Supporting Information. for. Visualization of Electrode-Electrolyte Interfaces in LiPF 6 /EC/DEC Electrolyte for Lithium Ion Batteries via In-Situ TEM

Transcription

1 Supporting Information for Visualization of Electrode-Electrolyte Interfaces in LiPF 6 /EC/DEC Electrolyte for Lithium Ion Batteries via In-Situ TEM Zhiyuan Zeng 1, Wen-I Liang 1,2, Hong-Gang Liao, 1 Huolin L. Xin 1, Yin-Hao Chu 2, and Haimei Zheng 1,3* 1 Materials Sciences Division, Lawrence Berkeley National Laboratory, Berkeley, CA 94720, USA 2 Department of Materials Science and Engineering, National Chiao Tung University, Hsinchu 30010, Taiwan 3 Department of Materials Science and Engineering, University of California, Berkeley, California 94720, United States * Correspondence should be addressed to: hmzheng@lbl.gov This file includes: Materials and Methods Figures S1- S5 Movies S1 and S2 Movie captions 1

2 Materials and Methods The schematic of biasing liquid cells we developed is shown in Figure 1, which is also reported in our previous work 1. These cells were fabricated at the Marvell Nanofabrication Laboratory of the University of California at Berkeley. The cells including bottom and top chips are fabricated using ultra-thin silicon wafers (200 µm, 4- inches, p-doped) purchased from Virginia Semiconductor (Fredericksburg, VA). A 35 nm thick low stress silicon nitride film was evaporated on the silicon wafer as a membrane of the viewing window. The viewing windows and two reservoirs were created by photolithographic patterning and etching with KOH solution (with water to KOH ratio of 2:1). The dimensions of the windows are 25 µm 6 µm. Two 120 nm-thick gold electrodes were deposited on the bottom chips with a face-to-face distance of 20 µm. The bottom and top chips were sealed by 150 nm thick sputtered indium spacer. The commercial electrolyte for lithium ion batteries, i.e. 1M lithium hexafluorophosphase LiPF 6 dissolved in 1:1 (v/v) ethylene carbonate (EC) and diethyl carbonate (DEC), was loaded into the reservoirs with a syringe. Liquid electrolyte flows into the viewing window by capillary force. We seal the cell including the reservoirs using Cu foil and epoxy. Since the window gap is no larger than 150 nm, no contamination from epoxy was observation during electrochemical experiments. The dimensions of a biasing cell are ~3 mm 3 mm square and ~400 µm thick. Both of the working and counter electrodes were extended to two gold pads in two reservoirs. Gold wires were bonded onto each gold pad. The bonded gold wires were connected to the two copper pads on the TEM holder tip. The TEM holder fits a JEOL 2100 TEM operated at 200 kv. Real-time videos of the electrochemical experiments were recorded with 2 ~ 15 frames per second. Faster recording up to 400 frames per second was also incorporated using a high speed camera provided by Direct Electron, LP (San Diego, CA) with a model number of DE-12. The electrochemical process was controlled by an electrochemical workstation (Model 660D series made by CH Instruments) connected with the TEM holder. It was used to perform in-situ cyclic voltammetry measurement. For ex-situ experiments, we first put a drop of Au nanowire solution (used as purchased from Nanopartz Inc.) onto the lacey carbon Cu grids and let the Cu grids dry out so that lots of Au nanowires were attach to the carbon film. Then, the TEM grids were immersed into the lithium electrolyte with short face-to-face distance. The two Cu girds were connected with an electrochemical working station and a cyclic voltammetry with the voltage range of 0 ~ -3 V and a scan rate 0.1 V/s were applied to the two Cu grids. (1.) Estimate the total lithium in electrolyte initially: The total volume in liquid cell is V = L 2 h, where L is the side length of liquid cell with length of 3 mm, h is the height and is equal to 150 nm. The concentration (c) of LiPF 6 in EC/DEC is 1 mol/l. Suppose all the lithium ions are reduced to Li metal, the total generated lithium metal in weight is W = V c M w, where M w is the molecular weight of Li and is equal to 6.94 g/mol. Then the calculated result shows that depletion of all lithium ions in liquid cell can deposit Li metal of g. (2.) Estimate the maximum thickness of lithium layer: In the extreme condition, suppose that all lithium ions in the electrolyte are electrodeposited on the Au electrodes with a uniform Li metal film covering Au 2

3 electrodes, the total metal film after the depletion of lithium ions can be estimated. The area of Au patterns exposed to electrolyte in the liquid cell is about S = 2(L 1 W 1 + L 2 W 2 ), where L 1, L 2, W 1 and W 2 are 1020 µm, 60 µm, 20 µm and 15 µm and S is calculated to be cm 2. Therefore, if all Li in the electrolyte is deposited on the electrodes, we can get Li metal film of 413 nm covering all Au electrode patterns. In this case, we think the existing Li source in battery electrolyte might be enough to sustain the SEI formation, Li dendrite growth and Li-Au alloy process. Figure S1. Cyclic voltammetry with a potential range of 4 ~ 0 V applied on the electrodes. Figure S2. Ex situ TEM results were obtained after the cyclic voltammetry (0 ~ -3V) was applied on two gold TEM grids immersed in the same lithium electrolyte in air. (A) TEM image of the reaction products. (B) The corresponding Selected Area Electron Diffraction Pattern (SAED) showing lithium oxides were obtained. It suggests that lithium metal must be achieved initially during charge-discharge cycles and lithium can be 3

4 subsequently oxidized in air. Figure S3. Ex situ TEM study of gold nanowires before and after the charge cycles in the same lithium electrolyte. (A) TEM image of gold nanowires before lithiation. (B) A magnified TEM image of the nanowire in (A). (C) the corresponding Selected Area Electron Diffraction Pattern (SAED) of Au nanowire before lithiation in (B); (D) TEM image of lithiated Au nanowires. (E) A magnified TEM image of a nanowire in (D). (F) The corresponding SAED of the nanowire in (E). Lithiation of gold nanowires led to the expansion and twists of nanowires and the Li-Au alloy formation. 4

5 Figure S4. Ex situ TEM study of gold nanowires after lithiation. (A) High Angle Annular Dark Field (HAADF) image of a lithiated Au nanowire. Bending and twisting of the nanowire were observed. (B) A higher magnification view (HAADF image) of a lithiated Au nanowire. The middle section is the heavily lithiated with some lithium enriched dots. (C) Selected Area Electron Diffraction Pattern (SAED) of lithiated Au nanowire corresponding to a Au 3 Li phase. (D) High resolution spectroscopic image of lithiated Au nanowire with a L1 2 ordered Li-Au structure. (001) plane is oriented parallel to the electron beam with alternating planes of Au 2 and Au 1 Li 1 as shown in the model structure (Au, yellow; Li, Blue); (E) HAADF image of a lithiated Au nanowire; The electron energy loss spectroscopic (EELS) line profile was recorded along the dash line; (F) Extracted lithium and gold concentration along the line profile in (E), the maximum intensity was normalized to 1. Lithium concentration was integrated from the Li K-edge (56-65 ev) (Au O edge contribution in the energy range is negligible. Additional information was included in Figure S5. Gold concentration was estimated from the ADF signals. 5

6 Figure S5. Comparison of the single scattering spectra of LiF and Au. Reference spectra were obtained from EELS atlas (CC Ahn, OL Krivanek 1983). Here Li K edge has a large cross section and a good near edge jump ratio, however Au O edge has a poor jump ratio and the cross section is an order of magnitute smaller. Movie Captions Movie S1: Lithiation of Au electrode under cyclic voltammetry with an applied voltage range of 0 ~ -3 V. It shows volume expansion, lithium dendritic growth and dissolution on the reacted area of Li-Au alloy. Movie S2: Solid electrolyte interface on Au electrode under cyclic voltammetry with an applied potential of 4 ~ 0 V. (1.) Sun M. H.; Liao H. G.; Niu K. Y.; Zheng H. M. Structural and Morphological Evolution of Lead Dendrites during Electrochemical Migration, Sci. Rep. 3, 3227 (2013). 6