Anisotropic manipulation of ferroelectric polarization in SrTiO3/(Co0.9Zn0.1)Fe2O4 heterostructural films by magnetic field
Bo Chen, Yong-Chao Li, Jun-Yong Wang, Jian-Guo Wan, and Jun-Ming Liu
Citation: Journal of Applied Physics 115, 044102 (2014); doi: 10.1063/1.4863144 View online: http://dx.doi.org/10.1063/1.4863144
View Table of Contents: http://scitation.aip.org/content/aip/journal/jap/115/4?ver=pdfcov Published by the AIP Publishing
Anisotropic manipulation of ferroelectric polarization in
SrTiO
3/(Co
0.9Zn
0.1)Fe
2O
4heterostructural films by magnetic field
Bo Chen, Yong-Chao Li, Jun-Yong Wang, Jian-Guo Wan,a)and Jun-Ming Liu National Laboratory of Solid State Microstructures and Department of Physics, Nanjing University, Nanjing 210093, People’s Republic of China
(Received 6 September 2013; accepted 11 January 2014; published online 24 January 2014) Multiferroic SrTiO3/(Co0.9Zn0.1)Fe2O4 (STO/CZFO) films with preferential crystallographic orientations were prepared by a sol-gel process. The films exhibited evident ferroelectricity and well-defined ferromagnetic characteristics with certain magnetic anisotropy. Remarkable suppression of ferroelectric polarization by in-plane magnetic field and great enhancement under out-of-plane magnetic field were realized, and large anisotropic magnetodielectric effect was observed. We showed that the ferroelectric polarization of the whole film was closely related to the defect dipoles in the STO layer. Based on the model where the dissociation or formation of defect dipoles is associated with the volume change of STO unit cells, we elucidated the mechanism of anisotropic magnetic-manipulation of ferroelectric polarization for the films, and attributed it to the rearrangement of oxygen vacancies in the STO layer, which was controlled by the ferromagnetic CZFO layer through interface coupling under external magnetic field.VC 2014 AIP Publishing LLC.
[http://dx.doi.org/10.1063/1.4863144]
I. INTRODUCTION
In recent years, anisotropic magnetic-manipulation of ferroelectric polarization, i.e., ferroelectric polarization is enhanced by external magnetic field in one direction while suppressed by magnetic field in another direction, has attracted much interest due to its potential applications in detecting vector magnetic field, designing multi-state stor- age, logical switching devices, etc.1–3 Many efforts have been made to search possible material systems, and some multiferroic compounds such as triangular-lattice helimag- nets (e.g., CuFe1xAlxO2) and copper-oxide superconductors (e.g., La2CuO4þx) have been found to exhibit such kind of effect.4–6 Unfortunately, in most compounds ever discov- ered, such anisotropic magnetic-manipulation of ferroelectric polarization can only be realized at quite low temperature (below 10 K) and very high magnetic field (above 5 T), while the manipulation is rather small (only about60.02lC/cm2), which is disadvantageous to actual applications.
It is known that at room temperature (RT) and low mag- netic field, the magnetoelectric (ME) coupling in a ferroelectric-ferromagnetic laminate is much stronger than that in a multiferroic compound,7 thereby large anisotropic response of ferroelectric polarization to magnetic field may be realized in the multiferroic laminate by strain engineering.
Herein, we propose an available avenue to achieve RT aniso- tropic magnetic-manipulation of ferroelectric polarization by combining a strained SrTiO3(STO) layer with a ferromag- netic oxide layer. Free STO is well known to be paraelectric even at low temperature, nevertheless, RT ferroelectric polarization can be observed in a STO thin film due to the combined effects of mismatch strain and defect dipoles.8–12 More interestingly, ferroelectric polarization induced by
defect dipoles is sensitive to the volume of STO unit cells since the formation or dissociation of defect dipoles is closely associated with the volume change of STO unit cells.13For a heterostructural film comprised of STO and fer- romagnetic oxide layers, anisotropic magnetostrictive response of ferromagnetic layer to external magnetic field makes it possible to modulate the volume of STO unit cells through interface stress coupling.14Accordingly, anisotropic response of ferroelectric polarization to magnetic field can be expected.
In this work, we report on the preparation of multiferroic STO/(Co0.9Zn0.1)Fe2O4 (CZFO) heterostructural films with preferential crystallographic orientations and the realization of RT anisotropic magnetic-manipulation of ferroelectric polarization in them. We show that the existence of defect dipoles in STO induces evident RT ferroelectric polarization.
Moreover, the films exhibit remarkable drop of ferroelectric polarization under in-plane magnetic field and great enhancement under out-of-plane magnetic field. In addition, large anisotropic magnetodielectric effect is observed. Based on the model where the dissociation or formation of defect dipoles is associated with the volume change of STO unit cells, we give a detailed analysis on such anisotropic manip- ulation of ferroelectric polarization by external magnetic field.
II. EXPERIMENTAL DETAILS
Three-layered STO/CZFO/STO heterostructural films were prepared on the Pt/Ti/SiO2/Si wafers. The bottom STO layer was used as a buffer layer to ensure both the middle CZFO layer and upper STO layer to grow well along prefer- ential crystallographic orientations. In addition, the usage of STO buffer layer is beneficial for weakening the substrate restriction on the upper layers to a certain degree. The films were prepared using a typical sol-gel and spin-coating
a)Author to whom any correspondence should be addressed. Email:
0021-8979/2014/115(4)/044102/6/$30.00 115, 044102-1 VC2014 AIP Publishing LLC
JOURNAL OF APPLIED PHYSICS115, 044102 (2014)
process.15,16 The 0.2 M STO and CZFO 2-methoxyethanol solutions were synthesized as the precursors. A 100 nm STO bottom layer was first grown on the Pt/Ti/SiO2/Si wafer as the buffer layer. It was rapidly annealed at 650C for 5 min under oxygen atmosphere. Then the CZFO layer and another STO layer were sequentially deposited onto the surface of the buffer STO layer by similar process, nevertheless, both layers were annealed under air atmosphere so as to produce a certain amount of oxygen vacancies. For comparison, we also prepared a single STO film with 100 nm thickness on the Pt/Ti/SiO2/Si wafer. The preparing process was the same as that of the STO buffer layer. For electric measurements, Au electrodes with diameter of 0.2 mm were deposited onto the film surface using ion-sputtering process.
The cross section and surface morphologies of the films were examined using a scanning electron microscopy (SEM). Structural characterizations were carried out by both x-ray diffraction (XRD) measurements on a D/MAX-RA dif- fractometer and Raman scattering measurements on a Ntegra Spectra Nanolaboratory (NT-MDT, Inc.). Ferroelectric polarization measurements were performed on a standar- dized ferroelectric test system (Precision Multiferroic, Radiant, Inc.) by two methods, i.e., ferroelectric hysteresis loop and Positive-Up-Negative-Down (PUND) process. The measuring frequency and width were fixed at 1.0 kHz and 1.0 ms, respectively. The leakage current vs voltage curves were measured using an amperemeter (Keithley 6517B), and field-dependent magnetization measurements were carried out on a superconducting quantum interference device magnetometer.
To demonstrate the magnetic-manipulation of ferroelec- tric polarization, we performed RT PUND measurements at various magnetic fields in both in-plane and out-of-plane directions of the films. To observe the magnetodielectric effect, we used an impedance analyzer (TH2828S) to mea- sure dielectric constants under various in-plane and out-of- plane magnetic fields at a fixed frequency of 1.0 kHz.
III. RESULTS AND DISCUSSION
According to the cross-sectional SEM image of the STO/CZFO heterostructural film in Fig.1(a), the upper STO layer, middle CZFO layer, and bottom STO layer are 60, 120, and 100 nm in thickness, respectively. The surface SEM image in Fig.1(b)shows that the surface of the upper STO layer is smooth and the distribution of crystallized grains is uniform. Fig. 1(c) presents the XRD spectra for both STO/CZFO heterostructure and single STO films. The single STO film, which can be analogous to the STO buffer layer in the STO/CZFO heterostructure, exhibits polycrystalline per- ovskite structures with no evident preferential crystallo- graphic orientation, indicating that the single STO film (or STO buffer layer) bears no lattice match relation to the Pt layer and it is structurally relaxed. Differently, in the STO/CZFO film the STO phase exhibits evident preferred (002) orientation with a (002) texture coefficient of 1.8.17 Since the bottom STO layer is polycrystalline with structur- ally relaxation, the preferential STO orientation should be attributed to the upper STO layer. In addition, one observes
that the CZFO layer is also grown on a preferred (311) orien- tation. In theory, the lattice match between STO (002) layer and CZFO (311) layer is only about 1%. (The lattice con- stants of STO and CZFO bulks are a¼b¼c¼3.905 A˚ and a¼b¼c¼8.395 A˚ , respectively.15,18) Such small mismatch allows the co-orientated growth of both STO and CZFO layers. Therefore, local epitaxy can be expected between the STO layer and CZFO layer. By refining the XRD spectrum in Fig.1(c), we obtained the out-of-plane and in-plane lattice constants of the STO layer to be 3.911 A˚ and 3.916 A˚, respectively. This means that the lattices of upper STO layer are under tensile state, and the volume of STO unit cell is expanded by 0.7%.
To examine whether RT ferroelectricity exists in the present STO/CZFO heterostructural films, we then per- formed ferroelectric measurements by ferroelectric hystere- sis loop and PUND methods. As shown in Fig.2(a), the film exhibits evident ferroelectric hysteresis characteristics with remanent polarization of Pr¼5.3lC/cm2 at Umax¼30 V,
FIG. 1. Cross-sectional SEM image (a) and surface SEM image (b) of the STO/CZFO film. X-ray diffraction spectra (c) and Raman scattering spectra (d) of the STO/CZFO film and single STO film.
FIG. 2. Ferroelectric hysteresis loop (a), PUND 2Prvs voltage curve (b), and current density vs voltage curve (c) of the STO/CZFO film. (d) In-plane and out-of-plane ferromagnetic hysteresis loops of the STO/CZFO film.
044102-2 Chenet al. J. Appl. Phys.115, 044102 (2014)
which is comparable to previous experimental values observed in the strained STO films containing oxygen vacan- cies.10,12From Fig. 2(b), the PUND 2Prvalue is almost the same as the 2Prvalue measured by ferroelectric hysteresis loops, suggesting that the influence of leakage current is neglectable. The current vs voltage curves in Fig. 2(c)also gives a high DC resistivity of1.5109Xcm at 10 V, indi- cating the whole film is indeed insulating. All the above results strongly indicate that the ferroelectric polarization does exist in the present STO/CZFO heterostructural films.
On the other hand, the PUND measurements demonstrated that the ferroelectric polarization was absent in the single STO film, suggesting that the STO buffer layer in the hetero- structure should have no ferroelectricity. This is understand- able since the single STO film is structurally relaxed and bears no strain. Accordingly, we suggest that the ferroelec- tric polarization of the STO/CZFO heterostructural film should be mainly attributed to the upper STO layer.
To explore the origin of ferroelectric polarization in the film, we further carried out Raman scattering measurements, as shown in Fig.1(d). An evident characteristic Raman scat- tering peak located at 480 cm1 is observed in the STO/CZFO film, which is a signature of the existence of fer- roelectric polarization caused by defect dipoles.22,23 For comparison, we also measured the Raman scattering spectra for the single STO film, as shown in Fig. 1(d). No polar mode corresponding to the ferroelectric polarization appears in the single STO film. Since the ferroelectric measurements also showed that the single STO film (STO buffer layer) was non-ferroelectric, we infer that the upper STO layer does play a key role on the generation of ferroelectricity in the whole STO/CZFO film.
In previous investigations, it is reported that a STO thin film will exhibit ferroelectric polarization along the tensile- strain direction only when the tensile-strain exceeds the so- called critical value of 0.54%.9Nevertheless, the calculations according to the XRD results show that the tensile-strains in the upper STO layer along both out-of-plane and in-plane directions are only 0.15% and 0.28%, respectively, much smaller than the critical value, indicating that the ferroelec- tricity in the present STO/CZFO film is not caused by the tensile-strain. Meanwhile, we notice that oxygen vacancies are easily introduced in the STO layer when the film is annealed under low oxygen press (e.g., under air atmos- phere).13It has been demonstrated that in a perovskite STO unit cell the oxygen vacancy can produce a defect dipole by combining an acceptor center on the Ti4þ site.19–21 Such defect dipole may induce the ferroelectricity in STO. To bet- ter understand how the defect dipoles form and how they induce the ferroelectric polarization in the STO layer con- taining oxygen vacancies, we give a detailed analysis as follows.
In a dielectric perovskite ABO3unit cell, when the B4þ site (e.g., Ti4þ or Zr4þ) is substituted by a X3þ impurity (such as Fe3þ or Mn3þ), an acceptor center of Fe0Ti;Zr or Mn0Ti;Zr can be generated. Such acceptor center can be fur- ther combined with the oxygen vacancies and form defect dipoles.19–21For a heterostructural film containing both per- ovskite STO and spinel CZFO layers, during the annealing
process a few Fe3þions in the CZFO layer could diffuse into the STO layer at high temperature, especially near the STO/CZFO interface. As a result, Fe0TiVO defect dipoles would be formed, which are comprised of a Fe0Ti acceptor center and some oxygen vacancies. Such defect dipoles can be switched to any orientation by the electric field. Due to the existence of defect dipoles around the STO/CZFO inter- face, the upper STO layer can be regarded as a bilayer con- sisting of ferroelectric defect-dipole sublayer near the STO/CZFO interface and the dielectric STO sublayer. Based on the theory of dielectric-ferroelectric bilayer,24 we deduced the out-of-plane ferroelectric polarization of the upper STO layer as
PSTO¼vSTO
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi tDD
2vSTOtSTOþ4tDD
r
PDD; (1) where PDD is the spontaneous polarization of defect-dipole sublayer and vSTO is the static electric susceptibility of dielectric STO sublayer. tSTOand tDDare the thicknesses of whole STO layer and ferroelectric defect-dipole sublayer, respectively. Equation(1)implies that the out-of-plane ferro- electric polarization can be modulated by changing the thick- ness ratio of tSTO and tDD as well as the amount of defect dipoles, even if the out-of-plane tensile-strain of the STO layer does not reach the critical strain value.
Fig. 2(d) shows the ferromagnetic hysteresis loops of the STO/CZFO film along both in-plane and out-of-plane directions. Well-defined ferromagnetic characteristics with easy magnetic axis along in-plane direction are observed.
The coercivityHcis 0.47 kOe for in-plane loop and 0.63 kOe for out-of-plane loop, while the in-plane and out-of-plane saturation magnetizations reach 420 and 380 emu/cc, respec- tively, which are larger than those of pure CoFe2O4films.15 Such magnetic anisotropy characteristics should be attributed to the preferential crystallographic orientations in the CZFO layer and its interior compressive stress caused by the local lattice mismatch between CZFO and STO.25 Noticing that the CoFe2O4—based ceramics belong to spine ferrites with negative magnetostrictive effect, when the STO/CZFO film is subject to an external magnetic field, the CZFO layer will produce a compressed strain along the magnetic field direc- tion, meanwhile it will be under tensional stress state in the direction perpendicular to the magnetic field.26By interface coupling and stress transferring, the strain produced in the CZFO layer can influence the volume expansion in the upper STO layer, causing the re-configuration of oxygen vacancies and consequently changing the ferroelectric polarization. In particular, distinctly different magnetomechanical responses of the CZFO layer under in-plane and out-of-plane magnetic fields may further cause anisotropic magnetic-modulation of ferroelectric polarization in the upper STO layer.
We then measured the variation of ferroelectric polariza- tion under external magnetic field (H) for the STO/CZFO films. Fig.3plots the䉭(2Pr) value as a function of H at vari- ous voltages. Here, we define the increment of PUND 2Pr values at different magnetic fields as 䉭(2Pr)¼2Pr(H) 2Pr(0), where 2Pr(H) and 2Pr(0) represent the PUND 2Pr value measured at certain magnetic field H and zero
magnetic field, respectively. Under in-plane magnetic field (H//), the PUND 2Pr value decreases with increasing H//. Typically, the 䉭(2Pr) value reaches 0.74lC/cm2 at 30 V when the H// is equal to 5 kOe. Differently, when an out-of-plane magnetic field (H?) is applied, the dependence of 2Pron H?is entirely opposite, i.e., the 2Prvalue increases with increasing H?. A typical䉭(2Pr) value ofþ0.80lC/cm2 at 30 V is obtained under H?¼5 kOe. In other word, aniso- tropic manipulation of ferroelectric polarization by external magnetic field is realized in the present STO/CZFO film.
Moreover, the modulation of ferroelectric polarization under H// or H? is much larger than previous results (60.02lC/cm2) reported in the multiferroic compounds.4–6
We attribute such anisotropic magnetic-manipulation of ferroelectric polarization to the response of oxygen vacan- cies in the upper STO layer to the strain transferred from the ferromagnetic CZFO layer. Under in-plane magnetic field (H//), the CZFO layer generates in-plane magnetostrictive strains k11 and k12 along x and y axis, respectively, which are further transferred into the upper STO layer through interface coupling. Based on the elastic model of magnetostrictive-piezoelectric laminate composite films,27 the strain transferred to the upper STO layer displays a non- linear distribution along the thickness direction, i.e., the strain mainly concentrates on the ferroelectric defect-dipole sublayer near the STO/CZFO interface while quickly decays in the dielectric STO sublayer far away from the interface.
According to the elastic model and Poisson’s equation,27,28 under external magnetic field, the expected volume expan- sion of STO unit cells in the STO layer can be deduced as
DV H;ð zÞ=V0¼DV0=V0
þ 1zþtCZFO=2 R
1c12
c11
kxþky
; (2) where z is the distance between a certain STO unit cell and STO/CZFO interface, DV0=V0 is the volume expansion of STO unit cell at H¼0, c11and c12are the elastic coefficients of STO, tCZFOis the thickness of CZFO layer, and R is the
curvature radius for the strain distribution in the STO layer.
R value is determined by the thickness and compliance coef- ficient of each layer,27 which is estimated to be 200 nm in this work. Under in-plane magnetic field, kx¼k11 and ky¼k12. Taking the typical parameters DV0=V0¼0.700%, c11¼336 GPa, c12¼107 GPa, k11¼ 250 ppm, and k12¼110 ppm into Eq.(2),28,29the volume expansion of the STO unit cells at the defect-dipole sublayer (z¼0) is decreased from 0.700% under H//¼0 to 0.693% under H//¼5 kOe, nevertheless, the volume expansion of STO unit cells far away from the defect-dipole sublayer (z¼tSTO) is hardly influenced by the magnetic field. It is previously reported that oxygen vacancies are more easily distributed in the STO unit cells with larger volume expansion.13 Therefore, under in-plane magnetic field, the oxygen vacan- cies will immigrate from the defect-dipole sublayer to the dielectric STO sublayer. Consequently, some oxygen vacan- cies will be dissociated from theFe0Tiacceptor centers, lead- ing to the decrease of ferroelectric polarization induced by Fe0TiVO defect dipoles. In another case, upon the applica- tion of out-of-plane magnetic field the CZFO layer generates compressive strain along z axis, i.e., kz¼k11, while tensile strains in the plane, i.e., kx¼ky¼k12. Based on the similar calculations by Eq.(2), we found that the volume expansion of STO unit cells at the ferroelectric defect-dipole sublayer increased from 0.700% under H?¼0 to 0.710% under H?¼5 kOe. This means that the oxygen vacancies immi- grate from the dielectric STO sublayer to the defect-dipole sublayer, so more oxygen vacancies combine with Fe0Ti acceptor centers. As a result, the ferroelectric polarization is enhanced under out-of-plane magnetic field. In a word, the anisotropic magnetic-manipulation of ferroelectric polariza- tion is attributed to the dissociation or formation of defect dipoles under in-plane or out-of-plane magnetic field.
According to the above analysis, we consider that the thickness of ferroelectric defect-dipole sublayer is important for the anisotropic magnetic-manipulation of ferroelectric polarization. Since the volume expansion of STO unit cell is very small (0.7%) in the present STO/CZFO film, the oxy- gen vacancy density in the STO unit cell can be treated to be linearly proportional to its volume expansion.12 Assuming that the probability for combination between oxygen vacan- cies andFe0Tiacceptor center is constant, the change of ferro- electric polarization in the STO layer manipulated by external magnetic field can thus be expressed as24,27
DP Hð Þ ¼vSTO
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi tDD
2vSTOtSTOþ4tDD
r
tSTOtDD
2tSTO
DV H;ð 0Þ DV H;ð tSTOÞ
DV H;ð tSTO=2Þ PDD: (3) Under a given external magnetic field, the ferroelectric polariza- tion in the STO layer is dominantly influenced by two items, i.e.,vSTO ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
tDD=ð2vSTOtSTOþ4tDDÞ
p andðtSTOtDDÞ=2tSTO. If
the thickness of defect-dipole sublayer is very thick (tDD tSTO),ðtSTOtDDÞ=2tSTO will tend to be zero. In this case, the oxygen vacancies are always localized in the defect-dipole layer, irrespective of external magnetic field. As a result, the fer- roelectric polarization of the STO layer cannot be manipulated
FIG. 3.䉭(2Pr) vs in-plane magnetic field (H//) curves and 䉭(2Pr) vs out-of-plane magnetic field (H?) curves measured at various voltages.
044102-4 Chenet al. J. Appl. Phys.115, 044102 (2014)
by external magnetic field. In another case, if the thickness of defect-dipole sublayer is very thin (tDD 0), vSTO ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
tDD=ð2vSTOtSTOþ4tDDÞ
p will tend to be zero, which
causes rather small ferroelectric polarization in the STO layer.
So the ferroelectric polarization of the STO layer almost does not change under external magnetic field. Only when the ratio of tDD/tSTOis moderate, external magnetic field might give rise to a great change in the ferroelectric polarization of the STO layer. Taking a typicalvSTOvalue of 100 into Eq.(3),9we fur- ther calculated 䉭P(H)/PDD using various tDD/tSTO values at in-plane and out-of-plane magnetic fields of 5 kOe, and plotted the䉭P(H)/PDDvalue as a function of tDD, as shown in Figure4.
One clearly observes that there exists an so-called optimized thickness ratio between tDDand tSTO, i.e., tDD/tSTO1/3, where the䉭P(H)/PDDvalue reaches the maximum. In actual experi- ments, the tDD/tSTOvalue may be adjusted by two ways. One is to change the thickness of defect-dipole sublayer by tuning the diffusion of Fe3þ ions into the STO layer during high-temperature annealing process. Another is to change the thickness of the STO layer during the spin-coating process.
Relative experiments will be carried out for further investigations.
Such anisotropic magnetic-manipulation of ferroelectric polarization may lead to anisotropic magnetodielectric (MD) effect. Generally, the response of ferroelectric polarization to magnetic field is accompanied with the evolution of ferro- electric domain activity.30 For instance, under out-of-plane magnetic field, the increase of ferroelectric polarization with increasing H?could cause the enhancement of ferroelectric domain movement, making the ferroelectric domains easier to rotate towards the magnetic field. So the dielectric response to electric field becomes easier, consequently lead- ing to the enhancement of MD coupling. Contrariwise, the application of in-plane magnetic field will cause a sup- pressed MD coupling. Figs.5(a)and5(b)present the dielec- tric constant (eH) as a function of bias voltage (Ubias) measured at various in-plane and out-of-plane magnetic fields for the present STO/CZFO film. The corresponding dielectric loss vs Ubiascurves are given in the insets of Figs.
5(a)and5(b). The low loss (0.08) implies that the leakage currents have little influence on the dielectric polarization.
Upon the application of H//, the dielectric constant monoto- nously decreases with increasing H// (shown in Fig. 5(a)).
Define the magnetodielectric coefficient as MD¼[e(H)–e(0)]/e(0), wheree(H) ande(0) are the dielectric constants under certain magnetic field and zero magnetic field, respectively. A negative MD value as large as 7.9%
at Ubias¼0 V is obtained under H//¼5 kOe. On the contrary, the out-of-plane magnetic field causes an increase of dielec- tric constant (shown in Fig. 5(b)), i.e., positive MD effect.
Typically, under H?¼5 kOe, the MD value reachesþ9.4%
at Ubias¼0 V. The above results indicate that the present het- erostructure possesses large anisotropic MD effect, which is undoubtedly related to the anisotropic response of ferroelec- tric polarization to external magnetic field. The change of dielectric loss with H// or H?(shown in the insets of Figs.
5(a) and 5(b)), which is opposite to that of dielectric con- stant, also supports this correlation. Finally, it is worth men- tioning that, since such anisotropic magnetic-manipulation of ferroelectric and dielectric polarizations in the present het- erostructures can be well realized at room temperature and low magnetic field, it may be promising to be used for devel- oping high-performance electronic and magnetic nanodevi- ces such as multi-state logical switching and vector magnetic field detector, which are superior to contemporary multifer- roic compounds.
IV. CONCLUSION
In summary, multiferroic heterostructural films consisting of STO and CZFO layers, both of which have preferential crystallographic orientations, have been grown on a buffered STO layer deposited on the Pt/Ti/SiO2/Si wafers. Evident fer- roelectric polarization induced by defect dipoles was observed in the STO layer. The anisotropic magnetic-manipulation of ferroelectric polarization was realized in such kind of hetero- structural films, i.e., the ferroelectric polarization was remark- ably suppressed by in-plane magnetic field while greatly enhanced by out-of-plane magnetic field. In addition, the films exhibited large anisotropic magnetodielectric effect. Such
FIG. 4. Calculated䉭P(H)/PDDas a function of tDD/tSTOfor the STO/CZFO film.
FIG. 5. Dielectric constant (eH) vs voltage bias (Ubias) curves under various in-plane magnetic fields (Hk) (a) and out-of-plane magnetic fields (H?) (b).
The insets of (a) and (b) are the corresponding dielectric loss vs Ubiascurves measured at 0 and 5 kOe, respectively.
anisotropic magnetic-manipulation of ferroelectric or dielec- tric polarizations was mainly attributed to the distinctly differ- ent re-arrangement of oxygen vacancies in the STO layer under in-plane and out-of-plane magnetic fields.
ACKNOWLEDGMENTS
This work was supported by the National Key Projects for Basic Research of China (Grant No. 2010CB923401), the National Natural Science Foundation of China (Grant Nos.
11134005 and 50972055), and the PAPD project.
1H. Greve, E. Woltermann, H. J. Quenzer, B. Wagner, and E. Quandt, Appl. Phys. Lett.96, 182501 (2010).
2Y. S. Oh, S. Crane, H. Zheng, Y. H. Chu, R. Ramesh, and K. H. Kim, Appl. Phys. Lett.97, 052902 (2010).
3Z. W. Li, M. X. Zhou, W. F. Ding, H. Zhou, B. Chen, J. G. Wan, J. M.
Liu, and G. H. Wang,Appl. Phys. Lett.100, 262903 (2012).
4S. Seki, H. Murakawa, Y. Onose, and Y. Tokura,Phys. Rev. Lett. 103, 237601 (2009).
5T. Nakajima, S. Mitsuda, S. Kanetsuki, M. Yamano, S. Iwamoto, Y.
Yoshida, H. Mitamura, Y. Sawai, M. Tokunaga, K. Kindo, K. Prokes, and A. Podlesnyak,Phys. Rev. B81, 014422 (2010).
6Z. Viskadourakis, I. Radulov, A. P. Petrovi0, S. Mukherjee, B. M.
Andersen, G. Jelbert, N. S. Headings, S. M. Hayden, K. Kiefer, S.
Landsgesell, D. N. Argyriou, and C. Panagopoulos, Phys. Rev. B 85, 214502 (2012).
7M. I. Bichurin, V. M. Petrov, and G. Srinivasan,Phys. Rev. B68, 054402 (2003).
8J. H. Haeni, P. Irvin, W. Chang, R. Uecker, P. Reiche, Y. L. Li, S.
Choudhury, W. Tian, M. E. Hawley, B. Craigo, A. K. Tagantsev, X. Q.
Pan, S. K. Streiffer, L. Q. Chen, S. W. Kirchoefer, J. Levy, and D. G.
Schlom,Nature430, 758 (2004).
9A. Antons, J. B. Neaton, K. M. Rabe, and D. Vanderbilt,Phys. Rev. B71, 024102 (2005).
10Y. S. Kim, D. J. Kim, T. H. Kim, T. W. Noh, J. S. Choi, B. H. Park, and J.
G. Yoon,Appl. Phys. Lett.91, 042908 (2007).
11A. Vasudevarao, S. Denev, M. D. Biegalski, Y. L. Li, L. Q. Chen, S. T.
McKinstry, D. G. Schlom, and V. Gopalan,Appl. Phys. Lett.92, 192902 (2008).
12Y. S. Kim, J. S. Choi, J. Kim, S. J. Moon, B. H. Park, J. Yu, J. H. Kwon, M. Kim, J. S. Chung, T. W. Noh, and J. G. Yoon,Appl. Phys. Lett.97, 242907 (2010).
13T. Ohnishi, M. Lippmaa, T. Yamamoto, S. Meguro, and H. Koinuma, Appl. Phys. Lett.87, 241919 (2005).
14P. E. Janolin, N. A. Pertsev, D. Sichuga, and L. Bellaiche,Phys. Rev. B 85, 140401 (2012).
15G. Srinivasan, E. T. Rasmussen, and R. Hayes,Phys. Rev. B67, 014418 (2003).
16L. Fang, W. Dong, F. G. Zheng, and M. R. Shen, J. Appl. Phys. 112, 034114 (2012).
17X. Mathew, J. P. Enriquez, C. M. Garcıa, G. C. Puente, M. A. C. Jacome, J. A. T. Antonio, J. Hays, and A. Punnoose,J. Appl. Phys.100, 073907 (2006).
18F. He, B. O. Wells, and S. M. Shapiro, Phys. Rev. Lett. 94, 176101 (2005).
19H. Mestriæ, R. A. Eichel, T. Kloss, K. P. Dinse, S. Laubach, P. C.
Schmidt, K. A. Sch€onau, M. Knapp, and H. Ehrenberg,Phys. Rev. B71, 134109 (2005).
20H. Mestric´, R. A. Eichel, K. P. Dinse, A. Ozarowski, J. V. Tol, L. C.
Brunel, H. Kungl, M. J. Hoffmann, K. A. Sch€onau, M. Knapp, and H.
Fuess,Phys. Rev. B73, 184105 (2006).
21L. X. Zhang, E. Erdem, X. B. Ren, and R. A. Eichel,Appl. Phys. Lett.93, 202901 (2008).
22H. W. Jang, A. Kumar, S. Denev, M. D. Biegalski, P. Maksymovych, C.
W. Bark, C. T. Nelson, C. M. Folkman, S. H. Baek, N. Balke, C. M.
Brooks, D. A. Tenne, D. G. Schlom, L. Q. Chen, X. Q. Pan, S. V. Kalinin, V. Gopalan, and C. B. Eom,Phys. Rev. Lett.104, 197601 (2010).
23Y. Y. Guo, H. M. Liu, D. P. Yu, and J. M. Liu,Phys. Rev. B85, 104108 (2012).
24Z. G. Ban, S. P. Alpay, and J. V. Mantese, Phys. Rev. B 67, 184104 (2003).
25R. C. O’handley, Modern Magnetic Materials: Principles and Applications, edited by G. Franklin, J. Falcone, and R. Farkas (John Wiley
& Sons, Inc., New York, 2000), p. 179.
26D. Fritsch and C. Ederer,Phys. Rev. B86, 014406 (2012).
27V. M. Petrov and G. Srinivasan,Phys. Rev. B78, 184421 (2008).
28F. S. A. Tostado, A. H. Gomez, J. C. Woicik, R. Droopad, Z. Yu, D. G.
Schlom, P. Zschack, E. Karapetrova, P. Pianetta, and C. S. Hellberg,Phys.
Rev. B70, 201403 (2004).
29Y. X. Zheng, Q. Q. Cao, C. L. Zhang, H. C. Xuan, L. Y. Wang, D. H.
Wang, and Y. W. Du,J. Appl. Phys.110, 043908 (2011).
30N. A. Pertsev, S. Prokhorenko, and B. Dkhil, Phys. Rev. B85, 134111 (2012).
044102-6 Chenet al. J. Appl. Phys.115, 044102 (2014)
