Three dimensional analytical study of thin film battery electrodes

Abstract

Li ions play a major role in batteries for energy storage. On the other hand, Li is notoriously challenging to be reliably detected with most microscpic techniques.Owing to its weak scattering form factor, low X-ray emission or peak overlaps in EELS spectroscopy, precise microscopic analysis of Li in battery materials is delicate. The aim of this study was to investigate a very sensitive analysis of the ionic transport on a microscopic, even on an atomic level scale. Well controlled amorphous lithium iron phosphate (LFP) thin films were prepared by ion-beam sputter deposition. In a subsequent annealing step, amorphous films were then crystallized. The electrochemical performance of both LFP phases is checked in cyclo-voltammetry, while structure and microstructure are confirmed by XRD and TEM, respectively. Cycling reversibility over 7000 cycles with a retention more than 92 % is accomplished for the crystalline LFP, whereas the amorphous phase is electrochemically nonfunctional. Intercalation of LFP thin films was studied as a function of film thickness (25 - 250 nm). The intercalation kinetics is systematically quantified over a wide range of scanning rates (0.004 to 400 mV s-1 in cyclic voltammetry experiments. Two different diffusion regimes for the material undergoing two phase reaction were explained with the help of the modified Randles-Sevcik equation. Slow Li diffusion in the thickest films was recorded. Dependence of the peak current on the layer thickness is explained in terms of increasing the grain boundary (GB) area. Opposite to the peak fluxes, the overpotential was interestingly found to be independent of the layer thickness. Less electrical driving force is required to force the same current in thick film. The grain boundaries represent an electroactive interface at which the overpotential appears. And hence, the grain boundaries work as fast conduction paths for faster Li ions diffusion. Thus, the total current is controlled by the total grain boundary area rather than the thin film surface. LiFePO4 (LFP) is then 3-dimensionally studied by laser-assisted Atom Probe Tomography (APT). The effects of laser power on the quantitative analysis of the amorphous phase by atom probe tomography were considered. The systematic investigation of amorphous samples presented herein demonstrates quantification of constituent elements, particularly lithium. Stoichiometric ratios relative to all elements (Li+Fe+P+O) and to the stable element (Fe) were calculated; P and O reveal reverse behavior against laser power. Li, on the other hand, after considering its migration, increases with rising laser power. Even though APT measurements at cryogenic temperatures (60 K) were performed, migration of Li ions in some LFP states was observed. In response to the applied measurement fields, Li ions are undoubtedly redistributed. In the amorphous LFP material, we observe a strong Li gradient towards the tip front, which hinders reliable analysis. Obviously, during measurement, Li is drawn towards the tip front and this effect increases with increasing laser power. The remaining host elements, Fe, P, and O, remain homogeneously distributed. New unique insights into the mechanisms of Li movements are provided. Li is pulled and Li enrichment/depletion regions are observed. A new term "Li shooting" is addressed to describe these Li movements. It is demonstrated that the ions indeed experience a field-dependent drift. By mathematically modelling the resulting composition profiles, the Li diffusivity is quantitatively evaluated. In a direct comparison between the amorphous and the crystalline LFP films of identical chemical composition, it is shown that the diffusivity of the amorphous structure is orders of magnitude faster than that of the crystalline state at a temperature of 60 K. Most notably, this is the first study to investigate the capabilities of APT in LFP at different de-/lithiation states. Li compositions show a wave-like distribution as a result of existence the Li rich/poor phases. 3D iso-surfaces and 2D orthoslices were provided to differentiate between the two phases. Li, in the fully lithiated phase, reaches its ideal stoichiometric ratio, while it is overestimated in the fully delithiated phase. Obviously, the thin films include inactive LFP regions. They were highlighted in this thesis for the first time by atom probe analysis. To quantify the two-phase nature of LFP, statistical analyses of the dis/charged LFP at different lithiation states were performed. Observed frequency distributions of the concentrations of small clusters were compared to the binomial distributions and discussed in detail. Deviations between observed and binomial distributions were represented in the Pearson coefficient to demonstrate the phase separation in the atom probe analysis. Our results provide an evidence to statistically understand the local microstructure evolution in battery materials, which is a pivotal characteristic of battery performance. On the other hand, APT has shown some constraints to microscopy differentiate between the two phases. Although APT measurements were performed at cryo-temperature, Li showed a displacement during measurements. Contrastingly, most of materials are fractured early at ultra low temperatures.