Synthesis and characterization of aurivillius phases in the Bi-Ag-Ti-O system

Article abstract of DOI:10.1111/j.1551-2916.2007.01770.x

New Aurivillius phases in the Bi-Ag-Ti-O system were investigated by means of a solid-state reaction and X-ray diffraction. We found that the oxygen partial pressure has a significant influence on the synthesis of the Aurivillius phases. The mixed-layer Aurivillius phase Ag0.5Bi8.5Ti7O27 was observed after firing in an O2 flow, but a single-phase material is difficult to obtain. A single-phase compound of the four-layer Aurivillius phase Ag0.5Bi4.5Ti4O15 was obtained on firing in an oxygen partial pressure of 10 bar (1 × 106 Pa). The dielectric properties (at 1 MHz) of the Ag0.5Bi4.5Ti4O15 compound were as follows: Tmax=687°C, εr=166 (～20°C), and tan δ=0.004 (～20°C).

Full text of DOI:10.1111/j.1551-2916.2007.01770.x

J. Am. Ceram. Soc., 90 [8] 2363–2366 (2007)
DOI: 10.1111/j.1551-2916.2007.01770.x
r 2007 The American Ceramic Society
ournal
J
Synthesis and Characterization of Aurivillius Phases in the Bi–Ag–Ti–O
System
w
ˇ
Xing Hu, SreWo D. Skapin, and Danilo Suvorov*
Advanced Materials Department, Jomef Stefan Institute, 1000 Ljubljana, Slovenia
New Aurivillius phases in the Bi–Ag–Ti–O system were inves-
tigated by means of a solid-state reaction and X-ray diffraction.
We found that the oxygen partial pressure has a significant in-
fluence on the synthesis of the Aurivillius phases. The mixed-
layer Aurivillius phase Ag0.5Bi8.5Ti O was observed after
It is possible that many compounds might exist, but have re-
mained undetected because of their extremely low free energy of
formation compared with neighboring compounds and the high
nucleation barrier to complex-phase formation compared with
8
that of a simpler structure. Morgan PED suggested that fast or
7
27
firing in an O flow, but a single-phase material is difficult to
2
slow firing of molecularly mixed or other unusual precursor
mixtures would be helpful in finding these compounds.
In the present study, we tried to synthesize Aurivillius phases
in the Bi–Ag–Ti–O system with the nominal compositions
obtain. A single-phase compound of the four-layer Aurivillius
phase Ag0.5Bi4.5Ti O was obtained on firing in an oxygen
4
15
6
partial pressure of 10 bar (1ꢀ 10 Pa). The dielectric properties
(
at 1 MHz) of the Ag0 Bi Ti O compound were as follows:
.5
Ag0.5Bi8.5Ti
7
O
27, Ag0.5Bi4.5Ti
4
O
15, and AgBi 18, where
5
Ti
5
O
4.5
4
15
Tmax 5 6871C, e 5 166 (B201C), and tan d 5 0.004 (B201C).
n 5 3.5, 4, and 5 corresponds to [Bi B O3n11] and
2
O
2
][Anꢁ1
n
r
n 5 3.5 corresponds to the intergrowth of 314. The syntheses
6
took place in an O flow and in 10 bar of O (1ꢀ 10 Pa). The
2
2
solid-state reaction method was used to prepare all the compo-
sitions. The dielectric and ferroelectric properties of single-phase
Ag0.5Bi4.5Ti O were also investigated.
I. Introduction
4
15
ECENTLY, a great deal of attention has been given to the
R
Aurivillius phase compounds because of their potential
for applications in ferroelectric non-volatile memories
(
lius phases are a family of layered bismuth oxides that have been
II. Experimental Procedure
The starting materials used in this study were Bi
Ag O (499%), and TiO (anatase, 99.9%). The TiO
were dried at 6001C. The powders, in nominal compositions, were
homogenized in ethanol by ball milling with 3-mm yttria-stabi-
lized zirconia balls. The powders were placed in alumina crucibles
ina tubefurnace, firedinanoxygenflow at 101C/min to9001C for
12 h. Two more firings in an oxygen flow were also performed on
the above reaction product with the nominal composition Ag0.5-
1
Fe-RAM) and high-temperature piezoelectrics. The Aurivil-
,2
O
(99.975%),
powders
2
3
3
known for 50 years. The structural formula of these com-
2
2
2
2
1
2ꢁ
3n11
]
pounds is usually described as [Bi
2
O
2
]
[Anꢁ1
n
B O
, which
layers interleaved
with [Bi O ] layers along the c-axis. The Aurivillius phases
2
ꢁ
consists of perovskite-like (Anꢁ1
n 3n11
B O )
21
2
2
also occur as mixed-layer compounds with the general formula
Bi 3m11]. The structure is
made up of an intergrowth of one half of the unit cell of the
usual bismuth layer-structured oxides Bi A B O and one
[
2
O
2 n 2 2 m
][Anꢁ1B O3n11][Bi O ][Amꢁ1B O
7
Bi8.5Ti O27. One was heating at 101C/min to 6001C, followed by
2
nꢁ1
n
3n13
half of the unit cell of Bi
2 mꢁ1 m
A B O
3m13 along each c-axis.
0.11C/min to 9001C and soaking for 6 h; the other was heating at
101C/min to9001C for 12 h. Eachtime the samples were takenout
of the furnace, they were reground and re-milled. The powders
The restriction placed on the choice of A and B ions in
Aurivillius phases is that electrical neutrality must be maintained
in the overall composition. Armstrong and Newnham suggest-
ed that the instability could be related to the stability of the
perovskite layer and the size mismatch between the [Bi
4
were also fired under an oxygen partial pressure ðPO Þ of 10 bar
2
6
(10 Pa)at10001Cfor6h(heatingrate51C/min).Thesingle-phase
21
2
]
2
O
4 2
Ag0.5Bi4.5Ti O15 was sintered at 10001C for 3 h in 10 bar of O .
and the perovskite layers. The A site can be occupied by large
The constituent phases were assessed using powder X-ray
diffraction (XRD) studies (Bruker AXS D4 Endeavor diffracto-
meter with CuKa radiation, Bruker AXS, Karlsruhe, Germany).
The unit-cell parameters were determined using Crysfire and
Topas R software. The bulk density of the sintered samples was
evaluated by measuring the dimensions and the weight. The di-
electric properties were measured with an impedance analyzer
(Model HP 4284A LF, Hewlett-Packard, Palo Alto, CA). For
the dielectric and ferroelectric property measurements, some sil-
ver paste was fired on the samples at 7501C for 10 min. Thermal
analyses (DSC) were performed with a Netzsch STA 449C ther-
mal analysis system (Netzsch, Selb, Germany) in an atmosphere
of air. The DSC was carried out from room temperature to
8201C at a heating rate of 21C/min. The ferroelectric properties
were examined with a hysteresis meter (Radiant Technologies,
Precision LC, Albuquerque, NM).
1
2-fold-coordinated cations, such as Na , K , Ca , Sr ,
1
21
21
1
Ba , Pb , Bi , or Ln , and the B site by sixfold-coordinat-
21
21
31
31
3
1
31
ed cations, such as Fe , Cr , Ti , Nb , Ta , or W
41
51
51
61
.
1
1
21
5
Although Ag has a radius similar to Na and Sr , Subbarao
failed to obtain Ag0.5Bi4.5Ti 15 and so suggested that the
4
O
electronic configuration or polarizability also played a role in the
stability of the Aurivillius phases. However, in reality, it is the poor
thermal stability of Ag O that prevents the synthesis of
2
Ag-based compounds via a conventional solid-state reaction
method. To synthesize the perovskite compound Ag1/2Bi1/2TiO ,
3
high-temperature, high-pressure techniques were used by
Park et al., and the oxidizing conditions generated by the
6
7
2
decomposition of Ag O were used by Inaguma et al.
D. Damjanovic—contributing editor
III. Results and Discussion
Manuscript No. 22222. Received September 7, 2006; approved February 14, 2007.
This work was financially supported by the Science and Education Foundation of the
Republic of Slovenia under contract No. OK-92P10SK.
The nominal compositions of Ag0.5Bi8.5Ti
7 4 15
O27, Ag0.5Bi4.5Ti O ,
and AgBi Ti 18 were heated in an oxygen flow at 101C/min
5
5
O
*
Member, American Ceramic Society.
w
to 9001C and soaked for 12 h. The XRD analysis of the
reaction mixtures revealed the presence of multiple phases
Author to whom correspondence should be addressed. e-mail: huxingmse@yahoo.
com.cn
2
363

2
364
a)
Journal of the American Ceramic Society—Hu et al.
Vol. 90, No. 8
9
Pb(Mg1/3Nb2/3)O
Ti
composition Ag0.5Bi8.5Ti O was heated at 101C/min to
3
. To increase the amount of the Ag0.5Bi8.5-
7
O27 phase, the above reaction product with the nominal
(
Bi Ti O
Bi Ti O
2 2 7
4
3
12
Ag0.5Bi8.5Ti O
7
27
7
27
6
001C, followed by 0.11C/min to 9001C, and soaked for 6 h,
8
the so-called slow-heating method. The amount of Ag0.5Bi8.5-
Ti O increased from 20% to 75% and the pyrochlore phase
disappeared, as shown in Fig. 2. Next, after the slow-heating
C
B
7
27
method, the reaction product was heated at 101C/min to 9001C
7
and soaked for 12 h. The amount of Ag0.5Bi8.5Ti O27 increased
from 75% to 85%.
It is well-known that the presence of a pyrochlore phase is
always unfavorable for the synthesis of a perovskite phase such
9
. It was initially thought that firing a mix-
as Pb(Mg1/3Nb2/3)O
3
A
ture of Bi Ti O , TiO , and Ag O corresponding to the nom-
4
3
12
2
2
inal composition of Ag0.5Bi8.5Ti
the Ag0.5Bi8.5Ti
es of the Ag0.5Bi8.5Ti
slow-heating method for the corresponding reaction product,
the amount of Ag0.5Bi8.5Ti 27 phase could be increased from
% to 45%.
It is possible that the poor thermal stability of Ag O may
7
O
27 might aid the formation of
1
0
20
30
40
50
60
7
O
27 phase. Disappointingly, however, only trac-
2
θ (degree)
7
O
27 phase were observed. Using the above
7
O
(
b)
5
Bi Ti O
Ag0.5Bi8.5Ti O
7 27
4
3
12
2
Bi Ti O
2
2
7
prevent the synthesis of Ag-based compounds such as Ag
0
.5-
6
,7
3
Bi0.5TiO via a conventional solid-state reaction method. To
reduce the influence of the poor thermal stability of the Ag
2
O
C
compound, an oxygen partial pressure of 10 bar was used during
the formation of the present compounds. The XRD results of
three nominal compositions fired at 10001C for 6 h in 10 bar of
B
A
O
2
are shown in Fig. 3. For the nominal composition Ag0.5Bi4.5-
Ti O , all the reflection lines can be matched to the SrBi Ti O
4
phase (PDF 43-0973). In Ag0.5Bi4.5Ti O15, no silver metal was
4
obtained after the calcinations and sintering, and the powders or
pellets remained yellow. For the nominal composition Ag0.5-
Bi8.5Ti O , the single phase that we expected was not obtained.
4
15
4
15
7
27
The dominant phases were the two Aurivillius compounds:
Bi Ti 15; we observed only the shoulders
of some peaks related to Ag0.5Bi8.5Ti O , as shown in Fig. 3.
1
5
16
17
18
2
29
θ (degree)
30
31
4
3 4
O12 and Ag0.5Bi4.5Ti O
7
27
Fig. 1. X-ray diffraction results of the A–C nominal compositions
fired at 9001C in an O flow: (a) in the 2y range 101–601; (b) in the 2y
ranges 151–191 and 291–311. (A, Ag0.5Bi8.5Ti 27; B, Ag0.5Bi4.5Ti
C, AgBi Ti 18.)
For the nominal composition AgBi
5
Ti O
18, which is on the line
, the following
15, Bi Ti , and
5
2
between Ag0.5Bi4.5Ti 15 and Ag0.5Bi0.5TiO
4
O
3
7
O
4 15
O ;
three phases were detected: Ag0.5Bi4.5Ti
silver metal.
4
O
2
2 7
O
5
5
O
The phase constituents of the nominal compositions fired un-
der different conditions are shown in Table I. In the present
(
(
shown in Fig. 1 and Table I). An Aurivillius phase, Bi
4 3 12
Ti O
study, there was no evidence for the existence of AgBi Ti O
18
PDF72-1019), anda pyrochlorephase, Bi Ti (PDF89-4732),
2
O
2 7
5
5
(
n 5 5 in [Bi O ][A B O
n
]). An attempt to synthesize Na-
were observed in all the compounds. Because the strongest peak
of the Ag phase overlapped with the peak of the Bi Ti 12 phase,
2
2
nꢁ1
3n11
5
5 5 5 5
Bi Ti O18 and KBi Ti O18 by Subbarao and Uchida and Kiku-
4
3
O
1
0
chi was also unsuccessful. The maximum value of n in the
Aurivillius phase is supposed to be 4 when the A site is occupied
by a monovalent ion and a Bi ion. The existence of the mixed-
it was difficult to be certain from the XRD results whether Ag is
present or not. The existence of Ag was expected in all the com-
pounds because of the dark color of the samples. Additional lines
in the 2y ranges 16.81ꢁ171, 30.31ꢁ30.41, and 391ꢁ39.31 were
found in all the compounds. Because the additional lines may be
associated with the Na0.5Bi8.5Ti O phase (PDF 32-1044), the
7
layer Aurivillius phase Ag0.5Bi8.5Ti O27 was confirmed by the
XRD results when the nominal compositions were fired in an
oxygen flow. It was difficult to obtain the single-phase mixed-
7
27
layer structure of Ag0.5Bi8.5Ti
simpler three-layer structure of Bi
small concentration of Ag can be incorporated into the Bi4-
Ti O phase, and the Bi Ti O phase with dissolved Ag has a
7
O27 because of the presence of the
new phase was assumed to be Ag0.5Bi8.5Ti O .
7
27
Ti 12. It is believed that a
3
O
7
The amounts of Ag0.5Bi8.5Ti O27 phase can be estimated
4
from the strongest peak of each phase using the following
equation:
3
12
4
3
12
7
composition close to Ag0.5Bi8.5Ti O27. A slow heating rate or a
%
Ag0:5Bi8:5Ti7O27 ¼ 100ꢀI
=
prolonged dwelling time should be used to enhance the forma-
tion of the Ag0.5Bi8.5Ti O phase because of the higher free
energy of formation of Ag0.5Bi8.5Ti O compared with that of
7 27
118ðAg0:5Bi8:5Ti7O27Þ
Bi8:5Ti7O27ÞþI
7
27
½
I
1
18ðAg0:5
117ðBi Ti3O12Þ
4
þI222ðBi Ti2O7Þꢂ
4 3 7
Bi Ti O12. The structure of the Ag0.5Bi8.5Ti O27 phase was
2
refined in the orthorhombic space group I2 cm, suggested by
Boullay et al., Z51, and the lattice parameters are a55.451(1)
11
A similar equation is frequently used for calculating the
amounts of pyrochlore phase during the fabrication of
˚
A, b55.419(1) A, and c536.751(8) A.
˚
˚
Table I. Starting Nominal Compositions of Aurivillius Phases and Phases Detected by XRD
6
Fired in 10 Pa O
Nominal composition
Fired in O
2
flow (at 9001C for 12 h)
2
(at 10001C for 6 h)
Ag0.5Bi8.5Ti O
7
Bi Ti O 1Bi8.5Ag0.5Ti O 1Bi Ti O 1Ag
7
Bi Ti O 1Ag0.5Bi4.5Ti O 1Ag0.5Bi8.5Ti O
4 3 12 4 15 7 27
27
4
3
12
7
27
2
2
Ag0.5Bi4.5Ti O
4
Bi Ti O 1Bi Ti O 1Ag0.5Bi8.5Ti O 1Ag
7
Ag0.5Bi4.5Ti O
4 15
15
4
3
12
2
2
7
27
AgBi Ti O
5
Bi Ti O 1Ag0.5Bi8.5Ti O 1Bi Ti O 1Ag
2
Bi Ti O 1Ag0.5Bi4.5Ti O 1Ag
2 2 7 4 15
5
18
4
3
12
7
27
2
7
XRD, X-ray diffraction.

August 2007
Synthesis and Characterization of Aurivillius Phases
a)
2365
(
Bi Ti O
4
3
12
Bi Ti O
2
2
7
Ag0.5Bi8.5Ti O
7
27
a+b+a
a+b
a
A
B
10
20
30
40
50
60
1
5
16
17
18
29
θ (degree)
30
31
2
θ (degree)
2
(b)
Fig. 2. X-ray diffraction results of the nominal composition
Ag0.5Bi8.5Ti 27 in the 2y ranges 151–191 and 291–311. (a, firing at
01C/min to 9001C for 12 h; b, 101C/min to 6001C, followed by
7
O
1
0
.11C/min to 9001C and soaked for 6 h).
When fired under a pressure of oxygen, single-phase Ag0.5-
Bi4.5Ti O (n 5 4) could be obtained. The structure of the
4
15
Ag0.5Bi4.5Ti
4
O
15 phase was refined in the orthorhombic space
1
group A2
2
1
am, Z 5 2, and the cell parameters are a 5 5.4586(4)
˚
Fig. 4. (a) X-ray diffraction results of Ag0.5Bi4.5Ti
B, reheating single phase in an O flow at 9001C for 6 h); (b) Rietveld
refinement results of Ag0.5Bi4.5Ti O . (The upper graph shows the fit
4
O15 (A, single phase;
˚
˚
A (Rwp 5 0.1634)
A, b 5 5.4293(7) A, and c 5 40.758(3)
4 15
Fig. 4(b)). We also heated the single-phase Ag0.5Bi4.5Ti O
2
(
4
15
between the experimental and calculated patterns, while the lower graph
shows the difference between these two patterns.)
(a)
Bi Ti O
Ag0.5Bi4.5Ti O
4
15
4
3
12
Bi Ti O
Ag
in flowing oxygen at 9001C for 6 h. The dominant phase was
2
2
7
Ag0.5Bi8.5Ti O
7
27
still Ag0.5Bi4.5Ti O , but the reflection of Ag0.5Bi Ti O
4
15
8.5
7
27
appears at the shoulder of some peaks, and the pyrochlore
phase is observed, as shown in Fig. 4(a). We concluded that
Ag0.5Bi4.5Ti O is not stable when fired in flowing oxygen.
4
15
Dense ceramics (494% of the theoretical density) were
obtained when sintering Ag0.5Bi4.5Ti 15 at 10001C for 3 h in
pressurized O . After sintering, the ceramic shows the same
yellow color as the powders after calcinations. A dielectric con-
stant, e , of 166 and a dielectric loss, tand, of 0.004 were
obtained at 1 MHz at room temperature. From room temper-
ature to 3001C, the dielectric constant was found to be almost
temperature independent, as shown in Fig. 5. A sharp peak was
found at 6871C, which coincides with the minimum of the di-
electric loss. No significant difference was found for the value of
Tmax, whether the temperature-dependence measurement was
C
4
O
2
B
A
r
1
0
20
30
40
50
60
2θ (degree)
(b)
Bi Ti O
Ag0.5Bi4.5Ti O
4
15
4
3
12
Bi Ti O
3.0
2
2
7
Ag0.5Bi8.5Ti O
7
27
1
1
200
000
0
0
0
0
0
.5
.4
.3
.2
.1
(at 1MHz)
2.5
2.0
1.5
1.0
0.5
Tmax=687°C
C
B
A
8
6
4
00
00
00
0.0
6
50 700 750 800
Temperature (°C)
200
0
1
5
16
17
18
29
30
31
0.0
0
100 200 300 400 500 600 700
2θ (degree)
Temperature (°C)
Fig. 3. X-ray diffraction results of the A–C nominal compositions fired
at PO2 with 10 Pa: (a) in the 2y range 10–601; (b) in the 2y ranges
7
Fig. 5. Temperature dependence of the dielectric properties of the
Ag0.5Bi4.5Ti 15 ceramic. (The inset shows DSC analysis results; Tmax
the temperature of emax).
7 4
151–191 and 291–311. (A, Ag0.5Bi8.5Ti O27; B, Ag0.5Bi4.5Ti O15; C,
4
O
,
AgBi Ti 18.)
5
5
O

2
366
Journal of the American Ceramic Society—Hu et al.
IV. Conclusions
Vol. 90, No. 8
8
6
4
2
0
2
4
6
8
The nominal compositions of Aurivillius phases (Ag0.5Bi8.5-
Ti O , Ag0.5Bi4.5Ti O , and AgBi Ti O ) in the Bi–Ag–Ti–O
7
27
4
15
5
5
18
system were investigated by means of a solid-state reaction and
XRD. The mixed-layer Aurivillius phase Ag0.5Bi8.5Ti 27 was
found after firing in an O flow. A slow heating rate and a pro-
7
O
2
longed dwelling time can increase the amount of the Ag0.5Bi8.5-
Ti
1ꢀ 10 Pa), a single-phase compound of the four-layer Aur-
ivillius phase Ag Bi Ti O was obtained. There was no ev-
7
O27 phase. On firing in an oxygen partial pressure of 10 bar
6
(
−
80
−40
0
40
80
0
.5 4.5
4
15
−
−
−
−
idence for the existence of AgBi 18. The dielectric properties
(
5 5
Ti O
Electric Field (kV/cm)
at 1 MHz) of the Ag0.5Bi4.5Ti O15 phase were as follows:
4
Tmax 5 6871C, e 5 166 (B201C), and tand 5 0.004 (B201C).
r
References
1
B. H. Park, B. S. Kang, S. D. Bu, T. W. Noh, J. Lee, and W. Jo, ‘‘Lanthanum-
Substituted Bismuth Titanate for Use in Non-Volatile Memories,’’ Nature, 401,
4
Fig. 6. P–E hysteresis loops of Ag0.5Bi4.5Ti O15 ceramic with increasing
6
82–4 (1999).
D. Damjanovic, ‘‘Materials for High Temperature Piezoelectric Transducers,’’
2
electric field (at 100 Hz).
Curr. Opin. Solid State Mater. Sci., 3 [5] 469–73 (1998).
B. Aurivillius, ‘‘Mixed Bismuth Oxides With Layer Lattices: I. The Structure
3
2 2 9
Type of CaNb Bi O ,’’ Ark. Kemi, 1 [54] 463–80 (1949).
4
performed at a heating rate of 51 or 21C/min. Usually, Tmax can
be approximated as T . The DSC results shown in the inset
R. A. Armstrong and R. E. Newnham, ‘‘Bismuth Titanate Solid Solutions,’’
Mater. Res. Bull., 7, 1025–34 (1972).
C
5
E. C. Subbarao, ‘‘Crystal Chemistry of Mixed Bismuth Oxides With Layer-
Type Structure,’’ J. Am. Ceram. Soc., 45 [4] 166–9 (1962).
J. H. Park, P. M. Woodward, J. B. Parise, R. J. Reeder, I. Lubomirsky, and O.
graph of Fig. 5 suggest that the phase transition at the T
15 is not of the first order, but of the second order.
For bismuth layer-structured ferroelectrics, the TC is strongly
influenced by the r of the A-site cations, and bismuth layer-
structured ferroelectrics with smaller A-site cations tend to show
C
of
Ag0.5Bi4.5Ti
4
O
6
3
Stafsudd, ‘‘Synthesis, Structure and Dielectric Properties of (Bi1/2Ag1/2)TiO ,’’
Chem. Mater., 11 [1] 177–83 (1999).
i
7
Y. Inaguma, T. Katsumata, R. P. Wang, K. Kobashi, M. Itoh, Y. J. Shan, and T.
,’’
1
1
Nakamura, ‘‘Synthesis and Dielectric Properties of a Perovskite Bi1/2Ag1/2TiO
Ferroelectrics, 264, 1785–90 (2001).
3
a higher T
C i C
. The r of Ag is larger than that of Na , but the T
8
(
6871C) of the Ag .5Bi Ti O is higher than that (6621C in
0
4.5
4
15
P. E. D. Morgan, ‘‘Preparing New Extremely Difficult-to-Form Crystal Struc-
tures,’’ Mater. Res. Bull., 19 [3] 369–76 (1984).
1
3
14
Ahn et al. and 6551C in Ben Jannet et al. ) of the Na0.5Bi4.5-
Ti
9
1
S. L. Swartz and T. R. Shrout, ‘‘Fabrication of Perovskite Lead Magnesium
Niobate,’’ Mater. Res. Bull., 17 [10] 1245–50 (1982).
4
O15. The large electronegativity of the Ag (1.93) may play a
role in the influence on the Curie temperature (the electron-
10
K. Uchida and T. Kikuchi, ‘‘Subsolidus Phase-Equilibria in System Na
–TiO at 1000 Degrees,’’ J. Am. Ceram. Soc., 61 [1–2] 5–8 (1978).
Ph. Boullay, G. Trolliard, D. Mercurio, J. M. Perez-Mato, and L. Elcoro,
‘Toward A Unified Approach to the Crystal Chemistry of Aurivillius-Type Com-
2
O–
1
egativity of Na is 0.93). The TC of both the Ag0.5Bi4.5Ti O
4
15
Bi
2
O
3
2
1
1
4 4
and the Na0.5Bi4.5Ti O15 is higher than that of the SrBi4.5Ti O
15
1
5
16
‘
4 4
(5201C) and lower than that of the CaBi Ti O15 (8201C).
pounds. I. The Structural Model,’’ J. Solid State Chem., 164 [2] 252–60 (2002).
C. H. Hervoches, A. Snedden, R. Riggs, S. H. Kilcoyne, P. Manuel, and
P. Lightfoot, ‘‘Structural Behavior of the Four-Layer Aurivillius-Phase Ferroelec-
The P–E hysteresis loops of the Ag0.5Bi4.5Ti O ceramic
12
4
15
are shown in Fig. 6. The distorted loops indicate a lossy non-
linear capacitor behavior. Because both the polar orthorhombic
structure and the sharp peak from the dielectric constant
measurement suggest a ferroelectric phase at room temperature
trics SrBi
4
Ti
C. W. Ahn, I. W. Kim, M. S. Ha, W. K. Seo, J. S. Lee, and S. S. Yi,
‘Dielectric and Piezoelectric Properties of Lead-Free Na0.5Bi4.5ꢁxLa Ti 15 and
Na0.5Bi4.5ꢁxNd Ti 15 Ceramics,’’ Ferroelectrics, 273, 2639–44 (2002).
D. Ben Jannet, M. El Maaoui, and J. P. Mercurio, ‘‘Ferroelectric
4 5 3
O15 and Bi Ti FeO15,’’ J. Solid State Chem., 164 [2] 280–91 (2002).
13
‘
x
4
O
x
4
O
14
4
in Ag0.5Bi4.5Ti O15, we believe that the loops are the result
of both the weak ferroelectric properties and the leakage
current. The leakage–current density measured at 10 kV/cm
4 4 4
Versus Relaxor Behaviour in Na0.5Bi4.5Ti O15-BaBi Ti O15 Solid Solutions,’’
J. Electroceram., 11 [1–2] 101–6 (2003).
15
ꢁ7
2
Y. Noguchi, M. Miyayama, and T. Kudo, ‘‘Ferroelectric Properties of Inter-
growth Bi Ti 12–SrBi Ti 15 Ceramics,’’ Appl. Phy. Lett., 77 [27] 3639–41 (2000).
M. Demartin maeder, D. Damjanovic, and N. Setter, ‘‘Lead Free Piezoelectric
Materials,’’ J. Electroceram., 13, 385–92 (2004).
was 8 ꢀ 10 A/cm . When the soaking time was increased,
a larger P and a round curve were obtained because of the
leakage current.
4
3
O
4
4
O
16
r
&