70 71 D I A M O N D L I G H T S O U R C E A N N U A L R E V I E W 2 0 2 1 / 2 2 D I A M O N D L I G H T S O U R C E A N N U A L R E V I E W 2 0 2 1 / 2 2 Figure 2: a) Measured greyscale profiles across the crack at the region indicated by the red lines in Fig. 1 b) Magnified image from Fig. 1(vi) showing where the greyscale values were determined for the spallation crack (red box) and different parts of the vertical crack (blue and yellow boxes) after passage of 1.0 mAh cm −2 ; c) Greyscale analysis showing the amount of lithium in the regions of the crack identified by the boxes after passage of 1.0 mAh cm −2 . Adapted by permission from Springer Nature: Nature Materials, Visualizing plating-induced cracking in lithium-anode solid-electrolyte cells, Z. Ning, D. Spencer Jolly, G. Li et al., Copyright 2021. Visualising dendrite-induced cracking in lithiumanode solid-state batteries Related publication: Ning, Z., Spencer Jolly, D., Li, G., deMeyere, R., Pu, S. D., Chen, Y., Kasemchainan, J., Ihli, J., Gong, C., Liu, B., Melvin, D. L. R., Bonnin, A., Magdysyuk, O., Adamson, P., Hartley, G. O., Monroe, C.W., Marrow, T. J., & Bruce, P. G. Visualizing plating-induced cracking in lithium- anode solid-electrolyte cells. NatureMaterials 20 , 1121–1129 (2021). DOI: 10.1038/s41563-021-00967-8 Publication keywords: Solid-state batteries; Imaging techniques; Materials for energy H igh energy density solid-state batteries, with ceramic solid electrolytes and lithium metal anodes, promise to address the range anxiety and safety issues of electric vehicles. However, practical application of solid-state batteries is limited by electrolyte cracking and short circuits at high charging rates due to thepropagationof lithiumfilaments calleddendrites through the ceramic electrolyte. Observing how lithium dendrites penetrate solid electrolytes and correlating this with crack propagation will guide the design of better ceramic electrolytes that can enable fast-charging solid-state batteries. Researchers used Diamond Light Source’s Joint Engineering, Environmental and Processing (JEEP) beamline (I12) to image dendrite-induced cracks using high spatial resolution X-ray Computed Tomography (XCT) and located lithium dendrites by spatially mapped X-ray diffraction. Combining XCT and diffraction provided reliable evidence of the correlation between cracks and dendrite growth into the cracks. At high charging rates, lithium dendrite ingress into the ceramic electrolyte induces spallations (conical ‘pothole’-like cracks) in the ceramic adjacent to the lithium electrode. Further lithium plating into the spallation cracks drives the propagation of transverse cracks across the electrolyte by widening the cracks from the rear. During charging, lithium plates into the dry transverse cracks, ultimately causing short- circuiting of the battery. Preventing lithium dendrites is key to enabling fast-charging solid-state batteries. This work suggests that an effective way to prevent dendrite growth in solid-state batteries is to inhibit the development of dry cracks in the ceramic electrolyte. Therefore, strategies that toughen the ceramic electrolyte, such as fibre reinforcement and transformation toughening, may help to enable fast-charging, safe and highly energy-dense solid-state batteries. Solid-state batteries pairing a lithium anode with a ceramic solid electrolyte are being considered for next-generation batteries as they promise to revolutionise the energy density and safety of cells 1 . Electric vehicles are an important market for next-generation batteries, and a key requirement is fast charging. However, charging at practical rates in the mA/cm 2 range proves challenging, as dendrites (filaments of Li metal) are observed to penetrate across the ceramic electrolyte, leading to short-circuit and cell failure 2 . In order to develop solid-state batteries that are able to charge at high rates, it is important to understand the mechanism by which dendrites grow into the ceramic electrolyte during the charging process. However, despite decades of research into this problem there is no clear consensus in the literature as to how soft Li metal can propagate dendritic cracks through ceramic electrolytes which have shear moduli that are orders of magnitude higher than that of Li metal 3 . In this study, in situ X-ray Computed Tomography (XCT) with high spatial resolution and phase contrast was combined with spatially mapped X-ray Diffraction (XRD) to follow the penetration of Li dendrites into the ceramic electrolyte of a solid-state battery during cycling. Imaging the dendrite as it grows provides new observations as to how dendritic cracks initiate and propagate into the ceramic electrolyte, giving new insights as to the mechanism of dendrite growth. For this study, a solid-state symmetric cell consisting of Li electrodes and an argyrodite-type Li 6 PS 5 Cl electrolyte was chosen, as Li 6 PS 5 Cl is a leading candidate solid electrolyte due to its high ionic conductivity. XCT was used to track crack propagation as a function of the state of charge of the cell. Early in the charging process, XCT revealed that a crack forms (Fig. 1ii), developing into a conical, ‘pothole’-like crack, termed a spallation crack, at the surface of the ceramic electrolyte near the interface with the plated electrode (Fig. 1iii). As more charge is passed, this surface cracking worsens and the spallation crack is observed to widen (Fig. 1iv). Analysis of the greyscale within the spallation crack shows that the grey-level within the crack gradually increases as a function of charge, revealing that Li is plating within the spallation, filling and widening the crack (Fig. 2a). As more charge still is passed, a crack which is perpendicular to the electrode/electrolyte interface, termed the transverse crack, is observed to propagate across the ceramic electrolyte from the position of the spallation crack. The transverse crack advances across the solid electrolyte, finally reaching the counter electrode after passing 0.8 mA·h/cm 2 of charge (Fig. 1v). Interestingly, even after passing 1 mA·h/cm 2 capacity the cell was not observed to short-circuit, indicating that the crack traversing the electrolyte was not completely filled with Li, but rather was propagating ahead of the Li metal. This conclusion was strengthened by further analysis of the greyscale within the crack after passing 1 mA·h/cm 2 (Fig. 2b-c), revealing that whereas Li has started to plate into the top of the transverse crack near to the spallation (blue), the lower region of the crack near the counter electrode contained no Li metal (yellow). The implication of these results is that dendrite propagation through the solid electrolyte is driven by the plating of Li not at the crack tip, but at the rear of the crack. As a result, cracks can traverse the electrolyte without the cell short-circuiting, but will critically fail once Li fills the cracks. These observations are contrary to some previous models for the formation of dendrites, and therefore considerably narrow the possible mechanisms for dendrite growth through ceramic electrolytes. The relationship between spallation cracking and dendrites was further investigated using spatially resolved XRD mapping to identify the location of Li dendrites. Diffraction was carried out with the X-ray monochromatic beam perpendicular to the Li/Li 6 PS 5 Cl interface, and the intensity of the diffraction peak corresponding to Li {110} was mapped onto the area of the ceramic electrolyte to identify where Li metal was present within the ceramic (Fig. 3a). This map of dendrite locations could then be superimposed on the XCT reconstruction of the ceramic electrolyte, revealing that Li dendrites had only penetrated into the ceramic at positions where surface spallations were observed. This result demonstrates that transverse dendritic cracks typically initiate from spallation cracks. That Li dendrite growth into ceramic electrolytes initiates with spallation cracking at the interface, and that these spallations develop into transverse cracks propagating across the electrolyte driven by Li plating at the rear of the crack rather than at the tip, are new observations that inform understanding of how solid-state batteries fail at high rates of charge. These insights lay the foundations for preventing dendrite growth during fast-charging using strategies such as engineering the ceramic electrolyte by transformation toughening and fibre reinforcement to block the propagation of dry-cracks. This study also reveals the power of synchrotron X-ray techniques such as in situ XCT and XRD mapping for revealing the failures occurring within a solid- state battery. References: 1. Janek, J. et al. A solid future for battery development. Nat. Energy 1 , 16141 (2016). doi: 10.1038/nenergy.2016.141 2. Kerman, K., et al. Review - Practical challenges hindering the development of solid State Li ion batteries. J. Electrochem. Soc. 164 , A1731–A1744 (2017). DOI: 10.1149/2.1571707jes 3. Pasta, M. et al. 2020 Roadmap on solid-state batteries. J. Phys. Energy 2 , 0–52 (2020) DOI: 10.1088/2515-7655/ab95f4 Funding acknowledgement: P.G.B. is indebted to the Faraday Institution All-Solid-State Batteries with Li and Na Anodes (FIRG007, FIRG008), as well as the Engineering and Physical Sciences Research Council, Enabling Next Generation Lithium Batteries (EP/ M009521/1), the University of Oxford experimental equipment upgrade (EP/M02833X/1) and the Henry Royce Institute for Advanced Materials (EP/ R0066X/1, EP/S019367/1, EP/R010145/1) for financial support. G.L. and C.W.M. acknowledge the Faraday Institution Multiscale Modelling (FIRG003) and the UK Industrial Strategy Challenge Fund: Materials Research Hub for Energy Conversion, Capture, and Storage, under grant EP/R023581/1, for financial support. J.I. is supported by the Swiss National Science Foundation (no. PZ00P2_179886). We thank Diamond Light Source (beamline I12, experiment no. EE20795-1 ) and the Paul Scherrer Institute (TOMCAT beamline X02DA, experiment no. 20182142 ) for beamtime, as well as technical and experimental support. Corresponding author: Dominic Spencer Jolly, University of Oxford,
[email protected]; Peter G. Bruce, University of Oxford,
[email protected] Imaging andMicroscopy Group Beamline I12 Figure 1: In situ propagation phase-contrast XCT virtual cross-sections during a single plating of a Li/Li 6 PS 5 Cl/Li cell. Cross-sections show (i) the pristine cell and (ii) after 0.2 mAh cm −2 ; (iii) 0.4 mAh cm −2 ; (iv) 0.6 mAh cm −2 ; (v) 0.8 mAh cm −2 and (vi) 1.0 mAh cm −2 of charge passed; plated electrode at the top. The red line across the crack corresponds with the greyscale analysis in Fig. 2a. Adapted by permission from Springer Nature: Nature Materials, Visualizing plating-induced cracking in lithium-anode solid-electrolyte cells, Z. Ning, D. Spencer Jolly, G. Li et al., Copyright 2021. Figure 3: Diffraction mapping showing distribution of lithium dendrites and their association with the spallation cracks. a) Diffraction intensity of lithium {110} peak revealing the distribution of lithium dendrites. The dashed black circle marks the position of the electrodes in the cell; b) XCT image slices from planes in the electrolyte parallel and adjacent to the two electrodes, (i) and (ii). The two most intense lithium peaks are marked with stars. Adapted by permission from Springer Nature: Nature Materials, Visualizing plating-induced cracking in lithium-anode solid-electrolyte cells, Z. Ning, D. Spencer Jolly, G. Li et al., Copyright 2021.