Effect of nano- and micron-sized K0.5Na0.5NbO3 fillers on the dielectric and piezoelectric properties of PVDF composites

Polymer nanocrystal composites were fabricated by embedding polyvinylidene fluoride (PVDF) with K0.5Na0.5NbO3 (KNN) nanocrystallites of different volume fraction using the hot-pressing technique. For comparison, PVDF–KNN microcrystal composites of the same compositions were also fabricated which facilitated the studies of the crystallite size (wide range) effect on the dielectric and piezoelectric properties. The structural, morphological, dielectric, and piezoelectric properties of these nano and micro crystal composites were investigated. The incorporation of KNN fillers in PVDF at both nanometer and micron scales above 10 vol% resulted in the formation of polar β-form of PVDF. The room temperature dielectric constant as high as 3273 at 100 Hz was obtained for the PVDF comprising 40 vol% KNN nanocrystallites due to dipole–dipole interactions (as the presence of β-PVDF is prominent), whereas it was only 236 for the PVDF containing the same amount (40 vol%) of micron-sized crystallites of KNN at the same frequency. Various theoretical models were employed to predict the dielectric constants of the PVDF–KNN nano and micro crystal composites. The PVDF comprising 70 vol% micron-sized crystallites of KNN exhibited a d33 value of 35 pC/N, while the nanocrystal composites of PVDF–KNN did not exhibit any piezoelectric response perhaps due to the unrelieved internal stress within each grain, besides the fact that they have less domain walls.

Amongst these, β-phase is polar, consisting of a planar zigzag all trans (TTTT) chain conformation with an orthorhombic structure (a = 8.58 Å, b = 4.91 Å, and c = 2.56 Å, space group Cm2m) [12,13]. It possesses spontaneous polarization and, as a result, exhibits piezoelectric and ferroelectric properties associated with high dielectric breakdown strength [14]. However, these properties are quite inferior to those of well known ferroelectric ceramics namely PZT, BaTiO 3 ,  K 0.5 Na 0.5 NbO 3 (KNN) crystallites at any length scales. Therefore, we thought that it was worth attempting to study the piezoelectric response of PVDF-KNN-based composites.
It is well known that the method of preparation of composites has a strong influence on their physical properties [18]. Several methods such as hot-pressing, compression molding, solvent casting, and tape casting (to fabricate in the form of thin sheets) besides the process associated with pellet shaped specimens were reported in the literature to fabricate several forms of polymer-based composites [17,18,[34][35][36][37][38][39][40]. Among them 0-3 composites are found to be easy to process and could be produced in large scale which consist of polymers and randomly distributed ceramic particles [5,37]. The method that was reported in the literature to fabricate PVDF comprising KNNLS, KNNLST, and K 0.48 Na 0.52 NbO 3 was based on cold-pressing which generally results in the formation of voids and pores within the composites that affect the physical properties [32,33]. However, there are no reports available in the literature to fabricate the above composites based on hot-pressing technique. The hot-pressing technique is expected to yield dense composites associated with uniform microstructures as compared to that of the other techniques [37,39]. Therefore, in the present work, we have adopted simple hot-pressing method to fabricate 0-3 type composites in the form of pellets by using crystalline KNN powder as fillers in the PVDF matrix. Interestingly, this method facilitated PVDF to crystallize in polar β-PVDF with the addition of KNN fillers at optimum pressure and temperature. Besides the phase transition behavior of PVDF, the effect of crystallite size and volume fraction on the dielectric and piezoelectric properties of PVDF-KNN composites were investigated. For this, nano and micro crystal composites were fabricated separately by dispersing KNN at nanometer and micron scales in the PVDF matrix. It is known in the literature that the volume fraction of filler plays a crucial role in obtaining high dielectric constant and large piezoelectric response [18,[41][42][43][44]. Therefore, we have varied the volume fraction of nano and micro crystalline KNN powder (as the case may be) in PVDF matrix in order to optimize the volume fraction that is required to obtain improved physical properties. In the case of microcrystal composites, we have varied the volume fraction from 10 to 70 vol%, whereas in the case of nanocrystal composites the volume was limited to 40 vol% due to difficulties involved in obtaining the samples in pellet form.

1 Preparation of KNN nanocrystalline powder
Fine nanocrystallites of KNN were synthesized via the citrate-assisted sol-gel method [45]. For this, analytical grade ammonium niobate (V) oxalate penta hydrate (NH 4 4 OH was added dropwise slowly to increase the pH up to about 8.5 and stirring continued for another 1 h in order to obtain a transparent viscous yellow niobate peroxo (V)-citrate solution. Sodium acetate (5 mmol) and potassium acetate (5 mmol) solutions were prepared separately using distilled water and these were added slowly to the above niobate peroxo (V)-citrate solution. The precursor solution was further heated at 60 ℃ and stirred for 4 h to get the water evaporated and obtain solid precursor powder. Nanocrystalline KNN powder was obtained by subjecting the precursor powder to calcination at 550 ℃ for 5 h.

2 Preparation of micron-sized KNN powder
Polycrystalline powder consisting of micron-sized crystallites of K 0.5 Na 0.5 NbO 3 (KNN) was synthesized by the conventional solid-state reaction route. Analytical grade K 2 CO 3 , Na 2 CO 3 , and Nb 2 O 5 were taken in a stoichiometric ratio and mixed with acetone and ball-milled for 5 h using zirconia balls to obtain a homogenous mixture. This mixture was dried and calcined at 900 ℃ for 5 h.

3 Preparation of PVDF-KNN nano and micro crystal composites
The PVDF with KNN microcrystal composites of different volume fraction (10-70 vol%) and PVDF nanocrystal composites comprising KNN nanocrystalline powder of 10-40 vol% were prepared by the hot-pressing technique. For this, the appropriate amounts of either micron-or nano-sized crystallites of KNN were added to the PVDF powder and mixed thoroughly using a mortar and pestle in an acetone medium for 1 h. These mixtures were dried at room temperature before making pellet shaped specimens. The dried mixtures were taken in a die of 10 mm in diameter and hot-pressed at optimized conditions. Nearly 1 mm thick samples of PVDF and PVDF with KNN (micron-or nano-sized crystallites) of different volume fraction were obtained. The densities of these pellets were measured by the Archimedes method using xylene as the liquid medium.

4 Characterization
The hot-pressed samples were subjected to X-ray powder diffraction (XRD, Pananalytical X-pert Pro) with Cu Kα radiation in the 2θ range of 20°-60°. A scanning electron microscope (SEM, FESEM, Inspect F50) was used to study the micro-structural aspects of the samples under investigation. In order to have an insight into the nature of dispersion of KNN in the PVDF matrix, cross-sectional SEM was carried out for fractured samples of PVDF-KNN nano and micro crystal composites. For making electrical measurements, the major surfaces of the pure KNN nanocrystalline ceramics and hot-pressed pellets were silver coated on both sides. The frequency dependent dielectric response was studied using an impedance analyzer (Agilent 4294A) by sweeping between 100 Hz and 100 MHz at room temperature. The piezoelectric coefficients of the samples were measured by the piezo-d 33 meter (KCF tech). The dielectric breakdown voltage measurements were carried out on the pellets as per the procedure outlined in ASTM D149. Figure 1 shows the XRD patterns obtained for pure KNN (nanocrystalline powder), hot-pressed pure PVDF, and PVDF with KNN nanocrystallites of different volume fraction (10-40 vol%). The XRD pattern obtained for pure KNN is indexed to the orthorhombic cell using ICSD data. The XRD pattern ( Fig. 1(b)) recorded for hot-pressed PVDF reveals sharp peaks at 2θ = 17.7° (100), 18.3° (020), 19.9° (110), 33.1° (130), and 35.8° (200) corresponding to the α-phase of PVDF, and the broad peaks at 2θ = 26.5° (022) and 38.6° (211) are assigned to the γ-phase of PVDF [11,[46][47][48][49][50][51]. In order to visualize the effect of the presence of nanocrystallites of KNN on the structural characteristics of PVDF, the XRD studies were carried out on PVDF comprising different volume fraction of KNN (Figs. 1(c)-1(f)). The PVDF embedded with KNN nanocrystalline powder below 20 vol% shows peaks corresponding to α-, γ-, and small traces of β-phases along with predominant peaks pertaining to KNN. Indeed, above 20 vol% of KNN nanocrystalline powder in the PVDF matrix shows interestingly a distinct diffraction peak at 2θ = 20.7° (200) corresponding to β-phase [8,46,47,51] in addition to KNN peaks ( Fig. 1(d)). This phase, being non-centrosymmetric, is expected to play a crucial role in facilitating these composites to exhibit piezoelectric and ferroelectric properties. The intensity of the peak corresponding to the β-phase (with respect to the αand γ-phases of PVDF) is predominant and found to increase with increase in the volume fraction (10-40 vol%) of nanocrystallites of KNN in the PVDF matrix. Interestingly, the peaks corresponding to α-and γ-phases of PVDF are absent in the XRD patterns obtained for the samples containing above 20 vol% of KNN. This result demonstrates that the presence of nanocrystallites of KNN in PVDF promotes strong electrostatic interaction between KNN at nanoscale and the PVDF polymer especially at the interface, leading to the alignment of the PVDF matrix in the extended TTTT conformation and resulting in the β-phase. Similar behavior is observed in the PVDF with various fillers such as BaTiO 3 , MWCNT, TiO 2 , and ferrite-based composites [41,[52][53][54][55]. It is known in the literature that high pressure favors β-phase formation of PVDF [10]. However, it is intriguing to note that the moderate pressure and temperature employed in the present study in the presence of nanocrystallites of KNN facilitate the formation of the β-phase of PVDF.

1 X-ray structural studies
In order to investigate and confirm the crystallite size effect of KNN on the transformation of α-phase to β-phase and its physical properties, systematic experiments were conducted using micron-sized KNN. Figure 2 shows the XRD patterns obtained for micron-sized crystallites of KNN and hot-pressed samples with micron-sized KNN of different volume fraction in the PVDF matrix. The peak corresponding to the β-phase is observed for PVDF containing 20-50 vol% of KNN (micron-sized crystalline powder) along with KNN peaks (Figs. 2(c)-2(f)). For the PVDF comprising 50 vol% KNN and above, the intensity of the β-phase peak is reduced drastically (Figs. 2(g) and 2(h)). Furthermore, the XRD patterns recorded show

2 Density and micro-structural studies
The densities of the PVDF-KNN nano and micro crystal composites were obtained both theoretically and experimentally, and these are depicted in Fig. 3. Theoretical values of PVDF-KNN nano and micro crystal composites were estimated using the mixture rule [18] given below: where  is the density of the composite, f It is known in the literature that the composites prepared through the hot-pressing technique give the maximum possible density [17,[37][38][39][40]. From the above equation, the composites with PVDF-KNN nanocrystals show the maximum relative density of 97% at lower volume fraction (10 vol%), whereas when the KNN nanocrystal content is increased from 10 to 20 vol%, the density decreases to 94% and remains almost constant above 20 vol%. This is attributed to the agglomeration of KNN nanocrystals at higher volume fraction (discussed in the later part). The microcrystal composites fabricated using the present method exhibit a maximum relative density of 96%-98% up to 50 vol%. For the KNN composition higher than 50 vol%, there is a sudden decrease in density. This is due to the poor dispersion of KNN microcrystals in the PVDF matrix at higher volume fraction of KNN micron-sized crystallites. The densities of the nano and micro crystal composites for all the composites under study were calculated by not taking the voids and defects into account. Figures 4 and 5 show the cross-sectional SEM micrographs recorded for nano and micro crystal composites, respectively. The SEM micrograph recorded for pure PVDF shows a smooth and continuous surface ( Fig. 4(a)), and the SEM images of pure KNN nano and micro crystalline powder show the presence of well developed grains in 40-90 nm and 0.8-4 µm size ranges respectively, exhibiting a cuboidal morphology. The SEM micrographs along with the grain size distribution obtained for these samples are depicted in Fig. S1 in the Electronic Supplementary Material (ESM). There is a uniform distribution of KNN nanocrystallites in the PVDF matrix at lower volume fraction loading. The agglomeration of the crystallites increases with the increase in the volume fraction (beyond 10 vol%) of nanocrystallites of KNN in the PVDF matrix. Similarly, Fig. 5 shows the SEM micrographs obtained for the PVDF with different volume fraction of micron-sized KNN crystallites, depicting a homogenous distribution of micron-sized KNN crystallites in the PVDF matrix at lower volume fraction of KNN. The agglomeration of these crystallites is not that pronounced as compared to that of PVDF-KNN nanocrystal composites at higher loadings of KNN (Figs. 5(b)-5(g)). The connectivity of the crystallites is found to increase with an increase in the volume fraction of KNN in the PVDF matrix. Figure 6 shows the room temperature frequency dependent dielectric constant ( r  ) and loss (D) for PVDF with KNN nanocrystallites of different volume fraction in the frequency range of 100 Hz-100 MHz (covered in the present study). There is a decrease in the dielectric constant with an increase in frequency of measurement. The dielectric constant for pure PVDF is   with the increase in volume fraction is attributed to high density and the presence of maximum amount of β-PVDF (which is polar) coupled with high dielectric constant KNN nanocrystallites. Similarly, the dielectric loss versus frequency is obtained and low frequency dispersion is found to increase with an increase in the volume fraction of KNN nanocrystallites (up to 30 vol%) in the PVDF matrix, above which there is a sudden drop in the loss value. The loss (D) versus frequency shows two relaxations: one at low frequency region and the other at higher frequency ( Fig. 6(b)). The relaxation at low frequency below 100 Hz is related to c  relaxation associated with molecular motion in the crystalline regions of PVDF, whereas above 10 6 Hz, a small relaxation peak is observed due to glass transition relaxation of the PVDF and is referred to as a  relaxation [46,47].

3 Dielectric properties
In order to study the effect of micron-sized crystallites of KNN on dielectric properties of PVDF, dielectric measurements were made at different frequencies for pure PVDF and PVDF with   (Fig. 6(c)). The dielectric constant decreases with an increase in frequency and, as expected, increases with an increase in the amount of KNN in the PVDF matrix. However, the value of the dielectric constant is low as compared to that of the PVDF-KNN nanocrystal composites. The increase in dielectric constant in the nanocrystal composites is chiefly attributed to an increase in dipole-dipole interaction between the PVDF polymer and the KNN nanocrystals at the interface and also due to ion-dipole interaction between the surface of the nanocrystallites and the PVDF matrix [18,[54][55][56][57][58].
It is well known that, as the size of KNN crystallites decreases, there is a tendency of the nanocrystal to form agglomerates which results in the formation of continuous random cluster [56]. Therefore, one would expect an increase in strong dipole-dipole interaction between the large aggregates of KNN nanocrystallites and PVDF polymer at the interface resulting in an enhanced dielectric constant [56]. For instance, the value of r  for 10 vol% of micron-sized crystallites of KNN in the PVDF matrix is around 19 at 100 Hz and increases to 236 for 40 vol% loading, which is significantly low as compared to that of PVDF nanocrystal composites at 40 vol% loading. On further increasing the volume fraction of KNN to 70 vol%, r  reaches a value as high as 1096. The increase in r  for 70 vol% of PVDF-KNN microcrystal composite is mainly attributed to the increase in the volume fraction of the KNN micron-sized crystallites in the PVDF matrix vis-à-vis connectivity and strong dipole-dipole interaction. This attribution is evidenced by SEM studies (Fig. 5(g)). The room temperature values of dielectric constant ( r  ) obtained at 10 kHz for PVDF-KNN microcrystal composites are summarized in Table 1. The value of loss (D) observed in the present case is around 1.5 at 100 Hz for 40 vol% loading (Fig.  6(d)), which is quite low as compared to that of their nanocrystal composite counterpart (at 40 vol%) due to less agglomeration of crystallites.
To rationalize the dielectric properties of current  composites, various models were employed to predict their effective dielectric constants. The effective dielectric constant of the composite depends on the dielectric constant of the polymer, filler, filler size, filler shape, its volume fraction and also depends mainly on the dielectric constant of the interphase region, and the volume of the interphase region of the composite [18,46,47]. Therefore, to predict the dielectric constant of the PVDF-KNN nano and micro crystal composites, various models were used and these are given below. According to the Maxwell and Garnett model, the effective dielectric constant ( r  ) of a composite comprising randomly dispersed crystallites having spherical morphology associated with a relatively high dielectric constant [42,59] is given by where 1  is the dielectric constant of the polymer (dielectric constant of PVDF, 1  = 13.26 at 10 kHz), 2  is the dielectric constant of the filler (dielectric constant of KNN nanocrystals, 2  = 385 at 10 kHz (Fig. S2 in the ESM) and for micron-sized KNN crystallites, it is 199 [60]), and q is the volume fraction of the filler.
The second model was proposed by Furukawa, in which, it was assumed that the filler crystallites are spherical and uniformly dispersed in the polymer matrix. It was proposed that the dielectric constant of the composite mainly depends on the dielectric constant of the matrix [44,61]. The effective dielectric constant in this model is given by where 1  is the dielectric constant of the polymer matrix and q is the volume fraction of the spherical crystallites.
The third model that was used to predict the dielectric constant is due to Rayleigh [42], developed using the Maxwell-Garnett and Furukawa models for composite containing minor spherical fillers. The effective dielectric constant ( r  ) is given by In order to account for the local variations of the electric field and the interaction with the applied electric field, the Bhimasankaram-Suryanarayana-Prasad (BSP) model was used. This model is very effective especially for the composite with a large fraction of piezoelectric crystallites [62]. The effective dielectric constant ( r  ) is given by Apart from other models, the logarithmic mixture rule, the effective medium theory (EMT), and the Yamada model were also used to predict the dielectric constant of the composites under study using Eqs. (6), (7), and (8), respectively [42]: where 1  and 2  are volume fractions of PVDF and KNN nano-/micron-sized fillers (as the case may be) respectively, 1  and 2  are the dielectric constants of PVDF and KNN nano-/micron-sized fillers respectively, and n is the fitting parameter or the morphology factor. The EMT model is similar to the Maxwell-Garnett theory proposed for filler particles which are homogenously distributed, non-interacting and roughly spherical, and are surrounded by a concentric matrix layer. However, the Yamada model [63] has an advantage as it helps in predicting the dielectric constant of the composite based on the dielectric constant of the individual materials and it also considers a factor n = 4π/m, where n is related to the shape parameter or fitting parameter attributed to the shape of the filler particles and m is the relative orientation of the filler [42]. The dielectric constant ( r  ) of the composite according to the Yamada model is as follows: Using the above models, the effective dielectric constants of PVDF-KNN nano and micro crystal composites for different volume fraction of KNN were estimated and are depicted in Fig. 7 along with the corresponding experimental values. Figure 7(a) clearly shows that the models described in Eqs. (2)-(5) are ineffective for the PVDF with KNN nano-sized crystallites as the experimental values are not in agreement with those predicted by these models. However, the experimental dielectric constant values of nanocrystal composites are very close to those predicted by Yamada and EMT models, especially at higher concentrations of KNN nanocrystals. In the case of microcrystal composites, the values obtained by the models (2)-(5) are in good agreement for lower volume fraction (< 40 vol%), whereas for higher volume fraction, the experimental values are in deviation. Indeed, the observed dielectric constants are in close agreement with those predicted by the Yamada model at higher volume fraction of KNN micron-sized crystallites in PVDF matrix (above 40 vol%) by taking into account the shape factor. The shape factor or fitting parameter is different for different particle size [18]. The shape factor reflects asymmetry, orientation, and distribution of the ceramic particles in the matrix. In the case of PVDF-KNN nanocrystal composites, n value is found to be high (n = 110) as compared to that of PVDF-KNN microcrystal composites (n = 16.85) as obtained from the best fit. This could be due to non-uniform distribution and/or clustering of KNN nanocrystallites in the PVDF matrix as compared to that of PVDF-KNN microcrystal composites [64].

4 Piezoelectric and breakdown voltage studies
The crystallite size and volume fraction dependent piezoelectric properties were probed for PVDF-KNN nano and micro crystal composites. All the samples under study were poled at 3 kV/mm at room temperature and the piezoelectric coefficients were determined. It is well known in the literature that the piezoelectric coefficient has strong dependence on crystallite size [18,43,65]. Therefore, we have compared the results of the piezoelectric coefficient (d 33 ) obtained for hot-pressed PVDF with nano-and micron-sized crystallites of KNN at 40 vol% (exhibiting better piezoresponse). In the present study, the PVDF with 40 vol% of KNN nanocrystals does not show any piezoelectric response. However, the 40 vol% of KNN micron-sized crystallites dispersed in PVDF exhibits the interesting piezoelectric response (d 33 ) of 10 pC/N. The reason for PVDF-KNN nanocrystal composites (below 100 nm) not exhibiting piezoelectric properties is owing to the fact that these contain only a few domain walls, thus contribution from domain wall motion through poling is difficult due to internal clamping by space charge layer at the interface (between the matrix and the KNN nanocrystallites). Besides, the nanocrystallites have a tendency to agglomerate, which may give rise to an additional domain wall pinning effect at the grain boundaries of nanocrystals within the agglomerates [66]. Also, the size of the agglomerates increases with an increase in the volume fraction of nanocrystallites of KNN in the PVDF matrix. It may create a conduction path, where these space charges will percolate easily, resulting in high leakage current [67] and hindering the piezoelectric response of the nanocrystal composites.
In the case of the microcrystal composites, the agglomeration of crystallites is less pronounced and the surface area contact between the micron-sized KNN crystallites and the PVDF is large. Therefore, the poling of microcrystal composites would be more efficient than nanocrystal composites due to the increased probability of a large number of micron-sized crystallites getting poled and hence results in high d 33 value [18] . In order to measure the piezoelectric response at different volume fraction, micron-sized KNN amounts are varied in different volume fraction (10-70 vol%) in PVDF matrix. The PVDF with micron-sized crystallites of KNN at lower volume www.springer.com/journal/40145 317 fraction (below 40 vol%) does not show significant piezoelectric response. When the volume fraction is increased to 40 vol% and above in PVDF, it exhibits enhanced piezoelectric response. The piezoelectric coefficient (d 33 ) is found to increase with an increase in the volume fraction of the micron-sized KNN crystallites in PVDF matrix. The PVDF microcrystal composite at 40 vol% shows a d 33 value of 10 pC/N and reaches a maximum value of 35 pC/N for 70 vol% of KNN in the PVDF matrix (Fig. 8). The increase in the d 33 coefficient with the increase in KNN content is attributed to an increase in the connectivity of crystallites, coupled with an enhanced degree of domain reorientation through poling [33].
In addition, in order to have a gross idea about the performance of piezo active PVDF-KNN microcrystal composites for energy harvesting applications, a figure of merit (FOM) is estimated using the following expression [33]: The figure of merit is found to increase with the increase in volume fraction of KNN up to 60 vol% in the PVDF matrix, and beyond this, there is a sudden drop in the value. This could be due to poor dispersion at higher volume fraction of KNN micron-sized fillers in the PVDF matrix and also due to a high dielectric constant ( r  ) value. Similar behavior was reported in the case of PVDF with 80 vol% of KNNLST composites [33].
The PVDF with different volume fraction of microcrystalline KNN (40-70 vol%) were subjected to breakdown tests according to ASTM D149 standard to estimate breakdown voltages and their dielectric strengths. These studies were carried out in air ambience. The sample was placed between two cylindrical electrodes (both top and bottom) of 6 mm in diameter and the AC (50 Hz) voltage was continuously increased at a rate of 500 V/s till the sample broke down [47]. The breakdown voltages V (kV) of the samples were recorded and the dielectric strengths E (kV/mm) were calculated as E = V/t, where t is the thickness of the sample in millimeter [47]. As expected, the breakdown voltage is found to decrease with an increase in volume fraction of KNN in the PVDF matrix. The breakdown voltage at 40 vol% of KNN in PVDF shows 8.56 kV and decreases to a value of 4.7 kV for 70 vol% of micron-sized crystallites. This is ascribed to charge leakage in the composites. It is known in the literature that addition of fillers to the polymer introduces defects in the sample which causes charge leakage and reduces the dielectric strength [67][68][69]. The dielectric strength, breakdown voltage, figure of merit, and their corresponding piezoelectric coefficient for PVDF-KNN microcrystal composites are summarized in Table 1.

Conclusions
X-ray structural studies carried out on KNN added PVDF confirmed the transformation of PVDF to polar β-phase; this was particularly effective when KNN was at nanoscale. More than 10 fold enhancement in the dielectric properties was achieved in 40 vol% KNN nanocrystallites added PVDF ( r  = 3273 at 100 Hz) as compared PVDF-KNN microcrystal composites. This was attributed to an increase in dipole-dipole interaction and ion-dipole interaction between the β-PVDF and KNN nanocrystals. The higher d 33 value (35 pC/N) exhibited by 70 vol% KNN (micron-sized crystallites)-PVDF composite was of technological importance. This was rationalized by invoking the presence of a large number of domain variants, besides the increased connectivity of the KNN in the PVDF matrix. Though the PVDF-KNN nanocrystal composites did not exhibit piezoresponse, it would be of some interest for dielectric applications as it exhibited higher dielectric constants.