English

• 0.   Introduction

  Crystalline spin-crossover (SCO) materials with tunable spin states under external perturbations such as temperature, light irradiation, pressure or magnetic field have been proved with attractive potential in display devices and data storage^[1-2]. Compared with SCO in the solid state, the spin state switching in solution at room temperature can offer intriguing applications in solution-based chemosensing and switchable MRI contrast agents^[3]. Nevertheless, for thermally induced spin transitions of molecules in solution, it often shows gradual and incomplete spin equilibrium due to the lack of cooperatively^[4]. In the past few years, some effective approaches have been developed to achieve the spin state switching in solution:(1) light induced cis-trans configuration transition of ligands and/or coordination number change to turn the ligand field strength and spin state of metal complexes^[5]; (2) redox process at a redox-active ligand to control spin state of metal centre in solution^[6]; (3) the special designed noncovalent intera-ctions, such as hydrogen bonding and anion binding, have emerged as powerful tools to affect the spin state of SCO complexes in solution^[7].

  For the research of anion binding and its influence on spin state of metal centres in solution, metal-organic cages with plentiful positive charges and intrinsic cavities may provide a unique platform.Since metal-organic cages have shown great affinity and selectivity in guest binding, the inner cavities of cages enable to establish abundant intermolecular interactions with the encapsulated guests^[8].

  Recently, solution SCO behaviours of supramolecular iron(Ⅱ) cubic and tetrahedral cages have been proved to be slightly affected by small molecules guest encapsulation^[9].

  However, to the best of our knowledge, the examples about encapsulated anionic guests in metal-organic cages to influence the SCO properties in solution are still scare[9b], and the realization of sensitive spin state change based on metal-organic cages from low-spin (LS) to high-spin (HS) in solution at room temperature is still a challenge.

  To develop anion dependent SCO cages, three main factors must be considered:(1) it is crucial to choose the assembly ligand with appropriate ligand field strength to insure metal centres undergoing SCO behaviour; (2) the cavity shape and volume of the rational designed cages should be suitable for capturing anionic guests; (3) the appending of hydrogen-bonding or other anion-binding groups inside the cage cavities would create chelating pockets capable of selective interaction with anions.With these in mind, three new iron(Ⅱ) tetrahedral cages with imidazole-imine (C=N) type ligands, which have appropriate ligand field strength to allow SCO for Fe(Ⅱ) centre, were prepared and characterized.Their solid state magnetic properties and halide triggered spin state switching in solution are demonstrated (Scheme 1).

  Scheme 1

  

  Scheme 1.  Schematic representation of the halide triggered spin state switching of iron(Ⅱ) tetrahedral cages

  下载: 全尺寸图片 幻灯片

  1.   Experimental

  1.1   Chemical materials and physical measurements

  All reagents and solvents were reagent grade, purchased from commercial sources and used without further purification.

  Caution:Although no problems were encoun-tered in this work, the perchlorate salt was potentially explosive.Thus, this starting material should be handled in small quantities and with great caution!

  Infrared spectra were measured with a Nicolet 6700 FT-IR spectrophotometer with ATR attachment in the range of 500~4 000 cm^-1 region.¹H NMR spectra were recorded on AVANCE Ⅲ (400 MHz) instrument at 298 K using standard Varian or Bruker software, and chemical shifts were reported downfield from tetramethylsilane.Element analyses were conducted on elementar corporation vario EL Ⅲ analyzer.UV-Vis absorbance spectra were collected on Shimadzu UV-2101 PC scanning spectrophotometer.Variable-temperature magnetic susceptibilities on crystalline samples were performed on a Quantum Design MPMS-XL-7 SQUID magnetometer with an applied magnetic field of 1 kOe over the temperature range of 2~400 K.The molar susceptibility was corrected for diamagnetic contributions using Pascal′s constants and the incre-ment method.Samples were restrained with petroleum jelly to prevent decomposing of the crystallites.Thermal gravimetric analysis (TGA) was carried out on a Waters TGA Q500 by heating the samples from 40 to 600 ℃ under nitrogen atmosphere at a heating rate of 15 ℃·min^-1.

  1.2   Preparation of the complexes

  1.2.1   Synthesis of 1, 2-di(imidazole-2-carboxaldehyde) ethane

  The 1, 2-di(imidazole-2-carboxaldehyde)ethane were synthesized according to the previously reported procedures with slightly modifications^[10]. All experi-mental details regarding the synthesis and the characterization of the compounds were presented in the supporting information.

  1.2.2   Synthesis of cages 1~3

  Cage 1:1, 2-di(imidazole-2-carboxaldehyde)ethane (43.6 mg, 0.2 mmol), (R)-1-phenylethylamine (48.9 mg, 0.4 mmol) and Fe(BF₄)₂·6H₂O (45.0 mg, 0.13 mmol) were added to a flask with 20 mL of acetonitrile in nitrogen atmosphere.The solution was stirred and heated at 80 ℃ for 2 h, cooled to room temperature.Then, the resulting purple solution was filtered.Cage 1 was precipitated as dark purple crystals through slow diffusion of diethyl ether into the filtrate at room temperature.Yield:61%.Anal.Calcd.for C₁₅₆H₁₆₈B₈F₃₂ Fe₄N₃₆(%):C, 54.07; H, 4.88; N, 14.55; Found(%):C, 54.51; H, 4.61; N, 14.74.IR (cm^-1):3 132, 3 030, 2 981, 1 614, 1 571, 1 531, 1 484, 1 441, 1 387, 1 364, 1 304, 1 054, 918, 838, 762, 706, 625.

  Cage 2:1, 2-di(imidazole-2-carboxaldehyde)ethane (43.6 mg, 0.2 mmol), (S)-1-(4-chlorophenyl)ethylamine (62.3 mg, 0.4 mmol) and Fe(ClO₄)₂·6H₂O (48.4 mg, 0.13 mmol) were added to a flask with 20 mL of acetonitrile in nitrogen atmosphere.The solution was stirred and heated at 80 ℃ for 2 h, cooled to room temperature.Then, the resulting purple solution was filtered.Cage 2 was precipitated as dark purple crystals through slow diffusion of diethyl ether into the filtrate at room temperature.Yield:48%.Anal.Calcd.for C₁₅₆H₁₅₆C_l20Fe₄N₃₆O₃₂(%): C, 47.08; H, 3.95; N, 12.67; Found(%): C, 47.39; H, 4.41; N, 12.86.IR (cm^-1): 3 134, 3 030, 2 978, 1 616, 1 566, 1 530, 1 489, 1 440, 1 385, 1 305, 1 073, 1 010, 970, 917, 831, 771, 711, 687, 619.

  Cage 3:the preparation was similar to that described for cage 2 except that (S)-1-(4-bromophenyl)ethylamine (80.0 mg, 0.4 mmol) instead of (S)-1-(4-chlorophenyl)ethylamine was used.Dark purple crystals of cage 3 were obtained.Yield:54%.Anal.Calcd. for C₁₅₆H₁₅₆Br₁₂C_l8Fe₄N₃₆O₃₂(%): C, 41.52; H, 3.48; N, 11.17; Found(%):C, 41.88; H, 3.90; N, 11.04.IR (cm^-1):3 134, 2 978, 1 622, 1 566, 1 526, 1 485, 1 434, 1 384, 1 300, 1 072, 1 008, 967, 914, 823, 759, 708, 688, 618.

  1.3   General procedure for halide titrations experiments

  Stock solutions of the cage complexes were made up in acetonitrile with the concentrations being 63 μmol·L^-1.And the stock solutions of the anionic guest were also made up in the acetonitrile with the concen-trations being 0.05 mol·L^-1 to avoid the dilution effects.The UV-Visible spectra data for the additions of the anionic guest solution using 10 μL syringe to a 3 mL of the cage complex solution (1.0 cm path length for the cuvette) were collected.The solutions were mixed by repeated inversion, and the absorption spectra were recorded after mixing for 1 min.

  1.4   General procedure for Ag(Ⅰ) back-titrations experiments

  Firstly, 50 μL tetrabutylammonium salts (0.05 mol·L^-1, CH₃CN) were added to 3 mL solution of cage complex (63 μmol·L^-1, CH₃CN).Waiting for one minute, a solution of AgBF₄ or AgClO₄ (0.05 mol·L^-1) was added step by step using 10 μL syringe, and the UV-Visible absorption spectra were recorded after mixing for 1 min.

  1.5   X-ray diffraction studies details

  The crystal structures were determined on a Siemens (Bruker) SMART CCD diffractometer using monochromated Mo Kα radiation (λ=0.071 073 nm).Cell parameters were retrieved using SMART software and refined using SAINT^[11] on all observed reflections.The highly redundant data sets were reduced using SAINT and corrected for Lorentz and polarization effects.Absorption corrections were applied using SADABS^[12] supplied by Bruker.Structures were solved by direct methods using the program SHELXL-97^[13].All of the non-hydrogen atoms except the anions were refined with anisotropic thermal displacement coefficients.Hydrogen atoms of organic ligands were located geometrically and refined in a riding model, whereas those of solvent molecules were not treated during the structural refinements.Disorder was modelled using standard crystallographic methods including constraints, restraints and rigid bodies where necessary.For cage 1, two tetrafluoroborate anions (the fluorine atom F(1B), F(2B), F(3B) and F(4B) bound to B(1B), the fluorine atom F(1F), F(2F), F(3F) and F(4F) bound to B(1F)) are disordered.For cage 2, despite rapid handling and long exposure times, the data collected are less than ideal quality, and one perchlorate anion no reasonable could be found.For cage 3, two 1-(4-bromophenyl)ethylamine group (C(33)~C(40) and Br(3)) are disorder, and the oxygen atoms bound to Cl(5) are disordered.The crystals of cages 1~3 decayed rapidly out of solvent.Nevertheless, the data for cages 1~3 are of more than sufficient quality to unambiguously establish the connectivity of the structures.Reflecting the instability of the crystals, there is a large area of smeared electron density present in the lattice.Despite many attempts to model this region of disorder as a combination of solvent molecules no reasonable fit could be found and accordingly this region was treated with the SQUEEZE^[14] function of PLATON^[15].Final crystallographic data for cages 1~3 are listed in Table 1, and the selected bond lengths (nm) and angles (°) are listed in Table S1 (Supporting Informa-tion) and anion binding interactions (nm) between the imidazole C-H groups and the encapsulated anion for cages 1~3 in Table S2.

  Table 1

  

  Table 1.  Crystallographic data for cages 1~3

  下载: 导出CSV

  	cage 1	cage 2	cage 3
  Formula	C156H168B8F32Fe4N36	C160H164Cl18Fe4N38O24	C156H156Br12Cl8Fe4N36O33.26
  Formula weight	3 465.14	3 864.79	4 533.29
  T/K	173(2)	173(2)	173(2)
  Crystal system	Triclinic	Orthorhombic	Orthorhombic
  Space group	P1	C2221	C2221
  a/nm	1.830 33(19)	2.005 18(11)	2.015 94(13)
  b/nm	1.877 5(2)	3.333 03(15)	3.347 33(17)
  c/nm	1.897 6(2)	3.434 6(2)	3.435 8(2)
  α/(°)	62.228(3)
  β/(°)	63.553(3)
  γ/(°)	77.863(3)
  V/nm3	5.1661(10)	22.954(2)	23.185(2)
  Z	1	4	4
  Dc/(Mg·m-3)	1.114	1.118	1.229
  μ/mm-1	0.354	0.516	2.468
  F(000)	1 788	7 968	9 080
  θ range/(°)	1.23~22.50	3.02~24.25	2.95~25.00
  Index ranges	-19 ≤ h ≤ 19, -19 ≤ k ≤ 20, -20 ≤ l ≤ 20	-23 ≤ h ≤ 23, -38 ≤ k ≤ 37, -35≤ l ≤39	-23 ≤ h ≤ 23, -37 ≤ k ≤ 36, -40≤ l ≤39
  Reflection collected	29 723	50 325	54 201
  GOF (F2)	1.039	1.064	1.106
  R1a, wR2b [I>2σ(I)]	0.096 3, 0.228 1	0.080 4, 0.202 8	0.065 9, 0.150 1
  R1a, wR2b (all data)	0.172 3, 0.258 5	0.115 8, 0.224 0	0.117 0, 0.166 8
  a R1=∑||Fo|-|Fc||/∑|Fo|; b wR2=[∑w(Fo2-Fc2)2/∑w(Fo2)]1/2.

  CCDC:1560377, cage 1; 1560378, cage 2; 1560379, cage 3.

  2.   Results and discussion

  2.1   Preparation and characterization of cages 1~3

  The three tetrahedral cages [Fe₄(L₁)₆](BF₄)₈ (1) (L₁=1, 2-di((imidazol-2-ylmethylene)-(R)-1-phenylethan-amine)ethane), [Fe₄(L₂)₆](ClO₄)₈ (2) (L₂=1, 2-di((imid-azol-2-ylmethylene)-(S)-1-(4-chlorophenyl)ethylamine)ethane) and [Fe₄(L₃)₆](ClO₄)₈ (3) (L₃=1, 2-di((imidazol-2 -ylmethylene)-(S)-1-(4-bromopheny)ethylamine)ethane) were obtained by the self-assembly reactions of flexible 1, 2-di(imidazole-2-carboxaldehyde)ethane, iron(Ⅱ) ions and amine components in a 3:2:6 molar ratio in aceto-nitrile solution.IR spectra of cages 1~3 show strong absorptions in the region around 1 570~1 616 cm^-1, which are typical for stretching of imidazole-imine (C=N) groups.The peaks at about 1 073 or 1 051 cm^-1 reveal the existence of ClO₄^- or BF₄^- (Fig.S3~S5).The TGA analyses showed that solvent molecules existed in the cage 1~3 (Fig.S6).

  Single crystal X-ray diffraction confirmed the edge-capped tetrahedral capsule structures for 1~3(Fig. 1).The iron(Ⅱ) centre coordinated to six nitrogen atoms from three different ligands forms a distorted octahedral FeN₆ geometry, and the four six-coordinated octahedral Fe(Ⅱ) metal nodes are bridged by six C₂-symmetric ligand linkers forming a tetrahedral cage.The metal centres occupies the vertices and the linker situates at the edges of the tetrahedron.The average Fe-N bond lengths of cages 1~3 range from 0.189 2 to 0.199 2 nm, which are typical for low-spin state Fe(Ⅱ) centre^[16].In each [Fe₄L₆]⁸⁺ cation, twelve intramolecular face-to-face π-π stacking interactions are generated between each parallel phenyl ring and imidazole ring of the adjacent ligand, which further stabilize the supramolecular structures (Fig.S7).The average center-center distances of π-πinteractions for cages 1~3 range from 0.382 1 to 0.384 7 nm.The average Fe…Fe separations along the edges lie in the range from 0.903 9 to 0.904 9 nm for cages 1~3.

  Figure 1

  

  Figure 1.  X-ray crystal structures of iron(Ⅱ) tetrahedral cages 1 (a), 2 (b) and 3 (c) showing one encapsulated anion in space-filling mode

  C:grey; N:blue; Fe:purple; B:orange; F:lime; Cl:bright green; O:red; Br:dark teal; All H atoms, solvent molecules and the remaining anions have been removed for clarity

  下载: 全尺寸图片 幻灯片

  Furthermore, it is worth noting that one of the ClO₄^- or BF₄^- anions is encapsulated at the central cavity, and the remaining anions are decorated around the periphery of cages 1~3.The encapsulated anion shows strong anion binding interactions with the cages (Fig. 2a).Each terminal F atoms (or O atoms) of the encapsulated tetrafluoroborate anion or perchlorate anion binds with three C-H protons of imidazole moieties at each tetrahedral vertex in the internal surface of the cage (Fig. 2b), while each C₂-symmetric ligand binds the encapsulated anion through two C-H…F or C-H…O interactions (Fig. 2c).The average anion binding of H…F or H…O distances for cages 1~3 range from 0.250 3 to 0.261 2 nm (Table S3).In addition, around the periphery, each tetrahedral vertex of cages decorated one anion through C-H…F or C-H…O binding between the terminal F atoms (or O atoms) of anions and the C-H protons of imidazole-imine groups (Fig.S8).The space-filling pictures derived from the crystallographic data show that there is a gap in the centre of the triangular faces and the encapsulated anion is clearly visible through the window in the centre of each face (Fig.S9~S10).Additionally, in the solid state the bridging ligands are substantially folded, and a more relaxed conforma-tion in solution would result in larger windows^[17], such that each face of the cage possesses a large enough window for the anion to diffuse into and out of the cavity.

  Figure 2

  

  Figure 2.  (a) Anion binding between the encapsulated anion and the C-H protons of imidazole moieties for cages 1~3; (b) Close-up view of one anion binding with three C-H protons of each vertex of the cages; (c) Close-up view of one anion binding with two C-H protons of each side of the cages

  All H atoms, solvent molecules, the remaining anions and the twelve substituted phenyl ring group around the periphery of four vertices have been removed for clarity

  下载: 全尺寸图片 幻灯片

  2.2   Magnetic properties

  The temperature dependent magnetic suscepti-bilities data on polycrystalline samples of cages 1~3 were collected in the cooling mode from 400 to 2 K.As shown in the Fig. 3, gradual and incomplete spin-crossover behaviours for cages 1~3 were observed.For cage 1, the χ_MT value is 10.35 cm³·K·mol^-1 at 400 K (Fig. 3a), which is lower than the expected value for four S=2 high spin Fe(Ⅱ) ions.Then, the χ_MT values gradually decreases to 0.75 cm³·K·mol^-1 at 200 K and almost constant at the low temperature, which suggests that the Fe(Ⅱ) centers are in the LS state below 200 K.For cages 2 and 3, the χ_MT values are 8.80 and 8.81 cm³·K·mol^-1 at 400 K, respectively (Fig. 3b~3c).Upon cooling from 200 to 20 K, the χ_MT value remains almost constant (proximity to 4.81 cm³·K·mol^-1 for cage 2 and 6.23 cm³·K·mol^-1 for cage 3).The change of χ_MT values below 20 K are probably due to the antiferromagnetic coupling of the objections^[18].

  Figure 3

  

  Figure 3.  Variable-temperature solid-state magnetic susceptibility for cages 1 (a), 2 (b) and 3 (c)

  下载: 全尺寸图片 幻灯片

  2.3   Spin state switching experiments

  Based on the detailed structural analysis, the three iron(Ⅱ) tetrahedral cages 1~3 possess multiple cationic charge, and the intrinsic cavities show abundant C-H…F or C-H…O binding interactions with the encapsulated anion for the stabilization of low-spin cages.These features provide a unique platform to influence the spin transition properties through anion binding in solution.The UV-Vis spectra of cages 1~3 in acetonitrile solution all exhibit metalto-ligand charge transfer (MLCT) processes with characteristic broad bands around 536 nm, which confirms the low-spin state of 1~3 in solution at room temperature (Fig.S11~S13).Since the low-spin iron(Ⅱ) is expected to feature more intense MLCT transitions, the spin state change in principle could be followed by monitoring the absorption changes of the MLCT through UV-Vis spectra.

  The thermally induced spin transitions of cages 1~3 in solution were investigated by variable tempera-ture UV-Vis spectra.It gives rise to a decrease in the intensity of the characteristic MLCT absorption band in UV-Vis spectra as the temperature warms from 20 to 80 ℃ (Fig.S14~S16), indicating that the iron(Ⅱ) center in 1~3 change from the low-spin state to high-spin state.However, the thermally induced spin transitions of cages 1~3 in solution are gradual, and the temperature range is wide.In order to develop a more effective approach to tune the spin state of cages 1~3 in solution, a series of anion titrations experiments were carried out.And it was found that the addition of halide (Cl^- and Br^-) had major influence on the spin state of cages 1~3.

  The solution color of cages 1~3 dramatically changed from violet red to pale yellow with titration of TBACl in acetonitrile, which was easily monitored by the naked eye, suggesting that the spin state transition in solution (Fig. 4).Additionally, tremendous spectral changes in the UV-Vis spectra were observed.The absorption intensity of characteristic broad absorption MLCT bands at about 536 nm significantly decreased for cages 1~3 after the step by step addition of Cl^- (Fig. 5a, 5c and 5e).Binding isotherms showed that the absorption intensity at 490, 506, 536 and 566 nm undergo a rapid decline with the addition of Cl^-, and then reach a platform (Fig. 5b, 5d and 5f).These results demonstrate that the sensitive and significant change with spin state from low-spin to high-spin of the metal center for iron(Ⅱ) cages 1~3 in solution upon addition of Cl^-.Similar color change and UV-Vis monitored results were also observed when titrations of cages 1~3 with TBABr (Fig.S17~S19).Ag(Ⅰ) back-titration experiments were carried out.The addition of AgBF₄ or AgClO₄ to the HS pale yellow solution led to a reproduction of LS red solution (Fig.S20), and the absorption intensity of MLCT bands was observed to gradually increase (Fig.S21~S23).However, the absorption intensity of MLCT bands in UV-Vis spectra showed a slight increase (Fig.S24) when titrations of cages 1~3 with TBAI.

  Figure 4

  

  Figure 4.  Photographs of cage 1 in CH₃CN solutions upon addition of tetrabutylammonium salts of Cl^-

  下载: 全尺寸图片 幻灯片

  Figure 5

  

  Figure 5.  Plot of the absorption changes for cages 1 (a), 2 (c) and 3 (e), and binding isotherms with fitted curves for cages 1 (b), 2 (d) and 3 (f) titrated with tetrabutylammonium salts of Cl in a CH₃CN solution

  c_complex=63 μmol·L^-1, V_complex=3 mL; c_TBACl=0.05 mol·L^-1

  下载: 全尺寸图片 幻灯片

  In order to gain further insight into the halide trigged spin state switching behaviors of cages 1~3, the ¹H NMR titration experiments were tried.Unfor-tunately, this experiment was infeasible due to the significant precipitation occurred upon addition of TBACl or TBABr at the concentration of cages 1~3 (65 mmol·L^-1) used for ¹H NMR measurements.Even so, the precipitates were collected, and the variable-temperature magnetic susceptibility measurements of the corresponding precipitates showed that they were all in high spin state (Fig.S25). These results further confirmed that cages 1~3 with halide anions (Cl^- and Br^-) in solid state were in high spin state.

  From the single crystal X-ray diffraction analysis of cages 1~3, efficient anion-cage interactions are formed through C-H…F or C-H…O interactions between the encapsulated BF₄^- or ClO₄^- anion and the ligands.These interactions can increase the σ-donating ability of the ligands bound to Fe(Ⅱ) ions, which thus contributes to stabilization of the low-spin state^[19].For titrations of cages 1~3 with halide in solution, we speculate that the halide anions displace the inner and peripheral BF₄^- or ClO₄^- anions due to the diffusional exchange and strong electrostatic interaction between the halide and cage host.However, the small volume of Cl^- (0.019 51 nm³) and Br^- (0.025 52 nm³) could not match well with the cavity size of cages 1~3^[20], so they will not be able to interact with the cage superstructure effectively and the anion binding interactions are weakened even disappeared, which causes the LS to HS spin state transition for cages 1~3.Since the volume of I^- (0.036 62 nm³) gets much larger than Cl^- and Br^-, the effective anion-π interac-tions may existence between the encapsulated I^- and the twelve imidazole groups in the internal surface of the cages^[21], which results to a slight more stabilization of the low-spin state for iron(Ⅱ) centers than BF₄^- or ClO₄^- anions.Therefore, it is implying that the anion binding interactions inside the cage cavity and around the periphery may play an important role in realizing the spin state transition in solution for cages 1~3.

  3.   Conclusions

  In summary, three SCO active tetrahedral iron(Ⅱ) cages were prepared and exhibited sensitive halide (Cl^- or Br^-) triggered spin state change from low-spin to high-spin in solution.Anionic guests induced spin state switching based on the host-guest chemistry in metal-organic cage opens new avenues for tuning the spin-crossover properties in solution at room temperature.

  Supporting information is available at http://www.wjhxxb.cn

• 1. [1]

     (a) Halcrow M A.Chem. Soc. Rev., 2011, 40: 4119-4142
     (b)Bousseksou A, Molnar G, Salmon L, et al.Chem. Soc. Rev., 2011, 40: 3313-3335
     (c)Cook Kershaw L J, Mohammed R, Sherborne G, et al.Coord. Chem. Rev., 2015, 289: 2-12
     (d)Tao J, Wei R J, Huang R B, et al.Chem. Soc. Rev., 2012, 41: 703-737
     (e)Manrique-Juárez M D, Mathieu F, Laborde A, et al.Adv. Funct. Mater., 2018: 1801970
     (f)WANG Yu-Xia(王玉侠), QIU Dan(邱丹), XI Sai-Fei(奚赛飞), et al.Chinese J. Inorg. Chem.(无机化学学报)., 2016, 32(11): 1965-1972

  2. [2]

     (a) Khusniyarov M M.Chem. Eur. J., 2016, 22: 15178-15191
     (b)Guionneau P.Dalton. Trans., 2014, 43: 382-393
     (c)Kahn O, Martinez C J.Scienc., 1998, 279: 44-48
     (d)Rosner B, Milek M, Witt A, et al.Angew. Chem. Int. Ed., 2015, 54: 12976-12980
     (e)Zenere K A, Duyker S G, Trzop E, et al.Chem. Sci., 2018, 9(25): 5623-5629

  3. [3]

     (a) Samanta S, Demesko S, Dechert S, et al.Angew. Chem. Int. Ed., 2015, 54: 583-587
     (b)Dommaschk M, Schutt C, Venkataramani S, et al.Dalton. Trans., 2014, 43: 17395-17405

  4. [4]

     (a) Shores M P, Klug C M, Fiedler S R.Spin-Crossover Materials: Properties and Applications. Halcrow M A Ed., Oxford: Wiley-Blackwel., 2013: 281-301
     (b)Weber B, Walker F A.Inorg. Chem., 2007, 46: 6794-6803

  5. [5]

     (a) Hasegawa Y, Kume S, Nishihara H.Dalton. Trans., 2009, 2: 280-284
     (b)Heitmann G, Schütt C, Herges R.Eur. J. Org. Chem., 2016, 22: 3817-3823
     (c)Venkataramani S, Jana U, Dommaschk M, et al.Scienc., 2011, 331: 445-448
     (d)Thies S, Sell H, Schütt C, et al.J. Am. Chem. Soc., 2011, 133: 16243-16250

  6. [6]

     (a) Schmitz M, Seibel M, Kelm H, et al.Angew. Chem. Int. Ed., 2014, 53: 5988-5992
     (b)Wilson R K, Brooker S.Dalton. Trans., 2013, 42: 12075-12078
     (c)Miller J S, Min K S.Angew. Chem. Int. Ed., 2009, 48: 262-272

  7. [7]

     (a) Young M C, Liew E, Ashby J, et al.Chem. Commun., 2013, 49: 6331-6333
     (b)Ono K, Yoshizawa M, Akita M, et al.J. Am. Chem. Soc., 2009, 131: 2782-2783
     (c)Ni Z, Shores M P.J. Am. Chem. Soc., 2009, 131: 32-33
     (d)Ni Z, McDaniel A M, Shores M P.Chem. Sci., 2010, 1: 615-621
     (e)Ni Z, Shores M P.Inorg. Chem., 2010, 49: 10727-10735

  8. [8]

     (a) Rissanen K.Chem. Soc. Rev., 2017, 46: 2638-2648
     (b)Zarra S, Wood D M, Roberts D A, et al.Chem. Soc. Rev., 2015, 44: 419-432
     (c)Ahmad N, Younus HA, Chughtai A H, et al.Chem. Soc. Rev., 2015, 44: 9-25
     (d)Ramsay W J, Ronson T K, Clegg J K, et al.Angew. Chem. Int. Ed., 2013, 52: 13439-13443
     (e)Ward M D, Hunter C A, Williams N H.Chem. Lett., 2017, 46: 2-9
     (f)Metherell A J, Cullen W, Williams N H, et al.Chem. Eur. J., 2018, 24: 1554-1560

  9. [9]

     (a) Struch N, Bannwarth C, Ronson T K, et al.Angew. Chem. Int. Ed., 2017, 56: 4930-4935
     (b)Bilbeisi R A, Zarra S, Feltham H L, et al.Chem. Eur. J., 2013, 19: 8058-8062

  10. [10]

      (a) Ren D H, Qiu D, Pang C Y, et al.Chem. Commun., 2015, 51: 788-791
      (b)Zhang F L, Chen Q J, Qin L F, et al.Chem. Commun., 2016, 52: 4796-4799

  11. [11]

      SAINT-Plus, Version 6.02, Bruker Analytical X-ray System, Madison, WI, 1999.

  12. [12]

      Sheldrick G M.SADABS, Bruker Analytical X-ray Systems, Madison, WI, 1996.

  13. [13]

      Sheldrick G M.SHELXTL-97, University of G?ttingen, G?ttingen, Germany, 1997.

  14. [14]

      van der Sluis P, Spek A L. Acta Crystallogr. Sect. A:Found. Crystallogr., 1990, A46:194-201 doi: 10.1007/s10593-016-1928-1

  15. [15]

      Spek A L. Acta Crystallogr. Sect. D:Biol. Crystallogr., 2009, D65:148-155

  16. [16]

      Brooker S. Chem. Soc. Rev., 2015, 44:2880-2892 doi: 10.1039/C4CS00376D

  17. [17]

      Paul R L, Argent S P, Jeffery J C, et al. Dalton. Trans., 2004, 21:3453-3458

  18. [18]

      (a) Vostrikova K E, Luneau D, Wemsdorfer W, et al.J. Am. Chem. Soc., 2000, 122: 718-719
      (b)Liu Y C, Hua S A, Cheng M C, et al.Chem. Eur. J., DOI: 10.1002/chem.201801325

  19. [19]

      (a) Luo Y H, Nihei M, Wen G J, et al.Inorg. Chem., 2016, 55: 8147-8152
      (b)Nemec I, Herchel R, Travnicek Z.Dalton. Trans., 2015, 44: 4474-4484

  20. [20]

      Darawsheh M, Barrios L A, Roubeau O, et al. Chem. Eur. J., 2016, 22:8635-8645 doi: 10.1002/chem.v22.25

  21. [21]

      Mooibroek T J, Gamez P. CrystEngComm, 2013, 15:1802-1805 doi: 10.1039/c2ce26853a

• 

  Scheme 1  Schematic representation of the halide triggered spin state switching of iron(Ⅱ) tetrahedral cages

  下载: 全尺寸图片 幻灯片
  

  Figure 1  X-ray crystal structures of iron(Ⅱ) tetrahedral cages 1 (a), 2 (b) and 3 (c) showing one encapsulated anion in space-filling mode

  C:grey; N:blue; Fe:purple; B:orange; F:lime; Cl:bright green; O:red; Br:dark teal; All H atoms, solvent molecules and the remaining anions have been removed for clarity

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a) Anion binding between the encapsulated anion and the C-H protons of imidazole moieties for cages 1~3; (b) Close-up view of one anion binding with three C-H protons of each vertex of the cages; (c) Close-up view of one anion binding with two C-H protons of each side of the cages

  All H atoms, solvent molecules, the remaining anions and the twelve substituted phenyl ring group around the periphery of four vertices have been removed for clarity

  下载: 全尺寸图片 幻灯片
  

  Figure 3  Variable-temperature solid-state magnetic susceptibility for cages 1 (a), 2 (b) and 3 (c)

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Photographs of cage 1 in CH₃CN solutions upon addition of tetrabutylammonium salts of Cl^-

  下载: 全尺寸图片 幻灯片
  

  Figure 5  Plot of the absorption changes for cages 1 (a), 2 (c) and 3 (e), and binding isotherms with fitted curves for cages 1 (b), 2 (d) and 3 (f) titrated with tetrabutylammonium salts of Cl in a CH₃CN solution

  c_complex=63 μmol·L^-1, V_complex=3 mL; c_TBACl=0.05 mol·L^-1

  下载: 全尺寸图片 幻灯片

  Table 1.  Crystallographic data for cages 1~3

  	cage 1	cage 2	cage 3
  Formula	C156H168B8F32Fe4N36	C160H164Cl18Fe4N38O24	C156H156Br12Cl8Fe4N36O33.26
  Formula weight	3 465.14	3 864.79	4 533.29
  T/K	173(2)	173(2)	173(2)
  Crystal system	Triclinic	Orthorhombic	Orthorhombic
  Space group	P1	C2221	C2221
  a/nm	1.830 33(19)	2.005 18(11)	2.015 94(13)
  b/nm	1.877 5(2)	3.333 03(15)	3.347 33(17)
  c/nm	1.897 6(2)	3.434 6(2)	3.435 8(2)
  α/(°)	62.228(3)
  β/(°)	63.553(3)
  γ/(°)	77.863(3)
  V/nm3	5.1661(10)	22.954(2)	23.185(2)
  Z	1	4	4
  Dc/(Mg·m-3)	1.114	1.118	1.229
  μ/mm-1	0.354	0.516	2.468
  F(000)	1 788	7 968	9 080
  θ range/(°)	1.23~22.50	3.02~24.25	2.95~25.00
  Index ranges	-19 ≤ h ≤ 19, -19 ≤ k ≤ 20, -20 ≤ l ≤ 20	-23 ≤ h ≤ 23, -38 ≤ k ≤ 37, -35≤ l ≤39	-23 ≤ h ≤ 23, -37 ≤ k ≤ 36, -40≤ l ≤39
  Reflection collected	29 723	50 325	54 201
  GOF (F2)	1.039	1.064	1.106
  R1a, wR2b [I>2σ(I)]	0.096 3, 0.228 1	0.080 4, 0.202 8	0.065 9, 0.150 1
  R1a, wR2b (all data)	0.172 3, 0.258 5	0.115 8, 0.224 0	0.117 0, 0.166 8
  a R1=∑||Fo|-|Fc||/∑|Fo|; b wR2=[∑w(Fo2-Fc2)2/∑w(Fo2)]1/2.

  下载: 导出CSV
•