Solid Electrolyte Formation in LI-Metal Batteries and LIFSI/TMP Electrolyte
Abstract
Rechargeable Li-ion batteries (LIB) are the most popular devices for energy storage but still a lot of research needs to be done to improve their cycling and storage capacity. LIB feature energies densities in the range of 100-265 Wh/kg which is very low if compared with gasoline in which the range is in the order of 12,000 Wh/kg. Therefore, Li-metal has been proposed as an anode material because the energy density of the battery could increase up to 2,600 Wh/kg for a Li-Sulfur battery and to 3,458 Wh/kg for a Li-air (O₂) battery. With the addition of Li-metal as an anode material a new set of batteries called lithium metal batteries (LMB) can be developed with the potential to increase the cell-level energy of the LIBs. Therefore, focus is needed on the lithiation process of Li-metal anodes where it is known the mechanical, electrochemical, and electric phenomena such as cracking, SEI formation and ionic-clustering, respectively, that occur during the charge/discharge cycles.
Performing molecular dynamics simulations of an electrolyte comprising trimethyl phosphate (TMP) solvent and a lithium bis(fluorosulfonyl)imide (LiFSI) salt, the effects of salt concentration on solvation and ion-transport are explored. Three LiFSI-TMP electrolyte salt concentrations of 0.7, 1.43 and 3.82 molar are simulated. A statistical analysis was performed to study ion-pairing, clustering, diffusivity, conductivity, and coordination of Li-ions, providing insights into relations between molecular structures and transport properties. Molecular structure of ionic components changes as concentration increases, from a predominant solvent separated ion pair (SSIP) and contact ion pair (CIP) to aggregate (AGG) salt and ionic cluster formation. The formation of ionic clusters suggests that the diffusion mechanism of Li-ions changes from a hopping/exchange to a vehicular mechanism as concentration increases; this is validated by a decrease of ionic conductivity. Ionicity was also calculated to reveal how the ionic motion changes from an uncorrelated to a correlated one as the salt concentration increases.
Identifying the mechanism of SEI formation at electronic and atomic levels is especially important to understand how the SEI formation affects the overall battery performance such as the decrease of active material, decrease of cell potential, and interfacial stability. Ab initio molecular dynamics simulations were performed for Li⁺-conducting electrolytes based on trimethyl phosphates (TMP) and lithium bis(fluorosulfonyl)imide (Li⁺FSI⁻) salt in contact with a Li-metal electrode. We focused on the transient-state behavior at the electrolyte, interfacial electrolyte−Li-metal electrode, and lithium reference electrode−electrolyte−Li-metal electrode to study dynamics and activation energy barriers of the Li⁺ ion, electrochemical and thermal stability of the interface electrode−electrolyte, and potential behavior of the Li-metal electrode, respectively.
An interfacial study is performed using ab initio molecular dynamics simulations to elucidate the solid electrolyte interphase (SEI) evolution formed between an electrolyte based on trimethyl phosphates (TMP) and lithium bis(fluorosulfonyl)imide(Li⁺FSI⁻) salt in contact with a Li-metal electrode. Going beyond the initial SEI composition generated due to the degradation of one counter-ion adding a second and third counter-ions ana analysis of how the initial SEI evolution is performed. The results indicate a different product formation due to the LiFSI salt dissociation as the SEI is formed. The products formed due to the dissociation of the 1st LiFSI salt when in direct contact with the Li-metal anode are Li₂O, Li₂S, Li₃N and LiF. These four Li-binary products compose the formed SEI. Then, a 2nd LiFSI is located at the electrolyte/SEI/Li-metal. The products formed due to the dissociation of the 2nd LiFSI when in contact with the SEI are Li2S, Li₂O, LiF, Li₃NSO₂. Finally, a 3rd LiFSI is located at the electrolyte/SEI/Li-metal. The products formed due to the dissociation of the 3rd LiFSI when in contact with the SEI are Li₂SO₂NSO₂ and LiF.
Computational techniques such as molecular dynamics (MD) simulations can simulate a large number of atoms, in the order of 10⁵ interacting through their forcefields. A nanobattery MD model is an accurate, yet simple model to study electrochemical phenomena occurring in in any rechargeable battery. A regression machine learning algorithm is proposed to overpass paramount timescale limitations of any atomistic MD model. The primary limitation of the nanobattery model is the extremely short charging time compared to the longer charging time in a real battery. Using data from several macro-scale commercial Li-ion batteries, and a nanobattery MD model, we constructed a scaling regression algorithm to scale the values obtained from the nanobattery MD model to a macro-scale Li-ion battery. The goal is to demonstrate that three transport properties: 1) the time, tLi, a Li-ion spend to travel from cathode to anode; 2) the superficial density frequency of arrival Li-ions, ALi (s⁻¹A⁻²); and 3) the frequency, fLi, of Li ions arrival to the anode, can be incorporated in one model that could predict the macro-scale variables having as an input the nano-scale variables.
Citation
Galvez Aranda, Diego Eduardo (2022). Solid Electrolyte Formation in LI-Metal Batteries and LIFSI/TMP Electrolyte. Doctoral dissertation, Texas A&M University. Available electronically from https : / /hdl .handle .net /1969 .1 /198125.