Ferroic Materials For Smart Systems From Fundamentals To Device Applications Dai Full Chapter

Ferroic materials for smart systems
from fundamentals to device

applications Dai
Visit to download the full and correct content document:
https://ebookmass.com/product/ferroic-materials-for-smart-systems-from-fundamental
s-to-device-applications-dai/
Ferroic Materials for Smart Systems
Ferroic Materials for Smart Systems
From Fundamentals to Device Applications
Jiyan Dai
Author All books published by Wiley-VCH
are carefully produced. Nevertheless,
Prof. Jiyan Dai authors, editors, and publisher do not
The Hong Kong Polytechnic University warrant the information contained in
Department of Applied Physics these books, including this book, to
Hung Hom be free of errors. Readers are advised
Kowloon to keep in mind that statements, data,
Hong Kong illustrations, procedural details or other
items may inadvertently be inaccurate.
Cover
© antoniokhr/Getty Images; Library of Congress Card No.:
(graph) Courtesy of Professor Jiyan Dai, applied for
The Hong Kong Polytechnic Univeristy
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>.
© 2020 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: 978-3-527-34476-5

ePDF ISBN: 978-3-527-81534-0
ePub ISBN: 978-3-527-81537-1
oBook ISBN: 978-3-527-81538-8
Typesetting SPi Global, Chennai, India

Printing and Binding
Printed on acid-free paper
10 9 8 7 6 5 4 3 2 1
v
Contents
1 General Introduction: Smart Materials, Sensors, and

Actuators 1
1.1 Smart System 2
1.2 Device Application of Ferroelectric Materials 5
1.2.1 Piezoelectric Device Applications 6
1.2.2 Infrared Sensor 7
1.2.3 Ferroelectric RAM (FeRAM) 8
1.3 Device Application of Ferromagnetic Materials 9
1.3.1 Spin-Transfer Torque Memory 9
1.3.2 Magnetic Field Sensor Based on Multiferroic Device 9
1.4 Ferroelastic Material and Device Application 10
1.5 Scope of This Book 12
References 13
2 Introduction to Ferroelectrics 15
2.1 What Is Ferroelectrics? 15
2.1.1 P–E Loop 15
2.1.2 Relationships Between Dielectric, Piezoelectric, Pyroelectric, and
Ferroelectric 16
2.1.2.1 Ferroelectric–Dielectric 16
2.1.2.2 Ferroelectric–Piezoelectric 17
2.1.2.3 Ferroelectric–Pyroelectric 18
2.2 Origin of Ferroelectrics 18
2.2.1 Structure-Induced Phase Change from Paraelectric to
Ferroelectric 18
2.2.2 Soft Phonon Mode 19
2.3 Theory of Ferroelectric Phase Transition 21
2.3.1 Landau Free Energy and Curie–Weiss Law 21
2.3.2 Landau Theory of First-Order Phase Transition 23
2.3.3 Landau Theory of a Second-Order Phase Transition 26
2.4 Ferroelectric Domains and Domain Switching 28
2.4.1 Domain Structure 28
2.4.2 Ferroelectric Switching 28
2.5 Ferroelectric Materials 29
vi Contents
2.5.1 From BaTiO3 to SrTiO3 29

2.5.2 From PbTiO3 to PbZrO3 32
2.5.3 Antiferroelectric PbZrO3 33
2.5.3.1 Pb(Zrx Ti1−x )O3 (PZT) 35
2.5.4 Relaxor Ferroelectrics 36
2.5.4.1 Relaxor Ferroelectrics: PMN-xPT Single Crystal 37
2.5.4.2 Polar Nano Regions 38
2.5.4.3 Morphotropic Phase Boundary (MPB) of PMN-PT Crystal 40
2.6 Ferroelectric Domain and Phase Field Calculation 41
References 42
3 Device Applications of Ferroelectrics 47

3.1 Ferroelectric Random-Access Memory 47
3.2 Ferroelectric Tunneling Non-volatile Memory 50
3.2.1 Tunneling Models 51
3.2.2 Metal–Ferroelectric–Semiconductor Tunnel Junction 55
3.2.3 Ferroelectric Tunneling Memristor 56
3.2.4 Strain Modulation to Ferroelectric Memory 57
3.3 Pyroelectric Effect and Infrared Sensor Application 58
3.3.1 Pyroelectric Coefficient 58
3.3.2 Pyroelectric Infrared Sensor 59
3.3.3 Pyroelectric Figures of Merit 60
3.4 Application in Microwave Device 63
3.5 Ferroelectric Photovoltaics 65
3.6 Electrocaloric Effect 67
References 68
4 Ferroelectric Characterizations 73
4.1 P–E Loop Measurement 73
4.2 Temperature-Dependent Dielectric Permittivity Measurement 76
4.3 Piezoresponse Force Microscopy (PFM) 77
4.3.1 Imaging Mechanism of PFM 77
4.3.2 Out-of-plane Polarization (OPP) and In-plane Polarization (IPP)
PFM 80
4.3.2.1 Electrostatic Force in PFM 83
4.3.2.2 Perspectives of PFM Technique 84
4.4 Structural Characterization 86
4.5 Domain Imaging and Polarization Mapping by Transmission Electron
Microscopy 87
4.5.1 Selected Area Electron Diffraction (SAED) 88
4.5.2 Convergent Beam Electron Diffraction (CBED) for Tetragonality
Measurement 91
References 92
5 Recent Advances in Ferroelectric Research 95

5.1 Size Limit of Ferroelectricity 95
5.2 Ferroelectricity in Emerging 2D Materials 96
Contents vii
5.3 Ferroelectric Vortex 99

5.4 Molecular Ferroelectrics 104
5.5 Ferroelectricity in HfO2 and ZrO2 Fluorite Oxide Thin Films 106
5.6 Ferroic Properties in Hybrid Perovskites 114
References 117
6 Piezoelectric Effect: Basic Theory 123

6.1 General Introduction to Piezoelectric Effect 123
6.2 Piezoelectric Constant Measurement 124
6.2.1 Piezoelectric Charge Constant 125
6.2.2 Piezoelectric Voltage Constant 126
6.2.3 Dielectric Permittivity 127
6.2.4 Young’s Modulus (Elastic Stiffness) 127
6.2.5 Elastic Compliance 127
6.2.6 Electromechanical Coupling Factor 128
6.2.6.1 How to Measure Electromechanical Coupling Factor? 129
6.3 Equivalent Circuit 132
6.4 Characterization of Piezoelectric Resonator Based on a Resonance
Technique 135
6.4.1 Length Extensional Mode of a Rod 135
6.4.2 Extensional Vibration Mode of a Long Plate 138
6.4.3 Thickness Shear Mode of a Thin Plate 139
6.4.4 Thickness Mode of a Thin Disc/Plate 140
6.4.5 Radial Mode in a Thin Disc 141
6.4.6 Mechanical Quality Factor 141
References 141
7 Piezoelectric Devices 143

7.1 Piezoelectric Ultrasonic Transducers 143
7.1.1 Structure of Ultrasonic Transducers 143
7.1.2 Theoretical Models of Ultrasonic Transducer (KLM Model) 145
7.1.3 Characterization of Ultrasonic Transducers 147
7.1.3.1 Bandwidth (BW) 147
7.1.3.2 Sensitivity of the Transducer 148
7.1.3.3 Resolution 148
7.1.4 Types of Ultrasonic Transducers 149
7.1.4.1 Medical Application 149
7.1.5 Piezoelectric Film Application in Ultrasound Transducers 149
7.1.6 Challenges and Trend of Developing New Advanced Ultrasound
Transducers 150
7.2 Ultrasonic Motor 150
7.2.1 Terminologies 151
7.2.2 Design of USM 153
7.3 Surface Acoustics Wave Devices 154
7.3.1 Interdigital Electrode in SAW Device 155
7.3.2 Acoustic Wave 155
7.3.3 Piezoelectric Property Considerations for SAW Devices 157
viii Contents
7.3.4 Characterization of SAW Devices 159

7.3.5 Lead-Free Piezoelectric Materials 161
References 163
8 Ferromagnetics: From Material to Device 165

8.1 General Introduction to Ferromagnetics 165
8.2 Ferromagnetic Phase Transition: Landau Free-Energy Theory 168
8.3 Domain and Domain Wall 169
8.4 Magnetoresistance Effect and Device 171
8.4.1 Anisotropic Magnetoresistance (AMR) 171
8.4.2 Giant Magnetoresistance (GMR) 172
8.4.3 Colossal Magnetoresistance (CMR) 175
8.4.4 Tunneling Magnetoresistance (TMR) 176
8.4.4.1 Spin-Transfer Torque Random-Access Memory (STT-RAM) 177
8.5 Magnetostrictive Effect and Device Applications 178
8.5.1 Magnetostrictive Properties of Terfenol-D 180
8.5.2 Magnetostrictive Ultrasonic Transducer 183
8.5.3 Magnetoelastic Effect 184
8.5.3.1 Magnetomechanical Strain Gauge 185
8.6 Characterizations of Ferromagnetism 186
8.6.1 Vibrating Sample Magnetometer (VSM) 186
8.6.2 Superconducting Quantum Interference Device (SQUID) 187
8.6.3 Magnetic Force Microscopy (MFM) 188
8.6.4 Magneto-Optical Kerr Effect (MOKE) 189
8.7 Hall Effect 191
8.7.1 Ordinary Hall Effect 191
8.7.2 Anomalous Hall Effect 191
8.7.3 Spin Hall Effect 192
References 193
9 Multiferroics: Single Phase and Composites 197

9.1 Introduction on Multiferroic 197
9.2 Magnetoelectric Effect 199
9.3 Why Are There so Few Magnetic Ferroelectrics? 199
9.4 Single Phase Multiferroic Materials 200
9.4.1 Switching Mechanism in BFO Films 204
9.5 ME Composite Materials 205
9.6 Modeling the Interfacial Coupling in Multilayered
ME Thin Film 207
9.6.1 PZT/CFO Multilayered Heterostructures 207
9.6.2 Ferroelectric Properties of PZT/CFO Multilayers 209
References 212
10 217
Device Application of Multiferroics
10.1 ME Composite Devices 217
10.1.1 Effect of Preload Stress 221
10.2 Memory Devices Based on Multiferroic Thin Films 223
Contents ix
10.3 Memory Devices Based on Multiferroic Tunneling 224

References 229
11 Ferroelasticity and Shape Memory Alloy 231

11.1 Shape Memory Alloy 231
11.1.1 SMA Phase Change Mechanism 232
11.1.2 Nonlinearity in SMA 233
11.1.3 One-Way and Two-Way Shape Memory Effect 235
11.1.4 Superelastic Effect (SE) 235
11.1.5 Application Examples of SMAs 236
11.2 Ferromagnetic Shape Memory Alloys 237
11.2.1 Formation of Twin Variants 238
11.2.2 Challenges for Ni–Mn–Ga SMA 242
11.2.3 Device Application of MSMA 243
References 244
Index 247
1
General Introduction: Smart Materials, Sensors, and

Actuators
The early twenty-first century has foreseen acceleration of innovations in

robotics and automations as well as artificial intelligence (AI), where sensors/
transducers and smart materials play very important roles. The concept of AI
has been around since the late 1950s; however, it’s only since the first decade of
the twenty-first century that excitement about it has really begun to grow due
to the ability of fast computation and abundant size of memory devices. A very
successful demonstration of AI is Google’s AlphaGo, which is the first computer
program to defeat a professional human Go player (see Figure 1.1). Another
successful application of AI is the unmanned vehicles and aircrafts where large
number of sensors and actuators are used.
One may ask, what is the relation between the key word “ferroic materials” of
this book and the mentioned robotics, automations, and AI? The answer is that
these smart and intelligence systems rely on large amount of data from sensors
and memories for machine learning and actuators for close-looped feedback con-
trol systems; and among these sensors, actuators, and memories, ferroic materials
play very important roles.
For example, the piezoelectric property of a ferroelectric material (one of the
most typical ferroic materials) can be used for ultrasound sensors to detect dis-
tance of your car from a wall for auto parking system. A ferroelectric material can
be used as the functional element for many kinds of sensors from pressure sensor
to acceleration sensor, infrared sensor, etc. Beyond that, a ferroelectric polariza-
tion and its switching can also be used in memory devices such as ferroelectric
random-access memory (FeRAM) where the ferroelectric layer acts as gate insu-
lator in a field-effect transistor (FET) structure. Ferroelectric tunneling-based
resistive random-access memory (RRAM) has also been demonstrated, and such
ferroelectric-based memory has been shown to be able to perform as an artificial
synapse. More interestingly, artificial neural networks (ANNs) based on these
ferroic synapses can realize brain-like computing and AI functions such as image
recognition. As shown in Figure 1.2, synapses with BiFeO3 (BFO) ferroelectric
layer has been successfully demonstrated.
This book will tell you fundamentals and characterization methods of ferroic
materials, physics, and technologies behind ferroic device design and applica-
tions as well as their recent advances.
Ferroic Materials for Smart Systems: From Fundamentals to Device Applications,

First Edition. Jiyan Dai.
© 2020 Wiley-VCH Verlag GmbH & Co. KGaA. Published 2020 by Wiley-VCH Verlag GmbH & Co. KGaA.
2 1 General Introduction: Smart Materials, Sensors, and Actuators
Figure 1.1 AI beats human chess player.
Synapse
Nucleus
Axon
Axon
Memristor Input
Co
BFO Synapses
CCMO To pre-neuron
To post-neuron Output
Figure 1.2 A cross-bar structure of synapse and artificial neuron networks based on a
cross-bar structure. Ferroelectric thin films such as BiFeO3 can be used as the junction material.
Source: Adapted from Boyn et al. (2017).
1.1 Smart System

A smart system, such as a self-driving car or remote-control aircraft, is a sys-
tem that relies on sensors and actuators to realize instant feedback of controlled
variables (CVs) such as speed, height, etc. The basic component of a smart sys-
tem usually contains sensors, actuators, and control system. An intelligent smart
system needs large amount of data processing and memories, while ferromag-
netic and ferroelectric materials have been implemented in realizing non-volatile
memristors. Beyond that, memories based on ferroelectric thin films may also
find application in electronic synapses as building blocks toward building ANN.
1.1 Smart System 3
Proportional
+ Integral Work
Sensor
Derivative
Actuator
(a) (b)
Figure 1.3 (a) Photo of a remote control copter and (b) diagram of a PID feedback control
system where sensors and actuators are implemented.
As an example, a smart system of remote control copter relying on

proportional–integral–derivative (PID) feedback control system is shown
in Figure 1.3. PID is a three-order feedback control system that has been widely
used in auto driving vehicles and auto-pilot airplanes to make the dynamic
system operate smoothly or being stable during video imaging. Equation (1.1)
illustrates the mechanism of PID control where three terms of proportional gain
(P), integral gain (I), and derivative feedback (D) can provide instant response to
cure the error (E) between the set point (speed of vehicle or height of the copter)
and controlled variable (CV):
t
dCV (t)
MV (t) = PE(t) + I E(t ′ )dt ′ − D (1.1)
∫0 dt
In this feedback control system, we can find applications of ferroic materials, for
example, piezoelectric-based gyroscope, surface acoustic wave device for wireless
communication, and ferroelectric-based infrared detector. In gyroscope, rotation
and acceleration can be sensed by measuring induced voltage generated by piezo-
electric effect; a surface acoustic wave device based on piezoelectric effect is used
for communication band selection; and a ferroelectric-based infrared detector
can be used as an intruder sensor making the copter able to find people for res-
cue mission. Figure 1.4 shows the finite element modeling (FEM) simulation of
three resonant motions in a Pb(Zrx, Ti1−x )O3 (PZT)-based gyroscope and photo
of a fabricated gyroscope (Chang and Chen 2017).
Sensors: Devices that can “sense” a change in some physical characteristics and
perform an electrical input function are commonly called sensors. For example,
a strain sensor converts mechanical strain into electrical signal.
Actuators: Devices that perform an output function are generally called actua-
tors. An actuator can be utilized to control some external moduli or output
mechanical movement such as ultrasonic wave. For example, an atomic force
microscope (AFM) uses piezoelectric actuators to realize scanning along three
directions.
Transducers: Both sensors and actuators are collectively known as transducers
because they are used to convert energy of one kind into energy of another
kind. Transducers can be used to sense a wide range of different energy forms
Y
MEMS scanner
X Horizontal scan
Z Vertical
scan
PZT on stainless steel
1 mm
Gyro sensors 4 mm
Resonant in Resonant in Resonant in
Y-direction Y-direction X-direction
Figure 1.4 FEM simulations of three resonant motions in a PZT-based gyroscope and photo of
a fabricated gyroscope. Source: Adapted from Chang and Chen (2017).
such as movement, electrical signals, thermal or magnetic energy, etc. The type
of input or output transducers being used really depends on the type of signal
or process being “sensed” or “controlled,” but we can define a transducer as a
device that converts one physical quantity into another.
A smart system needs sensors and actuators to realize the sensing functions
such as distance, movement, and acceleration as well as actions. These sensors
and actuators use smart materials to realize the conversion between different
energies and moduli to electrical signals such as voltage, current, and capaci-
tance. Of course, many sensor devices are made of semiconductors such as the
FET, but this is not the focus of this book.
“Smart material” is a very large concept, in fact, there is no stupid material (a
joke), i.e. all materials are smart in some way since they all have their own prop-
erties and response to external stimuli. But in this book, we restrict the “smart
materials” to those materials with “ferroic” characteristics. We focus on basic
physics, materials science, structures, devices, and applications of ferroic mate-
rials for smart systems. The ferroic materials are usually classified as possessing
one of the followings based on coupling of stimuli:
(i) Ferroelectric, which is also piezoelectric when electromechanically coupled
and pyroelectric when thermoelectrically coupled.
(ii) Ferromagnetic, which is also magnetostrictive when magnetomechanically
coupled.
(iii) Ferroelastic, which also includes shape memory when thermomechanically
coupled.
Among these ferroics, we can see that strain, electric polarization and magne-
tization, and their interplay or coupling are involved. We call a material as ferroic
material if it possesses at least one of the properties of ferroelectric, ferromag-
netic and ferroelastic.
If we look at the diagram shown in Figure 1.5, we can see that the coupling
and interplay between electricity, mechanics, magnetism, heat, and optics result
in many smart functions, such as ferroelectric, piezoelectric, pyroelectric, ferro-
magnetic, electromechanical, etc. One book cannot cover all of them, but those
belong to ferroic materials and devices especially in the form of thin films will be
Electricity
E
ic ity Ma lectr
Elec
ctr
ic
n gn om
alor
o ele rictio eto ag
ele ne
tro-o
iez st ctr tism
troc
P c tro ici
e
ptic
l ty
Elec
E
Mechanics El Magnetism
as Magnetostriction
to-
op ric
tic
s c alo
to
optic
The
Elas oelastic
ne
ag
rm
M
toca
eto-
elec city
Opto tovoltai
tricit
Pho
i
n
The roelectr
loric
ism
Mag
elec
net
ag
rmo
tricit
om
Py
e rm
y
cs
Th
Thermo-optics
Heat Optics
Figure 1.5 Diagram showing coupling between different moduli and the clarification of smart
materials.
extensively introduced in this book. Before going into details, some application
examples of ferroic materials in smart systems are given in this chapter.
1.2 Device Application of Ferroelectric Materials

When people talk about applications of ferroelectric materials, the first thing
jumps out is most possibly the PZT (lead–zirconium–titanate with chemical
formula Pb(Zrx Ti1−x )O3 ), which is known as an excellent piezoelectric material.
As the most popular ferroelectric, PZT is also the most important piezoelectric
material in commercial applications. Piezoelectric materials have very broad
applications in many fields, from medical ultrasound imaging to ultrasonic
wire bonding machine in semiconductor industry, from pressure sensors to
accelerometer, etc. The market size of piezoelectric materials is more than US
$1 billion now and is expected to be US $1.68 billion by 2025 (GRAND VIEW
RESEARCH).
Another field of application of ferroelectric materials is the infrared sensors
based on their pyroelectric property, which is also one of the most important
properties of a ferroelectric material. Beyond these well-known applications,
another important application based on the switching of ferroelectric polariza-
tion is the non-volatile memory device such as FeRAM. Examples are given in
the following and details will be introduced in the following chapters.
1.2.1 Piezoelectric Device Applications

An example of smart system using piezoelectric material is the distance radar
system in a car or a sonar system in submarines as shown in Figure 1.6, where the
key sensing element is based on piezoelectric material to realize the conversion
between electrical energy and acoustic energy for sending and receiving sound
waves. Other application examples of piezoelectric devices include active damp-
ing system, micro-scanning system in scanning probe imaging instrument (such
as AFM), force sensor, accelerometer, energy harvesting, etc.
Medical ultrasound imaging system with piezoelectric material as the trans-
ducer to convert electrical and acoustic energies is another very good example
of device application where the piezoelectric material plays the roles of sens-
ing and actuating functions. Figure 1.7 shows photos of ultrasound transducers
developed in our group. Knowledge in ultrasound transducer fabrication, char-
acterization, and applications will be intensively introduced in Chapter 6.
A very new application example is piezoelectric-based fingerprint ID system in
mobile phone. The currently used finger identification system is based on capac-
itance measurement to obtain two-dimensional (2D) information of fingerprint,
but it faces the problem of difficulty to identify the fingerprint when the finger is
dirty or wet. Ultrasound fingerprint identification system based on piezoelectric
ultrasonic transducer and imaging system can obtain a three-dimensional image
(a) (b)
Figure 1.6 Piezoelectric materials-based sonar system for car (a) and submarine (b).
(a) (b)
Figure 1.7 (a) Transducers and (b) B-mode image of a wire phantom acquired with
PolyU-made array ultrasound transducer.
PMUT unit
Capping layer
Coupling materials + MEMS
Piezoelectric layer
Cavity
CMOS wafer
Bottom electrode Top electrode
Figure 1.8 Illustration of concept of a ultrasonic transducer-based fingerprint ID system based

on complementary metal-oxide-semiconductor micro-electro-mechanical systems
(CMOS-MEMS) technology.
of fingerprint with a certain depth. This can overcome the problems of the cur-
rent fingerprint identification system in most mobile phones. InvenSense, Inc. is
one of the main suppliers of this solution, and Figure 1.8 is an illustration of the
ultrasonic fingerprint system.
1.2.2 Infrared Sensor

An infrared sensor is usually made of a ferroelectric material, which is also
pyroelectric that generates surface electric charges when exposed to heat
in the form of infrared radiation. A pyroelectric-based infrared sensor can detect
the temperature change but produce no response for a steady temperature since
the pyroelectric sensing element can only produce polarization change-induced
electric charge when the sensor is subject to temperature change. Figure 1.9a
shows a photo of a real infrared detector with its internal device structure
illustrated in Figure 1.9b, where the active element is made of pyroelectric
materials such as LiTaO3 . Those pyroelectric materials with their polarization
able to be switched are ferroelectrics. Therefore, pyroelectric sensors that are
widely used as infrared detectors are important device applications for ferroic
materials in a smart system.
Window
Absorbing electrode layer

Bottom electrode Pyroelectric plate
Circuit board
(a) (b)
Figure 1.9 A photo of an infrared detector (a) and illustration of its internal structure (b).
1.2.3 Ferroelectric RAM (FeRAM)

Ferroelectric RAM (FeRAM, F-RAM, or FRAM) is a random-access memory that
is similar to Dynamic Random Access Memory (DRAM) in structure but uses a
ferroelectric layer instead of a dielectric layer to achieve non-volatility. FeRAM
is one of a growing member of alternative non-volatile random-access memory
technologies that offers the same functionality as flash memory.
Advantages of FeRAM over flash memory include lower power usage, faster
write performance, and much greater maximum read/write endurance (about
1010 –1014 cycles). FeRAMs have data retention of more than 10 years at +85 ∘ C
(up to many decades at lower temperatures). Market disadvantages of FeRAM
are much lower storage densities than flash devices and higher cost.
A ferroelectric material has a nonlinear relationship between the applied elec-
tric field and the stored charge. Specifically, the ferroelectric characteristic has
the form of a hysteresis loop, which is very similar in shape to the hysteresis
loop of ferromagnetic materials. The dielectric constant of a ferroelectric is typi-
cally much higher than that of a linear dielectric because of the effects of electric
dipoles formed in the crystal structure of the ferroelectric material. When an
external electric field is applied across a dielectric, the dipoles tend to align them-
selves with the field direction. This alignment process is produced by small shifts
in the positions of ions and shifts in the distributions of electric charges in the
crystal structure. After the charges are removed, the dipoles retain their polariza-
tion state. Binary “0”s and “1”s are stored as one of the two possible electric polar-
izations in each data storage cell. For example, in Figure 1.10, a “1” is encoded
using the negative remnant polarization “−Pr ,” and a “0” is encoded using the pos-
itive remnant polarization “+Pr ”. In this FET structure with a ferroelectric layer as
gate dielectric, the two polarization states correspond to different V th , resulting
in a memory window within which the ON and OFF states of the FET can be read.
Spontaneous
polarization of
ferroelectric layer
Gate Memory
window
Source Drain – – – –
– – – – – – – –– –
N N
–––––––––
IDS P
–––––––––
– – – –
VG
Figure 1.10 Schematic diagram of field-effect transistor (FET) and the current–voltage (I–V)
characteristics induced by two different polarization state.
1.3 Device Application of Ferromagnetic Materials 9
FeRAM remains a relatively small part of the overall semiconductor market.

In 2016, worldwide semiconductor sales were US $338.93 billion (according to
WSTS, SIA), with the flash memory market accounting for US $59.2 billion
(according to IC insights (Cho 2018)). The 2017 annual revenue growth of
Cypress semiconductor, perhaps the major FeRAM vendor, were reported to
be US $2.33 billion. The much larger sales of flash memory compared with the
alternative FeRAMs support a much larger research and development effort.
Flash memory is produced using semiconductor linewidths of 15 nm at Renesas
Electronics Corporation (2017). Flash memory can store multiple bits per cell
(currently Samsung has announced the 64-layer 512-Gb in the NAND flash
devices). As a result of innovations in flash cell design, the number of bits per
flash cell is projected to increase to double or even to triple. As a consequence,
the areal bit densities of flash memory are much higher than those of FeRAM,
and thus the cost per bit of flash memory is orders of magnitude lower than that
of FeRAM.
1.3 Device Application of Ferromagnetic Materials

Among many successful applications of ferromagnetic-based devices, memory
device based on ferromagnetic material is one of the most successful examples,
especially in the thin film form. This is manifested by the very large market of
magnetic hard disc in computing systems. But in most recent years, solid state
memory (mainly flash memory) is superseding the magnetic hard disc. Never-
theless, ferromagnetic material also finds its application in non-volatile memories
such as spin-transfer torque memory.
1.3.1 Spin-Transfer Torque Memory

Spin-transfer torque can be used to flip the active elements in magnetic
random-access memory. Spin-transfer torque magnetic random-access memory
(STT-RAM or STT-MRAM) has the advantages of lower power consumption
and better scalability over conventional magnetoresistive random-access mem-
ory (MRAM), which uses magnetic field to flip the active elements. Spin-transfer
torque technology has the potential to make possible MRAM devices combining
low current requirements and reduced cost; however, the amount of current
needed to reorient the magnetization at present is too high for most commer-
cial applications, and the reduction of this current density alone is the basis
for present academic research in spin electronics. Figure 1.11 is a schematic
diagram of spin valve structure, while arrows indicate the magnetization
direction.
1.3.2 Magnetic Field Sensor Based on Multiferroic Device

If a material possesses more than one of the ferroic properties of ferroelectric,
ferromagnetic and ferroelastic, it is called multiferroics. Unfortunately, such
Free layer
Pin layer
Low resistance state High resistance state
Figure 1.11 Schematic diagram of spin valve structure where arrows indicate the
magnetization directions.
materials are rare and are usually strong in one property but very weak in
another, such as BiFeO3 , which is very strong in ferroelectrics but very weak in
ferromagnetic (it is antiferromagnetic in fact). This makes multiferroic materials
hard to be practically applied in devices. However, people have been trying to
make composite materials such as piezoelectric with magnetostrictive mate-
rials, where the mechanical coupling between them makes the “multiferroic”
meaningful for device application, for example, making very sensitive magnetic
field sensor.
The magnetoelectric (ME) effect is the phenomenon of inducing magnetiza-
tion by an applied electric field (E) or polarization by magnetic field (H). Many
efforts have been devoted to improve the limit of detection of the ME compos-
ite at low frequency range, and values of ∼10−7 Oe at 1 Hz has been reported
(Wang et al. 2011). Based on the magnetic–strain–electric coupling, scientists
have demonstrated dc magnetic field sensor with a detection limit of 2 × 10−5 Oe
to dc magnetic field with a nonlinear ME magnetic effect (Li et al. 2017). Ferro-
electric material PZT and magnetostrictive material Metglas have been imple-
mented in the composite device (see Figure 1.12).
1.4 Ferroelastic Material and Device Application

Shape memory alloys (SMAs) are the most typical ferroelastic material, which
is an important member of ferroics. The shape memory characteristics origi-
nates from the phase transition between high temperature austenite phase and
low temperature martensite phase, where the shape at cubic-austenite phase can
be resumed from low temperature martensite phase whose lattice can be largely
twisted (see Figure 1.13).
The SMAs have been widely used in devices from brace of orthodontia
and other medical applications, air jet, satellite antenna, etc. Research was
carried out in developing systems that would optimize the chevron “immersion”
1.4 Ferroelastic Material and Device Application 11
ID electrodes
Metglas
Piezofiber
Epoxy Kapton
VirginiaTech
Figure 1.12 Outline of ME device from Virginia Tech and schematic of the cross-section of the
ME composite. Source: Wang et al. (2011). Adapted with permission of John Wiley and Sons.
He
ati
ng
Stress/load
Loading
Cooling
Heating
Temperature
Figure 1.13 Phase transition between high temperature austenite phase and low
temperature martensite phase, where the shape at cubic-austenite phase can be resumed
from low temperature martensite phase whose lattice can be largely twisted.
into the jet flow based on the flight condition. As shown in Figure 1.14a,
SMAs activated by heat were developed that would allow for full chevron
immersion in jet flow during high thrust requirements (e.g. during take-off )
and not immersing it during cruise where the thrust efficiency is of greater
importance (Anon n.d.).
(a) (b)
Figure 1.14 (a) Brace of orthodontia using shape memory alloys and (b) arthrodesis device
developed by Karnes et al.
For broken bone rehabilitation, a SMA plate with a memory transfer tempera-
ture close to body temperature can be attached to both ends of the broken bone as
shown in Figure 1.14b. From body heat, the plate will contract and retain its origi-
nal shape, therefore exerting a compression force on the broken bone at the place
of fracture. After the bone has healed, the plate continues exerting the compres-
sive force and aids in strengthening during rehabilitation (Garlock et al. 2017).
1.5 Scope of This Book

In Chapters 2–5, fundamentals of ferroelectrics, applications of ferroelectric
materials, recent advances, and advanced measurement and testing techniques
in ferroelectrics will be introduced. In particular, device applications of fer-
roelectric materials in thin film form will be introduced including FeRAM,
ferroelectric tunneling-based resistive switching, etc. The recent advances
include ferroelectricity in emerging materials such as 2D materials and high-k
gate dielectric material HfO2 , while the advanced characterization technologies
include the piezoresponse force microscopy (by imaging and switching ferro-
electric domains) and Cs-corrected transmission electron microscopy (TEM)
where atomic level ionic displacement can be identified.
As the most important property application of ferroelectric materials,
fundamentals of piezoelectric physics and engineering considerations for device
design and fabrication are introduced in Chapters 6 and 7.
In Chapter 8, starting with a brief introduction on origin of ferromagnetism and
its analogy to ferroelectrics, device applications, particularly for magnetostrictive
devices, are introduced.
Chapters 9 and 10 will introduce the multiferroics of materials possessing both
ferromagnetic and ferroelectric orders including single phase and composite
materials. In particular, devices based on the integration of ferroelectric and
ferromagnetic materials such as multiferroic memory device and ME coupling
device for sensor applications will be introduced.
In Chapter 11, ferroelastic materials represented by SMA and magnetic SMAs
as well as their device applications will be introduced.
References 13
References
Boyn, S., Grollier, J., Lecerf, G. et al. (2017). Learning through ferroelectric domain
dynamics in solid-state synapses. Nature Communications 8: 1–7.
Chang, C.-Y. and Chen, T.-L. (2017). Design, fabrication, and modeling of a novel
dual-axis control input PZT gyroscope. Sensors 17 (11): 2505.
Cho, J. (2018). Amid contradictory forecast: IC insights: ‘Memory chips will grow at
annual rate of 5% only on average by 2022’. Seoul, Korea: BusinessKorea.
Garlock, A., Karnes, W.M., Fonte, M. et al. (2017). Arthrodesis devices for
generating and applying compression within joints. US 2017/0296241 A1,
Available at: https://patents.google.com/patent/US20170296241A1/en.
Li, M., Dong, C., Zhou, H. et al. (2017). Highly sensitive DC magnetic field sensor
based on nonlinear ME effect. IEEE Sensors Letters 1 (6): 1–4.
Renesas Electronics Corporation (2017). Renesas electronics achieves large-scale
memory operation in fin-shaped MONOS flash memory for industry’s first
high-performance, highly reliable MCUs in 16/14nm process nodes and beyond.
Wang, Y., Gray, D., Berry, D. et al. (2011). An extremely low equivalent magnetic
noise magnetoelectric sensor. Advanced Materials 23 (35): 4111–4114. Available
at: https://doi.org/10.1002/adma.201100773.
15
Introduction to Ferroelectrics
2.1 What Is Ferroelectrics?

Ferroelectric can be defined as switchable spontaneous polarization at
temperatures below the Curie point (T c ). The spontaneous polarization and
its switching make the ferroelectric material exhibit polarization–electric field
hysteresis loop (called P–E loop) as shown in Figure 2.1, where the Pr is called
remnant polarization, Ps is the saturated polarization, and Ec is the coercive field,
a field at which the polarization direction can be switched. A very important
characteristic of ferroelectric materials is the presence of domain structure
where each domain has its own polarization direction. The formation of domain
structure is to align dipoles in different directions in order to minimize elec-
trostatic and elastic energy, and it is the domain switching and domain wall
movement that produce hysteresis and saturation nonlinearities. As shown in
the inset of the figure, the polarization switching is accompanied by domain wall
movement and new domain formation.
It is worthy to note that, except with presence of charged defects, the polariza-
tions are usually aligned head-to-tail. However, in some nanostructured ferro-
electrics, or under a special strained situation, charged domain walls, topological
domains such as vertex type of domains can be formed. Details will be introduced
in Chapter 5.
2.1.1 P–E Loop

People usually use the term of P–E loop to prove the ferroelectricity. But, strictly
speaking, it should be called D–E loop since the so-called polarization is actually
obtained by measuring the electric displacement D through capacitance mea-
surement. Nevertheless, it is approximately correct, since for ferroelectric mate-
rial, the electric polarization is nearly equal to the electric displacement.
For a dielectric material, the presence of an electric field E makes the elec-
tron cloud slightly shifted inducing a local electric dipole moment that is called
electric displacement, while for a ferroelectric material, its polarization density P
(spontaneous polarization or induced by electric field) will dominate the electric
displacement D that can be defined as
D = 𝜖0 E + P (2.1)
Ferroic Materials for Smart Systems: From Fundamentals to Device Applications,

First Edition. Jiyan Dai.
© 2020 Wiley-VCH Verlag GmbH & Co. KGaA. Published 2020 by Wiley-VCH Verlag GmbH & Co. KGaA.
16 2 Introduction to Ferroelectrics
P Figure 2.1 Typical P–E loop from a

ferroelectric crystal. Insets illustrate
domain structures for the
Ps corresponding polarizations. As the
field is increased, the polarization of
Pr domains with an unfavorable direction
of polarization will start to switch in
the direction of the field.
E
–Ec +Ec
–Pr
where 𝜖 0 is the vacuum permittivity or called permittivity of free space and

P depends linearly on the electric field in a homogeneous and isotropic
material, i.e.
P = 𝜖0 𝜒 e E (2.2)
where 𝜒 e is electric susceptibility of the material. Therefore, we have
D = 𝜖0 (1 + 𝜒e )E = 𝜖E (2.3)
where the permittivity
𝜖 = 𝜖0 𝜖r (2.4)
and relative permittivity
𝜖r = 1 + 𝜒e (2.5)
For a ferroelectric material, its electric displacement is contributed mainly by
ionic displacement, which is much larger than the contribution from vacuum
dielectric; therefore, the polarization P is approximately equal to D. Practically,
P–E loop is measured by a Sawyer–Tower method that will be introduced in
Chapter 3.
2.1.2 Relationships Between Dielectric, Piezoelectric, Pyroelectric,

and Ferroelectric
2.1.2.1 Ferroelectric–Dielectric
Why does a ferroelectric material have a very large dielectric constant?
In a large category, a ferroelectric belongs to dielectrics. There are three pri-
mary contributions to electric polarization: electronic, ionic, and dipole reori-
entation; and the polarization is expressed quantitatively as the sum of electric
dipoles per unit volume (C/m2 ).
2.1 What Is Ferroelectrics? 17
Dipole
Electronic Ionic
reorientation
E=0
E>0
Center of negative charge
Electric field
Figure 2.2 Schematic diagram showing the origin of the electric polarization.
For all dielectric materials, the electron clouds deform under electric field,
forming electric dipoles. Electric polarization from this electron clouds defor-
mation is usually much smaller compared with ionic displacement. In ionic
crystals, when an electric field is applied, cations and anions are attracted to the
cathode and anode, respectively. While for a ferroelectric, electric polarization
reorientation (rotation or reversing) will result in a significant change of electric
displacement (see Figure 2.2).
This diagram also illustrates that a ferroelectric falls into a larger category called
dielectrics. Let’s look at their difference with an example by comparing refrac-
tive index of a typical dielectric material SiO2 and a typical ferroelectric mate-
rial BaTiO
√ 3 . By following relationship of refractive index with dielectric constant
n = 𝜖r , with dielectric constant of about 3.6 for SiO2 , its refractive index is
about 1.4–1.5. However, for a ferroelectric material, its dielectric constant is usu-
ally very large due to the reorientation of electric polarization, a few hundred for
BaTiO3 , for example, but we cannot
√ say that its refractive index is also very large
based on the relation of n = 𝜖r . In fact, the refractive index of a material is an
optical constant that only works in optical frequency regime in which electron
cloud can respond to the optical frequency of electric field (say 1014 Hz), but the
ferroelectric polarization
√ cannot follow the optical frequency to rotate or switch,
i.e. this equation n = 𝜖r does not work for ferroelectrics in optical frequency.
2.1.2.2 Ferroelectric–Piezoelectric
Many materials such as AlN, GaN, and ZnO exhibit electric polarization orig-
inated from their crystal structure symmetry breaking, but their polarizations
cannot be switched by external electric field. These electrically polarized mate-
rials are piezoelectric but not ferroelectrics. Figure 2.3 shows the relationships
between the piezoelectric and ferroelectric materials as well as pyroelectric and
dielectric materials. A ferroelectric is also piezoelectric since the electric polar-
ization change induced by mechanical strain will generate electric charges and
voltage at the surface of a ferroelectric plate, i.e. piezoelectricity.
Figure 2.3 Relationships of the

piezoelectric, ferroelectric,
pyroelectric, and dielectric.
Ferroelectric
Pyroelectric
Piezoelectric
Dielectric
2.1.2.3 Ferroelectric–Pyroelectric
As illustrated in Figure 2.3, a ferroelectric is also a pyroelectric, but a pyroelectric
material may not be a ferroelectric. Pyroelectricity is a phenomenon of tempera-
ture dependence of the spontaneous polarization. As the temperature of a crys-
tal is changed, electric charges corresponding to the change of the spontaneous
polarization appear on the surface of the sample. Among many pyroelectric mate-
rials, only those whose spontaneous polarization can be reversed by an electric
field belong to ferroelectrics.
2.2 Origin of Ferroelectrics

2.2.1 Structure-Induced Phase Change from Paraelectric
to Ferroelectric
The origin of ferroelectric comes from space inversion symmetry breaking, i.e.
the structure symmetry breaking results in spontaneous polarization of a crystal
and generates the ferroelectricity. A simple understanding is that there must be a
relative shift of cations and anions, i.e. non-coincidence of positive and negative
charge centers, in the crystal unit cell. This will not happen in a crystal of cubic
structure with central symmetry. To generate electric polarization in a perovskite
structure, lattice distortion along either ⟨100⟩, ⟨110⟩, or ⟨111⟩ is a must, plus rel-
ative shift of cations and anions. This only happens when the crystal is below its
Curie temperature. The following describes the formation of ferroelectricity from
the crystal symmetry point of view:
• Crystals can be classified into 32 point groups according to their crystallo-

graphic symmetry, and these point groups can be divided into two classes, one
with a center of symmetry and the other without, as indicated in Table 2.1.
• Among these 32 point groups, 21 of them do not have a center of sym-
metry. Except the point group 432, in crystals belonging to 20 of these
non-centrosymmetric point groups, positive and negative charges are gen-
erated on the crystal surfaces when a stress is applied; so these crystals are
known as piezoelectrics.
2.2 Origin of Ferroelectrics 19
Table 2.1 Crystallographic classification according to crystal centrosymmetry.
Crystal system
Symmetry Polarity
Cubic Hexagonal Tetragonal Rhombohedral Orthorhombic Monclinic Triclinic
Centro (11) m3m m3 6/mmm 6/m 4/mmm 4/m 3m 3 mmm 2/m
32 point group Non-polar
432 622 422
of (22) 23 6 4 32 222
crystallographic Non-centro 43m 6m2 42m
symmetry (22) 2
Polar (10) 6mm 6 4mm 4 3m 3 mm2 1
m
Inside are piezoelectrics (20)
Inside are pyroelectrics (10)
Source: Adapted from Uchino (2009).
Ti
Ti O Ti
Ps
Ps
Ba
T > Tc T < Tc
Figure 2.4 Illustration of crystal structure symmetry breaking-induced ferroelectricity in

BaTiO3 .
• Among these 20 point groups, 10 of them have a unique polar direction, and
these 10 polar classes are pyroelectrics. Among these pyroelectrics, only those
whose polar can be reversed by external electric field are ferroelectrics.
BaTiO3 is an example of a typical ferroelectric crystal with perovskite struc-
ture as illustrated in Figure 2.4. Above the Curie temperature T c , the paraelectric
cubic structure is centrosymmetric, whereas in the tetragonal phase below T c , it
is energetically favorable for the O2− ions to be shifted slightly below face centers,
and Ti4+ ions are shifted upward from the unit cell center. The relative change in
positions of the Ti4+ and O2− ions produces a spontaneous polarization Ps as well
as the non-centrosymmetric structure. More ferroelectric phases of BaTiO3 will
be introduced later.
2.2.2 Soft Phonon Mode

Ionic displacement can be expected through lattice vibrations at a finite temper-
ature. Figure 2.5 shows some of the possible lattice vibrations in a perovskite-like
crystal, where
Panel (a) shows an initial cubic (symmetrical) structure.
Panel (b) is a symmetrically elongated one.
Panel (c) has coherently shifted center cations exhibiting ferroelectricity. If the
shift is anti-polarized, it results in antiferroelectricity (AFE) that will be intro-
duced later.
Elongation
direction
(a) (b)
(c)
Figure 2.5 Starting from the original cubic structure (a), if (b) is stabilized, only oxygen
octahedra are distorted without generating dipole moments (acoustic mode). When (c) is
stabilized, dipole moments are generated (optical mode). The final stabilized state (c)
corresponds to ferroelectric state.
Since only oxygen octahedra are distorted in Figure 2.5b without generating
electric polarization, its vibration is a kind of phonon mode. However, if the par-
ticular mode Figure 2.5c becomes stabilized, with decreasing temperature, the
vibration frequency decreases (soft phonon mode), and finally at a certain phase
transition temperature (Curie temperature), this frequency becomes zero, i.e. the
electric polarization state is stabilized.
Since the soft modes in ferroelectrics lead to electric polarization, they are opti-
cally active and can be detected by means of optical spectroscopy in the spectra of
dielectric permittivity (real and imaginary parts). Therefore, the vibration shown
in Figure 2.5c is also called optical mode.
Spectroscopic studies of the soft phonon modes provide a very powerful tool
for investigating the ferroelectric transitions. Figure 2.6 shows Raman spectra of
BaTiO3 phase transition, where the 305 cm−1 peak reduces its sharpness as the
in situ temperature increases and becomes indistinct at 150 ∘ C, indicating that
tetragonal phase is transformed to cubic phase. This trend is consistent with the
results that the peak around 310 cm−1 appeared below Curie temperature (T c )
and vanished above T c (Hayashi et al. 2013).
The dielectric permittivity in the soft mode is governed by three laws (Cochran
1960; Shirane et al. 1970; Luspin et al. 1980):
(a) Static dielectric permittivity 𝜖 r (0) produced by the soft mode obeys the
Curie–Weiss law:
𝜖r (0) ∼ (T − Tc )−1 (2.6)
(b) The eigen frequency 𝜔(T) of the soft mode follows the Cochran behavior:
1
𝜔(T) ∼ (T − T0 ) 2 (2.7)
where T 0 is the soft mode condensation temperature.

Intensity (a.u.)
200 °C
150 °C
100 °C
50 °C
25 °C
200 300 400 500 600 700 800

Raman shift (cm–1)
Figure 2.6 In situ Raman spectra of BaTiO3 particles measured at different temperatures: 25,
50, 100, 150, and 200 ∘ C. Source: Hayashi et al. (2013). Adapted with permission of Elsevier.
(c) The static dielectric constant and the soft mode frequency are connected via
the Lyddane–Sachs–Teller (LST) relation:
𝜖∞ 𝜔2 (T)
(T) = T 2 (2.8)
𝜖r (0) 𝜔L
where 𝜖 ∞ is the high-frequency dielectric constant and 𝜔T and 𝜔L are the
transverse and longitudinal frequencies of the corresponding vibrations,
respectively.
2.3 Theory of Ferroelectric Phase Transition

2.3.1 Landau Free Energy and Curie–Weiss Law
Ferroelectric phase transition can be well described by thermodynamic theory.
Landau addressed this problem by considering free energy in the form of expan-
sion of ferroelectric polarization P, from which fundamental properties of the fer-
roelectric material such as polarization and dielectric permittivity can be derived.
For simplicity, we only consider one-dimensional case, so the expression of Lan-
dau free energy as a function of polarization and temperature T is
F = F0 (T) + F(P, T) (2.9)
1 2 1 4 1 6
F(P, T) = 𝛼P + 𝛽P + 𝛾P + … (2.10)
2 4 6
where F 0 does not depend on the order parameter P and it describes the tem-
perature dependence of free energy of high temperature phase near the phase
transition (Graz University of Technology n.d.; Uchino 2009). The coefficients
𝛼, 𝛽, and 𝛾 are temperature dependent in general, while their signs are different
for the first- and second-order phase transitions. The reason to have only even
number of powers is that when the polarization changes signs, the free energy of
the crystal will not change. Since this phenomenological formula describes the
free energy across the paraelectric to ferroelectric phase transition, it should be
applicable to all temperature ranges.
In our calculation, however, the coefficient 𝛼 is the only coefficient that is
assumed to be temperature dependent. The fraction numbers are for conve-
nience when we differentiate F in respect of P. To define the relation of 𝛼 and T,
the free energy for certain polarization should be considered. As in paraelectric
state (i.e. P = 0), the free energy should be zero (F(P, T) = 0) at any temperature
above its Curie point. In ferroelectric state, there are two situations for free
energy, either F(P, T) > 0 or F(P, T) < 0. For free energy greater than zero, the
paraelectric state should be realized because it always tends to the minimum
energy state. Therefore, the free energy for a certain polarization must be lower
than zero in order to stabilize the ferroelectric state. Thus, the coefficient 𝛼 of
the P2 term must be negative in ferroelectric state, while it is positive passing
through zero at temperature T 0 . According to this concept, a relation of 𝛼 and T
is formed as the following:
𝛼 = 𝛼0 (T − T0 ) (2.11)
where 𝛼 0 is a temperature-independent constant and T 0 the Curie–Weiss tem-
perature, which is equal to or lower than the actual transition temperature T c
(Curie temperature).
By differentiating the free energy, electric field E related to the equilibrium
polarization can be expressed as
𝜕F
E= = 𝛼P + 𝛽P3 + 𝛾P5 (2.12)
𝜕P
By considering the existence of polarization without external electric field, i.e.
E = 0, we can get
P(𝛼 + 𝛽P2 + 𝛾P4 ) = 0 (2.13)
This equation has two possible solutions, among them the trivial solution P = 0
that corresponds to a paraelectric
√
state is not our concern for ferroelectric state.
−𝛽± 𝛽 2 −4𝛼𝛾
Another finite solution P2 = 2𝛾
corresponds to a ferroelectric state.
From Landau free energy, the temperature-dependent dielectric constant
(relative permittivity) of a ferroelectric material at ferroelectric and para-
electric phases (below and above Curie temperature) can be derived. For
the first order of phase transition from paraelectric phase to ferroelectric
phase, the temperature-dependent Landau free energy, dielectric constant, and
spontaneous polarization are illustrated in Figure 2.7. One can see that the
characteristic of the first-order ferroelectric phase transition is that the order
parameter Ps drops to zero abruptly and the dielectric constant reaches a finite
T > T1
T c < T < T1
1/ϵr
Ps
T = Tc
T0 < T < Tc ϵr
T < T0
(a) (b) T0 Tc T1
Figure 2.7 (a) Free energy, (b) dielectric constant and spontaneous polarization, of first-order
phase transition.
peak value continuously; while 1/𝜖 r is linear at temperatures lower or higher

than Curie point. The meaning of T 0 and T 1 will be shown in the following
derivations.
2.3.2 Landau Theory of First-Order Phase Transition

For second-order phase transition, the order parameter grows continuously from
zero at the phase transition so the first two terms of the power series will dom-
inate. If the free energy is expanded to the sixth order in the order parameter,
the system will undergo a first-order phase transition with 𝛼 > 0, 𝛽 < 0, and 𝛾 > 0
(Wang 2010).
1 2 1 4 1 6
F(P, T) = 𝛼P + 𝛽P + 𝛾P (2.14)
2 4 6
When we try to find the existent polarization P without the external electric field
E, the free energy is minimized.
𝜕F
E= = 0 = 𝛼P + 𝛽P3 + 𝛾P5
𝜕P
P=0 or 𝛼 + 𝛽P2 + 𝛾P4 = 0
So, nonzero solution of P is
√
2
−𝛽 ± 𝛽 2 − 4𝛼𝛾
P =
2𝛾
Here the “−” sign is meaningless since it will result in P2 < 0. So,
√
2
−𝛽 + 𝛽 2 − 4𝛼𝛾
P =
2𝛾
Since 𝛼 = 𝛼 0 (T − T 0 ), the solutions of P are
√ √
−𝛽 + 𝛽 2 − 4𝛼0 (T − T0 )𝛾
P=± for T < T0
2𝛾
√ √
−𝛽 + 𝛽 2 − 4𝛼0 (T − T0 )𝛾
P = 0, ± for T0 < T < T1
2𝛾
P=0 for T > T1
The temperature T 1 is where the two nonzero solutions become unstable. For
T > T 1 , there is only one free energy minimum. For T 1 > T > T 0 , there are three
√ minima. P = 0 is the lowest free energy solution when T 1 > T > T c , while
potential
√
−𝛽+ 𝛽 2 −4𝛼0 (T−Tc )𝛾
P=± 2𝛾
exhibits the lowest free energy when T c > T > T 0 . For
T < T 0 , there are double potential minima of the free energy that correspond to
stable spontaneous polarization.
The difference between the Curie–Weiss temperature T 0 , the Curie tempera-
ture T c , and the ferroelectric limit temperature T 1 can be verified based on the
potential minima obtained in the first-order phase transition.
As the potential minima are obtained from
𝜕F
= E = 𝛼P + 𝛽P3 + 𝛾P5 = 0
𝜕P
which is valid for any temperature below or above Curie temperature, there are
three possible minima including P = 0 (i.e. F = 0). At T = T c , the free energy at
nonzero polarization must be equal to that of the paraelectric state, i.e.
F(P, T)|T=Tc = F(P, T)|P=0 = 0
So,
1 2 1 4 1 6
F(P, T) = 𝛼P + 𝛽P + 𝛾P = 0
2 4 6
then we get
𝛼 + 𝛽P2 + 𝛾P4 = 0
1 1
𝛼 + 𝛽P2 + 𝛾P4 = 0
2 3
By eliminating the P4 term from these two equations,
1 2
2𝛼 + 𝛽P = 0
2
4𝛼
P2 = −
𝛽
Putting P2 = − 4𝛼
𝛽
into the equations,
( ) ( )2
4𝛼 4𝛼
𝛼+𝛽 − +𝛾 − =0
𝛽 𝛽
or
( )2
1 4𝛼
𝛼− 𝛾 − =0
3 𝛽
Taking 𝛼 = 𝛼 0 (T − T 0 ) = 𝛼 0 (T c − T 0 ),
3 𝛽2
𝛼 = 𝛼0 (Tc − T0 ) =
16 𝛾
the Curie temperature is thus calculated as
3 𝛽2
Tc = T0 +
16 𝛼0 𝛾
This indicates that T c is little higher than T 0 . Meanwhile, when T = T 1 , the free
energy has only one solution at P = 0, i.e.
𝛽 2 − 4𝛼0 (T1 − T0 )𝛾 = 0
𝛽2
T1 = T0 +
4𝛼0 𝛾
For this equation, we can identify that T 1 is higher than T c .
As a result, we can conclude that
T0 < Tc < T1
To identify the dielectric constant, we first have to calculate the relative permit-
tivity 𝜖 r which is the response of the system to an electric field.
dP
𝜖r =
dE
In free energy equation, the field has been included to calculate 𝜖 r
𝜕F
=E
𝜕P
E = 𝛼0 (T − Tc )P + 𝛽P3 + 𝛾P5
dE = [𝛼0 (T − Tc ) + 3𝛽P2 + 5𝛾P4 ]dP

√
At the critical temperatures P = 0, −𝛽
2𝛾
, the susceptibility in the vicinity of the
phase transition can be obtained.
dP || 1
at Tc , 𝜖r = =
dE ||P=0 𝛼0 (T − Tc )
dP || 1
at T1 , 𝜖r = | √ −𝛽 =
dE |P= , 𝛼0 (T − T1 )
2𝛾
Based on the equations earlier, we can verify that in case of 𝛽 < 0, the permittivity
shows a maximum and a discontinuity of the spontaneous polarization appears
at T c .
2.3.3 Landau Theory of a Second-Order Phase Transition

At a second-order phase transition, the order parameter increases continuously
from zero starting at the critical temperature of the phase transition (i.e. T c = T 0 ).
It is assumed that 𝛽 > 0, so that the free energy has a minimum for finite values
of the order parameter. When 𝛼 > 0, there is only one minimum at P = 0. When
𝛼 < 0, there are two minima with P ≠ 0 (Figure 2.8).
For the ferroelectric materials following the second-order phase transition, it
can be derived as
1 1
F(P, T) = 𝛼P2 + 𝛽P4 where 𝛼0 > 0, 𝛽 > 0
2 4
It is assumed that 𝛽 > 0 so that the free energy has a minimum for finite values of
the order parameter.
Similar to the previous case, polarization can be found when E = 0.
𝜕F
= 0 = 𝛼P + 𝛽P3 = 𝛼0 (T − Tc )P + 𝛽P3
𝜕P
So,
P=0 or 𝛼0 (T − Tc ) + 𝛽P2 = 0
When T > T c , there is only one minimum at P = 0. When T < T c there are two
minima with P ≠ 0, i.e.
P=0 for T > Tc
√
𝛼0 (Tc − T)
P=± for T < Tc
𝛽
For dielectric constant,
𝜕F
=E
𝜕P
dE = [𝛼0 (T − Tc ) + 3𝛽P2 ]dP
T > Tc
T = Tc
Ps
1/ϵr
T < Tc ϵr
(a) (b) Tc
Figure 2.8 (a) Free energy, (b) dielectric constant and spontaneous polarization, of
second-order phase transition.
dP || 1 C
𝜖r = = = for T > Tc
dE ||P=0 𝛼0 (T − Tc ) (T − Tc )
dP || 1 C
𝜖r = | 𝛼 (T − T) = = for T < Tc
dE |P2 = 0 c 2𝛼0 (Tc − T) 2(Tc − T)
𝛽
where C = 𝛼1 is Curie constant.
0
Now we can see that the dielectric permittivity can be derived from
second-order differentiation of Landau free energy with boundary conditions of
E = 0 (without electric field). The temperature-dependent dielectric permittivity
for ferroelectric materials following the first-order phase transition can thus be
derived. In both cases, 1∕𝜖r linearly depends on temperature.
The classification of order of phase transition comes from thermodynamic free
energy according to Ehrenfest (Jaeger 1998; Blundell and Blundell 2009), where
it can be labeled as the lowest derivation (order parameter) of the free energy
that shows discontinuity as a function of other thermodynamic variables during
phase transition. For instance, if a discontinuity is exhibited in the first derivative
of Gibbs free energy, it is classified as first-order phase transition. If the discon-
tinues order parameter is the second order of differentiation of Gibbs free energy,
the corresponding phase transition is classified as second-order phase transi-
tion. Although Ehrenfest classification is clear to understand, it is not a complete
method as it is not suitable if the derivative of free energy diverges. Therefore,
a modern classification is made. Similar to the Ehrenfest classes, two broad cat-
egories are divided by taking the latent heat into account (Maris and Kadanoff
1978).
For example, various solid/liquid/gas transitions usually involve a discontinu-
ous change in density, which is the first derivative of free energy with respect to
pressure. Therefore, they are classified as first-order phase transition based on
Ehrenfest classification. In the view point of modern classification, during the
water to ice phase transition, latent heat is added while the temperature of the
system remains unchanged. The specific discontinuity with respect to energy and
temperature of this phase transition thus assorts to the first-order differentiation
of Gibbs free energy, and it belongs to first-order phase transition associated with
latent heat. While for β-brass (an alloy of copper and zinc, in equal amounts)
phase transition, there is no latent heat. However, when the temperature is lower
than transition temperature, the probability for each copper atom to have more
zinc nearest neighbors increases, which would result in a completely random
arrangement of copper and zinc atoms. To say, the material undergoes a contin-
uous phase transition in which the specific heat has a singularity and it belongs
to second-order phase transition.
For the ferroelectric phase transition, for example, BaTiO3 , there is latent
heat during paraelectric to ferroelectric phase transition, and the spontaneous
polarization abruptly (discontinuously) changes to zero, which is the first-order
differentiation of Landau free energy. Therefore, this phase transition belongs
to first-order phase transition. While for the second-order phase transition, the
Ps continuously changes to zero, but the dielectric permittivity, which is the
second-order of differentiation of Landau free energy, is discontinuous at T c .
Most ferroelectric phase transitions are first order, very few (such as triglycine
sulfate [TGS]) are second-order transition.
2.4 Ferroelectric Domains and Domain Switching

2.4.1 Domain Structure
In order to reduce electrostatic and elastic energy, a ferroelectric material usually
presents a domain structure where every single domain possesses a polarization
direction. Depending on the structure, the domain polarizations are formed into
different angles. For example, in a perovskite ferroelectric material BaTiO3 with
polarizations along ⟨100⟩ and ⟨110⟩ for a pseudocubic unite cell, the domains
are aligned with 180∘ or 90∘ as shown in Figure 2.9a. While if the polarizations
are along ⟨111⟩ directions, 71∘ and 109∘ domains will be formed (angles between
the ⟨111⟩ directions). However, for a hexagonal structure such as LiNbO3 with
polarizations along ⟨0001⟩, only 180∘ domains can be formed. As shown in the
figure, the polarizations in adjacent domains are usually arranged head-to-tail in
order not to form charged domain walls. But some charged domain walls can also
be formed in some crystals if defects exist.
There are also vortex types of domains formed in some ferroelectric materials,
for example, YMnO3 , or in nanostructures and thin films. Figure 2.9b is a vor-
tex domain structure in YMnO3 . These vortex domains have been theoretically
predicted and experimentally observed and will be introduced in Chapter 5.
2.4.2 Ferroelectric Switching

In a ferroelectric material, polarization switching can be realized by applying
electric field and/or mechanical stress. It should be noted that an electric field can
induce both 180∘ and 90∘ polarization switching, whereas a mechanical stress can
only induce 90∘ switching of polarization. This can be understood by considering
an ellipsoid representing a polarized unit cell as illustrated in Figure 2.10. One
can see that an electric field can force the ellipsoid from out-of-plane to in-plane
90° boundary 180° boundary
β+
γ– α–
α+ γ+
(a) (b) β–
Figure 2.9 (a) Ferroelectric domain structure in BaTiO3 and (b) cloverleaf domain patterns and
the enlarged single vortex domain in the ferroelectric YMnO3 . Source: Adapted from Choi et al.
(2010) and Reprinted with permission of Springer Nature: Zhang et al. (2013).
E-field
direction E-field
direction
Adding electric field or
90° switching 180° switching
Stress
Adding mechanical stress or
90° switching
Figure 2.10 Illustration of polarization switching by electric field and mechanical stress.
as well as up-to-down, i.e. 90∘ and 180∘ polarization rotations, respectively. How-
ever, a mechanical stress along the long axis can only press the ellipsoid into an
in-plane ellipsoid.
In a ferroelectric crystal, its polarization switching is accompanied by domain
switching as well as domain wall movement. This process will induce strain in the
crystal, where the non-180∘ switching results in large piezoelectric strain.
2.5 Ferroelectric Materials

The first ever ferroelectric material is Rochelle salt discovered in 1675 by an
apothecary, Pierre Seignette, of La Rochelle in France. But this Rochelle salt is
not suitable for engineering application due to its poor property and stability.
The most typical and applicable ferroelectric material is BaTiO3 , which has
been widely used in electronic components as capacitors due to its very large
dielectric constant and very low dielectric loss. Nevertheless, the most popular
and useful ferroelectric material is Pb(Zrx Ti1−x )O3 (PZT) due to its very large
piezoelectric constant and remnant polarization. The applications of PZT in
sensors and actuators are the implementation of its excellent piezoelectric
property and applications in infrared sensors utilizing its pyroelectric prop-
erty; while some small-scale application of resistive random access memory
(ReRAM) implement its ferroelectric property. Relaxor ferroelectrics such
as Pb(Mg1/3 Nb2/3 )O3 –xPbTiO3 (PMN-xPT) emerged in recent decades have
attracted great attention due to their outstanding ferroelectric and piezoelectric
properties.
2.5.1 From BaTiO3 to SrTiO3

BaTiO3 is a model ferroelectric material, when temperature decreases, it goes
through a series of phase transitions from cubic to tetragonal, orthorhombic, and
then rhombohedral. The spontaneous polarization directions corresponding to
these non-centrosymmetric structures are ⟨100⟩, ⟨110⟩, and ⟨111⟩, respectively,
indexed by pseudocubic structure. The so-called pseudocubic means we still use
a′
Ti
b′ 130 °C c
a′ –90 °C 0 °C
a
b
Ba
Rhombohedral Orthorhombic Tetragonal Cubic
(a = b = c; (a = b = c; (a = b ≠ c; (a = b = c;
α = β = γ ≠ 90°) α ≠ 90°, β = γ = 90°) α = β = γ = 90°) α = β = γ = 90°)
Figure 2.11 Illustration of BaTiO3 structures at different temperatures showing different

polarization directions at different phases.
cubic structure to index the atomic planes and directions for convenience, but in
fact, the crystal structure is not cubic any more due to its distortion.
You may be wondering, the so-called orthorhombic structure in Figure 2.11
should be a monoclinic structure when the cubic unit cell is stretched along
[110] direction. It seems to be right, but since there is relative shift of cations
and anions, the real Bravais unit cell is actually a larger orthorhombic unit cell
with the shadow plane as the a′ b′ plane with c′ perpendicular
√ to a′ b′ plane.
′ ′ ′
The orthorhombic unit cell has a = a, b = c = 2a, i.e. the volume of the
orthorhombic unit cell is doubled, and a′ , b′ , and c′ are not equal, where a is for
pseudocubic unite cell in the distorted structure.
Ceramics based on BaTiO3 are very important in electronic component of
capacitors due to its high dielectric constant and low dielectric loss. BaTiO3 thin
film is also one of the most attractive ferroelectric thin films for memory device
application. This will be introduced in Chapter 4.
Pure SrTiO3 is a quantum paraelectric material where quantum fluctuations of
atomic positions suppress a ferroelectric transition, leading to a so-called incipi-
ent ferroelectric (Müller and Burkard 1979; Zhong and Vanderbilt 1996; Barrett
1952; Hemberger et al. 1995). It means that SrTiO3 is supposed to be ferroelectric
but its Curie temperature is absolute 0 K, which can never be reached. There-
fore temperature-dependent dielectric constant of SrTiO3 increases continuously
when temperature decreases, but it can never have a peak like a normal ferroelec-
tric material (see Figure 2.12). However, one can see that when a bias voltage is
applied on the SrTiO3 crystal, a relaxor-like paraelectric–ferroelectric transition
happens above 0 K. This is because of the existence of polar nano regions (PNRs)
in SrTiO3 at low temperatures. These PNRs can form ferroelectric nano-domains
under electric field, giving rise to relaxor-like ferroelectricity that will be intro-
duced later in this section.
SrTiO3 single crystal has been used as many oxide thin films substrate material
due to its high chemical stability and good lattice matching with many perovskite
oxide materials. When SrTiO3 is doped with BaTiO3 , it becomes (Bax Sr1−x )TiO3
(called BST) and its Curie temperature can be tuned between 0 K to T c of BaTiO3
depending on how much BaTiO3 is added. The study of BST, especially BST thin
films, has been a very hot topic due to its large dielectric tunability which is very
useful in tunable microwave device.
BST is attractive for microwave device application mainly because of two rea-
sons. The first is that the dielectric permittivity of BST strongly depends on the
25
1
1- E = 0 kV/cm
20 2- E = 2 kV/cm
3- E = 5 kV/cm
4- E = 15 kV/cm
ϵ(T ) × 103
15
10
2
5 3
4
0
0 100 200 300
Temperature (K)
Figure 2.12 Dielectric constant of SrTiO3 single crystal as a function of temperature and
biasing field. Source: Vendik et al. (1999). Reprinted with permission of Springer Nature.
bias electric field; this property is termed as dielectric tunability. Another reason
is that BST possesses very low dielectric loss (tan 𝛿) at microwave frequency. In
the following paragraphs, properties and applications of BST are introduced.
In Figure 2.13, an experimental result shows the change in lattice parameters
when different compositions of Ba and Sr are added into the compound. When
the composition of Ba increases, it can be seen that the structure of BST varies,
changing from cubic to tetragonal, proofing the structural transition from SrTiO3
(STO) to BaTiO3 (BTO). Therefore, it can be concluded that the tetragonality
increases when the composition of Ba increases.
0.405
c
Lattice parameters (nm)
0.400
a T
0.395
C
0.390
0.0 0.2 0.4 0.6 0.8 1.0

SrTiO3 Composition BaTiO3
Figure 2.13 Lattice constant of (Bax Sr1−x )TiO3 (both bulk and films) as a function of
composition x. Source: Adapted from Abe and Komatsu (1995).
25 000
x=0
20 000 x = 0.34
x = 0.27
x = 0.47 x = 0.65
x = 0.87
15 000
ϵr(T )
10 000 x = 1.0
5000
0
0 100 200 300 400
Temperature (K)
Figure 2.14 The phase transition behavior (the dielectric constant as a function of
temperature) of (Bax Sr1−x )TiO3 single crystal for various content of Ba–x. Source: Vendik et al.
(1999). Reprinted with permission of Springer Nature.
Figure 2.14 shows a theoretical result that the composition of Ba in BST varies
the Curie temperature as well as the dielectric properties of the compound.
Owing to the increase of tetragonality, when more Ba is added, the Curie
temperature increases. One can also notice that the dielectric constant near T c
is a few orders of magnitude greater than the rest, indicating a paraelectric to
ferroelectric phase transition.
Figure 2.15 shows the BST dielectric constant as a function of bias electric field.
The dielectric constant has a huge difference when E-field is applied. In the case
of x = 0.24, the dielectric constant is greatly reduced from 𝜖 r > 1400 to 𝜖 r < 500
when a field of E = 10 MV/m is applied. The electric field-dependent dielectric
permittivity is commonly described as the dielectric tunability n, which is defined
as the ratio of the dielectric permittivity at zero electric field bias to the permit-
tivity under electric field bias E, i.e.
𝜖r (0) − 𝜖r (E)
n=
𝜖r (0)
Question to students: Why does dielectric constant decrease when an E-field is
applied?
2.5.2 From PbTiO3 to PbZrO3

PbTiO3 (PT) is another model ferroelectric material. As illustrated in Figure 2.16,
above its Curie temperature, the paraelectric cubic structure is centrosymmetric;
but below Curie temperature, it is energetically favorable for the O2− ions to shift
slightly below face centers and Ti4+ ions to shift upward from the unit cell center
forming tetragonal phase. The relative shift in positions of the Ti4+ and O2− ions
produces a spontaneous polarization Ps .
Figure 2.15 Bias field 1600

dependence of (Bax Sr1−x )TiO3
dielectric constant under the
condition of f = 100 kHz at room 1400
temperature. Source: Adapted
from Abe and Komatsu (1995). 1200
Dielectric constant ϵ
1000
800
0.24
600
0.44
400
0
0.68
200
1.0
0
–50 –40 –30 –20 –10 0 10 20 30 40 50
Bias field (MV/m)
Pb
Ti
Ti O
Ps
(a) T > Tc (b) T < Tc
Figure 2.16 (a) Perovskite structure of PbTiO3 in the cubic form above T c and (b) tetragonal
structure of PbTiO3 for T < T c presenting spontaneous polarization.
2.5.3 Antiferroelectric PbZrO3

The notion of antiferroelectric (AFE) dates back to the early 1950s. It describes
a state where chains of ions in the crystal are spontaneously polarized, but
with neighboring chains polarized in antiparallel directions. As a result, the
crystal does not display spontaneous macroscopic polarization or the piezo-
electric effect. PbZrO3 was the first compound identified as an antiferroelectric
crystal (Sawaguchi et al. 1951; Shirane et al. 1951). At that time, Kittel (1951)
proposed a theory of antiparallel ionic displacements in dielectrics using the
antiferromagnetism scheme. In parallel, Sawaguchi et al. assigned the perovskite
lead zirconate, PbZrO3 , as antiferroelectric because of its dielectric behavior.
Thereafter, materials that exhibit a structural phase transition between two
nonpolar phases with a strong dielectric anomaly at the high temperature side of
Another random document with
no related content on Scribd:
En later, toen hij werkelijk aan 't zetten toe was: "'t Zal wel een
beetje pijn doen, maar niet schreeuwen, hoor!"
Verschrikkelijke pijn had 't gedaan, Eduard rilde nog als hij er aan
dacht, en heel stijf had hij zijn tanden in zijn lippen gedrukt, en
zonder dat hij 't wilde waren telkens groote tranen in zijn oogen
gekomen; maar geschreeuwd had hij niet; zouden de andere
jongens het dapper van hem vinden, dat hij niet geschreeuwd had?
Och, maar ze hadden niet gevoeld hoeveel pijn het deed, en
natuurlijk hadden ze gezien dat hij huilde! Maar Vader zou hem toch
zeker wel flink gevonden hebben als Vader 't wist; hij had toch ook
zijn best gedaan een moedige jongen te zijn. — Wat was dat ook
weer? Wie had dat gezegd? Wanneer zou dat geweest zijn? Even
was 't, of hij niet meer kon denken, — wat beteekende het dan toch?
En in zijn gedachte herhaalde hij weer de woorden: "Zul je je best
doen een moedige jongen te zijn?" — Nee, 't was niet "een", 't was
iets anders, wát was 't? Eduard voelde weer tranen in zijn oogen
komen, en nog eens probeerde hij te denken — och, wie kon hem
dat nu toch zeggen, wie wist 't dan toch? Nee, hier wisten ze 't niet,
hier wisten ze niks — Vader wist 't, Vader begreep zooiets
[a330] altijd dadelijk — Vader, wat was dat dan toch, van die
moedige jongen?
Eduard deed zijn oogen dicht: ja, daar was Vaders gezicht, daar
was hij ook weer in Vaders kamer, daar begon Vader te praten; —
wat zei Vader? Nu goed luisteren — duidelijk zag hij Vaders lippen
bewegen: "Eddy, zul je je best doen Vaders moedige jongen te zijn?"
Daar was 't! Nu wist hij 't weer: "Zul je je best doen Vaders moedige
jongen te zijn?" — Heel, heel dikwijls was hij niet moedig geweest,
maar dit zou hij dan toch aan Vader kunnen vertellen, dat hij niet
geschreeuwd had toen zijn arm gezet werd; — dat was immers
moedig? Wie had gezegd dat het moedig was? Daarnet zei iemand
dat toch — o ja, de dokter had het gezegd, de dokter had het zélf
verteld aan Tante Lina — "hij heeft zich ferm gehouden, nu hebben
we het ergste gehad!" — 't Ergste was nu voorbij; o, een heerlijk
gevoel was 't geweest, toen hij eindelijk hier in bed lag, toen hij heel
stil, heel rustig kon blijven liggen; — 't had ook nog zoo vreeselijk
lang geduurd voordat hij thuis was, die akelige lange rit in het rijtuig,
waar maar geen eind aan kwam, met de dokter en Hugo, toen hij
maar aldoor zijn oogen dicht gehouden had; en 't was zoo
ongemakkelijk geweest! En toen die drukte hier in huis, ze waren
allemaal zoo geschrikt, en ze hadden toch dadelijk gedacht dat 't erg
was, al was van Merlen ook gaan zeggen dat 't niets beteekende.
[a331] Eduard hoorde beneden een deur open en dicht gaan, toen
volgde een zacht gestommel op de trap — er kwam iemand met licht
naar boven — o, dat was 't nachtlichtje waarover Tante Lina al iets
gezegd had — wie kwam 't nu brengen? Was 't Tante Lina zelf? Nee,
zeker was 't Juf. — Eduard hoorde hoe 't nachtlichtje heel zacht op 't
portaaltje werd neergezet — de juffrouw was straks ook zoo stil naar
boven gekomen, toen de kleintjes naar bed gingen, en bij 't
uitkleeden had hij ze ook bijna niet gehoord; anders maakten ze
altijd zoo'n lawaai, maar zeker waren ze ook geschrikt. Wat zouden
Lineke en Bep wel gezegd hebben, en de kleine jongens? Zouden
Hugo en Piet nog niet naar bed gaan? O nee, 't was pas half negen
... nu niet meer denken, maar slapen; — waar dacht hij daarnet nu
ook weer aan, voordat 't nachtlichtje daar stond — aan 't rijtuig —
nee, dat was toch niet 't laatste — o ja, van Merlen die 't was gaan
zeggen — van Merlen was een leuke jongen, hij was er ook niet bij
geweest toen ze hadden gezegd van dat naar beneden rijden — "ik
durf best, maar ik doe het niet" — o, had hij het ook maar niet
gedaan, had hij maar volgehouden dat hij niet wilde, dát was moedig
geweest, en dat hij gegaan was, was laf! Hij was heelemaal niet
Vaders moedige jongen geweest — had hij toen hij op het punt was
om naar beneden te rijden maar aan Vader gedacht, misschien had
dat hem wel geholpen, misschien had hij 't dan wel niet gedaan! Of
had het vanmorgen [a332] nu maar dadelijk zoo gewaaid, dan waren
ze heelemaal niet gaan fietsen, en dan was er ook niets gebeurd —
nu woei het, heele vlagen waren 't, maar 't was nu te laat — waar
zou de fiets gebleven zijn? Hadden de jongens hem vergeten? Zeker
was hij heelemaal kapot; hij had er nog niet eens aan gedacht om 't
te vragen — straks roepen tegen Hugo en Piet als ze boven
kwamen. Nee, alleen Hugo roepen — als de fiets kapot was moest
hij zoo gauw mogelijk gemaakt worden, — hij kon hem niet missen;
— of ja, nu wel, hij kon nu toch niet rijden — hoe lang zou 't wel
duren? Een dag, twee dagen, drie, vier, vijf, zes .... wat zes? — O ja,
dagen, zes dagen — zou 't zes dagen duren? Wat? 't Maken van de
fiets? Of 't in bed liggen? Wat zei de dokter? Hoeveel dagen? Zes
dagen, zeven dagen, acht, negen, tien ....
Onrustig flikkerde 't nachtlichtje, telkens bijna uitwaaiend als een

hevige windvlaag langs 't huis streek; hoe laat was 't nu? Waren
Hugo en Piet al naar bed? — Eduard luisterde — stil, akelig stil was
het in 't heele huis — had hij lang geslapen? 't Kon best midden in
den nacht zijn; buiten joeg de wind door de boomen en langs de
muren — 't was nu nog veel erger dan straks, zeker had hij toch
nogal lang geslapen — sloeg er nu maar eens een klok! En zijn
lippen waren zoo akelig droog — o, maar 't glas water stond naast
hem; — even probeerde [a333] Eduard zich op te richten, maar
dadelijk zakte hij weer in 't kussen terug — 't deed pijn; was 't hier
maar niet zoo donker, dan kon je tenminste op je horloge kijken — 't
nachtlichtje op 't portaaltje gaf haast niets; en hij had zoo'n dorst!
Toch nog eens probeeren om te drinken — voorzichtig haalde hij 't
glas naar zich toe; nu ging 't beter — he, dat was wel heerlijk frisch,
maar iedere beweging deed pijn — nu verder stil blijven liggen. —
Weer loeide een windvlaag langs 't raam, en weer flikkerde 't
lichtje. Nu stormde 't dan toch wel —zoo angstig hard was 't aldoor
— en wat zou 't op zee vreeselijk weer zijn! Eens was hij op een
stormachtigen dag aan 't strand geweest met Vader; hooge golven
met groote witte schuimkoppen, en daarboven die heel donkere,
zwarte lucht — maar dat was nog overdag geweest, nu was 't
midden in den nacht, nu was 't heelemaal pikdonker, niets zou je
zien, alleen zou je 't hooren, 't slaan van de golven en 't geloei van
den wind — en midden op zee was Vader, heel ver weg was 't schip
nog, midden in den storm — zou Vader ook de golven tegen 't schip
hooren slaan en zou Vader ook 't gieren van den wind hooren? Zou
Vader ook wakker zijn, en nu, op 't oogenblik, aan hem denken?
Zeker zou Vader denken dat hij kalm sliep, en Vader zou zelfs
heelemaal niet vermoeden hoe hij hier wakker lag, met een
gebroken arm! Wat zou Vader 't akelig vinden als Vader weer thuis
kwam — natuurlijk zou [a334] Vader denken dat hij zijn jongen al
aan 't station zou vinden; — maar Vader zou hem niet zien, en dan
zou Vader hier komen en hooren dat hij ziek was, en dan zou Vader
dadelijk naar boven gaan — wat kráákte daar op de tráp? —
Nee, 't was niets. —
Dat was zijn straf, dat Vader hem zóó zou moeten vinden, met zijn
gebroken arm, en 't was allemaal zijn eigen schuld — dat had de
dokter ook nog gezegd: "Eigen schuld natuurlijk, dat malle woeste
rijden ook, en dan zoo'n gevaarlijke helling! 't Is nog wonder dat het
zoo is afgeloopen!" — O, hij had best dood kunnen zijn; — wat
zouden ze hier dan wel gezegd hebben, als hij eens echt dood was?
Zou 't ze veel kunnen schelen? Een heel klein beetje misschien? —
Och, ze zouden 't morgen alweer vergeten zijn — alleen Vader niet,
arme Vader! Heel ver weg, midden op zee, was Vader —
"Vader," snikte Eduard zacht en groote tranen druppelden langs

zijn wangen op 't kussen, "Vader, ik voel me zoo eenzaam! Vader!"
Maar niemand hoorde hem. —
O, hij kon Vader niet missen, evenmin als Vader hem, waarom was
Vader toch weggegaan? — En als Vader nu terug kwam vond hij zijn
jongen met een gebroken arm — en dat was zijn eigen schuld, zijn
eigen onvoorzichtigheid; — hoe had Vader 't ook weer gezegd toen
Vader wegging: "Pepi, zul je maken dat ik een vroolijke, gezonde
jongen terug-[a335] vind?" En toch was hij niet voorzichtig geweest,
en als hij nu dood was .... "Nee, hij is niet dood, kijk maar, hij
beweegt zijn been!" — dat had een van de jongens gezegd, maar 't
was net of hij 't Linekes stem hoorde herhalen — hoe kon dat nu?
Zijn nichtje was er toch niet bij geweest? Waar kwam Lineke nu
opeens vandaan? Lineke had er niets mee te maken, en Bep ook
niet — "Nee, hij is niet dood" — wie was niet dood? — Nee, hij was
zelf niet dood, als je dood was kon je je niet meer bewegen, dan was
je koud en stijf, zooals — zooals 't kleine vogeltje — daar was 't, 't
kleine vogeltje, dat ze in het bosch gevonden hadden, dat hij
voorzichtig opgenomen had — "nee, 't is niet dood," zei Lineke, "kijk
maar, 't beweegt zijn pootje ...." Dat was het! — Arm klein vogeltje, 't
was toch gestorven, ver van zijn warme nestje, alleen ... En zonder
ophouden gleden Eduards tranen in 't kussen — "Vader, Vader, kom
toch gauw!"
Daar was de zee; aldoor loeide de storm, en donkere, zwarte

wolken hingen over de met witte schuimkoppen bedekte golven. —
Eduard rilde; — had hij toch maar een jas meegenomen, hij had 't
nog tegen Hugo gezegd, maar Hugo vond 't immers niet noodig. —
Daar was 't schip, duidelijk zag hij het in de verte, 't schip waarmee
Vader thuis kwam, en daar stond Vader ook en wuifde — "Vader!"
riep [a336] Eduard, en de menschen om hem heen keken hem met
malle, verbaasde gezichten aan — maar 't schip kwam niet verder,
aldoor bleef 't maar even ver weg; weer zoo'n hooge golf — waar
was 't schip nu? Weg was 't — overal de donkere zee met de witte
schuimkoppen — Vader — o, waarom zochten ze Vader nou niet?
Was Vader dood? Vader! Vader! En nog aldoor keek Eduard naar de
woeste zee daar voor hem .... ineens, daar was 't schip weer, nu
heel dicht bij, heel duidelijk kon hij Vader zien staan, en weer zag hij
Vader wuiven en hij zag Vaders lippen bewegen — wat zei Vader?
Och, kon hij 't nu maar verstaan, wat was 't dan toch? Weer opende
Vader zijn mond: "Eddy, zul je je best doen Vaders moedige jongen
te zijn?" ... Weg was alles weer, niets was meer te zien van Vader,
van het schip, alleen de donkere, zwarte zee — "Vader, Vader!" riep
Eduard wanhopig, en nog eens: "Vader!" —
Eduard schrikte wakker; 't was nu heelemaal donker in 't kleine

kamertje, want 't nachtlichtje op 't portaal was uitgewaaid — wat was
er ook weer voor akeligs gebeurd? Zijn arm was gebroken, maar wat
nog meer? Of had hij gedroomd? Wat stormde 't buiten! Ja, dat was
't, en wat nog meer? Buiten — op zee — het schip — Vader ....
Eduard was er weer; zoo akelig duidelijk was 't geweest, — o, als 't
eens echt waar was, als 't schip eens echt vergaan was — met
groote, angstige oogen keek Eduard in [a337] de donkere, zwarte
ruimte, en opeens was 't alsof hij hier niet in zijn kleine kamertje lag
maar in een heel groote, wijde, hooge zaal waarvan de muren
oneindig ver weg waren — en Eduard voelde plotseling een rilling
door zich heen gaan, alsof er iets heel griezeligs was — wat bewoog
daar in de verte, wat ging daar door de kamer heen — wat was dat?
Eduard kneep zijn oogen stijf dicht om 't maar niet meer te zien —
wat was dat nu weer? Allemaal groene en roode figuren die heen en
weer bleven zweven, eerst heel langzaam, toen vlugger, aldoor
vlugger — daar ging 't weer door de kamer ....
"Hugo!" riep Eduard, en hij schrikte van zijn eigen stem, vreemd
klonk 't door 't stille huis — maar niemand antwoordde ....
Ingespannen bleef Eduard luisteren: zeker had hij niet hard genoeg
geroepen, want alles bleef stil — alleen een zacht, zagend geluid
drong tot hem door; hoe kwam dat nu weer? — Ineens vond hij de
oplossing van 't raadselachtige gezaag — 't was 't snurken van Piet,
dat was altijd zoo hard en de muur tusschen de kamers in was niet
heel dik. —
Hoe kwam 't nu opeens weer dat hij met zijn hoofd aan 't
voeteneind lag? Hij had zich toch niet omgedraaid? Wat was alles
toch raar vannacht! Waar was de muur nu? Nee, daar niet —
voorzichtig voelend stak Eduard zijn hand uit — daar? Maar dat was
de verkeerde kant — o, wacht, dan lag hij toch goed, dan was 't
maar verbeelding geweest daarnet ... [a338] Nu aan iets anders
denken; wat had hij vanavond ook weer willen doen? O ja, een brief
aan Vader schrijven — dat had hij nu nog niet eens gedaan, en
morgen moest de brief uiterlijk weg.
Er moest toch een brief naar Vader toe, hij wilde Vader toch
vertellen dat hij zijn arm gebroken had, en dan vragen of Vader zoo
gauw mogelijk thuis kwam — en Beppie zou 't schrijven, ze kon
alleen maar drukletters maken, maar dan zou hij wel voorzeggen wat
ze zetten moest — aan de anderen wilde hij 't toch niet vragen. —
Nu bedenken wat hij schrijven zou ....

"Eindelijk wakker? Wel, je hebt een gat in den dag geslapen!" zei
Tante Lina, de gordijnen openschuivend, "weet je wel dat het al
kwart over negen is?"
Eduard wreef met zijn hand in zijn oogen. "Zoo laat al?" vroeg hij
verbaasd.
"Ja, zoo laat al!" lachte Tante Lina, "je hebt het er zeker maar eens
van genomen, he? Ik ben vannacht ook nog eens naar je komen
kijken, om een uur of twee, maar je sliep zoo lekker!"
Eduard gaf geen antwoord; om twee uur was hij dan zeker juist niet
wakker geweest. — En stil bleef hij er naar liggen kijken hoe Tante
Lina aan 't opruimen ging en water in de kom schonk.
"Nu zal ik de invalide eerst eens helpen met [a339] wasschen,"

praatte ze verder, "en dan zal ik je ontbijt boven sturen."
Over opstaan zei Tante Lina niets, en Eduard begon er ook niet
over. Hij had er nu heelemaal geen zin meer in, hij had hoofdpijn, en
verder voelde hij zich ook nog niks lekker. —
Piet kwam de boterhammen en 't glas melk boven brengen. "Wil ik

je soms helpen met eten?" vroeg hij.
"Nee dank je, ik kan 't best alleen, hoor!" zei Eduard, maar pas toen
Piet al lang weer beneden was stak hij den eersten hap in zijn mond.
Een halve boterham at hij op, toen had hij er genoeg van; die gelei
smaakte ook al zoo vervelend.
Eduard was blij toen Bep een poosje later boven kwam. "Hoe is 't
met je, Eetje?" vroeg ze, en met groote oogen keek ze naar 't witte
verband om den gebroken arm.
"Goed," antwoordde Eduard, "vind je niet dat mijn arm er mooi

uitziet zoo?"
Bep knikte, en keek aandachtig de kamer rond. "Je hebt je
boterhammen niet opgegeten!" zei ze toen.
"Ik heb geen trek meer."
"En heb je geen dorst ook?"
"Ja, dorst wel, geef me 't glas maar."
Voorzichtig, zonder morsen, gaf Beppie 't glas melk aan en haastig
dronk Eduard 't leeg. "Wil je wat voor me doen, Bep?" vroeg hij toen,
zijn mond aan zijn mouw afvegend.
"Ja, wat dan?"
[a340] "Een brief schrijven."
"Maar ik kan niet schrijven!"
"Da's niks, je kunt de letters toch wel maken, en dan zal ik 't je
verder wel voorzeggen."
"Goed," zei Beppie, en Eduard wees haar waar ze papier en een
potlood vandaan moest halen; toen sjouwde ze een paar
voetkussens aan en ging op den rand van 't bed zitten. "Wat moet ik
nu schrijven?" vroeg ze.
En Eduard dicteerde. "Lieve Vader! Een l, Bep, [a343] nu een i, e,

v, e; nu een nieuw woord, Bep, v .... en zoo ging 't langzaam verder:
Lieve Vader, ik heb mijn arm gebroken, ik ben gevallen, Bep schrijft
....
"Er kan niet meer op," zei Beppie, "kijk maar, 't papier is heelemaal
vol!"
Even lachte Eduard om de groote, scheeve drukletters. "Geef mij

nu 't papier nog even," zei hij, "dan zal ik er zelf mijn naam onder
zetten." En in een hoekje krabbelde hij, onduidelijk, "Pepi."
"Wat staat daar?" vroeg Bep, 't papier nog eens nieuwsgierig
bekijkend.
"Niks," antwoordde Eduard, en even kreeg hij een kleur, "niks

bizonders tenminste." En toen liet hij Bep de brief in 't couvert
stoppen en dichtplakken. "Neem hem nu maar mee naar beneden
en vraag of Oom Vaders adres er opschrijft," zei hij.
[a344]
XIX.
Stil lag Eduard voor zich uit te kijken.
't Raam stond open, en een zachte, warme zomerlucht kwam naar
binnen.
't Was nu al bijna een week geleden, dat hij zijn arm gebroken had,
en nog altijd lag hij hier in bed, en 't werd zoo vreeselijk vervelend!
Och, ze kwamen allemaal wel dikwijls boven om even een praatje te
maken, maar 't grootste deel van den dag waren ze naar school —
eergisteren niet, toen was 't Zondag, en toen had hij bijna aldoor
gezelschap gehad, maar gisteren was 't zoo echt vervelend
geweest, en vandaag begon 't alweer net zoo — na 't ontbijt waren
Bep en Broertje nog even bij hem geweest, maar al gauw was Juf ze
komen halen om met ze te gaan wandelen nu 't zulk mooi weer was,
en daarna had hij niemand meer boven gezien — hij mocht wel
lezen, en naast hem op 't tafeltje lag ook wel een boek, maar 't ging
zoo lastig, met dien eenen arm, en hij gaf ook nooit zoo heel veel om
lezen — en nog altijd was [a347] Vader niet thuis — al van Zondag
af had hij 't gedacht: misschien is de boot wel heel vroeg, misschien
komt Vader vandaag wel — maar Vader was niet gekomen —
Daar werd beneden gescheld; 't klingelde door 't heele huis;
misschien was het Vader — och nee, er werd 's morgens immers
altijd zoo dikwijls gebeld, de slager, en de groenteboer — maar
misschien was het de dokter wel, die zou vandaag nu toch zeker wel
weer eens komen! En misschien vertelde de dokter dan wel dat hij
op mocht staan! 't Was niet om zijn arm, dat hij nog in bed moest
blijven, maar de dokter had gezegd dat er van opstaan nog geen
sprake kon zijn zoolang de koorts niet heelemaal wegbleef — en
gisteren had hij bijna geen koorts gehad.
Nee, 't was de dokter niet, anders was hij nu al lang boven
geweest. — En Vader was 't ook niet, die zou ook wel dadelijk bij
hem gekomen zijn! — Maar Vader kwam nu toch gauw thuis, ieder
oogenblik kon Vader thuiskomen, 't was de derde Juni vandaag, en
de vier maanden waren om. Waren ze nu eigenlijk gauw omgegaan
of niet? Eduard wist 't niet. Soms was 't hem wel eens meegevallen,
dat er al zeven, of acht, of negen weken voorbij waren, maar
meestal toch had het akelig, vreeselijk akelig lang geduurd. — 't Was
immers ook nog midden in den winter toen Vader wegging, en nu
was 't heelemaal zomer, 't was warm en de zon scheen [a348] en de
boomen zagen er zoo prachtig groen uit! — Hoe laat was 't nu?
Eduard haalde zijn horloge onder 't kussen vandaan; tien minuten
voor twaalf; nee, dan kwam de dokter niet meer vanmorgen, de
anderen zouden wel haast uit school komen, Lineke en Tommy en
Hugo en Piet. — Even dacht Eduard er aan hoe 't op school altijd
was, als om twaalf uur de bel ging; de drukte in de klas, 't
geschreeuw en 't gestamp in de vestibule, 't harde gehol van de
jongens om maar zoo gauw mogelijk weg te komen, — en dan 't
vervelende gevoel als je school moest blijven en je alleen in die
leege, stille klas zat; met je tweeën was 't lang zoo erg niet, dan had
je nog wel eens pret, maar in je eentje was 't echt lam, en dan begon
die vervelende Snijders altijd te zaniken over dat malle
toelatingsexamen. —
Natuurlijk zou hij nu wel weer een heeleboel achter raken bij de
andere jongens. Wie weet hoe lang 't nog wel duurde voor hij weer
naar school mocht, en als hij nu weer terug kwam zouden de
sommen zeker nog wel weer slechter gaan dan vroeger. — Maar
dan was Vader ook weer terug, en dan zou Vader hem wel helpen,
als hij nu eerst maar aan Vader verteld had hoe ellendig 't rekenen
tegenwoordig ging en van die slechte cijfers! — Eduard duwde de
deken een eind van zich af; hij had 't zoo verschrikkelijk warm!
Daar werd weer gebeld. Was Vader daar? Eduard luisterde, bang
om zich te bewegen en een van de [a349] geluiden beneden te
missen. "Dag Maatje!" ving hij op; 't waren Bep en Broertje, met de
juffrouw, en even later drongen ook Tommy's bekende hooge kreten
met de lange uithalen tot hem door. — Waarom kwamen ze nu toch
niet een van allen boven? Niemand dacht er aan dat hij hier alleen
lag, en dat hij al de heele morgen alleen geweest was; allemaal
hadden ze hem vergeten! —
Daar kwam opeens iemand aan; haastige voetstappen wipten de

trap op, met een flinken zwaai werd de deur opengegooid en Lineke
stapte binnen, een groote tak jasmijn in haar hand. "Is hij niet
prachtig?" riep ze, "bijna alle bloemetjes er van zijn uit!" en toen ging
ze naast Eduards bed staan en hield den witten tak vlak bij zijn
gezicht.
"Ja, erg mooi!" bewonderde Eduard, voorzichtig een bloempje
vasthoudend om 't goed te bekijken, "en wat ruiken ze lekker!"
"Ik zal ze hier bij je zetten," zei Lineke, "dan heb je er den heelen
middag plezier van!" en toen haalde ze de karaf van de waschtafel
om den tak daarin te steken; een glas was niet groot genoeg.
Met een groot blad met boterhammen kwam Tommy naar boven
geschuifeld.
"Ik kom bij je koffiedrinken!" verklaarde hij; "vind je dat niet

vreeselijk prettig? En Juf zal straks de glazen melk brengen!"
"Natuurlijk vindt hij het prettig!" lachte Lineke, de kamer

uitgaande.[a350]
[a353] Tommy's gezelschap was heel vermakelijk. Hij deed
verhalen over gevechten, waarin hij jongens, die minstens tweemaal
zoo groot als hij waren, glansrijk verslagen had; over slooten, waar
hij over gesprongen was, en die zoowat even breed waren als de
lengte van Eduards bed ..."
"Zoo breed als de lengte van mijn bed?" herhaalde Eduard.
"Ja, zooiets," zei Tommy, "of misschien een klein beetje minder!" —
"Wil je eens uit het raam kijken wie er gescheld heeft?" vroeg
Eduard, die bij iedere bel ongeduldiger werd.
Tommy stak zijn hoofd naar buiten. "'t Is die jongen van Meertens!"
zei hij, "en Keetje heeft hem binnengelaten."
"O, leuk! laat hij gauw boven komen!" riep Eduard, "vraag aan
Tante of hij nu alsjeblieft eindelijk eens bij mij mag, anders is hij weg,
en ik moet hem noodig wat vragen ook!"
Zijn haast maakte indruk op Tommy, en de kleine jongen holde
naar beneden.
Dat "noodig" had er wel af kunnen blijven, bedacht Eduard, maar 't
was ook zoo leuk om eens over school en zoo te praten, en tot nu
toe waren Theo en Meertens en van Merlen, die ook eens was
komen vragen of hij bij hem mocht, allemaal weggestuurd, omdat er
geen bezoek mocht komen zoolang hij nog koorts had. Maar nu zou
't toch zeker wel mogen, hij was immers bijna beter! —
[a354] Een luidruchtig gestomp tegen de deur volgde; "Binnen!"

bromde Eduard met een lage stem, en Meertens kwam de kamer in.
"Zoo Kerner, leef je nog?" vroeg hij.
"Zooals je ziet!" zei Eduard, en op den stoel bij 't voeteneind van
zijn bed wijzend, "ga zitten!"
Meertens liet zich op den stoel neervallen. "Wanneer sta je weer

op?" informeerde hij.
"Zoodra 'k van die vervelende knul van 'n dokter m'n bed uit mag!"
"Da's nogal logisch!" lachte Meertens, "'t in bed liggen schijnt je

maar matig te bevallen!"
"'t Hangt me al lang meters ver de keel uit!" mopperde Eduard;

"hoe is 't op school?"
"Nou, sloom natuurlijk! Je boft dat je er niet bent; Snijders doet niks
als ons benauwd maken voor dat halve gare toelatingsexamen en
geeft hoopen huiswerk!"
"Prettig!"
"Ja, dat snap je!" zei Meertens, de kamer rondkijkend. — "Je Tante
wou me eerst niet naar boven laten gaan, maar toen kwam dat
kleine jong met die boodschap van je beneden en toen mocht ik,
even, zei ze. Wat had je voor interessants te vragen?"
"Ik? Nou, niks, ik wou eens wat anders hooren, je moet niet denken
dat ik hier voor m'n plezier den heelen dag in m'n eentje lig!"
"Nou, 't is een taaie boel op school," herhaalde Meertens, "en ik

kan niks doen of Snijders ziet 't nou jij niet voor me zit!"
[a355] "En wat zegt hij dan van het toelatingsexamen?"
"Nou, dat we er niet komen natuurlijk, we zijn lui, en dom, en nog

een heeleboel meer."
Even schrikte Eduard. Als mijnheer Snijders dat tegen álle jongens
zei, wat moest hij dan wel tegen hém zeggen! Hij was in den laatsten
tijd tóch bepaald al een van de minsten, en dan nog die weken dat
hij ziek was er bij —
"Zeg Meertens!"
"Nou?"
"Jij gaat naar de H. B. S. he?"
"Ja, naar de Hengelende Biggen Societeit," zei Meertens.
"Denk je dat je er komt?"
"Nou, dat hoop ik wel," zei Meertens, "'k zou 't nogal lam vinden als
ik zakte."
"Ik ga naar 't gymnasium." Eduard zweeg even. —"Denk je dat ik er

door kom?"
Meertens haalde zijn schouders op. "Hoe weet ik dat nou?"

antwoordde hij, "als 't aan mij lag wel!"
"Nou ja, doe nou niet flauw, geloof je dat ik zakken zal?"
"Ik zeg immers al dat ik 't niet weet!" herhaalde Meertens, "je kunt
soms wel nogal uilig doen tegenwoordig, maar misschien bof je bij 't
examen wel!"
"Da's opwekkend!" zei Eduard half lachend, en "nou, daar kan ik

toch niks aan doen!" beweerde Meertens, en toen begon hij een
verhaal te doen over Groeneveld, die 't bord schoon had moeten
maken, [a356] en er zoo hard tegen gebonkt had dat 't op den grond
geduikeld was. "Nou, Snijders tippelde er natuurlijk dadelijk naar
toe," vertelde hij, "en opspelen dat hij deed! 't Was om je een bult te
lachen!"
Eduard had dolle pret om 't verhaal en probeerde dadelijk 'n gezicht
te trekken zooals Groeneveld gezet moest hebben. Die kon zoo echt
onnoozel kijken. —
"Nou, zeg, 'k smeer 'm," zei Meertens eindelijk, en zich uitrekkend:
"eerste uur misselijke aardrijkskunde!"
"Dat vind ik juist altijd nogal moppig," antwoordde Eduard.
"'k Geef 't je cadeau!" bood Meertens aan; "kom, ik ga weer eens
naar school kuieren," en Eduard een klap op zijn uitgestoken hand
gevend: "ik kom nog wel eens!"
"Da's goed," zei Eduard, en toen Meertens al bij de deur was: "Doe
je 't heusch?"
"Ja zeker."
Op de trap hoorde Eduard hem nog roepen: "Beterschap!"
"Dank je!" riep Eduard terug. —
Eduard ging met zijn gezicht naar den muur liggen en deed zijn
oogen dicht.
"Misschien bof je wel!" had Meertens gezegd, en zeker had hij dus
ook niet gedacht dat Eduards kans om er door te komen heel groot
was; o, 't zou vreeselijk zijn als hij zakte, misschien zou mijnheer van
Eerde wel vinden dat hij heelemaal geen examen [a357] mocht
doen, en wat zou Vader dan wel zeggen! Had hij maar aan Meertens
gevraagd om zijn boeken mee te brengen! — Was hij nu heusch zoo
vreeselijk stom? Was hij nu heusch zooveel dommer dan de andere
jongens? En opeens was 't weer of hij mijnheer Snijders hoorde
zeggen: "'t Is niets dan luiheid, je kunt best als je maar wilt!" — Ja,
lui, dat was hij geweest! Nooit had hij zijn werk goed afgemaakt in de
laatste vier maanden, en iederen morgen was hij naar school
gegaan met halfgeleerde lessen — en de taak, die mijnheer Snijders
hem voor de Paaschvacantie gegeven had, was ook al niet
afgeweest, en het gedeelte, dat hij wél gemaakt had, zag er slordig
uit, en was vol fouten — en dat moest hij nu allemaal aan Vader
vertellen; — wat zou Vader wel zeggen!
Piet kwam nog even boven voor hij naar school ging. Eduard keek
om toen hij de deur hoorde opengaan, maar dadelijk draaide hij zijn
gezicht weer naar den muur.
"Wat mankeer jij?" vroeg Piet verbaasd.
"Niks," antwoordde Eduard onverschillig.
"Da's niet veel!" — en toen Eduard niets zei: "Nou, je kunt toch wel
antwoord geven!"
"Je vraagt me niks," zei Eduard met zijn oogen dicht, en Piet, die
zoo'n conversatie ál te gezellig vond, liep fluitend naar beneden. —
Eduard hoorde hem de voordeur achter zich dichttrekken. —
[a358] Nee, 't kon hem nu allemaal niks meer schelen, alles was
ook even lam, dat lange in bed liggen hier, en dat ellendige examen,
en Vader kwam maar niet thuis ... En stil bleef hij liggen soezen, uit
verveling de streepjes in 't behang tellend, en dan deed hij zijn
oogen weer dicht omdat ze zoo moe werden van het kijken naar die
warrelende figuurtjes. —
Opeens schrikte hij weer op. Daar werd gebeld! 't Was niet heel
hard, maar toch had hij het duidelijk gehoord — zou Vader daar zijn?
En weer bleef hij luisteren, half overeind, aldoor naar de deur
kijkend — wat duurde 't weer lang! Daar kwam iemand naar boven,
— nee, met hun tweeën kwamen ze naar boven, Eduard kon het
hooren aan de voetstappen, en ook hoorde hij ze praten — een lage
mannenstem — Vader! — Wat liepen ze langzaam! Nu antwoordde
Tante Lina, — vervelend dat die nu mee naar boven kwam, ze kon
toch wel begrijpen — wild bonsde het in Eduards hoofd en een
schok ging door hem heen toen de deurknop werd omgedraaid —
De dokter kwam binnen, en Tante Lina volgde hem en sloot de

deur weer. Met groote oogen keek Eduard de dokter aan, toen liet hij
zich weer achterover in het kussen vallen. —
Nog altijd was Vader er niet!
"Hoe is 't met onzen held?" vroeg de dokter vroolijk, maar Eduard
kon er niets aan doen dat zijn lip [a359] trilde bij het "Goed dokter,"
en dat er tranen in zijn oogen kwamen. De dokter zag het wel, maar
hij zei niets en begon naar het verband te kijken, alleen toen Tante
Lina even later naar beneden gegaan was om nieuw linnen te halen
keek hij Eduard aan en vroeg vriendelijk: "Wat scheelt mijn
patientje?"
"'t Duurt allemaal zoo vreeselijk lang!" snikte Eduard opeens, maar
de dokter liet hem niet verder praten: "Niet huilen, niet huilen!" zei
hij, "kom, je moet je dapper houden, tot nu toe ben je aldoor zoo'n
flinke jongen geweest! — Kijk, ik had je juist willen vertellen dat je
morgen eens mocht probeeren om op te staan, heb je daar zin in?"
Eduard knikte, en de dokter vervolgde: "Je moet niet zoo

ongeduldig worden, me dunkt dat we al weer een heel eind op streek
zijn! Je ligt hier bijna als een prins, de zomerlucht komt zoo lekker
naar binnen! Je hebt het heusch heel wat beter dan de meeste van
mijn patienten, geloof dat maar!" en tegen Tante Lina, die weer
binnengekomen was: "Is die prachtige jasmijn uit uw tuin, Mevrouw?
Wat 'n mooie groote bloemen!" —
"U mag wel een takje van de jasmijn hebben voor uw knoopsgat!"
zei Eduard, toen de dokter wegging, en hij keek er naar hoe Tante
Lina een schaar nam en een klein takje afknipte. — Hij mocht het
zelf in 't knoopsgat steken, 't ging wel een beetje lastig met één
hand, maar 't ging toch.
"Dus morgen mag hij eens probeeren op te staan!"[a360] zei de

dokter nog eens, en weer bleef Eduard alleen, en weer hoorde hij
Tante Lina en de dokter praten op de trap.
Hoe dom ook om daarnet te denken dat Vader naar boven kwam;
Vaders stem leek immers niets op de stem van den dokter! Och,
maar wanneer zou Vader nu toch ook eindelijk komen! — Met een
zucht draaide Eduard zich weer naar den muur toe. Maar weer
figuurtjes tellen, misschien viel hij dan wel in slaap, daarnet sliep hij
ook bijna toen die dokter opeens kwam ....
Hoe Iaat was 't nu? Eduard keek op zijn horloge; 'n minuut of zeven
voor vier; een uur zoowat had hij geslapen. Nu kwamen straks de
anderen weer uit school, en dan gingen ze zeker in den tuin spelen
of misschien kwamen ze ook wel boven, maar 't kon Eduard niks
schelen, ze mochten weg blijven ook, hij had nou toch geen zin om
naar die vervelende verhalen van ze te luisteren. —
Daar kwam er al weer eentje naar boven, nou kreeg hij zeker weer
zoo'n glas melk met een ei, akelig flauw goed; dan was 't altijd maar
't best om je slapend te houden; als je wakker was moest je 't nog
dadelijk opdrinken ook!
Eduard maakte zich gereed om Tante Lina, of de juffrouw, of wie er

anders komen mocht, slapend op te wachten, en weer kneep hij zijn
oogen dicht.
De voetstappen kwamen dichter bij, Eduard hoorde [a361] hoe de

knop van de deur omgedraaid werd en hoe er iemand naar 't bed
toekwam — nou, wat bleven ze nou staan zaniken, waarom zetten
ze nou dat glas niet neer, dan konden ze immers weer verdwijnen!

Ferroic Materials For Smart Systems From Fundamentals To Device Applications Dai Full Chapter

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

Ferroic Materials For Smart Systems From Fundamentals To Device Applications Dai Full Chapter

Uploaded by

Copyright:

Available Formats

Ferroic materials for smart systems

from fundamentals to device

From Fundamentals to Device Applications

Bibliographic information published by

© 2020 Wiley-VCH Verlag GmbH &

All rights reserved (including those of

Print ISBN: 978-3-527-34476-5

Typesetting SPi Global, Chennai, India

Printed on acid-free paper

1 General Introduction: Smart Materials, Sensors, and

2.5.1 From BaTiO3 to SrTiO3 29

3 Device Applications of Ferroelectrics 47

5 Recent Advances in Ferroelectric Research 95

5.3 Ferroelectric Vortex 99

6 Piezoelectric Eﬀect: Basic Theory 123

7 Piezoelectric Devices 143

7.3.4 Characterization of SAW Devices 159

8 Ferromagnetics: From Material to Device 165

9 Multiferroics: Single Phase and Composites 197

10.3 Memory Devices Based on Multiferroic Tunneling 224

11 Ferroelasticity and Shape Memory Alloy 231

General Introduction: Smart Materials, Sensors, and

The early twenty-ﬁrst century has foreseen acceleration of innovations in

Ferroic Materials for Smart Systems: From Fundamentals to Device Applications,

Figure 1.1 AI beats human chess player.

1.1 Smart System

As an example, a smart system of remote control copter relying on

1.2 Device Application of Ferroelectric Materials

1.2.1 Piezoelectric Device Applications

Coupling materials + MEMS

Bottom electrode Top electrode

Figure 1.8 Illustration of concept of a ultrasonic transducer-based ﬁngerprint ID system based

1.2.2 Infrared Sensor

Absorbing electrode layer

1.2.3 Ferroelectric RAM (FeRAM)

FeRAM remains a relatively small part of the overall semiconductor market.

1.3 Device Application of Ferromagnetic Materials

1.3.1 Spin-Transfer Torque Memory

1.3.2 Magnetic Field Sensor Based on Multiferroic Device

Low resistance state High resistance state

1.4 Ferroelastic Material and Device Application

1.5 Scope of This Book

2.1 What Is Ferroelectrics?

2.1.1 P–E Loop

Ferroic Materials for Smart Systems: From Fundamentals to Device Applications,

P Figure 2.1 Typical P–E loop from a

where 𝜖 0 is the vacuum permittivity or called permittivity of free space and

2.1.2 Relationships Between Dielectric, Piezoelectric, Pyroelectric,

Center of negative charge

Figure 2.3 Relationships of the

2.2 Origin of Ferroelectrics

• Crystals can be classiﬁed into 32 point groups according to their crystallo-

Table 2.1 Crystallographic classiﬁcation according to crystal centrosymmetry.

Inside are piezoelectrics (20)

Inside are pyroelectrics (10)

Source: Adapted from Uchino (2009).

Figure 2.4 Illustration of crystal structure symmetry breaking-induced ferroelectricity in

2.2.2 Soft Phonon Mode

𝜖r (0) ∼ (T − Tc )−1 (2.6)

where T 0 is the soft mode condensation temperature.

200 300 400 500 600 700 800

2.3 Theory of Ferroelectric Phase Transition

peak value continuously; while 1/𝜖 r is linear at temperatures lower or higher