Professional Documents
Culture Documents
Ferroic Materials For Smart Systems From Fundamentals To Device Applications Dai Full Chapter PDF
Ferroic Materials For Smart Systems From Fundamentals To Device Applications Dai Full Chapter PDF
https://ebookmass.com/product/molecular-beam-epitaxy-materials-
and-device-applications-asahi/
https://ebookmass.com/product/handbook-of-smart-photocatalytic-
materials-fundamentals-fabrications-and-water-resources-
applications-chaudhery-mustansar-hussain/
https://ebookmass.com/product/calculations-and-simulations-of-
low-dimensional-materials-tailoring-properties-for-
applications-1st-edition-ying-dai/
https://ebookmass.com/product/polymer-nanocomposite-based-smart-
materials-from-synthesis-to-application-rachid-bouhfid/
Self-Healing Smart Materials and Allied Applications
Inamuddin
https://ebookmass.com/product/self-healing-smart-materials-and-
allied-applications-inamuddin/
https://ebookmass.com/product/flow-batteries-from-fundamentals-
to-applications-christina-roth/
https://ebookmass.com/product/nickel-titanium-smart-hybrid-
materials-from-micro-to-nano-structured-alloys-for-emerging-
applications-micro-and-nano-technologies-1st-edition-sabu-thomas-
editor/
https://ebookmass.com/product/thermal-energy-storage-from-
fundamentals-to-applications-alexandra-soh/
https://ebookmass.com/product/spintronic-2d-materials-
fundamentals-and-applications-materials-today-wenqing-liu-editor/
Ferroic Materials for Smart Systems
Ferroic Materials for Smart Systems
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.
10 9 8 7 6 5 4 3 2 1
v
Contents
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
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
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
Index 247
1
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).
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.
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
1.2 Device Application of Ferroelectric Materials 5
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.
(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.
1.2 Device Application of Ferroelectric Materials 7
PMUT unit
Capping layer
Piezoelectric layer
Cavity
CMOS wafer
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.
Window
(a) (b)
Figure 1.9 A photo of an infrared detector (a) and illustration of its internal structure (b).
8 1 General Introduction: Smart Materials, Sensors, and Actuators
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
Free layer
Pin layer
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).
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.).
12 1 General Introduction: Smart Materials, Sensors, and Actuators
(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).
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
E
–Ec +Ec
–Pr
Dipole
Electronic Ionic
reorientation
E=0
E>0
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.
18 2 Introduction to Ferroelectrics
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.
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
Ti
Ti O Ti
Ps
Ps
Ba
T > Tc T < Tc
• 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.
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:
(b) The eigen frequency 𝜔(T) of the soft mode follows the Cochran behavior:
1
𝜔(T) ∼ (T − T0 ) 2 (2.7)
Intensity (a.u.)
200 °C
150 °C
100 °C
50 °C
25 °C
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.
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
2.3 Theory of Ferroelectric Phase Transition 23
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.
√ √
−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
2.3 Theory of Ferroelectric Phase Transition 25
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
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 .
26 2 Introduction to Ferroelectrics
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.
2.3 Theory of Ferroelectric Phase Transition 27
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 .
28 2 Introduction to Ferroelectrics
Most ferroelectric phase transitions are first order, very few (such as triglycine
sulfate [TGS]) are second-order transition.
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).
2.5 Ferroelectric Materials 29
E-field
direction E-field
direction
Stress
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.
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°)
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
2.5 Ferroelectric Materials 31
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
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).
32 2 Introduction to Ferroelectrics
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?
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
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.
FOOTNOTES:
[12] Major Harris, in his “Highlands of Ethiopia,” has made the
following assertions:—“In utter abhorrence of the country and its
inhabitants, the Moslem servants who accompanied the Embassy
from India all took their departure, willing to brave the dangers
and difficulties of a long journey through the inhospitable deserts
of Adaîel, rather than prolong a hateful sojourn in Abyssinia. One
half of the number were murdered on their way down, and the
places of all long remained empty.” This is most unjust both to the
Dankalli and the Abyssinian, for of the twenty native and Arab
servants, independent of the two tent Lascars mentioned as
having accompanied the Mission, eight only were dismissed in
Shoa—Sultaun, Hadjji Abdullah, Allee Chous, Berberah Allee,
Abbas, Mahudee, Hadjji Ohmed, and an Indian boy, whose name
I have forgotten. These servants had been led to expect, on their
arrival in Abyssinia, the payment of the high wages which, in
some of their cases, had alone induced them to accompany the
Mission through Adal. Their disappointment may be conceived
when they were then informed that a moiety only of their wages
would be paid to them in Shoa, and that the remainder would run
on in arrears until their return to Aden. This injustice, as it was
conceived to be, was resented, and the discharge of these eight,
in this remote country, was the consequence. The unfortunate
servants appealed to the Negoos for redress, who condescended
(but without avail) to intercede for their return to the Mission. This
affront to the royal dignity was never forgotten, whilst a very
injurious prejudice was raised by the conduct that was pursued by
our representative with reference to the non-performance of the
engagements entered into with these men. This being followed
shortly afterwards by the infliction of corporeal punishment upon a
soldier for a breach of martial law, when no other kind of discipline
was even pretended to be kept up, astonished the Abyssinians
not a little, and gave the finishing blow to all popular respect for
English civilization, or wishes for any connexion whatever with our
country.
But this is not all. Of the eight discharged servants, instead of the
whole of these men showing any abhorrence of the country, the
greater part of them took to themselves wives, and upon what
little they had saved lived near me in Aliu Amba. Three of them
however (Hadjji Ohmed, Mahudee, and the Indian boy), were
induced to attempt a passage to the sea-coast. The Kafilah they
accompanied was attacked on the eastern bank of the Hawash by
the Takalee tribe. The Indian boy was slain, but Hadjji Ohmed and
Mahudee, being mounted, fled different ways; the former
fortunately found protection and shelter for more than a month
with Omah Batta’s sub-division of the Sidee Ahbreu tribe, whilst
Mahudee contrived to reach a much more distant portion of Adal,
the country of Chur-Chur, on the road to Hurrah from Shoa. Here
he also remained several weeks, receiving the greatest attention
and kindness, and finally was restored, as was also Hadjji
Ohmed, to the Negoos of Shoa, who rewarded their Adal
entertainers for their hospitality to British subjects.
It may be naturally supposed that the author of the “Highlands of
Ethiopia” was ignorant of these facts, but this is impossible, for
Mahudee, who had visited Chur-Chur, was reinstated in his
situation as horsekeeper to Major Harris himself, in return for the
interesting information it was supposed he could give of the little-
known country where he had been living.
CHAPTER XXIX.
Conversation on medical matters with the Negoos.—Of Guancho.—
The State prison.—The construction of its defences.—Good
medicine for captives.—Its probable effect.—Of the Gallas.—Their
invasion.—Of the Gongas.—Abyssinian slaves.—Conclusion.
A very singular circumstance connected with our conversation
respecting the health of the brother of the Negoos was, that neither
Bethlehem or myself recollected at first, that all the near relations of
Sahale Selassee were incarcerated in prison, according to ancient
Abyssinian custom, and which, I believe, was also practised in the
kingdom of Judea, to secure, by this cruel policy, the monarch from
personal danger, and the country from the evils inflicted by civil war,
that might otherwise arise by the ambition or simplicity of the other
branches of the Royal family, either acting itself or admitting of being
acted upon by the arts of others.
The Shoan prison for these unfortunates is a high conical hill,
called Guancho, situated midway between Aliu Amba and Farree,
and is the residence of the Wallasmah Mahomed, who fills the office
of State gaoler, as well as collector of duties upon that frontier of the
kingdom. Here, at the period of this interview with the King, were
confined five princes of the blood Royal, some of whom had been
prisoners for as many as thirty, or thirty-four years.
From personal inspection of their apartments, an opportunity
afforded to no other European besides, I can state that the close and
rigorous confinement, said to have been imposed upon these
captives, is much exaggerated; and, although the separate sleeping
apartments at night were not more than seven feet in all their
dimensions, still they were only composed of sticks, such as the
common garden rods for raising peas in England, and a strong man
leaning hard against them must have fallen out through the wall of
his cell. Only two of the royal prisoners wore chains; these were on
one hand and leg of the same side, and were long enough to admit
of the freest motion. A long-thatched wort bait, or meat-house,
contained their families; for not only did the King remember his
captive brethren on days of festival, by sending them oxen, and
honey-wine, but they were allowed to marry, and their wives lived
with them in their confinement. I took a ground plan of the whole
establishment, and the Wallasmah, who was too old to accompany
me on my survey, when I was in the only place that looked like a
dungeon at all, a vault about twenty feet square, cut out of the
summit of the hill, stamped several times upon the roof to intimate
that his sitting-room was over this secure place. In this dismal
dungeon, however, no person had been confined for the last six or
seven years, although it was being then prepared, by a second door
being put up, for the occupation of the unfortunate Samma-negoos,
an ex-frontier governor, who had assisted his brother, a denounced
rebel, to escape to Argobba, where he is now entertained by the
Mahomedan Prince of that country, Beroo Lobo. When I visited
Guancho, this prisoner occupied a small den of sticks, not four feet
wide in any direction, and his hands and feet were chained close
together, so that his removal to the larger subterranean cell will, at all
events, afford him some opportunities of exercise, though he will
then be deprived of light and fresh air.
Although, therefore, the Royal prisoners did not enjoy life in a
valley of delight, they certainly did not drag out a miserable existence
upon the hill of despair. This would have been adding unnecessary
cruelty to an exigency of State policy; an evil that would, I am
convinced, have long before corrected itself, by the frequent escapes
that would have been attempted, especially in a place that afforded
such opportunities for obtaining personal freedom. An Abyssinian
Baron Trenck would only have to wrench open the thin bar of soft
iron which constitutes fetters in that country, and by three successive
jumps through, not over, as many fences of rotten sticks, he would
be as free as the wildest Galla, into whose country a walk of a few
hours would take him. I did not show any lucifer matches, for I
recollected that the Portuguese traveller Bermudez, had been
confined in this very prison, and I did not know whether an act of
incendiarism might not at a future time be available as a means of
escape; for, it must be understood, at the time of this visit I had been
threatened if I attempted to leave Shoa with the Embassy on their
return to Aden, to be confined in Guancho, so desirous was the
Negoos of detaining me with him.
Guancho, the State prisoners, nor the anxiety of his Majesty that I
should remain in his service, can be entered into now; this is
anticipating the occurrence of events, the relation of which must be
excused from the increasing contraction of my limits, that prevents
me from holding but a little more pleasant converse with my reader,
who, I hope, so far has been conducted with an amused interest
through the scenes of Dankalli and Abyssinia life, in which I was a
participator during my sojourn in those countries.
Recalled to a recollection of the circumstances of the condition of
the Royal relations, by a remark of the Negoos, Bethlehem turned to
me, and commented upon the sanitary observances I had been
recommending for the benefit of my supposed patient, as he was a
prisoner, and I then learnt, that the Negoos was consulting me upon
the subject of a disease, to which he was himself subject. The
symptoms that he had detailed plainly indicated a great
determination of blood to the head, and among other things which I
had suggested as preventives of the occasional giddiness, dimness
of sight, &c., which was complained of, was frequent exercise by
walking, and recommended that this should be for some distance
regularly every morning and evening. It was this which had led the
Negoos to make some playful remark in his character, as his
brother’s representative, that this indeed would be a pleasant
medicine for him, and which reminded the interpreter that if I went on
prescribing in that way I might say something unpleasant to the
Royal ear. Perhaps the unconscious shrug, with which I
acknowledged our error operated upon the mind of the monarch
more than any direct appeal that I could have made in favour of his
unhappy relatives, and expressed more real sorrow than the cold
interpretation of Bethlehem could have conveyed.[13]
The monarch soon after changed the subject by alluding to the
bad state of my own health, and of the necessity of my remaining
quiet in Aliu Amba until the termination of the wet season, which was
expected about the middle of the present month. He did not forget to
recommend to me the study of the Amharic language during that
time, so that in the next expedition against the Galla, to which he had
already invited me, I might be able to converse with him. The
Negoos did not detain me much longer, but after telling me not to
miss seeing the Muscal (at Debra Berhan on the 24th), if it were
possible for me to come, he dismissed me, glad enough to escape
from the fatiguing interview.
I returned to Musculo’s house a great deal too tired to think of
going on to Aliu Amba directly, but made up my mind to stay until an
hour or two before sunset, to arrive in that town just in time for bed,
and so escape the houseful of inquiring friends, who would have
thronged around me with compliments and congratulations on my
return.
To amuse me some portion of the time, Musculo introduced three
of four slaves who had been brought from the more interesting
countries around Shoa, and none of whom, as regards their political
relations with that country, demand a more particular notice than the
Gallas. These appear to surround Shoa on every side, except
towards the north, where the Amhara inhabitants of the Argobba
appear to have their country in that direction, continuous with the
Shoan province of Efat; but even here a narrow belt of debateable
land, by the mutual jealousies of the rulers of the two kingdoms, is
left to the undisputed possession of some unsettled Adal Galla
tribes.
I have several times, in the body of this work, represented these
people as being the mixed descendants of the Dankalli and Shankalli
people, and although this descent has been modified in some
situations by contiguity to nations differing very considerably, both
physically and morally, from each other, still all the numerous tribes
that stretch on the eastern side of the table land of Abyssinia, from
the neighbourhood of Massoah to an unknown distance in the south,
speak one language, and practise nearly similar customs. The first
disputed question respecting the Gallas is their origin, which is
generally supposed to be foreign to the continent they now occupy,
and from the name Calla resembling a Hebrew word signifying milk,
it has been presumed that they were a white people of that nation,
who have become changed in colour by a long residence in their
present inter-tropical possessions. Modern travellers continue in
supporting this supposition, but in recording my dissent I ask no one
to adopt my opinion, I owe it to my readers to state my ideas upon a
subject I have studied a little, and upon which I presume they require
information. It is not, therefore, to attract attention by opposing
received opinion, which I would much rather avoid, but for the sake
of exciting discussion among abler men than myself, that I here
throw out suggestions respecting the Gallas, as on other subjects I
have done before.
The origin of the name Galla, from the Arian word calla, black,
appears easy and natural, and I have therefore adopted it, but shall
feel greatly indebted to any learned ethnologist who will correct me if
I am in error. The country their presumed parents occupied, is that in
which, from its situation, no other complexioned people could reside,
whilst that law of nature continues to exist which has imposed a
black skin upon men living in a very hot country.
We find, however, these so-called blacks in geographical
situations, quite at variance with that betokened by the dark colour of
their skin, and more particularly upon the elevated plateau of
Abyssinia, the natural country of the pale yellow Gonga, where their
appearance presents an apparent anomaly, which, fortunately,
history enables us to explain. The first intrusion of the Amhara I have
in another place endeavoured to show was in the time of the
Egyptian king, Psammeticus, and to trace their history, in connexion
with the changes consequent upon their colonization of the left
banks of the Abi and the Abiah, would be most easy and interesting;
for the present generation possess sufficient documentary evidence,
to supply the necessary materials; but until some indefatigable
scholar takes upon himself this task, I have no hope of seeing that
obscurity dispelled which hangs over the earlier history of mankind,
and which is intimately connected with the earlier history of
Abyssinia. With this part of my subject, however, at present we have
nothing to do, and must call attention to the fact, that the first
recorded appearance of the Galla in Abyssinia, as hostile invaders,
was in 1537, during the reign of the Emperor David, otherwise called
Onag Segued. By this must be understood that it was at that time
they first found themselves able to assert their independence. A
more favourable opportunity could not have been afforded them than
that offered, when the Mahomedan King of Adal, Mahomed Grahne,
conquered and overran considerable portions of the ancient empire.
To the distractions and misfortunes that then harassed the Christian
Court the Gallas contributed, led on by sheer destiny, I believe, for
they quietly took possession with their herds of the countries that
had been devastated during the long civil sectarian war which, at the
time of Grahne, had assumed a national character from the divisions
of the Christian and Mahomedan Amhara, being then under two
distinct monarchial governments. These two kindred people mutually
destroying each other, were unable to offer any resistance to the
lawless and barbarous intruders who were alone benefited by the
struggle for supremacy between the professors of these two faiths.
The Adal conquerors, however, lost a great deal more by the war
than the defeated Christians of the table land; for occupying a
country of much less elevation than Abyssinia, the Gallas naturally
located themselves first upon the lands so much more suited to their
habits and constitutions, and accordingly, the Dankalli, closing from
the north, whilst the Shankalli came up from the south, their progeny
soon swept from the face of the country their Amhara predecessors;
and the red man of America retreats no faster before civilization,
than on this coast of Africa, the latter has been extinguished by the
advance of the barbarian Gallas. Only one town remains of the once
mighty kingdom of Adal, the city of Hurrah, the former capital of
Mahomed Grahne, before whose time Christianity was here at least
tolerated and professed by numbers of its inhabitants. Within the last
century another lingering remnant of this population of Adal has
been entirely driven out. Owssa, now exclusively Mahomedan
Dankalli, was formerly the capital of Amhara kings of Adal, and the
traditions of the present occupiers record the late residence in that
country of a Christian population. After the death of Mahomed
Grahne and the expulsion of the Jesuits from Abyssinia, the attention
of its princes was first directed to the increasing evil of Galla
intrusion, and they then endeavoured unsuccessfully, to recover
those portions of the table-land upon which they had established
themselves.
It is admitted that the Gallas entered Abyssinia, through the
natural breach in its surrounding rampart on the east, where the
denuding operation of the Hawash has constructed a favourable high
road for the journeyings of a nation. Had a similar facility existed to
the south, such as would be afforded, for example, by the débouché,
of a river from the table land in that situation, we may be assured
that the national integrity of the Gonga people, who, in the north,
were unable to contend against the intruding Amhara, would have
found it very difficult to contend with the more warlike Galla; yet who,
it will be found, have made less impression there than in any other
situation upon the whole table land.
It appears that Fatagar, Efat, Shoa, then Damot, (which at that
period extended to the south of the river Abi,) were successively
taken possession of, a succession of conquests which prove that the
course of the Hawash, was the principal natural direction this people
took in their wanderings.
In Shoa and Efat they appear to have been early civilized. One of
the most characteristic traits of the Galla people is, the facility with
which they appear to adopt the religious creeds of their neighbours;
and the adjoining kingdom of Amhara, the central stronghold of the
Christian religion, afforded numerous opportunities of conversion,
and perhaps other favourable circumstances then existed of which
we are now ignorant; but the result has been a closer amalgamation
of the Gallas with the Amhara people in Shoa than, I believe, any
other country of Abyssinia presents. Whilst, therefore, an
exceedingly corrupt dialect of the Amharic language is there spoken,
the dark colour of their skin attests their close consanguinity with the
Galla invaders, coming from the low hot country immediately at the
foot of the Abyssinian scarp in this situation.
The Galla, physically speaking, are a fine race of men, tall,
muscular, and well formed. In the colour of their skin they vary
considerably, as may be supposed, from the differences of situation
and of neighbourhood in which they have located themselves. The
Edjow Gallas, to the north of Angotcha, are, I understand, of a lighter
colour than the real Amhara or red man, but it is probable that some
mistake exists as regards this statement. The Gallas of Limmoo are
very dark-coloured, but they live in a country considerably more
elevated than that of the Edjow Gallas. The Shoans themselves, who
are considerably more Galla than Amhara, are a very dark brown,
although several light red individuals, not born in Shoa, but more to
the north, as I was told by Sheik Tigh, are to be found among them.
In the expression of the Galla countenance there is that which
reminds the observer more of their Shankalli than of their Dankalli
origin. The form of their heads is long, the sides being flat, with very
contracted but not receding foreheads. The lower parts of their faces
have the full negro-form development of the lips and jaws, although
the teeth are regular and well set, without the inclination forwards I
have observed in several negro skulls. Their hair is coarse and
frizzly. It is generally worn in long narrow plaits, that hang directly
down upon the neck and shoulders. In Shoa it is customary to dress
it with considerable care, and it is then sometimes arranged in most
fantastic forms, the head being adorned all over with numerous small
collected tufts, and at others, three monstrous heaps of hair on the
sides and top make the head and face look like a huge ace of clubs.
Their natural dispositions are very good, and their courage is
undoubted.
It is very interesting to remark how readily the Galla appear to
adapt their national habits to the circumstances in which they are
placed. This seems to be a kind of instinct in man, or perhaps is an
element of that moral development which seems to determine those
occasionally mysterious inroads of a new people, who seem to have
sprung up at once to exert the most extensive changes in the history
of nations, and which then subsides again for another term of ages.
Such was the appearance of the Mongols in Asia, and of the Goths
in Europe; such was the appearance of the Arabians after Mahomed;
and such are the Gallas of the present day, who are gradually
appropriating to themselves the whole of the Abyssinian empire. This
moral principle, however, whatever it may be, seems to promise an
abundant harvest of converts to the zealous and intelligent
missionary, who shall first appear as the professed apostle of
Christianity among them.
Besides the Gallas whom I saw at Musculo’s, were several
Zingero and Kuffah slaves, and as these are the principal
representatives of the Gonga people, of whom I have frequently
spoken, I shall take this opportunity of more particularly describing
them. The Gongas are a mysterious people, of whom rumours alone
had reached the civilized world in the remotest antiquity, and the
same obscurity continues at the present time to hang over this
interesting and secluded nation. With the evidence I collected during
my travels in Abyssinia, it will not be presumption in me to call
attention to a few facts that appear to me calculated to throw some
little light upon this subject, and which may probably excite a greater
desire to become better acquainted with the hidden secrets of man’s
history contained in the heart of Africa.
The Gongas, in the era of the celebrated Egyptian king
Psammeticus, occupied the whole table-land of Abyssinia. Neither
Amhara, or Galla, had, as yet, appeared upon their naturally
defended and very extensive fortress. In their social institutions the
great principle of foreign policy, was the exclusion of strangers; and
their isolated situation, easily enabled them to effect this. One
character of civilization, the geography of the desert-surrounded
table-lands of Africa, is eminently calculated to prosper and promote,
that peculiar social condition, the consistency and continuance of
which, requires little or no intercourse to be kept up with the rest of
mankind; the isolated members of which, live contented among
themselves, uninfluenced by wants which could only be gratified by
the products of other lands. In such African communities, no inland
seas, or navigable rivers, afford that facility of intercourse which is
enjoyed (as it is presumed) by the inhabitants of more highly
favoured countries. Protected also from foreign invasion by vast and
almost impassable deserts, individual enterprize could scarcely be
tempted to keep up a communication with a people so situated,
provided that they adhered to the principles of contentment, and did
not allow themselves to be seduced into a desire for foreign luxuries;
an unwise indulgence in which, first leads to molestation from
commercial intruders; who, breaking up the seclusion, open a path to
military invasion, which usually ends in the loss of country and of
personal freedom.
We hear of the Gongas in ancient history under various names,
but they were principally characterized by the cautious manner in
which they communicated with those merchants, with whom nature
imperatively commanded them, at least, to have some intercourse to
exchange the productions of their country, for what was an absolute
necessary of life to them, and of which they had no supply but from
abroad; I need scarcely mention, that this was salt. In return for this,
it appears, that gold was principally given to the traders; and for
ages, this commerce was carried on, with no more communication
than was necessary, through the medium of the following practice.
“This country of Sasu is very rich in gold mines. Every year the King
of Axum sends some of his people to this place for gold. These are
joined by many other merchants; so that, altogether, they form a
caravan of about five hundred people. They carry with them oxen,
salt, and iron. When they arrive upon the frontiers of that country
they take up their quarters, and make a large barrier of thorns. In the
meantime having slain and cut up their oxen, they lay the pieces of
flesh, as well as the iron, and salt, upon the thorns. Then come the
inhabitants, and place one or more parcels of gold upon the wares,
and wait outside the enclosure. The owners of the flesh, and other
goods, then examine whether this be equal to the price or not. If so,
they take the gold, and the others take the wares; if not, the latter still
add more gold, or take back what they had already put down. The
trade is carried on in this manner, because the languages are
different, and they have no interpreter; it takes about five days to
dispose of the goods which they bring with them.”[14] Heeren, in his
Historical Researches, connects the country where this system of
barter was practised, with that of the Macrobians, or long-lived
Ethiopians, mentioned by Herodotus. By an ingenious conclusion, he
supposes that the altar or table of the Sun which characterized the
latter people was the market-place, in which, at a later day, the trade
with the strangers was transacted. My observations have also led
me to the same conclusion, but I am able more distinctly to
authenticate this, and to suggest additional and more direct evidence
of its being the actual fact.
The worship of the Gongas, which has continued to the present
time, is the adoration of the river that flows through their country, as
being part of the sacred Nile. The Abi, or Nile of Bruce, is
worshipped by the modern Adjows, whilst the Gibbee, or Abiah, is
the object of a similar devotion among the Pagan Gongas of Zingero,
and of Kuffah. We are enabled from our knowledge of the former
river to presume, that its singular course determined in the first
instance, a reverence, which, when the increasing encroachments of
foreign foes had made this river a convenient defence to the pressed
Gongas, was soon elevated to the character of a protecting deity.
That its singular course should have thus attracted attention arises, I
believe, from the circumstance of its encircling an extensive
province, and going around it, as the sun was supposed to revolve
around the earth. The zodiac, or track of the sun through the
heavens, was typified by the form of a serpent, and this I have
always understood to have been the source of that serpent-worship
which characterized so many of the earlier and more civilized nations
of the earth. In no country, was this idolatry more prevalent than
upon the plateau of Abyssinia, and Arwè, the great serpent, it will be
recollected figures considerably in the earlier history of the Amhara,
who appear to have in some measure adopted the religion of the
Gongas, when they took possession of the countries upon the left
hand of their father, their king, their sun, by all of which names, it is
usual, even at the present day, to designate the river Abi.
The great serpents of classic mythological history, the Hydra, the
Python, and others unnamed, destroyed by Apollo and Hercules, all
allude evidently to the worship of the serpent in Africa being
superseded by that of the sun. The relation of these gods to that
luminary is generally admitted, and Hiero Calla, fortunately for my
derivation of the word Galla, the sun of the blacks, is the
interpretative analysis of the name of Hercules. In the modern
Dankalli language no other word is used for sun but Hiero, and it
enters into the name of several names of places; Hyhilloo, the scene
of the celebrated battle between the forces of Lohitu and the Muditu,
is translated by the Dankalli to mean the hill of the sun.
The head of a sculptured Hercules is invariably portraited with the
frizzly hair of the Dankalli, whilst antiquarian ethnologists will be
interested to observe the persistance of national character preserved
in the flowing locks and ample beard usually given to Jupiter, his
European counterpart.
That which increased the celebrity of the northern portion of the
table-land of Abyssinia, and established the superiority in dignity of
its stream, was the circumstance of its flowing through the lake
Tzana or Dembea. No little light breaks upon the subject when it is
understood, that the literal interpretation of these two words in very
different languages, is the same, both signifying the lake of the sun.
Dembea, let me observe, is a word in use in Abyssinia that belongs
to the same language as Abi, Assa, Galla, Nil, and others, that to
avoid confusion, I have called Arian. That so many proper names,
should all be derived from an Asiatic language in a country where no
representatives of the modern people who speak it can now be
found, is only to be accounted for, by supposing that the African
original of the Arian family of man yet continues in some of the
secluded oases of Intra-tropical Africa, to reward by their discovery
future enterprize.
Bahr Dembea, or the Lake of the Sun, would give a very
appropriate designation to the plateau upon which it is found. It was
that, and the course of the Abi, which occasioned the country visited
by the messengers of Cambyses to be called the Table of the Sun. It
was also the presence of these singularly situated geographical
features, and their supposed reference to the sun’s track in the
zodiac, that determined the reputed sanctity of this portion of
Ethiopia in the classic ages.
The connexion of the ancient Persian empire with its Ethiopic
tributary kingdoms, did not extend so far as the country of Sasu, and
the fate of Cambyses, in his attempted conquest of that country,
would be, I have no doubt, an instructive lesson to his successors.
The claims of these monarchs to supremacy in Ethiopia appears, in
fact, to have been founded upon former family connexion with some
father-land in Africa, not situated upon the plateau of Abyssinia, then
inhabited by the Gongas, but in another desert-surrounded country,
of the same character; probably, that which surrounds the sources of
the Bahr ul Abiad.
The African origin of other ancient nations can also be most easily
demonstrated, and the historical accounts of their descent from
gods, which have come down to us, although they consist of
exaggerated and distorted relations, in consequence of having been
derived from the ignorant translation of hieroglyphical records, in
which it would appear that the earlier history of Africa was preserved;
still we are able to gather from these mythological enigmas
everything that is necessary to connect their origin with a common
centre of divergence, which I believe to have been the country
around the sources of the Nile.
In the same manner the worship of the rivers in India, and of the
dragon monster in China, seem to have originated from Ethiopia; the
emigration which carried the first colonies of serpent worshippers to
these countries having probably flowed in a direction from the south,
as Europe and Western Asia appear to have been civilized by
colonists from the north of the same point of dispersion.
It is most interesting to trace the intimate connexion at an early
period of the, at present, widely separated and even physically
distinct varieties of man; and did not a cautious policy restrain me, I
would attempt to demonstrate the original unity of nations now the
most dissimilar upon novel evidence, which, to be satisfactory to
others, must, however, receive farther corroboration than my own
individual observations.
One illustration of the light African explorations promise to throw
upon this subject I cannot refrain from advancing, as it is such a
striking evidence of the presumed fact of even ourselves having
originated from a colony of African emigrants; and that the ancient
British temple of Abury, or Abibury, near Stonehenge, derived its
name from the same religious worship being there celebrated as was
once general on the plateau of Abyssinia, and which, in fact, is so
called from exactly the same cause. The deductions of classical
learning materially assist a traveller, whose pursuits, so different to a
closet student, do not allow him to assume the character of a learned
critic or commentator. Dr. Stukeley, known by his inquiries into the
ancient religion of the Druids, has proved, I think incontestably, the
true character of the temple at Abibury, and demonstrates it to have
been constructed in the form of a serpent, bearing upon its back a
circle. He referred the religion, that directed such a form to be
assumed in the sacred architecture of this people, to an Egyptian
origin, and freely speculates, in consequence, upon the African
origin of our ancestors, which is asserted by our most ancient
historians, but who have been in consequence considered to be
apocryphal. In these traditions it is affirmed that Britain was first
inhabited by a celebrated descendant of Shem, singularly enough
the same, who is considered by biblical ethnologists to have been
the common father of no less a respectable people, than the modern
Dankalli; Affer, the son of Abraham, having led a colony of Africans
to our shores, where he introduced the worship of the sun, and
established the religion of Druidism. I recommend to my reader the
perusal of Stukley’s work upon “Abury, a Temple of the Ancient
Britons,” and then to compare the parallel, but more magnificent
temple of nature upon the plateau of Abyssinia, where the serpent
Arwè, or in profane language the river Abi, bears upon its back the
lake of the sun, most curiously identifying the peculiar worship of that
luminary by the ancient Ethiopians with the same adoration which
was professed by the Druids in Britain, but who, from their situation,
were obliged to construct the winding avenue of stones at Abibury to
represent the same mystical hieroglyphic of the serpent and the sun.
The name given to this work was Abi, the father, or king, as it was
also of the river-symbol in Abyssinia; hence the name Abibury, the
latter portion of which word is of Saxon origin; and added,
subsequently to the decline of Druidism.[15]
Returning to the Gongas and their connexion with the Sasu of
Cosmas, it is singular to observe in what manner the seclusive
integrity of that country was first sapped, and then in a great
measure overthrown. Within the last two centuries, the Adjows of
Northern Abyssinia, the representatives of the Gongas in that
situation, were said to continue the original practice of their fathers
with respect to commercial transactions. But these must have been a
tribe now extinct, as, from what I can learn, it is only in the extreme
south where the custom is still persisted in, and it is among these
that the most ancient authentic record (uninspired) of the
antediluvian world will be found. It is here, too, that the original name
of this people, Sasu, is preserved in the modern word Susa, of
whom, as a nation, we scarcely possess any information more than
sufficient, to warrant the mere assertion that such a people now exist
highly civilized, and using a peculiar written character dissimilar to
any with which the literati of Europe are acquainted.
Of the Sasu traders in the time of Cosmas, we are told they
carried with them oxen which, on their arrival in the country, they
killed, and hung up the raw flesh on the thorns, as a kind of
merchandise. It will be remembered that I have previously stated the
intoxicating effects of this kind of food upon the Amhara, and I have
therefore no difficulty in supposing that the Gongas were tempted by
this kind of dissipation into the intercourse with the traders, just as in
modern times, “fire water” for Indians, and opium for the Chinese are
employed to effect a similar object. This receives further confirmation
from the fact, that the secluded Gongas of the present day live
entirely upon vegetables, the ensete plant and grain forming the
principal food. In Zingero and Enarea, broken in upon by the
Mahomedan and Christian religions, the inhabitants have adopted
the use of animal food, but even among them a party of the older
faith exists who continue the original mode of living of their fathers,
and who are contemptuously styled, for that reason, “grain eaters.”
The Gongas that I have seen are of short stature, not exceeding
five feet four inches, are delicately made, and of a pale yellow
complexion. The aperture of the eyelid in some were quite straight,
but in others it was obliquely divided. Their hair was straight and
strong. A triangular formed face, the forehead being low and long,
and the chins very pointed. I could not convince myself, as I looked
at their whole appearance, but that they were of the same race as
the Hottentots of the Cape, differing only in so much as that the latter
are in a very degraded state. Many remarkable customs practised by
both nations could not have been merely coincidental; and one, that
of voluntary semi-emasculation, is too extraordinary not to be
referred to the same origin of imposition. Of the identity of the two
people there can be no doubt, and there is no ethnological fact I
observed during my journey of which I am so well satisfied as this.
The remains of this interesting people in Northern Abyssinia are
the Adjows and the Falasha, and if future travellers will expend their
resources in exploring Northern Abyssinia, in preference to the far
more important examination of its southern portion, they cannot
occupy themselves more advantageously to science than by
examining into the customs and characters of the Adjows. I consider
it would be a waste of time that could be occupied much better in
another direction, or I would, for my own satisfaction, visit the
country for this purpose; but as it is far from difficult and constitutes
an excellent probationary journey, I recommend aspirants for fame in
the field of African discovery to make this their trial excursion.
One more remark upon the Southern Gongas of Enarea, Zingero,
and Kuffah, and I must close this notice of a very interesting race of
man; and that is to explain the apparent anomaly of their country,
situated at such an elevation above the level of the sea as I presume
it to be, producing cotton and grapes in profusion.
The observations of that indefatigable and enterprising traveller,
Dr. Beke, has proved that the river Abi, after flowing a distance of
scarcely one hundred and thirty miles, has excavated a valley five or
six thousand feet below the general level of the table land, whilst the
opposite summits of the bounding sides are distant between thirty
and forty miles. We may look in vain over every portion of the known
world for a similar effect of denudation, and this again illustrates the
wide field of novel facts which is promised to science, by an
examination of the unknown interior of Africa.
On the artificial terraces and natural slopes of these extensive
valleys the vegetables of all climates can be successfully cultivated,
and the theoretical centre of successive elevations from whence,
according to the hypothesis of Linnæus, all vegetation spread over