Microwave synthesis of lithium lanthanum titanate

Article abstract of DOI:10.1016/S0038-1098(02)00852-9

A study was performed on microwave synthesis of lithium lanthanum titanate (LLTO). The method was found to be fast, clean, economical and it gave phase pure product of high conductivity. The microwave method was also shown to be well suited for sintering of LLTO.

Full text of DOI:10.1016/S0038-1098(02)00852-9

Solid State Communications 125 (2003) 557–562
www.elsevier.com/locate/ssc
Microwave synthesis of lithium lanthanum titanate
Mulki H. Bhata, Anne Miurab, Phillipe Vinatierb, Alain Levasseurb, Kalya J. Raoa,
*
aSolid State and Structural Chemistry Unit, Indian Institute of Science, Bangalore 560 012, India
bGroupe Ionique du Solide, ENSCPB, 16 Avenue Pey Berland, 33607 Pessac Cedex, Bordeaux, France
Received 4 November 2002; accepted 27 November 2002 by C.N.R. Rao
Abstract
A microwave method has been developed for the preparation and sintering of lithium lanthanum titanate (LLTO) of the
formula Li0.35La0.55TiO3. The method is fast, clean, economical and gives phase pure product of high conductivity.
q 2003 Elsevier Science Ltd. All rights reserved.
PACS: 77.84.Dy
Keywords: B. Microwave synthesis; D. Solid electrolytes; D. Li-batteries
1. Introduction
these La ions (x) are substituted by thrice that number (3x) of
Liþ ions, the resulting LLTO (Li3xLa2/32xTiO3) is found to
exhibithighLi-ion conductivity. Itisnowwidely acceptedthat
Liþ ions make use of the available vacancies on the A sub-
lattice for transport, which is responsible for its high ionic
conductivity. It is easily seenthatthe maximumLi substitution
corresponds to x ¼ 0.166, or about 16%. Highest Liþ ion
conductivity of s ¼ 3.9 £ 1024 S cm21 (at 300 K) has been
observedinthesystemwithx ¼ 0.11 [1]. Liþ ionconductivity,
however,hasbeenfoundtobeinfluencedsignificantly,bothby
charge and size of the B ion (substitution of Ti by other
elements) [3–5,7–10]. Efforts have been made in literature to
correlate this influence to Li–O interactions in the lattice
[11–14]. LLTO therefore holds very high promise as an
electrolyte material in lithium batteries.
Depletion of fossil fuels and environmental concerns have
led to intense research in the area of solid state batteries. For
well-known advantages such as high energy densities, etc.,
lithium battery materials are at the forefront of the new
developments. Li-batteries also hold promise for the develop-
ment of thin film and small capacity energy sources which are
in high demand for applications in consumer electronics.
Discovery of high Li-ion conductivity in lithium lanthanum
titanate (LLTO) has generated new interest in this direction
[1–7]. LLTO is a member of a versatile class of oxide
materials, namely perovskites. Perovskites are oxides of the
general formula ABO3 whose structure can be visualized as
built of corner-sharing BO6 octahedra. A ions, which have
lower valency, are present in the cubic sites of high oxygen
coordination, as in the archetypal BaTiO3 (perovskite)
structure. It has been found possible to substitute both A and
B ions in ABO3 structure with a variety of other elements
[3–5,7–10] maintaining the overall electrical neutrality and
hence the chemical versatility of perovskite oxides. Substi-
tution of A ions alone with trivalent ions like La is
accomplished by insertion of 2/3 the number of A ions by La
keeping the remaining ion sites vacant. If now a fraction of
The preparation of LLTO is generally accomplished
through high temperature ceramic methods. The component
oxide materials (Li2CO3 for Li) are heated up to 1150 8C
with heating protocols requiring up to 24 h. Sintering of the
LLTO requires still higher temperatures of 1350 8C and
several hours. In fact, the high temperature (1350 8C) of
sintering has been widely known to result in evaporation
losses of lithia [14]. It is, therefore, essential to develop
more efficient methods of preparation and sintering of
LLTO.
*
Corresponding author. Tel.: þ91-80-360-2897; fax: þ91-80-
Microwaves have been increasingly used for the
synthesis and consolidation of materials in recent times
360-1310.
E-mail address: kjrao@sscu.iisc.ernet.in (K.J. Rao).
0038-1098/03/$ - see front matter q 2003 Elsevier Science Ltd. All rights reserved.
PII: S0038-1098(02)00852-9

558
M.H. Bhat et al. / Solid State Communications 125 (2003) 557–562
Fig. 1. XRD patterns of (a) conventionally and (b) microwave-prepared LLTO.
[15,16]. In general, microwave-assisted reactions are found
to be fast, clean, energy efficient and most remarkably occur
at temperatures much lower than in conventional processes.
It is now believed that the above features are germane to
microwave reactions because of the coupling of ionic
motions (through their charge) with the local electric fields
of penetrating microwaves. Such interactions give rise to
ponderomotive forces, which in turn are responsible for the
observed features of microwave reactions [17].
carbon, which acts as a secondary heater. The mixture is
pelletized and irradiated with microwaves in a domestic
microwave oven (Kenstar, 2.45 GHz, 900 W; tunable power)
using the maximum power available. In the second procedure,
the reactant mixture is pelletized, embedded in amorphous
carbon powder and irradiated by microwaves. In both
procedures, no more than 5–6 min were required for the
completion ofthereaction. Bothmethods leadtothe formation
of phase-pure LLTO. In the first procedure, excess carbon has
to be burned out by reheating the product at 700 8C for 3 h. In
the second procedure carbon sticking to the pellet was
removed by dusting. Microwave sintering of the pellet is
accomplished by placing the pellet between two pieces of
alumina inside a thin walled pre-sintered SiC pit. SiC acts as a
secondary heater for sintering. Sintering required just 50 min.
Sintered pellets have also been prepared by conventional
method by heating the pellets at 1350 8C for 6 h in a
Thermolyne furnace.
In this communication, we have reported a novel
microwave route for the preparation of LLTO. The structure
of microwave-prepared (MP-LLTO), its conductivity, its
siterability and microstructure, have been compared with
those of LLTO prepared by conventional method.
2. Experimental
The densities of the sintered pellets were determined
through apparent weight loss method. Xylene was used as the
immersion fluid. Powder X-ray diffraction patterns were
obtained using a Phillips 1390 X-ray diffractometer. Micro-
structural studies were carried out using a scanning electron
microscope (SEM-JEOL JSM 5600 LV). Differential thermal
analysis and thermo-gravimetric analysis (DTA–TGA) of the
reactant mixture was carried out on a Setaram TGDTA 92
thermal analyzer. Ionic conductivity data was recorded using
an HP-4192A Impedance Analyzer. 7Li NMR (MAS and
static) spectra were recorded on a Bruker MSL-300 solid-state
high-resolution spectrometer operating at 90.4 MHz (mag-
netic field 7.05 T). 908 pulses of 5 ms were employed with a
delay time of 10 s between pulses in all the experiments. In the
The composition of LLTO investigated in this work is
La0.55Li0.35TiO3. LLTO has been prepared by both conven-
tional method as described in Ref. [1] and microwave method
described below. In the conventional method, mixture of
stoichiometric amounts of La2O3, TiO2 and Li2CO3 (high
purity; Aldrich 99 þ %) is initially heated at 800 8C for 4 h in
order to completely decompose the Li2CO3 in the mixture.
This is followed by heating at 1150 8C for 36 h with
intermittent grinding after every 12 h. The heating and cooling
rate for each cycle was maintained at 10 K min21. Microwave
synthesis of LLTO of the same composition is achieved in two
ways. In the first, a similar initial mixture as in the
conventional procedure is further mixed with amorphous

M.H. Bhat et al. / Solid State Communications 125 (2003) 557–562
559
case of the MAS experiments, a spinning speed of 5–9 kHz
was employed. All the spectra were recorded at room
temperature using freshly powdered samples.
are consequences of sintering. However, MP-LLTO pow-
ders also contain large particles of faceted nature, but there
are no clear indications of junction formation. This is a
likely consequence of short residence time in microwaves
insufficient for recrystallization and grain growth. The SEM
microstructures of the conventionally sintered LLTO (CS-
LLTO) and microwave-sintered LLTO (MS-LLTO) are
shown in Fig. 3(a) and (b), respectively. SEM of CS-LLTO
reveals the formation of larger grains of cubic morphology
and good sintering. MS-LLTO samples observed under
identical conditions reveal the presence of faceted grains
with larger distribution of sizes and extensive junction and
interface formation. But in both processes, presence of
enclosed porosity is clearly observed. Using a theoretical
density of 4.999 (from XRD) and the observed densities of
4.850 of the MS and 4.754 of the CS samples, it is evident
that slightly better sintering (0.97 TD) is achieved in
microwaves than in the conventional method (0.95 TD).
Presence of large microwave fields at interparticle junctions
and in cavities lead to ponderomotive forces which brings
about rapid sintering and grain growth in microwaves.
The d.c. conductivities of the MS and CS samples are
3. Results and discussion
XRD pattern of LLTO prepared by conventional and
microwave methods are presented together in Fig. 1 for 2u
values between 20 and 1008. All the peaks compare
excellently with those reported in the literature (JCPDS
46-0465). The XRD of microwave-prepared (MP) and
conventionally prepared (CP) samples are also in excellent
agreement between themselves. The crystal structure
corresponds to the cubic Pm3m space group.
The SEMs of the CP and MP samples are given in Fig. 2.
The as-prepared LLTO have large distribution of sizes in
both CP and MP samples. The MP-LLTO appears to consist
of well-faceted particles while the conventionally prepared
LLTO (CP-LLTO) consists of well-rounded particles.
Therefore in CP-LLTO indications are that sintering has
already begun to occur. The presence of a massive particle A
and a triangular junction B are indicated in Fig. 2(a), which
Fig. 3. SEMs of (a) conventionally sintered and (b) microwave-
sintered LLTO.
Fig. 2. SEMs of (a) as-prepared and (b) microwave-prepared LLTO.

560
M.H. Bhat et al. / Solid State Communications 125 (2003) 557–562
Fig. 4. DC conductivity of conventionally and microwave-sintered LLTO.
shown in Fig. 4. The room temperature conductivities
of the MS and CS-LLTO are 8.63 £ 1024 and
8.95 £ 1024 S cm21, respectively. These values are in
good agreement with literature reports (3.9 £ 1024 S cm21
as reported in Ref. [1]). Although there is a slightly higher
enclosed porosity in the CS samples, the conductivities of
the CS and MS samples are in remarkably good agreement.
Therefore, the Liþ ion transport in LLTO is largely a bulk
phenomenon. The activation barrier of 0.23 eV observed in
the present studies are somewhat lower than the values
reported in the literature [1,2]. However, the temperature
range employed in the present studies is narrow compared to
those in the literature. The conductivities are also known to
exhibit VTF type behavior [1].
7Li NMR recorded at laboratory temperature with and
without MAS are shown together in Fig. 5. It is seen that
even without spinning the resonances are very narrow and
the MAS resonances are expectedly narrower as the dipolar
interactions have been removed. The FWHM values from
the powder patterns for the CP and MP samples are 2.17 and
1.94 ppm, respectively, and indicate motional narrowing
consistent with the high value of Liþ ion conductivity and
low activation barrier.
loss extends up to nearly 900 8C. There is clear correspon-
dence between the several stages of weight loss observed in
TGA and the endotherms in DTA. The thermal behavior is
curious as the only component of the mixture which can
exhibit features in DTA or TGA is Li2CO3 because it
undergoes thermal decomposition. But DTA of Li2CO3
obtained separately reveals that Li2CO3 ! Li2O þ CO2
decomposition occurs around 722 8C. Therefore, the origin
of the observed complex thermal features is unclear. We,
therefore, propose the following speculative explanation.
Li2O forms as a layer over Li2CO3 particles during
decomposition. The decomposition progresses inwards
from the surface. Since La2O3 and TiO2 are present in the
immediate neighborhood of the particle it leads to formation
of LLTO layers, which forms a coating around undecom-
posed part of Li2CO3 particles. Further decomposition of
Li2CO3 is inhibited till such temperatures when CO2
pressure builds up in Li2CO3 encapsulated by the LLTO
coating. At higher temperatures, therefore, the capsules
burst and further decomposition occurs. Such burst-caused
exposure of Li2CO3 occurs at stages and hence the observed
thermograms. This explanation is consistent with the
observation that reactants have to be ground at least twice
during conventional preparation of LLTO in order to
complete the reaction (see Section 2). The reaction
pathway—including Li2CO3 decomposition—in microwave
preparation is likely to be very different. The roto-
vibrational modes of oxy-anions like CO232 are excited in
microwaves and within the particles, and causes bulk
heating as is typical of microwave irradiation. The reaction
is not only rapid, but also continuous. Also, it is completed
in a single stage.
The thermal behavior of the reactant mixture used for
making LLTO has been examined using DTA and TGA in
order to understand the nature of the reactions (Fig 6). The
DTA of Li2CO3 itself has been included for comparison.
There are several endotherms observed in the DTA. TGA
reveals abrupt loss of weight around 320 8C. Above this
temperature and up to 700 8C there is further loss which is
semi-continuous. In the region above 700 8C, there is again a
relatively abrupt loss of weight and the process of weight

M.H. Bhat et al. / Solid State Communications 125 (2003) 557–562
561
Fig. 5. 7Li NMR (static and MAS) for (a) conventionally and (b) microwave-prepared LLTO.

562
M.H. Bhat et al. / Solid State Communications 125 (2003) 557–562
Fig. 6. (a) DTA and TGA of the reaction mixture for LLTO preparation (b) DTA of Li2CO3.
[4] Y. Inaguma, M. Itoh, Solid State Ionics 86–88 (1996) 257.
4. Conclusions
[5] H.T. Chung, J.G. Kim, H.G. Kim, Solid State Ionics 107
(1998) 153.
In conclusion, we have described a microwave method for
the preparation of LLTO which providesanelegant alternative
for the conventional procedure. The microwave methodisalso
shown to be well suited for sintering of LLTO. The properties
of the microwave-prepared products have been shown to be
entirely comparable with those of the CP-LLTO.
[6] J.L. Fourquet, H. Duroy, M.P. Crosnier-Lopez, J. Solid State
Chem. 127 (1996) 283.
[7] L. Latie, G. Villeneuve, D. Conte, G.L. Flem, J. Solid State
Chem. 51 (1984) 293.
[8] O. Bohnke, J. Emery, A. Veron, J.L. Fourquet, J.Y. Buzare, P.
Florian, D. Massiot, Solid State Ionics 109 (1998) 25.
[9] T. Brousse, P. Fragnaud, R. Marchand, D.M. Schleich, O.
Bohnke, K. West, J. Power Sources 68 (1997) 412.
[10] M. Nakayama, H. Ikuta, Y. Uchimoto, M. Wakihara, J. Mater.
Chem. 12 (2002) 1500.
Acknowledgements
The authors would like to thank the European Commission
Project (EEC-06) for supporting this work. One of the authors
(M.H.B) would like to thank the CSIR for financial support.
[11] Y. Inaguma, T. Katsumata, M. Itoh, Y. Morii, J. Solid State
Chem. 166 (2002) 67.
[12] Y. Inaguma, C. Liquan, M. Itoh, T. Nakamura, T. Uchida, H.
Ikuta, M. Wakihara, Solid State Commun. 86 (10) (1993) 689.
[13] M.A. Paris, J. Sanz, C. Leon, J. Santamaria, J. Ibarra, A.
Varez, Chem. Mater. 12 (2000) 1694.
References
[14] C.A. Kirk, A.R. West, J. Solid State Chem. 162 (2001) 379.
[15] K.J. Rao, B. Vaidhyanathan, M. Ganguli, P.A. Ramakrishnan,
Chem. Mater. 11 (1999) 882.
[1] O. Bohnke, C. Bohnke, J.L. Fourquet, Solid State Ionics 91
(1996) 21.
[2] Y. Inaguma, L. Chen, M. Itoh, T. Nakamura, Solid State Ionics
70/71 (1994) 196.
[16] W.H. Sutton, Am. Ceram. Soc. Bull. 68 (1989) 376.
[17] K.I. Rybakov, V.E. Semenov, Phys. Rev. B 52 (1995) 3033.
[3] M. Itoh, Y. Inaguma, W.H. Jung, L. Chen, T. Nakamura, Solid
State Ionics 70/71 (1994) 203.