Beam Effects in Synchrotron Radiation Operando Characterization of Battery Materials: X-Ray Diffraction and Absorption Study of LiNi0.33Mn0.33Co0.33O2 and LiFePO4 Electrodes

Operando synchrotron radiation-based techniques are a precious tool in battery research, as they enable the detection of metastable intermediates and ensure characterization under realistic cycling conditions. However, they do not come exempt of risks. The interaction between synchrotron radiation and samples, particularly within an active electrochemical cell, can induce relevant effects at the irradiated spot, potentially jeopardizing the experiment’s reliability and biasing data interpretation. With the aim of contributing to this ongoing debate, a systematic investigation into these phenomena was carried out by conducting a root cause analysis of beam-induced effects during the operando characterization of two of the most commonly employed positive electrode materials in commercial Li-ion batteries: LiNi0.33Mn0.33Co0.33O2 and LiFePO4. The study spans across diverse experimental conditions involving different cell types and absorption and scattering techniques and seeks to correlate beam effects with factors such as radiation energy, photon flux, exposure time, and other parameters associated with radiation dosage. Finally, it provides a comprehensive set of guidelines and recommendations for assessing and mitigating beam-induced effects that may affect the outcome of battery operando experiments.


Dose estimation:
The dose as defined in equation ( 1) is specific for each electrode material and cells configuration and depends on the photon flux (F), Photon energy (Ephoton), exposure time (t), transmittance at a given energy (T), the beam spot size (A), the electrode thickness (e) and density (d).
In order to calculate the photon flux measurements with a 10 µm thick Si photodiode of 10 x 10 mm 2 area were taken at the sample position using the same beamline configuration as for the coin cells measurements (including or not the different attenuators).In this way, the flux attenuation due to all the different elements along the beamline was considered.Furthermore, all the elements present in the coin cell prior to reaching the electrode were taken into account.The transmission of 75 µm of Kapton or 200 µm of Be from the cell window was calculated from XOP software.In the same way, the transmission through the different attenuators (100 µm or 250 µm Al foil) used in the beam optics path, prior to reach the sample, was also calculated using XOP and was compared with the direct measurements using the Si photodiode with and without those filters, validating the XOP calculations.XOP was also used to estimate the attenuation related to the absorption of the separator and the electrolyte.In this case, the transmission of 420 µm of SiO2 with 50% porosity with half of the thickness filled with electrolyte (210 µm) was considered to represent and the absorption of the electrolyte was estimated from its formulation.The transmission of the active material was calculated with XOP software based on the electrode composition and porosities specified by electrode supplier (NEI corporation).

Figure S1 .
Figure S1.Schematic layout, zenithal view and image of the setup of the operando cells Kapton window coin cell (a), Leriche 1.0 (b) and LeRiChe'S V2 (c).View of the carrousel with 4 coin cells mounted on the sample stage at NOTOS beamline (d) and a schematic representation of the protocol followed during combined XRD and XAS experiments involving measurements at two different spots.

Figure S3 .
Figure S3.Operando X-ray diffraction patterns (zoom in on reflection 003) of NMC111 (a) and LFP (b) acquired using Leriche operando cells with Be window corresponding to the first charge cycled in galvanostatic mode at C/10 rate.Measurements were conducted under continuous synchrotron light exposition at 5 spots in the cell.At four positions bothPXRD and EXAFS of  (Mn, Co, Ni)  and Fe were performed for NMC and LFP, respectively.Due to small distance between points and error in measurement macro dose per position can't be reliably assessed.

Figure S2 .
Figure S2.Operando X-ray diffraction patterns (center), X-ray absorption spectra (right) and electrochemical curve (left) of NMC111 (a) and LFP (b) electrodes measured in Kapton window coin cells.(The cells were continuously exposed to the synchrotron X-ray beam for the sequential acquisition of the XRD patterns and XAS spectra).

Figure S4 .
Figure S4.Operando X-ray diffraction patterns (zoom in on reflection 200) of LFP at spots A (up) and B (down) and corresponding electrochemical curve acquired in similar conditions in Leriche 1.0 cells cycled at C/15 (a) and in Kapton window coin cells cycled at C/7 (b).Spot B in cell (b) was exposed to the beam for 3 min after each 1 min PRDX at 11 keV.Cell (a) starts in discharge due to error in macro during the first charge, then the cell was kept at constant voltage (4.5 V) off the beam for 3h prior to discharge and measurement was conducted at a fresh spot.Dots marked in the electrochemical curve indicate measurement points.

Figure S6 .
Figure S6.Operando X-ray diffraction patterns (zoom in on reflection 003) of NMC111 at spots A (up) and B (center), together with the corresponding X-ray absorption spectra measured only at spots B (down) and their corresponding electrochemical curve acquired using Kapton window coin cells under different measurement conditions.Exposure time to 11 KeV radiation was 4 times longer per measurement in point A of cell (b) than for cells (a) and (c).At point B only K-edge spectra of Mn, Co and Ni were respectively measured on cells (a), (b) and (c).

Figure S7 .
Figure S7.Ni K-edge energy vs. % of charge capacity for NMC111 electrodes tested: in operando cells with different geometry (a), in coin cells under diverse photon flux and exposure times (b), and in LeRiChe´S v2 cell with different electrode thicknesses and with a cycling protocol involving voltage pulses (c).

Table S1 .
Calculated X-ray radiation dose per measurement