The ubiquitous paddle-wheel building block in two-dimensional coordination polymers with square grid structure

Abstract This work describes design of a series of new paddle-wheel binuclear clusters containing 2-D coordination polymers based on ditopic carboxylate linkers, 1,4-benzenedicarboxylate (BDC) or 2-amino,1,4-benzenedicarboxylate (Am-BDC). The strategic use of strongly coordinating base/solvent as blocking ligand to restrict the structure in 2-D space is explored, and the role of organic base on the overall structure formation is further elaborated. The isostructural [Zn(BDC)(Py)]n (1) and [Co(BDC(Py)]n (2) were formed by the use of strong base pyridine (Py) as a blocking ligand whereas reaction using N-methylimidazole (Mim) in place of pyridine gives [Co(BDC)(Mim)]n (3) with similar topology and coordination environment. The use of weak/non-coordinating base such as 2-chloropyrimidine, pyrazine, and tetramethylammoniumhexafluorophosphate [(CH3)4 N(PF6)] gives the DMF-coordinated 2-D frameworks, [Cu(BDC)(DMF)]n (4), [Zn(BDC)(DMF)]n (5), and [Zn(AmBDC)(DMF)]n (6). All the structures crystallize in monoclinic crystal system yielding 2-D nets with square grid 44 topology and solid state 3-D structure via extensive non-covalent supramolecular interactions. Surface area analysis via N2 adsorption of three representative 2-D coordination polymers, 1, 4, and 6, indicate that 4 has a surface area of 450 m2 g−1 which is a signature of microporosity, while 1 and 6 have moderate (161.6 m2 g−1) and negligible (33 m2 g−1) surface areas, respectively.


Introduction
Polytopic organic linkers have evolved in a multifaceted way with application in materials chemistry through the formation of metal-organic frameworks (MOFs) [1][2][3][4][5][6]. The ability to form diverse building blocks make them versatile and amenable for designer chemistry. For example, the rational design of structures employing the concept of reticular chemistry in a well-known class of coordination polymers/ MOFs is strongly dependent on the precise use of polytopic organic linkers [7,8]. Although variety in the wide range of structures and their corresponding properties are credited to the functionality of the linker, flexibility in coordination with metals to form different secondary building units (SBUs) and stability of final structures, ample challenges still exist in design of coordination polymers with a predetermined architecture specifically to establish the correct reaction conditions for desired materials [9]. The building of novel coordination polymers is in itself an exciting arena in supramolecular chemistry and draws allegiance to a separate branch of chemistry, namely crystal engineering [10]. One of the approaches frequently adopted in crystal engineering is to use SBUs as synthons to build a desired architecture. One of the most ubiquitously present SBUs is the so-called "paddle-wheel" (PW) motif, which can be found in molecular complexes and extended polymeric structures such as 1-D and 2-D networks and 3-D frameworks [11,12]. The PW building unit, with two metal ions (with or without a metal-metal bond) equatorially bridged by four carboxylates in a syn-syn configuration, plays an important role in developing functionality of the MOF. For example, creation of an open coordination site through removal of coordinated ligand from the axial position of PW motif can render the MOF to be catalytically active, transform it to a good adsorbent and a sensor for gasses. Some of the iconic MOFs, such as 3-D framework of HKUST-1 [13], 2-D network of square grid topology in MOF-2 [14], and the pillared square grid of 2-D networks [15], all contain PW SBU and display the above-mentioned properties. We, therefore, sought to develop synthesis protocols for 2-D square grids containing PW motif varying the axial ligands and their influence in the stacking of 2-D layers and subsequent effect on porosity. The design of 2-D architecture from the polytopic linker requires manipulation at the molecular level by truncating the growth in one specific direction with the use of blocking ligands (blockers) in the axial position. These blockers are most of the time strong donors such as pyridine, which enables the control of structural architecture in one-and two-dimensions for the same reactants [16]. in this context, the ditopic 1,4-benzenedicarboxylate is recognized as a suitable linker and known to form stable 2-D and 3-D porous MOFs with a wide variety of SBUs with versatile nodes [10,17]. Therefore, using the principle of incorporating a blocker in designing 2-D structures, we herein report the synthetic methodology and crystal structures of six 2-D coordination polymers with Zn, Co, and Cu, [Zn(BDC)(Py)] n (1), [Co(BDC(Py)] n (2), [Co(BDC)(Mim)] n (3), [Cu(BDC)(DMF)] n (4), [Zn(BDC)(DMF)] n (5), and [Zn(AmBDC)(DMF)] n (6) (BDC = 1,4-benzenedicarboxylate; Py = pyridine; Mim = N-methylimidazole; AmBDC = 2-amino-1,4-benzenedicarboxylate; DMF = N,N-dimethylformamide), containing binuclear PW metal clusters. The gas sorption experiments for the three representative MOFs (1, 4, and 6) are also reported and the development of microporosity in 4 is also discussed in this article.

Synthesis, characterization, and crystal structure
The PW binuclear clusters 1-5 containing 2-D frameworks were synthesized by reaction of equimolar amount of metal salt and BDC dissolved in DMF/etOH (see Experimental section for details) mixture whereas the optimized metal : BDC molar ratio in case of 6 was 2 that yielded single-phase product. Under the given condition, single-phase products were isolated for 1, 4, and 6; however for 2, 3, and 5, some unidentified impurities were present. The 2-D [Zn(BDC)(Py)] n (1) coordination polymer was synthesized by addition of BDC and pyridines to a solution of Zn(NO 3 ) 2 in a binary mixture of ethanol and DMF. The reaction produced off-white crystals that crystallized in monoclinic P2 1 /c space group (table  1) with cell parameters a = 7.4046(6) Å, b = 16.5754(13) Å, c = 11.0692(10) Å, β = 115.5850(10)°, and volume = 1225.36(18) Å 3 . The asymmetric unit consists of one Zn, one BDC, and one pyridine ( figure 1(a)) with a formula unit of [Zn(BDC)(Py)]. The polymeric unit consists of PW binuclear Zn(ii) tetracarboxylate units with inter-metallic Zn distance of 2.987(3) Å and two axial pyridines ( figure 1(b)). The internuclear Zn⋯Zn distance is almost comparable with other reported PW Zn 2 (COO) 4 [14], [Zn(L)(DeF)]·DeF (where L = 4,4′-(2,3,5,6-tetramethylbenzene-1,4-diyl)dibenzoic acid; DeF = N,N-diethylformamide) (2.9464(5) Å) [18], etc. The framework is neutral where eight carboxylate oxygens bind to each of the four coordination sites of Zn and pyridine occupy the axial positon. Such bimetallic PW arrangements give stability to the framework and are extensively reported with transition metals [11]. identical reaction using Co(OAc) 2 ·4H 2 O gave polymeric  [Co(BDC)(Py)] (2), isostructural to 1 (figure S1-a) as a pink solid with similar PW binuclear cluster (figure S1-b) and a 2-D square grid coordination network (figure S1-c). However, the internuclear Co distance in 2 is 2.7650(8) Å, much smaller than Zn⋯Zn in 1 but in the range of Co⋯Co distances measured in other structures with Co 2 (RCOO) 4 cluster [19][20][21][22]. The metal(ii) PW clusters in 1 and 2 are bridged with BDC to form infinite polymeric 2-D (4 4 ) layers parallel to (1 0 2) planes with M 2 (CO 2 ) 4 as nodes (figure 1(c)). each 2-D layer forms a square grid of ~ 10.58-10.58 Å 2 dimension and are stacked perpendicular to (1 0 2) planes with an offset which allows the pyridine molecules from the adjacent layers to fill the square aperture of the layer as shown in figure 2(a). Stacking of layers in both 1 and 2 leads to a 3-D supramolecular crystalline array, which when viewed along the a-axis, a 1-D diamond-shaped channel can be observed wherein protrudes the pyridine molecules (figure 2(b)). However, in reality, space-filling modes do not show pores when viewed along the a-axis, but small rhombic-shaped channels can be seen when viewed along the (6 1 10) planes (figure S2). The 3-D assembly is stabilized by non-covalent π⋯π and H-bonding interactions (tables S1-S4). interactions of π-electron clouds of pyridine ring with adjacent pyridine rings and with benzene ring of BDC give π⋯π interactions (figure 2(c)). The Cg (1)  properties to 1. The phase purity of the crystals acquired was analyzed and found to have formed in excellent bulk purity as seen from good matching of simulated and experimental PXRD patterns of 1 ( figure 3(a)). The FT-iR spectrum reveals a shift in the stretching frequency of ν(C = O) from 1720 cm −1 in uncoordinated BDC to 1631 cm −1 in 1, indicative of coordination of carboxylate and the formation of a metal chelate [23,24]. The presence of coordinated pyridine was confirmed by shift of the band at 403, 601, and 745 for free base to 420, 642, and 753 cm −1 for coordinated pyridine [25]. Furthermore, thermogravimetric analysis (TGA) supports the formula unit as proposed from the crystal structure. The TGA profile for 1 given in figure 3(b) shows that the observed weight loss of 25.89% between 240-320 °C accounts for near-complete displacement of pyridine (calculated weight loss = 25.65%) from the crystal structure. The initial weight loss is followed by a plateau till 420 °C, which strongly indicates the existence of a new stable phase upon removal of pyridine. On the other hand, no weight loss till 240 °C clearly shows that the structure is fairly stable and free from any non-coordinating solvent/guest molecules.

Base-directed solvent coordination
The coordination of pyridine truncated the growth of the PW SBU along the apical direction and turned out to be a strong blocker for the coordination polymer resulting in a 2-D framework. 2-D MOF-2 can be formed even in the absence of strong base, such as pyridine, but the coordination of pyridine to the apical position shows that any further growth of the 2-D structure in the presence of competitive auxiliary bridging ligands is restricted [14].      5(c)). Analogous to 1-3, the layers are stacked perpendicular to the (1 0 2) planes with an offset to accommodate the DMF in the square aperture of the layer, forming a 3-D supramolecular assembly, which is stabilized by H-bonding interactions (tables S6 and S7). The stacking of layers in 4 and 5 creates a 1-D channel along the a-axis wherein protrude the DMF molecules similar to pyridine in the case of 1, 2, and N-methyl imidazole for 3. As a representative case, the PXRD pattern and TGA thermogram of 4 are given in figure 6. A comparison of PXRD patterns of as-synthesized 4 (figure 6(a)) to the one simulated from single crystal shows high bulk purity. The TGA profile indicates stable thermogram till 180 °C and shows two steps of weight loss ( figure 6(b)). The observed first weight loss of 24.45% between 180-310 °C agrees well with the theoretical value (24.30%) calculated for loss of DMF from the structure whereas the second weight loss step from 360-490 °C can be due to decomposition of the structural framework ( figure 6(b)). The use of pyrazine or 2-chloropyrimidine instead of pyridine or Mim gives coordination of DMF to the open axial coordination site of the PW SBUs that resulted in formation of 4 and 5. earlier use of benzimidazole produced DMF-coordinated geometry in the case of 2-D Cu(ii) coordination polymer [Cu 2 (BDC) 2 (DMF)](DMF)·H 2 O·(C 2 H 5 -OH) 0.5 ] [28]. Therefore, the coordination of DMF may have been facilitated by either lower acidity (pk a ) of pyrazine/2-chloropyrimidine than that of pyridine (pKa = 5.24) or  The solid crystallizes in monoclinic C2/m space group and forms a 2-D layered structure with PW binuclear cluster bridged by BDC as the building unit ( figure 8(a)). The asymmetric unit consists of one zinc, one DMF, and half of the dicarboxylate unit ( figure 8(b)). Occupancy of N1(amino group) is 0.5 and has been modelled to avoid corresponding symmetry-generated amino group within the same ring. The solvent DMF is coordinated directly to Zn(ii), modelled as disordered over two equal sites (occupancy 0.5) across the crystallographic mirror plane. isostructural Zn(Am-BDC)(DMF)·(C 6 H 5 Cl) 0.25 (MOF-46) forms in DMF-chlorobenzene solvent containing chlorobenzene-trimethylamine (700 : 1, v/v) and incorporates a small amount of chlorobenzene into the structure [17]. However, in our synthesis there is no evidence of solvent intercalation either through crystallographic or TGA. The Zn(ii) PW clusters in 6 are bridged with Am-BDC to form infinite polymeric 2-D square grid (4 4 ) layers in the (2 0 1) planes as shown in figure 9(a) and the 2-D layers pack into a 3-D supramolecular structure by strong C-H⋯π and H-bonding interaction (tables S8 and S9). interlayer stacking in 6 is in coherence with other structures (1)(2)(3)(4)(5), however, the amino group in BDC ( figure 9(b)) is oriented to occupy the channels between the staggered layers; when viewed along the c-axis the 1-D channel is occupied by the disordered DMF molecules.  These 2-D MOFs (1)(2)(3)(4)(5)(6) are similar and essentially the same as MOF-2 reported by yaghi [14]. The identity of the axial ligand in the PW building unit and the nature of the intercalated solvent have subtle influences in the crystallization of theMOFs in different crystal systems, or different space groups in the same crystal system. Non-covalent interactions involving the axial ligand or the occluded solvent slightly alter the interlayer stacking. The interlayer arrangement in this class of square grid structures along different crystallographic axes gives diamond-shaped 1-D channels, wherein reside the coordinated solvent molecules essentially blocking the channels. in 1-5, the perpendicular interlayer distance is 5.18 Å, in agreement with earlier reported values [14,27] whereas the interlayer distance was 6.29 Å in 6. Non-covalent interactions facilitated by the axially coordinated molecules between the layers allow growth of characteristic staggered orientation of layers that gives low symmetry monoclinic P2 1 /c or C2/m in this MOF-2 group of structures preventing the formation of ideal high symmetry tetragonal space group P4/nbm formed from regularly spaced 2-D layers as in [Zn 2 (BDC) 2 [12,14,17,[26][27][28][29][30][31][32]. in fact, the origin of microporosity in 2-D Zn(BDC)(DMF)(H 2 O), MOF-2 structure discovered by yaghi et al. acted as a strong impetus to design methods to introduce similar porosity into the 3-D architecture via synthetic modification of the 2-D framework [15,33,34]. Formation of porous MOF-2 type PW structures has a strong influence on the dihedral angle (θ) between the planes of benzene ring and carboxylate group of BDC [17]. For 1-5 the dihedral angles are θ = 19.7° to 25.6°, relatively higher than MOF-2 (θ = 5.5°) (table S10). The dihedral angle (θ) between the planes of benzene ring and carboxylate group of BDC in 6 was comparable with its isostructural MOF-46 (table S10).

N 2 Adsorption isotherm
Since MOF-2 and [Cu(BDC)(DMF)]-C2/m-polymorph develop microporosity upon removal of the intercalated and coordinated solvent [14,26], we wanted to investigate the feasibility of introducing microporosity in 1-6. As a representative case, the adsorption and desorption isotherms of N 2 on 1, 4, and 6 at 77 K were collected after degassing the compounds at 120 °C for 12 h and are plotted in figure  10(a). The pattern of isotherms essentially follow type-i for 1 and 4, while for 6, there was negligible gas adsorption. Among all the 2-D polymers, 4 has the strongest N 2 affinity. The BeT surface area for 4 was calculated to be 449 m 2 g −1 with a micropore volume of 0.43 cm 3 g −1 . The observed surface area is relatively lower than that found for its disordered polymorph (C2/m) where BeT surface area was 625 m 2 g −1 [26] but higher than that found in MOF-2 [Zn(BDC)(H 2 O)] DMF which is known to have surface area of 270 m 2 g −1 and pore volume of 0.11 cm 3 g −1 for N 2 [14]. in contrast to 4, [Zn(BDC)(Py)] (1) shows moderate N 2 affinity which gives a surface area of 162 m 2 g −1 and pore volume of 0.17 cm 3 g −1 .
Micropores in [Zn(AmBDC)(DMF)] (6) were practically absent with a BeT surface area of 32 m 2 g −1 . Using HK method, the pore width of the micropores in 4 and 1 were calculated to be 0.52 and 0.48 nm ( figure  10(b)). Both MOF-2 and [Cu(BDC)(DMF)] (C2/m) [26] required removal of intercalated and coordinated solvent molecules at higher temperatures to create the porosity, while for 4, outgassing at 120 °C showed rapid condensation of N 2 gas and a high surface area. According to TGA, 4 does not show any weight loss before 180 °C, however, it may be possible that part of the DMF is removed at 120 °C under low pressure to create microporosity.

Conclusion
We have reported synthesis and crystal structure of six new 2-D coordination polymers with PW binuclear metal cluster classified as MOF-2 type solids. The role of base in directing the axial coordination to the metal center can be used to design materials with coordinatively unsaturated sites for catalytic and sensing applications. One of the 2-D materials, [Cu(BDC)(DMF)], 4, showed microporosity with accessible pores demonstrating zeolite-like behavior. The study also indicates that even if the differences are subtle in this class of materials, their impact on properties such as porosity can be significant.

Materials and physical methods
Cobalt nitrate (Co(NO 3 ) 2 ·6H 2 O), copper chloride (CuCl 2 ·2H 2 O), and zinc nitrate (Zn(NO 3 ) 2 ·6H 2 O) were purchased from Alfa Aesar. 1,4-Benzenedicarboxylic acid, 2-amino-1-4-benzenedicarboxylic acid and N-methylimidazole were purchased from Sigma Aldrich. All Chemicals and solvents were used without purification unless otherwise mentioned. Powder X-ray diffraction (PXRD) pattern was obtained from a PANalytical X'Pert Pro diffractometer equipped with a Cu Kα 1,2 anode and a linear array PiXcel detector over a 2θ range of 5 o -90 o with an average scanning rate of 0.0472 o s −1 . TGA has been performed on the sample with a TA Q50 TGA instrument with a scan rate of 10 °C min −1 under N 2 flow rate of 40 mL min −1 . FT-iR spectra were collected using a Thermo Nicolet Nexus 470 FT-iR spectrometer from 400 to 4000 cm −1 on a sample embedded in KBr pellets. N 2 gas adsorption and desorption studies were performed using a Quantachrome Autosorb-1 at 77 K.

Crystallography
Single-crystal X-ray diffraction intensity data-sets were collected on a Bruker Smart Apex diffractometer with monochromated Mo Kα radiation (0.7107 Å). Suitable crystal was selected and mounted on a glass fiber using epoxy-based glue. The data were collected employing a scan of 0.3 o in ω with an exposure time of 20 s per frame. The data-sets were collected using SMART software [35], the cell refinement and data reduction were carried out with SAiNT [31] and SADABS [36] was used for the absorption correction. The structures were solved by direct methods using SHeLX-97 and difference Fourier syntheses [37]. Full-matrix least-squares refinement against |F 2 | was carried out using the SHelXL-97/SHeLXL-2014 using the WiNGX programs suite [37]. Crystallographic data in the form of CiF files have been deposited with the Cambridge Crystallographic Data Center. These data can be obtained free of charge from the Cambridge Crystallographic Data Center via www.ccdc.cam.ac.uk/data_request/cif by quoting the following CCDC code numbers: CCDC 1444327 (1), CCDC 1444328 (2), CCDC 1444329 (3), CCDC 1444330 (4), CCDC 1444331 (5), and CCDC 1444332 (6).

Gas adsorption experiments
Before the start of the gas adsorption experiments, the as-prepared samples were outgassed at 120 °C under vacuum for 12 h. The specific surface areas of the samples were calculated using the Brunaueremmett-Teller (BeT) method. The pore size distribution curves were derived from the adsorption and desorption isotherms using the Horvath-Kawazoe (HK) and the Barrett-Joyner-Halenda (BJH) methods.

Supplementary material
Crystallographic information file (cif ) for 1-6. Tables of non-covalent interactions for 1-6, space filled model for 1, additional structural figures for 2 and 5 and table showing dihedral angles for 1-6.