Thermally Stable Anilate-Based 3D CPs/MOFs

Abstract

The synthesis and characterization of two novel redox-active MOFs/CPs based on 3d transition metal ions and 3,6-ditriazolyl-2,5-dihydroxybenzoquinone (trz₂An) are reported herein. By combining trz₂An with Ni^II and Mn^II ions via the hydrothermal method, two phases, formulated as [Ni₂(trz₂An)₂]·2.5H₂O (1) and [Mn(trz₂An)(H₂O)]·1.5H₂O (2), are obtained. Both compounds crystallize as neutral polymeric 3D frameworks, where the metal ions are coordinated through the oxygen atoms of the anilate linkers forming either straight (1) or zig-zag (2) 1D chains. In particular, (1) is a MOF, where these chains are connected through the nitrogen atom at the 4 position of the triazolyl group, which completes the coordination sphere of each metal ion, affording a 3D structure containing a void volume of 28.7% and voids that can be useful for the sorption of small molecules. Interestingly, (1) and (2) show a redox behavior due to the presence of the anilate linker, being reduced electrochemically in the −0.7 to −0.9 V range due to the benzoquinone–semiquinone one-electron reduction and magnetic behavior dominated by antiferromagnetic interactions in the anilate 1D chains.

1. Introduction

Coordination polymers (CPs) and metal-organic frameworks (MOFs) are fascinating classes of molecular materials that, in recent decades, have attracted noteworthy interest in the solid-state, crystal engineering, and coordination chemistry fields. Although CPs and MOFs are both formed by the self-assembly of an organic linker and a metal node, MOFs are defined as a subclass of CPs that is characterized by the presence of potential voids [1,2,3]. By a rational design, intriguing physical properties, such as magnetic, conduction and/or luminescence, and porosity, can be tailored in the MOFs for a wide range of applications, such as gas storage, catalysis, biomedicine, sensing, etc. [4,5,6,7], that are strictly dependent on the engineering of molecular building blocks. The metal node is generally a transition metal or lanthanide ion, which can provide magnetic and/or luminescence properties [8,9,10]. The organic linker is typically a molecule that must contain coordinating functional groups, such as oxalate, succinate, squarate, 1,4-benzene-dicarboxylate, pyridine, or pyrimidine-based bridging ligands (Chart 1). Quinone derivatives and especially 2,5-dihydroxy-1,4-benzoquinones, usually known as anilates, have been extensively investigated due to (i) their Janus-type capability of coordinating different metal ions on two opposite sides of the planar benzoquinone core, (ii) the possibility of tuning their chemical structure by changing the substituents at the 3,6 positions, (iii) the possibility of mediating magnetic exchange interactions due to their Janus capability, and (iv) their redox activity, which can afford stronger magnetic exchange interactions and/or favorable charge transport pathways [11].

In fact, emerging classes of RAMOFs (redox active) and magnetic MOFs are continuously developing in response to the continuous demand for new materials that can be exploited in energy (batteries, supercapacitors, photothermal devices, electro/photocatalysis, etc.) and electronic (spintronics, sensing, semiconductors, etc.) applications [12,13]. Several examples of RAMOFs based on anilates have already been reported, with studies mainly focusing on their conductivity, magnetism and/or use in batteries/supercapacitors (vide supra) [14,15].

In this context, in 2015, Harris et al. reported on an example of a 2D-MOF incorporating the 3,6-dichloro-2,5-dihydroxy-1,4-benzoquinone linker (namely chloranilic acid, H₂Cl₂An) in its semiquinoid form as a result of a spontaneous electron transfer from Fe^II, which was chosen as the metal node, to Cl₂An²⁻. Hence iron ions are present in a mixed valence state, i.e., Fe^II/Fe^III, as well as the chloranilate linker, which is simultaneously present in its benzoquinoid and semiquinoid forms and generated in situ. The resulting MOF is a porous semiconducting antiferromagnet showing spontaneous magnetization below 80 K and a conductivity value of σ = 1.4(7) × 10⁻² S cm⁻¹ due to the mixed valences given by both the Fe^II/Fe^III and benzoquinoid/semiquinoid couples.

By substituting the 3,6 positions of 3,6-dichloro-2,5-dihydroxy-1,4-benzoquinone with triazole groups, some of us synthesized and fully characterized two novel families of robust 3D MOFs based on 3,6-N-ditriazolyl-2,5-dihydroxy-1,4-benzoquinone (trz₂An). By combining the trz₂An linker with Ln^III (Ln^III = Er^III, Tb^III, Ho^III and Dy^III) ions, we obtained a series of flexible 3D MOFs showing a combination of luminescence and Single Ion Magnet (SIM) properties. In particular, in the Er^III-based MOFs, the structural flexibility is related to a sizeable change in the emission properties [16,17]. On the other hand, the combination of trz₂An with Co^II, a transition metal ion, leads to a robust and rigid 3D porous framework employed for CO₂ uptake and separation [18].

To enlarge the study of 3D MOFs with a rigid scaffold for gas sorption, in this study, we report the synthesis and structural, magnetic, and electrochemical characterizations of (i) a MOF based on Ni^II formulated as [Ni₂(trz₂An)₂]·2.5H₂O (1) and (ii) a CP based on Mn^II formulated as [Mn(trz₂An)(H₂O)]·1.5 H₂O (2), obtained by combining the trz₂An linker with Mn^II and Ni^II nodes.

2. Materials and Methods

2.1. Materials

Reagents of analytical grade were purchased from Zentek (TCI) and Sigma Aldrich and used without further purification. The synthetic protocol of the H₂trz₂An ligand was optimized with respect to the literature [19]. Elemental analyses (C, H, and N) were performed with a CE Instruments EA 1110 CHNS (CE Instruments Ltd., Wigan, UK).

2.2. Synthesis of [Ni(trz₂An)]·2.5H₂O (1) and [Mn(trz₂An)(H₂O)]·1.5H₂O (2)

Compounds (1) and (2) were synthesized as reported in the literature [16]. MCl₂ (Ni = 0.10 mmol, 16.6 mg; Mn = 0.10 mmol, 16.2 mg), H₂trz₂An (0.10 mmol, 27.4 mg), NaOH (0.2 mmol, 8 mg), and water (10 mL) were heated in a 20 mL autoclave via a hydrothermal reaction at 130 °C for 48 h. After being cooled to room temperature, microcrystalline powder for (1) and purple needle-like crystals suitable for XRD for (2) were obtained. Elemental analysis: Calcd % for C₁₀H₁₁N₆O_7.5Mn* (390.17) C, 30.78; H. 2.8; N. 21.5. Found % C. 30.99; H. 2.46; N. 21.50. Calcd % for C₁₀H₉N₆O_6.5Ni* (375.91) C, 31.95; H. 2.41; N. 22.36. Found % C. 29.14; H. 2.74; N. 22.23.

* Calculated for [Mn(trz₂An)H₂O]·2.5H₂O as suggested by elemental analysis and [Ni(trz₂An)]·2.5H₂O as suggested by TGA.

2.3. X-Ray Diffraction (Single Crystal and Powder)

X-Ray Powder Diffraction (XRPD) measurements were performed at room temperature on a Bruker D8 Advance diffractometer using Cu Kα (1.54059 Å) radiation (40 kV, 40 mA) and a LYNXEYE XE-T detector. Gently grinded crystals of (1) and (2) were placed on a silicon monocrystal zero-background sample holder. XRD data were collected between 5° and 40° (2θ), with a step size of 0.0205° and scan time lasting ca. 20 min for (1) (0.05°, 10 min for (2)). The full procedure of the structure solution from the XRPD data for (1) and the Rietveld refinement were performed with Expo14 v2.3 software [20]. The standard peak search procedure implemented in the program enabled us to locate 21 peaks. The indexing procedure was performed wit N-Treor [21], leading to orthorhombic P n n _, a = 9.4651, b = 10.0632, c = 7.8117, M20 = 26.00, FoM_new = 4.54, NIX = 0. The analysis of the systematic absence suggested a Pnnm space group (the same space group and similar cell parameters were reported for [Co(trz₂An)]) [18]. The structure solution was performed by a simulated annealing approach and, according to the symmetry constraints suggested for previously reported [Co(trz₂An)], the Ni atom was placed on the 2a Wyckoff position. Then, a fragment of the trz₂An linker, imported from the isostructural [Co(trz₂An)], was allowed to freely float with symmetry constraints on torsional angle. Later, we tried to include water molecules within the model without success, so most likely these molecules were located in a disordered manner within the structural voids, as also reported for [Co(trz₂An)]. In the final Rietveld refinement, the background was modelled by a Chebyshev polynomial and the peak shape as a Pearson VII. Restraints on the bond lengths and angle were applied. The H atoms were refined by using the riding model with the isotropic displacement parameters U_iso constrained to be 1.2 times of that to which they were attached. Figure S4 shows the final Rietveld plot and Table S2 summarizes the results.

A single crystal of (2) was maintained in contact with the mother liquor and transferred to oil before single crystal X-Ray Diffraction (SCXRD) measurement. A suitable crystal was selected, and data were collected at 120 K on a Supernova diffractometer equipped with a graphite-monochromated Enhance (Mo) X-ray Source (λ = 0.71073 Å). The program CrysAlisPro, Oxford Diffraction Ltd., was used for unit cell determination and data reduction. Empirical absorption correction was performed using spherical harmonics, implemented in the SCALE3 ABSPACK scaling algorithm. The structures were solved with the ShelXT structure solution program [22] and refined with the SHELXL-2013 program [23], using Olex2 [24]. Non-hydrogen atoms were refined anisotropically, and hydrogen atoms were placed in calculated positions, refined using idealized geometries (riding model) and assigned fixed isotropic displacement parameters. A summary of the crystallographic data and the structure refinement is given in Table S3. CCDC 2,447,564 contains supplementary crystallographic data for (2).

2.4. Topological Analysis

A topological analysis was performed with the topcryst.com standard representation algorithm [25], which is suitable for coordination compounds and valence-bonded MOFs; it was represented with ToposPro [26]. The representation of the structures of (1) and (2) were obtained by removing the 0- and 1- connected nodes that correspond to the H atoms and the water molecules. By removing the water molecules, the size and shape of the voids were calculated and represented using the contact surface method via the Mercury software [27] (Probe Radius = 1.20 Å and Grid Spacing = 0.3 Å). The pore analysis was performed with the same software using the following default parameters (Temperature = 298 K; He probe σ = 2.58 Å; N₂ probe σ = 3.314 Å; He probe ε = 10.22 K; cutoff distance 12.8 Å; cubelet size 0.2 Å; number of samples per atom = 500). The calculated pore parameters are summarized in Table S1.

2.5. Magnetic Measurements

Magnetic measurements were performed with a Quantum Design MPMS-XL-5 SQUID magnetometer in the 2−300 K temperature range with an applied magnetic field of 0.1 T at a scan rate of 2 K min⁻¹.

2.6. Electrochemical Measurements

The solid-state cyclic voltammetry experiments for (1) and (2) were performed using a Gamry electrochemical workstation (Gamry 1010E potentiostat/galvanostat). The powdered materials (2 mg) were mixed in 2 mL of Nafion-5% and ethanol (1:10) and deposited on a 3 mm diameter glassy carbon disc working electrode, which was polished sequentially with 0.3, 0.1, and 0.05 μm alumina powders and washed with deionized water before each experiment. A typical three-electrode experimental cell equipped with a platinum wire as the counter electrode and a Metrohm Ag/AgCl electrode as a reference electrode was used for the electrochemical characterization of the working electrodes. All measurements were carried out after nitrogen bubbling. The electrochemical properties were studied measuring CVs at a 50 mV/s scan rate in CH₃CN solution of TEA-BF4 0.1M.

Cyclic voltammetry for trz₂An was carried out with a Gamry electrochemical workstation (Gamry 1000E potentiostat/galvanostat), using a three-electrode cell equipped with a Glassy carbon working electrode, an Ag/AgCl (in KCl 3 M) reference, and a counter-electrode. The experiments were performed at room temperature (25 °C) in dry and argon-degassed CH₃CN solution containing 0.1 M [(n-Bu)₄N]PF₆ as a supporting electrolyte, at 100, 50, 20, 10, and 5 mV/s scan rates. Ferrocene was added as the internal standard upon completion of each experiment.

2.7. Spectroscopic Measurements

FT-IR spectra were collected using a Bruker Equinox 55 spectrometer, preparing the samples as KBr pellets, over a scan range of 400–4000 cm⁻¹ with an average 100 scans and a 4 cm⁻¹ step.

2.8. Thermogravimetric Analysis (TGA)

TGA was performed in alumina crucibles with the STA-6000 instrument under nitrogen flux (40 mL/min), in the 25–800 °C temperature range at a 10 °C/min heating rate.

3. Results

3.1. Synthesis

Due to the great affinity of the triazolyl group for transition metal ions, when trz₂An was combined with Ni^II and Mn^II metal ions via the hydrothermal method using a 1:1 stoichiometric ratio, two new materials, formulated as [Ni(Trz₂An)]·2.5 H₂O (1) and [Mn(trz₂An)(H₂O)]·1.5H₂O (2), were obtained, as shown in Scheme 1.

3.2. Crystal Structure

A strategy for obtaining new frameworks, often used in the literature, consists of the use of the same ligand and synthetic protocol by varying the metal ions. Indeed, variations in the coordination modes of the metal ion lead to the formation of nodes with different connectivity and geometries [11]. In our case, the same synthetic protocol applied to the anilate linker and different d-ions (Mn^II, Co^II and Ni^II) led to the formation of two distinct 3D networks. The compound [Co(trz₂An)] was recently reported by some of us and corresponds to a 3D ultramicroporous MOF with a 3D topology and excellent performance in capturing and separating CO₂ from gas mixtures [18]. By replacing Co^II with Ni^II, an orange microcrystalline powder formed, which precluded single crystal diffraction analysis. On the other hand, the powder diffraction pattern suggested that the two MOFs were isostructural, as confirmed by a preliminary XRPD structure solution and refinement procedure. Indeed, the experimental pattern acquired with a laboratory diffractometer was of good quality, despite the short range and acquisition time selected. The entire powder resolution procedure was performed with the EXPO v2.3 software [20]. The same orthorhombic Pnnm space group and similar cell parameters were easily derived from the indexing and space group determination procedure. Furthermore, no peaks attributable to another phase were identified. Subsequently, the structure was solved by means of the simulated annealing (SA) approach. The obtained structure, considering symmetry and structural restraints due to the triazole and anilate ring, was then used for the Rietveld refinement [28]. A comparison between the refined cell parameters of (1) and those of [Co(trz₂An)] is reported in

Table 1. Crystal data for (1) and [Co(trz₂An)].

Species	Symmetry	Space Group	a (Å)	b (Å)	c (Å)	V (Å3)
(1)	Orthorhombic	Pnnm	9.4650 (17)	10.0556 (19)	7.8015 (14)	742.5 (2)
[Co(trz2An)]	Orthorhombic	Pnnm	9.529 (2)	10.157 (2)	7.903 (2)	764.9 (3)

, while the details of the procedure and the results are reported in the experimental section and Supplementary Materials. The low value of R_wp% = 6.565 demonstrated the isostructural nature of the two frameworks and again emphasized the effectiveness of powder diffraction data analysis for the recognition of CPs/MOFs phases [29,30,31]. The cell parameters and the connectivity and topology of the structure, obtained from this analysis, were consistent and reasonable. Furthermore, an octahedral coordination was found for Ni^II, analogous to that of Co^II in [Co(trz₂An)], with Ni-O and Ni-N bond distances of around 2.05 Å. The water molecules inside the voids were probably disordered and could not be located by the SA procedure as they were for the SCXRD structure of the [Co(trz₂An)]. TGA (Figure S1) suggested a comparable amount of around 2.5 water molecules per formula unit (11% of weight loss before 100 °C) loosely bound in the channel. An elemental analysis showed a slightly higher hydrogen and lower carbon content than the calculated ones; this may have been due to the high hygroscopicity of the sample and, consequently, to water molecules on the surface (around one per formula unit) that were not present within the channel. Indeed, this extra water molecule was not detected by the TGA, so it was most likely surface hydration water that was retained by the sample from the moisture in the air and lost under the N₂ flow used in the TGA. It is worth noting that we did not identify impurities with the other experimental techniques. Consequently, we assumed that the obtained sample was of high purity, although the development of an optimized synthetic protocol is in progress in our laboratory.

Furthermore, TGA suggested a high thermal stability of the framework, with a decomposition onset at around 410 °C. The MOF (1) can therefore be better formulated as [Ni(trz₂An)]·xH₂O (where x = 2.5). The Rietveld refined structure of (1) is shown in Figure 1, together with the reported [Co(trz₂An)] (CCDC: 2091526) for comparison.

In contrast, when Mn^II was used, purplish-red crystals, suitable for single crystal structural determination, were obtained. The X-Ray structure revealed that this compound crystallized in the orthorhombic space group Fdd2 with one Mn^II ion, one trz₂An linker, one coordinating water molecule, and two solvated water molecules, one of them with an occupancy of 0.5, in the asymmetric unit (Figure 2a). The compound can therefore be formulated as [Mn(trz₂An)(H₂O)]·1.5H₂O (2), despite the fact that the elemental analysis was more consistent with the formula [Mn(trz₂An)(H₂O)]·2.5H₂O, which may be an indication of the hygroscopicity of (2). Since this extra water molecule was not detected by the TGA, it was attributed to surface hydration water, similar to (1).

The Mn^II ions had a distorted octahedral coordination geometry composed of the N4 of the trz₂An linker, one O from a water molecule, and four oxygen atoms from two independent trz₂An linkers that coordinated the metal ion in cis bidentate mode, as shown in Figure 2b. The crystal packing showed that Mn^II ions were linked together by trz₂An linkers along the a axis, giving rise to zigzagging 1D chains, in which the shortest intrachain Mn···Mn distance was 8.231(1) Å. Each chain was linked through trz₂An linkers to the other two neighboring chains through the N4 atoms of the 1,2,4-triazolyl (see Scheme 1) substituted pendant rings, in which the shortest interchain Mn···Mn distance was 5.961(1) Å. These interconnected chains through the trz₂An linkers gave rise to the 3D neutral polymeric framework, as shown in Figure 2c. Due to the larger ionic radius of the Mn^II ion, the mean Mn-O distance of (2) (2.202(4) Å) was longer than that one observed in [Co(trz₂An)] (2.090(6) Å) and almost the same of the one reported for [[Mn₂(trz₂An)₂]·CH₃OH] (2.202(2) Å) (CCDC 1568063) [32]. The polycrystalline sample of (2) was pure and homogeneous, as confirmed by the XRPD pattern, which was perfectly consistent with the one calculated from the CIF (Figure S2). The TGA (Figure S1) of (2) showed different behavior compared to (1), although the two compounds had almost the same amount of water in the formula unit. This was attributed to the different topology and the presence of a water molecule directly coordinated to Mn^II. Indeed, the compound showed a weight loss of about 10% at about 250 °C, which could be attributed to the loss of the 2.5 water molecules in the structure that were strongly bound by the coordination bond and hydrogen bonds. The decomposition onset of (2) was around 390 °C, highlighting its high thermal stability.

There were other two phases with Mn^II and trz₂An linker, which were deposited in the CCDC by Robson et al. in 2017 (CCDC 1568062 and CCDC 1568063) [33] without synthetic details. These compounds were formulated as [Mn(trz₂An)(H₂O)]·1.5H₂O and [Mn₂(trz₂An)₂]n·3CH₃OH and crystallizesd in the monoclinic Cc and P2₁/c space groups, respectively. Also in this case, the presence of triazolyl group as substituent afforded a 3D structure, with the 1D chains being interconnected by the N4 atoms of the 1,2,4-triazolyl pendant arms. The Mn-CP (2) had a different space group and topology compared to these compounds, so these phases were completely different, even though they had a similar coordination geometry around the metal ion. The connectivity of (1) and (2) was rationalized by a topological analysis using TopCryst [25]. The result, using the standard simplification method, was a cds topology for (1) where both the linker and the metal acted as 4-connected nodes (3c) that were interconnected with each other in an alternating manner, allowing a 3D topology to occur. Since [Co(trz₂An)] was isostructural to (1), the same cds topology was assigned to this framework. The Mn-CP exhibited a different utq topology. Indeed, only one of the triazoles of the linker was involved in the coordination of the metal ion, and so both Mn^II and trz₂An acted as 3-connected nodes (3c) that yielded this particular 3D connectivity at the network. A representation of the two topologies obtained with ToposPro [26] is shown in Figure 3. The cds topology was a three-periodic net with vertex symbol 6.6.6.6.6₂.* which, in this case, consisted of Ni^II and linker nodes, both simplified as square planar 4-c nodes that were interconnected in the three dimensions by means of an alternating 90° rotation. This allowed the formation of ideally square channels. The utq topogy was a three-periodic net with vertex symbol 10.10.10₃. It formed by the almost t-shaped 3c nodes of Mn^II and the linker that were interconnected in a complex manner by helical chains composed of the metal coordinated by the O of the bidentate anilate-moiety of the core and by the N of the triazolyl substituent along c, further interconnected by the bridging bidentate anilate core in the other dimensions. This allowed the formation of channels along c, as can be seen in Figure 3b, where water molecules resided in closed pores blocked by uncoordinated triazole substituents (vide infra).

In order to understand whether these two compounds were potentially ultramicroporous, the pore parameters of (1) and (2) were calculated using Mercury software [27] after removing the solvent molecules and compared (in Table S1) with the one calculated for [Co(trz₂An)]. The results clearly showed that the two isostructural structures had similar voids shape and parameters. The software calculated a void space in the unit cell of 28.7% and 30.0% for (1) and [Co(trz₂An)], respectively. Furthermore, the pore limiting and pore maximum diameter were 1.73 Å and 3.46 Å, respectively, for (1) and 1.72 Å and 3.43 Å for [Co(trz₂An)]. Due to similar pore size (comparable to those of CO₂, around 3.30 Å of diameter) [34], we assumed that (1) could potentially be used in the separation of gaseous mixtures, like isostructural [Co(trz₂An)]. The contact surface area of (1), calculated by Mercury, is reported in Figure S3a, showing that the water molecules were located inside approximately cylindrical voids that were almost interconnected along the three dimensions. The voids were restricted near the triazole moiety, and this prevented their interconnection, using a probe radius of 1.20 Å, due to the small limiting diameter. It is worth noting that the rotation of the triazole ring was free, and the framework may have exhibited some flexibility. This could lead to the opening of these voids and the formation of 3D hourglass-like channels filled with easily removable water molecules, as also suggested by TGA. In contrast, (2) had a void space of 14.3%, i.e., almost half of the value of (1), and a pore limiting and a pore maximum diameter of 0.83 Å and 3.33 Å, respectively. The contact surface area representation (Figure S3b) suggested that, in this case, the voids were more isolated and the small pore limiting diameter, combined with the densely packed utq topology, did not allow the formation of channels. In fact, the rotation of the triazole ring was partially blocked and could not lead to the opening of channels within the structure. For these reasons, water molecules were confined inside the structural voids and were lost at high temperatures. Accordingly, compound (1) is best described as a 3D ultramicroporous MOF and (2) as a 3D CP.

3.3. FT-IR Spectroscopy

Figure S5 reports the FT-IR spectra of (1) and (2) compared with the trz₂An and Co(trz₂An) spectra. According to Pawlukojć et al. and to previously reported [Co(trz₂An)] MOF [35], the band centered at 1650 cm⁻¹ was assigned to the νCO stretching vibrational mode of the uncoordinated C=O groups in the free ligand. In the FT-IR spectra of (1) and (2), this band was downshifted compared to the free ligand, and it could be attributed to the weakened double bond character of the terminal C=O groups when coordinating the metal ion. In the 1550–1450 cm⁻¹ range, a downshifted broad band was observed for (1) and (2), which could be assigned to a νC=C + νC=O combination band, where, again, the downshift observed was attributable to the metal–ligand coordination. Furthermore, in this range, a further band at ~1520 cm⁻¹ was attributable to the νN=N stretching mode [36]. The bands present in the 1400–1100 cm⁻¹ region were assigned to the vibrational stretching of the triazolyl aromatic rings and to the νC-N vibration, both of the C-N of the triazolyl groups and of the C-N bond between benzoquinone and triazolyl ring [36,37]. The bands assignable to νM-N and νM-O vibrations could be observed in the 480–430 cm⁻¹ range [38]. Furthermore, the IR spectrum in the O-H stretching region (around 3000 cm⁻¹) showed a broad band in the case of (1), which was consistent with the presence of disordered water molecules in the channels, while in the case of (2), it showed three defined bands centered at 3497, 3421, and 3330 cm⁻¹ that could be attributed to the three crystallographically independent water molecules in the structure, one of which was coordinated and the other two strongly bonded to the network.

3.4. Magnetic Properties

The thermal dependence of the product of the molar magnetic susceptibility with temperature (χT) of (1) and (2) is reported in Figure 4. At 300 K, values of 1.4 and 4.4 emu·K·mol⁻¹ were observed for (1) and (2), respectively. These values were expected for the isolated spins of each ion with an octahedral coordination geometry, as suggested by X-Ray structures, S = 1 and S = 5/2, respectively. When the temperature decreased, χT of the two compounds decreased as well, indicating an antiferromagnetic exchange coupling between the metal ion centers. The thermal dependence of χ of the two compounds showed a continuous increase with decreasing temperature, reaching a maximum at 3.5 K for (1), which was not observed for (2) This behavior is typical for single chains of transition metal ions connected through anilate ligands, like [(M)(Cl₂An)(H₂O)₂]_n and [(M)(Cl₂An)(H₂O)₂(phz)]_n (M = Fe^II, Co^II, Mn^II; phz = phenazine), as previously reported by Kawata et al. [39,40].

The magnetic data were analyzed using an isotropic model for a chain of interacting effective spins. We could calculate the magnetic susceptibility of this Heisenberg chain using a closed chain computational procedure with an increasing number of centers (N) via the following spin Hamiltonian:

$$H=-2J{S}_1·S_{\mathrm{N}}-2{J\sum }_{i=1}^{\mathrm{N}-1}{S}_i·S_{i+1}$$

The number of centers must be increased, making sure that the difference between the ring of N and N + 1 centers were small, i.e., less than 2–5%. The adjustments proved that this difference was less than 2% around the maximum of the susceptibility and higher temperatures for each compound. Therefore, the rings of N = 6 and N = 10 for Mn^II and for Ni^II could be considered practically identical to those of the infinite chain. Calculations were performed with the MAGPACK magnetism package [41]. This model very satisfactorily reproduced the magnetic data in the whole temperature range using the following parameters: J = –1.52 cm^–1, g = 2.07 and a paramagnetic impurity 6% for (1) and J = –0.249 cm^–1 and g = 2.00 for (2). The values of the exchange parameters(J) herein used were within the normal range observed for this kind of compound (see Figure 5 and Figure 6).

The isothermal field (H) dependence of the magnetization (M) was measured up to 5 T at 2 K (Figure 7) for all two compounds. It showed a linear increase at lower magnetic fields, as expected for an antiferromagnetic behavior.

3.5. Electrochemical Properties

Considering the redox activity of anilate-based materials, due to the presence of benzoquinone rings, solid state cyclic voltammetry was performed for trz₂An, (1), and (2) to study their electrochemical behavior in the −2.0–0.0 V (vs. Ag/AgCl) potential range. For the trz₂An linker, an irreversible reduction process was observed (Figure S6). As reported in the literature, the semiquinone form of the anilate based ligands could be stabilized by the complexation of metal ions [42]. Indeed, compounds (1) and (2) showed a quasi-reversible one-electron reduction to the semiquinoid form of the anilate at Ec = −0.86 V and−0.89 V for (1) and (2) respectively, as reported in Figure 8.

4. Conclusions

By combining a 3,6-N-ditriazolyl-2,5-dihydroxy-1,4-benzoquinone (trz₂An) linker with a 3d metal transition ion via the well-known hydrothermal approach, two different compounds, a MOF and a CP, formulated as [Ni(trz₂An)]2.5H₂O (1) and [Mn(trz₂An)(H₂O)]·1.5H₂O (2), were obtained. In (1), the Ni^II ions were equatorially coordinated to four oxygen atoms of two bis(bidentate) trz₂An ligands, leading to the formation of 1D straight chains running along the c axis. The coordination sphere of M^II ions was completed with two nitrogen atoms from the N4 atoms of the 1,2,4-triazolyl substituted pendant rings of anilate linkers from two neighboring chains. The bis(bidentate) trz₂An linkers of each chain were bonded to other two neighboring chains through the N4 atoms of the 1,2,4-triazole substituted pendant rings, yielding a 3D neutral polymeric framework with a cds topology, containing voids suitable for the sorption of small molecules, as reported for the analogue compound based on Co^II [18]. When trz₂An was combined with Mn^II ions, by using the same synthetic procedure, a different compound was obtained. In this case, Mn^II ions were coordinated to two trz₂An ligands in cis, and the ions were linked by trz₂An ligands along the a axis, giving rise to zigzagging 1D chains. These chains were connected through the N4 of the triazolyl group of trz₂An, which completed the distorted octahedral geometry of Mn(II), together with the oxygen of a water molecule, yielding a 3D neutral polymeric framework with an overall utq topology. The parameters of the voids within the two compounds were analyzed in detail, showing that (1) was potentially ultramicroporous, as suggested by the isostructurality with [Co(trz₂An)], with a pore diameter comparable to the kinetic diameter of CO₂, making this MOF a potential candidate for CO₂ capture and separation from gaseous mixtures if compared with (2), due to the different utq topology that did not allow the formation of channels to occur. Compounds (1) and (2) showed a magnetic behavior dominated by antiferromagnetic interactions in the anilate 1D chains, as previously reported by Kitagawa et al., with comparable J values [39]. Therefore, it seems that magnetic interactions through the triazolyl substituent were negligible due to the lack of electronic delocalization given by the N-bonding of triazolyl groups to benzoquinone rings. Finally, the M^II-trz₂An series further showed a redox activity due to the presence of the benzoquinone core, evidenced by cyclic voltammetry, showing a reduction potential (E_c) in the −0.7 to −0.9 V range, attributable to the one-electron reduction of benzoquinone to semiquinone, paving the way to tune the magnetic and conducting properties of these MOFs by post-synthetic chemical or electrochemical reduction.