3974 Chem. Mater.

2003, 15, 3974-3990

Reviews
Lithium Lanthanum Titanates: A Review
S. Stramare,† V. Thangadurai,* and W. Weppner
Chair for Sensors and Solid State Ionics, Faculty of Engineering,
Christian-Albrechts-University of Kiel, Kaiserstr. 2, D-24143 Kiel, Germany
Received March 14, 2003. Revised Manuscript Received July 22, 2003

To date, the highest bulk lithium ion-conducting solid electrolyte is the perovskite (ABO3)-
type lithium lanthanum titanate (LLT) Li3xLa(2/3)-x0(1/3)-2xTiO3 (0 < x < 0.16) and its related
structure materials. The x ≈ 0.1 member exhibits conductivity of 1 × 10-3 S/cm at room
temperature with an activation energy of 0.40 eV. The conductivity is comparable to that of
commonly used polymer/liquid electrolytes. The ionic conductivity of LLT mainly depends
on the size of the A-site ion cation (e.g., La or rare earth, alkali or alkaline earth), lithium
and vacancy concentration, and the nature of the B-O bond. For example, replacement of
La by other rare earth elements with smaller ionic radii than that of La decreases the lithium
ion conductivity, while partial substitution of La by Sr (larger ionic radii than that of La)
slightly increases the lithium ion conductivity. The high lithium ion conductivity of LLT is
considered to be due to the large concentration of A-site vacancies, and the motion of lithium
by a vacancy mechanism through the wide square planar bottleneck between the A sites. It
is considered that BO6/TiO6 octahedra tilting facilitate the lithium ion mobility in the
perovskite structure. The actual mechanism of lithium ion conduction is not yet clearly
understood. In this paper, we review the structural properties, electrical conductivity, and
electrochemical characterization of LLT and its related materials.

1. Introduction A wide variety of metal oxides are known to exhibit

At present, lithium ion secondary battery develop- high bulk lithium ion conductivity. Using temperature
ments are mainly based on LiCoO2 as a cathode (positive as scale, they can be divided mainly into two groups:
electrode), lithium ion-conducting organic polymer as (i) high-temperature ionic conductors, for example, Li2-
an electrolyte (LiPF6 dissolved in poly(ethylene oxide) SO4,3 Li4SiO4,4 and Li14ZnGe4O16 (Lithium Super Ionic
(PEO)), and Li metal or graphite as an anode (negative Conductor, LISICON)5 and (ii) low-temperature ionic
electrode). The formation of a solid electrolyte interface conductors, for example, γ-Li3.6Ge0.6V0.4O4,6 Li3N,7,8 Li-
(SEI) at the anode leads to a large irreversible capacity β-alumina,9 Li1+xTi2-xMx(PO4)3 (M ) Al, Sc, Y, La),10 and
loss during the discharge cycles. A further major concern Li0.34La0.5TiO2.94.11 They can also be classified into four
is the safety aspect of liquid and common polymeric groups according to the type of compounds: (i) lithium
electrolytes. The liquid-free batteries show some ad- oxyacid salts, for example, Li2SO4 and Li4SiO4; (ii) γ -Li3-
vantages over the currently commercialized ones. These PO4 solid solutions, for example, LISICON and γ-Li3.6-
include thermal stability, absence of leakage and pol- Ge0.6V0.4O4; (iii) NASICON-structured Li1+xTi2-xMx(PO4)3
lution, resistance to shocks and vibrations, and large (M ) Al, Ga, In, Sc), and (iv) Li ion-conducting A-site
electrochemical windows of application, thanks to the deficient perovskite solid solution (Li-ADPESS).
use of high-voltage (∼5.5 V/Li) cathode materials. To date, the fastest lithium ion-conducting solid
Furthermore, the development in microelectronic and electrolytes known are the perovskite-type (ABO3) oxide,
information technologies requires the search for a new with A ) Li, La and B ) Ti, lithium lanthanum titanate
generation of energy sources, among which a consider- (LLT) Li3xLa(2/3)-x0(1/3)-2xTiO3,11,12 and its structurally
able interest has been given to the all-solid-state lithium related materials.12-26 This paper aims at giving a
batteries. The main impediment is finding an appropri- comprehensive review of the available information about
ate solid electrolyte that has a reasonably high lithium LLT and its related materials (A site, B site, or both
ionic conductivity and a good chemical stability in substituted LLT), with special focus on composition-
contact with both electrodes, especially with metallic structure-ionic conductivity property.
lithium or LiAl alloy anode.1,2
2. General Considerations
* To whom correspondence should be addressed. E-mail: Historically, perovskite-type alkaline-earth titanates
[email protected]. Phone: 0049 431 880 6210. Fax: 0049 431 880 6203.
† Nanogate Technologies GmbH, Gewerbepark Eschbergerweg, (ATiO3, A ) Ca, Sr, Ba) aroused much interest due to
D-66121 Saarbrücken, Germany. their functional properties, for example, dielectricity and
10.1021/cm0300516 CCC: $25.00 © 2003 American Chemical Society
Published on Web 09/23/2003
Reviews Chem. Mater., Vol. 15, No. 21, 2003 3975

of pressure, and sintering conditions on their crystal

structure, conductivity, electrochemical properties, and
the mechanism for lithium ion conduction have been
widely investigated.12,19-26,32-100

3. Synthesis of LLT and Related Materials

LLTs and related compounds have been synthesized
mainly by three methods: (i) solid-state reaction, (ii)
sol-gel synthesis,49 and (iii) floating zone.48,85 Among
them, solid-state reactions are commonly used, allowing
the production of relatively large quantities of bulk
material. Sol-gel synthesis coupled with dip coating
was introduced to prepare thin films of LLT for applying
these materials in the solid-state devices.49 A floating
zone was introduced as a second synthesis step after
the solid-state reaction to produce single crystals of
Figure 1. Arrhenius plots of electrical conductivity of per-
Li0.27La0.59TiO348 and monocrystalline fibers of Li0.5La0.5-
ovskite-type lithium lanthanum titatate Li0.34La0.51TiO2.94,16 TiO385 to study anisotropic conductivity and the phase
along with several well-known solid lithium ion conductors. diagram under directional solidification, respectively.
The sharp increase in LiAlCl4 is due to melting.
4. Structural Characterizations of LLT
ferroelectricity, which arise from the displacement of Ti
atoms toward one of the octahedrally coordinated oxy- Various solid-state techniques came in aid of the
gen atoms along the c-axis.27a Brous et al.27b reported researchers for characterizing this family of materials.
for the first time the synthesis of cubic Li0.5La0.5TiO3 The evolution of the structure and its dependence on
by the replacement of A-site alkaline earth ions in ATiO3 the composition (Li/La ratio), substitution (A site, B site,
with a trivalent rare earth La ion and monovalent ions or both), synthesis/sintering conditions, and external
(Li, Na, K). Some years later, Patil and Chincholkar28,29 pressure was studied, mainly employing powder X-ray
confirmed this result and extended their investigation and neutron diffraction (XRD and ND), and in minor
in which the La was replaced by other rare earth ions. cases electron diffraction (ED). High-resolution electron
After these structural studies, the dielectric properties microscopy (HREM) has been used to study the super-
of these perovskite-type titanates were considered by structure of LLT. Scanning electron microscopy (SEM),
Varaprasad et al.,30 who found the hysteresis and the transmission electron microscopy (TEM), and secondary
dielectric anomaly peak at 67 °C in Li0.5La0.5TiO3. On ion mass spectroscopy (SIMS) were also employed for
the basis of their results, they concluded that Li0.5La0.5- the characterization of LLT and its related compounds.
TiO3 exhibits both ferroelectric and antiferroelectric 4.1. X-ray and Neutron Powder Diffraction Study.
ordering. They and Kochergina et al.31 assigned the The solid solution with the general perovskite
structure of Li0.5La0.5TiO3 to that of tetragonal tungsten (CaTiO3)-type98,99 formula Li3xLa(2/3)-x0(1/3)-2xTiO3 has
bronze-type.31b Increase of the capacitance with increas- turned out to be stable over a wide range of compositions
ing temperature, large dielectric loss, and dielectric (∼0.04 < x e ∼0.16).14,15a,32,33,42 The actual crystal
relaxation were observed from Inaguma et al.,11 who structure is still controversial. In fact, depending on the
considered these phenomena due to Li-ion motion. amount of lattice vacancies present and the synthesis
Indeed, the first studies on the conducting behavior method, in addition to the simple cubic unit
and on the stoichiometric range of stability of LLT were cell,11,12,14,16,27,28,33,45,56,57,66,68,73,82 hexagonal,75 tetra-
reported by Belous et al.14 They reported that oxides gonal12,14,15,29-33,35,42,44,51,59,66,68,69,73,83,87,89,100 and
with nominal composition Li3xLa(2/3)-x0(1/3)-2xTiO3 show orthorhombic perovskite-type distorted
a perovskite-like structure for x in the range 0.04 < x cells11,14-16,32,33,35,42,48,52,57,59,73,76,97 were reported. An
< 0.17, recognizing that the lanthanum ions, larger in overview of the chemical composition and crystal struc-
ionic size, are the main contributors to the stabilization ture data reported for LLT in the literature is listed in
of the perovskite-type structure and that the lithium Table 1.
ions are the charge carriers responsible for their high 4.1.1. Cubic Perovskite-Type Structure LLT. The cubic
electrical conductivity.14 perovskite-type unit cell with the lattice parameter a
However, the work of Inaguma et al.11 is referred to ) 3.8 Å (space group Pm3m and Z ) 1) was reported
as the first study on the ionic conductivity of LLT since for a specific composition14,27,28,33,45,68 and for samples
they reported a bulk lithium ion conductivity of 1 × 10-3 quenched from high temperature (>1150 °C).12,57,66,73,82
S/cm at room temperature (RT). LLT has attracted the The La3+ ions, Li+ ions, and vacancies are randomly
interest of many research groups around the world distributed over the A sites. Only one work reported the
because of its possible potential application as solid identification of a double F-centered supercell with a ≈
electrolyte in various electrochemical devices, for ex- 2ap and superstructure reflections, which have been
ample, all-solid-state lithium-ion batteries, sensors, and attributed to the ordering of the La3+ and Li+ and
electrochromic displays. Figure 1 shows the Arrhenius vacancies at the A sites.56 The lattice parameter a was
plots for ionic conductivity of LLT together with well- found to decrease with increasing x in LLT (Figure 2).57
known solid lithium ion conductors. Since 1993, the Moreover, in the case of LixLa0.57TiO3 the lithium was
influence of the composition, partial substitution or total found to evaporate at elevated temperature and the
substitution of La and Ti with other metal ions, effect lattice parameter was found to increase for the x ) 0.35
3976 Chem. Mater., Vol. 15, No. 21, 2003 Reviews

Table 1. Crystallographic Data for Li3xLa(2/3)-x0(1/3)-2xTiO3 (LLT)a′

lattice constant (Å)
x) a b c Z space group/crystal system reference
0.287 3.8742(9) 7.7673(5) 1 P4/mmm 59
0.207 3.8717(6) 7.7850(8) 1 P4/mmm 59
3.8697(8) 7.7756(5) 2 P4/mmmb 69
0.167 3.861 1 Pm3m 27
12.37 4.32 10 tetragonal 31
7.732 7.760 4 tetragonal 14
5.374 5.374 7.600 4 P4mm 35
∼2ap ∼2ap ∼2ap 4 orthorhombic 68
3.8688(4) 7.7463(2) 2 P4/mmm 73
3.8703(4) 1 Pm3m 73
5.4711(4) 13.404(1) 6 R3hc 75
0.146 3.8700(1) 7.7452(2) 2 P4/mmm 73
0.142 3.8673(4) 7.7528(26) 2 P4mmm or P4/mmm 33
0.130 3.8721(1) 7.7492(4) 2 P4/mmm 100
0.125 7.738 7.745 8 tetragonal 14
0.117 3.8717(1) 1 Pm3m 33
7.74894(3) 8 cubic 56
0.113 3.8719(2) 1 cubic 11
0.110c ap 1 Pm3m 66
0.110d ap ≈2ap 2 P4/mmm 66
0.104 3.8727(3) 7.7510(3) 2 P4/mmm 73
0.100 3.8748(1) 7.7409(4) 2 P4/mmm 100
0.097 7.75333(3) 8 cubic 56
0.090 3.8801(3) 3.8723(3) 7.7747(6) 2 orthorhombic 48
3.8684(1) 7.7746(1) 2 P4/mmm 100
0.087 3.8677(6) 7.7809(1) 1 P4/mmm 59
0.083 6.928 7.745 8.202 8 orthorhombic 14
0.079 3.8708(2) 7.7809(7) 2 P4mmm or P4/mmm 33
0.073 3.8671(5) 3.8654(5) 7.7727(4) 2 Pmmm 73
0.070 3.8714(1) 7.7789(3) 2 P4/mmm 42
0.067 6.928 7.724 8.8202 8 orthorhombic 14
7.75637(4) 2 cubic 56
0.062 3.8757(1) 3.8658(1) 7.7853(2) 2 Pmmm 73
0.060 3.8756(1) 3.8649(1) 7.7852(1) 2 Pmmm 73
3.8757(1) 3.8655(1) 7.7845(2) 2 Pmmm 73
0.050 7.7313(1) 7.7520(1) 7.7840(1) 8 Cmmm 97
0.0417 6.916 7.732 8.195 8 orthorhombic 14
0.04 3.8747(1) 3.8632(1) 7.7846(1) 2 Pmmm 73
0.03 3.8753(1) 3.8638(1) 7.7901(2) 2 Pmmm 73
a a ) cubic perovskite lattice parameter. b 27 °C. c Intermediate temperature polymorph (R), disordered phase (prepared by rapid
p
quenching from high temperature to room temperature).33 d Low-temperature polymorph (β), ordered phase (prepared by normal cooling
from the elevated temperature).33

4.1.2. Hexagonal Perovskite-Type Structure LLT. The

hexagonal unit cell has been reported for Li0.5La0.5TiO3-δ
(0e δ e 0.06) in a recent ND study,75 the distortion
being attributed to the tilting of the TiO6 octahedra.
Unit cell parameters were a ) 5.4711(4) Å and c )
13.404(1) Å, with space group R3 h c (Z ) 6), corresponding
to one of Glazer’s octahedral tilt schemes for perovskite-
related structures,101 which is adopted when the tilting
angle of the octahedra remains small (e.g., as in the case
of LaAlO3, LaCuO3, or LaNiO3). La, Ti, and O occupy
the Wyckoff position 6a site (0,0,0,25), 6b site (0,0,0),
and 18e site (x,0,0.25), respectively. Difference Fourier
map calculation allowed location of Li in the 18d
(0.5,0,0) position.75 The structure is constituted of nearly
Figure 2. Variation of the cubic perovskite lattice parameter regular TiO6 ochahedra (Ti-O 1.943 Å) and La is 12-
a, as a function of lithium content in Li3xLa(2/3)-x0(1/3)-2xTiO3.57 fold (La-O 2.559-2.911 Å) and Li 4-fold coordination
The samples were prepared by quenching the pellets from 1350 with oxygen (Figure 3). The Li ions are placed in the
°C into liquid nitrogen. The line passing through the data
middle of windows, formed by four TiO6 units, in square-
points is a guide to the eye.
planar configuration with Li-O bond lengths of 1.81-
and decrease for the x ) 0.30 member with increasing 2.07 Å.
sintering temperature from 1150 to 1350 °C.82 Lithium Recent investigation of Li0.5La0.5TiO3 by means of
titanium oxide and lanthanum oxide impurity phases TEM and SIMS resulted in the identification of a rigid
were found to appear when the sintering time was frame, constituted of La-Ti-O, with the ordering of the
shorter or longer than the optimum condition (6 h at La3+, resulting in a superstructure where unit cell edges
1350 °C).16 are doubled. The diffraction patterns of TEM mainly
Reviews Chem. Mater., Vol. 15, No. 21, 2003 3977

Figure 3. Schematic structure of pseudo-cubic perovskite-

type La0.5Li0.5TiO3. Ti atoms occupy the corners of the cube,
oxygen atoms are at the middle of the edges, La ions (or
vacancies) are the geometrical center of the cube, and Li ions
are at the center of each face. The La position is half occupied;
the Li position is one-sixth occupied.75

follow the rule for which the structure factor F ) 0 if (h

+ k + l) is even and F ) 0.5fLa if (h + k + l) is odd,
differentiating itself from the XRD response. This was
attributed to the uneven distribution of La3+ ions at the
A sites. Li2TiO3, as a second phase, was found at the
grain boundary junctions.68
4.1.3. Tetragonal Perovskite-Type Structure LLT. For Figure 4. Crystal structure of tetragonal Li3xLa(2/3)-x0(1/3)-2xTiO3
tetragonal phases, three types of cells have been pro- (x ) 0.11). (Top) ball and stick and (bottom) polyhedral view.
posed. Historically, for Li0.5La0.5TiO3-δ (0 e δ e 0.06),29-31 Unit cell constant: a ) 3.8741(1) Å and c ) 7.7459(5) Å: space
a tungsten bronze-like31b cell with lattice parameters group P4/mmm.42 The TiO6 octahedra are connected to each
a ) 12.16 Å and c ) 3.86 Å 30 or a ) 12.37 Å and c ) other by corners.
4.32 Å (Z ) 10)31a has been proposed. Recent studies,
employing both XRD and ND, identified two perovskite- ordered arrangements is found to increase with decreas-
like unit cells for Li3xLa(2/3)-x0(1/3)-2xTiO3 (0.04 < x < ing x.14,32,55 Irregular contrast, shown by 〈100〉 HREM
0.14): (i) x2ap × x2ap × c(∼2ap) (ap is the ideal cubic images, may be correlated to local La population fluc-
perovskite lattice parameter);35 (ii) ap × ap × c(∼2ap) tuations.42 For the x values in the range ∼0.15 > x >
with space group P4mmm (No. 99)12,32,33,42,73,100 or P4/ ∼0.04, the tetragonal phases differ from each other only
mmm (No. 123).14,66,68,69,89 in the intensity of the superstructure reflections.32 The
In the first case, the so-called “diagonal-distorted” unit tetragonal distortion c/2a nearly vanishes at x ) 0.08.
cell was suggested by Várez et al.35 for the Li3x- Ruiz et al.59,69 recently proposed that the actual unit
La(2/3)-x0(1/3)-2xTiO3-δ (∼0.06 < x < ∼0.16, 0 e δ e formulaforthiscompoundshouldbedoubled,La1.33-xLi3xTi2O6
0.06).32,33,42 The distortion was attributed to a 1:1 Li and (0.1 < x < 0.3), by considering the ordering of La3+ ions,
La ordering along the c-axis, and a tilting of the TiO6 Li+ ions, and vacancies. La and Li are 12-fold coordi-
octahedra was not excluded. This interpretation was nated with oxygen.
refused by Fourquet et al.,42 and they proposed a second 4.1.4. Orthorhombic Perovskite-Type Structure LLT.
type of cell on the basis of XRD and TEM studies, which The orthorhombic unit cell is observed mainly for a very
has currently been accepted12,42,44,51,59,73,83,87 (Figure 4). small amount of Li (x < 0.08).14,15,73,76 The distortion is
The TiO6 octahedra are distorted along the c-axis with generally attributed to the ordering of the A-site species
one short Ti-O2 (∼1.8 Å) bond opposed to one long Ti- (Li+, La3+, 0). Two types of unit cells have been
O1 (∼2 Å) bond and four equal Ti-O3 (∼1.94 Å) bonds. proposed: (i) a(∼2ap) × b(∼2ap) × c(∼2ap), where all
The tetragonal distortion is attributed to the unequal the parameters are doubled; (ii) a(∼ap) × b(∼ap) ×
distribution of vacancies, Li+ and La3+ cations on the c(∼2ap).
two possible crystallographic sites 1a and 1b (12-fold In the first case, the doubling of the c parameter is
coordinated with oxygen) in the space group P4/mmm,42 attributed to the ordering of the vacancies along the 001
analogous to the superstructure data reported for direction, while the doubling of the parameters a and b
La2/3TiO3-λ (λ ) 0.007-0.079).102 This phenomenon is is due to the ordering of La3+ and Li+ ions in the XY-
favored by the increase of the sintering time or of the plane [110].14 The proposed space groups are Cmmm (x
sintering temperature above 1200 °C. The extent of the ) 0.05) (No. 65)97a and Cm2m (x ) 0.04) (No. 38)97b for
3978 Chem. Mater., Vol. 15, No. 21, 2003 Reviews

The dependence of the spectra and of the relaxation

times T1 (spin-lattice relaxation), T2 (spin-spin relax-
ation), and T1F (spin-lattice relaxation in the rotating
frame) on both temperature and composition was in-
vestigated. The magnitude of T1 and T2 reflects the
strength of the interactions between the nuclei and their
environment. Static powder pattern, with or without
magic angle spinning (MAS), were recorded on
Li3xLa(2/3)-x0(1/3)-2xTiO3 (0.03 e x e 0.167) samples with
different crystal symmetries.40,41,47,51,54,55,61,66,67,73,76,80,81,95
5.1. NMR Studies on Cubic (ap) Structure LLT.
For the x ) 0.167 cubic LLT compound, absence of
satellite peaks allowed the conclusion that Li ions
occupy highly symmetric sites.40 At T < 200 K the line
width of the spectra has been attributed mainly to Li-
Li dipole interactions, while at T > 200 K ionic motion
Figure 5. Crystal structure of orthorhombic Li3x- was identified. Analysis of 1/T2 and 1/T1 vs 1000/T data
La(2/3)-x0(1/3)-2xTiO3 (x ) 0.05). Lattice parameters: a ) for x ) 0.167 yielded an activation energy for Li motion
7.7313(1) Å; b ) 7.7520(1) Å, and c ) 7.7840(1) Å; space group Es ) 0.15 eV. Further work showed that the spin-lattice
Cmmm.97
relaxation can be described by a stretched-exponential
correlation function.47 Ngai and León have demon-
the former and Pmmm16 for the latter. Recent ND strated, based on their calculation from the macroscopic
studies97 support the former double cell a(∼2ap) × electrical relaxation data, that this function is the same
b(∼2ap) × c(∼2ap), which contains a 3D framework of as the microscopic ion hopping conductivity correlation
corner-sharing TiO6 octahedra that are tilted along the function.64,65 Therefore, the property of crystalline LLT
b-axis, alternatively arranged La along the c-axis, and turns out to be a special case, distinctly different from
Li in two equivalent off-centered positions at the A site ionic glasses, which exhibit a large difference between
(Figure 5). Li+ ions are coordinated by four oxygen ions nuclear spin relaxation and ion conductivity relax-
with Li-O lengths in the range 1.75-2.21 Å, which are ation.105
in good agreement with the expected ones from the ionic
radii after Shannon (Li(IV) ) 0.59 Å and O(IV) ) 1.40 5.2. NMR Studies on Tetragonal (ap × ap × 2ap)
Å).103 Structure LLT. In tetragonal perovskites (a ) b ) 3.87
Å; c ) 2a),42,51,61 two kinds of Li ions exist (Figure 4
For the latter a(∼ap) × b(∼ap) × c(∼2ap) type, the
and section 4.1.3.) in different environments. One kind
distortion is attributed to small tilting of octahedra in
of Li is present in (or near) the 1a (0,0,0) site and
the ab-plane or along the c-axis.76 The Ti atoms are
preferentially surrounded by La3+, and another kind of
displaced from the center of the octahedral cavities
Li is present in (or near) the 1b (0,0,0.5) site and
toward the z/c ) 0.5 layer, perhaps to compensate for
surrounded by either La3+, other Li+, and /or vacancies
the asymmetric distribution of charges around the
in the space group P4/mmm. These may explain the
octahedral site. These structural features resemble
different mobilities recorded by NMR spectra. The spin-
those reported by Fourquet et al.42 and MacEachern et
lattice relaxation times T1 vs 1/T curves showed a strong
al.104 The XRD analysis of single crystals indicated an
asymmetry,51,61 which could not be explained by an ionic
orthorhombic distorted perovskite structure. Superlat-
motion occurring in lower dimensions (1D or 2D) as
tice lines were attributed to the difference in the
described by Richards.106 In a first approximation, this
occupancy of the La(1) (0,0,0) site and the La(2)
asymmetry was explained in terms of correlated
(0,0,1/2) site accompanied by a shift of the Ti ion from
motion107-110 and by assuming that the dynamics is
the symmetric center position of the TiO6 octahedron.48
governed by a distribution of activation energies for the
thermally activated Li+ ion hops.
5. Nuclear Magnetic Resonance (NMR)
The existence of a distribution of energy barriers is
Investigations
undoubtedly connected to the disorder present in these
Solid state 7Li NMR is a powerful tool for studying crystalline materials110,111and it is clearly revealed by
the motion of ions. 7Li is a quadrupolar nucleus (I ) the powder XRD analysis. The T1 (x ) 0.11) values41
3/ ), so static 7Li NMR signals depend on the interactions
2 are comparable with those measured by Latie et al.13
between the electrical quadrupole moment of the Li+ for Li0.05La0.33Nb0.95Ti0.05O3. Very recently, T1 and T1p
nucleus and the electric field gradient (EFG) at the site. studies, performed on the 6Li nucleus,95 clearly show
The EFG arises as a result of the other ions surrounding that just dipolar nuclear interaction is responsible for
the lithium ion. This makes the static spectra sensitive Li+ relaxation. This result is different from that previ-
to changes in lattice structure and motions. Usually, the ously put forward regarding the relaxation process in
static NMR spectrum of 7Li is formed by a central line, these compounds.
arising from the -1/2 T 1/2 transition, and by two The spin-lattice relaxation time T1 of 7Li was mea-
satellites lines, coming from the 1/2 T 3/2 and -1/2 T 3/2 sured for tetragonal structure LLT (x ) 0.11) together
transitions associated with the electric quadrupolar with cubic structure of the same position as a function
interactions. of the order parameter S.66,80 Major observations are a
The spin-lattice relaxation time (T1) is a measure of similar extent of the strong Li-Li correlation irrespec-
the local charge-density fluctuations due to ionic motion. tive of S and a decrease of the uniformity of the Li sites
Reviews Chem. Mater., Vol. 15, No. 21, 2003 3979

with increasing S. Plots of 1/T1 vs 1000/T as a function (160-500 K), but Arrhenius local fits can be made,
of the order parameter S show a minimum at ∼70 °C, giving an activation energy Ea ) 0.40 eV at low
shifting very slightly to a lower temperature with temperature and Ea ) 0.26 eV at high temperatures.
decreasing S.66 This observation roughly follows the
Bloemberger-Purcell-Pound expression, which as- 7. Electrical Conductivity and Electrochemical
sumes isotropic, 3D motion. The magnitude of the T1 Characterization of LLT
value (∼0.1 s at 250 K for x ) 0.11) is comparable to
that of León et al.40 but it was found to be much higher A major role in studying the electrical properties of
compared to the Emery et al. value (4 ms at 300 K for LLT and related materials has been performed by ac
x ) 0.08).51 This difference assumed due to the para- impedance spectroscopy. Different electrodes and at-
magnetic impurities in the samples.66 However, Emery mospheres have been employed; however, the interpre-
et al. related the spin-lattice relaxation (T1) to correla- tation of the evolution of the ionic conductivity with the
tion time, τc, of ionic motion and proposed two different temperature is not unique. dc measurements allowed
natures for lithium ion motion in LLT.51 determination of the prevalent ionic character of the
5.3. NMR Studies on Orthorhombic (ap × ap × conduction, providing a value of transference number
2ap) Structure LLT. In orthorhombic perovskites (a for electrons te ∼ 10-5. Electrochemical techniques, such
) 3.864 Å, b ) 3.875 Å, c ) 7.786 Å)73 two types of Li, as coulometric titration, galvanostatic cycling, and cyclic
with different mobility, have been ascribed to Li located voltammetry, have been used to collect information
in two different planes of the perovskite. The amount about the Li intercalation and deintercalation properties
of Li and vacancies observed in the plane z/c ) 0.5 is and the stability of LLT in electrochemical systems.
significantly higher than that in z/c ) 0. From this fact, 7.1. ac Impedance Measurements. The ionic con-
the mobility of Li ion should be favored in the ab-plane ductivity of LLT was measured by ac impedance mostly
(2D) with respect to the perpendicular c-axis.73 Li with over the frequency range from 5 Hz to 13 MHz and the
low mobility, located at z/c ) 0 planes, displays impor- temperature range 150-700 K. Influence of composition,
tant quadrupole constants while Li with high mobility, pressure, sintering, and current collectors was widely
located at z/c ) 0.5 planes, displays small constants. The investigated. When Au or Pt are employed as electrodes,
increase of Li mobility along the series Li3x- LLT shows a small bulk resistance at the high-
La(2/3)-x0(1/3)-2xTiO3 (0.03 e x e 0.167) is responsible for frequency side, a large grain boundary contri-
the progressive line narrowing of the central line bution,11,14-17,31,32,34,44,45,48,51,52,57,59,63,82,84,86 (Rgb ≈ 50Rb32
detected in lithium NMR spectra. for sintered samples), and a tail in the low-frequency
side, suggesting a blocking effect of the electrodes on
6. Electrical Conductivity Relaxation (ECR) the mobile ions (lithium) (Figure 6).11 All the solid
Measurements solutions showed a non-Debye response,112,114-117 typical
for ionic conductors, and the ac data can be interpreted
ECR studies have been conducted to obtain informa- by means of the Randles equivalent circuit.118a More-
tion about the dynamics of the transport process in over, it has been observed that the shape of the curves
LLT.40,47,64,65,76 It has been demonstrated that when in the complex impedance plane is strongly influenced
NMR and ECR measurements are conducted in the by the type of electrodes (current collectors) employed
same frequency and temperature range, the results can for the measurements. Figure 6b shows a typical imped-
be correlated and the activation energies for short- and ance plot for LLT with reversible lithium electrodes PP/
long-range motion in these perovskites can be deter- LiClO4/PC using Li metal electrodes.11 The appearance
mined. of the low-frequency intercept118b in this case is proof
The frequency dependence of the real part of the of lithium ion conduction in the perovskite-type materi-
conductivity shows a low-frequency plateau and a als.
crossover to power-law dependence at high frequency. 7.2. dc Measurements. dc experiments, with Li-ion
This behavior is characteristic of ion hopping and can blocking Ag electrodes, were conducted both on bulk and
be described using Jonscher’s expression112 thin film samples. Electronic conduction values in the
range 5 × 10-9 to 1 × 10-8 S/cm at room temperature
σ*(ω) ) σdc[1 + (jω/ωp)n] (1) were reported.11,31 The transference number (te) 45 of the
electrons in pure LLT was found to be less than 10-5 at
where σdc is the dc conductivity, ωp is a crossover room temperature. Measurements of the lithium ion
frequency, and the exponent n ≈ 0.6 is related to the transference number tLi+ by Tubandt’s method at 480
degree of correlation among moving ions. °C, following the procedure of Hladik,119 confirmed the
The dispersive behavior of the conductivity in the lithium ion conductivity and the ionic conductivity to
frequency domain can be interpreted in terms of a be entirely due to the lithium ion motion (tLi+ = 1).14
Kohlrausch-Williams-Watts113 (KWW) correlation func- 7.3. Lithium Ion Conductivity of LLT. The x ∼
tion φ(t) in the time domain, which takes the form of a 0.11 member of LLT exhibits the highest bulk lithium
stretched exponential: ion conductivity of 1 × 10-3 S/cm at room temperature.
A collection of bulk lithium conductivity and the respec-
φ(t) ) exp[-(t/τσ)βσ] (2) tive activation energies for LLT is given in Table 2
together with its related structure materials. Arrhenius
Here τσ is a temperature-dependent relaxation time, plots for electrical conductivity shows a bend/break of
thermally activated with the same Ea as the dc conduc- the curves at higher temperature.11 Figure 7 shows the
tivity. The temperature dependence of the dc conductiv- Arrhenius plots of the ionic conductivity of LLT deter-
ity is non-Arrhenius over the whole temperature range mined using ac impedance of polycrystalline (x ) 0.11)
3980 Chem. Mater., Vol. 15, No. 21, 2003 Reviews

Figure 7. Arrhenius plots for lithium ion conductivity of

Li3xLa(2/3)-x0(1/3)-2xTiO3 with different compositions. O: x ) 0.11
(poly crystalline); 0: x ) 0.09 (single crystal parallel to
c-axis);48 4: x ) 0.06 (dc method);76 and 1: x ) 0.167 (dc
Figure 6. Typical ac impedance plots obtained in the fre- method).47
quency range 5 Hz to 13 MHz for Li0.34La0.5100.15TiO2.94 in air.
(a) Lithium ion blocking Au pasted or sputtered electrodes at site, when the B-site ion occupies the 1a site) (Figure
27 °C and (b) lithium ion reversible electrodes at 17 °C. The 8).32 The temperature increase as well as the presence
appearance of a spike in a plot is typical for the ionic blocking
behavior and absence of a tail in b indicates that the LLT is a
of more vacancies in the structure may facilitate the
lithium ion conductor.11 The magnitudes of the capacitance of tilting of the octahedra, which favors a thermally
the bulk and grain boundary region are in the order of 100 pF assisted conduction mechanism.
and 10 nF, respectively. The ionic conductivity is highly sensitive to the
lithium content and a dome-shaped dependence of
and single crystal (x ) 0.09) and dc method for x ) 0.06 the conductivity on the Li content was ob-
and 0.167 members.11,47,48,76 At low temperature all the served15a,19,20,32,39,51,57,73 (Figure 9). The effect of ionic
compounds exhibit similar conductivity values, while at conductivity on lithium content is discussed in section
higher temperature, polycrystalline materials show 7.5.2. The application of an external isostatic pressure17
slightly higher conductivity than the single-crystal decreases the ionic conductivity due to lattice deforma-
compound. The bending at high temperature is consid- tions. Thin films of the same composition of bulk
ered from some authors due to a phase transition, materials showed a lower conductivity (σ120°C ) 10-8-
taking place at about 127 °C, which induces a conduc- 10-7 S/cm) with larger activation energy (1.05 eV).49
tion process with two different activation Both conductivity and activation energy were found to
energies.11,12,14-16,32,45,55,57,59,66,70,84 Most of the authors, be dependent on the sample preparation. dc measure-
anyway, describe the conduction behavior of LLT as ments perpendicular to the surface showed high resis-
thermally activated, like ionic conducting solid glasses120 tance, which may be due to the presence of pyrochlore
and define such a dependence as non-Arrhenius behav- La2Ti2O7 impurity phase in the sample.49
ior.40,41,44,47,48,51,61,64,65,73,76 7.3.1. Degree of Ordering of A-Site Cations and
At higher temperature (T > 127 °C), conductivity data Vacancy. The degree of ordering of the cations and
could be fitted to a Vogel-Tamman-Fulcher (VTF) type vacancies on the A sites influences strongly the crystal
relationship,121-123 talking then of a thermally assisted structure as well as lithium ion conductivity in LLT.
mechanism of conduction.44 Originally, the VTF equa- An extensive study has been conducted by Harada and
tion was developed to deal with viscosity properties of others,12,52,57,66,70 who introduced an order parameter S,
supercooled liquids124 and it has been used to explain which is defined for alternative arrangement in the
the temperature dependence of the ionic conductivity tetragonal phase, as57
of polymer electrolytes.125 The VTF behavior suggests
that the mechanism of conduction involves tilting or R(La-rich) - R(dis)
rotating of the TiO6 octahedra, leading to an opening S) (3)
1 - R(dis)
and closing of the bottleneck through which the lithium
ion has to move to enter an adjacent A-site vacancy, where R(La-rich) and R(dis) are the occupancies of the
accompanied by a positive volume of activation. The A sites by La3+ ions in the La-rich layers of the ordered
bottleneck is located in the space between the two form and in the (001) plane of the disordered form,
adjacent A sites, surrounded by four oxygens (in cubic respectively. The disordered cubic form (which can be
lattice with space group Pm3m corresponds to the 3c quenched from high temperatures) has a higher con-
Reviews Chem. Mater., Vol. 15, No. 21, 2003 3981

Table 2. Bulk Lithium Ion Conductivities at 27 °C and Activation Energies of Selected LLTs and Its Related Structure
Materials
compound σbulk (S/cm) Ea (eV) reference
Li0.34La0.51TiO2.94 1.0 × 10-3 0.40 11
Li0.27La0.59TiO3 6.8 × 10-4 ⊥c-axis 0.36 48
5.8 × 10-4 |c-axis 0.35
Li0.34La0.56TiO3 (cubic, ap) 1.53 × 10-3 0.33 57
Li0.34La0.56TiO3 (tetragonal, ap, 2ap) 6.88 × 10-4 0.35 57
Li0.10La0.63TiO3 7.9 × 10-5 0.36 15a
(Li0.1La0.5)0.9Sr0.1TiO3 1.5 × 10-3 15a
Li0.15La0.51Sr0.15TiO3 5.3 × 10-5 0.38 15b
Li0.25La0.41Sr0.25TiO3 7.6 × 10-5 0.35 15b
(Li0.1La0.63)(Mg0.5W0.5)O3 ∼10-6 0.39 15a
(La0.5Li0.5)1-x(La0.5Na0.5)xTiO3 (x > 0.4) <10-7 15a
(La0.5Li0.5)1-x(La0.5Na0.5)xTiO3 (x ) 0.25) 2 × 10-5 15a
Li0.5(La0.4Nd0.1)TiO3 ∼10-3 0.33 45
Li0.245La0.592Ti0.98Mn0.02O3 ∼10-3 58
La0.58Li0.36Ti0.95Mg0.05O3 2.1 × 10-4 0.29 78
La0.56Li0.36Ti0.95Al0.05O3 6.4 × 10-4 0.26 78
La0.55Li0.36Ti0.95Mn0.05O3 1.9 × 10-4 0.29 78
La0.55Li0.36Ti0.95Ge0.05O3 3.6 × 10-4 0.29 78
La0.55Li0.36Ti0.95Ru0.05O3 5.2 × 10-5 0.28 78
La0.51Li0.36Ti0.95W0.05O3 7.3 × 10-4 0.27 78
La0.54Li0.36TiO3 8.9 × 10-4 0.29 78
La0.55Li0.36Ti0.995Al0.005O3 1.1 × 10-3 0.28 78
Li0.067La0.64TiO2.99 7.9 × 10-5 0.36 91
Li0.06La0.66Ti0.93Al0.06O3 1.7 × 10-6 0.36 91
Li0.10La0.66Ti0.90Al0.10O3 7.3 × 10-6 0.35 91
Li0.15La0.66Ti0.85Al0.15O3 9.6 × 10-6 0.36 91
Li0.20La0.66Ti0.80Al0.20O3 4.3 × 10-5 0.33 91
Li0.25La0.66Ti0.75Al0.25O3 7.7 × 10-5 0.35 91
Li0.30La0.66Ti0.70Al0.30O3 1.7 × 10-5 0.33 91
La(1/3)-xLi3xNbO3
x ) 0.01 4.06 × 10-6 0.36 137
x ) 0.02 2.33 × 10-5 0.33 137
x ) 0.03 3.52 × 10-5 0.35 137
x ) 0.04 4.25 × 10-5 0.37 137
x ) 0.05 3.85 × 10-5 0.36 137
x ) 0.06 3.82 × 10-5 0.34 137
La0.67-xNa1.5xTiO3
x ) 0.08 <10-7 0.92 87b
x ) 0.14 <10-7 0.92 87b
Li0.25La0.25TaO3 1.4 × 10-5 0.35 136b

ductivity with a lower activation energy, Ea ) 0.33 eV, conduction, which is associated with the a-axis contrac-
than the ordered tetragonal phase (obtained after an- tion in the primitive cell. The ionic conductivity de-
nealing at high temperatures), Ea ) 0.36 eV. Cycling creases with increasing order parameter S (Figure 10).12
experiments have demonstrated that the order param- 7.3.2. Electrochemical Characterization. Due to the
eter S can be changed reversibly by annealing in a range very high lithium ion conductivity, LLT has been
from 600 to 1150 °C. The decrease in ionic conductivity investigated with regard to its stability against lithium
is attributed to an increase in activation energy for ionic and intercalation of lithium using several electrochemi-
cal methods that include galvanostatic insertion,34,38,44
coulometric titration,25,44,50 and discharge/charge char-
acteristics.34,44

Figure 8. Schematic representation of the bottleneck for

lithium ion conduction in the perovskite structure Li3x-
La(2/3)-x0(1/3)-2xTiO3. Li+, La2+, and vacancy are surrounded by Figure 9. Variation of the lithium ion conductivity at room
12 oxygen ions. The diameter of the bottleneck is estimated temperature of Li3xLa(2/3)-x0(1/3)-2xTiO3 as a function of lithium
to be 1.07 Å using the ideal perovskite cell, ap ) 3.871 Å, and concentration. Dome-shaped behavior with a maximum at x
the ionic radius 1.40 Å for O2- ion with six coordination.17 ≈ 0.1 is observed.32
3982 Chem. Mater., Vol. 15, No. 21, 2003 Reviews

Figure10. Bulkionicconductivityat25°CforLi3xLa(2/3)-x0(1/3)-2xTiO3 Figure 12. Coulometric titration curves of Li3xLa(2/3)-x0(1/3)-2x-

(x ) 0.11) as a function of the order parameter S.12 The line TiO3.25 The measurements were performed in an electrochemi-
passing through the data points is a guide to the eye. cal cell with a three-electrode configuration, using 1 M solution
of LiClO4 in PC as an electrolyte and metallic Li as counter
and reference electrodes. The thickness of the LLT pellets
employed for the measurements were 250-350 µm. The lines
passing through the data points are guides to the eye.

T2 of the 7Li nuclei clearly reveals a modification of the

Li+ environment during intercalation and explained by
the formation of polarons.67,129 Two relaxation mecha-
nisms that were observed in the ac impedance spectros-
copy of the Li-intercalated samples clearly indicate the
two conductive pathways for the ions and the elec-
trons.67,118b For the mixed conducting sample, the
electronic conductivity is predominant at low temper-
Figure 11. Schematic representation of the 001 plane of the
atures, and the ionic one becomes more and more
ideal perovskite ABO3. The solid and broken line represent
the A-3c-A and 3c-3c-3c paths, respectively.38 important at high temperatures.
Coulometric titration measurements, conducted at
The fast reduction of Ti4+ ions to a lower valence of room temperature on Lix+δLa(0.66-x)TiO3 samples with
3+ by lithium insertion in A and 3c site was rationalized x ) 0.14, 0.23, 0.32, and 0.35 showed a sharp drop of
by the lattice self-potential, site potential, and lattice the potential from 2.8 V to 1.7 vs Li/Li+ due to the
energy calculation.34,126,127 Interestingly, lithium inserts change in the stoichiometry from δ ) 0 to δ ) 1.5 ×
preferably first into vacant A sites and then into 10-3 25,50 (Figure 12). A contemporaneous change in color
interstitial 3c sites,16,17 suggesting that occupation of from ivory white to gray was observed. Intercalation of
A-site vacancies in the first stage of insertion is ener- more Li than the expected from the number of available
getically favorable compared to the insertion into 3c vacancies produced a non s-shaped titration curve,
sites.38 The pathway of the inserted Li+ migration can approaching a plateau attributed to the possible forma-
be assumed to follow two cases: (i) from an A site to a tion of a second phase at δ > 0.18.50 The intercalated
3c site (abbreviated A-3c-A below) for the A-site lithium could not be fully extracted from the sample
deficient perovskite and (ii) from a 3c site to another 3c anymore in this case and the sample remained slightly
site (abbreviated 3c-3c-3c below) (Figure 11). The site gray after deintercalation.
potential at the 3c site can be assumed to take up both The chemical diffusion coefficient, calculated by means
positive and negative values, depending on the valence of galvanostatic intermittent titration technique (GITT)
of A and B. For example, for the couple A ) 0 and B ) theory,130a was estimated to be D̃ ) 10-6-10-7 cm2/s for
6, it has negative value. The high valence of the B cation both insertion and extraction of lithium.50 The D̃ values
should be advantageous for lithium insertion and con- do not change markedly with respect to stoichiometric
duction in the perovskite ABO3. change δ. The slightly higher values of D̃ in the
Interestingly, preferential lithium insertion into in- stoichiometric range δ ) 0-0.02 may be attributed to
terstitial sites for A site deficient perovskite oxides128 the comparably large number of empty lattice sites,
has not been taken into account. The calculated ideal giving more insertion paths. Furthermore, the increase
capacity was 160 mA h g-1.34 The intermittent galvano- of D̃ over both stoichiometric ranges of δ ) 0-0.02 and
static discharge/charge measurement44 showed a pla- δ ) 0.12-0.14 may be due also to an increase of the
teau in the discharge curve around 1.5 V vs Li/Li+ and Wagner enhancement factor, due to the increased slope
a maximum intercalation ratio x ) 0.15 Li per mole of of the titration curve.130.
LLT. The charge curve showed that only 60% of the 7.4. Studies on Doped LLT. Chemical substitution
lithium could be extracted in this material and 50% very has been carried out mainly for two purposes: (i) to
easily at a potential equal to the intercalation potential. improve the ionic conductivity and (ii) to stabilize the
7Li NMR experiments clearly showed that during LLT against elemental Li, to use as solid electrolyte for
intercalation a modification of the Li+ ion environment secondary batteries. An overview of the influence of
takes place. The variation of the relaxation times T1 and substitution on the crystal structure is given in Tables
Reviews Chem. Mater., Vol. 15, No. 21, 2003 3983

Table 3. Crystallographic Data for A-Site Substituted Li3xLa(2/3)-x0(1/3)-2xTiO3 (LLT)a′

lattice constant (Å)
chemical formula a b c Z space group/crystal system reference
Li0.5Nd0.5TiO3 5.44 5.57 7.75 4 orthorhombic 29
Li0.5Nd0.5TiO3 12.43 4.24 10 tetragonal 31
Li0.5Nd0.5TiO3 7.652 7.672 8 tetragonal 55
β-Li0.3125Nd0.5625TiO3 3.8245(3) 7.6809(9) 2 P4mmm or P4/mmm 33
R′-Li0.3125Nd0.5625TiO3 5.4121(5) 7.6345(14) 8 P4mmm or P4/mmm 33
C-Li0.3125Nd0.5625TiO3 5.4007(14) 5.4345(6) 7.6534(21) 4 Pbnm 33
A-Li0.3125Nd0.5625TiO3 3.8356(2) 1 Pm3m 33
Li0.34Nd0.55TiO3 3.8204(6) 3.823(1) 7.679(1) 2 orthorhombic 16
Li0.35Nd0.55TiO3 5.43665(7) 7.6649(1) 5.39761(7) 4 Pnma 37
Li0.5Sm0.5TiO3 5.39 5.59 7.71 4 orthorhombic 29
Li0.5Sm0.5TiO3 5.37 5.87 7.55 4 orthorhombic 31
Li0.38Sm0.52TiO3 5.340(3) 7.619(3) 5.441(2) 4 orthorhombic 16
Li0.5Pr0.5TiO3 12.40 4.30 10 tetragonal 31
Li0.34Pr0.56TiO3.01 3.8372(6) 7.701(1) 2 tetragonal 16
Li0.38Pr0.54TiO3 5.43793(5) 7.66828(8) 5.40769(5) 4 Pnma 37
A-Li0.26Pr0.58TiO3 3.8434(1) 1 Pm3m 43
β-Li0.26Pr0.58TiO3 3.8332(1) 7.7226(4) 2 tetragonal 43
Li0.41Pr0.53TiO3 5.4384(5) 7.6716(7) 5.4064(5) 4 Pnma 43
Li0.5Gd0.5TiO3 5.34 5.61 7.67 4 orthorhombic 29
Li0.5Gd0.5TiO3 5.33 5.79 7.54 4 orthorhombic 31
Li0.5Dy0.5TiO3 5.30 5.59 7.62 4 orthorhombic 29
Li0.5Dy0.5TiO3 5.29 5.73 7.53 4 orthorhombic 31
Li0.5Y0.5TiO3 5.25 5.58 7.54 4 orthorhombic 29
Li0.5Y0.5TiO3 5.28 5.73 7.52 4 orthorhombic 31
Li0.5Eu0.5TiO3 5.36 5.88 7.55 4 orthorhombic 31
Li0.5Tb0.5TiO3 5.32 5.74 7.55 4 orthorhombic 31
Na0.5La0.5TiO3 3.865 1 Pm3m 27
Na0.5La0.5TiO3 3.8734(2) 1 cubic 11
K0.5La0.5TiO3 3.899 1 Pm3m 27
Rb0.5La0.5TiO3 3.895 1 Pm3m 27
Ag0.5La0.5TiO3 3.874 1 Pm3m 27
Tl0.5La0.5TiO3 3.875 1 Pm3m 27
(K0.5La0.5)xBa1-xTiO3
x ) 0.1 3.999 1 cubic 30
x ) 0.2 3.989 1 cubic 30
x ) 0.3 3.963 4.065 1 tetragonal 30
x ) 0.5 3.952 1 cubic 30
x ) 0.7 3.912 1 cubic 30
x ) 0.9 3.899 1 cubic 30
x ) 1.0 3.894 1 cubic 30
La(2/3)-xSrxLixTiO3
x ) 0.06 5.477(6) 5.47483) 7.767(4) 4 orthorhombic 15b
x ) 0.10 5.481(3) 5.476(2) 7.768(4) 4 othorhombic 15b
x ) 0.15 5.489(2) 5.485(4) 7.769(3) 4 orthorhombic 15b
x ) 0.20 5.493(4) 5.490(4) 7.771(3) 4 othorhombic 15b
x ) 0.25 5.495(3) 5.491(1) 7.772(4) 4 othorhombic 15b
x ) 0.275 5.497(3) 5.493(1) 7.774(3) 4 othorhombic 15b
x ) 0.30 5.499(3) 5.494(1) 7.775(1) 4 othorhombic 15b
(Li0.5La0.5)xBa1-xTiO3
x ) 0.2 3.995 1 cubic 30
x ) 0.3 3.986 1 cubic 30
x ) 0.4 3.979 1 cubic 30
x ) 0.5 3.975 1 cubic 30
x ) 1.0 12.160 3.86 10 tetragonal 30
Li0.33(La1-xNdx)0.56TiO3
x)0 3.873 1 cubic 70
x ) 0.25 3.863 1 cubic 70
x ) 0.5 3.848 1 cubic 70
x ) 0.75 3.832 7.686 2 tetragonal 70
x ) 1.0 3.824 7.708 2 tetragonal 70
[AgyLi1-y]3xLa(2/3)-xTiO3
x ) 0.09 y ) 0 3.873(3) 7.750(8) 2 P4/mmm 87
x ) 0.09 y ) 1/3 3.881(2) 3.868(7) 7.791(4) 2 Pmmm 87
x ) 0.09 y ) 0.5 3.881(2) 3.868(7) 7.791(4) 2 Pmmm 87
x ) 0.09 y ) 2/3 3.882(0) 3.867(1) 7.791(4) 2 Pmmm 87
x ) 0.09 y ) 1 3.887(6) 3.887(6) 7.746(5) 2 P4/mmm 87
a R′ indicates slightly different polymorph to R-phase.33 β indicates low-temperature polymorph.33 A indicates high-temperature

polymorph.33 C indicates orthorhombic distortion of the R′-phase.33

3 and 4. The electrical conductivity data of LLT and 7.4.1. Influence of A, B, or A and B Substitution on
several A- or B-site substituted LLT related compounds Structure and Electrical Conductivity. Taking Li3x-
are listed in Table 2. La(2/3)-x0(1/3)-2xTiO3 as reference composition, the sys-
3984 Chem. Mater., Vol. 15, No. 21, 2003 Reviews

Table 4. Crystallographic Data for Both A- and B- or B-Site Substituted Li3xLa(2/3)-x0(1/3)-2xTiO3 (LLT)a′
lattice constant (Å)
chemical formula a b c Z space group/crystal system reference
(Li0.1La0.63)(Mg0.5W0.5)O3 7.8075(20) 7.8349(20) 2 × 7.9145(15) 16 orthorhombic 15a
Li0.20La0.70Ti0.70Mn0.30O3 3.8803(4) 1 cubic 58
Li0.20La0.767Ti0.50Mn0.50O3 5.524(5) 5.479(5) 7.778(7) 4 orthorhombic 58
Li0.20La0.667Ti0.80Mn0.20O3 3.871(4) 7.763(9) 2 tetragonal 58
Li0.33La0.56M(IV)xTi1-xO3
M(IV) ) Zr x) 0.05 3.881 1 cubic 70
M(IV) ) Zr x) 0.10 3.887 1 cubic 70
M(IV) ) Hf x) 0.05 3.882 1 cubic 70
M(IV) ) Hf x) 0.10 3.887 1 cubic 70
Li0.36La0.58Ti0.95Mg0.05O3 3.884(7) 1 cubic 78
Li0.36La0.58Ti0.95Al0.05O3 3.870(4) 1 cubic 78
Li0.36La0.58Ti0.95Mn0.05O3 3.866(7) 1 cubic 78
Li0.36La0.58Ti0.95Ge0.05O3 3.872(2) 1 cubic 78
Li0.36La0.58Ti0.95Ru0.05O3 3.877(4) 1 cubic 78
Li0.36La0.58Ti0.95W0.05O3 3.877(5) 1 cubic 78
Li0.36La0.58Ti0.90W0.10O3 3.884(4) 7.770(1) 2 tetragonal 78
Li0.36La0.58Ti0.995Al0.005O3 3.870(1) 1 cubic 78
Li0.36La0.58Ti0.992Al0.008O3 3.868(1) 1 cubic 78
Li0.36La0.58Ti0.995W0.005O3 3.880(2) 1 cubic 78
Li0.36La0.58Ti0.995Ta0.005O3 3.869(2) 1 cubic 78
Li0.5-3xLa0.5+x+yTi1-3yCr3yO3
polymorph A 3.8775(7) 1 cubic 72
polymorph C 5.497(6) 5.476(3) 7.758(8) 4 orthorhombic 72
polymorph β 3.873(1) 7.761(2) 2 tetragonal 72
Li0.14La2/3Ti0.86Fe0.14O3 3.8769(1) 3.8904(2) 7.7821(3) 2 Pmmm 83
Li0.25La2/3Ti0.75Fe0.25 O3 3.8935(5) 7.7898(22) 2 P4/mmm 83
LixLa2/3Ti1-xAlxO3
x ) 0.06 5.479 5.471 7.772 4 orthorhombic 91
x ) 0.1 5.478 5.469 7.761 4 orthorhombic 91
x ) 0.15 5.469 5.466 7.729 4 orthorhombic 91
x ) 0.2 5.465 5.463 7.729 4 orthorhombic 91
x ) 0.25 5.457 5.454 7.717 4 orthorhombic 91
x ) 0.30 5.454 5.450 7.715 4 orthorhombic 91
Li0.20Pr0.667Ti0.80Cr0.20O3 3.8463(9) 1 cubic 60
Li0.05Pr0.917Ti0.80Cr0.20O3 5.4569(8) 7.711(2) 5.446(1) 4 orthorhombic 60
Li0.15Pr0.75Ti0.60Cr0.40O3 3.849(1) 7.705(2) 2 tetragonal 60
La(1/3)-xLi3xNbO3
x)0 5.535(5) 5.620(9) 7.851(5) 4 orthorhombic 137
x ) 0.01 5.532(6) 5.62(1) 7.844(8) 4 orthorhombic 137
x ) 0.02 5.532(5) 5.613(9) 7.848(6) 4 orthorhombic 137
x ) 0.03 5.529(4) 5.603(9) 7.837(5) 4 orthorhombic 137
x ) 0.04 5.525(6) 5.59(1) 7.837(7) 4 orthorhombic 137
x ) 0.05 5.516(7) 5.58(1) 7.833(4) 4 orthorhombic 137
x ) 0.06 5.523(4) 5.579(7) 7.829(5) 4 orthorhombic 137
Li0.25Nd0.25TaO3 3.8720 7.7300 2 tetragonal 136b
Li0.25Sm0.25TaO3 3.8590 7.7100 2 tetragonal 136b
Li0.25Y0.25TaO3 3.8220 3.8422 7.6540 4 orthorhombic 136b
Li0.33(Ca1-xSrx)0.56Ta0.56Ti0.44O3
x)0 5.487 5.381 3.857 2 orthorhombic 70
x ) 0.3 3.860 1 cubic 70
x ) 0.5 3.884 1 cubic 70
x ) 0.8 3.912 1 cubic 70
x ) 1.0 3.938 1 cubic 70
Li0.33(Ca1-xSrx)0.56Fe0.225Ta0.775O3
x)0 5.525 5.392 3.860 2 orthorhombic 70
x ) 0.2 5.531 5.434 3.870 2 orthorhombic 70
x ) 0.5 3.903 1 cubic 70
x ) 0.8 3.935 1 cubic 70
x ) 0.9 3.952 1 cubic 70
x ) 1.0 3.961 1 cubic 70
Li0.33Sr0.56M(III)0.225Ta0.775O3
M(III) ) Cr 3.948 1 cubic 70
M(III) ) Co 3.978 1 cubic 70
M(III) ) Ga 3.954 1 cubic 70
M(III) ) Y 4.000 4.014 1 tetragonal 70
LiCa1.65Ti1.3Nb1.7O9 5.363(1) 5.464(1) 7.66283) 4 orthorhombic 77b
LiCa1.65Ti1.3Ta1.7O9 5.363(1) 5.456(1) 7.661(1) 4 orthorhombic 77b
LiSr1.65Ti2.15W0.85O9 3.911(1) 1 cubic 77b
LiSr1.65Ti1.3Nb1.7O9 3.932(1) 1 cubic 138a
LiSr1.65Ti1.3Ta1.7O9 3.932(1) 1 cubic 138a
LiSr1.65Zr1.3Ta1.7O9 4.017(1) 1 cubic 138a
Li0.1Sr0.8Ti0.7Nb0.3O3 3.916(1) 1 cubic 138b
Li0.3Sr0.6Ti0.5Nb0.5O3 3.925(1) 1 cubic 138b
Li0.3Sr0.6Ti0.5Ta0.5O3 3.926(1) 1 cubic 138b
Li0.3Sr0.6Ti0.45Fe0.05 Ta0.5O3 3.933(1) 1 cubic 138b
Li0.3Sr0.6Ti0.40Fe0.10 Ta0.5O3 3.942(2) 1 cubic 138b
Li0.3Sr0.6Ti0.35Fe0.15Ta0.5O3 3.937(2) 1 cubic 138b
Li0.3Sr0.6Ti0.20Fe0.30Ta0.5O3 3.944(1) 1 cubic 138b
a A, C, and β: see Table 3.
Reviews Chem. Mater., Vol. 15, No. 21, 2003 3985

Figure 13. Variation of the ionic conductivity and the

activation energy as a function of a perovskite lattice param-
eter for Li0.34La0.51TiO2.94 (27 °C), Li0.34Pr0.51TiO3.01 (27 °C), Li0.34-
Nd0.55TiO3 (27 °C), and Li0.38Sm0.52TiO2.97 (127 °C).16 La com-
pound shows the highest ionic conductivity with the lowest Figure 14. Arrhenius plots for electrical conductivity of (a)
activation energy. The broken lines passing through the data Li0.34La0.51TiO2.94,16 (b) Li0.31La0.63Ti0.9Co0.1O3 (0),26,96 (c) Li0.31-
points are guides to the eye. La0.63Ti0.9Ni0.1O3 (O),26,96 (d) Li0.1La0.63Mg0.5W0.5O3,15a (e) Li0.34-
Nd0.55TiO3 (‚),16 (f) Li0.1La0.3NbO3,92 (g) Li0.25La0.25TaO3
(- ‚ -),136b (h) Li0.3Sr0.6Nb0.5Ti0.5O3 (∆),138b (i) Li0.3Sr0.6Ta0.5-
tems studied in the literature can be classified as Ti0.5O3 (b),138b and (j) Li0.5Sr0.5Fe0.25Ta0.75O3.70a
follows: (i) partial or complete substitution (Table 3) of
other lanthanides (Pr, Nd, Sm, Gd, Dy, Y) for A-site bottleneck size, which can be modified by appropriate
La;13,16,29,31,33,36,37,43,45,55,70,71,131,132 (ii) partial or complete chemical substitution.
substitution (Table 3) of other alkali ions Na, K, etc. or In the third case, it has been observed that the ionic
alkaline earth ions (Sr, Ba) or Ag for A-site size and amount of the substituent for Ti and the
Li;11,15,19,20,27,30,55,74,87,129,133,134 (iii) partial or complete number of vacancies have a strong influence on the
substitution (Table 4) of tri-, tetra-, penta-, and hexava- crystal structure and the conductivity of these materials.
lent ions for B-site Ti;15,39,58,62,63,70,72,78,83,91,92,132,135-137 Larger cations and few vacancies stabilize the cubic
and (iv) contemporaneous complete substitution (Table symmetry, which indeed has been observed for Co and
4) of other lanthanides or alkaline earths for A-site La Ni substitution for Ti.26,96 Interestingly, by keeping the
and partial or complete substitution of transition metals amount of Co constant and varying the amount of
for B-site Ti.18,60,70,71,77,88,93,138 vacancies, the symmetry went down to tetragonal with
lattice parameters and space group similar to those
In the first case, complete substitution of the La3+ with proposed by Fourquet et al. for LLT.42,96 Partial Co, Ni,
smaller Ln3+ leads to an orthorhombic distortion of the and Cu substitution do not improve the ionic conductiv-
lattice, and the ionic conductivity is found to decrease ity of LLT (Figure 14). It is interesting to point out that
accompanied by an increase in activation energy 16,31a it is not the type of substituent but the amount of
with decreasing ionic radius of the rare earth ion (Figure vacancies that has the major influence on the conduct-
13). Pr systems were found to show a smaller variation ing behavior. Similar to LLT, doped LLT also exhibits
of the conductivity with composition compared to the dome-shaped dependence of the conductivity on the
La compounds.43 Systems containing two different rare composition.26,91,96
earth elements (La + Nd or La + Gd) showed two more Partial B-site substitution with Cu, Mo, and Fe for
contributions: space charge polarization and Warburg Ti induced an orthorhombic distortion.83,96 Partial B-site
impedance at low frequencies.45 Studies by ND have substitution with Mn and Cr (Li0.5-3xLa0.5+x+yTi1-3yB3yO3)
shown that the Pr and Nd compounds show a structure showed two types of behavior depending on the amount
related to compounds such as GdFeO3, La1-xTiO3,104 and of substituent: (i) mixed conduction with greater ionic
(Ca,Zr)1-x(LiTa)xO3.139 than electronic conductivity for 0 < y e 0.07 (Mn) and
In the second case, partial A-site substitution of La 0 < y e 0.133 (Cr) and (ii) mixed conduction with greater
by monovalent alkali ions (Na, K) and divalent alkaline electronic than ionic conductivity for y > 0.3 (Mn) and
earth ion (Ba) did not improve much the conductivity. y > 0.133 (Cr). dc measurements showed that the
Indeed, Na substitution decreases the available space electronic conductivity depends on the amount of Mn
for Li+ ion motion, depressing the ionic conductivity, and or Cr content in LLT.58,72
Ba substitution expands the lattice but induces local A small amount of Al substitution for Ti caused a very
deformations, which decrease the ionic conductivity.15,55 small increase in the ionic conductivity, suggesting that
Substitution of Ag leads to a distortion of the bottlenecks the change in interatomic bond strengths (both A-O
that is favorable for the conduction mechanism because and B-O bonds) is one of the factors influencing the
it decreases the activation energy.87 Doping Sr up to 5% lithium ion mobility.78 Higher amounts of Al substitu-
increases the lithium ion conduction. 12,15,19,20,52,57,70,74 tion cause a decrease in the bulk ionic conductivity by
It stabilizes the cubic structure and disturbs the order- 2 orders of magnitude; for example, Li0.25La0.66Ti0.75-
ing of the Li0.3La0.57TiO3 perovskite.74 The results Al0.25O3 exhibits σbulk ) 7.6 × 10-5 S/cm at room
indicate that ionic conductivity strongly depends on the temperature.91 It is expected that substitution of higher
3986 Chem. Mater., Vol. 15, No. 21, 2003 Reviews

Table 5. Room Temperature Electrical Conductivity and at the moment, the exact dimensionality (2D or 3D) of
Activation Energy of Selected Perovskite-Type Materials the lithium mobility in LLT is still controversial.
without La; For Comparison, Data for Li0.34La0.51TiO2.94
Are Included
Evidence of both of them is given from different experi-
ments.
σRT Ea
compound (S/cm) (eV) reference 7.5.1. Structural Considerations: Vacancy, Bottleneck,
and TiO6 Octahedra Tilting. Most of the authors believe
Li0.34Pr0.56TiO3 10-6 0.47 16
Sm0.52Li0.34TiO3 <10-7 0.64 16 that the high ionic conductivity is due to the presence
Nd0.55Li0.34TiO3 <10-7 0.53 16 of A-site vacancies. 11,15-17,19,32,38,55,59,74,87 The exact
Nd0.25Li0.25TaO3 <10-9 0.60 136b location of Li in LLT is still not clear and many
Sm0.25Li0.25TaO3 <10-9 0.65 136b controversial results have been reported. It has been
Y0.25Li0.25TaO3 <1010 0.85 136b
Li0.5Sr0.56Fe0.25Ta0.75O3 1.0 × 10-4 0.36 70b observed by means of ND study that, depending upon
Li0.5Sr0.56Cr0.25Ta0.75O3 6.0 × 10-5 0.39 70b the composition, the structure changes its crystal sym-
Li0.33Sr0.56Cr0.225Ta0.775O3 1.0 × 10-4 70a metry and has different Li-oxygen coordination num-
Li0.33Sr0.56Co0.225Ta0.775O3 5.1 × 10-6 0.41 70a bers. For example, the Li ions are located at the center
Li0.33Sr0.56Ga0.225Ta0.775O3 7.7 × 10-6 70a
Li0.33Ca0.56Fe0.225Ta0.775O3 1.5 × 10-6 70a
of the A site (12-fold coordinated),42,69 the off-center
Li0.33(Ca0.8Sr0.2)0.56Fe0.225Ta0.775O3 5.8 × 10-7 70a position (4-fold coordinated),97 and the bottleneck posi-
Li0.33(Ca0.5Sr0.5)0.56Fe0.225Ta0.775O3 4.1 × 10-5 70a tion (4-fold coordinated)75 for Li0.32La0.56TiO3, Li0.16La0.62-
Li0.33(Ca0.2Sr0.8)0.56Fe0.225Ta0.775O3 9.8 × 10-5 70a TiO3, and Li0.5La0.5TiO3, respectively. The highest ionic
Li0.33(Ca0.1Sr0.9)0.56Fe0.225Ta0.775O3 1.3 × 10-4 70a conductivity has been observed for the Li-rich com-
Li0.33Sr0.56Fe0.225Ta0.775O3 8.5 × 10-5 70a
LiCaTiNbO6 <10-7 0.68 138a pounds (x ≈ 0.10), that is, the compounds (cubic or
LiSrTiNbO6 ∼10-6 0.42 138a tetragonal perovskites) with the Li ion at the center of
LiSrTiTaO6 5.5 × 10-4 0.33 138a the A sites.
LiSr2Ti2NbO9 <10-6 0.37 138a
LiBa2Ti2NbO9 <10-7 0.74 138a The substitution of different cations for La induces
LiSr2Ti2TaO9 3.2 × 10-5 0.27 138a deviations in conductivity and activation energy, which
LiCa1.65Ti1.3Nb1.7O9 <10-7 0.71 77b have been attributed to structural randomness along
LiCa1.65Ti1.3Ta1.7O9 <10-7 0.68 77b the conduction paths caused by the tilting and rotating
LiSr1.65Ti2.15W0.85O9 ∼10-6 0.49 77b
LiSr1.65Ti1.3Nb1.7O9 2.0 × 10-5 0.34 138a
of the TiO6 octahedra, disturbing the Li+ ion migra-
LiSr1.65Ti1.3Ta1.7O9 4.9 × 10-5 0.35 138a tion.16 Moderate lattice expansion accompanied by A-
LiSr1.65Zr1.3Ta1.7O9 1.3 × 10-5 0.36 138a site substitution would enhance the conductivity, as
Li0.1Sr0.8Ti0.7Nb0.3O3 ,10-7 0.31 138b observed for the substitution of up to 5% Sr for La,
Li0.3Sr0.6Ti0.5Nb0.5O3 5.4 × 10-6 0.31 138b which was found to increase the bulk lithium ion
Li0.3Sr0.6Ti0.5Ta0.5O3 1.7 × 10-4 0.27 138b
Li0.3Sr0.6Ti0.45Fe0.05 Ta0.5O3 6.0 × 10-5 0.36 138b conductivity (Table 2).15a,19,55,74,87 However, substitution
Li0.3Sr0.6Ti0.40Fe0.10 Ta0.5O3 3.6 × 10-5 0.36 138b of a higher amount of Sr (>5%) for La decreases the
Li0.3Sr0.6Ti0.35Fe0.15Ta0.5O3 2.8 × 10-5 0.39 138b ionic conductivity.15 A similar effect was observed for
Li0.3Sr0.6Ti0.20Fe0.30Ta0.5O3 <10-7 138b the substitution of Ba for La,19 which may be due to
Li0.34La0.51TiO2.94 1.0 × 10-3 0.40 11
local lattice deformation caused by narrowing the
bottleneck size (see section 7.3).142,143 These observations
valence B-site ions (Nb/Ta for Ti increases the B-O
are also applicable to other perovskite-type titanates
bond strength, which in turn weakens the A-O bond
with the general formula M3xRE2/3-xTiO3 (where RE )
in ABO3 perovskites.62 However, the substitution of Nb
rare earth element; M ) Li, Na, K, Ag).55,87 A compara-
or Ta for Ti and K for La also does not increase the
tive study on the pressure effect and the RE cation
conductivity.92 Complete substitution of the Ti by
substitution on the ionic conductivity of the LLT showed
(Mg0.5W0.5) in LLT, yields a new compound, isostructural
that the former (lattice expansion) is much smaller than
to the orthorhombic perovskite-type La2/3(Mg1/2W1/2)O3,140
the latter (lattice deformation) on their ionic conductiv-
which exhibits an ionic conductivity 1 order of magni-
ity behavior.17 The results have been expressed in terms
tude lower compared to that of the undoped LLT (x )
of activation volume.144
0.11).15
In the last case, contemporaneous complete substitu- In the cubic (disordered) and tetragonal perovskite-
tion of the La with Pr and partial B-site substitution like structure, the Li+ ion in the A site is surrounded
with Cr showed mixed conduction with greater ionic by 12 oxygen ions and can migrate to the vacancy in
than electronic conductivity for 0 < y e 0.10 (y ) Cr) the adjacent A site through the bottleneck surrounded
and mixed conduction with greater electronic than ionic by four oxygen ions (Figure 15).17,138a,145 The positive
conductivity for y > 0.10.60 Further studies on doped activation volume (∆V) value of 1.6-1.7 cm3/mol (smaller
LLT (both A and B site) pointed out an anti-correlation than that of Sodium Super Ionic Conductor, NASICON,
between the Ea and the bulk conductivity similar to that 2.8 cm3/mol146)17 implies that a dilatation of the bottle-
observed in Na-β-alumina.70,141 It must be mentioned neck occurs when a Li+ ion jumps from A site to
that complete replacement of La by other rare earths vacancy. It must be observed that due to the tilting of
or alkaline earths and partial or complete substitution TiO6 octahedra, the bottleneck is not constant all over
of Ti by other transition metal ions do not increase the the crystal lattice.
ionic conductivity (Table 5 and Figure 14). It has been suggested that d0 cations such as Ti(IV),
7.5. Mechanism of Li+ Ion Conduction. Several Zr(IV), Nb(V), and Ta(V) are the best choice for B cations
conduction mechanisms have been proposed for LLT on in lithium ion conducting perovskite-type oxides138a
the basis of structural considerations, conductivity because d0 cations are known to distort BO6 octahedra,
measurements, and theoretical models, to justify its giving rise to short and long B-O bonds.147 A dynamic
high ionic conductivity at room temperature. Anyway, fluctuation of B-O bond lengths caused by a distorted
Reviews Chem. Mater., Vol. 15, No. 21, 2003 3987

The highest ionic conduction in Li3xLa(2/3)-x0(1/3)-2xTiO3

was obtained when the total concentration of lithium
and A-site ion vacancy n′(0.33 + x) is approximately
equal to 0.44-0.45. This value is higher than the site
percolation threshold, nc, value of 0.3117 for simple
cubic lattice. This result suggests the presence of at least
one vacancy/Li+ adjacent to every Li+/vacancy. The
difference between the theory and experimental value
is considered to be due to the local distortion. Such a
distortion decreases the bottleneck and hence impedes
the lithium ion jump.19 Accordingly, the effective carrier
concentration, neff, has been proposed to explain the
conductivity behavior of LLT at various temperatures
and is defined as20

neff ∝ m(1 - m) (6)

where m is the ratio of lithium to the sum of lithium

and vacancy, N ) NLi + NV. The conductivity can be
described, using the percolation theory, by the equa-
Figure 15. Schematic structure of Li3xLa(2/3)-x0(1/3)-2xTiO3 tion20
showing the bottleneck for the Li+ ion migration. The Li, La,
and vacancy are distributed at A sites.138a
3x(0.33 - 2x)
σ) {(0.33 + x) - nc)} (7)
distribution of A-site species in the perovskite structure (0.33 + x)2
seems to be more favorable for facile conduction than a
frozen distribution of bond lengths.138a Since the crystal The conductivity data measured at 227 °C is similar to
structure and conductivity changes with Li-concentra- that calculated using eq 7, suggesting that ionic con-
tion (see section, 7.5.2.), there is a lack of a single ductivity depends on both site percolation and on the
structural model to explain the conductivity behavior ratio of lithium to vacancy concentration.20
over the wide range of compositions in LLT. 7.5.3. Theoretical Modeling. Recently, theoretical
7.5.2. Effect of Carrier Concentration and Site Perco- modeling of the Li-ion motion for the Li-poor composi-
lation. For single-phase LLT the ionic conductivity is tion Li0.2La0.6TiO3 (low conducting member of LLT) has
given by been carried out by means of molecular dynamics
(MD)148 and according to the concept of bond valence
σ ) ZeCµ (4) summation (BVS).94 The MD employing the partially
ionic model (PIM)90,148 and fully ionic model (FIM)90
where C is the concentration of the charge carriers (Li) could simultaneously reproduce the crystal structure
with charge Ze (Z ) 1 for Li+) and mobility µ. The and its physical properties such as thermal expansion,
mobility mainly depends on the activation energy for compressibility, and Li-ion migration. The properties,
ionic conduction. Since the activation energy for ionic for example, density and thermal expansion coefficients
conduction is nearly constant at ambient temperature obtained using PIM, were found to be close to experi-
(0.35 eV), the mobility of lithium is assumed to be mental data, while the FIM model found too poor
constant in the wide range of composition 0.06 < x < results. Both simulation and experimental data support
0.15. The charge carrier concentration involves the the displacement of Li in the 〈100〉 direction from the
lithium (NLi) and A-site ion vacancy (Nv) concentra- center of the A site for Li0.2La0.6TiO3. These simulations
tions.32 It has been assumed that all the lithium ions at high pressure as a function of the repulsion param-
(NLi ) 3x/Vs) in the LLT can move independently of each eter of the Li ion, aLi, reveal that the Li-ion conductivity
other through the A-site vacancies (Nv ) (0.33 - 2x)/ decreases with the lattice contraction. The ionic con-
Vs) for conduction. All the available A sites are presum- ductivity obtained by the MD was found to be smaller
ably energetically and symmetrically equal, so one can than the experimental values in the temperature range
write N ) NLi + NV. Accordingly, the lithium ion 27-727 °C. For example, at 27 and 127 °C, the conduc-
conductivity of LLT can be expressed as32 tivity values from MD are 8.67 × 10-4 and 6.56 × 10-4
S/cm, while the corresponding experimental values are
1.9 × 10-4 and 4.3 × 10-3, respectively.90 The chemical
(x - 6x2)
σ)e µ (5) diffusion coefficient was also found to be slightly lower
(0.33 + x)Vs than the experimental value of Li0.29+δLa0.57TiO3, de-
termined by the GITT method.50
where Vs is the perovskite subcell volume. Reporting Mazza et al.94 employed the BVS method, which
the value of C as a function of x, we obtain a dome- allowed tracing to a time-averaged pathway followed by
shaped curve for conductivity with a maximum at x ) the Li+ ion during long-range motion in the bulk of the
0.075. Interestingly, the measured lithium ion conduc- structure. By inspection of the saddle point of such
tivity as a function of x also shows the same shape32 pathway, it is possible to estimate the activation energy
with a maximum at x ≈ 0.1. Thus, the ionic conductivity for the ionic jump. From this calculation, the bulk
of LLT does not support the hypothesis of the ordering conductivity could be predicted on the basis of the
of the A-site vacancies and cooperative ionic motion. following three hypotheses for Li+ ions location in the
3988 Chem. Mater., Vol. 15, No. 21, 2003 Reviews

LLT structure: (i) the Li+ ions and vacancies are equally
distributed over the La1/La2 sites,42 (ii) all Li+ ions are
at the La2 sites, and (iii) the La2 site is completely
occupied by La and Li. It has been found that the third
possibility is clearly conflicting with the experimental
data, while the first, redistribution of Li ion and
vacancies with the same ratio on La1 and La2 sites,
gives high conductivity and is comparable to the ex-
perimental values. The second possibility also produces
similar results. Thus, it has been suggested that both
may contribute to the conduction in LLT.
7.5.4. Analysis of NMR and ECR Data. ECR and
NMR methods allowed separation of long- and short-
range Li motion with distinct Ea. 44,47,51,61,67,73,76 A
dispersive behavior of the conductivity, measured by
ECR methods, could be interpreted both in terms of ion
hopping (with correlated ion motion)149 or of a Kohl-
rausch-Williams-Watts (KWW)112 correlation function
Φ(t) in the time domain.150 NMR spin lattice relaxation
results, according to Ngai’s coupling model, could be
evaluated giving two activation energies: Es for long-
range motion and a microscopic Ea free of the effect of
cooperativity. On the basis of a thermally activated
relaxation mechanism, Es and Ea are related through
the β exponent by

Ea ) βsEs151-153 (8)

However, it has to be pointed out that such relation

between the activation energies holds only if both of
them are calculated in the same temperature and
frequency range.47,154
Ibarra et al.73 studied the influence of composition on
conductivity of these solid solutions with 0.03 e x e
0.167. In the NMR study of orthorhombic perovskites
(a ) 3.864 Å, b ) 3.875 Å, c ) 7.786 Å), two types of Li,
with different mobility, have been ascribed to Li located
in two planes of the perovskite.76 The accommodation Figure 16. Schematic representation of 2D and 3D mobility
of cation vacancies in alternating planes along the c-axis of lithium ions in the perovskite structure.86
should favor a 2D movement of Li in orthorhombic
samples.59,76 However, in tetragonal samples (a ) b ) at 27 °C) than the pure LLT, while complete substitu-
3.87 Å; c ) 2a), cation and vacancies turn progressively tion of other transition metal ions for Ti decreases the
disordered, favoring a 3D movement of Li ions at higher ionic conductivity. Complete replacement of La by other
temperature and at temperature below 200 K lithium rare earth and other metal ions (alkali or alkaline earth
ion hops between cages through the bottleneck in the ion) were also found to decrease the ionic conductivity.
ab plane (2D motion).81,86,95 Figure 16 shows the sche- (iii) The optimum total lithium and vacancy concentra-
matic representation of 2D and 3D lithium ion mobility tion for higher ionic conductivity is found to be in the
in LLT. This also explains the high activation energy range 0.44-0.45. Lithium ion conductivity is due to the
for the low-temperature regime and low activation motion of lithium along the A-site vacancies, the window
energy for the high-temperature region. The increase of four oxygens separating adjacent A sites constituting
of Li mobility along the series is responsible for the the bottleneck for lithium ion migration. (iv) Depending
progressive line narrowing of the central transition on the synthesis conditions and composition, LLT
detected by 7Li NMR spectra. crystallizes in simple cubic (a ) 3.87 Å), tetragonal
(a ) b ) 3.87 Å; c ) 2a), or orthorhombic (a ) 3.864 Å,
8. Conclusions b ) 3.875 Å, c ) 7.786 Å) perovskite-type structure.
In summary, (i) the high lithium ionic conducting Cubic structure exhibits slightly higher conductivity
phase is the A-site defect perovskite-type Li3x- than the ordered tetragonal structure for the same bulk
La(2/3)-x0(1/3)-2xTiO3 (0.06 < x < 0.14), the x ≈ 0.10 composition. The low ionic conductivity in ordered phase
member exhibits the highest bulk conductivity of 1 × is due to the unequal ordering of Li, La, and vacancies
10-3 S/cm at 27 °C with an activation energy of ∼0.40 along the c-axis. (v) NMR study reveals that at low
eV. (ii) The ionic conductivity depends mainly on the temperature (T > RT) the lithium ion hops between
size of the A-site cation, lithium ion, and vacancy cages through the bottleneck in the ab plane, and at
concentration. LLT substituted with 5 mol % Sr for La high temperature, the lithium ions hop in three direc-
exhibits slightly higher conductivity (1.5 × 10-3 S/cm tions.
Reviews Chem. Mater., Vol. 15, No. 21, 2003 3989

Although LLT exhibits the highest ionic conductivity (26) Stramare, S.; Weppner, W. Ionics 1999, 5, 405.
(27) (a) Bhalla, A. S.; Guo, R.; Roy, R. Mater. Res. Innovat. 2000, 4,
at room temperature, its use as an electrolyte is not 3. (b) Brous, J.; Fankuchen, I.; Banks, E. Acta Crystallogr. 1953,
favorable because LLT is not stable in direct contact 6, 67.
with elemental lithium and undergoes easy and fast Li (28) Patil, P. V.; Chincholkar, V. S. Curr. Sci. 1970, 39, 348.
(29) Patil, P. V.; Chincholkar, V. S. Indian J. Chem. 1978, 16, 161.
insertion with consequent reduction of Ti4+ to Ti3+, (30) Varaprasad, A. M.; Shashi Mohan, A. L.; Chakrabarty, D. K.;
leading to high electronic conductivity. Furthermore, the Biswas, A. B. J. Phys. C: Solid State Phys. 1979, 12, 465.
use as a lithium battery electrode is also limited because (31) (a) Kochergina, L. L.; Khakhin, N. B.; Porotnikov, N. V.; Petrov,
K. I. Russ. J. Inorg. Chem. 1984, 29, 506. (b) Magneli, A. Arkiv.
of very small lithium uptake and its practical capacity Kemi 1949, 1, 213; 1949, 1, 269.
is much smaller. It is worth mentioning that LLT finds (32) Kawai, H.; Kuwano, J. J. Electrochem. Soc. 1994, 141, L78.
(33) Robertson, A. D.; Garcia Martin, S.; Coats, A.; West, A. R. J.
application at room temperature as a separator for Mater. Chem. 1995, 5, 1405.
selective recovery and isotope separation of lithium ion (34) Shan, Y. J.; Chen, L.; Inaguma, Y.; Itoh, M.; Nakamura, T. J.
by electrolysis21 and as pH sensors.155 In contrast to Power Sources 1995, 54, 397.
(35) Várez, A.; Garcia-Alvarado, F.; Morán, E.; Alario-Franco, M. A.
oxide ion conductivity in ABO3 perovskite oxides, the J. Solid State Chem. 1995, 118, 78.
development of fast cation conductors, for example, Li+ (36) Kawai, H.; Kuwano, J. Ceramics: Charting the Future, Advances
and Na+ ion in the perovskite structure with appropri- in Science and Technology, Part 3, Proceedings of the 8th Cimtec
World Ceramic Congress, Florence, Italy; Vincenzini, P., Ed.;
ate properties for possible application in all-solid-state Techna: Faenza, Italy, 1995; p 2641.
batteries still remains a challenge. (37) Skakle, J. M. S.; Mather, G. C.; Morales, M.; Smith, R. I.; West,
A. R. J. Mater. Chem. 1995, 5, 1807.
(38) Shan, Y. J.; Inaguma, Y.; Itoh, M. Solid State Ionics 1995, 79,
Acknowledgment. Dr. V. Thangadurai would like 245.
to thank the Deutscher Akademischer Austausch Dienst (39) Belous, A. G. Solid State Ionics 1996, 90, 193.
(40) León, C.; Lucı́a, M. L.; Santamarı́a, J.; Paris, M. A.; Sanz, J.;
(DAAD), Bonn, Germany, for financial support. Várez, A. Phys. Rev. B 1996, 54, 184.
(41) Nairn, K. M.; Forsyth, M.; Greville, M.; MacFarlane, D. R.;
Smith, M. E. Solid State Ionics 1996, 86-88, 1397.
References (42) Fourquet, J. L.; Duroy, H.; Crosnier-Lopez, M. P. J. Solid State
Chem. 1996, 127, 283.
(1) Robertson, A. D.; West, A. R.; Ritchie, A. G. Solid State Ionics (43) Morales, M.; West, A. R. Solid State Ionics 1996, 91, 33.
1997, 104, 1. (44) Bohnke, O.; Bohnke, C.; Fourquet, J. L. Solid State Ionics 1996,
(2) (a) Adachi, G.-Ya; Imanaka, N.; Aono, H. Adv. Mater. 1996, 8, 91, 21.
127. (b) Gopalakrishnan, J.; Shukla, A. K.; Thangadurai, V. Curr. (45) Lee, J.-S.; Yoo, K. S.; Kim, T. S.; Jung, H. J. Solid State Ionics
Sci. 1999, 76, 1473. 1997, 98, 15.
(3) Kvist, A.; Lunden, A. Z. Naturforsch. 1965, 20a, 235. (46) Scharner, S. Untersuchungen zur Struktur, Elektrochemie und
(4) West, A. R. J. Appl. Electrochem. 1973, 3, 327. Farbe von substituirten Lithiumtitanaten fuer moegliche An-
(5) (a) Hong, H. Y.-P. Mater. Res. Bull. 1978, 13, 117. (b) Bruce, P. wendungen in der Elektrochromie Ph.D. Thesis, Technische
G.; West, A. R. Mater. Res. Bull. 1980, 15, 117. Fakultaet der Christian-Albrechts-Universität zu Kiel, Ger-
(6) (a) Kuwano, J.; West, A. R. Mater. Res. Bull. 1980, 15, 1661. (b) many, November 1997.
Alpen, V.u.; Bell, M. F.; Wichelhaus, W. Electrochim. Acta 1978, (47) León, C.; Santamarı́a, M. A.; Paris, M. A.; Sanz, J.; Ibarra, J.;
23, 1395. (c) Rodger, A. R.; Kuwano, J.; West A. R. Solid State Torres, L. M. Phys. Rev. B 1997, 56, 5302.
Ionics 1985, 15, 185. (48) Inaguma, Y.; Yu, J.; Katsumata, T.; Itoh, M. J. Ceram. Soc.
(7) Boukamp, B. A.; Huggins, R. A. Phys. Lett. A 1976, 56, 231. Jpn.: Int. Ed. 1997, 105, 597.
(8) Alpen, U.v.; Rabenau, A.; Talat, G. H. Appl. Phys. Lett. 1977, (49) Kitaoka, K.; Kozuka, H.; Hashimoto, T.; Yoko, T. J. Mater. Sci.
30, 621. 1997, 32, 2063.
(9) (a) Yao, Y. F. Y.; Kummer, J. T. J. Inorg. Nucl. Chem. 1967, 29, (50) Birke, P.; Scharner, S.; Huggins, R. A.; Weppner, W. J. Elec-
2453. (b) Whittingham, M. S.; Huggins, R. A. Solid State trochem. Soc. 1997, 144, L167.
Chemistry, Mater. Bur. Stand. Spec.; Pub. 364; U.S. GPO: (51) Emery, J.; Buzare, J. Y.; Bohnke, O.; Fourquet, J. L. Solid State
Washington D.C., 1972. Ionics 1997, 99, 41.
(10) Aono, H.; Sugimoto, E.; Sadaoka, Y.; Imanaka, N.; Adachi, G. (52) Harada, Y.; Kuwano, J.; Saito, Y. Abstracts of the 4th IUMRS
J. Electrochem. Soc. 1989, 136, L590. International Conference in Asia, Chiba, Japan; Elsevier: Am-
(11) Inaguma, Y.; Liquan, C.; Itoh, M.; Nakamura, T.; Uchida, T.; sterdam, 1997; p 706.
Ikuta, H.; Wakihara, M. Solid State Commun. 1993, 86, 689. (53) Klingler, M.; Chu, W. F.; Weppner, W. Ionics 1997, 3, 412.
(12) Harada, Y.; Hirakoso, Y.; Kawai, H.; Kuwano, J. Solid State (54) León, C.; Santamarı́a, J.; Paris, M. A.; Sanz, J.; Ibarra, J.; Várez,
Ionics 1999, 121, 245. A. J. Non-Cryst. Solids 1998, 235-237, 753.
(13) Latie, L.; Villeneuve, G.; Conte, D.; Le Flem, G. J. Solid State (55) Belous, A. G. Ionics 1998, 4, 360.
Chem. 1984, 52, 293. (56) Duncan, F. M.; Kirk, C. A.; Skakle, J. M. S. ISIS Experimental
(14) Belous, A. G.; Novitskaya, G. N.; Polyanetskaya, S. V.; Gornikov, Report, Technical Report; RB No 9708; Rutherford Appleton
Yu. I. Inorg. Mater. 1987, 23, 412. Lab., Oxfordshire, U.K., 1998.
(15) (a) Inaguma, Y.; Chen, L.; Itoh, M.; Nakamura, T. Solid State (57) Harada, Y.; Ishigaki, T.; Kawai, H.; Kuwano, J. Solid State Ionics
Ionics 1994, 70/71, 196. (b) Orrantia, A. M.; Martin, S. G.; 1998, 108, 407.
Alario-Franco, M. A. Chem. Mater. 2003, 15, 363. (58) Moreno, L.; Morales, M.; Martinez-Sarrion, M. L. J. Solid State
(16) Itoh, M.; Inaguma, Y.; Jung, W.-H.; Chen, L.; Nakamura, T. Chem. 1998, 140, 377.
Solid State Ionics 1994, 70/71, 203. (59) Ruiz, A. I.; Lopez, M. L.; Veiga, M. L.; Pico, C. Solid State Ionics
(17) Inaguma, Y.; Yu, J.; Shan, Y.-J.; Itoh, M.; Nakamura, T. J. 1998, 112, 291.
Electrochem. Soc. 1995, 142, L8. (60) Morales, M.; Martinez-Sarrion, M. L. J. Mater. Chem. 1998, 8,
(18) Inaguma, Y.; Matsui, Y.; Shan, Y.-J.; Itoh, M.; Nakamura, T. 1583.
Solid State Ionics 1995, 79, 91. (61) Bohnke, O.; Emery, J.; Veron, A.; Fourquet, J. L.; Buzare, J. Y.;
(19) Katsumata, T.; Matsui, Y.; Inaguma, Y.; Itoh, M. Solid State Florian, P.; Massiot, D. Solid State Ionics 1998, 109, 25.
Ionics 1996, 86-88, 165. (62) Chung, H.-T.; Kim, J.-G.; Kim, H.-G. Solid State Ionics 1998,
(20) Inaguma, Y.; Itoh, M. Solid State Ionics 1996, 86-88, 257. 107, 153.
(21) Kunugi, S.; Inaguma, Y.; Itoh, M. Solid State Ionics 1999, 122, (63) Kim, J.-G.; Kim, H.-G.; Chung, H.-T. J. Mater. Sci. Lett. 1999,
35. 18, 493.
(22) Klingler, M. Entwicklung neuartiger elektrochromer 3 Schicht (64) Ngai, K. L.; León, C. Solid State Ionics 1999, 125, 81.
Systeme durch insitu Erzeugung lokaler Inhomogeneitaeten in (65) Ngai, K. L.; León, C. Phys. Rev. B 1999, 60, 9396.
Halbleitern mittels elektrischer Spannungen, Ph.D. Thesis, (66) Hirakoso, Y.; Harada, Y.; Kuwano, J.; Saito, Y.; Ishikawa, Y.;
Technische Fakultaet der Christian-Albrechts-Universität zu Eguchi, T. Key Eng. Mater. 1999, 169-170, 209.
Kiel, Germany, April 1998. (67) Emery, J.; Bohnke, O.; Fourquet, J. L.; Buzare, J. Y.; Florian,
(23) Klingler, M.; Chu, W. F.; Weppner, W. Ionics 1995, 1, 251. P.; Massiot, D. J. Phys.: Condens. Matter 1999, 11, 10401.
(24) Klingler, M.; Chu, W. F.; Weppner, W. In Extended Abstracts of (68) Chung, H.-T.; Cheong, D.-S. Solid State Ionics 1999, 120, 197.
the 11th International Conference on Solid State Ionics, Hono- (69) Ruiz, A. L.; Lopez, M. L.; Veiga, M. L.; Pico, C. J. Solid State
lulu, Hawaii, Nov 1997; Liebert, B. E., Weppner, W., Eds.; Chem. 1999, 148, 329.
International Society for Solid State Ionics: Garmisch-Parten- (70) (a)Watanabe, H.; Kuwano, J. J. Power Sources 1997, 68, 421.
kirchen, Germany, 1997; p 431. (b) Harada, Y.; Watanabe, H.; Kuwano, J.; Saito, Y. J. Power
(25) Klingler, M.; Chu, W. F.; Weppner, W. Ionics 1997, 3, 289. Sources 1999, 81-82, 777.
3990 Chem. Mater., Vol. 15, No. 21, 2003 Reviews

(71) Ishikawa, Y.; Harada, Y.; Kuwano, J. Key Eng. Mater. 2000, (112) (a) Jonscher, A. K. Dielectric Relaxation in Solids; Chelsea
181-182, 171. Dielectric Press: London, 1983; Chapter 5. (b) León, C.; Rivera,
(72) Martinez-Sarrion, M. L.; Mestres, L.; Morales, M.; Herraiz, M. A.; Varez, A.; Sanz, J.; Santamarı́a, J.; Nagi, K. L. Phy. Rev.
J. Solid State Chem. 2000, 155, 280. Lett. 2001, 86, 1279.
(73) Ibarra, J.; Várez, A.; León, C.; Santamarı́a, J.; Torres-Martinez, (113) Kohlrausch, R. Ann. Phys. (Leipzig) 1847, 72, 393.
L. M.; Sanz, J. Solid State Ionics 2000, 134, 219. (114) Jonscher, A. K.; Reau, J. M. J. Mater. Sci. 1978, 13, 563.
(74) Wang, G. X.; Yao, P.; Bradhurst, D. H.; Dou, S. X.; Liu, H. K. J. (115) Bruce, P. G.; West, A. R.; Almomnd, D. P. Solid State Ionics 1982,
Mater. Sci. 2000, 35, 4289. 7, 57.
(75) Alonso, J. A.; Sanz, J.; Santamarı́a, J.; León, C.; Várez, A.; (116) MacDonald, J. R. Impedance Spectroscopy, Emphasising Solid
Fernández-Dı́az, M. T. Angew. Chem., Int. Ed. 2000, 39, 619. Materials and Systems; Wiley-Interscience: New York, 1987;
(76) Parı́s, M. A.; Sanz, J.; León, C.; Santamarı́a, J.; Ibarra, J.; Várez, Chapter 2.
A. Chem. Mater. 2000, 12, 1694. (117) Irvine, J. T. S.; Sinclair, D. C.; West, A. R. Adv. Mater. 1990, 2,
(77) (a) Harada, Y.; Ishikawa, Y.; Kuwano, J.; Saito, Y. Key Eng. 138.
Mater. 2000, 181-182, 175. (b) Sebastian, L.; Shukla, A. K.; (118) (a) Randles, J. E. B. Discuss. Faraday Soc. 1947, 1, 11. (b)
Gopalakrishnan, J. Proc. Indian Acad. Sci. (Chem. Sci.) 2001, Thangadurai, V.; Huggins, R. A.; Weppner, W. J. Power Sources
113, 427. 2002, 108, 64.
(78) Thangadurai, V.; Weppner, W. Ionics 2000, 6, 70. (119) Hladik, J. Physics of the Elements (Russian Translation); Mir:
(79) Mizumoto, K.; Hayashi, S. Solid State Ionics 2000, 127, 241. Moscow, 1978.
(80) Hirakoso, Y.; Harada, Y.; Saito, Y.; Yoshida, Y.; Kuwano, J. Key (120) Kincs, J.; Martin, S. W. Phys. Rev. Lett. 1996, 76, 70.
Eng. Mater. 2000, 181-182, 179. (121) Vogel, H. Phys. Z. 1921, 22, 645.
(81) Emery, J.; Bohnke, O.; Badot, J. C. In 16th International Meeting
(122) Tamman, G.; Hesse, W. Z. Anorg. Allg. Chem. 1926, 156, 245.
Hertzian Optics and Dielectrics, Sept 2001; Silly, G., Buzaré, J.
Y., Eds.; Université du Maine: Le Mans, France, 2001. (123) Fulcher, G. S. J. Am. Ceram. Soc. 1925, 8, 339.
(82) Ban, C. W.; Choi, G. M. Solid State Ionics 2001, 140, 285. (124) Cohen, H.; Turnbull, D. J. Chem. Phys. 1959, 31, 1164.
(83) Crosnier-Lopez, M. P.; Duroy, H.; Calage, Y.; Greneche, J. M.; (125) Ratner, M. A. Polymer Electrolytes Review Part I; Elsevier
Fourquet, J. L. Mater. Res. Bull. 2001, 36, 651. Applied Science, London, 1987.
(84) Chen, C. H.; Amine, K. Solid State Ionics 2001, 144, 51. (126) Shan, Y. J.; Chen, L.; Inaguma, Y.; Itoh, M.; Nakamura, T. Solid
(85) Várez, A.; Sanjuán, M. L.; Laguna, M. A.; Peña, J. I.; Sanz, J.; State Ionics 1994, 70/71, 429.
de la Fuente, G. F. J. Mater. Chem. 2001, 11, 125. (127) Van Gool, W.; Piken, A. G. J. Mater. Sci. 1969, 4, 95.
(86) Bohnke, O.; Badot, J. C.; Emery, J. In 16th International Meeting (128) Ho, C.; Raistrick, I. D.; Huggins, R. A. J. Electrochem. Soc. 1980,
Hertzian Optics and Dielectrics, Sept 2001; Silly, G., Buzaré, J. 127, 343.
Y., Eds.; Université du Maine: Le Mans, France, 2001. (129) Austin, I. G.; Mott, N. F. Adv. Phys. 1969, 18, 41.
(87) Bohnke, O.; Bohnke, C.; Ould Sid’Ahmed, J.; Crosnier-Lopez, (130) (a) Weppner, W.; Huggins, R. A Annu. Rev. Mater. Sci. 1978, 8,
M. P.; Duroy, H.; Le Berre, F.; Fourquet, J. L. Chem. Mater. 269. (b) Weppner, W.; Huggins, R. A. J. Electrochem. Soc. 1977,
2001, 13, 1593. 124, 1569.
(88) Harada, Y.; Kuwano, J.; Saito, Y. Key Eng. Mater. 2002, 214- (131) Takahashi, H.; Baba, Y.; Ezaki, K.; Okamoto, Y.; Shibata, K.;
215, 227. Kuroki, K.; Nakano, S. Jpn. J. Appl. Phys. 1991, 30, 2339.
(89) Bohnke, O.; Duroy, H.; Fourquet, J. L.; Ronchetti, S.; Mazza, D. (132) Inaguma, Y.; Matsui, Y.; Yu, J.; Shan, Y.; Nakamura, T.; Itoh,
Solid State Ionics 2002, 149, 217. M. J. Phys. Chem. Solid 1997, 58, 843.
(90) Katsumata, T.; Inaguma, Y.; Itoh, M.; Kawamura, K. Chem. (133) Inaguma, Y.; Sohn, J.-H.; Kim, I.-S.; Itoh, M.; Nakamura, T. J.
Mater. 2002, 14, 3930. Phys. Soc. Jpn. 1992, 61, 3831.
(91) Morata-Orrantia, A.; Garcı́a-Martı́n, S.; Morán, E.; Alario- (134) Belous, A. G. J. Eur. Ceram. Soc. 2001, 21, 1797.
Franco, M. A. Chem. Mater. 2002, 14, 2871. (135) Nazri, G.-A.; Julien, C. Solid State Batteries: Materials Design
(92) Nakayama, M.; Ikuta, H.; Uchimoto, Y.; Wakihara, M. J. Mater. and Optimization; Kluwer: Boston, 1994; p 146.
Chem. 2002, 12, 1500. (136) (a) Mizumoto, K.; Hayashi, S. J. Ceram. Soc. Jpn. 1997, 105,
(93) Sugiura, K.; Harada, Y.; Kuwano, J. Key Eng. Mater. 2002, 216, 713. (b) Mizumoto, K.; Hayashi, S. Solid State Ionics 1999, 116,
127. 263. (c) Mizumoto, K.; Hayashi, S. Solid State Ionics 2000, 127,
(94) Mazza, D.; Ronchetti, S.; Bohnké, O.; Duroy, H.; Fourquet, J. L. 241.
Solid State Ionics 2002, 149, 81. (137) Garcı́a-Martı́n, S.; Rojo, J. M.; Tsukamoto, H.; Morán, E.; Alario-
(95) Emery, J.; Bohnke, O.; Fourquet, J. L.; Buzaré, J. Y.; Florian, Franco, M. A. Solid State Ionics 1999, 116, 11.
P.; Massiot, D. J. Phys.: Condens. Matter 2002, 14, 523. (138) (a) Thangadurai, V.; Shukla, A. K.; Gopalakrishnan, J. Chem.
(96) Stramare, S. Synthesis and Characterization of Substituted Mater. 1999, 11, 835. (b) Thangadurai, V.; Weppner, W. J.
Lithium Lanthanum Titanates, with Possible Application in Electrochem. Soc., in press.
Electrochromic Systems. Ph.D. Thesis, Technische Fakultaet der (139) Smith, R. I.; West, A. R. J. Solid State Chem. 1994, 108, 29.
Christian-Albrechts-Universität zu Kiel, Germany, April 2002. (140) Yao, Y. Y.; Kummer, J. T. J. Inorg. Nucl. Chem. 1967, 29, 2453.
(97) (a) Inaguma, Y.; Katsumata, T.; Itoh, M.; Morii, Y. J. Solid State (141) Torii, Y. Chem. Lett. 1979, 1215.
Chem. 2002, 166, 67. (b) Sanz, J.; Alonso, J. A.; Varez, A.;
(142) Stauffer, D. Introduction to Percolation Theory, 1st ed.; Taylor
Fernandez-Diaz, M. T. J. Chem. Soc., Dalton. Trans. 2002, (7),
and Francis: London, 1985.
1406.
(98) White, J. J.; Segall, R. L.; Barry, J. C.; Hutchison, J. L. Acta (143) Stauffer, D.; Aharomy, A. Introduction to Percolation Theory,
Crystallogr., Sect. B: Struct. Sci. 1985, 41, 93. 2nd ed.; Taylor and Francis, 1992.
(99) Kay, H. F.; Bailey, P. C. Acta Crystallogr., Sect. A: Found. (144) Samara, G. A. Solid State Physics 1984, 38, 1.
Crystallogr. 1957, 10, 219. (145) Thangadurai, V.; Weppner, W. Ionics 2002, 8, 281.
(100) Bhuvanesh, N. S. P.; Bohnke, O.; Duroy, H.; Crosnier-Lopez, M. (146) Kafalas, J. A.; Cava, R. J. Fast Ion Transport in Solids; Elsevier/
P.; Emery, J.; Fourquet, J. L. Mater. Res. Bull. 2002, 33, 1681. North-Holland: New York, 1979; p 419.
(101) Glazer, A. M. Acta Crystallogr., Sect. B: Struct. Sci. 1972, 28, (147) Bhuvanesh, N. S. P.; Gopalakrishnan, J. J. Mater. Chem. 1997,
3384. 7, 2297.
(102) Abe, M.; Uchino, K. Mater. Res. Bull. 1974, 9, 147. (148) Katsumata, T.; Inaguma, Y.; Itoh, M.; Kawamura, K. J. Ceram.
(103) Shannon, R. D. Acta Crystallogr., Sect. A: Found. Crystallogr. Soc. Jpn.: Int. Ed. 1999, 107, 615.
1976, 32, 751. (149) León, C.; Lucı́a, M. L.; Santamarı́a, J. Phys. Rev. B 1997, 55,
(104) MacEachern, M. J.; Dabkowska, H.; Garret, J. D.; Amow, G.; 882.
Gong, W.; Liu, G., Greedan, J. E. Chem. Mater. 1994, 6, 2092. (150) León, C.; Lucı́a, M. L.; Santamarı́a, J. Philos. Mag. B 1997, 75,
(105) Ngai, K. L. Phys. Rev. B 1993, 48, 13481. 629.
(106) Richards, P. M. Physics of Superionic Conductors, Topics in (151) Ngai, K. L. Comments Solid State Phys. 1979, 9, 121.
Current Physics; Springer-Verlag: Berlin, 1979. (152) Ngai, K. L. Comments Solid State Phys. 1980, 9, 141.
(107) Ngai, K. L.; Rendell, R. W.; Jain, H. Phys. Rev. B 1984, 30, 2133. (153) Ngai, K. L. Effects of Disorder on Relaxational Processes,
(108) Ngai, K. L.; Martin, S. W. Phys. Rev. B 1989, 40, 10550. Springer-Verlag: Berlin, 1994.
(109) Ngai, K. L. Solid State Ionics 1993, 61, 345. (154) Funke, K.; Wilmer, D. Europhys. Lett. 1990, 12, 363.
(110) Walstedt, R. E.; Dupree, R.; Remeika, J. P.; Rodriguez, A. Phys. (155) Bohnke, Cl.; Duroy, H.; Fourquet, J. L. Sensors Actuators B 2003,
Rev. B 1977, 15, 3442. 89, 240.
(111) Kim, K. H.; Torgeson, D. R.; Borsa, F.; Cho, J.; Martin, S. W.;
Svare, I. Solid State Ionics 1996, 91, 7. CM0300516