architecture

facilitate

the

mass

transfer

during

charge/discharge

process

synchronously.

Keywords: perovskite La0.6Sr0.4CoO3-δ, polysulfide immobilizer, coxial nanocable, sulfur loading, DFT calculations, lithium-sulfur battery

* Corresponding author. Tel. & fax: 86-27-87558241 E-mail: [email protected] (L. X. Yuan), [email protected] (Y. H. Huang)

2

1 Introduction The development of electric vehicles strongly depends on rechargeable batteries with high energy density and long service life [1, 2]. Lithium-sulfur (Li-S) batteries have an unapproachable energy density of 2600 Wh kg1 [3, 4], which are believed as one of the most promising next-generation rechargeable batteries to satisfy the emerging large-scale applications [5]. However, the Li-S batteries are bedeviled by the inherently low conductivity of S and its final discharge products (Li 2S2/Li2S), the large volume expansion (~80％) from the oxidation state S8 to reduction state Li2S, and the dissolution and migration of the long-chain intermediate polysulfides (Sn2, 3 ≤n≤8). The end result is the drastically shortened lifespan of Li-S batteries. To tackle the bottleneck issues in Li-S technology, extensive research works have been conducted. The main strategies include constructing various composite nano-architectures to limit S species within the cathode [6-10], utilizing interlayer structure [11-13] or novel solid electrolytes [14, 15] to block the migration of polysulfides within the cathode side, protecting the anode to reduce the reactivity between Li and polysulfides [16-18]. Among these strategies, the most widely-used one is to develop various nano-structured carbon matrices to encapsulate S within the pores, and hence to restrict the migration of the dissolved polysulfides during the discharge-charge process. Numerous novel porous carbon materials have been designed as hosts for S, including ordered/disordered micro/mesoporous carbons [19, 20], hollow or yolk-shell architectures [21, 22], and various derived structures of

3

carbon nanofibers [23, 24], nanotubes [25-27] or graphenes [28, 29]. Although these C/S nanocomposites have achieved much success in increasing sulfur utilization and prolonging the cycle life, the capacity decay is still a challenge especially in long cycles. The weak interaction between the nonpolar carbon host and the polar polysulfides cannot afford a long-term cycle stability. More recently, it is reported that some polar materials, such as metal oxides [30-34], metal sulfides [35, 36], metal-organic frameworks [22, 37], can immobilize S species via strong chemical interaction and therefore show high efficiency in entrapping polysulfides. Many research works have been carried out to elucidate the “S immobilization” mechanism [38] and to develop novel nanostructured polar hosts for S species [39]. However, most of these polar host materials have poor electronic conductivity, which may impede the electron transport within the cathode and hence worsen the rate capability or even the S utilization. Therefore, it is desirable to develop composite hosts for S with high electronic conductivity and high entrapping capability by combining chemical trap materials with carbon matrices. In this work, we first report a perovskite-type host material, La0.6Sr0.4CoO3-δ (LSC), which has the highest conductivity in the family of La1-xSrxCoO3-δ (x = 0-1) [40, 41], as an effective high-efficiency polysulfide adsorbent. Experimental results and theoretical calculation show that the LSC exhibits strong immobilization capability for polysulfides, which comes mainly from the ionic interactions between S atoms and Co atoms. Furthermore, Sr-doping creates abundant oxygen vacancies and

4

valence variation of Co in LSC and therefore results in a largely improved conductivity of 52.63 S cm1 from pristine LaCoO3’s 0.12 S cm1, which can facilitate the electrochemical redox process of the insulating S. To make the best of the polysulfide immobilization capability of LSC, a novel coaxial yolk-shell structured host with carbon nanotubes (CNTs) filled with porous LSC nanofibers was further designed. With an areal loading of 2.1 mgsulfur cm2, the LSC/S@C cathode exhibits a high reversible capacity of 996 mAh g1 at 0.5 C and long cycle life with average decay of only 0.039% per cycle over 400 cycles. Even the areal loading is increased to 5.4 mgsulfur cm2, the LSC/S@C cathode can still show high capacity and excellent cyclability.

2 Experimental Synthesis of porous LSC nanofibers. LSC nanofibers were fabricated via an electrospinning route. In a typical process, 0.328 g Sr(CH3COO)2 and 0.996 g Co(CH3COO)24H2O were dissolved in 10 ml N,N-dimethylformamide (DMF). After stirring for 1 h, 0.758 g La(CH3COO)3 and 0.8 g of polyacrylonitrile (PAN) were added and kept at 80 C overnight. Then the obtained solution was injected into a plastic syringe with a 20-gauge stainless steel needle. The electrode-clamped needle was linked to a variable high-voltage power supply. An aluminum foil (15 cm away from the needle tip) serves as grounded counter electrode as well as current collector. The feeding rate was set to 0.8 ml h1 via a syringe pump. The hybrid precursor nanofibers were formed when 17 kV voltage was applied. The as-collected 5

electrospun precursor fibers were dried at 80 C for 6 h and then calcined in air at 250 C for 3 h, followed by another 2 h at 750 C with a heating rate of 2 C min1. Synthesis of porous LSC nanofibers@C. The as-prepared LSC nanofibers were dispersed in 50 ml alcohol with concentration of 1 mg ml1. A certain amount of cetyltrimethyl ammonium bromide (CTAB, 0.2 g), deionized water (DIW, 5 ml), and ammonium hydroxide (5 ml, 28-30%) were added into the alcohol suspension. The resulted suspension solution was ultrasonically treated for 30 min to form uniform dispersion. After that, 215 µl tetraethyl orthosilicate (TEOS) was added dropwise into the above solution and stirred for 6 h to wrap a layer of SiO2 on the surface of LSC nanofibers. To coat the carbon shell on the LSC@SiO2, 0.08 g LSC@SiO2 nanofibers was ultrasonically dispersed into a mixed solution composed of 30 ml ethanol, 9 ml DIW and 1.0 g CTAB, followed by addition of 0.105 g resorcinol and 0.12 ml ammonium hydroxide, and stirred for 30 min. Then, 0.15 ml formalin solution was added to the dispersion under stirring. After stirring for 6 h, the mixture was aged at room temperature overnight to form resorcinol-formaldehyde (RF). The product of LSC@SiO2@RF was collected and washed by centrifugation, and dried at 80 C in air. The as-prepared LSC@SiO2@RF fibers were carbonized at 700 C for 3 h under nitrogen atmosphere with a heating rate of 3 C min1. Then the pyrolyzed product (LSC@SiO2@C) was treated with 5% HF solution for 5 min to remove the SiO2 layer. The product was washed by water and ethanol for several times via vacuum filtration. The final LSC@C was obtained after drying at 70 C in air overnight. 6

Synthesis of LSC/S@C composite cathode. A mixture of sulfur powder and LSC@C (5:1, weight ratio) was sealed in a glass bottle under argon protection, and heated at 155 C for 12 h in an oven. Then, the product was heated at 250 C for 1 h under flowing argon atmosphere in a tube furnace to remove the redundant S outside LSC@C structure. The parallel sample of S@C was obtained by immersing LSC/S@C in hydrogen chloride (HCl) solution to remove LSC. The LCO/S@C cathode was fabricated by the same method as LSC/S@C.

Characterization. X-ray diffraction (XRD) patterns were collected on a X-ray powder diffractor (PANalytical X’pert PRO-DY2198, Holland) operating at 40 kV and 40 mA using Cu Kα radiation (λ= 0.15406 nm). The electronic conductivity was tested by a standard four-probe method. Nitrogen sorption isotherms were measured at -196 °C with Micrometritics ASAP 2010 analyzer (US). Before measurement, the sample

was

degassed

under

vacuum

at

300

°C

overnight.

The

Braunauer-Emmet-Teller (BET) surface area was calculated according to the adsorption data in the relative pressure range from 0.06 to 0.2. The morphologies of LSC and LSC/S@C were observed by scanning electron microscopy (SEM) (FEI, Sirion 200). High-resolution transmission electron microscopy (HR-TEM) images of LSC/S@C composites were obtained by Tecnai G2 F30 (FEI, Holland). The S content in LSC/S@C composite was determined by thermogravimetric (TG) analysis (PerkinElmer) in argon atmosphere with a heating rate of 10 °C min1 from 25 to 700 °C. The ultraviolet-visible (UV-Vis) absorption spectra were characterized by 7

UV-2550 spectrophotometer (Shimadzu), and the blank 1, 3-dioxolane (DOL)/1, 2-dimethoxyethane (DME) (1:1, v/v) solution was used as reference.

Electrochemical measurements. All the electrochemical measurements were tested with 2032 coin cells with Li foil as anode, and the electrolyte consists of 1 M bis(trifluoromethanesulfonyl)imide lithium (LiTFSI) in a mixed solution of DME and DOL (1:1 v/v) with LiNO3 (2 wt.%) as an additive. For the cathode preparation, LSC/S@C composite, super P, sodium carboxyl methyl cellulose (NaCMC) and styrene butadiene rubber (SBR) were dispersed in deionized water in weight ratio of 80: 10: 5: 5. The slurry was coated on carbon paper and dried at 80 °C overnight in vacuum. Then, the electrode was roll pressed, and punched into round discs of 8 mm in diameter with mass loading from 1 to 5.5 mgsulfur cm2. The ratio of sulfur to electrolyte is 0.1 mg l1 for all the cells. Cyclic voltammetry (CV) measurement was conducted on an electrochemical workstation (CHI614b) at a scan rate of 0.05 mV s1 with cutoff voltage range of 1.72.8 V at room temperature. Galvanostatic charge and discharge were tested in a voltage window of 1.72.8 V on a battery measurement system (Land, China). Experimental details of LSC as adsorbent and computational methods are shown in supporting information.

3 Results and discussion

8

The synthesis of the LSC/S@C nanofibers is illustrated in Figure 1. Firstly, porous LSC nanofibers were fabricated by a simple electrospinning route followed by high-temperature calcination, and then coated with the precursor of SiO2 and RF resin. After carbonized in N2 atmosphere at 700 C for 3 h, the resulted LSC@SiO2@C nanofibers were obtained. The nanofibers were then treated by HF solution (5%) to remove the SiO2 interlayer. Finally, S was encapsulated into the LSC@C host via the simple melt-diffusion process. As shown in Figure S1a and 1b, the as-electrospun LSC precursor exhibits ultrafine fibrous morphology, and the surface of the nanofibers is smooth and uniform. The fiber length reaches several micrometres and the average diameter is about 250 nm. After heat treatment in air oven, the porous LSC nanofibers were obtained (Figure 2a). Figure 2b displays typical TEM image for LSC nanofiber. It clearly shows the porous LSC nanofiber is composed of interlinked nanoparticles. The HR-TEM image in Figure 2c demonstrates that the LSC particle has typical polycrystalline structure with well-resolved lattice fringes of 0.27 and 0.22 nm corresponding to the d-spacing values of (110) and (111) planes of LSC, respectively. After coating with TEOS, the average diameter of LSC@SiO2 nanofibers increases to about 110 nm (Figure 2d). The thick SiO2 layer can generate sufficient internal void space to accommodate S, which aids to achieve high S content in the cathode. The thin RF layer grown on the LSC@SiO2 nanofibers serves as carbon precursor. After carbonization, the LSC@SiO2@C coaxial nanofibers well maintain the 1D morphology with an average diameter of ~150 nm, as shown in

9

Figure 2e. The thickness of the final carbon sheath is about 20 nm (Figures 2e and 2f). The TEM image of the sample after S loading verifies the successful encapsulation of S (Figure 2f). The gaps between the areas of C and S, S and La/Sr/Co in the multi-elemental overlapped mappings (Figure S2g-2i) clearly show three levels in the coaxial nanocable structure of LSC/S@C. The S content in the LSC/S@C composite is 66.3%, determined by thermogravimetric analysis (TG) (Figure S3). In Figure 2g, the XRD patterns of LaCoO3 (LCO), LSC, Li2S4-treated LSC (LSC/Li2S4) and LSC/S@C show the formed phases of LCO (JCPDS No. 75-0279), LSC (No. 48-0121) and sulfur (No. 99-0066). The five main diffraction peaks of the LSC sample can be indexed as (111), (200), (220), (311) and (222) reflections, indicative of the cubic phase of LSC (JCPDS No. 48-0121). In addition, all the diffraction peaks of LSC in LSC/S@C and LSC/Li2S4 agree well with those of the pure sample, suggesting stable physicochemical property of LSC in the S system. From the nitrogen absorption/desorption curve, the specific surface area of LSC was calculated as ~70.3 m2 g1 (Figure 2h), which will provide a high contact area for strong interactions between LSC and polysulfides. In addition, we tested the electronic conductivity by a standard 4-point probe method. As shown in Table S1, LSC/S@C achieves a high conductivity of 65.15 S cm1, which is much higher than that of other typical polysulfide absorbents like Ti4O7 (3.2±0.1 S cm1), MnO2 (<10-5 S cm1) and SnO2 (~10-2 S cm1), demonstrating that LSC and the carbon shell effectively offset the insulating nature of sulfur.

10

Coin cells were assembled to evaluate the electrochemical performance of the LSC/S@C cathode. Figure 3a shows the cyclic voltammogram (CV) curves of LSC/S@C cell at a scan rate of 0.05 mV s1. The cathodic peaks located at ~2.20 and 1.95 V correspond to the reduction reaction from S8 to high-order polysulfides (Li2Sx, x = 48), and then to low-order Li2S2/Li2S; in the subsequent anodic process, two oxidation peaks are observed at ~2.40 and 2.43 V, which relate to the reverse reaction of Li2S2/Li2S to higher-order polysulfides (Li2Sx, x = 48), then to S, respectively. It can also be observed that the curves from the first to fifth cycles are almost overlapped, indicative of excellent cycle stability of the LSC/S@C cathode. Figure 3b shows the discharge-charge profiles of the LSC/S@C cathode at different cycles (0.5 C). The voltage-capacity curves show two discharge plateaus located at 2.20 and 2.05 V, and two adjacent charge plateaus at 2.30 and 2.35 V, which are well consistent with the CV results. Figure 3c exhibits the rate performance of the LSC/S@C cell with 2.1 mg cm2 sulfur loading. The reversible capacities are 1475, 1210, 1059, 923, 822, 725, 621, 475 and 442 mAh g1 at 0.05, 0.1, 0.2, 0.5, 1, 1.5, 2, 3 and 4 C, respectively. When the rate is reset back to 0.1 C after running at various rates, a reversible capacity of 1077 mAh g1 is recovered, corresponding to a capacity retention of 89.0%. Figure 3d shows the voltage-capacity curves at different currents. The charge/discharge plateaus rise/drop as the current density increases, but two discharge plateaus maintain well throughout. Figure 3e shows the long-term cycle performance of the LSC/S@C cathode. With a mass loading of 2.1 mgsulfur cm2, it achieves an

11

initial capacity of 996 mAh g1 and maintains 852 mAh g1 after 400 cycle (0.5 C), corresponding to a capacity fading of only 0.039% per cycle, indicative of excellent capacity retention and cycle stability. In addition, the Coulombic efficiency (CE) of the LSC/S@C cell is stabilized at > 99% during the long-term cycles. In order to confirm the contribution from LSC itself to the total capacity of the cathode, a LSC cathode was prepared and tested at the same charge-discharge condition. The LSC cathode presents a capacity of only ~20 mAh g1 (Figure S4a and 4b), thus the contribution of LSC/C host itself to the reversible capacity can be ignored. The electrochemical performances of the coaxial S@C nanofiber cathode (without LSC) are also shown in Figure 3e and Figure S5. With similar S loading (2.0 mgsulfur cm2), the S@C cathode exhibits comparable S utilization in the initial 20 cycles, but in the long-term cycles, the specific capacity drops much more quickly than that of the cathode with LSC, which implies that LSC as polysulfide immobilizer plays a critical role in the long-term cycles. In addition, we can see from Figure S6a and 6b that the disassembled S@C cell shows drastic color change from initial colorless to bright yellow after 270 cycles, whereas the LSC/S@C cell keeps almost colorless throughout. The deeper color means the higher polysulfide concentration. The UV-Vis absorption analysis in Figure S6c gives the similar result. Sulfur loading is one of the key factors for actual energy density of Li-S battery. The electrochemical performance always deteriorates with increasing S loading. Compared with traditional S-based composites, the LSC/S@C architecture possesses

12

abundant interconnected conductive channels, which can promise high-efficiency electron and ion transfer and thus help to improve the electrochemical accessability of the insulating S; and the 1D structure can also aid to stabilize the electrode structure and therefore maintain the cycle stability. Here, the S loadings are improved from 1.4 to 5.4 mgsulfur cm2. As shown in Figure 4a, all the cathodes obtain reversible specific capacity about 1000 mAh g2 at 0.2 C, showing high S utilization. Figure 4b shows the areal discharge capacity of the LSC/S@C electrodes with various S loadings. A high areal discharge capacity of 4.0 mAh cm2 is achieved by the electrode with 3.3 mgsulfur cm2 loading (0.05 C), and 7.8 mAh cm2 is obtained by the cathode with 5.4 mgsulfur cm2 loading. The electrodes with sulfur loading from 1.4 to 3.3 mgsulfur cm2 show excellent cyclability (Figure 4c). For the electrode with 5.4 mgsulfur cm2 loading, the initial capacity is 915 mAh g1, and the reversible capacity is stabilized at 728 mAh g1, i.e. 4 mAh cm2, after 100 cycles (Figure 4d). Obviously, the 5.4 mgsulfur cm2 cathode does not achieve good long-term cycle performance like the 2.1 mgsulfur cm2 cathode. Here the anode degradation may be mainly responsible for the cell failing. The higher the sulfur loading, the higher current density the lithium anode bears [42], which makes the lithium anode degrade more seriously. It has been reported that Co metal can help improve the discharge capacity of sulfur at low current rates by effectively transforming long-chain polysulfides to short-chain ones [43]. In the LSC/S system, we found that the Sr doping plays a critical role in the improvement of the Li–S performance. On the one hand, Sr-doping

13

creates abundant oxygen vacancies and valence variation of Co in LSC and therefore results in a largely improved electron transport capability: the conductivity of LSC is 52.63 S cm1, much higher than that of LaCoO3 (LCO) (0.12 S cm1). On the other hand, Sr-doping also affects the adsorption performance of LSC. The visualized experiment in Figure S7 gives a straightforward evidence: The addition of LSC caused a dramatic color fading for the Li2S4 solution from deep orange-red to almost colorless, while LCO did not lead to any visible color change. Figure S8 also shows the colour change of Li2S4 solutions with the addition of LSC with different Sr content (x = 00.7). We can see that the samples with x = 0.30.6 lead to similar colour change. By considering that La1-xSrxCoO3-δ achieve high conductivity at around x = 0.30.4) [40, 41], we chose La0.6Sr0.4CoO3-δ as the polysulfide immobilizer in this work. Furthermore, Figure S9 proves that the LSC/S@C cathode attains better specific capacity and cycle stability than LCO/S@C, further confirming the higher conductivity and immobilizer ability of LSC/S@C cathode. In order to get insights into the interaction mechanism between LSC and polysulfide species, the first-principle spin-polarized density functional theory (DFT) calculations were performed via the Vienna ab initio Simulation Package (VASP). An energy cutoff of 400 eV was employed in all calculations. Brillouin zone was sampled using the Monkhorst-Pack k-points of (7 × 7 × 7) and (3 × 3 × 1) for bulk and slab calculations, respectively. The surfaces of LCO (110) and LSC (110) were modeled with a two-layer slab using a 4×2 supercell, in which the LSC (110) surface was

14

constructed by replacing La with Sr and producing surface oxygen vacancies in LCO (110) according to the experimental characterized oxygen vacancy ratio (Figure S10). Geometry optimizations in both bulk and surface calculations are performed until the Hellmann–Feynman force on each atom is smaller than 0.02 eV/Å. The adsorption energy (Ea) of Li2S4 was calculated using the equation: Ea = Esurface-molecule – (Esurface + Emolecule)

(1)

where Esurface-molecule, Esurface and Emolecule are the total energy of the whole system, the substrate surface and the Li2S4 molecule, respectively. The negativity of Ea indicates the adsorption strongth. The adsorption structures of Li2S4 on the LCO (110) and LSC (110) surfaces are presented in Figure 5 and Table 1. As shown in Table 1, Li atoms bond with surface O atoms with the similar bond lengths of 1.83 and 1.82 Å, respectively, on both surfaces, while the Co-S bond length is shortened from 2.24 Å on LCO (110) to 2.19 Å on LSC (110). The adsorption energies of Li2S4 are -2.59 eV on LSC (110) and -1.95 eV on LCO (110), indicating that stronger interaction happens between Li2S4 and LSC (110) than between Li2S4 and LCO (110). To understand the underlying reason for the larger adsorption strength of Li2S4 to LSC (110), the surface electron properties of Li2S4 adsorbed on LSC (110) and LCO (110) are further studied by charge density difference and Bader charge analysis. The plots of charge density difference in Figure S11 indicate that the S atoms of Li2S4 interact with the surface Co atoms with the feature of ionic interaction on both surfaces. Moreover, the Bader charge analysis

15

shows that the average valences of the two surface Co atoms at the adsorption site are +0.93 and +0.92 in LCO (110) and LSC (110), respectively, while the average valences of S atoms bonded with Co atoms are -0.71 in LCO (110) and -0.79 in LSC (110). Obviously, Sr-doping gives rise to a larger difference in electronic charge between Co and S in LSC (110), resulting in an increased surface binding energy of Li2S4 in LSC (110). The improvement of the adsorption energy implies stable adsorption between Li2S4 and LSC (110), suggesting that LSC can serve as a suitable polysulfide immobilizer. In addition, the interaction between Li2S4 and C was also calculated (Figure S12), which shows that Li2S4 adsorbs on carbon surface with an adsorption energy of -0.07 eV. This demonstrates that C immobilizes polysulfides via physical interaction. The physical interaction between C and S atoms is weak so that it cannot immobilize polysulfides effectively in the long-term cycle, which is why the S@C cell degrades quickly. The calculation results well agree with the experimental observation that LSC has better adsorption property to Li2S4 than LCO. Figure 5c and 5d show the XPS spectra of Co 2p and S 2p for LSC and LSC/Li2S4. The Co 2p3/2 peak at 780.2 eV for LSC matches well with previous reports [44, 45]. When mixed with Li2S4 solution, the Co 2p and S 2p peaks shift by -0.2 and +0.5 eV, respectively, which demonstrates that the valences of Co and S decrease, indicating the existence of Co-S bond. Figure S13a and 13b show XPS spectra of La 3d and Sr 3d for LSC and LSC/Li2S4. The XPS peaks of La 3d and Sr 3d locate at almost the same position in LSC and LSC/Li 2S4,

16

indicative of the chemical stability of La and Sr in the S system. Thus, we can draw a tentative conclusion that Sr doping results in valence variation in Co along with oxygen vacancy. The Co ions with mixed valence have strong adsorption to the polysulfide ions while the existence of oxygen vacancy enhances the binding strength between Li2S4 and LSC (110) rather than LCO (110). We use O 1s XPS spectra of LCO and LSC to further discuss the oxygen vacancies (Figure S14). The peak at low binding energy (BE) of 528.9 eV can be assigned to the lattice oxygen (O2), the peak at intermediate position (529.7 eV) is related to the surface-oxygen species (MO, M = Sr, La, Co), and the peak at high BE (531.4 eV) can also be ascribed to the surface-bound oxygen species [44]. Compared with LCO, the decrease of the lattice oxygen peak and the appearance of the surface-oxygen species (MO) peak in the LSC spectra demonstrate an increase of the oxygen vacancies in LSC. In addition, the accurate oxygen content (3-δ) and average valence of Co ions in LSC were determined by iodometric titration. As shown in Table S2, the average valence of Co in LSC is 3.11 and the amount of oxygen vacancy (δ) increases to 0.14 along with Sr-doping in LCO. In order to investigate the stability of LSC immobilizer in the sulfur system, TEM, XRD and XPS tests were conducted to compare the LSC/S@C cathodes before and after cycling (Figure S15). We can clearly see from the TEM images that the core-shell nanostructure for the LSC/S@C cathode keeps the same before and after 100 cycles. Moreover, no observable difference can be found in the XRD pattern before and after 100 cycles, indicating that the LSC immobilizer is quite

17

stable. In addition, there is no obvious peak shift in the La 3d, Sr 3d and Co 2p XPS spectra after long cycles, further confirming the stability of the LSC immobilizer. The experimental results and theoretical calculations demonstrate that Sr doping result in valence variation of Co along with oxygen vacancy; The Co ions with mixed valence have strong adsorption to the polysulfide ions while the existence of oxygen vacancy enhances the binding strength between Li2S4 and LSC. The chemical interaction between LSC and polysulfides plays a critical role in achieving excellent cycle stability for the LSC/C@S cathode. The carbon shell in LSC/C@S also contributes to the improvements via physical entrapment. Moreover, the highly conductive LSC@C host and the porous interconnected fiber web-like architecture also facilitate the mass transfer during charge/discharge process synchronously.

4 Conclusion In summary, we have demonstrated that the perovskite-type La0.6Sr0.4CoO3-δ can work well as a new type polysulfide immobilizer in Li-S batteries. Sr doping leads to valence variation of Co and meanwhile generates oxygen vacancies, which not only largely improves electronic conductivity, but also greatly facilitates the adsorption interaction between Co atoms and polysulfides. When the coaxial yolk-shell LSC@C nanofibers is used as hybrid host for S species, the resulted LSC/S@C composite cathode with a mass loading of 2.1 mgsulfur cm2 presents a high reversible capacity of 996 mAh g1 and long-term cycle stability with only 0.039% capacity decay per cycle over 400 cycles at 0.5 C. Even the S loading is increased to 5.4 mgsulfur cm2, the 18

LSC/S@C composite cathode can still achieve comparable S utilization and good cyclability. Our results thus demonstrate that the LSC/S@C composite cathode has great potential to be applied in high-energy Li–S batteries. Furthermore, perovskite oxides are a big family with high chemical stabilization, which component, cation valence and oxygen vacancy can be well tuned and controlled. Therefore, we believe this work will shed some new light on the discoveries of new high-efficiency polysulfide immobilizers and hence the design of high-performance sulfur cathode.

Acknowledgements Z.X.H. and R.Z. have contributed equally to this work. This work was supported by the 973 program (Grant No. 2015CB258400), and the National Science Foundation of China (Grant Nos. 51532005, 21273087, 21373098 and 51361130151). The authors acknowledge the Analytical and Testing Center of Huazhong University of Science and Technology (HUST) for XRD, XPS, FESEM and FTEM measurement (special thanks for Mr. Zhao’s FTEM operation), and the State Key Laboratory of Material Processing and Die & Mould Technology of HUST for TG and BET test.

References [1] Q. Pang, X. Liang, C. Y. Kwok, L. F. Nazar, Nature Energy 1 (2016) 16132. [2] D. Sun, Y. Hwa, Y. Shen, Y. H. Huang, E. J. Cairns, Nano Energy 26 (2016) 524-532. [3] Y. Sun, N. Liu, Y. Cui, Nature Energy 1 (2016) 16071. [4] A. Manthiram, S. H. Chung, C. Zu, Adv. Mater. 27 (2015) 1980-2006. [5] Y. Zhao, Y. Ding, Y. Li, L. Peng, H. R. Byon, J. B. Goodenough, G. Yu, Chem. Soc. Rev. 44 (2015) 7968-7996.

19

[6] Z. Li, L. Yuan, Z. Yi, Y. Sun, Y. Liu, Y. Jiang, Y. Shen, Y. Xin, Z. Zhang, Y. Huang, Adv. Energy Mater. 4 (2014) 1301473. [7] J. Yan, B. Li, X. Liu, Nano Energy 18 (2015) 245-252. [8] Z. Li, J. T. Zhang, Y. M. Chen, J. Li, X. W. Lou, Nat. Commun. 6 (2015) 8850. [9] R. Fang, S. Zhao, P. Hou, M. Cheng, S. Wang, H. M. Cheng, C. Liu, F. Li, Adv. Mater. 28 (2016) 3374-3382. [10] X. Ji, K. T. Lee, L. F. Nazar, Nat. Mater. 8 (2009) 500-506. [11] S. H. Chung, A. Manthiram, Adv. Mater. 26 (2014) 7352-7357. [12] Z. X. Hao, L. X. Yuan, Z. Li, J. Liu, J. W. Xiang, C. Wu, R. Zeng, Y. H. Huang, Electrochim. Acta 200 (2016) 197-203. [13] J. Liu, L. Yuan, K. Yuan, Z. Li, Z. Hao, J. Xiang, Y. Huang, Nanoscale 8 (2016) 13638-13645. [14] J. Zhang, J. Zhao, L. Yue, Q. Wang, J. Chai, Z. Liu, X. Zhou, H. Li, Y. Guo, G. Cui, L. Chen, Adv. Energy Mater. 5 (2015) 1501082. [15] Y. Zhang, Y. Zhao, D. Gosselink, P. Chen, Ionics 21 (2014) 381-385. [16] D. Lin, Y. Liu, Z. Liang, H. W. Lee, J. Sun, H. Wang, K. Yan, J. Xie, Y. Cui, Nat. Nanotech. 11 (2016) 626-632. [17] R. Zhang, X. B. Cheng, C. Z. Zhao, H. J. Peng, J. L. Shi, J. Q. Huang, J. Wang, F. Wei, Q. Zhang, Adv. Mater. 28 (2016) 2155-2162. [18] K. Yan, Z. Lu, H. W. Lee, F. Xiong, P. C. Hsu, Y. Li, J. Zhao, S. Chu, Y. Cui, Nature Energy 1 (2016) 16010. [19] Z. Li, L. Yuan, Z. Yi, Y. Liu, Y. Xin, Z. Zhang, Y. Huang, Nanoscale 6 (2014) 1653-1660. [20] N. Jayaprakash, J. Shen, S. S. Moganty, A. Corona, L. A. Archer, Angew. Chem., Int. Ed. 50 (2011) 5904-5908. [21] Z. Li, J. Zhang, X. W. Lou, Angew. Chem., Int. Ed. 54 (2015) 12886-12890. [22] J. Zhang, H. Hu, Z. Li, X. W. Lou, Angew. Chem., Int. Ed. 55 (2016) 3982-3986. [23] W. Li, Q. Zhang, G. Zheng, Z. W. Seh, H. Yao, Y. Cui, Nano Lett. 13 (2013) 5534-5540. [24] H. Xu, L. Qie, A. Manthiram, Nano Energy 26 (2016) 224-232. [25] Y. Zhao, W. Wu, J. Li, Z. Xu, L. Guan, Adv. Mater. 26 (2014) 5113-5118. [26] M. Q. Zhao, H. J. Peng, G. L. Tian, Q. Zhang, J. Q. Huang, X. B. Cheng, C. Tang, F. Wei, Adv. Mater. 26 (2014) 7051-7058. [27] X. B. Cheng, J. Q. Huang, Q. Zhang, H. J. Peng, M. Q. Zhao, F. Wei, Nano Energy 4 (2014) 65-72. [28] C. Tang, Q. Zhang, M. Q. Zhao, J. Q. Huang, X. B. Cheng, G. L. Tian, H. J. Peng, F. Wei, Adv. Mater. 26 (2014) 6100-6105. [29] R. Chen, T. Zhao, W. Wu, F. Wu, L. Li, J. Qian, R. Xu, H. Wu, H. M. Albishri, A. S. Al-Bogami, D. A. El-Hady, J. Lu, K. Amine, Nano Lett. 14 (2014) 5899-5904. [30] X. Wang, G. Li, J. Li, Y. Zhang, A. Wook, A. Yu, Z. Chen, Energy Environ. Sci. 9 (2016) 2533-2538. [31] X. Liang, C. Hart, Q. Pang, A. Garsuch, T. Weiss, L. F. Nazar, Nat. Commun. 6 (2015) 5682. [32] X. Tao, J. Wang, Z. Ying, Q. Cai, G. Zheng, Y. Gan, H. Huang, Y. Xia, C. Liang, W. Zhang, Y. Cui, Nano Lett. 14 (2014) 5288-5294.

20

[33] Q. Pang, D. Kundu, M. Cuisinier, L. F. Nazar, Nat. Commun. 5 (2014) 4759. [34] X. Han, Y. Xu, X. Chen, Y. C. Chen, N. Weadock, J. Wan, H. Zhu, Y. Liu, H. Li, G. Rubloff, C. Wang, L. Hu, Nano Energy 2 (2013) 1197-1206. [35] K. Sun, D. Su, Q. Zhang, D. C. Bock, A. C. Marschilok, K. J. Takeuchi, E. S. Takeuchi, H. Gan, J. Electrochem. Soc. 162 (2015) A2834-A2839. [36] J. Pu, Z. Shen, J. Zheng, W. Wu, C. Zhu, Q. Zhou, H. Zhang, F. Pan, Nano Energy 37 (2017) 7-14. [37] R. Demir-Cakan, M. Morcrette, F. Nouar, C. Davoisne, T. Devic, D. Gonbeau, R. Dominko, C. Serre, G. Ferey, J. M. Tarascon, J. Am. Chem. Soc. 133 (2011) 16154-16160. [38] X. Liang, C. Y. Kwok, F. Lodi-Marzano, Q. Pang, M. Cuisinier, H. Huang, C. J. Hart, D. Houtarde, K. Kaup, H. Sommer, T. Brezesinski, J. Janek, L. F. Nazar, Adv. Energy Mater. 6 (2016) 1501636. [39] Z. Cui, C. Zu, W. Zhou, A. Manthiram, J. B. Goodenough, Adv. Mater. 28 (2016) 6926-6931. [40] V. Øygarden, H. L. Lein, T. Grande, J. Solid State Chem. 192 (2012) 246-254. [41] Y. Tao, J. Shao, J. Wang, W. G. Wang, J. Power Sources 185 (2008) 609-614. [42] L. Qie, A. Manthiram, Adv. Mater. 27 (2015) 1694-1700. [43] Y. J. Li, J. M. Fan, M. S. Zheng, Q. F. Dong, Energy Environ. Sci. 9 (2016) 1998-2004. [44] D. H. Prasad, S. Y. Park, E. O. Oh, H. Ji, H. R. Kim, K. J. Yoon, J. W. Son, J. H. Lee, Appl. Catal., A: General 447-448 (2012) 100-106. [45] X. Cheng, E. Fabbri, M. Nachtegaal, I. E. Castelli, M. El Kazzi, R. Haumont, N. Marzari, T. J. Schmidt, Chem. Mater. 27 (2015) 7662-7672.

21

Figure 1 Schematic illustration of synthesis of LSC/S@C composite cathode.

22

Figure 2 SEM image of (a) LSC fibers. TEM images of (b,c) LSC fibers, (d) LSC@SiO2, (e) LSC@SiO2@C and (f) LSC/S@C. (g) XRD patterns of LCO (JCPDS No. 75-0279), LSC (No. 48-0121), LSC/Li2S4 and LSC/S@C (No. 99-0066). (h) N2 adsorption/desorption isotherms of LSC fibers.

23

Figure 3 Electrochemical performances for LSC/S@C cathode: (a) CV curves of initial five cycles at a scan rate of 0.05 mV s1, (b) voltage vs. specific capacity profiles at 0.5 C, (c) discharge capacity at various rates, and (d) voltage vs. specific capacity profiles at various rates. (e) Comparison in specific capacity and Coulumbic efficiency over cycling at 0.5 C between LSC/S@C and S@C cathodes.

24

Figure 4 Effects of S loading for LSC/S@C cathode: (a) voltage profiles at 0.2 C with various sulfur loading, (b) comparison of areal discharge capacity at 0.05 C, and (c) areal discharge capacity over cycling at 0.2 C. (d) Cycling performance of the thick LSC/S@C cathode with 5.4 mgsulfur cm2 loading at 0.2 C.

25

Figure 5 The most stable adsorption structures, after full relaxation, of Li2S4 on (a) LaCoO3 (110) and (b) La0.6Sr0.4CoO3-δ (110) surfaces (bottom panels for side views). (c) Co 2p XPS spectra of LSC and LSC/Li2S4, (d) S 2p XPS spectra of LSC and LSC/Li2S4.

26

Table 1. Adsorption structures and energies of Li2S4 on LCO (110) and LSC (110) Surface Li-O bond length (Å) Co-S bond length (Å) Ebind (eV) LaCoO3 (110) 1.83 2.24 -1.95 1.82 2.19 -2.59 La0.6Sr0.4CoO3-δ(110)

27

Author Information

Dr. Zhangxiang Hao received his BSc, MSc degree from Wuhan University of Technology (WHUT), and PhD degree from Huazhong University of Science and Technology (HUST). His current research focuses on rechargeable lithium-sulfur battery.

Mr Rui Zeng received his MSc degree from Huazhong University of Science and Technology (HUST) in 2013. He is now a PhD candidate in HUST. His current research focuses on solid oxide fuel cells.

Dr. Lixia Yuan received her BSc, MSc and PhD from Wuhan University. She is now a professor at Huazhong University of Science and Technology (HUST). Her research interest mainly focuses on rechargeable lithium-sulfur batteries.

28

Mr Qiming Bing received his BSc degree from Jilin University (JLU) in 2014. He is now a PhD candidate in JLU. His current research focuses on the theoretical study of formic acid catalytic decomposition mechanism.

Dr. Jingyao Liu received her BSc, MSc and PhD from Jilin University. In 2005, she became a professor in Institute of Theoretical Chemistry, Jilin University. Her current research focuses on the theoretical and computational studies on electronic structures of two-dimensional materials and the mechanisms of heterogeneous/homogeneous catalytic reactions.

Mr Jingwei Xiang received his BSc degree in School of Materials Science and Engineering from Huazhong University of Science and Technology. He is now a PhD candidate in HUST. His current research focuses on Li-S battery and Li metal anode.

29

Dr. Yunhui Huang received his BSc, MSc and PhD from Peking University. From 2002, he worked as a researcher in Tokyo Institute of Technology and University of Texas at Austin. In 2008, he became a chair professor of materials science in Huazhong University of Science and Technology. His research group work on the materials for energy storage and conversion. For details please see the lab website: http://www.sysdoing.com.cn.

Graphical Abstract

30

Highlights • A novel perovskite La0.6Sr0.4CoO3-δ (LSC) immobilizer is firstly introduced into the Li-S system. • A dual core-shell structure of LSC/S@C cathode have successfully designed and fabricated. • The chemical bond beweem LSC and polysulfide is confirmed by the first-principle spin-polarized density functional theory calculation, XPS spectra, visible absorbent test and iodometric titrations. • With a high sulfur loading of 5.4 mg cm-2, the LSC/S@C cathode still deliver comparable sulfur utilization and excellent cycling stability.

31