Heat Sources, Thermal Conductivity and Heat Distribution in Electrochemical Energy Storage Technologies

Abstract

The work presented in this thesis is based on taking thermal conductivity measurements of a range of materials used in three different electrochemical energy storage technologies. Specifically, two hydrogen technologies, the proton exchange membrane fuel cell (PEMFC) and the water electrolyzer (WE) as well as a secondary battery system, the lithium-ion (Li-Ion) battery, were investigated. All of these are based on a variation of a two-electrode process. A new measurement rig was constructed and automated to produce ex-situ experimental through-plane thermal conductivity values of high accuracy for porous and compressible materials. The new setup enabled measurements under conditions that have not been reported in the literature before, as the samples could be saturated with liquids or different gases during measurement.

The three technologies that were studied in this work are today based on the use of materials that are porous to a greater or lesser extent. Porous materials enable different simultaneous modes of transport through them. While a medium, gaseous or liquid, may be transported through the void space of the pores, electrons or ions may be transported through the solid phase. On top of that, heat can be transported through the material, in an adverse direction to the other two modes. The original substance that the porous material is based on governs the electrical and thermal conductivities while the transport properties of the void space can be manipulated by choosing the pore size and the overall porosity. Additionally, other substances can be introduced into the pores to enhance their transport properties even further. Porous materials respond differently to the application of mechanical pressure. Particle porous materials such as microporous layers (MPLs) and catalyst layers (CLs) are compressed more easily and plastically whereas fibrous porous materials such as gas diffusion layers (GDLs), sometimes also called porous transport layers (PTLs), compress less and elastically, almost regaining their pre-compression thickness after the mechanical stress is removed. This had to be considered and thus compression was one of the variables the new measurement rig could manipulate and monitor. A range of porous materials used in PEMFCs and WE were experimentally investigated under novel conditions and the results reported for the frst time in the literature. Thermal conductivity values for custom-made PEMFC GDL/MPL composite materials ranged from 0.14 to 0.4 WK-1m-1 depending on the base GDL material used. The thermal conductivity of two commercial PEMFC GDL materials increased by 7 -18% when saturated with hydrogen gas as compared to air. The graphitization of a PEMFC catalyst material was measured to cause a doubling of its thermal conductivity from 0.06 to 0.12 WK-1m-1. Four different titanium-based sintered transport layer materials for a proton exchange membrane WE (PEMWE) were measured to have thermal conductivities between 1.0 and 2.5 WK-1m-1.

The technologies that were investigated also share the need for some sort of separator in-between the two electrodes. In case of the hydrogen technologies a mechanically and chemically stable membrane is preferred. It is electronically insulating while it has high ionic conductivity. For the Li-Ion battery a liquid electrolyte soaked separator is commonly used, with ambitions to replace it by a solid electrolyte in the future. Being the centerpiece of these technologies, the thermal conductivity of these separators is of key interest. There are a number of studies reporting on the issue. This work was able to supplement the available data, especially for water electrolyzers and Li-Ion batteries. The HMT-PMBI-Cl- anion exchange membrane (AEM) studied has a thermal conductivity of around 0.2 WK-1m-1 in dry and wet conditions. Thermal conductivity for the sintered solid state electrolytes (SSEs) investigated was measured to around 0.5 WK-1m-1, while an unsintered SSE material was measured to have a thermal conductivity of around 0.2 WK-1m-1.

On the basis of the obtained thermal conductivity data, a discussion of the thermal signature of these technologies was enabled. The heat sources were decomposed and allocated to their respective material layer so as to gain an understanding of how the heat and temperature distributions emerge. These considerations led to the creation of a two-dimensional model of the heat flux for each of the technologies. The heat and temperature distribution in a single cell could then be simulated computationally. The results give thought-provoking impulses and insight as to where further and more detailed investigations of the thermal signatures are needed. Temperature differences of more than 14 K were shown to arise in PEMFC. Additionally, the results suggest an asymmetric heat distribution in the cell. With the WE models temperature dierences reached more than 7 - 17 K for the PEMWE depending on the heat source modeling and more than 18 K for the AEM WE. Also here the heat flux distribution is asymmetric, calling for different cooling needs for the two electrodes. In Li-Ion battery stacks temperature differences of up to 5 K between the center and the outer edge of the stack were shown to arise when using SSEs instead of traditional separators.

As a result, the work in this thesis demonstrates the increasing need for heat management in electrochemical energy storage technologies. As their energy densities increase further through research and development, the heat densities grow as well. Handling this heat will require more attention in the future to ensure safe operation and long lifetime of the commercial products based on the discussed technologies. The three mentioned established electrochemical energy storage technologies are investigated to determine the thermal conductivities of the materials involved and to identify the thermal gradients within them. Now it is upon others to increase the field of knowledge by implementing these results to understand better what these temperature gradients mean in detail.