Influence of Substituents in Terephthalate Linker on the Structure of MOFs Obtained from Presynthesized Heterometallic Complex

Abstract

The synthesis of new porous materials with desired properties is a challenging task. It becomes especially difficult if you need to combine several metals in one framework to obtain a heterometallic node. The use of presynthesized complexes for obtaining of new heterometallic metal–organic frameworks could be essential to solve the problem of tailored synthesis. In our study we use presynthesized heterometallic pivalate complex [Li₂Zn₂(piv)₆(py)₂] to obtain new MOFs with heterometallic core as a node of the framework. We are managed to obtain four new heterometallic MOFs: [H₂N(CH₃)₂]₂[Li₂Zn₂(bdc)₄]·CH₃CN·DMF (1), [Li₂Zn₂(H₂Br₂-bdc)(Br₂-bdc)₃]·2DMF (2), [H₂N(CH₃)₂][LiZn₂(ndc)₃]·CH₃CN (3) and [{Li₂Zn₂(dmf)(py)₂}{LiZn(dmf)₂}₂ (NO₂-bdc)₆]·5DMF (4). Moreover three of them contain starting tetranuclear core {Li₂Zn₂} and saves its geometry. We also demonstrate the influence of substituent in terephthalate ring on preservation of tetranuclear core. For compound 1 it was shown that luminescence of the framework could be quenched when nitrobenzene is included in the pores.

1. Introduction

Metal–organic frameworks (MOFs) represent a relatively new class of compounds that combine the advantages of inorganic compounds due to the presence of metal cations or clusters in the structure, together with almost unlimited possibilities of organic synthesis due to the use of organic ligands as linkers [1]. The large surface area and possibility of inner surface functionalization made MOFs a suitable material for gas storage and separation [2,3,4,5]. The use of MOFs as luminescent and electrochemical sensors, conductive materials and heterogeneous catalysts has significantly expanded the field of applications and posed new aims for the structural design of such compounds [6,7,8,9,10,11,12,13,14,15]. Synthesis of new porous materials with desired properties is challenging due to the many possible options for how inorganic nodes and organic linkers interact with each other and give the final structure [16]. The task becomes especially difficult if you need to combine several metals in one framework to obtain a heterometallic node [17,18]. The resulting materials are even more specialized for certain applications or can be used as multifunctional materials for the simultaneous execution of various tasks [19,20]. Thus using of presynthesized complexes could be essential to solve the problem of tailored synthesis. There are five main approaches for obtaining of heterometallic MOFs described in published investigations:

• postsynthetic modification of MOFs by the inclusion of metal ions in the pores of the framework [21,22,23];
• postsynthetic modification of MOFs by the partial or complete replacement of the metal cations in nodes with cations close in ionic radius [24,25,26];
• the use of salts of several metals in the initial reaction mixture to obtain heterometallic MOFs [27,28,29,30,31,32];
• the use of a metalloligand containing a metal cation other than the structure-forming metal framework [33,34,35];
• the use of presynthesized polynuclear heterometallic complexes to obtain MOFs with a given nodes [36,37,38,39,40];

The last approach looks the most promising for obtaining of new heterometallic MOFs with given geometry. In this case the synthesis of designed structure becomes rational and predictable. In our study we use previously obtained and described heterometallic pivalate complex [Li₂Zn₂(piv)₆(py)₂] to obtain new MOFs with given metal core as a node of the framework [41]. Terephthalic acid and its analogues with various substituents in the benzene ring are used as linkers (Scheme 1). We investigate the influence of the substituent on the formation of structure and changes occurring with the starting heterometallic core. Using this approaches we managed to obtain four new heterometallic MOFs: [H₂N(CH₃)₂]₂[Li₂Zn₂(bdc)₄]·CH₃CN·DMF (1), [Li₂Zn₂(H₂Br₂-bdc) (Br₂-bdc)₃]·2DMF (2), [H₂N(CH₃)₂][LiZn₂(ndc)₃]·CH₃CN (3) and [{Li₂Zn₂(dmf)(py)₂} {LiZn(dmf)₂}₂(NO₂-bdc)₆]·5DMF (4). Sorption and luminescence properties have been studied for 1 and 2.

2. Results

Synthesis and Crystal Structures

The compounds [H₂N(CH₃)₂]₂[Li₂Zn₂(bdc)₄]·CH₃CN·DMF (1), [Li₂Zn₂(H₂Br₂-bdc) (Br₂-bdc)₃]·2DMF (2), [H₂N(CH₃)₂][LiZn₂(ndc)₃]·CH₃CN (3) and [{Li₂Zn₂(dmf)(py)₂} {LiZn(dmf)₂}₂(NO₂-bdc)₆]·5DMF (4) were obtained under solvothermal conditions via the reactions of presynthesized pivalate complex [Li₂Zn₂(piv)₆(py)₂] with terephthalic acid or a substituted terephthalic acid in a solvent mixture of acetonitrile and DMF. The pivalate complex was synthesized according to previously published procedure [41]. Obtained crystals except for compound 2 were used for single crystal XRD to determine the crystal structures. In case of compound 2, crystals obtained ina reaction mixture of acetonitrile and DMA (instead of DMF) were used due to the smaller size of crystals obtained with DMF/acetonitrile mixtures.

The structure of starting pivalate complex [Li₂Zn₂(piv)₆(py)₂] presents a tetranuclear heterometallic block consisting of two Li⁺ and two Zn²⁺ cations all in a tetrahedral environment (Figure 1). Each Zn²⁺ atom is bound to the N donor of a pyridine ligand and to three O donors of three carboxylate groups, bond distances Zn–O_OCO are from 1.9297(18) to 1.9519(19) Å, Zn–N_py–2.067(2) Å. Li⁺ atoms coordinate to four O donors from four pivalates, bond distances Li–O_OCO vary from 1.889(5) to 1.985(5) Å. The center of symmetry of the tetranuclear block is at the intersection of the diagonals of the almost square [Li₂(μ₂-O)₂] moiety, and the {Li₂Zn₂} core has a planar structure. Three bridging pivalates link the Zn²⁺ and Li⁺ cations, with two carboxylates acting as bridges between two heterometallic centers and one acting as a μ₃-bidentate linker bridging two Li⁺ cations through one O atom. The six carboxylate anions are directed perpendicularly to the Zn–Zn axis of the node, while the Zn–N bonds are directed along this axis. During synthesis tetranuclear core {Li₂Zn₂} could saves it structure, fragmentizes to dinuclear {LiZn} node or decompose to individual cations (Figure 1).

According to single crystal XRD data compound [H₂N(CH₃)₂]₂ [Li₂Zn₂(bdc)₄]·CH₃CN·DMF (1) crystallizes in the monoclinic space group P2₁/n. The starting {Li₂Zn₂} core structure is conserved and becomes an 8-connected tetranuclear heterometallic [Li₂Zn₂(OOCR)₈] node linked by terephthalate moieties (Figure 2a). Thus both pivalate and pyridine moieties are replaced with terephthalate ones. Both Zn²⁺ cations are in tetrahedral environment of four O donors from carboxylates. Zn–O_OCO bond distances values vary from 1.944(2) to 1.974(2) Å which is close to Zn–O_OCO bond distances in starting complex. Li⁺ cations are also in a tetrahedral environment of four O donors, Li–O_OCO bond distances are from 1.860(6) to 1.953(6) Å which is slightly shorter than in the complex. Almost square angles O_OCO–Li–O_OCO and Li–O_OCO–Li conserve their values of 94.4(2) and 85.6(2) deg., respectively. Each [Li₂Zn₂(OOCR)₈] node is connected by eight terephthalate linkers to form a negatively charged 3D framework (Figure 2b). There are triangular pores with a aperture of approximately 3 Å in the framework. The guest accessible volume is 42% according to Mercury calculations (Figure S1d) [42]. The inner volume is filled with dimethylammonium ions and guest solvent molecules of DMF and acetonitrile. Dimethylammonium cations are formed during the synthesis in the reaction mixture due to the hydrolysis of DMF molecules. Dimethylammonium ions with partial occupancy (0.5; 0.5) fill pores of the framework and transform open pore structure into 0D cavities (Figure S1a). One of the dimethylammonium ions has a hydrogen bond with an O atom of a linker carboxylate (O…N distance is 3.007 Å). Taking into account the volume occupied with counterions the guest accessible volume ranges from 24.5% (with the first position of counterion fixed, Figure S1b) to 29.5% (with the second position of counterion fixed, Figure S1c).

According to single crystal XRD data compound [Li₂Zn₂(H₂Br₂-bdc) (Br₂-bdc)₃]·2DMA (2_DMA) crystallizes in the triclinic space group P–1. Although the unit cell parameters and space group of compound 2 differ from those of compound 1 these compounds are isoreticular. The starting {Li₂Zn₂} core also conserves its structure (Figure 3a). Zn–O_OCO bond distances values are from 1.933(4) to 1.984(4) Å, Li–O_OCO bond distances are from 1.879(13) to 1.956(12) Å which are almost the same as the bond distances in 1. Nevertheless the values of the square angles O_OCO–Li–O_OCO and Li–O_OCO–Li change to 92.0(5) and 88.0(5) deg., respectively. Each [Li₂Zn₂(OOCR)₈] node is linked by eight dibromoterephthalates, two of which are protonated. In the H₂Br₂-bdc linker the C–O_OH distance is 1.218(8) Å, which is slightly shorter than for C–O_OZn (1.270(8) Å). Moreover, the Zn–O_OCO bond distance for the H₂Br₂-bdc linker is the same as the Br₂-bdc²⁻ linker one. In this case a neutral framework is formed (Figure 3b). There are only 0D cavities (approx. 5 × 13 Å) in the structure due to the Br substituents in the terephthalate linker (Figure S2a). The guest accessible volume is 23.4% according to Mercury calculations [42]. The inner volume is filled with guest molecules of DMA. The O atom of a DMA molecule is connected with a protonated carboxylate group of the dibromo- terephthalate linkers by a hydrogen bond (O…O distance is 2.843 Å).

According to single crystal XRD data compound [H₂N(CH₃)₂][LiZn₂(ndc)₃]·CH₃CN (3) crystallizes in the monoclinic space group C2/c. In this structure the starting {Li₂Zn₂} core fragments into dinuclear {LiZn} blocks. In these blocks cations are connected by two bridging naphthalenedicarboxylic linker carboxylates. {LiZn} blocks engage individual Zn²⁺ cations from the reaction mixture and form trinuclear secondary building units {LiZn₂}. These SBUs bind with each other using pairs of bringing carboxylates to form chains along the c axis (Figure 4a). Thus all cations in the structure are in a tetrahedral environment of four O donors from the linkers. Zn–O_OCO bond distances values range from 1.922(2) to 2.001(2) Å, Li–O_OCO bond distances vary from 1.890(5) to 1.907(5) Å. Each SBU {LiZn₂} is connected with eight naphthalenedicarboxylic linkers, which are directed in pairs in four sides to build a negatively charged 3D framework with square cavities (Figure 4b). There are only 0D cavities with aperture 7 × 7 Å in the structure due to the bulky naphthalene moieties of the linker (Figure S2b). Guest accessible volume is 15.8% according to Mercury calculations and cavities are filled with dimethylammonium ions and acetonitrile guest molecules [42].

According to single crystal XRD data compound [{Li₂Zn₂(dmf)(py)₂} {LiZn(dmf)₂}₂(NO₂-bdc)₆]·5DMF (4) crystallizes in the triclinic space group P–1. There are two types of nodes in this structure: tetranuclear {Li₂Zn₂} and dinuclear {LiZn} cores. Tetranuclear cores conserve the structure of the starting complex and even coordinated pyridine as a terminal ligand (Figure 5a). Pivalate moieties are replaced with nitroterephthalate ones. Both Zn²⁺ cations are in a square-pyramid environment of three O donors from carboxylates, one O donor from DMF and one N donor from pyridine. Zn–O_OCO bond distances values range from 1.943(3) to 2.008(2) Å, Zn–O_dmf bond distance is 2.303(5) Å, Zn–N_py bond distance is 2.303(5) Å, which is slightly longer than in the starting complex. Li⁺ cations are in a tetrahedral environment of four O donors, Li–O_OCO bond distances are from 1.865(6) to 1.955(6) Å, which is slightly shorter than in the starting complex. Almost square angles O_OCO–Li–O_OCO and Li–O_OCO–Li conserve their values of 94.0(2) and 86.0(2) deg., respectively. Each [Li₂Zn₂(py)₂(OOCR)₆] node is connected by six nitroterephthalate linkers. Dinuclear {LiZn} blocks consist of Li⁺ and Zn²⁺ cations partly occupying the same positions (with a population of 0.5, Figure 5b). Both cations are in a tetrahedral environment of four O donors, three from carboxylates and one from DMF. Zn–O_OCO bond distances values range from 1.915(3) to 1.925(2) Å, Zn–O_dmf bond distance is 1.971(3) Å, Li–O_OCO bond distances range from 1.892(3) to 1.910(3) Å, Li–O_dmf bond distance is 1.967(3) Å. {LiZn} cores are connected with three nitroterephthalate linkers. Each {Li₂Zn₂} core binds to six {LiZn} cores using six nitroterephthalates, while the {LiZn} core is connected with three {Li₂Zn₂} cores (Figure 5c). Thus a layer with a biporous structure (6 × 6 Å and 8 × 4 Å) is formed. These layers have AB packing (Figure 5d) and form a 3D framework with a tortuous system of channels (Figure S2c). Guest accessible volume is 25.6% according to Mercury calculations and the cavities are filled with DMF guest molecules [42].

Compounds 1, 2 and 3 were obtained as individual products in good yields (>50%). The phase purity of the bulk samples was shown by powder XRD analysis (Figure 6). Chemical composition and structure were confirmed by elemental analysis and IR spectroscopy (Figures S3–S5). The stretching vibration bands of carboxyl groups of the organic linkers and aldehyde group of DMF are located around 1600 cm⁻¹. O–H, C–H stretching vibration bands of ligands and DMF are located from 3600 to 2900 cm⁻¹. C–H bending vibration bands of ligands and DMF are located in the 1450–1350 cm⁻¹ region.

For the safe evacuation of the guest molecules compounds were activated by soaking in acetone for two days and application of further evacuation and dynamic vacuum at 80 °C. To determine the texture characteristics and porosity of the obtained materials we recorded the N₂ and CO₂ adsorption isotherms at 77 K and 195 K, respectively. The experiments have shown very low N₂ sorption capacity for both compounds, demonstrating almost no porosity (Figure 7a). Meanwhile the PXRD patterns of the activated compounds (Figure 6) have shown the retention of the frameworks initial structure. Thus we additionally investigated the sorption of CO₂, which has a smaller kinetic diameter comparing to the N₂ molecule (Figure 7b). The BET surface area values calculated from these data are 61 and 10 m²/g for 1 and 2, respectively.

For compounds 1 and 2 solid state luminescence of as synthesized and inclusion compounds with benzene and nitrobenzene were investigated (Figure 8). To obtain inclusion compounds samples of as synthesized compounds were previously activated to remove guest solvent molecules and then immersed in the corresponding solvent (benzene or nitrobenzene). Both Zn^II and Li^I complexes are redox inactive and not known to interfere with the ligand-centered luminescence [43]. Compounds 1 and 2 have one broad band of emission with maximum at about 450 nm (λ_ex = 380 nm), which could be assigned to the intraligand π* → π transition in organic linkers. In the case of compound 1, the maximum is at 445 nm, which corresponds to the literature data for MOFs based on terephthalates [44,45]. For compound 2 emission maximum (460 nm) shifts to a longer wavelength region which corresponds to changes in the electronic structure of the ring when bromine is added as substituents. The soaking of compounds 1 in benzene and nitrobenzene results in changing of luminescence intensity and quantum yields (QY). For the as synthesized compound 1 QY is 41%. After immersing crystalline powder in benzene the QY of inclusion compounds decreases to 19.5%. This could be explained by a decrease in the rigidity of the framework due to the small amount of included benzene. This, in turn, significantly increases the probability of vibrational nonradiative transitions in the ligand and leads to a decrease in the quantum yield. Inclusion of nitrobenzene in framework 1 results in full quenching of the emission. In the case of compound 2 the QY for an as synthesized sample and an inclusion compound are almost the same: 6.8% for 2, 5.3% for benzene@2, and 4.7% for nitrobenzene@2. This may indicate that only a few molecules of benzene and nitrobenzene enter the inner volume of the framework.

Analysis of obtained samples by TGA allowed us to determine the quantity of benzene and nitrobenzene molecules located in the inner volume of frameworks 1 and 2 (Figure 9). For all samples of compound 1 the mass loss starts at about 100 °C. A second step for as synthesized 2 and its inclusion compounds starts at about 250 °C. Mass loss between these temperatures is attributed to the removal of solvent molecules from the inner volume. For as synthesized 1 the mass loss is 6.5% which corresponds to 0.9 molecules of DMF. For benzene@1 the mass loss is 5.0% which corresponds to 0.6 molecules of benzene and for nitrobenzene@1 the mass loss is 8.0% (0.6 molecules of nitrobenzene). As for compound 2 it is far more stable than compound 1. The first mass loss starts at about 250 °C, the second at 325 °C for all samples. For as synthesized 2 the mass loss is 3.5% which corresponds to 0.7 molecules of DMF. For benzene@2 and nitrobenzene@2 the mass loss is 3.5% (0.7 molecules of benzene) and 6.0% (0.8 molecules of nitrobenzene), respectively.

3. Discussion

Analysis of the obtained structures and synthesis conditions shows that terephthalic and dibromoterephthalic acids act almost the same as linkers. Despite the different unit cell parameters and space group compounds 1 and 2 are isoreticular (Figure 2b and Figure 3b). The starting tetranuclear core {Li₂Zn₂} fully retains its structure. The main difference in structures is a presence of dimethylammonium cation in the structure of compound 1.

We previously demonstrated that under a lower temperature of 100 °C it was only possible to obtain {Li₂Zn₂}-MOF [Li₂Zn₂(bdc)₃(bpy)]·3DMF·CH₃CN·H₂O based on terephthalic acid using an additional neutral ligand, 4,4′-bipyridine [41]. This fact was confirmed in the next compound [Li₂Zn₂(bpdc)₃(dabco)]·9DMF·4H₂O where diazabicyclo[2.2.2]octane (dabco) was used as a neutral ligand and biphenyldicarboxylic acid was used instead of terephthalic acid [41]. In other cases when dabco or any other coligand was not used in the reaction mixture, only dinuclear {ZnLi}-MOFs were obtained [46]. In this study we demonstrate that increasing the reaction temperature up 130 °C allowed us to obtain 3D frameworks with preserved {Li₂Zn₂} cores using only one type of ligands.

In the reaction of pivalate complex [Li₂Zn₂(piv)₆(py)₂] with naphthalenedicarboxylic acid, only structures with dinuclear {LiZn} cores were obtained. This could be explained by steric hindrance caused by the massive additional benzene ring substituent during structure formation. Structures containing {LiZn} blocks were previously obtained using naphthalenedicarboxylic, nitroterephthalic, trimesic and biphenyldicarboxylic acids when preservation of the tetranuclear core is difficult [47,48,49,50].

In the case of compound 4 one of the rarest examples of simultaneous presence of two different heterometallic cores in one structure is observed.

Investigation of the luminescence properties demonstrated that framework 1 is able to let in nitrobenzene molecules which quench the luminescence. At the same time, framework 2 contains only 0D cavities, which are difficult to penetrate. This fact results in very small changes of the QY and an almost complete absence of luminescence quenching by nitrobenzene.

4. Materials and Methods

All chemicals (terephthalic acid, Sigma-Aldrich, Germany, ≥98%), 2,5-dibromoterephthalic acid (TCI, Japan, >98.0%), 1,4-naphthalenedicarboxylic acid (TCI, >95.0%), nitroterephthalic acid (Sigma-Aldrich, Switzerland, ≥99%), N,N-dimethylformamide (“Reactiv”, Russia, >99%), N,N-dimethylacetamide (“Vecton”, Russia, 99.95%), CH₃CN (“Cryochrom”, Russia, 99.95%)) were of reagent grade and used as received without further purification. The complex [Li₂Zn₂(piv)₆(py)₂] was obtained according to a previously published procedure [41]. FTIR spectra were recorded in the range 4000–400 cm⁻¹ for the KBr-pelleted samples on a VERTEX 80 spectrometer. The powder X-ray diffraction (PXRD) data were collected with Cu-Kα radiation on a XRD 7000S powder diffractometer (Shimadzu). The elemental analyses on C, H and N were performed on a Vario Micro-Cube analyzer. The thermogravimetric analysis was carried out in He atmosphere on a TG 209 F1 thermoanalyzer NETZSCH with a heating rate of 10 deg/min. An analysis of the porous structure was performed by a nitrogen adsorption at 77 K and carbon dioxide at 195 K using Quantochrome’s AutosorbiQ system. The emission spectra of solid samples were recorded on a Cary Eclipse Varian fluorescence spectrophotometer at room temperature under λ_Ex = 380 nm.

4.1. Synthesis

4.1.1. Synthesis of [H₂N(CH₃)₂]₂[Li₂Zn₂(bdc)₄]·CH₃CN·DMF (1)

A mixture of complex [Li₂Zn₂(piv)₆(py)₂] (0.025 g, 0.027 mmol), obtained according to a previously published procedure [41], terephthalic acid (H₂bdc, 0.030 g, 0.181 mmol), N,N-dimethylformamide (DMF, 1 mL) and acetonitrile (CH₃CN, 4 mL) was heated in a closed vial at 130 °C for 2 days. The colorless block crystal was used for single crystal XRD analysis. Obtained crystals were collected and washed with DMF (3 × 1 mL) and dried at air. Yield: 0.015 g (55%). Elemental analysis calcd (%) for [H₂N(CH₃)₂]₂[Li₂Zn₂(bdc)₄]·CH₃CN·DMF, C₄₁H₄₂Li₂N₄O₁₇Zn₂: C 48.88, H 4.20, N 5.56; found: C 48.97, H 4.56, N 5.51. IR (KBr, cm⁻¹): 3425 m, 3065 w, 2663 w, 2544 w, 1965 w, 1682 s, 1597 s, 1504 s, 1387 s, 1294 s, 1136 w, 1113 w, 1018 m, 985 w, 936 m, 881 m, 825 s, 750 s, 530 s, 472 w, 415 w.

4.1.2. Synthesis of [Li₂Zn₂(H₂Br₂-bdc)(Br₂-bdc)₃]·2DMF (2)

Compound 2 was synthesized according to procedure used for 1. A mixture of complex [Li₂Zn₂(piv)₆(py)₂] (0.025 g, 0.027 mmol), 2,5-dibromoterephthalic acid (H₂Br₂-bdc, 0.057 g, 0.176 mmol), N,N-dimethylformamide (DMF, 1 mL) and acetonitrile (CH₃CN, 4 mL) was heated in a closed vial at 130 °C for 2 days. Obtained colorless crystals were collected and washed with DMF (3 × 1 mL) and dried in air. Yield: 0.034 g (78%). For further syntheses DMF was used instead of DMA to increase the yield. Chemical composition was obtained using IR spectra, TG and elemental analyses. Elemental analysis calcd (%) for [Li₂Zn₂(H₂Br₂-bdc)(Br₂-bdc)₃]·2DMF, C₃₈H₂₄Br₈Li₂N₂O₁₈Zn₂: C 28.88, H 1.53, N 1.77; found: C 28.50, H 1.86, N 1.73. IR (KBr, cm⁻¹): 3352 m, 3161 s, 3086 w, 1641 s, 1473 m, 1360 s, 1313 s, 1269 w, 1151 m, 1103 w, 1059 s, 936 w, 907 m, 893 m, 827 s, 789 s, 644 m, 555 s, 496 w, 432 m. Due to the insufficient quality of the crystals obtained for single crystal XRD analysis colorless plate crystals synthesized using DMA instead of DMF were used (2_DMA).

4.1.3. Synthesis of [H₂N(CH₃)₂][LiZn₂(ndc)₃]·CH₃CN (3)

Compound 3 was synthesized according to procedure used for 1 using 1,4-naphthalenedicarboxylic acid (H₂ndc, 0.038 g, 0.176 mmol) as a ligand. The colorless block crystal was used for single crystal XRD analysis. Obtained crystals were collected and washed with DMF (3 × 1 mL) and dried in air. Yield: 0.014 g (60%). Elemental analysis calcd (%) for [H₂N(CH₃)₂][LiZn₂(ndc)₃]·CH₃CN, C₄₀H₂₉LiN₂O₁₂Zn₂: C 55.39, H 3.37, N 3.23; found: C 55.18, H 3.32, N 3.38. IR (KBr, cm⁻¹): 3331 w, 3055 w, 2926 w, 1625 s, 1504 m, 1367 s, 1220 m, 1151 w, 1074 m, 1116 s, 878 w, 827 s, 748 s, 644 m, 530 s, 409 m.

4.1.4. Synthesis of [{Li₂Zn₂(dmf)(py)₂}{LiZn(dmf)₂}₂(NO₂-bdc)₆]·5DMF (4)

Compound 4 was synthesized according to procedure used for 1 using nitroterephthalic acid (H₂NO₂-bdc, 0.031 g, 0.149 mmol) as a ligand and at temperature of 100 °C. The yellow plate crystal was used for single crystal XRD analysis.

4.2. X-Ray Crystallography

Diffraction data for single crystals 1, 3 and 4 were obtained on an automated Agilent Xcalibur diffractometer equipped with an area AtlasS2 detector (graphite monochromator, λ(MoKα) = 0.71073 Å, ω-scans with a step of 0.5°). Integration, absorption correction, and determination of unit cell parameters were performed using the CrysAlisPro program package [51]. Diffraction data for single crystals 2_DMA were collected on the ‘Belok’ beamline (λ = 0.79313 Å, φ-scans with a step of 1.0°) of the National Research Center ‘Kurchatov Institute’ (Moscow, Russian Federation) using a Rayonix SX165 CCD detector. The data were indexed, integrated and scaled, absorption correction was applied using the XDS program package [52]. The structures were solved by dual space algorithm (SHELXT [53]) and refined by the full-matrix least squares technique (SHELXL [54]) in the anisotropic approximation (except hydrogen atoms). Positions of hydrogen atoms of organic ligands were calculated geometrically and refined in the riding model. The structures 3 and 4 contain large void volumes occupied with highly disordered solvent guest molecules (DMF and CH₃CN), which could not be refined as a set of discrete atoms. The final compositions of compounds 3 and 4 were defined according to PLATON/SQUEEZE procedure [55] (180 e− in 823 ų for 3, 206 e− in 836 ų for 4) and the data of elemental (C, H, N) analysis. The crystallographic data and details of the structure refinements are summarized in

Table 1. Crystal data and structure refinement for compounds 1–4.

Identification Code	1	2DMA	3	4
Empirical formula	C41H42Li2N4O17Zn2	C40H28Br8Li2N2O18Zn2	C40H29LiN2O12Zn2	C88H98Li4N18O46Zn4
M, g·mol−1	1007.40	1608.54	867.33	2433.08
T, K	150(2)	100(2)	150(2)	150(2)
Crystal system	Monoclinic	Triclinic	Monoclinic	Triclinic
Space group	P21/n	P–1	C2/c	P–1
a, Å	9.9681(3)	9.675(5)	15.6794(4)	13.4894(6)
b, Å	15.9931(6)	12.346(3)	13.2596(4)	14.8036(6)
c, Å	16.2939(6)	12.523(2)	19.9294(6)	17.0021(6)
α, deg	90	80.897(16)	90	64.769(4)
β, deg	92.160(3)	71.847(16)	104.551(3)	73.282(4)
γ, deg	90	70.278(16)	90	86.538(4)
V, Å3	2595.74(16)	1335.7(8)	4010.5(2)	2933.9(2)
Z	2	1	4	1
D(calcd), g·cm−3	1.289	2.000	1.436	1.377
μ, mm−1	0.990	9.131	1.260	0.899
F(000)	1036	772	1768	1252
Crystal size, mm	0.29 × 0.21 × 0.18	0.10 × 0.10 × 0.05	0.32 × 0.20 × 0.09	0.34 × 0.30 × 0.05
θ range for data collection, deg	2.43–28.51	1.9–31.0	2.68–28.74	2.37–28.35
Index ranges	−10 ≤ h ≤ 12, −20 ≤ k ≤ 16, −20 ≤ l ≤ 20	−12 ≤ h ≤ 12, −15 ≤ k ≤ 15, −16 ≤ l ≤ 16	−14 ≤ h ≤ 19, −16 ≤ k ≤ 13, −20 ≤ l ≤ 26	−17 ≤ h ≤ 18, −13 ≤ k ≤ 18, −19 ≤ l ≤ 22
Reflections collected/indep.	12,793/5802	15,935/5956	10,384/4510	26,115/12,851
Rint	0.0303	0.0585	0.0120	0.0330
Reflections with I > 2σ(I)	4603	5265	4127	9028
GOF on F2	1.093	1.063	1.087	1.067
Final R indices [I > 2σ(I)]	R1 = 0.0536, wR2 = 0.0713	R1 = 0.0624, wR2 = 0.0683	R1 = 0.0493, wR2 = 0.0530	R1 = 0.0633, wR2 = 0.0902
R indices (all data)	R1 = 0.1592, wR2 = 0.1717	R1 = 0.1663, wR2 = 0.1691	R1 = 0.1358, wR2 = 0.1383	R1 = 0.1922, wR2 = 0.2064
Largest diff. peak/hole, e·Å−3	0.916/−0.697	1.288/−1.267	1.441/−0.902	0.900/−0.726

. CCDC 2045318-2045321 contain the supplementary crystallographic data for this paper. These data can be obtained free of charge from The Cambridge Crystallographic Data Center at http://www.ccdc.cam.ac.uk/data_request/cif.

5. Conclusions

By applying a rare but efficient approach to the synthesis of heterometallic metal–organic frameworks we managed to obtain four new compounds: H₂N(CH₃)₂]₂[Li₂Zn₂(bdc)₄]·CH₃CN·DMF (1), [Li₂Zn₂(H₂Br₂-bdc)(Br₂-bdc)₃]·2DMF (2), [H₂N(CH₃)₂][LiZn₂(ndc)₃]·CH₃CN (3) and [{Li₂Zn₂(dmf)(py)₂}{LiZn(dmf)₂}₂(NO₂-bdc)₆]·5DMF (4). The approach involved the use of a presynthesized [Li₂Zn₂(piv)₆(py)₂] heterometallic complex which retains its geometry in the final structure during the reactions. This approach is one of the best options for so called rational design of desirable structures. We have shown the effect of substituents in the benzene ring of the linker on the preservation of {Li₂Zn₂} core. For frameworks 1 and 2 the luminescence properties were examined and efficient quenching of ligand luminescence was shown for inclusion compound nitrobenzene@1. The obtained results demonstrate the excellent future perspectives of using presynthesized complexes for the synthesis of new materials with tailored structures.