Scholarly article on topic '(K, Na)NbO3-based lead-free piezoceramics: Phase transition, sintering and property enhancement'

(K, Na)NbO3-based lead-free piezoceramics: Phase transition, sintering and property enhancement Academic research paper on "Materials engineering"

0
0
Share paper
Academic journal
J Adv Ceram
OECD Field of science
Keywords
{""}

Academic research paper on topic "(K, Na)NbO3-based lead-free piezoceramics: Phase transition, sintering and property enhancement"

Journal of Advanced Ceramics

2012, 1(1): 24-37

DOI: 10.1007/s40145-012-0003-3

Review

ISSN 2226-4108

(K, Na)NbO3-based lead-free piezoceramics: Phase transition,

sintering and property enhancement

Ke WANG, Jing-Feng LI

State Key Laboratory of New Ceramic and Fine Processing, Department of Materials Science and Engineering,

Tsinghua University, Beijing 100084, China

Received February 16, 2012; Accepted March 1, 2012 © The Author(s) 2012. This article is published with open access at Springerlink.com

Abstract: Most widely used piezoelectric ceramics are based on Pb(Zr,Ti)O3 (PZT) composition which has adverse environmental and health effects due to its high lead content. Environmental and safety concerns with respect to the utilization, recycling, and disposal of lead-based piezoelectric ceramics have induced a new surge in developing lead-free piezoelectric ceramics. Among all the lead-free ceramics, (K,Na)NbO3 (KNN) has drawn increasing attention because of its well-balanced piezoelectric properties and better environmental compatibility. On basis of the author's work, this review summarizes the progress that has been made in recent years on development of KNN-based piezoelectric ceramics, including crystallographic structure and phase transition analysis, pressurized solid-state sintering as well as liquid-phase-assisted sintering process, and poling treatment for property enhancement. All in all, KNN is a promising lead-free system, but more research is still required both from academic and industrial interests.

Key words: electrical properties; ferroelectric properties; piezoelectric properties; perovskites; niobates; lead-free

1 Introduction

In the 1880s, Pierre and Jacque Curie discovered that some crystalline materials, when compressed, produce a voltage proportional to the applied pressure. This characteristic is called piezoelectricity [1]. Although piezoelectricity is found in several types of natural materials, most modern devices use polycrystalline ceramics. Piezoelectric ceramics, which play an important role as functional electronic materials, have

* Corresponding author.

E-mail: jingfeng@mail.tsinghua.edu.cn

been widely used in various applications such as sensors, actuators transducers and so on. For decades, lead zirconate titanate (Pb(Zr1-xTix)O3 or PZT) ceramics has been market-dominating due to its excellent properties. Generally, the commonly used compositions are along a morphotropic phase boundary (MPB) around x = 0.48 separating tetragonal and rhombohedral phases, where dielectric, piezoelectric, and electromechanical coupling properties are optimized. The origin of the enhanced properties of PZT at the morphotropic phase boundary has been investigated in detail by many authors and is still a topic of intense interest. Several mechanisms have been invoked to explain the enhanced piezoelectric properties of PZT at MPB. However,

there is still no consensus in the literature on this issue. A widely accepted perspective believes that the increased properties are due to enhanced polarizability arising from the equivalent energy state between tetragonal and rhombohedral phases, which underlies optimal domain rotation during the poling process.

However, the large amount of lead contained in PZT materials has drawn much attention during the past decade, due to the environmental concern as well as government regulations against hazardous substances. In 2006, the European Union has already adopted the well-known directives Waste Electrical and Electronic Equipment (WEEE) and Restriction of the use of certain Hazardous Substances in electrical and electronic equipment (RoHS), with the purpose of protecting human health as well as environment by exclusion or substitution of hazardous substances used in electrical and electronic devices [2,3]. Similar regulations are also planned or established in North America, Japan, Korea and China. Lead ranks foremost among the hazardous substances list due to its notoriety, with maximum allowed concentration of mere 0.1 wt%. However, PZT, which contains more than 60 wt% lead oxides, is still allowed to be used in piezoelectric devices all over the world up to now, since no alternative lead-free counterparts is available. Therefore, tremendous efforts have been devoted to the development of competitive lead-free counterparts, such as BaTiO3, Bii/2Nai/2TiO3-based perovskites, bismuth layer-structured ferroelectrics (BLSF), and (K, Na)NbO3 (KNN)-based perovskites, and so on. At the moment, great embarrassment has been encountered in search for one single qualified lead-free system substituting PZT, which is so versatile in properties by utilization of various doping elements that it can fit almost all purposes. Therefore, it is reasonable to penetrate deeply into these various lead-free systems, and take advantage of them in distinct circumstances.

Among all the lead-free candidates, alkali niobate ceramics based on KNN have drawn much attention since Saito et al. made the breakthrough in Li, Ta, and Sb-modified KNN ceramics with textured structure [4]. Despite pure KNN ceramics is normally considered with a piezoelectric constant d33 as low as around 80 pC/N, the value reported by Saito et al. reached amazingly up to 416 pC/N, comparable to that of soft PZT ceramics. At first, it is believed that dopants like Li, Ta and Sb could consist in the constitution of MPB in KNN-based ceramics, which mimics that in PZT

ceramics and eventually results in the property enhancement. Thus, all kinds of dopants are exploited with special emphasis on the improvement of piezoelectric constant d33, which is usually measured using the convenient quasi-static method (Berlincourtmeter method) and taken as the small signal d33. Generally, the most effective approach for property enhancement of KNN-based ceramics is to form solid solutions with other species, such as LiNbO3 [5-12], LiTaO3 [7,13-20], LiSbO3 [21-23], BaTiO3 [24,25], CaTiO3 [26-30], (Bi0.5Na0.5)TiO3 [31-35], or a combination of multiple additives [30,36-51], and typical d33 values obtained are among 200~300 pC/N. Later on, it is proposed that the underlying mechanism for property enhancement of KNN-based ceramics by means of above-mentioned manners is not classical MPB but PPT, which is the abbreviation of polymorphism phase transition and exists in many materials such as BaTiO3 [15]. The key point of PPT theory claims that property enhancement in KNN-based ceramics is due to tetragonal-orthorhombic polymorphism phase transition point (7O-T) shifting downwards from around 200 °C to room temperature. However, MPB is actually more favorable over PPT since PPT has a problem of temperature instability [27,46]. As a result, the development of KNN-based ceramics is retarded for the time being, the situation of which has been encountered before by many great scientific innovations. Fortunately, several tentative ideas have already been brought out to further exploiting the potential of KNN-based materials as next generation lead-free piezoceramics [52-54].

Technically, it seems that the situation in KNN-based system is more complicated than that of traditional PZT series, which has invoked so many interesting scientific discussions. The authors have been working on KNN-based ceramics for almost 10 years, and would like to share these little tiny experiences with those who have already working on or will jump into this field. Certainly, there are already excellent reviews with regard to lead-free piezoceramics [55-60], but none of them is dedicated to KNN-based ceramics. The present review covers topics over a broad range, including crystallographic structure, sintering techniques and poling tricks. Nevertheless, up to now the perception of KNN-based materials is still not in-depth, thus divergence on these issues may be found in literature.

2 Polymorphism and Property Enhancement

KNN ceramics were investigated intensively several decades ago both in material properties and structures [61,62], as it was found as a piezoelectric material in as early as 1950s [63]. It is well known that the most interesting compositions are located with K/Na ratio approximately equal to 1:1 [63-65], and the KNN ceramics mentioned hereafter refer to these compositions if no specific explanations provided. KNN ceramics have an orthorhombic structure with space group Amm2 at room temperature; however, it is a little complicated because it holds an orthorhombic structure while the perovskite type ABO3 subcell possesses monoclinic symmetry [61]. Furthermore, there is no standard JAPDS-ICDD files for exact K0.5Na0.5NbO3 even up to now. As a result, crystallographic indexing is inconsistent in literature, especially for researchers mainly focusing on materials properties. So what really happens to the structure of KNN-based ceramics is not clear enough. Furthermore, the enhanced piezoelectric properties in the modified KNN were firstly attributed to the formation of the MPB similar to that observed in the PZT system [5,11,36,45], which was later on pointed out to be a different effect namely polymorphic phase transition (PPT) [15,22]. In any case, phase structure investigations in KNN-based ceramics are of paramount importance for obtaining excellent properties.

Generally, pure KNN ceramics have the following polymorphisms, low temperature (<123 °C) rhombo-hedral (R) phase, room temperature orthorhombic (O) phase, high temperature tetragonal (T, 200~410 C) and cubic (C, >410 C) phases [66]. Also, these phase transition temperatures can be tuned by introduction of dopants, such as Li, Ta, Sb, etc. In order to elucidate the crystalline structure of KNN and investigate the influence of dopants on the phase evolution as well as properties enhancement, (1-x)(K0.5Na0.5)NbO3-xLiNbO3 ceramics were taken as an example in the following context to correlate material structures with piezoelectric properties [67]. Hence, a better understanding of phase structure and its relation with properties in KNN-based ceramics may be attained, which could possibly benefit researchers for properties enhancement.

The structure illustration for room temperature KNN is shown in Fig. 1. Although the structure of KNN at room temperature is orthorhombic, the perovskite type ABO3 subcell, as shown in Fig. 1a,

(b) (c)

Fig. 1 The structure illustration for KNN of orthorhombic structure at room temperature (a) the perovskite type ABO3 subcell; (b) the projection of the subcell along b axis; (c) four adjacent subcell projections in an exaggerated manner to show the geometrical relationship, while omitting Nb and O atoms. (Reprinted with permission from Ref. [67]. © 2007, American Institute of Physics)

possesses monoclinic symmetry, with lattice parameters am = cm > bm while bm axis is perpendicular to amcm plane and angle P a little more than 90°. For the composition (K0.5Na0.5)NbO3, which is believed to lie along the orthorhombic-orthorhombic MPB and possess the best performance in KNbO3-NaNbO3 system, no superlattice reflections were found in previous XRD measurements [61], and thus no corner-linked O octahedra tilting should be taken into account [68]. Projection view of the subcell along bm axis is shown in Fig. 1b. Since angle Pis very close to 90°, it looks like that am axis is also perpendicular to cm axis. Therefore, in order to show the geometrical relationship more clearly, we could exaggerate angle P much more than 90°, and draw the projection of four adjacent perovskite subcells together, but omit Nb and O atoms, as shown in Fig. 1c. As am axis equals cm axis in length, the diagonals linked by dashed lines in Fig. 1c form a rectangle, which is the projection of unit cell of KNN along bm axis. Hence it could be easily understood that the perovskite type subcell of KNN is monoclinic while its unit cell has orthorhombic symmetry at room temperature. Furthermore, when temperature increases above 200 °C, a phase transition from orthorhombic to tetragonal occurs in KNN [69], which is similar to that of KNbO3, with space group

30 35 20 (°)

Fig. 2 The X-ray diffraction patterns of

(1 -x)(Ka5Naa5)NbO3-xLiNbO3 ceramics

changing from Amm2 to P4mm [70]. It is considered that dopants substitution, such as Li, in KNN would cause similar phase transition at room temperature. So in the tetragonal phase of Li-modified KNN, the perovskite type ABO3 subcell is the unit cell itself, with lattice parameters ct > at =bt and all three axes perpendicular with each other.

The above mentioned distinction between ortho-rhombic and tetragonal phases in KNN-based ceramics could be characterized by X-ray diffraction (XRD). The XRD patterns of (1-x)(K).5Na0.5)NbO3-xLiNbO3 ceramics are shown in Fig. 2, and the crystallographic indexing is according to the perovskite type ABO3 subcell, both for orthorhombic structure when x = 0 and tetragonal structure when x = 0.08, respectively. It is certain that the structure variation is due to the large distortion caused by doping Li element. The difference of peak profiles around 45° in X-ray diffraction pattern is usually used as an evidence for the phase transition in literature, though why these peaks would change in such a manner, or what really happens to the structures is hardly mentioned. The phase transition of (1-x)(K0.5Na0.5)NbO3-xLiNbO3 ceramics could be deduced from profiles of the N = 12 ({hkl} = {222}) and N =16 ({hkl} = {400}) line groups in XRD patterns, as shown in Fig. 3. The splitting of the N = 12 line group, as shown in Fig. 3a, indicates the existence of angle P in the monoclinic perovskite subcell. The two peaks around 84° in the XRD pattern when x = 0 tend to merge into one single peak with increasing Li amount, responding to the decrease in angle p. On the other hand, the N = 16 line group, as shown in Fig. 3b, shows the change of the three axes length in the subcell. It is obvious that in the orthorhombic phase when x=0, the intensity of the peak near 100° is larger than that of the one near 103°,

because am = cm > bm; while in the tetragonal phase the opposite situation occurs due to ct > at = bt. Therefore, when x = 0.06, the two contiguous peaks around 84° and the widely broadening peak around 99° obviously represent an intermediate state, showing the two phases co-existence condition. In addition, Raman spectroscopy is also an useful and convenient technique for phase identification of KNN-based piezoceramics [10,15,65,66,71-73].

As for property enhancement, shifting downward of tetragonal-orthorhombic transition (TO-T) point to room temperature by doping or forming solid solutions in KNN with other species do increase piezoelectric constant d33 to a level of 200~300 pC/N, while pure KNN ceramics suffering a relative low d33 around 80 pC/N. The fatal drawback is temperature instability [27, 46], which is absent in PZT systems due to existence of genuine MPB other than PPT. Apart from concerns about temperature instability of properties, it is noted that for identical or similar compositions of KNN-based ceramics, obtained properties could vary a lot among different research groups, which is possibly due to the exact location of TO-T. Recently, it is reported

Fig. 3

N = 16

20 (°)

The X-ray diffraction patterns of N = 12 and groups in (1-x)(K0.5Na0.5)NbO3-xLiNbO3

ceramics

that the TO_T could even be influenced as much as tens of degrees due to intergranular stress [73]. Besides, property enhancement in KNN-based ceramics appears to come a halt, which is actually the bottleneck effect of PPT. As a result, designing a PZT-like MPB in KNN systems seems more practical. The success of constructing a tricritical triple point in BaTiO3 system [74,75], as well as similar attempt in KNN-based ceramics [54], could probably offer fresh and innovative ideas for the future.

3 Sintering techniques

3.1 Spark plasma sintering (SPS)

Sintering has been among the most active topics for KNN-based materials ever since it was first discovered half a century ago [63,76,77], due to intrinsic volatility of alkaline species. Thus, normal sintering (in contrast to pressure-assisted sintering like hot-pressing or spark plasma sintering) of KNN-based materials often has a chance to encounter with low density and poor microstructure, which definitely degrades piezoelectric performance. Due to the poor sinterability and high volatility of KNN, pressure-assisted sintering techniques have been utilized for pursuing high density. Jaeger and Egerton [76] firstly fabricated highly dense KNN ceramics using hot pressing and reported that the hot-pressed samples possessed high piezoelectric constants and Curie temperature. Recently, spark plasma sintering (SPS) has been increasingly used instead of conventional hot pressing because of its advantages of a rapid heating rate and short soaking time. Our group has done a series of systematic studies on SPS of KNN-based ceramics, with special emphasis on property enhancement [78-83].

Sodium potassium niobate ceramics with the nominal composition of K0.5Na0.5NbO3 were fabricated successfully using SPS at a low temperature of 920 C. By post-annealing treatment in air, the SPS-processed K0.5Na0.5NbO3 ceramic samples showed typical ferroelectric and piezoelectric characteristics. The room temperature dielectric constants at 1, 10, 100, and 1000 kHz are 606, 584, 559, and 571, respectively, with dielectric loss values being in the range of 0.026 ~ 0.044. Although the grain sizes are small comparing with normally sintered samples (see Fig. 4), being on the order of 200 ~500 nm, the resultant Kx5Na0.5NbO3 ceramics show a considerably high d33 of 148 pC/N. It is suggested that the enhancement is the result of

extrinsic contributions to the polarizability associated with submicron grain sizes [55]. Similar result was also found in fine-grained BaTiO3 ceramics [84].

In addition to pure KNN samples, Li and Ta co-doped KNN ceramics were also investigated by SPS method [81]. Dense and fine-grained (Nao.535Ko.485)i-x-Lix(Nbo.8Tao.2)O3 (x=0.02~0.07) (abbreviated as NKLxNT) lead-free piezoceramics were fabricated by SPS at around 900 °C, followed by post-annealing in air. All the NKLXNT ceramics showed single perovskite structures with a phase transition from an orthorhombic symmetry to a tetragonal one across a composition region of x=0.04~0.05. The sample with a composition of x=0.05 had the maximum values of piezoelectric coefficient (d33=243 pC/N) and planar electromechanical coupling coefficient (¿p=46.1%), and other good properties such as Qm = 85, £r = 1240, and tan S= 0.023. Because of the enhanced densifica-tion and refined microstructure, the SPS-processed NKLxNT ceramics also had better fracture strength

2 i-im

Fig. 4 SEM micrographs of (a) SPS and (b) normally sintered (K0.5Na0.5)NbO3 ceramics

than that prepared by normal sintering.

3.2 Normal solid state sintering

Normal sintering is cost-effective and simple, which is of great attraction for mass production. So, investigation on normal sintering of KNN never ceases. It was found that in normal sintering of KNN-based ceramics, density increased greatly within a narrow temperature range but turned to decrease when the sintering temperature slightly exceeded the optimal one. Piezoelectric properties also showed similar relationship between the density and sintering temperature, but the highest piezoelectric strain coefficients were obtained at the temperatures lower than that for the highest density [85,86].

Figure 5 shows the change in the sintered densities of the KNN, Li-KNN (LKNN) and Li/Ta-KNN (LKNNT) samples as a function of sintering temperature. It was found that the change in density was very sensitive to the sintering temperature. Even though the sintering temperature was changed within a narrow range, the resultant density changed significantly for all the compositions. The density increased rapidly and then decreased slowly after reaching a maximum. Ta doping apparently shifted the maximum sintering point to higher temperature, whereas the densification behavior was almost the same both for the samples without and with Li doping. The highest density values obtained for the three compositions exceeded 95% of their theoretical densities. Fig. 5 also shows the changes in piezoelectric coefficient d33 of the three compositions as a function of sintering temperature. It is interesting that similar tendencies were obtained for the relationship between d33 and sintering temperature as the case of sintered density. All the three compositions showed their peak d33 values at their corresponding

1050 1100

Temperaure (°C)

Fig. 5 Density and piezoelectric coefficient d33 of KNN, LKNN and LKNNT samples as a function of sintering temperature

sintering temperatures. However, it must be noted that the temperature where the maximum d33 value was obtained differed from the point where the peak density was achieved. For example, the highest density was obtained at 1120 C for the LKNNT composition, whereas d33 of the resultant ceramics sample decreased apparently from the peak point if sintered at 1120 C; the highest d33 value was obtained at the temperature which is about 20~30 °C lower than that where the highest density can be reached. In other words, the d33 values dropped apparently when sintered at the temperatures exceeding the points that correspond to the peak density. On the other hand, the d33 values of the samples were relatively high, in spite of their low density when sintered at the temperatures below the point corresponding to the peak density.

The main question about the above results is why the optimal sintering temperature for the electrical properties differed from that for the density? As shown above, the piezoelectric and dielectric properties changed with a similar tendency as the relationship between density and sintering temperature. However, the sintering temperatures where the highest d33 values were obtained, were not consistent with that where the highest density was reached. To answer this question, we must take into account the volatilization of alkali components during the sintering, which occurred during the sintering of KNN-based ceramics and caused the composition deviation from the starting one. The volatilization may be accelerated when the liquid phase appears, although which also enhances the densification at the same time. Our study by inductively coupled plasma (ICP) analysis confirmed the serious volatilization above 1000 C, as shown in Fig. 6. Considering that the electric properties are sensitive to the composition change [6,79,87,88], and

Temperaure (°C) Fig. 6 Weight loss of alkaline elements in KNN-based ceramics as a function of sintering temperature

that the sample sintered to the highest density may not have the optimal composition, it is easy to understand why the highest piezoelectric and dielectric properties were obtained at somewhat lower sintering temperature than that for density. In addition, grain coarsening or abnormal grain growth occurred when the sintering temperature exceeded the peak point for the density, which may also not be favorable for the optimization of electrical properties [89].

3.3 Low-temperature liquid-phase sintering

Technically, the low-temperature sintering mentioned here is still in the discussion range of normal sintering. However, the densification mechanism during sintering is different due to emergency of a liquid phase, and the low-temperature sintering is greatly effective for property enhancement. As a result, we take out this topic as a separate section. As explained in the previous section, the properties of KNN-based ceramics are extremely sensitive to the processing conditions, especially the sintering temperature. The sintering temperatures are usually very high as compared with the melting points in the phase diagram, easily causing severe volatilization of alkali elements (see Fig. 6), and resulting in a composition deviation and hence property degradation. For example, in our previous study on Li-doped KNN ceramics, the d33 value decreases from a peak (314 pC/N) to about 220 pC/N if sintered at a temperature just 20 C higher than the optimized one (1060 C) for the KNN-ceramics containing 5.8 mol% LiNbO3 [6]. Therefore, it is necessary to develop low-temperature sintering processes for KNN-based ceramics. Also, low-temperature sintering could satisfy the demand for co-firing KNN-based ceramics with metal electrodes for device applications. Recently, our study revealed that Li-modified KNN lead-free piezoceramics can be sintered at a temperature as low as 950 C, whose d33 value is up to 280 pC/N [12,90]. This result shows that low temperature sintering and high performance could be achieved together in high-performance LiNbO3-KNN lead-free piezoceramics, which are of great potential for industrial applications. It is believed that the success of low-temperature sintering is related to the excess amount of Na2O added intentionally.

The addition of excess amount of Na2O could effectively reduce the sintering temperatures of Li-modified KNN ceramics. Figure 7 shows the shrinkage curve of (1-x)(Na0.535K0.48)NbO3-xLiNbO3 at x = 0.08, which clearly demonstrates that the sample

started to shrink at the temperature as low as 800 C, while the shrinkage rate had an abrupt change at 930 C . The nearly vertical shrinkage rate curve after 930 C suggests an unusually fast densification process, which is due to the formation of liquid phase. Although Na2O was added with excess amount, the effective sintering aid may be involved with other species. It was reported that A2O (A = Li, Na, K) could all form eutectic phases with Nb2O5, for mole ratio A:Nb more than 1:1, with melting points around 800 C; while A2CO3 (A = Li, Na, K), some of which probably remained in the matrix after calcination at 750 C since extra Na2CO3 was added initially, also has melting points lower than 950 °C. Considering that Li+, Na+, and K+ could all fill into the A site of perovskite subcell, the actual liquid phase may be eutectic phases formed between any of A2O (A = Li, Na, K) and Nb2O5, or remnant carbonates A2CO3 (A = Li, Na, K). The inner figures in Fig. 7 shows the SEM micrographs of (1-x)(Na0.535K0.4s)NbO3-xLiNbO3 with x = 0.08 sintered at 900 C and 950 °C, respectively. The average grain size of the sample sintered at 900 C is about 1-2 |m, and only a few of grains are adhered together with a low density of 3.95 g/cm3. The loose microstructure indicates no appearance of liquid phase at 900 C . By contrast, when the sample was sintered at 950 C, the microstructure changed significantly, with some large quadrate grains of 20 to 40 | m and an enhanced density of 4.36 g/cm3. Within a merely 50 C difference in sintering temperatures, the grain size increased from 1~2 |m to several tens of microns, confirming the effectiveness of liquid phase sintering. It has been recognized that the driving force for sintering is the decrease of surface energy of whole system. As for this case of liquid phase sintering, three

Temperaure (°C) Fig. 7 Shrinkage curve and SEM micrographs of 0.92 (Na0.535K048)NbO3 -0.08LiNbQ3

Table 1 Properties and phase structures of 0.92(Na0.535K0.48)Nb03-0.083LiNb03 samples

sintered at different temperatures

Temp. ( °C) 850 900 930 950 970 1000 1050 1100

p (g/cm3) 3.40 3.96 4.27 4.38 4.32 4.16 4.13 4.06

p, relative (%) 75.4 87.8 94.7 97.1 95.8 92.2 91.6 90.0

d33 (pC/N) 135 195 230 280 200 125 110 100

Phase structure* O_O-T O-T O-T O-T_T_T_

* O is abbreviation for orthorhombic, and T is for tetragonal. "-" means ambiguous phase identification due to emergence of impurities

distinct stages have been classified [91]: (1) particle rearrangement, (2) solution-precipitation process, and (3) coalescence process. The morphology of the 950 C-sintered sample demonstrates that it was at the final stage of liquid phase sintering, with substantially dense micro- structure and large faceted grains exhibiting shape accommodation characteristic for maximum density.

Table 1 shows the sintered density and piezoelectric coefficient d33 as a function of sintering temperature for the sample of x=0.080. The density increases significantly as the sintering temperature is increased from 850 to 950 C, and reaches a peak, then decreases gradually with further increasing sintering temperature. The bulk density (about 4.38 g/cm3) obtained at 950 C is higher than 95% of the theoretical density. The densification behavior can be also confirmed by the microstructure observation. This result is surprising because low-temperature sintering was realized by simply adding excessive Na2O, which is important for co-firing KNN-based multilayer with the Ag/Pd electrode. Although the low-temperature sintering mechanism is not yet clearly understood, liquid-phase sintering should be responsible for the densification. The grain growth behavior seems to be a clear indicator supporting our consideration about liquidphase sintering, because the grain sizes increased exceptionally as the sintering temperature was raised from 900 to 950 C . The density decreases when sintered above 950 C probably because the volatilization of alkali oxides is accelerated, which is not in favor of the formation of liquid phase. In fact, the weight loss at 950 C and 1050 C were 0.5% and 1.3%, respectively. In addition to the highest density, a peak d33 was obtained when sintered at 950 C for the x=0.080 composition, also because it consisted of orthorhombic and tetragonal phases. It should be noted that for the same x=0.080 composition, the resultant phase structure also changed from orthorhombic to

tetragonal with increasing sintering temperatures, as revealed in Table 1, which is the result of volatilization of alkaline species.

4 Poling tricks

It is known that poling is an important process to endow ferroelectric ceramics with macroscopic piezoelectric responses, by applying a high electric field up to several thousand voltages per millimeter. Inferior ferroelectric response is obtained by insufficient poling, which is caused by various factors such as defects. The effect of poling conditions on the dielectric and piezoelectric properties of KNN-based ceramics has been a subject of recent studies [92-97], most of which pointed out that poling temperature is of significant importance for property enhancement. Nevertheless, poling process could also be utilized to manipulate interior dipole movement, which might be tuned to favorable piezoelectric performance [52,53].

For a compositionally optimized Li-doped composition, its piezoelectric coefficient d33 can be increased up to 324 pC/N even from a considerably high value (190 pC/N) by means of re-poling treatment after room temperature aging; such a high d33 value is only reachable in literature for KNN ceramics with complicated modifications using Ta and Sb dopants [46,98,99]. Table 2 shows the piezoelectric coefficient d33 as well as tetragonal-orthorhombic phase transition temperature (TO-T) of Li-doped KNN ceramics with respect to Li content, after both the first and second poling, respectively. Even though the sintering temperature was as low as 950 C, a high piezoelectric constants d33 up to 280 pC/N was obtained at an optimal composition of x=0.083 after the first poling. The basic mechanism for piezoelectric property enhancement in the present study was generally considered due to two-phase coexistence [15], which was denoted by a downward of TO-T to around room

Table 2 Piezoelectric properties and tetragonal-orthorhombic phase transition temperature (TO-T) of (1-x)(K05Na05)NbO3-xLiNbO3 after both the first and second poling

x 5.8 6.5 7 7.5 7.8 8 8.3 8.5

1st poled d33 (pC/N) 125 130 145 170 185 190 280 180

2nd poled d33 (pC/N) 128 135 170 230 260 324 255 145

TO-T (C) 155 115 105 90 68 60 43 27

temperature, as also listed in Table 2. In spite of the considerably high piezoelectric performance of Li-doped KNN ceramics after the first poling treatment, the improvement of d33 after the second poling process is much more attractive, with a peak d33 of 324 pC/N at x=0.080, as shown in Table 2. It should be noted that more and more increment was made in d33 after the second poling with increasing Li content, with a maximum of 134 pC/N (from 190 to 324 pC/N) at x = 0.080. The maximum increment rate was amazingly more than 70%, even though no special treatment was given to the samples but just re-poling after aging for two months. However, the sample with an excellent d33 = 280 pC/N at x = 0.083 after the first poling showed no increase in d33 after re-poling two months later, but its d33 turned to decrease by 25 pC/N. It seems that aging and re-poling could not increase d33 for the x ^ 0.083 samples. In view of the TO-T and XRD measurements, it could be easily concluded that samples with coexisting phases at room temperature are favorable for the aging and re-poling induced enhancement of piezoelectricity, while those dominated by orthorhombic (with TO-T more than 120 C ) or tetragonal phases are neither sensitive nor profited. For the ferroelectric properties, it was found that the second poling resulted in an increasing Pr from 6.78 to 14.46 |C/cm2, and a reduced Ec from 22.52 to 17.58 kV/cm, respectively. The striking increase of d33 in the present study was due to large improvement of Pr [100].

As mentioned above, the increase of d33 was due to large improvement of Pr, which was accomplished by domain switching. Domain switching does not change the crystalline structure of the ceramics, but do make striking influences on macroscopic orientations of local atoms arrangement. XRD technique can provide accurate information concerning these variations by relative peaks intensity change, then domain evolution processes can be deduced. Figure 8 shows the XRD patterns of (222) and (004) lattice plane series for the x=0.080 sample, which were taken both before and

after the first poling, as well as after the second poling, respectively. The peak indexing is adopted for a tetragonal phase. These high-angle X-ray diffractions offer phase information more precisely than the low-angle ones usually employed in literatures [67].

As shown in Fig. 8, for the as-sintered sample, higher intensity of the (400)/(040) peak over (004) peak is a clear symbol for the tetragonal phase since c > a = b; however, slight split of the (222) peak, and over-lifted ka2 line of (004) peak indicate the presence of the orthorhombic phase (a = c > b with 90° < ¡5< 91°). After the first poling treatment, intensity contrast between (400)/(040) peak and (004) peak reverses, which is easily understood for the tetragonal phase due to an increase of c-domains. Besides, intensity increment of (004) peak after poling is also observed for the orthorhombic phase, despite the fact that <001> is not the spontaneous polarization direction. The increased intensity of (004) peak in orthorhombic phase can be easily deduced as a collateral consequence of domain rotation towards <110> directions. It should be emphasized that, significant

Fig. 8 Patterns of (222) and (004) lattice plane series for the sample 0.92(Naa535K0.48)NbO3-0.08LiNbO3, both before and after the first poling, as well as after the second poling, respectively. The Ps evolution inside domains is listed aside, assuming that the electric field during poling is parallel to <001> direction

changes, represented by distinct split of (222) peak and almost disappearance of (400)/(040) peak, took place after the second poling. The deduced Ps evolution inside domains is listed aside in Fig. 8, assuming that the electric field during poling is parallel to the <001> direction. The Ps parallel to <100> and <010> directions after the first poling, switched to <001> direction after the second poling, which indicates that the non-switched 90° domains successfully rotated and kept constant by means of the aging and re-poling process. A mechanism corresponding to details of spontaneous polarization change in the domain level was proposed, concerning the combined effect of migration of oxygen vacancies, and interaction between defect dipoles and spontaneous polarization inside domains, which can be found elsewhere in detail [52].

It is widely accepted that poling of KNN-based ceramics is usually companied with problems, such as large leakage current, easy breakdown, etc.; however, the above results demonstrate that poling is also a chance which could be resorted for property enhancement. If poling condition could be optimized for a specific composition, very high d33 can be obtained without complicated doping compositions.

5 Summary

KNN-based ceramics has been investigated as a potential candidate for a new group of lead-free piezoelectric materials, and recent advances related to phase transitions, sintering techniques and property enhancement are discussed. Despite inferior piezoelectric response than PZT, KNN system is still favored by high Curie temperature, complete environment-friendliness, etc. The most urgent issue concerning development of KNN-based materials should be a breakthrough, both theoretically reasonable and experimentally testified, in property enhancement in the absence of PPT effect, such as real PZT-like MPB design. Then problems like temperature instability could be solved thoroughly. Also, secondary considerations, such as poling conditions, processing optimization, domain influences, etc., would be favored as well, due to request both practically and scientifically. Moreover, from the viewpoint of industrial application, d33 is not the only figure of merit. Actually, preliminary reports on piezoelectric devices utilizing KNN-based materials already exhibit

promising results, including high-frequency ultrasonic transducers [101-103], surface acoustic wave (SAW) devices [29] and so on [104-107]. Although it is possible that KNN cannot carry out the historical assignment to substitute PZT everywhere, it must find suitable opportunities to fulfill itself eventually.

Open Access This article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.

Acknowledgement

The authors appreciate the support by Tsinghua University Initiative Scientific Research Program and National Nature Science Foundation of China (Grant Nos. 50921061 and 51028202).

References

[1] Jaffe B, Cook WR, Jaffe H. Piezoelectric Ceramics. New York: Academic Press, 1971.

[2] Eu-Directive 2002/96/Ec: Waste electrical and electronic equipment (WEEE). Off J Eur Union 2003, 46: 24-38.

[3] Eu-Directive 2002/95/Ec: Restriction of the use of certain hazardous substances in electrical and electronic equipment (RoHS). Off J Eur Union 2003, 46: 19-23.

[4] Saito Y, Takao H, Tani T, et al. Lead-free piezoceramics. Nature 2004, 432: 84-87.

[5] Guo YP, Kakimoto K, Ohsato H. Phase transitional behavior and piezoelectric properties of (Naa5Ka5)NbO3-LiNbO3 ceramics. Appl Phys Lett

2004, 85: 4121-4123.

[6] Zhao P, Zhang BP, Li JF. High piezoelectric d33 coefficient in Li-modified lead-free (Na,K)NbO3 ceramics sintered at optimal temperature. Appl Phys Lett 2007, 90: 242909.

[7] Hollenstein E, Davis M, Damjanovic D, et al. Piezoelectric properties of Li- and Ta-modified (Ka5Naa5)NbO3 ceramics. Appl Phys Lett 2005, 87: 182905.

[8] Matsubara M, Yamaguchi T, Kikuta K, et al. Effect of Li substitution on the piezoelectric properties of potassium sodium niobate ceramics. Jpn J Appl Phys

2005, 44: 6136-6142.

[9] Du HL, Tang FS, Liu DJ, et al. The microstructure and ferroelectric properties of (K0.5Na0.5)NbO3-

LiNbO3 lead-free piezoelectric ceramics. Mater Sci Eng B-Solid State Mater Adv Technol 2007, 136: 165-169.

[10] Klein N, Hollenstein E, Damjanovic D, et al. A study of the phase diagram of (K,Na,Li)NbO3 determined by dielectric and piezoelectric measurements, and raman spectroscopy. J Appl Phys 2007, 102: 014112.

[11] Song HC, Cho KH, Park HY, et al. Microstructure and piezoelectric properties of (1-x)(Naa5K05)NbO3-xLiNbO3 ceramics. J Am Ceram Soc 2007, 90: 1812-1816.

[12] Wang K, Li J F, Liu N. Piezoelectric properties of low-temperature sintered Li-modified (Na, K)NbO3 lead-free ceramics. Appl Phys Lett 2008, 93: 092904.

[13] Guo YP, Kakimoto K, Ohsato H. (Na0.5K0.5)NbO3-LiTaO3 lead-free piezoelectric ceramics. Mater Lett 2005, 59: 241-244.

[14] Zhao P, Zhang BP, Li JF. Enhancing piezoelectric d33 coefficient in Li/Ta-codoped lead-free (Na,K)NbO3 ceramics by compensating Na and K at a fixed ratio. Appl Phys Lett 2007, 91: 172901.

[15] Dai YJ, Zhang XW, Zhou GY. Phase transitional behavior in K05Na05NbO3-LiTaO3 ceramics. Appl Phys Lett 2007, 90: 262903.

[16] Kim MS, Jeong SJ, Song JS. Microstructures and piezoelectric properties in the Li2O-excess 0.95(Na05K05)NbO3-0.05LiTaO3 cramics. J Am Ceram Soc 2007, 90: 3338-3340.

[17] Lin D, Kwok KW, Chan HLW. Microstructure, phase transition, and electrical properties of (K0.5Na0.5)i-xLix(Nbi-yTay)O3 lead-free piezoelectric ceramics. J Appl Phys 2007, 102: 034102.

[18] Chang YF, Yang ZP, Ma DF, et al. Phase transitional behavior, microstructure, and electrical properties in Ta-modified [(K0.458Na0.542)0.96Li0.04]NbO3 lead-free piezoelectric ceramics. J Appl Phys 2008, 104: 024109.

[19] Wu L, Zhang JL, Shao SF, et al. Phase coexistence and high piezoelectric properties in (K0.40Na0.60V96Li0.04Nb0.80Ta0.20O3 ceramics. J Phys D: Appl Phys 2008, 41: 035402.

[20] Zuo RZ, Ye C, Fang XS, et al. Processing and piezoelectric poperties of (Na0.sK0.5)0.96Li0.04(Ta0.1Nb0.9)1-xCuxO3-3x/2 lead-free ceramics. J Am Ceram Soc 2008, 91: 914-917.

[21] Zang GZ, Wang JF, Chen HC, et al. Perovskite (Naa5K05)1-x(LiSb)xNb1-xO3 lead-free piezoceramics. Appl Phys Lett 2006, 88: 212908.

[22] Zhang SJ, Xia R, Shrout TR, et al. Piezoelectric properties in provskite 0.948(K05Na05)NbO3-0.052LiSbO3 lead-free ceramics. J Appl Phys 2006, 100: 104108.

[23] Wu JG, Xiao DQ, Wang YY, et al. Effects of K content on the dielectric, piezoelectric, and ferroelectric properties of 0.95(KxNa1-x)Nb03-0.05LiSbO3 lead-free ceramics. J Appl Phys 2008, 103: 024102.

[24] Park HY, Ahn CW, Song HC, et al. Microstructure and piezoelectric poperties of 0.95(Nao.5Ka5)NbO3-0.05BaTi03 ceramics. Appl Phys Lett 2006, 89: 062906.

[25] Lin D, Kwok KW, Chan HLW. Structure, dielectric, and piezoelectric properties of CuO-doped K05Na05Nb03-BaTi03 lead-free ceramics. J Appl Phys 2007, 102: 074113.

[26] Park HY, Cho KH, Paik DS, et al. Microstructure and piezoelectric properties of lead-free (1-x)(Na0 5K0 5)Nb03-xCaTi03 ceramics. J Appl Phys 2007, 102: 124101.

[27] Zhang SJ, Xia R, Shrout TR. Modified (K05Na05)Nb03 based lead-free piezoelectrics with broad temperature usage range. Appl Phys Lett 2007, 91: 132913.

[28] Chang RC, Chu SY, Lin YF, et al. The effects of sintering temperature on the properties of (Na05K05)Nb03-CaTi03 based lead-free ceramics. Sensor ActuatA-Phys 2007, 138: 355-360.

[29] Chang RC, Chu SY, Lin YF, et al. An investigation of (Na05K05)Nb03-CaTi03 based lead-free ceramics and surface acoustic wave devices. J Eur Ceram Soc 2007, 27: 4453-4460.

[30] Wu JG, Xiao DQ, Wang YY, et al. Improved temperature stability of CaTi03-modified [(K0.5Na0.5)0.96]Li0.04](Nb0.91Sb0.05Ta0.04)03 lead-free piezoelectric ceramics. J Appl Phys 2008, 104: 024102.

[31] Zuo RZ, Fang XS, Ye C. Phase structures and electrical properties of new lead-free (Na05K05)Nb03-(Bi0.5Na0.5)Ti03 ceramics. Appl Phys Lett 2007, 90: 092904.

[32] Zuo RZ, Fang XS, Ye C, et al. Phase transitional behavior and piezoelectric properties of lead-free (Na05K05)Nb03-(Bi0.5K0.5)Ti03 ceramics. J Am Ceram Soc 2007, 90: 2424-2428.

[33] Zuo RZ, Ye C. Structures and piezoelectric properties of (NaKLi)1-x(BiNaBa)xNb1-xTix03 lead-free ceramics. Appl Phys Lett 2007, 91: 062916.

[34] Du HL, Zhou WC, Luo F, et al. Structure and electrical properties' investigation of (K0.5Na0.5)Nb03-(Bi0.5Na0.5)Ti03 lead-free piezo electric ceramics. J Phys D: Appl Phys 2008, 41: 085416.

[35] Du HL, Zhou WC, Luo F, et al. Polymorphic phase transition dependence of piezoelectric properties in

(K0.5Na0.5)NbO3-(Bi0.5K0.5)TiO3 lead-free ceramics. J Phys D: ApplPhys 2008, 41: 115413.

[36] Chang YF, Yang ZP, Hou YT, et al. Effects of Li content on the phase structure and electrical properties of lead-free (K0.46-x72Na0.54-x/2Lix)-(Nb0 76Ta0 20Sb0 04)O3 ceramics. Appl Phys Lett 2007, 90: 232905.

[37] Chang YF, Yang ZP, Wei LL. Microstructure, density, and dielectric properties of lead-free (K0.44Na0.52Li004) (Nb096-xTaxSb004)O3 piezoelectric ceramics. J Am Ceram Soc 2007, 90: 1656-1658.

[38] Hagh NM, Jadidian B, Safari A. Property-processing relationship in lead-free (K,Na,Li)NbO3 solid solution system. JElectroceram 2007, 18: 339-346.

[39] Lin DM, Kwok KW, Chan HLW. Phase structures and electrical properties of K0.5Na0.5(Nb0.925Ta0.075)O3-LiSbO3 lead-free piezoelectric ceramics. J Phys D: Appl Phys 2007, 40: 6060-6065.

[40] Lin DM, Kwok KW, Chan HLW. Effects of BaO on the structure and electrical properties of 0.95K0.5Na0.5(Nb0.94Sb0.06)O3-0.05LiTaO3 lead-free ceramics. J Phys D: Appl Phys 2007, 40: 6778-6783.

[41] Qi P, Wang JF, Ming BQ, et al. Phase transition and high piezoactivity of Sb doped (Nao.53K0.435Li0.035) Nb0.94Ta0.06O3 lead-free ceramics. Chinese Phys Lett 2007, 24: 3535-3538.

[42] Rubio-Marcos F, Ochoa P, Fernandez JF. Sintering and properties of lead-free (K,Na,Li)(Nb,Ta,Sb)O3 ceramics. J Eur Ceram Soc 2007, 27: 4125-4129.

[43] Wu JG, Wang YY, Xiao DQ, et al. Effects of Ag content on the phase structure and piezoelectric properties of (K0.44-xNa0.52Li0.04Agx)-(Nb09iTa0.05Sb0.04)O3 lead-free ceramics. Appl Phys Lett 2007, 91: 132914.

[44] Wu JG, Xiao DQ, Wang YY, et al. Effects of K/Na ratio on the phase structure and electrical properties of (KxNa0.96-xLi0.04)(Nb0.91Ta0.05Sb0.04)O3 lead-free ceramics. Appl Phys Lett 2007, 91: 252907.

[45] Yang ZP, Chang YF, Wei LL. Phase transitional behavior and electrical properties of lead-free (K0.44Na0.52Li0.04)(Nb0.96-xTaxSb0.04)O3 piezoelectric ceramics. Appl Phys Lett 2007, 90: 042911.

[46] Akdogan EK, Kerman K, Abazari M, et al. Origin of high piezoelectric activity in ferroelectric (K0.44Na0.52Li0.04)(Nb0.84Ta0.1Sb0.06)O3 ceramics. Appl Phys Lett 2008, 92: 112908.

[47] Juan D, Wang JF, Zang GZ, et al. (Na0.52K0.44Li0.04)Nb0.9-xSbxTa0.1O3 lead-free piezoelectric ceramics with high performance and high Curie temperature. Chinese Phys Lett 2008, 25: 1446-1448.

[48] Lin DM, Kwok KW, Lam KH, et al. Phase structure and electrical properties of K0.5Na0.5(Nb0.94Sb0.06)O3-LiTaO3 lead-free piezoelectric ceramics. J Phys D: Appl Phys 2008, 41: 052002.

[49] Wu JG, Peng T, Wang YY, et al. Phase structure and electrical properties of (K0.48Na052)(Nb0.95Ta0 05)O3-LiSbO3 lead-free piezoelectric ceramics. J Am Ceram Soc 2008, 91: 319-321.

[50] Zhao P, Zhang BP, Tu R, et al. High piezoelectric d33 coefficient in Li/Ta/Sb-codoped lead-free (Na,K)NbO3 ceramics sintered at optimal temperature. J Am Ceram Soc 2008, 91: 3078-3081.

[51] Zuo RZ, Xu ZK, Li LT. Dielectric and piezoelectric properties of Fe2O3-doped (Na0.5K).5)0.96Li0.04Nb0.86-Ta04Sb0 04O3 lead-free ceramics. J Phys Chem Solids 2008, 69: 1728-1732.

[52] Wang K, Li JF. Domain engineering of lead-free Li-modified (K,Na)NbO3 polycrystals with highly enhanced piezoelectricity. Adv Funct Mater 2010, 20: 1924-1929.

[53] Wang K, Li JF, Zhou JJ. High normalized strain obtained in Li-modified (K,Na)NbO3 lead-free piezoceramics. Appl Phys Express 2011, 4: 061501.

[54] Zuo RZ, Fu J. Rhombohedral-tetragonal phase coexistence and piezoelectric properties of (NaK)(NbSb)O3-LiTaO3-BaZrO3 lead-free ceramics. J Am Ceram Soc 2011, 94: 1467-1470.

[55] Shrout TR, Zhang SJ. Lead-free piezoelectric ceramics: alternatives for PZT? J Electroceram 2007, 19: 111-124.

[56] Zhang SJ, Xia R, Shrout TR. Lead-free piezoelectric ceramics vs. PZT? J Electroceram 2007, 19: 251-257.

[57] Takenaka T, Nagata H, Hiruma Y, et al. Lead-free piezoelectric ceramics based on perovskite structures. J Electroceram 2007, 19: 259-265.

[58] Rödel J, Jo W, Seifert KTP, et al. Perspective on the development of lead-free piezoceramics. J Am Ceram Soc 2009, 92: 1153-1177.

[59] Panda PK. Review: Environmental friendly lead-free piezoelectric materials. J Mater Sci 2009, 44: 5049-5062.

[60] Safari A, Abazari M. Lead-free piezoelectric ceramics and thin films. IEEE Trans Ultrason Ferroelectr Freq Control 2010, 57: 2165-2176.

[61] Tennery VJ, Hang KW. Thermal and X-ray diffraction studies of NaNbO3-NbO3 system. J Appl Phys 1968, 39: 4749-4753.

[62] Ahtee M, Hewat AW. Structural phase-transitions in sodium-potassium niobate solid-solutions by neutron powder diffraction. Acta Crystallogr A 1978, 34: 309-317.

[63] Egerton L, Dillon DM. Piezoelectric and dielectric properties of ceramics in the system potassium sodium niobate. J Am Ceram Soc 1959, 42: 438-442.

[64] Dai YJ, Zhang XW, Chen KP. Morphotropic phase boundary and electrical properties of K1-xNaxNb03 lead-free ceramics. Appl Phys Lett 2009, 94: 042905.

[65] Zhou JJ, Li JF, Wang K, et al. Phase structure and electrical properties of (Li,Ta)-doped (K,Na)Nb03 lead-free piezoceramics in the vicinity of Na/K = 50/50. J Mater Sci 2011, 46: 5111-5116.

[66] Trodahl HJ, Klein N, Damjanovic D, et al. Raman spectroscopy of (K,Na)Nb03 and (K,Na)1-xLixNb03. Appl Phys Lett 2008, 93: 3.

[67] Wang K, Li JF. Analysis of crystallographic evolution in (Na,K)Nb03-based lead-free piezo-ceramics by X-ray diffraction. Appl Phys Lett 2007, 91: 262902.

[68] Ahtee M, Glazer AM. Lattice-parameters and tilted octahedra in sodium-potassium niobate solidsolutions. Acta Crystallogr A 1976, 32: 434-446.

[69] Shirane G, Newnham R, Pepinsky R. Dielectric properties and phase transitions of NaNb03 and (Na,K)Nb03. Physical Review 1954, 96: 581-588.

[70] Hewat AW. Cubic-tetragonal-orthorhombic-rhombohedral ferroelectric transitions in perovskite potassium niobate - neutron powder profile refine ment of structures. J Phys C: Solid State Phys 1973, 6: 2559-2572.

[71] Kakimoto K, Akao K, Guo YP, et al. Raman scattering study of piezoelectric (Na0.5K0.5)Nb03-LiNb03 ceramics. Jpn J Appl Phys 2005, 44: 7064-7067.

[72] Kano J, Sasanuma K, Tsukada S, et al. Piezoelectric (Na0.5K0.5)Nb03-SrTi03 ceramics in the tetragonal-orthorhombic phase boundary studied by Raman spectroscopy. Ferroelectrics 2007, 347: 55-59.

[73] Zhou JJ, Li JF, Zhang XW. 0rthorhombic to tetragonal phase transition due to stress release in (Li,Ta)-doped (K,Na)Nb03 lead-free piezoceramics. J Eur Ceram Soc 2012, 32: 267-270.

[74] Liu W, Ren X. Large piezoelectric effect in Pb-free ceramics. Phys Rev Lett 2009, 103: 257602.

[75] Xue D, Zhou Y, Bao H, et al. Large piezoelectric effect in Pb-free Ba(Ti,Sn)03-x(Ba,Ca)Ti03 ceramics. Appl Phys Lett 2011, 99: 122901.

[76] Jaeger RE, Egerton L. Hot pressing of potassium-sodium niobates. J Am Ceram Soc 1962, 45: 209213.

[77] Haertling GH. Properties of hot-pressed ferroelectric alkali niobate ceramics. J Am Ceram Soc 1967, 50: 329-330.

[78] Li JF, Wang K, Zhang BP, et al. Ferroelectric and

piezoelectric properties of fine-grained Nao.5K0.5NbO3 lead-free piezoelectric ceramics prepared by spark plasma sintering. J Am Ceram Soc 2006, 89: 706-709.

[79] Zhang BP, Li JF, Wang K, et al. Compositional dependence of piezoelectric properties in NaxK1-xNbO3 lerad-free ceramics prepared by spark plasma sintering. J Am Ceram Soc 2006, 89: 1605-1609.

[80] Wang K, Zhang BP, Li JF, et al. Lead-free Na0.5K0.5NbO3 piezoelectric ceramics fabricated by spark plasma sintering: Annealing effect on electrical properties. JElectroceram 2008, 21: 251-254.

[81] Shen ZY, Li JF, Wang K, Electrical and mechanical properties of fine-grained Li/Ta-modified (Na,K)NbO3-based piezoceramics prepared by spark plasma sintering. J Am Ceram Soc 2010, 93: 1378-1383.

[82] Liu N, Wang K, Li JF, et al. Hydrothermal synthesis and spark plasma sintering of (K,Na)NbO3 lead-free piezoceramics. J Am Ceram Soc 2009, 92: 1884-1887.

[83] Zhen Y, Li JF, Wang K, et al. Spark plasma sintering of Li/Ta-modified (K,Na)NbO3 lead-free piezoelectric ceramics: Post-annealing temperature effect on phase structure, electrical properties and grain growth behavior. Mater Sci Eng B-Adv Funct Solid-State Mater 2011, 176: 1110-1114.

[84] Arlt G, Hennings D, Dewith G. Dielectric-properties of fine-grained barium-titanate ceramics. J Appl Phys 1985, 58: 1619-1625.

[85] Zhen YH, Li JF. Normal sintering of (K,Na)NbO3-based ceramics: influence of sintering temperature on densification, microstructure, and electrical properties. J Am Ceram Soc 2006, 89: 3669-3675.

[86] Li JF, Zhen YH, Zhang BP, et al. Normal sintering of (K,Na)NbO3-based lead-free piezoelectric ceramics. Ceram Int 2008, 34: 783-786.

[87] Shen ZY, Zhen YH, Wang K, et al. Influence of sintering temperature on grain growth and phase structure of compositionally optimized highperformance Li/Ta-modified (Na,K)NbO3 ceramics. J Am Ceram Soc 2009, 92: 1748-1752.

[88] Shen ZY, Wang K, Li JF. Combined effects of Li content and sintering temperature on polymorphic phase boundary and electrical properties of Li/Ta Co-doped (Na, K)NbO3 lead-free piezoceramics. Appl Phys A 2009, 97: 911-917.

[89] Zhen YH, Li JF. Abnormal grain growth and new core-shell structure in (K,Na)NbO3-based lead-free piezoelectric ceramics. J Am Ceram Soc 2007, 90: 3496-3502.

[90] Wang K, Li JF. Low-temperature sintering of Li-modified (K,Na)NbO3 lead-free ceramics: Sintering behavior, microstructure, and electrical properties. J Am Ceram Soc 2010, 93: 1101-1107.

[91] Kingery WD. Densification during sintering in the presence of a liquid phase. I. theory. J Appl Phys 1959, 30: 301-306.

[92] Yoo J, Lee K, Chung K, et al. Piezoelectric and dielectric properties of (LiNaK)(NbTaSb)O3 ceramics with variation in poling temperature. Jpn J Appl Phys 2006, 45: 7444-7448.

[93] Du HL, Tang FS, Luo F, et al. Effect of poling condition on piezoelectric properties of (K0.5Na0.5) NbO3-LiNbO3 lead-free piezoelectric ceramics. Mater Sci Eng B-Solid State Mater Adv Technol 2007, 137: 175-179.

[94] Du HL, Zhou WC, Luo F, et al. An approach to further improve piezoelectric properties of (K0.5Nao.5)NbO3-based lead-free ceramics. Appl Phys Lett 2007, 91: 202907.

[95] Ogawa T, Furukawa M, Tsukada T. Poling field dependence of piezoelectric properties and hysteresis loops of polarization versus electric field in alkali niobate ceramics. Jpn J Appl Phys 2009, 48.

[96] Rubio-Marcos F, Romero JJ, Ochoa DA, et al. Effects of poling process on KNN-modified piezo-ceramic properties. J Am Ceram Soc 2010, 93: 318-321.

[97] Morozov MI, Kungl H, Hoffmann MJ. Effects of poling over the orthorhombic-tetragonal phase transition temperature in compositionally homogeneous (K,Na)NbO3-based ceramics. Appl Phys Lett 2011, 98.

[98] Zuo RZ, Fu J, Lv DY. Phase transformation and tunable piezoelectric properties of lead-free (Na0.52K0.48-xlix)(Nb1-x-ySbyTax)O3 system. J Am

Ceram Soc 2009, 92: 283-285.

[99] Gao Y, Zhang J, Qing Y, et al. Remarkably strong piezoelectricity of lead-free (K0.45Na0.55)0.98LWNb0.77Ta0.18Sb0.05)O3 ceramic. J Am Ceram Soc 2011, 94: 2968-2973.

[100] Safari A, Akdogan EK. Piezoelectric and Acoustic Materials for Transducer Applications. New York: Springer, 2008.

[101]Zhen YH, Li JF, Wang K. Fabrication and electrical properties of fine-scale 1-3 piezoceramic/ epoxy composites using (K,Na)NbO3-based lead-free ceramics. Ferroelectrics 2007, 358: 1043-1050.

[102]Shen ZY, Xu Y, Li JF. Fabrication and electromechanical properties of microscale 1-3-type piezoelectric composites using (Na,K)NbO3-based Pb-free piezoceramics. J Appl Phys 2009, 105.

[103]Shen ZY, Li JF, Chen R, et al. Microscale 1-3-type (Na,K)NbO3-based Pb-free piezocomposites for high-frequency ultrasonic transducer applications. J Am Ceram Soc 2011, 94: 1346-1349.

[104]Shung KK, Cannata JM, Zhou QF. Piezoelectric materials for high frequency medical imaging applications: a review. J Electroceram 2007, 19: 139-145.

[105]Guo M, Lam KH, Lin DM, et al. A rosen-type piezoelectric transformer employing lead-free K0.5Na0.5NbO3 ceramics. J Mater Sci 2008, 43: 709-714.

[106]Hagh NM, Jadidian B, Ashbahian E, et al. Lead-free piezoelectric ceramic transducer in the donor-doped K12Nai/2NbO3 solid solution system. IEEE Trans Ultrason Ferroelectr Freq Control 2008, 55: 214-224.

[107]Kwok KW, Hon SF, Lin D. Lead-free self-focused piezoelectric transducers for viscous liquid ejection.

Sensor Actuat a-Phys 2011, 168: 168-171.