Thermoelectric Properties of Non-Aqueous Electrochemical Cells - Local Reversible Heat Effects Relevant for Lithium Ion Batteries

The aim of this thesis was to investigate the thermoelectric properties and local reversible heat effects of cells relevant to lithium-ion batteries. The theory of non-equilibrium thermodynamics and experimental measurements of the thermoelectric power of these cells were used to study the system. The Seebeck coefficients, Peltier heats and transported entropies of Li+ in the cells have been estimated. The three different thermoelectric cells studied consisted of lithium cobalt oxide (LiCoO2), graphite (C6) and lithium iron phosphate (LiFePO4) electrodes and separators wetted with the same ternary electrolyte with lithium hexafluorophosphate (LiPF6), ethylene carbonate and diethyl carbonate. The time-evolution of the thermoelectric power measured indicated that two thermal diffusion phenomena of different time scales were present during the measurements for the LiCoO2 and LiFePO4 cells. These Soret effects in the electrolyte were found to contribute significantly to the electric potential. For LiCoO2, the Peltier heat at the electrode was estimated to be 84 +- 9 kJ/mol initially, -45 +- 6 kJ/mol at an intermediate state and 51 +- 6 kJ/mol at stationary state. For LiFePO4, the Peltier heat at the electrode was estimated to be 35 $\pm$ 9 kJ/mol initially, 26 +- 6 kJ/mol at an intermediate state and 127 +- 12 kJ/mol at stationary state. These values indicate large and time-dependent local reversible heat effects in Li-ion batteries during charging and discharging. These local heat effects are currently not included in thermal description of lithium-ion battery cells given in the literature. The transported entropy of Li+ in the LiCoO2 cell was found to be 200 +- 20 J/K mol, while in the LiFePO4 cell it was estimated to 460 +- 60 J/K mol. The discrepancy in S*_Li+ could indicate that the transported entropy depends on the electrode material in the thermoelectric cell. No consistent behaviour was observed for the C6 cells, probably due to irreversible changes inside the cells during the measurements. The large changes and uncertainties in the Seebeck coefficient and bias potential over time can indicate that structural changes occurred in the cell during and between the measurement