From 1D Coordination Polymers to Metal Organic Frameworks by the Use of 2-Pyridyl Oximes

The synthesis and characterization of coordination polymers and metal–organic frameworks (MOFs) has attracted a significant interest over the last decades due to their fascinating physical properties, as well as their use in a wide range of technological, environmental, and biomedical applications. The initial use of 2-pyridyl oximic ligands such as pyridine-2 amidoxime (H2pyaox) and 2-methyl pyridyl ketoxime (Hmpko) in combination with 1,2,4,5-benzene tetracarboxylic acid (pyromellitic acid), H4pma, provided access to nine new compounds whose structures and properties are discussed in detail. Among them, [Zn2(pma)(H2pyaox)2(H2O)2]n (3) and [Cu4(OH)2(pma)(mpko)2]n (9) are the first MOFs based on a 2-pyridyl oxime with 9 possessing a novel 3,4,5,8-c net topology. [Zn2(pma)(H2pyaox)2]n (2), [Cu2(pma)(H2pyaox)2(DMF)2]n (6), and [Cu2(pma)(Hmpko)2(DMF)2]n (8) join a small family of coordination polymers containing an oximic ligand. 9 exhibits selectivity for FeIII ions adsorption, as was demonstrated by a variety of techniques including UV-vis, EDX, and magnetism. DC magnetic susceptibility studies in 9 revealed the presence of strong antiferromagnetic interactions between the metal centers, which lead to a diamagnetic ground state; it was also found that the magnetic properties of 9 are affected by the amount of the encapsulated Fe3+ ions, which is a very desirable property for the development of magnetism-based sensors.


Introduction
The synthesis and characterization of metal coordination polymers has attracted an intense research interest over the recent years due to their applications in a variety of fields, including catalysis, drug delivery sensing, etc. [1][2][3][4][5][6][7]. The structure of such species is based on mononuclear or low nuclearity inorganic units that are held together through organic ligands forming multidimensional networks whose properties are strongly affected by the nature of the metal ions and the organic linkers. For example, 1D coordination polymers of paramagnetic metal ions can display single chain magnetism (SCM) behavior, i.e., they can exhibit slow relaxation of magnetization stemming from strong intrachain exchange interactions between high spin structural building units along the chain [8][9][10][11][12][13][14]. SCMs are

Materials, Physical, and Spectroscopic Measurements
All the manipulations were performed under aerobic conditions using materials (reagent grade) and solvents as received. Hmpko and H2pyaox was prepared as described elsewhere [105,106]. WARNING: Perchlorate salts are potentially explosive; such compounds should be used in small quantities and treated with utmost care at all times.
Elemental analysis (C, H, N) were performed by the in-house facilities of National University of Ireland Galway, School of Chemistry. IR spectra (4000-400 cm −1 ) were recorded on a PerkinElmer Spectrum 400 FT-IR spectrometer. Powder X-ray diffraction data (pxrd) were collected using an Inex Equinoz 6000 diffractometer. Solid-state, variable-temperature, and variable-field magnetic data were collected on powdered samples using an MPMS5 Quantum Design magnetometer operating at 0.03 T in the 300 to 2.0 K range. Diamagnetic corrections were applied to the observed susceptibilities using Pascal's constants. TGA experiments were performed on a STA625 thermal analyzer from Rheometric Scientific (Piscataway, New Jersey). The heating rate was kept constant at 10 °C/min, and all runs were carried out between 20-600 °C. The measurements were made in open aluminum crucibles, nitrogen was purged in ambient mode, and calibration was performed using an indium standard.

Materials, Physical, and Spectroscopic Measurements
All the manipulations were performed under aerobic conditions using materials (reagent grade) and solvents as received. Hmpko and H 2 pyaox was prepared as described elsewhere [105,106]. WARNING: Perchlorate salts are potentially explosive; such compounds should be used in small quantities and treated with utmost care at all times.
Elemental analysis (C, H, N) were performed by the in-house facilities of National University of Ireland Galway, School of Chemistry. IR spectra (4000-400 cm −1 ) were recorded on a PerkinElmer Spectrum 400 FT-IR spectrometer. Powder X-ray diffraction data (pxrd) were collected using an Inex Equinoz 6000 diffractometer. Solid-state, variable-temperature, and variable-field magnetic data were collected on powdered samples using an MPMS5 Quantum Design magnetometer operating at 0.03 T in the 300 to 2.0 K range. Diamagnetic corrections were applied to the observed susceptibilities using Pascal's constants. TGA experiments were performed on a STA625 thermal analyzer from Rheometric Scientific (Piscataway, New Jersey). The heating rate was kept constant at 10 • C/min, and all runs were carried out between 20-600 • C. The measurements were made in open aluminum crucibles, nitrogen was purged in ambient mode, and calibration was performed using an indium standard.

Synthesis of [Zn 2 (pma)(H 2 pyaox) 2 ] n (2)
Zn(ClO 4 ) 2 ·6H 2 O (0.149 g, 0.4 mmol) and H 2 pyaox (0.027 g, 0.20 mmol) were dissolved in H 2 O (10 mL). The resultant solution was put in the oven and heated at 100 • C for 1 h. Solid H 4 pma (0.025 g, 0.1 mmol) was then added and the vial was placed into the oven for 24 h, after which X-ray quality Materials 2020, 13, 4084 4 of 24 colorless crystal needles of 2 were formed. The crystals were collected by filtration, washed with cold MeCN (2 mL) and Et 2 O (2 × 5 mL), and dried in air. Yield 39%. Anal. Calc. for 2: C, 40 Method B: 9 (0.0760 g,~0.1 mmol) in DMF/H 2 O (10/10 mL) was placed into the oven (100 • C) for 4 h, after which time green crystals of 9 were formed; the crystals were collected by filtration washed with MeCN (2 × 5 mL) and dried under vacuum. Yield: 80%. The product was characterized by PXRD and IR comparison with the authentic material.

Single-Crystal X-ray Crystallography
Single-crystal diffraction data for 1, 5, 6, and 8 were collected in an Oxford Diffraction Xcalibur CCD diffractometer, whereas crystallographic data for 2-4, 7, and 9 were collected in an Oxford-Diffraction SuperNova A diffractometer. Mo Ka radiation (λ = 0.71073 Å) was used for 1, 3, 5, 6, and 8, and Cu Ka radiation (λ = 1.54184 Å) was used for 2, 4, 7, and 9. The structures were solved using SHELXT [107], embedded in the OSCAIL software [108]. The non-H atoms were treated anisotropically, whereas the hydrogen atoms were placed in calculated, ideal positions and refined as riding on their respective carbon atoms. Molecular graphics were produced with DIAMOND [109].
Unit cell data and structure refinement details for 1-9 are listed in Table 1. CIF files can be obtained free of charge from the Cambridge Crystallographic Data Centre, Cambridge, UK with the REF codes 2022403-2022411 for 1-9, respectively.

Metal Ion and 2-methyluracil Adsorption Kinetic and Thermodynamic Studies
The metal adsorption capacity of 9 was investigated as described below: a salt of a metal ion (0.027 g, 0.1 mmol for FeCl 3 ; 0.040 g, 0.1 mmol for Fe(NO 3 ) 3 ) was added to a glass vial containing distilled H 2 O (10 mL) and stirred until all solid is dissolved. Solid 9 (0.242 g) was then added and the mixture was left stirring at room temperature. For the kinetic study, small volumes of aliquots were taken at designated time intervals, centrifuged, and the metal content in the supernatant solution was determined by spectroscopic (UV-vis) techniques. For the thermodynamic study, the same procedure was repeated with varying 9: metal ratios; the mixture was stirred for 20 min, filtered, and the filtrate was analyzed for its metal content. The metal encapsulation was also confirmed by magnetism studies.
The 2-methyluracil adsorption capacity of 9 was investigated following the same method used for the metal adsorption studies.

Synthetic Discussion
We have developed an intense interest over the last years in the synthesis of metal clusters and SMMs by the use of 2-pyridyl oximes as bridging ligands. These research efforts have resulted in a large number of new homo-and heterometallic new species with interesting structural features and magnetic properties, e.g., Ni 5 , Ni 12 , Ni 16 , Ni 2 Ln 2 , Ni 8 Ln 8 , Ni 2 Mn 2 , Mn 8 , etc. [83][84][85][86][87][88][89]. We recently decided to explore the ability of such ligands to favor the formation of MOFs, when combined with di-and tricarboxylic ligands, such as 1,4-benzenedicarboxylic and 1,3,5-benzenetricarboxylic acid [110]. This study yielded a new family of 1D chains and prompted us to expand this work using a tetracarboxylic ligand with higher bridging capability, in order to further increase the potential of the reaction system to provide access to new MOFs. The ligand that was chosen, and discussed herein, is the pyromellitic acid, H 4 pma. A variety of experiments were performed, studying how the different synthetic parameters (presence/absence or kind of base, molar ratio of the reactants, metal sources, etc.) affect the identity of the isolated product. In particular, the use of other metal sources instead of perchlorates (e.g., chlorides, nitrates, acetates, etc.) led to the precipitation of amorphous materials that could not be further characterized.
In order to ensure the coordination of both the oxime and carboxylic ligand in the metal center, the solution containing the metal salt and the oxime was left under stirring for at least one hour before the addition of the H 4 pma. It is noteworthy that the appropriate ratio of oxime/H 4 pma, which is able to yield binary species, was found to be 2:1; a 1:1 ratio results in known compounds that contain only the teracarboxylic ligand, whereas a high excess of oxime leads to the formation of oximato metal complexes, preventing the coordination of the ligand H 4 pma either in its neutral or anionic form.
The  (3) were isolated, depending on the reaction time; 2 is an 1D polymer and is formed after 24 h, whereas 3 is a 2D MOF, which is formed after 2 h of reaction. The stoichiometric equation of the reactions that lead to the formation of 1-3 is represented in Equations (1)-(3).
Series of experiments were also performed in order to investigate the influence of the electronic properties of the oximic ligand on the identity and structural properties of the isolated products; to this end, 2-methyl pyridyl ketoxime (Hmpko) was used instead of H 2 pyaox. The reaction of  (6) and (7).
Following that, we investigated all the synthetic parameters that could lead to a different product and the deprotonation of the oximic ligand, i.e., the molar ratio of reactants, presence/kind of base, reaction conditions, metal source, etc. Indeed, by following the same experimental procedure, but by increasing the reaction time from 1 h to 4 h, green crystals of [Cu 4 (OH) 2 (pma)(mpko) 2 ] n (9) were isolated. The formation of 9 is summarized in the stoichiometric Equation (8). 9 can be also produced in good yield by reacting 8 and Cu(ClO 4 ) 2 ·6H 2 O in a 1:2 molar ratio, according to the stoichiometric Equation (9). Note that the OH − ions in the structure of 9 come from the dissociation of H 2 O molecules.  Tables S1-S9. 1 crystallizes in the triclinic space group Pī. Its structure ( Figure 1) consists of mononuclear [Zn(H 2 pma)(H 2 pyaox)(H 2 O) 2 ] species. The coordination sphere of the metal centre is completed by one terminal N,N -bidentate chelating H 2 pyaox ligand, two terminal water molecules, and one double deprotonated H 2 pma 2− ligand. Zn1 is five-coordinate adopting a distorted trigonal bipyramidal geometry (τ = 0.8) with O2 and N3 occupying the axial positions [111].
Following that, we investigated all the synthetic parameters that could lead to a different product and the deprotonation of the oximic ligand, i.e., the molar ratio of reactants, presence/kind of base, reaction conditions, metal source, etc. Indeed, by following the same experimental procedure, but by increasing the reaction time from 1 h to 4 h, green crystals of [Cu4(OH)2(pma)(mpko)2]n (9) were isolated. The formation of 9 is summarized in the stoichiometric Equation (8). 9 can be also produced in good yield by reacting 8 and Cu(ClO4)2·6H2O in a 1:2 molar ratio, according to the stoichiometric Equation (9). Note that the OH − ions in the structure of 9 come from the dissociation of H2O molecules.  Tables S1-S9. 1 crystallizes in the triclinic space group Pī. Its structure ( Figure 1) consists of mononuclear [Zn(H2pma)(H2pyaox)(H2O)2] species. The coordination sphere of the metal centre is completed by one terminal N,N′-bidentate chelating H2pyaox ligand, two terminal water molecules, and one double deprotonated H2pma 2− ligand. Zn1 is five-coordinate adopting a distorted trigonal bipyramidal geometry (τ = 0.8) with O2 and N3 occupying the axial positions [111].  A network of intermolecular hydrogen bonding interactions stabilizes the crystal structure of 1; this involves the protonated carboxylic groups (O4, O9), the amino group (N2), and the terminal H 2 O molecules (O10, O11) as donors, and carboxylic groups from neighboring compounds (O3, O5, O6, O7, and O8) as acceptors. Details of the metric parameters of the crystallographically established hydrogen bonds in 1 are listed in Table S10.
2 crystallizes in the triclinic space group Pī; its structure ( Figure 2) is a 1D double chain formed by the connection of [Zn(pma) 0.5 (H 2 pyaox)] repeating units through the µ 4 -κO; κO ; κO"; κO " pma 4− ligand. The coordination sphere of the metal ion is completed by two terminally ligated carboxylate groups from two different pma 4− ligands, and one neutral N,N -bidentate chelating H 2 pyaox ligand. Zn 2+ adopts a tetrahedral coordination geometry. Both hydrogen atoms from the NH 2 group participate in hydrogen bonding interactions in which carboxylate groups from neighboring chains act as acceptors  3 crystallizes in the monoclinic space group P21/n and is a 2D coordination polymer based on the secondary building unit [Zn(pma)0.5(H2pyaox)(H2O)] ( Figure 3). The latter possesses a similar formula to that of the repeating unit in 2 with the main difference being the ligation of a terminal H2O molecule. Similar to 2, the two oximic ligands are neutral adopting an N,N′-bidentate chelating coordination mode. The pma 4− ligands bridge four neighboring SBUs with two of the carboxylate groups adopting a chelating coordination mode, whereas the other two being terminally ligated. It is noteworthy that the different coordination mode of the pma 4-ligand in comparison to that in 2, in addition to the presence of the H2O molecule in the coordination sphere of the metal centre, result in the Zn being six-coordinated with a distorted octahedral coordination geometry. Note that the coordination geometry of the metal ion in 2 is tetrahedral. This difference in the coordination geometry of the metal ions in 2 and 3 leads to a different orientation of the ligands around the metal ion, favoring the formation of a two-dimensional framework in 3. Compounds 4 and 5 display related structures with their main difference being the type of the 3d metal ion (4, Co 2+ ; 5, Mn 2+ ). Thus, only the structure of 4 will be discussed in detail. Compound 4 crystallize in the monoclinic space group P21/n and its structure consists of the dinuclear complex [Co2(pma)(H2pyaox)2(H2O)6] (Figure 4). The dinuclear molecules lie on a crystallographic inversion center with the two Co 2+ atoms being bridged by the μ-κO;κO′ pma 4− ligand. The coordination sphere of each metal ion is completed by one neutral N,N′-bidentate chelating H2pyaox, and three terminal water molecules. The metal ion is six-coordinated displaying a slightly distorted octahedral geometry as a result of the relatively small bite angle of the chelating ligand (N1-Co1-N2 = 75.1(2)°).  Compounds 4 and 5 display related structures with their main difference being the type of the 3d metal ion (4, Co 2+ ; 5, Mn 2+ ). Thus, only the structure of 4 will be discussed in detail. Compound 4 crystallize in the monoclinic space group P21/n and its structure consists of the dinuclear complex [Co2(pma)(H2pyaox)2(H2O)6] (Figure 4). The dinuclear molecules lie on a crystallographic inversion center with the two Co 2+ atoms being bridged by the μ-κO;κO′ pma 4− ligand. The coordination sphere of each metal ion is completed by one neutral N,N′-bidentate chelating H2pyaox, and three terminal water molecules. The metal ion is six-coordinated displaying a slightly distorted octahedral geometry as a result of the relatively small bite angle of the chelating ligand (N1-Co1-N2 = 75.1(2)°).   Table S11 in the  Supplementary Material. 6 crystallizes in the triclinic space group Pī; its structure ( Figure 5) is based on the connection of [Cu(pma) 0.5 (H 2 pyaox)(DMF)] repeating units that result in the formation of a 1D double chain coordination polymer. The pma 4ligand bridges four neighboring repeating units, adopting µ 4 -κO; κO ; κO"; κO " ligation mode. The coordination sphere of the metal ion is completed by two terminally ligated carboxylate groups from two different pma 4− ligands, one neutral N,N -bidentate chelating H 2 pyaox ligand, and one terminal DMF molecule. Cu 2+ is five-coordinate with a square pyramidal coordination geometry (τ = 0.15) with O6 from the DMF molecule occupying the axial position [111]. Both hydrogen atoms from the NH 2 group form hydrogen bonds with oxygen atoms of carboxylate groups from neighboring chains (N2···O3  Table S11 in the Supplementary Material. 6 crystallizes in the triclinic space group Pī; its structure ( Figure 5) is based on the connection of [Cu(pma)0.5(H2pyaox)(DMF)] repeating units that result in the formation of a 1D double chain coordination polymer. The pma 4-ligand bridges four neighboring repeating units, adopting μ4-κO; κO′; κO″; κO‴ ligation mode. The coordination sphere of the metal ion is completed by two terminally ligated carboxylate groups from two different pma 4− ligands, one neutral N,N′-bidentate chelating H2pyaox ligand, and one terminal DMF molecule. Cu 2+ is five-coordinate with a square pyramidal coordination geometry (τ = 0.15) with O6 from the DMF molecule occupying the axial position [111]. Both hydrogen atoms from the NH2 group form hydrogen bonds with oxygen atoms of carboxylate groups from neighboring chains (N2···O3 = 2.813 Å, H2N2···O3 = 1.937 Å, N2-H2N2···O3 = 176.69°; N2···O4 = 2.873 Å, H1N2···O4 = 2.065 Å, N2-H1N2···O4 = 150.10°). Furthermore, an intrachain hydrogen bond between the oximic group (O1) and an oxygen atom of a carboxylate group (O4) provide additional stability to the crystal structure (O1···O4 = 2.701 Å, H1O1···O4 = 1.921 Å, O1-H1O1···O4 = 146.29°). Compound 7·2H2O crystallizes in the triclinic group Pī and its crystal structure is related to that of 4, with the main differences being: 1) the type of the 3d metal ion (4, Co 2+ ; 7, Zn 2+ ), 2) the type of the oximic ligand (4, H2pyaox; 7, Hmpko), and 3) the presence of two solvate H2O molecules in the crystal structure of 7. The coordination sphere of the metal ions and the coordination modes of the ligands in both compounds is similar ( Figure S3). The crystal structure of 7 is stabilized by two strong intramolecular H bonding interactions, which involve: 1) the oximic oxygen atom (O1) as a donor and an oxygen atom (O6) from the pma 4-ligand as an acceptor (O1···O6 = 2.656 Å, H1···O6 = 1.846 Å, O1-H1···O6 = 170.36°), and 2) the solvate H2O molecule (O9, donor) and a carboxylic oxygen atom (O6, acceptor), O9···O6 = 2.815 Å, H9B···O6 = 1.965 Å, O9-H9B···O6 = 176.96°. A network of intermolecular hydrogen bonding interactions is formed between the terminal H2O molecules (O2, O3, O4; donors) and carboxylic groups (O7, O8; acceptors) or the solvate H2O molecule (O9) from neighboring molecules. The metric parameters of the intermolecular hydrogen bonding interactions for 7 are listed in Table S12 in the Supplementary Material.
Compound 8 crystallizes in the monoclinic space group P21/c. Its structure consists of [Cu(pma)0.5(Hmpko)(DMF)]n repeating units that are held together to form a 1D-double chain. (Figure 6) The coordination sphere of the Cu 2+ atom is completed by one N,N'-bidentate chelating Hmpko ligand, one terminal DMF molecule, and two O atoms (O3 and O6) that come from two different pma 4-ions; the symmetry equivalent atom of O6 (O6′) links the neighboring building units forming a one-dimensional chain, whereas O3′ bridges two Cu atoms from two parallel chains, forming a 1D-double chain. Compound 7·2H 2 O crystallizes in the triclinic group Pī and its crystal structure is related to that of 4, with the main differences being: 1) the type of the 3d metal ion (4, Co 2+ ; 7, Zn 2+ ), 2) the type of the oximic ligand (4, H 2 pyaox; 7, Hmpko), and 3) the presence of two solvate H 2 O molecules in the crystal structure of 7. The coordination sphere of the metal ions and the coordination modes of the ligands in both compounds is similar ( Figure S3). The crystal structure of 7 is stabilized by two strong intramolecular H bonding interactions, which involve: 1) the oximic oxygen atom (O1) as a donor and an oxygen atom (O6) from the pma 4- Table S12 in the Supplementary Material.
Compound 8 crystallizes in the monoclinic space group P2 1 /c. Its structure consists of [Cu(pma) 0.5 (Hmpko)(DMF)] n repeating units that are held together to form a 1D-double chain. (Figure 6) The coordination sphere of the Cu 2+ atom is completed by one N,N'-bidentate chelating Hmpko ligand, one terminal DMF molecule, and two O atoms (O3 and O6) that come from two different pma 4ions; the symmetry equivalent atom of O6 (O6 ) links the neighboring building units forming a one-dimensional chain, whereas O3 bridges two Cu atoms from two parallel chains, forming a 1D-double chain. Cu1 is five-coordinate adopting a tetragonal pyramidal geometry (τ = 0.01) with O3, O6, N1, and N2 occupying the basal plane vertices and O2 the apical position [111]. There are strong intrachain hydrogen bonding interactions that stabilize the crystal structure of 8; these are formed between the oximic group (O1, donor), and the carboxylic group of the pma 4 − ion (O5), which acts as the acceptor (O1···O5 = 2.669 Å, H1O1···O5 = 1.912Å, O1-H1O1···O5 = 155.10°). The shortest Cu···Cu distance (6.5 Å) is between atoms of the same chain, whereas the shortest interchain metal···metal distance is 7.7 Å; it is noteworthy that the latter is shorter than the intrachain distance between metal atoms that belong to different chains of the 1D-double chain (8.8 Å) in 8. 9 crystallizes in the monoclinic space group I2/a. Its crystal structure contains a three-dimensional network based on a [Cu4(OH)2(pma)(mpko)2] repeating unit (Figure 7). The latter possesses a centrosymmetric "planar-butterfly" [Cu4(μ3-ΟΗ)2] 6+ core with a Cu4 rhombus topology; the μ3-OHions lie 0.841 Å above and below the plane formed by the four metal ions. Alternatively, the core of the SBU in 9 can be described as two face-sharing defective cubanes, i.e., with a metal atom missing from one vertex of each cubane; this motif is relatively common in metal cluster chemistry. The coordination sphere of the metal atoms in 9 is completed by two doubly bridging anionic mpkoligands and four doubly bridging carboxylate groups; the mpko-adopts the μ-κO; κN; κN' coordination mode. The pma-ligands link the neighboring SBUs, resulting in the formation of the 3D network, with each one bridging four Cu4 units. All metal atoms in 9 are five-coordinate with a tetragonal pyramidal geometry (τ = 0.28, Cu1, Cu1′; τ = 0.20, Cu2, Cu2′); O2 (for Cu1 and its symmetry equivalent) and O3 (for Cu2 and its symmetry equivalent) occupy the apical positions of the pyramids [111]. The structure of 9 is stabilized by an intramolecular hydrogen bond involving the Cu1 is five-coordinate adopting a tetragonal pyramidal geometry (τ = 0.01) with O3, O6, N1, and N2 occupying the basal plane vertices and O2 the apical position [111]. There are strong intrachain hydrogen bonding interactions that stabilize the crystal structure of 8; these are formed between the oximic group (O1, donor), and the carboxylic group of the pma 4− ion (O5), which acts as the acceptor (O1···O5 = 2.669 Å, H1O1···O5 = 1.912Å, O1-H1O1···O5 = 155.10 • ). The shortest Cu···Cu distance (6.5 Å) is between atoms of the same chain, whereas the shortest interchain metal···metal distance is 7.7 Å; it is noteworthy that the latter is shorter than the intrachain distance between metal atoms that belong to different chains of the 1D-double chain (8.8 Å) in 8. 9 crystallizes in the monoclinic space group I 2/a . Its crystal structure contains a three-dimensional network based on a [Cu 4 (OH) 2 (pma)(mpko) 2 ] repeating unit (Figure 7). The latter possesses a centrosymmetric "planar-butterfly" [Cu 4 (µ 3 -OH) 2 ] 6+ core with a Cu 4 rhombus topology; the µ 3 -OHions lie 0.841 Å above and below the plane formed by the four metal ions. Alternatively, the core of the SBU in 9 can be described as two face-sharing defective cubanes, i.e., with a metal atom missing from one vertex of each cubane; this motif is relatively common in metal cluster chemistry. Cu1 is five-coordinate adopting a tetragonal pyramidal geometry (τ = 0.01) with O3, O6, N1, and N2 occupying the basal plane vertices and O2 the apical position [111]. There are strong intrachain hydrogen bonding interactions that stabilize the crystal structure of 8; these are formed between the oximic group (O1, donor), and the carboxylic group of the pma 4 − ion (O5), which acts as the acceptor (O1···O5 = 2.669 Å, H1O1···O5 = 1.912Å, O1-H1O1···O5 = 155.10°). The shortest Cu···Cu distance (6.5 Å) is between atoms of the same chain, whereas the shortest interchain metal···metal distance is 7.7 Å; it is noteworthy that the latter is shorter than the intrachain distance between metal atoms that belong to different chains of the 1D-double chain (8.8 Å) in 8. 9 crystallizes in the monoclinic space group I2/a. Its crystal structure contains a three-dimensional network based on a [Cu4(OH)2(pma)(mpko)2] repeating unit (Figure 7). The latter possesses a centrosymmetric "planar-butterfly" [Cu4(μ3-ΟΗ)2] 6+ core with a Cu4 rhombus topology; the μ3-OHions lie 0.841 Å above and below the plane formed by the four metal ions. Alternatively, the core of the SBU in 9 can be described as two face-sharing defective cubanes, i.e., with a metal atom missing from one vertex of each cubane; this motif is relatively common in metal cluster chemistry. The coordination sphere of the metal atoms in 9 is completed by two doubly bridging anionic mpkoligands and four doubly bridging carboxylate groups; the mpko-adopts the μ-κO; κN; κN' coordination mode. The pma-ligands link the neighboring SBUs, resulting in the formation of the 3D network, with each one bridging four Cu4 units. All metal atoms in 9 are five-coordinate with a tetragonal pyramidal geometry (τ = 0.28, Cu1, Cu1′; τ = 0.20, Cu2, Cu2′); O2 (for Cu1 and its symmetry equivalent) and O3 (for Cu2 and its symmetry equivalent) occupy the apical positions of the pyramids [111]. The structure of 9 is stabilized by an intramolecular hydrogen bond involving the The coordination sphere of the metal atoms in 9 is completed by two doubly bridging anionic mpkoligands and four doubly bridging carboxylate groups; the mpkoadopts the µ-κO; κN; κN' coordination mode. The pma-ligands link the neighboring SBUs, resulting in the formation of the 3D network, with each one bridging four Cu 4 units. All metal atoms in 9 are five-coordinate with a tetragonal pyramidal geometry (τ = 0.28, Cu1, Cu1 ; τ = 0.20, Cu2, Cu2 ); O2 (for Cu1 and its symmetry equivalent) and O3 (for Cu2 and its symmetry equivalent) occupy the apical positions of the pyramids [111]. The structure of 9 is stabilized by an intramolecular hydrogen bond involving the hydroxo group (O2, donor) and one of the carboxylate groups (O4, acceptor) of the pma 4ion (O2···O4 = 2.722 Å, H2···O4 = 2.113 Å, O2-H2···O4 = 133.52 • ), Figure S4.
The original framework in 9 can be simplified to its underlying net following two different approaches, the so-called standard and cluster representations [112][113][114]. Applying the first one, each Cu atom and the center of the mass of the organic ligands are the nodes, which leads to a 3,4,5,8-c net with point symbol (3.4 2 ) 2 (3.4 3 .5 2 ) 2 (3.4 4 .5 3 .6 2 ) 2 (4 12 .10 16 ) [112][113][114]. This topology ( Figure S5) has not been observed in the past, thus, 9 exhibits a network with an unprecedented architecture. On the other hand, following the cluster representation, in which the Cu atoms with the organic ligands are considered as a single four-coordinated node (Z A , Z B ), a unimodal 4-c net is formed with lvt topology ( Figure S6) and a (4 2 )(8 4 ) point symbol.
1-9 are a new family of oximic metal compounds, coordination polymers and MOFs that display unprecedented structural features. Among them, 3 and 9 are the first MOFs based on a 2-pyridyl oxime with 9 possessing a novel framework topology. 2, 6, and 8 join a small family of coordination polymers containing an oximic ligand [110]. The purity and stability of these compounds has been verified by pxrd studies.

Adsorption Studies
Thermal stability studies ( Figure S17) in 9 indicated a promising adsorption potential (ca 30% mass loss below 100 • C corresponding to adsorbed solvent), which prompted us to assess its metal adsorption capacity in detail. The metal encapsulation studies for 9 were carried out by soaking MOF crystals into aqueous solutions of Fe(NO 3 ) 3 . 9H 2 O. The MOF crystals were activated prior to the Fe 3+ encapsulation in order to reduce the amount of solvent present; this was carried out by stirring the crystals in DMF for several hours and exchanging this solvent with volatile acetone, which is easily removed at 80 • C. The metal encapsulation was initially investigated by batch studies using UV-vis spectroscopy. The maximum loading capacity obtained for Fe 3+ in the case of Fe(NO 3 ) 3 .9H 2 O as a metal source is 104 mg Fe 3+ /g 9 (1.50 mol Fe 3+ /mol 9). The Fe 3+ adsorption by 9 exhibits fast kinetics; there is a smooth increase in the adsorption capacity over time, which after 5 min reaches a plateau (Figure 8, left). No metal adsorption is observed after 5 min. In order to get a better insight into the adsorption mechanism, the experimental kinetic data were fitted to a theoretical model; pseudo-first order and pseudo-second order kinetic models were used according to Equations (10) and (11), respectively [115][116][117].
ln q e − q t = ln q e − k 1 t (10) where k 1 and k 2 are the rate constants for the pseudo-first and pseudo-second kinetic models, respectively. A good fit was obtained for the pseudo-second kinetic model (Figure 8, right), which is indicative of a chemisorption mechanism, i.e., the formation of a strong interaction between the encapsulated Fe 3+ and 9. The corresponding fitting parameters are R 2 = 0.9968, q e = 107.83 mg Fe 3+ /g 9, which is in very good agreement with the experimental data.
where k1 and k2 are the rate constants for the pseudo-first and pseudo-second kinetic models, respectively. A good fit was obtained for the pseudo-second kinetic model (Figure 8, right), which is indicative of a chemisorption mechanism, i.e., the formation of a strong interaction between the encapsulated Fe 3+ and 9. The corresponding fitting parameters are R 2 = 0.9968, qe = 107.83 mg Fe 3+ /g 9, which is in very good agreement with the experimental data. The impact of the Fe 3+ source, and in particular the impact of the kind of counterion, on the kinetic and/or thermodynamic properties of the metal encapsulation by 9 was also studied by using FeCl 3 instead of Fe(NO 3 ) 3 . It was found that the maximum loading capacity is 105 mg Fe 3+ /g 9 (1.52 mol Fe 3+ /mol 9) and is reached after 7 min stirring. The metal adsorption equilibrium data are plotted in Figure 9. The best fitting of the data is provided by the Langmuir model [118,119], considering a monolayer adsorption with a finite number of homogeneous and equivalent active sites, Equation (12): where q e (mg/g) is the amount of metal ion per gram of 9 at the equilibrium concentration C e (ppm of metal ion remaining in solution), q s is the maximum adsorption capacity of 9, and K L is the Langmuir constant; the fitting parameters are q s = 129.87 mg Fe/g 9, K L = 5 × 10 2 (L mol −1 ) and R 2 = 0.9996. The corresponding kinetic plot is shown in Figure S8. Hence, the nature of the counterion of the Fe 3+ source does not affect the metal encapsulation capacity of 9.
Other metal ions were also tested, including Co 2+ , Ni 2+ , Mn 2+ , and Cr 3+ ; however, no adsorption was observed even after one day stirring, indicating that 9 exhibits high selectivity for Fe 3+ . The impact of the Fe 3+ source, and in particular the impact of the kind of counterion, on the kinetic and/or thermodynamic properties of the metal encapsulation by 9 was also studied by using FeCl3 instead of Fe(NO3)3. It was found that the maximum loading capacity is 105 mg Fe 3+ /g 9 (1.52 mol Fe 3+ /mol 9) and is reached after 7 min stirring. The metal adsorption equilibrium data are plotted in Figure 9. The best fitting of the data is provided by the Langmuir model [118,119], considering a monolayer adsorption with a finite number of homogeneous and equivalent active sites, Equation (12): where qe (mg/g) is the amount of metal ion per gram of 9 at the equilibrium concentration Ce (ppm of metal ion remaining in solution), qs is the maximum adsorption capacity of 9, and KL is the Langmuir constant; the fitting parameters are qs = 129.87 mg Fe/g 9, KL = 5 × 10 2 (L mol −1 ) and R 2 = 0.9996. The corresponding kinetic plot is shown in Figure S8. Hence, the nature of the counterion of the Fe 3+ source does not affect the metal encapsulation capacity of 9.
Other metal ions were also tested, including Co 2+ , Ni 2+ , Mn 2+ , and Cr 3+ ; however, no adsorption was observed even after one day stirring, indicating that 9 exhibits high selectivity for Fe 3+ . The Fe 3+ adsorption capacity of 9 was further investigated by EDX studies. Figure 10 shows the EDX spectra of 9 and Fe@9; Cu, C, and O are detected in both samples, with the the second sample displaying one additional peak corresponding to the Fe 3+ ion that has been adsorbed, hence confirming the metal uptake. It is noteworthy that 9 can be easily regenerated from Fe@9 by treatment with a 0.2 M EDTA solution. The activated MOF displays an identical sorption capacity with that of the original 9, which is retained for three cycles of regeneration/reuse experiments ( Figure S9). Thus, 9 is reusable, a property which is paramount for practical applications. The Fe 3+ adsorption capacity of 9 was further investigated by EDX studies. Figure 10 shows the EDX spectra of 9 and Fe@9; Cu, C, and O are detected in both samples, with the the second sample displaying one additional peak corresponding to the Fe 3+ ion that has been adsorbed, hence confirming the metal uptake. It is noteworthy that 9 can be easily regenerated from Fe@9 by treatment with a 0.2 M EDTA solution. The activated MOF displays an identical sorption capacity with that of the original 9, which is retained for three cycles of regeneration/reuse experiments ( Figure S9). Thus, 9 is reusable, a property which is paramount for practical applications. The good Fe 3+ adsorption capacity of 9 prompted us to evaluate its encapsulation performance for other species. 6-methyluracil is a compound for which the development of new sorbents or sensors is of great importance; it is a key component of many drugs, including anti-ulcer agents, radiation protective agents, and immunological adjuvants which help enhance immune responses, but on the other hand, it can also be encountered as a toxic byproduct in the synthesis of biologically important DNA adducts. If small impurities of such byproducts are not identified and removed, it could have devastating consequences for the corresponding biological assays and medical treatments. Thus, the development of efficient carriers/sorbents for the delivery or removal of these species would provide new insight into drug manufacturing. Considering also the fact that MOFs have not yet been used for drug purification, this would expand the range of potential applications of MOFs. With the above in mind, and following a similar process to the one used for metal adsorption, the 6-methylyracil capacity of 9 was investigated. This is an ongoing project, but preliminary results ( Figure S10) indicate that 9 possess a good performance in 6-methyluracil adsorption; the maximum uptake capacity is 167 mg 6-methyluracil/g 9 for a 1:1 ratio 6-methyluracil:9 in an aqueous solution. It exhibits fast kinetics with the maximum encapsulation being reached after 6 min stirring. It is noteworthy, though, that the regeneration of the MOF after the 6-methyluracil adsorption is not feasible. These initial results show the potential of 9, and MOFs in general, to be used in purification processes during drug manufacturing. Investigations are now in progress to assess the selectivity of 9 for 6-methyluracil and other toxic byproducts, which is a requirement for the use of MOFs in such applications.

Magnetism Studies
Solid state DC magnetic susceptibility measurements were performed on a polycrystalline sample of 9 in an applied field 0.03 T and temperature range 2-300 K. The obtained data are shown as a χMT vs. T plot in Figure 11. The χMT value at 300 K is 0.075 cm 3 ·mol −1 K, appreciably lower than the spin-only (g = 2) value of 1.5 cm 3 ·mol −1 K expected for four non-interacting Cu 2+ centres (S = 1/2), revealing strong antiferromagnetic coupling within the framework SBU. This is further supported by the overall profile of the plot, which is indicative of strong antiferromagnetic interactions and a diamagnetic ground state for 9. The good Fe 3+ adsorption capacity of 9 prompted us to evaluate its encapsulation performance for other species. 6-methyluracil is a compound for which the development of new sorbents or sensors is of great importance; it is a key component of many drugs, including anti-ulcer agents, radiation protective agents, and immunological adjuvants which help enhance immune responses, but on the other hand, it can also be encountered as a toxic byproduct in the synthesis of biologically important DNA adducts. If small impurities of such byproducts are not identified and removed, it could have devastating consequences for the corresponding biological assays and medical treatments. Thus, the development of efficient carriers/sorbents for the delivery or removal of these species would provide new insight into drug manufacturing. Considering also the fact that MOFs have not yet been used for drug purification, this would expand the range of potential applications of MOFs. With the above in mind, and following a similar process to the one used for metal adsorption, the 6-methylyracil capacity of 9 was investigated. This is an ongoing project, but preliminary results ( Figure S10) indicate that 9 possess a good performance in 6-methyluracil adsorption; the maximum uptake capacity is 167 mg 6-methyluracil/g 9 for a 1:1 ratio 6-methyluracil:9 in an aqueous solution. It exhibits fast kinetics with the maximum encapsulation being reached after 6 min stirring. It is noteworthy, though, that the regeneration of the MOF after the 6-methyluracil adsorption is not feasible. These initial results show the potential of 9, and MOFs in general, to be used in purification processes during drug manufacturing. Investigations are now in progress to assess the selectivity of 9 for 6-methyluracil and other toxic byproducts, which is a requirement for the use of MOFs in such applications.

Magnetism Studies
Solid state DC magnetic susceptibility measurements were performed on a polycrystalline sample of 9 in an applied field 0.03 T and temperature range 2-300 K. The obtained data are shown as a χ M T vs. T plot in Figure 11. The χ M T value at 300 K is 0.075 cm 3 ·mol −1 K, appreciably lower than the spin-only (g = 2) value of 1.5 cm 3 ·mol −1 K expected for four non-interacting Cu 2+ centres (S = 1/2), revealing strong antiferromagnetic coupling within the framework SBU. This is further supported by the overall profile of the plot, which is indicative of strong antiferromagnetic interactions and a diamagnetic ground state for 9. 18 Figure 11. χMT vs. T plot for 9 (circles) Fe@9-1 (triangles) and Fe@9-2 (squares). Solid line represents the best fit for 9. Inset: the exchange interactions in 9.
The isotropic Heisenberg spin Hamiltonian for 9 is given by Equation (12), where J1 is associated with the Cu1…Cu2′ interactions through a diatomic oximato group; J2 is associated with the Cu1…Cu2 interaction through one monoatomic carboxylate bridge; and J3 describes the coupling of Cu1…Cu1′ through a μ3-OH − group ( Figure 11).
The fitting of the experimental data to the Hamiltonian was performed using the PHI software [120] and is shown with a solid line in Figure 10; the fitting parameters are J1 = −508 cm −1 , J2 = −1.45 cm −1 , J3 = −8.77 cm −1 , and g = 2.07. J1 is significantly stronger than J2 and J3; this high value of antiferromagnetic coupling is common in complexes containing oximes and Cu ions; such complexes often exhibit diamagnetic behavior [82]. J2 and J3 are too small to take them as reliable values because at the temperature they start to operate, the complex is diamagnetic due to the extremely high value of J1.
An initial investigation of the potential of magnetism to be used for the development of novel MOF-based sensors was performed by exploring the impact of the encapsulated metal ion on the magnetic properties of 9. To this end, DC magnetic susceptibility studies were performed on the polycrystalline samples of Fe@9-1 and Fe@9-2, where Fe@9-1 is the Fe@9 aggregate after 1 min stirring, during which 9 is expected to have reached ca. its half metal encapsulation capacity, and Fe@9-2 is the aggregate at its maximum metal adsorption capacity. A closer inspection in Figure 11 shows that, although the overall profile of the three plots is similar, the increase of the paramagnetic component in the MOF pores have an effect on the observed χMT value at room temperature; the latter increases from 0.073 cm 3 ·mol −1 K (for 9) to 0.115 cm 3 ·mol −1 K (for Fe@9-1) and 0.135 cm 3 ·mol −1 K (for Fe@9-2), indicating that the magnetic properties of the Fe@9 aggregate are affected by the amount of metal encapsulated in the MOF pores.

Conclusions
The initial employment of a 2-pyridyl oxime (pyridine-2 amidoxime, H2pyaox; 2-methyl pyridyl ketoxime, Hmpko) in combination with 1,2,4,5-benzene tetracarboxylic acid (pyromellitic acid), H4pma, provided access to nine new compounds with interesting structural features, paving the way to the development of an alternative synthetic route towards new MOFs and coordination polymers.  The isotropic Heisenberg spin Hamiltonian for 9 is given by Equation (13), where J 1 is associated with the Cu1 . . . Cu2 interactions through a diatomic oximato group; J 2 is associated with the Cu1 . . . Cu2 interaction through one monoatomic carboxylate bridge; and J 3 describes the coupling of Cu1 . . . Cu1 through a µ 3 -OH − group ( Figure 11).
The fitting of the experimental data to the Hamiltonian was performed using the PHI software [120] and is shown with a solid line in Figure 10; the fitting parameters are J 1 = −508 cm −1 , J 2 = −1.45 cm −1 , J 3 = −8.77 cm −1 , and g = 2.07. J 1 is significantly stronger than J 2 and J 3 ; this high value of antiferromagnetic coupling is common in complexes containing oximes and Cu ions; such complexes often exhibit diamagnetic behavior [82]. J 2 and J 3 are too small to take them as reliable values because at the temperature they start to operate, the complex is diamagnetic due to the extremely high value of J 1 .
An initial investigation of the potential of magnetism to be used for the development of novel MOF-based sensors was performed by exploring the impact of the encapsulated metal ion on the magnetic properties of 9. To this end, DC magnetic susceptibility studies were performed on the polycrystalline samples of Fe@9-1 and Fe@9-2, where Fe@9-1 is the Fe@9 aggregate after 1 min stirring, during which 9 is expected to have reached ca. its half metal encapsulation capacity, and Fe@9-2 is the aggregate at its maximum metal adsorption capacity. A closer inspection in Figure 11 shows that, although the overall profile of the three plots is similar, the increase of the paramagnetic component in the MOF pores have an effect on the observed χ M T value at room temperature; the latter increases from 0.073 cm 3 ·mol −1 K (for 9) to 0.115 cm 3 ·mol −1 K (for Fe@9-1) and 0.135 cm 3 ·mol −1 K (for Fe@9-2), indicating that the magnetic properties of the Fe@9 aggregate are affected by the amount of metal encapsulated in the MOF pores.

Conclusions
The initial employment of a 2-pyridyl oxime (pyridine-2 amidoxime, H 2 pyaox; 2-methyl pyridyl ketoxime, Hmpko) in combination with 1,2,4,5-benzene tetracarboxylic acid (pyromellitic acid), H 4 pma, provided access to nine new compounds with interesting structural features, paving the way to the development of an alternative synthetic route towards new MOFs and coordination polymers. SBU. DC magnetic susceptibility studies revealed that there are strong antiferromagnetic interactions between the metal centers in 9, which lead to a diamagnetic ground state.
The metal encapsulation capacity of 9 was tested for a variety of different metal ions and showed that it exhibits selectivity for Fe 3+ adsorption. It is noteworthy that the magnetic properties of 9 depend on the amount of Fe 3+ present in to the MOF pores, revealing that magnetism can be an alternative technique for the detection of environmentally hazardous chemicals, and this can be especially useful for species that do not affect the photoluminescence properties or the color of a compound, hence they cannot be detected by the commonly used sensors. Treatment of the Fe@9 aggreagate with a 0.2 M EDTA solution leads to the removal of the metal ions from the MOF pores and the regeneration of the latter. 9 retains its metal adsorption capacity for three cycles of regeneration/reuse experiments.
Supplementary Materials: The following are available online at http://www.mdpi.com/1996-1944/13/18/4084/s1, Figure S1: Representation of the 3D network formed through hydrogen bonding interactions in 2. Figure S2: Representation of the molecular structure of the dinuclear complex 5. Figure S3: Representation of the molecular structure of the dinuclear complex 7. Figure S4: Representation of the intramolecular hydrogen bonding interactions in 9. Figure S5: Representation of the underlying 3,4,5,8-coordinated net in the standard representation in 9. Figure S6: The underlying net in the cluster representation with the lvt topology with Point Symbol (4 2 )(8 4 ) in 9. Figure S7: Comparison of the theoretical and experimental pxrd pattern for 9. Figure S8: Left: metal adsorption capacity (mg g −1 ) versus time (h) plot for the encapsulation of FeCl 3 by 9; right: simulation of the experimental data to the pseudo-second order kinetic model. The solid lines represent the fitting of the data. The corresponding fitting parameters are R 2 = 0.9937, q e = 103.09 mg Fe 3+ /g 9, in very good agreement with the experimental data. Figure S9: Comparison of the pxrd pattern of the initial 9 with that of the regenerated material. Figure S10: UV studies of the 6-methyluracil adsorption by 9; (a) 0.05 mmol 9/0.1 mmol 6-methyluracil and (b) 0.1 mmol 9/0.1 mmol 6-methyluracil (in 10 mL of water). Tables S1-S9: Selected interatomic distances (Å) and angles for 1-9. Table S10: Hydrogen bonding details for 1.

Conflicts of Interest:
The authors declare no conflict of interest.