Mesoporous TiO2 coating on carbon–sulfur cathode for high capacity Li–sulfur battery

In this paper, a meso-porous TiO2 (titania) coating is shown to effectively protect a carbon–sulfur composite cathode from polysulfide dissolution. The cathode consisted of a sulfur impregnated carbon support coated with a few microns thick mesoporous titania layer. The carbon–sulfur cathode is made using activated carbon powder (ACP) derived from biomass. The mesoporous titania coated carbon–sulfur cathodes exhibit a retention capacity after 100 cycles at C/3 rate (433 mA g −1) and stabilized at a capacity around 980 mA h g−1. The electrochemical impedance spectroscopy (EIS) of the sulfur cathodes suggests that the charge transfer resistance at the anode, (Ract) is stable for the titania coated sulfur electrode in comparison to a continuous increase in Ract for the uncoated electrode implying mitigation of polysulfide shuttling for the protected cathode. Stability in the cyclic voltammetry (CV) data for the first 5 cycles further confirms the polysulfide containment in the titania coated cathode while the uncoated sulfur electrode shows significant irreversibility in the CV with considerable shifting of the voltage peak positions. Raman spectroscopy and X-ray photoelectron spectroscopy (XPS) studies confirm the adsorption of soluble polysulfides by mesoporous titania.


Introduction
Lithium-air and lithium-sulfur batteries have attracted worldwide attention due to their potential for achieving higher energy density than lithium ion battery technology. 1 Sulfur as a cathode has an excellent theoretical gravimetric discharge capacity of 1672 mA h g À1 and Li-air has a theoretical discharge capacity of $3623 W h kg À1 corresponding to Li 2 O 2 (3840 mA h g À1 excluding O 2 mass). 1,2 Li-air batteries require pure O 2 at the cathode due to side reactions caused by the moisture and CO 2 present in air. These can lead to a range of issues such as cathode instabilities, catalyst driven side reactions, electrolyte instabilities and the reactivity of Li 2 O 2 and its intermediates. The majority of recent Li-O 2 reports have only assessed battery performance at limited depths of discharge using pure O 2 and impractical low mass loadings, and this makes it difficult to gauge their prospects in practical applications. 3 On the other hand, Li-S batteries seem closer to industrial readiness if poor cyclability is addressed through proper sulfur cathode formulations. 4 The poor cyclability of Li-S batteries is attributed to two main factors: poor electrical conductivity of sulfur and dissolution of polysuldes into the electrolyte during lithiation and delithiation. The lithiation of sulfur is facilitated by the formation of polysulde intermediates. Electrochemically, elemental sulfur rst reduces to S 8 2À and forms Li 2 S 8 in liquid form. Li 2 S 6 and Li 2 S 4 are formed thereaer. 5,6 As a result of polysulde dissolution, the phenomena known as polysulde shuttle will occur causing active material inaccessible for further electrochemical reactions. 7,8 Typically, polysuldes Li 2 S n (2 < n < 8) are known to dissolve in organic electrolytes. Poly-sulde shuttle phenomenon has been studied extensively. 7 Among organic solvents, 1,2-dimethoxyethane (DME) and 1,3dioxolane (DOL) based organic solvents are preferable due to their bulky anions which can effectively reduce polysulde solubility. 9 According to Barchasz et al. 10 dissolution of poly-sulde is also a necessity for the proper operation of the sulfur electrode. It is also claimed that dissolution of polysulde increases the viscosity of the electrolyte and hence less viscous solvents are preferable for Li-S batteries. There are three ways to improve the cyclability and durability of sulfur cathode: rst, the sulfur particles need to be encapsulated to minimize the leaking of polysuldes, second, the electrode material must be properly wetted by electrolyte, and lastly a good electronic conductivity must be maintained within the bulk electrode. In addition, there are some reports that indicate gas evolution in Li-S cells could also be problematic for their practical applications. 11 In the last few years, many techniques have been attempted to prevent dissolution of polysuldes from sulfur cathodes into the electrolyte. An excellent review of these techniques can be found in ref. 12. Use of meso/micro pore carbon structures has been attempted to trap soluble polysuldes. A method using micro pore structure was used to limit the chain length of resulting polysuldes during lithiation and avoid long chain soluble polysuldes. 13 The use of microporous carbon material has been found to yield high initial discharge capacity but resulted in continuous decay. 1 In another concept, a meso/ micro ordered carbon architecture-sulfur composite is synthesized by rst creating the carbon structure using a silicon template followed with diffusion of sulfur. 14 The advantage of such meso-porosity for a sulfur cathode is that electrolyte can channel into the micro-pore sites. In another technique, sulfur nano-particles were encapsulated by carbonized polymer coating. 15 Such encapsulated sulfur particles tend to expand when polysuldes are formed and are prone to crack with cycling. In order to avoid such cracking, yolk shell type coatings were studied. 16 Sulfur encapsulation using graphene has also been investigated. 17,18 Similarly, physical trapping of poly-suldes has been investigated using carbon nanotube mat electrodes. 19,20 This allowed the use of a ber matrix both as an active material for storage and as a barrier for leaking poly-suldes. In all the techniques prior to this, cyclability of sulfur electrodes had not been improved considerably, since the physical trapping mechanism did not support long-term encapsulation for dissolved polysuldes. Use of a ber matrix has improved both electronic conductivity and electrolyte channeling. Nanocomposites of sulfur chemically bonded with carbon were also investigated. 21 The chemically bonded sulfur however does not seem to participate in lithiation reaction while unbonded sulfur is found to be electrochemically active. Soluble polysuldes tapping by chemisorption of aminefunctionalized carbon has been reported in 22 with a signicant capacity retention. In addition to the above modications of cathode electrodes, the use of solid state electrolytes and gel polymer electrolytes are being investigated to alleviate the sulfur dissolution problem. 23,24 The development and study of solid or gel electrolytes with high lithium diffusion is a research topic of interest in itself. A good comprehensive review of electrolytes for lithium sulfur batteries can be found in. 25 There have been some reports on other liquid electrolytes to improve the cyclability of Li-S batteries. 26,27 Interestingly, the dissolution of polysuldes near the electrode-electrolyte interface is seen as a necessary step for complete lithiation of sulfur, i.e., 2Li + S / Li 2 S. Use of a polymer electrolyte would therefore hinder the intermediate polysulde formation, lowering the discharge capacity of sulfur. In ref. 28, discharge capacity for the all solid state Li-S battery has been demonstrated as 200 mA h g À1 at the 50 th cycle which is a considerably lower gravimetric capacity compared to organic electrolyte based Li-S batteries. Due to these reasons, numerous studies were focused on using liquid electrolytes, but trapping lithium polysuldes in porous carbon electrodes such as carbon nanotubes. [29][30][31][32][33] In our work, a mesoporous titania layer is utilized to trap the dissolved polysuldes along with an unique electrical bridging technique to improve the electrical conductivity within the bulk electrode. The titania particle coating is seen to improve the durability and high capacity retention of the sulfur cathode in two ways: (a) the meso-porosity will enable bulk diffusion of lithium through the pores; and (b) high surface area of titania will promote adsorption of polysuldes. As shown in ref. 34, S 2À ions can be adsorbed on titania surfaces. The role of the mesoporous titania coating of the sulfur cathode is thus investigated in order to understand the underlying mechanisms of sulde ion dissolution and its effect on capacity retention and durability. Sulfur cathodes comprised of titania/carbon have already been investigated by several other groups [35][36][37] for Li-S batteries. However, in this work, we have attempted a new approach of fabricating the sulfur electrode by simply coating the sulfur impregnated carbon matrix with meso-porous titania rather than mixing the carbon particles with titania.

Experimental method
The electrode material is used for coating titania is activated carbon (ACP) derived from bio mass. Sulfur is thermally diffused into the electrode. The synthesis of activated carbon from bio mass is described in the ESI. † The electrode is prepared by ball milling ACP with 10 ml of 60% PVDF (polyvinylidene uoride) in NMP (N-methyl-2-pyrrolidone) solvent for 12 hours. Then the ACP slurry is poured onto a clean glass surface to form a free-standing ACP sheet.
In the second step, sulfur (3-4 mg) is melted on a hotplate at 130 C and is impregnated into ACP free standing carbon structures by pressing them onto the melted sulfur. In the third step, ACP electrodes are coated with 200 nm titania paste by dipping the electrodes in titania suspension in ethanol, followed by air drying for 24 hours. A part of the back surface of the titania coating is scratched-off to expose the interior of the carbon/sulfur electrode (bridging) in order to make better electrical contact with the current collector. Then titania coated sulfur electrodes are pressed against a carbon black pellet forming the electrical bridge. The carbon black pellet is made by mixing 20 mg of acetylene carbon black and 20 ml of PTFE (polytetrauoroethylene). Next, it is placed on a stainless-steel mesh with a diameter of 15 mm and pressed under $300 kg of pressure (using a hydraulic press) to mount the entire assembly on the current collector. The thickness of the carbon black pellet is reduced to about 0.1-0.2 mm aer pressing. The cathode is then assembled in a CR2032 coin cell with pure lithium metal as the anode, inside an argon-lled glove box. Celgard 3401 polymer separator ($75 mm thick) is placed between the electrodes. The composition of the electrolyte used in this work is 1 : 1 ratio of 1,2-dimethoxyethane (DME Sigma Aldrich) and 1,3-dioxolane (DOL Sigma Aldrich) in 1 M bis(tri-uoromethane)sulfonimide lithium salt (LiTFSI) and 1% wt of LiNO 3 for a total of 0.5 ml of electrolyte. The ionic conductivity of the electrolyte is $14.7 mS cm À1 at 25 C. LiNO 3 is widely used as an additive in the electrolyte to form a protective lm on the lithium anode. Fig. 1 represents the side view of the cell including the SEM image of the titania coating, and the pore width distribution for anatase titania powder measured by BET technique. The mean pore width is found to be 40 nm which is mesoporous. The size of titania particles used in this experiment is around 200 nm. When the electrode is fabricated, we found that the particle coating maintains the meso-porosity.

Electrochemical and structural analysis
The cells were cycled between 1.5 and 2.8 V versus Li/Li + in galvanostatic mode using 16 channel Arbin battery test system. Cyclic voltammetry (CV) was performed at a scan rate of 0.3 V in the range of 1.5 to 2.8 V using the biologic sp-200 electrochemical system. AC impedance (EIS) of the cell was measured using the same electrochemical system over the 1 mHz to 1 MHz range. Both CV and EIS measurements were conducted by a swagelokbased three-electrode conguration with lithium as both the counter electrode and the reference electrode. All performances were carried out at 25 C.

Characterization
The ionic conductivity was measured by a biologic sp-200 system. The electrode surface morphology before and aer cycling was characterized by a TESCAN thermionic emission scanning electron microscope. X-ray photoelectron spectroscopy (VG scientic-MultiLab 3000) was employed to detect the chemical composition of the cathode. All XPS spectra were tted with Gaussian-Lorentzian functions and a shirley and linear type background. 2P 3/2 2P 1/2 peaks were tted using Lorentzian function. The binding energy values were all calibrated using carbon 1S 284.5 eV. Samples for SEM and XPS characterization were prepared by disassembling cells and rinsing with 1,2dimethoxyethane, 1,3-dioxolane. TGA studies were done by thermogravimetric analyzer TA 2050.

Results and discussion
Performance of the sulfur cathode is tested against lithium metal as the anode in a coin cell conguration over the voltage range of 2.8-1.5 V using an Arbin battery tester. The electrochemical performances of an uncoated sulfur cathode and a mesoporous titania coated cathode at C/3 rate are shown in Fig. 2. The areal sulfur loading is 2.65 mg cm À2 and mass of titania coating was approximately 1.5 mg cm À2 . In all cases of Fig. 2b, discharge curves showed two discharge plateaus at 2.4 and 2.0 V. The sudden drop of voltage in Fig. 2a from 2.6 V to 2.4 V is due to the polarization and IR drop of electrodes and electrolyte. The plateau at 2.4 V is believed to be due to the reduction of S 8 to high-order soluble lithium polysuldes (e.g. Li 2 S 4 ), and the plateau at 2.0 V is due to further reduction of Li 2 S 4 into insoluble Li 2 S. The uncoated sulfur electrode shows rapid decay of gravimetric discharge capacity within the rst 20 cycles (Fig. 2b). There, titania coated electrode shown in purple color however, exhibited stable discharge capacity in excess of 900 mA h g À1 even aer the 100 th cycle.
The inuence of proper electrical connection to the conductive matrix of the electrode is investigated by comparing three different activated carbon electrode systems: (i) uncoated, (ii) titania coated with poor electrical connectivity, and (iii) titania coated with improved electrical connectivity (electrical bridging) as shown in Fig. 2. Experiments were carried out at C/ 3 discharge and charge rate. Synthesis of activated carbon is described in the ESI. † A properly working sulfur electrode has two voltage plateaus at $2.4 V (formation of Li 2 S x polysuldes) and $2.0 V (formation of Li 2 S and Li 2 S 2 ). This is an indication that Li + transport has not been mitigated by the titania particle barrier. A stable discharge capacity of about 980 mA h g À1 for 100 cycles has been achieved for the titania coated ACP supported sulfur electrode with improved electrical conductivity. In contrast, the titania coated ACP supported sulfur electrode with poor electrical connectivity shows lower discharge capacity of Fig. 2 Gravimetric capacity results of ACP based sulfur electrodes; (a) potential vs. specific capacity curves of optimized ACP based titania coated sulfur cathode, (b) comparison of the discharge capacity of various ACP based cathodes; (i) titania coated sulfur cathode with (purple) and without (blue) improved electrical connection to the current collector and (ii) uncoated electrodes (red); left axis represent the discharge capacity and the right axis represents the coulombic efficiency for the optimized sulfur cathode (c) rate capability performances of ACP based titania coated sulfur electrode with improved electrical connection.
$700 mA h g À1 aer 100 cycles while the ACP supported sulfur electrode without a coating layer shows discharge capacity of only 265 mA h g À1 at 100 th cycle. The idea of coating the back side of the electrode with mesoporous titania is to prevent any leak of soluble polysuldes into the electrolytes when the battery is at idle between cycles.
Polysulde trapping by using metal oxide such as titania has been investigated in three different methods by other groups: in the rst method, sulfur cathodes were made by simply mixing titania particles with sulfur/carbon composites. [38][39][40] In the second method, sulfur was rst coated with titania followed by carbonization. 35,[41][42][43] In the third method, titania nanoparticles have been coated on the polymer separator, forming an effective polysulde adsorbing barrier. 44 In all three methods, electrical conductivity between current collector and active material is established only through the carbon matrix in the composite. However, in this work the electrical conductivity between current collector and active material is established through an electrical bridging technique (Fig. 3c). Its effect is further analyzed by 2-probe impedance tests as shown in Fig. 3. The electrical contacts were made to the current collector and the titania coating layer on the other side. It is found that the dc resistance for the electrically bridged cathode is 127.88 U in comparison to the dc resistance value of 1283.63 U for the electrode with titania coating on both sides. This is a signicant improvement in the net electrical resistance due to the electrical bridging leading to a high discharge capacity as seen in Fig. 2b. A complete impedance analysis is presented in Fig. S5 and Table  S1, † according to the equivalent circuits proposed in Fig. 3.
In order to investigate high-power performance of the titania coated sulfur electrode with proper electrical contact, rate capability was studied in the voltage range of 2.8-1.5 V with different current densities as shown in Fig. 2c. Five initial formation cycles have been shown at 130 mA g À1 current density followed by 5 cycles each at 260 mA g À1 , 650 mA g À1 , and 1.3 A g À1 current densities. It shows that discharge capacities at 130 mA g À1 , 260 mA g À1 , 650 mA g À1 , and 1.3 A g À1 are approximately 1000, 800, 700, and 450 mA h g À1 , respectively. When the current density is reduced back to 130 mA g À1 aer the rate performance testing, the sulfur cathode can retain the discharge capacity close to the formerly measured value of 900 mA h g À1 , indicating its good reversibility and high rate capability and demonstrating the recovery of the titania coated sulfur cathode aer subjecting it to different charge-discharge rates.
Next, cyclic voltammetry (CV) was carried out for the ACP based sulfur impregnated electrodes without and with titania coating as shown in Fig. 4 within 2.8 V and 1.5 V range at 0.3 mV s À1 rate. The lower end potential is chosen to be 1.5 V since LiNO 3 additive tends to be reduced irreversibly at the voltages below 1.5 V. 45 CV measurements are carried out for up to 5 cycles and both electrodes showed the complete two step redox reactions with two reduction peaks appearing at around 2.3 and 2.0 V and one oxidation peak at $2.4 V. The peak at $2.3 V is ascribed to the reduction of sulfur to form the higher order lithium polysuldes (Li 2 S n , n > 4), and the peak at $2.0 V corresponds to further reduction of these lithium polysuldes to lower order lithium polysuldes (Li 2 S n , n < 4) including Li 2 S 2 and Li 2 S. The oxidation peak at $2.4 V can be attributed to the oxidation of lithium polysuldes (Li 2 S n , n < 4) back to higher order lithium polysuldes (Li 2 S n , n > 4). Theoretically, two distinct oxidation peaks are expected for the sulfur cathode. However, in our case, the two oxidation peaks appear to merge into a single composite peak. We believe that, the resolution of the oxidation peaks in Li-S battery depend on the charge transfer resistance in the sulfur electrode. Sulfur cathode with better charge transfer properties, will allow all the polysulde species to oxidize in parallel reactions, result in one convoluted peak. The charge transfer resistance depends on the electronic conductivity, porosity, and the surface area of the conductive material of the sulfur cathode. The work reported in ref. 19 also suggests a similar nding that the oxidation peaks in the cyclic voltammetry curves become more convoluted and less resolved with increased surface area and better charge transfer properties. The CV curve shown in Fig. 4b and discharge curves of Fig. 2b for a titania coated sulfur cathode shows remarkable durability over the cycles. This is an indication of the reformation of sulfur within the bulk electrode and minimal leakage of polysuldes into the electrolyte. In contrast, the uncoated sulfur electrode shows signicant irreversibility in the CV diagram with shiing of peak positions and changes in current levels implying dissolution of polysuldes into the electrolyte. In this work, titania is expected to trap the polysuldes and the CV curves should not show a considerable current at the 2.8 V vertex as in Fig. 4b. This reasoning is conrmed by comparing the CV cycles of the uncoated sulfur electrode shown in Fig. 4a. It is noticeable that at 2.8 V vertex of Fig. 4a, there is a cathodic current $2 mA, implying existence of dissolved polysuldes in the electrolyte still undergoing oxidation.
Raman and X-ray photo electron (XPS) analysis were carried out to further conrm the trapping of polysuldes in titania layer. In Raman spectrum analysis, we investigated titania coated electrode before and aer discharge as shown in Fig. 5a. Both spectra show 3 clear peaks characteristic of crystalline anatase titania. An additional weak peak appearing at $742 cm À1 for the discharged electrode can be interpreted as due to the polysulde links (S x 2À , x ¼ 4-8). 46 Fig. 5b shows the sulfur 2P peak (S 2P ) with 2 distinct peaks at 160.4 and 161.9 eV corresponding to bridging sulfur and terminating sulfur respectively. 40 This is possible due to the efficient trapping of higher order soluble polysuldes in the mesoporous TiO 2 layer. The electrode which used in this analysis are washed with the 1 : 1 ratio of 1,2-dimethoxyethane (DME Sigma Aldrich) and 1,3-dioxolane (DOL Sigma Aldrich) to remove any dissolved polysulde from the surface of titania coating which might not have adhered to the titania surface. Thus, it is reasonable to conclude that the polysuldes detected by Raman and XPS are from the polysuldes which were adhered on titania particles. Fig. 5c compares XRD spectra for titania coated sulfur cathode before and aer discharge. XRD spectra for pristine titania and Li 2 S are also shown for comparison. It shows clear evidence of the presence of solid Li 2 S aer the rst cycle discharge.
In the uncoated sulfur electrode, soluble polysuldes are expected to dissolve into the electrolyte. In the case of meso-porous titania coating, the dissolved polysulde ions adsorb on titania surfaces and never reach bulk electrolyte beyond titania barrier. In the schematic diagram in Fig. 6, the processes of a Li-S battery with uncoated and titania coated cathodes have been categorized into several regions where key reactions take place.
EIS was used to determine impedances within coated and uncoated sulfur cathodes during cycling. Based on the Nyquist plots ( Fig. S4a and b in the ESI †) for coated and uncoated sulfur cathodes, an equivalent circuit has been proposed as presented in Fig. 6. Here, R e represents the electrolyte resistance as a single series resistance in the network. The loops in the Nyquist plot consisting of superposition of multiple semicircles are each represented by a combination of a resistance and a constant phase element (CPE) in parallel. A similar equivalent circuit modeling and electrochemical impedance analysis can be found in the ESI of the work reported in ref. 47.
The choice of a CPE instead of a capacitor is due to the nonideal behavior of the electrodes. Each semicircle represents (i) charge transfer at the cathode (R cct kCPE cct ) (ii) charge transfer at the anode (R act kCPE act ), and (iii) contact interphase at the Fig. 5 (a) Raman spectra of titania coated sulfur electrodes (ACP based) before and after discharge (b) XPS surface analysis for titania coating (c) XRD spectra of titania coated sulfur electrodes before and after discharge, spectra for Li 2 S and TiO 2 powders are also shown for comparison. The * represents the signature of the polymer bag and the dotted vertical lines represent the aluminum substrate.
cathode (R int kCPE int ) which is present in the bulk of the cathode representing the charge conduction between the cathode current collector and the redox sites in the cathode. The variation of R int in the case of titania coating is expected to be signicant. The contribution of the anode impedance is neglected because the anode impedance in an electrolyte with polysuldes is small. Fig. 7 shows the relevant impedance parameters extracted by tting the EIS data with the proposed equivalent circuit during (i) discharging, (ii) charging, and (iii) cycling processes.
The Nyquist plots of the impedance data for uncoated and titania coated sulfur electrode during the discharge are shown in Fig. S4a and b in the ESI. † Only a few selected sets of data are shown for clarity and a typical tting procedure is shown Fig. S4c † for a selected data set. The discharge curve has been categorized into three zones according to the key actions taking place in the cell. In zone 1, both electrodes are polarized, and solid sulfur starts to dissolve in the electrolyte. In zone 2, longer polysulde chains are shortened via further reduction (in the presence or absence of titania). In zone 3, solid Li 2 S and Li 2 S 2 are formed. These solid products are more ionic in nature. In Fig. 7a and b, tting parameters corresponding to charge transfer resistance (R cct ) and interphase resistance (R int ) at cathode for titania coated and uncoated sulfur electrode respectively during discharge are shown at various depth of discharge (DOD).
The behavior of the charge transfer resistance, R cct at the cathode is similar in both cases. They both show initial decrease of R cct reaching a minimum $40% DOD followed by a slow increase. The initial decrease of R cct can be interpreted as due to the improved electrochemical accessibility of solid sulfur (insulating) to undergo polysulde formation. The following increase in R cct is due to the formation of insulating and insoluble Li 2 S and Li 2 S 2 . For both coated and uncoated sulfur electrodes this charge transfer process is similar. However, R int shows distinctly different behaviors for coated and uncoated cathodes. In the case of titania coated cathodes, R int value is seen to increase in zone 1, presumably due to the adsorption of dissolved polysuldes at the defect sites of titania. It is interesting to observe that the interphase resistance drops in zone 2, where longer polysudes are reduced to shorter polysuldes.
During the charging of the cell, solid Li 2 S and Li 2 S 2 should eventually oxidize back to elemental sulfur through intermediate polysulde formation. The analysis of the variation of the R int during the charging process provides useful information Paper about the underlying mechanism of the titania coated electrode as shown in Fig. 7c. Once Li 2 S starts to oxidize to intermediate polysulde chains, the interphase resistance, R int is expected to decrease as the conductivity improves for the titania coated electrode (Fig. 7c). In contrast, the R int of the uncoated sulfur electrode is seen to increase as charging progresses (Fig. 7d). It is reasonable to assume that this conversion (Li 2 S/Li 2 S 2 to intermediate polysuldes) may take place at the electrode matrix-electrolyte interface since there are considerable amounts of dissolved polysuldes remaining in the electrolyte. As the sulfur growth takes place on the surface, the interphase resistance, R int continues to increase (Fig. 7d). It is also noted that, the effect on electrolyte resistance due to the dissolved polysulde is considerably small (Fig. 7e) in the case of titania coated cathode.
Dissolution of polysuldes increases the viscosity of the electrolyte causing an increase in electrolyte resistance (R e ) noticeable in Fig. 7e and f in different magnitudes. Titania coated sulfur electrodes show a stabilized R e , however, due to the adsorption activity of titania layer. It is evident from the relative magnitudes of the changes in R e that coating of the sulfur electrode with titania has signicantly limited the poly-sulde dissolution into the electrolyte. Finally, variation of, R act during cycling is presented in Fig. 7e and f for both coated and uncoated cathodes as evidence for polysulde shuttling and Li 2 S and Li 2 S 2 formation on the anode surface. Li 2 S and Li 2 S 2 are known to be formed on the anode by reducing the dissolved polysuldes (from cathode) in the electrolyte aer shuttling to anode. For the uncoated sulfur cathode, R act increases almost linearly until 20 th cycle and then shows an abrupt rise con-rming the continuous formation of Li 2 S/Li 2 S 2 on the anode. On the contrary, the titania coated cathode shows saturation of R act aer the 20 th cycle implying limited formation of Li 2 S/Li 2 S 2 as a result of encapsulation of soluble polysulde within the titania coating.

Conclusion
Titania coating of the sulfur electrode with proper electrical contact with the current collector has proven to be effective to enhance the cyclability of Li-S batteries by retaining a stable capacity of 980 mA h g À1 discharge prole over 100 cycles. The performance of mesoporous titania coated sulfur was compared with that of uncoated sulfur electrodes using EIS and CV techniques. The mechanism of trapping dissolved polysulde within the titania layer was veried by investigating in situ impedance measurements. R act of the cell with titania coated sulfur electrode was stabilized at 20 U while R act for uncoated Fig. 7 Fitting parameters of EIS data for ACP based sulfur cathode to an equivalent electrical circuit model: each plot contains charge transfer resistance at cathode (R cct ), charge transfer resistance at anode (R act ), electrolyte resistance (R e ), and interface resistance (R int ). Plots (a) and (b) represent results for titania coated and uncoated samples respectively against DOD; plots (c) and (d) represent results for coated and uncoated samples respectively against DOC. Plots (e) and (f) represent results for titania coated and uncoated samples respectively against cycle number.
sulfur electrode continued to rise beyond 20 U during charging and discharging. Such increase in charge transfer resistance at the anode in uncoated sulfur cathode is due to deposition of solid Li 2 S on lithium metal anode. The electrical bridging technique to improve the electrical conductance between the interior of the sulfur/carbon composite and the current collector is proven to contribute signicantly for the superior performance of titania coated sulfur electrodes. Otherwise, the role of titania to improve the cyclability of sulfur electrode with high discharge capacity will be undermined due to the poor electrical conductance between the interior of the electrode and the current collector. In addition, Raman and XPS analysis conrm the effective polysulde trapping by the mesoporous titania coating even though the isolation of different polysulde species was difficult. Finally, the XRD analysis concludes nonexistence of any phase changes in titania conrming that the polysulde is trapped only by adsorption onto titania.

Conflicts of interest
The authors declare no competing nancial interest.