Effect of Microstructure and Elemental Composition on Electrode Interfacial Impedance of Li7La3Zr2O12-based Batteries

Wednesday, 8 October 2014: 10:40
Sunrise, 2nd Floor, Galactic Ballroom 2 (Moon Palace Resort)
G. Hitz, J. Dai, J. A. Taillon, Y. Gong, X. Han (University of Maryland), V. Thangadurai (University of Calgary), L. Hu (University of Maryland, college park), and E. D. Wachsman (University of Maryland)

Solid state batteries offer increased safety, high voltage stability, and simpler electrolyte chemistry than organic electrolytes. Their inherent protection from dendrite growth gives the capability to use metallic lithium anodes. Recent advances have improved conductivity to within an order of magnitude of organic electrolytes at room temperature. However, interfacial resistance between electrodes and the electrolyte remains a pressing concern. In a typical planar battery, this high resistance per area is magnified by low interfacial surface area.

Elemental composition has a significant effect on the conductivity of a garnet electrolyte. The conductivity of the garnet fluctuates an order of magnitude based on a lithium stoichiometry of 6 to 7 in the Li7-xLa3Zr2-xTaxO12series. It has also been demonstrated that aluminium composition at the grain boundaries associated with higher temperature sintering increases conductivity. It stands to reason lithium activity extremes at the electrode interfaces effects would cause local variations in electrolyte composition and affect the conductivity.

This study aims to address some of these concerns using a systematic study of the anode-electrolyte interface and the cathode-electrolyte interface. We investigate the elemental composition of the interface, as well as the effect of greatly increased surface area through a highly microporous electrolyte scaffold containing the cathode and anode material.

Experimental Procedure

Li7La3Zr2O12 garnet was produced in a conventional solid state method. 10% excess LiOH-H2O and stoichiometric amounts of La2O3 and ZrO2were ball milled in ethanol for 24 hours. The milled materials were added to an alumina crucible, calcined at 900 ºC for 4 hours, and milled again.

A colloidal deposition of this slurry is prepared with polyvinyl butyral (PVB), benzyl butyl phthalate (BBP), and Solsperse dispersant in ethanol.

Calcined powder is added to toluene and ethanol solvents with a binder system containing PVB, BBP, and menhaden fish oil as a dispersant. This preparation is milled for two days, then degassed in preparation for tape-casting.

After tape-casting, tapes are drop-coated with the colloidal slurry and another tape is applied on top. This structure is held at 500ºC for 1 hour to burn off binders then heated to 1100 ºC for 1 hour. This produces a three layer porous-dense-porous scaffold of electrolyte that will maintain electronic isolation between electrodes.

Symmetric cells with two cathodes are produced by infiltrating a glycine/metal-nitrate solution into the pores of each side and combusting. For two-anode symmetric cells, lithium is lightly pressed into both sides at 350 ºC.

Electrochemical impedance spectroscopy was performed between room temperature and 300 ºC using a 10 mV amplitude between 10 MHz and 100 mHz.

Electron energy loss spectroscopy was performed on a 2100F TEM equipped with a Gatan Tridiem post-it column energy filter.


Figure 1 shows a cross section of a lithium-penetrated Li-garnet scaffold showing the fracture plane of the garnet(light) fully coated by the lithium (dark). Figure 2 shows the Nyquist plot of a full cell made with a lithium-penetrated scaffold. The bulk conductivity shown in the inset is fairly low at 20 Ω compared to the 1000 Ω overall cell impedance.