WO2009043639A1

WO2009043639A1 - Material comprising finely layered heterostructures of oxide materials

Info

Publication number: WO2009043639A1
Application number: PCT/EP2008/061109
Authority: WO
Inventors: Philippe Ghosez; Eric Bousquet; Patrick Hermet; Jean-Marc Triscone; Matthew Dawber; Céline LICHTENSTEIGER; Nicolas Stucki
Original assignee: Universite De Liege; Universite De Geneve
Priority date: 2007-09-29
Filing date: 2008-08-26
Publication date: 2009-04-09

Abstract

A material comprises finely layered heterostructures of at least two oxide materials, wherein the layered structure allows coupling of polar and non-polar vibrational modes or structural distortions at interfaces enhancing properties of the oxide materials, especially polarization, transition temperature and strain-polarization coupling. The material comprises layered perovskite-like ABO3 structures, wherein A and B are cations of different sizes. Especially there are layers of ABO3 and A'BO3, wherein A is Pb, A' is Sr and B is Ti.

Description

Material comprising finely layered heterostructures of oxide materials

Technical Field of the invention

The invention relates to epitaxial (monocrystalline) oxide su- perlattices (layers of thin films) .

Technical background of the invention

Ferroelectric thin films and superlattices are currently the subject of intensive research, because of the interest they raise for technological applications and also because their modified ferroelectric properties are of fundamental scientific interest. Ferroelectric superlattices allow the tuning of the ferroelectric properties while maintaining perfect crystal structure and avoiding the depolarization issues involved with ultra-thin ferroelectric thin films. This tuning is practically achieved by playing both with strain, to enhance the polarization, and with composition, to interpolate between the properties of the combined compounds.

Summary of the invention

It is one object of the present invention to disclose growth of superlattices with very short-periods yielding materials exhibiting unexpected new properties, which are particularly attractive for technological applications. Considering ferroelec- tric/paraelectric PbTiθ₃/SrTiθ₃ multilayers, it is shown from first-principles that the ground-state of the system is not purely ferroelectric but also primarily involves antiferrodis- tortive rotations of the oxygen atoms in a way compatible with improper ferroelectricity . It is then disclosed in this specification that there are experiments, showing that in contrast to pure PbTiθ₃ and SrTiθ₃ compounds the system according to the invention indeed behaves like a prototypical improper ferroelectric and exhibits a very large dielectric constant (ε_r -600) which, at the same time, is fairly temperature independent, both features being of direct practical interest for technological applications .

This behaviour is distinct from that of regular ferroelectrics for which the dielectric constant strongly evolves around the phase transition temperature and additionally to that of previously known improper ferroelectrics that exhibit only small dielectric constants.

It is possible to obtain these new and exceptional properties by reducing the thickness of each layer to a few atomic layers. It is also possible to tune these properties according to features mentioned in the annexed claims. Such properties are for example very high dielectric constants (at least a factor of 10 larger than the best dielectrics available in the prior art) . With such materials, it is possible to contemplate new gate materials for transistors and hence microprocessors.

Such finely layered heterostructures of oxide materials (including, but not limited to, epitaxial, polycrystalline, nanocrys- talline or amorphous structures) , produced by thin film deposition techniques (including, but not limited to, sputtering, pulsed laser deposition or molecular beam epitaxy) allow coupling of polar and non-polar vibrational modes or structural distortions at interfaces to induce a form of ferroelectricity or pyroelectricity in the structure or enhance the polarization, transition temperature or strain-polarization coupling in a material that is already ferroelectric. The inventors have demonstrated that a special artificial construction based on ferroelectric PbTiθ₃ and paraelectric SrTiθ₃ allows an improper ferroelectricity to be induced, which stands out as unique from other examples in bulk materials; not only do the functional forms of the temperature evolution bear much closer resemblance to the textbook definitions of improper ferroelectricity than any other material, but it is the first example of an improper ferroelectric exhibiting such a large polarization and dielectric constant. Further, these properties are a direct product of the interfaces in the artificially layered structure. This finding thus opens the door to further possibilities based on interface engineering, with potentially even more exciting properties, for example, in the field of magneto- electric and multiferroic compounds.

The invention can be used to tune (in which sense tuning can refer to either augmentation or reduction of) the ferroelectric polarization of a material, whether the parent materials are intrinsically ferroelectric or not.

The invention allows either inducing or enhancing of ferroelectricity in a material with magnetic properties, especially to couple pre-existing ferroelectric and magnetic properties in the parent materials.

The material according to the inventions comprises finely layered heterostructures of at least two oxide materials, wherein the layered structure allows coupling of vibrational modes or structural distortions at interfaces enhancing properties of the oxide materials. Such properties are from the group comprising polarization, ferroelectric transition temperature, strain- polarization coupling, transition temperature associated with magnetic ordering, temperature dependence of the dielectric con- stant, magnitude of the dielectric constant, voltage tunability of the dielectric constant, piezoelectric properties, piezomag- netic properties, magnetization, and magnetic susceptibility.

Short description of the drawings

The invention will now be described in connection with exemplary embodiments with reference to the enclosed drawings: Fig. 1 Schematic view of the prototype P4//n/n/n unit-cell of the 1/1 PbTiθ₃/SrTiθ₃ superlattice and atomic motions associated to different energy lowering distortions : (a) FE_Z (T₃ mode) giving rise to a polarization P_z, (b) AF- D_zo (M₄ mode) with oxygen rotation angle Φ_zo, and, (c) AFD₂₁ (M₂ mode) with oxygen rotation angle Φ_Z1.

Fig. 2 (a) Relative stability, in term of the epitaxial strain, of different phases compatible with the condensation of individual or coupled instabilities in a 1/1 PbTiO₃/SrTiO₃ superlattice. (b) Amplitude of the po-

_2 larization P_z (μC cm ) , rotation angles Φ_Z1 and Φ_zo (degrees) and energy gain ΔE with respect to the prototype P/4/nm/n phase (meV/supercell) in different low symmetry phases of a 1/1 PbTiO₃ZSrTiO₃ superlattice epi- taxially grown on SrTiO₃ (c) Evolution of the spontaneous polarization with the epitaxial strain in the FE_Z (P4iτ_m) and (AFD/FE) _z (Ε^>4bm) phases for a 1/1 PbTiO₃/SrTiO₃ superlattice epitaxially grown on SrTiO₃. (d) Evolution of the energy with the polarization in the FE_Z phase (Φ_Z1 = Φ_zo = 0) and in the (AFD/FE) _z phase (Φ_Z1 = 2.2° and Φ_zo = 4.0°) for a 1/1 PbTiO₃/SrTiO₃ superlattice epitaxially grown on SrTiO₃.

Fig. 3 (a) Ferroelectric polarization of a 100 nm thick 9/3 PbTiO₃/SrTiO₃ superlattice as a function of tempera- 1 / 9 ture. The line is a fit, P_z = 1.39 (893-T) (extrapolating to P_z = 0 at T_c measured from Fig. 3c) (b) Ferroelectric polarization of a 100 nm thick 2/3 PbTiθ₃/SrTiθ₃ superlattice as a function of temperature. The line is a fit, P_z=0.053 (500-T) , the inset shows the same measurement for a lOOnm 1/1 PbTiO₃/SrTiO₃ superlattice, (c) Tetragonality of a 100 nm thick 9/3 PbTiθ₃/SrTiθ₃ superlattice (measured using x-ray diffraction) as a function of temperature. The line is a fit, c/a=l .0215+2.2IxIO^"5 (893-T) , (d) Tetragonality of a lOOnm thick 2/3 PbTiθ₃/SrTiθ₃ superlattice as a function of temperature. The line is a fit, c/a=l.01437+2.39xlO^"8 (500-T)² (e) Dielectric constant of a 100 nm thick 9/3 PbTiO₃/SrTiO₃ superlattice as a function of temperature. The line is a fit ε=320, 000/ (893-T) , (f) Dielectric constant of a 100 nm thick 2/3 PbTiO₃/SrTiO₃ superlattice as a function of temperature. The inset is a close-up of the same data highlighting the step in the dielectric constant at the transition temperature.

Detailed description of exemplary embodiments

The alternation of ultrathin layers in a ferroelectric- dielectric superlattice allows the imposition of a coherent strain state on the ferroelectric while the electrostatic coupling between ferroelectric and dielectric layers is able to polarize the latter. When the dielectric is sufficiently polariz- able, this yields a uniformly and highly polarized ground-state, the polarization of which is predictable from simple electrostatic considerations, starting from a review as in Dawber, M., Lichtensteiger, C, Cantoni, M., Veithen, M. , Ghosez, Ph. , Johnston, K., Rabe, K. M. & Triscone, J. -M. in "Unusual behavior of ferroelectric polarization in PbTi(WSrTiOs superlattices", published in Phys. Rev. Lett. 95, 177601 (2005) . In PbTiO₃/SrTiO₃ superlattices made of the repetition of np unit cells of PbTiO₃ and ns unit cells of SrTiO₃, hereafter denoted np/ns, control of the PbTiO₃ volume fraction (np/ (np+ns) ) allows tuning of the polarization and phase transition temperature in a predictable way over a wide range of compositions. However, in the limit of ul- trathin PbTiO₃ layers, an unexpected recovery of ferroelectricity was observed that cannot be understood considering only the ferroelectric degree of freedom and basic electrostatic arguments. The invention is based on the insight obtained from a detailed study of the properties of short-period PbTiO₃/SrTiO₃ superlattices and shows that the recovery of ferroelectricity is the signature of a new type of coupling between different structural instabilities leading to unique properties.

The reference cubic structure of ABO₃ perovskite compounds can be unstable to different kinds of energy-lowering distortions. In PbTiO₃, the cubic phase is unstable not only to a polar zone- center distortion responsible for the ferroelectric (FE) ground- state but also to a zone-boundary distortion involving tilts of the oxygen octahedra. At the bulk level, the latter is suppressed when the polar distortion freezes in but both can coexist at surfaces as shown by Bungaro, C. and Rabe, K. M. in "Coexistence of antiferrodistortive and ferroelectric distortions at the PbTiO₃ (001) surface", published in Phys. Rev. B 71, 035420 (2005) . In bulk SrTiO₃, oxygen rotation is conversely responsible for a non-polar antiferrodistorted (AFD) ground-state and ferroelectricity is suppressed by quantum fluctuations but, again, both distortions can coexist under pressure or appropriate epitaxial strains, yielding complex phase diagrams. From these considerations, it is expected here, that FE and AFD dis- tortions will strongly compete in PbTiθ₃/SrTiθ₃ superlattices .

To determine theoretically the ground-state structure and properties of such superlattices, we performed density functional theory calculations within the local density approximation. We used norm conserving pseudopotentials and a plane-wave basis set, as implemented in the ABINIT package, a software project published as "First-principles computation of material properties: the ABINIT software project" in Comput . Mat. Sci . 25, 478- 492 (2002) . Similar to Dieguez, 0., Rabe, K. M. and Vanderbilt, D. in "First-principles study of epitaxial strain in perovskites" published in Phys . Rev. B 72, 144101 (2005), the inventors adopted a supercell approach that allows for (i) the implicit treatment of the mechanical constraint imposed by the substrate by fixing the in-plane lattice constant of the super- lattice and (ii) imposes short-circuit electrical boundary conditions through the use of periodic boundary conditions. However, additionally the size of the supercell in-plane was doubled in order to allow for AFD oxygen motions as based on Lin, C-H., Huang, C-M. and Guo, G. Y. in "Systematic ab initio study of the phase diagram of epitaxially strained SrTiOs" published in J. Appl. Phys. 100, 084104 (2006) . Starting then from the prototype paraelectric P4/mmm structure corresponding to the highest achievable symmetry, instabilities from the inspection of the phonon dispersion curves were identified and, according to these, the symmetry was lowered and new structural relaxations were performed. This was repeated until no instabilities were present.

Fig. 1 shows a schematic view of the prototype P4//nm/n unit-cell of the 1/1 PbTiθ₃/SrTiθ₃ superlattice and atomic motions associated to different energy lowering distortions : Pb cations have received the reference numeral 11, Sr cations have received the reference numeral 12, Ti cations have received the reference numeral 13 , 0 anions have received the reference numeral 14. Fig. 1 (a) shows FE_Z (I^ mode) giving rise to a polarization P_z according to arrow 21. Fig. 1 (b) shows AFD_ZO (M₄ mode) with oxygen rotation angle Φ_zo in the plane 15 according to arrow 22 and, Fig. 1 (c) shows AFD₂₁ (M₂ mode) with oxygen rotation angle Φ_Z1 in the plane 15 according to arrow 23.

For the simulation of a PbTiθ₃/SrTiθ₃ 1/1 superlattice grown on a SrTiθ₃ (001) substrate, inspection of the phonon dispersion curves of the P4/mmm phase reveals the existence of zone-center FE instable modes with polarization P out-of-plane (T₃ mode, Fig. Ia) or in-plane (T₅ mode) and hereafter called respectively FE_Z and FE_xγ. Moreover, significantly larger AFD instabilities are also present at the M (1/2,1/2,0) point, which correspond to different kinds of tilts of the oxygen octahedra. This includes (i) tilts around the (001) axis with successive octahedra along (001) moving out-of-phase (M₄ mode, Fig. Ib) or in-phase (M₂ mode, Fig. Ic) hereafter called AFD_ZO and AFD₂₁, and (ii) tilts around an axis perpendicular to (001) (M₅ mode) hereafter called AFD_xy. Fig. 1 shows the superlattice unit cell of a material composed of repetitions of one unit cell thick layers of SrTiC>₃ and one unit cell thick layers of PbTiθ₃. Materials according to the invention may have layer thicknesses ranging from 1 unit cell to around 20 unit cells.

The material comprises preferably single layers based on ABO₃ perovskites, though other oxide crystal structures are also envisaged. Variation in composition from one layer to another can be in either one or both of the A or B ions and within a layer the A or B site can be occupied by a mixture of ions (ie. a solid solutuion) . A can be chosen from the group Pb, Sr, Ba, La, Ca, K, Li, Y, Dy, Gd, Eu, Yb, Ce, Pr, Nd and B can be chosen from the group Ti, Zr, Hf, Mn, Mg, Fe, Ta, Ru, Sc, Al . Within a preferred embodiment A can be Pb in one layer and Sr in the other, while B can be Ti in both layers.

Structural optimizations under symmetry constraints compatible with the condensation of individual or coupled unstable modes identified above have then been performed for various epitaxial strains, yielding the energy diagram of Fig. 2. Inspection of the curves shows that the ground-state is strongly dependent on the epitaxial constraint and always involves the coupling of FE and AFD distortions.

At the epitaxial strain corresponding to a SrTiθ₃ substrate (SrTiC>₃ misfit strain equal to zero), it is shown in Fig. 2a that allowing a FE_Z distortion alone would produce a negligible gain of energy, while a much more significant energy gain can be achieved when allowing for a AFD_ZO distortion. Surprisingly the condensation of any of the individual distortions does not suppress the other instabilities and proper energy minimization yields an unexpected (AFD/FE)_Z ground-state combining FE_Z, AFD₂₁ and AFD_ZO distortions. Amplitudes of the related oxygen rotation angles Φ_Z1 and Φ_zo and of the polarization P_z (at first order proportional to the amplitude of the FE_Z distortion through the mode effective charge) are reported in Fig. 2b.

The following table groups the different curves of Fig. 2a with the reference numerals.

As further revealed in Fig. 2c, over a wide range of epitaxial strains, the AFD/FE coupling significantly enhances the spontaneous polarization compared to that of a purely FE_Z state. Curve 42 relates to the FE_Z distortion and curve 47 relates to the (AFD/FD) _z distortion. Moreover, it can even produce a sizeable polarization in cases where it would not be observed otherwise. Calculations at other thicknesses (3/3,5/3,7/3) also reveal the existence of a similar AFD/FE coupling. However, although oxygen tilts of similar amplitude are present at all thicknesses in the PbTiθ₃ interface unit cell, these tilts rapidly disappear in the interior of the layer, revealing that the AFD/FE coupling is essentially an interfacial effect. All this strongly suggests that the unexpected recovery of ferroelectricity in PbTiθ₃/SrTiθ₃ multilayers experimentally reported by Dawber, M., Lichtensteiger, C, Cantoni, M., Veithen, M. , Ghosez, Ph. , Johnston, K., Rabe, K. M. & Triscone, J. -M. "Unusual behavior of ferroelectric polarization in PbTiO₃/SrTiO₃ superlattices . Phys. Rev. Lett. 95, 177601 (2005), in the limit of ultrathin PbTiO₃ layers originates from the coupling of FE and AFD distortions.

Since the ground-state of short-period superlattices is not purely ferroelectric, we can expect a phase transition mechanism distinct from that of ordinary ferroelectrics . In the 1/1 super- lattice on SrTiθ₃ for instance, the ground-state results from the condensation of three distortions FE_Z, AFD₂₁ and AFD_ZO related to independent order parameters P_z, Φ_Z1 and Φ_zo. Since the FE_Z instability is virtually suppressed, P_z will certainly not appear as the primary order parameter. Different phase transition mechanisms (such as improper or triggered ferroelectricity) can be observed when independent order parameters are involved and the exact nature of the transition depends on the way order parameters are allowed to couple into a Landau-type free energy expansion. In the present case, it is worth realizing from the inspection of the irreducible representations of FE_Z (T₃ ) , AFD₂₁

(M₂ ⁺) and AFDzo (M₄ ^") in the P4/mmm phase that the product Φ_Z1. Φ_zo transforms as P_z so that, in contrast to ordinary ferroelec- trics that only allow even-power terms in P_z, the lowest-order Pz invariant term in the Landau expansion is linear and has the form - g Φ_Z1.Φ_ZO.P_Z. Calculation of the energy for the 1/1 super- lattice as a function of P_z around P_z= 0 with fixed Φ_Z1 and Φ_zo

(shown in Fig. 2d; reference numeral 57 showing the (AFD/FE) _z distortion) not only demonstrates the existence of such a linear term in P_z but also highlights the sizable value of the coupling parameter g. Levanyuk A. P. & Sannikov D. G. have disclosed within "Improper Ferroelectrics" in Usp . Piz. Nauk 112, 561-589

(1974), that this is compatible with an improper ferroelectric behaviour with primary order parameters Φ_Z1 and Φ_zo- We note that Φ_Z1 and Φ_zo here are not different components of a two- dimensional order parameter but instead belong to different one- dimensional irreducible representations and can a priori give rise to two independent phase transitions meaning that an improper ferroelectric behaviour of the kind discussed by Levanyuk will be observed only if the transition temperatures are very close to each others, which is demonstrated by experiment.

Improper ferroelectrics show critical behaviour different to that of ordinary ferroelectrics and the fingerprints of improper ferroelectricity can be tracked in the distinct temperature dependencies of the polarization and dielectric constant. Therefore we experimentally compared the behaviour of PbTiθ₃/SrTiθ₃ superlattices corresponding to the two regimes identified in the article prepared by Dawber, M., Lichtensteiger, C, Cantoni, M., Veithen, M. , Ghosez, Ph. , Johnston, K., Rabe, K. M. & Tris- cone, J. -M. "Unusual behavior of ferroelectric polarization in PbTiO₃/SrTiO₃ superlattices. Phys . Rev. Lett. 95, 177601 (2005) . The first regime corresponds to a range of compositions and thicknesses producing an ordinary behaviour (illustrated by a 9/3 sample) and the second regime corresponding to superlattices exhibiting an anomalous recovery of ferroelectricity (such as a 2/3 or 1/3 sample) .

These samples were prepared by off-axis magnetron sputtering on (001) SrTiO₃ single crystal substrates with Tiθ₂ termination. In addition to the superlattices, high quality epitaxial top and bottom SrRuO₃ electrodes (important for the accurate determination of electrical properties) were deposited in-situ. X-Ray diffraction confirmed that the entire structure was epitaxially constrained in plane to the SrTiO₃ lattice constant, and Atomic Force Microscopy analyses showed that the unit cell steps originally present on the substrate were carried through to the surface of the structures. More details on the sputtering process and growth conditions can be found by someone skilled in the art in Dawber, M., Lichtensteiger, C, Cantoni, M., Veithen, M. , Ghosez, Ph. , Johnston, K., Rabe, K. M. & Triscone, J. -M. "Unusual behavior of ferroelectric polarization in PbTiO₃/SrTiO₃ su- perlattices. Phys. Rev. Lett. 95, 177601 (2005) .

In PbTiθ₃/SrTiθ₃ superlattices, it was shown that epitaxial strain makes the ferroelectric phase transition second order. If the system behaves like an ordinary mean-field ferroelectric, P_z is therefore expected from Landau theory to evolve like (T_c-T) 1/2 and such a behaviour is indeed observed for the 9/3 sample as shown in Fig. 3a. In improper ferroelectrics however, the order parameter is no longer the polarization and it is instead Φ_zi and Φ_zo that, if they are mean-field, are expected to behave like

(Tc-T) 1/2. Due to the linear P_z term in the Landau expansion, this yields P_z ^ Φ_zi . Φ_zo --"■^■ (T₀-T) . This is exactly what is observed for the 2/3 sample and the 1/1 sample (shown as an inset) in Fig. 3b. Moreover, the experimentally measured polarization for the 1/1 sample extrapolates to 23 μC cm at zero temperature, in

_2 close agreement with the theoretical prediction of 26 μC cm . The distinct evolution of the polarization in the two kinds of samples is further confirmed in the X-ray measurement of the tetragonality (c/a) which shows a linear dependence on (T_c-T) for the 9/3 sample (Fig. 3c) and a quadratic dependence on (T_c-T) for the 2/3 sample (Fig. 3d) in agreement with the P_z behaviour expected from the strain-polarization coupling.

In improper ferroelectrics, it is furthermore expected that the dielectric susceptibility will not obey a conventional Curie- Weiss law at the phase transition because the order parameter is no longer polar and will not directly couple with an external electric field. Instead the dielectric susceptibility will remain largely independent of temperature, with a step discontinuity at the transition temperature. Distinct behaviours are again coherently reported in Fig. 3e and 3f for the two kinds of sam- pies. While a typical Curie-Weiss law is observed for the 9/3 sample, the dielectric constant of the 2/3 sample remains fairly independent over a wide range of temperature, the most noteworthy feature being a small step in the dielectric constant at the transition temperature, as expected for an improper ferroelectric transition.

The high values of the spontaneous polarization and dielectric constant of the 2/3 sample (P₃ = 11 μC . cm and ε_r ~ 600 at 300 K) have also to be highlighted since they strongly contrast with what was previously reported for typical improper ferroelectrics

_2

(in gadolynium molybdate, P₃ = 0.2 μC . cm and ε_r ~ 10) . In our short period superlattices, although ferroelectricity is unambiguously improper, the polarization and dielectric constant are comparable to those in usual proper ferroelectrics due to the intrinsically high polarizability of PbTiθ₃ and SrTiC^. Inducing improper ferroelectricity in artificial superlattices appears therefore as a new and particularly promising approach for producing improved materials for technical applications requiring high dielectric response, at the same time stable over a very wide range of temperature.

Calculations have been performed within density functional theory and the local density approximation using the ABINIT package mentioned above. Highly transferable Teter pseudopotentials were used. Strontium 4s, 4p and 5s electrons, Ba 5s, 5p and 6s electrons, Ti 3s, 3p, 3d and 4s electrons and O 2s and 2p electrons were treated as valence states. The wavefunction was expanded in plane-waves up to a kinetic energy cutoff of 45 Hartrees. Phonon dispersion curves have been deduced from quantities (dynamical matrix, Born effective charges, optical dielectric constant) obtained using a linear response approach. Differences of polarization were computed using the Berry phase formalism. For the 1/1 PbTiO₃/SrTiO₃ superlattice grown along the (001) direction, the primitive unit-cell of the prototype P4/mmm phase contains 10 atoms but the calculations used a supercell containing 20 atoms, doubled along (110) in order to allow the condensation of AFD modes. All the calculations concerning the relative stability of the different phases in Fig. 2a have been performed within this supercell and integrals over the Brillouin- zone have been replaced by sums over a 4x4x4 Monkorst-Pack mesh of special k-points, which was checked to be sufficiently accurate .

In the prototype P4/mmm phase, calculations of the phonon dispersion curves for an epitaxial strain corresponding to a SrTiO₃ substrate allow the identification of different kinds of instable phonon modes at V and M (1/2,1/2,0) . In the P4/mmm phase (D_4h) , the irreducible representations at V can be written, following Miller and Love notations, as:

Z Y₁ ⁺B iy Θ 3 ΓΓ Φ 6 Yf m r₄ ^"θ T ΓΓ and the irreducible representations at the M (1/2,1/2,0) point are :

Mf% KI:^*θ M₃ ^÷θ 2 M₄ ^*θ 2 M₅ ^÷θ Mfθ 4 M₂ ^"θ 3 MfΦ M₄ ^"θ 6 M₅ ^"

The unstable modes at r and M as well as their main characteristics are summarized in the Table below. It can be checked that the product of AFD_ZO (M₄ ) and AFD_zi (M₂ ) modes transforms like the FE_Z (T₃ ^") .

Table: Main characteristics of the unstable phonon modes at r and M in the P4/mmm phase of a 1/1 PbTiO₃/SrTiO₃ superlattice on a SrTiO₃ substrate.

On the experimental side, the deposition temperature used for the PbTiθ₃ and SrTiθ₃ was approximately 530° C with an oxygen/argon flow rate of 14 sccm/32 seem and a pressure of 180 mTorr. For the SrRuθ₃ electrodes the deposition temperature was approximately 560° C with an oxygen/argon flow rate of 6 sccm/32 seem and a pressure of 100 mTorr. The polarization and dielectric constant were experimentally measured using an Aixacct TF- Analyzer 2000 and an Agilent 4284A LCR meter. The voltage used for switching was 2V applied across a lOOnm film (the coercive voltage of the films was very low) . The data displayed was measured at 1000Hz, both the polarization and the dielectric constant were found to be virtually independent of measurement frequency in the range 100Hz-IOOkHz . The x-ray diffraction measurements were carried out on a Panalytical X' Pert Pro MRD equipped with an Anton-Parr DHS900 domed hot stage.

Claims

1. Material comprising finely layered heterostructures of at least two oxide materials, wherein the layered structure allows coupling of vibrational modes or structural distortions at interfaces enhancing properties of the oxide materials from the group comprising polarization, strain-polarization coupling, ferroelectric transition temperature, temperature dependence of the dielectric constant, magnitude of the dielectric constant, voltage tunability of the dielectric constant, piezoelectric properties, piezomagnetic properties, magnetization, transition temperature associated with magnetic ordering and magnetic susceptibility.

2. Material according to claim 1 being ferroelectric or having induced ferroelectricity or pyroelectricity in the structure, when the parent materials are not intrinsically ferroelectric.

3. Material according to claim 1 or 2, wherein the heterostructures are including epitaxial, polycrystalline, nanocrystalline or amorphous structures.

4. Material according to one of claims 1 to 3, wherein the layered structure comprises alternating layers of perovskite-like ABO₃ structures with variation on either the A and/or B site from layer to layer, including the situation where A and/or B is a solid solution within one or both layers.

5. Material according to one of claims 1 to 4, comprising layered perovskite-like ABO₃ structures, wherein A and B are cations of different sizes.

6. Material according to claim 5, comprising single layers of ABO₃ and AΕO₃, wherein A is Pb, A' is Sr and B is Ti.

7. Material according to one of claims 1 to 6, wherein the het- erostructures are produced by thin film deposition techniques including sputtering, pulsed laser deposition or molecular beam epitaxy.

8. Dielectric device, especially for high k dielectric applications, including but not limited to capacitors in memory devices, comprising the material according to one of claims 1 to 7.

9. Magnetic device achieving electric control of magnetic properties, magnetic control of electrical properties or multi-state electronic memories comprising the material according to one of claims 1 to 7.

10. Non-volatile memory device, in which information is stored using either polarization or magnetization, comprising the material according to one of claims 1 to 7.

11. Varactor type device, in which the tunability of the dielectric constant is used, comprising the material according to one of claims 1 to 7.