Interface Investigations and Bulk Behaviors of an All-Solid-State Lithium Microbattery By Means of Electrochemical Impedance Spectroscopy

Monday, 6 October 2014: 14:40
Sunrise, 2nd Floor, Star Ballroom 4 (Moon Palace Resort)
S. Larfaillou (Institut de Chimie Moléculaire et des Matériaux d'Orsay (ICMMO), UMR CNRS-UPS 8182, Université Paris Sud, STMicroelectronics), D. Guy-Bouyssou (ST Microelectronics), F. Le Cras (CEA LETI), and S. Franger (Institut de Chimie Moléculaire et des Matériaux d'Orsay (ICMMO), UMR CNRS-UPS 8182, Université Paris Sud)
Constant miniaturization of electronic devices leads nowadays to the conception of smaller batteries. In this domain, using an all solid state thin film technology has many advantages over conventional lithium cells. These microbatteries are thin, bendable, with a long service lifetime and can also be produced with customizable shapes for better integration in electronic devices. Moreover, without liquid electrolyte, they are safe and comply well with environmental standards. Unfortunately their design and relative small dimensions imply some difficulties to accurately characterize them (from a physico-chemical point of view) without damages.

This paper will present the use of Electrochemical Impedance Spectroscopy to understand the behavior of thin film batteries built by stacking several active layers thanks to physical vapor deposition technics (with LiCoO2 as positive electrode, lithium phosphorus oxynitride vitreous glass as electrolyte (LiPON) and lithium metal as negative electrode). The first step of this study was to distinguish the impedance response between each individual active layer and interface by testing different systems. Metal-insulator-metal stack let us access to the ionic transport properties in LiPON [1]. Half-cells composed with electrolyte and LiCoO2 or lithium, give us information about the electrolyte/active material interfacial behavior. At last, the insertion electrode bulk electronic conductivity can be obtained with a Pt/LiCoO2/Pt symmetrical cell. Thanks to all these results it is thus possible to determine an electrical equivalent model for the complete battery. The model is composed of serial contact and collector resistivity (A), bulk electrolyte (B), capacitance and resistance interfaces (C, D), semi-infinite diffusion in LiCoO(E) and charges accumulation phenomena (F). 

The Second step was to explain the reasons for spectrum evolutions in function of state of charge, aging or other external conditions. EIS was systematically performed after battery manufacturing and after aging, from 1 MHz to 50 mHz and between 3V and 4.2V. The most obvious observation is the evolution of the “D” contribution with cell potential. Iriyama et al. already attributed it to the charge transfer process at the LiCoO2/LiPON interface [2]. But after aging a new potential dependent R//C contribution appears and could correspond to bulk phenomena in the insertion electrode created by crystalline evolution during aging.

To conclude, thanks to electrochemical impedance spectroscopy, we will propose an electrical equivalent model of the operating microbattery to predict its evolution, failure and aging.

[1] X.Yu, J. Bates, G. Jellison anf F. X. Hart, J. Electrochem Soc., 144, 524 (1997)

[2] Y. Iriyama, T. Kako, C. Yada, T. Abe and Z. Ogumi, Solid State Ionics, 176, 2371 (2005).