Understanding interface stability in solid-state batteries

PDF Publication Title:

Understanding interface stability in solid-state batteries ( understanding-interface-stability-solid-state-batteries )

Previous Page View | Next Page View | Return to Search List

Text from PDF Page: 011

in excellent agreement with the DFT prediction. Again, the reduced transition-metal cation (Mn3+ in LaMnO3) was observed at high temperature. Stability under cycling conditions. During battery cycling, the LLZO/cathode interface is predicted to decompose via chemical mixing145 or LLZO oxidation, because many of the charged cathodes have poten- tials above the oxidation stability limit of LLZO (2.9 or 3.2 V)64,66. In fig. 4d, the computed driving force for chemical mixing between various cathodes and LLZO or LLTaO is plotted as a function of voltage. Among the reactions between LLZO and three common cathodes (LiCoO2, LiMnO2 and LiFePO4) in their typical cycling range of 2.5–4.5 V (refs128,165), the LLZO/LiCoO2 inter- face has the lowest driving force for chemical mixing (<~50 meV/atom), whereas the LLZO/LiFePO4 interface is the most reactive. However, experimental data for garnet/cathode interfaces under battery-cycling condi- tions remain ambiguous. In a LiCoO2|LLZO|Au|Li cell, a small irreversible capacity (~5 mAh g−1) was observed between 2.7 and 3.8 V (ref.128), which is consistent with the predicted LLZO oxidation or chemical mixing with LiCoO2 in this voltage range65,145. However, after cycling a solid-state LiCoO2|Nb-doped LLZO (LLZNO)|Li cell (fabricated by depositing a thin film of LiCoO2 on an SE pellet) between 2.5 and 4.2 V at room temperature166, an excellent first coulombic efficiency of 99% and capacity retention of 98% were reported after 100 cycles, indicat- ing that the extent of the reactions at both the LLZNO/ LiCoO2 and LLZNO/Li interfaces under cycling are small and/or passivating. Therefore, more direct exper- imental analysis of the garnet/cathode interface under battery operation is required to determine whether this interface is kinetically stabilized or passivated under long-term cycling. Summary. In summary, although LLZO has often been claimed to be stable with Li metal and to voltages above 5 V, the collective theoretical and experimental data sug- gest a more nuanced picture. Whereas the Zr-containing garnet only has a minor thermodynamic driving force to react with Li metal, the Nb-containing garnet can clearly be reduced by Li, as evidenced by both DFT calculations and experimental data. Strongly reducible dopants such as Fe3+ further deteriorate the reduction stability. The Zr and Ta systems have high barriers for topotactic Li insertion, which likely kinetically stabilize these systems against a Li-metal anode. However, if Li insertion occurs in cubic garnets, a tetragonal phase (stabilized by the higher Li content) forms and increases the interfacial impedance. In principle, the slight reduction of Zr4+ in LLZO by Li metal would also increase the electronic conductivity of the interphase and slowly propagate into the bulk electrolyte. The observed oxidation decomposition at approx- imately 4 V indicates that LLZO cannot be paired with a high-voltage cathode such as LiNi0.5Mn1.5O4 (~4.7 V)167,168. Stability investigations with classic lay- ered cathodes such as LiCoO2 and NCM provide a less clear picture. Although Li loss from LLZO, either at high temperature or from extraction from a highly charged cathode, appears to lead to the formation of La2Zr2O7 and other cathode-related decomposition products, experimental data indicating the significance of this reaction under normal cycling conditions are missing. In this context, we want to stress that the long-term operation of SSBs will require a very high coulombic efficiency and that even minor continuing reactivity at the interface must be prevented. Chemical mixing of garnets with oxide cathodes is much less severe than that of sulfide SEs; however, the high-temperature sintering required for processing not only destabilizes LLZO by Li loss but also promotes elemental interdiffusion and transition-metal reduc- tion. For example, the reaction products La2CoO4 and LaMnO3 both contain a reduced transition-metal cation (Co2+ and Mn3+) from the cathode and La3+ from the gar- net SE. Therefore, techniques such as low-temperature and/or short-time sintering and interfacial modification such as coating are desirable for garnet SEs. LiPON Amorphous LiPON has been successfully used as an SE for thin-film solid-state microbatteries, owing to its acceptable ionic conductivity (~10−6 S cm−1)169,170, low electronic conductivity (10−12–10−14 S cm−1)171,172 and apparent wide electrochemical stability window9. Capacity retention of 90% has been observed for a Li/LiPON/LiNi0.5Mn1.5O4 solid-state cell over 10,000 cycles between3.5and5.1V(ref.3),withthestabilitywindow of LiPON determined using CV ranging from 0 to 5.5 V (ref.9). Such outstanding electrochemical per- formance has been used to argue that LiPON is stable against a Li-metal anode and possesses excellent high- voltage stability3,9,169,173. However, DFT calculations predict the decomposition of LiPON by oxidation of N above 2.6 V to form N2 gas and Li3PO4 (or Li4P2O7), and reduction of P below 0.68 V to form Li3P (refs64–66). This apparent discrepancy can be explained by the for- mation of passivating SEIs at both high and low voltage, as none of the decomposition products are electron- conductive64,66. Indeed, gas evolution was observed in a LiPONthin-filmcellchargedto5.8V(ref.9),consistent with the predicted N2 generation above 2.6 V (ref.66). When in contact with Li metal, thermodynamic DFT analysis predicts LiPON to be fully reduced to Li3P, Li2O and Li3N (ref.64). Explicit interface calculations also point towards the instability of LiPON against Li metal88, with Li atoms observed to be inserted into LiPON dur- ing the structural relaxation, reducing P5+ and break- ing P–N and P–O bonds. In experiments, in situ XPS analysis indeed revealed the presence of Li3P, Li3N and Li2O at the LiPON/Li interface73. These decomposition products are favourable as they not only block electron conduction but also permit Li-ion diffusion across the interphase174,175. The chemical reactions at the LiPON/LiCoO2 inter- face was investigated using XPS during LiPON sputtering and subsequent annealing176. As LiPON was sputtered onto LiCoO2, LiNO2 and likely some Li2O formed, with Co3+ being reduced to Co2+ in LiCoO2. During the stepwise annealing, LiNO2 disappeared by 300 °C and Co3O4 and Li3PO4 formed at higher temperature. Reviews Nature reviews | Materials

PDF Image | Understanding interface stability in solid-state batteries

PDF Search Title:

Understanding interface stability in solid-state batteries

Original File Name Searched:

2019_xiao_nature_review.pdf

DIY PDF Search: Google It | Yahoo | Bing

Sulfur Deposition on Carbon Nanofibers using Supercritical CO2 Sulfur Deposition on Carbon Nanofibers using Supercritical CO2. Gamma sulfur also known as mother of pearl sulfur and nacreous sulfur... More Info

CO2 Organic Rankine Cycle Experimenter Platform The supercritical CO2 phase change system is both a heat pump and organic rankine cycle which can be used for those purposes and as a supercritical extractor for advanced subcritical and supercritical extraction technology. Uses include producing nanoparticles, precious metal CO2 extraction, lithium battery recycling, and other applications... More Info

CONTACT TEL: 608-238-6001 Email: greg@infinityturbine.com (Standard Web Page)