Enhancing Li-S battery performance via functional polymer binders for polysulﬁde inhibition

The commercialization of lithium-sulfur (Li-S) batteries faces several challenges, including poor conductivity, unexpected volume expansion, and continuous sulfur loss from the cathode due to redox shuttling. In this study, we introduce a novel polymer via a simple cross-linking between poly(ether-thioureas) (PETU) and poly(3,4-ethylenedioxythiophene)- poly(styrenesulfonate) (PEDOT:PSS) as a bifunctional binder for Li-S batteries (devotes as ‘‘PPTU”). Compared to polyvinylidene ﬂuoride (PVDF), as-prepared PPTU exhibits signiﬁcantly higher electrical conductivity, facilitating electrochemical reactions. Additionally, PPTU demonstrates effective adsorption of lithium polysulﬁdes, leading to improved cycling stability by suppressing the shuttling effect. We investigate this behavior by monitoring morphological changes at the cell interface using synchrotron X-ray tomography. Cells with PPTU binders exhibit remarkable rate performance, desired reversibility, and excellent cycling stability even under stringent bending and twisting conditions. Our work represents promising progress in functional polymer binder development for Li-S batteries.

Besides the challenges posed by the shuttling effect, the conductivity of the S cathode plays a pivotal role in achieving optimal battery performance.The relationship between electrode conductivity and electrochemical reactions is straightforward, and efficient charge and discharge processes rely on seamless electron and ion movement within the electrode material.Unfortunately, the intrinsic insulating nature of sulfur (S) itself hampers electrode conductivity.Additionally, traditional binders used in Li-S batteries contribute to this limitation.In this response, researchers have diligently explored novel carbon materials and composite architectures to enhance the S cathode [38].While progress has been made in electrode materials, the quest for an ideal binder remains and attracts constant interest.For example, a cell with polythiourea-triglycerol (PTTG) exhibited a stable electrochemistry for 80 cycles at 0.2 C [39].Wang et al., also presented that a cell with zwitterion polymer (ZIP) displayed a significant improvement on Li-S battery performance [40].Despite these advancements, a binder that combines substantial conductivity with effective shuttling inhibition remains elusive.Researchers continue to explore innovative strategies, seeking the holy grail of binders to bridge the gap between conductivity and functionality.
In the previous report, a novel polymer of PPTU via simple cross-linking between poly(ether-thioureas) (PETU) and poly(3,4ethylenedioxythiophene)-poly(styrenesulfonate) (PEDOT:PSS) has been developed as a functional electrode material in siliconbased Li-ion battery [41].Considering its good conductivity and the ability to inhibit volume expansion, in this work, we explore and develop it as a bifunctional binder for Li-S batteries.Compared to the commonly used PVDF, as-prepared PPTU possesses a much higher electrical conductivity.In addition, it exhibits a good adsorption effect on LiPSs due to its positively charged functional groups, which can increase the cycling stability by mitigating the shuttling effect.Moreover, to delve deeper into the interface evolution during cell cycling [42], synchrotron X-ray computed tomography (SXCT) À an advanced, nondestructive technique with high spatiotemporal resolution is employed.Surprisingly, there have been few reports of SXCT in Li-S systems [43].Our study pioneers its application, shedding light on the mechanism in the shuttling effect inhibition.In this work, a Li-S cell incorporating PPTU as the binder demonstrates remarkable cycle stability (exceeding 200 cycles at 1 C), a good rate performance (754.3 mA h g À1 at 5 C), and a desired reversibility (a high-capacity retention rate of 86% after 100 cycles at 0.5 C with a high S loading of 5.8 mg cm À2 ).Furthermore, the electrochemical stability of PPTU-involved cells under stringent bending and twisting conditions is rigorously evaluated.We envision PPTU finding applications in other energy storage systems dealing with shuttling effects.
Li foil with a thickness of 500 lm was purchased from China Energy Lithium Co., Ltd.Polypropylene (PP) separator (Celgard 2400) was purchased from Liaoyuan Hongtu LIBS Technology Co., Ltd.Nickel foam with a thickness of 500 lm was purchased from Hefei Saibo New Materials Co., Ltd.

PETU preparation
1.93 g (0.013 mol) of BAEE and 2.23 g (0.013 mol) of 1,1 0 -thiocar bonyldiimidazole were added to 7 mL of DMF slowly and stirred at 30 °C for 24 h. 10 mL of CH 2 Cl 2 was then added to the mixed solution above, which was dumped into 80 mL of Et 2 O and the precipitate was obtained in a diluted mixture.As-obtained precipitate was further washed with a mixture of CH 2 Cl 2 /EtOH (v/v = 1:3) for three times, which was collected and then dried in a vacuum oven at 140 °C for 10 h.Therefore, the resulting PETU was obtained.The synthetic chemical formula for PETU is shown in Fig. S1 in Supporting Information, and the successful synthesis is confirmed by nuclear magnetic resonance ( 1 H NMR) in Fig. S2.

PPTU preparation
As-obtained PETU and PEDOT:PSS were mixed with a weight ratio of 1:1 with 15 mL of NMP and then stirred at 60 °C for 20 h to generate PPTU.A schematic diagram for the synthesis process is illustrated in Fig. S3.

S@C cathode preparation
A preparation for S@C composites follows a process [44].S powder and carbon 350 with a weight ratio of 7:3 were grounded using a mortar and pestle for 30 min, and then heated at 155 °C for 20 h in an Ar atmosphere to obtain S@C composites [45].As-prepared S@C composites, PPTU and SP were mixed in the different ratios for 8:1:1, 7:2:1, and 6:3:1 (w/w/w) and stirred for 24 h to make a slurry, which was coated on an Al foil with a diameter of 13 mm, denoted as PPTU-based cathode-1, PPTU-based cathode-2, and PPTU-based cathode-3, respectively.The results show that the optimum performance for a cell with PPTU-based cathode-2 was achieved at 0.2 C in Fig. S4.Therefore, we selected PPTUbased cathode-2 as a representative sample in this study, which was denoted as ''S@PPTU" cathode as an abbreviation.In order to study the PPTU potential, we also prepared a reference cathode with the traditional PVDF in a similar method (i.e., an electrode consisted of S@C composite:PVDF:SP with a weight ratio of 7:2:1), which was denoted as ''S@PVDF" cathode.Note that a wide weight density range of 1.0-9.0mg cm À2 for S loading was employed to investigate the sulfur amount influence on battery performance.

Characterizations
1 H NMR was performed on a Bruker Avance Ⅲ HD (600 MHz) with CDCl 3 as solvent.Fourier transform infrared (FT-IR, Thermo-Scientific Nicolet 6700) was performed to identify the molecular structure for the various samples in a wavelength range of 400-4000 cm À1 .The morphologies, particle sizes, and element analyses for the different samples were characterized by scanning electron microscopy (SEM, Hitachi S-5500) coupled with an EDS (Ametek).The detailed crystal and morphological information was examined by transmission electron microscopy (TEM) and high resolution transmission electron microscopy (HRTEM) (JEOL-2100F).180°p eel and tensile properties were determined using the WBE-9010A 180°peel mode.Ultraviolet-visible spectroscopy (UV-Vis, HITACHI U-3900) was used to study the adsorption properties on LiPSs by the binder.The measured spectra were in a range of 300-600 nm.The surface composition and chemical state for the different cathodes were analyzed by X-ray photoelectron spectroscopy (XPS) on an ESZALB 250 XL XPS system with Al K a radiation (hm = 1486.7 eV) and an emission angle of 90°.The binding energy scale was calibrated with C 1s peak assigned to CÀH bond at 285.0 eV.Li-S tomography cells with S@PVDF and S@PPTU cathodes were measured using SXCT after 5 cycles with in situ mode (i.e., test the cell without disassembly) in Fig. S5.The experiments were conducted at BL13WB beamline at Shanghai Synchrotron Radiation Facility (SSRF) of China.By utilizing a double crystal monochromator (DMM), the synchrotron beam was monochromatized to 20 keV.An X-ray CCD detector was employed to achieve a spatial resolution of 1.625 lm pixel À1 , while the temporal resolution was set to 1 ms.The view field encompassed a 3.328 mm Â 3.
328 mm area, allowing for a full synchrotron X-ray tomography by rotating 180°.A total of 1280 projections were recorded during one test, and each exposure time was 80 ms.Subsequently, the crosssectional X-ray tomographic slices were analyzed using Avizo Studio 2021.1.
The diffusion coefficient of Li + can be calculated from the low frequency region of EIS plots by Eq. ( 1) as following: where R is the gas constant, T is the absolute temperature, A is the surface area of the cathode, n is the number of electrons per molecule during oxidization, F is the Faraday constant, C is the concentration of Li + , and r is the Warburg factor which obeys the following Eq.( 2): where R s is the resistance between the electrolyte and electrode, R ct is the charge transfer resistance, and x is the angle frequency, respectively.

Cell assembly and electrochemical characterizations
Li-S cell applies CR2025 coin type, which consisted of S@PVDF or S@PPTU as a cathode, PP as a separator soaked with 1 M LiTFSI with 1 wt% LiNO 3 in DOL/DME (1:1, v/v), and Li foil as an anode.Li-S pouch cell was assembled with S@PPTU as a cathode, PP as a separator soaked with 1 M LiTFSI with 1 wt% LiNO 3 in DOL/DME (1:1, v/v), and ultra-thin Li foil with a thickness of 50 lm as an anode.
The home-made transparent cell contained S@PVDF or S@PPTU as a cathode, 1 M LiTFSI with 1 wt% LiNO 3 in DOL/DME (1:1, v/v) as the electrolyte, and Li foil as an anode.The tomography cell consisted of S@PVDF or S@PPTU as a cathode, PP as a separator soaked with 1 M LiTFSI with 1 wt% LiNO 3 in DOL/DME (1:1, v/v), and Li foil as an anode.The electrolyte addition for each cell was 25 lL mg À1 .
Electrochemical impedance spectroscopy (EIS) tests were conducted on a PARSTAT 4000 potentiostat with a voltage amplitude of 10 mV and a frequency window of 10 À1 -10 5 Hz.The conductivity for as-prepared cathodes was measured using a four probe tes-ter (RTS-8).Galvanostatic discharge/charge measurements were performed using the NEWARE battery tester within a potential range of 1.6-2.8V vs. Li/Li + at the different currents of 0.1-5 C.

Results and discussion
PPTU was prepared via a facile cross-linking between PETU (poly(ether-thioureas) synthesized from BAEE and 1,1 0 -thiocarbo nyldiimidazole) and PEDOT:PSS, which was proved by FT-IR measurement (Fig. 1a and Fig. S6).Note that sulfur distribution in S@PPTU cathode is uniform with little aggregation (Fig. 1b), which is also confirmed by EDS elemental mapping in Fig. 1(c).The detailed morphology for as-prepared S@PPTU cathode was also studied by TEM and HRTEM.As a comparison with S@PVDF reference cathode (Fig. S7), it can be seen that an obvious layer with a thickness of $7 nm is attached well on the surface (Fig. 1d).This could provide a favorable potential on decreasing or inhibiting the shuttling of LiPSs [46].To study the peeling forces for the different cathodes, we conducted a 180°peel test.In Fig. 1e, it can be seen that the adhesion force of S@PPTU cathode is 7.2 N, which is 2.4 times that of S@PVDF (3.0 N).This could be explained that a cross-linking network generated on S surface and interlocking layer formed between thiourea groups in PPTU and the current collector promote the elongation rate for as-prepared S@PPTU [41].
EIS tests were performed on cells with S@PPTU and S@PVDF, and the corresponding resistances were obtained via equivalent circuit fitting.The two semicircles in the high frequency region and the middle frequency region and the straight line in the low frequency region respectively represent the interface film resistance (R s ), charge transfer resistance (R ct ) and Li + diffusion in the body electrode when Li + passes through the PP.A cell with S@PPTU possesses a better conductivity and a faster Li + migration than the one with S@PVDF, as shown in Fig. 2(a and b), Fig. S8, Tables S1 and  S2.It is clear that cells with S@PVDF show a significant decrease in impedance after 10 and 50 cycles, which is actually due to the loss of active material.This agrees well with the results from four probe tests, in which that S@PPTU's conductivity of 14.3 S cm À1 is ten times higher than S@PVDF's conductivity of 1.3 S cm À1 .These results suggest the considerable advantage on charge transfer for as-prepared S@PPTU, which favors electrochemical reactions [44,47].The stability of electrochemical reaction for cells with the different cathodes was studied through CV test, as shown in Fig. S9.It can be seen that the electrochemical stability for a cell with S@PPTU is better than that for a cell with S@PVDF.The home-made transparent cells were used to verify the strong adsorption of the different binders on LiPSs during discharging.In Fig. 2(c) and Fig. S10, it has been well visualized that PPTU binder plays an important role on restraining LiPSs, which is beneficial to the enhancement of cycling stability in Li-S cells.
The influence of the different binders on the electrochemical performance was investigated by galvanostatic charge-discharge (GCD) tests.Compared to the cycling performance for a cell with S@PVDF, the one in a cell with S@PPTU displays a higher capacity, as shown in Fig. 3(a) and Fig. S11.This can be attributed to the strong adsorption of LiPSs by the functional groups in PPTU.In order to study the magnitude of the specific capacity, we analyzed the contribution for the different discharge platforms.In principle, the discharge profile exhibits two identifiable reaction plateaus with adjacent sloping regions.The plateau at $2.3 V is attributed to the transformation of S 8 to Li 2 S x (4 x 8), representing the dis-charge capacity for the high discharge voltage plateau known as Q H ; while the lower one at $2.1 V is associated to the conversion of Li 2 S 4 to Li 2 S 2 /Li 2 S, corresponding the discharge capacity of the low voltage plateau denoted as Q L (Fig. S12) [23,45].Note that the ratio for Q L /Q H reflects the effective conversion of LiPSs, i.e., the bigger Q L /Q H is, the higher LiPSs conversion exists in the cell [48].All the values of Q H , Q L , and Q L /Q H for a cell with S@PPTU cathode are higher than those for a cell with S@PVDF cathode (Fig. 3b), confirming the promoting potential for PPTU in converting and holding LiPSs during discharge.In addition, a cell with S@PPTU exhibits a much better rate performance than the one with S@PVDF, as seen in Fig. 3(c) and Fig. S13.These improvements by as-prepared S@PPTU can be attributed to the conductivity enhancement and the inhibition for the LiPSs shuttling.
Although increasing S mass loading is expected to obtain the higher energy density in Li-S applications, it has been hardly achieved thus far due to the limited cathode conductivity [10,49,50].In Fig. 4(a), it can be seen that S@PPTU-containing cells with high S loadings of 3.0 and 5.8 mg cm À2 display a good cycling stability.To be adaptable to the practical application, a pouch cell is employed to study the potential for a higher S loading in a Fig. 2. Nyquist plots for cells with (a) S@PVDF and (b) S@PPTU cathodes at 0.5 C; (c) the photos for home-made transparent cells with S@PVDF and S@PPTU cathodes during discharging at 0.2 C within a potential range of 1.6-2.8V vs. Li/Li + .S@PPTU-containing cell, as seen in Fig. 4(b).A considerable cycling stability is maintained even when a high S loading of 20 mg is applied.Moreover, a cell shows a highly electrochemical stability even under stringent bending and twisting conditions (Fig. 4c-g), confirming the potential application for S@PPTU-containing cells in foldable and flexible electronic devices.These results represent a promising progress in the development of functional binder in Li-S batteries.
SXCT was employed to reveal interface evolution in cells with different cathodes, as seen in Fig. 5. Compared to the pristine interface image, the cell with S@PVDF shows a serious corrosion on Li anode, resulting from the LiPSs shuttling.In the previous report, the dissolved LiPSs intermediates defuse to the Li metal anode, where they can be reduced to LiPSs with shorter chains into the electrolyte or become solid products via side reactions [51].This could be the main reason of ''dead Li", which leads to a capacity fading or even sudden death.However, little change can be observed on the interface between Li anode and the separator in the cell with S@PPTU.This could be explained that PPTU inhibits the shuttle effect of LiPSs, thereby improving the cycling stability.In addition, the cross-sectional SEM images for the different cathodes before and after 10 cycles at 0.5C within a potential range of 1.6-2.8V vs. Li/Li + provide a direct evidence that PPTU addition can contribute to controlling the volume expansion of the cathode, as seen in Fig. S14.
The morphologies on the different cathodes before and after 10 cycles are studied by SEM.In Fig. 6(a) and Fig. S14, it can be seen that the S in S@PPTU is much evenly dispersed than that in S@PVDF at the initial state due to the considerable adhesion of PPTU.This is in a good agreement with the EDS elemental mapping of S (Fig. 1c).Additionally, as-prepared S@PPTU maintains well after 10 cycles, while the severe aggradation appears in S@PVDF.The good mechanical property of S@PPTU is also proved by the multifolding tests, as seen in Fig S15 .XPS is used to explore the surface composition for S@PPTU cathode, as seen in Fig. 6 ), N-S, and N-Li bonds, respectively [25].This suggests the presence of chemical interaction between PPTU binder and LiPSs [25].The adsorption of LiPSs by PPTU was also proved in S 2p spectrum.Compared to the peaks at the pristine state, the ones correspond to the formation of complexes of thiosulfate (159-166 eV) and polythioate (167-172 eV) after 10 cycles, further confirming that as-prepared PPTU could adsorb LiPSs [36].In addition, compared to C 1s before cycling, two extra peaks at 292.9 and 531.2 eV exist after cycling, which assign to C-F and C=O (Fig. S16a) [52].This suggests the SEI formation through electrochemical decomposition of LiTFSI and/or ether group [52].Moreover, Li 1s spectrum reveals the presence of Li in at least three different states at binding energies of 54.6, 55.1, and 55.5 eV, demonstrating the Li-N, Li-S, and Li-O bonds, which suggests an induced action of electro-negative N, S, and O atoms (Fig. S16b) [53,54].The XPS results reveal the strong LiPSs adsorption capability of PPTU binder, which is one of its positive functions for maintaining the S utilization and enhancing cycling stability.
Density functional theoretical calculations were performed in order to further verify the suppression of shuttle effect on soluble LiPSs by PPTU.All calculations were carried out using the GAUS-SIAN09 program [55].The geometries of PPTU, soluble Li 2 S x (x = 4, 6, and 8), and PPTU-Li 2 S x (x = 4, 6, and 8) were optimized at the M062x/6-31G(d) level.The computational results exhibit that hydrogen bonding in PPTU is formed through the pairwise mutual hydrogen bonds among PSS, PEDOT, and PETU.When Li 2 S x dissolves into the electrolyte, the weak hydrogen bonds between  7a), which are stronger than that of hydrogen bonding [56].In addition, these adsorption energies are similar to the ones in previous reports in the current state for Li-S batteries [55].Therefore, such strong interaction between PPTU and Li 2 S x (x = 4, 6, and 8) is beneficial for inhibiting the shuttle of soluble LiPSs and further improving the cycling stability.As mentioned above, the asprepared PPTU binder offers several advantageous features, including an easy-to-operate synthesis route, enhanced conductivity, and significant adsorption capacity for LiPSs.These attributes collectively contribute to improved electrochemical performance, specifically enhancing cycling stability and rate capability.The straightforward synthesis for the PPTU binder streamlines largescale production, ensuring reproducibility.Moreover, its ability to enhance both ionic and electronic conductivity within the electrode facilitates charge transfer during battery cycling.Notably, the PPTU binder's strong affinity for LiPSs mitigates their dissolution and loss, resulting in enhanced overall battery performance.As-prepared PPTU binder demonstrates promise as a valuable component in Li-S batteries.Fig. 7(b) provides schematic illustrations depicting the distinct mechanisms involved in cells using PVDF and PPTU binders.
(b).Compared to N 1s spectrum on the pristine state, three extra peaks at 397.1, 398.6, and 404.1 eV after 10 cycles are observed, which correspond to N-O (NO 3 À

Fig. 3 .
Fig. 3. (a) Cycling performances; (b) Q H , Q L ,and Q L /Q H for cells with S@PVDF and S@PPTU cathodes at 1 C; (c) rate capabilities for cells with S@PVDF and S@PPTU cathodes at various current rates from 0.2 to 5 C within a potential range of 1.6-2.8V vs. Li/Li + .

Fig. 4 .
Fig. 4. (a) Cycling performances for S@PPTU-containing cells with different S loadings of 3.0 and 5.8 mg cm À2 at 0.5 C; (b) cycling performance for a Li-S pouch cell using S@PPTU with a S loading of 20 mg at 0.1 C within a potential range of 1.6-2.8V vs. Li/Li + ; Photos for Li-S pouch cell with S@PPTU cathode under the different test conditions of (c) 0°, (d) 90°, and (e) 180°bending, and (f) 90°and (g) 180°twisting.

Fig. 5 .Fig. 6 .
Fig. 5. SXCT slice images and the corresponding 3D rendering images of cells using S@PVDF and S@PPTU after 5 cycles at 0.2 C within a potential range of 1.6-2.8V vs. Li/Li + .