Mixed Doping Effect of Al and Te on the Structure and Ionic Conductivity of Garnet-Type Li7La3Zr2O12
Dawei Wang,Guiming Zhong,Zhengliang Gong,Yixiao Li,Yong Yang
DOI: https://doi.org/10.1149/ma2014-03/2/108
2014-01-01
ECS Meeting Abstracts
Abstract:Lithium-ion batteries have become the most popular battery system for portable devices such as smart p hones and laptop computers in recent years, and are being widely explored as electric power systems for hybri d and pure electric vehicles and storage systems for smar t grid. However, large-scale application of the batteries i s hampered by the safety concerns because of flammabl e organic electrolyte used in conventional lithium io n batteries. Solid electrolytes are hopeful to replac e organic electrolyte because of intrinsic safety. Garnet type Li 7La3Zr2O12 is the most promising solid electrolyte due to its high conductivity after dopi ng and attracts intensive research. It was reported that Li 7La3Zr2O12 can react with Al 2O3 crucibles and Al plays as doping element on stabilizing cubic garnet phase , resulting in good lithium ionic conductivity . In this study, we have investigate the co-doping effect of Al and Te, i.e. mixed doping effect of Al and Te on the st ruc ure and ionic conductivity of Li 7La3Zr2O12. A series of Te-doped Li 7La3Zr2O12 with the formula of Li7-2xLa3Zr2-xTexO12 (0 ≤ x ≤ 1) were synthesized by conventional solid state reaction in Al 2O3 and ZrO2 crucibles, respectively. The crystal structure was characterized by X-ray powder diffraction (XPRD). L i7xLa3Zr2-xTexO12 (0 ≤ x ≤ 1) synthesized in Al 2O3 crucibles show pure cubic structure up to x = 0.75, the composition with x = 1 exhibits cubic structure, too, but impurities such as La 0.5Zr0.5O1.75 and La2O3 are detected. The crystal structure of samples synthesized in ZrO 2 crucibles are as follows, the composition with x = 0 shows tetragonal structure, x = 0.25 shows cubic structure with some amount of La0.5Zr0.5O1.75 impurity, while both x = 0.5 and 0.75 show pure cubic structure. EDS results show th at Al contents of samples synthesized in Al 2O3 crucibles are y = 0.25 and 0.07 in the formula of Li 7-2x-3yyAlLa3Zr2-xTexO12 while x = 0 and 0.25, respectively. The signal of Al is no t detectable while x ≥ 0.5. Al MAS NMR spectra corroborates this result and demonstrates Al occupi es tetrahedral sites as the chemical shift of Al is around 64ppm. We conclude that both Al and Te can stabilize cubic garnet phase, and Te is much easier to dope i nto crystal lattice of garnet than Al in our experiment al condition. While x = 0.5 and 0.75, the Te can stabilize cubic phase by itself and restrain the reaction of samples with Al, but when x = 0 and x = 0.25, the content of Te is not enough to stabilize cubic phase, thus tetragona l a d impurity phase La 0.5Zr0.5O1.75 forms (ZrO2 crucible), respectively, the samples have to react with Al 2O3 crucibles to stabilize pure cubic phase (Al 2O3 crucibles). The sample with the composition of x = 0.25 synthesized in Al 2O3 crucible shows the highest ionic conductivity σ = 4 × 10 S cm and lowest activation energy Ea=0.33eV at room temperature. The ionic conductivity decreases, while the activation energy increases with Te content. At the composition of x = 0.25, the sample with Al contained exhibits higher ionic conductivity and lower activation energy than that of Al free sample, due to the existence of La 0.5Zr0.5O1.75 impurity phase in Al free sample, which blocks the transport of lithium ions pass through the grain boundaries. The differences between x = 0.5, 0.75 and 1 are quite small. The relative densities of samples increase with Te content, shows contrary trend with ionic conductivity. It discloses the decrease of ionic conductivity of the garnet is due to structure reas ons rather than better grain boundaries. In addition, we successfully separate Li1 and Li2 i n garnet by Li MAS NMR technique. The lithium peak at high field side is attributed to tetrahedral Li1 si te and lithium peak at low field side is attributed to oct ahedral Li2 sites. We set total number of lithium as stoich iometry of formula, the number of Li1 and Li2 are calculate d by the percentage of integral area of each peak, respe ctively. The Li1 sites are about 12% occupied and Li2 sites are about 45% occupied. Full width at half maximum (fwh m) is closely related with lithium jump rate. For Li1, it has the smallest fwhm value at the composition of x = 0.25, and increases with Te content, which shows a simila r trend with ionic conductivity. For Li2, the fwhm decreases with Te content, which is contrary to the trend of ionic conductivity. Thus we conclude that Li1 si te are the key lithium jump sites and determines the lithi um jump rate and ionic conductivity of Li 7-2xLa3Zr2-xTexO12. Although the lithium transport mechanism is not fir med yet, we tentatively propose that 24d-96h-48g-96h-24 d channel as Li diffusion route , that lithium jump route does not bypass Li1 sites.