The Electrochemical Society hosted Prof. Raphaële J. Clément’s live online webinar, “From order to disorder: NMR insights into ionic conduction in battery materials,” on October 27, 2021. Below are answers to questions posed during the presentation.
Raphaële Clément is an Assistant Professor in the Materials Department at the University of California, Santa Barbara (UCSB), U.S. She received her PhD in Chemistry in 2016 from the University of Cambridge, UK, working under the supervision of Prof. Clare Grey. Her doctoral work focused on the study of layered sodium transition metal oxide cathodes for Na-ion secondary batteries. She then joined Prof. Gerbrand Ceder’s group at the University of California, Berkeley (UC Berkeley), U.S., focusing on cation-disordered rock salt oxyfluorides for Li-ion battery applications. She joined the UCSB faculty in 2018. Her primary research focus is the development and implementation of magnetic resonance techniques (experimental and computational) for the study of battery materials and beyond, with a strong emphasis on operando tools. She is an Associate Editor for Battery Energy, a new open access journal by Wiley.
Q&A
Could you please comment on the hybrid system and if it’s able to combine both advantages?
In the heterogeneous polyzwitterionic liquid electrolyte, the amorphous regions confer ductility and good mechanical properties, while the crystalline domains enhance Li+ conductivity and selectivity.
In a LiI-doped solid system, conductivity is found to be very high. Can it be used as a good electrolyte of Lithium ion battery?
Iodides (including LiI doping) are promising Li+ superionic conductors but their electrochemical stability is extremely limited. Hence, they would likely only be feasible in dual electrolyte designs and/or paired with cathode coatings to ensure good cycling stability.
In ball milled LYC, the conductivity enhancement due to stacking faults is expected to remain after charge/discharge cycles?
We have not tested this. Great suggestion!
Do you think polymeric solid state electrolytes like you’ve developed will allow flexible solid state batteries?
A challenge in applying polyzwitterionic liquids to flexible solid state batteries is that these electrolytes are not elastic over long timescales as a cross-linked network would be. It may be possible to embed these materials in an elastic matrix of some sort that has long term elasticity, but directly cross-linking the materials may lead to a decrease in the ability of the materials to generate the ordered domains that we anticipate are crucial towards their conduction properties. While this challenge may need to be addressed, there certainly could be some potential in this area.
How easy is this solid electrolyte to be assembled in an all solid state battery? Can this kind of SE tolerate the cathode volume expansion?
We have not sufficiently explored this area to provide a good answer to this question. The flexibility and ductility of the amorphous polymer matrix is promising in terms of buffering cathode volume expansion.
On slide 26, showing the improved performance with the NYZrC catholyte, specifically the efficiency, does that come from better stability of the NYZrC against the cathode active material, or due to the difference in Na conduction? What is the conductivity of the NaPS electrolyte compared to the NYZrC?
It comes from the better stability of the NYZrC electrolyte against the NaCrO2 cathode. The conductivity of NaPS (~3×10-3 S/cm for tetragonal phase) is higher than that of NYZrC75 (6×10-5 S/cm).
In the LYC temperature-mediated conductivity tuning, you said LiCl is produced; what is the effect on this directly on the conductivity versus ordering of the crystal?
The conductivity and ordering of the crystal (annealing out of planar defects) is simultaneous and related. At temperatures as low as 60°C, Li defect layers are removed, resulting in fewer Li site linkages in the LYC structure and poorer conduction. The structure also becomes more “ordered” as a result.
How does the phonon influence transport of ions in battery materials?
Phonons can lower the energy barrier for ion diffusion through the lattice. Please check: 10.1103/PhysRevB.56.11593.
On slide 20, how do you separate the effects of 0.05 ppm chemical shift changes from frequency shifts due to slight changes in the magnetic susceptibility? Do you use an internal standard?
We did not use an internal standard. Since all of these samples have similar compositions and we do not expect a significant change in particle shape, we attributed the change in chemical shift to a change in the bulk structure. This is consistent with results obtained on the solid state LYC sample (more negative 7Li shift) and first principles CASTEP calculations. Besides, our assignment to Li with a greater number of next nearest neighbor Y atoms is in good agreement with the loss of Li only defect layers observed by synchrotron XRD and the formation of LiCl at elevated temperature.
On slide 34, it was indicated that size disparity leads to higher crystallinity in the polymer. Can you elaborate on this?
We generally found that the anion size is important for crystallization and that the fluorous nature of the anion also seems to play a role.
Was the x-ray experiment carried out without exposing to the air?
Assuming that this question refers to the polyzwitterionic work, the samples were exposed to x-rays in vacuo.
As compared to oxide/phosphite-based solid electrolytes, LiYCl-based solid electrolytes are not stable chemically as well as moisture, so for real application, which will be better to go with?
LYC is not thermally stable nor stable against moisture. While the former is an issue for commercial applications, the latter can be circumvented through dry room processing since LYC is stable in dry air.
EIS is supposed to measure bulk conductivity and grain boundary conductivity separately and therefore grain boundary activation energy can be determined. Did you observe the grain boundary activation energy (intergrain conduction) is higher than the bulk activation energy (intragrain conduction)?
We were unable to separate intergrain and intragrain contributions for LYC. In fact, all reports on this system have failed to separate these contributions through EIS.
