(Maxim Borowski) Perovskites Structure, Propertie PDF

CHEMICAL ENGINEERING METHODS AND TECHNOLOGY
PEROVSKITES: STRUCTURE,
PROPERTIES AND USES
No part of this digital document may be reproduced, stored in a retrieval system or transmitted in any form or
by any means. The publisher has taken reasonable care in the preparation of this digital document, but makes no
expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No
liability is assumed for incidental or consequential damages in connection with or arising out of information
contained herein. This digital document is sold with the clear understanding that the publisher is not engaged in
rendering legal, medical or any other professional services.
CHEMICAL ENGINEERING METHODS
AND TECHNOLOGY
Additional books in this series can be found on Novas website

under the Series tab.
Additional E-books in this series can be found on Novas website

under the E-books tab.
CHEMICAL ENGINEERING METHODS AND TECHNOLOGY
PEROVSKITES: STRUCTURE,
PROPERTIES AND USES
MAXIM BOROWSKI
EDITOR
Nova Science Publishers, Inc.

New York
Copyright 2010 by Nova Science Publishers, Inc.
All rights reserved. No part of this book may be reproduced, stored in a retrieval system or
transmitted in any form or by any means: electronic, electrostatic, magnetic, tape, mechanical
photocopying, recording or otherwise without the written permission of the Publisher.
For permission to use material from this book please contact us:
Telephone 631-231-7269; Fax 631-231-8175
Web Site: http://www.novapublishers.com
NOTICE TO THE READER

The Publisher has taken reasonable care in the preparation of this book, but makes no expressed or
implied warranty of any kind and assumes no responsibility for any errors or omissions. No
liability is assumed for incidental or consequential damages in connection with or arising out of
information contained in this book. The Publisher shall not be liable for any special,
consequential, or exemplary damages resulting, in whole or in part, from the readers use of, or
reliance upon, this material. Any parts of this book based on government reports are so indicated
and copyright is claimed for those parts to the extent applicable to compilations of such works.
Independent verification should be sought for any data, advice or recommendations contained in
this book. In addition, no responsibility is assumed by the publisher for any injury and/or damage
to persons or property arising from any methods, products, instructions, ideas or otherwise
contained in this publication.
This publication is designed to provide accurate and authoritative information with regard to the
subject matter covered herein. It is sold with the clear understanding that the Publisher is not
engaged in rendering legal or any other professional services. If legal or any other expert
assistance is required, the services of a competent person should be sought. FROM A
DECLARATION OF PARTICIPANTS JOINTLY ADOPTED BY A COMMITTEE OF THE
AMERICAN BAR ASSOCIATION AND A COMMITTEE OF PUBLISHERS.
Additional color graphics may be available in the e-book version of this book.
LIBRARY OF CONGRESS CATALOGING-IN-PUBLICATION DATA
Perovskites : structure, properties, and uses / editor, Maxim Borowski.

> p. cm.
> Includes index.
> ISBN 978-1-61668-870-7 (eBook)
> 1. Perovskite. I. Borowski, Maxim.
> QE391.P47P474 2010
> 549'.528--dc22
> 2010007616
Published by Nova Science Publishers, Inc. New York

CONTENTS
Preface vii
Chapter 1 Controlled Fabrication and Catalytic Applications of Specifically
Morphological and Porous Perovskite-type Oxides 1
Jiguang Deng, Lei Zhang, Yuxi Liu and Hongxing Dai
Chapter 2 Perovskites and Their Nanocomposites with Fluorite-Like Oxides as
Materials for Solid Oxide Fuel Cells Cathodes and Oxygen-
Conducting Membranes: Mobility and Reactivity of the
Surface/Bulk Oxygen as a Key Factor of Their Performance 67
V.A. Sadykov , S. N. Pavlova, T. S. Kharlamova,
V. S. Muzykantov, N.F. Uvarov, Yu. S. Okhlupin,
A. V. Ishchenko, A. S. Bobin, N. V. Mezentseva,
G.M. Alikina, A.I. Lukashevich, T.A. Krieger, T. V. Larina,
N. N.Bulgakov, V. M. Tapilin, V.D. Belyaev, E.M. Sadovskaya,
A. I. Boronin, V. A. Sobyanin, O. F. Bobrenok, A. L. Smirnova,
O. L. Smorygo and J. A. Kilner
Chapter 3 On the Nature of Low-Temperature Resistive Peak in Colossal
Magnetoresistant Materials 179
A.I. Tovstolytkin
Chapter 4 Structural, Magnetic and Electron Transport Properties of
Ordered-Disordered Perovskite Cobaltites 213
Asish K. Kundu and B. Raveau
Chapter 5 Electrically Tunable Dielectric and Conduction Properties in
Perovskite Thin Films 251
Yidong Xia, Xiangkang Meng and Zhiguo Liu
Chapter 6 Microwave Dielectric Properties of Ca[(Li1/3A2/3)1-xMx]O3- [A=Nb,
Ta and M=Ti, Zr, Sn] Complex Perovskites: A Review 283
George Sumesh and Thomas Sebastian Mailadil
vi Contents
Chapter 7 Perovskite-type Oxides: Synthesis and Application in Catalysis 319

Junjiang Zhu and Jinfa Chen
Chapter 8 Mechanism of Formation of Perovskite-type Layered Oxides 345
Irina A. Zvereva and Alexander B. Missyul
Chapter 9 Some Theoretical Aspects of Magnetic Structure, Spin Excitations
and Magnetization of Manganites 377
B. V. Karpenko
Chapter 10 Correlations between the Structure of Ordered Solid Solutions and
Parameters of Pair Interactions in Perovskites 437
A. Yu. Gufan and K. Yu. Gufan
Chapter 11 ABX3-Type Oxides and Halides: Their Structure and Physical
Properties 463
A. S. Verma and V. K. Jindal
Chapter 12 Ferroelectric PbTiO3: From a Single-domain State to Composite
Components 481
V. Yu. Topolov and A. V. Krivoruchko
Chapter 13 Double Perovskites with Structure-disordered Oxygen Sublattice as
High-temperature Proton Conductors 501
I. Animitsa
Chapter 14 Short- and Long-range Ferromagnetic Orders in La1-xPbxMnO3:
Critical Behavior Study 525
The-Long Phan and Seong-Cho Yu
Chapter 15 Reddish Ceramic Pigments with Perovskite Structure 539
C. P. Pia
Index 559
PREFACE
Perovskite is a calcium titanium oxide mineral species composed of calcium titanate.

This new book discusses perovskite thin films, which are widely employed in today's
advanced technology. The broad range of physical properties in such category of materials
has offered various functionalities in devices ranging from dynamic random access memories
and ferroelectric nonvolatile memories to piezoelectric and optical devices. The structural,
magnetic and electron transport properties of ordered-disordered perskovite cobaltites is
discussed as is perovskite-type oxides, ferroelectric PbTiO3 from a single-domain state to
composite components.
Chapter 1-Perovskite-type oxides and perovskite-like oxides can tolerate substantial
substitutions in one or both cationic sites while maintaining their original crystal structures.
Such a feature renders the chemical tailoring of the materials via partial replacement of the
cationic site(s) with foreign metal ions, hence modifying their physicochemical properties.
Perovskite-type oxides and perovskite-like oxides have been found many applications in
physics and chemistry. The physicochemical properties of these materials are dependent on
the crystal structure, lattice defect, exposed lattice plane, surface morphology, particle size,
and specific surface area as well as the pore structure. Up to now, a large number of
perovskite-type oxides and perovskite-like oxides have been generated and investigated for
clarifying their physicochemical properties. In the past years, although several books and
reviews relevant to such a kind of materials appeared in the literature, a systematic
summarization on the newly published works related to the fabrication, characterization, and
catalytic applications of perovskite-type oxides and perovskite-like oxides with the particles
being nanosized, specifically morphological or porous has not been available. Due to the
unique properties of these nano- and/or porous materials with low dimensionality, author
shall summarize the results of such functional solid materials recently reported in the
literature, and aim to gain insights into the relationship between structure and property and
the potential applications in catalysis. Therefore, author shall focus the author attention in
Section 2 on the controlled fabrication of perovskite-type oxides and perovskite-like oxides
with specific morphologies, such as nano/microwires, nano/microrods, nano/microtubes,
nano/microfibers, nano/
microcubes, nano/microspheres, nano/microflowers, and etc. Due to the superiority in
physicochemical property of porous materials to their bulk counterparts, author shall also
include the synthesis of macro- and/or mesoporous perovskite-type oxides and perovskite-like
oxides (although limited works have been done in the aspect) in this section. In Section 3,
viii Maxim Borowski
author shall highlight the physicochemical properties of perovskite-type oxides and

perovskite-like oxides fabricated by the researchers mentioned in Section 2, so that one can
generalize some clear correlations between the fabrication strategy and the crystal or porous
structure or morphology. In Section 4, author shall concentrate on the applications of these
addressed materials in catalysis, such as NOx elimination, CH4 complete oxidation, and
volatile organic compounds removal, and etc. Finally, author shall propose the perspectives
and concluding remarks of the research in the future in Section 5.
Chapter 2- This paper presents results of research aimed at elucidating factors controlling
mobility and reactivity of the surface and bulk oxygen of complex perovskite-like AnBmOz
oxides (manganites, ferrites, cobaltites, nickelates of La/Sr) and their nanocomposites with
fluorite-like oxides (doped ceria or zirconia) as related to their performance in the
intermediate temperature solid oxide fuel cells (IT SOFC) cathodes and oxygen separation
membranes. Real/defect structure and surface properties of these materials as a function of
both composition, preparation route and sintering temperature were studied by combination of
TEM with EDX, XRD, EXAFS, UV-Vis and XPS methods. Lattice oxygen mobility, strength
of oxygen species bonding with the surface sites and their reactivity for both dispersed and
dense materials were estimated by oxygen isotope exchange, temperature-programmed
desorption and reduction by H2 and CH4. Coefficients of oxygen diffusion in dense materials
were estimated by analysis of dynamics of the oxygen loss monitored by measurements of the
weight loss or conductivity relaxations. For analysis of atomic-scale factors controlling
oxygen bonding strength and mobility, quantum-chemical approaches including the Density
Functional Theory (DFT) in spin-polarized SLDA approximation and semiempirical
Interacting Bonds Method were applied.
For simple perovskites, along with well-known effects of anion vacancies/interstitial
oxygen atoms and high- charge transition metal cations generated due to the oxygen non-
toichiometry, pronounced effects of the surface faces termination, ordering of oxygen
vacancies and appearance of extended defects were found to affect the oxygen mobility and
reactivity as well. From the microscopic point of view, decreasing the Me-O bonding strength
and increasing the distance between A and B cations due to the lattice parameter increase
helps to decrease the activation barriers for the oxygen diffusion in the perovskite-like lattice.
For perovskite-fluorite nanocomposites, interfaces were shown to be paths for fast
oxygen diffusion due to redistribution of cations between contacting phases.
The most promising combinations of perovskites and perovskite-fluorite nanocomposites
possessing a high lattice oxygen mobility and conductivity were used for synthesis of
functionally graded cathodes of IT SOFC with thin layers of doped zirconia electrolytes as
well as for manufacturing asymmetric Ni-Al compressed foam supported membranes for
oxygen separation and syngas generation. High and stable performance of these materials
promising for the practical application was demonstrated.
Chapter 3- Strong interplay between electronic behavior, structure and magnetism in
doped perovskite manganites gives rise to the enhanced sensitivity of transport properties to
magnetic and structural peculiarities. The optimally doped manganite compounds are metallic
and ferromagnetic at low temperatures, while their conductivity displays activated behavior at
high temperatures. The metal-insulator transition between these two states is strongly
associated with the magnetic ordering one, and this manifests itself in the formation of a
resistivity peak in the vicinity of Curie temperature TC. However, in some peculiar cases,
manganite samples exhibit an additional low-temperature resistivity peak (LT-peak) (or a
Preface ix
tendency to its emergence in a form of a plateau), which is not associated with any magnetic
or structural transition and nature of which remains unclear hitherto. In this chapter, based on
a simple phenomenological model, author demonstrate that the formation of the LT-peak may
be a result of a strong broadening of the ferromagnetic transition. The approach developed
examines the electrical properties of a system consisting of paramagnetic and ferromagnetic
phases which exhibit opposite trends in the temperature dependences of resistivity and coexist
over a wide temperature range. It is shown that, as the region of the phase coexistence
increases, the resistive transition first broadens, then, in addition to the resistive peak near TC,
the LT-peak emerges and begins to grow, later on it becomes dominant and, eventually, only
a weak shoulder-like feature remains at TC. Such an approach is shown to be applicable to the
description of the magnetic and electric properties of a number of inhomogeneous and phase-
separated manganite systems, including micro- and nanocrystalline samples, ultrathin
epitaxial films, etc.
Chapter 4- Rare earth perovskite cobaltites are increasingly recognized as materials of
importance due to rich physics and chemistry in their ordered-disordered structure for the
same composition. Apart from colossal magnetoresistance effect, like manganites, the
different forms of cobaltites exhibit interesting phenomena including spin, charge and orbital
ordering, electronic phase separation, insulator-metal transition, large thermoelectric power at
low temperature. Moreover, the cobaltites which display colossal magnetoresistance effect
could be used as read heads in magnetic data storage and also in other applications depending
upon their particular properties. The A-site ordered-disordered cobaltites exhibit
ferromagnetism and metal-insulator transitions as well as other properties depending on the
composition, size of A-site cations and various external factors such as pressure, temperature,
magnetic field etc. Ordered cobaltites, having a 112-type layered structure, are also reported
to have an effectively stronger electron coupling due to layered A-site cationic ordering. Most
importantly for the present article author focus on La-Ba-Co-O based ordered-disordered
perovskite phases, which exhibit interesting magnetic and electron transport properties with
ferromagnetic transition, TC ~ 177K, and it being the first member of lanthanide series. Zener
double exchange mechanism considered to be crucial for understanding basic physics of the
ferromagnetic-metallic phase, yet does not explain clearly the insulating-type phase. In terms
of electron transport the ferromagnetic-metallic or insulating/semiconducting states have been
discussed in the present article with different types of hopping model.
Chapter 5- Perovskite thin films are widely employed in todays advanced technology.
The broad range of physical properties in such category of materials has offered various
functionalities in devices ranging from dynamic random access memories and ferroelectric
nonvolatile memories to piezoelectric and optical devices. In addition to the aforementioned
applications, perovskites have been revealed to be promising candidate materials in many
novel applications including electrically tunable microwave devices and resistive switching
memories. Thin films of the perovskite ferroelectrics are being used to develop the tunable
microwave devices owing to their strong electric field dependent dielectric constant in the
paraelectric state. Reversible switching between high- and low- resistance states induced by
electric field makes perovskite thin films suitable for the applications in resistive switching
memories. This chapter focuses on the author recent progress on these two kinds of
electrically tunable properties in perovskite thin films. Based on the origin of dielectric
tunability that is defined as the variation of dielectric constant in a certain electric field range,
approaches to improve the tunable dielectric properties will be introduced. On the other hand,
x Maxim Borowski
although resistive switching behaviors have been discovered in many perovskites during the
last decade, the origin of such switching is not well understood yet. In this chapter, some
proposed mechanisms responsible for the resistive switching will be discussed.
Chapter6- The perovskite family of materials is of considerable technological importance
for its excellent temperature stable microwave properties for dielectric resonator based filters,
oscillators and antenna applications. In this chapter author review the preparation, charact-
erization and the microwave dielectric properties of Ca[(Li1/3A2/3)1-xMx]O3- [A=Nb, Ta and
M=Ti, Zr, Sn] dielectric ceramics. This family of perovskite materials shows relative
permittivity in the range 20 to 56 with a quality factor up to 45000 GHz and temperature
coefficient of resonant frequency (f) in the range from 21 to +83 ppm/C. The f can be
tailored by adjusting the titanium content. The sintering temperature can be lowered below
950 oC to suit LTCC by the addition of low melting glasses.
Chapter 7- Two topics of perovskite-type oxides were browed and discussed in this
chapter: one is the synthesis method and the other is its application in catalysis. Because of
the diverse applications of perovskite-type oxides, special properties are usually required
depending on the ultimate use, while the desired properties could be achieved by choosing a
proper synthesis method. For example, using coprecipitation method could lead to a well
crystallized structure, while by a nanocasting method perovskite with high surface area could
be achieved. Thus, an overview of synthesis method for perovskite-type oxides reported in
the publications was made for better understanding the advantage of each method.
Efforts on the application of perovskite-type oxides were made exclusively in catalysis in
this part. Attentions on using these oxides as catalyst are increasing year by year not only
because of their low manufacturing cost but also due to the unique properties, such as high
thermal stability, exchangeability of cations and of course, considerable activity. Many
catalytic reactions using perovskite-type oxides as catalyst has been attempted, and although
for a given reaction catalyst with different ions, either at A- or B-site, show different catalytic
activities, a common point is that the presence of oxygen vacancy is necessary for the reaction
to proceed. From the nature of catalytic reaction, that is the difficulty of electron transferring
between reactant and catalyst, it was proposed, exampled by the reactions of NO
decomposition and NO+CO reduction, that the crucial pre-requisite for a catalytic reaction to
occur is that the full redox cycle of catalyst can proceed easily. Thus only when this
requirement is fulfilled, other parameters, such as the amount of active sites, could come into
the focus of catalyst design.
Chapter 8- The chapter presents general analysis and results of experimental data on the
structure-chemical investigation of process of formation of perovskite type structures in
different groups of layered oxides, which exhibit a number of practically important properties.
In many cases the layered perovskite-like phases are formed in a multistage way. The
composition and stability of the intermediates can strongly affect both reaction rate and the
structure of the final compound. Examples of such impact for several Ruddlesden-Popper and
Aurivillius phases are considered. Formation of the double-perovskite layered phases for the
series of high temperature aluminates Ln2SrAl2O7 (Ln=La, Nd, Sm, Eu, Gd, Ho) is discussed
in terms of the dependence of mechanism of reaction on the nature of rare earth elements. For
the double-layered manganites LnSr2Mn2O7 (Ln=La, Nd, Gd) the structure and stability of the
intermediate perovskite phase affects both the reaction speed and the microstructure of the
resulting Ruddlesden-Popper phase, including phase separation in the CMR manganites. The
double-layered ferrites Ln2SrFe2O7 (Ln=La, Nd, Gd) exhibit the evolution of oxidation state
Preface xi
of iron along the process of synthesis. Mechanism of formation of the ferroelectric Aurivillius
phase Bi3TiNbO9 demonstrates the limitations on the BiLn and TiMn substitution in this
structure. The stability of the intermediate phase results in the substitution limitations.
Investigations of the processes occurring in the reaction mixture during synthesis of the
perspective ionic conductors and catalyst (Na,Nd)n+1TinO3n+1 allow to explain the necessity of
the excess of alkali metal containing starting compound, optimize the conditions of synthesis
and significantly decrease the size of particles.
Chapter 9- The behavior of the system La1 xCa x MnO3 is considered theoretically at the
absolute zero temperature (magnetic configurations), at the low temperatures (spin waves)
and at the high temperatures (magnetization).
The appearance of the different magnetic configurations in the row 0 x 1 is studied.
Superexchange, double exchange and anisotropy energy are taken into account. The spin
3+
operator of double exchange interaction in crystal between ions with different valence Mn
4+
and Mn is the straight generalization of two - spin operator problem of Anderson
Hasegawa molecule. Ground state energy minimization gives 11 types of magnetic
configurations: two ferromagnetic, three collinear antiferromagnetic and six non-collinear.
When the concentration of Ca ions changes, then the spin configurations replace one another
as the ground state. As a whole the sequence of configurations when x changes from 0 to 1
can be brought in correspondence to those observed at the experiment. The appearance of the
angle configurations in the systems may occur if the Heisenberg and non Heisenberg
exchange exist simultaneously. The comparison with experiment was made by means of
mixed procedure: part of the numerical values of interaction parameters and transition
concentrations x were taken from the experiment while the missing values of these
parameters were calculated with the help of the corresponding theoretical relations. The
proposed dynamical alloy model of different spins can be used for description of real
perovskite-like systems behavior.
The elementary spin excitations were studied in ferromagnets and antiferromagnets for
the compounds with the special contents and theoretical dispersion curves were compared
with experimental ones obtained by neutron inelastic scattering method for different
directions in the first full Brillouin zone. Theoretical expressions are obtained for the spin-
wave spectrum in a ferromagnetic cubic crystal with variable-valence ions. Interactions up to
the fourth coordination shell are taken into account. Two kinds of exchange are assumed to
occur in the system. As a result, the numerical values of total exchange parameters for each of
the coordination shells can be found. These studies are made for ferromagnets at
x = 0.17,0.25,0.30 . Theoretical expressions are obtained for energy of spin waves in
layered antiferromagnetic structure of the perovskite-like compound LaMnO3 . The quadratic
exchange interactions of the central spin moment of the trivalent manganese ion with its
eighteen neighbors from five coordination shells and single-ion magnetic anisotropy energy
are taken into account. For the three principal crystallographic directions, the theoretical
results are compared with the experimental data. It is shown that the in-plane ferromagnetic
interaction can be more than sixteen times as great as the interplanar antiferromagnetic
interaction.
xii Maxim Borowski
The high temperature dependence of spontaneous magnetization at a separate site in the

compounds with 0 x 1 is investigated in the molecular field approximation using a
homogeneous model for the magnetic sub-lattice with allowance for the Heisenberg and
double exchange between nearest neighbors. Unlike with the usual magnets having one type
of magnetic ions, in this case there appears a system of equations with two unknowns mean
spins of manganese ions. In the vicinity of the ordering temperature, this system can be
solved analytically and the expression for the ordering temperature can be obtained as well.
To find the temperature dependence of magnetization in the whole temperature range the
exact expression for the double exchange operator containing spin operators under square root
was used. The numerical solutions were found in this case.
Chapter 10- In the chapter two approaches to the crystals ordered states description are
used. The first is basing on the group theoretical considerations incorporated in theory with
the help of Landau potential. The second way of ordering description consist of accounting
different models of interatomic interactions. The author goal is to elucidate correlations
between these two approaches. Complex oxides with composition

AB / B // O
c 1 c 3 and
perovskite like structure are considered as an example. Six supercells which may be realized
as stable states in these solid solutions in accord with Landau theory are analyzed in the
framework of both approaches. The potential of interatomic coupling includes van der Waals
energy and terms following from pseudopotential model. The sum of interatomic energy in
twelve coordination spheres was accounted to get correlation between the parameters of
Landau potential and coefficients of interatomic energy and to define the domain of different
ordering state stability in the space of phenomenological parameters. In particular it is shown
that under certain conditions the type of ordering in a surface area may differ from the type of
ordering in a bulk of crystal. Using a new approach to X-ray diffraction pattern interpretation
there was established definite correlations between the parameters of pair interactions
potential and the peculiarities of ordered states which may be defined from the experiment.
Chapter 11- The term `perovskite` is used to denote a category of inorganic crystalline
solids with the general formula ABX3, where A and B are cations and X (oxides or halides) is
an anion. The perovskite structure occupies a prominent place under all the known ternary
systems of composition ABX3. This is due not only to its wide occurrence, but also to a series
of interesting and useful properties associated with this structural type. Crystallographic
behaviour of perovskites is of great interest because most structures are close to an ideal cubic
structure, however, frequently they are slightly distorted resulting in structures with lower
symmetry such as orthorhombic, tetragonal, rhombohedral, trigonal systems and so on. These
lattice distortions and structure changes are governed by temperature, pressure, chemical
composition, and in some cases, electric field. As a rule, with rising temperature, perovskites
tend to undergo a series of transitions to progressively higher symmetry, culminating in the
cubic structure where experimentally accessible. They are widely investigated, not only
because of interest in their crystal structural behaviour, but perovskites also exhibit a variety
of interesting electronic, electromechanical and conductive properties, which are the basis for
many existing and potential applications. Such properties are often symmetry dependent. In
this chapter, a summary of the scientific work performed on the ABX3 (X = Oxides and
Halides) system during the last few years is presented.
Chapter 12- Regular ferroelectric lead titanate (PbTiO3) is one of the well-known
perovskite-type compounds with physical properties studied intensively in the last decades.
Preface xiii
Of specific interest is PbTiO3 as an element of a hierarchical chain single-domain single

crystal polydomain single crystal poled ceramic composite, and the last element,
composite, can be based on other elements of the aforementioned chain. In this chapter
examples of modern piezo-active composites based on either PbTiO3 or PbTiO3-type solid
solutions are considered. These composites are regarded as matrix ones with 03 and 13
connectivity patterns and with a regular distribution of inclusions. To characterize the
effective electromechanical properties of the composite, two methods are applied. The
effective properties of the 03 composite with spheroidal inclusions and of the 13 composite
with circular cylindric inclusions are predicted by means of the effective field method. The
effective properties of the 13 composite with inclusions in the form of rectangular
parallelepiped are determined within the framework of the matrix method. In both the
methods, an electromechanical interaction between components is taken into account. The
role of the PbTiO3-containing component is emphasized in connection with the large
piezoelectric anisotropy and considerable piezoelectric sensitivity of the composite in certain
volume-fraction ranges.
Chapter 13- Materials with high proton or oxide ion conductivity are widely studied for
their potential application as electrolytes in fuel cells, electrolysers, batteries, sensors, etc. [1-
5]. Acceptor doped perovskite-type oxides ABO3- where A is Sr, Ba and B is Ce, Zr are well-
known high temperature proton conductors in wet atmosphere. It was found that the main
factor responsible for the appearance of protons in the structure was the presence of oxygen
vacancies. Upon hydration by equilibration with water vapor the oxygen vacancies may be
filled by oxygen from water and, in accordance with dissociative mechanism of water
dissolution, the hydroxide ions are formed.
Chapter 14- Author have studied the influence of Pb-doping on critical behavior of
polycrystalline manganites of La1-xPbxMnO3 (x = 0.1, 0.2 and 0.3) around their
ferromagnetic-paramagnetic phase transition temperature (TC) by means of analyzing
isothermal magnetization data M-H. According to the critical criterion proposed by Banerjee,
it was found that all three samples undergo the second-order phase transition as showing a
positive slope of isotherm plots of H/M versus M2. The critical parameters were determined
by the modified Arrott method, used the Arrott-Noakes equation of state M1/ versus (H/M)1/
and the asymptotic relations in the critical region. The obtained results reveal that with
increasing the Pb-doping concentration in La1-xPbxMnO3, the TC value increases from 162.0 K
(for x = 0.1) to about 346.3 K (for x = 0.3). Concurrently, there is a remarkable variation in
their critical exponents: = 0.498, = 1.456 and = 3.92 for x = 0.1; = 0.499, = 1.241
and = 3.49 for x = 0.2; and = 0.502, = 1.063 and = 3.12 for x = 0.3. Having compared
to the critical parameters expected from the theoretical models of mean-field theory, 3D-Ising
model, and 3D-Heisenberg model, author have found an existence of the short-range
ferromagnetic order in the samples with x = 0.1 and 0.2, and the long-range ferromagnetic
order in the x = 0.3 sample. Under such circumstances, magnetic interaction mechanisms
between Mn3+-Mn4+ in the samples would be changed, particularly in the paramagnetic
region.
Chapter 15 - In the past (1980), the ceramic pigments, 51 pigments in all, were classified
into 14 structures of industrial interest and these were classified into black, white and colored
pigments. These structures are oxides, materials used in ceramic as pigments because of their
greater stability in oxygen-containing ceramic systems. However, thanks to modern and more
xiv Maxim Borowski
refined methods of scientific investigation, new structures, new pigments, new synthetic
methods and new applications were also included. Among these author have the development
of new red ceramic pigments as perovskite structure of formula A3+B3+O3 as ferrites
LnFeO3 and chromium aluminates LnAl1-xCrxO3 (Ln, lanthanide). These lanthanides
aluminates ceramic pigments have been synthesized by different methods and
characterization studies were realized by X ray diffraction, SEM, UV-VIS spectra, etc. The
properties and uses of these lanthanides are appropriate as ceramic pigments; nevertheless,
new possibilities for these pigments are its use for luminescent coatings in a diversity of
devices or lasers.
In: Perovskites: Structure, Properties and Uses ISBN 978-1-61668-525-6
Editor: Maxim Borowski, pp. 1-66 2010 Nova Science Publishers, Inc.
Chapter 1
CONTROLLED FABRICATION AND CATALYTIC

APPLICATIONS OF SPECIFICALLY MORPHOLOGICAL
AND POROUS PEROVSKITE-TYPE OXIDES
Jiguang Deng, Lei Zhang, Yuxi Liu and Hongxing Dai*

Laboratory of Catalysis Chemistry and Nanoscience, Department of Chemistry and
Chemical Engineering, College of Environmental and Energy Engineering, Beijing
University of Technology, Beijing 100124, China
ABSTRACT
Perovskite-type oxides and perovskite-like oxides can tolerate substantial
substitutions in one or both cationic sites while maintaining their original crystal
structures. Such a feature renders the chemical tailoring of the materials via partial
replacement of the cationic site(s) with foreign metal ions, hence modifying their
physicochemical properties. Perovskite-type oxides and perovskite-like oxides have been
found many applications in physics and chemistry. The physicochemical properties of
these materials are dependent on the crystal structure, lattice defect, exposed lattice plane,
surface morphology, particle size, and specific surface area as well as the pore structure.
Up to now, a large number of perovskite-type oxides and perovskite-like oxides have
been generated and investigated for clarifying their physicochemical properties. In the
past years, although several books and reviews relevant to such a kind of materials
appeared in the literature, a systematic summarization on the newly published works
related to the fabrication, characterization, and catalytic applications of perovskite-type
oxides and perovskite-like oxides with the particles being nanosized, specifically
morphological or porous has not been available. Due to the unique properties of these
nano- and/or porous materials with low dimensionality, we shall summarize the results of
such functional solid materials recently reported in the literature, and aim to gain insights
into the relationship between structure and property and the potential applications in
catalysis. Therefore, we shall focus our attention in Section 2 on the controlled
*
Corresponding author: Tel.: +86-10-6739-6588, fax: +86-10-6739-1983; E-mail address: [email protected]
2 Jiguang Deng, Lei Zhang, Yuxi Liu et al.
fabrication of perovskite-type oxides and perovskite-like oxides with specific

morphologies, such as nano/microwires, nano/microrods, nano/microtubes,
nano/microfibers, nano/microcubes, nano/microspheres, nano/microflowers, and etc. Due
to the superiority in physicochemical property of porous materials to their bulk
counterparts, we shall also include the synthesis of macro- and/or mesoporous perovskite-
type oxides and perovskite-like oxides (although limited works have been done in the
aspect) in this section. In Section 3, we shall highlight the physicochemical properties of
perovskite-type oxides and perovskite-like oxides fabricated by the researchers
mentioned in Section 2, so that one can generalize some clear correlations between the
fabrication strategy and the crystal or porous structure or morphology. In Section 4, we
shall concentrate on the applications of these addressed materials in catalysis, such as
NOx elimination, CH4 complete oxidation, and volatile organic compounds removal, and
etc. Finally, we shall propose the perspectives and concluding remarks of the research in
the future in Section 5.
1. INTRODUCTION
Catalysts play an important role in modern chemical industry and environmental
protection [1]. The majority of the catalysts are based on single metal and/or metal oxides.
Among the mixed metal oxides, perovskite-type oxides and perovskite-like oxides occupy a
prominent status. They can tolerate substantial substitutions in one or both cationic sites while
maintaining the original crystal structures [24]. Such a feature renders the chemical tailoring
of the materials via partial replacement of the cationic site(s) with foreign metal ions, hence
modifying the physicochemical properties. Perovskite-type oxides and perovskite-related
oxides have been found many applications in physics and chemistry. The physicochemical
properties of these materials are dependent on the crystal structure, lattice defect, exposed
lattice plane, surface morphology, particle size, and specific surface area as well as the pore
structure. Up to now, a large number of perovskites have been generated and investigated for
clarifying their physicochemical properties. In the past years, although several books and
reviews relevant to such a kind of materials appeared in the literature [2,3,5,6], a systematic
summarization on the newly published works related to the fabrication, characterization, and
catalytic applications of perovskite materials with the particles being nanosized, specifically
morphological or porous has not been available. Due to the unique properties of these nano-
and/or porous materials with low dimensionality, we summarize the results of such functional
solids recently reported in the literature, and aim to gain insights into the relationship between
structure and property and the potential applications in catalysis. Therefore, we shall focus
our attention in Section 2 on the controlled fabrication of perovskite-type oxides and
perovskite-like oxides with specific morphologies, such as nano/microwires, nano/microrods,
nano/microtubes, nano/microfibers, nano/microcubes, nano/microspheres, nano/microflowers,
and etc. Owing to the superiority in physicochemical property of porous materials to their
bulk counterparts, we shall also include the synthesis of macro and/or mesoporous
perovskites (although fewer works have been done in the aspect) in this section. In Section 3,
we shall highlight the physicochemical properties of perovskites fabricated by the researchers
mentioned in Section 2, so that one can generalize some clear correlations between the
fabrication strategy and crystal or porous structure or morphology. In Section 4, we shall
Controlled Fabrication and Catalytic Applications 3
concentrate on the applications of these addressed materials in catalysis, such as NOx

elimination, CH4 combustion, and volatile organic compounds removal. Finally, we shall
propose the perspectives and concluding remarks of the research in the future in Section 5.
2. PREPARATION OF NANO/MICROSCALE PEROVSKITES
2.1. Nanoparticle
Nanocrystals exhibit physical and chemical properties that are distinct from their bulk
counterparts [7,8]. Controllable fabrication of nanometer perovskites is critical for
fundamental studies as well as for potential applications. In the past decade, a lot of methods,
such as novel solid-state method [8,9], sol-gel [1018], hydrothermal synthesis [19], co-
precipitation [20,21], reactive grinding method [22,23], microemulsion method [2426], and
nanocasting method [27], have been used to synthesize nanometer perovskites.
In the synthesis of titanate and zirconate nanomaterials, several groups have obtained
interesting results. By using barium titanium ethyl hexano-isopropoxide (BaTi(O2CC7H15)
[OCH(CH3)2]5) as precursor and oleic acid as stabilizing agent, OBrien and co-workers
fabricated monodispersed cubic BaTiO3 nanoparticles with diameters ranging from 6 to 12
nm (Figure 1) through a sol-gel process at 100140 oC in an argon or nitrogen atmosphere
[7]. Morse and co-workers reported a novel vapor-diffusion sol-gel method without the use of
a structure-directing template to generate well-defined cubic BaTiO3 nanocrystals (6 nm in
size) at low temperatures [28]. The nanocrystals are readily generated by the kinetically
controlled vapor diffusion of H2O(g)/HCl(g) into a non-aqueous solution of the bimetallic
alkoxide BaTi[OCH2CH(CH3)OCH3]6 at 16 oC. Via the interaction of alkaline earth metals
dissolved in benzyl alcohol with transition metal alkoxides at lower temperatures (200220
o
C), Niederberger et al. obtained well-grown cubic or tetragonal BaTiO3 and BaZrO3
nanoperovskites [29]. The average size of BaTiO3 nanoparticles is about 6 nm, whereas the
primary BaZrO3 particles are not isolated but form worm-like agglomerates with a diameter
of 23 nm and a length of up to 50 nm. With the aid of hydrothermal treatment at 140200 oC
under strong alkaline conditions, Wang et al. generated BaTiO3 nanoparticles (1720 nm in
diameter) [19]. Through a simple solid-state reaction between metal oxalates and anatase
TiO2 in a molten NaCl medium at 820 oC in the presence of a nonionic surfactant NP-9
(nonylphenyl ether), Mao and Wong produced the Ca1xSrxTiO3 (x = 01) nanomaterials with
a mean size of < 100 nm [8]. The substitution of smaller Ca2+ (radium = 0.134 nm) with
larger Sr2+ (radium = 0.144 nm) induces a reduction in symmetry from cubic to orthorhombic.
As we know, among the perovskites used as catalyst, LaMO3 (M = Co, Mn, Fe, Cr, Ni)-
based compounds have been received much attention due to their satisfactory catalytic
performance [2]. In the past years, our group have developed a novel strategy, citric acid
complexing-hydrothermal coupled method, to prepare nanoscale perovskites [30,31].
Figure 1. TEM micrographs and the SAED pattern (inset) of (a) a high-resolution lattice image of an
individual 8 nm diameter nanoparticle, (b) an ensemble of discrete BaTiO3 8 nm nanoparticles, (c) self-
organization of BaTiO3 nanoparticles, and (d) superlattice of 8 nm diameter BaTiO3 nanoparticles[7].
Using this method, we have obtained a number of highly active nanoperovskite catalysts,
including rhombohedral or orthorhombic La1xSrxMO3 (M = Co, Mn, Fe) and
La1xSrxM1yFeyO3 (M = Co, Fe). The typical preparation procedure of LaCoO3 is as follows:
lanthanum nitrate (0.01 mol), cobalt nitrate (0.01 mol), and a certain amount of citric acid
(metal/citric acid molar ratio = 2/1) are mixed in deionized water under constant stirring for
complete complexing of the metal ions. Then a certain amount of ammonia (28 wt%) solution
is added dropwise to adjust the pH value of the solution to 9.5 at which a sol is formed. The
sol is transferred to a 50-mL Teflon-lined stainless steel autoclave with a packed volume of
40 mL, and the autoclave is placed in an oven (for hydrothermal treatment) at 180 oC for 20 h.
Without filtering and washing, the obtained precursors are dried at 120 oC for 5 h. The dried
materials are well ground and calcined in air at a ramp of 1 oC/min in a muffle furnace to 400
o
C and kept at this temperature for 2 h, and then heated to 650 oC and remained at this
temperature for 2 h [32]. Shown in Figure 2 is the HRSEM image of the as-fabricated
LaCoO3 sample. The particles are relatively uniform in size with the typical diameter ranging
from 30 to 140 nm.
Figure 2. SEM image of LaCoO3 nanoparticles [32].
Figure 3. TEM image of the LaCoO3/SBA15 sample[39].
In order to overcome the drawback of lower surface areas of perovskite-type oxides, a

solid framework with a large surface area can be used to support them. For example,
supported LaCoO3 perovskites with 10 and 20 wt% loadings are fabricated by incipient
wetness impregnation of Ce1xZrxO2 (x = 00.3; 2338 m2/g) in a citrate solution [3335].
Recently, mesoporous silica materials (e.g., MCM-41 and SBA-15) with high surface area,
large pore volume, and narrow pore size distribution have been widely investigated as
support. Using an impregnation method with mixed LaCo citrate complex precursors, Nguyen
et al. loaded LaCoO3 on mesoporous silica [36], and Makshina et al. deposited LaCoOx into
the mesopores of MCM-41 [37,38]. Compared to MCM-41, SBA-15 with a well-ordered
mesopore structure is larger in pore size and higher in hydrothermal stability. Via a novel
microwave-assisted process with La-Co citrate complex as precursor, Cao and co-workers
embedded crystalline LaCoO3 nanoparticles in the mesopores of SBA-15 [39], as shown in
Figure 3. It is clearly seen that the LaCoO3 particles with an average size of 58.5 nm are
uniformly dispersed within the nanotubular channels of SBA-15. Recently, our group have
generated the yLaCoO3/SBA-15 (y = 1050 wt%) catalysts by an in-situ hydrothermal
method [40]. At a loading of less than 35 wt%, the LaCoO3 species can be highly dispersed
on SBA-15. Compared to the impregnation method, the in-situ hydrothermal strategy is more
suitable for the fabrication of well-dispersed LaCoO3 on SBA-15.
Figure 4. SEM images of LaNiO3 nanowires with the AAO template partly dissolved: (a, b) top view in
low magnification and (c) cross-section in high magnification[57].
2.2. One-Dimensional Nano/Microstructure
Owning to their unique physicochemical properties and potential applications in various

fields, one-dimensional (1D) nanostructured materials have been received great attention
[4144]. Considering the material diversity, cost, throughput, and large-scale production
potential, we here summarize the chemical fabrication of perovskite-type oxides with
morphologies of nano/microsized wires, rods, fibers, and tubes [45].
2.2.1. Hard-Templating Synthesis

Among the various methods, hard-templating synthesis is one of the straightforward
routes to obtain 1D nanostructural materials [4144,46]. The hard template adopted serves as
a scaffold, within or around which a different material is generated in-situ and shaped into a
nanostructure with its morphology complementary to that of the template. The advantage of
hard-templating strategy is that the shape and size of the product can be controlled by those of
the template. Considering the factors that (i) the pores of the hard template can be filled with
a variety of materials [47,48], (ii) the pores of the hard template are uniform and dense
(offering an opportunity to prepare monodisperse nanomaterials in high yield), and (iii) the
shape and size of the product can be controlled [41], one can employ the micro/nanoporous
hard templates (e.g., anodic aluminum oxide (AAO) , carbon nanotubes (CNTs), and porous
polycarbonate (PC) films [41,42,4953]) to fabricate 1D nanomaterials.
For the process using AAO as hard template, it is often combined with the sol-gel
method. The general procedure includes the following steps: (i) preparation of porous AAO
templates; (ii) fabrication of an appropriate concentration sol; (iii) filling the pores with the
precursor solution by dipping; (iv) calcination at a desired temperature; and (v) removing the
AAO template with a NaOH solution. Following the procedure, one can generate highly
ordered and densely packed polycrystalline rhombohedral LaCoO3 [54], orthorhombic
LaFeO3 [55,56], cubic LaNiO3 [57], BiFeO3 [58], monoclinic La1xCaxMnO3 [59], and
Pb(Zr0.53Ti0.47)O3 [60] nanowires. Shown in Figure 4 are the typical SEM images of the as-
prepared perovskite nanowires [57]. It can be seen that LaNiO3 nanowires are abundant,
uniform, and highly ordered in a large area. The length of LaNiO3 nanowires is about 50 m,
corresponding to the thickness of the AAO template adopted; the external diameter of the
nanowires is about 100 nm, equivalent to the pore diameter of the template employed.
Similarly, a series of perovskite nanotubes, such as LaNiO3 [61], BiFeO3 [62,63], monolithic
La0.6Sr0.4CoO3 [64], ATiO3 (A = Pb, Ba) [6568], Pb(Zr0.52Ti0.48)O3 [69], cubic Ba0.6Sr0.4TiO3
[70], and (Pb0.25Ba0.15Sr0.6)TiO3 [71] nanotubes, have also been fabricated. As shown in
Figure 5, tetragonal PbTiO3 nanotubes synthesized using the 200-nm AAO membrane as hard
template display a diameter of 200400 nm and length of 210 m [67]. Due to the
heterogeneity of the membrane used, the produced nanotube array is also inhomogeneous, as
can be seen from a side view (Figure 5b). Although the nanotubes are bundled together during
the drying process after the membrane removal, they can detach from each other easily by
sonication (Figure 5c and d). Nevertheless, because the only driving force in the process is the
capillary action, it is hard to pour the viscous sol in the nanochannels with a small pore
diameter [7274]. Although a rise in concentration of the sol has a promotional effect, too
high a concentration will result in either a viscous sol (causing the filling of the pores to be
difficult) or destabilization of the sol [75]. In order to overcome this drawback, several
strategies, such as spin-coating technique [76], electrophoretic deposition [77,78], and liquid-
phase deposition [79], have been developed to facilitate the sols to enter the pores. Hu et al.
fabricated pseudocubic LaMnO3 nanowires by means of a novel technique, i.e., through the
converse diffusion or convection process, metal ions and precipitation agents were first
transported in the nanochannels of the AAO template, and then reacted in-situ to generate a
1D nanostructure of the LaMnO3 perovskite [72].
Similar to AAO, CNTs with at least one open end can also be filled with liquids (molten
salts or solutions) with the aid of capillary action [8082]. However, it is hard to achieve
complete filling of CNTs with a high yield of the desired product due to its small diameter.
Moreover, CNTs in water or even in an organic solvent is apt to aggregate, hindering their
further manipulation. Recently, Gao et al. found that the addition of the surfactant P123
(poly(ethylene glycol)-block-poly(propylene glycol)-block-poly(ethylene glycol), EO20-
PO70EO20) and H2O2 could disperse the CNTs thoroughly and homogeneously in the
hydrothermal fabrication of La2CuO4 nanofibers with single-walled carbon nanotubes
(SWNTs) as the template [8385]. The formation of a perovskite structure often requires
high-temperature (> 650 oC) calcination [31], but CNTs can be completely oxidized to CO2 in
air at about 650 oC [86]. In order to obtain 1D perovskite nanomaterials by CNTs-templating
methods, it is necessary to add some reagents (such as citrate) to reduce the crystallization
temperature of the catalyst. For example, Zhu and co-workers generated single-phase
polycrystalline rhomhohedral LaCoO3 and LaSrCuO4 nanowires with CNTs as template in
the presence of citrate at 600 oC [87,88].
Figure 5. SEM images of nanotubes synthesized in 200 nm membranes: (a) nanotube ends, (b) side
view revealing growth inhomogeneties, (c) the nanotubes bundled during drying after dissolving the
membrane, and (d) Free-standing nanotubes after sonication[67].
In addition to AAO and CNTs, porous PC films have also been adopted to fabricate 1D
perovskite-type oxides. Leyva and co-workers prepared La0.67Ca0.33MnO3 [89], La0.325-
Pr0.300Ca0.375MnO3 [90,91], and La1xSrxCo1yFeyO3 [92] nanotubes via a porous PC film-
templating route in an adequate precursor system for syringe filtration. The typical SEM
image of La0.6Sr0.4CoO3 nanotubes obtained with porous PC (diameter = 800 nm) as template
[92] is shown in Figure 6. Clearly, the shape and size of the perovskite tubes are determined
by the characteristics of the template pores. The typical dimensions of the tubes are 770 nm in
diameter, 80 nm in wall thickness, and 34 m in length. As mentioned above,
electrophoretic deposition is one of effective approaches to overcome the problems
encountered in the direct sol-filling method. In the former method, an applied electric field
can induce electrophoretic motion of the nanoclusters in the sol, and pull charged
nanoclusters into the template pores, favoring the formation of a dense nanostructure (Figure
7) [75,93]. By means of a combination of sol-gel processing and electrophoretic deposition,
Cao and co-workers obtained Pb(Zr0.52Ti0.48)O3 nanorods with the PC membranes as template
[75,93]. However, since the force from the electric field is determined by the zeta-potential
magnitude of the nanoclusters, acidic or alkaline sols are usually required for the generation
of a good nanostructure. Moreover, a small absolute value of the zeta potential may result in
the formation of a poorly-structured product. Hence this method is unfavorable for the
fabrication of biologically active materials that are sensitive to acidic or alkaline
environments. In view of this consideration, the same group has developed a novel strategy,
i.e. sol-gel processing coupled with centrifugation. With the assistance of adjustable
centrifugation force, sol nanoclusters can be effectively pulled into the template pores. The
so-obtained Pb(Zr0.52Ti0.48)O3 nanorods show a dense and uniform structure (Figure 8).
Furthermore, such a method is suitable for a variety of sol systems, including proteins or
enzymes that may be sensitive to highly acidic or alkaline environments [74].
Figure 6. SEM image of the La0.6Sr0.4CoO3 nanotubes[92].

Figure 7. Schematic illustration of the progression of the growth process. Positively charged sol
particles are moving electrophoretically towards the negative electrode, depositing at the bottom of the
pore, while the negatively charged counter ions are moving in the opposite direction[75].
Figure 8. SEM images of the top view (a) and side view (b) of Pb(Zr0.52Ti0.48)O3 nanorods grown from
the precursor sol by centrifugation at 1400 rpm for 1 h[74].
Figure 9. Schematic illustrations showing the formation of nanowires by templating against

mesostructures self-assembled from surfactant molecules: (A) formation of a cylindrical micelle; (B)
formation of the desired material in the aqueous phase encapsulated by the cylindrical micelle; (C)
removal of the surfactant molecules with an appropriate solvent (or by calcination) to obtain an
individual nanowire; and (DF) similar to the process illustrated in (AC), except that the exterior
surface of an inverted micelle serves as the physical template[42].
As we know, carbon nanofibers have been widely used in various areas, such as in
electronic components, as polymer additives, for gas storage, and as catalyst support [94]. Via
the decomposition of hydrocarbons or carbon monoxide over Ni, Co or Fe catalysts, a large
quantity of carbon fibers can be produced. Moreover, these carbon fibers with different sizes
and shapes, such as straight, bent, thin and helical, etc., could be synthesized by controlling
the reaction conditions [9599]. With carbon nanofibers as template, Ueda and co-workers
prepared a number of single and binary transition metal oxide nanofibers and/or nanotubes,
including LaMnO3 and LaFeO3 nanofibers and nanotubes [99102].
Although hard-templating synthesis provides a simple and versatile pathway to fabricate
1D perovskite nanomaterials, the quantity produced in each run is relatively small. Removal
of the template through a post-synthesis process may also cause damage to the product.
Furthermore, the so-obtained samples are often polycrystalline, which may limit their use in
device fabrication and fundamental studies [42]. Therefore, it is highly demanded to develop
alterative methods that are effective to synthesize 1D perovskite nanomaterials.
2.2.2. Surfactant-Templating Synthesis

Surfactant-templating method has been widely used to synthesize nanoscale materials.
Up to now, most of the 1D nanomaterials prepared by surfactant-templating method are single
metal and metal oxides, and only a few works deal with the fabrication of 1D mixed metal
oxides via this pathway [103105]. The principle of surfactant-templating synthesis is that
when the concentration of the surfactant molecules reaches a critical value, they will
spontaneously organize into rod-shaped or inverse micelles (Figure 9), which can be
immediately used as a soft template to promote the formation of 1D nanomaterials [42].
Based on this principle, Xu et al. [106] synthesized tetragonal PbZr0.52Ti0.48O3 nanorods
hydrothermally at 200 oC for 12 h by using tetrabutyl titanate, zirconium oxychloride, and
lead nitrate as starting materials, potassium hydroxide as mineralizer, and poly(vinyl alcohol)
(PVA) as surfactant. The surfactant plays an important role in the preferential nucleation and
growth of the final product. Without PVA addition, the as-prepared particles are cubic in
shape [107]. As indicated by the arrows in Figure 10, there are much more transparent films
of amorphous PVA gel attached to the sides than to the head of the PbZr0.52Ti0.48O3 nanorods.
It provides a direct evidence for the templating role of a surfactant. In addition, the coupling
of sol-gel process with surfactant-templating method can give rise to single-crystalline
BaTiO3 nanorods [103].
Compared to the preparation of 1D single metal or metal oxides with a surfactant as
template, there are much more difficulties in coordinating the interaction between the
surfactant molecules and the A-/B-site metal ions for the generation of the stoichiometric
product ABO3. Moreover, the preparation and removal of the micellar phase is often difficult
and tedious. All of the problems encountered in the surfactant-templating process limit its
wide applications in fabricating 1D perovskites.
Figure 10. (a) TEM image of the PbZr0.52Ti0.48O3 nanorods prepared by the hydrothermal method at 200
o
C for 12 h assisted only by PVA; and (b) magnification of (a) [106].
(a) (b) (c)
400 nm 60 nm 1 m
(d) (e) (f) 4 nm

0.275 nm
0.275 nm
(110)
(110)
200 nm 5 nm 4 nm
Figure 11. HRSEM (a, c), HRTEM (b, d, e, f) images and SAED patterns (insets) of the
La0.6Sr0.4CoO3 sample calcined (a, b, e) at 650 oC and (c, d, f) at 650 oC[116].
Figure 12. TEM images of (a) TiO2, (b) BaTiO3, and (c) SrTiO3 nanotubes. Inset in (b) and (c) show
higher magnification images (50 nm 50 nm in (b) and 40 nm 60 nm in (c)) of single tubes and
electron diffraction patterns, respectively[132].
2.2.3. Hydrothermal Synthesis

In the past several years, an important advancement has been achieved in the synthesis of
1D nanomaterials via hydrothermal treatments [19,42,44,108]. The typical procedure is that
precursors and proper agents capable of regulating or templating the crystal growth are added
to distilled water with appropriate ratios, and then the mixture is placed in an autoclave for
the crystallization at an elevated temperature, time, and pressure. The major advantage of this
technique is that lots of materials can be made soluble in water by heating and pressurizing
the system close to the critical points [42]. As a result, a number of raw materials are usable
to synthesize the desired products by means of this approach. It has been reported that, after
hydrothermal treatment at a temperature (120275 oC) for a desired time (3 h7 day) and
without the assistance of a soft template (surfactant), one can directly generate a series of 1D
single-crystalline perovskite nanomaterials, including cubic MTiO3 (M = Ba, Sr, Pb)
[109112], MNbO3 (M = Na, K) [113115], cubic La0.6Sr0.4CoO3 [116], and Ag2MO4 (M =
Cr, Mo, W) [117] nanowires, and Zn2SnO4 [118] nanorods. Representative HRSEM and
HRTEM images as well as SAED patterns of the hydrothermally derived cubic
La0.6Sr0.4CoO3 nanowires/nanorods are shown in Figure 11 [116]. It is observed that the
sample calcined at 650 oC consists of a large quantity of uniform nanowires (typical diameter
= 1530 nm) (Figure 11a and b). A mixture of nanorods and nanowires (Figure 11c and d) is
obtained after calcination at 800 oC, and the typical diameter of the nanorods and nanowires
fall into the range of 1535 and 2085 nm, respectively. The corresponding SAED patterns
(insets) show regularly aligned lines of bright spots, disclosing that the La0.6Sr0.4CoO3
nanowires and nanorods are well-grown single crystallites. According to the HRTEM
photographs (Figure 11e and f), the d spacing of the (110) plane is estimated to be 0.275 nm,
not far away from the corresponding value (0.271 nm) of polycrystalline LaCoO3 (JCPDS
No.: 75-0279). It should be pointed out that all of the 1D single-crystalline perovskite
nanomaterials (except for Ag2MO4 (M = Cr, Mo, W)) mentioned above are hydrothermally
fabricated in the alkaline environment with NaOH, KOH, NH3H2O or N2H4H2O as
mineralizer. The pH value of the reactant solution plays an important role in formatting the
1D nanostructure, which will be discussed later. Based on the results of OBriens group
[119], Park and co-workers reported the synthesis of single-crystalline cubic BaTiO3 and
SrTiO3 nanorods through the hydrothermal decomposition of bimetallic alkoxide precursors
(BaTi(OCH(CH3)2)6 and SrTi(OCH(CH3)2)6) at 280 oC for 6 h [120122]. However, the
preparation of such bimetallic alkoxide precursors is rather difficult and expensive. Therefore,
a wide application of this approach is greatly limited. Other groups have also claimed to
successfully fabricate single-crystal perovskite La0.5A0.5MnO3 (A = Sr, Ba, Ca) nanowires
hydrothermally [123126], however, the very recently reported results reveal that these
nanowires are actually wire-like La(OH)3 nanocrystals, rather than the perovskite
nanocrystals [127131].
It is interesting that there are works on the hydrothermal generation of 1D titanate
perovskites with 1D TiO2 and Na2Ti3O7 nanoparticles as sacrificial template. For example,
cubic ATiO3 (A = Ba, Sr, Pb) nanotubes (Figure 12) are hydrothermally synthesized using the
TiO2 nanotubes as precursor [132,133], in which the BaTiO3 and SrTiO3 nanotubes are 815
nm in outer diameter, 47 nm in inner diameter, and 50500 nm in length (Figure 12). By
utilizing the highly active TiO2 nanofibers as raw material, orthorhombic CaTiO3 microtubes,
cubic SrTiO3 and BaTiO3 microfibers can be generated hydrothermally in basic solutions at
150 oC for 24 h. The formation of microtubular CaTiO3 involves the initial dissolution of the
TiO2 nanofibers and the conversion to microsized fiber-bundles after an Ostwald ripening
process, while the fabrication of SrTiO3 and BaTiO3 microfibers is associated with the ion
exchange reaction and in-situ growth process at the TiO2 nanofiber framework (Figure 13)
[134]. A variety of single-crystalline nanostructured BaTiO3 with morphologies of nanowire-,
nanosheet-, nanocube-, coral-, starfish-, and sword-like shapes as well as hexagonal
nanoparticles (Figure 14) have been produced hydrothermally employing the Na2Ti3O7
nanotubes and nanowires as precursor via the control of reaction temperature and
concentration of Ba(OH)2 [135]. Kang et al. observed the formation of single-crystalline

BaTiO3 nanorods from the K2Ti4O9 nanorod/wire precursor, the former aggregated into
nanoparticles with a rise in hydrothermal time from 42 to 72 h [136]. Although only 1D
titanates mentioned above have so far been successfully obtained through a hydrothermal
route, this strategy could be applied to fabricate other 1D perovskites in the future. Moreover,
taking into account the fact that the controllable synthesis techniques of 1D single metal
oxides have gained a great progress in the last decade [19,41,42,44,137,138], it is possible to
generate 1D or other specifically morphological perovskites (other than titanates) by
templating against 1D oxide of the A- and/or B-site metal(s). Furthermore, it is even likely to
obtain 1D multiphase-paragenous compounds with a template-cored and perovskite-shelled
structure. Such symbiotic materials may show interesting physicochemical properties as
compared to the single phase of template or perovskite.
In addition, hydrothermal method coupled with other techniques has been received great
attention in preparing 1D perovskites [139142]. For example, by adopting sol-gel and
hydrothermal coupled method, Hou and co-workers successfully synthesized single-
crystalline K0.5Bi0.5TiO3 and (Na0.8K0.2)0.5Bi0.5TiO3 nanowires [140,143]. Compared to the
conventional hydrothermal method, the microwave-hydrothermal combined process allows
the internal heat generated from the microwave irradiation to heat the reactants very rapidly
for crystallization, thus saving time and energy. By using this coupled strategy, single-
crystalline PbTiO3 nanowires can be readily produced [142].
Figure 13. Schematic illustration of the possible formation mechanism of MTiO3 (M = Ca, Sr and Ba)
perovskites with various morphologies[134].
Figure 14. Influence of the concentration of Ba(OH)2 and the hydrothermal temperature on the
morphology and structure of BaTiO3 products prepared from (a) Na2Ti3O7 nanotubes and (b) Na2Ti3O7
nanowires[135].
(a) (b)
(c)
Figure 15. (a) TEM, (b) HRTEM, and (c) SEM images of the BaTiO3 nanowires. Inset of (b) is the
SAED pattern of a BaTiO3 nanowire[145].
(a) (b)
(c) (d)
Figure 16. TEM images of (a) BaMnO3 at 200 oC for 120 h, (b) BaTi0.5Mn0.5O3 at 200 oC for 48 h, (c)
BaMnO3 at 170 oC for 72 h, and (d) BaMnO3 at 200 oC for 120 h, adding 1 mL of water[151].
Although a number of perovskite nanowires and nanorods can be generated via the
hydrothermal route, there are still no reports on the fabrication of micro/nanotubular
perovskites (except for titanites) by means of this method. Moreover, since a lot of
parameters, such as hydrothermal temperature and time, pressure, precursor nature, base
strength, and solvent, etc., can influence the physicochemical properties (including the
morphology) of the final product, further work on optimizing this preparation strategy is
highly needed to achieve controlled fabrication of 1D perovskites in the future.
2.2.4. Novel Solid-State Method

In recent years, a great progress in solution synthesis of 1D perovskites has been
achieved. To extend the method for 1D perovskite preparation, the solid-state process has
been developed. In 1986, Hayashi et al. reported the generation of rod-shaped BaTiO3 from
rod-like TiO2nH2O and BaCO3 in molten chloride [144]. With barium oxalate and TiO2
(anatase) as precursor and in the presence of NaCl and nonionic surfactant NP-9, Mao et al.
obtained single-crystalline cubic BaTiO3 after calcination of the well-grounded reactant
intermediate mixture at 820 oC for 3.5 h [145]. The BaTiO3 particles mainly possess straight,
smooth, and crystalline wire-like morphologies (Figure 15), in which the nanowires are
5080 nm in diameter and 1.510 m in length. Very recently, by employing the mixed
chlorides (NaCl/KCl molar ratio = 1:1) instead of NaCl as reaction medium, 1D MTiO3 (M =
Ba, Sr, Pb) nanostrips were generated at 950 oC for 5 h without the assistance of any
surfactants [146]. With La2O3, Sr(NO3)2, and a Mn(NO3)2 solution ( 50 wt%) as starting
materials, Feng and co-workers prepared La0.7Sr0.3MnO3 nanorods by means of a combined
strategy of sol-gel and molten salt synthesis in the NaCl-KCl (mass ratio = 1:4) flux molten
salts [147]. The sol-gel treatment step can guarantee the formation of homogeneous precursor
mixture. In addition, the authors also synthesized La0.7Sr0.3MnO3 nanorods by adopting the
two steps: (i) KMnO4 and MnCl2 (molar ratio = 2:3) are well mixed and grounded to prepare
a precursor of MnO2 fibers, and (ii) stoichiometric amounts of La2O3, SrCl2, and the fibrous
MnO2 precursor are well mixed and grounded with 1 mL of ethanol, then dried at 70 oC, and
finally calcined at 950 oC for 5 h [148,149]. It is believed that the MnO2 fibers function as the
crystal seed in the growth of La0.7Sr0.3MnO3 nanorods.
(a)
(b)
Figure 17. (a) TEM and (b) HRTEM images as well as the SAED pattern (inset) of La0.5Ba0.5MnO3
nanocubes prepared at 300 oC for 24 h[127].
An approach, called the composite-hydroxide-mediated method, was developed by Wang

and co-workers. It has been employed to synthesize single-crystalline hexagonal BaMnO3
nanorods, BaTiO3, Ba1xSrxTiO3, and Fe3O4 nanocubes, BaTi0.5Mn0.5O3, CoFe2O4, and
Pb2V2O7 nanobelts, and BaCeO3 nanoflowers [150152]. This method is based on the
reaction between a metallic salt and a metal oxide in a molten mixed solution of NaOH and
KOH (molar ratio = 51.5:48.5) at 200 oC. One of the key factors for the successful synthesis
of the above morphological materials at 200 oC or lower temperatures is that an eutectic point
as low as 165 oC can be achieved in the alkali molten solution at a NaOH/KOH molar ratio of
51.5:48.5 [150], although the melting points of pure KOH and NaOH are much higher (360
and 323 oC, respectively). The possible mechanism involving in the formation of perovskite
phase (take BaTixMn1xO3 as an example) can be described as the following:
BaCl2 + NaOH Ba(OH)2 + NaCl

xTiO2 + (1x)MnO2 + 2NaOH Na2TixMn1xO3 + H2O
Ba(OH)2 + Na2TixMn1xO3 BaTixMn1xO3 + 2NaOH
Due to the big viscosity of hydroxide, formation of a perovskite nanostructure is slow,

thus minimizing the extent of particle agglomeration. It is the reason why dispersive single-
crystalline nanostructured perovskites could be obtained during the reaction without the use
of a surface capping agent, as shown in Figure 16. The BaMnO3 nanorods obtained at
170200 oC for 72120 h display a width of 50100 nm and a length of 2001000 nm
(Figure 16a and c). The morphology of BaTi0.5Mn0.5O3 is different from that of BaMnO3. The
particles of the former material show an architecture of elliptical nanobelts with about 40 nm
in width, 20 nm in thickness, and 500 nm in length (Figure 16b). It should be noted that such
a fabrication approach is carried out in the absence of water. The addition of little (1 mL)
water to the reaction system would change the morphology of the product from nanorods to
nanoparticles (Figure 16d) [151].
2.2.5. Other Methods

In addition to the above fabrication strategies, there are other methods to produce 1D
perovskites. For example, single-crystalline tetragonal Ba1xSrxTiO3 nanorods can be obtained
via a simple ultrasonic radiation and sol-gel combined approach [153]. The electrospinning
technique has been used to generate single-crystalline tetragonal PbTiO3 nanowires [154],
polycrystalline Pb(Zr0.52Ti0.48)O3 [155], cubic Ba0.6Sr0.4TiO3 [156], and BaTiO3 [157]
nanofibers. Zhou and co-workers developed a pulsed laser deposition strategy to produce
novel transition metal oxide core-shell nanowires, including MgO/La0.67Ca0.33MnO3,
MgO/La0.67Sr0.33MnO3, MgO/YBa2Cu3O6.66, MgO/PbZr0.58Ti0.42O3, and MgO/Fe3O4 [158
160]. With the ion-exchangeable titanate nanotubes as starting material, thin films of single-
crystalline SrTiO3 nanorod arrays can be formed by annealing the Sr2+-introduced titanate
nanotubes [161]. Alkalides are crystalline ionic salts consisting of crown ether or cryptand
complexed alkaki metal cations. When being dissolved in nonreducible solvents, the alkalides
can give alkaki metal cations, which are nearly as a thermodynamically powerful reducer as a
solvated electron, the most powerful reducer possible in any given solvent, and are capable of
transferring two electrons simultaneously. Alkalide reduction of metal salts leads to the
generation of a nanoscale colloid (215 nm in diameter). Through such an alkalide reduction,
Wagner and co-workers generated single-crystalline orthorhombic NaTaO3 nanorods [162].

Highly anisotropic pseudo-cubic KNbO3 nanocrystallites with morphologies of nanoneedles
and nanoplatelets can be obtained after the crystallization of the precursor gel derived from a
polymerized complex process, which is based on the Pechini-type reaction route that starts
from a precursor solution of the respective metal-citrate complexes [163].
2.3. Nano/Microcube
It is well known that optical, magnetic, electronic, and catalytic properties of a material
are dependent on its particle size and shape [164166]. Owning to the unique properties
arising from the exposing well-defined crystal planes and the advantages in serving as a
building block for further construction of multifunctional devices [167], cube-shaped
nano/microcrystals have received great interests. Up to now, a number of works have been
devoted to synthesizing the perovskites with a cube-like morphology, among which mixed
valence manganites have been paid considerable attention due to their colossal
magnetoresistivity (CMR) [168,169]. In the past years, our group [130,131,170] and several
other groups [124,127,128,170175] have developed a facile hydrothermal strategy to
prepare the single-crystalline manganite A1xA'xMnO3 (A = La, Pr, Nd; A' = Ca, Sr, Ba; x =
0.30.6) micro/nanocubes without calcination at high temperatures. The hydrothermal
synthesis, which involves the reaction between MnO4 and Mn2+ in the presence of lanthanum
and alkaline earth metal salts to give the desired manganese oxidation state, is carried out at a
temperature of 220300 oC for 2472 h in a concentrated KOH solution. Shown in Figure 17
are the representative TEM images and the SAED pattern of the pseudocubic La0.5Ba0.5MnO3
sample obtained hydrothermally at 300 oC for 24 h [127]. Obviously, the products are single
crystalline and exhibit a nanocube-like (50100 nm in edge length) morphology. As
mentioned above, reaction parameters, such as solution pH value, hydrothermal temperature
and time as well as pressure, have great influence on the particle size and morphology of the
final product. In other words, it is possible to controllably grow a crystal with the desired
particle size and shape by carefully adjusting the above experimental parameters. A typical
example is from the work done by Chai et al. [173]. They successfully synthesized single-
crystalline La0.5Ba0.5MnO3 with flower- and micro/nanocube-like shapes by controlling the
reaction conditions, especially the concentration of KOH (Table 1 and Figure 18). During the
preparation process of the A1xA'xMnO3 micro/nanocubes, KOH actually took part in the
reaction leading to perovskite structure formation. Hence, KOH concentration may influence
or even change the growth mechanism [173,176], and further alter the shape and size of the
final product.
Table 1. The influence of fabrication parameter on the morphology of La0.5Ba0.5MnO3

[173]
Sample KOH Filling Hydrothermal Hydrothermal Morphology

no. concentration fraction temperature (oC) time (h)
(mol/L) (vol%)
1 6 60 270 25 Layered flower
2 10 60 270 25 Microcube
3 14 60 270 25 Nanocube
4 6 40 270 25 Nonlayered
flower
5 6 80 270 25 Irregular
6 6 60 250 25 Nonlayered
flower
7 6 60 230 25 Amorphous
(a) (b) (c)
Figure 18. SEM images of La0.5Ba0.5MnO3 prepared at 270 oC for 50 h under (a) 6 mol/L, (b) 10 mol/L,
and (c) 14 mol/L KOH conditions[173].
Figure 19. (a) Typical SEM image of the CaSn(OH)6 precursor prepared by using 2 mL PVP; (b)
typical TEM image of the CaSnO3 product after heat treatment of the precursor as shown in (a) at 500
o
C for 5 h[181].
Figure 20. SEM images of (a) CaTiO3 microcubes prepared at 180 oC for 15 h, (b) two individual
CaTiO3 cubes, (c) a broken cube showing hollowed interior, (d) powder XRD patter of the product
compared with reference to CaTiO3 JCPDS card no. 82-0229[182].
In addition to A1xA'xMnO3, LaFeO3 [177] and K1xNaxNbO3 (x = 01) [114,178]

micro/nanocubes can also be fabricated hydrothermally in the absence of a surfactant.
Although many single metal and metal oxide nanoparticles with a cube-like morphology have
been prepared with the adding of a surfactant to control the growth process [167,179], only a
few cube-like perovskites were generated with the aid of a surfactant [180,181]. For example,
in the presence of ployvinylpyrrolidone (PVP), Li and co-workers obtained orthorhombic
CaSnO3 micro/nanocubes after hydrothermal tretament at 140 oC for 10 h [181]. The PVP
acts as a shape-directing agent. It is preferentially adsorbed on different crystal faces of the
precursor, thus favoring formation of the cube-like morphological product. The SEM image
of the CaSn(OH)6 precursor obtained using 2 mL of PVP is shown in Figure 19. Apparently,
all of the particles are uniform both in shape (cube) and size (~1 m in edge length). The
uniform CaSnO3 micro/nanocubes can be readily generated after calcination of the
CaSn(OH)6 precursor at 500 oC for 5 h. The challenge of such a surfactant-assisted
hydrothermal strategy is to properly regulate the interaction between the surfactant molecules
and the A- and B-site metal ions in ABO3, and hence to control the composition,
stoichiometry, and crystal structure of the ABO3 material. It deserves further investigations.
Interestingly, by adopting a solvothermal process coupled with the surfactant-directing

technique in a poly(ethylene glycol) solvent, Yang et al. fabricated single-crystalline
orthorhombic CaTiO3 hollow microcubes (Figure 20) [182]. Furthermore, the authors also
observed that the morphology of the as-prepared product is highly sensitive to the crystal
growth condition. With an increase in water amount used, the morphology of the product
changes from hollow microcube to prismatic rectangular rod, and from butterfly-shaped
crystal to irregularly shaped crystal.
Compared to the conventional solid-state reaction, molten salt synthesis is believed to be
one of the simplest, most versatile, and highly cost-effective approaches available for
producing functional materials at lower temperatures [145,147149,183,184]. Wong and co-
workers have made many efforts in preparing perovskites with various morphologies via the
molten salt synthesis route [8,145,183,185]. They have shown that parameters, such as nature
and amount of salt, surfactant, reaction temperature and duration, precursor type and molar
ratio, heating/cooling rate, exert influence on the purity, particle size and shape, and
morphology of the product. Moreover, due to the effects of solubility and reactivity
associated with the salt, which may greatly alter the synthesis process and the resultant
particle size and shape, selection of the salt is a key issue. Through adjusting the preparation
parameters properly, one can generate cubic SrTiO3 [145] and cubic BaZrO3 [183,185]
nanocubes and nanospheres, BaTiO3 nanowires [145], and Ca1xSrxTiO3 nanoparticles [8]. As
mentioned in Section 2.2.4, single-crystalline BaTiO3 and BaxSr1xTiO3 nanocubes (Figure
21) can be obtained via the composite-hydroxide-mediated approach [150]. Very recently, by
using the titanium foils and BaCO3 instead of TiO2 and BaCl2 as precursor, He et al. [186]
synthesized single-crystalline cubic BaTiO3 nanocubes using the same method. As compared
with the comprehensive investigations on shape control in liquid-phase reactions [179,187],
however, fewer works have been done on particle shape control by means of the mentioned
novel solid-state process. Therefore, more efforts are needed to put on achieving the rational
synthesis of desired products by this technique.
2.4. Other Specific Morphologies
Perovskites with other morphologies, such as octahedral and dodecahedral, solid and
hollow spherical, dendritic, bur-, multipod-, and star-like shapes have also been fabricated.
By regulating the reaction conditions (e.g., hydrothermal temperature and time, reactant
concentration, and solvent constitution), single-phase perovskite BaZrO3 microcrystals with
controllable shapes of truncated rhombic dodecahedron and spheres (Figure 22) can be
synthesized [188]. It is found that the solvent polarity, which is governed by the solvent
composition, plays an important role in determining the geometric structure of the product. A
rise in the volume ratio of ethanol to water from 0 to 3 can bring about the evolution of
morphology from truncated dodecahedral to spherical (Figure 22d). Based on the inherent
properties of the templates, many methods have been proposed for the synthesis of hollow
spherical materials. For example, by templating against amorphous TiO2 spheres, Zhang and
co-workers have prepared monodispersed porous cubic SrTiO3 and BaTiO3 spheres after
hydrothermal treatment at 180 oC for 12 h [189]. Since the adopted template is amorphous, a
relatively high-temperature treatment is required to transform the amorphous oxide shells into
crystallites. However, such a processing step might result in deterioration of the shell and
tendency for perforation and fragmentation. In view of this, Yang and co-workers tried to
prepare hollow spherical BaTiO3 and SrTiO3 by employing the sulfonated polystyrene (PS)
spherical gel as template [190]. The typical procedure includes (i) the preparation of
sulfonated PS gel spheres, (ii) the adsorption of tetrabutyl titanate (TBT) by dispersing the as-
prepared template into a mixture of absolute ethanol and TBT (1:1 v/v, 2 mL) for 24 h, and
(iii) the treatment with Ba(OH)2 or Sr(OH)2 (0.3 mol/L, 4 mL) at 70 oC for 12 h under
stirring. Because the preparation is carried out under mild conditions, intact shells are
guaranteed, as shown in Figure 23. Using sodium dodecylbenzene sulfonate (SDBS) or
poly(sodium 4-styrenesulfonate) (PSS) as surfactant, bur-like three-dimensionally
hierarchical PbTiO3 nanostructures consisting of a microsphere core with an outer shell of
nanorods (Figure 24) are obtained by in-situ self-assemblying of nanoparticles under
hydrothermal conditions [191]. As can be observed from the SEM images (Figure 24), the
microspheres are 15 m in diameter and the nanorods are 50100 nm in diameter with a
length in the range from hundreds of nanometers to 2 m. The possible mechanism for
formation of the tetragonal bur-like PbTiO3 nanoparticles is illustrated in Figure 25. It can be
seen that only part of the generated PbTiO3 nanoparticles directly aggregate into microsphere,
whereas some of them become nonspherical building blocks on the microspheres and then
ripen to nanorods.
As mentioned in Section 2.2.3, the microwave-hydrothermal method is one of the
versatile and cost-effective approaches to fabricate crystalline and pure micro/nanoscale
materials in shorter reaction time at lower temperatures. By means of the microwave-
hydrothermal method, microsized decaoctahedral cubic BaZrO3 powders (Figure 26) are
firstly synthesized at 140 oC for 40 min [192]. By using a molten K2CO3 flux, Loye and co-
workers synthesized single crystals of RERhO3 (RE = La, Pr, Nd, Sm, Eu, Tb) for the first
time [193]. All of the compounds are found to crystallize in orthorhombic space group Pbnm,
which is confirmed by the truncated octahedral geometry of the LaRhO3 single crystal. Since
the multipod-shaped crystals are only synthesized under non-equilibrium conditions, it is hard
to fabricate perovskites with a multipod shape by conventional methods. In order to overcome
these problems, Toshima et al. developed a novel strategy, i.e., the self-propagating high-
temperature synthesis method, for the production of multipod cubic SrTiO3 crystals by using
titanium powder, NaClO4, and SrCO3 as the raw materials [194]. The key step for successful
fabrication is the choosing of irregularly shaped titanium powders. Using spherical powders
with smooth surface as the Ti source, cube- and octahedron-shaped SrTiO3 crystals are
obtained (Figure 27). However, hexapod- and decapod-shaped SrTiO3 crystals (Figure 28) are
generated by employing irregularly shaped Ti powders. Very recently, the preparation of
dendritic BaTiO3 cubic single crystallites has been reported first by Wang et al. [195]. Shown
in Figure 29 are the SEM images of the hydrothermally derived BaTiO3 products that consist
of homogeneous dendritic structures. It is found that the KOH concentration plays a key role
in the formation of BaTiO3 dendrites. With a decrease in KOH concentration from 1.0 to 0.7
mol/L, and then to 0.1 mol/L, the particle morphology changes from spherical to star-shaped,
and then to dendritic, accordingly. In addition, periodic arrays of nanorings of morphotropic
phase boundary lead zirconium titanate can be fabricated using a complicated physical
technique [196].
Figure 21. (AC) BaTiO3 and (DF) BaxSr1xTiO3 nanocubes synthesized by the composite-hydroxide-
mediated approach. (A, D) XRD patterns of BaTiO3 and BaxSr1xTiO3 nanopowder. (B, E) SEM images
of BaTiO3 and BaxSr1xTiO3 nanocubes; the inset is the EDS pattern. (C,F) TEM images of BaTiO3 and
BaxSr1xTiO3 nanocubes; the insets are the SAED pattern and HRTEM images of a nanocube, showing
its single-crystal structure[150].
Figure 22. (a, b) SEM and (c) TEM images of slightly truncated rhombic dodecahedral BaZrO3
microcrystals. The arrow shown in (b) is indication of truncated corners. (d) Simulated morphology
evolution process of the three states of regular rhombic dodecahedron, truncated dodecahedron and
sphere[188].
Figure 23. (a, d) SEM and (b, e) TEM images of (a, b) BaTiO3/PS and (d, e) SrTiO3/PS composite
hollow spheres; inset of (b): microtome image. (c, f) HRTEM images of a typical particle on the shell
of (c) BaTiO3/PS and (f) SrTiO3/PS composite hollow spheres; inset of (c, f): SAED patterns. (g, h)
SEM and TEM (inset) images of (g) BaTiO3 and (h) SrTiO3 hollow spheres obtained after BaTiO3/PS
and SrTiO3/PS composite hollow spheres were calcined in air at 450 oC[190].
Figure 24. SEM images of PbTiO3 nanostructures prepared by the hydrothermal method from Pb-Ti-
citrate sol for 48 h. (ad) With different SDBS concentration at 180 oC: (a, b) 0.1 mol/L, (c) 0.2 mol/L,
and (d) 0.3 mol/L. (e) Without surfactant at 180 oC. (f, g) With 0.2 mol/L SDBS concentration at
different temperatures, (f) 150 oC and (g) 200 oC. (h) 0.2 mol/L PSS concentration at 180 oC. Insets in
images are the selected parts imaged at higher magnification[191].
Figure 25. Schematic mechanism for the formation of bur-like nanostructure[191].
Generally speaking, adopting the above-mentioned hard/soft-templating method,

hydrothermal method, novel solid state method, sol-gel method, and their coupled technique
can give rise to a number of single-crystalline or polycrystalline perovskites with various
morphologies. Even so, however, a lot of works needs to be done. For example, catalysts with
the same chemical composition derived from various methods may exhibit greatly different
catalytic activities for the same reaction due to the presence of difference in exposed crystal
planes [197,198]. Therefore, the development of a strategy to control crystal growth
(maximizing the exposed ratios of more reactive crystal planes, minimizing those of less
reactive ones, and thus optimizing the desired structure of active sites) should be paid enough
attention.
(a) (b)
(c)
(d) (e)
Figure 26. SEM images of nearly spherical BaZrO3 powders obtained for (a) 10 and (b) 20 min, and of
decaoctahedral BaZrO3 powders obtained for (c) 40, (d) 80, and (e) 160 min at 140 oC[192].
2.5. Macro/Mesoporous Perovskite Synthesis
Because of the attractive physicochemical properties and potential applications in

catalysis, biotechnology, adsorption, and separation, fabrication of hierarchically porous
(macro/mesoporous) materials, especially for the three-dimensional ordered macro/
mesoporous (3DOM) materials, has been a focus in the research on materials science and
engineering in recent years [99,199,200]. By using close-packed arrays of monodisperse
spheres, such as polystyrene (PS), poly(methyl methacrylate) (PMMA), and silica as
template, metals [201,202], metal oxides [203208], metal chalcogenides [209], silica
[204,210,211], carbon [212,213], polymers [214,215], and hydroxyapatite [216] with 3DOM
structures have been generated.
Most of 3DOM metal oxides are fabricated by an alkoxide-based sol-gel process. The
typical procedures include: (i) preparation of a colloidal crystal template by ordering
monodispersive microspheres (e.g., PS, PMMA or silica) into a face-centered close-packed
array; (ii) infiltration of a metal alkoxide solution into voids in the colloidal crystals, and in-
situ solidification of the precursor via a sol-gel process; and (iii) template removal via a
calcination or extraction pathway [99,217219]. Using the alkoxide-based sol-gel method,
3DOM oxides of Si, Ti, Zr, Al, Sb, W, Fe, and 3DOM mixed oxides of some of them can
synthesized due to the moderate reactivity of their alkoxide precursors [99]. However, most
of the other metal alkoxides exhibit high reactivity and their reactions are too quick to be
controlled. Moreover, it is usually difficult and expensive to obtain the alkoxides of transition
metals and lanthanide metals. Due to the low melting points, commercially available salts of
these metals are usually not suitable for the use as starting material. At a temperature where
the polymer template decomposes, these metal alkoxide salts melt, and hence causing a
destruction of the 3DOM structure. Therefore, it is very important to solidify the transition or
lanthanide metal elements before decomposition of the polymer template. To resolve these
problems, it is a good strategy to add oxalic acid [220], ammonia [221] or EDTA [222] to the
preparation system because they can first react with the metal salts (acetates or nitrates) in the
voids of the colloidal crystals and then solidify the metal components. By means of these
techniques, a number of 3DOM metal oxides (e.g., MgO, Cr2O3, Mn2O3, Fe2O3, Co3O4, NiO,
and ZnO) have been successfully synthesized [220]. Unfortunately, since metal species show
different reactivity towards oxalic acid, ammonia or EDTA, and the generated oxalate salts or
metal hydroxides have different solubility in the reacting media, controllably fabricating a
mixed metal oxide with a desired metal ratio using the above methods is still difficult.
Figure 27. SEM images of SrTiO3 crystal with (A) cubic shape and (B) octahedral shape. Scale bar is 1
m[194].
Figure 28. SEM images of SrTiO3 crystals with (A) hexapod and (B) decapod shape. Insets are models
of configuration of the SrTiO3 legs. All of the red-colored legs in panel b lie on the cyan-colored disk
(bc plate). Scale bar is 2 m[194].
To obtain 3DOM perovskites, several groups have developed the synthesis procedures,
which ensure chemical homogeneity of the product. The only example without a
solidification process is the fabrication of 3DOM Sm0.5Sr0.5CoO3 from mixed metal nitrates
[223]. Using a mixed metal alkoxide precursor derived from a stoichiometric mixture of
lanthanum acetate, calcium acetate, strontium acetylacetonate, and manganese acetate in
acidified 2-methoxyethanol, Hur and co-workers generated 3DOM La0.7Ca0.3xSrxMnO3 with
PMMA as a colloidal template [224,225]. As shown in Figure 30, the as-prepared perovskite
displays a very smooth surface and a long-range ordered porous structure over large areas.
The size of the spherical voids is almost the same as that of the PMMA microspheres used,
indicating that little shrinkage of the overall structure occurs during the sintering process.
Using PS beads (diameter = 3 m) as template, a crystallized 3DOM Li0.35La0.55TiO3
membrane can be obtained by annealing the intermediate containing mixed Li-La-Ti-O and
PS gel at 1000 oC. The Li-La-Ti-O sol is derived from the mixed solution of CH3COOLi,
La(CH3COO)31.5H2O, Ti(OCH(CH3)2)4, CH3COOH, i-C3H7OH, and H2O with a molar ratio
of 0.35:0.55:1:10:20:140 [226228]. The 3DOM La0.75Sr0.25Cr0.5Mn0.5O3YSZ (yttria-

stabilized zirconia) and Ba0.5Sr0.5Co0.8Fe0.2O3YSZ composites are fabricated via a novel
route, which comprises (i) the preparation of a mixture consisting of solid powders of the
targeted phase and PMMA microspheres in an appropriate ratio, and (ii) the addition of
polyvinyl alcohol to the mixture as a binder to increase the cohesion between the particles and
secure the stability of the structure at high temperatures [229].
It is worth pointing out that Ueda and co-workers generated a number of 3DOM single or
mixed metal oxides, such as Al2O3, Mn2O3/Mn3O4, Fe2O3, Cr2O3, ZnAl2O4, LaAlO3, LaCrO4,
ZnCr2O4, MFe2O4 (M = Co, Ni, Zn, ZnxNi1x (x = 0.20.8)), La1xSrxFeO3 (x = 00.4), and
LaMnO3, by a facile one-pot synthesis route without the need of any alkoxide precursors
[99,217219]. Such an interesting and relatively versatile strategy is to use an ethylene glycol
(EG)-methanol solution of metal nitrates, which can be converted to a mixed metal glyoxylate
salt during in-situ nitrate oxidation at a lower temperature before the template PMMA or PS
is removed. Further calcination at higher temperatures causes the glyoxylate salt to
decompose to the desired 3DOM metal oxides, in the meanwhile removes the polymer
template. A typical procedure for the fabrication of La1xSrxFeO3 perovskites [218] is as
follows: desired amounts of mixed metal nitrates (total metal concentration = 12 mol/L) is
dissolved with about 5 mL of EG in a 100-mL beaker at room temperature for 2 h under
stirring, and the produced EG solution is poured to a 25-mL volumetric flask. Methanol (40
vol%) and EG are added in the amount necessary to achieve the desired concentration. Then
about 1.0 g of the PS colloidal crystals (161 10 nm in size) is soaked in the produced dark
brown solution for 5 h. The excess solution is removed from the impregnated PS template by
filtration. The obtained sample is allowed to dry in air at room temperature overnight. 0.5 g of
the desired sample is mixed with 2.5 g of quartz sands (1015 mesh) and calcined in an air
flow of 50 mL/min at a rate of 1 oC/min from room temperature to 600 oC and kept at this
temperature for 5 h in a tube furnace. Shown in Figure 31 are the typical SEM images of the
as-prepared perovskites. Clearly, well-ordered air spheres and interconnected walls create a
three-dimensional inverse-opal macroporous architecture with a pore size of 7294 nm,
which corresponds to a shrinkage of 4558% (with respect to the initial PS size) due to the
melting of PS microspheres and sintering of the formed perovskite. Addition of methanol to
the EG solution is a key step for successful fabrication of the desired 3DOM material. With a
rise in methanol content from 0 to 40 vol%, the complete penetration time shortens from 10 to
5 h, thus resulting in an increase in the fraction of the 3DOM structured particles from
6070% to 8090%. The possible formation mechanism of the 3DOM structure derived by
this method is illustrated in Figure 32 [99,217]. At first, the reaction taking place between the
mixed metal nitrates and EG gives birth to the mixed metal glyoxylate or oxalate salts in the
voids of the template microspheres at a low temperature. Then during the thermal treatment
process, the pore size is reduced due to the melting of the template and the air space in void
(shell structure) is generated after evaporation of methanol and remaining EG. Further
calcination results in a total removal of the template and formation of the desired crystal
phase. The subsequent crystallite growth causes the 3DOM materials to transform from the
shell structure to the skeleton structure.
Figure 29. SEM images of single-crystal BaTiO3 synthesized at 200 oC for 6 h with 0.1 mol/L KOH.
Inset is the SAED pattern[195].
Figure 30. (a) A typical SEM image of the PMMA template. (b) A low-magnification SEM image of
the macroporous La0.7Ca0.2Sr0.1MnO3 film showing the presence of cracks and defects produced after
removal of template. (c) A higher magnification SEM image of the same porous sample[224].
(a) (b)
Figure 31. SEM images of LaFeO3 obtained with PS infiltrated by EG-methanol (40 vol%) mixed
solution of Fe(NO3)3 and La(NO3)3 (total metal concentration = 2 mol/L) and calcined at 600 oC for 5
h[218].
Figure 32. Formation mechanism of 3DOM metal oxide[99].
(a) (b) (c)
50 nm 20 nm 5 nm
(d) (e) (f)
50 nm 10 nm 5 nm
Figure 33. HRTEM images of a SrTiO3 mesoporous ordered network (ac) before crystallization,
obtained at 590 oC, and (de) after crystallization, obtained at 630 oC[233].
Although 3DOM metal oxides can be synthesized following the above mentioned
methods, their surface areas are still relatively low (< 100 m2/g) and the pores are only
macropores without mesopore formation on the walls. These drawbacks inevitably inhibit
wide applications of such porous materials. Therefore, high-surface-area 3DOM metal oxides
with mesoporous walls are highly desirable. Recently, our group used P123 as soft template,
regularly packed PMMA microspheres as hard template, and tetraethyl orthosilicate as Si
source to successfully make 3DOM silica with 2D ordered mesoporous walls and unusually
large pore volume (2.1 cm3/g) [230]. Following this idea, we introduced P123 or F127
(EO106PO70EO106) to the mixed alcoholic solution containing PMMA-metal nitrates, and
generated a number of high-surface-area 3DOM MgO (243 m2/g), -Al2O3 (145 m2/g), and
Ce1xZrxO2 (x = 0.3, 0.4) (100 m2/g) with mesoporous crystalline walls [231]. This strategy is
promising in the fabrication of 3DOM perovskites with mesoporous walls.
Based on the evaporation-induced self-assembly (EISA) method reported by Yang et al.
[232], a number of mesoporous single and mixed metal oxides have been produced. However,
only one pervoskite-type oxide, i.e. SrTiO3 with mesoporous structure was prepared using the
EISA method with a specially formulated non-ionic block-copolymer KLE3739 (PBH79-
block-PEO89, PBH = hydrogenated poly(butadiene)) template [233]. As shown in Figure 33,
the walls are composed of nanoparticles (1020 nm in size). If the annealing treatment at the
crystallization temperature is stopped, amorphous mesoporous networks can be crystallized
without extensive degradation of the meso-organization. The porosity of SrTiO3 is about 31
vol% (estimated using the Bruggeman effective medium approximation model applied to
ellipsometric data). Unfortunately, utilization of such a methodology is greatly hindered due
to the use of extremely expensive and noncommercial template. Very recently, an alternative
method to prepare perovskites with a mesoporous structure has been reported by Vilarionho
and co-workers [234,235]. They adopted cetyltrimethylammonium chloride (C16TMAC, 25
wt% sol in water) as the cationic surfactant to generate cubic BaTiO3 single crystallites with
wormhole-like mesopores via the sol-precipitation route [235]. Further investigations made
by the same group reveal that nanoporous cubic BaTiO3 single crystallites can be obtained
with the assistance of nonionic block-polymer Pluronic PE 10300 (EO15PO70EO15) but
without subsequent thermal treatments [234]. Disordered nanoporous structures (Figure 34)
are formed in the sample calcined at 300 or 400 oC. These materials show a BET surface area
of about 70 m2/g and a nanopore volume fraction of 44.7%, larger than those of porous
BaTiO3 derived with a cationic surfactant C16TMAC [235]. Because of the closure and
collapse of partial pores, calcination at a higher temperature (e.g., 600 oC) results in an
obvious decrease in surface area from 70 to 44 m2/g. Therefore, this method seems to be not
suitable for the generation of mesoporous perovskites because a high-temperature (> 650 oC)
treatment is needed for perovskite phase formation. Nevertheless, it provides a thinking line
to directly synthesize mesoporous perovskites with the aid of a surfactant.
It is well known that the nanocasting strategy with mesoporous silica or carbon as hard
template is an effective technique to fabricate mesoporous metal and metal oxides [236238].
Figure 34. TEM images of (a) as-prepared BaTiO3 crystallites and the SAED pattern (inset),
nanoporous BaTiO3 calcined (b) at 300 oC and (c) at 400 oC[234].
Compared to mesoporous single metal oxides, reports on the synthesis of mesoporous

binary or mixed oxides via a nanocasting route are few [237240]. To the best of our
knowledge, there is no report claiming the successful preparation of mesoporous perovskites
following such a nanocasting route. By carefully controlling the calcination temperature for
perovskite structure formation and fully filling the mesopores of the hard template with a
homogenous precursor solution, however, it is possible to make mesoporous perovskites
through the nanocasting process.
3. PHYSICOCHEMICAL PROPERTIES
3.1. Magnetic Property
In an ideal cubic perovskite (ABO3) structure, each oxygen atom is shared by two B-site
ions (B-O-B angle = 180o), which is favorable for the super-exchange interactions between
magnetic B cations. Such an exchange usually leads to antiparallel coupling of nearest-
neighbor magnetic moments. Because the alignment of Fe moments is not strictly antiparallel
but slightly canted on LnFeO3 (Ln = lanthanide), a small net magnetization is generated.
Among the orthorhombic LnFeO3 (Ln = La, Pr, Nd, Sm, Eu, Gd) nanocrystallites prepared by
a general nanosized heterobimetallic precursor approach at 500 oC, SmFeO3 is paramagnetic
rather than superparamagnetic, ascribable to the paramagnetic susceptibility of Sm ions and a
minor ferrimagnetic contribution from the canting of the Fe sublattice, whereas the other
ferrites are weak ferromagnets. Moreover, the EuFeO3 and GdFeO3 nanocrystallites exhibit
better weak ferromagnetic behavior with the corresponding saturation magnetization intensity
being 3.99 and 3.89 emu/g, much higher than those (0.3251.335 emu/g) of the EuFeO3 and
GdFeO3 materials prepared via the sol-gel and solution-based chemical routes [9]. Below the
magnetic ordering temperature, compared to the CaMn1xNbxO3 (x 0.08) samples obtained
through a conventional solid-state reaction at 1400 oC for 6 h, the ultrafine manganite
nanoparticles derived from ultrasonic spray combustion display a pronounced ferromagnetic
suppression coupled to an abrupt increase of the absolute thermopower value. This might be
associated with the different compositions on grain boundaries and/or a weakening of the
double exchange ferromagnetism phenomena due to the small size of the nanoparticles [241].
On the basis of the experimental results, Feng and co-workers deduced that the disorder in the
grain boundary region of the rod-shaped La0.7Sr0.3MnO3 sample is larger than that of the
particle in the same size range, which leads to a decrease in the Curie temperature (Tc) and
effective magnetization (m) [147,148]. The Tc of La0.7Sr0.3MnO3 nanorods fabricated by the
molten-salt method is 11 oC, much lower than those (87~93 oC) of La0.7Sr0.3MnO3
synthesized via a solid-state reaction route. The m of La0.7Sr0.3MnO3 nanorods is 2.075 B,
about half of the ideal effective moment (4.17 B). At room temperature, BiFeO3
nanoparticles (80120 nm in size) and BiFeO3 nanotubes (20 nm in wall thickness) exhibit a
weak ferromagnetic order, quite different from the linear relationship of magnetization vs
magnetic field in the bulk BiFeO3. The weak ferromagnetic order in the BiFeO3 nanoparticles
and nanotubes is attributable to the grain size confinement effect [11,62]. By comparing the
magnetic behaviors of La0.325Pr0.300Ca0.375MnO3 nanotubes and powders with the same
average particle size, Curiale et al. pointed out that the nanostructure plays an important role
in the magnetic properties [91]. The tube-like structure favors the orientation of the moments
by the action of magnetic field, which should be directly related to the presence of magnetic
interactions between the grains. Although the general feature of magnetization vs temperature
has been reproduced, the La1xBaxMnO3 (x = 0.3, 0.5, 0.6) nanocubes, compared to those of
the bulk counterparts, show a slight reduction in saturation magnetization as well as a
suppression of the phase transition temperature. It may originate from the enhanced surface-
to-volume ratio or the presence of a superparamagnetic phase in these nanocrystalline
samples [127]. Based on the results mentioned above, one can find that both the particle size
[127,241] and crystal morphology [91,147,148] have important effects on the magnetic
property. Recently, Chai et al. pointed out that the product particle size exerts a bigger impact
than its morphology on the magnetic properties of the flower-, microcube-, and nanocube-like
La0.5Ba0.5MnO3 perovskites [173]. In other cases, however, it is hard to say which one (crystal
particle size or morphology) plays a more important role in influencing the magnetic
performance.
In addition, remarkable metal-insulator phase transitions and pronounced colossal
magnetoresistance have been observed in the core-shell MgO/La0.67Ca0.33MnO3 and
MgO/La0.67Ca0.33MnO3 nanowires [158160]. The results are likely related to the relaxation
of the stress in the radial direction unique in core-shell nanowires. Interestingly, the magnetic
moment of porous La0.7Ca0.3MnO3 perovskite is smaller than that of its single crystalline
counterpart [225]. The reduced moment in the porous perovskite can be attributed to the
presence of a superparamagnetic phase resulting from nanocrystalline domains in the
ferromagnetic matrix. On the other hand, porous La0.7Ca0.3MnO3 shows a much higher Tc
value than its single-crystalline counterpart. The possible reason for Tc enhancement is the
occurrence of composition fluctuation and/or an oxygen annealing effect. The porous material
can adsorb much more oxygen than the bulk one. Thus, oxygen saturation may result in an
enhancement of Tc. Although the Tc value of macroporous La0.7Ca0.2Sr0.1MnO3 is close to that
of its corresponding bulk polycrystallite, the resistivity of the porous perovskite throughout
the temperature region (248~102 oC) is about two orders of magnitude higher than that of its
polycrystalline counterpart due to the connectivity between the porous grains and nanosized
grains [224]. The magnitude of magnetoresistance observed in the porous perovskite is about
twice as much as that observed in the polycrystalline one, a result due to the spin-polarized
tunneling at the nanosized grain boundaries.
3.2. Electrical Property
Electrical properties of perovskites have been received special attention due to their
superconductivity [242]. The electrical behaviors in perovskites depend on the outermost
electrons, which may be localized at specific atomic sites or collective. Because localized
electrons can carry a spontaneous moment, there is a strong correlation between the electrical
and magnetic property of the perovskite material [2]. It is found that the ferroelectric
properties (e.g., Tc, mean polarization, area of hysteresis loop, coercive electric field, and
remnant polarization) of BaTiO3 nanowires alter linearly with the diameter of nanowires only
when the diameter is less than 20 nm [243]. Due to the nanosize effect in the BaTiO3
nanowires, the dielectric constant measurement shows a shift in the transition temperature
(105 oC) compared to the bulk transition temperature (132 oC) [110]. As for an individual
NaNbO3 nanowire with a width of 200 nm and a length of 80 nm, the piezoelectric coefficient
(dzz) along the vertical direction is 0.95 pm/V, whereas it is 4.26 pm/V for the one with 100
nm in width and 120 nm in length, suggesting that a nanowire with a larger length-to-width
ratio vibrates more easily under an applied electric field [114]. Polycrystalline PbTiO3 [66]
and Pb0.25Ba0.15Sr0.6TiO3 nanotubes [71], PbZr0.53Ti0.47O3 nanowires [60], and BiFeO3
nanowires [58] or nanotubes [63] display highly ordered ferroelectric behaviors. In the
coercive field in the amplitude and phase loops of a single PbZr0.53Ti0.47O3 nanowire
piezoresponse, a positive shift is observed. This result may be caused by the surface change at
the electrode and PbZr0.53Ti0.47O3 interface or the space change stored in the alumina template
walls [60]. The remnant polarization (11 C/cm2) of the BiFeO3 nanowires is much higher
than that (6 C/cm2) of the bulk BiFeO3. The hysteresis phenomenon is associated with the
polarization switching and ferroelectric properties of the BiFeO3 sample [58]. Due to a high
surface area of nanotubes, there is an enhancement in ferroelectric polarization of
Pb0.25Ba0.15Sr0.6TiO3 nanotubes compared to that of the nanofilm counterpart [71]. Very
interestingly, the dielectric constant of BaTiO3 greatly decreases after nanoporosity
introduction, which can be modified by varying the initial block copolymer type and/or
content and by changing the post thermal treatment conditions. As the frequency increases
from 10 kHz to 1 MHz, the dielectric constant of nanoporous BaTiO3 decreases from 375 to
110 whereas that of nonporous BaTiO3 decreases from 2530 to 700. That is to say, it provides
an effective way to manipulate the dielectric constant in a wide range, especially where low
dielectric constant is desired to achieve high figures of merit in pyroelectric and piezoelectric
applications [234].
3.3. Optical Property
Perovskites are a good candidate of optical materials for the generation of new
optoelectronic devices with high performance. Recently, photoluminescence (PL) of
perovskites arising from the imperfection and defect has attracted increasing attention
[182,185]. It is observed from the room-temperature PL spectra of single-crystalline PbTiO3
nanowires that there is a blue light emission band at a wavelength of about 471 nm (2.63 eV),
and the signal intensity increases with a rise in excitation wavelength [111,112]. At an
excitation wavelength of 488 nm, the room-temperature PL spectra of PbTiO3 nanotubes
show a strong green emission band at 550 nm (2.25 eV) [133]. Since the excitation energy
used is lower than the band gap energy (about 4.15 eV) of PbTiO3 nanowires, the light
emissions observed from 420 to 560 nm are not related to a direction electron transition from
the valence band to conduction band. Actually, oxygen vacancies of PbTiO3 nanowires,
which can produce a heterogeneous band gap structure with a tail of localized states and
possible mobile edges, might be responsible for the light emission [111,112]. A similar
explanation can be used in the case of PbTiO3 nanotubes. Due to the high surface area,
numerous local defects (e.g., oxygen vacancies and dangling bonds) inevitably exist in a
highly ordered nanotubular structure. Such defects are an emission source, resulting in a
unique PL behavior of PbTiO3 nanotubes [133]. As for the BaZrO3 perovskite with a
decaoctahedral morphology, the structural feature associated with distorted octahedral (ZrO6)
and dodecahedral (BaO12) clusters plays an important role in determining the PL property
[192]. It is known that BaTi0.5Mn0.5O3 is a high dielectric material and BaMnO3 is not a
luminescent material. Nevertheless, Wang and co-workers found that BaTi0.5Mn0.5O3
nanorods exhibit an excellent PL performance, showing blue and green fluorescence with the
corresponding emission bands at = 465 and 593 nm in the PL spectrum of this nanorod-like
perovskite [151].
In addition, perovskites are interesting when they are used as a host material for rare earth
photoluminescence. The room-temperature PL spectra of Eu3+-doped spherical, cubic or
dodecahedral, and agglomerated BaZrO3 are shown in Figure 35 [185,188]. The PL spectral
feature in the range of 570660 nm can be ascribed to emissions from the excited 5D0 state
level to lower lying 7FJ (J = 0, 1, 2, 3, 4) levels of the 4f6 configuration of Eu3+. The most
prominent band at 612 nm is due to the 5D0 7F2 transition, as expected from the Judd-Ofelt
selection rule. Although similar in spectral shape, their intensities differ in the Eu3+-doped
BaZrO3 perovskites with various morphologies. Since the intrinsic geometry of spherical
BaZrO3 particles minimizes the scattering of light from the sample surface, as compared with
cubic, polyhedral, and agglomerated BaZrO3 particles, it is understandable that the spherical
particles show the highest emission intensity. On the other hand, the aggregated particles
possess more bulk defects, surface cracks, and intraparticle residual elastic strains, hence
giving rise to a reduced luminescence yield. Such a morphology-dependent PL property has
also been observed in the case of CaTiO3:Pr3+ [182]. Under the same conditions, the
CaTiO3:Pr3+ mesocubes exhibit considerably stronger red emission at about 612 nm than the
prismatic rectangular CaTiO3:Pr3+ rods.
Very recently, the photoelectrochemical performance of microtubular CaTiO3 and fibrous
SrTiO3 and BaTiO3 has been reported. CaTiO3 microtubes show better photoelectrochemical
responsing ability relative to SrTiO3 and BaTiO3 microstructures. Beside the smaller
diameter, the 1D structure can effectively reduce the contact resistance at grain boundaries to
improve electron transport, leading to an enhancement in photocurrent and conversion
efficiency [134]. The current vs voltage characteristics of 1D Ag2MO4 (M = Cr, Mo, W)
nanowires exhibit a unique rectifying behavior with the fast and reversible photo-switching
response under on/off light exposure conditions [117]. Oxygen ions play an important role in
the photoconductivity. The light-to-dark conductivity ratios of these 1D compounds are
correlated with the ionic potential of the metal. This result suggests the possible application in
photosensitive devices in the future.
Figure 35. Room temperature PL emission (ex = 256 nm) spectra of the 2% Eu3+-doped BaZrO3
samples with different morphologies[185].
Figure 36. The ageing dependence of the CONOx cross-over point conversion over the
LaFe0.57Co0.38Pd0.05O3 and Pd/-Al2O3 catalysts[257].
(a) (b)
(c) (d)
(e)
Figure 37. SEM photographs of the LaCoO3 catalysts obtained after calcination: (a) SS; (b) COP; (c)
CIT; and (d) COPRG. (e) Steady-state conversions obtained for the oxidation reaction of 0.25% CH4 as
a function of the reaction temperature over LaCoO3: () COPRG, (z) RG, () CIT, () COP, and
() SS[258].
(g) (h)
Figure 38. SEM images of (ac) NdSrCu1xCoxO4 and (df) Sm1.8Ce0.2Cu1xCoxO4+ at x = 0 (a, d), 0.2
(b, e), and 0.4 (c, f). CH4 conversion as a function of reaction temperature over (A) NdSrCu1xCoxO4
and (B) Sm1.8Ce0.2Cu1xCoxO4+ as well as over LaCoO3 under the conditions of CH4/O2 molar ratio =
1/20 and space velocity = 10000 h1[263].
4. APPLICATIONS IN CATALYSIS
4.1. Nitrogen Oxide Elimination
Nitrogen oxides (NOx) emitted from vehicles and industrial activities are one of severe
pollutants to the environment. Direct decomposition of NOx and reduction of NOx with H2,
CO, NH3, and hydrocarbons are the good ways to eliminate such a kind of pollutant. Several
reviews relevant to this topic can be seen in the literature [244246]. However, a
summarization on catalytic performance of the specifically morphological perovskites for
NOx removal is still unavailable.
As a green process, NOx decomposition has been received much attention. By using
CNTs as template under hydrothermal conditions, Gao et al. [83] fabricated La2CuO4 (i.e.,
La2Cu0.882+Cu0.12+O3.94) nanofibers with a surface area of 105 m2/g. Below 300 oC, NO can be
totally decomposed to N2 and O2 over the cuprate nanofibers. Over the bulk La2CuO4 catalyst
(i.e., La2Cu0.922+Cu0.083+O4.04) with a surface area of 2.7 m2/g prepared by a solid-state
reaction method, only 78% NO can be decomposed even at 800 oC and the products are N2,
O2, and NO2. The authors attributed the good catalytic performance of the nanofibrous
cuprate material to the co-presence of nanofibrous morphology, high surface area, appropriate
Cu+/Cu2+ molar ratio, and uncompetitive O2 and NO adsorptions [83]. Via a highly
exothermic and self-sustaining reaction route, the LaBO3 (B = Cr, Mn, Fe, Co) nanoparticles
with a typical size of 5575 nm can be prepared [247]. The performance of these perovskites
for the decomposition of N2O to N2 and O2 depends on the nature of the B-site ions, with the
LaCoO3 catalyst showing the best activity. By adopting the sol-gel route, colloidal crystal
templating, and reactive grinding methods, Granger and co-workers prepared nanosized
rhombohedral LaCoO3 catalysts. The morphology of the product has a great impact on the
catalytic activity for N2O decomposition. The LaCoO3 catalyst fabricated by the reactive
grinding method exhibits the best performance, which may be related to a higher surface area
and a higher oxygen vacancy density as well as the presence of trace amounts of Fe and Zn
impurities introduced during the preparation process [248].
The improvement in physicochemical properties (including surface area) of perovskites
can enhance their efficiency of NOx removal. By using the reverse microemulsion method,
LaMO3 (M = Mn, Fe) [24], La1xAxFeO3 (A = Sr, Ce) [25], and La0.8Ce0.2B0.4Mn0.6O3 (B =
Cu, Ag) [26] nanoparticles with a diameter of 50150 and 4050 nm can be prepared,
respectively. Such high-surface-area (1461 m2/g) nanocatalysts showed better catalytic
activity for (NO + CO) reaction than the catalysts derived from the common ceramic and sol-
gel methods. In the case of LaMnO3, besides the high surface area, the presence of
Mn3+OMn4+ pairs (favorable for adsorption-desorption flip-flops) is a factor contributing to
the excellent performance [24]. The sequence in catalytic activity is La0.8Ce0.2FeO3 >
La0.8Sr0.1Ce0.1FeO3 > La0.85Sr0.15FeO3 > LaFeO3, in accordance with the order in increment of
their surface areas [25]. It is reported that La0.8Ce0.2Cu0.4Mn0.6O3 outperforms
La0.8Ce0.2Ag0.4Mn0.6O3 for the selective reduction of NO with CO [26]. This result is
understandable because the dominant Cu+ ions and their redox reaction Cu+ Cu2+ favor the
NO adsorption and reduction by CO. Nanosized LaFe1xCoxO3 catalysts can be obtained using
a porous carbon as hard template [27]. The surface area of LaFeO3 and LaFe0.6Co0.4O3 is 30.5
and 49.3 m2/g, much higher than that (5.6 and 3.6 m2/g) of the citrate-complexing derived
counterparts, respectively. The larger surface area allows the exposure of a higher number of
accessible active sites in the nanocasting-derived perovskites, resulting in a better catalytic
activity for the reduction of NO with CO.
In the last years, Kaliaguine and co-workers fabricated a series of nanoscale perovskites
LaCo1xCuxO3, LaMn1xCuxO3, LaFe1x(Cu, Pd)xO3, and La0.88Ag0.12FeO3 (surface area =
2348 m2/g) via a reactive grinding route, and investigated their catalytic behaviors for the
(NO + CO) and (NO + C3H6 + O2) reactions [249256]. The catalytic mechanism of NO
reduction with CO or C3H6 is different. The NO reduction with C3H6 as a reducer and in the
presence of O2 gives rise to the NO3 species over the catalyst surface, which can be reduced
by C3H6 to form organo nitrogen compounds; in the reduction of NO by CO, however,
nitrosyl species (NO) is the main adsorbed species [249]. The partial substitution with Cu
ions can increase the anion vacancy density (facilitating the adsorption and dissociation of
NO) and the lattice oxygen mobility (favoring the regeneration of the anion vacancies). Such
a substitution brings about a significant improvement in catalytic performance of the
perovskite for the (NO + CO) reaction [249,250]. In the selective catalytic reduction of NO
by C3H6, a marked enhancement in N2 yield can be achieved after Cu incorporation to the
perovskite lattice since the doping of copper ions promotes the formation of NO3 and
subsequent generation of organo nitrogen compounds (the rate-determining step)
[250,251,256]. Owning to the outstanding redox property of LaFe0.97Pd0.03O3, 96% NO and
86% CO conversions at 300 oC for the (NO + CO) reaction and 67% N2 yield and 68% C3H6
conversion at 350 oC for the (NO + C3H6 + O2) reaction can be achieved [249,251]. In
addition, this research group also examined the sensitivity to water vapor and poisoning to
SO2 of nanosized La(Co, Mn, Fe)1x(Cu, Pd)xO3 materials for the reduction of NO by C3H6
[254,255]. A moderate deactivation at 300500 oC and a slighter deactivation at 700 oC over
these catalysts are observed in the presence of 10% H2O in the feed gas mixture. Due to the
abundant anion vacancies, LaCoO3 shows a lower sensitivity to water vapor compared to
LaFeO3 and LaMnO3. Further improvement in resistance to H2O deactivation of LaCoO3 can
be achieved by Cu-doping. The introduction of palladium to the LaFeO3 lattice enhances the
ability of resistance to H2O deactivation. It should be pointed out that the deactivation of the
above perovskite catalysts induced by water vapor is reversible and can be well recovered
after removal of H2O from the feed [255]. The effect of SO2 on NO reduction by C3H6 over
LaFe0.8Cu0.2O3 is somewhat complicated. At SO2 concentration 20 ppm, most of the SO2
are chemisorbed at anion vacancies of the perovskite, and a minor amount of gaseous SO2 is
oxidized to generate sulfite and sulfate on the perovskite surface. The competitive adsorption
between SO2 and reactant molecules results in a slight modification on the physicochemical
property of the perovskite. After removal of SO2 from the feed, the induced deactivation is
reversible. With a rise in SO2 concentration up to 80 ppm, however, a partial destruction takes
place due to the formation of significant amount of sulfate, causing a great decrease in
catalytic activity. After being treated first in a 5% H2/He flow and then in an oxidized
atmosphere, the deactivated perovskite catalyst can be partially regenerated [254].
It is worth mentioning that the doping of noble metal ions to a perovskite framework has
a promotional effect on catalytic performance. As mentioned above, LaFe0.97Pd0.03O3 shows
excellent catalytic activity for NOx removal [249,251]. The incorporation of Pd to the
perovskite lattice can also circumvent the aggregation problem encountered in supported
noble metal catalysts. Nishihata et al. reported a LaFe0.57Co0.38Pd0.05O3 catalyst that possesses
a self-regeneration character [257]. The authors believed that Pd can move into and out of the
perovskite lattice via the process Pd2+ PdO when the perovskite is exposed to a fluctuation
atmosphere of the emission exhaust. A stable catalytic activity (Figure 36) of
LaFe0.57Co0.38Pd0.05O3 confirms the above viewpoint.
4.2. Methane Oxidation
Taking into consideration the serious concerns on environmental pollution and energy
shortage, most people think that natural gas (methane content 8090%) is a relatively clean
and rich energy resource owing to its abundant storage, low price, convenient utilization, high
thermal efficiency, and minor pollution. The direct combustion of methane usually takes
place at high temperatures (> 1200 oC), giving rise to a significant amount of NOx that is
harmful to the atmosphere. Catalytic combustion of methane, however, can avoid the
formation of NOx since its oxidation undergoes at low temperatures (< 900 oC) over a
catalyst. Therefore, such a catalytic strategy represents one of the best ways for methane
combustion.
Traditional catalysts employed for methane combustion are supported noble metals (Pd,
Pt, Rh) [5], which show excellent light-off activity and anti-sulfur poisoning ability but the
shortcomings (such as rather high cost and easy sintering at high temperatures) limit their
wide applications. Perovskite-type or perovskite-like oxides are cheap and exhibit good
catalytic performance for methane combustion. The partial substitution in one or both cationic
sites by heterovalent ions can stabilize the B-site ions with unusual oxidation states and
generate structural defects (such as cationic or anionic vacancies), thus bringing about a
significant enhancement in catalytic performance. The catalytic activities of some perovskite-
type oxides and perovskite-like oxides for methane combustion are summarized in Table 2.
Table 2. Surface morphologies, BET surface areas, preparation and reaction

parameters, and catalytic performance of some perovskite-type oxide catalysts
Catalyst Morphology Surface area/ CH4/O2/BGa T10% T50% T90% Ref.

calcination molar ratio; space (oC) (oC) (oC)
temperature velocity
(m2/g / oC)
60 wt% Mesoporous 510 / 700 1/4/95; 20000 h1 406 510 550 39
LaCoO3/SBA-15 structure
LaCoO3 Nanoparticle 6.0 / 700 1/4/95; 20000 h1 521 596 640 39
La0.5Sr0.5MnO3 Nanocube 3.73 / 1/10/39; 100000 360 492 570 128
mL g1 h1
La0.5Sr0.5MnO3 Nanoparticle 7.57 / 700 1/10/39; 100000 440 541 620 128
mL g1 h1
La0.5Sr0.5MnO3 Nanoparticle 9.28 / 700 1/21/78; 60000 365 478 558 267
mL g1 h1
La0.5Sr0.5MnO3 Nanocube 4.50 / 1/21/78; 60000 453 522 572 267
mL g1 h1
LaCoO3-SSb Nanoparticle 0.4 / 1000 0.25/1.0/98.75; 510 615 258
5625 h1
LaCoO3-COPRGc Nanoparticle 18.7 / 500 0.25/1.0/98.75; 305 370 418 258
5625 h1
LaMnO3 Nanoparticle 11.5 / 700 1/4/95; 60000 h1 460 565 655 260
LaSrCuO4 Spindle-like 11.9 / 850 1/10/39; 50000 485 618 668 268
mL g1 h1
YBa2Cu3O7 Nanorod 6.1 / 850 1/10/39; 50000 468 640 711 269
mL g1 h1
YBa2Cu3O7 Nanosphere 4.5 / 850 1/10/39; 50000 500 660 722 269
mL g1 h1
a
BG represents the balance gas; b the catalyst was obtained by the solid-state reaction of the metal
oxides; and c the catalyst was fabricated via the reactive grinding of amorphous La-Co precursor.
It is generally accepted that preparation strategy has a strong influence on the

performance of a catalyst. Royer et al. [258] adopted various methods, such as the solid-state
reaction of the metal oxides (SS), coprecipitation of the metal nitrates (COP), citrate
complexation (CIT), reactive grinding of the metal oxides (RG) and of amorphous La-Co
precursor (COPRG), to fabricate LaCo1xFexO3 with the particle size of 100 nm10 m
(Figure 37). The authors observed that, among the five catalysts, the COPRG one shows the
highest activity (T50% (the temperature required for 50% conversion of methane) = 375 oC at
5625 h1) for methane oxidation, and the specific reaction rate is intimately related to the
preparation procedure adopted. By adopting an annealing method with metal oxides as the
precursor, Petrovi et al. fabricated LaMO3 (M = Mg, Ti, Fe) and LaMg0.5TiyFexO3 catalysts,
and found that the LaMg0.5TiyFexO3 (0.13 x 0.53, 0.1 y 0.4) materials show good
catalytic performance, with the LaMg0.5Ti0.2Fe0.4O3 one performing the best (T50% = 660 oC at
100000 h1) for methane combustion [259]. Through a citric acid gel process and after
calcination at 600 oC, Hammami et al. [260] obtained lanthanum manganite (surface area =
13 m2/g) that gives a T50% of 565 oC at 60000 h1 for methane oxidation. Via the malic acid-
complexing and dry-wet impregnation pathways, LaMn0.4Fe0.6O3-loaded Pd (diameter = 515
nm) catalysts can be prepared. It is observed that the 5 wt% Pd/LaMn0.6Fe0.4O3 catalyst (T100%
(the temperature required for 100% conversion of methane) = 500 oC at 120000 mL/(g h))
outperforms the catalysts with other Pd loadings [261]. Over the LaMn1xMgxO3 (x = 00.5)
nanocatalysts (size = 2580 nm) synthesized by means of a citrate-complexing method [262],
50% methane conversion can be achieved at 435 oC and 4320 h1 over LaMn0.8Mg0.2O3
obtained after 800 oC-calcination for the complete oxidation of methane. Such a good
catalytic performance is due to the rich oxygen adspecies. Recently, our group have prepared
NdSrCu1xCoxO4 (containing Cu3+/Cu2+ ions and oxygen vacancies) and Sm1.8Ce0.2Cu1
2+ +
xCoxO4+ (x = 00.4) (containing Cu /Cu and excess oxygen) using citric acid complexing
method [263]. It is found that there is a morphological evolution from macroporous woven-
like to adobe-like architecture in NdSrCu1xCoxO4 at elevated Co doping, whereas petal-like
or porous morphologies are formed in Sm1.8Ce0.2Cu1xCoxO4+. Among the two series of
catalysts, the NdSrCuO3.70 and Sm1.8Ce0.2Cu0.6Co0.4O4.19 ones show the best catalytic
performance, with the T50% value being 550 and 560 oC at a space velocity of 10000 h1 for
methane combustion (Figure 38), respectively. The high-oxygen nonstoichiometry and strong
Cu3+/Cu2+ or Cu2+/Cu+ redox ability account for the good activity. By adopting the citrate-
complexing sol-gel, flame-hydrolysis or flame-spray pyrolysis approach, Forni and co-
workers prepared SrTiO3 and Sr1xMxTiO3 (M = Ag, K, Gd; x = 0, 0.1) [264266], among
which the Sr0.9K0.1TiO3 (particle size = 100250 nm) catalyst derived from the flame-
hydrolysis route and the Sr0.9Ag0.1TiO3 catalyst fabricated by the citrate-complexing sol-gel
method (particle size = 30100 nm and surface area = 24 m2/g) exhibit the best catalytic
performance (T50% = 560 and 400 oC at 6000 mL/(g h), respectively) for methane oxidation.
The good catalytic activity is associated with the formation of suprafacial O3 species and
O/Ti4+, O2/Ti4+, O3/Ti4+, O/Sr2+, and Ox/Ag2+ or Ox/Ag ion pairs induced by the doping
of potassium or silver ions.
Figure 39. Light-off curves of methane conversion as a function of temperature measured with (a)
MLC-60 (microwave-assisted method), (b) CLC-60 (conventional method), and (c) bulk LaCoO3
perovskites. The inset shows the corresponding Arrhenius plot[39].
It is well known that a catalyst with high surface area usually exhibits a good catalytic
activity for the total oxidation of an organic compound. Perovskite-type or perovskite-like
oxides prepared by means of conventional methods are low in surface area (< 20 m2/g), which
is not favorable for the improvement in catalytic performance. One of good ways to solve this
problem is to disperse the low-surface-area and active perovskite-type or perovskite-like
oxide on a high-surface-area support. For example, introducing LaCoO3 nanocrystals (5.08.5
nm) to the mesopores of high-surface-area SBA-15 via the microwave-assisted route can
enhance its catalytic performance for methane combustion, with the T50% value being 510 oC
over LaCoO3/SBA-15 and 596 oC over LaCoO3 [39] (Figure 39). Using a citric acid-
complexing-assisted impregnation method, Nguyen et al. [36] prepared mesoporous silica
supported LaCoO3 catalysts, in which the LaCoO3 nanoparticles are highly dispersed in the
mesopores of silica. The authors observed that the 38.5 wt% LaCoO3/silica catalyst displays
the highest activity (T50% = 390 oC at SV = 4500 h1) for the oxidation of methane, which is
better than that (T50% = 416 oC) over the bulk LaCoO3 catalyst under similar conditions.
It is known that particle surface morphology of a material has a great impact on the
catalytic performance. Zhu et al. [128,267] fabricated La0.5Sr0.5MnO3 microcubes (edge
length = 1.5 m) and nanoparticles (diameter = 3050 nm) using hydrothermal treatment (at
240 oC for 24 h) and citrate-complexing methods, respectively. A high conversion of CH4 can
be achieved over microcube-like perovskite (T90% (the temperature required for 90%
conversion of methane) = 572 oC) and the nanosized perovskite (T90% = 558 oC) at 12000 h1
(Figure 40). Recently, we adopted a hydrothermal method to generate nano/mircosized
single-crystalline perovskite-like oxides La2xSrxCuO4 (x = 0, 1) with spindle-, rod-, and short
chain-like morphologies (Figure 41) [268]. It is shown that at a space velocity of 50000
mL/(g h), the rod-like LaSrCuO4 (particle size = 150300 nm and length = 13 m) exhibits
the highest catalytic activity (T100% = 675 oC and the corresponding reaction rate = 40.9
mmol/(g h)). The good catalytic performance of LaSrCuO4 nano/microrods is associated with
the higher surface oxygen vacancy concentration, unique single-crystalline structure, and
specific surface morphology. Nano/microsized YBa2Cu3O7 single crystallites with spherical
and rod-like (0.51.0 m in diameter and 520 m in length, respectively) morphologies can
be fabricated hydrothermally at 260 oC [269] (Figure 42). The YBa2Cu3O7 catalyst (derived
hydrothermally with the nanoplatelet Y2O3 as yttrium source and Cu(NO3)2 and Ba(OH)2 as
the copper and barium sources, respectively) shows a 90% methane conversion at 711 oC and
SV = 50000 mL/(g h). Factors such as more oxygen adspecies, better reducibility, single
crystallinity, and unique morphology are responsible for the good catalytic performance of
the hydrothermally derived YBa2Cu3O7 single crystallite.
4.3. Volatile Organic Compounds Removal
Volatile organic compounds (VOCs), such as aldehydes, ketones, alcohols, esters,

carboxylic acids, aromatics, and aliphatic hydrocarbons, emitted from industrial and
transportation activities are major pollutants to the atmosphere. Most of them are malodorous,
toxic, and contribute to the formation of photochemical smog [270]. A large number of
technologies, including adsorption, absorption, condensation, pyrolysis, and catalytic
oxidation, have been developed to remove VOCs. Catalytic oxidation is capable of
completely destroying VOCs with a wide range of concentrations at relatively low
temperatures. Thus, such a strategy is considered as one of the most effective pathways to
eliminate VOCs. The removal efficiency is determined by the activity, selectivity, and
stability of the catalyst employed. Among the catalysts investigated so far, perovskites are
promising materials for the catalytic combustion of VOCs.
(a) (b)
(c) (d)
Figure 40 (Continued)
(e)
Figure 40. SEM, TEM, and HRTEM images as well as the SAED pattern (inset) of (ac) single-crystal
La0.5Sr0.5MnO3 cubes and (d) La0.5Sr0.5MnO3 nanoparticles. (e) Light-off curves of CH4 oxidation over
the La0.5Sr0.5MnO3 catalysts (C: cube; P: nanoparticle)[128]
(a) (b) (c)
2 m 2 m 2 m
(d) (e) (f)
500 nm 500 nm 1 m
Figure 41. (Continued)

(g)
Figure 41. SEM and TEM images as well as SAED patterns (insets) of (a) CuO_LCO, (b) La2O3_LCO,
(c) Nitrate_LCO, (d) CuO_LSCO, (e) La2O3_LSCO, and (f) Nitrate_LSCO. (g) CH4 consumption rate
as a function of reaction temperature over the La2xSrxCuO4 catalysts. LSO: La2CuO4; LSCO:
LaSrCuO4. The CuO, La2O3, and Nitrate in front of the _ symbol represents the precursor for the
perovskite LCO or LSCO preparation, respectively[268].
Most of researchers believe that the B-site cations play an important role in governing the
catalytic performance of perovskite-type oxides (ABO3) for the complete oxidation of organic
compounds [2,3]. The effect of the B-site cations on the physicochemical and catalytic
properties of lanthanum-based perovskites has been intensively studied, and the perovskite-
type oxides with Mn, Co, Fe or Cr as the B-site element are found to be the most active in
hydrocarbon combustion [2,12,18,20,271]. The catalytic activity of a perovskite is associated
with the chemical composition, lattice defect, redox ability, surface area, and particle
morphology, among which the latter four factors are usually influenced by the preparation
method. For example, Agarwal and Goswami studied the catalytic oxidation of toluene over
LaMO3 (M = Co, Fe, and Cr) prepared by a decomposition method, and observed a catalytic
activity order of LaCoO3 > LaFeO3 > LaCrO3 [272]; Levasseur and Kaliaguine prepared a
series of LaBO3 (B = Co, Mn, Fe) catalysts by adopting an approach called as the reactive
grinding method, and found that the catalytic performance for methanol oxidation follows
the sequence of LaMnO3 > LaCoO3 > LaFeO3 [23]. Working on the catalytic removal of
hexane over LaMO3 (M = Co, Mn, Fe) fabricated via calcination of the metal citrate
precursors at 800 oC for 5 h, however, Spinicci et al. reported a catalytic activity order of
LaFeO3 > LaCoO3 > LaMnO3 [273]. Very recently, we have demonstrated that the catalytic
performance for toluene oxidation follows the sequence of LaFeO3 > LaCoO3 > LaMnO3
[274]. The above discrepancy could be due to the differences in preparation method and the
nature of VOCs.
(a) (b)
5 m 10 m
100
(c)
80 Citrate-YBCO
CH4 conversion (%)
Citrate-HT-YBCO
60 HT-YBCO
HT-sp_Y2O3-YBCO
40
20
0
400 450 500 550 600 650 700 750
o
Temperature ( C)
Figure 42. SEM images and SAED patterns (insets) of (a) YBa2Cu3O7 fabricated hydrothermally at 260
o
C with Y2O3 as yttrium source (HT-sp_Y2O3-YBCO) and (b) YBa2Cu3O7 synthesized by citric-acid-
complexing-assisted hydrothermal treatment at 180 oC (Citrate-HT-YBCO). (c) CH4 conversion versus
reaction temperature over the as-fabricated YBCO catalysts. Citrate-YBCO: YBa2Cu3O7 fabricated by
citric-acid-complexing method; HT-YBCO: YBa2Cu3O7 fabricated hydrothermally at 260 oC with metal
nitrates as precursor [269].
The partial substitution of the A- and/or B-site cations can modify the catalytic behavior
of a perovskite. Up to now, various metal ions, including alkali metal ions (e.g., Na+, K+),
alkaline earth metal ions (e.g., Ba2+, Sr2+, Ca2+), and noble metal ions (e.g., Ag+, Pt4+, Pd2+),
have been used to substitute the A-site cations [17,20,275]. The impact of such a substitution
on catalytic performance may be positive or negative. For example, the partial substitution of
Sr2+ for La3+ can improve the redox process of Mn+ M(n+1)+ (M = Fe, Co, Mn; n = 2, 3), and
then enhance the catalytic activities of La1xSrxMO3 nanoparticles for the combustion of
toluene, ethyl acetate, cyclohexane, and propanol [21,3032]. The incorporation of a noble
metal ion to the perovskite lattice not only improves the performance of the catalyst but also
stabilizes the highly active component (noble metal) against sintering and volatilization. For
instance, Ag-doped LaMnO3 perovskite catalysts show higher activities than the LaMnO3
catalyst in the oxidation of ethanol, n-heptane, and toluene [20]. A dual substitution at the B-
site can give rise to the changes in the B-site redox ion couples and crystal structure stability.
Such conclusions have been confirmed by the results obtained over LaFe1xNixO3 [276],
LaMn1xCoxO3 [277], LaFe1xCoxO3 [13,278,279], and LaNi1xCoxO3 [280] for the catalytic
oxidation of VOCs. Similarly, dual substitutions at both cationic sites can also regulate the
catalytic performance of a perovskite material. For example, Rousseau et al. adopted a sol-gel
method to prepare La1xSrxCo1yFeyO3 nanomaterials, and observed that a significant
improvement is achieved by substituting 30% of La3+ by Sr2+ in LaCoO3 and a slight increase
in catalytic performance for toluene combustion is observed by substituting 20% of cobalt by
iron in La0.7Sr0.3CoO3 [14]. Similar results have been obtained by our group. Due to the
presence of Fe3+OFe4+ couples and a transition of electronic structure, the doping of a
small amount of iron to the La0.6Sr0.4MO3 (M = Co, Mn) lattice (i.e., La0.6Sr0.4M0.9Fe0.1O3)
can greatly improve its catalytic activity [274]. In the case of La1xCexCo1yFeyO3, however,
the Fe-doped catalysts are inferior to the Fe-free counterparts in performance for VOCs
catalytic oxidation. It is likely to be a result due to the deterioration in oxygen mobility and
B-site cation reducibility induced by Fe doping [22].
The major drawback of traditional perovskites is low in surface area. As mentioned in
Section 2, there are various methods to fabricate high-surface-area perovskites. For example,
via a novel strategy of citric acid complexing coupled with a hydrothermal process, high-
surface-area (1633 m2/g) nanosized La1xSrxMO3 (M = Co, Mn, Fe; x = 0, 0.4) and
La0.6Sr0.4M0.9Fe0.1O3 (M = Co, Mn) catalysts have been generated by our group
[3032,274]. All of these nanocatalysts show excellent catalytic performance for the
complete oxidation of typical VOCs (e.g., toluene and ethyl acetate). Under the reactions of
toluene/O2 (ethyl acetate/O2) = 1/4001/100, space velocity (SV) = 500080000 h1, toluene
or ethyl acetate can be totally converted to CO2 and H2O below 400 oC without generation of
any by-products over the La1xSrxMO3 (M = Co, Mn, Fe) and La0.6Sr0.4M0.9Fe0.1O3 (M = Co,
Mn) nanoperovskites. Moreover, these nanoperovskite catalysts perform better than the low-
surface-area (7.29.8 m2/g) counterparts synthesized by the conventional citrate-complexing
method [31]. Usually, a nanoperovskite highly dispersed on a high-surface-area support is
desired for catalytic performance improvement. For example, Alifanti et al. observed
improved catalytic performance over Ce1xZrxO2 (2338 m2/g) supported LaCoO3
nanomaterials for benzene, toluene or methane removal [3335]. Makshina et al.
demonstrated the specific activity over MCM-41-supported LaCoOx catalyst superior to that
over the bulk LaCoOx catalyst [37,38]. Recently, we have examined the effects of SBA-15
mesoporous structure and LaCoO3 loading on catalytic performance for the combustion of
toluene and ethyl acetate. It is observed that in terms of specific activity, the 3540 wt%
LaCoO3/SBA-15 catalysts fabricated in-situ by the hydrothermal method outperform the 35
wt% LaCoO3/SBA-15 catalyst generated by the incipient wetness impregnation method and
the bulk LaCoO3 catalyst derived from the sol-gel method. The excellent performance is
attributed to the good dispersion of highly reducible LaCoO3 species in the LaCoO3/SBA-15
catalyst [40].
Most of the La1xSrxMO3 (M = Fe, Co, Mn, Cr, Ni) catalysts reported in the literature are
polycrystalline, and it is rare to come across single-crystalline La1xSrxMO3 in catalytic
applications. Recently, single-crystalline La1xSrxMnO3+ microcubes and La0.6Sr0.4CoO3
nanowires/nanorods fabricated hydrothermally have been reported by Teng et al [128] and
our group [116,129131]. Such single-crystalline La1xSrxMnO3+ and La0.6Sr0.4CoO3
materials show excellent catalytic performance for the complete oxidation of methane and
VOCs (toluene and ethyl acetate), and their catalytic behaviors are related to the unique
crystalline structure and surface morphology. Under similar reaction conditions, the single-
crystalline La0.5Sr0.5MnO3+ microcubes fabricated by the hydrothermal method exhibit the
T50% and T100% values of 252 and 280 oC, much lower than those (296 and 368 oC,
respectively) over the polycrystalline counterpart prepared by the citric acid-complexing
method [129]. Similarly, the single-crystalline La0.6Sr0.4CoO3 nanorods are also much
superior in catalytic activity (T50% = 235 and T100% = 245 oC) to the polycrystalline
La0.6Sr0.4CoO3 catalyst (T50% = 306 and T100% = 339 oC) for toluene combustion [116].
Obviously, the single-crystalline catalyst outperforms its corresponding polycrystalline one.
Furthermore, the hydrothermal temperature and time as well as the base strength also have a
remarkable effect on the surface composition, Mn oxidation state distribution, oxygen
nonstoichiometry, and surface area of the manganite, and hence modifying the catalytic
performance of single-crystalline La0.5Sr0.5MnO3+ and La0.6Sr0.4MnO3+ [129,131]. A higher
hydrothermal temperature, a shorter hydrothermal time, and a proper amount of KOH favor
the generation of a La0.6Sr0.4MnO3+ material with more amounts of nonstoichiometric
oxygen, surface Mn4+, and better low-temperature reducibility. These factors are responsible
for the improvement in catalytic activity for the removal of VOCs [131].
4.4. Other Applications
Perovskites with specifically morphologies also exhibit other excellent properties in gas
sensoring, adsorption, photocatalysis, CO oxidation, and methanol reforming.
Because of the stability in thermal and chemical atmospheres, a number of perovskite-
type oxides, including LnMO3 (Ln = La, Sm, Nd; M = Co, Mn, Cr, Ni, Fe, Ti, V) have been
used as gas-sensing materials [2]. Owning to the presence of oxygen vacancies in single-
crystalline BaMnO3 nanorods, the conductance peak of the sample in O2 atmosphere shows
an n-type semiconductive behavior. The ultrasensitivity towards O2 of BaMnO3 nanorods can
be attributed to the smaller size and higher specific surface area, especially the clean surface
on the nanorods synthesized without using capping reagents [152]. The sensors based on the
cube-like CaSnO3 show much stronger responsing ability to ethanol, ammonia, and acetone
than the sensor based on the irregularly morphological bulk CaSnO3 particles, with the
sensitivity of about 1.45, 1.3, and 1.15, respectively [180]. The gas-sensing performance is
mainly mediated by the surface properties. On the one hand, there are more active centers for
trapping the target gas in microcube-like CaSnO3 (due to their higher surface area) than those
in bulk CaSnO3. On the other hand, the cube-like morphology provides a high fraction of
well-defined (100) crystal planes that are more sensitive to ethanol. Therefore, CaSnO3
microcubes exhibit sensitivity towards ethanol higher than those towards ammonia and
acetone. Due to a smaller size, the sensor based on BaTiO3 nanocubes shows a high
sensitivity towards a trace amount of hydrogen peroxide [186].
So far, the adsorption of NOx on superconducting compounds, such as Y-Ba-Cu-O [281]
and Mn-Zr-Y [282], has been reported. A large amount of NOx can be removed by these
active adsorbents. However, in the presence of CO2, the formation of stable carbonate leads to
a deactivation of these adsorptive materials. Interestingly, incorporation of alkali metal ions to
the perovskite lattice is capable of inhibiting carbonation. Perovskites like BaSnO3 [283],
La0.8Cs0.2MnO3 [284], and La0.8Cs0.2Mn0.8Ga0.2O3 [284] display excellent performance for
NOx adsorption even in the presence of 5% CO2. The discrepancy in carbonation behavior of
bulk alkali metal oxide and alkali metal-doped perovskite can be reasonably explained by the
fact that strongly held carbonates are formed on the former, whereas this is not the case on the
latter. Actually, nitrate adspecies rather than carbonate adspecies tend to be formed over the
alkali metal-doped perovskite material [283]. In addition, owing to its higher surface area and
cotton-like powders, the La0.8Cs0.2Mn0.8Ga0.2O3 sample containing nanorods show a higher
NOx adsorption capacity than the nanrod-free La0.8Cs0.2MnO3 sample [284].
Over the BiFeO3 nanosized photocatalyst, more than 90% methyl orange (MO) can be
decolorized after 8 h under UV-visible irradiation. Nevertheless, the MO degradation is
maintained at about 70% over the bulk BiFeO3 photocatalyst after 16 h of UV-visible
illumination. The better degradation efficiency of the nanoperovskite catalyst is attributed to
its higher surface area. Moreover, compared to the conventional TiO2 material, nanosized
BiFeO3 shows visible-light photocatalytic activity [11]. It is reported that single-crystalline
KNbO3 nanowires perform much better than the KNbO3 nanoparticles and nanocubes in the
photocatalytic water splitting reaction, giving a highest H2 evolution rate of 5.17 mmol/(gcat
h) from a CH3OH aqueous solution. The excellent performance of nanowire-like
photocatalyst is associated with its high surface area and good crystallinity [115]. Although
SrTiO3 is known to be inactive in photoinduced hydrophilic reactions, the surface of single-
crystalline SrTiO3 nanorods can become superhydrophilic by the oxidation of the surface
organic contaminants under UV illumination because their initial water contact angles are less
than 3o [161]. Therefore, the superhydrophilicity of the SrTiO3 nanorod arrays originates from
the presence of surface nanostructure.
As mentioned in Section 2.1, the sol-gel method with citric acid as the complexing agent
is a simple pathway to prepare polycrystalline nanoparticles showing good catalytic activities.
Unfortunately, such a nanocatalyst tends to sinter upon calcination at high temperatures and
in a long-duration run. Hence, fabricating a catalyst with good thermal stability is still a
challenge. Recently, Zhu and co-workers claimed that polycrystalline nanowire-like catalyst
shows much better thermal stability than the polycrystalline irregularly morphological
nanosized counterpart fabricated by citrate-complexing method [87,88]. A complete
conversion of CO over polycrystalline LaCoO3 nanowires and nanoparticles calcined at 600
o
C for 5 h can be achieved at 190 and 170 oC, respectively. Obviously, the nanoparticles
outperform the nanowires, a result due to a higher surface area (20.5 m2/g) of the latter than
that (14.6 m2/g) of the former. Calcination at 750 oC for 48 h causes the surface area of
LaCoO3 nanowires and nanoparticles to drop to 13.9 and 12.1 m2/g, respectively, resulting in
a catalytic performance (T100% = 220 oC) over the former much better than that (T100% = 310
o
C) over the latter [87]. Similar results are also observed over the polycrystalline LaSrCuO4
nanowires and nanoparticles [88]. For the steam reforming of methanol to CO2 and H2, the
nanofibrous La2CuO4 catalyst shows a much better activity than the bulk one synthesized by
citrate-complexing method [84]. At 150 oC and steam/methanol molar ratio = 1.3, methanol
can be completely converted to H2 and CO2 without CO generation. The excellent catalytic
performance is associated with the higher surface area, the co-presence of Cu2+/Cu+, and
oxygen vacancies in La2CuO4 nanofibers.
5. PERSPECTIVES AND CONCLUDING REMARKS

It is known that perovskite-type oxides and related materials with various morphologies,
such as tubular, spherical, wire-, rod-, and cube-like micro/nanoarchitectures, can be
fabricated successfully using the hard-templating, hydrothermal, and molten salt synthesis
routes. However, no reports are seen in the literature on the synthesis of micro/nanotubular
perovskites (except for titanates) by means of the hydrothermal and molten salt approaches.
With close-packed PMMA, PS or silica microspheres as hard template in an EG-methanol
solution of the metal nitrates, one can generate three-dimensionally ordered macroporous
perovskites. Nevertheless, such a method is not suitable for the preparation of mesoporous
perovskites. It has been demonstrated that the perovskites and related materials with specific
morphologies or porous structures possess unique physicochemical (e.g., magnetic, electrical,
optical, adsorption, and gas sensing) properties, and show outstanding catalytic activity for
the removal of nitrogen oxides, methane, and VOCs. The unusual performance is associated
with the factors, such as particle size, morphology, nonstoichiometric oxygen, exposed crystal
plane, crystallinity, surface area, and pore structure.
Since catalytic reaction usually takes place on the surface of a catalyst, a more amount of
exposed active crystal planes of the catalyst favors the heterogeneous catalysis process and
hence results in an improvement in catalytic performance. The traditional preparation
methods usually give rise to catalysts with most of the exposed crystal planes being low in
surface energy. Such stable crystal planes do not favor the enhancement in catalytic activity.
The 1D perovskites derived from hard-templating methods mentioned above are
predominantly polycrystalline. It has been observed that the single-crystalline perovskites
exhibit better catalytic activity than the polycrystalline counterparts, although the inherent
reasons are not fully clarified. As we know, the quantity of crystal planes (especially active
crystal planes) exposed on the surface of a single-crystalline material with specific
morphology is fewer than that of the corresponding polycrystalline one with irregular
morphology. To optimize the construction of active sites of a single crystallite by maximizing
the exposure ratio of more reactive crystal planes and minimizing the exposure ratio of less
reactive crystal planes, it is highly needed to develop a fabrication strategy to accurately
control the growth of such a single-crystalline material with the desired architecture.
As pointed out in the previous sections, the hydrothermal (or solvothermal) and molten
salt routes are promising in specifically morphological perovskite fabrication. Due to the
variety of influencing factors (precursor nature, solvent, base strength, temperature and time,
molten salt nature, etc.), controlled morphofabrication of perovskites is still a big challenge
even by adopting the two synthesis techniques. Although the surfactant-templating method is
difficult in generating stoichiometric perovskites, the surfactant employed may play a
morphology-oriented role. An appropriate modification on the synthesis parameters might
lead to the successful production of desired perovskites. Therefore, it is possible to fabricate

single-crystalline perovskites and related compounds with specific morphologies via a
surfactant-directing solvothermal or molten salt process. Further work is needed on the
sophisticated control of high percentage of exposed active crystal planes. It could be achieved
through the adjusting of crystal morphology. In addition, extensive and intensive researches
are also required to clarify the relationship between the morphology, crystal plane exposure
ratio or single crystallinity and catalytic performance.
It has been reported that, compared to the bulk transition metal oxides, their mesoporous
counterparts exhibit much better catalytic performance due to the mesopore structure and high
surface area. Based on this viewpoint, one can envisage that macro- or mesoporous perovskite
materials would show higher catalytic activity than the bulk ones. Therefore, it is greatly
desirable to establish an effective strategy to fabricate porous perovskites. The polymer or
silica microsphere-templating method with a surfactant (e.g., P123 or F127) as soft template
is promising in the fabrication of three-dimensionally ordered macroporous perovskites with
mesoporous walls. The rise in glassy temperature of the polymer microbeads is beneficial for
the generation of macro- or mesoporous perovskites because formation of a perovskite
structure needs a high-temperature calcination process. The successful generation of
mesoporous SrTiO3 due to the use of the KLE3739 [233] might be useful in fabricating other
mesoporous perovskites. Of course, if the mesopores of a hard template could be fully filled
by the homogenous precursor solutions, one would be able to obtain mesoporous perovskites
and related materials through the nanocasting pathway.
REFERENCES
[1] Bell, A. T. Science 2003, 299, 1688-1691.
[2] Pea, M. A.; Fierro, J. L. G. Chem. Rev. 2001, 101, 1981-2017.
[3] Tejuca, L. J.; Fierro, J. L. G. Properties and Application of Perovskite Type Oxides;
Marcel Dekker: New York, 1993.
[4] Voorhoeve, R. J. H.; Jr., Johnson, D. W.; Remeika, J. P.; Gallagher, P. K. Science
1977, 195, 827-833.
[5] Zwinkels, M. F. M.; Jrs, S. G.; Menon, G. P.; Griffin, T. A. Catal. Rev. -Sci. Eng.
1993, 35, 319-358.
[6] Zhang, R. D.; Kaliaguine, S. In Heterogeneous Catalysis Research Progress; Gunther,
M. B.; Ed.; Nova Publishers: New York, 2008; pp. 1-79.
[7] OBrien, S.; Brus. L.; Murray, C. B. J. Am. Chem. Soc. 2001, 123, 12085-12086.
[8] Mao, Y. B.; Wong, S. S. Adv. Mater. 2005, 17, 2194-2199.
[9] Xu, H.; Hu, X. L.; Zhang, L. Z. Cryst. Growth Des. 2008, 8, 2061-2065.
[10] Shabbir, G.; Qureshi, A. H.; Saeed, K. Mater. Lett. 2006, 60, 3706-3709.
[11] Gao, F.; Chen, X. Y.; Yin, K. B.; Dong, S.; Ren, Z. F.; Yuan, F.; Yu, T.; Zou, Z. G.;
Liu, J. M. Adv Mater. 2007, 19, 2889-2892.
[12] Spinicci, R.; Faticanti, M.; Marini, P.; Rossi, S. D.; Porta, P. J. Mol. Catal. A 2003,
197, 147-155.
[13] Merino, N. A.; Barbero, B. P.; Ruiz, P.; Cads, L. E. J. Catal. 2006, 240, 245-257.
[14] Rousseau, S.; Loridant, S.; Delichere, P.; Boreave, A.; Deloume, J. P.; Vernoux P.
Appl. Catal. B 2009, 88, 438-447.
[15] Blasin-Aub, V.; Belkouch, J.; Monceaux, L. Appl. Catal. B 2003, 43, 175-186.
[16] Sinquin, G.; Petit, C.; Hindermann, J. P.; Kiennemann, A. Catal. Today 2001, 70, 183-
196.
[17] Bialobok, B.; Trawczyski, J.; Mita, W.; Zawadzki, M. Appl. Catal. B 2007, 72, 395-
403.
[18] Irusta, S.; Pina, M. P.; Menndez, M.; Santamara, J. J. Catal. 1998, 179, 400-412.
[19] Wang, X.; Zhuang, J.; Peng, Q.; Li, Y. D. Nature 2005, 437, 121-124.
[20] Musialik-Piotrowska A.; Landmesser, H. Catal. Today 2008, 137, 357-361.
[21] Huang, H. F.; Liu, Y. Q.; Tang, W.; Chen, Y. F. Catal. Commun. 2008, 9, 55-59.
[22] Levasseur, B.; Kaliaguine, S. Appl. Catal. B 2009, 88, 305-314.
[23] Levasseur, B.; Kaliaguine, S. Appl. Catal. A 2008, 343, 29-38.
[24] Giannakas, A. E.; Ladavos, A. K.; Pomonis, P. J. Appl. Catal. B 2004, 49, 147-158.
[25] Giannakas, A. E.; Leontiou, A. A.; Ladavos, A. K.; Pomonis, P. J. Appl. Catal. A 2006,
309, 254-262.
[26] He, H.; Liu, M.; Dai, H. X.; Qiu, W. G.; Zi, X. H. Catal. Today 2007, 126, 290-295.
[27] Lima, R. K. C.; Batista, M. S.; Wallau, M.; Sanches, E. A.; Mascarenhas, Y. P.;
Urquieta-Gonzlez, E. A. Appl. Catal. B 2009, 90, 441-450.
[28] Bruechey, R. L.; Morse, D. E. Angew. Chem. Int. Ed. 2006, 45, 6564-6566.
[29] Niederberger, M.; Pinna, N.; Polleux, J.; Antonietti, M. Angew. Chem. Int. Ed.
2004,43, 2270-2273.
[30] Niu, J. R.; Liu, W.; Dai, H. X.; He, H.; Zi, X. H.; Li, P. H. Chin. Sci. Bull. 2006, 51,
1673-1681.
[31] Niu, J. R.; Deng, J. G.; Liu, W.; Zhang, L.; Wang, G. Z.; Dai, H. X.; He, H.; Zi, X. H.
Catal. Today 2007, 126, 420-429.
[32] Deng, J. G.; Zhang, L.; Dai, H. X.; He, H.; Au, C. T. Ind. Eng. Chem. Res. 2008, 47,
8175-8183.
[33] Alifanti, M.; Florea, M.; Somacescu, S.; Prvulescu, V. I. Appl. Catal. B 2005, 60, 33-
39.
[34] Alifanti, M.; Florea, M.; Prvulescu, V. I. Appl. Catal. B 2007, 70, 400-405.
[35] Alifanti, M.; Florea, M.; Cortes-Corberan, V.; Endruschat, U.; Delmon, B.; Prvulescu,
V. I. Catal. Today 2006, 112, 169-173.
[36] Nguyen, S. V.; Szabo, V.; Trong On, D.; Kaliaguine, S. Micropor. Mesopor. Mater.
2002, 54, 51-61.
[37] Makshina, E. V.; Nesterenko, N. S.; Siffert, S.; Zhilinskaya, E. A.; Aboukais, A.;
Romanovsky, B. V. Catal. Today 2008, 131, 427-430.
[38] Makshina, E. V.; Sirotin, S. V.; van den Berg, M. W. E.; Klementiev, K. V.;
Yushchenko, V. V.; Mazo, G. N.; Grnert, W.; Romanovsky, B. V. Appl. Catal. A
2006, 312, 59-66.
[39] Yi, N.; Cao, Y.; Su, Y.; Dai, W. L.; He, H. Y.; Fan, K. N. J. Catal. 2005, 230, 249-253.
[40] Deng, J. G.; Zhang, L.; Dai, H. X.; Au, C. T. Appl. Catal. A 2009, 352, 43-49.
[41] Hurst, S. J.; Payne, E. K.; Qin, L. D.; Mirkin, C. K. Angew. Chem. Int. Ed. 2006, 45,
2672-2692.
[42] Xia, Y. N.; Yang, P. D.; Sun, Y. G.; Wu, Y. Y.; Mayers, B.; Gates, B.; Yin, Y. D.;
Kim, F.; Yan, H. Q. Adv. Mater. 2003, 15, 353-389.
[43] Murphy, C. J.; Jana, N. R. Adv. Mater. 2002, 14, 80-82.

[44] Rao, C. N. R.; Deepak, F. L.; Gundiah, G.; Govindaraj, A. Prog. Solid State Chem.
2003, 31, 5-147.
[45] Xia, Y. N.; Rogers, J. A.; Paul, K. E.; Whitesides, G. M. Chem. Rev. 1999, 99, 1823-
1848.
[46] Cao, G. Z.; Liu, D. W. Adv. Colloid Interface Sci. 2008, 136, 45-64.
[47] Martin, C. R. Chem. Mater. 1996, 8, 1739-1746.
[48] Martin C. R. Science 1994, 266, 1961-1966.
[49] Zhou, Y. K.; Shen, C. M.; Huang, J.; Li, H. L. Mater. Sci. Eng. B 2002, 95, 77-82.
[50] Keller, N.; Pham-Huu, C.; Estourn, C.; Grenche, J. M.; Ehret, G.; Ledoux, M. J.
Carbon 2004, 42, 1395-1399.
[51] Ji, G. B.; Tang, S. L.; Xu, B. L.; Gu, B. X.; Du, Y. W. Chem. Phys. Lett. 2003, 379,
484-489.
[52] Zhang, G. Y.; Guo, B.; Chen, J. Sensor. Actuat. B 2006, 114, 402-409.
[53] Zhao, L. L.; Steinhart, M.; Yosef, M.; Lee, S. K.; Schlecht, S. Sensor. Actuat. B 2005,
109, 86-90.
[54] Yang, Z.; Huang, Y.; Dong, B.; Li, H. L.; Shi, S. Q. Appl. Phys. A 2006, 84, 117-122.
[55] Yang, Z.; Huang, Y.; Dong, B.; Li, H. L. Appl. Phys. A 2005, 81, 453-457.
[56] Yang, Z.; Huang, Y.; Dong, B.; Li, H. L. Mater. Res. Bull. 2006, 41, 274-281.
[57] Yang, Z.; Huang, Y.; Dong, B.; Li, H. L. J. Solid State Chem. 2005, 178, 1157-1164.
[58] Zhang, X. Y.; Dai, J. Y.; Lai, C. W. Prog. Solid State Chem. 2005, 33, 147-151.
[59] Ma, X. Y.; Zhang, H.; Xu, J.; Niu, J. J.; Yang, Q.; Sha, J.; Yang, D. R. Chem. Phys.
Lett. 2002, 363, 579-582.
[60] Zhang, X. Y.; Zhao, X.; Lai, C. W.; Wang, J.; Tang, X. G.; Dai, J. Y. Appl. Phys. Lett.
2004, 85, 4190-4192.
[61] Tagliazucchi, M.; Sanchez, R. D.; Troiani, H. E.; Calvo, E. J. Solid State Commun.
2006, 137, 212-215.
[62] Wei, J.; Xue, D. S.; Xu, Y. Scripta Mater. 2008, 58, 45-48.
[63] Zhang, X. Y.; Lai, C. W.; Zhao, X.; Wang, D. Y.; Dai, J. Y. Appl. Phys. Lett. 2005, 87,
143102(1)-143102(3).
[64] Wang, J.; Manivannan, A.; Wu, N. Q. Thin Solid Films 2008, 517, 582-587.
[65] Hernandez B. A.; Chang, K. S.; Fisher, E. R.; Dorhout, P. K. Chem. Mater. 2002, 14,
480-482.
[66] Liu, L. F.; Ning, T. Y.; Ren, Y.; Sun, Z. H.; Wang, F. F.; Zhou, W. Y.; Xie, S. S.;
Song, L.; Luo, S. D.; Liu, D. F.; Shen, J.; Ma, W. J.; Zhou, Y. L. Mater. Sci. Eng. B
2008, 149, 41-46.
[67] Rrvik, P. M.; Tadannaga, K.; Tatsumisago, M.; Grande, T.; Einarsrud, M. A. J. Eur.
Ceram. Soc. 2009, 29, 2575-2579.
[68] Hua, Z. H.; Li, D.; Fu, H. Acta Phys. -Chim. Sin. 2009, 25, 145-149.
[69] Hua, Z. H.; Yang, P.; Huang, H. B.; Wan, J. G.; Yu, Z. Z.; Yang, S. G.; Lu, M.; Gu, B.
X.; Du, Y. W.; Mater. Chem. Phys. 2008, 107, 541-546.
[70] Singh, S.; Krupanidhi. S. B. Phys. Lett. A 2007, 367, 356-359.
[71] Liu, W.; Sun, X. H.; Han, H. W.; Li, M. Y.; Zhao, X. Z. Appl. Phys. Lett. 2006, 89,
163122(1)-163122(3).
[72] Hu, Z. A.; Wu, H. Y.; Shang, X. L.; L, R. J.; Li, H. L. Mater. Res. Bull. 2006, 41,
1045-1051.
[73] Hsu, M. C.; Leu, I. C.; Sun, Y. M.; Hon, M. H. J. Solid State Chem. 2006, 179, 1420-
1424.
[74] Wen, T. L.; Zhang, J.; Chou, T. P.; Limmer, S. J.; Cao, G. Z. J. Sol-Gel Sci. Technol.
2005, 33, 193-200.
[75] Limmer, S. J.; Seraji, S.; Forbess, M. J.; Wu, Y.; Chou, T. P.; Nguyen, C.; Cao, G.. Z.
Adv. Mater. 2001, 13, 1269-1272.
[76] Kim, J.; Yang, S. A.; Choi, Y. C.; Han, J. K.; Jeong, K. O.; Yun, Y. J.; Kim, D. J.;
Yang, S. M.; Yoon, D.; Cheong, H.; Chang, K. S.; Noh, T. W.; Bu, S. D. Nano Lett.
2008, 8, 1813-1818.
[77] Nourmohammadi, A.; Bahrevar, M. A.; Hietschold, M. Mater. Lett. 2008, 62, 3349-
3351.
[78] Nourmohammadi, A.; Bahrevar, M. A.; Hietschold, M. J. Alloy. Compd. 2009, 473,
467-472.
[79] Hsu, M. C.; Leu, I. C.; Sun, Y. M.; Hon, M. H. J. Solid State Chem. 2006, 179, 1421-
1425.
[80] Ugarte, D.; Chtelain, A.; de Heer, W. A. Science 1996, 274, 1897-1899.
[81] Ajayan, P. M.; Lijima, S. Nature 1993, 361, 333-334.
[82] Dujardin, E.; Ebbessen, T. W.; Hiura, H.; Tanigaki, K. Science 1994, 265, 1850-1852.
[83] Gao, L. Z.; Chua, H. T.; Kawi, S. J. Solid State Chem. 2008, 181, 2804-2807.
[84] Gao, L. Z.; Sun, G. B.; Kawi, S. J. Solid State Chem. 2008, 181, 7-13.
[85] Gao, L. Z.; Wang, X. L.; Chua, H. T.; Kawi, S. J. Solid State Chem. 2006,179, 2036-
2040.
[86] Eder, D.; Kinloch, I. A.; Windle, A. H. Chem. Commun. 2006, 1448-1450.
[87] Teng, F.; Liang, S. H.; Gaugeu, B.; Zong, R. L.; Yao, W. Q.; Zhu, Y. F. Catal.
Commun. 2007, 8, 1748-1754.
[88] Teng, F.; Gaugeu, B.; Liang, S. H.; Zhu, Y. F. Appl. Catal. A 2007, 328, 156-162.
[89] Dolz, M. I.; Bast, W.; Antonio, D.; Pastoriza, H.; Curiale, J.; Snchez, R. D.; Leyva, A.
G. J. Appl. Phys. 2008, 103, 083909(1)-083909(5).
[90] Leyva, A. G.; Stoliar, P.; Rosenbusch, M.; Lorenzo, V.; Levy, P.; Albonetti, C.;
Cavallini, M.; Biscarini, F.; Troiani, H. E.; Curiale, J.; Sanchez, R. D. J. Solid State
Chem. 2004, 177, 3949-3953.
[91] Curiale, J.; Snchez, R. D.; Troiani, H. E.; Pastoriza, H.; Levy, P.; Leyva, A. G.
Physica B 2004, 354, 98-103.
[92] Sacanell, J.; Bellino, M. G.; Lamas, D. G.; Leyva, A.G.. Physica B 2007, 398, 341-343.
[93] Limmer, S. J.; Hubler, T. L.; Cao, G. Z. J. Sol-Gel Sci. Technol. 2003, 26, 577-581.
[94] De Jong, K. P.; Geus, J. W. Catal. Rev. -Sci. Eng. 2000, 42, 481-510.
[95] Yang, S.; Chen, X.; Kusunoki, M.; Yamamoto, K.; Iwanaga, H.; Motojima, S. Carbon
2005, 43, 916-922.
[96] Otsuka, K.; Kobayashi, S.; Takenaka, S. Appl. Catal. A 2001, 210, 371-379.
[97] Takenaka, S.; Shigeta, Y.; Tanabe, E.; Otsuka, K. J. Phys. Chem. B 2004, 108, 7656-
7664.
[98] Takenaka, S.; Ishida, M.; Serizawa, M.; Tanabe, E.; Otsuka, K. J. Phys. Chem. B 2004,
108, 11464-11472.
[99] Ueda, W.; Sadakane, M.; Ogihara, H. Catal. Today 2008, 132, 2-8.
[100] Ogihara, H.; Sadakane, M.; Nodasaka, Y.; Ueda, W. Chem. Mater. 2006, 18, 4981-
4983.
[101] Ogihara, H.; Sadakane, M.; Nodasaka, Y.; Ueda, W. Chem. Lett. 2007, 36, 258-259.
[102] Ogihara, H.; Sadakane, M.; Wu, Q.; Nodasaka, Y.; Ueda, W. Chem. Commun. 2007,
4047-4049.
[103] Zhang, S. Y.; Jiang, F. S.; Qu, G.; Lin, C. Y.; Mater. Lett. 2008, 62, 2225-2228.
[104] Kwan, S.; Kim, F.; Akana, J.; Yang, P. D. Chem. Commun. 2001, 447-448.
[105] Li, M.; Schnablegger, H.; Mann, S. Nature 1999, 402, 393-395.
[106] Xu, G.; Ren, Z. H.; Du, P. Y.; Weng, W. J.; Shen, G.; Han, G. R. Adv. Mater. 2005, 17,
907-910.
[107] Cheng, H.; Ma, J.; Zhu, B.; Cui, Y. J. Am. Ceram. Soc. 1993, 76, 625-629.
[108] Vayssieres, L. Int. J. Nanotechnol. 2004, 1, 1-41.
[109] Joshi, U. A.; Lee, J. S. Small 2005, 1, 1172-1176.
[110] Joshi, U. A.; Yoon, S.; Baik, S.; Lee, J. S. J. Phys. Chem. B 2006, 110, 12249-12256.
[111] Gu, H. S.; Hu, Y. M.; Wang, H.; Yang, X. R.; Hu, Z. L.; Yuan, Y.; You, J. J Sol-Gel
Sci. Technol. 2007, 42, 293-297.
[112] Gu, H. S.; Hu,Y. M.; You, J.; Hu, Z. L.; Yuan, Y.; Zhang, T. J. J. Appl. Phys. 2007,
101, 024319(1)-024319(7).
[113] Magrez, A.; Vasco, E.; Seo, J. W.; Dieker, C.; Setter, N.; Forr, L. J. Phys. Chem. B
2006, 110, 58-61.
[114] Ke, T. Y.; Chen, H. A.; Sheu, H. S.; Yeh, J. W.; Lin, H. N.; Lee, C. Y.; Chiu, H. T. J.
Phys. Chem. C 2008, 112, 8827-8831.
[115] Ding, Q. P.; Yuan, Y. P.; Xiong, X.; Li, R. P.; Huang, H. B.; Li, Z. S.; Yu, T.; Zou, Z.
G.; Yang, S. G. J. Phys. Chem. C 2008, 112, 18846-18848.
[116] Deng, J. G.; Zhang, L.; Dai, H. X.; He, H.; Au, C. T. Catal. Lett. 2008, 123, 294-300.
[117] Cheng, L.; Shao, Q.; Shao, M. W.; Wei, X. W.; Wu, Z. C. J. Phys. Chem. C 2009, 113,
1764-1768.
[118] Zhu, H. L.; Yang, D. R.; Yu, G. X.; Zhang, H.; Jin, D. L.; Yao, K. H. J. Phys. Chem. B
2006, 110, 7631-7634.
[119] OBrien, S.; Brus, L.; Murray, C. B. J. Am. Chem. Soc. 2001, 123, 12085-12086.
[120] Urban, J. J.; Yun, W. S.; Gu, Q.; Park, H. J. Am. Chem. Soc. 2002, 124, 1186-1187.
[121] Urban, J. J.; Spanier, J. E.; Ouyang, L.; Yun, W. S.; Park, H. Adv. Mater. 2003, 15,
423-426.
[122] Yun, W. S.; Urban, J. J.; Gu, Q.; Park, H. Nano Lett. 2002, 2, 447-450.
[123] Zhu, D. L.; Zhu, H.; Zhang, Y. H. J. Cryst. Growth 2003, 249, 172-175.
[124] Liu, J. B.; Wang, H.; Zhu, M. K.; Wang, B.; Yan, H. Mater. Res. Bull. 2003, 38, 817-
822.
[125] Zhu, D.; Zhu, H.; Zhang, Y. H. J. Phys.-Condens. Mater. 2002, 14, L519-L524.
[126] Zhang, T.; Jin, C. G.; Qian, T.; Lu, X. L.; Bai, J. M.; Li, X. G. J. Mater. Chem. 2004,
14, 2787-2789.
[127] Urban, J. J.; Ouyang, L.; Jo, M. H.; Wang, D. S.; Park, H. Nano Lett. 2004, 4, 1547-
1550.
[128] Teng, F.; Han, W.; Liang, S. H.; Gaugeu, B.; Zong, R. L.; Zhu, Y. F. J. Catal. 2007,
250, 1-11.
[129] Deng, J. G.; Zhang, Y.; Dai, H. X.; Zhang, L.; He, H.; Au, C. T. Catal. Today 2008,
139, 82-87.
[130] Deng, J. G.; Zhang, L.; Dai, H. X.; He, H.; Au, C. T. J. Mol. Catal. A 2009, 299, 60-67.
[131] Deng, J. G.; Zhang, L.; Dai, H. X.; Au, C. T. Catal. Lett. 2009, 130, 622-629.
[132] Mao, Y. B.; Banerjee, S.; Wong, S. S. Chem. Commun. 2003, 408-409.
[133] Yang, Y.; Wang, X. H.; Sun, C. K.; Li, L. T. J. Am. Ceram. Soc. 2008, 91, 3820-3822.
[134] Li, Y.; Gao, X. P.; Li, G. R.; Pan, G. L.; Yan, T. Y.; Zhu, H. Y. J. Phys. Chem. C 2009,
113, 4386-4394.
[135] Bao, N. Z.; Shen, L. M.; Srinivasan, G.; Yanagisawa, K.; Gupta, A. J. Phys. Chem. C
2008, 112, 8634-8642.
[136] Kang, S. O.; Park, B. H.; Kim, Y. I. Cryst. Growth Des. 2008, 8, 3180-3186.
[137] Pinna, N.; Niederberger, M. Angew. Chem. Int. Ed. 2008, 47, 5292-5304.
[138] Jun, Y. W.; Choi, J. S.; Cheon, J. Angew. Chem. Int. Ed. 2006, 45, 3414-3439.
[139] Wang, H.; Wang, L.; Liu, J. B.; Wang, B.; Yan, H. Mater. Sci. Eng. B 2003, 99, 495-
498.
[140] Hou, L.; Hou, Y. D.; Song, X. M.; Zhu, M. K.; Wang, H.; Yan, H. Mater. Res. Bull.
2006, 41, 1330-1336.
[141] Hou, Y. D.; Zhu, M. K.; Hou, L.; Liu, J. B.; Tang, J. L.; Wang, H.; Yan, H. J. Cryst.
Growth 2005, 273, 500-503.
[142] Zhu, X. H.; Wang, J. Y.; Zhang, Z. H.; Zhu, J. M.; Zhou, S. H.; Liu, Z. G.; Ming, N. B.
J. Am. Ceram. Soc. 2008, 91, 2683-2689.
[143] Hou, Y. D.; Hou, L.; Zhang, T. T.; Zhu, M. K.; Wang, H.; Yan, H. J. Am. Ceram. Soc.
2007, 90, 1738-1743.
[144] Hayashi, Y.; Kimura, T.; Yamaguchi, T. J. Mater. Sci. 1986, 21, 757-762.
[145] Mao, Y. B.; Banerjee, S.; Wong, S. S. J. Am. Chem. Soc. 2003, 125, 15718-15719.
[146] Deng, H.; Qiu, Y. C.; Yang, S. H. J. Mater. Chem. 2009, 19, 976-982.
[147] Qi, S. Y.; Feng, J.; Xu, X. D.; Wang, J. P.; Hou, X. Y.; Zhang, M. L. Chem. Res. Chin.
Univ. 2008, 24, 672-674.
[148] Feng, J.; Qi, S. Y.; Hou, X. Y.; Zhang, M. L. Solid State Sci. 2007, 9, 1033-1035.
[149] Qi, S. Y.; Feng, J.; Xu, X. D.; Wang, J. P.; Hou, X. Y.; Zhang, M. L. J. Alloy. Compd.
2009, 478, 317-320.
[150] Liu, H.; Hu, C. G.; Wang, Z. L. Nano Lett. 2006, 6, 1535-1540.
[151] Hu, C. G.; Liu, H.; Lao, C. S.; Zhang, L. Y.; Davidovic, D.; Wang, Z. L. J. Phys.
Chem. B 2006, 110, 14050-14054.
[152] Wang, N.; Hu, C. G.; Xia, C. H.; Feng, B.; Zhang, Z. W.; Xi, Y. Appl. Phys. Lett. 2007,
90, 16311(1)-16311(3).
[153] Jiang, W. Q.; Gong, X. Q.; Chen, Z. Y.; Hu, Y.; Zhang, X. Z.; Gong, X. L. Ultrason.
Sonochem. 2007, 14, 208-212.
[154] Lu, X. F.; Zhang, D. L.; Zhao, Q. D.; Wang, C.; Zhang, W. J.; Wei, Y. Macromol.
Rapid Commun. 2006, 27, 76-80.
[155] Wang,Y.; Santiago-Avils, J. J. Nanotechnology 2004, 15, 32-36.
[156] Maensiri, S.; Nuansing, W.; Klinkaewnarong, J.; Laokul, P.; Khemprasit, J. J. Colloid
Interf. Sci. 2006, 297, 578-583.
[157] Yuh, J.; Nino, J. C.; Sigmund, W. M. Mater. Lett. 2005, 59, 3645-3647.
[158] Li, C.; Lei, B.; Luo, Z. C.; Han, S.; Liu, Z. Q.; Zhang, D. H.; Zhou, C. W. Adv. Mater.
2005, 17, 1548-1553.
[159] Lei, B.; Li, C.; Zhang, D. H.; Han, S.; Zhou, C. W. J. Phys. Chem. B 2005, 109, 18799-
18803.
[160] Han, S.; Li, C.; Liu, Z. Q.; Lei, B.; Zhang, D. H.; Jin, W.; Liu, X. L.; Tang, T.; Zhou,
C. W. Nano Lett. 2004, 4, 1241-1246.
[161] Miyauchi, M. J. Phys. Chem. C 2007, 111, 12440-12445.

[162] Nelson, J. A.; Wagner, M. J. J. Am. Chem. Soc. 2003, 125, 332-333.
[163] Priboi, I.; Makovec, D.; Drofenik, M. Chem. Mater. 2005, 17, 2953-2958.
[164] Peng, X. G.; Manna, L.; Yang, W. D.; Wickham, J.; Scher, E.; Kadavanich, A.;
Alivisatos, A. P. Nature 2000, 404, 59-61.
[165] Buhro, W. E.; Colvin, V. L. Nat. Mater. 2003, 2, 138-139.
[166] Chu, X. F.; Jiang, D. L.; Zheng, C. M. Sensor. Actuat. B 2007, 123, 793-797.
[167] Sun, Y. G.; Xia, Y. N. Science 2002, 298, 2176-2179.
[168] Jin, S.; Tiefel, T. H.; McCormack, M.; Fastnacht, R. A.; Ramesh, R.; Chen, L. H.
Science 1994, 264, 413-415.
[169] von Helmolt, R.; Wecker, J.; Holzapfel, B.; Schultz, L.; Samwer, K. Phys. Rev. Lett.
1993, 71, 2331-2333.
[170] Chen, Y.; Yuan, H. M.; Tian, G.; Zhang, G. H.; Feng, S. H. J. Solid State Chem. 2007,
180, 167-172.
[171] Spooren, J.; Walton, R. I. J. Solid State Chem. 2005, 178, 1683-1691.
[172] Spooren, J.; Walton, R. I.; Millange, F. J. Mater. Chem. 2005, 15, 1542-1551.
[173] Chai, P.; Liu, X. J.; Wang, Z. L.; Lu, M. F.; Cao, X. Q.; Meng, J. Cryst. Growth Des.
2007, 7, 2568-2575.
[174] Li, J. Q.; Sun, W. A.; Ao, W. Q.; Tang, J. N. J. Magn.Magn. Mater. 2006, 302, 463-
466.
[175] Spooren, J.; Rumplecker, A.; Millange, F.; Walton, R. I. Chem. Mater. 2003, 15, 1401-
1403.
[176] Wang, X.; Li, Y. D. Angew. Chem. Int. Ed. 2002, 41, 4790-4793.
[177] Zheng, W. J.; Liu, R. H.; Peng, D. K.; Meng, G. Y. Mater. Lett. 2000, 43, 19-22.
[178] Sun, C.; Xing, X. R.; Chen, J.; Deng, J. X.; Li, L.; Yu, R. B.; Qiao, L. J.; Liu, G. R.
Eur. J. Inorg. Chem. 2007, 1884-1888.
[179] Wiley, B.; Sun, Y. G.; Mayers, B.; Xia, Y. N. Chem. Eur. J. 2005, 11, 454-463.
[180] Cheng, H.; Lu, Z. G. Solid State Sci. 2008, 10, 1042-1048.
[181] Lu, Z. G.; Liu, J. F.; Tang, Y. G.; Li, Y. D. Inorg. Chem. Commun. 2004, 7, 731-733.
[182] Yang, X. F.; Williams, I. D.; Chen, J.; Wang, J.; Xu, H. F.; Konishi, H.; Pan, Y. X.;
Liang, C. L.; Wu, M. M. J. Mater. Chem. 2008, 18, 3543-3546.
[183] Zhou, H. J.; Mao, Y. B.; Wong, S. S. Chem. Mater. 2007, 19, 5238-5249.
[184] Chen, J.; Yu, R. B.; Li, L. H.; Sun, C.; Zhang, T.; Chen, H. W.; Xing, X. R. Eur. J.
Inorg. Chem. 2008, 3655-3660.
[185] Zhou, H. J.; Mao, Y. B.; Wong, S. S. J. Mater. Chem. 2007, 17, 1707-1713.
[186] He, X. S.; Hu, C. G.; Xi, Y.; Wan, B. Y.; Xia, C. H. Sensor. Actuat. B 2009, 137, 62-
66.
[187] Pileni, M. P. Nat. Mater. 2003, 2, 145-150.
[188] Lu, Z. G.; Tang, Y. G.; Chen, L. M.; Li, Y. D. J. Cryst. Growth 2004, 266, 539-544.
[189] Wang, Y. W.; Xu, H.; Wang, X. B.; Zhang, X.; Jia, H. M.; Zhang, L. Z.; Qiu, J. R. J.
Phys. Chem. B 2006, 110, 13835-13840.
[190] Xu, H. F.; Wei, W.; Zhang, C. L.; Ding, S. J.; Qu, X. Z.; Liu, J. G.; Lu, Y. F.; Yang, Z.
Z. Chem. Asian J. 2007, 2, 828-836.
[191] Wang, G. Z.; Sterli, R.; Rrvik, P. M.; van Helvoort, A. T. J.; Holmestad, R.; Grande,
T.; Einarsrud, M, -A. Chem. Mater. 2007, 19, 2213-2221.
[192] Moreira, M. L.; Andrs, J.; Varela, J. A.; Longo, E. Cryst. Growth Des. 2009, 9, 833-
839.
[193] Macquart, R. B.; Smith, M. D.; zur Loye, H. C. Cryst. Growth Des. 2006, 6, 1361-
1365.
[194] Toshima, T.; Ishikawa, H.; Tanda, S.; Akiyama, T. Cryst. Growth Des. 2008, 8, 2066-
2069.
[195] Wang, Y. G.; Xu, G.; Yang, L. L.; Ren, Z. H.; Wei, X.; Weng, W. J.; Du, P. Y.; Shen,
G.; Han, G. R. Mater. Lett. 2009, 63, 239-241.
[196] Byrne, D.; Schilling, A.; Scott, J. F.; Gregg, J. M. Nanotechnology 2008, 19,
165608(1)-165608(5).
[197] Xie, X. W.; Li, Y.; Liu, Z. Q.; Haruta, M.; Shen, W. J. Nature 2009, 458, 746-749.
[198] Zhou, K. B.; Wang, X.; Sun, X. M.; Peng, Q.; Li, Y. D. J. Catal. 2005, 229, 206-212.
[199] Zhou, W.; Ran, R.; Shao, Z. P. J. Power Sources 2009, 192, 231-246.
[200] Liu, Q.; Chen, F. L. Mater. Res. Bull. 2009, 44, 2056-2061.
[201] Velev, O. D.; Tessier, P. M.; Lenhoff, A. M.; Kaler, E. W. Nature 1999, 401, 548-548.
[202] Yan, H. W.; Blanford, C. F.; Smyrl, W. H.; Stein, A. Chem. Commun. 2000, 1477-
1478.
[203] Holland, B. T.; Blanford, C. F.; Stein, A. Science 1998, 281, 538-540.
[204] Yang, P. D.; Deng, T.; Zhao, D. Y.; Feng, P. Y.; Pine, D.; Chmelka, B. F.; Whitesides,
G. M.; Stucky, G. D. Science 1998, 282, 2244-2246.
[205] Holland, B. T.; Blanford, C. F.; Do, T.; Stein, A. Chem. Mater. 1999, 11, 795-805.
[206] Sorensen, E. M.; Barry, S. J.; Jung, H. K.; Rondinelli, J. R.; Vaughey, J. T.;
Poeppelmeier, K. R. Chem. Mater. 2006, 18, 482-489.
[207] Woo, S. W.; Dokko, K.; Kanamura, K. Electrochim. Acta 2007, 53, 79-82.
[208] Tonti, D.; Torralvo, M. J.; Enciso, E.; Sobrados, I.; Sanz, J. Chem. Mater. 2008, 20,
4783-4790.
[209] Vlasov, Y. A.; Yao, N.; Norris, D. J. Adv. Mater. 1999, 11, 165-169.
[210] Velev, O. D.; Jede, T. A.; Lobo, R. F.; Lenhoff, A. M. Nature 1997, 389, 447-448.
[211] Holland, B. T.; Abrams, L.; Stein, A. J. Am. Chem. Soc. 1999, 121, 4308-4309.
[212] Zakhidov, A. A.; Baughman, R. H.; Iqbal, Z.; Cui, C. X.; Khayrullin, I.; Dantas, S. O.;
Merti, J.; Ralchenko, V. G. Science 1998, 282, 897-901.
[213] Woo, S. W.; Dokko, K.; Sasajima, K.; Takei, T.; Kanamura, K. Chem. Commun. 2006,
4099-4101.
[214] Johnson, S. A.; Ollivier, P. J.; Mallouk, T. E. Science 1999, 283, 963-965.
[215] Park, S. H.; Xia, Y. N. Chem. Mater. 1998, 10, 1745-1747.
[216] Zhou, L. Y.; Wang, D. P.; Huang, W. H.; Yao, A. H.; Xia, C. H.; Duan, X. Mater. Res.
Bull. 2009, 44, 259-262.
[217] Sadakane, M.; Horiuchi, T.; Kato, N.; Takahashi, C.; Ueda, W. Chem. Mater. 2007, 19,
5779-5785.
[218] Sadakane, M.; Asanuma, T.; Kubo, J.; Ueda, W. Chem. Mater. 2005, 17, 3546-3551.
[219] Sadakane, M.; Takahashi, C.; Kato, N.; Ogihara, H.; Nodasaka, Y.; Doi, Y.; Hinatsu,
Y.; Ueda, W. Bull. Chem. Soc. Jpn. 2007, 80, 677-685.
[220] Yan, H. W.; Blanford, C. F.; Holland, B. T.; Smyrl, W. H.; Stein, A. Chem. Mater.
2000, 12, 1134-1141.
[221] Sokolov, S.; Bell, D.; Stein, A. J. Am. Ceram. Soc. 2003, 86, 1481-1486.
[222] Zhang, Y. G.; Lei, Z. B.; Li, J. M.; Lu, S. M. New J. Chem. 2001, 25, 1118-1120.
[223] Chen, F. L.; Xia, C. R.; Liu, M. L. Chem. Lett. 2001, 30, 1031-1033.
[224] Kim, Y. N.; Kim, S. J.; Lee, E. K.; Chi, E. O.; Hur, N. H.; Hong, C. S. J. Mater. Chem.
2004, 14, 1774-1777.
[225] Chi, E. O.; Kim, Y. N.; Kim, J. C.; Hur, N. H. Chem. Mater. 2003, 15, 1929-1931.
[226] Dokko, K.; Akutagawa, N.; Isshiki, Y.; Hoshina, K.; Kanamura, K. Solid State Ionics
2005, 176, 2345-2348.
[227] Hara, M.; Nakano, H.; Dokko, K.; Okuda, S.; Kaeriyama, A.; Kanamura, K. J. Power
Sources 2009, 189, 485-489.
[228] Kanamura, K.; Akutagawa, N.; Dokko, K. J. Power Sources 2005, 146, 86-89.
[229] Ruiz-Morales, J. C.; Canales-Vzquez, J.; Pea-Martnez, J.; Marrero-Lpez, D.;
Irvine, J. T. S.; Nez, P. J. Mater. Chem. 2006, 16, 540-542.
[230] Li, H. N.; Dai, H. X.; He, H. J. Sci. Conf. Proc. 2009, 1, 186-189.
[231] Li, H. N.; Zhang, L.; Dai, H. X.; He, H. Inorg. Chem. 2009, 48, 4421-4434.
[232] Yang, P. D.; Zhao, D. Y.; Margolese, D. I.; Chmelka, B. F.; Stucky, G. D. Nature
1998, 396, 152-155.
[233] Grosso, D.; Boissire, C.; Smarsly, B.; Brezesinski, T.; Pinna, N.; Albouy, P. A.;
Amenitsch, H.; Antonietti, M.; Sanchez, C. Nat. Mater. 2004, 3, 787-792.
[234] Hou, R. Z.; Ferreira, P.; Vilarinho, P. M. Chem. Mater. 2009, 21, 3536-3541.
[235] Hou, R. Z.; Ferreira, P.; Vilarinho, P. M. Micropor. Mesopor. Mater. 2008, 110, 392-
396.
[236] Yue, W. B.; Zhou, W. Z. Prog. Nat. Sci. 2008, 18, 1329-1338.
[237] Tysz, H.; Lehmann, C. W.; Bongard, H.; Tesche, B.; Schmidt, R.; Schth, F. J. Am.
Chem. Soc. 2008, 130, 11510-11517.
[238] Xia, Y. D.; Mokaya, R. J. Mater. Chem. 2005, 15, 3126-3131.
[239] Cabo, M.; Pellicer, E.; Rossinyol, E.; Castell, O.; Suriach, S.; Bar, M. D. Cryst.
Growth Des. 2009, 9, 4814-4821.
[240] Li, G. S.; Zhang, D. Q.; Yu, J. C. Chem. Mater. 2008, 20, 3983-3992.
[241] Bocher, L.; Rosa, R.; Aguirre, M. H.; Malo, S.; Hbert, S.; Maignan, A.;Weidenkaff,
A. Solid State Sci. 2008, 10, 496-501.
[242] Bednorz, J. G.; Mller, K. A. Rev. Mod. Phys. 1988, 60, 585-600.
[243] Hong, J. W.; Fang, D. N. Appl. Phys. Lett. 2008, 92, 012906(1)-012906(3).
[244] Roy, S.; Hegde, M. S.; Madras, G. Appl. Energy 2009, 86, 2283-2297.
[245] Hu, Y. H.; Griffiths, K.; Norton, P. R. Surf. Sci. 2009, 603, 1740-1750.
[246] Zhu, J. J.; Thomas, A. Appl. Catal. B 2009, 92, 225-233.
[247] Russo, N.; Mescia, D.; Fino, D.; Saracco, G.; Specchia, V. Ind. Eng. Chem. Res. 2007,
46, 4226-4231.
[248] Dacquin, J. P.; Lancelot, C.; Dujardin, C.; Costa, P.; Djega-Mariadassou, G.; Beaunier,
P.; Kaliaguine, S.; Vaudreuil, S.; Royer, S.; Granger, P. Appl. Catal. B 2009, 91, 596-
604.
[249] Zhang, R. D.; Alamdari, H.; Kaliaguine, S. J. Catal. 2006, 242, 241-253.
[250] Zhang, R. D.; Villanueva, A.; Alamdari, H.; Kaliaguine, S. J. Mol. Catal. A 2006, 258,
22-34.
[251] Zhang, R. D.; Villanueva, A.; Alamdari, H.; Kaliaguine, S. J. Catal. 2006, 237, 368-
380.
[252] Zhang, R. D.; Villanueva, A.; Alamdari, H.; Kaliaguine, S. Appl. Catal. B 2006, 64,
220-233.
[253] Zhang, R. D.; Alamdari, H.; Kaliaguine, S. Catal. Lett. 2007, 119, 108-119.
[254] Zhang, R. D.; Alamdari, H.; Kaliaguine, S. Appl. Catal. A 2008, 340, 140-151.
[255] Zhang, R. D.; Alamdari, H.; Kaliaguine, S. Appl. Catal. B 2007, 72, 331-341.
[256] Zhang, R. D.; Villanueva, A.; Alamdari, H.; Kaliaguine, S. Appl. Catal. A 2006, 307,
85-97.
[257] Nishihata, Y.; Mizuki, J.; Akao, T.; Tanaka, H.; Uenishi, M.; Kimura, M.; Okamoto,
T.; Hamada, N. Nature 2002, 418, 164-167.
[258] Royer, S.; Brub, F.; Kaliaguine, S. Appl. Catal. A 2005, 282, 273-284.
[259] Petrovi, S.; Terlecki-Barievi, A.; Karanovi, L.; Kirilov-Srefanov, P.; Zduji, M.;
Dondur, V.; Paneva, D.; Mitov, I.; Raki, V. Appl. Catal. B 2007, 79, 186-198.
[260] Hammami, R.; Assa, S. B.; Batis, H. Appl. Catal. A 2009, 353, 145-153.
[261] Koponen, M. J.; Venlinen, T.; Suvanto, M.; Kallinen, K.; Kinnunen, T. J.; Hrknen,
M.; Pakkanen, T. A. J. Mol. Catal. A 2006, 258, 246-250.
[262] Saracco, G.; Geobaldo, F.; Baldi, G. Appl. Catal. B 1999, 20, 277-288.
[263] Deng, J. G.; Zhang, L.; Dai, H. X.; He, H.; Au, C. T. Appl. Catal. B 2009, 89, 87-96.
[264] Oliva, C.; Cappelli, S.; Rossetti, I.; Kryukov, A.; Bonoldi, L.; Forni, L. J. Mol. Catal. A
2006, 245, 55-61.
[265] Oliva, C.; Bonoldi, L.; Cappelli, S.; Fabbrini, L.; Rossetti, I.; Forni, L. J. Mol. Catal. A
2005, 226, 33-40.
[266] Fabbrini, L.; Kryukov, A.; Cappelli, S.; Chiarello, G. L.; Rossetti, I.; Oliva, C.; Forni,
L. J. Catal. 2005, 232, 247-256.
[267] Liang, S.; Teng, F.; Bulgan, G.; Zhu, Y. J. Phys. Chem. C 2007, 111, 16742-16749.
[268] Zhang, Y.; Zhang, L.; Deng, J. G.; Wei, L.; Dai, H. X.; He, H. Chin. J. Catal.
(Chinese) 2009, 30, 347-354.
[269] Zhang, Y.; Zhang, L.; Deng, J. G.; Dai. H. X.; He, H. Catal. Lett. 2010, in press.
[270] Armor, J. N. Appl. Catal. B 1992, 1, 221-256.
[271] Wang, W.; Zhang, H. B.; Lin, G. D.; Xiong, Z. T. Appl. Catal. B 2000, 24, 219-232.
[272] Agarwal, D. D.; Goswami, H. S. React. Kinet. Catal. Lett. 1994, 53, 441-449.
[273] Spinicci, R.; Tofanari, A.; Faticanti, M.; Pettiti, I.; Porta, P. J. Mol. Catal. A 2001, 176,
247-252.
[274] Deng, J. G.; Dai, H. X.; Jiang, H. Y.; Zhang, L.; Wang, G. Z.; He, H.; Au, C. T.
Environ. Sci. Technol. 2009, under review.
[275] Lee, Y. N.; Lago, R. M.; Fierro, J. L. G.; Corts, V.; Sapia, F.; Martnez, E. Appl.
Catal. A 2001, 207, 17-24.
[276] Pecchi, G.; Reyes, P.; Zamora, R.; Cads, L. E.; Fierro, J. L. G. J. Solid State Chem.
2008, 181, 905-912.
[277] Pecchi, G.; Campos, C. React. Kinet. Catal. Lett. 2007, 91, 353-359.
[278] Szabo, V.; Bassir, M.; Neste, A. V.; Kaliaguine, S. Appl. Catal. B 2002, 37, 175-180.
[279] Szabo, V.; Bassir, M.; Gallot, J. E.; Neste, A. V.; Kaliaguine, S. Appl. Catal. B 2003,
42, 265-277.
[280] Pereguez, R.; Hueso, J. L.; Holgado, J. P.; Gaillard, F.; Caballero, A. Catal. Lett.
2009, 131, 164-169.
[281] Mizuno, N.; Yamato, M.; Misono, M. J. Chem. Soc., Chem. Commun. 1988, 887-888.
[282] Matsukuma, I.; Kikuyama, S.; Kikuchi, R.; Sasaki, K.; Eguchi, K. Appl. Catal. B 2002,
37, 107-115.
[283] Hodjati, S.; Vaezzadeh, K.; Petit, C.; Pitchon, V.; Kiennemann, A. Catal. Today, 2000,
59, 323-334.
[284] Zahir, M, H.; Suzuki, T.; Fujishiro, Y.; Awano, M. Appl. Catal. A 2009, 361, 86-92.
In: Perovskites: Structure, Properties and Uses ISBN: 978-1-61668-525-6
Editors: Maxim Borowski, pp. 67-178 2010 Nova Science Publishers, Inc.
Chapter 2
PEROVSKITES AND THEIR NANOCOMPOSITES WITH

FLUORITE-LIKE OXIDES AS MATERIALS FOR SOLID
OXIDE FUEL CELLS CATHODES AND OXYGEN-
CONDUCTING MEMBRANES: MOBILITY AND
REACTIVITY OF THE SURFACE/BULK OXYGEN AS A
KEY FACTOR OF THEIR PERFORMANCE
V.A. Sadykov 1,2, S. N. Pavlova1, T. S. Kharlamova1,

V. S. Muzykantov1, N.F. Uvarov 2,3, Yu. S. Okhlupin3,
A. V. Ishchenko1, A. S. Bobin1, N. V. Mezentseva1, G.M. Alikina1,
A.I. Lukashevich1, T.A. Krieger1, T. V. Larina1, N. N.Bulgakov1,
V. M. Tapilin1, V.D. Belyaev1, E.M. Sadovskaya1, A. I. Boronin1,
V. A. Sobyanin2, O. F. Bobrenok4, A. L. Smirnova5,
O. L. Smorygo6 and J. A. Kilner7
1
Boreskov Institute of Catalysis SB RAS, Novosibirsk, Russia
2
Novosibirsk State University, Novosibirsk, Russsia
3
Institute of Solid State Chemistry and Mechanochemistry SB RAS, Novosibirsk, Russia
4
Institute of Thermophysics SB RAS, Novosibirsk, Russia
5
Eastern Connecticut State University, Willimantic, CT, USA
6
Institute of Powder Metallurgy, Minsk, Belarus
7
Department of Materials, Imperial College, London, United Kingdom
ABSTRACT
This paper presents results of research aimed at elucidating factors controlling
mobility and reactivity of the surface and bulk oxygen of complex perovskite-like
68 V.A. Sadykov, S. N. Pavlova, T. S. Kharlamova et al
AnBmOz oxides (manganites, ferrites, cobaltites, nickelates of La/Sr) and their

nanocomposites with fluorite-like oxides (doped ceria or zirconia) as related to their
performance in the intermediate temperature solid oxide fuel cells (IT SOFC) cathodes
and oxygen separation membranes. Real/defect structure and surface properties of these
materials as a function of both composition, preparation route and sintering temperature
were studied by combination of TEM with EDX, XRD, EXAFS, UV-Vis and XPS
methods. Lattice oxygen mobility, strength of oxygen species bonding with the surface
sites and their reactivity for both dispersed and dense materials were estimated by oxygen
isotope exchange, temperature-programmed desorption and reduction by H2 and CH4.
Coefficients of oxygen diffusion in dense materials were estimated by analysis of
dynamics of the oxygen loss monitored by measurements of the weight loss or
conductivity relaxations. For analysis of atomic-scale factors controlling oxygen bonding
strength and mobility, quantum-chemical approaches including the Density Functional
Theory (DFT) in spin-polarized SLDA approximation and semiempirical Interacting
Bonds Method were applied.
For simple perovskites, along with well-known effects of anion vacancies/interstitial
oxygen atoms and high- charge transition metal cations generated due to the oxygen non-
toichiometry, pronounced effects of the surface faces termination, ordering of oxygen
vacancies and appearance of extended defects were found to affect the oxygen mobility
and reactivity as well. From the microscopic point of view, decreasing the Me-O bonding
strength and increasing the distance between A and B cations due to the lattice parameter
increase helps to decrease the activation barriers for the oxygen diffusion in the
perovskite-like lattice.
For perovskite-fluorite nanocomposites, interfaces were shown to be paths for fast
oxygen diffusion due to redistribution of cations between contacting phases.
The most promising combinations of perovskites and perovskite-fluorite nano-
composites possessing a high lattice oxygen mobility and conductivity were used for
synthesis of functionally graded cathodes of IT SOFC with thin layers of doped zirconia
electrolytes as well as for manufacturing asymmetric Ni-Al compressed foam supported
membranes for oxygen separation and syngas generation. High and stable performance of
these materials promising for the practical application was demonstrated.
1. INTRODUCTION
Perovskites comprise a large class of complex oxides. Perovskite structure has ideal
formula ABO3 and consists of large-sized 12-coordinated cations at the A-site and small-sized
6-coordinated cations at the B-site. Perovskite compounds with different combinations of
cations in the A and B sites, e.g., 1-5, 2-4 or 3-3 and their more complex combination are
known. A remarkable feature of perovskites is a possibility of cationic substitutions in both A
and B-sites over a wide concentration ranges. Cationic substitution by lower charge cations
results in generation of oxygen vacancies in the oxygen sublattice that favors ionic
conductivity. Usually ABO3 compounds crystallize in polymorphic structures, which show
small distortion from the most symmetrical form of the perovskite structure. There are many
perovskite-related structures, like brownmillerite, K2NiF4-type structure, RuddlesdenPopper
phases An+1BnO3n+1, etc., which can be obtained as a result of the ordering of cations, defects,
inclusion of other crystallographic units into the ideal perovskite lattice, etc. [1, 2]. Because
Perovskites and Their Nanocomposites with Fluorite-Like Oxides 69
of the variety of structures and chemical compositions, perovskite oxides exhibit a large
diversity of properties, such as ferroelectricity, piezoelectricity, magnetic, catalytic properties
and even superconductivity. In addition to these well-known properties, several perovskite
oxides exhibit high ionic, electronic and mixed ionic-electronic conductivity required for such
applications as solid oxide fuel cells (SOFC) and permselective membranes [3, 4]. Perovskite
oxides containing transition metal cations exhibit high catalytic activity in various reactions,
in particular, oxidation reactions [5, 6].
The most efficient electrodes for SOFC are usually composites comprised of perovskites
and oxygen-conducting electrolytes (doped ceria, zirconia etc) as only composite materials
can meet the most of requirements for proper functioning of the electrode. For SOFC
cathodes, main requirements are [3]
morphological and thermodynamic stability at relevant temperatures and oxygen

pressures;
thermal expansion coefficient (TEC) should be close to that of other stack materials;
good adhesion and stable electrical (ionic) contact to the electrolyte;
chemical compatibility with other cell and stack components;
high catalytic activity in the reaction of the dissociative adsorption of oxygen;
low (less than 0.1 Ohm/cm2) polarization resistance;
high electronic (usually, p-type) conductivity, > 100 S/cm;
low cost of materials and the fabrication process.
For Intermediate Temperature (IT) operation, materials should also possess a high ionic
conductivity, i.e. to be mixed ionic-electronic conductors.
Most of these requirements are also applied to materials used for manufacturing of
oxygen permeable membranes. However, in conditions of a very high oxygen chemical
potential gradient such as in the case of syngas generation reactors, where membrane operates
under contact with methane from one side and air from the other side, functionally graded
multilayer (at least, bi-layer) design is required to prevent complete destruction of complex
perovskites in reducing conditions [7]. Note, though, that a similar concept of functionally
graded cathode materials with spatially varying composition and porosity of layers
successively deposited on a thin layer of electrolyte was shown to be the most promising in
design of anode-supported SOFC operating at decreased temperatures [8]. Tailor-made design
of these functionally graded composite materials and their performance modeling should be
based upon fundamental studies of the oxygen mobility and reactivity of both dense ceramics
and a porous/dispersed samples of both perovskites and their composites with electrolytes [9].
However, traditional techniques for estimation of oxygen diffusion and surface reactivity
parameters such as Secondary Ions Mass-Spectrometry (SIMS) oxygen isotope depth
profiling, conductivity or weight relaxation analysis can be applied only to dense ceramics,
and their sensitivity to the real atomic-scale structure of the surface (defects, impurities etc) is
a matter of debate [3, 9]. For powders, mass-spectrometric (MS) analysis of the gas phase
composition during isotope exchange allows the estimation of parameters characterizing both
the surface reaction and oxygen diffusion in the bulk, though this method is not yet
commonplace for the latter application [4-8]. Combination of this technique with
temperature-programmed oxygen desorption (TPD O2) and temperature programmed
reduction by H2 or CH4 (TPR) [6] allows to characterize transport properties and reactivity of
mixed ionic-electronic conducting materials under variable gradient of oxygen chemical
potential which is important for the proper modeling of their performance in a broad range of
operation parameters. Certainly, for design of mixed ionic-electronic conducting materials
with graded porosity, these studies are to be carried out for composite materials progressively
sintered from dispersed state to a dense ceramics, so that in the boundary region
characterization techniques for powdered and dense materials can be applied simultaneously.
Moreover, such a study will also allow to elucidate the role of interfaces suggested to be
important for transport properties of mixed ionic-electronic conducting oxides [10] or
composites [11]. To provide a noticeably impact on transport properties of a dense material,
interfaces should be sufficiently extended, so the sizes of respective domains should remain in
the nanorange. Certainly, this requires development of special procedures for synthesis of
MIEC nanocomposites as well as application of modern sophisticated techniques for
characterization of the structure and composition of these nanodomains and their boundaries.
This work reviews results of such research aimed at elucidating factors controlling
mobility and reactivity of the surface and bulk oxygen of complex perovskite-like AnBmOz
oxides (manganites, ferrites, cobaltites, nickelates of La/Sr) and their nanocomposites with
fluorite-like oxides (doped ceria or zirconia) prepared using new experimental approaches.
The first part considers characteristics of complex perovskites prepared via polymerized
polyester precursors (Pechini) route with a due attention paid to known concepts and
published data on the specificity of their real/defect structure as related to transport properties
and reactivity. In the second part properties of nanocomposites comprised of selected
perovskites with doped ceria or zirconia and prepared via powerful ultrasonic treatment of
suspensions in organic solvents with additions of surfactants will be discussed. The last part
presents results of application of these materials for design of functionally graded cathodes of
IT SOFC and oxygen-conducting membranes for syngas generation and their pilot-scale
testing.
2. PEROVSKITES
2.1. Theory and Experimental Procedures
Defect formation mechanisms and concentrations. In the perovskite structure ABO3 there
are three sublattices: two cationic and one anionic. Therefore, the total defect equilibrium
conditions include contribution of all possible defects (vacancies and substituting ions) in
each sublattice. For simplicity let us consider the defect formation in typical perovskite solid
solutions of the La1xSrxBO3- composition widely used in SOFC. As the perovskite lattice is
formed by cubic close packing of mixed AO3 layers, intrinsic defects are Schottky ones which
together with electron-hole pairs are formed in both pure and doped perovskites.
Quasichemical reactions of formation of intrinsic defects may be written as follows [12, 13]:
0 VLa + VB+ 3VO (1)

0 n + p (2)
The cationic vacancies are much less mobile than oxygen vacancies. The oxygen
vacancies together with electrons and holes determine transport properties of perovskites. For
practical purposes it would be desirable to obtain a solid solution with high concentrations of
both electronic defects and oxygen vacancies. The first option is to increase the concentration
of Sr-cations in A-sublattice through dissolution of SrO in the perovskite lattice:
2 SrO 2 SrLa + 2OO + VO (3)
The second approach is the doping of B-sublattice with low-charged cations, for example,
substitution of Ga3+ cations in LaGaO3:
+ 2OO + VO
2 MgO 2Mg La (4)
Reactions (3) and (4) proceed in solid electrolytes, for instance, in solid solutions based
on lanthanum gallate of the general formula: La1-xAxGa1-yByO3-(x+y)/2 (A = Ca, Sr, Ba; B = Mg,
Zn). These systems have high ionic conductivity caused by high concentration of oxygen
vacancies [14, 15] and they are wide gap semiconductors, hence, concentrations of intrinsic
electronic defects in these compounds are very low.
Other situation takes place in the perovskites with transition metals in B-sites. In frames
of crystal-field theory, the electronic structure of these materials is described in terms of
partially delocalized band states formed by t2g and eg levels [16]. For x > 0.1 the Fermi level
falls within the half-filled band, therefore, the compound exhibits metallic or semimetallic
conductivity at high temperature. The generation of electron-hole pairs proceeds via electron
excitation from the effectively neutral B3+ cation with formation of two different electronic
defects, corresponding to B2+ and B4+ cations. Hence Eqn. (2) can be rewritten as:
2 BB BB + BB (5)
Besides the process (5) the direct oxidation of B-cations also can take place following
possible reaction [17]:
3
O2 + 6 BB VA+ VB+ 3OO + 6 BB (6)
2
From (1) and (6) it follows that when this reaction prevails, concentration of oxygen
vacancies strongly decreases.
Oxidation process may be realized by another mechanism:
1
SrO + O2 + 2 BB 2 SrLa + 3OO + 2 BB (7)
2
Finally, in La1-xSrxBO3-d perovskites five basic processes expressed by reactions (1, 3, 5,

6 and 7), together with electrical neutrality condition
3[VA] + 3[VB] + [ SrLa ] + [ BB ] = 2[VO ] + [ BB ] (8)
define equilibrium concentrations of all ionic and electronic defects. Concentrations of

electrons [n] and holes [ p ] will be equal to [ BB ] and [ BB ] , respectively. Substituting

corresponding concentrations into the expressions for equilibrium constants and solving a
system of non-linear equations, one can find numerous values of all concentrations as a
function of the oxygen activity or partial pressure PO2. The analysis shows that general
dependence of concentration on pO2 can be divided into several domains, where some
particular neutrality conditions are fulfilled [18]:
I. Very high oxygen pressure or strongly oxidizing conditions. The process (6)
dominates corresponding to electroneutrality condition: 3[VA] + 3[VB] = [ p ] . As a

result, perovskite contains cation vacancies and, consequently, the excess of oxygen
ions so that > 0;
II. High oxygen pressure, oxidizing conditions when reaction (7) prevails. In this case
[ SrLa ] = [ p ] , strontium behaves like acceptor impurity in semiconductors and =
0;
III. Intermediate oxygen pressures or weakly oxidizing conditions. Here two competitive
reactions (7) and (3) proceed simultaneously so that resulting neutrality condition
[ SrLa ] = 2[VO ] + [ p ] should be held. The material becomes non-stoichiometric,
contains oxygen vacancies and the parameter varies within the range x/2 < < 0;
IV. Pressure range where neutral (neither oxidizing nor reducing) conditions are
fulfilled, where the charge of dopant cations is compensated by oxygen vacancies as
] = 2[VO ] , the parameter is fixed at
in the case of purely ionic conductors: [ SrLa

= x/2;
V. Low oxygen pressure, reducing conditions, the electrical neutrality is provided by
oxygen vacancies and electrons [n] = 2[VO ] , the parameter < x/2 is negative

and its absolute value increases with the decrease of pO2. Finally, at very low pO2 the
perovskite may irreversibly decompose due to the reduction process with formation
of other phases. This decomposition pressure depends on the material type and
temperature.
The five regions I-V are shown in Fig. 1 for A-doped RE1-xAxBO3 perovskite with the
rare earth ion (RE) in the A site. The comparison of model diagrams with experimental data
for acceptor doped La1-xSrxBO3- perovskites with B = Mn, Fe, and Co shows [19, 20] that at
high oxygen pressures the manganite is hyperstoichiometric ( > 0, region I). As pO2
decreases, the parameter falls to zero and does not change within the pressure range of 10-4-
10-10 atm (region II). This behavior implies that the oxygen vacancy concentration will be
very low, and this remains true even for quite heavily acceptor doped material, while
concentration of cation vacancies is high. No appreciable concentration of oxygen vacancies

is formed until oxygen partial pressures approaching 10-10 atm (region III). Thus, under
normal SOFC cathode operating conditions, the manganite materials are expected to have low
vacancy concentrations and, consequently, low oxygen diffusion coefficients. However,
rather low (~ 120 kJ/mol O2) values of the partial enthalpy of oxygen removal from the lattice
of undoped or Sr-doped lanthanum manganite not depending upon the oxygen excess as
estimated by Kuo et al from the results of detailed gravimetric studies [20] suggest that under
large gradient of oxygen concentration (i.e. big cathodic polarization) oxygen flux in the
lanthanum manganite lattice can be rather high.
Region V IV III II I
/
Neutrality
[p ] = [A ] - / ///
/ Re
condition [n ] = 2[V ] 2[V ] = [A ]
[p ] = [A ] [p ] = 6[V ]
O Re Re B
2[V ]

+
0

-1 0 -1 -1 -1
d{log[V
O
]}
/6 /4 /2 /8
d{log(PO2)}
conductivity
n-type ionic p-type p-type p-type
type
log PO2
Figure 1. Schematic representation of dependences of the non-stoichiometry parameter , the
concentration of oxygen vacancies and total conductivity on the oxygen partial pressure pO2 in an
acceptor-doped RE1-xAxBO3- oxide.
In contrast to the manganites, cobaltites and ferrites at normal pressure are in the mixed
compensation region III, where the value of lies between 0 and x/2. Thus, high oxygen
vacancy concentrations and, consequently, oxygen diffusivities are expected under normal
cathode conditions. For the ferrite, a plateau in the oxygen stoichiometry is observed at low
PO2 (Fig. 2) that corresponds to the range IV, indicating that the acceptor is vacancy
compensated and the material would exhibit ionic conductivity [9]. From this analysis, it is
obvious that there are significant changes in the defect populations in these acceptor-doped
materials. The properties of perovskites depend on the temperature, oxygen partial pressure,
and first of all, on the type of the transition metal in the B-cation site.
Figure 2. Oxygen stoichiometry in (a) LSF (x = 0.1) and (b) LSC (x = 0.3) as a function of O2 and
temperature.
Electrical properties. As it was mentioned above, to provide a low electrode resistance of

SOFC, the electrode materials should have sufficiently high electronic conductivity, > ~
100 S/cm. To provide such a high conductivity the concentration of mobile carriers (holes,
electrons) should be sufficiently high. Electronic structure of undoped perovskites LaBO3,
where B is the transition metal cation, is determined by the energy diagram for B cations
octahedrally coordinated by the oxygen anions. In terms of the ligand field model the
electronic bands are formed by bonding and antibonding molecular orbitals resulted form the
overlapping of cations d-orbitals (Bd) with 2p-orbitals of ligand anions (O2p). Bonding
molecular orbitals are mainly composed of O2p levels, whereas antibonding levels are mainly
formed by d-orbitals split into three t2g and two eg crystal field levels (d and d electron
states, respectively). In systems with unpaired electrons, especially in B cations occurring in
the high-spin state, the degeneracy of t2g or eg may be removed due to the Jahn-Teller effect
resulting also in the lattice distortion and lowering its symmetry. In perovskites, energies of d
and d states lie near the bonding orbitals (top of O2p bands). Moreover, metal cations may
exist in several valence states which differ not strongly by energy. In this case d-electron
levels of the high-charged state may be located a little below the top of O2p band creating
antibonding states at the top of the O2p bands that results in a pinning of high-valence state
that transforms the electrons and holes from localized to itinerant electronic behavior [21].
As the most effective cathode materials for SOFC are based on perovskites containing
Mn, Fe, Co and Ni, let us shortly characterize transport properties of these compounds. In
La1-xSrxMnO3 (LSM) and La1-xSrxFeO3 (LSF) the metal (III) cations have high-spin t3e1
and t3e2 configurations. Long-range Jahn-Teller spin-orbital ordering leads to the
orthorhombic distortion of their perovskite structures and splitting of dg levels. As a result,
there are several half-filled bands leading to strong antiferromagnetic coupling and insulating
properties. The electronic transport proceeds by thermally activated small polaron
mechanism, the activation energy decreasing with temperature from 0.55 to 0.15 eV [21, 22].
At high temperatures or on doping, the Jahn-Teller effect decreases, the structure transforms
from the orthorhombic to a cubic one, and the oxide becomes ferromagnetic. For perovskite-
type solid solutions La1-xAxMnO3 and La1-xAxFeO3 (A = Ca, Sr, Ba), the introduction of
A-cations leads to the formation of Mn (IV) and Fe(IV) cations. In manganites, Jahn-Teller
effect for Mn(IV) cation is not displayed that leads to less distortion of the structure. At some
critical concentration of x = ~0.1, the oxide becomes ferromagnetic with metallic-type
conductivity provided by itinerant electron-hole pairs via Zener small polaron mechanism
[21]. In ferrites Fe(IV) t3e1 state is pinned to the top of O2p band, so that d electrons and holes
are itinerant resulting in conductivity of the metallic type occurring via de Gennes small
polaron mechanism [22, 23]. In LaNiO3 Ni(III) ions occur in the low-spin state t6e1 with a
single electron per Ni cation, d-electrons form a narrow conducting band providing metallic
conductivity of order of 103 S/cm decreasing with temperature. Ni (III) state is pinned at the
top of the O2p band that leads to the stabilization of this oxidation state of nickel; however, at
high temperatures LaNiO3 is not stable. The lanthanum cobaltite LaCoO3 exhibits unusual
behavior: at low temperatures Co(III) occurs in a low-spin t6e0 state, and the compound is a
diamagnetic insulator. As the energy difference between the low-spin and high-spin t4e2 states
is small, a transition is observed to a high-spin state at elevated temperatures. Due to strong
Jahn-Teller effect, cobalt cations are first stabilized in a phase with intermediate spin state t5e1
of cobalt cation. At high temperatures this phase transforms into a ferromagnetic highly-
conducting phase where metal cations are in a high-spin state. Substitution of lanthanum for
alkali earth metal leads to a phase segregation with formation of the phase with Co (III)
intermediate spin and metallic state.
Hence, there are many specific mechanisms controlling conductivity of binary
perovskites LaBO3 as a function of the B-cation type and temperature. Substitution of
lanthanum for alkali-earth metal leads to generation of B cations in a higher oxidation state
stabilized by pinning of the O2p band. In the case of manganites and ferrites, it leads to
weakening of Jahn-Teller effect and antiferromagnetic coupling. It results in delocalization of
electrons and increase in the hole conductivity. Introduction of two transition metals into B-
sublattice allows fine tuning the electronic structure of perovskites. For example, solid
solutions of La0.8Sr0.2Co1-yFeyO3 (LSCF) have more complex electronic structure but
generally exhibits more semiconductor-like behavior at a high iron content (y > 0.8) while
being more metallic with a high Co content (y < 0.1) [9].
More complex defect models were also considered for LaMnO3 suggesting antisite
disordering (i.e. incorporation of Mn cations into La vacancies) [24, 25].
Oxygen diffusion. The total conductivity of Sr-doped lanthanum manganites, cobaltites
and ferrites is much higher than the oxygen ionic conductivity. Therefore, to ensure good
electrode performances it would be desirable to increase the ionic conductivity (i) or
diffusion coefficients which are interrelated via the Nernst-Einstein relation
fkT
D* = Dv [VO ] = i N (9)
4e 2
Here D* is the tracer diffusion coefficient; Dv is the diffusion coefficient of oxygen

vacancy; f is the correlation factor; N is the total volume concentration of oxygen ions in the
substance. Eqn. (9) includes only free oxygen vacancies, i.e. not associated with the dopant
cation. The diffusion coefficient is defined by ionic mobility mechanism and may be
expressed in terms of ion hopping parameters:
1 S H
D* = f [VO ]a 2 0 exp m exp m (10)
6 k kT
where a is the hopping distance; 0 is the attempt frequency; Sm and Hm is the migration
entropy and enthalpy, respectively. It is evident that the presence of oxygen vacancies is
necessary for the oxygen diffusion. Let us consider other factors affecting the diffusion
coefficient. As shown in the literature [26, 27], doping of LaBO3 perovskite in both La- and
B-sublattices leads to a considerable, by several orders of magnitude, increase in the oxygen
diffusion coefficient. The most probable reason of low diffusion coefficients in undoped
perovskites is a high formation energy of oxygen vacancies by Eqn. (1). Substitution of La (or
B-cation) for low-charged cations may lead to the formation of vacancies, but even in this
case they may be associated with doping cations to form immobile complexes which do not
participate in the oxygen transport. The contribution of association effects is the most
pronounced in lanthanum chromites and manganites, where the activation energy for
diffusion increases with x. In contrast, in lanthanum cobaltites and ferrites the activation
energy of oxygen diffusion monotonically decreases with x due to decrease of vacancy
formation energy, Hf, which is accompanied by a huge increase in the oxygen diffusion
coefficients [26, 27]. Fig. 3 shows temperature dependences of the diffusion coefficient of
oxygen vacancies in different perovskite materials [18]. One sees that Dv values are very high
and comparable with fluorite-type solid solutions known as typical ionic conductors. It
makes perovskite solid solutions promising materials for oxygen separation membranes or
electrode materials.
Figure 3. Typical temperature dependence of the oxygen vacancy diffusion coefficient in perovskites
and fluorite-like oxides.
For oxygen diffusion under gradient of its chemical potential , the chemical diffusion
% is used instead of D*. In this case the Ficks law will be expressed using the
coefficient D
Wagner equation as a function of the oxygen chemical potential in the form:
cO2 1 O2 cO2
J O2 = D% = 2 2 e i (11)
x 4 F e + i cO2 x
where F is the Faraday constant, and the expression in round brackets is an ambipolar
conductivity of electronic and ionic charge carriers. From this equation it follows that
1 e i O2
D% = (12)
8F 2 e + i cO
If e >> i, Eqn. (12) may be represented as
D* ln aO
D% = (13)
f ln cO
where the last multiplier is the thermodynamic factor which is equal to unity for ideal
solutions and may strongly differ from 1 for real systems. According to the theoretical
1
analysis [28], the limiting value of the thermodynamic factor is VO

and from Eqn. (9) it
% =D .
follows that in this case D V
Surface exchange reaction. The oxygen surface exchange coefficient, k, is another

important kinetic parameter characterizing oxygen transport rates in the oxide materials. It is
a measure of the rate of the neutral oxygen flux JO crossing the surface of the oxide as a result
of the process [26]:
1 VO n 2 aO1/2
OO = O2 + VO + 2n ' ; k = x 2 (14)
2 [OO ]
As follows from equation (14), this flux depends on the surface vacancy concentration,
the surface electron concentration, and the dissociation rate of O2 molecule. The expression
for k can be formally written in the simple form [29]
JO
k= (15)
[O]
The flux can be expressed through the exchange frequency and the concentration of the
oxygen vacancies. Assuming that the exchange frequency is thermally activated
Se He
= e exp exp (16)
k kT
the following expression for the oxygen surface exchange coefficient was obtained [29]:
S H
k = a0 m 1/3 [VO ]2/3 e exp e exp e (17)
k kT
where a0 is the cubic cell parameter; m is the number of oxygen sites per cell; e, Se and He
are the attempt frequency, the entropy and the enthalpy of the elementary process of the
surface exchange. This expression bears similarities to the atomistic expression for the
diffusion (10) and explains the experimental observation that higher vacancy concentration
leads to faster ion transport and surface exchange. Hence, the vacancy concentration is clearly
a very important parameter. Other important parameter is the electronic conductivity. It was
observed that oxygen exchange correlates with electronic conduction: materials with high
concentrations and/or more mobile electronic carriers tend to have much better surface
catalytic properties (k) than purely ionic conductive materials with similar rates of bulk ion
transport (D*) [9].
The influence of the surface exchange on cathode performance may be expressed in
terms of the characteristic length Lc defined as the ratio D/k. This length is the point at which
the permeation behavior of oxygen through a mixed conducting membrane (such as a
cathode) changes from diffusion-controlled to one limited by the surface exchange process
[30, 31].
Experimental Techniques for Investigation of Transport Properties and

Surface Exchange
Transport properties (separate determination of electronic and ionic conductivity, oxygen
tracer diffusion and chemical diffusion) and surface stages (rate/constant of exchange)
parameters of dense ceramics can be studied by several methods, such as electronic blocking
polarization methods [32-34], 18O tracer profile analysis by SIMS [26, 27, 29, 35], study of
the 18O/16O isotope exchange kinetics by gas-phase analysis of the isotope composition [27,
35] or even by monitoring the change in weight of an oxide during the isotope exchange by a
microbalance [36]; relaxation methods involving the measurement of change in conductivity,
mass or other properties after a sudden change of pO2 [37-40]; or by direct study of
permeation of oxygen through a dense membrane [41-47]. For powdered samples, oxygen
isotope heteroexchange with mass-spectometric control of the gas-phase isotope compositions
(both in static and flow modes) is a well-established technique allowing to estimate main
parameters of the surface reaction steps and surface/bulk oxygen diffusion [48-62]. For
powdered samples, the strength of oxygen bonding with the surface sites, amount of different
surface/bulk oxygen forms and dynamics of their migration from the bulk of particles to their
surface can be estimated by using temperature--programmed desorption (TPD) of oxygen or
temperature--programmed reduction (TPR) by a reducing agent (H2, CH4) [11, 60-63].
Lets consider some of these techniques in detail.
Electrical conductivity relaxation (ECR) method. ECR method is rather simple, it does
not require expensive equipment and provides high precision in determination of the oxygen
chemical diffusion and surface exchange coefficients in mixed ionic-electronic conductors
[64-66].
When a mixed ionic-electronic conductor (i.e. La1-xSrxBO3-) is subjected to an oxygen
chemical potential gradient, there arises a chemical diffusion of oxygen ions down the
gradient (or the diffusion of oxygen vacancies in the opposite direction) and redistribution of
electronic carriers. At typical values of x (0.1 < x < 0.5), the charge neutrality condition
stipulates that
[ SrLa ] = 2 VO n ' (18)
for n-type regime, or
[ SrLa ] = 2 VO + p (19)
for p-type regime; the contribution of other defects may be neglected. The defect

concentration gradient will be determined by ambipolar diffusion of VO and e (in parallel
direction) or h (in counter direction for a p-type regime). The total conductivity is a sum of
the electronic and ionic contributions and may be written as
= eun n '+ 2euv VO (20)
(contributions of other carriers are neglected), uk denote the mobility of the k-type carriers
[37]. Taking time derivatives and assuming that [R B ] / t = 0 , one has
V
= 2e(u n + uV ) O
[ ] (21)
t t
Due to Fick's second law,
VO
t
( )
= D% O VO , (22)
~
where DO is the oxygen chemical diffusion coefficient. In relaxation experiments, the
~
oxygen activity windows are chosen to be small enough (typically |log(aO2)| < 1) so that DO
may be assumed to be constant. The effect of the surface reaction should be considered as the
boundary condition for obtaining a solution to Ficks second law. Then an analytical solution
may be obtained in the approximation that the surface reaction kinetics is linear with respect
to the oxygen concentration at the surface. For a rectangular sample with dimensions 2h, 2w
and 2l the exact solution has a form [65]:
i2 Dt% m2 Dt
% n2 Dt
%
2 L 1
2
exp 2 L 2
2
exp 2 L3
2
exp
(t ) 0

2
h
2
w
2
l
= 1 2 2 2 2
0 i =1 m =1 n =1 i ( i + L1 + L1 )
2
m ( m + L2 2 + L2 ) n2 ( n2 + L32 + L3 )
(23)
where (t) is the timedependent conductivity, 0 and are initial (t 0) and limiting (t
) conductivity values, respectively. The dimensionless parameters L1, L2, and L3 are defined
using the surface exchange coefficient k and the chemical diffusion coefficient as
L1 L2 L3 k
= = = (24)
h w l D%
which can be used for the estimation of relative contribution of surface reaction to the
conductivity relaxation; coefficients i, m, and n are i-th, m-th or n-th roots of the equations:
i tan i = L1
m tan m = L2 (25)
n tan n = L3
For a specimen in the form of the infinite bar with a cross-section of 2a2a one obtains
[66]:
2
(t ) 2 L2 n2 D% O t
= 2 2 exp (26)
0 n =1 n ( n + L2 + L) a2
with n given by
k
n tan n = L a % (27)
D
Particularly when L<<1 the conductivity relaxation is diffusion controlled and Eqn. (23)
takes the form [66]
2
(t ) 64 1 (2 j + 1)2 2 D% O t
= 4 exp (28)
0 j =0 (2 j + 1) 2 4a 2
An example of the electrical conductivity relaxation curve for La0.8Sr0.2Fe0.8Ni0.2O3- is

shown in Fig. 4.
Figure 4. Conductivity relaxation curve for La0.8Sr0.2Fe0.8Ni0.2O3- perovskite at 800C after changing
pO2 from 0.210 to 0.070 atm. Points experiment, line fitting to Eqn. 28.
Oxygen isotope exchange. For dense ceramics, such technique as SIMS depth profiling
can be applied. The sample with a natural oxygen isotope composition is first kept for a fixed
time at the operation temperature under contact with oxygen enriched with the 18O isotope.
Then the 18O oxygen profile across the sample depth is measured by SIMS and analyzed
using the following equation [26, 66]
C ( x) C x x
C '( x) = erfc exp ( hx + h 2 D*t ) erfc + h D*t (29)
Cg C *
2 D t
*
2 D t
where C(x) is the normalized fraction of 18O, C(x) is the isotopic fraction as a function of
depth x which is obtained from SIMS measurements, C is the background level of 18O in the
sample, Cg is the fraction of 18O in the gas phase, and D* is the bulk tracer diffusion
coefficient. The parameter h is given by h = k/D*, where k is the oxygen surface exchange
coefficient. The values of D* (cm2/s) and k (cm/s) may be determined using a nonlinear least
squares fit.
For powdered samples, mass-spectrometric (MS) analysis of the gas phase composition
during isotope exchange is used for the estimation of parameters characterizing both the
surface reaction and oxygen diffusion in the bulk.
In the analysis of O2 isotopic redistribution between molecules of the gas phase, and
isotopic heteroexchange between the gas phase and surface oxygen atoms of a solid oxide
(O)S, three routes of exchange must be taken into account [52-56]:
18
O2 + 16O2 = 2 18O16O (I)
18
O2 + (16O)S = 18O16O + (18O)S (II)
18
O2 + 2 (16O)S = 16O2 + 2 (18O)S (III)
The distribution of the molecules 18Oi16O2-i (i = 0, 1, 2), characterized by their molar

fractions xi, is determined by two independent composition variables (isotope variables),
since x0 + x1 + x2 = 1. The 18O fraction of the oxygen gas, = 0.5 x1 + x2 , is usually selected
as the first variable, while the second independent variable can be chosen as a fraction of any
of the oxygen molecules, i.e. 18O16O (x1)
With these variables, the redistribution of the isotopes in the system under consideration
is described by the following isotope-kinetic equations [52-56]:
N d
= R( s ) (30)
S dt
N dx1
= K1[2 (1 ) x1 ] + K 2 [ (1 s ) + s (1 ) x1 ] + K3[2 s (1 s ) x1 ] (31)
S dt
Here N is the number of 2 molecules, S the surface area of oxide, t time,

R = 0.5 K 2 + K 3 the rate of heteroexchange, K1, K2 and K3 the rates of exchange of the
1st, 2nd and 3rd type, respectively, their sum giving the total rate of exchange K.
The system of isotope-kinetic equations (30)-(31) is not closed since, along with the
variables describing the gas phase composition, they contain the 3rd independent variable, s ,
the fraction of 18O in the oxide surface.
For a complete description of the isotope redistribution, differential equations reflecting
the isotope transfer in the solid phase should be added to this system
N s s N v
= R( s ) rD e
S t S =0
(32)
v 2 v
= rD
t 2 (33)
with the initial and boundary conditions:
= 0 , s = v = s
0
t=0:
= 0 : v = s
where v is the fraction of 18O in the oxide bulk, N S and Ne are the number of oxygen
atoms in the surface layer of oxide and the number of exchangeable atoms in the bulk of the
oxide, respectively; is the reduced depth (z) of the oxide layer =z/h, where h is the
D
characteristic size of oxide particles; rD = is the effective diffusion coefficient (diffusion
h2
relaxation constant), D is the oxygen self-diffusion coefficient.
The isotope-kinetic equations (30) - (31) allow the elucidation of the mechanism of
surface reactions (adsorption-desorption and other processes) and the estimation of the bond
strength of the surface oxygen [48-56] considered to be important parameters in analysis of
porous MIEC cathodes performance [9].
In the case of equivalent exchangeable oxygen atoms (Ne), the time dependence of the
gas phase isotopic fraction, (t), is well described by an exponential function. By using the
isotope balance equation 2 N + N e s = 2 N + N e s = (2 N + N e ) (where s0 is the
0 0
initial value of s , -isotope fraction at equilibrium), the variable s can be excluded from
Equation (30). This allows the estimation of the rate of heteroexchange R by using the
integral form of isotope-kinetic equation (30)
S
ln = (1 + ) Rt (34)

0
N
Here, = 2 N / N e
The complete system of equations (30) - (34) can be used to estimate the self diffusion
coefficient for oxygen in the bulk of the solid oxide. For this purpose, numerical analysis of
the time dependence of the isotope variables, which corresponds to the solution of the inverse
problem of the system of equations (30)-(34), was carried out through the minimization of the
next objective functional:
tend tend
F ( ) = ( ( calc
(t ) exp
(t )) dt +
2
(x
calc
1 (t ) x1exp (t )) 2 dt ) min
o o
where exp (t ) , x1exp (t ) are experimental and calc (t ) , x1calc (t ) - calculated

dependencies of the isotope fractions in the gas phase oxygen.
To evaluate the quality of the fit between experimentally determined transient curves and
calculated ones, the least square deviation was evaluated as follows:
1/ 2
N N

Sq = ( icalc iexp ) 2 / N + ( x1calc i x1exp i ) 2 / N
i =1 i =1
where N is the number of experimental points.

In the case of temperature-programmed oxygen exchange (TPIE), data analysis was
carried out using specially derived equations. Thus, by transformation of isotope equations
(30) and (31), the time-independent isotope-mechanistic equation was obtained [51-60]
ln[(v + b)/(v0 + b)] = a ln(s/s0) (35),
Here v = z/s2, s = s , z = x2 2, x2 fraction of 18O2 molecules, and parameters a and

b are determined by the type of mechanism (vide supra):
k + 2 3 1 k3 3
a = 2 = 2 ; b= =
r 0.5 2 + 3 2r k 2 + 2 3 1
Hence, Equation (35) can be used for estimation of the share of different types of
exchange characterized by parameters K2 and K3.
Cheselske, Hall et al [67, 68] derived an equation to estimate the apparent activation
energy of heteroexchange in experiments with a linear temperature ramp and equivalent
exchangeable atoms:
E = - [RT2(d/dT)/( - )]I (36)
Here, -d/dT is taken at the inflection point where it achieves a maximum value, is the
respective isotope fraction at this point (vide infra).
TPIE experiments also allow a semi-quantitative characterisation of the oxygen mobility.
Thus, the depth of isotope penetration from the gas phase into the oxide in the course of TPIE
can be characterized by the value NX determined by the isotope balance equation in the closed
system:
2 N 0 + N X S0 = (2 N + N X ) (37)
This quantity varying from 0 to the total number of exchangeable atoms is used as the
measure of the number of exchanged oxygen atoms in the solid oxide. It is more conveniently
expressed in relative units
NX NX
XS = and X V = N (38)
NS V
corresponding, respectively, to the number of exchanged oxygen monolayers Xs, (1

monolayer = 1.41019 atoms/m2 ) and the exchanged fraction of the bulk oxygen, Xv. For a
given series of samples, these parameters known as the dynamic extent of isotope exchange
[63] correlate with the oxygen diffusion coefficients [11, 60-63].
Temperature-Programmed Desorption of Oxygen From Powdered Samples (Oxygen

TPD). Temperature-programmed desorption of oxygen from oxides is used to characterize the
bonding strength of oxygen with the surface (as estimated by the temperature of the
noticeable oxygen desorption into the stream of He or by position of the maximum of
respective desorption peak) as well as the amount of easily desorbed oxygen [69]. For Sr-
doped perovskite oxides (cobaltites, nickelates, manganites etc) high-temperature peaks of
oxygen evolution (so called -form) are usually assigned to a loss of oxygen from the lattice
accompanied by a change of the charge state of transition metal cations [6]. For perovskite-
like oxides or their composites with fluorite-like oxides (GDC etc) with a high lattice oxygen
mobility, the specific rate of oxygen desorption estimated at Tmax of respective peak varies in
parallel with the lattice oxygen mobility estimated by the oxygen isotope exchange [62].
Apparently, the maximum rate of desorption in the temperature-programmed mode is
achieved when the effect of the increase of the oxygen mobility in the lattice with temperature
(determined by the activation energy of the oxygen chemical diffusion) is prevailed by the
effect of depletion of the amount of oxygen stored in the bulk of particle. Hence, this
method allows to estimate (at least, qualitatively) the lattice oxygen mobility in conditions of
a moderate oxygen chemical potential gradient between the surface and the bulk of particles.
Oxygen permeability. The measurement of oxygen permeability in a gradient of the
oxygen pressure or activity is a direct method for determination of the oxygen diffusion
coefficients in solids [70-73]. The permeability is numerically estimated from the value of
stationary oxygen flux JO2 through a dense oxygen-permeable membrane. The value of JO2
should satisfy the Ficks law, Eqn. (11), which can be rewritten in a more suitable form [72]:
ln P"
1 O2

J O2 = 2 2 e i d ln PO2 (39)
4 F L ln P' e + i
O2
For the case of high electronic conductivity (e >> i) when ionic conductivity depends
on PO2 by the power law i ipO2n, after integration of Eqn. (32) one finds:
0 RT
J O2 = i ( PO2 )n ( PO2 )n (40)
4 F nL
2 2
This equation does not take the surface reaction into account. More correct expression for
the oxygen flux [72] has the form:
L t t O2
J O2 = e2 i 2 (41)
L + 2 Lc 4 F L
where Lc is the characteristic length (or the membrane thickness). If L >> Lc, then the flux
obeys ordinary Ficks law. In the opposite case the permeability is limited by the surface
exchange reaction. It is seen that the oxygen flux across the membrane is reduced by a factor
(1 + 2Lc/L)1, relative to that in the absence of transfer limitations across the interfaces. This
problem is solved by supporting porous layers of the same or different composition on the
oxygen-lean and oxygen rich sides of a dense membrane [7, 70, 71]. For oxygen separation, a
thin porous layer of a complex perovskite (Sr-Co-Fe etc) with a high catalytic activity
supported on the air side of membrane allows to change membrane operation mode from the
surface-controlled to the bulk controlled one. For membranes designed for methane selective
oxidation into syngas by oxygen separated from air, supporting a thin layer of Sm-doped ceria
on the fuel side of perovskite membrane allows to prevent destruction of perovskite structure
caused by reduction. Supporting a porous layer of Ni or Pt catalyst allows to increase the rate
of CH4 oxidation and syngas selectivity.
There are several comprehensive reports [7, 74-79] devoted to mixed ionicelectronic
conducting ceramic membranes for gas separation and selective catalytic reactions. The
requirements to membrane materials differ from those to SOFC cathodes. For membranes, the
most promising concept is based upon supported asymmetric functionally graded design
combining a dense permselective layer and layers with different porosity. Cathode materials
may be not dense, they operate in air and may be unstable in a reducing atmosphere. On the
other hand, cathode materials should be chemically inert towards other SOFC components
and keep their characteristics under the working conditions.
Temperature-Programmed Reduction. Temperature-programmed reduction of powdered

perovskite samples by H2 allows to estimate the reactivity of surface/bulk oxygen species,
determine oxygen stoichiometry and elucidate specificity of the reaction mechanism [80].
Usually, reduction of perovskites containing redox transition metal cations (Co, Mn, Fe)
proceeds in steps corresponding to Me4+-Me3+, Me3+-Me2+, or Me2+-Meo transitions [7, 80].
Thus, peaks corresponding to Mn4+ and Co3+ or Ni3+ reduction are usually situated at
temperatures below 600 oC, and reduction proceeds by cooperative rearrangement of
perovskites structure without new phases nucleation. In this case, variation of the lattice
oxygen mobility with the composition/defect structure of perovskites affects both positions of
peaks and their intensity (rates at Tmax.). Reduction with formation of MnO or Meo particles
dispersed in matrix of irreducible rare-earth oxides proceeds via nucleation/growth stages
following typical topochemical kinetics (contracting spheres, etc) in isothermal experiments
[80].
Synthesis of perovskites
In frames of this work the following systems were synthesized via Pechini route:
perovskite-type compounds and solid solutions: La0.8Sr0.2Fe1-xNixO3- (x = 0.1-0.4)

(LSFNx), LaMnO3, La0.8Ca0.2MnO3+x , La0.8Sr0.2MnO3+x (LSM), La0.8Sr0.2-
Fe0.8Co0.2O3- (LSFC), La0.8Sr0.2Fe0.5Co0.5O3 - (LSFC), LaNiO3 (LN);
-Ruddlesden-Popper perovskite-related structures: La2NiO4, La2Ni0.8Cu0.2O4, Pr2Ni-
0.75Fe0.25O4 and Pr4Ni3O10.
In Pechini method, metal nitrates (MeN), citric acid (CA), ethylene glycol (EG) and
ethylenediamine (ED) were used as reagents. The molar ratio of CA:EG:ED:MeN was
3.75:11.25:3.75:1. CA and MeN were dissolved in ethylene glycol at 80 C and in distilled
water at room temperature, respectively. The prepared solutions were mixed together at room
temperature under stirring followed by addition of ED. The prepared solution was stirred for
60 min and then heated at 70 for 24 h allowing for the gel formation. The gel was calcined
at 25-700 .
Complex oxides in the Sr-Ba-Fe-Co-(Zr) -O system were prepared by 5 min mechanical
milling of the starting reagents (SrCO3, BaCO3, Fe2O3, Co3O4, zirconium hydroxide) blends
in an AGO-2 planetary mill with the powder/steel ball ratio of ~ 12/200 followed by
calcinations at 1000-1100 oC [60]. For comparison, LaCoO3 and LaFeO3 samples were
prepared via ceramic route [60]. The specific surface area of samples obtained via ceramic
and MA routes was in the range of 0.2-2 m2/g.
Characterization methods. X-ray diffraction (XRD) patterns were obtained with an URD-
6 an ARLXTRA diffractometers using Cu K monochromatic radiation (=1.5418 ). The
2 scanning region was 5-900. Transmission electron microscopy (TEM) micrographs were
obtained with a JEM-2010 instrument (lattice resolution 1.4 ) and acceleration voltage 200
kV. The morphology of sintered cell samples was studied by SEM using a BS-350 Tesla
scanning electron microscope. X-ray photoelectron spectra (XPS) were acquired using an ES-
300 electron spectrometer (Kratos Analytical, UK, AlK and MgK anodes). Samples were
mounted on a holder using double-side scotch tape. Spectra were calibrated relative C1s
binding energy (BE) 284.8 eV 0.2 eV. Quantitative analysis of the surface layers
composition was made using integral line intensities and known values of atomic sensitivities
factors with the relative accuracy ~10%. BET specific surface area (Ssp) was determined from
the Ar thermal desorption data.
Isotope oxygen exchange was carried out in a static installation with MS control both in
the temperature-programmed and isothermal modes as described elsewhere [60-63]. The
surface/bulk oxygen mobility for samples pretreated in O2 at 500 C was characterized by the
temperature-programmed desorption (TPD) in He flow. Reactivity of perovskites was
estimated by temperature-programmed reduction (TPR) by H2 or CH4 (1% in Ar, heating rate
of 5 K/min from 25 up to 880 ) carried out in kinetic installations equipped with GC and
PEM-2M analyzer (IR absorbance and electrochemical sensors) [11, 60-63].
Conductivity studies of samples were carried out on pellets using two-terminal scheme
with silver paste electrodes at temperatures from 50 to 750 in air. Electric properties were
measured with a Hewlett Packard 4284A Precision LCR Meter in the frequency range 20
Hz 1MHz. Powders of perovskites and composites were pressed into disks (15 mm in
diameter and about 1 mm thickness) and calcined at 1100C-1300 oC.
Electrical conductivities of cathode materials and relaxation profiles were measured using
the four-probe Van der Pauw method with platinum wires as current and voltage electrodes.
The measuring cell included Zr(Y)O2 test tube with the electrochemical oxygen pump.
Instrument/ device control and data acquisition used LabVIEW (National Instruments). For
the abrupt change in the oxygen pressure at the beginning of the reducing/oxidation step the
sample inside the electrochemical pump was blown by a pulse of the helium/oxygen flow
with a maximal rate of 200 ml/min-1. Then the pressure was regulated by the PID regulator of
the electrochemical oxygen pump with an accuracy of P=10-5 bar. The holder
accommodates a sample with the maximum length of the side of a square sample of 0.5 cm
and holds all four platinum electrodes and an K-type thermocouple.
2.2. Phase Composition, Morphology and Structural Features of Perovskites
XRD patterns of some samples of perovskites and perovskite-like oxides prepared via
Pechini route are shown in Figures 5-7. According to XRD data, mainly La2O2CO3 reflections
are present in XRD patterns of samples calcined at 500C. Phases of perovskites and K2NiF4-
type structures are formed after polymeric precursor calcination at 700C (Figs. 5, 6). LSM
calcination at higher temperatures results only in ordering of its structure, which leads to
splitting of reflections due to lattice distortion caused by cooperative Jan-Teller effect.
7300
La 0.8 Sr 0.2 Fe 0.5 Co 0.5 O 3
La 0.8Sr 0.2 Fe 0.8 Co 0.2 O 3
La 0.8Sr 0.2 Fe 0.6 Ni 0.4 O 3

Intensity, a.u.
La 0.8 Sr 0.2 FeO 3
LaNiO 3
La 2 Ni 0.8 Cu 0.2 O 3
La 2 NiO 4
0
0 10 20 30 40 50 60 70 80
2
Figure 5. XRD patterns of complex perovskites after calcination at 700C.
Figure 6. Powder XRD patterns showing the phase and structure evolution of LSFN (left) and LSM
(right) samples with temperature of calcination.
Annealing at 900C results in the disappearance of SrCO3 reflections in the

corresponding XRD patterns of Sr-doped samples. For LSNF perovskites, weak reflections of
NiO are still present and a new weak reflection of tetragonal La(Sr)2NiO4 phase appears (Fig.
6). The perovskite structure of LSFN0.3 remains orthorhombic after annealing at 900C, but
its lattice parameters decrease (Table 1). The symmetry of perovskite improves becoming
rhombohedral after calcination at 1100C, with admixtures of La(Sr)2NiO4 and NiO
appreciably decreasing.
Table 1. Lattice parameters and symmetry of LSFN0.3. perovskite phase
Sample T, C Lattice parameters

a, b, c, Symmetry1
LSFN0.3 700 5.5299 7.8259 5.5389 O
900 5.5271 7.8217 5.5365 O
1100 5.5033 5.5033 13.3295 R
1
O orthorhombic, R rhombohedral
Figure 7. Typical powder XRD patterns showing the phase evolution of LN (left) and L2N (right)
samples with temperature of calcination
60.4
60.0 700
900
59.6
59.2 1200
V/Z
58.8
58.4
0.1 0.2 0.3 0.4

XNi
Figure 8. Effect of Ni content in La0.8Sr0.2Fe1-xNixO on the lattice cell volume related per the formula
unit V/Z (3).
For LN sample some NiO is present after sample calcination at 700C. After calcination
at 1100 oC (Fig. 7), decomposition of the perovskite LaNiO3 phase was observed which is
explained by instability of Ni3+ cations at high temperatures [81]. In the case of L2N and
L2NCu samples mainly La(OH)3 and LaNiO3 are formed after calcination at 700 oC, with
only traces of a target tetragonal phase La2Ni(Cu)O4 being observed (Fig. 7, right). However,
single-phase La2Ni(Cu)O4 samples are formed after calcination at 1100 oC
Chiba at el shown that under air LSNF perovskites are stable single phase systems even
at 1400C when Fe mole fraction is higher than 0.5 [82]. In agreement with their data, the
structure of La1ySryNi1xFexO3 perovskite is rhombohedral at a high Ni content at all
calcination temperatures (Table 2). The increase of Ni content and sintering temperature leads
to LSFN lattice contraction (Fig. 8), as was earlier observed by other authors [83, 84] and
explained by a smaller size of Ni3+ cation and removal of residual Sr carbonates accompanied
by Sr incorporation into the lattice. For samples sintered at 1200 0C (Fig. 8), the trend is less
strong in the range of Ni content 0.3-0.4. This clearly correlates with some segregation of
NiO and tetragonal R-P type phase for LSFN0.4 sample at 1200 oC (Fig. 9).
o NiO; + tetragonal phase * perovskite + tetragonal phase o NiO

*
Intensity, a.u.
+ *
+o+ *
Intensity, a.u.
o
LSFNi0.4
+ +o+
+ o * LSFNi0.4
+ LSFNi0.3
LSFNi0.2
LSFNi0.3
LSFNi0.1
20 30 40 50 60 70 25 30 35 40 45
2 2
Figure 9. XRD patterns of La0.8Sr0.2Fe1-xNixO3 calcined at 1200 0C (left) and the patterns of the samples
with x = 0.3 and 0.4 with reflections of the impurity phases (right).
Note (Table 2) that a strong decline of specific surface area of LSFC and LSFN samples
with increasing calcination temperature is accompanied by rather moderate increase of X-ray
particle sizes (domains), so the extension of domain boundaries apparently increases with the
sintering temperature.
Table 2. Some characteristics of LSFN0.4 and LSFC0.2 samples
Structural characteristics*
Sample /
SBET (m2/g) Lattice parameter ()
Tc (C) d XRD, nm
a b c
LSFN /700 21.6 20.5 5.516 5.516 13.364
LSFN / 900 12 30.8 5.502 5.502 13.364
LSFN / 1200 1.4 74.0 5.503 5.503 13.336
LSFC/700 18.0 32.0 5.4935 7.7903 5.5307
LSFC/900 9.2 38.5 5.4946 7.7907 5.5335
LSFC/1200 2.1 58.5 5.4983 7.7856 5.5371
Rhombohedral structure of LSNF and orthorhombic structure of LSFC
Figure 10. Morphology and HREM image of the structure of La0.8Sr0.2Fe0.6Ni0.4O3 particle synthesized
via Pechini method after calcination at 700C for 2 h.
Figure 11. Morphology and structure of La0.8Sr0.2MnO3 sample synthesized via Pechini method after
calcination at 700C for 2 h.
Figure 12. Morphology and structure of La0.8Sr0.2Fe0.8Co0.2O3 sample synthesized via Pechini method
after calcination at 700C for 2 h.
a b
Figure 13. High resolution TEM images of particles of complex perovskites (a) BaFe0.4Co0.4Zr0.2Oy
and (b) Sr3.5Ba0.5Fe2.5Co1.5O10 prepared via MA route.
According to TEM data, perovskites prepared via Pechini route consist of crystalline
nano-sized (10-50 nm) particles (Figures 10-12). There are a lot of structural defects and
domain boundaries in such samples. This resembles the domain structure of disordered
complex Sr-Ba-Fe-Co-(Zr) -O perovskites prepared via MA route (Fig. 13) [60]. For all these
samples but those containing Zr, a cubic structure was observed, while particles were
comprised of disorderly stacked microdomains. The BaFe0.4Co0.4Zr0.2Oy sample particles
were comprised of the coherent intergrowth of domains with hexagonal BaFeO3-x [JCPDS 23-
1024] and monoclinic Ba2Fe2O5 [JCPDS 43-0256] structures.
C1s
284.8
288.8
285 290
Binding Energy (eV)
Figure 14. C1s spectrum of La0.8Sr0.2Fe0.6Ni0.4O3 sample calcined at 700C.

2.3. Surface Composition and Chemical State of Elements in

Complex Perovskites
For all perovskites at room temperature after calcination in air, XPS analysis revealed
substantial accumulation of carbonate species in the surface layer. In addition, the surface of
the samples contains considerable amount of adsorbed hydrocarbon species trapped from the
air. Typical C1s spectrum (sample La0.8Sr0.2Fe0.6Ni0.4O3- 700oC) presented in Fig 14 consists
of two components. The more intense component with Eb(C1s) = 284.8 eV is attributed to the
elementary carbon or hydrocarbons on the surface of samples. The component with Eb(C1s) ~
288.5 eV is unambiguously relates with carbonate species, and its share increases with the
calcination temperature. Since samples after calcination were cooled under air, and XPS
spectra were recorded without any pretreatment of samples, these carbonates are clearly
formed due to capture of CO2 from air by the surface layer of perovskites forming carbonate
species.
For Sr-containing samples, the increase of calcination temperature from 700 to 1100 oC
was accompanied by the surface enrichment with Sr, as is shown for the case of
La0.8Sr0.2Fe0.6Ni0.4O3 sample in Fig. 15. While the relative content of Fe and Ni cations varies
rather slightly (in the range of 10-20 relative %) (Fig. 16, 17), more pronounced effects were
observed for the charge state of both lanthanum and oxygen (Fig. 18, 19).
B - La(I)/La(II)
3.0 C - Sr/La
D - O(I)/O(II)
Ratio of components
2.5
2.0
1.5
1.0
0.5
900 1000 1100 1200 1300 1400

Temperature, K
Figure 15. XPS data for the surface composition of La0.8Sr0.2Fe0.6Ni0.4O3- sintered at different
temperatures; La(I) and La(II) are La cations with a lower and normal binding energy, respectively;
O(I) and O(II) are oxygen ions with a higher and normal covalency, respectively.
Sr3d La3d+Ni2p
500C
133.1 900C
1100C Ni2p1/2
871.9
833.0 833.9
133.5
130 135 140 825 840 855 870 885
Binding energy, eV
Figure 16. XPS spectra of La0.8Sr0.2Fe0.6Ni0.4O3- sample in the range of Sr 3d (left) and La and Ni
(right) binding energies.
Fe2p
500C
900C
1100C
709.5 709.9
710 720 730
Binding energy, eV
Figure 17. XPS spectra of La0.8Sr0.2Fe0.6Ni0.4O3- sample in the range of Fe 2p binding energies.
La3d5/2 La3d5/2 L3d5/2

832.8 834.6 832.5 834.4
834.3
832.7
830 835 840 830 835 840 830 835 840

Binding Energy (eV)
Figure 18. La3d5/2 spectra of La0.8Sr0.2Fe0.6Ni0.4O3 calcined at 700C (a), 900oC (b) and 1100oC(C).
O1s O1s O1s

528.6 530.9 530.4 530.4
a) 527.9 c)
b) 527.8
525 528 531 534 525 528 531 534 525 528 531 534
Binding Energy (eV)
Figure 19. O1s spectra of La0.8Sr0.2Fe0.6Ni0.4O3 calcined at 700C (a), 900oC (b) and 1100oC (c).
The curve fitting of the La3d5/2 lines shape (Fig. 18) resulted in two doublets with Eb ~
832.7 eV (LaI) and ~ 834.5 eV (LaII). The observed La(I) and La(II) states are assigned to
La3+ cations located in different oxygen environment.
The ratio of La(I) to La(II) increases with the calcination temperature (Fig. 15). Similarly,
two states of La cations were observed for other perovskites studied here.
Two states of oxygen anions are also clearly seen from O1s spectra presented in Fig. 19.
These oxygen states with Eb(O1s) ~528 eV and ~530.5 eV can be interpreted in terms of ionic
and covalent bonding of oxygen in two different cations sublattices. As is shown in many
papers, the O1s band shift reflects ionicity of metal- oxygen bonding. In general, oxygen with
a lower Eb(O1s) is characterized by more negative charge.
All these XPS features suggest that Sr enrichment in the surface layer favors
accumulation of oxygen vacancies, which decreases effective charge of La cations and
increases covalence of oxygen forms filling these vacancies. As is seen from Fig. 19, the
increase of the calcination temperature results in preferential growth of the surface oxide
layer where oxygen is in the covalent state. Moreover, increased intensity of component with
Eb(O1s) ~530.5 eV can be also caused by a higher content of carbonate species CO3
characterized by the same value of Eb(O1s) ~530.5 eV. For correct calculation of quantitative
dependence of ratio O(I)/O(II) (O(I) Eb(O1s) ~ 528 eV; O(II) - Eb(O1s) ~ 530.5 eV), the
contribution of carbonate species was subtracted from the total integral intensity of O1s
component with Eb(O1s) ~530.5 eV giving ratio OI/OII presented in Fig. 15.
Hence, for Sr-containing complex perovskites, substantial accumulation of oxygen
vacancies in the surface is suggested by XPS data. The same feature can take place within
domain boundaries known to possess less dense structure as compared with the bulk of
domains.
2.4. Conductivities of Perovskites
A high level of the total conductivity (>102 S/cm) at ~ 1000 K was revealed for LSFC
and LSFN samples sintered at sufficiently high (1100-1300 oC) temperatures (Fig. 20-21). A
close level of conductivity was determined for LSM in agreement with known results [85,
86]. The conductivity of LSFN perovskites is thermally activated exhibiting a polaronic
behavior within the studied temperature range. The character of the temperature dependence
of conductivity of doped nickelates is similar to that reported earlier [87-90]. According to
Huang et al [91], this polaronic conductivity is explained by an overlap of Fe4+/Fe3+ and
Ni3+/Ni2+ couples which provides two types of polaronic conduction, p-type and n-type, with
the p-type carriers moving on the Fe-atom array, and the n-type carriers moving on the Ni-
atom array.
o
La0.6Sr0.4Co0.2Fe0.8O3 (1500 C)
3
S/cm)
2 La0.6Sr0.2Co0.5Fe0.5O3
log (,
La0.8Sr0.2Co0.2Fe0.8O3
1
0 La0.8Sr0.2Ni0.4Fe0.16O3
1.5 2.0 2.5 3.0 3.5

1000/T, K -1
Figure 20. Temperature dependence of conductivity for LSFC samples and LSNF sample with a layered
perovskite- related structure sintered at 1500 (LSFC0.2) or 1100C (all other samples).
In general, different R-P type phases possess a lower conductivity than LSNF or LSFC
(Fig. 22). The temperature dependence of conductivity observed for La2Ni0.8Cu0.2O4 is similar
to that reported by Kharton et al [92], with values being somewhat lower. The latter might
be caused by lower sintering temperatures used for the pellets preparation in our case. Hence,
the data obtained suggest that La2Ni0.8Cu0.2O4 as well La2NiO4 and La0.8Sr0.2Ni0.4Fe0.6O3
provide a semiconductor-type behavior at low temperatures and a pseudometallic conduction
at high temperatures. In agreement with [93, 94], a partial substitution of Ni for Cu and/or Fe
increases conductivity of R-P type phases. La substitution for Sr and increase of the transition
metal cations content in R-P type phase up to Ln4Ni3O10 composition was shown to increase
conductivity as well [95, 96]. However, in our case conductivity of the R-P type mixed oxide
with the excess of transition metal cations La0.8Sr0.2Ni0.4Fe0.16O3- = (La0.8Sr0.2)2Ni0.8Fe0.32O4+
= (La0.8Sr0.2)4(Ni+Fe)2.24O8+ is not improved appreciably (Fig. 20), perhaps, due to
insufficient content of transition metal cations. A low pellet density could be important as
well, since conductivity of this sample is rather close to the specific conductivity of LSFC0.2
perovskite sintered at the same (1100 oC) temperature.

2.5
LSFN0.3
log(, S/cm)
2.0
1.5 LSFN0.4
1.0 LSFN0.2
0.5
LSFN0.1
0.0
1.0 1.5 2.0 2.5 3.0 3.5
1000/T, K -1
Figure 21. Temperature dependence of conductivity for LSFN samples with different Ni content
sintered at 13000C.
1.2
0.8
log(s, S/cm)
0.4
0.0 La2Ni0.8Cu0.2O4
La2NiO4
-0.4 Pr2Ni0.75Fe0.25O4
Pr4Ni3O10
-0.8
1.0 1.5 2.0 2.5 3.0 3.5
1000/T, K-1
Figure 22. Temperature dependence of conductivity for samples with R-P perovskite-type structure
sintered at 11000C.
In general, perovskite-like complex oxides Ba(Sr)-Fe(Co)-O comprised of disordered

intergrown perovskite- braunmillerite domains possessing a high lattice oxygen
mobility/oxide ionic conductivity demonstrate also not too high (1-10 S/cm) specific total
conductivity at 1000 K [97, 98].
2.5. Oxygen Mobility and Reactivity of Perovskites
2.5.1. Oxygen Isotope Exchange

As shown in Fig. 23, for Co- and Ni-containing perovskites calcined at 700 oC, noticeable
oxygen exchange starts at 300C. The dynamic extent of exchange (DEE) exceeds the
monolayer capacity at ~ 400 oC and reaches 10-15 monolayers at 550C, somewhat higher
values of DEE (and, hence, oxygen mobility) being observed for LSNF. This clearly
demonstrates a high lattice oxygen mobility for these perovskites, with the rate of bulk
diffusion exceeding that of the surface reaction. The increase of LSNF sintering temperature
to 900 oC shifts the start of exchange to higher temperatures and somewhat decreases DEE in
all studied temperature range (Fig. 24). After annealing at 1100 oC the surface steps
apparently proceed with a high rate only at temperatures exceeding 500 oC. However, at
temperatures exceeding 600 oC bulk diffusion occurs faster than for samples sintered at lower
temperatures, so DEE exceeds 30 monolayers at 700 oC. This increase of the lattice oxygen
mobility with sintering temperature can be explained by removal of carbonates from the
surface as well as from the vicinity of domain boundaries thus providing coherently stacked
domain boundaries as paths for the rapid oxygen migration. Moreover, as was revealed by
XRD data (vide supra), the extension of domain boundaries increases with the sintering
temperature, so this feature could indeed explain the increase of the lattice mobility in
perovskites with sintering temperature.
16 LaSrFeNi0.4
La2NiO4
XS, monolayers
12 LaSrFeCo0.2
8
LaSrNi0.4Fe0.16O3
4
0
200 300 400 500 600 700 800
o
Temperature, C
Figure 23. Temperature dependence of the dynamic degree of oxygen exchange for perovskites
calcined at 700o C.
Fig. 25 compares XS at 650 oC for some studied perovskites. It varies in a rather broad
range with the chemical composition and sintering temperature. Clearly, doped chromite
demonstrates rather low lattice oxygen mobility in the oxidized state. Rather low DEE values
were obtained for Gd-Sr-Co-O perovskite [99], despite report about a low polarization
resistance of this oxide as a cathode material [100]. As was considered by Mogensen et al
[101], the highest oxygen mobility in perovskites is provided by La in A sites due to the
minimum lattice distortion, which can explain the results obtained for Gd-Sr-Co-O
perovskite. Similarly, Cu-doped Pr2NiO4 oxide demonstrates a lower oxygen mobility than
Cu-doped La2NiO4 oxide. La2NiO4 shows a higher oxygen mobility than LaNiO3, which
correlates with a higher ionic conductivity of the former oxide due to interstitial conductivity
mechanism [102]. In agreement with the data of SIMS depth profile analysis [103, 104], the
oxygen mobility in La2NiO4 characterized by DEE exceeds that of LSFC calcined at the same
temperature. Moreover, decline of ionic conductivity in La2NiO4 due to Cu doping [105] is
also reflected in the decrease of DEE. LSFC annealed at 1100 apparently possesses the
highest lattice oxygen mobility which agrees with known data for dense samples [106].
However, LSFN sintered at the same temperature has at least comparable level of the lattice
oxygen mobility promising for the practical application. DEE values for these complex
perovskites are close to the values estimated for well-known complex perovskite systems with
the highest lattice oxygen mobility such as BaFe0.4Co0.4Zr0.2Ox and Sr4FeCo3O10 (Fig. 26)
[60].

1100
30
25
Xs, monolayers
20
700
15
10 900
5
0
100 200 300 400 500 600 700
Temperature, C
Figure 24. Temperature dependence of the dynamic degree of exchange for La0.8Sr0.2Fe0.6Ni0.4 O3
calcined at 700-1100C.
17
800 La0.8Pr0.2Mn0.2Mg0.2Cr0.6O3
1100
900 La0.8Sr0.2Fe0.5Co0.5O3
900
700 La0.8Sr0.2Fe0.8Co0.2O3
700 CoGd0.5Sr0.5O3
composition
700 Pr2Ni0.75Cu0.2Fe0.05O3
700 La2Ni0.8Cu0.2O4
700 La2NiO4
700 LaNiO3
700 La0.8Sr0.2Ni0.4Fe0.16O3-x
1100
900 La0.8Sr0.2Ni0.4Fe0.6O3
700
0 2 4 6 8 10 12 14 16 18 20 22 24156 157 158 159 160
o
Xs, monolayers, 650 C
Figure 25. Dynamic degree of the oxygen heteroexchange at 650 oC and 4 Torr O2 versus samples
composition and sintering temperature.
The lowest values of the oxygen mobility estimated by oxygen isotope exchange for
powders were found for lanthanum manganite either pure or doped by Ca or Sr (Fig. 26).
Manganites are well known to possess a low oxygen mobility in oxidizing conditions due to
the absence of oxygen vacancies in the lattice [24, 25]. Doping by Ca somewhat increases
oxygen mobility though not substantially.
500
400
300
200 BaFe0.4Co0.4Zr0.2Ox
XS, monolayers
Sr4FeCo3O10
100
4
3 La0.8Ca0.2MnO3
2
1 LaMnO3
0
350 400 450 500 550
o
Temperature, C
Figure 26. Temperature dependence of XS for lanthanum manganites and complex perovskites
BaFe0.4Co0.4Zr0.2Ox and Sr4FeCo3O10.
Isothermal exchange experiments carried out for the same powdered samples allowed to
estimate specific rates of the oxygen heteroexchange (vide supra). In general, these rates tend
to increase with sintering temperature for Sr-doped complex perovskites (Fig. 27, 28)
reflecting, perhaps, accumulation of oxygen vacancies in the surface layer of perovskite
particles revealed by XPS (vide supra). For LSNF sample, this is true for temperatures of
exchange exceeding 550 oC, while for lower temperatures specific rate of exchange decreases
with the increase of sintering temperature from 900 to 1100 oC (Fig. 27). This explains
variation of the temperature dependence of dynamic degree of exchange for this sample with
the sintering temperature (Fig. 25). With a due regard for segregation of Ni oxide species in
this sample after high-temperature annealing (vide supra), decreased low-temperature specific
rate of exchange can be assigned to this phenomenon, so annealing of some surface defects as
active sites of low-temperature exchange could be important as well. However, in the
intermediate temperature range of SOFC operation (600-700 oC), specific rate of the oxygen
exchange for LSNF sample is close to that of LSCF (Fig. 27). All other Ni- or Co-containing
complex perovskites studied here demonstrate a lower specific rate of exchange (Figs. 27,
28). In general, the scale of the specific rate of exchange variation is smaller than that of DEE
(cf. Figs. 25 and 28, 29). Indeed, LaMnO3 having much lower lattice oxygen mobility as
compared with mixed ferrite-cobaltites of Sr (Fig. 26) has rather close values of the specific
rate of oxygen heteroexchange (Fig. 29). From the point of view of surface chemistry and
heterogeneous catalysis [6, 55, 59, 80], comparable rates of the surface oxygen
heteroexchange are expected if the surface concentration of transition metal cations and
bonding strength of surface oxygen forms are close for different perovskites. While for simple
oxides of Mn, Co and Ni the bonding strength of the most reactive on-top (terminal) Me-O
oxygen forms was shown to be indeed comparable [107, 108], the situation is not so
straightforward for perovskites due to strong effect of the perovskite structure and chemical
composition on the types and fraction of the most developed surface planes (faces), different
terminations of the same face and possibility of its reconstruction/relaxation depending on the
oxygen stoichiometry and gas phase composition [108, 109]. Though usually observed
correlation between the oxygen surface exchange and bulk diffusion coefficients is explained
by importance of both surface and bulk oxygen vacancies for the surface reaction and bulk
diffusion, strong deviations are possible if detailed mechanism of the oxygen exchange is
considered [48-56].
17.6 1100 3-La0.8Sr0.2Fe0.5Co0.5O3

17.4 1100 3
lg R [O2/m2s], 4 Torr
17.2 1100 2
900 2-La0.8Sr0.2Fe0.8Co0.2O3
17.0
900
2
16.8
16.6 900 1
3
16.4
1-La0.8Sr0.2Fe0.6Ni0.4O3
16.2 1
1.0 1.1 1.2 1.3 1.4 1.5

1000/T, K-1
Figure 27. Temperature dependence of the specific rate of the oxygen heteroexchange at 4 Torr O2 for
La0.8Sr0.2Fe0.6Ni0.4O3 (1), La0.8Sr0.2Fe0.8Co0.2O3 (2) and La0.8Sr0.2Fe0.5Co0.5O3 (3) sintered at 900 or 1100
o
C (numbers at the left side of curves).
18
800 La0.8Pr0.2Mn0.2Mg0.2Cr0.6O3
1100
900 La0,8Sr0,2Fe0,5Co0,5O3
900 La0.8Sr0.2Fe0.8Co0.2O3
700
composition
700 CoGd0.5Sr0.5O3
700 Pr2Ni0.75Cu0.2Fe0.05O3
700 La2Ni0.8Cu0.2O4
700 La2NiO4
700 LaNiO3
700 La0.8Sr0.2Ni0,4F0,16O3-x
1100
900 La0.8Sr0.2Ni0.4Fe0.6O3
700
0
16,0 16,2 16,4 16,6 16,8 17,0 17,2 17,4 17,6 17,8 18,0
2 o
lg R [O2/m s], 4 torr, 650 C
Figure 28. Specific rate of the oxygen heteroexchange at 650 oC and 4 Torr O2 versus samples
composition and sintering temperature.
lg R, molec. O2/m2s
Sr4FeCo3O10
17
La0.8Ca0.2MnO3
LaMnO3
16
1.1 1.2 1.3 1.4 1.5 1.6
1000/ T
Figure 29. Temperature dependence of the specific rate of the oxygen heteroexchange at 4 Torr O2 for
Sr4FeCo3O10 calcined at 1000 oC, LaMnO3 and La0.8Ca0.2MnO3 calcined at 700 oC.
Sr4FeCo3O10 (T = 823 K) 0,9

1,0 La0.8Ca0.2MnO3 (T = 773 K)
0,8
fraction of isotopic molecules
fraction of isotopic molecules
18 16
0,8 O2 O2 0,7
18 16
0,6 O2 O2
0,6
0,5
0,4 0,4
18 16
0,3 O O
0,2
18 16 0,2
O O
0,1
0,0
0,0
0 10 20 30 40 50 60 70 80 0 10 20 30 40 50 60 70 8
time/min time/min
a b
Figure 30. Typical dynamics of the oxygen exchange for perovskite catalysts in isothermal experiments.
The possibility of exchange mechanism discrimination is clearly illustrated by Fig. 30, in

which the variation of the mole fractions of isotope dioxygen molecules with time is shown
for two samples greatly differing by the lattice oxygen mobility. In both cases the initial gas
was entirely comprised of the 18O2 molecules. However, for the Sr-Fe-Co-O sample (Fig.
30a), asymmetric isotope 18O16O molecules were not formed while for La-Ca-Mn-O sample
they appeared in considerable amounts (Fig. 30b). The complete absence of the asymmetric
isotope 18O16O molecules is the unequivocal evidence that only the 3rd type of exchange
operates (3 = 1). Indeed, in this case, according to Eq. (III) (vide supra) the dioxygen
molecule is exchanged for two atoms of oxygen from the oxide (which in the initial state was
characterized by the natural oxygen isotope abundance, i.e. the heavy isotope was practically
absent), and the number of emerging 16O2 molecules is equal to the number of consumed 18O2
molecules. For Ca-doped lanthanum manganite (as well as for LaMnO3) the 2nd type of
exchange apparently dominates since, according to Eq. (II), it generates asymmetric

molecules.
Analysis of the mechanism of exchange for various perovskites revealed that contribution
of the 1st type of exchange is very small being noticeable (~0.4) only for LaMnO3 at rather
low (~ 723 K) temperatures. The 3rd type of exchange usually dominates, while in some cases
(manganites, doped manganites) the contribution of the 2nd type (~ 0.6) becomes substantial
as well. The higher is the lattice oxygen mobility (the bigger amount of exchangeable
oxygen), the larger is the contribution of the 3rd type, which entirely determines exchange at
high temperatures.
The interrelation between the types of exchange and possible mechanism of this process
is considered in detail in [56]. The results obtained in the present work agree well with the
most commonly encountered two-step mechanism of dioxygen exchange with oxides [56]:
Step 1: O2 + Zads + ( )s ZOads + (O)s (42)
Step 2: ZOads + ( )s Zads + (O)s (43)
Here, a weakly bound form of adsorbed oxygen ZOads and a strongly bound oxygen of the
surface layer (O)s participate.
When this mechanism operates, oxygen exchange of the 1st type is absent, and
contributions of the 2nd and 3rd types are determined by the ratio of rates of steps (1) and (2):
2 = 1/(1 + 2) and 3 = 2/(1 + 2); (44)
Here i (i = 1, 2) the unidirectional rate of the ith step (the number of events per unit
of time) at equilibrium, (i.e. the rate of forward and backward reactions in conditions when
they are equal).
When the 1st type of exchange is observed, a reversible dissociative adsorption of oxygen
into a weakly bound form is suggested:
O2 + 2 Zads 2 ZOads (45)
At last, a high oxygen mobility in the lattice suggests that there is a reversible transfer
between the surface and bulk oxygen atoms:
(O)s + [ ]v ( )s + [O]v (46)
By itself, the isotope-exchange method does not allow to elucidate the physical nature of
the participating oxygen forms. However, it allows one to make certain conclusions about
their amount, atomic composition, bonding strength with the surface, mobility, types of
elementary steps and relation between their rates. To elucidate the nature of the active oxygen
forms, the isotope exchange method should be applied in combination with other
experimental (O2 TPD, TPR) and theoretical techniques (vide infra).
So, for manganites, operation of the 1 and 2 type of exchange at relatively low
temperatures provides a high overall rate of exchange despite a low concentration of
surface/bulk vacancies.
2.5.2. Oxygen TPD

Typical O2 TPD spectra are shown in Fig. 31-35, while their parameters (Wmax and ,
amount of oxygen desorbed in the TPD run) are given in Table 3 and in Fig. 36.
4
, molec. O2/m s
2
2
-15
W*10
1100
1
900
700
0 Isotherm
70 min
400 T, OC 800 880 880
Figure 31. Typical TPD spectra for La0.8Sr0.2MnO3-x perovskite calcined at 700, 900 and 1100C.
3.00E-007
Rate of oxygen desorption, mol/(g s)
2.50E-007
2.00E-007
1.50E-007 La-Mn
1.00E-007
5.00E-008 La-Ca-Mn
0.00E+000
200 400 o 600

T, C
Figure 32. Typical TPD spectra for LaMnO3+x and La0.8Ca0.2MnO3+x perovskites calcined at 700 C.
16
14
WO210 , molec. O2/m s

2
12
1
10
8 2
-14
6
3
4
2
4
Isotherm
0 70 min
400 800 880 880
O
Temperature, C
Figure 33. Typical TPD spectra for some complex perovskites. 1-La0.8Sr0.2Fe0.8Co0.2O3, 2-
La0.8Sr0.2Fe0.5Co0.5O3, 3-La0.8Sr0.2Fe0.6Ni0.4O3 calcined at 700C, 4-La0.8Sr0.2Fe0.6Ni0.4O3 calcined at
1000C (x0.1)
Figure 34. Effect of Ni content in La0.8Sr0.2Fe1-xNixO3 perovskites annealed at 1200 0C on O2 TPD

spectra.
Table 3. Parameters of O2 TPD for some perovskites
Temperature of Wmax10-15, molec.

System Tmax, C ,* monolayers
calcination, C O2/(m2s)
La0.8Sr0.2Fe0.8Co0.2O3 700 0.5 876 0.36
La0.8Sr0.2Fe0.5Co0.5O3 700 0.7 790 0.61
La0.8Sr0.2Fe0.6Ni0.4O3 700 0.6 880 0.20
1100 4.4 872 2.88
La0.8Sr0.2MnO3 700 0.7 817 0.76
La0.8Sr0.2MnO2.98 1100 2.9 723 1.32
La2Ni0.8Cu0.2O4 700 3.4 764 0.56
*Desorbed in the course of TPD run

300
1200
200
W, 1014molec. O2/m2s
25
20
15
900
10
0
700
Isotherm 70 min
-5
0 400 800880 880
o
Temperature, C
Figure 35. Effect of sintering temperature on O2 TPD spectra of La0.8Sr0.2Fe0.7Ni0.3O3 perovskites.
15 Wmax, 1014 molec. O2 /m2sec

, monolayers
14 300 LSFNi0.3
LSFNi0.3
13 LSFNi0.2
2 LSFNi0.2
25
LSFNi0.1 LSFNi0.1
0 0
700 800 900 1000 1100 1200

700 800 900 1000 1100 1200
Temperature, 0C Temperature, 0C
a b
Figure 36. Effect of La0.8Sr0.2Fe1-xNixO3 perovskites sintering temperature on the amount of oxygen
removed during TPD run (a) and the maximum rate of oxygen desorption (b).
According to these data, perovskites of a different chemical composition containing Co,

Ni, and Mn cations able to change their charge state in studied temperature range
accompanied by the loss of oxygen from the lattice show similar rates of O2 desorption in the
main peak situated at 700-880C. Traditionally, this rather narrow high-temperature peak is
termed as -form [6] related (at least, in part) to oxygen desorption from the bulk of
perovskite particles. Oxygen forms desorbed at lower temperatures (-forms) are assigned to
surface oxygen species [6]. Though in our experiments is below monolayer for samples
calcined at 700C (Table 3), this does not mean that only the surface oxygen is desorbed.
Indeed, for R-P type phase La2Ni0.8Cu0.2O4 calcined at 700 oC a higher maximum rate of O2
desorption can be explained by a higher lattice oxygen mobility provided by the oxygen
interstitials (vide supra). Usually the surface terminal oxygen forms with the heats of
adsorption up to 60 kcal/mol are desorbed at temperatures up to 600-700 oC [60, 107, 108].
Hence, several peaks situated in the 300-700 oC temperature range (Fig. 33-35) can be
assigned to desorption of various surface oxygen forms including those bound with the
regular and defect surface sites. Indeed, for manganites (Table 4), the surface coverage by
forms desorbed at low temperatures is quite small, so they are to be bound with defect sites.
The surface coverage by oxygen MO forms which can be desorbed before the main peak is in
the range of 10 % of monolayer, which agrees with results for simple transition metal oxides
[107, 108]. Doping lanthanum manganite with Ca cation decreases the amount of desorbed
oxygen which is mainly determined by respective decrease of the oxygen excess controlled
by the defect chemistry of this perovskite (Table 4). Doping with bigger Sr cation apparently
increases lattice oxygen mobility, especially for sample sintered at high temperatures.
For Co-containing system the intense oxygen desorption starts at lower temperatures as
compared with nickelates calcined at the same temperature, and a larger amount of oxygen is
desorbed correlating with a lower bonding strength of oxygen with Co cations than with Ni,
Fe or Mn cations [60, 107, 108]. Calcination of samples at higher (up to 1200C) temperature
results in increasing the rates of O2 desorption and amount of desorbed oxygen as well as in
shift of Tmax to lower temperatures. This suggests that sintering at higher temperatures
facilitates oxygen desorption and increases the lattice oxygen mobility which agrees with the
results of oxygen isotope exchange obtained in conditions of oxygen chemical equilibrium
between the gas and solid phases. LSNF perovskite in these conditions again demonstrates a
high oxygen mobility promising for the application as cathode material. Since the increase of
the desorption rate and amount of desorbed oxygen is clearly sensitive to the Ni content being
the most pronounced for La0.8Sr0.2Fe0.7Ni0.3O3 sample (Fig. 36), this suggests that indeed high
temperature sintering really accelerates the lattice oxygen mobility, perhaps, by providing
developed domain boundaries as paths for fast oxygen chemical diffusion. Since the amount
of desorbed oxygen is usually smaller than the amount of easily exchanged oxygen (vide
supra), preferential removal of oxygen from the vicinity of domain boundaries can be
suggested. Hence, O2 TPD data are in favor of domain boundaries as paths for fast oxygen
diffusion.
Table 4. Amount of oxygen desorbed from lanthanum manganite samples
Sample/Sspecific, Amount of oxygen desorbed, % monolayer

m2/g Weak surface formsa Strong surface formsb Lattice oxygenc
LaMnO3.19//20 0.7 11 40
La0.8Ca0.2MnO3.07/24 0.8 6.5 25
a
-removed up to 500 oC;b removed before the main desorption peak; c -removed in the main desorption
peak and during isothermal desorption at 880 oC.
2.5.3. Theoretical Analysis of The Heats of Oxygen Adsorption and Migration

To understand molecular-scale factors controlling oxygen mobility in perovskites,
theoretical analysis of possible types of surface adsorption sites and paths for oxygen
migration is very helpful [60, 102, 109-112]. For this purpose, semiempirical Interacting
Bonds method [60, 110] and DFT method [112] were applied for analysis of these features for
manganites, ferrite and cobaltites of La and Sr. Calculations by the Interacting Bonds method
have revealed the lowest surface energy for the (100) and (111) faces of perovskites (Table
5), which are thus expected to be the most developed under equilibrium conditions. For the
same face of a perovskite, different terminations can be realized.
Table 5. Surface energies (Hs) and heats (qO2) of oxygen adsorption (kcal/mol) on
perovskites
Parameter Face
(100) (110) (111)
Hs (LaMnO3) 69.6 126.9 110.7
Face composition MeO2 LaO LaMeO O2 LaO3 Me
qO2
O-Mn 53 65 68
O-La 72 79 46 58 64
O-La2 177
O-La2Mn 234
O-MnLa 160
O-Fe 51 68 72
O-Co 43 60 63
Due to steric reasons, bridging (O-Men) oxygen forms could not exist on a (100) face
with MeO2 termination, on a (110) face with O2 termination and on a (111) face of LaO3
composition. For the regular sites, the minimum heats of oxygen adsorption in the terminal
(O-M) form on Mn and La cations are comparable (Table 5), being in the range of 50
kcal/mol. This value is close to the activation energy of O2 desorption estimated for the main
TPD peak (-form, vide supra). However, since oxygen desorption from oxides is associated
with the reduction of cations, desorption of oxygen forms bound with La cations is considered
to be impossible. Bridging forms are too strongly bound to be desorbed up to 800 oC. Since
the enthalpy of oxygen removal from the bulk of lanthanum manganite or LSM is ~ 30
kcal/mol [20], this suggests that indeed desorption of the terminal Mn-O oxygen in the main
TPD peak is accompanied by fast replenishment of the surface vacancy by oxygen transferred
from the near-surface layers. All desorption forms observed at lower temperatures are
assigned to oxygen atoms bound with defect centers.
As a model for such a center, Mn cation migrating from the subsurface layer and filling
the La vacancy on the (100) face is suggested. Such a process was estimated to be
energetically favorable (~ 40 kcal/mol is gained). It could be even more favorable in the
vicinity of extended defects and in lanthanum manganites with a big oxygen excess. Indeed,
calculations by DFT [112] revealed that for lanthanum manganite with a big concentration of
La vacancies (La0.75MnO3), a stable local structure is formed by the shift of Mn cation from
the regular position towards the defect La-O plane. As the result, Mn4+ cation appears, and
oxygen bound with this cation decreases its effective charge to O-. In the surface defect
position, a Mn cation is 5-coordinated. The oxygen adsorption on such a center should
proceed with preservation of 5-fold Mn coordination, which can only be satisfied by
simultaneous breaking of the Mn-O bond with one of the surface oxygen. As the result, the
heat of oxygen adsorption on such a defect center decreases to 35 kcal/mol.
For Ca- or Sr-modified samples, Ca (Sr) migration from the bulk to the surface (100)
faces was found to be energetically favorable as well, which agrees with SIMS data [110]. As
a result, a part of the terminal O-Mn oxygen forms is converted into the bridging (Mn-O-Ca)
ones. This explains a decline of intensity of the high-temperature O2 TPD peak observed for
these samples.
For lanthanum ferrite and cobaltites, the energies of the oxygen adsorption were found to
be comparable with those in lanthanum manganite (Table 5).
Since lanthanum manganites with the oxygen excess are characterized by the presence of
cation vacancies in the lattice, it is worth to consider their effect on oxygen diffusion. It is
well known that the saddle-point for the oxygen atom jump in the perovskite lattice is a
triangle formed by two La and one Me cations [26, 27]. Though no diffusion is apparently
possible in stoichiometric perovskites without anion vacancies, however, in the course of
oxygen desorption or reduction, anion vacancies are generated at the surface and rapidly
migrate into the bulk. At moderate temperatures in dynamic conditions of
desorption/reduction, cation vacancies are not annealed due to a low mobility of cations, so
they could facilitate the oxygen atoms migration by decreasing the activation barrier.
Moreover, domain boundaries are usually characterized by less dense structure than domains,
which can be due to a high concentration of cation and anion vacancies.
To quantify this hypothesis, semiempirical Interacting Bonds method was applied to
estimate the oxygen migration energies for different configurations of the saddle point for
simple perovskites (Table 6).
Table 6. Migration energies for different saddle-point configurations
Metal Em, kcal/mole

Regular VMe VLa 2VLa
Cr 47 35 34 21
Mn 43 36 30 16
Fe 45 36 31 17
Co 43 36 28 14
As follows from the results shown in Table 6, cation vacancies indeed considerably
decrease the potential barrier for the oxygen migration in the lattice. For the case of two
lanthanum vacancies, calculated values are rather close to the experimental enthalpies of
oxygen vacancies migration [35].
To elucidate the effect of the local structure of perovskite-like Sr-Fe(Co)-O systems with
the oxygen deficiency on their oxygen mobility, SrCoOx (x=2.875, 2.75, 2.5 and 2) and
SrFeOx (x=2.875, 2.75 and 2) systems were analyzed by DFT technique. For SrCoOx two
kinds of oxygen vacancy ordering were revealed with energies 0.17 and 0.39 eV lower in
comparison with random vacancy distribution (Fig. 37). As the result of such ordering, pairs
of five-coordinated Co cations and string of 4-coordinated Co cations are formed. Transition
between these ordered vacancy systems could be a step in ionic conductivity mechanism with
activation energies of 0.61 and 0.78 eV depending on the direction of transformation. For
SrFeOx the oxygen vacancy formation is more energy consuming, and only one type of
vacancy ordering is preferable than a random distribution. Hence, these results agree in
general with hypothesis about interstitial mechanism of oxygen conduction in mixed Sr-Fe-
Co-O oxides with intergrown structure, where variation of the local coordination of Fe and
Co cations provides low energy barriers for the oxygen migration [111, 113].
a b c
Figure 37. Two types of oxygen vacancy ordering in SrCoOx perovskite-like oxide and a transition
configuration between them. a-pairs of 5-coordinated Co, E5-E4=0.18 eV; b-transition configuration, Et-
E4=0.78 eV; c-4-coordinated Co, E4=0.39 eV. Oxygen atoms outside the plane of Figure are not shown.
Oxygen atoms are marked by red, oxygen vacancies by white.
2.5.4. H2 TPR
The temperature-programmed reduction by H2 characterizes the surface oxygen reactivity
and lattice oxygen mobility in oxide systems under large gradient of the oxygen chemical
potential typical for operation of membranes for syngas production. Typical H2 TPR spectra
for basic types of studied perovskites are shown in Figs. 38-43. Usually, several peaks are
observed corresponding to main steps of transition metal cations reduction up to Me2+ (Mn)
or Meo (Ni, Co, Fe) state.
12
10
H2 consumption, a.u.
La-Mn La-Ca-Mn
8
6
4
2
0
0 200 400 600 800 1000
o
T, C
Figure 38. Typical H2 TPR spectra for LaMnO3.19 and La0.8Ca0.2MnO3.07 samples calcined at 700 oC.
The first peak of the lanthanum manganite sample bulk reduction at ~ 350 oC (Fig. 38)
corresponds to reduction of Mn4+ cations into Mn3+ state without destruction of perovskite
lattice. The second TPR peak corresponds to the LaMnO3 perovskite reduction into a mixture
of simple oxides (MnO +La2O3). Ca addition broadens the first TPR peak, which correlates
with the decline of the intensity of O2 TPD peak (Fig. 32). Sr doping produces similar but
more pronounced effect (Fig. 39), so the first peak splitting is apparent, which agrees with
earlier published results [114]. While position of the low-temperature component of this peak
at ~ 300 oC does not vary with the sintering temperature, the second component position is
shifted to higher temperatures. This suggests some heterogeneity of the state of easily
reducible Mn4+ cations in doped lanthanum manganite. To clarify the reason of this
heterogeneity, the amount of oxygen removed from the sample at different stages of reduction
was calculated.
-6
2.0x10
-6
1.6x10
mol H2/g s
-6
1.2x10
8.0x10
-7 900 1100
-7
4.0x10
0.0 500
0 200 400 600 800 1000
o
Temperature, C
Figure 39. Typical H2 TPR spectra for La0.8Sr0.2MnO3 samples calcined at 500 -1100 oC.
Figure 40. Typical H2 TPR spectra for LaCoO3 sample calcined at 1100 oC.
-6
8.0x10
-6
6.0x10
mol H2/g*K -6
4.0x10
-6
2.0x10
0.0
200 400 600 800 1000

o
Temperature, C
Figure 41. Typical H2 TPR spectra for LaFeO3 sample calcined at 1100 oC.
10
12 0 0 o C
mol/m *s
2
8
-7
W H2, 10
2 900oC
700oC
0
200 40 0 6 0 0 8 0 0 1 0 0 0
T e m p e ra tu re , o C
Figure 42. Effect of sintering temperature on H2 TPR spectra for La0.8Sr0.2Fe0.7Ni0.3O3 perovskites.
In estimation of the amount of oxygen required to be removed for reduction of all Mn4+
cations in a sample to the Mn3+ state, Mn3+ disproportionation into Mn2+ and Mn4+ by the
model of Roosmalen et al [115] was taken into account. For La1-xCaxMnO3+ the degree of
disproportionation was estimated by the equation [110, 115]:
(x+2+)/(1-x-2-2)2 = K; here K = 1.414

Comparison of experimental and theoretical amounts of oxygen removed in the low-
temperature peak revealed their good agreement for LaMnO3.15 and La0.8Ca0.2MnO3.07
samples [110]. For these samples, the total amount of Mn4+ due to the oxygen excess, Ca
substitution and disproportionation is ~ 6*10-4 mol O/g (Table 7). Complete reduction of all
Mn4+ cations in the bulk of particles at moderate temperatures in a narrow peak suggests a
high lattice oxygen mobility, which is not expected in stoichiometric undoped lanthanum
manganite where oxygen vacancies are absent (vide supra). As follows from Table 7, a
similar conclusion on a high rate of oxygen diffusion from the surface to the bulk of
lanthanum ferrite and cobaltite particles in the course of reduction without destruction the
perovskite lattice can be made. The maximum rates of reduction are comparable for
lanthanum manganite and ferrite, which agrees with similar heats of oxygen adsorption in the
reactive MO form (Table 5) and close values of oxygen migration energy (Table 6). A higher
reduction rate for lanthanum cobaltite agrees with a lower bonding strength of the MO
oxygen form (Table 5) as well as a lower oxygen migration energy (Table 6). Hence, under
conditions of very high oxygen chemical potential gradient between the surface and the bulk
of perovskites, the oxygen flux is determined by fast migration of oxygen vacancies from the
surface into the bulk of particles facilitated also by a lower bonding strength of oxygen with
transition metal cations in the high oxidation state as well as by the presence of cations
vacancies decreasing the oxygen migration energies in strongly non-equilibrium conditions of
TPR experiments [110].

WH2, 10-8 mol /m2s
900
4
700
0 1100
0 200 400 600 800

o
T, C
Figure 43. Effect of sintering temperature on H2 TPR spectra for La0.8Sr0.2Fe0.8Co0.2O3 perovskites.
For Sr-doped manganites the amount of oxygen removed in peaks situated in 300-500 oC
range exceeds that for LaMnO3.15 and La0.8Ca0.2MnO3.07 samples (Table 7). Hence, for Sr-
doped samples a part of Mn3+ cations is apparently reduced in the low-temperature range as
well. The increase of the maximum reduction rate with sintering temperature (Table 7)
correlates with respective increase of oxygen desorption rates (Fig. 31). This suggests that
facile paths of oxygen vacancies migration from the surface of LSM particles into the bulk
are associated with domain boundaries since their density increases with sintering (vide
supra). Generation of bulk oxygen vacancies due to Sr doping and sintering at higher
temperatures (Table 7) could facilitate reduction as well. However, since SrMnO3 is more
stable at low oxygen partial pressures than LaMnO3 [20], Mn cations having Sr cations as
next nearest neighbors could be less reactive than those neighboring with a cation vacancy
or generated by Mn3+ disproportionation. This seems to explain superposition of two H2 TPR

peaks of LSM reduction in the low-temperature range.
Table 7. Some characteristics of perovskites reduction by hydrogen
Sample Amount of oxygen removed in 1st peak(s), Rate at Tmax,

10-4 mol/g 1016 molec. H2/m2s
LaMnO3.19 5.8 4.3*
La0.8Ca0.2MnO3.07 5.5 2.7
La0.8Sr0.2MnO3, 500 oC 6.8 1.8
La0.8Sr0.2MnO3, 900 oC 10 3.5
La0.8Sr0.2MnO3, 1100 oC 8.2 27
LaFeO3 1100 C 17% of Fe3+ to Fe2+ 17 (480 oC)
LaCoO3 All Co3+ to Co2+ 2000 (470 oC)
*An order of magnitude higher rate at Tmax ~ 450 oC was obtained for a sample with Ssp. = 1 m2/g
The reduction of LSFN perovskite is characterized by two peaks located in the

temperature range 200400C followed by continuous reduction at temperatures exceeding
600 oC (Fig. 42). Similar pattern was earlier observed for LaNixFe(1-x)O3 perovskites [80,
116]. The low-temperature peaks were shown to correspond to reduction of Ni3+ and Fe3+ into
2+ state thus producing a defect perovskite structure, while the high-temperature peak to
reduction of Ni and Fe cations to Meo state forming an alloy dispersed on lanthana particles.
As judged by the balance of removed oxygen for LSNF sample (not shown for brevity), this
scheme is also valid in our case. The increase of LSFN calcination temperature also increases
the peak reduction rate (Fig. 42) and shifts reduction peak to higher temperatures. This
correlates with the increase of the specific rate of oxygen desorption from this perovskite with
the increase of sintering temperature (Fig. 36b). Hence, for this system the oxygen mobility in
the oxidized and reduced state apparently correlates.
For La0.8Sr0.2Fe0.8Co0.2O3 perovskites (Fig. 43) H2 TPR spectra are more complex
consisting of several overlapping peaks due to simultaneous reduction of Co and Fe cations.
The peak rate of reduction in the middle-temperature range goes through the maximum with
the sintering temperature, perhaps, reflecting rearrangement of the real structure.
2.5.5. CH4 TPR

Reduction of perovskites by methane (Figs. 44-47) usually proceeds at higher
temperatures as compared with the reduction by H2 (Figs. 38-43) due to strong C-H bond in
CH4 molecule. Isothermal reduction kinetics corresponds to that of a typical topochemical
process with nucleation of reduced phases (oxides or metals), growth of nuclei with time
described by the Avrami- Erofeyev equation followed by their overlapping and decline of
observed reduction rate described by the contracting spheres model [80] (vide infra examples
of isothermal reduction for LSFN-GDC composites). So the maximum rate of perovskites
reduction in isothermal experiments is expected to be determined both by the density of
possible nucleation sites (usually, sites with enhanced reactivity associated with defects such
as dislocations, including those located at domain boundaries etc, vide supra), as well as the
bulk oxygen mobility. Since O2 TPD and oxygen isotope exchange studies suggest that
oxygen mobility correlates with the extension of domain boundaries, the maximum rate of
perovskites reduction by methane is expected to correlate with the bulk oxygen mobility as
well. For CH4 as reductant, another important parameter is selectivity of methane
transformation into products of deep oxidation (H2O and CO2) and selective oxidation (CO
and H2) [11, 60, 63]. It is determined both by bonding strength of reactive surface oxygen
forms (determines primary selectivity of CH4 activation) and by the flux of oxygen from the
bulk of particles to their surface (determines possibility of CO and H2 combustion).
Moreover, for some perovskites containing Ni, Fe and Co cations, dissociation of CH4
molecules yielding H2 and surface carbon proceeds on metal particles even after complete
reduction of samples [117]. All this determines a complex shape of CH4 TPR spectra shown
in Figs. 44-47. Reduction process starts by formation of deep oxidation products at 400-500
o
C. Several overlapping peaks of CO and H2 evolution are usually observed reflecting step-
wise reduction of perovskites discussed above. Similarly, several peaks of products formation
are also observed in CH4 TPR spectra of MnO2, apparently reflecting sequence of phase
transitions MnO2-Mn2O3-Mn3O4-MnO in the course of reduction (Fig. 45). A higher
selectivity of CO + H2 formation in the course of La-Sr-Mn-O reduction by CH4 (Fig. 46)
versus that for the MnO2 case is the most probably explained by dilution effect of irreducible
La cations and a higher, in general, bonding strength of oxygen forms bound with Mn cations
in perovskite. Coordinatively unsaturated Mn2+ cations located at the surface of lanthana
particles continue to dissociate CH4 molecules even after complete reduction of manganite
forming carbonaceous deposits.

1.25 H2
CH4
1.00
0.75
%
0.50
0.25 H2O CO
CO2
0.00
Isotherm70 min
400
T, oC 800880 880
Figure 44. CH4 TPR spectra for La0.8Sr0.2MnO3 sample calcined at 900 oC.
Maximum rates of CO and CO2 evolution for manganites (Table 8) are usually achieved
in the isothermal part of TPR run at 880 oC. Maximum rates of CO and CO2 formation are not
sensitive to the LaMnO3 surface modification by Pt, which agrees with suggestion about their
control by the rate of oxygen diffusion from the bulk of particles to their surface. Higher
maximum rates of both CO and CO2 formation for Sr and Ba- containing systems (Table 8)
correlate with their higher lattice oxygen mobility (vide supra). For these systems as well as
for LaCoO3, a lower CO selectivity seems to be determined by a high oxygen flux leading to
oxidation of syngas into CO2 and H2O.
0.75
H2O
0.50
H2
%
0.25
CO2
CO
0.00
Isotherm 70 min
T, oC
400 800880 880
Figure 45. CH4 TPR spectra for sample MnO2 (Ssp. 10 m2/g).
Table 8. Maximum rates of some perovskites reduction by CH4.
Sample LaMnO3.15 0.3%Pt/ La0.8Ca0.2 LaCoO3. LaFeO3 Sr4FeCo3 Sr3.5Ba0.5

Wmax, 1016 at. LaMnO3 MnO3.07 Fe2.5Co1.5
O/m2s, 880 oC
CO 1.1 1.2 0.6 3.7 15 10 20
CO2 1.5 1.3 1.4 100 9.4 51 90
2.0
H2
1.5
1.0 CH4
%
0.5 CO
CO2 H2O
0.0 Isotherm 70 min
T, oC
400 800 880 880
Figure 46. CH4 TPR spectra for LSFC sample calcined at 900 oC.
For Ni- or Co-substituted La-Sr-ferrites (Figs. 46, 47) selectivity of CH4 transformation
into deep oxidation products is clearly lower than that for LSM. At 880 oC, complete
conversion of CH4 is accompanied by formation of stoichiometric amount of H2 (2% H2 for
1% CH4 in He) while CO content is substantially lower (~ 0.5% or less. Hence, carbon
deposition on the surface of these samples takes place.
2.0 H2
Concentration, %
1.5
CH4
1.0 CO
0.5
CO2 H2O
0.0 Isotherm 70 min
400 800880 880

Temperature, oC
Figure 47. CH4 TPR spectra for LSFN0.3 sample calcined at 1200 oC.
The increase of the maximum rates of CO and CO2 evolution with sintering temperature
of perovskites (Fig. 48) clearly correlates with the increase of the oxygen mobility estimated
by oxygen isotope exchange, O2 TPD and H2 TPR (vide supra).
70 LSFN0.3
LSFN0.3
24
Wmax, 1016 at. O/m2s
60
Wmax.CO2, 10 at. O/m s
2
50 20
LSFC
16 LSM
40
16
30 12 LSFC
20 LSM 8
10 LSFN0.2 4 LSFN0.2
0 700 800 900 1000 1100 1200

700 800 900 1000 1100 1200 o
o Tsint., C
Tsint., C
Figure 48. Effect of perovskites sintering temperature on the maximum rates of CO (a) and CO2 (b)
formation in the course of CH4 TPR.
Hence, among studied perovskites, such systems as LSM, LSFC and LSFN appear to be
promising for their application as electronic-conducting components in nanocomposite
systems with a high mixed ionic electronic conductivity in combination with ionic-
conducting oxides (doped ceria or zirconia), since they possess a high conductivity along with
reasonably high lattice oxygen mobility in conditions of broadly varying gradient of oxygen
chemical potential.
3. PEROVSKITE-FLUORITE NANOCOMPOSITES
As it has been mentioned above, composite electrodes have many advantages compared
to single-phase materials. Combination of high catalytic activity, mixed electronic and ionic
conductivity can provide excellent electrochemical properties of such materials. That is why
composite materials based on perovskite solid solutions are regarded as the most promising
cathode materials. In order to be used in real SOFC, the material should be compatible with
solid electrolyte, interconnect and sealing materials. Mechanical compatibility of two or
several densely contacting solid state materials is limited by difference in the thermal
expansion coefficients which should be as small as possible. Mechanical stresses accumulated
in the bulk of the material may lead to the break and pilling of cathode layers. The use of
composites allows to redistribute stress directions by variation of the electrode microstructure
thus diminishing the probability of the accumulation of the critical stress and avoiding fast
degradation of the material.
Though in literature there are a lot of data on application of MIEC composites as cathode
materials of SOFC and oxygen-conducting membranes [118-129], their synthesis procedures
were usually based on mixing and milling of rather coarse-grained oxides. As the result, the
interface between oxides with different predominant type of conductivity is usually not
developed even in dense materials. On the other hand, the most pronounced non-additive
effects related to the oxygen mobility are expected in the case of nanocomposites. It is
explained by morphological instability of nanosystems at elevated temperatures. Indeed,
nanocrystals or nanoceramics of pure oxides rather easily recrystallize into coarse-grained
powders or ceramics. It is caused by large thermodynamic affinity to recrystallization and
sintering due to high value of excess surface energy. In nanocomposite materials there are
many interfaces and the excess surface energy of both components is partially compensated
by the adhesion energy between the phases. It results in much lower affinity to the
recrystallization process and a higher stability of the morphology of nanocomposites. With a
due regard for the possible advantage of these materials, new procedures for synthesis of
perovskite fluorite nanocomposites were elaborated [11, 61, 62, 98, 99, 130-134], and results
on studies of their structural characteristics and transport properties are presented below.
3.1. Preparation and Characterization of Nanocomposites
The methods of the preparation of perovskites are given above. Fluorites were
synthesized similarly to perovskites using Pechini route. The ethylene glycol and distilled
water solutions prepared from metal nitrates were mixed together at room temperature under
stirring followed by addition of ED. The prepared solution was stirred for 60 min and then
heated at 70C for 24 h allowing for the gel formation. The gel was calcined at 25- 500C in
the case of GDC. Commercially available powders of ScCeSZ (DKKK corporation, Japan) or
other fluorite phase and nano-sized powder of GDC, LSM, LSFC and LSFN synthesized via
Pechini route were used for nanocomposites preparation via mixing in 1:2 (GDC:LSFC) or
1:1 ratio (all other composites) by weight.
It was found that the use of organic surfactants leads to smaller crystallite size of fluorite.
Powders were ultrasonically dispersed in a solvent (distilled water, acetone) with addition of
2 wt% of a surfactant (polymethylmetacrylate, polyethylene glycol, polyvinyl butiral) using a
Ika T25 ULTRA-TURRAX basic (IKA, Germany) generator. The dispersed mixture was
filtered, dried in air at 25-200C, pressed into pellets (d=15 mm, h=1 mm) under ~150
kgf/cm2 and calcined at temperatures up to 1350C for 5 h.
For comparison, La0.6Sr0.4Fe0.8Co0.2O3+x Ce0.9Gd0.1O2-x (1:2) composite (LS0.4FC-GDC)

prepared at Imperial College (London, UK) was characterized as well. For this composite
synthesis, commercially available powders of La0.6Sr0.4Co0.2Fe0.8O3- (Praxair Specialty
Ceramics, Seattle, USA) and Ce0.9Gd0.1O1.95 (Fuelcellmaterials.com, Ohio, USA) were mixed
in the appropriate amounts by ball-milling in ethanol. After drying, the composite powders
were pressed into pellets and densified by sintering at 1250oC for 4 hours. Further details
have been given in a previous publication [118, 119, 135]. Some pellets were crushed and
ground with a pestle and mortar into powder (Ssp. 0.2 m2/g, average particle diameter 3
microns).
The samples of nanocomposites were mainly characterized by the same techniques as
pure perovskites. In addition, UV-vis spectra were recorded relative BaSO4 standard in the
1100054000 cm1 range using a UV-2501 PC Shimadzu spectrometer equipped with the
diffuse reflectance ISR-240 unit. Powdered samples were put into quartz cell of 2 mm optical
length. Spectra were presented as Kubelka- Munk function of the wave number.
For comparison with results of conductivity relaxation studies, values of the oxygen
chemical diffusion coefficient, Dchem, were estimated by following the pellet weight
relaxation using a STA 409 PC "LUXX" NETZSCH machine after a step-wise change of O2
content in the N2 stream from 14 to 1.4%.
The heats of oxygen absorption in the surface layers of perovskites and composites
reduced by high-temperature pretreatment in He were estimated using a SEN SYS TG-DSC
flow microcalorimeter and pulses of O2 in He at 450 oC.
P
Intensity (a.u.)
P
O O
P
P P
P
P O
P
1100
900
700
20 30 40 50 60 70
2 (deg)
Figure 49. XRD patterns of LSM +ScCeSZ nanocomposite sintered at 700-1100 oC. P-perovskite, O-
fluorite phase reflections.
3.2. Phase Composition and Structure
XRD. As follows from XRD data (Figs. 49-54), all composites are mainly comprised of a
mixture of perovskite and fluorite phases. However, the structural peculiarities of samples
depend on their composition, preparation conditions and annealing temperature. Results of

structural investigations showed that in all composites sintered at temperatures below
1000C, only initial perovskite and fluorite-like phases are observed. In LSM-ScCeSZ (Fig.
49) and LSFC-GDC (Fig. 50) composites new phases were not revealed even at high sintering
temperatures. However, lattice parameters of both phases vary with the sintering temperature
(Table 9) reflecting redistribution of elements between phases. The increase of the lattice
parameters of LSM in LSM-ScCeSZ nanocomposite with Tsint. clearly demonstrates
incorporation of bigger cations into the perovskite lattice. This can be explained only by
incorporation of large Sc3+ and Zr4+ cations into the B sublattice of perovskite.
Figure 50. Typical XRD patterns of composite 70%La0.8Sr0.2Fe0.8Co0.2O3-x+30%Ce0.9Gd0.1 calcined at

different temperatures (a) and expanded range 2 46-48o (b).
XRD patterns of 50%LSFN0.3 + 50%GDC composite calcined at different temperatures

are shown in Fig. 51. In the composites dried after ultrasonic treatment in solvent and
calcined at 700 C, reflections corresponding to the cubic lattice of GDC and orthorhombic or
rhombohedral lattice of perovskite are observed. For both phases in this composite, X-ray
particle sizes are close to those for starting powders (Table 9). Sintering of composites at 900
C increases the intensity of all reflections and narrows them (Fig. 51) due to lattice ordering
and particle size increase (Table 9), without any apparent change of peaks positions.
At sintering temperature 1200 C, for composites with LSFNx (x=0.1-0.2) only reflections of
perovskite and fluorite phases are observed (Fig. 51). For LSFN0.3 GDC composite, weak
NiO reflections appear, while for LSFN0.4-GDC composite intensity of NiO reflections
increases, which is determined by the thermodynamic instability of corresponding perovskites
[136, 137]. Note that in contrast to individual perovskites LSFNx (x=0.3-0.4) (vide supra), in
the patterns of their composites the reflections of tetragonal phase are absent (Figs. 51-53).
This can be explained by incorporation of La into GDC leading, among other consequences,
to La2NiO4 consumption. Indeed, reflections corresponding to Ce0.9Gd0.1O1.9 phase are shifted
to lower angles (Figs. 50, 53) and the lattice parameter of fluorite phase in composites
increases in all cases (Table 9) that is determined by incorporation of La and Sr cations [128]
as well as a part of Ni into the fluorite lattice [138]. Similarly, parameters of perovskite phase
in composites are changed as well (Table 9). Furthermore, the structure of perovskite in the
composite with LSFN0.3 changes from the orthorhombic to the rhombohedral type.
Table 9. Some structural characteristics of nanocomposite samples
Fluorite Perovskite
Sample / SBET
2 Lattice parameter ()
Tc (C) (m /g) d XRD (nm) Lattice parameter a () d XRD (nm)
a b c
GDC/500 32 9 5.413
GDC/700 32 24 5.420
GDC/900 12 40 5.418
GDC/1200 3 72 5.418
LSFNi0.1+GDC/700 24 18 5.420 27 5.516 5.566 7.860
LSFNi0.1+GDC/900 9.4 32 5.434 42 5.538 5.522 7.813
LSFNi0.1+GDC/1200 2.6 65 5.444 65 5.525 5.545 7.850
LSFNi0.2+GDC/700 23 19 5.418 22 5.548 5.498 7.814
LSFNi0.2+GDC/900 22 32 5.417 31 5.540 5.495 7.796
LSFNi0.2+GDC/1200 2 44 5.438 59 5.540 5.501 7.806
LSFN0.3+GDC/700 30 14 5.418 21 5.493 5.560 7.790
LSFN0.3+GDC/900 12 30 5.424 29 5.480 5.532 7.789
LSFN0.3+GDC/1200 2.4 39 5.445 50 5.491 5.535 7.802
LSFN0.4+GDC/700 27 18 5.418 20 5.508 5.508 13.364
LSFN0.4+GDC /900 9.5 36 5.423 35 5.500 5.500 13.363
LSFN0.4+GDC /1200 1.0 37 5.454 59 5.501 5.501 13.344
LSFC+GDC /700 23 23 5.421 27 5.523 5.523 13.448
LSFC+GDC /900 9.2 29 5.425 41 5.519 5.519 13.452
LSFC+GDC /1100 2.1 100 5.432 54 5.517 5.517 13.450
LSM-ScCeSZ /700 26 13 5.093 33 5.499 5.499 13.365
LSM-ScCeSZ /900 8.1 24 5.097 48 5.513 5.513 13.378
LSM-ScCeSZ /1100 1.8 52 5.099 71 5.526 5.526 13.381
LSFN0.2-ScCeSZ
700 5.513 5.545 7.827
900 5.548 5.522 7.808
1200 5.631 5.576 7.879
For composite with ScCeSZ sintered at high temperatures (Fig. 54), along with NiO, a
pyrochlore-type La2Zr2O7 and a layered perovskite LaSrFeO4 phases were formed as well.
Hence, for Ni-containing composites, both Ni content in the perovskite phase and sintering
temperature should be optimized to avoid segregation of low-conducting pyrochlore-type
phases. The lattice parameters of LSNF phase increased with composite sintering
temperature, which can be explained by a partial reduction of small Ni3+ cations to larger Ni2+
cations accompanied by the oxygen loss at high temperatures as well as by a partial
incorporation of larger Zr and Sc cations into the perovskite lattice.
I * perovskite * perovskite NiO

I fluorite
NiO
*
*
Intensity, a.u.
I
Intensity, a.u.
I I
* *
* * I I
* * ** 0 * 0
1200 C * 1200 C
* 900 C
0
0
900 C *
0
0
* 700 C
700 C
20 40 60 36 38 40 42 44
2 2
Figure 51. XRD patterns of composite La0.8Sr0.2Fe0.7Ni0.3O3 +Ce0.9Gd0.1O1.95 sintered at different

temperatures.
*-perovskite; I-fluorite; o-NiO * perovskite o NiO

I *
I
*
o
I
* I o *
* *I * LSFNi0.4
* * ** LSFNi0.4
o
** o
*
Intensity, a.u.
Intensity, a.u.
LSFNi0.3
LSFNi0.3
LSFNi0.2
LSFNi0.2
LSFNi0.1 LSFNi0.1
20 30 40 50 60 70 36 38 40 42 44
2 2
Figure 52. XRD patterns of composites La0.8Sr0.2Fe1-xNixO3 + Ce0.9Gd0.1O1,95 calcined at 1200 C.
TEM and EDX. According to XRD (Table 9) and TEM data (Fig. 55), the increase of
calcination temperature is accompanied by porosity annealing and the increase of particle size
of both components. Estimation of the particle sizes for separate phases in nanocomposites
from X-ray diffraction patterns using Sherrer equation revealed that for composite systems
domain sizes of P and F phases remain on a nanoscale level even after sintering at 1200
(Table 9). As judged by TEM data, the nanocomposites prepared via ultrasonic dispersion
route are characterized by uniform spatial distribution of P and F domains in composites (Fig.
55- 57). Such spatial uniformity allows formation of developed nearly coherent P/F interfaces
at high temperatures, though sizes of P and F domains increase. As follows from EDX data,
pronounced redistribution of elements between P and F domains takes place (Fig. 56- 58):
transition metal cations are incorporated into F domains (mainly, surface layers), while Sc, Ce
and Zr cations migrate into P domains (much stronger for LSNF phase). Domain interfaces
appear to be rather coherent demonstrating a good epitaxy between P and F phases.
* perovskite I fluorite + tetragonal phase

I
Intensity, a.u.
I
Intensity, a.u.
I composite
I
* * *I
* * * composite
LSFNi0.3
+
LSFNi0.3
GDC GDC
20 30 40 50 60 22 28
2 2
Figure 53. XRD patterns of La0.8Sr0.2Fe0.7Ni0.3O3, Ce0.9Gd0.1O1.95 and composite

50%La0.8Sr0.2Fe0.7Ni0.3O3 + 50%Ce0.9Gd0.1O1.95 calcined at 1200 C.
LSFN0.2
Zr0.89Sc0.1Ce0.01Ox
La2Zr2O7
5000
NiO
LaSrFeO4 LSFNi0.2+ ScCSZ
Intensity, ab. un.
4000

1200 0C

3000
900 0C
2000
1000
700 0C
0
20 30 40 50 60 70 80 90
2
Figure 54. X-ray diffraction patters of LSFNi0.2 -ScCeSZ composite sintered at different temperatures.
Figure 55. Typical morphology of LSNF-GDC composite particles sintered at 700 (a), 900 (b) and 1200
(c).
Figure 56. Typical morphology of particles in LSNF-GDC composite sintered at 1200 C and
respective EDX spectra from regions 1-3. Elemental composition: 1 - 19 wt.% La0.97Sr0.03Fe0.68Ni0.32O3
+ 81 wt.% Ce0.93Gd0.07O2, 2 - 21 wt.% La0.94Sr0.06Fe0.74Ni0.26O3 + 79 wt.% Ce0.93Gd0.07O2, 3 - 82 wt.%
La0.98Sr0.02Fe0.63Ni0.37O3 + 18 wt.% Ce0.68Gd0.32O2.
UV-Vis. Spectra of La0.8Sr0.2MnO3-x samples are shown in Fig. 59. The bands in the
visible range can be assigned to d-d 4A2g4T2g (16000 cm-1) and 4A2g4T1g (F) (19000 cm-1)
transitions in Mn4+ cations (d3 configuration, basic term 4A2g in the octahedral crystal field)
[139]. Presence of Mn4+ cations in LSM can be explained both by charge compensation due
to Sr doping and Mn3+ disproportionation (vide supra). Note, though, that a proper
interpretation of LSM spectra requires taking into account spin dependent hybridization
between Mn d states and O p states [140, 141]. As the result, the majority Mn d bands overlap
the O p bands while the minority Mn d bands are separated by a gap from the O p bands.
Figure 57. High resolution image of LSM-ScCeSZ nanocomposite sintered at 900 oC and composition
of neighboring domains. ffluorite phase, p-perovskite phase.
Figure 58. Lattice spacing and elemental composition of neighboring domains in LS0.4FC-GDC
composite by TEM with EDX. Left: 70%wt. La0.82Sr0.18Fe0.78Co0.22O3 + 30%wt. Ce0.82Gd0.18O2, d
=1.96, 3.90 (1) and 17%wt. La0.87Sr0.13Fe0.66Co0.34O3 + 83%wt. Ce0.92Gd0.08O2, d= 2.76 (2). Right :
86%wt. La0.59Sr0.41Fe0.78Co0.22O3 + 14%wt. Ce0.60Gd0.40O2; d = 2.78 (1) and 80%wt.
La0.63Sr0.37Fe0.79Co0.21O3 + 20%wt. Ce0.55Gd0.45O2, d =2.72 (2).
LSM samples calcined at 700 and 900C (curves 1-2) are characterized by a broad
absorption band at 13700 cm1 corresponding to Mn cations in weakly distorted octahedral
coordination of pseudocubic disordered perovskite phase (vide supra XRD data). For samples
sintered at 1100-1200C (curves 3-4) the oxygen excess declines, structure becomes more
ordered (orthorhombic) and domain sizes increase (vide supra). This is reflected in the
decrease of the intensity of band in the visible range, its shift to 15000 cm1 and appearance of
a shoulder at 19800 cm-1 caused by cooperative Jan-Teller distortion of octahedra in LSM.
13700
20 1
2
16 15000
19800
F(R)
12 3
4
8
0
10000 20000 30000 40000 50000
Wavenumber, cm-1
Figure 59. UV-vis spectra of La0.8Sr0.2MnO3-x calcined at 700C (1), 900C (2), 1100C (3) and 1200C
(4).
12 14000
10 1
19000
8 2 15700
F(R)
6
3
4
0
10000 20000 30000 40000 50000
Wavenumber, cm-1
Figure 60. UV-vis spectra of LSMn-ScCeSZ composite sintered at 700C (1), 900C (2), 1100C (3).
Similar variations with increasing the sintering temperature were observed in UV Vis
spectra of LSM-ScCeSZ composite (Fig. 60). However, in composite the blue shift and
splitting of the band is observed already after sintering at 700 oC. After sintering at 1100 oC
the shift of band in composite is more pronounced than in LSM, suggesting a stronger
distortion of oxygen octahedra around Mn cations. This can be explained by incorporation of

bigger Sc and Zr cations into the B sublattice of LSM.
In spectra of LFN0.3 samples a band at 12500-14100 dominates shifting to higher
wavenumbers with the increase of sintering temperature (Fig. 61). This band is absent in the
spectra of LaFeO3 sample (Fig. 62), so it can be assigned to d-d transitions in Ni3+ (d7)
cations. Indeed, this band is present in the spectra of NiO calcined at 900 oC due to some
oxygen excess (and, hence, presence of Ni3+) in this sample (Fig. 63). Blue shift of this band
position with sintering temperature can be explained by ordering of the LFN structure.
Substitition of La for Sr does not affect appreciably the shape and position of this band (Fig.
64). Decrease of Ni content in LSFN leading to the lattice expansion, and, hence, increasing
Ni-O bond length and decreasing its strength is accompanied by the red shift of the band
position (Fig. 65).
24 12500
1
20
16 13300
2
F(R)
14100
12
3
8
0
10000 20000 30000 40000 50000
Wavenumber, cm-1
Figure 61. UV-Vis spectra LaFe0.7Ni0.3O3-x sintered at 700C (1), 900C (2) and 1200C (3).
8
7
6
5
F(R)
4
3
2
1
0
10000 20000 30000 40000 50000
Wavenumber, cm-1
Figure 62. UV-Vis spectrum of LaFeO3 calcined at 1100 oC.

7 29400
6
5
26400
4
F(R)
3 24000
21800
13900
2 15400
1
0
10000 20000 30000 40000 50000
Wavenumber, cm-1
Figure 63.UV-Vis spectrum of NiO calcined at 900 oC.
12800
20 1
16
12 14400
F(R)
2
8
0
10000 20000 30000 40000 50000
Wavenumber, cm-1
Figure 64. UV-Vis spectra of La0.8Sr0.2Fe0.7Ni0.3O3-x sintered at 700C (1) and 1200C (2).
Spectra of Ce0.9Gd0.1O1.95 sintered under air are comprised of charge transfer bands (Fig.
66). The increase of sintering temperature improves spectra resolution without any shift of the
absorption edge position.
24
12500
20
16 13700 3
2
F(R) 12 1
8
0
10000 20000 30000 40000 50000
Wavenumber, cm-1
Figure 65. UV-vis spectra of LSFN0.1 (1) and LSFN0.2 (2, 3) sintered at 900 C (2) and 1200 oC (1, 3).
8
36600
7 29500
6 28600 1
5 2
F(R)
4
3
2
1
0
10000 20000 30000 40000 50000
Wavenumber, cm-1
Figure 66. UV-vis spectra of Ce0.9Gd0.1O1.95 sintered at 500C (1) and 1200C (2).
In the spectra of LSFN0.3 + GDC composite sintered at 700-900 C (Fig. 67) the bands in
the UV range are due to GDC. A poor resolution of these bands suggests disordering of GDC
lattice due to smaller domain size and incorporation of La and Sr cations from the perovskite
domains. The most striking feature is a strong red shift of bands corresponding to d-d
transition in Ni3+ cations (Figs. 67, 68) even for samples calcined at moderate temperatures.
This suggests considerable weakening of Ni-O bond due to disordering of LSFN structure as
well. At high sintering temperatures (Fig. 68) a shift of GDC absorption observed as well
indicating a strong disordering of GDC surface layer due to incorporation of transition metal
cations along with La and Sr .
8
29700 35200
7
6 1
5
F(R)
4 2
3
2
1
0
10000 20000 30000 40000 50000
Wavenumber, cm-1
Figure 67. UV-Vis spectra of La0.8Sr0.2Fe0.7Ni0.3O3-x+Ce0.9Gd0.1O1.95 composite sintered at 700C (1) and
900C (2).
8 29600 30000
28000
7
6 20000
3
5
F(R)
4
3 2
2 1
1
0
10000 20000 30000 40000 50000
Wavenumber, cm-1
Figure 68. UV-vis spectra of LSFN0.2+GDC (1) and LSFN0.1+GDC (2, 3) composites sintered at 900C
(1, 2) and 1200C (3).
3.3. Surface Composition of Composites
LSM-ScCeSZ. Mn 2p spectra (Fig. 69) can be decomposed into two doublets with
Mn2p3/2 BE ~642 eV and ~645 eV (20%). Their position and relative intensity are
independent upon the sintering temperature, so they can be tentatively assigned to co-existing
Mn3+ and Mn4+ cations in the LSM lattice and/or on the surface of fluorite-like doped zirconia
phase. Indeed, according to Choudhary et al [142], Mn4+ cations incorporated into the surface
layer of cubic ZrO2 are characterized by Mn2p3/2 BE 642.3 eV, hence, it is not possible to
differentiate location of Mn cations in this system. For Sr and La cations (Fig. 69, 70), both
position and intensity of components at a higher BE vary with Tsint.. This suggests
redistribution of these cations between sites with different coordination environment in
coexisting P and F phases. An apparent increase with Tsint. of high-energy component of Sr 3d
peak (~ 135 eV) (Fig. 69b) and high-energy shift of La 3d peak from 835.4 to 836.4 eV (Fig.
70a) suggest progressive incorporation of these cations into F phase in agreement with XRD
data (Table 9). Appearance of high-charge Zr4+ cations as neighbors and change of the
average coordination number (CN) by oxygen from 12 in perovskite to 7-8 in fluorite
decrease the electron density on Sr and La cations, and, hence, increase their effective charge
and BE, respectively.
Sr3d
133.1 135.3
1100 C
900 C 135.8
135.0
700 C
132.8
132 136 140

EB, eV EB, eV
a b
Figure 69. XPS spectra of Mn (a) and Sr(b) cations in the surface layer of LSM-ScCeSZ nanocomposite
sintered at different temperatures.
In Zr 3d spectra, a low- energy component at ~ 182.6 eV corresponds to Zr cations in the

surface layer of cubic zirconia [142], while the high energy component at ~ 184 eV should
be assigned to Zr cations with a higher effective charge. Hence, the increase of the relative
intensity of high BE component of Zr 3d spectra with Tsint. (Fig. 70) can be explained by
progressive transfer of Zr cations from ScCeSZ (CN 8) into B positions of LSM (CN 6).
A much larger content of La and Mn as compared with Zr within the probing depth of
XPS ~ 20 (Fig. 71a) reveals preferential presence of the perovskite phase in the surface
layers of nanocomposite particles, especially those sintered at high temperatures. As the
result, mainly Zr cations incorporated into the LSM surface layer and characterized by the
high BE component are detected by XPS in the nanocomposite annealed at 1100 oC (Fig. 70).
La3d
836.4
834.6
1100 C
835.9
900 C
835.4
700 C
834.2
830 835 840 845 b

a EB, eV
Figure 70. XPS spectra of La (a) and Zr (b) in the surface layer of LSM-ScCeZs nanocomposite
sintered at different temperatures.
Figure 71. Sintering temperature effect on the concentration of elements in the surface layer of LSM-
ScCeSZ (a) and LSFNi0.4-GDC (b) composites.
LSFN-GDC. For better presentation all spectra for this composite are related to the same
area of integral spectrum Ce3+ and Ce4+. Calibration of spectra by binding energy was carried
out by assigning the same binding energy 916.7 eV for the 3d1/2 Ce4+ component.
The shake-down processes in the course of photoemission from Ce4+ cations determine
the shape of spectrum as well-resolved multiple structure (Fig. 72). For composites annealed
at different temperatures, the gaps between neighboring peaks somewhat differ, which is
caused by the varying content of Ce3+ cations [143]. A slight difference observed between
spectra of composite sintered at increasing temperatures indicates that this content is rather
small to cause any substantial effect.
Ce3d La3d
834.3
916.7
833.9
o
1100 C
900C
700C
o
1100 C
o
900 C
o
700 C
880 890 900 910 920 930 830 840 850 860
Binding Energy (eV)
Figure 72. XPS spectra of Ce3d and La3d regions for LSFN-GDC composite sintered at 700-1100 oC.
Sr3d Gd4d
133.8 142.0
133.5
1100C
o
132.9 900 C
o
1100 C
o
700 C
128 131 134 137 140 135 140 145 150 155
Binding Energy (eV)
Figure 73. XPS spectra of Sr 3d Gd 4d regions for LSFN-GDC composite sintered at 700-1100 oC.
Spectral data for La 3d (Fig. 72) demonstrate small shifts and broadening of peaks as well
as increase of La/Ce ratio with the sintering temperature. By analogy with LSM-ScCeSZ
system this suggests transfer of La from the perovskite into the surface layers of GDC
particles, which agrees with XRD and EDX data.
As judged by the spectra of Gd4d (Fig. 73), both the chemical state of this element as
well as the Gd/Ce ratio are not affected by the temperature of composite sintering. This
suggests small (if any) incorporation of Gd into the perovskite phase, which agrees with well-
known higher stability of La cations in the mixed LnO3 layers of the perovskite structure
[101].
Sr 3d peaks increase by intensity and shift to higher energy with the sintering temperature
(Fig. 73) following similar trend for LSM-ScCeSZ composite (Fig. 69). Based upon EDX
results (vide supra), this feature can be assigned to incorporation of Sr cations into the surface
layer of fluorite particles.
O1s Fe2p
529.3
710.5
531.6
529.1
531.5
o
710.2
1100 C
o
900 C
o
700 C
525 530 535 710 720 730

Binding Energy (eV)
Figure 74. XPS spectra of O1s and Fe2p regions for LSFN-GDC composite sintered at 700-1100 oC.
In Fe2p spectra (Fig. 74), peaks position and their shape are not affected by sintering
temperature. The increase of the spectra intensity with sintering temperature suggests
segregation of Fe cations in the surface with formation of new oxide phase. Concomitant
increase of the intensity of O1s line with EB ~531.5 eV implies that this new phase is enriched
by Fe and Sr. Based upon EDX results (vide supra), such a new phase can be assigned to the
surface layers of some GDC particles with incorporated La, Sr and Fe cations. In Sr-
containing perovskites (SrFeO3, LSCF etc), a part of Fe cations is known to be in the 4+ state
[144]. At least, position of Fe 2p3/2 peak at ~ 710 eV does not exclude coexistence of Fe3+ and
Fe4+ cations in LSFN sample [144], though more detailed studies are further required.
Resulting variation of the surface concentration of elements in LSFN-GDC composite
with sintering temperature is given in Fig. 71. In general, for both composites effects are
rather weak agreeing with reasonable stability of perovskite and fluorite phases and their
moderate interaction in studied temperature range, though exceeding the relative error of XPS
analysis of the surface concentration ~ 10 rel. %. In all cases a clear trend in increasing the
surface content of transition metal cations and La with sintering temperature was observed.
This agrees with EDX data thus demonstrating that certainly transition metal cations along
with La are transferred into the surface layers of F phases. Since composition of the surface
layer of composite particles comprised of perovskite and fluorite domains reflects processes
of elements redistribution between this domains, it is apparent that similar (and even stronger)
redistribution takes places at P-F interfaces in the bulk of composite particles.
2.0 1.5 1300 C

1.5 1100 C
1100 1.0 700 C
1.0
log(, S/cm)
log (, S/cm)
0.5 900 0.5

0.0
-0.5 0.0
700
-1.0
-0.5
-1.5
-2.0 1.0 1.5 2.0 2.5 3.0 3.5
-1
1.0 1.5 2.0 2.5 3.0 3.5 1000/T, K
1000/T, K-1
b
a
Figure 75. Temperature dependences of conductivity of LSM-GDC (a) and LSFN0.4 GDC (b)
composites sintered at different temperatures.
3.4. Electronic and Oxygen Transport in Nanocomposites
3.4.1. Electronic Conductivity

Total conductivity of nanocomposites is mainly determined by electronic conductivity of
P phase and usually increases with the sintering temperature (Fig. 75) due to densification of
samples that favors percolation and improves contacts between domains provided isolating
pyrochore phase is not formed [125, 145]. For LSFN-based composites, the total conductivity
tends to increase with the Ni content in perovskite (Fig. 76), though for pure perovskites this
trend is not so straightforward (Fig. 21) due to NiO segregation at a high Ni content (vide
supra). Hence, incorporation of La, Sr, Ni and Fe cations into the surface layers of GDC
generating layers of mixed cerates with a high MIEC conductivity seems to prevent
segregation of NiO as a low conducting inter-granular phase. In general, conductivity of even
dense nanocomposite remains several times lower than conductivity of pure P phase (Fig. 77)
both due to the effect of dilution with F phase possessing a lower specific conductivity as well
as due to incorporation of Zr, Ce or Sc cations into the P lattice. For LSM-ScCeSZ (Fig. 75),
LS0.2FC0.2 -GDC and LSFN0.4-GDC composites sintered at optimal (1100 -1200 oC)
temperatures (Table 10), the total specific conductivity is reasonably high for their application
as IT SOFC cathode materials, which agrees with results of Kharton et al [127, 129]. A higher
specific conductivity in air (lg ~ 2.4 at 1000 K) was measured for LS0.4FC-GDC (30%)
composite sintered at 1250 oC [61] (Table 10), which can be explained by a higher Sr and
lower GDC content. Substantially lower (lg ~ -21 at 1000 K) specific conductivities were
found for GdMnO3-GDC and La(Pr,Nd)MnO3-Ce(Pr,Nd,Sm)O2 composites [11, 133, 134]
and Ce0.8Gd0.2O1.9 - Gd0.7Ca0.3CoO3+ composites [128]. For LSFN-GDC and LSM-ScCeSZ
composites the activation energy of conductivity at low-temperatures is slightly higher than in
the corresponding pure perovskites. It may be caused by influence of the intergrain barriers or
by decrease in the Ni or Mn content in P due to redistribution of the elements between the
phases in nanocomposites. For LSNF-ScCeSZ nanocomposite sintering at high (1330 oC)
temperature, formation of isolating pyrochlore phase (vide supra) is reflected in a drastic
decline of conductivity (Fig. 78).
LSFNi0.4+GDC
1.6
1.2
log(, S/cm)
0.8 LSFNi0.3+GDC
0.4
LSFNi0.2+GDC
0.0
LSFNi0.1+GDC
-0.4
1.0 1.5 2.0 2.5 3.0 3.5
1000/T, K-1
Figure 76. Temperature dependence of conductivity for La0.8Sr0.2Fe1-xNixO3 -GDC composites sintered
at 1330 oC.
2
LSFN0.2
log (, S/cm)
1
LSFN0.1
0
LSFN0.1+GDC
LSFN0.1+GDC
-1
1.0 1.5 2.0 2.5 3.0 3.5
-1
1000/T, K
Figure 77. Temperature dependences of conductivity of composites La0.8Sr0.2Fe0.8NixO3- GDC

(LSFNxGDC, x = 0.1, 0.2) sintered at 1300 C in comparison with corresponding pure LSFNix
perovskites.
Table 10. Transport properties of perovskite and nanocomposite materials
Sample/* lgi XSii lg(R)iii O2iv Wmax O2v Wmax. CH4 TPRvi
Tc (C) (S/cm) (ML) (O2/m2s) (ML) CO CO2
Comp1/700 15 0.3 0.4 0.5 0.7
Comp1/900 0.9 44 16.7 1.4 1.5 1.8 1.5
Comp1/1200 1.9 100 17.9 15 14 14 9.3
Comp2/700 0 20 0.1 0.1 0.5 0.1
Comp2/900 0.8 43 17.2 0.7 0.9 1.0 0.3
Comp2/1100 1.5 175 17.6 3.9 3.6 3.8 1.6
Comp3/1250 2.4 450 17.8 17 17 5 6
* Comp1-LSFN0.4 GDC; Comp2- LSFC0.2 GDC; Comp3- LS0.4FC GDC
i
-specific conductivity at 1000K;
ii
-dynamic degree of exchange XS (monolayers);
iii
- the rate of oxygen isotope exchange at 650C;
iv
-amount of oxygen desorbed in TPD run (monolayers);
v
- rate of O2 desorption at Tmax of TPD peak (1015 molecules O2/m2s);
vi
-rate of CO or CO2 evolution at Tmax of CH4 TPR peaks (1017 atoms O/m2s).
-1
-2
log (, S/cm)
-3
-4
1.0 1.2 1.4 1.6 1.8 2.0 2.2

1000/T, K-1
Figure 78. Temperature dependence of conductivity for LSFNi0.2-ScCeSZ composite sintered at 1330
o
C.
The dependences of conductivity on the oxygen partial pressure for composites

La0.8Sr0.2Fe0.8Ni0.2O3- - 50 wt.% Ce0.9Gd0.1O1.95 and La0.8Sr0.2Fe0.9Ni0.1O3- - 50 wt.%
Ce0.9Gd0.1O1.95 sintered at 1330 C in comparison with the conductivities of pure perovskite
phase are shown in Fig. 79. The slope dlog()/dlog(PO2) ~ 0.03 is very low, that corresponds
to the Region II on a general pressure dependence (vide supra Fig. 1) where reaction (7)
] = [ p ] , i.e. strontium behaves like acceptor impurity.
prevails and [ SrLa

LSFNi0.2
2.4
LSFNi0.1
2.2
log( / S.cm-1)
2.0
1.8
1.6
LSFNi0.2 + GDC
1.4
LSFNi0.1 + GDC
1.2
-2.0 -1.5 -1.0 -0.5 0.0
log(PO2/ atm)
Figure 79. Dependences of electrical conductivity of La0.8Sr0.2Fe0.8Ni0.2O3- and La0.8Sr0.2Fe0.9Ni0.1O3- s

well as the composites of the same compounds with 50 wt.% of Ce0.9Gd0.1O1.95 on the oxygen pressure
at 800C.
1.0 1.0
18
0.8 O2 16 0.8
O2
Mole fraction
0.6 0.6
z,
0.4 0.4
z
16 18
0.2 O O 0.2
0.0 0.0
0 10 20 30 0 10 20 30
Time, min
Time, min
Figure 80. Time dependence of concentration of isotope molecules (left) and isotope variables (right) in
the isothermal experiments (650 oC, 1.7 Torr) for 70%La0.8Sr0.2Fe0.8Co0.2Ox+ 30%Ce0.9Gd0.1Ox
composite sintered at 900 oC.
3.4.2. Oxygen Isotope Exchange

Figs. 80 and 81 show typical variation of the fractions of oxygen isotope molecules and
respective isotope variables in the course of isothermal (Fig. 80) or temperature-programmed
(Fig. 81) oxygen isotope exchange for LSFC-GDC composite. In both cases, the fraction of
16 18
O O molecules is quite low which is typical for samples with a high lattice oxygen
mobility (cf. Fig. 30). Estimation of the type of exchange using results of isothermal
experiments and TPIE experiments fitted in coordinates of isotope-mechanistic Equation (35)
(Fig. 82) revealed that for all composites a share of the 3rd type of exchange is in the range of
0.75 0.05. Since for manganites the share of the second type of exchange can be up to 0.6
(vide supra), a big share of the 3rd type of exchange for LSM-ScCeSZ composite apparently
correlates with much higher lattice oxygen mobility (Fig. 83), in agreement with Eq. (42)-
(46).
18 16
1.0 O2 O2 1.0

0.8 0.8
Mole fraction
0.6 0.6
z,
0.4 0.4
16 18
z
0.2 0.2
O O
0.0 0.0
300 400 500 600 700 300 400 500 600 700
o o
Temperature, C Temperature, C
Figure 81. Temperature dependence of isotope molecules fraction (left) and isotope variables (right) in
the course of temperature-programmed isotope exchange experiment for 70%La0.8Sr0.2Fe0.8Co0.2Ox+
30%Ce0.9Gd0.1Ox composite sintered at 900 oC. Temperature ramp 5o/min from 100 to 750 oC, pO2 1.7
Torr.
Figure 82. Estimation of the oxygen exchange type for LaSMn-ScCeSZ composite sintered at 1100 oC
from results of TPIE experiments using Eq. (35).
Figure 83. Temperature dependence of XS for LSM-ScCeSZ composites sintered at 700o -1100 oC. PO2
1.8 Torr.
Figure 84. The dynamic degree of oxygen isotope exchange (Xs) for: 1 Ce0.9Gd0.1O1.95, 2
La0.8Sr0.2Fe0.6Ni0.4O3, 3 composite 50% La0.8Sr0.2Fe0.6Ni0.4O3+ 50% Ce0.9Gd0.1O1.95 calcined at different
temperatures.
For doped ceria the noticeable exchange in the dynamic mode starts from ~ 600 C (Fig.
84) which is apparently explained by a lower specific rate of the oxygen heteroexchange [63].
For this oxide, high-temperature sintering apparently facilitates lattice oxygen mobility due to
annealing of extended defects (domain boundaries) and more uniform distribution of Gd
cations in the lattice increasing concentration of free anion vacancies [63, 146]. For doped
ceria, the 3rd type of exchange (share ~ 90%) dominates as well [63].
For LSNF-GDC composite samples, exchange starts at nearly the same temperatures as
for LSNF (Fig. 84). This suggests close specific rates of oxygen heteroexchange for LSNF
and composite samples sintered at temperatures 900 C, when interaction between
perovskite and fluorite components is pronounced. Already at temperatures 550 C, the
dynamic degree of exchange for composite sintered at 1200 C (> 10 monolayers, Fig. 84,
Table 10) considerably exceeds the additive value of individual components contributions.
This result suggests very high oxygen mobility in the bulk of composite particles. It can be
explained only by pronounced role played by interfaces between the fluorite and perovskite
domains providing fast diffusion paths. Indeed, the high-temperature sintering of composite
into nearly dense ceramics provides a plenty of such interfaces, since domain sizes of both
perovskite and fluorite particles remain in the nano-scale (Table 9). The values of XS obtained
for LSFN-GDC, LSFC-GDC and LSM-ScCeSZ composite sintered at high temperatures
(Figs. 83-84, Table 10) far exceed values for earlier studied GdMnO3 GDC or Ln-Mn
perovskite fluorite nanocomposites (below 20 monolayers) [11], while being close to those
for typical mixed ionic electronic conductor Sr-Fe-Co-O (>40-100 monolayers) [60]. As
shown in Fig. 85, for LSFC-GDC composite with a high lattice oxygen mobility practically
all oxygen in the solid is equilibrated with the gas phase oxygen at rather moderate
temperatures (XV~ 1).
1.2 700
900
1.0
1100
0.8
XV
0.6
0.4
0.2
0.0
100 200 300 400 500 600
o
Temperature, C
Figure 85. Temperature dependence of XV for LS0.2FC-GDC composite sintered at 700o -1100 oC. PO2
1.8 Torr.
Figure 86. Temperature dependence of the specific rate of oxygen heteroexchange for samples of
LSMn-ScCeSZ nanocomposite sintered at 700-1100 oC.
lg R [O2/m2s], 4 Torr
17.5 16
2 1
1100
3
lg R, molec. O2/m2s
17.0
900
16.5 15
700
16.0
1.0 1.1 1.2 1.3 1.4
1000/T, K-1
14
1.0 1.2 1.4
1000/ T, K-1
Figure 87. Temperature dependence of the specific rate of oxygen heteroexchange for samples of
LSFNi0.4-GDC nanocomposite sintered at 700-1100 oC (a) and fluorite-like oxide electrolytes (b): Sm
(20%)-doped ceria (1), 2- ScCeZr, 3-ScZr.
The specific rate of oxygen heteroexchange R characterizing ability of surface sites to

dissociate O2 molecules also increases with the sintering temperature (Figs. 86-87, Table 10)
achieving a level exceeding that of pure perovskites (cf. Fig. 27, 28). This occurs despite
dilution of perovskite with electrolyte possessing a lower specific rate of exchange (Fig. 87).
This can be explained with a due regard for EDX and XPS data demonstrating pronounced
redistribution of transition metal cations between perovskite and fluorite domains. As the
result, new active surface sites are created due to incorporation of transition metal cations in
the surface of fluorite domains. This is reflected in the increase of apparent activation energy
of exchange with the composite sintering temperature.
To estimate the oxygen diffusion coefficient from the results of oxygen isotope exchange
experiments for powders, the isothermal isotope exchange curves were fitted using a
complete system of isotope-kinetic and diffusion equations (30-34). In modeling, the values
of diffusion coefficients (coefficients of diffusion relaxation rD = D/h2) and the share of the
3rd type of exchange 3 = K3 / (K2 + K3) were varied. The (t) curves were found to be not
sensitive to any variation of the fitting parameters, which agrees with the exponential-type
kinetics of the exchange of equivalent oxygen atoms (Fig. 80). The second isotope variable x1
(fraction of 16O18O molecules) is much more sensitive to the modeling parameters increasing
more steeply with a decrease in the oxygen diffusion coefficient and a share of the 3rd type of
exchange. The minimum values of k3 thus estimated are in the range of 0.75-0.9.
For composite LS0.4FC-GDC prepared in Imperial College and characterized in details by
variety of techniques [39, 40, 118, 119, 135], coefficients of oxygen self-diffusion (Dmin)
estimated by fitting the results of isothermal isotope exchange for a powder with the
characteristic size ~3*10-4 cm are compared in Fig. 88 with the results obtained by isotope
exchange depth profiling (IEDP) technique for the same composite as a dense material (D*)
[119]. With a due regard for the inherent problems of fitting the results of isotope exchange
for powders, the agreement between both methods is quite satisfactory.
-9
-10
D, cm2/s
-11
1
-12
2
-13
1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7

1000/T, K-1
Figure 88. Comparison between coefficients of oxygen self-diffusion in LS0.4FC-GDC composite

estimated for powders (Dmin) (1) and extrapolated values of D* for dense material (2).

-6.0
-6.4 powder
logk(cm/s)
-6.8
-7.2
dense
-7.6
1.0 1.1 1.2 1.3 1.4 1.5 1.6
1000/T
Figure 89. Comparison between coefficients of oxygen exchange in LS0.4FC-GDC composite estimated
for powdered and dense material.
Another important parameter estimated by IEDP -the surface exchange coefficient k is

related to the rate of heteroexchange R (estimated for powders) by the relationship k =
6*R/nO*VX derived by transformation of diffusion Equation (32) to parameters used in [119].
Here, nO is the number of oxygen atoms per unit volume (3.6*1022 at/cm3), VX the fraction of
easily exchangeable oxygen in the bulk (an averaged value in these experiments is equal to
0.15). Comparison of k and k* values related to the same oxygen pressure (~ 1 bar) as used in
IEDP SIMS studies [40, 119] revealed that k values estimated for powder exceed the k*
values (estimated by IEDP ) across the whole range of temperatures studied, being close only
at the highest temperatures. This can be explained by the much higher sensitivity of the
isotope exchange experiments on powders carried out at low pressures to the highly active
surface sites which occupy perhaps only a small part of the surface due to their association
with defects either generated by grinding or situated in the vicinity of the perovskite-fluorite
interfaces, domain boundaries etc. Due to a lower oxygen bonding strength, these defects
could play a prominent role at lower temperatures, while their impact is not substantial at high
temperatures. This would be consistent with a lower activation energy of heteroexchange (~
80 kJ/mol) estimated for this composite as a powder (by using results of isothermal
experiments as well as temperature programmed exchange data analyzed by Eq. 36)
compared to the activation energy of the surface exchange coefficient obtained by IEDP (~
130 kJ/mol) [119].
3.4.3 Oxygen Temperature-Programmed Desorption (TPD) From Powders

O2-TPD spectra for LSFNx-GDC composites show that the increase of Ni content in
LSFNx shifts beginning of desorption to lower temperatures and increases the rate of oxygen
desorption (Fig. 90). Since the amount of oxygen desorbed in TPD run exceeds several
monolayers (Table 10), the oxygen release from the sample caused by reduction of Ni3+ to
Ni2+ is apparently controlled by the bulk diffusion. Direct estimation of heat of oxygen
adsorption on perovskite or LSFN0.3-GDC composite after high-temperature oxygen
desorption by using microcalorimetric installation (Fig. 91) confirmed that indeed it varies in
the range of 120-260 kJ/mole O2 corresponding to terminal M-O forms including those fixed
at defect centers (vide supra). Indeed, desorption of oxygen from LSFN perovskite in a broad
temperature range (Fig. 35) suggests existence of surface sites with broadly varying oxygen
bonding strength. Identical dependence of adsorption heat on the amount of desorbed oxygen
for perovskite and perovskite-fluorite composite, despite twice as low content of perovskite in
composite, demonstrates very fast redistribution of oxygen between perovskite and fluorite
phases.

W*10-14 , molecules O2 /m2s
150 LSFNi0.4+GDC
100
50 LSFNi0.3+GDC
LSFNi0.2+GDC
0 LSFNi0.1+GDC Isotherm 70 min
400 800 880 880
Temperature, oC
Figure 90. Effect of Ni content in LSFN perovskite on the O2 TPD spectra for LSFN-GDC composites
sintered at 1200 oC.
300 LSFN0.3
250
- H (kJ/mol O2)
200
150
LSFN0.3 +GDC
100
50
0
0.0 0.2 0.4 0.6 0.8 1.0
N, monolayers
Figure 91. Heat of oxygen adsorption vs. amount of desorbed O2 for LSFN0.3 and LSFN0.3 + GDC
composite sintered at 1200 oC and pretreated in He at 750oC. Pulses of O2 in He at 450 oC.
The increase of the peak rate of oxygen desorption from composites with the sintering
temperature (Fig. 92) apparently correlates with the porosity annealing thus increasing P-F
interface. This suggests that P-F interface indeed provides a path for fast oxygen migration in
the bulk and surface sites for reversible dissociation/recombination of oxygen
molecules/atoms. In this case, a higher Ni content in perovskite provides a higher rate of
oxygen desorption (Fig. 92), so for the air side of membrane composites with a higher Ni
content could be more efficient in the oxygen activation. Hence, sintering temperature is to
be optimized if porous composite materials are used in design of functionally graded SOFC
cathodes.
150 x=0.4
W*10 ,molecules O2/m s
2
100
x=0.3
50 x=0.2
-14
x=0.1
0
800 1000 1200
o
Calcination Temperature, C
Figure 92. Temperature dependence of the maximum rate of O2 desorption for LSFNx-GDC
composites.
The same trend of lattice oxygen mobility increasing with sintering temperature was
observed for other nanocomposites (LSCF-GDC, LSM-ScCeSZ) both by oxygen isotope
exchange and O2 TPD (Fig. 93, Table 10). This suggests that P-F interface could indeed
provide a path for fast oxygen migration. The maximum rates of oxygen desorption are
comparable for composites of GDC with LSFC and LSFN (Table 10), while being several
times lower for LSM-ScCeSZ composites (Fig. 93). This certainly correlates with a higher
lattice oxygen mobility in the former composites estimated by the oxygen heteroexchange
(vide supra).
0.05
WO2, 1017 molec. O2 m2s
0.04
/
0.03
0.02 1100
0.01 900
700
0.00 Isotherm
70 min
0 400 800880 880
T, OC
Figure 93. Temperature dependence of specific rates of O2 TPD for LSM+ScCeSZ nanocomposite
sintered at 700-1100 oC.
To verify whether the increase of the specific rate of oxygen desorption with the sintering
temperature of nanocomposites can be indeed explained by increasing the oxygen bulk
mobility, modeling of oxygen temperature-programmed desorption from powdered samples
has been carried out. The model of oxygen temperature-programmed desorption accompanied
by the oxygen diffusion from the bulk of oxide particles to the surface is represented by the
system of differential equations:
1 dCO2 1 gSN OS
= CO2 + k DESO2S (47)
dT NG
1 d OS N O V OBULK
= 2k DESO2S + BULK D (48)
dT N OS S h =0
OBULK 2OBULK
=D (49)
t h 2
Here is the heating rate (grad/s); NG the total number of gas molecules in the volume
of a flow reactor, CO2 is the concentrations of gen in the gas phase; T is the reactor
temperature; (s) is the contact time (a ratio between the reactor volume, VR, m3, and the
feed rate of inert sweeping gas, VHe, m3/s); g weight of sample, S specific surface area
(m2/g); V volume of sample (m3); N OS and N OBULK - the maximum oxygen capacity per the
unit of the surface and the bulk, respectively; k DES is the rate constant of desorption; D is
the diffusion coefficient of oxygen in the oxide bulk (m2s-1) ; h is the characteristic size of
oxide particle (m).
Calculations carried out within the frames of this model using some typical values of kDES
known for transition metal oxides and considering surface and bulk oxygen atoms to be
uniform by bonding strength and mobility revealed that:
In the absence of oxygen diffusion from the bulk of oxide particles, the maximum
concentration of oxygen in the gas phase monitored continuously decreases
proportionally to the specific surface area provided the surface adsorption capacity
remains constant (Fig. 94 a), while the position of desorption peak remains the same.
This provides independence of the maximum specific rate of desorption defined for
the flow reactor as the product of the CO2. and the flow rate of He related to the oxide
surface unit.
The oxygen desorption from the surface accompanied by a slow oxygen diffusion
from the bulk of oxide particles is characterized by two desorption peaks situated at
low temperatures (corresponds to oxygen desorption from the surface) and at high
temperatures (corresponds to oxygen desorption from the bulk). With the increase of
the bulk oxygen mobility in the oxide approaching the rate of diffusion to that of
desorption from the surface (or even exceeding it), position of the high-temperature
peak shifts to lower temperatures superimposing with the low-temperature peak (Fig.
94b). With decreasing specific surface area position of this peak shifts to higher
temperatures due to decreasing the efficient diffusion coefficient D/h2 as well as
decline of the oxygen flux at gas-solid interface determined by expression

N OBULKVCAT OBULK
D .
N OS S h =0
In this case, the maximum oxygen content in the gas phase decreases rather slightly (Fig.
94b) which corresponds to some increase of the maximum specific rate of desorption. Hence,
identical position of oxygen desorption peak for samples with broadly varying specific
surface areas (particle sizes) observed in the case of perovskite-fluorite composites calcined at
different temperatures (Figs. 92, 93) are expected only in the case when both oxygen bulk
diffusion coefficients and specific constants of oxygen desorption increase with the particle
size (sintering temperature). This apparently agrees with the data of oxygen isotope
heteroexchange demonstrating the increase of the dynamic degree of exchange XS (Figs. 83-
85) and specific rates of heteroexchange (Figs. 86, 87) with the nanocomposites sintering
temperature.
a 1
0.30
b
0.020 1
0.25 2
0.015 0.20
O2
O2
0.15
C
0.010
0.10
0.005
2 0.05
0.000 0.00
200 300 400 500 600 700 200 300 400 o 500 600
o TC
T C
Figure 94. Oxygen TPD spectra calculated for the case of very slow (a) and fast (b) oxygen chemical
diffusion in the bulk for samples with specific surface area 10 m2/g (1) and 1 m2/g (2).
3.4.4. Dynamics of Oxygen Desorption from Dense Samples

Gravimetric studies. Typical weight loss curves for the pellet of LS0.4FC-GDC composite
after step-wise changing the oxygen content in the feed are shown in Fig. 95. As judged by
the whole weight loss (Fig. 96), only a small part of oxygen (~1-3%) is removed from sample
in this temperature range, which agrees with O2 TPD data for powder (Table 10). Kinetic
curve (Fig. 97) is comprised of two characteristic parts intersecting approximately at the point
of half conversion. The initial stage is described by the dependence proportional to the square
root of time, while the final stage is described by the exponential time dependence. This
picture qualitatively agrees with the model of pure diffusion controlled kinetics of oxygen
removal from the pellet of a finite size. The initial stage corresponds to the movement of
diffusion profiles from the two sides of pellet until they meet in the middle. At this stage,
kinetics is described by the equation
Dchemt
M = 4M max (50)
d2
where M is the current change of mass, M max - the change of mass at complete
conversion (desorption), D chem the chemical diffusion coefficient, d - pellet thickness.
This equation is valid for the plate with infinite thickness at Dt<<d2 but approximation is
satisfactory up to 30% conversion. At the final stage, kinetics is determined by the
exponential decay of the most long-wave Fourier component of the concentration profile
(with wavelength 2d). Respective temperature dependence is given by the equation
2 D ( t t0 )
M = M max 1 exp (51)
d2

where t0 - an efficient time of start-up of kinetics of this type provided it operates from the
very beginning and is responsible for the whole mass loss M max . Fitting curves
corresponding to equations 50 and 51 are shown in Fig. 97. Diffusion coefficients estimated
from these two parts 1 and 2 were found to be close to each other and equal to 3*10-5 and
2*10-5 cm2/s, resp.
Figure 95. Weight loss curves for the pellet of LS0.4FC-GDC composite after switching the stream of
14%2+53%N2+ 33%He to 1.4%2+5.3%N2 +93.3% He.
0.16
0.14
0.12
Mmax , % 0.10
0.08
0.06
0.04
0.02
0.00
650 700 750 800 850 900
o
T, C
Figure 96. Temperature dependence of the maximum weight loss for gravimetric experiments of Fig.
95.
99.90
-5 2
D~3*10 cm /s
99.85
99.80
Mass change, %
99.75
99.70
99.65
-5 2
D~2*10 cm /s
99.60
99.55
99.50
-20 0 20 40 60 80 100 120 1
Time, s
Figure 97. Typical results of fitting the weight loss curve for the pellet of LS0.4FC-GDC composite at
900 oC by Equations (47-48).

-4.5
-4.7
lg(D) (cm2/s)
-4.9
-5.1
-5.3
0.85 0.90 0.95 1.00 1.05 1.10
1000/T, K-1
Figure 98. Temperature dependence of the oxygen chemical diffusion coefficient for the pellet of
LS0.4FC-GDC composite.
The temperature dependence of oxygen chemical diffusion coefficients for this composite
is shown in Fig. 98. At 1073 K, the value of Dchem.~ 10-5 cm2/s and activation energy ~ 110
kJ/mol are identical to those estimated for LS0.4FC perovskite by the weight relaxation
method [41] and conductivity relaxation [39]. At lower temperatures, lower (~ 30 kJ/mol)
activation energies of oxygen chemical diffusion provides higher values of Dchem. for the
composite found in this work (Fig. 98) as compared with pure LS0.4FC [39, 44]. This effect
can be explained by impact of perovskite-fluorite interfaces as paths for fast oxygen diffusion.
Indeed, this agrees with a small amount of oxygen desorbed in this temperature range (Fig.
96), which apparently differs by migration and/or bonding energy with majority of the lattice
oxygen. Note that in the same intermediate temperature range low activation energies of
oxygen self-diffusion were estimated for powdered composite by analysis of oxygen isotope
exchange data (Fig. 88). From the structural point of view, diffusion of La and transition
metal cations onto the surface of GDC domains revealed by TEM with EDX and XPS studies
(vide supra) favours the formation of surface cerate layers with high mixed ionic-conductivity
and, hence, oxygen mobility.
The values of the oxygen self-diffusion coefficient, D*, calculated from Dchem. using
known relations [39, 41, 44, 119] and taking the same value for the thermodynamic factor
=1/2(lnpO2 /lnCo) for LSCF-GDC composite as for LSCF (~ 102) [41], are shown in Fig.
99. In the high-temperature range (~ 1100 K), the values of D (~ 10 -7cm2/s) and their
activation energy (~ 110 kJ/mol) estimated by the weight relaxation method correspond well
to the values found by Esquirol et al for the same composite using IEDP-SIMS method [119].
-6
TG
-7
-8
D,cm /s
SIMS
2
-9
-10
-11
1.0 1.2 1.4 1.6
1000/T
Figure 99. Comparison of oxygen self-diffusion coefficients for the pellet of LS0.4FC-GDC composite
estimated by SIMS [119] and TG (this work).

-5.0 LSFNi0.3+GDC
lg(D),cm2/s
-5.2
-5.4
LSFNi0.4+GDC
-5.6 LSFNi0.3
0.80 0.85 0.90 0.95

-1
1000/T, K
Figure 100. Temperature dependence of chemical diffusion coefficient for dense pellets of composites
and perovskites sintered at 1330C.
Fig. 100 shows temperature dependence of chemical diffusion coefficient for dense
pellets of LSFN0.3 perovskites and LSFNx GDC composites sintered at 1330C. In this case,
higher values of Dchem. for composites as compared with those for perovskite were found in
the high-temperature range, which can be explained by a higher content of GDC (50%) in
these composites (and, hence, more developed perovskite-fluorite interface) as compared with
LSFC-GDC composite (30%). The activation energy of the oxygen chemical diffusion for
LSFNx GDC composites (~ 120-140 kJ/mol) is close to that for LSFC-GDC composite,
though values of Dchem. are somewhat lower (cf. Figs. 98 and 100). A higher value of Dchem
for composite with a lower Ni content can be due to some segregation of NiO phase from
LSFN0.4 perovskite under high-temperature sintering (vide supra).
Conductivity relaxation studies. In agreement with weight relaxation studies, electrical
conductivity relaxation method revealed that oxygen chemical diffusion coefficients in
composites are approximately an order of magnitude higher than those in pure perovskite
phases (Fig. 101). This increase of chemical oxygen diffusivity is easily explained in terms of
the effect of fast diffusion along the perovskite-fluorite interfaces. Note that the absolute
values of these coefficients are somewhat higher that the values obtained by gravimetric
studies (cf. Figs. 100 and 101). Hence, either conductivity relaxation method is more sensitive
to the loss of oxygen from the surface layers of samples, or effect of perovskite-fluorite
interfaces should be considered in frames of more detailed model considering fast transfer of
oxygen along domain boundaries with a much slow diffusion within domains [10, 147]. At
least, the difference between estimations made by different techniques could not be explained
by the pressure difference effect, since the dependence of the oxygen chemical diffusion
% on the oxygen pressure is rather moderate for both perovskites and composites
coefficient D
(Fig. 102).

-4.5
LSFN0.1-GDC
LSFN0.2-GDC
log (D/ cm2.s-1)
-5.0
-5.5 LSFN0.1
-6.0
LSFN0.2
-6.5
perovskites composites
Figure 101. Oxygen diffusion coefficient values at 800C for perovskites LSFNx and their composites
with GDC.
LSFN0.1-GDC
-4.8
log (D / cm2.s-1)
-5.2
LSFN0.1
-5.6
-2.0 -1.5 -1.0 -0.5

log(PO2/atm)
Figure 102. The dependence of chemical diffusion coefficients on the oxygen pressure for LSFN0.1 and
LSFN0.1-GDC composite an 800 oC.
3.4.5. Temperature--Programmed Reduction

H2 TPR spectra for studied composites are presented in Fig. 103-106. The spectra profiles
and peak positions depend on the composition and calcination temperature of samples. For
composites sintered at high temperatures, the low-temperature peaks are situated at lower
temperatures as compared with perovskites (Fig. 104). Furthermore, for LSFN0.3-0.4 GDC
composites sintered at 1200 , several peaks in the region 400-800 C appear (Figs. 104-
105). These features can be assigned to bulk reduction of GDC domains accelerated due to H2
activation on small Nio clusters formed by reduction of Ni cations on the GDC surface (vide
supra). The fastest reduction in the low-temperature range is observed for LSFN0.3 GDC
composite possessing the highest oxygen mobility (Fig. 105). In general, specific rates of
composites reduction increase with the sintering temperature thus correlating with the lattice
oxygen mobility (vide supra). For LSFN0.3 -GDC and LSFC -GDC nanocomposites sintered
at 1200 oC, the maximum rate of reduction by H2 exceeds at least several times that for Mn-
containing fluorite perovskite composites [11]. Hence, a high oxygen flux in composites
under high oxygen chemical potential gradient makes these materials attractive for design of
membranes for syngas generation from hydrocarbons via their selective oxidation with
oxygen transferred from the air side to the fuel side of these membranes.
CH4-TPR. Typical curves of product evolution in CH4 temperature-programmed
reduction of LSFN0.3-GDC composite calcined at 1200 are shown in Fig. 107. First CO2
appears at ~ 700 oC, then CO and H2 simultaneously evolve. Isothermal experiments (Fig.
108) revealed that CO and H2 formation is described by typical topochemical kinetics even at
700 oC, i.e. reduction of nanocomposite by methane proceeds via nucleation and growth of
reduced phases (Ni or Ni alloy particles). CO2 appears immediately after CH4 contact with the
catalyst surface evidencing a high reactivity of the surface oxygen of these composites.
6 1200oC 3.5
0
3.0 1100 C
WH2, 10-7 mol H2/m2*s
W(H2), 10 mol/m s
2
2.5
4
2.0
-7
1.5
0
900 C
1.0
2
0
0.5 700 C
900oC
0.0
0 700oC 0 200 400 600 800 1000

200 400 600 800 1000
Temperature, C
b
a
Figure 103. H2 TPR spectra of LSFNi0.3 GDC (a) and LSFNi0.4 GDC (b) calcined at different
temperatures.
1.6
W(H2), 10-6 mol/m2s

1.4 composite
1.2
1
perovskite
0.8
0.6
0.4
0.2 x5
GDC
0
200 400 600 800 1000
Temperature, oC
Figure 104. H2 TPR spectra for La0.8Sr0.2Fe0.7Ni0.3O3 (perovskite), Ce0.9Gd0.1O1.95 (GDC) and their
composite calcined at 1200 .

LSFN0.4
W(H2), 10-6 mol/m2s
2.0
1.5
LSFN0.3
1.0
LSFN0.2
0.5
LSFN0.1
0
200 400 600 800 1000
Temperature, oC
Figure 105. H2 TPR spectra for LSFNix GDC composites calcined at 1200 .
15
1200oC
13
W, 10-7mol/m2s
11
9
7
5
3 900oC
1 700oC
0 200 400 600 800 1000
o
Temperature, C
Figure 106. H2 TPR spectra for LSFC-GDC composite calcined at 700 oC, 900 oC and 1100 oC.
Figure 107. CH4 TPR spectra for LSFN0.3-GDC composite sintered at 1200 oC.
1.00 He 1%CH4 in He through reactor He 1%CH4 in He

Concentration, %
1.00 H2 CH4
CH4
0.75
Concentration, %
0.75
0.50 H2
0.50
CO
0.25 CO 0.25
CO2 CO2
0.00 0.00
10000 15000 20000 25000
Time, s 7500 10000 12500 15000 17500
Time, s
a
b
Figure 108. Dynamics of isothermal reduction of sintered at 1200 oC LSFN0.3-GDC composite in the
flow of 1% CH4 in He at 700 oC (a) and 800 oC (b).
The peak rates of CO/CO2 evolution during CH4-TPR for LSFNx - GDC composites (Fig.
109) and LSFC-GDC composites (Table 10) increase with their sintering temperature. This
certainly correlates with the increase of the lattice oxygen mobility characterized by oxygen
isotope exchange or O2 TPD (Table 10).
The dependence of peak rate of CO/CO2 evolution during CH4-TPR on Ni content in
composites calcined at 1200C is presented in Fig. 110. Increasing Ni content up to x=0.3
leads to a little increase of the peak rate values for both CO and CO2 evolution, and only at
x=0.4 they increase substantially. In all cases, CO peak rates exceed those of CO2 evolution.
Similarly, peak rates of CO2, CO and H2 formation in the course of CH4 TPR for LSM-
ScCeSZ composite sample increase with the sintering temperature (Fig. 111).
C O 2 (L S F N 0 .1 )
C O (L S F N 0 .1 )
at. O/m s
12
2
C O 2 (L S F N 0 .2 )
C O (L S F N 0 .2 )
C O 2 (L S F N 0 .3 )
C O (L S F N 0 .3 )
17
8 C O 2 (L S F N 0 .4 )
Wmax, 10
C O (L S F N 0 .4 )
0
800 1000 1200
Figure 109. Dependence of the peak rate of CO/CO2 evolution during CH4-TPR on calcination
temperature of composites 50%LSFNx+50%GDC.
Wmax, 1017 at. O/m2s

12
CO
8
4 CO2
0
0.1 0.2 0.3 0.4
xNi, at. fraction
Figure 110. Dependence of peak rate of CO/CO2 evolution during CH4-TPR on Ni content in LSFNx
composite calcined at 1200C.
100
50
LSM+ScCeSZ
,molecule/m s
2
CO2
CO
15
H2
10
-16
W*10
o
1100 C
5
o
900 C
o
700 C Isotherm
0 70 min
400 800 880 880
O
T, C
Figure 111. Typical curves of CO2, CO and H2 evolution during CH4 TPR for LaSrMn- ScCeSZ
composites calcined at 700, 900 and 1100C.
Figure 112. Schematic illustration of the SOFC test rig and the pellet assembly: cross sectional view
(a), top (b) and bottom (c) of the cell (1.3 cm2).
4. DESIGN AND TESTING OF IT SOFC CATHODES

Details of button-size cells assembling and their testing procedures are given in [148].
NiO-YSZ/YSZ(ScYSZ)/cathode fuel cells with and without GDC functional layer between
electrolyte and cathode were assembled using prefabricated anode-supported planar half cells
manufactured either in Research Center Juelich, Germany, or at the Institute of
Thermophysics SB RAS (Novosibirsk). In the last case, either YSZ or ScYSZ layers (10
microns) were supported by chemical vapor deposition (CVD).
Schematic sketch of the setup is shown in Fig. 112. Humidified (5% RH) hydrogen at
0.03-0.2 L/min and air at ambient conditions were used as a fuel and an oxidant, respectively.
Pt paste and Pt mesh from the cathode and Pt paste and Ni mesh from the anode side were
used as current collectors.
The morphological characterization and chemical composition of the tested SOFCs was
performed using SEM equipped with EDX analyzer. The SOFC samples were embedded in
epoxy resin and polished using Cr2O3. The SEM images of SOFCs cross sections were made
with a JSM-6460 LV (Jeol, Japan) microscope.
Compositions of cells are given in Table 11.
4.1. Single Layer Cathodes
As can be seen from data presented in Figures 113, 114, performance of cells with a
single layer cathode (without GDC functional layer) is low. Using LSNF+GDC composite
results in some improvement of performance (Fig. 113). However, the cell performance
decreases gradually after increase of the temperature of testing (Fig. 113).
Table 11. Some characteristics of cells
Fuel cell # Cathode layer Functional layer

1 50% GDC + 50% LSFN No
2 LSFN No
3 LSFN GDC
4 50% GDC + 50% LSFN GDC
5 LSFN; 50% GDC + 50% LSFN (two- layered cathode) No
6 LSFN; 50% GDC + 50% LSFN (two- layered cathode) GDC
7 LSCF No
8 LSCF GDC
9 LSFN; 50% GDC + 50% LSFN (two- layered cathode) GDC
10 LSFN LSM-ScCeSZ
11 LSFN LSFN-ScCeSZ
In the case of LSCF cathode, a much lower power density of the cell (Fig. 114) as
compared with that provided by LSNF cathode (Fig. 113) is obviously caused by peeling of
the cathode layer due to a higher LSCF thermal expansion coefficient (TEC), and, hence, its
bigger mismatch with YSZ. However, decline of performance of cell with LSCF cathode in
isothermal conditions clearly indicates on the pronounced chemical interaction between YSZ
electrolyte and LSCF with formation of low conducting La2Zr2O7 and SrZrO3 layers. Such an
interaction clearly takes place with LSNF cathode as well, which is confirmed by SEM/EDX
analysis of cells after testing (Fig. 115). According to this data, diffusion of Zr atoms through
the cathode layer takes place. In the case of the cell 1 with LSFN cathode, Zr atoms are
distributed throughout the cathode layer (Fig. 115).
Figure 113. (a) -voltage and power density vs. current density for SOFCs at 700 oC with single layer
cathode: LSFN, LSNF-GDC composite; (b)- performance of the cell 2 with composite cathode
at 750: in 1 h; in 2.5 h.
Besides, a new rather dense layer mainly containing Zr and La cations was formed
between the electrolyte and porous cathode layers. Being morphologically different from the
porous cathode, this new dense layer was well visible in the SEM image (Fig. 115). This new
layer can be a reason for delamination of the cathode layer during heating-cooling cycles.
However, it should be mentioned that peeling observed in the SEM images is caused by the
samples preparation including their fixing in epoxy resin.
1.0
2.0
0.8
Power Density (mW/cm )

2
1.5
0.6
E (V)
1.0
0.4
0.2 0.5
0.0 0.0
0 2 4 6 8 10
2
I (mA/cm )
Figure 114. Performance of the SOFC with LSCF as the single layer cathode at 700 C after load of 0.5
V for 1 h () and 4.5 h () at 700 C.
Figure 115. SEM micrographs for the cell 1 after testing: cross-section (a) and EDX analysis data in
different spots near the cathode-electrolyte interface (b); CC current collector, C cathode, NL new
low conducting layer, E electrolyte.
Hence, main problems for using LSFN and LSCF complex peroskites as cathode
materials for IT SOFC based on YSZ as electrolyte layer are their chemical interaction and
TEC mismatch with the electrolyte. To prevent strong chemical interaction of the cathode
material and YSZ resulting in formation of undesirable phases and/or destruction of dense
electrolyte layer, a GDC functional layer was used.
4.2. Cathodes with GDC Functional Layer
As follows from data of Fig. 116, supporting of GDC functional layer considerably
improved performance of cells with all types of cathode materials. The cell with LSFN
cathode has the best performance among tested cells. According to the data of SEM with
EDX (Fig. 117), a positive effect of GDC interlayer seems to be primary associated with
hampering of Zr cations migration into cathodes layers this suppressing formation of dense
La-Zr-O layers with a low conductivity. Another factor is in situ formation of nanocomposite
when supporting perovskite layers on porous GDC layer. This clearly helps to match the
thermal expansion coefficients of composite cathodes and YSZ, thus improving contact and
providing developed three-phase boundary required for a good performance. However,
another important factor could be a chemical interaction between LSNF and GDC
nanoparticles under load (Fig. 118) leading to increase of the power density with time. This
feature can be assigned to migration of Ni cations onto the surface of GDC and/or YSZ
creating new sites for O2 molecules dissociation as earlier revealed for nanocomposite
powders by oxygen isotope exchange studies (vide supra). A driving force for this migration
could be an inherent instability of Ni2+ cations in the cubic perovskite structure reflected also
in appearance of some traces of NiO phase in the course of LSNF-GDC nanocomposites
sintering (vide supra). From these point of view, a lower performance of cells with
LSNF+GDC nanocomposite cathode and porous GDC functional layer can be assigned to a
lower LSNF concentration at GDC interface decreasing mixed ionic-electronic conductivity
and weakening positive effects of LSNF-GDC interaction under load.
1.0 600
1.0
400
0.9
500
350 0.8
0.8
0.7
300
2
2
400
0.6 250 0.6
E (V)
E (V)
0.5 300
200
0.4 0.4
150
200
0.3
100
0.2 0.2
100
50
0.1
0.0 0 0.0 0
0 200 400 600 800 1000 0 200 400 600 800 1000 1200
2
I (mA/cm ) 2
I (mA/cm )
a b
Figure 116. Performance of cells with the GDC functional layer between the cathode and YSZ at 700
(a) and 750C (b) after 0.5 V load for 15 h at 700C: - LSFN; 50% LSFN + 50% GDC; - LSCF.
Figure 117. SEM micrographs of a cell cross-section for the cell #4 with composite LSNF+GDC
cathode after testing (a) and data of the EDX analysis in different spots near the cathode electrolyte
interface (b); CC current collector, C cathode, E electrolyte. Note that the current collector was
peeled off during the sample preparation for SEM study.
On the contrary, for LSCF cathode interaction of perovskite with GDC results in
performance decline with time-under-load (Fig. 119). Perhaps, in this case a negative effect of
time-under-load can be assigned to a higher Sr content in used sample of
La0.58Sr0.4Co0.2Fe0.8O3-, so mainly Sr migration on the surface of CDC occurs leading to
formation of surface cerate layers with a low conductivity. A small difference between the
maximum power densities at different temperatures (Fig. 119) suggests progressing transfer
of Sr with the temperature increase.
1.0 600
450
0.9
500
430 0.8
0.7
Power Density (mW/cm)

2
400
410
Power Dencity (mW/cm)
2
0.6
E (V)
390 0.5 300
0.4
370
200
0.3
350 0.2
100
330 0.1
0.0 0
310 0 200 400 600 800 1000 1200
0 2 4 6 8 10 12 14 2
I (mA/cm )
Time (h)
a b
Figure 118. Time dependence of the power density of the SOFC with the GDC functional layer between
the LSFN cathode and the electrolyte under load of 0.5 V at 700 C (a) and performance of the SOFC
with the GDC functional layer between the cathode and LSFN electrolyte at different temperatures after
0.5 V load for 15 h at 700 C (b): 700 C; - 750 C and 800 C.
240
1.0 250
230
0.8 200
2

220
2
0.6 150
E (V)
210
0.4 100
200
0.2 50
190
180 0.0 0
0 2 4 6 8 10 12 14 0 200 400 600 800
2
Time (h) I (mA/cm )
a b
Figure 119. Time dependence of the power density of the SOFC with the GDC functional layer between
the LSCF cathode and the electrolyte under load of 0.5 V at 700 C and performance of the SOFC with
the GDC functional layer between the LSCF cathode and the electrolyte after 0.5 V load for 15 h at 700
C (); 750 C () and 800 C ().
4.3. SOFCs with Thin Nanocomposite Functional Layers
In designing functionally graded cathodes, one of the most important problems is

providing a good sintering of GDC interlayer with YSZ dense layer required for ensuring a
low area specific resistance (ASR) at intermediate temperatures. Though different sintering
agents (Li salts, Co oxide etc) were suggested [149, 150], the problem is still far from being
solved.
As alternative approach, nanocomposites LSM-ScCeSZ and LSNF-ScCeSZ were
supported as thin (ca 5 microns) layers on thin dense YSZ or ScYSZ layers (Table 11).
Additional advantage of these nanocomposites is their good sinterability at reasonable low
(1000 oC) temperatures.
As follows from Figs. 120 and 121, performance of cells with these functional layers and
LSFN as thick porous cathode layers is rather good even exceeding that for cells with GDC
interlayers. For our graded design of cathodes, the maximum power density values (Figs. 120,
121) are comparable with the best published results obtained with composite cathode layers in
cells with thin YSZ electrolytes [126, 151, 152]. In all cases, performance was stable at least
for 10-30 hours. SEM studies with EDX analysis of cells discharged after tests has not
revealed formation of isolating pyrochlore layers between composite interlayers and
electrolyte, though some additional transfer of Zr cations into the interlayer takes place.
700
1.0 800C
600
0.8
Power Density, mW/cm2

700C 500
0.6 400
E, V
300
0.4
600C 200
0.2 100
0.00 500 1000 1500

0
2000
i, mA/cm2
Figure 120. Typical performance of button-type cell with LSM-ScCeSZ interlayer and LSFN0.2 cathode.

Figure 121. Typical performance of button-type cell with LSNF-ScCeSZ interlayer and LSFN0.2
cathode.
Hence, interlayers comprised of GDC or nanocomposites of perovskites with ScCeSZ

allow to provide a high and stable performance of anode-supported SOFC in the IT range.
5. DESIGN OF ASYMMETRIC MULTILAYER MEMBRANES BASED ON

MIXED IONIC-ELECTRONIC CONDUCTING COMPOSITES
It is well known that for thick dense oxygen permeable membranes the oxygen flux is
limited by the oxygen diffusion [4, 44-47]. A number of studies clearly showed the promising
features of asymmetric supported membranes with a thin dense layer for high-performance
oxygen separation [44, 153-156]. However, fabrication of such membranes is still a challenge
because of difficulty in obtaining porous supports with desirable properties in terms of pore
size, gas permeability, thermal/mechanical stability and compatibility with MIEC materials
by thermal expansion coefficients (TEC) [157, 158]. One of the promising solutions is to use
a robust macroporous substrate on which MIEC nanocomposite layers with graded porosity
and chemical composition would be consecutively deposited [157]. Combination of meso-

and micro-porous MIEC layers with a thin gas-tight layer provides a maximal oxygen flux
under mixed surface exchange kinetics/bulk diffusion control [159, 160].
In our research, LSFN-GDC nanocomposites (vide supra) were supported as layers with a
graded porosity onto macroporous metallic plates made from Ni-Al alloy compressed foam
possessing a high thermal stability and mechanical strength due to segregated corundum
surface layer [117]. From the fuel side, a dense layer of MnFe2O4 (MF) - GDC
nanocomposite possessing required phase stability and oxygen mobility in strongly reducing
conditions [161] was supported. This layer was covered by a porous layer of La-Ni-Pt/Ce-Zr-
Pr-O catalyst known to be efficient and stable to coking in the reaction of methane
transformation into syngas in stoichiometric CH4/H2O feeds [117, 162].
5.1. Preparation of Membrane
Macroporous open-cell nickel foam substrate was manufactured by the nickel

electroplating of the polyurethane foam samples (cell density 60 ppi) followed by sintering in
the dissociated ammonia atmosphere at 1100oC for 1 h. The foam samples were then
deformed by a uniaxial compression to 1 mm thickness modifying the cell morphology and
decreasing porosity from 95.5 % to 60-80 %. Deformed foams were subjected to the pack
aluminizing and then annealed at 1000 for 1 h under air to form a thin -alumina layer over
the foam cell walls/ribs for a better adhesion of composite layers.
Fabrication of a supported multilayer membrane followed the next steps. First, a disc cut
from membrane was pressed and spark-welded into the stainless steel ring to prevent its
bending under subsequent manipulation procedures. Then the cells of a foam support from
one side of a membrane were filled with macroporous layer of a coarse (fraction up to 10
microns) particles of LSFN0.3+GDC composite presintered at 1200 oC and ball milled. A
thick viscous slurry of these particles in isopropanol with addition of PVB was supported on
the substrate via slip casting. This procedure was repeated several times with intermediate
drying and calcination at 900-1000C after each supporting until required uniform filling of
foam cells for depth ~ 0.1 mm was achieved. For filling macropores between the coarse
particles of composite (thus forming a mesoporous layer with thickness 10-20 microns), a
highly dispersed LSFN0.3+GDC composite precalcined at 700C was next supported several
times from suspension prepared by ultrasonic dispersion of nanocrystalline powders in
isopropanol with addition of PVB. Next the gas-tight layer (thickness ~ 10 micron) comprised
of GDC+MnFe2O4 (MF) 1:1 nanocomposite was supported via slip casting of a slurry in
isopropanol with addition of PVB prepared using nanocrystalline MF synthesized via Pechini
route. A special smoothing procedure was applied to restrict the thickness of this layer, thus
keeping its upper surface at the outside level of the foam cell ribs. To attain complete
densification of this layer by sintering at 1200C under Ar flow, such additives as Cu, Bi or
Ag nitrates were used. The gas-tight layer was washcoated by a porous catalytic layer
comprised of Pr0.3Ce0.35Zr0.35O2-x complex oxide with La-Ni-Pt active component (~ 10 wt.%
of LaNi0.9Pt0.1O3) supported by the incipient wetness impregnation with the mixed solution of
H2PtCl6 and La and Ni nitrates followed by calcination at 1000 oC. Finally, the main bulk of
foam substrate and the surface of macropores in the coarse nanocomposite layer were covered
with a thin layer of LSFN0.4 by membrane impregnation from the air side by a suspension of
perovskite in isopropanol with addition of PVB followed by drying and calcination at 1000
o
C.
5.2. Membrane Reactor Testing
To ensure the gas-tight sealing of supported disc membrane in titanium lab-scale reactor,
it was first placed in specially designed copper washers, pre-sealed using Ag paste, pressed
and preheated under Ar flow at 900 oC to provide a hermetic junction between the membrane
steel ring and a copper washer/ring (Fig. 122a, b). Such a disc was inserted between two
halves of specially designed Ti reactor (Fig. 122c), and the copper ring was compressed by
using bolts and nuts ensuring required sealing.
Figure 122. Typical images of small-size membrane disc embedded into the copper washer (a, b) and
titanium lab-scale reactor for lab-scale membrane testing (c). a-fuel side, b-air side of membrane.
The testing results were obtained for membrane with the working surface area of 1.1 cm2.
At the outlet from the fuel side of the reactor, the reactor was loaded by a layer (about 1 m3)
of the catalyst with the same active component LaNi0.9Pt0.1O3/Pr0.3Ce0.35Zr0.35O2-x supported
on the microspherical -alumina fixed between two Fechraloy gauzes protected by corundum
layer supported by dust blasting technique [162].
The membrane reactor was tested in the flow kinetic installations using feeds containing
4.5-20 vol.% CH4 (1-15 L/h) and purified air (0.7-60 L/h) supplied into the fuel and air parts
of the reactor, respectively. Concentrations of reagents and products were determined by
using gas chromatographic analysis and on-line IR absorbance gas analyzer PEM-2M
equipped with an electrochemical H2 sensor [162].
In all experiments, periodical checking of the exit feed by gas chromatographic analysis
on the N2 admixture revealed its absence within the limit of analysis sensitivity (~ 0.1%).
Hence, membrane retained its integrity and oxygen permselectivity in studied conditions.
As follows from Fig. 123, for feeds with a low CH4 content, at a low feed rate mainly
deep oxidation of CH4 occurs. The increase of the flow of the fuel mixture results in the
increase of selectivity of methane conversion into syngas, which is associated with the change
of balance between the oxygen flux through the membrane and methane supply. H2/CO ratio
~ 2 observed at high flow rates agrees with the stoichiometry of methane partial oxidation
into syngas, though a part of CH4 is also combusted into CO2 and H2O.
2.5 70
Concenration (CO2, CO, H2),%
CH4 conversion, %
2.0 60
CO2
1.5 50
1.0 H2 40
0.5 30
CO
0.0 20
1 2 3 4 5 6 7
Flow rate of methane, l/h
Figure 123. Dependence of methane conversion, CO and H2 concentration on the flow rate of 4.5%CH4
in He at 900C. The air flow rate 1.33 L/h.
At a fixed ratio of methane and air flows, the increase of methane concentration in the
fuel mixture from 1 to 5 vol.% insignificantly affects the degree of its conversion, while
increasing syngas selectivity (Fig. 124). Hence, in this case, the increase of the rate of CH4
activation on the layer of catalyst supported on membrane helps to consume the oxygen
transferred through membrane forming mainly selective oxidation products.
In the process of membrane reactor start-up via its heating up to 900 o in the flow of
4.5% CH4 in He with the flow rate of 1 L/h followed by its increasing up to 5 L/h, both
methane conversion degree and syngas yield were 1.5-2 times higher than stationary values.
This suggests that at a low air flow rate nitrogen accumulation in meso- micropores of the
membrane occurs, which hampers the oxygen transfer. Indeed, the increase of the air feed rate
has resulted in the increase of the fraction of deep oxidation products due to a higher oxygen
flux through the membrane (Fig. 125). According to well-known experience in design of
supported membranes, limitation of oxygen viscous flow in pores of supports caused by
nitrogen accumulation is indeed the vital problem which is to be dealt with by optimizing the
pore structure of substrate and tuning the air feed rate [154].
100
methane conversion/selectivity,%
CO
80
H2
60
40 XCH
4
20
0
1 2 3 4 5
methane concentration, vol.%
Figure 124. Dependence of methane conversion, CO and H2 selectivity on methane concentration in the
feed at 900C. The fuel flow rate 5 L/h, the air flow rate 1.2 L/h.
100
methane conversion/selectivity, %
CO
90
80
H2
70
60
50 XCH
4
40
0.5 1.0 1.5 2.0 2.5 3.0 3.5

the air feed, l/h
Figure 125. Dependence of methane conversion, CO and H2 selectivity on the air flow rate at 900C.
The flow rate of 4.5% CH4 in He 5 L/h.
100
methane conversion/selectivity,%
90
CO
80
70
60 H2
50
40
30 XCH
4
20
4 6 8 10 12 14 16 18 20 22
methane concentration,vol.%
Figure 126. Dependence of methane conversion, CO and H2 selectivity on methane concentration in the
feed at 900C. The flow rate of CH4+He 5 L/h. The air flow rate 2 L/h.
Fig. 126 shows effect of methane concentration in the feed on its conversion and CO and
H2 selectivity. While selectivities change rather moderately, methane conversion declines as
expected for the constant air feed rate. In part, decline of methane conversion can be
explained by a partial coking of both supported and inserted microspherical catalytic layers.
Indeed, as follows from Fig. 127, for a high CH4 content in the feed, the increase of air flow
rate favors the increase of methane conversion and CO selectivity at a constant methane feed
rate.
18
Exit gas composition, %

15 CH4
12
H2
9
CO
6
3
0 5 10 15 20
Air feed rate, l/h
Figure 127. Effect of air feed rate on exit gas composition. 980oC, feed 20% CH4 in He at 5 L/h.
15
CH4
14
Exit composition, %
H2
7
6
CO
5
4
880 930 o 980
T, C
Figure 128. Temperature dependence of exit gas composition for transformation of 20% CH4 in He at
feed rate 5 L/h. Air feed rate 3.2 L/h.
25 CO
20
Exit concentration,%
15 H2
10
5
CH4
0.5
C2H6
0.0
0.0 0.2 0.4 0.6 0.8 1.0
Air feed rate, l/min
Figure 129. Effect of air feed rate on exit gas composition for oxi-dry reforming of CH4 in membrane
reactor. 20%CO2 + 20%CH4 in N2, feed rate 0.25 L/min, 930 oC.
Balance estimation revealed that under given conditions the maximum oxygen flux
through the membrane is within the limits of 1-2 mlO2/m2min for 4.5% CH4 in He and ~ 5-7
mlO2/m2min for 20% CH4 in He, i.e. increasing with methane concentration. That is
associated with the increase of gradient of oxygen chemical potential across the membrane
with the increase of reducing potential of the fuel mixture, which is a driving force for its
transfer through the membrane. Therefore, at methane concentration in the input mixture
about 50-100% provided a proper tuning of the air feed, the oxygen flow is expected to
increase at least up to 10-15 ml/m2min, which corresponds to the criterion of membrane
suitability for the practical application [4, 160]. The increase of operation temperature further
increases methane conversion and syngas conversion due to higher oxygen flux through
membrane (Fig. 28).
For CH4 dry reforming, a well-known problem is the catalyst coking leading to its
deactivation. Additional supply of oxygen through dense membrane could alleviate this
problem helping to remove coke precursors from the surface of catalysts supported on
membrane. Indeed, in the case of feed 20% CH4 +20% CO2 in N2, the increase of the air feed
through reactor results in increasing CH4 conversion and syngas content in the exit stream
(Fig. 129), which can be explained by a positive effect of a higher oxygen flux through
membrane.
6. CONCLUSION
For nanocrystalline perovskites (manganites, cobaltites, ferrites and nickelates of La with
a partial substitution by Sr) prepared via Pechini route and their nanocomposites with fluorite-
like oxides (doped ceria or zirconia) mobility and reactivity of the surface/lattice oxygen was
studied under a broad variation of oxygen chemical potential gradient applying variety of
experimental techniques (oxygen isotope exchange, isothermal and temperature programmed
desorption/reduction) for both powdered samples and dense materials. The real/defect
structure and surface composition of these materials characterized by applying TEM with
EDX, XRD, UV-Vis and XPS revealed redistribution of elements between perovskite and
fluorite domains remaining in nanorange even in dense nanocomposites, while isolating
pyrochlore-like phases were not observed. Analysis of the interrelation between the transport
characteristics of these materials and their structural features including theoretical quantum-
chemical methods allowed to demonstrate a pronounced role played by nanostructuring of
these systems (especially, by developed perovskite-fluorite interface in nanocomposites) in
providing enhanced oxygen mobility and reactivity. Procedures of perovskite fluorite
nanocomposites preparation via the ultrasonic dispersion of powders in organic solvents with
addition of surfactants were successfully applied for design of functionally graded cathodes
for IT SOFC and asymmetric supported oxygen-conducting membranes which demonstrated
high and stable performance promising for the up-scaling and practical application.
ACKNOWLEDGMENTS
Different parts of this work were supported by NATO SFP 980878 and ISTC 3234
Projects, Integration Project 57 SB RAS-NAN of Belarus, Project 57 of RAS Presidium
Program No.27, RFBR-ofi_m 09-03-12317 and Russian-French Associated European
laboratory on catalysis (RFBR CNRS 09-03-93112 Project).
REFERENCES
[1] Well, A.F.; Structural Inorganic Chemistry, 5-th Edition, Oxford University Press
(1986), Pt. 2, Chapter 13.
[2] Aleksandrov, K.S.; Beznosikov, B.V. Perovskite-related crystals (structural hierarchy,
variety of physical properties, synthesis of new compounds). Siberian Branch of the
Russian Academy of Sciences Pbls., Novosibirsk, 1997.
[3] Mogensen M.; Primdahl S.; Jorgensen M. J.; Bagger C. J. Electroceramics. 2000, 5,
141-152.
[4] Bouwmeester, H. J. M. Catal. Today. 2003, 82, 141-150.
[5] Arai, H.; Yamada, T.; Eguchi, K.; Seiyama, T.; Appl. Catal. 1986, 26, 265-276.
[6] Yamazoe, N.; Teraoka, Y. Catal. Today. 1990, 8, 175-199.
[7] Wachsman, E.D. Solid State Ionics. 2002, 152- 153, 657- 662.
[8] Wang, W. G.; Mogensen, M. Solid State Ionics. 2005, 176, 457-462.
[9] Adler, S.B. Chem. Rev. 2004, 104, 4791-4843.
[10] Goldberg, E.; Nemudry, A.; Boldyrev, V.; Schllhorn, R. Solid State Ionics. 1999, 122,
17-22.
[11] Sadykov, V.A.; Borchert, Yu.V.; Alikina, G.M.; Lukashevich, A.I.; Mezentseva, N.V.;
Muzykantov, V.S.; Moroz, E.M.; Rogov, V.A.; Zaikovskii, V.I.; Zyuzin, D.A.; Uvarov,
N.F.; Ishchenko, A.V.; Zyryanov, V. V.; Smirnova, A. Glass Physics and Chemistry.
2007, 33, 320-334.
[12] De Souza, R.A. Ionic Transport in Acceptor Doped Perovskites. PhD Thesis,
University of London, London (1996).
[13] De Souza, R.A.; J. Maier. Phys. Chem. Chem. Phys. 2003, 5, 740-748.
[14] Ishihara, T.; Matsuda, H.; Takita, Y. J. Am. Chem. Soc. 1994, 116 3801-3803.
[15] Huang, K.; Tichy, R.; Goodenough, J.B. J. Am. Ceram. Soc. 1998, 81, 2565-2575.
[16] Bhide, V.G.; Rajoria, D.S.; Rao, C.N.R.; Rao, G.R.; Jadhao, V.G. Phys. Rev. B. 1975,
12, 2832-2843.
[17] Van Roosmalen, J.A.M.; Cordfunke, E.H.P.; Helmholdt, R.B.; Zandberg H.W. J. S olid
State Chem. 1994, 110, 100-105.
[18] Anderson, H.U. Defect chemistry of p-type perovskites. In: Proc. 14th Riso In.
Symp.on Materials Science. (Poulsen F.W.; Bentzen J.J.; Jacobsen, T.; Skou, E.;
Ostergard, M.J.L. Eds.) Riso National Laboratory, Roskilde, 1993, 1-18.
[19] Mizusaki, J.; Minima, Y.; Yamauchi, S.; Fueki, K.; Tagawa H. J. Solid State Chem.
1989, 80, 102-111.
[20] Kuo, J.H.; Anderson, H.U.; Sparlin, D.M. J. Solid State Chem. 1989, 83, 52-60.
[21] Goodenough, J.B. Rep. Prog. Phys. 2004, 67, 1915-1993.
[22] Vechersky, S.I.; Batalov, N.N.; Esina, N.O.; Shehtman, G.S. Fizika Tverdogo Tela,
2004, 46, 1433-1440.
[23] De Gennes, P.G. Phys. Rev. 1960, 118, 141-154.
[24] Grundy, A.N.; Hallstedt, B.; Gauckler, L.J. Solid State Ionics. 2004, 173, 17-21.
[25] Nakamura, K. in Mixed Ionic Electronic Conducting Perovskites for Advanced Energy
System (N. Orlovskaya, N. Browning Eds), Kluwer Academic Publ., Boston,
Dordrecht, London (2004), pp. 191-204.
[26] Kilner, J.A.; Berenov, A.; Rossiny, J. Diffusivity of the Oxide Ion in Perovskite Oxides,
In: Perovskite Oxide for Solid Oxide Fuel Cells (Ishihara, T. Ed.), Springer, 2009,
95-116.
[27] Ishigaki, T.; Yamauchi, Sh.; Kishio, K.; Mizusaki, J.; Fueki, K. J. Solid State Chem.
1988, 73, 179-187.
[28] Chebotin, V.N. Chemical Diffusion in Solids. Nauka Pbls., Moscow, 1989.
[29] Kilner, J.A.; De Souza, R.A.; Fullarton, I.C. Solid State Ionics. 1996, 86-88, 703-709.
[30] Adler, S.B.; Lane J.A.; Steele B.C.H. J. Electrochem. Soc. 1996, 143, 3554-3564.
[31] Steele, B.C.H. Solid State Ionics. 1997, 94, 239-248.
[32] Teraoka, Y.; Zhang, H. M.; Okamoto, K.; Yamazoe, N. Mater. Res. Bull. 1988, 23, 51-
58.
[33] Teraoka, Y.; Nobunaga T.; Okamoto, K.; Miura N.; Yamazoe, N. Solid State Ionics
1991, 48, 207-212.
[34] Diethelm, S.; Closset, A.; Nisancioglu, K.; van Herle, J.; McEvoy, A. J.; Gur, T. M. J.
Electrochem. Soc. 1999, 146, 2606-2612.
[35] Ishigaki, T.; Yamauchi, Sh.; Kishio, K.; Mizusaki, J.; Fueki, K. J. Solid State Chem.
1984, 54, 100-107.
[36] Komatsu, W.; Ikuma, Y.; Kato, M.; Uematsu, K. J. Amer. Ceram. Soc. 1982, 65, C211-
212.
[37] Yoo, H.-I.; Lee, Ch.-E. Solid State Ionics. 2009, 180, 326-337.
[38] Diethelm, S.; Closset. A.; Van Herle, J.; Nisancioglu, K. Solid State Ionics. 2000, 135,
613-618.
[39] Lane, J.A.; Kilner, J.A. Solid State Ionics. 2000, 136-137, 997-1001.
[40] Lane, J.A.; Benson,S.J.; Waller, D.; Kilner, J.A. Solid State Ionics. 1999, 121, 201-208.
[41] Katsuki M.; Wang S.; Dokiya M.; Hashimoto T. Solid State Ionics. 2003, 156, 453-461.
[42] Lankhorst, M.H.R.; Bouwmeester, H.J.M. J. Electrochem. Soc. 1997, 144, 1261-1272.
[43] Van Hassel, B.A.; Kawada, T.; Sakai, N.; Yokokawa, H.; Dokiya, M.; Bouwmeester,
H.J.M. Solid State Ionics. 1993, 66, 295-305.
[44] Bouwmeester, H.J.M.; Kruidhof, H.; Burggraaf, A.J. Solid State Ionics. 1994, 72, 185-
194.
[45] Chen, C.H.; Bouwmeester, H.J.M.; Van Doorn, R.H.E.; Kruidhof, H.; Burggraaf, A.J.;
Solid State Ionics. 1997, 98, 7-13.
[46] Lee, T.H.; Yang, Y.L.; Jacobson, A.J.; Abeles, B.; Zhou, M. Solid State Ionics. 1997,
100, 77-85.
[47] Kim, S.; Yang, Y.L.; Jacobson, A.J.; Abeles, B. Solid State Ionics. 1998, 106, 189-195.
[48] Klier, K.; Novakova, J.; Jiru, P. J. Catal. 1963, 2, 479-484.
[49] Klier, K.; Kucera, E. J. Phys. Chem. Solids. 1966, 27, 1087-1095.
[50] Sadovskaya, E.M.; Ivanova, Yu.A.; Pinaeva, L.G.; Grasso, G.; Kuznetsova, T.G.; Van
Veen, A.; Sadykov, V.A.; Mirodatos, C. J. Phys. Chem. A: Gen.. 2007, 111, 4498-4505.
[51] Muzykantov, V.S.; Kemnitz, E.; Sadykov, V.A.; Lunin, V.V. Kinetika i Kataliz. 2003,
4, 349-352.
[52] Muzykantov, V.S.; Popovsky, V.V.; Boreskov, G.K. Kinetika i Kataliz. 1964, 5, 624-
629.
[53] Muzykantov, V.S.; Jiru, P.; Klier, K.; Novakova, J. Collect. Czech. Chem. Commun.
1968, 33, 829-835.
[54] Muzykantov, V.S.; Ewald, G.; von Levis, G. Kinetika i Kataliz. 1974, 15, 1512-1518.
[55] Boreskov, G.K.; Muzykantov V.S. Ann. N.Y. Acad. Sci. 1975, 213, 137-160.
[56] Muzykantov, V.S. React. Kinet. Catal. Lett. 1987, 35, 437-447.
[57] Galdikas, A.; Descorme, C.; Duprez D. Solid State Ionics. 2004, 166, 147-155.
[58] Galdikas, A.; Duprez, D.; Descorme, C. Appl. Surf. Sci. 2004, 236, 342-355.
[59] Duprez, D. Isotopes in Heterogeneous Catalysis, Imperial College Press, London,
2006, p. 133.
[60] Sadykov, V.A.; Bulgakov, N.N.; Muzykantov, V.S.; Kuznetsova, T. G.; Alikina, G. M.;
Lukashevich, A.I.; Potapova, Yu.V.; Rogov, V.A.; Burgina, E. B.; Zaikovskii, V. I.;
Moroz, E.M.; Litvak, G.S.; Yakovleva, I.S.; Isupova, L.A.; Zyryanov, V.V.; Kemnitz,
E.; Neophytides, S. in: Mixed Ionic Electronic Conducting Perovskites for Advanced
Energy Systems (Orlovskaya, N.; Browning, N., Editors), Kluwer Academic
Publishers, Boston/Dordrecht/London, 2004, p. 53-70.
[61] Sadykov, V.; Kharlamova, T.; Batuev, L.; Mezentseva, N.; Alikina, G.; Muzykantov,
V.; Krieger, T.; Pavlova, S.; Zaikovskii, V.; Ishchenko, A.; Zarubina, V.; Rogov, V.;
Bobrenok, O.; Uvarov, N.; Kilner, J.; Druce, J.; Smirnova, A. Mater. Res. Soc. Symp.
Proc. 2008, 1098, HH07-06. 1-6.
[62] Sadykov, V.; Kharlamova, T.; Batuev, L.; Muzykantov, V.; Mezentseva, N.; Krieger,
T.; Alikina, G.; Lukashevich, A.; Rogov, V.; Zaikovskii, V.; Ishchenko, A.; Salanov,
A.; Boronin, A.; Koscheev, S.; Pavlova, S.; Uvarov, N.; Smirnova, A.; Vasylyev, O.
Composite Interfaces. 2009, 16, 407-431.
[63] Sadykov, V.A.; Frolova,Yu. V.; Alikina, G.M.; Lukashevich, A.I.; Muzykantov, V.S.;
Rogov, V.A.; Moroz, E.M.; Zyuzin, D.A.; Ivanov V.P.; Borchert, H.; Paukshtis, E. A.;
Bukhtiyarov, V. I.; Kaichev, V. V.; Neophytides, S.; Kemnitz, E.; Scheurell K. React.
Kinet. Catal. Lett. 2005, 85, 367-374.
[64] Yasuda, I.; Hishinuma M. J. Solid State Chem. 1996, 123, 382-390.
[65] Song, C.-R.; Yoo, H.-I. Solid State Ionics. 1999, 120, 141-153.
[66] Ji, Y.; Kilner, J.A.; Carolan, M.F. Solid State Ionics. 2005, 176, 937-943.
[67] Cheselske, F.J.; Wallage, W.E.; Hall, W.K. J. Phys. Chem. 1959, 63, 505-512.
[68] Hall, W.K.; Leftin, H.P.; Cheselske, F.J.; OReilly, D.E. J. Catal. 1963, 2, 506-517.
[69] Gaillard, F.; Joly, J.-P.; Li, N.; Boreave, A.; Deloume, J.-P. Solid State Ionics. 2008,
179, 941-945.
[70] Huang, K.; Goodenough, J.B. J. Electrochem. Soc. 2001, 148, 203-214.
[71] Schroeder, M.; Huang, K.Q.; Goodenough, J.B. Ceramic Bilayer Membranes for High
Temperature Oxidation of Methane. In: Mass and Charge Transport in Inorganic
Materials. Fundamentals to Devices (Vincencini, P, Buscaglia, V, Eds.), Jesolo di Lido,
Venice, Italy, 2000, p. 1507-1514.
[72] Bouwmeester, H.J.M.; Burggraaf , A.J. Dense Ceramic Membranes for Oxygen
Separation. In: The CRC Handbook of Solid State Electrochemistry (P.J. Gellings,
H.J.M. Bouwmeester, Eds), CRC Press, Inc. 1997, 485-557.
[73] Sunarso, J.; Baumann, S.; Serra, J.M.; Meulenberg W.A.; Liu S.; Lin Y.S.; Diniz da
Costa J.C. J. Membrane Sci. 2008, 320,13-41.
[74] Saracco, G.; Specchia, V. Catal. Rev.-Sci. Eng. 1994, 36, 305-384.
[75] Balachandran, U.; Dusek.; J.T.; Mieville, R.L.; Poeppel, R.B.; Kleefisch, M.S.; Pei, S.;
Kobylinski, T.P.; Udovich, C.A.; Bose, A.C. Appl. Catal. A:Gen. 1995, 133, 19-29.
[76] Tsai, Ch.-Y.; Dixon, A.G.; Moser, W.R.; Ma, Y.H. Ceramics Processing. 1997, 43,
2741-2750.
[77] Li, Sh.; Jin, W.; Huang, P.; Xu, N.; Shi, J.; Hu, M. Z.-C.; Pauzzant, E.A.; Ma, Y.H.
AIChE J. 1999, 45, 276-284.
[78] Tong, J.; Yang, W.; Cai, R.; Zhu, B.; Lin, L. Catal. Lett. 2002, 78, 129-137.
[79] Liu, Y.; Tan, X.; Li, K. Catal. Rev. 2006, 48, 145-198.
[80] Pea, M.A.; Fierro, J.L.G. Chem. Rev. 2001, 101, 1981-2017.
[81] Cotton, F.A.; Wilkinson, G. Advanced Inorganic Chemistry, Interscience, New York,
1972.
[82] Chiba, R.; Yoshimura, F.; Sakurai, Y. Solid State Ionics. 2002, 152-153, 575-582.
[83] Rida, K.; Benabbas, A.; Bouremmad, F.; Pena, M.A.; Martinez-Arias, A. Catal.
Commun. 2006, 7, 963-968.
[84] [84].
[85] Van Roosmale, J. A.M.; Huijsmans, J.P.P.; Plomp, L. Solid State Ionics. 1993, 66, 279-
284.
[86] Hildrum, R.; Aasland, S.; Johannesen, . Solid State Ionics. 1993, 66, 207-218.
[87] Chiba, R.; Yoshimura, F.; Sakurai, Y. Solid State Ionics. 1999, 124, 281-288.
[88] Kumar, R.; Choudhary, R. J.; Khan, M. W.; Srivastava, J. P.; Bao, C.W.; Tsai, H. M.;
Chiou, J. W.; Asokan, K.; Pong, W. F. J. Appl. Phys.. 2005, 97, 093526.
[89] Chiba, R.; Yoshimura, F.; Sakurai, Y. Solid State Ionics. 1999, 124, 281-288.
[90] wierczek, K.; Marzec, J.; Paubiak, D.; Zajc, W.; Molenda, J. Solid State Ionics.
2006, 177, 1811- 1817.
[91] Huang, K.; Lee, H.Y.; Goodenough, J.B. J. Electrochem. Soc. 1998, 145, 3220-3227.
[92] Kharton, V.V.; Yaremchenko, A.A.; Shaula, A.L.; Patrakeev, M.V.; Naumovich, E.N.;
Logvinovich, D.I.; Frade, J.R.; Marques, F.M.B. J. Solid State Chem. 2004, 177, 26-37.
[93] Miyoshi, S.; Furuno, T.; Matsumoto, H.; Ishihara,T. Solid State Ionics. 2006, 177, 2269
-2273.
[94] Ishihara, T.; Nakashima, K.; Okada, S.; Enoki, M.; Matsumoto, H. Solid State Ionics.
2008, 179, 1367-1371.
[95] Vashook, V.; Zosel, J.; Wen, T.-L.; Guth, U. Solid State Ionics. 2006, 177, 1827-1830.
[96] Amow, G.; Au, J.; Davidson, I. Solid State Ionics. 2006, 177, 1837-1841.
[97] Liu, W.; Zhang, G.G.; Xie, S.; Chen, C.S.; Meng, G.Y.; Peng, D.K . Solid State Ionics.
2000, 135, 727-730.
[98] Zyryanov, V.V.; Sadykov, V.A.; Uvarov, N.F.; Alikina, G.M.; Lukashevich, A.I.;
Ivanovskaya, M.I.; Neophytides, S.; Criado, J.M. Solid State Ionics. 2005, 176, 2813-
2818.
[99] Sadykov, V.; Mezentseva, N.; Muzykantov, V.; Sammes, N.; Kilner, J.; Smirnova, A.
Proc.Int. Symp. on Innovative Materials for Processes in Energy Systems (IMPRES),
28-31 October, 2007, Kyoto, Japan, A. 1-5.
[100] Rossignol, C.; Ralph, J.M.; Bae, J.M.; Vaughey, J.T. Solid State Ionics 2004, 175,
59-61.
[101] Mogensen, M.; Lybye, D.; Bonanos, N.; Hendriksen, P.V.; Poulsen, F.W. Solid State
Ionics. 2004, 174, 279-286.
[102] Frayret, C.; Villesuzanne, A.; Pouchard, M. Chem. Mater. 2005, 17, 6538-6544.
[103] Zhao, H.; Mauvy, F.; Lalanne, C.; Bassat, J.-M.; Fourcade, S.; Grenier, J.-C. Solid State
Ionics. 2008, 179, 2000-2005.
[104] Skinner, S.J.; Kilner, J.A.; Solid State Ionics. 2000, 135, 709-712.
[105] Mauvy, F.; Boehm, E.; Bassat, J.-M.; Grenier, J.-C.; Fouletier, J. Solid State Ionics.
2007, 178, 1200-1204.
[106] Liu, J.; Co, C.A.; Paulson, S.; Birss, V.I. Solid State Ionics. 2006, 177, 377-387.
[107] Sadykov, V.A.; Tikhov, S.F.; Tsybulya, S.V.; Kryukova, G.N.; Veniaminov, S.A.;
Kolomiichuk, V.N.; Bulgakov, N.N.; Paukshtis, E.A.; Ivanov, V.P.; Koshcheev, S.V.;
Zaikovskii, V.I.; Isupova, L. A.; Burgina, E.B.; J. Molec. Catal. A. 2000, 158, 361-365.
[108] Sadykov, V.A.; Isupova, L. A.; Zolotarski, I. A.; Bobrova, L. N.; Noskov, A. S.;
Parmon, V. N.; Brushtein, E. A.; Telyatnikova, T. V.; Chernyshev, V. I.; Lunin, V. V.
Appl. Catal. A. 2000, 204, 59-87.
[109] [Read, M. S. D.; Islam, M. S.; Watson, G. W. and Hancock, F. E. J. Mater. Chem.
2001, 11, 2597-2602.
[110] Sadykov, V.A.; Kuznetsova, T.G.; Bulgakov, N.N.; Lunin, V.V., Kemnitz, E. Mater.
Res. Soc. Symp. Proc. 2003, 751, Z3.27.1-6.
[111] Fisher, C.A.J.; Islam, M.S. J. Mater. Chem. 2005, 15, 3200-3207.
[112] Tapilin, V.M. J. Struct. Chem. 2007, 48, 212-218.
[113] Guggilla, S.; Manthiram, A. J. Electrochem. Soc. 1997, 144, 120-121.
[114] Wang, W.; Zhang, H.; Lin, G.; Xiong, Zh. Appl. Catal. B: Envir. 2000, 24, 219-232.
[115] van Roosmalen, A. M.; Gordfunke, E. H. P. J. Solid State Chem. 1994, 110, 113-117.
[116] Provendier, H.; Petit, C.; Kiennemann, A. Chemistry, 2001, 4, 57-66.
[117] Sadykov, V.; Mezentseva, N.; Alikina, G.; Bunina, R.; Pelipenko, V.; Lukashevich, A.;
Tikhov, S.; Usoltsev, V.; Vostrikov, Z.; Bobrenok, O.; Smirnova, A.; Ross, J.R.H.;
Smorygo, O.; Rietveld, B.; Catal. Today. 2009, 146, 132-140.
[118] Dusastre, V.; Kilner, J. A. Solid State Ionics. 1999, 126, 163-174.
[119] Esquirol, A.; Kilner, J.; Brandon N. Solid State Ionics. 2004, 175, 63-67.
[120] Murray, E. P.; Barnett, S. A. Solid State Ionics. 2001, 143, 265-273.
[121] Figueiredo, F. M.; Frade, J. R.; Marques F. M. B. Solid State Ionics. 2000, 135, 463-
467.
[122] Matsuzaki, Y.; Yasuda, I. Solid State Ionics. 2002, 152-153, 463- 468.
[123] Wang, W. G.; Mogensen, M.; Solid State Ionics 2005, 176, 457-462.
[124] Oishi, N.; Yoo, Y.; Davidson, I. Electrochem. Soc. Proc. 2005, 07, 1645-1651.
[125] Jorgensen, M.J.; Primdahl, S.; Bagger, C.; Mogensen, M. Solid State Ionics. 2001, 139,
1-11.
[126] Wang, Zh.; Cheng, M.; Dong, Y.; Zhang, M.; Zhang, H. Solid State Ionics. 2005, 176,
2555 - 2561.
[127] Kharton, V. V.; Kovalevsky, A. V.; Biskup, A. P.; Shaula, A. L.; Figueiredo, F. M.;
Naumovich, E. N.; Marques, F. M. B. Solid State Ionics. 2003, 160, 247-258.
[128] Nigge, U.; Wiemhfer, H.-D.E.; Rmer, W. J.; Bouwmeester, H. J. M.; Schulte T. R.
Solid State Ionics. 2002, 146, 163-174.
[129] Kharton, V. V.; Kovalevsky, A. V.; Biskup, A. P.; Figueiredo, F. M.; Yaremchenko,
A.A.; Naumovich, E. N.; Marques, F. M. B. J. Electrochem. Soc. 2000, 147, 2814-
2821.
[130] Sadykov, V.; Pavlova, S.; Zarubina ,V.; Bobin, A.; Alikina, G.; Lukashevich, A.;
Muzykantov, V.; Usoltsev, V.; Kharlamova, T.; Boronin, A.; Koscheev, S.; Krieger, T.;
Ishchenko, A.; Mezentseva, N.; Salanov, A.; Smirnova, A.; Bobrenok, O.; Uvarov, N.
ECS Trans. 2009, 25, 2403-2412.
[131] Sadykov, V.; Muzykantov, V.; Bobin, A.; Batuev, L.; Alikina, G.; Lukashevich, A.;
Boronin, A.; Krieger, T.; Ishchenko, A.; Bobrenok, O.; Uvarov, N.; Smirnova, A.;
Vasylyev, O. Mater. Res. Soc. Symp. Proc. 2009, 1126, S13-03. 1-6.
[132] Kharlamova, T.; Pavlova, S.; Sadykov, V.; Krieger, T.; Mezentseva, N.; Muzykantov,
V.; Alikina, G.; Boronin, A.; Zaikovskii, V.; Ishchenko, A.; Rogov, V.; Uvarov, N.;
Frade, J.; Argirusis, Chr. Mater. Res. Soc. Symp. Proc. 2008, 1056. HH03-64. 1-6.
[133] Frolova-Borchert, Yu.V.; Sadykov, V.A.; Alikina, G.M.; Lukashevich, A.I.; Moroz,
E.M.; Kochubey, D.I.; Kriventsov, V.V.; Zaikovskii, V.I.; Zyryanov, V.V.; Uvarov,
N.F. Solid State Ionics. 2006, 177, 2533-2538.
[134] Isupova, L.A.; Obyskalova, E.A.; Rogov, V.A.; Tsybulya, S.V.; Dovlitova, L.S.;
Burgina, E.B.; Ischenko, A.V.; Zaikovskii, V.I.; Sadykov, V.A.; Orlovskaya, N. Mater.
Res. Soc. Symp. Proc. 2006, 885E, A03-04.1-6.
[135] Druce, J.; Kilner, J.A. ECS. Trans. 2009, 16, 327-336.
[136] Bannikov, D.O.; Cherepanov, V.A. J. Solid State Chem. 2006, 179, 2721-2727.
[137] Zinkevich, M.; Aldinger, F. J. Alloys and Compounds. 2004, 375, 147-161.
[138] Dong, W.-S.; Roha, H.-S.; Jun, K.-W.; Park, S.-E.; Ohb, Y.-S. Appl. Cata. A: Gen.
2002, 226, 63-72.
[139] Lever, A.B.P. Inorganic Electronic Spectroscopy. 2nd Ed. Amsterdam-Oxford-New
York-Tokyo: Elsevier. 1987. Part 1 and 2.
[140] Pickett, W.E.; Singh, D.J. Phys. Rev. B.1996, 53, 1146-1160.
[141] Ravindran, P.; Kjekshus, A.; Fjellva, H.; Delin, A.; Eriksson, O. Phys. Rev. B. 2002, 65,
064445.1-19.
[142] Choudhary, V.R.; Uphade, B.S.; Pataskar, S.G. Appl. Catal. A:Gen. 2002, 227, 29-41.
[143] Borchert, H.; Borchert, Yu.; Kaichev, V.V.; Prosvirin, I.P.; Alikina, G.M.;
Lukashevich, A.I.; Zaikovskii, V.I.; Moroz, E.M.; Paukshtis, E.A.; Bukhtiyarov, V.I.;
Sadykov, V.A. J. Phys. Chem. 2005, 109, 20077-20086.
[144] Mineshige, A.; Abe, J.; Kobune, M.; Uchimoto, Y.; Yazawa, T. Solid State Ionics.
2006, 177, 1803-1806.
[145] Zheng, Ya.; Ran, R.; Gu, H.; Cai, R.; Shao, Z. J. Power Sources. 2008, 185, 641-648.
[146] Borchert, H.; Borchert, Yu.; Kaichev, V.V.; Prosvirin, I.P.; Alikina, G.M.;
Lukashevich, A.I.; Zaikovskii, V.I.; Moroz, E.M.; Paukshtis, E.A.; Bukhtiyarov, V.I.;
Sadykov, V.A. J. Phys. Chem. 2005, 109, 20077-20086.
[147] Zhogin, I.L.; Nemudry, A.P.; Glyanenko, P.V.; Kamenetsky, Yu.M.; Bouwmeester,
H.J.M.; Ismagilov, Z.R. Catal. Today. 2006, 118, 151-157.
[148] Kharlamova, T.; Smirnova, A.; Sadykov, V.; Zarubina, V.; Krieger, T.; Batuev, L.;
Ishchenko, A.; Salanov, A.; Uvarov, N.; ECS Trans. 2008, 13, 275-284.
[149] Nicholas, J. D., and De Jonghe, L. C. Mater. Res. Soc. Symp. Proc. 2007, 1023, JJ05-
09.1-6.
[150] Lewis, G.S.; Atkinson, A.; Steele, B.C.H.; Drennan, J. Solid State Ionics. 2002, 152
153, 567 573.
[151] Leng, Y. J.; Chan, S. H.; Khor, K. A.; Jiang, S. P. Int. J. Hydrogen Energy. 2004, 29,
10251033.
[152] Yoon, S. P.; Han, J.; Nam, S.W.; Lim, T.-H.; Oh, I.-H.; Hong, S.-A.; Yoo, Y.-S., Lim,
H. C. J. Power Sources. 2002, 106, 160166.
[153] Kovalevsky, A. V.; Kharton, V. V.; Snijkers, F. M. M.; Cooymans, J. F. C.; Luyten,
J.J.; Frade, J. R. Solid State Ionics. 2008, 179, 61-65.
[154] van Hassel, B. A. Solid State Ionics, 2004, 174, 253-260.
[155] Watanabe, K.; Yuasa, M.; Kida,T.; Shimanoe, K.; Teraoka, Y.; Yamazoe, N. Solid
State Ionics. 2008, 179, 13771381.
[156] Takamura, H.; Kobayashi, T.; Kasahara, T.; Kamegawa, A.; Okada, M. J. Alloys Comp.
2006, 408-412, 1084-1089.
[157] Zyryanov, V. V.; Sadykov,V. A.; Alikina, G. M. Separ. Sci. Tech. 2007, 42, 2849-2861.
[158] Li, Ch.; Hu, T.; Zhang, H.; Chen, Y.; Jin, J.; Yang, N.; J. Membr. Sci. 2003, 226, 17.
[159] Deng, H.; Zhou, M.; Abeles, B. Solid State Ionics. 1994, 74, 75-84.
[160] Liu, Y.; Tan. X.; Li, K. Catalysis Reviews. 2006, 48, 145-198.
[161] Takamura, H.; Ogawa, M.; Suehiro, K.; Takahashi, H.; Okada, M. Solid State Ionics.
2008, 179, 1354-1359.
[162] Sadykov, V.; Sobyanin, V.; Mezentseva, N.; Alikina, G.; Vostrikov, Z.; Fedorova, Y.;
Pelipenko, V.; Usoltsev, V.; Tikhov, S.; Salanov, A.; Bobrova, L.; Beloshapkin, S.;
Ross, J. R. H.; Smorygo, O.; Ulyanitskii, V.; Rudnev, V. Fuel, 2010, 89, 12301240,
doi:10.1016/j.fuel.2009.12.015.
Chapter 3
ON THE NATURE OF LOW-TEMPERATURE

RESISTIVE PEAK IN COLOSSAL
MAGNETORESISTANT MATERIALS
A.I. Tovstolytkin*
Institute of Magnetism, 36-b Vernadskogo Blvd., Kyiv 03142, Ukraine
ABSTRACT
Strong interplay between electronic behavior, structure and magnetism in doped
perovskite manganites gives rise to the enhanced sensitivity of transport properties to
magnetic and structural peculiarities. The optimally doped manganite compounds are
metallic and ferromagnetic at low temperatures, while their conductivity displays
activated behavior at high temperatures. The metal-insulator transition between these two
states is strongly associated with the magnetic ordering one, and this manifests itself in
the formation of a resistivity peak in the vicinity of Curie temperature TC. However, in
some peculiar cases, manganite samples exhibit an additional low-temperature resistivity
peak (LT-peak) (or a tendency to its emergence in a form of a plateau), which is not
associated with any magnetic or structural transition and nature of which remains unclear
hitherto. In this chapter, based on a simple phenomenological model, we demonstrate that
the formation of the LT-peak may be a result of a strong broadening of the ferromagnetic
transition. The approach developed examines the electrical properties of a system
consisting of paramagnetic and ferromagnetic phases which exhibit opposite trends in the
temperature dependences of resistivity and coexist over a wide temperature range. It is
shown that, as the region of the phase coexistence increases, the resistive transition first
broadens, then, in addition to the resistive peak near TC, the LT-peak emerges and begins
to grow, later on it becomes dominant and, eventually, only a weak shoulder-like feature
remains at TC. Such an approach is shown to be applicable to the description of the
magnetic and electric properties of a number of inhomogeneous and phase-separated
*
E-mail: [email protected]
180 A. I. Tovstolytkin
manganite systems, including micro- and nanocrystalline samples, ultrathin epitaxial

films, etc.
INTRODUCTION
The doped perovskite manganites La1-xAxMnO3 (A = Ca, Sr, Ba, ) have attracted
much recent attention primarily due to the discovery of negative colossal magnetoresistance
(CMR) around the Curie temperature TC [1,2]. Although it was recognized long ago that the
essential physics of these materials mainly stems from the double exchange resulting from the
motion of an electron between Mn3+ and Mn4+ ions [3], recent studies have highlighted the
important role of the Jahn-Teller distortion associated with the Mn3+ ion as well as charge
ordering effects opposing double exchange (for a review see [4,5]). In view of the competing
character of these interactions, a change in the Mn4+/Mn3+ ratio, caused either by heterovalent
substitutions or by the introduction of cation (anion) vacancies, may greatly affect the
properties of doped manganites, leading in some cases to the appearance of extremely
complex electronic/magnetic orderings [6] or peculiar phase-separated states [7,8].
The optimally doped (x 0.3) manganite compounds are metallic and ferromagnetic
(FM) at low temperatures, while their conductivity displays activated behavior at high
temperatures. The metal-insulator transition between these two states is strongly associated
with the magnetic ordering one and this manifests itself in the formation of a resistivity peak
in the vicinity of TC. The origin of this peak was discussed in detail in papers [9,10,11]. In
many cases, however, an additional low-temperature peak (LT-peak) (or a tendency to its
emergence in a form of a plateau) appears on the resistivity vs. temperature T dependences
for thin film [12,13,14] as well as bulk polycrystalline [15,16,17] samples of manganites,
which is not related to any kind of phase transitions, neither a magnetic nor a structural one.
In this chapter, based on a simple phenomenological model, we demonstrate that the
formation of the LT-peak may be a result of a strong broadening of the ferromagnetic
transition. The approach developed examines the electrical properties of a system consisting
of paramagnetic (PM) and FM phases which exhibit opposite trends in the temperature
dependences of resistivity and coexist over a wide temperature range. It is shown that, as the
region of the phase coexistence increases, the resistive transition first broadens, then, in
addition to the resistive peak near TC, the LT-peak emerges and begins to grow, later on it
becomes dominant and, eventually, only a weak shoulder-like feature remains at TC. Such
approach is shown to be applicable to the description of the magnetic and electric properties
of a number of inhomogeneous and phase-separated manganite systems, including micro- and
nanocrystalline samples, ultrathin epitaxial films, etc.
ELECTRONIC TRANSPORT IN MAGNETICALLY

INHOMOGENEOUS MANGANITE MATERIALS
The complexity of the electronic and magnetic phase diagrams of doped manganites is a
result of a strong interaction between charge, spin and orbital subsystems, with a key role
being played by aliovalent manganese ions [4,7,18]. In such a complex situation there is some
On the Nature of Low-Temperature Resistive Peak 181
attraction in attempting to use averaged parameters such as valency, tolerance factor, MnO
distance or MnOMn bond angle to describe the electric, magnetic and magnetoresistive
properties of doped manganites [5,7,19]. Although there are many examples demonstrating
apparently systematic trends (for a review see [4,5,7]), there is also clear evidence that these
averaged quantities mask the effects resulted from a deviation of an actual parameter from its
average value [20,21,22]. To isolate these effects, one should carry out additional studies on a
series of carefully prepared and characterized samples with fixed values of averaged
parameters.
A manifestation of one such effect is clearly seen from Figure 1. The figure shows
evolution of the vs. T curves for La0.775Sr0.225MnO3 ceramic samples, which is observed
upon a change in a sintering temperature Ts (the details of the samples preparation and
characterization can be seen in [21,23,24]). All the curves exhibit a maximum at Tp 340 K,
which is related to the magnetic transition from PM to FM state [24]. Both a shape of (T)
and Tp value are in good compliance with reference data, including those obtained on single
crystalline samples [4,25]. However, an interesting feature of the (T) curves is an emergence
of additional low-temperature peak observed as Ts decreases from 1350 to 1250 oC. As
follows from our investigations, as well as from the works of other authors [21,23,26,27], the
appearance of LT-peak is a phenomenon which is often encountered in doped manganites, but
its nature remains unclear hitherto.
In recent years, an extra low-temperature peak (in addition to the peak near the Curie
point) on the temperature dependence of electric resistance R(T) was reported for a number of
samples belonging to the (La,Sr)MnO3- and (La,Ce)MnO3 systems [23,26,27]. Tang et al.
[28] observed similar R(T) curves in polycrystalline (La,A)(Mn,Cu)O3- (A = Sr, Ba) samples.
Zhang et al. [16] attributed the low-temperature resistivity peak formed as a result of a
decrease in a granule size in the La0.85Sr0.15MnO3 perovskite to the tunneling between
granules occurring due to different values of the Curie temperature at a granule surface and its
core. Steenbeck et al. [29] studied films with one grain boundary (on a bicrystalline substrate)
and associated the two-peak R(T) dependence as arising from a reduced TC value in the
boundary region. However, such approaches are unable to explain the peculiar behavior of
ultrathin epitaxial films investigated, among others, by Izumi et al. [13]. No explanation was
offered in this paper for the emergence of the second peak upon a decrease in the film
thickness. It should be emphasized that in all the above cases, no magnetic or structural
peculiarities which could lead to anomalies in electrical properties were observed below the
Curie point. It is such kind of resistive anomalies that is a subject of the discussion in this
chapter.
An absence of low-temperature resistive anomalies in bulk single crystalline counterparts
of the samples studied in the above papers points to the idea that the formation of LT-peak is
associated with the presence of defects, structural imperfections, strained regions and/or
chemical inhomogeneities, all of them being inevitably present in polycrystalline samples and
strained films. A number of experimental data testifies in favor of such point. So, there is
ample evidence that the LT-peak is likely to emerge upon a decrease in a grain size (in both
bulk and thin film polycrystalline samples) [16,21,26,30] or film thickness [13], in other
words, upon an increase in the number of various kinds of imperfections (intergrain
boundaries, structural defects, strained regions) or/and enhancement of chemical, structural or
magnetic inhomogeneity. One can expect that such imperfections directly or indirectly
(through a defect-induced magnetic disordering) influence the electric conductivity. The

features of the influence of various kinds of imperfections on a local magnetic order in
manganites will be analyzed below. At this stage, let us only pay attention to the fact that the
increase in the defect number gives rise to the enhancement of magnetic inhomogeneity [20],
and consider the evolution of the electric properties of such system with the increase in the
degree of magnetic inhomogeneity.
A number of experimental studies have indicated that in doped manganites, the PM and
FM phases can coexist over a wide temperature range below the Curie point [27,31,32,33,34].
Chechersky et al. [32] and Simopoulos et al. [35] have described the temperature dependence
of the fraction of PM phase (T < TC) for some Ca-based manganites. An analysis of these data
shows that the magnetic transition may be stretched down to the lowest temperatures (in the
samples studied by Chechersky et al. [32] the fraction of the PM phase is about 20% even for
T/TC = 0.4), and a part of the sample may remain paramagnetic even at absolute zero
[13,16,28]. It is reasonably to assume that the enhancement in magnetic inhmomogeneity
results in a widening of the temperature range where two phases coexist. A distinguishing
feature of the CMR manganites is a strong dependence of their electric properties on magnetic
state: the FM phase has the metallic-type conductivity (dR/dT > 0), while the PM one displays
the activated-type conductivity (dR/dT < 0) [4,5]. Accounting for the fact that the characters
of the R(T) dependence have opposite trends in these phases, the behavior of the resulting
resistance may be quite complex in the regime of the phase coexistence. It is this situation
that will be analyzed in detail below.
Figure 1. Resistivity vs. temperature dependences for La0.775Sr0.225MnO3 ceramic samples synthesized at
different sintering temperatures.
MODEL CALCULATIONS:
ELECTRIC CONDUCTIVITY OF A TWO-PHASE MANGANITE SYSTEM
Consider the behavior of a two-phase system consisting of a mixture of PM and FM
phases which have different types of conductivity (paramagnetic activated, ferromagnetic
metallic). Denote the resistance of the material in the two phases by Rp and Rf, respectively.
Consider a case where the concentration of each the phase (p and f, respectively) is a
function of temperature, and both the phases coexist in a certain temperature interval below a
temperature T0 of the FM phase nucleation (p + f = 1). Having defined specific form of the
Rp(T) and Rf(T) dependences, let us analyze the evolution of the resulting resistance of the
system upon a variation of one of the parameters, p(T) or f(T).
Keeping in mind that we are dealing with the doped manganite systems, choose the
expressions for Rp(T) and Rf(T) as most suitable for these materials. An analysis of extensive
experimental material [4,25,36,37] shows that the temperature variation of the resistance of
the FM phase can be written with a good accuracy in the form
Rf = B + C t2, (1)
which reflects the important role of the electron correlations in these materials [4,25]. Here, t
is the temperature normalized to T0, and B and C are constants.
Papers [38,39] reported on the results of a thorough experimental investigation of the
electric properties of the films and bulk samples of La0.67A0.33MnO3 (A = Ca, Sr) over a wide
temperature range above the Curie point (up to 1200 K). The authors examined the range of
applicability of various models, viz., variable range hopping conductivity, semiconductor-
type conductivity, and small polaron hopping conductivity, the latter in both adiabatic and
nonadiabatic approximations, for a description of the resistance of the PM phase of these
materials. It was shown that the temperature dependence of Rp can be best described by the
formula
Rp = A t exp(E/t), (2)
which corresponds to the small polaron hopping conductivity in the adiabatic approximation
(E is the activation energy of a polaron and A is a constant).
The measurements of the temperature dependence of the volume fraction of the PM
phase were carried out in works [32,35] with the use of Mssbauer spectroscopy for La1-
xCaxMnO3 (x = 0.2; 0.3) samples. Some indirect data, for example, for samples investigated
by Chen and De Lozanne [40], can be obtained by comparing the temperature dependence of
the magnetization of epitaxial and polycrystalline films having the same composition and
prepared under identical conditions but on different substrates. Assuming that the fraction of
the FM phase in the epitaxial film is 100% below TC, we can estimate f in the polycrystalline
film from the ratio between the magnetizations of the films. A similar procedure can be
adopted for samples studied by Izumi et al. [13].
The analysis of the p(T) dependences, adopted from papers [13,40,41], has shown that
the probability of the emergence of LT-peak on the R(T) curve increases with an extension of
the temperature interval of the two phase coexistence. The simplest function enabling a
variation of the width of this interval with the help of a single variable is an exponential
function. Thus, the temperature dependence of the fraction of the PM phase is chosen to be
p = 1 (t > 1) (3a)
p = exp((t - 1)/d) (t 1), (3b)
where the parameter d characterizes the transition width. It should be noted that the values of
p obtained in Refs. [16,28,32,40] do not equal zero even at the lowest temperatures. Hence
the theoretical analysis should cover the cases of d < 1 as well as d 1. A case of a more
complex expression for the temperature dependence of p will be considered separately.
For the calculations of the total resistance R we exploit the percolation model, which was
developed in paper [42] to describe the resistance of the two-phase medium and was
experimentally verified on a large number of binary media. The author of [42] obtained the
equation which relates the resistance R of the system to the resistances of each of the
coexisting phases, Rp and Rf, their volume fractions, p and f, and two parameters the
critical volume fraction of the metallic (in our case ferromagnetic) phase fc, at which the
percolation occurs, and the effective demagnetization coefficient L, which is determined by
the shape of the inclusions of one phase in another one.
Being adapted to our case, this formula can be written as
p (R1/g - Rp1/g)/(R1/g + K Rp1/g) +
+ (1 - p)(R1/g - Rf1/g)/(R1/g + K Rf1/g) = 0, (4)
where
K = (1 - fc)/ fc; g = (1 - fc)/(1 - L). (4a)
The equation (4) is valid until the condition
Rf << Rp (4b)
is satisfied [42].
For simplicity, let us make calculations, assuming that the shape of the inclusions is close
to spherical and, thus, L = 1/3. The estimations made in [42] showed that for the ellipsoids of
rotation, oriented along the current flow, the changes in L are small enough for a wide range
of ratios between the ellipsoid axes, 1/2 < c/a < 2. For this reason, the results of the
calculations are not particularly sensitive to the morphology of the mutual distribution of the
phases and such the choice of L is well-grounded.
The critical volume fraction fc of the metallic phase at which percolation occurs is
chosen to be 0.16, which is in accord with the experimental results and the theoretical
analysis of the behavior of CMR manganites [43]. Such a choice of fc and L gives g = 1.26
and K = 5.25 in Eq. (4).
Figure 2. Temperature dependences of the volume fraction of PM phase p (a) and electric resistance R
(b) in a two-phase system for different values of the parameter d.
Figure 2 shows the temperature dependences of the volume fraction of the PM phase p
and the resulting resistance R of a two-phase system calculated by using Eq. (4) for a set of
the values of the parameter d. The coefficients included in Eqs. (1) (4) are:
B = C = 10-4, A = 1, E = 3. (5)
Such a choice of the coefficients was made, on the one hand, based on the results of the
analysis of experimental data obtained in Refs. [27,38,39] and, on the other hand, in view of
the constraint imposed by the inequality (4b) on the functions Rp and Rf.
It is clearly seen from Figure 2(b) that for a sufficiently narrow (d 0.5) transition, the R
vs. t dependence exhibits a peak near t =1, as is usually observed in CMR manganites [4,5].
An increase in d gives rise to a broadening of the resistive transition. However, beginning
from d 3.2, a second peak is formed at t < 1, in addition to the former one. A sharp rise in
the LT-peak is observed upon the further increase in d. This peak becomes dominant and
eventually, for d > 4.2, only a weak shoulder-like feature, which may be hard to detect
experimentally, remains at t = 1. Concerning experimental results, the situation where the
temperature of the resistance maximum strongly differed from the Curie temperature has been
reported in works [16,27,28,30,31,32,44]. In such cases, great care should be exercised while
interpreting the experimental results and comparing the resistive data with the magnetic ones.
As our calculations show, the characteristic features of the evolution of the R(t)
dependences remain almost unchanged for other values of the coefficients included in Eq. (4),
as well for a more complex form of the temperature dependence of the volume fraction of PM
phase. The first stage is the broadening of the resistive transition, the second one the
formation and growth of the LT-peak, and, finally, the third stage almost complete
disappearance of the peak near the temperature of the FM phase nucleation. Figure 3
illustrates the results of the calculations performed for the case where p has the form
p = 0 + (1 - 0)exp((t - 1)/d) (t < 1), (6)
and the values of the corresponding parameters1 are:
pc = 0.5; 0 = 0.49. (7)
It should be noted that the strong temperature broadening of the magnetic transition is a
phenomenon encountered quite frequently in doped manganites. Apparently, this is due to a
competition between different types of interactions characteristic of these materials, as well as
to a strong dependence of the magnetic properties on microstructure, mechanical stresses,
degree of disorder, and chemical inhomogeneities. Strongly broadened transitions were
observed experimentally in works [16,27,28,30,31,32,44]. Regarding the topic of our
investigations, the most illustrative results were reported in Refs. [13,21,28,40], where a
correlation between the magnetic transition width and the emergence of the low-temperature
resistive anomaly could be traced clearly.
Figure 3. Temperature dependences of resistance calculated according to the expressions (4), (6) (7).
d is the width of the transition.
1
The calculations show that for the case where the expression for p has the form (6), the features of the R(T)
evolution are determined by the difference between 0 and pc, rather than by their absolute values.
Thus, we have theoretically analyzed the electrical properties of a system consisting of

PM and FM phases which have different types of conductivity and coexist over a wide
temperature interval. It is shown that in spite of the monotonic variation of magnetic and
electric properties of each the phase, the resulting resistance may exhibit anomalies in the
region of the phase coexistence. A short review of experimental data available points towards
high likelihood of the emergence of such anomalies in doped manganites.
EFFECT OF MICROSTRUCTURE ON TRANSPORT

PROPERTIES OF MANGANITE POLYCRYSTALS
Transport and magnetoresistance properties of manganite polycrystals have been
observed to differ from those of single crystals and epitaxial films [45,46,47]. Contrary to the
perfect crystals, both enhanced low-field magnetoresistance over a wide temperature range
below TC and increased overall resistivity are typical of polycrystals, and also non-
coincidence of ferromagnetic and metal-insulator transition temperatures has repeatedly been
reported [16,27,28,30,31,44]. In order to deeper understand the physics of such materials, one
should investigate not only the properties intrinsic to single crystalline state but also those
caused by microstructure, strain, disorder, and compositional variations.
This section focuses on the detailed investigation of the transport, magnetic and
magnetoresistive properties of bulk manganite samples depending on microstructural changes
such as grain size and intergrain connectivity. It is aimed at the elucidation of the nature of a
peak on a temperature dependence of electric resistance, which emerges at TLT 0.7TC. It is
shown that the decrease in a grain size and/or deterioration of intergrain conductivity can
result in a formation of such the peak. The data obtained can be well explained in terms of the
coexistence of the PM and FM phases which exhibit opposite trends in the temperature
dependences of resistivity, with the features of the phase coexistence being dependent on a
sample microstructure.
Microstructure effects were studied in a series of La0.825Sr0.175MnO3- polycrystalline
samples prepared through a two-stage solid state reaction in air according to the procedure
described in Refs. [21,24]. At the final stage of the samples preparation, the obtained
products were sintered for 2 hours at Ts = 1360 (pellet A), 1300 and 1200 oC (pellets B and C,
respectively), and then furnace cooled down to room temperature. The details of the samples
characterization and measurements were described elsewhere [21].
Figure 4 (a-c) shows the scanning electron microscopy (SEM) pictures of the cleaved
surfaces for the samples A, B and C, respectively. As the sintering temperature decreases, the
size of the grains becomes smaller. The average intercept length over large number of grains,
calculated according to the method suggested by Mendelson [48], is equal to 3.6 m for the
sample A, 2 m for B and 1.25 m for C. It is seen that the decrease in Ts gives rise to the
decrease in the average grain size, and thus, to the increase in the fraction of intergrain area.
As the latter is structurally and magnetically disordered [41,49], one can expect that a degree
of magnetic inhomogeneity will be enhanced upon a transition from A to C. In what follows,
we will present the experimental evidence of this statement.
Figure 4 (a-c). SEM images of the cleaved surfaces of the samples A, B, and C, respectively.
Figure 5 (a-c) shows the results of the measurements of electric resistance R and
magnetic susceptibility for the samples A, B and C. The character of the temperature
dependences of indicates that all the samples become FM upon the temperature lowering.
The Curie temperature TC was calculated as a temperature of a maximum on a -d/dT vs. T
curve. The TC value is almost the same for all the samples and is about 250 K. For all the
plots of Figure 5, the temperature is normalized to the TC value determined in above way.
Figure 5 (a-c). Temperature dependences of electric resistance R and magnetic susceptibility for the
samples A, B, and C, respectively. Bold solid lines denote R(T) dependences extrapolated from a high-
temperature region. R resistance measured at zero external magnetic field; R15 kOe resistance under
an applied field of H = 15 kOe.
Figure 6. Normalized dependences of -d/dt vs. t for the samples A, B and C.

As is seen from the figure, the R(T) dependences are similar in the PM region but
markedly differ from each other around and below Curie temperature. The resistivity of the
sample A grows with lowering temperature, reaches a maximum at Tp slightly higher than TC,
and displays a broad peak at TLT 0.7TC (~ 165 K). While external magnetic field H = 15 kOe
suppresses the first peak, giving rise to the maximum in magnetoresistance MR = (R RH)/R,
there is no visible MR anomaly around the latter. No any peculiarity near TLT is also observed
on the dependence of magnetic susceptibility vs. temperature. When going from Figure 5(a)
to Figure 5(c), the evolution of the transport and magnetoresistance behavior has the
following features:
(a) the low-temperature resistance peak gradually grows relative to the peak at Tp, until
the latter completely disappears in the sample C;
(b) while magnetoresistance at T << TC progressively increases, the MR peak near TC
becomes smaller and broader, and finally vanishes, as well.
The grain-size dependent magnetic and transport properties of polycrystalline manganites

were studied in Refs. [16,47,49]. It has been found that the reduction in a grain size gives rise
to the broadening of the FM transition and often results in the decrease in saturation
magnetization. Our data show similar trend. Figure 6 presents the dependences of -d/dt vs. t,
normalized to the maximal ordinate value, for the samples under investigation. The width of
the transition measured at an ordinate value y = y0/e 0.37y0 (y0 is a peak value of -d/dt for
each of the samples; e = 2.718) grows from 0.05 to 0.19 TC, as the average grain size
decreases from 3.6 m to 1.25 m. These effects can be understood taking into account
evolution of the ratio of grain surface area to grain volume. The structure and composition of
the surface layer differ from those of a grain core, and stronger magnetic disorder is expected
far from the center of the grain [16,47]. As the particle size decreases, the contribution from
the surface area increases. As a result, this brings about a degradation of the FM phase, leads
to a broadening of the magnetic transition and often to a decrease in saturation magnetization.
Consider the behavior of electric resistance in more detail. It was mentioned above that at
T > TC, the electronic transport in manganite materials is best described by the model of small
polaron hopping [39]. In adiabatic approximation the activated conduction of small polarons
follows Eq. (2). We have analyzed our data using this model. Shown by bold solid lines in
Figure 5 are the curves fitted according to the above expression. The results of the analysis
demonstrate that activation energy E grows from 105 to 137 meV, when going from the
sample A to the sample C.
As temperature decreases, the behavior of experimental data deviates from that predicted
by adiabatic small polaron hopping model. This deviation is accompanied by a sharp growth
of both susceptibility and magnetoresistance. Further temperature lowering leads to the
formation of the peaks near TC on the R(T) and MR(T) dependences for the samples A and B.
It is noteworthy that, although in the case of the sample C the resistivity at TC is far from its
maximum value, the conduction below TC does not obey adiabatic small polaron hopping law.
Figure 7. (a) R vs. t dependences calculated according to Eq. (4). (b) Experimental R vs. t dependences
for the samples A, B and C. Shaded are the regions where the LT-peak emerges.
Close to and below the Curie temperature electronic behavior of the samples under study
considerably differs from each other (see Figure 5). A and B exhibit a relatively sharp
resistance peak near TC and a broad one at TLT < TC, the latter only beginning to emerge in the
case of the sample A. When going from A to C, the height of the LT-peak on R(T) curve
grows relative to the resistance at TC. This effect is accompanied with a decrease in
magnetoresistance near TC and, as was shown above, with a broadening of magnetic
transition. There is no any resemblance to the high-temperature (near TC) peak in the sample
C.
The main trends in the transformation of the R(T) and (T) curves, observed upon the
decrease in grain size, are in compliance with the results obtained in other research groups
[40,45,47]. The analysis of these data confirms the ideas that the emergence of the LT-peak is
a result of a broadening of magnetic transition and that it is usually observed in a region of the
phase coexistence. The behavior of doped manganites for the case of the broadened magnetic
transition was discussed in detail in the previous section. Figure 7 illustrates the results of the
calculations made according to Eq. (4) for the following parameters:
(a) a critical volume fraction of the metallic phase, at which percolation occurs, is equal
to 0.5;
(b) the temperature dependence of the PM fraction p is described by the expression p =
0.49 + exp((t-1)/d), where t = T/TC and the parameter d characterizes the transition
width in units of TC.
As is seen from Figure 7, the character of the evolution of the R(t) dependences, observed
upon the increase in the magnetic transition width, is similar for the experimental and
calculated data. With regard to the more detailed computations, it does not make sense to
aspire a complete coincidence of the calculated data with the experimental ones, because the
model requires Rf << Rp, but this is not the case for the samples under study.
The conclusions of the theory are directly applicable to the description of the transport
and magnetoresistance properties of the sample C. It should be noted that, although there is
no visible anomaly near TC on the R(t) curve, the resistance at and below Curie point is far
from that extrapolated from PM region.
It follows from the theoretical calculations that, as the LT-peak begins to develop, the
conduction of the system becomes strongly dependent on subtle features of the electric
properties of each of the phases, which, in turn, are very sensitive to the intergrain
connectivity [14]. This fact can explain the evolution of the resistivity of manganite samples,
which was observed to occur when the number of cooling-heating cycles increases [21].
As a result, we have correlated transport properties of La0.825Sr0.175MnO3- polycrystalline
samples with their magnetic and microstructure features. It has been shown that both grain
size and intergrain connectivity strongly affect electric conductivity, leading in some cases to
the formation of LT-peak. The analysis of the experimental data leads to the conclusion that
the low-temperature resistance anomaly arises as a result of a wide-temperature-region
coexistence of the phases which exhibit opposite trends in the temperature dependences of
resistivity.
SELF-DOPED LANTHANUM MANGANITES AS A PHASE-SEPARATED

SYSTEM: TRANSFORMATION OF TRANSPORT PROPERTIES
WITH DOPING AND HYDROSTATIC COMPRESSION
Self-doped manganites LaxMnO3+ (0.815 x 1) are suitable model compounds for the
examination of the approach developed, since for the samples synthesized under identical
conditions, the character of the R(T) dependences changes from activated (x = 1) to well
pronounced double-peaked (x = 0.90, 0.815) with a simultaneous transformation of magnetic
structure from cluster spin glass to FM [50,51,52]. At the same time, as x changes from 1 to
0.815, the concentration of the metallic phase crosses the percolation limit. Such a wide range
of variation of transport and magnetic properties is brought about by a variation of both the
concentration of Mn4+ ions and cation vacancies in lanthanum and manganese sublattices
[52]. The application of high pressures makes it possible to change the character of
conductivity from activated to metallic for the sample with x = 0.94 [51,52]. This section
analyzes the electric properties of LaxMnO3+ (0.815 x 1.0) polycrystalline samples at

normal and high pressures up to 11.5 kbar.
A series of LaxMnO3+ (x = 0.97; 0.94; 0.90; 0.815) polycrystals (samples L0.97, L0.94,
L0.90, and L0.815, respectively) was synthesized by a conventional technology of solid state
reactions in air [50,52,53]. The details of the measurements carried out are described
elsewhere [8,50-53].
Magnetic and resonance properties of the LaxMnO3+ (0.815 x 1.0) polycrystalline
samples were studied as functions of temperature, La/Mn ratio, and hydrostatic compression
in works [37,51,52]. It was shown that upon lowering the temperature, all the samples
undergo a transition from the high-temperature PM state to a low-temperature one, with the
latter being either FM for x = 0.815 0.94, or cluster spin glass for x = 0.97 and 1.0. Both the
temperature of magnetic transition and saturation magnetization diminish and the transitions
become more broadened as the La/Mn ratio grows. The hydrostatic compression strongly
affects the value of the Curie temperature, which reflects the enhanced sensitivity of the
double exchange interaction to the interatomic distance. It is demonstrated that in all the
samples under consideration, the PM phase persists down to the lowest temperatures and its
volume fraction rises with the increase in x.
The vs. T dependences for the L0.97, L0.94, L0.90, and L0.815 samples are depicted in
Figure 8. For the L0.97 sample, the electric conductivity exhibits the activated character (d/dT
< 0) over the whole temperature range under discussion (77 300 K). For this sample, the
application of high pressure strongly reduces resistivity, but does not change the character of
its temperature dependence. On the contrary, a strong transformation of the (T) curve under
the influence of pressure is observed for the L0.94 sample (Figure 9). At normal pressure, a
broad maximum at TLT 105 K and slightly noticeable tendency towards metallic (d/dT > 0)
conductivity below TLT are characteristic of the vs. T dependence. Only a slight shoulder-
like anomaly is observed at TC. High pressures diminish resistivity approximately by one
order of magnitude at low temperatures and essentially widen the region of the metallic state.
It is noteworthy that a similar result was also reported in work [54], where the application of a
pressure of 9.4 kbar induced the metallic state in La0.94Mn0.98O3 single crystal below TC 210
K.
For the case of the L0.90 and L0.815 samples, the vs. T dependence shows two strongly
pronounced peaks: the first one near TC and the second, the LT-peak, at 200 and 173 K for the
former and latter samples, respectively. High pressures do not change the character of the
curves and to a lesser extent affect the absolute values of resistivity.
To gain a deeper insight into the phenomena observed, let us analyze the behavior of
these materials based on the approach described above. To calculate the vs. T dependence
within the region of the PM-FM phase coexistence, we use the f(T) and p(T) dependences
obtained experimentally at low and high temperatures, respectively. For each the sample, the
volume fraction of the PM phase at low temperatures, 0, (see Eq. (6)) can be found from the
magnetic measurements [37,52]. For calculations, we set fc = 0.45, as follows from the
analysis of the experimental data for the samples of this system [37].
Figure 8. vs. T dependences for the L0.97, L0.94, L0.90, and L0.815 samples.
Before proceeding further, let us discuss how strongly the intrinsic transport properties of
the samples under investigation might be affected by porosity effects.
To date, various aspects of the porosity effects on electric and heat transport properties
have been analyzed in Refs. [55,56,57,58,59] with the use of a number of approaches and
calculation techniques. In particular, a simple analytical model developed by Mizusaki et al.
[56] relates the conductivity of highly porous but simply connected ceramics to its porosity. A
numerical method developed by Zinchenko [57] is capable of calculating the effective
conductivity of randomly packed granular materials with a high particle-to-medium
conductivity ratio. The features of the influence of tunneling effects on the electric
conductivity of porous materials were specified in papers [60,61]. All the above results can be
summarized as follows. For the cases where the tunneling effects do not play an important
role in the charge transfer processes, the porosity affects the absolute value of resistivity and
scarcely transforms its temperature dependence. Otherwise, it is the tunneling effects that
govern the temperature, electric field and stress dependences of the electric resistivity
[46,60,61]
Figure 9. vs. T curves at different pressures for the L0.94 sample.

Figure 10. Simulated vs. T dependence (dash line) along with the experimental data (bold solid line)
for the L0.90 sample at normal pressure.
Figure 11. The same as in Figure 10 for the L0.815 sample.
As follows from the data of work [37], the characteristic feature of the microstructure of
the LaxMnO3+ samples with x = 0.815 and 0.90 is the presence of simply connected domains
containing a number of grains. The porosity of these samples does not exceed 7 %. In both
the cases, the peak resistivity is comparable to the values which were observed in
polycrystalline samples for which the negligible role of tunneling effects was proved
experimentally [26,62,63]. All these facts imply that for the L0.815 and L0.90 samples, the
porosity affects the absolute value of resistivity but hardly modifies its temperature
dependence. For the other samples, however, the role of porosity as well as the contribution
of tunneling effects to the conductivity should be discussed separately. This will be done
below while discussing the resistivity of the L0.94 sample.
The L0.90 sample is most suitable for the comparison of the experimental and calculated
vs. T dependences, since for this, a complete set of the parameters necessary for calculations
is deduced from experiment [33] and only the parameter d, which characterizes a transition
width, should be fitted. As follows from the calculations, it is for d = 0.11 that the calculated
curve is closest to the experimental one, concerning both the LT-peak position and its height.
Figure 10 compares the results of the calculations with the experimental (T) dependence at
normal pressure. The fact that the calculated (TC) value is somewhat higher than the
experimental one can be explained by the presence of a small fraction of the FM phase at T
TC. To date, there has been ample direct evidence that in doped manganites, the FM clusters
may be present above TC [64,65].
The results obtained for the L0.90 sample are in compliance with the SEM investigations
[37]. The latter show that, due to a relatively large average grain size and ample amount of
simply connected metallic domains, the role of intergrain charge transfer in the total
conductivity does not prevail. In fact, it turns out that the double-peaked character of vs. T
dependence can be naturally explained within the frames of the simple model which accounts
for the coexistence of two magnetic phases with different characters of temperature
dependence of resistivity.
Figure 11 shows the results of a similar simulation for the L0.815 sample. It can be seen
that for d = 0.09 (the best choice), the calculated dependence quite well describes the
experimental data. It should be noted that the fact that the fitted value of the parameter d
decreases from 0.11 to 0.09 upon a transition from L0.90 to L0.815 sample is in accordance with
the data of magnetic susceptibility measurements [52]: the magnetic transition becomes
sharper as the La content diminishes.
A comparison of the calculated and experimental data for the L0.94 sample is made in
Figure 12. The best agreement in the range 130 300 K is achieved for d = 0.11. The
calculated data describe well both the shoulder-like anomaly at TC = 205 K and the activated
character of resistivity down to 130 K. However, at lower temperatures, the difference
between the calculated and experimental data is essential: for all reasonable values of d, the
calculations show a distinctive metallization, instead of a plateau, and, as a result, the
calculated and experimental data differ by about 5 times. As was noted above, the analysis of
the SEM pictures is indicative of the increased contribution of intergrain charge transfer
processes in the samples with greater x. Thus, although the grain cores may display metallic
conductivity at low temperatures, the contribution from the surface layers, which usually have
degraded magnetic properties and raised resistivity, impedes a strong resistivity reduction,
which eventually results in a formation of a plateau.
Let us consider how well the approach under consideration can describe the (T)
dependence at high hydrostatic pressures. Analyze the experimental curves, obtained at a
maximal pressure for two characteristic samples: the L0.90 sample, which displays metallic
conductivity with negligible changes under high pressure, and the L0.94 sample whose
metallization under high pressures is most pronounced. A comparison of the concentration
dependences of -1(La/Mn) at normal and high pressure (11 kbar) [37] shows that the value
of the La/Mn ratio, at which percolation occurs, is not changed with pressure and
approximately equals 0.95. This, however, does not automatically mean that also the value
of fc remains the same, i.e. 0.45. As follows from the experimental data [66,67], in
manganites, the application of even moderately high pressures (<10 kbar) gives rise not only
to the TC growth, but also to the increase in f. So, the pressure-effect coefficient for the FM
volume fraction of LaMnO3.05 manganite was shown to be 8.710-3 kbar-1 [66]. As follows
from the recent data [67], the application of the pressure of 9.3 kbar to La0.7Sr0.3MnO2.85
manganite leads to the increase in the f value from 13 % to 18 %. For this reason, in further
calculations, we should take into account the likely effect of the high pressure on the change
of the FM volume fraction and vary not only the parameter d, but also f. Based on the data
available, we set a restriction that the increase in f, caused by the application of a pressure
11 kbar, cannot be less than 5 % and more than 10 %.
As follows from the calculations, the activation energy E (see (2)) for the L0.90 sample
decreases from 130 meV at normal pressure to 126 meV at high pressure. This change in E
corresponds to the rate dE/dP = -0.34 meV/kbar, which is close to the value -0.6 meV/kbar
obtained in paper [68] for La0.75Ca0.25MnO3. The best agreement between the calculated and
experimental data is achieved when the increment in both 0 and fc is the same and equals 5
%. Figure 13 compares the results of the calculations with the experimental data obtained at P
= 10.8 kbar. The best correspondence is observed for d taking the value 0.11. It should be
noted that the correspondence becomes far worse when only the increment in 0 is taken into
account with fc set to be constant, or vice versa.
Figure 12. The same as in Figure 10 for the L0.94 sample.
for the L0.90 sample at a pressure of 10.9 kbar.
for the L0.94 sample at a pressure of 11.8 kbar.
Thus, the approach employed not only allows us to describe the effect of high pressure on
the magnetic and transport properties of doped manganites, but also provides a deeper insight
into the nature of the changes induced by the pressure. So, as follows from the calculations,
for the case of the L0.90 sample, both the relative volume fractions of the coexisting phases
and the parameters of each of the phases turn out to be drastically affected. In particular, the
changes the FM phase undergoes are the 5 % increment in the volume fraction and 3 times
resistivity drop under a pressure of 10.9 kbar.
For the case of the L0.94 sample, the activation energy displays a slight increase with the
pressure rise with a rate dE/dP = 0.29 meV/kbar. At none of the fitting parameter values can a
satisfactory agreement be achieved between the calculated and experimental vs. T
dependences obtained at P = 11.8 kbar (Figure 14). At d 0.1, the LT-peak resulted from
simulations is a few times higher than its actual value. The smaller d values neither go along
with the results of the calculations at normal pressure (see Figure 12), nor agree with the data
of magnetic susceptibility measurements [52]. Thus, the fact that neither at normal pressure,
nor at high one, can the behavior of the resistivity be satisfactory described within the frames
of the model that we successfully employed earlier, testifies to the enhanced role of the
processes of intergrain charge transfer in L0.94 and this is likely to be associated with a small
grain size and poor intergrain connectivity. Therefore, a pronounced metallization under
high pressure results from the transformation of the processes of charge transfer through the
intergrain area, rather than from the change in the character of the grain core resistivity.
As a result, the electric properties of LaxMnO3+ (0.815 x 1.0) polycrystalline samples
have been studied in the temperature range 77 370 K and at high pressures up to 11.8 kbar.
It is shown that the increase in La/Mn ratio gives rise to a change in the low temperature
magnetic state from FM to cluster spin glass, as well as to a drastic transformation of electric
properties. A peculiar double-peaked shape is characteristic of the resistivity vs. temperature
curves for the intermediate range of x values. Within this range, the application of high
pressures drastically changes both the value of resistivity and the character of its temperature
dependence. It is shown that the approach, which regards the state of the LaxMnO3+
polycrystals as a mixture of interpenetrating PM insulating and FM metallic phases, is able to
successfully describe the peculiarities of the temperature dependence of total resistivity, as
well as the features of its transformation under hydrostatic compression. It is demonstrated

that the formation of a low temperature resistance peak is a result of a wide-temperature-
region coexistence of the phases which exhibit opposite trends in the temperature
dependences of resistivity. The conclusion is made that not only does the hydrostatic
compression result in drastic changes in the relative volume fractions of the coexisting
phases, but it also affects the intrinsic parameters of each of the phases.
TRANSPORT AND MAGNETORESISTANCE PROPERTIES

OF NANOCRYSTALLINE MANGANITE SAMPLES
It was shown above that the properties of doped manganites are very sensitive to the
microstructural features of these materials. It is appealing to find the ways for a wide-ranging
governing of the material microstructure. In most cases, however, manganese oxides are
synthesized at high temperatures using a standard ceramic technique, which allows one to
prepare the samples with a grain size of the order of 1 m. Recently, chemical preparation
methods such as sol-gel and coprecipitation from solutions have been introduced to
synthesize the CMR oxides [49,63]. All these methods maintain the metal precursors in a
highly dispersed state and offer unique advantages including more purity, homogeneity and
precise composition control. What is more, such methods make it possible to synthesize the
samples with a grain size of the order of tens nanometers [49,63].
In this section, we report the magnetic, transport and magnetoresistance properties of
nanocrystalline La0.7Ca0.3MnO3 synthesized by a method of coprecipitation from solutions.
The details of the sample preparation and characterization were described elsewhere [24,63].
The average grain size estimated from the tunneling electron microscopy studies is about 20
nm [63].
Figure 15 presents the temperature dependences of resistivity of La0.7Ca0.3MnO3 sample
measured at H = 0 and 15 kOe. The (T) curve shows a broad peak at TLT 230 K and a
shoulder-like feature at Tp 264 K. Magnetic field does not change the nature and position of
the broad peak very much, but suppresses the second one greatly. The difference in the
magnetic field response of the two peaks indicates that they may have different origins. Near
Tp, the low-field AC susceptibility exhibits a sharp increase with a decrease of temperature
(Figure 16), indicating an occurrence of magnetic ordering. The onset of the magnetic
ordering corresponds to a PM-FM transition at TC 266 K, which is consisted with other
reports [4,7]. As can be seen from Figure 16, the susceptibility changes monotonically with
further lowering temperature, and neither nor -d/dT displays any anomaly in the vicinity of
TLT. Similarly, the magnetization M measured at 10 kOe as a function of temperature (not
shown here) does not exhibit peculiarities below TC.
Figure 17 shows the magnetic hysteresis loop M(H) at T = 5 K. As a function of field, M
tends to saturate at H 6 kOe, and the saturation magnetization Ms is equal to 3.4 B/Mn.
This value, being lower than that predicted by theory for the given composition (3.7 B/Mn),
suggests that not all the volume of the sample is in FM state even at temperatures close to
absolute zero. Such result is not surprising for materials with sizeable amount of intergrain
boundaries area. A significant decrease in Ms upon reducing the particle size was reported by
Huang et al. [30] and Balcells et al. [69] for granular La0.7Sr0.3MnO3 perovskites. This
phenomenon may be related to a discrepancy in magnetic behavior between the grain core
and intergrain region. Indeed, since the grain boundary is expected to be similar to the glassy
or amorphous state and the magnetic configuration in the grain surface is more chaotic than
that in the core, the coupling between the magnetic ions in the surfaces should be weaker than
that in the cores.
Magnetoresistance ratio MR shows a steady decline with increasing T, interrupted by an
increase near Tp, before finally dropping to zero above 285 K (see Figure 15). Two
contributions to the magnetoresistance are clearly seen from the figure, one is the cusplike
MR near TC, related to the intrinsic properties of the system, and the other is low-temperature
MR, extrinsic in nature, whose origin lies in the existence of interfaces and grain boundaries
[45,46,47]. In our case, the latter component reaches a value of 24 % at T = 80 K, indicating a
substantial contribution of intergrain area to the transport and magnetotransport of the
samples under investigation.
Figure 15. Temperature dependences of resistivity and magnetoresistance for the nanocrystalline
La0.7Sr0.3MnO3. resistivity measured at zero external magnetic field; 15 kOe resistivity under an
applied field of H = 15 kOe.
Figure 16. Dependences of magnetic susceptibility and -d/dT vs. T.

Figure 17. Magnetic hysteresis loop at T = 5 K.
With regard to the mechanisms lying behind the low-temperature component of MR, two
different interpretations have been given to date, namely, spin-polarized intergrain tunneling
[46,47] and spin-dependent scattering at the grain boundaries [45]. Gross et al. studied
electrical transport properties of artificial grain boundaries in La2/3Ca1/3MnO3 films and
explained the results within a model implying the presence of a disordered PM grain
boundary layer that is depleted below TC [46]. The authors concluded that the transport across
the depleted layer is dominated by multi-step tunneling via localized states. Nevertheless, for
tunneling to be a dominant conduction mechanism, one should expect strongly non-Ohmic
current I vs. voltage V characteristics [46,47]. As can be seen from Figure 18, this is not the
case for the nanocrystalline La0.7Ca0.3MnO3 samples. The linear I(V) dependence over more
than two orders of V is strong evidence excluding tunneling as the dominant conduction
mechanism in the samples under study.
Figure 18. Current vs. voltage characteristics at 77 K. Inset shows the same dependence on a
logarithmic scale.
In a recent paper devoted to spin-polarized transport in magnetic oxides Gupta et al. [70]
pointed out that the grain boundary conductivity might be quite inhomogeneous, with
insulating microscopic regions separated by more conducting regions. Conduction in this
parallel channel network can then occur by a combination of spin-scattering and inelastic
tunneling processes, with the relative contribution varying for different materials. In a general
case, it is difficult to directly convert this picture into mathematical equations. However, this
is possible for the cases where one of the processes is prevailing. When neglecting tunneling
processes, the approach developed in this chapter, which regards the samples under
discussion as a random mixture of the high-resistive PM and the low-resistive FM phases (or
regions), is able to describe all the peculiarities observed.
Thus, transport, magnetic and magnetoresistance properties of La0.7Ca0.3MnO3
nanocrystalline samples have been systematically studied. It is shown that temperature
dependences of both electric resistivity and magnetoresistance exhibit peaks near Curie
temperature TC 266 K. In addition, a broad resistivity peak, which is not accompanied with
any anomaly on (T) and MR(T) curves, is observed at TLT 230 K. Relatively low value of
the saturation magnetization at 5 K suggests that FM metallic phase does not occupy the
whole sample. Current-voltage I(V) characteristics are linear over more than two orders of V
at T < TC, in rather contrast to the picture when intergrain tunneling is the dominant
conduction mechanism. The results are well explained in terms of an inhomogeneous
transport through a mixture of high-resistive PM regions separated by more conducting FM
regions, where increased amount of the PM phase is provided by the expanded intergrain
area.
PECULIAR FEATURES OF ELECTRIC PROPERTIES

OF ULTRATHIN EPITAXIAL FILMS
This section analyzes features of the transformation of magnetic and transport properties
of epitaxial manganite films upon a decrease in the film thickness. It is shown that the
anomalous resistive behavior of very thin films may result from the degradation of magnetic
transition and development of inhomogeneous magnetic state consisting of PM high-resistive
and FM metallic phases coexisting over a wide temperature region. Theoretical calculations
performed within the frames of the approach developed in this chapter are concluded to well
describe the main features of the properties of the objects under consideration.
Recent investigations conducted on epitaxial films of doped manganites have shown that
these films have degraded magnetic and electrical properties near a substrate-film interface
[12,13,71]. Depending on substrate material, interface roughness and lattice mismatch
between a film and substrate, the thickness of the dead layer was estimated to vary from 1
to 11 nm at T = 77 K. Furthermore, a thin surface layer with degraded magnetic properties
was revealed by spin-polarized photoemission measurements carried out on La0.67Sr0.33MnO3
epitaxial film [72]. In other work, Blamire et al. [73] concluded that magnetic state of
ultrathin (5 6 nm) La0.7Ca0.3MnO3 epitaxial films can be modeled on the basis of a high
density of small magnetically ordered regions separated by a matrix of disordered spins. All
these observations show that even the properties of high quality epitaxial films are far from
those of ideal homogeneous systems, and at least the magnetic inhomogeneity should be
taken into account to correctly analyze the behavior of manganite systems.
Izumi et al. [13] reported on the growth, structure, and properties of La0.6Sr0.4MnO3 thin
films fabricated on SrTiO3 (001) substrates using pulsed laser deposition. Such analyses as
reflection high-energy electron diffraction, coaxial impact collision ion scattering
spectroscopy, atomic force microscopy, and four-circle x-ray diffraction were carried out to
characterize the surface and crystal structure of the films. By atomic force microscopy, the
surface of resulting films was confirmed to be extremely flat, showing atomically smooth
terraces and 0.4 nm high steps. The direction and spacing of the steps were the same as those
for the original substrate. The authors concluded that perfect layer-by-layer growth took place
for the films used for further investigation. Transport and magnetic properties were compared
for the two films with the thicknesses of 100 and 4 nm (denoted here as films A and B,
respectively). The thick film showed saturation magnetization Ms of about 3.5 B/Mn atom at
5 K and relatively sharp change in magnetization near TC = 310 K. The properties of a 4-nm-
thick film were considerably different both reduced Ms at 5 K and broadened magnetic
transition were characteristic of the film. Two bumps, one near TC and another one at T
0.48TC, were observed on resistivity vs. temperature curve. It was noted by Izumi et al. that
such modification of the properties might be attributed to interface and/or surface effects but
the detailed analysis of the phenomena was not performed.
Let us analyze the results by Izumi et al. from the viewpoint of the approach described
above. At T > TC the resistance of the film B can be well fitted by the expression:
RpB(t) = 9.710-4texp(3.6/t). (8)
A fraction of the PM phase persisting in the film B below TC can roughly be estimated from
the ratio between the magnetization of the film B (MB) and that of the film A (MA). Taking
into account that Ms of the thick film is close to the theoretical value (3.6 B/Mn for the given
composition) and that magnetic transition is sufficiently sharp in the film A, it is reasonably to
assume that below TC the whole volume of the film A is in FM state (corresponds to 0 = 0
and d << 1 in Eq (6)). Since the magnetization of the PM phase is sufficiently smaller than
that of the FM one, a ratio of MB/MA corresponds to the fraction of the FM phase in the film
B, and the temperature dependence of the PM fraction can be expressed as
pB(t) = 1 - MB(t)/MA(t) (t < 1). (9)
Inset of Figure 19 shows the dependence of pB vs. T/TC calculated from the experimental
data of Ref. [5] (circles) and that fitted according to Eq. (6) (solid line):
pB(t) = 0.5 + 0.5exp((t - 1)/0.2) (10)
As a next step, let us estimate the critical volume fraction of the FM phase at which
percolation occurs. There is good reason to think that, as temperature decreases, fB(t) reaches
its critical value fc at a temperature, at which a sharp drop in resistance is observed (shown
by bold arrow in Figure 19). The value of fc determined in such a way is equal to 0.46.
It was noted above that, to calculate the total resistance of the two-phase system, we used
the formula (4). To correctly apply this formula, the following condition has to be satisfied:
Figure 19. Resistance vs. T/TC for the 4-nm-thick La0.6Sr0.4MnO3 epitaxial film (circles experimental
data adopted from Ref. [5], solid line calculated as described in text). Inset: temperature dependence
of the fraction of PM phase p for the same film (circles calculated from Ref. [5], solid line fitted
according to Eq. (6)).
Rf << Rp.
It follows from the experimental data by Izumi et al. that this condition is not valid for
the film B. Nevertheless, our calculations indicate that over the temperature region where two
phases coexist, the total resistance is largely determined by the value of Rp and is weakly
sensitive to the specific form of Rf. Thus, we have put Rf of the film B in the following form:
RfB(t) = 10-6(1 + t2). (11)
As can be seen from Figure 19, such choice of RfB(t) leads to a discrepancy between the
experimental data (circles) and calculated ones (solid line) at very low temperatures.
However, at intermediate and high temperatures, the behavior of the total resistance derived
from Eqs. (4), (8) (11) well reflects the main features of the resistance of the film B.
Thus, we have analyzed the transport and magnetic behavior of La0.6Sr0.4MnO3 ultrathin
epitaxial films. It is shown that over a wide temperature region, the state of the film can be
well modeled as a mixture of the PM high-resistive and FM metallic phases. In this state, the
resistance is preferentially determined by magnetically disordered regions and does not reflect
the behavior of magnetization which measures an average spin alignment. Calculations
performed in the framework of the approach developed in this chapter have been shown to
well describe the main features of magnetic and transport properties of the film under
consideration.
ANALYSIS OF EXISTENT APPROACHES TO

DESCRIBE ELECTRIC PROPERTIES OF DOPED MANGANITES
It was stressed above that the properties of doped CMR manganites cannot be in full
measure understood without accounting for the effects induced by various kinds of
imperfections (grain boundaries, chemical inhomogeneities, strained regions, etc.). It is this
topic that has been under intensive investigation in a number of recent papers.
The role of intergrain boundaries in the formation of the properties of doped CMR
manganites has been investigated in detail in papers [20,45,46,70,74,75]. There has been
ample experimental evidence of the perturbations of deformation fields and variations in
stoichiometry, induced by boundaries. So, Chisholm et al. [76] calculated stress distribution
near a low-angle intercrystallite boundary and showed that the region within which the
deformation exceeds 1 % extends at least a few nanometers (4.4 nm for the 5-angle
boundary). Browning et al. [77] came to a conclusion that the peculiar features of intergrain
boundaries in perovskites include the variations of oxygen stoichiometry over the area
exceeding several nanometers. Recently, Gross et al. presented strong arguments in favor of
the important role of the effects of energy band bending and charge depletion in the formation
of electric and magnetoresistive parameters of the boundaries in doped manganites [46].
It is noteworthy that hydrostatic pressure, which causes a 1 %-deformation in
La2/3Ca1/3MnO3- for example, gives rise to the change in the Curie temperature TC 10 K
[78]. At the same time, the minor changes in the parameter lead to TC 15 K [79]. For this
reason, the local values of the Curie temperature and magnetization in the vicinity of grain
boundaries strongly differ from those characteristic of a homogeneous phase. What is more,
the perturbations grow with the rise in the angle between crystalline lattices on different sides
of the boundary [20].
Similar effects are expected to result from other defects of crystalline structure, strained
regions, chemical inhomogeneities, interfaces between a film and substrate, etc. [20,71,80].
So, even a high-quality single crystal displays a low-field MR component near TC, which is
thought to originate from heterogeneity caused by local clustering [20,70]. Thus, the
enhancement in the defectiveness of doped manganite materials gives inevitably rise to the
increase in their magnetic inhomogeneity. At the same time, the characteristic size of the
regions where magnetic parameters deviate from their average values exceeds, as a rule, the
dimensions of the defects themselves. Accounting for a strong interplay between magnetic
and electric parameters in manganites, a similar picture is also characteristic of the
distribution of conductivity. Moreover, in this case one should expect even stronger
inhomogeneity because, in addition to the indirect influence through the fluctuations of
magnetization, the defects will serve as additional centers for the charge carrier scattering.
As indicated above, there has been clear evidence that the inhomogeneity of magnetic
and electric parameters is an important integral part of all the samples of doped manganites
(including high-quality single crystals), and it has to be taken into account while considering
the processes of charge transfer in such materials. Let us analyze the theoretical approaches
which include the consideration of the effects induced by various kinds of imperfections.
Paper [81] reports the approach whose idea is very close to that developed in this chapter.
Within the frames of the effective medium theory, Song et al. calculated the resistance of a
two-phase system which is a mixture of the high-resistive PM and low-ohmic FM phases. The
calculations are based on an assumption that a manganite sample is not characterized by a

single Curie point; there is a Gauss distribution of the TC around an average value. This model
was shown to fit very well to the measured resistance-temperature curves for polycrystalline
La0.7Ca0.3MnO3. Unfortunately, the authors did not consider the case of a highly broadened
magnetic transition. It is in the latter case that the temperature dependences of resistance
should necessarily exhibit the LT-peak.
The model developed by Yuan et al. [82] is close in concept to the preceding one but
differs from it in a computational approach used. The authors employ a random resistor-
network model in which FM metallic particles with the number density p randomly fill the
space of a sample. The computations are carried out with the use of the Monte Carlo method.
The authors obtained the results close to those we described in this chapter, but they did not
carry out the detailed analysis of the data.
The different sources and types of defectinduced spin disorder are discussed in paper
[20] and an assessment is made of transport phenomena that contribute to the
magnetoresistive response. The authors succeeded in an isolation of the intergrain-boundary-
induced contributions to electric- and magnetoresistance. They concluded that spinpolarized
tunnelling and spin scattering at interfaces play a relatively minor role, the dominant low
field contribution to magnetoresistance coming from the response of mesoscale magnetic
inhomogeneity induced by the grain boundary. A successful description of magnetoresistance
as a function of magnetic field is achieved in the paper. However, the temperature
dependence of electric resistance was not analyzed in this work. It should be noted that
experimental results obtained in Ref. [20] correlate with those we reported in this chapter:
although the low-field magnetoresistance changes monotonously with temperature, the R vs.
T curves exhibit a peak below TC. The authors analyze alternative approaches to describe
electronic transport in magnetically inhomogeneous manganites and show that such
approaches are unable to describe the whole complex of experimental data, but usually reflect
only one side of the phenomena observed.
The grain-size-dependent transport properties in the granular perovskite La0.85Sr0.15MnO3
were investigated by Zhang et al. [16]. The authors show that the decrease in a grain size
from ~1 m to 50 nm results in both the magnetization reduction and magnetic transition
broadening. The LT-peak was observed to emerge when the grain size becomes less than 190
nm. Zhang et al. developed an approach to describe such behavior of electric resistance. Their
model is based on following assumptions: (a) the conductivity of a grain core differs from
that of a surface layer; (b) the Curie temperature of the latter is much lower than that of the
former; (c) the mechanism of charge transfer through a surface layer is a spin-dependent
tunneling. The authors succeeded in a satisfactory description of electric and
magnetoresistance properties of the samples with nanoscale grain size. However, they do not
provide evidence for the statement that it is the tunneling that is the dominant charge transfer
mechanism through the surface layer.
The approaches developed in Refs. [7,9,83,84,85] deal with the two-phase systems, for
which the conductivity of each the phases depends on magnetic state and is of activated
character for the case of non-ferromagnetic ordering. However, these works mainly focus on
the elucidation of the nature of the resistivity peak near TC, on the description of the behavior
of magnetoresistance or concentration dependence of electric resistance. So, the author of
paper [85] admits the existence of magnetically multiphase state in perovskite manganites and
considers magnetoresistive properties of a mixture of FM metallic and antiferromagnetic
semiconducting phases. The model developed makes it possible to calculate the

magnetoresistance of the system for the cases where one of the phases is prevailing. However,
the author did not perform the analysis of the temperature dependences of electric resistance.
As a result, the approach we developed in this chapter is based on the ideas which are
close to those suggested by other authors, but it also covers the case of a strongly broadened
magnetic transition, i.e. the topic which has not been considered by them.
Taking into account a considerable interest of researchers in understanding the CMR
effect in doped manganites, a number of experimental and theoretical works have been
devoted to the investigation of these materials (see review papers [4,5,7]). However, the
theoretical works have largely been aimed at the development of the approaches enabling a
description of particular behavioral features of these materials. In most cases, the ranges of
applicability of these approaches to describe other physical properties have been analyzed
superficially. Review and critical analysis of the available views on the nature of unusual
properties of manganites, as well as a comparison of the conclusions of the theories with the
experiment has been made in Refs. [7,20,46,70,86].
In general, the approaches, developed for a description of electric and magnetoresistive
properties of doped manganites with account for inhomogeneity of magnetic and electric
parameters, can be divided into three groups according to the dominant mechanism of charge
transfer through the high-resistive regions:
1. Spin-polarized tunneling of charge carriers through weakly conducting barriers

(intergrain boundaries, defect-induced regions of degraded magnetic ordering and
depressed conductivity, and others).
2. Spin-dependent scattering of polarized charge carriers (a scattering of conductivity
electrons on phonons or impurities into corresponding available states of spin-
polarized energy bands).
3. Activated charge transfer through the weakly conductive regions.
A critical analysis of the ranges of applicability of each of the above models has been
made in Refs. [7,20,46,70,86,87]. The authors have come to the conclusion that each of the
above mechanisms may occur in manganite materials, but the question of which of these
mechanisms is dominant is not easy to be answered. For this reason, in each particular case,
one should perform a comprehensive analysis of the behavior of a sample, find a prevailing
charge transfer mechanism, and only after that it may become possible to reliably predict the
regularities of the behavior upon a change of one or a few parameters.
Let us illustrate such conclusion with the results obtained in work [64]. Fth et al. used
scanning tunneling spectroscopy to investigate single crystals and thin films of La1-xCaxMnO3
with x near 0.3. Below the bulk transition temperature TC, a phase separation picture was
observed, where inhomogeneous structures of metallic and insulating areas coexisted and
were strongly field dependent in their size and structure. Insulating areas were found to persist
far below TC. In this case, the charge transfer through the high-resistive regions can occur by
a combination of spin-scattering and inelastic tunneling processes (including multi-step
tunneling via localized states), with the relative contribution depending on a number of
factors, including field and temperature. It is almost impossible to convert this picture into a
system of mathematical equations and, thus, one should look for adequate approaches, one of
which is developed in this chapter.
CONCLUSION
The chapter focuses on the nature of low-temperature resistive peak, which is often
observed in colossal magnetoresistant manganites below Curie temperature. Based on a
review of the experimental data available, it is concluded that the presence of the LT-peak is a
consequence of the enhanced inhomogeneity of manganite materials. An approach, which
regards the state of doped manganites as a mixture of interpenetrating paramagnetic insulating
and ferromagnetic metallic phases, is shown to be able to successfully describe the
peculiarities of the temperature dependence of total resistivity. It is shown that in spite of the
monotonic variation of magnetic and electric properties of each the phase, the resulting
resistance may exhibit resistive anomaly (the LT-peak) within the region of the phase
coexistence. Such an approach is concluded to be applicable to the description of the
magnetic and electric properties of a number of inhomogeneous and phase-separated
manganite systems, including micro- and nanocrystalline samples, ultrathin epitaxial films,
etc.
ACKNOWLEDGMENTS
The author expresses deep gratitude to all those who contributed to the works underlying
this chapter. These include: G. Bondarkova, S. Cherepov, I. Lezhnenko, A. Matviyenko, D.
Podyalovskii, A. Pogorily, A. Vovk (Institute of Magnetism); A. Belous, E. Pashkova, O.
Vyunov, O. Yanchevskii (V.I. Vernadsky Institute of General and Inorganic Chemistry); V.
Borodin, V. Doroshev, V. Kamenev, A. Mazur, and T. Tarasenko (O.O. Galkin Institute of
Physics and Engineering).
REFERENCES
[1] Helmolt, R.; Wecker, J.; Samwer, K.; Haupt L.; Brner, K. J. Appl. Phys. 1994, 76,
6925 6928.
[2] Jin, S.; Tiefel, T.H.; McCormack, M.; Fastnacht, R.A.; Ramesh, R. Science 1994, 264,
413 415.
[3] Zener, C. Phys. Rev. 1951, 82, 403 405.
[4] Haghiri-Gosnet, A.-M. and Renard, J.-P. J. Phys. D: Appl. Phys. 2003, 36, R127
R150.
[5] Drr, K. J. Phys. D: Appl. Phys. 2006, 39, R125 R150.
[6] Yoo, Y.-K.; Duewer, F.; Yang, H.; Yi, D.; Li, J.-W.; Xiang, X.-D. Nature 2000, 406,
704 708.
[7] Dagotto, E.; Hotta, T.; Moreo, A. Phys. Rep. 2001, 344, 1 153.
[8] Tovstolytkin, A.; Pogorily, A.; Vovk, A.; Podyalovskii, D.; Lezhnenko, I.; Matviyenko,
A. J. Magn. Magn. Mater. 2004, 272-276, 1839 1840.
[9] Cen, C.; Alvarez, G.; Dagotto, E. Phys. Rev. Lett. 2007, 98, 127202 (1 4).
[10] Verges, J.A.; Martin-Mayor, V.; Brey, L. Phys. Rev. Lett. 2002, 88, 136401 (1 4).
[11] Kumar, S.; Mayumdar, P. Phys. Rev. Lett. 2006, 96, 016602 (1 4).
[12] Tovstolytkin, A.I. Mater. Science Forum 2001, 373-376, 613 616.
[13] Izumi, M.; Konishi, Y.; Nishihara, T.; Hayashi, S.; Shinohara, M.; Kawasaki, M.;
Tokura, Y. Appl. Phys. Lett. 1998, 73, 2497 2499.
[14] Tovstolytkin, A.I.; Pogorily, A.N.; Kovtun, S.M. Low. Temp. Phys. 1999, 25, 962
965.
[15] Arulraj, A.; Manesh, R.; Subbara, G.N.; Mahencliran, R.; Raychaundhure, A.K.; Rao,
C.N.R. J. Sol. St. Chem. 1996, 127, 87 91.
[16] Zhang, N.; Ding, W.; Zhong, W.; Xing, D.; Du, Y. Phys. Rev. B. 1997, 56, 8138
8142.
[17] Troyanchuk, I.O.; Khomchenko, V.A.; Tovar, M.; Szymczak, H.; Brner, K. Phys. Rev.
B 2004, 69, 054432 (1 12).
[18] Millis, A.J.; Darling, T.; Migliori, A. J. Appl. Phys.1998, 83, 1588 1591.
[19] Hwang, H.Y.; Cheong, S-W.; Radaelli, P. G.; Marezio, M.; Batlogg, B. Phys. Rev. Lett.
1995, 75, 914 917.
[20] Evetts, J.E.; Blamire, M.G.; Mathur, N.D.; Isaac, S.P.; Teo, B.-S.; Cohen, L.F.;
MacManus-Driscoll, J.L. Philos. Trans. Roy. Soc. Lond. A 1998, 356, 1593 1615.
[21] Tovstolytkin, A.I.; Pogorily, A.N.; Cherepov, S.V.; Bondarkova, G.V.; Silantiev, V.I.
Metallofizika i Noveishie Tekhnologii 2000, 22, 23 31.
[22] Sun, J.R.; Rao, G.H.; Liang, J.K. Appl. Phys. Lett. 1997, 70, 1900 1902.
[23] Pogorilyi, A.N.; Belous, N.A.; Tovstolytkin, A.I.; Belous, A.G.; Yanchevski, O.Z. Low
Temp. Phys. 1999, 25, 74 75.
[24] Belous, A.G.; Vyunov, O.I.; Pashkova, E.V.; Yanchevskii, O.Z.; Tovstolytkin, A.I.;
Pogorily, A.N. Inorg. Mater. 2003, 39, 161 170.
[25] Urushibara, A.; Moritomo, Y.; Arima, T.; Asamitsu, A.; Kido, G.; Tokura, Y. Phys.
Rev. B 1995, 51, 14103 14109.
[26] Surthi, S.; Kotru, S.; Pandey, R.K.; Fournier, P. Sol. St. Commun. 2003, 125, 107 110.
[27] Mandal, P.; Das, S. Phys. Rev. B. 1997, 56, 15073 15080.
[28] Tang, Y.; Shaltout, I.; Braunstrin, R.; Helmolt, R.; Haupt, L.; Brner, K. Phys. Stat.
Sol. (b) 1994, 182, 509 519.
[29] Steenbeck, K.; Eick, T.; Kirsch, K.; Schmidt, H.-G.; Strinbei, E. Appl. Phys. Lett.
1998, 73, 2506 2508.
[30] Huang, Y.-H.; Xu, Z.-G.; Yan, C.-H.; Wang, Z.-M.; Zhu, T.; Liao, C.-S.; Gao, S.; Xu,
G.-X. Solid State Commun. 2000, 114, 43 47.
[31] Archibald, W.; Zhou, J.-S.; Goodenough, J.B. Phys. Rev. B 1996, 53, 14445 14449.
[32] Chechersky, V.; Nath, A.; Ju, H.; Greene, R.L. Low Temp. Phys. 1997, 23, 536 540.
[33] Savosta, M.M.; Doroshev, V.D.; Kamenev, V.I.; Borodin, V.A.; Tarasenko, T.N.;
Mazur, A.S.; Maryko, M. JETP 2003, 97, 573 586.
[34] Tovstolytkin, A.I.; Pogorily, A.M.; Dzhezherya, Yu.I.; Dzyublyuk, V.V.; Mapps, D.J.
J. Phys.: Condens. Matter. 2009, 21, 386003 (1 7).
[35] Simopoulos, A.; Kallias, G.; Devlin, E.; Panagiotopoulos, I.; Pissas, M. J. Magn.
Magn. Mater. 1998, 177-181, 860 861.
[36] Rubinstein, M.; Gillespie, D.J.; Snyder, E.J.; Tritt, M.T. Phys. Rev. B 1997, 56, 5412
5423.
[37] Doroshev, V.D.; Borodin, V.A.; Kamenev, V.I.; Mazur, A.S.; Tarasenko, T.N.;
Tovstolytkin, A.I.; Trukhanov, S.V. J. Appl. Phys. 2008, 104, 093909 (1 13).
[38] Snyder, G. J.; Hiskes, R.; DiCarolis, S.; Beasley, M.R.; Geballe, T.H. Phys. Rev. B
1996, 53, 14434 14444.
[39] Worledge, D.C.; Snyder, G.J.; Beasley, M.R.; Geballe, T.H.; Hiskes, R.; DiCarolis, S.
J. Appl. Phys. 1996, 80, 5158 5161.
[40] Chen, C.C.; De Lozanne, A. Appl. Phys. Lett. 1998, 73, 3950 3952.
[41] Zhang, L.W.; Gong, F.; He, L.; Cao, B.S.; Zhu M.; Zhao Y.G. J. Magn Magn. Mater.
2000, 219, 236 240.
[42] McLachlan, D.S. J. Phys. C: Sol. St. Phys. 1987, 20, 865 877.
[43] Gorkov, L.P. Physics-Uspekhi 1998, 41, 589 594.
[44] Mitchell, J.F.; Argurion, D.N.; Potter, C.D.; Hinks, D.G.; Jorgensen, J.D.; Bader, D.
Phys. Rev. B 1996, 9, 6172 6183
[45] Gupta, A.; Gong, G.Q.; Xiao, G.; Duncombe, P.R.; Lecoeur, P.; Trouilloud, P.; Wang,
Y.Y. Phys. Rev. B. 1996, 54, R15629 R15632.
[46] Gross, R.; Alff, L.; Bchner, B.; Freitag, B. H.; Hfener, C.; Klein, J.; Yafeng, Lu,
Mader, W.; Philipp, J. B.; Rao, M. S. R. J. Magn. Magn. Mater. 2000, 211, 150 159.
[47] Hwang, H.Y.; Cheong, S.-W.; Ong, N.P.; Batlogg, B. Phys. Rev. Lett. 1996, 77, 2041
2044.
[48] Mendelson, M.I. J. Amer. Cer. Soc. 1969, 52, 443 446.
[49] Sanchez, R.D.; Rivas, J.; Vasquez-Vasquez, C.; Lopez-Quintela, A.; Causa, M.T.;
Tovar, M.; Oseroff, S. Appl. Phys. Lett. 1996, 68, 134 136.
[50] Savosta, M.M.; Kamenev, V.I.; Borodin, V.A.; Novak, P.; Maryko, M.; Hejtmanek, J.;
Drr, K.; Sahana, M. Phys. Rev. B 2003, 67, 094403 (1 9).
[51] Savosta, M.M.; Doroshev, V.D.; Kamenev, V.I.; Borodin, V.A.; Tarasenko, T.N.;
Mazur, A.S.; Maryko, M. JETP 2003, 97, 573 586.
[52] Borodin, V.A.; Doroshev, V.D.; Kamenev, V.I.; Mazur, A.S.; Tarasenko, T.N. High-
Pressure Physics and Technology 2007, 17, 52 64.
[53] Tovstolytkin, A.I.; Pogorily, A.M.; Podyalovskii, D.I.; Kalita, V.M.; Lozenko, A.F.;
Trotsenko, P.O.; Ryabchenko, S.M.; Belous, A.G.; V'yunov, O.I.; Yanchevskii, O.Z. J.
Appl. Phys. 2007, 102, 063902 (1 7).
[54] Markovich, V.; Rozenberg, E.; Gorodetsky, G.; Greenblatt, M.; McCarroll, W.H. Phys.
Rev. B 2001, 63, 054423 (1 4).
[55] Batchelor, G.K.; OBrien, R.W. Proc. R. Soc. Lond. A 1977, 355, 313 318.
[56] Mizusaki, J.; Waragai, K.; Tsuchiya, S.; Tagawa, H.; Arai, Y.; Kuwayama, Y. J. Amer.
Cer. Soc. 2005, 79, 109 116.
[57] Zinchenko, A.Z. Phil. Trans. R. Soc. Lond. A 1998, 356, 2953 2978.
[58] Gonzo, E.E. Chem. Eng. J. 2002, 90, 299 306.
[59] Kupkova, M.; Kupka, M. Metalurgija 2004, 43, 97 102.
[60] Sheng, P. Phys. Rev. B 1980, 21, 2180 2183.
[61] Balberg, I. Phys. Rev. Lett. 1987, 59, 1305 1307.
[62] Li, X.W.; Gupta, A.; Xiao, G.; Gong, G.Q. Appl. Phys. Lett. 1997, 71, 1124 1126.
[63] Pogorily, A.N.; Tovstolytkin, A.I.; Vovk, A.Ya.; Bondarkova, G.V.; Silantiev, V.I.;
Cherepov, S.V.; Belous, A.G.; Yanchevski, O.Z. Mater. Sci. Forum 2001, 373-376,
621 624.
[64] Fth, M.; Freisem, S.; Menovsky, A.A.; Tomioka, Y.; Aarts, J.; Mydosh, J.A. Science
1999, 285, 1540 1542.
[65] Matveev, V.V.; Ylinen, E.; Zakhvalinskii, V.S.; Laiho, R. J. Phys.: Cond. Matt. 2007,
19, 226209 (1 4).
[66] Fita, I.M.; Szymczak, R.; Baran, M.; Markovich, V.; Puzniak, R.; Wisniewski, A.;
Shiryaev, S.V.; Varyukhin, V.N.; Szymczak, H. Phys. Rev. B 2003, 68, 014436 (1 7).
[67] Trukhanov, S.V.; Troyanchuk, I.O.; Trukhanov, A.V.; Fita, I.M.; Vasilev, A.N.;
Maignan, A.; Szymczak, H. JETP Lett. 2006, 83, 33 38.
[68] Kozlenko, D.P.; Ovsyannikov, S.V.; Shcheninnikov, V.V.; Voronin, V.I.; Savenko,
B.N. JETP Lett. 2007, 85, 203 209.
[69] Balcells, Ll.; Martinez, B.; Sandiumenge, F.; Fontcuberta, J. J. Magn. Magn. Mater.
2000, 211, 193 199.
[70] Gupta, A.; Sun, J.Z. J. Magn. Magn. Mater. 1999, 200, 24 43.
[71] Liang, S.; Sun, J.R.; Wang, J.; Shen, B.G. Appl. Phys. Lett. 2009, 95, 182509 (1 3).
[72] Park, J.-H.; Veskovo, E.; Kim, H.-J.; Kwon, C.; Ramesh, R.; Venkatesan, T. Nature
1998, 392, 794 796.
[73] Blamire, M.G.; Teo, B.-S.; Durrell, J.H.; Mathur, N.D.; Barber, Z.H.; MacManus-
Driscoll, J.L.; Cohen, L.F.; Evetts, J.E. J. Magn. Magn. Mater. 1999, 191, 359 367.
[74] Fontcuberta, J.; Bibes, M.; Martnez, B.; Trtik, V.; Ferrater, C.; Snchez, F.; Varela, M.
J. Magn. Magn. Mater. 2000, 211, 217 225.
[75] Wang, X.L.; Dou, S.X.; Liu, H.K.; Ionescu, M.; Zeimetz, B. Appl. Phys. Lett. 1998, 73,
396 398.
[76] Chisholm, M.F.; Pennycook, S.J. Nature 1991, 351, 47 49.
[77] Browning, N.D.; Chisholm, M.F.; Pennycook, S.J.; Norton, D.P.; Lowndes, D.H. Phys.
C.: Supercond. 1993, 212, 185 190.
[78] Laukhin, V.; Fontcuberta, J.; Garca-Muoz, J.L.; Obradors, X. Phys. Rev. B 1997, 56,
R10009 R10012.
[79] Malde, N.; de Silva, P.; Hossain, A.; Cohen, L.F.; Thomas, K.A.; MacManus-Driscoll,
J.L.; Mathur, N.D.; Blamire, M.G. Solid State Commun. 1998, 105, 643 648.
[80] Kovalenko, V.F.; Tychko, A.V. Functional Materials 2004, 11, 521 527.
[81] Song, H.; Kim, W.J.; Kwon, S.-J. J. Magn. Magn. Mater. 2000, 213, 126 134.
[82] Yuan, S.L.; Liu, M.H.; Li, Z.Y.; Peng, G.; Xia, Z.C.; Yang, Y.P.; Tu, F.; Zhang, G.Q.;
Liu, J.; Liu, L.; Tang, J.; Zhang, G.H.; Zhang, L.J.; Feng, W.; Xiong, C.S.; Xiong, Y.H.
Solid State Commun. 2002, 121, 291 294
[83] Nagaev, E.L. Phys. Rep. 2001, 346, 387 531.
[84] Tovstolytkin, A.I.; Pogorily, A.N.; Lezhnenko, I.V.; Matviyenko A.I.; Podyalovski,
D.I.; Kravchik, V.P. Phys. Sol. St. 2003, 45, 1952 1956.
[85] Krivoruchko, V.N. Low Temp. Phys. 1996, 22, 798 802.
[86] Ahn, K.H.; Lookman, T.; Bishop, A.R. Nature 2004, 428, 401 404.
[87] Shenoy, V.B.; Rao, C.N.R. Philos. Trans. Roy. Soc. Lond. A 2008, 366, 63 82.
Chapter 4
STRUCTURAL, MAGNETIC AND

ELECTRON TRANSPORT PROPERTIES OF
ORDERED-DISORDERED PEROVSKITE COBALTITES
Asish K. Kundu1 and B. Raveau2

1
Indian Institute of Information Technology Design & Manufacturing, Dumna Airport
Road, Jabalpur 482005, India
2
Laboratoire CRISMAT, ENSICAEN UMR6508, 6 Boulevard Marchal Juin, Cedex 4,
Caen-14050, France
ABSTRACT
Rare earth perovskite cobaltites are increasingly recognized as materials of
importance due to rich physics and chemistry in their ordered-disordered structure for the
same composition. Apart from colossal magnetoresistance effect, like manganites, the
different forms of cobaltites exhibit interesting phenomena including spin, charge and
orbital ordering, electronic phase separation, insulator-metal transition, large
thermoelectric power at low temperature. Moreover, the cobaltites which display colossal
magnetoresistance effect could be used as read heads in magnetic data storage and also in
other applications depending upon their particular properties. The A-site ordered-
disordered cobaltites exhibit ferromagnetism and metal-insulator transitions as well as
other properties depending on the composition, size of A-site cations and various external
factors such as pressure, temperature, magnetic field etc. Ordered cobaltites, having a
112-type layered structure, are also reported to have an effectively stronger electron
coupling due to layered A-site cationic ordering. Most importantly for the present article
we focus on La-Ba-Co-O based ordered-disordered perovskite phases, which exhibit
interesting magnetic and electron transport properties with ferromagnetic transition, TC ~
177K, and it being the first member of lanthanide series. Zener double exchange
mechanism considered to be crucial for understanding basic physics of the ferromagnetic-
metallic phase, yet does not explain clearly the insulating-type phase. In terms of electron
transport the ferromagnetic-metallic or insulating/semiconducting states have been
discussed in the present article with different types of hopping model.
214 Asish K. Kundu and B. Raveau
INTRODUCTION
Last few decades, there has been extensive research on ABO3-type perovskite oxides of
the general formula Ln1-xAkxBO3 (Ln = trivalent lanthanide, Ak = divalent alkaline earth, B =
transition metal) [1]. Particularly the perovskite cobaltites were discovered in 1950s and the
magnetic ordering was first reported in 1960s [2], since then there are lots of interesting
phenomena reported in the literature. Some of the novel properties of the perovskite cobaltites
are known for some time, especially the crystal structure transformation, the paramagnetic
(PM) to ferromagnetic (FM) transition at Curie temperature (TC) and the associated insulator-
metal transition (TIM) and so on. The discovery of colossal magnetoresistance (CMR) in doped
manganites has renewed great interest in perovskite oxides since the early 90s [1]. In 1997,
large value of magnetoresistance was reported for newly discovered ordered cobaltite,
LnBaCo2O5.4 (Ln=Eu, Gd) known as layered 112-phases [3]. This leads to enhanced interest
because of their potential applications in improving magnetic data storage. Moreover, the
perovskite cobaltites have also attracted attention because of their possible applications as
oxidation catalysts, gas sensors, solid oxides fuel cells and oxygen separation membranes due
to their stability in a wide range of oxygen contents, high oxygen diffusivity and electron
conduction [4]. Consequently superconductivity was discovered in hydrated sodium cobaltite
phase in 2003, since then cobaltites have received even more attention [5]. Aside from
potential applications, the cobaltites exhibit rich phase diagram spanning a wide range of
magnetic properties and phenomena like charge ordering, orbital ordering, spin/cluster-glass
behavior, electronic phase separation etc [6-11]. These phenomena represent a combined
interaction between the spin, the lattice, the charge and the orbital degrees of freedom, which
will provide better understanding of strongly correlated electronic behavior. Such interactions
are manifested in single crystal, polycrystalline samples as well as in thin films. The
properties of perovskite cobaltites can be tuned either by external factors or by chemical
means. In certain critical range of cation doping, at the A-site of ABO3-perovskite, the rare
earth cobaltites exhibit simultaneous occurrence of ferromagnetism and metallicity, along
with CMR in the vicinity of TC or TIM [7].
Some of the perovskite-based cobaltites are known to exhibit electronic inhomogeneities
arising from the existence of more than one phase in crystals of nominally monophasic
composition. This is understood in terms of electronic phase separation described recently in
the literature [1]. Such a phenomenon occurs because of the comparable free energies of the
different phases. The phase-separated hole-rich and hole-poor regions give rise to anomalous
properties such as weak FM moments in an antiferromagnetic regime [1]. A variety of
magnetic inhomogeneities manifest themselves in Ln1-xAkxCoO3 depending on the various
factors such as the average radius of the A-site cation and size-disorder as well as external
factors such as temperature, magnetic field etc [6-11]. In the last few years electronic phase
separation in cobaltites has attracted considerable attention.
In this article, we discuss the A-site cationic ordering and disordering effects on magnetic
and electron transport properties for rare earth cobaltites. Perovskite cobaltites have two
possible forms of the A-site cations distribution depending on the type of cations or the
synthesis procedures [9, 12]. The first reported compounds on perovskite cobaltites are the A-
site disordered structure [2, 13], which have been investigated for last few decades, and the
other one is A-site ordered perovskites possessing a layered 112-type structure [3]. The latter
Structural, Magnetic and Electron Transport Properties 215
one discovered few years back, consists of oxide layers [LnO]-[CoO2]-[AkO]-[CoO2]

alternating along the c-axis [3, 9, 12, 14, 15]. The ordering of Ln3+ and Ak2+ ions is favorable
if the size difference is large between the A-site cations, hence smaller size Ln3+ and bigger
Ak2+ ions easily form a layered 112-structure and till date it is reported only for Ak2+ = Ba2+
cation i.e. LnBaCo2O5+ (0 1) series [3, 9, 12, 14-16]. In contrast, as the size difference
becomes smaller, as for example between La3+ and Ba2+, the disordered cubic perovskite
becomes more stable and special treatment is required to obtain the 112 ordered structure, as
reported recently for the order-disorder phenomena observed in the perovskites of the system
La-Ba-Co-O [12, 16]. The CoO2 layers in ordered cobaltites are free from the random
potential which would otherwise arise from the Coulomb potential and/or local strain via the
random distribution of A-site cations (Ln3+/Ak2+). Both ordered and disordered cobaltites
exhibit similar features like ferromagnetism, insulator-metal transition and magnetoresistance
in a certain temperature range, yet prominent differences are evidenced in their properties.
The ferromagnetism in these cobaltites is considered to be due to favorable Co3+ - O - Co4+
interactions. The effect of random distribution of A-site cations or disordered perovskite
structure arising from the chemical disorder as observed in the conventional random doping at
the A-site in ABO3-perovskite. For rare earth manganites, Attfield et al. [17] have reported the
role of A-site cationic radius, rA, and size disorder parameter, 2, on the TC and/or TIM with
variation of these parameters. In the present article we have also discussed the role of rA,
and 2 for disordered cobaltites Ln0.5Ak0.5CoO3+ (Ln = La, Nd, Gd and Ak = Ba, Sr) which
unusually influence the magnetic and electronic properties [13, 18, 19]. These interesting
phenomena are related to structural disordering caused by the substitution of Ak2+ ions in
place of Ln3+ and we have briefly presented the effect of size disorder 2 for Ln = Nd, Gd [18,
19]. The structural disorder in cobaltites is 3D, while a layered 2D structure is adopted by the
112-phase ordered cobaltites, LnBaCo2O5+. The crystallographic structures for perovskite
cobaltites with an integral number of oxygen ions per formula unit are well known, no
consensus has been reached for non-integer compounds. Basic knowledge of the
crystallographic structure of a compound is of particular importance for determining magnetic
structures, since these two properties are closely related. Hence we have presented first the
crystal structure description for both the (ordered-disordered) cobaltites before discussing the
physical properties.
CRYSTAL STRUCTURE
In general, the rare earth ordered-disordered cobaltites crystallize in the perovskites
structure with various types of superstructures also evidenced. The disordered ABO3-
perovskite is a simple cubic structure (Pm-3m) as shown in Figure 1. However, many
perovskite deviate a little from this structure even at room temperature. The perovskite
structure is most stable when the Goldschmidt tolerance factor, t, is unity (for cubic structure),
which is defined by, t = (rA + rO)/ 2 (rB + rO) where, rA, rB and rO are the average ionic radius
of the A, B and O ions respectively. Deviation of t from unity leads to the structural
distortion. For a small deviation in t (i.e. t < 1), the crystal structure changes from cubic to
tetragonal, rhombohedral or orthorhombic etc. In this situation the C-O-Co bond angle
decreases from 180. The perovskite structure occurs only within the range 0.75 t 1.00.
The stability of the perovskite structure of cobaltites depends on the relative size of the Ln/Ak
and Co ions in Ln1-xAkxCoO3. In disordered cobaltites Ln/Ak cation is surrounded by eight
corner sharing CoO6 octahedra, which build a 3D network. The smaller ionic radius of the
cations results in a lower value of t, consequently increasing the lattice distortion. The
increase in lattice distortion significantly decreases the C-O-Co bond angle from 180,
which strongly affects the physical properties of cobaltites. When t < 1, there is a
compression of the Co-O bonds, which in turn induces a tension on Ln-O bonds. A
cooperative rotation of the CoO6 octahedra and a distortion of the cubic structure counteract
these stresses. The rare earth cobaltites can be crystallized not only in cubic or orthorhombic
structure (as reported mostly), but also in tetragonal, hexagonal, rhombohedral, and
monoclinic structures as well. As a consequence, for the disordered cobaltite series
Ln0.5Ba0.5CoO3+ the crystal structure is reported as cubic or rhombohedral for Ln = La,
whereas for other lanthanides such as Pr, Nd, Gd, Dy the systems crystallize in the
orthorhombic structure with different space groups [13, 18]. The structural Rietveld analysis
for disordered Ln0.5Sr0.5CoO3 series reveals that the structure is rhombohedral for Ln = La, Pr
Nd and that for Gd, the structure is orthorhombic as reported for Ba-doped compounds [13].
B
O
A
Figure 1. Rare earth ABO3 perovskite with disorder structure.
Similarly, the layered 112-type ordered cobaltites LnBaCo2O5+ can be found in several
crystallographic symmetries at room temperature. The variation of oxygen stoichiometry (0
1) in ordered 112-type cobaltites lead to various structures as well as different cobalt and
oxygen coordinations such as pyramidal, octahedral and/or the mixing of both environments
for the Co-ions [14, 15]. In the following, the room temperature structures for different values
of will be discussed. The crystal structure of the stoichiometric LnBaCo2O5 (Ln = Eu, Gd,
Tb, Ho) cobaltite is tetragonal with P4/mmm space group (unit cell ap ap 2ap ; where ap is
defined as pseudo cubic cell parameters) [14, 15]. This corresponds to a doubling of the
original perovskite unit cell along the c-direction due to alternating BaO and LnO layers
(Figure 2). The layered structure is best observed in the most oxygen deficient case
LnBaCo2O5, because it is assumed that the oxygen ions are absent only in the Ln-layer [14,
15]. For = 0 the Co2+ and Co3+ ions (ratio 1:1) are all within square base pyramids formed
by their five oxygen neighbors.
(a) LnBaCo2O5 (b) LnBaCo2O5.5 (c) LnBaCo2O6

(ap ap 2ap) (ap 2ap 2ap) (ap ap 2ap)
Figure 2. Layered 112-type ordered LnBaCo2O5+ (0 1) cobaltites with (a) =0.0; LnBaCo2O5, (b)
=0.5; LnBaCo2O5.5 and (c) =1.0; LnBaCo2O6.
a)
b)
c)
Figure 3. Rietveld refinement of X-ray diffraction pattern for (a) disordered La0.5Ba0.5CoO3 (b) ordered
LaBaCo2O5.5 and (c) ordered LaBaCo2O6 (Taken from Ref. 12 & 16).
For non-stoichiometric cobaltites LnBaCo2O5+ (0 < <1) the crystal structure is more
complex: due to oxygen vacancy ordered super-structures can arise, which vary with oxygen
content. The oxygen content is strongly dependent on the size of the lanthanides [14]. Clearly,
a strong correlation exists between the Ln3+ radius and the amount of oxygen the compound
can accommodate. The first structural study of an ordered oxygen deficient perovskite of type
LnBaCo2O5+ was presented for the series Ln = Pr to Ho [14]. Although, the series has been
investigated in details for almost all lanthanide elements of the periodic table, interestingly
there was no such report on the first member of this series, i.e. for LaBaCo2O5.5. Very
recently we have reported LaBaCo2O5.5 and characterized by neutron diffraction, electron
microscopy and magnetic studies, showing that at room temperature the structure is 112-
layered orthorhombic ap 2ap 2ap supercell with Pmmm space group [16]. In contrast, for
Ln = Pr Ho, all of the x-ray diffraction patterns could be indexed using a tetragonal
structure ap ap 2ap with P4/mmm symmetry [14]. However, electron diffraction
measurements additionally revealed two kinds of superstructures, depending on the radius of
the lanthanide (or the oxygen content). For larger lanthanides (Pr, Nd, Sm, Eu, Gd and Tb) a
doubling of one lattice parameter is observed, corresponding to an orthorhombic Pmmm ap
2ap 2ap supercell, whereas for smaller lanthanides (Ho and Dy) a tripling of two lattice
parameters is noticed, as in a 3ap3ap2ap supercell. It was suggested [14] that ordering of
oxygen vacancies is at the origin of the observed superstructures. Importantly, the authors
found that the oxygen vacancies are located uniquely in the LnO layers (apical positions). By
changing the oxygen content the superstructures either vanish or change from one to the
other: Reducing oxygen content in GdBaCo2O5+ from = 0.4 to 0 destroys the
superstructure. On the other hand in HoBaCo2O5.3 an increase of from 0.3 to 0.4 leads to a
change in superstructure from 3ap3ap2ap to ap2ap2ap. High resolution electron
microscopy supported the assumption that the ordering of the oxygen vacancies is responsible
for the superstructures. Burley et al. and Pralong et al. [14] have reported NdBaCo2O5+ for
various oxygen stoichiometries. The = 0 compound has a tetragonal structure with ap ap
2ap supercell where the Nd layer incorporates all the oxygen vacancies. For the slightly
higher oxygen content = 0.38 the structure is derived from the = 0 one, but with oxygen
ions inserted randomly into the NdO layers. Oxygen vacancy ordering in b-direction resulting
in an orthorhombic ap 2ap 2ap supercell unit cell with Pmmm symmetry is reported for
= 0.5, in agreement with other reported structure [14]. The oxidized material with = 0.69 is
again described by a tetragonal ap ap 2ap supercell unit cell, but a very weak peak
originating from a doubling of the unit cell along the b-direction was detected.
Finally, for stoichiometric cobaltites LnBaCo2O6 all Co-ions (Co3+ and Co4+; ratio 1:1)
are in octahedral environment. In the LnBaCo2O6 series, the first member Ln = La could be
synthesized in the two forms, ordered and disordered as shown by neutron diffraction and
electron microscopy [9, 12]. At room temperature and in normal synthesis condition the
crystal structure is cubic for the disordered cobaltite, which means the La3+ and Ba2+ ions are
distributed randomly on the A-site. This is due to the smaller size differences between the
cations as discussed earlier. Nevertheless, the ordered 112-phases LaBaCo2O5.5 and
LaBaCo2O6 are obtained for the same composition in special synthesis conditions [12, 16],
which crystallize at room temperature in orthorhombic and tetragonal structures respectively.
Therefore, a layered 112-type ordered perovskite structure LnBaCo2O6 with a smaller
lanthanide also exist, again with a tetragonal unit cell ap ap 2ap with P4/mmm symmetry,
but require special synthesis conditions as reported by Pralong et al. [14]. To conclude the
ordered cobaltites phase, a summary of the most commonly used models for compounds with
oxygen content = 0, 0.5, 6 at room temperature is given (Figure 2):
1. LnBaCo2O5 has the tetragonal structure with ap ap 2ap supercell (P4/mmm

symmetry). Ln3+ layers alternate with BaO layers along the c-axis (oxygen vacancies
in Ln layer).
2. LnBaCo2O5.5 has an orthorhombic structure with ap 2ap 2ap supercell (Pmmm
symmetry). LnO0.5 layers alternate with BaO layers along the c-axis. Along the b-
direction the oxygen vacancies are ordered. This leads to an alternation of CoO5
pyramids and CoO6 octahedra along the b-direction.
3. LnBaCo2O6 is again tetragonal unit cell ap ap 2ap with P4/mmm symmetry. LnO
layers alternate with BaO layers along the c-axis (no oxygen vacancies in LnO layer).
We will now discuss in brief the various parameters to obtain different phases (ordered-
disordered) of La-Ba-Co-O cobaltites. The synthesis of LaBaCo2O5.5, maintaining the O5.5
stoichiometry and a perfect layered ordering of La3+ and Ba2+ cations, is delicate due to their
small size difference which favors their statistical distribution. Moreover, the larger size of
La3+ compared to other lanthanides allows large amounts of oxygen to be inserted, so that the
disordered La0.5Ba0.5CoO3 perovskite is more easily formed under normal conditions. Thus,
the successful synthesis of ordered phase requires several steps, using soft chemistry method,
and the strategy was to control the order-disorder phenomena in this system by means of two
synthesis parameters, temperature and oxygen partial pressure. In order to favour the ordering
of the La3+ and Ba2+ cations, the synthesis temperature was as low as possible, and
consequently a soft-chemistry synthesis route was used since it allows a high reactivity at low
temperature. However, this condition is not sufficient alone to achieve a perfect ordering of
these cations. The formation of La3+ layers is in fact favoured by the intermediate creation of
ordered oxygen vacancies, leading then to the 112-type layered non-stoichiometric cobaltites
LaBaCo2O5+, built up of layers of CoO5 pyramids between which the La3+ smaller than Ba2+
cations can be interleaved. In this process, it is rather difficult to control the oxygen
stoichiometry to O5.5. For this reason, the synthesis of the ordered LaBaCo2O6 phase was
carried out initially, using high purity argon gas, followed by annealing in an oxygen
atmosphere at specific temperature. Thereafter, the layered 112 cobaltite LaBaCo2O5.5 was
obtained from the ordered LaBaCo2O6 phase by employing temperature controlled oxygen
depletion method in inert atmosphere [16].
The X-ray powder diffraction (XRD) patterns of the three perovskite phases (Figure 3)
show their excellent crystallization. The disordered phase La0.5Ba0.5CoO3, (Figure 3a) and the
ordered phase LaBaCo2O6 (Figure 3c) exhibit sharp peaks, the patterns are indexed with the
cubic Pm-3m and tetragonal P4/mmm space groups, respectively [12]. The latter corresponds
to a doubling of the cell parameter along the c-axis related to the 1:1 ordering of the LaO/BaO
layers (Figure 2c). More importantly, the XRD pattern of the ordered LaBaCo2O5.5 (Figure
3b) is completely different from that observed for the ordered phase (Figure 3c) and for this
reason it is indexed and refined using the orthorhombic structure Pmmm space group [16].
The result for ordered-disordered cobaltites is slightly different from that obtained by
Nakajima et al. [9], who observed a smaller cell volume for the ordered phase (58.66 3/Co)
compared to the disordered one (58.77 3). Remarkably, the La/Ba ordering also involves a
slight deformation of the perovskite sublattice with a dilatation of the ap parameter within the
LaO/BaO layers and a compression along the LaO/BaO layers stacking direction.
Transmission electron microscopy (TEM) investigations on the disordered La0.5Ba0.5CoO3
and ordered LaBaCo2O6 cobaltites confirm their cubic and tetragonal structures respectively.
The selected area electron diffraction (SAED) patterns of the disordered cobaltite and the
corresponding high resolution electron microscopy (HREM) images (Figure 4a) are indeed
characteristic of a classical cubic perovskite (Pm-3m) with a ap 3.9 . For ordered
LaBaCo2O6, the reconstruction of the reciprocal space from the SAED patterns leads to a
tetragonal cell, with a ap and c 2ap, compatible with the space group P4/mmm. The
doubling of one cell parameter with respect to the simple perovskite cell is clearly observed
on the HREM image displayed in Figure 4b and on the corresponding SAED [100] zone axis
patterns (inset of Figure 4b). Beside the SAED investigation which gives a better view of the
microstructure, bright field images are also of great interest to obtain information about the
size of the 112-type domains, taking into consideration the possibility of twinning. In the case
of the disordered cubic cobaltite no twinning is observed, as expected from the symmetry. In
contrast, in the long-range ordered 112-perovskite LaBaCo2O6 twinning is clearly evidenced
(Figure 5b).
(a)
(b)
Figure 4. The HREM images and the corresponding SAED patterns for (a) disordered La0.5Ba0.5CoO3
and (b) ordered LaBaCo2O6 (Taken from Ref. 12).
b)
Figure 5. Bright field images for ordered cobaltites (a) LaBaCo2O5.5 and (b) LaBaCo2O6 (Taken from
Ref. 12 & 16).
Moreover, Rietveld refinements of the neutron powder diffraction (NPD) data of

LaBaCo2O5.5, shows a deviation from the ideal orthorhombic Pmmm structure due to partial
disorder of the oxygen atoms at the LaO0.5 layers. One indeed observes occupancy of about
90% and 10% of the O6 and O7 sites by oxygen, respectively, instead of the expected 100%
and 0% [16]. Therefore, an investigation of ordered LaBaCo2O5.5 cobaltite by TEM, in order
to identify possible secondary phase, superstructures, and/or particular microstructural
features is of particular interest. In the Figure 6a, SAED confirms that the observed spots can
be indexed with the Pmmm ap 2ap 2ap structure refined from powder diffraction.
However, two additional features are also evidenced: First, one observes twinned domains at
a microscale level, all over the investigated crystals regions, corresponding to the Pmmm ap
2ap 2ap structure (112-type I), second, in few areas of some investigated crystals, very weak
extra reflections are evidenced that cannot be indexed considering only the Pmmm ap 2ap
2ap structure. Considering the literature on the 112-type compounds, a part of these extra
spots can be indexed considering the Cmmm 2ap 4ap ap cell corresponding to the
LaBaMn2O5.5 compound [20]. This is in agreement with the results of the NPD refinements
where the occupancies obtained for the O6 and O7 atomic positions can be locally attributed
to a different vacancy/oxygen ordering (Figure 6b) leading notably to the existence of faulted
zones having a centered structural motif (112-type II). This tendency is evidenced by SAED
but also visible in HREM for instance looking at one of the directions (Figure 6a). In
summary, the LnBaCo2O5.5 structure consists mainly of 112 type I type domains (90%)
combined with 112 type II (manganites) domains (10%). Finally, it is worth mentioning that
no 3ap 3ap 2ap superstructure observed by TEM for ordered LaBaCo2O5.5 cobaltite, as
evidenced for HoBaCo2O5.5 and YBaCo2O5.44 compounds [14, 21].
Figure 6a. The HREM images and the corresponding SAED patterns for ordered LaBaCo2O5.5.
Figure 6b. Layered 112-type ordered LaBaCo2O5+ cobaltites with (a) =0.5; LnBaCo2O5,(type-I) (b)
=1.0; LnBaCo2O6 and (c) =0.5; LnBaCo2O5.5 (type-II) (Taken from Ref. 12 & 16).
MAGNETIC AND ELECTRON TRANSPORT PROPERTIES

There are two characteristic structural distortions which influence the physical properties
of doped rare earth cobaltites. The first one is cooperative tilting of the CoO6 octahedra which
is essentially established due to doping effect. This distortion is a consequence of the
mismatch of the ionic radius and various factors as discussed in the previous section. The
other distortion arises from the Jahn-Teller (JT) effect due to Co3+ ion, which distorts the
CoO6 octahedra in such a way that there are long and short Co-O bonds (Figure 7). This
occurs below a characteristic temperature for particular compounds, as for instance at 180K
for the disordered La0.5Ba0.5CoO3 cobaltite [8]. This is well understood in terms of crystal
field theory which describes how the d-electron of transition metal ions is perturbed by the
chemical environment. The most effective distortion is the basal plane distortion (called Q2
mode), with one diagonally opposite oxygen-pair displaced outwards and the other pair
displaced inward. It is well established that a JT distortion involving a displacement of
oxygen ions 0.1 can split the eg-band of the cobaltites (which forms the conduction band)
and opens a gap at the Fermi level. The magnitude of the crystal field splitting of d-orbital
determines whether the Co-ion occurs in the low-spin (LS), intermediate-spin (IS) or high-
spin (HS) configuration. Figure 7, shows a scheme of the band diagram of LaCoO3 to
elucidate how the JT distortion splits the conduction band and makes the material insulating.
The octahedral ligand environment around Co-ion splits the five d-orbitals into t2g-triplet (dxy,
dyz and dzx) and eg-doublet (dx2-y2 and dz2) state. In this system, the resulting crystal-field
splitting, cf, between t2g and eg orbital is around 2.06 eV as reported by Korotin et al. [22] for
theoretical observation, although experimentally obtained values are around 1.2 eV and 0.9eV
[22]. Further splitting of the eg orbitals due to the JT effect opens a gap at the Fermi level.
The intra-atomic exchange energy responsible for Hunds highest multiplicity rule, ex (or
JH), is smaller than cf i.e. ex < cf for Co3+ ion. Therefore Co-ions are always in low spin
state below 100 K for LaCoO3 [22]. This perturbation induced electronic spin-state transition
in rare earth cobaltites has been of great interest in recent years. The thermally driven spin-
state transition in cobaltites is a consequence of the subtle interplay between the crystal field
splitting (cf) and the Hunds coupling energy (ex). The cf usually decreases as the
temperature is increased, whereas ex is insensitive to temperature since it is an atomic
quantity. The spin-state of undoped LaCoO3 (Co3+ ion) exhibits a gradual crossover with
increasing temperature from the LS state (t2g6eg0; S = 0) to IS state (t2g5eg1; S = 1) at around
100 K and finally to HS state (t2g4eg2; S = 2) [22]. This results from the competition of the
crystal field with energy cf (t2g- eg splitting) and the interatomic (Hund) exchange energy
ex, leading to redistribution of electrons between t2g and eg levels.
Figure 7. Field splitting of the five-fold degenerate atomic 3d levels for perovskite cobaltites.
In Ln1-xAkxCoO3 the doping by Ak2+ cation induces the formal presence of Co4+, which
slightly reduces the JT distortion leading towards cubic structure. The JT distortion of CoO6
octahedra has been reported by Fauth et al. [8], which is favored due to the IS state of Co3+
(t2g5eg1) and Co4+ (t2g4eg1). The transfer interaction of eg-electrons is greater in rhombohedral
or pseudocubic phase than in the orthorhombic phase because the C-O-Co bond angle
becomes closer to 180. Clearly, the doping effect or the presence of Co4+ ion plays an
important role in this material to provide the FM and metallic behavior by suppressing the JT
distortion. The electron transport properties of cobaltites also depend on oxygen
stoichiometry, and belong to the class either of metals or of semiconductors. Doped cobaltites
are mixed valence materials, which mean that the Co-ions can carry different charges in the
same site. The ratio between different Co-ion configurations is also determined by the oxygen
stoichiometry as well as by the doping concentration, following the charge neutrality
condition. Changes in the crystallographic and magnetic structures as well as the other
physical properties can be induced by varying several parameters of either intrinsic
hole/electron (doping, oxygen content) or extrinsic nature (temperature, pressure, magnetic
field). Since only small changes of parameters (intrinsic/extrinsic) can cause a structural,
magnetic or electronic transition, hence a variety of contradictory models or results exist for
cobaltites. Detailed discussions are therefore required to clarify the influence of each
parameter on the physical properties of the perovskite cobaltites. In this article, special
emphasis has been given on the cationic ordering and oxygen stoichiometry onto the
crystallographic, magnetic and electronic properties of few particular cobaltite systems. The
magnetic structures, especially of the layered cobaltite systems with oxygen content of O5.5
per unit formula, are still debated for some of the reported compounds. The determination of
the magnetic structure is a complex task contrary to other materials, because the Co-ion in
these materials can be in different spin states as discussed earlier.
I. Disordered Cobaltites
Disordered rare earth cobaltites Ln1-xAkxCoO3 have been investigated for several years
due to their novel magnetic and electronic properties which include temperature-induced
spin-state transitions, cluster-glass like behavior, electronic phase separation,
magnetoresistance (MR) and so on [6-8, 11, 13]. The physical properties of perovskite
cobaltites are sensitively dependent on the doping concentration of the rare-earth site. Doping
brings up mixed valences in the Co-ions due to charge neutrality such as
(Ln3+Ak2+)(Co3+Co4+)O3. Therefore, substitution of Ln3+ by Ak2+ in Ln1-xAkxCoO3 will favor
the transformation of Co3+ into Co4+ in same ratio of doping, as a result Co3+ and Co4+ will
interact ferromagnetically obeying the Zener double-exchange (DE) mechanism [23]. The
simultaneous observation of ferromagnetism and metallicity in cobaltites is explained by this
mechanism, where the hopping of an electron from Co3+ to Co4+ via oxygen ion, i.e. where
the Co3+ and Co4+ ions exchange takes place. The integral defining the exchange energy in
such a system is non-vanishing only if the spins of the two d-orbitals are parallel. That is the
lowest energy of the system is one with a parallel alignment of the spins on the Co3+ and Co4+
ions. Due to this, the spins of the incomplete d-orbitals of the adjacent Co-ion are
accompanied by an increase in the rate of hopping of electrons and therefore by an increase in
electrical conductivity. Thus, the mechanism which leads to enhanced electrical conductivity
requires a FM coupling. On the other hand, Co3+-Co3+ and Co4+-Co4+ couple

antiferromagnetically due to super-exchange interactions. Super-exchange interaction
generally occurs between localized moments of ions in insulators. Goodenough et al. pointed
out that the FM interaction is governed not only by the DE interaction, but also by the nature
of the super-exchange interactions [23]. Whether the ferromagnetism in cobaltites (similar to
manganites) is mediated by a DE mechanism or not is clearly not understood at present.
However, the absence of half filled t2g orbitals is providing core spin and strong Hunds rule
coupling, unlike manganites, making this mechanism less feasible. It seems, the FM-metallic
phase in cobaltites can be explained by the Zener-DE mechanism whereas super-exchange
will fit for insulating state. Hence, there will be always a competition between these two
interactions to dominate one over another giving rise to a tendency of electronic phase
separation in the system [1]. The growth of interest in perovskite cobaltites is due to the
expectation that, in addition to the lattice, charge and spin degrees of freedom found in many
other transition metal oxides, the cobalt oxides also display a degree of freedom in the spin-
state at the Co-site.
(a)
(b)
15 Ln0.5Ba0.5CoO3-
Ln = La 0
10
Ln = Gd
(.cm)
Ln = Nd
10 -2
10
M (emu/g)
Ln = Nd
5 -4
10 Ln = La
Ln = Gd
-6
0 10
100 200 300 100 200 300 400
T(K) (c) T(K) (d)
Figure 8. Temperature dependent Magnetization and Resistivity for disordered cobaltites

Ln0.5Ak0.5CoO3 (Taken from Ref. 13 & 18).
The physical properties of the cobaltites are sensitive to doping concentration at the rare
earth site. Accordingly, the other parameters such as the average radius of the A-site cation,
rA,, and size-disorder parameter, 2, also vary due to doping at the perovskite A-site. These
parameters crucially control the physical properties of disordered cobaltites. Disordered
cobaltites of the type Ln0.5Ak0.5CoO3, especially those with Ln= La, Nd, Gd and Ak = Ba, Sr
are FM, many of them showing a metallic behavior as shown in Figure 8 [13, 18, 19]. These
properties arise because of the major influence of Co3+-O- Co4+ interactions in these
cobaltites. The FM TC increases with the increase in the size of the A-site cations, rA,. In the
case of Ak = Ba, ferromagnetism occurs when Ln = La (TC ~ 190 K) and Nd (TC ~ 130 K), but
for Ln = Gd, the material shows unusual magnetic behavior (Figure 8c). Furthermore,
Gd0.5Ba0.5CoO3 is an insulator and exhibits an electronic transition around 350 K with
cationic ordering [13], whereas La0.5Ba0.5CoO3 is metallic below FM TC (Figure 8d).
Gd0.5Ba0.5CoO3 which is charge-ordered at room temperature, shows a FM-like feature around
280 K, without reaching a saturation value of the magnetic moment or the highest moment
achieved is rather low even in the higher applied field conditions [13, 18]. The magnetic
transition around 280 K in Gd0.5Ba0.5CoO3, has been classified in the literature as FM or meta-
magnetic, and other cobaltites of this family such as La0.5(Nd0.5)Ba0.5CoO3 have not been
distinguished, and have all been treated as FM transitions. There are, however, considerable
differences amongst these cobaltites. The magnetic transitions in La0.5(Nd0.5)Ba0.5CoO3 are
distinctly FM, showing a sharp increase in magnetization at TC, and the value is rather low
(130-190 K) [13, 18]. Gd0.5Ba0.5CoO3, with a much smaller A-site cation, should have been
associated with an even lower TC. In Figure 9, we present the temperature variation of the
magnetization of a few compositions of Ln0.5Ak0.5CoO3. The TC values are plotted against
<rA> from the literature data, increasing up to <rA> value of 1.40 and decreasing thereafter.
The decrease in TC for <rA> 1.40 is likely to arise from the size-disorder. Indeed, the cations
size mismatch, 2, is known to play an important role in determining the properties of rare
earth cobaltites [13]. It appears that the large value of 2 in the Gd0.5Ba0.5CoO3 (0.033 2)
compared to that of La0.5Ba0.5CoO3 (0.016 2) could be responsible for the absence of

ferromagnetism and metallicity in the former [13, 18].
To understand the role of cationic size, we will now discuss the magnetic and electrical
properties of several series of cobaltites. The temperature dependent magnetization and
resistivity behavior of Gd0.5-xNdxBa0.5CoO3- series is presented Figure 10. With increasing
the Nd substitution or the <rA> value, the evolution of ferromagnetism is observed. It is
interesting to note that the 280 K magnetic transition of Gd0.5Ba0.5CoO2.9 disappears even
when x 0.1. Moreover, for x = 0.1, a complex behavior is observed with a magnetic
transition around 220 K. The x = 0.3 composition shows a weak increase in magnetization
value around 125 K, and the effect is more prominent for x = 0.4. Similarly, for Gd0.5-
xLaxBa0.5CoO3-, there is no clear FM transition in the 200-280 K region for 0.1 < x < 0.25. A
distinct FM transition occurs at x = 0.5 in the case of Nd, and at x = 0.4 in the case of La. It is
interesting that the FM characteristics start emerging at low temperatures (< 150 K) in these
cobaltite compositions around a rA, value of 1.30 . Clearly with increasing x, the size of
ferromagnetic clusters increases, eliminating the phase separation at lower value of x, caused
by size disorder. It is noteworthy that in the Ln0.7Ca0.3CoO3 (Ln = La, Pr, Nd) series the spin
glass like behavior has been reported at low temperatures, the system approaching toward
spin glass state [11] for the higher size disorder. Spin glass behavior is also reported for La1-
xSrxCoO3 (x < 0.1), but with increase in x ferromagnetism manifests itself [13].
Gd0.5-xNdxBa0.5CoO3- exhibit an insulating behavior throughout the temperature range,
but the electrical resistivity decreases significantly with increase in x, particularly x = 0.5
composition exhibiting the lowest resistivity (Figure 10). Similarly for Ln = La, the resistivity
value decreases with increasing x, becoming metallic for x = 0.5. All the other compositions
are semiconducting. Considering that with increase in x, there is significant increase in <rA>
values in these disordered Gd0.5-xLnxBa0.5CoO3- (Ln = La, Nd) series of cobaltites, the
changes observed can essentially be attributed to the effects of cation size which indirectly
control the electronic band width and as a result the energy band gap varies.
Figure 9. Variation of FM TC with <rA> for disordered cobaltites Ln0.5Ak0.5CoO3 (Taken from Ref. 13).
0
10 Gd0.5-xNdxBa0.5CoO3-
10
x = 0.0 -1
M (emu/g)
x = 0.1 10
x = 0.2
x = 0.3 -2
5 x = 0.4 10
(.cm)
x = 0.5
-3
10
-4
0 10
100 T(K) 200 300 100 T(K)200 300
Figure 10. Temperature dependent Magnetization and Resistivity for Gd0.5-xNdxBa0.5CoO3.
-1
200 10
2 2
(A )
TC(K)
15
160 0.0239
-2
10 0.0235
(.cm)
120 0.0207
10 0,016 0,020 0,024
M(emu/g)
0.0200
2 2 -3
(A ) 10 0.0170
5
-4
10
0
100 0 100 200 300
T(K) 200 300 T(K)
a) b)
Figure 11. Temperature dependent Magnetization and Resistivity for a fixed <rA> of 1.317 , inset
figure (a) shows the TC -2 plot (Taken from Ref. 18).
20 2 2
(A )
1 Constant <rA> of 1.289 A
0.0329 10
15 0.0276
M (em u/g)
(.cm)
0.0248
-1
0.0211 10
10
0.0184
-3
5 10
0 0 100 200 300

100 200 300 T(K)
T(K)
a) b)
Figure 12. Temperature dependent Magnetization and Resistivity for a fixed <rA> of 1.289 (Taken
from Ref. 18).
In order to understand the role of size-disorder due to cation size mismatch, 2, we have
discussed two series of cobaltites with fixed <rA> values of 1.317 and 1.289 , corresponding
to those of Nd0.5Ba0.5CoO3 and Gd0.5Ba0.5CoO3 respectively. The data, for a fixed <rA> of
1.317 , show that the FM TC decreases with increasing 2 (inset Figure 11a), eventually
destroying ferromagnetism at a high value of 2 ( 0.024 2). Similarly, the data in Figure 12,
for a fixed <rA> of 1.289 , are interesting. It is observed that with decreasing 2, the
magnetic behavior of this system changes markedly. Thus, when 2 = 0.028 2, there is no
such magnetic anomaly as observed in the case of Gd0.5Ba0.5CoO2.9 at 280 K. When 2 =
0.021 2, we observe a FM transition with a TC ~ 160 K and for 2 = 0.018 2, the TC value
reaches 220 K, a value higher than that of La0.5Ba0.5CoO3. These data clearly demonstrate that
the absence of prominent ferromagnetism in Gd0.5Ba0.5CoO3, and the unusual magnetic
properties, like magnetic anomaly around 280 K, is almost entirely due to the disorder arising
from the cation size mismatch. Such size-disorder can give rise to electronic phase separation
as reported for other cobaltites [18]. The electrical resistivity of these series of cobaltites
corroborates the results from the magnetic measurements. In Figure 12b, we present the
electrical resistivity data to demonstrate how the resistivity increases with increase in 2.
Interestingly, disorder-induced insulator-metal transitions are noticed in both the series of
cobaltites and the compositions with 2 < 0.02 2 showing metallic behavior. While disorder-
induced meal-insulator transitions are common, size variance-induced insulator-metal
transitions are indeed novel.
Further support of the previous discussion that cation size-disorder crucially determines
the properties of Gd0.5Ba0.5CoO2.9 is provided in this section by the study of Gd0.5Ba0.5-
xSrxCoO3- series of cobaltites as reported in the literature [18]. Here, x = 0.5 composition,
corresponding to Gd0.5Sr0.5CoO3, has a smaller <rA> than Gd0.5Ba0.5CoO2.9, and yet it shows
ferromagnetic features. The 280 K magnetic anomaly of Gd0.5Ba0.5CoO3 disappears even
when x = 0.1 and the apparent TC increases with increasing x in this series. This behavior is
clearly due to size disorder effect, since 2 decreases with increase in x. Accordingly, the
present system exhibits an insulator-metal transition with increase in x or decrease in 2. It
appears that a 2 value larger than 0.02 2 generally destroys ferromagnetism in the cobaltites
and changes the metal into an insulator as discussed earlier.
In view of the interesting magnetic and electrical properties of the disordered
Ln0.5Ba0.5CoO3 cobaltites, we have also presented a comparative study of Sr-doped disordered
cobaltites Ln0.5Sr0.5CoO3 with Ln = La, Pr, Nd, and Gd. Magnetic properties of La0.5Sr0.5CoO3
are fairly well understood. Although it shows a sharp FM TC around 240 K, it exhibits a
significant divergence between the FC and ZFC magnetization data (Figure 13a) and also
shows a frequency-dependent AC susceptibility maximum around 165 K, suggesting a glassy
nature [19]. It was realized a few years ago that La0.5Sr0.5CoO3, which was considered to be a
good ferromagnetic metal, was actually a magnetic cluster-glass with some long-range
magnetic ordering wherein frustration arose from inter-cluster interactions at low
temperatures [6]. The magnetic behavior of these systems has generally been interpreted in
terms of short-range magnetic ordering. Cluster-glass behavior occurs at a high concentration
of Sr (x > 0.3), where the coalescence of short-range ferromagnetic clusters is proposed to
occur [6, 13]. 139La NMR studies confirm the coexistence of ferromagnetic, paramagnetic, and
cluster-glass phases in La1-xSrxCoO3 [6, 13]. Electronic phase separation in the La1-xSrxCoO3
systems is associated with the formation of isolated nanoscopic ferromagnetic clusters. Thus,
the cobaltites comprise ferromagnetic clusters, paramagnetic matrices and spin-glass-like

phases, all contributing to the glassy magnetic behavior. More importantly, this glassy
ferromagnetism is accompanied by phase separation wherein the ferromagnetic clusters exist
within a antiferromagnetic (AFM) matrix. In these cobaltites, the FM phase is conducting and
the AFM phase is insulating. Depending on x or the carrier concentration, we can have a
situation such as that shown in Figure 14. The phase separation scenario here is somewhat
complex because the transition from the metallic to the insulating state is not sharp. In the
presence of Coulomb interaction, the microscopically charged inhomogeneous state is
stabilized, giving rise to clusters of one phase embedded in another. Electronic phase
separation with phases of different charge densities is generally expected to give rise to
nanometer scale clusters [19]. This is because large-phase separated domains would break up
into small pieces because of Coulomb interactions. The shapes of these pieces could be
droplet or stripes (see Figure 14a-d). The domains of the two phases can also be sufficiently
large to give rise to well-defined signatures in neutron scattering or diffraction. One can
visualize phase separation arising from disorder as well. The disorder can arise from the size
mismatch of the A-site cations in the disordered perovskite structure [1]. Such phase
separation is reported for (La1-yPry)1-xCaxMnO3 system in terms of a metal-insulator transition
induced by disorder [1]. The size of the clusters depends on the magnitude of disorder. The
smaller the disorder, the larger would be the size of the clusters. This could be the reason why
high MR occurs in systems with small disorder.
Figure 13. Temperature dependent FC-ZFC Magnetization for (a) La0.5Sr0.5CoO3 and (b) Gd0.5Sr0.5CoO3
(Taken from Ref. 19).
(a) (b) (c) (d)
Figure 14. Schematic representation of electronic phase separation (Taken from Ref. 1).
Figure 15. Variation of (a) FM TC and (b) Magnetization with <rA> for disordered cobaltites
Ln0.5Sr0.5CoO3 at 120 K (Taken from Ref. 19).
Electronic phase separation in the La1-xSrxCoO3 systems is associated with the formation
of isolated nanoscopic FM clusters [6]. Evidence for the occurrence of phase separation in
La0.5Sr0.5CoO3 is provided by the NMR studies [6]. The Mssbauer spectra reported by Bhide
et al. [13] show the presence of a paramagnetic signal in addition to the six-finger pattern due
to the FM-type species over a range of compositions and a wide range of temperatures. The
temperature variation of the ferromagnetic-paramagnetic (FM/PM) ratio was reported
recently for La0.5Sr0.5CoO3 [19], showing an increasing ratio with decreasing temperature, as
one would expect. While the FM/PM ratio increases with a decrease in temperature, the PM
phase continues to exist well below TC in La0.5Sr0.5CoO3. Clearly, this observation is a direct
evidence for phase separation in La0.5Sr0.5CoO3. It is of interest to examine the glassy phases
and phase separation occurring in the other rare earth cobaltite compositions of the type
Ln0.5Sr0.5CoO3. The temperature-dependent DC magnetization data of Gd0.5Sr0.5CoO3 exhibits

divergence in the ZFC and FC behavior just as in magnetically disordered systems. What is
particularly noteworthy is that the magnetization value in Gd0.5Sr0.5CoO3 is much smaller than
the Pr and La counterparts below their TCs, especially at temperatures below the relatively
sharp transition around 110 K. Since the carrier concentration or the Co3+/Co4+ ratio remains
constant in these cobaltites, the only possible explanation for such a decrease in
magnetization is that the proportion of PM species relative to that of the FM species increases
with the decrease in the size of the rare earth ion. It is entirely understandable that the
proportion of the clusters responsible for ferromagnetism decreases with decrease in <rA>. In
the Ln0.5Sr0.5CoO3 series, the FM TC decreases with decreasing <rA> (Figure 15) which is a
well-established behavior for disordered cobaltites [13, 19]. Since the magnetic moment
below TC is related to the proportion of the FM clusters, the variation of magnetization with
<rA> at 120 K in a series of Ln0.5Sr0.5CoO3 compounds is shown in Figure 15. We have
presented the size disorder effect in the Ln0.5Sr0.5CoO3 system as well. For this purpose, the
data were presented for fixed <rA> = 1.196 as equal to that of Dy0.5Sr0.5CoO3 and varying
the 2 values for different cobaltite compositions. The plot of TC values against 2 is shown in
Figure 16. The FM TC increases with a decreasing 2 and the TC0 value corresponding to the
disorder-free case (2 =0.0) is 217 2 K. The linear relation between TC and 2 gives a slope
of 9667 192 K -2, which is comparable to that reported in the literature [13, 18]. Thus, it
can be concluded for disordered cobaltites that, increasing size disorder favors electronic
phase separation, giving rise to magnetic clusters of different size ranges, while decreasing
size disorder increases the FM/PM ratio.
Figure 16. Magnetic properties for a fixed <rA> of 1.196 (same as Dy0.5Sr0.5CoO3) (Taken from Ref.
19).
Figure 17. Magnetization and Resistivity behavior for disordered La0.7-x(Pr/Nd/Gd/Dy)xCa0.5CoO3 series
with the variation of <rA> at 50 K (Taken from Ref. 11).
The value of MR for disordered cobaltites (even at TC/TIM) is much smaller in magnitude
than for the manganites [7]. However, in disordered cobaltites the large and negative MR has
been reported in the insulating compositions of La0.5Sr0.5CoO3 [7]. In such system, the
maximum MR is observed where it shows SG-like behavior [7, 13]. Disordered cobaltites of
the type Ln1-xCaxCoO3 show no long-range ferromagnetism or insulator-metal transition,
instead they exhibit electronic phase separation and/or glassy magnetic behavior at low
temperatures [11, 13]. Studies on La1-xCaxCoO3 samples have suggested that there are no
major differences from the Sr-doped system. Ferromagnetism is observed in both systems,
with the Curie temperature being lower in Ca-doped materials at a fixed doping level.
Magnetic and electron transport properties of different rare earth cobaltites have been
investigated to examine the effect of rA, and 2 on these systems as presented in Figure 17.
Thus, while La0.7Ca0.3CoO3 (rA, = 1.354 ) shows glassy ferromagnetism associated with
metallicity at low temperature, Ln0.7Ca0.3CoO3 with a smaller rA, of 1.179 (Ln = Pr) and
1.168 (Ln = Nd) shows no long-range ferromagnetism or insulator-metal transition [11].
Instead, the latter two systems exhibit electronic phase separation and/or spin-glass like
behavior at low temperatures. The electronic phase separation and associated magnetic
properties of Pr0.7Ca0.3CoO3 and Nd0.7Ca0.3CoO2.95 arise because of the small average size of
the A-site cations [11]. In these two cobaltites, the average radius (for orthorhombic structure)
is less than 1.18 , which is the critical value only above which long-range ferromagnetism
manifests itself. A detailed study on disordered rare earth cobaltites has shown the occurrence
of electronic phase separation and glassy magnetic behavior for small rA, and a large 2
value [11].
Figure 18. Temperature dependence ZFC (open symbol) and FC (solid symbol) Magnetization (H = 0.1
Tesla) for (a) disordered La0.5Ba0.5CoO3 and (b) ordered LaBaCo2O6. The inset figure shows dMFC/dT
vs temperature plot.
Figure 19. Temperature dependence electrical resistivity, , for (a) disordered La0.5Ba0.5CoO3 and (b)
ordered LaBaCo2O6 cobaltites (Taken from Ref. 12).
II. Ordered Cobaltites
The cationic order-disorder phenomena in the perovskite cobaltites do not affect much
their PM-FM transition temperature TC [9], in contrast to the ordered LaBaMn2O6 and
disordered La0.5Ba0.5MnO3 manganites [20]. However, the influence of the cationic ordering
upon the TC seems to be reverse since, according to Nakajima et al. [9], the disordered
cobaltites exhibit a higher TC of 190 K than the ordered phase (175 K). Figure 18, shows the
temperature dependent magnetization measurements for the disordered La0.5Ba0.5CoO3 and
ordered LaBaCo2O6 also indeed exhibit TC of 177 K and 174 K respectively [12]. Moreover,
other magnetic behaviors such as the field and the temperature dependent magnetization are
similar to expected for PM/FM phases [9, 12]. The coercive field, HC, for disordered
La0.5Ba0.5CoO3 and ordered LaBaCo2O6 are 0.08 and 0.05 Tesla (T) respectively (at 10 K); the
low value of HC signifies the nature of soft FM material [12]. There is one more cobaltite
compound reported in the literature as ordered-disordered phases, with a FM TC of 250 K in
the ordered NdBaCo2O6 phase [14] whereas for disordered Nd0.5Ba0.5CoO3 phase [13] the
value is 130 K. Since those compounds are reported by two different groups so we will not
elaborate our discussion on those phases for the present article.
Figure 19 shows the temperature dependence of electrical resistivity for both ordered-
disordered phases in the presence and absence of an applied magnetic field of 7 Tesla(T).
Note that the resistivity, (T), behavior of the disordered La0.5Ba0.5CoO3 and ordered
LaBaCo2O6 are different from those reported by Nakajima et al. [9], and corroborate the
result reported by Fauth et al. [8] for the disordered La0.5Ba0.5CoO3. The (T) curve in 10-400
K temperature range depicts that at high temperature (T>300K) the disordered La0.5Ba0.5CoO3
(Figure 19a) phase is semi-metallic, whereas the ordered LaBaCo2O6 is semiconducting down
to 190 K (Figure 19b). This feature is explained by the fact that the disordered cobaltite
exhibits 180 C-O-Co bond angle in this temperature range, in agreement with its cubic or
pseudo cubic structural symmetry, favoring a perfect overlapping of the Co 3d orbitals and
oxygen 2p orbitals. This is in contrast to the ordered LaBaCo2O6, where the C-O-Co bond
angles of 174 in the equatorial planes of the [CoO2] layers are observed at room
temperature [9]. In this case the conduction of charge carriers will be more favorable for
linear bond angle, as a result metallic type of conductivity is noticed for the disordered
cobaltite. With decreasing temperature a transition to a nearly metallic state is observed for
both phases. It is characterized by a change in slope of (T) at TC for the disordered
La0.5Ba0.5CoO3 (Figure 19a), or by a flat maximum at TC for the ordered LaBaCo2O6 (Figure
19b). Thus, these results show that irrespectively of their structural nature, the different forms
of cobaltites exhibit a FM metallic behavior below TC. Moreover, both phases depict an
upturn in the resistivity behavior at low temperature. This feature is interpreted as a weak
localization contribution associated with electron-electron interaction. In the present case, the
magnetoresistance measurements, that will be discussed later, suggest that the upturn is rather
due to grain boundary effects.
It is now important to discuss about the oxygen deficient ordered cobaltites LaBaCo2O5.5,
which have been of great interest due to their rich physical properties and interesting
structural phenomena associated with them. In these phases the oxygen stoichiometry is
5.5, hence the average valency of cobalt ion is Co3+ unlike the presence of mixed valences
in the other two phases. This is particularly interesting because of the ordering of Co3+ ions in
two different crystallographic sites corresponding to pyramidal and octahedral oxygen
coordination as discussed earlier. The magnetization and susceptibility curves versus
temperature, M (T), measured in the range 10-400 K, under external fields between 0.01 and 5
T (Figure 20) were reported recently by Rautama et al. [16] to be similar to those previously
observed for other lanthanides [3, 14]. In the whole temperature range the system exhibits
several magnetic transitions from PM to FM-like to antiferromagnetic as the temperature
decreases from 400 K to 10 K. The sudden increase of the magnetic susceptibility at TC = 326
K indicates a PM/FM transition and a sharp decrease in the magnetization at TN = 295 K
indicates a FM/AFM transition. The interactions between Co3+ ions, both in pyramidal and
octahedral coordination, are found to be AFM at low temperature. Moreover, the
thermomagnetic irreversibility between ZFC and FC in the low temperature AFM state
remains well discernible even at higher field (H 5 T). It is reported that a high field
basically affects the FM-AFM competition, which in turn suppresses the magnetization drop
below the TN. Thus, with increasing the external magnetic field, the FM state (260 K T
326 K) becomes more stable (indeed TC increases and the FM region expands in the
temperature scale), but TN shifts to lower temperature. In the AFM phase, the nonzero value of
magnetization down to low temperature signifies the presence of some kind of FM-like
interactions, where some weak magnetic transition is evidenced. Inset of Figure 20 depicts the
inverse magnetic susceptibility versus temperature plot in the temperature range of 220-400
K. This follows a simple Curie-Weiss behavior in the 335 K T 400 K range giving a PM
Weiss temperature (p) of ~ -290 K and an effective PM magnetic moment (eff ) of 5.27
B/f.u. The large negative p value for ordered LaBaCo2O5.5 sample indicates the existence of
AFM type interactions in the high temperature region. The obtained eff value (5.27 B/f.u.)
from the high temperature region magnetic data corresponds to a situation where the Co3+
ions are most probably in the IS state, which corroborate the results of NPD [16]. The sharp
drop in inverse susceptibility near TC, is similar to that observed for YBaCo2O5.5 and
GdBaCo2O5.5 [21, 24].
The FM-like features below TC ~ 326 K, have been confirmed by the magnetic field
dependent isotherm magnetization, M(H), studies at six different temperatures as shown in
Figure 21. The M(H) curve at 300 K shows a prominent hysteresis loop with a remanent
magnetization, Mr, and a coercive field, HC, values of 0.02 B/f.u and 0.1T respectively,
indicating a FM-like state below TC. Nevertheless, the maximum value of the magnetic
moment measured in 5T (0.23B/f.u.) at 275 K is much smaller than the theoretical spin-only
value (4B/f.u.) of Co3+ in IS state. Therefore, the FM-like behavior of this compound is most
probably due to the canting of the magnetic spin alignment in the AFM phase, often observed
in other systems [14]. Although there are some controversies in the literature to explain these
FM-like features, yet this behavior is prominent for this ordered cobaltite and subject to
further investigations. It is worth pointing out that at low temperature in the AFM state, some
FM-like phase is still present as evidenced from the M(H) behavior, with a finite value of the
coercive field. Interestingly, the highest moment of 0.23 B/f.u. (at 275 K) and highest
coercive field value of 0.4 T (at 245 K) are obtained in the AFM region compared to the
values of 0.16 B/f.u. and 0.1 T in the FM region (at 300 K). For comparison the HC values
have been re-plotted for lower fields at 245 and 300 K in the inset of Figure 21. The HC values
in the AFM region (275-200 K) are larger than in the FM region and finally at 10 K, the M(H)
behavior becomes linear akin to AFM state.
Figure 22 shows (T) for LaBaCo2O5.5 in the temperature range of 10-400 K (re-plotted
M(T) for comparison), which is plotted for heating and cooling cycle of the measurements in
the presence and absence of the external magnetic field of 7 T. The zero field (T) curve
shows a significant change in slope corresponding to the semiconductor-semiconductor
transition (TSC) around 326 K (Figure 22a). This type of transition is previously reported for
the other series of LnBaCo2O5.5 [14, 24, 26], and referred to as TIM albeit the true nature of
this transition is semiconducting to semiconducting type. For LaBaCo2O5.5 system, in contrast
to a metallic behavior, the slope of the resistivity curve (d/dT) is negative above the
transition temperature (T > TSC). Furthermore, for LaBaCo2O5.5 there is absence of any
significant change in the resistivity behavior through out the temperature range, apart from a
slight decrease in the magnitude below TSC. It is noticed that, the electronic and magnetic
transition temperatures for LaBaCo2O5.5 are almost the same (TC & TSC 326 K), in contrast
to other ordered cobaltites LnBaCo2O5.5, which exhibit a large shift between them [14, 24, 26,
27]. Hence, the sample is magnetic-semiconductor below TSC and the resistivity increases
exponentially with decreasing temperature.
0.12
T ~295K (a)
p~ -290K 200 N
0.10 eff~5.27B/f.u.
(emu mol Oe)

150
TC~326K
0.08 100
-1
50
(emu mol Oe )
-1
-1
0.06
-1
250 300 350 400

T(K)
0.04
0.02
H=0.01T
0.00 Open symbol ZFC
Solid symbol FC (b)
-1
10 H=5T
-2 H=2T
M(/f.u.)
10
-3
H=0.1T
10
-4 H=0.01T
10
0 50 100 150 200 250 300 350 400
T(K)
Figure 20. Temperature dependent ZFC (open symbol) and FC (solid symbol) Magnetization behavior
of ordered LaBaCo2O5.5; (a) magnetic susceptibility, , under H=0.01 Tesla (inset figure shows the
inverse magnetic susceptibility, -1, versus temperature plot and solid line is Cuire-Weiss fitting) and (b)
Magnetic moment in different magnetic fields (H=0.01, 0.1, 2 and 5 Tesla).
In general, the semiconductor or insulator like transport properties in perovskite

cobaltites were characterized by three possible models [11, 25] namely thermal activation
(TA): log 1/T, Efros-Shklovskii type hopping (ESH): log T-1/2 and Motts variable range
hopping (VRH): log T-1/4. To understand the transport mechanism for LaBaCo2O5.5, the
data have been analyzed based on these models. The zero field resistivity data is plotted in
Figure 23a, which shows that the VRH model fits better than the other two models in 40 K T
270 K range, and is consistent with the preceding studies on perovskite cobaltites [11, 25].
This type of transport mechanism is typical for disordered systems as shown in Figure 23b,
where the charge carriers move by hopping between localized states [11]. Taskin et al [26]
have described the formation of localized states in terms of oxygen defects in 112-type
ordered GdBaCo2O5.5, which inevitably generate electrons or holes in the CoO2 planes.
Furthermore, we consider ESH and TA models, which are expected to describe the dominant
conduction process. The transport mechanism is very complicated in the whole temperature
region and does not satisfy entirely any of the above models for LaBaCo2O5.5. From the
nearly linear region of the TA model ( 250 K) we have calculated the approximate activation
energy (Ea) of around 21 meV, yet less than the value reported for GdBaCo2O5.5 [26].
The magnetoresistance (MR) behavior of the stoichiometric ordered-disordered phases
(Figure 24) shows a clear magnetic field dependent change in the resistivity below TC. The
MR value is calculated as MR (%) = [{(7)-(0)}/(0)]x100, where (0) is the sample
resistivity at 0 T and (7) under an applied field of 7 T. For the disordered La0.5Ba0.5CoO3
(Figure 24a), the maximum MR value is obtained around TC and the corresponding value is
indeed 7 % at 179 K. The ordered LaBaCo2O6 exhibits (Figure 24b) a rather close value of 6
% around TC (at 179 K). But importantly, the ordered LaBaCo2O6 depicts an MR value up to
14.5% at 10 K in an applied field of 7 T which is much larger than an MR value of 4% at the
same temperature (10 K) for the disordered phase. This difference suggests that at low
temperature (T < 50 K), the grain boundary effect plays an important role in the anisotropic
MR behavior for the ordered LaBaCo2O6 perovskite. This is in agreement with its much larger
upturn of resistivity, which is almost 10 times larger (at 10 K) than the disordered
La0.5Ba0.5CoO3 cobaltites. Thus, the larger MR observed for the ordered LaBaCo2O6 is
interpreted as tunnelling magnetoresistance (TMR) effect due to the increase of the intergrain
insulating barriers [28], rather than an intrinsic effect, and is dominant at 10 K over the TMR
effect for the disordered phase. Therefore the spin-polarized tunnelling of carriers across the
insulating boundaries occurring at the interfaces between polycrystalline grains give rise to
the TMR effect in these phases. Figure 25, shows the MR effect for ordered LaBaCo2O5.5 at
five different temperatures. Unfortunately, in the case of temperature dependent resistivity
data (with field), there is no major change in the (T) curve even though the M(T) behavior
reveals some kind of FM and AFM ordering. The charge transport for this kind of system is
expected to be very sensitive due to the co-existence of FM and AFM state and the external
magnetic fields readily induces an MR effect by affecting the subtle balance between FM-
AFM phases. Unlike, disordered La0.5Ba0.5CoO3 and ordered LaBaCo2O6 cobaltites, the
highest negative MR value is obtained at 245 K around 5 % in an applied field of 7 T and near
room temperature the value is only 1.6 %. At low temperatures, the MR values are only 2.8 %
(at 150 K) and 2.5 % (at 50 K) respectively. The evidence of negative MR at low
temperatures, similar to the TMR observed usually in polycrystalline samples, in the case of
ordered cobaltites, is considered to be related to the spin-dependent scattering at grain
boundaries. Nevertheless, in the case of ordered LaBaCo2O5.5, the highest MR value is noticed
near the FM-AFM phase boundary, hence the grain boundary effect can be ignored and
considered to be as an intrinsic effect.
Interestingly, the isothermal MR behavior at 245 K exhibits an irreversible effect
analogous to those of isothermal magnetization, M(H), behavior (inset Figure 25), also
present in 300 K isothermal MR data. The peak in the isothermal MR data occurs around the
coercive field value, which corresponds to the state of maximum disorder in the orientation of
the neighbouring magnetic spins.
0.08 T=300K
M(/f.u.)
0.2
0.04 T=245K
0.00
0.1 -1.0 -0.5 0.0 0.5 1.0
H(T)
-0.04
M(/f.u.)
-0.08
0.0
10 K
200K
245K
-0.1 275K
300K
325K
-0.2
-4 -2 0 2 4
H(Tesla)
Figure 21. Magnetic field dependent isotherm magnetization, M(H), of ordered LaBaCo2O5.5 at six
different temperatures. The inset figure shows the expanded version for lower magnetic fields at 245
and 300K (Taken from Ref. 25).
Hence, the field dependent MR data that is indirectly related to the alignment between
magnetic spins, reaches a maximum value. This effect is prominent for 300 K data, compared
to 245 K as evidenced from the inset Figure 25 (dotted vertical lines), which may be due to
FM-like state near 300 K whereas latter one corresponds to magnetic phase boundary.
Additionally, the isothermal MR data exhibit hysteresis effects that resemble the butterfly-
like feature, although the effect is rather weak at low temperature (50 K). It is clear from the
obtained data that the MR effect is strongly irreversible near the FM-AFM phase boundary.
The butterfly-like feature appears only near the magnetic phase boundary. Correspondingly,
the magnetic field dependent isotherm MR behavior at 10 K for stoichiometry ordered-
disordered phases exhibits an anisotropic effect similar to those of the magnetization
behavior. Nevertheless, the isotherm MR data at low temperature for all three phases exhibit
hysteresis effects, which resemble the butterfly-like feature, although the effect is rather
weak for the disordered phase. It is noticed that the MR effect is almost isotropic for
temperatures near or above the TC. Hence, the butterfly-like feature appears only at low
temperatures, which is prominent for ordered LaBaCo2O5.5 and LaBaCo2O6 phases. The
origin of the magnetic field induced maximum MR near TC for the disordered La0.5Ba0.5CoO3
and ordered LaBaCo2O6 are explained by the mechanism of suppression of spin fluctuations
below TC. On the other hand, the highest obtained MR value, obtained at 10 K for the ordered
LaBaCo2O6 is explained by the TMR effect due to the presence of more insulating grain
boundaries. Hence, the occurrence of irreversible MR behavior nearly at similar temperatures
as those for magnetic field variation isothermal M(H) studies suggest the strongly correlated
nature of field-induced magnetic and electronic transitions [12, 25].
6
(a)
10
2
TSC~326K
( cm)
4
10 1
( cm)
T(K)
2
10 280 300 320 340
0T(Cooling)
0 0T(Heating)
10 7T(Heating)
0.08
(b) TC~326K
H = 0.5 Tesla
0.06
M(/f.u.)
Solid symbol FC
Open symbol ZFC
0.04
0.02
0.00
0 50 100 150 200 250 300 350 400
T(K)
Figure 22. Temperature dependent physical properties for ordered LaBaCo2O5.5: (a) electrical
resistivity, (T), in the presence (solid symbol) and absence (open symbol) of magnetic field (7 Tesla)
during heating and cooling cycles (inset shows the expanded version near the transition temperature,
TIM), and (b) ZFC (open symbol), FC (solid symbol) Magnetization in an applied field of 0.5 Tesla
(Taken from Ref. 25).
(a)
6 6
TA
5 4
log
Ea~21meV
2
4
log
0
-1/2 -1/2
3 0.005 0.010 0.015 0.020
T (K )
-1 -1
0.04 0.08 0.12 0.16
T (K ) 6
2
4
log
1
2
0 0 ESH
VRH fit: 270-40K
-1
0.22 0.24 0.26 0.28 0.30 0.32 0.34 0.36 0.38 0.40 0.42
-1/4
-1/4
T (K )
(b)
Figure 23. Logarithm of the resistivity versus T-1/n plots for (a) ordered LaBaCo2O5.5 and (b) disordered
Ln0.7Ca0.3CoO3 cobaltites (where n = 1, 2 or 4): open symbols and solid lines represent the experimental
data and apparent fit to the hopping models as described in the text (Taken from Ref. 11 & 25).
Figure 24. Temperature variation marnetoresistance for (a) disordered La0.5Ba0.5CoO3 (b) ordered
LaBaCo2O6 cobaltites (Taken from Ref. 12).
In order to get more insight into the nature of the conduction mechanism below TSC, we
have also discussed the thermopower (Seebeck coefficient), S(T), and thermal conductivity,
(T), measurements for ordered LaBaCo2O5.5. These are sensitive to the magnetic and
electrical nature of charge carriers (hole/electron) and can give some valuable information,
which are absent in the magnetotransport measurements. Additionally, the S(T) data is less
affected due to the presence of grain boundaries, which often complicates the (T) data
interpretation for polycrystalline samples. We noticed that, at room temperature the system
shows a relatively large positive value of the thermoelectric power (91 V/K), while it has a
low resistivity (~1.14 cm at 300 K). The S(T), for cooling and heating cycles is shown in
Figure 26. It is observed that, the S(T) value is positive below TSC and with decreasing
temperature the value increases gradually and reaches a maximum value of ~303 V/K at
around 120 K (marked by an arrow in the figure). Below that, the S(T) value decreases
continuously and similar to the literature behaviors (i.e. for NdBaCo2O5.5, GdBaCo2O5.5 and
HoBaCo2O5.5 the peak values are 105, 88 and 70 K respectively [26, 27]).
H(Tesla)
-6 -4 -2 0 2 4 6
0
-1 T=300K
-2 50K
150K
-3 245K
T=245K
275K
300K
MR(%)
-4
H(T) H(T)
-5 -1.0 -0.5 0.0 0.5 1.0
0.1
-4 -2 0 2 4
0.2
0.0 0
M(/f.u.)
M(/f.u.)
MR(%)
MR(%)
-6 -1
-0.2 0.0 -2 0.0

-7
-3
T=300K T=245K
-8 -0.4 -0.1 -4 -0.2
-9
Figure 25. Magnetic field dependent isotherm magnetoresistance, MR, effect for ordered LaBaCo2O5.5
at five different temperatures (H= 7 Tesla). Inset figures show the isotherm magnetization, M(H), and
MR plot at 245 and 300 K for comparison (Taken from Ref. 25).
The S(T) data implies that the LnBaCo2O5.5 systems show a semiconducting type
behavior of the thermopower down to low temperature akin to resistivity behavior and below
the S(T) behavior is complex, decreasing abruptly at low temperature. In fact, with decreasing
temperature the S(T) value should increase due to trapping or localization of charge carriers.
This type of S(T) behavior is quite unexpected for semiconducting thermoelectric materials
and there is no general explanation till date. Following the general approach in analyzing the
semiconducting behavior we have plotted the S(T) data in the T-1/n scale similar to (T). Since
for semiconductors the S(T) is expected to be linear in T-1 behavior (according to TA model)
or to follow either of the described hopping models similar to resistivity behavior [11, 25]. It
is noticed that the S(T) data can not be described by the previous mentioned models and one
observes p-type conductivity throughout the temperature range. Many authors have already
explained the S(T) behavior at higher temperature and their sign reversal near the electronic
transition [26, 27]. However, the low temperature S(T) evolution, basically the appearance of
the broad maximum in the metastable AFM phase and the decreasing nature with temperature
(in spite of semiconducting behavior) has not been explained properly. We have analysed the
obtained thermopower data using an expression S(T) = S0 + S3/2T3/2 + S4T4 defined by P.
Mandal [29], which can be understood on the basis of electron magnon scattering (spin wave
theory). In ferro- and antiferromagnets, electrons are scattered by spin wave as explained
earlier for perovskite manganites [29] and hence it is expected that this theory may explain
the S(T) behavior for the present system. However, the obtained thermopower value is higher
than the one reported for manganites. The S(T) data in 60 K T 105 K range follows T3/2
behavior and in the 120-320 K range, it fits linearly with the T4 behavior (insets of Figure 26).
At low temperature the second term (S3/2) will dominate over S4 (S3/2>>S4), hence the S(T)
will depict downward trend at low temperature. Although we do not have sufficient
experimental data points below the broad peak, yet the S(T) curve fits linearly to the T3/2 term
(inset of Figure 26b), as expected from the spin wave theory. Additionally, the downward
feature in S(T) data is present in all studied LnBaCo2O5.5 cobaltites [25-27]. Hence, the broad
peak at low temperature and downward trends for layered cobaltites are considered to be the
result of the electron magnon scattering akin to perovskite manganite as explained by P.
Mandal [29].
300 (a)
Tp~120K
250
200 300 liear fit:120-320K

S(V K )
-1
-8 -5
S4~-2.7x10 VK
150
-1
S0~311VK
S(V K )
200
-1
100 100
50 4
T (10 K )
9 4
0
0 2 4 6 8 10
0 6
1.2 liear fit:60-105K (b)
300
5
S(V K )
-1
280 S3/2~0.065VK
-5/2
0.8 S0~239VK
-1
4
(W K m )
S (W K m )
-1
3/2 2 3/2
260 T (10 K )
4 6 8 10 3
-2
-1
0.4
-1
2
Cooling
2
heating
1
0.0
0
50 100 150 200 250 300
T(K)
Figure 26. Temperature dependent transport measurements for ordered LaBaCo2O5.5: (a) thermoelectric
power, S(T), during cooling (solid triangle) and heating (open triangle) cycles and inset shows the S-T4
plot in the 120-320K range, (b) thermal conductivity, (T), and power factor, S2(T), in the temperature
range of 60-320K (inset shows the S-T3/2 plot in the 60-105K range) (Taken from Ref. 25).
Thermal conductivity, (T), measurements of LaBaCo2O5.5 during cooling and heating

cycles are shown in Figure 26b. The (T) value increases with increasing temperature,
indicating an usual phonon mediated scattering mechanism of charge carriers, which is the
sum of phonon (p) and electronic (e) contribution [i.e. (T) = p(T) + e(T)]. Therefore, from
the Wiedemann-Franz law we have calculated the electronic contribution (e) at 250 K, which
is around ~1.5x10-4 WK-1m-1 [whereas (250 K) ~ 2.8 WK-1m-1]. This implies that, the lattice
modulated phonon contribution (p) is dominant near room temperature i. e. p(T) >> e(T).
We have also calculated the temperature dependence of power factor (S2) from the obtained
resistivity and thermopower data as shown in Figure 26b, which depicts a clear maximum
near 270 K and decreases rapidly on both sides of the temperature scale (60K T 320K). It
is important to note that, this maximum is in the region of FM-AFM phase boundary. This
suggests that the electron magnon scattering at low temperature strongly affects the S(T)
behavior of LaBaCo2O5.5 system. The calculated figure of merit (ZT) shows too small values
close to room temperature (~10-5) for applications as expected from 112-type cobaltite
systems.
CONCLUSION
The discussion in the earlier sections clearly brings out the varied effects of
magnetization, magneto-transport properties, electronic phase separation, thermopower etc, in
the ordered-disordered rare earth cobaltites. This paper shows that the method of synthesis
precursor reactivity and oxygen pressure plays a crucial role for controlling the cationic
order-disorder phenomena in the non-stoichiometry ordered LaBaCo2O5.5 and the
stoichiometric disordered La0.5Ba0.5CoO3 and ordered LaBaCo2O6. From the viewpoint of the
transport properties, the ordered LaBaCo2O5.5 is semiconducting throughout and the
disordered La0.5Ba0.5CoO3 exhibits a semi-metal to metal transition around TC, whereas the
layer-ordered LaBaCo2O6 is characterized by semi-conductor to metal transition around TC.
As a consequence, they exhibit a similar intrinsic magnetoresistance, maximum at the vicinity
of TC (6 to 7 %), whereas the ordered LaBaCo2O6 exhibits a much larger MR value, indicating
TMR at low temperature due to different grain boundary effects. It appears that the
phenomenon is much more common than anticipated, to the extent that some workers suggest
that even the physical properties of these cobaltites, showing different types of interesting
behavior, is a consequence of cationic ordering. It is noteworthy that the large thermopower
value caused by lowering temperature in ordered LaBaCo2O5.5 may indeed be due to
semiconducting behavior. The unusual S(T) behavior and the appearance of a broad peak at
low temperature with a maximum value of ~303 V/K is explained by electron magnon
scattering mechanism. This is expected to be applicable to all series of LnBaCo2O5.5 cobaltites
at low temperature. It would be worthwhile to study the relevance of this phenomenon to
other transition metal oxide systems.
ACKNOWLEDGMENTS
AKK gratefully acknowledges Prof. A. Ojha, Director of IIITDM-J, for sanctioning
faculty research grant.
REFERENCES
[1] Rao C. N. R.; B. Raveau (Eds.), Colossal Magnetoresistance, Charge-Ordering and
Related Properties of Manganese Oxides, World Scientific, Singapore, 1998; Dagotto
E., Nanoscale Phase Separation and Colossal Magnetoresistance: The Physics of
Manganites and Related Compounds, Springer-Verlag, Berlin, New York, 2003;
Rao, C. N. R.; Kundu, A. K.; Seikh, M. M.; Sudheendra, L. Dalton Trans. 2004, 19,
3003.
[2] Jonker, G. H.; van Santen, J. H.; Physica 1953, 19, 120; Goodenough, J. B. J. Phys.
Chem. Solids 1958, 6, 287; Jonker, G. H. J. Appl. Phys. 1966, 37, 1424.
[3] Martin, C.; Maignan, A.; Pelloquin, P.; Nguyen, N.; Raveau, B. Appl. Phys. Lett.
1997, 71, 1421.
[4] Teraoka Y.; Nobunaga T.; Okamoto K.; Miura N.; Yamazoe N. Solid State Ionics
1991, 48, 207; Kruidhof H.; Bouwmeester H. J. M.; van Doorn R. H. E.; Burggraaf
A. J. Solid State Ionics 1993, 63, 816; van Doorn R. H. E.; Burggraaf A. J. Solid
State Ionics 2000, 128, 65; Kim, G.; Wang, S.; Jacobson, A. J.; Reimus, L.;
Brodersen, P.; Mims, C. A. J. Mater. Chem. 2007, 17, 2500.
[5] Takada K.; Sakurai H.; Takayama-Muromachi E.; Izumi F.; Dilanian R.; Sasaki T.
Nature 2003, 422, 53.
[6] Itoh M.; Natori I. ; Kubota S.; Matoya K. J. Phys. Soc. Jpn.1994, 63, 1486; Anil
Kumar P. S. ; Joy P. A. ; Date S. K. J. Phys.: Condens. Matter 1998, 10, L487; Nam,
D. N. H.; Jonason, K.; Nordblad, P.; Khiem, N. V.; Phuc, N. X.; Phys. Rev. B, 1999,
59, 4189; Kuhns, P. L.; Hoch, M. J. R.; Moulton, W.G.; Reyes, A. P.; Wu, J.;
Leighton, C. Phys. Rev. Lett. 2003, 91, 127202; Burley, J. C.; Mitchell, J. F.; Short,
S. Phys. Rev. B 2004, 69, 054401; Troyanchuk, I. O.; Kasper, N. V.; Khalyavin, D.
D.; Szymczak, H.; Szymczak, R.; Baran, M. Phys. Rev. Lett.1998, 80, 3380.
[7] Mahendiran, R.; Raychaudhuri, A. K. ; Chainani, A. ; Sarma, D. D. J. Phys.:
Condens. Mater. 1995, 7, L561; Mahendiran, R.; Raychaudhuri, A. K. Phys. Rev. B
1996, 54, 16044.
[8] Fauth, F.; Suard, E.; Caignaert, V. Phys. Rev. B 2001, 65, 60401.
[9] Nakajima, T.; Ichihara, M.; Ueda, Y. J. Phys. Soc. Jpn. 2005, 74, 1572.
[10] Wu, J.; Lynn, J.W.; Glinka, C.J.; Burley, J.; Zheng, H.; Mitchell, J.F.; Leighton, C.
Phys. Rev. Lett. 2005, 94, 037201.
[11] Kundu, A. K.; Ramesha, K.; Seshadri, R.; Rao, C. N. R. J. Phys. Condens. Matter
2004, 16, 7955; Kundu, A. K.; Sampathkumaran, E. V.; Gopalakrishnan, K. V.; Rao,
C. N. R. J. Mag. Mag. Mater. 2004, 281, 261; Kundu, A. K.; Nordblad, P.; Rao, C.
N. R. Phys. Rev. B 2005, 72, 144423; Kundu, A. K.; Nordblad P.; Rao C. N. R.; J.
Solid State Chem. 2006, 179, 923.
[12] Rautama, E-L; Boullay, P.; Kundu, A. K.; Pralong, V.; Caignaert, V.; Karppinen, M.;
Raveau, B. Chem. Mat. 2008, 20, 2742.
[13] Rao, C. N. R.; Bhide, V. G.; Mott, N. F. Phil. Mag. 1975, 32, 1277; Bhide, V. G.;
Rajoria, D. S.; Rao, C. N. R.; Rao, G. R.; Jadhao, V. G. Phys. Rev. B 1975, 12, 2832;
Rao, C. N. R.; Prakash, O.; Bahadur, D.; Ganguly, P.; Nagabhushana, S. J. Solid
State Chem. 1977, 22, 353; Senaris-Rodriguez, M.A.; Goodenough, J.B. J. Solid
State Chem. 1995, 118, 323; Moritomo, Y.; Takeo, M.; Liu, X. J.; Akimoto, T.;
Nakamura, A. Phys. Rev. B 1998, 58, R13334; Troyanchuk, I.O.; Kasper, N.V.;
Khalyavin, D.D.; Szymczak, H; Szymczak, R.; Baran, M. Phys. Rev. B 1998, 58,
2418; Vanitha, P. V.; Arulraj, A.; Santosh, P. N.; Rao, C. N. R. Chem. Mater. 2000,
12, 1666; Ganguly, R.; Maignan, A.; Martin, C.; Hervieu, M.; Raveau, B. J. Phys.:
Condens. Matter 2002, 14, 8595.
[14] Maignan A, Martin C, Pelloquin D, Nguyen N and Raveau B J. Solid State Chem.
1999, 142, 247; Moritomo, Y.; Akimoto, T.; Takeo, M.; Machida, A.; Nishibori, E.;
Takata, M.; Sakata, M.; Ohoyama, K.; Nakamura, A. Phys. Rev. B 2000, 61, R13325;
Roy, S.; Khan, M.; Guo, Y. Q.; Craig, J.; Ali, N. Phys. Rev. B 2002, 65, 064437;
Burley, J. C.; Mitchell, J. F.; Short, S.; Miller, D.; Tang, Y. J. Solid State Chem.
2003, 170, 339; Roy, S.; Dubenko, I. S.; Khan, M.; Condon, E. M.; Craig, J.; Ali, N.
Phys. Rev. B 2005, 71, 024419; Pralong, V.; Caignaert, V.; Hebert, S.; Maignan, A.;
Raveau, B. Solid State Ionics 2006, 177, 1879; Seikh, M. M.; Simon, C.; Caignaert,
V.; Pralong, V.; Lepetit, M. B.; Boudin, S.; Raveau, B. Chem. Mat. 2008, 20, 231.
[15] Vogt, T.; Woodward, P. M.; Karen, P.; Hunter, B. A.; Henning, P.; Moodenbaugh,
A. R. Phys. Rev. Lett. 2000, 84, 2969; Suard, E.; Fauth, F.; Caignaert, V.; Mirebeau,
I.; Baldinozzi, G. Phys. Rev. B 2000, 61, R11871.
[16] Rautama, E-L.; Caignaert, V.; Boullay, P.; Kundu, A. K.; Pralong, V.; Karppinen,
M.; Ritter, C.; Raveau, B.; Chem. Mat. 2009, 21, 102.
[17] Rodriguez-Martinez, L. M.; Attfield, J. P. Phys. Rev. B 1996, 54, R15622; Attfield, J.
P. Chem. Mater. 1998, 10, 3239.
[18] Kundu, A. K.; Rao, C. N. R. J. Phys.: Condens. Matter 2004, 16, 415; Kundu, A. K.;
Sampathkumaran, E. V.; Rao, C. N. R. J. Phys. Chem. Solids 2004, 65, 95.
[19] Kundu, A. K.; Sarkar, R.; Pahari, B.; Ghoshray A.; Rao C. N. R. J. Solid State Chem.
2007, 180, 1318 and references therein.
[20] Millange, F.; Caignaert, V.; Domengs, B.; Raveau, B.; Suard, E. Chem. Mater.
1998, 10, 1974.
[21] Akahoshi, D.; Ueda, Y. J. Solid State Chem. 2001, 156, 355.
[22] Abbate, M.; Fuggle, J. C.; Fujimori, A.; Tjeng, L. H.; Chen, C. T.; Potze, R.;
Sawatzky, G. A.; Eisaki, H.; Uchida, S. Phys. Rev. B 1993, 47, 16124; Korotin, M.
A. ; Ezhov, S. Y. ; Solovyev, I. V. ; Anisimov, V. I.; Khomskii, D. I.; Sawatzky, G.
A. Phys. Rev. B 1996, 54, 5309; Wu, H. Phys. Rev. B 2000, 62, R11953.
[23] Zener, C. Phys. Rev. 1951, 82, 403; Goodenough, J. B.; Wold, A.; Arnott, R. J.;
Menyuk, M. Phys. Rev. 1961, 124, 373.
[24] Taskin, A. A.; Lavrov, A. N.; Ando, Y. Phys. Rev. Lett 2003, 90, 227201.
[25] Kundu, A. K.; Raveau, B.; Caignaert, V., Rautama, E-L.; Pralong, V. J. Phys.:
Condens. Matter 2009, 21, 056007.
[26] Taskin, A. A.; Lavrov, A. N.; Ando, Y. Phys. Rev. B 2005, 71, 134414; Taskin, A.
A.; Lavrov, A. N.; Ando, Y. Phys. Rev. B 2006, 73, R121101.
[27] Maignan, A.; Caignaert, V.; Raveau, B.; Khomskii, D.; Sawatzky, G. Phys. Rev. Lett.
2004, 93, 26401.
[28] Niebieskikwiat, D.; Prado, F.; Caneiro, A.; Sanchez, R. D. Phys. Rev. B 2004, 70,
132412.
[29] Mandal, P. Phys. Rev. B 2000, 61, 14675.
Chapter 5
ELECTRICALLY TUNABLE DIELECTRIC AND

CONDUCTION PROPERTIES IN
PEROVSKITE THIN FILMS
Yidong Xia*, Xiangkang Meng and Zhiguo Liu

Department of Materials Science and Engineering, and National Laboratory of Solid
State Microstructures, Nanjing University, Nanjing 210093, Peoples Republic of China
ABSTRACT
Perovskite thin films are widely employed in todays advanced technology. The
broad range of physical properties in such category of materials has offered various
functionalities in devices ranging from dynamic random access memories and
ferroelectric nonvolatile memories to piezoelectric and optical devices. In addition to the
aforementioned applications, perovskites have been revealed to be promising candidate
materials in many novel applications including electrically tunable microwave devices
and resistive switching memories. Thin films of the perovskite ferroelectrics are being
used to develop the tunable microwave devices owing to their strong electric field
dependent dielectric constant in the paraelectric state. Reversible switching between
high- and low- resistance states induced by electric field makes perovskite thin films
suitable for the applications in resistive switching memories. This chapter focuses on our
recent progress on these two kinds of electrically tunable properties in perovskite thin
films. Based on the origin of dielectric tunability that is defined as the variation of
dielectric constant in a certain electric field range, approaches to improve the tunable
dielectric properties will be introduced. On the other hand, although resistive switching
behaviors have been discovered in many perovskites during the last decade, the origin of
such switching is not well understood yet. In this chapter, some proposed mechanisms
responsible for the resistive switching will be discussed.
*
Author to whom correspondence should be addressed; electronic mail: [email protected]
252 Yidong Xia, Xiangkang Meng and Zhiguo Liu
I. DIELECTRIC TUNABLE PROPERTIES IN

PEROVSKITE THIN FILMS BASED ON (BA,SR)TIO3
In the last two decades, much attention has been paid to ferroelectrics thin films based on
(Ba,Sr)TiO3 (BST) and SrTiO3 (STO) for their applications in high density dynamic random
access memory because of its high dielectric constant [1]. With the development of these
storage capacitor dielectrics for the memory applications, it has been revealed that such
perovskites show strong electric-field dependent dielectric constant, where typically
decreases with increasing field [2-4]. Such behavior is usually described by a quantity termed
as tunability, which is defined as [5-7]
( max min ) / max (1)
to characterize the variation of dielectric constant in a certain electric field range. Here, max
is the maximum dielectric constant, generally obtained at zero applied bias, and min is the
minimum dielectric constant measured at the bias applied to the dielectrics. The field-
dependent dielectric properties are essential in designing a new class of electrically tunable
microwave devices, such as phase-shifters, tunable filters, and voltage-controlled oscillators
[5-15].
Generally, the perovskite thin films are integrated in capacitor configuration for these
applications. In most practical cases the tunability of the capacitance
TC (V ) = (Cmax Cmin ) / Cmax rather than the tunability of dielectric constant T (V ) defined
by Eq. 1 is used to estimate the effective parameters of the materials [7]. Two capacitor
structures are usually employed [13]. One is the sandwich-type with parallel-plate electrode,
i.e. a layer of ferroelectric thin film sandwiched between two electrodes (metal or conducting
oxide electrodes). The other is the planar capacitor with interdigital electrode, i.e. a gap
between electrodes on the surface of a ferroelectric film. For sandwich-type capacitors TC (V )
is approximately the same as T (V ) , while it is smaller than T (V ) for planar capacitors. It
has to be concerned for the planar capacitors that measured effective dielectric constant,
tunability, and dielectric loss are often lower than those of the film itself [7]. This is because
in planar structures, a considerable part of the field is outside the ferroelectric films and
locates in the substrates. The field outside the ferroelectric layers with given film thickness
depends on the gap between the planar electrodes, which makes the applied field inefficient to
adjust the dielectric constant and higher control voltages are accordingly necessary to obtain
the desired tunable parameters as compared with the sandwich capacitors. As a result, in
general the sandwich-type capacitors are especially attractive in these electrically tunable
devices although the planar capacitors are simpler to fabricate and integrate into circuits
[7,13].
Electrically Tunable Dielectric and Conduction Properties 253
1. Thickness-dependent Dielectric Characteristics of Perovskite Thin Films
A critical issue as to the sandwich capacitors based on ferroelectric thin films is the
thickness-dependence of dielectric properties. The development of portable and agile devices
requires the reducing size of capacitor components. Unlike their predecessor, bulk ceramics,
however, thin films show far lower effective dielectric constant and tunablity as what is
expected [16-28]. The dielectric properties of ferroelectric thin films degrade with decreasing
thickness, markedly in the case of metals used as the electrodes [16,17]. Several mechanisms
have been proposed to account for the microscopic origin of this thickness dependence.
Among them, the most employed one is the so-called interfacial dead layer model [17-
19,24-28], which assumes the existence of an interfacial layer with low dielectric constant at
film/electrode interface, acting as a smaller capacitance in series with the film bulk
capacitance. The effective dielectric constant in thin film capacitors is usually experimentally
derived from the measured capacitance density. Consequently, according to the principle of
dead layer model, the reduction in effective dielectric constant can be expressed in terms of
the limited overall capacitance density C as [28]
1 1 1 1
= + + , (2)
C Ci Cb Ci
where Ci and Cb stand for the low capacitance density of the electrode/film interfacial regions
and the capacitance density of the bulk film, respectively.
But, the origins of such dead layer are not well understood yet. Debates focused on
whether such dead layer exists at all, and if so, whether it is an intrinsic property as films
contacted with electrodes or a result of processing issues such as defects and strains [28].
First-principles calculations, recently reported by Stengel and Spaldin, conjectured that
excellent screening properties at the interface are crucial to reducing the dead layer [28]. They
demonstrated the existence of a dielectric dead layer by calculating the dielectric profile
across the interface and analyse its origin by extracting the ionic and electronic contributions
to the electrostatic screening. They further predicted that metals, e.g. Pt, not the metallic
oxide electrodes (e.g. SrRuO3), should be the best candidates for reducing the dead layer,
because of their shorter electronic screening length. Accordingly, the experimentally observed
dielectric degradation with film thickness for Pt/BST/Pt structure is attributed to the structural
mismatching or defects at interfaces.
Studies on the interfacial microstructure via high resolution transmission electron
microscopy (HRTEM) have been performed to verify the existence of dead layer between
BST films and Pt or SrRuO3 electrodes. Formations of misfit dislocations, planar fault, and
disordered or amorphous regions have been reported [19,27,29-33]. More recently, we have
observed two distinct types of interfaces between BST films and Pt electrodes via HRTEM
[34]. One case is the presence of interfacial layer with distorted structure (type I), as shown in
Figure 1 (a). Another case, called type II, is exhibited in Figure 1 (b), in which good
interfacial match with no interfacial layer is observed.
(a)
(b)
Figure 1. (Ref. [34])
The origin of such dissimilar interfaces has been analyzed based on the crystallographic
relationships between BST and Pt. It is concluded that the atomic arrangement in the plane
parallel to the interfacial plane plays an important role in determining the feature of the
interfaces. For BST films, the directions perpendicular to the BST/Pt interfaces in the two
cases are both determined to be [001], while for Pt electrodes the normal direction is [111], as
illustrated in the insets in Figs. 1 (a) and (b). This indicates that the influence of the out-of-
plane lattice constants is negligible due to the same mismatch of lattice constants between
BST and Pt in the normal direction in these two cases. However, the BST orientations parallel
to the BST/Pt interfaces are dissimilar. For type I, the atomic arrangement is along the [010]
direction, whereas [110] is the direction parallel to the interface for type II.
No amorphous areas at the interfaces can be detected in both Figs. 1 (a) and (b). The
amorphous regions were demonstrated to be the Ti-excess amorphous phase compared with
the BST film matrix [32]. The absence of such amorphous regions indicates that the effect of
excess titanium caused by the deposition process can be ruled out. On the other hand, Yoo et
al. reported that the crystallization of amorphous BST at the interfaces can distort the lattice
structure during the annealing process [31]. By calculating the angle between [001] and [010]
(89.6o, for type I) and the angle between [001] and [110] (89.5o, for type II), respectively, no
distorted structure is observed in the regions approaching the BST/Pt interfaces for both
samples. Along the [001] direction from the BST/Pt interfaces, the separation of the two
planes next to the interface is consistent with the separation of the planes more far away from
the interface. It is not the case reported by Jin et al. [27] In their work, the separation of the
last two planes was observed to be much larger than those in the film bulk, which was
attributed to Ruddlesden-Popper planar fault [35]. As a result, the above-mentioned effects
cannot fit this case very well.
Taking into account the above discussions, the study on the origin of interfacial layer is
hence focused on the relationships of the in-plane lattice structure between BST film and Pt.
Pt is an elemental metal with fcc structure. A schematic representation of [111] projected
lattice structure of Pt is shown in Figure 2 (a). The Pt atoms are arranged according to
equilateral triangle lattice structure in Pt (111) plane. The atomic separation between the
nearest neighbor Pt atoms is calculated to be 0.274 nm along the <110> direction. BST is a
typical perovskite dielectric with ABO3 structure. In such crystal structure, Ti ions are located
at the body center (B site), each surrounded by six oxygen ions at the face centers to structure
the oxygen octahedral, while the A site ions (Ba and Sr) occupy the corners. A-O planes
alternate with B-O planes along <001> direction to form the BST crystal structure. Figure 2
(b) and (c) show the schematic diagram of [001] projected lattice structure of BST, depicting
the B-O plane and the A-O plane, respectively. In B-O plane, B and O atoms are arranged
along the <010> direction with the B atomic separation of 0.394 nm. In A-O plane, the A
atomic separation is twice of the (110) interplanar distance that is determined to be 0.276 nm.
For ABO3 films, the first atomic plane grown onto the Pt can be either A-O plane or B-O
plane, as demonstrated by the atomic resolution HAADF/STEM investigations [19]. So, it can
be speculated that, if B-O plane is the first plane onto the Pt plane, according to the obtained
atomic separations, the two in-plane lattices are evidently mismatched. However, in the case
of A-O plane as the first plane, good matching can be expected with the mismatch smaller
than 1%, as the A atomic separation along <110> can be adequate to the twice that in Pt. The
effects of terminations on the interface energetics were theoretically demonstrated [36], where
different cases of Pt occurred when Pt films were grown on SrO- and TiO2-terminated
surfaces of SrTiO3.
Back to the crystallographic relationships between BST and Pt in Figs. 1 (a) and (b), the
origin of two distinct interfacial cases with identical normal crystal orientation can be
proposed based on the in-plane mismatch. The first grown atomic plane is crucial to
determining the crystal growth. As B-O plane is the first growth plane, [010] is the direction
parallel to the interface. Similar case was found in BST/LaAlO3 structures by the Z-contrast
STEM dark field image [33]. Therefore, large mismatch gives rise to a distorted lattice in
several atomic layers adjacent to the interface. It is the case of type I interfacial layer, as
illustrated in Figure 1 (a). When the growth initiates with A-O plane, type II interface occurs.
Good matching takes place as the atoms are arranged along the [110] direction to form the in-
plane lattice. However, it is also noticed that there is no good matching in the perpendicular
direction, while no indication of interfacial layer exists in this case. Thus, the results
presented herein imply that the quality of the BST/Pt interface can be improved when there is
a good film/electrode lattice matching at least in one direction. It can be expected that
appropriate fabrication of films on Pt electrodes can avoid the presence of interfacial layers
and obtain atomically sharp interfaces [28].
2. Approaches to Enhance the Tunability of BST Thin Films
Dielectric tunability and dielectric loss are the basic parameters characterizing
ferroelectrics for applications in electrically tunable devices [7,13-15]. Earlier viewpoint
considered the loss as the more critical issue in device applications. Much effort has focused
on lowering the dielectric loss in thin films, but no substantial reduction of loss has been
achieved yet. Moreover, such lowering of loss is usually at the expense of reduced tunability
[37,38].
More recently, Pervez et al. pointed out that [39], as in phase shifter circuit, higher
tunability allows for fewer tuning elements, which are cascaded to achieve the desired phase
shifting. The decrease of the amount of the tuning units directly reduces the net circuit loss.
As a result, the tunability should be more crucial in practical applications than the dielectric
loss. However, as discussed in the above section, dielectric properties degrade dramatically in
ferroelectric thin films. How to enhance the dielectric tunability is consequently the key issue
to be solved for the tunable devices based on ferroelectric thin films. In the following, we will
introduce some approaches to enhance the tunable properties based on the origin of the
tunability of perovskites.
2.1. Origin of Dielectric Tunability in Perovskites

Tunability of perovskites is defined according to the dielectric nonlinearity of as
functions of electric field above the Curie temperature. Ferroelectrics for applications in
electrically tunable devices are generally in the paraelectric phase [3,4,7-10,13]. The reason is
that most of the ferroelectrics in polar phase are also piezoelectric. Piezoelectric
transformations cause large losses at relatively low microwave frequencies, and additional
losses in polar phase are associated with the domain wall movements. Another reason
hindering the applications of a ferroelectric in a polar phase is the hysteresis in field-
dependent dielectric characteristics [7].
It is known that the displacive ferroelectrics have quite perceptible dielectric nonlinearity
at the temperature range of 50oC ~ 100oC above Tc [4]. Marked tunability takes place near the
phase transition point, which decreases with the increasing temperature. In particular, a wide
temperature range of nonlinearity can be obtained in material with a diffused phase transition.
The explicit functional relationship between dielectric constant and electric field for a
ferroelectric in the paraelectric state is derived by K. M. Johnson in 1961 based on Slaters
and Devonshires work [3,40,41]. In the light of the phenomenological theory of Devonshire
[41], the free energy F is expanded as [3]
(T ) 2 B(T ) 4
F (T , P) = F (T ,0) + P + P + ... , (3)
2 0 C 12
where T is the absolute temperature, P is the polarization developed by the external field,
F (T ,0) is the zero-field free energy, C and are the Curie-Weiss constant and temperature,
B(T) is a phenomenological constant, and 0 is the dielectric constant of vacuum. The
dielectric constant r follows Curie-Weiss law above the Curie temperature in the
paraelectric state. Since the dielectric constant of BST r (T , E ) >> 1 , we can express
r (T , E ) as [10]
1 E 1
0 = + 0 B (T ) P 2 + ... , (4)
r (T , E ) P T r (T ,0)
where the electric field is given by
F (T ) B(T ) 3
E = = P+ P + ... (5)
P T 0C 3
Figure 3.
In the case of small field, we can assume that P = 0 r E and neglect terms in P4 or
greater. Based on Eqs. 4 and 5, Johnson proposed the variation of dielectric constant r with
the electric field E as
r (T ,0)
r (T , E ) = . (6)
{1 + [ 0 r (T ,0)]3 B (T ) E 2 }1 / 3
This expression demonstrates the decreasing dielectric constant with increasing E. The
comparison of experimental and calculated data of r E characteristics is shown in Figure
3. The agreement between experimental and theoretical results, especially the data under
lower electric field, verifies the validity of illuminating the field-dependent dielectric constant
by using Eq. 6. If we rewrite Eq. 6 as
r (T , E ) 1
tunability = 1 =1 , (7)
r (T ,0) {1 + [ 0 r (T ,0)]3 B(T ) E 2 }1 / 3
we thus have the expression of tunability which shows dependence on r (T ,0) . The larger
r (T ,0) in the right site of Eq. 7 directly results in the higher tunability.
The above analyses is based on the thermodynamical theory, while G. Rupprecht et al.
and Johnson also discussed the origin of such field-dependent behavior according to the ion
oscillating in a lattice anharmonic potential in perovskites [3,42]. It has been concluded that
the anharmonic restoring forces on the Ti ion when it is displaced from its equilibrium
position, is responsible for the dielectric tuning in BST and STO. G. Rupprecht et al. also
evaluated the nonlinear constant of A( = 0 B(T )) in Eq. 6. Their theoretical results indicated
3
that the nonlinearity of the dielectric constant has a marked dependence on the direction of an
external field with respect to the crystallographic axes [42]. A is largest for the <100>
directions, while A takes the smallest value along the <111> directions. This direction-
dependent nonlinearity is consistent with the studies on ion displacements by Neville et al.
[43] They considered the crystal structure of perovskite STO and arrived at the conclusions
that Ti ion motion determines the nonlinearity of dielectric constant. Their results revealed
that the nearest-neighbor Ti-O distance decreases fastest if the Ti ion is displaced in the [001]
direction, and least if the Ti ion is displaced in the [111] direction. Consequently, the greatest
nonlinearity can be expected if Ti ion displacements are significant for electric fields applied
in the [001] direction, and the smallest one takes place for fields applied in the [111]
direction. It is thus evident that perovskite BST and STO films with (100) texture can exhibit
strongest tunable behaviors as the fields are generally applied to the capacitors in <100>
directions.
According to the above discussions, the dielectric tunable properties of perovskites are
highly associated with the phase transition of ferroelectrics, crystal texture, and lattice
anharmonic interactions. These accordingly provide us approaches to enhance the tunable
properties of perovskites thin films.
2.2. Effects of Electric Annealing on the Tunable Properties of BST Thin Films
It has been discussed in section 2.1 that the anharmonic restoring forces on the Ti ion
when it is displaced from its equilibrium position, is responsible for the dielectric tuning in
perovskites. For BST, the Ti ion is displaced toward the oxygen ions at the apices of the
octahedron, even in the paraelectric state, resulting in asymmetry of the Ti-O bond lengths
[44]. A large dielectric constant is often associated with a high value of tunability, because
when the dielectric constant is larger, the asymmetry of the bonds is greater to begin with, so
further displacement becomes even more difficult [44]. In addition, Hyun et al. reported that
the tensile strain along the direction of the applied field increases the ionic displacement, and
hence enhance the tunability [45]. All of these indicate that the displacement of the Ti ion
from its equilibrium position plays a crucial role in determining the tuning properties of BST
films.
Since the applied voltage has an impact on the ionic displacement, Xia et al. has
introduced an electric field to the process of thermal annealing in an attempt to improve the
tunable properties of BST thin films [46], whereafter this approach was further confirmed by
Cho et al. [47] Ba0.5Sr0.5TiO3 thin films derived by rf magnetron sputtering technique were
rapid thermal annealed with electric field applied. The set-up composed of two parallel plates
acting as electrodes for applying the electric field is placed in a quartz tube furnace to
experience the thermal annealing. The bottom plate is electric grounded, and voltages with
different values were applied onto the top plate. The voltage was still applied during the
cooling process. In addition, the annealing under 0 V is simplified as conventional annealing,
and the annealing under non-zero voltage is called as electric annealing.

The effects of electric annealing on the film microstructures have been investigated based
on the x-ray diffraction (XRD) patterns. The in-plane lattice constant is barely changeable as
the applied electric field [47], whereas the out-of-plane lattice constants are increased after
the films were subjected to electric annealing under both the positive and the negative voltage
applied, as shown in Figure 4. As the applied voltage increases, the out-of-plane lattice
constants also increase. But the degree of the (200) orientation, 200 , has no obvious change
with respect to the electric annealing. Here, 200 is defined as the relative intensity ratio
200 = I 200 /( I 200 + I110 ) , where I hkl is the integrated intensity of the (hkl) peaks recorded in
2 scans. This indicates that the electric annealing has no contribution for the
improvement of the film orientation and the crystallization compared to the conventional
annealing, but just results in a slight elongation of the lattice constant along the surface
normal direction. This elongation could be attributed to the enlarged ionic displacement in the
field direction caused by the electric field, considering the only difference between the
electric and conventional annealing is the applied electric filed. Furthermore, AFM
observations reveal that the electric annealing also affects the surface morphology of the BST
films. The electric annealing leads to slightly smooth surface compared to the conventional
annealing [47].
The dielectric properties of the BST films are correlated to the electric annealing. Figure
5 shows the capacitance-bias (C-V) characteristics of BST films annealed with applied
voltage of 200 V and 0 V, measured at room temperature. The zero-capacitance is found to
increase with the introduction of applied voltage when annealed. The calculated zero-voltage
dielectric constant are 345, 424 and 415 for the annealing voltage of 0 V, +200 V and 200 V,
respectively. This increase could be attributed to the elongated ionic displacement. A larger
zero-electric-field dielectric constant is often associated with a higher value of tunability [44].
The tunability is calculated from the C-V curves, and the tunability measured at +5 V as the
function of the annealing voltage is exhibited in Figure 6. The tunability of electric-annealed
BST films shows a noticeable increase compared to that of BST annealed without electric
field. The field dependence of the dielectric constant at paraelectric state is caused by the
anharmonic interaction of Ti ion. For the samples subjected to electric annealing, which have
elongated ionic displacement compared to those subjected to conventional annealing, the
asymmetry of the bonds is greater to begin with, so the ionic vibration and further
displacement become even more difficult when voltage is applied, and hence enhances the
tunability. Cho et al. also studied the dielectric loss of BST films at microwave range (1-3
GHz). As the applied voltage increases, the dielectric loss slightly increases, which can be
explained by mechanism of energy scattering on the ferroelectric mode that is connected with
the relaxation of phonon distribution function [47].
It is reported that the sputtered films deposited from a stoichiometric ceramic are often
Sr-deficient films and have O vacancies [48]. Ti4+ ions located near the O vacancy can
capture one electron and change to Ti3+. For the Sr-deficient films, some Ti3+ ions will occupy
the Sr vacancies [48,49]. When the films are annealed under an applied electric field, the Ti3+
ions could trend to distribute against the direction of electric field compared to the Ti4+ ions.
This results in the different distributing regions of Ti3+ and Ti4+ ions in the films, which
changes the former random ionic distribution. It can be hypothesized that the regions at which
Ti3+ ions located could be Ba-rich because the Ti3+ can occupy the Sr site, and display some
BaTiO3-like behavior and become polar, whereas other regions remain paraelectric at room
temperature. This indicates that a polar nano-region phase might appear in the BST films
during the electric annealing. The polar-regions have the tetragonal phase structures with
larger c-lattice constants, which is consistent with the calculated out-of-plane lattice
constants. The existence of the nano-polar-regions could be related to the relaxor-like
characteristics, and result in the hysteresis loops even in the paraelectric state at room
temperature [50]. The existence of polar phases in the annealed films has been confirmed by
the observed polarization hysteresis loops even though the electric-annealed films are still in
paraelectric state at room temperature. And hence, larger tunability could be obtained, as
expected, due to the large dielectric constants that relaxors usually have.
Furthermore, it is also noticed, in Figure 5, that the symmetry of C-V curves is destroyed
after the films electric-annealed. For the films annealed at +200 V, the negative part of the C-
V curve exhibits slightly stronger tuning property than the positive part, whereas the films
annealed at 200 V shows the opposite phenomenon. Such asymmetry of the C-V
characteristics might also result from the appearance of the nano-polar-regions in the
paraelectric state, since the similar asymmetries observed in ferroelectric films due to the
hysteresis effects.
2.3. Effects of A-site Doping on the Tunable Properties

The phase transition temperature (Tc) plays a critical role in determining the tunable
properties of perovskites. O. G. Vendik pointed out that marked tunability takes place near
the phase transition point [4], which decreases with the increasing temperature. So, if the Tc
can be modulated to be closer to the operation temperature, higher tunability could be
obtained. On the other hand, the microwave devices made of these tunable materials would be
operated at environment with different temperatures, requiring the tunability of thin films to
be low temperature dependent over a broad temperature range.
An effective approach to modulate the Tc is chemical substitution in such films.
Moreover, it is also known that adding more kinds of cations to the A-site in perovskite
structure (ABO3) usually results in diffuse phase transition, which can broaden the transition
peak and induce low temperature-dependence. Recently, the A-site doping of PbO, Bi2O3, and
CaO in (Ba,Sr)TiO3 thin films has been reported to modulate the Tc and result in enhanced
diffuse phase transition in BST [51-54]. Herein, Pb doping is introduced as a typical example.
Figure 7 shows the dielectric constant-temperature ( T ) characteristics of the Pb
doped BST (BPST) thin films at 1MHz. For comparison, the T behavior of BST films is
also shown in the inset of Figure 7. It is noticed that the dielectric constant of BPST exhibits a
maximum at the temperature of 7 oC, whereas, in the case of BST, the maximum occurs at
79 oC. The temperature corresponding to this peak of dielectric constant is considered to be
Tc. As is known, BST is a continuous solid solution between BaTiO3 (BTO) and STO.
PbTiO3 (PTO) has a higher Tc of 485 oC than that of BTO of 120 oC. Thats why, with the
adding of PTO into BST system, the Tc of the solid solution BPST could be adjusted to shift
to a higher temperature.
As more kinds of cations were added to the A-site in perovskite structure, enhanced
relaxor behavior and diffuse phase transition are to be expected for BPST films. Figure 7 also
gives the characteristics of T at different frequencies for BPST films. An obvious
increase of peak temperature with increasing frequency, together with the broadening of the
transition peak, is observed. This indicates a typical behavior of relaxor ferroelectrics.
However, no clear shift of Tc of BST films is found, as illustrated in the inset of Figure 7. It is
now known that the relaxor behavior is related to the heterogeneity of the nanoscale
composition [50]. As a result, adding Pb ions into the A-site of pervoskite structure of BST
system could be responsible for the relaxor behavior of BPST films. The regions with A-site
rich in Pb2+ would display some PTO-like behavior and become polar at room temperature,
whereas other regions could have a tendency to remain paraelectric (BST with Sr rich is
paraelectric at room temperature). The existence of nanopolar regions in paraelectric
environment results in the distribution of Tc in BPST films, giving rise to the broadening of
the transition peak and low temperature-dependent dielectric constant.

As the Tc of BPST films is much closer to the room temperature, the enhanced tunable
properties would be obtained at room temperature. The tunability is calculated from the
dielectric constant-voltage ( V ) characteristics by using the Eq. 1. Figure 8 shows the
V curves of the BPST and BST films, measured at room temperature and 1MHz. As
expected, the calculated tunabilty of BPST is much larger than that of the undoped BST.
Furthermore, such enhanced tunability could also be related with the existence of nanopolar
regions in BPST films, since the nanodomains may be affected by the external field through
changing the nonferroelectric states into ferroelectric ones [2,50,55]. In addition, the
dielectric loss is shown in the inset of Figure 8. The loss tan increases slightly, also due to
the Tc of BPST closer to the measured temperature.

As aforementioned, microwave devices require the materials to be stable against the

change of temperature. It has been discussed in section 2.1 that the dependence of tunability
on the temperature can be elucidated according to the phenomenological theory of
Devonshire [41]. Eq. 7 indicates that the decreased r (T ,0) directly results in the lower
tunability. The temperature dependence of the tunability (at 5 V) is presented in Figure 9, and
the inset shows the T characteristics (T larger than room temperature) without applied
bias. It is worth noting that, in the temperature range from 25 oC to 100 oC, the BPST films
still exhibits relatively high tunability from 60 % slightly down to 54 %. This no steep
variation of tuning property with regard to temperature could be attributed to the broadened
dielectric constant peak caused by the diffuse phase transition, as illustrated in the T
curve. Since the dielectric constant decreases slowly with the increasing temperature, the
relatively stable tunability can be obtained. Figure 9 also gives the theoretical data based on
Eq. 7 and T curve. The agreement between the experimental and the theoretical results
verifies the validity of illuminating the low temperature-dependent tunability by using Eq. 7.
These results demonstrated that Pb substitution can be a simple and effective method to
obtain high tunability and low temperature-dependence simultaneously. This stability of high
tuning property bodes well for tunable devices made of BPST films, which would be operated
at environment with different temperature.
2.4. Effects of Bilayer Thin Films on the Tunable Properties

The (100) orientation has been found to produce large dielectric constant and high
tunability [44,56]. BST films have been successfully fabricated on Pt coated silicon substrates
by sputtering [56], chemical vapor deposition (CVD) [57], and pulsed laser deposition (PLD)
[58]. Compared to other deposition methods, the sputtering process offers some special
characteristics, for example, controlling the Ar/O2 ratio and the total (Ar + O2) pressure
during the deposition could attain the (100) preferred orientation [56] and adjust the
stoichiometry [59]. A highly (100) preferred orientation has been developed with the
introduction of the compositional bilayer structure.
BST films with different Ba/Sr and (Ba + Sr)/Ti ratios have been fabricated using rf
magnetron sputtering by controlling the deposition conditions. The variation in thin film
composition can be controlled by the gas pressure in sputtering process [59,60]. BST films
with a Ba/Sr ratio of ~ 0.80 and (Ba + Sr)/Ti ratio of ~ 0.91 were deposited at 600 oC at a total
process gas pressure of 3 Pa, while at a total pressure of 9 Pa a Ba/Sr ratio of ~ 0.97 and (Ba +
Sr)/Ti ratio of ~ 0.98 was obtained. This makes it advantageous to fabricate BST
compositional multilayer films by stacking different chemical compositional layers under
different sputtering conditions.
In this approach, a bilayer film with the same individual thicknesses has been fabricated
according to the following procedure. A first layer with a thickness of 120 nm was deposited
at 600 oC at a total process gas pressure of 3 Pa on the Pt bottom electrode, and was
immediately annealed in situ at 600 oC in the same ambience for 30 min. Then, an upper layer
was deposited at 600 oC at a total process gas pressure of 9 Pa onto the lower layer with the
same thickness of 120 nm. For comparison, some monolayer BST films were deposited at
600 oC at a total pressure of 9 Pa. They have the same thickness as the 240 nm bilayer films.
The crystalline structures of BST bilayer and monolayer films, analyzed by x-ray
2 scans, are shown in Figure 10. The bilayer case exhibits a polycrystalline structure
with a highly (100) preferential orientation, while in the monolayer case the structure was
characterized by a strong (110) orientation. The (100)-orientation parameter, 100 , is
calculated from the relative heights of the (100) and (110) diffraction peaks, i.e.,
100 = I100 /( I100 + I110 ) , and 100 = 0.84 was obtained for the bilayer BST thin films. The
change in the orientation indicates that the crystallization and growth of the films are
significantly influenced by the introduction of the bottom layer. The (100) preferred
orientation of BST films could be attained by controlling the Ar/O2 ratio, the deposition gas
pressure, and the deposition rate during deposition in rf sputtering. As can be seen in Figure
10 (c), the bottom layers with an in situ annealing, after being deposited at a total gas pressure
of 3 Pa, show a structure with a strong (100) peak and a weaker (110) peak, which is similar
to the report of P. Padmini et al. [56] It is well known that at higher deposition pressures, the
atom mean free path can be reduced, which results in the lower deposition rate. The average
energies of the atoms forming the films will also be reduced, so that the growth orientation
may be more determined by the substrate structure [61,62]. So when the upper layer was
deposited onto the (100) preferred layer at a total gas pressure of 9 Pa, the bottom layer could
act as a nucleation site to determine the texture of the upper layer and a highly developed
(100) orientation could be obtained.
The (100) texture leads to high tunability in sputtered BST on Pt/TiO2/SiO2/Si substrates
[44,56]. Figure 11 (a) shows the C-V characteristics measured at 1 MHz of BST bilayer and
monolayer films. The zero-bias dielectric constant in bilayer case as calculated from the result
shown in Figure 11 (a) is 600, which shows a significant increase compared to that of the
monolayer of 387. The (100) orientation has a larger dielectric constant because Ti ion is
preferentially displaced along the (100) direction towards the oxygen ions at the apices of the
octahedron, even in the paraelectric state, resulting in the maximum polarization and
dielectric constants along this direction [44]. So in bilayer case the highly (100) oriented film
exhibits a considerable enhancement of the dielectric constant. A large dielectric constant is
often associated with a high value of the tunability. The tunability of the bilayer films
calculated from Figure 11 (a) was about 44% at an applied electric field of 210 kV/cm, which
is double the value of 21% for the monolayer films. It has been demonstrated that the
vibration of Ti ion in oxygen octahedral is responsible for the tuning of the BST [3,63]. The
displacement of the Ti ion induces the asymmetry in the Ti-O bond, which results in the
anharmonic restoring forces on the Ti ion. The more highly (100) texture is, the greater the
asymmetry of the bonds will be. As the Ti ion is displaced more from the center position, its
vibration is softened more easily with increasing applied voltage, and hence the tunability
increases.
(a)
(b)

(c)
Figure 12.
On the other hand, the interface between two successive dielectric layers must also be
considered for the dielectric enhancement, which was associated with an increased loss tan ,
from 0.014 to 0.024, as shown in Figure 11 (b). A compositional gradient can be formed at
this interface zone due to the different composition of the two layers. It will induce more
space charges to be located in this region. Hence, the resistivity of the interfacial material is
reduced, resulting in the increased leakage current density, as shown in Figure 11 (c). These
results can be correlated to a Maxwell-Wagner (MW) like dielectric enhancement effect. As
proposed by Catalan and ONeill et al [64,65], the dielectric enhancement of the mulitilayer
film system can be associated with the onset of MW behavior as long as a low resistivity
interfacial region is formed between two bulk-like insulating dielectric layers. The effective
structure is thus treated as consisting of interfacial, semiconducting regions intercalated
between a bulk-like, insulating ferroelectric [65]. Such a system is well described by the MW
series capacitor, which yields imaginary part of dielectric constant
( ) =
1 ( )
+ 0 2 2 , (8)
C0 (Ri + Rb ) 1 +
where is frequency, C0 is a geometric factor, Ri and Rb are resistances of the two dielectric
components, 0 and are dielectric constant at zero and infinite frequency, and is the
relaxation time of the entire MW capacitor [64]. Figure 12 illustrates the dependence of
on frequency. The of bilayer films increases with the decreasing frequency, which is the
typical MW behavior according to Eq. 8, whereas for the monolayer case does not exhibit
such dependence.
Some of the samples of bilayer were subjected to a long high-temperature furnace post-
annealing. Such a long annealing could result in the destruction of the bilayer structure. The
diffusion caused by the thermal annealing gradually eliminates the non-uniform composition
distribution at the interface, and so destroys the interface between the two layers. As
expected, the enhanced tuning properties were degraded after post-annealing at 800 oC for 1.5
h due to the destruction of the bilayer structure. This indicates that the interface has a more
dominating influence on the dielectric characteristics of the bilayer structure compared with
the orientation. The longer annealing time is, the lower the tunability and zero-bias dielectric
constant will be. The films after 2.5 h of annealing have a tunability and zero-bias dielectric
constant of 28% and 460, respectively, as shown in Figure 13, which are still larger than those
of the monolayer films because of the remaining highly (100) orientation.

3. Conclusions
Perovskites based on BST, because of their strong electric-field dependent dielectric

constant in the paraelectric state, are being used to develop a new class of electrically tunable
microwave devices. The dielectric tunability is a critical parameter for such applications. The
anharmonic restoring forces on the Ti ion when it is displaced from its equilibrium position in
the lattice of perovskites is responsible for the dielectric tuning in BST. Unfortunately, the
tunable dielectric properties degrade dramatically in perovskite thin films. As the dielectric
tunable properties are highly associated with the phase transition of ferroelectrics, crystal
texture, and lattice anharmonic interactions, in this section some approaches to enhance the
tunable properties of BST thin films have been introduced. It has been demonstrated that
post-annealing in proper electric field can increase the dielectric constant and the tunability
remarkably and destroy the symmetry of C-V characteristics of the films. The increased out-
of-plane lattice constant and the appearance of the hysteresis loops in the electric-annealed
films indicated the formation of small polar regions with tetragonal structure, which are
responsible for the increased dielectric constant and tunability. The substitution of Pb for Ba
in BST system has been revealed to raise the transition temperature close to room
temperature, which results in the enhanced tuning property. The adding of lead into BST
leads to the relaxor behavior and diffuse phase transition. The tunability of BPST films
consequently displays good stability against the temperature. This stability of high tuning
property bodes well for tunable devices made of BPST films, which would be operated at
environment with different temperature. In addition, a highly (100) preferred orientation can
be developed, and enhanced tunability is obtained with the introduction of compositional
bilayer structure. A Maxwell-Wagner like effect is also considered to result in enhanced
dielectric properties and increased loss tan . These results provide effective approaches to
obtain high tunability, and are of importance to develop BST thin films in the dielectric
tunable applications.
II. RESISTIVE SWITCHING IN PEROVSKITE THIN FILMS

Leakage behaviors in perovskite thin films are usually treated as an adverse factor for
memory applications. Efforts have focused on the reduction of leakage current in perovskite
thin films for DRAM and ferroelectric memory applications. This is because high leakage
current (or low resistance) generally leads to the loss of data. But, recently the leakage
behavior has been utilized to construct a novel memory concept of resistive switching
memory (RRAM), which has triggered scientific and technological attentions. The basic
principle of RRAM is that the conduction behaviors of a metal-insulator-metal (MIM) device
can be electrically modulated in a simple way. Consequently the different resistance states
(normally high- and low- resistances) controlled by the applied electric fields can act as data
of 0 and 1 in the memory processes. RRAMs, due to the nonvolatile characteristics and
the simplified device structure, are considered to be the most promising candidates for the
next-generation nonvolatile memory [66-69].
The candidate materials for RRAMs include chalcogenide semiconductors, organic
materials, and transition-metal oxides. Recently, such reversible switching behaviors were
also discovered in perovskite type materials, such as manganates, zirconates, and titanates,
which are of increasing interest during the last decade [70-75]. Well-developed two state or
multilevel switching performances have been reported not only in epitaxial thin films but also
in polycrystalline thin films. Efforts have been put forth to make clear the origin of the
switching characteristics in perovskite thin films. Several models have been proposed: (1)
charge-transfer processes via donor and acceptor levels (Cr3+ and Cr4+), (2) conductivity
changes related to the spontaneous polarization, (3) formation of local conducting filaments,
and (4) process of charge accumulation [70-78]. However, some inconsistent results still
remain in previous reported papers (for example, whether the interface or the film bulk
dominates the resistive switching) [74]. Several conduction mechanisms control the leakage
characteristics in dielectric thin films, and consequently can determine the resistance features.
Therefore, its worthwhile to further comprehend the switching mechanisms with regard to
the conduction characteristics in such films.
1. Resistive Switching Based on Space-charge-limited Current
More and more evidence indicates that space-charge-limited current (SCLC) plays an
important role in the conduction transition between high- and low- resistance states [79-83].
The Lamperts theory of SCLC predicts that valuable information can be obtained from the
current-voltage (I-V) dependence of the material [84]. A typical I-V characteristic of Zr doped
BST (BSZT) polycrystalline film is illustrated as a semilogarithmic plot in Figure 14. By
sweeping the voltage V, hysteresis of I-V curve is obtained, indicating different resistance
states occur during the voltage sweeping. It is noteworthy that a shift of the current minimum
to negative from zero voltage exists in forward I-V curve, demonstrating the presence of
internal field in the film due to nonuniform distribution of space charges contained in the
film. But, this shift vanishes when reverse swept, indicating the charge neutralization takes
place during the scan cycle. It can be assumed that negative charges, i.e. electrons, are trapped
in the film to diminish this negative shift.
Nonlinear I-V characteristics occur at higher voltages in forward curve, as shown in

Figure 14. Classic nonlinear conduction mechanisms include SCLC, Schottky emission, and
Poole-Frenkel (PF) emission. These mechanisms can be distinguished via the isothermal
logarithmic plot of I-V correlation. Figure 15 exhibits the replot of the positive branch of the
forward curve (from 0 to +3 V) in Figure 14 as a log(I)-log(V) curve. There are three regions
observed in this logarithmic curve: two linear regions for smaller voltages (< 0.7 V) and
higher voltages (>1.3 V) and a nonlinear increase region for the in-between voltages. This is
the well-known SCLC feature as described in Refs. [85 87]. According to the SCLC theory,
linear dependence for voltages smaller than 0.7 V is corresponded to the ohmic law region
[85,88]:
qnAV
I= , (9)
d
where q is the electronic charge, d is the film thickness, n is the concentration of charge
carriers, is the mobility, and A is the area of the electrode. This indicates that the injected
carrier concentration is lower than the equilibrium concentration at lower fields. At higher
fields (voltages larger than 1.3 V in Figure 15), a strong injection of the charge carriers in the
bulk of the film takes place, and current follows a square dependence on voltage: I V ,
2
which is corresponded to the Childs law region. In the presence of shallow trapping level, the
current is given by [85,88]:
9A r 0 2
I = V , (10)
8d 3
where is the ratio of the free electron density and the density of the filled trapping sites, r
is the dielectric constant of the film and 0 is the permittivity of free space. This equation
also holds well for an exponential distribution of traps. If sufficient charge is injected into the
insulator, all the traps will become filled. Further injected charge then exists as free charge in
the conduction band and contributes to the current [85].
The voltage values of 0.7 and 1.3 V denote the ohmic-boundary voltage and the square-
law onset voltage, respectively. The intermediate region represents the transition between
ohmic and SCLC behaviors, providing information about the trap distribution in the films.
Electron traps formed by defects in polycrystalline films locate below the conduction band
can absorb injected electrons [86]. Such defects in pervoskite films mainly come from oxygen
vacancies, and also from other types of defects, such as A-site vacancies and Ti4+ interstitials
[74]. When the voltage reaches the value to fill the traps, a large-voltage square-law
conduction occurs, and this voltage is defined as the so-called trap-filled-limit voltage (VTFL).
Such electron trapping is in accord with the above assumption about the shift of the current
minimum. Instead of the clearly defined VTFL, a small voltage range is found in Figure 15,
indicating the presence of traps exponentially distributed in energy [89], such as different
trapping levels below the conduction band edge in the B-O plane and the A-O plane in
perovskite dielectrics [90].
The slope of the log(I)-log(V) curve, i.e. d log( I ) / d log(V ) , is used for further
confirmation of the SCLC mechanism, as presented in Figure 16. Numbers from 1 to 4
represent four different regions in this slope curve. For small voltage (< 0.7 V), the slopes
approximately equal unity, indicating the ohmic behavior. When the voltage larger than 1.3 V
(region 3), d log( I ) / d log(V ) values are around 2, demonstrating the domination of SCLC.
The increasing slope value in region 2 is corresponded to the trap-filled process. Similar
treatment is executed for the positive branch of reverse sweep in Figure 14, and the
d log( I ) / d log(V ) values are ploted in the inset of Figure 16. These values around 1
through most of the voltage range are consistent with those in the highest voltage region
(region 4) in Figure 16, indicating that the conduction mechanism in these two voltage ranges
might be the same. This means that the difference between forward and reverse scan is
eliminated in region 4, and the low-resistance state is formed in this region. The dominant
conduction mechanism for the low-resistance state is thus concluded to be ohmic conduction,
which is coincident with the conclusions in other literatures (Refs. [72] and [81]). The other
two conduction models (PF emission and Schottky emission) are characterized by linear
ln(I / V ) V for PF and linear ln (I ) V for Schottky emission. No straight lines in
both ln ~ V curves can be found in the inset of Figure 15, indicating that these two
mechanisms can be ruled out or they play minor role in determining the I-V characteristics
than SCLC, if they exist. SCLC is a bulk-limited mechanism, demonstrating the bulk-limited
effect dominates the transition of resistance states in BSZT films.
Two resistance states are illustrated in Figure 17 as a V / I (resistance) V plot. The

zero voltage data were deleted in this figure due to V/I value equals 0 (at 0 V). It is clearly
seen that high- and low-resistance states exist during the voltage sweeping. The state
transition is characterized by the resistance drop over the voltage range from 0.7 to 2.2 V,
corresponded to region 2 and 3 in Figure 16. A nonlinear region is found in Figure 17 and
enlarged in the inset figures. The V/I vs V curve exhibits an inverse proportion characteristic
from 1.3 to 2.2 V, indicating the I V characteristic and further demonstrating the SCLC
2
mechanism. Beyond this region, the two resistance states are superposed, and the sample is
positioned in the low-resistance state.
2. Behaviors of Oxygen Vacancies Responsible for the Resistive Switching
It is now accepted that oxygen vacancies play critical role in the conduction in oxides
[75,79,91-93]. As above discussed, electron traps in films, mostly the oxygen vacancies in
perovskite oxides, were employed to structure the SCLC model for the elucidation of bulk-
limited effect. There are still issues to be further interpreted. As the BSZT films investigated
are polycrystalline in nature, it is known that the electron traps can locate in the grain bulk
and/or at the grain boundaries. It is required to make clear whether the grain bulk or the grain
boundary dominates the conduction. On the other hand, we still need to know the exact
behaviors of oxygen vacancies responsible for the high- and low- resistance states,
respectively.
Impedance spectroscopy (IS) is a powerful tool used to study the electrical characteristics
in oxide dielectrics [94-97], and is progressively utilized to analyze the resistive switching
[70,93,98]. Complex impedance spectroscopy can be employed to identify the response of
grain bulk from those of grain boundary and electrode interface by representing the samples
as an analogous circuit consisting of a series of parallel RC combinations. Different responses

of the grain bulk, grain boundary and electrode interface, each contributing to a parallel
combination of a resistance and a capacitance, correspond to the individual semicircle in the
complex plots [94-96]. Figure 18 shows the complex impedance plots of BSZT samples set at
low-resistance state in the temperature range from 20 oC to 400 oC. Although whole
semicircle cannot be obtained at room temperature due to the measurements at low frequency
beyond the frequency range of our equipment, only one semicircular arc can be detected in
the Z vs Z curves at higher temperatures. The absence of more than one semicircle certainly
demonstrates that the response arising from the electrode interface can be excluded, and
furthermore, indicates the conduction in such polycrystalline films is more controlled by the
grain bulk rather than the grain boundary effect [96].
The activation energies for the conduction in the low- and high- resistance states can be
derived from the variation of conductivity as a function of temperature. The conductivity
increases with the increasing temperature, as shown in Figure 19 (a). According to the
relationship
Ea
= 0 exp( ) (11)
kT
where is the conductivity, 0 is the pre-exponential factor, Ea is the activation energy, k

is the Boltzmanns constant, and T is the absolute temperature, the curve in Figure 19 (a) can
be fitted with two linear lines in low and high temperature regions. The calculated activation
energy for the low temperature region is 0.093 eV, and 0.44 eV for the high temperature
region.
(a)
(b)
It is well known that in the titanate-based perovskite oxides, the ionization of the oxygen
vacancies will create the conducting electrons
Oo Vo + 1 / 2 O2 (12)
Vo Vo. + e' (13)
Vo. Vo.. + e' (14)
or these electrons might be bonded to Ti4+ in the form of
Ti4+ + e' Ti3+. (15)
. ..
Here, the notation defined by Krger and Vink is adopted; Vo and Vo represent the
oxygen vacancies carrying one and two positive charges, respectively [99-101]. It has been
concluded by Chen Ang et al. that the oxygen vacancies lead to shallow level electrons,
which may be trapped by Ti4+ ions or oxygen vacancies and are easy to be activated
becoming conducting electrons [100].
In perovskite oxides, the activation energy of the first ionization of oxygen vacancies as
expressed in Eq. 13 was reported to be 0.1 eV [100-102]. In BSZT films, for the samples set
at low-resistance state, the obtained activation energy of 0.093 eV in the low temperature
region indicates that at room temperature, the first ionization of oxygen vacancies is
responsible for the conduction.
For comparison, the conductivity of BSZT samples positioned at high-resistance state is
also plotted in Figure 19 (b). Larger activation energies for conduction are obtained from the
ln vs 1000/T curve. If the electrical conduction is governed by the thermal excitation of

carriers from the second ionization of oxygen vacancies (as described in Eq. 14) to the
conduction band, the activation energy for conduction was demonstrated to be around 0.7 eV
[99,100], near the activation energy of 0.63 eV obtained in the low temperature region. This
implies that at high-resistance state, the conducting electrons at room temperature are
dominated by the second ionization of oxygen vacancies. In the higher temperature region,
the activation energy is calculated to be 0.98 eV. According to the literatures, the activation
..
energy for the motion of Vo in Ti-based perovskite materials is 0.9 - 1.1 eV [94,96]. This
means that the extra contribution to the conduction from the doubly ionized oxygen vacancies
leads to the increase in , which is the typical behaviors in such perovskite oxides.
Consequently, it can be concluded that in the low-resistance state, the first ionization of
oxygen vacancies determines the conduction in the low temperature region. The conduction
in the high-resistance state is the normal case for perovskite oxides, i.e. the second ionization
of oxygen vacancies is responsible for the conduction at low temperature and the motion of
Vo.. contributes to the conduction at higher temperature.
The electron trapping accountable for the transition of resistance states can be deduced in
the light of the defect formations and the obtained activation energies. Based on the above
discussions, at room temperature, the higher Ea for the second ionization of oxygen vacancies
compared to that for the first ionization indicates that it is more difficult to activate the
conducting electrons without any applied field, and hence the low free electron concentration
..
is in agreement with the high-resistance state at which the initial samples were set. Vo acts as
the carriers in such case and p-like behaviors can be expected [101]. After the samples were
set at low-resistance state, according to the first ionization process, the existence of Vo is in
accord with the assumption about the electrons trapped by the oxygen vacancies. The
activated conducting electrons increase the free electron concentration, representing the n-like
conduction. Such conductive transition can be confirmed by the different dc conduction
characteristics. The transition of SCLC, Ohmic conduction, and Schottky emission were also
demonstrated in other reports [81]. Accordingly this electron trapping effect and the transition
of conduction behavior can be expected to fit the resistive switching observed in other
perovskite oxides.
3. CONCLUSIONS
In this section, hysteretic I-V characteristics and resistive switching have been
demonstrated in perovskite BST films, which provides the potential applications of
perovskites in resistive switching memory. Detailed study of I-V characteristics reveals SCLC
is responsible for the transition of resistance states. The electron trapping and detrapping
process is proposed for interpreting the resistive switching. The domination of the grain bulk
effect on the resistive switching is demonstrated by the complex impedance spectroscopy.
The analyses of activation energies for the conduction in the low- and high- resistance states
have been executed comparatively. It is proposed that the first and the second ionization of
oxygen vacancies are responsible for the conduction in the low- and high- resistance states,
respectively. This electron trapping effect and the transition of conduction behavior can be
expected to fit the resistive switching observed in other perovskite oxides. It can be
hypothesized that resistive switching can take place in various oxides as long as electron-
trapping-assisted SCLC exists.
ACKNOWLEDGMENTS
This work was supported by a grant from the State Key Program for Basic Research of
China (2004CB619305-2010CB631004), National Natural Science Foundation of China
(10804048), and Research Fund for the Doctoral Program of Higher Education of China
(200802841003).
REFERENCES
[1] J. F. Scott, Annu. Rev. Mater. Sci. 28, 79 (1998).
[30] H. Diamond, J. Appl. Phys. 32, 909, (1961).
[31] K. M. Johnson, J. Appl. Phys. 33, 2826 (1962).
[32] O. G. Vendik, Ferroelectrics 12, 85 (1976).
[33] Q. X. Jia, J. R. Groves, P. Arendt, Y. Fan, A. T. Findikoglu, S. R. Foltyn, H. Jiang,
and F. A. Miranda, Appl. Phys. Lett. 74, 1564 (1999).
[34] L. A. Knauss, J. M. Pond, J. S. Horwitz, D. B. Chrisey, C. H. Mueller, and R. Treece,
Appl. Phys. Lett. 69, 25 (1996).
[35] S. S. Gevorgian, and E. L. Kollberg IEEE Trans. Microwave Theory Tech. 49, 2117
(2001).
[36] V. K. Varadan, D. K. Ghodgaonker, V. V. Varadan, J. F. Kelly, and P. Glikerdas,
Microwave J. 35, 116 (1992).
[37] R. W. Babbitt, T. E. Koscica, and W. C. Drach, Microwave J. 35, 63 (1992).
[38] A. Outzourhit, J. U. Trefny, T. Kito, B. Yarar, A. Naziripour, A. M. Hermann, Thin
Solid Films 259, 218 (1995).
[39] De Flaviis, N. G. Alexopoulos, and M. Staffsudd, IEEE Trans. Microwave Theory
Tech. 45, 963 (1997).
[40] S. Zafar, R. E. Jones, P. Chu, B. White, B. Jiang, D. Taylor, P. Zurcher, and S.
Gillepsie, Appl. Phys. Lett. 72, 2820 (1998).
[41] R. York, A.Nagra, E. Erker, T. Taylor, P. Periaswamy, J. Speck, S. Streiffer, O.
Auciello, Proceedings of the 2000 12th IEEE International Symposium on
Applications of Ferroelectrics 1, 195 (2000).
[42] X. X. Xi, H.-C. Li, W. Si, A. A. Sirenko, I. A. Akimov, J. R. Fox, A. M. Clark, and J.
Hao, J. Electroceramics 4, 393 (2000).
[43] X.H. Zhu, J. M. Zhu, S. H. Zhou, Z. G. Liu, N. B. Ming, S. G. Lu, H. L.-W. Chan,
and C.-L. Choy, J. Electron. Mater. 32, 1125 (2003).
[44] J. C. Shin, J. Park, C. S. Hwang, and H. J. Kim, J. Appl. Phys. 86, 506 (1999).
[45] C. S. Hwang, J. Appl. Phys. 92, 432 (2002).
[46] J. Q. He, E. Vasco, C. L. Jia, and R. H. Wang, Appl. Phys. Lett. 87, 062901 (2005)
[47] S. Schmidt, D. O. Klenov, S. P. Keane, J. Lu, T. E. Mates, and S. Stemmer, Appl.

Phys. Lett. 88, 131914 (2006).
[48] L. J. Sinnamon, M. M. Saad, R. M. Bowman, and J. M. Gregg, Appl. Phys. Lett. 81,
703 (2002).
[49] B. H. Park, E. J. Peterson, Q. X. Jia, J. Lee, X. Zeng, W. Si, and X.X. Xi, Appl. Phys.
Lett. 78, 533 (2001).
[50] I. A. Akimov, A. A. Sirenko, A. M. Clark, J. H. Hao, and X. X. Xi, Phys. Rev. Lett.
84, 4625 (2000).
[51] C. T. Black and J. J. Welser, IEEE Trans. Electron Devices 46, 776 (1999).
[52] C. Zhou and D. M. Newns, J. Appl. Phys. 82, 3081 (1997).
[53] L. J. Sinnamon, R. M. Bowman, and J. M. Gregg, Appl. Phys. Lett. 78, 1724 (2001).
[54] B. Chen, H. Yang, L. Zhao, J. Miao, B. Xu, X. G. Qiu, B. R. Zhao, X. Y. Qi, and X.
F. Duan, Appl. Phys. Lett. 84, 583 (2004).
[55] H. Z. Jin, J. Zhu, P. Ehrhart, F. Fitsilis, C. L. Jia, S. Regnery, K. Urban, and R.
Waser, Thin Solid Films 429, 282 (2003).
[56] M. Stengel and N. A. Spaldin, Nature 443, 679 (2006).
[57] C. L. Chen, J. Shen, S. Y. Chen, G. P. Luo, C. W. Chu, F. A. Miranda, F. W. Van
Keuls, J. C. Jiang, E. I. Meletis, and H. Y. Vhang, Appl. Phys. Lett. 78, 652 (2001).
[58] J. H. Chen, C. L. Lia, K. Urban, and C. L. Chen, Appl. Phys. Lett. 81, 1291 (2002).
[59] D. C. Yoo and J. Y. Lee, J. Cryst. Growth 224, 251 (2001).
[60] Y. L. Qin, C. L. Jia, K. Urban, R. Liedtke, and R. Waser, Appl. Phys. Lett. 80, 2728
(2002).
[61] H. -J. Gao, C. L. Chen, B. Rafferty, S. J. Pennycook, G. P. Luo, and C. W. Chu,
Appl. Phys. Lett. 75, 2542 (1999).
[62] Y. D. Xia, K. B. Yin, C. Xu, Y. Zhang, B. Xu, W. Y. He, X. K. Meng, J. Yin, and Z.
G. Liu, Appl. Phys. Lett. 92, 102906 (2008).
[63] S. N. Ruddlesden, and P. Popper, Acta Cryst. 11, 54 (1958).
[64] A. Asthagiri and D. S. Sholl, J. Chem. Phys. 116, 9914 (2002).
[65] M. W. Cole, P. C. Joshi, and M. H. Ervin, J. Appl. Phys. 89, 6336 (2001).
[66] M. W. Cole, P. C. Joshi, M. H. Ervin, M.C. Wood, and R. L. Rfeffer, Thin Solid
Films 374, 34 (2000).
[67] N. K. Pervez, P. J. Hansen, and R. A. York, Appl. Phys. Lett. 85, 4451 (2004).
[68] J. C. Slater, Phys. Rev. 78, 748 (1950).
[69] A. F. Devonshire, Phli. Mag. 40, 1040 (1949).
[70] G. Rupprecht, R. O. Bell, and B. D. Silverman, Phys. Rev. 123, 97 (1961).
[71] R. C. Neville, B. Hoeneisen and C. A. Mead, J. Appl. Phys. 43, 2124 (1972).
[72] B. A. Baumert, L. H. Chang, A. T. Matsuda, T. L. Tsai, C. J. Tracy, R. B.
Gregory, P. L. Fejes, N. G. Cave, W. Chen, D. J. Taylor, T. Otsuki, E. Fujii, S.
Hayashi, and K. Suu, J. Appl. Phys. 82, 2558 (1997).
[73] S. Hyun and K. Char, Appl. Phys. Lett. 79, 254 (2001).
[74] Y. D. Xia, J. B. Chen, B. Pan, D. Wu, X. K. Meng, and Z. G. Liu, Appl. Phys. Lett.
87, 052902 (2005).
[75] K.-H. Cho, C.-H. Lee, C.-Y. Kang, S.-J. Yoon, and Y.-P. Lee, Appl. Phys. Lett. 90,
162905 (2007).
[76] X. Wang, U. Helmersson, L. D. Madsen, I. P. Ivanov, P. Munger, S. Rudner, B.
Hjorvarsson, and J. E. Sundgren, J. Vac. Sci. Technol. A 17, 564 (1999).
[77] K. W. Blazey, R. Koch, and K. A. Muller, Mater. Res. Bull. 16, 1149 (1981).
[78] J. Zhai, X. Yao, L. Zhang, and B. Shen, Appl. Phys. Lett. 84, 3136 (2004).
[79] P. Sun, N. Matsuura, and H. E. Ruda, J. Appl. Phys. 96, 3417 (2004).
[80] Y. D. Xia, C. Cai, X. Y. Zhi, B. Pan, D. Wu, and Z. G. Liu, Appl. Phys. Lett. 88,
182909 (2006).
[81] W. Chen, X. Yao, and X. Y. Wei, Solid State Communications 141, 84 (2007).
[82] W. Qin, W. B. Wu, J. R. Cheng, and Z. Y. Meng, Materials Letters 61, 5161 (2007).
[83] S. G. Lu, X. H. Zhu, C. L. Mak, K. H. Wong, H. L. W. Chan, and C. L. Choy, Appl.
Phys. Lett. 82, 2877 (2003).
[84] P. Padmini, T. R. Taylor, M. J. Lefevre, A. S. Nagra, R. A. York, and J. S. Speck,
Appl. Phys. Lett. 75, 3186 (1999).
[85] H. J. Cho, and H. J. Kim, Appl. Phys. Lett. 72, 2820 (1998).
[86] S. G. Yoon, J. C. Lee, and A. Safari, J. Appl. Phys. 76, 2999 (1994).
[87] Im Jaemo, O. Auciello, P. K. Baumann, and S. K. Streiffer, D. Y. Kaufman, and A.
R. Krauss, Appl. Phys. Lett. 76, 625 (2000).
[88] Y. D. Xia, D. Wu, and Z. G. Liu, J. Phys. D 37, 2256 (2004).
[89] B. Chapman, Glow Discharge Processes Wiley (New York) p.203 (1980).
[90] J. A. Thornton, Ann. Rev. Mater. Sci. 7, 239 (1977).
[91] J. W. Liou, and B. S. Chiou, J. Am. Ceram. Soc. 80, 3093 (1997).
[92] D. ONeill, R. M. Bowman, and J. M. Gregg, Appl. Phys. Lett. 77, 1520 (2000).
[93] G. Catalan, D. ONeill, R. M. Bowman, and J. M. Gregg, Appl. Phys. Lett. 77, 3078
(2000).
[94] International Technology Roadmap for Semiconductors, 2007 Edition, www.itrs.net/
Links/2007ITRS/ExecSum2007.pdf.
[95] G. I. Meijer, Science 319, 1625 (2008).
[96] R. Waser, M. Aono, Nat. Mater. 6, 833 (2007).
[97] A. Sawa, Mater. Today, 11, 28 (2008).
[98] S. Tsui, A. Baikalov, J. Cmaidalka, Y. Y. Sun, Y. Q. Wang, Y. Y. Xue, C. W. Chu,
L. Chen, and A. J. Jacobson, Appl. Phys. Lett. 85, 317 (2004).
[99] A. Beck, J. G. Bednorz, Ch. Gerber, C. Rossel, and D. Widmer, Appl. Phys. Lett. 77,
139 (2000).
[100] J.-W. Park, K. Jung, M. K. Yang, J.-K. Lee, D.-Y. Kim, and J.-W. Park, J. Appl.
Phys. 99, 124102 (2006).
[101] Y. Watanabe, J. G. Bednorz, A. Bietsch, Ch. Gerber, D. Widmer, A. Beck, S. J.
Wind, Appl. Phys. Lett. 78, 3738 (2001).
[102] D. Choi, D. Lee, H. Sim, M. Chang, and H. Hwang, Appl. Phys. Lett. 88, 082904
(2006).
[103] K. Szot, W. Speier, G. Bihlmayer, R. Waser, Nat. Mater. 5, 312 (2006).
[104] R. Oligschlaeger, R. Waser, R. Meyer, S. Karthuser, and R. Dittmann, Appl. Phys.
Lett. 88, 042901 (2006).
[105] Y. Watanabe, Phys. Rev. B 59, 11257 (1999).
[106] M. Hamaguchi, K. Aoyama, S. Asanuma, Y. Uesu, and T. Katsufuji, Appl. Phys.
Lett. 88, 142508 (2006).
[107] Y. D. Xia, W. Y. He, L. Chen, X. K. Meng, and Z. G. Liu, Appl. Phys. Lett. 90,
022907 (2007).
[108] R. Dong, D. S. Lee, W. F. Xiang, S. J. Oh, D. J. Seong, S. H. Heo, H. J. Choi, M. J.

Kwon, S. N. Seo, M. B. Pyun, M. Hasan, and H. Hwang, Appl. Phys. Lett. 90,
042107 (2007).
[109] P. W. M. Blom, R. M. Wolf, J. F. M. Cillessen, and M. P. C. M. Krijn, Phys. Rev.
Lett. 73, 2107 (1994).
[110] L.-E. Yu, S. Kim, M.-K. Ryu, S.-Y. Choi, and Y.-K. Choi, IEEE Electron Device
Letters 29, 331 (2008).
[111] N. Das, S. Tsui, Y. Y. Xue, Y. Q. Wang, and C. W. Chu, Phys. Rev. B 78, 235418
(2008).
[112] M. A. Lampert and P. Mark, Current Injection in Solids (Academic, New York,
1970).
[113] P. C. Joshi and S. B. Krupanidhi, J. Appl. Phys. 73, 7627 (1993).
[114] G. R. Fox and S. B. Krupanidhi, J. Appl. Phys. 74, 1949 (1993).
[115] C. Sudhama, A. C. Campbell, P. D. Maniar, R. E. Jones, R. Moazzami, C. J. Mogab,
and J. C. Lee, J. Appl. Phys. 75, 1014 (1994).
[116] J. G. Simmons, in Handbook of Thin Film Technology, edited by L. I. Maissel and R.
Glang (McGraw-Hill, New York, 1970), p. 14.
[117] C. Sudhama, A. C. Campbell, P. D. Maniar, R. E. Jones, R. Moazzami, C. J. Mogab,
and J. C. Lee, J. Appl. Phys. 75, 1014 (1994).
[118] J. D. Baniecki, R. B. Laibowitz, T. M. Shaw, C. Parks, J. Lian, H. Xu, and Q. Y. Ma,
J. Appl. Phys. 89, 2873 (2001).
[119] M. Janousch, G. I. Meijier, U. Staub, B. Delley, S. F. Karg, B. P. Anderson, Adv.
Mater. 2007, 19, 2232.
[120] Y. B. Nian, J. Strozier, N. J. Wu, X. Chen, A. Ignatiev, Phys. Rev. Lett. 2007, 98,
146403.
[121] Y. D. Xia, Z. G. Liu, Y. Wang, L. Shi, L. Chen, J. Yin, X. K. Meng, Appl. Phys. Lett.
2007, 91, 102904.
[122] D. Wu, A. D. Li, and N. B. Ming, Appl. Phys. Lett. 84, 4505 (2004).
[123] D. C. Sinclair and A. R. West, J. Appl. Phys. 66, 3850 (1989).
[124] S. Saha and S. B. Krupanidhi, J. Appl. Phys. 87, 849 (2000).
[125] M. M. Kumar and Z. G. Ye, J. Appl. Phys. 90, 934 (2001).
[126] Y. H. You, B. S. So, J. H. Hwang, W. Cho, S. S. Lee, T. M. Chung, C. G. Kim, and
K. S. An, Appl. Phys. Lett. 89, 222105 (2006).
[127] K. Morii, H. Kawano, I. Fujii, T. Matsui, and Y. Nakayama, J. Appl. Phys. 78, 1914
(1995).
[128] Chen Ang, Zhi Yu, and L. E. Cross, Phys. Rev. B 62, 228 (2000).
[129] C. C. Wang and L. W. Zhang, Phys. Rev. B 74, 024106 (2006).
[130] J. Daniels and K. H. Hardtl, Philips Res. Rep. 31, 480 (1976).
Chapter 6
MICROWAVE DIELECTRIC PROPERTIES OF CA[(LI1/3A2/3)1-

XMX]O3- [A=NB, TA AND M=TI, ZR, SN] COMPLEX
PEROVSKITES: A REVIEW
George Sumesh and Thomas Sebastian Mailadil*

Materials and Minerals Division, National Institute for Interdisciplinary Science and
Technology (CSIR), Thiruvananthapuram - 695 019, India
ABSTRACT
The perovskite family of materials is of considerable technological importance for its
excellent temperature stable microwave properties for dielectric resonator based filters,
oscillators and antenna applications. In this chapter we review the preparation,
characterization and the microwave dielectric properties of Ca[(Li1/3A2/3)1-xMx]O3-
[A=Nb, Ta and M=Ti, Zr, Sn] dielectric ceramics. This family of perovskite materials
shows relative permittivity in the range 20 to 56 with a quality factor up to 45000 GHz
and temperature coefficient of resonant frequency (f) in the range from 21 to +83
ppm/C. The f can be tailored by adjusting the titanium content. The sintering
temperature can be lowered below 950 oC to suit LTCC by the addition of low melting
glasses.
1. INTRODUCTION
Microwave integrated electronics play a major role in communication systems. The
growth of satellite communication and cellular radio systems has led to the requirement of
compact, temperature stable, and low cost filters. In the early seventies, it was difficult to find
a device which could perform as filter, oscillator, and antenna with weight and size features
suitable for integration in miniature circuits. For instance, for frequency stabilization, one has
to rely upon the bulky expensive metallic cavities which were not integrable. These problems
*
Email: [email protected]
284 George Sumesh and Thomas Sebastian Mailadil
were alleviated with the advent of ceramic microwave dielectric resonators and have been
essential in eliminating metallic cavities and waveguides. The past 25 years have seen a rapid
development in the miniaturization of microwave circuitry and this has led to the
development of dielectric resonators which can replace the bulky metallic cavities. Recently,
microwave dielectric materials designed for wireless communication has become a booming
area of growth. Latest communication technologies including mobile multimedia systems,
ultra high speed wireless local area network and intelligent transport systems require highly
intelligent device that enable communication effectively between people and machines. The
dielectric resonators used in filters oscillators and antennas have become indispensable
component in microwave communication systems for its integration potential, size reduction,
reliability and temperature stability [1]. The materials for dielectric resonators are required to
excel in three characteristics, a high dielectric constant, a high quality factor, and a nearly
zero temperature coefficient of resonant frequency. High dielectric constant is essential since
it is the key factor for miniaturization and can thus lower the cost of production. The higher
the quality factor, the lower is the insertion loss and steeper the cut off. Moreover, a high
quality factor provides a narrow band width and hence enables the allocation of more number
of channels within a given frequency range [1-4].
From a historical perspective, guided electromagnetic wave propagation in dielectric
media received much attention in the early days of researching microwaves [5]. The
possibility of using unmetallized dielectrics as resonators was first proposed by Ritchmyer [6]
in 1939. The search for dielectrics for microwave applications starts from 1960. The first
reported work on the adaptability of a high dielectric constant dielectric rod to generate
standing wave resonance phenomena was proposed by Schlike [7]. Okaya reported that a
piece of rutile (TiO2) acted as a resonator and subsequently Okaya and Barash analyzed the
different modes of a dielectric resonator [8, 9]. The first microwave filter proposed by Cohen
[10] in 1968 where TiO2 was used as the dielectric resonator. The rutile TiO2 has a high
dielectric constant (r = 100), high quality factor (Qxf = 10000 GHz), and high temperature
coefficient of resonant frequency (f = +450 ppm/OC). This filter was not put into practical use
because of its large f. A real breakthrough in microwave ceramic technology occurred when
the first temperature stable ceramic dielectric barium tetratitanate developed by Raytheon
[11] in 1970. Later, a modified barium titanate with improved performance was reported from
BEL laboratories [12]. The next major breakthrough came from Japan when the Murata
Manufacturing Company produced (Zr-Sn)TiO4 ceramics with excellent microwave dielectric
properties [13, 14]. These positive results led to an intensive research and actual
implementation of dielectric resonator as the indispensable component in the microwave
communication. With the revolution in mobile phone and satellite communication systems
using microwave as propagation media, the research and development in the field of
microwave dielectrics for device miniaturization and excellent performance has been one of
the biggest challenges in the area of materials science.
Microwave Dielectric Properties of Ca[(Li1/3A2/3)1-xMx] 285
Table 1. Microwave dielectric properties and sintering temperatures of

representative materials
Material Permittivity Qxf (GHz) f Sintering Reference

Temperature (oC)
MgTiO3 CaTiO3 20 55000 0 1400 [15]
BaTi4O9 37 22700 15 1300 [16, 17]
Ba2Ti9O20 39 32000 2 1350 [12]
(Zr1-xSn x)TiO4 39 51500 0.7 1600 [13, 14]
Ba(Zn1/3Ta2/3)O3 28 168000 0.5 1500 [77-80]
Ba(Mg1/3Ta2/3)O3 24 430000 8 1640 [29, 78-80]
BaO-Sm2O3-TiO2 77.5 11200 -3.4 1360 [19]
BaO-Nd2O3-TiO2 83 10500 70 1450 [81]
CaTiO3 - NdAlO3 41 33000 -17 1450 [21, 22]
SrTiO3 - LaAlO3 39 32000 2 1450 [20, 23]
The microwave dielectric resonator materials with dielectric constant from 20 to about 80
with high quality factor and low f are now available for practical use. The extensive research
in the last three decades provides several extremely low loss ceramics suitable for resonator
applications. The properties of the important DR materials such as MgTiO3CaTiO3 [1, 15],
BaTi4O9 [16, 17], BaTi9O20 [12], (Zr1-xSnx)TiO4 [1, 13, 14], Ba(Zn1/3Ta2/3)O3 and
Ba(Mg1/3Ta2/3)O3 [18], BaO-RE2O3-TiO2 [19], SrTiO3-LaAlO3, CaTiO3-LaAlO3 [20-23] are
shown in table 1. Thirty years ago Ichinose et al. [24] investigated the A site modified (Sr1-x
Cax)[(Li1/4Nb3/4)1-yTiy]O3 complex perovskite ceramics. This was the first low loss lithium
based complex perovskite system and x = y = 0.08 composition has good dielectric properties
of r = 43.2, Q = 3500 GHz at 9 GHz, and f = -30 ppm/C when sintered at 1300 C.
The current trend and the state of the art of microwave dielectric materials for
telecommunication applications is discussed in the book Dielectric materials for wireless
communications [1] authored by Sebastian. This book reports the sintering temperature and
microwave dielectric properties of more than 2000 low loss dielectric ceramic compositions
and discusses the ways to tailor their properties by possible substitution, dopant and glass
additions. This book covers the important low loss microwave dielectric materials of the type
zirconium tin titanate, BaO-TiO2 systems, pseudo-tungsten bronze, perovskite type ceramics
(1:1 cation ordering, 1:2 cation ordering, and the cation deficient complex perovskites) and
the other materials such as tungstates, silicates, aluminates and spinals etc. Among the various
type of microwave dielectric ceramics, perovskite structure has attracted much attention for
their excellent microwave dielectric properties. The complex perovskite family shows the
relative permittivity in the range 25-40 with high quality factor and low temperature
coefficient of resonant frequency. Compared to other complex perovskite materials,
Ba(Zn1/3Ta2/3)O3 and Ba(Mg1/3Ta2/3)O3 show excellent microwave dielectric properties and
have been extensively studied [1]. However, they are usually made up of expensive chemicals
such as tantalates and a high sintering temperature (1500 oC) is necessary for obtaining good
densification and microwave dielectric properties. It is well known that the cation ordering
affects the dielectric properties, especially the quality factor of the complex perovskites [1].
In order to obtain a high quality factor, a high sintering temperature together with a long
annealing time is necessary for these complex perovskite systems. Hence, these resonators are
very expensive and not cost effective. The above problem can be solved by the development
of low temperature sinterable low cost materials or by lowering the sintering temperature of
the microwave ceramics with reasonably good dielectric properties. Most of the low loss
dielectric materials have high sintering temperature [1]. Earlier reports show that addition of
sintering aid is the best method to reduce the sintering temperature. Low temperature co-fired
ceramics (LTCC) has been identified as an enabling solution for the fabrication of
miniaturized microwave devices and modules. An advantage of LTCC technology includes
shrinking circuit dimensions, high level of passive integration, exceptional high frequency
performance, excellent reliability and low cost [25]. However, from table 1 it can be observed
that the currently available high Q materials are having high processing temperature.
However, in most of the high temperature processed high Q complex perovskites, a drastic
degradation of microwave dielectric properties is observed with the reduction of sintering
temperature by the glass addition [26, 27]. Reports show that, large amount of glass is
necessary to lower the sintering temperature of the ceramics having high processing
temperature [26-31]. As the amount of glass increases, the quality factor decreases. Hence the
selection of the low loss ceramics having low processing temperature is important for glass
addition. From the device design point of view, a good combination of high permittivity, high
quality factor, nearly zero temperature coefficients of resonant frequency and low processing
temperature are required for practical applications. Hence the development of good
microwave dielectric materials with optimal balance of these properties is one of the major
challenges in the microwave communication industry.
Recently, Choi et al. [32, 33] proposed a nonstoichiometric Ca(Li1/3Nb2/3)O3- complex
perovskite ceramic which can sinter at 1150oC with excellent microwave dielectric properties.
This system is multiphase and the substitution of Sn4+, Zr4+ and Ti4+ for [Li1/3Nb2/3]3.67+ by
forming Ca[(Li1/3Nb2/3)1-xMx]O3- [M=Ti, Zr, Sn] give single phase perovskites. This family
of materials with excellent dielectric properties can be a promising candidate for low
temperature sintering ceramics because they can be sintered at 1150oC without sintering aid.
However, the sintering temperature of this material is relatively high to use silver or copper as
an internal electrode in multilayered structures and further there exist the possibility of
evaporation of volatile lithium at such high sintering temperatures. Although some of the Li
based ceramics have excellent dielectric properties, the volatile Li has a deleterious effect on
the dielectric properties. The deficiency of Li leads to decrease in density and lattice defects
which adversely affect the dielectric properties. This problem can be minimised by muffling
the pellets in its own calcined powder or by lowering the sintering temperature by using a
sintering aid. Several authors [34-45] tried to minimize the escape of volatile lithium by
lowering the sintering temperature by the addition of low melting additives such as B2O3,
Bi2O3, and glass frit and also tried to tailor the dielectric properties by suitable substitution.
The present report reviews the microwave dielectric properties of Ca[(Li1/3A2/3)1-xMx]O3-
[A=Nb, Ta and M=Ti, Zr, Sn] ceramics and the possible substitution and glass addition.
2. EXPERIMENTAL
The Ca[(Li1/3Nb2/3)1-xTix]O3- [CLNT] and Ca[(Li1/3Ta2/3)1-xTix]O3- [CLTT] ceramic
samples were prepared by the conventional solid state ceramic route. High purity CaCO3,
Li2CO3, TiO2 (99.9+%, Aldrich chemical company, Inc, Milwaukee, WI, USA) Ta2O5 and
Nb2O5 (99.9+%, NFC Hyderabad, India) were used as the starting materials. Stoichiometric
amounts of the powder samples were mixed and ball milled using zirconia balls in ethanol
medium for 24 hours. Li2CO3 dissolve in water and hence ethanol was used as the mixing
medium. The resultant slurry was then dried and calcined. The calcinations were carried out
at temperatures in the range 900 to 1050 oC for four hours. The calcination temperature is
optimized for the best density and the dielectric properties. The sintering temperature of the
CLNT/CLTT ceramics were lowered by the addition of different borosilicate glasses such as
lithium borosilicate (35.5:Li2O, 31.66:B2O3, 33.2:SiO2)(LBS), lead borosilicate
(40:PbO,40:B2O3, 20:SiO2) (PBS), barium borosilicate (30:BaO,60:B2O3, 10:SiO2)(BBS),
zinc borosilicate (60:ZnO,30:B2O3, 10:SiO2)(ZBS), bismuth zinc borosilicate (35:Bi2O3,
32:ZnO, 27:B2O3, 6:SiO2)(BZBS) and lithium magnesium zinc borosilicate (20:Li2O,
20:MgO, 20:ZnO, 20:B2O3, 20:SiO2) (LMZBS). The corresponding glass powders were
synthesized from high purity oxide chemicals of B2O3, Li2CO3, SiO2, PbO, BaCO3, Bi2O3,
(MgCO3)4 Mg(OH)2 5H2O and ZnO (99.9 % Aldrich chemical company, U.S.A). These
oxides were weighed stoichiometrically and were mixed in an agate mortar for 2 hours using
ethanol as the medium. They were dried, melted in platinum crucible above their melting
point, quenched and powdered. Different weight percentages of various glasses were added to
the fine powder of calcined CLNT and CLTT ceramics. Polyvinyl alcohol (PVA) (Molecular
Weight 22000, BDH Lab Suppliers, England) solution was then added to the powder, mixed,
dried and ground well and pressed into cylindrical disks of about 14 mm diameter and 7mm
thickness, by applying a pressure of about 100 MPa. These compacts were muffled by CLNT
or CLTT powder of the same composition and sintered at different temperatures in the range
875oC 1200 oC for 4 hours. The sintering temperature was optimized for the best density
and the dielectric properties. The samples were muffled with powders of same composition to
prevent the escape of volatile lithium at elevated sintering temperatures. The crystal structure
and phase purity of the powdered samples were investigated by X- ray diffraction technique
using Ni-filtered Cu-K radiation (Rigaku Dmax-I, Japan, diffractometer). The
microstructures of the sintered samples were studied using scanning electron microscope
(JEOL-JSM 5600 LV, Tokyo, Japan). The sintered density of the specimen was measured by
the Archimedes method. The microwave dielectric properties were measured in the frequency
range 4 to 6 GHz by a Vector Network Analyzer (8753 ET, Agilent Technologies). The
dielectric constant and unloaded quality factor of the samples were measured by Hakki and
Coleman [46] and cavity method [47] respectively. Shrinkage characteristics were studied
using Dilatometer (TMA- 60 H, Shimadzu, Kyoto, Japan).
CLNT Muffled
CLNT Without muffling
CLNT+ 5 wt% of Li2CO3(without muffling)
CLNT+10 wt% of Li2CO3(without muffling)
CLNT+15 wt% of Li2CO3(without muffling)
21200
19200
19150
18450
38.1
38.4
38.3
28.8
38.6
9000
95.6
96.6
95.6
96.0
89.6
r Quxf (GHz) Density (%)
Figure 1. Relative density and dielectric properties of Ca[(Li1/3Nb2/3)0.8Ti0.2]O3- ceramics with different
sintering condition.
96.5 (a) 39 97 (b) 36.8
96.0
Relative Permittivity (r)
Relative Density (%)

38
95.5 96 36.4
37
95.0
36
94.5 95 36.0
35
94.0
Relative Permittivity (r) Relative Density
93.5 34 94 35.6
Relative Permittivity
22 3
Quxf (x10 GHz) 3
o
f (ppm/ C)
(a) 10
(b) Quxf (x 10 GHz)
o
-2
21 29 f ppm/ C
Quxf (x 103GHz)
20 -4
Quxf (x103GHz)
8
f (ppm/ C)
28
f (ppm/ C)
19 -6
o
6
18
27 -8
17 4
-10
16 26
850 900 950 1000 1050 900 925 950 975 1000 1025 1050
o
Calcination Temperature ( C) Calcination Temperature ( C)
o
Figure 2. Variation of densification and microwave dielectric properties of (a) CLNT and (b) CLTT
ceramics as a function of calcination temperature.
44 99
96 (a) (b) 36
40 96

93

93 34
36
90 90
32 Relative Permittivity (r)
32
87 Relative Permittivity (r) 87
28
84 30
84 24
81
22 3 30 Quxf (GHz)
Quxf (x10 GHz)
o
o f ppm/ C
f (ppm/ C) 27
20 10
(b) -3
24
18 (a)
Quxf (x103GHz)
f (ppm/ C)
Quxf (x10 GHz)
f (ppm/ C)
21
o
3
16
-6
18
14 6
15
12 -9
4 12
10
1125 1150 1175 1200 1225 1125 1150 1175 1200 1225
o o
Sintering Temperature ( C) Sintering Temperature ( C)
Figure 3. Variation of densification and microwave dielectric properties of (a) CLNT and (b) CLTT
ceramics as a function of sintering temperature.
3. CA(LI1/3A2/3)O3- COMPLEX PEROVSKITE [A=NB, TA]
The Ca(Li1/3A2/3)O3- [32, 38] is a nonstoichiometric perovskite and is multiphase when

prepared by the conventional solid state ceramic route. The X- ray diffraction pattern of
Ca(Li1/3A2/3)O3- [A = Nb, Ta] [32, 38] can be indexed based on CaTiO3 orthorhombic
perovskite structure with a small amount of second phase. However, substitution of Sn4+ and
Ti4+ for [Li1/3Nb2/3]3.67+ by forming Ca[(Li1/3Nb2/3)1-xMx]O3- [M=Ti, Sn] give single phase
perovskites in the vicinity of x=0.2 [32]. The Ca[(Li1/3Nb2/3)1-xMx]O3- [M=Ti, Sn] dielectric
resonators were prepared by calcining the ball milled raw powders at temperatures in the
range 900-1000 oC/4h and sintering at about 1150 oC/4hrs [32, 38]. It was reported that
lithium is volatile and escape during sintering at elevated temperature [1, 32, 38, 43].
Muffling of the samples with calcined powders of the same composition or keeping the
pellets in closed box is the best method to avoid the volatilization of lithium. It is also
possible to control the escape of lithium by adding extra lithium in the parent composition.
Recently, Sumesh et al. [48] compares the microwave dielectric properties of the
Ca[(Li1/3Nb2/3)0.8Ti0.2]O3- ceramics with different sintering condition such as open sintering,
muffling and different wt% of lithium addition to the parent material. Figure 1 shows the
densification and the microwave dielectric properties of Ca[(Li1/3Nb2/3)0.8Ti0.2]O3- ceramics
with different sintering condition such as muffling, without muffling, and addition of extra
Li2CO3 to the parent material. It is found that the sample sintered without muffling has
inferior dielectric properties compared to other sintering conditions. Addition of extra lithium
in the parent material shows dielectric properties comparable to that of muffling. However,
there is a possibility of the formation of second phase with the addition of extra lithium.
Hence muffling is the best method to avoid the volatilization of low melting elements.
1. 3.1. Optimization of Calcination and Sintering Temperatures
The microwave dielectric properties not only depend on the chemical composition but
also on the calcination and sintering temperatures. Figures 2 and 3 show the optimization of
calcination and sintering temperatures of Ca[(Li1/3Nb2/3)0.8Ti0.2]O3- (CLNT) and
Ca[(Li1/3Ta2/3)0.7Ti0.3]O3- (CLTT) ceramics. These compositions have been selected for its
phase purity and excellent temperature stable microwave dielectric properties. The calcination
and sintering temperatures were optimized for best density and dielectric properties. It can be
observed that the increase in the calcination temperature improved the physical and electrical
properties. The best physical and dielectric properties are observed for the compounds
calcined at 975 oC/4hrs. Further increase in the calcination temperature degraded the physical
and dielectric properties. Unreacted materials in under-calcined powder acted as an inhibitor
of grain growth during sintering [43]. It is reported that higher calcination temperature
lowered the reactivity of ceramic powders resulting in the increase in the number of pores.
These could be the reason for the degradation of densification and dielectric properties below
and above the optimized calcination temperature. Figure 3 shows the variation of the density
and dielectric properties of Ca[(Li1/3Nb2/3)0.8Ti0.2]O3- and Ca[(Li1/3Ta2/3)0.7Ti0.3]O3- ceramics
as a function of sintering temperature. It is found that the density and dielectric properties
improve with increase in the sintering temperature. The increase in the permittivity and
quality factor with sintering temperature is attributed to the improvement in densification. On
the other hand the dielectric properties were degraded due to the formation of pores and
escape of volatile lithium at higher sintering temperature (>1000 oC). The optimized sintering
temperature of Ca[(Li1/3Nb2/3)0.8Ti0.2]O3- (CLNT)and Ca[(Li1/3Ta2/3)0.7Ti0.3]O3- (CLTT)
ceramics are 1175 oC and 1200 oC respectively for the best density and dielectric properties. It
may be noted that the calcination and sintering temperatures may vary slightly with the
particle size, purity and origin of the initial raw materials.
3.2. Phase Analysis of Ca[(Li1/3A2/3)1-xMx]O3- [A=Nb, Ta and M=Ti, Zr, Sn]

Ceramics
Figures 4, 5, and 6 show the X ray diffraction patterns of Ca[(Li1/3Nb2/3)1-xMx]O3-

[M=Ti, Zr, Sn] ceramics sintered at 1150 oC with increasing amount of Ti, Sn, and Zr
respectively. The XRD pattern could be indexed based on the CaTiO3 type orthorhombic
perovskite with four formula per unit cell. At x = 0, the composition Ca(Li1/3Nb2/3)O3- is
multiphase. However, the stability of perovskite phase increased with increasing the Ti, Sn, or
Zr content and a single phase material is obtained in the range 0.2 x 0.3. Further increase
of Ti, Sn, or Zr concentration results in the formation of second phases. A systematic shift in
the peak positions towards higher 2 values observed with increasing Ti substitution (see
figure 4) [32]. However, the XRD peaks shift to lower angle with an increase in Zr and Sn
substitution (see Figure 5 and Figure 6) [33]. This could be due to the higher ionic radii of
(Li1/3Nb2/3) (0.066 nm) compared to Ti (0.056 nm) and lower ionic radii compared to Sn
(0.069 nm) and Zr (0.072 nm). The tantalum analogue also has interesting microwave
dielectric properties [38]. The Ca(Li1/3Ta2/3)O3- ceramics is also a multiphase compound with
a secondary phase of Li3TaO4. In order to tune the f and to lower the sintering temperature,
Liu et al. [38] made a partial replacement of (Li1/3Ta2/3) by Ti and added B2O3 in
Ca(Li1/3Ta2/3)O3-. The substitution of smaller Ti4+ for (Li1/3Ta2/3) decreased the cell volume.
X-ray diffraction study showed that substitution of Ti for (Li,Ta) stabilized the orthorhombic
perovskite phase but decreased the 1:2 ordering (see Figure 7). The ordering peak disappeared
with Ti substitution. In addition to the formation of single phase compound, the substitution
of Ti4+ in the B site of Ca[(Li1/3A2/3)]O3- [A = Nb, Ta] has an additional advantage of tuning
the temperature coefficient of resonant frequency to zero.
Figure 4. X-ray diffraction pattern of Ca[(Li1/3Nb2/3)1-xTix]O3- ceramics sintered at 1150oC as a function

of (1) x=0.0, (2) x=0.1, (4) x=0.2, (6) x=0.5. (after ref . 32).
Figure 5. X-ray diffraction pattern of Ca[(Li1/3Nb2/3)1-xSnx]O3- ceramics sintered at 1150oC as a

function of (1) x=0.0, (2) x=0.1, (3) x=0.15 (4) x=0.2, (5) x=0.3. (after ref. 32)
3.3. Microwave Dielectric Properties of Ca[(Li1/3A2/3)1-xMx]O3- [A=Nb, Ta

and M=Ti, Zr, Sn] Ceramics
Figure 8 shows the relative permittivity of Ca[(Li1/3Nb2/3)1-xMx]O3- [M=Ti, Zr, Sn] [32,
33, 38] ceramics sintered at 1150 oC with increasing concentration of Ti, Sn, and Zr. The
relative permittivity decreases from 29.6 to 23.5 with increasing Sn concentration from 0 to
0.3. It is expected that the decrease of relative permittivity by the substitution of Sn occurs
because a slightly large Sn4+ ion (0.069 nm) is substituted to a B site (Li1/3Nb2/3) having
slightly smaller ionic radii (0.066 nm). However, the relative permittivity increases with an
increase in the substitution of Ti ions in the B site (Li1/3Nb2/3). For Ti substitution, the relative
permittivity increases from 29.6 to 55.2 with an increase in x from 0 to 0.50. The increase in
the relative permittivity could be due to the smaller ionic radii of Ti4+ ion (0.056 nm)
incorporated to a B site (Li1/3Nb2/3) with larger ionic radii (0.066 nm) which may distort the
unit cell. Hence it is easily displaced under electric field and the relative permittivity
increases. A similar trend has been reported in the literature [49-52]. A slight decrease in the
relative permittivity is expected for the Zr substituted system because of the higher ionic radii
of Zr4+ (0.072 nm) compared to (Li1/3Nb2/3) (0.066 nm). However, a slight increase of relative
permittivity from 29.6 to 31.1 has been observed with an increase in x from 0 to 0.30. This
could be due to the higher ionic polarizability of Zr4+ (0.325 nm) compared to (Li1/3Nb2/3)
(0.305 nm).
Figure 9 shows the quality factor of Ca[(Li1/3Nb2/3)1-xMx]O3- [M=Ti, Zr, Sn] ceramics
with increasing the amount of Ti, Sn, and Zr substitution. The quality factor increased from
40000 to 52100 GHz with increasing Sn substitution from x = 0 to 0.15 and then decreased
from 52100 to 46000 GHz with increasing Sn substitution from x = 0.15 to 0.3. It is worth
noting that Sn substitution significantly improves the quality factor. The quality factor
decreased from 40000 to 18600 GHz with increasing Ti for x= 0 to 0.5. Similarly for Zr
substitution, the quality factor decreases from 40000 to 27100 GHz with increase in x from 0
to 0.30.
Figure 6. X-ray diffraction pattern of Ca[(Li1/3Nb2/3)1-xZrx]O3- ceramics sintered at 1150oC as

a function of (2) x=0.1, (4) x=0.2, (5) x=0.3. (after ref. 33)
Figure 7. X-ray diffraction pattern of Ca[(Li1/3T2/3)1-xTix]O3- ceramics doped with 3 wt% of B2O3
sintered at 1050oC as a function of (a) x=0.05, (b) x=0.1, (c) x=0.15, (d) x=0.2, (e) x=0.3, (f) x=0.5.
(After ref. 38)
60 Sn
Ti
Zr
50
40
30
20
0.0 0.1 0.2 0.3 0.4 0.5
x
Figure 8. Variation of Relative permittivity of Ca[(Li1/3Nb2/3)1-xMx]O3- [M=Ti, Sn, Zr] ceramics
sintered at 1150oC as a function of x.
50000
Sn
Ti
Zr
Quality factor (Qxf) GHz

40000
30000
20000
0.0 0.1 0.2 0.3 0.4 0.5
Figure 9. Variation of quality factor of Ca[(Li1/3Nb2/3)1-xMx]O3- [M=Ti, Sn, Zr] ceramics sintered at
1150oC as a function of x.
Figure 10 shows the variation of temperature coefficient of resonant frequency (f) of the
Ca[(Li1/3Nb2/3)1-xMx]O3- [M=Ti, Zr, Sn] ceramics sintered at 1150 C as a function of x. The
f value slightly improved from 21 to 15 ppm/C with an increase of Sn substitution (x
=0.1) and then degraded from 15 to 40 ppm/C with an increase of x > 0.1. The f of the
Ca[(Li1/3Nb2/3)1-xZrx]O3- improved from -25.1 to-4.9 ppm/oC with increasing Zr content from
(x = 0 to 0.10) and then degraded from -4.9 to -15.2 ppm/oC with x = 0.10 to 0.30. The f of -
4.9 ppm/oC is realized for x = 0.10 of Zr substitution. This composition has r of 29.8 and Qxf
value of 36300GHz. As expected, the f value increased from 21 to +83 ppm/C with an
increase in Ti substitution from 0 to 50 mole%. A f value of 0 ppm/C was realized for
Ca[(Li1/3Nb2/3)0.8Ti0.2]O3 (x = 0.2). This composition has r and Qxf value of 38.6 and
26100 GHz, respectively. The Qxf value of this composition is relatively lower than the
commonly used materials such as Ba2Ti9O20 (32000 GHz) and (Zr,Sn)TiO4 (49000 GHz) but
this composition has a low sintering temperature (1150 C). The temperature coefficient
resonant frequency of the titanium substituted Ca[(Li1/3Nb2/3)1-xBx]O3- (B=Ti, Zr, Sn) shifts
from negative to positive while the Sn and Zr substituted Ca[(Li1/3Nb2/3)1-xBx]O3- (B= Zr, Sn)
remains as negative. The Ca(Li1/4Nb3/4)O3- ceramics sintered at 1275oC has r =26 Qxf =
13000 GHz and f =-49 ppm/oC. The solid solution of 0.7Ca(Li1/4Nb3/4)O3-- 0.3CaTiO3 [43]
ceramics sintered at 1300 oC showed r = 44, Qxf = 12000 GHz and f = -9 ppm/oC. These
properties are different from non-stoichiometric Ca[(Li1/3Nb2/3)1-xTix]O3- where the zero f is
for x=0.2. Table 2 shows the microwave dielectric properties of Ca[(Li1/3Nb2/3)1-xBx]O3-
(B=Ti, Zr, Sn) complex perovskites. The tantalum analogue of Ca[(Li1/3Nb2/3)1-xTix]O3-
ceramics has received less attention [38]. Liu et al. [38] reported the low temperature
sintering and microwave dielectric properties of Ca[(Li1/3Ta2/3)1-xTix]O3- ceramics. As x
increased to 0.5, r increased linearly from 24 to 48 and f became positive. At x=0.3,
Ca[(Li1/3Nb2/3)0.7Ti0.3]O3- +3 wt% B2O3 sintered at 1050 oC/4hrs showed r =35, Qxf = 22800
GHz and f = -4 ppm/oC.
Sn
75 Ti
Zr
50
f (ppm/ C)
o 25
-25
0.0 0.1 0.2 0.3 0.4 0.5

x
Figure 10. Variation of temperature coefficient of resonant frequency of Ca[(Li1/3Nb2/3)1-xMx]O3-

[M=Ti, Sn, Zr]ceramics sintered at 1150oC as a function of x.
Table 2. Microwave dielectric properties of Ca[(Li1/3Nb2/3)1-xBx]O3- (B=Ti, Zr, Sn)

complex perovskites with tolerance factor
Material x Tolerance r Quxf (GHz) f (ppm/oC

factor
0.0 0.931656 29.6 40000 -21
0.1 0.935010 34.6 27200 -2.3
Ca[(Li1/3Nb2/3)1-xTix]O3- 0.2 0.938388 38.6 26100 0
0.3 0.941792 44.7 22500 20
0.5 0.948672 55.2 18600 83
0.10 0.931191 25.2 48200 -14
0.15 0.930958 24.8 49100 -25
Ca[(Li1/3Nb2/3)1-xSnx]O3-
0.20 0.930725 23.3 50600 -30
0.30 0.930261 22.6 46300 -39
0.05 0.930750 29.9 41800 -20
0.1 0.929850 30.1 36000 -5
Ca[(Li1/3Nb2/3)1-xZrx]O3-
0.2 0.928051 30.5 33000 -12
0.3 0.926258 31.1 27100 -15
4. LOW TEMPERATURE CO-FIRED CERAMICS

Telecommunication industry today requires high volume efficiency without the
degradation of electrical performance of the devices. In order to reduce the size of microwave
devices in communication systems, the dielectric component must also be miniaturized. The
development of multilayered structures has led to considerable volume efficiency [53, 54]. In
order to use a material for incorporation into multilayer type elements, the dielectric material
should be capable of being sintered along with electrode, in addition to having dielectric
properties suitable for desired applications. Low Temperature Co-fired Ceramic (LTCC)
technology is one of the promising technology for advanced packaging and multilayered
structures with buried passive components for fast digital applications [28, 55]. LTCC is well
suited for realizing the compact, multifunctional RF and microwave modules. One of the
most important processes in manufacturing multilayered LTCC component involves the co-
firing of electrode materials such as silver with low loss ceramics. A good combination of
high permittivity, high quality factor, nearly zero temperature coefficients of resonant
frequency and low processing temperature are required for the possible application of
microwave dielectric materials in LTCC based devices. There are many ceramic materials
with excellent microwave dielectric properties but most of them have high sintering
temperature [1]. In general, the sintering temperature of most of the microwave dielectric
ceramics is much higher than the melting point of commonly used electrode materials such as
silver (961oC). Hence in multilayered structure, it is necessary to lower the sintering
temperature of the ceramics below the melting point of electrode materials such as silver.
Generally, chemical synthesis, use of starting powder with smaller particle size and liquid
phase sintering are the three methods employed to lower the sintering temperature of the
ceramics [28, 56-59]. Among these, liquid phase sintering using glass addition is known to be
the most effective and least expensive method of achieving high density sintered ceramics at
low sintering temperatures [26, 28, 60, 61]. In the LTCC systems, addition of low melting
glasses to the ceramics is commonly used to lower the sintering temperature. However, in the
majority of ceramic- glass systems, the reduction of sintering temperature is usually
accompanied by an abrupt degradation of dielectric properties and is due to the loss
mechanism associated with the glass and the formation of second phases. To weaken the
destructive effect of the glass additives on the microwave dielectric properties of the
ceramics, researchers have to find a good glass flux which could decrease the sintering
temperature without affecting the microwave dielectric properties. It is well known that the
sintering aids containing boron oxide and silicon dioxide promotes densification by liquid
phase sintering and have low dielectric loss [62]. Silica is not an effective flux for microwave
dielectrics if it is used alone. To lower the melting point, the rigid bonds in silicon dioxide
may be broken by modifiers such as B2O3, and alkali ions. In general, the B2O3 modifier could
increase the fluidity of glass and the pores were correspondingly coalesced during melting.
Reports show that single component low melting additives easily forms second phase with
parent materials which may have high dielectric loss [1, 26-28, 63]. Similar phenomenon has
been reported in high quality factor materials like Ba2Ti9O20 (BT), Ba(Mg1/3Ta2/3)O3 (BMT),
Ba(Zn1/3Ta2/3)O3 (BZT) and (Zr1-xSnx)TiO4 (ZST) etc [26, 27, 64-69]. Earlier reports show
that multi component glasses are more effective than single component glasses to reduce the
sintering temperature of the ceramics without much degradation of the microwave dielectric
properties [28, 61]. When glass is used in microwave ceramics, the quality factor invariably
degrades due to the loss mechanism associated with the glass. Therefore the development of
low temperature sinterable ceramics with high relative permittivity, low dielectric loss and
nearly zero temperature coefficient of resonant frequency is one of the major challenges in the
electronic industry.
It has been reported that the A site and M site modified solid solution of Ca[(Li1/3A2/3)1-
xMx]O3- [A=Nb, Ta and M=Ti, Zr, Sn] [32, 33, 38] complex perovskite ceramics show
excellent microwave dielectric properties at a relatively low sintering temperatures of about
1150 oC/4hrs. However, the sintering temperature of this material is still too high to use silver
or copper as an internal electrode in multilayered structures. In order to develop lithium based
complex perovskite as a new type of LTCC, it is essential to reduce the sintering temperature
without considerably affecting the microwave dielectric properties. Therefore this section
focused on the sinterability and microwave dielectric properties of Ca[(Li1/3A2/3)1-xMx]O3-
[A=Nb, Ta and M=Ti, Zr, Sn] complex perovskite ceramics with glass addition. Among the
Ca[(Li1/3A2/3)1-xMx]O3- [A=Nb, Ta and M=Ti, Zr, Sn] complex perovskites, the Ti4+
substitution has a wide range of f extending from -21 to +83 ppm/oC. A single phase
perovskite with zero f is observed in the vicinity of x = 0.2. Hence the majority of the work
based on LTCC has been done on the titanium substituted Ca[(Li1/3A2/3)1-xTix]O3- [A=Nb,
Ta] complex perovskite system.
4.1. Effect of Glass Addition in Ca[(Li1/3A2/3)1-xTix]O3- [A=Nb, Ta]
Several authors [34-45] tried to control the escape of volatile lithium by lowering the
sintering temperature and by the addition of low melting additives such as B2O3, Bi2O3, glass
frit or LiF. Effects of B2O3 and/or Bi2O3 on the sinterability and microwave dielectric
properties of nonstoichiometric Ca(Li1/3Nb2/3)O3 (CLN) ceramics were investigated by Liu
et al. [37]. With the addition of B2O3 from 0.5 wt% to 4.0 wt%, the r and Qxf value of CLN
sintered at 990 oC were comparable to those of pure CLN sintered at 1150 oC. For the
specimens with 1.0 wt% B2O3 and 9 wt% Bi2O3, the sintering temperature of the sample was
further reduced to 900 oC, while the Qxf value decreased with the increase of Bi2O3 content.
These results are due to the formation of a low-melting-point second phase between Li2O and
B2O3. The dielectric properties of the samples with 6.0 wt% B2O3 degraded significantly due
to the formation of the Li2B4O7 secondary phase. The sintering temperature of CLN ceramics
was reduced by Bi2O3 at the expense of the Qxf value. Typically, r of 30.6, Qxf of 31000
GHz, and f of 17.5 ppm/C were obtained for the samples with 4.0 wt% B2O3 sintered at
1000 oC. Lui et al. [39, 40] also studied the effect of B2O3 addition on the microwave
dielectric properties of Ca[(Li1/3Nb2/3)1-xTix]O3-. Addition of 2 wt % B2O3 and 6 wt % Bi2O3
to Ca[(Li1/3Nb2/3)0.7Ti0.3]O3- ceramics sintered at 920 oC has r = 43.1, Qxf = 10600 GHz, and
f = 10.7 ppm/oC.
Ha et al. [34-36] reported the microwave dielectric properties of Ca[(Li1/3Nb2/3)1-xTix]O3-
ceramics with Bi2O3 and B2O3-ZnO-SiO2-PbO glass. Secondary phase has been observed
when more than 5 wt% of B2O3 was added. The Ca[(Li1/3Nb2/3)0.8Ti0.2] O3- ceramics with 5
wt.% Bi2O3 sintered at 900 oC for 3 hours has r = 35, Qxf = 11000 GHz, and f = 13 ppm/oC.
Addition of glass frit (B2O3ZnOSiO2PbO) improved the densification and decreased the
sintering temperature of Ca[(Li1/3Nb2/3)1-xTix]O3- dielectric ceramics from 1150 oC to 900oC.
The dielectric properties of Ca[(Li1/3Nb2/3)0.8Ti0.2] O3- with 12 wt% glass sintered at 900 oC
for 3 h were r = 40, Qxf = 12500 GHz, and f = - 8 ppm/oC respectively.
Tong et al. [41, 42] reported the low temprature firing and microwave dielectric
properties of Ca[(Li1/3Nb2/3)0.9Ti0.1]O3- ceramics with LiF addition and the
Ca[(Li1/3Nb2/3)0.84Ti0.16] O3- with the combination of LiF and ZBS glass. The
Ca[(Li1/3Nb2/3)0.84Ti0.16] O3- ceramics with 2 wt% LiF and 3 wt% ZBS sintered at 900oC for 2
h show excellent dielectric properties: r = 34.3, Qxf = 17400 GHz, and f = - 4.6 ppm/oC.
Table 6 shows the microwave dielectric properties of low temperature sintered CLNT/CLTT
ceramics.
4.1.2. Effect of Borosilicate Glasses in Ca[(Li1/3A2/3)1-xTix]O3- [A=Nb, Ta] Ceramics

The low temperature sintering and microwave dielectric properties of Ca[(Li1/3A2/3)1-
xTix]O3- [A=Nb, Ta] complex perovskite with different borosilicate glasses such as lithium
borosilicate (LBS), lead borosilicate (PBS), zinc borosilicate (ZBS), barium borosilicate
(BBS), lithium magnesium zinc borosilicate (LMZBS) and bismuth zinc borosilicate (BZBS)
[70-73] is discussed in detail in this section. Table 3 shows the physical and dielectric
properties of various borosilicate glasses [28, 30, 44, 60, 70-75]. The Ca[(Li1/3Nb2/3)0.8Ti0.2]
O3- and Ca[(Li1/3Ta2/3)0.7Ti0.3] O3- ceramic compositions have been taken to investigate the
effect of glass addition on the sinterability and microwave dielectric properties for its phase
purity and excellent temperature stable microwave dielectric properties.
Table 3. Physical and dielectric properties of various borosilicate glasses
Glass Composition (mole Density Softening r Tan Reference

code %) (g/cc) Temperature
(oC)
PBS 40:PbO,40:B2O3, 4.31 448 12.9 7.1x10-3 ( 28
20:SiO2 12 GHz)
BBS 30:BaO,60:B2O3, 3.40 627 7.20 4.4x10-3 28

10:SiO2 (15 GHz)
ZBS 60:ZnO,30:B2O3, 3.60 582 7.50 1.07x10-2 28

10:SiO2 (15 GHz)
LBS 35.5:Li2O, 2.34 513 6.44 3x10-3 (1 28

31.66:B2O3, MHz)
33.2:SiO2
BZBS 35:Bi2O3, 32:ZnO, 4.30 950 8.80 1.3x10-3 28
27:B2O3, 6:SiO2 (1 MHz)
LMZBS 20:Li2O, 20:MgO, 2.75 900 6.90 1x10-3 28
20: ZnO, 20:B2O3, (1 MHz)
20:SiO2
CLNT CLTT
0 0
BBS
Shrinkage (%) -3 (a) BBS -3 (b) LBS
Shrinkage (%)
-6 -6
PBS PBS
-9
-9 BZBS LMZBS
ZBS -12
-12 LMZBS
ZBS
BZBS
-15
-15 LBS
200 400 600 800 1000 200 400 600 800 1000 1200
o o
Temperature ( C) Temperature ( C)
Figure 11. Shrinkage characteristics of (a) CLNT and (b) CLTT ceramics with various borosilicate
glasses.
4.1.3. Shrinkage Characteristics

To identify whether the glass additive would be effective on the low temperature firing of
CLNT / CLTT ceramics, the linear shrinkage of as pressed pellets as a function of
temperature is to be initially measured. The dilatometry study provides insight into the
densification process for the glass ceramic systems. Figure 11 (a) and (b) shows the shrinkage
characteristics of CLNT / CLTT ceramics and with 12 wt% of various borosilicate glasses. It
is found that there is no linear shrinkage for pure CLNT and CLTT in the temperature range
of 800-1100oC. This indicates that sintering does not start even at 1100 oC for both the
ceramics. However, a linear shrinkage is observed as shown in Figure 11 for all the glass
added ceramics which implies that all the borosilicate glass acts as a good sintering aid.
Among the various glasses, BZBS, ZBS and LMZBS glasses in CLTT and LBS, LMZBS,
and ZBS glasses in CLNT show a linear shrinkage of more than 12% in the temperature range
800-1000 oC.
4.1.4. Densification
Figure 12 shows the variation of sintering temperature, and the relative density of CLNT
/ CLTT ceramics mixed with different wt% of various borosilicate glasses. It is found that all
the glasses decreases the sintering temperature. However, among the various six borosilicate
glasses, LBS, LMZBS and ZBS glass could reduce the sintering temperature of CLNT
ceramics. The ZBS and BZBS glasses reduced the sintering temperature of CLTT ceramics
below the melting point of silver. The theoretical density (D) of the glass added CLNT was
calculated using the equation
W1 + W2
D=
W1 + W2
D1 D2
97 100
96 (a) (b)
98
95

Relative density (%)
94 96
93
BZBS
92 BZBS 94
ZBS
91 ZBS PBS
PBS BBS 92
90 BBS LBS
89 LBS LMZBS
90
LMZBS
88
BZBS
1200 BZBS
(a) ZBS (b) ZBS
PBS 1150
Sintering Temperature ( C)
PBS
Sintering Temperature ( C)
1150 BBS
o
BBS
o
LBS 1100
LBS
1100 LMZBS
LMZBS
1050
1050
1000
1000 950
950 900
0 2 4 6 8 10 12 0 2 4 6 8 10 12
wt % of Glasses wt% of Glasses
Figure 12. Variation of sintering temperature and relative density of (a) CLNT and (b) CLTT ceramics
with different wt% of various borosilicate glasses.
where W1 and W2 are the weight percentage of ceramics and glass with densities D1 and D2
respectively.
The relative density of the glass-ceramic composite was calculated using the equation:
Dth
Densificat ion =
Dexp t
Figure 12 also shows the variation of relative density as a function of glass content. The
relative density decreases with increase of glass content in CLTT ceramics. For CLNT-glass
composite, the relative density increases slightly with an increase in the glass. However,
higher wt% of glass addition decreases the densification of CLNT-glass composite. The
relative density and the densification temperature is different for different glasses in ceramics.
The effectiveness of sintering aid depends on several factors such as sintering temperature,
viscosity, solubility, and glass wettability. The densification of glass ceramic systems is due
the formation of liquid phase. The glassy liquid phase at the grain boundary effectively
eliminates pores and this could be the reason for the increase in the relative density of CLNT-
glass system. Trapped porosity associated with grain growth and formation of pores by the
evaporation of excess glass components are the reasons for the reduction in the density for
higher wt% of glass fluxing [27, 28, 60]. It may be noted that the glasses have lower density
as compared to CLNT / CLTT ceramics. The required densification for practical application
could be obtained by the addition of glasses. However, undesirable decomposition of host

dielectric material and the formation of second phase may occur which may degrade the
microwave dielectric properties. Hence the amount of glass frit has to be optimized to get the
best dielectric properties.
4.1.5. Phase Analysis

Figs. 13 and 14 show the X-ray powder diffraction patterns of sintered CLNT / CLTT
ceramics and the different borosilicate glass added CLNT / CLTT ceramics. The XRD
patterns are indexed based on the CaTiO3 type orthorhombic perovskite structure. Higher
wt% of glasses resulted in the formation of a small amount of second phase which was
prominent for LBS glass added CLTT ceramics. For co-firing applications, the glass ceramic
composite should not react with silver. Figure 13 (h), (i) and (j) show the X-ray diffraction
patterns of sintered CLNT ceramics with 20 wt% of silver and 5 wt% of LBS, 12 wt% of
LMZBS, and 8 wt% of ZBS glasses respectively. Similarly, Figs. 14(h) and (i) show the X-
ray diffraction patterns of 20 wt% of silver added CLTT ceramics with 12 wt% of ZBS and
BZBS glasses sintered at the optimized sintering temperature. It is clear from the X-ray
diffraction patterns that there is no additional phases other than the orthorhombic CLNT /
CLTT ceramics and metallic silver (ICDD File No: 03-0921). This indicate that there is no
chemical reactivity between the glass added CLNT/CLTT and the electrode material silver.
# - Silver
# * - Second phase
# # (j)
# # # (i)
Intensity (arb. unit)
#
# # (h)
* (g)
(f)
(e)
(d)
(c)
*
(b)
042 240
400 242
004
204323
232
002121
040
202
222
321
430
101
020
221
013
200
250
161
(a)
031
30 40 50 60 70 80
2 Degree
Figure 13. The X-ray powder diffraction patterns of sintered (a) CLNT ceramics, CLNT ceramics with
(b) 12 wt% of BBS, (c) 12 wt% of PBS, (d) 5 wt% of LBS, (e) 12 wt% of LMZBS, (f) 8 wt% of ZBS,
(g) 12 wt% of BZBS glasses, (h) CLNT ceramics with 5 wt% of LBS glass and 20 wt% of silver and (i)
CLNT ceramics with 8 wt% of ZBS glass and 20 wt% of silver and (j) CLNT ceramics with 12 wt% of
LMZBS glass and 20 wt% of silver using Ni-filtered Cu-K radiation.
4.1.6. Microstructure Analysis

Figures 15 (a) and 16 (a) shows the scanning electron micrographs of thermally etched
Ca[(Li1/3Nb2/3)0.8Ti0.2]O3- ceramics and Ca[(Li1/3Ta2/3)0.7Ti0.3]O3- ceramics. The CLNT and
CLTT ceramics sintered at 1175oC and 1200oC give an average grain size of 2-4 m. It is
evident from the micrographs that a small amount of glass addition helps the sintering of
CLNT ceramics by liquid phase. The densification improved initially up to 1 wt% of LBS
glass addition and further addition of glass degraded the densification which can be clearly
seen from Figure 15 (c). Figure 15 (e), and (f) shows the SEM picture of 8 and 12 wt% of
ZBS and BZBS glass added ceramics. It is also evident from both the micrograph (see Fig 15
and 16) that the glass addition decreases the grain size which is in agreement with previous
reports [37]. At higher sintering temperature the glasses melts and may evaporate and
generate porosity which can be seen in Figure 16 (b) and (c). Figure 15 (d) and 16 (d) shows
the scanning electron micrographs of CLNT+ 5 wt% of LBS+ 20 wt% of Ag and CLTT+12
wt% of BZBS+ 20 wt% of Ag respectively. It is observed that the silver does not react with
glass-ceramic composite and the absence of such reaction is required for the material for
LTCC application.
#- Silver
# * - Second Phase
# # (i)
# # # (h)
Intensity (arb.units)
* (g)
(f)
(e)
*
* * * * (d)
* * *
(c)
* * (b)
250 004
042
430 161
400 242
321 240
204 323
121
002
040
222
020
200
202
013
101
232
221
(a)
30 40 50 60 70 80
2 Degree
Figure 14. The X-ray powder diffraction patterns of sintered (a) CLTT ceramics, CLTT ceramics with
12 wt% of (b) BBS, (c) PBS, (d) LBS, (e) LMZBS, (f) ZBS, (g) BZBS glasses, (h) CLTT ceramics with
12 wt% of ZBS glass and 20 wt% of silver and (i) CLTT ceramics with 12 wt% of BZBS glass and 20
wt% of silver using Ni-filtered Cu-K radiation.
Figure 15. The microstructure of sintered Ca[(Li1/3Nb2/3)0.8Ti0.2]O3- ceramics with (a) 0wt%, (b) 1 wt%
LBS (c) 5 wt% LBS, (d) 5 wt% of LBS + 20 wt% Ag, (e) 8 wt % ZBS and 12 wt% LMZBS glass.
4.1.7. Microwave Dielectric Properties

Addition of glasses not only promotes the sinterability but also affect the microwave
dielectric properties. It is well known that relative density and dielectric properties are inter-
related. Figure 17 shows the microwave dielectric properties CLNT / CLTT ceramic as a
function of different wt% of various glasses. The relative permittivity of glass added CLNT /
CLTT ceramics decreases with increase in the glass content as shown in Figure 17. The
relative permittivity of all the borosilicate glasses is much lower than that of CLNT / CLTT
ceramics as given from table 3. Moreover, higher wt% of glass addition increased the porosity
and decreased the densification of ceramics. The low permittivity of glass and the poor
densification are the reasons for the decrease in the relative permittivity of CLNT / CLTT
ceramics with higher wt% of glass addition.
Figure 16. The microstructure of sintered Ca[(Li1/3Ta2/3)0.7Ti0.3]O3- ceramics with (a) 0wt %, (b) 12
wt% BZBS (c) 12 wt% ZBS, and (d) 12 wt% of BZBS + 20 wt% Ag.
The microwave dielectric properties of CLNT/CLTT ceramics with different amount of

glass are shown in table 6. Figure 17 shows the optimized quality factor for the addition of 0-
12 wt% of different borosilicate glasses in CLNT ceramics. The sintering temperature versus
wt % of glass in CLNT/CLTT composite shows (see figure 12) that, all the glasses could not
act as good sintering aid to lower the sintering temperature below 950 oC. Among the six
borosilicate glasses, only the LBS, LMZBS, and ZBS glass could reduce the sintering
temperature of CLNT ceramics with reasonable dielectric properties. It is observed that the
addition of small amount of LBS, ZBS and LMZBS glasses increased the quality factor
slightly. Maximum Quxf value of 22900 GHz, 24500 GHz, and 23500 GHz (f = 4.4 GHz) was
found for the addition of 1 wt% of LBS, 3 wt% of LMZBS, and 1 wt% of ZBS glasses
respectively. These could be due to the increased densification of CLNT ceramics by liquid
phase sintering with small amount of glass fluxing. These compositions have relative
permittivity of 34.2, 34 and 38.3 respectively with a f of around -10 ppm/ oC. Higher wt% of
glass addition in CLNT ceramics lowered the sintering temperature with further lowering of
the quality factor. The CLNT ceramics with 5 wt% of LBS glass sintered at 950 oC shows r =
30.5, Qu x f = 14700 GHz (f = 4.6 GHz) and f = -18 ppm/ oC. A similar trend is also observed
with the addition of ZBS and LMZBS glasses in CLNT ceramics. Addition of 12 wt% of ZBS
glass in CLNT ceramics gives a relative permittivity of 33.5 with Qxf of about 13200 GHz
and f of -28 ppm/oC. Hence for CLNT ceramics, amoung the various six borosilicate glasses
the LBS, LMZBS, and ZBS glass acts as good sintering aid to lower the sintering temperature
with reasonably good microwave dielectric properties.
38
40 BZBS BZBS
ZBS ZBS
38 PBS PBS 36
Relative Permittivity (r) BBS BBS

36 LBS LBS
34
34 LMZBS LMZBS
32 32
30
30
28
26 (a) (b) 28
BZBS BZBS 30
24 ZBS
ZBS
PBS PBS
21 25
BBS BBS
Quxf (x10 GHz)

LBS LBS
Quxf (x 103GHz)
18 LMZBS LMZBS 20
3
15 15
12 10
9 5
(a) (b)
0
0 2 4 6 8 10 12 0 2 4 6 8 10 12
wt % of Glasses wt % of glasses
Figure 17. Variation of relative permittivity and quality factor of (a) CLNT and (b) CLTT ceramics
with different wt% of various borosilicate glasses.
It can be observed from figure 17 that the quality factor of CLTT ceramics decreased
with increase in borosilicate glass additions. For CLTT ceramics, the ZBS and BZBS glasses
acts as good sintering aids and the glasses such as BBS, PBS, LBS, and LMZBS could not
lower the sintering temperature below 950 oC with reasonably good dielectric properties. The
CLTT ceramics with 12 wt% of ZBS glass sintered at 950 oC showed r = 31.8, Qu x f = 9000
GHz (f = 4.6 GHz). Addition of 12 wt% of BZBS glass in CLTT ceramics sintered at 950 oC
showed r = 32.0, Qu x f = 9000 GHz (f = 4.7 GHz). It may be noted that the dielectric loss of
glasses is much higher than that of pure CLTT ceramics. The added glasses not only acts as
the sintering aid but also as an impurity that can adversely affect the microwave dielectric
properties. It is reported that the microwave dielectric loss was mainly caused not only by the
lattice vibrations but also by the pores, the grain morphology and the second phase [76].
These could be the reason for the decrease in the quality factor with glass additions.
Figure 18 shows the variation of temperature coefficient of resonant frequency of CLNT /
CLTT ceramics with different wt% of various borosilicate glasses. Addition of various
borosilicate glasses decreased the temperature coefficient of resonant frequency (f) of CLNT
ceramics +4 ppm/o C to -22 ppm/o C and that of CLTT ceramics from -3 ppm/o C to -41 ppm/o
C. This is due to the high negative f of all the borosilicate glasses compared to CLTT
ceramics. The CLNT / CLTT ceramics sintered below 950 oC with selected borosilicate
glasses show excellent dielectric properties. However, the temperature coefficient of resonant
frequency is slightly higher (> -20 ppm/o C). Hence, it is necessary to lower the f close to
zero for applications in LTCC based devices. It is well known that rutile has high positive f
and are widely used for tuning the temperature coefficient of resonant frequency of dielectric
resonators [49-52]. In order to lower the f value, the Ca[(Li1/3Nb2/3)0.8Ti0.2]O3- ceramics were
modified to Ca[(Li1/3Nb2/3)0.75 Ti0.25]O3- and the Ca[(Li1/3Ta2/3)0.7Ti0.3]O3- ceramics were
modified to Ca[(Li1/3Ta2/3)0.6 Ti0.4]O3- by increasing the rutile content.
5 BZBS
ZBS
0 PBS
BBS
-5 LBS
f (ppm/ C)
LMZBS
o
-10
-15
-20
-25
-30
BZBS
0
ZBS
PBS
-10 BBS
LBS
f (ppm/ C)
LMZBS
o
-20
-30
-40
0 2 4 6 8 10 12
wt% of glasses
Figure 18. Variation of temperature coefficient of resonant frequency of (a)CLNT and (b) CLTT
ceramics with different wt% of various borosilicate glasses.
Table 4. Densification and microwave dielectric properties of LBS and LMZBS glass
added Ca[(Li1/3Nb2/3)0.75Ti0.25]O3- ceramics and the BZBS glass added Ca[(Li1/3Ta2/3)0.6
Ti0.4]O3- ceramics (after Ref 44, 82)
Material Sintering r Quxf f ppm/o

Temperature (oC) (GHz) C
Ca[(Li1/3Nb2/3)0.75 Ti0.25]O3- 1175 41.9 17500 +23
Ca[(Li1/3Nb2/3)0.75 Ti0.25]O3- + 5 wt% 950 33.0 11500 -2
LBS
Ca[(Li1/3Nb2/3)0.75 Ti0.25]O3- + 8 wt% 940 30.5 9750 -5
LBS
Ca[(Li1/3Nb2/3)0.75 Ti0.25]O3- + 12 wt% 920 28.5 11000 -3
LMZBS
Ca[(Li1/3Nb2/3)0.75 Ti0.25]O3- + 15 wt% 900 27.4 11100 -8
LMZBS
Ca[(Li1/3Ta2/3)0.6Ti0.4]O3- 1200 38.7 15000 +15
Ca[(Li1/3Ta2/3)0.6Ti0.4]O3- + 12 wt% of 950 34.9 6500 -6
BZBS glass
Table 5. Sintering temperature and microwave dielectric properties of Ca[(Li1/3A2/3)1-

xMx]O3- [A=Nb, Ta and M=Ti, Zr, Sn] complex perovskites
SI Materials Sintering r Qxf (GHz) f (ppm/oC) Ref

No Temperature (oC)
1 Ca[(Li1/3 Nb2/3)1-x 1150/5h 29-55 40000- -21 to+83 32
Tix]O3- x=0- 0.5 18000
x = 0.00 1150/5h 29.6 40000 -21 32
0.05 1150/5h 32.4 29800 -4.5
0.10 1150/5h 34.6 27200 -2.3
0.20 1150/5h 38.6 26100 0
0.30 1150/5h 44.7 22500 +20
0.50 1150/5h 55.2 18600 +83
2 Ca[(Li1/3 Nb2/3)1- 1150/3h 29-23 40000- -21 to -40 32
xSnx]O3- (x=0- 0.3) 46000
x = 0.00 1150/3h 29.6 40000 -21 32
0.10 1150/3h 25.2 48200 -14
0.15 1150/3h 24.8 49100 -25
0.20 1150/3h 23.3 50600 -30
0.30 1150/3h 22.6 46300 -39
3 Ca[(Li1/3 Nb2/3)1- 1150/3h 29.5- 31 40000- -25 to -15 33

xZrx]O3- (x=0- 0.5) 27000
x = 0.00 1150/3h 29.5 40000 -25 33
0.10 1150/3h 29.8 36000 -4.9
0.20 1150/3h 30.5 32000 -13.2
0.30 1150/3h 31.0 27100 -15.2
4 1-x 1200-1300/2h 36-48 16000- -33to +40 43

Ca[(Li1/4Nb3/4)O3- 11000
xCaTiO3 ,x=0- 0.5
x = 0.00 1275/2h 26.0 13000 -49 43
0.20 1275/2h 36.0 15400 -25
0.25 1275/2h 39.0 12500 -21
0.30 1275/2h 43.0 9000 -18
0.35 1275/2h 45.0 8500 +10
0.40 1275/2h 48.0 8500 +40
5 Ca[(Li1/3 Ta2/3)1-x 1200/4 24-42 42000- -40 to +50 38
Tix]O3- x=0- 0.5 11000
x=0.0 1200/4 24.0 42000 -40 38
0.1 1200/4 26.3 - -
0.2 1200/4 28.4 - -
0.3 1200/4 35.0 22800 -4.1
0.5 1200/4 42.0 11000 +50
Table 6. Low temperature sintered Ca[(Li1/3A2/3)1-xTix]O3- [A=Nb, Ta] complex

perovskites with sintering aid
SI Materials wt % of Sintering r Qxf f Ref

No Glass (y) Temperature (GHz) (ppm/oC)
(oC)
1 Ca[(Li1/3 Nb2/3)O3- 0 980/4 24.0 10000 -16 37
+1 wt% B2O3+ y Bi2O3 2 980/4 28.0 17000 -20
4 980/4 29.0 12000 -22
6 980/4 30.0 11000 -23
8 980/4 30.2 10000 -24
2 Ca[(Li1/3 Nb2/3)1-x 10 900/3 23.0 17000 -75 35
Tix]O3- + y 12 900/3 27.0 16000 -60
(B2O3+ZnO+SiO2+PbO 15 900/3 28.0 13000 -50
x=0
3 x=0.1 10 900/3 25.0 16000 -58 35
12 900/3 30.5 15000 -52
15 900/3 34.0 11000 -42
4 x=0.2 10 900/3 32.5 14000 -22 35
12 900/3 35.5 11000 -20
15 900/3 38.0 7000 -18
5 x=0.3 10 900/3 41.0 10000 +20 35
12 900/3 43.0 7500 +30
15 900/3 46.0 7000 +50
6 Ca[(Li1/3 Nb2/3)1-x 5 900/3 20.0 3500 -10 36
Tix]O3- + y (Bi2O3) 7 900/3 22.5 5000 -
x=0.05 10 900/3 24.5 6000 -
Table 6. Continued
7 x=0.1 5 900/3 23.0 4500 +5 36

7 900/3 24.0 5000
10 900/3 27.0 6500
8 x=0.2 5 900/3 35.0 7000 +12 36
7 900/3 37.5 8000
10 900/3 40.0 10500
9 x=0.3 5 900/3 37.0 6500 +18 36
7 900/3 40.0 7500
10 900/3 42.5 8000
10 Ca[(Li1/3 Nb2/3)1-x 2 wt% 1050/3 30.0 30000 -20 39
Tix]O3- x=0.0 B2O3
11 x=0.1 2 wt% 1000/3 34.0 25000 -10 39
B2O3
12 x=0.2 2 wt% 1000/3 37.0 20000 +5 39
B2O3
13 x=0.3 2 wt% 1000/3 42.0 18000 +15 39
B2O3
14 x=0.5 2 wt% 1000/3 52.0 15000 +80 39
B2O3
15 Ca[(Li1/3 Nb2/3)0.7 1 wt% 960 43.9 16600 +35.5 39
Ti0.3]O3- + 2 wt% B2O3 Bi2O3
16 Ca[(Li1/3 Nb2/3)0.7 3 wt% 940 43.1 12900 +53.7 39
Ti0.3]O3- + 2 wt% B2O3 Bi2O3
17 Ca[(Li1/3 Nb2/3)0.7 6 wt% 920 43.1 10600 +10.7 39
Ti0.3]O3- + 2 wt% B2O3 Bi2O3
18 Ca[(Li1/3 Nb2/3)0.9 0 1150 35.0 16000 41
Ti0.1]O3-d + y LiF
19 Ca[(Li1/3 Nb2/3)0.9 2 975 31.0 15000 +17 41
Ti0.1]O3- + y LiF
20 Ca[(Li1/3 Nb2/3)0.9 6 940 29.0 17000 +2 41
Ti0.1]O3- + y LiF
21 Ca[(Li1/3 Nb2/3)0.9 10 900 25.0 19000 -15 41
Ti0.1]O3-d + y LiF
22 Ca[(Li1/3 Nb2/3)0.9 20 840 19.0 20000 -25 41
Ti0.1]O3- + y LiF
23 Ca[(Li1/3 Nb2/3)0.84 0 42
Ti0.16]O3- 1180/3 37.8 14270 -6.2
24 Ca[(Li1/3 Nb2/3)0.84 1 940/3 34.5 16000 +28 42
Ti0.16]O3- + 2 wt% 2 940/3 34.0 15000 +7
ZBS+ y wt%LiF 3 940/3 32.5 14500 -5
5 940/3 30.5 13000 -18
25 Ca[(Li1/3 Nb2/3)0.84 1 940/3 34.0 18000 +20 42
Ti0.16]O3- + 2 wt% LiF 2 940/3 33.5 16000 +7
+ y wt% ZBS 3 940/3 33.0 15000 +5
5 940/3 32.4 14500 -8
26 Ca[(Li1/3 Nb2/3)0.8 5 1025 33.4 18500 -12 42
Ti0.2]O3- + y wt% ZBS 8 950 32.6 18000 -21

12 925 33.5 13200 -28
27 Ca[(Li1/3 Nb2/3)0.8 3 1025 34.2 24500 -5 44
Ti0.2]O3- + y wt% 5 975 30.7 22600 -18
LMZBS 8 975 28.4 13590 -20
12 900 26.2 13000 -20
15 900 24.1 11500 -21
28 Ca[(Li1/3 Nb2/3)0.8 3 1025 32.0 20000 -15 44
Ti0.2]O3- + y wt% LBS 5 950 30.5 14700 -18
8 940 27.8 11100 -22
12 920 26.5 7000 -18
29 Ca[(Li1/3 Ta2/3)]O3+ y 0 1200/2 24.0 42000 -40 38
wt% B2O3 1 1100/2 24.1 38900 +7
3 1050/2 23.4 40300 +5
6 1050/2 23.4 27900 -8
30 Ca[(Li1/3 Ta2/3)1-x Tix]O3- 3 wt% 1100/4 26.1 22000 -97 38
B2O3
x=0.05
31 x=0.15 3 wt% 1050/4 29.5 20700 -57 38
B2O3
32 x=0.20 3 wt% 1050/4 28.4 12900 -14.7 38
B2O3
33 x=0.30 3 wt% 1050/4 35.0 22800 -4.1 38
B2O3
34 x=0.50 3 wt% 1050/4 45.0 12300 75 38
B2O3
35 Ca[(Li1/3 Ta2/3)0.7 5 1050 34.0 15250 -30 82
Ti0.3]O3- + y wt% ZBS 8 1000 32.2 20800 -32
12 950 32.3 11600 -38
36 Ca[(Li1/3 Ta2/3)0.7 5 1025 32.0 11300 -36 82
Ti0.3]O3- + y wt% 8 975 32.0 9000 -42
BZBS 12 950 31.8 7350 -41
Table 4 shows the densification and microwave dielectric properties of LBS and LMZBS
glass added Ca[(Li1/3Nb2/3)0.75Ti0.25]O3- ceramics and the BZBS glass added Ca[(Li1/3Ta2/3)0.6
Ti0.4]O3- ceramics. The Ti substitution increased the dielectric constant of the CLNT / CLTT
ceramics. This can be due to the incorporation of small Ti4+ ion (0.605 ) to the B site [Li1/3
Nb/Ta2/3]3.67+ with slightly larger ionic radius (0.66 ) [52]. The Ca[(Li1/3Ta2/3)0.6 Ti0.4]O3-
ceramics with 12 wt% of BZBS glass sintered at 950 oC has r = 34.9, Qxf =6500 GHz (f =
4.6 GHz) and f -6 ppm/oC. The Ca[(Li1/3Nb2/3)0.75 Ti0.25]O3- ceramics with 12 wt%
LMZBS glass sintered at 900o C has r =28.5, Qu x f =11000 GHz (f = 4.7 GHz) and f = -3
ppm/oC and that of LBS glass addition of 5 wt% sintered at 950oC has r = 33, Qxf =11500
GHz (f = 4.6 GHz) and f -2 ppm/oC. The CLNT ceramics shows good dielectric properties
at low sintering temperature compared to CLTT ceramics. Hence the Ca[(Li1/3Nb2/3)0.75
Ti0.25]O3- ceramics with suitable amount of glass addition can find application in LTCC
based devices.
Table 5 shows the sintering temperature and microwave dielectric properties of

Ca[(Li1/3A2/3)1-xMx]O3- [A=Nb, Ta and M=Ti, Zr, Sn] complex perovskites and table 6 shows
the low temperature sintered Ca[(Li1/3A2/3)1-xTix]O3- [A=Nb, Ta] complex perovskites with
sintering aid.
5. CONCLUSIONS
The present review reports the microwave dielectric properties of Ca[(Li1/3A2/3)1-xMx]O3-

[A=Nb, Ta and M=Ti, Zr, Sn] ceramics and their composites with glass for the possible
applications in LTCC based devices. Ca[(Li1/3A2/3)1-xMx]O3- [A=Nb, Ta and M=Ti, Zr, Sn]
complex perovskites show excellent temperature stable microwave dielectric properties when
sintered at about 1150oC. The Ca[(Li1/3A2/3)1-xTix]O3- [A=Nb, Ta] solid solution has an
advantage of tuning the temperature coefficient of resonant frequency close to zero by
adjusting the titanium content. Compared to CLTT-glass composites, the CLNT-glass
composites show excellent dielectric properties with low sintering temperature suitable for
LTCC applications.
ACKNOWLEDGMENT
The authors are grateful to the CSIR, New Delhi, India for the financial assistance.
REFERENCES
[1] Sebastian M. T. Dielectric materials for wireless communications, Elseiver
Publishers, Oxford U. K (2008).
[2] Freer R. Microwave Dielectric Ceramics- An Overview, Silicate Industriels, (1993),
9-10, 191-197.
[3] Wersing W. Microwave ceramics for resonators and filters, Ceramics, composites
and intergrowths, (1996), 715-731.
[4] Kajfez D. Guillon P. Dielectric Resonators, Noble Publishing Corporation, Atlanta,
US Atlanta (1998).
[5] Fiedziuszko S. J., Hunter I. C., Itoh T., Kobayashi Y., Nishikawa T., Stitzer S. N. and
Wakino K. Dielectric Materials, Devices, and Circuits, IEEE Trans. Microwave
Theory and Tech., (2002), 50, 706-720.
[6] Richtmyer R. D. Dielectric Resonators, J. App. Phys., (1939), 15, 391-398.
[7] Schlicke H. M. Quasidegenerate modes in high-e dielectric cavities, J. Appl. Phys.,
(1953), 24, 187-191.
[8] Okaya A. The Rutile microwave resonator, Paper presented at the Proc. IRE,
vol.(1960)
[9] Okaya A. and Barash L. F. The dielectric microwave resonator, Paper presented at
the Proc. IRE. (1962)
[10] Cohn S. B. Microwave bandpass filters containing high Q dielectric resonators, IEEE
Trans. Microwave Theory Tech., (1968), MTT-16, 218-227.
[11] Masse D. J. A new low loss high-k temperature compensated dielectric for
microwave applications, Paper presented at the Proc. IEEE.(1971)
[12] Plourde J. K., Linn D. F., OBryan H. M. J. and Thompson J. J. Ba2Ti9O20 as a
microwave dielectric resonator, J. Am. Ceram. Soc., (1975), 58, 418-420.
[13] Wolfram G. and Goebel H. E. Existence range, structural and dielectric properties of
ZrxTiySnzO4 ceramics (x+y+z = 4). Mat. Res. Bull., (1981), 16, 1455-1463.
[14] Michiura N., Tatekawa T., Higuchi Y. and Tamura H. Role of donor and acceptor
ions in the dielectric loss tangent of (Zr0.8Sn0.2)TiO4 dielectric resonator material., J.
Am. Ceram. Soc., (1995), 78, 793-796.
[15] Huang C.-L. and Weng M.-H. Improved high Q value of MgTiO3-CaTiO3
microwave dielectric ceramics at low sintering temperature, Mater. Res. Bull.,
(2001), 36, 2741-50.
[16] Nishigaki S., Yano S., Kato H., Hirai T. and Nonomura T. BaO-TiO2-WO3
microwave ceramics and crystalline BaWO4., J. Am. Ceram. Soc., (1988), 71, C-11-
17.
[17] OBryan H. M. J., Thomson J. J. and Plourde J. K. A new BaO-TiO2 compound with
temperature-stable high permittivity and low microwave loss, J. Am. Ceram. Soc.,
(1974), 57, 450-453.
[18] Roy R. Multiple ion substitution in perovskite lattice, J. Am. Ceram. Soc., (1954), 27,
581-588.
[19] Nishigaki S., Kato H., Yano S. and Kamimura R. Microwave dielectric properties of
(Ba,Sr)O-SmO -TiO ceramics, Am. Ceram. Soc. Bull., (1987), 66, 1405-1410.
[20] Huang C.-L. and Weng M.-H. Dielectric properties of B2O3 doped (1-x)LaAlO3-
xSrTiO3 ceramic system at microwave frequency range, Mater. Res. Bull., (2002),
37, 1941-1948.
[21] Jancar B., Valant M. and Suvorov D. Solid state reactions occurring during the
synthesis of CaTiO3-NdAlO3 perovskite solid solutions, Chem. Mater., (2004), 16,
1075-1082.
[22] Moon J. H., Jung H. M., Park H. S., Shin J. Y. and Kim H. S. Sintering behaviour
and microwave dielectric properties of (Ca,La)(Ti,Al)O3 ceramics, Jpn. J. Appl.
Phys., (1999), 38, 6821-6827.
[23] Sun P., Nakamura T., Shan Y. J., Inaguma Y., Itoh M. and Kitamura T. Dielectric
behaviour of (1-x)LaAlO3-xSrTiO3 solid solution system at microwave frequencies,
Jpn. J. Appl. Phys, (1998), 37, 5625-5629.
[24] Inchinose I., Mizutani H. and Ookuma H. Ceramics International, (1977), 3, 100.
[25] Imanaka Y. Multylayered Low Temperature Cofired Ceramics (LTCC) Technology,
Spinger, Japan (2005).
[26] Surendran K. P., Mohanan P. and Sebastian M. T. The effect of glass additives on the
microwave dielectric properties of Ba(Mg1/3Ta2/3)O3 ceramics, J. Sol. State Chem,
(2004), 177, 4031-4046.
[27] Bijumon P. V. and Sebastian M. T. Influence of glass additives on the microwave
dielectric properties of Ca5Nb2TiO12 ceramics, Mater. Sci. Engg. B, (2005), 123, 31-
40.
[28] Sebastian M. T. and Jantunen H. Low loss dielectric materials for LTCC
applications: a review, International Materials Review, (2008), 53, 57-90.
[29] Surendran K. P. Investigations on Low Loss Dielectric Materials for Wireless
Communication Faculty of Science,(2004), (Kerala, University of Kerala).
[30] Thomas S. and Sebastian M. T. Effect of B2O3-Bi2O3-SiO2-ZnO glass on the
sintering and microwave dielectric properties of 0.83ZnAl2O4-0.17TiO2, Mater. Res.
Bull., (2008), 43, 843-51.
[31] Anjana P. S. and Sebastian M. T. Microwave Dielectric Properties and Low-
Temperature Sintering of Cerium Oxide for LTCC Applications, J. Am. Ceram. Soc.,
(2009), 92, 96-104.
[32] Choi J.-W., Kang C.-Y., Yoon S.-J., Kim H.-j. and Jung H.-J. Microwave dielectric
properties of Ca[(Li1/3Nb2/3)1-x M x]O3- (M=Sn,Ti) Ceramics, J. Mater. Res., (1999),
14, 3567-3570.
[33] Choi J.-W., Ha J. Y., Yoon S.-J., Kim H.-j. and Yoon K. H. Microwave dielectric
properties of Ca[(Li1/3Nb2/3)1-x Zr x]O3- ceramics, Jan. J. Appl. Phys, (2004), 43, 223-
225.
[34] Ha J.-Y., Choi J.-W., Kang C.-Y., Yoon S.-J., Choi D. J. and Kim H.-j. Low
Temperature Sintering of Ca[(Li1/3Nb2/3)1-x Tix]O3- Based Microwave Dielectric
Ceramics with Glass Frit, Jpn. J. Appl. Phys., (2005), 44, 1322-1325.
[35] Ha J.-Y., Choi J.-W., Kang C.-Y., Yoon S.-J., Choi D. J. and Kim H.-j. Microwave
Dielectric Properties of Ca[(Li1/3Nb2/3)1-x Tix]O3- ceramics with glass, J.
Electroceramics, (2006), 17, 399-403.
[36] Ha J.-Y., Choi J.-W., Yoon S.-J., Choi D. J., Yoon K. H. and Kim H.-j. Microwave
Dielectric Properties of Bi2O3 doped Ca[(Li1/3Nb2/3)1-x Tix]O- ceramics, J. Eur.
Ceram. Soc., (2003), 23, 2413-2416.
[37] Liu P., Kim E. S. and Yoon K. H. Low-Temperature Sintering and Microwave
Dielectric Properties of Ca(Li1/3Nb2/3)O3 Ceramics, Jpn. J. Appl. Phys., (2001), 40,
5769-5773.
[38] Liu P., Ogawa H., Kim E. S. and Kan A. Low temperature sintering and microwave
dielectric properties of Ca(Li1/3Ta2/3)O3- -CaTiO3 ceramics, J. Eur. Ceram. Soc.,
(2003), 23, 2417-2421.
[39] Liu P., Ogawa H., Kim E. S. and Kan A. Microwave dielectric properties of low-
temperature sintered Ca[(Li1/3Nb2/3),Ti]O3- ceramics, J. Eur. Ceram. Soc., (2004),
24, 1761-1764.
[40] Liu P., Zhang H., Ogawa H. and KIM E. S. Low temperature cofired
Ca[(Li1/3Nb2/3)]O3- Ceramics for microwave applications, phys. stat. sol. (a), (2004),
201, 2175-2179.
[41] Tong J. X., Zhang Q. L., Yang H. and Zou J. L. Low-temperature sintering and
microwave dielectric properties of Ca[(Li1/3Nb2/3)0.9Ti0.1]O3- ceramics with LiF
addition, Mater. Lett., (2005), 59, 3252-3255.
[42] Tong J. X., Zhang Q. L., Yang H. and Zou J. L. Low Temperature Firing and
Microwave Dielectric Properties of Ca[(Li1/3Nb2/3)0.84 Ti 0.16]O3- Ceramics for LTCC
applications, J. Am. Ceram. Soc., (2007), 90, 845-849.
[43] Yoon S. O., Kim D. M., Shim S. H., Park J. K. and Kang K. S. Microwave dielectric
properties of Ca(Li1/4Nb3/4)O3CaTiO3 ceramic systems, J. Eur. Cer. Soc., (2006),
26, 2023-2026.
[44] George S. and Sebastian M. T. Effect of lithium based glass addition on the
microwave dielectric properties of Ca[(Li1/3Nb2/3)1-xTi x]O3- ceramics for LTCC
Applications, J. All. and Comp., (2009), 473, 336-340.
[45] Pang L.-X., Wang H., Zhou D., Zhou H.-F. and Yao X. Low temperature firing and
microwave dielectric properties of Ca[(Li1/3Nb2/3)0.8Ti0.2]O3- ceramics with ZnB2O4
glass addition., Int. J. Appl. Ceram. Technol., (2008), 5, 341-346.
[46] Hakki B. W. and Coleman P. D. A dielectric resonator method of measuring
inductive capacitance in the millimeter range, IRE Trans. Microwave Theory Tech.,
MTT., (1960), 8, 402-410.
[47] Krupka J., Derzakowski K. D., Riddle B. and Jarvis J. B. A Dielectric Resonator for
Measurements of Complex Permittivity of Low Loss Dielectric Materials as a
Function of Temperature, Meas. Sci. Technol., (1998), 9, 1751-56.
[48] George S. and Sebastian M. T. Unpublished data, Unpublished data.
[49] Dai S. X., Huang R. F. and Wilcos Sr D. L. Use of titanates to achieve a temperature
stable low temperature cofired ceramic dielectric for wireless application, J. Am.
Ceram. Soc., (2002), 85, 828-832.
[50] Fukuda K., Kitoh R. and Awai I. Microwave characteristics of TiO2-Bi2O3 dielectric
resonator, Jpn. J. Appl. Phys., (1993), 32, 4584-4587.
[51] Haga K., Ishi T., Mashiyama J. and Ikeda T. Dielectric properties of two phase
mixtures ceramic composed of rutile and its compounds, Jpn. J. Appl. Phys., (1992),
31, 3156-3159.
[52] Kagata H., Kato J., Nishimoto K. and Inoue T. The dielectric properties of Pb based
perovskite substituted by Ti for B site at microwave frequencies, Jpn. J. Appl. Phys.,
(1993), 32, 4332-4334.
[53] Wang Y., Zhang G. and Ma J. Research of LTCC/Cu, Ag multilayer substrate in
microelectronic packaging, Mater. Sci. Eng. B, (2002), 94, 48-53.
[54] Zhang Q. L., Yang H., Zou J. L. and Wang H. P. Sintering and Microwave Dielectric
Properties of LTCC Zinc Titanate Multilayers, Mater. Lett., (2005), 59, 880-884.
[55] Shimoda H., Ishitobi N., Kawamura K. and Kobayashi M. Development of a High-Q
Multilayer Ceramic Resonator, Jpn. J. Appl. Phys., (1992), 31, 3160-3163.
[56] Liang M. H., Wu S.-Y., Hu C.-T. and Li I.-N. Enhancing the sinterability of
Ba(Mg1/3Ta2/3)O3 dielectrics by using chemically-derived powders, Mater. Chem.
Phys., (2003), 79, 276-281.
[57] Lu C. H. and Tsai C. C. Homogeneous precipitation synthesis and sintering behavior
of microwave dielectrics: Ba(Mg1/3Ta2/3)O3, Mater. Sci. Eng. B., (1998), 55, 95-101.
[58] Renoult O., Boilot J. P., Chaput F., Papiernik R. and Hubert-Pfalzgraf L. G. Sol-Gel
Processing and Microwave Characteristics of Ba(Mg1/3Ta2/3)O3 Dielectrics, J. Am.
Ceram. Soc., (1992), 75, 3337-3340.
[59] Shirey H. M. Low temperature synthesis of the microwave dielectric material Barium
Magnesium Tantalate (BMT), (2002), (University of Pittsburgh).
[60] George S., Anjana P. S., Deepu V., Mohanan P. and Sebastian M. T. Low-
Temperature Sintering and Microwave Dielectric Properties of Li2MgSiO4 Ceramics,
J. Am. Ceram. Soc., (2009), 92, 1244-1249.
[61] Jantunen H. A novel low temperature cofiring ceramic (LTCC) material for
telecommunication devices, Ph.D. thesis, Faculty of Science, (2001), (Finland,
University of Oulu).
[62] Ho I.-C. Semiconducting barium titanate ceramics prepared by Boron-containing

liquid-phase sintering, J. Am. Ceram. Soc., (1994), 77, 829-832.
[63] Zhang Q. L. and Yang H. Low temperature firing and microwave dielectric
properties of ZnO-Nb2O5-TiO2-SnO2 ceramics with CuO-V2O5, Mater. Res. Bull.,
(2005), 40, 1891-1898.
[64] Huang C.-L., Weng M.-H., Lion C.-T. and Wu C.-C. Low temperature sintering and
microwave dielectric properties of Ba2Ti9O20 ceramics using glass additions, Mater.
Res. Bull., (2000), 35, 2445-2456.
[65] Huang G., Zhou D., Xu J., Chen X., Zhang D., Lu W. and Li B. Low-temperature
sintering and microwave dielectric properties of (Zr,Sn)TiO4 ceramics, Mater. Sci.
Eng. B., (2003), 99, 416-420.
[66] Jantunen H., Rautioaho R., Uusimaki A. and Leppavuori S. Temperature Coefficient
of Microwave Resonance Frequency of a Low-Temperature Cofired Ceramic
(LTCC) System, J. Am. Ceram. Soc., (2002), 85, 697-699.
[67] Kim D. W., Lee D.-G. and Hong K. S. Low-temperature firing and microwave
dielectric properties of BaTi4O9 with Zn-B-O glass system, Mater. Res. Bull., (2001),
36, 585-595.
[68] Lee Y.-C., Lee W.-H. and Shieu F.-S. Low temperature sintering and microwave
dielectric properties of Ba2Ti9O20 ceramics with 3ZnO-B2O3 addition, Jpn. J. Appl.
Phys., (2002), 419, 6049-6053.
[69] Wang S.-F., Chiang C.-C. and Wang C.-H. Phase stability of B2O3 added Ba2Ti9O20
ceramic: Processing effects, J. Mater. Res., (2003), 18, 201-207.
[70] Jantunen H., Rautioaho R., Uusimaki A. and Leppavuori S. Composition of MgTiO3-
CaTiO3 ceramic with two borosilicate glasses for LTCC technology, J. Eur. Ceram.
Soc., (2000), 20, 2331-2336.
[71] Jean J. H. and Lin S. C. Effect of borosilicate glass on densification and properties of
borosilicate glass + TiO2 ceramics, J. Mater. Res., (1999), 14, 1359-1363.
[72] Shin H., Shin H. K., Jung H. S., Cho S. Y. and Hong K. S. Phase evalution and
dielectric properties of MgTi2O5 ceramic sintered with lithium borosilicate glass,
Mater. Res. Bull., (2005), 40, 2021-2028.
[73] Takada T., Wang S. F., Yoshikawa S., Jang S.-J. and Newnham R. E. Effect of glass
additions on BaO-TiO2-WO3 microwave ceramics, J. Am. Ceram. Soc., (1994), 77,
1909-1916.
[74] Park J. H., Choi Y. J. and Park J. H. Low fire dielectric compositions with
permittivity 20-60 for LTCC applications, Mater. Chem. Phys., (2004), 88, 308-312.
[75] Wu J.-M. and Huang H. L. Microwave properties of zinc, barium and lead
borosilicate glasses, J. Non Cryst. Solids., (1999), 260, 116-124.
[76] Wersing W. Microwave ceramics for resonators and filters, Curr. Opin. Solid State
Mater. Sci., (1996), 1, 715-731.
[77] Kawashima S., Nishida M., Ueda I. and Ouchi H. Ba(Zn,Ta)O3 ceramic with low
dielectric loss, J. Am. Ceram. Soc., (1983), 66, 421-423.
[78] Kawashima S., Nishida M., Ueda I., Ouchi H. and Hayakawa S. Dielectric properties
of Ba(Zn,Ta)O3-Ba(Zn,Nb)O3 ceramic, Paper presented at the Proc.
Ferroelect.Mater. Applications.(1977).
[79] Nomura S., Toyama K. and Tanaka K. Ba(Mg1/3Ta2/3)O3 ceramics with temperature-
stable high dielectric constant and low microwave loss, Jpn. J. Appl. Phys., (1982),
21, L624-626.
[80] Surendran K. P. and Sebastian M. T. Low loss dielectrics in Ba[(Mg1/3Ta2/3)1-x Tix]O3
and Ba[(Mg1-x Znx)1/3]Ta2/3]O3 systems, J. Mater. Res., (2005), 20, 1-7.
[81] Wakino K., Minai K. and Tamura H. Microwave characteristics of (Zr,Sn)TiO4 and
BaO-PbO-Nd2O3-TiO2 dielectric resonator, J.Am. Ceram. Soc., (1984), 67, 278-281.
[82] George S. and Sebastian M. T. Effect of borosilicate glass addition on the
sinterability and microwave dielectric properties of Ca[(Li1/3Ta2/3)1-xTix]O3-
ceramics for LTCC Applications, J. Mater. Sci. Mater. Elect. Communicated (2010)
Chapter 7
PEROVSKITE-TYPE OXIDES:
SYNTHESIS AND APPLICATION IN CATALYSIS
Junjiang Zhu1,2,* and Jinfa Chen3

1Technical University Berlin, Englische Str. 20, 10587 Berlin Germany
2Max Planck Institute of Colloids and Interfaces, Colloid Chemistry, Am Mhlenberg 1,
14476 Potsdam/Golm, Germany
3National Engineering Research Center of Chemical Fertilizer Catalyst, Fuzhou
University, Gongye Road 523, Fuzhou, 350002, Fujian, China
ABSTRACT
Two topics of perovskite-type oxides were browed and discussed in this chapter: one
is the synthesis method and the other is its application in catalysis. Because of the diverse
applications of perovskite-type oxides, special properties are usually required depending
on the ultimate use, while the desired properties could be achieved by choosing a proper
synthesis method. For example, using coprecipitation method could lead to a well
crystallized structure, while by a nanocasting method perovskite with high surface area
could be achieved. Thus, an overview of synthesis method for perovskite-type oxides
reported in the publications was made for better understanding the advantage of each
method.
Efforts on the application of perovskite-type oxides were made exclusively in
catalysis in this part. Attentions on using these oxides as catalyst are increasing year by
year not only because of their low manufacturing cost but also due to the unique
properties, such as high thermal stability, exchangeability of cations and of course,
considerable activity. Many catalytic reactions using perovskite-type oxides as catalyst
has been attempted, and although for a given reaction catalyst with different ions, either
at A- or B-site, show different catalytic activities, a common point is that the presence of
oxygen vacancy is necessary for the reaction to proceed. From the nature of catalytic
reaction, that is the difficulty of electron transferring between reactant and catalyst, it was
*
320 Junjiang Zhu and Jinfa Chen
proposed, exampled by the reactions of NO decomposition and NO+CO reduction, that

the crucial pre-requisite for a catalytic reaction to occur is that the full redox cycle of
catalyst can proceed easily. Thus only when this requirement is fulfilled, other
parameters, such as the amount of active sites, could come into the focus of catalyst
design.
1. INTRODUCTION
Perovskite, which is first used to describe the oxides CaTiO3, is now extensively referring
to oxides with ABO3 structure, and sometimes also to the oxides with A2BO4 structure. In
some cases, carbides, nitrides, halides, and hydrides also have the ABO3 structure,[1] but
practically perovskite refers to the oxides compounds. In this context, if not specified,
perovskite means an oxide compound having the ABO3 or A2BO4 structure.
Ideal structure of perovskite oxides with ABO3 or A2BO4 structure are shown in Figure 1.
The former is cubic with space group Pm3m-Oh,[2] while the latter is composed of ABO3
structure plus an AO layer. Detailed descriptions of these structures have been made in the
original publications and in some previous reviews,[1-4] here we just give an initial concept
of perovskite. As shown in Figure 1, A is a larger cation that locates on the edge of the
structure, and B is a smaller cation locating in the center of the octahedron. The A- and/or B-
site cations could be substituted by a foreign one, which would lead to the distortions of
structure, however not destroying the matrix structure, if the cations in the structure could
satisfy the requirement of t = ( rA + rO ) /{ 2 (rB + rO )} , where t is in the range of 0.75 ~ 1.0.[5]
Figure 1. Ideal models of perovskite-type oxides with ABO3 and A2BO4 structure. The red cycle means
that A-site cation was substituted by a foreign one; the blue square means the oxygen vacancy. Note:
the oxygen symbol was not shown in the A2BO4 structure for simplification. Reprinted from ref. 73,
copyright 2009, with permission from Elsevier
Perovskite-Type Oxides: Synthesis and Application in Catalysis 321
Figure 2. Schematic presentation of particle formation by precipitation in a microemulsion. Reprinted

from ref. 24, copyright 2004, with permission from Elsevier
Attentions on perovskite oxides have been largely paid in the last decades due to their
special structures and unique properties. It was mentioned that 90% of the metallic natural
elements of the Periodic Table are stable in a perovskite structure and multicomponent
perovskites could be synthesized by partial substitution of cations at either A- or B-site
without destroying the matrix structure.[1] This allows tuning the oxidation state of cataions
(especially for these at B-site) or creating oxygen vacancies with desired properties, thus
offering the possibility of correlating these properties to the chemical performances directly.
As a result, these oxides are of highly concerned in scientific research since clear and
confident correlation could be plotted simply by partial substitution of the cations.
Application of perovskite oxides have been extensively studied [6-13] in such as, solid
state chemistry, physics, advanced materials, and catalysis. Because of these diverse
applications and special requirements in each application, perovskite-type oxides with special
properties are required depending on the ultimate end use. For example, materials-oriented
applications require densification by high temperature sintering to minimize both surface area
and surface free energy in order to maximize mechanical strength. In contrast, catalytic
materials have to maintain sufficiently high surface area in order to maximize their
participation and activity in chemical reactions.[14]
The requirement for these special properties mostly could be achieved by choosing a
proper synthesis method, such as the solid-state reaction, citric acid combustion or chemical
vapor deposition. The synthesis of new perovskite materials has been an ongoing interest for
materials scientists and a combination of perovskites with other materials has proved to be an
efficient approach to enhance the desired properties and develop new applications. Therefore,
to find a synthesis method with novel processing undoubtedly is an interesting subject of

future tasks. [3]
In this chapter, synthesis methods and applications of perovskite-type oxides in catalysis
were reviewed. However, since the reported literatures on these two topics are really
extensive, we do not have the attempt in making an exhaustive coverage for them. Here, only
a series of typical examples is given to illustrate the synthesis method and efforts made on the
application of perovskites concentrated mainly in the field of catalysis.
2. SYNTHESIS METHODS FOR PEROVSKITE OXIDES

The preparation route plays a critical role on the physical and chemical properties of
perovskites. The approach to synthesis of perovskites must be examined in function of the
specific demands concerning activity and/or selectivity, both of which depend on the
arrangement of atoms at a small scale. A general objective is to locate the different
constituting atoms at precise positions. The main difficulty is to transform the starting
precursor into a highly porous solid without segregation of different elements that would
produce tiny parts with different properties.
A. Solid-state Method
The simplest method for synthesizing perovskite is by grinding or ball-milling the mixed
oxides, either in dry or wet slurry. The mixture was then calcined at temperatures of 600
1000 C to obtain the final product. This operation sometimes needs to be repeated several
times to obtain a pure pervoskite. However, it should be noted that the grinding conditions,
such as the rotate rate, grinding time and grinding atmosphere, have significant effect on the
surface area and catalytic performances.[15] The solid-solid reaction can be accelerated by
priming or catalyzing by a small amount of additive pre-mixed with the reactant oxides,
presumably for increasing surface area and nucleating the desired phase.[16, 17] However,
when an additive is used, the perovskite product should be thoroughly washed, in order to
free the sample from the residues of the additive.
Nanoparticle perovskite could also be synthesized by this method with the assistance of
oriented substrate. The surface energy of perovskite changes with the orientation of the
substrates due to variation in atomic arrangements. The resulting particle shape is polyhedral,
and is composed of planes that minimize the surface and interfacial energies between the
individual components and that between the components and the substrates.[18]
Whereas it should be noted that the solid-state method is difficult to make the mixed
oxides precursors in a uniform state, therefore, in order to prepare perovskite with
homogeneous phase, other techniques, such as ultrasonic treatment, could be used to
homogenize the mixture to improve the quality of the formed perovskite.
In all, this method could be easily operated but is not so preferable, especially in
catalysis, as it gives wide grain size distributions, bad homogametic phases and further needs
very high calcination temperatures, which is energy consuming. Also, due to the
agglomeration at high temperature, it shows a relatively low surface area.
B. Wet Method
Other than that in solid-state method, the precursors could be homogeneously mixed if
they are presenting in gaseous or liquid phase. Precursor in liquid phase is more convenient,
because it has no long-range structure, permitting a statistical distribution at molecular scale.
In addition, the movements of molecule in liquid state is of the order of magnitude of Van der
Waals forces or hydrogen bond forces, that is a few kJ/mol, while diffusion in solids
necessitates the loosening of ionic forces or the breaking of covalent bonds, the strength of
which is much higher than that in liquid state. Therefore, perovskite synthesized by chemical
solution methods show fine and homogeneous particles with relatively high specific surface
areas.
B1. Coprecipitation
Coprecipitation is a common method for the preparation of mixed-oxide catalyst.
Generally, nitrates, carbonates and oxalates could be used as the metal source, while NaOH,
KOH or (NH3)aq. as precipitant agent. Aqueous ammonia is preferable for precipitation,
especially in the case that foreign ions (e.g. Na+, K+) have negative effect on the synthesized
catalyst. An exception to this is when (NH3)aq has the possibility of forming metal-ammine
complexes (e.g. Ni, Co, Cu, Zn). Following is an example using coprecipitation method for
the preparation of La1-xCaxFeO3 oxides: [19] 0.1N sodium hydroxide solution was dropped
one by one to 0.1N of corresponding nitrates precursors at room temperature until the pH
value reaches 10, which was then aged for 48 h under the same conditions. The resulting
precipitate was filtered off, washed with distilled water to remove excess sodium ions until
reaching constant conductivity value, and then dried in air at 110 C overnight. The obtained
solids were finally calcined in air at 700 C for 6h.
In this method two major steps are involved in, namely, nucleation and growth. To obtain
a highly dispersed precipitate, it is necessary to accelerate the nucleation rate while suppress
the growth rate, which could be achieved by the following techniques:[20]
(1) A vigorous stirring of the solution when adding the chemical compound. This allows
the ions reaching the over-saturation status rapidly, and hence accelerates the
nucleation speed.
(2) A rapid addition of precipitating agent ensures a rapid reaching of supersaturation in
the whole volume of the solution, leading to a maximum nucleation rate.
(3) Because higher temperature will promote the growth rate of new phase, precipitation
is usually performed at the lowest temperature, though the situation practically is
different from case to case.
(4) Other specific steps, for example the addition of an extra component, re-dispersion of
hydroxide, coprecipitation in an aqueous medium by sonication, etc.
However, for compound oxides, because the precipitation point for each component is
different, segregation almost invariably occurs when liquids are reacted to give solids. Hence
it is difficult to control the stoichiometry and homogeneity of cationic distribution in the
grains.
B2. Citrate Method

Compared to coprecipitation method, the citrate one which consists of preparing stable
amorphous or glassy precursors guarantees stoichiometry and homogeneity of cation
distribution and tolerates wide range of compositions. The starting solution is a mixture of
metallic salts and a hydroxyacid such as citric, malic, tartaric, lactic, or glycolic acid. Citric
acid so far has been used most frequently because it is cheap and stable, and hence the name
citrate is attached to the method.
Typically, the concentrations of organic acid are higher than that of the total metals, and
the solution is concentrated by rapid evaporation under vacuum, until a gel is obtained. So far,
the solution can be dehydrated under vacuum at about 80 C to give a rigid-foam (meringue)
precursor, which is amorphous and hygroscopic, and exhibits the color of the metallic ions.
The amorphous precursor is some sort of a three-dimensional polymer in which the
multifunctional organic acids are linked to two or more cations by covalent or hydrogen
bonds. The thermal decomposition of the amorphous mass could proceed continuously or in
two distinct steps at 80100 and 250400 C. The resulting powders, amorphous citrate
precursor, were crushed and sieved to obtain the required particle size. Finally, the solids
were calcined at a certain temperature for several hours to obtain the product.
According to literature [21], the hydroxy acids with one hydroxy group and two or more
carboxylic groups seem to be more preferable. While at this time the properties of perovskite
would be changed depending on the metal source. Metal nitrates were considered to be the
best one since it will not leave any residues during thermal decomposition. Also, the use of
nitrate would promote the surface area and porosity of sample, due to the large amount of gas
produced during the decomposition. The metal acetates or chlorides were not as effective as
nitrates in producing the perovskite phase at lower temperatures.
However, because of the violent reactions during the thermal decomposition of
amorphous mass, it should be carried out carefully, taking all necessary precautions. The
preparation of oxides containing metallic ions which are highly active in oxidation needs very
stringent precautions and is generally not advisable, especially for copper and silver salts. In
these cases, the freshly formed metal particles could be powerful catalysts for some
unexpected reactions, e.g. with ammonia. Also, secondary reactions can occur even at room
temperature resulting in the formation of a small amount of lethal cyanogen gas.[20]
Many attempts have been made to modify the citrate method, and a promising one is
called Pechini method, in which polyalcohol is introduced to the initial solution.[22] The
role of polyalcohol is supposed to promote the formation of reticulation from the starting
solution to the rigid foam. The reticulation or polymer formation is due to the esterification
reaction between citric acid and the poly-alcohol.
B3. Microemulsions Method

Microemulsions are transparent, thermodynamically stable multicomponent fluids,
normally are composed of an aqueous component, an oily component, an amphiphile (used as
emulsifying agent) and frequently a co-surfactant. There are two different types of
microemulsions: (a) oil in water microemulsion (o/w-microemulsion), in which the
hydrophobic part of the surfactant is absorbed on the surface of oil microdroplets while the
hydrophilic part is oriented toward the water; and (b) water in oil microemulsion (w/o-
microemulsion), in which the hydrophilic part of the surfactant is absorbed at the surface of
water microdroplets while the hydrophobic part is oriented toward the oil. The latter is also
called reverse microemulsion. With increasing the water concentration the nanodroplets of
microemulsion increases and eventually forms aggregates or infinite clusters.[23] At this
stage the microemulsion processes a bicontinuous structure. The reverse microemulsion
method, utilizing aqueous solution nanodrops surrounded by surfactant molecules, has been
used extensively for the synthesis of many kinds of materials including perovskites.
Figure 2 shows a schematical double microemulsion technique, where two
microemulsions with equal watersurfactantoil ratio are mixed. One contains metal ions and
the other the precipitating agent.[24] Due to the dynamics of the system,[25] the water in the
droplets mixed rapidly, leading to the formation of particles by precipitating inside the
microemulsion droplets. The perovskite phase can then be formed at a low calcination
temperature, due to the homogenous mixing of the metals and the small-size primary
particles. Low calcination temperature avoids the particle growth and thus preserves high
surface area. Compared to these prepared by co-precipitation in water, the materials by
microemulsion technique show more homogenous morphology consisting of smaller
crystallites.
B4. Combustion Method

The combustion method, which is used to synthesize oxides with fine particles, represents
an interesting alternative due to the following advantages: relatively cheap starting reactants,
simple and easily available organic molecules as fuel and highly exothermic and self-
sustaining reactions. In this method, the appropriate amount of metal nitrates and organic
fuels were first added to de-ionized water, the mixtures were then thoroughly stirred to
eliminate the water at 6070 C until homogeneous sol-gel was formed, subsequently the sol-
gel was ignited in air and fine oxides particles were obtained. These fine particles were finally
calcined at desired temperature to form the perovskite structure.
Organic molecules (e.g. glycine, alanine, glycerol and urea) play a double role in this
process: (i) as complexing agent to homogenize metal cations during evaporation and (ii) as
fuel to react with the internal oxidants (nitrogen oxides formed during the decomposition of
the nitrates) and/or atmospheric oxygen to form the fine oxide particles. It has been reported
that LaMnO3 and LaCoO3 prepared by combustion technique showed the highest thermal
stability and the best catalytic activity for methane combustion than that by other synthesis
routes.[26] While perovskites synthesized by different organic fuels result in different
temperature attained in the combustion process, and hence different BET surface areas,
residual carbon contents and catalytic activities. For example, Specchia et al. [27] found that
the use of glycine as a fuel for the synthesis of LaMnO3-based catalysts is more preferable
than that of alanine and glycerol. They also investigated the effect of fuel/oxidizer ratio and
found that the optimum value is at ~1:1. Besides, Wang et al [28] recently investigated the
effect of complex on the catalytic performances of perovskite for CO oxidation reaction,
finding that LaSrNiO4 catalyst prepared by using oxalic acid showing the best and stable
activity, while that by citric acid is the lowest one.
However, the specific surface areas of perovskites synthesized by combustion method
were not dramatically higher than those obtained by traditional methods. Further the sample
synthesized by this way was found to be rapidly deactivated due to sintering, even by a short
thermal treatment. Therefore, in this method NH4NO3 was usually chosen as an additive,
because it can enhance the specific surface area and resistance to thermal treatment, while
without leaving any residues and altering the proportion of the components.[29] The main
drawback of using NH4NO3 is that hazardous or polluting compounds are emitted during the
synthesis (mainly NH3 or NOx).
B5. Freeze-drying
Usually slow evaporation of a homogeneous solution containing two different precursors
leads to two separate salts in the solid product. To prevent this, the elimination of physical
diffusion process would be useful, which could be achieved by a rapid cooling technique, i.e.
freeze-drying. Freeze-drying is a versatile method that can produce mixtures of phases with
high homogeneity, although there is, in principle, no atomic level homogeneity.
Following is an example for the preparation of NdCoO3 perovskite by freeze-drying
method: [30] 0.005 mol neodymium sesquioxide was first dissolved in 5 ml of concentrated
nitric acid, and then added to a solution of 0.01 mol of cobalt (II) acetate in 25 ml of water.
The pH of the resulting solution was adjusted to 4.5 by addition of dilute ammonia. Finally,
the resulting reddish solution were frozen in liquid N2 and transferred into a freeze-dryer
operated at a pressure of 7.6 torr. The resulting blue-pink powder was placed into an alumina
boat and pretreated at 400 C for 2 h under vacuum. The "ashes" were finally ground and
treated for 24 h under a dynamic oxygen atmosphere at 700 and 900 C.
The use of the freeze-drying allows obtaining a relatively high surface area and
precursors show perfect homogeneity in grain composition, hence the product could be easily
synthesized at lower temperatures. However, the important parameters in freeze-drying are
cooling rate and the final temperature of the solidified solution. These can be controlled but
only with a small scale. Hence the method is not well suited to large-scale manufacture of
catalysts.
B6. Template Method

By filling the interspaces of the template with the material (precursor) to be cast, and
subsequently calcinating and removing the mold, materials with replica structure can be
obtained. The use of a confined space is of great importance because it restricts the growth of
the material during the synthesis and facilitates the formation of nanostructures. By this way,
numerous mesoporous metal oxides, including perovskite, could be synthesized.[31-33] For
the synthesis of perovskite, complex (e.g. citrate acid) usually is used to guarantee
stoichiometry and homogeneity of cation distribution in the mother solution.
One typical procedure for nanocasting synthesis of mesoporous LaCoO3 perovskite is as
below:[33] stoichiometric lanthanum and cobalt nitrates were first dissolved in the mixed
citrate solution of deionized water and ethanol, with the molar ratio of citric acid to total
metal cations (La + Co) at 1:1. So far, 4 mmol lanthanum nitrate, 4 mmol cobalt nitrate and 8
mmol citric acid were dissolved in a mixture of 15 mL ethanol and 5 mL deionized water. A
homogeneous solution was obtained after stirring at room temperature for one night. Then 1.0
g of hard template was added to this solution, the mixture was stirred for another several
hours at 40 C until it became viscous, then dried at 80 C for 6 h and calcined at 500 C for
4h to decompose the nitrate species. To achieve higher loadings, the impregnation could be
repeated with the same procedure. The resulting material was calcined at 700 C for 4 h
(ramp, 2 C /min) in air. The silica framework was then etched by using 2M NaOH aqueous
solution at room temperature for three times. After being centrifuged, washed with water and
ethanol, and dried at 80 C overnight, finally the black LaCoO3 powder was obtained.
The advantage of this strategy is that the synthesis of the porous materials takes place in a
confined space formed by the porosity of the template. Thus-synthesized perovskite has a
high surface area with three-dimensional ordered mesoporous channels, which can not be
achieved by conventional methods. Moreover, template-based growth is commonly a solution
or colloidal dispersion based process. It is less expensive and readily scalable to mass
production. It also offers the advantages of less contamination and is environmentally benign.
[34]
However, the removal of template is difficult and requires severe conditions, either strong
basic medium (e.g. NaOH) or acidic medium (e.g, NH4HF2), which may destroy the structure
of the produced perovskite. Further, not all of perovskites can be successfully synthesized by
templating. It is difficult to find appropriate templates with pore channels of desired diameter,
length and surface chemistry and to remove the template completely without compromising
the integrity of the nanostructured perovskites. Gao et al [35] concluded that the rules of
choosing templates to get highly textured perovskites are: (1) Template with large anisotropy
dimension can be easily prepared; (2) The elements of templates composition should be
involved in the composition of matrix perovskites and can be easily removed; (3) Lattice
structure of templates should be similar to that of matrix perovskites; (4) Templates should
have high reactivity or diffusion with matrix particle to ensure a single phase; (5)The
morphology and the size of template should be uniform for tape casting.
3. APPLICATION IN CATALYSIS
A. Photocatalysis for Water Splitting
It has been reported long time ago that some of perovskite-type oxides, e.g. SrTiO3, can
decompose H2O into H2 and O2 with no external bias potential.[36] The principle for this
process is that the conduction band (CB) edges of some of the perovskite oxides are more
negative than the H+/H2 energy level.[1] However, only a few work on using perovskite
oxides as photocatalyst is reported up to date and study on this field is still in its infancy.
In most cases however the perovskite alone shows no or little activity for water splitting
and thus introduction of a foreign metal on the substrate is required, e.g. NiO/SrTiO3.[37] The
introduction of the foreign metal is to decrease the H2 overvoltage in the process, since high
H2 overvoltage leads to low efficiency for water reduction. Accordingly, it was proposed that
the oxidation and reduction processes in water splitting occur on two different sites: the
oxidation of oxygen (2O O2) is on the perovskite semiconductor and the reduction of
hydrogen (2H H2) is on the supported metal (Note: electron transferring was not marked
for simplification).[37]
As expected the activity of perovskite catalyst for water splitting could be altered through
partial substitution of cation by a foreign one. A typical example is the substitution of Ta5+ by
Ti4+ in KTaO3.[38] As shown in Figure 3, the activity of KTa1-xTixO3-x/2 for water splitting
changes with the amount of Ta being substituted and the best activity appears at x = 0.08.
Other parameters, such as calcination conditions and preparation methods, also have
great effect on the activity of the catalyst. In a recent work Hu et al [39] investigated the
preparation methods (Solgel, hydrothermal and solid-state methods) on the crystalline
structure and photocatalytic activity of NaTaO3 perovskite for water splitting, finding that the
catalyst prepared by sol-gel method, which has a monoclinic phase, shows the best activity
for water splitting, as shown in Figure 4. Also they concluded that the photocatalytic activity
of the NaTaO3 specimens in water splitting has the same order as that of the Ta-O-Ta bond
angle observed in the structure.
Other catalytic reactions, especially for those using hydrogen as reactant, are also
explored in photocatalysis, due to the production of H2 from the photocatalytic water
splitting. Two typical cases are the reduction of CO2 [40-42] by H2O and the synthesis of NH3
from N2 and H2O.[43, 44] The later indicated that the thermodynamically unfavorable
reaction of NH3 formation from N2 and H2O can be produced under band-gap irradiation of
these catalysts.
B. Electrocatalysis
The solid oxide fuel cell (SOFC) is an electrochemical device that can be used for either
stationary or mobile generation of electrical energy from a gaseous fuel.[45] Perovskite
oxides could be good cathode materials for SOFC with an efficiency ranging from 50 to
65%,[46-48] because they possess high-electronic and ionic conductivities, porosity, thermal
and chemical compatibilities with the electrolyte, and long-term stability, which are essential
for cathode of SOFC. Also, some unique properties of perovskite oxides, such as the
exchangeability of cations and the possibility of creating oxygen vacancy without destroying
the matrix structure, make them a prominent candidate in fabricating the cathode of SOFC.
Figure 3. Formation rate of H2 and O2 as a function of Ti content in KTaO3 (loaded with 1wt% NiO).
Reprinted from ref. 38, copyright 1999, with permission from Chemical Soc. Japan.
Sr-doped LaMnO3 (La1-xSrxMnO3) perovskite has been extensively studied for decades as
a cathode material in ZrO2-based SOFC, due to its relatively high electrocatalytic activity for
O2 reduction and good thermal and chemical compatibility with ZrO2- based electrolyte. [1,
49] However the Mn-based perovskite usually are lacking of oxygen vacancy, thus the oxide-
ion conduction at the working temperature, makes it necessary to use thick, porous electrodes
containing an array of triple-phase boundaries where gas, electrolyte and electrode meet.
Also, because the conventional SOFC using La1-xSrxMnO3 as the cathode material
usually requires high operating temperature (~1000 C), which leads to some serious
problems, such as a high possibility of interfacial reaction between the electrode and
electrolyte to form insulating phase(s), and the densification of the electrode layer due to
high-temperature sintering, thus the search for new cathode candidate that can work at a
relative low temperature range is essential.[50] Shao and Haile [51, 52] recently found that
Ba0.5Sr0.5Co0.8Fe0.2O3 (BSCF) perovskite oxide can show high electrochemical activity at
temperature of 600 C and a detailed review regarding the using of BSCF in SOFC was given
more recently.[50]
Another application of perovskite in electrocatalysis is for electrochemical reduction of
some pollutes such as NO. For example, Simonsen et al [53] investigated the simultaneously
reduction of NO and O2 over La2-xSrxNiO4 based electrodes. They found that, based on CV
measurement, the reduction of NO is more preferred than O2 at -100 mV (see Figure 5) and
the electrode made by LaSrNiO4 showed the best result. This offers an option for NO removal
in the presence oxygen, which practically is a challenging topic in heterogeneous catalysis.
Figure 4. Photocatalytic H2 (empty) and O2 (full) evolution from 900 mL pure water suspended with 1 g
of the solgel, hydro-thermal, and solidstate NaTaO3 photocatalysts, which have surface areas of 23,
12, and 0.6 m2/g, respectively. Reprinted from ref. 39, copyright 2009, with permission from Wiley
C. Heterogeneous Catalysis
Different from the two above-mentioned catalysis, heterogeneous catalysis does not
involve extra substance (such as photos, electrons), and all the behaviors performed in the
reaction are done by the catalyst itself. A review of literature shows that the amount of works
reporting the application of perovskite catalyst in heterogeneous catalysis is far more than that
in photocatalysis or in electrocatalysis. Possible reasons could be the various reactions in
heterogeneous catalysis and no restriction for the catalyst is required at this time, which is
different from that for Photocatalysis or Electrocatalysis, where special properties, such as
negative energy level, high-electronic and ionic conductivities are usually required.
Figure 5. CVs obtained for LaSrNiO4 electrode at 400 C. The measurements are performed in 10% O2
in Ar and 1% NO in Ar at 1 mV/s. Reprinted from ref. 53, copyright 2009, with permission from
Springer
Figure 6. Temperature dependence of the intrinsic activity in the catalytic CO oxidation on LaCuO3-
catalysts: (a) LaCuO2.55; (b) LaCuO2.65; (c) LaCuO2.73; (d) LaCuO2.95 and (e) La2CuO4. Reprinted from
ref. 63, copyright 2000, with permission from Elsevier
C1. Liquid Phase Reaction

Liquid phase reaction using solid catalyst is an interesting topic in fine chemicals.
Comparing with homogeneous catalysis where liquid catalyst is used, heterogeneous reaction
has the great advantage in separating catalyst from the product. Perovskite oxides due to its
low cost, diverse properties and of course, considerable activity thus received great attentions.
Another fact that should be pointed out is that the active site of perovskite oxides, other than
that of supported noble metal catalyst, will not be leached in the reaction.
Liu et al [54] first investigated the catalytic behavior of La2-xSrxCuO4 perovskite oxides
for liquid hydroxylation of phenol using H2O2 as the oxidant. A considerable value for phenol
conversion (50.9% for La2CuO4) was observed and especially, they found that the activity of
catalyst was unchanged even after ten runs of the reaction, indicating that perovskite oxides
are very stable catalyst for liquid phase reaction. Notably, they pointed out that copper with
low oxidation state is more favorable for the hydroxylation reaction, while side reactions
became prominent when copper has high oxidation state.
Recently, Resini et al [55] investigated the catalytic behaviors of La0.8Sr0.2Mn0.98O3 for
catalytic wet oxidation of phenol through tuning the temperature, catalyst loading, phenol
concentration and stirrer speed used in the experiment. A significant activity was observed at
low oxygen pressure and only a very small formation of toxic side products, e.g.
hydroquinone, p- and o-benzoquinone. They showed that the reaction is first-order with
respect to phenol concentration but there has a transport limitation of phenol to the catalyst
surface, which however could be overcome by increasing the catalyst loading.
Also, comparison of preparation method on the catalytic behavior of perovskite oxides
for liquid reaction was investigated. Kulkarni et al [56, 57] prepared a series of La1-xSrxFeO3
catalyst for reduction of nitrobenzene to aniline by microwave irradiation method, finding
that the activity is similar to that prepared by conventional methods. The advantage in using
microwave method is that it does not involve intermittent grindings and calcinations at
elevated temperatures.
C2. Gas Phase Reaction
C2.1. CO Oxidation
The oxidation of CO over perovskite oxides has been largely investigated not only with
the aim for removing this harmful gas, but also for using as a probe reaction to study the
defect chemistry of these compounds. The first report using perovskite oxides as catalyst for
CO oxidation was carried out by Parravano in the early 50s, [58] while it is still extensively
investigated up to date.
Generally it was believed that the CO oxidation over perovskite catalyst proceeds via a
suprafacial process, in which the electronic band structure near the Fermi level plays a key
role. [59, 60] That is, the reaction proceeds via the adsorbed CO and oxygen. However, the
lattice oxygen also plays an important role in the reaction, since it could affect the oxygen
mobility in the bulk, the redox behavior of the metal ion and in some certain conditions, it
could also be transferred into adsorbed oxygen.
Because molecular oxygen usually is inert, it must be activated and dissociated into
atomic oxygen before participating in the reaction.[61] 18O isotope experiment gave evidence
that the oxygen adsorption and activation on perovskite catalyst proceeds on the oxygen
vacancy, which is formed due to the losing of lattice oxygen.[62] Hence, the amount of
oxygen vacancy or the mobility of lattice oxygen would have great effect on the CO oxidation
activity. Indeed, Falcon et al [63] have investigated this relationship by testing the CO
oxidation activity over LaCuO3- ( = 0.05 0.45) catalyst system, in which the concentration
of oxygen vacancies can be gradually changed by controlled annealing under high oxygen
pressures at temperatures from 800 to 1000 C. The results shown in Figure 6 indicate that
although all the samples show measurable and similar activity at 150 C, an obvious
difference in the activities is observed with increasing the reaction temperature. The
orthorhombic LaCuO2.55 with the largest oxygen defect shows the best activity for CO
oxidation, while La2CuO4 with no oxygen vacancy shows the lowest one.
Au et al [64] investigated the influence of oxygen (both adsorbed and lattice) on the CO
oxidation activity by using ACo1-xBixO3+ as the catalyst. They found that the incorporation of
Bi could lead to the lattice oxygen desorbing at a lower temperature and improve the mobility
of lattice oxygen, as evidenced by O2-TPD and H2-TPR experiments (see Figure 7). They also
suggested that the redox reaction of cobalt ions plays an important role in the reaction, and in
the present case the activity could be attributed to the extent of the redox processes:
2Co4+ + O2-
lattice
2Co3+ + (1/2) O2
(1)
2Co3+ + O2-
lattice
2Co2+ + (1/2) O2
(2)
Co2+ + O2-
lattice
Co0 + (1/2) O2
(3)
Bi5+ + O2-
lattice
Bi3+ + (1/2) O2
(4)
2Bi3+ + 3 O2-
lattice
2B0 + (3/2) O2
(5)
Effect of electrochemical properties of perovskite catalyst on CO oxidation reaction was

recently investigated through comparing with CH4 complete oxidation reaction, which is
usually considered to be an intrafacial reaction. Results indicate that for suprafacial CO
oxidation, its activity depends mainly on the redox peak area of the catalyst, while for
intrafacial CH4 oxidation, the activity depends on the symmetry of redox potentials of the
catalyst.[65, 66] Dependence of CO oxidation activity on the redox peak area was shown in
Figure 8, which clearly showed that they have the same trends, indicating that the redox peak
area of catalyst is a very important parameter in deciding the activity of CO oxidation.
Mechanism of CO oxidation over perovskite catalyst, according to the results of Tilset et
al, [67] must involve reduction and re-oxidation of the catalyst. That is, the catalyst is first
oxidized by oxygen and then reduced by CO. At this time the oxygen must be in adsorbed
state while the CO could be in gas (i.e. Eley-Rideal mechanism) or adsorbed state (i.e.
Langmuir-Hinshelwood mechanism), as below: [68]
CO(g) + O(s) CO2 (g) (1)
CO(g) CO (s) (2)
CO(s) + O(s) CO2 (s) CO2 (g) (3)
However, because carbonate is usually observed in the experiment, which suggests that
the reaction process is not as simple as mentioned above, a new mechanism containing
carbonate formation and decomposition steps were thus proposed, denoted as reactions (4)
and (5): [68]
CO(s) + 2O(s) COOO (s) (4)
COOO (s) COO (s) + O (s) COO (g) + O (s)) (5)
C2.2. NO Removal
Although the reaction of nitrogen oxide to nitrogen and oxygen (NO = 1/2N2 + 1/2O2) is
thermodynamically favored it is kinetically hindered.[69] For NO decomposition it involves
the dissociation of NO to N2 and O2 on a catalytic surface and that nitrogen readily desorbs as
N2 while oxygen is strongly retained on the surface. Therefore, the catalyst used for NO
decomposition should possess the ability not only for NO dissociation, but also for oxygen
desorption. This thus requires the presence of oxygen vacancy that are using for oxygen
migrating and colliding to form oxygen molecule (i.e. O + O O2).
Presence of oxygen vacancy in the catalyst is very important for oxidation reaction
includes NO removal reaction. Besides the above-mentioned function for oxygen formation
and desorption, oxygen vacancy also is the site for NO adsorption and activation. Therefore, a
competitive adsorption between O2 and NO usually is unavoidable for most of catalysts, in
particular when NO decomposition reaction is carried out in the presence of oxygen. Thus,
the preparation of perovskite catalyst, which possesses oxygen vacancy and with highly
regeneration frequency is of great importance for NO decomposition. Zhao et al [70] reported
that in NO decomposition La2NiO4 without oxygen vacancy showed less than 20% NO
conversion even at 900 C, while a ~95% NO conversion was observed over LaSrNiO4+ that
possesses large amount of oxygen vacancy ( = -0.16). However, it should be pointed out that
for Ca2Fe2O5+ that have lots of oxygen vacancies, it also can not show activity for NO
decomposition, but high activity could be shown when Ca was replaced by Sr, i.e. Sr2Fe2O5+.
The reason is that the oxygen vacancy in the former is in ordered structure, while that in the
latter is disordered one.[71] That is, only for oxygen vacancy with disordered structure, can it
work for NO decomposition.
Figure 7. TPR and O2-TPD profiles of La0.8Ba0.2CoO2.94 (a, a) and ACo0.8Bi0.2O2.87 (b, b). Reprinted
from ref. 64, copyright 2001, with permission from Springer
o
100 250 C 200
o
300 C
Peak Area
75 150
Peak Area (a.u.)
CO conversion (%)
50 100
25 50
0 0
La 2CuO 4 La 1.5Sr0.5CuO 4 LaSrCuO 4
Figure 8. Relationship between CO conversion and redox peak area over La2-xSrxCuO4 (x = 0.0, 0.5,
1.0). Reprinted from ref. 66, copyright 2006, with permission from Springer
Figure 9. Active site structure of La1-xCexSrNiO4 for NO decomposition (: oxygen vacancy).

Reprinted from ref. 77, copyright 2005, with permission from Elsevier
Substitution of both A- and B-site cation was largely investigated [72-76] in order to
correlate the relationships of physiochemical properties and NO decomposition behaviors.
Usually the substitution of A-site cation is to change the physical properties of catalyst, such
as the ability to adsorb NO, the amount of oxygen vacancy and/or the oxidation state of B-site
cation; while the substitution of B-site cation is to change the chemical properties of catalyst,
i.e. the redox behavior of catalyst. That is, the A-site cation of perovskite oxides is
catalytically inactive. However, this conclusion is obtained only when the A-site cation has a
permanent oxidation state (e.g. La3+, Sr2+) and things maybe different if the A-site cation has
changeable oxidation state. Actually, it was reported [77] that the substitution of Ce for La in
La1-xCexSrNiO4 leads to a new active site for NO decomposition, as shown in Figure 9, which
indicates that the A-site cation is not inactive and it can be one component of the active site
for NO decomposition if it has changeable oxidation state.
Figure 10. Two proposed active site structures for NO decomposition reaction carried out on
perovskite-type catalyst. Model a was proposed by Shin et al [78]; model b was by Zhu et al. [79].
Reprinted from ref. 78 and 79, copyright 1979 and 2007, with permission from Elsevier and American
Chemical Society.
The above discussion indicates that both oxygen vacancy and metal ions are component
of active site, but the exact form of active site is till in controversial at present. Shin et al [78]
based on IR results proposed that the active site contains two adjacent oxygen vacancies and
one transitional metal, while recently Zhu et al [79] based on a theoretical calculation and the
catalytic activity of LaSrCuO4 with T and T* phases suggested that the active site should
contain two oxygen vacancies, two transitional metals and one lattice oxygen. A comparison
of the two active site structure was shown in Figure 10. In route (b) there involves a lattice
oxygen between the two transition metals, which suggests that the properties (e.g. mobility)
of lattice oxygen would have great effect on the reaction. Actually, there have lots of work
suggest that the NO decomposition activity has close relation to the properties of lattice
oxygen based on the measurements such as O2-TPD and H2-TPR.
Anyway, it means that the active site has several components and hence how to
characterize it suitably becomes difficult, as the traditional method was used only for oxygen
vacancy or for transitional metal. In order to solve this problem and to check the overall redox
properties of catalyst, cyclic voltammetry (CV) measurement were introduced.[80-82] The
use of CV has the advantage of measuring the catalytic behavior of the whole catalyst,
irrespective of the influence of each component. Table 1 listed some parameters calculated
from the CV curves of several perovskite catalysts, together with their activity for NO
decomposition and for NO+CO reduction. Results indicate that the NO decomposition
activity depends mainly on the symmetry of redox potentials, while the NO+CO activity
depends on the redox peak area. Concerning the major difference in these two reactions is the
presence of CO in NO+CO reaction, the change of activity in Table 1 could be explained by
the following reasons: 1) for NO decomposition, in which no reducing agent (i.e. CO) was
present, the closing cycle (steps of reduction and re-oxidation) depends fully on the redox
ability of catalyst. The smaller the difference in the redox potentials, the faster the reaction
proceeds. As a result, LaSrNiO4 with the smallest E shows the best activity for NO
decomposition. 2) While for NO+CO reaction, the apparent redox potentials of catalyst,
according to Nernst Equation, could be reduced due to the presence of CO reducing agent,
that is, the reducibility of catalyst could be compensated at this time. Therefore, the role of
E becomes weak and that of redox peak area (which represents the amount of active sites)
turns to be the major factor in deciding the activity. As a result, LaSrCuO4 with the largest
redox peak area shows the best activity for NO+CO reduction.
Table 1. Electrochemical properties and catalytic performance of various catalysts for

NO decomposition. Reprinted from ref. 80, copyright 2005, with permission from Elsevier
Catalysts NO conv. NO conv. Redox peak Red. Pot. Ox. Pot. E (V)e
(%)a (%)b area (a.u.) (V)c (V)d
La2CuO4 1.0 1.9 5 2.49 3.35 0.86
La1.5Sr0.5CuO4 60.0 9.8 40 2.0 2.9 0.9
LaSrCuO4 96.8 34.3 188 1.30 3.13 1.77
LaSrCoO4 39.9 20.3 2.15
LaSrNiO4 80.1 94.7 22 2.75 3.25 0.5
a
NO+CO reduction at 400C; b NO decomposition at 850C; c Red. Pot. = Reductive Potential;
d
Ox. Pot = Oxidative Potential; E: Difference between oxidative and reductive potential.
EOx
EOx 2
2
Current / mA
Current / mA
Eair
0
0
ECO Air atmosphere
CO atmosphere -2
-2 ERed
ERed
0.0 0.2 0.4 0.6

0.0 0.2 0.4 0.6
Potential / V
Potential / V
EOx
2
Current / mA
0
ENO
NO atmosphere
-2
ERed
0.0 0.2 0.4 0.6
Potential / V
Figure 11. An imaged profile for the change of redox potentials in oxidative (e.g. NO) and reductive
(e.g. CO) atmosphere. Here the change of the profiles in different atmospheres were speculated based
on the Nernst Equation. Reprinted from ref. 73, copyright 2009, with permission from Elsevier
The essence of CV measurements indicates that the crucial pre-requisite for catalytic
reaction to occur is that the full redox cycle can proceed easily, that is, the electrons
transferring between reactant and catalyst is not a problem. Only when this requirement is
fulfilled, other parameters, such as the amount of active sites, could come into the focus of the
catalyst design. The apparent redox ability of catalyst would be changed depends on the
reaction environment, e.g. CO or NO atmosphere, as shown in Figure 11. This explains why
in NO+CO reaction the redox peak area plays the major role in the reaction, since the
difference in redox potentials (i.e. the difficulty in electron transfer) could be compensated
due to the presence of CO.
It was recently proposed that the redox ability of perovskite catalyst could also be
evaluated from the reduction peak area measured from H2-TPR experiments, represented as:
E = (RT/nF) ln(AF/AS).[83] At this time, the ratio of the two reduction peak areas could be
taken as a criterion of the difficulty of electrons transferring in M(n+1)+ Mn+. Results in
Figure 12 indicate that the E calculated by this way indeed has a close relation to the NO
decomposition activity, suggesting that this new method could be an indication in predicting
the catalytic performance of catalyst for NO decomposition.
100 100
80 80
E
[200 x E] / V
N2 yield / %
60 60
40 40
20 20
0 0.2 0.5 0.8 1
x in LaSrMn1-xNixO4+
Figure 12. Relationships between the difference in redox potentials (E) and NO decomposition
activity (in 2.5v/v% O2, at 850 C) of LaSrMn1-xNixO4+. Reprinted from ref. 83, copyright 2009, with
permission from Springer
4. CONCLUSION AND PERSPECTIVES

Perovskite-type oxides are one kind of promising materials in practical application not
only because of its low cost, but also due to its controllable properties. Synthesis of
perovskite-type oxides with desired properties is still an ongoing topic in materials science
and select a proper method is certainly of importance. More and more methods were explored
to synthesize perovskite with desired properties to satisfy the diverse applications. One
potential application of perovskite-type oxides is as catalyst for catalytic reaction, which has
been broadly investigated. Although at present there have not yet found application as
commercial catalysts, theses oxides offer the possibility of correlating solid-state properties
with catalytic performances, and hence getting insight into the nature of catalytic reaction. By
correlating the electrochemical properties with the catalytic performances, it could be
concluded that the difficulty of electron transferring between catalyst and reactant (i.e. the
difference in redox potentials) is the crucial pre-requisite for a catalytic reaction to occur.
Thus further studies are proposed to be concentrated on searching new method for
synthesizing perovskite oxides with suitable ability for electron denotation and reception,
especially when used as catalyst for catalytic reactions.
REFERENCES
[1] Pena MA, Fierro JLG. (2001). Chemical structures and performance of perovskite
oxides. Chem Rev. 101, 1981-2017.
[2] Catalytic Chemistry of Solid state Inorganics. New York: New York Academy of
Sciences 1976.
[3] Properties and Applications of Perovskite-type Oxides. New York: CRC Press 1993.
[4] Smyth DM. (1985). DEFECTS AND ORDER IN PEROVSKITE-RELATED
OXIDES. Annu Rev Mater Sci. 15, 329-57.
[5] Goldschmidt VM. Skr. Nor. Viedenk. -Akad., K1. I: Mater. Naturvidensk. K1. 1926.
[6] Kim JH, Kim Y, Connor PA, Irvine JTS, Bae J, Zhou WZ. (2009). Structural, thermal
and electrochemical properties of layered perovskite SmBaCo2O5+d, a potential
cathode material for intermediate-temperature solid oxide fuel cells. J Power Sources.
194, 704-11.
[7] Wang HC, Wang CL, Zhang JL, Zhao ML, Liu J, Su WB, et al. (2009). Cu Doping
Effect on Electrical Resistivity and Seebeck Coefficient of Perovskite-Type LaFeO3
Ceramics. Chinese Phys Lett. 26.
[8] Takashima H, Shimada K, Miura N, Katsumata T, Inaguma Y, Ueda K, et al. (2009).
Low-Driving-Voltage Electroluminescence in Perovskite Films. Adv Mater. 21, 3699-
+.
[9] Wang H, Zhao Z, Liang P, Xu CM, Duan AJ, Jiang GY, et al. (2008). Highly active
La1-xKxCoO3 perovskite-type complex oxide catalysts for the simultaneous removal
of diesel soot and nitrogen oxides under loose contact conditions. Catal Lett. 124, 91-9.
[10] Liu J, Zhao Z, Xu CM, Duan AJ, Jiang GY. (2008). The structures, adsorption
characteristics of La-Rb-Cu-O perovskite-like complex oxides, and their catalytic
performances for the simultaneous removal of nitrogen oxides and diesel soot. J Phys
Chem C. 112, 5930-41.
[11] He H, Dai HX, Au CT. (2001). An investigation on the utilization of perovskite-type
oxides La1-xSrxMO3 (M = Co0.77Bi0.20Pd0.03) as three-way catalysts. Appl Catal B-
environ. 33, 65-80.
[12] Zhu JJ, Mao DH, Li J, Xie XF, Yang XG, Wu Y. (2005). Recycle - new possible
mechanism of NO decomposition over perovskite(-like) oxides. J Mol Catal A-chem.
233, 29-34.
[13] Zhu JJ, Zhao Z, Xiao DH, Li J, Yang XG, Wu Y. (2005). CO oxidation, NO
decomposition, and NO plus CO reduction over perovskite-like oxides La2CuO4 and
La2-xSrxCuO4: An MS-TPD study. Ind Eng Chem Res. 44, 4227-33.
[14] Tejuca LG, Fierro JLG. Properties and Applications of Perovskite-type Oxides. New
York: CRC Press 1993.
[15] Szabo V, Bassir M, Gallot JE, Van Neste A, Kaliaguine S. (2003). Perovskite-type
oxides synthesised by reactive grinding - Part III. Kinetics of n-hexane oxidation over
LaCo(1-x)FexO3. Appl Catal B-environ. 42, 265-77.
[16] Tien-Thao N, Zahedi-Niaki MH, Alamdari H, Kaliaguine S. (2007). Effect of alkali
additives over nanocrystalline Co-Cu-based perovskites as catalysts for higher-alcohol
synthesis. J Catal. 245, 348-57.
[17] Tien-Thao N, Zahedi-Niaki MH, Alamdari H, Kaliaguine S. (2007). Conversion of

syngas to higher alcohols over nanosized LaCo0.7Cu0.3O3 perovskite precursors. Appl
Catal A-gen. 326, 152-63.
[18] Chiodelli G, Malavasi L, Tealdi C, Barison S, Battagliarin M, Doubova LM, et al.
(2009). Role of synthetic route on the transport properties of BaCe1-xYxO3 proton
conductor. J Alloy Compd. 470, 477-85.
[19] Pecchi G, Reyes P, Zamora R, Campos C, Caduus LE, Barbero BP. (2008). Effect of
the preparation method on the catalytic activity of La1-xCaxFeO3 perovskite-type
oxides. Catal Today. 133, 420-7.
[20] Delmon B. (2007). Preparation of heterogeneous catalysts. J Therm Anal Calorim. 90,
49-65.
[21] Teraoka Y, Kakebayashi H, Moriguchi I, Kagawa S. (1993). Hydroxy Acid-Aided
Synthesis of Perovskite-Type Oxides with Large Surface-Areas. J Alloy Compd. 193,
70-2.
[22] McCarty JG, Wise H. (1990). Perovskite catalysts for methane combustion. Catal
Today. 8, 231-48.
[23] Moulik SP, Paul BK. (1998). Structure, dynamics and transport properties of
microemulsions. Adv Colloid Interfac. 78, 99-195.
[24] Wallin M, Cruise N, Klement U, Palmqvist A, Skoglundh B. (2004). Preparation of
Mn, Fe and Co based perovskite catalysts using microemulsions. Colloid Surface A.
238, 27-35.
[25] Fletcher PDI, Howe AM, Robinson BH. (1987). The Kinetics of Solubilisate Exchange
between Water Droplets of a Water-in-Oil Microemulsion. J Chem Soc Farad T 1. 83,
985-1006.
[26] Campagnoli E, Tavares A, Fabbrini L, Rossetti I, Dubitsky YA, Zaopo A, et al. (2005).
Effect of preparation method on activity and stability of LaMnO3 and LaCoO3
catalysts for the flameless combustion of methane. Appl Catal B-environ. 55, 133-9.
[27] Specchia S, Civera A, Saracco G. (2004). In situ combustion synthesis of perovskite
catalysts for efficient and clean methane premixed metal burners. Chem Eng Sci. 59,
5091-8.
[28] Wang KJ, Zhong P, Zhu JJ. (2009). Preparation of Highly Active and Stable
Perovskite-like Catalyst by Combustion Method: Effect of Complex. Catal Lett. 131,
672-5.
[29] Civera A, Pavese M, Saracco G, Specchia V. (2003). Combustion synthesis of
perovskite-type catalysts for natural gas combustion. Catal Today. 83, 199-211.
[30] Gonzalez A, Tamayo EM, Porter AB, Corberan VC. (1997). Synthesis of high surface
area perovskite catalysts by non-conventional routes. Catal Today. 33, 361-9.
[31] Tian BZ, Liu XY, Yang HF, Xie SH, Yu CZ, Tu B, et al. (2003). General synthesis of
ordered crystallized metal oxide nanoarrays replicated by microwave-digested
mesoporous silica. Adv Mater. 15, 1370-+.
[32] Jiao F, Bruce PG. (2007). Mesoporous crystalline beta-MnO2- a reversible positive
electrode for rechargeable lithium batteries. Adv Mater. 19, 657-+.
[33] Wang YG, Ren JW, Wang YQ, Zhang FY, Liu XH, Guo Y, et al. (2008). Nanocasted
synthesis of mesoporous LaCoO3 perovskite with extremely high surface area and
excellent activity in methane combustion. J Phys Chem C. 112, 15293-8.
[34] Cao GZ, Liu DW. (2008). Template-based synthesis of nanorod, nanowire, and
nanotube arrays. Adv Colloid Interfac. 136, 45-64.
[35] Gao F, Hong RZ, Liu JJ, Yao YH, Tian CS. (2008). Effect of different templates on
microstructure of textured Na0.5Bi0.5TiO3-BaTiO3 ceramics with RTGG method. J
Eur Ceram Soc. 28, 2063-70.
[36] Wrighton MS, Ellis AB, Wolczanski PT, Morse DL, Abrahamson HB, Ginley DS.
(1976). Strontium-Titanate Photoelectrodes - Efficient Photoassisted Electrolysis of
Water at Zero Applied Potential. J Am Chem Soc. 98, 2774-9.
[37] Domen K, Naito S, Onishi T, Tamaru K. (1982). Study of the Photocatalytic
Decomposition of Water-Vapor over a Nio-Srtio3 Catalyst. J Phys Chem-Us. 86, 3657-
61.
[38] Mitsui C, Nishiguchi H, Fukamachi K, Ishihara T, Takita Y. (1999). Photocatalytic
decomposition of pure water over NiO supported on KTa(M)O-3 (M=Ti4+,Hf4+,Zr4+)
perovskite oxide. Chem Lett. 1327-8.
[39] Hu CC, Tsai CC, Teng H. (2009). Structure Characterization and Tuning of Perovskite-
Like NaTaO3 for Applications in Photoluminescence and Photocatalysis. J Am Ceram
Soc. 92, 460-6.
[40] Vijayakumar KM, Lichtin NN. (1984). Reduction of Co2 by H2 and Water-Vapor over
Metal-Oxides Assisted by Visible-Light. J Catal. 90, 173-7.
[41] Lichtin NN, Vijayakumar KM, Rubio BI. (1987). Photoassisted Reduction of Co2 by
H-2 over Metal-Oxides in the Absence and Presence of Water-Vapor. J Catal. 104,
246-51.
[42] Lichtin NN, Vijayakumar KM. (1983). Photocatalyzed Light-Driven Reduction of Co2
and N2 on Semiconductor Surfaces. J Electrochem Soc. 130, C108-C.
[43] Miyama H, Fujii N, Nagae Y. (1980). Heterogeneous Photocatalytic Synthesis of
Ammonia from Water and Nitrogen. Chem Phys Lett. 74, 523-4.
[44] Li Q, Domen K, Naito S, Onishi T, Tamaru K. (1983). Photocatalytic Synthesis and
Photo-Decomposition of Ammonia over Srtio3 and Batio3 Based Catalysts. Chem Lett.
321-4.
[45] Huang YH, Liang G, Croft M, Lehtimaki M, Karppinen M, Goodenought JB. (2009).
Double-Perovskite Anode Materials Sr2MMoO6 (M = Co, Ni) for Solid Oxide Fuel
Cells. Chem Mater. 21, 2319-26.
[46] Huang K, Lee HY, Goodenough JB. (1998). Sr- and Ni-doped LaCaO3 and LaFeO3
perovskites - New cathode materials for solid-oxide fuel cells. J Electrochem Soc. 145,
3220-7.
[47] Weston M, Metcalfe IS. (1998). La0.6Sr0.4Co0.2Fe0.8O3 as an anode for direct
methane activation in SOFCS. Solid State Ionics. 115, 247-51.
[48] Sfeir J, van Herle J, McEvoy AJ. (1999). Stability of calcium substituted lanthanum
chromites used as SOFC anodes for methane oxidation. J Eur Ceram Soc. 19, 897-902.
[49] Kuo JH, Anderson HU, Sparlin DM. (1990). Oxidation Reduction Behavior of
Undoped and Sr-Doped Lamno3 - Defect Structure, Electrical-Conductivity, and
Thermoelectric-Power. J Solid State Chem. 87, 55-63.
[50] Zhou W, Ran R, Shao ZP. (2009). Progress in understanding and development of
Ba0.5Sr0.5Co0.8Fe0.2O3-delta-based cathodes for intermediate-temperature solid-
oxide fuel cells: A review. J Power Sources. 192, 231-46.
[51] Shao ZP, Haile SM. (2004). A high-performance cathode for the next generation of
solid-oxide fuel cells. Nature. 431, 170-3.
[52] Shao ZP, Haile SM, Ahn J, Ronney PD, Zhan ZL, Barnett SA. (2005). A thermally
self-sustained micro solid-oxide fuel-cell stack with high power density. Nature. 435,
795-8.
[53] Simonsen VLE, Norskov L, Hagen A, Hansen KK. (2009). Electrochemical reduction
of NO on La2-xSrxNiO4 based electrodes. J Solid State Electr. 13, 1529-34.
[54] Liu CB, Zhao Z, Yang XG, Ye XK, Wu Y. (1996). Superconductor mixed oxides La2-
xSrxCuO4+/-lambda for catalytic hydroxylation of phenol in the liquid-solid phase.
Chem Commun. 1019-20.
[55] Resini C, Catania F, Berardinelli S, Paladino O, Busca G. (2008). Catalytic wet
oxidation of phenol over lanthanum strontium manganite. Appl Catal B-environ. 84,
678-83.
[56] Kulkarni AS, Jayaram RV. (2003). Liquid phase catalytic transfer hydrogenation of
aromatic nitro compounds on perovskites prepared by microwave irradiation. Appl
Catal A-gen. 252, 225-30.
[57] Kulkarni AS, Jayaram RV. (2004). Liquid phase catalytic transfer hydrogenation of
aromatic nitro compounds on La1-xSrxFeO3 perovskites prepared by microwave
irradiation. J Mol Catal A-chem. 223, 107-10.
[58] Parravano G. (1953). Catalytic Activity of Lanthanum and Strontium Manganite. J Am
Chem Soc. 75, 1497-8.
[59] Tejuca LG, Fierro JLG, Tascon JMD. (1989). Structure and Reactivity of Perovskite-
Type Oxides. Adv Catal. 36, 237-328.
[60] Voorhoeve RJH, Remeika JP, Trimble LE. (1976). Defect Chemistry and Catalysis in
Oxidation and Reduction over Perovskite-Type Oxides. Ann Ny Acad Sci. 272, 3-21.
[61] Sokolovskii VD. (1990). Principles of Oxidative Catalysis on Solid Oxides. Catal Rev.
32, 1-49.
[62] Anshiz AG, Sokolovskii VD, Boreskov GK, Boronin AI. (1977). Participation of
Molecular-Oxygen in Formation of Complete Oxidation-Products Via a Concerted
Mechanism. React Kinet Catal L. 7, 87-92.
[63] Falcon H, Martinez-Lope MJ, Alonso JA, Fierro JLG. (2000). Defect LaCuO3-delta
(delta=0.05-0.45) perovskites - Bulk and surface structures and their relevance in CO
oxidation. Appl Catal B-environ. 26, 131-42.
[64] Dai HX, He H, Li W, Gao ZZ, Au CT. (2001). Perovskite-type oxide
ACo(0.8)Bi(0.2)O(2.87) (A = La0.8Ba0.2): a catalyst for low-temperature CO
oxidation. Catal Lett. 73, 149-56.
[65] Zhu JJ, Yang XG, Xu XL, Wei KM. (2006). Cyclic voltammetry for predicting
oxidation process in heterogeneous catalysis. Z Phys Chem. 220, 1589-94.
[66] Zhu JJ, Yang XG, Xu XL, Wei KM. (2006). Study of CH4 and CO oxidation from
electrochemical method. Chinese Sci Bull. 51, 374-6.
[67] Tilset BG, Fjellvag H, Kjekshus A, Slagtern A, Dhal I. (1996). Properties of LaCo1-
tCrtO3 .3. Catalytic activity for CO oxidation. Appl Catal A-gen. 147, 189-205.
[68] Zhang-Steenwinkel Y, van der Zande LM, Castricum HL, Bliek A. (2004). Step
response and transient isotopic labelling studies into the mechanism of CO oxidation
over La0.8Ce0.2MnO3 perovskite. Appl Catal B-environ. 54, 93-103.
[69] Kubaschewski O, Alcock CB, Spencer PJ. Materials Thermochemistry. 6th Edition ed.
Oxford: Pergamon Press 1993.
[70] Zhao Z, Yang XG, Wu Y. (1996). Comparative study of Nickel-based perovskite-like
mixed oxide catalysts for direct decomposition of NO. Appl Catal B-environ. 8, 281-
97.
[71] Shin S, Hatakeyama Y, Ogawa K, Shimomura KY. (1979). Catalytic Decomposition of
No over Brownmillerite-Like Compounds, Ca2fe2o5 and Sr2fe2o5. Mater Res Bull.
14, 133-6.
[72] Zhu YJ, Wang D, Yuan FL, Zhang G, Fu HG. (2008). Direct NO decomposition over
La2-xBaxNiO4 catalysts containing BaCO3 phase. Appl Catal B-environ. 82, 255-63.
[73] Zhu J, Thomas A. (2009). Perovskite-type mixed oxides as catalytic material for NO
removal. Appl Catal B-environ. 92, 225-33.
[74] Ishihara T, Shinmyo Y, Goto K, Nishiyama N, Iwakuni H, Matsumoto H. (2008). NO
decomposition on Ruddlesden-Popper-type oxide, Sr3Fe2O7, doped with Ba and Zr.
Chem Lett. 37, 318-9.
[75] Iwakuni H, Shinmyou Y, Yano H, Matsumoto H, Ishihara T. (2007). Direct
decomposition of NO into N-2 and O-2 on BaMnO3-based perovskite oxides. Appl
Catal B-environ. 74, 299-306.
[76] Ishihara T, Ando M, Sada K, Takiishi K, Yamada K, Nishiguchi H, et al. (2003). Direct
decomposition of NO into N-2 and N-2 over La(Ba)Mn(In)O-3 perovskite oxide. J
Catal. 220, 104-14.
[77] Zhu JJ, Xiao DH, Li J, Yang XG, Wu Y. (2005). Effect of Ce on NO direct
decomposition in the absence/presence of O-2 over La1-xCexSrNiO4 (0 <= x <= 0.3).
J Mol Catal A-chem. 234, 99-105.
[78] Shin S, Arakawa H, Hatakeyama Y, Ogawa K, Shimomura K. (1979). Absorption of
No in the Lattice of an Oxygen-Deficient Perovskite Srfeo3-X and the Infrared
Spectroscopic Study of the System No - Srfeo3-X. Mater Res Bull. 14, 633-9.
[79] Zhu JJ, Yang XG, Xu XL, Wei KM. (2007). Active site structure of NO decomposition
on perovskite(-like) oxides: An investigation from experiment and density functional
theory. J Phys Chem C. 111, 1487-90.
[80] Zhu JJ, Zhao Z, Xiao D, Li J, Yang XG, Wu Y. (2005). Application of cyclic
voltammetry in heterogeneous catalysis: NO decomposition and reduction.
Electrochem Commun. 7, 58-61.
[81] Zhu JJ, Zhao Z, Xiao DH, Li J, Yang XG, Wu Y. (2005). Study of La2-xSrxCuO4
(x=0.0, 0.5, 1.0) catalysts for NO plus CO reaction from the measurements of O-2-
TPD, H-2-TPR and cyclic voltammetry. J Mol Catal A-chem. 238, 35-40.
[82] Zhu JJ, Zhao Z, Xiao DH, Li J, Yang XG, Wu Y. (2005). CO oxidation over the
perovskite-like oxides La2-xSrxMO4 (x=0.0, 0.5, 1.0; M = Cu, Ni): A study from
cyclic voltammetry. Z Phys Chem. 219, 807-15.
[83] Zhu JJ, Xiao DH, Li J, Yang XG. (2009). Perovskite-Like Mixed Oxides (LaSrMn1-x
Ni (x) O4+delta, 0 a parts per thousand currency sign x a parts per thousand currency
sign 1) as Catalyst for Catalytic NO Decomposition: TPD and TPR Studies. Catal Lett.
129, 240-6.
Chapter 8
MECHANISM OF FORMATION OF
PEROVSKITE-TYPE LAYERED OXIDES
Irina A. Zvereva* and Alexander B. Missyul

Saint-Petersburg State University, Department of Chemistry,
Universitetsky pr., 26, Petrodvoretz, Saint-Petersburg, 198504, Russia
ABSTRACT
The chapter presents general analysis and results of experimental data on the
structure-chemical investigation of process of formation of perovskite type structures in
different groups of layered oxides, which exhibit a number of practically important
properties. In many cases the layered perovskite-like phases are formed in a multistage
way. The composition and stability of the intermediates can strongly affect both reaction
rate and the structure of the final compound. Examples of such impact for several
Ruddlesden-Popper and Aurivillius phases are considered. Formation of the double-
perovskite layered phases for the series of high temperature aluminates Ln2SrAl2O7
(Ln=La, Nd, Sm, Eu, Gd, Ho) is discussed in terms of the dependence of mechanism of
reaction on the nature of rare earth elements. For the double-layered manganites
LnSr2Mn2O7 (Ln=La, Nd, Gd) the structure and stability of the intermediate perovskite
phase affects both the reaction speed and the microstructure of the resulting Ruddlesden-
Popper phase, including phase separation in the CMR manganites. The double-layered
ferrites Ln2SrFe2O7 (Ln=La, Nd, Gd) exhibit the evolution of oxidation state of iron
along the process of synthesis. Mechanism of formation of the ferroelectric Aurivillius
phase Bi3TiNbO9 demonstrates the limitations on the BiLn and TiMn substitution in
this structure. The stability of the intermediate phase results in the substitution
limitations. Investigations of the processes occurring in the reaction mixture during
synthesis of the perspective ionic conductors and catalyst (Na,Nd)n+1TinO3n+1 allow to
explain the necessity of the excess of alkali metal containing starting compound, optimize
the conditions of synthesis and significantly decrease the size of particles.
*
346 Irina A. Zvereva and Alexander B. Missyul
1. INTRODUCTION
Perovskite type layered structures and especially layered oxides are intensively studied
because they exhibit a wealth of "modern" physical phenomena, such as high-Tc
superconductivity, colossal magnetoresistance, ferroelectricity and quantum critical behavior.
Due to this layered oxides create the basis for many technological applications including
information storage and transmission, micro-electronics and micro-manipulation.
And moreover layered structures attract particular interest, not only from the point of
view of promising properties but also due to broad fundamental aspects of crystal chemistry
and thermodynamics. Layered compounds represent the intergrowth of blocks of different
simple structure-types. The phase stability in such systems depends on extent of constrains
between different types of structure and determined by equilibrium in multicomponent
(different cations) microheterogeneous (different layers) system.
The best examples of layered compounds are the intergrowth of layers of more
outstanding and well investigated type of structure perovskite and slabs of the different
type of structure [1,2]. It could be slabs with rock salt or calcium fluorite type or slabs with
cations of metals. Some phenomena play important role in structure of layered compounds.
The first one is isomorphism of non-isovalent atoms (usually lanthanide Ln3+ and alkali-earth
A2+ or alkaline A1+ atoms) in matrix of layered structure [3,4] resulting in the distribution of
cations of different charge over one (two) equivalent sites. The variations of cationic
substitution (isovalent and non-isovalent) in this type of compounds ensure such outstanding
properties as superconductivity, colossal magnetoresistance and catalytical ability. The
second one is the nanoscaled thickness of perovskite layers and consequently specific 2D
character of electrical, magnetic and optical properties.
The influence of particularities of structure on physical and chemical properties is
considered in many respects, but the mechanism of the process of formation of such
compounds was hardly investigated. Moreover the knowledge of formation could be useful
for the explanation of the stability of compounds and for the determination the optimal
methods of synthesis.
In this chapter we will consider the structure-chemical mechanism of formation of several
groups of layered oxides being among most important for applications: aluminates
Ln2SrAl2O7 (Ln=La Ho) with high termal stability, CMR manganites LnSr2Mn2O7 (Ln=La,
Nd, Gd), ferrites Ln2SrFe2O7 (Ln=La, Nd, Gd), ferroelectric titanoniobates Bi3TiNbO9 and
perspective catalysts titanates NaNdTiO4 and Na2Nd2Ti3O10.
We investigate in details the most typical way for synthesis of the intergrowth structures
(solid state reaction starting from oxides and carbonates) by quenching-cooling method and
X-ray phase analysis. The time-temperature dependence of the composition of reaction
mixtures was used for interpretation of mechanism and kinetics of reactions.
Mechanism of Formation of Perovskite-Type Layered Oxides 347
Figure 1. Layered perovskite-like oxides (n = 3).
2. PEROVSKITE-TYPE LAYERED PHASES

Perovskite-type layered compounds are the intergrowth of perovskite layers (P) 3
and slabs of the differ type of structure (rock salt, calcium fluorite type, cations of metals).
Depending on the nature of slabs between perovskite blocks, layered compounds belong to
three big groups: Ruddlesden-Popper phases, Aurivillius phases, Dion-Jacobson phases.
Ruddlesden-Popper phases are the most well known and intensively investigated
perovskite-type layered structures. In 1957 R.S. Ruddlesden and P. Popper synthesized
complex oxide Sr2TiO4 [5] with K2NiF4-type of structure and then Sr3Ti2O7 and Sr4Ti3O10 [6]
which gave the name of new structure types. These compounds formed new class of complex
oxides Srn+1TinO3n+1 and appearance of the term Ruddlesden-Popper phases. Particularity of
Ruddlesden-Popper phases is the intergrowth of perovskite layers (P) and slabs of rock salt
(RS) with consequence of layers RS (P--P) RS (P--P) RS. Schematic
representation of these phases is given in Figure 1a.
Great interest to compounds with 2D character of physical properties in 70-th [7-14] and
discovering of high temperature superconductivity in 1986 for K2NiF4-type oxides [15] gave
rise to the strong growth of investigation of perovskite-type layered oxides containing rare-
earth elements LnAMO4 (Ln lanthanoid or Y; A Sr,Ca; M Al or 3d-cations) [16-26]. At
that time compounds Ln2SrAl2O7 (Ln La, Nd, Gd) with Sr3Ti2O7-type of structure [27,28]
have been obtained. Later the series was prolongated till Ho2SrAl2O7 [3,29]. In the end of 90-
th the search for materials with ionic conductivity and ion-exchange properties resulted in
synthesis of new titanates LnATiO4 [30-32] and Ln2A2Ti3O10 [33-39], Na2Ca2Ta3O10 [40],
even though LaNaTiO4 have been obtained by G. Blasse as early as in 1965 [41]. Discovering
the phenomena of colossal magnetoresistance moved the interest to oxides Sr2LnMn2O7 [42]
and Sr3Fe2O7 [43] and solid solutions with partial cationic substitution [44-48]. Recently the
search for new high temperature superconductors among bismuthates and plumbates results in
synthesis of Ban+1(Pb1-xBix)nO3n+1 and Ba3Pb2O7, (Ba, K)2BiO4 and (Ba, K)3Bi2O7 [49].
Thus, today Ruddlesden-Popper phases include complex oxides containing metals from
groups 1, 2, 13, 14 and 15 as well as transition (d- and 4f-) elements. The general formula of
such oxides can be written as AO(AO3)n, where A means alkaline, alkali earth or rare earth
element, while B is d-element, Al, Ga, In, Pb or Bi. This formula shows the main structural
feature of this class of the layered perovskite-like compounds, that is the intergrowth of the
perovskite blocks AMO3 () and rock salt blocks (RS) in a consequence -Pn-RS-Pn-RS-.
In case of n = the perovskite structure itself is obtained. Figure 1 shows schematically the
structure of the Ruddlesden-Popper phases (n=3) as compared to the other types of the
layered perovskite-like compounds.
In perovskite type layered structure the consequence of perovskite layers and other blocks
is not usual geometrical one, it is really the intergrowth of different structures. Stability of
intergrowth depends on the extent of geometrical constrains between different types of
structure, and thermodynamic equilibrium between different layers.
The Dion-Jacobson phases are chemically close to the Ruddlesden-Popper phases. They
have been synthesized for the first time in 1980 by M. Dion [50], but it was not until A.
Jacobsons studies [51,52] that they attracted significant attention for high reactivity of these
compounds. The general formula of the Dion-Jacobson phases is An-1BnO3n+1, where is
usually alkaline ion, is alkali earth or rare earth ion, and B is group 5 ion or group 4 ion, for
example KCa2Nb3O10 or MLa2Ti2TaO10. In these compounds the perovskite layers are
divided by the layers of the alkaline or alkali earth ions (Figure 1b). The Dion-Jacobson
phases show high catalytic activity as well as high ionic conductivity [39,53-54], and that is
why they are very popular investigation objects in the last years.
The Aurivillius phases An-1Bi2nO3n+3 have been discovered by Aurivillius in 1949 [55].
In 1961 G. A. Smolenski, V. A. Isupov and A. I. Agranovskaya [56] discovered
ferroelectricity in this class of oxides. Today the Aurivillius series includes a large number of
oxides An-1Bi2nO3n+3, where A = Bi3+, Ba2+, Sr2+, Ca2+, Pb2+, K+, Na+, Ln3+ etc., = Ti4+,
Fe3+, Nb5+, Ta5+, Mo6+, W6+ [57-60]. The perovskite layers in the structure of the Aurivillius
phases are interleaved by the fluorite-type layers Bi2O2 (Figure 1c). That is why in the last
time their composition is usually written as (Bi2O2)An-1BnO3n+1.
3. FORMATION OF THE DOUBLE-PEROVSKITE LAYERED ALUMINATES

LN2SRAL2O7 (LN=LA HO)
The mechanism of formation of the double-perovskite layered aluminates was
investigated for the series of Ln2SrAl2O7 (Ln=La, Nd, Sm, Eu, Gd, Ho) [61-65,29] in the
temperature range of 1000-1500C for the reaction of the binary oxides and strontium
carbonate:
Ln2O3 + Al2O3 + 2 SrCO3 = Ln2SrAl2O7 (1)
In case of La, Nd, Sm-containing oxides, the perovskite phases LnAlO3 and 1/1
intergrowth structures LnSrAlO4 were obtained as intermediate phases [61,62]. Their
interaction results in forming of the 2/1 intergrowth of oxides Ln2SrAl2O7 (Ln=La,Nd,Sm).
Thus intermediate reactions of overall process of Ln2SrAl2O7 are the following:
Ln2O3 + Al2O3 = 2 LnAlO3 (2)
Ln2O3 + Al2O3 + 2 SrCO3 = 2 LnSrAlO4+ 2 CO2 (3)
LnAlO3 + LnSrAlO4 = Ln2SrAl2O7 (4)
Figure 2. The structure transformation during the last stage of Ln2SrAl2O7 formation.
To support this mechanism the consequence of reaction (2)-(4) have been performed. We
can say that this is the usual way of formation of the Ruddlesden-Popper phases through the
more simple perovskite-types structure if they stable at temperature of synthesis. This
mechanism works for the beginning of lanthanide raw LaSm. The structure-chemical
reaction of the last stage of Ln2SrAl2O7 formation which determines the rate of overall
process is presented in Figure 2.
In the second part of the lanthanides raw (starting from gadolinium) the perovskite type
structure is not stable at temperature of synthesis. Due to the instability of GdAlO3 oxide
SrAl2O4 with spinel type of structure occurs as the intermediate phase. Till the end of reaction
the gadolinium oxide is also found in the reaction mixture. Interaction between these two
compounds is the limiting reaction for the overall process of 2/1 intergrowth Gd2SrAl2O7
formation [63]:
Al2O3 + SrCO3 = SrAl2O4 + CO2 (5)
SrAl2O4+ Gd2O3 = Gd2SrAl2O7 (6)
Mechanism (5)-(6) works for the second part of lanthanides Gd-Ho. Holmium brings
some novel phenomena. Its smallest atoms form the phase SrHo2O4 with tunnel structure of
the CaFe2O4 type at temperature higher than 1300 and simultaneously with reactions (5)-
(6) intermediate reactions (7)-(8) take place [29].
Ho2O3 + SrCO3 = SrHo2O4 + CO2 (7)
SrHo2O4+ Al2O3 = Ho2SrAl2O7 (8)
Of the special interest is the formation of the Eu-containing compound, Eu2SrAl2O7,

located between two groups with different formation mechanisms. The mechanism of
formation of the Eu-containing phase depends on the temperature. Below 1400 the
intermediate reaction is
SrAl2O4 + Eu2O3 Eu2SrAl2O7 (9)
while above 1400 the reaction goes through the stage
EuAlO3 + EuSrAlO4 Eu2SrAl2O7 (10)
This means that Eu2SrAl2O7 formation goes through the P/RS (EuSrAlO4) and P
(EuAlO3) phases only if the last one is stable at the synthesis temperature, while at lower
temperatures the mechanism of the reaction changes from the one characteristic for the first
members of the lanthanide series (reactions 2-4) to the one typical for the last members
(reactions 5-6).
Figure 3. Temperatures and mechanisms of formation of Ln2SrAl2O7.
Figure 3 demonstrates the difference in mechanism of formation in the raw La Ho.

Diagram also indicates the temperature of the beginning of formation of Ln2SrAl2O7 and
temperature of obtaining of single phase without impurities of intermediate compounds. It is
important to mention that difference in character of Ln2SrAl2O7 melting follows the
difference in the mechanism of formation.
In the rare-earth series La-Gd there is a large increase of the formation rate of the
intergrowth oxides. In Figure 4(a-d) it is possible to see that after three hours heating at 1500
C, in La-containing system there is a mixture of perovskite, 1/1 and 2/1 intergrowth phases.
Under the same conditions, in Nd and Gd-containing systems there is the only 2/1 intergrowth
oxide.
After Gd the mechanism of formation of the Ln2SrAl2O7 phases remains the same,
SrAl2O4 being the intermediate compound, but the formation rate decreases significantly.
SrAl2O4 is formed at 1000 in all cases, but interaction of SrAl2O4 with Gd2O3 starts already
at 1100, while Ho2O3 reacts only after 1300. At 1500 Gd2SrAl2O7 formation is
finished already after 1 hour heating, while pure Ho2SrAl2O7 can be obtained only after 24
hours (Figure 4e).
The effect of the cationic substitution on the formation of the Ruddlesden-Popper
aluminates has been investigated for the (LaxHo1-x)2SrAl2O7 and (NdxHo1-x)2SrAl2O7 solid
solutions. Figure 5 shows change of the formation rate of (La0.1Ho0.9)2SrAl2O7 and
(Nd0.15Ho0.85)2SrAl2O7 compared to Ho2SrAl2O7. It is clear that even small amounts of La or
Nd strongly affect the reaction rate. The presence of Nd results in the increase of the
formation rate, while the La-containing phase is formed at smaller rate. This difference
corresponds to the difference in the reactivity in the systems Ln2O3SrCO3Al2O3 (Ln=La,
Nd, Ho).
Figure 4. Degree of formation of Ln2SrAl2O7 vs. time dependence for different temperatures: a
Ln=La; - 1200C; - 1300C; - 1400C; - 1500C; - 1530C; b Ln=Nd; - 1000C;
- 1100C; - 1200C; - 1300C; - 1400C; - 1500C; c Ln=Sm; - 900C; - 1000C;
- 1100C; - 1200C; - 1300C; - 1400C; d Ln=Gd; - 1200C; - 1300C; - 1400C; e
Ln=Ho; - 1200C; - 1300C; - 1400C.
- (La0.1Ho0.9)2SrAl2O7; - Ho2SrAl2O7; - (Nd0.15Ho0.85)2SrAl2O7.
Figure 5. Degree of formation of Ho2SrAl2O7-based solid solutions vs. time dependence (1400C).
Figure 6. Comparison of the structures of 4H-SrMnO3 (a) and Sr7Mn4O15 (b).

4. FORMATION OF THE DOUBLE-PEROVSKITE LAYERED

MANGANITES LN2SRMN2O7 (LN=LA, ND, GD).
The formation of the NdSr2Mn2O7 from the Nd2O3, SrCO3 and MnO2 has been
investigated under isothermal conditions at 1200C and under constant heating rate conditions
by the DSC-TGA method.
The qualitative phase analysis of the reaction mixture after 3 hrs heating at 1200C shows
that the following phases present: Nd2O3, Sr7Mn4O15, NdSr1-MnO3 and Nd1+Sr2-Mn2O7.
Using the analogy with the previously found mechanism of formation of the Ruddlesden-
Popper phases Ln2SrAl2O7 and Ln2SrFe2O7 [61-67] the following mechanism had been
proposed:
7SrCO3 + 4MnO2 Sr7Mn4O15 + 7CO2 (11)
3Nd2O3 + 2SrCO3 + 8MnO2 8Nd0.75Sr0.25MnO3 + 2CO2 + 1.5O2 (12)
Nd2O3 + 2Sr7Mn4O15 8Nd0.25Sr1.75MnO4 + 0.5O2 (13)
Nd0.75Sr0.25MnO3 + Nd0.25Sr1.75MnO4 NdSr2Mn2O7. (14)
Absence of the peaks of the Nd0.25Sr1.75MnO4 phase on the powder pattern means that its
formation is the slowest stage in the reaction while the last stage is much faster.
The possibility of the proposed reaction sequence (11-14) has been confirmed by carrying
out each of them independently. It should be noted that the intermediate phases are formed at
higher temperatures when synthesized apart. This is analogous to the case of the formation of
the Ln2SrAl2O7 phases where the temperature of the separate formation of the intermediate
perovskite phase LnAlO3 is higher than the temperature of its formation as an intermediate
compound. [61]
The fact that Sr7Mn4O15 appears in the reaction mixture instead of the K2NiF4-type phase
can be understood by analysis of the structure of SrMnO3. While possessing the same formula
as perovskites, this compound has different structure with face-sharing MnO6 octahedra
(Figure 6a). This structural feature remains unchanged when the layered structure is formed,
leading to the different chemical composition (Figure 6b). It is interesting to note that 4H-
SrMnO3 transforms to the orthorhombic -SrMnO3 [68,69] at approximately the same
temperature (about 1400), as the temperature of stabilization of Sr2MnO4 compared to
Sr7Mn4O15 [70]. This gives the additional confirmation to the interconnection of between the
structure of SrMnO3 and the layered phase.
The quantitative analysis of the reaction mixture shows that the proposed mechanism
does not account for the reaction completely. The molar ratio (Sr7Mn4O15) : (Nd2O3)
appears to be 0.77:1 instead of 2:1, as required by the proposed mechanism. In order to
explain this fact it was proposed that the perovskite phase contains less Nd, while the excess
of the Nd2O3 reacts with the perovskite in the same way as in case of the formation of the
Gd2SrFe2O7 [67]. This means that the perovskite phase reacts with Nd2O3 forming the
K2NiF4-type phase and this K2NiF4-type phase reacts with the perovskite phase again yielding
the double-layered Ruddlesden-Popper phase. Again, the slowest stage of the reaction is the
formation of the K2NiF4-type phase.
The processes that take place in the system during heating were investigated by the DSC-
TGA method (Figure 7). Comparison with the literature data for the pure initial compounds
and previously published results concerning the formation of the LnSr2Al2O7 series [61-62] as
well as X-ray phase analysis gives rise to the following explanation of the observed thermal
effects. At 360C small amounts of water (about 0.5%) absorbed by Nd2O3 are released. At
650C MnO2 decomposes yielding Mn2O3. The broad peak at 950-1030C appears due to the
series of processes: Mn2O3 decomposes yielding Mn3O4, SrCO3 undergoes the phase
transition and the formation of the intermediate products starts. These processes result in a
large decrease in mass (14% approximately). At higher temperature the mass of the reaction
mixture increases again because the formation of the intermediate compounds from Mn3O4
implies absorption of oxygen from air. The small peak at 1190C corresponds to the phase
transition in Mn3O4.
The complete mechanism of the reaction can be written as follows:
Common stages
2MnO2 Mn2O3 + 0.5O2 (15)

3Mn2O3 2Mn3O4 + 0.5O2 (16)
21SrCO3 + 4Mn3O4 + 4O2 3Sr7Mn4O15 + 21CO2 (17)
3xNd2O3 + 6(1-x)SrCO3 + 2Mn3O4 + (4-3x)O2 6NdxSr1-xMnO3 + 6(1-x)CO2 (18)
Mechanism (I)
Nd2O3 + 2Sr7Mn4O15 8Nd0.25Sr1.75MnO4 + 0.5O2 (19)

NdSr1-MnO3 + Nd0.25Sr1.75MnO4 Nd0.25+xSr2.75-xMn2O7 (20)
Mechanism (II)
2NdSr1-MnO3 + Nd2O3 2Nd1+xSr1-xMnO4 + 0.5O2 (21)

NdSr1-MnO3 + Nd1+xSr1-xMnO4 2Nd1+2xSr2-2xMn2O7 (22)
In both reaction pathways the slowest stage is the formation of the K2NiF4-type phase
(19,21). It should be noted that the cationic composition of the double-layered Ruddlesden-
Popper phases obtained by these two pathways is not the same. This means that the proposed
mechanism gives explanation to the formation of two phases observed by P. D. Battle et al.
[42,43]
The stoichiometry of the reaction allows to find the perovskite phase composition and the
proportions of the product formed by each of the mechanisms simultaneously with the kinetic
calculations. The following relationship between the x value in NdxSr1-xMnO3 and the molar
fraction of the Ruddlesden-Popper phase obtained by the mechanism II (y) can be derived:
(1 - y)(x + 0.25) + y(1 + 2x)=1, (23)

100
20
10 90
Heat flow, mW
0
Weight, %
80
-10
70
-20
60
-30
-40 50
200 400 600 800 1000 1200
Temp, C
Figure 7. The DSC-TGA analysis of the reaction mixture during synthesis of NdSr2Mn2O7.
meaning that the Nd0.25+xSr2.75-xMn2O7 and Nd1+2xSr2-2xMn2O7 phases are formed at the molar
ratio (2):(0.75 - ).
Using the quantitative X-ray diffraction analysis the kinetics of the reaction at 1200C
has been investigated. As long as the amount of the perovskite phase can not be determined
accurately from the X-ray diffraction data due to the peak overlapping, it was assumed that
the conversion degree according to both mechanisms is almost the same after 3 hours heating.
The error of such assumption is of the same order as the error of the X-ray quantitative
analysis.
mechanism I; - mechanism II.
Figure 8. The reaction rate of two mechanisms of the NdSr2Mn2O7 formation at 1200C.
mechanism I; mechanism II; total amount.
Figure 9. Amount of Nd1-xSr2+xMn2O7 formed by each of the reaction mechanisms and total amount of
the Ruddlesden-Popper phases at 1200C.
The composition of the intermediate perovskite phase appears to be Nd0.42Sr0.58MnO3.

Approximately 72% of the double-layered phase is formed according to the mechanism (I).
The comparison of the reaction rate for each mechanism is shown on Figure 8. The amounts
of the Ruddlesden-Popper phases obtained for each mechanism are shown on Figure 9.
In case of LaSr2Mn2O7, the intermediate compounds observed in the reaction mixture
after 3 hours heating at 1200 are La2O3, LaSr1-MnO3, LaSr2Mn2O7, La1-xSr1+xMnO4 and
Sr7Mn4O15. However, after 5 hours Sr7Mn4O15 cannot be observed in the reaction mixture,
probably due to high rate of the interaction with La2O3. Mechanism of formation of
LaSr2Mn2O7 is the same as for NdSr2Mn2O7 but the comparative rates of the reaction steps
are different: the K2NiF4-type compound forms and reacts with approximately the same rate.
This holds for the Gd-containing compound too. The overall formation rate of LaSr2Mn2O7
and GdSr2Mn2O7 at 1200 is shown at Figure 10.
The DSC-TGA data of the reaction mixture during the formation of LaSr2Mn2O7 and
GdSr2Mn2O7 are qualitatively the same as for NdSr2Mn2O7. This supports the conclusion that
the formation process goes the same way or all lanthanides investigated.
Figure 10. Degree of formation vs. time dependence for LaSr2Mn2O7 () and GdSr2Mn2O7 () at
1200C.
Figure 11. The SEM microphotograph of the reaction mixture during the formation of GdSr2Mn2O7.
The reaction mixture obtained during the formation of GdSr2Mn2O7 has been investigated
by SEM. The interacting particles have the size of 1-5m (Figure 11). The microprobe
analysis of the sample has shown that the intermediate perovskite phase is Gd0.5Sr0.5MnO3,
the composition being close to the one obtained for NdSr2Mn2O7 by means of the X-ray
diffraction data.
Investigation of the formation of manganites LnSr2Mn2O7 (Ln=La, Nd, Gd) allowed to
understand the new aspects in nature of the phase separation in the double-layered
Ruddlesden-Popper phases, namely that cationic composition in these compounds depends on
pathways of reaction. The difference in the cationic composition of two phases obtained by
the one-step synthesis is due to the difference in the mechanisms of formation.
4. FORMATION OF THE DOUBLE-PEROVSKITE LAYERED LN2SRFE2O7

(LN=LA, ND, GD)
Ferrite ceramics is used in memory chips of computer equipment, as catalysts in chemical
industry, and also in power engineering and electronics. To develop effective technology of
ferrite products with the required magnetic and other characteristics, it is necessary to provide
not only the preset crystal-chemical characteristics (structure and distribution of cations over
various structural positions), but also stability of the valence state of iron atoms.
Figure 12. The DTA-TGA analysis of the reaction mixture during synthesis of La2SrFe2O7.
Polycrystalline samples of Ln2SrFe2O7 (Ln = La, Nd, Sm, Gd) were prepared by the
solid-phase synthesis in an oxygen atmosphere [71], whereas single crystals of Eu2SrFe2O7
were prepared from a solution in a PbO melt [72]. However, data on the mechanism and
kinetics of formation of Ln2SrFe2O7 are lacking.
The mechanism of formation of La2SrFe2O7 and Nd2SrFe2O7 has been investigated in
[66,67]. An initial oxides mixture with the stoichiometry corresponding to the compound
La2SrFe2O7 was subjected to thermal analysis up to 1400C (Figure 12); two strong
endothermic effects at 344-422 and 489-578C, accompanied by weight loss of 4.59 and
1.58%, respectively, were revealed. These effects correspond to the stepwise dehydration of
La(OH)3 with the formation of lanthanum oxyhydroxide and oxide, respectively. The
presence of lanthanum hydroxide in the starting mixture was caused by the specific procedure
of sample preparation, involving powdering and mixing of components in distilled water, and
by the reactivity of lanthanum oxide toward water. Starting from 730C, three extra
endothermic effects accompanied by a weight loss are clearly seen. These effects result from
the decomposition of strontium carbonate conjugated to the formation of LaSrFeO4, as
follows from the X-ray diffraction patterns. As the temperature is increased to 1200-1300C,
a sharp increase in the rate of formation of LaSrFeO4 and LaFeO3 is observed, the starting
oxides completely disappear from the mixture, and the formation of La2SrFe2O7 begins. The
formation of Nd2SrFe2O7 goes the same way.
It is known that at temperatures exceeding 1380C iron(III) oxide dissociates to form
Fe3O4 [73]. The thermal analysis of the mixture of the stoichiometric composition La2SrFe2O7
(Figure 12) in air revealed no effects attributable to changes in the iron(III) oxidation state.
The results of the thermal analysis and the data of the X-ray phase analysis of the samples
obtained by heat treatment in the annealing-quenching mode show that the intermediate
compounds LnFeO3 and LnSrFeO4 are formed fairly rapidly at temperatures up to 1200C.
This fact suggests that the intermediate compounds are formed at temperatures lower than the
temperature of the Fe2O3 conversion to Fe3O4 and that the iron(III) state is stabilized in
LnFeO3 and LnSrFeO4.
When the mixtures are sintered in the range 1400-1500C, the rate of the Ln2SrFe2O7
formation increases. The solid-phase reaction is complete after 3 hours heating at 1500C.
In [67] the mechanism of formation of the complex oxide Gd2SrFe2O7 has been examined
by means of X-ray phase analysis and Mossbauer spectroscopy. The phase analysis of the
samples obtained after heat treatment for 3 h at 1200, 1300, and 1400C shows that Gd2O3,
GdFeO3, SrFeO3-x, and Gd2SrFe2O7 present in the reaction mixture. The mechanism of
formation of the complex oxide Gd2SrFe2O7 is not similar to that of the aluminate
Gd2SrAl2O7. The complex aluminate Gd2SrAl2O7 is formed by the reaction of an intermediate
compound of the spinel structure with gadolinium oxide (6).
The composition of the reaction mixture and the structural characteristics of intermediate
compounds indicate that reaction is a complex multistep process, as well as the majority of
solid-phase reactions, and allows to suggest the following mechanism for the solid-phase
synthesis of the complex oxide Gd2SrFe2O7:
1 - 1400C; 2 - 1300C.
Figure 13. Degree of formation of La2SrFe2O7 vs. time dependence for different temperatures.
Figure 14. Degree of formation of GdSrFeO4 vs. time dependence at 1300C.

Figure 15. Fe(III) content in the reaction mixture during the formation of GdSrFeO4 at 1300C.
Gd2O3 + Fe2O3 2GdFeO3 (24)
2SrCO3 + (1 - 2x)O2 + Fe2O3 2SrFeO3-x + 2CO2 (25)
Gd2O3 + 2SrFeO3-x 2GdSrFeO4 + (1 - 2x)O2 (26)
GdFeO3 + GdSrFeO4 Gd2SrFe2O7 (27)
The fact that the compound GdSrFeO4 is not detected in the reaction mixture by X-ray
diffraction can be accounted for by a high rate of its reaction with GdFeO3.
The degree of conversion to the final product La2SrFe2O7 was determined for
temperatures of 1300 and 1400C in relation to the reaction time in the range 1-24 h. This is
illustrated by kinetic curves in Figure 13.
A comparative analysis of the kinetics of La2SrFe2O7 and La2SrAl2O7 formation shows
that the degree of the aluminate conversion in 24 h at 1300C does not exceed 0.2, and for the
ferrite it is 0.8. This fact indicates that the rate of the ferrite formation considerably exceeds
the rate of the aluminate formation and requires lower temperatures. This does not hold for
Gd-containing compounds: Gd2SrAl2O7 is formed much faster than Gd2SrFe2O7.
The mechanism and kinetics of formation of the single-layered Ruddlesden-Popper phase
GdSrFeO4 has been investigated in [74] by means of X-ray diffraction and Mossbauer
spectroscopy. The X-ray phase analysis shows that SrFeO3-x acts as an intermediate
compound in this reaction. According to the Mossbauer spectrum, x is equal to 0.360.05.
The kinetics of the formation of GdSrFeO4 has been characterized by the phase
composition determined from the Xray diffraction data (Figure 14) and by the Fe3+ fraction
in the reaction mixture determined from the Mossbauer data (Figure 15). Both data sets are in
a good agreement.
5. FORMATION OF THE DOUBLE-PEROVSKITE LAYERED BI3TINBO9

The kinetics of the formation of Bi3NbTiO9 from the initial oxides was investigated at
600C. The interaction between intermediate compounds was investigated at 700-900C. The
DTA has been performed in order to confirm the data of the isothermal investigation.
The results of phase analysis show that, like the Ruddlesden-Popper phases, the
Aurivillius phases are formed in a multistage way with more simple phases of the same class
as intermediates. The peaks of Bi3NbTiO9, Bi12TiO20, Bi4Ti3O12 and Bi5Nb3O15 can be
observed at the X-ray powder pattern. This means that during the formation of the n=2 phase
Bi3NbTiO9 two other Aurivillius phases, Bi4Ti3O12 (n=3) and Bi5Nb3O15 (n=1.5) act as
intermediates (Figure 16). The Bi12TiO20 compound has been observed earlier [75] during the
investigation of the formation of Aurivillius phases in the Bi2O3-TiO2-Fe2O3 system. It
appears to be the -form of the bismuth oxide stabilized by titanium dioxide [76]. It is the
formation of stabilized -Bi2O3 that leads to the activation of the solid state reactions in such
systems. In general, the formation of the Bi3NbTiO9 phase can be represented as follows:
Bi2O3 -Bi2O3 (or Bi2O3 + TiO2 Bi12TiO20) (28)
-Bi2O3 +TiO2 Bi4Ti3O12 (29)
-Bi2O3 +Nb2O5 Bi5Nb3O15 (30)
Bi4Ti3O12 + Bi5Nb3O15 Bi3NbTiO9 (31)
Figure 16. The structure transformation during the last stage of Bi3NbTiO9 formation.
The thermal analysis of the reaction mixture reveals two endothermic effects at 740
and 860 (Figure 17). The temperature of the first effect corresponds to the temperature of
formation of the TiO2-stabilized -Bi2O3 (28). The second effect corresponds to the peritectic
melting of Bi12TiO20. [77] The results of the X-ray phase analysis show that after the second
effect almost pure Bi3NbTiO9 is obtained.
The X-ray data of the sample quenched after reaching the temperature of the first
endothermic effect show that besides the previously observed compounds another bismuth
niobate, BiNbO4, presents in the reaction mixture. This fact could be explained by the
assumption that the formation of Bi5Nb3O15 is a two-step reaction:
-Bi2O3 +Nb2O5 BiNbO4 (30a)
-Bi2O3 + BiNbO4 Bi5Nb3O15 (30b)
The second step is much faster than the first one. This results in a very low concentration
of BiNbO4 after long heating time. The reaction (30a) is the limiting stage of the overall
process.
Figure 17. The DTA analysis of the reaction mixture during synthesis of Bi3NbTiO9.
Proposed reaction mechanism accounts for the observed limits of obtaining the
substituted Aurivillius phases. It was found that the lanthanide-substituted phases
Bi2LnNbTiO9 can be obtained only for Ln=La-Gd, while the Mn-substituted Bi3NbTi1-
xMnxO9 can be obtained only for x<0.3. These limits coincide with the limits of the cationic
substitution in the intermediate compound Bi4Ti3O12. Namely, investigations [78] of the
substitution of Bi3+ for Ln3+ in triple-layered Aurivillius Bi4Ti3O12 have shown that
Aurivillius phases are not formed for the lanthanides smaller than Gd. In case of the Mn-
substituted Bi4Ti3O12, the Mn+4 cations are fully ordered in the middle perovskite layer [79].
It means that 1/3 Ti is substituted by Mn in Bi4Ti3O12 and Mn content in Bi3NbTi1-xMnxO9 is
limited by the value x=0.3. So the Bi3NbTiO9 derivatives that would require formation of the
unstable derivatives of Bi4Ti3O12 as intermediates can not be obtained by straightforward
synthesis from the binary oxides.
The kinetics of formation of Bi3NbTiO9 from the initial oxides has been investigated at
600C. The conversion degree versus time plot is shown at Figure 18. The formation of the
Bi3NbTiO9 from the intermediate compounds (synthesized in advance by 6 hours annealing at
1000C) according to the reaction (31) was studied at 700-900C (Figure 19).
Figure 18. Degree of formation of Bi3NbTiO9 from the initial oxides vs. time dependence at 600C.
The results show that Bi3NbTiO9 is formed much faster if the metal oxides are used as the
initial materials than if Bi3NbTiO9 is synthesized from the intermediate compounds. This fact
can be explained in terms of the microstructure of the intermediate compounds in these two
cases. If Bi3NbTiO9 is synthesized from the metal oxides, intermediate compounds react
immediately after they are formed, so they have less particle size and higher concentration of
defects than if they are synthesized in advance.
The mechanism of formation of the Ln-substituted phases Bi2LnNbTiO9 appears to be the
same as for the non-substituted phase. However, the limiting stage changes: for the formation
of the non-substituted Aurivillius phase the reaction rate is limited by the formation of
BiNbO4 (reaction 30b), while for the substituted ones the reaction rate is limited by the
formation of the intermediate Aurivillius phase Bi4LnNb3O15 (reaction 34):
Bi2O3 -Bi2O3 (or Bi2O3 + TiO2 Bi12TiO20)
-Bi2O3 + Ln2O3+ TiO2 Bi2Ln2Ti3O12 (32)
700C; 800C; 900C.
Figure 19. Degree of formation of Bi3NbTiO9 from the intermediate compounds vs. time dependence.
Figure 20. Degree of formation of Bi2LnNbTiO9 at 800C.
-Bi2O3 + Nb2O5 BiNbO4 (33)
BiNbO4 + -Bi2O3 + Ln2O3 Bi4LnNb3O15 (34)
Bi2Ln2Ti3O12 + Bi4LnNb3O15 Bi2LnNbTiO9 (35)
The kinetics of the formation of the Ln-substituted Aurivillius phases has been
investigated at 800 (Figure 20). The reaction rate is much slower than or the non-
substituted phase and decreases along the lanthanide row.
Figure 21. The DTA-TGA analysis of the reaction mixture during synthesis of NaNdTiO4.
6. FORMATION OF THE RUDDLESDEN-POPPER PHASES NANDTIO4

AND NA2ND2TI3O10
The synthesis of NaNdTiO4 is usually carried out using reaction of the sodium carbonate,
neodymium oxide and titanium oxide at a temperature of 900C for 6 h. The experiments
showed that Na2CO3 melts at temperatures of the order of 850C and leaves the reaction zone.
Due to this fact the excess of Na2CO3 is necessary to obtain the pure compound and extra
stage of washing out the residue of the sodium carbonate has to be performed. The synthesis
at higher temperatures leads to the formation of the impurity compounds Na2Nd2Ti3O10 at
950C and Nd2Ti2O7 at 1100C. The study of the thermal stability of the NaNdTiO4 complex
oxide showed that, at a temperature of 1150C, the oxide almost completely decomposes into
the phases Nd2Ti2O7 and Nd2Ti3O9, which do not contain sodium.
The derivatographic study [81] of reaction in the temperature range 25800C
demonstrated that the main weight loss of the reaction system, which is equal to
approximately 10% (Figure 21), is observed in the temperature range 600780C and
accompanied by endothermic effects at temperatures of 600, 660, and 780C. The
endothermic effects at temperatures of 300 and 500C with a small weight loss correspond to
the decomposition of neodymium hydroxide with the formation of neodymium oxyhydroxide
and neodymium oxide, which is explained by the activity of neodymium oxide in the reaction
with water. The thermal analysis of the binary mixture of sodium carbonate and titanium
oxide reproduces the endothermic effects at 600 and 660C.
Figure 22. Electron micrograph of the NaNdTiO4 sample synthesized at 750C.

The derivatographic data agree with the results obtained by the annealingquenching
method [81]. The Na2CO3TiO2 system has been investigated separately, motivated by the
low reactivity of the neodymium oxide at the temperature of synthesis. The X-ray diffraction
patterns of the reaction mixture indicate that sodium titanate Na4Ti5O12 is formed at 600C.
At 720C, the X-ray diffraction patterns contain reflections of the Na8Ti5O14 phase. The
fraction of the Na8Ti5O14 phase increases at 780C and reflections of Na2Ti3O7 appear. At
900C, virtually only the Na8Ti5O14 and Na2Ti3O7 phases are present in the reaction mixture.
In case of the Nd2O3Na2CO3TiO2 system the same processes have been observed up to
720C. After this temperature NaNdTiO4 starts to form. At 780C only NaNdTiO4 is observed
in the reaction mixture.
The determination of the temperature intervals of the phase transformations in the
Na2CO3TiO2 system and the determination of the temperature range of the reaction with
Nd2O3 make it possible to propose the following reaction sequence of the formation of the
NaNdTiO4 oxide:
600C 4Na2CO3 + 10TiO2 2Na4Ti5O12 + 4CO2 (36)
660C Na4Ti5O12 + 2Na2CO3 Na8Ti5O14 + 2CO2 (37)
Na4Ti5O12 l/7Na8Ti5O14 + 10/7Na2Ti3O7, (38)
720780C
{ 10/7Na2Ti3O7 + 2Na2O 6/7Na8Ti5O14,
2Na8Ti5O14 + 5Nd2O3 10NaNdTiO4 + 3Na2O.

(39)
(40)
Reaction (39) requires sodium oxide (possibly, partially transformed into carbonate),
which is formed in the course of reaction (40). This reaction mechanism of the formation of
NaNdTiO4 explains the necessity of using at least 20% excess of sodium carbonate for the
reaction. The synthesis of NaNdTiO4 without a sodium carbonate excess at temperatures
below 780C is practically impossible (instead, the formation of NaNdTi2O6 slowly occurs)
because Na8Ti5O14 is hardly formed in the system, whereas the decomposition reaction of
Na4Ti5O12 proceeds very slowly. With 60% excess of sodium carbonate, only reactions (36),
(37) and (40) sequentially occur.
The analysis of the phase composition of the reaction mixture at 780C enables one to
claim that the formation of the layered NaNdTiO4 oxide is almost complete after 30 minutes
heating in the temperature range 720780C, which is considerably lower than the melting
point of Na2CO3 (830C). Therefore, the losses of sodium carbonate at this temperature are
negligible.
It should be noted that the particles of the material obtained by this procedure have
substantially smaller sizes and a more uniform shape than those obtained according to the
methods described in the literature. Thus, the synthesis of NaNdTiO4 at 750C makes it
possible to obtain flaked particles 200 nm thick (Figure 22). At the same time the mean size
of the particles obtained by the standard procedure is 14 m. This fact cannot be
underestimated taking into account the prospect of the use of NaNdTiO4 as a catalyst.
Figure 23. The DSC-TGA analysis of the reaction mixture during synthesis of Na2Nd2Ti3O10.
Another Ruddlesden-Popper phase in the Nd2O3Na2CO3TiO2 system, Na2Nd2Ti3O10,

attracted interest in the last years due to possible photocatalytic activity and ability to take
part in the metathesis reactions [82,83]. The formation of this phase has been investigated at
temperatures up to 1100C by means of the X-ray diffraction and thermal analysis.
No chemical reactions take place in the Nd2O3TiO2Na2CO3 mixture up to 500C.
Above 600C Na4Ti5O12 phase appear together with the traces of the perovskite-type
Na1/2Nd1/2TiO3. At 700C Na8Ti5O14 starts to form and the amount of Na1/2Nd1/2TiO3
increases significantly. When the reaction mixture is heated further to 850C the Na4Ti5O12
disappears completely. At 900C the Ruddlesden-Popper phases NaNdTiO4 and
Na2Nd2Ti3O10 appear. In the temperature range of 9501050C the amount of Na2Nd2Ti3O10
increases at the expense of NaNdTiO4. At 1100C the formation of Na2Nd2Ti3O10 is
complete.
The DSC-TGA analysis (Figure 23) supports the data obtained under the isothermal
heating conditions. The main mass loss (8.75%) is observed 550820C and can be attributed
to the CO2 evolution during the reactions involving Na2CO3. Four thermal effects can be
found at the DTA curve. First of them appears at 325C and can be attributed to the Nd2O3
dehydration. The second one at 594C corresponds to the beginning of the solid-state
reactions. The large exothermal peak at 845 appears due to the series of reactions including
the decomposition of Na4Ti5O12 and the formation of NaNdTiO4 and Na2Nd2Ti3O10. The last
effect at 957 corresponds to the decomposition of NaNdTiO4.
The following reactions can be proposed for the formation of Na2Nd2Ti3O10:
Figure 24. The structure transformations during the last stage of Na2Nd2Ti3O10 formation.
595 2 Na2CO3 + 5 TiO2 Na4Ti5O12 + 2 CO2 (36)
Na4Ti5O12 + 2 Na2CO3 Na8Ti5O14 + 2 CO2 (37)

695
{ 7 Na4Ti5O12 Na8Ti5O14 + 10 Na2Ti3O7 (38)
600820 Nd2O3 + 4 TiO2 + Na2CO3 4 Na1/2Nd1/2TiO3 + CO2 (41)
10/7Na2Ti3O7 + 2Na2O 6/7Na8Ti5O14, (39)
825895
{ 2Na8Ti5O14 + 5Nd2O3 10NaNdTiO4 + 3Na2O.
4 Na1/2Nd1/2TiO3 Na2Nd2Ti3O10 + TiO2

(40)
(42)
9601100 2 NaNdTiO4 + TiO2 Na2Nd2Ti3O10 (43)

c d
e
Figure 25. The reaction lines for the formation of different types of the layered perovskite-like oxides.
a double-layered Ruddlesden-Popper aluminates; b double-layered Ruddlesden-Popper manganites;
c double-layered Ruddlesden-Popper ferrites; d double-layered Aurivillius phases; e single- and
triple-layered Ruddlesden-Popper titanates.
The formation of Na2Nd2Ti3O10 goes by two ways: about 900C Na1/2Nd1/2TiO3

decomposes to form Na2Nd2Ti3O10 and TiO2, while at 960C Na2Nd2Ti3O10 is formed by
interaction of NaNdTiO4 and TiO2 appearing at the previous stage. The structure
transformations during these processes are shown at Figure 24.
It is interesting to note that during the formation of Na2Nd2Ti3O10 it is possible to observe
the same reactions (36-40) as during the formation of NaNdTiO4. However, the temperature
necessary for these reactions is approximately 100C higher (825895C instead of 780C).
7. CONCLUSION
The formation of the layered perovskite-like oxides usually goes by a complex multistage
way. The set of the intermediate compounds may strongly affect the properties of the final
product or even the possibility of obtaining the pure phase. Some of the typical reaction
pathways are shown at Figure 25. In the most cases the intermediates are compounds of the
same family with simpler structure. However, if these compounds are not stable under the
conditions of synthesis, the mechanism may become quite complex and unpredictable.
In the series of the phases containing chemically analogous elements the reaction rate
changes regularly provided that the mechanism remains unchanged. On the other hand,
change in the mechanism might result in the strong jump of the reaction rate.
On the whole, investigations of the process of formation of the layered perovskite-like
compounds are very important for optimization of the synthesis conditions and in many cases
may give explanation to the properties of the final product.
ACKNOWLEDGMENT
This work was partly supported by Russian Foundation for Basic Research (grant 09-03-
00853).
REFERENCES
[1] Wells, A.F., Structural Inorganic Chemistry. Oxford: Oxford Univ. Press, 1986.
[2] Rao, C.N.R., Raveau, B. Transition metal oxides: structure, properties, and synthesis
of ceramic oxides. Wiley-VCH, New York, Chichester, Weinheim, Brisbane,
Singapore, Toronto, 1998.
[3] Zvereva, I., Smirnov, Yu., Gusarov, V., Popova, V., Choisnet, J. Complex
aluminates RE2SrAl2O7 (RE = La, Nd, Sm-Ho):Cation ordering and stability of the
double perovskite slab-rocksalt layer P2/RS intergrowth. Solid State Sciences. 2003,
5, 343-349.
[4] Smirnov, Yu. E., Zvereva, I. A., Cation Distribution and Interatomic Interactions in
Oxides with Heterovalent Isomorphism: III. Complex Aluminates LnCaAlO4 (Ln =
Y, La, Nd, Gd, Ho, Er, Yb). Russian Journal of General Chemistry. 2001, 71, 845
852.
[5] Ruddlesden, S.N., Popper, P. New compounds of the K2NiF4 type. Acta Crystallogr.
1957, 10, 538-539.
[6] Ruddlesden, S.N., Popper, P. The compounds Sr3Ti2O7 and its structure. Acta
Crystallogr. 1958, 11, 54-55.
[7] Oudalov, Y.P., Daoudi, A., Jourbort, J.C., Le Flem, G., Hagenmuller P. Sur une
nouvelle serie dalluminates double de calcium et delements lanthanique de
structure K2NiF4. Bull. Soc. Chem. Fr. 1970, 10, 3408-3411.
[8] Fava, J., Oudalov, Y.P., Reau, J.-M., Le Flem, G., Hagenmuller, P. Sur une nouvelle
famille dalluminates double de strontium ou deuropium divalent et de terres rares.
Compt. Rend. 1972, C274, 1837-1839.
[9] Longo, J.M., Raccah, P.M. The structure of La2CuO4 and LaSrVO4. J. Solid State
Chem. 1973, 6, 526-531.
[10] Fava, J., Le Flem, G. Etude des oxide avec une structure lamellaire. Mat. Res. Bull.
1975, 10, 75-80.
[11] Demazeau, G., Pouchard, M., Hagenmuller P. Sur quelque nouveaux composes
oxyges du Nickel +III de structure K2NiF4. J. Solid State Chem. 1976, 18, 159-162.
[12] Demazeau, G., Courbin, P., Main, J.C., Le Flem, G. Sur une nouvelle serie doxides
du cobalt +III de structure K2NiF4. Compt. Rend. Acan. Sci. Paris. 1976, 233, 61-62.
[13] Fava, J., Danot, M., Dinh, N.., Daoudi, A., Le Flem, G. Etude de
radiocrystallographique et par resonance Mossbauer de CaLaFeO4. Solid State
Comm. 1977, 22, 733-736.
[14] Benabad, A., Daoudi, A., Salmon, R., Le Flem, G. Les phases SrLnMnO4,
BaLnMnO4 MnxLa1-xMnO4. J. Solid State Chem. 1977, 22, 121-126.
[15] Bednorz, J.G., Muller, K.A. Possible high-Tc superconductivity in the Ba-La-Cu-O
system. J. Phys. B. Condesed Matter. 1986, 64, 189-193.
[16] Soybeyroux, G.-L., Delmas, C. The 2D magnetic oxides. J. Magn. Mat. 1980, 15-18,
1315-1316.
[17] Le Flem, G., Demazeau, G., Hagenmuller, P. Relations between Structure and
Physical properties in K2NiF4-type Oxides. J. Solid State Chem. 1982, 44, 82-88.
[18] Le Flem, G., Courbin, P., Delmas, C., Soybeyroux, J.-L. Magnetic properties of the
2D oxides. Z. Anorg. Allg. Chem. 1981, 476, 69-88.
[19] Ganguly, P., Rao, .N.R. Crystal chemistry and magnetic properties of layered metal
oxides possessing the K2NiF4 or related structure. J. Solid State Chem. 1984, 53, 193-
216.
[20] Le Flem, G. Origine des couplages magnetique tridimensionelles dans les oxydes de
type K2NiF4. Compt. Rend. Acad. Sci. Paris. 1982, 233, S.., 29-32.
[21] Le Flem, G. Magnetic properties of ternary compounds. Il nuova cimento. 1983, 2B,
1814-1822.
[22] Mohan Ram, R.A., Singh, K.K., Madhusadan, W.H., Ganguly, P., Rao, C.N.R.
Electronic and spin configurations of Co3+ and Ni3+ ions in oxides of K2NiF4 type
structure: a magnetic susceptibility study. Mat. Res. Bull. 1983, I 8, 703-712.
[23] Kovba, M.L., Lukova, L.N., German, M.V., Antipov, E.V. Oxides with perovskite-
type structure. Russian Journal of General Chemistry. 1986, 56, 1006-1014.
[24] Buttrey, D., Honig, J.M., Rao, C.N.R. Magnetic properties quasi-two-dimensional
La2NiO4. J. Solid State Chem. 1986, 64, 287-295.
[25] Jorgensen, J.D., Dabrovski, ., Pei, S., Richrds, D.R., Hikks D.G. Structure of the
interstitial oxygen defect in La2NiO4+. Phys. Rev. Bull. 1989, 40, 2187-2199.
[26] Wada, S., Furukawa, Yu., Kaburagi, M., Kajitani, ., Mosoya, S., Yamada, Yo.
Magnetic and electronic structure of antiferromagnetic LaNi oxide (La2NiLa2NiO4+)
and LaSrNi oxide (La2-xSrxNiO4+). J. Phys.: Condens. Mater. 1993, 5, 765-780.
[27] Fava, J., Le Flem, G., Les phases SrLa2Al2O7 et SrGd2Al2O7. Mat. Res. Bull. 1975,
10, 75-80.
[28] Oudalov, Y.P., Salmon, R, Bondar, I.A. System SrO-Nd2O3-Al2O3. Russian Journal
of Inorganic Chemistry. 1976, 21, 541-546.
[29] Popova, V. F., Tugova, E. A., Isaeva, A. S., Zvereva, I. A., Gusarov, V. V. Phase
Equilibria in the Ho2O3SrAl2O4 System. Glass Physics and Chemistry. 2007, 33,
498
[30] Toda, K., Kameo, Y., Kurita, S., Sato, M. Crystal structure determination and ionic
conductivity of layered perovskite compounds NaLnTiO4 (Ln = rare earth). Journal
of Alloys Compd. 1996, 234, 19-25.
[31] Byeon, S.-H., Park, K., Itoh, M. Structure and ionic conductivity of NaLnTiO4;
comparison with those of Na2Ln2Ti3O10 (Ln = La, Nd, Sm, and Gd). J. Solid State
Chem. 1996, 121, 430-436.
[32] Schaak, R.E., Mallouk, T.E. KLnTiO4 (Ln = La, Nd, Sm, Eu, Gd, Dy): A New Series
of Ruddlesden-Popper Phases Synthesized by Ion-Exchange of HLnTiO4. J. Solid
State Chem. 2001, 161, 225-232.
[33] Sato, M., Toda, K., Watanabe, J., Uematsu, K. Structure determination and silver ion
conductivity of layered perovskite compounds M2La2Ti3O10 (M = K and Ag). J.
Chem. Soc. Japan. 1993, 640-646.
[34] Toda, K., Watanabe, J., Sato, M. Crystal structure determination of ion-exchangeable
layered perovskite compounds, K2La2Ti3O10 and Li2La2Ti3O10. Materials Research
Bulletin. 1996, 31, 1427-1435.
[35] Toda, K., Kameo, Y., Fujimoto, M., Sato, M. Crystal structure and ionic conductivity
of a layered perovskite, Na2La2Ti3O10. Journal of Ceram. Soc. Jpn. 1994, 102, 737-
741.
[36] Park, K., Byeon, S.-H. Correlation between structures and ionic conductivities of
Na2Ln2Ti3O10 (Ln = La, Nd, Sm, and Gd). Bulletin of the Korean Chemical Society.
1996, 17, 168-172.
[37] Wright, A.J., Greaves, C. A neutron diffraction study of structural distortions in the
Ruddlesden-Popper phase Na2La2Ti3O10. Journal of Materials Chemistry. 1996, 6,
1823-1825.
[38] Lalena, J.N., Falster, A.U., Simmons, W.B., Carpenter, E.E. Synthesis and
characterization of new mixed-metal triple-layered perovskites, Na2La2Ti3-xRuxO10 (x
1.0). Chemistry of Materials. 2000, 12, 2418-2423.
[39] Hong, Y.-S., Kim, K. New-type ordering behaviour in the layered perovskite
compound RbLa2Ti2NbO10. Chemistry Letters. 2000, 29, 690-691.
[40] Toda, K., Teranishi, T., Takahashi, M., Zuo-Guang, Ye Z.-G., Sato, M. Structural
chemistry of new ion-exchangeable tantalates with layered perovskite structure: new
reduced Ruddlesden-Popper phase, Na2Ca2Ta3O10. Solid State Ionics. 1998, 113-115,

501-508.
[41] Blasse, G. New compounds with K2NiF4 structure. J. Inorg. Nucl. Chem. 1965, 27,
2683-2684.
[42] Battle, P.D., Green, M.A., Laskey, N.S., Millburn, J.E., Murphy, L., Rosseinsky,
M.J., Sullivan, S.P., Vente, J.F. Layered Ruddlesden-Popper Manganese oxides:
Synthesis and Cationic Ordering. Chem. Mater. 1997, 9, 552-559.
[43] Battle, P.D., Rosseinsky, M.J. Synthesis, structure, and magnetic properties of n = 2
Ruddlesden-Popper manganites. Current Opinion in Solid State and Materials
Science. 1999, 4, 163-170.
[44] Dann, S.E., Welle,r M.T., Currie, D.B. Structure and Oxygen Stoichiometry in
Sr3Fe2O7-y, 0 y 1.0. J. Solid State Chem. 1992, 97, 179-185.
[45] Seshadri, R., Martin, C., Maignan, A., Hervieru, M., Raveau, B. Structure and
magnetotransport properties of layered manganites RE1.2Sr1.8Mn2O7 (RE = La, Pr,
Nd). J. Mater. Chem. 1996, 6, 1585-1590.
[46] Floros, N., Hervieu, M., Van Tendeloo, G., Michel, C., Maignan, A., Raveau, B. The
layered manganate Sr4-xBaxMn3O10: synthesis, structural and magnetic properties.
Solid State sciences. 2000, 2, 1-9.
[47] Siruguri, V., Ganguly, R., Krishna, P.S.R., Sastry, P.U., Paranjpe, S.K.,
Gopalakrishnan, I.K., Yakhmi, J.V. Stability of Sr3Ti2O7 structure in La1.2(Sr1-
xCax)1.8Mn2O7 and Ca3-yLayMn2O7. J. Mater. Chem. 2001, 11, 11581161.
[48] Yu, R.C., Zhu, J.L., Li, S.Y., Li, F.Y., Zhang, Z., Jin, C.Q., Voigt-Martin, I.G.
HTEM studies on La2-2xCa1+2xMn2O7 (x = /0.5 and 0.25). Materials Science and
Engineering. 2003, A345, 344-349.
[49] Kazanova, N.R., Kovba, M.L., Putilin, S.N., Antipov, E.V., Lebedev, O.I., Van
Tendeloo, G. Synthesis, structure and properties of layered bismuthates:
(Ba,K)3Bi2O7 and (Ba,K)2BiO4. Solid State Communications. 2002, 122, 189-193.
[50] Dion, M., Ganne, M., Tournoux, M., Nouvelles familles de phases MIMII2Nb3O10 a
feuillets perovskites. Materials Research Bulletin. 1981, 16, 1429-1435.
[51] Jacobson, A.J., Lewandowski, J.T., Johnson, J.W. Ion exchange of the layered
perovskite KCa2Nb3O10 by protons. Journal of the Less-Common Metals. 1986, 116,
137-145.
[52] Jacobson, A.J., Johnson, J.W., Lewandowski, J.T. Interlayer chemistry between thick
transition-metal oxide layers: synthesis and intercalation reactions of K[Ca2Nan-
3NbnO3n+1]. Inorg. Chem. 1985, 24, 3727-3729.
[53] Thangadurai, V., Schmid-Beurmann, P., Weppner, W. Synthesis, Structure, and
electrical conductivity of A[A2B3O10] (A = Rb, Cs; A = Sr, Ba; B = Nb, Ta): new
members of Dion-Jacobson type layered perovskites. J. Solid State Chem. 2001, 158,
279-289.
[54] Hong, Y.-S., Han, C.-H., Kim, K. Structural characterization of new layered
perovskites MLa2Ti2TaO10 (M = Cs, Rb) and NaLa2Ti2TaO10xH2O (x = 2, 0.9, 1.0).
J. Solid State Chem. 2001, 158, 290-298.
[55] Hong, Y.-S., Kim, K. New-type ordering behaviour in the layered perovskite
compound RbLa2Ti2NbO10, Chemistry Letters. 2000, 29, 690-691.
[56] Aurivillius, B. Mixed bismuth oxides with layer lattices. I, II, and III. Ark. Kemi.
1949, 1, 463-471; 1949, 1, 499-512; 1950, 1, 519-528.
[57] Smolenski, G. A., Isupov, V. A., Agranovskaya, A. I., Ferroelectrics of the Oxygen-
Octahedral Type With Layered Structure. Sov. Phys. Solid State. 1961, 3, 651-655
[58] Hervoches, H., Snedden, A., Riggs, R., Kilcoyne S.H., Manuel P.,
Lightfoot P. Structural Behavior of the Four-Layer Aurivillius-Phase Ferroelectrics
SrBi4Ti4O15 and Bi5T i3FeO15. J. Solid State Chem. 2002, 164, 280-291.
[59] Pineda-Flores, J.L., Chavira, E., Huanosta-Tera, A. Ferroelectric characteristics in
Aurivilius solid solutions: Bi4-xLnxTi3O12 (Ln = Gd, Dy), 0 < x < 0.8. Physica C.
2001, 364-365, 674-677.
[60] Boullay, Ph., Trolliard, G., Mercurio, D., Perez-Mato, J.M., Elcoro, L. Toward a
unified approach to the crystal chemistry of Aurivillius-type compounds. J. Solid
State Chem. 2002, 164, 252-271.
[61] Subbanna, G.N., Guru Row, T.N., Rao, C.N.R. Structure and dielectric properties of
recurrent intergrowth structures formed by the Aurivillius family of bismuth oxides
of the formula Bi2An-1BnO3n+3. J. Solid State Chem. 1990, 86, 206-211.
[62] Zvereva, I. A., Popova, V. F., Vagapov, D. A., Toikka, A. M., Gusarov, V. V.
Kinetics of Formation of Ruddlesden-Popper Phases: I. Mechanism of La2SrAl2O7
Formation. Russian Journal of General Chemistry. 2001, 71, 1181-1185.
[63] Zvereva, I. A., Popova, V. F., Pylkina, N. S., Gusarov, V. V. Kinetics of Ruddlesden-
Popper Phase Formation: II. Mechanism of Nd2SrAl2O7 and Sm2SrAl2O7 Formation.
Russian Journal of General Chemistry. 2003, 73, 433-447
[64] Zvereva, I. A., Popova, V. F., Missyul, A. B., Gusarov, V. V. Kinetics of
Ruddlesden-Popper Phase Formation: III. Mechanism of Formation of Gd2SrAl2O7.
Russian Journal of General Chemistry. 2003, 73, 684-688
[65] Isaeva, A. S., Kozhina, I. I., Toikka, A. M., Zvereva, I. A. StructuralChemical
Mechanism of Formation of Solid Solutions in the La2SrAl2O7Ho2SrAl2O7 System.
Glass Physics and Chemistry. 2006, 32, 106112.
[66] Missyul, A. B., Marchenko, E. M., Popova, V. F., Zvereva, I. A. Mechanism and
Kinetics of Formation of Solid Solutions in the Nd2SrAl2O7Ho2SrAl2O7 System.
Glass Physics and Chemistry. 2003, 29, 608612.
[67] Tugova, E. A., Popova, V. F., Zvereva, I. A., Gusarov, V. V. Mechanism and
Kinetics of Formation of La2SrFe2O7 and Nd2SrFe2O7. Russian Journal of General
Chemistry. 2007, 77, 979-981.
[68] Zvereva, I. A., Otrepina, I. V., Semenov, V. G., Tugova, E. A., Popova, V. F.,
Gusarov, V. V. Mechanism of Formation of the Complex Oxide Gd2SrFe2O7.
Russian Journal of General Chemistry. 2007, 77, 973-978.
[69] Majewski, P., Benne, D., Epple, L., Aldinger, F. Phase equilibria in the system
La2O3-SrO-Mn3O4 in air. Intern. J. of Inorg. Materials. 2001, 3, 12571259.
[70] Cherepanov, V. A., Barkhatova, L. Yu., Voronin, V. I. Phase Equilibria in the La-Sr-
Mn-O System. J. Solid State Chem. 1997, 134, 3844
[71] Kriegel, R., Feltz, A., Walz, L., Simon, A., Mattausch, H.-J. ber die Verbindung
Sr7Mn4O15 und Beziehungen zur Struktur von Sr2MnO4 und alpha-SrMnO3. Z. anor.
allg. Chem. 1992, 617, 99104.
[72] Sharma Indu, B., Singh, D., Magotra, S.K. Effect of substitution of magnetic rare
earths for La on the structure, electric transport and magnetic properties of
La2SrFe2O7. J. Alloys Comp. 1998, 269, 13-16.
[73] Drofenik, M., Kolar, D., Golic, L. Crystal growth and crystallographic study of
Eu2SrFe2O7. Journal of Crystal Growth. 1973, 20, 75-76.
[74] Darken, L.S., Gurry, R.W. The system iron-oxygen. I. The wstite field and related
equilibria. J. Am. Chem. Soc. 1945, 67, 1398.
[75] Otrepina, I. V., Volodin, V. S., Zvereva, I. A., Liu, J.-Sh. Investigation of the
Formation of the GdSrFeO4 Oxide. Glass Physics and Chemistry. 2009, 35, 423430.
[76] Morozov, M. I., Gusarov, V. V. Synthesis of Am1Bi2MmO3m+3 Compounds in the
Bi4Ti3O12BiFeO3 System. Russian Journal of Inorganic materials. 2002, 38, 723-
729.
[77] Craiq, D. C., Stephenson, N. C. Structural studies of some body-centered cubic
phases of mixed oxides involving Bi2O3: The structures of Bi25FeO40 and Bi38ZnO60.
J. Solid State Chem. 1975, 15, 1-8.
[78] Bruton, T. M. Study of the liquidus in the system Bi2O3-TiO2. J. Solid State Chem.
1974, 9, 173-175.
[79] Armstrong, R. A., Newnham, R. E. Bismuth titanate solid solutions. Mat. Res. Bull.
1972, 7, 1025-1034.
[80] Yu, W. J., Kim, Y. I., Ha, D. H., Lee, J. H., Park, Y. K., Seong, S., Hur, N. H. A new
manganese oxide with the Aurivillius structure: Bi2Sr2Nb2MnO12-. Solid State
Communications. 1999, 111, 705709.
[81] Zvereva, I. A., Silyukov, O. I., Markelov, A. V., Missyul, A. B., Chislov, M. V.,
Rodionov, I. A., Liu, J. Sh. Formation of the Complex Oxide NaNdTiO4. Glass
Physics and Chemistry, 2008, 34, 749755.
[82] Richard, M., Brohan, L., Tournoux, M. Synthesis, characterization, and acid
exchange of the layered perovskites: A2Nd2Ti3O10 (A=Na, K). Journal of Solid State
Chemistry. 1994, 112, 345-354.
[83] Tai, Y.-W., Chen, J.-S., Yang, Ch.-C., Wan, B.-Z. Preparation of nano-gold on
K2La2Ti3O10 for producing hydrogen from photo-catalytic water splitting. Catalysis
Today. 2004, 97, 95-101.
Reviewed by Professor Jacques Choisnet, Laboratoire CRISMAT, CNRS ENSICAEN

and Caen University, France
Chapter 9
SOME THEORETICAL ASPECTS OF MAGNETIC

STRUCTURE, SPIN EXCITATIONS AND
MAGNETIZATION OF MANGANITES
B. V. Karpenko*
Institute of Physics of Metals, Ural Division, Russian Academy of Science,
Ekaterinburg 620041, Russia
ABSTRACT
The behavior of the system La1 xCa x MnO3 is considered theoretically at the
absolute zero temperature (magnetic configurations), at the low temperatures (spin
waves) and at the high temperatures (magnetization).
The appearance of the different magnetic configurations in the row 0 x 1 is
studied. Superexchange, double exchange and anisotropy energy are taken into account.
The spin operator of double exchange interaction in crystal between ions with different
3+ 4+
valence Mn and Mn is the straight generalization of two - spin operator problem
of Anderson Hasegawa molecule. Ground state energy minimization gives 11 types of
magnetic configurations: two ferromagnetic, three collinear antiferromagnetic and six
non-collinear. When the concentration of Ca ions changes, then the spin configurations
replace one another as the ground state. As a whole the sequence of configurations when
x changes from 0 to 1 can be brought in correspondence to those observed at the
experiment. The appearance of the angle configurations in the systems may occur if the
Heisenberg and non Heisenberg exchange exist simultaneously. The comparison with
experiment was made by means of mixed procedure: part of the numerical values of
interaction parameters and transition concentrations x were taken from the experiment
while the missing values of these parameters were calculated with the help of the
*
Data for correspondence: Karpenko Boris Victorovich. Senior scientific worker. Laboratory of Magnetic
Semiconductors, Institute of Physics of Metals, Ural Division, Russian Academy of Science, 18 S.
Kovalevskaya Street, Ekaterinburg 620041, Russia. E-mail: [email protected]
378 B. V. Karpenko
corresponding theoretical relations. The proposed dynamical alloy model of different

spins can be used for description of real perovskite-like systems behavior.
The elementary spin excitations were studied in ferromagnets and antiferromagnets
for the compounds with the special contents and theoretical dispersion curves were
compared with experimental ones obtained by neutron inelastic scattering method for
different directions in the first full Brillouin zone. Theoretical expressions are obtained
for the spin-wave spectrum in a ferromagnetic cubic crystal with variable-valence ions.
Interactions up to the fourth coordination shell are taken into account. Two kinds of
exchange are assumed to occur in the system. As a result, the numerical values of total
exchange parameters for each of the coordination shells can be found. These studies are
made for ferromagnets at x = 0.17,0.25,0.30 . Theoretical expressions are obtained for
energy of spin waves in layered antiferromagnetic structure of the perovskite-like
compound LaMnO3 . The quadratic exchange interactions of the central spin moment of
the trivalent manganese ion with its eighteen neighbors from five coordination shells and
single-ion magnetic anisotropy energy are taken into account. For the three principal
crystallographic directions, the theoretical results are compared with the experimental
data. It is shown that the in-plane ferromagnetic interaction can be more than sixteen
times as great as the interplanar antiferromagnetic interaction.
The high temperature dependence of spontaneous magnetization at a separate site in
the compounds with 0 x 1 is investigated in the molecular field approximation
using a homogeneous model for the magnetic sub-lattice with allowance for the
Heisenberg and double exchange between nearest neighbors. Unlike with the usual
magnets having one type of magnetic ions, in this case there appears a system of
equations with two unknowns mean spins of manganese ions. In the vicinity of the
ordering temperature, this system can be solved analytically and the expression for the
ordering temperature can be obtained as well. To find the temperature dependence of
magnetization in the whole temperature range the exact expression for the double
exchange operator containing spin operators under square root was used. The numerical
solutions were found in this case.
INTRODUCTION
It is already sixty years herovskite-like manganites continue to surprise the investigators
by their unusual properties. The fundamental works [1-7], experimental and theoretical, gave
the ground for subsequent studies. Up to now there are innumerable publications on this
theme. The review articles [8-24] may help to navigate in this topic. So far there is no more or
less common point of view on the physical essence of manganites. Therefore many results
and approaches are in controversy.
In present paper it will be considered some particular problems concerning magnetic
phases, elementary excitations and magnetization in manganites.
Regularly the perovskite-like manganese compounds were began to study as long as1950
years. Zener was the first who proposed the mechanism of indirect interaction of localized
moments in crystal via conduction electrons [3, 4] and called it as double exchange. The
experimental results by Jonker and van Santen [1, 2], that connected ferromagnetism and
conductivity, were in full qualitative concordance with Zeners theory. The general idea of
Some Theoretical Aspects of Mangetic Structure, Spin Excitations 379
Zener gave rise to appear two main approaches. The first is double exchange (DE)
mechanism in the sense of Anderson and Hasegawa [6]. The second is the mechanism of
indirect interaction by Ruderman-Kittel-Kasuya-Yosida (RKKY) [25-27].
DE and RKKY mechanisms use opposite initial approaches: if the DE is believed that the
kinetic energy of the electron is small comparing with potential energy (interatomic
exchange), but in RKKY - on the contrary, the kinetic energy (Fermi energy) is supposed to
be substantially higher than the interatomic interactions. As a result RKKY mechanism leads
to the effective interaction of localized moments of Heisenberg type, and DE to non-
Heisenberg one. So we can say that, DE is more suitable for semiconductors and RKKY - for
metals. RKKY can lead both to ferromagnetic and antiferromagnetic interactions (the
interaction of an oscillating type), while DE only to ferromagnetic.
It must be said that the idea of a correlation between mobile and localized electrons
appeared in the thirties of the last century and it was proposed by two soviet scientists
Schubin and Vonsovsky in the paper [28]. In this model it is supposed that the localized
electrons (d-electrons) are in magnetic order due to, for example, direct exchange interaction,
while conduction electrons (s-electrons) are magnetized due to s-d interaction, which leads to
a number of peculiarities in the behavior of the last. A detailed description of this model and
the various effects that appear when it is used can be found in the monograph [29].
As stated above, there are large differences between the values of kinetic and potential
energies in the different models, but in principle it is possible to imagine a situation where
both the energy of one order of magnitude and then there appears a need to formulate new a
mechanism for interaction.
With regard to the specific compounds of perovskite-like manganite, it is difficult in
advance to give preference to any kind of exchange interactions that are realized in them. We
can assume the existence of direct exchange interactions between magnetic cations, indirect
interactions through anions (superexchange), the double exchange interaction and the
interaction via phonons (as evidenced by isotopic effect). Below, if necessary, we will
provide two types of exchange interactions: of Heisenberg and non-Heisenberg types. In
doing so, for us it is unimportant by what interaction (direct or indirect) the Heisenberg
effective spin Hamiltonian is caused. Under the non-Heisenberg Hamiltonian we will bear in
mind the effective spin operator of the double exchange interaction. Only in the consideration
of the dispersion curves for the spin-wave spectrum in LaMnO3 the double exchange
operator will not be available due to natural causes.
PART 1. MAGNETIC STRUCTURES

This part is written following the results of the papers [1, 2, 5-7, 30-40].
Since below the notion of the double exchange will be often used then it seems helpful to
give the main features of this interaction mechanism.
380 B. V. Karpenko
1.1. Anderson-Hasegawa Molecule and Effective Spin Hamiltonian
So we shall briefly set main results for Anderson-Hasegawa molecule [6]. Let we have
two centers a and b with localized on them S a = Sb = S and one additional electron
(spin = 1 2 ), which can transfer from one center to another with transition integral B > 0 .
r r r
On each site spin is bound with spins S a or Sb into rigid spin S + 1 / 2 due to Hunds
rule (the exchange integral J ). Then as it is shown in Ref. [6], the lower group of energy
levels E at J is given by the expression
1 2 S0 + 1
E= B, (1.1)
2 2S + 1
where S0 is the total molecule spin;
r r r r
S 0 = S a + Sb + . (1.2)
S0 takes 2 S + 1 values:
S0 = 1 / 2,3 / 2,..2 S + 1 / 2 . (1.3)
Each of these numbers is realized twice.

Lets consider the identity
1 2 r r r
S 0 = + 1 + S (2 S + 3) 1 + ( S a + ) Sb . (1.4)
2 S (2 S + 3)
Then energy Eq. (1.1) can be treated as the eigenvalues of an effective spin operator
) r r r
H D = pB 1 + q( S a + ) Sb , (1.5)
where
1 + S (2S + 3) 2
p= ,q = . (1.6)
2S + 1 1 + S (2S + 3)
(It is clear, that Eqs. (1.4) and (1.5) can be presented in symmetrized form.)
)
The operator H D from Eq. (1.5) we shall call the double exchange interaction operator.
)
Its non-Heisenberg form is obvious. The operator H D represents interaction between rigid
spins S + 1 / 2 and S . Just in this form we shall use it considering the crystal.
c,x
1 1
1 1
3 3
3 3
1 1 b,z
1 1
a,y
Figure 1.1. The elementary cell of La1-xCaxMnO3. Only Mn-sites are denoted. Figures 1-4 number
Bravais lattices.
1.2. Hamiltonian for the Crystal
Lets write down the model Hamiltonian for the crystal. We shall use the homogeneous
model for magnetic manganese sub-lattice in La1 xCa x MnO3 , that is we shall suppose that
4+
each site in the manganese sub-lattice is occupied with probability x by ion Mn and has
382 B. V. Karpenko
spin S ( Mn 4 + ) = 3 / 2 and with probability 1 x - by ion Mn3+ and has

3+
spin S ( Mn ) = 2 . Thus our magnetic system represents a homogeneous dynamical spin
alloy of rigid spins of S1 = 2 and S 2 = 3 / 2 . Further the nearest neighbor approximation
will be used when each spin interacts with its nearest environment by means of two
mechanisms: Heisenberg (quadratic) and non-Heisenberg (double exchange).
Lets assume that the compound La1 xCa x MnO3 has the GdFeO3 type of structure in
the whole range 0 x 1 . The elementary cell of orthorhombic lattice is presented on Figure
1.1 where only manganese ions sites are presented. Numbers 1, 2, 3, 4 enumerate four Bravais
lattices. The base vectors of Bravais magnetic lattices (sub-lattices) are
r r 1 r r r 1r r 1 r r r (1.7)
1 = 0, 2 = (a + b ), 3 = c , 4 = (a + b + c ),
2 2 2
r r r
where a , b and c are vectors of primitive translations for a simple orthorhombic lattice.
Vectors of six nearest neighbors are given by the expressions
r 1 r r r 1 r r r 1 r r r 1 r r
1 = (a + b ), 2 = (a + b ), 3 = (a + b ), 4 = (a b ), (1.8)
2 2 2 2
r 1r r 1 r
5 = c , 6 = c.
2 2
3+
Lets denote the integral of Heisenberg exchange between ions Mn Mn3 + in (ab) -
r r r r
plane as I1 (the nearest neighbors 1 , 2 , 3 , 4 ), while that along c - axis as I 2 (the nearest
r r 4+
neighbors 5 and 6 ). Similarly for interaction Mn Mn 4 + in (ab) - plane as I 3 ,
3+
while along c - axis as I 4 . Interactions Mn Mn 4 + in (ab) - plane we shall denote
3+ 4+
as I 5 , and along c - axis as I 6 . The transfer integral between ions Mn and Mn in (ab)
- plane lets be B1 , and along c - B2 . Besides we shall take into consideration anisotropy
energy choosing b - axis as an easy one. We shall denote the anisotropy constant for ions
Mn3 + as K1 > 0 , and for ions Mn 4 + as K 2 > 0 . Thus 10 interaction parameters are taken
into account in our problem: I1 , I 2 , I 3 , I 4 , I 5 , I 6 , B1 , B2 , K1 and K 2 .
)
With regard for all stated above lets write down spin Hamiltonian H in the form:
) ) ) )
H = H1ex + H 2 ex + H a , (1.9)
where
) 4 6 2 2 r r r r r r r r
H1ex = I nl ( i ) S n (m + k ) Sl (m + k + i ), (1.10)
r
m k =1 i =1 n =1 l =1
) r r r r r r r
4 6 2 2 r 2 S n ( m + k ) Sl ( m + k + i )
H 2 ex = Bnl ( i ) 1 + , (1.11)
r
m k =1 i =1 n =1 l =1 1 + S (2 S + 3)
) 4 2
r r
H a = An ( S nz ( m + k )) 2 . (1.12)
r
m k =1 n =1
) )
Here H1ex is the operator of Heisenberg exchange interaction, H 2ex is the operator of
) r
double exchange and H a is the operator of anisotropy energy. The summation over m in
Eqs. (1.10-1.2) is led upon N sites of sub-lattice (the whole number of sites is 4 N ). Sum
over index k means summing over four sub-lattices. Index i numbers the nearest neighbors.
3+ 4+
Indexes n and l distinguish ions Mn ( n, l = 1) and Mn (n, l = 2) . In this case we have
S1 = S + 1 / 2 (spin on ion Mn 3+ ) and S 2 = S (spin of ion Mn 4 + , S = 3 / 2 ). The following
notations are introduced also in Eqs. (1.10-1.2):
r
I11(i ) = (1 x)2 I1;i = 1,2,3,4. (1.13)
r
I11(i ) = (1 x)2 I 2;i = 5,6. (1.14)
r
I 22 (i ) = x2 I3;i = 1,2,3,4. (1.15)
r
I 22 (i ) = x2 I 4;i = 5,6. (1.16)
r r
I12 (i ) = I21(i ) = x(1 x)I5;i = 1,2,3,4. (1.17)
r r
I12(i ) = I21(i ) = x(1 x)I6;i = 5,6. (1.18)
r r
B11(i ) = B22 (i ) = 0. (1.19)
r r S +1 (1.20)
B12 (i ) = B21(i ) = x(1 x) B ;i = 1,2,3,4.
2S +1 1
384 B. V. Karpenko
r r S +1 (1.21)
B12 (i ) = B21(i ) = x(1 x) B ;i = 5,6.
2S +1 2
A1 = (1 x) K1, A2 = xK 2. (1.22)
)
The double exchange operator H 2 ex from Eq. (1.11) in the limit case of two sites goes to
operator of Eq. (1.5) of Anderson-Hasegawa molecule.
For the following analysis it is convenient to introduce local coordinate systems
according to Euler formulas for direction cosines ij :
11 = sin cos cos cos sin ,

12 = cos cos cos sin sin ,
13 = sin sin ,
21 = sin sin ,
22 = cos sin cos sin cos , (1.23)
23 = sin cos ,
31 = sin cos ,
32 = sin sin ,
33 = cos .
Later we shall not investigate dependence upon angle , leaving only polar and azimuth
angles and , leaving = 0 . This is our model suggestion. So, we have
r
xi r r 3 j ( k ) r r (1.24)
Sn (m + k ) = ji (k )Sn (m + k ), x1 x, x2 y, x3 = z.
j =1
When substitute Eq. (1.24) into Eqs. (1.10-1.12), we obtain
) 4 6 2 2 r 3 3 r r
H1ex = r I nl ( i ) ( k , i ) , (1.25)
m k =1 i =1 n =1 l =1 =1 =1
r
( k ) r r ( r k + r i ) r r r
Sn (m + k )Sl (m + k + i )
) 4 6 2 2 r
H 2ex = r Bnl ( i )
m k =1i =1 n =1l =1 (1.26)
3 r r ( k ) r r ( k + r i ) r r r
r r
2 ( k , i )S n (m+ k )Sl (m + k + i )
, =1
1+ ,
(S +1)(2S +1)
) 4 2 3 r r r r
H a = r An ( 3 ( k )Sn ( k ) (m + k )) 2 , (1.27)
m k =1 n =1 =1
where
r r 3 r r r (1.28)
( k , i ) = r ( k ) r ( k + i ).
r =1
Eqs. (1.9, 1.25-1.28) present the final form of our Hamiltonian.
1.3. Ground State Energies of Different Magnetic Configurations
Here we shall confine ourselves only to the case of absolute zero temperature. In order to
consider energies of ground states we shall use the formula
r
(k ) r r
Sn j (m + k ) = j 3 S n , n = 1,2 . (1.29)
Substituting Eq.(1.29) into Eqs.(1.9, 1.25-1.27), we obtain for the ground state energy E
the expression
E = N{a[ (12) + (34)] + b[ (13) + (24)] +

(1.30)
+ c[ 2 (1) + 2 (2) + 2 (3) + 2 (4)] +
+ 8Q[2( 1+ (12) + 1+ (34) ) B1 +
+ ( 1+ (13) + 1+ (24) )B2 ]},
where
a = 8[(1 x) 2 S12 I1 + x 2 S 22 I 3 + 2 x(1 x)S1S 2 I 5 ], (1.31)
b = 4[(1 x)2 S12 I 2 + x2 S22 I 4 + 2 x(1 x)S1S2 I6 ], (1.32)
c = (1 x)S12 K1 + xS22 K 2 , (1.33)
S +1 S (1.34)
Q = x(1 x) , = ,
2S +1 S +1
(i) = cos ,
i
(ij) = sini sin j cos(i j ) + cosi cos j ;i, j = 1,2,3,4. (1.35)
386 B. V. Karpenko
Indexes i, j number sub-lattices.

The energy of Eq. (1.30) is minimal at angles and , satisfying eight equations:
E E E E E E E E
= = = = = = = = 0. (1.36)
1 2 3 4 1 2 3 4
The system (1.36) has the following solutions.
A :1 = 2 = 3 = 4 = ;1 = 2 = 0,3 = 4 = . (1.37)
1
A1 :1 = 2 = 3 = 4 = ,1 = 2 = arccosh,3 = 4 = 1. (1.38)
2
1
A2 :1 = 2 = ,3 = 4 = ,1 = 2 = 3 = 4 = arccos(h1) . (1.39)
2
B :1 = 2 = 3 = 4 = ,1 = 2 = 3 = 4 = 0 . (1.40)
B:1 = 2 = 3 = 4 = ,1 = 2 = 3 = 4 = / 2 . (1.41)
C :1 = 2 = 3 = 4 = ,1 = 3 = 0,2 = 4 = . (1.42)
1
C1 :1 = 2 = 3 = 4 = ,1 = 3 = arccosg,2 = 4 = 1 . (1.43)
2
1
C2 :1 = 3 = ,2 = 4 = ,1 = 2 = 3 = 4 = arccos( g1) . (1.44)
2
G : = = = = , = = 0, = = . (1.45)
1 2 3 4 1 4 2 3
1
G1 :1 = 2 = 3 = 4 = ,1 = 4 = arccosh2 ,2 = 4 = . 1
(1.46)
2
1
G2 :1 = 4 = ,2 = 3 = ,1 = 2 = 3 = 4 = arccos( g2 ) . (1.47)
2
(Angle is arbitrary. Here and below symbols A, B, C , G introduced by Wollan and

Koehler in Ref. [5] are used in order to denote spin configurations.) The following notations
are input above:
2 ,
1 f (1.48)
h = 1
b c

2
1 f , (1.49)
h1 = 1
b+ c

1 d 2 , (1.50)
g = 1
a c

2
1 d , (1.51)
g1 = 1
a + c

2
1 d + f , (1.52)
h2 = 1
a +b c

2 , (1.53)
1 d + f
g2 = 1
a +b + c

Where
8Sx(1 x) B1 , (1.54)
d=
(S +1)(2S +1)
4Sx(1 x)B2 . (1.55)
f=
(S +1)(2S +1)
We have for energies of the corresponding states = E / 4 N
1 B2
( )

, (1.56)
A = a b + 2c +8 2B1 + x 1 x
2
2S +1

388 B. V. Karpenko
1 S +1 2S +1 16B22 S
A = a b+ c +16B1 x(1 x) x 2 (1 x) 2 ,
1 2 S S (2S +1)(bc)
b c < 0, (1.57)

1 S +1 1 16SB22 2
A = a b c +16 B1 x(1 x) 2
x (1 x) ,
2 2 S S (2S +1)(b + c)

b + c < 0, (1.58)
B = 1 a +b + 2c +8(2B1 + B2 )x(1 x) , (1.59)

2
B = a +b+8(2B1 + B2 )x(1 x) ,
1 (1.60)
2
1 2B1
, (1.61)
C = a +b+ 2c +8 + B2 x(1 x)
2
2S +1

1 S +1 2S +1 64SB12
C = a +b+ c x 2 (1 x) 2 +8B2 x(1 x),
1 2 S S (2S +1)(a c)
a c < 0, (1.62)

1 S +1 1 64SB12 2 2
C = a +b c x (1 x) +8B2 x(1 x),
2 2 S S (2S +1)(a + c)

a + c < 0, (1.63)
1 8
(2B1 + B2 )x(1 x), (1.64)
G = a b + 2c +
2 2S +1
2
1 S +1 S +1 2S +1 16S (2B1 + B2 ) 2
G = a b+ c x (1 x) 2 ,
1 2 S S S (2S +1)(a +bc)

a + b c < 0, (1.65)
c c
A B
3 3
3 4 3 4
3 3 3 3
1 1 b 1 1 b
2 2
1 1
1 1
a a
c
B
3
3 4
3 3
1 1 b
2
1
1
c a c
C G
3 3
3 4 3 4
3 3 3 3
1 1 b 1 1 b
2 2
1 1
1 1
a a
Figure 1.2. Collinear magnetic configurations A, B, B , C , G . Arrows denote the site spins.
390 B. V. Karpenko
A1, C1, G1

1, 1, b
A2, C2, G2
Figure 1.3. The ordering of sub-lattice moments in angle configurations A1 , C1 , G1 and A2 , C 2 , G 2 .

200
100
x
0
0 0,2 0,4 0,6 0,8 1
,K
-100
A
B
-200
B'
C
G
-300
-400
Figure 1.4. Energies (x) for collinear structures A, B, B , C , G .
x
-260
0,14 0,16 0,18 0,2 0,22 0,24
-270
-280
-290
,K
-300
-310
-320
Figure 1.5. Energies ( x) for non-collinear structures A1 (thin line) and A2 (thick line).
392 B. V. Karpenko
1 S +1 S +1 1 16S (2 B + B ) 2
G = a b c 1 2
x 2 (1 x ) 2
2
2 S S S (2S + 1)(a + b + c ) , (1.66)
a + b + c < 0.
Thus we have obtained five collinear magnetic configurations two of which ( B and B )
are ferromagnetic while three are antiferromagnetic ( A, C and G ), and also six non-
collinear structures ( A1 , A2 , C1 , C2 , G1 , G2 ). In the ferromagnetic phase B vector of
magnetization is directed along the easy axis b , while in ferromagnetic phase B vector of
magnetization is perpendicular to axis b . The collinear phases A, B, B, C , G are depicted in
Figure 1.2. In A state each spin is surrounded by four parallel to it spins and two antiparallel.
In C state the surrounding consists of two parallel and four antiparallel spins. In
configuration G all six nearest neighboring spins are antiparallel to the central spin. The
figure shows that in B state spins are parallel to c -axis, however this is not necessarily
they must only lie in the ( ac) -plane. In non-collinear phases A1 , C1 , G1 vectors of spin sub-
lattices are situated symmetrically with respect to the b -axis normal, while in configurations
A2 , C2 , G2 the sub-lattices spins are situated symmetrically with respect to b -axis as it is
shown in Figure 1.3. On this Figure is the angle of non-collinearity (the angle of canting),
and is the angle between sub-lattice vectors.
The energies in Eqs. (1.56-1.66) as the functions of concentration x are shown in
Figures 1.4, 1.5, 1.6, 1.7 at certain values of interaction parameters (see the next section).
1.4. Comparison with Experiment
Numerous experiments demonstrate that with increase of x in La1 x Ca x MnO3

magnetic configurations change each other. Being guided the results of Refs. [1, 2, 5, 30-38]
we can suppose the following scheme of magnetic transitions when x changes from 0 till 1 :
A A1 A2 B C2 C1 C G . (1.67)
The following equations should be held in the transition points

-290
0,45 0,5 0,55 0,6 0,65 0,7 0,75
-300
-310
-320
,K
-330
-340
-350
-360
Figure 1.6. Energies (x) or non-collinear structures C1 (thin line) and C2 (thick line).
x
-210
0,55 0,6 0,65 0,7 0,75 0,8
-230
-250
,K
-270
-290
-310
-330
Figure 1.7. Energies (x) or non-collinear structures G1 (thin line) and G2 (thick line).
394 B. V. Karpenko
A ( x1) = A ( x1) , (1.68)

1
A ( x2 ) = A ( x2 ) , (1.69)
1 2
A ( x3) = B ( x3) , (1.70)

2
B ( x4 ) = C ( x4 ) , (1.71)
2
C ( x5) = C ( x5) , (1.72)

2 1
C ( x6 ) = C ( x6 ) , (1.73)
1
C ( x7 ) = G ( x7 ) . (1.74)
Substituting energies of Eqs.(1.56-1.66) in Eqs.(1.68-1.74), then we obtain
4B2 Sx1(1 x1) (1.75)

b( x ) c( x ) + = 0, A A1.
1 1 2S +1
16S 2 B22 x22 (1 x ) 2 (1.76)
b ( x2 ) c ( x2 )
2 2 2 = 0, A A .
1 2
(S +1)(2S +1)
4SB2 x3 (1 x3 ) (1.77)
b( x3) + c( x3) + = 0, A2 B.
2S +1
8SB1 x4 (1 x4 ) (1.78)
a( x4 ) + c( x4 ) + = 0, B C2
2S +1
64 S 2 B12 x52 (1 x5 ) 2 (1.79)

a 2 ( x5 ) c 2 ( x5 ) = 0, C2 C1
(S +1)(2S +1)
8B Sx (1 x6 ) (1.80)
a( x6 ) c( x6 ) + 1 6 = 0,C1 C.
2S +1
1 (1.81)
b( x7 ) + 4 B2 1 x (1 x7 ) = 0, C G.

2S +1 7
0 0,2 0,4 0,6 0,8 1

-100
-150
-200
,K
-250
-300
-350
-400
Figure 1.8. Energy (x) , consisting of energies A (x) , A ( x) , A ( x) , B (x) , C ( x) ,

1 2 2
C ( x) , C (x)
1
and G (x) . Vertical lines are drawn through transition points x1 - x7 .
So we have seven Eqs. (1.75-1.81) with seventeen unknowns: ten parameters I1 I 6 ,

K1 , K 2 , B1 , B2 and seven transition concentrations x1 x7 . In order to use the system of
Eqs. (1.75-1.81) in practice we should determine some ten values independently, while the
missing seven ones we shall find solving the system of seven Eqs. (1.75-1.81). Let us do it in
the following way. The values I1 = 9.6 K , I 2 = 6.7 K , K1 = 1.92 K were obtained in
Ref. [30] during the analysis of spin-wave spectrum in LaMnO3 . We arbitrarily (not having
experimental data) assumed K1 = K 2 = 1.92 K . If one uses data on Nel temperature for [5,
38] we shall put approximately I 3 = I 4 = 8.7 K . Keeping in mind the results of papers [5,
34-38] lets put some mean
experimental values for transition concentrations
x3 = 0.2, x4 = 0.5, x6 = 0.7, x7 = 0.85 .When substitute these ten parameters
I1 , I 2 , I 3 , I 4 , K1 , K 2 , x3 , x4 , x6 , x7 into the system of Eqs. (1.75-1.81), we have obtained for
the missing parameters the following values:
I5 = 12.673K , I 6 = 9.603K , B1 = 144.571K , B2 = 116.085K , (1.82)

396 B. V. Karpenko
45
40
35
30
25
1,
20
15
10
0
0 0,2 0,4 0,6 0,8 1
x
Figure 1.9. Dependence of angle 1 upon concentration x .
x1 = 0.173, x2 = 0.197, x5 = 0.55 . (1.83)
Using these values we have depicted plots of energies in the whole range of x values in
Figures 1.4-1.7. Figure 1.8 shows the continuous curve (x) , made up of the pieces of
functions A , A , A , B , C , C , C and G , closing each other consecutively in the
1 2 2 1
transition points x1 x7 . Figure 1.9 shows the dependence 1 ( x) . Figure 1.10 depicts the
dependence of angle between sub-lattices moments upon x , Figure 1.11 is the plot of
canting angle (x) , while Figure 1.12 presents the concentration dependence of
ferromagnetic momentum (on one ion of manganese ) m = 2 B S F , where B is Bohr
magneton and S F is given by formula
1
S F = (4 x) cos . (1.84)
2 2
180
150
120
,
90
60
30
0
0 0,2 0,4 0,6 0,8 1
x
Figure 1.10. Dependence of angle between sub-lattices vectors upon concentration x.
90
80
70
60
,
50
40
30
20
10
0
0 0,2 0,4 0,6 0,8 1
x
Figure 1.11. Theoretical dependence of canting angle upon concentration x.

398 B. V. Karpenko
3,5
2,5
2
m, B
1,5
0,5
0
0 0,2 0,4 0,6 0,8 1
x
Figure 1.12. Theoretical dependence of ferromagnetic momentum m upon concentration x.
Functions (x) for collinear states A, B, C , G have positive second derivatives

2 / x 2 in the whole range 0 x 1 . The energies of non-collinear states
A1 , A2 , C1 , C 2 , G1 , G2 are defined each in the own region of definition
A : 0.173 x 0.239, (1.85)

1
A : 0.147 x 0.2, (1.86)

2
C : 0.474 x 0.7, (1.87)

1
: 0.5 x 0.721, (1.88)

C2
G : 0.594 x 0.786, (1.89)

1
G : 0.617 x 0.797. (1.90)

2
The states A1 , A2 , G1 , G2 have negative second derivative in their definition regions. The
state C1 has negative second derivative for x < 0.515 and positive one for x > 0.515 .
Respectively for C2 the point of inflection for (x) is equal to 0.54. Thus in the range
x5 x x6 for C1 the second derivative is positive, while for C2 the second derivative
changes sign from negative to positive with the increase of x in the range x4 x x5 .
It was experimentally shown in Ref. [34], that transition from antiferromagnetic phase A
at small x with increase of concentration to ferromagnetic state with 1 = = 90o (in our
terminology B ) occurs rather smoothly (see Figure 1.13 [34] which should be compared
with our Figure 1.11) and only through the sequence of one-type angle states (analog of our
configuration A1 ). So we see that the results of our theory and this experiment differ
essentially: state B does not realize in our theory at all and two intermediate angle states A1
and A2 exist between antiferromagnetic state A and ferromagnetic state B ( = 0 ).
Besides the range of transition between anti- and ferromagnetic states in Ref. [34] is much
broader than we have obtained. We do not have enough experimental data about the
dependence of canting angle upon concentration (x) for values x 0.5 in order to
compare in details theory and experiment. Only Ref. [35] informs about angle magnetism at
x = 0.667 .
Our theoretical curve for m (see Figure 1.12) can be compared with experimental data
from Refs. [1, 5] that are depicted on Figure 1.14. The type of the dependence is
approximately the same while the qualitative coincidence is not good enough.
90
75
60
,
45
30
15
0
0 0,05 0,1 0,15 0,2 0,25
x
Figure 1.13. Experimental dependence of angle upon concentration in La1 xCax MnO3 . The plot
is taken from Ref. [34].
400 B. V. Karpenko
3,5
2,5
m, B 2
1,5
0,5
0
0 0,1 0,2 0,3 0,4 0,5 0,6 0,7
x
Figure 1.14. Experimental dependence of ferromagnetic momentum m per one Mn ion upon
concentration x for La1 xCa x MnO3 . (See Figure 12 of Ref. [5]). Clear circles are data of Ref.1,
while dark ones were taken from Ref. [5].
1.5. Some Discussions
The proposed model of homogeneous dynamical spin alloy for compound

La1 xCa x MnO3 has in principle allowed to explain the different magnetic configurations
observed experimentally and the sequence of their change depending on the concentration x .
The numerical values of interaction parameters obtained theoretically are rather plausible.
Taking account of simultaneous presence of Heisenberg-type direct or superexchange and
non-Heisenberg-type double exchange leads to the appearance of non-collinear (canted)
structures. Heisenberg interaction can be both positive (ferromagnetic) and negative
(antiferromagnetic). Double exchange is essentially positive. Canted structures appear at
values of concentration x , when balance takes place between the total Heisenberg
antiferromagnetic exchange interaction and ferromagnetic double exchange. Heisenberg
interaction and double exchange taken separately cannot lead to canted configurations.
Presence of anisotropy energy (which is small) leads to the fact that angle states
symmetrical with respect to the normal of b -axis turn at certain x values into states
symmetrical with respect to b -axis. At finite values of anisotropy energies ferromagnetic
state B , which is the limit case of angle states ( = 90 ), symmetrical with respect to
o
normal of b -axis turns out to be unattainable. Ferromagnetic state B always has lower
energy than that of ferromagnetic state B and their difference is just anisotropy energy.
Lets note that the canting angle (x) has infinite derivatives with respect to x in the points
of transitions between collinear and non-cllinear structures ( x1 , x3 , x4 , x6 ).
Lets note that magnetic phase transitions depending upon concentration in systems with
double exchange were investigated earlier by many authors with the use of band
approximation which was proposed by de-Gennes in his famous work [7]. The typical feature
of this approximation is the proposal that additional electron realizing exchange coupling
between localized spins is the conduction electron whose transfer integral depends upon
mutual orientation of localized electrons spins. The band description is used for it so it is
natural to call such an approach to double exchange theory the band approximation of double
exchange. Then the approximation used in our paper can be relevantly called the localized
approximation of double exchange though the conditionality of such a title is obvious. We did
3+
not use the concept of bands and only different spins S + 1 / 2 and S for ions Mn and
Mn 4 + are present in our approach. (However this does not mean that our model forbids
electronic conductivity. The conductivity is hiding here in probabilities x and 1 x which
we have introduced. Of course they allow electron migration from site to site. It seems that in
our approach its more proper to speak about not band but jump conductivity.) The difference
3+ 4+
between ions Mn and Mn disappears in band model and correspondingly different
3+ 4+
superexchange parameters between ions Mn and Mn are absent while in our approach
these parameters are present separately. This essential difference does not allow compare
literally the band and localized approximations.
The following moments should be associated with difficulties of our theory. The
negativity of the second derivative for angle states A1 , A2 and partly C2 indicates non-
stability of these states. This question needs an additional investigation. Its impossible to
obtain phenomenon of charge ordering in our approach because we are using homogeneous
approximation. In connection with this we have simplified the situation supposing that only
configurations C1 ,C2 or C are present in the range 0.5 < x < 0.85 while experiment shows
that CE states are also present in this range of x . We have called the object under
investigation hypothetical compound La1 x Ca x MnO3 because of the fact that really at x
increasing from 0 till 1 both the lattice symmetry change ( LaMnO3 has orthorhombic
symmetry while CaMnO3 is cubic) and the lattice parameters hence the exchange parameters
I , B, K also change. Unfortunately the dependence of interaction parameters upon x is
unknown. It is of course inconsistently to assume parameters I1 and I 2 to be different while
I 3 and I 4 equal ones as we have done above at numerical evaluations. Thus numerical
values for I , B, K obtained above are approximate and averaged.
402 B. V. Karpenko
PART 2. SPIN-WAVE EXCITATIONS

The content of this part is based on the papers [30-34, 41-59].
2.1. Some General Formulas. Ferromagnets
After considering the some magnetic configurations at zero temperature and

concentration transitions between them in Part 1 then the magnon spectrum will be studied
now. Many papers were devoted to this problem [30-34, 41-46]. As before well assume that
there are two kind of exchange in a system Heisenberg and double. The energy operator for
conducting ferromagnetic manganites is
) ) ) )
H = H1ex + H 2ex + H a , (2.1)
where the first term represents Heisenberg interaction, the second term stands for the double
exchange operator, and the third term is the anisotropy energy. These are expressed as
) 2 r r r r r r
H1ex = I kl ( ) S k (m) Sl (m + ), (2.2)
r r
m k , l =1
r r r r r
) 2 r 2 S k (m ) S l ( m + )
H 2ex = Bkl ( ) 1 + , (2.3)
r r
m k , l =1 1 + S (2S + 3)
) 2
r
H = Ak ( S kz (m)) 2 , (2.4)
r
m k =1
respectively. The sense of the designations is the same as in the Part 1.
We then come from spin operators to second quantization operators in a standard manner,
expand the square root in Eq. (2.3) in a bilinear operator form, and, finally, perform the
Fourier transform. As a result, we obtain the following expression for the spin wave energy:
r n ~
[ r
]
E (q ) = 2 z p J p p (q ) J p + A, (2.5)
p =1
r
where q is the spin-wave quasi-momentum, subscript p numbers n coordination shells,
z is the number of sites in the p -th shell,
p
r zp r r (2.6)
p (q) = exp(iq j p ),
j =1
1
J p = (1 x)2 (S + )(11I p ) + x2S (22I p ) +
2 (2.7)
Bp
+ x(1 x) 2S (2S +1) 12 I p + 2
,
(2 S + 1)
~ Bp
J p = J p + f (S ) x(1 x) 12 I p + ,
(2S +1) 2
1 (2.8)
f (S ) = 2S + 2S (2S +1) ,
2
and
1 (2.9)
A = (1 x)(S + ) K1 + xSK 2.
2
To derive Eq. (2.5), we used the relations
r rr n r (2.10)
I
r 11( ) exp( i q ) = (1 x)2 11I p p (q ),
p =1
r rr n r (2.11)
r I 22() exp(iq) = x I p p (q),
2 22
p=1
r rr n r
r I12 () exp(iq) = x(1 x) I p p (q ),
12 I12 = I 21, (2.12)
p =1
r rr n r
r B12 () exp(iq) = x(1 x) B p p (q ),
p =1
B12 = B21, B11 = B22 = 0. (2.13)
A1 = (1 x) K1, A2 = xK2. (2.14)
Due to symmetry of the system, we assumed that all integrals of the same type inside the
same coordination shell are equal to each other. Therefore we introduced for the new
designations I ij .
p
404 B. V. Karpenko
Table 1.
p r
zp j
p j
p p
1 6 (a 00),(0a0),(00a) a 1 2(cos aqa + cos aqb + cos aqc )
2 12 (aa 0), (a0a ), (0aa), a 2 4(cos aqa cos aqb +
( aa0),(a0a),(0 aa) + cos aqa cos aqc +
+ cos aqb cos aqc )
3 8 (aaa),(aa a), a 3 8 cos aqa cos aqb cos aqc
(a aa),(aaa)
4 6 (2a 00), (02a 0),(002a ) a 4 2(cos2aqa + cos 2aqb + cos 2aqc )
6
E, meV
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4

q(-1)
Figure 2.1. Dispersion curves for the (circles) [001], (triangles) [110] and [101] and (squares) [111]
directions borrowed from [44]. The solid curves were calculated Eqs. (2.15)-(2.17).
2.5
1.5
1
J ,m e V
0.5
0
1 1.2 1.4 1.6 1.8 2
-0.5
-1
-1.5
r/a
Figure 2.2. Dependences of the exchange integrals of the reduced distance r/a (a is the lattice
parameter).
As will be explained below, we consider the case of four coordination shells. The main
parameters for this situation are given in the table (where a is the lattice parameter).
2.2. Comparison with Experiment. Ferromagnets
First we compare above theoretical formulae with experiment from paper [44] for
La0.83Ca0.17 MnO3 . Dispersion curves for spin waves obtained by neutron inelastic studies
for three principal crystallographic directions are shown in Figure 2.1. We now compare them
with our theoretical calculations and have
E 100 (q) = a1 + b1 cos(aq) + c1 cos(2aq), (2.15)
aq 2aq (2.16)
E 110 (q) = a2 + b2 cos + c2 cos ,
2 2
aq 2aq 3aq (2.17)

E 111 (q) = a3 + b3 cos + c3 cos + d3 cos ,
3 3 3
where
a1 = 2 x1 + 2 x2 + 2 x4 + x5,
b1 = x1 + 4 x2 + 4 x3, c1 = x4, (2.18)
a2 = x1 + x2 + 2 x3 + x4 + x5,
b2 = 2 x1 + 4 x2, c2 = x2 + 2 x3 + 2 x4 , (2.19)
a3 = 3x2 + x5, b3 = 3x1 + 3x3,

c3 = 3x2 + 3x4, d3 = x3, (2.20)
~ ~ ~ ~
x1 = 4 J1, x2 = 4 J 2 , x3 = 4 J 3, x4 = 4 J 4 , (2.21)
~
x5 = 2 z p J p + A .
4
(2.22)
p=1
406 B. V. Karpenko
100
LCMO25
80
Energy, meV 60
40
20
(1/2, 1/2, 0) (0, 0, 0) (1/2, 0, 0) (1/2, 1/2, 0)(1/2, 1/2, 1/2)(0, 0, 0)
Figure 2.3. LCMO25. Dispersion curves for the main crystallographic directions. Circles
experimental points from paper [46]. Thin line is calculated curve in [46], thick line is calculated by
Eqs. (2.27).
100 LCMO30
80
Energy, meV
60
40
20
0
(1/2, 1/2, 0) (0, 0, 0) (1/2, 0, 0) (1/2, 1/2, 0) (1/2, 1/2, 1/2) (0, 0, 0)
Figure 2.4. LCMO30. Dispersion curves for the main crystallographic directions. Circles
experimental points from paper [46]. Thin line is calculated curve in [46], thick line is calculated by
Eqs. (2.28).
5
x
4 0,15 0,2 0,25 0,3 0,35
0
3
J1,
2 -0,5
J2,
1
-1
0
0,15 0,2 0,25 0,3 0,35
x -1,5
0,8 0,8
0,6
0,6
0,4
J4,
J3,
0,4
0,2
0,2
0
-0,2 0
0,12 0,17 0,22 0,27 0,32 0,15 0,2 0,25 0,3 0,35
x x
Figure 2.5. Dependences of the exchange integrals on the concentration x.
To find the unknowns x1 x5 we use the least squares method; to this end, we set up the
functional
F = (Etheor Eexp )2 + (Etheor Eexp )2 +

100 110 . (2.23)
+ (Etheor Eexp )2
111
We impose the minimum conditions
F F F F F (2.24)
= = = = = 0,
x1 x2 x3 x4 x5
on this functional and obtain the set of five equations for five unknowns.The lattice parameter
is taken to be the experimental value ( a = 3.89 , see below), and we have (in meV)
J1 = 2.405, J 2 = 1.088, J 3 = 0.605, J 4 = 0.384, x = 17.19. (2.25)
408 B. V. Karpenko
The calculated dispersion curves are presented in Figure 2.1. Figure 2.2 shows the
distance dependence of exchange integrals J n . The energy gap is expressed as
4 Bp (2.26)
= 2 A + 2 f (S ) x(1 x) z p 12 I p + .
p =1

(2S +1)
2
At the numerical values of the parameters from Eq. (2.25), this gap is 0.136.
Let us discuss the approximations and results obtained.
First, we should explain why we took into account interactions up to fourth nearest
neighbors (i.e., neighbors from the fourth coordination shell). Indeed, as follows from the
experimental curve for the [100] direction (Figure 2.1), the dispersion curve has an inflection
point at q < 0.4 , whereas the nearest-neighbor approximation with the cos(aq) can give an
inflection only at q > 0.4 (at a = 3.89 ). The required inflection point can be induced by
the cos(2aq ) harmonic. However, the next and next-next nearest neighbors cannot generate
this harmonic, and it appears only when fourth-nearest neighbors are taken into account.
Second, we used in spite of the fact that the compound under study has an orthorhombic
structure. However, such compounds are often approximately considered as cubic or
pseudocubic crystals because of numerical relations between their lattice parameters. The
compound under study was taken to be cubic, and this approximation should inevitably affect
the accuracy of our calculations.
Third, the lattice parameter was assumed to be 3.89 which is average of the values of
parameters a, b and c at low temperatures for La0.83Ca0.17 MnO3 (see [44], Figure 4).
Fourth, when analyzing the dispersion curves, we can only determine total exchange
ij
integrals J1 , J 2 , J 3 , and J 4 rather than I p and B p separately. It is still unclear from
which experiments these exchange components could be determined.
Fifth, note that, in our model an energy gap in spin-wave spectrum appears due to both
anisotropy and (quadratic and double) exchange energies, which is caused by the presence of
ions with different ( S , S + 1 / 2 ) spins in the system.
Sixth, La0.83Ca0.17 MnO3 is a semiconductor, and a tendency to the metallic temperature
dependence of electrical resistivity appears only in the vicinity of the Curie temperature.
Compounds with 0.125 < x < 0.220 are attributed to a precursor metallic phase [44]. The
compound La0.83Ca0.17 MnO3 can serve as a basis for analyzing the dependence of the
exchange parameters on concentration x in the range of more metallic compounds. The
approach proposed in this work is suitable for any x provided that charge ordering is absent
in the system.
Seventh, the system La0.83Ca0.17 MnO3 is more complex that than it is assumed here. In
reality, Hennion and Moussa [44] showed that dispersion law for the lower curve in Figure
2.1 is only valid for the [001] direction and that experimental points in the range q 0.125
in the [100] direction do not fall on the general curve. This behavior cannot be described in
terms of our model.
The next objects of our attention will be the results obtained in the papers [45, 46]. Ye et
al studied by the method of inelastic neutron scattering spin-wave dispersion curves in
ferromagnetic metallic compounds with x = 0.25 and x = 0.30 . It considered as cubic.
Their data for different directions in Brillouin zone are represented in Figure 2.3 and Figure
2.4. We again will compare these experimental results with our theoretical calculations. As
above the presence of Heisenberg and double exchanges are assumed to be realize in these
compounds. So, we can use the formulae Eqs. (2.1-2.14). According the data of [45, 46] well
assume that anisotropy energy is absent ( A = 0 ). Using the standard square-mean method we
obtain in four-shell coordination approximation and for the directions [100] + [110] + [111] :
J1 = 4.402, J 2 = 0.097, J 3 = 0.068, J 4 = 0.3, J = 26.44 (2.27)
for LCMO25 and
J1 = 3.635, J 2 = 0.087, J 3 = 0.081, J 4 = 0.586, J = 23 .755 (2.28)
for LCMO30. The values are given in meV-units. The calculated curves are represented in
Figure 2.3 and Figure 2.4 by solid lines.
The authors of [45, 46] also treated their data from theoretical point of view to obtain the
numerical magnitudes of exchange parameters. Their calculated curves are given in Figure
2.3 and Figure 2.4 by thin lines. They obtained
I1 = 2.088, I 2 = 0, I3 = 0, I 4 = 0.136 (2.29)
for LCMO25
and
I1 = 1.719, I 2 = 0, I3 = 0, I 4 = 0.335 (2.30)
for the compound LCMO30.

The additional explanations are needed to understand the difference in the approaches
our and in [45, 46].
Although the authors of [45, 46] claimed, that they use the double exchange interaction,
as a matter of fact usual Heisenberg spin-operator was treated. Contrary, in our approach
Heisenberg and double exchanges are present in evident form.
3+
In order to take into account the existence of two different manganese ions Mn (spin 2)
4+
and Mn (spin 3/2) the notion of average spin was introduced in [45, 46]. In our approach
two different spins S and S + 1 / 2 present separately without any averaging.
In contrast to our approach the authors of [45, 46] suppose J 2 = J 3 = 0 in advance.
In our study the parameters J1 , J 2 , J 3 and J 4 have an internal structure, whereas these
merely represent certain effective quantities in [45, 46].
410 B. V. Karpenko
Our magnitudes J p are the analogues of I p from [45, 46], but literally it is impossible to
compare due to their different physical meaning.
Note, that really interactions with second and third coordination shells are small in
comparing with first and fourth ones, as supposed the authors of [45, 46]. However since the
number of second order neighbors is equal 12 and ones of third order 8 as to compare with 6
neighbors of fourth order then total contribution from second and third order may be
comparable with fourth order contribution in magnetic behavior.
Let us tell about one difficulty which appears at numerical estimations of exchange
parameters. The total ferromagnetic interaction the measure of which may serves the
magnitude
4 ~
J = zpJ p (2.31)
p =1
appears larger for LCMO25 then for LCMO30. This is in contradiction with the fact that
Curie temperature of LCMO25 is lower then one of LCMO30. The same situation in [45, 46]:
I1 + I 4 and average spin are larger for LCMO25 then for LCMO30. Hence the problem of
concordance of low-temperature and high-temperature properties demands additional
investigation.
At last using results for three compounds with x = 0.17,0.25,0.30 we obtain the
dependence of exchange integrals from concentration x . These dependences are shown in
Figure 2.5. Take notice the nonmonotonic dependence of J1 and J 4 and also large magnitude
of J 2 and J 3 in LCMO17 as compare with LCMO25 an LCMO30. These facts need
additional consideration also.
2.3. Some General Formulas. Antiferromagnet
Above we considered spin-wave spectrum in the ferromagnetic conducting compounds

LCMO17, LCMO25 and LCMO30. Now we will study magnon energy in antiferromagnetic
dielectric LaMnO3 . Naturally the mechanism of double exchange is absent here. Only the
three-valence manganese ions with spin of 3 / 2 are existed in this system. The crystal is
orthorhombic, space group no. 62, GdFeO3 structure type. The unit cell is shown in Figure
1.1. We are interested in the antiferromagnetic phase A , which is shown in Figure 1.2
together with other magnetic structures. The phase A consists of ferromagnetic (ab) planes
with the total magnetic moments directed alternately along and opposite to the axis b . The
antiferromagnetic structure is a two-sublattice system. One sublattice includes the first and
second Bravais lattices, and the other sublattice includes third and fourth Bravais lattices.
Most frequently, an analysis of the exchange interactions has been performed with
allowance made only for the interactions of the nearest neighbors, even though this inclusion
can prove to be insufficient. On the other hand, it is impossible to predict how many
neighbors should be taken for describing the situation adequately. Let us, fairly arbitrarily,
choose the interactions between eighteen neighbors from four coordination shells. In this
case, we focus our attention only on the interstitial distances, even though it is known that the
direction of the neighbor can play a decisive role. The position vectors of the neighbors can
be represented as follows:
r r
1 = 2 = (0, a 2,b 2) ,
r
3 = 4 = (0, a 2,b 2), (2.32)
r r
5 = 6 = (0, a,0), (2.33)
r r
7 = 8 = (0,0,b), (2.34)
r r
9 = 10 = (c 2,0,0), (2.35)
r r
11 = 12 = (c 2, a 2,b 2),
r r
13 = 14 = (c 2, a 2, b 2),
r
15 = 16 = (c 2, a 2,b 2),
r r
17 = 18 = (c 2, a 2,b 2). (2.36)
In these formulas, a, b and c are the lattice constants. According to [30], these lattice
constants are (in )
a = 5.5333,b = 5.7461, c = 7.6637. (2.37)
Then, the magnitudes of the vectors from relationships (2.32)-(2.36) can be written as
r r r r (2.38)
1 = 2 = 3 = 4 = 3.9886,
r (2.39)
5 = 6 = 5.5333,
r r (2.40)
7 = 8 = 5.7461,
r r (2.41)
9 = 10 = 3.8319,
r s r r r r r r (2.42)
11 = 12 = 13 = 14 = 15 = 16 = 17 = 18 = 5.531.
412 B. V. Karpenko
The next neighbors are at a distance of 6.7306; however, we ignore them in our
consideration.
In terms of symmetry, here, we have five different parameters of the exchange
interaction:
r r
I for the vectors , (2.43)
1 1 4
r r
I for the vectors , , (2.44)
2 5 6
r r
3 7 8
r r
4 9 10
r r
I for the vectors . (2.47)
5 11 18
In addition to these five parameters, we also introduce the anisotropy constant D , which
orients antiferromagnetic axis along the crystallographic direction b .
As was said above we assume that the exchange interaction is described by the
Heisenberg Hamiltonian and that the anisotropy energy has single-ion nature. Then, we have
) 2 18 r r r r r r r r 2
(2.48)
H= r I (m i
, mi
+ j
) S (mi
) S (mi
+ j
) D zr ) 2 ,
r (S m
i =1 j =1mi i =1mi i
r
where i denotes the sublattice number and mi runs over N -values of radius-vectors of sites
for each of sublattices. It is assumed that D > O Further, following the routine of passing
from the spin operators to the second quantization operators and performing the
diagonalization by the Fourier transform, we obtain the following expression for the magnon
r
energy w as a function of the reduced quasi-momentum q :
r
w(q ) = t12 t22 , (2.49)
t1 = 2S[ I1( 1 4) + I 2 ( 2 2) + I3( 3 2) + 2I 4 + 8I5 D] , (2.50)
t2 = 2S[ I 4 4 + I5 5 ] , (2.51)
9
8
7
w , THz 6
5
4
3
2
1
0
1,5 1,6 1,7 1,8 1,9 2 2,1 2,2 2,3 2,4 2,5 2,6 2,7 2,8 2,9 3
qy
Figure 2.6. Dispersion curve for the [100] direction. Experiment from [30].
4,5
4
3,5
3
w , THz
2,5
2
1,5
1
0,5
0
0 0,2 0,4 0,6 0,8 1 1,2 1,4 1,6 1,8 2 2,2 2,4 2,6 2,8 3
qy
414 B. V. Karpenko
1 = 4cosqy cosqz , (2.52)
2 = 2cos 2qy , (2.53)
3 = 2 cos 2qz , (2.54)
4 = 2cos 2qx , (2.55)
5 = 8cosqx cosq y cosqz . (2.56)
The expression for w is valid over entire Brillouin zone.
2.4. Comparison with Experiment. Antiferromagnet
Moussa et al. [30] investigated inelastic neutron scattering and obtained the dispersion
curves for principal crystallographic direction in LaMnO3 compound. Figures 2.6-2.8
present the experimental points. Now, we compare our results with these experimental data
for the directions [100],[110] and [001] . From expressions Eqs. (2.49) - (2.56), we obtain
1.4
1.3
1.2
1.1
1
w, THz
0.9
0.8
0.7
0.6
0.5
-3 -2.5 -2 -1.5 -1
qx
Figure 2.9. The directed lobes of wave functions. The plane picture. From [49].
I4
I5
b
I3
I2 I1
Figure 2.10. The directed lobes of wave functions. The volume picture.
w = (a1 + b1 cosqy + c1 cos2 qy + d1 cos3q y + f1 cos4 q y )1 2, (2.57)

[100]
w = (a2 + b2 sin 2 q y + c2 sin 4 q y )1 2, (2.58)

[110]
w = (a3 + b3 cos2 qx )1 2, (2.59)

[001]
where
416 B. V. Karpenko
a1 = 16[16(I1 + I2 )2 +
+ (2I 4 + 8I5 D)(2I 4 + 8I5 D 8I1 8I 2 ) 4I 42 ], (2.60)
b1 = 128[ I1(4I1 4I 2 + 2I 4 + 8I5 D) 4I 4 I5 ], (2.61)
c1 = 128[2I12 + I 2 (4I1 4I 2 + 2I 4 + 8I5 D) 8I52 ], (2.62)
d1 = 512 I1I 2 , (2.63)
f1 = 256I22, (2.64)
a2 = 16D[4(I 4 + 4I5 ) D], (2.65)
b2 = 128[( I1 + I 2 + I3 )(2I 4 + 8I5 D) + 4( I 4 + 4 I5 ) I5 ], (2.66)
c2 = 256[( I1 + I 2 + I3)2 4 I52 ], (2.67)
a3 =16(2I4 + 8I5 D)2, (2.68)
b3 = 64( I 4 + 4I5 )2. (2.69)
Here we have set S = 2 .
In order to find numerical values of the parameters I1 I 5 .and D , we employ the least-
square method by minimizing the standard deviation over these parameters and using all fifty
four experimental points for all three crystallographic directions simultaneously. The
resulting nonlinear of six equations in six unknowns has the following solution (in THz):
I1 = 0.588, I 2 = 0.010, I 3 = 0.011, I 4 = 1.226, I 5 = 0.271, D = 0.040. (2.70)
The dispersion curves theoretically calculated with these parameters are shown in Figures
2.6-2.8 by solid lines together with the experimental points. In order to estimate the error in
the calculation and the stability of the method, we assume that the relative experimental error
for the energy at any quasi-momentum has a moderate value of 10% (the error are not given
in [30]). Then, we scatter the experimental data (points in Figures 2.6 - 2.8) by 10% with a
random number generator and recalculate the values of I1, I 2 , I 3 , I 4 , I 5 and D . We have
repeated this procedure 20 times. Our parameters averaged over 20 sets are as follows:
I1 = 0.550, I 2 = 0.009, I 3 = 0.011,

I 4 = 1.142, I 5 = 0.250, D = 0.039. (2.71)
Determining the root-mean-square deviations, we find the following relative errors :
( I1) = 18%, ( I 2 ) = 59%, ( I3) = 32%,

( I 4 ) = 20%, (I5 ) = 23%, (D) =11%. (2.72)
The values of the parameters from solutions Eq. (2.70) are within the confidence
intervals.
2.5. Discussion of Results
Let us discuss the results. First, we compare our solution Eq. (2.70) with the results
reported in [30]. In [30], it is assumed, just as we do, that I1 > 0 and I 4 < 0 (here, we use
the notations accepted in present work). However, the authors of those studies allowed only
for the nearest neighbor interactions ( I1 and I 4 parameters), while we consider the next-
nearest neighbors ( I 2 , I 3 and I 5 ) as well. The performed in [30] mathematical treatment of
the experimental data (shown here in Figures 2.6-2.8) yields the following results (in THz):
I1 = 0.2, I 2 = 0.14, I 3 = 0.04. (2.73)
A comparison of these results with the values of solutions Eq. 2.70 demonstrates that
only the value of D is in agreement, while the values of I1 and I 4 differ very greatly from
solutions Eq. 2.70. Note that the authors of [30] indicated difficulties in applying their method
to the determination of the I 4 parameter: if its value is determined for the [001] direction, it
amounts to -0.14 THz, whereas the corresponding value determined from curves in the (ab)
plane is equal to zero.
Feiner and Ole [47, 48] also determined I1 and I 4 . They did not analyze the dispersion
curves but calculated I1 and I 4 from their theoretical model. They obtained 0.28 THz for I1
and -0.21 THz for I 4 . These values also disagrees with our data (solutions Eq. 2.70).
Reasons for disagreement of our results with those obtained in [30] are most likely as
follows: first, we have taken into account a large number of interaction parameters, and,
second, we have determined these parameters for all the three crystallographic directions
simultaneously rather than separately. As regard the discrepancies with results obtained in
[47, 48], we can only remark that the model microscopic calculation and the results of
mathematical treatment of dispersion curves have not reached a fairly good agreement.
418 B. V. Karpenko
It is evident from solutions Eqs. (2.70) and (2.71) that the exchange interaction is
significantly anisotropic. This fact agrees well with the orbital ordering phenomenon, which
was revealed in the LaMnO3 compound by resonant x-ray scattering in [49]. The
magnitudes of the exchange integrals are determined by both the distance between ions and
the mutual orientation of the 3d functions. We see in Figure 2.9, reproduced from the work by
Murakami et al [49] the direction of wave function is shown. In volume picture Figure 2.10 it
is seen that I1 and I 5 correspond to the mutual perpendicularity of the lobes of the wave
functions, whereas I 2 , I 3 and I 4 correspond to their parallelism. This is responsible for the
strong anisotropy of the exchange interaction. This effect is especially strong for I 2 and I 5
interactions: the distances are practically the same (5.5333 and 5.531 ), but they have
different signs and different absolute values.
Now, we note the following circumstance. The Nel temperature TN calculated from the
known formula of the molecular field theory with the use of our solutions Eq. (2.70) is equal
to 500 K, whereas the corresponding experimental temperature is 140 K [30]. In our opinion,
such a large difference can be explained as follows. First, the molecular field theory is known
to lead to overestimate values for the ordering temperature. Second, this theory should not be
applied to our case at all. Indeed, we introduce two quantities J1 and J 2 as follows:
J1 = 4 I1 + 2I 2 + 2I3, J 2 = I 4 + 4 I5. (2.74)
The quantity J1 is the total interaction in the ferromagnetic plane, and the quantity J 2 is
the interplanar interaction. Using solution Eq. (2.70) we obtain J1 = 2.354, J 2 = 0.142 ,
and y J 2 / J1 = 0.06 . Such a small value of the ratio y allows us to treat our system as a
quasi-two-dimensional magnet. In this case, the ordering temperature should be calculated by
a method that differs from the method used in the molecular field theory. In [50-53], the
approach based on the temperature Greens function method was developed to calculate the
temperatures TN and TC in layered quasi-two dimensional structures. It was shown that, as
the ratio between the interplanar exchange interaction parameter changes from 1 to 0,
ordering temperature changes from a maximal temperature to zero. Unfortunately, we cannot
use these calculations, because were performed for a particular case of S = 1 / 2 and included
only two interaction parameters. However, we can assume with confidence that a small value
of y in our case, undoubtedly, leads to a decrease in the calculated temperature TN
(possible, even a severalfold decrease).
For the parameters corresponding to solution Eq. (2.70), the calculated paramagnetic
Nel temperature N amounts to 400 K. One should not expect that this value can be
confirmed in the experiment. The point is that, in order to determine the Nel
temperature N , it is necessary to perform high-temperature measurements of the magnetic
susceptibility. However, an increase in the temperature results in a change in the lattice
parameters and occurrence of phase transitions. Therefore, there arises an uncertainty in the
determination of the Nel temperature N . Two paramagnetic Nel temperatures, i.e., the
low Nel temperature (75 K) and the high Nel temperature (205 K) were obtained in
[54-56].
The check of solution Eq. (2.70) for stability has shown that the average values Eq. (2.71)
are close to the initial values Eq. (2.70). The relative errors ( I 2 ) and ( I 3 ) are very large;
however, the values of I 2 and I 3 are small and do not make a substantial contribution to the
total exchange interaction. The exchange integrals I1 , I 4 and I 5 determine the magnetism of
the LaMnO3 compound.
Note, when we use the mean square method to determine the I1 I 5 , D , the very
complicated nonlinear system of equations was obtained. The procedure of solving is rather
cumbersome and therefore is not present here.
So, the performed analysis of the dispersion curves for spin waves in the LaMnO3
compound has demonstrated that the exchange interactions are characterized by a strong
anisotropy: the interactions are opposite in sign for the nearest neighbors and can both be
opposite in sign and differ in absolute value by many factors for the next-nearest neighbors.
Such anisotropy, which is caused by orbital ordering phenomenon, leads to the quasi-two
dimensionality of the magnet. By and large, the theoretical curves agree with experimental
data.
PART 3. MAGNETIZATION
In this part the high temperature magnetization in systems with double exchange will be
considered. The molecular field method will be used. Formally, this method is applicable at
all temperatures, but since at low temperatures it gives incorrect results, we call our approach
as a high-temperature one. As above, we will use our model of homogeneous spin alloy with
spins S1 = S + 1 / 2 for Mn3+ and S 2 = S for Mn 4 + where S = 3 / 2 . Only collinear
configurations will be considered here: one ferromagnetic B and three antiferromagnetic
A, C and G (see Figure 1.2). The nearest neighbor approximation will be used and two
interaction mechanisms Heisenberg and double exchange are to be present. The needed
coordinates of sites and operators of the exchange interactions are given in Part 1.
3.1. The Near Ordering Temperatures

)
If one expands the square root in Hamiltonian H 2ex (see Eqs. (1.11) and (1.26)) up to the
first degree of spin operator scalar product then in Weis molecular field approximation we
obtain
Slz (k ) = Sl BS l ( Z lk ), (3.1)
420 B. V. Karpenko
where
(3.2)
2S +1 2S +1 1 Z
BS (Z ) = coth Z coth
2S 2S 2S 2S
3+ 4+
is the Brillouin function. (As above l = 1,2 differs Mn and Mn ; k = 1,2,3,4 numerates
3+ 4+
the sublattices.) Let us introduce the following notations for spins of ions Mn and Mn ,
being in one of four magnetic sublattices:
S z1(1) = x1, S z1(2) = x3 , S1z (3) = x5, S1z (4) = x7 ,

z
S 2z (1) = x2 , S2z (2) = x4, S2z (3) = x6 , S 2 (4) = x8. (3.3)
Then we obtain the following system of equations for sublattices magnetizations
x1 = S1BS (Z11),
1
x2 = S2 BS (Z 21),
2
x3 = S1BS (Z12),
1
x4 = S2 BS (Z22 ),
2
x5 = S1BS (Z13),
1
x6 = S2 BS (Z 23),
2
x7 = S1BS (Z14),
1
x8 = S2 BS (Z 24 ). (3.4)
2
The following notations for molecular fields are accepted here:
2S1 ~ ~
Z11 = {(1 x)2[4I1x3 + 2I2 x5] + x(1 x)[4I5 + 2I6 x6 ]},
T
2S ~ ~
Z21 = 2 {x2[4I3 x4 + 2I4 x6 ] + x(1 x)[4I5 x3 + 2I6 x5 ]},
T
2S ~ ~
Z12 = 1 {(1 x)2[4I1x1 + 2I2 x7 ] + x(1 x)[4I5 x2 + 2I6 x8 ]},
T
2S ~ ~
Z22 = 2 {x2[4I3x2 + I4x8] + x(1 x)[4I5x1 + 2I6x7 ]},
T
2S1 ~
Z13 = {(1 x)2[4I1x7 + 2I2 x1] + x(1 x)[4I5 x8 + 2I6 x2 ]},
T
2S2 2 ~ ~
Z23 = {x [4I3 x8 + 2I4 x2 ] + x(1 x)[4I5 x7 + 2I6 x1]},
T
2S ~ ~
Z14 = 1 {(1 x)2[4 I1x5 + 2 I 2 x3 ] + x(1 x)[4 I5 x6 + 2 I6 x4 ]},
T
2S ~ ~ (3.5)
Z24 = 2 {x2[4I3x6 + 2I4 x4 ] + x(1 x)[4I5 x5 + 2I6 x3]},
T
where
~ B1 S +1 (3.6)
I5 = I5 + ,
1+ S (2S +3) 2S +1
~ B2 S +1 (3.7)
I6 = I6 +
1+ S (2S +3) 2S +1
To find the temperature dependence of spontaneous magnetization at temperature

approaching the ordering temperature from below lets expand all Brillouin functions at small
Z
(2S +1) 2 1 Z (2S +1) 4 1 Z

3
(3.8)
BS (Z ) = , Z << 1.
(2S ) 2 3 3(2S ) 4 45
The system of equations for sub-lattices magnetizations Eq. (3.4) is essentially simplified
if one chooses the concrete type of ordering.
B -structure
As it can be seen from Figure 1.2, at ferromagnetic ordering one has
x3 = x5 = x7 = x1, x4 = x6 = x8 = x2. (3.9)
As a result system Eq. (3.4) will consist of two equations
x1 = S1BS (Z11), x1 = S1BS1 ( Z11 )

1
x2 = S2 BS (Z 21), (3.10)
2
where
422 B. V. Karpenko
Z11 = x1 + x2 , Z 21 = x1 + x2, (3.11)
2S1 (3.12)
= (1 x)2 (4I1 + 2I2 ),
T
2 ~ ~
= S1 x(1 x)(4I5 + 2I6 ), (3.13)
T
S2 ~ ~
=2 x(1 x)(4I5 + 2I6 ), (3.14)
T
2
= S2 x2 (4I3 + 2I4 ). (3.15)
T
As a result we obtain the following system for magnetizations of the first sub-lattice x1
and x2 near Curie temperature
x1 = t1(x1 + x2 ) a1(x1 + x2 )3, (3.16)
x2 = t2(x1 + x2) a2(x1 + x2)3, (3.17)
where
S1 +1 S +1 (3.18)
t1 = , t2 = 2 ,
3 3
(S1 +1)[1+ (2S1 +1) 2 ] , (S 2 +1)[1+ (2S 2 +1) 2 ] (3.19)

a1 = a = .
45(2S1) 2 2
45(2S 2 ) 2
The iteration method gives the following solution of system Eqs. (3.16), (3.17)
b + b 2 4ac x 2 = b+ b 4ac
2
(3.20)
x12 = , 2
2a 2 2a 2
where
a = 3a1a2t22 2, (3.21)
b = [a1t23 2 + a2t1(1 t2 )2 ], (3.22)
(1t2 )2 (3.23)
c= d,

a = 3a1a2t12 2, (3.24)
b = [a2t13 2 + a2t2 (1 t1 )2 ], (3.25)
(1t1 ) 2
(3.26)
c = d,

d = t1t2 (1 t1 )(1 t2 ). (3.27)
When temperature approaches the ordering temperature from below magnetization x1

and x2 should turn to zero. One can see that this is carried out when the condition d = 0 is
fulfilled. Solution of this equation leads to the following expression for Curie temperature
1 (3.28)
TC = {M B + N B + PB },
3
M B = (1 x)2 S1(S1 +1)(4I1 + 2I 2 ) + x2S2 (S2 +1)(4I3 + 2I 4 ), (3.29)
N B = [(1 x)2 S1(S1 +1)(4I1 + 2I 2 ) x2S2 (S2 +1)(4I3 + 2I 4 )]2, (3.30)
~ ~
PB = 4x2 (1 x)2 S1(S1 +1)S2 (S2 +1)(4I5 + 2I6 )2. (3.31)
One can easy see that magnetization tends to zero at approaching TC according to the
law
1 c (3.32)
x12 (T TC ).
(TC )b(TC ) T TC
2
A - Structure
At antiferromagnetic ordering of A type we have (see Figure 1.2)
x3 = x1, x5 = x7 = x1, x4 = x2 , x6 = x8 = x2. (3.33)

424 B. V. Karpenko
The iteration method analogous to that for ferromagnetic B -ordering leads to the
following expression for Nel temperature
1 (3.34)
TNA = {M A + N A + PA },
3
where
M A = (1 x)2 S1(S1 +1)(4I1 2I2 ) + x2S2 (S2 +1)(4I3 2I 4 ), (3.35)
N A = [(1 x)2 S1(S1 +1)(4I1 2I 2 ) x2S2 (S2 +1)(4I3 2I 4 )]2, (3.36)
~ ~
PA = 4x2 (1 x)2 S1(S1 +1)S2 (S2 +1)(4I5 2I6 )2. (3.37)
C -structure
As it can be seen from Figure 1.2, the following correlations between sub-lattices
magnetizations take place in this case
x3 = x1, x5 = x1, x7 = x1, x4 = x2 , x6 = x2 , x8 = x2. (3.38)
Then the Nel temperature has the form
1
TNC = {M C + N C + PC } , (3.39)
3
where
M C = (1 x)2 S1(S1 +1)(4I1 + 2I 2 ) + x2 S2 (S2 +1)(4I3 + 2I 4 ), (3.40)
NC = [(1 x)2 S1(S1 +1)(4I1 2I2 ) x2S2 (S2 +1)(4I3 2I 4 )]2, (3.41)
~ ~
PC = 4x 2 (1 x)2 S1(S1 +1)S2 (S2 +1)(4I5 2I 6 )2. (3.42)
G -structure
With the use of conditions (see Figure 1.2)
x3 = x1, x5 = x1, x7 = x1, x4 = x2 , x6 = x2 , x8 = x2 , (3.43)
we obtain the following expression for ordering temperature

1 (3.44)
TNG = {M G + N G + PG },
3
where
M G = (1 x)2 S1(S1 +1)(4I1 + 2I 2 ) x2S2 (S2 +1)(4I3 + 2I 4 ), (3.45)
NG = [(1 x)2 S1(S1 + 1)(4I1 + 2I 2 ) x 2S2 (S2 + 1)(4I3 + 2I 4 )]2 , (3.46)
~ ~
PG = 4x2 (1 x)2 S1(S1 +1)S2 (S2 +1)(4I5 + 2I6)2. (3.47)
For all three collinear antiferromagnetic structures A, C and G the temperature

dependence of separate sub-lattice magnetization near ordering temperature has the form
x1 TNi T , i = A, C , G. (3.48)
3.2. The Wide Temperature Range
When we investigated the temperature dependence of magnetization near ordering

temperature where magnetization tends to zero the square root in the expression of double
exchange operator was expanded into series. If we are far from ordering temperature such an
expansion is not righteous. The magnetization is expressed via Brillouin function only at
Heisenberg form of the Hamiltonian. In our case magnetization should be calculated with the
help of overall formula
)
Sp( J z exp( H T )) (3.49)
Jz = ) .
Sp(exp( H T )
In the molecular field approximation the systems of equations for mean spin values of
Mn3 + and Mn 4 + ions, belonging to the first sub-lattice at different types of ordering have the
following form ( i = B, A, C , G ):
2 2(1 x)
m exp [ Ki + Li ] (3.50)
x1(i) =
m = 2 T ,
2 2(1 x)
exp [Ki + Li ]
m = 2 T
426 B. V. Karpenko
32 2x
m exp [ Ri +Qi ] (3.51)
x2 (i) =
m= 3 2 T ,
32 2x
exp [ Ri +Qi ]
m = 3 2 T
K B = m(1 x)(4 I1 + 2I 2 ) x1 + x(4 I5 + 2I6 ) x2, (3.52)
S2 +1 (3.53)
LB = x (4B + 2B2 ) 1+x2 m ,
2S2 +1 1
RB = m(1 x)(4I5 + 2I6 ) x1 + x(4I3 + 2I 4 ) x2 , (3.54)
S 2 +1 (3.55)
QB = (1 x) (4B + 2B2 ) 1+x1m ,
2S 2 +1 1
K A = m(1 x)(4I1 2 I2 ) x1 + x(4I5 2I6 ) x2 , (3.56)
S +1 (3.57)
LA = x (4B 1+x2 m + 2B2 1x2m ),
2S +1 1
RA = m(1 x)(4I5 2I6 ) x1 + x(4I3 2I 4 ) x2 , (3.58)
S +1 (3.59)
QA = (1 x) (4B 1+x1m + 2B2 1x1m ),
2S +1 1
KC = m(1 x)(4I1 2 I 2 ) x1 + x(4 I5 2 I6 ) x2 , (3.60)
S +1 (3.61)
LC = x (4B 1x2 m + 2B2 1+x2 m ),
2S +1 1
RC = m(1 x)(4 I5 2 I6 ) x1 + x(4 I3 2 I 4 ) x2 , (3.62)
S +1 (3.63)
QC = (1 x) (4B 1x1m + B2 1+x1m ),
2S +1 1
K = m(1 x)(4I1 + 2I2 ) x1 + x(4I5 + 2I6 ) x2 , (3.64)

G
S +1 (3.65)
LG = x (4B + 2B2 ) 1x2 m ,
2S +1 1
RG = m(1 x)(4I5 + 2 I6 ) x1 + x(4 I3 + 2 I 4 ) x2, (3.66)
S +1 (3.67)
QG = (1 x) (4B + 2B2 ) 1x1m ;
2S +1 1
2 (3.68)
= .
1+ S (2S +3)
The systems of equations (3.50), (3.51) were solved numerically with the use of exchange
parameters and transfer integrals values found in paper Part 1. At that for each type of
ordering B , A , C or G some concentration was chosen from the corresponding interval.
Figures 3.1-3.4 display the temperature dependences of mean values x1 and x2 of ions
Mn3 + and Mn 4 + spins, correspondingly. Figures 3.5-3.8 describe temperature dependences
3+
of contributions to magnetization on separate site from ions Mn
1 = (1 x) x1 (3.69)
4+
and Mn
2 = xx2 . (3.70)
In the same place on Figures 3.5-3.8 temperature dependence of the total magnetization
of a separate site
= 1 + 2 (3.71)
is presented for all investigated collinear configurations. It should be noted that for
3+
configurations B and C contributions to separate site magnetization from ions Mn and
4+
Mn have one and the same sign while for configurations A and G the signs are different.
428 B. V. Karpenko
2,0
x1, x2 1,5
1,0
0,5
0,0
0,0 0,2 0,4 0,6 0,8 1,0
T/TC
Figure 3.1. Temperature dependence of mean spin values for ions Mn 3+ (short dashcurve) and
Mn 4+ (dash dot curve) at ferromagnetic ordering. The data are obtained with concentration x = 0.3 ,
in this case Tc = 55.9 K .
2,0
1,5
1, 2,
1,0
0,5
0,0
0,0 0,2 0,4 0,6 0,8 1,0
T/T C
Figure 3.2. Temperature dependence of contributions to magnetization of a separate site from ions
Mn 3+ 4+
(short dash curve) and Mn (dash dot curve) and the total magnetization of a separate site
(solid curve) at ferromagnetic ordering.
2,0
1,6
1,2
0,8
x1, x2
0,4
0,0
-0,4
-0,8
-1,2
0,0 0,2 0,4 0,6 0,8 1,0

A
T/T N
Mn 4+ (dash dot curve) at antiferromagnetic ordering of A-type ( x = 0.1, TN = 168.7 K ).
A
2,0
1,6
1, 2,
1,2
0,8
0,4
0,0
0,0 0,2 0,4 0,6 0,8 1,0

A
T/T N
Mn 3+ 4+
(solid curve) at antiferromagnetic ordering of A-type.
430 B. V. Karpenko
2,0
1,5
x1, x2
1,0
0,5
0,0
0,0 0,2 0,4 0,6 0,8 1,0
C
TN
Mn 4+ (dash dot curve) at antiferromagnetic ordering of C-type ( x = 0.75, T N = 37.2 K ).
C
1,6
1,2
1, 2,
0,8
0,4
0,0
0,0 0,2 0,4 0,6 0,8 1,0
C
T/TN
Mn 3+ 4+
(solid curve) at antiferromagnetic ordering of C-type.
2,0
1,5
1,0
x1, x2 0,5
0,0
-0,5
-1,0
-1,5
0,0 0,2 0,4 0,6 0,8 1,0

G
T/TN
Mn 4+ (dash dot curve) at antiferromagnetic ordering of G-type ( x = 0.9, TN = 106.5K ).
G
0,5
0,0
1, 2,
-0,5
-1,0
-1,5
0,0 0,2 0,4 0,6 0,8 1,0
G
T/TN
Mn 3+ 4+
(solid curve) at antiferromagnetic ordering of G-type.
432 B. V. Karpenko
3.3. Some Notes
The magnetic properties of four collinear magnetic configurations of

compound La1 xCax MnO3 are investigated in molecular field approximation. Expressions
for ordering temperatures are found. They have the form analogous to Nel temperature for
4+
ferrimagnets, and contain evident dependence upon concentration x of Mn ions.
The analytical temperature dependences of spontaneous site magnetization were obtained
at temperatures near the ordering temperature. For all collinear structures under investigation
this dependence has the form
Tord T . (3.72)
Temperature dependences of site magnetization are calculated in the whole temperature

range from zero till ordering temperature. We used the results of papers [60, 61] in this Part.
CONCLUSION
Some problems of physics of lanthanum-calcium manganites were considered in this
paper from theoretical point of view. In the frame of proposed model of spin alloy the system
behavior was studied at different temperatures in the dependence of concentration
heterovalent manganese ions.
In Part 1 at zero temperature the magnetic structures change one another with
concentration variety and this process may be described as the pure concentration effect. No
structural changes need for explanation of magnetic configuration transitions. But, of cause, it
is not so in reality: with concentration variation both the crystallographic and electron
structures change also. Therefore we called our object as a hypothetical compound. It seems
that this is impossible just now to give the complete picture of magnetic transitions in the
whole concentration range.
In Part 2 the spin-wave excitations were considered and in broad terms the concordance
between theory and experiment was obtained. As to numerical values of interaction
parameters then two difficulties are exist. First difficulty is the large spread of experimental
data. Second difficulty is the shortcoming of theoretical treatment methods. Above used the
mean square method they consider as incorrect approach. This method can not give the
evidence of truth of some theory, but only gives the preference to certain set of parameter
values rather then to any other. Therefore the good concordance between theory and
experiment may turn out illusory. Nevertheless, after these, not very optimistic notes, say, that
comparison theory and experiment must continue with attraction different theories and more
sophisticated treatment methods. As to present paper we say, once more, that our results have
rather quantitative not qualitative character.
In Part 3 the temperature dependences of spontaneous magnetization of sublattices were
obtained. Not comparison with experiment was done, we hope to make it in future. The
theoretical formulas have general character and obtained numerical values and graphical
matter are given as peculiar examples. It is interesting is it possible to distinguish by some
measurements the contributions in general magnetization from three and four valence
manganese ions.
In the conclusion must be said that present approach to the above considered problems is
not unique; the different points of view, of cause, may exist.
ACKNOWLEDGMENTS
I thank L. P. Karpenko, L. V. Karpenko, E. A. Volkova, A. V. Kuznetsov, V.V. Dyakin,
S. V. Naumov and L. D. Falkovskaya for help in the work.
The author is indebted to Department of Physical Science and Presidium of Ural Branch
of the Russian Academy of Sciences, Program The Physics of new materials and structures,
the Russian Foundation for Basic Research (project N 07 02 000 68).
REFERENCES
[1] Jonker, G. H.; van Santen, J. H. Physica 1950, 16, 337.
[2] Jonker, G. H.; van Santen, J. H. Physica 1950, 16, 599.
[3] Zener, C. Phys. Rev. 1951, 81, 440.
[4] Zener, C. Phys. Rev. 1951, 82, 403.
[5] Wollan, E. O.; Koehler, W. C. Phys. Rev. 1955, 100, 545.
[6] Anderson, P.W.; Hasegawa, H. Phys. Rev. 1955, 100, 675.
[7] de Gennes, P.-G. Phys. Rev. 1960, 118, 141.
[8] Ramirez, A. P. J. Phys.; Conds. Matter. 1997, 9, 81711.
[9] Dagotto, E.; Yunoki, S.; Malvezzi, A. L.; Hu, J.; Capponi, S.; Poilblanc, D.;
Furukawa, N. Phys. Rev. B, 1998, 58, 6414.
[10] Imada, M.; Fujimori, A.;Tokura, Y. Rev. Mod. Phys. 1998, 70, 1039.
[11] Moreo, A.;Yunoki, S.; Dagotto, E. Science 1998, 283, 2034.
[12] Coey, J.; Viret, M; von Molnar, S. Adv. Phys. 1999, 48, 167.
[13] Tokura, Y.; Tomioka, Y. J. Magn. Magn. Matter. 1999, 200, 1
[14] Ibarra, M. R.; De Teresa, J. M. Mater. Sci. Forum 1999, 302-303, 125.
[15] Loktev, V. M.; Pogorelov, Yu. G. Fizika Nizkhih Temper. 2000, 26, 231. (In
Russian)
[16] Dagotto, E.; Hotta, T.; Moreo, A. Phys. Rep. 2001, 344, 1.
[17] Moreo, A. Electron Spectrosc. Relat. Phenom. 2001, 117-118, 251.
[18] Salamon, M. B.; Jaime, M. Rev. Mod. Phys. 2001, 75, 583.
[19] Kagan, M. Yu.; Kugel, K. I. Usp. Fiz. Nauk 2001, 171 B, 577. (In Russian)
[20] Dagotto, E. Nanoscale Phase Separation and Colossal Magnetoresistance; Springer:
Berlin, 2002; p.p. 448.
[21] Haghiri-Gosnet, A.-M.; Renard, J.-P. J. Phys. D: Appl. Phys. 2003, 36, R127.
[22] Mathur, N.; Littlewood, P. Phys. Today 2003, 56, 25.
[23] Weisse, A; Fehske, H. New J. Phys. 2004, 6, 158.
[24] Dagotto, E. New Journal of Physics 2005, 7, 67.
[25] Ruderman, M. A.; Kittel, C. Phys. Rev.1954, 96, 99.
434 B. V. Karpenko
[26] Kasuya, T. Progr. Theor. Phys.1956, 16, 45.

[27] Yosida, K. Phys. Rev. 1957, 106, 893.
[28] Schubin, S. P.; Vonsovsky S. V. Proc. Roy. Soc. 1934, A145, 159.
[29] Vonsovsky, S. V. Magnetism; Nauka; Moscow, 1971, p.p. 805. (In Russian).
[30] Moussa, F.; Hennion, M.; Rodriguez-Carvajal, J.; Mouden, H.; Pinsard, L.;
Revcolevschi, A. Phys. Rev. B 1996, 54, 15149.
[31] Hennion, M.; Moussa, F.; Rodriguez-Carvajal, J.; Pinsard, L.; Revcolevschi, A.
Phys. Rev. B 1997, 56, R497.
[32] Hennion, M.; Moussa, F.; Bioteau, G.; Rodriguez-Carvajal, J.; Pinsard, L.;
Revcolevchi, A. Phys. Rev. Lett. 1998, 81, 1957.
[33] Hennion, M.; Moussa, F.; Bioteau, G.; Rodriguez-Carvajal, J.; Pinsard, L.;
Revcolevchi, A. Phys. Rev B 2000, 61, 9513.
[34] Bioteau, G.; Hennion, M.; Moussa, F.; Rodriguez-Carvajal, J.; Pinsard, L.;
Revcolevschi, A.; Mukovskii, Y. M.; Shulyatev, D. Phys. Rev. B 64, 2001, 104421.
[35] Radaelli, P. G.; Cox, D. E.; Capogna, L.; Cheong, S.-W.; Marezio, M. Phys. Rev. B
1999, 59, 14440.
[36] Pissas, M.; Kallias, G.; Hofmann, M.; Tbbens, D. M. Phys. Rev. B 2002, 65,
064413.
[37] Pissas, M.; Kallias, G. Phys. Rev. B 2003, 68, 134414.
[38] Heffner, R. H.; Sonier, J. E.; Machlaughlin, D. E.; Nieuwenhuys, G. M.; Luke, G.
M.; Uemura, Y. J.; William Ratcliff II; Cheong, S. W.; Balakrishnan, G. Phys. Rev. B
2001 63, 094408.
[39] Karpenko, B.V.; Falkovskaya, L. D.; Kuznetsov, A.V. (2005) Magnetic structure and
double exchange in hypothetical compound La1 x Ca x MnO3 , cond-mat0509507.
[40] Karpenko, B.V.; Falkovskaya, L. D.; Kuznetsov, A.V. The Physics of Metals and
Metallography 2008, 105, 443.
[41] Moussa, F.; Hennion, M.; Wang, F.; Kober, P.; Rodrigez-Carvajal, J.; Reutler, P.;
Pinsard, L.; Revcolevcshi, A. Phys. Rev. B 2003, 67, 214430-1.
[42] Perring, F.G.; Aeplli, G.; Hayden, S. M.; Carter, S. A.; Remeika, J. P.; Cheong, S-W.
Phys. Rev. Lett. 1996, 77, 711.
[43] Hwang, H. Y.; Dai, P.; Cheong, S. W.; Aeppli, G.; Tennant, D. A.; Mook, H. A.
Phys. Rev. Lett. 1998, 80, 1316.
[44] Hennion, M.; Moussa, F. New Journal of Physics 2005, 7, 84 .
[45] Ye, F.; Dai, P.; Fernandez-Baca, J. A.; Hao Sha; Lynn, J. W.; Kawano-Furukawa, H.;
Tomioka, Y.; Tokura, Y.; Zhang, J. Phys. Rev. Lett. 2006, 96, 047204-1.
[46] Ye, F.; Dai, P.; Fernandez-Baca, J. A.; Adroja, D.T.; Perring, T. G.; Tokura, Y. Phys.
Rev. B 2007, 75, 144408-1.
[47] Feiner, L. F.; Ole, A. M. Physica B 1999, 259-261, 796.
[48] Ole, A. M.; Feiner, L. F. Phys. Rev. B 2002, 65, 052414-1.
[49] Murakami, Y.; Hill, J. P.; Gibbs, D.; Blume, M.; Koyama, I.; Tanaka, M.; Kawata,
H.; Arima, Y.; Tokura, Y.; Hirota, K.; Endoh, Y. Phys. Rev. Lett. 1998, 81, 582.
[50] Lines, M. E. Phys. Rev. 1964, 133, A841.
[51] Lines, M. E. Phys. Rev. 1964, 135, A1336.
[52] Karpenko, B. V.; Falkovskaya, L. D.; Dyakin, V. V.; Kaybitcheva, S. L. Low-Dim.
Struct. 2002, 7/8, 57.
[53] Karpenko, B. V.; Falkovskaya, L. D.; Dyakin, V. V.; Kuznetsov, A. V.; Kaybitcheva,
S. L. Low-Dim. Struct. 2003, 5/6, 53.
[54] Prado, F.; Zysler, R.; Morales, L.; Caneiro, A.; Tovar, M.; Causa, M. T. J. Magn.
Magn. Mater. 1999, 196-197, 481.
[55] Causa, M. T.; Alejandro, G.; Zysler, R.; Prado, F.; Caneiro, A.; Tovar, M. J. Magn.
Magn. Mater. 1999, 196-197, 506.
[56] Tovar, M.; Alejandro, G.; Butera, A.; Caneiro, A.; Causa, M.T.; Prado, F.; Snchez,
R. D. Phys. Rev. B 1999, 60, 10199.
[57] Karpenko, B. V.; Falkovskaya, L. D.; Kuznetsov, A. V. The Physics of Metals and
Metallography 2008, 106, 136.
[58] Karpenko, B. V.; Falkovskaya, L. D.; Kuznetsov, A. V. Physics of the Solid State
2008, 50, 2102.
[59] Karpenko, B. V.; Kuznetsov, A. V. (2008). The Exchange Parameters in Lanthanum
Manganites, cond-mat arXiv:0805. 3006 v1.
[60] Karpenko, B. V.; Falkovskaya, L. D.; Kuznetsov, A. V. (2006). Magnetic Properties
of Collinear Structures of Compound La1 xCax MnO3 , cond-mat/0604161.
[61] Karpenko, B. V.; Falkovskaya, L. D.; Kuznetsov, A. V. Israel Journal of Chemistry
2007, 47, 397.
Chapter 10
CORRELATIONS BETWEEN THE STRUCTURE OF

ORDERED SOLID SOLUTIONS AND PARAMETERS OF
PAIR INTERACTIONS IN PEROVSKITES
A. Yu. Gufan1 and K. Yu. Gufan2

1
Institute of Physics, South Federal University, Rostov-on-Don, Russia
2
Institute of Security Computing Devices and Automatics, Rostov-on-Don, Russia
ABSTRACT
In the chapter two approaches to the crystals ordered states description are used. The
first is basing on the group theoretical considerations incorporated in theory with the help
of Landau potential. The second way of ordering description consist of accounting
different models of interatomic interactions. Our goal is to elucidate correlations between
these two approaches. Complex oxides with composition

AB / B // O
c 1 c 3 and perovskite
like structure are considered as an example. Six supercells which may be realized as
stable states in these solid solutions in accord with Landau theory are analyzed in the
framework of both approaches. The potential of interatomic coupling includes van der
Waals energy and terms following from pseudopotential model. The sum of interatomic
energy in twelve coordination spheres was accounted to get correlation between the
parameters of Landau potential and coefficients of interatomic energy and to define the
domain of different ordering state stability in the space of phenomenological parameters.
In particular it is shown that under certain conditions the type of ordering in a surface
area may differ from the type of ordering in a bulk of crystal. Using a new approach to X-
ray diffraction pattern interpretation there was established definite correlations between
the parameters of pair interactions potential and the peculiarities of ordered states which
may be defined from the experiment.
438 A.Yu Gufan and K. Yu Gufan
1. INTRODUCTION
The physical properties of multi-component oxides with the perovskite structures (OPS)
are strongly correlated with the cationic composition. The magnetoresistive effect exists in
almost all manganates containing rare-earth elements similar to La 2/3 Ca 1/3 MnO 3- [[1] the
high-Tc conductivity occurs in many kinds of cuprates, containing Ba and rare-earth element
(R) or Yttrium similar to R1 / 3 Ba 2 / 3 CuO3 y [2],[3]; the relaxor properties exists in many Pb-
based, complex OPS similar to PbMg 1/3 Nb 2/3 O 3 , etc.
The cationic composition determines the specific properties of the structure. Usually, the
structure of multi-component OPS is described by comparison with the structure of parent
1
phase. The OPS parent phase of ABO 3 compound is characterized by Oh -symmetry. The
structure of parent phase is determined by the distribution of , and ions in vertexes of the
1
system of regular points (SRP) of Oh -symmetric unit cell by the following way: A-ions are
located in 1(b) ( 1 , 1 , 1 ) SRP, B-ions are in 1(a) ( 0,0,0 ) SRP and oxygen O occupies
2 2 2
3(d) SRP. The composition of the multi-component OPS
/ // // / // // (i) (i)
is ( A A A ...)( B B B ...)O3 , where A
x1 x2 x3 y1 y2 y3 and B ions are distributed, regularly or
randomly, in 1(b) and 1(a) SRP respectively. In complex OPS, the order of the cationic
distribution in a crystal is strongly correlated with the crystals composition. For example, the
3+ 2+
structures of all OPS similar to R1 / 3 Ba 2 / 3 CuO3 are marked by the following ordered
distribution of R and Ba cations along the axis of 4-fold parent phase symmetry [[4] in A-
block: R, Ba, Ba, R, Ba, Ba . The structure of Ba-based OPS, that show anomaly high
transparency in the range of millimeter-waves (for instance, BaMg1/3 Nb 2/3 O 3 or
BaZr1/3 Ta 2/3 O 3 [[5]) is characterized by triple parent phase period caused by the ordered
1
distribution of cations along the axis of 3-fold parent phase Oh .-symmetry in 1 (a) SRP (in
B-sublattice), too. In the Pb-based relaxors featured by giant electrostriction such as
PbB1// 3 B2/// 3 O3 , the parent phase period is doubled because of the ordered distribution of the
cations along the axis of 3-fold symmetry in B-sublattice [6].
The diversity of the observed types of the ordered cationic distribution inside crystal
lattice put a question about the characteristics of structural elements of complex OPS, the
combination of which determines a particular type of the order in the cationic distribution. In
other words, it is interesting to build a sufficiently founded mathematical model of
/ //
interactions between structural elements in crystals similar to A( B x B1 x )O3 and to clear
(elucidate) conditions on the model parameters determining the type of the order of the
cations distribution inside of crystal.
In this chapter, the model of the effective ion pair interaction spreads on not less than 11
coordinate spheres has been taken as a base model of interactions between structural
components. Regardless of the criticism of the model, the usage of approximation of effective
pair interactions has many advantages that become conclusive during the choice of the proper
Correlations between the Structure of Ordered Solid Solutions 439
model of interactions. First of all, this model gives possibility to conduct the numeric and
analytical researches simultaneously and, also provides us with an ability to obtain some
parameters of the model on the base of the first principles. Moreover, the numeric
parameters of the model have clear physical meaning and can be compared with real features
of the studied objects independently from the experiment conditions.
These arguments, mainly practical, in favor of using the approximation of effective pair
interactions, can not let us disregard very substantial arguments against the model. So, in the
next section of this chapter, the results obtained by using the model of effective pair
interactions for the interpretation of data received from X-ray diffraction pattern of
PbFe1 / 2 Nb1 / 2O3 sample are given as an additional confirmation of usefulness of the model
in case of a perovskite-like structure. Although the application of the model has to be
legitimated during further discussion, the obtained result is of independent interest.
2. INTERPRETATION OF X-RAY DIFFRACTION PATTERNS OF

COMPLEX OXIDES WITH A PEROVSKITE-TYPE STRUCTURE
2.1. Theoretical Aspects
The primitive cell of the ideal perovskite ABO3 structure is cubic. It is characterized by
the following positions of ions in the arbitrarily chosen first unit cell: A , B and O ions
occupy, respectively, the positions 1(a) (0, 0, 0) , 1(b) (a/2, a/2, a/2) and
3(c) [(0, a/2/a/2), (a/2, 0, a/2), (a/2, a/2,0)] . A ions are generally have a large ionic
2+ 2+
radius, ( Pb , Ba , Ca 2 + ... ), whereas B ions ( Ti 4+ , Mg 2+ , Nb 5+ etc) generally have a
small ionic radius. In an perfect crystal of the ABO3 composition , the static electric field
induced by all ions at any site belonging to the regular system of points 1(a ) is zero due to
the symmetrical arrangement of the occupied regular systems of points 1(b) and 3(c) . Such
compensation occurs at any pair interactions between ions if these interactions depend only
on distance.
/ //
If ions of two types, B and B , are distributed (regularly and irregularly) over the
system of regular points 1(b) , there will necessarily be an uncompensated electric field at
some (or all) positions belonging to the regular system of points 1(a) . The potential energy
minimum determining the average position of A ions cannot be located at positions with an
uncompensated field. Thus, in solid solutions of oxides with perovskite structure, A ions are
spontaneously shifted from 1(a ) positions. The shift of A ions in turn leads to the shift of
/ //
other ions. Since shifts of B , B and O ions are secondary, this process is referred to as
relaxation of the structure. In view of partial (or complete) randomness of the distribution of
B / and B // ions over the actual lattice sites, X-ray diffraction patterns of crystals of solid
solutions (and compounds within the homogeneity range in the phase diagram) differ from
the X-ray diffraction patterns of perfect crystals [[7]-[13]. In particular, an X-ray diffraction
/
pattern contains information about the degree and type of order in the distribution of B and
B // , ions, the values and anisotropy of mean static and dynamic (thermal) shifts of
A, B / , B // , O ions, and the electron-density distribution in the unit cell. The question is on
the basis of what models one can obtain reliable information about the above-mentioned
specific features of an actual structure.
Let us describe some class of models used to interpret X-ray diffraction patterns of lead
containing solid solutions of perovskites of the PbB x/ B1// x O3 /
composition, where B and
B // ions are generally several times lighter than Pb (atomic weight 207,2) ions:: Mg (24),
Nb (93), Zn (65,4), Sc (44) etc. It is well known that the scattering factor (amplitude) of
an ion is approximately proportional to its atomic weight (more strictly, to the total number of
electrons localized at the ion). This circumstance makes it possible to develop a model for
interpreting X-ray diffraction patterns by the method of successive approximations [[7]-[13]].
The zero-order approximation [[7]-[13]] is based on the assumption that all atoms
reliably occupy all sites which are allowed for them in the ideal perovskite structure and the
/
scattering factor f of the ions located in positions ( 1b ), is calculated for a mean ion. If B
and B // atoms are irregularly distributed over the 1(b) positions,
f ( X ) = f ( B / ) c + f ( B // ) (1 c) . If the distribution of B / and B // " atoms is ordered
as in compounds (as for example in daltonides [[14]]) or partially ordered as in ordering solid
solutions, X-ray diffraction patterns are interpreted using a unit cell of lower symmetry,
whose positions are occupied by Y atoms with the scattering factor
f (Y ) = P1 f ( B / ) + P2 f ( B // ), where P1 , P2 = 1 P1 are the occupation
/ //
probabilities for B and B atoms. The fitting parameters used to fit the X-ray diffraction
pattern calculated within the zero-order approximation model to experimental patterns, along
with the tentative structure, are the Debye-Waller factors W . The factors W are proportional
to the mean square shift of each ion ( Pb , X , O ) from its equilibrium position. In the
calculation formulas for scattering intensity, they enter as factors at the scattering amplitudes
of ions. The scattering amplitude modified by the factor W is f / = f exp( (W / kT ))
. Here - is the type of an atom ( = Pb, O, X ). Within the zero-order approximation, the
factors W are assumed to be spherically symmetric:
W ~ (< x2 > + < y2 > + < z2 >) , where < x > - is the mean-square shift of an
ion along the x axis, etc. When the ideal perovskite structure is considered and the mean-
X atom approximation is used to calculate f , the assumption about the isotropic Debye-
Waller factor is valid only for oxygen atoms having tetragonal environment. The mean-square
r
shift of an atom from the equilibrium position: < r2 >=< x2 > + < y2 > + < z2 > ,
includes both dynamic components due to thermal atomic vibrations with meansquare
amplitudes proportional to k , and static components < rst > , due to the random
2
/ //
distribution of B and B ions over the regular system of points 1(b) in the ideal perovskite
structure: < r 2 >=< rst2 > + < rd2 > .
Generally, the model in which all lead atoms in the solid solution are statically shifted
from their equilibrium positions in the ideal structure is considered in the first-order
approximation [10],[13] . In each unit cell, lead atoms are shifted in the same directions and
at the same distance < r > . In the first-order approximation models, shifts in asymmetric
directions are sometimes chosen [[8]]. In this case, the model acquires three' additional fitting
parameters: x Pb , y Pb , z Pb .
The models assuming that Pb ions are shifted with equal probabilities in several
nonequivalent directions should also be assigned to the first-order approximation models. If
we do not assume that shifts in nonequivalent directions are equiprobable, such models also
include additional phenomenological parameters related only to Pb ions: shifting
probabilities in a specified direction and the shift values.
The next-level models allow for some relaxations of the structure, i.e., take into account
/ //
shifts of X or ( B and B ) and O [7]-[13]. In these models, it is assumed that the system
of unit cells of a crystal is divided into two or more subsystems. A specific model is put into
correspondence with each subsystem and the concentrations of cells belonging to a particular
subsystem are used as an additional fitting parameter [8],[12].
Thus, in the approach of current use, refinement of the X-ray diffraction pattern
interpretation results in an increase in the number and variety of fitting parameters in the
theoretical model of a structure studied. Besides that, most of these parameters are introduced
ad hoc and have an unclear physical meaning outside the context of the interpretation of
diffraction patterns.
However, as will be shown below, the zero-order approximation model has a potentiality
of direct analytical calculation of a larger part of phenomenological (fitting) model
parameters of the next approximations. Calculation is performed on the basis of theoretical
probabilistic analysis of the chemical compositions of unit cells of a solid solution crystal,
which are possible in principle. Consideration of this fact should simplify and refine the
theory of X-ray diffraction patterns.
For example, the model of a perfect crystal suggests that all positions allowed for B
/ // / //
atoms, are occupied by B and B with only one ( B or B ) atom per position. Therefore,
/ //
the distribution of B and B atoms over the 1(b) position obeys a binomial law. In more
complex oxides ( A A ... A1nc1c 2 ... )( Bc/ 3 Bc//4 ...B1nc1c 2 ... )O3 the Fermi statistics is
/
c1
//
c2
valid for independent distribution of A and B atoms. Note that vacancies in solid solutions
can be considered as the third clement of A or B types.
/ //
Let us assume that all B and B atoms are randomly distributed over the sites
belonging to the regular system of points 1(b) of the ideal perovskite structure. In this case,
22 structural versions of the inveronment of Pb ions are possible. They include: (i) eight
/ //
versions in which 0, 1, 2, and 3 B ions and 8, 7, 6, and 5 B ions are arranged in
geometrically different ways around the lead ion; (ii) eight versions with interchanged
/ // /
positions of B and B ions; and (iii) six geometrically distinguishable positions of four B
//
and four B ions. If only interaction with the nearest neighbors is taken into account, all 14
geometrically different environments of Pb ion form nine types of crystal field of different
symmetry at 1(b) positions: Oh , Td , D3d , D2 h , C 3v , C 4 v , C 2 v , C s , C 2 (Figure 1). If the
crystal field has any of the C3v , C 4 v , C 2 v , C s , C 2 types of symmetry, the Pb , ion will be
necessarily shifted from its position in the ideal perovskite lattice toward the symmetry axis
of the four above-mentioned point groups. If the field symmetry at the 1(b) position is C s ,
the Pb , ion can be generally shifted in an arbitrary direction in one of the 12 planes of the
[1,1,0] family of the ideal cubic cell. However, if the interaction forces between lattice ions
are central and pair, the situation is simplified. Let us consider a unit cell with seven sites
/ //
occupied by B atoms and one site occupied by B . The forces acting on the Pb ion from
/
the B ions located at the ends of one of the body diagonals of the cube balance each other.
/ //
Therefore, the Pb ion can be shifted only along the body diagonal connecting B and B
ions. Let us denote the corresponding shift as u . Obviously, the ion Pb will be shifted by
the same value u and in the same direction (along the threefold symmetry axis) in the cell
// /
where seven sites are occupied by B and one site is occupied by B . Let us now consider a
/ //
unit cell with two sites in the 1(b) position, occupied by B and six sites occupied by B
/ /
ions. There may be three geometrically different versions of arrangement of B . If B ions
occupy the lattice sites located along the same body diagonal of the unit cell, the field at the
Pb ion has the D3d symmetry and, therefore, the Pb ion will not be shifted from the 1()
/
position. If B ions are located in the lattice sites connected by the same edge of the unit cell,
/ //
the symmetry of the crystal field induced by B and B ions will be C 2 v at the center of the
unit cell, and Pb ion should be shifted along the twofold axis of the (110) type. The value of
this shift can easily be calculated from geometric considerations, = u (2 2 / 3) . All
/ //
other shifts of Pb ions for all 256 possible geometries of arrangement of B and B ions in
the ideal perovskite lattice are calculated similarly using the fitting parameter of the theory.
For a partially ordered crystal, the number of fitting parameters increases as a result of
consideration of the degree and type of order. However, these additional parameters can be
determined and/or calculated in interpretation of the results of independent experiments
[[15]. Moreover, the dependence of the probability distribution of shifts of Pb ions on the
/ //
degree of order in the distribution of B and B ions and the type of ordered distribution of
B / and B // , ions (calculated, for example, in [[16]), makes it possible, minimizing the factor
of incompatibility of theoretical and experimental X-ray diffraction patterns ( R -factor), to
independently establish the crystal symmetry in the ordered state and the degree of order.
Figure 1. All possible types of geometrically different nearest environments of A -ion in

/ // / //
AB B O3 . Solid and empty circles denotes B and B ions.
x 1 x
In what follows we present some evident results of calculation of the probabilities of

implementing particular mean shifts, obtained within the model based on a binomial
/ //
distribution of B and B ions over 1(b) positions.
Let us consider a disordered PbBc/ B1//c O3 solution and try to determine (i) the
probability of existing of a cell in which the Pb ion is shifted along one of 12 equivalent
directions of [1,1,0] type and (ii) the value of this shift, averaged over the crystal.
Such a shift of Pb ions can be implemented only when the nearest environment of the
Pb ion includes 2, 4, or 6 B // ions located at ideal 1(b) positions. Therefore, the
probability of shift of Pb ions in the above-mentioned direction is
(110) = (c (1 c) (c (1 c) 6 )) /(c 2 (1 c) 2 ) . In these cases, the crystal-field
2 2 6
symmetries at the center of the cubic unit cell are different: C 2 v , C 3 or. However, in the
approximation of central pair forces of interatomic interactions, Pb ions will be always
shifted in the [1,1,0] direction; their mean shift in this direction is P (2 2u 0 ) / 3 .
The results obtained for a disordered crystal are shown in Figure 2 which demonstrates
the probabilities of shifts of Pb ions along the [1,1,1] , [1,1,0] , [0,1,0] directions. To
establish the relationship with the conventional models I [7]-[13], the dependences of the
/
crystal averaged shifts of lead ions along the x, y, z axes on the concentration of B are
shown in Figure 3. Note that within the calculation scheme proposed here the static mean-
square shifts are no more the fitting parameter of the theory and they are calculated with the
same accuracy as if all shifts of lead ions would be identical and uniformly distributed over
the crystal: < x > + < y > + < z >= 8u 0 (c c ) .
2 2 2 2
Figure 2. Probabilities of shifts of Pb ions along the [1,1,1] (1), [1,1,0] (2), [0,1,0] (3)
directions.
Figure 3. Dependences of the crystalaveraged shifts of lead ions along the x (1), y (2),
/
z (3) axes on the concentration of B ions.
/ //
If the distribution of B and B ions over the system of points 1(b) is more ordered,
calculation of the probabilities of implementing particular shifts of lead ions or crystal-
averaged shifts becomes more cumbersome. For brevity, we will present only the results for
/ //
one of possible types of ordering of B and B ions. Let us consider the ordered state in
which 1(b) sites of ideal perovskite are split into two nonequivalent sublattices and, thus, the
volume of the perovskite primitive cell is larger by a factor of two. For clearness, we will use
below a Bravais cell whose volume exceeds that of an ideal perovskite cell by a factor of
eight. The model considered here neglects spontaneous strains and the Bravais cell is assumed
to be cubic. The periods of the lattice composed of extended unit cells are
r r r
d1 (2a,0,0), d 2 (0,2a,0), d 3 (0,0,2a) . The eight 1(b) positions belonging to the extended
Bravais cell have the coordinates:
1.[0,0,0] 2.[a,0,0] 3.[0, a,0] 4.[a, a,0]

(2.1)
5.[0,0, a ] 6.[ a,0, a ] 7.[0, a, a ] 8.[ a, a,0]
/
We consider the ordering at which the probabilities of occupying these positions by B
ions are determined by the relations:
P1 = P8 = c + 3 , P2 = P3 = P4 = P5 = P = c (2.2)
/
Here, the parameter characterizes the degree of order in the distribution of B ions
over positions (2.1).
It is possible to calculate both the probabilities of occurring cells with different directions
of shifts of Pb ions and all mean parameters used to interpret X-ray diffraction patterns. For
example, the dependences of mean-square shifts of Pb ions along the x, y , z axes on the ion
concentration and degree of order of ion distribution in a crystal are fairly cumbersome:
< y 2 >= 8 * u02 (3c 7 + (50 / 3 + 14c) 6 + (33 - 58c - 25c 2 ) 5 + (18 - 76c + 68c 2 + 20c3 ) 4
+ (8/3 - 16c + 38c 2 - 76/3c3 - 5c 4 ) 3 + (-3 - 2c 4 - 2c 5 - 12c 2 + 12c3 + 4c) 2 + (-2c5 + c 6 + c 4 )
20/3c3 + 8/3c6 + 10c 4 + 1/3c - 8c5 + 5/3c 2
(2.3)
< z 2 >= 8 * u 02 (-36 8 + (-99 + 192c) 7 + (-400c2 - 334/3 + 414c) 6 + (384c3 - 63 + 326c - 601c 2 ) 5 +
(92c - 14 - 120c 4 - 220c 2 + 260c 3 ) 4 + (-64c5 - 16c + 155c 4 - 460/3c 3 + 70c 2 + 8/3) 3 +
(190c 4 + 48c 6 + 52c 2 - 146c 5 + 1 - 132c 3 - 12c) 2 + (-2c5 + c 6 + c 4 ) +
28/3c 3 + 16c 7 - 22c 4 - 4c 8 + 32c 5 - 7/3c 2 - 88/3c 6 + 1/3c
(2.4)
< x 2 > + < y 2 > + < z 2 >= 8u 0 [9 7 + (6 + 42c ) 6 + (75c 2 + 3 + 18c) 5 12c(1 5c 2 + c) 4 +
+ (18 15c 2 12c)c 2 3 + (6c 5 12c 3 + 18c 4 3) 2 + 3c 4 (c 1) 2 + c c 2 ]
(2.5)
Expressions (2.3)-(2.5) are pictorially illustrated in Figure 4a (dependences of x, y and

z on at c = 0.5 Figure 4b (dependences x, y, z on c at = 0.1 ) and Figure 4
(dependences of x, y and z on c at = 0.2 ). In addition, bearing in mind the possibility
of applying the results obtained here to description of X-ray diffraction patterns of the
PbB1// 2 B1/// 2 O3 and PbB1// 3 B2/// 3 O3 compounds, we present the mean-square shifts x, y, z
and mean-square static shifts, calculated for c = 1 / 2 and c = 1 / 3 and two val1les of ,
allowable at a given concentration:
=0,1 =0,125 =0,1 =0,125

<x2> (c=1/2) 0,439 0,423 <x2> (c=1/3) 0,406 0,384
<y2> (c=1/2) 1,136 1,023 <y2> (c=1/3) 0,952 0,844
<z2> (c=1/2) 0,193 0,183 <z2> (c=1/3) 0,181 0,176
<r2> (c=1/2) 1,767 1,630 <r2> (c=1/3) 1,539 1,404
Thus, the theoretical probabilistic structural model makes it possible to express the
probabilities of the distributions of shifts of A ions of the distributions of shifts of A ions in
the crystal lattice, both in magnitudes and directions, in terms of only one positional fitting
parameter and calculate the contribution of these spontaneous shifts to the anisotropic factor
W . Within these models, consideration of the anisotropy of the factor W does not lead to an
increase in the number of fitting parameters of the theory.
Figure 4. Illustration of expressions (2.3)-(2.5). Dependences of < x2 > (1), < y2 > (2) and
< z 2 > (3) on at c = 0.5 (a), on c at = 0.1 (b) and on c at = 0.2 (c).
2.2. Interpretation of PbFe1 / 2 Nb1 / 2 O3 X-ray Diffraction Pattern
The model presented above with a single phenomenological parameter u 0 is not seen
founded enough. To make it more precise it is possible to apply following consideration. It is
evident, that free volume in unit cell open for shifting of ion A in it, depends in some way on
/ //
nominal radius of B and B ions, and on the first approximation is proportional to the
number of ions with less radius, which were placed in the given unit cell. Thereby it is
reasonable to introduce a correction to the shifting value of A ion that will be proportional to
//
concentration of B ions in its nearest surrounding. Thus, on Figure 1(f) Bravais
parallelepiped will be characterized by shifting projection r (C s ) = u 0 + 3 / 8v 0 . If ions
( 2)
B / will be substituted by B // and vice versa, then projection r (C 3v(1) ) on reference axis will
be equal to u 0 5 / 8v 0 . Taking into account this correction, it is possible to evaluate the
/ //
width of Bragg reflexes component associated with random distribution of B and B ions
according to common rules. While evaluating the complete intensity of reflexes, we
considered ions thermal vibrations in isotropic approximation.
Intensity of 870 X-ray reflexes from PFN crystal annealed at 1600, measured on
diffractometer CAD-4 using MoK -emission and graphite monochromator was used as
experimental base to define of u 0 and v 0 values. Within the limits of measurement error, a
unit cell appears to be a cube with edge length equal to 4.0090.001, i.e. it contains one
PFN formula unit. Set of 870 experiment modules of structure amplitude was split into 59
groups; inside each several group modules of structure amplitude should have equal value
according to symmetry of paraelectric phase G0 = Oh .
1
Lines R = const that are presented at Figure 5, were calculated with step R = 0.02
and alteration of u 0 from 0.00 up to 0.40; and v 0 from -0.40 up to 0.40. We found two main
ravines, in one of which (the primary) at 0.0 u 0 0.2 the smallest values of R are
concentrated. At Figure 6 there is the tail of primary ravine [17], where value of R -factor
is higher than in its center.
As a result of minimization of R -factor the following values of positional parameters
( u 0 and v 0 ) and thermal parameters were calculated (recall that while making interpretation
of X-ray diffraction pattern, thermal parameters are understood as values 8 < u > , where
2
< u 2 > - is a mean-square amplitude of thermal oscillation of atoms):
u0 v0 BPb BNb BFe BO R%

0.0326 0.0001 0.568 0.730 0.781 1.42 2.45
Deducted value of R -factor keep pace with the best results of adapting with use of
traditional multiparametric models.
Figure 5. Constant R-factor lines in (u0 , v0 ) - space with step R = 0.02 .

Figure 6. The tail of primary ravine in (u0 , v0 ) - space.
Small value of v 0 parameter and, what is more important, clearly visible at Figure 5-6
highly correlating values of v 0 and u 0 under conditions of constant R -factor allows
supposing that the correction v 0 doesnt contribute anything to the model. Thus, the model,
based on an assumption of pair-wise interaction of closest neighbors, allows us to describe an
average local structure of the object investigated equally to experiment.
3. THEORY OF ORDERING IN COMPLEX OPS
3.1. General Aspects
Discussions on Theory and numerical modeling of ordering in complex oxides with

perovskite structure (and determination of ordering type) are commonly encountered in the
literature. In [[19], the state of Pb, Zr , Ti and O ions in was calculated in the one-electron
approximation. Both all possible electrostatic interactions (up to the quadrupolequadrupole
one) and short-range repulsive forces were taken into account in the effective pair interaction
energy. Then, the minimum energy for nine ordered structures was calculated in this
maximally complete model of interactions by successive decreasing of the energy of clusters
16 16 16a03 and 18 18 18a03 in size (hereinafter, a0 is the unit-cell parameter of the

parent phase). It was predicted in [19] that the structure with a quadruple period along one of
the four-fold axes must have the minimum energy. Purely coulomb interactions between point
ions were taken into account in [[20] . The charge of the ions was assumed to be equal to their
maximum valence, and a cluster 6 6 6a0 in size was considered. As a result, the stability
3
of stoichiometrically ordered structures, characteristic of heterovalent ordering cations, was

justified. However, the model proposed in [[20] made it impossible to explain why
PbMg1 / 3 Nb2 / 3O3 and BaMg1 / 3 Nb2 / 3O3 are characterized by different ordered structures.
The other studies in this line of research [[21],[22] have similar drawbacks.
In what follows, we used a phenomenological approach to solve the problem of the most
stable ordered structure. As in all studies on the ordering in solid solutions known to us [[19]-
[22], we will restrict ourselves to the approximation of effective pair interactions whose form
is not specified in deriving the general relations. Within this approximation, we will calculate
the coefficients in the Landau potential ( ), which describes the ordering in terms of the
(i )
pair interaction energies of ions located at different distances.

It should be noted that in the approximation of effective pair interactions the ordered
structure formed at the highest temperature ( T = Tord ), remains stable in the entire
temperature range from T = Tord to T = 0 [[23] , which makes all calculations much easier.
Thus, our problem is reduced to the calculation of the dependence of the coefficient in the
Landau potential for the squared order parameter on the effective pair interactions. It is this
coefficient that determines the temperature of the loss of stability of the high-temperature
phase with respect to the stabilization of the order in the 1 arrangement of cations in the
1
RPS Oh .
3.1.1. Conditions for the Stabilization of Ordered State Observed in PbMg1 / 3 Nb2 / 3O3
The ordering of cations of different valence, which occurs in lead-containing relaxors of
/ +2 // + 5 / //
the Pb( B1 / 3 B2 / 3 )O3 type is characterized by the alternation of B and B ions according
/ // / //
to the scheme 1 : 1 ( ...B B B B ... ) along the three-fold symmetry axis of the parent phase.
In terms of the Landau theory, such ordering is characterized by the one-arm star vector
k13 = (b1 + b2 + b3 ) / 2 [[25] and the one-component order parameter denoted by R [[26].
The ordering described by R , leads to the partition of the - sublattice of perovskite into two
/
equivalent sublattices (1 and 2). The energy of effective pair interactions between B and
B // ions with allowance for the partition into sublattices, can be written as:
E2 = {VB / 1B / 1 ( PB2/ 1 + PB2/ 2 ) + VB // 1B // 2 ( PB2// 1 + PB2// 2 ) + VB11B / 2 PB / 1 PB / 2 + VB // 1VB // 2 PB // 1 PB // 1 PB // 2 +

N
+ VB11B // 1 ( PB / 1 PB // 1 + PB / 2 PB // 2 ) + VB / 1B // 2 ( PB / 1 PB // 2 + PB / 2 PB // 1 )}
2
(3.1)
where N is the number of sites in the RPS 1(a). Each separate term in (3.1), for
example, V B / 1B / 2 PB / 1 PB / 2 , is a double sum (over all sites of sublattices 1 and 2 and the
coordination spheres of each site [[24]):
4
V B / 1B / 2 PB / 1 PB / 2
N2
V
i1, j2 r1i = r2 j
B / 1i , B // 2 j
(| r1i r2 j |) PB / 1i PB // 2 j (3.2)
of effective pair interaction energies VB / 1i , B // 2 j r1i r2 j ( ) - of B / and B // ions, located,

respectively, in sites 1i and 2 j , which belong, respectively, to sublattices 1 and 2. Here
PB / 1i - is the probability of occupation of site i, belonging to the first sublattice, by B / ions.
The approximation used in (3.1) assumes for all probabilities that
i, j , k , l : PB / 1i = PB / 1l , PB // 2 j = PB // 2 k . This circumstance allows us to introduce the
/ //
concept of a probability for a B (or B ) ion to be located at any site of sublattice 1 or 2:
PB / 1 , PB / 2 , PB // 1 , PB // 2 . It is also assumed in (3.2) that the energy of pair interactions between
(
ions VB / 1i , B // 2 j r1i r2 j ) depends only on the distance between them. Thus, V B / 1, B / 2
are
averaged and normalized per atom sums of the energies of effective pair interactions of all B
//
ions/, located in sublattice 1, with all B ions located in sublattice 2. Generalization of this
consideration to all other terms in (3.1) is obvious.
/ //
The interaction energy of B and B in (3.2) atoms in (3.2) is given in terms of local
(generalized) coordinates of the state of the crystal PB / 1i , PB // 2 j ... . In (3.1), the generalized
coordinates PB / 1 , PB // 1 averaged over sublattices 1 and 2 are collective variables. The

generalized coordinates PB / 1 , PB // 1 can be related to the symmetric coordinates; i.e., the
Landau order parameters R [[24],[28] and Frenkel order parameter ( x ) [[24],[26],[30]:
PB // 2i = 1 PB / 1i ; PB // 2 j = 1 PB / 1 j ; (3.3a)
PB // 1 = 1 PB / 1 ; PB // 2 = 1 PB / 1 ; PB / 1 P1 ; PB / 2 P2 ; (3.3b)
x = ( P1 + P2 ) / 2 ; R = ( P1 P2 ) / 2 (3.3c)
In the homogeneous state x = 2c . However, the interaction energy of B and B

/ //
ions may lead to the decomposition of oxides with perovskite structure into two phases
[[24],[26],[27]. In the case of the decomposition, the order parameter x determines both the
/
concentration of B ions in each phase and the fractions of averaged perovskite cells
belonging to the new phases [[21],[29]. Substituting designations (3.3) into (3.1) and (3.2)
/ //
we can introduce into consideration the ordering energies of B and B ions at each
coordination sphere:
( ) ( ) (
ukk / ij = VB / kiB / k / j ri rj + VB // kiB // k / j ri rj 2VB / kiB // k / j ri rj ) (3.4)
where k , k are the numbers of sublattices and the value of ri rj determines the number
/
of the coordination sphere. The local ordering energies (3.4) are averaged according to
designations (3.2) and yield the effective ordering energies u (R) . With u (R) , the quadratic
part of the Landau potential takes the form:
(21) (k13 ) = 2 x 2 + a1R R 2 = 1 ( P12 + P22 ) + K 2 P1P2 (3.5)
With allowance for the structures of the parent phase and the ordered state, 2 and a1R
can be written as:
2 = 3u (1) + 6u ( 2 ) + 4u ( 3 ) + 3u (2) + 12u ( 5 ) + 12u ( 6 ) + .... + 4u (2 3 ) + ...

(3.6) a1R = 3u(1) + 6u( 2 ) 4u( 3) + 3u(2) 12u( 5 ) + 12u( 6 ) + .... + 4u(2 3) + ...
(3.7)
In (3.6) and (3.7) the distance between ions is expressed in terms of the edge lengths a0
of the parent phase unit cells. According to (3.5), it follows from the Landau theory that, at
T 0 , 2 < 0 , a1R > 0 a complex oxide decomposes into two phases of different
composition. If 2 > 0 , a1R < 0 , the ordering of the 1 : 1 type (observed in
PbMg1 / 3 Nb2 / 3O3 ) occurs in complex oxides with perovskite structure.
3.1.2. Conditions for the Stabilization of Ordered State Observed in

Ba( Mg1 / 3 Nb2 / 3 )O3
Ba( B1//+32 B2///+35 )O3 The ordering of the described by the order parameter whose
translational characteristics coincide with those of cos[2 (b1 + b2 + b3 )r ]/ 3 [5]. The
ordered structure is characterized by the presence of identical packets consisting of three
/
planes. One of the packet planes contains predominantly (with the probability P1 ) B ions
and the other two planes (with equal probabilities P2 = P3 ) contain predominantly B
//
ions.
It is convenient to represent the Landau potential describing such ordering as a function of the
two components of the effective order parameter
( 1 ~ cos[2 (b1 + b2 + b3 ) r ]/ 3; 2 ~ sin [2 (b1 + b2 + b3 ) r ]/ 3 ) and x .
Repeating the calculations of the previous section, we obtain the part of the Landau
potential quadratic in the components 1 , 2 in the form:
(22) k 9 ) = 2/ 2 + a1 (12 + 22 ) = Q1 ( P12 + P22 + P32 ) + Q2 ( P1 P2 + P2 P3 + P3 P1 )

(3.8)
Here, the quantity ( P1 + P2 + P3 ) / 3 plays the same role of the nonequilibrium

/
concentration of B ions as x in (3.5), accordingly 2/ = 2 (ui (r )) (3.6). The relation
between the generalized state coordinates P1, P2, P3 averaged over each sublattice and the
components of the effective order parameter 1 and 2 is given by the expressions:
1 = (2 P1 P2 P3 ) / 6 ; 2 = ( P2 P3 ) / 2 ; (3.9)
The ordering in Ba ( Mg1 / 3 Nb2 / 3 )O3 occurs when 2 = P2 P3 = 0 . On the basis of

(3.8) and (3.9), we obtain:
1
a1 = Q1 Q2 = 3u (1) + 3u ( 2 ) u ( 3) 3u (2) + 6u ( 5 ) + ... 0.5u (2 3 ) + ...
2
(3.10)
3.1.3. Conditions for the Stabilization of Ordered State Observed in Y1 / 3 Ba2 / 3CuO3
In the copper-containing oxides with perovskite structure of the R1 / 3 Ba2 / 3CuO3 type,
the arrangement of the planes occupied by rare-earth R and Ba ions is of the 1 : 2
type: ...RBaRBaRBa... . These planes are perpendicular to the four-fold axis of the parent
1
phase. The ordered R1 / 3 Ba2 / 3CuO3 state is described by one component of the six-
component order parameter [[32]:
1 = 2 = ... = 5 = 0, 6 0 (T ) cos(2 )b1r / 3 (3.11)
The quadratic part of the effective Landau potential, dependent on
1
Real phase transitions to the state of equiprobable distribution of R and Ba ions over the parent phase cell in
R1 / 3 Ba2 / 3CuO3 compounds have not been observed. Nevertheless, different models describing various
YBa Cu O
2 3 7 y
properties of on the basis of the concepts about the parent phase in the form of an ideal
perovskite have been multiply considered [[31],[32]]. In the general case, the model of a parent phase or a
virtual crystal does not imply necessary realization of real phase transitions.
= ( P1 + P2 + P3 ) / 3; 5 = ( P2 P3 ) / 2 ; 6 = (2 P1 P2 P3 ) / 6 (3.12)
takes the form:
(23) = 2// 2 + a1 (25 + 26 ) (3.13)
at 2// = 2 (6) and
a1 = [3u (1) 4u ( 3 ) + 3u (2) + 12u ( 6 ) + ... 2u (2 3) + ... ] / 2 (3.14)
The condition for the stabilization of the order observed in Y1 / 3 Ba2 / 3CuO3 is expressed
by the inequality a1 < 0 .
3.1.4. Conditions for the Stabilization of the Suggested Order in Arrangement of

Cations in A -sublattice in Ca1 / 3 La2 / 3 MnO3
The unit cell parameters of Ca1 / 3 La2 / 3 MnO3 in the ordered state [[1] suggest that the
stable order is reached as a result of several transitions of different physical nature. The
structure of Ca1 / 3 La2 / 3 MnO3 can be used to establish two order parameters, which, even in
the first-order approximation, determine the difference of all real structures of
A1//23+ A2///33+ MnO3 compounds from the ideal perovskite structure. One of these order
parameters is determined by the three-arm vector star k11 = (b2 + b3 ) / 2 , and the other order
(1)
parameter is determined by the three-arm vector star k10 = b1 / 2 . In both cases, the
(1)
symmetry of the order parameter is described by the group L = C 48 [[24], [33]]. The part
of the Landau potential (which describes both types of ordering), quadratic in the components
of the order parameter, has the form:
(24) = 2/// x 2 + a1x ( X 12 + X 22 + X 32 ) + a1M ( M 12 + M 22 + M 32 ) (3.15)
Here 2/// = 2 (3.6), X i are the components of the order parameter describing the
doubling of the parent-phase period along the four-fold axes, and the translational
(1)
characteristics M i are determined by the rays of the star k10 [[24] . Calculation analogous to
the previous one leads to the expressions:
a1x = u (1) 2u ( 2 ) 4u ( 3 ) + 3u (2) + 4u ( 5 ) + ... + 4u (2 3 ) + ... (3.16)
a1M = u (1) 2u ( 2 ) + 4u ( 3 ) + 3u (2) 4u ( 5 ) + ... + 4u (2 3 ) + ... (3.17)

The conditions for the stabilization of each considered type of order in the arrangement of
cations at low temperatures follow from (3.16) and (3.17).
4. CONSIDERATION OF RESULTS
4.1. Pair Interactions and Coordinate Spheres
To achieve actual results on the base of (3.6, 3.7, 3.10, 3.14, 3.16, 3.17) first of all it is
necessary to solve the question of choosing the type of model potential of pair interactions.
The interaction between two non-bonded particles called the van der Waals potential
is expressed as a function of the distance between the particles. The common form of this
function is the following:
C
U VDW (r ) = DVDW e Br (4.1)
r
where DWDV , B, C are phenomenological parameters. There are other functions, which are
traditionally taken as approximations to the van der Waals potential. The most frequently
used are the Lennard-Jones, the Rydberg and the Morse potentials:
R 12 R
6
U LJ (r ) = D LJ 2 (4.2)
r r
U R = D R [1 + R (r R )]e R ( r R ) (4.3)
[
U M ( r ) = DM 1 e M ( r R ) ]
2
(4.4)
Here Di , i are phenomenological parameters, and R is the distance between the

particles, corresponding to the minimum energy of their interaction. Such potentials, mostly
because of their simplicity, are widely used in different calculations, e.g. in such computer-
intensive applications, as in [[34]-[42].
As all these functions approximate the energy of the same interaction, in main aspects the
differences between them are nonessential. The explicit relationships between some
generalizations of these functions are demonstrated in the papers [[43],[44]. To consider
general problems which are observed in this chapter it is enough to confine ourselves to
investigation of one representative of this class of pair interactions potentials. As a
representative it is convenient to choose the (1) potential without the DVDW multiplier, that
does not play any role in the comparison of the Landau potential coefficients.
Another way of taking into account the particles interaction is the pseudopotential
method, that as its known [[45]] may be reduced to the approximation of the effective pair
interactions. As an approximation we chose the following sum:
1 Acos(qr)
U PP (r ) = - + (4.5)
r6 (qr) 3
The first summand here is stipulated by the interaction of fluctuations of atom dipole
moments. And the second summand takes the form predicted in the pseudopotential's
approximation [[45]].
The second important problem which must be solved to proceed to the computations is
limitation of the accounting coordinate spheres number. There are two functions expressing
the energy of state Hamiltonian and Landau potential. In the second order respect to
generalized coordinates these functions can be written as
n
H=
k i 0
ik Pi Pk (4.6)
m
= j 2j (4.7)
j =0
where Pi is probability of occupation of sublattice i by appointed ion, j are the Landau

order parameters (symmetric coordinates). These two forms of energy are equivalent. There
are linear relationships between symmetric coordinates and probabilities of sublattices
occupation which leads to the linear relationships between coefficients ik and j . So the
numbers of ik and j must be equal. It is easy to see that to fulfill this condition it is
necessary and sufficient to take into account all pair interactions up to the interaction between
two nearest ions situated in equivalent positions. In our case (with the possibility of tripling
the lattice period) it means that we must take into account 11 coordinate spheres.
4.2. ORDERING CONDITIONS IN A BULK CRYSTAL

The Figure 7(a,b) displays the areas in the space of potentials (4.1) and (4.5) parameters,
corresponding to the different types of ordering. The figures 7(a,b) were based on calculations
that were made formally, without accounting the realistic restrictions on the parameters'
values. So in fact only the segments of the phase diagrams shown on these figures have the
physical meaning.
Figure 7. Phase diagrams of bulk sample of OPS in space of pair interaction potentials parameters: (a)
van der Waals potential (4.1), (b) pseudopotential approximation (4.5).
Figure 8. Phase diagrams of near-surface layer of OPS in space of pair interaction potentials
parameters: (a) van der Waals potential (4.1), (b) pseudopotential approximation (4.5). Grey lines
shows borders between phases in bulk sample.
Phenomenological parameters of the potential can vary with the external conditions. For
example, obviously, with the increase of the external pressure the distances between the
particles, corresponding to the minimum potential energy ( R parameter in (2)-(4) or
BC
2W / B for the potential (4.1), where W(x) is the Lambert W function) are usually
2

decreased. Without going into details of definite type of parameters of model potentials of the
pair interactions dependence on external conditions, let us keep in mind the fact that this
dependence can be determined. This means that it is possible to construct thermodynamic
paths, corresponding to the defined variations of the external conditions on the phase
diagrams in Figure 7(a,b). The conditions of ordering type variations can also be established.
When comparing the Figure 7(a) and (b) it is seen, that in the Figure 7(b) there is a border line
between two areas corresponding to the two phases of Ca1 / 3 La2 / 3 MnO3 while there is no
such line in the Figure 7(a). The fact that such variation of the ordering occurs on the
experimental phase diagram is a strong argument in favor of a better match reality of the
pseudopotential approximation in comparison with the model potentials (4.1)-(4.4).
4.3. ORDERING IN A NEAR-SURFACE LAYER OF A CRYSTAL

The effect of variation of the conditions of phase transition in the near-surface layer of a
crystal relating to the conditions of phase transition in a bulk crystal is well known. For the
first time such effect was estimated for magnetic phase transitions [[46], and later for
structural phase transitions [[47]. Similar effect is observed also for the transitions like
ordering in a near-surface layer [[48].
The expressions for the coefficients of the Landau potential in terms of the energies of
pair interactions (3.6, 4.7, 4.10, 4.14, 4.16, 4.17) are valid only for the description of ordering
in a bulk sample. For the positions of the SRP, located near the crystal surface, coordination
spheres of radius r > h , where h is the depth of the position relating to the crystal surface,
situated partially outside the crystal boundaries, and total energy of interaction between the
ion, located in the considering position, and the ions, located on the coordination sphere of
such radius, will differ from the similar total energy for the position, located at depth H > r .
To describe the ordering of cations in a near-surface layer correctly it is necessary to take this
difference into account in a course of total interaction energies calculation, corresponding to
several of the initial coordination spheres. Figure 8(a,b) shows the areas of spaces of the
potentials coefficients (4.1) and (4.5), where the type of ordering in a near-surface area of a
crystal should differ from the one in a bulk sample. In Figure 8(a,b) there are also border lines
of areas, corresponding to the different types of ordering in a bulk crystal. Among others, on
the Figure 8(a,b) it is seen, that some types of ordering display high stability relating to near-
surfaces effects, while others tend to the strong narrowing of the area in the space of model
potentials parameters, in which they are energetically favorable in near-surface layer.
REFERENCES
[1] Salamon M.B. Reviews of Modern Phys. 2001, vol.73. p.583.
[2] Bednorz G., Mller K. A., Z. Phys. B, 1986, vol. 64, p.189.
[3] Drake A., Abell J. S., Sutton S. D., J. Less. Comm. Met., 1990, 164-165, 187.
[4] Burns Gerald, Dacol F.N. Solid State Communications. 1983. vol.48. no. 10. P.853.
[5] Davies P.K., Tong J., Negas.T. Jorn. Of Am.Ceram.Soc. 1977. vol.80.no7. P.1727.
[6] Yan Y., Pennycook, Xu Z., Viehland D., Applied Physics Letters.1998, vol.72. no.24.
p.3145.
[7] Kolesova R. V., Kolesov V. V., Kupriyanov M. F., Lavrova O. A.Bulletin of the
Russian Academy of Sciences - Physics. vol.64 no.6,pp. 883-885.
[8] Kolesova R., Kolesov V., Kupriyanov M., Skulski R. Phase Transitions, 1999, vol
.68. P.621.
[9] Kupriyanov M., Kogan V., Ferroelectrics. 1991. vol.124. p.213.
[10] Lebedinskaya A.R., Kupriyanov M.F., Skulski R., Material Science and Engineering
, 2001. vol. 83, p.119.
[11] Vakhrushev S., Zhukov S., Fetisov G., Chernyshov V., J.Phys. Condens.Matter.
1994. vol.6. p.4021.
[12] Vakhrushev S., Nabereznov A., Sina S.K., Feng Y.P., Egami T. J. Phys. Chem.
Solids. 1996. vol.57. no10. P.1517.
[13] Bonneau P., Garnier P., Calvarin G. J.of Solid State Chemistry. 1991. vol.91. p.350.
[14] Bokij G.B., Proc. Nat. Inst. Sci. India Part A 24 (1958), pp 142-150, available
online:
http://www.new.dli.ernet.in/rawdataupload/upload/insa/INSA_1/20005ac9_142.pdf
[15] Jona F., Shirane G., Ferroelectric Crystals; Pergamon Press: Oxford and London,
1962.
[16] K. Yu. Gufan , Physics of the Solid State, vol. 47, no. 3, pp. 459465.
[17] M. Gelfand, Vul E. B., Ginzburg S.L., Fedorov Yu. G., The Ravine Method in the
Problems of the X-Ray Structural Analysis [in. Russian]; Nauka: Moscow, 1966.
[18] Yan Y., Pennycook, Xu Z., Viehland D.Applied Physics Letters,1998.vol.72. no.24.
p.3145.
[19] V.I.Zinenko, S.N.Safronova, Physics of the Solid State, 2004, vol.46, no.7, p.1291.
[20] Bellaiche L., Vanderbilt D., Phys.Rev.Lett.1998. vol.81. no.6. p.1318.
[21] Burton B.P., Cockayne E. Phys.Rev.B.1999. vol.60. no18. P.R12542.
[22] Burton B.P. Phys.Rev.B. 1999 . vol.59. no9 . P.6087.
[23] Smart J. S., Effective Field Theories of Magnetism; Saunders: Philadelphia, 1966.
[24] Gufan Yu. M., Structural Phase Transitions [in Russian], Nauka: Moscow, 1982.
[25] Kovalev O. V., Irreducible and Induced Representations and Corepresentations of
the Fedorov Groups [in Russian], Nauka: Moscow, 1986.
[26] Bouckaert L.P., Smoluchowski R., Wigner E., Phys. Rev. 1936, vol. 50. p. 58.
[27] Gufan A.Yu., Klimova E.N., Prus Yu.V., Stryukov M.B., Bulletin of the Russian
Academy of Science: Physics, 2001, vol.65 no.6, pp.850-854.
[28] Landau L.D., Lifshitz E.M., Statistical Physics, 3ed, Pergamon Press: Oxford and
London, 1980.
[29] Gufan A Yu. Bulletin of the Russian Academy of Science: Physics, 2005, vol.68 no.5,
pp.734-739.
[30] Frenkel Ya. I., Introduction to the Theory of Metals [in Russian], 3d. ed. Fizmatgiz:
Moscow, 1958.
[31] Toldano P., Dmitriev V., Reconstructive Phase Transitions, World Scientific:
Singapore, 1996.
[32] Gufan Yu M, JETP Lett., 1968, no. 8, p.167.
[33] Stokes H. T., Hatch D. M. Isotropy Subgrpoups of the 230 crystallographic space
groups. World Scientific: Singapore, 1988.
[34] Lifson S., Haggler A. T., Dauber P., J. Amer. Chem. Soc. 1979, vol. 101, p. 5111.
[35] Brooks R., Bruccoleri R. E., Olafson B. D., States D. J., Swaminathan S., Karplus
M., J. Comput. Chem. 1983, vol. 4, p. 187.
[36] van Gunsterenm W. F. , Berendsen H. J. C., Groningen Molecular Simulation
(GROMOS) library manual, 1987.
[37] Clark M., Cramer R. D. III, van Opdenbosch N., J. Comput. Chem, 1989, vol. 10, p.
982.
[38] Mayo S. L., Olafson B. D., Goddard W. A. III, J. Phys. Chem, 1990, vol. 94, p. 8897.
[39] Allured V. S., Kelly C. M., Landis C. R., J. Amer. Chem. Soc. 1991, vol.113, p.
9493.
[40] Rappe K., Casewit C. J., Colwell K. S., Goddard W. A. III, Skiff W. M., J. Amer.
Chem. Soc. 1992, vol. 114, p. 10024.
[41] Nemethy G., Gibsen K.D., Palmer K. A., Yoon C.N., Paterlini G., Zagari A., Rumsey
S., Sheraga H. A., J. Phys. Chem., 1992, vol. 96, p.6472.
[42] Cornell W. D., Cieplak P., Bayly C. I., Gould I.R., Merz K. M. Jr., Ferguson D. M.,
Spellmeyer D. C., Fox T., Caldwell J.W., Kollman P. A., J. Amer. Chem. Soc. 1995,
vol. 117, p.5179.
[43] Lim T. C., (2003) Zeitschrift fuer Naturforschung, A: Physical Sciences, 2003, vol.
58 no. 11, p. 615.
[44] Lim T. C., Acta Chim. Slov. 2005, vol. 52, 149152.
[45] Mott N.F., MetalInsulator Transitions; Taylor and Francis Ltd.: London, 1974.
[46] Mills D. L., Phys. Rev. B 1971, no. 3, p. 3885.
[47] Binder K., Hohenberg P. C., Phys. Rev. B 1974, no. 9, 2194.
[48] Dosch H., Critical Phenomena at Surfaces and Interfaces, Springer Tracts in Modern
Physics Vol. 126, Springer-Verlag: Berlin, 1992.
Chapter 11
ABX3-TYPE OXIDES AND HALIDES:

THEIR STRUCTURE AND PHYSICAL PROPERTIES
A. S. Verma* and V. K. Jindal

Department of Physics, Panjab University Chandigarh, 160014, India
ABSTRACT
The term `perovskite` is used to denote a category of inorganic crystalline solids with
the general formula ABX3, where A and B are cations and X (oxides or halides) is an
anion. The perovskite structure occupies a prominent place under all the known ternary
systems of composition ABX3. This is due not only to its wide occurrence, but also to a
series of interesting and useful properties associated with this structural type.
Crystallographic behaviour of perovskites is of great interest because most structures are
close to an ideal cubic structure, however, frequently they are slightly distorted resulting
in structures with lower symmetry such as orthorhombic, tetragonal, rhombohedral,
trigonal systems and so on. These lattice distortions and structure changes are governed
by temperature, pressure, chemical composition, and in some cases, electric field. As a
rule, with rising temperature, perovskites tend to undergo a series of transitions to
progressively higher symmetry, culminating in the cubic structure where experimentally
accessible. They are widely investigated, not only because of interest in their crystal
structural behaviour, but perovskites also exhibit a variety of interesting electronic,
electromechanical and conductive properties, which are the basis for many existing and
potential applications. Such properties are often symmetry dependent. In this chapter, a
summary of the scientific work performed on the ABX3 (X = Oxides and Halides) system
during the last few years is presented.
*
Corresponding author: e-mail: [email protected], Mob: +91 9412884655, Phone: +91 565 2423417
464 A. S. Verma and V. K. Jindal
1. INTRODUCTION
The ternary nature appears at first to be clearly defined by the chemical composition. It is
therefore more meaningful to speak of ternary structures (i.e. an anion and two significantly
different cation sites within the structure) than it is to speak of ternay phases in the purely
chemical sense (i.e. containing three different elements). Ternary structure is not comparable
with ternary phase. For ceramists, other materials scientists and earth scientists the ternary
structural families hold a specify the significance. To the best of our knowledge, most of the
solid earth is made up of a material with a ternary structure. In modern technology we are
concerned with large quantities of refractories, or with ferroelectric devices or computer-
memory cores. In this chapter we hope especially to be able to point out the interrelationships
between the various structral types, the transitions from one structure to another, and physical
properties. The rational design of advanced materials assumes an understanding of the
correlation between basic crystal chemistry and simple structure-property relationships.
Implicit in such a relationship is an understanding of the connection between chemical
composition and crystal structure in solid-state materials. Properties such as electronegativity,
radii, valence electron count, and principal quantum number can help establish a link between
composition, structure, and properties for different classes of materials. Establishing such a
link, however, is particularly difficult due to the large number of parameters which affect
structures including the charge and coordination preferences of metals, the existence of
energetically close-lying structure types, and synthetic parameters, e.g., pressure and
temperature. The use of high-pressure or low-temperature synthesis complicates the problem
through the introduction of kinetic considerations and the possibility of metastable phases.
Numerous attempts have been made to correlate structural stability with chemical or physical
variables derived from atomic properties of the constituent elements; however, for different
classes of materials, different atomic properties must be used to achieve good corre1ation [1-
6].
Although several predictive methods exist for ionic compounds [2,4-7], each has
noticeable shortcomings. Very few systems allow the study of the relationship between
structural changes and physical properties in such a clear way as ABX3 perovskites (X =
oxide and halide) [4-10]. In this review, we present past approaches that have been used for
predicting composition-structure relationships for ABX3 compounds.
Synthesis
Several methods have been described for the preparation of perovskite-type oxides. Many
of them are analyzed in detail in several references [11, 12] and for this reason only a brief
account on this aspect will be given here. Practically all the standard methods known for the
synthesis of solid materials [13] can be applied to the preparation of perovskite-type oxides.
The classical route is the reaction between intimate mixtures of the constituent binary oxides
or their precursors (nitrates, carbonates, etc) in the solid state at high temperatures. These
reactions are slow as a result of large diffusion distances, and repeated grinding and heating
are required in order to obtain pure and fine powders. In some cases, special precautions must
be taken. For example, compounds containing Pb (II) in the A-site of the perovskite structure
ABX3-Type Oxides and Halides 465
are often difficult to obtain due to the volatility of lead oxide. This problem can be overcome
by heating the reactants in a lead oxide atmosphere, by using an excess of lead oxide or by
working at relatively low temperatures in the primary reaction phase before the final high
temperature treatment. In some cases, other structural types can be stabilized in the
intermediate reaction stages, which sometimes are difficult to transform later into the desired
perovskite phase.
Other useful methods are the so-called "precursors-techniques" in which usually the
carbonates of the metals were coprecipitated in adequate proportions and subsequently
decomposed by heating [14]. This method allows obtaining a better and more homogeneous
starting mixture and, usually, a lowering of the final reaction temperature. Similar procedures
involve the coprecipitation of oxalates or citrates and other similar materials, a technique that
recently has been often used for the synthesis of the new high Tc, superconducting oxides
[15]. Other procedures often used in the last years are the "Sol-Gel" methods [16-18] in which
the hydrolysis of certain chemical species under controlled conditions is used to generate the
necessary precursors for the synthesis of the oxide materials. The most common systems are
metallic alkoxides, which are very useful because of their ability to form homogeneous
solutions in a variety of solvents and in the presence of other alkoxides or metallic
derivatives. These solutions are then transformed into gels which by desiccation or heat
treatment produce glass or ceramic materials. Among non-alkoxide precursors, nitrates,
carboxylates, acetylacetonates, chlorides and other inorganic species are appropriate to be
used in Sol-Gel preparations.
Structure
The ABX3 general group special attention has been paid to those compounds with the
ideal cubic perovskite structures or with structures which are slightly distorted from the
former, either by cation displacements, tilting of BX6 octahedra or distortion in the shape of
the X6 group which surrounds the B cation. More recent attention has been paid to the group
formed by compounds with hexagonal-close-packed structures which are closely related with
the cubic ones [19]. The ideal perovskite (ABO3) structure (Figure 1) has a cubic unit cell of
side about 3.9 , space group Pm3m and contains one formula unit. The B ions have an
octahedral oxygen coordination and the A ions have 12-fold coordination. The oxygen ions
are linked to six cations (4A + 2B). If the unit cell is chosen with a B ion at the body centre,
then the oxygen ions occupy the face centers and the A ions occupy the comers of the unit
cell. When A ion is assigned the body centre position, the oxygen ions are at the middle of the
edges and B ions at the comers of the unit cell. The structure can be visualized in several
ways [20]. The most useful approach is to consider that it is based on that of ReO3 with comer
sharing octahedra arranged along the cube axis with the A ions occupying the interstices. The
structure can also be viewed as a body-centered cubic (b.c.c.) arrangement of the A and B
cations (as in CsCI) with the oxygen ions added at the centers of the edges. It can also be
considered as a face-centered close-packed AO3 array, as in the ordered Cu3Au structure, with
B ions occupying the octahedral hole in the centre of the unit cell (Figure 2).
Figure 1.
In the ideal structure, where the atoms are just touching one another, the B-X distance is
equal to a/2 whereas the A-X distance is 2(a/2) (a = cubic unit cell length) and the following
relation between the ionic radii holds:
(rA + rX) = 2 (rB + rX)
Notwithstanding, it was found that the cubic perovskite structure, or slightly distorted
variants of it, was still retained in ABX3 compounds even where this relation is not exactly
obeyed. As a measure of the deviation form ideality Goldschmidt [21], introduced a
tolerance factor t, defined as,
t = (rA + rX) / 2 (rB + rX)
That is applicable at room temperature to the empirical ionic radii.

For an ideal perovskite, t is unity; however, the perovskite structure is also found for
lower t-values (0.75 < t 1). In such cases, the structure distorts to tetragonal, rhombohedral
or other lower symmetries. Indeed, many perovskite oxides are polymorphs. In ideal
perovskite structure, the atoms are in the following positions; A; 000, B; X; 0,
0 , 0 . Moreover, the above given geometric relations are not a sufficient condition for
the stabilization of the perovskite lattice, since the A and B cations must, in themselves, be
stable in twelvefold and sixfold coordination respectively. This latter condition sets lower
limits for the cationic radii. In oxide systems these limits are rA > 0.90 and rB > 0.51 [19].
Cu3Au Structure ABO3 Structure
Figure 2. Perovskite structure and its relationship to CsCl, ReO3 and Cu3Au structures.
The concept of oxide and halide analogue compounds was first introduced by
Goldschmidt [21], on the basis of a number of crystal chemical considerations: (a) the
similarity in ionic radii of O-2 and F-; (b) the applicability of the rigid ion model to
compounds containing the O-2 and F- ions, which have relatively low polarizabilities resulting
in their ionic radii being almost independent of coordination number; and (c) the
correspondence of the crystal structures of oxide and halide compounds in which the cations
are also of comparable ionic radii. Notable examples of this modeling concept are LiF-MgO
(rocksalt structure) CaF2 ThO2 (flurite structure), MgF2 TiO2 (rutile structure).
Goldschmidt [21], also suggested that because of their lower ionic charge fluorides should be
weakened models of their oxide analogues and thus be characterized by lower melting
temperature, lower hardness and lower refractive index.
Some oxide and halide materials that adopt this ideal structure are BaPbO3, CaZrO3,
PbHfO3, SrSnO3, CdTiO3, PbSnO3, KMgF3, KMnCl3, CsPbBr3, CsSnI3 etc. [22, 23]. Besides
chalcogenides and halogenides, there are a few examples of perovskite hydrides (X=H). In
the case of A and B being a monovalent alkali metal (M) and a divalent alkaline earth metal,
respectively, the ternary metal hydrides have a saline bonding character. Such hydrides are of
interest for hydrogen storage at high temperature. There are five reported hydrides with the
ideal Pm3m perovskite structure: CsCaH3, RbCaH3, KMgH3, BaLiH3, and SrLiH3 [24-26].
The letter is sometimes referred to as inverse perovskites since the singly charged M cation
occupies the octahedral voids. Additionally, it has been proposed that the light metal hydride
LiBeH3 adopts the cubic perovskite structure and displays high-temperature
superconductivity [27]. It is interesting to remark that many compounds present the ideal
cubic structure only at high temperature.
As the most important components in the Earths mantle, ABO3 (orthorhombic)
perovskites have been intensively studied [7,8-10]. The understanding of the mechanisms that
lead to the stabilization of the various perovskite structures has been a long-standing problem
in material science. Due to the frequent occurrence of perovskites, and also to the great
variety of properties presented by these systems, the accurate and transparent description of
the driving forces toward the high or low symmetry crystal structures is of general interest
[28]. The orthorhombic structures are derived from the ideal cubic structure via the titling and
distortion of the BO6 octahedra [5]. The degree of distortion of these perovskites can be
correlated to the tilting of BO6 octahedra, and even where a structural transition is not
observed the tilt angle decreases as the temperature increases. The octahedral tilting,
originating from finite rotations of the BO6 octahedra, is realized by preserving the central
symmetry of the B cation, while the A cation prefers off-center displacements. A number of
workers have used such changes to infer the presence of thermally induced phase transitions.
From the crystal chemistry point of view, the two different perovskites of ABO3 compounds
have very similar local structures, in which each B atom is octahedrally co-ordinated by O
atoms, as shown in Figure 3 [2, 29-34]. Orthorhombic perovskite is basically built up from
the corresponding cubic perovskite.
The rare earth ortho-ferrites RFeO3 (R is a rare earth element) have a distorted perovskite
crystal structure [35]. All the Fe+3 ions are crystallographically equivalent and lie in an
octahedral oxygen environment. Each iron ion is surrounded by six Fe+3 ions and the FeO
Fe bond angle is close to 180 [36]. This results in a strong negative indirect Fe+3OFe+3
exchange interaction, which causes the formation of two magnetic sublattices whose moments
are almost anti-parallel. The small angle between the moments of these sub-lattices produces
a small ferromagnetic moment [37, 38]. The rhombic ortho-ferrites RFeO3 (space group
Pbnm) are usually attributed to the perovskite-like structures obtained as a result of all the
possible distortions of the cubic CaTiO3 perovskite lattice (space group Pm3m) on replacing
the Ca cations by R and Ti by Fe. These distortions can change the crystal symmetry. A
transition from the cubic perovskite structure to a rhombic structure in the rare earth ortho-
ferrites has been established experimentally [39, 40]. RMO3 compounds with M = Cr, Sc, V,
Ga, and Al also have this type of structure [41-45]. It is interesting to note that the
deformation which develops for the rhombic distortions has almost no effect on the oxygen
octahedron surrounding the iron ions, while the coordination polyhedron of the R+3 cations is
significantly distorted. The average inter-atomic FeO and OO distances are essentially
constant for the entire rare earth series RFeO3, at 2.011 and 2.844 , respectively [39]. As R
is changed, only the mutual positions of the oxygen octahedral change and this leads to a
change in the angle of the FeOFe valence bond. The deviation from cubic symmetry
increases as the R+3 ionic radius decreases; it is smallest in LaFeO3 and greatest in LuFeO3.
Figure 3. Schematic structures of cubic perovskite (a), orthorhombic perovskite (b) of ABO3
compounds large filled blue circles represent A, small yellow B and small red O atoms.
In several materials the cubic cell may have a small deformation to rhombohedral
symmetry. This deformation does not enlarge the unit cell. It is possible to index it on a unit
cell containing either one or two formula units with rhombohedral angles ~ 90 or ~ 60
respectively. However, the anions are generally displaced as requires the larger unit cell with
~ 60.
The best known example of a tetragonal perovskite is probably the room temperature
form of the ferroelectric BaTiO3, with a = 3.994 , c = 4.038 and Z = 1. In this case the
TiO6-octahedra are somewhat distorted (one Ti-O bond at 1.86 , four at 2.00 and one
longer at 2.17 ). Barium is coordinated by four oxygens at 2.80 , four at 2.83 and four
more at 2.88 . In the isotypic PbTiO3, the TiO6-polyhedra are more distorted than in
BaTiO3, a fact which can surely be related to the greater polarizing power of Pb (II), which
has often been discussed in relation to other structural and bonding problems in systems
containing this cation [46]. A number of other tetragonal perovskites (PbHfO3, SrPbO3,
SrZrO3, AgTaO3, KCoF3 etc.) are isotypic with BaTiO3 and possess uni-molecular cells. The
fact that most of them are tetragonal only at high temperatures makes accurate
crystallographic studies very difficult. In some cases, more complex tetragonal unit cells have
been found; also superstructure reflections can be observed in some tetragonal perovskite,
pointing to multiple cells or to cells of lower symmetries [47]. KCuF3 has a tetragonal
superstructure with a = a, c~2a and Z = 4 (a = pseudocubic cell-length) due to a shift of the
fluoride ions, caused by the Jahn-Teller effect. As expected, the CuF6-octahedra are
elongated, with two significantly longer Cu-F distances [48].
As stated in the preceding paragraphs, a great number of perovskite-like materials present
several polymorphic modifications. Some of them are very important in relation to their
physical properties and applications. For example, in BaTiO3 and KNbO3 the following
transformations are observed with increasing temperatures:
rhombohedral orthorhombic tetragonal cubic
and the three lower temperature forms are ferroelectric. It should be noted that these phase
changes reversible and all the polymorphic forms possess a pseudo-cubic unit cell with a~4
. Obviously, these polymorphic changes further complicate the crystal chemistry of this type
of material [47, 48].
2. MODELS FOR THE ANALYSIS OF PROPERTIES

OF PEROVSKITES (ABX3)
2.1. Structural Properties
First, Mooser and Perason [49] applied a two-dimension graphic to study the stability of
different compounds, the two factors they used were the difference of electronegativity
between the cation and the anion and the average principle quantum number. They succeeded
to discriminate the crystal structures of AB-type compounds, of AX2-type halides, and the
metallic or non-metallic ternary fluorides ABX3. Similar methods were called the structural
map technology, and more parameters were used to draw the graphic. Muller and Roy [50]
proposed to plot structural map, which took the ionic radius of A and B as coordinates to
study the distribution of different structures for many ternary structural families. Furthermore,
the schematic distribution map of different crystal structure for A+1B+2X3, A+2B+4X3 and
A+3B+3X3 systems separately, were given by the same method [50, 51]. However, the criterion
for perovskite formability was not discussed, possibly due to the lack of accurate data of
crystal structure of some ABX3 compounds at that time. The ionic radius is the most
important ionic parameter that dominates the crystal structure of ionic compound. The ideal
cubic structure form may be seen as a network of BX6 octahedra, where all the octahedra are
corner sharing. The A ions occupy the cubo-octahedral holes in between the octahedra. Many
derivatives of the ideal structure are found and the distortion of the structure from cubic is
often considered to be determined by the relative sizes of the different ions of the compound.
The much used tolerance factor (t) is defined as
t = rA + rX / 2 (rB + rX) (1)
where rA, rB and rX are the ionic radii of A, B and X3 respectively. t = 1 corresponds to an
ideal perovskite, while a t < 1 indicates tilting or rotation of the BX6 octahedra. A t > 1
indicates preference for hexagonal perovskite structures, where the octahedra at least to some
extent share faces. The face-sharing octahedra then form chains along the hexagonal c axis,
bound to each other by A-X bonds. The face sharing octahedra give relatively short B-B
distances, thus increasing the electrostatic repulsion between the B-site cations. Hexagonal
stacking and face sharing is therefore introduced in stages with increasing size of the A cation.
Alkaline-earth-metal manganese oxides show this importance of size well [52]. CaMnO3
forms an orthorhombic derivative of the cubic structure with slightly tilted octahedra [53]. At
high temperatures SrMnO3 takes a cubic perovskite structure with corner-sharing octahedra
only [54]. At low temperatures, on the other hand, SrMnO3 forms a four-layer hexagonal
structure with both face-sharing and corner-sharing octahedra corresponding to hexagonal and
cubic stacking, respectively. BaMnO3 forms a hexagonal structure with face sharing of
octahedra only. The stabilization of the hexagonal structures at low temperatures has long
been ascribed to the formation of metal-metal interactions across the shared face [55].
However, the Mn-Mn d-orbital overlap has recently been shown to be minimal [52, 56]
Instead covalent interactions between manganese and oxygen atoms are indicated. Both the
relative sizes of the cations and the covalency of the Mn-O bonds appear to be important for
stabilizing the hexagonal structures of SrMnO3 and BaMnO3 at low temperatures [52, 56].
Tolerance factor (t) has been widely accepted as a criterion for the formation of the perovskite
structure, many researchers have used it to discuss the cubic perovskites stability, and
therefore, it is an important factor for the stability of cubic perovskites, so tolerance factor (t)
constructs one axe of the structure map. Another important octahedron factor (rB/rO) [57] is as
important as the tolerance factor to form cubic perovskites, so octahedron factor constructs
another axe of the two-dimension structure map. Which are shown in figure 4.
According to Goldschmidts point, t values of cubic perovskites are in the range of 0.8
0.9. From Figure 4, it is indicated that t values of cubic perovskites, except BaMoO3, are in
the range of 0.8571.032, which is wider than Goldschmidts range. However, t = 0.857
1.032 is a necessary but not a sufficient condition for the formation of the cubic perovskite
structure. The 85 systems, the t values of which are in the range (0.8571.032), can not form
cubic perovskite structure [1]. The correct knowledge of lattice constants of unknown
compounds is necessary for analyzing the structure and properties of materials. This helps in
the selection of an appropriate material in many industries related problems [58]. Some
empirical models were established that can predict lattice constant of perovskites from
selected atomic properties of their constituent elements.
0.9
0.8
octahedral Factor
0.7
0.6
0.5
0.4
0.3
Cubic Perovskite
0.2
Non Cubic Perovskite
0.1
0.6 0.7 0.8 0.9 1 1.1 1.2
Tolerance Factor
Figure 4. Classification of cubic (ABO3) perovskite oxides [1].

Recently, the methodology developed by Jiang et al. [59], which allows one to predict the
lattice constant of cubic perovskites by using the known ionic radii of the cations and anion.
According to him the lattice constant may be determine by the following relation,
a (in ) = 1.8836(rB + rX) + 1.4898 rA + rX 2(rB + rX) 1.2062 (2)
where rA, rB and rX are the ionic radii of A, B and X3 (X =O, F, Cl, Br and I), respectively.
According to Ye et al. [60], the lattice constant of the ideal perovskite oxides (ABO3) can
be linearly correlated to some atomic parameters as,
a = 0.3166rA + 1.422rB 0.1708XA + 0.0562XB 0.0066(ZB ZA) +2.706 (3)
where rA, rB, XA, XB, ZA and ZB are the ionic radii, electronegativity and valence number of
ion A and B, respectively.
Equation (3), depends on the number of valence electrons and ionic charge also depends
on the number of valence electrons, which changes when a metal forms a compound.
Furthermore, perovskite oxide solids can be defined in different systems like as A1+B5+O3,
A2+B4+O3 and A3+B3+O3. Recently [9,10], the authors have presented alternative explanation
for lattice constants of orthorhombic and cubic perovskites in terms of valence electrons and
ionic charges. The lattice constants of cubic perovskite solids exhibit a linear relationship
when plotted against average ionic radii rav (), but fall on different straight lines according
to the valence electron product of the compounds, which are presented in Figure 5. To obtain
better agreement between experimental and theoretical data for cubic perovskite type
compounds, relations (1) and (2) have been extended to [9],
Lattice constant (a in ) = K (V1V2V3)S rav (4)
Figure 5.
0.1
0.09
0.08
0.07
0.06
log rav
0.05
0.04
0.03
0.02
0.01
0
0.71 0.72 0.73 0.74 0.75 0.76 0.77 0.78
log a
Figure 6. Plot of log a (lattice constant in ) against log rav (average ionic radii in ) for A+2B+4O3 and
A+3B+3O3 orthorhombic perovskites and A+2B+4O3 and A+3B+3O3 are represented by and
respectively.
Where V1, V2 and V3 are the valence electron on the A, B and X3 respectively, K and S
are constants, rav is the average ionic radii of ABX3 in and it can be calculated by rav = (rA +
rB + rX) / 3.
The lattice constants of orthorhombic perovskite solids exhibit a linear relationship when
plotted against average ionic radii rav (), but fall on different straight lines according to the
ionic charge product of the compounds, which are presented in Figs. 68. The lattice constant
of orthorhombic perovskite-type crystal structure compounds can be evaluated using their
ionic charge by following relations [10],
lattice constant (a in ) = 4(Z1Z2Z3)0.054 rav (5a)
lattice constant (b in ) = 4(Z1Z2Z3)0.055 rav (5b)
lattice constant (c in ) = 6(Z1Z2Z3)0.040 rav (5c)
where Z1, Z2 and Z3 are the ionic charge on the A, B and O3 respectively, rav is the average
ionic radii of ABO3 in . The correct ionic radii (ra, rb and ro) are taken from Shanons work
[47] and other Ref. [58]. The main advantage of Equations (4) and (5a)(5c) are the
simplicity of the formula, which does not require any experimental data expect the ionic radii
of orthorhombic perovskite solids and the evaluated values are in close agreement with the
experimental and theoretical data as compared to the values reported by previous researchers
so far. The presented method is quite reasonable and can give us a useful guide in calculating
and predicting of the more complex class of cubic and orthorhombic perovskite solids.
0.25
0.2
log rav 0.15
0.1
0.05
0
0.71 0.73 0.75 0.77 0.79
log b
Figure 7. Plot of log b (lattice constant in ) against log rav (average ionic radii in ) for A+2B+4O3 and
respectively.
0.1
0.08
log rav
0.06
0.04
0.02
0
0.86 0.87 0.88 0.89 0.9 0.91 0.92 0.93 0.94
log c
Figure 8. Plot of log c (lattice constant in ) against log rav (average ionic radii in ) for A+2B+4O3 and
respectively.
2.2. Mechanical Properties
Yamanaka et al [63] have studied mechanical properties of alkaline-earth perovskites.

According to him, the elastic moduli and Debye temperature of BaUO3, BaZrO3, BaCeO3,
BaMoO3, SrTiO3, SrZrO3 and SrCeO3 are corrected to the values for 100% of the theoretical
density by using a finite element method, while those of SrMoO3, SrHfO3 and SrRuO3 are not
corrected. It has found that BaZrO3 and SrZrO3 show the highest Youngs modulus in Ba and
Sr series perovskites, respectively. It is known that the Debye temperature (D) can be related
to the melting temperature (Tm in K), the molar mass (M), and the molar volume (Vm) by the
Lindemann relationship [64]. This relationships were re-examined for the perovskites, and the
ratio of D to q5/6(Tm/(MVm2/3))1/2 was evaluated to be 1.60 [65-68], where q is the number of
atoms in the chemical formula.
2.3. Miscellaneous Properties and Applications
The measurement of electronic, optical and thermal properties has often been used to
characterize perovskite materials [69-71] and also to identify phase transitions [11,72-74] M.
Akaogi et al. [70], results shows that the recovered run products from conditions of 21.2-25.7
GPa and 1,673-2,073 K contained lithium niobate-structured phase. Based on the study by
Leinenweber et al. [75], concluded that the phase was the quench product of MgGeO3
orthorhombic perovskite stable at high pressure. According to M. Rousseau [74], E. P. R.
experiment clearly indicates that an order-disorder local behaviour occurs in SrTiO3 near Tc
[76]; in the same way, 87Rb NMR study in RbCaF3 seems to indicate a local symmetry
lowering far above Tc. In this context, RbCaF3 with its large tilt angles in one of the most
suitable compound for testing the validity of this description; to this purpose NMR spin-
lattice relaxation time measurements are being carried out. Eglitis et al [69], quantum
chemical INDO calculations have given additional evidence for the existence of electron
polarons in BaToO3 crystals which are expected to be stable at low temperatures. These
calculations for a series of perovskite materials-SrTiO3, BaTiO3, KNbO3 and KTaO3- give a
strong support to the green luminescence in these crystals as a result of the recombination of
the electrons and holes forming the charge transfer vibronic exciton rather than due to the
electron transitions in a MeO6 complex.
Luana et al [77], have presented a systematic analysis of the topological properties of the
electronic density in a family of ionic crystals: the halide perovskites AMX3. AIM (atoms in
molecules) theory provides a rigorous quantum-mechanical foundation to key concepts of the
chemical structure, such as which bonds do exist, how the shape of an atom is and what is its
charge, or how the properties of a functional group change form compound to compound.
Although it has been extensively applied to a number of organic gas-phase molecules, not
many studies have been published concerning solid-state systems [78-84] and Victor Luana et
al [77] also studied formation of antiperovskites. Antiprovskites, MAX3 have the same
structure as perovskites, exvept that M2+ and A+ exchange their position. A+ ions are
octahedrally coordinated by the anions and the A-X bonds will dominate. The only
antiperovskites that is simultaneously more stable than the perovskite and MX2+AX are
CaLiF3, SrLiF3 and BaLiF3. Even more, the perovskite structure should be unstable in the
three compounds. BaLiF3 is the only antiperovskite experimentally observed among the
AMX3 compounds [85-87]
Applications
An especially important field of application is their use as hosts for laser systems. The
perovskite structure appears as appropriate for this purpose because it contains a large A site
suitable for the incorporation of Nd (III) and Sm (III), which are some of the most popular
laser ions. And, on the other hand, the smaller B sites can receive the small Cr (III) ion which
is also frequently used in such systems. Many perovskite oxides present high electrical
resistivities, which make them useful as dielectric materials. However, another group of
materials such as LaVO3, LaTiO3, SrMoO3, which contain B-ions in an oxidation state lower
than their most stable one or which contain B-ions in two different oxidation states, are fairly
good conductors or semiconductors. The best-known metallic conductors are probably the
tungsten bronzes with the cubic-perovskite structure [4, 8, 11]. Perovskites abound both in
nature and in the laboratory, and their wide compositional range renders a variety of useful
properties such that perovskites are encountered in applications as disparate as
electroceramics, superconductors, refractories, catalysts, magnetoresistrors, and proton
conductors. The design of such advanced materials requires an understanding of the
relationship between chemical composition and crystal structure.
Some perovskite-oxides can be used as oxygen sensors. These types of sensors utilize
electrical conductivity changes due to oxygen adsorption. The conductivity is proportional to
oxygen partial pressure. Among these materials, SrTiO3 showed a high sensitivity to oxygen.
On the other hand, SrSnO3 appears as promising for use as a combustion monitoring-sensor
[61]. Finally, it is interesting to remember that perovskite materials are probably the
predominant minerals in the earths lower mantle. Among these important geological
perovskites, two phases seem to be the most important ones, namely (Mg, Fe)SiO3 and
CaSiO3. Both materials adopt the perovskite structure at pressures of several hundred
thousand atmospheres. Studies carried out on upper mantle silicate minerals, such as
pyroxenes, spinels, garnets and olivines, show that at the high pressures and temperatures of
the deep interior of the earth, these silicates are transformed, combining with some other
minerals, to form rocks that are dominated by perovkite-like silicates [62].
CONCLUDING REMARKS
Generally ABO3 oxides are electronic conductors. Ionic conduction is not favored
because of the close-packed nature of the perovskite structure. However, it has recently been
found that substitution in some ABO3 oxides can lead to ionic conduction. Thus partial
substitution by aluminium in CaTiO3 gives rise to ionic conductivity close to that observed in
oxides with the fluorite structure [88]. Substitution by lower valence ions of fixed valency at
the B site in these ABO3 oxides creates oxygen vacancies for charge neutrality. This increase
in the oxygen ion mobility and oxygen ionic conduction is observed. Quantum chemical
methods are a very efficient tool for the study of optical properties of advanced perovskite
materials.
ACKNOWLEDGMENTS
One of the authors (Dr. Ajay Singh Verma, PH/08/0049) is thankful to the University
Grant Commission New Delhi, India for supporting this research under the scheme of U.G.C.
Dr. D. S. Kothari Post Doctoral Fellowship.
REFERENCES
[1] A. Kumar, A. S. Verma and S. R. Bhardwaj, Open Applied Physics Journal, 1, 11,
(2008).
[2] Liu Liang, Lu Wencong and Chen Nianyi, J. Phys. Chem. Solids, 65, 855, (2004).
[3] A. B. A. Schippers, V. Brandwijk and E. W. Gorter, J. Solid State Chemistry, 6, 479,
(1973).
[4] R. Ubic, J. Am. Ceram. Soc, 90, 3326, (2007).
[5] R. P. Liferovich and R. G. Mitchell, J. Solid State Chemistry, 177, 2188, (2004).
[6] S. Yamanaka, K. Kurosaki, T. Maekawa, T. Matsuda, S. Kobayashi and M. Uno, J.
Nuclear Materials, 344, 61, (2005).
[7] L. Chonghe, T. Yihao, Z. Yingzhi, W. Chunmei and W. Ping, J. Phys. Chem. Solids,
64, 2147, (2003).
[8] R. Ubic and G. Subodh, J. Alloys and Compounds, (2009), doi:10.1016/
j.jallcom.2009.08.139.
[9] A. S. Verma and V. K. Jindal, J. Alloys and Compounds, 485, 514, (2009).
[10] A. Kumar and A. S. Verma, J. Alloys and Compounds, 480, 650, (2009).
[11] F.S.Galasso, Structure, Properties and Preparation of Perovskite-Type Compounds,
Pergamon Press, Oxford, (1969).
[12] P.Courty and C.Marcilly, in Preparation of Catalysts (B. Delmon, P.A. Jacobs and
G.Poncelet,Eds.), Elsevier,Amsterdam, (1976).
[13] A.K.Cheetham and P.Day, Solid State Chemistry.Techniques, Oxford University Press,
Oxford, (1987).
[14] J.M.Longo, H.S.Horowitz and L.R.Clavenna, in Solid State Chemistry (S.L.Holt,
J.B.Milstein and M.Robbins,Eds.), Adv.Chem. Series 186, American Chemical Sot.,
Washington DC, (1980).
[15] M.Kanimoto and T.Ozawa, Thermochim. Acta 148, 219 (1989).
[16] R.Roy, Science 238, 1664 (1987).
[17] L.G.Hubert-Pfalzgraf, New J. Chem. 11, 663 (1987).
[18] M. Guglielmi and G.Carturan, J.Non-Cryst.Solids 100, 16 (1988).
[19] J. B. Goodenough and J. M. Longo, in H. H. Landolt and R. Bornstein (Eds.)
Zahlenwerte und Funktionen aus Naturwissenschaften und Technik, New Series, Group
III, Vol. 4 (part a) (1970).
[20] F. Galasso, Structure and Properties of Inorganic Solids, Pergamon, New York,
(1970).
[21] V. M. Goldschmidt, Skrifter Norske Videnskaps-Akad. Oslo. I Mathemat. Naturwiss.
Klasse No 8, (1926).
[22] O Fukunaga and T. Fujita, J. Solid State Chemistry, 8, 331, (1973).
[23] C. Li, X. Lu, W. Ding, L. Feng, Y. Gao and Z. Guo, Acta Crystallographica, B64, 702,
(2008).
[24] H. H. Park, M. Pezat, B. Darriet, Rev. Chim. Min., 23, 323, (1986), C.R.Acad. Sc. 306,
963, (1988).
[25] R. Schuhmacher, A. Weiss, J. Less-Common Met. 163, 179, (1990).
[26] C.E.Messer, J. C. Eastman, R. G. Mers, A. J. Maeland, Inorg. Chem. 3, 776, (1964).
[27] A. W. Overhauser, Phys. Rev. B, 35, 411, (1987).
[28] I. Kuzmanovski and S. Alcksovska, Chemometrics and Intelligent Laboratory Systems,

67, 167, (2003).
[29] H. Horiuchi, T. Shishido, A. Saitow, M. Tanaka and S. Hosoya, Materials Science and
Engg. A, 312, 237, (2001).
[30] P. Wu and C. Li, Computer Coupling Phase Diagrams Thermochemistry, 27, 201,
(2003).
[31] S. Sasaki, J. D. Bass Prewitt and W. A. Schulze, Acta Crystal. C43, 1668, (1987).
[32] S. Sasaki, C. T. Prewitt and R. C. Liebermann, Am. Miner. 68, 1189, (1983).
[33] S. K. Saxena, L. S. Dubrovinsky, P. Lazor, Y. Cerenius, P. Haggkvist, M. Hanfland, J.
Z. Hu, Science, 274, 1357, (1996).
[34] R. A. Oganow, J. P. Brodholt and G. D. Price, Phys. Earth Planet Interiors, 122, 277
(2000).
[35] S. Geller, J. Chem. Phys. 24, 1236 (1956).
[36] W. C. Koehler and E. O. Wollan, J. Phys. Chem. Solids 2, 100 (1957).
[37] V. E. Nash and E. A. Turov, Fiz. Met. Metalloved. 11, 161, 321 (1961).
[38] V. E. Nash and E. A. Turov, Fiz. Met. Metalloved. 9, 10 (1960).
[39] M. Marezio, J. P. Remeika, and P. D. Dernier, Acta Crystallogr., Sect. B: Struct.
Crystallogr. Cryst. Chem. 26, 2008 (1970).
[40] P. Coppens and V. Eibschutz, Acta Crystallogr. 19, 524 (1965).
[41] S. Geller and E. A. Wood, Acta Crystallogr. 9, 563 (1956).
[42] S. Geller and V. B. Bala, Acta Crystallogr. 9, 1019 (1956).
[43] S. Geller, Acta Crystallogr. 10, 243 (1957).
[44] S. Geller, Acta Crystallogr. 10, 248 (1957).
[45] M. A. Gilleo, Acta Crystallogr. 10, 161 (1957).
[46] E. J. Baran and P. J. Aymonino, An Asoc. Quim. Argent, 56, 11, (1968).
[47] R. D. Shannon, Acta Crystallogr. A32, 751, (1976).
[48] A. Okazaki and Y. Suemune, J. Phys. Soc. Japan. 16, 176, (1961)
[49] E. Mooser, W.B. Pearson, Acta Cryst. 12, 1015, (1959).
[50] O. Muller, R. Roy, The Major Ternary Structural Families, Springer, New York,
1974.
[51] A. S. Bhalla, R. Y. Guo, R. Roy, Mater. Res. Innovations, 4, 3, (2000).
[52] R. Sondena, S. Stolen, P. Ravindran, T. Grande, and N. L. Allan, Phys. Rev. B 75,
184105, (2007).
[53] F. F. Fava, P. DArco, R. Orlando, and R. Dovesi, J. Phys.: Condens. Matter 9, 489,
(1997).
[54] Y. Syono, S. I. Akimoto, and K. Kohn, J. Phys. Soc. Jpn. 26, 993, (1969).
[55] R. H. Mitchell, Perovskites: Modern and Ancient (Almaz Press, Thunder Bay, Canada,
(2002).
[56] R. Sondena, P. Ravindran, S. Stolen, T. Grande, and M. Hanfland, Phys. Rev. B 74,
144102 (2006).
[57] Z.L. Wang, Z.C. Kang, Functional and Smart Materials: Structural Evolution and
Structure Analysis, Plenum Press, New York, 1998.
[58] R. L, Moreira and A. Dias, J. Phys. Chem. Solids, 68, 1617, (2007).
[59] L. Q. Jiang, J. K. Guo, H. B. Liu, M. Zhu, X. Zhou, P. Wu, and C. H. Li, J. Phys.
Chem. Solids, 67, 1531, (2006).
[60] C. Z. Ye, J. Yang, L. X. Yao, N. Y. Chen, Chinese Sci. Bull., 47, 725, (2002).
[61] Y. Shimizu, Y. Fukuyama, T. Narikiyo, H. Arai and T. Seiyama, Chem. Letters 377,
(1985).
[62] R. M. Hazen, Scient. Amer. 258, 75, (1988).
[63] S. Yamanaka, K. Kurosaki, T. Maekawa, T. Matsuda, S. Kobayashi and M. Uno, J.
Nuclear Materials, 344, 61, (2005).
[64] F. A. Lindemann, Physikalische Zeitschrift, 14, 609, (1910).
[65] The Chemical Society of Japan ed., Kikan Kagaku Sousetsu, Perovskite-Related
Compounds, No. 32, p. 37, (1997).
[66] Y. Hanajiri, T. Sato, T. Matsui, Netsu Sokutei, 26, 92, (1992).
[67] E. R. Naimon, Phys. Rev. B, 9, 737, (1974).
[68] A. Bartolotta, G. Carini, G. DAngelo, A. Fontana, F. Rossi, G. Tripodo, J. Non-Cryst.
Solids, 245, 9, (1999).
[69] R. I. Eglitis, E. A. Kotomin, G. Borstel, S. E. Kapphan and V. S. Vikhnin,
Computational Materials Science 27, 81, (2003).
[70] M. Akaogi, H. Kojitani, H. Yusa, R. Yamamoto, M. Kido and K. Koyama, Phys.
Chem. Minerals, 32, 603, (2005).
[71] T. Nakamura, J. Solid State Chemistry, 38, 229, (1981).
[72] W. J. Merz, Phys. Rev. 76, 1221, (1949).
[73] H. El-Mallah, B. E. Watts and B. Wanklyn, Phase Transit, 9, 235, (1987).
[74] M. Rousseau, Journal DE Physique Letters, 40, L439, (1979).
[75] K. Leinenweber, Y. Wang, T. Yagi and H. Yusa, Am. Mineral, 79, 197, (1994).
[76] A. D. Bruce, K. A. Muller and W. Berlinger, Phys. Rev. Lett. 42, 185, (1979).
[77] V. Luana, A. Costales and A. Martin Pendas, Phys. Rev. B, 55, 4285, (1997).
[78] P. Zou and R. Bader, Acta Crystallogr. A50, 714, (1994).
[79] V. Tsirelson, P. Zou, T. Tang and R. Bader, Acta Crystallogr. A51, 143, (1995).
[80] M. Eberhart, M. Donovan, J. MacLaren and D. Clougherty, Prog. Surf. Sci. 36, 1,
(1991).
[81] M. Eberhart, M. Donovan and R. Uutlaw, Phys. Rev. B, 46, 12744, (1992).
[82] M. Eberhart, J. MacLaren and D. Clougherty, J. Mater. Res., 8, 438, (1993).
[83] C. Mei, K. Edgecombe, V. Smith and A. Heilingerbrunner, Int. J. Quantum Chem. 48,
287, (1993).
[84] G. Grosch, B. Freytag, K. Range and U. Rosler, J. Chem. Phys., 101, 6782, (1994).
[85] S. Baldochi, V. Mazzocchi, C. Parcente and S. Morato, Materials Research Bulletin,
29, 1321, (1994).
[86] A. Boumriche, J. Gesland, A. Bulou, M. Rousseau, J. Fourquct and B. Hennion, Solid
State Commun. 91, 125, (1994).
[87] M. Duarte, M. Vieira and S. Baldochi, Mater. Sci. Eng. B, 25, 133, (1994).
[88] K. W. Browall, O. Muller and R. H. Doreonous, Mater. Res. Bull., 11, 1475 (1976).
Reviewed by Prof. B. K. Sarkar. Department of physics, VIT University, Vellore, Tamil

Nadu 632014, India.
Chapter 12
FERROELECTRIC PBTIO3: FROM A SINGLE-DOMAIN

STATE TO COMPOSITE COMPONENTS
V. Yu. Topolov* and A. V. Krivoruchko

Department of Physics, Southern Federal University, 5 Zorge Street,
344090 Rostov-on-Don, Russia
ABSTRACT
Regular ferroelectric lead titanate (PbTiO3) is one of the well-known perovskite-type
compounds with physical properties studied intensively in the last decades. Of specific
interest is PbTiO3 as an element of a hierarchical chain single-domain single crystal
polydomain single crystal poled ceramic composite, and the last element, composite,
can be based on other elements of the aforementioned chain. In this chapter examples of
modern piezo-active composites based on either PbTiO3 or PbTiO3-type solid solutions
are considered. These composites are regarded as matrix ones with 03 and 13
connectivity patterns and with a regular distribution of inclusions. To characterize the
effective electromechanical properties of the composite, two methods are applied. The
effective properties of the 03 composite with spheroidal inclusions and of the 13
composite with circular cylindric inclusions are predicted by means of the effective field
method. The effective properties of the 13 composite with inclusions in the form of
rectangular parallelepiped are determined within the framework of the matrix method. In
both the methods, an electromechanical interaction between components is taken into
account. The role of the PbTiO3-containing component is emphasized in connection with
the large piezoelectric anisotropy and considerable piezoelectric sensitivity of the
composite in certain volume-fraction ranges.
*
E-mail: [email protected], [email protected]
482 V. Yu. Topolov and A. V. Krivoruchko
INTRODUCTION
In the last decades, advanced functional materials based on lead titanate (PbTiO3) are
intensively studied for a variety of reasons. First, PbTiO3 is the regular ferroelectric
perovskite-type oxide [13] that undergoes the first-order phase transition near TC = 765 K
[4]. Features of physical properties, crystal and domain structures of PbTiO3 single crystal
(SC), and its behavior in the phase-transition region are subjects of experimental and
theoretical studies providing new information on the ferroelectric state. Second, full sets of
elastic, piezoelectric and dielectric constants (i.e., electromechanical constants) are known for
single-domain [4, 5] and polydomain [6] SCs of PbTiO3 at room temperature. At the same
time, no full sets of electromechanical constants were experimentally determined on
ferroelectric ceramic (FC) samples of PbTiO3. Third, PbTiO3 is the important component of
the perovskite-type ferroelectric solid solutions, such as Pb(Zr1-xTix)O3, (Pb1-xLax)(Zr1-yTiy)O3,
(Pb1-xCax)TiO3 (widespread FCs with important piezoelectric and other properties) [1, 2, 7]
and (1 x)Pb(A1/3Nb2/3)O3 xPbTiO3 (highly piezo-active SCs with so-called domain
engineered structures), where A = Mg or Zn [810]. These and other PbTiO3-based solid
solutions are of interest due to remarkable piezoelectric, optical, dielectric, and other
properties. Fourth, FCs of modified PbTiO3, (Pb1-xCax)TiO3 etc. and relaxor-ferroelectric SCs
of (1 x)Pb(A1/3Nb2/3)O3 xPbTiO3 are used as components of modern piezo-active
composites [1116] with 03, 13, 33 and other connectivity patterns. The piezoelectric
anisotropy, activity, sensitivity, electromechanical coupling factors, and other parameters of
there composites [1216] are of value for sensor, actuator, hydrophone, and transducer
applications. Fifth, electromechanical properties of materials based on PbTiO3 were analysed
within the framework on the hierarchical chain single-domain SC polydomain SC poled
FC [1719]. A subsequent element of the chain, the composite based on PbTiO3 (either SC
or FC, including modified compositions), is of an independent interest due to various
interrelations [16, 2022] between the properties and microgeometry of components.
In the present chapter we discuss examples of performance of modern PbTiO3-based
composites and important interrelations between the electromechanical properties of the
elements in the aforementioned hierarchical chain. Our aim is to show the role of the PbTiO3-
based components in forming the anisotropic piezoelectric response, considerable
piezoelectric sensitivity and non-monotonic volume-fraction dependence of the effective
properties of the piezo-active composites. Below we consider the FC / polymer and SC / FC
matrix composites with 3 connectivity ( = 0 or 1).
03 FERROELECTRIC CERAMIC / POLYMER COMPOSITE

AND ITS PERFORMANCE
Modelling of Effective Electromechanical Properties
The 03 PbTiO3-type FC/epoxy resin composite is an example of the piezo-active

material prepared by electric-field structuring [11]. The microstructure and anisotropic
electromechanical properties of this composite are attained due to use of dielectrophoresis for
the structural modification. At means that a dispersion of FC particles embedded in a liquid
Ferroelectric PbTiO3: From a Single-domain State to Composite Components 483
polymer or pre-polymer is exposed to a moderate as field that promotes the formation of

chain-like structures in a polymer matrix. After a sodification of the polymer medium, these
structures remain stable of the composite sample [11].
An analysis of the composite microstructure suggests that this material is characterized
by 03 connectivity (in terms introduced by Newnham et al. [20]), and the shape of the FC
inclusions therein approaches the spheroidal shape.
To predict the effective electromechanical properties of the 03 composite with the
spheroidal inclusions, different averaging methods are applied [15, 16, 2325]. Our current
study is carried out within the framework of the model [15] that has first been applied to the
03 composite based on the (Pb1-xCax)TiO3 FC with the large piezoelectric anisotropy. In this
model the shape of each inclusion (Figure 1, component 1) is described by the equation
(x1 / a1)2 + (x2 / a1)2 + (x3 / a3)2 = 1 (1)
relative to the axes of the rectangular coordinate system (X1X2X3), where the semi-axes of the
spheroid are a1 = a2 and a3. A remanent polarization vector of each inclusion is Pr(1) OX3,
and the OX3 axis is the poling axis of the composite. We consider a regular distribution of the
spheroidal inclusions occupy sites of a simple tetragonal lattice with unit-cell vectors parallel
to the OXj axes. It is assumed that the electrical conductivities (1) of FC and (2) of polymer
are linked with an inequality (2) (1) and that surface charges caused by the ferroelectric
polarization fully screened by free carriers coming to interfaces in the composite sample. The
( n ), E
components of the composite are characterized by sets of elastic moduli cab measured at
(n)
electric field E = const, piezoelectric coefficients eij and dielectric permittivities
),
(n
pp measured at mechanical strain = const, where n = 1 and 2. The effective
electromechanical properties determined in the long-wavelength limit depend on the volume

fraction m of FC and the aspect ratio = a1 / a3 where aj are the semi-axes of the FC inclusion
(Figure 1).
Figure 1. Schematic of the 03 composite with spheroidal inclusions. The polarization vector of the
inclusion is denoted by an arrow therein. X1X2X3 is the rectangular coordinate system, ai are semiaxes of
the spheroid, m is the volume fraction of component 1, and 1 m is the volume fraction of component
2.
The calculation procedure is based on the effective field method (EFM), i.e., the Mori
Tanaka method [26] generalized for heterogeneous piezoelectric media [16, 23]. The EFM
based on Eshelbys concept of spheroidal inclusions [27, 28] is a variant of the self-consistent
scheme for the calculation of effective constants of the piezoactive composites. Following the
EFM, we take into account the electromechanical interaction between the piezo-active
inclusions in the matrix and related coupled effects. The effective electromechanical
properties of the 03 composite are represented in the matrix form as
E t
c* e*
|| C* || = || C*(m, ) || = (2)
e * *

where the superscript t denotes the transposition. It should be noted that the matrices || c*E ||
(6 6), || e* || (3 6) and || * || (3 3) from Eq. (2) correspond to mm symmetry of the 0
3 composite. The effective electromechanical properties from Eq. (2) are determined by
means of the EFM, and the matrix of these properties is written [16] as
|| C* || = || C(2) || + m (|| C(1) || || C(2) ||){|| K || + (1 m) || S || || C(2) ||-1 (|| C(1) || || C(2) ||)}-1, (3)
where matrices of the electromechanical properties of the components || C(1) || and || C(2) ||
have the form similar to that shown Eq. (2), || K || is identity matrix, and || S || is the matrix that
comprises components of the electroelastic Eshelby tensor [28]. C omponents of the
(2)
Eshelby tensor depend on elements of || C || and the aspect ratio of the inclusion. The
* * *
effective piezoelectric coefficients dij , gij and hij are determined using the effective
constants from Eq. (3) in accordance with formulae [29]
d ij* = eif* s*fjE = *pi g *pj and eij* = *pi h*pj . (4)
In Eqs. (4), *pi = *pi + e*pf d if* is dielectric permittivity measured at mechanical stress
= const.
Effective Electromechanical Properties: Volume-fraction and Aspect-ratio

Dependences
As is known from various experimental data (Figure 2), important features of the PbTiO3-
(1)
type FCs are three positive piezoelectric coefficients eij (ij = 15, 31 and 33) and the large
(1) (1) (1)
anisotropy of the piezoelectric coefficients d 3 j , i.e., d 33 / | d 31 | >> 1. At the first stage of
calculations, the full sets of room-temperature experimental electromechanical constants of
FC I (see data I from Figure 2) and epoxy [32] are used.
( n ), E
Figure 2. Room-temperature experimental values of elastic moduli cab (in 1010 Pa), piezoelectric
coefficients eij( n ) (in C / m2) and relative dielectric permittivities 33* / 0 of FCs I [30], II IV [31] and
V [2]. FCs I and II with equal chemical formulae are distinguished by processing conditions.
Results on the calculated effective parameters of the 03 composite studied are shown in
Figure 3. To avoid differences between electromechanical constants of the PbTiO3-type FC
used in experimental work [11] and constants of FC I from Figure 2, we consider normalized
dielectric permittivity = 33* / 33(1), and normalized piezoelectric coefficients d = d 33* /
d 33(1) and g = g 33
* (1)
/ g 33 . Good agreement between the experimental [11] and calculated
volume-fraction dependences of is observed in the aspect-ratio range = 0.13 0.19 (see
Figure 3. a). It means that the structured 03 composite can be regarded as an aggregate of
prolate inclusions in a matrix so that the semi-axes of the inclusions are linked by a3 / a1 5
7. The similar microstructure of the 03 composite is observed in experimental work [11].
The volume-fraction dependence of for the unstructured 03 composite can be
approximated by pieces of curves calculated for 0.32 0.50 (Figure 3, b).
As follows from our analysis of the piezoelectric response of the composite studied,
* * * * *
among the piezoelectric coefficients d 3 j , e3 j , g 3 j , and h3 j (j = 1 and 3) from Eqs. (4), g 31
*
and g 33 are most sensitive to changes in the m and values. For example, in the aspect-ratio
* (1)
range = 0.01 0.20, one can reach maxg 3 10 and min g 31 / g 31 9 40. A
difference between the aforementioned ratios is concerned with the influence of the system of
the prolate FC inclusions on forming the anisotropic piezoelectric response of the composite.
Due to this system, the volume-fraction dependence of d and g (Figure 3, c and d) resembles
* *
behaviour of the piezoelectric coefficients d 33 and g 33 , respectively, in the 13 PbTiO3-type
FC / polymer composite [33] with the regularly-distributed FC rods. An advantage of the 03
composite based on the PbTiO3-type FC consists in the relatively large maxg values that are
attained near the volume fraction m = 0.1 at various aspect ratios (Figure 3, d). It is
( n ), E
concerned with a favourable balance of elastic moduli cab of components and relatively
*
low dielectric permittivity 33 at 0 < m 0.1. Eqs. (4) suggest that 33* considerably
*
influences g 33 and g of the composite. Examples of behaviour of two factors of the
* * * * * *E * *E * *E *E
piezoelectric anisotropy, d 33 / d 31 and e33 / e31 = (2 d 31 c13 + d 33 c33 ) / [ d 31 ( c11 + c12 )
* *E * *
+ d 33 c13 ], is shown in Figure 3, e and f, respectively. It is seen that the ratio of e33 / e31 is
varied in the wider range, and this behaviour is apparently concerned with an active influence
*
of elastic properties of the FC and polymer components on the piezoelectric coefficients e3 j
and with the presence of the prolate FC inclusions aligned on the poling direction OX3 (Figure
1).
(a) (b)
(c) (d)
(e) (f)
Figure 3. Volume-fraction dependence of normalized dielectric permittivity (a and b), normalized

* *
piezoelectric coefficients d (c) and d (d), and factors of the piezoelectric anisotropy d 33 / d 31 (e) and
* *
e33 / e31 (f) of the 03 PbTiO3-based FC / epoxy composite. Graphs (a) and (b) correspond to the
structured composite and unstructured composite, respectively. Experimental points are from dielectric
measurements [11].
At the second stage of calculations, we compare the normalized parameters , d and g

that are evaluated using the full sets of electromechanical constants of FC from Figure 2.
Examples of the volume-fraction dependence of these parameters in the composite with the
prolate spheroidal inclusions are shown in Figure 4. The composites based on different FC
compositions with the large piezoelectric anisotropy have the similar performance. Values of
max g are expected at volume fractions 0.05 < m < 0.10, i.e., in the range where the influence
of dielectric permittivity of polymer remains strong. In this connection, we mention that
dielectric properties of the FC and polymer components along the OX3 axis are interrelated by
a ratio 33(1), / 33( 2) ~ 10.
Transition to 13 Connectivity
Now we consider a transition from 03 connectivity to 13 one. In case of the 13

composite, we assume that = 0 and the composite contains the infinitely long FC rods being
parallel to the poling axis OX3. They have the shape of a circular cylinder with a fixed radius
and are aligned on the axis OX3 of the rectangular co-ordinate system (X1X2X3). Centres of
sections of these cylinders constitute a square lattice in the (X1OX2) plane, i.e., square patterns
in the arrangement of the FC rods are considered. The EFM is also applied for calculating the
effective electromechanical properties of the 13 composite. Moreover, it should be added
that the transition from 03 to 13 means continuous changes [28] in elements of the || S ||
matrix from Eq. (3). As a consequence, we see a monotonic variation of the effective
properties of the composite (Figure 5). To determine its effective properties, we used the full
sets of electromechanical constants of (Pb0.75Ca0.25)TiO3 FC [15] with the large piezoelectric
anisotropy and elastic and dielectric constants of piezo-passive araldite [15].
As follows from Figure 5, the decrease of the effective parameters 33 ( = e, d, g, h, ,
*
and ) of the 03 composite as compared to 33 of the 13 composite at m = const is caused

*
by a discrete distribution of the piezo-active component along the OX3 axis and by a lowering
of the piezoelectric activity. This lowering is concerned with an increase of the depoling field
* *
in the FC inclusions of the 03 composite. A slight increase of e31 and a decrease of | d31 | as
the aspect ratio increases at m = const is observed in the presence of the FC component with
(1)
that the large anisotropy of the piezoelectric coefficients d 3 j and with the moderate
(1) *
anisotropy of the piezoelectric coefficients e3 j > 0. The more pronounced increase of h31 as
*
compared to e31 at m = const is accounted for by the influence of the dielectric permittivity
33* of the composite in accordance with Eqs. (4).
Figure 4. Volume-fraction dependence of normalized parameters , d and g of the 03 composite

PbTiO3-type FC / epoxy. Electromechanical constants of FCs I V were taken from experimental
data in Figure 2.
Figure 5. Volume-fraction dependence of the piezoelectric coefficients e3* j (in C / m2), d 3* j (in pC /
N), g 3* j (in mV.m / N), and h3* j (in GV / m) and dielectric permittivities 33* / 0 and 33* / 0 of the
(Pb0.75Ca0.25)TiO3 FC / araldite composite at 0 0.05 (reprinted from Topolov et al. [33], with
permission from IOP Publishing).
Figure 6. Volume-fraction dependence of the piezoelectric coefficients e3* j (in C / m2), d 3* j (in pC /
N), g 3* j (in mV.m / N), and h3* j (in GV / m), dielectric permittivities 33* / 0 and 33* / 0, and
*E
elastic moduli c ab (in GPa) which have been calculated for the 13 (Pb0.75Ca0.25)TiO3 FC / araldite
composite by means of the EFM and finite element metod (reprinted from Topolov et al. [33], with
permission from IOP Publishing).
Below we compare data on the effective parameters (Figure 6) calculated using two
different methods, the EFM and the finite elemet method [33]. In general, the good agreement
between the calculated values of the effective parameters (e.g., piezoelectric coefficients e3* j ,
*
d 3* j , g 3* j , and h3* j , dielectric permittivities 33 / 0 and 33 / 0) is attained. The difference
*
*
between the the effective parameters 33 of the 13 composite does not exceed 1 % in the
volume-fraction range of 0.1 m 0.6. At the largest volume fraction of FC m = 0.6 (Figure
*
6), a difference between values of the piezoelectric coefficient e31 evaluated using different
methods equals about 6 %. The same difference takes place for the piezoelectric coefficient
*
h31 evaluated using different methods.
Figure 7. Schematic of the 13 SC / polymer composite with planar microgeometry. (X1X2X3) is the
rectangular co-ordinate system. m and 1 m are volume fractions of SC and polymer, respectively. Ps(1)
and Pr(2) are the spontaneous polarisation vector of SC and the remanent polarization vector of polymer,
respectively.
13 FERROELECTRIC SINGLE CRYSTAL / FERROELECTRIC POLYMER

COMPOSITE AND ITS PEFORMANCE
Effective Electromechanical Properties and Components
The ferroelectric SC / polymer composite with 13 connectivity is the novel piezo-active

material that has attracted attention in the last decade due to efforts to improve the performace
of the composite by using the relaxor-ferroelectric component with the very high piezoelectric
activity [1214, 16]. In this section we consider the 13 composite (Figure 7) containing SC
rods in the form of rectangular parallelepipeds. The SC rods have a square base and are
characterized by a square arrangement in the (X1OX2) plane. The rods (either single-domain
or polydomain in the whole composite sample) are poled along the OX3 axis, and the
spontaneous polarisation vector Ps(1) OX3 in each rod. The main crystallographic axes X, Y
and Z of SC are parallel to the following perovskite unit-cell directions: X || [001], Y || [010]
and Z || [001]. The composite comprises one of the following groups of the SC rods poled
along the [001] direction of the perovskite unit cell:
(i) single-domain PbTiO3,

(ii) (1 x)Pb(Mg1/3Nb2/3)O3 xPbTiO3 with domain engineered structure (four 71
domain types that are separated by the (100) and (010) domain walls) [3, 34] and
compositions near the morphotropic phase boundary, or
(iii) polydomain PbTiO3 with equal volume fractions of two types of 90 domains that are
regularly distributed and separated by stressfree (001) domain walls [6]. The
spontaneous polarization vectors of these domain types are Ps(Ps / 2 ; 0; Ps / 2 )
and Ps(Ps / 2 ; 0; Ps / 2 ) in the co-ordinate system (X1X2X3) shown in Figure 7.
It is assumed that the composite matrix is made of polyvinylidene fluoride (PVDF). This
ferroelectric polymer can be poled so that its remanent polarisation vector Pr(2) OX3, and
( 2) ( 2)
in this case the piezoelectric coefficients dij of polymer obey condition sgn d31 =
( 2) ( 2) (1)
sgn d33 =sgn d15 < 0. For the SC component (i) or (ii), the analogous condition sgn d 31 =
(1) (1)
sgn d 33 =sgn d15 < 0 is also true [46, 12]. Conditions for the antiparallel orientation of the
Ps(1) and Pr(2) vectors (Figure 7) can be satisfied on poling the composite sample in an electric
field E || OX3 due to the significant difference between the coercive fields [14, 35] of SC and
polymer. In case of the composite based on single-domain PbTiO3 SC, a bias electric field Eb
|| OX3 is applied to stabilize the single-domain state with Ps(1) OX3 in each SC rod.
The determination of the effective electromechanical properties of the 13 composite
with planar interfaces (Figure 7) is carried out using a matrix approach [16]. An averaging
procedure within the framework of this approach leads to averaging of the electromechanical
constants of the SC rods and the polymer matrix in the OX1 and OX2 directions, in which the
periodical structure of the composite is observed. In our calculations, the full sets of elastic
( n ), E (n )
compliances s ij (measured at E = const), piezoelectric coefficients d kl and dielectric
permittivities (ffn ), (measured at mechanical stress = const) of both components are used.
The electromechanical properties are averaged by taking into account boundary conditions
[16] for the electric and mechanical fields. These boundary conditions at the interface x1 =
const (Figure 7) mean the continuity of normal components of mechanical stress 1v (v = 1, 2
and 3), tangential components of mechanical strain 22, 23 and 33, normal component of
electric displacement D1, and tangential components of electric field E2 and E3. At the
interface x2 = const (Figure 7), the continuity of 2v (v = 1, 2 and 3), 11, 13, 33, D2, E1, and
E3 is required. The matrix of the effective electromechanical properties of the composite [16]
is given by
|| C* || = [|| C1 ||.|| M || m + || C2 || (1 m)] [|| M || m + || I || (1 m)]-1 (5)
In Eq. (5), || Cn || are matrices of electromechanical constants of SC (n = 1) and polymer

(n = 2), and these matrices are represented in the following form:
|| s(n),E || || d(n) ||t

.
|| Cn || = (n) (n),
|| d || || ||
The matrix || M || from Eq. (5) is related to the aforementioned boundary conditions at
interfaces xi = const (i = 1 and 2), and || I || is the identity matrix. The effective
electromechanical properties determined on the basis of Eq. (5) are regarded as homogenised
properties in the longwave approximation, when the wavelength of an external acoustic field
is much longer than the size of the rod in the composite sample. We neglected a concentration
of mechanical stress near the edge of the rod because of the small area of this concentration
[36] in comparison to the area of the square section of the rod in the (X1OX2) plane. Our
recent results [37] on the effective parameters calculated for the similar 13 SC / polymer
composite by using the matrix method and the finite element method suggest that the good
agreement between the parameters is observed in the whole volume-fraction range.
Figure 8. Room-temperature elastic compliances elastic compliances s ij( n ), E (in 10-12 Pa-1), piezoelectric
coefficients d kl(n) (in pC / N) and relative dielectric permittivities (ffn ), of (1 x)Pb(Mg1/3Nb2/3)O3
xPbTiO3 SCs: x = 1, single-domain and polydomain [6], x = 0.33 and 0.30, with domain engineered
structure [38, 39]. The full set of electromechanical constants of polydomain PbTiO3 SC has been
determined [6] for case of motionless domain walls separating the 90 domains.
(a)
(b)
Figure 9. Volume-fraction dependence of effective parameters calculated for the 13 single-domain

*
PbTiO3 SC / PVDF composite with planar microgeometry: a, piezoelectric coefficients d33 (in pC /
*
N), g33 and gh* (in mV.m / N) and b, squared figures of merit
* 2
(Q33 ) and (Qh* ) 2 (in 10-15 Pa-1).
Maxima of Effective Parameters
Now we consider volume-fraction behaviour of the effective piezoelectric coefficient

3 3

* * *
g33 and its hydrostatic analog g h = g kl* . Based on d kl being calculated within the
k =1 l =1
*
framework of || C* || from Eq. (5) and g kl being calculated using Eq. (4), we study the
* 2 *
volume-fraction dependence of squared figure of merit (Q33 ) = d 33 g 33 = ( d 33 )2 /
* *
33* and
* 2 * *
its hydrostatic analog (Qh ) = d h g h = ( d h )2 /
*
33* . The parameters g 33
* * 2
and (Q33 ) are
* * 2
important for sensor applications, the parameters g h and (Qh ) are of value for hydrophone
* 2 * 2
and other hydroacoustic applications. The parameters (Q33 ) and (Qh ) are used to describe
the sensor signal-to-noise ratio of the piezoelectric element [16]. Figure 8 contains the full
sets of electromechanical constants of components of the studied 13 composites.
* * * 2 * 2
It is seen that g 33 , g h , (Q33 ) , and (Qh ) of the 13 composite have maxima at
volume fractions 0.045 < m < 0.20 (Figure 9), i.e., in the range where the piezoelectric
*
coefficient d33 increases intensively (see curve 1 in Figure 9, a) and dielectric permittivity
33* remains relatively low. Volume fractions m 0.05 and more are attainable when
manufacturing the 13 composite, and this circumstance is to be taken into consideration for
*
further applications. Of particular interest is the piezoelectric coefficient g33 : the value of
*
max g 33 = 641 mV.m / N at m 0.045 0.050 (see curve 2 in Figure 9, a) remains larger than,
*
for example, max g 33 found for the 13 (1 x)Pb(Mg1/3Nb2/3)O3 xPbTiO3 SC / araldite
composite (x = 0.30, 0.33 and 0.42) [40]. The presence of piezopassive araldite instead of
piezo-active PVDF with approximately equal stiffness leads to decreasing the piezoelectric
sensitivity of the 13 composite, even if it is based on the highly piezo-active SC component.
Figure 10 contains data on maxima of the aforementioned effective parameters of the
*
related composites. Contrary to the composite based on single-domain PbTiO3 SC, max g33
*
and max g h of these composite are lower than those shown in Figure 9, a, and these maxima
are attained at small volume fractions m < 0.02. In our opinion, such a performance is
accounted for by a less-favourable balance of elastic constants of components (x = 0.33 and
0.30) or by a relatively low piezoelectric activity of polydomain PbTiO3 along the poling axis
(1)
OX3. As follows from data in Figure 8, the piezoelectric coefficient d 33 of polydomain
( 2)
PbTiO3 is about 1.8 times more than d 33 of PVDF, and this factor may promote the low
piezoelectric sensitivity of the composite.
Figure 10. Maximum values of effective parameters * of the 13 (1 x)Pb(Mg1/3Nb2/3)O3 xPbTiO3

SC / PVDF composite with planar microgeometry shown in Figure 7. In parentheses there are values of
the SC volume fraction m that corresponds to max*.
* *
Figure 11. Volume-fraction dependence of effective piezoelectric coefficients d33 (in pC / N), g33 (in
* *
mV.m / N) and anisotropy d33 / d 31 which have been calculated for the 03 single-domain PbTiO3 SC
/ FC V composite with oblate spheroidal inclusions.
03 FERROELECTRIC SINGLE CRYSTAL / FERROELECTRIC

CERAMIC COMPOSITE AND ITS PERFORMANCE
The ferroelectric SC / FC composite is the novel piezo-active material that has attracted
attention after research [41]. Smotrakov et al. [41] prepared samples of the ferroelectric SC /
FC composite based on the perovskite-type solid solution of (Pb1xCax)TiO3 with x 0.24 and
studied microstructure of this composite. The 03 0.71Pb(Ni1/3Nb2/3)O3 0.29PbTiO3 SC /
0.5Pb(Ni1/3Nb2/3)O3 0.15PbZrO3 0.35PbTiO3 FC composite with perovskite-type relaxor-
ferroelectric components and the regular arrangement of the SC inclusions was prepared and
studied by Takahashi et al. [42]. First attempts to predict the effective electromechanical
properties of the SC / FC composite were made on the basis of models of the 03 composite
[21, 22] with isolated inclusions being either single-domain or polydomain. Examples of the
performance of the 03 composite with the relaxor-ferroelectric SC inclusions surrounded by
the FC matrix were described in paper [43]. In this section we consider an example of the 03
single-domain PbTiO3 SC / FC V composite and its piezoelectric performance.
It is assumed that the 03 composite comprises the system of oblate spheroidal FC
inclusions and they are distributed regularly in the FC matrix. The inclusions are single-
domain with the spontaneous polarisation vector Ps(1) OX3, and the crystallographic axes
of each inclusion are oriented along the co-ordinate axes shown in Figure 1. The FC medium
surrounding the inclusions has the remanent polarization Pr(2) OX3. To determine the
effective electromechanical properties and other parameters of the composite, we apply the
EFM (see Eq. (3)) and use the full sets of electromechanical constants of single-domain
PbTiO3 [6] and FC V [2] (see also data in Figs. 2 and 8). Now the volume fractions of SC
(component 1 in terms of Figure 1) and FC (component 2 in terms of Figure 2) equal m and 1
m, respectively.
The studied 03 PbTiO3-based composite exhibits the considerable stability of the
piezoelectric properties with respect to changes in the aspect ratio . For example, changes in
from 10 to 103 do not lead to appreciable changes in the piezoelectric coefficients of the
* * * * *
composite d 3 j (m), g3 j (m) and in the piezoelectric anisotropy d33 (m) / d 31 (m) = g 33 (m) /
*
g 31 (m) (Figure 11). It can be accounted for by the presence of FC with the large piezoelectric
anisotropy and by comparable values of electromechanical constants of the SC and FC
* *
components in the composite. Values of the piezoelectric coefficients d33 and g 33 (see
curves 1, 2, 4, and 5 in Figure 11) are comparable to those predicted in earlier papers [21, 22]
for the related SC / FC composites.
CONCLUSION
This chapter has been devoted to the piezoelectric performance of modern composites
based on perovskite ferroelectric PbTiO3 or its solid solutions. Modelling and effective-
property predictions have been carried out for the matrix 3 composites with structures
shown in Figs. 1 and 7. It should be noted that elements of the hierarchical chain single-
domain SC of PbTiO3 polydomain SC of PbTiO3 poled FC of modified PbTiO3 are
*
attractive as components of the composites with the large piezoelectric anisotropy ( d33 /
* * * *
| d 31 | >> 1 and/or e33 / | e31 | >> 1), high piezoelectric sensitivity ( g33 102 mV.m / N and
* 2
(Q33 ) 10-11 Pa-1), and considerable hydrostatic parameters ( g h* 102 mV.m / N and
(Qh* )2 10-11 Pa-1). This performance is attainable due to favourable relations between the
electromechanical properties of the components and to the presence of the system of the
prolate FC inclusions or long SC rods oriented along the poling axis. Contrary to this, the
system of the oblate SC inclusions does not promote high piezoelectric sensitivity, as follows
from data on the 03 single-domain PbTiO3 SC / PbTiO3-type FC composite.
The results discussed in this chapter can promote the selection of the proper SC and FC
components for novel composites with high piezoelectric sensitivity, large piezoelectric
anisotropy, considerable hydrostatic response, and other remarkable effective parameters. It is
believed that the development of high-performance composites will be attractive for modern
piezotechnical applications (sensors, hydrophones, transducers, etc.).
ACKNOWLEDGMENTS
The authors would like to thank Prof. Dr. A.V. Turik, Prof. Dr. A. E. Panich, Prof. Dr. I.
A. Parinov, Dr. V.V. Eremkin, and Dr. V.G. Smotrakov (Southern Federal University,
Russia), Prof. Dr. L.N. Korotkov (Voronezh State Technical University, Russia), Prof. Dr. R.
Stevens and Dr. C.R. Bowen (University of Bath, UK), and Prof. Dr. P. Bisegna (University
of Rome Tor Vergata, Italy) for their continued interest in the research problems. This work
was partially supported by the administration of the Southern Federal University (Project No.
11.1.09f on basic research), and the support is gratefully acknowledged.
REFERENCES
[1] Smolensky, G. A.; Bokov, V. A.; Isupov V. A.; Krainik, N. N.; Pasynkov, R. Ye.;
Sokolov, A. I.; Yushin, N. K. Physics of Ferroelectric Phenomena; Nauka:
Leningrad, 1985; 396 p (in Russian).
[2] Xu, Y. Ferroelectric Materials and Their Applications; North-Holland: Amsterdam,
London, New York, NY, Toronto, 1991; 391 p.
[3] Cross, L. E. In Piezoelectricity. Evolution and Future of a Technology; Heywang.
W.; Lubitz, K.;Wersing, W.; Eds.; Springer Series in Materials Science 114;
Springer: Berlin, Heidelberg, 2008; pp 131156.
[4] Fesenko, E. G.; Gavrilyatchenko, V. G.; Semenchev, A. F. Domain Structure of
Multiaxial Ferroelectric Crystals; Rostov University Press: Rostov-on-Don, 1990;
192 p.
[5] Gavrilyatchenko, V.; Semenchev, A.; Fesenko, E. Ferroelectrics 1994, 158, 3135.
[6] Turik, A. V.; Topolov, V. Yu.; Aleshin, V. I. J. Phys. D: Appl. Phys. 2000, 33, 738
743.
[7] Haertling, G. J. Am. Ceram. Soc. 1999, 82, 797818.
[8] Damjanovic, D. In Piezoelectric and Acoustic Materials for Transducer
Applications; Safari, A.; Akdoan, E. K.; Eds.; Springer: New York, NY, 2008; pp
5979.
[9] Noheda, B. Curr. Opinion i. Solid State a. Mater. Sci. 2002, 6, 2734.
[10] Davis, M. J. Electroceram. 2007, 19, 2345.
[11] Wilson, S. A.; Maistros, G. M.; Whatmore, R. W. J. Phys. D: Appl. Phys. 2005, 38,
175182.
[12] Ritter, T.; Geng, X.; Shung, K. K.; Lopath, P. D.; Park, S.-E.; Shrout, T. R. IEEE
Trans. Ultrason., Ferroelec., a. Freq. Contr. 2000, 47, 792800.
[13] Ren, K.; Liu, Y.; Geng, X.; Hofmann, H. F.; Zhang, Q. M. IEEE Trans. Ultrason.,
Ferroelec., a. Freq. Contr. 2006, 53, 631638.
[14] Wang, F.; He, C.; Tang, Y. Mater. Chem. Phys. 2007, 105, 273277.
[15] Glushanin, S. V.; Topolov, V. Yu.; Krivoruchko, A. V. Mater. Chem. Phys. 2006,
97, 357364.
[16] Topolov, V. Yu.; Bowen, C. R. Electromechanical Properties in Composites Based
on Ferroelectrics; Springer: London, 2009; 202 p
[17] Topolov, V. Yu.; Bondarenko, E. I.; Turik, A. V.; Chernobabov, A. I. Ferroelectrics

1994, 152, 237242.
[18] Turik, A. V.; Topolov, V. Yu. J. Phys. D: Appl. Phys. 1997, 30, 15411549.
[19] Topolov, V. Yu.; Turik, A. V.; Chernobabov, A. I. Crystallogr. Repts 1994, 39, 805
809.
[20] Newnham, R. E.; Skinner, D. P.; Cross, L. E. Mater. Res. Bull. 1978, 13, 525536.
[21] Topolov, V. Yu.; Glushanin, S. V. Tech. Phys. Lett. 2002, 28, 279282.
[22] Glushanin, S. V.; Topolov, V. Yu., Turik, A. V. Crystallogr. Repts. 2003, 48, 491
499.
[23] Dunn, M. L.; Taya, M. Internat. J. Solids Struct. 1993, 30, 161175.
[24] Levin, V. M.; Rakovskaja, M. I.; Kreher, W. S. Internat. J. Solids Struct. 1999, 36,
26832705.
[25] Fakri, N.; Azrar, L.; El Bakkali, L. Internat. J. Solids Struct. 2003, 40, 361384.
[26] Mori, T.; Tanaka, K. Acta Metall. 1973, 21, 571574.
[27] Eshelby, J. D. Proc. Roy. Soc. (London) 1957, A241, 376396.
[28] Huang, J. H.; Yu, S. Compos. Engin. 1994, 4, 11691182.
[29] Ikeda, T. Fundamentals of Piezoelectricity; Oxford University Press: Oxford, New
York, NY, Toronto, 1990; p 17.
[30] Ikegami, S.; Ueda, I.; Nagata, T. J. Acoust. Soc. Am. 1971, 50, 10601066.
[31] Nagatsuma, K.; Ito, Y.; Jyomura, S.; Takeuchi, H.; Ashida, S. In Ferroelectricity and
Related Phenomena. V. 4: Piezoelectricity; Taylor, G. W.; Ed.; Gordon and Breach
Sci. Publ.: New York, NY, London, Paris, Montreux, Tokyo, 1985; pp 167176.
[32] Cao, W.; Zhang, Q. M.; Cross, L. E. IEEE Trans. Ultrason., Ferroelec., a. Freq.
Contr. 1993, 40, 103109.
[33] Topolov, V. Yu.; Bisegna, P.; Krivoruchko, A. V. J. Phys. D: Appl. Phys. 2008,
41, 0354068 p.
[34] Yin, J.; Jiang, B.; Cao, W. IEEE Trans. Ultrason., Ferroelec., a. Freq. Contr. 2000,
47, 285291.
[35] Sessler, G. M. J. Acoust. Soc. Am. 1981, 70, 15961608.
[36] Kar-Gupta, R.; Venkatesh, T.A. J. Appl. Phys. 2008, 104, 02410517 p.
[37] Topolov, V. Yu.; Krivoruchko, A. V.; Bisegna, P.; Bowen, C. R. Ferroelectrics
2008, 376, 140152.
[38] Zhang, R.; Jiang, B.; Cao, W. J. Appl. Phys. 2001, 90, 34713475.
[39] Cao, H.; Schmidt, V. H.; Zhang, R.; Cao, W.; Luo, H. J. Appl. Phys. 2004, 96, 549
554.
[40] Bezus, S.V.; Topolov, V.Yu.; Bowen, C.R. J. Phys. D: Appl. Phys. 2006, 39, 1919
1925.
[41] Smotrakov, V. G.; Eremkin, V. V.; Aleshin, V. A.; Tsikhotsky, E. S. Izv. RAN. Ser.
Fiz. 2000, 64, 12201223 (in Russian).
[42] Takahashi, H.; Tukamoto, S.; Qiu, J.; Tani, J.; Sukigara, T. Jpn. J. Appl. Phys. Pt. 1
2003, 42, 60556058.
[43] Topolov, V. Yu.; Kamlah, M. J. Phys. D: Appl. Phys. 2004, 37, 15761585.
Chapter 13
DOUBLE PEROVSKITES WITH STRUCTURE-

DISORDERED OXYGEN SUBLATTICE
AS HIGH-TEMPERATURE PROTON CONDUCTORS
I. Animitsa*
The Ural State University, av. Lenin 51, Ekaterinburg 620083, Russia
INTRODUCTION
Materials with high proton or oxide ion conductivity are widely studied for their potential
application as electrolytes in fuel cells, electrolysers, batteries, sensors, etc. [1-5]. Acceptor
doped perovskite-type oxides ABO3- where A is Sr, Ba and B is Ce, Zr are well-known high
temperature proton conductors in wet atmosphere. It was found that the main factor
responsible for the appearance of protons in the structure was the presence of oxygen
vacancies. Upon hydration by equilibration with water vapor the oxygen vacancies may be
filled by oxygen from water and, in accordance with dissociative mechanism of water
dissolution, the hydroxide ions are formed. In this case the dissolution of water may be
written [6]:
H2O(g) + Oo+ Vo 2OHo (1)
In late 1990s it was shown that the proton transport can be also realized in perovskite-
related compounds with structural oxygen vacancies. The niobates and tantalates of alkaline-
S
earth metals of a general formula A4(B2B2)O11[V O ]1 are an examples of such structures.
These phases can uptake a significant quantity of water and show a dominant proton transport
below 500-600C. In this work we summarized a principal relationships of the proton transfer
and hydration processes for this novel class of high-temperature proton conductors.
*
502 I. Animitsa
STRUCTURE
Because of the high tolerance of the perovskite structure ABO3, both A- and B-sublattices
can adopt cations with different oxidation state and size, therefore numerous members of this
family may form a various types of superstructures.
The structure of double perovskites is characterized by alternate nonequidimensional
octahedrons containing cations in different oxidation states; at that the B-sublattice is
partitioned into two sublattices (B and B) and the edges are doubled as compared with the
initial perovskite structure. For example, the elpasolite structure (named after the mineral
K2[NaAlF6]) with general composition A2BBO6 has predominantly ordered arrangement of
cations in the B-sublattice (the ratio 1:1 provides the staggered ordering of a small [BO6] and
big [BO6] octahedra , Figure 1); when A and B are identical atoms we come to the cryolite
structure A2(AB)O6. The cryolite structure (named after the mineral Na2[NaAlF6]) may be
considered as an ordered perovskite too. A statistical distribution of two sorts of B-cations
leads to the perovskite structure. The idealized cryolite/ elpasolite structure is the cubic
system in space group Fm3m and contains four formula units per unit cell. In this structure
the 8c crystallographic sites are occupied by larger cations (A-sublattice), B and B-ions
occupy two crystallographically nonequivalent positions: 4a and 4b. Oxygen anions (or F -)
occupy 24e sites.
Figure 1. The structure of double perovskite (elpasolite/ cryolite).

Double Perovskites with Structure-Disordered Oxygen Sublattice 503
Figure 2. Hypothetical perovskite-related structure with structural oxygen vacancies [7].
At a proper combinations of oxidation states of atoms in B-sublattices (for the given A-

cation) an oxygen deficient structures can be exist. The most oxygen-deficient composition
(among double perovskites) corresponds to the general formula A2BBO5.5 or A4B2B2O11,
where if A is nominally +2 cation, then B and B cations may have the charge combinations
(1 + 6), (2 + 5) and (3 + 4). In general, such ordered perovskites (cryolite/ elpasolite) have a
twice cubic perovskite unit cell, implying that oxygen vacancies are random. Thus, 22 oxygen
atoms are distributed over 24e sites, i.e. 8.33% of oxygen sites remains unoccupied (Figure 2.
[7]). These oxygen vacancies are inherent defects of the structure and do not require charge
compensation. When the oxygen vacancies are formed, the metal atoms are shifted and
simultaneously the coordination number of metal is lowered. The octahedrons are
transformed to tetrahedrons or tetrahedral pyramids; therefore, the formation of such phases is
possible when the atoms in the B-sublattice can have different coordination with respect to
oxygen. A typical members of this family are Me6+2M2+5O11, where Me+2 -represents an
alkaline-earth metal and M+5-niobium or tantalum.
The niobates and tantalates are the most widely studied as oxide-ion and proton
conductors. The examples are Sr6Nb2O11 [8, 20], Sr6Ta2O11 [9-13], Ba4Ca2Nb2O11 [14-19] ,
Ba4Ca2Ta2O11 [21]. The homogeneity regions are described too: Sr6-2Nb2+2O11+3 [8,],
Sr6-2Ta2+2O11+3 (0x0.33) [9-10], (Ba1Ca)6Nb2O11 (0.230.47) [16-18] . The
homogeneity range of Sr6-2Nb2+2O11+3 is markedly dependent on the temperature, and the
composition Sr6Nb2O11 does not include in homogeneity range below 800oC (two-phase
region with SrO-impurity) [22]. Other compositions are stable in wide range of the
temperatures, pO2 and pH2O.
Such structure can provide predominant oxygen ion transport at high temperatures and
proton transport at low temperatures and in wet atmospheres. The presence of coordination-
unsaturated polyhedrons makes possible the water dissociative insertion from gas phase.
504 I. Animitsa
Upon hydration (by equilibration with water vapor) the oxygen vacancies may be filled by the
oxygen from the water molecule. When the oxygen vacancies are completely occupied, the
achievement of the theoretically possible hydration limit (50 mol % 2 per formula unit
A4B2B2O11) becomes possible.
Other structural analogues with two octahedral sublattices are also known. For example,
for charge combination of B-cations (1+6). The structure of the compositions
A4+2(B+1B+6)2O11 (A+2 =Sr, Ba, B+1=Li, Na, B+6=Mo, W) was described in [23,24].
WATER INCORPORATION IN OXYGEN-DEFICIENT NIOBATES AND

TANTALATES OF ALKALINE-EARTH METALS
The presence of oxygenhydrogen groups in the solids can be detected by
thermogravimetry method. Upon heating these groups are decomposed and evacuated from
the complex oxide as a water molecules. Therefore, the number of moles of water per one
formula unit of the compound can be estimated. The notation of the total composition as
ABO3nH2O is formal, but convenient for comparing.
The phases Sr6Nb2O11, Sr6Ta2O11, Ba4Ca2Nb2O11 , Ba4Ca2Ta2O11 as well as solid
solutions Sr6 2xNb2 + 2xO11 + 3x (I), Sr6 2xTa2 + 2xO11 + 3x (II), (Ba1 xCax)6Nb2O11 (III) have been
studied. In the systems (I) and (II) within of their homogeneity range the oxygen vacancies
concentration changes, while in the system (III) the oxygen deficiency level does not change:
it remains maximally high for the given cation-charge combinations. It follows from the
thermogravimetry data that during the heatingcooling treatment in wet oxygen or argon
(pH2O = 0.02 atm) the samples of the composition Sr6 2xNb2 + 2xO11 + 3x , Sr62xTa2+2xO11+ 3x ,
and (Ba1 xCax)6Nb2O11 show a reversible mass change. Thus observed effects were explained
by the removal (during the heating) or insertion (during the cooling) of water in the structure
and corroborated by thermal desorption and mass-spectrometry measurements. Figure 3.
demonstrates thermogravimetry data for solid solution Sr62xTa2+2xO11+ 3x [9,10]. Water is
removed in two stages, that reflects the presence of nonequivalent oxygenhydrogen groups.
The correlation between water uptake and the number of vacant oxygen sites was established
(Figure 4) : the water uptake in the composition range Sr6 2xTa2 + 2xO11 + 3x increases with
increasing oxygen vacancy concentration, and the highest reversible water uptake, close to the
theoretical hydration limit, was found for the composition Sr6Ta2O11 [10]. Similar trend was
observed for solid solution Sr6 2xNb2 + 2xO11 + 3x [8]. The general view of TG-curves for other
compositions, such as Ba4Ca2Nb2O11 [16], Ba4Ca2Ta2O11 [21], is the same.
So, the phases with nominally high contents of vacant oxygen sites and the same high
proton concentration are more interesting for detailed studies.
Figure 3. Thermogravimetric data for solid solution Sr6 2xTa2 + 2xO11 + 3x : (a) x= 0 , (b) x= 0.04, (c) x=
0.10, (d) x=0.15, (e) x=0.23; (1) heating, (2) cooling in wet atmosphere (pH2O = 2102 atm), (3)
cooling in dry atmosphere (pH2O = 3 105 atm ) after the first stage of heating [10].
506 I. Animitsa
Figure 4. Dependence of water uptake vs. oxygen vacancies concentration in Sr6 2xTa2 + 2xO11 + 3x
(200C, pH2O = 2 102atm), (1) theoretical and (2) experimental dependencies [10].
The measurement of some properties at constant temperature as a function of pH2O gives

a better understanding of the behaviour of proton incorporation. In terms of quasichemical
formalism, the process of water insertion can be represented by the following equation:
Vo + H 2 O + 2Oo OV''o + 2(OH ) o (2)
where Vo is the oxygen structural vacancy, is the oxygen in its regular position, Vo
is the oxygen in the position of the oxygen structural vacancy, and () is the hydroxo-
group in the oxygen sublattice. In terms of the quasichemical approach the oxygen structural
vacancies can be formally interpreted as ionized defects because the exchange between the
lattice oxygen and the oxygen structural vacancies exists: Vo + Vo + Vo . By
applying the active mass law to equation (2) and bearing in mind the electroneutrality
[OHO]= 2[OVo], it was found that the proton concentration is proportional to pH2O1/3 [11].
Therefore, the dependence of log=f(logpH2O) must be linear with a slope of 1/3, where CH
is the bulk concentration of protons. Thus, the apparent 1/3 slope in the double logarithmic
plot of CH vs pH2O may represent dominance of protons. If oxygen vacancies dominate, for
minority concentrations of protons it was obtained the commonly observed [OH o]
Kp(H2O)1/2 dependency. When all vacancies are filled up, the uptake of water and
concentration of protons will have to saturate and become independent of p(H2O). The
description of the defect concentrations vs. pH2O was given in work [18] too.
Figure 5. Dependence of (a) relative mass change and water uptake and (b) proton concentration in
Sr6Ta2O11 vs. the water vapor pressure at temperature (C) of: (a) (1) 300, (2) 400, (3) 450, (4) 475, (5)
500, (6) 550; (b) (1) 400, (2) 450, (3) 500 [12, 19].
Figure 5. shows the dependences of the water uptake and proton concentration vs. the
water vapor pressure for composition Sr6Ta2O11 [12, 19]. The general view of the
dependencies agrees well with the expected one: over a certain range, the proton
concentration is proportional to p(H2O)1/3. However, at a high pH2O values, that is, when
proton concentrations are significant, the dependences show a sharp increase. This behaviour
reflects a change in crystal structure as a result of water incorporation. The plateau on the
thermogravimetry curves corresponds to approximately complete filling of the oxygen
508 I. Animitsa
vacancies by the oxygen from water molecules, and the concentration of protons will have to
saturate and become independent of pH2O.
X-ray diffraction measurements confirm that incorporation of H2O up to ~0.50 mol H2O
per formula units Sr6Ta2O11 takes place without change of the crystal structure, which could
be described under the cubic space group Fm3m with a cell parameters of approximately 8.32
. Further incorporation of water leads to structure transformation to the orthorhombic space
group Fmmm. The temperature of phase transition increases with increasing pH2O. It is
around 475450oC at the humidity of surrounding atmosphere ( pH2O=0.02 atm) [12].
This phase transition is induced by pH2O (T = const) and temperature (pH2O= const) and
can be observed as a jump on gravimetry curves vs. pH2O or T. At 200oC, the most hydrated
phases correspond to the compositions Sr6Ta2O11 0.96H2O (close to the theoretical value of
1.0). The hydrated sample with the composition approaching the limiting one Sr6Ta2O10(OH)2
had the following crystal lattice parameters: a = 0.8223(6) nm, b = 0.8317(2) nm, and c =
0.8365(9) nm; = = = 90 [12].
A fragment of Sr6Ta2O11 Sr6Ta2O11H2O phase diagram is proposed (wet conditions
pH2O=0.02 atm), Figure 6. [12, 25]. The results may be summarized as follows.
(i) At temperatures higher than 1000oC, dry double oxide Sr6Ta2O11 with cubic
structure exists.
(ii) Over the temperature range of 5001000 oC, intercalation of water with formation of
a solid solution Sr6Ta2O11mH2O takes place. The quantity of intercalated water
depends on the temperature and reaches m 0.5 mol ( the line di in the diagram
corresponds to equilibrium of solid solution of Sr6Ta2O11 mH2O with water vapor).
This process occurs without serious structure reconstruction and, evidently, such
dissolution may be described in the frame of commonly used quasi-chemical
approach, where water is regarded as an impurity.
(iii) Further water incorporation at T500 oC results in change in crystal structure and
formation of a new hydrated compound the hydroxo-phase Sr6Ta2O10 n(OH)2n (n
1) with orthorhombic symmetry. The quantity of intercalated water increases when
the temperature decreases and it indicates that this compound has a homogeneity
range. In the phase diagram the peritectoid line ic corresponds to the compound
decomposition, where three phases are in equilibrium: two solids and water vapor.
The point c is the peritectic point in which one solid phase transforms into another.
However, by contrast to this, the hydration of the nominal composition Sr6Nb2O11 (not
included in homogeneity range of Sr6-2Nb2+2O11+3) is accompanied by the formation of the
secondary phase Sr(OH)2 [20].
For solid solution (Ba1yCay)6Nb2O11 the thermogravimetric studies carried out in wet
argon or oxygen atmosphere showed that the water uptake is nearly the same: 0.850.95
mole, which approached the theoretical hydration limit [16,17]. The thermogravimetric
measurements confirm that the dependences log=f(logpH2O) are linear, with a slope of 1/3
(Figure 7. represents limited ranges of the dependences without phase transformations [16]).
Figure 6. Quasibinary phase diagram of Sr6Ta2O11 Sr6Ta2O11H2O, pH2O = 0.02 atm [12].
It was shown by X-ray diffraction analysis that in the row of solid solution (Ba1
xCax)6Nb2O11 only one composition Ba4Ca2Nb2O11 (x= 0.33) changed its initial cubic
symmetry upon hydration. During the process of water incorporation the compound
Ba4Ca2Nb2O11 undergoes structural changes due to formation of a hydroxo-phase of
approximate composition Ba4Ca2Nb2O10()2 with a lower, monoclinic symmetry (the space
group C2/m) [14, 16]. The composition Ba4Ca2Nb2O11 is characterized by equal number (1:1)
of calcium and niobium cations in the B-sublattice. Under such chequer-wise alternation of
the large and small octahedrons, the crystal lattice energy appears being lower than for the
phases without long range ordering in the cation sublattice. Obviously, any displacement of
the cations in the 1:1 ordered phases, e.g., during the water insertion, results in
macroscopically evident lattice distortion that is detected by the X-ray diffraction analysis.
510 I. Animitsa
Figure 7. Dependence of proton concentration in a Ba4Ca2Nb2O11 on the water vapor pressure at a

temperature (C) of: (1) 300, (2) 350, (3) 400, (4) 500 [16].
STATES OF H+-CONTAINING SPECIES IN HYDRATED NIOBATES AND

TANTALATES OF ALKALINE-EARTH METALS
WITH A PEROVSKITE-RELATED STRUCTURE
The oxygenhydrogen groups can be identified by using the IR-spectroscopy and proton
magnetic resonance (1H NMR) methods. Generally, OHn- groups can exist in solid inorganic
materials as OH, H2O or H3O+ groups.
IR-spectra of hydrated niobates and tantalates of alkaline-earth metals confirm the
presence of the oxygenhydrogen groups: the IR spectra for all studied phases contain a broad
band at ~3500 cm1 (absorption at 28003700 cm1 should be assigned to stretching
vibrations of oxygen-hydrogen groups).
However, in the stretching vibration region it is impossible to separate the bands related
to vibrations in different types of the oxygenhydrogen groups. Therefore, of interest is the
studying of deformation vibrations. It is known that the frequency of deformation vibrations
of isolated water molecule equals 1595 cm1. Association of the H2O molecules (in particular,
for reason of hydrogen bonds) results in the shift of the band to region 16001680 cm1. The
hydroxonium ion H3O+ absorbs at higher frequencies (16801800 cm1); the hydroxo- group
OH ( MOH) absorbs below 1500 cm1. The number of components in the spectrum, as a
rule, corresponds to the number of different types of the oxygenhydrogen groups.
Figure 8. IR-spectra of hydrated niobates and tantalates of alkaline-earth metals: (a) Ba4Ca2Nb2O11, (b)
Sr5.92Ta2.08O11.12, (c) Sr6Ta2O11 [7].
Figure 8. shows IR-spectra of hydrated niobates and tantalates of alkaline-earth metals. In

all spectra we can distinguish a very intense band at ~1450 cm1, which can be related to the
OH group vibrations, moreover, this band is a superposition of several contributions. It was
shown [7, 13] that the IR band at ~1450 cm1 can be correlated with the water uptake and is
not observed for compounds without water uptake. The band at ~1650 cm1 disappears after
heating at 100oC and this indicates that this band corresponds to absorbed water.
To identify the OHn -groups, the wide-line 1H-NMR method was used too (the samples
were hydrated down to 200 oC in order to eliminate the presence of absorbed water). Figure 9.
presents 1H-NMR spectra recorded at -150 oC for some samples. The spectra were complex
and represented a superposition of several components caused by the presence of different
oxygenhydrogen groups [13]. Three components may be distinguished in a general case.
512 I. Animitsa
Figure 9. 1H NMR-spectra for hydrated samples with the composition: (a) Sr6Ta2O11 , (b)
Ba4Ca2Nb2O11 (150C). Points: experimental spectrum, line: model spectrum [ 13].
(i) The narrow signal (defined as HI) can be described as a Gaussian with a broadening
parameter HI = 0.40.8 G. It represents isolated OH groups with an inter-proton
distance RHH = 3.03.5 .
(ii) The Gaussian line with the parameter HII = 1.52.0 G ( HII) is due to closely
spaced OH groups. The distance between the protons is RHH = 2.22.4 . They are
connected probably with the same metal atom and belong to the same edge of an
octahedron, for example:
OH
M<
OH
(iii) A broad signal at HIII =4.55.5 G (defined as HIII) is modeled as a two-spin system
(when the interaction of two nearest neighbors dominates). However, it cannot be
related to water molecules. The distance between protons was estimated at 2.0 ,
which is much larger than the normal distance between water protons in crystals of
1.521.65 . This signal may be caused by the appearance of close-spaced OH
groups in a structural element of the coordination-unsaturated polyhedrons, shown in
Figure 10 [13]. Probably, such strong interaction of protons in OH- groups reflects an
intermediate state HOHO O HOH.
Figure 10. Scheme of formation of close-spaced OH- groups. () oxygen vacancy; () hydroxyl [13].
Figure 11. 1H-NMR spectra recorded at -150 oC for Ba4Ca2Nb2O11 with different water uptake. (a)
Ba4Ca2Nb2O110.15H2O; (b) Ba4Ca2Nb2O11 0.25H2O; (c) Ba4Ca2Nb2O11 0.80H2O [13].
514 I. Animitsa
Figure 12. Dependence of total conductivity of solid solutions Sr6 2xTa2 + 2xO11 + 3x vs. the oxygen partial
pressure in (1) wet atmosphere (pH2O = 2 102 atm) and (2) dry atmosphere (pH2O =3 105 atm ) at (a)
x = 0, (b) x = 0.15, (c) x = 0.23 , T = 600C [19].
An 1H-NMR analysis of partially hydrated sample Ba4Ca2Nb2O11 containing 0.15 and

0.25 mol water was performed in order to determine the distribution of proton sites depending
on the concentration of intercalated water (Figure 11., [13]). Separation of the NMR spectra
into components allows to estimate the quantitative ratios of protons in each component of a
spectrum. The results are given in Table 1. The following conclusions may be drawn:
the fraction of protons in isolated OH- groups (the HI line) was a maximum in
Sr6Ta2O11;
in the niobates (independently of the nature of alkaline-earth metals), most of the
protons were present as closely spaced OH- groups (the HII line) and protons, which
were defined as HIII.
the fraction of protons of the HIII component diminishes with decreasing degree of
hydration, i.e. the larger the degree of hydration, the more probable the appearance of
closely spaced OH- groups.
The difference between tantalates and niobates is due probably to different coordination
tendencies of Ta and Nb. Since all phases have an oxygen vacancies, the structure must
contain polyhedra with coordination number less than 6. Ta+5 favors a squarepyramidal
coordination [TaO5] rather than a tetrahedral coordination, whereas a tetrahedral coordination
for Nb+5 is more common. In the case of full hydration (i.e. if water molecules occupy all
oxygen vacancies), an initial tetrahedron is restored to an octahedron in accordance with
appearance of the structural element [NbO4(OH)2], in which two OH- groups belonging to the
same metal atom are formed. The formation of an octahedron from a pyramidal coordination
is accompanied probably by appearance of [TaO5(OH)], leading to formation of isolated OH-
groups.
Table 1. The ratios of protons in each component of 1H NMR spectrum [13]
Composition HI HII HIII

Sr6Ta2O111.0 H2O 0.23 0.23 0.54
Ba4Ca2Nb2O11 0.95H2O 0.01 0.17 0.82
Ba4Ca2Nb2O11 0.80H2O 0.07 0.23 0.70
Ba4Ca2Nb2O11 0.25H2O 0.10 0.50 0.40
Ba4Ca2Nb2O11 0.15H2O 0.20 0.80 -
Comparing the TG and 1H-NMR data, it is possible to propose a dehydration scheme,

because, the ratio of crystallographically (and, consequently, energetically) nonequivalent
protons changes with temperature. During heating the hydrated sample, the beginning of an
intensive weight loss corresponds to removal most of closely spaced protons. This process
takes place in a relatively narrow temperature interval and is reflected in TG curves as a well-
resolved step. As the sample is heated further, the dehydration deals with the removal of
relatively isolated hydroxyl groups. This dehydration process is characterized by a continuous
decrease in the weight over a relatively wide temperature interval (up to 800 oC).
Electrical Properties
By the studying of the total conductivity of the solid solution Sr6 2xTa2 + 2xO11 + 3x under
the varying oxygen partial pressure pO2 (wet and dry conditions) the general relationships
were shown. Figure 12. demonstrates the evolution of the conductivity isotherms [19, 26]. For
the compositions with high concentration of oxygen vacancies 0<x0.20 (that is, with high
conductivity) the effect of humidity is most pronounced over a wide rang of pO2 . The total
conductivity of these samples was found to be independent of the oxygen partial pressure for
all investigated temperatures. This indicates that the electronic component of the conductivity
is negligible. Some contribution of the p-type electronic conductivity was observed for
compositions x>0.20 at high oxygen partial pressure (dry atmosphere), in this case the
equation may be written:
Vo + 1 / 2O2 2h + O ' 'V (3)

o
In wet atmosphere the hole concentration decreases as a result of the shift of the
equilibrium (3) to the left. The summary process that includes the equilibria (2) and (3) can be
written as follows:
H 2 O + 2h + 2Oo 1 / 2O2 + 2(OH ) o (4)

516 I. Animitsa
Figure 13. Temperature dependence of conductivity of Sr5.92Ta2.08O11.12 in wet atmosphere (pH2O = 2

102 atm). (1) proton , (2) oxygen-ion (estimated from EMF method) and (3) total conductivities. Curve
(4) is the calculated value of the protonic conductivity, H = (ion)wet (ion)dry [19].
Therefore, depending on the ratio of the hole and proton concentrations, the total
conductivity can remain constant or even decrease at high pO2 (air) and wet conditions.
The comparison of temperature dependences of the total conductivity in dry and wet
atmospheres demonstrates the most typical situation: the conductivity in wet atmospheres is
higher than in dry as a result of the contribution of proton conductivity in wet atmosphere ;
the differences and, thus, the significance of protonic conduction, decrease with increasing
temperature, and mainly disappear at temperatures above 800900oC.
Figure 13. shows the partial conductivities obtained from the EMF-type transport number
measurements for the composition Sr5.92Ta2.08O11.12 (x = 0.04) from the homogeneity region
of Sr6 2xTa2 + 2xO11 + 3x [19]. The magnitude and temperature dependence of the partial proton
conductivity confirm that for wet conditions protons become dominating defects and also
dominate the conductivity below 550 C. The transition from predominantly oxygen to proton
conductivity is accompanied by an overall change in activation energy from 0.96 to 0.56 eV,
the latter being a typical value for proton mobility in oxygen deficient perovskites [11, 19].
For the phases demonstrating a phase transition during water incorporation a local
maximum on temperature dependences of conductivity is observed (Figure 14, 15) [11, 14].
Figure 14. Temperature dependence of conductivity of Sr5.92Ta2.08O11.12 in (1) wet atmosphere (pH2O =
2 102 atm) and (2) dry atmosphere (pH2O =3 105 atm ) [11].
Figure 15. Temperature dependence of conductivity of single crystal Ba4Ca2Nb2O11 in wet (pH2O = 2
102 atm) and dry atmospheres (pH2O =3 105 atm ) [14].
518 I. Animitsa
Figure 16. Dependence of proton conductivity vs. proton concentration in the solid solution Sr6 2xTa2 +
2xO11 + 3x (
= 550) [26].
The proton conductivity of the solid solution Sr6 2xTa2 + 2xO11 + 3x as a function of their
composition was analyzed [26]. Figure 16. shows the dependence of the proton conductivity
vs. the concentration of protons for a variety of x . It is seen that with the decreasing of the
parameter x, that is, with the increasing of the proton concentration, the value of proton
conductivity increases monotonically. This result can be explained by the increase not only in
the proton concentration but also in the proton mobility. Figure 17. shows the concentration
dependence of the proton mobility [26], calculated in according with the relation:
H = ZeC H H , (5)
where Ze is the charge, CH is the proton bulk concentration, H is the proton mobility.
The correlations between some experimental data are represented in Figure 18 [26]. The
same tendency was observed for the solid solution Sr6 2xNb2 + 2xO11 + 3x [8].
The effect of cationic sublattice on the value of proton conductivity was demonstrated by
the example of the solid solution (Ba1 yCay)6Nb2O11 with equal concentration of oxygen
vacancies. With the decreasing of the parameter y (that is, with the increasing of BaO content)
the oxygen-ion conductivity increased (Figure 19); the proton conductivity increased in the
same way [19]. This effect can be explained by the increasing of the current carrier mobility.
Because the cell parameter increases with the increasing of BaO content, the increase in the
interatomic distance results in the decrease of the energy consumed in the oxygen ion
migration, which leads to the increase in its mobility. The proton mobility also increases with
the increasing of the oxygen ion mobility. Thus, the dynamics of the oxygen sublattice affects
the proton conductivity.
The studies of the total conductivity under the changing pH2O allowed revealing the
principal trends (Figure 20 shows the typical dependences of = f(pH2O) for Sr6Ta2O11 [11]).
Similar plots were obtained for (Ba1 yCay)6Nb2O11 [26]):
At temperatures higher than 600oC the conductivity is almost independent on

p(H2O). This indicates that protons are only present as minority defects and that
proton conduction is insignificant.
At lower temperatures (with the increasing of pH2O) over a certain range of
intermediate water vapour pressures, the conductivity is proportional to p(H2O)1/3.
Increasing the water vapour pressure further, the conductivity shows a sharp increase
near p(H2O) = 0.02 atm in the temperature range 450-475 C. This behaviour agrees
with the thermogravimetry data and reflects a change in crystal structure as a result
of water incorporation. The breaking of monotonic character of the dependences is
caused by the formation of hydroxo-phase.
At low temperatures and high p(H2O), when all vacancies are filled up, the
conductivity exhibits a weak dependence on water vapour pressure.
The recent investigation [27] demonstrates that the double perovskites are promising
materials for electrochemical devices.
Figure 17. Dependence of proton mobility vs. proton concentration in the row of solid solution Sr6
2xTa2 + 2xO11 + 3x (
= 550) [26].
520 I. Animitsa
Figure 18. Concentration dependencies of (a) proton (1) and oxygen-ion (2) conductivities , (b) proton
transport number, (c) unit cell in the row of solid solution Sr6 2xTa2 + 2xO11 + 3x [26].
Figure 19. Dependence of (1) oxygen-ion conductivity and (2) proton conductivity on the composition
of solid solution (Ba1 yCay)6Nb2O11 (T = 400C) [19].
Figure 20. Dependences of total conductivity vs. partial water pressure for Sr6Ta2O11 . Temperature
(C): (1) 400, (2) 450, (3) 475, (4) 500, (5) 600, [11].
CONCLUSION
Data on the thermogravimetry, spectroscopy, and electrical properties as functions of T,
p(H2O), p(O2) for niobates and tantalates of alkaline-earth metals with structure disordering
of the oxygen sublattice are summarized. These phases reversibly incorporate of water and
exhibit high-temperature proton transport.
522 I. Animitsa
Water uptake is close to the theoretical value expected from complete filling of oxygen
vacancies by water. Incorporation of significant quantity of water leads to a phase
transformation owing to formation of hydrated compositions close to Me6+2M2+5O10n(OH)2n.
The phase transformation is induced by both temperature and pH2O changes. For the phases
demonstrating a phase transition during water incorporation a local maximum on temperature
dependences of conductivity and a jump on pH2O dependences are observed.
A wide-line 1H NMR study indicates that the proton signal consists of three components,
because protons were present in the form of energetically unequivalent OH- groups, namely
(i) relatively isolated OH- groups, (ii) paired OH- groups located on one edge of an
octahedron, and (iii) closely spaced OH - groups, which are ordered within some defect
cluster. The niobates and tantalates have different ratios of those groups that may be
explained by different coordination tendencies of niobium and tantalum and as a consequence
a different arrangement of oxygen vacancies. The presence of different OH- groups
determines staging of water removal.
It was shown that the mobility of protons depends on the dynamics of the oxygen
sublattice: with the increasing of the lattice parameter and, as a consequence, the oxygen-ion
conductivity, proton mobility increases.
REFERENCES
[1] Iwahara H., Asakura Y., Katahira K., Tanaka M. Solid State Ionics 168 (2004) 299.
[2] Sundmacher K., Rihko-Struckmann L.K., Galvita V. Solid electrolyte membrane
reactors: Status and trends//Catalysis Today 104 (2005) 185199.
[3] Athanassioua C., Pekridisc G., Kaklidisc N., Kalimeric K., Vartzokac S., Marnellosb
G. Hydrogen production in solid electrolyte membrane reactors (SEMRs)
//International Journal of Hydrogen Energy 32 (2007) 38 54.
[4] Guangyao Meng , Guilin Ma, Qianli Ma , Ranran Peng , Xingqin Liu. Ceramic
membrane fuel cells based on solid proton electrolytes//Solid State Ionics 178 (2007)
697703.
[5] Marques F.M.B., Kharton V.V., Naumovich E.N., Shaula A.L., Kovalevsky A.V.,
Yaremchenko A.A. Oxygen ion conductors for fuel cells and membranes: selected
developments// Solid State Ionics 177 (2006) 16971703.
[6] Norby T. Korean J. Ceram. 1998. V. 4. P. 128.
[7] Colomban Ph., Romain F., Neiman A., Animitsa I. Double perovskites with oxygen
structural vacancies: Raman spectra, conductivity and water uptake.// Solid State
Ionics. 2001. V.145/1-4, p.339-347.
[8] Glckner R., Neiman A., Larring Y., Norby T. Protons in Sr3(Sr1+xNb2-x)O9-3x/2
perovskite // Solid State Ionics, 1999. V.125. P.369-376.
[9] Animitsa I., Neiman A, Sharafutdinov A, Nochrin S. Strontium tantalates with
perovskite-related structure.// Solid State Ionics. 2000. V.136-137, p.265-271.
[10] Animitsa I. E., Neiman A. Ya., Sharafutdinov A. R., Kazakova M. G. Strontium
Tantalates with Perovskite Structure: Their Conductivity and High-Temperature
Interaction with Water// Russian Journal of Electrochemistry, Vol. 37, No. 3, 2001,
pp. 266272.
[11] Animitsa I., Norby T., Marion S., Glockner R., Neiman A. Incorporation of water in
strontium tantalates with perovskite-related structure.// Solid State Ionics. 2001.
V.145/1-4, p.357-364.
[12] nimitsa I., Neiman A., Titova S., Kochetova N., Isaeva E., Sharafutdinov,
Timofeeva N., Colomban Ph. Phase relations during water incorporation in the
oxygen and proton conductor Sr6Ta2O11// Solid State Ionics. 2003. V.156, 95-102.
[13] Animitsa I., Denisova T., Neiman A., Nepryahin A., Kochetova N., Zhuravlev N.,
Colomban Ph. States of H+-Containing Species and Proton Migration Forms in
Hydrated Niobates and Tantalates of Alkaline-Earth Metals with a Perovskite-
Related Structure// Solid State Ionics. 2003. V.62-163, 73-81.
[14] nimitsa I., Neiman ., Kochetova N., Melekh B., Sharafutdinov . Proton and
oxygen-ion conductivity of Ba4Ca2 Nb2O11// Solid State Ionics. 2003. V. 162-163,
63-71.
[15] Animitsa I., Neiman A., Kochetova N., Korona D., Sharafutdinov A. Chemical
diffusion of water in the double perovskites Ba4Ca2Nb2O11 and Sr6Ta2O11 // Solid
State Ionics. 2006. V.177. .2363-2368.
[16] Kochetova N. A., Animitsa I. E., Neiman A. Ya. The Synthesis and Properties of
Solid Solutions Based on Ba4Ca2Nb2O11// Russian Journal of Physical Chemistry A,
2009, Vol. 83, No. 2, P. 203208.
[17] Ashok A., Kochetova N., Norby T., Olsen A. Structural study of the perovskite
system Ba6 yCayNb2O11 hydrated to proton conducting Ba6yCayNb2O10(OH)2//Solid
State Ionics 179 (2008) 18581866.
[18] Korona D. V., Neiman A. Ya., Animitsa I. E., Sharafutdinov A. R. Effect of
Humidity on Conductivity of Ba4Ca2Nb2O11 Phase and Solid Solutions Based on
This Phase//Russian Journal of Electrochemistry, 2009, Vol. 45, No. 5. P. 586592.
[19] Animitsa I.E. High-Temperature Proton Conductors with Structure-Disordered
Oxygen Sublattice// Russian Journal of Electrochemistry, 2009, Vol. 45, No. 6. P.
668676.
[20] Jalarvo N., Haavik C., Kongshaug C., Norby P., Norby T. Conductivity and water
uptake of Sr4(Sr2Nb2)O11nH2O and Sr4(Sr2Ta2)O11nH2O// Solid State Ionics (2009).
[21] Baliteau S., Mauvy F., Fourcade S., Grenier J.C. Investigation on double perovskite
Ba4Ca2Ta2O11// Solid State Sciences 11 (2009) 15721575.
[22] Spitsyn, V.I., Ippolitova, E.A., Kovba, L.M., et al., Zh. Neorg. Khim., 1982, vol. 27,
p. 827.
[23] Hikichi Y., Suzuki S. Preparation of cubic perovskites A(B2/5W3/5)O3 (A=Ba or Sr,
B=Na or Li). // Journal of the American Ceramic Society. 1987. V.70. N 5. P. 99-
100.
[24] Hoffmann R., Hoppe R. Ein neuer perowskit mit Lucken im Anionenteil:
Ba4[Na2W2O11]. // Z.anorg.allg.chem. 1989. V.575. P.154-164.
[25] Schober T. Phase diagrams in the proton conductor systems Sr6Ta2O11nH2O and
Sr5.92Ta2.08O11.12nH2O// Solid State Ionics, V. 177. Issues 5-6. 2006. P. 471-474. 26.
Kochetova N.A., Animitsa I. E., Neiman A. Ya. Electrical properties of solid
solutions based on strontium tantalate with perovskite related structure // Russian
Journal of Electrochemistry. 2010, Vol. 46, No. 2.
524 I. Animitsa
[26] Bhella S. S., Thangadurai V. Synthesis and characterization of carbon dioxide and
boiling water stable proton conducting double perovskite-type metal oxides //
Journal of Power Sources 186 (2009) 311319.
Chapter 14
SHORT- AND LONG-RANGE FERROMAGNETIC

ORDERS IN LA1-XPBXMNO3:
CRITICAL BEHAVIOR STUDY
The-Long Phan* and Seong-Cho Yu

BK-21 Physics Program and Department of Physics, Chungbuk National University,
Cheongju 361-763, South Korea
ABSTRACT
We have studied the influence of Pb-doping on critical behavior of polycrystalline
manganites of La1-xPbxMnO3 (x = 0.1, 0.2 and 0.3) around their ferromagnetic-
paramagnetic phase transition temperature (TC) by means of analyzing isothermal
magnetization data M-H. According to the critical criterion proposed by Banerjee, it was
found that all three samples undergo the second-order phase transition as showing a
positive slope of isotherm plots of H/M versus M2. The critical parameters were
determined by the modified Arrott method, used the Arrott-Noakes equation of state M1/
versus (H/M)1/ and the asymptotic relations in the critical region. The obtained results
reveal that with increasing the Pb-doping concentration in La1-xPbxMnO3, the TC value
increases from 162.0 K (for x = 0.1) to about 346.3 K (for x = 0.3). Concurrently, there is
a remarkable variation in their critical exponents: = 0.498, = 1.456 and = 3.92 for x
= 0.1; = 0.499, = 1.241 and = 3.49 for x = 0.2; and = 0.502, = 1.063 and = 3.12
for x = 0.3. Having compared to the critical parameters expected from the theoretical
models of mean-field theory, 3D-Ising model, and 3D-Heisenberg model, we have found
an existence of the short-range ferromagnetic order in the samples with x = 0.1 and 0.2,
and the long-range ferromagnetic order in the x = 0.3 sample. Under such circumstances,
magnetic interaction mechanisms between Mn3+-Mn4+ in the samples would be changed,
particularly in the paramagnetic region.
*
Corresponding author: Email: [email protected]; Phone: +82-43-261-2269 ; Fax: +82-43-275-6416
526 The-Long Phan and Seong-Cho Yu
1. INTRODUCTION
Recently, the discovery of magnetoresistance and magnetocaloric phenomena in hole-
doped perovskite-type manganites of Ln1-xAxMnO3 (Ln = La, Pr, Nd; A = Ca, Sr, Ba, Pb etc.)
around their ferromagnetic (FM) to paramagnetic (PM) phase transition temperature (TC) has
attracted interest in the community of solid-state physics [1-9]. Among these manganite
compounds, the systems of La1-xCaxMnO3, La1-xSrxMnO3, and La1-xPbxMnO3 with x = 0.1 -
0.4 have received intensive attention as showing colossal magnetoresistance (CMR) and large
magnetocaloric (MC) effects around room temperature [4, 5, 10]. This could lead to
development of some manganites-based electronic devices, such as magnetic sensitive
sensors, reading/recording heads, and cooling systems (magnetic refrigerators).
Basically, the CMR effect in perovskite manganites has been explained by means of the
mechanism of double-exchange interaction taking place between Mn3+ and Mn4+ ions, which
was supposed by Zener [11]. In some cases, however, this model only is not enough to
explain fully the magneto-transport properties observed by experimental physicists,
particularly for the material system La1-xCaxMnO3. The addition of the polaron effect (known
as the Jahn-Teller effect) caused by strong electron-phonon interactions is necessary [12].
With such opinions, the magnetic and magneto-transport properties of perovskite manganites
are definitely dependent on both the length bond Mn-O and Mn-O-Mn angle bond. As a
matter of fact, these parameters are influenced strongly by concentration and the radius of
divalent ions substituted into the A site (i.e., Ca, Sr, Ba, or Pb). Their presence in manganite
compounds Ln1-xAxMnO3 would lead to a lattice distortion (where the angle bond Mn-O-Mn
is smaller than 180o, and the length bond Mn-O becomes longer or shorter than its primitive
value), and to the change in the concentration ratio of Mn3+/Mn4+. This is a key point that it
could be exploited to control the magnetic, electrical and magneto-transport properties of
perovskite manganites as expected. Notably, the possibility of utilizing perovskite manganites
in electronic devices is currently still limited, probably since their physical behaviours are
unstable and very sensitive to oxygen-concentration change [13]. This feature makes
perovskite manganites become interesting in the field of fundamental research rather than the
application aspect. The improvement of weaknesses in these materials is thus necessary, and
currently an issue of interest.
To better understand the CMR and MC effects, it is necessary to understand the nature of
the FM-PM phase transition and interaction mechanism taking place between Mn ions in the
manganite host lattice in the vicinity of TC. Besides the study of magneto-transport behaviors,
one of the most effective approaches is to consider the critical exponents , , and , which
are associated with the spontaneous magnetization Ms(T,0), initial magnetic susceptibility
0(T), and the critical isotherm M(TC,H), respectively. Based on these critical parameters in
comparison with those predicted by the mean-field theory, 3D Ising and 3D Heisenberg
models [14], one can figure out the magnetic-interaction type (or interaction mechanism)
existing in a material, i.e long- and/or short-range ferromagnetic interactions. In this chapter,
we present a careful study of critical behavior of polycrystalline perovskite manganites of
La1-xPbxMnO3 (x = 0.1, 0.2 and 0.3) prepared by conventional solid-state reaction. We shall
point out that the increase of Pb-doping concentration in La1-xPbxMnO3 influences
remarkably their critical parameters of TC, , , and . For example, TC increases from about
162.0 K (for x = 0.1) to about 346.3 K (for x = 0.3), and the exponents obtained is = 0.498,
Short- and Long-range Ferromagnetic Orders in La1-xPbxMnO3 527
= 1.456 and = 3.92 for x = 0.1; = 0.499, = 1.241 and = 3.49 for x = 0.2; and =
0.502, = 1.063 and = 3.12 for x = 0.3. Having compared to those expected by the
theoretical models, we have found an existence of the short-range ferromagnetic order in the
samples with x = 0.1 and 0.2, and the long-range ferromagnetic order in the x = 0.3 sample.
This is due to an increase in Mn4+ concentration in La1-xPbxMnO3 as increasing the Pb-doping
concentration, in which magnetic interaction mechanisms between Mn3+ and Mn4+ would be
changed.
Figure 1. Field dependences of magnetization for (a) La0.8Pb0.2MnO3, and (b) La0.7Pb0.3MnO3 measured
at various temperatures.
2. EXPERIMENT
Three polycrystalline samples of La1-xPbxMnO3 (x = 0.1, 0.2 and 0.3) were prepared by
the conventional solid-state reaction method. The precursors of commercial powders La2O3,
PbO and MnCO3 (99.9 % in purity) combined with stoichiometrical quantities were well
mixed, ground and pressed into pellets. These pellets were then pre-sintered at 900 C for 15
hrs. After several times of intermediate grinding and pressing, they were sintered at 920 C
for 15 hrs in air. The cooling was carried out slowly to room temperature. It should be noticed
that these pellets were annealed at a quite low temperature comparing to other perovskite
manganites (usually at temperatures 1100 -1300 oC), because of the presence of PbO (its
melting point is about 830 oC). The single phase in an orthorhombic structure of the final
products was confirmed by an X-ray diffractometer [10, 15]. The dependence of
magnetization on the magnetic field, M-H, at various temperatures were performed on a
vibrating sample magnetometer (VSM) with the applied field up to 17 kOe. An incremental
temperature (T) for each M-H measurement is 3 K near TC, and 5 K for the other regions.
3. RESULTS AND DISCUSSION

Figure 1 shows representatively the M-H data for La1-xPbxMnO3 (x = 0.2 and 0.3)
measured at different temperatures in the vicinity of TC. It appears that the M-H curves versus
temperature separate into two characteristic regions. In the first region T < TC, magnetization
increases nonlinearly with increasing the magnetic field, and does not reach to a saturation
value. Meanwhile, in the second region T > TC, the field dependence of magnetization is
linear. These features seem popular and could be found in most perovskite manganites in the
ferromagnetic and paramagnetic regimes [1, 5, 16], excepting for the system La1-xCaxMnO3
where its magnetic phase transition is usually accompanied with structural changes, and thus
its M-H curves become anomalous, see Refs. [3, 4, 7]. This could explain why the MC effect
in La1-xCaxMnO3 (x 0.3) becomes largest among perovskite manganite compounds [8].
To assess the nature of magnetic phase transition in the samples, we firstly pay attention
to the free energy G near the critical temperature, which can be expressed in terms of the
order parameter M (magnetization) as follows [9]
G (T , M ) = G0 + aM 2 + bM 4 MH , (1)
where a and b are temperature-dependent parameters. From the equilibrium condition of

energy minimization, G / H = 0 , it is easily deduced the magnetic equation of state
(corresponding to the mean-field theory) [14]
H / M = 2a + 4bM 2 . (2)
Figure 2. The Arrott plots of M2 versus H/M for the sample La0.7Pb0.3MnO3.
With this equation, the representation curves of H/M versus M2 should be straight lines,
and the intercept of these straight lines on the H/M axis gives the order of magnetic transition.
The straight lines with a positive slope indicates for the second-order transition, and with a
negative slope indicates for the first-order transition, known as the Banerjee criterion [17]. It
could be generalized Eq. (2) to the form of (H/M)1/ = (T-TC)/T1 + (M/M1)1/, the so-called
Arrott-Noakes equation of state [18]. The representation of (H/M)1/ versus M1/ is then
known as the modified Arrott plots. In the case of the mean-field theory, the values and
are 0.5 and 1.0, respectively, which leads to the regular Arrott plots [19]. Concerning our
work, Figure 2 shows the performance of H/M versus M2 for a representative sample of
La0.7Pb0.3MnO3. One can see clearly that the plots of H/M versus M2 give a positive slope,
similar to those of La0.7Sr0.3MnO3 reported by Mira et al. [4]. This feature was also met in the
other samples with x = 0.1 and 0.2, not shown here. It means that a second-order transition
exists in our samples. Concerning the first-order transition, it can be found in compositions
La0.7Ca0.3MnO3 and La1.2Sr1.8Mn2O7 [4, 20].
Mathematically, around the critical temperature the second-order transition could be
described by a set of power functions through the exponents , , and , which are obtained
basing on the following asymptotic relations [21]
Ms(T)=limH0(M)=M0(-) < 0, (3)

Figure 3. Temperature dependences of spontaneous magnetization, Ms(T), and the inverse initial
magnetic susceptibility, 0
1
(T ) , for the samples (a) La0.9Pb0.1MnO3, (b) La0.8Pb0.2MnO3 and (c)
La0.7Pb0.3MnO3 are fitted to the power laws of Eqs. (3 and 4). respectively.
0-1(T)=limH0(H/M)=(h0/M0) > 0, (4)
M = DH1/, =0, (5)

where M0, h0, and D are the critical amplitudes, and = (T-TC)/TC is the reduced temperature.
According to scaling theory [5, 21], on the other hand, the critical exponents could be also
obtained via the magnetic equation of state in the critical region written as
M ( H , ) =| | f ( H / | | + ) , (6)
in which scaling functions of f+ and f- represent for variations in the temperature regions of
T>TC and T<TC, respectively. This relation implies that M/|| as a function of H/||+ falls on
two different branches, one for temperatures T>TC and the other for T<TC corresponding to
the paramagnetic and ferromagnetic regions, respectively. Accordingly, the critical exponents
and determined from Eqs. (3 and 4) must be in agreement with the descriptions of Eq. (6).
Table 1. The critical parameters (TC, , , and ) obtained for La1-xPbxMnO3 are
compared to those expected by theoretical models (Mean-field theory, 3D Heisenberg
model, 3D Ising model, and Tricritical mean-field theory) and perovskite manganites
(and cobaltites)
Material TC (K) Ref.

La0.9Pb0.1MnO3 0.498 1.456 3.92 162.0 This work
Mean-field theory 0.5 1.0 3.0 [14]
3D Heisenberg model 0.365 1.336 4.80 [14]
3D Ising model 0.325 1.241 4.82 [14]
Tricritical mean-field theory 0.25 1.0 5.0 [21]
La0.8Sr0.2MnO3 0.50 1.08 3.13 315 [5]
La0.8Sr0.2CoO3 0.46 1.39 4.02 198.5 [23]
La0.75Sr0.25CoO3 0.46 1.39 4.02 222.3 [23]
La0.7Sr0.3CoO3 0.43 1.43 4.38 223.4 [23]
LaMnO3.14 0.415 1.47 NA 141.0 [22]
La2NiMnO6 0.408 1.295 4.174 271.1 [16]
Nd0.6Pb0.4MnO3 0.374 1.329 4.54 156.5 [26]
La0.7Sr0.3MnO3 0.37 1.22 4.25 354 [1]
La0.875Sr0.125MnO3 0.37 1.38 4.72 186.1 [6]
La0.67Ba0.33MnO3 0.365 1.22 NA 306.1 [24]
La0.8Ca0.2MnO3 0.36 1.45 5.03 174.0 [2]
Pr0.77Pb0.23MnO3 0.344 1.352 4.69 167.0 [25]
La0.5Sr0.5CoO3 0.321 1.351 4.39 223.2 [27]
La0.6Ca0.4MnO3 0.25 1.03 5.0 255.5 [3]
La0.9Te0.1MnO3 0.201 1.27 7.14 239.5 [9]
La0.7Ca0.3MnO3 0.14 0.81 1.22 222.2 [1]
Figure 4. The plot of ln(M) versus ln(H) in the vicinity of TC for (a) La0.8Pb0.2MnO3, and (b)
La0.7Pb0.3MnO3. The solid lines are fitting curves of the isotherm to Eq. (5) at T TC.
Experimentally, to obtain these parameters, we followed the modified Arrott plots [18].
Firstly, Ms(T,0) and 0-1(T) data were determined by plotting M1/ versus (H/M)1/ from trial
exponents with = 0.365 and = 1.336 (not shown); they are the exponents of the 3D
Heisenberg model [14]. The linear extrapolation from the part of high fields of the curves to
the intercepts with the axes M1/ and (H/M)1/ would introduce the values of Ms(T) and 0-1(T),
respectively. These values for three samples around TC are shown in Figure 3. It should be
noticed that if using the set of other trial exponents (for example, the exponents of the mean-
field theory or 3D Ising model), the steps would be also carried out similarly to lead to Ms(T)
and 0-1(T) values close to those obtained by the previous trial exponents. With Ms(T) and 0-
1
(T) data obtained, when fitting the Ms(T) data of the samples to Eq. (3), we determined the
parameters TC = 161.99 K and = 0.498 for La0.9Pb0.1MnO3, TC = 191.54 K and = 0.499 for
La0.8Pb0.2MnO3, and TC = 346.54 K and = 0.502 for La0.7Pb0.3MnO3. Meanwhile, the fitting
of the 0-1(T) data to Eq. (4) gave the results of TC = 162.00 K and = 1.456 for
La0.9Pb0.1MnO3, TC = 191.31 K and = 1.241 for La0.8Pb0.2MnO3, and TC = 346.05 K and =
1.063 for La0.8Pb0.2MnO3, see Figure 3. Here, the TC values determined by the above ways are
quite close to each other. Having used the Widom scaling relation of =1+/ [21], it was
deduced to be 3.92, 3.49, and 3.12 for La0.9Pb0.1MnO3, La0.8Pb0.2MnO3, and
La0.7Pb0.3MnO3, respectively. On the other hand, if considering the M-H curves at
temperatures in the vicinity of TC and fitting the critical isotherm M(TC,H) data to Eq. (5), we
could also obtained the values of the x = 0.1, 0.2 and 0.3 samples to be 3.90, 3.17 and 3.33,
respectively, as shown representatively in Figure 4 for the x = 0.2 and 0.3 samples. One can
see that these values are in good accordance with those determined from the Widom law. All
the critical parameters obtained for the samples are shown in Table 1. For the following
performance and discussion, we shall use average values of TC taken from those obtained by
fitting the Ms(T) and 0-1(T) data for the samples, as displayed in Table 1.
With the critical parameters of TC, and , we have performed the scaling plots of
ln(M/||) versus ln(H/||+) for the samples at temperatures around their phase transition TC.
The results of this performance reveal that the M-H data fall on two branches, one for
temperatures T<TC and the other for T>TC, as can be seen in Figure 5. This is in good
agreement with the descriptions of Eq. (6) as detailed above, revealing the reliability of the
critical parameters obtained in our case when using the modified Arrott plots. The successful
determination of the critical parameters based on the critical equations, i.e., Eqs. (3-5), proves
an existence of the second-order phase transition in our samples of La1-xPbxMnO3 (x = 0.1-
0.3), in accordance with the preliminary judgment based on the Banerjee criterion.
From the results displayed above, one can see that with increasing the Pb-doping
concentration in La1-xPbxMnO3, TC increases from about 162.0 K (for x = 0.1) to 346.3 K (for
x = 0.3), see Table 1. This can be explained by means of changes in their structural
parameters, particularly the average length bond Mn-O and angle bond Mn-O-Mn. Chau et
al. [15] found that Mn-O was decreased with increasing the Pb-doping concentration, which
is 1.968, 1.966 and 1.962 for x = 0.1, 0.2 and 0.3, respectively. And, Mn-O-Mn was
slightly increased from 162o (for x = 0.1) to 166o (for x = 0.3). Under such circumstances, the
magnetic structure of the x = 0.2 and 0.3 samples becomes more stable than that of the x = 0.1
sample versus the thermal energy [10]. As a result, TC increases with increasing the Pb
concentration in La1-xPbxMnO3. However, the Pd doping does not much influence the
exponent 0.5 of three samples. This proves magnetic interaction of our samples in the
ferromagnetic regime can be described by the mean-field theory with = 0.5 [12]. Having
reviewed earlier reports on perovskite manganites and cobaltites as well, there was only one
compound of La0.8Sr0.2MnO3 having = 0.50 [5], which is very close to our material system.
Most perovskite compounds have = 0.408 - 0.46 [16, 22, 23] lying between the mean-field
theory and 3D Heidenberg model, or = 0.344 - 0.374 [1, 2, 6, 24-26] that is very close to the
value of the 3D Heidenberg model ( = 0.365), see Table 1. Few other compounds [3, 9, 27]
displayed in Table 1 have close to the 3D Ising theory ( = 0.325) or tricritical mean-field
theory ( = 0.25). For La0.7Ca0.3MnO3, which usually exhibits the first-order transition, the
using of the asymptotic relation of Eq. (1) for fitting would give = 0.14 that is not close to
any above models [1].
Figure 5. Scaling plots of ln(M/||) versus ln(H/||+) below and above TC for (a) La0.8Pb0.2MnO3, and
(b) La0.7Pb0.3MnO3 using the critical exponents listed in Table 1.
Regarding the mean-field theory, its main idea is to consider all magnetic interactions in
a multi-body system as an effective interaction of one body system [21]. This can happen
when our samples exhibit the long-range ferromagnetic order. However, to give the better
conclusion of what type of magnetic order existing in our samples, it is necessary to take into
account for the exponent characteristic for the paramagnetic region. As shown in Table 1,
the x = 0.1 and 0.2 samples have = 1.456 and 1.241, respectively, while that of the x = 0.3
sample is 1.063. The values of two former samples are close to that of the 3D Ising model (
= 1.241) with a short-range ferromagnetic order. This means that ferromagnetic interactions
between Mn3+ and Mn4+ ions still persist in La0.9Pb0.1MnO3 and La0.8Pb0.2MnO3 at
temperatures above TC, which could be from ferromagnetic clusters, as observed by electron
spin resonance [28]. However, a similar result is absent in La0.7Pb0.3MnO3, where = 1.063 is
very close to that of the mean-field theory with = 1.0 (see Table 1). Together with the above
data, it is made sure that the long-range ferromagnetic order is responsible for the magnetic
phase transition in La0.7Pb0.3MnO3, similar to the case of La0.8Sr0.2MnO3 [5]. Meanwhile, the
other compositions La0.9Pb0.1MnO3 and La0.8Pb0.2MnO3 exhibit the short-range ferromagnetic
order, meaning an existence of a ferromagnetic phase above their TC. Clearly, with increasing
the Pb concentration in La1-xPbxMnO3, besides an enhancement in the phase-transition
temperature TC, the ferromagnetic interaction also varies from the short-to-long range
interaction, with a threshold value of x = 0.3. This variation is characterized by the set of the
critical exponents for each compound. The reason for the changes in the ferromagnetic orders
can be explained as follows: it has been known that the antiferromagnetic Mn3+-Mn3+
superexchange interaction plays an important role in the parent compound LaMnO3. The Pb
substitution for La leads to an appearance of Mn4+, and thus introduces the ferromagnetic
double-exchange interaction between Mn4+ and Mn3+ ions. At a low Pb-doping concentration
with x = 0.1, the ferromagnetic Mn3+-Mn4+ interaction could exist in ferromagnetic clusters
surrounded by an antiferromagnetic host lattice due to the dominant interactions of Mn3+-
Mn3+ and Mn4+-Mn4+. Here, a compound exhibits the feature of ferromagnetic clusters would
be considered as a short-range ferromagnetic order. With increasing the Pb-doping
concentration to x = 0.2, the area of these ferromagnetic clusters would be enlarged, which
competes with the antiferromagnetic phase. However, this ferromagnetic phase is not strong
enough to be representative of the whole the x = 0.2 compound. As a result, the exponent of
the x = 0.2 sample still stays round the value of the 3D Ising model with the short-range
ferromagnetic interaction. If further doping Pb into La1-xPbxMnO3 with x = 0.3, this introduces
more Mn3+ concentration and enhance the ferromagnetic Mn3+-Mn4+ interaction that dilutes
strongly the antiferromagnetic interactions existing in La0.7Pb0.3MnO3. In this case, the
ferromagnetic clusters could coalesce to form a uniform ferromagnetic phase dominant in the
whole compound La0.7Pb0.3MnO3. Consequently, its critical exponents and are close to
those expected by the mean-field theory, and the compound exhibits the long-range
ferromagnetic order.
4. CONCLUSION
We studied in detail the influence of Pb-doping on critical behavior of La1-xPbxMnO3 (x =
0.1, 0.2 and 0.3) in the vicinity of their TC, based on analyzing M-H data. It was revealed that
both samples exhibited the second-order phase transition, where the critical parameters were
determined according to the modified Arrott method. Basically, the increase of the Pb-doping
concentration in La1-xPbxMnO3 led to the enhancement in TC, and also caused the remarkable
change in the critical exponents , and of the samples. We found that the short-range
ferromagnetic order existing in the x = 0.1 and 0.2 samples, while the long-range
ferromagnetic order is present in the x = 0.3 sample.
ACKNOWLEDGMENTS
Authors would like to thank Dr. S. G. Min for performance of magnetic measurements.
This work was supported by the research grant of the Chungbuk National University in 2009.
REFERENCES
[1] Ghosh, K.; Lobb, C. J.; Greene, R. L.; Karabashev, S. G.; Shulyatev, D. A.; Arsenov,
A. A., and Mukovskii, Y. Phys. Rev. Lett. 1998, vol 81, pp 4740-4743.
[2] Hong, C. S.; Kim, W. S., and Hur, N. H. Phys. Rev. B 2001, vol 63, pp 092504.
[3] Kim, D.; Revaz, B.; Zink, B. L.; Hellman, F.; Rhyne, J. J., and Mitchell, J. F. Phys.
Rev. Lett. 2002, vol 89, pp 227202.
[4] Mira, J.; Rivsa, J.; Rivadulla, F.; Vazquez, C. V., and Quintela, M. A. L. Phys. Rev. B
1999, vol 60, pp 2998-3001.
[5] Mohan, C. V.; Seeger, M.; Kronmuller, H.; Murugaraj, P., and Maier, J. J. Magn.
Magn. Mater. 1998, vol 183, pp 348-355.
[6] Nair, S.; Banerjee, A.; Narlikar, A. V.; Prabhakaran, D., and Boothroyd, A. T.
arXiv:cond-mat/0305598v1 [cond-mat.str-el] 2003.
[7] Shin, H. S.; Lee, J. E.; Nam, Y. S.; Ju, H. L., and Park, C. W. Solid State Commun.
2001, vol 118, pp 377-380.
[8] Yang, J.; Lee, Y. P., and Li, Y. J. Appl. Phys. 2007, vol 102, pp 0333913.
[9] Yang, J.; Lee, Y. P., and Li, Y. Phys. Rev. B 2007, vol 76, pp 054442.
[10] Phan, T. L.; Min, S. G.; Phan, M. H.; Ha, N. D.; Chau, N., and Yu, S. C. Phys. Stat.
Sol. (b) 2007, vol 244, pp 1109 - 1117.
[11] Zener, C. Phys. Rev. B 1951, vol 82, pp 403-405.
[12] Millis, A. J.; Shraiman, B. I., and Mueller, R. Phys. Rev. Lett. 1996, vol 77, pp 175.
[13] Souza, J. A.; Neumeier, J. J.; Bollinger, R. K.; Mcguire, B.; Dos Santos, C. A. M., and
Terashita, H. Phys. Rev. B 2007, vol 76, pp 024407.
[14] Stanley, H. E. Introduction to phase transitions and critical phenomena; Oxford
University Press, London, 197.
[15] Chau, N.; Nhat, H. N.; Luong, N. H.; Minh, D. L.; Tho, N. D., and Chau, N. N. Physica
B 2003, vol 327, pp 270-278.
[16] Luo, X.; Sun, Y. P.; Wang, B.; Zhu, X. B.; Song, W. H.; Yang, Z. R., and Dai, J. M.
Solid State Commun. 2009, vol 149, pp 810-813.
[17] Banerjee, S. K. Phys. Lett. 1964, vol 12, pp 16-19.
[18] Arrott, A. and Noakes, J. E. Phys. Rev. Lett. 1967, vol 19, pp 786-789.
[19] Arrott, A. Phys. Rev. 1957, vol 108, pp 1394-1396.
[20] Potter, C. D.; Swiatek, M.; Bader, S. D.; Argyriou, D. N.; Mitchell, J. F.; Miller, D. J.;
Hinks, D. G., and Jorgensen, J. D. Phys. Rev. B 1998, vol 57, pp 72-75.
[21] Huang, K. Statistical Mechanics; Wiley, New York, 1987, 2nd Edition.
[22] Freitas, R. S.; Haetinger, C.; Pureur, P.; Alonso, J. A., and Ghivelder, L. J. Magn.
Magn. Mater. 2001, vol 226-230, pp 569-571.
[23] Mira, J.; Rivas, J.; Vazquez, M.; Beneytez, J. M. G.; Arcas, J.; Sanchez, R. D., and
Rodriguez, M. A. S. Phys. Rev. B 1999, vol 59, pp 123-126.
[24] Moutis, N.; Panagiotopoulos, I.; Pissas, M., and Niarchos, D. Phys. Rev. B 1999, vol
59, pp 1129 - 1133.
[25] Padmanabhan, B.; Bhat, H. L.; Elizabeth, S.; Rszler, S.; Rszler, U. K.; Drr, K., and
Mller, K. H. Phys. Rev. B 2007, vol 75, pp 024419.
[26] Sahana, M.; Rossler, U. K.; Ghosh, N.; Elizabeth, S.; Bhat, H. L.; Dorr, K.; Eckert, D.;
Wolf, M., and Muller, K. H. Phys. Rev. B 2003, vol 68, pp 144408.
[27] Mukherjee, S.; Raychaudhuri, P., and Nigam, A. K. Phys. Rev. B 2000, vol 61, pp
8651-8653.
[28] Phan, T. L.; Min, S. G.; Yu, S. C., and Oh, S. K. J. Magn. Magn. Mater. 2006, vol 304,
pp e778-e780.
Chapter 15
REDDISH CERAMIC PIGMENTS WITH

PEROVSKITE STRUCTURE
C. P. Pia
Divisin de Estudios de Posgrado, Facultad de Qumica, Mxico, D. F.
ABSTRACT
In the past (1980), the ceramic pigments, 51 pigments in all, were classified into 14
structures of industrial interest and these were classified into black, white and colored
pigments. These structures are oxides, materials used in ceramic as pigments because of
their greater stability in oxygen-containing ceramic systems. However, thanks to modern
and more refined methods of scientific investigation, new structures, new pigments, new
synthetic methods and new applications were also included. Among these we have the
development of new red ceramic pigments as perovskite structure of formula A3+B3+O3 as
ferrites LnFeO3 and chromium aluminates LnAl1-xCrxO3 (Ln, lanthanide). These
lanthanides aluminates ceramic pigments have been synthesized by different methods and
characterization studies were realized by X ray diffraction, SEM, UV-VIS spectra, etc.
The properties and uses of these lanthanides are appropriate as ceramic pigments;
nevertheless, new possibilities for these pigments are its use for luminescent coatings in a
diversity of devices or lasers.
1. INTRODUCTION
Perovskite materials have many interesting and spectacular properties such as high
temperature superconductors, magneto resistance, ferroelectricity, charge ordering, spin
dependent transport, high thermo power, optical properties, etc. As well as their applications
in catalyst electrodes in certain types of fuel cells, memory devices, etc.
In the case of optical properties, nowadays the use of perovskites materials is very
important in ceramic pigments overcoat in the reddish range. The reddish pigments must have
suitable properties of high thermal stability at high temperature and chemical stability but are
540 C. P. Pia
restricted to zircon structure of hematite or cadmium sulfoselenide and malayaite (type

sphene). In the last years ceramic pigments of cerianite solid solutions and ferrites of
lanthanides type perovskites have been synthesized. Moreover in the case of ferrites, type
perovskites LnFeO3 (Ln= La, Ho, Nd, Er and Sm) show a reddish dark color mainly in the
presence of iron in the chemical formula and these ferrites are mostly used for their magnetic
properties.
In the reddish perovskite LnAlO3 (Ln = lanthanides) lattice the color is obtained
substituting Al3+ for Cr3+ into AlO6 octahedral of perovskite lattice forming solid solutions
resulting in an orthorhombic or rhombohedra distortion. These pigments are very interesting
because adequate tinting strength and purity of color can be obtained. These materials are
oxides showing greater stability in oxygen containing ceramic systems and are resistant to
high temperature and corrosive environment in the firing process. The specific properties of
inorganic pigments result from an interplay of solid state properties, particle size and particle
shape.
2. CERAMIC PIGMENTS CLASSIFICATION

Most of the materials used in ceramic as pigments are oxides because of their great
stability in oxygen-containing ceramic systems. In the past, ceramic pigments have been
classified as colorants into: black, white and colored pigments because pigments depend for
its color and its ability to reflect or absorb light. There are classifications on the majority of
oxide materials, which can be used as inorganic pigments according to their color. Black
color pigments absorb all wavelengths of visible range, white color pigments show no
absorption in the visible range and colored pigments absorb certain wavelengths of visible
light. In the electromagnetic spectrum the visible light is placed between the infrared and
ultraviolet in the range of 400 to 700 nm. [1, 2]
The oxides pigments were classified in 1982 by the Dry Color Manufacturers Association
DCMA system based in a structural approach. These structures are classified in 14 categories
and each category has a number of determined pigments, 51 in total, and each pigment has a
classification number. Perovskite structure is not included in this classification.
The structures must have several principles, such as the nearest neighbor cation-anion
distances determine the lattice energy of an ionic solid material at room temperature. The
preferred coordination polyhedral of the anions surrounding each cation in ionic solid phases
is determined usually by the radio of the ionic size of the cation to that of the anion. The
structures of any combination of cations and anions are subject to the rules of electrostatic
neutrality, meaning that the ions replacing each other must have the same charge and must be
fairly similar in size, with a difference of a maximum of 15% in the radii of the ionic
elements. This allows a wide range of cation substitution, in particular the case of transition
elements or lanthanides; this may be done if the ionic sizes of the substituted ions are similar.
In this way the color obtained can vary. Each structure can take doped elements, may be
incorporated by the host crystal lattice and in this way alter the color or other physical
properties of the pigment without changing its structure [2-5]
Among the reddish pigments included in the DCMA classification we can find:
Reddish Ceramic Pigments with Perovskite Structure 541
Category III. Corundum-Hematite Structure, (Al2O3-Fe2O3)
This category contains 4 pigments: chrome alumina pink corundum CrAlO3 (3-03-5
DCMA number), manganese alumina pink corundum MnAlO3 (3-04-5 DMCA number),
Hematite -Fe2O3, Cr2O3 and Mn2O3 have the same corundum structure Al2O3 with
approximately hexagonal close packed array of oxygen in which Al3+ ions occupy two thirds
of the octahedral holes. The structure is complex because there is a sharing of vertices, edges
and each octahedron FeO6, CrO6, MnO6 or AlO6, shares a face with another in the layer above
or below. All these pigments have the same ABO3 perovskite stochiometry with radium size
0.6 < r A r B < 0.9 . When corundum (Al2O3) has 0.1% of chrome oxide is a ruby natural
gemstone. [2, 5]
Category XI. Rutile Cassiterite TiSnO4
ABO4 stoichiometry, contains 12 pigments Chrome tin orchid cassiterite (11-23-5

DCMA number). It is prepared with equimolecular quantities of titanium and TiO2 and SnO2
chromium doped and heated at high temperatures. [2, 4-6]
Category XII. Sphene: CaSnSiO5
In this category there is one pigment only: tin pink sphene chromium doped CaSnSiO5
(12-25-5 DCMA number). It is used as ceramic pigment, was prepared with calcium oxide,
tin oxide and silica doped with chromium oxide at 1200C for 24 hours on the formation of
malayaite or calcium tin silicate CaSnSiO5. The color of this pigment depends on the ratio of
the concentration of chromium oxide to that of tin oxide. Generally when the ratio is 1:15 the
resulting color is purple, 1:17-20, is red or maroon, and 1:25 is pink. The valences of Cr
doped in sphene type (CaTiSiO5-CaSnSiO5), were directly related to the color of the pigments
and were dependent on the Ti/Sn ratio in the mother crystals. In CaSnSiO5, Cr4+ is detected
causing a reddish purple color in 2mol% Cr doped. When it is Cr3+, the color changes to red
brown and finally to brown.[2, 6-9]
Category XIII. Spinel Structure AB2O4
Contains 19 pigments Chrome alumina pink spinel CrAl2O4, (13-32-5 DCMA number).
The color is explained by the ligand field theory of the ion oxygen tetrahedral of the spinel
lattice. This ligand field splits the originally equivalent d levels of the ion chromium and
allows transitions between the split energy levels as a result of light absorption in the visible
range. When there are transition elements in the structures as: V, Cr, Mn, Fe, Co, Ni, or Cu
forming the structures or as doped the compounds have color.
The synthesis of spinel structure is realized by solid state with the oxides of Al3+ and Cr3+
at 1200C for 48 hours. [2, 3-5]
542 C. P. Pia
Category XIV. Zircon Pink and Coral Colors ABO4
Contains 3 pigments. The zircon, orthosilicate of zirconium contains SiO4- ions, the
cations of Zr4+, occupying interstices which are surrounded by a number of eight O2- ions
number appropriate to their size. The iron zircon pink ZrSiO4-Fe, (14-44-5 DCMA number),
was patented in 1961 and made with zirconium oxide, silica, iron oxide and used a mixture of
sodium fluoride, sodium chloride and sodium nitrate as a mineralize. The calcinations
temperature was realized at 880-1000C. This pigment has been studied by several authors
but the formation of iron zircon pink pigment synthesized by ceramic method with 2% weight
LiF as a mineralize indicates the formation of a substitution solid solution with about 1.5
mol% irons. In this compound the shades extend from coral to pink. The iron zircon pinks are
more difficult to control and develop than other zircon colors, vanadium zircon blue, (14-42-2
DCMA number) and praseodymium zircon yellow (14-42-4 DCMA number). [2, 6-12, 14]
Encapsulation in Zircon ZrSiO4
The last red pigment based on zircon lattice was the inclusion of red cadmium
sulfoselenide Cd (S Se) crystals. The cadmium sulfoselenide pigments are not of an oxide
system but orange, red and dark red colors are obtained. The shades result from varying the
ratio of the concentration of cadmium sulfide to that of cadmium selenide. Due to the fact that
these pigments cannot be used in higher temperature applications its pigments are
incorporated in a zircon lattice during manufacture, imparting stability of zircon to the
pigment. That means the pigment is encapsulated. [13]
Hematite - Fe2O3
It is important to indicate that between natural inorganic pigments, hematite - Fe2O3 red
is included. In ancient times the chemical analysis of caves paintings indicates that hematite
among the iron oxides, was used as pigment. This explains the use of these and other minerals
to create cave paintings, decoration of ceramic objects for the preservation of the history of
human existence. [2, 14]
3. LANTHANIDES
3.1. Perovskite Structure
The perovskite structure is any material with the same type of crystal structure as
CaTiO3. It was discovered in the Ural mountain of Russia by Gustav Rose in 1839 and is
named after the Russian mineralogist, L. A. Perovski (1792-1856).
Perovskites oxides1 have the general formula ABO3. This structure is reported as pigment
[15] but is not included in the DCMA system. Their idealized structure is cubic, it is found in
only a limited number of compounds. Each A of the formula, corresponds to an ion of large
size and is 12-coordinated by oxygen and each B of the formula corresponds to an ion of
small size and is 6-coordinated by oxygen. Each oxygen is linked to four A ions and two B
ions. The large A ions are comparable in size with oxygen ions forming a close packing array
and B small ions have a radius appropriate to octahedral coordination. See figure.
Ideally, there is a relationship between the radii of the ions. That is:
(1) rA + rO = t ( rB + rO )
(2) t = rA + rO / ( rB + rO )
There are a few cases of the cubic ideal structure: t is a Goldschmidt tolerance factor and
in the cubic ideal perovskite must be 1, but the approximated limits are 0.7 1.0. For oxides
the radii of the A and B ions must lie within the ranges 1.0-1.4 and 0.45-0.75, In the
lanthanides the radii trivalent found in the range from Cerium 1.07 to Samarium1.00. The
rest of radii trivalent lanthanides are in the range from Europium 0.98 to Lutetium 0.85 ,
all radii comparable in size with calcium ion radii 1.00, thus the lanthanides correspond to
A ion of large size. In aluminates and ferrites the Al3+ and Fe3+ ions are of small size,
corresponding to B ion, having 0.51 and 0.64, respectively. This means that in the
perovskites oxides A and B ions are between the indicated ranges according to the tolerance
factor.
The most common perovskites have distortions derived from small displacements of the
atoms. In perovskites structures there are distortions in orthorhombic and rhombohedra (Hex)
symmetries; in particular there are rare distorted symmetries in the tetragonal, monoclinic and
triclinic perovskite structures. In addition the number of these structures is very large and
1
There are perovskite structures of: halides, hydrides, carbides, nitrides.
544 C. P. Pia
some compounds show polymorphism capable of existing in two or more forms with different
crystal structures. [3-5]
It is important to remark that among oxides of perovskite structure, A and B ions besides
having appropriate radii must confer electrical neutrality, that means that pairs of cations must
have 1+ and 5+ (KNbO3), 2+ and 4+ (CaCeO3) and 3+ and 3+ (LaFeO3) oxidation states. The
last oxides correspond to lanthanides aluminates and lanthanides ferrites.
In some cases the perovskite structure is more complicated due to the fact that some of
the unoccupied A sites or to the fact that there are non-stoichiometry (ABO3-). [16]
The perovskite structure is also found in a number of oxides in which A and or B sites are
not all occupied by atoms of the same kind; vg. AABO3 (K0.5La 0.5TiO3), ABBO3
(SrGa05Nb0.5O3), AABBO3 (BaK)(TiNb)O3. These compounds are called mixed oxides.
There have been reported double perovskite lanthanides structure 2(ABO3) or (A2B2O6)
and A2BBO6 in which two kinds of ions are found in B sites in alternating octahedral
Ba2LnSbO6 and Sr2LnSbO6 (Ln = Ho, Dy, Gd). [17] Other oxides consisting of a single
cation in different oxidation states are classified as complex oxides.
Due to the diversity of this complex oxides resulting from a thorough selection of the A
and B ions, their fractions, ionic sizes, valences, spin states and orbital ordering, as well as the
oxygen content and vacancy ordering, the distortions of the ideal cubic symmetry. These
compounds show a large number of different properties: electronic, magnetic, ferroelectric,
thermoelectric, optics mixed conducting. The best known examples are triple perovskites as
the spectacular 123 high temperature superconducting cuprates as LnBa2Cu3O7- . These
present Cu ions with different oxidation states and oxygen vacancies. [16, 18, 26]
3.2. Optical Properties
3.2.1. Reddish Perovskites

Colorants may be pigments or dyes, a colored pigment is opaque and insoluble a dye is
colored and transparent and often soluble. Pigments are solid particles with solid state
properties resulting from structure, chemical composition, the color, stability as heat, light
and insolubility in solvents and binders, etc. For their chemical composition the pigments
may be divided in two categories, organic and inorganic. In ceramic usually concerning to
inorganic pigments. The particle size, distribution and shape including the hue and the light
scattering are very important. If a pigment is red, it reflects radiation of 630-700 nm and
absorbs the rest.
The need to developed ceramic pigments with great stability resulted in a thorough
research during the 1960s. Some lanthanides ferrites were reported for their magnetic
properties and not as pigments [19], for this reason the zircon have gained increasing
acceptance in the industry for their resistance under high temperatures and corrosive
atmosphere during firing. The zircon (Zr Si O4) was discovered and patented in 1948 as
ceramic pigment. In the case of red zircon iron there is a solid solution formation. [10]
A number of new high temperature pigments used in ceramic has led the search for the
perovskite pigments of lanthanides. This system has been first developed in 1984-1986 in the
red perovskite pigments lanthanides and recently in 2000 red, reddish and pink color
pigments with perovskite structure of LnAlO3 (Ln= lanthanide) chromium doped have been
reported. [20]
Chromium doped yttrium aluminates perovskite type have been reported. In 2006s red
perovskite ceramic pigments have been developed efficiently in different ceramic
applications. [21]
In 2005 and 2008 lanthanides ferrites perovskite structure LnFeO3 (Ln = La, Ho) were
synthesized by precursor and ceramic methods and reported as reddish ceramic pigments. [22,
23]
In 2000s Jansen and Letschert demonstrated that solid solutions of the perovskites of
oxy-nitrides: CaTaO2N and LaTaON2 constitute promising pigments in the yellow-red range
without toxic metals. [25]
The preparation of these perovskites system was oriented in the study of other optical
properties such as: red photoluminescence emitted by crystalline NdAlO3 or
photoluminescence of homogeneous dispersion of NdAlO3 nanocrystals in an Al2O3 matrix
synthesized by sol gel process. Fast ultraviolet emission was observed in NdAlO3 and
SmAlO3 and strong photoluminescence was observed in LaAlO3 and YAlO3 crystals in visible
and ultraviolet wavelength regions. [26-28]
The combination of luminescence and electrical conductivity in perovskite oxides make
them promising materials for optical, electronic and structural application, for novel
optoelectronic devices, window materials and a variety of lamps, for solid state lasers with
fiber optics. In communication silica fluoride doped Eu3+ up to date is very important
commercially and has been exploited. NdAlO3 films are useful insulating dielectric for
microelectronic applications. Superconductors layers grow by sputtering over neodymium
aluminates layers, they are also used for the substrates for the epitaxial growth of various
perovskite type films such as ferroelectrics, piezoelectric and magneto resistance oxides.
Other perovskites were synthesized as catalysts precursors and are being incorporated into
automobile catalytic converters. The perovskite GdAlO3 is a material for the neutron
absorption and control rod applications [26, 28, 29, 32]
3.2.2. Fluorite Structure

The fluorite structure of the red pigment of praseodymium doped cerianite CeO2 was
known in 1960, the cerianite fluorite structure was patented in France and published in the
years 1900-2000 as cerianite doped with Pr4+ as red ceramic pigment. [30]
Recently, in 2001 cerianite doped with lanthanides has been reported because of the low
toxicity red pigments, as well as Pr with Tb, Eu, Pr-Mo (reddish color), Pr-Fe, (orange color)
average size of 300nm and also doped with chromium or indium. [10]
Cerianite doped Pr is synthesized at high temperatures 1300-1500 C for 15-48 hours to
obtain the red color but when it is applied on ceramic the color disappears. [31]
In 1971 rare earths were used in fluorescent lamps. Red-Y2O3-Eu+3, blue-BaMg2Al16O27-
Eu2+, green-(Ce-Tb)MgAll1O19 are deposited on lamps walls as fine particles. The research on
photoluminescence materials doped lanthanides is actually an area of great demand.
546 C. P. Pia
4. METHODS USED IN PEROVSKITE SYNTHESIS

The scientific interest on the pigments used in the ceramic industry overcoat in reddish
ceramic pigments has developed a series of possibilities in new systems of preparation of
LnFeO3 and LnAlO3 perovskite lanthanides systems with a large and spectacular application.
The selected synthesis depends on the knowledge of: reactivity of solids, crystalline
chemistry and the necessary tools as reactants, material laboratory, equipment and the
objective in the preparation of known compounds with specific properties, unknown or new
compounds.
4.1. Conventional Method
In solid state chemistry the conventional method is denominated ceramic method where
solids are mixed continuously, in general oxides. These were submitted afterwards to various
thermal processes, generally at high temperatures in ranges 800C, 1000C, 1300C, 1400C,
1500C. The iron zircon pink was patented in 1961 and made with zirconium oxide, silica,
iron oxide and a mixture of sodium fluoride, sodium chloride and sodium nitrate was used as
a mineralize. The calcinations temperature was realized at 880-1000C. The iron zircon pinks
are more difficult to control and develop than other zircon colors. The iron zircon pigments
provide pink and coral colors. This pigment has been studied by several authors but the
formation of iron zircon pink pigment synthesized by the ceramic method with 2% weight
LiF as a mineralize, indicate the formation of substitution solid solution with about 1.5mol%
iron. [3, 3b, 4, 8] without to forget: tri perovskites (superconductors 123) [18], layered
perovskites with titanium and lanthanides [4].
Nowadays lanthanum ferrites and chromium doped lanthanide aluminates with
perovskite structure are used as ceramic pigments because they have colorimetric properties,
high thermal stability, fine particle size and chemical resistance. Also, the ferrites have
interesting magnetic properties and the aluminates are materials having photoluminescence
properties.
4.2. Precursor Method
This method has been realized for synthesis of perovskite solids. Precursor compounds
with anions as oxalates, acetates, cyanides, nitrates, hydroxides have been employed as
precursors for the perovskite synthesis. Examples are: LaCoO3 from LaCo(CN)6.5H2O and
LaNiO3 and NdNiO3 from tri hydroxide of lanthanide in both examples can be a substitution
of cations (lanthanides or transition elements), LnFeO 3, (Ln = Er, Sm, Nd). Sometimes in this
method, it is necessary to find a suitable precursor compound for the synthesis of the desired
composition because it is limited by the stoichiometric ratio of cations in the precursor and in
the final oxide. [32, 33, 44]
4.3. Co Precipitation Method
This method is a precipitation together with two or more different compounds. For the
preparation of perovskites, co precipitation as hydroxides, carbonates, oxalates, citric acid
have been used. The precipitating agents were KOH, NH4OH for hydroxides, K2CO3,
NH4CO3 for carbonates, oxalic acid H2C2O4 or ammonium oxalate (NH4)2C2O4 for the co-
precipitation of oxalates. The carbonates and oxalates required higher temperatures than the
hydroxides. The presence of alkaline ions could contaminate the product and must be
eliminated. The presence of ion ammonium could form ammonium complexes with transition
ions. Citric acid was employed for preparing LaCrO3, LnAlO3 (Ln = Y, La Sm) and LaMO3
(M = V, Cr, Mn, Fe, Co Ni)
4.4. Sol Gel Synthesis
This method is useful in the preparation of perovskite type oxides and other complex
oxides as they allow to obtain pure phase products and to control their stoichiometry.
Neodymium aluminates NdAlO3 and gadolinium strontium aluminates Gd1-XSrXAlO3 were
synthesized [26, 27, 34-36]
4.5. Ion Exchange Method
Reported by proton exchange of LiNbO3 to give HNdO3 where a structural

transformation of the rhombohedra LiNbO3 structure to the cubic HNdO3 structure occurs.
[37]
4.6. Spray and Freeze Method
Procedure of spray and freeze techniques for the preparation of Cr-doped sphene
pigments from an aqueous solution of precursors salts followed by calcinations. Chromium
doped CaSnSiO5 pigments, developed pink hues. In Cr-doped CaTiSiO5 have brown
colorations due to the combination of Cr (III) and Cr (IV) cations dissolved in the sphene
matrix. [38]
4.7. Alkali Flux Method
This method is employed for the synthesis of solids using alkaline compounds as solid
fluxes or in aqueous alkali solutions. Perovskites of KxBa1-xBiO3 and spinels of lanthanides
have been synthesized.
548 C. P. Pia
4.8. Nano Explosion Synthesis
This method was used for the preparation of nanoparticles with high surface energy and
chemical activity. In this procedure sol gel process, coprecipitation, micellar nanoreactor,
hydrothermal synthesis, sonochemically and/or microwave and combustion synthesis have
been included [39]
4.9. Combustion Synthesis
This method was used for the preparation of NdAlO3. When the sample is irradiated by
unfocused ultraviolet pulsed laser light red light visible is emitted. [27]
5. STUDY OF FERRITES AND CHROMIUM DOPED ALUMINATES

REDDISH CERAMIC PIGMENTS WITH PEROVSKITE STRUCTURE
The advances in most synthesis of lanthanides new solids have been realized for the
ceramic method and the precursor method; for this reason these methods remain very
important for solid state chemistry. Examples of studies realized using these methods are
ferrites and lanthanides aluminates perovskites presented here below.
1. Synthesis
1.1. Ceramic Method of Ferrites: LnFeO3 (Ln = La, Ho)

The raw materials used in the synthesis of LnFeO3 (Ln = La, Ho) were Ln2O3 oxides
Aldrich and Fe2O3 Baker R.A. This ferrites were prepared mixing the starting materials in
stoichiometric relations 1:1 in an agate mortar, after were fired in an electric furnace at
900C, 950C by 20h each one and 1000C, in the case of lanthanum ferrite by 24h and in the
case of holmium ferrite by 30h. LnFeO3 (Ln = Nd, Sm, Er) were synthesized by the precursor
method. (See forward: Precursor method)
1.2. Precursor Method of Ferrites: LnFeO3 (Ln = Nd, Sm, Er)

The LnFeO3 (Ln = Nd, Sm, Er) were synthesized using Ln (NO3)3.H2On, Rhone Poulenc.
They were heated to dryness in a stove at 110C; once dried they were weighed and dissolved
in deionized water and gauged to 100ml. Aliquots of each one were taken and the
stoichiometric amounts of FeCl3.6H2O R.A. Baker, were added in a 1:1 ratio each of them.
These samples were put in a platinum crucible and heated to complete dryness in an electric
plate. The dry samples were fired in an electric furnace at 800C, 1000C and finally at
1300C for 12 hours, every time with intermittent grinding in an agate mortar in each of the
thermal processes.
1.3. Ceramic Method of Chromium Doped Lanthanide Aluminates: Ln Al0.93Cr0.07O3

(Ln = Nd, Dy, Ho, Er)
The raw materials used in the synthesis of Ln Al0.93Cr0.07O3 (Ln = Nd, Dy, Ho, Er) were:
Nd2O3 and Er2O3, Rhone Poulenc, Dy2O3 and Ho2O3 Apl. Eng. Materials Inc.,
Al(NO3)3.9H2O E. Merck, Cr2O3 J.T.Baker and LiF Aldrich
The lanthanides perovskites have been prepared by mixing the starting materials in
relations according to the chemical formula LnAl0.93Cr0.07O3 and 5% LiF as mineralize, in an
agate mortar. These samples were put in the platinum crucibles and heated to complete
dryness in an electric plate, afterwards were fired in an electric furnace at 900C, 1000C by
24 h each one and at 1300C for 48 h with grinding intermitting.
2. Characterization
The characterization of ferrites and aluminates was carried out by X-ray diffraction
powders in a Siemens D5000 diffractometer using Cu, K radiation ( = 1.5406) and a Ni
filter in a 5 (2) 85 range. Details of the microstructure and morphology were done by a
(SEM) scanning electron microscope JEOL model JSM-5900LV. Energy Dispersive X-ray
Analysis (EDX) shows the microanalysis of the elements in the samples. The UV-Vis spectra
were realized in a Cary-5000 spectrophotometer.
3. Results and Discussion
The examination of X-ray diffractograms shows that the perovskites structure LnFeO3
(Ln = La, Nd, Sm, Ho, Er) and LnAl0.93Cr0.07O3 (Ln = Nd, Dy, Ho and Er) have orthorhombic
symmetry and the space group is Pbnm, which is a distorted perovskite where each iron
remains essentially octahedral. Different opinion is that perovskites with larger tolerance
factors have rhombohedra phases (Pr, Nd). In this work this occurs only in NdAl0.93Cr0.07O3
with rhombohedra (Hex) symmetry but non in neodymium ferrite and praseodymium
aluminates. The incorporation of the different lanthanides is related to the lanthanide
contraction, decrease of the ionic radium and the tolerance factor; lanthanides with smaller
tolerance factors have orthorhombic phases, this changes the symmetry lattice and the
distortion increases. See Figures 1a, 1b.
The results from the observation of SEM micrographs reveal that the morphology of the
pigments with agglomerates: such as flakes, in the case of holmium ferrite, the particles are
approximately from 0.5 m to 1.0 m in size. Neodymium, samarium and erbium ferrites are
agglomerates like globules of approximately 1.0 m, and in the case of erbium ferrite
approximately 2.0 m. See Figure 2a.
The chromium doped lanthanide aluminates micrographs of Pr, Nd, Dy, Ho, and Er by
SEM show the same type of agglomerates like globules of 1.0 m, to 2.0 m in size. The
particle size, particle shape and particle morphology are very near between ferrites and
aluminates and the materials are appropriate as pigments used in ceramic materials. See figure
2b.
550 C. P. Pia
800 700
700 NdFeO3 SmFeO3

600
600
500
500
Intensidad (Cps)
Intensidad (Cps)
400
400
300
300
200
200
100
100
0 0
10 20 30 40 50 60 70 10 20 30 40 50 60 70
2 Theta 2 Theta
1400
1400
2.7730 1200 HoFeO3
LaFeO3 2.6981
1200
1000
1000
Intensidad (Cps)
Intensidad (Cps)
800
800
600
600
400
400
200
200
0
0
20 40 60 80
20 40 60 80
2 Theta
2 Theta
Figure 1a. X-ray diffraction patterns of LnFeO3 (Ln= Nd, Sm, La, Ho)
All SEM micrographs have the same morphology and similar particle sizes in the range of
1-2 m.
The reddish pigments with perovskite structure of type doped chromium lanthanides
aluminates LnAl1-XCrXO3 is when Al3+ is substituted by Cr3+ occupying octahedral sites in the
structure forming solid solution, because Cr3+ radium is 12.3% in size bigger than the size of
Al3+, (0.63) and (0.51) respectively. This makes that the Cr-O distance decreases and the
Al-O distance and the perovskite distortion increase.
The UV-VIS of chromium doped lanthanide aluminates of Pr, Nd, Dy, Ho, and Er show
that octahedral absorption in the visible range Cr+3 present two absorption bands due to the
4
A2g 4T1g in blue range and 4A2g 4T2g in yellow range transitions. Absorption bands are
displaced towards higher energies thus the two transitions forms a broad absorption green
band region and consequently there is red reflection. See Figure 3b.
The reddish color variation, in the case of ferrites LnFeO3 (Ln = La, Nd, Sm, Ho, Er)), is
very dark in Pr, in Nd is a bright red color, in the rest is only reddish. See figure 3a.
800
Intensity
600
400
200
0
200 0 50 100
2
Dy
1500
1000
500
0
0 20 40 60 80
2
Er
1500
Intensity
1000 HoAlO3
500
0
0 20 40 60 80
2
Figure 1b. X-ray diffraction patterns of LnAl0.93Cr0.07O3 (Ln= Dy, Er, Ho)
500
450
400
350
300
Intens ity
250
200
150
100
50
0
-50 0 20 40 60 80
Nd
400
350
300
250
200
150
100
50
0
50 0 20 40 60 80
Pr
Figure 1b. X-ray diffraction patterns of Ln Al0.93Cr0.07O3 (Ln = Nd, Pr)

552 C. P. Pia
HoFeO3 LaFeO3
NdFeO3 SmFeO3
Figure 2a. Microstructures of LnFeO3 (Ln = Ho, La, Nd, Pr, Sm).
Pr Al0.93Cr0.07O3 Ho Al0.93Cr0.07O3
Dy Al0.93Cr0.07O3 Er Al0.93Cr0.07O3
Figure 2b. Microstructure of LnAl0.93Cr0.07O3 (Ln = Pr, Ho, Dy, Er)

0.8
0.7
0.6
0.5 NdFeO3
Absorb.
0.4
0.3
0.2
0.1
0
0 10000 20000 30000 40000 50000
cm1
1.2
1
0.8
S SmFeO
mF eO33
0.6
0.4
Abs orb
0.2
0
0 10000 20000 30000 40000 50000
-1
c m -1
cm
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0 10000 20000 30000 40000 50000
Figure 3a. UV-VIS of LnFeO3 (Ln = Nd, Sm, Er)

554 C. P. Pia
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0 10000 20000 30000 40000 50000
1.2
0.8
0.6
0.4
0.2
0
0 10000 20000 30000 40000 50000
1
0.8
0.6
0.4
0.2
0
0 10000 20000 30000 40000 50000
Figure 3b.UV-VIS of LnAl0.93Cr0.07O3 (Ln = Nd, Ho, Dy)
In the chromium doped lanthanide aluminates Ln Al0.93Cr0.07O3 (Ln = Nd, Dy, Ho, Er),
the reddish color variation is red in Nd, pink shades lighter in color from Dy to Er.
That means that the development of the red color is related to different phenomena,
between this: the change of the Cr-O distance of the crystal field, the symmetry that increases
from orthorhombic phase (Er) to rhombohedra (Nd) and the lanthanides ions types revealed
their influence associated to the variation of the color shades. [39, 40, 41, 42, 43]
Figures 4a, 4b, shows energy dispersive X-ray spectrum of the ferrites and aluminates of
lanthanides compounds to confirm the assumptions made from the X-ray diffraction powder
results.
Figure 4a. EDX of LnFeO3 (Ln = Nd, La, Er).
Figure 4b. E DX of LnAl0.93Cr0.07O3 (Ln = Pr, Ho, Dy, Er)

556 C. P. Pia
REFERENCES
[1] Streatfield, G. R. Trans.Br.Ceram.Soc.1990, 89, No.5, 177-180.
[2] Burgyan, A.; Eppler, R.A. Am. Ceram. Soc. Bulletin. 1983, 62, No.9, 1001-1003.
[3] Evans, R. C. An Introduction to Crystal Chemistry, Cambridge, Univ. Press, London N.
W., 1966, 1676-170
[4] West, A. R. Solid State Chemistry, John Wiley & Sons, Chichester, 1985.
[5] Wells, Wells, Structural Inorganic Chemistry. Oxford University Press 1975, London,
483, 484, 486.
[6] 6. Eppler, R. A. J. of the Am. Ceram. Soc. 1976, No. 9-10, 455.
[7] Masahiro, K.; Hidero, U.; Minoru, T., J. of the Ceram. Soc. Soc. of Japan 200, 108, No.
1257, 478-481
[8] Harriet, G. F. J. of Chem. Educ. 1981, 58 No. 4, 291-294.
[9] Kato, M.; Unuma H.; and Takahashi, M., Jour. of the Ceramic Soc. of Japan, 2000,
108, No.1257, 478-481.
[10] Eppler R. A. J. of the Am. Ceram. Soc. 53, No. 8.
[11] Eppler R. A. Am. Ceram. Soc. Bull. 1977, 56, No. 2, 213-2118.
[12] Carreto, E.; Pia, C.; Arriola, H.; Barahona, C.; Nava, N.; Castao, V. J. of Radioanal.
and Nucl. Chem. 2001, 250 No. 3, 453-458.
[13] Rincn, J.Ma.; Carda, J.; Alarcn, J., Nuevos productos y tecnologas de esmaltes y
pigmentos cermicos, Faenza, Editrice Ibrica, S. L.Madrid, 1992, 171-180.
[14] Garca, A.; Llusar, M.; Sorl, S.; Calbo, J.; Ten A, M. A.; Monrs G., Tour. of the
European Ceramic Sopc. 2003, 23, 11, 1829-1838.
[15] Pia, P. C., Pigmentos cermicos, Cuadernos de Posgrado 5, Qumica Inorgnica II,
1982, 107-124.
[16] Dabrowski, B.; Kolesnik, S.; Chmaissem, O.; Suescun, L.; Mais, J., Acta Phys.
Polonica A. 2007, 111 No.1, 15-24.
[17] Karunadasa, H. ; Huang, Q.; Ueland, B.G.; Schiffer, P.; Cava, R.J.PNAS 2003, 100
No.14, 8097-8102.
[18] Bednorz, J. G. and Muller, K. A. ZPhys B. Condens. Matter, 1986, 64, 189.
[19] Eibschtz, M.; Shtrikman, S. Treces, D., Physical Review 1967, 156, No.2, 562-577.
[20] Kato, M.; Takahashi, M., 2004, J. of Material Science Letters 2004, 20 No.5, 413-414.
[21] Matteucci, F.; Lepri, N.; Dondi, M.; Cruciani, G.; Baldi, G. Boschi, A. O.,Advances in
applies ceramics 2006, 105 No.2, 99-106.
[22] Cunha, M.; Melo, D.; Martinelli, A.; Melo. M.; Maia, I; Cunha, S., Dyes Pigm. Elsevier
2005, 65, 11-14.
[23] Pia, P.; Buentello, H.; Arriola, H.; Nava E., Hyperfine Interact 2008, 185, 173-177.
[24] Dohnalova, Z.; Sulcova, P.; Trojan, M. J. of Thermal Analysis and Calorimetry. 2008,
91, No. 2, 559-563.
[25] Jansen, M.; Letschert, H.P., Nature 2000, 404, 980-082.
[26] Mathur, S.; Veith, M.; Shen, H.; Hufner, S.; Jilavi M.H., Chem Mater. 2002, 14 (2),
568-582.
[27] Maciel, G.; Rakov, N.; Zanon, R.A.; Fellows, C.E.; Guimaraes, R.B.; Rodrguez, J.J.,
Chemical Physical Letters 2008, 465 (4-6), 258-260.
[28] Yutaka, K.; Akio, Y.; Elichi, H.; Tsuyoshi, K.; Yoshinori, T.; Yoshihiro, T.,Hirofumi,
K., Proceedings Inorganic Optical Materials II 2000, 4102, 144-151.
[29] Rauch, W.; Benher, H.; Gieres, G.; Sipos, B.; Seeboeck, R. J.; Kerner, R.; Soelkner, G.;
Gornik, E., Applied Phys. Letters 1992, 60 No.26, 3304-3306.
[30] Kalarical, J.; Cheruvathoor, P.; Balachandran, U.; Ramasami T., Solar energy Mat. and
solar cells, 2008, 92 No.11, 1462-1467.
[31] Garca, A.; Llusar, M.; Calbo, J.; Tena, M.A., Monros, G., Green Chemistry 2001, 3,
238-242.
[32] Goldwasser, M.; Rivas, M.; Pietri, E. ; Prez-Zurita, M.; Cubeiro, M.; Gingembre L.;
Leclercq, L.; Leclercq, G., Applied Catalysis 2003, 255, No.1, 45-57
[33] Lepe, F. J.; Fernndez, J.U.; Mestres, L.; Martnez M.L., Journal of Powders Sources
2005, 151, 74-78.
[34] Feldhoff A.; Arnold, M.; Martynczuk, J.; Gesing, Th. M.; Wang, H., et al. Solid State
Sciences 2008, 10, 689-701.
[35] Chinie, A.M.; Stefan, A.; Georgescu, S., Romanian Reports in Physic 2005, 57, No.3,
412-417.
[36] Cizauskaite, S.; Reichlova, V.; Nenartaviciene, G.; Beganskiene, A.; Pinkas, J.;
Kareiva, A., et al. Materials Science Poland 2007, 25, No. 3, 755- 7--.
[37] Rice, C.; Jackel, J. L., Jour. of Solid State Chemistry 1982, 41, 308-314..
[38] Stoyanova, T. L.; Matteucci, F.; Costa, A.; Dondi, M.; Carda, J. Powder Technology
2009, 193 No.1, 1-5.
[39] Valsykiv, O.; Sakka, Y., Skorokhod V.V. J. of the European Ceram. Soc. 2006, 04, 157
19.
[40] Pavlov, R.; Blasco, B.; Cordoncillo, E.; Escribano, P.; Carda, J. B., Boletn de la
Sociedad Espaola de Cermica y Vidrio 2000, 39, No. 5, 609-616.
[41] Cruciani, G.; Matteucci, F.; Dondi, M.; Baldi, G.; Barzanti A., Zeitschrift fur
Kristallographie 2005, 220 No.11, 930-937
[42] Matteucci, F.; Dondi, M; Cruciani, G; Baldi, G.; Barzanti, A., Engineering Materials
2004, 264-268.
[43] Dalziel, J. Journal of the Chemical Society 1959, 1993-1998.
[44] Bucko, M.; Stobierska, E.; Lis, J.; Gubernat, A. Materialy Ceramiczne 2008, 60(1), 20-
27.
INDEX
anisotropy,xi,xiii,327,377,378,382,383,400,402,
A 408,409,412,418,419,440,446,481,482,483,
484,485,486,487,496,497
absorption,48,119,125,128,129,355,510,540,
annealing,19,31,35,37,46,88,98,100,120,122,
541,545,550
140,145,220,255,259,261,262,266,269,270,
accounting,xii,205,437,456
285,332,359,363,367
acetone,53,118
aqueousalkalisolution,547
achievement,504
argon,3,220,504,508
acid,3,4,30,46,47,51,52,53,54,86,321,324,
aromatics,48
325,326,376,547
assessment,206
activationenergy,74,76,84,85,108,136,142,144,
assumptions,206
151,152,183,190,197,198,240,276,277,278,
asymmetricmolecules,103
516
asymmetry,259,262,267
activecenters,53
atomicforce,203
activeoxygen,103
atomicpositions,222
activesite,x,28,43,55,100,320,331,335,336,
authors,18,23,43,44,46,47,90,181,183,201,
337
203,206,207,219,245,286,298,312,401,409,
adaptability,284
410,417,472,476,498,542,546
additives,11,166,286,297,298,313,339
averaging,409,483,492
adhesion,69,118,166
adsorption,24,29,43,44,48,53,54,55,69,83,
103,106,107,108,109,113,144,145,147,331, B
333,339,476
AFM,231,237,238,240,241,245,247,261 background,81
ageing,40 bandgap,39,228
aggregates,325 barium,3,17,48,284,287,299,316
alanine,325 barriers,viii,68,109,136,207,240
alcohol,3,12,287,324,339 basicresearch,498
aldehydes,48 batteries,xiii,340,501
alkalineearthmetals,3 behaviors,ix,37,38,43,53,236,244,251,259,271,
aluminium,476 273,275,278,330,331,335,526
aluminumoxide,7 bending,166,205
ammonia,4,30,53,166,323,324,326 benign,327
ammonium,547 benzene,52
amplitude,38,440,447,448 bias,252,261,265,266,269,327,492
anatase,3,17 binaryoxides,349,363,464
aniline,331 binding,87,93,94,132
bindingenergy,87,93,132
560 Index
biotechnology,29 chemicalproperties,3,322,335,346
birth,32 chemicalreactions,321,368
bismuth,287,299,362,374,375 chemicalreactivity,302
blends,86 chemicalstability,539
blocks,346,347,348 chemicalvapordeposition,159,265,321
bonds,39,216,224,259,262,267,297,323,324, chromium,xiv,539,541,545,546,549,550,554
470,475,510 classes,464
buildingblocks,24 classification,540
butadiene,35 closure,35
byproducts,52 clusters,39,154,196,228,230,232,325,449,535
coatings,xiv,539
cobalt,4,52,75,216,226,237,326,332,372
C
cohesion,32
coke,171
cadmium,540,542
color,iv,324,540,541,544,545,550,554
calcinationtemperature,36,45,90,93,95,114,
combustion,3,36,45,46,47,48,50,52,53,115,
122,154,157,287,288,290,322,325
321,325,340,476,548
calcium,vii,31,341,346,347,372,432,509,541,
commonrule,447
543
communication,283,284,286,296,545
candidates,253,271
communicationsystems,283,284,296
capillary,7,8
communicationtechnologies,284
carbides,320,543
community,526
carbon,7,8,11,29,35,43,93,115,116,325,524
compatibility,69,118,165,329
carbondioxide,524
compensation,73,124,439,503
carbonmonoxide,11
competition,186,224,226,237
carbonnanotubes,7,8
complexity,180
carboxylicacids,48
compliance,181,191,196
carboxylicgroups,324
compression,166,193,199,216,220
carrier,205,231,233,239,273,518
concordance,378,410,432
cast,326
concrete,421
catalyst,x,xi,3,8,11,43,44,45,46,47,48,52,53,
condensation,48
54,55,86,154,166,167,168,171,319,323,325,
conduction,39,78,96,109,190,192,201,202,214,
327,330,331,332,333,335,336,337,338,342,
224,237,240,244,271,272,273,274,275,276,
345,367,539
277,278,327,329,378,379,401,476,516,519
catalyticactivity,43,44,46,47,50,52,53,55,56,
conductor,79,141,247,340,523
69,85,117,325,336,340,348
confidence,417,418
catalyticproperties,20,50,69,78
confidenceinterval,417
catalyticreaction,x,55,86,319,328,337,338
configuration,31,39,110,124,200,224,252,392,
cathodematerials,69,74,86,87,118,135,161,
399,432
162,328,341
confinement,37
cation,52,71,72,73,74,75,76,90,107,108,109,
connectivity,xiii,37,187,192,198,481,482,483,
113,180,192,214,215,216,225,227,228,230,
487,491
285,320,324,326,327,335,464,465,468,469,
connectivitypatterns,xiii,481,482
470,503,504,509,540,544
consensus,215
chalcogenides,29,467
construction,20,55
challenges,284,286,297
consumption,50,121
channels,5,284,327
contacttime,147
character,44,96,180,188,192,193,196,198,206,
contamination,327
346,348,351,432,468,519
continuity,492
chemicalindustry,2,358
contradiction,410
chemicalinteraction,160,161,162
Index 561
control,14,22,23,28,55,56,78,87,115,166,199, dataset,361
220,227,228,252,271,289,298,323,526 decay,149
controversies,238 decomposition,x,11,14,30,42,43,50,72,89,302,
conversion,14,39,40,42,44,46,47,51,54,116, 320,324,325,333,335,336,337,338,339,341,
148,149,167,168,169,171,331,333,334,356, 343,359,366,367,368,451,508
360,361,363 defectformation,70,278
cooling,23,161,192,238,242,244,246,247,259, defects,viii,33,39,45,68,69,70,71,72,79,92,
326,346,504,505,526,528 100,108,114,140,144,181,205,240,253,273,
coolingprocess,259 286,364,503,506,516,519
coordination,xi,xii,108,109,125,131,216,237, deficiency,109,286,504
378,402,403,405,408,409,410,411,437,451, definition,398,399
452,459,464,465,466,467,468,503,512,514, deformation,205,220,468,469,510
522,540,543 degenerate,224
copper,44,48,167,286,298,324,331,453 degradation,35,54,118,190,202,253,286,290,
copyright,iv,320,321,328,329,330,334,335,336, 296
337,338 dehydration,359,368,515
correlation,xii,38,75,101,186,218,272,321,379, dendrites,24
437,464,504 density,43,44,97,113,114,131,160,162,163,
cost,x,7,23,24,45,69,283,286,319,331,338 164,166,202,206,252,253,268,273,286,287,
cotton,54 288,290,297,300,301,304,342,343,440,474,
Coulombinteraction,231 475
coupling,ix,xii,12,36,74,75,200,213,224,226, densityfunctionaltheory,343
401,437,482 densityvalues,164
covalency,93,471 deposition,8,9,19,116,159,203,255,265,266,
covalentbond,95,323 321
criticalanalysis,207 depositionrate,266
criticalvalue,11,203,235 deposits,115
criticism,438 derivatives,79,363,398,401,465,470
crown,19 desiccation,465
crystalgrowth,13,23,28,256 desorption,viii,43,68,69,78,83,84,87,106,107,
crystalstructure,vii,1,2,22,25,52,203,214,215, 108,109,113,114,137,144,145,146,147,148,
217,218,219,255,259,287,464,467,468,470, 149,172,333,504
473,476,507,508,519,542,544 destruction,30,44,69,85,110,113,162,269
crystalline,xii,5,12,14,15,17,19,20,23,24,28, deviation,83,181,190,215,222,416,466,468
35,37,38,47,53,54,55,56,92,181,187,205, DFT,viii,68,107,108,109
266,313,327,340,463,545,546 dielectricconstant,ix,38,251,252,253,257,258,
crystallinesolids,xii,463 259,261,262,263,264,265,266,269,270,273,
crystals,xii,24,29,31,32,172,187,205,207,214, 284,285,287,311,317,482,487
222,359,408,437,438,439,475,512,541,542, dielectricpermittivity,484,485,486,487,495
545 dielectrics,252,273,275,284,297,315,317
cubicsystem,502 differentialequations,82,146
cuprates,438,544 diffusereflectance,119
currency,343 diffusionprocess,326
CVD,159,265 diffusivity,153,214
cycles,161,192,242,244,246,247 dimensionality,vii,1,2,419
dipolemoments,456
discrimination,102
D
disorder,37,186,187,190,206,214,215,216,220,
222,227,228,230,231,233,236,240,247,475
darkconductivity,39
disorderedsystems,233,239
dataanalysis,83
562 Index
dispersion,xi,53,122,166,172,323,327,378,379, electrons,19,38,71,72,74,75,207,224,225,240,
408,409,414,416,417,419,482,545 245,271,273,277,278,330,337,378,379,401,
displacement,224,259,261,267,492,509 440,472,475
dissociation,44,77,115,145,162,333 electroplating,166
distilledwater,13,86,118,323,359 electrospinning,19
distortions,xii,223,320,373,463,468,543,544 elongation,261
distributionfunction,262 elucidation,83,187,206,275
divergence,230,233 email,463
diversity,xiv,7,69,438,539,544 emission,38,39,40,44,272,274,278,447,545
domainstructure,92,482 endothermic,359,362,366
dominance,506 engineering,29,358
doping,xiii,44,46,52,71,74,76,99,111,113,124, entropy,76,78
214,215,223,225,227,234,263,525,526,533, environment,14,42,95,131,219,224,262,263,
535,536 265,270,337,382,440,444,468
drying,7,8,119,166,326 environmentalprotection,2
DSC,119,354,355,356,357,368 enzymes,9
DTAcurve,368 epitaxialfilms,ix,180,181,187,202,204,208
duration,23,54 epitaxialgrowth,545
dyes,544 equilibrium,24,70,72,83,103,107,108,113,258,
dynamics,viii,68,78,102,325,340,518,522 259,270,273,346,348,440,441,508,515,528
equipment,79,276,358,546
erbium,549
E
ethanol,18,23,52,53,119,287,326
ethylacetate,52,53
earth,ix,20,39,51,72,75,86,213,214,215,216,
ethylene,8,23,32,86,118
223,225,227,232,234,247,346,348,351,373,
ethyleneglycol,8,23,32,86,118
438,453,464,468,470,474,476,501,503,504,
europium,372
510,511,514,521
evaporation,32,35,286,301,324,325,326
education,279
evolution,x,23,25,46,54,84,88,89,115,117,
eigenvalues,380
137,154,157,158,181,182,183,185,186,190,
electricconductivity,182,192,193,194
192,228,245,329,345,368,515
electricfield,ix,xii,9,38,194,251,252,257,258,
EXAFS,viii,68
259,261,262,267,270,271,292,439,463,483,
excitation,38,71,278
492
exciton,475
electricalconductivity,80,138,152,225,374,476,
exposure,39,43,55,56
545
externalmagneticfields,240
electricalproperties,ix,179,180,181,187,202,
extraction,29
228,230,290,521
extrapolation,532
electrocatalysis,329,330
electrodes,69,87,117,252,253,254,256,259,
329,342,539 F
electrolyte,69,118,142,159,160,161,163,164,
328,329,522 fabrication,vii,1,2,3,5,7,8,9,11,14,17,19,21,
electromagnetic,284,540 24,29,31,32,35,55,56,69,165,256,286
electrondiffraction,13,203,219,220 family,x,227,283,285,286,371,375,442,475,
electronmicroscopy,87,187,199,218,219,220, 502,503
253 Fermilevel,71,224,331
electronstate,74 ferrimagnets,432
electronicstructure,52,71,75,373 ferrite,73,100,107,109,113,358,361,548,549
ferroelectrics,ix,251,252,256,257,259,263,270,
545
Index 563
ferromagnetism,ix,36,213,214,215,225,227, groups,x,3,14,20,31,191,207,216,220,236,345,
228,230,231,233,234,378 346,347,348,350,442,447,461,491,504,510,
fiber,14,545 511,512,513,514,522
fiberoptics,545 growth,8,10,12,14,18,20,22,32,55,86,95,114,
fieldtheory,xiii,71,224,418,525,526,528,529, 154,186,190,196,203,226,256,266,283,290,
531,532,533,535,541 301,323,325,326,327,348,376
filmthickness,181,202,252,253,273 growthmechanism,20
filters,x,252,283,312,313,316 growthrate,323
finetuning,75
finiteelementmethod,474,493
H
flame,46
fluctuations,205,241,456
Hamiltonian,379,380,381,382,385,412,419,425,
fluorescence,39
456
fluorideions,469
hardness,467
foams,166
heat,15,21,108,144,194,359,360,465
foils,23
heatingrate,87,147,354
formula,xii,xiv,68,71,89,183,184,203,214,215,
height,191,196
225,290,348,354,375,385,396,418,425,447,
helium,87
463,465,469,473,475,501,502,503,504,508,
heptane,52
539,540,543,549
heterogeneity,7,111,205,263
freeenergy,257,321,528
heterogeneouscatalysis,55,100,329,330,342,343
freevolume,447
hexane,50,339
freedom,214,226
holmium,548,549
frequencies,257,263,313,315,510
homogeneity,31,199,323,324,326,439,503,504,
frustration,230
508,516
fuel,xiii,85,154,159,166,167,168,171,178,214,
host,39,302,526,535,540
325,328,341,342,501,522
HRTEM,13,14,16,18,25,26,34,49,253
hue,544
G humidity,508,515
hybridization,124
gadolinium,350,360,547 hydrides,320,467,543
gassensors,214 hydrocarbons,11,42,48,93,154
gel,3,7,9,12,15,18,20,24,28,29,31,36,43,46, hydrogen,54,114,159,323,324,327,328,376,
52,54,86,118,199,324,325,327,329,545,548 468,504,510,511
gelformation,86,118 hydrogenbonds,324,510
gene,455 hydrogenperoxide,54
generation,viii,9,12,14,17,19,35,38,44,52,53, hydrogenation,342
55,56,68,69,70,71,75,154,271,328 hydrolysis,46,465
glasses,x,283,287,297,299,300,301,302,303, hydroquinone,331
304,305,306,307,316 hydrothermalprocess,52
glycerol,325 hydrothermalsynthesis,3,20,548
glycine,325 hydroxide,xiii,12,19,23,25,86,323,359,366,501,
glycol,8,23,32,86,118 546
gold,376 hydroxyapatite,29
grainboundaries,36,37,39,200,201,205,240, hydroxyl,513,515
241,244,275 hydroxylgroups,515
grains,37,187,195,240,323 hypothesis,109
granules,181 hysteresis,38,199,201,238,241,257,262,270,
graphite,447 271
hysteresisloop,38,199,201,238,262,270
564 Index
IRspectra,510,511
I
IRspectroscopy,510
Islam,176
ideal,xii,36,68,77,202,222,439,440,441,444,
isolation,206
445,453,454,463,465,466,467,468,470,472,
isothermalheating,368
543,544
isotherms,515
identity,380,484,492
isotope,viii,68,69,78,81,82,83,84,85,100,102,
illumination,54
103,107,114,117,137,138,139,140,142,143,
image,4,5,9,12,21,22,26,33,91,125,161,221,
146,148,151,157,162,171,331
256
issues,253,275
implementation,284
iteration,422,424
impregnation,5,46,47,52,166,326
impurities,43,69,207,351
inclusion,68,410,483,484,497,542 J
incompatibility,442
independence,147 judgment,533
independentvariable,82
indication,25,256,337
K
indium,545
industry,286,296
ketones,48
inelastic,xi,202,207,378,405,409,414
kineticcurves,361
inequality,185,454,483
kineticequations,82,83
infancy,327
kinetics,78,79,86,114,142,148,149,154,166,
infinite,80,149,269,325,401
346,356,359,361,363,365
inhibitor,290
KOH,14,19,20,21,24,33,53,323,547
inhomogeneity,181,187,202,205,206,207,208
inhomogeneties,8
initialstate,102 L
insertion,284,503,504,506,509
insight,193,198,244,300,338 Landautheory,xii,437,450,452
instability,89,118,120,162,350 lanthanide,ix,xiv,30,36,213,214,218,219,346,
insulators,226 350,363,365,539,545,546,549,550,554
integration,85,283,286 lanthanum,4,20,31,46,50,71,73,75,76,93,100,
integrity,167,327 102,107,108,109,110,113,192,326,341,342,
intercepts,532 359,432,546,548
interface,38,118,145,146,148,152,161,162,163, lasers,xiv,539,545
172,202,203,253,255,256,268,269,271,275, latticeparameters,88,120,121,219,401,408,418,
492 508
interfaciallayer,253,255,256 lattices,205,255,374,381,382,383,386,392,396,
interrelations,482 397,410,421,424,468
interrelationships,464 laws,530
interval,183,184,187,427,515 layerbylayergrowth,203
iontransport,78,503 leakage,268,271
ionexchange,19,348,373 ligand,74,224,541
ionicconduction,476 lightscattering,544
ionicforces,323 likelihood,187
ionicity,95 limitation,168,331,456
ionization,277,278 line,35,81,87,134,167,195,197,198,203,204,
iron,x,52,75,345,358,359,376,468,540,542, 239,391,393,406,450,459,508,511,512,514,
544,546,549 522
irradiation,15,54,328,331,342 lineardependence,273
Index 565
liquidphase,297,301,303,305,323,331 membranes,viii,8,9,68,69,70,76,85,86,110,
liquids,8,323 118,154,165,168,172,214,522
lithium,285,286,287,289,290,298,299,315,316, memory,252,271,278,358,464,539
340,475 memoryprocesses,271
localorder,452 metalhydroxides,30
localization,237,245 metaloxidenanofibers,11
lowtemperatures,viii,xi,3,45,48,75,96,104,107, metaloxides,2,11,12,15,29,32,35,36,45,46,56,
147,179,180,193,196,204,228,230,234,240, 107,147,226,271,326,364,371,372,524
241,377,408,419,455,465,471,475,503,519 metalsalts,19,20,30
luminescence,39,475,545 metals,3,29,30,45,71,75,114,225,253,324,325,
lying,39,201,464,533 336,346,347,348,379,464,465,501,510,511,
514,521
methanol,32,33,50,53,54,55
M
methodology,35,472
methylmethacrylate,29
macropores,35,166
micelles,12
magnesium,287,299
microemulsion,3,43,321,324,325
magnet,418,419
microscopy,87,199,203,219,220
magneticfield,ix,37,189,190,199,200,206,213,
microspheres,vii,2,24,29,31,32,35,55
214,225,237,238,239,240,241,242,528
microstructure,x,118,186,187,192,195,199,221,
magneticmoment,36,37,227,233,238,410
253,304,305,341,345,364,482,483,485,496,
magneticproperties,37,186,192,196,202,203,
549
214,230,235,372,374,375,432,540,544,546
microstructurefeatures,192
magneticresonance,510
microtome,26
magneticstructure,192,215,225,410,432,533
microwaves,284
magnetism,viii,179,399,419
migration,76,78,98,107,108,109,113,145,146,
magnetoresistance,ix,37,180,187,190,191,192,
151,162,163,401,518
199,200,202,206,213,214,215,225,237,240,
miniature,283
245,247,346,348,526
minority,124,506,519
majority,2,124,151,297,298,360,540
misfitdislocations,253
manganese,xi,xii,20,31,180,192,199,376,378,
mixing,118,216,287,325,359,548,549
381,382,396,409,410,432,433,470,541
mobilephone,284
manipulation,8,166,346
modeling,69,142,146,449,467
mantle,468,476
modules,286,297,447
manufacturing,viii,x,68,69,297,319,495
modulus,474
massloss,149,368
molarvolume,475
materialsscience,29,284,338
mold,326
matrix,xiii,37,86,202,231,255,320,321,327,
mole,90,102,109,144,295,299,508
328,346,481,482,483,484,485,487,492,496,
molecularoxygen,331
497,545,547
molecules,11,12,22,44,81,82,84,102,115,137,
meanfieldtheory,xiii,525,526,528,529,531,532,
138,139,142,145,147,162,325,475,504,508,
533,535
510,512,514
measurement,38,78,85,329,336,447,475,506,
momentum,396,398,400,402,412,416
528
monolayer,84,98,106,107,266,269
measures,204
MonteCarlomethod,206
mechanicalproperties,474
morphology,vii,1,2,7,16,17,19,20,21,22,23,
mechanicalstress,186,484,492,493
24,25,37,39,43,47,50,53,55,87,118,124,
media,30,184,284,484
166,184,261,306,325,327,549,550
melt,30,359
motif,222
meltingtemperature,467,475
motion,9,180,259,278
566 Index
movement,148
multilayerfilms,266
O
multilayeredstructure,286,297,298
observations,202,261
multimedia,284
oil,324,325
multiplier,77,455
operator,xi,xii,377,378,379,380,381,383,384,
402,409,419,425
N opticalproperties,346,476,539,545
optimization,290,371
nanobelts,19 orchid,541
nanocomposites,viii,68,70,118,119,122,135, ores,464
141,146,148,154,162,164,165,166,171 organiccompounds,viii,2,3,48,50
nanocrystals,3,14,47,118,545 organicsolvents,70,172
nanofibers,8,11,14,19,43,55 orientation,37,240,256,261,265,266,269,270,
nanomaterials,3,7,8,11,13,52 322,401,418,492
nanometer,3,231 oscillation,448
nanometerscale,231 Ostwaldripening,14
nanoparticles,3,4,5,14,19,22,23,24,35,36,43, overlap,96,124,471
47,49,52,54,162,548 oxalate,17,30,32,547
nanorods,9,10,12,14,15,17,18,19,24,37,39, oxidationproducts,115,116,168
53,54 oxideelectrodes,252,253
nanoscalematerials,11,24 oxidenanoparticles,22
nanostructuredmaterials,7 oxygenabsorption,119
nanostructures,24,27,326 oxygensensors,476
nanosystems,118
nanotube,7,8,341
P
nanowires,6,7,8,11,14,15,16,17,19,23,37,38,
39,53,54
packaging,297,315
naturalgas,44,340
palladium,44
negativity,401
parallel,79,85,202,225,252,254,256,259,276,
neglect,258
392,468,483,487,491
neodymium,326,366,367,545,549
parallelism,418
network,34,202,206,216,284,470
particlemorphology,24,50,549
nextgeneration,342
partition,450
nickel,75,166
passive,286,297,487
niobium,503,509,522
pathways,46,48,355,358,371
nitrates,30,31,32,35,46,51,55,86,118,166,323,
percolation,135,184,192,196,203
324,325,326,464,465,546
perforation,24
nitrides,320,543,545
PeriodicTable,321
nitrobenzene,331
permeability,85,165
nitrogen,3,43,55,168,325,333,339
permeablemembrane,69,85,165
nitrogencompounds,43
permeation,78
nitrogenoxides,55,325,339
permission,iv,320,321,328,329,330,334,335,
NMR,230,232,475,510,511,512,513,514,515,
336,337,338,489,490
522
permittivity,x,273,283,285,290,292,294,297,
noblemetals,45
304,305,306,313,316,484,485,486,487,495
nonequilibrium,453
perovskiteoxide,69,84,214,275,277,278,279,
nonlinearequations,72
320,321,327,328,329,331,335,338,339,341,
nucleation,12,86,114,154,183,186,266,323
343,466,471,472,476,545
nuclei,114
peroxide,54
numericalanalysis,83
Index 567
phaseboundaries,329 probe,87,331
phasediagram,180,214,439,456,459,508,509 processgas,266
phasetransformation,367,508,522 production,7,24,56,110,284,327,328,522
phasetransitions,37,115,180,401,418,453,459, project,433
468,475,536 propagation,284
phenol,331,342 propylene,8
phonons,207,379 proteins,9
photocatalysis,53,328,330 protons,xiii,374,501,506,508,512,514,515,516,
photoconductivity,39 518,519,522
photoemission,132,202 pulse,87
photographs,14,41 purewater,329,341
photoluminescence,38,39,545,546 purity,23,199,220,287,290,299,528,540
physicalproperties,vii,ix,xii,172,207,215,216, PVA,12,287
223,225,227,237,242,247,251,335,348,438, PVP,21,22
464,470,481,482,540 pyrolysis,46,48
physicochemicalproperties,vii,1,2,7,15,17,29,
43
Q
physics,vii,ix,1,2,180,187,213,321,432,479,526
piezoelectricproperties,497
quantization,402,412
piezoelectricity,69
quantumchemicalmethods,172
PLspectrum,39
quartz,32,119,259
platinum,87,287,548,549
PMMA,29,31,32,33,35,55
polarization,38,69,73,78,98,257,262,266,271, R
483,491,492,497
pollutants,42,48 radiation,19,87,287,302,303,544,549
pollution,44 radio,283,540
poly(methylmethacrylate),29 radium,3,541,549,550
polycarbonate,7 radius,214,215,218,223,227,235,311,412,439,
polymer,11,30,32,35,56,324,482,483,485,487, 447,459,468,470,487,526,543
491,492,493 Ramanspectra,522
polymermatrix,483,492 rareearthelement,348,438
polymorphism,544 rawmaterials,13,24,290,548,549
polystyrene,24,29 reactant,x,14,17,23,44,319,322,328,337,338
polyurethane,166 reactionmechanism,86,357,363,367
polyurethanefoam,166 reactionmedium,17
polyvinylalcohol,32 reactionrate,x,46,47,345,351,356,357,364,365,
poor,129,198,214,304 371
porosity,35,69,86,122,145,165,166,194,195, reactiontemperature,14,23,41,42,50,51,332,
301,303,304,324,327,328 465
porousmaterials,vii,1,2,35,194,327 reactiontime,24,361
potassium,12,46 reactionzone,366
power,ix,85,160,162,163,164,213,244,246, reactions,x,23,29,43,52,54,69,70,72,83,103,
247,342,358,469,529,530 193,313,320,321,324,325,330,331,333,336,
praseodymium,542,545,549 338,346,349,350,360,362,367,368,371,374,
precipitation,3,8,35,315,321,323,325,547 464
preference,379,432,470 reactivity,viii,23,29,67,68,69,70,86,110,114,
priming,322 154,171,220,247,290,302,327,348,351,359,
probability,118,183,381,442,444,451,452,456 367,546
probabilitydistribution,442 reading,526
568 Index
reagents,8,53,86,167 saturation,36,37,190,193,199,202,203,227,323,
reality,408,432,459 528
reason,19,37,76,111,161,184,196,203,205, scaling,172,531,533
207,220,231,257,290,301,306,333,464,510, scanningelectronmicroscopy,187
535 scatter,416
recall,448 scattering,xi,39,201,202,203,205,206,207,231,
reception,338 240,245,247,262,378,409,414,418,440
recombination,145,475 screening,253
reconstruction,101,220,508 search,284,329,348
recrystallization,118 seed,18
redshift,127,129 segregation,75,90,100,121,134,135,152,322,
redistribution,viii,68,79,81,82,120,123,131,135, 323
136,142,144,172,224 selectedareaelectrondiffraction,220
reflection,88,203,550 selectivity,48,86,115,116,167,168,169,322
reflectionhighenergyelectrondiffraction,203 selfassembly,24,35
reflexes,447 selforganization,4
refractiveindex,467 SEMmicrographs,161,163,549,550
regeneration,44,333 semicircle,276
relationship,vii,1,2,37,56,143,257,276,332, semiconductor,75,96,183,238,239,327,408
355,444,464,467,472,473,475,476 sensing,53,55
relativesize,216,470,471 sensitivity,viii,xiii,44,53,69,143,167,179,193,
relaxation,37,69,78,79,80,81,83,87,101,119, 476,481,482,495,497
142,151,152,262,269,439,475 sensors,xiii,53,87,214,476,497,501,526
relevance,247,342 separation,viii,ix,x,29,68,76,85,86,165,207,
reliability,284,286,533 213,214,225,228,230,232,234,247,255,345,
reparation,viii,68,118,120 358
replacement,vii,1,2,291 shape,7,9,12,20,22,23,24,30,31,39,95,115,
requirements,69,86,321 127,132,134,181,184,198,322,367,465,475,
residues,322,324,325 483,487,540,544,549
resolution,4,87,92,125,128,129,219,220,253, sharing,216,354,465,470
255 shortage,44
resonator,x,283,284,285,312,313,315,317 signaltonoiseratio,495
respect,32,79,220,259,261,331,346,392,400, signs,418,427
450,456,497,503 silica,5,29,35,47,55,56,326,340,541,542,545,
rods,7,39,485,487,491,492,497 546
roomtemperature,32,37,86,93,118,187,215, silicon,265,297
216,218,219,227,237,240,244,247,261,262, silver,46,87,286,297,298,300,302,303,324,373
263,264,265,270,276,277,278,323,324,326, simulation,196
466,469,482,526,528,540 singlecrystals,24,187,205,207,359
rootmeansquare,417 skeleton,32
rotations,468 smog,48
roughness,202 smoothing,166
rutile,284,307,315,467 sodium,24,214,323,366,367,542,546
sodiumhydroxide,323
solarcells,557
S
solgel,3,7,9,12,15,18,19,28,29,36,43,46,52,
54,199,325,328
salt,18,19,23,32,37,55,346,347,348
solidoxidefuelcells,viii,68,69,339
samarium,549
solidphase,82,107,342,508,540
satellite,283,284
Index 569
solidsolutions,xii,xiii,70,71,74,75,76,86,118, successiveapproximations,440
313,348,351,353,375,376,437,439,440,441, sulfur,45
450,481,482,497,504,514,523,540,545 superconductivity,38,69,214,346,348,372,468
solidstate,28,118,187,193,287,289,321,346, superiority,vii,2
362,464,540,541,544,545,546,548 superlattice,4
solidification,29,31 supply,167,171
solubility,23,30,301 suppression,36,241
solvents,19,70,172,465,544 surfacearea,vii,x,xii,1,2,5,35,38,39,43,45,46,
space,xii,24,32,38,42,45,46,47,52,206,216, 47,50,52,53,54,55,56,82,87,90,147,148,
217,219,220,268,271,273,320,326,327,410, 167,190,319,321,322,323,324,325,326,327,
437,448,449,456,457,458,459,461,465,468, 329,340,437,459
502,508,509,549 surfacechemistry,100,327
species,vii,viii,5,30,43,46,53,68,86,93,95,100, surfaceenergy,55,107,118,322,548
106,232,326,465 surfacelayer,82,87,93,95,100,103,108,119,123,
specificsurface,vii,1,2,48,53,87,90,147,148, 129,130,131,132,134,135,153,166,190,196,
323 202,206,458,459
spectroscopy,183,203,207,275,278,360,361, surfacemodification,115
510,521 surfaceproperties,viii,53,68
spectrum,xi,92,93,127,128,132,361,378,379, surfacereactions,83
395,402,408,410,510,512,514,515 surfacestructure,342
speed,x,284,323,331,345 surfactant,3,8,11,12,14,17,22,23,24,27,35,55,
spindle,47 56,118,324,325
squarelattice,487 susceptibility,36,188,189,190,196,198,199,200,
stability,x,xii,5,32,48,52,53,54,69,118,134, 230,237,239,372,418,526,530
165,166,214,216,265,270,284,290,316,328, suspensions,70
340,345,346,358,371,401,416,419,437,450, switching,ix,38,39,149,251,271,275,278
459,464,470,471,497 SWNTs,8
stablestates,xii,437 symbols,243,386
standarddeviation,416 symmetry,xii,3,74,88,89,219,220,221,237,262,
statistics,441 270,332,336,401,403,412,438,440,442,447,
steel,4,86,166,167 450,454,463,468,469,475,484,508,509,544,
stoichiometry,22,73,74,86,101,167,205,216, 549,554
220,225,237,241,247,266,323,324,326,355,
359,541,544,547
T
storage,ix,11,45,213,214,252,346,468
strain,187,215,259,483,492
tantalum,291,295,503,522
strategy,viii,2,3,5,7,9,15,17,18,19,20,22,24,
TCC,315
28,30,32,35,45,46,48,52,55,56,220,327
TEM,viii,4,5,12,13,16,17,18,20,21,25,26,36,
strength,viii,17,53,55,68,78,83,84,100,103,
49,50,68,87,92,122,125,151,172,220,222
107,113,115,127,144,147,166,321,323
temperatureannealing,100
stress,37,118,194,205,492
tension,216
stretching,510
terraces,203
stronginteraction,180,512
testing,70,159,160,161,163,167,332,475
strontium,31,72,137,342,349,359,372,523,547
tetragonallattice,483
structuralchanges,432,464,509,528
texture,259,266,270
structuralcharacteristics,118,121,360
TGA,354,355,356,357,359,365,368
structuraldefects,45,92,181
thermalactivation,239
structureformation,20,36
thermalanalysis,359,362,366,368
structuring,482
thermaldecomposition,324
substrates,183,203,252,265,266,322,545
thermalenergy,533
570 Index
thermalexpansion,69,118,160,162,165
thermalproperties,475
U
thermalstability,x,54,166,319,325,366,539,546
uncertainty,418
thermaltreatment,32,35,38,325
uniform,4,7,9,14,22,122,140,147,166,269,
thermodynamicequilibrium,348
322,327,367,535
thermodynamics,346
urea,325
thermogravimetry,504,507,519,521
thinfilms,vii,ix,19,202,203,207,214,251,252,
253,256,257,259,262,263,266,270,271 V
thinking,35
threshold,535 vacuum,257,324,326
tin,285,541 valence,xi,20,39,74,225,358,377,378,410,433,
tinoxide,541 450,464,468,472,473,476
titanate,vii,xii,3,12,14,19,24,277,284,285,316, vanadium,542
367,376,481,482 vapor,xiii,3,44,159,265,321,501,504,507,508,
titanium,vii,x,3,23,24,167,255,283,295,298, 510
312,362,366,541,546 variables,82,83,138,139,451,464
toluene,50,52,53 variance,230
totalenergy,459 variations,126,187,205,346,459,531
toxicmetals,545 vector,392,412,450,454,483,491,492,497
toxicity,545 vehicles,42
transducer,482 velocity,42,45,46,47,52
transformation,83,109,115,116,143,166,170, vibration,262,267,510
191,192,193,198,202,214,225,349,362,508, viscosity,19,301
522,547 volatilization,52,289,465
transitionelements,540,541,546
transitionmetal,viii,3,11,19,30,56,68,69,71,73, W
74,75,85,86,96,100,107,110,113,123,129,
135,142,147,151,214,224,226,247,336 watervapor,xiii,44,501,504,507,508,510
transitionmetalions,224 wavenumber,119
transitiontemperature,xiii,37,38,187,207,236, wavepropagation,284
238,242,262,270,525,526,535 wavelengths,540
transitions,ix,xii,86,124,127,186,193,213,225, wealth,346
227,230,237,242,392,401,402,432,453,454, weightloss,viii,68,148,150,359,366,515
459,463,464,475 wettability,301
transmission,253,346 windows,79
transmissionelectronmicroscopy,253 wires,7,87
transparency,438 workers,3,5,8,9,11,14,15,18,19,22,23,24,31,
transportation,48 32,35,37,39,43,46,54,247,468
treatmentmethods,432 workingconditions,86
trends,ix,179,180,181,182,187,191,192,199,
246,332,519,522
trial,532 X
tungsten,285,476
tunneling,37,181,194,195,199,201,202,206,207 XPS,viii,68,87,93,94,95,100,131,132,133,134,
tunnelingeffect,194,195 135,142,151,172
twinning,221 Xraydiffraction,xii,87,122,123,218,291,292,
293,294,302,356,358,359,361,367,368,437,
439,440,441,442,446,448,508,509,549,550,
551,554
Xraydiffractiondata,356,358
Index 571
Y Z
yttrium,48,51,545 zinc,287,299,316
zirconia,viii,32,68,69,70,117,130,131,171,287
zirconium,12,24,86,285,542,546
ZnO,30,287,298,299,309,314,316

(Maxim Borowski) Perovskites Structure, Propertie PDF

Uploaded by

Copyright:

Available Formats

(Maxim Borowski) Perovskites Structure, Propertie PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

(Maxim Borowski) Perovskites Structure, Propertie PDF

Uploaded by

Copyright:

Available Formats

CHEMICAL ENGINEERING METHODS AND TECHNOLOGY

Additional books in this series can be found on Novas website

Additional E-books in this series can be found on Novas website

Nova Science Publishers, Inc.

NOTICE TO THE READER

LIBRARY OF CONGRESS CATALOGING-IN-PUBLICATION DATA

Perovskites : structure, properties, and uses / editor, Maxim Borowski.

Published by Nova Science Publishers, Inc. New York

Chapter 7 Perovskite-type Oxides: Synthesis and Application in Catalysis 319

Perovskite is a calcium titanium oxide mineral species composed of calcium titanate.

author shall highlight the physicochemical properties of perovskite-type oxides and

The high temperature dependence of spontaneous magnetization at a separate site in the

between these two approaches. Complex oxides with composition

Of specific interest is PbTiO3 as an element of a hierarchical chain single-domain single

CONTROLLED FABRICATION AND CATALYTIC

Jiguang Deng, Lei Zhang, Yuxi Liu and Hongxing Dai*

fabrication of perovskite-type oxides and perovskite-like oxides with specific

concentrate on the applications of these addressed materials in catalysis, such as NOx

2. PREPARATION OF NANO/MICROSCALE PEROVSKITES

Figure 2. SEM image of LaCoO3 nanoparticles [32].

Figure 3. TEM image of the LaCoO3/SBA15 sample[39].

In order to overcome the drawback of lower surface areas of perovskite-type oxides, a

2.2. One-Dimensional Nano/Microstructure

Owning to their unique physicochemical properties and potential applications in various

2.2.1. Hard-Templating Synthesis

Figure 6. SEM image of the La0.6Sr0.4CoO3 nanotubes[92].

Figure 9. Schematic illustrations showing the formation of nanowires by templating against

2.2.2. Surfactant-Templating Synthesis

(a) (b) (c)

(d) (e) (f) 4 nm

2.2.3. Hydrothermal Synthesis

concentration of Ba(OH)2 [135]. Kang et al. observed the formation of single-crystalline

2.2.4. Novel Solid-State Method

An approach, called the composite-hydroxide-mediated method, was developed by Wang

BaCl2 + NaOH Ba(OH)2 + NaCl

Due to the big viscosity of hydroxide, formation of a perovskite nanostructure is slow,

2.2.5. Other Methods

Wagner and co-workers generated single-crystalline orthorhombic NaTaO3 nanorods [162].

Table 1. The influence of fabrication parameter on the morphology of La0.5Ba0.5MnO3

Sample KOH Filling Hydrothermal Hydrothermal Morphology

(a) (b) (c)

In addition to A1xA'xMnO3, LaFeO3 [177] and K1xNaxNbO3 (x = 01) [114,178]

Interestingly, by adopting a solvothermal process coupled with the surfactant-directing

2.4. Other Specific Morphologies

Figure 25. Schematic mechanism for the formation of bur-like nanostructure[191].

Generally speaking, adopting the above-mentioned hard/soft-templating method,

2.5. Macro/Mesoporous Perovskite Synthesis

Because of the attractive physicochemical properties and potential applications in

of 0.35:0.55:1:10:20:140 [226228]. The 3DOM La0.75Sr0.25Cr0.5Mn0.5O3YSZ (yttria-

Figure 32. Formation mechanism of 3DOM metal oxide[99].

(a) (b) (c)

(d) (e) (f)

Compared to mesoporous single metal oxides, reports on the synthesis of mesoporous

3.1. Magnetic Property

3.2. Electrical Property

3.3. Optical Property

4.1. Nitrogen Oxide Elimination

4.2. Methane Oxidation

Table 2. Surface morphologies, BET surface areas, preparation and reaction