Lithium–sulfur pouch cells with 99% capacity retention for 1000 cycles

Abstract

The lithium–sulfur (Li–S) battery is a highly promising candidate for next-generation battery systems. However, the shuttle effect of polysulfides or the dendrites and side reactions of lithium metal anodes limit the cycle life of batteries. In particular, at the pouch cell level, achieving long-term cycling stability is extremely challenging. Here, we have constructed a Li–S pouch cell with sulfurized pyrolyzed poly(acrylonitrile) (SPAN) as the cathode and graphite (Gr) as the anode, introducing lithium-ions through a facile in situ pre-lithiation method. In carbonate-based electrolytes, the SPAN cathode can avoid the shuttle effect, while the Gr anode can exclude the interference of lithium metal. By rationally controlling the cycling conditions to suppress the loss of active lithium and the increase in resistance, a SPAN‖Gr pouch cell with 1000 cycles and 99% capacity retention rate can be ultimately obtained. The A h-level pouch cell can stably cycle for 1031 times with 82% capacity retention rate and pass multiple safety tests. This design is expected to fundamentally improve the long-term cycling stability of Li–S pouch cells and it has great potential in the field of large scale energy storage due to its absence of transition metal elements.

---

Broader context

Lithium-ion batteries have been widely applied in the field of energy storage. However, even though the technology has been continuously optimized, it is still difficult for lithium-ion batteries to meet the indicators of high energy density, low cost, and high safety. Therefore, lithium–sulfur (Li–S) batteries with high energy density and lower material costs have received widespread attention. But unfortunately, due to the shuttle effect of polysulfides, lithium dendrites, and numerous side reactions of the lithium metal anode, the cycle life of Li–S batteries is relatively short. The Li–S pouch cell model, which is closest to commercial application, often only has dozens of cycles. In order to improve the cycle life of the Li–S pouch cell, this work designed a pouch cell using sulfurized pyrolyzed poly(acrylonitrile) (SPAN) as the cathode without the shuttle effect and graphite (Gr) as the anode, which excluded the interference of lithium metal. Through the study of the capacity decay mechanism and optimization of the capacity decay factors, a long-term cycling stability SPAN‖Gr pouch cell was ultimately obtained, which provides a new and highly competitive candidate for large scale energy storage.

Introduction

With the gradual increase in energy demand, indicators such as cost, energy, power, cycle life, safety, and environmental compatibility have become the most important concerns for rechargeable batteries.¹ Lithium–sulfur (Li–S) batteries have become one of the promising next-generation batteries due to their lower material cost and higher energy density.² Typical Li–S batteries include a sulfur-based cathode, a lithium metal anode, and an organic electrolyte. During the lithiation process, sulfur undergoes a conversion reaction to form lithium sulfide. The intermediate product polysulfides will dissolve, leading to a continuous decrease in capacity.³ Besides, the volume expansion of the sulfur cathode is large (80%) during lithiation, which will cause damage to the electrode structure. Additionally, the existence of dendrites and side reactions in the lithium metal anode also result in the failure of batteries. The above issues will all lead to a rapid cycle decrease in Li–S batteries.

In order to improve the cycling stability of Li–S batteries, methods including sulfur cathode surface coating, host modification, separator design, and using electrolyte additives have been developed.^4–6 However, most studies focus on lab-scale coin cells, which have a significant gap between current production. In contrast, pouch cells have attracted a lot of attention due to their closer proximity to practical applications.⁷ Compared to coin cells, pouch cells have higher capacity and energy density, but the problems of gas production, thermal runaway, and uneven reaction are more serious. Therefore, many effective improvement strategies in coin cells have little effect in pouch cells.

The Li–S pouch cells have been studied intensively in the past decade, but this area is still at an early stage compared with the research of Li–S batteries. Until now, the energy density of the Li–S pouch cells has been significantly improved, but in sharp contrast, the cycle life of the Li–S pouch cells has not made a substantial breakthrough. According to statistics, it is difficult for the cycle life of Li–S pouch cells with capacity below 0.1 A h to exceed 300 cycles with a capacity retention rate above 80%.^8–15 The same goes for the Li–S pouch cells with capacity between 0.1 A h and 1.0 A h,^16–23 while the pouch cells exceeding 1.0 A h often have only a few dozen cycles (Fig. 1a).^24–31 It is acknowledged that the cycle life of the Li–S pouch cells is far below the commercial standard. Research into improving the cycling stability and cycle life of Li–S pouch cells is accordingly essential and encouraged.

Fig. 1 (a) Cycle number and capacity retention of the Li–S pouch cells with different capacity. (b) Comparison of the energy density between LFP, NCM₈₁₁, and SPAN.

In an effort to develop commercial long-term cycling stability Li–S pouch cells, simply inhibiting the growth of lithium dendrites and side reactions of lithium metal anode, or inhibiting the shuttle effect of polysulfides may not be effective. For the anode, a graphite (Gr) anode is currently the most stable anode and has been commercially applied.^32–35 Replacing the lithium metal anode with a Gr anode could avoid dendrites and side reactions in the lithium metal anode, which will significantly prolong the cycle life of Li–S batteries. For the cathode, a sulfurized pyrolyzed poly(acrylonitrile) (SPAN) cathode is believed to have no shuttle effect of polysulfides in a carbonate electrolyte.^36,37 Remarkably, comparing the SPAN cathode with a commercially applied LiFePO₄ (LFP) cathode and a LiNi_0.8Mn_0.1Co_0.1O₂ (NCM₈₁₁) cathode (Fig. 1b and Table S1, ESI†), the SPAN cathode has the highest theoretical energy density (1184 W h kg⁻¹), and there are no transition metal elements in the SPAN cathode, so its cost is lower, and it is more conducive to the recycling and reutilization of materials. In summary, the combination of a Gr anode and SPAN cathode is expected to fundamentally extend the cycle life of the Li–S pouch cells while ensuring sufficient energy density and lower material cost.

In this work, a SPAN‖Gr pouch cell was constructed by using a Gr anode and SPAN cathode, which could be briefly in situ pre-lithiated through using thin lithium foil. Subsequently, it is found that the main reasons for the capacity decay are the loss of active lithium and the increase in resistance, which are closely related to the generation of dead lithium (dead Li) and the continuous thickening of the solid electrolyte interphase on the anode (SEI) and cathode (CEI). By controlling the cycling conditions, including depth of discharge (DOD), current density, and temperature, SPAN‖Gr pouch cells with nearly zero cycling decay can be obtained. The A h-level SPAN‖Gr pouch cells could cycle stably for 1031 cycles and maintain a capacity retention rate of 82%. Besides, the A h-level SPAN‖Gr pouch cells could pass various safety tests, indicating that this battery system has a certain safety. The design of this battery system is expected to fundamentally improve the cycling stability of the Li–S pouch cells, and it has advantages in material cost, making it a highly promising candidate for future energy storage systems.

Results and discussion

Characteristics of the SPAN‖Gr pouch cell

Since the SPAN‖Gr pouch cell lacks lithium, it needs to be pre-lithiated. Here, a simple in situ lithiation method is adopted, which involves attaching a thin lithium foil to the surface of the Gr anode or the SPAN cathode during the assembly of the pouch cells (Fig. S1, ESI†). When the battery is assembled, the anode or cathode can be spontaneously lithiated. When the thin lithium foil is attached to the Gr anode, the pouch cell should discharge first. During the first discharge, the SPAN cathode undergoes the first lithiation, some Li⁺ are irreversibly consumed in the SPAN cathode, and then it shows an increase in voltage plateau during subsequent cycles (Fig. S2a, ESI†). When the thin lithium foil is attached to the SPAN cathode, the pouch cell should charge first. If there is still residual lithium after lithiation of the SPAN cathode, the charging profile will show more capacity, and the subsequent discharge process will not have a lower voltage plateau or higher irreversible discharge capacity because the SPAN cathode has already been lithiated (Fig. S2b, ESI†). Placing the thin lithium foil on the cathode can ensure more thorough lithiation of SPAN. However, it is worth noting that, taking 1.4 A h pouch cells as examples, attaching a thin lithium foil to the Gr anode produces less heat than that of the SPAN cathode (Fig. S3, ESI†), which might be safer for mass production. Therefore, the placement of a thin lithium foil on which electrode depends on specific requirements. Both methods were adopted in this work. It is worth noting that when the mass loading of the SPAN cathode is 4.5 mg cm⁻², the 20 μm thin lithium foil will just meet the lithiation needs of the SPAN cathode and Gr anode. When the mass loading of the SPAN cathode is in the range of 4 ≤ SPAN < 4.5 mg cm⁻², a small portion of lithium will remain and compensate the active lithium loss during subsequent cycling (Table S2, ESI†).

To investigate the characteristics of the SPAN‖Gr pouch cell, a single-layer stacked pouch cell (about 20 mA h) was assembled (Fig. 2a and Fig. S4, ESI†). According to the rate performance (Fig. 2b), as the rate increases (1C = 1000 mA g⁻¹), the cycling stability of the SPAN‖Gr pouch cell decreases, which indicates that the kinetics of this type of battery at room temperature are not very good. Comparing the capacity decay rates of the SPAN‖Gr pouch cells rested for one month (about 720 h) under different states of charge (SOC), it is found that at 0% SOC, 50% SOC, and 100% SOC, the capacity retention of the SPAN‖Gr pouch cells was 100% (Fig. S5a, ESI†), 99% (Fig. 2c), and 99% (Fig. S5b, ESI†), respectively. Voltage profiles at different temperatures show that the SPAN‖Gr pouch cell can retain 97% of its initial capacity at 0 °C, while only 78% and 60% can be retained at −10 °C and −20 °C, respectively (Fig. S6, ESI†), indicating that the low-temperature performance of this battery still needs to be improved. According to the calculation of the voltage profiles at 0.1C (Fig. S7, ESI†), the energy efficiency of this battery is about 82%. The cycling performance of the SPAN‖Gr pouch cell at 0.5C indicates that this battery system greatly extends the cycle life of the Li–S pouch cell, but there is also a certain degree of capacity decay, with a capacity decay rate of 20% at the 665th cycle (Fig. 2d). Remarkably, the capacity does almost not decay during the rest, so it can be considered that capacity decay is caused by the cycling process of the pouch cell. Ultrasonic non-destructive scanning was conducted on the SPAN‖Gr pouch cells, to convert the peak to peak values of ultrasonic transmission waves into colors ranging from blue to red.³⁸ The ultrasonic transmission image of the long cycle SPAN‖Gr pouch cell at different capacity decay stages indicates that this battery system hardly exhibits gassing or electrolyte drying-out (Fig. 2e and Fig. S8, ESI†). The ultrasonic transmission image of the SPAN‖Gr pouch cells rested for one month at different SOCs presented the same result (Fig. S9, ESI†).

Fig. 2 Characteristics of the SPAN‖Gr pouch cell. (a) Structural schematic of the SPAN‖Gr pouch cell. (b) Rate performance. (c) Voltage profiles of the SPAN‖Gr pouch cell rested for one month at 50% SOC. (d) Cycling performance and corresponding capacity decay stages. (e) The ultrasonic transmission image of the long cycle SPAN‖Gr pouch cell at different capacity decay stages.

Capacity decay mechanism of the SPAN‖Gr pouch cell

To enhance the cycling stability, it is important to first understand the cycle decay mechanism. For the sake of finding out the reasons for the capacity decay, the different stages (0%, 5%, 10%, 15%, and 20%) of capacity decay in the cycled SPAN‖Gr pouch cell are characterized. As can be seen from the differential capacity-potential plots of different capacity decay stages (Fig. 3a), as the capacity decay intensifies, the peak area gradually decreases, and the peak position during the charge process gradually shifts towards higher voltage, while the peak position during the discharge process gradually shifts towards lower voltage. This indicates that during cycling, the active lithium gradually decreases while the polarization also gradually increases.^39,40 The impedance evolution of different capacity decay stages in electrochemical impedance spectroscopy (EIS) shows that the overall impedance of the SPAN‖Gr pouch cell is relatively small (Fig. 3b). The intercept between the high-frequency region and the real axis is considered as the Ohmic impedance (R_Ω), which does not show significant changes at different decay stages.⁴¹ The diagonal of the low-frequency region originates from the solid diffusion process of lithium-ions (Li⁺), and its slope shows no marked change. However, the impedance in the high-frequency to intermediate frequency region increases with the decay of the capacity, which is allocated to the electron/ion transfer resistance caused by SEI/CEI.⁴²

Fig. 3 Capacity decay mechanism of the SPAN‖Gr pouch cell. (a) Differential capacity-potential plot of the long cycle SPAN‖Gr pouch cell. (b) EIS data for the SPAN‖Gr pouch cell. (c) ⁷Li NMR spectrum of delithiated Gr anodes at different decay stages. (d) and (e) XRD patterns of delithiated (d) Gr anodes and (e) SPAN cathodes at different decay stages. (f) FTIR spectra of SPAN cathodes. (g) The capacity loss ratio of Gr anode and SPAN cathode. (h) ICP data for delithiated electrodes.

Then, the SPAN‖Gr pouch cells cycled to different capacity decay stages were disassembled, and the electrode surfaces were observed at delithiated status. It is found that there is no significant change on the surfaces of the cathodes, but different degrees of silver white dead Li appeared on the surfaces of the anodes (Fig. S10, ESI†). The generation of dead Li may be related to the decreased kinetics of Li⁺ deposition and the obstruction of the insertion process. The solid-state nuclear magnetic resonance (NMR) spectroscopy of the delithiated Gr anodes confirmed the existence of dead Li (Fig. 3c). The peak near 265 ppm corresponds to pure Li metal (Li⁰), the peak near 44 ppm corresponds to LiC₆, and the peak near 0 ppm corresponds to Li⁺ from SEI.^43,44 Compared to the decay stage of 0%, the Gr anodes in the subsequent decay stages all showed Li⁰ signals, indicating that after cycling, the Gr anodes will exhibit varying degrees of dead Li. During the discharge process, this part of Li⁰ cannot be removed, which may be related to the increased resistance causing the precipitated lithium to lose electrical contact. On the other hand, compared to 0%, the signal of LiC₆ in other Gr anodes is significantly enhanced, which may be related to the increase in resistance, so that more Li⁺ cannot be stripped and remains intercalated in the Gr anodes when discharging to the same voltage. The production of dead Li leads to a decrease in active lithium, while the increase in LiC₆ indicates that the increased resistance inhibits the intercalation and deintercalation of active Li⁺. These reasons all lead to a continuous decay in the capacity of the SPAN‖Gr pouch cell.

The X-ray diffraction (XRD) patterns of delithiated Gr anodes and SPAN cathodes at different decay stages indicate that there is no significant change in the crystal structure of electrodes during cycling (Fig. 3d and e). And the Fourier-transformed infrared spectroscopy (FTIR) spectra of the SPAN cathodes also indicate that the chemical structure of the SPAN cathodes does not change apparently during cycling (Fig. 3f). The scanning electron microscopy (SEM) images of the delithiated electrodes at different decay stages maintain a similar surface morphology (Fig. S11 and S12, ESI†). These data all confirm that the electrode did not undergo severe damage to its morphology and structure during cycling. In order to compare the difference in capacity loss caused by the anode and cathode, the SPAN‖Gr pouch cells with 0% SOC cycled to different decay stages were disassembled, and the SPAN cathodes were reassembled with lithium metal to form the SPAN‖Li pouch half cells. The difference between the first charging capacity of the half cells and the final capacity of the SPAN‖Gr pouch cells is considered the process capacity loss. The revised discharge capacity of the half cells can be obtained by adding the process capacity loss to the discharge capacity of the half cells. Since the lithium of the half cells is sufficient, the electrolyte is also sufficient, and the fresh lithium anode does not have large impedance from the SEI of long cycle cells, so the difference between the initial capacity and revised discharge capacity is thought to be caused by the capacity loss of the cathode, and the difference between the revised discharge capacity and final capacity is thought to be caused by the capacity loss of the anode (Fig. S13 and Table S3, ESI†). As the capacity continues to decay, the capacity loss caused by the anode gradually becomes dominant (Fig. 3g). The inductively coupled plasma (ICP) data for delithiated electrodes indicate that the lithium content on the anode gradually increases and eventually exceeds that on the cathode (Fig. 3h). This result may be related to the continuous generation of dead Li in Gr anodes.

The thickness variation of the SEI/CEI was observed by high-resolution transmission electron microscopy (HRTEM) (Fig. 4a–d). The distribution of F elements from the SEI/CEI can be observed by the energy-dispersive X-ray spectroscopy (EDS) measurements. When the capacity decays to 5%, the thickness of the SEI on the Gr anode is about 15 nm, but when the capacity decays to 20%, a fragmented and uneven SEI of about 88 nm on the Gr anode can be recognized. As for the SPAN cathodes, when the capacity decay increases from 5% to 20%, the thickness of the CEI increases from 108 nm to 184 nm. The dark field images and corresponding EDS line scan also confirm that the SEI/CEI is constantly thickening (Fig. S14, ESI†). The high-resolution F 1s X-ray photoelectron spectroscopy (XPS) spectra of the delithiated electrodes at different decay stages reveal the evolution pattern of the SEI/CEI (Fig. 4e–h). The prominent signal of 684.8 eV belongs to LiF, and the signal of 686.8 eV belongs to Li_xPO_yF_z.⁴⁵ Using argon ion (Ar⁺) sputtering for different times to obtain the depth profiles, which indicates that LiF is more distributed in the depths of SEI/CEI, and compared to the decay stage of 5%, the relative content of LiF in the SEI/CEI at the decay stage of 20% has increased. The increase in LiF further indicates a decrease in active lithium, and the thickening of the SEI/CEI also means an increase in resistance, and the fragmented SEI/CEI after cycling will exacerbate its own irregular growth.⁴⁶

Fig. 4 Characteristics of SEI/CEI. (a)–(d) HRTEM images of delithiated (a) Gr anode at 5% capacity decay stage, (b) Gr anode at 20% capacity decay stage, (c) SPAN cathode at 5% capacity decay stage, and (d) SPAN cathode at 20% capacity decay stage. (e)–(h) F 1s XPS spectra of delithiated (e) Gr anode at 5% capacity decay stage, (f) Gr anode at 20% capacity decay stage, (g) SPAN cathode at 5% capacity decay stage, and (h) SPAN cathode at 20% capacity decay stage.

The in situ optical fiber sensor derived monitoring technique is used to observe the strain evolution in electrodes during charging and discharging processes for finding out the reasons for SEI/CEI thickening. The stress can be detected by attaching fiber Bragg grating (FBG) sensors onto the surface of the electrodes (Fig. 5a).⁴⁷ When the SPAN‖Gr pouch cell is charged, the stress on the Gr anode gradually increases, while the stress of the SPAN cathode gradually decreases, which may be related to the Li⁺ escaping from the cathode to the anode (Fig. 5b). Notably, the stress change at the SPAN cathode (about 6.3 MPa) is larger than that at the Gr anode (about 2.5 MPa), which further indicates that SPAN undergoes greater volume changes and more severe electrode deformation during the lithiation process.⁴⁸ The cross-sectional morphologic images of the electrodes were obtained via SEM. The DOD decreased from 100% to 0%, the thickness of the Gr anode increased from 97 μm to 106 μm (Fig. 5c and Fig. S15a, ESI†), and the thickness of the SPAN cathode decreased from 111 μm to 99 μm (Fig. 5d and Fig. S15b, ESI†), confirming that the cathode has a greater volume change. During the cycling process, due to the changes in electrode volume, the SEI/CEI are continuously destroyed and self-repaired. If the formed SEI/CEI is not compact and uneven, the increase in thickness will become more severe.

Fig. 5 Volume changes and stress evolution of electrodes in the SPAN‖Gr pouch cell. (a) Schematic of in situ optical fiber sensor derived monitoring. (b) Stress curves of electrodes in the SPAN‖Gr pouch cell and corresponding voltage profile. (c) and (d) SEM cross-sectional images of (c) Gr anodes and (d) SPAN cathodes at 100% DOD and 80% DOD. Scale bars are 50 μm. (e) Schematic of the capacity decay mechanism for the SPAN‖Gr pouch cell.

To sum up, the main reasons for the capacity decay for the SPAN‖Gr pouch cell are the loss of active lithium and the increase in resistance. The loss of active lithium is related to the thickening of the SEI/CEI and the generation of dead Li. Besides, due to the continuous changes in volume of the electrodes during cycling, the SEI/CEI will continuously thicken, and a fragmented SEI/CEI will exacerbate its thickening, which will lead to an increase in resistance. An increase in resistance makes it increasingly difficult for the intercalation and deintercalation of Li⁺. Most Li⁺ are inside the electrodes, resulting in capacity decay (Fig. 5e).

High cycling stability SPAN‖Gr pouch cell

On the basis of understanding the capacity decay mechanism of the SPAN‖Gr pouch cell (about 20 mA h), the cycling stability can be improved by controlling the cycling conditions. For electrodes with significant volume changes, the severe thickening of the SEI/CEI can be avoided by controlling the DOD.⁴⁹ On the other hand, the current density affects the Li⁺ ionic mass transfer rate and the state of the SEI/CEI, and a lower current density can frequently form a compact and more complete SEI/CEI and also reduce the generation of dead Li. Additionally, properly increasing the temperature can enhance the kinetics of Li⁺ transport, thereby reducing the generation of dead Li.⁵⁰

To reduce the impact of significant volume changes, it is reasonable and effective to control the DOD at 80%. The condition of 80% DOD is achieved by controlling the discharge capacity to 80%, while charging is still cut off to 3.0 V. At 80% DOD, the thickness of the Gr anode is 100 μm and the SPAN cathode is 102 μm, which greatly avoids the rapid thickening of the SEI/CEI caused by significant volume changes (Fig. S16 and S17, ESI†). By combining the thickness changes of other different DOD (Fig. S18 and S19, ESI†), it can be found that SPAN cathodes have a greater thickness variation from 80% DOD to 100% DOD. Therefore, cutting off to 80% DOD can effectively relieve the electrode expansion issues, greatly restrain the continuously destroyed and self-repaired CEI, then slower its thickening during cycling. The cycling of the SPAN‖Gr pouch cell at 80% DOD will lead to a decrease in the discharge cut-off voltage (Fig. S20, ESI†). When the cut-off voltage drops to 0.8 V, which is the cut-off voltage of the pouch cell that is cycled at 100% DOD, it can be considered as reaching the capacity decay stage of 20%. The cell was first cycled at 100% DOD for 3 cycles to determine the capacity, and then cycled at 80% DOD starting from the 4th cycle. Therefore, the cycle life of the pouch cell at 80% DOD is nearly three times that of the pouch cell at 100% DOD (Fig. 2d), and it can stably cycle through 1900 cycles (Fig. 6a).

Fig. 6 High cycling stability SPAN‖Gr pouch cells. (a) Cycling performance of the SPAN‖Gr pouch cell at 80% DOD. (b) and (c) Cycling performance of the SPAN‖Gr pouch cells at (b) 0.1C and (c) 0.2C. (d) Cycling performance of the SPAN‖Gr pouch cells at different temperatures.

In terms of current density regulation, when the current density is reduced to 0.1C, the capacity retention rate of the pouch cell is 98% after 500 cycles (Fig. 6b and Fig. S21a, ESI†). Furthermore, when the current density is 0.2C, the pouch cell can stably cycle 500 times, and the capacity retention rate is 99% (Fig. 6c and Fig. S21b, ESI†). The Coulombic efficiency (CE) is close to 100%. Impressively, to the best of our knowledge, this is currently the longest cycle life and highest cycling stability reported for the Li–S pouch cell (Fig. 1a). Encouragingly, when the temperature was increased from 25 °C to 45 °C, the cycling stability of the pouch cell was significantly improved (Fig. 6d and Fig. S21c, ESI†). The rapid decay of the capacity at 25 °C is closely related to the abundant generation of dead Li (Fig. S22, ESI†). After 300 cycles at 1C, a large amount of silver white dead Li appeared on the surface of the delithiated Gr anode (25 °C). In contrast, at 45 °C, due to the optimization of the kinetic characteristics, there was almost no generation of dead Li, resulting in higher cycling stability, and it approached a capacity retention rate of 99% in 1000 cycles at 1C. Therefore, it can be indicated that the SPAN‖Gr battery is a battery system with great potential for long-term cycling, and by optimizing the cycling conditions, it is easy to obtain a battery with high cycling stability.

A h-level SPAN‖Gr pouch cell

The 20 mA h SPAN‖Gr pouch cells have excess electrolyte and a high cycling stability. However, in commercial applications, A h-level pouch cells are needed, and the amount of electrolyte used will be significantly reduced. Therefore, it is necessary to investigate the A h-level SPAN‖Gr pouch cells with lower electrolyte usage. After scaling up production, the A h-level SPAN‖Gr pouch cells have been prepared (Fig. S23, ESI†). When the electrolyte is limited and the temperature is 25 °C, the 1.4 A h SPAN‖Gr pouch cell (Fig. S24a, ESI†) decays rapidly at 100% DOD. After combining the strategy of DOD and current density to control the cycling conditions, the cycling stability is significantly improved. Specifically, after every 100 cycles of 80% DOD, a lower current density (0.05C) at 100% DOD was conducted for 3 cycles. Although the cycling stability was not as good as that of the 20 mA h pouch cells, it could still be stably cycled for thousands of times. At the 1031st cycle, the capacity retention rate was 82% (Fig. 7a and Fig. S25a, ESI†), which is also currently the best cycling performance among A h-level Li–S pouch cells. As for the 2.8 A h SPAN‖Gr pouch cell (Fig. S24b, ESI†), it can also achieve higher cycling stability under the same cycling conditions, with a capacity retention rate of 90% at the 211th cycle (Fig. 7b and Fig. S25b, ESI†). The jump point is caused by the conversion of the current program (Fig. S26, ESI†). And the rate performance of the 1.4 A h SPAN‖Gr pouch cell using LB-015 electrolyte still needs further improvement (Fig. S27, ESI†).

Fig. 7 A h-level SPAN‖Gr pouch cells. (a) and (b) Cycling performance of the (a) 1.4 A h and (b) 2.8 A h SPAN‖Gr pouch cells. (c) Digital photo of smartphone charging by the pouch cells. (d) The 6.1 A h SPAN‖Gr pouch cell. (e) Safety testing of the 1.4 A h SPAN‖Gr pouch cells.

Connecting two 2.8 A h pouch cells in series can charge the smartphone (Fig. 7c). Subsequently, a 6.1 A h pouch cell was assembled to measure its energy density. Through actual measurements, the 6.1 A h pouch cell has an energy of 10.47 W h and a mass of approximately 80.30 g (0.08 kg). Therefore, the specific energy density of this pouch cell is 131 W h kg⁻¹ (Fig. 7d and Fig. S28, ESI†). Remarkably, there is still much room for improvement in energy density. The calculation of energy density based on the theoretical capacity for the 6.1 A h SPAN‖Gr pouch cell can reach over 180 W h kg⁻¹ (Table S4, ESI†). In addition, the mass loading of the electrodes and the specifications of auxiliary materials can be further optimized, which is expected to achieve higher energy density. Last but not least, the high safety of batteries is also indispensable for commercial applications, so the safety tests were conducted on the 1.4 A h pouch cells. Tests include the deep discharge test, the over charge test, the heating test, the short circuit test, and the nail penetration test. Impressively, the A h-level SPAN‖Gr pouch cells have passed all tests (Fig. 7e and Fig. S29, ESI†), which may be related to using a Gr anode instead of a lithium metal anode. Accordingly, it can be considered that the SPAN‖Gr pouch cell also has a certain degree of safety.

Conclusions

In summary, in order to fundamentally improve the long-term cycling stability of the Li–S pouch cells, we assembled the SPAN‖Gr pouch cell, which can avoid the shuttle effect and exclude the interference of lithium metal. It is revealed that the capacity decay mechanism of this pouch cell is the loss of active lithium and the increase in resistance. The continuous thickening of SEI/CEI and the generation of dead Li will lead to a decrease in active lithium. In addition, the thickened SEI/CEI which is related to the changes in electrode volume can also lead to an increase in resistance. Therefore, by controlling the cycling conditions including DOD, current density, and temperature, a SPAN‖Gr pouch cell (20 mA h) with 1000 cycles and 99% capacity retention rate can be successfully obtained. The A h-level pouch cells have not only passed multiple safety tests, but can also cycle stably for over 1000 cycles. This battery system has lower material cost due to its absence of transition metal elements, and it also has exciting cycling stability while ensuring energy density and safety. As a consequence, it may have enormous potential in the application of future large scale energy storage devices.

Methods

Materials

All reagents were commercially available and used as supplied without further purification. Polyacrylonitrile (PAN) (M_w = 150 000 g mol⁻¹) was purchased from Sigma-Aldrich. Sulfur was purchased from Aladdin. LB-015 (1 M LiPF₆ in DEC : EC = 1 : 1 vol% with 5% FEC) electrolyte was purchased from DodoChem. The thin lithium foil (20 μm) and lithium foil (100 μm) were purchased from China Energy Lithium Co., Ltd. The Gr anode was provided from Calb Technology Co., Ltd. The polypropylene (PP) separator (25 μm) was purchased from Canrd Technology Co. Ltd.

Preparation of the SPAN materials

SPAN materials were synthesized by heating a mixture of sulfur and PAN with a mass ratio of 2 : 1 at 350 °C for 6 h with an Ar atmosphere. To remove excess sulfur, the sample was treated again at 300 °C for 6 h. The sulfur content is 40–45 wt% (Table S5, ESI†) determined by element analysis (EA).

Material characterization

XRD patterns were measured by the equipment (Holland, PANalytical X’pert PRO-DY2198) with Cu Kα radiation (λ = 0.15406 nm). FTIR spectra were recorded by a Bruker VERTEX 70 FTIR spectrometer. The HRTEM data were collected through an FEI Talos f200x TEM. XPS data was collected using an AXIS-ULTRA DLD-600 W with Al Kα radiation. The SEM analysis was from an FEI Nova NanoSEM450 microscope operated at 10 kV. The EA data was from a Vario Micro cube element analyzer and the ICP analysis was carried out on an ELAN DRC-e inductively coupled plasma mass spectrometer. The ⁷Li NMR data were obtained from a Bruker Ascend 400WB (9.4 T) spectrometer.

Assembly of the SPAN‖Gr pouch cell and the SPAN‖Li pouch half cell

For the mA h-level SPAN‖Gr pouch cell, the SPAN cathode slurry consists of SPAN powder, LA133 binder and Super P with a mass ratio of 8 : 1 : 1, and then is dried in a 70 °C drying oven. The scale of the SPAN cathode is 3.0 × 3.0 cm, the scale of the thin lithium foil is 3.5 × 3.5 cm, and the scale of the Gr anode is 3.5 × 3.5 cm. The SPAN cathode, PP separator, thin lithium foil, and Gr anode were assembled to obtain the mA h-level SPAN‖Gr pouch cell. The SPAN cathode, PP separator and 100 μm lithium foil (3.5 × 3.5 cm) anode were assembled to obtain the SPAN‖Li pouch half cell. The electrolyte is LB-015 and the ratio of the electrolyte volume to the SPAN mass (E/SPAN) is 15 μl mg⁻¹. The separator of the pouch cell for in situ optical fiber sensor derived monitoring is glass fiber paper (Whatman GF/D, 650 μm), which could avoid crosstalk caused by thickness changes. For the A h-level SPAN‖Gr pouch cell, the scale of the SPAN cathode is 7.7 × 7.8 cm, the scale of the thin lithium foil is 8.1 × 8.1 cm, and the scale of the Gr anode is 8.1 × 8.1 cm. The electrolyte is also LB-015 but E/SPAN = 1.5 μl mg⁻¹. The double-sided coated electrode is prepared, and the assembly method is the same as the mA h-level pouch cell. For the 1.4 A h pouch cell, the layers of cathodes and anodes are 4 and 5, respectively, for the 2.8 A h pouch cell, the layers of cathodes and anodes are 7 and 8, respectively, and for the 6.1 A h pouch cell, the layers of cathodes and anodes are 16 and 17, respectively. The mass loading of all SPAN cathodes is 4–4.5 mg cm⁻² each side, and the negative to positive capacity ratio (N/P) is 1.05. All mA h-level pouch cell assemblies are carried out in an argon-filled glove box, and the A h-level pouch cells assembled in the dry room.

Electrochemical characterization

The differential capacity-potential plots and all cycling performance tests of the pouch cells were collected using a LAND (CT3002A) and NEWARE Battery Test System (MIHW-200-160CH, Shenzhen, China). For the pouch cells cycling at 100% DOD, the cut-off voltage window is 0.8–3.0 V. The calculation of specific capacity is based on the mass of overall SPAN and 1C = 1000 mA g⁻¹. EIS plots were carried out on an electrochemical workstation (CHI660E) and the spectra were recorded with a frequency ranging from 0.01 Hz to 100 kHz.

Author contributions

H. Z., Z. L., and Y. H. conceived and designed the research. H. Z. carried out experiments and measurements. Y. Z. (Yidan Zhang) performed part of the synthesis and electrochemical tests. C. C. and W. Z. conducted safety testing on the pouch cells. K. H. and Y. Z. (Yi Zhang) conducted part of the characterizations. Y. S., Z. L., and Y. H. gave instructive advice to revise the full text and supported the perfection of the manuscript. All authors discussed the results and contributed to the final version of the paper.

Data availability

The data supporting this article have been included as part of the ESI.†

Conflicts of interest

The authors declare no competing interests.

Acknowledgements

This work was supported by the National Key R&D Program of China (2021YFB2400300) and the National Natural Science Foundation of China (5202780089). The authors acknowledge the Analytical and Testing Center of Huazhong University of Science and Technology for the XRD, FTIR, HRTEM, XPS, EA and ICP test, and the State Key Laboratory of Materials Processing and Die & Mould Technology for the SEM measurements. The authors also acknowledge the Wuhan National High Magnetic Field Center (WHMFC) for providing NMR testing services. The Test Center of CALB Group Co., Ltd was acknowledged for providing the safety tests. The NEWARE Battery Test System (MIHW-200-160CH, Shenzhen, China) and LAND (CT3002A) were acknowledged for providing the performance testing of the batteries.

References

1. A. Manthiram, Y. Fu, S.-H. Chung, C. Zu and Y.-S. Su, Chem. Rev., 2014, 114, 11751–11787.
2. P. G. Bruce, S. A. Freunberger, L. J. Hardwick and J.-M. Tarascon, Nat. Mater., 2012, 11, 19–29.
3. Z. W. Seh, Y. Sun, Q. Zhang and Y. Cui, Chem. Soc. Rev., 2016, 45, 5605–5634.
4. Q. Pang, A. Shyamsunder, B. Narayanan, C. Y. Kwok, L. A. Curtiss and L. F. Nazar, Nat. Energy, 2018, 3, 783–791.
5. R. Liu, Z. Wei, L. Peng, L. Zhang, A. Zohar, R. Schoeppner, P. Wang, C. Wan, D. Zhu, H. Liu, Z. Wang, S. H. Tolbert, B. Dunn, Y. Huang, P. Sautet and X. Duan, Nature, 2024, 626, 98–104.
6. W. Yao, J. Xu, L. Ma, X. Lu, D. Luo, J. Qian, L. Zhan, I. Manke, C. Yang, P. Adelhelm and R. Chen, Adv. Mater., 2023, 35, 2212116.
7. S. Dörfler, H. Althues, P. Härtel, T. Abendroth, B. Schumm and S. Kaskel, Joule, 2020, 4, 539–554.
8. J. He, Y. Chen and A. Manthiram, Energy Environ. Sci., 2018, 11, 2560–2568.
9. C. Zhao, A. Daali, I. Hwang, T. Li, X. Huang, D. Robertson, Z. Yang, S. Trask, W. Xu, C.-J. Sun, G.-L. Xu and K. Amine, Angew. Chem., Int. Ed., 2022, 61, e202203466.
10. X. Qi, X. Jin, H. Xu, Y. Pan, F. Yang, Z. Zhu, J. Ji, R. Jiang, H. Du, Y. Ji, D. Yang, L. Qie and Y. Huang, Adv. Mater., 2024, 36, 2310756.
11. S. Nanda, A. Bhargav, Z. Jiang, X. Zhao, Y. Liu and A. Manthiram, Energy Environ. Sci., 2021, 14, 5423–5432.
12. F. S. Hoffmann, F. Schmidt, J. Müller, K. Schönherr, S. Dörfler, T. Abendroth, H. Althues, A. Kwade, N.-L. Wu and S. Kaskel, Batteries Supercaps, 2023, 6, e202300093.
13. Y. He, Z. Chang, S. Wu, Y. Qiao, S. Bai, K. Jiang, P. He and H. Zhou, Adv. Energy Mater., 2018, 8, 1802130.
14. A. Rafie, R. Pai and V. Kalra, J. Mater. Chem. A, 2021, 9, 26976–26988.
15. H.-S. Kang, E. Park, J.-Y. Hwang, H. Kim, D. Aurbach, A. Rosenman and Y.-K. Sun, Adv. Mater. Technol., 2016, 1, 1600052.
16. C. Zhao, G.-L. Xu, T. Zhao and K. Amine, Angew. Chem., Int. Ed., 2020, 59, 17634.
17. L. Luo, S.-H. Chung, H. Y. Asl and A. Manthiram, Adv. Mater., 2018, 30, 1804149.
18. J. Chen, H. Lu, X. Kong, J. Liu, J. Liu, J. Yang, Y. Nuli and J. Wang, Angew. Chem., Int. Ed., 2024, e202317923.
19. A. Hu, W. Chen, X. Du, Y. Hu, T. Lei, H. Wang, L. Xue, Y. Li, H. Sun, Y. Yan, J. Long, C. Shu, J. Zhu, B. Li, X. Wang and J. Xiong, Energy Environ. Sci., 2021, 14, 4115–4124.
20. S. Nanda, A. Bhargav and A. Manthiram, Joule, 2020, 4, 1121–1135.
21. Y. Liu, X. Meng, Z. Wang and J. Qiu, Nat. Commun., 2022, 13, 4415.
22. F. Pei, L. Wu, Y. Zhang, Y. Liao, Q. Kang, Y. Han, H. Zhang, Y. Shen, H. Xu, Z. Li and Y. Huang, Nat. Commun., 2024, 15, 351.
23. B. Shen, B. Sarkodie, L. Zhang, H. Jiang, C. Li and Y. Hu, Energy Storage Mater., 2022, 45, 687–695.
24. W. Xue, Z. Shi, L. Suo, C. Wang, Z. Wang, H. Wang, K. P. So, A. Maurano, D. Yu, Y. Chen, L. Qie, Z. Zhu, G. Xu, J. Kong and J. Li, Nat. Energy, 2019, 4, 374–382.
25. C.-X. Zhao, X.-Y. Li, M. Zhao, Z.-X. Chen, Y.-W. Song, W.-J. Chen, J.-N. Liu, B. Wang, X.-Q. Zhang, C.-M. Chen, B.-Q. Li, J.-Q. Huang and Q. Zhang, J. Am. Chem. Soc., 2021, 143, 19865–19872.
26. X.-Y. Li, S. Feng, C.-X. Zhao, Q. Cheng, Z.-X. Chen, S.-Y. Sun, X. Chen, X.-Q. Zhang, B.-Q. Li, J.-Q. Huang and Q. Zhang, J. Am. Chem. Soc., 2022, 144, 14638–14646.
27. L. Shi, S.-M. Bak, Z. Shadike, C. Wang, C. Niu, P. Northrup, H. Lee, A. Y. Baranovskiy, C. S. Anderson, J. Qin, S. Feng, X. Ren, D. Liu, X.-Q. Yang, F. Gao, D. Lu, J. Xiao and J. Liu, Energy Environ. Sci., 2020, 13, 3620–3632.
28. Y. Guo, Z. Jin, J. Lu, L. Wei, W. Wang, Y. Huang and A. Wang, Energy Environ. Sci., 2023, 16, 5274–5283.
29. C. Zhao, G.-L. Xu, Z. Yu, L. Zhang, I. Hwang, Y.-X. Mo, Y. Ren, L. Cheng, C.-J. Sun, Y. Ren, X. Zuo, J.-T. Li, S.-G. Sun, K. Amine and T. Zhao, Nat. Nanotechnol., 2021, 16, 166–173.
30. Y. Huang, M. Shaibani, M. J. Abedin, D. J. Mendoza, Z. Xu, T. D. Gamot, M. C. D. Cooray, M. Lin, G. Garnier, M. R. Hill and M. Majumder, Adv. Energy Mater., 2022, 12, 2202474.
31. Z. Li, I. Sami, J. Yang, J. Li, R. V. Kumar and M. Chhowalla, Nat. Energy, 2023, 8, 84–93.
32. J. Xu, J. Zhang, T. P. Pollard, Q. Li, S. Tan, S. Hou, H. Wan, F. Chen, H. He, E. Hu, K. Xu, X.-Q. Yang, O. Borodin and C. Wang, Nature, 2023, 614, 694–700.
33. C.-Y. Wang, T. Liu, X.-G. Yang, S. Ge, N. V. Stanley, E. S. Rountree, Y. Leng and B. D. McCarthy, Nature, 2022, 611, 485–490.
34. S. Tu, B. Zhang, Y. Zhang, Z. Chen, X. Wang, R. Zhan, Y. Ou, W. Wang, X. Liu, X. Duan, L. Wang and Y. Sun, Nat. Energy, 2023, 8, 1365–1374.
35. R. Zhang, C. Wang, P. Zou, R. Lin, L. Ma, T. Li, I.-H. Hwang, W. Xu, C. Sun, S. Trask and H. L. Xin, Nat. Energy, 2023, 8, 695–702.
36. J. Wang, J. Yang, J. Xie and N. Xu, Adv. Mater., 2002, 14, 963–965.
37. J. Wang, Y.-S. He and J. Yang, Adv. Mater., 2015, 27, 569–575.
38. Z. Deng, Z. Huang, Y. Shen, Y. Huang, H. Ding, A. Luscombe, M. Johnson, J. E. Harlow, R. Gauthier and J. R. Dahn, Joule, 2020, 4, 2017–2029.
39. S. S. Zhang, Energies, 2014, 7, 4588–4600.
40. M. Dubarry, C. Truchot and B. Y. Liaw, J. Power Sources, 2014, 258, 408–419.
41. L. Xu, Y. Xiao, Y. Yang, S.-J. Yang, X.-R. Chen, R. Xu, Y.-X. Yao, W.-L. Cai, C. Yan, J.-Q. Huang and Q. Zhang, Angew. Chem., Int. Ed., 2022, 61, e20221036.
42. J. Lian, W. Guo and Y. F. Liaw, J. Am. Chem. Soc., 2021, 143, 11063–11071.
43. X. Lin, Y. Shen, Y. Yu and Y. Huang, Adv. Energy Mater., 2024, 14, 2303918.
44. W.-C. Oh, Bull. Korean Chem. Soc., 2001, 22, 367–371.
45. G. Yang, S. Zhang, S. Weng, X. Li, X. Wang, Z. Wang and L. Chen, Nano Lett., 2021, 21, 5316–5323.
46. J. Holoubek, H. Liu, Z. Wu, Y. Yin, X. Xing, G. Cai, S. Yu, H. Zhou, T. A. Pascal, Z. Chen and P. Liu, Nat. Energy, 2021, 6, 303–313.
47. Y. Zhang, X. Xiao, W. Chen, Z. Zhang, W. Li, X. Ge, Y. Li, J. Xiang, Q. Sun, Z. Yan, Y. Yu, H. Yang, Z. Li and Y. Huang, Small, 2024, 2311299.
48. Z. Miao, Y. Li, X. Xiao, Q. Sun, B. He, X. Chen, Y. Liao, Y. Zhang, L. Yuan, Z. Yan, Z. Li and Y. Huang, Energy Environ. Sci., 2022, 15, 2029–2038.
49. V. Chakrapani, F. Rusli, M. A. Filler and P. A. Kohl, J. Power Sources, 2012, 205, 433–438.
50. X.-B. Cheng, R. Zhang, C.-Z. Zhao and Q. Zhang, Chem. Rev., 2017, 117, 10403–10473.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4ee02149e

---