Combined 4D-STEM and EELS for Mapping Chemical-structural Heterogeneity in Cathode Materials

Nanoscopic structural and chemical analysis is critical for developing our understanding of performance degradation in energy storage materials. One such technique is scanning transmission electron microscopy (STEM) which can be used to elucidate chemical ordering, strain motifs, and structural rearrangement during battery cycling [1,2,3]. Such studies have expanded our understanding of degradation pathways in cathode materials, showing the formation of consequential phases which increase the energy barrier for Li ion intercalation [4]. Recent efforts have included the application of scanning electron nano diffraction (SEND), or 4D-STEM, to probe larger areas with nanoscopic precision while providing diffraction patterns at each incremental step. These patterns can be data-mined for structural information such as strain or orientation mapping [5]. Previous studies utilizing SEND have provided insight into structural phase domain formation and strain degradation from sodium ion intercalation and the resultant change in ion diffusivity [6]. This type of analysis can be paired with traditional spectroscopy methodology to describe chemical changes within the material driving the structural rearrangement. Here, we use SEND, along with position resolved electron energy loss spectra (EELS), to analyze chemical heterogeneity within a nickel rich battery cathode material (LixNi0.8Mn0.1Co0.1O2, NMC811), before (Figure 1) and after electrochemical cycling. By analyzing chemical oxidation states within the cathode, we show that the Li concentration in the oxide becomes heterogeneous after fast charge cycling of a NMC811/Graphite cell. EELS maps of the particles show regions of oxidation state heterogeneity as the Ni oxidation transitions from 2.875+ in the particle interior to 2+ at some of the particle edges because of oxygen loss. Simultaneously interior particle regions show higher Ni oxidation states to compensate for Li-loss from the oxide during the cycling. We note that by accounting for all transition metal oxidation states, we can estimate local lithium concentrations in the oxide. To analyze the structural effects of primary particle chemical heterogeneity, we overlap the EELS maps with SEND data. For example, by analyzing the SEND patterns corresponding to Oxygen K-edge pre-peaks we observe diffuse scattering between the (-120) and (-12-3) Bragg peaks. Taking a virtual dark field image using an aperture between these peaks, we find contrast consistent with Oxygen K-edge EELS mapping (Figure 2). Using correlation methods between EELS and SEND allows us to describe the chemical-structural relationship across micron sized particles with nanoscopic accuracy. (Left panel) EELS L3/L2 ratio maps corresponding to oxidation state for Ni, Mn, and Co from a pristine oxide cathode. (Right panel) Corresponding raw core loss EELS. a) O-K edge map showing the integrated pre-peak over the integrated edge intensities. B) Representative O-K edge spectra. c) Corresponding diffraction patterns of the regions used to collect EELS data. e) Virtual dark field images collected using the region marked in c.