Preparation of Bacterial Cellulose/Ketjen Black-TiO2 Composite Separator and Its Application in Lithium-Sulfur Batteries

Lithium-sulfur batteries (LSBs) have attracted extensive attention due to their high energy density and low cost. The separator is a key component of LSBs. An excellent LSBs separator requires not only good electrolyte wettability, but also high thermal stability, good tensile mechanical properties, green environmental protection potential and enough inhibition of shuttle effect. In this paper, composite separator Bacterial cellulose/Ketjen black-TiO2 (BKT) was prepared by coating the green and environmentally friendly bacterial cellulose (BC) substrate with KB-TiO2 material. BKT not only demonstrates higher electrolyte wettability, but also displays thermal stability and tensile resistance to enhance the safety of the battery. The high ratio of TiO2 and KB on the BKT surface provides chemical and physical adsorption of lithium polysulfides (LiPSs), thereby inhibiting the shuttle effect and increasing the cycle life of LSBs. The secondary current collector formed by TiO2 and KB can also reactivate the adsorbed LiPSs, further improving the capacity retention rate of the battery. Therefore, the LSBs assembled with the BKT separator exhibited an initial discharge capacity of 1180 mAh × g−1 at a high current density of 0.5 C, and maintained a specific discharge capacity of 653 mAh × g−1 after 100 cycles was achieved. Even at 2.0 mg × cm−2 sulfur areal density and 0.1 C current density, the BKT separator based battery still has an initial discharge specific capacity of 1274 mAh × g−1. In conclusion, BKT is a promising lithium-sulfur battery separator material. sulfur areal densities.


Introduction
The growing demand for energy storage forces the development of more advanced batteries. LSBs with higher specific capacity, energy density and lower cost are ideal alternatives to lithium ion batteries (LIB) [1,2]. However, some problems need to be solved before commercialization, i.e., the low conductivity of S and discharge products, volume expansion caused by density changer from S to Li 2 S during the redox reaction and shuttle effect [3][4][5]. Among the above shuttle effect, shuttle effect is an urgent problem. In the shuttle effect, long chain LiPSs are produced in the initial discharge process and then dissolved in the electrolyte and diffuse to the anode side through the separator to form short chain LiPSs. Moreover, the short chain LiPSs will pass through the separator and return to the cathode under the electric field [6], causing an irreversible loss of battery active materials and rapid decline of battery life [7,8]. Therefore, inhibiting the shuttle effect in LSBs and correspondingly reducing the loss of sulfur active materials is an urgent need for advancing the usage of LSBs.
The separator is one of the core components of LSBs. LiPSs generated in the reaction must pass through the separator to reach the anode [9]. Modifying the separator is a common strategy to suppress the shuttle effect [10]. As a mature commercial LIB separator, the PP separator is usually used in LSBs. Its pore size is about 200 nm and the LiPSs produced by the reaction of LSBs are about 2 nm. It is obvious that the PP separator cannot into the KB, the as prepared S/KB mixture kept in a reaction kettle was put in an electric heating blast drying oven at 155 • C for 12 h. The treated powder is further ground with mortar and is called S/C composite. First, mix S/C, super li and PVDF uniformly at a mass ratio of 7:2:1, then add an appropriate amount of NMP to grind to form slurry. The prepared slurry was coated on aluminum foil with a scraper. The modified aluminum foil was heated at 60 • C for 12 h in a vacuum drying oven to volatilize NMP and cut into a Φ 12 mm positive electrode sheet with a microtome. Three kinds of positive plates with different sulfur loading were prepared and the surface density of sulfur was 1 mg × cm −2 , 1.5 mg × cm −2 and 2 mg × cm −2 , respectively.

Preparation of KB-TiO 2
A solvent mixture of absolute ethanol and deionized water with a volumetric ratio of 1:3 was prepared to disperse TiO 2 and KB. Add 0.1 g of pre weighed TiO 2 powder into 200 mL ethanol/deionized water solvent and conduct ultrasonic treatment for 1.5 h to form a uniform TiO 2 suspension. Then 0.9 g of KB powder was added to TiO 2 dispersion and the dispersion was ultrasonicated for another 2 h to obtain a homogenous a KB/TiO 2 dispersion. Dry KB-TiO 2 powder was finally obtained by filtration followed by freeze-drying.

Preparation of BKT Separator
To prepare the BKT separator, BC wet film was firstly purified: BC was immersed in 1 moL/L NaOH solution at 60 • C and washed with deionized water after one hour immersion. This combined step was repeated several times until full removal of the residual impurities in BC. Then the purified BC was vacuum-dried to obtain a BC membrane. The as-prepared KB-TiO 2 , super p li and PVDF with a weight ratio of 7:1:2 and an appropriate amount of NMP solution was mixed to form a uniform slurry. The slurry was directly coated on the BC film by using a coating machine and then placed in a vacuum drying oven at 60 • C for 12 h to obtain a KB/TiO 2 -coated BC composite membrane (BKT). The prepared BKT was cut into Φ 19 mm discs by using a microtome.

Li 2 S 6 Solution Synthesis
To obtain 0.3 M Li 2 S 6 brown-red solution, Li 2 S (0.14 g) and S powder (0.48 g) were dissolved in 10 mL of DME/DOL (volume ratio 1:1) solution under stirring at 65 • C for 48 h in an argon atmosphere glove box. The solution was further diluted to 3 mM with the same type of DOL/DME solution used for dissolution. The Li 2 S 6 solution was further used to demonstrate the adsorption of KB-TiO 2 to LiPSs and the inhibitory effect of BKT separator on the shuttle effect of LiPSs.

Characterization Techniques
The XRD of the samples was measured by an X-ray diffract meter. After placing the sample powder into the groove of the test matrix, it was flattened with a glass plate to ensure that the surface of the sample is flat. (D 8 -ADVANCE, from Bruker Co., Ltd., Bremen, Germany, voltage: 40 KV, current: 30 mA, 2θ range: 5-90 • , wavelength: K<alpha>1: 1.540593, K<alpha>2: 1.54441, step size: 0.02 • ). Place the sample to be tested in a vacuum drying oven at 60 • C for 12 h. At the same time, place an appropriate amount of phosphorus pentoxide to prevent the generation of moisture and ensure that the sample to be tested is completely dry. After the samples were sprayed with gold, the microscopic morphology of the samples was analyzed by scanning electron microscope (SEM, Regulus8220, from Hitachi High-Tech Co., Ltd., Tokyo, Japan, EHT: 5.0KV). The element distribution on the surface of the sample was collected by an energy spectrometer (EDS, Xflash 6160, from Hitachi High-Tech Co., Ltd.). Determination of mechanical properties of samples was performed by texture analyzer (Ta. XT Plus, from Stable Micro System, Godalming, UK). Measuring the contact angle (OCA50, from Germany Dataphysics Instrument Co., Ltd., Stuttgart, Germany) between the separator and the electrolyte was carried out with a Polymers 2022, 14, 5559 4 of 12 contact angle meter. The thermal stability of the sample was analyzed by thermogravimetry (TGAQ50, from TA Instrument Co., Ltd., Newcastle, DE, USA).

Battery Assembly and Electrochemical Testing
A CR2025 lithium battery was assembled in an argon protective atmosphere glove box with both water and oxygen content lower than 1.0 ppm. The CR2025 cell used sulfur electrode as the cathode, lithium as the anode, BKT as the separator and 1.0 M LTFSI in DOL: DME = (1:1) Vol% with 1.0% LiNO 3 as electrolyte, respectively. At the same time, the battery assembled with KB coated BC substrate separator (BK) was used as a contrast experiment. The assembled half-cell was placed in the glove box for 24 h to ensure adequate infiltration of the electrolyte. Long-cycle and different rate (0.1-1 C) constant-current charge-discharge tests were performed on the two batteries using a Land system at room temperature, with a voltage range of 1.7-2.8 V.
In order to test the ionic conductivity of the separator, a stainless steel (SS) symmetrical battery was assembled and placed in it for 24 h, so that the electrolyte can fully soak the battery. Then, a PARSTAT electrochemical workstation (4000 A) was used to test the AC impedance of the battery at 10 5 -10 −2 Hz and 5 mV amplitude. The following formula 1 was used to obtain the ionic conductivity: σ = L/(R b × S), where σ is the ionic conductivity, L is the thickness of the separator, S is the active area of the stainless steel electrode and R b is the volume resistance of the symmetrical battery [33]. The electrochemical stability of the separator was studied by line scanning voltammetry (LSV). Generally, PARSTAT is used to collect the LSV curve of SS/electrolyte immersion separator/Li half battery at a scanning rate of 5 mV s −1 from 1.5 to 5 V.  [35]. The XRD patterns of KB-TiO 2 composites are a combination of the pristine KB and TiO 2 , indicating the successful preparation of KB-TiO 2 composites. The surface morphology of KB-TiO 2 composites was observed by SEM. Figure S2 (see Supplementary Materials) shows that the material is in a loose and porous state and its abundant pores are conducive to the infiltration of electrolyte and the blocking of LiPSs. The KB-TiO 2 was prepared into a slurry and coated on the BC substrate to obtain the BKT separator.

Preparation and Microstructure Characterization of BKT Separator
The morphology of the KB-TiO 2 coated and non-coated BKT separator are shown in Figure 1b,c, separately. The uncoated BKT separator shows that the BC membrane surface is relatively flat and the long fiber network with porous structure forms a rich pore structure. A successfully KB-TiO 2 coating was confirmed from the change of a fibril network to the porous particle distribution (Figure 1b). The secondary current collectors can reactivate the blocking LiPSs and enhance the performance of the battery [36,37]. The cross-section SEM test of the BKT separator was carried out. The test results are shown in Figure 1d. The coating composed of KB-TiO 2 is uniformly loaded on the BC substrate and the coating is closely bound to the BC substrate. It can be seen from Figure 1d that the coatings of BKT separator are closely stacked and the BC base presents a layered structure. The thickness of the entire separator is 27 µm. The coating thickness is 12 µm. The morphology of the KB-TiO2 coated and non-coated BKT separator are shown in Figure 1b,c, separately. The uncoated BKT separator shows that the BC membrane surface is relatively flat and the long fiber network with porous structure forms a rich pore structure. A successfully KB-TiO2 coating was confirmed from the change of a fibril network to the porous particle distribution (Figure 1b). The secondary current collectors can reactivate the blocking LiPSs and enhance the performance of the battery [36,37]. The crosssection SEM test of the BKT separator was carried out. The test results are shown in Figure  1d. The coating composed of KB-TiO2 is uniformly loaded on the BC substrate and the coating is closely bound to the BC substrate. It can be seen from Figure 1d that the coatings of BKT separator are closely stacked and the BC base presents a layered structure. The thickness of the entire separator is 27 μm. The coating thickness is 12 μm.

Inhibition of LiPSs by BKT Separator
The working mechanism of the BKT separator is shown in Figure 2a. When LiPSs dissolved in the electrolyte pass through the secondary collector composed of KB and TiO2, the physical and chemical adsorption generated by KB and TiO2, respectively, strongly limits the shuttle of LiPSs and improves the performance of the battery. In order to verify the adsorption capacity of KB and KB-TiO2 for LiPSs, a typical adsorption experiment of Li2S6 was carried out. Transfer 20 mg of KB and KB-TiO2 powders into vials respectively, add 2 mL of 3 mM Li2S6 DOL/DME solution, and seal the vials for 12 h. The adsorption capacity was characterized by observing the color change and UV-Vis spectrum after adsorption. The experimental results are shown in Figure 2b, the color of the blank Li2S6 sample is the darkest and the color of the solution after static adsorption of the KB sample changes obviously compared with the color of the original solution, indicating that KB has an adsorption effect on Li2S6. However, the color of the KB-TiO2 sample after static adsorption is lighter than that of the KB sample, indicating that KB-TiO2 has the best adsorption effect on Li2S6. Therefore, it is more conducive to inhibiting the shuttle effect and improving the performance of LSBs. The adsorption difference may be due to the fact that KB only provides physical adsorption; however, KB-TiO2 not only provide physical adsorption, but also polar molecule bonding sites for LiPSs [38]. The UV-Vis spectrum results also correlate with the adsorption experiments. A lower absorption peak of Li2S6 is observed with KB-TiO2, which also indicates that KB-TiO2 has a higher absorption capability for LiPSs.

Inhibition of LiPSs by BKT Separator
The working mechanism of the BKT separator is shown in Figure 2a. When LiPSs dissolved in the electrolyte pass through the secondary collector composed of KB and TiO 2 , the physical and chemical adsorption generated by KB and TiO 2, respectively, strongly limits the shuttle of LiPSs and improves the performance of the battery. In order to verify the adsorption capacity of KB and KB-TiO 2 for LiPSs, a typical adsorption experiment of Li 2 S 6 was carried out. Transfer 20 mg of KB and KB-TiO 2 powders into vials respectively, add 2 mL of 3 mM Li 2 S 6 DOL/DME solution, and seal the vials for 12 h. The adsorption capacity was characterized by observing the color change and UV-Vis spectrum after adsorption. The experimental results are shown in Figure 2b, the color of the blank Li 2 S 6 sample is the darkest and the color of the solution after static adsorption of the KB sample changes obviously compared with the color of the original solution, indicating that KB has an adsorption effect on Li 2 S 6 . However, the color of the KB-TiO 2 sample after static adsorption is lighter than that of the KB sample, indicating that KB-TiO 2 has the best adsorption effect on Li 2 S 6 . Therefore, it is more conducive to inhibiting the shuttle effect and improving the performance of LSBs. The adsorption difference may be due to the fact that KB only provides physical adsorption; however, KB-TiO 2 not only provide physical adsorption, but also polar molecule bonding sites for LiPSs [38]. The UV-Vis spectrum results also correlate with the adsorption experiments. A lower absorption peak of Li 2 S 6 is observed with KB-TiO 2, which also indicates that KB-TiO 2 has a higher absorption capability for LiPSs.
The concentration gradient of LiPSs on cathode and anode is the main cause of the shuttle effect of LSBs. To verify the ability of the BKT separator to suppress the shuttle effect, the diffusion behavior of Li 2 S 6 solution was simulated in a glove box using an H-type electrolytic cell, as shown in Figure 2c. Add an appropriate amount of 3 mM Li 2 S 6 solution and DOL/DME (volume ratio 1:1) solution to the left and right sides of the H-type electric cell. Then place the BKT and BK separators in the middle of the electrolytic cell. At 0 h, the right sides of both systems are in a clear and transparent state. At 6 h, the color on the right side of BK separator electrolytic cell changed, while the color on the right side of BKT separator electrolytic cell did not change. Over time, the color on the right side of BK membrane electrolytic cell gradually deepened, while the right side of BKT  The concentration gradient of LiPSs on cathode and anode is the main cause of the shuttle effect of LSBs. To verify the ability of the BKT separator to suppress the shuttle effect, the diffusion behavior of Li2S6 solution was simulated in a glove box using an Htype electrolytic cell, as shown in Figure 2c. Add an appropriate amount of 3 mM Li2S6 solution and DOL/DME (volume ratio 1:1) solution to the left and right sides of the Htype electric cell. Then place the BKT and BK separators in the middle of the electrolytic cell. At 0 h, the right sides of both systems are in a clear and transparent state. At 6 h, the color on the right side of BK separator electrolytic cell changed, while the color on the right side of BKT separator electrolytic cell did not change. Over time, the color on the right side of BK membrane electrolytic cell gradually deepened, while the right side of BKT membrane electrolytic cell did not change until a slight change 24 h later, indicating that BKT membrane can effectively prevent the shuttle of LiPSs. Figure 3a,b reflect the wettability of the electrolyte on the uncoated surface and the coated surface of the BKT separator using the contact angle measuring instrument. The wettability of the un-coated surface of the BKT separator to the electrolyte is reflected using a contact angle meter. The size of the contact angle reflects the wettability. The smaller the contact angle, the better the wettability of the separator. Good wettability helps to reduce the interface resistance [39]. The contact angle results against electrolyte solution are shown in Figures 3a,b. For uncoated BKT surface and coated BKT surface, the contact angle is 32.5 degrees and 21.9 degrees, respectively, at 0s. At 60s, the contact angle is 14.9 degrees and 3.2 degrees, respectively. It can be concluded that BKT separator has excellent  Figure 3a,b reflect the wettability of the electrolyte on the uncoated surface and the coated surface of the BKT separator using the contact angle measuring instrument. The wettability of the un-coated surface of the BKT separator to the electrolyte is reflected using a contact angle meter. The size of the contact angle reflects the wettability. The smaller the contact angle, the better the wettability of the separator. Good wettability helps to reduce the interface resistance [39]. The contact angle results against electrolyte solution are shown in Figure 3a,b. For uncoated BKT surface and coated BKT surface, the contact angle is 32.5 degrees and 21.9 degrees, respectively, at 0 s. At 60 s, the contact angle is 14.9 degrees and 3.2 degrees, respectively. It can be concluded that BKT separator has excellent electrolyte wettability. The excellent electrolyte wettability of BKT membrane is mainly attributed to the rich oxygen containing groups on its uncoated surface having affinity for electrolyte [39] and the rich pore structure of BKT coating.

BKT Separator Performance Test
Sufficiently high thermal stability of the separator is crucial to the safety of the battery. In order to explore the thermal stability of the BKT separators, a thermal shrinkage test was conducted (Figure 3c). The separator with a diameter of 19 mm is kept at 20 • C and 180 • C for 30 min respectively, then the size change of the separator is observed. It can be seen from Figure 3c that, as temperature rises, no significant dimensional change is observed for the BKT separator even at 180 • C, indicating that the BKT separator has high enough thermal dimensional stability. The excellent thermal dimensional stability of BKT is mainly due to the high crystallinity of oriented cellulose chain in BC [40]. Figure 3d shows the TGA of BKT separator and BK separator, respectively. It can be seen from the figure that, before 100 • C, two kinds of separator will experience a weight loss, which is mainly caused by the evaporation of water molecules. With the increase of temperature, the membranes degraded and the initial degradation temperature of both membranes was 299 • C. At 800 • C, the coke residue of BKT separator is 27% and that of BK separator is 19%.
Polymers 2022, 14, x FOR PEER REVIEW 7 of 13 electrolyte wettability. The excellent electrolyte wettability of BKT membrane is mainly attributed to the rich oxygen containing groups on its uncoated surface having affinity for electrolyte [39] and the rich pore structure of BKT coating. Sufficiently high thermal stability of the separator is crucial to the safety of the battery. In order to explore the thermal stability of the BKT separators, a thermal shrinkage test was conducted (Figure 3c). The separator with a diameter of 19 mm is kept at 20 °C and 180 °C for 30 min respectively, then the size change of the separator is observed. It can be seen from Figure 3c that, as temperature rises, no significant dimensional change is observed for the BKT separator even at 180 °C, indicating that the BKT separator has high enough thermal dimensional stability. The excellent thermal dimensional stability of BKT is mainly due to the high crystallinity of oriented cellulose chain in BC [40]. Figure  3d shows the TGA of BKT separator and BK separator, respectively. It can be seen from the figure that, before 100 °C , two kinds of separator will experience a weight loss, which is mainly caused by the evaporation of water molecules. With the increase of temperature, the membranes degraded and the initial degradation temperature of both membranes was 299 °C . At 800 °C , the coke residue of BKT separator is 27% and that of BK separator is 19%.
The mechanical properties of the BKT separator were tested to verify the stability of the coating and the flexibility of the separator. The results are shown in Figure 3e. After the BKT separator is bent and immersed in electrolyte, the coating and BC substrate are still tightly bonded. No mechanical separation, powder or slag removal occurred, indicating that the BKT separator still has good flexibility and stability under various conditions. This excellent performance not only blocks the physical barrier of the LSBs during the cycle, but also ensures the cycle performance of the battery. The separator was subjected The mechanical properties of the BKT separator were tested to verify the stability of the coating and the flexibility of the separator. The results are shown in Figure 3e. After the BKT separator is bent and immersed in electrolyte, the coating and BC substrate are still tightly bonded. No mechanical separation, powder or slag removal occurred, indicating that the BKT separator still has good flexibility and stability under various conditions. This excellent performance not only blocks the physical barrier of the LSBs during the cycle, but also ensures the cycle performance of the battery. The separator was subjected to tensile test using a texture analyzer. The results are shown in Figure 3f. The tensile stress at the fracture point of BKT and BK separators is 112 MPa and 107 MPa, respectively, with little difference between them. The good tensile properties can ensure battery assembly [40]. The stability of the separator during the process ensures the safety of the battery. Figure 3g shows LSV curves of BKT separator and BK separator, respectively. LSBs will experience a chemical reaction during operation, so it is necessary to ensure that the separator has a certain chemical inertia. Therefore, linear scanning voltammetry is used to evaluate the electrochemical stability between the membrane and electrolyte. It can be seen from the figure that the response current of the two separators remains stable between 1.5-3.08 V, which can ensure electrochemical inertia in the charge discharge range of 1.7-2.8 V of lithium sulfur battery.
Next, Nyquist plots of SS/SS symmetric cells was collected to evaluate the ion conductivity of separators. Good ionic conductivity is very important to the rate performance of the battery. The test results are shown in Figure 3h, the volume resistance of the separator is represented by the real intercept, and the ionic conductivity of the separator can be obtained by equation. The volume resistances of BKT and BK separators are 9.5 Ω and 11.2 Ω, respectively. The ionic conductivity of the separator can be calculated by Formula 1.
The calculated values of BKT and BK separators are 0.14 mS × cm −1 and 0.119 mS × cm −1 .
In this work, KB/S composites were prepared. As shown in Figure S3 (see Supplementary Materials), the actual sulfur content in the KB/S composite obtained through TGA test is 67 wt%, close to the theoretical value (70%, mass fraction). The cathode was prepared with KB/S composite material and the battery was assembled with BKT separator to test the cycle performance and rate capability of the battery. Figure 4a,b show the charge discharge curves of the first and second cycles of the battery based on BKT separator and BK separator when the current density is 0.5 C. It can be observed that the specific discharge capacities of the BKT separator based battery in the first and second cycles are 1180.8 mAh × g −1 and 1114.5 mAh vg −1 , respectively, which are higher than the 1059.4 mAh × g −1 and 1021.8 mAh × g −1 of the BK separator based battery. There are two discharge plateaus in its discharge curve, among which the 2.3 V high-voltage discharge plateau corresponds to the transformation of the cyclic S 8 molecule to Li 2 S n (4 ≤ n ≤ 8). However, the low-voltage discharge plateau of the LSBs is mainly provided by the discharge capacity. First, the low-voltage discharge plateau mainly represents the reduction of Li 2 S 4 to short-chain Li 2 S 2 /Li 2 S [41] and the charging process is mainly the transformation from Li 2 S 2 /Li 2 S to S 8 . The reason why the BKT separator battery obtains such excellent performance is mainly because the coating composed of TiO 2 and KB can not only limit LiPSs through KB physical adsorption and TiO 2 chemical adsorption, but also the conductive layer can be used as a secondary collector to activate the adsorbed LiPSs. Table S1 (see Supplementary Materials) shows the electrochemical performance of batteries equipped with different separators. It can be seen from the table that the BKT separator has better performance than other separators. As shown in Figure 4a,b, the phase conversion coefficient (denoted Q 2 /Q 1 ) of BKT (2.32) is higher than that of BK (1.83), where Q 1 denotes the solid-liquid conversion process of S 8 to Li 2 S 4 and Q 2 corresponds to the liquid-solid-solid conversion process of Li 2 S 4 to Li 2 S 2 /Li 2 S, implying the high sulfur utilization ability and effective LiPS conversion during the redox process [42]. Figure 4c is the 100-cycle test diagram of the BKT separator battery and the BK separator battery at a large current density of 0.5 C. It can be seen from the figure that the discharge specific capacity of both BKT separator-based and BK separator-based batteries gradually decreases with the increase in cycle numbers, but the capacity of the BKT separator battery always maintains an advantage during the cycle. The initial discharge specific capacity of the BKT separator battery is 1180.8 mAh × g −1 , with a sulfur utilization rate of 70%. The discharge specific capacity after 100 cycles drops to 653.7 mAh × g −1 and the capacity retention rate is 55.4%. The initial discharge specific capacity of the BK separator-based battery is 1159.4 mAh × g −1 , with a sulfur utilization rate of 69%. After 100 cycles, the discharge specific capacity is 535.9 mAh × g −1 and the capacity retention rate is 46.2%. Figure 4d shows the magnification test diagram of BKT separator and BK separator battery. At the current rate of 0.1-1 C, the first discharge specific capacities of the BKT separator-based battery at different rates are 1520.3, 1053.9, 783.5 and 652.4 mAh × g −1 , respectively. The BK-based batteries are 1364, 948.7, 708, 581.6 mAh × g −1 . When the current rate is recovered from 1 C to 0.1 C, the discharge specific capacity of the BKTbased battery is 866 mAh × g −1 and the specific capacity of the BK-based battery is 782.6 mAh × g −1 , the results showing that the battery assembled by the BKT separator has good reversibility. The above results show that the BKT separator inhibits the shuttle effect of LiPSs to a certain extent and shows better rate performance. Polymers 2022, 14, x FOR PEER REVIEW 9 of 13  Figure 4c is the 100-cycle test diagram of the BKT separator battery and the BK separator battery at a large current density of 0.5 C. It can be seen from the figure that the discharge specific capacity of both BKT separator-based and BK separator-based batteries gradually decreases with the increase in cycle numbers, but the capacity of the BKT separator battery always maintains an advantage during the cycle. The initial discharge specific capacity of the BKT separator battery is 1180.8 mAh × g −1 , with a sulfur utilization rate of 70%. The discharge specific capacity after 100 cycles drops to 653.7 mAh × g −1 and the capacity retention rate is 55.4%. The initial discharge specific capacity of the BK separator-based battery is 1159.4 mAh × g −1 , with a sulfur utilization rate of 69%. After 100 cycles, the discharge specific capacity is 535.9 mAh × g −1 and the capacity retention rate is 46.2%. Figure 4d shows the magnification test diagram of BKT separator and BK separator battery. At the current rate of 0.1-1 C, the first discharge specific capacities of the BKT separator-based battery at different rates are 1520.3, 1053.9, 783.5 and 652.4 mAh × g −1 , respectively. The BK-based batteries are 1364, 948.7, 708, 581.6 mAh × g −1 . When the current rate is recovered from 1 C to 0.1 C, the discharge specific capacity of the BKT-based battery is 866 mAh × g −1 and the specific capacity of the BK-based battery is 782.6 mAh × g −1 , the results showing that the battery assembled by the BKT separator has good reversibility. The above results show that the BKT separator inhibits the shuttle effect of LiPSs to a certain extent and shows better rate performance.

Electrochemical Characterization
Theoretically, the LSBs has an energy density far exceeding the existing battery system. However, due to the low conductivity of the sulfur element, a large amount of inactive conductive matrix must be introduced to increase the conductivity of the LSBs. In addition, in order to achieve better electrical conductivity, the coating thickness of the positive electrode material is low and the sulfur loading is also often low, which makes it difficult for the energy density of the battery to meet people's requirements.
Therefore, this experiment explores the electrochemical performance of BKT separator batteries at high sulfur areal densities of 1.5 mg × cm −2 and 2.0 mg × cm −2 . The current density is small at 0.1 C. Compared with the charge and discharge at a large current density, the single charge and discharge time is longer and the soluble long-chain LiPSs stay in the electrolyte for a long time, resulting in a shuttle effect. The greater the possibility of Theoretically, the LSBs has an energy density far exceeding the existing battery system. However, due to the low conductivity of the sulfur element, a large amount of inactive conductive matrix must be introduced to increase the conductivity of the LSBs. In addition, in order to achieve better electrical conductivity, the coating thickness of the positive electrode material is low and the sulfur loading is also often low, which makes it difficult for the energy density of the battery to meet people's requirements.
Therefore, this experiment explores the electrochemical performance of BKT separator batteries at high sulfur areal densities of 1.5 mg × cm −2 and 2.0 mg × cm −2 . The current density is small at 0.1 C. Compared with the charge and discharge at a large current density, the single charge and discharge time is longer and the soluble long-chain LiPSs stay in the electrolyte for a long time, resulting in a shuttle effect. The greater the possibility of this, the more the charge-discharge cycles at small current densities can verify the ability of the separator to suppress the shuttle effect. It can be seen from Figure 4e that the initial discharge specific capacity of the BKT separator battery decreases with the increase of the sulfur-loaded surface density. The initial discharge specific capacities of 0.1 C current density are 1440 mAh × g −1 and 1274 mAh × g −1 , respectively, at 1.5 mg × cm −2 and 2.0 mg × cm −2 sulfur loading surface densities of electrodes. After 100 cycles, the battery still has a discharge specific capacity of 809 mAh × g −1 and 566 mAh × g −1 . These results indicate that the BKT separator can also effectively inhibit the shuttle of LiPSs under the high sulfur-loading areal density of the battery, thereby enabling the LSBs to improve discharge capacity and battery energy density.
As shown in Figure 4f, the BKT separator based battery was tested for lighting the LED lamp. This figure shows that the BKT separator based battery can make the LED light bright at the beginning, but the brightness decreases significantly with time. We also monitored the open circuit potential of the battery before and 1 h after discharge. After continuously lighting the LED lamp for hours, the battery potential decreased slightly from 2.72 V to 1.83 V.

Characterization of BKT Separator before and after Cycling
In order to further characterize the inhibitory effect of the BKT separator on the shuttle effect, SEM together with EDS tests were performed on the BKT separator before cycling and after 100 charge-discharge cycles. According to the SEM diagram before and after BKT separator cycling in Figure 5a,b, a new layer was formed on the surface of BKT separator after cycling, which indicates that the BKT separator effectively captured LiPSs during the battery charging and discharging cycle. In addition, according to the EDS diagram before and after BKT separator circulation in Figure 5a,b, sulfur was found in the separator after the charge-discharge cycle. However, no sulfur was observed on the separator prior to circulation. The appearance of sulfur element further demonstrated the retention of LiPSs on the BKT membrane. bright at the beginning, but the brightness decreases significantly with time. We also monitored the open circuit potential of the battery before and 1 h after discharge. After continuously lighting the LED lamp for hours, the battery potential decreased slightly from 2.72 V to 1.83 V.

Characterization of BKT Separator before and after Cycling
In order to further characterize the inhibitory effect of the BKT separator on the shuttle effect, SEM together with EDS tests were performed on the BKT separator before cycling and after 100 charge-discharge cycles. According to the SEM diagram before and after BKT separator cycling in Figures 5a,b, a new layer was formed on the surface of BKT separator after cycling, which indicates that the BKT separator effectively captured LiPSs during the battery charging and discharging cycle. In addition, according to the EDS diagram before and after BKT separator circulation in Figure 5a,b, sulfur was found in the separator after the charge-discharge cycle. However, no sulfur was observed on the separator prior to circulation. The appearance of sulfur element further demonstrated the retention of LiPSs on the BKT membrane.

Conclusions
In conclusion, in this paper, green BC is used as a substrate, and KB-TiO 2 material is coated on its surface to form a BKT separator, which is used to suppress the shuttle effect of LSBs. The BKT separator not only has excellent thermal stability and good wettability to electrolyte, but also shows excellent electrochemical performance of LSBs. The results show that the initial discharge specific capacity of the battery based on the BKT separator with 0.5 C high current density is 1180 mAh × g −1 under the surface density of 1 mg × cm −2 sulfur carrying electrode. At the same time, the initial discharge specific capacity of the battery based on BKT separator at 0.1 C low current density is 1440 and 1274 mAh × g −1 , respectively, under the high sulfur carrying surface density of 1.5 and 2 mg × cm −2 electrodes. The above results prove that the separator can suppress the shuttle effect of LSBs well and demonstrates a good application prospect for the separator.