Correlation between Phase Transitions and Piezoelectric Properties
in Lead-Free (K,Na,Li)NbO3–BaTiO3 Ceramics
Cheol-Woo AHN, Chee-Sung PARK, Dwight VIEHLAND, Sahn NAHM1,
Dong-Heon KANG2, Kyoo-Sik BAE2, and Shashank PRIYA
Center for Energy Harvesting Materials and Systems, Materials Science and Engineering, Virginia Tech, Blacksburg, VA 24061, U.S.A.
1Department of Materials Science and Engineering, Korea University, 1-5 Ka, Anam-dong, Sungbuk-ku, Seoul 136-713, Korea
2Department of Electronic Materials Engineering, Suwon University, Hwaseong, Gyeonggi-do 445-743, Korea
(Received August 1, 2008; accepted October 1, 2008; published online December 19, 2008)
In this manuscript, we report the polymorphic phase transitions, structural changes and piezoelectric properties of alkali
niobate based lead-free ceramics. The phase transitions were characterized as a function of alkali niobate content at room
temperature. The results clearly demonstrate that in this system high piezoelectric properties are achieved for a specific
fraction of ferroelectric orthorhombic (O) and tetragonal (T) phases. Using Rietveld and powder diffraction analysis, a
correlation was established among the piezoelectric response, the fraction of O and T phases, and KNN ratio for three different
systems of (K,Na)NbO3–BaTiO3 (KNN–BT), KNN–LiNbO3 (LN), and (K,Na,Li)NbO3 (KNLN)–BT.
[DOI: 10.1143/JJAP.47.8880]
KEYWORDS: lead-free piezoelectric ceramics, phase transition, (K,Na,Li)NbO3–BaTiO3
1. Introduction
Recently, lead-free piezoelectric ceramics have been
widely studied as alternatives to Pb(Zr,Ti)O3 (PZT).1–14)
Among the various possible candidates, (K,Na)NbO3 (KNN)
based materials have received considerable attention owing
to their excellent piezoelectric properties and high Curie
temperature (TC). The enhancement in piezoelectric proper-
ties has been previously associated to existence of morpho-
tropic phase boundary (MPB) in KNN based systems similar
to that of PZTs between orthorhombic (O) and tetragonal (T)
phase. However, this increase is correlated to the polymor-
phic phase transition rather than MPB.1–12) Recent work has
also shown that the enhancement in piezoelectric properties
is related to the coexistence of orthorhombic and tetragonal
phases at room temperature.13–15) Figure 1 illustrates the
variation of room temperature longitudinal piezoelectric
constant (d33) for KNN based ceramics as a function of O !
T phase transition temperature (TO{T) for solutions with
various substituents. The data in this figure scales well to
linear relationship with TO{T: d33,RT ðpC/NÞ ¼ 306:21
1:02TO{T. This indicates that higher piezoelectric properties
for KNN might be obtained by shifting TO{T towards room
temperature. However, this relationship is misguiding
because ceramic compositions with similar d33 values can
have different TO{T as shown in Fig. 1 for systems KNN–
BaTiO3, KNN–CaTiO3, and KNN–LiNbO3. This indicates
that the magnitude of piezoelectric response is related to
additional factors.
Here, we investigated two very important factors: namely,
(i) polymorphic phase transitions in KNN based systems at
room temperature, and (ii) correlation between the piezo-
electric response and fractional ratio of orthorhombic and
tetragonal phases. It has been suggested in the literature that
stable piezoelectric properties can be obtained until high
temperatures by shifting TO{T to just below room temper-
ature. Our results agree with this concept, but we illustrate
here a systematic methodology for tuning of TO{T. The
ceramic composition of (1 x)(K0:48Na0:48Li0:04)NbO3–
xBaTiO3 [(1 x)KNLN–xBT], investigated in this study,
was used as a model system to illustrate the findings.
2. Experimental Procedure
Ceramics in the system (1 x)(K0:48Na0:48Li0:04)NbO3–
xBaTiO3 were synthesized from oxides of >99% purity by
conventional solid-state route. The powders of K2CO3,
Na2CO3, Nb2O5, Li2CO3, BaCO3, and TiO2 (all obtained
from Alfa Aesar) were mixed for 24 h in a polypropylene jar
with zirconia balls. Mixed powders were dried and then
calcined at 950 C for 3 h. Calcined powders were milled for
48 h, dried and pressed into disks under pressure of 100 kgf/
cm2 and sintered in the range of 1070–1140 C for 2 h.
Figure 2(a) shows the X-ray diffraction (XRD) patterns
of (1 x)KNLN–xBT ceramics sintered at 1080 C for 2 h
(Philips Xpert Pro). All the peaks were indexed to the
perovskite structure. The inset of Fig. 2(a) shows that
tetragonality increases with increase of the BT ratio in
(1 x)KNLN–xBT ceramics. The coexistence of ortho-
rhombic and tetragonal phase was observed for ceramics
sintered at 1080 C in all of these compositions. In order
to observe phase variations with BT content, the ratios of
0 50 100 150 200 250
0
100
200
300
400
d 3
3
(p
C/
N)
TO-T (°C)
: KNN
: KNN-LiTaO3
: KNN-LiNbO3
: KNN-LiSbO3
: KNN-Li(Nb,Ta,Sb)O3
: KNN-BaTiO3
: KNN-CaTiO3
: KNN-LiNbO3-BaTiO3
Fig. 1. (Color online) Piezoelectric constants (d33) of KNN based ceram-
ics as a function of orthorhombic to tetragonal phase transition
(TO{T).2,4,6–15)
E-mail address: spriya@mse.vt.edu
Japanese Journal of Applied Physics
Vol. 47, No. 12, 2008, pp. 8880–8883
#2008 The Japan Society of Applied Physics
8880

tetragonal peak intensities were calculated by the sum of the
peak’s intensity marked in the inset of Fig. 2(a).
3. Results and Discussion
Figure 2(b) illustrates the variation of d33 with substitu-
tion fraction in KNN based ceramics. The fractional ratio of
tetragonal peak intensity with respect to the orthorhombic
one was calculated from the XRD patterns, while the
tetragonality of KNN–BT ceramics was measured by
Rietveld analysis as seen in our previous work.7) In this
study, we used the equation given as: FT ¼ SIT=ðSIT þ SIOÞ;
in order to calculate the approximate fraction of tetragonal
phase (FT), here SIT and SIO are the sums of peaks
intensities for tetragonal and orthorhombic phases. It was
found that the calculated values of fractional ratio using
XRD peak intensity showed a similar trend as that
determined from Rietveld analysis. Thus, an approximation
of the amount of the two phases (T and O) can be made by
using the XRD patterns shown in Fig. 2(a). The tetragonal
peak ratio was calculated to be 9.6% for pure KNN ceramics
and 88.5% for 0.9KNN–0.1BT ceramics. Three interesting
observations can be immediately made from Fig. 2(b) for
KNN–BT ceramics, which are: (i) d33 shows a maxima over
a narrow range of compositions; (ii) the maxima in d33
corresponds to a tetragonal peak ratio of 65–75%; and (iii)
the maxima in d33 corresponds to a fractional peak ratio of
70%. In conjunction with the results in Fig. 1, we can
conjecture that KNN–BT ceramics will have a maximum d33
coefficient at room temperature with tetragonal phase ratio
of 75%.
The variation of d33 with KNN ratio for the KNN–LN and
KNLN–BT systems is similar to that for KNN–BT and
explained in more detail in Fig. 2(c). In Li-modified
systems, the maximum in d33 occurs over a wider range of
tetragonal phase ratios. In the case of KNN–LN and KNLN–
BT, a maximum in d33 was obtained for a tetragonal phase
fraction of 50–70%. It is important to mention here that the
compositions in this wide range have different TO{T [as
shown in Fig. 2(d)], however the magnitude of d33 is quite
similar. Thus, based on the results of Figs. 2(b) and 2(c)
(c)
Rietveld Result
0 20 6040 80 100
0
50
100
150
200
250
300
0% T 96.1% T
100% T
% T
41.8% Td
33

(p
C/
N)
Ratio of Peak Intensity
[T/(T+O), %]
: KNN-BT
: KNN-LN
: KNLN-BT 71.6
T-rich
(d)
0
100
200
300
400
500
1.00 0.99 0.98 0.97 0.96 0.95 0.94 0.93
0
100
200
300
400
500
T O
-T

(°C
)
: KNN
: KNN-LN
: KNN-BT
: KNLN--BT
T C

(°C
)
KNN Ratio
1.00 0.98 0.96 0.94 0.92 0.90
0
20
40
60
80
100 0
20
40
60
80
100
50
100
150
200
250
Te
tr
ag
on
al
ity
(%
)
KNN Ratio
: KNN-LiNbO3
: KNN-BaTiO3
: KNLN-BaTiO3
R
at
io
o
f P
ea
k
In
te
ns
ity


[T
/(T
+O
), %
]d
33

(p
C/
N)
(b)
20 30 40 50 60
44 46 48
x=0.020
x=0.015
x=0.010
x=0.005
x=0.000
In
te
ns
ity
(A
rb
.U
nit
s)
2θ
(21
1) C
(21
0) C
(20
0) C
(11
1) C(1
10
) C
x=0.020
x=0.015
x=0.010
x=0.005
(10
0) C
x=0.000
In
te
ns
ity
(A
rb
.U
nit
s)
2θ
(a)
O
T
O: Orthorhombic
T: Tetragonal
Fig. 2. (Color online) (a) XRD patterns of (1 x)KNLN–xBT specimens sintered at 1080 C for 2 h, (b) the variations of d33 with
substitution rates, the fraction of tetragonal peaks intensities in XRD patterns, and the tetragonality measured by Rietveld analysis in
KNN based ceramics, (c) variation of d33 with fraction of tetragonal peak intensities in XRD patterns, and (d) the variations of TC
and TO{T in KNN based ceramics.7,8)
Jpn. J. Appl. Phys., Vol. 47, No. 12 (2008) C.-W. AHN et al.
8881