Improved-performance lithium–sulfur batteries modified by magnetron sputtering

Abstract

The cathode material S/AC for lithium–sulfur batteries was synthesized with elemental sulfur as the active material and activated carbon (AC) as the conductive matrix. Al and Ti were respectively deposited onto the surface of S/AC electrodes by the method of radio-frequency magnetron sputtering to modify the electrodes and improve the battery performance. The properties of S/AC and the sputtered cathode materials, labeled S/AC/Ti and S/AC/Al, were characterized by XRD and FESEM. Electrochemical performances of the Li/S batteries with the three cathode materials were determined by alternating-current impedance, cyclic voltammetry (CV) and constant-current charge and discharge. Experiments showed that S/AC, S/AC/Ti and S/AC/Al delivered initial specific capacity of 1197 mA h g⁻¹, 1255 mA h g⁻¹ and 1257 mA h g⁻¹ respectively under the current rate of 0.5C. And the modified batteries operated reversibly over 100 cycles and maintained a discharge specific capacity of 722 mA h g⁻¹ and 977 mA h g⁻¹ after 100 cycles, superior to 634 mA h g⁻¹ of S/AC. Besides, the coulombic efficiencies of the sputtered electrodes were over 0.97 after 100 cycles.

---

1. Introduction

In recent years significant progress has been achieved in the research into rechargeable lithium–sulfur (Li–S) batteries which are promising candidates for power sources for electronic communication and transportation. The theoretical specific capacity of a Li–S battery is 1675 mA h g⁻¹ which is more than 5 times that of lithium-ion batteries based on traditional insertion compound cathodes.^1–4 Despite the advantages of elemental sulfur – abundance, low price and nontoxicity, there are several main problems plaguing Li–S batteries^5–7 including (1) poor electronic conductivity of sulfur, (2) the shuttle effect associated with dissolution of intermediate polysulfides into electrolyte, and (3) large volumetric expansion of the redox reaction products, which result in rapid capacity decay and low coulombic efficiency. Pioneering work has been conducted to address these problems, during which various strategies have been made to enhance the electrical conductivity of sulfur which involve combining active materials with various carbon materials,^8–11 encapsulating sulfur particles with conducting materials^12,13 such as graphene oxide,^14–16 and so on. Recently, Moreno,¹⁷ Liu¹⁸ and coworkers utilized activated carbons derived from olive stones and wheat straw activated with KOH as a conductive matrix for elemental sulfur. Moreno obtained an initial capacity of about 670 mA h g⁻¹ meanwhile Cheng reported that the first and the 200th specific discharge capacity of 582 mA h g⁻¹ and 445 mA h g⁻¹ are retained at 1C. Raja et al.¹⁹ have employed sisal as biomass precursors to develop activated carbons for lithium–sulfur batteries which delivered an initial discharge capacity of 1050 mA h g⁻¹ at 0.05C-rate but suffered from rapid capacity fading and low coulombic efficiency.

Lots of research work concentrates on chemical approaches for modification of lithium–sulfur batteries, nevertheless the reports on the application of physical methods are quite scanty. In our work, lithium–sulfur batteries were modified by magnetron sputtering, an effective physical vaporous deposition method, and a stable cycle performance was achieved. Cathode material S/AC was synthesized with commercial AC (pore diameter 2.0–2.2 nm, specific surface area 1800 m² g⁻¹, JCNANO, Nanjing) as the conductive matrix. The large specific surface makes AC a suitable carrier for elemental sulfur, which can restrain S and intermediate polysulfides in the interspace, improve the absorption of electrolyte and provide adequate room for the products of redox reaction. Then the S/AC electrodes were coated with “thin films” by magnetron sputtering, which may be the key to improve the conductivity and mitigate the shuttle effect, consequently leading to improved cycle performance. Magnetron sputtering has been widely used in deposition of metal and metalloid materials attributing to the advantages as low deposition temperature, good compactness and remarkable adhesion between thin films and substrates.^20,21 At present magnetron sputtering has been a mature technology with moderate cost and environmental friendliness, which could be expected to be employed for innovation of lithium sulfur batteries and show a new prospect of interdisciplinarity. In this experiment, Al and Ti were deposited onto the surface of S/AC electrodes to develop composites S/AC/Al and S/AC/Ti as the cathode materials for Li–S batteries (Fig. 1). Comparison between the batteries with the three cathode materials was conducted in order to analyze the feasibility of improvement for Li–S batteries based on magnetron sputtering and lay the foundation for all-solid-state lithium sulfur batteries.^22–25

Fig. 1 Structure of the modified Li–S battery.

2. Experimental details

2.1 Preparation of S/AC cathode material

AC and elemental S were mixed with mass ratio of 1 : 4, then grinded adequately in agate mortar. The mixture was heated in a two-step way in a vacuum tubular furnace under flowing argon atmosphere. Firstly, it was heated to 155 °C and kept for 6 h to make the sulfur melt fully and spread into the carbon interspace. Secondly, it was raised to 300 °C and kept for 1 h, so that the residual elemental sulfur on the AC surface could be partly removed. Finally, the mixture was cooled to room temperature in the furnace, and the S–C composite powder was obtained. The composite was mixed with acetylene black (AB) as conductive additive, polyvinylidene fluoride (PVDF) as adhesive and N-methyl-2-pyrrolidone (NMP) as solvent to prepare S/AC cathode material. The mass ratio of the S–C composite, AB and PVDF was 7 : 2 : 1. The mixture was fully grinded to black slurry, then evenly coated onto the current collector of Al foil with the scraper. Subsequently, the Al foil was fully dried at 60 °C for 6 h. After being pressed by twin screw roller press, the Al foil coated with S/AC cathode material was placed into the vacuum chamber of the magnetron sputtering apparatus.

2.2 Magnetron sputtering deposition of Al and Ti

Magnetron sputtering deposition was performed on the ultrahigh vacuum magnetron sputtering equipment (FJL560B1, SKY Technology Development Co., Ltd. Chinese Academy of Sciences). Al and Ti were deposited onto the Al foil coated with S/AC by radio-frequency magnetron sputtering with pure Al and Ti targets (purity 99.99%) at room temperature. The Al foil coated with S/AC was employed as the substrate. The base pressure of sputtering chamber was 1.0 × 10⁻⁴ Pa and the deposition was carried out under Ar atmosphere (purity 99.99%) with the flow rate of 20 mL min⁻¹ measured accurately by the gas mass flowmeter (D08-3B/ZM, Beijing Sevenstar Huachuang Electronics Co. Ltd.). Pre-sputtering of the Al and Ti targets was conducted for 15 min before deposition in order to remove the impurities on the surface of the targets. The sputtering pressure was 1.0 Pa, power 25 W, and sputtering time 9 min. The sample holders rotated at a constant speed in order to obtain homogeneous distribution of Al and Ti particles on the surface of substrates. After the sputtering was accomplished, the substrates were taken out and cut into circular electrode slices with diameter of 14 mm. The slices were labeled as S/AC, S/AC/Al and S/AC/Ti respectively and placed in a vacuum oven of 50 °C.

2.3 Battery assembly

CR-2032 button cells were assembled in dry glove box (volume fractions of H₂O and O₂ less than 0.1 × 10⁻⁶, MBRAVN LABMAS-TER 130) filled with Ar, S/AC, S/AC/Al and S/AC/Ti as working electrodes respectively, lithium as counter electrode, Celgrad 2400 membrane as separator and 1 mol L⁻¹ lithium bis(trifluoromethanesulphonyl)imide (LiTFSI)/1,2-dimethoxyethane (DME) + 1,3-dioxolane (DOL) (volume ratio 1 : 1) as electrolyte.

2.4 Performance testing

X-Ray diffractometer (X'-pert PROPA Nalytical, Netherlands) was employed to analyze the structure of sublimed S, AC and three cathode materials with Cu Kα source (λ = 1.542 Å). The working voltage was 40 kV and the operating current was 40 mA, scanning continuously from 10° to 90° with the rate of 8° min⁻¹. Field emission scanning electron microscope (FESEM Hitachi SU8020) were used to characterize the morphology of cathode materials and the distribution of S, C, Al and Ti. Cyclic voltammetry and alternating-current impedance were obtained on electrochemical workstation (CHI604D, Shanghai Chenhua). The scanning voltage was ranged from 1.5 V to 3.0 V and scanning speed was 0.1 mV s⁻¹. The alternating-current impedance frequency was ranged from 10⁻² Hz to 10⁵ Hz with the amplitude of ±5 mV. Constant current charge–discharge curves were performed on battery performance testing system (NEWARE BTS3000).

3. Results and analysis

Fig. 2 is the XRD patterns of AC, elemental S and three kinds of cathode materials. The amorphous AC show low and broad peaks at about 21° and 43°. Elemental sulfur presents (222), (026) and (040) multicrystal peaks at about 23.1°, 25.7° and 27.3°, manifesting the structure of orthorhombic system. Three cathode materials all show the diffractive peaks of S at identical angles, which means that after being grinded and heated, elemental S still exists in single phase. However, the diffractive peak intensity of S were weakened and broadened which signifies that elemental S was partly absorbed into the interspace of AC. Moreover the diffractive peaks of Al and Ti certify that the particles of Al and Ti were just coated onto the substrates instead of reacting to them.

Fig. 2 XRD patterns of AC, S and cathode materials, (a) S, AC and S/AC, (b) S/AC/Al and S/AC/Ti.

Fig. 3 is FESEM patterns of three cathode materials. The conductive matrix AC has abundant interspace and large specific surface area, which will absorb sulfur and polysulfides. After sputtering, there are some small tablets and blocks of sulfur appear on the surface of electrode, showed in Fig. 3(d) and (h), which is probably due to the sublimation and re-condensation of sulfur caused by the temperature rising as a result of the collision between sputtering particles and the cathode materials.^26,27 Both Al and Ti particles may improve the conductivity of the electrodes and fill in the interspace of AC to hinder the dissolution and diffusion of polysulfides, thus alleviated the shuttle effect to a certain degree. As the sputtering rate of Al is greater than that of Ti,²⁷ there should to be more particles of Al on the surface of the electrode than Ti, thereby superior performances of S/AC/Al could be expected.

Fig. 3 FESEM patterns of S/AC, S/AC/Al and S/AC/Ti, (a) morphology of S/AC, (b) distribution of S in S/AC, (c) distribution of C in S/AC, (d) morphology of S/AC/Ti, (e) distribution of S in S/AC/Ti, (f) distribution of C in S/AC/Ti, (g) distribution of Ti in S/AC/Ti, (h) morphology of S/AC/Al, (i) distribution of S in S/AC/Al, (j) distribution of C in S/AC/Al, (k) distribution of Al in S/AC/Al.

Fig. 4 demonstrates the electrochemical impedance spectroscopy before charge–discharge cycles. Three samples all show typical characteristics of Li–S batteries, exhibiting semicircles in the high frequency region and inclined lines at low frequencies. The first point of intersection of the semicircle and abscissa axis denotes intrinsic resistance (R_in) of the battery, which is probably determined by ionic resistance of electrolyte, electronic resistance of active materials and contact resistance at the interface of active material and current collector.²⁸ The similar R_in of three samples reflects the approximately equal dynamics resistance of the electrochemical reaction took place at the interface of electrode and electrolyte.²⁹ However the semicircle radius of S/AC/Al is the shortest, revealing the least charge transfer resistance (R_ct) among them. The R_ct may attribute to the structure of the electrodes and the presence of Al and Ti on the surface which improves the conductivity effectively. Taking the resistivity of Al (2.83 × 10⁻⁸ Ω m) and Ti (4.20 × 10⁻⁷ Ω m) into consideration, the conductivity of S/AC/Al is the best among the three. The slopes of the inclined short lines of S/AC/Al and S/AC/Ti in low frequency region are both close to 1, standing for the analogous diffusing ability of lithium ions in the solid phase of them,³⁰ superior to that of S/AC. Overall, S/AC/Al presents the best impedance characteristic.

Fig. 4 EIS of S/AC, S/AC/Al and S/AC/Ti.

The results of CV tests are presented in Fig. 5. For the three samples, two reduction peaks of CV curves appear at about 2.3 V and 2.0 V respectively. The former is corresponding to the process in which elemental sulfur is reduced to soluble polysulfides with long chains (Li₂S_n, n ≥ 4), controlled by chemical reaction mechanism, and the latter is corresponding to the diffusion-controlled process in which polysulfides with long chains are further reduced to insoluble polysulfides (Li₂S₂ and Li₂S) with short chains (Li₂S_n, n < 4). The oxidation peak is corresponding to the oxidation processes, in which insoluble Li₂S₂ and Li₂S are firstly oxidized to polysulfides with long chains and further to elemental sulfur. The reduction current of S/AC at about 2.0 V is the least in the three batteries and the peak position moves to the left, revealing a low reductive reaction speed. Furthermore, the low current, wide shape and the asymmetry of oxidation peak of S/AC indicate serious polarization and bad electrochemical reversibility.³¹ In comparison, the CV curve of S/AC/Al exhibits the greatest peak currents and sharpest shapes of the oxidation and reduction peaks, showing the least polarization and best electrochemical properties in three batteries.³² The results of EIS and CV tests illustrate that the deposition of Al and Ti on the surface of electrodes can effectively improve the conductivity, reduce the polarization and retard shuttle effect to some extent.

Fig. 5 CV profiles of S/AC, S/AC/Al and S/AC/Ti.

The 1st, 100th and 200th charge and discharge curves of the three batteries at different current rates are displayed in Fig. 6. There are two platforms corresponding to the two reduction peaks of the CV curve. The platforms in high voltage region occur at about 2.3–2.4 V and the low ones appear in the region of 1.9–2.1 V. More than 65% of specific capacity is provided by the low voltage platform. It has been proved that the specific capacity is directly related to the utilization of active materials. The discharge platforms at about 2.4 V of S/AC are the most unconspicuous at 0.5C and 1C current rates, which relates to dissolution and diffusion of polysulfides resulting in the wastage of active materials and shuttle effect.³³ Furthermore, the reaction between polysulfides and lithium will also lead to irreversible loss of capacity and coulombic efficiency decreasing, which coincides with the following analysis about the curve of coulombic efficiency. At the current rate of 0.5C, three samples show initial specific capacities of 1197 mA h g⁻¹, 1255 mA h g⁻¹ and 1257 mA h g⁻¹ severally, and 634 mA h g⁻¹, 722 mA h g⁻¹ and 977 mA h g⁻¹ at the 100th discharging. The result reveals that the utilization of sulfur in S/AC/Al is the highest in three cells due to the presence of Al thin films. At 1C rate, S/AC/Al provides a specific capacity of 1171 mA h g⁻¹ in the 1st discharging and 818 mA h g⁻¹ in the 100th cycle. The cycle performance at 0.5C is better than that at 1C because lower circulating rate leads to less polarization of the electrodes and fewer side-reactions between electrolyte and electrode materials.³⁴ The voltage difference between charge and discharge platforms of S/AC is the greatest among them which is caused by polarization.³⁵ For S/AC/Al, the discharging platforms are much wider and plumper than the other two, indicating the best conductivity, least polarization and highest energy density.³⁶ Therefore, S/AC/Al is considered to be superior to the other two samples in terms of reversible specific capacity and cycling stability.

Fig. 6 The charge and discharge characteristics of S/AC, S/AC/Al and S/AC/Ti, (a) 0.5C, (b) 1C.

Fig. 7 illustrates the coulombic efficiency and specific discharge capacity profiles of three samples at 0.5C and 1C rate. Coulombic efficiency is an important indicator to measure the severity of shuttle effect. As shown in Fig. 7, irregular fluctuations in the curves of coulombic efficiency are resulted from shuttle effect. Moreover, the polysulfides produced in cycles dissolve and diffuse in electrolyte and react with lithium anode. As the products of irreversible discharging, Li₂S₂ and Li₂S deposit on the surface of lithium and form a passive film. In each cycle, the self-discharging and generating of “dead lithium” on anode will lead to decrement of the utilization of sulfur and irreversible loss of capacity.³⁷ Three batteries all keep the coulombic efficiency over 0.96 in 200 cycles at 0.5C and 1C respectively. Furthermore, Fig. 7(c) reflects that S/AC/Al has considerable superiority of stability and reversibility of electrochemical performance owing to the presence of Al on the surface of S/AC substrate. Despite that Al and Ti can both improve the electrical conductivity of the electrode, S/AC/Al may benefit more from the sputtering than S/AC/Ti in consideration of greater specific conductance of Al compared to Ti. Moreover, the sputtering rate of Al is higher than that of Ti under the same sputtering conditions, thus more Al particles were coated on the surface of the substrates and lead to a distinct improvement of the electrical conductivity, which may be the main reason for the best performance of S/AC/Al.

Fig. 7 Specific capacity and coulombic efficiency curves of S/AC, S/AC/Al and S/AC/Ti, (a) 0.5C, (b) 1C, (c) variable rates.

4. Conclusions

S/AC composite was synthesized with AC as the conductive matrix. Al and Ti particles were deposited respectively on the surface of S/AC electrodes by the method of radio-frequency magnetron sputtering to enhance the electrical conductivity of Li–S battery, and stable cycles were achieved as a result. The button cell with S/AC/Al as cathode material exhibits the best performance with the specific capacity of 1257 mA h g⁻¹ and 977 mA h g⁻¹ in the 1st and 100th discharging at the current rate of 0.5C and coulombic efficiency over 0.97 even after 200 cycles. The results show that it is an effective method to coat Al on the electrodes by magnetron sputtering for modification of Li–S battery.

Acknowledgements

The authors gratefully acknowledge the support of the “Strategic Priority Research Program” of the Chinese Academy of Science (Grant No. XDA03040000), and the Earth-Panda Advance Magnetic Material Co. Ltd Fund (Grant no. 13-332).

References

1. M. Wang, H. Zhang and Y. Zhang, J. Solid State Electrochem., 2013, 17, 2243.
2. P. Bruce, S. Freunberger and L. Hardwick, Nat. Mater., 2012, 11, 19.
3. S. Evers and L. Nazar, Acc. Chem. Res., 2013, 46, 1135.
4. A. Fotouhi, D. Auger and K. Propp, Renewable Sustainable Energy Rev., 2016, 56, 1008.
5. L. Li, L. Li and X. Guo, J. Solid State Electrochem., 2013, 17, 115.
6. B. Ding, C. Yuan and L. Shen, Chem.–Eur. J., 2013, 19, 1013.
7. S. Evers, T. Yim and L. Nazar, J. Phys. Chem. C, 2012, 116, 19653.
8. H. Nersisyan, S. Joo and B. Yoo, Carbon, 2016, 103, 255.
9. J. Qu, S. Lv and X. Peng, J. Alloys Compd., 2016, 671, 17.
10. S. Seo, C. Choi and D. Kim, Mater. Lett., 2016, 172, 119.
11. Z. Tang, X. Fei and Y. Liang, New Carbon Mater., 2015, 95, 1083.
12. L. Li, X. Liu and K. Zhu, J. Solid State Electrochem., 2015, 19, 1.
13. L. Smith, P. Malati and J. Fang, Mater. Horiz., 2015, 3, 137.
14. D. Wang, D. Xie and T. Yang, J. Power Sources, 2016, 313, 233.
15. H. Lee, Y. Kim and S. Jung, J. Nanosci. Nanotechnol., 2016, 16, 2692.
16. Y. Zhang, Y. Zhao and Z. Bakenov, Ionics, 2014, 20, 1047.
17. N. Moreno, A. Caballero and L. Hernán, Carbon, 2014, 70, 241.
18. J. Liu, Y. Cheng and S. Ji, RSC Adv., 2015, 5, 100089.
19. M. Raja, N. Angulakshmi and A. Stephan, RSC Adv., 2016, 6, 13772.
20. Y. Huang, H. Li and M. Zuo, J. Magn. Magn. Mater., 2016, 409, 39.
21. L. Tao, H. Li and J. Shen, J. Magn. Magn. Mater., 2015, 375, 124.
22. Z. Hu, Modified Solid State Electrolyte Membrane for Lithium/Sulfur batteries, National University of Defense Technology, Changsha, 2011.
23. F. Han, Y. Jie and X. Fan, Nano Lett., 2016, 16, 4521.
24. H. Nagata and Y. Chikusa, J. Power Sources, 2016, 329, 268.
25. H. Marceau, C. Kim and A. Paolella, J. Power Sources, 2016, 319, 247.
26. X. Hao, Equipment for Electronic Products Manufacturing, 2013, vol. 220, p. 57.
27. M. Tian, Thin film technologies and materials, Tsinghua University press, Beijing, 2006, p. 484.
28. M. Wang, H. Zhang and Y. Zhang, J. Solid State Electrochem., 2013, 17, 2243.
29. J. Brückner, S. Thieme and H. Grossmann, J. Power Sources, 2014, 268, 82.
30. Z. Deng, Z. Zhang and Y. Lai, J. Electrochem. Soc., 2013, 160, A553.
31. S. Chung and A. Manthiram, Electrochim. Acta, 2013, 107, 569.
32. Y. Mikhaylik, J. Akridge and N. White, Solid State Ionics, 2004, 175, 243.
33. Z. Zhao, D. Qin and S. Wang, Electrochim. Acta, 2014, 127, 123.
34. S. Zheng, F. Yi and Z. Li, Adv. Funct. Mater., 2014, 24, 4156.
35. S. Xiong, X. Kai and X. Hong, J. Natl. Univ. Def. Technol., 2012, 34, 150.
36. J. Chen, W. Feng and R. Chen, Carbon, 2014, 70, 428.
37. Z. Liu, S. Bertolini and P. Balbuena, ACS Appl. Mater. Interfaces, 2016, 8, 4700.

---