rapid mapping of lithiation dynamics in transition metal oxide particles with operando x-ray absorption spectroscopy

by:Yucai     2019-12-18
Since the commercial lithium ion batteries (LIBs)
Layered transition metal oxide (
LiMO2, where m = u2009 Co, Mn, Ni or its mixture)
Has become the preferred material for LIB cathode.
The transition metal changes its oxidation state during the cycle, and this effect can be detected by X-
Ray absorption near the edge structure (XANES)spectrum. X-
Ray Absorption spectrum (XAS)
Thus, it can be used to visualize and quantify the lithium kinetics in the cathode of transition metal oxides; however, in-
In-situ measurements are usually subject to time resolution and X-
Radiation doses, need to be compromised in the conditions of the electrochemical cycle used or in the materials examined.
We reported a reduction in measurement time and X-
Radiation exposure to the lithium ion battery studied by operando XAS.
Highly discrete energy resolution plus advanced rear
The treatment can identify the oxidation state quickly and reliably. A full-
The field microscope setting provides
The particle resolution on a large area of the battery electrode enables the oxidation state within many transition metal oxide particles to be tracked at the same time.
Here we apply this approach to get a deeper understanding of business, mixing-
Nickel-cobalt alumina metal oxide cathode material (NCA), during (dis)
Charge during overcharge and its degradation.
Experimental by microXAS (X05LA)
The beam line of the Swiss Light Source at Paul Scherrer Institute.
The incoming photon flux at the beam line is about 3 × 10 phs at the sample after passing through the dual-crystalline Silicon monochrome system (Kohzu Co. )
It provides 0 energy resolution ΔE/E. 02%.
After passing through the bag cells, the X-
The Ray hit YAG: Ce-
Doping the evicator, where X-
Convert ray photons to visible light, zoom in 10-
CCD for folding and image (pco.
2000,2048 × 2048 pixels, 7.
The pixel size is 4 um and 6 rms e rms at 10 mhz).
This configuration provides View fields for ca.
1500x1500, a pixel corresponding to about 0. 74u2009×u20090. 74u2009μm.
Calibration and reference measurement for nickel standards (
Nickel foil, nickel and nickel)
The NCA electrode high resolution XANES scan with full charge and discharge was obtained from 8 scans. 31u2009keV to 8. 5u2009keV in 0.
In the fluorescence mode, the 5 ev step near the edge and the 2 ev step above the edge (
Silicon Drift diode (SDD)
Single element detector (
Ketek Limited, Germany))
And in Athena IFEFFIT ().
For measurements, take images under 12 energy: 8. 31 and 8.
322kev thousand electronic volts in the pre-edge region; 8. 343, 8. 346, 8. 350, 8. 353, 8. 356, 8. 360, 8. 372, and 8.
388 KV in the near-edge area; and 8. 447 and 8.
460f Kevin for calibration above the edge.
A total of 20 images were taken at each energy, with exposure times of: 10 samples per 15 MS, and then 10 samples were moved out of the beam (flats).
This results in a sample exposure time of 1.
5 s, with the time to move the sample out of the beam, the total time for each energy is 50 s.
Therefore, the current measurement is limited by the sample movement time and the camera dwell time.
In future experiments, the measurement time can be increased to less than a minute.
Image processing is done with Fiji Image j.
10 Sample images are divided by the average unit to explain the fluctuation of the beam.
Align the mean image through scale invariant feature transformation (SIFT)
Algorithm to correct any shift caused by sample movement, x-
Phase Y, or X-
Light source and optics.
To verify this method, 12-
The point method is compared to a fine scan of 47 images taken with 0.
5 ev step size and 2 ev step size from 8 above the edge. 3 to 8. 46u2009keV. Full-
Processing on-site transmission images in MATLAB. A particle-
Choose free area to calculate non
NCA components for bag batteries: aluminum foil, adhesive, carbon black, electrolyte, PE separator, multi-layer foil and lithium ().
For example, the aluminum foil contains traces of iron, which is absorbed in the energy window of interest and needs to be corrected.
Then subtract this background attenuation from all).
In order to associate pixels with particles, we set the standard that the edge absorption difference must be four times higher than the background noise.
Then, the absorption spectra of each pixel are normalized by subtracting the pre-matched linear fitedge region (8. 310 & 8. 322u2009keV)and the post-edge region (8. 447 & 8. 460u2009keV)
And determined the different pre-edge and post-edge to 1.
This eliminates the influence of the thickness of different active materials on the electrode.
A good correspondence between the gray-scale transport image of the particle dark and the SOC graph showing the false color particles confirms the validity of the method ().
According to Equation 1, the discrete spectra conform to the references.
For each pixel, different combinations of reference Ni, Ni and background contributions are considered in the 2% steps.
The best fit in 1326 possible combinations is determined by selecting a combination of the values representing the ratio of the reference state representing the minimum difference to the normalized absorption data)
For all the energy, index according to the scan situation, where the values = u2009 12 or 47. We need an R-value below 0.
02 accept the combination of curves as fit.
The representative diagram of the fitting value is given in.
Three types of maps convey the location information of lithium in the sample.
Oxidation state diagram (e. g. c and b)
Give the percentage of Ni (100 × )
In the material slice of the image in a given pixel. SOC maps (e. g. b and b)
By drawing the fraction of the colored and de-colored materials that appear in a given pixel, the materials can then be plotted linearly on a scale of 0 to 1 to show the local SOC.
Subtract the graph (e. g. )
Shows the SOC difference between the two time steps and has a value starting from-1 (less lithiated)to 1 (more lithiated).
Custom message