Figure 2. Multiple length scale x-ray tomography of NMC622 electrode. Panel (a) is microtomography (0.65 μm pixel size) data of a piece of NMC electrode. Panel (b) is the 3D rendering of the nano-tomography data (100 nm pixel size) of an arbitrarily selected region of interest over the electrode, with its central slice in the lateral direction shown in panel (c). Three of the representative particles are highlighted in red (panel (d)), green (panel (e)), and blue (panel (f)). They are labeled as severely damaged, mildly damaged, and least damaged particles, respectively. The relative frequency of these three kinds of particles, which changes as a function of the cycling history, are summarized in the radar chart in panel (g). The scale bar in panel (a), (b) and (c), (d) are 150 μm, 20 μm, and 3 μm, respectively.
The x-ray phase contrast tomography data covers over a thousand active particles with spatial resolution down to ~50 nm. Such unprecedented amount of morphological details can facilitate a more detailed quantification. We show in Figures 3a-3d two lateral virtual slices (xy plane) at different z position. Figure 3a represents the slice near the top surface of the electrode (close to separator), while Figure 3b is the data near bottom surface (near the aluminum current collector). Figures 3c and 3d are the same slices with color-coding to the degree of particle fracturing. The corresponding relative probability distribution are shown in the insets in Figures 3c and 3d, respectively. It is evident in our observation that a depth-dependent particle fracturing pattern has been developed in this electrode.
To understand the chemical implications of the different degrees of morphological degradation at the top and the bottom of the electrode, we carried out surface sensitive soft x-ray absorption spectroscopic (soft XAS) measurements of the top and the bottom of the electrode. The total electron yield (TEY, probing depth at ~10 nm) and total fluorescence yield (TFY, probing depth at ~100 nm) signals are acquired over the absorption L-edges of Ni (Figures 3e and 3f), Mn, Co, and the K-edge of O (Supplementary Figures S1), respectively. In Ni rich NMC cathode, Ni’s redox reaction is the major mechanism for the charge compensation during repeated lithium (de)intercalation. We, therefore, present the Ni L-edge data in Figures 3e and 3f for in-depth discussions and leave the rest of the spectroscopic data, whose implications are less significant, in the Supplementary Information. In the TEY mode, Ni2+ component is clearly observed on both the top and bottom of the electrode (Figure 3e), suggesting that notable surface reconstruction\cite{batteries} happens throughout the entire electrode under fast charging conditions. The difference shown in Figure 3e indicates that the top of the electrode experiences more severe undesired local phase transition from the layered structure to a mixture of spinel and rocksalt structure. On the other hand, the bulk sensitive TFY signal shows only little difference between the top and the bottom of the electrode. This observation suggest that, at the electrode level, we do not observe significant depth heterogeneity in the SoC. While the mesoscale SoC heterogeneity could persist after long term relaxation\cite{Gent_2016}, it doesn’t seem to be the case at the electrode level as the charge transfer among the active particles could be relatively easy through the interconnected liquid electrolyte and conductive carbon network.