7Li NMR Chemical Shift Imaging To Detect Microstructural Growth of Lithium in All-Solid-State Batteries

All-solid-state batteries potentially offer safe, high-energy-density electrochemical energy storage, yet are plagued with issues surrounding Li microstructural growth and subsequent cell death. We use 7Li NMR chemical shift imaging and electron microscopy to track Li microstructural growth in the garnet-type solid electrolyte, Li6.5La3Zr1.5Ta0.5O12. Here, we follow the early stages of Li microstructural growth during galvanostatic cycling, from the formation of Li on the electrode surface to dendritic Li connecting both electrodes in symmetrical cells, and correlate these changes with alterations observed in the voltage profiles during cycling and impedance measurements. During these experiments, we observe transformations at both the stripping and plating interfaces, indicating heterogeneities in both Li removal and deposition. At low current densities, 7Li magnetic resonance imaging detects the formation of Li microstructures in cells before short-circuits are observed and allows changes in the electrochemical profiles to be rationalized.


Electrochemical measurements.
The voltage profile of the short-circuited Li-LLZTO-Li cell discussed in Figure 1b is shown below (Figure S2). The cell was first cycled at a current density of 0.2 mA•cm -2 up to a capacity of Intensity / a.u.

S3
1.0 mAh•cm -2 for 3 cycles, then at a current density of 0.5 mA•cm -2 up to a capacity of 1.0 mAh•cm -2 for 3 cycles, followed by 3 cycles at a current density of 1.0 mA•cm -2 up to a capacity of 1.0 mAh•cm -2 and finally, again, at a current density of 0.5 mA•cm -2 for one discharge until short-circuiting occurred. Due to the increased thickness of this pellet compared to all other pellets used (3.95 mm), the cell lasted for several cycles above 0.5 mA•cm -2 . Nevertheless, the voltage curves at 1.0 mA•cm -2 do show significant fluctuations in cell voltage, which suggest extensive dendrite growth is occurring at 1.0 mA•cm -2 , but that these dendrites have not yet fully reached the other electrode to cause a short circuit due to the increased thickness of the electrolyte. However, applying a current density of 0.5 mA•cm -2 after cycling at 1.0 mA•cm -2 instantaneously led to short circuit of the cell.

S4
The relationship between LLZTO thickness and the time the cell takes to short circuit at 0.5 mA•cm -2 is shown in Figure S3. The thicker the LLZTO pellet, the longer the cell can be discharged at a current density of 0.5 mA·cm -2 before short circuit. Figure S2. Voltage profile of the short circuited Li-LLZTO-Li symmetrical cell discussed in Figure 1c of the main text cycled at 0.2 mA·cm -2 up to a capacity of 1.0 mAh·cm -2 for 3 cycles (black), followed by 0.5 mA·cm -2 up to a capacity of 1.0 mAh·cm -2 for 3 cycles (blue), 1.0 mA·cm -2 up to a capacity of 1.0 mAh·cm -2 for 3 cycles (green) and finally at 0.5 mA·cm -2 for one discharge only until short circuit (red).    Table S2. EIS analysis of the Li-LLZTO-Li cells in Figure 3 of the main text at various time points during their cycling, where the resistances, R1-R3 (bulk, grain boundary and interface, respectively) and capacitances C1-C3, for the three constant phase elements are defined in Figure  S4.

S8
An equivalent circuit composed of a series of three resistors and constant-phase elements in parallel to account for the bulk (R1), grain boundary (R2), and interface (R3) resistances present within the cell was used to fit the data (see schematic below) The capacitance values for each resistance are labeled as C1, C2, and C3, respectively. The total resistance of the cell (Rtotal) for each time point is also calculated.
Significant increases in the interface resistance (R3) are labelled in red in Table S2. It should be noted that low frequency behaviors in Nyquist plots such as the interface resistance are much harder to fit than the bulk and grain boundary resistances, at least for our data, and therefore should not be regarded as absolute. Additionally, from our EIS analysis and voltage profiles of the cells in Figure 3, it is only just before cell failure that major changes in the EIS or cell voltage are observed. No information on the initial growth of microstructural Li can be extracted. Hence, 7 Li CSI could, in the long run, be more useful in providing a new dimension to study dendrite growth and cell failure, especially at visualizing Li metal growth at the very early stages of cycling.
All Nyquist plots obtained during cycling of cells in Figure 3f and 3g of the main text are shown in Figure S5. These plots show alternating higher and lower total impedances depending Figure S4. Schematic of the equivalent circuit composed of a series of three resistors and constantphase elements in parallel to account for the bulk (R1), grain boundary (R2) and interface (R3) resistances present within the cell was used to fit EIS data.
S9 on the current direction for the cell in Figure 3f, and a clear decrease in total impedance for the cell in Figure 3g as current is passed. Both Nyquist plots after 600 minutes and 720 minutes for the cell in Figure 3g suggest that a short-circuit has occurred.  S10 7 Li chemical shift imaging. Figure S6 illustrates the effect of orientation on the shift of the Li metal signal. If the electrodes were oriented parallel to B0, the resolution between the dendrites and the Li metal electrodes was significantly worse than the perpendicular orientation ( Figure S6). For each sample, single pulse 7 Li NMR spectra were collected with a recycle delay of 1 s and 2048 scans. Typical 90° radiofrequency (rf) pulses were approximately 12 μs for 7 Li. 7 Li CSI were recorded using a single phase encoded spatial dimension. 1 Typical experiment times for 7 Li CSI ranged from ca. 7-15 h.
Suitable signal could be achieved within 1-2 h for most samples, but longer experiments were run to collect additional signal from low intensity Li dendrite species (and to ensure their absence at low cycling times).    Figure 1c). 7 Li MRI was performed in the y dimension to show lateral information.