Enhancing the Reversibility of Mg/S Battery Chemistry through Li + Mediation Tao Gao, Malachi Noked, * Alex J Pearse, Eleanor Gillette, Xiulin Fan, Yujie Zhu, Chao Luo, Liumin Suo, Marshall A Schroeder, Kang Xu, Sang Bok Lee, * Gary W. Rubloff, * Chunsheng Wang * Department of Chemical and Bimolecular Engineering, Department of Material Science and Engineering, Department of Chemistry and Biochemistry, University of Maryland, College Park,College Park, MD 20740, USA Electrochemistry Branch, Power and Energy Division Sensor and Electron Devices Directorate, U.S. Army Research Laboratory, Adelphi, 20783, USA S1
Figure S1. SEM images of the morphology of deposits prepared by holding Mg foil at -1.2V vs Mg RE in 0.1 M Mg-HDMS + 1.0 M LiTFSI electrolyte for 1 hour. Deposition Plasticine Support intensity Li Mg Li 6.46 Mg 30 32 34 36 38 40 2-Theta Figure S2. XRD of deposits. Pt foil was used as the deposition substrate in this test to avoid confusion in XRD results analysis. Deposit was prepared by holding Pt foil at -1.2V vs Mg RE in 0.1 M Mg- HDMS+1.0 M LiTFSI electrolyte for 1 hour. The deposit is a very thin layer on Pt surface and it exfoliates from Pt easily. Thus, the XRD sample was prepared by placing the exfoliated deposits onto a plasticine support. The black pattern here refers to the XRD of plasticine support instead of the Pt foil. Before XRD tests, the sample was thoroughly washed with TEGDME to remove any residual electrolyte components. There is no detectable Li crystal, Li 6.46 Mg crystal or Li compound like Li 2 O or LiOH. S2
a) Survey O 1s b) Mg 1s Intensity (a.u.) Mg 1s F 1s/KLL Mg KLL Intensity (a.u.) O KLL C 1s Mg 2p Cl 2p Pt 4f 1400 1200 1000 800 600 400 200 0 1308 1305 1302 1299 1296 c) Li 1s/Mg 2p Binding Energy (ev) Binding Energy (ev) Intensity (a.u.) Location of Li 1s Mg 2p 66 64 62 60 58 56 54 52 50 48 Binding Energy (ev) Figure S3. XPS of deposits on a Pt foil prepared at a potential of -0.5 V vs. Mg RE (the potential of Mg anode in the Mg/S cell, see Figure S5.) The deposition took the form of small domains of visual deposits under the experimental conditions, so a small spot size (110 microns) was used to reduce signal from the substrate. (a) shows a survey spectrum of the deposit, showing the presence of Mg as well as other elements associated with oxidized Mg and incorporation of components of the electrolyte and salt anions. (b) is a higher resolution scan of the Mg 1s peak (20 scans) and (c) shows the region containing both the Mg 2p and the Li 1s (20 scans). No Li 1s peak is clearly identifiable above the noise of the measurement. Based on the area of the Mg 1s peak, the signal-to-noise ratio in the Li 1s region, and the photoelectron cross sections of Mg and Li, the ratio Mg:Li in the deposit must be >10. S3
a) b) S4
intensity ACC/Sulfur 20 40 60 80 2-Theta sulfur powder c) Figure S4. a) SEM and EDS of the ACC/S composite electrode showing clear dispersion of sulfur to the interiors part of the fiber b)tga of the ACC/S composite, showing 15% sulphur content. c) XRD of the ACC/S cathode. No peaks attributed to crystalline sulfur were detected, indicating that sulfur is in a highly dispersed state inside the pores. S5
Potential (V) 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0-0.5 S Mg Full Cell 0 50000 100000 150000 200000 TestTime(s) -0.1 max Over potential (V) -0.2-0.3-0.4-0.5-0.6 0 2 4 6 8 10 12 14 16 18 Cycle # Figure S5. a) First cycle potential of sulfur cathode, Mg anode vs. Mg RE and full cell potential in 0.2 M Mg-HMDS + 0.5 M LiTFSI electrolyte and b) Overpotential for Mg deposition during cycling S6
Figure S6: XPS of Mg foil aged in TEGDME/sulfur, clearly demonstrating MgS layer on the Mg surface. transmitance Li2S8-TEGDME S8-TEGDME-LiTFSI Mg-S8-TEGDME-LiTFSI 1600 1400 1200 1000 800 Wavenumber (cm -1 ) a) S7
70000 60000 50000 Intensity 40000 30000 20000 10000 0 Mg-S8-TEGDME-LiTFSI S8-TEGDME-LiTFSI Li2S8-TEGDME 100 200 300 400 500 600 Raman Shift (cm -1 ) b) Figure S7. a) FT-IR and b) Raman Spectroscopy of Li-PS solution (Li 2 S 8 -TEGDME) and Mg-PS solution (Mg/S 8 -TEGDME-LiTFSI). The image of Li-PS and Mg-PS are shown in Figure S8 and Figure S9. We cannot discriminate Li-PS from Mg-PS by FT-IR since their corresponding peaks are overlapping. Raman Spectroscopy is not helpful since the Raman peaks of Mg-PS solution (black line) are dominated by dissolved sulfur (blue line). We cannot discriminate the discharged product from XRD due to no peaks corresponding to crystalline sulfur can be detected in the ACC/carbon composite (Figure S4c). Li + S in TEGDME 12 hours Figure S8. Image of 100 mm Li 2 S 8 solution prepared by soaking Li metal in a sulfur powder dispersion in TEDGME and stirring for 12 hours. S8
Mg + S in TEGDME Three weeks Mg + S + LiTFSI in TEGDME 12 hours Mg + S + LiTFSI in TEGDME 54 hours a) b) c) Figure S9. Images of a) Mg + sulfur powder in TEGDME (atomic ratio of added Mg/S is 8:1, corresponding to MgS 8 ); b) Mg + sulfur powder + 100 mm LiTFSI in TEGDME for 12 hours; c) Mg + sulfur powder + 100 mm LiTFSI in TEGDME for 54 hours. Table S1 Theoretical Capacity and Energy Density Comparison of different Battery Systems Full Cell system Cathode Anode Full Cell Theoretical Capacity Voltage (V) Theoretical Capacity Voltage (V) Capacity (mah/g) (mah/g) (mah/g) Full Cell Energy (Wh/kg) LiCoO 2 /graphite 140 4 370 0.1 102 396 Li-S 1675 2.1 3828 0 1165 2447 Li 2 S/Silicon 1165 2.1 4211 0.1 913 1825 Li 2 S/graphite 1165 2.1 372 0.1 454 908 Mg/Mo 6 S 8 128 1.2 2233 0 121 145 Mg/S 1675 1.8 2233 0 957 1722 Mg/S in this study _ 874 For a Mg/S battery, complete sulfur reduction corresponds to reaction below. Theoretical Full cell capactiy (only consider the mass of cathode and anode, assume all cathode and anode participate reaction) is calculated by: S9
= = 2 26800 h 24 =957 h/ 32 Theoretical Full cell energy density is calculated by: = =957 h 1.8 =1722 h/ The theoretical capacities and energy densities of other battery system are calculated using the same procedure. In our Mg/S system, as demonstrated Figure 1, two plateaus at 1.75 V and 1.2V last for ~600 mah/g and ~400 mah/g, which correspond to 19200 mah/mol-sulfur and 12800 mah/molsulfur. Thus the real energy density is calculated below. = 19200 h 1.75 12800 h 1.2 24 32 =874 h/ As can be seen, the Mg/S battery here obtains about half of the theoretical energy density of Mg/S system. Note the mass of inactive components(binder, conductive carbon, seperator and electrolyte, cell package, etc) will furthur lower this value, however, it is beyond the scope of this study. The calculation shows that Mg/S system has the potential to replace LiCoO 2 /graphite chemistry in terms of obtainable energy density. S10