Continuous developments in nanotechnology require new approaches to materials synthesis that can produce novel functional structures. Here, we show that nanoscale defects, such as nonstoichiometric nanoregions (NSNRs), can act as nano-building blocks for creating complex electrical polarization structures in the prototypical multiferroic BiFeO3. An array of charged NSNRs are produced in BiFeO3 thin films by tuning the substrate temperature during film growth. Atomic-scale scanning transmission electron microscopy imaging reveals exotic polarization rotation patterns around these NSNRs. These polarization patterns resemble hedgehog or vortex topologies and can cause local changes in lattice symmetries leading to mixed-phase structures resembling the morphotropic phase boundary with high piezoelectricity. Phase-field simulations indicate that the observed polarization configurations are mainly induced by charged states at the NSNRs. Engineering defects thus may provide a new route for developing ferroelectric- or multiferroic-based nanodevices.
All Science Journal Classification (ASJC) codes
- Physics and Astronomy(all)
Access to Document
Defect-Induced Hedgehog Polarization States in Multiferroics. / Li, Linze; Cheng, Xiaoxing; Jokisaari, Jacob R.; Gao, Peng; Britson, Jason; Adamo, Carolina; Heikes, Colin; Schlom, Darrell G.; Chen, Long Qing; Pan, Xiaoqing.In: Physical review letters, Vol. 120, No. 13, 137602, 27.03.2018.
Research output: Contribution to journal › Article › peer-review
TY - JOUR
T1 - Defect-Induced Hedgehog Polarization States in Multiferroics
AU - Li, Linze
AU - Cheng, Xiaoxing
AU - Jokisaari, Jacob R.
AU - Gao, Peng
AU - Britson, Jason
AU - Adamo, Carolina
AU - Heikes, Colin
AU - Schlom, Darrell G.
AU - Chen, Long Qing
AU - Pan, Xiaoqing
N1 - Funding Information: Li Linze 1,2 Cheng Xiaoxing 3 Jokisaari Jacob R. 2 Gao Peng 2 Britson Jason 3 Adamo Carolina 4 Heikes Colin 4 Schlom Darrell G. 4,5 Chen Long-Qing 3 Pan Xiaoqing 1,6 ,* Department of Chemical Engineering and Materials Science, 1 University of California–Irvine , Irvine, California 92697, USA Department of Materials Science and Engineering, 2 University of Michigan , Ann Arbor, Michigan 48109, USA Department of Materials Science and Engineering, 3 Penn State University , University Park , Pennsylvania 16802, USA Department of Materials Science and Engineering, 4 Cornell University , Ithaca, New York 14853, USA 5 Kavli Institute at Cornell for Nanoscale Science , Ithaca, New York 14853, USA Department of Physics and Astronomy, 6 University of California–Irvine , Irvine, California 92697, USA * Corresponding author. email@example.com 27 March 2018 30 March 2018 120 13 137602 27 October 2017 © 2018 American Physical Society 2018 American Physical Society Continuous developments in nanotechnology require new approaches to materials synthesis that can produce novel functional structures. Here, we show that nanoscale defects, such as nonstoichiometric nanoregions (NSNRs), can act as nano-building blocks for creating complex electrical polarization structures in the prototypical multiferroic BiFeO 3 . An array of charged NSNRs are produced in BiFeO 3 thin films by tuning the substrate temperature during film growth. Atomic-scale scanning transmission electron microscopy imaging reveals exotic polarization rotation patterns around these NSNRs. These polarization patterns resemble hedgehog or vortex topologies and can cause local changes in lattice symmetries leading to mixed-phase structures resembling the morphotropic phase boundary with high piezoelectricity. Phase-field simulations indicate that the observed polarization configurations are mainly induced by charged states at the NSNRs. Engineering defects thus may provide a new route for developing ferroelectric- or multiferroic-based nanodevices. U.S. Department of Energy 10.13039/100000015 DESC0014430 DEAC0205CH11231 National Science Foundation 10.13039/100000001 DMR-1420620 DMR-1506535 DMR0723032 OCI0821527 EEC-1160504 ECCS-1542081 University of California-Irvine’s Materials Research Institute U.S. Department of Energy 10.13039/100000015 FG0207ER46417 Quasi-1D ferroic states arising from the interplay of spin, charge, orbital, and/or lattice degrees of freedom in strongly correlated oxides usually exhibit complex polarization or spin textures and unique properties that hold great promise for the development of new device paradigms  . Examples include vortices, hedgehogs, and Skyrmions that have been discovered and extensively studied in magnetic materials [2–4] . These magnetic states can be efficiently manipulated by magnetic fields [5,6] or driven by a spin-transfer torque at ultralow current densities  and thus can be potentially applied for high-density magnetic memories or current-driven devices. At the same time, considerable interest has also been given to exploring new polarization states in ferroelectrics, because the use of ferroelectric states that are much smaller than their counterparts in magnets as functional elements in nanodevices is particularly attractive. For example, advances in the synthesis of ferroelectric thin-film heterostructures have enabled the creation of periodic patterns of polarization vortices through tuning the boundary conditions at film interfaces [8–10] . It is also theoretically predicted that Skyrmion-like polarization patterns can be stabilized in ferroelectric nanocomposites through the application of external electric fields  . Yet, another type of quasi-1D ferroelectric state known as the polarization hedgehog, in which all the polarization vectors diverge from the center of the structure, remains relatively unexplored. This is partially because such a hedgehog state is generally considered unstable, unless the inevitable accumulation of a large amount of polarization bound charge at the center of the structure can be compensated by some strong driving force. An intriguing theoretical work has predicted that the hedgehog states can be stabilized in BaTiO 3 nanowires under specific boundary conditions, where surface effects play a critical role  . But these hedgehog states have not yet been experimentally confirmed. On the other hand, the recent observation of a strong interaction between ferroelectric polarization and built-in fields induced by charged impurity defects embedded in the ferroelectric matrix suggests that these defects may be used as nano-building blocks that can provide a strong electrostatic-driven force for stabilization of complex domain structures [13,14] . For example, it has been shown that nanoscale planar charged defects in BiFeO 3 thin films can induce novel head-to-head mixed-phase polarization structures  . Nevertheless, these observed impurity defects were accidentally introduced in those films, and the possiblity to precisely control the formation of the defects and thus to generate the desired domain structures has not been studied. Here, we show that, by controlling the substrate temperature during the growth of a BiFeO 3 thin film, an array of nonstoichiometric nanoregions (NSNRs) can be deliberately introduced as charged defects into the host material matrix, which leads to the stabilization of novel hedgehog or antihedgehog nanodomains. These nanodomains present exotic patterns of polarization rotation that form mixed-phase structures and can be coupled with polarization vortices. Their novel properties could be useful for nanoelectronic and electromechanical applications. BiFeO 3 is a room-temperature multiferroic with a rhombohedral lattice structure, exhibiting coupled ferroelectric ( T C ∼ 1103 K ) and antiferromagnetic ( T N ∼ 650 K ) order  . In pseudocubic unit cells [Fig. 1(a) ], the oxygen octahedra and the central Fe cation are displaced from their respective positions at the face and body centers, giving rise to a large spontaneous polarization ( ∼ 100 μ C cm − 2 ) along the ⟨ 111 ⟩ PC directions [15–19] , where the PC subscript represents pseudocubic indices. Using aberration-corrected high-angle annular dark-field (HAADF) scanning transmission electron microscopy (STEM) imaging, the atomic-scale structure within a BiFeO 3 thin film can be directly visualized, and the polarization vector within each pseudocubic unit cell can be measured. For example, in the HAADF image shown in Fig. 1(b) , where the Bi columns appear as the brighter dots, the Fe columns show weaker contrast, and the oxygen atoms are not visible, a vector, D F B , can be defined as the atomic displacement in the image plane of the Fe cation from the center of the unit cell formed by its four Bi neighbors. This D F B vector is exactly opposite to the projection of the polarization vector in the image plane and can be directly measured by fitting the atomic columns as two-dimensional (2D) Gaussian peaks to locate their centers and recording the offset [8,20] . 1 10.1103/PhysRevLett.120.137602.f1 FIG. 1. Atomic structure and defects in a BiFeO 3 thin film. (a) Atomic model of the pseudocubic structure of BiFeO 3 . Polarization is shown by a green arrow. (b) HAADF STEM image of a BiFeO 3 thin film viewed in cross section along the [ 100 ] PC zone axis. The D F B vector, defined as the atomic displacement in the image plane of the Fe cation from the center of the unit cell formed by its four Bi neighbors, is opposite to the polarization vector ( P ). (c) HAADF STEM image of the BiFeO 3 ( BFO ) / TbScO 3 ( TSO ) interfacial region, where the interface is indicated by the yellow dashed line and an array of defects are observed at a uniform height of 3 nm above the interface. The growth of BiFeO 3 films is very sensitive to deposition parameters, and by most deposition techniques phase-pure BiFeO 3 can be obtained only within a narrow combination of growth parameters  . An alternate method enabling the growth of phase-pure BiFeO 3 for a broader range of growth parameters, referred to as the “growth window,” is what we use when synthesizing BiFeO 3 by molecular-beam epitaxy (MBE). In this process, BiFeO 3 is formed by flooding the growing film with excess bismuth and oxidant and relying on thermodynamics to desorb the excess of both of these constituents, resulting in a phase-pure BiFeO 3 film for conditions within the growth window. Our prior MBE work mapping out the adsorption-controlled growth conditions for BiFeO 3 indicates that a change of the substrate temperature during film growth, given a constant bismuth and oxygen overpressure at a fixed Bi:Fe flux ratio, can result in three different products: (I) BiFeO 3 + γ - Fe 2 O 3 (at a higher than optimal temperature), (II) BiFeO 3 (within the growth window, i.e., at an optimal temperature), and (III) BiFeO 3 + Bi 2 O 2.5 (at a lower than optimal temperature)  . The formation of the Fe-rich or Bi-rich defects can be directly monitored during film growth by in situ RHEED, since these defects can generate additional spots in the RHEED pattern  . In this work, we grew epitaxial ( 001 ) PC -oriented BiFeO 3 thin films on ( 110 ) O TbScO 3 single crystal substrates (where the O subscript represents orthorhombic indices) by MBE, monitoring the growth by RHEED. The films were grown at constant bismuth (Bi flux = 1.4 × 10 14 Bi / cm 2 s ) and oxygen ( O 2 + ∼ 10 % O 3 background pressure = 1 × 10 − 6 Torr ) with pressure over a fixed Bi:Fe flux ratio of 8 ∶ 1 . Film growth was initiated at a substrate temperature of ∼ 680 ° C (higher than the optimal temperature) until additional spots in the RHEED pattern, indexed to diffraction from (111)-oriented γ - Fe 2 O 3 , were observed (Supplemental Fig. S1  )  . Upon seeing these spots, the substrate temperature was decreased to 610 ° C and the RHEED pattern, which is very surface sensitive, went back to that of phase-pure BiFeO 3 . As a result, an array of Fe-rich defects were inserted into the BiFeO 3 matrix at a uniform height along the film thickness, several unit cells above the bottom interface of the film, evident in the HAADF STEM image shown in Fig. 1(c) . In this way, control of the temperature lends control of the placement of defects in the film. The atomic-scale structure of the defects is examined in Fig. 2(a) , where a magnified HAADF image of an interfacial region containing one typical semicircular loop defect in the BiFeO 3 matrix is shown. As there are no other elements introduced during film growth, the defect should also be composed of Bi, Fe, and O atoms, with the brighter dots in the HAADF image corresponding to the heavier Bi columns and the weaker dots corresponding to the lighter Fe columns. This defect is generally composed of two parts. The first part is a short vertical segment on the right (highlighted in brown) with a distorted pseudocubic perovskite lattice, where an apparent lattice expansion is observed along the horizontal direction. The second part is a long curved or inclined segment on the left (highlighted in yellow and green) with a structure and stoichiometry different from the BiFeO 3 matrix, forming NSNRs. Furthermore, these NSNRs in the defect are composed from two structural units—planar units (highlighted in green) and stepped units (highlighted in yellow). While the planar unit is oriented on the horizontal ( 001 ) PC plane of BiFeO 3 and adopts the structure of one pair of Bi atoms alternating with one pair of Fe atoms, the stepped unit has an inclined orientation and is formed from one pair of Bi atoms alternating with two pairs of Fe atoms. 2 10.1103/PhysRevLett.120.137602.f2 FIG. 2. Atomic structure associated with a semicircular loop defect in the BiFeO 3 thin film. (a) HAADF STEM image of the BiFeO 3 ( BFO ) / TbScO 3 ( TSO ) interfacial region containing one typical semicircular loop defect. The defect is composed of a vertical segment (highlighted in brown) with distorted pseudocubic perovskite lattices and nonstoichiometric planar (highlighted in green) and stepped (highlighted in yellow) structural units. (b) Magnified image of the nonstoichiometric planar unit in the defect [red rectangular highlighted region “B” in (a)] and a model of its corresponding atomic structure on the right. (c) Magnified image of the nonstoichiometric stepped unit in the defect [red rectangular highlighted region “C” in (a)] and a model of its corresponding atomic structure on the right. The same types of planar and stepped NSNRs have also been observed in previous studies of Nd- and Ti-doped antiferroelectric BiFeO 3 by MacLaren et al. [24,25] . Their detailed structural study suggests that the Fe atoms within the NSNRs are coordinated by six oxygens in the form of edge-sharing oxygen octahedra, resembling the structure of γ - Fe 2 O 3 and forming the planar and stepped structural units shown schematically in Figs. 2(b) and 2(c) , respectively. As a result, there is an excess of anions (more oxygen) at the NSNRs, which produces a net local negative charge density. Assuming Bi 3 + , Fe 3 + , and O 2 − at the defects, the charge density is calculated to be − 1.4 C / m 2 at the planar unit and − 1.1 C / m 2 at the stepped unit. Another feature associated with the NSNRs is an atomic shift of one-half of a pseudocubic unit cell of the BiFeO 3 lattice across the planar units in the [ 001 ] PC direction and a shift of one-half of a pseudocubic unit cell of the BiFeO 3 lattice across the stepped units in both the [ 010 ] PC and [ 001 ] PC directions, as evident in the Fourier-filtered images shown in Supplemental Fig. S3  . These shifts, however, do not induce obvious strain in the BiFeO 3 matrix, as shown by the quantitative analysis of the high-resolution HAADF images using geometric phase analysis (GPA) in Supplemental Fig. S4  . The shift in the lattice occurs mainly through atomic rearrangement at the boundary of the NSNRs and the shifted BiFeO 3 lattices are separated, accommodating the strain and forming an isolated region enclosed by the loop defect. The polarization distribution in the BiFeO 3 lattice adjacent to the defect was determined by mapping the − D FB vectors [Fig. 3(a) ]. Within the region enclosed by the semicircular defect (marked as region “1”), a hedgehog polarization state can be identified. While the polarization vectors of each unit cell in the middle of this region are attenuated and oriented along multiple directions, the unit cells that are in contact with the defect forming the periphery of the NSNRs (the unit cells highlighted by red rectangles) show significant lattice distortions with enhanced polarization pointing outward, towards the NSNRs. In contrast, the BiFeO 3 lattice outside of the defect presents a continuous rotation of polarization, with polarization vectors generally pointing inward towards the defect, except for the right-side region where the polarization vectors are pointing outward. This arrangement leads to the formation of six different nanodomains [marked “2”–“7” in Fig. 3(a) ] surrounding the defect. Some of these domains also experience a change in lattice symmetry, forming mixed-phase structures. Particularly, the “2,” “4,” and “6” domains possess normal rhombohedral-like ( R -like) structures with polarizations oriented along the diagonal directions. In contrast, the “3” and “5” domains adopt polarizations along [ 010 ] PC and [ 001 ] PC directions, respectively, resembling horizontal or vertical tetragonal-like ( T -like) structures  . Within region 7, significantly attenuated polarizations were observed. 3 10.1103/PhysRevLett.120.137602.f3 FIG. 3. Polarization map of the nanodomains surrounding the semicircular loop defect in the BiFeO 3 thin film. (a) The same HAADF STEM image as Fig. 2(a) overlaid with the polarization vectors ( − D F B , shown by yellow arrows). The white dashed lines represent the positions of domain walls. (b) Phase field simulation of the polarization distribution in the image plane of the domain structure stabilized by a charged defect with a configuration similar to the experimental observation. The polarization vectors are overlaid on a color map of the charge distribution, where the red, blue, and gray colors represent positive ( + 1.1 C / m 2 ), negative ( − 1.1 C / m 2 ), and zero charge densities, respectively. Since the polarization directions are generally pointing toward the NSNRs in the defect and these NSNRs are negatively charged, a plausible mechanism is that the polarization distortions of the BiFeO 3 lattices are modified by the built-in fields induced by the charge on the NSNRs. To gain further insight of the driving force for the formation of the nanodomains, we employ phase field simulations to calculate the stable polarization configuration in the presence of a charged defect. In the simulated BiFeO 3 thin film, we created a defect with a similar shape to what we observed in the experiment in Fig. 2 and then applied an additional charge at the defect, with no additional eigenstrain introduced by the defect from the lattice mismatch between the defect and the matrix, as the defect-induced strain has a minimal effect on the polarization (see Supplemental Material  for a discussion). Since the left curved or inclined part of the defect is mostly composed of the stepped units, for simplicity a uniform charge density of − 1.1 C / m 2 was applied there. Although the defect chemistry of the vertical segment on the right side of the defect cannot be directly determined by STEM imaging, its lattice expansion along the horizontal direction could indicate a local accumulation of oxygen vacancies and thus a positive charge density [27,28] . Simulations allow the effect of charge density to be studied, and a comparison of simulated results with different values of charge densities ranging from − 1.1 to + 5.5 C / m 2 at the right-side segment of the defect is shown in Supplemental Fig. S5  . The result with a charge density of + 1.1 C / m 2 [Fig. 3(b) ] is most consistent with the experimental observation, indicating that the existence of a moderate positive charge is the most probable situation. Note that in each case a random polarization distribution was set as the initial structure, and simulations repeated with different initial random polarization distributions were found to yield consistent final configurations. In the simulated polarization configuration around the defect [Fig. 3(b) ], attenuated polarization is observed within the semicircular region enclosed by the defect (region 1), with the exception of the very large polarization of a few unit cells at the region boundary. This could be a result of the exclusion of the experimentally observed distorted unit cells at the region boundary [red rectangular highlighted unit cells in Fig. 3(a) ] from the simulation. Nevertheless, the simulated results reproduced almost all of the features in the polarization configuration outside of the defect captured in the experiment, including the polarization reorientation around the defect with the formation of six different polarized regions (regions 2–7), the change of lattice symmetries into T -like structures in region 3 and 5, and the attenuated polarization in region 7. These results indicate that the charge on the NSNRs should be the major driving force inducing the observed nanodomain structures, in which the gradient energy from the polarization rotation is counterbalanced by the corresponding reduction in overall electrostatic energy. The hedgehog and antihedgehog domains are ubiquitous around the defects composed of NSNRs in the BiFeO 3 thin films, and they can also induce polarization vortices, as shown by three examples in Figs. 4(a)–4(c) . Similar to the previous case shown in Fig. 2 , these defects are mainly comprised of the same types of nonstoichiometric planar units and stepped units, with the same excess of oxygen and thus local net negative charge. Some additional variations were noted. For example, within the circular defect in Fig. 4(a) , an inclined segment is observed on the right side. It is composed of five consecutive pairs of Fe atoms with no Bi atoms in between them. Examination of the polarization shows that the hedgehog and antihedgehog domains around the defects are also accompanied by the formation of multiple different polarized regions and mixed-phase structures. In Fig. 4(a) , a tiny hedgehog polarization state with a diameter of 4 unit cells was observed within the nanoregion encircled by the defect. In both Figs. 4(a) and 4(b) , antihedgehog polarization states are stabilized at the shell regions surrounding the defects. On larger length scales outside of these defects, the overall continuous polarization rotation patterns lead to flux-closure vortex structures. In Fig. 4(c) , a pair of polarization semivortex and semiantivortex structures form at the left and right sides of this nearly linear defect as a result of the formation of a “head-to-head” polarization configuration above and below the defect. 4 10.1103/PhysRevLett.120.137602.f4 FIG. 4. Polarization maps of nanodomains induced by three different defects in the BiFeO 3 thin film. (a)–(c) HAADF STEM images of three different defects above the BiFeO 3 ( BFO ) / TbScO 3 ( TSO ) interface, where the polarization vectors ( − D F B , shown by yellow arrows) are overlaid on the BiFeO 3 lattice. The dashed red line marks the domain walls that penetrate to the top surface of the thin film. (d) Magnified map of the polarization vectors in the nanoregion enclosed by the loop defect in (a) and in the three different nanoregions highlighted by the blue rectangles in (b) and (c). In conclusion, we have demonstrated the stabilization of novel hedgehog and antihedgehog nanodomains in multiferroic BiFeO 3 thin films by making use of nanoscale charged defects. These nanodomains offer new opportunities to explore the complex interplay of spin, charge, orbital, and lattice degrees of freedom in systems with strong electron correlations. As the hedgehog states are confined within the nanosized structure defects, they can approach sizes down to several unit cells and have a much higher density than the bulk domains in the film. Our finding therefore opens the door for creating high-density nanodomains in relatively thick ferroelectric films, which can overcome the limit on domain size or density imposed by the scaling law that states the domain width is proportional to the square root of the sample’s thickness [29,30] . The polarization component perpendicular to the hedgehog plane (the image plane in our study) may be switchable. Therefore, further study of the electrical and magnetoelectric properties and dynamics under applied fields may assist future development of nanoelectronic devices such as high-density memories. On the other hand, it is well known that defect engineering through adding small amounts of donor dopant to a ceramic can create randomly oriented defect dipoles between donors and cation vacancies in the crystal structure, which promotes domain wall motion leading to enhanced piezoelectric properties  . Here, the antihedgehog polarization states stabilized by the nanoscale impurity defects possess mixed-phase structures that are reminiscent of the ferroelectric domains near a morphotropic phase boundary. These states may also exhibit enhanced properties such as giant piezoelectricity and thus are potentially attractive for electromechanical applications. As it is known that nonstoichiometric phases can commonly occur in functional oxides due to different mechanisms, such as cation doping [24,25] , stoichiometric fluctuation  , and temperature variation during film growth  , our observation of topological states induced by NSNRs in BiFeO 3 should stimulate efforts to design functional patterns of ferroelectric or multiferroic nanodomains or nanopolymorphs with a focus on controlling the intrinsic defect structures and motivate a search for a similar mechanism in other functional oxide systems. While conventional chemical engineering methods (composition tuning) and interface engineering methods (boundary condition tuning) produce only structure changes uniformly distributed within the whole material, defect engineering can generate functional structures at local nanoregions without the disruption of the bulk structural patterns, which is desirable as the dimensions of individual elements in nanodevices continue to shrink. Further developments may allow the NSNRs to be introduced in a fine controlled fashion, enabling their configurations to be adjusted during material synthesis, or allow the charge states of the NSNRs to be modified by external charge injection methods, opening the door to the control of nanodomain structures and the customized creation of topological states in complex materials through defect engineering. The experimental work was mainly supported by the Department of Energy (DOE) under Grant No. DESC0014430 (L. Z. L., J. R. J., P. G., and X. Q. P.) and partially by the National Science Foundation (NSF) under Grants No. DMR-1420620 and No. DMR-1506535. The TEM used for this work was funded by the National Science Foundation under Grant No. DMR0723032, and by the University of California-Irvine’s Materials Research Institute (IMRI). The theoretical component of this work at the Pennsylvania State University was supported by the U.S. Department of Energy, Office of Basic Energy Sciences, Division of Materials Sciences and Engineering under Award No. FG0207ER46417 (X. X. C. and L. Q. C.). Calculations at the Pennsylvania State University were performed on the Cyberstar Linux Cluster funded by the National Science Foundation through Grant No. OCI0821527. The work at Cornell University was supported by the National Science Foundation (Nanosystems Engineering Research Center for Translational Applications of Nanoscale Multiferroic Systems) under Grant No. EEC-1160504 (C. A., C. H., and D. G. S.). The authors also acknowledge the National Center for Electron Microscopy at Lawrence Berkeley National Laboratory for their support under DOE Grant No. DEAC0205CH11231 for user facilities. This work was performed in part at the Cornell Nanoscale Facility, a member of the National Nanotechnology Coordinated Infrastructure (NNCI), which is supported by the National Science Foundation (Grant No. ECCS-1542081). The authors also acknowledge Ramamoorthy Ramesh at University of California–Berkeley for discussing the results and providing insightful comments.  1 J. Seidel , Topological Structures in Ferroic Materials Domain Walls , Vortices and Skyrmions preface ( Springer , New York, 2016 ).  2 A. Wachowiak , Science 298 , 577 ( 2002 ). SCIEAS 0036-8075 10.1126/science.1075302  3 S. Y. Xu , Nat. Phys. 8 , 616 ( 2012 ). NPAHAX 1745-2473 10.1038/nphys2351  4 X. Z. Yu , Y. Onose , N. Kanazawa , J. H. Park , J. H. Han , Y. Matsui , N. Nagaosa , and Y. Tokura , Nature (London) 465 , 901 ( 2010 ). NATUAS 0028-0836 10.1038/nature09124  5 S. Heinze , K. von Bergmann , M. Menzel , J. Brede , A. Kubetzka , R. Wiesendanger , G. Bihlmayer , and S. Blügel , Nat. Phys. 7 , 713 ( 2011 ). NPAHAX 1745-2473 10.1038/nphys2045  6 T. Schulz , R. Ritz , A. Bauer , M. Halder , M. Wagner , C. Franz , C. Pfleiderer , K. Everschor , M. Garst , and A. Rosch , Nat. Phys. 8 , 301 ( 2012 ). NPAHAX 1745-2473 10.1038/nphys2231  7 F. Jonietz , Science 330 , 1648 ( 2010 ). SCIEAS 0036-8075 10.1126/science.1195709  8 C. T. Nelson , Nano Lett. 11 , 828 ( 2011 ). NALEFD 1530-6984 10.1021/nl1041808  9 Y. L. Tang , Science 348 , 547 ( 2015 ). SCIEAS 0036-8075 10.1126/science.1259869  10 A. K. Yadav , Nature (London) 530 , 198 ( 2016 ). NATUAS 0028-0836 10.1038/nature16463  11 Y. Nahas , S. Prokhorenko , L. Louis , Z. Gui , I. Kornev , and L. Bellaiche , Nat. Commun. 6 , 8542 ( 2015 ). NCAOBW 2041-1723 10.1038/ncomms9542  12 J. W. Hong , G. Catalan , D. N. Fang , E. Artacho , and J. F. Scott , Phys. Rev. B 81 , 172101 ( 2010 ). PRBMDO 1098-0121 10.1103/PhysRevB.81.172101  13 L. Z. Li , Y. Zhang , L. Xie , J. R. Jokisaari , C. Beekman , J.-C. Yang , Y.-H. Chu , H. M. Christen , and X. Pan , Nano Lett. 17 , 3556 ( 2017 ). NALEFD 1530-6984 10.1021/acs.nanolett.7b00696  14 L. Xie , Adv. Mater. 29 , 1701475 ( 2017 ). ADVMEW 0935-9648 10.1002/adma.201701475  15 F. Zavaliche , S. Y. Yang , T. Zhao , Y. H. Chu , M. P. Cruz , C. B. Eom , and R. Ramesh , Phase Transitions 79 , 991 ( 2006 ). PHTRDP 0141-1594 10.1080/01411590601067144  16 Y. H. Chu , L. W. Martin , M. B. Holcomb , and R. Ramesh , Mater. Today 10 , 16 ( 2007 ). MATOBY 1369-7021 10.1016/S1369-7021(07)70241-9  17 G. Catalan and J. F. Scott , Adv. Mater. 21 , 2463 ( 2009 ). ADVMEW 0935-9648 10.1002/adma.200802849  18 R. R. Das , Appl. Phys. Lett. 88 , 242904 ( 2006 ). APPLAB 0003-6951 10.1063/1.2213347  19 F. Kubel and H. Schmid , Acta Crystallogr. Sect. B 46 , 698 ( 1990 ). ASBSDK 0108-7681 10.1107/S0108768190006887  20 L. Z. Li , P. Gao , C. T. Nelson , J. R. Jokisaari , Y. Zhang , S.-J. Kim , A. Melville , C. Adamo , D. G. Schlom , and X. Pan , Nano Lett. 13 , 5218 ( 2013 ). NALEFD 1530-6984 10.1021/nl402651r  21 H. Deniz , A. Bhatnagar , E. Pippel , R. Hillebrand , A. Hähnel , M. Alexe , and D. Hesse , J. Mater. Sci. 49 , 6952 ( 2014 ). JMTSAS 0022-2461 10.1007/s10853-014-8400-3  22 J. F. Ihlefeld , Appl. Phys. Lett. 92 , 142908 ( 2008 ). APPLAB 0003-6951 10.1063/1.2901160  23 See Supplemental Material at http://link.aps.org/supplemental/10.1103/PhysRevLett.120.137602 for transmission electron microscopy and phase-field simulation methods, and discussion on strain effects of defects.  24 I. MacLaren , Appl. Mater. 1 , 021102 ( 2013 ). 10.1063/1.4818002  25 I. MacLaren , Appl Mater. 2 , 066106 ( 2014 ). 10.1063/1.4884684  26 R. J. Zeches , Science 326 , 977 ( 2009 ). SCIEAS 0036-8075 10.1126/science.1177046  27 Y.-M. Kim , A. Morozovska , E. Eliseev , M. P. Oxley , R. Mishra , S. M. Selbach , T. Grande , S. T. Pantelides , S. V. Kalinin , and A. Y. Borisevich , Nat. Mater. 13 , 1019 ( 2014 ). NMAACR 1476-1122 10.1038/nmat4058  28 Y.-M. Kim , J. He , M. D. Biegalski , H. Ambaye , V. Lauter , H. M. Christen , S. T. Pantelides , S. J. Pennycook , S. V. Kalinin , and A. Y. Borisevich , Nat. Mater. 11 , 888 ( 2012 ). NMAACR 1476-1122 10.1038/nmat3393  29 C. W. Huang , L. Chen , J. Wang , Q. He , S. Y. Yang , Y. H. Chu , and R. Ramesh , Phys. Rev. B 80 , 140101 ( 2009 ). PRBMDO 1098-0121 10.1103/PhysRevB.80.140101  30 Y. B. Chen , M. B. Katz , X. Q. Pan , R. R. Das , D. M. Kim , S. H. Baek , and C. B. Eom , Appl. Phys. Lett. 90 , 072907 ( 2007 ). APPLAB 0003-6951 10.1063/1.2472092  31 C. H. Hong , J. Mater. 2 , 1 ( 2016 ). JMLSAM 0022-2453 10.1016/j.jmat.2015.12.002
PY - 2018/3/27
Y1 - 2018/3/27
N2 - Continuous developments in nanotechnology require new approaches to materials synthesis that can produce novel functional structures. Here, we show that nanoscale defects, such as nonstoichiometric nanoregions (NSNRs), can act as nano-building blocks for creating complex electrical polarization structures in the prototypical multiferroic BiFeO3. An array of charged NSNRs are produced in BiFeO3 thin films by tuning the substrate temperature during film growth. Atomic-scale scanning transmission electron microscopy imaging reveals exotic polarization rotation patterns around these NSNRs. These polarization patterns resemble hedgehog or vortex topologies and can cause local changes in lattice symmetries leading to mixed-phase structures resembling the morphotropic phase boundary with high piezoelectricity. Phase-field simulations indicate that the observed polarization configurations are mainly induced by charged states at the NSNRs. Engineering defects thus may provide a new route for developing ferroelectric- or multiferroic-based nanodevices.
AB - Continuous developments in nanotechnology require new approaches to materials synthesis that can produce novel functional structures. Here, we show that nanoscale defects, such as nonstoichiometric nanoregions (NSNRs), can act as nano-building blocks for creating complex electrical polarization structures in the prototypical multiferroic BiFeO3. An array of charged NSNRs are produced in BiFeO3 thin films by tuning the substrate temperature during film growth. Atomic-scale scanning transmission electron microscopy imaging reveals exotic polarization rotation patterns around these NSNRs. These polarization patterns resemble hedgehog or vortex topologies and can cause local changes in lattice symmetries leading to mixed-phase structures resembling the morphotropic phase boundary with high piezoelectricity. Phase-field simulations indicate that the observed polarization configurations are mainly induced by charged states at the NSNRs. Engineering defects thus may provide a new route for developing ferroelectric- or multiferroic-based nanodevices.
UR - http://www.scopus.com/inward/record.url?scp=85044779048&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85044779048&partnerID=8YFLogxK
U2 - 10.1103/PhysRevLett.120.137602
DO - 10.1103/PhysRevLett.120.137602
M3 - Article
C2 - 29694202
AN - SCOPUS:85044779048
VL - 120
JO - Physical Review Letters
JF - Physical Review Letters
SN - 0031-9007
IS - 13
M1 - 137602