Tunable room temperature magnetoelectric response of SmFeO₃/poly(vinylidene fluoride) nanocomposite films

Abstract

We report the results of our studies on the tailored magnetoelectric response of a polymer based nanocomposite system comprising of varying SmFeO₃ filler volume percentage embedded in a poly(vinylidene fluoride) (PVDF) matrix. The ME induced voltage was observed to increase with an increase in the SmFeO₃ filler concentration from 1% to 20%. For composite films with 20% SmFeO₃ concentration an induced voltage of ∼5.2 mV cm⁻¹ Oe⁻¹ was obtained at 1 kHz. Such a large magnetoelectric response is attributed to intensive interfacial interaction caused by dense packing of magnetostrictive SmFeO₃ in the piezoelectric matrix of PVDF. Further, poling of composite films by application of an electric field was observed to lead to a saturated magnetization hysteresis loop caused by reorientation of the magnetization of SmFeO₃. This electric field mediated manipulation of magnetic properties in SmFeO₃/PVDF holds great potential for low-energy-consuming spintronics devices.

---

1. Introduction

Magnetoelectric (ME) multiferroics are fascinating materials and are being actively investigated for use in multifunctional integrated devices.^1,2 Most of the single phase multiferroic materials suffer from the drawback of weak ME coupling and that too at temperatures below room temperature.³ To address this limitation composites comprising of magnetostrictive and piezoelectric phases are being actively investigated. In these materials the magnetostriction induced mechanical deformation leads to piezoelectric effect induced dielectric polarization variation, allowing ME coupling^4,5 which is many orders of magnitude higher than observed in single-phase materials, making them attractive for device applications.⁶ Although ceramic based composites have been investigated, these suffer from the drawback of low electrical resistance and high dielectric losses, hindering sustainable device applications.^7,8 The above mentioned problems can be circumvented in polymer based ME composites in 0–3 configuration, that is particles of one material (0) embedded in a 3 dimensional matrix of another material (3).^7,9 Further, polymer based multiferroic materials offer several other advantages like easy processing, the ability to be molded into desired configurations, light weight, low cost and scalable production methods compatible with industrial requirements for flexible structures.^10–13

Theoretical studies as well as some recent experiments have shown that piezoelectric polymer based multiferroic composites can provide colossal ME response due to large piezoelectric stress coefficients and great displacement transfer capability of the polymer.^7,14–16 The strength of ME coupling in these composites can be tailored by selecting the ferroelectric and ferromagnetic constituent materials according to their strain transfer efficiency.^17,18 Further, in piezoelectric polymer materials one can also take advantage of electric field poling to achieve a higher piezoelectric coefficient and consequently stronger ME coupling in the polymer based multiferroic composites. Few reports are available on the effect of poling on the ferroelectric/ferromagnetic ceramic composites, however the effect of electric field induced poling in the polymer based multiferroic composite films have not been reported.

We have therefore investigated the effect of electric field poling in the polymer based multiferroic composites with 0–3 configuration. PVDF (β-phase) is the most widely studied piezoelectric polymer material due to its high piezoelectric coefficient, excellent mechanical flexibility, low production cost compared to that of ceramics,¹⁹ which makes it a good candidate for developing flexible multiferroic composites used in magnetic control of electric polarization. For magnetostrictive material we chose SmFeO₃ (SFO) because SmFeO₃ it has a high magnetostriction coefficients,²⁰ high magnetic ordering temperature (T_N ∼ 670 K)²¹ and a spin reorientation (SR) transition temperature (T_SR) around 480 K.²² Because of its anomalous magnetoelastic properties near SR region and SR temperature higher than room temperature, SFO is a very attractive candidate for magneto electric applications.

2. Experimental section

PVDF solution was prepared by dissolving PVDF powder (molecular weight: 530 000 procured from Aldrich) in dimethyl sulphoxide. The nano SFO powder was prepared by glycine assisted autocombustion. Samarium nitrate (Sm(NO₃)₃·5H₂O) and ferric nitrate (Fe(NO₃)₃·9H₂O) were dissolved in water in stoichiometric proportion and then glycine was added. The solution was heated to ∼200 °C on a hot plate with continuous stirring. SFO nanoparticles were formed after combustion of gel. All precursor chemicals are of analytical grade and used as received without further purifications. To develop SFO/PVDF composite films, varying volume percentages of SFO nanoparticles (1%, 5%, 10% and 20%) were added to the PVDF solution and ultrasonicated (using ultrasonic probe vibrator) for 2 h. After that the solution was dispersed on a glass substrate by casting at 90 °C and annealed at 90 °C for 5 h. The SFO/PVDF composite films of thickness ∼ 10 μm were peeled from the glass substrate and crystallization was achieved by cooling down the samples to room temperature.

The phase of the powder was determined by powder X-ray diffraction (Instrument: Rigaku, X-ray source: Cu K_α radiation, λ = 1.5406 Å) operated in θ–2θ configuration. The particle size was analyzed using Zeiss field emission scanning microscope (FESEM). X-ray photoemission spectroscopy (XPS) was performed using Al-K_α (λ = 0.834 nm) lab-source.

The low field amplitude dielectric and conductivity measurements were carried out on the samples using Novocontrol Alpha-A impedance analyzer in the frequency range (10 Hz to 10⁶ Hz). The d₃₃ coefficients of these polymer nano composite films were measured using d₃₃ piezo meter test system (Piezotest). Magnetization measurements were carried out by employing a commercial 7 Tesla SQUID-vibrating sample magnetometer (SVSM; Quantum Design Inc., USA). Magnetostriction measurements were carried out using strain gauges and an electromagnet. Magnetoelectric measurements were done by varying the bias magnetic field (1 T) under a superimposed ac magnetic field of 1 Oe (with a resonating frequency of 1 kHz generated by Helmholtz coils). A dc bias and ac magnetic fields were simultaneously applied perpendicular to the sample surface (out of plane measurements). The output voltage generated from the composite was measured by lock in amplifier (Stanford, SR 532); the reference signal was taken from the signal generator feeding the Helmholtz coils. In ME coupling measurement the generated output voltage due to application of dc magnetic field is order of nanovolt. To separate out ME coupling voltage from noise due to other effects, an ac magnetic field was applied parallel to dc magnetic field with 1 kHz (resonant frequency of the Helmholtz coil) so that the signal generated due to ME coupling at 1 kHz was filtered out by lock in amplifier which was tuned to frequency 1 kHz. So the ME coupling voltage (V_ME) is proportional to [α × (H_dc + H_ac sin wt)]. Since H_dc ≫ H_ac, the |V_ME| mainly depends on H_dc and signal filtered out due to the frequency of H_ac.

3. Results and discussion

Rietveld refined X-ray diffraction (XRD) pattern of SFO (Fig. 1(a)) reveal its orthorhombic structure with space group Pnma. The calculated diffraction pattern matches well with the observed pattern indicating the absence of impurity phases. The XRD pattern of SFO/PVDF composite films with varying volume percentage of SFO is shown in Fig. 1(b). All the XRD patterns indicate that PVDF in SFO/PVDF composite films is in ferroelectric beta phase and the crystal structure of SFO is not affected in PVDF matrix.

Fig. 1 (a) Rietveld refined X-ray diffraction pattern of SmFeO₃; (b) X-ray diffraction pattern of SmFeO₃ nanoparticles and SmFeO₃/PVDF composite films.

The morphology and particle size of SFO nanoparticles were characterized using FESEM. The FESEM image (ESI file Fig. S1†) indicates nearly spherical SFO nanoparticles with diameter varying from ∼50 to 100 nm. The morphology of SFO/PVDF composite films with varying volume percentage of SFO is shown in Fig. 2. The insets of the Fig. 2 shows photo graphs of the composite films with different concentration of SFO. It is evident that for low filler volume percentage of 1% and 5% the SFO nanoparticles are uniformly distributed in the PVDF matrix (Fig. 2(a) and (b)). However with increasing filler volume percentage agglomeration is observed, which is expected for dense packing of filler (Fig. 2(c) and (d)).

Fig. 2 FESEM images of (a) 1%; (b) 5%; (c) 10% and (d) 20% SmFeO₃/PVDF composite films. The inset shows photo graph of the composite films.

To visualize the oxidation state of Sm and Fe, XPS measurements were carried out for pure SFO (ESI file Fig. S2†). The XPS spectra of SFO exhibits 2p core level binding energy of 711.5 eV (Fe 2p_3/2) and 723.6 eV (Fe 2p_1/2), corresponding to those of Fe³⁺ ions.^23,24 The peaks corresponding to (Sm 3d_3/2) and (Sm 3d_1/2) are positioned at 1082.4 eV and 1107.8 eV respectively. XPS studies confirm that the oxidation state of Sm and Fe are purely +3 states.

The dielectric permittivity is important properties of composites because it depends on the dielectric properties of the constituents, their mixing ratio and microstructure of composites etc. The dielectric properties depend on the frequency and also reflect the dynamical, transport and relaxation processes of the composites. To study the dielectric performance of the SFO/PVDF composites for ME coupled device application and to compare it with individual constituents the frequency dependent measurement of the dielectric permittivity of bulk SFO, pure PVDF and composite films were carried out at room temperature. The results are shown in Fig. 3(a). The dielectric constant of SFO/PVDF nanocomposite films is lower compared to pure PVDF and bulk SFO. In the composite films, SFO nanoparticles are surrounded by insulating PVDF, forming a capacitor like arrangement in the matrix.²⁵ In polymer based nanocomposites, the polymer chain movement around nano particles gets reduced due to the strong bonding of polymer chains and the particle surface.^26,27 Hence an immobile layer gets formed around nanoparticles which lead to reduction in permittivity of SFO/PVDF composite films compared to PVDF.

Fig. 3 (a) Frequency dependent permittivity for pure PVDF, pure SmFeO₃ and 20% SmFeO₃/PVDF composite films; (b) frequency dependent permittivity for 1%, 5%, 10% and 20% SmFeO₃/PVDF composite films; (c) frequency dependent ac conductivity for pure PVDF, pure SmFeO₃ and SmFeO₃/PVDF composite films; (inset of Fig. 3(c)) shows determination of η by fitting linear curve with log(σ) vs. log(ν) (experimental graph); (d) frequency dependent ac conductivity for 1%, 5%, 10% and 20% SmFeO₃/PVDF composite films.

The permittivity of the SFO/PVDF composite films with varying volume percentage of SFO is shown in Fig. 3(b). In the low frequency region the dielectric constant of composite films increases with decreasing frequency. This might be due to the interfacial polarization arises due to immobile layer formed around the particle surface (behaves like microcapacitors).²⁸ As the volume percentage of the SFO increases capacitor network also increases, which results in an increase in the dielectric constant from 2.5 for 1% SFO/PVDF at 1 kHz to 5.6 for 20% SFO/PVDF nano composite. Since the dielectric loss tangent remains almost same for all volume fraction of SFO in PVDF (0.04 at 1 kHz), the formation of conducting network of capacitors is ruled out. This indicates that the added volume of SFO in PVDF is well below the percolation threshold.

The measurements carried out on ac conductivity for the composite films are shown in Fig. 3(c). The conductivity of the composite films is lower in comparison to that of pure SFO as well as pure PVDF. In the 0–3 type composite geometry, the polymer matrix serves as dielectric capacitor while nanoparticles are isolated by thin polymer layers and act as resistors. This particular geometry of the composite film leads to the reduction in the ac conductivity. The frequency dependent component of ac conductivity is given by σ_fre(T) = A(T) × ω^η(T).²⁹ The parameter η was found to be temperature and frequency dependent and the exponent η is such that (0 < η < 1). The value of η for SFO/PVDF (20%) composite films is found to be 0.86 ± 0.018 at room temperature (inset Fig. 3(c)). Fig. 3(d) shows the conductivity of SFO/PVDF composite films with varying volume percentage of SFO. The ac conductivity of all the composite film follows the universal dynamic response, as with other conductor–insulator systems.³⁰ The observed low electrical conductivity of the composite films in comparison to the pure PVDF and SFO samples is due to the fact that SFO nanoparticles are capped by insulating polymer shells and show homogeneous dispersion^31,32 in the polymer matrix, which prevent electron conduction resulting in the lower leakage current and lower dielectric loss of the nanocomposite films.

Since PVDF is a piezoelectric polymer and the piezoelectric coefficient depends on the volume fraction of filler which in turn affects the ME coupling coefficient, the piezoelectric response (d₃₃) of the composite films was measured for different SFO concentrations content. As filler concentration increases, the piezoelectric response increase and the value of d₃₃ was found to be 9, 11, 15 and 17 for 1%, 5%, 10% and 20% SFO/PVDF composite films. The increase in piezoelectric response indicates that the increase in filler concentration up to 20% favors the polar phase formation (beta phase) of the PVDF composite film.³³ The good piezoelectric properties of SFO/PVDF nanocomposites might be related to the uniform dispersion of the filler nanoparticles.^34,35

Fig. 4(a) shows magnetization behavior as a function of temperature for pure SFO measured during field cooled process at magnetic field of 500 Oe. The magnetization curve clearly demonstrates two transitions one at 640 K and another around 450 K. The transition at 640 K corresponds to T_N, below which SmFeO₃ becomes a canted antiferromagnet and the drop in magnetization below 450 K (T_SR) is due to spin-reorientation transition (the spontaneous magnetization rotates continuously from a to c axis as the temperature is raised through a finite region). Such spin reorientation in orthoferrites is known to be brought by anisotropic magnetic interactions between R³⁺ and Fe³⁺ ions.^36,37 These results are consistent with the previous reports, and confirm the single phase nature of SFO.¹⁹

Fig. 4 (a) Temperature dependent magnetization for SmFeO₃ and inset of figure shows room temperature magnetization vs. magnetic field hysteresis loops for SmFeO₃; (b) room temperature magnetization vs. magnetic field hysteresis loops for 1%, 5%, 10% and 20% SmFeO₃/PVDF composite films. Inset of figure shows enlarge view of comparison of coercivity of pure SmFeO₃ and SmFeO₃/PVDF composite films.

The inset of Fig. 4(a) shows room temperature magnetic hysteresis loop (M vs. H) for pure SFO. The M–H loop behavior of pure SFO shows the antiferromagnetic behavior with weak ferromagnetism at room temperature. The hysteresis behavior of SFO is like a two phase magnetic system which could be decomposed into two additive loops. One of the loops reveals low coercive field and magnetic saturation, whereas the other loop shows high coercive field and no signature of saturation up to the applied field of 7 T. The observed two-phase like pinched magnetic hysteresis behavior is interesting and can be attributed to magnetization switching in this system around room temperature that originated from the strong interplay Sm-4f and Fe-3d electrons. SFO features two inequivalent magnetic sublattices, namely, Sm sublattice (4f electron) and Fe sublattice (3d electron) which are antiferromagnetically coupled. Non collinear antiferromagnetism in the Fe sublattice gives rise to weak ferromagnetism at relatively high temperatures (∼300 K), while the Sm sublattice typically orders antiferromagnetically at much lower temperatures (∼135 K).³⁸ Earlier reports reveal that SFO possesses an extremely useful magnetization that is very sensitive to small field perturbations even at room temperature.^20,38

Fig. 4(b) shows room temperature magnetic hysteresis loop (M vs. H) of SFO/PVDF composite films. The M–H curves for SFO/PVDF composite films exhibit significantly different behavior in comparison to the pure SFO. For comparison an enlarged view of M–H curves of pure SFO and SFO/PVDF composite films is shown in the inset of the Fig. 4(b).

The observed coercive field was 65 Oe and 260 Oe for pure SFO and SFO/PVDF composite films respectively. This might be due to the geometry of the composite films where dispersion of the SFO nanoparticles in the polymer matrix leads to modification in interparticle interactions among the nanoparticles. It is known that in ME composites homogeneously dispersed fraction of magnetic component lead to the interrupted magnetic interaction between grains, resulting in increased coercivity.³⁹ The coupling between magnetic polarization and mechanical strain was verified for the composite films by determining magnetostriction (λ) and piezomagnetic coefficients (q). Fig. 5 shows in-plane parallel magnetostriction (λ₁₁) and in-plane perpendicular magnetostriction (λ₁₂) as a function of magnetic field for SFO/PVDF composite films. The piezomagnetic coefficient (q_ij) is derived for composite films by differentiating the magnetostriction parameter with respect to the magnetic field (q₁₁ = dλ₁₁/dH and q₁₂ = dλ₁₂/dH). The λ₁₁ of 1% SFO/PVDF composite decreases with the increase in magnetic field and saturates at 2.1 kOe with a maximum value of −1.4 ppm (Fig. 5(a)). Similarly, the λ₁₂ saturates at 1.9 kOe with a maximum value of 1.5 ppm. A similar behavior was observed for the 5%, 10% and 20% SFO/PVDF composite films (Fig. 5(b)–(d)) with λ₁₁ = −4, −9.3, −12.1 and λ₁₂ = 4, 8.9, 12 ppm with saturation fields of 2, 2.4 and 2.6 kOe, respectively. The inset of Fig. 5 presents the magnetic field dependence of q₁₁ and q₁₂ for SFO/PVDF composite films. The longitudinal piezomagnetic coefficient (q₁₁) shows maxima with a value of −1.6, −4.2, −8.4, −10 ppm Oe⁻¹ at H = 0.8, 1.1, 1.2, 1.3 kOe for 1%, 5%, 10% and 20% SFO/PVDF composites respectively. The position of peaks in q₁₁ and q₁₂ is an important factor as this corresponds to the field required to obtain the maximum ME coefficient.

Fig. 5 Room temperature longitudinal and transverse magnetostriction (a) 1%, (b) 5%, (c) 10% and (d) 20% SmFeO₃/PVDF composite films. Magnetic field dependence of piezomagnetic coefficients is shown in the inset of figures.

In order to investigate the ME coupling of the SFO/PVDF composite films, magnetoelectric coupling coefficient (α) was measured for 1%, 5%, 10% and 20% SFO/PVDF composite films at 1 kHz under H_ac = 1 Oe as shown in Fig. 6(a). The ME coefficient (α_ME) initially increases with the magnetic field, shows a maximum around 1 kOe and decreases thereafter with further increase in the magnetic field. The dependence of α on H tracks the variation of the piezomagnetic coupling between piezoelectricity and magnetostriction.

Fig. 6 (a) Magnetoelectric coupling coefficient (α) vs. dc magnetic field for 1%, 5%, 10% and 20% SmFeO₃/PVDF composite films at 1 kHz under H_ac = 1 Oe; (b) α_ME vs. volume fraction of the SmFeO₃ in the PVDF matrix.

The ME coefficients observed at the maxima for 1%, 5%, 10% and 20% SFO/PVDF composites are 0.7, 0.82, 3.2 and 5.2 mV cm⁻¹ Oe⁻¹, respectively. The increase in α_ME with increase in SFO concentration in polymer matrix can be attributed to the increase in magnetostrictively induced strain of composites with the volume fraction of the magnetic phase. Fig. 6(b) shows increase in α_ME with increasing volume percentage of the SFO in the PVDF matrix.

The achieved α_ME values are compared favorably to the values of the reported multiferroic PVDF based polymer composites.⁷ The ME effect in polymer composites can be explained by the following equation.⁹

where are given byp and m indicating the piezoelectric and magnetostrictive phase respectively; d_3n the piezoelectric coefficients; ε the dielectric constant, ϕ the volume fraction of the magnetostrictive component; T and H the stress and applied magnetic field, respectively; ξ the magnetic permeability and M the magnetization. The ME voltage coefficient, α₃₃, in these composites can be further increased by increasing filler content, ϕ, dielectric constant of the composite, ε, or by suitably modifying the elastic modulus.

To further test the strain coupling between electric and magnetic orders, M–H curves were measured for electrically poled and unpoled composite films. The results are shown in Fig. 7(a) for 20% SFO/PVDF composite film. Electric poling was performed using an electric field of +1 MV cm⁻¹ on the same piece of sample on which the MH measurements were performed without poling. Interestingly, the poled samples reveal a drastic change in the behavior of M–H curves (Fig. 7(a)). Poled composite films showed ferromagnetic features with well saturated M–H loop at an applied magnetic field of only 4 T. However, M–H loop of unpoled samples were not saturated even with 7 T applied magnetic field. Moreover, remnant magnetization (M_R) for poled samples was reduced by half and coercivity increased by a factor of two. The poling process induced significant changes in the magnetization due to strong strain mediated interfacial interactions between PVDF and SFO phases. Due to the converse piezoelectric effect, an electric field applied on the ferroelectric phase produces a strain which modifies the magnetic properties of ferromagnetic material through the inverse magnetostrictive effect.

Fig. 7 (a) Room temperature magnetization vs. magnetic field hysteresis loops; (b) frequency dependent permittivity for electrically poled and unpoled 20% SmFeO₃/PVDF composite films.

Dielectric constant of the composite films was also measured for unpoled and poled composite films and the results are shown in Fig. 7(b). Poled film shows increases in the permittivity at high frequencies indicating better alignment of dipoles and polarization ordering.⁴⁰ In the composite films, there is strong bonding of polymer chains and the particle surface and interface bonding is sensitive to atomic displacements at the interface. This strong interface bonding might lead to the change in the magnetization when the ordering of ferroelectric polymer layer gets modified. The observed results clearly demonstrate strong strain mediated coupling between magnetic and electric order parameters (ME coupling) in SFO/PVDF composite films.

In Fig. 8 we show schematically the possible mechanism of coupling between electric and magnetic properties after electrical poling of SFO/PVDF composite films. In SFO/PVDF composite films, the electric field poling in z direction induces strain in PVDF matrix in x and y direction and consequently magnetic moments get locked in x and y direction. These strain induced changes in magnetic properties are attributed to the changes in magnetic anisotropy which can be clearly understood from the sign of measured magnetostriction coefficients for the composite films (Fig. 5). Interestingly, the magnetostriction in SFO reaches to maximum in the SR region (i.e. nearly above room temperature).²⁰ It shows temperature and/or magnetic field induced fast magnetic spin switching and an easy axis rotation transition.^20,22 The magnetostriction in SFO is proportional to the rotation angle of the magnetization in SR region.²⁰ When the composite films were electrically poled, the electric diploes inside ferroelectric layer get aligned in the applied field direction which in turn is expected to modify the magnetization alignment of SFO nanoparticles, which is reflected in the magnetization hysteresis curves of the poled films.

Fig. 8 Schematic representation of possible mechanism of coupling between electric and magnetic properties after electrical poling of SmFeO₃/PVDF composite films.

4. Conclusions

Nanocomposite films having highly magnetostrictive SFO nanoparticles dispersed in the piezoelectric PVDF matrix in volume% varying from 1 to 20% were prepared by a simple film casting method. These films have shown strong ME coupling at room temperature with ME coefficient increasing with volume fraction of SFO. Compared to unpoled films poled polymer based composite films were observed to show well saturated M–H loop, a factor of two reduction in remnant magnetization and an increase in coercivity by a factor of two. This suggests electric field control of magnetization which arises because of spin reorientation. We expect that the described phenomenon of poling in polymer based composites (magnetostrictive/piezoelectric) presents simple and easy method to strongly couple the magnetic and electric properties at room temperature. SFO/PVDF composite films are attractive candidates for applications such as ME sensors, memory storage and spintronics devices etc.

Acknowledgements

The authors would like to acknowledge DST for financial support. MMS thanks China Postdoctoral Science Foundation (Grant no. BH2340000067) and the National Natural Science Foundation of China (Grant No. 21427804) for the financial support.

References

1. M. Fiebig, J. Phys. D: Appl. Phys., 2005, 38, R123.
2. M. M. Shirolkar, C. Hao, X. Dong, T. Guo, L. Zhang, M. Li and H. Wang, Nanoscale, 2014, 6, 4735.
3. J. F. Scott, NPG Asia Mater., 2013, 5, e72.
4. C. W. Nan, Phys. Rev. B: Condens. Matter Mater. Phys., 1994, 50, 6082.
5. B. Y. Wang, H. T. Wang, S. B. Singh, Y. C. Shao, Y. F. Wang, C. H. Chuang, P. H. Yeh, J. W. Chiou, C. W. Pao, H. M. Tsai, H. J. Lin, J. F. Lee, C. Y. Tsai, W. F. Hsieh, M.-H. Tsai and W. F. Pong, RSC Adv., 2013, 3, 7884.
6. A. Kulkarni, K. Meurisch, I. Teliban, R. Jahns, T. Strunskus, A. Piorra, R. Knöchel and F. Faupel, Appl. Phys. Lett., 2014, 104, 022904.
7. P. Martins and S. L. Méndez, Adv. Funct. Mater., 2013, 23, 3371.
8. S. Liu, S. Xue, W. Zhang, J. Zhai and G. Chen, J. Mater. Chem. A, 2014, 2, 18040.
9. P. Martins, V. Y. Kolenko, J. Rivas and S. Lanceros-Mendez, ACS Appl. Mater. Interfaces, 2015, 7, 15017.
10. O. D. Jayakumar, B. P. Mandal, J. Majeed, G. Lawes, R. Naik and A. K. Tyagi, J. Mater. Chem. C, 2013, 1, 3710.
11. J. Jin, S. G. Lu, C. Chanthad, Q. Zhang, M. A. Haque and Q. Wang, Adv. Mater., 2011, 23, 3853.
12. S. Kasisomayajula, N. Jadhav and V. J. Gelling, RSC Adv., 2016, 6, 967.
13. P. Martins, X. Moya, L. C. Phillips, S. K. Narayan, N. D. Mathur and S. L. Mendez, J. Phys. D: Appl. Phys., 2011, 44, 482001.
14. J. Jin, F. Zhao, K. Han, M. A. Haque, L. Dong and Q. Wang, Adv. Funct. Mater., 2014, 24, 1067.
15. P. Martins, A. Larrea, R. Gonçalves, G. Botelho, E. V. Ramana, S. K. Mendiratta, V. Sebastian and S. Lanceros-Mendez, ACS Appl. Mater. Interfaces, 2015, 7, 11224.
16. T. H. L. Nguyen, L. Laffont, J.-F. Capsal, P. J. Cottinet, A. Lonjon, E. Dantras and C. Lacabanne, Mater. Chem. Phys., 2015, 153, 195.
17. P. Martins, A. Lasheras, J. Gutierrez, J. M. Barandiaran, I. Orue and S. Lanceros-Mendez, J. Phys. D: Appl. Phys., 2011, 44, 495303.
18. P. Martins, R. Gonçalves, S. Lanceros-Mendez, A. Lasheras, J. Gutiérrez and J. M. Barandiarán, Appl. Surf. Sci., 2014, 313, 215.
19. O. D. Jayakumar, B. P. Mandal, J. Majeed, G. Lawes, R. Naik and A. K. Tyagi, J. Mater. Chem. C, 2013, 1, 3710.
20. M. Abe, K. Kaneta, M. Gomi, Y. Mori and S. Nomura, Jpn. J. Appl. Phys., 1977, 10, 1799.
21. E. N. Maslen, V. A. Streltsov and N. Ishizawa, Acta Crystallogr., Sect. B: Struct. Sci., 1996, 52, 406.
22. J. H. Lee, Y. K. Jeong, J. H. Park, M. A. Oak, H. M. Jang, J. Y. Son and J. F. Scott, Phys. Rev. Lett., 2011, 107, 117201.
23. A. H. M. Gonzalez, A. Z. Simões, L. S. Cavalcante, E. Longo, J. A. Varela and C. S. Riccardi, Appl. Phys. Lett., 2007, 90, 052906.
24. A. Z. Simões, A. H. M. Gonzalez, L. S. Cavalcante, C. S. Riccardi, E. Longo and J. A. Varela, J. Appl. Phys., 2007, 101, 074108.
25. D. Bhadra, M. G. Masud, S. Sarkar, J. Sannigrahi, S. K. De and B. K. Chaudhuri, J. Polym. Sci., Part B: Polym. Phys., 2012, 50, 572.
26. M. Roy, J. K. Nelson, R. K. M. Crone, L. S. Schadler, C. W. Reed, R. Keefe and W. Zenger, IEEE Trans. Dielectr. Electr. Insul., 2005, 12, 629.
27. A. Ahlawat, S. Satapathy, S. Bhartiya, M. K. Singh, R. J. Choudhary and P. K. Gupta, Appl. Phys. Lett., 2014, 104, 042902.
28. O. L. Pamies, T. Goudarzi, A. B. Meddeb and Z. A. Ounaies, Appl. Phys. Lett., 2014, 104, 242904.
29. A. K. Jonscher, Nature, 1975, 253, 717.
30. L. He and S. C. Tjong, Nanoscale Res. Lett., 2013, 8, 132.
31. T. Yamada, T. Ueda and T. Kitayama, J. Appl. Phys., 1982, 53, 4328.
32. J. B. Ngoma, J. Y. Cavaille, J. Paletto, J. Perez and F. Macchi, Ferroelectrics, 1990, 109, 205.
33. R. Gonçalves, P. M. Martins, C. Caparrós, P. Martins, M. Benelmekki, G. Botelho, S. Lanceros-Mendez, A. Lasheras d, J. Gutiérrez and J. M. Barandiarán, J. Non-Cryst. Solids, 2013, 361, 93.
34. B. Chu, M. Lin, B. Neese, X. Zhou, Q. Chen and Q. M. Zhang, Appl. Phys. Lett., 2007, 91, 122909.
35. J. X. Zhang, J. Y. Dai, L. C. So, C. L. Sun, C. Y. Lo, S. W. Or and H. L. W. Chan, J. Appl. Phys., 2009, 105, 054102.
36. T. Yamaguchi, J. Phys. Chem. Solids, 1974, 35, 479.
37. G. Gorodetsky and L. M. Levinson, Solid State Commun., 1969, 7, 67.
38. S. Cao, H. Zhao, B. Kang, J. Zhang and W. Ren, Sci. Rep., 2014, 4, 5960.
39. A. Gupta and R. Chatterjee, J. Eur. Ceram. Soc., 2013, 33, 1017.
40. G. T. Davis, T. Furukawa, A. J. Lovinger and M. G. Broadhurst, Macromolecules, 1982, 15, 329.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra01152g

---