Functional Metal Oxides E-Book
162,99 €
Mehr erfahren.
- Herausgeber: Wiley-VCH
- Kategorie: Fachliteratur
- Sprache: Englisch
Functional oxides are used both as insulators and metallic conductors in key applications across all industrial sectors. This makes them attractive candidates in modern technology ? they make solar cells cheaper, computers more efficient and medical instrumentation more sensitive. Based on recent research, experts in the field describe novel materials, their properties and applications for energy systems, semiconductors, electronics, catalysts and thin films.
This monograph is divided into 6 parts which allows the reader to find their topic of interest quickly and efficiently.
* Magnetic Oxides
* Dopants, Defects and Ferromagnetism in Metal Oxides
* Ferroelectrics
* Multiferroics
* Interfaces and Magnetism
* Devices and Applications
This book is a valuable asset to materials scientists, solid state chemists, solid state physicists, as well as engineers in the electric and automotive industries.
Sie lesen das E-Book in den Legimi-Apps auf:
Seitenzahl: 932
Veröffentlichungsjahr: 2013
Ähnliche
Table of Contents
Related Titles
Title Page
Copyright
List of Contributors
Part I: Magnetic Oxides
Chapter 1: Introduction to Magnetic Oxides
1.1 Oxide Structures and Crystal Chemistry
1.2 Oxide Growth
1.3 Magnetic Properties of 3d and 4f Ions
1.4 Magnetic Interactions in Oxides
1.5 Concentrated Magnetic Oxides
1.6 Dilute Magnetic Oxides
1.7 Conclusions
Acknowledgments
References
Chapter 2: Magnetic/Multifunctional Double Perovskite Oxide Thin Films
2.1 Introduction
2.2 Thin-Film Deposition
2.3 Structure and Morphology
2.4 Electronic Structure
2.5 Physical Properties
2.6 Applications of Multifunctional Oxides
2.7 Future Directions
Acknowledgments
References
Part II: Dopants, Defects and Ferromagnetism in Metal Oxides
Chapter 3: Magnetic Oxide Semiconductors: on the High-Temperature Ferromagnetism in TiO2- and ZnO-Based Compounds
3.1 Introduction
3.2 Properties of (Ti,Co)O22
3.3 Properties of Transition-Metal-Doped ZnO
3.4 Discussion
3.5 Summary and Outlooks
Acknowledgments
References
Chapter 4: Effect of Ta Alloying on the Optical, Electronic, and Magnetic Properties of TiO2 Thin Films
4.1 Introduction
4.2 Ta Substitution in TiO2: Doping or Alloying?
4.3 Diluted Magnetic Semiconductors (DMS)
4.4 Defect-Mediated Ferromagnetism
4.5 Magnetic Impurity Analysis in Ti1−xTaxO2 System
4.6 Defect-Induced Ferromagnetism in Ti1−xTaxO2 Film
4.7 First-Principles Spin-Polarized GGA + U Calculations
4.8 Mechanism of Defect-Mediated FM
4.9 Optimization of Ferromagnetism
4.10 Outlook for Defect-Mediated Properties of Ti1−xTaxO2
References
Chapter 5: Defect-Induced Optical and Magnetic Properties of Colloidal Transparent Conducting Oxide Nanocrystals
5.1 Introduction
5.2 Colloidal Transition-Metal-Doped Transparent Conducting Oxide Nanocrystals
5.3 Native Defects in Colloidal Transparent Conducting Oxide Nanocrystals
5.4 Summary and Outlook
Acknowledgments
References
Part III: Ferroelectrics
Chapter 6: Structure–Property Correlations in Rare-Earth-Substituted BiFeO3 Epitaxial Thin Films at the Morphotropic Phase Boundary
6.1 Introduction
6.2 Combinatorial Discovery of a MPB in Sm-substituted BiFeO3 (Sm-BFO)
6.3 Structural Evolution across the MPB in Sm-BFO
6.4 Universal Behavior in RE-Substituted BFO
6.5 Structural Fingerprint of MPB in RE-Substituted BFO
6.6 Chemical-Substitution-Induced Polarization Rotation in BFO
6.7 Concluding Remarks and Future Perspectives
Acknowledgments
References
Chapter 7: Antiferroelectricity in Oxides: a Reexamination
7.1 Introduction
7.2 Definition and Characteristic Properties
7.3 Microscopic Origins of Macroscopic Behavior
7.4 Antiferroelectric Materials: Structure and Properties
7.5 Relation to Alternative Ferroelectric Phases
7.6 Antiferroelectricity in Thin Films
7.7 Properties for Applications
7.8 Prospects
Acknowledgments
References
Part IV: Multiferroics
Chapter 8: Probing Nanoscale Electronic Conduction in Complex Oxides
8.1 Scanning-Probe-Based Transport Measurements
8.2 Domain Wall Conductivity
8.3 Photovoltaic Effects at Domain Walls
8.4 Local Characterization of Doped Oxides and Defects
8.5 Local Electronic Probing of Oxide Interfaces
8.6 Nanoscale Electronic Properties of CMR Manganites
8.7 Future Directions
References
Chapter 9: Multiferroics with Magnetoelectric Coupling
9.1 Introduction: Ferroic Materials
9.2 Principles of Symmetry Analysis
9.3 Magnetoelectric Couplings and a Landau-Like Theory
9.4 Chemical Considerations
9.5 Classification of Multiferroics
9.6 Multiferroic Materials and Mechanisms
9.7 Microscopic Mechanisms of Magnetic Ferroelectricity: Type II Multiferroics
9.8 Domains and Metal–Insulator Transition
Summary
References
Part V: Interfaces and Magnetism
Chapter 10: Device Aspects of the SrTiO3–LaAlO3 Interface; Basic Properties, Mobility, Nanostructuring, and Potential Applications
10.1 Introduction
10.2 The LaAlO3/SrTiO3 Interface: Key Characteristics and Understanding
10.3 Charge-Carrier Mobility
10.4 Micro/Nanostructuring
10.5 Electric Field Gating
10.6 Applications
Acknowledgments
References
Chapter 11: X-Ray Spectroscopic Studies of Conducting Interfaces between Two Insulating Oxides
11.1 Introduction
11.2 Photoemission Measurements of Interfaces
11.3 Interfaces between a Mott Insulator and a Band Insulator: LaAlO3/LaVO3 and LaTiO3/SrTiO3
11.4 Interfaces between Two Band Insulators: LaAlO3/SrTiO3
11.5 Summary
References
Chapter 12: Interfacial Coupling between Oxide Superconductors and Ferromagnets
12.1 Introduction
12.2 Experimental Results
12.3 Materials Considerations
12.4 Conclusions
References
Part VI: Devices and Applications
Chapter 13: Metal-Oxide Nanoparticles for Dye-Sensitized Solar Cells
13.1 TiO2: Polymorphism, Optoelectronic Properties, and Bandgap Engineering
13.2 Principle and Basis of Dye-Sensitized Solar Cell (DSC) Technology
13.3 Progress in TiO2 Engineering for Improved Charge-Collection Efficiency and Light Confinement
13.4 Development of Molecular Sensitizers Suited to TiO2 Optoelectronic Properties
13.5 Development of Redox Mediators
13.6 Conclusions
Acknowledgments
References
Chapter 14: Hybrid Solar Cells from Ordered Nanostructures
14.1 Introduction
14.2 Working Mechanisms of Hybrid Solar Cells
14.3 Nanostructures for Hybrid Solar Cells
14.4 Metal Oxide Modifications
14.5 Conclusion and Outlook
List of Abbreviations
References
Chapter 15: Electric Field Effects in Functional Metal Oxides
15.1 Introduction
15.2 Developments of Gate Dielectrics: from High-k Oxides to Ionic Electrolytes
15.3 Electric-Field-Induced Modulation of Ferromagnetism
15.4 Electrostatic Modulation of Superconductivity
15.5 Transparent Amorphous Metal-Oxide Field-Effect Transistors
15.6 Solution-Processable Transparent Amorphous Metal Oxides for Thin-Film Transistors
15.7 Challenges and Opportunities
References
Chapter 16: Resistive Switchings in Transition-Metal Oxides
16.1 Introduction
16.2 Classification of Current–Voltage Hystereses
16.3 Bipolar Continuous Switching
16.4 Toward Device Applications and Summary
Acknowledgments
References
Index
Related Titles
Collins, T.T., Reynolds, D.D., Litton, C.C. (eds.)
Zinc Oxide Materials for Electronic and Optoelectronic Device Applications
2011
ISBN: 978-0-470-51971-4
Geckeler, K.E., Nishide, H. (eds.)
Advanced Nanomaterials
2010
ISBN: 978-3-527-31794-3
Duò, L., Finazzi, M., Ciccacci, F. (eds.)
Magnetic Properties of Antiferromagnetic Oxide Materials
Surfaces, Interfaces, and Thin Films
2010
ISBN: 978-3-527-40881-8
Bruce, D.D., Bruce, D.W., O'Hare, D., Walton, R.I. (eds.)
Functional Oxides
2010
Print ISBN: 978-0-470-99750-5
Kumar, C.S.S.R. (ed.)
Nanostructured Oxides
2009
ISBN: 978-3-527-32152-0
The Editors
Prof. Satishchandra B. Ogale
National Chemical Laboratory (CSIR-NCL)
Physical & Materials Chemistry
Dr. Homi Bhabha Road, Pashan
411 008 Pune
India
Prof. Thirumalai V. Venkatesan
National University of Singapore
Faculty of Engineering
4 Engineering Drive 3, Block A
117576 Singapore
Singapore
Prof. Mark G. Blamire
University of Cambridge
Department of Materials Science
Pembroke Street
CB2 3QZ Cambridge
United Kingdom
All books published by Wiley-VCH are carefully produced. Nevertheless, authors, editors, and publisher do not warrant the information contained in these books, including this book, to be free of errors. Readers are advised to keep in mind that statements, data, illustrations, procedural details or other items may inadvertently be inaccurate.
Library of Congress Card No.: applied for
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library.
Bibliographic information published by the Deutsche Nationalbibliothek
The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at <http://dnb.d-nb.de>.
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Boschstr. 12, 69469 Weinheim, Germany
All rights reserved (including those of translation into other languages). No part of this book may be reproduced in any form — by photoprinting, microfilm, or any other means — nor transmitted or translated into a machine language without written permission from the publishers. Registered names, trademarks, etc. used in this book, even when not specifically marked as such, are not to be considered unprotected by law.
Print ISBN: 978-3-527-33179-6
ePDF ISBN: 978-3-527-65489-5
ePub ISBN: 978-3-527-65488-8
Mobi ISBN: 978-3-527-65487-1
oBook ISBN: 978-3-527-65486-4
Preface
Complex functional oxides exhibit a fascinating range of behaviors: from colossal magnetoresistance and high-temperature superconductivity to ferroelectricity and multiferroicity. Oxides therefore can potentially provide solutions to many of the technological challenges facing the world, including lower energy consumption devices and the shift to renewable sources of energy. Unlike their semiconducting counterparts such as silicon, GaAs, oxides have more degrees of freedom in terms of the properties at the molecular level, which when exploited could lead to unusual functionalities. For example, the oxygen octahedra in the perovskite oxides can affect the material properties dramatically when the octahedral tilt angles are changed by a variety of means, strain, for example. Indeed, the molecular orbitals of the atoms play a very strong role in determining the properties at oxide interfaces and these could be affected by strain and, electric and magnetic fields. Hence, their functional behavior is intimately coupled to structure and stoichiometry. On one hand, this implies that the oxide properties can be readily tuned by several means, but on the other, it also makes oxides more sensitive to precise processing conditions and strain within device structures than the metals or semiconductors that they might aim to replace. With advances in growth techniques in recent years and in situ monitoring of growth, the field of oxide research is moving steadily to the domain of realistic oxide electronics. In the course of this research, several novel phenomena are being constantly discovered, which are challenging our understanding about the structure–property relationships in these systems. Integration of oxide systems with nonoxide materials such as other classes of inorganic materials (sulfides, nitrides), functional carbon (graphene, CNT), and organics (molecules, polymers) is promising to emerge as a frontier area of research, with significant implications for applications in the key areas of energy, environment, and health. In the field of energy, the ability to tune the position of the conduction and valence bands of the material has a crucial bearing on our ability to use the material in applications such as photocatalysis, water splitting, or CO2 sequestration. Very often in such energy applications, the chemical stability of the material becomes the limiting factor as to whether a material may be deployed or not. Once again, oxides emerge as winners here, as many of the oxides are quite stable in corrosive environments.
The application and understanding of oxides require a multidisciplinary approach and the problems would be best solved by teams of chemists, physicists, materials, and device engineers working together. This book provides an overview of the current state of developments and suggestions as to how the field might move forward. Several eminent experts worldwide and their collaborators/students have contributed to this effort through their chapters. The editors greatly appreciate their time and effort in putting together such excellent chapters without which the book would not have been possible. We believe that this book will prove to be a good reading for young students joining research as well as for young research scientists working in the field. The editors have tried their best to see that several areas of current interest are adequately covered. The intent is to transmit the current excitements in the field to the readership and we hope that the book will serve this purpose.
One of us, SBO, would like to acknowledge the strong support of his family in all his endeavors including this book effort, the parent institute CSIR (Director General Prof. Sameer Brahmachari) and the laboratory CSIR-NCL (Director Dr. Pal and past Director Dr. Sivaram), and students/collaborators. MGB would like to thank the past and present members of the Device Materials Group, University of Cambridge, for their enthusiasm and commitment to novel electronic materials and his wife for support away from the laboratory. TVV would like to thank his team at NUSNNI-NanoCore for their extraordinary dedication to this field of research, his wife for overlooking his not being able to be with her all the time, and his daughter for coming over to Singapore to finish her high schooling here.
Finally, and most importantly, we would like to thank the Wiley publishing team including Dr. Esther Levy with whose invitation this book got initiated and was later followed up by Dr. Gudrun Walter, Ms. Lesley Belfit, and their colleagues. Their patience and pursuance were important for the successful conclusion of this process.
CSIR-NCL, India
Satishchandra B. Ogale
NUS, Singapore
Thirumalai V. Venkatesan
Cambridge, UK
Mark G. Blamire
List of Contributors
Part I
Magnetic Oxides
Chapter 1
Introduction to Magnetic Oxides
J. M. D. Coey, M. Venkatesan, and Hongjun Xu
Oxides are ubiquitous. The Earth's crust and mantle are largely made up of compounds of metal cations and oxygen anions. Looking at the composition of the crust in Figure 1.1, we see that oxygen is the most abundant element and the most common metals are aluminum and silicon. Most rocks are therefore aluminosilicates. The next most abundant element, and the only transition metal other than titanium to feature among the top ten, which account for over 99% of the crust, is iron (Table 1.1). Remarkably, the same electronic configuration, 2p6, is shared by five of the top ten ions, which account for 92% of the atoms in the crust. Usually, only iron, with its two common charge configurations, Fe2+ (3d6) and Fe3+ (3d5), forms ions with a partially filled shell containing electrons of unpaired spin that exhibit a net magnetic moment. At 2.1 at. % (5.7 wt.%), iron is 40 times as abundant as all the other magnetic elements put together; the runners up – manganese, nickel, and cobalt – trail far behind.
Figure 1.1 (a) Abundance of elements in the Earth's crust in atomic % and (b) abundance of magnetic elements.
Table 1.1 Elements in the Earth's Crust (Atomic %)
Ion
Abundance
Configuration
O
2–
60.7
2p
6
Si
4+
20.6
2p
6
Al
3+
6.1
2p
6
Na
+
2.6
2p
6
Fe
2+/3+
2.1
3d
6/5
H
+
2.1
1s
0
no electrons
Ca
2+
1.9
3p
6
Mg
2+
1.8
2p
6
K
+
1.5
3p
6
Ti
4+
0.3
3p
6
For over 20 centuries, up until about 1740 [1], the only useful permanent magnets known to man were lodestones. These prized natural magnetic rocks were largely composed of impure magnetite, the black spinel-structure oxide Fe3O4 with a ferrimagnetic structure that had been magnetized by a fortuitous lightning strike [2]. The other common rock-forming iron oxide is hematite, the reddish corundum-structure sesquioxide αFe2O3. Hematite is also magnetically ordered, in a canted antiferromagnetic structure, but its magnetization is about 200 times weaker than that of magnetite.
The weak remanent magnetism imparted to rocks as they cooled in the Earth's magnetic field has allowed us to read the record of fluctuations of the magnitude and direction of the field at the Earth's surface. The remanence is largely due to segregated nano crystallites of titanomagnetite in the rock [3]. The Earth's field is a precious shield that has protected us from the solar wind and allowed life to develop on our planet over the past 3.5 billion years. We learn from the magnetic record that it has reversed numerous times on a geological timescale. The tectonic movements of the plates were thereby pieced together, leading to the first unified theory of Earth sciences. It was actually a quest to understand the magnetism of rocks and baked clay that motivated Louis Néel to formulate the molecular field theory of ferrimagnetism [4], completing the theory of antiferromagnetism that he had developed in his thesis, which was based on the original molecular field theory of ferromagnetism that his mentor and PhD supervisor Pierre Weiss had formulated in 1906, 30 years previously.
Ferrimagnetic oxides are quite strongly magnetic and insulating when the only iron ions they contain are ferric ions. This was a winning combination at the time because there was a need for magnetic materials to be used at microwave frequencies. Metallic alloys based on iron, or even magnetite itself (which contains both Fe2+ and Fe3+), were unsuitable because of prohibitive eddy-current losses associated with their conductivity. Oxides from the cubic structural families of spinels and garnets, especially nickel–zinc ferrite, manganese–zinc ferrite, and yttrium–iron garnet (YIG), proved to be the champions here.
Another winning combination was ferrimagnetism and strong uniaxial anisotropy. This was the basis of the success of the hexagonal ferrites with the magnetoplumbite structure, which now account for 90% of the tonnage of permanent magnets produced worldwide. Discovered at the Philips laboratory in the Netherlands around 1950, these were the first true permanent magnets, in the sense that their coercivity could exceed their magnetization, which meant that they could remain magnetized whatever be their shape. A nice, square hysteresis loop is the icon of permanent magnetism. Individual magnet ownership has progressed from one or two per person 60 years ago to a 100 times as many today.
Magnetic recording, which demanded less permanently magnetized materials that could be remagnetized in the reverse direction, as required, spurred the next big development. This included the production of acicular (needlelike) particles of the other ferric sesquioxide, spinel-structure γFe2O3, which is known as maghemite. Subsequently, Fe2O3 was surface-doped with cobalt, and then acicular CrO2, the only simple oxide that is a ferromagnetic metal was developed. These micrometer-sized particles made up the magnetic media on hard and floppy disks and tapes used for audio, video, and data recording for several decades. However, the relentless march of progress characterized by the magnetic version of Moore's law, which states that the areal density of recorded information doubles every 18–24 months, has now rendered particulate oxide media obsolete. Thin film metallic media are now used to store everything that is downloaded from the Internet.
Intellectually too, scientific interest in magnetic oxides has waxed and waned. The first Golden age was the 1950s and the 1960s, when the ferrites were explored and their properties optimized [5]. The phenomenology of the exchange interactions among localized electrons in 3d shells was systematically investigated [6]. Discovery of the first family of ferromagnetic oxides, the mixed-valence manganites, also dates from this period [7]. The interest in them revived when the copper oxide high-temperature superconductors were found to have a similar, perovskite-related structure, and the magnetoresistance of thin film mixed-valence manganites was shown to be “colossal”. Ferromagnets have a higher magnetization than ferrimagnets, but the double-exchange mechanism that allows them to order magnetically above room temperature entails electrons hopping among localized cores (3d3 for manganites and 3d5 for magnetite) so that the oxides are conducting and not insulating. The manganites illustrate nicely the progression from studies on bulk ceramics to single crystals to thin films and other nanostructures which marks the historical evolution of research in magnetic oxides.
We may now be entering a new Silver age for oxide research, where multifunctionality, control of defects, interfaces, and thin-film device structures are the new challenges [8]. These ideas are discussed in the later chapters of the book. Here we do the groundwork, by introducing some of the basic ideas and magnitudes relating to structure and properties of magnetic oxides [9, 10]. SI units are used consistently throughout this chapter, but tables of cgs conversions can be found in textbooks on magnetism [11]. The merits of using SI units are compelling. Not only is it possible to check that equations are dimensionally correct, but ideas about the shape of M(H) hysteresis loops, for example, clear when M and H are measured in the same units (ampere per meter).
1.1 Oxide Structures and Crystal Chemistry
Figure 1.2 A close-packed oxygen lattice showing octahedral (blue) and tetrahedral (green) cation sites.
Table 1.2 Ionic Radii of Cations in Oxides
A variety of defects can be found in oxide structures [12]. Point defects include vacant oxygen or metal sites, and metal cations in interstitial sites, which are unoccupied in the perfectly-ordered structure. The former may trap one or two electrons, in which case the defect is known as an F center. Planar defects include grain boundaries in polycrystalline material, and missing planes of oxygens such as those found in the Magnéli phases TinO2n−1, which are based on rutile-structure TiO2. Defects modify the electronic structure and may influence the optical and magnetic properties.
The ionic picture is, of course, an oversimplification. The chemical bond between metal and oxygen has part-ionic and part-covalent character, governed by the electronegativity of the atoms involved. The covalent character is more pronounced in tetrahedral sites and for cations (such as V4+) with a high formal charge state.
Next we briefly present ten representative structures encountered in magnetic oxides. Of course there are many more than ten oxide structures, but these examples serve to illustrate the structural principles and cover the most common materials. We refer to them by the name of a mineral type, which is not necessarily an oxide. In each case a picture of the structure is included in Figure 1.3, and structural information on specific nonmagnetic oxides is provided in Table 1.3.
Figure 1.3 Some common oxide structure types (a) halite, (b) wurzite, (c) spinel, (d) corundum, (e) rutile, (f) perovskite, (g) garnet (h) magnetoplumbite, and (i) pyrochlore.
Table 1.3 Properties of Nonmagnetic Oxides
We conclude this section with a few remarks on the electronic structure of oxides. Normally, the oxides are insulators, with a valence band derived from the filled 2p6 oxygen levels, and a conduction band derived from unoccupied metal orbitals. Examples where no unpaired transition- element electrons are present are MgO, ZnO, MgAl2O4, Al2O3, and TiO2. This is also a list of commonly-used substrates for thin-film growth of the magnetic counterparts. Properties of these baseline nonmagnetic oxides are given in Table 1.3. All have a wide bandgap and are optically transparent as single crystals and the powders are white.
The structure more frequently encountered when a transition metal is present in the structure, is for the 3dn level to fall in the gap. Narrow d bands are formed by 3d(M) – 2p(O) hybridization, which frequently leads to optical absorption in the visible – leading to the observed colors (red for hematite, green for YIG, and black for magnetite) [14].
When the 3d level lies in the gap and the energy of the excitation 2(dn)→ dn−1 + dn+1 is less than the 3d or 4d bandwidth, the oxide is a d-band metal. Examples include CrO2 and SrRuO3. However, if the energy of this excitation exceeds the d bandwidth, the material may be a Mott insulator, but when 3dn level lies below the top of the 2p band, the relevant low-energy electronic excitation is p6dn→p5 + dn+1. According to whether this is less than or greater than the d bandwidth, the material is a p/d metal or a charge-transfer insulator [13] (Figure 1.4).
Figure 1.4 Electronic structure of a 3d oxide for an early transition metal (a) and a late transition metal (b). The low-energy d–d and p–d electronic excitations are shown.
There are a few ferromagnetic metal oxides where the 3d bands are spin split to the extent that a spin gap appears in either the ↑ or ↓ subband, and the electrons at the Fermi level are completely spin-polarized. Such materials are known as half metals. Stoichiometric half-metals exhibit a spin moment per unit cell which is an integral number of Bohr magnetons.
Oxides may not be precisely stoichiometric, but small deviations often have little influence on the electronic properties because the electrons or holes create a distortion of the lattice, forming immobile polarons that have a large effective mass.
1.2 Oxide Growth
1.2.1 Polycrystalline Materials
The easiest methods of oxide growth involve physical or chemical reactions that yield bulk polycrystalline material.
1.2.1.1 Precipitation
Precipitation is a widely used wet-chemical technique for synthesizing ultrafine ceramic powders of simple or complex oxides with a narrow particle size distribution [15]. The method avoids complex steps such as refluxing of alkoxides, and it is faster than other techniques. Precipitation or coprecipitation from aqueous salt solutions (nitrate, sulfate, chloride, perchlorate, etc.) by fine control of the pH by adding NaOH or NH4OH solutions yields a poorly-crystallized hydroxide or an intimate mixture of the hydrated oxides. The particle size of the precipitate is strongly dependent on pH and the molarity of the precursor solution; it is usually converted to oxide by heating in air. The coprecipitation method offers simple and rapid preparation of complex oxides with control of particle size and good overall homogeneity.
1.2.1.2 Sol–Gel
1.2.1.3 Solid-State Reaction
Solid-state diffusion is often used to make complex oxides from an intimate mixture of microcrystalline precursor powders [17]. Neither a solvent medium nor controlled vapor-phase interactions are involved. Sometimes known as shake ‘n’ bake, it is the most widely used method for the preparation of complex oxides from a mixture of solid starting materials. Interdiffusion takes place at an appreciable rate at high pressure and/or high temperature (1000–1500 °C for several days). The rate of the solid-state reaction also depends on the ambient atmosphere, structural properties of the reactants, their surface area, reactivity, and the thermodynamic free energy change associated with the reaction. Several cycles of grinding the powder and refiring at progressively higher temperatures are usually necessary to achieve a pure phase. Solid-state synthesis is used to create ceramic powders or dense polycrystalline sintered masses. It may be advantageous to use precursors such as carbonates or oxalates, which decompose at low temperature to give fine-grained oxides.
1.2.1.4 Combustion Synthesis
The term covers a group of methods related to solid-state synthesis and coprecipitation where a fuel is incorporated in the mixture to be fired [18]. An exothermic reaction promotes uniform heating of the solid mass, and the production of oxide powders with a narrow crystallite size distribution in the range 10–100 nm. The solution method typically involves dissolving the precursors such as metal nitrates (the oxidizer) and a compound such as urea or glycine (the fuel) in water, followed by a self-sustaining reaction between the dried constituents, initiated at a relatively low temperature of 500 °C. Microwave irradiation may be used to modify the reaction conditions using a simple microwave oven.
1.2.2 Single Crystals
Single crystals, which may range in size from micrometers to centimeters, are needed for detailed characterization of the physical properties including the elementary excitations (magnons, phonons, and excitons). The key to growing them is control of the nucleation process.
1.2.2.1 Bridgeman Method
The Bridgeman method is a slow, controlled freezing process taking place under liquid–solid equilibrium conditions by allowing the solid–liquid interface to move slowly until the entire molten charge is solidified [19, 20]. The growth takes place in a temperature gradient, and the idea is to create a single nucleus from which a single crystal will grow. The method involves melting polycrystalline material in a crucible with a pointed end and slowly cooling it from the bottom where a seed crystal nucleates at the tip. The crystal grows progressively up the length of the crucible. Either the crucible itself or a tube furnace with a temperature gradient can be moved. Compared to other growth methods, the Bridgeman method is rather simple but it cannot be applied if the system decomposes before it melts, or to oxides of elements with a high vapor pressure.
1.2.2.2 Czochralski Method
The Czochralski process is widely used to grow large single-crystal boules of semiconductors such as silicon, germanium, and gallium arsenide, but it can also be applied for many oxide crystals including Al2O3 (sapphire), LaAlO3 (LAO), Y3Fe5O12 (YIG), Y3Al5O12 (YAG), and Gd3Ga5O12 (GGG) [21]. Single-crystal material is pulled out of a slightly undercooled melt by dipping in a single-crystal seed and then slowly withdrawing it. The seed crystal rod is rotated at the same time as it is drawn out, and by precisely controlling the atmosphere, temperature gradient, rate of pulling, and speed of rotation, it is possible to extract a large, single-crystal oxide boule from the melt, which may then be sliced and polished to make substrates for thin film growth of other oxide materials.
1.2.2.3 Zone Melting (Image Furnace)
In the float-zone technique, the sample is in the form of a vertical polycrystalline rod, clamped only at its ends, a short segment of which is melted by a local heating [22]. The molten zone is suspended as a drop between the two solid parts of the rod, and it is moved along the rod by slow motion of either the heater or the rod itself. The optical floating zone technique, which makes use of an infrared image furnace, has been extensively utilized to grow single crystals of oxides. Early designs had one or two mirrors, but now, four mirrors are generally used to obtain more uniform sample heating. Ellipsoidal mirrors are used to focus the light from a halogen or xenon lamp onto the sample to produce the molten zone, making the technique suitable for both conducting and nonconducting materials. Optical heating is particularly convenient and efficient for oxides that absorb easily in the infrared. The sample is protected from its environment by a large diameter, clear quartz tube, which prevents evaporated material from settling on the mirror and allows control of the growth atmosphere and gas pressure around the growing crystal [23]. Optimizing crystallization rate, atmosphere, gas pressure, and temperature is the key to achieving stable growth and, good crystal quality. This technique is commonly used to produce pyrochlore, perovskite, and double-perovskite single crystals. Zone melting also helps to purify the oxide.
1.2.2.4 Flux Method
The flux method involves crystal growth by slow cooling of a high-temperature solution. It is suitable for growing crystals of incongruently-melting compounds, but virtually any stable oxide may be grown from a suitable solvent. The flux and the oxide or its constituents are melted in a platinum crucible, which is then cooled extremely slowly. Times of order a month may be required to complete the crystal growth cycle. Oxides (B2O3, Bi2O3, BaO, and PbO), hydroxides (KOH and NaOH), or halides such as PbF2 can be used as solvents [24]. However, eutectics, found in binary (PbO–PbF2, Li2O–MoO3, Li2O–B2O3, etc.) or ternary diagrams, are generally preferred owing to their low temperature of melting and low viscosity. Flux growth is useful whenever the melting temperature is high and when the vapour pressure at the melting point is elevated. The main disadvantage of the technique is the low-growth rate, more than 100 times slower than for crystals pulled from the melt. Crystals may take weeks to grow.
1.2.2.5 Chemical Vapor Reactions
Transport of material in the gas phase is used to grow crystals in a sealed quartz tube that is placed in a temperature gradient in a tube furnace [25]. A powder of the oxide to be grown is included with a transport agent such as I2 or TiCl2 with which it can react. When the reaction is exothermic, the oxide is transported from the cool zone of the furnace to the hot zone where the compound decomposes and the oxide crystals grow. The temperature gradient must be carefully controlled, and the process may take several days, but the quality of the small oxide crystals is often excellent.
1.2.3 Thin Films
Figure 1.5 Methods for preparing thin films by a vapor or plasma condensing on a substrate: (a) thermal evaporation, (b) e-beam evaporation, (c) pulsed-laser deposition, and (d) sputtering.
1.2.3.1 Physical Methods
Thermal Evaporation
Resistive evaporation is a commonly-used vacuum deposition process in which electrical energy is used to heat a boat containing the charge or a filament that heats the material to be deposited up to the point of evaporation [27, 28]. The vapour condenses in the form of a thin film on the cold substrate surface. The method is restricted to materials with moderately low melting points to avoid contamination by the boat, which is usually made of graphite, molybdenum, or tungsten.
E-Beam Evaporation
Electron beam (e-beam) evaporation is a variant that is used, both for research and on an industrial scale, for making thin film coatings [28]. The technique involves bombarding a target of the material to be evaporated with a beam of high-energy electrons that may be swept over the surface of the target in a specific pattern with the help of beam focussing coils. The target is placed either directly in a water-cooled crucible or in a crucible liner that is made of a different material such as graphite or tungsten, which has a higher melting point than the target and does not form an alloy with it. As a result of the localized melting of the target, the material evaporates and is transported to a substrate. Another geometry utilizes a conducting target in the shape of a thin rod and the e-beam is electrostatically accelerated toward the end of the rod. As the material is evaporated, the rod is fed manually or automatically to keep a constant deposition rate. This geometry is used for materials with a very high melting point such as molybdenum or tungsten. E-beam evaporation can be used in a reactive environment to make thin films of oxides by evaporating a metallic target in the presence of reactive gas. In our case, the oxygen is either directly released into the evaporation chamber or introduced from a low-energy divergent-beam plasma ion source that is directed at the substrate. The latter technique, also known as ion-beam-assisted deposition (IBAD), produces more stoichiometric and better quality oxide films. Very high deposition rates can be achieved by e-beam evaporation compared to sputtering, which makes it advantageous for coating thick films. This technique is widely used to produce SiO2, Al2O3, and some transition-metal oxides (TiO2, HfO2, and ZrO2).
Molecular Beam Epitaxy (MBE)
Molecular beam epitaxy (MBE) is a sophisticated version of vacuum evaporation [29]; it is a method of laying down layers of materials a few atoms thick in ultrahigh vacuum (UHV). Molecular beams of the constituent elements are generated from heated sources and travel without scattering to a substrate where they combine to form an epitaxial film. The most common type of MBE source is the effusion cell (K-cell). The growth rate depends on the flux of material in the molecular beams, which can be controlled by the evaporation rate and, most importantly, switched on and off with shutters in a fraction of the time required to grow one monolayer. Typical growth rates are a monolayer per second or a micrometer per hour. MBE can produce high-quality layers with very abrupt interfaces and good control of thickness, doping, and composition. Because of the high degree of control, MBE is a valuable tool for development of sophisticated electronic and optoelectronic devices, but it is not used suited for industrial production.
Sputtering
Sputtering is the preferred industrial thin film vacuum deposition technique, but it is also widely used in research laboratories. Sputtered films exhibit excellent, reproducibility, uniformity, density, purity and adhesion. It is possible to make oxides, nitrites, and other compounds of precise composition by reactive sputtering from metal targets [28, 30]. In dc sputtering, substrates are placed in the vacuum chamber, and it is evacuated to high vacuum before a low pressure (0.05–1 Pa) of the process gas, usually argon, is introduced. Sputtering starts when a negative potential of a few hundred volts is applied to the target material to be deposited, causing a plasma or glow discharge. Positively charged Ar+ ions generated in the plasma collide with the negatively biased target. The momentum transfer ejects atomic-scale particles from the target, which traverse the chamber and are deposited as a thin film on the surface of the substrate. A magnetic field is usually created near the target surface by means of an arrangement of permanent magnets, known as a magnetron, in order to improve the ionization efficiency. Oxygen is mixed with the argon sputtering gas to produce oxides from metal targets. Alternatively, to make oxide or other insulating films directly, the radio-frequency method of rf sputtering is employed. Here the power supply commonly operates at 13.56 MHz. For part of the cycle, Ar ions bombard the target; for the rest of the cycle, electrons neutralize the build up of positive charge. Electrons also ionize the argon to create the plasma. Sputtering systems often have multiple targets, which permit the fabrication of complex thin film stacks used for spin electronic applications. An argon pressure of 0.02 Pa is usually sufficient to maintain a radio-frequency discharge.
Pulsed Laser Deposition (PLD)
A problem with the method is that liquid droplets or particulates may contaminate the plume and settle on the surface of the film. This can be avoided by off-axis deposition, where the substrate is parallel to the plume. Alternatively, the droplets may also be trapped by rapidly rotating mechanical filters, which feature in the PLD tools that are being developed for the industrial production of ferroelectric films.
1.2.3.2 Chemical Methods
Chemical Vapor Deposition (CVD)
The growth of thin films by chemical vapor deposition (CVD) is an industrially significant process with good stability and reproducibility, which is used in a wide array of applications. CVD involves depositing a solid film from a gaseous molecular precursor [27]. Different energy sources, precursor gases, and substrates are used, depending on the desired product, but the precursors must be volatile, yet stable enough to be able to be delivered to the reactor where the volatilized precursor (such as silane, an organometallic, or a metal coordination complex) is passed over a heated substrate. Thermal decomposition of the precursor produces a thin film deposit, and ideally, the ligands associated with the precursor are cleanly lost to the gas phase as reaction products. Pressure and temperature are the important variables.
Chemical Vapor Transport
The method, akin to that used to grow single crystals, entails the reversible conversion of nonvolatile elements and chemical compounds into volatile derivatives [25]. The volatile derivative migrates throughout a sealed reactor, typically a sealed, evacuated glass tube heated in a tube furnace. Because the tube is in a temperature gradient, the volatile derivative reverts to the parent solid, which is deposited as a thin film on the substrate and the transport agent is released at the opposite end of the tube to where it originated.
Atomic Layer Deposition (ALD)
Atomic layer deposition (ALD) is based on sequential, self-limiting surface chemical reaction [32]. This unique growth technique can provide atomic layer control and allow ultrathin conformal films to be deposited on very high aspect ratio structures. ALD deposits films using pulses of gas that produce one atomic layer at a time. Within fairly wide process windows, the deposited film thickness is only dependent on the number of deposition cycles providing extremely high uniformity and thickness control. ALD reactions are typically carried out in the range 200–400 °C. In particular, ALD is currently used in the semiconductor industry for high-k gate dielectrics (HfO2, ZrO2).
Dip Coating
This is the main wet-chemical method to produce wide variety of oxide thin films such as ZnO, SnO2, TiO2, and (Sn1−xInx)O2 (ITO). A substrate is dipped into a liquid coating solution of the chemical precursor and then gently withdrawn at a controlled rate. The thickness is determined by the balance of forces at the stagnation point on the liquid surface. A faster withdrawal rate pulls more fluid up onto the surface of the substrate before it has time to flow back into the solution. The thickness is also affected by fluid viscosity, density, and surface tension. Finally, the coating is cured by a conventional thermal treatment, or else UV, or IR irradiation.
Spray Pyrolysis
A water-based precursor solution is sprayed through a nozzle onto a substrate where the atomized solution is dried and the metal oxide film is formed. It provides an easy way to dope any element in a required ratio through the solution medium. This method is convenient for preparing pinhole-free, homogenous, smooth thin films of oxides such as TiO2 with controllable thickness. Spray pyrolysis can also be used to produce oxide powders.
1.3 Magnetic Properties of 3d and 4f Ions
1.1
This quantity, equal to 9.27 × 10−24 A m2, is the fundamental unit of atomic-scale magnetism.
1.2
From the electron's point of view, it sees the nucleus orbiting around it, creating a magnetic field proportional to ml, which interacts with the electronic spin moment. The celebrated spin–orbit interaction that is responsible for much that is important and useful in magnetism is represented by the Hamiltonian
1.3
The magnetic moment of the free ion can be written as gμBJ, where the g-factor
1.4
Two examples:
1.5
and
1.6
The above discussion concerned only the magnetism of free atoms or ions. About two-thirds of the atoms in the periodic table, and their isoelectronic free ions, have unpaired electrons and a net magnetic moment. However, when the ions are placed in the crystalline environment of a solid oxide, there are some drastic changes due to the influence of the oxygen neighbors. This is the crystal-field or ligand-field interaction. For 3d and 4d ions, it is stronger than the spin–orbit interaction, whereas for the 4f ions, which are well shielded from the outside world by the large filled 5s and 5p shells, the crystal field is screened so that it is only a perturbation on 0 + so. (Table 1.4). For the rare earths, J is therefore a good quantum number, and the main effect of cf is to introduce single-ion anisotropy, whereby the ion may have one or more easy axes of magnetization in the crystal lattice.
Table 1.4 Relative Magnitudes of Energy Terms for 3d and 4f Ions
1.7
where N0 is Avogadro's number. In numerical terms,
1.8
We focus first on the 3d ions, in order to explain how the crystal field influences both the electronic structure and magnetic properties of oxides. We begin with the one-electron model [33, 34], which ignores the onsite 3d–3d Coulomb interactions. This is a good approximation for d1, d4, d6, and d9 ions when Hund's first rule applies, as these ions have a single electron or hole outside an empty, a half-filled, or a filled shell.
When the ion is found in an undistorted tetrahedral or octahedral site in an oxide, the local environment has cubic symmetry unlike the spherical symmetry of a free ion. The free ion eigenfunctions ψ0, ψ±1, and ψ±2, where the subscripts denote ml, must be replaced by suitable linear combinations, which reflect the cubic symmetry. They are
1.9
This basis set of d orbitals is illustrated in Figure 1.6.
Figure 1.6 The 3d orbitals for a free ion (a) and the orbitals for the ion in a cubic crystal field (b).
We consider the influence of a crystal field due to an octahedron or tetrahedron of oxygen neighbors on the 3d wave functions. Considering the disposition of the oxygen orbitals in the octahedron, it is obvious that the xy, yz, and zx orbitals are degenerate – they are labeled as t2g orbitals. It is less obvious that the x2 − y2 and z2 orbitals, labeled as eg, are degenerate, but they are clearly higher in energy because the electron density is maximum near the negatively charged anions. The crystal-field splitting is illustrated in Figure 1.7.
Figure 1.7 Crystal-field splitting for an ion in sites with octahedral, tetrahedral, and cubic coordination.
In the tetrahedral site, the splitting is reversed. The e orbitals are lower and the t orbitals are higher (the “g” subscript is dropped when there is no center of symmetry). The splitting in cubic coordination is similar (Figure 1.7).
The splitting cannot be entirely explained in terms of the electrostatic potential due to oxygen anions; about half is attributable to the different overlap of the t and e orbitals with the lower-lying 2p oxygen orbitals which introduces, in an octahedral site for example, greater bonding–antibonding splitting for the 2p-eg σ-bonds than the 2p-t2g π-bonds. This is known as the ligand-field effect. The magnitude of Δoct is of order 1 eV (Figure 1.8).
Figure 1.8 Illustration of the splitting of the t2g and eg orbitals in an octahedral environment. Δcf is the crystal-field splitting calculated in a point charge model. The extra ligand-field splitting due to the hybridization of 3d and sp orbitals gives the total splitting Δ.
The influence of a distortion, trigonal, tetragonal, or monoclinic, is to raise the degeneracy of the one-electron energy levels, as shown in Figure 1.9 for tetragonal distortion. The splitting preserves the center of gravity of the sets of orbitals.
Figure 1.9 Influence of a tetragonal elongation or flattening of an octahedral site on the one-electron energy levels.
