Waterstable lithium ion conducting solid electrolyte of the LAGTP system with NASICONtype structure

Water-stable lithium ion conducting solid electrolyte of the Li 1.4Al 0.4Ti 1.6?x Ge x (PO 4)3system (x =0–1.0)with NASICON-type structure

Peng Zhang,Masaki Matsui,Atsushi Hirano,Yasuo Takeda,Osamu Yamamoto ?,Nobuyuki Imanishi

Graduate School of Engineering,Mie University,Tsu,Mie 514-8507,Japan

a b s t r a c t

a r t i c l e i n f o Article history:

Received 14July 2013

Received in revised form 11September 2013Accepted 11September 2013Available online xxxx Keywords:

Solid electrolyte Lithium conductor Lithium –air battery

Solid ionic conductors composed of Li 1.4Al 0.4Ti 1.6?x Ge x (PO 4)3(x =0–1.0)with the NASICON-type structure were synthesized with a precursor prepared using the sol –gel method.The electrical conductivity was ex-amined as a function of x in Li 1.4Al 0.4Ti 1.6?x Ge x (PO 4)3at various sintering temperatures and for various sintering periods.The highest electrical conductivity was obtained for Li 1.4Al 0.4Ti 1.4Ge 0.2(PO 4)3sintered at 900°C for 11h in air.The total and bulk conductivities of the sintered pellet were 1.29×10?3and 2.35×10?3S cm ?1at 25°C,respectively.The grain boundary resistance of Li 1.4Al 0.4Ti 1.4Ge 0.2(PO 4)3was signi ?cantly increased to 552Ωcm 2from 12.2Ωcm 2by immersion in distilled water at 50°C for one week,whereas the bulk resistance was not increased.However,no signi ?cant increase of the bulk and grain boundary resistance was observed after immersion in a saturated aqueous solution of LiOH and LiCl.

?2013Elsevier B.V.All rights reserved.

1.Introduction

The lithium –air battery is a promising power source for electric vehicles because of its high theoretical speci ?c energy density [1].Two types of Li –air batteries,non-aqueous [2]and aqueous [3],have been developed.The speci ?c energy density of the non-aqueous system (3460Wh kg ?1)is higher than that of the aqueous system (1910Wh kg ?1);however,the non-aqueous system has some severe problems that must still be addressed,such as lithium corrosion by water and CO 2ingression when operated in air,precip-itation of high resistance reaction products (lithium oxide)on the air electrode,and high polarization during charge process.The aqueous system could avoid these problems observed for the non-aqueous system by using a water-stable lithium electrode and an aqueous elec-trolyte.Lithium metal reacts intensely with water;therefore,the key material of a water-stable lithium electrode is the water-stable lithium ion conducting solid electrolyte,which is used to prevent direct contact of lithium with water [4].At present,only two types of high lithium ion conducing solid electrolytes are known to be stable in water;one is the Li analog of NASICON,Li 1+x A x M 2?x (PO 4)3(LAMP;A =Al,Sc,M =Ti or Ge)[5],and the garnet-type Li 7La 3Zr 2O 12(LLZ)[6].The reaction product of the aqueous lithium –air system is LiOH and the concen-tration of LiOH increases with increasing discharge https://www.360docs.net/doc/1e4542608.html,MP and LLZ are unstable in concentrated LiOH aqueous solution,but

are stable in aqueous solution saturated with LiOH and LiCl [7,8].Therefore,these compounds could be used as the protective layer of the lithium metal electrode for lithium –air batteries with a LiCl saturated aqueous solution.The electrical conductivities of the sintered LLZ and Nb-or Y-doped LLZ were reported to be 2×10?4to 8×10?4S cm ?1at room temperature [6,8–11]and LLZ is stable in contact with lithium metal [6],but it is dif ?cult to prepare a pore-free dense plate of LLZ due to the evaporation of lithium during high temperature sintering.The electrical conductivity of LAMP is dependent on the composition and the preparation method.The re-spective conductivities of the Li 1+x Al x Ti 2?x (PO 4)3glass ceramic [12]and Li 1.5Al 0.5Ge 1.5(PO 4)3[13]were as high as 1.3×10?3and 5.08×10?3S cm ?1at room https://www.360docs.net/doc/1e4542608.html,MP is unstable in contact with lithium metal;therefore,an interface layer should be used between the lithium metal and LAMP.Zhang et al.[14]reported the fabrication of a cell that consisted of a lithium metal anode,a lithium conducting polymer electrolyte composed of polyethylene oxide (PEO)with Li(CF 3SO 2)2N (LiTFSI)-BaTiO 3and a water-stable NASICON-type lithium ion conducting Li 1+x +y Al y (Ti,Ge)2?x P 3?y Si y O 12glass ceramic (Ohara plate),which had an electrode resistance of 173Ωcm 2at room temperature.The elec-trical conductivity of the Ohara plate is around 1×10?4S cm ?1at room temperature.

To reduce the resistance of the water-stable lithium metal elec-trode,a solid lithium ion conductor with a total conductivity above 10?3S cm ?1is desirable.However,the total ionic conductivity of polycrystalline lithium ion conductors with the NASICON-type struc-ture is usually in the order of 10?4S cm ?1at room temperature,ex-cept for the Li 1+x Al x Ti 2?x (PO 4)3[12]and Li 1.5Al 0.5Ge 1.5(PO 4)3

Solid State Ionics 253(2013)175–180

?Corresponding author.Tel.:+81592319420.

E-mail address:Yamamoto@chem.mie-u.ac.jp (O.

Yamamoto).0167-2738/$–see front matter ?2013Elsevier B.V.All rights reserved.

https://www.360docs.net/doc/1e4542608.html,/10.1016/j.ssi.2013.09.022

Contents lists available at ScienceDirect

Solid State Ionics

j o ur n a l h o m e p a g e :w ww.e l s e v i e r.c om /l o c a t e /s s i

glass ceramics[13].The Li1.5Al0.5Ge1.5(PO4)3glass ceramic is attrac-tive,because the conductivity of a pelletized sample of the sintered glass ceramic powder was reported to be4.62×10?3S cm?1at 27°C[13].The preparation of the glass ceramics is a little complex and germanium is an expensive element.Fu[15]also studied the Li1+x Al x Ge2?x(PO4)3glass ceramic and found the highest electri-cal conductivity of4.0×10?4S cm?1for Li1.5Al0.5Ge1.5(PO4)3.The solid solution of Li1+x Al x Ge y Ti2?x?y(PO4)3(0.2≤x≤0.8,0.8≤y≤1.0)was synthesized and studied by Maddonado-Manso et al.

[16].The highest bulk and total conductivities at room temperature were7×10?4and1.1×10?4S cm?1for Li1.5Al0.5Ti0.7Ge0.8(PO4)3,re-spectively.Xu et al.[17]have investigated the Li1.4Al0.4Ti1.6?x Ge x(PO4)3 glass ceramics and reported a maximum room temperature conductivity of6.21×10?4S cm?1for Li1.4Al0.4Ti0.53Ge1.07(PO4)4.In this study, the effect of substitution of Ge for Ti on the electrical conductivity have been examined for the whole Li1.4Al0.4Ti1.6?x Ge x(PO4)3 (x=0–1.0)system.The highest total electrical conductivity of 1.29×10?3S cm?1at25°C was found for Li1.4Al0.4Ti1.4Ge0.2(PO4)3. The stability of this high conducting solid electrolyte in aqueous so-lutions was also studied for application as the protective layer of the water-stable lithium electrode for lithium–air batteries.

2.Experimental

The precursor of Li1.4Al0.4Ti1.6?x Ge x(PO4)3was prepared by the sol–gel method using citric acid,as reported previously[18].Stoichio-metric amounts of Ge(OC2H5)4(Aldrich)and Ti(OC4H9)4(Aldrich) were dissolved in ethylene glycol,and then added to a0.2M aqueous solution of citric acid and stirred continuously with a magnetic stirrer at120°C for12h to obtain a homogeneous solution.After preparation of the gel was completed,stoichiometric amounts of LiNO3,Al(NO3)3·9-H2O,and NH4H2PO4were added to the gel solution.The molar ratio of citric acid to Li++Al3++Ge4++Ti4+was4:1.After a homoge-neous solution was formed,the gel was kept at170°C for several hours to allow evaporation of the water and to promote esteri?cation and polymerization.The gel was then heated at500°C for4h to com-plete the chemical decomposition of nitrates and organic compounds. The black product with residual carbon was ground to a uniformly ?ne powder with an agate mortar and pestle to increase the surface area of the samples before sintering at800°C for5h to complete the chemical reaction.The precursor was then grounded by wet milling with a planetary mill(Fritsch)for15h to make?ne powders,which are necessary to achieve dense ceramic sintering.The?ne powder was isostatically pressed(150MPa)into a pellet(ca.12mm diameter,ca.

1.5mm thick)and then sintered at various temperatures on a gold sheet to protect the samples from reaction with the alumina refractory board.

The crystal structure of the synthesized samples was analyzed using X-ray diffraction analysis(XRD;Rigaku RINT2500)with Cu Kαradia-tion in the2θrange from10to90°at a scanning step rate of0.02°s?1 and using Si powder as an internal standard.The microstructure and morphology of the pellets was observed using scanning electron mi-croscopy(SEM;Hitachi SEM S-4000).

The AC impedance of sintered pellets with sputtered gold electrodes was measured in air using a frequency response analyzer(Solartron 1260)in the temperature range from20to80°C and the frequency range of0.1Hz to1MHz.The impedance pro?les were analyzed to es-timate the bulk and grain boundary conductivities using a non-linear in-stant?t program in the ZView software package.The stability of the sintered pellets was investigated by immersion into distilled water,sat-urated LiOH,saturated LiCl,and saturated LiOH and LiCl aqueous solu-tions at50°C for one week.The pellets immersed into these solutions were then carefully washed with distilled water and dried in a vacuum at220°C for20h before measuring electrical conductivity and XRD patterns.3.Results and discussion

The sol–gel precursor of Li1.4Al0.4Ti1.4Ge0.2(PO4)3was sintered in the temperature range of880to950°C for11h.The XRD patterns of the samples sintered at880and900°C indicate the NASICON-type struc-ture without an impurity phases,although an impurity phase of AlPO4 was observed for those samples sintered at950°C,as reported by Xu et al.,[17]and Mariappan et al.[19]for the glass ceramics.The pellet sintered at880°C had a low relative density of77.9%;therefore,the sintering temperature was?xed at900°C for this study.

Fig.1shows the XRD patterns of the Li1.4Al0.4Ti1.6?x Ge x(PO4)3 (x=0–1.6)series sintered at900°C for11h.These diffraction pat-terns were indexed according to the NASICON-type structure and impurity phases such as Li3PO4,AlPO4,GeO2,TiO2,GeP2O7,and Li3PO4,which were observed in previous studies[19–21],were below the detection limit.The intensity ratio of the diffraction peaks are changed slightly with x,especially the peaks for the (137)and(2110)planes were diminished for the samples with x=1.2and1.6.Fig.2shows the change in the lattice constants with x for Li1.4Al0.4Ti1.6?x Ge x(PO4)at room temperature,where the lattice constants were calculated using the hexagonal unit cell. The lattice constants of both a-and c-axes were decreased by substi-tution of Ti with Ge,which has a smaller ionic radius(0.053nm) than Ti(0.0605nm);however,the change does not obey Vegard's law in a range of x=0to1.6.The lattice parameter change suggests that Li1.4Al0.4Ti1.6(PO4)3may have a slightly different crystalline structure with Li1.4Al0.4Ge(PO4)3and in the range of x=0.3to1.2, two phases of the solid solution of Li1.4Al0.4Ti1.3Ge0.3(PO4)3and that of Li1.4Al0.4Ti0.4Ge1.2(PO4)3may exist at room temperature.The miscibility gap was con?rmed by the electrical conductivity dependence on x in Li1.4Al0.4Ti1.6?x Ge x(PO4)3,as discussed later.Fig.3shows the relative densities of Li1.4Al0.4Ti1.6?x Ge x(PO4)3pellets sintered at900°C for 11h as a function of x.The relative densities were calculated from the ratios of the observed densities calculated using the sintered pellet di-mensions and weight and those calculated using the lattice parameters. The relative densities of the sintered pellets were increased by the sub-stitution of Ti with Ge and a maximum density of94.4%was obtained for Li1.4Al0.4Ti1.4Ge0.2(PO4)3.Further substitution of Ge for Ti provided the second maximum density for Li1.4Al0.4Ti0.4Ge1.2(PO4)3.This relative density behavior also suggests that there is a miscibility gap for the Li1.4Al0.4Ti1.6?x Ge x(PO4)3system.

Fig.4shows impedance pro?les of sintered samples for Li1.4Al0.4Ti1.2Ge0.2(PO4)3sintered at various temperatures and periods measured at25°C.These impedance pro?les have a semicircle that

is Fig.1.XRD patterns of Li1.4Al0.4Ti1.6?x Ge x(PO4)3(x=0–1.6)sintered at900°C for11h.

176P.Zhang et al./Solid State Ionics253(2013)175–180

attributed to the grain boundary resistance [22].The intercept of the semicircle on the real axis at high frequency represents the bulk resis-tance,and the diameter of the semicircle indicates the grain boundary resistance.The grain boundary and bulk resistances are dependent on the sintering temperature and period.The highest total conductivity was found for the sample sintered at 900°C for 11h;the total electrical conductivity for as-sintered Li 1.4Al 0.4Ti 1.2Ge 0.2(PO 4)3at 25°C was as high as 2.25×10?3S cm ?1.However,the electrical conductivity of the as-sintered samples was subject to an aging effect,and the grain boundary resistance was signi ?cantly increased with the aging peri-od.Fig.5shows the total,grain boundary,and bulk conductivities at 25°C for Li 1.4Al 0.4Ti 1.44Ge 0.16(PO 4)3,Li 1.4Al 0.4Ti 1.40Ge 0.20(PO 4)3,and Li 1.4Al 0.4Ti 1.3Ge 0.30(PO 4)3as a function of the aging period.The bulk conductivities of these samples showed no signi ?cant aging effect,but the grain boundary conductivities of all samples decreased sig-ni ?cantly with the aging period.Stable conductivity was observed after approximately one month.The Li 1.4Al 0.4Ti 1.40Ge 0.20(PO 4)3sample stored in a glove box also showed a similar increase in grain boundary resistance by aging,as shown in Fig.5(b).The grain boundary conductivity is affected by impurity phases at the grain boundary.Impurity phases such as Li 2O and AlPO 4exhibit both blocking and a space charge effect for lithium ion transport [21].The increase in grain boundary resistance with the aging period could be explained by a metastable impurity phase with high con-ductivity in the as-sintered sample,which may change to the stable low conductivity phase by aging.Similar aging behavior was ob-served for the grain boundary resistance of Li 1.4Al 0.4Ti 1.6(PO 4)-3wt.%ZrO 2(and 3wt.%Al 2O 3)[23].

Fig.6shows the electrical conductivity (measured at 25°C)of as-sintered (900°C for 11h)Li 1.4Al 0.4Ti 1.6?x Ge x (PO 4)3as a function of

x and that stored in air for one month.The conductivity measured at 25°C for the as-sintered Li 1.4Al 0.4Ti 1.6(PO 4)3(4.79×10?4S cm ?1)is comparable to that of Li 1.4Al 0.4Ti 1.6(PO 4)3prepared using sol –gel precursors (6.13×10?4S cm ?1)[24].The conductivity vs.x curves shows two conductivity peaks at x =0.2and 1.2,which suggest the formation of a miscibility gap at room temperature in the solid solution of Li 1.4Al 0.4Ti 1.6?x Ge x (PO 4)3,as observed in the lattice constant change with x.The highest total conductivity of 2.25×10?3S cm ?1at 25°C was observed for as-sintered Li 1.4Al 0.4Ti 1.4Ge 0.2(PO 4).The conductivity was decreased with the aging period and a stable total conductivity of 1.29×10?3S cm ?1was observed after one month storage;the stable bulk and grain boundary conductivities were 2.35×10?3and 2.87×10?3S cm ?1at 25°C,respectively.The bulk conductivity of Li 1.4Al 0.4Ti 1.6(PO 4)3was enhanced approximately 3times and the grain boundary conductivity by one order when Ge was substituted for Ti.The high total steady state conductivity of Li 1.4Al 0.4Ti 1.4Ge 0.2(PO 4)could be explained by the formation of a high lithium ion mobility phase in the grains and a high grain boundary conductivity phase by the substitution of Ge for Ti.The second conductivity peak was observed at x =1.2in Li 1.4Al 0.4Ti 1.6?x Ge x (PO 4)3.The steady state total,bulk,and grain boundary conductivities of Li 1.4Al 0.4Ti 0.4Ge 1.2(PO 4)3were 3.43×10?4,6.67×10?4and 7.65×10?4S cm ?1at 25°C,respectively.

Fig.7shows the temperature dependence of the bulk,grain boundary and total conductivities for Li 1.4Al 0.4Ti 1.4Ge 0.2(PO 4)3and Li 1.4Al 0.4Ti 0.4Ge 1.2(PO 4)3in the temperature range of 20to 80°C,where the samples were stored at room temperature for one month.These Arrhenius plots are linear in the examined tempera-ture range;therefore,the activation energies for the bulk,grain boundary and total conductivities were calculated from the slopes of the Arrhenius plots and are summarized with the conductivity data in Table 1.The activation energies for Li 1.4Al 0.4Ti 1.4Ge 0.2(PO 4)3were calculated to be 20.2kJ mol ?1for bulk conduction and 55.0kJ mol ?1for grain boundary conduction,and those for Li 1.4Al 0.4Ti 0.4Ge 1.2(PO 4)

3

https://www.360docs.net/doc/1e4542608.html,ttice parameters for Li 1.4Al 0.4Ti 1.6?x Ge x (PO 4)3(x =0–1.6)measured at room

temperature.

Fig.3.Relative density of Li 1.4Al 0.4Ti 1.6?x Ge x (PO 4)3(x =0–1.6)pellets sintered at 900°C for 11

h.

Fig.4.Impedance pro ?les (measured at 25°C)for Li 1.4Al 0.4Ti 1.4Ge 0.2(PO 4)3(a)sintered at various temperatures for 11h and (b)sintered at 900°C for various periods.

177

P.Zhang et al./Solid State Ionics 253(2013)175–180

were 25.7kJ mol ?1for bulk conduction and 53.3kJ mol ?1for grain boundary conduction.The activation energy for bulk conduction in Li 1.4Al 0.4Ti 1.4Ge 0.2(PO 4)3is slightly lower than that for Li 1.4Al 0.4Ti 1.6(PO 4)3(21.1kJ mol ?1).These activation energies for bulk conduction are lower than those for the high lithium ion conducting glass ceramics of Li 1+x Al x Ti 2?x (PO 4)3(24.4kJ mol ?1)[25],and for Li 1.5Al 0.5Ge 1.5(PO 4)3(31.8kJ mol ?1)[20]and Li 1.6Al 0.6Ti 0.7Ge 0.8(PO 4)3(33.8kJ mol ?1)[16].Li 1.4Al 0.4Ti 1.4Ge 0.2(PO 4)3has the lowest activation energy for bulk conduction of the NASICON-type lithium ion conducting solids.The activation energy for bulk conduction in Li 1.4Al 0.4Ti 0.4Ge 1.2(PO 4)3is slightly higher than that for Li 1.4Al 0.4Ti 1.4Ge 0.2(PO 4)3and lower than that for Li 1.4Al 0.4Ge 1.6(PO 4)3.The difference in the activation energy for bulk conduction between Li 1.4Al 0.4Ti 0.4Ge 1.2(PO 4)3and Li 1.4Al 0.4Ti 1.4Ge 0.2(PO 4)3could be explained by the miscibility gap of the Li 1.4Al 0.4Ti 1.6?x Ge x (PO 4)3system at room temperature,as ob-served by the change in the lattice constant with x shown in Fig.2.

The NASICON-type structure of LiM 2(PO 4)3(M =Ti or Ge)consists of a three dimensional framework of PO 4tetrahedra corner-sharing with MO 6octahedra,Li +ions partially occupying at intermediate of the framework [26].The high bulk conductivity of Li 1.4Al 0.4Ti 1.4Ge 0.2(PO 4)3suggests that the rigid [(Ti 1.4Ge 0.2,Al 0.4)(PO 4)3]skeleton provides easier pathways for migration than that in the [(Ti 1.6Al)(PO 4)3]skeleton,because the covalency of the Ge \O bond is stronger than that of the Ti \O bond [20].The slightly higher bulk activation energy for Li 1.4Al 0.4Ti 0.4Ge 1.2(PO 4)3than that of Li 1.4Al 0.4Ti 1.4Ge 0.2(PO 4)3may be due to their slightly different crystal structures.The detailed differences in the crystal structures should be analyzed to con ?rm the change in ac-tivation energy.The activation energy for the grain boundary conduction of as-sintered Li 1.4Al 0.4Ti 1.4Ge 0.2(PO 4)3was as low as 30kJ mol ?1and the grain boundary conductivity was as high as 1.03×10?2S cm ?1,while the activation energy of Li 1.4Al 0.4Ti 1.4Ge 0.2(PO 4)3stored at

room

Fig. 5.Bulk,grain boundary,and total conductivities measured at 25°C for (a)Li 1.4Al 0.4Ti 1.44Ge 0.16(PO 4)3,(b)Li 1.4Al 0.4Ti 1.4Ge 0.2(PO 4)3,and (c)Li 1.4Al 0.4Ti 1.3Ge 0.3(PO 4)3as a function of the aging period.Grain boundary and total conductivities of Li 1.4Al 0.4Ti 1.4Ge 0.2(PO 4)3stored in a glove box (glove)were shown in Fig.

5-(b).

Fig.6.Total and bulk electrical conductivities measured at 25°C for as-sintered Li 1.4Al 0.4Ti 1.6?x Ge x (PO 4)3(x =0–1.6)and the total electrical conductivity measured at 25°C for stored Li 1.4Al 0.4Ti 1.6?x Ge x (PO 4)3(x =0–1.6)for one month in

air.

Fig.7.Temperature dependence of the bulk,grain boundary,and total conductivity of (a)Li 1.4Al 0.4Ti 0.14Ge 0.2(PO 4)3and (b)Li 1.4Al 0.4Ti 0.4Ge 1.2(PO 4)3.

178P.Zhang et al./Solid State Ionics 253(2013)175–180

temperature for one month increased to 55.0kJ mol ?1,while the grain boundary conductivity was decreased to 2.87×10?4S cm ?1.The activation energies for the grain boundary conduction of Li 1.4Al 0.4Ti 1.4Ge 0.2(PO 4)3and Li 1.4Al 0.4Ti 0.4Ge 1.2(PO 4)3are slightly lower than those of Li 1.4Al 0.4Ti 1.6(PO 4)3and Li 1.4Al 0.4Ge 1.6(PO 4)3.The grain boundary conduction may be in ?uenced by impurity phases such as Li 2O and AlPO 4[21],and the similar grain boundary activation energies for Li 1.4Al 0.4Ti 1.4Ge 0.2(PO 4)3and Li 1.4Al 0.4Ti 0.4Ge 1.2(PO 4)3suggest that the grain boundary impurity phase is similar in these phases.

The stability of the Li 1+x +y Al x (Ti,Ge)2?x Si y P 3O 12glass ceram-ic (Ohara plate)in aqueous solutions has been extensively studied by Imanishi and coworkers [7,27].The glass ceramic decomposes in strong acid solution and dissolves in strong alkaline solution.They found that the glass ceramic is stable in saturated LiCl and LiOH aqueous solution,due to the low dissociation of LiOH to ions in the presence of a high Li +ion content [7].The stability of the high lithi-um ion conducting solid electrolyte in saturated LiOH and LiCl aque-ous solution is a key issue for application as the protective layer of the lithium metal electrode in aqueous lithium –air batteries,because the reaction product for aqueous lithium –air batteries is LiOH.The sta-bility of Li 1.4Al 0.4Ti 1.4Ge 0.2(PO 4)3stored in distilled water,saturated LiCl aqueous solution,saturated LiOH solution,and saturated LiCl and LiOH solution at 50°C for one week was examined.The XRD patterns of Li 1.4-Al 0.4Ti 1.4Ge 0.2(PO 4)3immersed in these solutions revealed no impurity phases and there was no change in the ratio of diffraction peaks before and after immersion in these solutions.However,the impedance pro-?les showed a change after immersion in these solutions,as shown in Fig.8,where the as-sintered samples were used.The grain boundary resistance of Li 1.4Al 0.4Ti 1.4Ge 0.2(PO 4)3immersed in aqueous solution increased from 12.2to 552Ωcm 2,while the bulk resistance in-creased slightly.A similar increase in grain boundary resistance was observed for the Li 1+x +y Al x (Ti,Ge)2?x P 3?y Si y O 12glass ceramic (Ohara plate)[27].This could be explained by the ion-exchange

reaction between Li +and H +,as observed in Li 3x La 2/3T i O 3[28].The mechanical strength of the pellet also became very poor by immer-sion in distilled water at 50°C for 1week.An increase in the grain boundary resistance was observed for Li 1.4Al 0.4Ti 1.4Ge 0.2(PO 4)3im-mersed in saturated LiOH aqueous solution.On the other hand,the change in the electrical conductivity of Li 1.4Al 0.4Ti 1.4Ge 0.2(PO 4)3by im-mersion in saturated LiCl aqueous solution and in saturated LiOH and LiCl aqueous solution was not signi ?cant.The conductivity of Li 1.4Al 0.4-Ti 1.4Ge 0.2(PO 4)3immersed in saturated LiOH and LiCl aqueous solution at 50°C for one week was 1.66×10?3S cm ?1at 25°C,which is comparable to that of the sample after one week aging shown in Fig.5.It could be concluded that Li 1.4Al 0.4Ti 1.4Ge 0.2(PO 4)3with high lithium ion electrical conductivity is stable in LiOH and LiCl saturated aqueous solution,and is thus acceptable for use as the protective layer in a water-stable lithium electrode.4.Conclusion

The NASICON-type lithium ion conducting solid electrolyte,Li 1.4Al 0.4Ti 1.6?x Ge x (PO 4)3,was prepared using a sol –gel precursor at 900°C for 11h.The electrical conductivity of the as-sintered sample exhibited an aging effect and stable conductivity was obtained after storage for approximately one month.The electrical conductivity vs.x curve for Li 1.4Al 0.4Ti 1.6?x Ge x (PO 4)3has two peaks at x =0.2and 1.2.The highest total lithium ion conductivity of 1.29×10?3S cm ?1at 25°C was found for Li 1.4Al 0.4Ti 1.4Ge 0.4(PO 4)3with a high relative den-sity of 94.4%.The bulk and grain boundary conductivities of this sample at 25°C are 2.35×10?3and 2.87×10?3S cm ?1,respectively,and the total conductivity of 1.29×10?3S cm ?1is approximately one order higher than that of Li 1.4Al 0.4Ti 1.6(PO 4)3.The high lithium ion conducting solid electrolyte exhibited stability in LiOH and LiCl saturated aqueous solution.The high lithium ion conductivity and stability in LiOH and LiCl aqueous solution of Li 1.4Al 0.4Ti 1.4Ge 0.2(PO 4)3suggests that this com-pound is an attractive candidate as the protective layer for the lithium metal electrode in aqueous lithium –air batteries.Acknowledgment

This study was supported by the Japan Science and Technology Agency (JST)under the “Advanced Low Carbon Technology Research and Development Program ”.References

[1]M.Armand,J.M.Tarascon,Nature 451(2008)652.

[2]T.Ogasawara,A.Debart,M.Holzapgel,P.Novak,P.G.Bruce,J.Am.Chem.Soc.128

(2006)1390.

[3]T.Zhang,N.Imanishi,S.Hasegawa,A.Hirano,J.Xie,Y.Takeda,O.Yamamoto,N.

Sammes,J.Electrochem.Soc.155(2008)A965.

[4]S.J.Visco,E.N.,B.Katz,L.C.D.Jonghe,M.Y.Chu,Abst.#53,The International Meeting

on Lithium Batteries,Nara,Japan,2004.

[5]H.Aono,E.Sugimoto,Y.Sadaoka,N.Imanaka,G.Adachi,J.Electrochem.Soc.136

(1989)590.

[6]R.Murgan,V.Thangadurai,W.Weppner,Angew.Chem.Int.Ed.46(2007)

7778.

[7]Y.Shimonishi,T.Zhang,N.Imanishi, D.Im, D.J.Lee,A.Hirano,Y.Takeda,O.

Yamamoto,N.Sammes,J.Power Sources 196(2011)5128.

Table 1

Bulk (σb ),grain boundary (σgb ),and total(σt )conductivity at 25°C and activation energy for bulk (E b ),grain boundary (E gb ),and total (E t )conduction for samples stored for one month in air.

Compound

Conductivity (S cm ?1)Activation energy (kJ mol ?1)σb

σgb σt E b E gb E t Li 1.4Al 0.4Ti 1.6(PO 4)3

8.04×10?4 2.27×10?4 1.77×10?421.160.436.4Li 1.4Al 0.4Ti 1.4Ge 0.2(PO 4)3 2.35×10?3 2.87×10?3 1.29×10?320.255.029.0Li 1.4Al 0.4Ti 0.4Ge 1.2(PO 4)3 6.67×10?47.65×10?4 3.43×10?425.753.332.2Li 1.4Al 0.4Ge 1.6(PO 4)3

4.20

×10?4

3.40

×10?4

1.77

×10?4

27.4

58.5

34.4

Fig.8.Impedance pro ?les at 25°C for (a)pristine Li 1.4Al 0.4Ti 1.44Ge 0.16(PO 4)3,and Li 1.4Al 0.4Ti 1.44Ge 0.16(PO 4)3immersed in (b)water,(c)saturated LiCl aqueous solu-tion,(d)saturated LiCl and LiOH aqueous solution,and (e)in saturated LiOH aqueous solution at 50°C for one week.

179

P.Zhang et al./Solid State Ionics 253(2013)175–180

[8]Y.Shimonishi,A.Toda,T.Zhang,A.Hirano,N.Imanishi,O.Yamamoto,Y.Takeda,

Solid State Ionics183(2011)48.

[9]S.Ohta,T.Kobayashi,T.Asaoka,J.Power Sources196(2011)3342.

[10]R.Murugan,S.Ramakumar,N.Jansmi,https://www.360docs.net/doc/1e4542608.html,mun.13(2011)1373.

[11] E.Rangasmy,J.Wolfenstin,J.Sakamoto,Solid State Ionics206(2012)29.

[12]J.Fu,Solid State Ionics96(1997)195.

[13]J.S.Thokchom,B.Kumar,J.Power Sources185(2008)480.

[14]T.Zhang,N.Imanishi,Y.Shimonishi,A.Hirano,Y.Takeda,O.Yamamoto,J.Electrochem.

Soc.157(2010)A214.

[15]J.Fu,Solid State Ionics104(1997)191.

[16]P.Maddonado-Manso,E.R.Losilla,M.Martinez-Lara,M.A.G.Aranda,S.Bruque,F.E.

Mouahid,M.Zahir,Chem.Mater.15(2003)1879.

[17]X.Xu,Z.Wen,Z.Gu,X.Xu,Z.Lin,Solid State Ionics171(2004)207.

[18]K.Takahashi,J.Ohmura,D.Im,D.L.Lee,T.Zhang,N.Imanishi,A.Hirano,M.B.

Phillipps,Y.Takeda,O.Yamamoto,J.Electrochem.Soc.159(2012)A142.[19] C.R.Mariappan,C.Yada,F.Rosciano,B.Roling,J.Power Sources195(2011)

6456.

[20]T.Salkus,A.Mindune,Z.Kamepe,J.Ronis,A.Ureinskas,A.Kezionis,A.F.Orliukas,

Solid State Ionics178(2007)1282.

[21] B.Kumar,D.Thomas,J.Kumar,J.Electrochem.Soc.159(2009)A506.

[22]P.C.Bruce,A.R.West,J.Electrochem.Soc.130(1983)662.

[23]K.Takahashi,P.Johnson,N.Imanishi,N.Sammes,Y.Takeda,O.Yamamoto,

J.Electrochem.Soc.159(2012)A1065.

[24]X.Xu,Z.Wen,J.Wu,X.Yang,Solid State Ionics178(2007)29.

[25]J.S.Thokchon,B.Kumar,J.Electrochem.Soc.154(2007)A331.

[26]J.B.Goodenough,H.Y.P.Hong,J.A.Kafalas,Mater.Res.Bull.11(1976)203.

[27]S.Hasegawa,N.Imanishi,T.Zhang,J.Xie,A.Hirano,Y.Takeda,O.Yamamoto,

J.Power Sources189(2009)371.

[28]O.Bohnke,Q.N.Pham,A.Boulant,J.Emery,T.Salkus,M.Barre,Solid State Ionics188

(2011)144.

180P.Zhang et al./Solid State Ionics253(2013)175–180