Download Paper

CHARACTERIZATION OF CERAMIC MATERIALS BY
ACOUSTIC EMISSION
M. Roth, E. Dul'kin and E. Mojaev
Department of Applied Physics, The Hebrew University, Jerusalem
91904, Israel.
The acoustic emission (AE) method has been used as a powerful
characterization tool for studying the structure and preoperties of
technologically important ceramic materials, such as high-Tc
superconductors and relaxor ferroelectrics. With regard to the superconducting YBCO (YBa2Cu3Ox) ceramics and BISCCO (Bi2Sr2CaCu2Ox)
composite tapes, we show that by monitoring the acoustic emission
bursts it is possible to measure the temperature hysteresis of phase
transitions and to reveal their order, to determine the temperature of
maximal oxygen absorption as well as to measure the lower critical
magnetic field Hc1 and the full penetration field under electrical current
transport. The restored strain energy in PbZn1/3Nb2/3O3 (PZN) and
9%PbTiO3-doped PZN (PZN-9%PT) relaxor crystals has been studied
by means of AE as well. Two types of AE activity signals have been
recorded: (i) related to temperature- or electric field-induced
macroscopic
phase
transitions
and
(ii)
associated
with
formation/disappearance of intrinsic polar nanoregions. Monitoring of
AE under varying [001] electric fields has allowed a unique in situ
observation of a low-field (1 kV/cm) irreversible orthorhombic-to-MC
phase transition within the morphotropic phase boundary region of PZN9%PT. The cumulative results demonstrate that acoustic emission
method is an indispensable tool for studying the structure and properties
of ceramic materials.
1. Introduction
Currently, there is a growing interest in the class of phenomena
whereby transient elastic waves (in the ultrasound range) are generated
by the rapid release of energy from localized sources within a material.
It is known as acoustic emission (AE), which is associated with
structural reconstructions within the solid state under the influence of
external forces [1]. AE is a nondestructive method for investigating the
kinetics of defect production, such as movement and accumulation of
133
dislocations accompanying plastic deformation and their annihilation,
twinning and movement of twin walls and of phase boundaries (PB) as
well as the generation and propagation of cracks in solid state materials
subjected to mechanical stress [2].
Another extensively studied source of AE includes martensitic
phase transitions (PT) in metals and alloys under thermal ramping [3-6].
The greatest contribution to the AE accompanying martensitic PT is
made by processes associated with the generation (or annihilation) and
movement of dislocations originating from crystallographic mismatch
between the original and the new phases at the PB. The AE activity, N
(sec-1), measured by means of a piezoelectric transducer reveals the
temperature of the martensitic PT. It can be also a measure of dislocation
density changes in course of durable thermal cycling through the PT,
called phase work hardening (PWH).
Martensitic-like AE responses are observed in course of phase
transitions in ferroelectric and ferroelastic crystals as well. By
employing the AE method it has been possible to detect all ferroelectricferroelectric-paraelectric PTs in the BaTiO3 and SrTiO3 ceramic
materials [7]. AE has been applied to studying the dependence of the
dislocation density in PbTiO3 crystals on the PB orientation, in relation
to the direction of the thermal field gradient; an AE maximum has been
found when the PB is oriented at about 45° angle relatively to the
direction of thermal field gradient [8]. Similarly to the martensitic PT in
NiTi-based alloys [6], PWH has been detected in PbTiO3 crystals [9] as
well as in (Na1-xLix)NbO3 binary solid solution ceramics [10] during
prolonged thermal cycling. In (Na1-xLix)NbO3, AE accompanies two
high-temperature ferroelectric-ferroelectric and ferroelectric-paraelectric
PT, which have not been detected by the traditional dielectric method
[11]. A similar uniqueness of the AE method has been demonstrated in
the case of Ba0.85Sr0.15TiO3 posistor ceramics, where the PT cannot be
detected by resistance measurements [12]. On the other hand, in PbZrO3
and PbHfO3 crystals, a strong AE signal accompanies only the
ferroelectric-paraelectric PT due to the corresponding incoherent PBs
[13]. A similar AE effect is observed in Pb(Fe0.5Nb0.5)O3 crystals
through the ferroelectric-ferroelectric-paraelectric PT [14]. The
appearance of PWH is also observed in the relaxor ferroelectric
Pb(Mg1/3Nb2/3)O3 crystal through a diffusive PT [15]. It is well known,
134
that high-Tc superconductors also undergo PTs, including the
superconducting transition. Some of the high-Tc superconductors have a
perovskite-like crystallographic structure, like the ferroelectric crystals,
and they undergo structural PTs similar to the martensitic-like transitions.
This explains the extensive efforts that have been made during the last
decade to apply the AE method to investigating the high-Tc
superconducting phenomena.
The phenomena described above show clearly that AE is an
indispensable method of studying many kinetic features of the structural
PTs in solid state materials, such as determining the Tc values and
characterizing the ΔTc hysteresis, identifying the order of PTs, estimate
the degree of inter-phase coherence at phase boundaries, defining the
degree of hardening based on the PWH data [16], etc. Below, we review
the main results recently obtained with the best characterized
YBa2Cu3Ox (YBCO) and Bi2Sr2CaCu2Ox (BISCCO) high-Tc
superconductors as well as Pb(Zn1/3Nb2/3)O3 with PbTiO3 (PZN-xPT) as
a representative of relaxor ferroelectrics.
2. Experimental
The common experimental procedure of AE measurements is
simple, and the basic setup is presented schematically in Fig. 1. Due to
an external force of mechanical, thermal or electromagnetic nature, the
investigated material produces elastic (ultrasonic) waves, which are
converted to electrical signals by direct coupling to a piezoelectric
sensor. Then the output of the piezoelectric sensor is amplified through a
frequency-selective low-noise preamplifier, filtered and additionally
amplified through an amplitude discriminating amplifier and converted
to voltage pulses through an amplifier-multivibrator, which are counted
and displayed in time units. Usually, three parameters of the AE are
being measured: (i) total signal amplitude ∑A, (ii) total number of
pulses ∑N and (iii) activity ΔN/Δt = N (s-1). The latter parameter is
most commonly determined.
135
Fig. 1. Basic setup for AE measurements under mechanical, temperature or
electromagnetic loading
It is noteworthy that both in the case of low- and high-temperature
experiments it is undesirable to subject the AE sensor to nonambient
temperatures. Therefore, a quartz glass waveguide is usually introduced
as a buffer transmitting the ultrasonic waves from the studied material to
the AE sensor [2]. In the high-temperature setup, the sample is glued
with a high-temperature epoxy resin to the polished end of the fused
quartz acoustic waveguide. A piezoelectric PZT-19 ceramic sensor is
glued to the opposite end of the waveguide and connected to a 500 kHz
band-pass preamplifier. The sample comprising the top part of the
waveguide is mounted in a resistance furnace. A Ch/Al thermocouple is
attached to the waveguide near the sample. Two pinned rods connected
to an external differential dilatometer are monitoring the sample size.
The AE activity N and thermal expansion ΔL can be simultaneously
measured during heating and cooling at a rate of about 1-2 K/min with
or without an oxygen flow. In the low-temperature setup, the
configuration of the experiment is the same, but liquid nitrogen vapor is
used for cooling the sample and the temperature is monitored by a Cu-K
thermocouple attached to the waveguide near the sample. An induction
136
coil is added to measure the magnetic susceptibility χ at the frequency of
1 MHz. The AE activity N , susceptibility χ and thermal expansion ΔL
are then simultaneously measured during heating and cooling at a rate of
about 1-2 K/min. In another low-temperature setup, the sample is glued
to the bottom end of the acoustic waveguide, while the piezoelectric
sensor is adhered to its top end. A similar 500 kHz band-pass
preamplifier is used. The sample with the lower part of the waveguide is
submerged into liquid nitrogen. DC electric current is applied through
two silver epoxy contacts on opposite side ends of the sample. The
liquid nitrogen Dewar flask is mounted between the two poles of a DC
magnet. AE activity N is then measured at 77K in the presence of an
electric current flow or magnetic field.
3. Sintering and oxygenation of high-Tc superconductors
Practical application of oxide superconductors requires bulk
materials with greatly improved current-carrying capacities. Since it is
well known that the critical current density of sintered YBCO ceramics
appears to be limited by intergrain resistance, it is essential to control the
grain growth. During the process of ceramics sintering, spontaneous
grains growth is observed experimentally. In YBCO, the grains appear
in the temperature range of about 800-900ºC and they continue to grow
in size with a rate proportional to the temperature gradient in the
material volume [17]. The anisotropic expansion and contraction of the
grains during thermal processing produce an AE signal, which carries
information about the size of the grains formed. The AE activity of
YBCO displays three characteristic stages [18]. AE is initiated above
810°C, and a relatively sharp activity peak is observed in the 820-840ºC
temperature range. This is followed by a narrow second stage, where
essentially no AE can be detected. However, above 850ºC, the AE
activity reemerges and increases nearly linearly with further temperature
ramping. Just above this temperature, regular grain growth commences.
These results may be interpreted within the framework of a qualitative
model [19] suggesting that sintering proceeds in three stages. During the
first stage, the pellet of pressed powder shrinks, and the material density
increases. The shrinkage is accompanied by considerable mechanical
stresses in the powder generating AE. During the second stage, a fluid
glass phase, a "lubricant", appears at a higher temperature facilitating the
137
material’s further shrinkage. Such "lubrication" decreases the friction in
the system, which is confirmed by absence of AE. During the third stage,
grain growth starts due to recrystallization. Unstrained crystallites take
up material and grow into the neighboring strained (heavily plastically
deformed) areas of the same phase, being gradually increased in size.
This results in an increase in the area of the stressed intergranular
boundaries, which involves climb or cross-slip of dislocations as they
rearrange into the moving boundary. For larger grains, or larger
intergranular area, the relief of plastic deformation strain is accompanied
by an increased AE associated with the recrystallization process.
Sintering is only the first important step of the YBCO preparation.
The second necessary step is oxygenation, which is crucial for obtaining
the material in its superconducting state. Oxygen content determines the
actual superconducting PT temperature, or so-called critical temperature
(Tc). On cooling, after the sintering, of the initially tetragonal YB2Cu3OX
phase in oxygen atmosphere the supercondicting orthorhombic II-phase
(O-II) appears at about 650ºC. In this process, inflowing oxygen ions
engage in occupying some of the vacant sites to form O-Cu-O chains.
When the oxygen stoichiometric coefficient x ~ 6.5, most of the
alternating O-Cu-O chains are filled with oxygen. The Tc of O-II is close
to 60 K. The lattice strain associated with the incorporation of extra
oxygen ions and the consequent T→O-II phase transition is
accommodated by crystallographic twinning in the (110) plain. The 60K
O-II phase nucleates and grows gradually in the tetragonal matrix. On
further oxygenation, oxygen ions fill the vacant sites in the O-Cu-O
chains completely as x reaches the value of 7. The lattice parameter a
becomes half of that in the O-II phase implying that the O-II phase
transforms to a new phase, O-I, with the Tc increasing to about 90K.
Since the O-II→O-I phase transition is similar to the martensitic-like PT
in FE materials, AE is expected to be a suitable method of studying the
oxygenation processes in the YB2Cu3OX material. Our studies have
revealed that the AE has actually two peaks of N on cooling after
sintering: at around 650ºC and 605ºC [20]. The first peak corresponds to
the T→O-II PT, while the second is obviously attributed to the
maximum adsorption of oxygen associated with the O-II→O-I phase
transition.
138
AE exhibits interesting features also during sintering of
superconducting BISCCO tapes [21]. Tapes composed of the highest-Tc
Bi-2223 phase ceramics enclosed in silver cladding have been studied
most extensively. The specific feature of such tapes is that cracks arise
between primary and secondary sintering, due to intermediate rolling.
Therefore, the secondary sintering of Bi-2223/Ag tapes after rolling up
to a strain of  ~ 18% has been studied by both the AE and magnetic
susceptibility methods for comparison. During the second (post-rolling)
thermal treatment of the tape, a broad band of AE from is detected in the
temperature range from 570 to 660°C. This broad band can be
interpreted on the basis of other data obtained during in situ studies of
crack generation and healing in Bi-2223/Ag tapes [22]. According to the
BISCCO phase diagram, a liquid phase exists in the 400-660°C
temperature range. Magnetic susceptibility (χ") measurements have been
carried out during heating through this temperature interval. The results
show a clear narrowing of the χ" peak, which is characteristic of
enhanced electrical connectivity between the ceramic grains due to
healing of the cracks [23]. Consequently, the process of liquid-phase
healing of the rolling-induced cracks in the tapes can be regarded as the
source of AE.
On cooling the Bi-2223/Ag tape after secondary sintering at
730°C, an AE signal is detected below 230°C as well [24]. The AE
activity is weak near 230°C, but increases in intensity with decreasing
temperature down to room temperature. The appearance of AE signals
has been explained in terms of the relative deformation of the silver clad
and the ceramic core of the Bi-2223/Ag composite tape. The thermal
expansion coefficient (α) of silver and Bi-2223 ceramics are αAg =
20.5·10-6 K-1 and αc = 13.6·10-6 K-1 respectively, which implies stronger
contraction of silver on cooling. In view of the high plasticity of silver
(yield stress 65 MPa) and sufficiently large strength of the sintered
ceramic (yield stress 150 MPa), the silver envelope cannot contract the
underlying ceramics effectively and experiences considerable tensile
strain. The associated plastic deformation causes the formation and
movement of dislocations and is accompanied by AE. The observed
strong AE activity ( N values reach several hundreds of s-1 at RT), is
139
about an order of magnitude larger than that of martensitic phase
transitions in metals.
4. Superconducting phase transition in YBCO
We address now the results of AE measurements employed for
studying the superconducting PT in YBa2Cu3OX. Over a decade ago, one
of us has conducted a series of combined measurements comprising AE,
thermal expansion (ΔL) and magnetic susceptibility (χ) using a lowtemperature set up described above [25]. The results of these
measurements are shown in Fig. 2. The dilatometric curve has two
inflection points, at 83 and 92K, exhibiting no discontinuity. The first
inflection is accompanied by AE, the second - by a 'jump' of χ. The
relative change in χ is considerably large, about 50%, and the
superconducting PT width is about 1K. The magnitude of N in the AE
spectrum is too small to represent a 1st-order PT. These results imply,
therefore, that a 2nd-order structural PT takes place around 82 K, below
the Tc, in agreement with the results of ultrasonic wave velocity
measurements [26]. This structural PT can be understood in terms of the
interaction between ultrasonic phonons and conduction electrons, which
produces attenuation in the normal state and decay below Tc as the
electrons become paired [27]. The symmetry of the subsystem
associated with paired electrons may coincide or not with the symmetry
of the crystallographic lattice [28]. Therefore, the superconducting
transition is able to induce certain lattice instability and, as a
consequence, lead to a structural PT below Tc [28]. A fit of the BCS
theory to experimental results of elastic energy dissipation due to
electron-phonon interaction gives an average value of Tc  86.5K, while
the electrical resistivity measurements performed on the same YBCO
ceramic sample yield Tc = 94K. This provides an additional explanation
to our AE results indicating that a 2nd-order PT occurs at 82K, below the
Tc, in the YBa2Cu3OX material.
140
Fig. 2. Simultaneous measurements of AE, thermal expansion and
magnetic susceptibility in the superconducting PT range of
YBa2Cu3OX ceramics
5. Flux penetration and mixed state in YBCO
In his fundamentalc paper, Abrikosov has demonstrated [29]
that the magnetic field (MF) flux lines (FL) start penetrating the
type-II superconductor when the applied MF reaches the lower
critical value, Hc1. Above this field, each FL is accompanied by a
vortex of persistent current (supercurrent) surrounding a normal
resistance core. These FLs thread the material’s volume along the
applied MF direction. Upon entering the material, the FLs are
trapped by the pinning centers (defects, including grain boundaries,
twins, etc.), initially at the surface and then in the volume of the
sample. Under the transport current (I) condition, the repulsive
Lorentz force (IμH) acts on the FLs in such a way as to drive
them deeper into the material if they can overcome the opposing
pinning forces, and the mixed state begins to form [30]. On the
other hand, trapping at pinning centers forces the FLs to transfer
their kinetic energy to the defects, and the latter begin to vibrate
inducing the acoustic waves in the material as a whole. Excitation
141
of AE may be expected therefore as the MF penetrates the material.
Indeed, AE is observed in Nb-Ti wire under electrical current
transport [31].
Recently, we have applied the AE method to studying the FL
penetration process in high-Tc superconductors, such as YBa2Cu3OX.
The low-temperature experimental setup has been used. The selfinduced MF enhances as the DC current I flowing through the sample
wire increases (external MF is absent). Fig. 3 shows that two AE activity
peaks appear at 0.5 and 2.7A values of the DC transport current. The
first peak is less intense, and it is assumed to mark the onset of MF
penetration into the sample [32]. We have shown [33] that the magnetic
field H(r) induced by the transport current just outside the wire, or the
sample surface (r  r0), is given by
H
2I
ln r  r0 .
r0
(1)
This result implies that the MF tends to infinity, yet the flux can
penetrate the sample as deep as the coherence length, ξ = 1.2·10-9m [34].
Substituting the current value of 0.5A determined by AE (Fig. 3) into eq.
1 and taking (r–r0)  ξ, we obtain H  100 Oe. This value of the
magnetic field is in a good agreement with Hc1 = 90 Oe deduced from
the internal friction measurements [31]. Thus, AE is a very convenient
and useful tool for determining the Hc1 values in superconductors.
The second AE activity peak in Fig. 3 is about four times more
intense than the first peak, and it appears slightly above the critical
current (Ic). Obviously, the difference in the N magnitude is related to
the difference in the number of FLs. The first peak reflects only FL
pinning in the near-surface region, while FLs forming the mixed state
interact with numerous pinning sites at a deeper volume. As the
transport current reaches the Ic value, the Lorentz force overcomes the
pinning forces, and FLs start moving through the sample [30]. The FL
movement is accompanied by AE due to atomic movement, as in the
case of FL jumps in Nb-Ti wires quoted above. Moreover, the
movement of FLs is perpendicular to I and, therefore, active energy
losses sharply increase due to the evolution of Joule heat within the
normal core of the moving FLs. Such sharp increase of power losses has
142
been registered during electrical current transport through Bi-2223/Ag
tapes at I ~ (1.1-1.3)·Ic values [35]. Thus, AE can be successfully
applied also to detecting the current overload in high-power
superconducting circuits.
Fig. 3. AE activity as a function of electrical transport current flowing through
YBa2Cu3OX at 77K
6. Relaxor ferroelectrics
Single crystals of relaxor-based solid solutions, such as
PbMg1/3Nb2/3O3PbTiO3 (PMNPT) and PbZn1/3Nb2/3O3PbTiO3 (PZNPT), are known to exhibit exceptionally large electromechanical
coupling and piezoelectric responses when poled along the pseudocubic
[001] direction [36,37]. These unusual properties occur in the so-called
morphotropic phase boundary (MPB) region, which is related to a lowsymmetry monoclinic phase separating a rhombohedral PT-poor phase
from a tetragonal PT-rich phase of PZN-PT. In the case of pure PZN,
which is considered as a prototypical relaxor ferroelectric material with
a complex perovskite structure, Zn2+ and Nb5+ cations are present in the
correct ratio for charge balance, but they are randomly distributed on the
B-site of the ABO3 perovskite structure, which has only short range
order. These chemically ordered regions are expected to give rise to
locally quenched random electric fields which can be strong enough to
prevent the phase transition towards a true long-range polar ferroelectric
143
phase. Relaxors are in a way self-organized nanostructured materials
consisting of short-range polar nanoregions (PNRs) embedded into a
polar (PZN) or nonpolar (PMN) matrix. Recently, we have shown that
AE signal can be related also to the electric-field- induced phase
transitions in PMN [38]. In the following we report on detection of AE
activity in PZN and PZN-PT crystals (330.3 mm3 samples) under
varying temperature and applied electric field conditions.
Fig. 4. AE activity as a function of temperature during thermal cycling of a PZN
crystal; arrows indicate the heating and cooling directions
Fig. 4 shows the AE activity variation as a function of temperature
for pure PZN in the heating and cooling regimes. A sharp AE signal
with a thermal hysteresis of about 10K occurs around Tc  413K. This
signal corresponds to the structural phase transition from the cubic
(high-temperature) to rhombohedral symmetry, as originally described
by Kuwata et al. [39], and it shows the I-order character related to this
transition. The more striking and peculiar features are the two higher
temperature AE signals detected in the cubic phase at T  505K and
730K. These unusual anomalies are quite different from the AE activity
corresponding to the phase transition; they are broader (at least three
times) and do not show any thermal hysteresis. The highest temperature
144
peak (730K) corresponds to the Burns temperature, Td, associated with
formation of locally polarized nanoregions, or PNRs, in relaxors [40].
This result is also in agreement with our recent data on PMN [41]. The
505K peak, which we denote as T*, exhibits the same AE features as the
Td peak, e.g. in terms of the peak width, and we presume that it is also
related to the PNRs and especially to the initiation of increase of their
correlation length, or of size of the polar nanoregions. The intensity of
the AE activity at T* is three times stronger than that at Td reflecting,
therefore, larger elastic strains stored. Apparently, enhanced growth of
the PNRs occurs around this temperature, and the interaction between
larger PNRs of different polarizations generates intense transient elastic
waves during the reorientational motion of polar nanoregions. This
assumption is supported by an earlier analysis of the diffuse neutron
scattering data [42].
Fig. 5. AE activity as a function of temperature during thermal cycling of a
PZN-9%PT crystal; arrows indicate the heating and cooling directions
We now compare the AE response of pure PZN with a solid
solution crystal, especially with a PZN-9%PT composition
corresponding to the MPB. The AE activity for such compound is shown
145
in Fig. 5. PZN-9%PT undergoes two macroscopic phase transitions
evidenced by sharp AE signals. The peak close to 450K (with about a
10K thermal hystersis, similarly to the 413K peak in PZN) is I-order and
corresponds to an abrupt cubic-to-tetragonal (C-T) phase transition. An
additional transition is characterized by a 380K peak on heating and a
340K peak on cooling exhibiting a very large thermal hysteresis of about
40K. Similar features have been observed in course of low-angle light
scattering and Brillouin backscattering spectroscopy measurements. The
PZN-9%PT composition belongs to the MPB, and there is a competition
between the structures of different symmetries with free energies very
close to each others, which hampers the straightforward assignment of
the transition involved. It has been shown that the room temperature
phase of PZN-9%PT can be either monoclinic (MC) or orthorhombic or a
mix of both phases. For the sake of clarity we assume in the following
that the room temperature phase is orthorhombic being the limiting case
of the monoclinic MC-type phase. The strength of the AE response for
such tetragonal-to-monoclinic (T-MC) phase transition is two times
weaker than that corresponding to the C-T transition (at ~ 450K). The
large thermal hysteresis of 40K shows that the restored elastic strain
energy is small and may be indicative of a rather gradual growth of the
tetragonal into the monoclinic (MC) phase. We thus assume that there is
a continuous change of the polarization from the [001] tetragonal to,
ultimately, the [011] orthorhombic direction in the monoclinic (010)
plane, which is consistent with the diffuse I-order transition model. A
surprising result of the AE measurements during thermal cycling is the
persistence of the two high-temperature responses, namely the T* and Td
peaks, in the cubic paraelectric phase of the PZN-9%PT crystal at
temperatures very close to those of pure PZN. Only a weak increase of
Td is observed for PZN-9%PT (Td  736K). The T* value remains
practically constant, although it is expected to increase through the T* =
Tc + T relationship, with T increasing at higher PT concentrations [42].
This discrepancy may be sample dependent, but also other PZN-xPT
compositions must be studied before the dynamics of polar region
reorientation is understood in more depth.
146
Fig. 6. AE activity as a function of electric filed applied in the [001] direction at
room temperature
More insight into the MPB composition and phase transitions can
be gained from studying the AE activity under an external dc electric
field. The AE measurements at room temperature on PZN-9%PT crystal
sample under a dc electric field applied along the [001] direction is
shown on Figure 6. When the field value increases, two sharp AE
activity peaks are detected around 1 and 17 kV/cm. X-ray studies of
PZN-9%PT demonstrate that the room temperature structure of this
material within the MPB is orthorhombic [43] in the absence of an
electric field. Single crystals of PZN-9%PT poled along the [001]
direction become monoclinic, and complex monoclinic domain states are
obtained with the polar axis making an angle of about 45° relatively to
the poling direction. Due to the low static elastic energy of these domain
states in PZN-9%PT, a 3D network of small size (<100 mm) monoclinic
phase domains with a large interfacial is formed in the orthorhombic
matrix. The emergence, expansion and disappearance of such interphase
domain boundaries are accompanied by an acoustic emission activity
peaking at 1 kV/cm under the [001] electric field. The relatively low
field required reflects the very low energy barrier existing in PZN-xPT
147
between the O and MC states, which also allows the O polar axis [101] to
rotate easily in the monoclinic (010) plane [44]. The strong second peak
of AE activity, at 17 kV/cm, emerging upon further increase of the
externally applied [001] field clearly corresponds to the MC-T phase
transition. This transition is very energetic and corresponds to high strain
restitution at this field value, as it is common for highly piezoelectric
materials. Two weaker sub-bands detected on the right side of the 17 kV
peak may be related to surface effects or responses from other parts of
the sample due to the sample inhomogeneity. Fields of similar
magnitude have been determined for the same structural transition in
PZN-8%PT single crystals from unipolar measurements of electric-field
dependences of the polarization and strain responses [45]. The latter
show a hystersis loop, namely a reverse T-MC phase transition at fields a
few kV/cm lower upon decrease of the applied field. A corresponding
hysteresis peak at 14 kV/cm appears on the AE activity curve, in the
decreasing field regime, as shown in Fig. 3. It is noteworthy that no
other peak appears upon the field decrease down to 0 kV/cm, which
implies that the MC phase remains stable under zero field at room
temperature, without an immediate reverse MC-O transition. However,
repeated measurements of field-induced AE with a time gap of 24 hours
reveal the 1 kV/cm again, showing that the MC phase relaxes into the
more stable orthorhombic phase in terms of hours rather than weeks as
reported for the MA-X room temperature transition in PZN-8%PT [45].
7. Conclusions
The results reviewed above demonstrate that AE is an efficient
and inexpensive method suitable for studying many aspects of ceramic
materials, from their processing to fundamental investigations of
structural phase transitions and performance in devices. We have
described comprehensively how AE can be applied to investigation of
grain growth and phase work hardening of high-Tc superconductors as
well as to determination of the lower critical magnetic field Hc1 and of
the mixed state magnetic field. The generation of strong AE under
current transport in superconducting wires emphasizes one of the more
important industrial applications of the method, namely an early current
overload alarm in high-power superconducting circuits.
It has been also shown that the AE method is a powerful tool for
studying the relaxor-based systems. It allows to detect the structural
148
changes accompanied by strain relief due to all macroscopic phase
transitions (thermally activated or induced by application of an electric
field) and especially due to the formation and interaction of PNRs that
are responsible for the exceptional piezolelectric and electrostrictive
properties of relaxor ferroelectrics. The peculiar AE signature related to
the PNRs has allowed to determine accurately the temperatures of their
nucleation and enhanced growth of their correlation length. Electricfield-induced phase transitions in the PZN-9%PT crystal, corresponding
to the MPB composition, have been studied at room temperature as well.
The transition sequence follows the path O-MC-T on the field increase,
but only the reversible T-MC transition is observed in the AE
measurement. The irreversible O-MC transition at a low field of 1 kV/cm
can be revealed explicitly only by the AE method.
Thus, AE is becoming a universal method of studying the entire
range of phenomena associated with high-Tc superconductivity and
turning into a routine characterization tool of a growing number of
ceramic materials. In particular, the present results will stimulate further
research of phase transition as well as nucleation, growth and interaction
of PNRs in various relaxor and similar systems (with embedded
nano/microregions in the parent matrix) using the AE technique.
References
1. J. Kaiser, Archiv Eisenhuttenwessn Bd 24 (1953) 42.
2. V.S. Boyko, R.I. Garber, A.M. Kosevich, Reversible Crystal Plasticity
(AIP, NY, 1994).
3. J. Baram, Y. Gefen, M. Rosen, Scripta Met. 15 (1981) 835.
4. C. Picornell, C. Segni, V. Torra, F.C. Lovey, R. Rapacioli,
Thermochim. Acta 113 (1987) 171.
5. E. Vives, J. Ortín, L. Mañosa, I. Ràfols, R. Pérez–Magrané, A.
Planes, Phys. Rev. Lett. 72 (1994) 1694.
6. V.A. Plotnikov, D.V. Kokhanenko, Tech. Phys. Lett. 28 (2002)
833.
7. V.A. Kalitenko, V.M. Perga, I.N. Salivonov, Phys. Solid State
22 (1980) 1067.
8. E.A. Dul’kin, V.G. Gavrilyachenko, Tech. Phys. Lett. 23 (1997)
835.
9. E.A. Dul’kin, V.G. Gavrilyachenko, A.F. Semenchev, Phys.
Solid State 37 (1995) 668.
149
10. E.A.
Dul’kin,
L.V. Grebenkina,
I.V. Pozdnyakova,
L.A.Reznichenko, V.G. Gavrilyachenko, Tech. Phys. Lett.
1999, 25, 70 – 71.
11. E.A. Dul’kin, L.V. Grebenkina, V.G. Gavrilyachenko, Tech.
Phys. Lett. 27 (2001) 873.
12. E.A. Dul’kin, I.P. Raevskii, Tech. Phys. Lett. 22 (1996) 201.
13. E.A. Dul’kin, V.G. Gavrilyachenko, O.E. Fesenko, Phys. Solid State
39 (1997) 654.
14. E.A. Dul’kin, I.P. Raevskii, S.M. Emel’yanov, Phys. Solid State 39
(1997) 316.
15. E.A. Dul’kin, I.P. Raevskii, S.M. Emel’yanov, Phys. Solid State 45
(2003) 158.
16. E.A. Dul’kin, Mat. Res. Innovat. 2 (1999) 338.
17. M. Reissner, W. Steiner, R. Stroh, S. Hörhager, W. Schmid, W.
Wruss, Physica C 167 (1990) 495.
18. E.A. Dul’kin, Superconductivity 7 (1994) 104.
19. T. Ikegami, Acta Metall. 35 (1987) 667.
20. E.A. Dul’kin, Tech. Phys. Lett. 21 (1995) 533.
21. E. Dul’kin, V. Beilin, E. Yashchin, M. Roth, L.V. Grebenkina, Tech.
Phys. Lett. 27 (2001) 387.
22. K. Krishan, D.V. Natarajan, B. Purniax, J. Pramana, J. Phys. 31
(1988) L327.
23. A. Polyanskii, V. Beilin, A. Goldgirsh, E. Yashchin, M. Roth, D.
Larbalestier, Proc. 2000 Int. Workshop on Superconductivity (June
19 – 22, 2000, Shimane, Japan) 253.
24. E. Dul’kin, V. Beilin, E. Yashchin, M. Roth, L.V. Grebenkina, Tech.
Phys. Lett. 28 (2002) 489.
25. E. Dul’kin, Superconductivity 6 (1993) 246.
26. R. Abd-Shukor, Jpn. J. Appl. Phys. 31 (1992) L1034.
27. D.P.J. Almond, Phys. C : Solid State Phys. 21 (1988) L1137.
28. Yu.M. Poluektov, Low Temp. Phys. 20 (1994) 161.
29. A.A. Abrikosov, JETP 5 (1957) 1174.
30. M. Nikolo, Supercond. Sci. Technol. 6 (1993) 618.
31. G. Pasztor, C.J. Schmidt, Appl. Phys. 53 (1982) 3918.
32. E. Dul’kin E., Tech. Phys. Lett. 30 (2004) 200.
33. E. Dul’kin, V. Beilin, E. Yashchin, M. Roth, Supercond. Sci. Technol.
15 (2002) 1081.
150
34. M. Turchinsraya, D.T. Smith, A.L. Roytburd, D.L.J. Kaiser, Appl.
Phys. 88 (2000) 1541.
35. N. Savvides, J. Herrmann, D. Reilly, Physics C 306 (1998) 129.
36. S.-E. Park, T.R. Shrout, J. Appl. Phys. 82 (1997) 1804.
37. S.-F. Liu, S.-E. Park, T.R. Shrout, L.E. Cross, J. Appl. Phys. 85
(1999) 2810.
38. E. Dul’kin, M. Roth, B. Dkhil, and J.M. Kiat, J. Appl. Phys. 98 (2005)
023520.
39. J. Kuwata, K. Uchino, S. Nomura, Ferroelectrics 22 (1979) 863.
40. G. Burns, F.H. Dacol, Solid State Commun. 48 (1983) 853.
41. E. Dul’kin, I.P. Raevskii, S.M. Emel’yanov, Phys. Solid State 45
(2003) 158.
42. D. La-Orauttapong, J. Toulouse, Z.-G. Ye, W. Chen, R. Erwin, J.L.
Robertson, Phys. Rev. B67 (2003) 134110; D. La-Orauttapong, J.
Toulouse, J.L. Robertson, Z.-G. Ye, ibid. B64 (2001) 212101.
43. D.E. Cox, B. Noheda, G. Shirane, Y. Uesu, K. Fujishiro, Y. Yamada,
Appl. Phys. Lett. 79 (2001) 400.
44. D. La-Orauttapong, B. Noheda, Z.G. Ye, P.M. Gehring, J. Toulouse,
D.E. Cox, G. Shirane, Phys. Rev. B65 (2002) 144101.
45. K. Ohwada, K. Hirota, P.W. Rehrig, Y. Fujii, G. Shirane, Phys. Rev.
B67 (2003) 094111; B. Noheda, Z. Zhong, D.E. Cox, G. Shirane, SE. Park and P. Rehrig, ibid. 65 (2002) 224101.
151