Nonlinear Frequency Response Analysis for process identification and characterization of Li-ion batteries

锂离子电池是电动汽车和缓冲间歇性可再生能源的关键(未来)组件。为了提高它们的能量密度并节省成本，改进的电极和电池设计被开发出来。此外，现有电池最好在其性能极限下运行，这会导致老化和安全问题。在这两种情况下，都迫切需要非破坏性的操作数分析，以便深入、特别是定量地了解电池中的物理和化学过程，以及电池设计和操作模式如何影响它们。只有电压和电流可以很容易地测量，而电化学阻抗谱，一种线性分析方法，经常用于这一目的。然而，不同的状态和效应可能会导致相似的光谱。为了解决这些局限性，本项目使用了非线性方法，特别是非线性频率响应分析(NFRA)。分析的高次谐波响应包含了关于非线性过程的信息，特别是关于高度非线性反应动力学的信息，以及关于表面态的信息。因此，NFRA是诊断电池的一个极其强大和有前途的工具，但目前人们对此知之甚少。目前，对于如何测量和分析NFRA光谱--更重要的是--分析，还没有达成共识。这些都是阻碍社区更广泛地接受这种方法的主要障碍。在这个项目中，在KIT和MPI开发的不同的NFR方法将被用于电池分析。将首次研究阳极和阴极的NFR光谱，以区分它们在电池信号中的贡献；此外，还将评估先前未分析的相位信息和直流分量的信息含量。将确定类似于阻抗的高阶频率响应函数，其不依赖于幅度，但包含关于非线性响应的信息。模型的解析和数值解将被用来再现和解释实验光谱。因此，该项目建立了通过非线性频率响应分析优化提取电池(状态)信息的系统方法，并为这一新技术的广泛分析铺平了道路。

Li-ion batteries are crucial (future) components for electromobility, and for buffering intermittent renewables. To increase their energy density and save costs, improved electrode and cell designs are developed. Furthermore, existing cells are preferably operated at their performance limits, which leads to aging and safety issues. In both cases, there is urgent need for non-destructive operando analysis that provides in-depth and especially quantitative understanding of the physical and chemical processes in the batteries, and how cell design and operating mode impact them. Only voltage and current can be readily measured, and electrochemical impedance spectroscopy, a linear analysis method, is frequently used for this purpose. Yet, various states and effects can cause similar spectra. To address these limitations, nonlinear methods, in particular nonlinear frequency response analysis (NFRA), is used in this project. The analysed higher harmonic responses contain information on nonlinear processes, especially the highly nonlinear reaction kinetics, and on surface states. NFRA is thus an extremely powerful and promising tool to diagnose batteries, but currently barely understood. Presently, there is no common understanding how NFRA spectra should be measured and - more importantly - analysed. These are main obstacles hindering broader acceptance of this method in the community. In this project different NFR approaches, developed at KIT and MPI will be employed for battery analyses. For the first time, NFR spectra of anode and cathode will be studied to discriminate between their contributions in the cell signal; further, the previously not analysed phase information and DC component is evaluated regarding information content. Higher order frequency response functions, which analogously to impedance are not amplitude-dependent but contain information on the nonlinear response will be determined. Analytical and numerical solutions of models will be used to reproduce and interpret the experimental spectra. Resultingly, the project establishes a systematic methodology for optimal extraction of battery (state) information by nonlinear frequency response analysis, and paves the way for a wide-spread analysis of this new technique.