PDF (10 MB) - ACS Publications - American Chemical Society

Jun 3, 2019 - for the conversion of polysulfides to insoluble end products (S or Li2S/Li2S2), and its deployment as a cathode-integrated interlayer wa...
0 downloads 0 Views 10MB Size
Download PDF
www.acsnano.org

A Cathode-Integrated Sulfur-Deficient Co9S8 Catalytic Interlayer for the Reutilization of “Lost” Polysulfides in Lithium−Sulfur Batteries Downloaded via UNIV OF SOUTHERN INDIANA on July 17, 2019 at 13:09:07 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

Haibin Lin, Shengliang Zhang, Tianran Zhang, Sheng Cao, Hualin Ye, Qiaofeng Yao, Guangyuan Wesley Zheng,* and Jim Yang Lee* Department of Chemical and Biomolecular Engineering, National University of Singapore, 10 Kent Ridge Crescent, Singapore 119260, Singapore S Supporting Information *

ABSTRACT: Lithium−sulfur batteries, with their high theoretical energy density and the low material cost of sulfur, are highly promising as a post-lithium ion battery contender. Their current performance is however compromised by sulfur loss and polysulfide shuttle to result in low energy efficiency and poor cycle stability. Herein, a catalytic material (Co9S8−x/CNT, nanoparticles with a metallic Co9S8 core and a sulfur-deficient shell on a CNT support) was applied as an interlayer on the sulfur cathode to retain migratory polysulfides and promote their reutilization. The Co9S8−x/CNT catalyst is highly effective for the conversion of polysulfides to insoluble end products (S or Li2S/Li2S2), and its deployment as a cathode-integrated interlayer was able to retain the polysulfides in the cathode for reuse. The accumulation of polysulfides in the electrolyte and the polysulfide shuttle were significantly reduced as a result. Consequently, a host-free sulfur cathode with the Co9S8−x/CNT interlayer had a low capacity fade rate of 0.049% per cycle for 1000 cycles at a 0.3C rate, a significant improvement of the capacity fade rate without it (0.28% per cycle for 200 cycles). The results here provide not only direct evidence for the contributions of sulfur deficiencies on the catalytic activity of Co9S8 in polysulfide conversion reactions but also the methodology on how the catalyst should be deployed in a Li−S battery for the best catalytic outcome. KEYWORDS: catalytic interlayer, sulfur deficiency, cobalt sulfide, polysulfide conversions, lithium−sulfur batteries

A

host, an electrically conducting interlayer may be inserted between the sulfur cathode and the separator and serves as a physical barrier against the polysulfide diffusion to the lithium metal anode.9−11 Polysulfide-anchoring materials can be added to the interlayer to increase the polysulfide capture efficiency.12,13 We envisage “polysulfide reutilization” in the interlayer would be an even more effective concept than “polysulfide retention” in decreasing the polysulfides in the electrolyte. The reutilization of polysulfides in the interlayer has to rely on the presence of an effective “catalyst”, which can accelerate the conversion of dissolved polysulfides to insoluble products (and vice versa), and adequate substrate electrical conductivity. Since “polysulfide catalysis” is a fairly recent concept, the rational design and integration of the polysulfide conversion catalyst with the interlayer and their localization in

mong the various battery chemistries introduced as potential successors to the highly successful lithiumion batteries, lithium−sulfur batteries have drawn a very strong following because of the high theoretical capacity (about 1675 mAh g−1) and energy density (2600 Wh kg−1) of the sulfur cathode and its low cost and environmental impact.1−3 The practical performance of lithium−sulfur batteries is however affected by several difficult technical issues. The most challenging among them is the loss of electrochemically active sulfur as soluble intermediate products (lithium polysulfides, Li2Sx, 2 < x ≤ 8). The polysulfides are migratory and reactive toward the Li anode. Their redox shuttling in lithium−sulfur batteries causes low Coulombic efficiency and capacity losses during use4,5 and even internal short circuiting in extreme cases of anode passivation.4−6 The most common method to “contain” the sulfur loss is sulfur entrapment in a host material.7,8 This often involves the design of an elaborate cathode host structure, a complex sulfur encapsulation process, and introduces an inevitable “dead weight” effect. As an alternative to the use of a sulfur cathode © 2019 American Chemical Society

Received: March 27, 2019 Accepted: May 30, 2019 Published: June 3, 2019 7073

DOI: 10.1021/acsnano.9b02374 ACS Nano 2019, 13, 7073−7082

Article

Cite This: ACS Nano 2019, 13, 7073−7082

Article

ACS Nano

Scheme 1. Schematic illustration of the synthesis of Co9S8−x/CNT with a metallic Co9S8 core and a sulfur-deficient shell on a CNT support and the conversion of Li2Sx on the Co9S8−x/CNT surface.

Figure 1. Characterizations of Co9S8/CNT composites heat-treated at different temperatures in a hydrogen atmosphere. (a) XRD patterns, (b) ideal crystal structure of Co9S8, (c, d) expanded view of XRD patterns in 2θ regions of interest; and (e) high-resolution XPS Co 2p spectra.

produce such a sophisticated construct can be challenging and costly. Furthermore, the integrity of such a composite catalytic system can be undermined by the volume change in the S ↔ Li2S transformations occurring in the sulfur core. Therefore, the search for more catalyst candidates with high activity, high electrical conductivity, and low material cost and the maximization of their utility in lithium−sulfur batteries through a rational deployment approach are the success factors for a rewarding “polysulfide catalysis” solution. We envisage the localization of an effective electrocatalyst in a cathode-integrated interlayer as the best strategy. The interlayer can be applied as a thin surface coating on the solution side of the sulfur cathode. It serves as a cathode extension and also the chokepoint for polysulfide passage. The localization of an efficient polysulfide catalyst in the interlayer, with increased catalyst−polysulfide contacts, is more able to

the Li−S battery for the best outcome are still unexplored ideas as of now. Transition metal sulfides such as molybdenum disulfide (MoS2) and tungsten disulfide (WS2) form the largest class of polysulfide conversion catalysts investigated to date. Good intrinsic catalytic activity and low material cost are their strongest features.14,15 Their main disadvantage is low electrical conductivity, which has affected the effectiveness of their deployment and their practical performance. Almost all of the polysulfide electrocatalysts in use today are administered as additives for the sulfur cathode. The polysulfide capture efficiency is then limited by the number and quality of electrochemical contacts between the polysulfides and the electrocatalyst during battery charging and discharging. While sulfur particles coated with the catalyst may appear to be a solution to the problem, the specific chemistry required to 7074

DOI: 10.1021/acsnano.9b02374 ACS Nano 2019, 13, 7073−7082

Article

ACS Nano

Figure 2. Morphology and structure of Co9S8−x/CNT obtained by heating Co9S8/CNT in a hydrogen atmosphere at 700 °C. (a) FESEM and (b−e) TEM images. (f) HRTEM image, the corresponding FFT pattern, and the schematic of atomic arrangement. (g, h) EDS elemental maps.

convert the retained polysulfides effectively to either sulfur or Li2S. The concept will be demonstrated in this paper with the use of a catalytic interlayer of CNTs decorated with Co9S8 nanoparticles with a sulfur-deficient shell (Co9S8−x/CNT) and a Co9S8 core. The use of Co9S8 as a catalyst in lithium−sulfur batteries was inspired by their catalysis of the polysulfide mediators in quantum-dot-sensitized solar cells.16 While Co9S8 has been used as a polysulfide host for the sulfur cathode,17 its catalytic effect on polysulfide conversion is still unelucidated due to limited research on the electrochemical behavior of Co9S8. Co9S8 was also deployed within a complex host structure which is not conducive to adoption for practical use.18 On the other hand, there is evidence to suggest that sulfur deficiencies in transition metal sulfides may increase the efficacy of polysulfide conversion catalysis due to the higher electrochemical activity of sulfur deficiencies.19,20 Sulfur deficiencies can be easily incorporated in Co9S8 via a core− shell construction (see Methods section). The design then takes advantage of the sulfur-deficient shell for polysulfide adsorption and catalysis, and the crystalline Co9S8 core for electrical conductivity and particle stability.21 As an added assurance we also dispersed the core−shell Co9S8 nanoparticles on carbon nanotubes (CNTs) to inhibit the aggregation of the core−shell nanoparticles, to improve interparticle electron conduction, and also to leverage the ease of 1D nanostructure entanglements to increase the polysulfide encounter and the capture efficiency.22,23

The Co9S8−x/CNT composite was prepared by a facile hydrothermal process followed by heat treatment in hydrogen at elevated temperature to generate the requisite sulfur deficiency (Scheme 1). It was then used as a catalyst and applied directly to the sulfur cathode exterior surface (the solution side) without any additional material. This method of application addresses not only the volume density penalty associated with the use of a standalone interlayer but also an inherent disadvantage of the latterthat the effectiveness of catalyst coating is dependent on the interlayer pore system and potential polysulfide leakage through the larger pores in the interlayer. The experimental results corroborated the effectiveness of the design heuristics: that polysulfide diffusion in lithium−sulfur batteries was effectively suppressed in the presence of the cathode-integrated Co9S8−x/CNT interlayer. The effective reutilization of polysulfides enabled a host-free sulfur cathode to exhibit good cycle stability for an impressive number of cycles (capacity fade rate of 0.049% per cycle for 1000 cycles at the 0.3C rate). Even with the unoptimized integrated interlayer design in this study, good application performance can still be obtained from this rational approach to the interlayer design and deployment.

RESULTS AND DISCUSSION A simple hydrothermal procedure was first used to form Co9S8/CNT as the catalyst precursor. It was then heated in hydrogen at different temperatures to vary the sulfur-deficiency 7075

DOI: 10.1021/acsnano.9b02374 ACS Nano 2019, 13, 7073−7082

Article

ACS Nano

Figure 3. Elucidation of the polysulfide conversion kinetics in lithium−catalyst cells. (a) Current−time curves from the potentiostatic discharging from the open-circuit condition (2.2 V) to 2 V. (b) XPS S 2p spectra of the working electrodes after the potentiostatic discharge. (c) Current−time curves from the potentiostatic charging to 2.4 V. (d) XPS S 2p spectra of the working electrodes after potentiostatic charging. (e) Cyclic voltammograms at 0.1 mV s−1 in the 1.7−2.6 V windows. (f) Electrochemical impedance spectra of lithium−catalyst cells under open-circuit conditions.

content in Co9S8/CNT. The XRD characterization of the sulfur-deficient Co9S8/CNT microstructure (Figure 1a) yielded similar XRD patterns which can be indexed to the cubic Fm3̅m space group of Co9S8 (PDF#65-6801, Figure 1b).24 Heat treatment in hydrogen had therefore not affected the Co9S8 crystal structure significantly. Some minor differences in the patterns were however detected in the enlarged views of selected 2θ regions (Figure 1c,d). The slight shifts of the peaks between 29° and 32° and those between 51° and 53° to lower 2θ values can be understood in terms of lattice parameter expansion. Lattice expansion occurred because of the increasing presence of Co in low oxidation states after hydrogen reduction, which have larger atomic radii.25

The surface compositions of sulfur-deficient Co9S8 prepared under different heat-treatment temperatures were determined by X-ray photoelectron spectroscopy (XPS). Table S1 is a summary of the cobalt (Co) and sulfur (S) contents as determined by XPS. The Co:S ratio of as-synthesized Co9S8 prior to the hydrogen treatment was 52.89:47.11, which is close to the 9:8 ratio in stoichiometric Co9S8. The Co:S ratio increased with the increase in hydrogen treatment temperature due to the progressive removal of the surface sulfur. The Co 2p XPS spectrum in Figure 1e shows the typical Co 2p3/2 and Co 2p1/2 spin−orbit doublets and their associated shakeup satellites (identified as “Sat.”). The Co 2p3/2 doublet at about 780 eV could be deconvoluted into two peaks, with the peak at 778.4 eV assignable to Co+ and the peak at 781.1 eV to 7076

DOI: 10.1021/acsnano.9b02374 ACS Nano 2019, 13, 7073−7082

Article

ACS Nano

Figure 4. Evaluation of the catalytic action by the symmetric cell method. (a) Cyclic voltammograms of symmetric cells with different electrodes (Co9S8−x/CNT, Co9S8/CNT, or CNT). (b) Tafel plots of the symmetric cells.

Co2+.26,27 A higher heat temperature increased the amount of Co+ along with the increase in sulfur deficiency. Although the 900 °C sample had the highest sulfur deficiency content, the severity of the treatment also resulted in increased Co9S8 nanoparticle aggregation (Figure S1). The Co9S8−x/CNT composite treated at 700 °C with the next highest sulfur deficiency content and a stable structure was therefore selected for further investigations. The morphology of this stable Co9S8−x/CNT composite with high sulfur deficiency was examined by both field emission scanning electron microscopy (FESEM) and transmission electron microscopy (TEM). The FESEM and TEM images in Figure 2a,b show a porous CNT network with fairly evenly distributed Co9S8−x nanoparticles. Energy dispersive spectroscopy (EDS) of the Co9S8−x/CNT composite in Figure 2g,h also verified the uniform distribution of all constituent elements. The microstructure of individual Co9S8−x nanoparticles is shown in Figure 2c−e. The Co9S8 nanoparticles all showed a lower crystallinity shell and an intimate contact with the CNT (Figure S2). The different degrees of crystallization between the core and the shell were due to the formation of sulfur deficiencies by the H2 corrosion of the Co9S8 surface in heat treatment. The lattice spacings of 0.29 nm (Figure 2d) and 0.17 nm (Figure 2e) in the deeper subsurface regions agree well with the (311) and (440) diffractions of cubic Fm3̅m Co9S8, indicating that Co9S8 crystallinity was intact in the core area. The high crystallinity of the Co9S8 core was also confirmed by the HRTEM image and the corresponding fast Fourier transform (FFT) pattern in Figure 2f. The sulfurdeficient Co9S8 shell could catalyze the polysulfide conversion, while the crystalline Co9S8 core served as a robust catalyst support to maintain catalyst stability.28 Furthermore, the intimate and electrically conducting interface between the Co9S8 core and CNT would facilitate charge separation and charge conduction from the reaction sites. The key feature of the catalytic interlayer in lithium−sulfur batteries is the accelerated conversion of polysulfides to considerably decrease the presence of polysulfides in the electrolyte and, hence, their diffusion across the separator to the Li metal anode. The catalysis should be proficient for both polysulfide reduction and oxidation in order to work between the two solid end-products (Li2S and S). The effectiveness of Co9S8−x/CNT for polysulfide catalysis was first evaluated in lithium−catalyst cells. The lithium−catalyst cells consisted of a sulfur-free catalyst-only working electrode, a lithium metal

counter electrode, and a 0.2 M Li2S6 electrolyte. The cell assembled as such had an open-circuit voltage of 2.2 V. The kinetics of polysulfide conversion was measured by potential step methods. Two potential steps were used: one for discharging the lithium−catalyst cell from 2.2 V to 2 V to accumulate Li2S on the working electrode and one for charging the lithium−catalyst cell from 2.2 V to 2.4 V to accumulate sulfur on the working electrode.29,30 Figure 3a shows the current−time response from the potentiostatic discharge of two lithium−catalyst cells. The current from the lithium−Co9S8−x/CNT cell first decreased with time and then increased to form a characteristic peak before it decreased again. The discharge current, which included both the instantaneous capacitive discharge current and the Li2S6 reduction current, was initially dominated by the former. The electrochemical deposition of solid Li2S after a nucleation lag caused a current peak to form after the capacitive transient.31 Since Li2S was derived from Li2S6 reduction, the polysulfide reduction kinetics would have a strong impact on the deposition and could be altered through catalysis. The absence of a current peak from the lithium− CNT cell is then an indication of no substantial Li2S deposition on the working electrode without the Co9S8−x/ CNT catalyst. For a more direct proof the working electrodes after the potentiostatic discharge were examined by XPS. Figure 3b shows the S 2p XPS spectra of the two disassembled working electrodes. The ∼162 and ∼165 eV peaks are due to the Li2S deposited on the working electrode after literature consultation.32 A comparison of peak intensities clearly suggests more Li2S deposition on the Co9S8−x/CNT electrode than on the CNT electrode. With this a greater extent of polysulfide reduction (to Li2S) on Co9S8−x/CNT may be inferred. The oxidation of polysulfide to sulfur was examined likewise by the potentiostatic charging of the lithium−catalyst cells (Figure 3c). The higher oxidation peak that prevailed on a Co9S8−x/CNT electrode after the initial capacitive transient likewise suggests a more facile oxidation of Li2S6 on the Co9S8−x/CNT surface. The post-mortem analysis of the working electrodes was again based on the S 2p XPS spectrum (Figure 3d), where the 161.9 and 162.9 eV peaks can be assigned to the Co−S bonds in Co9S8 and the 163.5 and 164.7 eV peaks to the sulfur deposit.14 The increased presence of the sulfur peaks on Co9S8 indicates the greater extent of polysulfide oxidation (to sulfur) on the Co9S8−x/CNT surface. 7077

DOI: 10.1021/acsnano.9b02374 ACS Nano 2019, 13, 7073−7082

Article

ACS Nano

Figure 5. Electrochemical performances of lithium−sulfur batteries with and without a catalyst interlayer. (a) Voltage−time self-discharge curves. (b) Cycle stability and (c, d) corresponding galvanostatic charge−discharge curves. FESEM images of the lithium anode after 100 cycles in the batteries (e) without an interlayer and (f) with a Co9S8−x/CNT interlayer. (g) Cell stability test for 1000 cycles (1C = 1600 mA g−1).

The contributions of sulfur deficiency to the polysulfide redox reactions on Co9S8−x/CNT were also confirmed by the symmetric cell method.14 The symmetric cells contained identical working and counter electrodes and a 0.2 M Li2S6 electrolyte (Figure 4a). A Co9S8/CNT composite heat-treated in N2, which contained very little sulfur deficiency, was used to provide the baseline performance of deficiency-free Co9S8 (Figure S3). The voltammogram of the Co9S8−x/CNT electrode shows four distinct redox peaks at −0.07, −0.37, 0.07, and 0.37 V indicative of high electrochemical reversibility. The electrode reactions that contributed to these peaks are summarized in Figure S4. The deficiency-free Co9S8/CNT

The same conclusion was obtained from cyclic voltammetry and electrochemical impedance (EIS) measurements. Figure 3e shows the cyclic voltammograms of lithium−catalyst cells with different working electrodes. The sharper redox peaks and narrower peak separation for each redox couple in the voltammogram of the lithium−Co9S8−x/CNT cell indicate higher electrochemical reversibility and more facile polysulfide conversions in the presence of Co9S8−x/CNT. EIS measurements (Figure 3f) also registered a considerably smaller resistance for the lithium−Co9S8−x/CNT cell, as shown by the smaller size of the high-frequency semicircle in the Nyquist plot. 7078

DOI: 10.1021/acsnano.9b02374 ACS Nano 2019, 13, 7073−7082

Article

ACS Nano

Figure 6. Effects of the Co9S8−x/CNT interlayer on polysulfide dissolution. In situ UV−vis spectra of the sulfur cathodes (a) without an interlayer and (b) with a Co9S8−x/CNT interlayer at 1.7 V for different amounts of time. (c) Schematic illustration of the Co9S8−x/CNT interlayer effect in a lithium−sulfur battery.

deduced from the Tafel plots,36 where the Co9S8−x/CNT electrode showed higher anodic and cathodic branches. Encouraged by these displays of effective polysulfide catalysis, we proceeded to develop integrated catalytic interlayers based on the Co9S8−x/CNT catalyst. The performance of these interlayers was evaluated in lithium−sulfur coincell batteries with a host-free sulfur cathode and a lithium metal anode. The interlayer was implemented as an extension of the sulfur cathode by coating a thin film of Co9S8−x/CNT on the sulfur cathode exterior, as shown in Figure S5a. Figure S5b−d show the cross-sectional FESEM images of a sulfur cathode with an optimally integrated Co9S8−x/CNT interlayer as a uniform ∼3 μm thick film on a ∼25 μm thick sulfur cathode. A series of electrochemical tests were conducted to verify the benefits of the Co9S8−x/CNT catalytic interlayer in polysulfide

electrode, on the other hand, showed only remnants of these peaks as broad redox features at −0.19, −0.51, 0.19, and 0.51 V, to suggest some decreases in the reversibility of the electrode reactions. The CNT electrode was even less distinctive, showing only a very drawn-out reduction peak at −0.6 V and a very drawn-out oxidation peak at 0.6 V. The lower reversibility and slower kinetics on the defect-free Co9S8/CNT electrode indicated evidently the contributions of surface sulfur deficiencies. It is believed that the sulfur deficiencies rendered the Co9S8−x surface more electron-rich. The adsorption of polysulfides on an electron-rich surface could elongate the S−S bonds to facilitate the bond fissure in reduction.14,33 The sulfur deficiencies in the Co9S8−x surface could also facilitate the formation of metastable Sx•, an important intermediate in the polysulfide reduction and oxidation.34,35 The facile reaction kinetics could also be 7079

DOI: 10.1021/acsnano.9b02374 ACS Nano 2019, 13, 7073−7082

Article

ACS Nano

The catalytic action of the Co9S8−x/CNT interlayer in lithium−sulfur batteries could also be visually demonstrated by UV−vis spectroscopy using the setup in Figure S6. The test cell consisted of a sulfur working electrode with or without an integrated Co9S8−x/CNT interlayer. The light beam was set to pass close to the sulfur cathode so as to detect polysulfides effused from the sulfur cathode. The cells were discharged potentiostatically to 1.7 V for different periods of time to obtain a time-course of the UV−vis spectra. Sulfur is expected to reduce to Li2S through the intermediate polysulfides at 1.7 V (vs Li+/Li).40 Figure 6a shows the UV−vis spectra of the solution around the sulfur cathode without an interlayer and digital photos of the cell before and after the test. Figure 6b shows the corresponding spectra for the sulfur cathode with a Co9S8−x/CNT interlayer. The peak at 425 nm could be assigned to S42−, and the peaks at lower wavelengths to the long-chain polysulfide anions.35 The monotonic increase of the intensity of the long-chain polysulfides with time is an indication of the increase in the extent of polysulfide dissolution (Figure 6a), which was also indicated by the change in the color of the electrolyte during the test. Both spectroscopic measurements and visual inspection clearly demonstrated the inhibition of polysulfide dissolution in the presence of the Co9S8−x/CNT interlayer. Figure 6c is a schematic showing how the presence of a cathode-integrated interlayer can better contain the dissolved polysulfides within the cathode area and uses the embedded catalyst to accelerate the polysulfide conversion reactions and increase the polysulfide reuse.

reutilization and polysulfide diffusion inhibition. Figure 5a shows the self-discharge profiles of Li−S batteries with and without the catalytic interlayer. All the cells were cycled at 0.3C for two cycles and charged to 2.6 V at the end of the cycling. The cell voltage during battery storage was monitored with time. The self-discharge during cell storage can be used to indicate the severity of “chemical cross-talk” between the two electrodes. In the context of the Li−S batteries this would result in reactions between the migratory polysulfides and the Li anode. The cell voltage of the Li−S battery without any interlayer would decrease to 2.2 V, while that of the Li−S battery with an integrated CNT interlayer would decrease at a slower rate to 2.22 V. By comparison, the Li−S battery with the integrated Co9S8−x/CNT interlayer did not show any selfdischarge, an indication of the successful suppression of polysulfide diffusion to the Li anode by the Co9S8−x/CNT interlayer. Figure 5b shows the cycle stability of the sulfur cathode with and without the catalytic interlayer at the 0.3C rate. The higher cell performance of the Li−S battery with a Co9S8−x/CNT catalytic interlayer is quite evident: an initial discharge capacity of 1272.9 mAh g−1, which decreased to 1052.9 mAh g−1 after 100 cycles. The Li−S cell with an integrated CNT interlayer started with a similar initial discharge capacity, but the capacity decreased quickly to 746.5 mAh g−1 after 100 cycles. The Li−S battery without any interlayer performed the least satisfactorily: a lower initial capacity of 1085.7 mAh g−1, which decreased to 621.8 mAh g−1 in the same time span. Galvanostatic discharge−charge curves were also recorded during the cycle stability tests (Figure 5c,d). A pair of redox features (at ∼1.9 V during discharge and ∼2.2 V during charge; pointed by arrows), which was more noticeable in the presence of the interlayer, could be associated with the redox reactions between Li2S2 and Li2S.37−41 As this redox feature was only accessible in the presence of the Co9S8−x/CNT interlayer, it can be regarded as a derived benefit from the catalytic function of the Co9S8−x/ CNT interlayer. They correspond well with the peaks in the cyclic voltammogram of the Co9S8−x/CNT electrode in Figure 3e and provide the direct experiment proof for polysulfide reutilization in the interlayer. Figure 5e,f show the FESEM images of the lithium anode of the cycled lithium−sulfur cells with and without the cathode-integrated Co9S8−x/CNT interlayer. The corrugated surface of the latter was caused by the passivating effects of polysulfides on Li deposition (Figure 5e). The smooth Li surface in Figure 5f can then be taken as evidence for the suppression of polysulfide diffusion by the Co9S8−x/CNT interlayer. The stability of the Co9S8−x/CNT interlayered cell for long-term cycling at 0.3C rate was also determined (Figure 5g). The discharge capacity was 648.5 mAh g−1 after 1000 cycles of continuous cycling, or a capacity fade rate of 0.049% per cycle. The good cycling stability confirmed that the Co9S8−x/CNT interlayer can function at the cell level to catalyze the conversion of polysulfides, thereby inhibiting the infamous polysulfide shuttle. The facile reactions from polysulfides to their end-products (Li2S and sulfur) and vice versa improved the reutilization of polysulfides and effectively suppressed the major capacity and efficiency loss mechanisms caused by the migratory polysulfides. In comparison with other implementations of interlayers for the lithium−sulfur batteries (Table S2), the cathode-integrated Co9S8−x/CNT interlayer is a significantly improved selection with a strong performance in almost all functional categories.

CONCLUSIONS In summary, we have developed a catalytic interlayer based on a sulfur-deficient Co9S8−x/CNT catalyst that can be integrated with the sulfur cathode of lithium−sulfur batteries. During the charge and discharge of the sulfur cathode, the catalytic interlayer effectively blocks the polysulfides from leaving the cathode area and uses the catalyst within to reutilize the soluble polysulfide reaction intermediates. The preparation of sulfur-deficient Co9S8−x/CNT is relatively simple, and the catalyst is highly effective for accelerating the polysulfide conversion reactions. It has a Co9S8−x core−shell structure with the most catalytically active phase (sulfur-deficient Co9S8) in the shell and leverages the metal-like Co9S8 core and the intimately integrated CNT for charge transfer. The catalytic action of Co9S8−x/CNT on the polysulfide conversion reactions was demonstrated in symmetric cells, lithium− catalyst cells, spectroelectrochemical cells, and Li−S coin cells. Lithium−sulfur batteries with an integrated Co9S8−x/CNT interlayer have also shown higher capacity and enhanced cyclability. This study presents not only an easily synthesized low-cost catalyst for the polysulfide conversion reactions but also a rational catalyst design approach and the best deployment method to optimize the catalyst performance in lithium−sulfur batteries. METHODS Chemicals and Materials. Ethylene glycol (EG, 99.8 wt %), cobalt(II) acetate tetrahydrate ((CH3COO)2Co·4H2O, 98 wt %), thiourea (NH2CSNH2, 99 wt %), polyvinylidene fluoride (PVDF, 99.5 wt %), lithium sulfide (Li2S, 99.98 wt %), sulfur (99.5 wt %), lithium bis(trifluoromethanesulfonyl)imide (LiTFSI, 99.95 wt %), 1,3-dioxolane (DOL, 99.8 wt %), 1,2-dimethoxyethane (DME, 99.5 wt %), N-methyl-2-pyrrolidone (NMP, 99.5 wt %), and lithium 7080

DOI: 10.1021/acsnano.9b02374 ACS Nano 2019, 13, 7073−7082

Article

ACS Nano nitrate (LiNO3, 99.99 wt %) from Sigma-Aldrich; Super-P carbon (SP, 99.5 wt %) from Timcal; and single-wall carbon nanotubes (96 wt %) from Carbon Solutions, Inc. were used as received. Materials Synthesis. The Co9S8/CNT composite precursor was prepared first by a hydrothermal method. In a typical synthesis, 0.25 mmol of (CH3COO)2Co·4H2O and 0.25 mmol of thiourea were dissolved in 30 mL of ethylene glycol. Then 0.01 g of CNT was added to the mixture, which was then sonicated for 2 h. The content was then transferred to a 50 mL Teflon-sealed autoclave and heated at 180 °C for 24 h to form the catalyst precursor Co9S8/CNT. The solid recovered after washing with water and ethanol three times was heattreated in a 10% H2/Ar mixture at different temperatures for 3 h to form sulfur-deficient Co9S8−x/CNT. Materials Characterizations. The crystal structure of catalysts was determined by X-ray diffraction (XRD) on a Bruker D8 ADVANCE (Germany) using a Cu Kα source. The catalyst morphology was examined by TEM and high-resolution TEM (HRTEM) on a JEOL 2100F microscope and by FESEM on a JEOL JSM-6700F microscope. The catalyst surface chemical state was analyzed by XPS on a Kratos AXIS Ultra DLD surface analyzer using a monochromatic Al Kα radiation source at 15 kV (1486.71 eV). UV−Vis Spectra Measurement. UV−vis spectroscopy was carried out on an ASD LabSpec 4 vis/NIR spectrometer using the test setup shown in Figure S6. The incident radiation was directed to pass through the electrolyte in the sulfur cathode proximity. The polysulfides leaving from the sulfur cathode could then be detected by their absorption peaks in the UV/vis spectra. An Al foil was first coated with a slurry of 70 wt % sulfur, 20 wt % Super P, and 10 wt % PVDF in NMP and then with a different slurry of 90 wt % Co9S8−x/ CNT and 10 wt % PVDF in NMP to form a catalyst-integrated overlayer. The Al foil was then cut into a 1 cm × 3 cm sheet to be used as the sulfur cathode of the test cell, which was ionically connected to the lithium metal anode via the 1 M LiTFSI in DOL/ DME (1:1 v/v) electrolyte. Electrochemical Measurements. Symmetric electrochemical cells were constructed as follows: 90 wt % of active material (Co9S8−x/CNT or CNT) and 10 wt % of PVDF binder were homogenized in NMP to form a slurry, which was uniformly applied to an Al foil. The foil was cut into 1 cm × 1 cm sheets. Two of the coated Al sheets, a Celgard 2400 separator, and 40 μL of an electrolyte of 1 M LiTFSI and 0.2 M Li2S6 in a 1:1 (v/v) DOL/DME mixture were assembled into a CR2025 coin cell in an Ar-filled M Braun glovebox. The lithium−catalyst cells were configured differently: the cathode was a 1 cm × 1 cm sheet of cut Al foil coated with a different slurry (90 wt % active material (Co9S8−x/CNT or CNT) and 10 wt % PVDF binder in NMP), and the anode was a lithium metal foil. The cathode after testing was disassembled from the cell, rinsed with DOL three times to remove the lithium salt on the surface, and then evacuated overnight at room temperature for ex situ analysis on the following day. Lithium−sulfur batteries were assembled by a slightly different procedure: an NMP slurry of 70 wt % sulfur, 20 wt % Super P, and 10 wt % PVDF in NMP was used to coat the Al foil to a sulfur loading of about 2 mg cm−2. For the preparation of the catalytic interlayer-integrated cathode, the sulfur cathode was coated with a slurry of 90 wt % active material (0.1 g of Co9S8−x/CNT or CNT) and 10 wt % PVDF binder (0.0111 g) in 2 mL of NMP by the doctor blade method. Figure S5a shows a typical coated cathode with a coating thickness of ∼3 μm. The cathode, together with a lithium metal foil anode, a Celgard 2400 separator, and 40 μL of the 1 M LiTFSI electrolyte in a DOL/DME mixture (1:1 v/v) with the added presence of 2 wt % LiNO3, was assembled into a CR2025 coin cell. The electrolyte to sulfur ratio in the cell was about 20 μL mg−1 (sulfur). Battery charging and discharging were regulated by a Neware battery tester. Specific capacities were normalized by the mass of sulfur only, as per the common practice. Cyclic voltammetry and electrochemical impedance measurements were carried out on an Autolab type III electrochemical workstation. All tests were performed at a room temperature of 25 °C.

ASSOCIATED CONTENT S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsnano.9b02374. Additional SEM, TEM, XRD, XPS, polysulfide adsorption test, and electrochemical performance (PDF)

AUTHOR INFORMATION Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. ORCID

Haibin Lin: 0000-0003-0088-9447 Shengliang Zhang: 0000-0001-6368-1147 Tianran Zhang: 0000-0003-2837-4971 Sheng Cao: 0000-0002-6203-9088 Qiaofeng Yao: 0000-0002-5129-9343 Guangyuan Wesley Zheng: 0000-0003-0286-5908 Jim Yang Lee: 0000-0003-1569-9718 Notes

The authors declare no competing financial interest.

ACKNOWLEDGMENTS H.L. acknowledges the National University of Singapore for his research scholarship. REFERENCES (1) Pang, Q.; Liang, X.; Kwok, C. Y.; Nazar, L. F. Advances in Lithium−Sulfur Batteries Based on Multifunctional Cathodes and Electrolytes. Nature Energy 2016, 1, 16132. (2) Li, B.-Q.; Zhang, S.-Y.; Kong, L.; Peng, H.-J.; Zhang, Q. Porphyrin Organic Framework Hollow Spheres and Their Applications in Lithium−Sulfur Batteries. Adv. Mater. 2018, 30, 1707483. (3) Zhang, L.; Chen, X.; Wan, F.; Niu, Z.; Wang, Y.; Zhang, Q.; Chen, J. Enhanced Electrochemical Kinetics and Polysulfide Traps of Indium Nitride for Highly Stable Lithium-Sulfur Batteries. ACS Nano 2018, 12, 9578−9586. (4) Hwang, J.-Y.; Kim, H. M.; Shin, S.; Sun, Y.-K. Designing a HighPerformance Lithium−Sulfur Batteries Based on Layered Double Hydroxides−Carbon Nanotubes Composite Cathode and a DualFunctional Graphene−Polypropylene−Al2O3 Separator. Adv. Funct. Mater. 2018, 28, 1704294. (5) Kong, L.; Li, B.-Q.; Peng, H.-J.; Zhang, R.; Xie, J.; Huang, J.-Q.; Zhang, Q. Porphyrin-Derived Graphene-Based Nanosheets Enabling Strong Polysulfide Chemisorption and Rapid Kinetics in Lithium− Sulfur Batteries. Adv. Energy Mater. 2018, 8, 1800849. (6) Xiao, Z.; Yang, Z.; Zhang, L.; Pan, H.; Wang, R. Sandwich-Type NbS2@S@I-Doped Graphene for High-Sulfur-Loaded, UltrahighRate, and Long-Life Lithium-Sulfur Batteries. ACS Nano 2017, 11, 8488−8498. (7) Ye, C.; Zhang, L.; Guo, C.; Li, D.; Vasileff, A.; Wang, H.; Qiao, S.-Z. A 3D Hybrid of Chemically Coupled Nickel Sulfide and Hollow Carbon Spheres for High Performance Lithium-Sulfur Batteries. Adv. Funct. Mater. 2017, 27, 1702524. (8) Huang, S.; Meng, Y.; He, S.; Goswami, A.; Wu, Q.; Li, J.; Tong, S.; Asefa, T.; Wu, M. N-, O-, and S-Tridoped Carbon-Encapsulated Co9S8 Nanomaterials: Efficient Bifunctional Electrocatalysts for Overall Water Splitting. Adv. Funct. Mater. 2017, 27, 1606585. (9) Zhou, G.; Li, L.; Wang, D. W.; Shan, X. Y.; Pei, S.; Li, F.; Cheng, H. M. A Flexible Sulfur-Graphene-Polypropylene Separator Integrated Electrode for Advanced Li-S Batteries. Adv. Mater. 2015, 27, 641. (10) Abbas, S. A.; Ding, J.; Wu, S. H.; Fang, J.; Boopathi, K. M.; Mohapatra, A.; Lee, L. W.; Wang, P. C.; Chang, C. C.; Chu, C. W. Modified Separator Performing Dual Physical/Chemical Roles to 7081

DOI: 10.1021/acsnano.9b02374 ACS Nano 2019, 13, 7073−7082

Article

ACS Nano Inhibit Polysulfide Shuttle Resulting in Ultrastable Li-S Batteries. ACS Nano 2017, 11, 12436−12445. (11) Dong, Y.; Zheng, S.; Qin, J.; Zhao, X.; Shi, H.; Wang, X.; Chen, J.; Wu, Z. S. All-Mxene-Based Integrated Electrode Constructed by Ti3C2 Nanoribbon Framework Host and Nanosheet Interlayer for High-Energy-Density Li-S Batteries. ACS Nano 2018, 12, 2381−2388. (12) Park, J.; Yu, B.-C.; Park, J. S.; Choi, J. W.; Kim, C.; Sung, Y.-E.; Goodenough, J. B. Tungsten Disulfide Catalysts Supported on a Carbon Cloth Interlayer for High Performance Li-S Battery. Adv. Energy Mater. 2017, 7, 1602567. (13) Hwang, J.-Y.; Kim, H. M.; Lee, S.-K.; Lee, J.-H.; Abouimrane, A.; Khaleel, M. A.; Belharouak, I.; Manthiram, A.; Sun, Y.-K. HighEnergy, High-Rate, Lithium-Sulfur Batteries: Synergetic Effect of Hollow TiO2-Webbed Carbon Nanotubes and a Dual Functional Carbon-Paper Interlayer. Adv. Energy Mater. 2016, 6, 1501480. (14) Lin, H.; Yang, L.; Jiang, X.; Li, G.; Zhang, T.; Yao, Q.; Zheng, G. W.; Lee, J. Y. Electrocatalysis of Polysulfide Conversion by SulfurDeficient MoS2 Nanoflakes for Lithium−Sulfur Batteries. Energy Environ. Sci. 2017, 10, 1476−1486. (15) Babu, G.; Masurkar, N.; Al Salem, H.; Arava, L. M. Transition Metal Dichalcogenide Atomic Layers for Lithium Polysulfides Electrocatalysis. J. Am. Chem. Soc. 2017, 139, 171−178. (16) Guo, W.; Chen, C.; Ye, M.; Lv, M.; Lin, C. Carbon Fiber/ Co9S8 Nanotube Arrays Hybrid Structures for Flexible Quantum DotSensitized Solar Cells. Nanoscale 2014, 6, 3656−3663. (17) Dai, C.; Lim, J.-M.; Wang, M.; Hu, L.; Chen, Y.; Chen, Z.; Chen, H.; Bao, S.-J.; Shen, B.; Li, Y.; Henkelman, G.; Xu, M. Honeycomb-Like Spherical Cathode Host Constructed from Hollow Metallic and Polar Co9S8 Tubules for Advanced Lithium−Sulfur Batteries. Adv. Funct. Mater. 2018, 28, 1704443. (18) Chen, T.; Ma, L.; Cheng, B.; Chen, R.; Hu, Y.; Zhu, G.; Wang, Y.; Liang, J.; Tie, Z.; Liu, J.; Jin, Z. Metallic and Polar Co9S8 Inlaid Carbon Hollow Nanopolyhedra as Efficient Polysulfide Mediator for Lithium−Sulfur Batteries. Nano Energy 2017, 38, 239−248. (19) Li, H.; Tsai, C.; Koh, A. L.; Cai, L.; Contryman, A. W.; Fragapane, A. H.; Zhao, J.; Han, H. S.; Manoharan, H. C.; AbildPedersen, F.; Nørskov, J. K.; Zheng, X. Activating and Optimizing MoS2 Basal Planes for Hydrogen Evolution through the Formation of Strained Sulphur Vacancies. Nat. Mater. 2016, 15, 48−53. (20) Wang, H.; Zhang, Q.; Yao, H.; Liang, Z.; Lee, H.-W.; Hsu, P.C.; Zheng, G.; Cui, Y. High Electrochemical Selectivity of Edge Versus Terrace Sites in Two-Dimensional Layered MoS2 Materials. Nano Lett. 2014, 14, 7138−7144. (21) Wang, L.; Holewinski, A.; Wang, C. Prospects of PlatinumBased Nanostructures for the Electrocatalytic Reduction of Oxygen. ACS Catal. 2018, 8, 9388−9398. (22) Nam, D. H.; Lee, S.; Lee, Y. J.; Jo, J. H.; Yoon, E.; Yi, K. W.; Lee, G. D.; Joo, Y. C. Gaseous Nanocarving-Mediated Carbon Framework with Spontaneous Metal Assembly for Structure-Tunable Metal/Carbon Nanofibers. Adv. Mater. 2017, 29, 1702958. (23) Zhang, X.; Yu, X.; Zhang, L.; Zhou, F.; Liang, Y.; Wang, R. Molybdenum Phosphide/Carbon Nanotube Hybrids as Ph-Universal Electrocatalysts for Hydrogen Evolution Reaction. Adv. Funct. Mater. 2018, 28, 1706523. (24) Dou, S.; Tao, L.; Huo, J.; Wang, S.; Dai, L. Etched and Doped Co9S8/Graphene Hybrid for Oxygen Electrocatalysis. Energy Environ. Sci. 2016, 9, 1320−1326. (25) Kim, H.-S.; Cook, J. B.; Lin, H.; Ko, J. S.; Tolbert, S. H.; Ozolins, V.; Dunn, B. Oxygen Vacancies Enhance Pseudocapacitive Charge Storage Properties of MoO3−X. Nat. Mater. 2017, 16, 454− 460. (26) Liu, X.; Liu, H.; Zhao, Y.; Dong, Y.; Fan, Q.; Kuang, Q. Synthesis of the Carbon-Coated Nanoparticle Co9S8 and Its Electrochemical Performance as an Anode Material for Sodium-Ion Batteries. Langmuir 2016, 32, 12593−12602. (27) Sun, S.; Luo, J.; Qian, Y.; Jin, Y.; Liu, Y.; Qiu, Y.; Li, X.; Fang, C.; Han, J.; Huang, Y. Metal−Organic Framework Derived Honeycomb Co9S8@C Composites for High-Performance Supercapacitors. Adv. Energy Mater. 2018, 8, 1801080.

(28) Zhuang, T.-T.; Liang, Z.-Q.; Seifitokaldani, A.; Li, Y.; De Luna, P.; Burdyny, T.; Che, F.; Meng, F.; Min, Y.; Quintero-Bermudez, R.; Dinh, C. T.; Pang, Y.; Zhong, M.; Zhang, B.; Li, J.; Chen, P.-N.; Zheng, X.-L.; Liang, H.; Ge, W.-N.; Ye, B.-J.; et al. Steering Post-C−C Coupling Selectivity Enables High Efficiency Electroreduction of Carbon Dioxide to Multi-Carbon Alcohols. Nat. Catal. 2018, 1, 421− 428. (29) Peng, H.-J.; Zhang, Z.-W.; Huang, J.-Q.; Zhang, G.; Xie, J.; Xu, W.-T.; Shi, J.-L.; Chen, X.; Cheng, X.-B.; Zhang, Q. A Cooperative Interface for Highly Efficient Lithium−Sulfur Batteries. Adv. Mater. 2016, 28, 9551−9558. (30) Zhou, T.; Lv, W.; Li, J.; Zhou, G.; Zhao, Y.; Fan, S.; Liu, B.; Li, B.; Kang, F.; Yang, Q.-H. Twinborn TiO2−Tin Heterostructures Enabling Smooth Trapping−Diffusion−Conversion of Polysulfides Towards Ultralong Life Lithium−Sulfur Batteries. Energy Environ. Sci. 2017, 10, 1694−1703. (31) Fan, F. Y.; Carter, W. C.; Chiang, Y.-M. Mechanism and Kinetics of Li2S Precipitation in Lithium−Sulfur Batteries. Adv. Mater. 2015, 27, 5203−5209. (32) Liu, J.; Qian, T.; Wang, M.; Zhou, J.; Xu, N.; Yan, C. Use of Tween Polymer to Enhance the Compatibility of the Li/Electrolyte Interface for the High-Performance and High-Safety Quasi-Solid-State Lithium-Sulfur Battery. Nano Lett. 2018, 18, 4598−4605. (33) Cheng, F.; Zhang, T.; Zhang, Y.; Du, J.; Han, X.; Chen, J. Enhancing Electrocatalytic Oxygen Reduction on MnO2 with Vacancies. Angew. Chem., Int. Ed. 2013, 52, 2474−2477. (34) Wang, Q.; Zheng, J.; Walter, E.; Pan, H.; Lv, D.; Zuo, P.; Chen, H.; Deng, Z. D.; Liaw, B. Y.; Yu, X.; Yang, X.; Zhang, J.-G.; Liu, J.; Xiao, J. Direct Observation of Sulfur Radicals as Reaction Media in Lithium Sulfur Batteries. J. Electrochem. Soc. 2015, 162, A474−A478. (35) Cañas, N. A.; Fronczek, D. N.; Wagner, N.; Latz, A.; Friedrich, K. A. Experimental and Theoretical Analysis of Products and Reaction Intermediates of Lithium−Sulfur Batteries. J. Phys. Chem. C 2014, 118, 12106−12114. (36) Faber, M. S.; Park, K.; Caban-Acevedo, M.; Santra, P. K.; Jin, S. Earth-Abundant Cobalt Pyrite (CoS2) Thin Film on Glass as a Robust, High-Performance Counter Electrode for Quantum DotSensitized Solar Cells. J. Phys. Chem. Lett. 2013, 4, 1843−1849. (37) Moy, D.; Manivannan, A.; Narayanan, S. R. Direct Measurement of Polysulfide Shuttle Current: A Window into Understanding the Performance of Lithium-Sulfur Cells. J. Electrochem. Soc. 2015, 162, A1−A7. (38) Li, G.; Wang, S.; Zhang, Y.; Li, M.; Chen, Z.; Lu, J. Revisiting the Role of Polysulfides in Lithium-Sulfur Batteries. Adv. Mater. 2018, 30, No. e1705590. (39) Li, Q.; Song, Y.; Xu, R.; Zhang, L.; Gao, J.; Xia, Z.; Tian, Z.; Wei, N.; Rümmeli, M. H.; Zou, X.; Sun, J.; Liu, Z. Biotemplating Growth of Nepenthes-Like N-Doped Graphene as a Bifunctional Polysulfide Scavenger for Li−S Batteries. ACS Nano 2018, 12, 10240−10250. (40) Song, Y.; Zhao, W.; Kong, L.; Zhang, L.; Zhu, X.; Shao, Y.; Ding, F.; Zhang, Q.; Sun, J.; Liu, Z. Synchronous Immobilization and Conversion of Polysulfides on a VO2−VN Binary Host Targeting High Sulfur Load Li−S Batteries. Energy Environ. Sci. 2018, 11, 2620− 2630. (41) Song, Y.; Zhao, W.; Wei, N.; Zhang, L.; Ding, F.; Liu, Z.; Sun, J. In-Situ Pecvd-Enabled Graphene-V2O3 Hybrid Host for Lithium− Sulfur Batteries. Nano Energy 2018, 53, 432−439.

7082

DOI: 10.1021/acsnano.9b02374 ACS Nano 2019, 13, 7073−7082