A Rechargeable Li-Air Fuel Cell Battery Based on Garnet Solid Electrolytes

Non-aqueous Li-air batteries have been intensively studied in the past few years for their theoretically super-high energy density. However, they cannot operate properly in real air because they contain highly unstable and volatile electrolytes. Here, we report the fabrication of solid-state Li-air batteries using garnet (i.e., Li6.4La3Zr1.4Ta0.6O12, LLZTO) ceramic disks with high density and ionic conductivity as the electrolytes and composite cathodes consisting of garnet powder, Li salts (LiTFSI) and active carbon. These batteries run in real air based on the formation and decomposition at least partially of Li2CO3. Batteries with LiTFSI mixed with polyimide (PI:LiTFSI) as a binder show rechargeability at 200 °C with a specific capacity of 2184 mAh g−1carbon at 20 μA cm−2. Replacement of PI:LiTFSI with LiTFSI dissolved in polypropylene carbonate (PPC:LiTFSI) reduces interfacial resistance, and the resulting batteries show a greatly increased discharge capacity of approximately 20300 mAh g−1carbon and cycle 50 times while maintaining a cutoff capacity of 1000 mAh g−1carbon at 20 μA cm−2 and 80 °C. These results demonstrate that the use of LLZTO ceramic electrolytes enables operation of the Li-air battery in real air at medium temperatures, leading to a novel type of Li-air fuel cell battery for energy storage.

A new class of solid-state electrolytes, garnet-type Li 7 La 3 Zr 2 O 12 (LLZO) ceramics, were first reported by Murugan et al. 29 in 2007. LLZO ceramics have several promising properties, including chemical stability against Li metal, large electrochemical windows (above 5 V), and high Li + conductivity (above 10 −4 S cm −1 at room temperature) 30 . One problem associated with LLZO is the difficulty in sintering high-density ceramics due to the volatility of Li. Recently, through hot-press sintering and optimized processes, we successfully obtained lamellar Li 6.4 La 3 Zr 1.4 Ta 0.6 O 12 (LLZTO) electrolytes with densities as high as 99.6% and improved the ionic conductivity to 1.6 × 10 −3 S cm −1 at room temperature (as depicted in Figure S1). Based on these LLZTO disk electrolytes (~0.1 cm in thickness), we designed a new type of SSLAB architecture. An air cathode mainly composed of Ketjen black (KB), LLZTO particles and LiTFSI in polyimide (PI:LiTFSI) or polypropylene carbonate (PPC:LiTFSI) was coated on one side of the LLZTO disk. The metallic Li anode was pressed on the other side of the LLZTO disk, which was subsequently sealed airtight. The operating temperature was elevated to improve the interface contact between the Li and LLZTO as well as between the cathode and LLZTO. These two-part SSLABs can be operated in real air with good rechargeability. By replacing PI:LiTFSI with PPC:LiTFSI as the conductive binder, the batteries can be operated at reduced temperatures while displaying improved performance. Figure 1a shows the schematic configuration of a typical SSLAB. A 0.1-cm-thick lamellar LLZTO disk is used as the electrolyte. On the anode side, a Li foil is attached on the LLZTO, which is sealed in a stainless steel container with high-temperature resistive sealant. This part can withstand up to 250 °C as indicated by the thermogravimetric-differential scanning calorimetry (TG-DSC) analysis shown in Figure S2. On the cathode side, 0.2-μ m LLZTO particles are used as the ion conducting framework. As demonstrated in Figure S3, the plate consisting of LLZTO particles and the PI:LiTFSI composite shows an ionic conductivity of 1.9 × 10 −5 S cm −1 at 25 °C and 3.5 × 10 −5 at 80 °C, while the PPC:LiTFSI counterpart shows 9.3 × 10 −5 S cm −1 and 1.6 × 10 −4 S cm −1 , respectively. The KB used for the cathode is responsible for constructing the electronic network and the occurrence of both ORR and OER 31 . Figure 1b shows a cross-section image of a typical battery. Composition analysis of the air cathode by energy dispersive X-ray analysis (EDX) reveals that La, Zr and C are homogeneously distributed in the whole region (as shown in Figure S4). Figure 1c shows the discharge and charge behaviours of the batteries with PI:LiTFSI as the binder. These batteries show rechargeability only at 200 °C. Nevertheless, it is worth noting that the coulombic efficiency can be as high as 97.1%, which is likely due to the influence of elevated temperature on the Li anode and the interfacial resistance. To lower the operation temperature of the batteries, PI:LiTFSI is replaced with PPC:LiTFSI, which can reduce the interfacial resistance as discussed above. The resulting batteries can be cycled at 80 °C with a current density of 20 μ A cm −2 , as shown in Fig. 1d. After a large discharge plateau at approximately 2.78 V, the curve drops suddenly at a specific potential. The specific discharge capacity reaches approximately 20300 mAh g −1 carbon , which is approximately 6430 mAh g −1 cathode when being calculated according to the overall cathode mass. Upon charging, a plateau at 3.01 V with capacity of approximately 650 mAh g −1 carbon is initially observed. Then, a large plateau at 3.87 V with a capacity of approximately 15000 mAh g −1 carbon appears. To clarify the origin of the charge reaction at 3.87 V, the composites with PPC:LiTFSI and KB as the cathodes were charged in air at 80 °C (curve displayed in Figure S5). These composites display decomposition only when the charge potential is greater than 4.45 V. This indicates that the capacity during charging at 3.87 V is solely related to decomposition of the discharge products rather than decomposition of the composite cathode.

Results
The morphology and composition of the reaction products were characterized by both ex-situ and in-situ techniques. Figure 2 shows SEM images of changes in the air cathode during cycling. Compared to the pristine cathode ( Fig. 2a), the average size of granular particles increases at the half discharge state (Fig. 2b). After full discharge (Fig. 2c), rod-like particles can be observed, indicating that the discharge products grow upward and merge with each other to form agglomerated particles 32 . Upon charging, the large particles identified in Fig. 2c clearly shrink, as shown in Fig. 2d. After the fifth discharge (Fig. 2e), rod-like particles similar to those formed at the initial discharge ( Fig. 2c) and during the shrinkage (Fig. 2f) can also be observed, indicating repetitive formation and decomposition of the discharge product. Fourier transform infrared spectroscopy (FTIR) was also performed on the air cathodes after discharge and charge processes at 80 °C. Considering that PPC has complex FTIR peaks between 2000 cm −1 and 1250 cm −1 33 , where the characteristic peaks of Li 2 O 2 and Li 2 CO 3 lie, we used the batteries with PI:LiTFSI in the FTIR measurement. As shown in Fig. 3, the peak intensity corresponding to the Li 2 CO 3 reference greatly increases after discharge, which indicates that the identified discharge product is Li 2 CO 3 . The same phenomenon was observed by transmission electron microscopy (TEM) of the air cathodes, in which carbon nanotubes (CNT) were used as the active material instead of KB to clearly characterize the product composition upon cell operation. As shown in Figure S6, the Li 2 CO 3 grows around the CNT during discharging and decomposes upon charging.
To obtain more information on the change in chemical composition on the cathode surfaces, in-situ X-ray photoelectron spectroscopy (XPS) of the discharge and charge process was performed under ambient pressure of 0.1 mbar. The cathodes used in the in-situ XPS measurement were carbon materials that were deposited on the LLZTO disk by sputtering (details can be found in Methods section). As shown in Fig. 4, Li 2 O 2 peak intensity first increases and then decreases upon discharge. Meanwhile, the Li 2 CO 3 and Li-C-O peak intensities steadily increase with increasing depth of discharge. These results indicate that the Li 2 O 2 is formed at the initial stage of discharge, and reacts with trace H 2 O and CO 2 in ambient air to form a large quantity of Li 2 CO 3 on the cathode surfaces. The appearance of the Li-O-C peak can be attributed to the decomposition of the carbon cathode (i.e., KB) when the cell is charged, according to ref. 34. Note that in-situ measurement allows the observation of transient products (i.e., Li 2 O 2 ) 35,36 . The ex-situ measurement can only provide information on the stabilized products in air, such as Li 2 CO 3 . While charging, the peak intensity of Li 2 CO 3 greatly decreases, indicating that the Li 2 CO 3 is most  probably decomposed upon charging, at least partially. Rechargeable Li-CO 2 batteries based on the formation and decomposition of Li 2 CO 3 have already been demonstrated in previous reports 37,38 , which also indicated that the decomposition of Li 2 CO 3 upon charge was dependent on the type of carbon materials in the cathodes. The KB carbons used here are most likely not the optimum components, causing partial decomposition of Li 2 CO 3 upon charge and thus accumulation of passivated Li 2 CO 3 products after multiple cycles.
Based on the existing data and previous reports, we propose the following reaction route in the present Li-air battery. During discharging, O 2 is electrochemically reduced to form Li 2 O 2 (equation 1) in the cathode initially. Li 2 O 2 then reacts with CO 2 in air to form Li 2 CO 3 via a chemical reaction (equation 2). During charging, Li 2 CO 3 is electrochemically decomposed at least partially, according to the release of CO 2 during charge (as shown in Figure S8). In fact, preliminary DEMS (differential electrochemical mass spectrometry) results indicate that the reaction mechanism upon charging is very complicated. Further work with isotope labelling is necessary and ongoing in our laboratory.
According to the above analysis, the discharge products grow on the cathode surfaces during discharge and partially decompose during charging. This situation is similar to Li-O 2 batteries based on Li 2 O 2 formation and decomposition 39 . The insulating deposits have a passivating influence on active sites for ORR 40,41 . In the case of full discharge, a number of thick discharge products form and coat the carbon surfaces, which allows only a few cycles, as shown in Figure S7. To increase the cycle number, limiting the depth of discharge is necessary. As shown in Fig. 5a, the batteries can be run for more than 50 cycles at 20 μ A cm −2 when the discharge capacity is cut off at 1000 mAh g −1 carbon , which is approximately 316 mAh g −1 cathode if calculated according to the overall cathode mass.

Discussion
Here we demonstrate the fabrication of rechargeable SSLABs that can be operated in real air at elevated temperatures. The newly synthesized LLZTO ceramic electrolytes play essential roles in these batteries. They are dense (99.6% in relative density), highly conducting (1.6 × 10 −3 S cm −1 at room temperature) 42 , and are able to prevent gas species from diffusing through and reacting with the Li anodes. LLZTO ceramic electrolytes are also chemically and electrochemically stable against Li 29,43,44 , which allows direct contact with Li without introducing an additional resistive intermediate layer. Previous reports found that LLZTO might react with H 2 O in air, causing degradation of ionic conductivity 45,46 , but the moisture content in air can be greatly reduced when the surrounding temperature is elevated to above 80 °C 47 . This means that the problem of LLZTO degradation can be greatly relieved at elevated temperatures. Other important issues include construction of the air cathode and interfacial concerns. Air cathodes cannot be wetted in the solid-state electrolytes as they do in the liquid electrolytes. As a result, it is necessary to artificially build the electronic and ionic network. Because the discharge products are solids deposited on the cathode surfaces, sufficient space to accommodate these solids is also required for large capacities, which is why porous composite cathodes are needed. The LLZTO particles provide a framework that is ionically conducting and able to create space for air diffusion, while PPC:LiTFSI enhances ion conduction at the LLZTO interfaces.
The use of elevated temperatures is especially advantageous in terms of improving ion conduction at the Li/ LLZTO and electrode/LLZTO interfaces. However, too high operating temperature may cause the decomposition of carbon-based cathodes. As a result, selecting suitable carbon materials, and optimizing cathode configuration to reduce the interfacial resistance must be considered to improve battery performance. Moreover, searching for suitable materials with effective catalysts that can promote the decomposition of Li 2 CO 3 upon recharge is also crucial for the development of SSLABs.

Conclusions
A novel type of solid-state Li-air battery based on LLZTO ceramic electrolytes is demonstrated here. After optimizing the interfacial resistance, the batteries show a specific capacity of approximately 20300 mAh g −1 carbon at 20 μ A cm −2 (or 6430 mAh g −1 cathode if referring to the mass of cathode) in real air at 80 °C. With a cutoff discharge capacity of 1000 mAh g −1 carbon at 20 μ A cm −2 , the batteries can be operated for 50 cycles while maintaining their initial capacity. Analysis of the cell chemistry reveals that the batteries operate through formation and decomposition of Li 2 CO 3 at least partially. Cathode design related to the interfacial issue and materials research to promote decomposition of the discharge product are two key problems worth investigating in the near future. The performance achieved in the SSLABs presented here demonstrates the possibility of operating Li-air batteries in real air, which opens new avenues for the use of Li-air fuel cell battery for energy storage.

Methods
Material preparation and cell assembly. LLZTO powders in the cubic phase were prepared via a conventional solid-state reaction. More details can be found in our previous paper 42 . The bulk ceramics were fabricated in lamellar disks with a diameter of 1.2 cm and a thickness of 0.1 cm by the hot-press sintering technique.
Composite cathodes consisted of Ketjen black (KB), polyimide (PI) or polypropylene carbonate (PPC), Li(CF 3 SO 2 ) 2 N (LiTFSI) (99.95%, Sigma-Aldrich, Shanghai, China), and LLZTO powders (D 50~2 00 nm). The PI suspension was prepared by dispersing PI powder in N-methyl-2-pyrrolidone (NMP) and was stirred before use; the PPC solution was prepared by dissolving PPC in NMP. LiTFSI was dissolved in NMP before mixing with KB and LLZTO powders in an agate mortar and grinding for 1 h. The obtained slurries were coated on one side of the LLZTO ceramic membrane by blades, then pre-dried in an oven at 80 °C for 2 h to remove NMP and baked at 80 °C in vacuum overnight. The typical loading mass of each cathode was approximately 1 mg cm −2 . The weight ratio between KB, LLZTO, LiTFSI/and PI or PPC was fixed at 6.3:6.3:6.3:1. Li anodes were attached on the other side under high pressure in an Ar-filled glove box with oxygen and moisture levels below 0.1 ppm. The Li-side was sealed in a stainless steel container with high-temperature resistive sealant. Stainless steel (SS) foils were selected as current collectors. Both the specific capacity and the current density were normalized to the mass of KB.
Electrochemical measurement. Galvanostatic charge and discharge behaviours of the batteries were investigated using an Arbin battery cycler at 80 °C or 200 °C that was calibrated by a thermocouple attached to the measured cell. Before testing, the cells were first rested in a thermostatic oven for 6 h to reach equilibrium. Electrochemical impedance spectroscopy (EIS) was performed using an Autolab instrument.
Characterization. Crystalline structures of the LLZTO powders and ceramics were examined by X-ray diffraction (XRD, Bruker D2 Phaser) using Cu Ka radiation. Ionic conductivities of LLZTO ceramics were measured using an impedance analyzer (Novocontrol Beta High Performance Impedance Analyzer) at a temperature range of 20 to 100 °C with frequency ranging from 0.1 to 20 MHz with an amplitude of 10 mV. Surface and cross-section morphologies of the LLZTO ceramics and the cathodes were analyzed by scanning electron microscopy (SEM, FEI Magellan 400, Hitachi S-4800). High-resolution transmission electron microscopy (HR-TEM, JEOL JSM-6700F) and selected area electron diffraction (SAED) measurements were performed to investigate the solid products.
In-situ XPS. XPS data were collected from the top cathode surface of the SSLAB under air atmosphere during cell operation at SIMIT 35 . The cell was placed onto an insulating sample holder. The air cathodes were carbon materials, which were coated on the LLZTO disks by sputtering. The bottom and surface sides of the cells were in contact with two silver wires that served as current collectors. Two copper cables were used to connect the negative and positive electrodes to an external Bio-Logic SP-300 potentiostat. No other precleaning was performed. XPS data were collected under an atmosphere of 10 −4 atm at an operating temperature of 80 °C with a current density of 20 μ A cm −2 . The cell was discharged and charged for 6 h each. XPS spectra were collected at a photon energy of 1486.6 eV. The binding energy of all the spectra was calibrated to the Pt 4 f peak of sputtered Pt foil. Shirley background subtraction was applied to the photoemission lines, which were fitted using a combined Gaussian-Lorentzian line shape (CasaXPS).