Thesis presented December 20, 2022

Abstract:
Here, we propose an original battery separator that allows to free oneself from the need to maintain Lithium Metal Polymer batteries at high temperatures (80°C). Freeing oneself from this constraint is a major challenge, particularly for the automotive industry.
To date, lithium-ion technology is a mature and widely used technology. However, it has some limitations, in particular relatively modest energy densities and two major safety risks: i) a phenomenon of dendritic growth of the lithium can lead to short-circuiting of the two electrodes which can induce spontaneous combustion of the device and i) environmental risks in the event of assembly failure and dissemination of lithium hydroxide.
Equipped with pure lithium metal anodes, the All-Solid batteries allow to reach high energy densities. A canonical electrolyte is Polyethylene Oxide (PEO) doped with lithium salts. These are known as "Lithium Metal Polymer" (LMP) batteries. The ionic conductivity of the electrolyte is closely linked to the dynamics of the polymer chains, which act as a solid solvent. To ensure the mechanical strength of the assembly, high molecular weight POE (35 kg.mol^-l) is used, which in its creep regime has a high viscosity (n) that limits the conductivity of the lithium (Stokes-Einstein relation: the diffusion coefficient is as 1/η).
This technology faces a strong physical constraint: for good conduction performances it is necessary to maintain the battery at 80°C, a temperature higher than the melting point of POE bulk (60°C).

In this context, we propose a battery separator consisting of a composite polymer membrane whose porosity is ensured by macroscopically oriented Carbon NanoTubes (CNT). This device allows to combine several effects:
- The mechanical strength of the device is ensured by the membrane, the use of liquid POE (500 g.mol^-1) at room temperature. Its low viscosity allows rapid diffusion of lithium ions
- One-dimensional (1D) ionic conduction from one electrode to the other through the pores.
- A significant lowering of the melting point by the Gibbs-Thomson effect, i.e. shifting T_M. the melting point of a material towards low temperatures according to ΔT_M≃ 1/d where d is the characteristic size of a pore.

We characterise the membrane morphology by a coupled Scanning Electron Microscopy (SEM) and Small Angle Neutron Scattering (SANS) study. We show, by neutron imaging, that the CNT core allows an efficient transport of the electrolyte from one side of the membrane to the other. The dynamics of the POE is probed from the molecular to the mesoscopic scale by Quasi-Elastic Neutron Scattering (QENS) and Nuclear Magnetic Resonance (NMR). Conductivity is characterised by i) Electrochemical Impedance Spectroscopy (EIS) and cycling on a Swagelok assembly ii) an alternative method in an electrochemical cell on a membrane of which only a small fraction of the CNT are made accessible by a FIB (Focused Ion Beam) etching.
Thanks to a factor of 3 gain in conductivity offered by the 10 confinement and the use of liquid PEO, we equal at room temperature the conductivity offered by PEO 35 kg. mol^-1 at 80°C. We also obtain a lowering of the melting point of the confined electrolyte by 17 K. This work provides proof of concept that our proposed battery separator is relevant as a separator for All-Solid Lithium-Metal-Polymer batteries functional down to 5°C.

Keywords:
Electroyte, Carbon NanoTube, battery, caracterisation