Journal of Power Sources 160 (2006) 752–757 Ionic conductivity measurements of molten iodide-based electrolytes Patrick Masset a,b,c, ∗ , Antoine Henry a , Jean-Yves Poinso a, 1 , Jean-Claude Poignet c a CEA Le Ripault, BP 16, 37260 Monts, France b ASB-Aerospatiale Batteries, Alle´ e Saint-H´ el` ene, 18021 Bourges, France c LEPMI, BP 75, 38402 Saint-Martin d’H` eres, France Received 4 December 2005; received in revised form 4 January 2006; accepted 4 January 2006 Available online 17 February 2006 Abstract A number of iodide-based electrolytes (LiCl–LiI, LiI–KI, LiCl–LiBr–LiI, LiCl–LiI–KI and LiF–LiCl–LiI) were considered to be used in thermal batteries. Ionic conductivity of multi-cation and all-lithium electrolytes were evaluated. They were compared to the classical electrolytes (LiCl–KCl, LiF–LiBr–KBr and LiF–LiCl–LiBr) used in thermal battery applications. Measurements were realised by electrochemical impedance spectroscopy (E.I.S.). It was shown that some all-lithium iodide-based electrolytes exhibit interesting ionic conductivities and are suitable as electrolytes in thermal batteries. It was pointed out that activation energy E a was close to 11 kJ mol -1 for multi-cation electrolytes, whereas it was equal to 7 kJ mol -1 for all-lithium electrolytes. © 2006 Elsevier B.V. All rights reserved. Keywords: Molten salt; Ionic conductivity; Iodides; Electrolytes; Thermal batteries 1. Introduction Thermal activated batteries are suitable electrical gen- erators for military applications [1] due to their reliability and robustness. Installed on the dedicated devices, they can remain more than 20 years without performance degradation. Electrical needs increase more and more, especially for long time operations. To reach the future performance requirements, various improvements are under investigation. The performance benefits expected from a temperature increase in the initial stage are rather limited by the thermal decomposition of pyrite (see the Fe–S phase diagram [2,3] and experimental studies under helium atmosphere [4,5] and in molten salts [6]). In this frame, iodide-based electrolytes were investigated. Molten iodide mixtures present low melting points (m.p.) compared to the chloride, bromide or fluoride-based mixtures. Thus, a global performance improvment is expected from their use. In the past, iodide-based electrolytes have been envisaged and abandoned for three reasons: ∗ Corresponding author. Present adress: Karl Winnacker Institut der Dechema e.V., Theodor-Heuss Allee 25, 60486 Frankurt am Main, Germany. Tel.: +49 69 7564 362; fax: +49 69 7564 388. E-mail address: masset@dechema.de (P. Masset). 1 Present adress: CEA Valduc, 21120 Is-sur-Tille, France. (i) The relative high cost of lithium iodide LiI in comparison to other lithium halides [7]. (ii) Iodide salts are very hygroscopic [8] and water up-take kinetic is unfavourable compared to other halides [8,9], therefore, a careful drying must be performed in order to avoid hydrolysis [10]. (iii) Lithium iodide forms easily low temperature hydrates with water which are stable at highest temperatures [12,13] com- pared to other lithium halides LiCl [13–16] or LiBr [13–15]. Earlier, our group reported [11] partial results based on ionic conductivity measurements of iodide-based electrolytes. At the meantime, Guidotti and Reinhardt [17] published results showing that iodide-based electrolytes (LiF–LiCl–LiI ternary eutectic and LiF–LiCl–LiBr–LiI: two quaternary compositions) are attractive electrolytes for thermal batteries. The quaternary LiF–LiCl–LiBr–LiI (15.4–21.7–32.9–30 mol%) has already been studied by Borger et al. [7] for high temperature batteries. In molten salt electrolytes, the “current” transportation through single cells is ensured by ionic species migration. Usually, the mobility follows Arrhenius-type behaviour. The resulting ionic conductivity is then equal to: κ = κ ◦ exp  -E a RT  (1) 0378-7753/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.jpowsour.2006.01.014