Takayoshi
Hara
,
Jun
Kurihara
,
Nobuyuki
Ichikuni
and
Shogo
Shimazu
*
Department of Applied Chemistry and Biotechnology, Graduate School of Engineering, Chiba University, 1-33 Yayoi, Inage, Chiba 263-8522, Japan. E-mail: shimazu@faculty.chiba-u.jp; Fax: +81 43 290 3379; Tel: +81 43 290 3379
First published on 19th September 2014
The interlayer spacing of layered Ni–Zn mixed basic salts (NiZn) can be precisely controlled by the intercalation of various long alkyl chain carboxylate anions into the NiZn interlayer. The butyrate-exchanged NiZn (C3H7COO−/NiZn) catalyst effectively promotes the epoxidation of various cyclic enones with aqueous hydrogen peroxide in remarkably high yields. The C3H7COO−/NiZn-catalysed epoxidation of 2-cyclohexen-1-one with an equimolar amount of H2O2 proceeds in a highly efficient manner, with 97% efficiency of H2O2 utilization. This C3H7COO−/NiZn catalyst can be reused without any loss of its catalytic activity and selectivity.
Layered hydroxy double salts (HDSs), which consist of positively charged layers and exchangeable interlayer anions, have also received considerable interest as anion-exchangeable layered compounds.8 Compounds in the HDS family follow the general formula MA2+1−xMB2+2xAn−2x/n(OH)2·mH2O (0.15 < x < 0.25), where MA2+ and MB2+ are divalent metal cations such as Zn2+, Cu2+, Co2+, or Ni2+, and An− represents various interlayer anions. The anionic guests that can be used range from simple inorganic anions (NO3−, ClO4−, Cl−, MnO4−, SO42−, etc.) to anionic metal complexes to organic anions including drugs and dyes.9 It can be considered that HDSs are also powerful candidates for advanced nanoscaled catalysts or catalyst supports that allow for control over the location of the catalytically active metal species.
The epoxidation of CC double bonds is one of the platform synthetic processes in basic research and industrial applications, because epoxides and their intermediates are widely employed in the synthesis of high-valued chemicals, such as cosmetics, pharmaceuticals, and polymer materials.10 The epoxidation of electron-deficient olefins in particular, such as α,β-unsaturated ketones, has been an important target for the functionalisation of ketones. It is well known that this type of epoxidation generally requires a nucleophilic oxidant, such as OOH− species.11 Many epoxidation procedures employ H2O2/Bu4NF, TBHP/DBU, mCPBA/KOH, NaOCl/cinchona alkaloid, NaBO3/THAHS, iodosyl benzene, dioxirane, or TBA2S2O8 as an oxidant,12 however, these oxidants often yield stoichiometric amounts of their deoxygenated forms as a waste by-product. In contrast, hydrogen peroxide is an ideal oxidant because water is the only by-product. Although organometallic complexes, heteropolyacids, organic molecules, and their immobilized materials can act as effective catalysts for the epoxidation of enones under neutral conditions,13 utilization of homogeneous reaction systems with H2O2 under alkaline conditions (NaOH, KOH, LiOH, Na2CO3, or K2CO3) is the most common procedure. To replace these standard homogeneous systems with more practical heterogeneous alternatives, a number of solid bases such as Mg–Al hydrotalcite, KF/alumina, hydroxyapatite, or sodalite catalysts have been developed.14
We have previously developed a Ni–Zn mixed basic salt (NiZn) as a solid catalyst or catalyst support based on the design idea of “Intercalation Catalysts”.15 Ni1−xZn2x(OCOCH3)2x(OH)2·nH2O (0.15 < x < 0.25; NiZn), classified as a layered HDS, consists of positively charged layers and exchangeable interlayer anions. Our selection of NiZn as a catalyst material is motivated by its unique characteristics such as (i) simple preparation, (ii) high crystallinity, (iii) strong electrostatic interactions between guest anions and Zn2+ cations, and (iv) high anion exchange capacity, as well as (v) the ability to fine-tune the interlayer space by modifying the size of the guest anion. Recently, interlayer space-controlled NiZn with alkyl carboxylate anions intercalated into the interlayer has been found to be an effective heterogeneous Brønsted base catalyst for Knoevenagel condensation in water.15c Here, we report the successful epoxidation of cyclic enones using hydrogen peroxide in the presence of NiZn as a basic catalyst.
The powder X-ray diffraction (XRD) profiles of the synthesised NiZn catalysts are shown in Fig. 1. From the XRD results, the layered structure of NiZn was maintained in each case; however, the calculated clearance spaces of each material (C.S. = basal spacing (d001) – thickness of layer (0.46 nm)) were changed after the intercalation procedure, also shown in Fig. 1. The C.S. of each was proportional to the length of the alkyl groups intercalated (slope: 0.17 nm per CH2, Fig. 2). Taking the C.S. and the size of the pillared alkyl carboxylate anions into account, a suitable conformation of Cn−1H2n−1COO−/NiZn can be a bilayer arrangement with a tilting angle of approximately 40 degrees.
Fig. 2 The relation between C.S. and carbon number of alkyl groups in the carboxylate anion of Cn−1H2n−1COO−/NiZn catalysts. |
To explore the catalytic abilities of the anion-exchanged NiZn, the epoxidation of 2-cyclohexen-1-one (1) into 1,2-epoxycyclohexanone (2) was carried out in the presence of 4 equivalents of aqueous hydrogen peroxide (30 wt%) relative to the amount of 1. The results and the C.S. values of the Cn−1H2n−1COO−/NiZn catalysts are shown in Table 1.
Entry | Catalyst | C.S.b (nm) | Time (h) | Conv.c (%) | Yieldc (%) |
---|---|---|---|---|---|
a 1 (0.5 mmol), NiZn catalyst (0.05 g), DMF (2 mL), 30 wt% H2O2 aq. (4 eq. relative to the substrate), 60 °C. b Calculated from XRD data. c Determined by GC using an internal standard technique. d Ni catalyst (Ni: 0.26 mmol). e Zn catalyst (Zn: 0.13 mmol). f Hydrotalcite (0.05 g), purchased from WAKO, was used as a catalyst. g C3H7COONa (1.47 mmol) was used as a catalyst. | |||||
1 | HCOO−/NiZn | 0.45 | 1 | 7 | 6 |
2 | CH3COO−/NiZn | 0.84 | 1 | 45 | 45 |
3 | C3H7COO−/NiZn | 1.06 | 1 | 55 | 55 |
4 | C5H11COO−/NiZn | 1.49 | 1 | 28 | 28 |
5 | C7H15COO−/NiZn | 1.74 | 1 | 24 | 22 |
6 | C9H19COO−/NiZn | 2.11 | 1 | 19 | 16 |
7d | Ni(OCOCH3)2·4H2O | — | 3 | trace | trace |
8e | Zn(OCOCH3)2·2H2O | — | 3 | trace | trace |
9d | Ni(OH)2 | — | 3 | trace | trace |
10e | Zn(OH)2 | — | 3 | trace | trace |
11f | Mg6Al2(CO3)(OH)16·4H2O | 0.30 | 3 | 18 | 18 |
12g | C3H7COONa | — | 3 | 35 | 35 |
13 | Blank | — | 3 | no reaction |
In the case of HCOO−/NiZn, the catalytic activity was low (entry 1), most likely due to the mass transfer limitation within the narrow interlayer space. The parent CH3COO−/NiZn catalysed this epoxidation efficiently, yielding 45% of 2 after 1 h (entry 2). It should be noted that the C3H7COO−/NiZn catalyst exhibited the highest activity, yielding only 2 in 55% yield after 1 h (entry 3).18 On the other hand, the catalysts which have more enlarged interlayer spaces such as C5H11COO−/NiZn, C7H15COO−/NiZn, and C9H19COO−/NiZn did not promote this epoxidation reaction effectively (entries 4–6). It is likely that the substrate could not diffuse smoothly into the interlayer space of these catalysts because these spaces are occupied by the alkyl chains of the guest anions. Ni(OCOCH3)2·4H2O, Zn(OCOCH3)2·2H2O, Ni(OH)2, and Zn(OH)2, the precursors of NiZn and some of the components of the brucite-like layer of NiZn, did not show any significant catalytic activity under the test reaction conditions (entries 7–10). It is well known that Mg–Al hydrotalcite can heterogeneously catalyse the epoxidation of electron-deficient olefins with H2O2.14a–c However, the reaction rate over the Mg6Al2(CO3)(OH)16·4H2O catalyst was slow under the same reaction conditions (entry 11). When C3H7COONa (1.47 mmol) was used as a catalyst under the same reaction conditions, only 35% yield of 2 was obtained after 3 h (entry 12). The epoxidation did not proceed at all without the catalyst (entry 13).
The screening of solvents to determine their effect on the epoxidation of 1 over the C3H7COO−/NiZn catalyst at 50 °C showed that N,N-dimethylformamide (DMF) was superior to the other solvents tested, giving only 2 as a product in 83% yield within 3 h (Table 2, entry 1), whereas N,N-dimethylacetamide (DMA), acetonitrile, methanol, n-hexane, 1,2-dichloroethane, toluene, and water were not effective (entries 2, 4–5, and 7–10). When a combined oxidant of benzonitrile, acetonitrile, or acetamide and hydrogen peroxide was used in the presence of the C3H7COO−/NiZn catalyst, the enone epoxidation did not proceed effectively (entries 3, 4 and 6). The fact that no benzamide or acetamide was formed in nitrile solvents indicates that the active peroxy intermediates from nitrile or amide, which were used as oxidants, were not involved in this catalytic system.19
Entry | Solvent | Conv.b (%) | Yieldb (%) |
---|---|---|---|
a 1 (0.5 mmol), C3H7COO−/NiZn catalyst (0.05 g), solvent (4 mL), 30 wt% H2O2 aq. (2 eq. relative to the substrate), air, 50 °C, 3 h. b Determined by GC using an internal standard technique. c With acetamide (2 mmol). | |||
1 | N,N-Dimethylformamide | 83 | 83 |
2 | N,N-Dimethylacetamide | 39 | 39 |
3 | Benzonitrile | 7 | 5 |
4 | Acetonitrile | 57 | 57 |
5 | Methanol | 38 | 38 |
6c | Methanol | 42 | 38 |
7 | n-Hexane | 39 | 34 |
8 | 1,2-Dichloroethane | 3 | 1 |
9 | Toluene | 9 | 8 |
10 | Water | 68 | 68 |
In order to further demonstrate the requirement of the heterogeneous C3H7COO−/NiZn catalyst, the catalyst was removed by filtration when epoxidation of 1 had reached almost 70% conversion (Fig. 3). After removal of the catalyst, the reaction was monitored for an additional hour, and no additional formation of 2 was observed. These results demonstrate that oxidation proceeds in the NiZn interlayer and that any dissolved anionic species are not involved in this reaction. After the epoxidation was completed, it was possible to easily separate the C3H7COO−/NiZn catalyst from the reaction mixture by simple centrifugation (Fig. 4). From the XRD profile of the recovered C3H7COO−/NiZn catalyst, it was found that the crystallinity of NiZn had decreased and its C.S. expanded from 1.06 nm to 1.36 nm (Fig. 1d) due to the intercalation of water molecules and/or DMF solvent during the reaction.20
Fig. 3 Hot filtration experiment. Reaction conditions were as follows: 1 (0.5 mmol), C3H7COO−/NiZn catalyst (0.1 g), DMF (2 mL), 30 wt% H2O2 aq. (4 eq. relative to the substrate), 60 °C. |
Fig. 4 The photograph of the C3H7COO−/NiZn catalyst (a) in the reaction mixture during catalytic epoxidation of 1 and (b) after phase separation by centrifugation. |
The epoxidation of various cyclic enones using the C3H7COO−/NiZn catalyst was carried out in DMF at 60 °C (Table 3). 2-Cyclopenten-1-one and 1 were successfully oxidised into their corresponding epoxyketones within 3 and 6 h, respectively (entries 1 and 3). The products expected from an aldol-type condensation were scarcely observed under the reaction conditions when performed over the C3H7COO−/NiZn catalyst. The recovered C3H7COO−/NiZn catalyst was found to be reusable without any loss of its activity and selectivity at least four times (Table 1, entry 3 and Table 3, entries 4–7). After the 4th recycling experiment (Table 3, entry 7), any nickel and zinc species and butylate anion leached into the reaction mixture were in amounts small enough to be undetectable by atomic absorption spectroscopy and ion chromatography. The reactivity is affected by various factors such as the orbital energy in LUMO of substrates, contact probability, the formation rate of HOO−, inhibition of side reaction, and so on. Especially, the reactivity of enones can be explained by using the orbital energies in LUMO of the substrates.21 Because the nucleophilic attack of HOO− to the substrate is known as the rate-determining step,22 LUMO energies of substrates are critical to explain the difficulty of the epoxidation of enones. In the case of the epoxidation of β-substituted cyclic enones,23 such as 3-methyl-2-cyclopenten-1-one (LUMO energy: −0.128 eV) and 3-methyl-2-cyclohexen-1-one (−0.121 eV), the reaction rates were decreased (entries 2 and 8). The lower reactivity of these compounds compared with the non-substituted ones (1: −0.116 eV, 2-cyclopenten-1-one: −0.126 eV) might be explained by the steric hindrance about the CC double bond of substrates. The epoxidation of α-isophorone hardly occurs (entry 11) due to the steric hindrance of the methyl group at the β-position and the gem-dimethyl group.24
Entry | Substrate | Product | Time (h) | Conv.b (%) | Yieldb (%) |
---|---|---|---|---|---|
a Substrate (0.5 mmol), C3H7COO−/NiZn catalyst (0.05 g), DMF (2 mL), 30 wt% H2O2 aq. (4 eq. relative to the substrate), 60 °C, air. b Determined by GC using an internal standard technique. c Value in parenthesis is isolated yield. d 1st reuse. e 2nd reuse. f 3rd reuse. g 4th reuse. h Methanol (2 mL) was used as a solvent. | |||||
1 | 3 | 97 | 97 | ||
2 | 12 | 62 | 62 | ||
3 | 6 | 98 | 98 (91)c | ||
4d | 1 | 55 | 55 | ||
5e | 1 | 56 | 55 | ||
6f | 1 | 54 | 54 | ||
7g | 1 | 53 | 52 | ||
8 | 12 | 70 | 70 | ||
9h | 5 | 97 | 96 | ||
10 | 9 | 16 | 11 | ||
11 | 24 | 5 | n.d | ||
12 | 24 | 13 | 10 | ||
13 | 24 | 20 | 20 |
The β,β-disubstituted acyclic enone, (R)-(+)-pulegone, with a high LUMO energy of 0.273 eV, was not transformed into an epoxyketone, even after 24 h. In contrast, the C3H7COO−/NiZn-catalysed epoxidation of 1,4-naphthoquinone, which has a low LUMO orbital energy (−1.529 eV), totally proceeded, and the corresponding product was formed almost quantitatively (entry 9). The high reactivity of the substrates possessing low LUMO orbital energies supports the hypothesis that the catalytic epoxidation process occurs via the nucleophilic attack of a perhydroxyl species. Conversely, 3-nonen-2-one with a high LUMO orbital energy25 gave an extremely low yield of the corresponding epoxyketone (entry 10). In addition, the epoxidation of bulky substrates such as benzalacetone and chalcone resulted in only 10% and 20% yields of 2,3-epoxy-1,3-diphenylpropanone and 2,3-epoxy-1,3-diphenyl propanone, respectively (entries 12 and 13).
Moreover, the C3H7COO−/NiZn-catalyzed epoxidation of 1 with an equimolar amount of H2O2 proceeded in a highly effective manner,26 giving 97% of 2 at 60 °C in 8 h, with 97% efficiency of H2O2 utilization, as highlighted in Scheme 1.27
Based on these results, the most plausible mechanism for C3H7COO−/NiZn-catalysed epoxidation is as follows. A perhydroxyl anion, which is formed by the reaction between the base site of NiZn and H2O2, attacks the β-olefinic carbon atom of an α,β-unsaturated ketone. Following the ring enclosure, the corresponding epoxyketone is formed and anion exchange between OH− and C3H7COO− takes place.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4cy01063a |
This journal is © The Royal Society of Chemistry 2015 |