1072
Impurity Formation of Garnet-like Solid Electrolyte Densified By Spark Plasma Sintering

Thursday, 23 June 2016
Riverside Center (Hyatt Regency)
H. Yamada and T. Ito (Graduate School of Engineering, Nagasaki University)
Garnet-like lithium-ion conducting solid electrolytes attracts great interests for their potential application to all solid state batteries with lithium anode. One of drawbacks of this material is severer sintering condition (typically >1200°C and > 10 h). It has been demonstrated that spark plasma sintering (SPS) successfully accelerates the sintering of garnet-like solid electrolytes like other materials. On SPS, it has been also pointed out that impurity phases are included in the obtained pellets. However, mechanism of the impurity formation is not revealed. In this paper, we prepared pellets of garnet-like Li6.5La3Zr1.5Ta0.5O12(LLZT) with various conditions and investigated the plausible mechanism.

LLZT powder was synthesized by solid state reactions using LiOH, La2O3, ZrO2 and Ta2O5. Except for Li source, other starting materials were mixed with a stoichiometric ratio. Amount of LiOH was by 10% higher than the stoichiometric ratio to compensate the loss of Li on the following calcination, which was conducted at 900°C for 10 h in air. Calcined powder was sintered using graphite dies by SPS with a sintering temperature of 900-1100°C, pressure of 12.5-50 MPa and time of 1-30 min under vacuum.

Crystalline structure of LLZT powder and sintered pellets were confirmed by XRD. To confirm distribution of phases along depth, XRD profiles were repeatedly recorded after surface of the pellets was polished for certain thickness (Figure 1). On one side of pellets, which corresponds to anode of pulse current of SPS, La2Zr2O7 was observed. This impurity was confirmed only on the anode and disappeared with polishing the surface, suggesting inhomogeneous decomposition, possibly caused by electrolysis. Interestingly, on cathode, no impurity phase was confirmed by XRD. When the La2Zr2O7 is formed as a result of electrolysis, reduction should occur on cathode. The cathode product is possibly amorphous carbon reduced by Li2CO3 that had been produced on surface of LLZT particles by absorption of CO2.

The Li2CO3 is supposed to play another role on SPS. When LLZT powder with smaller particle size was used for SPS, more La2Zr2O7 was produced, suggesting the resistance between electrodes were reduced by molten Li2CO3.

In addition to the La2Zr2O7, asymmetric impurity formation was visually confirmed. On anode side, the pellets were white, while cathode sides looked dark grey to black. However, the XRD did not show any impurity phases, indicating the origin of the dark color was amorphous. By taking account of the electrolysis of LLZT on anodes to form La2Zr2O7, it is plausible that amorphous carbon was produced on cathodes due to the electrochemical reduction of CO32− of Li2CO3.

These impurities were formed in the region within ca. 100-300 μm from the surface of pellets for both cathode and anode sides. Therefore, it is necessary to polish pellets for at least 300 μm for both sides before the pellets are used for electrochemical cells. When these impurities were removed, the obtained pellets exhibited very high density of 96% and high ionic conductivity of 8×10−4 S cm−1 at 298 K, which are attractive properties for all solid state batteries.