Construction of a Symmetrical Bi-Hydroxamate Metal–Organic Framework with Chemical Robustness

Abstract

Recently, the emerging class of hydroxamate-based metal–organic frameworks (MOFs) has demonstrated significant structural diversity and chemical robustness, both essential for potential applications. Combining the favorable hard–hard Bi-O interactions and chelating chemistry of hydroxamate groups, a rigid and symmetrical three-dimensional bismuth-hydroxamate metal–organic framework was successfully prepared via solvothermal synthesis and structurally elucidated via X-ray crystallography. The MOF, namely SUM-91 (SUM = Sichuan University Materials), features one-dimensional Bi-oxo secondary building blocks (SBUs), which are bridged by chelating 1,4-benzenedihydroxamate linkers. With the demonstrated permanent porosity and molecular sieving effect (CO₂ vs. N₂), SUM-91 was also found to be stable under harsh chemical conditions (aqueous solutions with pH = 2–12 and various organic solvents). As the structural robustness of SUM-91 could be attributed to the finetuning of the coordinative sphere of Bi centers, this work shed light on the further development of (ultra-)microporous materials with high stability and selective adsorption properties.

1. Introduction

Metal–organic frameworks (MOFs), which have been around for a few decades now, are still constantly growing in terms of structural and chemical diversity as well as ranges of application [1,2]. Featuring tailorable chemical composition, topology, and porosity, MOFs have shown promising applications in gas storage and separation, heterogeneous catalysis, and many other energy- and environment-related fields [1,3,4,5]. For all avenues of material research and development, one common prerequisite for useful MOFs is stability. Therefore, the creation or modification of MOFs with elevated degrees of stability has been one of the central research questions in the community [6,7]. One traditional rule of thumb for obtaining stable MOFs is the “hard and soft acids and bases” (HSAB) principle from coordination chemistry. Therefore, for the majority of MOF ligands that have carboxylate groups, the “go-to” choice of metal would be one of the high valent ones (e.g., Zr/Hf/Cr) [8,9]. Another routinely used strategy for enhancing MOF stability is to increase the connectivity between individual building blocks within the crystal structure. This is applicable for primary building units (PBUs, a.k.a., standalone metal atoms) as well as secondary building units (SBUs) of different dimensionalities. One extreme example would be the incorporation of one-dimensional (1D) infinite SBUs, which connects with neighboring SBU chains at a large number of sites, increasing the overall structural stability as a result [9,10,11].

Recently, inspired by naturally occurring siderophores secreted by fungi and bacteria to capture Fe^III from the environment, chelating hydroxamate groups have been successfully incorporated in MOF (or related material) chemistry, yielding a series of hydrolytically stable structures [12,13,14,15,16,17,18,19]. In all these examples, hydroxamate chelators endowed the materials with exceptional chemical robustness. In fact, the power of hydroxamate group, as a hard base itself, could go well beyond the HSAB guidelines, manifesting in stable structures with softer acids (e.g., Zn^II, Cu^I/II) [20,21,22]. Additionally, since hydroxamate-based linkers have similar bridging capabilities with carboxylate-based linkers, they are also compatible with 1D SBUs, again forming stable MOFs [23,24,25,26]. The category of hydroxamate-based MOFs has been reviewed [27], and it is clear that there is still a vast expanse of chemical diversity left underexplored, including bismuth-hydroxamate MOFs. Because of the low toxicity, abundance, and rich coordination chemistry of Bi, its incorporation in MOFs has certainly triggered research attention [28]. Concurrently, several bismuth-based hydroxamate complexes have been reported [29,30,31,32], highlighting their structural robustness under water.

Herein, we report the rational design and preparation of a Bi-hydroxamate MOF, taking advantage of the “hard-hard” interactions following the HSAB guidelines and the chelating hydroxamates. Particularly, as some conventional Bi-MOFs suffer from structural flexibility, which limits the exploration of their porosity, we propose that, when Bi centers are part of 1D SBUs in MOFs, the regularly distributed points of extension along the 1D SBU chains could be interconnected with robust chelating hydroxamate linkers. Learning from previously reported MOFs with 1D SBUs and high-valent metals, we were interested in studying the crystalline structure and properties of the product.

2. Materials and Methods

Dimethyl terephthalate (>98%), methanol (MeOH, 99.9%), N,N-dimethylformamide (DMF, >99%), sodium bicarbonate (>99%), hydroxylamine hydrochloride (>99%), sodium hydroxide (>99%), and Bi(NO₃)₃·5H₂O (>99%) were purchased from Adamas Co., Ltd. (Shanghai, China) and used without further treatment unless otherwise noted. Ultrapure water was obtained from a Millipak Express 40 system (Merk-Millipore, Darmstadt, Germany).

2.1. Synthesis of H₄-BDHA

The synthesis of ligand 1,4-benzenedihydroxamic acid (H₄-BDHA) was carried out according to the procedure reported in the literature [26]. A solution of hydroxylamine hydrochloride (12.6 g, 180 mmol) and sodium hydroxide (14.4 g, 360 mmol) in deionized water (90 mL) was added dropwise to a suspension of dimethyl terephthalate (11.6 g, 60 mmol) in methanol (100 mL). The mixture was stirred at 40 °C for 72 h, cooled to room temperature, and acidified to pH 5.5 using 5% HCl (~90 mL). The resulting white precipitate was collected by filtration, washed sequentially with deionized water (60 mL), saturated NaHCO₃ solution (3 × 30 mL), and deionized water (2 × 30 mL), and then dried under vacuum overnight to yield H₄-BDHA as a white solid. The identity and purity of the ligand were confirmed by ¹H nuclear magnetic resonance (NMR) spectroscopy, shown in Figure S1 of the Supporting Information.

2.2. Synthesis of SUM-91

In total, 12.0 mg H₄-BDHA and 7.3 mg Bi(NO₃)₃·5H₂O were ultrasonically dissolved in 1.0 mL DMF and 1.0 mL MeOH in a 4 mL glass vial that was later tightly sealed. The mixture was heated in an isothermal oven at 100 °C for 12 h. Rod-shaped single crystals were then collected and washed with fresh DMF (3 × 1 mL). Yield: 4.6 mg (62%). The crystal structure was determined via single-crystal X-ray crystallography (CCDC deposition number 2432919) and named SUM-91 (SUM = Sichuan University Materials).

2.3. Activation of SUM-91

The as-synthesized SUM-91 crystals were subjected to a rigorous activation process to remove guest molecules for subsequent characterization. Initially, the crystals were washed with dry DMF (3 × 20 mL) to remove residual solvents, followed by repeated washing with acetone (6 × 20 mL) to exchange the high-boiling-point DMF for a more volatile solvent. The crystals were then dried in an oven at 80 °C for 6 h to remove the majority of the acetone. Finally, the material was activated under high vacuum (<10 Pa) at 100 °C for 12 h to completely evacuate the pores.

3. Results and Discussion

As described in Section 2.2, the solvothermal reaction between Bi(NO₃)₃·5H₂O and 1,4-benzenedihydroxamic acid (H₄-BDHA) in a mixed solvent system of DMF and MeOH (1:1, v/v) (Figure 1a) yielded rod-shaped single crystals (Figure 1b). Single-crystal X-ray diffraction (SC-XRD) revealed that SUM-91 crystallizes in the monoclinic space group P2₁/n (CCDC deposition number 2432919, Table S1 of Supporting Information), with the asymmetric unit comprising two independent Bi³⁺ centers and two BDHA ligands (Figure S2 of Supporting Information). Overall, the formula of SUM-91 was determined to be Bi₂(H₂-BDHA)₃, indicating that it is an electroneutral framework. Moreover, there was a presence of approx. an additional 273 electrons in the SUM-91 unit cell corresponding to free solvent molecules (e.g., 7 DMF molecules) in the cavities. The phase purity of SUM-91 was confirmed by powder X-ray diffraction (PXRD), comparing the experimental and simulated patterns (Figure 1c and Figure S3 of Supporting Information), with noticeable discrepancy in relative peak intensities at 2θ = 17.3°, indicating the presence of unequal distribution of different orientations for the crystallites.

One of the Bi³⁺ positions (denoted as Bi1 in the caption of Figure 1) is coordinated by seven oxygen atoms from three chelating hydroxamate groups and one bridging hydroxamate group (in monodentate mode) from four BDHA moieties. The other Bi³⁺ position (denoted as Bi2 in the caption of Figure 1) is coordinated by six oxygen atoms from two chelating and two bridging hydroxamate groups (in monodentate mode), again from four BDHA moieties. These two Bi³⁺ positions with distinct coordination environments alternate on the 1D SBU chain (Figure 1d). The coordination mode of the BDHA linker in SUM-91, while significantly differing from other reported BDHA-linked MOF structures featuring PBUs (e.g., MUV-11 [13], SUM-1 [17], SUM-9 [18], etc.), showed a certain degree of structural similarity to a couple of other examples featuring 1D SBUs and BDHA linkers [25,26]. The 1D chains are further interconnected by BDHA moieties to form a three-dimensional framework, with narrow triangular channels along the [100] direction (Figure 1e and Figure S4 of Supporting Information). On the sides of the 1D chains, multiple N–H···O hydrogen bonding interactions (d = 1.932 Å) are present between neighboring hydroxamate groups on BDHA moieties (Figure S7 of Supporting Information). From this angle, the cross-section of SUM-91 conforms to the two-dimensional hxl (hexagonal lattice) net, as shown in Figure 1f (also illustrated in Figure S8 as a rod-packing net). The structural features of 1D SBUs with regularly spaced points of extension and chelating hydroxamate linkers enable SUM-91 to inherit the stability that has been previously observed in hydroxamate-based MOFs [20,21,22,23,24,25,26,33], as evidenced by its thermogravimetric analysis profile (Figure 1g). We note that a thermal decomposition temperature of approx. 230 °C for SUM-91 is very close to previously reported entries of the same MOF category [20,21,22,23,24,25,26,33].

Traditionally, one major challenge that MOFs constantly face is maintaining the original crystalline structures in solvents differing from the synthetic mother liquor. Therefore, PXRD was used to study SUM-91’s crystallinity after soaking in aqueous solutions with varied pH values and a range of organic solvents. Experimental results confirmed that SUM-91 could completely maintain its structural integrity across a wide range of organic solvents (Figure 2a,b) and also in acidic and basic aqueous solutions (Figure 2c,d, pH ranging from 2 to 12).

To investigate the porosity of SUM-91, a Connolly representation (probe radius = 1.65 Å, corresponding to CO₂) of the crystal structure was generated in Materials Studio 2020 first (Figure 3a). It was revealed that 1D channels run in the [100] direction of the SUM-91 unit cell, measuring approx. 3.40 Å across, closely matching a calculated pore size distribution plot (Figure 3b) produced by Zeo++ [34]. Next, N₂ adsorption experiments that were conducted at 77 K revealed only negligible N₂ uptake (<2 cm³ g⁻¹) for SUM-91 after activation under high vacuum (<10 Pa) at 100 °C for 12 h (described in Section 2.3). In contrast to several previously reported Bi-based MOFs [35,36,37,38,39,40], which exhibited negligible gas adsorption (N₂ or CO₂), the CO₂ sorption isotherms of SUM-91 exhibited a saturated uptake of 25 cm³ g⁻¹ (Figure 3c), indicating an appreciable molecular sieving effect (over N₂), even though the absolute uptake value was only moderate compared to benchmark MOF materials designed for CO₂ sorbents (Table S2 of Supporting Information). Additionally, the CO₂/N₂ adsorptive behavior of SUM-91 also differed significantly from other BDHA-based MOFs (i.e., SUM-1 [17], SUM-9 [18]). We attributed the selective adsorption to the different kinetic diameters of N₂ and CO₂ molecules (3.64 Å vs. 3.30 Å) considering SUM-91’s narrow channels (vide supra), which would significantly restrict the diffusion of N₂ molecules. Moreover, the 1D triangular channels are rich in -NH- groups, which could form N−H⋯O hydrogen bonds with the CO₂ molecules, enhancing adsorption [41]. But, the narrow channels and small cavities in SUM-91 would ultimately limit the CO₂ uptake in the materials when compared to benchmark CO₂ sorbents (vide supra). After N₂/CO₂ sorption experiments, the SUM-91 sample remained highly crystalline, as revealed by PXRD, with a diffractogram aligning well with that of an as-synthesized sample (Figure 3d).

4. Conclusions

In this study, SUM-91, a novel bismuth-hydroxamate MOF with a rigid three-dimensional structure, has been successfully constructed. Single-crystal crystallography shows that SUM-91 is formed by interconnecting 1D infinite Bi-oxo SBU chains with linear bidentate hydroxamate linkers. These structural and chemical features endowed the MOF with marked thermal stability and particularly exceptional chemical robustness (crystallinity retention in aqueous solutions with pH 2–12). Moreover, the highly symmetrical and ultra-microporous triangular channels in SUM-91, which run along the [100] direction and measure approx. 3.40 Å in diameter, exhibited a significant molecular sieving effect for CO₂ over N₂. The rational design and successful preparation of the robust Bi-hydroxamate MOF have shed light on new directions for environmentally friendly yet chemically robust porous matrices.