Speaker
Description
Despite having been used as anode materials in commercial Li-ion batteries for decades, Lithium-graphite intercalation compounds (Li-GICs) are still not sufficiently understood at the atomistic level. This difficulty is, in part, due to meso-scale domain ordering and non-equilibrium phenomena, the investigation of which by predictive-quality first-principles methods like by density-functional theory (DFT) would incur prohibitive computational cost.
To overcome the computational bottleneck, we developed a complete model based on density-functional tight-binding (DFTB) with a machine-learned repulsion potential [1], which matches state-of-the-art dispersion corrected DFT in performance, at a fraction of the cost and for the entire range of states of charge (SOC) [2]. This model allows us to fully relax supercells with thousands of atoms, assess the stability of the different compositions appearing during the charging cycle, and compute elementary jump rates at virtually any SOC and local configuration. Further, it naturally includes long-range electrostatic interactions, allowing for the extraction of dielectric properties as a function of SOC [3]. In ongoing work, the elementary Li migration barriers, as well as the dielectric permittivities are now used as inputs for charged kinetic Monte Carlo (kMC) models to resolve the meso-scale diffusion dynamics under operational conditions.
In order to experimentally validate the kMC models, a suitable, well defined reference is required. As such, we decided to investigate the upper limit of lithium intercalation in the morphologically quasi-ideal highly oriented pyrolytic graphite (HOPG), with a LiC6 stoichiometry. This corresponds nominally to 100% SOC and is the best understood configuration of Li-GICs. We prepared a sample by immersion in liquid lithium at ambient pressure and investigated it by static 7Li nuclear magnetic resonance. Surprisingly, this resolved unexpected signatures of so-called “superdense” intercalation compounds, LiC6-x. The latter have been ruled out for decades, since the highest geometrically possible composition, LiC2, can only be prepared under high pressure. However, the inaccessibility of intermediate overlithiated compositions was never substantially verified. Our work thus challenges the widespread notion that any additional intercalation beyond LiC6 is not possible under ambient conditions. Taking spontaneous overintercalation into account may shed light onto some hitherto poorly understood phenomena, such as the partial reversibility of plating, transport limitations during fast charging, and the apparent loss of a fraction of lithium between charging cycles.
[1] C. Panosetti, A. Engelmann, L. Nemec, K. Reuter, and J.T. Margraf, J. Chem. Theo. Comp. 16, 2181 (2020).
[2] C. Panosetti, S.B. Anniés, C. Grosu, S. Seidlmayer, and C. Scheurer, J. Phys. Chem. A 125, 691 (2021).
[3] S. Anniés, C. Panosetti, M. Voronenko, D. Mauth, C. Rahe, and C. Scheurer, Mater. 14, 6633 (2021).
| Abstract Number (department-wise) | TH 20 |
|---|---|
| Department | TH (Reuter) |
