Zuo-Guang. Ye Advanced Dielectric Piezoelectric and Ferroelectric Materials: Synthesis, Characterisation and Applications

Подождите немного. Документ загружается.

Handbook of dielectric, piezoelectric and ferroelectric materials1000

Figure 32.21(d) shows Q–V curves of the SZO/STO superlattices measured

at 132.5Hz. It should be stressed that the [SZO

/STO

]

superlattice shows

a clear hysteresis loop in the Q–V relation. The absolute value of the remanent

polarization could not be determined because of the use of interdigital

electrodes, but the Q–V relations indicate that the remanent polarization

decreases with increasing periodicity. The hysteresis loop was only observed

around or below the frequency of the dielectric relaxation. The ferroelectricity

in the SZO/STO superlattices is induced artificially by the lattice distortion

because both materials do not have a ferroelectric nature. Abe and coworkers

62,63

also reported that the ferroelectricity of BTO–STO solid solution films was

enhanced by the anisotropic lattice distortion. From the results of Figs 32.8

and 32.20, it is obvious that the artificial ferroelectricity depended on the

anisotropic lattice distortion, i.e. the [SZO

/STO

]

superlattice with the

highest lattice distortion showed the largest remanent polarization with a

clear hysteresis curve. It is interesting that the remanent polarization is not

maintained after the measuring cycle in the artificially induced ferroelectricity.

In the measurement using interdigital electrodes, the electric field is mainly

applied along the film plane, which is perpendicular to that of lattice distortion

along the c-axis. Therefore, the spontaneous polarization should rotate 90°

in order to show the hysteresis loops. The 90° switching of the spontaneous

polarization tends to relax after the electric field is removed, which may be

the reason for the reduced remanent polarization after a measuring cycle.

The dielectric loss peak with the Debye-like dielectric relaxation may be due

to the rotation of the spontaneous polarization induced by the lattice distortions.

The artificial ferroelectricity found in the artificial superlattices is unique

because it is observed only in the low-frequency region accompanied with

the dielectric relaxation. In normal ferroelectrics, dielectric relaxation is not

observed up to the GHz region and ferroelectricity is retained up to much

higher frequencies. The 90° switching of spontaneous polarization in the

superlattices may have a limit of switching speed.

Considering the lattice parameters of bulk specimens, the stress induced

in the SZO/STO superlattices must be larger than that induced in the BTO/

STO superlattices as mentioned in Section 32.3. The artificial ferroelectricity

was not observed in the BTO/STO superlattices. This indicates that the

anisotropic lattice distortion is more effective at inducing the ferroelectricity

than the ferroelectric nature of BTO. The BTO/STO superlattices showed a

higher dielectric permittivity than the SZO/STO

superlattices. It seems that

a large lattice distortion enhances the ferroelectricity but reduces the dielectric

permittivity in the perovskite superlattices, which is analogous with the

behaviors of normal ferroelectrics and relaxors. The anisotropic lattice distortion

induced by the lattice mismatch between the film and substrate and between

the constituent layers in superlattices played an important role in giving rise

to the high permittivity and the artificial ferroelectricity.

WPNL2204

Dielectric and optical properties of perovskite 1001

32.6 Conclusions and future trends

The aim of our study of perovskite artificial superlattices was to answer the

following simple question: are the dielectric properties of perovskite artificial

superlattices different from the thin films with the same composition and

thickness? We can now say that the answer is ‘yes’. In this chapter, important

studies on the preparation and the properties of the perovskite artificial

superlattices were introduced, and the advantage and disadvantage of deposition

processes were discussed. As far as the film quality is concerned, the reactive

MBE process seems to be superior to other processes. The quality and perfection

of the superlattice structure are the most important factors for research into

superlattices as pointed out by Shlom et al.

We have succeeded in preparing

BTO/STO and SZO/STO superlattices using reactive MBE system but the

films obtained still suffer from quality issues. The deposition of perfect

superlattices containing metal elements with high melting temperatures requires

further improvement of the whole system. The lattice distortion induced by

the lattice mismatch between the film and substrate and also between the

constituent layers in the superlattices is the origin of the dielectric anomalies

observed in the superlattices. In both BTO/STO and SZO/STO superlattices,

the highest anisotropic distortion was induced in the superlattices when the

periodicity of the superlattices is 10. The refractive index of superlattices

measured with a spectroscopic ellipsometer showed maximum values when

the highest anisotropic distortion was induced into the superlattices. The

permittivity of the superlattices was measured using interdigital electrodes

and the EM analysis. The BTO/STO superlattices showed a marked

enhancement of the permittivity when the superstructure periodicity was 10.

The similar enhancement was confirmed in the superlattice prepared by the

RF-sputtering process. The mechanism of the dielectric enhancement was

discussed in terms of the anisotropic lattice distortion. Ferroelectricity was

artificially induced in the SZO/STO superlattices with the highest anisotropic

lattice distortion even though both compounds were non-ferroelectrics. This

ferroelectricity was interpreted by 90° rotations of the spontaneous polarizations

induced by the lattice distortion into the superlattices.

The future trend of this subject will be the manipulation of one molecular

layer in the superlattices. The insertion of magnetic active layers in ferroelectric

superlattices may produce a new interaction between magnetic and ferroelectric

properties. The conductive layers in the dielectric superlattices may enhance

the interfacial polarization to obtain high permittivity. The composition

graduation in the superlattice is also a challenging topic for study. This kind

of artificially modulated structures will produce special stress fields in the

films, which may give additional enhancements of the properties.

The history of oxide superlattices is only 20 years old. In spite of the

recent progresses in the deposition, characterization and property evaluation

WPNL2204

Handbook of dielectric, piezoelectric and ferroelectric materials1002

techniques, the study of oxide superlattices is still the state of the art for most

materials and ceramics scientists. We believe that the oxide superlattices

have definitely opened a new field of materials science and nanotechnology,

in which young or established materials scientists can make a significant

impact. We hope this chapter provides them with useful guidance to investigate

this challenging but very fruitful research field.

32.7 References

1. Esaki L and Tsu R (1970), ‘Superlattice and negative differential conductivity in

semiconductors’, IBM J Res Development, Jan., 61–65.

2. Tsu R and Esaki L (1973), ‘Tunneling in a finite superlattice’, Appl Phys Lett, 22,

562–564.

3. Esaki L and Chang L L (1974), ‘New transport phenomenon in a semiconductor

superlattice’, Phys Rev Lett, 33, 495–498.

4. Iijima K, Terashima T, Bando Y, Kamigaki K and Terauchi I (1992), ‘Atomic layer

growth of oxide thin films with perovskite-type structure by reactive evaporation’,

J Appl Phys, 72, 2840–2845.

5. Iijima K, Terashima T, Yamamoto K, Hirata K and Bando Y (1990), ‘Preparation of

ferrolectrics BaTiO

thin films by activated reactive evaporation’, Appl Phys Lett,

56, 527–529.

6. Ohkubo H, Kanda N, Yoshimoto M and Koinuma H (1993), ‘Fabrication of high-

quality perovskite oxide films by lateral epitaxy verified with RHEED oscillation’,

Jpn J Appl Phys, 32, 689–691.

7. Yoshimoto M, Maeda T, Gonda S, Shomozono K, Koinuma H, Ishiyama O and

Shomomura M (1994), ‘Molecular layer epitaxy of perovskite-based oxides by laser

MBE’, Mat Res Soc Symp Proc, 341, 133–138.

8. Kawasaki M, Takahashi K, Maeda T, Tsuchiya R, Shonohara M, Ishiyama O, Yonezawa

T, Yoshimoto M and Koinuma H (1994), ‘Atomic control of the SrTiO

crystal

surface’, Science 266, 1540–1542.

9. Yoshimoto M, Maeda T, Ohnishi T, Lee G H and Koinuma H (1996), ‘Atomic scale

contral of epitaxial growth and interface in oxide thin films for advanced oxide

lattice engineering’, Mat Res Soc Symp Proc, 401, 21–31.

10. Tabata H, Tanaka H and Kawai T (1994), ‘Formation of artificial BaTiO

/SrTiO

superlattices using pulsed laser deposition and their dielectric properties’, Appl Phys

Lett, 65, 1970–1972.

11. Tsurumi T, Suzuki T, Yamane M and Daimon M (1994), ‘Fabrication of barium

titanate-strontium titanate artificial superlattice by atomic layer epitaxy’, Jpn J Appl

Phys, 33, 5192–5195.

12. Kim J, Kim Y, Kim Y-S, Lee J, Kim L and Jung D (2002), ‘Large nonlinear dielectric

properties of artificial BaTiO

/SrTiO

superlattices’, Appl Phys Lett, 80, 3581–3583.

13. Kim L, Jung D, Kim J, Kim Y, Kim Y-S and Lee J (2003), ‘Strain manipulation in

BaTiO

/SrTiO

artificial lattice toward high dielectric constant and its nonlinearilty’,

Appl Phys Lett, 82, 2118–2120.

14. Kim J, Kim L, Jung D, Kim Y-S, Kim I-W, Je J-H and Lee J (2003), ‘Thickness

dependent dielectric property of BaTiO

/SrTiO

artificial superlattices’, Jpn J Appl

Phys, 42, 5901–5903.

WPNL2204

Dielectric and optical properties of perovskite 1003

15. Guerville F De, Marssi M El, Raevski I P, Karkut M G and Yuzyuk Yu I (2006), ‘Soft

mode dynamics and the reduction of Ti

disorder in ferroelectric/relaxor superlattices

BaTiO

/BaTi

0.68

0.32

’, Phys Rev, B74, 064107.

16. Christen H-M, Knauss L A and Harshavardhan K S (1998), ‘Field-dependent dielectric

permittivity of paraelectric superlattice structure’, Mater. Sci. Engineer., B56, 200–

203.

17. Tong F O, Yu W X, Liu F, Zuo Y and Ge X (2003), ‘Microstructural study of

BaTiO

/SrTiO

superlattice’, Mater Sci Engineer, B98, 6–9.

18. Qu B D, Evstigneev M, Johnson D J and Prince R H (1998), ‘Dielectric properties

of BaTiO

/SrTiO

multilayered thin films prepared by pulsed laser deposition’, Appl

Phys Lett, 72, 1394–1396.

19. Katiyar R S, Yuzyuk Yu I, Das R R, Bhattacharya P and Gupta V (2005), ‘Polarization

Raman spectra of BaTiO

/SrTiO

superlattices’, Ferroelectrics, 329, 907–916.

20. Choi T and Lee J (2005), ‘Dielectric properties of PbZrO

/PbTiO

artificial superlattices

grown by pulse laser deposition’, Ferroelectrics, 328, 799–804.

21. Seo S S A, Lee H N and Noh T W (2005), ‘Infrared spectroscopy of CaTiO

, SrTiO

BaTiO

, Ba

0.5

TiO

, thin films, and (BaTiO

)

/(SrTiO

)

superlattice grown on

SrRuO

/SrTiO

(001. substrates’, Thin Solid Films, 486, 94–97.

22. Jiang J C, Pan X Q, Tian W, Theis C D and Schlom D G (1999), ‘Abrupt PbTiO

SrTiO

superlattices grown by reactive molecular beam epitaxy,’ Appl Phys Lett, 74,

2851–2853.

23. Schlom D G, Haeni J H, Lettieri J, Theis C D, Tian W, Jiang J C and Pan X Q (2001),

‘Oxide nano-engineering using MBE’, Mater Sci Eng, B87, 282–291.

24. Tian W, Jiang J C, Pan X Q, Haeni J H, Li Y L, Chen L Q, Schlom D G, Neaton J

B, Rabe K M and Jia Q X (2006), ‘Structural evidence for enhanced polarization in

a commensurate short-period BaTiO

/SrTiO

superlattice’, Appl Phys Lett, 89, 092905.

25. Tsai H-N, Liang Y-C and Lee H-Y (2005), ‘Characteristics of sputter-deposited

BaTiO

/SrTiO

artificial superlattice films on an LaNiO

-coated SrTiO

substrate’,

J Cryst Growth, 284, 65–72.

26. Katiyar R S and Yuzyuk Yu I (2006), ‘Stress relaxation effects in ferroelectric thin

films and superlattices’, Ferroelectrics, 334, 487–491.

27. Wu T-B and Hung C-L (2005), ‘Constrained ferroelectricity in the (001)-textured

superlattices of PbZrO

/BaZrO

’, Appl Phys Lett, 86, 112902.

28. Hung C-L, Chueh Y-L, Wu T-B and Chou L-J (2005), ‘Characterization of constrained

ferroelectricity in PbZrO

/BaZrO

superlattice films’, J Appl Phys, 97, 034105.

29. Shimuta T, Nakagawara O, Makino T, Arai S, Tabata H and Kawai T (2002),

‘Enhancement of remanent polarization in epitaxial BaTiO

/SrTiO

superlattices

with “asymmetric” structure’, J Appl Phys, 91, 2290–2294.

30. O’Neill D, Bowman R M and Gregg J M (2000), ‘Dielectric enhancement and

Maxwell–Wagner effects in ferroelectric superlattice structures’, Appl Phys Lett, 77,

1520–1522.

31. Neaton J B and Rabe K M (2003), ‘Theory of polarization enhancement in epitaxial

BaTiO

/SrTiO

superlattices’, Appl Phys Lett, 82, 1586–1588.

32. Kim L, Kim J, Jung D, Lee J and Waghmare U V (2005), ‘Polarization of strained

BaTiO

/SrTiO

artificial superlattice: first-principles study’, Appl Phys Lett, 87,

052903.

33. Kim L, Kim J, Waghmare U, Jung D and Lee J (2005), ‘Lattice instability of BaTiO

SrTiO

artificial superlattice’, Integrated Ferroelectrics, 73, 3–10.

34. Kim L, Kim J, Waghmare U, Jung D and Lee J (2005), ‘Structural transition and

WPNL2204

Handbook of dielectric, piezoelectric and ferroelectric materials1004

dielectric response of an epitaxial strained BaTiO

/SrTiO

superlattice: a first-principles

study’, Phys Rev, B72, 214121.

35. Yang K, Wang C L, Li J C, Zhao M L and Wang X Y (2006), ‘Strain induced

ferroelectricity in the SrZrO

/SrTiO

superlattice: first principles study,’ Solid State

Comm., 139, 144–147.

36. Lo V C and Jiang W (2006), ‘Simulation of stress-induced enhanced dielectric

permittivity in ferroelectric superlattice,’ Integrated Ferroelectrics, 78, 35–43.

37. Tsurumi T, Miyasou T, Ohashi N and Ishibashi Y (1998), ‘Preparation and dielectric

property of BaTiO

/SrTiO

artificially modulated structures’, Jpn J Appl Phys, 37,

5104–5107.

38. Ishibashi Y, Tsurumi T, Ohashi N and Fukunaga O (1998), ‘Epitaxial growth of

barium titante thin films at low temperatures by low-energy positive oxygen ion

assisting’, Solid State Ionics, 108, 91–97.

39. Ishibashi Y, Ohashi N and Tsurumi T (2000), ‘Structural refinement of x-ray diffraction

profile for artificial superlattices’, Jpn J Appl Phys, 39, 186–191.

40. Nishizawa S, Tsurumi T, Hyoudo H, Ohashi N and Fukunaga O (1997), ‘Structural

changes in ZnO/NiO artificial superlattice made by ion beam sputtering’, Thin Solid

Films, 302, 133–139.

41. Collins R W (1990), ‘Automatic rotating element ellipsometers: calibration, operation

and real-time applications’, Rev Sci Instrum, 61, 2029–2062.

42. Shirafuji T, Motomura H and Tachibana K (2004), ‘Fourier transform infrared phase-

modulated ellipsometry for in situ diagnostics of plasma surface interactions’, J

Phys, D37, R49–R73.

43. Tolmachev V A (1999), ‘Adsorption-ellipsometry method of studying the optical

profile, thickness, and porosity of thin films’, J Opt Tech, 66, 596–607.

44. Jenkins T E (1999), ‘Multiple-angle-of-incidence ellipsometry’, J Phys, D32, R45–

R56.

45. Topcu S, Chassagne L and Alayli Y (2003), ‘A new type of fiber-optic-based

interferometric ellipsometer for in situ and real-time measurements’, Rev Sci Instrum,

74, 4442–4447.

46. Lee J, Koh J and Collins R W (2001), ‘Dual rotating-compensator multichannel

ellipsometer: instrument development for high-speed Mueller matrix spectroscopy

of surfaces and thin films’, Rev Sci Instrum, 72, 1742–1754.

47. Azzam R M A and Bashara N M (1977), Ellipsometry and Polarized Light, Amsterdam,

North Holland, Elsevier Science Publishers BV.

48. Sik J, Schubert M, Leibiger G, Gottschalch V and Wagner G (2001), ‘Band-gap

energies, free carrier effects, and phonon modes in strained GaNAs/GaAs and GaNAs/

InAs/GaAs superlattice heterostructures measured by spectroscopic ellipsometry’, J

Appl Phys, 89, 294–305.

49. Sik J, Schubert M, Leibiger G, Gottschalch V, Kirpal G and Humlicek J (2000),

‘Near-band-gap optical functions spectra and band-gap energies of GaNAs/GaAs

superlattice heterostructures measured by spectroscopic ellipsometry’, Appl Phys

Lett, 76, 2859–2861.

50. Logothetidis S, Kalfagiannis N, Sarakinos K and Patsalas P (2006), ‘Investigation of

bilayer period and individual layer thickness of CrN/TiN superlattices by ellipsometry

and X-ray techniques’, Surf Coat Tech, 200, 6176–6180.

51. Bellani V, Stella A, Chen C and Wang X (2005), ‘Spectroscopic ellipsometry study

of Cd

1–x

Te/CdTe superlattices’, J Appl Phys, 98, 103523.

52. Losurdo M, Giuva D, Giangregorio M M, Bruno G and Brown A S (2004),

WPNL2204

Dielectric and optical properties of perovskite 1005

‘Spectroscopic ellipsometry characterization of interface reactivity in GaAs-based

superlattices’, Thin Solid Films, 455–456, 457–461.

53. Ma Y, Tsurumi T, Nishizawa S, Ohashi N and Fukunaga O (1998), ‘Spectroscopic

ellipsometry of NiO/ZnO artificial superlattice’, J Am Ceram Soc, 81, 2125–2129.

54. Nijs J M M and Silfhout A (1988), ‘Systematic and random errors in rotating-

analyzer lllipsometry,’ J Opt Soc Am, 5, 773–781.

55. Kim S Y (1996), ‘Simultaneous determination of refractive index, extinction coefficient,

and void distribution of titanium dioxide thin film by optical methods’, Appl Opt 35,

6703–6707.

56. Trolier-McKinstry S, Chindaudom P, Vedam K and Hiremath B V (1995),

‘Characterization of optical thin films by spectroscopic ellipsometry’, J Am Ceram

Soc, 78, 2412–2416.

57. Chindaudom P and Vedam K (1992), ‘Determination of the optical constants of an

inhomogeneous transparent LaF

thin film on a transparent substrate by spectroscopic

ellipsometry’, Opt Lett, 17, 538–540.

58. Gustin K M (1987), ‘Optical characterization of low-index transparent thin films on

transparent substrates by spectroscopic ellipsometry’, Appl Opt, 26, 3796–3802.

59. Farnell G W, Cermak I A, Silverster P and Wong S K (1970), ‘Capacitance and field

distributions for interdigital surface-wave transducers’, IEEE Trans. UFFC, SU-17,

188–195.

60. Li J, Kakemoto H, Wada S, Kawaji H and Tsurumi T (2006), ‘Dielectric relaxation

in GHz region and phase transition of BaTiO

-based ceramics’, J Appl Phys, 100,

024106.

61. Tsurumi T, Harigai T, Tanaka D, Nam S-M, Kakemoto H, Wada S and Saito K

(2004), ‘Artificial ferroelectricity in perovskite superlattices’, Appl Phys Lett, 85,

5016–5018.

62. Abe K, Komatsu S, Yanase N, Sano K and Kawakubo T (1997), ‘Ferroelectric

properties in heteroepitaxial Ba

0.6

0.4

TiO

thin films on SrRuO

/ SrTiO

substrates’,

Jpn J Appl Phys, 36, 5575–5579.

63. Kawakubo T, Komatsu S, Abe K, Sano K, Yanase N and Fukushima N (1998),

‘Ferroelectric properties of SrRuO

/(Ba, Sr)TiO

/SrRuO

epitaxial capacitor’, Jpn J

Appl Phys, 37, 5108–5111.

WPNL2204

1006

33.1 Introduction

The study of ferroelectrics offers unparalleled opportunities to advance the

fundamental physics and materials design of dielectric materials, and to

develop innovative semiconductor-based applications such as nonvolatile

memories, piezoelectric and electro-optic devices and uncooled infrared

detectors [1]. Ferroelectric bismuth titanate (Bi

, BiT) [2] has been

regarded as a promising material because of its high Curie temperature (T

)

of 675 °C [3], large spontaneous polarization (P

) of ~50µC/cm

[4, 5] and

large electro-optic coefficient [6, 7].

In the crystal structure of BiT, the perovskite layers composed of three

TiO

octahedral layers with Bi at the A site are sandwiched between Bi

layers, as shown in Fig. 33.1. In the paraelectric state above T

, BiT has a

parent tetragonal structure with I4/mmm symmetry. The ferroelectric transition

results in descending symmetry, and the crystals at room temperature exhibit

a monoclinic distortion with B1a1 space group [8]. Polarization measurements

of BiT single crystals have revealed a P

of 50µC/cm

along the a-axis and

of 3–5µC/cm

along the c axis [4, 5]. Rae et al. [8] have clearly shown

through single-crystal structural analysis at room temperature that the shape

of TiO

octahedra in the paraelectric state remains nearly intact in the

ferroelectric state. The TiO

octahedra rotate in the a–b plane as well as

incline away from the c-axis. The displacement of the TiO

octahedra with

respect to Bi ions along the a-axis mainly contributes to the ferroelectric

polarization in the BiT system and off-center Ti displacements in the octahedra

play a minor role in P

[8]. These structural features of BiT are markedly

different from the ABO

-type ferroelectrics such as BaTiO

and PbTiO

where

the ferroelectricity results from the relative displacements of A and B cations

with respect to oxygen-octahedral cage.

It is recognized that defects are closely related to the leakage current and

polarization properties of BiT [9–16]. Takahashi et al. [17, 18] have revealed

the conduction mechanism of BiT single crystals and polycrystalline ceramics

Crystal structure and defect control in

-based layered ferroelectric

single crystals

Y NOGUCHI and M MIYAYAMA,

The University of Tokyo, Japan

WPNL2204

Crystal structure and defect control 1007

at high temperatures. The electrical conductivity along the a-axis is higher

by several orders of magnitude than that along the c-axis. Oxide-ion migration

ascribed to a large number of oxygen vacancies in the perovskite layers

controls the high electrical conduction along the a-axis. With decreasing

temperature, the contribution of oxide ions to the electrical conduction becomes

less significant, and electron holes become the dominant carrier at low

temperatures. Recently, Noguchi et al. [12, 19] have clarified the origin of

the leakage current of BiT crystals at room temperature. Electron holes

arising from the incorporation of oxygen from ambient at the oxygen vacancies

act as detrimental carriers of leakage current up to high electric fields.

La- and Nd-substituted BiT (BLT and BNT) have received a great deal of

attention since these substitutions led to better memory properties such as a

large remanent polarization (P

), a high fatigue endurance and a superior

insulating property [20–32]. These rare-earth (RE) elements have been reported

to preferentially occupy the perovskite A site [33]. The presence of RE at the

A site reduces lattice distortion in the perovskite layers [34–36] as well as

providing an additional domain-nucleation site, an antiphase-domain boundary

in the lattice [30]. Recently, considerable effort has been made to realize a

ferroelectric memory using BLT and BNT, and the device technology related

to processing and integration is being developed.

In this chapter, the polarization switching and leakage current properties

of the single crystals of BiT and RE-substituted BiT are discussed in terms

of the crystal structure, electronic structure, domain structure and defect

structure. At first, the ferroelectric distortion and phase transition are viewed

(b) Ferroelectric

(a) Paraelectric

Bi2

Ti2

Ti1

Bi1

Bi2

Ti2

Ti1

Bi1

33.1

Crystal structures of Bi

of (a) the paraelectric phase at

700°C and (b) the ferroelectric phase at 25 °C.

WPNL2204

Handbook of dielectric, piezoelectric and ferroelectric materials1008

from the electronic and structural features. Next, the defect structure in BiT

is shown, and the vacancies of Bi at the A site and the adjacent oxygen

vacancies (the resultant electron hole) are shown to dominate the overall

polarization and leakage current properties of BiT crystals. The role of La

and Nd at the A site is discussed from the structural and electronic points of

view. Finally, a guiding principle of defect control in fabrication process for

achieving high-performance BiT-based devices is proposed.

Here, the materials properties along the a-axis (the major polarization

direction) are focused, and therefore the simplified crystal symmetry of

orthorhombic B2cb instead of exact monoclinic B1a1 is applied to discuss

the structure–property relations in the BiT system.

33.2 Crystal structure

33.2.1 Ferroelectric distortion and spontaneous

polarization

Figure 33.1 shows the crystal structure of BiT in (a) the paraelectric state

(700°C) and (b) the ferroelectric state (25°C) determined by neutron diffraction

study. The perovskite layers in the paraelectric state are composed of a

regular stacking of three TiO

octahedra with Bi at the A site, while a large

structural distortion is stored in the crystal lattice in the ferroelectric state.

The TiO

octahedra rotate in the a–b plane as well as incline away from the

c-axis [8, 10, 34, 36–38]. These octahedral distortions are induced by the

freezing of several kinds of displacive-type ferroelectric soft modes. The

dipole moments caused by the ionic displacements along the b-axis are

cancelled due to the presence of glide plane normal to the b-axis. Thus, the

constituent ions are displaced cooperatively along the a and c-axes, leading

to a spontaneous polarization (P

) along the both axes. It has been reported

that the P

along the a-axis is ~50µC/cm

[4, 5, 10], which is much larger

than that along the c-axis (3–5µC/cm

[4]. Hereafter, we pay attention only

to the P

along the a-axis.

Assuming the constituent ions to have their formal charge (+3 for Bi, +4

for Ti, –2 for O), we can calculate the ionic spontaneous polarization (P

s,ion

)

by the following equation:

s,ion

= ∑

× ∆x

× Q

e)/Ω [33.1]

where m

is the site multiplicity, ∆x

is the displacement along the a-axis

from the corresponding position in the parent tetragonal (I4/mmm) structure,

e is the ionic charge for the ith constituent ion, and Ω is the volume of the

unit cell. Figure 33.2 depicts (a) the actual crystal structure at 25°C projected

along the b-axis, (b) the ∆x

along the a-axis and (c) the dipole moment (∆x

× Q

e). It is interesting to note that the three TiO

octahedral layers move

WPNL2204

Crystal structure and defect control 1009

cooperatively along the a-axis with respect to Bi ions, which leads to a P

s,ion

opposite to the a-axis. In our analysis, P

s,ion

is estimated to be 37.6µC/cm

at 25°C. This value is smaller than that of the experimental P

, because the

electronic contribution is not taken into account in the calculation of P

s,ion

[39]. Although the relative displacements of Ti ions with respect to the TiO

octahedral center are seen, the off-center Ti displacements play a minor role

in P

s,ion

. The entire shift of the TiO

octahedra with respect to heavy Bi ions

results in the large dipole moments of oxide ions in the perovskite layers,

which gives the dominant contribution to the ferroelectric polarization in the

BiT system (see Fig. 33.2c).

32.2.2 Bond valence analysis and the role of Bi in the

ferroelectric phase transition

To understand the distorted structure in the ferroelectric phase, bond valence

analysis [40, 41] has been used as a powerful approach, especially for

understanding the crystal chemistry in layered ferroelectrics [38, 42, 43].

The bond valence sum (V

) is assumed to be the apparent valence of an atom

in the crystal structure. Figure 33.3 shows the V

of cations as a function of

0.02nm

(b)

10µC/cm

(c)

Dipole moment

= 37.6µC/cm

Displacement (∆

)

Bi2

Bi1

Ti2

Ti1

(a)

Center

of gravity

Ti1

Bi1

Ti2

Bi2

33.2

Detailed structural distortion at 25 °C along the polarization

direction (the

-axis): the actual crystal structure projected along the

-axis, (b) the atomic displacement (∆

) along the

-axis from the

corresponding position in the parent tetragonal (

I4/mmm

) structure

and (c) the dipole moment obtained from the ∆

and formal charge.

The

along the

-axis at 25°C is estimated to be 37.6 µC/cm

WPNL2204