An important element in the reduction of CO₂ is the change of vehicles with internal combustion engines to electric battery powered vehicles. The as such produced renewable energy can be used for individual mobility as well as for a temporary intermediate storage of excess energy. A viable electric mobility concept requires however stable cycle batteries with high specific energy (minimising weight, maximising driving range).

Li ion batteries are widely used in applications such as mobile phones and laptops, and will likely be key to future electromobility. An alternative promising battery is the lithium sulfur battery with a potential twofold energy density increase. The requirements for such batteries present major challenges; e.g. energy capacity, deactivation/stability and safety. A detailed understanding of the charge, discharge and deactivation mechanisms are thus required, preferably quantitative and spatially resolved. X-ray absorption spectroscopy (XAS) is a characterisation technique which provides detailed electronic and structural information on the material under investigation, in a time- and spatially resolved manner.

Here, I will explain the strengths and limitations of XAS for battery research. A novel operando XAS cell design will be described [1], including the challenges to perform reliable experiments (electrochemically and spectroscopically). The cell allows time and spatial resolved XAS, providing insights in the type, location and reversibility of the intermediates formed in electrodes and electrolyte separately. Obtained insights in cycling and deactiviation mechanisms for the different battery types will be discussed [1-6] and future research directions described.

[1] Y. Gorlin, A. Siebel, M. Piana, T. Huthwelker, H. Jha, G. Monsch, F. Kraus, H.A. Gasteiger, M. Tromp, J. Electrochem. Soc. 162(7): A1146-A1155, 2015.

[2] Y. Gorlin, M. U. M. Patel, A. Freiberg, Q. He, M. Piana, M. Tromp, H. A. Gasteiger, J. Electrochem. Soc. 2016, 163(6), A930-A939.

[3] J. Wandt, A. Freiberg, R. Thomas, Y. Gorlin, A. Siebel, R. Jung, H. A. Gasteiger, M. Tromp, J. Mater. Chem. A 2016, 4, 18300-18305.

[4] A. T. S. Freiberg, A. Siebel, A. Berger, S. M. Webb, Y. Gorlin, H. A. Gasteiger, M. Tromp, J. Phys. Chem. C 2018, 122, 10, 5303-5316.

[5] A. Berger, A. T. S. Freiberg, R. J. Thomas, M. U. M. Patel, M. Tromp, H. Gasteiger, Y. Gorlin, J. Electrochem. Soc. 2018, 165(7), A1288-A1296.

[6] R. Jung, F. Linsenmann, R. J. Thomas, J. Wandt, S. Solchenbach, F. Maglia, C. Stinner, H. A. Gasteiger, M. Tromp, J. Electrochem. Soc. 2019, 166(2): A378-A389.

A set of sodium iron titanite samples with general formula Na_xFe⁺²_x/2Ti_2–x/2O₄ was prepared using solid-state synthesis in an inert atmosphere to test for application as cathode materials for Na-ion batteries. These materials have several advantages over analogues with Fe³⁺, demonstrating better sodium ion conductivity and higher Na+ ions capacity without phase transition or destruction of the structure. [1] In the course of the investigation, several compositions of a new compound were obtained with NSIT-like structure type similar to Na_0.9Fe³⁺_0.9Ti_1.1O₄. Among them, composition with x = 0.9 was selected due to its electrochemical performance and structural peculiarity. From the crystallographic point of view, formation of phases in Na_xFe⁺²_x/2Ti_2–x/2O₄ system with Na_0.9Fe³⁺_0.9Ti_1.1O₄ structure (having Fe³⁺ ions mixed with Ti⁴⁺ ions) is rather unusual due to different radii of mixing ions (Fe²⁺ = 0.92 Å, Fe³⁺ = 0.785 Å, Ti⁺⁴ = 0.745 Å (CN=6) [2]).

The Na_0.9Fe³⁺_0.9Ti_1.1O₄ was further studied by operando XANES spectroscopy. The sample was placed as a cathode inside custom electrochemical cell with glassy carbon X-Ray transparent windows. Li foil was used as anode and 1M LiPF6 in 1:1 EC:DMC commercial solution (Sigma) was used as electrolyte. The cell was cycled in 1.6 to 4.5 V range with current of C/20. Operando Fe K-edge XANES spectra were measured with the R-XAS Looper (Rigaku, Japan) laboratory X-Ray absorption spectrometer.

In total 200 spectra were collected for Na_xFe⁺²_x/2Ti_2–x/2O₄ sample with x = 0.9 during 10 consecutive cycles, which were further analyzed by PCA to extract spectra of phases participating the electrochemical process and corresponding phase content diagrams. 2 components were successfully extracted (fig. 1). Component corresponding to a Fe²⁺ phase shows good agreement with FeTiO₃ reference in terms of absorption edge position and overall profile of the spectrum. On the other hand, agreement with the spectrum of as-prepared sample is far from decent. Component corresponding to a Fe³⁺ phase shows good agreements with a reference compound (reference sample, fully oxidized in air). Comparison with theoretical spectra for various structural models have shown decent agreement of Fe²⁺ phase with the spectrum of freudenbergite. Fig. 2 shows the cell potential and phase concentrations from PCA as a function of time. One can clearly see the decrease in Fe²⁺/Fe³⁺ conversion rate during first 3 cycles, after that the rate remains stable and conversion is highly reversible. This decrease in conversion rate, as well as lack of agreement between Fe²⁺ phase from PCA and as prepared sample, might be accounted by electrochemical substitution of Na with Li that takes place during cycling in Li-based cell, which causes the change in local atomic and electronic structure of material, possibly leading to partially blocked or collapsed ion transfer channels. The degree and details of such substitution are subject for study by operando XRD and Mössbauer spectroscopy.

[1] Rajagopalan et al., (2017) Adv. Mater. 29

[2] Shannon, R.D. (1976), Acta Crystallogr. Sect. A 32, 751-767

Authors would like to acknowledge the financial support of Russian Foundation for Basic Research in the framework of grant 20-32-70227