Various Structural Types of Cyanide-Bridged FeIII–MnIII Bimetallic Coordination Polymers (CPs) and Polynuclear Clusters Based-on A New mer-Tricyanoiron(III)Building Block: Synthesis, Crystal Structures, and Magnetic Properties

Four cyanide-bridged FeIII–MnIII complexes {[Fe(qxcq)(CN)3][Mn(L1)(H2O)]}[Mn(L1)(H2O)(CH3OH)](ClO4)·1.5MeOH·0.5H2O (L1 = N,N′-bis(3-methoxy-5-bromosalicylideneiminate) (2), {[Fe(qxcq)(CN)3][Mn(L2)]}2·0.5H2O (L2 = N,N′-ethylene-bis(3-ethoxysalicylideneiminate)) (3), [Fe(qxcq)(CN)3][Mn(L3)] (L3 = bis(acetylacetonato)ethylenediimine) (4), [Fe(qxcq)(CN)3][Mn(L4)]·1.5MeOH·0.5CH3CN·0.25H2O (L4 = N,N′-(1,1,2,2-tetramethylethylene)bis(salicylideneiminate)) (5), were prepared by assembling a new structurally characterized mer-tricyanoiron(III) molecular precursor (Ph4P)[Fe(qxcq)(CN)3]·0.5H2O (qxcq− = 8-(2-quinoxaline-2-carboxamido)quinoline anion) (1) and the corresponding manganese(III) Schiff base compound. Complexes 2and 3containa cyanide-bridged heterobimetallic dinuclear entity, which can be further dimerized by self-complementary H-bond interactions through the coordinated water molecule from one complex and the free O4unit from the adjacent complex. Complexes 4 and 5 area one-dimensional coordination polymer (CP) comprised of the repeated [Mn(Schiffbase)-Fe(qxcq)(CN)3] units. Complex 4 shows a linear-chain conformation with two trans-located cyano groups bridgingthe neighboring Mn units, while complex 5 is a zigzag-like 1D CP, where the two cyano groups in cis configurationfunction as bridges. In bothcomplexes 4 and 5, the inter-chain π–πstack interactions within the aromaticrings of cyanide precursor extend the 1D chain into the supermolecular 2D networks. The magnetic property has been experimentally studied and theoretically fitted over the four Fe(III)-Mn(III) complexes, revealing the antiferromagnetic interaction in complexes 2 and 4 and the unusual ferromagnetic coupling in complexes 3 and 5 between the Fe(III) ion and the Mn(III) ion bridged by the cyano group. Furthermore, the different magnetic coupling nature has been analyzed on the basis of the magneto-structure correlation of the mer-tricyanometallate-based Fe(III)-Mn(III) magnetic system.


Introduction
The research on molecule-based magnetic materials has always been within the hottest topics of interests for chemical and physical researchersdue to their potential values in molecule devices, information memorizer, and quanta computers, et. al. [1][2][3][4]. Among which, since their magnetic coupling nature between neighboring paramagnetic metal centers through the cyanogroup and the molecular topological structures can be relatively easily predicted and controlled, the cyanide-bridged magnetic systems have been widely and intensely studied. Recently, with the aim not only of further disclosing the magneto-structure correlation but also more importantly for the great potential in many high-tech fields, a large numberof molecular magnetic materials containing a cyanide linkagewith different structural types have been synthesized and magnetically studied.Many of the reported cyanide-bridged magnetic assemblies showed very interesting magnetic properties [5][6][7][8][9][10][11][12][13][14] including single-molecule magnet (SMM) and single-chain magnet (SCM), which behave as nanosized magnets under the blocking temperature because of the energy barrierscoming from the large ground spin state (ST) and the strong easy axial anisotropy (D) [15][16][17].

General Procedures and Materials
The preparation reaction for all compounds was finished at room temperaturein air, and all of the solvents and chemicals used were reagent grade without any further purification. The mer-[Fe III (qxcq)(CN) 3 ] − precursor was preparedby referencing the method used to prepared other mer-tricyanoiron(III) compounds [35]. The Schiff base ligands can be obtained with acceptable yields by condensing the corresponding amine and aldehyde or ketone with the mole ratio 1:2. The four Mn(III) Schiffbase compounds could be obtained by referencing the procedure in the previous literature [36,37].
Caution! KCN is hazardous and hypertoxic. Metal complexes containing perchlorate are possibly explosive sothese materials should be treated with great caution.

Synthetic General Procedure for the Hqxcq Ligand and the Complexes 1-5
The Hqxcq ligand [38]: To the solution of 2-quinoxalinecarboxylicacid (0.87g, 5mmol) in pyridine (2mL) was added 8-aminoquinoline (0.72 g, 5 mmol) in pyridine (2mL) at 373K. The mixture was stirred gently for 10 min, and triphenylphosphite (1.55 g, 5 mmol) was added rapidly. Then, the resultant solution was heated with stirring in a water bath for 8 h. On cooling of the mixture, a dark yellow solid resulted, with a weight of 1.2 g (80%). An analytical sample was twice recrystallized from CH 2 Cl 2 to give yellow needles. The recrystallized product was dissolved in CH 2 Cl 2 and purified over silica gel withthe mixture of petroleum ether and ethyl acetate [3:1, v:v] as the eluent. The major band was collected, giving the bright yellow colormicrocrystallinetarget compound ( Figure S1, Supplementary Materials) with a yield of about 60% (meltingpoint: 190-192 • C), which was structurally characterized withInfrared Spectrum and Nuclear Magnetic Resonance ( Figures S2 and S3). 1 10.0 mmol)dissolved in 20 and 60 mL methanol, respectively, was refluxed for 1h before KCN (3.91 g, 60.0 mmol) diluted in distilled water (40 mL) was then added. The solution was refluxed for 8 h until the color of the solution changed into dark blue. After cooling and filtering, the filtrate was concentrated under reduced pressure to about 30 mL. [Ph 4 P]Br (4.20 g, 10.0 mmol) was poured slowly into the filtrate, resulting in the target blue crystalline solid, which was collectedby suction filtration, washed with distilled water (2 × 10 mL), and dried for 12 h at r.tunder vacuum.
Dark-blue crystals used for single x-ray diffraction were obtained by slowly evaporating the methanol solution of complex 1 for about one week at 298 K. (

Single-Crystal X-Ray Diffraction and Structural Refinement
For the reported complexes 1-5, the corresponding single-crystal for x-ray analyses with a suitable sizewas mounted ona glass rod, and the crystal diffraction data were collected by using the ω scan mode on a Bruker APEX2 CCD diffractometer with a Mo Kα sealed tube (λ = 0.71073 Å) at room temperature. All structures were resolved using the direct method and further expanded employing Fourier difference techniques with the SHELXTL-2018 program package. All of the non-hydrogen atoms were anisotropically refined with anisotropic displacement coefficients. Hydrogen atoms were distributed to the isotropic displacement coefficients with U(H) =1.2 or 1.5U(C), and the coordinates were permitted to ride on the corresponding carbons utilizing SHELXL-2018, except for some solvent H atoms. The H atoms from the solvent molecules were isotropically refined with fixed U values, during which the DFIX command was utilized for the purpose of rationalizing the bond parameter. The supplementary crystallographic data for the reported complexes 1-5 were deposited in The Cambridge Crystallographic Data Center (CCDC) with the number 1902326-1902330, which is freely available from the Cambridge Crystallographic Data Centre [39]. The crystal parameters, data collection, and refinement are summarized in detail in Table S1 (Supplementary Materials). Some important bond parameters for complexes 1-3 and 4-5 are listed in Tables S2 and S3, respectively. The measured and calculated powder x-ray diffraction (PXRD) patterns for compounds 1-5 are given in Figures S4-S8 (Supplementary Materials). The calculated and experimental diffraction peaks were basically in the same position, demonstrating the phase purity of these compounds.

Physical Measurements
Elemental analysis for C, H, and N were carried out by using an Elementary Vario El and the infrared spectra were measured by employing KBr disks on a Shimadzu FTIR-8600 spectrophotometer. Magnetic susceptibilities were measured by a Quantum Design SQUID MPMS-XL magnetometer, where the dc measurements were collected within the temperature range 2-300 K and the field range of 0-50 KOe. All of the experimental susceptibilities were corrected for the diamagnetism of the constituent atoms (Pascal's tables).

Preparation and General Characterization
The Mn III quadridentate Schiffbase compounds with always sizable anisotropic characteristics have been widely utilized to synthesize the CN-bridged magnetic compounds not only due to their easy preparation but also to the big spin (S = 2) and the intrinsic Jahn-Teller effectalong the axial direction [36]. Reactions of the four quasi planar tetradentate Schiffbase ligands based manganese(III) compounds with the new mer-tricyanoiron(III) building block provided four CN-bridged heterobimetallic compounds with different structural conformations, for which the corresponding synthesis scheme for the organic ligand and complexes 1-5 are provided in Scheme S1, indicating that the structure of the different Schiffbase has an obvious influence on the structural type of the target cyano-bridged complexes. Additionally, the obtained results also imply that the mer-tricyanoiron(III) precursor [Fe(qxcq)(CN) 3 ] − is more favorable for constructing a low-dimensional molecular magnetism system.
The IR spectroscopy characterization (Figures S9-S13) for the CN-bridged Fe(III)-Mn(III) complexes2-5 as well as PPh 4 [Fe(qxcq)(CN) 3 ]·0.5H 2 O werecarried out. In the infrared spectra of the cyano precursor, the absorption positioned at about2115 cm −1 was distributed to the uncoordinated cyanide group. For complexes 2-5, two sharp peaks originating from the CN group stretching vibration could be observed within 2120-2160 cm −1 , implying the existence of both bridging and non-bridging cyano groups in these complexes.  The intramolecular Fe-Mn distances through the cyano linkage were 5.169 Å in 2and 5.259, 5.281 Å in 3, respectively. Both of the cyanide-bridged binuclear entity and the monocnuclear unit can be self-assembledby the coordinated solvent molecule from one complex and the free O 4 unit from the adjacent complex, therefore giving the supramolecular dimerized structure. Furthermore, the dimeric moieties are linked into a one-dimensional supramolecular architecture by the π-π stacking interactions between the qxcq ligands.

Crystal
Complexes 4 and 5. Figures 4 and 5 show the independent binuclear, one-dimensional structure and the cell packing diagram presenting the π-π interaction along a axis for complex 4. For complex 5, its binuclear and one-dimensional structure is given in Figure 6.      It should be pointed out that either of the linear or the zigzag chains could be stacked in parallel by the π-π stacking interaction between the [Fe III (qcq)(CN) 3 ] − moieties, leading to the 2D supramolecular sheet, where the solvent contents (for complex 5) were filled in the channels along the a or c axial. Figure 6. The independent binuclear and 1Dchain structure of complex 5. The solvent content has been omitted for clarity.

Magnetic Properties of Complexes 2-5
The temperature dependent magnetic susceptibility with the form of per Fe III Mn III 2 for complex 2 and per Fe III Mn III for complexes 3-5 measured from 2-300 K under the applied magnetic field of 2000 Oe using the corresponding single crystal with the quantity about 10-0 mg are shown in Figures 7 and 8, respectively. The variation trends of the χ m T-T curves for compounds 2 and 4 were similar to each other, which was also similar for complexes 3 and 5. The χ m T values at r.t. for these four complexes were about 6.16, 3.27, 3.32, and 3.31 emu·K·mol -1 , respectively, which were basically in agreement with the spin-only value of 6.375 for the two isolated Mn(III) ions (S = 2), one low spin Fe(III) ion (S = 1/2) for 2 and 3.375 emu·K·mol -1 for one Mn(III) ion, and one Fe(III) ion for 3-5, respectively. For compounds 2 and 4, the χ m T values decreased slowly, accompanying the temperature down to about 50 K, then the χ m T values began to decrease quickly and reached their lowest value at about 1.47 and 0.56 emu·K·mol -1 at 2 K, respectively, primarily indicating the antiferromagnetic interaction between the Mn(III)-Fe(III) unit through the bridging cyanide group. The χ m T-T curves for complexes 3 and 5 presenteda basically contrary change tendency to the above two compounds. By decreasing the temperature, the χ m T value for these two complexes remained almost constant until 75 K, then began to increase smoothly and attained their maximum value at about 3.50 emu·K·mol -1 for 3 and 3.72 emu·K·mol -1 for 5, respectively, and then decreasedwith high speed to the minimum value of about 1.50 and 1.24 emu·K·mol -1 for 3 and 5, respectively, at 2 K. The magnetic behavior of the complexes at very low temperature can probably be due to the intermolecular anti-ferromagnetic interaction resulting from the inter-molecular weak interactions and/or the zerofieldsplitting (ZFS) effect of the Mn(III) ion. The described change tendency for the χ m T-T curves above primarily implied the antiferromagnetic and ferromagnetic coupling between the cyanide-bridged Mn(III) ion and Fe(III) ion in complexes 2 and 4 and 3 and 5, respectively. The magnetic susceptibilities for complexes 2-5 obey the Curie-Weiss law well in 10-300 K. The obtained negative Weiss constant θ= −5.2 K, Curie constant C= 6.26 emu·K·mol -1 for 2, θ= −2.65 K, C= 3.30 emu·K·mol -1 for 4, positive θ= 0.65 K, C= 3.33 emu·K·mol -1 for 3, and θ= 1.35 K, C= 3.32 emu·K·mol -1 for 5, further confirming the overall antiferromagnetic or ferromagnetic interaction between the Mn(III) ion and Fe(III) ion in these complexes. The insets of Figures 7 and 8 show that the field-dependent magnetization was measured up to 50 kOe at 2 K for the corresponding complexes. The magnetization for all complexes increasedquickly up to about 10 kOe. After that, the curve becamecomparatively even and gradually reached their largest value. The values at 50 kOe were 6.1 and 2.9 Nβ for complexes 2 and 4, which were significantly lower than the expected 9 and 5 Nβfor the uncoupled two or one high spin Mn(III) ion and one low spin Fe(III) ion based on g = 2.0, proving again the existence of overall antiferromagnetic coupling in complexes 2 and 4. The maximum field-dependent magnetization value for complexes 3 and 5 were about 4.9 Nβ, very close to the saturated value 5 Nβcorresponding to the spin state S = 5/2 from the high-spin Mn(III) ion andlow-spin Fe(III) ion, further revealing the intramolecular ferromagnetic interaction in them.
The theoretical magnetism analysis for complexes 2-5 was carried out according to their structural types.Complex 2 can be structurally considered as the cyanide-bridged binuclear entity by adding an isolated free Mn(III) ion compound, while complex 3 belongs to only a pure cyanide-bridged binuclear structure. Basedon the dinuclear Mn III -Fe III (S = 2 and 1/2) model, the magnetic susceptibilities for these two compounds can be simulated by the following expression originatingfrom the isotropic exchange spin HamiltonianĤ = −2JŜ FeŜMn . However, an additional isolated Mn(III) ion (χ Mn = Ng 2 β 2Ŝ Mn (Ŝ Mn +1)/3kT, S = 2) needs to be included for complex 2.
The magnetic data for these two compoundswere fitted in the temperature range of 10-300 K by introducing the mean-field approximation (zJ'), accounting for the effects other than the intramolecular magnetic coupling. The obtained best-fit parameters were J = −5.56 cm −1 , g = 1.99, zJ' = −0.81 cm −1 and J = 2.15 cm −1 , g = 1.98, zJ' = −0.58 cm −1 for complexes 2 and 3, respectively. The negative zJ' values indicate that antiferromagnetic interactions occur between dimeric units, which are related to the presence of the intermolecular H-bonds. These results are basically comparable to those values found in the cyanide-bridged Fe III Mn III polynuclear complex previously reported [35]. The magnetic data for the two one-dimensional chain-like complexes 4 and 5 werefitted by the HamiltonianĤ =Ĥ ex +Ĥ anis + gβĤŜ, where the first term corresponds to the isotropic interactions (the magnetic couplings through the single cyano bridge). The second term deals with the local anisotropy of the high spin Mn(III) ion, and the last one is the Zeeman effect. Considering that the two independent Fe(III)-CN-Mn(III) linkages in these two complexes were not markedly different to the two Mn-N≡C bond angles 154.7(10), 158.5(10) • in 4 and 173.4(4), 172.4(4) • in 5, respectively, these two compounds could be treated as a chain containing alternative spins 1/2 and 2 with approximately one exchange coupling J (Scheme 2). In this case, the magnetic susceptibilities of the two 1D chains could be fitted and calculated rationally on the basis of a closed loop model comprised by four pairs of 1/2-2 spin pairs. The simulation was done by numerical matrix diagonalization techniques employing a Fortran program [40].

Magneto-Structural Correlation
As has been proven, due to the strict orthogonality of the magnetic orbitals of [d xy /d xz /d yz ] 1

Conclusions
In summary, a new mer-tricyanometallate wasfirst prepared and structurally characterized, and based on which and the different Schiff base manganese(III) compounds, a new series of cyanide-bridged magnetic complexes were obtained. Single crystalx-ray diffraction analysis revealed that their different structural types belonged to discrete polynuclear clusters and 1D infinite linear or zig-zag coordination polymers, which could be further dimerized by the intermolecular hydrogen bond interactions and assembled into a 2D supramolecular structure by the inter-chain π-π interactions, respectively, indicating that the structure of the Schiff base ligand could play an important rolein the structure of the target cyanide-bridged complexes. Systematic magnetic study on all complexes with the combination of the magneto-structural correlation disclosed the antiferromagnetic coupling in complexes 2 and 5 with smaller Mn-N≡C bridge bond angles between the cyanide-bridged Fe III -Mn III , while the ferromagnetic magnetic interaction in the other two complexes containing the larger Mn-N≡C bridge bond angle, which for the latter has been seldom been previously reported with regard to the mer-tricyano-based Fe(III)-Mn(III) magnetic assemblies. The results reported here offer meritorious information for the rational design of newmer-tricyano-containing precursors used to constructmolecular magnetism systems in coordination chemistry with various structural types and different magnetic coupling nature. The obtained new series of bimetallic complexes can further enrich the members of the cyanide-bridgedfamily, which are very helpful for thoroughly discovering the magneto-structural correlation in this field from the molecular magnetism perspective.