**1. Introduction**

It is known that lead-based ferroelectric systems exhibit very good properties for differ‐ ent kinds of applications [1-6] and also that high-level ferroelectric and piezoelectric activities have remained confined to these materials. The only drawback in the technolo‐ gy as a whole has been the environmentalist's nightmare of its dependence on a high leadcontaining family of materials [7]. Therefore, over the last 20 years there has been an enormous effort made in developing lead-free ferroelectric systems in order to obtain better dielectric, ferroelectric, piezoelectric and pyroelectric properties than those of convention‐ al ferroelectric ceramics based on lead, such as lead zirconate titanate (PZT) [8-10].

An important group of lead-free ferroelectric materials belong to the Aurivillius family {[Bi2O2] 2+[Am−1BmO3m+1] 2−}, which was discovered by Bent Aurivillius in 1949 [11]. These compounds have a complex structure, which is composed of perovskite blocks ([Am-1BmO3m +1] 2-) interleaved between bismuth and oxygen layers ([Bi2O2] 2+), where *m* is the number of perovskite blocks in the structure. The *A* sites of the structure are typically occupied by elements such as Sr2+, Ba2+, Ca2+ and Bi3+, with low valence; the *B* sites are typically occupied by elements with high valence such as Ti4+, Nb5+ and W6+ [12-16]. These materials have received great attention due to their large remanent polarization, low real dielectric permittivity at room temperature, lead-free nature, relatively low processing temperatures, high Curie tempera‐ tures, high electromechanical anisotropy and coercive fields, and excellent piezoelectric properties [8-9], which have suggested them as good candidates for high-temperature piezoelectric applications and memory storage. The bismuth layers [Bi2O2] 2+ constrain the size of perovskite blocks establishing a limit for the incorporation of elements into them and providing the mixing of different elements between *A* sites and bismuth sites in the layered structure [17]. The ferroelectricity depends strongly on the crystallographic orientation of these materials, which is the subject of continuing researches. The main contribution to its sponta‐ neous polarization comes from the displacement of the *A* cation in the perovskite block, which is quite different for the perovskite structure. It is well known that these have the majority polarization vector along the a-axis in a unit cell and that the oxygen vacancies prefer to stay in the Bi2O2 layers, where their effect upon the polarization is thought to be small, and not in the octahedral site that controls polarization [18].

Figure 1 shows the structure for some Aurivillius systems with m=1, m=2 and m=3, at the paraelectric phase, as examples. The structural studies on these materials have shown a relation between the number of perovskite blocks and the symmetry of the cell, i.e., the number of perovskite blocks is related to the crystallographic orientation and to the plane of polarization in these materials [12-16, 18-19]. The polarization vector has also shown a relation to the number of perovskite blocks [18]. For even-layered systems, it has been reported to be a restriction on the polarization to the *a-b* plane of the cell and an orthorhombic symmetry with *A21am* space group [18]. For odd-layered systems, the polarization has shown a component in c and orthorhombic phase with *B2cb* space group [18]. Other results have shown a strong relation between the elements in *A* sites of the structure and the symmetry of the cell [13-14, 17].

**Figure 1.** Structure of some Aurivillius materials with m=1, m=2 and m=3, at the paraelectric phase.

**1. Introduction**

86 Ferroelectric Materials – Synthesis and Characterization

{[Bi2O2]

+1]

2+[Am−1BmO3m+1]

It is known that lead-based ferroelectric systems exhibit very good properties for differ‐ ent kinds of applications [1-6] and also that high-level ferroelectric and piezoelectric activities have remained confined to these materials. The only drawback in the technolo‐ gy as a whole has been the environmentalist's nightmare of its dependence on a high leadcontaining family of materials [7]. Therefore, over the last 20 years there has been an enormous effort made in developing lead-free ferroelectric systems in order to obtain better dielectric, ferroelectric, piezoelectric and pyroelectric properties than those of convention‐

al ferroelectric ceramics based on lead, such as lead zirconate titanate (PZT) [8-10].

2-) interleaved between bismuth and oxygen layers ([Bi2O2]

piezoelectric applications and memory storage. The bismuth layers [Bi2O2]

the octahedral site that controls polarization [18].

An important group of lead-free ferroelectric materials belong to the Aurivillius family

compounds have a complex structure, which is composed of perovskite blocks ([Am-1BmO3m

perovskite blocks in the structure. The *A* sites of the structure are typically occupied by elements such as Sr2+, Ba2+, Ca2+ and Bi3+, with low valence; the *B* sites are typically occupied by elements with high valence such as Ti4+, Nb5+ and W6+ [12-16]. These materials have received great attention due to their large remanent polarization, low real dielectric permittivity at room temperature, lead-free nature, relatively low processing temperatures, high Curie tempera‐ tures, high electromechanical anisotropy and coercive fields, and excellent piezoelectric properties [8-9], which have suggested them as good candidates for high-temperature

of perovskite blocks establishing a limit for the incorporation of elements into them and providing the mixing of different elements between *A* sites and bismuth sites in the layered structure [17]. The ferroelectricity depends strongly on the crystallographic orientation of these materials, which is the subject of continuing researches. The main contribution to its sponta‐ neous polarization comes from the displacement of the *A* cation in the perovskite block, which is quite different for the perovskite structure. It is well known that these have the majority polarization vector along the a-axis in a unit cell and that the oxygen vacancies prefer to stay in the Bi2O2 layers, where their effect upon the polarization is thought to be small, and not in

Figure 1 shows the structure for some Aurivillius systems with m=1, m=2 and m=3, at the paraelectric phase, as examples. The structural studies on these materials have shown a relation between the number of perovskite blocks and the symmetry of the cell, i.e., the number of perovskite blocks is related to the crystallographic orientation and to the plane of polarization in these materials [12-16, 18-19]. The polarization vector has also shown a relation to the number of perovskite blocks [18]. For even-layered systems, it has been reported to be a restriction on the polarization to the *a-b* plane of the cell and an orthorhombic symmetry with *A21am* space group [18]. For odd-layered systems, the polarization has shown a component in c and orthorhombic phase with *B2cb* space group [18]. Other results have shown a strong relation between the elements in *A* sites of the structure and the symmetry of the cell [13-14, 17].

2−}, which was discovered by Bent Aurivillius in 1949 [11]. These

2+), where *m* is the number of

2+ constrain the size

SrBi2Nb2O9 (SBN) is a member of the Aurivillius family in which the ferroelectric properties can be affected by the crystallographic orientation due to their anisotropic crystal structure [13-14]. This system has received particular attention due to its large fatigue resistance, which has been associated with the migration of oxygen vacancies in the material [20]. The Sr2+cation, which is located between the corner-sharing octahedral, can be totally or partially replaced by other cations, as barium is an important element for improving fatigue resistance [20]. The studies on the barium-modified SrBi2Nb2O9 system have shown interesting results from the structural and dielectric point of view [13-14, 20-25]. Structural studies have shown an orthorhombic symmetry with *A21am* space group for pure and doped SBN samples [22]. The mixing of different elements between *A* sites and bismuth sites, which occurs to equilibrate the lattice dimensions between the (Bi2O2) 2+ layers and the perovskite blocks, has been also analysed [22]. The oxygen vacancies, which are the results of Bi3+ for Ba2+/Sr2+ substitution, could have an important influence in the properties of these compositions [22].

For the Sr1-xBaxBi2Nb2O9 system (x= 0, 15, 30, 50, 70, 85, 100 at%), the barium concentration dependence of *Tm*, as well as the temperature of the corresponding maximum for the real part of the dielectric permittivity, has suggested a cation site mixing among atomic positions, which has been supported by structural analysis [22]. For compositions with x ≤ 30 at%, *Tm* increased with the barium concentration; for x ≥ 50 at%, a decrease of *Tm* and a widening of the curves was observed with the increase of the barium concentration. The structural studies have shown the mixing of Sr2+, Ba2+ and Bi2+ into *A* sites and the bismuth sites of the structure [22]. For lower barium concentrations (x ≤ 30 at%), the presence of bismuth into *A* sites and the increasing of the strontium concentration into this site, has been discussed as the principal reason for the increase of the *Tm* value. The higher barium concentration into *A* sites was obtained for the compositions with x ≥ 50 at% [22], supporting the decreasing of *Tm* [22].

On the other hand, a change from normal ferroelectric-paraelectric phase transition to relaxor behaviour has been observed when the barium concentration is increased [22]. For the compositions showing relaxor behaviour, an increase of the frequency dispersion degree was also observed with the increase of barium concentration. The relaxor behaviour is typical of materials with a disorder distribution of different ions in equivalent sites of the structure, which is called compositional disorder. For the studied materials, the relaxor behaviour has been explained with reference to the positional disordering of cations at *A* sites of the structure, which delays the evolution of long-range polar ordering [23, 26].
