EELS Determination of Li Distribution and Fe Valence Mapping in Lithiated FeOF/C Nanocomposite Battery Materials

TLDR: In this article, the authors used scanning transmission electron microscopy (STEM) combined with EELS to determine the Li spatial distribution, its chemical state and the Fe valence state in FeOF/C nanocomposite electrodes during charge and discharge processes.

Abstract: Amongst the techniques to investigate Li-ion battery materials, electron energy loss spectroscopy (EELS) play a unique role as Li distribution, chemical state and valence of transition metals (charge transfer) can all be determined with nanometer scale spatial resolution. In this work, we use EELS to investigate new positive electrodes for Li-ion batteries based on transition metal fluoride (FeF3, FeOF, FeF2, CuF2....)/C nanocomposites [1]. The high specific capacity in these new electrodes is obtained by using all the oxidation states of Fe from Fe to Fe during discharge cycles via a complete conversion process. In this study, we used Scanning Transmission Electron Microscopy (STEM) combined with EELS to determine the Li spatial distribution, its chemical state and the Fe valence state in FeOF/C nanocomposite electrodes during charge and discharge processes. This STEM-EELS analysis was done using a JEOL 2010F equipped with a Gatan GIF 200 spectrometer and with a Hitachi 2700 STEM equipped with an Enfina spectrometer. In order to minimize electron beam damage and F loss, the samples were cooled to LN2 temperatures and imaged with a total electron dose not exceeding 10 C/cm. Both lithiated (discharged) and delithiated (re-charged) FeOF/C nanocomposites electrodes were analyzed by EELS. The Fe valence state was obtained by measuring the Fe L3/L2 intensity ratio [2,3]. The L line intensities were obtained using either a 4.5 eV window or by taking the positive component of the EELS spectra second derivative. An ADF STEM image of a FeOF/C cathode material discharged to 1.5V is shown in Fig.1a with the corresponding low energy EELS signal (c.f. Fig. 1b) taken from area marked A revealing the superposition of the Li-K and Fe-M edges. The extracted Li-K edge has two prominent peaks whose energies are separated by 6.6 eV. In addition to the two prominent Li peaks, there is a third one located at a distance of 4.2 eV from the first peak. The existence of these peaks is indicative of the presence of two Li-base compounds (LiF) and a new Li-Fe-O-F cubic phase. The Li-K/Fe-M intensity map shown in Fig.1c from the area depicted in Fig.1a reveals the presence of Li and Fe rich phases with a spatial distribution in the 3-5 nm range. At this voltage the expected phases are LiF+Fe+LixFeOyFz [3]. At the surface, a 10-20 nm thick Li rich phase (c.f. Fig 1d) is observed corresponding to a mixed LiF-Li2CO3 solid electrolyte interface (SEI) surface layer. Upon lithiation, the Fe valence state decreases as represented by a decrease in the Fe-L3/L2 intensity ratio. At the lowest voltage of 0.8V, all Fe is in the metallic state. At the intermediate discharge voltage of 1.5 V, the Fe valence state is not uniform and the microstructure is composed of a mixture of high and low valence state phases as depicted in Fig. 2b. The O-K concentration map shown in Fig.2a has a similar distribution as the valence map of Fig. 2b which indicates that the oxygen rich phase is also the phase with highest valence state. A quantitative analysis of the Fe L3/L2 intensity ratios using standard model compounds (Fe, FeF2 and FeOF) as reference indicate a Fe valence state of 2.3 for this cubic LixFeOyFz phase. Upon recharge to 4.5 V, all the Fe in the electrode returns to its initial Fe valence state with the electrode material converting back to its initial rutile FeOF phase. [4].