B′ = Cr3+, Fe3+ - American Chemical Society

Jul 10, 2014 - Structural Transition in the Perovskite-like Bimetallic Azido. Coordination Polymers: (NMe4)2[B′·B″(N3)6] (B′ = Cr3+, Fe3+; B″...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/crystal

Structural Transition in the Perovskite-like Bimetallic Azido Coordination Polymers: (NMe4)2[B′·B″(N3)6] (B′ = Cr3+, Fe3+; B″ = Na+, K+) Published as part of the Crystal Growth & Design virtual special issue IYCr 2014 - Celebrating the International Year of Crystallography Zi-Yi Du,†,‡ Ying-Ping Zhao,‡ Chun-Ting He,† Bao-Ying Wang,† Wei Xue,† Hao-Long Zhou,† Jie Bai,† Bo Huang,† Wei-Xiong Zhang,*,† and Xiao-Ming Chen*,† †

MOE Key Laboratory of Bioinorganic and Synthetic Chemistry, School of Chemistry and Chemical Engineering, Sun Yat-Sen University, Guangzhou 510275, China ‡ College of Chemistry & Chemical Engineering, Gannan Normal University, Ganzhou 341000, China W Web-Enhanced Feature * S Supporting Information *

ABSTRACT: Through in situ variable-temperature singlecrystal X-ray diffraction analysis, a solid−solid structural phase transition induced by the successive displacements of the [NMe4]+ guest and a subsequent abrupt order−disorder transform for both the cationic guest and the cage-like framework was tracked in four perovskite-like bimetallic azido coordination polymers: (NMe4)2[B′·B″(N3)6] (B′ = Cr3+, Fe3+; B″ = Na+, K+). Such structural transition was also confirmed by differential scanning calorimeter measurement, variable-temperature powder X-ray diffraction analysis, and variable-temperature dielectric permittivity measurement, as well as molecular dynamics simulation. Conclusively, these compounds provide a good host−guest model for understanding and modulating the thermal motion behavior of the [NMe4]+ guest in various confined spaces constructed by the perovskite-like azido frameworks.



INTRODUCTION Solid−solid structural phase transition in materials represents an important critical phenomenon, and it often can impart functionalities to the system, such as thermal energy storage, charge ordering, ferroelectricity, dielectric switching, and nonlinear optical switching, most of which have found practical and high-technological applications.1−9 Among various types of the solid−solid structural phase transition materials, the inorganic perovskite oxide materials are listed among the most important ones, which have been utilized as diversified functional materials such as ferroelectrics (especially multiferroics), piezoelectrics, superconductors, memory devices, sensors and catalyst electrodes in fuel cells, and base materials for high-efficiency photovoltaics.4,10 Considering the versatility of the well-studied inorganic perovskites, molecule-based ABY3type coordination polymers (CPs) mimicking the inorganic ABO3-type perovskite structure have aroused great interest, from both a theoretical and a practical point of view. For instance, very recently some perovskite-like CPs as the promising light harvester were reviewed by Ahmad et al.11 The perovskite-like CPs are usually assembled by the inclusion of guest species into well-matched host cages, and upon external stimulus such as temperature or pressure, the © XXXX American Chemical Society

induced order−disorder change of the cation guest in the wellmatched cage-like frameworks often can be the dominant driving force for structural phase transitions. The designable/ tunable characteristics by chemical modifications make them promising candidates for developing new types of solid−solid phase transition materials with multifunctionality. However, up to now, reports on the phase transition behaviors of such CPs remain very scarce.7,11−19 The ligands employed mainly contain the monatomic I−, diatomic CN−, and multiatomic HCOO−, N3− anions, which act as monovalent end-to-end bridging ligands. In the current studies, we are interested in searching new types of solid−solid structural phase transition materials in the perovskite-like CPs, and our recent efforts have yielded an example of such compounds, exhibiting unique ferroelastic phase transition.20 In this contribution, we present a series of perovskite-like bimetallic azido coordination polymers, namely, (NMe 4 ) 2 [CrNa(N 3 ) 6 ] (1), (NMe 4 ) 2 [CrK(N 3 ) 6 ] (2), (NMe4)2[FeNa(N3)6] (3), and (NMe4)2[FeK(N3)6] (4), all Received: April 7, 2014 Revised: June 26, 2014

A

dx.doi.org/10.1021/cg5004676 | Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design

Article

Table 1. Summary of Crystal Data and Structural Refinements for Compound 4 at Different Temperatures 4, formula = C8H24FeKN20, fw = 495.42 T/K CCDC space group a/Ǻ V/Å3 Z Dcalcd/g cm−3 μ/mm−1 GOF on F2 R1, wR2 [I > 2σ(I)]a R1, wR2 (all data)a a

150(2) 987337 Pa3̅ 13.1026(4) 2249.4(1) 4 1.463 0.896 1.086 0.0300, 0.0810 0.0346, 0.0848

200(2) 987338 Pa3̅ 13.1319(4) 2264.6(1) 4 1.453 0.890 1.091 0.0322, 0.0882 0.0380, 0.0917

240(2) 987339 Pa3̅ 13.1676(5) 2283.1(2) 4 1.441 0.883 1.081 0.0337, 0.0920 0.0427, 0.0966

265(2) 987340 Pa3̅ 13.2019(13) 2301.0(4) 4 1.430 0.876 1.061 0.0366, 0.0944 0.0497, 0.1045

290(2) 987341 Pa3̅ 13.2235(6) 2312.3(2) 4 1.423 0.872 1.084 0.0408, 0.1060 0.0549, 0.1160

303(2) 987342 Pa3̅ 13.2400(6) 2320.9(2) 4 1.418 0.868 1.056 0.0465, 0.1275 0.0608, 0.1382

311(2) 987343 Pa3̅ 13.2530(6) 2327.8(2) 4 1.414 0.866 1.061 0.0480, 0.1189 0.0630, 0.1305

330(2) 987344 Fm3m ̅ 13.351(1) 2379.9(3) 4 1.383 0.847 1.055 0.0322, 0.0735 0.0338, 0.0755

348(2) 987345 Fm3m ̅ 13.3607(8) 2385.0(2) 4 1.380 0.845 1.047 0.0359, 0.0852 0.0381, 0.0865

R1 = ∑||F0| − |Fc||/∑|F0|, wR2 = {∑w[(F0)2 − (Fc)2]2/∑w[(F0)2]2}1/2.

of which contain a cationic [NMe4]+ guest in their cage-like frameworks. It is noted that the structure of 3 (obtained as a byproduct) at 123 K has been reported by Zhou and coworkers,21 and the Mö ssbauer spectra of it had been documented much earlier.22 An unprecedented structural transition upon temperature stimulus is found in these compounds, in which the successive displacements of the [NMe4]+ guest, and a subsequent abrupt, dynamic rotation of the [NMe4]+ guest accompanying by a synergic sway of the host [B′0.5·B″0.5(N3)3]+ framework (B′ = Cr3+, Fe3+; B″ = Na+, K+) trigger the structural phase transition. Herein, we report their syntheses, crystal structures, structural transition mechanisms, and dielectric properties.



Syntheses of Compounds 3 and 4. Iron(III) nitrate nonahydrate (0.3 mmol) was dissolved in an aqueous solution of tetramethylammonium nitrate (4 mL, 2.0 mmol), then an aqueous solution of sodium azide (2 mL, 3.0 mmol) was added dropwise into the above solution, and the resultant deep red solution was allowed to stand at room temperature. A few minutes later, red block-shaped crystals of 3 were deposited, in a ca. 85% yield based on Fe. IR data (KBr, cm−1): 3382(m), 3028(m), 2061(s), 1483(s), 1446(m), 1349(s), 1284(m), 949(s), 638(m), 607(m). When a similar procedure was performed but with potassium azide in place of sodium azide, red block-shaped crystals of 4 were obtained in a ca. 80% yield based on Fe. IR data (KBr, cm−1): 3375(m), 3026(m), 2056(s), 1481(s), 1446(m), 1347(s), 1277(m), 949(s), 638(m), 607(m). Caution! Although our samples never exploded during handing, azide metal complexes are potentially explosive. Only a small amount of material should be prepared, and it should be handled with great caution. Single-Crystal X-ray Crystallography. The in situ variabletemperature single-crystal X-ray diffraction intensities for compounds 1−2 and 3−4 were collected on an Oxford Gemini S Ultra CCD diffractometer or a Rigaku R-AXIS SPIDER IP diffractometer, respectively, with graphite-monochromated Mo−Kα radiation (λ= 0.71073 Å). The measurement sequence for each crystal sample at different temperatures followed a continuous heating process, except that the crystallographic data of compound 4 at 311 K were collected after the experience of a slow cooling course starting from 323 K. The measurement temperatures were controlled by a dry nitrogen open flow using a Cryojet controller of Oxford Instrument or a Rigaku Gas Flow GN2 apparatus and corrected by a thermal couple at the crystal position. Absorption corrections were applied by using CrysAlisPro or ABSCOR.23,24 The structures were solved with direct methods and refined with a full-matrix least-squares technique with the SHELXTL program package.25 Anisotropic atom displacement parameters were applied to all non-hydrogen atoms. All the hydrogen atoms were generated geometrically, and the hydrogen atoms of disordered tetramethylammonium are not included in the refinements. Crystallographic data and structural refinements for 1−4 under different temperatures are summarized in Table 1 and Tables S1−S3. Selected bond lengths are listed in Table 2 and Tables S4−S6. More details about the crystallographic data have been deposited as Supporting Information. Classical Molecular Dynamics (MD) Simulation. The classical MD simulations for compound 4 were performed using the Forcite program implemented in the Materials Studio 5.0 package,26 in a constant-volume and constant-temperature ensemble (NVT) with the temperature controlled at 270 and 430 K, respectively. The total MD simulation time was 4 ns.

EXPERIMENTAL SECTION

Materials and Instrumentations. All chemicals were obtained from commercial sources and used without further purification. Fourier transform infrared (FT-IR) spectra were recorded on a PerkinElmer Spectrum One FT-IR spectrometer using KBr pellets from 4000 to 400 cm−1. Powder X-ray diffraction (PXRD) patterns (Cu−Kα) were collected on a Bruker Advance D8 θ−2θ diffractometer. Thermogravimetric analyses (TGA) were carried out on a TA Q50 system at a heating rate of 5 K min−1 under a nitrogen atmosphere. Differential scanning calorimeter (DSC) measurements for compounds 1 and 2−4 were performed by heating and cooling the powder samples on a NETZSCH DSC 204F1 instrument and a PerkinElmer DSC-7 instrument, respectively. The measurements were carried out under a nitrogen atmosphere in aluminum crucibles with a heating/cooling rate of 5 K min−1, in the temperature range of 280− 340 K. The complex permittivities were measured under a nitrogen atmosphere, using a Tonghui TH2828A LCR meter in a Mercury iTC cryogenic environment controller of Oxford Instrument, and the samples were ground and pressed into tablets under a pressure of ca. 9 MPa. Syntheses of Compounds 1 and 2. Chromium(III) nitrate nonahydrate (0.2 mmol) was dissolved in an aqueous solution of tetramethylammonium nitrate (4 mL, 1.5 mmol), then sodium azide (2.0 mmol) was added into the above solution under stirring, and the resultant turbid liquid was filtered to afford a dark green solution, which was allowed to stand at room temperature. One day later, green block-shaped crystals of 1 were deposited from the filtrates, in a ca. 65% yield based on Cr. IR data (KBr, cm−1): 3371(m), 3028(m), 2067(s), 1481(s), 1446(m), 1354(s), 1288(m), 952(s), 652(m), 605(m). When a similar procedure was performed but with potassium azide in place of sodium azide, green block-shaped crystals of 2 were obtained in a ca. 60% yield based on Cr. IR data (KBr, cm−1): 3379(m), 3027(m), 2062(s), 1484(s), 1444(m), 1352(s), 1300(m), 950(s), 660(m), 606(m).



RESULTS AND DISCUSSION Thermal Analysis. Compounds 1−4 are stable when exposed to air, and the TGA analyses show that they all are B

dx.doi.org/10.1021/cg5004676 | Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design

Article

centrosymmetric space group Pa3̅ as the α phase, and in Fm3m ̅ as the β phase. Such phase transition is also clearly confirmed by the in situ variable-temperature PXRD (Figure 2 and Figure

Table 2. Selected Bond Lengths and The Shortest Fe···K Distance (Å) for Compound 4 at Different Temperatures T/K

Fe−N

K−N

N−N

Fe···K

150 200 240 265 290 303 311 330 348

2.083(1) 2.082(1) 2.082(1) 2.083(2) 2.083(2) 2.081(2) 2.079(2) 2.078(6) 2.076(9)

2.775(1) 2.780(2) 2.786(2) 2.793(2) 2.795(2) 2.797(3) 2.797(3) 2.781(8) 2.783(9)

1.161(2), 1.196(2) 1.158(2), 1.193(2) 1.156(2), 1.191(2) 1.154(3), 1.191(3) 1.149(3), 1.192(3) 1.145(4), 1.196(3) 1.142(4), 1.205(4) 1.182(8), 1.18(2) 1.17(2), 1.18(2)

6.551 6.566 6.584 6.601 6.612 6.620 6.626 6.676 6.680

stable up to 200 °C. (Note: To avoid the potential explosive decomposition of the azide metal complexes, a further heating process over 200 °C was not performed.) DSC measurement was used to detect the possible phase transition and to confirm the existence of heat anomaly. As shown in Figure 1, the DSC

Figure 2. Variable-temperature PXRD patterns of compound 4.

S1, Supporting Information), as shown by the systematic absences of some diffraction peaks in the β phase, i.e., (3, 0, 0), (4, 1, 0), and (4, 2, 1). The experimental patterns in both phases match well with the simulated ones based on the single crystal structures, but the experimental 2θ values of the corresponding diffraction peaks at 283 K all are slightly lower (∼0.16°) than those of the simulated one at 150 K, implying that the crystal structure of the α phase varies a little bit at different temperatures. The PXRD patterns also indicate that the powders of all these compounds are thermally stable at 393 K as the β phase, upon which a further test was not performed. Compounds 1−4 are isomorphous and all have similar phase transition behavior; hence only the structure of 4 will be discussed in detail as a representative. The crystal structure of 4 in both phases can be roughly described as a distorted perovskite-like structure, with a general formula of A(B′0.5B″0.5)Y3 (A = monovalent guest cation, B′/B″ = trivalent/monovalent metal cations, arranged orderly and alternately over two types of B sublattices, in a chess-board manner, and Y = monovalent anion). Each Fe3+ and K+ ion here is surrounded by six N atoms from six azido ions, all of which act as end-to-end bridging ligands between the heterometallic Fe3+ and K+ ions, thus leading to a threedimensional cage-like framework (Figure 3). The common structural feature of compound 4 is the anionic [Fe0.5K0.5(N3)3]− cage enclosed by 12 Fe−N−N−N−K fragments, within which the guest [N(CH3)4]+ cation resides. It is noted that the Fe−N bond is strong and coordinative, but the K−N bond is much weaker and ionic in compound 4, with their bond lengths being 2.083(2) and 2.795(2) Å at 290 K, respectively. The structural difference between the α and β phases of compound 4 can mainly be attributed to the sway of the N3− bridge and the rotation of the [NMe4]+ guest (Figure 3). In the α phase, both of the Fe3+ and K+ ions locate at inversion centers and 3-fold axes, and the [NMe4]+ guest lies on a 3-fold axis, while the crystallographically independent N3− ligand situates at a general position. The above three kinds of cations and one unique anion all are ordered as a “frozen” state in the α phase. However, in the β phase, the [NMe4]+ cation and the N3− ligand turn disordered as a “melt-like” state: (1) The tetrahedral-like [NMe4]+ cation locates at the cage center and rotates dynamically, nearly over arbitrary orientation; (2) The rod-like N3− ligand blurs over eight positions, as required by the

Figure 1. DSC curves of compounds 1−4 (from top to bottom, respectively), obtained on heating/cooling cycles.

curves of all compounds show a pair of endothermic/ exothermic peaks at heating/cooling runs, revealing that each of them can undergo a reversible phase transition above room temperature (vide infra). It is noted that the endothermic/ exothermic peaks in compound 2 consist of two highly overlapped peaks, implying that such phase transition experienced a two-step process. For convenience, we label the phase below and above TC(heating)/TC(cooling) as the α and β phase, respectively. For such phase transitions, the sharp peak shape, prominent enthalpy change, and a thermal hysteresis of ca. 2−4 K between the heating and cooling runs reveal the discontinuous character of the transition, being indicative of a first-order phase transition. Crystal Structures, Structural Transitions, and Mechanism. In order to understand the phase transition, the crystal structures of compounds 1−4 were determined at different temperatures. The in situ variable-temperature single-crystal Xray diffraction analysis reveals that they all crystallize in the C

dx.doi.org/10.1021/cg5004676 | Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design

Article

Figure 3. Crystal structures of compound 4 at 150 K (a) and 348 K (b). Fe, K, N, and C atoms are shaded in green, orange, blue, and black, respectively. The dynamically disordered bonds at 348 K are represented by thin lines. The diagonal lines in the cage unit, with/without a 3-fold symmetry, are drawn as solid/dotted purple lines, respectively.

Figure 4. (a) Temperature dependence of d[OM−OTA] in the structure of compound 4; (b) temperature-dependent void space for ammonium guest per cage in 4α, and temperature-dependent Ueq (equivalent isotropic atomic displacement parameter) for C1/C2 atoms (□/Δ) of ammonium guest in 4α, based on in situ variable-temperature single-crystal X-ray diffraction analysis.

pseudocubic Fm3̅m imposed symmetry. As a result, the two heterometallic Fe3+ and K+ cations bridged by the N3− ligand also rotate dynamically to cater for the sway of the N3− bridges. The synergic disorder of these components leads to the change of symmetry elements in the perovskite-like structure, with the total symmetric elements of the crystallographic point group increasing from 24 (E, 8C3, 3C2, i, 8S6, 3σh) to 48 (E, 8C3, 3C2, 6C′2, 6C4, i, 8S6, 3σh, 6σd, 6S4). Correspondingly, the occupancy factors of all components are reduced to 1/48 occupied for Fe3+ and K+ ions, 1/8 occupied for N3− ligand, and 1/24 occupied for [NMe4]+ cation, respectively. Furthermore, spatial symmetric operation numbers decrease from 48 to 24 during the symmetry breaking process of β → α (see Figure S2, Supporting Information),27 in good agreement with macro-

scopic symmetry breaking. Overall, the host−guest interaction accompanying with the synergic order−disorder transitions of the [Fe0.5K0.5(N3)3]− cage and the [NMe4]+ guest during the heating/cooling process results in such structural phase transition. It is also worth noting that although compound 4 crystallizes in the α phase for a wide temperature range, a close inspection reveals that the [NMe 4 ] + guest confined in the [Fe0.5K0.5(N3)3]− cage unit shows successive displacements during the heating/cooling process, implying that slight changes of dipole moment exist between the host cage anion and the guest cation. As shown in Figures 3a and 4a, the positive charge center of the metal ions within a cage unit (denoted as OM) just locates at the intersection of four D

dx.doi.org/10.1021/cg5004676 | Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design

Article

Table 3. Summary of Thermal Properties for Compounds 1−6, by DSC Measurements, at a Heating/Cooling Rate of 5 K·Min−1

a

compound

1

2

3

4

5a

6a

TC(heating)/K TC(cooling)/K thermal hysteresis/K ΔH(heating)/KJ·mol−1 ΔS(heating)/J·K−1·mol−1 N(heating) (ln N = ΔS/R)

312.1 308.2 3.9 14.84 47.56 305

313.7 311.8 1.9 14.46 46.10 256

304.0 301.2 2.8 15.82 52.05 524

315.1 312.6 2.5 14.92 47.36 298

308 303 5 5.72 18.6 9.4

322 318 4 3.25 10.1 3.4

Here the thermal properties of 5 and 6 refer to the phase transition of Pm3̅m ↔ P21/m.19,20

Figure 5. Temperature dependence of the real part (ε′) and imaginary part (ε″) of the complex permittivity for powder-pressed pellets of compounds 2 (a) and 4 (b), at 100 (black), 300 (red), 600 (blue), and 800 (green) kHz, on a very slow cooling process of ∼8 h.

ments, as quantitative changes, to generate the final qualitative change. To further track the critical phenomenon, it is also worth mentioning that a spontaneous “merohedral twinning” phenomenon occurs from the β phase to the α phase during the cooling process. For instance, an as-synthesized singlecrystal of 4α was randomly chosen and measured at 298 and 311 K, and no twinning phenomenon was detected. However, when we raised the test temperature for this single-domain crystal of 4α to 323 K, at which 4β is stable, then slowly cooled it down to 311, 298, 283, and 250 K, respectively, four crystallographic data sets were collected for 4α again. The resultant crystal data sets proved to be twinning structures, in which a twin law of “0 1 0 1 0 0 0 0 −1” was identified, and the refined BASF values were 0.444(3), 0.553(4), 0.444(3), 0.557(3), respectively.25 Such a phenomenon can be ascribed to the formation of crystal domains of different orientations as the F-lattice of β phase turns to the P-lattice of α phase, leading to two twin domains. The driving force for phase transition α → β is mainly ascribed to the high disorder of the N3− bridges as well as the [NMe4]+ guest. The entropy change (ΔS) of these phase transitions during the heating process can be estimated from the DSC measurement to be around 50 J·K−1·mol−1 (Table 3). From the Boltzmann equation, ΔS = R ln N, where N represents the ratio of possible configurations and R is the gas

diagonals, while the positive charge center of the tetramethylammonium guest (denoted as OTA) nearly situates on its N atom, and for charge balance, the negative charge center of the azido bridges in a cage unit (denoted as OAz) dwells between OM and OTA. In the α phase, OM and OTA do not coincide, and as the temperature increases, OTA moves along the 3-fold axis toward OM, with d(OM−OTA) decreasing from 0.602(1) Å at 150 K to 0.521(3) Å at 311 K. Meanwhile, according to the calculations by Platon,28 the void space that the framework provides for each [NMe4]+ guest expands slightly from 142.5 Å3 at 150 K to 153.0 Å3 at 311 K (Figure 4b), and the thermal ellipsoids for the C atoms of [NMe4]+ guest also expand gradually, with the values of Ueq (equivalent isotropic atomic displacement parameter) for C1/C2 atoms increasing from 0.035/0.035 at 150 K to 0.107/0.099 at 311 K. By and large, the [Fe0.5K0.5(N3)3]− cage exhibits a high flexibility, with its volume expanding 3.5% as the temperature increases from 150 to 311 K, and a further expansion of 2.5% happens as the temperature reaches up to 348 K of the β phase. Overall, a solid−solid structural phase transition induced by the successive displacements and a subsequent abrupt order− disorder transform mechanism were uncovered for the unprecedented perovskite-like bimetallic azido CPs, and such a phase transition is a rare type of phase transition for CPs, which requires a gradual accumulation of dynamic displaceE

dx.doi.org/10.1021/cg5004676 | Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design

Article

constant. It is found that the value of N ranges from 256 to 524 in compounds 1−4. The very large configuration values indicate an extremely high degree of disorder in the β phase, which originates from the dynamical rotation of the [NMe4]+ cation as well as the intense swing of the N3− anions. In addition, the phase transition behavior is also confirmed by the classical molecular dynamics simulations. As visually shown in Movie S1, all the atoms in crystal 4 vibrate slightly in the vicinity of their equilibrium positions at 270 K. In contrast, the sway of the N3− bridges and the rotation of the [NMe4]+ guest are obviously observed at 430 K (Movie S2). Dielectric Properties and Their Relevance to Structural Transitions. The temperature-dependent traces of the complex dielectric permittivities were measured for compounds 1−4 on powder-pressed pellets, at 100, 300, 600, and 800 kHz (Figure 5 and Figure S3). The dielectric permittivities (ε′) for 1−4 at 100 kHz and 296 K are 4.71, 6.03, 9.34, and 5.49, respectively. Since each of the two pairs of compounds, i.e., 1 and 2, as well as 3 and 4, has similar temperature dependence for the dielectric permittivity, only those of 2 and 4 will be discussed in detail. As shown in Figure 5, upon a very slow cooling process, the dielectric permittivities (ε′) of 2 and 4 at all frequencies show turning points at 313.6 and 313.1 K, respectively, corresponding well to the phase transition points mentioned above. In the α phase, ε′ of 2 under 100/300 kHz decreased gradually from 6.10/4.88 at 303 K to 5.43/4.55 at 213 K. However, an opposite trend can be observed for ε′ under 600/800 kHz, which increases slowly from 4.06/3.81 at 303 K to 4.23/4.19 at 213 K. The temperature-dependent permittivity within the wide temperature range of the α phase is consistent with the fact that the guest cation shifts slightly but successively during the heating/cooling process, as mentioned above. Although a quantitative analysis is beyond the scope of this article, the opposite variation trends at different frequencies at least indicate that multiple factors, such as the guest cation and the inorganic framework, contribute to ε′ but asynchronously. For compound 4, ε′ at 100/300/600 kHz have a similar decreasing trend upon cooling within the α phase, but varies slightly at 800 kHz. On the other hand, the imaginary part (ε″) of the complex permittivities for all compounds at 600/800 kHz are remarkable and decreased gradually on cooling, but become very small and temperature-independent at 100/300 kHz, showing limited dielectric loss at the relatively low frequency band. On the whole, although compounds 1−4 are isomorphous and all have a similar phase transition behavior, the detailed dielectric properties of them are diversified and are hard to be predicted to a certain extent, reflecting that subtle changes in the coordination parameters of these compounds can sharply influence their dielectric properties. Comparisons of the Different Perovskite-like Azido CPs with [NMe4]+ as the Cation Guest. Up to now, three homometallic perovskite-like azido CPs with [NMe4]+ as the cation guest,19,20,29,30 namely, (NMe4)[B(N3)6] (B = Mn2+, Cd2+ or Ca2+, labeled as compounds 5, 6, and 7, respectively), and four bimetallic ones here, namely, (NMe4)2[B′·B″(N3)6] (B′ = Cr3+, Fe3+; B″ = Na+, K+), have been obtained; hence the comparisons of these different perovskite-like azido CPs are meaningful: (i) The general formula of the bimetallic A(B′0.5B″0.5)Y3 can be derived from the classical homometallic ABY3, where the combined charges of B′3+0.5·B″+0.5 in the former are equal to B2+ of the latter. It is also worth noting that the heterometallic ions, B′3+ and B″+, are orderly arranged over two types of sublattices,

which is essentially different from the substitutional disorder in some doping system. Hence, such compounds can be viewed as double-perovskite-like CPs. As a result, the cell lengths in the heterometallic 1−4 are roughly twice as long as those in the cubic phase of the homometallic 5−7. More importantly, the incorporation of bimetallic ions in the perovskite-like system provides an additional variable to induce a greater degree of conformational changes during the phase transition process, which is confirmed by the larger value of entropy change in the former ones (Table 3). (ii) Different from the totally coordination framework of the homometallic 5−6, the Cr−N or Fe−N bonds are strong and coordinative, but the Na−N or K−N bonds are much weaker and ionic in the heterometallic 1−4. One advantage of introducing the ionic Na−N or K−N bonds into the perovskite-like azido system is to utilize their less directional characteristic which is different from the specific directions of the coordination bonds, thus facilitating the sway of the N3− bridges more readily. It is noted that although the Ca−N bond in the homometallic 7 is also ionic, such compound decomposed slowly under ambient conditions.30 (iii) At the temperature range of 150−393 K, compound 6 undergoes three reversible phase transitions, while compounds 1−4 and 5 each only experiences one reversible phase transitions upon heating/cooling runs. In addition, different phase transition behaviors can be found for compounds 1−4, 5, and 6, which crystallize in the space groups Pa3̅, P21/m, and C2/c at 150 K, respectively. After phase transitions, compounds 1−4 and 5−6 crystallize in the space groups Fm3̅m and Pm3̅m at 333 K, respectively. For 7, it was known to crystallize in the space groups P4/nmm at room temperature, but its phase transition behavior has not been investigated yet. (iv) The cumulative bond lengths of M−N−N−N−M′ in compounds 1−5 at 333 K and 6 at 350 K alter in a wide range of 6.732−7.226 Å, while the corresponding M···M′ distances fluctuate in a somewhat narrow range of 6.450−6.676 Å (Table S7, Supporting Information), reflecting that these cage-like frameworks can automatically adjust their host size to match the [N(CH3)4]+ guest. In fact, if the N3− bridge is replaced by a somewhat shorter CN− linker, perovskite-like CPs with the same [N(CH3)4]+ guest can also be afforded in a wide range. 31−36 Surprisingly, although many of the cyanocompounds crystallizing in various space groups have been documented at the last century, their phase transition behaviors were seldom investigated. We anticipate that these compounds may also exhibit rich phase transition behaviors, if the [N(CH3)4]+ guest can motion/rotation in the confined cagelike space upon temperature stimulus. (v) The phase transition mechanisms of these compounds are substantially similar, owing to the sway of the rod-like N3− bridge as well as the rotation of the tetrahedral-like [NMe4]+ guest. However, the [NMe4]+ guest itself in these crystal structures can show various symmetries, such as 2-, 3-, or 4-fold axis, and mirror symmetry, and even in the disordered state, the degree of disorder in them is also distinguishable based on the different crystallographic symmetries of the various structures. Similarly, the N3− bridge also can exhibit different degrees of disorder, such as 2-fold, quadruple, and octuplet disorder. Thus, their detailed mechanisms differ with diversified manifestations, reflecting that the subtle changes in the coordination parameters can modulate the flexibility of the perovskite-like azido systems and tune its structural phase transitions as well as physical properties. F

dx.doi.org/10.1021/cg5004676 | Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design



Article

(9) Chen, S.; Shang, R.; Hu, K.-L.; Wang, Z.-M.; Gao, S. Inorg. Chem. Front 2014, 1, 83. (10) Properties and Applications of Perovskite-Type Oxides; Tejuca, L. G., Fierro, J. L. G., Eds.; Marcel Dekker, Inc.: New York, 1993. (11) Kazim, S.; Nazeeruddin, M. K.; Nazeeruddin, M. K.; Grätzel, M.; Ahmad, S. Angew. Chem., Int. Ed. 2014, 53, 2. (12) Mautner, F. A.; Cortes, R.; Lezama, L.; Rojo, T. Angew. Chem., Int. Ed. 1996, 35, 78. (13) Wang, Z.; Zhang, B.; Otsuka, T.; Inoue, K.; Kobayashi, H.; Kurmoo, M. Dalton Trans. 2004, 2209. (14) Jain, P.; Dalal, N. S.; Toby, B. H.; Kroto, H. W.; Cheetham, A. K. J. Am. Chem. Soc. 2008, 130, 10450. (15) Zhang, W.; Cai, Y.; Xiong, R.-G.; Yoshikawa, H.; Awaga, K. Angew. Chem., Int. Ed. 2010, 49, 6608. (16) Fu, D.-W.; Zhang, W.; Cai, H.-L.; Zhang, Y.; Xiong, R.-G.; Huang, S. D.; Nakamura, T. Angew. Chem., Int. Ed. 2011, 50, 11947. (17) Takahashi, Y.; Obara, R.; Lin, Z.-Z.; Takahashi, Y.; Naito, T.; Inabe, T.; Ishibashi, S.; Terakura, K. Dalton Trans. 2011, 40, 5563. (18) Asaji, T.; Ito, Y.; Seliger, J.; Ž agar, V.; Gradiŝek, A.; Apih, T. J. Phys. Chem. A 2012, 116, 12422. (19) Zhao, X.-H.; Huang, X.-C.; Zhang, S.-L.; Shao, D.; Wei, H.-Y.; Wang, X.-Y. J. Am. Chem. Soc. 2013, 135, 16006. (20) Du, Z.-Y.; Zhao, Y.-P.; Zhang, W.-X.; Zhou, H.-L.; He, C.-T.; Xue, W.; Wang, B.-Y.; Chen, X.-M. Chem. Commun. 2014, 50, 1989. (21) Zhou, A.-J.; Chen, G.-Q.; Liang, J.-J.; Chen, H.-Y.; Mai, Q.-W.; Li, M. J. Guangzhou University (Natural Science Edition) 2012, 11, 21. (22) Ekkehard, F.; Peter, K. Z. Anorg. Allg. Chem. 1967, 350, 263. (23) CrysAlisPro; Agilent Technologies: Yarnton, Oxfordshire, England, 2010. (24) Higashi, T. ABSCOR; Rigaku Corporation: Tokyo, Japan. 1995. (25) Sheldrick, G. M. SHELX-96 Program for Crystal Structure Determination, 1996. (26) Accelrys, Materials Studio Getting Started, release 5.0; Accelrys Software, Inc.: San Diego, CA, 2009. (27) International Tables for Crystallography, 5th ed.; Hahn, T., Ed.; Springer: Dordrecht, Netherlands, 2002; Vol. A. (28) Spek, A. L. Platon: A Multi-purpose Crystallographic Tool; Utrecht University: Utrecht, The Netherlands, 2001. (29) Hanna, S.; Mautner, F. A.; Koppelhuber-Bitschnau, B.; AbuYoussef, M. A. M. Mater. Sci. Forum 2000, 321−324, 1098. (30) Mautner, F. A.; Krischner, H.; Kratky, C. Monatsh. Chem. 1988, 119, 1245. (31) Babel, D. Z. Naturforsch. 1982, 37b, 1534. (32) Morales, A. D.; Romero, R. G.; Rodriguez, J. D.; Hernandez, R. P.; Bertran, J. F. Transition Met. Chem. 1990, 15, 106. (33) Ziegler, B.; Witzel, M.; Babel, D. Z. Anorg. Allg. Chem. 1991, 600, 239. (34) Peschel, S.; Babel, D. Z. Naturforsch. 1994, 49b, 1373. (35) Schwarten, M.; Ziegler, B.; Witzel, M.; Babel, D. Z. Naturforsch. 1997, 52b, 391. (36) Schwarten, M.; Babel, D. Z. Anorg. Allg. Chem. 2000, 626, 1921.

CONCLUSION Through in situ variable-temperature single-crystal X-ray diffraction analysis, combined with DSC measurement, variable-temperature PXRD analysis, and variable-temperature dielectric permittivity measurement, a solid−solid structural phase transition induced by the successive displacements of the [NMe4]+ guest and a subsequent abrupt, dynamic rotation of the [NMe4]+ guest accompanied by a synergic sway of the host [B′0.5·B″0.5(N3)3]+ framework (B′ = Cr3+, Fe3+; B″ = Na+, K+) were uncovered for the perovskite-like bimetallic azido coordination polymers. Obviously, variable-temperature single-crystal X-ray diffraction analysis is a powerful tool to track and disclose the fine structural changes in crystalline solid− solid structural transition materials. In addition, these compounds provide a good host−guest model for understanding and modulating the thermal motion behavior of the [NMe4]+ guest in various confined spaces constructed by the perovskite-like azido frameworks. Future efforts will be focused on the thermal motion and especially the polarization behavior of other strong-dipole-moment cation guests in such cage-like coordination polymers.



ASSOCIATED CONTENT

S Supporting Information *

Details of crystal data and characterizations. CCDC 987319− 987345. These materials are available free of charge via the Internet at http://pubs.acs.org. W Web-Enhanced Features *

Movies in mpg format of the classical molecular dynamics simulations are available.



AUTHOR INFORMATION

Corresponding Authors

*(W.-X.Z.) E-mail: [email protected]. *(X.-M.C.) E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was supported by the NSFC (21290173, 21121061 and 21301198), the 973 Project (2012CB821706), and NSF of Guangdong (S2012030006240). W.-X.Z. is grateful for “100 Talents Program of SYSU” startup funding. Z.-Y.D. is thankful to the NSFC (21361002) and China Postdoctoral Science Foundation funded project (2013M531888).



REFERENCES

(1) Collings, I. E.; Cairns, A. B.; Thompson, A. L.; Parker, J. E.; Tang, C. C.; Tucker, M. G.; Catafesta, C.; Levelut, C.; Haines, J.; Dmitriev, V.; Pattison, P.; Goodwin, A. L. J. Am. Chem. Soc. 2013, 135, 7610. (2) Kenisarin, M. M.; Kenisarina, K. M. Renewable Sustainable Energy Rev. 2012, 16, 1999. (3) Coskun, A.; Banaszak, M.; Astumian, R. D.; Stoddart, J. F.; Grzybowski, B. A. Chem. Soc. Rev. 2012, 41, 19. (4) Peña, M. A.; Fierro, J. L. G. Chem. Rev. 2001, 101, 1981. (5) Zhang, W.; Xiong, R.-G. Chem. Rev. 2012, 112, 1163. (6) Fu, D.-W.; Cai, H.-L.; Liu, Y.; Ye, Q.; Zhang, W.; Zhang, Y.; Chen, X.-Y.; Giovannetti, G.; Capone, M.; Li, J.; Xiong, R.-G. Science 2013, 339, 425. (7) Zhang, W.; Ye, H.-Y.; Graf, R.; Spiess, H. W.; Yao, Y.-F.; Zhu, R.Q.; Xiong, R.-G. J. Am. Chem. Soc. 2013, 135, 5230. (8) Sun, Z.; Luo, J.; Zhang, S.; Ji, C.; Zhou, L.; Li, S.; Deng, F.; Hong, M. Adv. Mater. 2013, 25, 4159. G

dx.doi.org/10.1021/cg5004676 | Cryst. Growth Des. XXXX, XXX, XXX−XXX