Unprecedented Ferroelectric–Antiferroelectric–Paraelectric Phase

Article pubs.acs.org/JACS

Unprecedented Ferroelectric−Antiferroelectric−Paraelectric Phase Transitions Discovered in an Organic−Inorganic Hybrid Perovskite Peng-Fei Li, Wei-Qiang Liao, Yuan-Yuan Tang, Heng-Yun Ye, Yi Zhang,* and Ren-Gen Xiong* Ordered Matter Science Research Center, Jiangsu Key Laboratory for Science and Applications of Molecular Ferroelectrics, Southeast University, Nanjing 211189, P. R. China S Supporting Information *

ABSTRACT: As a promising candidate for energy storage capacitors, antiferroelectric (AFE) materials have attracted great concern due to their congenital advantages of large energy storage ability from double polarization versus electric field (P−E) hysteresis characteristics in contrast to ferroelectrics and linear dielectrics. However, antiferroelectricity has only been discovered in inorganic oxides and some hydrogen-bonded molecular systems. In view of the structural diversity and unique physical properties of organic−inorganic hybrid system, it remains a great opportunity to introduce antiferroelectricity into organic−inorganic hybrid perovskites. Here, we report that polarizable antiparallel dipole arrays can be realized in an organic−inorganic hybrid perovskite, (3-pyrrolinium)CdBr3, which not only exhibits an excellent ferroelectric property (with a high spontaneous polarization of 7.0 μC/cm2), but also presents a striking AFE characteristic revealed by clear double P−E hysteresis loops. To the best of our knowledge, it is the first time that such successive ferroelectric− antiferroelectric−paraelectric phase transitions have been discovered in organic−inorganic perovskites. Besides, a giant dielectric constant of 1600 even at high frequency of 1000 kHz and a bulk electrocaloric effect with entropy change of 1.18 J K−1 kg−1 under 7.41 kV/cm are also observed during the phase transition. Apparently, the combined striking AFE characteristic and giant dielectric constant make (3-pyrrolinium)CdBr3 a promising candidate for next generation high-energy-storage capacitors.

■

INTRODUCTION Electrical energy storage has attracted worldwide concern for the present threat of energy crisis and global warming. To effectively reduce the production of greenhouse gases along with the rapid development of electronic industry, there has been increasing demand for high-energy-storage capacitors over the past decades.1 Among the homogeneous-structured dielectrics used for energy storage capacitors, antiferroelectric (AFE) material is considered as one of the most promising candidates due to its congenital advantages of large energy storage ability from polarization versus electric field (P−E) characteristics in contrast to ferroelectrics and linear dielectrics.2 Structurally, AFE materials with adjacent dipoles (in a single domain) polarized in antiparallel directions display a typical double P−E hysteresis loop when subjected to a strong electric field.3 The electric field induced AFE to ferroelectric phase transition generally leads to a large lattice strain and high electrostatic energy density.4 In addition, the high dielectric constant and the distinct phase transition in AFE materials also provide great opportunities for the realization of high-energy storage density and fast discharging rates in high performance capacitors. Homogeneous-structured AFE materials have so far been discovered mainly in inorganic oxides, such as PbZrO3, PbHfO3, and NaNbO3, with perovskite structure5 and some © 2017 American Chemical Society

molecular-based crystals like ammonium dihydrogen phosphate (ADP).6 Compared with the abundant structural diversity of dielectrics and ferroelectrics, such a small amount of existing AFE materials has become a big hindrance for further development of technical implementation in high-energystorage capacitors. Therefore, it remains both a great opportunity and challenge to explore new AFE materials with unique properties like low-density, easy processing, and flexible structural tunability.7 Currently, active research has been focused on hybrid organic−inorganic perovskites for various applications like electric, magnetic, luminescent, and photovoltaic devices.8 Constructing new electric ordering materials based on organic−inorganic hybrids has also been proven to be one of the most feasible strategies.9 Very recently, we have successfully designed a three-dimensional organic−inorganic perovskite ferroelectric, (3-ammoniopyrrolidinium)RbBr3, which shows extraordinary high Curie temperature and ultrafast polarization switching.10 In addition, organic−inorganic perovskites with one-dimensional hexagonal stacking frameworks can also be structurally engineered to show excellent ferroelectricity.11 However, reports of organic−inorganic perovskite AFE systems have still remained sparse. This situation is quite Received: May 8, 2017 Published: June 8, 2017 8752

DOI: 10.1021/jacs.7b04693 J. Am. Chem. Soc. 2017, 139, 8752−8757

Journal of the American Chemical Society

■

RESULTS AND DISCUSSION Colorless rod-like crystals of (3-pyrrolinium)CdBr3 (1) with size 3 × 3 × 15 mm3 were grown by slow evaporation of a clear aqueous solution containing equal molar amounts of CdBr2 and pyrroline hydrobromide at room temperature (Supporting Information, Figure S1). Phase purity of 1 was confirmed by powder XRD and infrared (IR) analysis (Figures S2 and S3). To confirm the phase transition behavior of 1, we first carried out DSC measurement. As shown in Figure 1, the DSC curves

similar to aforementioned inorganic oxides, indicating that polarizable antiparallel dipole arrays can hardly form in homogeneous crystalline materials. On the other hand, as a vital criterion for AFE, double P−E hysteresis loops have always been hard to demonstrate in AFE materials since their switching field is usually larger than the breakdown field. Although some new compounds were claimed to have AFE property, double P−E hysteresis loops are rarely observed experimentally. In the course of exploring high-performance electric ordering materials based on organic−inorganic hybrids, we found a new compound (3-pyrrolinium)CdBr3, which not only exhibits an excellent ferroelectric property (with a high spontaneous polarization of 7.0 μC/cm2), but also presents a striking AFE characteristic revealed by clear double P−E hysteresis loops. Besides, a giant dielectric constant of 1600 even at high frequency of 1 MHz and a bulk electrocaloric effect with entropy change of 1.18 J K−1kg−1 under 7.41 kV/cm are also observed during the ferroelectric-to-antiferroelectric phase transition. Apparently, the combined striking AFE characteristic and giant dielectric constant make (3-pyrrolinium)CdBr3 a promising candidate for next generation high-energy-storage capacitors. To the best of our knowledge, it is the first time that such successive ferroelectric−antiferroelectric−paraelectric phase transitions have been discovered in organic−inorganic perovskites. In view of the structural diversity of organic− inorganic hybrid system, the present work will undoubtedly stimulate a research upsurge about structure−property relationships toward high performance ferroelectric or AFE devices.

■

Article

Figure 1. DSC curves of 1.

clearly present two pairs of reversible heat anomalies at 246.6/ 244.2 K (T1) and 237.9/235.2 K (T2) in the heating/cooling runs, with the narrow temperature hysteresis of 2.4 and 2.7 K, respectively, indicative of two typical second-order phase transitions. For convenience, the three solid phases of 1 are marked according to these two phase transitions, as the lowtemperature phase (LTP), the intermediate-temperature phase (ITP), and the high-temperature phase (HTP), respectively. To deeply understand the phase transition mechanism, we determined the crystal structures of 1 at 293, 238, and 203 K by single-crystal XRD, respectively (Figure 2). The structure of the HTP is solved in the orthorhombic space group Cmcm, which is isostructural to our recently reported AMX3 analogues.12 It contains infinite columns of face-sharing CdBr6 octahedra separated by the 3-pyrrolinium cations (Figure 2a), arranged in the same manner as the NiO6 unities and Ba cations in the hexagonal 2-H BaNiO3.13 The geometric parameters of the organic and inorganic parts are unexceptional. In terms of polarization, the characteristic feature of the HTP is that the organic cation is located on a crystallographic mirror plane with a disordered fashion, and thus, the polarization is canceled out, which leads to no alignment of polarization in the c-direction. The ITP and the LTP have similar packing structures as that of the HTP, showing the pseudosymmetry to the HTP. Therefore, the other possible space groups different from Cmcm were chosen for the ITP and LTP based on the antiferroelectriciy and ferroelectricity, respectively. The ITP assumes the space group “C 2/m 1 1” (we chose the nonstandard unit cell setting so that the directions of the axis are close to those in the HTP, which facilitates the comparison and the understanding of the transition mechanisms). The structure can also be refined in the space group of the HTP (Cmcm) or the LTP (Cmc21); we did not refine it in the space group Cmcm or Cmc21 because the antiferroelectricity for this phase requires antiparallel alignment of the dipoles. The

EXPERIMENTAL SECTION

DSC, SHG, and XRD Measurements. Differential scanning calorimetry (DSC) measurements were performed on a PerkinElmer Diamond differential scanning calorimeter under nitrogen atmosphere in aluminum crucibles with a heating or cooling rate of 10 K/min. For second harmonic generation (SHG) experiments, an unexpanded laser beam with low divergence (pulsed Nd:YAG at a wavelength of 1064 nm, 5 ns pulse duration, 1.6 MW peak power, 10 Hz repetition rate) was used. The instrument model is Ins 1210058, INSTEC Instruments, while the laser is Vibrant 355 II, OPOTEK. The numerical values of the nonlinear optical coefficients for SHG have been determined by comparison with a KDP reference. Variable-temperature X-ray diffraction (XRD) analysis was carried out using a Rigaku Saturn 724+ CCD diffractometer with Mo−Kα radiation (λ = 0.71073 Å). Data collection, cell refinement, and data reduction were performed using Rigaku CrystalClear 1.3.5. The structures were solved by direct methods and refined by the full-matrix method based on F2 using the SHELXTL software package. All non-hydrogen atoms were refined anisotropically, and the positions of all hydrogen atoms were generated geometrically. Dielectric, Ferroelectric, and Pyroelectric Measurements. For dielectric, ferroelectric, and pyroelectric measurements, the samples were made with single-crystals cut into the form of thin plate perpendicular to the crystal c-axis. Silver conduction paste deposited on the plate surfaces was used as the electrodes. Complex dielectric permittivity was measured with a TH2828A impedance analyzer over the frequency range from 1−1000 kHz with an applied electric field of 0.5 V. Dielectric hysteresis loops were recorded on a Radiant Precision Premier II. Pyroelectric property was measured with an electrometer/high resistance meter (Keithley 6517B) with a heating or cooling rate of 10 K/min. The bias electric field dependence of dielectric constant was measured with an Alpha-A High performance Frequency Analyzer at 10 kHz. 8753

DOI: 10.1021/jacs.7b04693 J. Am. Chem. Soc. 2017, 139, 8752−8757

Article

Journal of the American Chemical Society

Figure 3. SHG intensity of 1 as a function of temperature.

the SHG conclusions match well with the results of crystal structure analyses. Furthermore, the continuous increase of SHG intensity near T2 suggests a second-order phase transition feature. The temperature dependence of the complex dielectric constant (ε = ε′ − iε″, where ε′ is the real part of the complex constant, and ε″ is the imaginary part of the complex constant) has been proved to be another effective way to detect structural phase transition, the reason for which is that the macroscopic dielectric response is sensitive to the variation of the microscopic structure. The real part (ε′) as a function of temperature, measured at different frequencies along the c-axis, is shown in Figure 4a. As the temperature decreases to in the vicinity of the T1, ε′ increases slowly and presents a tiny frequency independent dielectric anomaly. At the lower temperature, the ε′ shows rapidly changing, accompanied by a giant frequency dependent dielectric responses at T2. The peak values increase from 1600 to 2500 as the measurement frequencies decrease from 1000 kHz to 1 kHz. In addition, dielectric anisotropy measurement is one of the most feasible strategies to further investigate the intrinsic structural changes in the phase transition processes. As shown in Figure S4, the values of ε′ along the c-axis are obviously larger than those along the b- and a-axes. The remarkable dielectric anisotropy is related to the placement and motion of the cations. It can be seen from the structure in the ITP that the disordered 3pyrrolinium cation takes two positions, arousing the motions both in the plane parallel to the bc-plane and perpendicular to the a-axis. The in-plane motion only yields a very weak dipole moment component in the a-axis, which leads to a small dielectric change. Furthermore, by comparing the structure of the HTP and LTP, the N atoms show obvious shift along the cdirection, which will lead to large dielectric response. Significantly, besides large dielectric anomalies, antiferroeletrics and ferroelectrics display strong nonlinear dielectric response because of switchable spontaneous polarization. Hence, the bias-field dependence of ε′ was investigated in the three phases. As presented in Figure 4b, ε′ remains unchanged with the changes of the electric field and shows linear dielectric properties in the HTP, whereas ε′ in the ITP shows an exciting double butterfly shape under circular function of external DC electric field, indicating a striking feature of AFE. As the electric-field increase from zero to maximum values, ε′ increases first and reaches the first peak value at the phase switching from AFE to FE state and then decreases gradually. Similarly, as electric field decreases from maximum values to zero, dielectric constant also increases first and then declines. A second peak value at the phase transformation from FE to AFE

Figure 2. Perspective views of 1 in the (a) HTP, (b) ITP, and (c) LTP, showing the similarities of the crystal packings and differences of the orientational states of the cations and the lattices. In panel b, the atoms with the probability of N of 70% are shown in blue, while the atoms with probability of C of 70% are shown in brown. The pink planes indicate the (pseudo) crystallographic mirror planes.

modeled 3-pyrrolinium cation remains disordered. However, the ratios the N and C for the two atoms bonded by the double bond are 0.7:0.3 and 0.3:0.7. As shown in Figure 2b, such a model indicates that there are antiparallel electric dipole moments along the c-axis. The LTP assumes the space group Cmc21. As shown in Figure 2c, the orientations of the cations become the same, and thus, the crystal should display a net polarization. The SHG signal is sensitive to the lack of inversion symmetry; thus, the SHG technique has been developed as an indispensable supplementary approach to detect the symmetry change accompanying the structural phase transition.14 To further confirm the reasonable space groups, determined from the structural analyses, the temperature dependent SHG measurements of 1 were performed. As shown in Figure 3, SHG intensity is zero above T1, suggesting that 1 is centrosymmetric in the HTP, since only noncentrosymmetric solid is SHG active in principle. With the temperature decreasing between T1 and T2, SHG intensity remains zero, indicating that 1 also belongs to centrosymmetric in the ITP. When the temperature deep cooled below T2, nonzero values of the SHG intensity emerged, unambiguously revealing that 1 is noncentrosymmetric in the LTP. Therefore, 8754

DOI: 10.1021/jacs.7b04693 J. Am. Chem. Soc. 2017, 139, 8752−8757

Article

Journal of the American Chemical Society

Figure 4. Dielectric properties in 1. (a) Real part of the complex dielectric constant (ε′) of 1 along the c-axis at different frequencies. (b) Plot showing the dependence of the real part ε′ of the complex dielectric constant of 1 on the bias electric field.

Figure 5. (a) P−E hysteresis loops of 1 measured along the c-axis at HTP, ITP, and LTP. (b) I−E curves of 1 measured along the c-axis at HTP, ITP, and LTP.

state is obtained in this course. Dramatically, under the same external DC electric field modulate, the dielectric constant changes from a double butterfly shape to single butterfly shape in the process of transforming from ITP to LTP. The butterfly curve in the LTP indicates that the ε′ shows nonlinearity in response to the polarization switching, which is one of the characters for other well studied ferroelectrics. The measurement of P−E hysteresis loop is one of the most direct methods to determine the ferroelectricity and antiferroelectricity. As depicted in Figure 5a, the polarization response to the applied field is linear at 248 K in the HTP, as expected for ordinary dielectrics under a smaller field. With the temperature decreasing below T1, a characteristic double hysteresis P−E loops become visible at 245 K, showing one distinct features of the antiferroelectrics. The forward phase switching field (EF) and the backward switching field (EA) can be estimated by extrapolating the two steepest sections of the hysteresis loop and obtaining their intersections with the horizontal axis. Moreover, the better double hysteresis loops can be obtained with larger saturated polarization and smaller switching field as the temperature decreases (Figure 5a and Figure S5a). However, when the temperature drops down below T2, the double hysteresis loops change to single hysteresis loop in the LTP, revealing a transition from

antiferroelectric to ferroelectric phase. At the same time, the saturated polarization Ps reaches the maximum value of about 5.65 μC cm−2 at 223 K, which is fairly consistent with the estimated Ps value (5.48 μC·cm−2) according to point electric charge model (Supporting Information, Table S1). The corresponding coercive field is about 3.5 kV cm −1 . Furthermore, according to the P−E loop measured at 245 K in the ITP (Figure S6), we estimate that the maximum energy storage density is about 0.4−0.6 J cm−3; the energy-storage efficiency of (C4NH10)CdBr3 was 85% at 245 K for an applied electric field of 32 kV cm−1. Ferroelectric as a special kind of pyroelectric shows a sensitive spontaneous polarization response to the temperature. Obviously pyroelectric effect can be generated during the ferroelectric−antiferroelectric and ferroelectric−paraelectric phase transition. The polarization and current as a function of temperature for compound 1 are shown in Figure 6. As the temperature is above T2, the electric current keeps unchanged, while a sharp change of the current occurs in the vicinity of T2. By integrating the pyroelectric current, we can see that the polarization values increase with decreasing temperature below T2, basically consistent with those measured by the P−E hysteresis loops. Moreover, the tendency of polarization versus temperature is similar as that of the SHG signal, which can be 8755

DOI: 10.1021/jacs.7b04693 J. Am. Chem. Soc. 2017, 139, 8752−8757

Article

Journal of the American Chemical Society

between the ferroelectric to antiferrroelectric configuration is constructed based on the crystal structure obtained from the single-crystal XRD. Accordingly, the structure of low temperature ferroelectric phase at 223 K is used as the ferroelectric configuration (λ = 1), while the antiferrroelectric (λ = 0) configuration is obtained from the matrix transformation of the coordinates of λ = 1 considering both the rotation and displacement of the organic cations. The variation of polarization and total energy as a function of the dynamic path are shown in Figure 8, from which the ferroelectric

Figure 6. (a) Temperature dependence of spontaneous polarization by integrating the pyroelectric current.

explained by the Landau theory equation, χ(2) = 6ε0β′Ps, where the ε0 and β′ are temperature independent. Electrocaloric (EC) effect, the reverse course of pyroelectric effect, is the change in temperature of a material upon the application or removal of an electric field. According to the measurement results of P−E loops as a function of temperature (Figure S7), we obtain the relation of polarization change under different external electric fields (Figure 7a). The entropy

Figure 8. Variation of (a) polarization and (b) total energy as a function of the dynamic path connecting the ferroelectric (λ = 1) to antiferrroelectric (λ = 0) configuration.

polarization with 6.0 μC/cm2 along c-axis can be extracted from the initial point of λ = 1 configuration, while the value of polarization gradually decreases to zero toward the λ = 0 AFE state with an energy difference of 0.15 eV per unit cell, in good accordance with the experimental pyroelectric and P−E hysteresis measurements.

■

CONCLUSION In summary, on the basis of our preliminary study on hybrid metal halide perovskite-type compounds, by deliberate selection of organic cations and by modifying the bridging halogen atoms, we successfully obtained a new hybrid compound (3-pyrrolinium)CdBr3 with typical ferroelectric and antiferroelectric properties. This finding opens a new avenue for design ferroelectric/antiferroelectric multifunctional materials based on organic−inorganic hybrid compounds. The excellent ferroelectric and antiferroelectric performances of (3pyrrolinium)CdBr3 may render them to be promising materials for potential applications in information storage, energystorage, and sensor field. Furthermore, to realize room/higher temperature ferroelectricity or antiferroelectricity, more efforts should be paid on increasing the energy barrier of the phase transition through delicate structural design. Empirically, strengthening the hydrogen bonding effect between the organic and inorganic parts or tuning the fit tolerance of the two parts with more rigid stereohindrance effect maybe two effective strategies for pursuing ferroelectric/antiferroelectric materials with higher Curie temperature. Considering the flexible

Figure 7. (a) Relation of polarization change and (b) electrocaloric entropy change (ΔS) as a function of temperature for different external electric fields.

change can be obtained through the variation of the Maxwell 1

equation, ΔS = − ρ ∫

E2

E1

( ∂∂TP ) dE. As shown in Figure 7b, the

entropy change with temperature during in the process of phase transition shows a peak value of entropy change about 1.18 J K−1 kg−1 under 7.41 kV/cm, which is comparable to those bulk inorganic ferroelectric oxides.15 To gain deep insight into the ferroelectric-to-antiferroelectric phase transition, density functional theory calculation is carried out to evaluate the polarization dynamics. A dynamic path 8756

DOI: 10.1021/jacs.7b04693 J. Am. Chem. Soc. 2017, 139, 8752−8757

Article

Journal of the American Chemical Society

(10) Pan, Q.; Liu, Z.-B.; Tang, Y.-Y.; Li, P.-F.; Ma, R.-W.; Wei, R.-Y.; Zhang, Y.; You, Y.-M.; Ye, H.-Y.; Xiong, R.-G. J. Am. Chem. Soc. 2017, 139, 3954. (11) (a) Ye, H. Y.; Zhang, Y.; Fu, D. W.; Xiong, R. G. Angew. Chem., Int. Ed. 2014, 53, 11242. (b) Zhang, Y.; Liao, W.-Q.; Fu, D.-W.; Ye, H.-Y.; Chen, Z.-N.; Xiong, R.-G. J. Am. Chem. Soc. 2015, 137, 4928. (c) Ye, H.-Y.; Zhou, Q.; Niu, X.; Liao, W.-Q.; Fu, D.-W.; Zhang, Y.; You, Y.-M.; Wang, J.; Chen, Z.-N.; Xiong, R.-G. J. Am. Chem. Soc. 2015, 137, 13148. (d) Zhang, Y.; Liao, W. Q.; Fu, D. W.; Ye, H. Y.; Liu, C. M.; Chen, Z. N.; Xiong, R. G. Adv. Mater. 2015, 27, 3942. (12) Zhang, Y.; Ye, H.-Y.; Zhang, W.; Xiong, R.-G. Inorg. Chem. Front. 2014, 1, 118. (13) Lander, J. J. Acta Crystallogr. 1951, 4, 148. (14) Shi, P.-P.; Tang, Y.-Y.; Li, P.-F.; Liao, W.-Q.; Wang, Z.-X.; Ye, Q.; Xiong, R.-G. Chem. Soc. Rev. 2016, 45, 3811. (15) Moya, X.; Kar-Narayan, S.; Mathur, N. Nat. Mater. 2014, 13, 439.

structural tunability of organic−inorganic hybrid system, new ferroelectric/antiferroelectric materials with exotic physical properties can be expected.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.7b04693. Supplementary methods and figures (PDF) Crystallographic information (CIF) Crystallographic information (CIF) Crystallographic information (CIF)

■

AUTHOR INFORMATION

Corresponding Authors

*[email protected] *[email protected] ORCID

Yi Zhang: 0000-0002-6375-1712 Ren-Gen Xiong: 0000-0003-2364-0193 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by 973 project (2014CB932103) and the National Natural Science Foundation of China (21290172, 91422301, 21427801, and 21522101).

■

REFERENCES

(1) (a) Xu, B.; Chakraborty, H.; Remsing, R. C.; Klein, M. L.; Ren, S. Adv. Mater. 2017, 29, 1605150. (b) Dai, Q.; Patel, K.; Donatelli, G.; Ren, S. Angew. Chem., Int. Ed. 2016, 55, 10439. (2) (a) Yao, Z.; Song, Z.; Hao, H.; Yu, Z.; Cao, M.; Zhang, S.; Lanagan, M. T.; Liu, H. Adv. Mater. 2017, 29, 1. (b) Love, G. R. J. Am. Ceram. Soc. 1990, 73, 323. (c) Dang, Z. M.; Yuan, J. K.; Yao, S. H.; Liao, R. J. Adv. Mater. 2013, 25, 6334. (d) Lu, X.-Z.; Rondinelli, J. M. Nat. Mater. 2016, 15, 951. (3) (a) Lines, M. E.; Glass, A. M. Principles and Applications of Ferroelectrics and Related Materials; Oxford University Press, 1977. (b) Kittel, C. Phys. Rev. 1951, 82, 729. (4) (a) Hao, X.; Zhai, J.; Yao, X. J. Am. Ceram. Soc. 2009, 92, 1133. (b) Zhang, S. T.; Kounga, A. B.; Jo, W.; Jamin, C.; Seifert, K.; Granzow, T.; Rodel, J.; Damjanovic, D. Adv. Mater. 2009, 21, 4716. (c) Tagantsev, A.; Vaideeswaran, K.; Vakhrushev, S.; Filimonov, A.; Burkovsky, R.; Shaganov, A.; Andronikova, D.; Rudskoy, A.; Baron, A.; Uchiyama, H.; et al. Nat. Commun. 2013, 4, 2229. (5) (a) Samara, G. Phys. Rev. B 1970, 1, 3777. (b) Zhelnova, O.; Fesenko, O. Ferroelectrics 1987, 75, 469. (6) Lasave, J.; Koval, S.; Dalal, N.; Migoni, R. Phys. Rev. Lett. 2007, 98, 267601. (7) Jain, P.; Dalal, N. S.; Toby, B. H.; Kroto, H. W.; Cheetham, A. K. J. Am. Chem. Soc. 2008, 130, 10450. (8) (a) Liao, W. Q.; Zhao, D. W.; Yu, Y.; Grice, C. R.; Wang, C. L.; Cimaroli, A. J.; Schulz, P.; Meng, W. W.; Zhu, K.; Xiong, R. G.; Yan, Y. F. Adv. Mater. 2016, 28, 9333. (b) Liao, W. Q.; Zhao, D. W.; Yu, Y.; Shrestha, N.; Ghimire, K.; Grice, C. R.; Wang, C. L.; Xiao, Y. Q.; Cimaroli, A. J.; Ellingson, R. J.; Podraza, N. J.; Zhu, K.; Xiong, R. G.; Yan, Y. F. J. Am. Chem. Soc. 2016, 138, 12360. (c) Li, W.; Wang, Z.; Deschler, F.; Gao, S.; Friend, R. H.; Cheetham, A. K. Nat. Rev.Mater. 2017, 2, 16099. (d) Stroppa, A.; Quarti, C.; De Angelis, F.; Picozzi, S. J. Phys. Chem. Lett. 2015, 6, 2223. (e) Saparov, B.; Mitzi, D. B. Chem. Rev. 2016, 116, 4558. (9) Liao, W.-Q.; Zhang, Y.; Hu, C.-L.; Mao, J.-G.; Ye, H.-Y.; Li, P.-F.; Huang, S. D.; Xiong, R.-G. Nat. Commun. 2015, 6, 7338. 8757

DOI: 10.1021/jacs.7b04693 J. Am. Chem. Soc. 2017, 139, 8752−8757