Abstract
A number of non-aqueous electrolyte formulations based on sodium hexafluorophosphate (NaPF6) were evaluated for performance in activated carbon supercapacitor devices, with the best performer found to be NaPF6 in a propylene carbonate-dimethoxyethane solvent mixture of 8:2 by mole fraction, yielding a stable specific capacitance of 63.19 Fg-1. This system was also shown to suffer the least amount of initial capacity loss of 25% with a final energy efficiency of 58%. In addition, this system was compared alongside a number of other electrolyte salts in the same solvent system including tetraethylammonium (TEA+) and lithium analogues, whereby it was shown that the initial system possessed a superior final discharge capacitance upon cycling, albeit at the expense of lower energy efficiency when compared to the tetraethylammonium, lithium and other sodium-based systems. However, the systems based on group 1 metals suffered an initial capacity loss of approximately 70% in the case of lithium hexafluorophosphate (LiPF6) and approximately 60% in the case of sodium bis(trifluoromethanesulfone)imide (NaTFSI) with the TEA+ systems showing the least amount of loss upon initial cycling, resulting in better stability over the cycling period overall. As this capacity loss occurred over the first few cycles an attempt to understand the interfacial processes at the electrode was made by use of electrical impedance spectroscopy, where modelling of particular components is possible, allowing for a determination of the mechanisms involved in charge storage at the electrode surface and whether pore inclusion plays a significant role in this process. As a result, it was found that for the sodium and TEA+ systems charge transfer and storage mechanism appeared to be limited to the surface of the particles that comprise the electrode material, whereas for the lithium systems the results suggest that a more subtle pore inclusion is the primary mechanism.