All-solid-state batteries with lithium metal as anode material are attracting more and more attention because of their potentially higher energy and power density. In particular, they promise to be safer than lithium-ion batteries, which use flammable liquid electrolytes. The property of the solid electrolyte determines the success of future all-solid-state batteries. Garnet-type solid electrolytes based on Li₇La₃Zr₂O₁₂ are widely regarded as one of the most promising solid electrolyte class due to its relatively high lithium ion conductivity, wide voltage stability window and chemical compatibility with lithium metal. However, in operation, lithium dendrites can penetrate into the garnet-type solid electrolyte, especially along grain boundaries. This is one of the main obstacles for further development and the origin is still unclear [1].

Other groups addressed this topic by performing simulations or traditional electrochemical characterizations on macroscale to characterize grain boundaries conduction properties. However, the conclusions drawn are contradictory. For example, some works reported that grain boundaries in garnet-type solid electrolytes have lower ionic conductivity, while other studies reported the opposite [2-3]. Thus, for finding the origin for lithium dendrite growth, we used in-operando scanning force microscopy with high spatial resolution for the first time.

In this work, smooth cuts though surface of a representative garnet-type solid electrolyte Li_6.25Al_0.25La₃Zr₂O₁₂ were prepared by ion milling and SFM tests were operated in an argon-filled glove box. First, we measured that grain boundaries have similar lithium ionic conduction property with grain parts. Second, an electrical potential drop at grain boundaries close to the counter lithium electrode results from grain boundaries’ stronger electron capture ability. Third, electron beam irradiation on Li_6.25Al_0.25La₃Zr₂O₁₂ to induce metal lithium expulsion was used to quantitatively analyze the different electron conduction properties in grain boundaries. Based on our experimental findings, we provide a model, which explains why grain boundaries are “hot spots” for lithium dendrite formation [4].

References:

[1] Krauskopf T, Dippel R, Hartmann H, Peppler K, Mogwitz B, Richter FH, Zeier WG, Janek J. Lithium-metal growth kinetics on LLZO garnet-type solid electrolytes. Joule. 2019 Aug 21; 3(8):2030-49.

[2] Lu Z, Yang Z, Li C, Wang K, Han J, Tong P, Li G, Vishnugopi BS, Mukherjee PP, Yang C, Li W. Modulating Nanoinhomogeneity at Electrode–Solid Electrolyte Interfaces for Dendrite‐Proof Solid‐State Batteries and Long‐Life Memristors. Advanced Energy Materials. 2021 Apr; 11(16):2003811.

[3] Cheng L, Chen W, Kunz M, Persson K, Tamura N, Chen G, Doeff M. Effect of surface microstructure on electrochemical performance of garnet solid electrolytes. ACS applied materials & interfaces. 2015 Jan 28; 7(3):2073-81.

[4] Zhu C, Fuchs T, Weber S, Butt H-J, Richter F. H, Janek J, Berger R. Submitted (2022).