Structure determination and magnetic studies of triazole chelated Co(II) coordination polymers
Two cobalt(II) halide complexes with 1,2,4‐triazole as a ligand were synthesized. Their structures were determined by extended x‐ray absorption fine structure (EXAFS) and powder x‐ray diffraction (XRD). Both complexes [Co(Htrz)Cl2]n (1) and {[Co(Htrz)2(trz)]BF4}n (2) form one‐dimensional polymeric chain and the distances of Co⋯Co are 3.3521(2) Å and 3.8629(2) Å, respectively. The Htrz and Cl− are bridging ligands to connect two Co(II) ions in 1, and the local environment of Co site is in a distorted octahedron with {CoN2Cl4} core. In complex 2, two Htrz and one trz are bridging ligands to connect two Co(II) ions, and the local geometry of Co is in a pseudo octahedron with {CoN6} core. The analysis of Co LII,III‐edge XAS indicates that the Co(II) of both complexes are at high spin state with t2g5eg2 configuration and the crystal field strength (10Dq) is about 1.2 eV. The broken‐symmetry DFT calculations indicate that antiferromagnetic coupling state of Co⋯Co is the most stable state in both complexes; and the coupling constants of 1 and 2 are −0.32 cm−1 and −3.70 cm−1, respectively. Based on the distances of Co⋯Co and coupling constants, such antiferromagnetic interaction is achieved through triazole ligands.
Keywords: antiferromagnetic coupling; broken symmetry; DFT; EXAFS; XRD
The broken‐symmetry DFT calculations indicate that antiferromagnetic coupling state of Co⋯Co is the most stable state in [Co(Htrz)Cl2]n (1) and {[Co(Htrz)2(trz)]BF4}n (2); and the coupling constants of 1 and 2 are −0.32 cm−1 and −3.70 cm−1, respectively.
Based on the distances of Co⋯Co and coupling constants, such antiferromagnetic interaction is achieved through triazole ligands.
INTRODUCTION
Developing multifunctional materials with controllable properties like conductivity, luminescence, and magnetism is one of the demands in scientific research studies nowadays. In general, transition metal complexes can provide such properties, especially in the 3d4–3d7 metal complexes that can adapt to high spin (HS) or low spin (LS) states depending on the splitting energy (10Dq) of the ligand field strength and the spin pairing energy (Π). By combining with organic polymers and transition metal complexes, functional composite materials can be made for applications. For example, Shepherd et al. prepared composite bilayer actuators by mixing [Fe(trz)(Htrz)2](BF4) (Htrz = 4H‐1,2,4‐triazole; trz = 1,2,4‐triazole) and poly(methylmethacrylate) (PMMA), where [Fe(trz)(H‐trz)2](BF4) is a spin‐crossover (SCO) compound by using triazole as bridging ligands to connect Fe(II) centers to form the extended one‐dimensional (1D) polymeric structure.[[1]] Due to the poor crystallinity of such fiber‐like sample, the crystal structure was determined by powder x‐ray diffraction (XRD). From application point of view, such 1D coordination polymer, which connects the identical metals or diverse metals with multidentate ligand, may have more interesting magnetic or electronic properties through the cooperative effects. However, such research only focus on the known Fe(II) complexes, but lack of Co(II) case. The study of Co(II) complex may provide the structural information and the behavior of magnetic interactions for the reference of future applications.
In 1995, Zaydoun et al. synthesized Mn(II)/Fe(II)/Co(II)/Ni(II)/Cu(II) halide complexes with 1,2,4‐triazole ligands and proposed a 1D polymeric structure for these complexes based on spectroscopy methods.[[2]] In these studies, only nickel complex exhibits ferromagnetic interaction and the others are antiferromagnetic exchange coupling. However, due to the formation of poor crystalline materials, only cell constants were reported in these complexes so that the magnetic interactions between metal sites cannot be further studied through theoretical calculations. In addition, the structure may behave very differently even due to small differences in sample synthesis.[[[3]]] For example, Hsu et al. reported that [Co(Htrz)Cl2]n is crystalized in orthorhombic system by XRD, but Gautier et al. obtained a single crystal in monoclinic system through a hydrothermal method by using more than 10 times concentration of Zn source in strong acidic (37% HCl) condition.[[2]] Thus, it is worth to reexamine the structure directly from the polycrystalline powder sample based on normal synthetic methods.
In order to study the magnetic interactions between metal sites in such 1‐D coordination polymer, we will focus on two complexes [Co(Htrz)Cl2]n (1) and {[Co(Htrz)2(trz)]BF4}n (2), which will be synthesized by typical methods. Then, x‐ray absorption spectroscopy (XAS) of Co and Cl K‐edge will be applied to characterize the oxidation states and local structure of Co site. The complete 3D geometric structure will be determined based on the high‐resolution XRD data through direct method (DM) or global optimization method.[[5]] Finally, the magnetic interaction between Co ions will be discussed based on the broken‐symmetry density functional theory (DFT) calculations. Our studies will not only provide the crystal structures and magnetic coupling between metal sites, but also suggest some hybrid methods to accelerate structure determination from powder diffraction (SDPD) such as to combine with periodic DFT calculation or structural databases (e.g. Cambridge Structure Database).
EXPERIMENTAL SECTION
Synthesis
The synthetic method for compounds [Co(Htrz)Cl2]n (1) and {[Co(Htrz)2(trz)]BF4}n (2) were similar to those in previous reports.[[[2], [6]]] All chemicals were purchased from company without further treatment.
Preparation of [Co(Htrz)Cl 2 ] n (1)
Complex 1 was obtained as follows: a solution of 0.3453 g (5 × 10−3 mol) of 1H‐1,2,4‐triazole in 20 ml of acetone was added to a solution of 1.18965 g (5 × 10−3 mol) of cobalt(II) chloride hexahydrate with 0.09606 g (5 × 10−4 mol) of L‐ascorbic acid as antioxidants in 20 ml of acetone at room temperature. The color of the reaction solution was blue at the beginning, and then turned colorless after 30 min while a violet precipitate appeared. It was allowed to stand for 6 hr, and then the precipitate was filtered out of the solution. The obtained precipitate was further washed with acetone and dried under vacuum in a sample vial. Anal. Calcd for C2H3CoN3Cl: C, 12.08%; H, 1.52%; N, 21.13%. Found: C, 12.21%; H, 1.13%; N, 21.22%. IR(ATR, cm−1): 1,060 (s), 1,130 (m), 1,207 (w), 1,278 (w), 1,308 (s).
Preparation of {[Co(Htrz) 2 (trz)]BF 4 } n (2)
The complex 2 is synthesized in a similar way as that of complex 1, a solution of 1.70315 g (5 × 10−3 mol) of Cobalt (II) tetrafluoroborate hexahydrate with 0.08806 g (5 × 10−4 mol) L‐Ascorbic acid as antioxidants in 10 ml of H2O was added into a solution of 1.03605 g (1.5 × 10−2 mol) of 1H‐1,2,4‐triazole in 10 ml of H2O. A precipitate appeared after 5 min. At the beginning, its color is white and then turned into orange. It was allowed to stand for 6 hr. Because the particle size is too small to filter out by filter paper with 11 μm pore size, and the precipitate was obtained by centrifuging. The obtained orange powder was further dried under vacuum in a sample vial. Anal. Calcd for C6H8CoN9BF4: C, 20.47%; H, 2.30%; N, 35.82%. Found: C, 20.11%; H, 2.29%; N, 35.84%. IR(ATR, cm−1): 1,035 (s), 1,048 (s), 1,143 (m), 1,160 (m), 1,184 (w), 1,210 (w); 1,264 (w), 1,278 (m), 1,305 (s).
Characterizations and computational studies
IR spectroscopy
The IR spectra of powder sample were measured by JASCO FT/IR‐4600. Data were collected in attenuated total reflectance (ATR) mode, and the scan range is from 4,000 to 650 cm−1.
Powder x‐ray diffraction measurements
All synthesized samples were initially measured at Malvern Panalytical X'Pert3 Powder diffractometer using Cu‐Kα radiation equipped with PIXcel1D detector. The high‐resolution powder x‐ray diffraction (XRD) data were collected either at TPS19A, TLS01C2, and TLS17A in the National Synchrotron Radiation Research Center (NSRRC) in Taiwan or at BL02B2 in SPring‐8 in Japan. The wavelengths using at TPS19A, TLS01C2, TLS17A, and BL02B2 are 0.82656, 1.03321, 1.320793, and 0.798287 Å, respectively. Each powder sample was packed into a glass capillary with 0.3 mm diameter. Each capillary was spinning during measurement to avoid anisotropic effects. All XRD data were indexed by N‐TREOR and DICVOL04 in EXPO2014 package.[[[7]]]
X‐ray absorption spectroscopy measurements
All x‐ray absorption experiments were performed at NSRRC in Taiwan. Both Co and Cl K‐edge were recorded at room temperature. Co K‐edge measurements were performed in transmission mode at the TLS17C X‐ray Wiggler beamline with a double crystal Si(111) monochromator. The energy resolution ΔE/E was estimated to be about 2 × 10−4. High harmonic photons were rejected by Rh‐coated mirrors. The spectra were scanned from 7,500 to 8,600 eV using a gas‐ionization detector. A reference Co foil was measured simultaneously for the calibration of photon energy. The first inflection point at 7,709 eV of the Co foil absorption spectrum was used for energy calibration. The ion chambers used to measure the incident (I0) and transmitted (It) photon intensities were filled with a mixture of N2 and He gases and a mixture of N2 and Ar gases, respectively. The EXAFS data analysis of Co K‐edge is following the standard procedure as previous published by Chan et al.[[[9], [11], [13]]]
The Cl K‐edge data were measured by using powder sample at TLS16A with DCM Si(111). The energy resolution ΔE/E is 1 × 10−4. The energy is scanned from 2,700 to 3,000 eV using Lytle detector in fluorescence mode, and the detector is purged with pure N2 gas during data collection. The sample chamber is filled with high purity He gas to avoid air absorption. The photon energy was calibrated to the maximum of the first pre‐edge peak of Cs2CuCl4 at 2,820.2 eV.[[14]]
The Co LII,III‐edge XAS were collected at TLS20A beamline equipped with high‐energy spherical grating monochromator (HSGM). The absorption spectra of Co LII,III‐edge were taken in the energy range from 760 to 815 eV. The powder sample was stuck to conducting tape and kept in an ultrahigh vacuum chamber (10−9 Torr) during measurements. The data were collected in total electron yield mode by measuring the sample current, and the beam size was set by 10 × 10 μm2 slits.
Magnetic measurements
The magnetic measurement was carried out from 2 to 300 K by SQUID magnetometer (SQUID‐VSM Quantum Design Company) under 0.2 T external magnetic field. The magnetic susceptibility data were corrected with ligands' diamagnetism by the tabulated Pascal's constants.
Computational details
Density functional theory (DFT) is used to do geometry optimization on the obtained crystal structure and magnetic interaction study. To simulate the IR spectra of 1H‐1,2,4‐triazole, 4H‐1,2,4‐triazole, and 1,2,4‐triazole anion, the initial geometry is optimized by B3LYP exchange functional and def2‐tzvp basis set in ORCA 4.0.0.2 program.[[15]] The vibrational frequency is calculated on the optimized geometry using the same exchange functional and basis set. There is no imaginary frequency in the calculated results.
The periodic DFT (pDFT) calculations were carried out with CRYSTAL17 package.[[16]] The exchange functional and basis set are PBEh‐3c and def2‐msvp, respectively, which is modified from PBE0‐D3 and a modification of double‐zeta basis set def2‐svp.[[17]] An unrestricted open shell calculation is performed and the spin‐polarization factor is set based on the magnetic measurement by SQUID. The shrinking factors of IS1, IS2, and IS3 are set as 4, 4, and 4, respectively. The number of k points in the irreducible part of the Brillouin Zone (IBZ) is 24. The convergence on energy is set as 10−7 Eh.
To study the magnetic interaction between Co(II) ions in complexes 1 and 2, the coupling calculations are performed with UB3LYP exchange functional and 6‐31G* basis set by G09 program.[[18]] The anti‐ferromagnetic (AFM) or ferromagnetic (FM) coupling constant Jab are calculated based on broken symmetry method with Yamaguchi formula, where the stabilization at AFM or FM is represented in terms of Jab < 0 or Jab > 0, respectively.[[19]] In the dimer models, the multiplicities to calculate the AFM and FM coupling states are 1 and 7, respectively. In tetramer models, the multiplicities to calculate the AFM and FM coupling states are 1 and 13, respectively. In this study, the convergence criterion (CONVER) of simulations is set as 4, 6 and 7. First, the convergence criterion of 4 is achieved. The results obtained with a convergence criterion of 4 were used as the starting point to obtain the results at a convergence criterion of 6. The final convergence criterion of 7 is achieved by using the results of convergence criterion of 6 as the starting point. The other detailed information is described in Data S1.
RESULTS AND DISCUSSION
Structure determination of [Co(Htrz)Cl 2 ] n and {[Co(Htrz) 2 (trz)]BF 4 } n
IR spectroscopy
The experimental IR spectra of complexes 1 and 2 are displayed in Figure S1 and S2, respectively. To resolve the differences of IR spectra between ligand and complexes, the IR spectra of 1H‐1,2,4‐triazole are also depicted in Figure S1 and S2 for comparison. In 5‐membered N‐heterocyclic derivatives, the stretching of ring‐skeletal is around 1,640–1,335 cm−1 and ring‐breathing coupling CH in‐plane deformation is around 1,300–900 cm−1.[[20]] Thus, it is not straightforward to assign the CN and NN stretching modes only based on experimental data. However, the NN symmetric stretching mode in 1,2,4‐triazole anion can be assigned around 1,029 cm−1 based on DFT calculations published by El‐Azhary et al.[[21]] Following the same concept, the DFT calculated IR spectra of 1H‐1,2,4‐triazole, 4H‐1,2,4‐triazole, and 1,2,4‐triazole anion are displayed in Figure S3 where the frequencies around 1,076, 1,030, and 1,017 cm−1 are assigned to the NN symmetric stretching in 1H‐1,2,4‐triazole, 4H‐1,2,4‐triazole, and 1,2,4‐triazole anion, respectively. Based on the change of this vibrational frequency, the triazole ligands chelated to cobalt ions can be identified once the free ligand of 1H‐1,2,4‐triazole was changed to 4H‐1,2,4‐triazole or 1,2,4‐triazole anion in complexes 1 and 2. Therefore, the frequency around 1,060 cm−1 of complex 1 is assigned to NN stretching of 4H‐1,2,4‐triazole. The frequency around 1,048 and 1,035 cm−1 of complex 2 are assigned to NN stretching of 1,2,4‐triazole anion and 4H‐1,2,4‐triazole, respectively. Definitely, the vibrational frequencies of ring‐breathing related to CN are also shift compared with those of 1H‐1,2,4‐triazole ligand. As a matter of fact, the selected region 1,250–1,000 cm−1 of simulated IR spectra of 4H‐1,2,4‐triazole and 1,2,4‐triazole anion can even be correlated to the IR spectra of complexes 1 and 2, as shown in Figure S4 and S5, respectively. Based on the simulated IR spectra, ligands are successfully chelated to Co ions.
Analysis by x‐ray absorption spectroscopy
In complex 1, the Cl− can be a counter anion or bonded to cobalt ion. Taking the advantage of element‐specific character of x‐ray absorption spectroscopy, we measure the Cl K‐edge XAS to study the characteristic of Cl atom in complex 1. Using Oh crystal field as an example, the 3d orbitals of cobalt ion is split into t2g and eg orbitals, with 10Dq energy difference. If CoCl bond exists in complex 1, the cobalt ion and chloride will generate the hybrid orbitals as 3p(Cl)‐3d(Co). When chloride 1s core electron was excited by hard x‐ray, the transition from 1s(Cl) to 3p(Cl)‐3d(Co) is allowed, and this transition signal will show in the pre‐edge region of Cl K‐edge XAS. In Figure 1, a pre‐edge peak at 2,823.0 eV appears in Cl K‐edge XANES of complex 1, which indicates the CoCl bond indeed exists.
The Co K‐edge XAS is displayed in Figure 2 and the pre‐edge region is enlarged in the inset graph. Based on the first derivative of the spectrum, the threshold energy E0 of complexes 1 and 2 are assigned at 7,717.77 and 7,717.70 eV, respectively, which is close to the E0 of CoO at 7,718.2 eV. Thus, we conclude the oxidation state of Co is +2 in both complexes. In addition, the weak intensity of pre‐edge peak at ~7,709.17 eV (complex 1) or ~7,709.30 eV (complex 2) indicated that the local environment of Co is at distorted octahedral symmetry because the transition of 1 s to 3d is forbidden in octahedral symmetry.
According to the results derived from FTIR spectra, Cl, and Co K‐edge XANES, we can expect that the coordination number of Co(II) in complex 1 is six and at least one Cl− and one triazole ligand are bonded to Co(II). Using these informations as queries to search in Cambridge Structure Database (CSD), we found three possible structures as shown in Figure 3. Those complexes codes are GOHNOH, FOVHAA, and ROLYUN, each have some different features but also fit our queries. In GOHNOH structure, two chlorides are bridging ligands, which connect each cobalt ion, and the triazole‐like ligand is perpendicular to the plane of cobalt and chloride. These connections lead to a distance ~3.57 Å between two Co atoms. In FOVHAA structure, the triazole‐like ligand is a bridging ligand between two cobalt ions, and the Cl− are also bridging ligands. This makes a pseudo tetrahedral geometry of Co site, and interatomic distance between two Co is ~3.75 Å. In the ROLYUN structure, both tetrazole‐like and chloride are bridging ligands. This implies a pseudo octahedral geometry of Co site and the interatomic distances between two Co is ~3.36 Å. In order to figure out which model is the most suitable to represent the local structure of Co of complex 1, the EXAFS data analysis will be performed up to ~4.4 Å.
The EXAFS fitting results are displayed in Figure 4. The results showed that the peaks around 2.15(2) Å and 2.48(1) Å were the contributions of two nitrogen and two chlorine atoms, respectively. The third peak in R‐space is mainly the contribution of two N atoms. An attempt to fit the peak by the two N and two C atoms is failed. This implies that the triazole is also a bridging ligand like model FOVHAA or ROLYUN. In combination with the pseudo‐octahedral geometry from the weak intensity of pre‐edge peak, we adapt ROLYUN model and modified its tetrazole ligand to triazole ligand denoted as ROLYUNm for our further EXAFS fitting. The final obtained bond lengths, Debye–Waller factors, and fitting information are listed in Table 1. Thus, the final structural model of complex 1 is displayed in Figure 3 as "ROLYUNm."
1 TABLE EXAFS fitting results of complex 1.
| R (Å) | [1.8–4.4] |
|---|
| k(Å−1) | [2.5–11.4] |
| Rf | 1.7% |
| S02 | 0.8 |
| ΔE(eV) | 4.70 |
| Co‐N(Å) | 2.15(2) |
| Co‐Cl(Å) | 2.48(1) |
| Co⋯Co(Å) | 3.35(4) |
| σ12 (Co‐N) (Å2) | 0.005(2) |
| σ22 (Co‐Cl) (Å2) | 0.004(1) |
| σ32(Co–Co) (Å2) | 0.007(4) |
Structure determination by direct method
The XRD patterns of complex 1 are initially indexed by DICVOL06 in EXPO2014 software. After getting possible cell constants, we inspect the systematic absence by CMPR program to visualize the calculated individual Bragg position and its matching with experimental data. The obtained cell constants are a = 7.1041(5) Å, b = 11.7324(9) Å, c = 6.7043(5) Å, β = 90.579(4)°, V = 558.76(9) Å3 and the space group is I 2/a. The indexing results of the XRD pattern are displayed in Figure S6, in which the blue lines and yellow lines are Bragg peaks and systematic absence peaks, respectively. Based on the obtained cell constants and space group, the data collected at TPS19A (λ = 0.82656 Å; 2θ = [4°, 56.6°]) are used for structure determination. The structural model is obtained by direct method in reciprocal space based on resolution bias correction algorithm (RBM) and the covariance‐principle‐based completion method (COVMAP) in EXPO2014 program.[[22]] The hydrogen atoms are positioned with idealized geometry. The obtained structural model is refined by Rietveld method in GSAS program.[[23]] The steric restraints are set to the rigid triazole ring. However, the bond distances of CN and NN are not good enough to represent the delocalization of triazole ring. Thus, this structural model obtained by Rietveld refinement is further optimized by pDFT calculation in CRYSTAL17 program, in which the cell constants are fixed and only atomic positions are allowed to vary. Then, the optimized structural model is further performed by Rietveld refinement again. An attempt to apply March–Dollase formulation to refine the preferred orientation peak of (020) and (031) is failed, indicating that some strain or stress exists in such fiber‐like sample. The final Rietveld refinement plots are displayed in Figure 5, and the crystallographic data information is listed in Table 2. The thermogravimetric analysis (TGA) (Figure S8) indicates that there is no solvent molecule inside the structure, which is consistent with no missing atom in the difference Fourier map.
2 TABLE Crystallographic data of complex 1.
| Compound | CoII(Htrz)Cl2 |
|---|
| Formula | C2 H3 Cl2 Co N3 |
| FW (g/mo1) | 198.91 g/Mol |
| Radiation source | Synchrotron TPS19A |
| λ (Å) | 0.82656 |
| Crystal system | Monoclinic |
| Space group | I 2/a |
| Z | Z: 4 Z′: 0.5 |
| a (Å) | 7.1041(5) |
| b (Å) | 11.7324(9) |
| c (Å) | 6.7043(5) |
| β (°) | 90.579(4) |
| V (Å3) | 558.76(7) |
| T (K) | Room temp. |
| F(000) | 388 |
| 2θ (deg) | 4.0–41.9 |
| sinθ/λ (Å−1) | 0.0422–0.4326 |
| Number of measure points | 10,108 |
| Parameters | 13 |
| Rietveld refinement program | GSAS |
| Limiting indices | 0 ≤ h ≤ 8; 0 ≤ k ≤ 14; −8 ≤ l ≤ 8 |
| Observe unique reflections | 193 |
| Rp | 4.95% |
| wRp | 7.80% |
| R(F2) | 4.63% |
| Restrains | 6 |
| Δρ min; Δρ max (e/Å3) | −0.353;0.561 |
| GOF | 0.80 |
The final structure of complex 1 is displayed in Figure 6, where the cobalt and ligand planes are perpendicular to a‐axis and Co⋯Co distance is 3.3521(2) Å. The Co⋯Co⋯Co distance is almost equal to the length of c‐axis (6.7043(5) Å), indicating that Co⋯Co⋯Co 1‐D chain is almost parallel to c‐axis. The distances of Co‐N1 is 2.093(4) Å, and Co‐Cl are 2.482(4) / 2.476(4) Å. These results are comparable to the distances obtained by EXAFS analysis. Moreover, the distance between two nearest 4H‐1,2,4‐triazole ring is around 3.572 Å. These structural data are consistent with the single‐crystal results reported by Gautier et al., which the bond lengths of Co‐N1 and Co‐Cl are 2.101 Å and 2.492/2.490 Å, respectively. This single crystal was obtained in hydrothermal method by a mixture of Co3O4 (0.66 mmol), Zn (7.64 mmol), and 1,2,4‐triazole (7.24 mmol) under strong acidic condition (37% HCl, 35.95 mmol).[[2]]
Structure determination by global optimization
The indexing XRD patterns of complex 2 are similar to that of 1. The obtained cell constants are a = 9.398(2) Å, b = 17.442(3) Å, and c = 7.729(1) Å in orthorhombic system. After inspection of the systematic absence by CMPR program, the space group is Pbnm. The XRD indexing results are displayed in Figure S2. Based on the cell constants and space group, the data collected at TLS01C2 (λ = 1.03321 Å; 2θ = [5.05°, 38.86°]) are used for structure determination. By using the cell constants as the queries to search in CSD, the code FIBCEA with formula (C6H8FeN9)(BF4) is obtained. In this structure, two 4H‐1,2,4‐triazole ligands and one 1,2,4‐triazole are bridging ligands chelated to iron ions, and the tetrafluoroboric is counterion. As the cell constants and synthetic conditions are similar to our compound, an initial structural model is built by replacing the iron with cobalt. Then this model is used to search the best fitting model in real space by comparing with experimental XRD data, where the simulated annealing algorithm is used to obtain a new structure in each step for global optimization. The final structural model is an isostructure of FIBCEA. Based on the best model from simulated annealing method, a further Rietveld refinement is performed in GSAS program to complete the structure. The number of electrons in hydrogen atom is added into its bonding C or N atom, so that the number of total electrons in the unit cell can be conserved. The steric restraints are set to the rigid triazole ring. The Rietveld refinement results and perspective packing diagram are displayed in Figures 7 and 8, respectively. The crystallographic data information is listed in Table 3. The TGA results shown in Figure S9 indicate no solvent molecule inside the structure, which is consistent with no missing atom in the difference Fourier map. The Co⋯Co⋯Co distance is around 7.726 Å almost equal to c‐axis = 7.7259(3) Å, so that the 1‐D polymeric chain is parallel to c‐axis. There are two 4H‐1,2,4‐triazole and one 1,2,4‐triazole ligands in this structure. Due to the data resolution, it is hard to assign which one is 1,2,4‐triazole ligand. The bond distances between Co and N are 2.139(6) Å, 2.087(7) Å and 2.067(7) Å.
3 TABLE Crystallographic data of complex 2.
| Compound | [CoII(Htrz)2(trz)](BF4) |
|---|
| Formula | C6 Co N9.29 B F4 |
| FW (g/mo1) | 351.97 g/Mol |
| Radiation source | Synchrotron BL01C |
| λ (Å) | 1.03321 |
| Crystal system | Orthorhombic |
| Space group | P b n m |
| Z | 4 |
| a (Å) | 9.4006 (2) |
| b (Å) | 17.4331(4) |
| c (Å) | 7.7259(3) |
| V (Å3) | 1,266.13(4) |
| T (K) | 300 (room temp.) |
| F(000) | 700 |
| 2θ(deg) | 5.05–38.86 |
| sinθ/λ(Å−1) | 0.043–0.322 |
| Number of measure points | 3,382 |
| Parameters | 16 |
| Rietveld refinement program | GSAS |
| Limiting indices | 0 ≤ h ≤ 6; 0 ≤ k ≤ 12; 0 ≤ l ≤ 5 |
| Observe unique reflections | 202 |
| Rp | 2.26% |
| wRp | 3.25% |
| R(F2) | 6.22% |
| Δρ min; Δρ max(e/Å3) | −0.302; 0.260 |
Electronic structure and magnetic properties
Co L II,III ‐edge XAS
The Co LII,III edge spectrum of complexes 1 and 2 together with multiplet simulation results by CTM4XAS program are displayed in Figures 9 and 10, respectively.[[24]] The parameters used for the simulations are listed in Tables S1 and S2. Both simulations are calculated based on the transition 2p63d7 ➔ 2p53d8 with the assumption of pseudo‐octahedral environment. The best results to reproduce the feature of experimental spectrum is to set 10Dq = 1.2 eV. Thus, the ground state of Co(II) ion is at high spin state with t2g5eg2 configuration in both complexes.
Magnetic measurement
The magnetic measurement of complex 1 with temperature from 300 K to 2 K is displayed in Figure 11. The μeff value remains around at 4.75 B.M. from 300 to 100 K and gradually decreases to 2.6 B.M. at 2 K. Based on spin only model, the theoretical μeff value of Co(II) (t2g5eg2) should be 3.87 B.M. In addition to the orbital contribution to magnetic moment, the 1‐D polymeric chain may enhance the cooperative effect due to the shorter distance (~3.35 Å) between cobalt and cobalt, so that μeff may be higher than the theoretical value. The SQUID data can be fitted by Curie–Weiss law in χ−1 versus temperature plot. The Curie–Weiss Law fitting results are depicted in Figure 12, where C and θ are Curie constant and Weiss constant, respectively. The negative value of θ/C (−4.8(3)) indicates the weak antiferromagnetic interactions exist between each polymer chains.
The magnetic measurement of complex 2 is displayed in Figure 13. The μeff value remains around at 4.25 B.M with temperature from 300 K to 100 K and gradually decreases to ~1.50 B.M. from 100 K to 2 K, which indicates t2g5eg2 (Stotal = 3/2) configuration at room temperature. A small peak around 4 K may be caused by spin canting effect during extreme low temperature. The overlap of data between cool down mode and heat up mode indicates the reversibility and no hysteresis in this compound. This also indicates that there is no SCO behavior in complex 2 even though its structure is similar to [Fe(trz)(H‐trz)2](BF4) case.
Theoretical studies on magnetic interactions
The broken‐symmetry DFT calculations are based on the geometry obtained by XRD results. The calculations started with dimer model and were extended to tetramer model. In the case of complex 2, the calculation without and with counterion effects are also included. In the discussion of dimer model, the magnetic interactions of Co1 and Co2 are divided into ferromagnetic (FM) (↑↑) and antiferromagnetic (AFM) (↑↓) coupling states. In the tetramer model, the magnetic interactions of Co1, Co2, Co3, and Co4 are divided into AFM1(↑↓↑↓), AFM2(↓↑↑↓), AFM3(↓↓↑↑), and FM(↑↑↑↑) states. Overall, the results of dimer model and tetramer model are comparable. The case with and without counterion effects are consistent in complex 2. Thus, the results of dimer model are discussed in context and the others are described in Data S1. The Mulliken population analysis of complex 1 in FM and AFM coupling states is listed in Table 4. The Mulliken population indicates that both Co(II) are at the same charge and in antiferromagnetic coupling fashion with spin densities 2.81 and −2.81. The coupling constant Jab value is −0.32 cm−1, indicating that complex 1 will be stabilized at AFM state. In the calculation of tetramer model, the most stable state is at AFM3.
4 TABLE The Mulliken charge and spin densities of complex 1 after CONVER = 7.
| FM Mulliken | AFM Mulliken |
|---|
| Atom | Charge | Spin densities | Charge | Spin densities |
|---|
| Co1 | 0.506088 | 2.815056 | 0.505941 | 2.810110 |
| Co2 | 0.483208 | 2.811907 | 0.483240 | −2.806275 |
In the calculations of complex 2, the Mulliken population analysis without and with counterion effects are listed in Tables 5 and 6, respectively. The results indicate that both Co(II) are at the same charge and in antiferromagnetic coupling fashion with spin densities 2.77 and −2.77. The Jab values are −3.98 and −3.70 cm−1 for the case without and with counterion effects, respectively. It seems that the counterion does not affect so much on the magnetic property. Thus, complex 2 will be stabilized in AFM state. The calculations performed with tetramer model indicate that AFM1 state is the most stable state. The Co⋯Co distances of 1 (3.3522(2) Å) is shorter than that of 2 (3.8629(2) Å), and the coupling constants of Jab of 1 (−0.32 cm−1) is smaller than that of 2 (−3.98 cm−1). These results may indicate that the AFM coupling is through the triazole bridging ligands.
5 TABLE The Mulliken charge and spin densities of complex 2 without counterion effect after CONVER = 7.
| FM Mulliken | AFM Mulliken |
|---|
| Atom | Charge | Spin densities | Charge | Spin densities |
|---|
| Co1 | 0.887149 | 2.767812 | 0.886994 | 2.772015 |
| Co2 | 0.887145 | 2.767796 | 0.886991 | −2.772004 |
6 TABLE The Mulliken charge and spin densities of complex 2 with counterion effect after CONVER = 7.
| FM Mulliken | AFM Mulliken |
|---|
| Atom | Charge | Spin densities | Charge | Spin densities |
|---|
| Co1 | 0.867896 | 2.766932 | 0.867740 | 2.771082 |
| Co2 | 0.867869 | 2.766840 | 0.867715 | −2.771002 |
CONCLUSIONS
In this work, two 1‐D coordination polymers of complexes [Co(Htrz)Cl2]n (1) and {[Co(Htrz)2(trz)]BF4}n (2) were successfully synthesized by typical synthetic methods. The structures of 1 and 2 are determined from XRD with direct method and simulated annealing method, respectively. In complex 1, Htrz and Cl− are bridging ligands to connect two Co(II) ions to form one‐dimensional polymeric chain and the distance of Co⋯Co is 3.3521(2) Å. The local environment of Co site is in a distorted octahedron with {CoN2Cl4} core; and the bond lengths of CoN and CoCl are 2.093(4) Å and 2.483(4)/2.477(3) Å, respectively. These structural data can also be obtained by Co K‐edge EXAFS fitting. Complex 2 is also in one‐dimensional polymeric structure, where two Htrz and one trz are bridging ligands to connect two Co(II) ions in a distance of 3.8629(2) Å. The local geometry of Co is in a pseudo octahedron with {CoN6} core; and the bond lengths of CoN are 2.1514(1) Å, 2.1020(1) Å and 2.0830(1) Å. In the studies of electronic structure and magnetic interactions, the analysis of Co LII,III‐edge XAS indicates that the Co(II) of both complexes are at high spin state with t2g5eg2 configuration and the crystal field strength (10Dq) is about 1.2 eV. The broken‐symmetry DFT calculations indicate that antiferromagnetic coupling state of Co⋯Co is the most stable state in both complexes; and the coupling constants of complex 1 and 2 are −0.32 and −3.70 cm−1, respectively. Based on the distances of Co⋯Co and coupling constants, such antiferromagnetic interaction is achieved through triazole ligands. In the methodology of structure determination from powder samples, a combination with Cambridge structure database may assist the EXAFS analysis (complex 1) and the determination of initial model (complex 2) for further structure solution by XRD in real space. Meanwhile, a combination with periodic DFT geometry optimization (complex 1) can improve the results of Rietveld refinement due to the resolution limit in XRD data.
ACKNOWLEDGMENTS
We gratefully acknowledge financial support from National Science and Technology Council (NSTC), Taiwan and research fellowship from National Taipei University of Technology. The authors thank the support of SPring‐8 with the proposal number: 2013B0084 for the measurements at BL02B2 beamline. One of the authors (Yasutaka Kitagawa) acknowledges a financial support from the Grants‐in‐Aid for Scientific Research (JP22H02050). We also thank Dr. Ling‐Yun Jang and National Synchrotron Radiation Research Center of Taiwan (NSRRC) for their support on the hardware and software applied in this work.
ASSOCIATION CODES
CCDC number 2238978 and 2238981 contains the supplementary crystallographic data for this paper. These data can be obtained free of charge via www.ccdc.cam.ac.uk/data%5frequest/cif or by emailing data_request@ccdc.cam.ac.uk.
GRAPH: Data S1. Supporting information.
[
Footnotes
1
[Correction added on 19 April 2023, after first online publication: The name of the first author has been corrected in this version.]
]
[
REFERENCES
(a) H. J. Shepherd, I. A. Gural'skiy, C. M. Quintero, S. Tricard, G. Molnar, A. Bousseksou, Nat. Commun. 2013, 4, 2607. (b) I. A. Gural'skiy, C. M. Quintero, J. S. Costa, P. Demont, G. Molnar, L. Salmon, H. J. Shepherd, A. Bousseksou, J. Mater. Chem. C 2014, 2, 2949.
2
(a) S. Zaydoun, M. S. idrissi, A. Zrineh, B. Agricole, C. Garrigou‐Lagranage, Polyhedron 1995, 14, 1477. (b) I. J. Hsu, B. H. Chen, Y. C. Chuang, L. Y. Jang, J. M. Chen, J. J. Lee, J. F. Lee, Acta Cryst 2016, A72, s312. (c) R. Gautier, R. Clérac, Cryst. Growth Des. 2017, 17, 864.
3
J. Zarembowitch, O. Kahn, New J. Chem. 1991, 15, 181.
4
(a) O. Kahn, J. Krober, C. Jay, Adv. Mater. 1992, 4, 718. (b) Y. Garcia, P. van Koningsbruggen, E. Codjovi, R. Lapouyade, O. Kahn, L. Rabardel, J. Mater. Chem. 1997, 7, 857.
5
(a) W. I. F. David, K. Shankland, L. B. McCusker, C. Baerlocher, Structure determination from powder diffraction data, Oxford University Press, New York 2002. (b) C. J. Gilmore, J. A. Kaduk, H. Schenk, International tables for crystallography volume H: Powder diffraction, 1st ed., WILEY, New York 2019.
6
J. Krober, J. P. Audiere, R. Claude, E. Codjovi, O. Kahn, J. G. Haasnoot, F. Groliere, C. Jay, A. Bousseksou, J. Linares, F. Varret, Chem. Mater. 1994, 6, 1404.
7
A. Boultif, D. Louer, J. Appl. Crystallogr. 2004, 37, 724.
8
A. Altomare, C. Cuocci, C. Giacovazzo, A. Moliterni, R. Rizzi, N. Corriero, A. Falcicchio, J. Appl. Crystallogr. 2013, 46, 1231.
9
S. I. Chan, Y. J. Lu, P. Nagababu, S. Maji, M. C. Hung, M. M. Lee, I. J. Hsu, P. D. Minh, J. C.‐H. Lai, K. Y. Ng, S. Ramalingam, S.‐F. Yu, M. K. Chan, Angew. Chem. Int. Ed. 2013, 52, 3731.
F. C. Lo, J. F. Lee, W. F. Liaw, I. J. Hsu, Y. F. Tsai, S. I. Chan, S.‐F. Yu, Chem. Eur. J. 2012, 18, 2565.
Y. C. Fang, H. C. Lin, I. J. Hsu, T. S. Lin, C. Y. Mou, J. Phys. Chem. C 2011, 115, 20639.
Z. S. Lin, F. C. Lo, C. H. Li, C. H. Chen, W. N. Huang, I. J. Hsu, J. F. Lee, J. C. Horng, W. F. Liaw, Inorg. Chem. 2011, 50, 10417.
M. Li, Y. S. Huang, U. S. Jeng, I. J. Hsu, Y. S. Wu, Y. H. Lai, C. H. Su, J. F. Lee, Y. Wang, C. C. Chang, Biophys. J. 2009, 97, 609.
B. Hedman, K. O. Hodgson, E. I. Solomon, J. Am. Chem. Soc. 1990, 112, 1643.
F. Neese, Wiley Interdiscip. Rev. Comput. Mol. Sci. 2012, 2, 73.
R. Dovesi, R. Orlando, A. Erba, M. Claudio, Z. Wilson, B. Civalleri, S. Casassa, L. Maschio, M. Ferrabone, M. D. L. Pierre, P. D'Arco, Y. Noël, M. Causà, M. Rérat, B. Kirtman, Int. J. Quantum Chem. 2014, 114, 1287.
S. Grimme, J. G. Brandenburg, C. Bannwarth, A. Hansen, J. Chem. Phys. 2015, 143, 054107.
G. W. T. M. J. Frisch, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. J. A. Montgomery, J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, T. Keith, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Revision C.01, Gaussian, Inc., Wallingford CT 2010.
Y. Kitagawa, S. Nakano, T. Kawakami, K. Mashima, K. Tani, K. Yamaguchi, Mol. Cryst. Liq. Cryst. 2002, 379, 525.
(a) C. N. R. Rao, R. Venkataraghavan, Can. J. Chem. 1964, 42, 43. (b) L. W. Hartzel, F. R. Benson, J. Am. Chem. Soc. 1954, 76, 667.
A. A. El‐Azhary, H. U. Suter, J. Kubelka, J. Phys. Chem. A 1998, 102, 620.
(a) A. Nicola, C. Corrado, C. Aurelia, F. Anna, M. Rosanna, Cryst. Res. Technol. 2015, 50, 737. (b) A. Carmelo, G. Corrado, C. Aurelia, F. Anna, M. Rosanna, J. Appl. Crystallogr. 2012, 45, 789.
(a) B. H. Toby, J. Appl. Crystallogr. 2001, 34, 210. (b) A. C. Larson, R. B. Von Dreele. Los Alamos National Laboratory Report LAUR, 86–748. 2000. (c) L. B. McCusker, R. B. Von Dreele, D. E. Cox, D. LoueЀrd, P. Scardie, J. Appl. Crystallogr. 1999, 32, 36.
G. S. Henderson, F. M. De Groot, B. J. Moulton, Rev. Mineral. Geochem. 2014, 78, 75.
]
By Bo‐Hao Chen; Jun‐Jia Xu; Wei‐Ren Lai; Chung‐Kai Chang; Jeng‐Lung Chen; Jyh‐Fu Lee; Jin‐Ming Chen; Hwo‐Shuenn Sheu; Jey‐Jau Lee; Yoshiki Kubota; Ming‐Hsi Chiang; Yasutaka Kitagawa; Yu‐Chun Chuang and I‐Jui Hsu
Reported by Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author
