1. Introduction

The new development of structurally well-defined catalytic systems that emulate nature’s capacity to directly create stereochemical compounds is of great interest in the chemical synthesis era. Au nanoparticles have been proved to be a class of excellent catalysts for alkenes epoxidation implying propene, styrene, or cyclohexene epoxidation processes,¹ however, the development of efficient stereoselective epoxidation still presents a great challenge, despite the steric demands of epoxides from non-terminal alkenes are readily differentiated for pharmaceuticals, agrochemicals, and fine chemicals. Hughes et al. reported the formation of trans- epoxide and cis-epoxide from the selective oxidation of cis-stilbene on Au/C catalyst in the presence of TBHP, whereas its application is only restricted in a small-scale laboratory experiment process by the very low yield (5.5% in toluene as a solvent).² Additionally Lignier et al. found the high activity of Au/TiO₂ in the stereoselective epoxidation of trans-stilbene in methylcyclohexane as a solvent, however the mechanistic description was concentrated on the effect of methylcyclehexane in the free-radical reaction, and made no mention of the catalysis of gold catalyst on trans-stilbene epoxidation.³ Recently we disclosed the effective epoxidition of styrene using atomically precise Au_n(SR)_m nanoclusters as catalysts and proposed a understanding of the nanocatalysis of Au_n nanoclusters at atomic level, wherein each active site on Au_n surface conducts selective access to epoxidation. It is realized that the site-specific catalysis of Au_n nanoclusters can be employed to transfer a variety of non-terminal alkenes for stereoselective epoxidation. Herein, we report the atomic level catalysis of Au_n nanoclusters for efficiently stereoselective epoxidation of cis-stilbene and trans-stilbene.

Such Au_n nanoclusters are composed of metal core with a precisely controlled, specific number of gold atoms, formulated as Au_n(SR)_m, where, n and m represent the respective numbers of gold atoms and thiolate ligands in the particle. Here, atomically monodisperse Au_n(SR)_m nanoclusters (where R stands for −CH₂CH₂Ph), involving Au₂₅(SR)₁₈ (~1 nm in diameter), Au₃₈(SR)₂₄ (1.3 nm in diameter), and Au₁₄₄(SR)₆₀ (1.6 nm in diameter) were synthesized by wet chemistry approaches.^4-7

2. Experimental

2.1. Synthesis of Au_n(SR)_m catalysts (R= CH₂CH₂Ph)

Au₂₅(SR)₁₈ nanoclusters: HAuCl₄·3H₂O (0.4 mmol, dissolved in 5 ml nanopure water), and tetraoctylammonium bromide (TOAB, 0.47 mmol, dissolved in 10 ml toluene), were combined in a 25 ml tri-neck round bottom flask. The solution was vigorously stirred for 15 min, and the aqueous was then removed. The toluene solution of Au (III) was purged with N₂ and cooled down to 273 K in an ice bath over a period of 30 min under constant magnetic stirring. PhCH₂CH₂SH (0.17 ml) was added to the flask, and stirring was reduced to a very low speed (~50 rpm). After the solution turns to clear (~ 1 hr), NaBH₄ (4 mmol, ~7 ml aqueous solution) was rapidly added all at once. The reaction was allowed to proceed overnight under N₂ atmosphere. After aging overnight, ethanol was added to separate Au₂₅ clusters from TOAB and other side-products. The Au₂₅ clusters were collected after removing the supernatant.

Au₃₈(SR)₂₄ nanoclusters: HAuCl₄·3H₂O (0.3 mmol) was first dissolved in methanol (20 mL) and phenylethanethiol (1.2 mmol, PhCH₂CH₂SH) was dissolved in nanopure water (10 mL). The two solutions were then mixed and then the mixture was cooled to ca. 273 K in an ice bath. NaBH₄ (3 mmol, dissolved in 6 mL nanopure water) was added to the suspension under vigorous stirring. After reaction for another hour, the resulting precipitation (Au:SR) was collected by centrifugation and washed with methanol. The above Au:SR clusters were dissolved in H₂O (4 mL). Then acetone (6 mL) and phenylethanethiol (8 mL) were added and the PhCH₂CH₂SH phase forms the upper layer and the aqueous solution of Au:SR clusters forms the bottom layer. The solution was heated to 353 K and allowed to react for several hours under vigorous stirring. After 3 h, the Au clusters were completely transferred to the organic phase. The isolated phenylethanethiol phase was washed thoroughly with excess ethanol and then with acetone to remove excess phenylethanethiol and byproducts. The major fraction containing Au₃₈ clusters was extracted with toluene from the crude product.

Au₁₄₄(SR)₆₀ nanoclusters: Toluene solution (20 mL) of TOAB (0.116 mmol) was mixed with an aqueous solution of HAuCl₄·3H₂O (0.1 mmol, 20 mL). After phase transfer of gold(III) into toluene, the aqueous layer was removed; PhCH₂CH₂SH (1.5 mmol) was added to the toluene solution, followed by stirring for 30 min. The mixture was then cooled to 0 ^oC. NaBH₄ (0.1 mmol, dissolved in 20 mL 0 ^oC nanopure water) was rapidly injected into the solution under vigorous stirring. After reaction overnight, the organic phase was washed thoroughly with water and then rota-evaporated to dryness. The residue was washed with methanol to remove excess PhCH₂CH₂SH, TOAB, and other byproducts. The dried samples (~20 mg) were incubated in pure PhCH₂CH₂SH at 80℃ for 24 hr under vigorous stirring. The final solution was washed thoroughly with methanol and then acetone to remove by-products such as gold(I): PhCH₂CH₂SH complexes. The Au₁₄₄(SCH₂CH₂Ph)₆₀clusters were extracted from the products using a mixed toluene/acetone (1:1, v:v) solvent.

Supported Au_n catalysts: The supported Au_n catalysts were obtained by stirring a solution of Au_n(SR)_m nanoclusters in dichloromethane with a calculated amount of SiO₂ for one day and then the solution was removed in flowing nitrogen. The synthesis of Au_n/CeO₂ is similar with that of Au_n/SiO₂.

2.2. Catalytic Test

Catalytic reaction was carried out at atmospheric pressure. Supported Au_n catalysts (100 mg powder, 1%wt loading Au_n(SR)_m), stilbene (5 mmol), and TBHP (t-butyl hydroperoxide, 17 mmol) were added in a solvent (15 mL) in a 100 mL sealed glass reactor. Then the above mixture was heated to 82℃ under constant vigorous stirring in an oil bath for a certain reaction time. The solvent was removed with rotation evaporation and the products were analyzed by ¹H NMR spectroscopy. The conversion and selectivity were calculated by quantitative NMR.

2.3. Characterization

¹H NMR spectra were measured at Bruker Avance^TM 300 MHz. UV-vis absorption spectra (190-1100 nm) were recorded using a Hewlett-Pachard (HP) 8453 diode array spectrophotometer. Electrospray ionization mass spectra were acquired using a Waters Q-TOF mass spectrometer equipped with Z-spray source.

3. Results and Discussion

These as-prepared Au_n nanoclusters possess a non- crystallographic atom packing structures (Figure 1). In detail, the crystal structure of Au₂₅(SR)₁₈ nanocluster is arranged in six “V-shaped” –S(R)-Au-S(R)-Au-S(R)– staples, based on a centered icosahedral Au₁₃ kernel encapsulated by twelve gold atoms shell.^8,9 The entire particle adopts a quasi-D_2h symmetry and is protected by 18 thiolate ligands. The crystal structure of Au₃₈(SR)₂₄ nanocluster is based on a face-fused biicosahedral Au₂₃ core which adopts a D_3h symmetry. The Au₂₃ core is structurally strengthened by three monomeric staples –S(R)-Au-S(R)– distributed at the waist of Au₂₃ and six dimeric staples –S(R)-Au-S(R)-Au-S(R)– evenly distributed on the two icosahedral units of Au₂₃ core.¹⁰ The atom packing structure of Au₁₄₄(SR)₆₀ nanocluster is theoretically predicted to be arranged in –S(R)-Au-S(R)– staples with an icosahedral Au₁₁₄ core capped by 30 gold atoms shell.¹¹ Such precisely controlled Au_n nanocatalysts offers unique opportunity for investigating nanocatalysis and the correlation of structure with their catalytic properties at the atomic level.

Figure 1. The crystal structures of Au25 and Au38 nanoclusters. (for clarity, only Au and S atoms were shown)

The stereoselective epoxidation of cis-stilbene and trans-stilbene was chosen as a target to evaluate the selectivity and activity of Au_n(SR)_m nanoclusters catalysts, where, TBHP is used as an effective oxidant for epoxidation of olefins. Considering for the commercial interest, such as reuse of catalysts or easy-separation from reaction system, supported Au_n(SR)_m nanocatalysts are investigated on inert silica and non-inert metal oxides, respectively. Cis-stilbene and trans-stilbene as aromatic non-terminal alkenes are not apt to be oxidized compared with terminal alkenes. In the absence of TBHP, no stilbene conversion was observed, when the experiment was carried out using O₂ as an oxidant. In the presence of TBHP, no epoxide products were found without gold catalyst. This indicates that hydroperoxide and gold catalyst are necessary for stilbene epoxidation.

Firstly, Au_n(SR)_m/SiO₂ catalysts were investigated which is based on the indeed identification of gold active species without reference from the interaction of Au_n and supports. Table 1 shows the catalytic performance of Au_n(SR)_m/SiO₂ catalysts for stereoselective epoxidation of cis-stilbene. With Au₂₅(SR)₁₈/SiO₂ catalyst, the selectivity of trans-epoxide up to 98% was observed at 26% conversion of cis-stilbene. The other products, cis-stilbene epoxide and benzaldehyde, have only negligible amounts(~1% selectivities for both). Au₃₈(SR)₂₄/SiO₂ catalyst achieved near 100% selectivity to trans-epoxide and 15% conversion of cis-stilbene, and Au₁₄₄(SR)₆₀/SiO₂ catalyst gave rise to 14% conversion of cis-stilbene and 91% selectivity to trans-epoxide, 2% selectivity to cis-stilbene epoxide and 7% selectivity to benzaldehyde. In Table 1, the catalytic results of Au_n/CeO₂ show that the conversion of cis-stilene is efficiently promoted by the supported Au_n on CeO₂ catalysts. Noted Only CeO₂ support can not convert stilbene. Au₂₅/CeO₂ catalyst achieved 54% conversion of cis-stilbene and 96% selectivity of trans-stilbene epoxide. Au₃₈/CeO₂ catalyst gave rise to 48% conversion with 97% selectivity of trans-stilbene epoxide. Au₁₄₄/CeO₂ catalyst converted 49% cis-stilbene with 97% selectivity toward trans-stilbene epoxide. Test results indicated the important role that CeO₂ support can play cooperating in the activity of suported gold.^12,13 Interestingly, CeO₂ support has no obvious effect on the selectivity of cis-stilbene epoxidation.

With respect to the stereoselective epoxidation of trans- stilbene on supported Au_n catalysts, trans-stilbene epoxide is a major product, benzaldehyde is a second product and cis-stilbene epoxide is a minor product. All these Au_n catalysts exhibited a higher activity for trans-stilbene oxidation that they converted more amount of trans-stilbene comparing to cis-stilbene system (Table 2). As Au_n were supported on SiO₂, 51% conversion with 94% selectivity of trans-stilbene epoxide was obtained on Au₂₅; 43% conversion with 93% selectivity of trans-stilbene epoxide was achieved on Au₃₈; 36% conversion with 93% selectivity of trans-stilbene epoxide was obtained on Au₁₄₄. Au_n/CeO₂ promoted the trans-stilbene oxidation: Au₂₅ catalyzed trans-stilbene into trans-stilbene epoxide in 95% selectivity at 83% conversion; Au₃₈ gave 95% of trans-stilbene epoxide at 81% conversion; Au₁₄₄ achieved 95% selectivity toward trans-stilbene epoxide at 76% conversion.

Table 1. Catalytic results of cis-stilbene epoxidation with supported Aun nanoclusters as catalysts. (100 mg catalysts, 1%wt Aun(SR)m loading, 5 mmol cis-stilbene, 17 mmol TBHP, 15 mL CH3CN as solvent, 82℃, 24 hr.)

Table 2. Catalytic results of trans-stilbene epoxidation catalyzed by supported Aun catalysts. (100 mg catalysts, 1%wt Aun(SR)m loading, 5 mmol trans-stilbene, 17 mmol TBHP, 15 mL CH3CN as solvent, 82 oC, 24 hr.)

Figure 2. The conversions and selectivities of (a) cis-stilbene and (b) trans-stilbene epoxidation on Au25/SiO2 catalyst at different reaction time, respectively

Scheme 1. Proposed mechanism of the stereoselective epoxidation of stilbene on site-specific catalysis of Au25 nanocluster. Golden: Au atoms of the core; blue: Au atoms of the shell. (Where, when R1 is phene and R2 is H, the reagent is cis-stilbene; when R1 is H and R2 is phene, the reagent is trans-stilbene.)

Moreover, the kinetic study of the reaction profile and reaction time suggested the rate of trans-epoxide formation is much higher than that of cis-epoxide due to the sterical hinderance and potential-energy barriers (Figure 2). At the beginning, the formation of trans-stilbene epoxide is earlier than other products. With the reaction time, the worm increase of cis-stilbene epoxide selectivity indicates a rotation about the C−C bond following the formation the two C−O bonds of trans-stilbene epoxide. The formation of benzaldehyde results from the sequential oxidation at high temperature. When the reaction time is longer (>30 h), the increase in the benzaldehyde selectivity and decrease in the epoxide with increasing the reaction time are caused by the overoxidation of C=C double bond of stilbene at high temperature. Over 25 h reaction time, the conversion of stilbenes, including cis-stilbene and trans-stilbene, has no significant increase, indicating that a reaction equilibrium is reached.The specific activity of Au_n nanoclusters can be explained by their core-shell-type geometrical and electronic structures. Scheme 1 presents a proposed mechanism for epoxidation of stilbene catalyzed by Au₂₅ nanocluster. In the first step, the free radical process involving the interaction of TBHP and electron-rich Au₁₃ core generates a hydroperoxy species A, and then species A undergoes one-electron reduction with a loss of H₂O molecule and rearranges to give rise to the peroxo intermediate species B (Au₂₅−O−O). Meanwhile, the presence of partial positive charges on the surface gold atoms (Auδ⁺, 0<δ<1) of the Au₁₂ shell, that is electrophilic Au atoms on the shell, should be responsible to active the nucleophilic C=C double bond of stilbene (species B). Then the peroxidic oxygen is unstable and labile to attack the activated nucleophilic C−C bond through side-by-side interaction on Au₂₅ surface sites, in which a transient species C is formed. When reagent is trans-stilbene (where R₁ is H and R₂ is phene), trans-epoxide is formed through the transfermation of one oxygene to C−C bond and the remove of the other oxygene from species D. When reagent is cis-stilbene, that is, R₁ is phene and R₂ is H, stereoselectivity of trans-epoxide and cis-epoxide is controlled through the difference in the potential-energy barriers of single reaction channel.¹⁴ While one oxygene transfer to olefinic band, R₁−C−R₂ group rotates around the C−C bond since the activation energies of trans-epoxide and cis-epoxide makes the trans-epoxide channel become favored. Therefore, the trans-epoxide product is preferable to form and then Au₂₅ returns to active starting-point for another cycle to trigger the catalytic reaction.

4. Conclusions

In summary, we have developed a highly efficient stereoseletive epoxidation of stilbene using atomically precise Au_n nanoclusters catalysts at an atomic level. The site-specific activity of Au_n nanoclusters is attributed to their core-shell-type geometrical and electronic structures. The site-specific catalysis of Au_n nanoclusters is promising to develop reaction classes that are widely viewed as important within the field of stereoselective catalysis.

Acknowledgements

This work is financially supported by AFSOR, NIOSH and NSF.