Origins of Heterogeneous Li Insertion at the Porous Electrode, Single Particle, and Atomic Length Scales

Abstract

The insertion of a guest ion to a host crystal is the fundamental reaction underpinning all Li-ion batteries. Understanding the mechanisms of how the ions enter the host, and how the host crystal responds, are keys to rational design of battery materials with higher energy density, power density, and cycle life. Our work directly visualizes the ion insertion mechanism in the model phase-separating material LiFePO4 at multiple length scales: porous electrode, single particles, and atomic. We show LiFePO4 is highly heterogeneous in terms of Li insertion, does not necessarily phase separate, and is not a one-dimensional conductor during charge and discharge. We begin by investigating lithiation at the porous electrode length scale (i.e., microns). Using ptychography (Shapiro et al. Nat. Photon. 8, 765, 2014), a synchrotron X-ray imaging technique with a spatial resolution of 10 nm, we study which particles in an electrode lithiates first, we show that the connectivity between the particles and the carbon black network controls the sequence of lithiation. When electronic transport is uniform (achieved via high carbon black loading), smaller particles preferentially lithiate. This further shows that the rate-limiting process in a porous electrode is electronic transport (Li et al. Adv. Mater.27, 6591, 2015). Next, we decrease the length scale to that of single particles (i.e., hundreds of nm), and investigate the fraction of particles actively intercalating. Our work shows that the active particle fraction increases with rates of lithiation, indicating that higher cycling rates improve the uniformity of Li insertion (Li et al. Nat. Mater.13, 1149, 2014). This rate-dependent heterogeneity arises as a result of the unique thermodynamics of phase-separating battery electrodes, whereby the local current density on each particle is mostly invariant with the global C-rate. To understand the role of ion insertion within individual particles, we zoom in further to study heterogeneity at the sub-particle length scale (i.e., tens of nm), We developed operando liquid X-ray imaging platforms to track the Li migration within individual particles (Li et al. Adv Funct. Mater. 25, 3677, 2015; Nelson Weker et al. ChemElectroChem 2, 1765, 2015; Lim, Li, et al. Submitted). By tracking the same particles in real time, we quantified the local current density. Our work shows that phase separation is suppressed at higher rates of discharge, yielding a uniform solid solution pathway. On charge, however, heterogneeous phase separation is much more prevalent, a result of the composition-dependent exchange current density.5 Finally, we study ion insertion at the atomic length scale. We combine experimental and computational tools to study how a solid-solution particle separates. Our results demonstrate that, in the presence of a liquid electrolyte, LiFePO4 rapidly phase separates within individual particles, rather than creating a mosaic of lithiated and delithiated particles. This indicates the presence of a surface diffusion layer perpendicular to the fast bulk diffusion direction that is activated by the electrolyte, and that LiFePO4is not strictly a one-dimensional conductor under dynamic conditions. By directly visualizing Li insertion across diverse length scales, we obtain a comprehensive understanding of ion insertion in LiFePO4. These insights are likely applicable to a broad range of phase-separating materials for lithium-ion batteries and beyond.