In Situ X-ray Absorption Spectroscopic Investigation of the Capacity

Mar 7, 2019 - ABSTRACT: The Mg/S battery is attractive because of its high theoretical energy density and the abundance of Mg and S on the earth. Howe...
0 downloads 0 Views 741KB Size
Subscriber access provided by Bibliothèque de l'Université Paris-Sud

Communication

In-situ X-ray absorption spectroscopic investigation of the capacity degradation mechanism in Mg/S batteries Yan Xu, Yifan Ye, Shuyang Zhao, Jun Feng, Jia Li, Hao Chen, Ankun Yang, Feifei Shi, Lujie Jia, Yang Wu, Xiaoyun Yu, Per-Anders Glans, Yi Cui, Jinghua Guo, and Yuegang Zhang Nano Lett., Just Accepted Manuscript • DOI: 10.1021/acs.nanolett.8b05208 • Publication Date (Web): 01 Apr 2019 Downloaded from http://pubs.acs.org on April 2, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Nano Letters

Page 1 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

In-situ X-ray absorption spectroscopic investigation of the capacity degradation mechanism in Mg/S batteries Yan Xu,†,§,♣ Yifan Ye,♯,♣ Shuyang Zhao,‖ Jun Feng,♯ Jia Li,‖ Hao Chen,‡ Ankun Yang,‡ Feifei Shi,‡ Lujie Jia,◈ Yang Wu,◈ Xiaoyun Yu,‡ Per-Anders Glans-Suzuki,♯ Yi Cui,‡ Jinghua Guo,♯,* Yuegang Zhang†, ◈,*

†i-Lab,

Suzhou Institute of Nano-Tech and Nano-Bionics, Chinese Academy of Science,

Suzhou, Jiangsu, 215123, China; §School

of Nano-Tech and Nano-Bionics, University of Science and Technology of

China, Hefei, Anhui, 230026, China; ♯Advanced

Light Source, Lawrence Berkeley National Laboratory, Berkeley,

California 94720, United States; ‖Laboratory

for Computational Materials Engineering, Shenzhen Tsinghua University,

Shenzhen, Guangdong 518055, China; ‡Department

of Materials Science and Engineering, Stanford University, Stanford, CA

94305, USA; ◈Department

of Physics, Tsinghua University, Beijing, 100084, China.

ABSTRACT: Mg/S battery is attractive because of its high theoretical energy density and the abundance of Mg and S on the earth. However, its development is hindered by the lack of understanding to the underlying electrochemical reaction mechanism of its charge-discharge processes. Here, using an unique in-situ X-ray absorption spectroscopic tool, we systematically study the reaction pathways of the Mg/S cells in Mg(HMDS)2-AlCl3 electrolyte. We find that the capacity degradation is mainly due to 1

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

the formation of irreversible discharge products, such as MgS and Mg3S8, through a direct electrochemical deposition or a chemical disproportionation of intermediate polysulfide. In light of the fundamental understanding, we propose to use TiS2 as a catalyst to activate the irreversible reaction of low order MgSx and MgS, which results in an increased discharging capacity up to 900 mAh·g-1 and a longer cycling life.

KEYWORDS: Mg-S battery · in-situ XAS · fade mechanism · catalyst

Mg/S battery features a high theoretical energy density (3,200 Wh·l-1), a dendritefree Mg anode,1-4 and a low material cost.5-10 Earlier studies of rechargeable Mg/S batteries have been mainly focused on finding a proper electrolyte that is compatible with both Mg and electrophilic S electrodes.11-14 Hexamethyldisilazide (HMDS)-based electrolyte has been a top performer since its discovery by John Muldoon et al. in 2011,11 though the earlier Mg/S cells based on the HMDS electrolyte have suffered with the problems of overcharging, ultra-short cycle life (< 2 cycles), and low initial discharge voltage (~1 V). These problems are ascribed to the solubility of S and polysulfide in the electrolyte, as revealed by Muldoon’s group using ex-situ X-ray photoelectron spectroscopy (XPS).12 Recently, Manthiram’s group has showed that the shuttle effect of polysulfides can be partially mitigated by introducing pre-activated carbon nanofiber into cathodes using a modified Mg(HDMS)2 electrolyte. However, the problems of high polarization and short cycle life still exist, despite of the efforts dedicated to improve the electrolytes10,15,16 and cathodes14,17,18. The solution for 2

ACS Paragon Plus Environment

Page 2 of 22

Page 3 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

achieving high performance in Li/S batteries do not seem to be readily applicable to the Mg/S system, owing to the different electrochemical behaviors of Mg and Li ions. It is thus necessary to gain a thorough understanding of the specific electrochemical processes in Mg/S batteries. Traditional characterization tools, such as ex-situ X-ray photoelectron spectroscopy (XPS),11,12 ultraviolet-visible spectroscopy (UV-vis),13 are not fully capable of revealing the reaction pathways and the capacity degradation mechanism in Mg/S battery. Here, we employ the synchrotron X-ray absorption spectroscopy (XAS), which can provide the detailed chemical environment and oxidation state information of a specific element in a complex system. Furthermore, an in-situ XAS method can investigate the real-time change of a material’s electronic structure under a close-toreal battery operation condition.19-21

RESULTS AND DISCUSSION As a subject of our in-situ XAS study, Mg/S cell was assembled using the HMDSbased electrolyte. The typical galvanostatic charge-discharge voltage profiles and the cycle stability of the cell are shown in Figure S1a. The first discharge profile is composed of an apparent plateau at around 1.5 V followed by a sharp slop between 1.5 V and 1.0 V and a long slope between 1.0 V and 0.3 V, giving an initial specific capacity of 1080 mAh·g-1 by the weight of S. Unlike the first discharge process, the first charge profile shows no clear voltage plateau. The polarization voltage of the first chargedischarge cycle is as high as 1.5V. In the second discharge profile, the voltage plateau 3

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

becomes undistinguishable, and the specific capacity is significantly reduced to about 400 mAh·g-1. In the following cycles, the cell’s capacity rapidly drops to a merely 200 mAh·g-1 (Figure S1b). To reveal the origin of high polarization and fast capacity degradation of the Mg/S cell, we firstly conducted EDS analysis of the Mg anode retrieved from Mg/S cells after the 1st cycle. As shown in Figure S2, no sulfur signal can be detected, indicating that the high polarization and fast capacity decay in the first two cycles are not due to the shuttle effect of polysulfides, as previously proposed by Aurbach22,23 and Fichtner24. Therefore, we perform the investigation on the evolution of chemical species in the cathode by in-situ and ex-situ XAS experiments. For comparison, MgS8 and MgSx samples were prepared and used as references. The Mg: S ratios of MgS8 and MgSx (x~3) were determined by Energy Dispersive X-Ray Spectroscopy (EDX) (Figure S3 and S4). Figure 1a shows the S K-edge XAS spectra of the reference samples. The preedge feature at 2470.3 eV in both spectra of MgS8 and MgSx (x~3) is assigned to the negatively charged terminal S atoms of the polysulfide chain. The observation of this pre-edge feature is consistent with that of lithium polysulfides Li2Sx.25 The XAS spectrum of MgS8 shows a strong peak at 2472.0 eV, which is assigned to the neutrally charged internal S atoms in S82-. The broad peak near 2475.0 eV can be ascribed to the absorption of negative bivalent S atoms (Figure S5).25 The peak at 2472.8 eV can be assigned to the absorption of the neutral S atoms in Mg3S8. According to our theoretical calculation result, where a blue shift of 0.8 eV from the absorption energy of MgS8 is due to the distinct chemical environment and atomic coordination of S atoms in Mg3S8 4

ACS Paragon Plus Environment

Page 4 of 22

Page 5 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

(Figure 1b). To reveal the details of this difference, we analyzed the extended X-ray absorption fine structure (EXAFS) of Mg3S8 and MgS8 (Figure S6). The Fourier transformed S K-edge spectrum of Mg3S8 exhibits a S-S first shell distance with respect to 1.30 Å, while that of MgS8 exhibits a S-S first shell distance with respect to 1.67 Å. This indicates that the S-S bond in Mg3S8 is 0.37 Å shorter than that in MgS8. A shorter S-S bond means that Mg3S8 is a more stable species, and more energy is needed to excite the S 1s orbital electron, which results in the 0.8 eV blue shift of the XAS peak (Figure 1b). This result is quite different from that in the Li-S system, where the S-S bond lengths are almost identical (2.0 Å) in all lithium polysulfides, 27 suggesting a similar chemical stability of lithium polysulfides. In order to gain the possible structures of charge/discharge products at various charge/discharge stages, density functional theory (DFT) calculations were performed (Figure 1b & Figure S7). We found that the stable structures of MgSx are amorphous, while the most stable structure of MgS is crystalline. The calculated stable structures with the lowest formation energies are depicted in Figure 1b. The amorphous structures of polysulfides (MgxS8, x=1, 2, 3, 8) were searched by using a particle swarm optimization algorithm as implemented in the CALYPSO (Crystal structure AnaLYsis by Particle Swarm Optimization) code.37 The black dots represent the formation energies of possible amorphous structures of magnesium polysulfides and the red dot represents the formation energy of the crystalline MgS. The in-situ XAS measurement results in Figure 2a show the S K-edge spectral evolution of the S cathode during the first discharge process at C/50 with a cutoff 5

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

voltage of 0 V. During the first 6h discharge (corresponding to 201 mAh g-1) the main peak intensity at 2472.0 eV decreased rapidly because of a fast transformation of S8 to MgSx and the formation of polysulfides can also be evidenced by the appearance of the signature pre-edge feature at 2470.3 eV (Figure 2b), which is contributed from the terminal S atoms of polysulfides. Figure 2b also shows a main peak shift from 2472.0 eV to 2472.8 eV and the appearance of the broad peak near 2475.0 eV after the initial 6h discharge. According to the analysis of reference samples (Figure 1a), the blue shift of the main peak can be attributed to the formation of Mg3S8 species. At C/50 rate, 6 hours discharge corresponds to a nominal capacity of 201 mAh·g-1; at such discharging state, MgS8 is the expected discharge product. The appearance of strong XAS feature of Mg3S8 at the early discharging stage can be explained by the disproportionation of unstable high order polysulfide MgS8. In the subsequent discharge stage, however, the XAS peak shapes and intensities keep almost unchanged, which provides a strong evidence for a sluggish solid-state reaction from Mg3S8 to MgS. The in-situ XAS measurement results during the 1st charge process and the 2nd discharge process are shown in Figure 2c and 2d. In contrast to our expectation, the feature at 2472.8 eV does not shift back to 2472.0 eV even in the fully charged state. Moreover, the intensity of this peak at 2472.8 eV only slightly decreases in the 2nd discharge (Figure 2c), implying that Mg3S8 is irreversible species and the discharging capacity from the 2nd cycle is mainly from the sluggish reaction of low order MgSx to MgS. Also, the feature at 2475.0 eV decrease mildly at the early charge state (~340 mAh·g-1) and then keep unchanged which indicate only a small portion of MgS is 6

ACS Paragon Plus Environment

Page 6 of 22

Page 7 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

oxidized to Mg3S8 at early charge state, consistent with the high diffusion barrier of Mg-ion in MgS of 1000 meV by Ceder et al., which further verified that MgS is hard to be oxidized.28 The almost unchanged features at 2472.8 eV and 2475.0 eV from the 2nd cycle suggests the high stability of Mg3S8 and MgS that is responsible for the high polarization and disappearance of charge plateau from the 2nd cycle. The ex-situ XAS spectra of the sulfur cathode sample at different states are shown in Figure S8, which is consistent with the in-situ XAS measurements. The contents of discharging products and unreacted S can be derived from the fitting parameters of spectra as a function of discharging and charging time (Figure 2e). A total discharge capacity of 1116 mAh·g−1 according to the first discharge voltage profile indicates incomplete conversion to MgS. The spectral fitting result also shows the content of S in the charging process keeps constant, indicating that the discharged product can’t be reversed back to elemental S in the charging process. To obtain more information about the evolution of S species in different discharge/charge depth, we also use Mg3S8, MgS8, and S as starting active materials for the cathodes; the 1st discharge profiles of these cathodes are compared in Figure 3a. The MgS8 based cathode displays a typical discharging profile that contains a long discharge plateau around 1.5V, a sharp slope between 1.5V and 1.0V and a slope between 1.0 V and 0.3 V. In contrast, in the Mg3S8 based cathode, there is no plateau at 1.5V. Instead, a very short plateau at 1.1 V and a slope starting from 1.1 V to 0.3 V are observed. The Mg/S batteries using different cathodes were tested without resting in order to avoid the high polarization voltage at the Mg anode. Due to a possible electrolyte wettability 7

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

issue in a starting electrode, the discharge curve of Mg3S8 cathode is shorter than the same discharge plateau in MgS8 and S cathodes. Here, the plateau near 1.5 V corresponds to the transformation of high-order polysulfides MgS8 to MgS4, the sharp slope between 1.5 V and 1.0 V can be related to MgS4 to low order MgSx according to the calculated energy (Table S1), while the sloping region starting from ~1.1 V can be assigned to the reduction of low-order MgSx to MgS. Based on these results, we propose there are three major electrochemical processes in Mg/S cells (Figure 3b): (1) Formation of high order MgSx (MgS8, MgS4) intermediates at a fast reaction rate, because both MgS8 and MgS4 are soluble in the electrolyte. (2) Reduction of high order MgS4 to Mg3S8. (3) Further reduction of low-order Mg3S8 to MgS. The third stage is much more sluggish because the reaction occurred in the solid phase rather than in the liquid phase. While in the charging process, although a small portion of MgS can be oxidized to Mg3S8 at the early charge state, the discharge products Mg3S8 and MgS cannot be reversed back to high order polysulfides or elemental S as revealed in the insitu XAS results. The overcharged capacity in the rest charge process probably comes from the decomposition of electrolyte. As shown in Figure S9, the Mg/S pouch cells become inflated after 2 cycles, which indicate gas generated because of the decomposition of electrolyte. The generated gas species included H2 (m/z=2), C2H6 (m/z=30), H2S (m/z=33) and SO2 (m/z=64 and m/z=48) as revealed by gas chromatography-mass spectrometry (GC-MS) (Figure S10). Above analysis suggests that the irreversible discharge products Mg3S8 and MgS are the major cause for irreversible capacity loss in Mg/S batteries. 8

ACS Paragon Plus Environment

Page 8 of 22

Page 9 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Based on above results and analysis, we conclude that it is a critical to find a way to decompose the inactive MgS and low-order polysulfides in order to tackle the issue of dramatic capacity decay in Mg/S cells. Inspired by previous works in Li/S battery and Mg-ion system,5,29,30,31,32 we choose TiS2 as catalyst to activate the low-order magnesium polysulfides. Figure 4a shows Mg/S cell configuration with the TiS2 catalyst. The cathode was prepared by diffusing 68 μL of 0.014 mg μL−1 S dissolved in tetrahydrofuran into the carbon nanofiber; and Mg foil was used as a counter electrode. In order to obtain a good conductivity, the S: C ratio is at 1:3. The separator was covered by TiS2 with a mass loading of 0.15 mg/cm2. The modified cell has a discharging capacity of 900 mAh·g-1 over 30 cycles at a current density of 83 mA·g-1, which is a significant improvement compared with that without TiS2 (Figure 4b & Figure S1b). In order to figure out capacity contributed by TiS2, Mg/TiS2 cells have been tested. As shown in Figure S11, the discharge capacity of -1

-1

Mg/TiS2 cells is ~20 mAh·g when cycled at 83 mA·g This negligible capacity -1

compared with ~900 mAh·g for Mg/S cells reveals that the discharge capacity shown in Figure 4b is mainly contributed by the redox reaction of S. Ex-situ X-ray photoelectron spectroscopy (XPS) was conducted to verify the effect of using TiS2. Figure S12 shows S 2p spectra of the retrieved sulfur cathode after 5 cycles at fully discharged and charged states. In the fully charged state, the elemental S peak (164 eV) and polysulfide peak (162.2 eV) show up, which indicates that TiS2 is helpful for the oxidation of MgS and Mg3S8 to higher order polysulfides and elemental S. The peak positioned at binding energy of 160.6 eV can be assigned to MgS, which is shown up 9

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

at fully discharged state. Also, a small amount of MgSx can be detected at fully discharged state because of the unreacted MgSx. The interaction between TiS2 and Mg in MgS species favors the decomposition of Mg-S bond (Figure 4c) and consequently lowers the reaction overpotential. The decomposition energy barriers of chemisorbed MgS on the surface of TiS2 and graphene are calculated to be 1.40 eV and 2.28 eV, respectively (See calculation details in Supporting Information Figure S13). Therefore, the reduced energy barrier on TiS2 facilitates the conversion of MgS to low-order MgSx. CONCLUSION The whole reaction process and corresponding discharging products in Mg/S battery during the discharge/charge processes in HMDS-based electrolyte are thoroughly investigated by the in-situ XAS method. In the first discharging process, the conversion of S can be divided into three stages: formation of high order MgSx (MgS8, MgS4) at a fast reaction rate, reduction of MgS4 to Mg3S8, and a sluggish further reduction of Mg3S8 to MgS. Unfortunately, according to the in-situ XAS analysis in the charging process, Mg3S8 and MgS are more electrochemically inert and cannot be reversed back to high order polysulfides or S. These discharge products with slow kinetics are responsible for the rapid capacity decay in the following cycles and greatly shorten the cycling life. By successfully revealing the reaction pathway in Mg/S chemistry by in-situ XAS analysis, we improved the cell performance in Mg/S cell by using TiS2 to active the unreacted MgS and Mg3S8 species. Thus, this study highlights the effective pathway toward great breakthroughs in Mg/S battery.

METHODS 10

ACS Paragon Plus Environment

Page 10 of 22

Page 11 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Electrolyte preparation. Magnesium bis(hexamethyldisilazide) [Mg (HMDS)2, 97%], aluminum chloride (99%) and anhydrous diglyme (DG) were purchased from SigmaAldrich. All the samples were handled in a dry argon-filled glovebox with water and oxygen concentrations below 10 ppm. HMDS-based electrolyte was synthesized according to our previous paper.32 In an argon-filled glove box, Mg(HMDS)2 (517.6 mg, 1.5 mmol) and anhydrous diglyme (3 mL) were added to a 10 mL glass vial, which was then vigorously stirred at 140°C for 1 h. Then AlCl3 (400 mg, 3 mmol) was added with stirring overnight at 90°C to give the 0.5 M [Mg(DG)2][(HMDSAlCl3)]2/DG electrolyte.32 Synthetize MgS8 and MgSx. Treatment of 0.167 g of Mg (6.88mmol), 1.801 g of sulfur (7.02 mmol), and 30 mL of N-MeIm at 95°C for 12 h gave a red solution. Filter the solution by the 45 μm pore size filter. Put the red solution at ambient atmosphere for 48h, red crystal precipitation will be separated out and vacuum dried at 50°C for 3 days. Heating MgS8 at 200°C for 8h, re-dissolving it in N-MeIm, and storing it in glovebox can form MgSx crystal. Preparation of TiS2@separator. The coated separator was prepared by depositing of a layer of TiS2 onto a piece of separator. First, 500 mg TiS2 was dispersed in 100 mL N-methyl-2-pyrrolidone (NMP) and sonicates for 6h. Then the deposition of TiS2 powder was performed by a filtration process. After a vacuum-filtration process, the coated separator was dried at 50℃ under vacuum for 24 h. The areal mass of the TiS2 coated on the separator was controlled at 0.15 mg cm-2. Simulation Methods. Spin-polarized density functional theory calculations were 11

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

performed using the Vienna ab initio simulation package (VASP). The projected augmented wave (PAW) method was used to describe the electron-ion interaction and the generalized gradient approximation (GGA) with the function of Perdew-BurkeErnzerhof (PBE) was used to describe exchange-correlation energy. The cutoff energy of 400 eV was used with a plane-wave basis set. We searched stable and metastable structures of MgxS8 (x=1,2,3,8) with an evolutionary method using a particle swarm optimization (PSO) algorithm as implemented in the CALYPSO code.36 The top 20 structures with relatively low-energy in the CALYPSO structure searches are fully relaxed by using the Vienna Ab initio Simulation Package (VASP).

Two-phase fitting method of XAS spectra by Athena software. We used the spectra of the initial and the fully discharged S cathode as references. The S content of the initial state S cathode was set as 1 and the S content of the fully discharged S cathode was set as 0. The MgS and Mg3S8 content of the initial state S cathode was set as 0 and the MgS and Mg3S8 content of the fully discharged S cathode was set as 1. The changes of S and MgS/Mg3S8 contents of intermediate state cathode were calculated as a function of time based on above setting.33,34 In-situ Mg-S Cell Construction and Electrochemistry Test. The sulfur electrodes were prepared by grinding a mixture containing the commercial sulfur, carbon black (super P) and PVDF (S: C: PVDF = 20:70:10, by weight) for 10 min, then dispersing the mixture in NMP to make a slurry, coating the slurry onto a commercial carbon paper 12

ACS Paragon Plus Environment

Page 12 of 22

Page 13 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

and evaporating the solvent at 50°C overnight. 0.5 M Mg(HMDSAlCl3)2 in diethylene glycol dimethyl ether was used as electrolyte. CR2325 coin cells were assembled in an argon-filled glove box. Electrochemical performance of the cells was evaluated between 0V and 3.0V or 1675 mAh·g-1 capacity control in charging process. X-ray absorption spectroscopy measurements. The in-situ sulfur K-edge XAS spectra were measured at beamline 5.3.1 and 10.3.2 at the Advanced Light Source, Lawrence Berkeley National Laboratory. The X-ray beam size is ~100 μm × 100 μm. The XAS spectra were collected in total fluorescence yield mode and calibrated using elemental sulfur spectra by setting the position of the white line to 2472.0 eV. All the XAS spectra were measured under constant helium flow in the sample chamber and acquired continuously during the discharge/charge processes. The cells used to perform in-situ XAS experiments were adapted from the CR2325 coin cells: R=2 mm round hole was drilled at the sulfur (cathode) side. The hole was then sealed with a 100-nmthick Si3N4 film to make air-proof while allowing X-ray beam penetration (the X-ray transmission ratio of 100 nm thick Si3N4 film at 2470 eV is 97%).35 In addition, we applied carbon paper as the holder for the sulfur electrode materials to allow the direct detection of sulfur species by incident X-rays.

ASSOCIATED CONTENT Supporting Information Available. Voltage profile and cycling stability of Mg/S cells in HMDS-based electrolyte. EDX spectroscopy of MgS8 and MgSx crystal. S K-edge XAS spectra of MgS, Mg3S8, MgS8 and S. EXAFS data of MgS8 and Mg3S8. Ex-situ 13

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

XAS spectra of the different state of sulfur cathode sample. S 2p XPS spectra of the retrieved sulfur cathode after 5 cycles. These materials are available free of charge via the internet at http://pubs.acs.org.

AUTHOR INFORMATION Corresponding Author *E-mail: [email protected] *E-mail: [email protected] Author Contributions ♣ These

authors contributed equally to this work

Notes The authors declare no competing interests. ACKNOWLEDGMENTS This work was supported by the National Key Research and Development Program of China (2016YFB0100100, 2017YFB0701600), the National Natural Science Foundation of China (No. 21433013, U11874036, U1832218), the CAS-DOE Joint Research Program (121E32KYSB20150004), CAS-Queensland Collaborative Science Fund (121E32KYSB20160032). The Local Innovative and Research Teams Project of Guangdong Pearl River Talents Program (Grant No. 2017BT01N111) and Basic Research Project of Shenzhen (JCYJ20170412171430026). S K-edge XAS experiment was performed on BL 5.3.1 at Advanced Light Source and BL 4-3 at Stanford Linear 14

ACS Paragon Plus Environment

Page 14 of 22

Page 15 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Accelerator Center. This research used resources of the Advanced Light Source, which is a DOE Office of Science User Facility under contract no. DE-AC02-05CH11231. JHG was also supported by the Joint Center for Energy Storage Research (JCESR). REFERENCE (1) Yoo, H. D.; Liang, Y.; Dong, H.; Lin, J.; Wang, H.; Liu, Y.; Ma, L.; Wu, T.; Li, Y.; Ru, Q.; Jing, Y.; An, Q.; Zhou, W.; Guo, J.; Lu, J.; Pantelides, S. T.; Qian, X.; Yao, Y. Nat. Commun., 2017, 8 (1), 339-349. (2) Koketsu, T.; Ma, J.; Morgan, B. J.; Body, M.; Legein, C.; Dachraoui, W.; Giannini, M.; Demortière, A.; Salanne, M.; Dardoize, F.; Groult, H.; Borkiewicz, O. J.; Chapman, K. W.; Strasser, P.; Dambournet, D. Nat. Mater., 2017, 16, 1142-1148. (3) Muldoon, J.; Bucur, C. B.; Gregory, T. Chemical Reviews 2014, 114 (23), 1168311720. (4) Aurbach, D.; Gofer, Y.; Lu, Z.; Schechter, A.; Chusid, O.; Gizbar, H.; Cohen, Y.; Ashkenazi, V.; Moshkovich, M.; Turgeman, R.; Levi, E. J. Power Sources, 2001, 9798, 28-32. (5) Zhou, G.; Tian, H.; Jin, Y.; Tao, X.; Liu, B.; Zhang, R.; Seh, Z. W.; Zhuo, D.; Liu, Y.; Sun, J.; Zhao, J.; Zu, C.; Wu, D. S.; Zhang, Q.; Cui, Y. Proc. Natl. Acad. Sci., 2017, 114 (5), 840-845. (6) Du, A.; Zhang, Z.; Qu, H.; Cui, Z.; Qiao, L.; Wang, L.; Chai, J.; Lu, T.; Dong, S.; Dong, T.; Xu, H.; Zhou, X.; Cui, G. Energy Environ. Sci., 2017, 10 (12), 2616-2625. (7) Zhang, Z.; Chen, B.; Xu, H.; Cui, Z.; Dong, S.; Du, A.; Ma, J.; Wang, Q.; Zhou, X.; Cui, G. Adv. Funct. Mater., 2018, 28 (1), 1701718. 15

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(8) Zhang, Z.; Cui, Z.; Qiao, L.; Guan, J.; Xu, H.; Wang, X.; Hu, P.; Du, H.; Li, S.; Zhou, X.; Dong, S.; Liu, Z.; Cui, G.; Chen, L. Adv. Energy Mater., 2017, 7 (11), 1602055. (9) Qiu, Y.; Rong, G.; Yang, J.; Li, G.; Ma, S.; Wang, X.; Pan, Z.; Hou, Y.; Liu, M.; Ye, F.; Li, W.; Seh, Z. W.; Tao, X.; Yao, H.; Liu, N.; Zhang, R.; Zhou, G.; Wang, J.; Fan, S.; Cui, Y.; Zhang, Y. Adv. Energy Mater., 2015, 5 (23), 1501369. (10) Li, W.; Cheng, S.; Wang, J.; Qiu, Y.; Zheng, Z.; Lin, H.; Nanda, S.; Ma, Q.; Xu, Y.; Ye, F.; Liu, M.; Zhou, L.; Zhang, Y. Angew. Chem. Int. Ed., 2016, 55 (22), 64066410. (11) Kim, H. S.; Arthur, T. S.; Allred, G. D.; Zajicek, J.; Newman, J. G.; Rodnyansky, A. E.; Oliver, A. G.; Boggess, W. C.; Muldoon, J. Nat. Commun., 2011, 2, 427. (12) Zhao-Karger, Z.; Zhao, X.; Wang, D.; Diemant, T.; Behm, R. J.; Fichtner, M. Adv. Energy Mater., 2015, 5 (3), 1-9. (13) Yu, X.; Manthiram, A. ACS Energy Lett., 2016, 1 (2), 431-437. (14)Vinayan, B. P.; Zhao-Karger, Z.; Diemant, T.; Chakravadhanula, V. S. K.; Schwarzburger, N. I.; Cambaz, M. A.; Behm, R. J.; Kubel, C.; Fichtner, M. Nanoscale, 2016, 8 (6), 3296-3306. (15) Gao, T.; Noked, M.; Pearse, A. J.; Gillette, E.; Fan, X.; Zhu, Y.; Luo, C.; Suo, L.; Schroeder, M. A.; Xu, K.; Lee, S. B.; Rubloff, G. W.; Wang, C. J. Am. Chem. Soc., 2015, 137 (38), 12388-12393. (16) Gao, T.; Hou, S.; Wang, F.; Ma, Z.; Li, X.; Xu, K.; Wang, C. Angew. Chem. Int. Ed., 2017, 56 (43), 13526-13530. 16

ACS Paragon Plus Environment

Page 16 of 22

Page 17 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

(17) Zhou, X.; Tian, J.; Hu, J.; Li, C. Adv. Mater., 2018, 30 (7), 1704166. (18) Canepa, P.; Sai Gautam, G.; Hannah, D. C.; Malik, R.; Liu, M.; Gallagher, K. G.; Persson, K. A.; Ceder, G. Chemical Reviews 2017, 117, (5), 4287-4341. (19) Robba, A.; Vizintin, A.; Bitenc, J.; Mali, G.; Arčon, I.; Kavčič, M.; Žitnik, M.; Bučar, K.; Aquilanti, G.; Martineau-Corcos, C.; Randon-Vitanova, A.; Dominko, R. Chem. Mater., 2017, 29 (21), 9555-9564. (20) Gao, T.; Ji, X.; Hou, S.; Fan, X.; Li, X.; Yang, C.; Han, F.; Wang, F.; Jiang, J.; Xu, K.; Wang, C. Adv. Mater., 2018, 30 (3), 1704313. (21) Ye, H.; Ma, L.; Zhou, Y.; Wang, L.; Han, N.; Zhao, F.; Deng, J.; Wu, T.; Li, Y.; Lu, J. Proc. Natl Acad. Sci., 2017, 114 (50), 13091-13096. (22) Salama M, Rosy, Attias R, et al. Metal–Sulfur Batteries: Overview and Research Methods. ACS Energy Letters, 2019, 4 (2), 436-446. (23) Salama M, Attias R, Hirsch B, et al. On the Feasibility of Practical Mg-S Batteries: Practical Limitations Associated With Metallic Magnesium Anodes. ACS Energy Letters. 2019, 4, 436-446. (24) Zhao-Karger Z, Liu R, Dai W, et al. Toward Highly Reversible Magnesium–Sulfur Batteries with Efficient and Practical Mg[B(hfip)4]2 Electrolyte. ACS Energy Letters, 2018, 3 (8), 2005-2013. (25) Pascal, T. A.; Wujcik, K. H.; Velasco-Velez, J.; Wu, C.; Teran, A. A.; Kapilashrami, M.; Cabana, J.; Guo, J.; Salmeron, M.; Balsara, N.; Prendergast, D.

J.

Phys. Chem. Lett., 2014, 5 (9), 1547-1551. (26) Farrell, S.; E. Fleet, M.; Stekhin, I.; Kravtsova, A.; Soldatov, A.; Liu, X. Y., Am. 17

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Mineral., 2002; 87, 1321-1332. (27) Dominko, R.; Patel, M. U. M.; Lapornik, V.; Vizintin, A.; Koželj, M.; N. Tušar, N.; Arčon, I.; Stievano, L.; Aquilanti, G. J. Phys. Chem. C, 2015, 119 (33), 1900119010. (28) Canepa, P.; Bo, S.-H.; Sai Gautam, G.; Key, B.; Richards, W. D.; Shi, T.; Tian, Y.; Wang, Y.; Li, J.; Ceder, G. Nature Communications, 2017, 8 (1), 1759. (29) Chung, S.-H.; Luo, L.; Manthiram, A. ACS Energy Lett., 2018, 568-573. (30) Ma, L.; Wei, S.; Zhuang, H. L.; Hendrickson, K. E.; Hennig, R. G.; Archer, L. A. J. Mater. Chem. A, 2015, 3 (39), 19857-19866. (31) Sun, X.; Bonnick, P.; Duffort, V.; Liu, M.; Rong, Z.; Persson, K. A.; Ceder, G.; Nazar, L. F. Energy & Environmental Science, 2016, 9 (7), 2273-2277. (32) Sun, X.; Bonnick, P.; Nazar, L. F. ACS Energy Letters, 2016, 1 (1), 297-301. (33) Xu, Y.; Li, W.; Zhou, G.; Pan, Z.; Zhang, Y. Energy Storage Materials, 2018, 14, 253-257. (34) Zhang, L.; Sun, D.; Feng, J.; Cairns, E. J.; Guo, J. Nano Lett., 2017, 17 (8), 50845091. (35) Zhang, L.; Sun, D.; Kang, J.; Wang, H.-T.; Hsieh, S.-H.; Pong, W.-F.; Bechtel, H. A.; Feng, J.; Wang, L.-W.; Cairns, E. J.; Guo, J. Nano Lett., 2018, 18 (7), 4506-4515. (36) Wang, Y., Lv, J., Zhu, L. & Ma, Y. Comput. Phys. Commun., 2012, 183, 2063– 2070. (37) S. Grimme, J. Comput. Chem. 2006, 27, 1787.

18

ACS Paragon Plus Environment

Page 18 of 22

Page 19 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

TOC graphic

Figure 1. (a) The S K-edge XAS spectra of the MgS8 and MgSx (x~3) reference samples. The preedge feature at 2470.3 eV, identified as the fingerprint of polysulfides, is from the terminal S atoms. The features at 2472.0 and 2472.8 eV are from neutral S atoms in cyclic MgS8 and Mg3S8, respectively. (b) The calculated structures of magnesium polysulfides and their formation energies.

19

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 2. In-situ S K-edge XAS spectra and fitting results from the cathode of Mg/S cell. (a, b) S K-edge XAS spectra collected during the first discharging process of the Mg/S cell at a C/50 rate. (c, d) S K-edge XAS spectra collected during the first charging and second discharging processes of the Mg/S cell at a C/30 rate. (e) Contents of S, MgS and Mg3S8 calculated from two-phase fitting of the XAS spectra as a function of time by Athena software. The values of the contents are obtained directly from the fitting parameters of the two-phase fitting spectra. The black line is the voltage profile of the Mg/S cell from the in-situ experiment.

20

ACS Paragon Plus Environment

Page 20 of 22

Page 21 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Figure 3. Discharge reaction path in Mg/S cell. (a) The 1st cycle discharge profile of Mg3S8, MgS8, and S cathodes, respectively. (b) The voltage profile and the corresponding reaction processes in Mg/S cell.

21

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 4. The catalytic effect of TiS2 in Mg/S cell. (a) Schematic diagram of Mg/S cell with TiS2 coated separator. (b) Cycling performance of the Mg/S cell with TiS2 separator. (c) Schematic diagram of MgS decomposition on TiS2 surface.

22

ACS Paragon Plus Environment

Page 22 of 22