Three new cobalt(II) coordination polymers based on 1,3-bis(4-pyridyl)propane: Syntheses, structures and magnetic properties

Highlights

• Three Co(II) coordination polymers have been prepared under hydrothermal conditions.

• Their crystal structures have been investigated.

• The magnetic properties were studied in the solid state.

Abstract

Three dinuclear cobalt(II) compounds with different fluoro-substituted benzoate ligands, named as [Co₂(3,4-dfba)₄(bpp)₂]_n (1), [Co₂(3-fba)₄(bpp)₂]_n (2) and {[Co₂(2,4-dfba)₄(bpp)₂(H₂O)]·(bpp)(H₂O)}_n (3) (3,4-Hdfba = 3,4-difluorobenzoic acid, 3-Hfba = 3-fluorobenzoic acid, 2,4-Hdfba = 2,4-difluorobenzoic acid, bpp = 1,3-bis(4-pyridyl)propane), have been successfully synthesized. In all compounds, Co(II) centers feature hexa-coordinated environments with distorted octahedrons. Among them, compounds 1 and 2 are isomorphic which perform three-dimensional (3D) network, whereas 3 displays a 2D layer structure. This modulation leads to different antiferromagnetic interactions between the metal ions (J₁ = −6.43 cm⁻¹, J₂ = −2.28 cm⁻¹ for 1, J₁ = −19.44 cm⁻¹, J₂ = −6.08 cm⁻¹ for 2, J = −12.65 cm⁻¹ for 3).

Introduction

Metal-organic frameworks (MOFs) that exhibit entangled architectures are not only attractive for their aesthetic structures and topological characteristics, but also for their potential applications in multifarious areas including antimicrobial materials, catalysis, ion exchange, sensors, selective adsorption, gas storage, photochemistry and magnetic materials [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15]. In recent years, a good deal of MOFs with variable dimensionality and different coordination styles have been designed and fabricated [16], [17], [18], [19], [20], [21]. As it turns out, constructing the CPs with different dimensional frameworks is an effective and controllable strategy to design functional polymeric materials with new topologies [22], [22](a), [22](b). Furthermore, the key factor for a control of the structure and properties of the resulting CPs is the judicious selection of both metal cations and ligands [23], [23](a), [23](b), [23](c), [24], [24](a), [24](b). Among the preferred Co, Ni and Cu compounds, the exchange coupled magnetic cobalt(II) compounds are most attractive. Because the systems exhibit substantial orbital contributions to the magnetic moments and therefore possess strong magnetic anisotropy [25], [25](a), [25](b), [25](c), [25](d). The indagation of magnetic cobalt (II) compounds is advantageous to understand fundamental magnetic phenomena and will provide an effective platform for the further progress of new functional magnetic materials [26]. Although some molecular-based magnetic cobalt(II) compounds have been reported, it is difficult to design and construct molecular cobalt(II) compounds with predictable magnetism because the structural factors controlling the exchange coupling are abstruse and sophisticated.

It has by far been proven that the modification of some structural factors, such as the ligand-field strength, the replacement of a substituent in a ligand, is capable of slightly adjusting the coordination microenvironment, which in turn can make effect on the magnetic properties [27], [27](a), [27](b), [27](c), [27](d). Therefore, the selection of organic ligand and introduction of auxiliary ligand are very effective ways to design and tune the structure of CPs. As we know, among various types of spacing ligands, the flexible ligand 1,3-bis(4-pyridyl)propane (bpp) not only exhibits good adaptability to the coordination requirements of transition-metal ions and benzoate, but also is able to mold either interpenetrating or noninterpenetrating ring-opening polymerization [28], [29]. In addition, as a kind of classic ligand, fluoro-substituted benzoate are widely employed for their adequate coordinate sites and prominent angle orientation between carboxyl groups [30]. Although a number of CPs with various structures have been prepared from fluoro-substituted benzoate, systematic investigations on the external stimuli based on auxiliary ligands are still rare.

Taking the above into consideration, we choose a bpp ligand and three different fluoro-substituted benzoates, 3,4-difluorobenzoic acid (3,4-Hdfba), 3-fluorobenzoic acid (3-Hfba) and 2,4-difluorobenzoic acid (2,4-Hdfba) as coligands. And then, two novel 3D CPs and one 2D CP, [Co₂(3,4-dfba)₄(bpp)₂]_n (1), [Co₂(3-fba)₄(bpp)₂]_n (2) and {[Co₂(2,4-dfba)₄(bpp)₂(H₂O)]·(bpp)(H₂O)}_n (3) have been successfully synthesized. The important effects of different ligands on the final structures are discussed. The Co(II) centers in all instances manifest distorted octahedral geometry, which are further bridged by carboxyl groups to generate the dimers in 1–3. Magnetic investigation reveals that three compounds display distinct intermetallic antiferromagnetic exchange.

Resorting to hydrothermal synthesis, a classical and feasible method for CPs synthesis [31], [32], [33], compounds 1–3 were assembled of CoCl₂·6H₂O, 1,3-bis(4-pyridyl)-propane, and three different fluoro-substituted benzoates (details see ESI). The resulting compounds were measured by IR spectrum, EA, PXRD and TGA. Compounds 1–3 are air-stable, insoluble in common organic solvents, and can maintain the crystalline integrity at ambient conditions. TGA data shows that their main frameworks are able to remain unchanged until nearly 400°(details see ESI), indicating good thermal stability of 1–3.

Single-crystal X-ray diffraction analysis reveals that compounds 1 and 2 are isomorphous and crystallize in monoclinic space group P2₁/c (Table S1). As shown in Figs. 1 and S1, both 1 and 2 feature dinuclear clusters with nearly identical structures, consisting of two Co(II) ions, four dianionic base ligands of 3,4-dfba⁻ for 1 or 3-fba⁻ for 2, and two bpp ligands. The coordination environments of the two Co(II) ions are equivalent to each other. All Co(II) atoms form distorted octahedral geometry surrounded by six atoms for compounds 1 and 2. Two axial positions of Co(II) center are occupied by two nitrogen atoms from two independent bpp ligands, while the equatorial plane is formed by four O atoms from three fluoro-substituted benzoate ligand (Figs. 1a–c). The NCoN bond angles are very close to linearity, being 177.54(12) and 176.14(12)° for 1, 178.01(15) and 176.14(12)° for 2, respectively. The Co-N lengths [2.145(3) and 2.164(3) Å for 1, 2.143(4) and 2.173(4) Å for 2] are significantly longer than the Co-O lengths [2.008(3) and 2.224(3) Å for 1, 2.001(3) and 2.238(3) Å for 2], resulting a significantly axially elongated octahedral coordination geometries in 1 and 2. In both compounds, two Co(1) ions or Co(2) ions are bridged by the carboxyl groups to yield dimeric Co(1)···Co(1) and Co(2)···Co(2) units [Co(1)-Co(1A) = 4.132 Å and Co(2)-Co(2A) = 4.288 Å for 1, Co(1)-Co(1A) = 4.089 Å and Co(2)-Co(2A) = 4.215 Å for 2] (Figs. 1d and S1d). As shown in Fig. 1d, Co(1) dimer and Co(2) dimer are further linked by bpp ligands with conformation which contributes to flexible a helical metal chain in both compounds (Fig. 1e). The bpp linkers interconnect those dimers to form a 3D framework (Fig. 2a). The topological analysis shows that the overall structure of compounds can be rationalized to a uninodal 4-connected cds net with the point symbol of (6⁵·8) by denoting the dimetallic units as four-connected nodes (Fig. 2b).

Compound 3 crystallizes in space group P-1. The asymmetric unit is composed of two crystallographically unique Co(II) ions, four 2,4-dfba⁻ anions and two bpp ligands (Fig. 3). The Co(1) and Co(2) ions reside in a distorted hexacoordinated octahedral geometry completed with N4, O1W, O5 and N1 in axial direction and orther N, O atoms in equatorial plane. The average CoN length is 2.161 Å, while CoO lengths range from 2.045 to 2.189 Å (Table S4). Each Co1 ion is linked to one adjacent Co2 by a μ₂-H₂O and two syn-syn-carboxyl groups to form a dinuclear unit with the Co1···Co2 distance of 3.606 Å. The 2,4-Hdfba ligand in the dinuclear compound exhibits two different coordination modes of monodentate (O(4)) and μ₂-bidentate (O(5), O(6)) carboxylates bridging two Co(II) ions. Differing from 1 and 2, the coordination of flexible bpp ligand weaves the dimers of Co(1) and Co(2). The bpp ligands link the dinuclear Co(II) ions to form a limitless 1D chain (Fig. S2a), which are further fabricated into a 2D netwok parallel to the ab plane (Fig. S2b). Some lattice water and bpp molecules are located in the empty volume of whole framework.

Magnetic measurements were carried out on pure samples confirmed by powder X-ray diffraction (Fig. S3). The magnetic susceptibilities were collected with an applied field of 1000 Oe in the temperature range of 1.8–300 K for 1 and 3, 2–300 K for 2, respectively. As shown in Fig. 4, the χ_MT values of 1–3 at 300 K are 6.16, 5.207 and 6.41 cm³ mol⁻¹ K, respectively, clearly larger than the spin-only value of 3.75 cm³ mol⁻¹ K for two high-spin Co(II) ions (S = 3/2 and g = 2), indicating a considerable contribution from the orbital angular momentum. In three compounds, the values of χ_MT decrease smoothly until ~50 K and then drop down quickly to the minimum values of 2.61 and 0.31 cm³ mol⁻¹ K for 1 and 3 at 1.8 K, 1.552 cm³ mol⁻¹ K at 2 K for 2, respectively. Such tendency is mainly attributed to spin-orbit coupling of Co(II) ion and/or intradimeric antiferromagnetic coupling between Co(II) ions. Fitting of χ_M⁻¹-T using the Curie-Weiss law$({\chi }_M=C/\left(T-\theta \right))$, gives the Curie constant C = 6.46, 5.42, and 6.97 cm³ mol⁻¹ K and the Weiss constant θ = −13.74, −25.60, and −25.60 K for 1–3, respectively. The negative θ values usually indicate the domination of the antiferromagnetic coupling between the Co(II) centers. Moreover, the magnetic exchange interaction (J) between the cobalt centers in the present compounds are fitted to a Lines’ model by use of MagSaki(A) Software [34], [34](a), [34](b). The best fit of the data gives J₁ = −6.43 cm⁻¹, J₂ = −2.28 cm⁻¹, J’= −1.25 cm⁻¹, λ = −74 cm⁻¹, R = 4.8 × 10⁻⁴ for 1, J₁ = −19.44 cm⁻¹, J₂ = −6.08 cm⁻¹, J’ = −0.11 cm⁻¹, λ = −57 cm⁻¹, R = 1.62 × 10⁻⁴ for 2, and J = −12.65 cm⁻¹, J’= −0.61 cm⁻¹, λ = −78 cm⁻¹, R = 3.7 × 10⁻⁴ for 3 (R is the agreement factor defined as $\sum {\left[{\left({\chi }_{M}T\right)}_{\mathit{obsd}}-{\left({\chi }_{M}T\right)}_{\mathit{calcd}}\right]}^2/\sum {\left[{\left({\chi }_{M}T\right)}_{\mathit{obsd}}\right]}^2$). The orbital reduction factors (κ) are 0.82, 0.77 and 0.74 for 1–3, and axial splitting parameters Δ = 406.5 cm⁻¹ for 1, Δ = 339.7 cm⁻¹ for 2 and Δ = 365.2 cm⁻¹ for 3, temperature independent paramagnetism α = 1.4 × 10⁻³, 3.38 × 10⁻³, 4.59 × 10⁻³ cm³ mol⁻¹ for 1–3. The fitting result of a small negative value for J further reveals that the three compounds has antiferromagnetic interaction between the nearest Co(II) ions.

The magnetizations (M) vs. field data (H) at ~2 K are collected with fields up to 5 T for compounds 1–3, respectively. The magnetization rises quickly to the value of 4.52 Nβ for 1 and 4.2 Nβ for 2 at high fields, which are slightly smaller than the saturation value of two Co(II) ions, supporting the presence of antiferromagnetic interaction (Figs. S6a and b). In contrast, the metamagnetism is performed sigmoidal shape of the magnetization curve, from which the critical field is estimated to be 14 kOe in compound 3 (Fig. S6c). The magnetization curve increases linearly under low field, subsequently climbs up quickly until 50 kOe with an effective moment of 3.8 Nβ, also far below the common saturation value of two Co(II) centers (at 1.8 K, the effective spin is S = 1/2), exhibiting the existence of large anisotropy and antiferromagnetic interaction. Unfortunately, under the oscillating field of 2 Oe and the frequencies of 1000 Hz, the alternating current (AC) magnetic susceptibility experiments with a 0 Oe static field for 1 was determined in the range of 1.8–20 K and 2–3 were determined in the range of 1.8–15 K, no out-of phase (χ″_M) signals were observed until the temperature drop down to 1.8 K (Fig. S7).

Three dinuclear Co(II) compounds 1–3 are synthesized with bpp ligand and various benzoates under the same reaction condition. For isomorphic 1 and 2, two Co(1) ions or Co(2) ions are bridged by the carboxyl groups to yield dimeric Co(1) ··Co(1) and Co(2)···Co(2) units with different distances, whereas, structural analyses show that compound 3 presents a 2D layer network in which adjacent Co(II) ions are triply linked through the connection of two syn-syn-carboxylates and μ₂-H₂O. Although the Co(II) ions all possess hexacoordinated geometries, the substituents are clearly different in these compounds, implying divergent magnetic behaviors. For 1 and 2, the possible pathway of magnetic transmission between the dinuclear cobalt units is provided by double OCO bridges, leading to internuclear antiferromagnetic exchange with the magnetic coupling constant (J) of −6.43 cm⁻¹, −2.28 cm⁻¹ for 1 and −19.44 cm⁻¹, −6.08 cm⁻¹ for 2. In contrast, the antiferromagnetic interactions with the J value of −12.65 cm⁻¹ between Co(1)···Co(2) ions in 3 is guided by two syn-syn-carboxylates and μ₂-H₂O. Compared to the previously reported Co-MOFs with dinuclear cluster units [35], [35](a), [35](b), [35](c), [35](d), [35](e), [35](f), [35](g), the CoCo distances in 1–3 are relatively longer due to the bidentate-bridging mode of carboxylate groups, indicating similar antiferromagnetic behaviors (Table 1).

In general, employing fluoro-substituted benzoate ligands, three novel Co(II)-containing CPs have been successfully constructed with 1,3-bis(4-pyridyl)propane (bpp) coligand. Compounds 1 and 2 are isomorphic, which perform interesting 3D topological networks, whereas compound 3 is a 2D structure. The Co(II) centers in three compounds exhibit hexacoordinated distorted octahedral geometry, which are then bridged by diversely fluoro-substituted benzoates groups to form binuclear units. The result determines that subtle modification in the structure of the auxiliary ligands is a feasible approach to get access to extremely different structures. And then, their magnetic properties were also investigated, variable-temperature magnetic studies demonstrate that compounds 1–3 exhibit antiferromagnetic interactions mediated by carboxylic group, and the dominant antiferromagnetic coupling arises between two Co(II) ions. Unfortunately, the performances of slow magnetic relaxation and long-range magnetic ordering are not observed in the system.