Synthesis of sinter-resistant Au@silica catalysts derived from Au₂₅ clusters

Abstract

Gold clusters exhibit remarkable catalytic activity for many reactions such as carbon monoxide oxidation, alcohol, alkene, and hydrocarbon oxidations, and reduction reactions at low temperatures. However, several previous studies show that Au clusters undergo problematic sintering at temperatures above 250 °C, which makes them unsuitable catalysts for high-temperature oxidation reactions. Here we report the coating of Au₂₅(11-MUA)₁₈ clusters (where 11-MUA = mercaptoundecanoic acid) by silica to produce sinter-resistant Au@SiO₂ catalysts. The structure of the resulting materials before and after calcination at temperatures up to 650 °C was followed by TEM and extended X-ray absorption fine structure spectroscopy (EXAFS) analyses, which showed that the Au@SiO₂ catalysts created were much more stable to sintering compared to control materials; with average particles sizes of 2.2 nm after calcination at 250 °C and just over 3 nm after calcination at 650 °C. In addition, we explored the activity of the resulting materials for the 4-nitrophenol reduction and styrene epoxidation reactions; results clearly showed that the Au surfaces are accessible for reactants and that the kinetics of 4-nitrophenol reduction was directly related to the dispersion of the Au particles, as measured via the first shell Au–Au coordination numbers by EXAFS. Styrene epoxidation results show that the Au@SiO₂ materials have excellent activity and recyclability.

---

Introduction

Gold catalysts, especially below 2 nm in size, have attracted much attention in catalysis owing to their well-understood structures and broad catalytic activity.^1–5 Several theoretical studies have shown that small clusters (Au₁₄, Au₂₅, Au₂₈, Au₃₈, etc.) can dissociate molecular oxygen by transferring electron charge to 2π* of dioxygen,^6,7 which is a crucial step in oxidation reactions. For example, ceria supported Au₂₅ clusters have shown moderate catalytic activity for carbon monoxide oxidation at room temperature.⁸ A number groups have shown that Au₂₅(SR)₁₈ (in the form of free clusters or supported clusters) are effective catalysts for oxidation and reduction reactions at moderate temperatures.^9,10 However, activation of pre-synthesized clusters without destroying their integrity remains a significant challenge in the field. Many groups have shown that Au catalysts showed sintering and corresponding drops in activity for supported-Au catalysts after heating at moderate temperatures.^11,12 For example, Haruta et al. observed that the Au supported on TiO₂ catalyst calcined at 500 °C showed lesser catalytic activity for CO oxidation than the catalyst calcined at 200 °C, due to sintering of Au clusters at higher calcination temperatures.¹³ The melting point of metals decreases with particle size,¹⁴ and most metals show significant surface mobility and sintering when approaching their Tammann temperature (half the melting point of a metal, in K).¹⁵ Thermal stability of catalysts is crucial because many industrially important reactions, such as hydrocracking,¹⁶ partial oxidation,¹⁷ and complete combustion¹⁸ are carried out at higher temperatures of 300 °C to 600 °C. Thus due to their thermal instability, application of Au catalysts in industry is restricted.

Previously our group studied the thermal stability of Au₂₅(SC₈H₉)₁₈ clusters on carbon supports by EXAFS and TEM analysis which showed that at temperatures above 200 °C, thiol stabilizers start to decompose and Au clusters begin to sinter.¹⁹ The size of the Au clusters increased up to 1.9 nm and 2.1 nm after heating at 250 °C and 350 °C for 1.5 h, respectively. Other groups have shown similar results on a variety of supports, particularly at high loadings.^20,21 Zhu et al. carried out thermogravimetric analysis (TGA) studies which showed that Au_n(SR)_m begins to lose its stabilizing ligands at 200 °C and decomposes completely at 250 °C.²⁰ Sintering in nanoparticle catalysts is mainly due to the agglomeration of small particles owing to their mobility on supports at a higher temperature. The agglomeration of Au (and other metal) nanoparticles can be prevented by a number of strategies. One popular strategy is to encapsulate the nanoparticles with a metal oxide shell (either core@shell or yolk@shell structures),^22,23 overcoating with metal oxides via atomic vapour deposition,²⁴ or spatially isolating them in mesoporous silica²⁵ or MOF frameworks.²⁶

Nearly all attempts to make core@shell or yolk@shell clusters have involved much larger nanoparticles, which have lower metal surface areas and thus are less practical for catalysis. Two previous attempts have been made to encapsulate Au clusters with a silica shell.^27,28 Pradeep et al. previously synthesized glutathione-protected Au₂₅ clusters encapsulated with silica shell, with many clusters embedded in a silica matrix, but did not study the thermal stability of the resulting materials.²⁷ Devi et al. synthesized silica colloids containing many ca. 1.5 nm Au clusters stabilized by mercaptoalkylammonium chloride stabilizers.²⁸ However, while the resulting particles were catalytically active for H₂O₂ oxidation of a peroxidase substrate, the Au clusters were found to undergo significant sintering at 250 °C thermal treatments, with 90% of the particles above 3 nm in size.

In this paper, we synthesized Au₂₅(11-MUA)₁₈ clusters and then encapsulated the clusters with silica shells. The thermal stability of the resulting materials was probed by TEM and EXAFS analysis. TEM images of Au@SiO₂ catalysts revealed that majority of the particles are still less than 2 nm in size after calcination at 650 °C, although there is some minor sintering of particles to 2–4 nm in size which is due to agglomeration of multiple Au₂₅(11-MUA)₁₈ clusters in single silica particles. In order to show that the Au clusters are much more sinter-resistant in the silica matrix, we coated Au₂₅(11-MUA)₁₈ clusters onto silica spheres and calcined at two different temperatures (250 °C and 650 °C) and compared the samples and their catalytic activity with Au₂₅@SiO₂ catalysts. There was dramatic differences in the catalytic activity of silica-encapsulated vs. the non-encapsulated clusters, which was due to the tremendous sintering seen for non-encapsulated clusters. For encapsulated clusters, only a small reduction of the catalytic activity for 4-nitrophenol reduction reaction over Au₂₅@SiO₂ catalysts was seen upon increased calcination temperature, which correlates well with the slight increase in EXAFS first shell Au–Au coordination numbers. Styrene epoxidation reactions using Au₂₅@SiO₂ particles show that materials calcined at 250 °C retain excellent activity and strong recyclability, while materials calcined at 650 °C significantly outperform their non-encapsulated counterparts

Experimental section

Materials

Hydrogen tetrachloroaurate(III) trihydrate (HAuCl₄·3H₂O, 99.9% on metal basis, Aldrich), 11-mercaptoundecanoic acid (11-MUA, 95%, Sigma-Aldrich), tetraoctylammonium bromide (TOAB, 98%, Aldrich), sodium borohydride (NaBH₄, 98%, EMD), tetraethylorthosilicate (TEOS, 98%, Aldrich), 4-nitrophenol (PNP, 99%, Alfa Aesar), tert-butyl hydroperoxide (TBHP, 70 wt% in H₂O, Sigma-Aldrich), acetonitrile (CH₃CN, 99.9% Fischer Scientific), and styrene (99%, Sigma-Aldrich) were used as received. Ammonia (30%), tetrahydrofuran (THF, high purity) and ethanol (100%) were purchased from EMD, Fischer Scientific and Commercial Alcohol respectively. Milli-Q water was used for synthesis.

Synthesis of Au₂₅(11-MUA)₁₈ clusters

The synthetic procedure for Au₂₅(11-MUA)₁₈ clusters has been documented previously by our group.²⁹ In a typical synthesis, TOAB (333.5 mg, 1.2 eq.) was added to HAuCl₄·3H₂O solution (200.0 mg) in 20 ml THF in a round bottom flask. The solution was stirred until it turned from yellow to orange-red. After that 11-MUA (545.9 mg, 5 eq. in 5 mL THF) was added to the flask and the solution was stirred until it became colourless. The solution was then cooled using an ice bath followed by the addition of ice-cold NaBH₄ (75.66 mg, 4 eq. in 2 mL water) dropwise and the solution was monitored using UV-vis spectroscopy. The addition of NaBH₄ was continued until the characteristic absorption peaks of Au₂₅(11-MUA)₁₈ clusters were seen by UV-vis spectroscopy. The solution was then centrifuged at 10 000 rpm for 2 min, and the precipitate of larger particles was discarded. The remaining Au₂₅(11-MUA)₁₈ cluster solution in THF was then cooled in an ice bath and additional NaBH₄ (28.38 mg, 1.5 eq. in 2 mL water) was added dropwise until a brown precipitate was obtained. The residue was centrifuged and washed with THF twice, followed by re-dissolution in water. Au₂₅(11-MUA)₁₈ clusters were precipitated out with a few drops of dilute acetic acid with an approximate pH of 3. The precipitate was washed with water twice and redissolved in THF. One drop of NaBH₄ (1.5 eq. in 2 mL water) was added to the THF solution, which re-precipitated out Au₂₅(11-MUA)₁₈ clusters. The precipitate was then collected using centrifugation and washed with THF. 10 mg of dried Au₂₅(11-MUA)₁₈ clusters were dispersed in Millipore water (6 mL) and diluted with 25 mL ethanol.

Synthesis of silica encapsulated Au₂₅(11-MUA)₁₈ clusters

The silica encapsulation strategy is depicted in Scheme 1. In order to minimise the formation of multiple clusters in a single silica shell, 1.0 ml of the above Au₂₅(11-MUA)₁₈ cluster solution was diluted with 10 mL ethanol. Then the diluted Au₂₅(11-MUA)₁₈ clusters and ammonia (1.0 mL) were added to a solution of TEOS (10 μL) in ethanol (10 mL) dropwise under moderate stirring. After 12 h of stirring the material was collected by centrifugation and washed several times with ethanol and water. The silica encapsulated clusters are labelled as Au₂₅@SiO₂ in this manuscript for brevity. The Au₂₅@SiO₂ materials were dried at 100 °C and then calcined at different temperatures 250 °C, 350 °C, 450 °C, 550 °C and 650 °C for 3 h in a stream of air, to give the resulting Au@SiO₂ materials.

Scheme 1 General scheme for the synthesis of silica encapsulated Au₂₅(11-MUA)₁₈ clusters.

Synthesis of Au₂₅(MUA)₁₈ clusters on silica spheres (control samples)

Synthesis of silica spheres: to synthesize silica spheres, we followed the same procedure as above, without Au₂₅(11-MUA)₁₈ clusters. In a typical synthesis, 6 mL water is diluted with 25 mL ethanol, and then 1 mL of a water/ethanol mixture is added into the vial and the whole mixture was diluted with 10 mL ethanol. Then this very diluted water/ethanol mixture and ammonia (1 mL) was added to the TEOS (10 μL) and ethanol (10 mL) solution dropwise under moderate stirring. The solution was then kept stirring for 12 h, then collected by centrifugation and washed several times with ethanol and water.

Deposition of Au₂₅(11-MUA)₁₈ clusters on silica spheres: 1.0 ml of the aqueous Au₂₅(11-MUA)₁₈ cluster solution was added to the dried silica sphere sample and stirred. After 2 h stirring, the solvent was evaporated using a rotary evaporator followed by drying on a Schlenk line apparatus. The metal loading was maintained as 2.0 wt% Au. Au₂₅(11-MUA)₁₈ clusters supported on silica spheres were dried at 100 °C and then calcined at two different temperatures 250 °C and 650 °C for 3 h in a stream of air. The non-encapsulated catalysts have been labelled as Au₂₅/SiO₂.

Characterization

UV-vis absorption spectra of the Au₂₅(11-MUA)₁₈ clusters were analysed using a Varian Cary 50 Bio UV-vis spectrometer. The morphology of the Au₂₅@SiO₂ materials was analysed by a HT7700 TEM operating at 100 kV. The size of 200 particles was measured for each sample using ImageJ software to calculate average particle sizes and standard deviations and create size distribution histograms.³⁰ Extended X-ray absorption fine structure (EXAFS) spectroscopic analysis was performed at HXMA beamline 061D-1 (energy range 5–30 keV, resolution 1 × 10⁻⁴ ΔE/E) at the Canadian Light Source. The storage ring electron energy and ring current were 2.9 GeV and 250 mA, respectively. All data was collected in transmission mode. The energy for the Au-L₃ edge (11 919 eV) was selected by using a Si(111) double crystal monochromator with Rh-coated 100 nm long KB mirror. Higher harmonics were removed by detuning the double crystal monochromator. Data fitting was carried out using the Demeter software package.³¹ In order to fit the data, the amplitude reduction factor was fixed at 0.9 for all the data, which was the value obtained from fitting of the Au foil data. Metal loadings of Au₂₅@SiO₂ and Au₂₅/SiO₂ were analysed by a Varian Spectra AA 55 Atomic Absorption Spectroscope. Nitrogen adsorption/desorption isotherms were collected by using a Micromeritics ASAP2020 system (Norcross GA). The BET model was used to calculate the specific surface area of catalysts.

Catalytic activity for 4-nitrophenol reduction

2.0 mL of 0.10 mM 4-nitrophenol and 1.0 mg Au₂₅@SiO₂ were mixed in a quartz cuvette and the UV-vis spectra were collected. After adding ice cold 0.5 mL 0.1 M NaBH₄ in water to the cuvette, the spectrum was analysed immediately, to get the initial concentration of 4-nitrophenolate, and the progress of the reaction was monitored at 2 min time intervals. Catalytic results were reproduced several times for each sample, with errors <5%.

Styrene epoxidation

Catalytic tests were carried out in a 100 mL round bottom flask fitted with a reflux condenser. 20 mg catalyst, 920 μL styrene, and 2.0 mL of the TBHP solution were added to 5.0 mL acetonitrile and then kept for stirring for 24 h at 82 °C.³² The substrate to Au mole ratio is 3920 : 1. Products were analysed by a gas chromatograph (7890A, Agilent Technologies) equipped with a HP-5 column using a flame ionization detector. For recyclability tests, the catalyst was removed from the catalytic mixture by centrifugation, washed with acetone, and dried at 50 °C before reuse.

Results and discussion

UV-vis absorption studies of the as-synthesised Au₂₅(11-MUA)₁₈ clusters in water/ethanol mixtures show three peaks at 680 nm, 446 nm, and 398 nm which are attributed to HOMO–LUMO transitions as shown in Fig. 1.³³ The UV-vis spectrum gives clear evidence for the formation of Au₂₅ clusters. Previously we have also used MALDI-MS analysis to further strengthen this assignment.²⁹ The clusters were copiously purified by taking advantage of the poor solubility of the carboxylate-terminated clusters in THF in basic conditions as well as poor solubility in water under acidic conditions. TEM images (Fig. S1, ESI†) of the Au₂₅(11-MUA)₁₈ clusters show cluster cores with a size of ca. 1 nm in agreement with previous results.³⁴

Fig. 1 UV-vis spectrum of Au₂₅(11-MUA)₁₈ clusters in water/ethanol mixture.

The Au₂₅(11-MUA)₁₈ clusters were then coated with silica via sol–gel synthesis using NH₃ as a catalyst. Au₂₅ cluster solutions were significantly diluted before silica coating to minimize the number of clusters per silica sphere in the final materials and control the final metal loading. Fig. 2 shows TEM images of the as-synthesized silica coated Au₂₅(11-MUA)₁₈ clusters. The TEM images suggest that the clusters are encapsulated in the silica matrix, stronger evidence for encapsulation comes from comparisons of calcinations of encapsulated clusters vs. control samples shown below. It should be noted that many silica spheres contain multiple clusters, although some particles seem to have only a single Au₂₅(11-MUA)₁₈ cluster. From TEM analysis, the average silica sphere diameter is 40 nm and clusters are ca. 1.1 ± 0.3 nm in size (the slightly larger sizes are likely an artefact of the inability to focus on each individual particle in the three-dimensional silica spheres, causing some particles to appear larger as they are out of focus). The Au loading in silica matrix was maintained as 2.0 wt% which was confirmed by atomic absorption spectroscopy (AAS). Control samples were also made which consisted of Au₂₅(11-MUA)₁₈ clusters decorated on the surface of silica spheres, as shown in Fig. 3a.

Fig. 2 TEM image of as-synthesized Au₂₅(11-MUA)₁₈@SiO₂ (inset; enlarged image).

Fig. 3 TEM images of Au₂₅/SiO₂ control samples a) as synthesized, b) calcined at 250 °C, c) calcined at 650 °C.

A major issue with many Au systems is the propensity for Au clusters or nanoparticles to sinter at moderate calcination temperatures; previous studies have shown that temperatures of at least 250 °C are needed to remove thiol stabilizers from the Au cluster surfaces, with even higher temperatures needed to completely oxidize the disulphide biproducts.¹⁹ To investigate the sinter-resistance of the Au₂₅@SiO₂ catalysts, they were calcined at temperatures up to 650 °C followed by TEM analysis. Fig. 4 shows TEM images of the samples upon sintering at temperatures between 250 °C and 650 °C (2.2 ± 1.0 nm at 250 °C and 3.2 ± 2.0 nm at 650 °C); very little growth in the average particle size was seen. Particle size histograms showed a high population of <2 nm sized particles even after calcination at higher temperatures (Fig. 5). As shown in Fig. 3b and c, control samples consisting of Au₂₅(11-MUA)₁₈ clusters on the surface of silica spheres showed a much greater degree of sintering, with average particle sizes of 3.2 ± 1.7 nm at 250 °C and 15.5 ± 10.0 nm at 650 °C. The particle size histogram for the non-encapsulated control samples is shown in Fig. S2 (ESI†), and is quite distinct from the Au@SiO₂ histogram in Fig. 5, with a number of particles >6 nm in size even after calcination at 250 °C, and nearly all particles >5 nm in size after calcination at 650 °C. Thus, it is quite evident that the encapsulation of the Au₂₅ clusters in silica greatly promotes their stability to sintering at higher temperatures.

Fig. 4 TEM images of Au₂₅@SiO₂ calcined at a) 250 °C, b) 350 °C, c) 450 °C, d) 550 °C, e) and f) 650 °C (inset; enlarged image).

Fig. 5 Histogram of Au₂₅@SiO₂ catalysts calcined at different temperatures.

The Au₂₅@SiO₂ materials were also examined by nitrogen adsorption/desorption isotherms, which are shown in Fig. S3,† along with the corresponding BJH pore-size distributions. The BET surface area of the catalyst increased from 23 m² g⁻¹ to 70 m² g⁻¹ after calcination at 250 °C. Calculated BJH pore-size distributions show a broad distribution of mesopores centred at 9 nm after calcination at 250 °C that are not present in the initial as-synthesized sample. Pore volumes of micropores increased from 0.0012 cm³ g⁻¹ to 0.0092 cm³ g⁻¹ after calcination at 250 °C, while mesopore pore volumes increased from 0.051 cm³ g⁻¹ to 0.111 cm³ g⁻¹. We believe the resulting porosity was caused by removal of the MUA as well as residual solvent from the silica spheres. Importantly, the silica spheres have some porosity, which may allow for catalytic accessibility of the Au clusters.

To further understand changes to the Au clusters before and after calcination, Au L₃-edge EXAFS analysis was carried out in transmission mode on the hard X-ray microanalysis beamline (HXMA) at the Canadian Light Source (CLS). Fig. 6 shows the Au L₃ edge EXAFS k-space and phase-corrected R-space spectra of the as-synthesized Au₂₅(11-MUA)₁₈@SiO₂ clusters prior to calcination. The black line represents the experimental Fourier transformed EXAFS spectra, and the red line represents the simulated EXAFS fit for Au₂₅(11-MUA)₁₈ clusters. Others have previously collected X-ray crystallographic data of Au₂₅(SR)₁₈ clusters which have shown that they have a core–shell morphology, in which the core is composed of an Au₁₃ icosahedron, in which the central atom is surrounded by 12 Au atoms while the shell consists of six S–Au–S–Au–S staple motifs.³⁵ Twelve out of the twenty faces of the icosahedron are surrounded by six staple motifs, with sulfur atoms directly attached to 12 Au atoms of the icosahedron core. A multishell fitting approach which has previously been documented by Zhang's group was used to fit the Au–S and three Au–Au contributions of the clusters.³⁶ In order to fit the data, we first fit the Au–S contribution by using Au–S model data obtained from the standard Au₂₅(SR)₁₈ structure.^36,37 After fitting the parameters for Au–S contribution, we fixed those values followed by fitting all first shell Au–Au coordination modes (there are three Au–Au first shell interactions: the first two involve Au atoms in core and the third involves Au staple atoms). Coordination number (CN) values for all Au–Au contributions were fixed based on the crystal structure of the clusters. The final EXAFS fitting parameters are shown in Table 1. The Au–S bond length was found to be 2.31(1) Å which matches well with crystallographic data of Au₂₅L₁₈ clusters in which L = phenylethanethiol (see Table S1, ESI†).³⁸ For the Au first shell fit the Au–Au(core) R value of 2.76(2) Å is attributed to the distance between the central Au atom and the surface Au atom of the icosahedron core and 6 of the Au surface atom pairs, whereas 2.91(6) Å is the bond distance between the other surface Au atom pairs of the icosahedron (Au–Au(surf)). The last Au–Au bond length, 3.3(1) Å is attributed to the distance between the surface Au atoms and the staple Au atoms. These values agree well with those seen for other Au clusters,¹⁹ and thus the EXAFS data fitting clearly shows that the basic core structure of Au₂₅(11-MUA)₁₈ clusters are similar to that of Au₂₅ clusters using other thiolate ligands.

Fig. 6 Au L₃ edge EXAFS fitting in a) k space and b) phase-corrected R space of as-synthesized Au₂₅(11-MUA)₁₈@SiO₂.

Table 1 EXAFS fitting parameters of as-synthesized Au₂₅(11-MUA)₁₈@SiO₂

Type	CN	R/Å	σ ^2/Å^2	E o shift (eV)
Au–S	1.3	2.31(1)	0.001	0.73
Au–Au(core)	1.44	2.76(2)	0.002(2)	2.5(3)
Au–Au(surf)	1.92	2.91(6)	0.01(1)	2.5(3)
Au–Au(staple)	2.88	3.3(1)	0.10(3)	2.5(3)

---

The Au₂₅@SiO₂ materials calcined at different temperatures were also analyzed by EXAFS spectroscopy at the Au L₃ edge. Phase-corrected R-space EXAFS spectra are shown in Fig. 7, while fitted data are found in Table 2. Individual k and phase-corrected R space spectra of Au₂₅@SiO₂ calcined at different temperatures are shown in Fig. S4 and S5 (ESI†) respectively. Au–S contributions are seen around 2.3 Å, while Au–Au contributions are present between 2.8 Å and 3.2 Å. An fcc Au model was used for fitting the single shell Au–Au fit, as some growth in cluster sizes was seen which prevented the use of the previous Au₂₅ model. After calcination at 250 °C, an Au–S contribution was not observed in the data, which suggests that thiol stabilizers are completely removed from the gold surface, which is consistent with previous work using other thiolate ligands.¹⁹ After calcination at 250 °C, the first shell coordination number for the Au–Au (N_Au–Au) contribution is 9.6(7) which suggests that the average number of atoms per particle is ∼300, which works out to a ca. 2.1 nm average particle size.³⁹ As many of the silica spheres contain multiple Au clusters, this is consistent with some sintering of clusters within silica spheres, and is also consistent with TEM analysis above. EXAFS data suggest that on average ca. 12 clusters sinter together by 250 °C. As the calcination temperature increases, the average N_Au–Au slightly increases and the N_Au–Au of catalysts calcined at 350 °C, 450 °C, 550 °C, and 650 °C are 9.8(7), 10.0(7), 10.2(7), and 10.5(6), respectively which suggest that the average number of atoms in the Au particles grows slowly at higher calcination temperatures. Thus, EXAFS results suggest that there is some sintering of particles, albeit from TEM analyses there are still a large number of <2 nm particles at 650 °C.

Fig. 7 Au L₃-edge EXAFS data in phase-corrected R space of Au@SiO₂ catalysts calcined at different temperatures.

Table 2 EXAFS fitting parameters of as-synthesized Au₂₅@SiO₂ catalysts calcined at different temperatures

Catalyst	CN (Au–Au)	R/Å (Au–Au)	σ ^2/Å^2	E o shift (eV) (Au–Au)
Au25@SiO2 calcined at 250 °C	9.6(7)	2.84(1)	0.010(2)	6.1(8)
Au25@SiO2 calcined at 350 °C	9.8(7)	2.843(8)	0.010(1)	5.7(7)
Au25@SiO2 calcined at 450 °C	10.0(7)	2.851(7)	0.009(1)	5.6(7)
Au25@SiO2 calcined at 550 °C	10.2(7)	2.860(8)	0.009(1)	6.7(7)
Au25@SiO2 calcined at 650 °C	10.5(6)	2.856(6)	0.009(1)	6.5(5)

---

Catalytic activity

4-Nitrophenol reduction reaction. One of the major concerns upon coating clusters or nanoparticles with silica overlayers is whether or not the metal surfaces are still catalytically accessible. To probe whether the Au clusters could be accessed prior to and after sintering the Au₂₅@SiO₂ materials, the materials were probed as catalysts for the reduction of 4-nitrophenol. The catalytic activity for the 4-nitrophenol reduction reaction is an excellent gauge of the catalytically accessible surface area of metallic catalysts.⁴⁰ A typical UV-vis spectra reaction profile is shown in Fig. S6;† the original solution has a peak at 317 nm which corresponds to 4-nitrophenol.⁴¹ After adding NaBH₄ to the solution, the peak shifted to 400 nm which indicates the deprotonation of 4-nitrophenol to form nitrophenolate. The intensity of peak due to 4-nitrophenolate decreased as a function of time and a new peak appeared at 300 nm, which has been previously attributed to the formation of 4-aminophenol.⁴² In the absence of Au₂₅@SiO₂ catalysts, no activity for nitrophenol reduction was seen. Table 3 shows the rate constants for 4-nitrophenolate reduction reactions, which are calculated using pseudo-first-order kinetics.^43,44 All materials showed a linear relationship between ln[Ct]/[C_o] and reaction time (min), as shown in Fig. S7.† As the calcination temperature increases, the rate constants for the 4-nitrophenol reduction reaction was slightly reduced from 1.1 × 10⁻¹ min⁻¹ at 250 °C to 1.78 × 10⁻² min⁻¹ at 650 °C, which is likely due to the small increase in Au particle size, and thus decrease in catalytically available surface area. Fig. 8 shows the correlation of reaction rate constant with the coordination number of Au as determined by EXAFS at each calcination temperature; a strong relationship was seen between N_Au–Au and the rate constant for 4-nitrophenol reduction. There is a drop in activity beyond calcination temperatures of 350 °C that cannot be explained by coordination number changes alone; this may be due to partial condensation of the silica matrix which leads to mass-transfer issues.

Table 3 Rate constant k (min⁻¹) for 4-nitrophenol reduction reaction

Catalyst	k (min^−1)
Au25@SiO2 calcined at 250 °C	1.1 × 10^−1
Au25@SiO2 calcined at 350 °C	1.0 × 10^−1
Au25@SiO2 calcined at 450 °C	3.75 × 10^−2
Au25@SiO2 calcined at 550 °C	2.80 × 10^−2
Au25@SiO2 calcined at 650 °C	1.78 × 10^−2
Au25/SiO2 calcined at 250 °C	1.75 × 10^−1
Au25/SiO2 calcined at 650 °C	4.82 × 10^−4

---

Fig. 8 Plot of first shell coordination number of Au@SiO₂ catalysts vs. rate constant for 4-nitrophenol reduction reaction and calcination temperature.

Control samples in which silica spheres were decorated with Au₂₅(11-MUA)₁₈ clusters on the surface, followed by calcination at 250 °C and 650 °C were also examined (Fig. S8†). The encapsulated clusters (e.g. Au₂₅@SiO₂) showed a slightly lower activity than the Au₂₅/SiO₂ control samples upon calcination at 250 °C; this is likely due to mass transfer issues caused by the silica shell.⁴⁵ The rate constant for the control samples for 4-nitrophenol reduction dropped from 1.75 × 10⁻¹ min⁻¹ for samples treated at 250 °C to 4.82 × 10⁻⁴ min⁻¹ for samples treated at 650 °C, which was due to the tremendous sintering of gold clusters on the surface of the silica support as shown earlier by TEM. The Au₂₅@SiO₂ sample calcined at 650 °C had a rate constant that was over 35 times that of the control samples treated at the same temperature. Thus, this cluster encapsulation strategy allows for the formation of well-dispersed, supported-Au catalysts that can withstand high temperature operating and/or calcination conditions. We do note that for mild catalytic reactions such as CO oxidation, Au₂₅ clusters can be activated under mild thermal conditions without significant adverse sintering of the Au cluster catalysts.⁴⁶

Styrene epoxidation. Results for styrene epoxidation reactions over Au₂₅@SiO₂ and Au₂₅/SiO₂ catalysts calcined at two different temperatures (250 °C and 650 °C) are shown in Table 4. The Au₂₅@SiO₂ catalyst calcined at 250 °C showed 70.0% conversion for this reaction, with a selectivity towards styrene oxide (SO) of 92.3% (the other product is benzaldehyde (BA)). This is higher than selectivities seen for larger Au particles,³² but in line with that reported for Au₂₅ clusters supported on hydroxyapatite (HAP) by others.⁴⁷ After calcination at 650 °C, conversion was slightly reduced to 62.3% with similar selectivity for SO, which is consistent with the drop in catalytic activity seen in the 4-nitrophenol reduction reactions above due to the slightly lower surface area of catalysts calcined at higher temperatures. Recyclability studies for Au₂₅@SiO₂ catalysts calcined at 250 °C are shown in Table S2;† after five cycles the catalyst showed the same activity for the reaction. This is significant, as others have shown that Au clusters supported on HAP or silica showed slight deactivation over multiple cycles.²⁰ Au₂₅@SiO₂ catalysts calcined at 250 °C and 650 °C had average turnover frequencies (TOFs) of 116 ± 9 h⁻¹ and 103 ± 10 h⁻¹, respectively. These catalysts also show excellent catalytic activity for the styrene epoxidation reaction, in line with the best results in the literature.^32,47–49 For example, Tsukuda et al. reported that Au₂₅/HAP had an average TOF of 114 h⁻¹.⁴⁷ Wu et al. demonstrated that silica supported Au nanoparticles with an average particle size of 6.4 nm had an average TOF of 99.4 h⁻¹.³² Non-encapsulated catalysts (Au₂₅/SiO₂) calcined at 250 °C had slightly higher conversion of styrene, with very slightly higher selectivity for styrene oxide (>93%); however, conversions dropped to 15.1% upon calcination of the control samples at 650 °C. These results are also consistent with 4-nitrophenol reduction results, such that non-encapsulated catalyst show slightly higher activities at moderate calcination temperatures due to the absence of mass-transfer issues, while higher calcination temperatures lead to tremendous sintering and a corresponding loss of catalyst surface area and activity. In order to compare TONs of the catalysts, we also took total surface Au atoms into account. When accounting for the relative surface areas; samples calcined at 650 °C showed higher adjusted TONs compared to Au₂₅@SiO₂ calcined at 250 °C. This may be due the removal of disulphide by-products from the catalyst at higher temperatures.

Table 4 Conversion (%) and TONs for styrene epoxidation reactions over 24 h

Catalyst	Conversion (%)	Selectivity		TON^a	Adj. TON^b
		SO (%)	BA (%)
a TON = moles of product/moles of Au. b TON = moles of product/moles of surface Au.
Au25@SiO2 calcined at 250 °C	70.0	92.3	7.6	2800	5100
Au25@SiO2 calcined at 650 °C	62.3	91.8	8.3	2500	6500
Au25/SiO2 calcined at 250 °C	75.6	93.7	6.3	3000	8000
Au25/SiO2 calcined at 650 °C	15.1	93.4	6.6	600	7500

---

Conclusions

Herein we have synthesized silica encapsulated Au₂₅(11-MUA)₁₈ clusters, followed by calcination to give active catalysts which have minimal sintering of the Au clusters. To the best of our knowledge, sinter-resistant Au₂₅(11-MUA)₁₈ clusters encapsulated with silica shells have not been documented elsewhere. Even after calcination at 650 °C, the majority of the particles are below 2 nm. Catalytic studies for 4-nitrophenolate reduction reaction and styrene epoxidation reactions over the catalysts show that the active sites on the gold surface are accessible to the reactants. As the calcination temperature increases, the rate constant of the reaction slightly decreased, which is likely due to the increased coordination number of gold atoms, as individual clusters within the silica spheres sinter to form slightly larger particles. Finally, Au₂₅@SiO₂ catalysts calcined at 250 °C showed remarkable catalytic activity, selectivity and recyclability for styrene epoxidation reactions.

Acknowledgements

We would like to acknowledge the National Sciences and Engineering Research Council of Canada (NSERC) for financial support. We thank Ning Chen at the Canadian Light source (CLS) for assistance with XAS measurements. Hiwa Salimi and Leila Dehabadi are acknowledged for AAS and N₂ adsorption analysis respectively. EXAFS experiments described in this paper were performed at the Canadian Light Source, which is supported by the Natural Sciences and Engineering Research Council of Canada, the National Research Council Canada, the Canadian Institutes of Health Research, the Province of Saskatchewan, Western Economic Diversification Canada, and the University of Saskatchewan.

Notes and references

1. Y. Zhu, H. Qian, B. A. Drake and R. Jin, Angew. Chem., Int. Ed., 2010, 49, 1295–1298.
2. S. Sreedhala, V. Sudheeshkumar and C. P. Vinod, Catal. Today, 2015, 244, 177–183.
3. O. Lopez-Acevedo, K. A. Kacprzak, J. Akola and H. Häkkinen, Nat. Chem., 2010, 2, 329–334.
4. M. Haruta and M. Daté, Appl. Catal., A, 2001, 222, 427–437.
5. Y. Zhao, A. Mpela, D. I. Enache, S. H. Taylor, D. Hildebrandt, D. Glasser, G. J. Hutchings, M. P. Atkins and M. S. Scurrell, Stud. Surf. Sci. Catal., 2007, 163, 141–151.
6. B. Yoon, H. Häkkinen and U. Landman, J. Phys. Chem. A, 2003, 107, 4066–4071.
7. A. Sanchez, S. Abbet, U. Heiz, W. D. Schneider, H. Häkkinen, R. N. Barnett and U. Landman, J. Phys. Chem. A, 1999, 103, 9573–9578.
8. X. Nie, H. Qian, Q. Ge, H. Xu and R. Jin, ACS Nano, 2012, 6, 6014–6022.
9. Y. Zhu, H. Qian, M. Zhu and R. Jin, Adv. Mater., 2010, 22, 1915–1920.
10. A. Shivhare, S. J. Ambrose, H. Zhang, R. W. Purves and R. W. J. Scott, Chem. Commun., 2013, 49, 276–278.
11. C. M. Yang, M. Kalwei, F. Schüth and K. J. Chao, Appl. Catal., A, 2003, 254, 289–296.
12. G. M. Veith, A. R. Lupini, S. Rashkeev, S. J. Pennycook, D. R. Mullins, V. Schwartz, C. A. Bridges and N. J. Dudney, J. Catal., 2009, 262, 92–101.
13. F. Boccuzzi, A. Chiorino, M. Manzoli, P. Lu, T. Akita, S. Ichikawa and M. Haruta, J. Catal., 2001, 202, 256–267.
14. P. Buffat and J. P. Borel, Phys. Rev. A, 1976, 13, 2287–2298.
15. J. Lu, B. Fu, M. C. Kung, G. Xiao, J. W. Elam, H. H. Kung and P. C. Stair, Science, 2012, 335, 1205–1208.
16. K. C. Park and S. K. Ihm, Appl. Catal., A, 2000, 203, 201–209.
17. S. S. Bharadwaj and L. D. Schmidt, Fuel Process. Technol., 1995, 42, 109–127.
18. O. Deutschmann, L. I. Maier, U. Riedel, A. H. Stroemman and R. W. Dibble, Catal. Today, 2000, 59, 141–150.
19. A. Shivhare, D. M. Chevrier, R. W. Purves and R. W. J. Scott, J. Phys. Chem. C, 2013, 117, 20007–20016.
20. Y. Zhu, H. Qian and R. Jin, Chem. – Eur. J., 2010, 16, 11455–11462.
21. J. E. Martin, J. Odinek, J. P. Wilcoxon, R. A. Anderson and P. Provencio, J. Phys. Chem. B, 2003, 107, 430–434.
22. S. H. Joo, J. Y. Park, C.-K. Tsung, Y. Yamada, P. Yang and G. A. Somorjai, Nat. Mater., 2009, 8, 126–131.
23. S. Wang, M. Zhang and W. Zhang, ACS Catal., 2011, 1, 207–211.
24. A. Cao, R. Lu and G. Veser, Phys. Chem. Chem. Phys., 2010, 12, 13499–13510.
25. H. Song, R. M. Rioux, J. D. Hoefelmeyer, R. Komor, K. Niesz, M. Grass, P. Yang and G. A. Somorjai, J. Am. Chem. Soc., 2006, 128, 3027–3037.
26. G. Lu, S. Li, Z. Guo, O. K. Farha, B. G. Hauser, X. Qi, Y. Wang, X. Wang, S. Han, X. Liu, J. S. DuChene, H. Zhang, Q. Zhang, X. Chen, J. Ma, S. C. J. Loo, W. D. Wei, Y. Yang, J. T. Hupp and F. Huo, Nat. Chem., 2012, 4, 310–316.
27. M. A. Habeeb Muhammed and T. Pradeep, Small, 2011, 7, 204–208.
28. A. Samanta, B. B. Dhar and R. N. Devi, J. Phys. Chem. C, 2012, 116, 1748–1754.
29. A. Shivhare, L. Wang and R. W. J. Scott, Langmuir, 2015, 31, 1835–1841.
30. X. Liu, M. Atwater, J. Wang and Q. Huo, Colloids Surf., B, 2007, 58, 3–7.
31. B. Ravel and M. Newville, J. Synchrotron Radiat., 2005, 12, 537–541.
32. J. Liu, F. Wang, S. Qi, Z. Gu and G. Wu, New J. Chem., 2013, 37, 769–774.
33. Y. Lu and W. Chen, Chem. Soc. Rev., 2012, 41, 3594–3623.
34. Z. Wu, J. Suhan and R. Jin, J. Mater. Chem., 2009, 19, 622–626.
35. J. Akola, M. Walter, R. L. Whetten, H. Häkkinen and H. Grönbeck, J. Am. Chem. Soc., 2008, 130, 3756–3757.
36. M. A. Macdonald, D. M. Chevrier, P. Zhang, H. Qian and R. Jin, J. Phys. Chem. C, 2011, 115, 15282–15287.
37. G. A. Simms, J. D. Padmos and P. Zhang, J. Chem. Phys., 2009, 131, 214703.
38. M. Zhu, C. M. Aikens, F. J. Hollander, G. C. Schatz and R. Jin, J. Am. Chem. Soc., 2008, 130, 5883–5885.
39. H. G. Fritsche and R. E. Benfield, Z. Phys. D: At., Mol. Clusters, 1993, 26, 15–17.
40. S. Wunder, F. Polzer, Y. Lu, Y. Mei and M. Ballauff, J. Phys. Chem. C, 2010, 114, 8814–8820.
41. F. Dong, W. Guo, S. K. Park and C. S. Ha, Chem. Commun., 2012, 48, 1108–1110.
42. K. Kuroda, T. Ishida and M. Haruta, J. Mol. Catal. A: Chem., 2009, 298, 7–11.
43. M. Schrinner, F. Polzer, Y. Mei, Y. Lu, B. Haupt, M. Ballauff, A. Göldel, M. Drechsler, J. Preussner and U. Glatzel, Macromol. Chem. Phys., 2007, 208, 1542–1547.
44. B. Baruah, G. J. Gabriel, M. J. Akbashev and M. E. Booher, Langmuir, 2013, 29, 4225–4234.
45. J. Lee, J. C. Park and H. Song, Adv. Mater., 2008, 20, 1523–1528.
46. Z. Wu, D. E. Jiang, A. K. P. Mann, D. R. Mullins, Z. A. Qiao, L. F. Allard, C. Zeng, R. Jin and S. H. Overbury, J. Am. Chem. Soc., 2014, 136, 6111–6122.
47. Y. Liu, H. Tsunoyama, T. Akita and T. Tsukuda, Chem. Commun., 2010, 46, 550–552.
48. N. S. Patil, R. Jha, B. S. Uphade, S. K. Bhargava and V. R. Choudhary, Appl. Catal., A, 2004, 275, 87–93.
49. Y. Jin, P. Wang, D. Yin, J. Liu, H. Qiu and N. Yu, Microporous Mesoporous Mater., 2008, 111, 569–576.

---

Footnote

† Electronic supplementary information (ESI) available: TEM images of Au₂₅(SR)₁₈ clusters, particle size histograms, gas adsorption data, and EXAFS fitting plots for calcined samples and nitrophenolate reduction and styrene oxidation catalytic results. See DOI: 10.1039/c6cy01822j

---