Effects of Atmospheric Gases on Li Metal Cyclability and Solid-Electrolyte Interphase Formation

For Li–air batteries, dissolved gas can cross over from the air electrode to the Li metal anode and affect the solid-electrolyte interphase (SEI) formation, a phenomenon that has not been fully characterized. In this work, the impact of atmospheric gases on the SEI properties is studied using electrochemical methods and ex situ characterization techniques, including X-ray photoelectron spectroscopy, X-ray diffraction, Fourier transform infrared spectroscopy, and scanning electron microscopy. The presence of O2 significantly improved the lithium cyclability; less lithium is consumed to form the SEI or is lost because of electrical disconnects. However, the SEI resistivity and plating overpotentials increased. Lithium cycled in an “air-like” mixed O2/N2 environment also demonstrated improved cycling efficiency, suggesting that dissolved O2 participates in electrolyte reduction, forming a homogeneous SEI, even at low concentrations. The impact of gas environments on Li metal plating and SEI formation represents an additional parameter in designing future Li-metal batteries.

A Li metal foil was used as the counter and reference electrode. Celgard 2400 swollen with 1M LiTFSI in tetraglyme was used as the separator and electrolyte. Copper mesh used as the substrate and working electrode for Li plating/stripping tests. A flat mesh was used, and current densities were calculated for the projected top-view area. A mesh was used to allow gas dissolution into the electrolyte.

Supplemental figure 2: Method for calculating coulombic efficiencies
Coulombic efficiency was calculated using the Aurbach method. 1 This method considers an average coulombic efficiency over the whole experiment. First an initial quantity of lithium is plated onto the Cu (qi). A fraction of this reservoir (qc) is subjected to N cycles of symmetric plating and stripping. Finally, all available Lithium is stripped from the Cu (qf). The coulombic efficiency, X, can then be calculated using the expression shown on the top left-hand side. Plating overpotentials were similar to that in an Ar environment. This is consistent with our hypothesis that O2 reduction occurs on the Cu electrode prior to Li plating, thus promoting uniform initial Li nucleation and growth.

Supplemental figure 4:
Electrochemical Impedance spectroscopy measured after 20 plating/stripping cycles in O2, Ar, and N2 environments (top) and equivalent circuit model used to fit the data (bottom).
A symmetric cell was made using the Cu working electrodes from two cycled cells.
Potentiostatic EIS was performed from 1MHz to 1Hz using 10 mV amplitude.  , was necessary to fit the data, suggesting  an additional component to the SEI layer with uniquely large resistance values in the O2 sample  (table S1 below). This RC unit has significantly higher capacitance ( ~10 -5 F) than a typical grain boundaries capacitance ( ~10 -8 -10 -8 F), 3 however since the high-frequency semi-circle is usually assigned to the contribution of grain boundaries, this component likely corresponds to the LiOH layer with large grain sizes (as seen in the SEM, fig S12). The high capacitance could be the result of higher porosity compared to the typical porosity of ceramic electrolytes. 4 A third capacitor/resistor element was used to model the charge transfer resistance (CT) and double layer capacitance (dl) at the SEI/electrolyte interface. A Warburg diffusion element was used to model Li ion diffusion. Constant phase elements were used instead of capacitors in order to capture the distribution of relaxation times caused by the complexity of the SEI layer. 5 These elements give a pseudocapacitance (Q) and the dispersion coefficient (a) refers to the phase deviation from 90 (an ideal capacitor).   Histograms showing the relative atomic % from XPS measurements for the SEI formed in O2, Ar, and N2 at 0 nm and 10 nm sputtering depth. A large C-C peak is visible for all samples, indicative of adventitious carbon as well as the organic SEI layer. Small amounts of salt decomposition products such as LiF (purple) and Li2S (cyan) and higher concentrations of F, N, and S were already visible, for Ar and N2 samples. Table 2: Total atomic % for the SEI formed in O2, Ar, and N2 at various sputter depths.

Supplemental
Supplemental figure 8: EDX spectra for SEI formed in O2, Ar, and N2. Significantly higher F and S content were observed for Ar samples.

Supplemental table 3:
Relative atomic weight % from EDX point measurements of the SEI formed in various gas conditions. EDX spectra for Li metal exposed to O2 and Ar at OCV with electrolyte were also measured.
For the electrochemically formed SEI in O2, the O content was dramatically increased, while the chemically formed surface showed much less oxidation. The relatively high F content for Li metal exposed to O2 and Ar at OCV is likely due to residual electrolyte.  As current is applied, various electrolyte species are reduced on the copper electrode until the potential drops to below 0 when Li + reduction begins. In the presence of O2, less charge is needed to form a stable and electrically insulating initial layer. The cell in an O2 environment shows consistent coulombic efficiencies for many cycles (black). In an Ar environment, coulombic efficiency drops rapidly after the first 3 cycles due to dead Li (red). Cycle life could be greatly improved when 5 plating/stripping cycles were first performed under an O2 environment before purging the cell with Ar gas and continuing to cycle in an Ar environment (green). However, even with the pre-treatment, the coulombic efficiency drops after 30 cycles. LiOH is visible on the surface of the Li nuclei and the FTIR show peaks for both -OH (~3650 cm -1 ) and -CO3 (~1620 cm -1 ) in mixed O2/N2 environments. (1) Oxygen is reduced into Li-oxide species which can nucleophilically attack ethers and form Li-carbonates, Li-alkoxys, and LiOH. 6 (2) Ether reduction to form a radical which is readily stabilized by O2 forming a peroxy-radical. DFT calculations performed by Assary et al. indicate that the O2 reaction with ether radical is favorable. 7 The radical then decomposes into Li-carbonates, Li-alkoxys, and LiOH. Additional experiments adding 1% H2O into the electrolyte and then plating/stripping lithium in an Ar environment were performed. Cells were only able to complete 9 symmetric plating/stripping cycles before a final stripping step to remove all available lithium: coulombic efficiencies calculated using the Aurbach method were 40% and 42% for the Li cycled in Ar and Ar with H2O additive, respectively. This suggests that although H2O contamination may form LiOH in the SEI, it does not appear to affect Li plating/stripping efficiencies in the same manner as O2 crossover species. Future experiments using deuterated glymes may be explored to confirm the proton source in the LiOH found in the SEI. Dendrites are still visible in the SEM image even with the addition of water, explaining the low coulombic efficiencies even with water additive. Electrolyte is first reduced until the Cu electrode is uniformly covered by an electrically insulating layer that prevents further electrolyte reduction (a, f). The O2 crossover species may help form a uniform coating of LiOH and Li carbonates (g) compared to an SEI composed of heterogeneous salt decomposition products (b). Lithium ions then migrate through the SEI in order to be reduced to Li metal (c-d, h-i). The SEI formed with O2 is more robust and uniform, therefore promotes spherical growths (e, j).
Supplemental Table 4: Peaks in the C 1s spectra for SEI formed in O2, Ar, and N2 at various sputter depths. Li2CO3 is seen as an SEI component ~290 eV and is in larger relative concentrations for samples in O2. The CF3 peaks seen at the surface is due to residual TFSIsalt anion or decomp products.