Yuqian Liu1, Peiran Zhao1, Chunying Duan1, Cheng He1. 1. State Key Laboratory of Fine Chemicals, Dalian University of Technology Dalian 116024 P. R. China hecheng@dlut.edu.cn.
Two- or three-dimensional porous metal–organic frameworks (MOFs) are an important kind of functional materials, which have an inherent advantage of being readily self-assembled via suitable organic linkers and inorganic metal ions/clusters.[1] Due to the unlimited coordinated bonds through the metal ions/clusters and the organic linkers, they usually appear as infinite multiple dimension networks, which endow the MOFs with diversified structures and functional applications.[2] Meanwhile they usually have high porosity, large surface area and considerable capacity,[3] which offer a wide range of potential applications in gas storage,[4] gas capture,[5] separation,[6] chemical catalysis,[7] luminescence,[8] magnetism,[9] drug delivery,[10] and so on.One special type of MOF, the lanthanide MOFs (Ln-MOFs),[11] have attracted extensive attention. Firstly, Ln-MOFs have unique luminescence properties[12] such as linear emission, high color purity, long fluorescence lifetime and so on. Secondly, the high coordination numbers and the flexible coordination geometries endow the Ln-MOFs with abundant topological structures.[13] In addition, the organic linkers can be designed and selected to synthesize MOFs, which can be used as target materials.[14] Thirdly, inside the channels of the MOFs, it contains functional sites such as open metal sites, hydrogen bonding sites and organic coordinate sites, which are highly ordered.[15] Some of the Ln-MOFs are stable and even more stable than their transition-metal-based analogues,[16] which are easy to activate the substrates without its framework collapse.[17] In addition, Ln-MOFs as attractive Lewis acid catalysts have well-proportioned Lewis acid sites within its pores/channels which afford a platform to activate the substrates and make the catalyst reaction continue.[18] Compared to the homogeneous catalysts, the heterogeneous Ln-MOFs can provide uniformly dispersed microenvironment, which is beneficial for the substrates to be close to the functional sites and easy to be separated.[19] Compared to the other heterogenous materials such as zeolites,[20] which have no organic moiety and are generally inactive for reactions in water, the Ln-MOFs can be well designed and tuned.[21]On the other hand, cyanohydrins are very important for chemistry, which could be widely used for starting materials for the synthesis of many kinds of pivotal intermediates in chemistry, such as α-amino acids, β-amino alcohols, α-hydroxy acids, α-hydroxy ketones, and so on.[22] The product of cyanosilylation of aldehydes with cyanides, such as trimethylsilyl cyanide (Me3SiCN or TMSCN), which allows the cyanohydrins to be prepared as the corresponding trimethylsilyl ether, could be readily converted into cyanohydrins.[23] As a result, cyanosilylation between aldehydes and Me3SiCN has attracted a lot of attention from the scientists in the past few years.Thermally stable MOFs (for example MIL-47 (V), MIL-53 (Al), MIL-101 (Cr), and UiO-66 (Zr)) have been used for catalyzing the cyanosilylation based on their Lewis acid sites.[24] In this paper, the luminescent lanthanide ion Tb(iii) was selected to assemble with the rigid ligand H2sbdc with bi-carboxyl acid group to construct a stable 3D porous MOF, and the Tb-MOF with coordinative unsaturated Tb centers are able to sever as an efficient heterogeneous catalyst for the cyanosilylation of a series of aldehyde substrates.[25]
Experimental section
Materials and instrumentation
All the chemicals were purchased from commercial sources and used without any further purification. 1H-NMR spectra were recorded with BRUKER AVANCE III 500 NMR instrument (CDCl3 solution) at 500 MHz. The single-crystal structure was tested on a D8 Venture SMART-CCD instrument with Mo-Kα radiation designed by Bruker from Germany. X-ray powder diffraction (XRD) measurements were performed with Cu-Kα radiation source (λ = 0.15406 nm) at 40 kV voltage and 25 mA current. IR spectra were performed on an IS50 FT-IR instrument. Thermogravimetric analysis was finished on a SDTQ600 thermogravimetric analysis instrument under N2 condition. The content of C, H and N was analysed by vario EL III elemental analysis instrument designed by Elementar Company from Germany.
Synthesis of Tb-MOF
0.3 mmol (0.091 g) H2sbdc and 0.2 mmol (0.091 g) Tb(NO3)3·6H2O were mixed in a 9 mL bottle with 5 mL EtOH and 1 mL H2O stirring at room temperature for 3 h. Then sealed and placed in a pre-heated oven, and kept it at 140 °C for 4 days. The product was cooled to room temperature in 2 days. The crystals were formed under static conditions for 10 days. The resulting solid was filtered, washed with methanol and diethyl ether. Finally, the colorless block crystals were obtained and dried in vacuum at room temperature for one day (elemental analysis C: 37.31%; H: 2.679%).
The method of catalysis by Tb-MOF
The typical experiment was performed as the following condition: substrate (0.5 mmol), trimethylsilyl cyanide (1.2 mmol), catalyst (12.79 mg, 4% mol of the substrate, based on the metal ions), and DCM (2 mL) in a 6 mL glass bottle in Ar environment at room temperature under continuously stirring. The corresponding yields of targeted products were successfully monitored by applying the NMR approach, which was added 0.5 mmol 1,3,5-trimethoxybenzene as an internal standard substance.
Crystallography
X-ray intensities of the complexes were collected on a Bruker D8 Venture CCD diffractometer with graphite-monochromated Mo-Kα (λ = 0.71073 Å) using the SMART and SAINT programs. The structures were solved by direct methods and refined on F2 by full-matrix least-squares methods with SHELXTL-2018.
Crystal data for Tb-MOF
[Tb2(sbdc)3(H2O)3], M = 1278.63, triclinic, space group P1̄, colorless block, a = 13.907 (1) Å, b = 14.243 (1) Å, c = 14.994 (1) Å, α = 118.151 (1), β = 99.052 (2), γ = 103.126 (2), V = 2426.3 (3) Å3, Z = 2, Dc = 1.750 g cm−3, μ(Mo-Kα) = 0.71073 Å, T = 296 (2) K. 52 159 unique reflections [Rint = 0.0603], final R1 [with I > 2σ(I)] = 0.0337, wR2 (all data) = 0.0928, GOOF = 1.079. CCDC number: 2100650.†In the structural refinement of Tb-MOF, except the partly occupied parts, the other non-hydrogen atoms were refined anisotropically and hydrogen atoms were fixed geometrically at calculated distances and allowed to ride on the parent non-hydrogen atoms. The highly disordered solvent molecules could not be located, and hence in the final refinement, the electron density was treated with the SQUEEZE routine in the PLATON program package.
Results and discussion
Three-dimensional (3D) Tb-MOF was prepared using a rigid ligand 5,5-dioxo-5H-dibenzo[b,d]thiophene-3,7-dicarboxylic acid (namely H2sbdc) and Tb(NO3)3·6H2O by using a solvothermal method to synthesize a novel material.[26] Tb-MOF can be considered as a highly efficient heterogeneous catalyst which was anticipated to catalyze the cyanosilylation of aldehyde based on its Lewis acid sites and porous structure.The single crystal X-ray diffraction data showed that the obtained Tb-MOF crystallizes in P1̄ space group (CCDC 2100650†). The asymmetric unit of Tb-MOF contained two Tb(iii) ions, three H2sbdc organic ligands, and three coordinated H2O molecules. As seen in Fig. 1a, Tb(1) and Tb(2) had two different coordination modes, of which coordinated numbers were both eight. Tb(1) was coordinated with eight oxygen atoms (Tb–O = 2.27–2.52 Å) from seven di-monodentate carboxyl groups in different ligands, and one water molecule. Tb(2) was coordinated with eight oxygen atoms (Tb–O = 2.26–2.53 Å) from four di-monodentate carboxyl groups in different ligands, one bi-dentate carboxyl group, and two water molecules, which was similar to some other works of Tb-MOFs reported before.[27] Tb(1) and Tb(2) were linked through two di-monodentate carboxyl groups and on μ2-oxygen atom of the didentate carboxy group. The two adjacent Tb(1) centers were bridged by four di-monodentate carboxyl groups, forming a paddle wheel dinuclear Tb(2) core by which the two Tb(1)–Tb(2) units were linked to form a zigzag tetra-nuclear unit. And the tetra-nuclear unit was further extended by the bi-carboxyl ligands, resulting in a 3D framework having channels along the a and b direction with the area of the pore was about 7.5 × 5.0 Å2 and 9.0 × 5.5 Å2 respectively, which were available for guest accommodation and exchange. The metal centers having removable water molecules were well-positioned in the channels. The results obviously demonstrated that Tb-MOF not only had lots of open metal Tb(iii) sites but also contained lots of cavities as Lewis acids sites simultaneously, resulting in its potential applications in the catalytic field. The PXRD profile of the synthesized solid sample (Fig. 2) was carried out at ambient temperature and 200 °C, respectively. The characteristic diffraction lines of prepared Tb-MOF and Tb-MOF heated to 200 °C were similar with the simulated from the single crystal data of Tb-MOF, revealing the purity and stability of the product Tb-MOF.
Fig. 1
(a) The asymmetric unit of Tb-MOF; (b) Tb4 units in Tb-MOF; (c) view of the 3D structure along a direction; (d) view of the 3D structure along b direction. Hydrogen atoms were omitted for clarity.
Fig. 2
The PXRD spectra. (black bar) Simulated Tb-MOF; (red bar) Tb-MOF; (blue bar) Tb-MOF Heated to 200 °C; (green bar) cycle one for catalysis of m-methoxybenzaldehyde; (purple bar) cycle two for catalysis of m-methoxybenzaldehyde; (yellow bar) cycle three for catalysis of m-methoxybenzaldehyde; (light blue bar) cycle four for catalysis of m-methoxybenzaldehyde.
Thermogravimetric analysis (TGA) demonstrated that in the range of 25–800 °C which increased 10 °C per minute in N2 atmosphere. As obviously displayed in Fig. 3, the TGA result of the fresh synthesized Tb-MOF demonstrated a slow weight loss of 8% before 200 °C, corresponding to the loss of water and EtOH molecules. Tb-MOF became metal oxide and its structure collapsed when the temperature raised to 500 °C. It indicated that the framework of Tb-MOF was stable below 500 °C.
Fig. 3
The TGA spectra of Tb-MOF.
After treating 10 mg Tb-MOF with 2′,7′-dichlorofluorescein in DCM for 24 h, the product was filtered and washed with DCM. Then HCl was used to destroy the structure of Tb-MOF so that the quantity of 2′,7′-dichlorofluorescein in the channels of Tb-MOF could be tested by UV-vis adsorption. The result in Fig. 4 showed that Tb-MOF can adsorb 6.3% 2′,7′-dichlorofluorescein of its weight, which could prove that Tb-MOF had enough pores and a good ability to adsorb small molecules.[28]
Fig. 4
(a) The UV-vis absorption spectra of released 2′,7′-dichlorofluorescein from Tb-MOF; (b) the concentration of 2′,7′-dichlorofluorescein was determined by comparing the UV-vis absorption with a standard curve.
Since the addition of cyanide to a carbonyl compound to form a cyanohydrin was one of the fundamental C–C bond-forming reactions in organic chemistry and had frequently been at the forefront of advances in chemical transformations,[29] the catalytic ability of Tb-MOF for the cyanosilylation of the aromatic aldehyde was further studied in the detail due to the uniformly dispersed Lewis acid sites and the size of its pores in the whole structure.[30]Therefore, p-methoxybenzaldehyde with TMSCN catalyzed by Tb-MOF was studied in different solutions. As we could see in Table 1, Tb-MOF had no catalytic ability in EtOH for this reaction, even the mixture was stirred for 12 h. And the yield of the reaction in MeOH was much lower than those in DCM, THF, and CH3CN. Considering the stability of the MOF catalyst and the separation of the product, DCM was chosen as the solvent of the reaction for further study. Consequently, the reaction condition was optimized as the addition of 2.4 equivalent of TMSCN (Fig. S63†) and 12 h (Fig. S64†) under Ar at room temperature.
The yields in different solutions (0.5 mmol p-methoxybenzaldehyde, 1.2 mmol TMSCN, 4 mol% Tb-MOF, in 2 mL solution, under Ar condition)
Entry
Solvent
Conv. (%)
1
MeCN
85
2
THE
82
3
DCM
78
4
MeOH
14
5
EtOH
0
Therefore, different aromatic aldehydes, which had different electronic effect, spatial effect, and fatty aldehyde were selected. Table 2 showed the results based on 12 different aldehyde substrates and trimethylsilyl cyanide in the presence of Tb-MOF. Most of them had a higher yield. Even, the yield of 1-naphthalene formaldehyde reached 77%. Table 3 showed the catalytic cyanosilylation of o-, m- and p-methoxybenzaldehyde in different conditions. The final yield of 3-methoxybenzaldehyde and 4-methoxybenzaldehyde were 85% and 81% after reacting 12 h. The reaction was monitored after reacting 4 h. The yield of o-, m-, and p-methoxybenzaldehyde was 28%, 30%, and 41%, respectively. The corresponding initial TOF was about 1.8 h−1, 1.9 h−1, and 2.6 h−1, respectively. The catalyst Tb-MOF was filtered after reacting 4 h and let the reaction continue. The yield of 4-methoxybenzaldehyde increased a little to 47% after reacting 12 h, of which the catalyst Tb-MOF was removed after reacting 4 h. The yield of this reaction was 0 without any catalyst for 12 h. The yield of the reaction catalyzed by the ligand without Tb-MOF was 0 either. The yield had no significant change when the catalyst was doubled. These consequences demonstrated that the ligand couldn't catalyze this reaction and Tb-MOF was necessary for this reaction. The Lewis acid sites in the channel of Tb-MOF could activate the substrates, which could make the reaction continue. Furthermore, as the yields of o-, m- and p-methoxybenzaldehyde after 12 h are 82%, 79%, 78%, catalyst Tb-MOF had no prominent electronic effect on this reaction. However, the yield of p-benzyloxybenzaldehyde and 4-(diethylamino)salicylaldehyde were 66% and 57%, which were lower than other substrates. It seemed that the substrates which had larger sizes didn't fit the pores of Tb-MOF well, indicating that the activation of substrates should be occurred in the pores of Tb-MOF. Table 2 also showed that all the TON of the 12 product were about 20, which were not at a high level. Consequently, 5 mmol p-methoxybenzaldehyde, 12 mmol TMSCN, and 0.2% mmol Tb-MOF with no solvent were stirred at room temperature under Ar condition. Notably, the yield of solvent free condition was about 92%, and the TON could reach about 460 with the MOF catalyst being 0.2 mol%, which demonstrated that Tb-MOF had a good ability of catalyzing cyanosilylation. The results above demonstrated that the Tb-MOF had a good ability for catalyzing most kinds of aldehydes for higher yields as a heterogeneous catalyst.
Catalytic cyanosilylation of different substrates in the presence of Tb-MOF. Reaction conditions: (CH3)3SiCN: 1.2 mmol, aldehyde: 0.5 mmol, Tb-MOF catalysts (4 mol%), room temperature under Ar for 12 h
Entry
Substrate
Product
Conversion (%)
TON
1
78
20
2
85
21
3
81
20
4
77
19
5
82
21
6
79
20
7
78
20
8
74
19
9
85
21
10
83
21
11
57
14
12
66
17
13a
92
460
5 mmol p-methoxybenzaldehyde, 12 mmo1 TMSCN, and 0 2% mmol Tb-MOF with no solvent were stirred at room temperature under Ar condition.
Catalytic cyanosilylation o-, m- and p-methoxybenzaldehyde in different conditions
Substrate
12 h
4 h
4–12 h
Double catalyst
Without catalyst
Ligand
82
28
46
81
0
0
79
30
43
83
0
0
78
41
47
80
0
0
5 mmol p-methoxybenzaldehyde, 12 mmo1 TMSCN, and 0 2% mmol Tb-MOF with no solvent were stirred at room temperature under Ar condition.To test the recyclability of the catalyst Tb-MOF, four cycles of cyanosilylation reaction in DCM were performed. Tb-MOF was separated by filtration from the reaction mixture after 12 hours, washed with DCM, dried under vacuum overnight, and reused in the next reaction run. The second, the third, and the forth cycles were completed under the same condition as the first cycle. From the PXRD measurements of the recovered catalyst in Fig. 2, Tb-MOF retained its structural framework even after three cycling reaction process. However, the decrease of the PXRD could be found in the forth cycle. As shown in Fig. 5, reactions catalyzed by Tb-MOF showed a slight decrease after three cycles, which demonstrated that the heterogeneous Tb-MOF had a better advantage than the homogeneous catalyst. In the forth cycle, the yields decreased more than the first three cycles, which fitted the PXRD results. TON of o-, m-, and p-methoxybenzaldehyde for all the four cycles were 80, 74, and 69, respectively. As a result, Tb-MOF had a good ability for recyclable catalyzing cyanosilylation for at least three times without the yields' going down.
Fig. 5
The yields for three cycles (0.5 mmol p-methoxybenzaldehyde, 1.2 mmol TMSCN, 0.01 mmol Tb-MOF, in 2 mL DCM at room temperature under Ar condition).
To test whether Tb-MOF could adsorb the substrate, 0.01 mmol Tb-MOF, 3 mmol substrate with 2 mL DCM in a 6 mL glass bottle were under continuously stirring for 24 h. Then the solid was filtered, washed by DCM and dried in vacuum. Fig. 6 was the IR spectrogram, which showed the differences between Tb-MOF and Tb-MOF with the substrate inside. The peak of the CO bond of 2-naphthaldehyde was 1701 cm−1, while the peak moved to 1710 after 2-naphthaldehyde was treated with Tb-MOF. It demonstrated that the substrate could be absorbed by the pores of Tb-MOF and had interaction with Tb-MOF.
Fig. 6
The IR spectra of Tb-MOF with different p-nitrobenzaldehyde.
To test the ability for activating the substrates of Tb-MOF, 2 mg Tb-MOF with 2-naphthaldehyde from 0 to 0.001 M was treated. A significant quenching of the fluorescence intensity could be seen in Fig. 7. It was important to note that the sorption of 2-naphthaldehyde caused the luminescence quenching. The value of I0/I and the concentration exhibited a linear relationship at a lower concentration range (0–0.5 mM), which fitted the Stern–Volmer (SV) equation, I0/I = Ksv[M] + 1 (where I0 represented the initial intensity of the peak at 615 nm, I represented the intensity at different concentrations of 2-naphthaldehyde suspension, [M] represented the concentration of 2-naphthaldehyde, and Ksv represented the quenching constant (M−1)), with a correlation coefficient (R2) of 0.98.[31] The KSV value for 2-naphthaldehyde was counted to be 1.26 × 103 M−1.[32] The high KSV allowed us to easily identify the existence of a small amount of aldehyde in solution. As previously mentioned, the relationship between I0/I and the concentration of substrate could be well fitted into the SV equation at lower concentrations, suggesting the interaction with quenching substrates, forming complexes and influencing the luminescence spectra of luminescent substances.[33] The fluorescence titration proved that the catalysis of the reaction was in the channel of Tb-MOF, which could also recognize the substrates. The Lewis-acid sites played a dormant role in the cyanosilylation. In this case, the luminescence behavior of Tb-MOF suggested the sorption of the substrate into the MOFs and the possible interaction between the aldehyde substrate and the Tb centers. Based on the experimental results above and the works reported before, a possible reaction mechanism was presented to explain the reaction process of cyanosilylation catalyzed by Tb-MOF: the substrates were activated by the coordinatively unsaturated Tb centres to react with TMSCN, which was shown in Fig. S65.† The products were removed by other substrates, meanwhile, Tb-MOF continued to activate the substrates in the next catalytic turn.[34]
Fig. 7
(a)Luminescent spectra of Tb-MOF emulsion in DCM upon addition of substrates up to 0.001 M (excited at 335 nm). (b) The plot of I0/I vs. the lower concentration (0–0.5 mM) of 2-naphthaldehyde which showed intensities at 615 nm as a function of substrates (excited at 335 nm).
Conclusions
In conclusion, a novel Tb(iii)-based metal–organic framework was successfully synthesized, which could be considered as an efficient heterogeneous catalyst for the cyanosilylation of different aldehydes at room temperature because of its pores and the activation ability for the substrates of its Lewis acid sites. Meanwhile, Tb-MOF exhibited a high TON (TON = 460) in solvent free condition and a good recyclability for these reactions in solvent condition at least three times. Compared to other transition metal organic frameworks,[24] which were designed for catalyzing cyanosilylation in solvent free condition, Tb-MOF had a higher TON. As a result, Tb-MOF was considered as a good heterogeneous catalyst for cyanosilylation and had a good prospect for further study. We hope that this work can offer a useful and meaningful method to use such ligands to synthesize MOFs for sensing or absorbing organic molecules towards environmental and catalyzing different reactions.
Authors: Sergiy I Vasylevs'kyy; Ganna A Senchyk; Andrey B Lysenko; Eduard B Rusanov; Alexander N Chernega; Julia Jezierska; Harald Krautscheid; Konstantin V Domasevitch; Andrew Ozarowski Journal: Inorg Chem Date: 2014-03-18 Impact factor: 5.165
Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7