Supporting Information

Size: px
Start display at page:

Download "Supporting Information"

Transcription

1 Supporting Information Garnet electrolyte with an ultra-low interfacial resistance for Li-metal batteries Yutao Li, Xi Chen, Andrei Dolocan, Zhiming Cui, Sen Xin, Leigang Xue, Henghui Xu, Kyusung Park, and John B. Goodenough 1 * Materials Science and Engineering Program and Texas Materials Institute, The University of Texas at Austin, Austin, TX 78712, USA *Corresponding Author: jgoodenough@mail.utexas.edu S1

2 EXPERIMENTAL Garnet preparation and characterization. The preparation of garnet Li 6.5 La 3 Zr 1.5 Ta 0.5 O 12 (LLZT) was the same as was reported in our previous paper 19. The fresh and aged LLZT pellets were covered with carbon and fired at 700 or 850 o C for 10 h in Ar atmosphere. Powder X-ray diffraction (Philips PW1830, Cu Kα) was employed to monitor the phase formation in the 2θ range of 10 to 70. A field-emission scanning electron microscope (Quanta 650) was used to obtain the microstructure of the garnet and the surface of metallic anode. Ionic conductivity was measured with an Auto Lab workstation with a applied frequency range from 10 6 to 1 Hz; the Au electrode was sputtered on the garnet surface for the Li-ion conductivity measurement. The conductivity has been calculated based on the surface area and the thickness of the pellet. A thick pellet of 3 mm was used for the Li-ion conductivity testing to improve the accuracy; a pellet with a thickness of 300 μm was used in all the battery testing. Cyclic voltammetry was carried out on an Auto Lab workstation at a scan rate of 1 mv s -1. Thermogravimetric analysis (TGA), differential thermal analysis (DTA) and mass spectrometric analyses were conducted on the aged garnet powders and aged garnet powders mixed with carbon in Ar atmosphere from room temperature to 1100 o C. Surface chemical states of garnet were characterized with X-ray photoelectron spectroscopy (XPS; Kratos AXIS Ultra DLD). Raman spectroscopy was carried out with a WITec Alpha 300 Raman microscope instrument (WITec, Germany) equipped with a 532 nm wavelength laser for excitation. Symmetric Li/garnet/Li cell: The lithium metal size is 0.3 cm -2 for all the battery preparation and testing. Two lithium metal foils were directly put on the both sides of the garnet pellets, and one nickel mesh was put on the top of lithium foil to prepare the 2032 coin cell. The obtained Li/garnet/Li cell was tested at room temperature and 65 o C at a current density of 100 μa cm -2. S2

3 TOF-SIMS analysis: For depth profiling and chemical analysis, we used a TOF.SIMS 5 instrument (ION-TOF GmbH, Germany, 2010). During depth profiling the sputtering ion beam (Cs + at 2keV ion energy, 70 na measured sample current and 1 μm beam size) was raster scanned typically over an area of μm 2. The analysis ion beam consisting of Bi 1 + pulses (30 kev ion energy, 20 ns pulse duration, 3.8 pa measured sample current) was set in the high current bunched (HC) mode and raster scanned over a μm 2 area centered within the Cs+ sputtered area at the regressing surface. The depth profiles were acquired in noninterlaced mode, that is, sequential sputtering and analysis, at a base pressure of 10 9 Torr. All mass spectra were acquired in negative polarity while the mass resolution was >7000 (m/δm) for all fragments of interest. The sputtering rate was calculated at 0.12 nm/s by measuring the depth of a crater formed in the garnet (~22 µm) after 50 hours of Cs+ sputtering with a Wyko NT 9100 Optical Profilometer (Supplementary Fig. 10). For sample transfer we used an in-house designed air-free capsule to transfer the sputtered samples to and from an argon-filled glovebox in vacuum or argon environment, respectively. For high-resolution imaging we employ the burst alignment mode with bursts, which can achieve a lateral resolution of <100 nm while providing a mass resolution of >5000 (m/δm). The local corrugation was directly visualized with an Asylum Research MFP-3D atomic force microscope (Supplementary Fig. 12) All-solid-state battery: The LiFePO 4 cathode was prepared with the same method as was described in our previous paper 14. The active material LiFePO 4, carbon black, cross-linked polyethylene oxide (CPEO), and LiTFSI with a weight ratio of 60:12:20:8 were dispersed in DMF and stirred overnight. The LiFePO 4 cathode with a loading of 2 to 3 mg cm -2 and a surface area of 0.3 cm -2 was dried at 90 o C for 12 h under vacuum before use type coin cells were fabricated with S3

4 lithium foil and garnet as the anode and the electrolyte in an Ar-filled glove box. The all-solid-state Li/garnet/LiFePO 4 battery was tested at 65 o C with an Arbin station between the 2.5 and 3.8 V. The preparation of all-solid-state Li/LiNi 0.8 Co 0.1 Mn 0.1 O 2 battery was the same as the Li/LiFePO 4 battery. Hybrid Li-S Battery: The cathode preparation of Li-S battery was the same as our previously report 35. The hybrid lithium-sulfur batteries were assembled in 2032-type coin cells with a polysulfide-impregnated ACNF paper (surface area 0.15 cm -2 ) as the cathode, 1.0 M LiCF 3 SO 3 salt in a mixture of 1, 2-dimethoxyethane (DME) and 1,3-dioxolane (DOL) (1 : 1 in volume) as the liquid electrolyte, and lithium-metal disk with a surface area of 0.3 cm -2 as the anode. The polysulfide catholyte was prepared by dissolving stoichiometric sulfur powder and Li 2 S into the above liquid electrolyte at 70 C overnight to give 1.5 M sulfur in the form of 0.1 M Li 2 S 6 clear solution. ~ 11 µl of the as-prepared Li 2 S 6 solution was then dropped onto the ACNF paper, corresponding to a sulfur content of 50 % and a sulfur loading of 1 mg cm -2. Then, a solid-state LLZT pellet with a diameter of 12 mm and a thickness of 0.3 mm was placed on top of the cathode, followed by a piece of TF40 membrane (NKK) soaked with the liquid electrolyte to prepare Li/organic electrolyte/garnet/organic electrolyte/s batteries. For the Li/garnet/organic electrolyte/s batteries, the lithium foil was directly attached on the surface of garnet electrolyte. The assembled hybrid Li-S batteries were galvanostatically charged and discharged with an Arbin station between 1.8 and 2.85 V. All the capacities were calculated based on the mass of sulfur in the catholyte. Electrochemical impedance spectroscopy (EIS) data were collected with an Auto Lab workstation at the open-circuit voltage. S4

5 Supplementary Figure 1. The electrochemical impedance plots of the fresh and aged LLZT pellets. S5

6 Supplementary Figure 2. The TGA, DSC and mass spectrometric analyses of the LLZT powders aged in air for three months. S6

7 Supplementary Figure 3. a Photographs of the LLZT pellets before and after reaction with the surface-coated carbon at 700 o C in Ar atmosphere and b the corresponding XRD patterns c XRD pattern of aged garnet with carbon sintered at 700 o C; Li 2 O with a diffraction peak at 2θ = 33.5 o was observed as a reaction product.. S7

8 Supplementary Figure 4. a The electrochemical impedance plot and b the temperature dependence of Li + -ion conductivities of the LLZT-C pellet. S8

9 Supplementary Figure 5. a Raman spectrum of the LLZT pellet fired at 850 o C for 6 h in Argon and b Charge/discharge voltage curves of Li/LLZT-treated at 850 o C/Li cell, which died in 40 hours at 100 µa cm -2 at 65 o C. S9

10 Supplementary Figure 6. a The electrochemical impedance plot of LLZT-C with symmetric Li electrodes at 25 o C. b Charge/discharge voltage curves of Li/LLZT/Li cell died in several hours at 50 µa cm -2. c Charge/discharge voltage profiles Li/LLZT-C/Li at 25 o C at 100 µa cm -2 d Charing Li/LLZT-C/Li cell at 50 µa cm -2 for 150 h at 65 o C. S10

11 Supplementary Figure 7. Charge and discharge voltage profiles of a Li/LLZT-C/Li at 0.4 ma cm -2 at 65 o C and b Li/CPEO/garnet-C/CPEO/Li at 1 ma cm -2 at 65 o C. The voltage fluctuation in Figure a is caused by the temperature change of oven. S11

12 Supplementary Figure 8. The surface a and b-c cross-section of the LLZT-C pellet after cycling the Li/LLZT-C/Li for 450 h. d-e SEM image of LZT-C pellet after cycling the Li/LLZT-C/Li for 450 h, no lithium metal was found inside the LLZT-C pellet. S12

13 Supplementary Figure 9. a SEM images of surface of LLZT-C(scale bar: 8 µm). b Long range roughness of LLZT-C pellet. The LLZT-C pellet has a rough surface to improve the contact area with metallic lithium; it has a R p (the height difference between the mean line and the highest point) and R b (the height difference between the mean line and the lowest point) value of 1.75 and 2.45 µm, respectively. c-d SEM images of Li metal after cycling the symmetric Li/LLZT-C/Li cell for 450 h (scale bar: 8 µm) S13

14 Supplementary Figure 10. Cyclic voltammogram of a Li/LLZT-C/Au and b Li/organic-electrolyte/LLZT+carbon (or carbon) at 1 mv s -1. S14

15 Supplementary Figure 11. Sputtering for ~50 hours with a Cs + ion beam at 2keV ion energy, 70 na measured sample current, 1 μm beam size and raster scanned over an area of μm 2 produces a crater depth of ~22 µm, which translates into a sputtering rate of ~0.12 nm/s. The crater measurement was performed with a Wyko NT 9100 Optical Profilometer. S15

16 Supplementary Figure 12. TOF-SIMS cross-sectional mapping of a LLZT C pellet after cycling in a Li/LLZT C/Li cell for 450 h. The Li wetting layer is visible at the surface while penetrating several tens of microns in the cycled LLZT-C. S16

17 Supplementary Figure 13. 2D a and 3D b atomic force microscopy (AFM) maps of the cycled LLZT-C surface showing the local RMS corrugation is larger than 100 nm. Peak-to-peak corrugation is >1 µm. S17

18 Supplementary Figure 14. a Comparative CO 3 mass spectra integrated over the first 40 s of Cs + sputtering (~5 nm). The CO 3 removal following the carbon treatment is evident. b The LiOH signal at the garnet surface is dropping by ~25% after the carbon treatment. S18

19 Supplementary Figure 15. Charge and discharge voltage profiles of the all-solid-state Li/LLZT/LiFePO 4 cell at 100 μa cm -2. S19

20 Supplementary Figure 16. a The electrochemical impedance plots of Li/LLZT-C/LiNi 0.8 Co 0.1 Mn 0.1 O 2 cell and b charge and discharge voltage profiles of the cell at 100 μa cm -2. S20

21 Supplementary Figure 17. Electrochemical characterization of Li/organic-electrolyte/LLZT/organic -electrolyte/s cell. a the electrochemical impedance plot. b charge/discharge voltage profiles at 200 μa cm -2, and c capacity retention and cycling efficiency. Electrochemical characterization of Li/organic-electrolyte/LLZT-C/organic-electrolyte/S cell. d the electrochemical impedance plot. e charge/discharge voltage profiles. f capacity retention and cycling efficiency. S21

22 Supplementary Figure 18. Cycling stability and Coulombic efficiency of the Li-S cell without solid electrolyte. S22