magnetic properties of batio3/la0.7sr0.3mno3 thin films ... · pdf filemagnetic properties of...
TRANSCRIPT
Magnetic properties of BaTiO3/La0.7Sr0.3MnO3 thin films integrated on Si(100)Srinivasa Rao Singamaneni, Wu Fan, J. T. Prater, and J. Narayan Citation: Journal of Applied Physics 116, 224104 (2014); doi: 10.1063/1.4903322 View online: http://dx.doi.org/10.1063/1.4903322 View Table of Contents: http://scitation.aip.org/content/aip/journal/jap/116/22?ver=pdfcov Published by the AIP Publishing
[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
152.14.71.179 On: Wed, 10 Dec 2014 17:10:36
Magnetic properties of BaTiO3/La0.7Sr0.3MnO3 thin films integrated on Si(100)
Srinivasa Rao Singamaneni,1,2,a) Wu Fan,2 J. T. Prater,1,2 and J. Narayan2
1Materials Science Division, Army Research Office, Research Triangle Park, North Carolina 27709, USA2Department of Materials Science and Engineering, North Carolina State University, Raleigh, North Carolina27695, USA
(Received 24 September 2014; accepted 22 November 2014; published online 10 December 2014)
Two-phase multiferroic heterostructures composed of room-temperature ferroelectric BaTiO3 (BTO)
and ferromagnetic La0.7Sr0.3MnO3 (LSMO) epitaxial thin films were grown on technologically
important substrate Si (100). Bilayers of BTO/LSMO thin films display ferromagnetic Curie
transition temperatures of �350 K, close to the bulk value, which are independent of BTO films
thickness in the range of 25–100 nm. Discontinuous magnetization jumps associated with BTO
structural transitions were suppressed in M(T) curves, probably due to substrate clamping effect.
Interestingly, at cryogenic temperatures, the BTO/LSMO structure with BTO layer thickness
of 100 nm shows almost 2-fold higher magnetic coercive field, 3-fold reduction in saturation
magnetization, and improved squareness compared to the sample without BTO. We believe that the
strong in-plane spin pinning of the ferromagnetic layer induced by BTO layer at BTO/LSMO
interface could cause such changes in magnetic properties. This work forms a significant step forward
in the integration of two-phase multiferroic heterostructures for CMOS applications. VC 2014AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4903322]
INTRODUCTION
Multiferroic materials offer the possibility of switching
the magnetization with an electric field and switching of
electric polarization with a magnetic field, which would rep-
resent a major advance for information storage and low
power non-volatile device applications. In that respect, there
have been numerous recent reports of ferromagnetic/ferro-
electric (FM/FE) heterostructures such as La0.7Sr0.3MnO3/
BaTiO3 (LSMO/BTO),1–3 La0.7Ca0.3MnO3/BTO (LCMO/
BTO),3,4 Fe/BTO,5 EuO/BTO,6 and Fe3O4/BTO7 hetero-
structures which show the coupling effect manifested as a
change in magnetization, magnetic anisotropy, and electrical
resistance due to strain- and interface-charge mediated cou-
pling. Very recently,8,9 Udalov et al. have discovered that
long range Coulomb interaction could explain the magneto-
electric coupling in composite granular multiferroics.
Among several materials, LSMO/BTO has emerged as the
prototypical system as both materials show strong ferroic
order parameters well above room temperature. Of particular
interest, LSMO deposited on BTO single crystals exhibited
large magneto-electric coupling due to strain associated
with BTO structural transitions as a function of tempera-
ture.1–3 For instance, Eerenstein et al.1 also demonstrated
electrically induced giant, sharp, and persistent magnetic
changes at the interface of LSMO/BTO. In another intrigu-
ing work,2 E-field induced magnetic modulation was
reported on BTO/LSMO heterostructure. BTO exhibits3
three distinct structural phase transitions: cubic-tetragonal
(C-T) at a temperature T¼ 393 K (also, ferroelectric
Curie point), tetragonal-orthorhombic (T-O) at T¼ 278 K,
and orthorhombic-rhombohedral (O-R) at T¼ 190 K,
respectively.
These materials possess a negligible lattice mismatch
which favours epitaxial growth of BTO/LSMO interfaces
when the thin films of those were grown on lattice matched
SrTiO3 (STO) substrates. In the present day microelectronics
industry, STO is unsuitable for fabricating the devices based
on the multiferroics. The integration of this important class
of multiferroic heterostructures on the CMOS compatible
substrate Si (100) is needed to advance the field and realize
its tremendous future prospects for data storage applications.
To address this, a few studies10,11 have reported on the depo-
sition of BTO thin films on buffered (with SrTiO3) Si (100)
by taking the advantage of molecular beam epitaxy (MBE)
technique to avoid the formation of an interfacial SiOx layer
that would disrupt epitaxial growth. For instance, recently, a
group12 from IBM have reported on the deposition of a BTO
thin film on Si (100), in which, they have used MBE to
deposit a STO buffer layer on Si (100). As early13 as 1995,
the epitaxial growth of BTO on Si (100) has been reported,
deposited by PLD using TiN as a buffer layer, in which, TiN
was used as both top and bottom electrodes. In another
report,14 vertically aligned thin film nanostructures of BTO-
CeO2 have been deposited on silicon substrates. However,
none of the above studies reported on the magnetic proper-
ties of two-phase epitaxial multiferroic BTO/LSMO on
Si(100). In that spirit, we demonstrate the epitaxial integra-
tion of BTO/LSMO thin film heterostructures on Si (100)
using TiN/MgO as buffer layers and report on their ferro-
magnetic properties as a function of BTO layer thickness.
Magnetization jumps associated with the BTO structural
transitions were suppressed in M(T) curves, probably due to
substrate clamping. The ferromagnetic Curie temperature
(TC) of the bilayers is close to that of the parent LSMO film,
and, found to be independent of the BTO layer. More inter-
estingly, we find that the BTO/LSMO bilayer structure hav-
ing a BTO layer thickness of 100 nm shows almost 2-timesa)[email protected]
0021-8979/2014/116(22)/224104/6/$30.00 VC 2014 AIP Publishing LLC116, 224104-1
JOURNAL OF APPLIED PHYSICS 116, 224104 (2014)
[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
152.14.71.179 On: Wed, 10 Dec 2014 17:10:36
higher magnetic coercive field, 3-fold reduction in saturation
magnetization, and improved squareness compared to the
sample without BTO.
EXPERIMENTAL DETAILS
To describe the deposition procedure briefly, TiN, MgO,
LSMO, and BTO targets were ablated sequentially in the
same chamber using a rotating target assembly. The deposi-
tion of TiN layer was performed at 625 �C in vacuum
(1� 10�6 Torr). Following TiN deposition, the first few
mono layers (for about 500 pulses) of MgO were deposited
under vacuum (1� 10�6 Torr) at 575 �C. The remaining
MgO was then deposited at the same temperature under an
oxygen pressure of 6� 10�4 Torr. Finally, LSMO and BTO
depositions were carried out under a O2 partial pressure of
2� 10�1 Torr at a substrate temperature of 750 �C. The
energy density and pulse frequency were 1.5–2.5 J/cm2 and
10 Hz, respectively. Once completed, the samples were
cooled slowly to ambient temperature under the same O2
partial pressure. For ferroelectric measurements, capacitor
structures were made separately using SrRuO3 (SRO) and Pt
as bottom and top electrodes, respectively. Three samples of
BTO/LSMO were prepared by varying the thicknesses
(25, 50, and 100 nm) of BTO layer. We found that the mag-
netization data of 25- and 50-nm thick BTO samples are
similar. Hence, in this letter, we present the data pertaining
to two samples, in which, the thicknesses of BTO layer are
100 and 50 nm, labeled as samples A and B, respectively.
For comparison, the magnetization data of the sample with-
out BTO is also included, labeled as sample C. The thickness
of the LSMO layer is kept constant in all the studied samples
at 217 nm. No post growth oxygen annealing was performed
on any of the structures discussed here. All the magneti-
zation measurements are performed in the plane of the film.
The structures of BTO samples were characterized by
XRD h-2h scans using a Rigaku x-ray diffractometer with
Cu Ka radiation (k¼ 1.5418 A�). XPS was conducted on a
SPECS FlexMod system equipped with an Al Ka
monochromatic x-ray source (1486.7 eV) to identify (if
present) any surface metal contamination and to probe the
oxidation states of the respective elements from BTO. High
temperature XRD measurements were performed by heating
the sample in a high temperature displex up to a maximum
of 700 �C, and a temperature control of 60.5 �C. The micro-
structures of these films were characterized using a JEOL-
2000FX transmission electron microscope (TEM). A detailed
atomic-resolution study at BTO/LSMO/MgO interfaces was
performed, using a JEOL-2010F high resolution TEM
(HRTEM), equipped with a Gatan image filter tuning attach-
ment, which has a point-to-point resolution of 0.18 nm.
RESULTS AND DISCUSSION
In Fig. 1(a), we present the h-2h XRD pattern of the
BTO/LSMO/MgO/TiN/Si (100) heterostructure. For clarity,
we show the zoomed version of the Bragg peaks correspond-
ing to the LSMO and BTO (002) reflections in Figs. S1(a)
and S1(b).15 It is evident from this pattern that all the layers
show preferential (00l) orientation, suggesting either the tex-
tured or epitaxial growth of the multilayered structure sug-
gesting the presence of 90� ferroelectric domains in BTO.
From the 2h XRD data for the (002) peak, we determined the
out-of-plane (OOP) lattice parameter of BTO layer for sam-
ple A ands B as 3.993 and 3.985 A, respectively, less than
the bulk value16 of 4.036 A, indicating that these BTO layers
are under compression in OOP, suggesting that the film is
under tension in the plane of the film. The c-axis lattice con-
stants of LSMO layers in samples A, B, and C are estimated
as 3.869, 3.857, and 3.854 A, respectively, while the bulk
LSMO shows 3.870 A, with the residual strain of about
1.5%–0.5%. The epitaxial growth and the in-plane (IP) ori-
entation of all the three layers were studied in detail by
means of u-scan XRD. As depicted in Fig. 1, the u-scan
patterns of (111) reflection for BTO, LSMO, and Si were
collected. This pattern shows 4 peaks separated by �90�
indicating its cubic symmetry and establishing the cube-on-
cube relationship of the BTO with the underlying substrate
FIG. 1. (a) Typical h-2h (out of plane) XRD patterns of BTO/LSMO/MgO/TiN/Si (100) heterostructures (samples A, B, and C) showing high quality, single
phase and only (00l) reflections. (b) u-scan XRD patterns of BTO, LSMO, and Si of (111) reflection collected from sample A at 2h¼ 38.90�, x¼ 19.45�, and
v¼ 55.07� for BTO; 2h¼ 40.20�, x¼ 20.10�, and v¼ 55.07� for LSMO; and 2h¼ 28.46�, x¼ 14.23�, and v¼ 54.74� for Si (100). This pattern shows 4 peaks
separated by �90� indicating its pseudo cubic/rhombohedral symmetry, establishing the cube-on-cube relationship with the underlying substrate Si (100).
224104-2 Singamaneni et al. J. Appl. Phys. 116, 224104 (2014)
[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
152.14.71.179 On: Wed, 10 Dec 2014 17:10:36
Si (100), confirming that all 3 layers are grown epitaxially
cube-on-cube, i.e., (001)epilayer//(001)buffer and [110]epi-
layer//[110]buffer. Figure S1(c) shows15 the rocking curve
(to evaluate mosaicity, which is, angular dispersion along the
growth direction) of the BTO (002) peak; where the full
width at half maximum (FWHM) is about 1� to 1.3�.Fig. 2(a) is a typical bright-field cross-section TEM
image of the sample B, in which the BTO/LSMO/MgO/TiN/
Si (100) layers are labeled. The thicknesses of BTO, LSMO,
MgO, and TiN are determined as �50, 217, 175, and
133 nm, respectively. It should be emphasized that the
epitaxial growth of BTO/MgO on Si (100) is made possible
due to the epitaxial growth of large mismatched system
based on the DME paradigm, e.g., TiN on Si (100) where
four lattice constants of TiN match with three of Si
(100).10,11 An important feature of DME concept10,11,17 is
that most of the strain is relieved almost immediately upon
initiation of growth, i.e., within the first couple of mono-
layers of growth. In this way, lattice misfit strain accommo-
dation is confined to the interface making it possible for the
rest of the film to be grown free of defects and lattice strain.
TiN has good lattice match with MgO, which in turn, has
good lattice match with LSMO. In addition, TiN also pro-
vides a good diffusion barrier. More details on TiN/Si depo-
sition can be found in our earlier work.10,11 Figures 2(b) and
2(c) present typical high resolution electron microscopic
(HREM) images taken at the LSMO/MgO and BTO/LSMO
interfaces, showing clean and coherent growth. The tempera-
ture dependent x-ray diffraction, Raman, and ferroelectric
polarization measurements clearly show the room tempera-
ture ferroelectric nature of BTO films. Temperature depend-
ent I-V measurements show hysteric non-leaky ferroelectric
behavior. The in-detail experimental findings are communi-
cated in another paper.18 For clarity, we present the
polarization-voltage (P-V) hysteresis measurements col-
lected for several BTO devices are shown in Fig. S2.15
The magnetic response of the BTO/LSMO heterostruc-
ture with variation in temperature, measured by the use of
a superconducting quantum interference device (SQUID)
magnetometer (Quantum Design, MPMS-XL), is presented
in Fig. 3, for all the three samples. The temperature depend-
ence of the in-plane magnetization (M vs T), was measured
in small magnetic fields of 20 and 100 Oe after cooling the
samples from T¼ 400 K to T¼ 4 K in the plane of the sam-
ple, i.e., in the (100) plane of BTO under no magnetic field
(zero field cooled). As M(T) curves measured both at 20
and 100 Oe show similar behavior, we show the data corre-
sponding only to 100 Oe measuring field. The data were
collected15 (see Fig. S3) while cooling and heating the sam-
ple and found no thermal hysteresis. All the samples show
clear saturation at low temperature, though with varying
saturation magnetization (discussed later). The saturation
FIG. 2. (a) Bright field cross-section TEM image taken from sample B, where BTO (�50 nm) film was grown at 750 �C. All 4-layers are marked. The scale
bar is 100 nm (b) HRTEM image of LSMO (217 nm)/MgO (175 nm) interface. The scale bar is 5 nm (c) HRTEM image of BTO (50 nm)/LSMO (217 nm) inter-
face. The scale bar is 2 nm. The two interfaces are clean and sharp without inter diffusion and secondary phases.
224104-3 Singamaneni et al. J. Appl. Phys. 116, 224104 (2014)
[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
152.14.71.179 On: Wed, 10 Dec 2014 17:10:36
magnetization (�380 emu/cc) of LSMO without BTO is in
close agreement with the reported values.3 Two worthy
observations can be noted: (a) the ferromagnetic Curie tem-
perature (TC), the point at which the magnetization rises, is
found to be independent of BTO layer thickness, consistent
with the earlier reports and (b) surprisingly, the large, sharp,
and discontinuous jumps in the temperature dependent mag-
netization data associated with BTO structural transforma-
tions are absent in both samples A and B. We noticed that
such behavior is observed when the data were collected dur-
ing both the heating and cooling cycles. In contrast, such
jumps were observed when thin films of Fe,5 Fe3O4,7
LCMO3,4 and LSMO1–3 were deposited on BTO single crys-
tal substrates.
In Figure 4, we present the isothermal M-H data col-
lected on three samples at 4 K, when the samples were
cooled under 100 Oe. Three profound observations can be
noted: (i) the squareness (ratio of saturation to remnant mag-
netization) of M-H loop is improved with the increase of
BTO film thickness varying from 0–100 nm, (ii) the coercive
field (Hc) of sample A (�365 Oe) is almost twice that of
sample C (�163 Oe), and (iii) the saturation magnetization
of sample A is reduced to about 1/3 that of sample C. We
have noticed that the same trend (data not shown) is followed
when the measurements were performed at several higher
cooling fields such as 6300, 6500, 61000, and 62000 Oe
collected at 4 K, 170 K, 190 K, 270 K, and 290 K, below the
ferromagnetic Curie temperature of (�350 K) LSMO. The
observed trend is found to be independent of the polarity of
the field cooling. This indicates that the observed magnetic
features as a function of BTO film thickness are not associ-
ated with the BTO structural variations caused by the tem-
perature sweep and applied magnetic fields.
Now, we discuss the possible sources that might lead to
strong BTO-thickness dependence on LSMO magnetic prop-
erties. M vs T heating and cooling branches do not exhibit
distinct anomalies and concomitant thermal hysteresis at
BTO T-O and O-R structural transitions. The arrest of mag-
netization jumps could be due to the clamping of BTO/
LSMO heterostructure to the underlying Si (100) as it was
recently19 argued in the case of Fe/BTO fully epitaxial films.
The substrate may constrain the expansion accompanying
the structural transitions. This is a kind of clamping effect is
common for thin films. If the BTO goes into a multi-domain
state during the cubic to tetragonal transition where 50% of
the structural domains have a c-axis in the plane and 50% of
the structural domains have a c-axis normal to the plane, the
effect may be diminished or absent.
In addition, the cross-sectional HREM images (see
Figs. 2(b) and 2(c)) did not manifest the appearance of any
defects close to the interface in the investigated films. In
addition, the temperature-dependent magnetization curves
suggest the existence of just a single-phase magnetic state. If
there was any phase separation or magnetic inhomogeneity
in the bulk of the LSMO film, one would expect a clear ther-
mal hysteresis between the cooling and heating runs of M(T)
curves. Oxygen vacancies are known to severely affect the
magnetic and magneto transport properties of LSMO/BTO
bilayers and superlattices deposited on STO substrate.20
Oxygen vacancies are not believed to explain the observed
change in magnetic properties since the same O2 partial pres-
sure has been used in all the bilayer samples deposition. The
newly discovered8,9 long range Coulomb interaction mecha-
nism recently forwarded to explain magneto electric cou-
pling for granular multiferroic materials could not be applied
in the present case as the geometry of our thin film hetero-
structures is significantly different.
Similar to our case, Alberca and co-authors previously
have reported21 that magnetization at 10 K was reduced
almost 2-fold and coercive field was enhanced 8-fold when
12 nm-thick ferromagnetic LCMO layer is in contact with a
(001)-oriented ferroelectric BTO substrate measured, com-
pared to LCMO deposited on bare STO substrates. In addi-
tion, they found that the ferromagnetic Curie temperature of
FIG. 3. Magnetization vs. temperature (M-T) curves of sample A (in black),
sample B (in red), and sample C (in blue) for all the structures. The data
were collected during the cooling cycle with the measuring field of 100 Oe.
As it can be noticed, the Curie temperature (TC) of all the samples is found
to be the same at �350 K. The magnetic field is applied along h100i direc-
tion of the sample.
FIG. 4. Comparison of isothermal (4 K) M-H curves measured on all the
samples cooled from 400 K under 100 Oe. As one can notice, sample A
shows 2-times higher Hc, much improved squareness (Ms/Mr) and 3-times
lesser saturation magnetization (Ms) in comparison with sample C (with no
BTO). All the experiments were conducted under the same conditions as
mentioned above. The magnetic field is applied along h100i direction of the
sample.
224104-4 Singamaneni et al. J. Appl. Phys. 116, 224104 (2014)
[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
152.14.71.179 On: Wed, 10 Dec 2014 17:10:36
LCMO was reduced from 270 K to 200 K. They have
explained these magnetic features to be a result of non-
uniform strain field in LCMO/BTO caused by the corruga-
tion of the ferroelectric domains in the rhombohedric phase
of the BTO bulk crystal. Though at the microscopic level,
we cannot rule out the above possibility, this situation is
unlikely in the present case since we did not observe thermal
hysteresis between cooling and heating M(T) runs (see
Fig. S3), caused by inhomogeneous strain fields that would
generate inhomogeneous magnetic phases. The bright field
transmission electron microscopy and high resolution elec-
tron microscopy images (see Figs. 2(a)–2(c)) show continu-
ous epitaxial films without granularity. We observed the
Curie temperature (350 K) close to the value reported
(360 K) for bulk single crystalline LSMO.22 Furthermore, we
performed zero field cooled (ZFC) and field cooled (FC)
temperature dependent magnetization (M-T) measurements
on sample A (see Fig. S4). We see no indication of a block-
ing temperature, a characteristic feature of inhomogeneous
magnetic phases. Thus, our comprehensive experimental
findings provide no evidence for the presence of inhomoge-
neous phases/grain boundary effects. If a significant inhomo-
geneous phase is present, the Curie temperature is
expected23 to be widely different from the bulk. Hence, in
the present case, our experimental observations led us to
believe that grain boundaries do not affect the magnetic na-
ture of the sample. The nature of thin film epitaxy minimizes
the large angle grain boundaries unlike bulk polycrystalline
materials.
In addition, we did not see a clear strain effect of the
BTO layer on the ferromagnetic properties of the LSMO
layer either in the form of discontinuous magnetization
jumps or varying ferromagnetic Curie temperature as the
structural transitions of BTO layer seem to be arrested or
broadened. Our results are consistent with the findings of a
recent spin wave resonance study,24 reported on
LSMO(330 nm)/BTO(270 nm) epitaxial heterostructure de-
posited on LaAlO3. This particular study concludes that the
BTO layer itself induces in-plane surface pinning at the
BTO/LSMO interface that modifies the in-plane bulk mag-
netic anisotropy of the LSMO film. Due to strong in-plane
pinning between BTO and LSMO induced by the ferroelec-
tric BTO layer, a strong reduction in saturation magnetiza-
tion and enhanced coercive field are observed, and appear to
vary as a function of the stress state of the BTO. We believe
that the stress state of the BTO layer directly influences the
reordering of the atoms at the BTO-LSMO interface and
thus affects the extent of magnetic pinning in the LSMO
layer. To support our hypothesis and to prove that the ferroe-
lectricity in BTO causing the changes in LSMO magnetic
properties, we have inserted a thin (�10–15 nm) insulating-
non-magnetic-non-ferroelectric SrTiO3 layer between BTO
and LSMO and measured its magnetic properties (data not
shown). We observed that the corresponding M-H and M-T
curves behave very similar to the sample C without BTO
layer and retains all the pristine (magnetization, ferromag-
netic Curie temperature, and coercive field) characteristics of
LSMO layer, thus confirming that it is most likely that the
ferroelectric nature of BTO causing the changes in the
magnetic properties of LSMO. To explore this further, we
are planning to perform temperature dependent MOKE
measurements on our samples to probe the super exchange
coupling between Mn and Ti ions, which could lead to the
observed reduction in magnetization and enhancement in co-
ercive field.
SUMMARY
In summary, in the present work, we have demonstrated
the effect of ferroelectric BTO layer thickness on the ferro-
magnetic properties of LSMO layer, which have been epitax-
ially deposited on the technologically important substrate Si
(100). Bilayers of BTO/LSMO thin films preserve good fer-
romagnetic properties with a ferromagnetic Curie tempera-
ture of �350 K, close to the expected value, and found to be
independent of BTO films thickness of 25–100 nm. Our data
show that the discontinuous magnetization jumps associated
with BTO structural transitions were suppressed in M(T)
curves, probably due to substrate clamping effect. More
interestingly, the bilayer structure with BTO layer thickness
of 100 nm shows almost 2-fold higher magnetic coercive
field, 3-fold reduction in saturation magnetization, and
improved squareness compared to the sample without BTO.
We believe that the strong in-plane spin pinning by BTO
ferroelectric layer at the BTO/LSMO interface could lead to
the observed magnetic properties of LSMO in contact with
BTO. This work demonstrates that we could manipulate the
magnetic properties of ferromagnetic layer by conjoining
with a ferroelectric layer. Our work makes a promising step
in the realization of multiferroic materials for CMOS
applications.
ACKNOWLEDGMENTS
S.R.S. acknowledges financial support from the
National Academy of Science (NAS), USA in the form of
NRC postdoctoral research associate fellowship. We thank
Chrstian Binek, University of Nebraska and N. D. Mathur,
University of Cambridge for stimulating discussion. Also,
the authors are pleased to acknowledge the support of the
Army Research Office under Grant W911NF-04-D-0003.
Also, the authors acknowledge the use of the Analytical
Instrumentation Facility (AIF) at North Carolina State
University, which was supported by the State of North
Carolina and the National Science Foundation.
1W. Eerenstein, M. Wiora, J. L. Prieto, J. F. Scott, and N. D. Mathur, Nat.
Mater. 6, 348 (2007).2M. K. Lee, T. K. Nath, C. B. Eom, M. C. Smoak, and F. Tsui, Appl. Phys.
Lett. 77, 3547 (2000).3H. Lu, T. A. George, Y. Wang, I. Ketsman, J. D. Burton, C.-W. Bark, S.
Ryu, D. J. Kim, J. Wang, C. Binek, P. A. Dowben, A. Sokolov, C.-B.
Eom, E. Y. Tsymbal, and A. Gruverman, Appl. Phys. Lett. 100, 232904
(2012).4A. Alberca, N. M. Nemes, F. J. Mompean, T. Feh0er, F. Simon, J. Tornos,
C. Leon, C. Munuera, B. J. Kirby, M. R. Fitzsimmons, A. Hernando, J.
Santamaria, and M. Garcia-Hernandez, Phys. Rev. B 88, 134410 (2013).5S. Sahoo, S. Polisetty, C.-G. Duan, S. S. Jaswal, E. Y. Tsymbal, and C.
Binek, Phys. Rev. B 76, 092108 (2007).6S. Cao, P. Liu, J. Tang, H. Lu, C.-W. Bark, S. Ryu, C. B. Eom, A.
Gruverman, and P. A. Dowben, Appl. Phys. Lett. 102, 172402 (2013).
224104-5 Singamaneni et al. J. Appl. Phys. 116, 224104 (2014)
[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
152.14.71.179 On: Wed, 10 Dec 2014 17:10:36
7H. F. Tian, T. L. Qu, L. B. Luo, J. J. Yang, S. M. Guo, H. Y. Zhang, Y. G.
Zhao, and J. Q. Li, Appl. Phys. Lett. 92, 063507 (2008).8O. G. Udalov, N. M. Chtchelkatchev, and I. S. Beloborodov, Phys. Rev. B
89, 174203 (2014).9O. G. Udalov, N. M. Chtchelkatchev, A. Glatz, and I. S. Beloborodov,
Phys. Rev. B 89, 054203 (2014).10J. Narayan, Acta Mater. 61, 2703 (2013).11J. Narayan, P. Tiwari, X. Chen, R. Chowdhury, and T. Zheleva, Appl.
Phys. Lett. 61, 1290 (1992).12C. Dubourdieu, J. Bruley, T. M. Arruda, A. Posadas, J. Jordan-Sweet, M.
M. Frank, E. Cartier, D. J. Frank, S. V. Kalinin, A. A. Demkov, and V.
Narayanan, Nat. Nanotechnol. 8, 748 (2013).13M.-B. Lee, M. Kawasaki, M. Yoshimoto, and H. Koinuma, Appl. Phys.
Lett. 66, 1331 (1995).14F. Khatkhatay, A. Chen, J. H. Lee, W. Zhang, H. Abdel-Raziq, and H.
Wang, ACS Appl. Mater. Interfaces 5, 12541 (2013).15See supplementary material at http://dx.doi.org/10.1063/1.4903322 for the
shift of LSMO (002) XRD reflection collected for all three samples and
shift of BTO (002) XRD reflection collected for all three samples. The
rocking curve with FWHM of 1�–1.3� of BTO (002) diffraction, P-V hys-
teresis measurements collected for several BTO devices, M(T) curves col-
lected from sample A while cooling and heating runs measured under 100
Oe, showing no thermal hysteresis and discontinuous magnetization
jumps.16M. El Marssi, F. Le Marrec, I. A. Lukyanchuk, and M. G. Karkut, J. Appl.
Phys. 94, 3307 (2003).17S. S. Rao, J. T. Prater, F. Wu, C. T. Shelton, J.-P. Maria, and J. Narayan,
Nano Lett. 13, 5814 (2013).18S. R. Singamaneni, S. Punugupati, J. T. Prater, F. Hunte, and J. Narayan,
J. Appl. Phys. 116, 094103 (2014).19G. Radaelli, D. Petti, M. Cantoni, C. Rinaldi, and R. Bertacco, J. Appl.
Phys. 115, 172604 (2014).20P. Murugavel and W. Prellier, J. Appl. Phys. 100, 023520 (2006).21A. Alberca, N. M. Nemes, F. J. Mompean, N. Biskup, A. de Andres, C.
Munuera, J. Tornos, C. Leon, A. Hernando, P. Ferrer, G. R. Castro, J.
Santamaria, and M. Garcia-Hernandez, Phys. Rev. B 84, 134402
(2011).22A. Khapikov, L. Uspenskaya, I. Bdikin, Y. Mukovskii, S. Karabashev, D.
Shulyaev, and A. Arsenov, Appl. Phys. Lett. 77, 2376 (2000).23A. Alberca, C. Munuera, J. Tornos, F. J. Mompean, N. Biskup, A. Ruiz, N.
M. Nemes, A. de Andres, C. Leon, J. Santamar0ıa, and M. Garc0ıa-
Hernandez, Phys. Rev. B 86, 144416 (2012).24Y.-L. Zhao, Y. Sun, L.-Q. Pan, K.-S. Li, and D.-B. Yu, Appl. Phys. Lett.
102, 042404 (2013).
224104-6 Singamaneni et al. J. Appl. Phys. 116, 224104 (2014)
[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
152.14.71.179 On: Wed, 10 Dec 2014 17:10:36