Surface and Interface Phenomena in
Magnetic Nanostructures and Thin Films
Dissertation Proposal by Natalie A. Frey
Department of Physics
University of South Florida
Nanostructured systems composed of two or more technologically important materials
are useful for device applications and intriguing for the new fundamental physics they
may display. Magnetism at the nanoscale is dominated by size and surface effects which
combined with other media lead to new spin dynamics and interfacial coupling
phenomena. These new properties may prove to be useful for optimizing sensors and
devices as well as for biomedical applications such as drug delivery and hyperthermia
treatment for cancer. In this project we examine the magnetic and dielectric properties of
composite and multilayer systems in which one of the phases is magnetic. The other
phases have been chosen such that the entire structure is multifunctional, retaining the
magnetism while displaying ferroelectricity, biocompatibility. In some cases, the joining
of the materials results in magnetic coupling such as exchange bias. This is observed
when an antiferromagnet and a ferromagnet interacts or in magnetic nanoparticles with a
considerable amount of spin disorder. The structures that combine magnetism and
ferroelectricity will be measured for the magnetoelectric effect. Measurements to be used
include basic magnetic characterization as well as transverse susceptibility (TS), a
resonant radio frequency method that extracts the effective anisotropy constant (Keff) of
the sample. These measurements are taken inside of our commercial Physical Properties
Measurement System (PPMS) which can apply magnetic fields as large as ±7 T in
temperatures ranging from 350K down to 2K. Additionally, we have designed an
additional probe for the PPMS that can interface with a low frequency impedance
analyzer. This will enable us to take impedance measurements of these systems over a
range of temperatures and in the presence of an external magnetic field. With such a
probe, it is possible to glean even more information about the properties of these
composite materials and put them in a context to improve upon as well as add to existing
sensors and device technology.
Table of Contents
1. Introduction 3
1.1 Background and Motivation 3
1.2 Measurement Techniques 6
2. Nanoparticle Composite Systems 11
2.1 PVDF/Magnetic Nanoparticle Composite Thin Films 11
2.2 Au-Fe3O4 Core-Shell Nanoparticles 14
2.3 Au-Fe3O4 Bifunctional Nanoparticles 16
3. Multilayer Thin Film Systems 23
3.1. BaM/BSTO Multilayer Thin Films 23
3.2. CrO2 and CrO2/Cr2O3 Bilayer Thin Films 25
3.3. LCMO/YBCO Heterostructures 29
4. Conclusion 34
1.1 Background and Motivation
The term “nanostructured” describes materials with structure on the length scales
from 1 to 100 nm1. For magnetic particles, this leads to new phenomena as the length
scale becomes smaller than a typical magnetic domain. At this point,
superparamagnetism can be achieved in which thermal fluctuations prevent a stable
magnetization from occurring1. Particles showing superparamagnetism show no
coercivity above a characteristic blocking temperature, TB. At these scales, the surface to
volume ratio becomes so high that surface effects drastically alter the magnetic
properties. In nanoparticles, surface atoms are missing one or more exchange bond
leading to spin disorder. These disordered spins make up an outer layer that can interact
in certain ways with the interior atoms. One effect is that the disordered spins can couple
to the interior spins increasing the anisotropy of the particles below TB.
Similarly, when magnetic films are grown in a bilayer or multilayer configuration,
the interactions between the spins at the interface can play a big role in the overall
behavior observed. For example, exchange bias (EB) is a shift of the hysteresis loop
along the field axis which occurs when a ferromagnetic film is in contact with an
antiferromagnetic film2. The shift occurs when the AFM is ordered in the presence of a
field or an already ordered FM material. EB in thin films has technological applications
in devices such as magnetoresistive sensors. Even though the phenomenon was
discovered almost 50 years ago, its microscopic origin is not completely understood. The
shift in the hysteresis loop (HE) is also accompanied by an increase in the hysteresis loop
width enhancing the coercivity (HC). It is believed that both the exchange bias and HC
are the results of interfacial exchange coupling of the AFM and FM materials. While
initial studies of exchange coupling has been limited to thin AFM/FM films, core-shell
nanoparticle systems have also been shown to display exchange coupling3.
Trends toward device and sensor miniaturization have led to an interest in
combining two or more nanostructures into multifunctional materials, so that a single
device component can perform more than one task. Examples of combined
nanostructures include multilayer thin films, nanocomposites consisting of a continuous
matrix in which nanoparticles are embedded and nanoparticles of two different materials
grown in a core-shell structure. Moreover, the interactions between the properties can
lead to additional functionalities4.
Another such coupling effect is the magnetoelectric (ME) effect. The ME effect
is defined as the induction of a magnetization by an electric field and/or of a polarization
by a magnetic field4. Several “multiferroic” (possessing any combination of
ferromagnetism, ferroelectricity or ferroeleasticity) crystals have since been found and
studied in depth due to the rich new physics they display. However, it was not long after
that initial discovery that physicists gave up on the application of multiferroics since the
ME effect was relatively weak in single phase materials.
Recently, advances in thin film and nanocomposite technology have led to a
renaissance for multiferroic materials. This is because the optimization of
heterostructures of two ferroic materials has led to breakthroughs where multiferroic
behavior is observed. Such tailored multiferroics display both properties of the
constituent materials, but in some cases they also display the coupled effects, such as
ME, that were originally discovered in single crystals. In this case, it is interfacial
phenomena that provide the coupling necessary to produce cross manipulation of
properties. A good example of this is multilayers of lead zirconium titanate (PZT) and
Terfenol-D, an alloy of terbium dysprosium and iron (Tb1-xDyxFe2)5. PZT is a strong
piezoelectric and Terfenol-D is a magnetostrictive material. When an electric field is
applied to the whole system, the PZT responds with strain, which in turn induces a
magnetic moment in the Terfenol-D.
Figure 1. In a multiferroic, the coexistence of at least two ferroic forms of
ordering leads to additional interactions. In a magnetoelectric multiferroic, a
magnetic field may control P or an electric field may control M (green arrows).
(From reference 5)
Scientists have been looking to exploit ME and EB for various applications and
have even tried combining two. Borisov et al.6 recently demonstrated how the ME effect
can be used to switch the sign of HE in heterostructures of Cr2O3 and CoPt. It seems
plausible that such combinations could lead to even more opportunities for the future of
The overall goal of this proposed research is to measure the magnetic and
dielectric properties of multilayer thin films and nanocomposite materials in which one or
both of the phases is magnetic. This has two purposes. First, the nanostructures will
possess mulifunctionality meaning the properties of both phases can be exploited by
sensor and device technology. Second, the possible coupling between the phases can be
explored both for technological purposes and from a fundamental physics perspective.
We will use traditional magnetic and electrical measurements to characterize the samples,
some of which have been supplied through collaborations and some grown in our lab. In
addition, we will use our own novel resonant method to probe the magnetic anisotropy of
these systems which should certainly be affected by interactions among the phases. To
look closely at electrical response, we aim to build a specialized probe with which to
measure the impedance over a range of frequencies as a function of bias voltage,
temperature and even applied magnetic field. This should give us a good understanding
of the dynamics in the chosen systems.
In keeping with the goals of the IGERT (Interdisciplinary Graduate Education and
Research Training) fellowship, these materials will be explored in the context of
biological sensing and biomedical applications. This can be achieved in a number of
ways. For example, in the case of multilayer thin films, the coupling mechanism can be
evaluated for signal transduction. Biocompatible polymer-nanoparticle composites could
lend themselves well for external sensing. Lastly the joining of Au and Fe3O4 in core-
shell structure or as a composite nanoparticle will be looked at in for use such
applications as drug delivery, MRI contrast agents or hyperthermia.
1.2 Measurement Techniques
Static Magnetic Measurements
In any magnetic system, there are basic measurements that are undertaken to
characterize the properties. The magnetization versus field (M-H) curve is an important
indicator of many properties such as saturation magnetization (MS) and coercive field
(HC). How the magnetization of a material evolves with temperature, M(T) is also useful
information. For characterizing nanoparticles, M(T) can be taken while the sample is
cooled in an external field (an FC curve) or without an external field (a ZFC curve).
When plotted together, ZFC-FC curves give us information about the blocking
temperature and the size distribution of the particles. These experiments are done
routinely in the Functional Materials Laboratory with our Physical Properties
Measurement System (PPMS) from Quantum Design. It is a 7 Tesla superconducting
electromagnet cooled with liquid helium. We are able to adjust the temperature from 2K
to 350K. Besides the static magnetic measurements included with the commercial
system, we have been able to build our own probes to integrate into the PPMS to meet
various specialized needs. An existing “homebuilt” probe measures the transverse
susceptibility (TS) of thin films and nanoparticles. It has provided us with a wealth of
information about the magnetic anisotropy of these materials. We also propose for this
research building an impedance probe to measure the electrical response of these
materials over a range of temperatures and in the presence of a magnetic field.
Resonant methods have the advantage of precision and high sensitivity when it
comes to detecting changes in the physical properties of materials as a function of
temperature and magnetic field. This is due to the fact that frequency can be measured
with a high degree of accuracy. In a typical resonant technique based on an LC tank
circuit, the capacitor or inductor couples to the material under study and acts as a
transducer of physical parameters. Any change in material properties will induce a
change in the capacitance or inductance, which in turn results in a shift in the resonant
frequency. Thus, measurement of the frequency shift translates to directly probing the
electronic, dielectric, or magnetic response of the material to the oscillating signal.
Tunnel diode oscillators (TDOs) which operate based on this principle have been used in
the past to study a wide variety of material properties7.
The principle of the TDO can be explained as follows. An LC tank circuit is
maintained at a constant amplitude resonance by supplying the circuit with external
power to compensate for dissipation. This power is provided by a tunnel diode that is
forward biased with a voltage in the region of negative slope of its current-voltage (I-V)
curve, or “negative resistance region”. Such an arrangement makes it a self-resonant
circuit as the power supplied by the diode maintains continuous oscillation of the LC tank
operating at a frequency given by the expression7
When a sample is inserted into the oscillator tank coil, there is a small change in
the coil inductance ∆L. If ∆L/L <<1, one can differentiate equation (1) and obtain the
The inductance change is related to material properties. In the case of a magnetic
material, this is proportional to the real part µ’ of the complex permeability
The inductance coil in this experimental setup serves as the sample space in
which a gel cap containing the sample can fit. This entire coil is inserted into the sample
chamber of our PPMS using a customized radio frequency (RF) co-axial probe. The
static and magnetic fields are varied using the PPMS. The oscillating RF field HRF
produced by the RF current flowing in the coil windings, is orientated perpendicular to
the static field (HDC) and this arrangement sets up the transverse geometry. When the RF
field is perpendicular to the varying DC field, the change in inductance is actually
determined by the change in transverse permeability µT of the sample. Thus, knowing the
precise geometrical parameters we can derive an absolute value for the transverse
susceptibility (TS) χT=µT-1. The transverse susceptibility ratio can be written as:
∆χ /χ (%) =
[∆χ T ( H ) − χ T × 100
where χTsat is the transverse susceptibility at the saturating field Hsat. This quantity,
which represents a figure of merit, does not depend on geometrical parameters and is
useful for comparing the transverse susceptibility data for different samples, or for the
same sample under different conditions8.
The theory of reversible TS was first studied theoretically in detail by Aharoni et
al. in 19579. In his work, the expression for the TS was analytically derived for a singe
domain particle based on the Stoner-Wolfarth model. When ∆χT/χT is plotted as a function
of applied magnetic field, singularities are observed at ±HK, which correspond to the
anisotropy fields and are referred to as anisotropy peaks. The effective anisotropy
constant can be extracted from the relation
where MS is the saturation magnetization. This experiment is unique in that it uses a
resonant method to directly extract HK rather than indirect or numerical techniques.
Many known coupling phenomena are characterized by an overall increase in
effective anisotropy of the system, exchange coupling being the best known example. In
these systems, an increase in coercivity is almost always observed10. We have shown
how TS can also be used to compliment traditional static magnetic measurements for
systems whose effective anisotropy has increased due to the presence of a second phase.
In the examples outlined in subsequent sections we explain how the TS curves for well
understood systems are drastically altered when another system is introduced, either as a
multilayer thin film or a composite.
Although the majority of the experiments done for this project are probing the
magnetic response of the systems, it is also important to look at how these materials are
affected by electrical stimulus. Our Hewlitt Packard low frequency impedance analyzer
can be used to measure complex impedance of a material over a range of frequencies (5
Hz < f < 13MHz). In addition, the dielectric constant, ε, can be derived from the real and
imaginary parts of the complex impedance. We expect that in materials where we have
seen coupling between the phases, ε too should be affected.
By integrating the impedance analyzer into the PPMS, the parameter space for
dielectric measurements will be greatly enhanced. We can measure the impedance, Z,
over the range of temperatures available to us in the PPMS as well as in the presence of
an external magnetic field. Besides seeing how these materials respond under a wide
variety of conditions, this experiment should provide direct evidence for any
magnetoelectric (ME) coupling found the systems. For this project, we propose to build
another probe specifically for this purpose. We have already verified that this is possible
by extending the coaxial cables from the impedance analyzer by a length equivalent to
the distance into the PPMS sample chamber. We then used a bulk sample of Barium
Titanate (BaTiO3) provided by a collaborator and verified that the measured Z graphs
were the same regardless of cable length. Next we temporarily altered an existing probe
by rewiring it with SMA cables and connecting those to the coaxial cables external to the
PPMS. We then had a working prototype for the impedance probe. We again
reproduced accurate impedance measurements for the BaTiO3 and varied the temperature
within the PPMS to measure the ferroelectric Curie temperature of the material. Our
results matched the results obtained by the collaborator who provided us the sample.
We have used the TS probe design as a template and designed a probe that can
integrate into the PPMS just for impedance measurements. While the majority of the
parts have been machined, there is still some work that needs to be completed before
accurate measurements can be taken. We expect that once the probe is completed, we
can compliment many of our magnetic experiments on the systems under study with
dielectric measurements as well.
Since several systems are composed of a magnetic material and a ferroelectric
material it is imperative to directly evaluate the ferroelectric properties of the composites.
We are equipped with a commercial ferroelectric tester from Radiant Technologies and a
high voltage amplifier for measuring polarization as a function of voltage (P-V), the
ferroelectric equivalent of an M-H curve. With this instrument we can learn more about
the ferroelectric components present in the multilayers and composites to be studied. By
comparing the ferroelectric properties of the multilayers/composites to the single phase
material we can determine what role the magnetic phase of the system may be playing.
As of right now, the ferroelectric tester is set up for room temperature use only.
While ideally we would like to take P-V measurements inside the PPMS (to vary
temperature as well as add an external H field), the high voltage wiring may impose some
instrumentation limits on any probe we can build. This will be explored with future
2. Nanoparticle Composite Systems
2.1. PVDF/Magnetic Nanoparticle Composite Thin Films
Polyvinylidene fluoride (PVDF) is the best known in a small class of polymers
displaying piezo-, pyro-, and ferroelectricity. It is thought that these properties originate
from the strong molecular dipoles within the polymer chain11. PVDF has gained
considerable attention due to its biocompatibility and close impedance matching with
human skin12. These properties have already allowed people to make numerous
biomedical devices mainly in the form of external sensors (flexible bands, wrappings or
strips) used for monitoring mechanical events of the cardiovascular and pulmonary
systems12. However, little work has been done to use piezoelectric polymers for
biochemical sensing, mainly because they are so inert that sensing has to be done
mechanically rather than chemically. But recently, Inácio et al developed a biochemical
sensor that uses a combination of PVDF and a porous membrane with a high protein
binding capacity to be used as an oscillatory resonant device13.
PVDF and its copolymers have already been exploited as the matrix phase of a
multiferroic composite along with Terfenol-D, a highly magnetostrictive material14. Our
laboratory has sufficient experience in polymer/magnetic nanoparticle composites to
explore PVDF as a host matrix for other nanoparticle systems.
PVDF films have been formed from dissolving polymerized powder in dimethyl
sulfoxide (DMSO) and then spin coating the solution onto various substrates. The films
are then baked at 130°C for an hour to let the solvent evaporate. Resulting films do not
stick well onto glass or Si/SiO2 substrates, and usually peel off intact. While
measurements can still be performed on these films, optimizing a substrate and PVDF
film thickness will remain part of the ongoing research in this area.
Fe3O4, CoFe2O4 and NiFe2O4 nanoparticles have all been added separately to
PVDF/DMSO solutions and then spun coat to create composite thin films. Each type of
nanoparticle was synthesized in our lab by chemical methods and then thoroughly
characterized. Some of the composites have then been measured using the PPMS.
Preliminary findings reveal that the nanoparticles retain their superparamagnetic
properties even when embedded in the PVDF. This is likely due to the coating of oleic
acid on the nanoparticles, a surfactant which has been shown to keep nanoparticles from
agglomerating in polymer/nanoparticle composite thin films15.
The composites also display blocking temperatures that are the same as for the
plain nanoparticles previously measured. While the blocking temperatures are the same,
the HC values at low temperatures for the composites are much higher than for just the
nanoparticles (figure 2). This effect is most likely due to the presence of a matrix rather
than the specific piezoelectric matrix as this has been witnessed in other systems.
Cross-sectional and surface TEM images of the composites have revealed that the
nanoparticles do not seem to disperse well in the PVDF (figure 3). Finding a dispersion
method either via changing the polymer solvent or altering the surface chemistry of the
nanoparticles will be necessary.
Fe3O4 in PVDF
-20 0 20
-1 0 1
Figure 2. Detail graph of M-H curve showing the coercivity of PVDF/Fe3O4
composite compared to just Fe3O4 nanoparticles. Inset: full M-H curve.
Figure 3. TEM image the surface of a PVDF/CoFe2O4 film. The darkest regions
are where the nanoparticles have collected.
Ongoing and Proposed Research
The ongoing research for this system is still in the optimization of the composites.
We will have to better understand the chemistry involved in successfully getting the
nanoparticles to disperse in the matrix as well as successfully adhering the composite to a
substrate. Dielectric measurements will also require electrodes being affixed to the films
either by capacitor geometry or possibly a series of four surface electrodes. While
capacitor geometry will require metallization of a preexisting substrate, this might
actually help with the adhesions issues. We also are curious to look at nanoparticles of
Terfenol-D as micron size particulates are all that have been studied for the magnetic
phase in such composites. Since we have seen an increase in HC with the composites,
this tells us that TS will be a useful measurement for evaluating the overall magnetic
Once this system is optimized, it will be a good candidate for dielectric
measurements due to the outstanding electrical properties of the PVDF. We expect that
with the new probe, taking and interpreting the dielectric results for this system will be
time consuming. Systematic P-V measurements as well as impedance measurements will
help us look for any sign of ME coupling.
2.2. Au-Fe3O4 Core-Shell Nanoparticles
Au nanoparticles are often used for biological and chemical sensing because their
surface chemistry allows one to easily tag them with an antigen. The functionalized
particles will then have selectivity for a specific antibody. Immunoassays operate on this
principle with surface plasmon resonance being the measurement of choice for detecting
a change in spectra associated with an antigen-antibody bond. When the Au is grown as
a shell around a Fe3O4 nanoparticle another sensing mechanism is introduced, namely the
By themselves, magnetic nanoparticles are being explored for targeted drug
delivery. This would reduce the amount of a cytotoxic drug and the associated side
effects. It would also reduce the amount of drug required because the targeting is
localized. In magnetically targeted therapy, the cytotoxic drug is attached to a
biocompatible magnetic nanoparticle carrier. The drug/carrier complexes are usually
used as a ferrofluid that is injected into a patient’s circulatory system. After the particles
have entered the bloodstream, high gradient external magnetic field can be used to
concentrate the complex at a specific site within the body. Once at the targeted site, the
drug can be released either via enzymatic activity or changes in physiological conditions
such as pH, osmolality or temperature and taken up by the tumor cells16. Having an Au
coating on the Fe3O4 particles should facilitate the drug attachment and delivery while the
magnetic properties of the particles should not be compromised.
Another application for magnetic nanoparticles is MRI contrast enhancement16.
The basic idea behind MRI is to measure the changes in the magnetization of hydrogen
protons in water molecules sitting in a strong magnetic field. Protons in different tissues
react differently which create the images of different anatomical structures. The images
can be enhanced using “contrast agents” which sharpen the contrast by affecting the
behavior of protons in the vicinity17. Since MRI contrast relies on the differential uptake
of different tissues, the nanoparticles are coated with different target-specific agents. For
instance, coating a magnetic nanoparticle with dextran allows them to be selectively
taken up by the liver, thus making the liver much easier to image16. Having the
nanoparticles coated with an Au shell would make them much more versatile for
functionalization and thus differential uptake by biological systems. Again, the principle
behind using nanoparticles for MRI contrast enhancement should be compromise the
magnetic properties of the Fe3O4.
Lastly, the use of magnetic nanoparticles has been investigated for hyperthermia
treatment in cancer treatment. It has been shown that certain types cancer cells are more
sensitive to temperatures in excess of 41° C than normal cells. Hyperthermia refers to the
localized heating of tumor cells by injecting magnetic fluids into the affected area. In the
past, this method has relied on the theory that metallic objects when placed in an
alternating magnetic field will have induced currents flowing through them, which is
proportional to the size of the magnetic field and the size of them object17. As the
currents flow within the metal inductive heating occurs. If the metal is also magnetic, the
phenomenon is greatly enhanced. While traditional magnetic metals have been proposed,
the issues of toxicity and material stability are of concern. By using Au-Fe3O4 core-shell
particles, these limitations may be overcome. It is also possible that using the Au surface
properties, drugs can be delivered to the site in conjunction with the hyperthermia.
Since Au-Fe3O4 core-shell particles show potential to improve many facets of the
medical industry, we propose to work closely with our IGERT collaborators to synthesize
Au-Fe3O4 core-shell nanoparticles and use our facilities to characterize them. Traditional
magnetic measurements as well as dynamic testing such as AC susceptibility and TS can
help determine optimize the core diameter and shell thickness for use in magnetic
applications. Then we propose to use the knowledge of our collaborators to functionalize
the particles to evaluate them for drug delivery. By utilizing the knowledge gained in our
lab and taking advantage of what our collaborators have to offer, this proposed research
has the potential to contribute to the interdisciplinary field of bionanotechnology.
2.3. Au-Fe3O4 Bifunctional Nanoparticles
As discussed previously, nanoparticles of Fe3O4 and Au have been extensively
studied as individual systems -Fe3O4 for its magnetic properties and Au for its optical and
electronic properties as well as its desirable surface chemistry. The joining of these two
materials in a unique nanostructure is interesting from applications perspective as well as
for the opportunity to learn more about the fundamental physics of composite
nanostructures. Due to the lattice constant of Fe3O4 (a = 8.35A) being very nearly double
that of Au (a = 4.08A), it turns out that Au and Fe3O4 can be grown as epitaxial
composite nanoparticles. Based on the chemistry of the reaction, Fe3O4 can grow on one
facet or multiple facets of an Au seed particle. The former results in a so-called
“dumbbell-like” nanoparticle and the latter in a “flower-like” nanoparticle. Besides being
composed of two technologically important materials thus rendering them “bifunctional”,
such structures represent new surface spin configurations for the much studied Fe3O4
nanoparticles. Both the dumbbell and flower Au-Fe3O4 structures described above were
synthesized by Dr. Sun and his research group at Brown University18. Our work has
focused on the magnetic characterization and the influence of surface and shape
anisotropy on the collective magnetization of these particles. We have demonstrated that
in the case of the dumbbell particles, the behavior is consistent with superparamagnetic
particles with blocking temperatures (TB) and low temperature coercivities (HC) that vary
with Au particle size. However, in flower particles, the magnetic properties are
drastically altered, with prominent shifts in the hysteresis loops and anomalously high
coercivities. This research has been written up and submitted for publication:
Natalie A. Frey, Sanyadanam Srinath, Hariharain Srikanth, Chao Wang and
Shouheng Sun. Static and dynamic magnetic properties of composite Au-Fe3O4
nanoparticles. Submitted to IEEE Transactions on Magnetics.
We are also currently preparing two additional manuscripts for submission.
We have performed extensive static and dynamic magnetic measurements on both
the dumbbell and the flower nanoparticles. Zero field cooled-field cooled (ZFC-FC)
measurements of the dumbbell particles (figure 4a) revealed behavior consistent with
superparamagnetic particles with a blocking temperature (TB) that varied with Au particle
size (Fe3O4 particle size has not been varied). M-H measurements taken above TB the
particles show zero coercivity, while below TB the particles acquire some hysteresis. The
coercivity for all the dumbbell samples is still larger than previously reported for Fe3O4
nanoparticles. Since coercivity is an indicator of magnetic anisotropy, it makes sense that
a particle possessing asymmetry should display higher coercivity.
The ZFC-FC measurements for the flower particles (figure 4b) showed a sharper
peak at the blocking temperature. The M-H measurements for the flower particles
indicated extremely large coercivities in the blocked state. We also performed M-H
loops after cooling from above TB in the presence of a magnetic field. Exchange bias
(EB) was observed for this configuration and persisted until a temperature significantly
below TB. We believe there are two characteristic temperatures for the flower particles-
the first being a freezing temperature reminiscent of spin glasses or cluster glasses and
the other being the traditional superparamagnetic blocking temperature. Not only did EB
persist as temperature was increased, it also persisted through several hysteretic cycles,
with the HC values decreasing through each cycle. This is the so-called “training effect”
and has been observed for exchange bias in several nanoparticles systems19. Exchange
bias in nanostructures is a topic of current interest3 and has been reported in several core-
shell particles. We believe our results on these flower particles represent the first report
of exchange bias in nanoparticles with an exotic geometrical configuration.
ZFC H = 100 Oe ZFC
H = 100 Oe 2 FC
Au-Fe3O4 dumbbell nanoparticles
Au-Fe3O4 flower nanoparticles
0 50 100 150 200 250 300 0 50 100 150 200 250 300
T (K) T (K)
Figure 4. ZFC-FC curves for (a) dumbbell particles and (b) flower particles.
The insets show TEM images of each type of particle. In each case the Au is 8
nm and the Fe3O4 is 18 nm.
-4 -2 0 2 4
Figure 5. M-H curves showing exchange bias in flower particles at 2K (closed
circles) and 40K (open circles).
TS measurements were also preformed on all of the Au-Fe3O4 particles to further
study the magnetic anisotropy. At low temperatures the dumbbell particle showed two
sets of peaks of asymmetric height which is consistent with nanoparticles in the blocked
state (figure 6a). The anisotropy peaks disappear as TB is crossed. The flower particles
(figure 6b) showed some properties that reinforce the highly anisotropic picture given by
the M-H measurements. The two sets of peaks were both asymmetric in peak width,
height and field position. The most important difference between the two sets of particles
however is the striking difference in the anisotropy fields for the flower particles
compared to the dumbbell particles. This is consistent with the flower particles having
higher shape and surface anisotropies than the dumbbell particles.
We were also able to see exchange bias in the TS of the flower nanoparticles after
field cooling from above TB. The shape of the curves and relative position of the peaks
remain the same, but the entire curves were clearly horizontally shifted. The shift
persisted up to the hypothesized spin freezing temperature.
Several groups have focused on anomalous magnetic behavior in ferrite
nanoparticles due to surface spin disorder such as exchange-biased hysteresis loops, and
high field irreversibility. An explanation for this behavior is that when a large enough
fraction of atoms reside on the surface of a particle, the broken exchange bonds are
sufficient to induce surface spin disorder thus creating a core-shell structure made of the
ferrite core with a shell of disordered spins. These disordered spins can take on a number
of configurations, one of which can be chosen by field-cooling the particle to induce an
exchange-bias effect20. It is thought that the lowest energy configuration of surface spins
in the zero field cooled condition of a spherical particle is the one in which the spins
point in the radial direction from the particle. The energy required to rotate these spins
contributes to the enhanced coercivity below the spin freezing temperatures as well as
“open”, irreversible hysteresis up to high fields. Although Bodker et. al.21 have
demonstrated via symmetry arguments that a perfectly spherical particle should have a
zero net contribution from surface anisotropy, it is important to note that in both
dumbbell and flower particles the spherical symmetry is broken, resulting in a net surface
anisotropy. For the dumbbell particles, the surface spins will play less of a role since the
Fe3O4 particles are 18 nm, their surface spins will be a smaller fraction of the overall net
magnetization. In fact, in both experimental and theoretical literature, exchange coupling
is observed when the particle’s diameter is below 10 nm, though we still expect the
asymmetry of the dumbbell particles to influence the anisotropy. The flower particles
present a much more complicated case of surface anisotropy in which much of the
particle is comprised of surface atoms and if each “petal” is considered, the distance from
surface to core is small enough to be greatly influenced by these effects. Kachkachi et
al.22 argue that that magnetic disorder on the surface caused by surface anisotropy is long
range, implying that spins in the core of a very small particle (~2nm) render a
magnetization that deviates from the bulk value. The interactions between the moments
on separate “petals” around a single Au core should certainly not be overlooked in this
case and merit further investigation
-1000 -500 0 500 1000 0.00
-10000 -5000 0 5000 10000
H (Oe) H (Oe)
Figure 6. Low temperature TS for the (a) dumbbell and (b) flower nanoparticles.
Current and Ongoing Research
We are currently performing a series of experiments to better understand the time
dynamics of these systems. Preliminary results on the dumbbell and flower particles
have revealed the existence of a “memory effect” which is described as follows. A
traditional field-cooled curve is taken in a low field, referred to as the reference curve.
The sample is then warmed up, and the measurement is again cooled in a field. This
time, the measurement is stopped at several temperatures and the field is turned off for a
set amount of time. After each stop and wait period the field is turned back on and
cooling is resumed. This procedure produces a step-like M(T) curve (figure 7). After
reaching the base temperature the sample temperature is raised continuously in the same
low field and the magnetization is again measured. In this system, the step-like curve is
reproduced upon re-warming. It is as if the system “remembers” its thermal history.
Understanding this phenomenon and the overall time dependent magnetic response of
this system is a goal this research.
0 50 100
Figure 7. The memory effect in dumbbell nanoparticles. Black line indicates
field cooled measurements which are stopped at three temperatures. The red line
indicates field warming after reaching the lowest temperature.
Bifunctional nanoparticles made of Au and Fe3O4 could also be extremely
attractive to the medical community. In this particular configuration, the Au could be
tagged with a biological entity using gold-thiol chemistry while the Fe3O4 lends itself to
site-specific targeting when guided by an external magnet. It may even be possible to use
the surfaces of both the Au and Fe3O4 for functionalization, allowing the particles to
deliver two different types of treatment to a specified area while using two separate
release mechanisms for the drugs.
The exceptional magnetic properties displayed by these configurations would
cause them to give off a large signal for MRI contrast while, as mentioned above, the Au
could be tagged with the differential uptake agent.
These composite particles would also be an intriguing option for exploring
hyperthermia as the magnetic properties seem to be heavily influenced by Au size. By
adjusting the Au and/or Fe3O4 sizes, we could even explore making the particles
ferrimagnetic at room temperature which would allow hysteretic losses to add to the
inductive heating explained previously. Such tunability of the magnetic properties along
with the known biomedical applications of Au nanoparticles makes composites
nanoparticles worth studying systematically these biological applications.
3. Multilayer Thin Film Systems
3.1 BaM/BSTO Multilayer Thin Films
The ferroelectric ceramic Barium Strontium Titanate (BaxSr1-x)TiO3 or BSTO is
an attractive candidate for many applications due to its exceptionally high tunability, high
breakdown field and relatively low loss tangent at microwave frequencies23. BSTO is
one of the most widely studied ferroelectric materials, which makes it ideal for
experimentation with multilayer growth.
Similarly, Barium Hexaferrite (BaFe12O19 or BaM) is a well-studied ferrimagnet
with an extremely large anisotropy and coercivity as well as a high Curie temperature
(350°C) making it magnetic at room temperature24. BaM has also been looked at
extensively for microwave applications and high-density recording media.
The joining of these two materials in multilayer thin films formed the basis of
my master’s thesis (Microstructure and Magnetism in Ferrite-Ferroelectric Multilayer
Thin Films) as well as two publications:
1.) N.A. Frey, R. Heindl, S. Srinath, H. Srikanth, and N.J. Dudney.
Microstructure and Magnetism in Barium Strontium Titanate (BSTO)-Barium
Hexaferrite (BaM) multilayers. Materials Research Bulletin, 40, 1286 (2005).
2.) S. Srinath, N.A. Frey, R. Hajndl, H. Srikanth, K.R. Coffey, and N.J. Dudney.
Growth and Characterization of Sputtered BST/BaF Multilayers. Journal of
Applied Physics. 97, 10J115 (2005).
Several samples of BSTO/BaM multilayer thin films as well as BSTO and BaM
control samples were grown via magnetron sputter deposition onto Si/SiO2 and Al2O3
substrates. These samples were grown both at Oak Ridge National Laboratory through a
summer Southeastern Universities Research Association fellowship and at the University
of Central Florida through a collaboration. The films were subsequently annealed at
USF. The microstructure of the films was characterized using x-ray diffraction (XRD),
surface and cross-sectional scanning electron microscopy (SEM) and atomic force
microscopy (AFM). XRD showed that BSTO and BaM polycrystalline phases were
present in all multilayers as well as an impurity phase (Sr3Si3O9) in the samples grown on
Si/SiO2. AFM and SEM surface scans of the BaM showed the characteristic needle-like
BaM grains and cross-sectional SEM images confirmed that the multilayer structure was
preserved even after annealing (figure 8).
Figure 8. Cross sectional SEM image of a BaM/BTO multilayer thin film taken
after annealing. The inset shows a surface scan of the top layer which is BaM.
A comparative study of magnetic properties was performed on the BaM films and
the multilayer films. While the BaM films showed behavior quite consistent with bulk
BaM, the multilayers displayed an increase in coercivity with increase in temperature,
contrary to other published results (table 1). We concluded that the presence in BSTO
was affecting the microstructure in the BaM enough to alter the grain shape and thus the
effective anisotropy of the system.
Material Hc (10K) Hc (300K)
Pure BaM 2100 Oe 1900 Oe
BaM/BSTO 1750 Oe 2100 Oe
Table 1. Comparison of the temperature dependent coercivity for BaM and
BaM/BSTO multilayer thin films.
While these ferroelectric-ferrite multilayers have the potential to show
magnetoelectric coupling, measurements to explore this possibility were beyond the
scope of my masters’ work. With the realization of the proposed impedance probe, it will
be possible to look for anomalous behavior which could be indicative of ME coupling.
3.2. CrO2 and CrO2/Cr2O3 Bilayer Thin Films
Cr2O3 is antiferromagnetic with a Néel temperature of 307 K. It was also the first
multiferroic material discovered possessing ME coupling dating back to 196125. CrO2 is
a ferromagnet that is predicted to be 100% spin-polarized, making it of interest for
spintronic devices such as magnetic tunnel junctions and spin valves26. Dual
functionality combining a known multiferroic and a half-metal is intriguing from both a
basic and applied materials prospective.
Through our collaboration with Dr. Gupta and his group at the University of
Alabama’s MINT Center, we were able to obtain both CrO2 thin films and Cr2O3 bilayer
thin films for magnetic investigation. We used a variety of experiments (basic magnetic
characterization as well as transverse susceptibility) to confirm the presence of coupling
between the two layers. This research to date has resulted in two published papers:
1.) N.A. Frey, S. Srinath, H. Srikanth, M. Varela, S. Pennycook, G.X. Miao and
A. Gupta. Magnetic anisotropy in epitaxial CrO2 and CrO2/Cr2O3 bilayer
thin films. Physical Review, B, 74, 024420 (2006).
2.) S. Srinath, N.A. Frey, H. Srikanth, G.X. Miao, and A. Gupta. Exchange bias
in CrO2/Cr2O3 bilayer thin films. Advances in Science and Technology, 45,
The films were grown using a CVD and annealing technique perfected by Dr.
Gupta and his group. A CrO2 film is first grown on (100) TiO3 and then the film is
controllably converted to Cr2O3 via high temperature annealing. By annealing for
different times, different ratios of CrO2:Cr2O3 can be formed allowing a systematic
investigation of the magnetic interaction between these two materials. High resolution
TEM was performed at Oak Ridge National Laboratory of a cross section of one of the
films revealing epitaxial layers.
First, we used static magnetic measurements and TS to characterize the CrO2
films to fully understand the thickness dependence of the magnetic properties, and if
necessary decouple them from the bilayer effects (figure 9). The CrO2 films displayed an
easy axis of magnetization that changed orientation with film thickness. This has been
previously documented by our collaborators27 and is believed to result from an
inhomogeneous distribution of strain in the films, where a portion of the film remains
highly strained while the balance is partially relaxed. We used TS to verify these results
and for the first time gained a view of the temperature dependence of the double
switching phenomenon. We also used the anisotropy fields obtained from TS
measurements to note that for a critical thickness around 200 nm the temperature
dependence of HK is nearly constant. This would imply that films of this thickness could
be optimized for uniform properties in devices such as magnetic tunnel junctions that are
operational over a broad range in temperature.
(10 ) ∆χT/χT (%)
3 21.5 nm 1.5
(10 ) ∆χT/χT (%)
-4000 -2000 0 2000 4000
Figure 9. TS curves for CrO2 films of varying thicknesses.
Next we performed M-H measurements on the CrO2/Cr2O3 bilayer thin films.
When we compared the temperature dependent coercivity of the bilayers with the CrO2
films, we saw an overall increase in coercivity with the addition of the Cr2O3 (figure 10).
This is a well known result of exchange bias in ferromagnetic/antiferromagnetic bilayer
thin films. However, the usual accompanying horizontal shift (exchange bias) in the M-
H loops after field cooling was not present. This discrepancy can be explained in terms
of a model proposed by Shulthess and Butler28 in which the spin-flop coupling between
the ferromagnetic and antiferromagnetic bilayers with a perfectly flat interface gives rise
to uniaxial anisotropy (increase in HC) but no unidirectional anisotropy (HE). Since high
resolution TEM images show near-perfect epitaxy between the layers, we strongly
believe that this model describes our situation.
CrO2 100 %
300 CrO2 64.3%
HC (Oe) 200
0 50 100 150 200 250 300 350
Figure 10. Temperature dependent of coercivity in CrO2 films and CrO2/Cr2O3
bilayer films with two different percentages of CrO2.
The TS results for the CrO2/Cr2O3 bilayers showed very different results from
CrO2 single layers. Most noticeably, the anisotropy peaks were shifted towards much
higher fields as the Cr2O3 content increases (figure 11). There were also qualitative
features present in the bilayers that were not seen in the single layers. To verify that the
increase in anisotropy was due to the Cr2O3 and not simply thickness dependence of
CrO2, we determined the effective anisotropy constant, Keff, of the bilayers via equation
(5). Then we compared the Keff values for the bilayers to Keff values of the corresponding
thickness of CrO2 films found by Miao et al28. When these values were compared it
became clear that the Keff we observed was consistently larger than what it would be
without the Cr2O3 lending more evidence for CrO2/Cr2O3 coupling.
What we have learned so far is that this is an incredibly rich, complex system
whose magnetic properties are dominated by exchange coupling and interfacial strain.
This system in particular will be a very interesting candidate for the proposed dielectric
measurements. Since Cr2O3 is a known ME material, we anticipate that imposing both
electric and magnetic fields on the system will produce new and exciting results.
20 CrO2 0%
(10 ) ∆χT/χT (%)
(10 ) ∆χT/χT (%)
-4000 -2000 0 2000 4000
Figure 11. TS curves for CrO2/Cr2O3 bilayers of varying CrO2 content.
3.3. LCMO/YBCO Heterostructures
Recently, the interplay of ferromagnetism and superconductivity has been a topic
of interest. The competition between generally mutually exclusive types of long-range
order has led to investigations of changes in superconducting transition temperatures and
the dependence of transition temperature on direction of magnetization. Heterostructures
of highly spin polarized CMR oxides and cuprate superconductors have shown to be
exceptional systems with which to explore these questions. Since both classes of
materials have similar perovskite structures with comparable lattice constants in the basal
plane, it is possible to combine them into structurally coherent superlattices with very
sharp interfaces. Many experiments have shown that cuprate/manganite based
superlattices exhibit long ranging proximity effects29, suppressed magnetization30, and
Oxx et al32 have shown that TS can be used to measure the penetration depth, λ, in
the superconductor YNi2B2C. Their results showed that λ can yield important
information about the superconducting order parameter, vortex dynamics in the mixed
state and can even delineate critical fields in superconductors. In this experiment, the
change in resonant frequency is related to λ by the following:
δλ(T, H, θ) ≡ -g[f(T, H, θ)-f0(T, H)] (6)
where here T is the temperature, H the DC magnetic field, θ is the angle between H and
the a-b plane and g is a geometrical factor set by the sample dimensions. As before, f0
and f refer to the initial and final resonant frequency of the TDO respectively. The
analog of the ∆χT/χT graph then becomes ∆λ/λ where
∆λ ≡ δλ(T, H, θ) – δλ(T, H=0, θ) (7)
We have been given samples of La0.7Ca0.4MnO3/YBa2Cu3O7 (LCMO/YBCO)
bilayers and LCMO/YBCO/LCMO trilayers as well as single layers of each material by
Dr. Santamaria and his group at Universidad Complutense de Madrid. They were grown
on SrTiO3 substrates by high pressure DC sputtering with the thicknesses of the LCMO
and YBCO being 40 u.c. and 15 u.c. respectively. We have used our TS method to
simultaneously probe the dynamic magnetic susceptibility and the vortex penetration
depth for these samples.
The TS data for the pure LCMO films was consistent with soft ferromagnetic thin
films measured along the hard axis of magnetization. We see two anisotropy peaks in a
unipolar scan when measured with the external field parallel to the hard axis of
magnetization. In figure 12 we present the TS data for the pure YBCO film. The
geometry of this measurement places HDC perpendicular to the film plane. This means
HRF lies in the plane. The bipolar scan was plotted to show the hysteretic properties of
the film at high fields. What is immediately noticeable about this graph are the three
different regimes corresponding to the superconducting state, the normal state and an
intermediate state. While a definitive analysis of these results has yet to conducted, we
have seen that the configuration of just the superconductor plays a large role. We know
that for superconductors in a large magnetic field (H > HC1), the dependence of λ on
magnetic field arises from two competing factors: the depairing of condensed electrons
and dynamic vortex response. The former dominates when JRF is parallel to HDC (this
corresponds to HRF perpendicular to HDC, the typical transverse geometry) and the latter
when JRF is perpendicular to HDC (or, when HDC and HRF are perpendicular, the so-
called longitudinal geometry, not shown).
-60 -40 -20 0 20 40 60
Figure 12. TS curves for YBCO in the transverse geometry.
In figures 13 and 14 we present TS data for the bilayer sample and the trilayer
sample both measured in the transverse geometry. Recall that in these samples the
magnetic phase and superconducting phase are both present. In the bilayers, the behavior
is almost completely dominated by the superconducting phase as the anisotropy peaks of
the LCMO can barely be made out below TC. In both configurations there is complex
behavior that changes quickly with temperature around TC. This suppression of the
anisotropy peaks is consistent with results obtained by Hoffman et al30.
In the trilayers, when H is applied along LCMO’s hard axis of magnetization the
anisotropy peaks are clearly present and persist even in the superconducting state. In this
sample, TC is 60K due to the presence of the LCMO. What these results tell us is there is
a dramatic shift in the ferromagnetic/superconducting competition when the layering
scheme of the heterostructures changes from S/F to F/S/F. In the longitudinal geometry,
no anisotropy peaks are present. However, it is evident that the mere existence of the
ferromagnetic phases in both the bilayers and the trilayers play an important role in the
overall vortex response since the longitudinal susceptibility graphs (not shown) of the
heterostructures vary greatly from the pure YBCO.
The sharply decreasing behavior at lower fields seen at intermediate temperatures
in both the YBCO and the bilayers is reminiscent of the behavior noted by Srikanth et al33
when TS was performed on RuSr2GdCuO8. They attributed this behavior with the
variation of penetration depth in response to vortices entering the sample in the mixed
state (HC1 < H < HC2). Note this behavior is suppressed in the trilayers.
There is still much to be understood about this complex system in order to
properly put the results into context with other LCMO/YBCO studies. The ongoing
research for this part of the project will be to gain a deeper understanding of the joining
of ferromagnetism and superconductivity and realize how our TS measurements can add
to the existing picture.
-30 -20 -10 0 10 20 30
Figure 13. TS curves for LCMO/YBCO bilayer film in the transverse
-20 0 20
Figure 14. TS curves for LCMO/YBCO/LCMO trilayer film in the
Composites and thin film multilayers composed of two functional materials have
the potential to contribute greatly in the coming years for device and sensor technology.
Not only are such structures compact which can lead to even smaller devices, the
coupling phenomena that are possible add more possibilities for functionality.
The main goal of this research is to measure and evaluate several such systems in
the context of device and sensor technology. Basic magnetic and electric measurements
paired with more specialized methods allow us to verify their usefulness as well as probe
the underlying interactions present at the interface between these materials. Basic
magnetic measurements consist of magnetization as a function of temperature, time and
applied field. Our transverse susceptibility is being used for the first time to observe
interfacial coupling in multilayers as well as surface anisotropy induced from a core-
shell-like spin configuration in exotic particles. The ferroelectric tester will be used to
measure P-V characteristics of the ferroelectric phases of multilayers and composites. To
add to the existing techniques, a low frequency impedance probe will be built and
integrated into our PPMS to measure the electrical response under a range of conditions
including changing temperature and magnetic field.
Composite made of magnetic nanoparticles embedded with a ferroelectric
polymer matrix have the potential to contribute to device and sensor technology due to
the excellent transduction properties of PVDF. Embedding magnetic nanoparticles within
them keeps with the theme of exploring possible coupled phenomena while evaluating
the composites for use in biosensing. There is much work to be done on this particular
system as synthesis parameters such as solvents and surface treatments for the
nanoparticles still need to be optimized. After optimization, there are several
measurements that can be done including basic and novel magnetic measurements as well
as ferroelectric and impedance measurements. This system promises to posses many
useful and possibly new properties.
Composites in which both parts are nanoparticles are the smallest multifunctional
materials known. Au and Fe3O4 are two nanoparticle systems that are well-studied and
currently under investigation for a myriad of biological applications. Au is known for
being able to attach via thiol chemistry to biological entities while remaining
biocompatible. Fe3O4 is being explored for magnetic site-specific drug delivery and
hyperthermia. We have already done many experiments to study Au-Fe3O4 composite
nanoparticles grown into dumbbell and flower shapes. These systems have shown to
have extremely complex magnetic properties. The flower particles display exchange
biased hysteresis loops and TS curves as well as the training effect. Both types display
the memory effect, a time dependent magnetization phenomenon. Further tests of
magnetic relaxation are planned to better understand the time dynamics of these particles.
We propose to do more research with Au and Fe3O4 as a core-shell system which could
have simultaneous application as a site specific drug delivery tool and as a tool for
Preliminary results for this project include the growth, characterization, and
magnetic measurement of BaM/BSTO multilayer thin films. Testing this system with our
ferroelectric tester will help complete the overall characterization while impedance
measurements will be used to explore any possible ME coupling present in these films.
Another multilayer thin film system that has been measured consists of
CrO2/Cr2O3 bilayers. We have used systematic M-H measurements to deduce exchange
coupling present in these films without the presence of exchange bias. This is a result of
the epitaxial growth of the films leaving no uncompensated spins in the AFM layer.
Instead, the increase in HC implies a uniaxial anisotropy without a unidirectional
anisotropy. This finding was reaffirmed with several TS measurements which show
behavior much different than pure CrO2 films. Again, there are electrical measurements
which should be undertaken to try to understand the dynamics better. Since Cr2O3 is a
ferroelectric as well as an antiferromagnetic, P-V curves will be useful to add to the
overall picture of this system. Impedance measurements in a magnetic field can also help
us determine if there is ME coupling present on top of the already confirmed exchange
coupling seen in these films.
A different kind of coupling has been observed in bilayers and trilayers of LCMO
and YBCO below the superconducting temperature. While many TS measurements have
revealed behavior already seen in superconductors such as electron depairing and vortex
penetration, extensive analysis is needed to better understand the interplay of
ferromagnetism and superconductivity especially in the context of the proximity effect.
Future sensors and devices will not only need to be miniaturized and composed of
several different materials to improve upon existing technology, but perhaps even take
advantage of the interesting behavior that occurs when they are combined.
Diandra L. Leslie-Pelecky and Ruben D. Rieke. Chem. Mater. 8, 1770 (1996).
W. H. Meiklejohn and C. P. Bean, Phys. Rev. 102, 1413 (1956).
J. Nogués, J. Sort, V. Langlais, V. Skumryev, S. Suricach, J. S. Mucoz, and M. D. Bary
Phys. Rep. 422, 65 (2005).
Nicola A. Spaldin and Manfred Fiebig. Science. 309, 391 (2005).
Gang Liu, Ce-Wen Nan, Ning Cai, and Yuanhua Lin. J. Appl. Phys. 95, 2660 (2004).
Pavel Borisov, Andreas Hochstrat, Xi Chen, Wolfgang Kleemann, and Christian Binek.
Phys. Re. Lett. 94, 117203 (2005).
H. Srikanth, J. Wiggins and H. Rees. Review of Scientific Instruments 70, 3097 (1999).
L. Spinu, C.J. O’Connor and H. Srikanth. IEEE Transactions on Magnetics 37, 2188
A. Aharoni, E.H. Frei, S. Shtrikman and D. Treves. Bull. Res. Counc. Isr.l 6A, 215
J. Nogués and I. K. Schuller. J. Magn. Magn. Mater. 192, 203
Siegrfried Bauer, Simona Bauer-Gogonea, Mario Danaschmüller, Pilles Dennler, Ingrid
Graz, Martin Kaltenbrunner, Christoph Keplinger, Howard Reiss, Niyazi Serdar
Sariciftci, Thokchom Birenda Singh and Schwödiauer. Mater. Res. Soc. Symp. Proc.
889, W01.02.1 (2006).
H. S. Nalwa (editor) (1995). Ferroelectric Polymers: Chemistry, Physics and
Applications. Marcel Dekker, Inc.
P. Inácio, J. N. Marat-Mendes, and C. J. Dias. Materials Science Forum. 455-456, 411
Ce Wen Nan, Ming Li, Xiqiao Fend and Shouwen Yu. Appl. Phys. Lett. 78, 2527
J. Gass, P. Poddar, J. Almand, S. Srinath, and H. Srikanth. Adv. Fun. Mater. 16, 71
Q.A. Pankhurst, J. Connolly, S.K. Jones and J. Dobson. J. Phys. D. 36, R167 (2003).
Catherine C. Berry and Adam S.G. Curtis. J. Phys. D. 36, R198 (2003).
H. Yu, M. Chen, P. Rice, S. X. Wang, R. L. White, and S. Sun. Nano Lett. 5, 379
R.K. Zheng, G.H. Wen, K.K. Fung, and X.X. Zhang. Phys. Rev. B. 69, 214431 (2004).
R. H. Kodama, A. E. Berkowitz, E. J. McNiff, Jr., and S. Foner. Phys. Rev. Lett. 77,
F. Bodker, S. Mørup, and S. Linderoth. Phys. Rev. Lett. 72, 282 (1994).
H. Kachkachi, A. Ezzir, M. Nogues, and E. Tronc. Eur. Phys. J. B. 14, 681 (2000).
J. Im, O. Auciello, P.K. Baumann, and S.K. Streiffer. Appl. Phy. Lett. 76, 625
T.L. Hylton, M.A. Parker, M. Ullah, K.R. Coffey, R. Umphress, and J.K. Howard,
J. Appl. Phys. 75, 5960 (1994).
G. T. Rado and V. J. Flen, Phys. Rev. Lett. 7, 310 (1961).
A. Gupta and J. Z. Sun. J. Magn. Magn. Mater. 200, 24 (1999).
G. Miao, G. Xiao, and A. Gupta, Phys. Rev. B 71, 094418 (2005).
T. C. Schulthess and W. H. Butler. Phys. Rev. Lett. 81, 4516 (1998).
Z. Sefrioui, D. Arias, V. Peña, J.E. Villegas, M. Varela, P. Prieto, C. León, J.L.
Martinez, and J. Santamaria. Phys. Rev. B. 67, 214511 (2003).
A. Hoffmann, S.G.E. te Velthuis, Z. Sefioui, J. Santamaria, M.R. Fitzsimmons, S. Park
and M. Varela. Phys. Rev. B. 72, 140407(R) (2005).
V. Peña, Z. Sefrioui, D. Arias, C. León, J. Santamaria, J.L Martinez, S.G.E. te Velthuis
and A. Hoffmann. Phys. Rev. Lett. 94, 057002 (2005).
S. Oxx, D.P. Choudhury, Balam A. Willemsen, H. Srikanth, S. Sridhar, B.K. Cho, and
P.C. Canfield. Physica C. 264, 103 (1996).
H. Srikanth, L. Spinu, T. Kodenkandath, J.B. Wiley and J. Tallon. J. Appl. Phys. 89,