Amarajothi
Dhakshinamoorthy
*ab,
Abdullah M.
Asiri
c,
Mercedes
Alvaro
b and
Hermenegildo
Garcia
*bc
aSchool of Chemistry, Madurai Kamaraj University, Madurai-625 021, Tamil Nadu, India. E-mail: admguru@gmail.com
bDepartamento de Quimica and Instituto Universitario de Tecnologia Quimica (CSIC-UPV), Universitat Politecnica de Valencia, Av. De los Naranjos s/n, 46022 Valencia, Spain. E-mail: hgarcia@qim.upv.es
cCenter of Excellence in Materials Research, King Abdulaziz University, Jeddah, Saudi Arabia
First published on 9th October 2017
Metal organic frameworks (MOFs) are being intensively studied as solid catalysts for organic reactions in liquid media. This review focuses on those reports in which these materials have been used as catalysts in the absence of solvents or embedding ionic liquids (ILs). One of the major roles of solvents in liquid phase reactions is to desorb reagents and products from the active sites, facilitating the turnover of the active sites. For this reason, it is a general observation that most solid catalysts undergo strong deactivation and poisoning in the absence of solvents. In the present review, examples are presented showing that, due to their large porosity and framework flexibility, MOFs can be used as catalysts in the absence of solvents for several reaction types including cyanosilylations, condensations, cyloadditions and CO2 insertions, among others, and that they show even better performance than in the presence of conventional organic solvents. This review also describes the synergy that arises from the combination of ILs, frequently with suitable task-specific chains, and MOFs due to the cooperation with the catalysis when two centres in MOF and ionic liquid are present and due to the change in the microenvironment of the active sites. By embedding an ionic liquid in the MOF pores or by synthesising the ionic liquid covalently attached to the ligand in satellite positions, reusable and efficient catalysts requiring the minimum amount of ionic liquid can be obtained. Both complementary strategies increase the greenness of MOFs as heterogeneous catalysts and have advantages from the environmental point of view. Finally, the last section describes the catalytic activity of hierarchical porous MOFs in some selected reactions.
MOFs are porous crystalline materials whose lattice is constituted by metallic nodes coordinated with rigid multipodal organic linkers, mostly aromatic polycarboxylates.10–12 The catalytic activity of these materials derives from the presence of coordinatively unsaturated sites (CUS) around the metal ions at the nodes and structural defects, from the presence of acidic or basic substituents at the ligands or from the presence of guest species located within the internal voids of the structure.13–15 The extensive use of MOFs as catalysts is understandable considering that there is a wide choice in the nature of the metal that can form a part of the MOF and the possibility to adapt the properties of the MOF after the synthesis. Other attractive features of MOFs as catalysts are their large surface area and high pore volume and the flexibility in their synthesis. In this way, MOFs have become one of the preferred solid catalysts for liquid phase reactions including oxidations,16,17 condensations,18 and C–C couplings,9 among others.19 The use of MOFs as solid catalysts can contribute to the greenness of the process, but other important factors to be considered regardless of the catalyst are the nature of the reagents and solvents employed in the reaction. Particularly, volatile organic solvents are one of the main sources of undesirable environmental impact, especially those containing halogens which have been shown to have adverse effects on the ozone layer in the atmosphere as well as on human health. For this reason, research is being carried out in one of the lines in green chemistry aimed at finding suitable alternatives to conventional low boiling point organic solvents that can be emitted into the atmosphere or can become a waste product after the reaction.
One of the simplest approaches to avoid the use of volatile organic solvents is to perform the reaction in the absence of any solvent, just by using substrates/reagents that are in liquid form under the reaction conditions. While solvent-free conditions are favourable from the point of view of avoiding the emission of vapours into the atmosphere, the absence of solvents generates considerable stress on the catalyst, particularly when using solid porous materials. This is because one of the main roles of solvents is to assist in the desorption of reagents, products or byproducts from active sites by competing with them. In other words, in the absence of solvents, it is a general observation that solid catalysts become easily deactivated by strong adsorption onto the active sites of chemicals, including reagents, products and byproducts.
The use of solvent-free conditions on a solid catalyst is, as previously commented, very convenient from the green chemistry point of view; however, it is not very practical due to the rapid deactivation of the catalyst. In the present review, examples will be shown which illustrate that there are several general reaction types that have been proved to be successful under solvent-free conditions in the presence of MOFs. The present state of the art has shown that silylations, condensations, including aldolic reactions and formation of heterocycles, as well as selective hydrogenations and CO2 insertions can be carried out by MOFs in the absence of organic solvents. Although solventfree conditions are not exclusive to MOFs and other classes of catalysts have also been used in the absence of solvents, the purpose of this review is to show that some general reaction types can be carried out advantageously using MOFs as catalysts without solvents.
Besides solvent-free conditions, another alternative to the use of volatile organic compounds that has also become a research front in catalysis is the use of ILs.20–22 They have a negligible vapour pressure and it has been shown that they can dissolve a large variety of organic compounds and can also be a convenient liquid media for organic reactions. One of the main problems of ILs is their viscosity that considerably complicates the extraction of products and chemicals after the reaction as well as the recovery of the catalysts. In this context, one of the strategies that has been developed to take advantage of the benefits of the use of ILs as the reaction medium as well as to diminish as much as possible the volume of ILs and simplify the reaction work-up has been to deposit these ILs as a thin film on the surface of large surface area materials.23 The presence of porosity in MOFs makes it possible to incorporate ILs inside these solid catalysts and in this way, the advantages of heterogeneous catalysis are combined with the benefits of ILs in terms of cooperation with the reaction mechanism, while still being recoverable together with the solid after the reaction. For these reasons, this review also covers those reports combining MOFs and ILs in minimal amounts. Particularly relevant are those examples in which task-specific ILs, i.e. those ILs that have been designed having functional groups to promote organic reactions while maintaining their solubility properties, will also be commented on. As will be shown, the combination of ILs that have Brønsted acidity/basicity with the active sites of MOFs can serve to preserve the catalytic activity of MOFs in Knoevenagel condensations and CO2 insertions and also to avoid the particle size increase that is a common deactivation mechanism observed when supported metal nanoparticles (NPs) are used as catalysts. In addition, another possibility to increase the catalytic activity of MOFs is the use of hierarchical structures in which the diffusion of reagents and substrates is facilitated by the presence of meso- and micropores.
The aim of the present review is to show the current state of the art of the use of MOFs as catalysts in the absence of solvents, incorporating ILs within the pores in such a way that they form a system with the MOF or exhibiting a hierarchical structure with the presence of meso- and micropores. While there is abundant literature reporting on the use of MOFs as catalysts in the presence of solvents, this review aims at demonstrating that MOFs can also perform with high activity and selectivity in the absence of solvents without compromising on their stability, thus fulfilling one of the principles of “Green Chemistry”. In addition, the present review also shows the utility of ILs attached to MOFs to provide a microenvironment to the active sites or to provide additional sites to enhance the activity of these materials compared to the activity of the parent MOFs. A final aspect included in the present review is hierarchical structuration to facilitate the accessibility of substrates and reagents to the active sites.
Table 1 lists the reactions and the MOF catalysts that will be included in the present review. Emphasis is placed on the evidence provided on the stability of the MOF structure under different reaction conditions.
Catalyst | Reaction | Stability evidences | Ref. |
---|---|---|---|
InPF: indium polymeric framework; hfipbb: hexafluoroisopropylidene bisbenzoate; bipy: bipyridine; BTT: 1,3,5-benzene tristetrazolate; RPF: rare earth polymer framework; DSB: 3,5-disulfobenzoate; Phen: 1,10-phenathroline; tipp = 5,10,15,20-tetrakis(4-(imidazol-1-yl)phenyl)-porphyrin; bpdc: biphenyl-4,4′-dicarboxylic acid; DMA: N,N′-dimethylacetamide; H2LOMe: 5,5′-(2,3,6,7-tetramethoxyanthracene-9,10-diyl)diisophthalic acid; NMP: N-methylpyrrolidine; 4,4′-bpe = 1,2-bis(4-pyridyl)ethylene; m-bds = 1,3-benzenedisulfonic acid; PTA: phosphotungstic acid; TPA: tetrapropylammonium; BDC: 1,4-benzenedicarboxylate; DMA: dimethylamine; BTC: 1,3,5-benzenetricarboxylate; ptz: 2,4,6-tri-(4-pyridyl)-1,3,5-triazine; nsa: sodium 2,6-naphthalene disulfonate; BTB: 1,3,5-tris(4-carboxyphenyl)benzene; NDC: naphthalenedicarboxylic acid; ABIL: amino-functionalized basic ionic liquid; DAIL: dual amino-functionalized ionic liquid. | |||
Vanadium–salen based Cd-MOF | Cyanosilylation of benzaldehyde | XRD, UV-Vis | 24 |
In-PF-6 | -do- | XRD, reuse | 25 |
[In4(OH)4(hfipbb)4(4,4′-bipy)] | -do- | Reuse, hot-filtration | 26 |
[(Cu4O0.27Cl0.73)3(H0.5BTT)8] | -do- | XRD, reuse | 27 |
[Nd(3,5-DSB)(Phen)] (RPF-19) | -do- | XRD, reuse | 28 |
[Cd3(tipp)(bpdc)2]·DMA·9H2O | -do- | Leaching, reuse | 29 |
Ba2(H2LOMe)2(NMP)·NMP | -do- | Reuse | 30 |
M(4,4′-bpe)2(H2O)4(m-bds) (M: Zn2+ or Cd2+) | Synthesis of 2-amino-6-(arylthio)pyridine-3,5-dicarbonitriles | Reuse | 31 |
IRMOF-3 | Synthesis of dihydropyrimidinones and dihydropyridines | — | 32 |
IRMOF-3 | Synthesis of polyhydroquinoline | — | 33 |
PTA@MIL-101 | Synthesis of 3,4-dihydropyrimidin-2(1H)-one | Hot-filtration, reuse, FT-IR, FESEM | 34 |
[TPA]2[Zn2-(BDC)3(DMA)2] | Michael addition of pyrrole to β-nitrostyrenes | XRD, reuse | 35 |
Cu3(BTC)2 | Michael addition between β-nitrostyrene and indole | XRD, FT-IR and reuse | 36 |
[TPA]2[Zn2-(BDC)3(DMA)2] | Knoevenagel condensation reaction between benzaldehyde and malononitrile | XRD, reuse | 37 |
Cu(ptz)(nsa)0.5·H2O | Azide–alkyne cycloaddition | XRD, reuse | 38 |
Cu(I)-MOF | Synthesis of propargylamine | XRD, reuse, leaching | 39 |
MIL-101(Cr) | Coupling of CO2 with styrene oxide | Reuse, XRD | 40 |
F-IRMOF-3 | Coupling of CO2 and propylene oxide | XRD | 41 |
[Zn4O(2,6-NDC)(BTB)4/3] | -do- | XRD, reuse | 42 |
ABIL-OH/HKUST-1 | Knoevenagel condensation between benzaldehyde and malononitrile | XRD, hot-filtration | 43 |
PdCl2-ILs/Cu3(BTC)2 | Oxidation of cyclohexene | Reuse, XPS, XRD | 44 |
IL@UiO-66 | Oxidative desulfurization | Reuse | 45 |
PW/DAIL/MIL-101(Cr) | Oxidation of benzyl alcohol | XRD, reuse, leaching | 46 |
PW4/DAIL/MIL-100(Fe) | Oxidation of benzyl alcohol | XRD, reuse, leaching | 47 |
DAIL-Fe3O4@NH2-MIL-88B(Fe) | Esterification of oleic acid with ethanol | Reuse | 48 |
[SO3H-(CH2)3-HIM]3PW12O40@MIL-100(Fe) | -do- | Reuse | 49 |
[NMP]+CH3SO3−@MIL-101(Cr) | Esterification of acetic acid with amyl alcohol | Reuse, leaching, TEM | 50 |
TEDA-BAIL/MIL-101 | Acetalization of benzaldehyde with glycerol | Reuse, leaching | 51 |
Pd/IL/Cu3(BTC)2 | Hydrogenation of acetylene | Reuse, XRD, TEM | 52 |
Cu-MOF | Hydrocarboxylation of cyclohexane | XRD, leaching, reuse | 53 |
CuI@UiO-67-IM | Azide–alkyne cycloaddition | Reuse, leaching, XPS | 54 |
MIL-101(Cr)-X(n-Bu)3Br (X: N or P) | Coupling of CO2 and propylene oxide | ICP, FT-IR, XRD | 55 |
UiO-67-IL | Coupling of CO2 to epichlorohydrin | XRD, leaching | 56 |
IL-ZIF-90 | Coupling of CO2 with propylene oxide | Reuse, XRD, leaching | 57 |
Recently, the synthesis of a chiral vanadium–salen based Cd-MOF with a chiral ligand (R,R)-(−)-1,2-cyclohexanediamino-N,N′-bis(3-tert-butyl-5-(4-pyridyl)salicylidene) has been reported under solvothermal conditions and its catalytic activity has been examined in the solvent-free cyanosilylation of benzaldehyde (Scheme 1).24 Chiral VO–salen complexes were used as homogeneous asymmetric catalysts for various types of organic reactions.64,65 Under optimized conditions, the reaction of benzaldehyde with TMSCN catalyzed by the chiral VO–salen attached to Cd-MOF gave 95% yield with 78% ee. In contrast, the same reaction afforded 91% yield with 57% ee when VO–salen was used as the homogeneous catalyst. These data clearly indicate the accessibility of substrates to the chiral centers located inside the pores of the Cd-MOF. On the other hand, the observed activity with this Cd-MOF under solvent-free conditions is comparable with that of other related chiral MOFs in the presence of solvents.66,67 Furthermore, control experiments performed in the presence of chloroform and acetonitrile detected the occurrence of leaching, thus showing the convenience of carrying out organic reactions under solvent-free conditions, although the exact reasons why leaching does not occur in the absence of solvents remain unknown. The activity of this catalyst was maintained in two runs, although a decrease in the yield and ee was noticed in the third run. This deterioration in the catalytic activity was believed to be due to the modification of the surrounding environment of chiral centers or/and the undesirable adsorption of organic compounds. However, in spite of the decrease in activity, the catalyst maintained its crystallinity after the third run and the vanadium oxidation state did not show remarkable changes.
The use of a new flexible ligand in the synthesis of two In-MOFs with ancillary coordination resulted in one one-dimensional and two isoreticular 2D structures, their catalytic activity being evaluated in the cyanosilylation of carbonyl compounds.25 It was found that those In-MOFs with donor ancillary ligands in their structures show the best activity, the one having 1,10-phenanthroline being the most active. The turnover frequency (TOF) value reached in the cyanosilylation of benzaldehyde with the most active In-MOF was 2149 h−1 with 99% yield at 50 °C under solvent-free conditions. Furthermore, the cyanosilylation of acetophenone at 80 °C with the best-performing In-MOF resulted in 39% yield with a TOF value of 19 h−1. The recyclability data indicated that the In-MOF could be reused at least four times, maintaining, according to powder XRD patterns, its crystallinity, exhibiting, however, a small loss in its catalytic activity.
Later, four new indium MOFs, namely [In2(hfipbb)3(1,10-phen)2]·2H2O (phen: phenanthroline) (InPF-12), [In2(hfipbb)3(2,2′-bipy)2]·2H2O (InPF-13), [In2(hfipbb)3(4,4′-bipy)] (InPF-14) and [In4(OH)4(hfipbb)4(4,4′-bipy)] (InPF-15), were hydrothermally synthesised and their activity in the solvent-free cyanosilylation of carbonyl compounds was also studied.26 Among these catalysts, InPF-15 exhibited better activity in the cyanosilylation of benzaldehyde and acetophenone compared to other analogous catalysts (Table 2). As an example, the cyanosilylation of cyclohexanone using TMSCN in the presence of In-PF-15 resulted in a TOF value of 12048 h−1 at 80 °C with 0.1 mol% catalyst loading under a nitrogen atmosphere. InPF-15 maintains its crystallinity after seven catalytic cycles and hot filtration experiments, supporting the heterogeneity of the process.
Catalyst | Benzaldehyde | Acetophenonec | ||
---|---|---|---|---|
Yieldb (%), (h) | TOFd (h−1) | Yieldb (%), (h) | TOFd (h−1) | |
a Without solvent, 2.5 mol% of catalyst under a N2 atmosphere, 50 °C (benzaldehyde) or 80 °C (acetophenone). b Based on GC-MS. c All reactions were completed at 7 h. d TOF: mmol substrate/mmol catalyst × h. | ||||
InPF-11 | 99 (0.67) | 105 | 65 (4) | 109 |
InPF-12 | 99 (2) | 284 | 24 (4) | 1 |
InPF-13 | 93 (2) | 298 | 41 (2) | 38 |
InPF-14 | 99 (0.75) | 332 | 45 (1) | 50 |
InPF-15 | 80 (1.5) | 265 | 77 (4) | 140 |
Recently, the solvothermal reaction between Cu(OAc)2 and H3BTT·2HCl produced a sodalite-type MOF with the following structural formula: [(Cu4O0.27Cl0.73)3(H0.5BTT)8(H2O)12]·3MeOH·9DMF.27 This MOF exhibits a porous 3D (3,8)-connected framework constructed by square [Cu4(μ4-O/Cl)] units and triangular BTT ligands. Furthermore, this material can be dehydrated by removing solvent molecules to form [(Cu4O0.27Cl0.73)3(H0.5BTT)8], possessing coordinatively unsaturated Cu2+ centres. Interestingly, the activity of [(Cu4O0.27Cl0.73)3(H0.5BTT)8] at 1 mol% for the cyanosilylation of benzaldehyde led to a 96% conversion at 40 °C under a N2 atmosphere, which is one-eleventh of the loading used in the related Mn–BTT catalyst.68 On the other hand, the reactivity of [(Cu4O0.27Cl0.73)3(H0.5BTT)8] was higher than that of Sc-MOF and Ln-MOFs in which the loading of catalysts was 5 mol%.28,69 Leaching tests under optimized reaction conditions demonstrated the heterogeneity of the process. Notably, the use of 1-naphthaldehyde as the substrate gave only 78% of the product, while the use of 9-anthracenecarboxaldehyde did not afford any conversion. These experimental data are in clear agreement with the expectations of size selective catalysis, implying that the cyanosilylation occurs within the pores of [(Cu4O0.27Cl0.73)3(H0.5BTT)8]. The catalyst could be reused five times for the reaction of furfuraldehyde and TMSCN without noticeable loss of activity.
Three Ln-based 2D MOFs with the formula [Ln(3,5-DSB)(Phen)] (Ln = La, Pr (RPF-18); Ln = Nd (RPF-19)) were hydrothermally synthesized and their activity in the cyanosilylation of carbonyl compounds was examined.28 RPF-19-Nd exhibits the highest activity for the cyanosilylation of benzaldehyde with 5 mol% catalyst loading at 50 °C under a N2 atmosphere, showing a TOF value of 12.94 h−1. Furthermore, the reactivity of RPF-19-Nd (benzaldehyde, 2 h, 94.8% yield) was higher than that of Mn-MOF (benzaldehyde, 9 h, 98% yield),68 with the latter requiring a solvent. In contrast, the activity of RPF-19-Nd was slightly lower than that of Ln-MOF (benzaldehyde, 1 h, 88% yield);70 however, the latter requires a pre-activation treatment. The RPF-19-Nd catalyst was reused in four consecutive runs without observing appreciable changes. The synthesis of a porous 4-fold interpenetrated MOF, namely [Cd3(tipp)(bpdc)2]·DMA·9H2O, has been reported and its activity tested for the solvent-free cyanosilylation of carbonyl compounds.29 The reaction of benzaldehyde with TMSCN afforded 94% yield of the desired product with 6 mol% of [Cd3(tipp)(bpdc)2]·DMA·9H2O at room temperature with a turnover number (TON) value of 156.7. Moreover, no catalytic activity was observed for acetophenone and 4-methylacetophenone in the cyanosilylation reactions with [Cd3(tipp)(bpdc)2]·DMA·9H2O. The solvothermal reaction of 5,5′-(2,3,6,7-tetramethoxyanthracene-9,10-diyl)diisophthalic acid (H4L, Scheme 2) and Ba(NO3)2 resulted in the formation of a microporous MOF, namely Ba2(H2LOMe)2(NMP)·NMP, with high thermal stability. Upon heating at 325 °C, Ba2(H2LOMe)2(NMP)·NMP transformed to [Ba(H2LOMe)0.5(H2O)·4H2O]n through a single-crystal-to-single-crystal conversion.30 Furthermore, the formation of an intermediate was observed and its catalytic activity studied in the cyanosilylation of carbonyl compounds with TMSCN. The intermediate Ba MOF catalyst showed 100% yields for acetophenone, p-methylacetophenone, p-chloroacetophenone, p-nitroacetophenone, and 4-phenylbut-3-en-2-one at room temperature. The intermediate Ba MOF exhibited higher activity than Sc-MOF, Cu-DDQ and CPO-27-Mn catalysts for the addition of TMSCN with acetophenone.71–73 Moreover, the TON and TOF for the cyanosilylation of benzaldehyde were 192 and 384 h−1, respectively, which was much higher than those achieved for other MOF catalysts mentioned above. Also, the activity of this intermediate catalyst was higher than for polymer-supported metal complexes and SbCl3-based catalysts.74,75 The Ba MOF catalyst maintained its activity in the cyanosilylation of benzaldehyde in three consecutive cycles.
The activity of M(4,4′-bpe)2(H2O)4(m-bds) (M = Zn2+ or Cd2+) was investigated in the synthesis of 2-amino-6-(arylthio)pyridine-3,5-dicarbonitriles under solvent-free conditions using benzaldehyde, malononitrile, and thiophenol as reagents, as shown in Scheme 3.31 The use of Zn and Cd MOFs as catalysts gave 86 and 87% yield, respectively, under solvent-free conditions, while the same MOFs required twenty times longer time to achieve comparable yields using toluene under identical reaction conditions. These experiments clearly demonstrate the advantage of performing organic transformations under solvent-free conditions by reducing the time as well as energy consumption associated with the process. It could be that the excellent packing already documented for alkyl aromatics inside the cavities of certain MOFs disfavors the accessibility for substrates inside MOFs,76,77 just the opposite role that the solvents are supposed to play. Furthermore, these catalysts also afforded the desired product with substituted thiophenols in high yields (81–87%) under solvent-free conditions.
Dihydropyrimidinones (DHPMs) exhibit a variety of important biological activities, acting as calcium channel blockers, antihypertensives, adrenergic antagonists, and so forth. On the other hand, dihydropyridines (DHPs) are among the most widely used drugs, some relevant examples being shown in Scheme 4. These heterocyclic rings are also known as neuroprotectants because they exhibit anti-platelet aggregation properties and, therefore, they are important in the treatment of Alzheimer's disease and as anti-ischemic agents.78–81
Most of these nitrogen heterocycles are synthesised using a large variety of homogeneous catalysts. In this context, the use of isoreticular MOFs (IRMOF-3) as heterogeneous catalysts to promote the Hantzsch and Biginelli coupling towards DHPMs and DHPs has been reported.32 The reaction of benzaldehyde, methyl acetoacetate and ammonium acetate using IRMOF-3 as a catalyst under solvent-free conditions afforded DHP with 90% yield (Scheme 5). This catalyst was also tested for the synthesis of DHPMs, and high yields of the desired products were observed here also (Scheme 5). It is however, worth noting that leaching and reusability data, necessary to support the heterogeneity of the process as well as catalyst stability, were not studied.
In another precedent, IRMOF-3 was reported as a catalyst for the synthesis of hydroquinoline (Scheme 6) with moderate to high yields under solvent-free conditions.33 The catalyst was reused four times without any change in the yield for the first three runs; however, the yield from the third to the fourth run decreased from 90 to 72%. Unfortunately, there was no data available on the used catalyst that could shed light on the reasons for the catalyst deactivation.
MIL-101(Cr) is a robust MOF whose crystal structure defines two types of cages of 29 and 34 Å diameter that can be accessed through five- and six-membered ring windows. Keggin-type phosphotungstic acid (H3PW12O40, PTA) was encapsulated in the mesocages of MIL-101 (PTA@MIL-101) and its catalytic activity was studied in the three-component Biginelli condensation between aldehydes, 1,3-dicarbonyl compounds and urea or thiourea under solvent-free conditions.34 Interestingly, PTA@MIL-101 was extremely active in promoting the reaction of the substituted benzaldehydes with methyl acetoacetate and thiourea to provide various DHPM derivatives. However, a hot filtration test revealed the presence of PTA in the liquid phase, thus contributing to the yield increase after the removal of the catalyst. Nevertheless, the reusability data indicated that there was no decline in the efficiency of the catalyst for up to three runs.
In this context, in another precedent, Cu3(BTC)2 has been reported as a heterogeneous catalyst for the Michael addition between β-nitrostyrene and indole, achieving 98% yield of the adduct under solvent-free conditions (Scheme 8).36 This activity of Cu3(BTC)2 is comparable with those of UiO-67-Squar/bpdc MOF83 and Cu(dbda) using a squaramide ligand84 in chloroform.
Scheme 9 Knoevenagel condensation reaction between benzaldehyde and malononitrile using [TPA]2[Zn2-(BDC)3(DMA)2] as a catalyst. |
The next section describes the incorporation of ILs inside the pores of the MOF. Organic tetraalkylammonium halides have been used in combination with MOFs to promote CO2 insertion in epoxides. Thus, MIL-101(Cr) has been reported as a heterogeneous catalyst promoting the solvent-free coupling of CO2 with styrene oxide to produce styrene carbonate.40 The presence of tetrabutylammonium bromide as a co-catalyst was essential for the formation of the styrene carbonate. On the other hand, the cycloaddition of CO2 to propylene oxide resulted in the formation of propylene carbonate. The catalyst stability was, however, insufficient and the catalyst lost its crystallinity after being reused twice, as evidenced from powder XRD patterns.
An amino group functionalized IRMOF-3 (F-IRMOF-3; Scheme 12) was reported as a heterogeneous catalyst for the solvent-free synthesis of cyclic carbonates without any co-catalyst, reaching TOF values of 778 mmol g−1 h−1.41 Furthermore, F-IRMOF-3-4d resulted in better catalytic activity than other reported catalysts like MIXMOF/Et4NBr (499 mmol g−1 h−1),112 MgO (0.9 mmol g−1 h−1),113 SmOCl (4.1 mmol g−1 h−1),114 cross-linked polymeric nanoparticles (127 mmol g−1 h−1),115 hydroxyl IL grafted onto cross-linked divinylbenzene polymer (70 mmol g−1 h−1)116 and polystyrene-supported threonine (16 mmol g−1 h−1).117 F-IRMOF-3-4d was reused for four runs without much change in its catalytic activity.
A 3D MOF, [Zn4O(2,6-NDC)(BTB)4/3] (MOF-205, BET = 4200 m2 g−1), was synthesised by microwave heating and its catalytic activity was examined in the cycloaddition of CO2 to propylene oxide to produce the corresponding five-membered cyclic carbonates under solvent-free conditions.42
All the above data show that MOFs can frequently be used as solid catalysts in the absence of solvents for condensations, Michael addition, cycloadditions and couplings, giving yields that are even better than in the presence of solvents. Although it should be remarked that catalysis in the absence of solvent is not specific to MOFs and most of the above discussed reactions have also been conducted with other solid catalysts, the present review aims to illustrate that under solvent-free conditions, MOFs can exhibit even higher catalytic activity than in the presence of solvents. An additional advantage is the environmental benefits derived from the diminution of wastes associated with the use of solvents.
As an example to illustrate this approach, a size- and shape-selective porous catalyst embedding IL, namely HKUST-1 incorporating an amino-functionalized basic ionic liquid (ABIL-OH/HKUST-1), was synthesized by impregnation (Scheme 13).43 The characterization of this catalyst revealed that the ABIL-OH is likely confined within the nanocavities of HKUST-1 via a Cu–NH2 coordination bond. The fact that the ABIL-OH/HKUST-1 catalyst exhibits a pale blue colour that can be visually perceived was taken as evidence of the change in the coordination of copper ions as the colour is associated with the d–d transition of Cu2+ ions. Furthermore, XPS studies of ABIL-OH/HKUST-1 indicate that the Cu 2p3/2 and Cu 2p1/2 binding energy shifted from 933.8 eV to 933.3 eV and from 953.5 eV to 953.0 eV, respectively. This slight reduction of binding energy is in agreement with the occurrence of an electron transfer from the NH2 to the Cu2+ centre. The loading of ABIL-OH in ABIL-OH/HKUST-1 was determined to be 1.07 mmol g−1 by elemental analysis.
Scheme 13 Structure and proposed interaction of ABIL-OH with HKUST-1 and Knoevenagel condensation catalyzed by ABIL-OH/HKUST-1. Reproduced from ref. 43 with permission from Elsevier. |
The catalytic activity of ABIL-OH/HKUST-1 was evaluated in the Knoevenagel condensation between benzaldehyde and malononitrile. The as-synthesized HKUST-1 resulted in 55.3% benzaldehyde conversion after 240 min. On the other hand, the ABIL-OH/HKUST-1 catalyst exhibited complete benzaldehyde conversion after 210 min under the same conditions. Meanwhile, the use of ABIL-OH as a homogeneous organocatalyst resulted in complete benzaldehyde conversion after 120 min. This high activity of ABIL-OH could be due to the availability of the free NH2 group in ABIL-OH, thus acting as a Lewis base catalyst. ABIL-OH/HKUST-1 showed high catalytic activity (>90% conversion) with 100% selectivity after being reused for five recycles; however, a slight decline in conversion is observed, again indicative of the occurrence of some deactivation. On the other hand, comparison of the powder XRD patterns between the fresh and reused catalysts revealed that the main diffraction peaks are still observed in the reused catalyst, although the disappearance of several peaks at high angles was observed, indicating that there were some changes in the MOF structure. It has frequently been observed that HKUST-1 has an unstable structure that can collapse due to the strong binding of Cu2+ with substrates, reagents or solvents.128,129 However, a hot filtration test indicated the heterogeneity of the process without clues of Cu leaching. No catalytic activity was observed for the reaction between o-hydroxybenzaldehyde (6.65 × 5.41 Å) and 3-methoxy-4-hydroxybenzaldehyde (7.60 × 6.43 Å) with malononitrile in the presence of ABIL-OH/HKUST-1, whereas these substrates were completely converted to their respective products with ABIL-OH as the homogeneous catalyst under identical reaction conditions. This difference in activity was attributed to the confinement of the active site within the microenvironment of HKUST-1.
Recently, UiO-66 was loaded with different percentages of 1-methylimidazolium-3-propylsulfonate hydrogensulfate and its activity was studied in the oxidative desulfurization using hydrogen peroxide as the oxidant.45 It was observed that at 40 °C the removal of sulfur is higher than 94%, with an O/S molar ratio of 7:1 using 40% 1-methylimidazolium-3-propylsulfonate hydrogen sulfate supported on UiO-66. Increasing the content of 1-methylimidazolium-3-propylsulfonate hydrogen sulfate lowered the activity to 92.9% under identical conditions. Reusability experiments showed that the removal of sulfur was 90.6% after the sixth cycle, thus, showing the stability of the catalytic activity under the experimental conditions used. However, no data were given on leaching or the structural integrity of the reused sample compared to the fresh catalyst.
Recently, phosphotungstic acid (H3PW12O40, PW) was immobilized within the nanocages of DAIL-modified MIL-101(Cr) to obtain PW/DAIL/MIL-101(Cr) and its activity in the oxidation of benzyl alcohol using TBHP was examined.46 An analogous catalyst, namely PW/MIL-101(Cr), was also prepared as a control and its activity was compared with that of PW/DAIL/MIL-101(Cr) to gain information on the role played by DAIL during this oxidation reaction. Activated MIL-101(Cr) was treated with DAIL to obtain DAIL/MIL-101(Cr) in which a strong coordination bond between the metal centres of the MOF and the amino groups of the DAIL was proposed to occur. The immobilization of PW onto DAIL/MIL-101(Cr) was performed through anion exchange of the corresponding polyoxometal and the DAIL. The PW cluster has an appropriate dimension to diffuse through the larger12 Å windows of MIL-101(Cr). The preparation procedure followed to obtain PW/DAIL/MIL-101(Cr) and the chemical structure of DAIL are illustrated in Fig. 1. The powder XRD pattern of DAIL/MIL-101(Cr) showed good agreement with that of MIL-101(Cr), confirming that the crystal structure remains mostly intact after the functionalization of the MOF with DAIL cations. It was observed that the thermal stability of PW/DAIL/MIL-101(Cr) increased compared to that of DAIL/MIL-101(Cr) and this enhanced stability was proposed to be due to the additional interactions between DAIL and PW molecules. The XRF measurements indicated that the loading values of PW were 0.18 mmol g−1 and 0.23 mmol g−1 for HPW/MIL-101(Cr) and PW/DAIL/MIL-101(Cr), respectively. PW/DAIL/MIL-101(Cr) exhibited higher catalytic activity (95% conversion and 1900 TON) than HPW/MIL-101(Cr) (70% conversion and 1400 TON) in the oxidation of benzyl alcohol to benzaldehyde. On the other hand, the conversion of benzyl alcohol was influenced by the loading of DAIL, as was observed from the fact that benzyl alcohol conversion gradually increased when DAIL loading increased from 0.3 to 0.9 mmol g−1. These data suggest that the presence of the DAIL group is essential to achieve high activity. In contrast, MIL-101(Cr) gave 43% benzyl alcohol conversion under identical conditions. The hot filtration test proved the heterogeneity of the process. The catalyst was reused five times with no significant decrease in the catalytic activity. The similar powder XRD patterns of the fresh material and reused PW/DAIL/MIL-101(Cr) catalyst indicate that the system is stable under the reaction conditions. Although high conversion of benzyl alcohol was observed with this catalytic system in chloroform, it is clear that the use of volatile organic solvents, particularly chlorinated solvents, do not comply with the principles of green chemistry and further work is necessary to replace these liquids with more environmentally tolerable solvents. In another precedent, the catalytic performance of PW4/DAIL/MIL-100(Fe) was compared with that of the DAIL free catalyst (HPW4/MIL-100(Fe)) in the oxidation of benzyl alcohol using TBHP as an oxidant.47 It was interesting to note that the TON value of the former catalyst is 613, which is superior to that of the latter catalyst, with the value of 433, under identical conditions and this higher activity is ascribed to the ability of the DAIL groups to form hydrogen bonds, enhancing the accessibility of TBHP.
Fig. 1 Structure of DAIL and synthesis of PW/DAIL/MIL-101(Cr) catalyst. Reproduced from ref. 46 with permission from the Royal Society of Chemistry. |
Fig. 2 Preparation of functionalized magnetic MOF incorporating the acid IL DAIL-Fe3O4@NH2-MIL-88B(Fe). Reproduced from ref. 48 with permission from the American Chemical Society. |
A new strategy designed to construct the IL, POM, and MOF composite is shown in Fig. 3. In this approach, POM was first loaded within the cages of MIL-100(Fe) to play the role of a bridge. Then, the IL was introduced to encapsulate the heteropolyanion.49 The POM-IL-functionalized MOF has several advantages such as a simple preparation process, high stability, high content of active components, ease of separation and high reusability. Furthermore, the preparation method did not use toxic or harmful reagents, such as HF and DMF, thus exhibiting advantages from the environmental point of view. The catalytic activity of [SO3H-(CH2)3-HIM]3PW12O40@MIL-100(Fe) was studied as a catalyst in the esterification of oleic acid with ethanol. The pristine MIL-100(Fe) showed some catalytic activity of 15.5% conversion at 111 °C, probably due to the presence of coordinatively unsaturated Fe3+ ions as Lewis acid sites. On the other hand, encapsulated phosphotungstate HPW12O40@MIL-100(Fe) (acidity: 0.39 mmol H+ g−1) gave 40.3% conversion of oleic acid. Importantly, 94.6% conversion was achieved using [SO3H-(CH2)3-HIM]3PW12O40@MIL-100 as a catalyst (acidity: 1.74 mmol H+ g−1). This enhanced activity of the composite catalyst can be attributed to the sulfonic acid groups in the IL cation increasing the number of Brønsted acid sites compared to the heteropolyanion. Also, the existence of a cooperative effect between Lewis and Brønsted acid sites improving the activity of the catalyst was proposed.140 Although the catalytic activity of the pure IL was higher (95.1%), the heterogeneous catalyst could be easily separated from the reaction medium. In contrast, the [SO3H-(CH2)3-HIM][HSO4]@MIL-100(Fe) catalyst that lacked POM and was prepared by the direct impregnation of MIL-100(Fe) with IL exhibited 90.2% conversion under identical condition; however, its activity decreased upon reuse (Fig. 4). Overall, the [SO3H-(CH2)3-HIM]3PW12O40@MIL-100(Fe) catalyst showed optimal activity for the esterification of oleic acid compared with other catalysts.135,141–144 The catalyst was reused for six cycles without any appreciable decline in its activity.
Fig. 3 Synthesis of [SO3H-(CH2)3-HIM]3PW12O40@MIL-100(Fe) and its use as an esterification catalyst for the preparation of biodiesel. Reproduced from ref. 49 with permission from Wiley. |
Fig. 4 Activity of [SO3H-(CH2)3-HIM]3PW12O40@MIL-100(Fe) (blue) and [SO3H-(CH2)3-HIM][HSO4]@MIL-100(Fe) (red) upon recycling for the esterification of oleic acid with ethanol. Reproduced from ref. 49 with permission from Wiley. |
A Brønsted IL acid, namely N-methyl-2-pyrrolidonium methyl sulfonate ([NMP]+CH3SO3−), was immobilized on MIL-101(Cr) and its catalytic activity was evaluated in the esterification of acetic acid with amyl alcohol.50 This catalyst exhibited 82.4% conversion which is comparable to those of [NMP]+CH3SO3− (97%)145 and HPW/MIL-101 (92.3%)146 catalysts for the esterification reaction. Although the Brønsted IL acid loaded on MIL-101(Cr) showed slightly lower activity compared to those of [NMP]+CH3SO3− and HPW/MIL-101, the merit of MIL-1001(Cr) supported IL is to be a heterogeneous catalyst, avoiding tungstic acid. Furthermore, the catalyst was reused six times without any considerable decline in its activity.
Another Brønsted acidic IL (BAIL) confined inside the nanocages of MIL-101(Cr) was reported for the acetalization of benzaldehyde with glycerol (Scheme 14).51 The BAIL functionalized MIL-101(Cr) catalyst was prepared by heating at reflux temperature imidazole (IMIZ) or triethylene diamine (TEDA) with the dehydrated MIL-101(Cr). These precursors were further treated with 1,4-butane sultone and exchanged with H2SO4, as shown in Fig. 5. The parent MIL-101(Cr) showed 32% conversion of benzaldehyde at 90 °C after 3 h, this low conversion being probably due to its mild Lewis acidity. Interestingly, the TEDA-BAIL/MIL-101(Cr) catalyst afforded 88% conversion of benzaldehyde while 58% conversion of benzaldehyde is achieved with the IMIZ-BAIL/MIL-101(Cr) catalyst under identical conditions. Furthermore, the catalytic activity achieved by TEDA-BAIL/MIL-101(Cr) was comparable with that of the benchmark catalyst, namely p-toluenesulfonic acid, which gave 90% conversion of benzaldehyde. The catalyst was recycled six times without any significant decline in its catalytic activity or leaching of active sites.
Fig. 5 Procedure for the synthesis of BAIL confined in MIL-101 nanocages by post-synthetic modification. Reproduced from ref. 51 with permission from the Royal Society of Chemistry. |
Fig. 6 Synthesis of Pd/IL/Cu3(BTC)2 catalyst and its catalytic activity for the selective partial hydrogenation of phenylacetylene. Reproduced from ref. 52 with permission from the Royal Society of Chemistry. |
Fig. 7 Synthesis of HL and formation of the corresponding MOF by reaction with copper nitrate. Reproduced from ref. 53 with permission from the Royal Society of Chemistry. |
Fig. 8 Synthesis of L, UiO-67-IM and CuI@UiO-67-IM. Reproduced from ref. 54 with permission from the American Chemical Society. |
A strategy was developed to combine together coordinatively unsaturated metal sites based MOFs with functional IL to form heterogeneous bifunctional catalysts (Fig. 9) and its activity was tested in the cycloaddition of CO2 with epoxides under mild conditions in the absence of a co-catalyst. In this context, two bifunctional heterogeneous catalysts incorporating IL, MIL-101(Cr)-X(n-Bu)3Br (X: N or P) were synthesised with CUS as Lewis acid centres and embedding ILs as additional active sites.55 Characterization data by powder XRD and IR spectroscopy indicated the formation of MIL-101-NH2 and its modification by ILs. MIL-101-N(n-Bu)3Br and MIL-101-P(n-Bu)3Br converted CO2 and propylene oxide into propylene carbonate (PC, Scheme 18) with 99.1 (110.1 TON) and 98.6% (109.6 TON) yields, respectively. Control experiments using mixtures of pure components of MIL-101-NH2, (nBu)4NBr or (n-Bu)4PBr showed a comparatively lower yield of PC under the same conditions, which can be rationalized due to the existence of only monofunctional catalytic sites in these materials. Furthermore, although the physical mixtures of MIL-101-NH2 with (n-Bu)4NBr and MIL-101-NH2 with (n-Bu)4PBr exhibited 98.4 and 97.9% yields, respectively, the values of which are similar to those achieved by MIL-101-N(n-Bu)3Br and MIL-101-P(n-Bu)3Br, the advantage of the IL-functionalized MOFs was that there was no decline in activity even after three cycles, while the catalysts consisting of the physical mixture of IL and MOF showed a dramatic decrease in yield in the second cycle. In one case, ICP analysis revealed the presence of 0.012 mg of phosphorus in solution. IR spectroscopy and powder XRD patterns for the fresh and recycled MIL-101(Cr)-X-(n-Bu)3Br catalysts did not show appreciable changes, thus, supporting the stability of the material under the experimental reaction conditions. Furthermore, the advantage of these bifunctional MOF catalysts was confirmed by comparing their activity with some benchmark MOFs153–157 under the same reaction conditions. As shown in Fig. 10, MIL-101-X-(n-Bu)3Br showed the highest yield to PC (>98% conversion). In contrast, other MOFs exhibit only a very low conversion, with 2.5, 5.2, 5.4, 20.9 and 23.2% for MOF-5, IRMOF-3, HKUST-1, MIL-101 and Mg-MOF-74, respectively. This enhanced catalytic activity of MIL-101(Cr)-X-(n-Bu)3Br was believed to be due to the synergetic catalytic roles of Lewis acid sites and the Br− ion of ILs as a nucleophile. These experiments clearly demonstrate that bifunctional catalysts offer enhanced activity compared to monofunctional catalysts.
Fig. 9 Structure of MIL-101(Cr)-X-(n-Bu)3Br (Cr: green, C: gray, O: red, N: blue, Br: amaranth, and R = N or P). Reproduced from ref. 55 with permission from the Royal Society of Chemistry. |
Fig. 10 Comparison of PC yield in the cycloaddition CO2 to propylene oxide and catalyzed by bifunctional MIL-101(Cr)-X-(n-Bu)3Br compared to other MOFs. Reproduced from ref. 55 with permission from the Royal Society of Chemistry. |
A new imidazolium-based UiO-67 type MOF (UiO-67-IL) was synthesised using ligand functionalization (Fig. 11).56 The catalytic activity of UiO-67-IL was studied in the cycloaddition of CO2 to epichlorohydrin (Scheme 19) with or without TBAB as co-catalyst under 1 atm of CO2. The powder XRD pattern indicates that UiO-67-IL is highly crystalline and its structure is similar to that of pristine UiO-67. The use of UiO-67-IL afforded 99% yield (45.08 h−1 TOF) of the desired product with TBAB as a co-catalyst at 90 °C, while the same catalyst required, under identical conditions, a much longer time to reach 88% yield (13.66 h−1 TOF) without TBAB. On the other hand, no catalytic activity was observed with UiO-67 in the absence of IL even at 90 °C (<3% yield); however, the carbonate yield increased to 8% yield for longer time. In contrast, UiO-67-IL with imidazolium moieties afforded 95% yield under similar conditions. On the other hand, the UiO-67-TBAB catalytic system showed 98% yield, whereas UiO-67-IL gave 99% yield in less than half the time for epichlorohydrin CO2 cycloaddition. These data clearly illustrate the pivotal role played by the embedded imidazolium group in UiO-67-IL catalyzing this reaction. No leaching of active sites was detected under this experimental condition. Also, the catalyst was reused for five catalytic runs. Powder XRD patterns of the fresh and five-times reused catalyst showed that the structural integrity is maintained.
Fig. 11 Synthesis of UiO-67-IL by using an IL-functionalized biphenyldicarboxylic acid as linker. Reproduced from ref. 56 with permission from the American Chemical Society. |
Recently, the covalent post-functionalization of zeolitic imidazolate framework-90 (ZIF-90) with a pyridinium based IL was performed to obtain IL-supported ZIF-90 (IL-ZIF-90) and its activity was examined in the solvent-free cycloaddition of epoxides with CO2.57 2-Imidazolecarboxaldehyde (ICA) showed 19% conversion while aminopyridinium iodide (AMPyI) afforded 44% conversion with 100% selectivity, whereas the combined physical mixture of all the starting materials (zinc nitrate, ICA and AmPyI) employed for the synthesis of IL-ZIF-90 resulted in slightly higher conversion of 55% under identical conditions. On the other hand, the use of ZIF-90 as a catalyst resulted in 51% conversion which is slightly lower than that of the combination of ZIF-90 and AmPyI (65% conversion). Interestingly, IL-ZIF-90 exhibited 97% conversion with 98% selectivity under identical conditions. These data clearly demonstrate the advantage of post-functionalization in MOFs, enhancing the activity of ILs within the frameworks of MOFs. Among the various epoxides screened, propylene oxide afforded a maximum conversion of 97%, whereas cyclohexene oxide showed 9% conversion under identical conditions. This lower activity was due to the steric hindrance caused by the cyclohexene ring. IL-ZIF-90 was reused in four successive runs without any significant decline in its activity. Powder XRD patterns of the fresh catalyst did not show any noticeable changes in the four-times reused catalyst.
Recently, a facile route has been reported for the preparation of carbon nitride (CN) foams as structural templates with micrometer-sized pores with the nitrogen content of 25.6 wt% by the fast carbonization of melamine foam (Fig. 12). The nitrogen functionalities of CN foam provide the feasibility of chemical anchoring and growth of ZIF-8 crystals, thus leading to the development of hierarchical porous MOFs.160 Furthermore, the growth of ZIF-8 crystals also changes the nature of the CN foam from hydrophilic to highly hydrophobic, thus creating an effective shield against water for the MOF crystals. As a result, the introduction of ZIF-8 crystals onto the CN foam provides selective absorption of oils up to 58 wt% from water/oil mixtures and also facilitates the highly efficient conversion of CO2 to 3-chloropropene carbonate in a quantitative yield with selectivity. In contrast, the use of ZIF-8-based catalysts for this reaction resulted in low product selectivity to 3-chloropropene carbonate (52%) and also produced by-products including diols (23.7%) and a dimer (24.3%).161 Also, ZIF-8 catalysts showed a significant loss in their catalytic activity after the first cycle whereas ZIF-8/CN foam retained high selectivity and conversion up to the fourth cycle. Hence, this approach could be extended to those MOFs that have poor stability in water.
Fig. 12 Schematic representation of the synthesis of ZIF-8/CN foam with hierarchical porosity. Reproduced from ref. 160 with permission from Wiley. |
The ability of MOFs to gelate under specific experimental conditions provides further opportunities in the preparation and shaping of hierarchically porous MOF monoliths, which could be used in catalysis and adsorption applications. Recently, xero- or aerogel monoliths consisting solely of nanoparticles of several prototypical Zr4+-based MOFs such as UiO-66-X (X = H, NH2, NO2, (OH)2), UiO-67, MOF-801, MOF-808 and NU-1000, have been reported.162 Furthermore, UiO-66 gels were shaped into monolithic spheres of 600 mm diameter using an oil drop method, thus creating different materials which could find applications in packed-bed catalytic or adsorptive applications, where hierarchical pore systems can greatly weaken mass transfer limitations.
A pyridyl-decorated MOF-505 analogue, [Cu2(L)(H2O)2]·Gx (H4L = 5,5′-(pyridine-2,5-diyl)-diisophthalic acid, G: solvent molecule), has been reported with rare hierarchical meso- and microporosity with exposed unsaturated CuII sites and Lewis basic pyridyl sites. The catalytic activity of this catalyst was reported in the cyanosilylation reaction under solvent-free conditions.163 The dehydrated catalyst showed 99% conversion in the cyanosilylation of benzaldehyde at 40 °C. The catalytic activity of this hierarchical MOF is comparable to that of other Cu or Zr-based MOFs72,164,165 in the cyanosilylation of benzaldehyde but is considerably higher than the catalytic acitvity of other catalysts like Cu3(BTC)2 (50–60%),129 Cd-MOF (77%),166 and Ce-MOF (94%, 2 h).167 This enhanced activity is due to the presence of a high density of the exposed metal sites and the improved molecular accessibility of the hierarchical meso- and microporosity. Catalysis was heterogeneous in nature and MOF-505 was recycled up to five times with only a minor loss of activity.
In another precedent, a continuous phase transformation processing strategy has been applied for fabricating and shaping MOFs into shaped bodies and MOF foams that exhibit reversible transformation among these states. In this context, a cup-shaped reactor based on a Cu-MOF composite and hierarchically porous MOF foam were prepared and their activities were compared in the oxidation of diphenylmethane with TBHP as an oxidant.168 The use of a cup-shaped reactor afforded 76% conversion with 93% selectivity, while using porous foam enhances the conversion to 92% with 97% selectivity under identical conditions. Furthermore, quenched solid-state density functional theory calculation revealed that the foam possesses hierarchically distributed porosity with the pore size distribution consisting of both micro- and mesopores expanding up to 40 nm. The hierarchically porous MOF foam was reused for three cycles with no significant decline in its activity.
Recently, a coordination replication strategy has been applied to construct hierarchical Cu-based MOF nanoarrays and their activity has been examined in the reduction of 4-nitrophenol to 4-aminophenol using sodium borohydride as a reducing agent.169 In this approach, Cu(OH)2 nanorods were the source for Cu that coordinated with a series of organic ligands to form different MOF crystals, and they also served as a 3 D substrate to support the growth of the Cu-based MOF (Fig. 13). Hierarchical MOF nanoarrays exhibited enhanced activity in the reduction of 4-nitrophenol compared with those on MOF-2 nanoparticles and Cu(OH)2 nanorod arrays. This is due to the unique 3D hierarchical nanoarray architecture of the nanocomposite. There was no significant loss in activity for eight successive catalytic cycles thus showing the long term stability of hierarchical MOFs.
Fig. 13 Preparation of hierarchical Cu-based MOF nanoarrays by the coordination replication strategy. Reproduced from ref. 169 with permission from Wiley. |
Two hierarchically porous MOFs namely, NUS-6 (NUS: National University of Singapore) composed of either zirconium (Zr) or hafnium (Hf) clusters were synthesised using a modulated hydrothermal method with high stability and strong Brønsted acidity.170 These MOFs exhibit BET surface areas of 550 and 530 m2 g−1 for NUS-6(Zr) and NUS-6(Hf), respectively, and a hierarchically porous structure of coexisting micropores (∼0.5, ∼0.7, and ∼1.4 nm) and mesopores (∼4.0 nm) with dangling sulfonic acid groups. Structural analysis showed that the hierarchical porosity of NUS-6 derives from missing linkers and clusters of the parent UiO-66 framework. The activity of these MOFs was tested in the dehydration of fructose to 5-hydroxymethylfurfural (HMF), in which NUS-6(Hf) showed 98% yield which is superior than NUS-6(Zr) (84% yield) after 1 h at 100 °C. This enhanced activity is due to the stronger Brønsted acidity contribution from Hf-μ3-OH groups as well as the availability of smaller pore sizes to restrict undesirable reactions.
The above examples illustrate how the preparation of hierarchical MOFs with controlled micro-/mesoporosity and particle shapes can enhance the catalytic activity of these solids by facilitating the accessibility of substrates and reagents to the active sites. This particle shaping is of great importance in commercial catalysts used in industry, but is an aspect that is frequently ignored in most catalytic studies. Shaping MOF particles with an adequate morphology and ordering is a necessary step to be considered in any potential application of these materials as adsorbents and catalysts.
One point that has not been, in our opinion, sufficiently addressed is MOF stability as a catalyst under solvent-free conditions. It should be considered that the most robust MOFs have pore sizes of 1 nm or below and structure stability decreases as the pore size increases. Besides the lack of structural stability that is a general issue to be carefully addressed when using MOFs as catalysts, specific points that need to be investigated are how product adsorption affects catalytic activity and what are the causes of deactivation under solvent-free conditions. A general claim in most of the reports presented is the lack of the catalyst deactivation without paying attention to why product inhibition or byproduct formation do not seem to occur. In other words, all the solid catalysts should deactivate upon reuse and solvent-free conditions are specially favourable for poisoning due to the high concentration of products/byproducts and the absence of desorption by solvent molecules. It would be important to establish an in-depth study of the reasons why this deactivation process does not happen and the system always appears stable.
Another important aspect that has been frequently neglected is particle shaping and structuring resulting in the formation of hierarchical MOF catalysts. The presence of porosity combining micro- and mesoporosity is known to facilitate mass transport and this results in an enhanced activity using heterogeneous catalysts. Examples have been presented showing that the flexibility in MOF synthesis allows developing different strategies like linker labilization, gelation or synthesis on foams that have resulted in hierarchical MOFs.
Overall, the present review has shown that MOFs can contribute to the greenness of chemical processes not only by acting as efficient and selective solid catalysts, but also by avoiding the use of, in some cases, volatile solvents without the penalty of fast deactivation. Similarly, decoration or coating of the internal pores of MOFs by IL units has shown that it is possible to combine the effects of molecular ILs with that of a solid catalyst with the minimum amount of IL and with advantages in terms of recovery of IL and workup of the reaction mixture. It has also been shown that the preparation of MOFs exhibiting a hierarchical structuring enhances their catalytic activity by facilitating mass transport. Since MOF catalysis is under intense investigation, it can be predicted that additional efforts will be made to expand the use of MOFs as heterogeneous catalysts using the three strategies commented above to other types of reactions.
This journal is © The Royal Society of Chemistry 2018 |