Synthesis of MCF-supported AuCo nanoparticle catalysts and the catalytic performance for the CO oxidation reaction

Abstract

The oxidation of carbon monoxide under low temperature is increasingly becoming an important process, and supported gold nanoparticles have generated an immense interest in this field due to their extremely high reactivity. In this paper, we have synthesized MCF-supported AuCo nanoparticles, and through heating the AuCo/MCF in an O₂ atmosphere, we have developed Au–CoO_x heterostructured catalysts for CO oxidation. The structure of the Au–CoO_x/MCF hybrid catalysts was investigated by using a combination of XRD, TEM, HR-TEM, EDX, SEM, XPS and in situ FTIR experiments. Various pretreatment conditions were required to form a highly active and stable Au–CoO_x/MCF catalyst to achieve 100% CO conversion under low temperature. The AuCo/MCF catalyst calcined at 500 °C for 1 h was found to produce the most active and stable catalyst for CO oxidation with the highest activity at a reaction temperature of 30 °C for 15 h on-stream. Furthermore, XRD results of the used Au–CoO_x/MCF catalyst showed its good resistance to sintering during catalytic process. However, by heating the Au–CoO_x/MCF catalyst in H₂ at 400 °C for 1 h to reduce the CoO_x back to Co to reform the AuCo catalyst, it was found that the AuCo/MCF catalyst was much less active for CO oxidation. This was explained by the in situ FTIR results, which showed that CO molecules could be chemisorbed and activated on the Au–CoO_x/MCF catalyst more than on the AuCo/MCF catalyst. It was likely that the increased interfacial contact between the Au and CoO_x formed the most active site on the catalyst and was responsible for the enhanced catalytic properties when compared with pure Au/MCF.

---

1. Introduction

Gold has a full 5d band and a high ionization potential making it a relatively inert material, however, supported Au nanoparticles with sizes smaller than 5 nm exhibit remarkable performance in many potential applications such as environmental catalysis, chemical synthesis, clean energy processing and so on.^1–3 Among them, supported gold catalysts for the oxidation of carbon monoxide (CO) have attracted more and more attention since Haruta and coworkers demonstrated that small dispersed Au particles on TiO₂ created highly active catalysts for the oxidation of CO at low temperatures (200 K).⁴ Haruta showed that Au/TiO₂ synthesized using the deposition–precipitation (DP) method yielded high turnover frequencies for the CO oxidation reaction due to the formation of hemispherical particles that were strongly attached to the support. Most of gold catalysts are prepared by loading gold on metal oxide supports by DP, coprecipitation and impregnation,^5,6 and the reducibility of supports (e.g., TiO₂, Fe₂O₃, and CeO₂) is often considered as a key factor to determine the activity of Au catalysts because of their ability to activate oxygen through oxygen-vacancy defects and charge transfer from the surface of metal oxides to gold nanoparticles.^7,8

As known to all, the catalytic performance of Au catalysts is highly dependent on the Au particle size^9,10 and other factors such as the nature of support materials^11,12 and preparation methods.^13,14 Recent advances in catalysis have shown that merging of metal or oxide phases into closely coupled heterostructured nanoparticles can result in altered catalytic activity, selectivity, and stability.¹⁵ For example, our group has previously explored the use of Au–Fe₃O₄ dumbbell heterostructures on SiO₂ and TiO₂ supports to stabilize gold nanocatalysts and demonstrated the high activities and thermal stability of these catalysts in CO oxidation.¹⁶ Along this line, our group has also dispersed colloidal AuNi particles onto SiO₂ supports and demonstrated that upon oxidative pretreatment at elevated temperatures, AuNi nanoparticles could transform into coupled Au–NiO heterostructures that were active and stable catalysts for low-temperature CO oxidation.¹⁷ Through some experiments it was reasoned that the Au nanoparticles were decorated with small NiO particles, and the interface between the two materials helped to enhance the overall catalytic properties. In a separate study, intermetallic AuSn nanoparticles were also found to be stable and active for CO oxidation after undergoing oxidative pretreatment to form Au–SnO₂.¹⁸ It is believed that this phase segregation process forms a close interaction between the Au nanoparticle and the respective oxide, which can stabilize the catalyst and prevent sintering and thus preserve the small particle size of gold. Recently, our group studied the catalytic properties of Au–Cu alloy nanoparticles and determined that their most active form for the CO oxidation reaction was Au–CuO_x,¹⁹ and after calcinations at 300 °C, complete CO conversion at room temperature could be obtained. However, in conjunction with FTIR, it was concluded that when Cu was alloyed with Au, the catalyst was inactive. Similar findings were obtained by Liu et al., who additionally showed that the selective oxidation of CO in the presence of H₂ was greatly improved with the Au–CuO_x catalyst.²⁰

Excellent catalytic performance over supported Au–Cu, Au–Ni, and Au–Fe nanocatalysts has been observed, in which the Au-oxide ensembles have been identified as the active structure as well. Furthermore, in this work, we used the Au–Co alloy as an example to demonstrate that an oxidative dealloying process to phase separate the AuCo alloy into a Au–CoO_x heterostructure was a viable method to design more complex nanoparticle catalysts. The catalytic properties of the Au–CoO_x heterostructure were examined to study the catalytic affects on activity and stability for the CO oxidation. The ability to use mesoporous cellular foam silica (MCF) as a catalyst support is attractive because of its high surface-area, thermal stability, mesoporous structure and relatively inert,²¹ which will have little influence on the catalytic activity of the AuCo alloy. However, the high pH required to hydrolyze HAuCl₄·3H₂O to Au(OH)₄ and the low isoelectric point of silica causes a weak interaction with Au, and traditional DP methods often lead to particle aggregation and large catalytically inactive particles.^22,23 So, we reported a previously method to convert supported nanoparticles into an intermetallic nanocrystal catalyst by solution synthesis.¹⁹

In this paper, gold nanoparticles were first supported on MCF using a modified DP method in which a cationic Au species (Au(ethylenediamine)₂Cl₃) strongly interacted with the negative SiO₂ surface at a high pH to form small and stable Au particles. The MCF supported Au nanoparticles were then added to a cobalt acetate solution and under the appropriate conditions Co²⁺ reduced to Co⁰ and diffused into the supported gold nanoparticles to form a partially alloy structure. The structure of these gold catalysts were probed by powder X-ray diffraction (XRD), transmission electron microscopy (TEM), high-resolution TEM (HR-TEM), scanning electron microscopy (SEM), EDX (energy dispersive X-ray spectrometry), X-ray photoelectron spectroscopy (XPS), thermogravimetry (TG) and in situ FTIR. The catalytic properties of these catalysts were evaluated by carbon monoxide oxidation reaction. The effects of treatment conditions on the catalytic performance were investigated, and the optimum activity has been observed by maximizing the density of boundaries between CoO_x patches and Au surfaces, which could be achieved through suitable redox pretreatments of the AuCo/MCF catalysts. This paper aims to study the synthesis and pretreatment conditions necessary for the activation and stability of the Au–CoO_x/MCF catalyst for the oxidation of carbon monoxide.

2. Experimental

2.1 Materials

All chemicals were used as received and purchased from Aldrich unless otherwise stated: cobalt(II) acetate, 98%; gold(III) chloride trihydrate, ACS reagent; 1-octadecene, tech, 90%; oleylamine, approximate C18 content 80–90%; oleic acid technical grade, 90%; ethylenediamine, Reagentplus, >99%.

2.2 Synthesis of Au(en)₂Cl₃ (en = ethylenediamine)

To synthesize the Au(en)₂Cl₃ precursor, ethylenediamine (0.45 mL) was slowly added to an aqueous solution of HAuCl₄·3H₂O (1.0 g in 10.0 mL of DI H₂O) to form a transparent brown solution. After stirring for 30 min, 70.0 mL of ethanol was added to induce precipitation. The final product was centrifuged, washed in ethanol, and dried overnight.

2.3 Synthesis of Au/MCF catalyst

The mesoporous material host (MCF) was synthesized as described by Schmidt-Winkel et al.²⁴ According to our previously reported procedure,²⁵ 80 mg of Au(en)₂Cl₃ was dissolved in 100 mL of DI H₂O to make 3.6 wt% Au loading on MCF. A 1.0 M solution of NaOH was added dropwise to raise the pH to 10.5. A 1.0 g MCF was added, and the pH rapidly decreased. Over the next 30 min, 1.0 M NaOH solution was added dropwise to maintain the pH at 10.5. The mixture was then transferred to a 60 °C water bath for 2 h. The final product was collected by centrifugation, washed in H₂O, dispersed by a vortexer and centrifuged four times. The yellowish product was dried in a vacuum oven for 5 h at 70 °C and reduced at 150 °C in 10% H₂/Ar for 1.0 h to obtain a red powder.

2.4 Synthesis of AuCo/MCF catalyst

According to our previously reported procedure of AuCu alloy,¹⁹ AuCo alloy nanoparticles supported on MCF were prepared by first dissolving Co(C₂H₃O₂)₂ (0.05 mmol) into 1-octadecene (20 mL), oleic acid (1.68 mmol), and oleylamine (1.68 mmol) in a 100 mL, three-neck, round-bottom flask. Then the as-prepared Au/MCF (180 mg) was added to the solution, and the mixture was magnetically stirred under flowing Ar gas. The temperature was first raised to 120 °C for 20 min to remove water, and then increased to 305 °C for 1.5 h. After that, the heating mantle was removed, and the reaction mixture was cooled to room temperature, diluted in ethanol, and centrifuged at 7500 rpm for 7 min. The final product was washed by suspending the powder in ethanol and centrifuging four times before drying in air.

The as-prepared AuCo/MCF sample was calcined in 10% O₂/Ar at different temperatures (200, 300, 400, and 500 °C) for 1 h, or at 400 °C 10% O₂/Ar for 1 h followed by 400 °C 10% H₂/Ar for 1 h. Then, different catalysts were obtained, and the final Au loading of all the samples was about 2.0 wt% with the molar ratio of Au to Co almost 1 : 1, based on inductively coupled-plasma (ICP) analysis.

2.5 Characterization of catalysts

The powder XRD data were recorded with a PANalytical Empyrean diffractometer, operated at 45 kV and 40 mA. TEM and HR-TEM experiments were conducted using HITACH HD2000 microscopes with an accelerating voltage of 200 kV. The surface morphology and chemical composition were studied by SEM and EDX using a SU8010 instrument from Hitachi, Japan at 15 kV. Elemental analysis of the samples was done by inductively coupled-plasma atomic emission spectroscopy using an Optima 2100 DV spectrometer (PerkinElmer Corporation). XPS data were collected using a PHI 3056 spectrometer with an Al anode source operated at 15 kV and an applied power of 350 W. Adventitious carbon and Si⁴⁺ from the SiO₂ were used to calibrate the binding energy shifts of the sample (C 1s 284.8 eV, Si⁴⁺ 103.7 eV). Thermogravimetry was conducted on NETZSCH TG 209F3 instrument, performed under air atmosphere with a heating rate of 20 °C min⁻¹ from room temperature to 800 °C.

An in situ FTIR study was conducted to study CO adsorption on a diffuse reflectance cell (HC-900, Pike Technologies, cell volume of about 6 cm³) which was used in a Nicolet Nexus 670 FTIR spectrometer with a MCT/A detector with a spectral resolution of 4 cm⁻¹. Approximately, 10.0 mg of catalyst was used and after the desired pretreatments, a background spectrum was collected from the sample before CO adsorption at room temperature using 256 scans and 4 cm⁻¹ resolution. Diffuse reflectance FTIR (DRIFTS) spectra were collected at room temperature and obtained by subtracting the background spectrum from subsequent spectra.

2.6 Catalytic experiments

Prior to the catalytic test, the catalysts were pretreated at different temperatures (200, 300, 400 or 500 °C) for 1 h in 10% O₂/Ar or at 400 °C in 10% O₂/Ar, followed by 400 °C in 10% H₂/Ar for 1 h. Catalytic CO oxidation was carried out in a fixed-bed reactor (U-type quartz tube with inner diameter of 4 mm) at atmospheric pressure. Typically, a 20 mg catalyst supported by quartz wool was loaded in the reactor. The feed gas of 1% CO balanced with dry air (<4 ppm water) passed though the catalyst bed at a flow rate of 10 mL min⁻¹ corresponding to gas hourly space velocity (GHSV) of 30 000 mL (h g_cat)⁻¹. The concentrations of CO and CO₂ in the reactor effluent were analyzed by a Buck Scientific 910 gas chromatograph equipped with a dual molecular sieve/porous polymer column (Alltech CTR1) and a thermal conductivity detector. The reaction temperature was controlled using a tubular furnace for temperatures higher than 20 °C or by immersing the reactor in a dry ice-acetone cold trap. The data were reported as light-off curves displaying CO conversion as a function of reaction temperature. In a typical steady-state experiment, the system was adjusted to the 30 °C for 15 h, and the activity was recorded during the whole process.

3. Results and discussion

3.1 Characterization of catalysts

3.1.1 XRD. The phase structures of as-synthesized Au/MCF and AuCo/MCF catalysts were characterized by wide-angle XRD. From Fig. 1, the broad diffraction peak at about 23° in the XRD patterns of all the samples are attributed to the MCF (SiO₂). The face-centered cubic (fcc) structure of gold in Au/MCF after calcination at 500 °C for 1 h was shown in the XRD pattern in the Fig. 1a, and the Au diffraction peaks were located near 38°, 45°, 65° and 78° corresponded to (111), (200), (220) and (311) crystal planes²⁶ with the most intense peak (111) located at 2θ = 38.2°. When Co diffused into the Au particles, the fcc structure remained, but the Au lattice contracted to a certain extent, and the XRD peaks shifted to the right, as seen in the XRD patterns in Fig. 1b. Though these features indicated the characteristics of a AuCo intermetallic structure, the intensities of these particular peaks were weak and indicated that only a partially ordered alloy was formed.²⁷ Thus, the observed small shift in our case suggested that only a small part of the surface Co species has been alloyed with Au nanoparticles. In addition, neither Co nor CoO_x phases were detected in the AuCo/MCF catalysts, suggesting that Co might exist as highly dispersed species. On the other hand, in Fig. 1, the average sizes of Au nanoparticles for all the samples was very small (4–7 nm) using Scherrer's equation, and this suggested that metallic Au nanoparticles with a small particle size were formed on the MCF support even after high-temperature treatment conditions despite that the Au loading of all the samples was about 2.0 wt%, based on ICP analysis.

Fig. 1 XRD patterns of catalysts ((a) Au/MCF + O₂ (500 °C, 1 h), (b) AuCo/MCF-before calcinations, (c) AuCo/MCF + O₂ (200 °C, 1 h), (d) AuCo/MCF + O₂ (300 °C, 1 h), (e) AuCo/MCF + O₂ (400 °C, 1 h), (f) AuCo/MCF + O₂ (500 °C, 1 h), (g) AuCo/MCF + O₂ (400 °C, 1 h) + H₂ (400 °C, 1 h)).

To form the Au–CoO_x heterostructures, the AuCo alloy catalysts were oxidized at different temperatures. Fig. 1 showed that when the catalyst AuCo/MCF was heated above 200 °C the peaks shifted toward smaller angles again and they reached the positions that corresponded to pure gold at 38.2°, signifying that the Co was oxidized and segregated from the alloy structure to slowly form a composite of Au and amorphous CoO_x. This observation was in agreement with our previous studies that indicated that AuCu/SiO₂ alloy catalysts began to oxidize above 150 °C to form Au–CuO_x structures.¹⁹ However, when the oxidized Au–CoO_x nanoparticles were exposed to a 10% H₂/Ar atmosphere at 400 °C to reduce the CoO_x again, the XRD pattern in Fig. 1g showed that the (111) peak of Au shifted towards a higher angle indicating that Co partially diffused into Au to form the alloy again. The reversible structural changes in the AuCo/MCF with response to the redox treatments clearly showed that the surface Co atoms could diffuse inside of Au nanoparticles when annealing in a H₂ atmosphere, while the alloyed Co atoms might segregate to the particle surfaces driven by the oxidation treatment, which was consistent with previous reports of AuCu nanoparticles.²⁸

3.1.2 TEM and HR-TEM. TEM technique is one of the most direct and efficient approaches for the detection and analysis of nano-sized particles. The TEM images of representative 500 °C – pretreated Au/MCF, 500 °C – pretreated AuCo/MCF and 300 °C – pretreated AuCo/MCF catalysts were illustrated in Fig. 2A, B and C, respectively. The HR-TEM images of representative samples such as AuCo/MCF-before calcinations and AuCo/MCF + O₂ (500 °C, 1 h) were exhibited in Fig. 2D and E, respectively. The results of Fig. 2 showed that gold metal nanoparticles in the catalysts exhibited small particle size and narrow size distribution, which corresponded to the broad XRD diffraction bands in Fig. 1. Previous studies of Au–Cu/SiO₂ alloy nanoparticles revealed that oxidizing the system to form Au–CuO_x improved the stability and reduced particle sintering, even after calcination at high temperatures.¹⁹ Thus, the resistance to particle sintering and the stability of the catalyst was examined by TEM after the catalysts were subjected to calcinations pretreatments at 300 °C or 500 °C.

Fig. 2 TEM and HR-TEM results of catalysts (TEM: (A) Au/MCF + O₂ (500 °C, 1 h); (B) AuCo/MCF + O₂ (500 °C, 1 h); (C) AuCo/MCF + O₂ (300 °C, 1 h) HR-TEM: (D) AuCo/MCF-before calcinations; (E) AuCo/MCF + O₂ (500 °C, 1 h)).

Fig. 2A and B showed TEM images of as-synthesized Au/MCF and AuCo/MCF after calcination in O₂/Ar at 500 °C for 1 h, respectively. The average size of Au nanoparticles for the Au/MCF catalyst was about 6.5 nm in diameter after calcination at 500 °C. Interestingly, for the AuCo/MCF catalysts, shown in Fig. 2B and C, the mean size appeared to have a smaller diameter as well as showed a tighter size distribution after the same treatment. In the case of the AuCo catalyst, a smaller average particle diameter 4.6 nm was observed after calcination at 500 °C (Fig. 2B), which was almost the same as the average particle diameter observed after calcinations at 300 °C (4.5 nm, Fig. 2C). It indicated that the high temperature treatment had little influence on the size of gold nanoparticles.

In order to confirm the Au–CoO_x interaction, HR-TEM was done for the representative samples. Fig. 2D showed a HR-TEM image of a typical AuCo alloy particle before calcinations, in which the Au and Co atoms are homogeneously distributed throughout the nanocrystal. While after oxidation at 500 °C in O₂ for 1 h, a thin CoO_x shell was formed around the Au particle to create a core–shell structure, which was indicated by the less intense contrast layer, as seen in Fig. 2E. As shown in Fig. 8 of stability tests, when the Au–CoO_x structure is formed after calcinations in O₂, little to no deactivation is observed, and the catalyst remains stable for 15 h period. It is due to the fact that CoO_x has a stronger interaction with silica than Au, which can anchor Au particles to the MCF support, preventing sintering of the Au nanoparticles at elevated temperatures and keeping the particles in an optimal size range to efficiently catalyze the CO oxidation reaction.

3.1.3 SEM and EDX. To discuss the Au–CoO_x structure further, SEM and EDX to a representative sample, 500 °C – oxidized Au–CoO_x/MCF that showed the best performance for CO oxidation, were conducted, as shown in Fig. 3. Although we cannot distinguish the micrograph of Au from that of CoO_x because of the large size in SEM image, the EDX results of different pots on the same SEM image showed that the Au and Co atoms are not homogeneously distributed throughout the crystal. In the Fig. 3A, the atom ratio of Au to Co is 4.70, meaning that Au atom is much more than Co atom in this area, while in the Fig. 3B, the atom ratio of Au to Co is 0.43, suggesting that Co atom is much more than Au atom in this area. So it can be concluded that AuCo alloy in the sample of AuCo/MCF-before calcinations has separated to Au and CoO_x after treatment in 500 °C O₂ for 1 h. Actually, the results of XPS in Fig. 4 gave the evidence of Co(0) changing to Co(x+) after calcinations in O₂, and the XRD results in Fig. 1 also indicated that Au is metallic and CoO_x is amorphous for the samples after calcinations in O₂ under high temperature because no CoO_x peaks were observed in XRD patterns.

Fig. 3 SEM and EDX results of AuCo/MCF + O₂ (500 °C, 1 h) (A and B are the different pots on the same SEM image).

Fig. 4 XPS results of catalysts ((a) AuCo/MCF + O₂ (400 °C, 1 h); (b) AuCo/MCF + O₂ (400 °C, 1 h) + H₂ (400 °C, 1 h)).

3.1.4 XPS. XPS was used to study the influence of different treatment conditions on the chemical states of the MCF supported AuCo catalysts, and the Au 4f and Co 2p XPS spectra of AuCo/MCF + O₂ (400 °C, 1 h) and AuCo/MCF + O₂ (400 °C, 1 h) + H₂ (400 °C, 1 h) catalysts were shown in Fig. 4a and b, respectively. In the investigated AuCo/MCF catalysts, the Au components always kept metallic with Au 4f located at 83.9 eV, irrespective of the various pretreatments, which was consistent with the formation of metallic Au⁰ species in the two catalysts.²⁹ The slight shift in binding energy, compared to metallic gold foil (84.0 eV) was probably due to the well documented initial and final state effects that occurred with a reduced metal coordination for the surface gold atoms.

However, reduction of the Au–CoO_x/MCF catalyst in 10% H₂/Ar for 1 h changed the Co 2p spectrum slightly during the process. The data in Fig. 4a showed a Co²⁺/Co³⁺ – like species (782.1 and 786.5 eV) in the 400 °C O₂/Ar annealed sample and a shift to a more reduced form (777.9 and 779.8 eV) after the AuCo/MCF + O₂ (400 °C, 1 h) catalyst further calcined in 400 °C H₂/Ar for 1 h (Fig. 4b). This result indicated that after calcinations in O₂, the main peaks of Co 2p shifted to 782.1 and 786.5 eV, corresponding to the CoO_x structure, and after further calcinations in H₂, Co⁰ species (777.9 and 779.8 eV) were mostly presented in the reduced samples with part survival of CoO_x structure (783.7 eV in Fig. 4b). The reported works on Pt–Ni/carbon black catalysts have shown that Ni species at the surfaces of support and Pt nanoparticles were prone to getting oxidized when exposed to air, while the Ni atoms at the subsurface regions of Pt nanoparticles remained metallic due to the protection of surface noble metal overlayers.³⁰ Therefore, we suggested that the observed CoO_x in the AuCo catalysts after calcination in O₂/Ar should be from the Co species on the surfaces of MCF support or Au nanoparticles, and the metallic Co species were due to alloyed Co atoms inside of Au nanoparticles, which were free from oxidation in the air. It was reported in literature that the CuO_x–Au interface that formed under a higher calcination temperature had an important influence on the dehydrogenation of ethanol in an oxygen-deficient environment,²⁸ so it was suggested that the CoO_x–Au interface that formed from higher temperature calcinations would probably played a key role in the oxidation of carbon monoxide.

3.1.5 TG. To put our results in better perspective, we compared the MCF supported AuCo catalysts with different calcination treatment. As shown in the part of catalytic performance to the oxidation of carbon monoxide (seeing 3.2), AuCo/MCF showed very low activity when pretreated at 200 °C O₂/Ar for 1 h, possibly because such a temperature was not high enough to burn off organic species used during the process of catalyst preparation, which was proven by the TG data shown in Fig. 5. To activate the catalyst, we also pretreated AuCo/MCF at 300 °C, 400 °C and 500 °C in O₂/Ar for 1 h, and the TG data in Fig. 5 showed that oxidative pretreatment at 500 °C was sufficient to remove almost all the organic species. It was probably one of the reasons that the 500 °C – pretreated AuCo/MCF catalyst was the most active for the CO oxidation reaction (seeing 3.2).

Fig. 5 TG result of AuCo/MCF before calcinations.

3.1.6 In situ FTIR. In situ CO absorbed FTIR spectroscopy was conducted to investigate the structure change of AuCo/MCF with the redox treatments and to clarify the catalytic behavior observed in Fig. 7 (seeing 3.2). It showed that when the alloyed AuCo/MCF catalyst was heated in 10% O₂/Ar at 250 °C, 300 °C, and 350 °C, two broad absorption components near 2168 and 2134 cm⁻¹ were observed (Fig. 6d–f). These results similarly agreed with the reported FTIR spectra of gas phase CO peak (2168 cm⁻¹)¹⁹ and CO absorbed onto oxidized Co species (2134 cm⁻¹).³¹ It was noted that with the increasing of treatment temperature, the two peaks both became stronger and stronger, especially the peak near 2134 cm⁻¹. This indicated that the higher treatment temperature in O₂/Ar resulted in the more changing of Co to CoO_x, and then the peak attributed to the CO absorbed onto oxidized Co species became stronger.

Fig. 6 In situ FTIR results of AuCo/MCF catalysts with different treatments ((a) 350 °C O₂ + 400 °C H₂; (b) 350 °C O₂ + 350 °C H₂; (c) 350 °C O₂ + 400 °C H₂ + 400 °C O₂; (d) 250 °C O₂; (e) 300 °C O₂; (f) 350 °C O₂).

After the sample was oxidized at 350 °C, it was then treated with 10% H₂/Ar at 350 °C or 400 °C for 1 h. As shown earlier in our group, CuO_x was reduced under these conditions and the Cu diffused back into the Au particles to reform the weakly ordered intermetallic phase,¹⁹ and exposure of CO at room temperature upon this reduced AuCu/SiO₂ resulted in weak adsorbed CO. Similarly, the reduced samples AuCo/MCF also exhibited the phenomena that the adsorbed CO bands obviously became weaker after further treatment of Au–CoO_x/MCF with 10% H₂/Ar at 350 or 400 °C, and CO absorbed onto metallic Au (2118 cm⁻¹)³¹ were seen with the disappearance of the peak of CO absorbed onto oxidized Co species (Fig. 6a and b). The broad features observed were probably due to adsorbed CO that was accompanied by gas phase CO peaks that also appeared near 2168 cm⁻¹ or near 2115 cm⁻¹ that underlied the surface CO bond.³² However, as shown in Fig. 6c, the band near 2134 cm⁻¹ characteristic of surface adsorbed CO onto oxidized Co species appeared much obviously and strongly after the reduced AuCo/MCF was calcined in 10% O₂/Ar at 400 °C again for 1 h to form a phase segregated Au–CoO_x catalyst.

Therefore, the FTIR spectra in Fig. 6 indicated that CO would not sufficiently adsorb onto AuCo alloy nanoparticles, but would easier adsorb onto the Au–CoO_x/MCF materials, and this fact was responsible for the different activities of different catalysts shown in Fig. 7 (seeing 3.2). It should be pointed out that the surfactant at the MCF surface may affect the structural change with the treatment, and if the as-prepared AuCo/MCF sample was not annealed in O₂ without initial calcinations, the catalyst presents quite low activity to CO oxidation.

Fig. 7 Catalytic activities of AuCo/MCF samples with different pretreatments ((a) Au/MCF + O₂ (500 °C, 1 h); (b) AuCo/MCF-before calcinations; (c) AuCo/MCF + O₂ (200 °C, 1 h); (d) AuCo/MCF + O₂ (300 °C, 1 h); (e) AuCo/MCF + O₂ (400 °C, 1 h); (f) AuCo/MCF + O₂ (500 °C, 1 h); (g) AuCo/MCF + O₂ (400 °C, 1 h) + H₂ (400 °C, 1 h)).

3.2 Catalytic performance for the CO oxidation reaction

3.2.1 Effect of calcination conditions on the catalytic activity. To run gas-phase catalytic reactions, it is necessary for the pretreatment of catalysts at elevated temperatures to remove adsorbed water and volatile contaminants, and avoiding the simultaneous growth of Au nanoparticles during the high-temperature process is essential for the supported Au catalysts.³³ To investigate how AuCo alloy and the Au–CoO_x heterostructure affected the catalytic activity compared with Au/MCF, the oxidation of carbon monoxide reaction was carried out, and the light-off curves of AuCo/MCF catalysts under different pretreatment conditions were reported in Fig. 7.

As seen in Fig. 7a, Au/MCF calcined in 500 °C O₂ for 1 h began to catalyze the oxidation of CO at room temperature, however, after reaching its maximum conversion of 80% at 50 °C it began to deactivate as the temperature increased. The AuCo/MCF catalyst calcined at 200 °C displayed T₅₀ (temperature required for 50% CO conversion) at approximately 260 °C, which was almost the same as that of AuCo/MCF catalyst without any pretreatment (T₅₀ = 265 °C). The incomplete removal of organic surfactants under low-temperature calcinations which was already demonstrated in TG result might hinder access to the catalytically active sites and be partly responsible for the latent CO oxidation activity. Interestingly, with the increasing of calcinations temperature to 300 °C, the catalytic activity of CO oxidation was increased accordingly and the T₅₀ decreased to 150 °C. It was worthy to note that when the samples were calcined at 400 and 500 °C, the T₁₀₀ (temperature required for 100% CO conversion) occurred at dramatically lower temperatures, which was 180 and 102 °C, respectively (shown in Fig. 7e and f). Furthermore, in the case of the AuCo/MCF sample that was calcined at 400 °C O₂ to remove the organic surfactants to produce Au–CoO_x/MCF catalyst followed by 400 °C H₂ to obtain the intermetallic AuCo nanoparticle structure, it could be observed that the catalytic activity obviously decreased (T₁₀₀ = 240 °C) in comparison of that of AuCo/MCF catalyst calcined in 400 °C O₂ (T₁₀₀ = 180 °C). In summary, the Au–CoO_x/MCF catalysts after O₂ calcinations were significantly more active for the CO oxidation reaction, as compared with AuCo/MCF catalyst without any pretreatment.

As shown in Fig. 7a, the pure Au/MCF catalyst exhibited a U-shape dependence of activity with the reaction temperature and in fact, the pure gold catalyst never reached complete CO conversion before 320 °C, which was similar to the previous results.³⁴ It has been reported that CO oxidation over Au nanoparticles occurs via low-temperature and high-temperature reaction mechanisms. In the low-temperature range, CO molecules mainly reacted with weakly absorbed O₂ or surface hydroxyl groups. An increase of reaction temperature decreased the coverage of the weakly absorbed oxygen-containing surface species and, consequently, lowered the catalytic activity. In contrast, O₂ could dissociate on the Au surface in the high temperature region, in which CO oxidation activity showed a normal Arrhenius-type behavior. Thus, a U-shape reaction curve was exhibited with the transition from the low-temperature mechanism to the high-temperature mechanism.³⁵

The as-synthesized AuCo/MCF catalyst before calcinations was not active for CO oxidation when the reaction temperature was below 230 °C, and the conversion increased sharply between 250 and 270 °C. This observation indicated that the activation phenomenon was due to the combustion and desorption of organic species when the temperature was higher. The removal of residual organic residue was confirmed by TG analysis (Fig. 5), which showed a large decrease in weight between 250 and 400 °C. Therefore, it was necessary to pretreat the catalysts under higher temperature in order to remove residual organic species. In addition, it was apparent from Fig. 7 that the MCF-supported Au–CoO_x catalysts after calcinations under O₂ showed a higher rate of CO oxidation than the AuCo/MCF catalyst before calcinations. Similar observations were also found in Au@Fe₂O₃ core@shell nanoparticles for the oxidation of CO.³⁶ From previous studies in our group,²⁸ it was known that the AuCu alloy was resistant to oxidation up to 150 °C in air before it started to phase-segregate into Au–CuO_x, and by the time 100% conversion was reached, the Au–CuO_x structure was already formed.

Next, the AuCo/MCF catalyst oxidized at 400 °C in O₂ was further reduced at 400 °C in H₂, and the reaction results of Fig. 7g showed that the reduced catalyst presented decreased CO oxidation activity compared to the oxidized Au–CoO_x/MCF catalyst. The treatment in 400 °C H₂ can reduce the surface Co oxide to metallic Co. The surface energy of Au is much lower than that of Co, therefore, surface Co atoms tend to diffuse into the Au subsurface regions while Au atoms segregate to the topmost surface when annealing in the reductive atmosphere.³⁷

It is generally recognized that small gold particles are necessary for achieving high activity in CO oxidation, and the question arises as to how certain additives can stabilize metal particles. These results of Fig. 7 indicated that when the AuCo alloy particles were oxidized, the Au and CoO_x interfaces were in very close contact, and it was possible that a strong metal–metal oxide interaction existed between the Au and CoO_x. As expected, both the two catalysts in the absence of O₂, Au/MCF and AuCo/MCF catalyst before calcinations, exhibited a lower catalytic activity. Similar Au nanoparticle heterostructures, such as Au–NiO catalysts, have also shown an enhancement in the oxidation of CO, which was parallel to the observations in this work.³⁸ In addition, based on the literature,³⁹ the support effect of oxide-supported catalyst system is largely related to the surface defects (i.e. oxygen vacancies) abundant on the reducible oxides, which play a crucial role in controlling the nucleation, growth, and stabilization of nanosized Au particles and consequently the generation of low-coordinated Au sites responsible for the activation of molecular oxygen. Obviously, the highly dispersed CoO_x in the Au–CoO_x/MCF catalyst plays a key role in this process by providing a high surface-defect concentration.

3.2.2 Stability tests as a function of reaction time. As pointed out by Bond and Thompson, one of the unsatisfactory aspects of current gold catalysis research is that possible structural changes of Au catalysts in conversion with time on stream.⁴⁰ In addition, the high-temperature calcinations is a very stringent test for the stability of gold nanocatalysts against sintering, which is related to the thermal stability of gold catalysts in practical applications where high temperature environments may be encountered. Therefore, the stability of Au–CoO_x/MCF catalysts was tested for the oxidation of carbon monoxide at a GHSV of 30 000 mL (h g_cat)⁻¹ as a function of time on stream for 15 h after being subjected to different calcination temperatures. To test the stability properly, a constant reaction temperature of 30 °C for 15 h was conducted, in order to make sure that the conversion was not 100% on stream, and the structure of catalysts would not change during this reaction temperature. The results of stability tests were shown in Fig. 8.

Fig. 8 Conversions of CO over different catalysts as a function of time on stream ((a) AuCo/MCF + O₂ (200 °C, 1 h); (b) AuCo/MCF+ O₂ (300 °C, 1 h); (c) AuCo/MCF + O₂ (400 °C, 1 h); (d) AuCo/MCF + O₂ (500 °C, 1 h); (e) AuCo/MCF + O₂ (400 °C, 1 h) + H₂ (400 °C, 1 h)).

As shown in Fig. 8, although the catalytic activities of different catalysts were different due to the different pretreatment conditions, the conversion of CO over all the catalysts still seemed to be very stable over 15 hours on stream. Au–CoO_x/MCF was found to be the most active when it was calcined at 500 °C for 1 h in 10% O₂/Ar and maintained the CO conversions at 50–55% for 30 °C reaction over 15 h time. The Au–CoO_x/MCF catalyst calcined at 400 °C O₂ showed slightly less catalytic activity, with ∼40% CO conversions over the same time frame. In case of the AuCo/MCF catalyst calcined at 200 °C O₂, the conversions of CO always kept very low during the 15 h reaction time because of the incomplete removal of organic residue used in the preparation process, which was agreed with the results in Fig. 7.

As reported,⁴¹ the stability of SiO₂-supported Au catalyst exhibited declined trend with a certain fluctuation in CO conversion during the reaction time. The most likely reason for the drop in CO conversion with the Au/SiO₂ catalyst over the time may be the sintering of particles because of the weak interaction between Au nanoparticles and SiO₂ support. With the subsequent oxidation treatment at 300, 400 or 500 °C, the Co atoms alloyed with Au segregated and formed highly dispersed CoO_x nanopatches, and a Au–CoO_x/MCF heterostructure was formed. Transition metals such as Fe, Co, and Ni have a stronger oxygen affinity than Au, so it is expected that the transition metal tends to segregate to surfaces and get oxidized under oxidative conditions.^20,42

As shown in Fig. 8, when the Au–CoO_x was formed, no deactivation was observed, and the catalysts remained very stable for 15 h period. It was because that CoO_x had a stronger interaction with silica than Au, so it could anchor Au particles to the support, keeping the particles in an optimal size range to efficiently catalyze the reaction. Similarly, Liu and Li have speculated that CuO_x may donate O to the CO that was adsorbed onto the surface of Au particle and this effect may contribute to the improved ethanol oxidation performance of the Au–Cu catalysts.^20,42 In addition, as reported, the deactivation of gold catalysts during the process of CO oxidation was often ascribed to the accumulation of carbonate and/or formate species,⁴³ whereas the increase in CO conversion before reaching a steady state was due to the accumulation of trace amount of water that promoted CO oxidation and the more sufficient reduction of gold during CO oxidation.⁴⁴ On the basis of the results, we concluded that the Au–CoO_x structure was highly active for the CO oxidation, and the active sites may be located at the perimeters of CoO_x nanopatches. The ensemble at the Au–CoO_x boundaries may provide dual sites for O₂ activation and CO adsorption.

3.3 Characterization of used catalyst after reaction

As known to all, the metal sintering was a crucial factor to affect the performance of catalysts. In order to further investigate the metal sintering of the catalyst during the reaction, XRD analysis of the spent AuCo/MCF + O₂ (500 °C, 1 h) catalyst was performed after CO oxidation activity test for different temperatures, which were shown in Fig. 9. The used catalyst AuCo/MCF + O₂ (500 °C, 1 h) collected after reaction were characterized by XRD compared with the corresponding fresh catalyst. As shown in Fig. 9, the only features observable in the XRD were due to metallic gold and no cobalt diffraction features were detected, suggesting that Co might still exist as highly dispersed species. The average particle size of Au calculated using the XRD line width at 2θ = 38° by Scherrer formula was 6.5 nm, which was a little larger than that of Au particle size of the fresh catalyst (4.6 nm). This observation further suggested that the CoO_x was helpful to the resistant capacity of Au nanoparticles to sintering during the CO oxidation reaction. When Au was supported on a reducible oxide support, such as TiO₂, CeO₂, Fe₂O₃, etc., the reducibility of the support was increased at the interface, which increased the mobility of the lattice oxygen to participate in the reaction.^7,8 Therefore, from the above results, it was concluded that the interfacial contact between Au and CoO_x resulted in a highly reducible and active CoO_x phase that worked in synergy with the Au particles.

Fig. 9 XRD of the used AuCo/MCF + O₂ (500 °C, 1 h) catalyst (b) compared with the fresh catalyst (a).

4. Conclusions

Au–Co alloy nanoparticles were synthesized through the diffusion of reduced Co into MCF-supported Au nanoparticles, and these formed heterostructures produced Au–CoO_x/MCF catalysts that were capable of obtaining high activity and stability toward carbon monoxide oxidation, in which AuCo/MCF catalyst calcined at 500 °C O₂ for 1 h provided the highest CO conversions. According to the above studies in the structure–activity relationship, the enhanced catalytic performance can be attributed to the synergistic effect between Au and CoO_x components. It was probable that the close proximity between the CoO_x phase and the Au metal allowed the two materials to work in synergy to convert CO into CO₂ than if Au was on its own. In addition, the CoO_x increased the resistance to sintering, thus allowing the particles to remain in an optimal size range to catalyze the oxidation of carbon monoxide. Therefore, oxidative dealloying of supported alloy nanoparticle catalysts is a valuable method of forming heterostructured catalysts to help promote catalytic activity and stability for the CO oxidation.

Acknowledgements

Financial supports of this work by the National Natural Science Foundation of China (21403304) are greatly appreciated. At the same time, the research was sponsored by the Division of Chemical Sciences, Geosciences, and Biosciences, Office of Basic Energy Sciences, US Department of Energy. The electron microscopy at ORNL was supported by the U.S. Department of Energy, Office of Science, Basic Energy Sciences, Materials Sciences and Engineering Division and performed in part as a user project at the ORNL Center for Nanophase Materials Sciences, which is a DOE Office of the Science User Facility.

References

1. M. Haruta and M. Daté, Appl. Catal., A, 2001, 222, 427.
2. A. S. K. Hashmi and G. J. Hutchings, Angew. Chem., Int. Ed., 2006, 45, 7896.
3. P. Zhang, J. Guo, P. Zhao, B. Zhu, W. Huang and S. Zhang, RSC Adv., 2015, 5, 11989.
4. (a) M. Haruta, N. Yamada, T. Kobayashi and S. Iijima, J. Catal., 1989, 115, 301; (b) M. Haruta, S. Tsubota, T. Kobayashi, H. Kageyama, M. J. Genet and B. Delmon, J. Catal., 1993, 144, 175.
5. J. Yu, G. Wu, G. Lu, D. Mao and Y. Guo, RSC Adv., 2014, 4, 16985.
6. T. V. Choudhary and D. W. Goodman, Top. Catal., 2002, 21, 25.
7. A. Sanchez, S. Abbet, U. Heiz, W. D. Schneider, H. Hakkinen, R. N. Barnett and U. Landman, J. Phys. Chem. A, 1999, 103, 9573.
8. U. Heiz and W. D. Schneider, J. Phys. D: Appl. Phys., 2000, 33, R85.
9. M. Valden, X. Lai and D. W. Goodman, Science, 1998, 281, 1647.
10. M. S. Chen and D. W. Goodman, Science, 2004, 306, 252.
11. S. Arrii, F. Morfin, A. J. Renouprez and J. L. Rousset, J. Am. Chem. Soc., 2004, 126, 1199.
12. M. Comotti, W. C. Li, B. Spliethoff and F. Schüth, J. Am. Chem. Soc., 2006, 128, 917.
13. M. Okumura, S. Nakamura, S. Tsubota, T. Nakamura, M. Azuma and M. Haruta, Catal. Lett., 1998, 51, 53.
14. F. Moreau and G. C. Bond, Appl. Catal., A, 2006, 302, 110.
15. Z. P. Liu, S. J. Jenkins and D. A. King, Phys. Rev. Lett., 2004, 93, 156102.
16. H. F. Yin, C. Wang, H. G. Zhu, S. H. Overbury, S. H. Sun and S. Dai, Chem. Commun., 2008, 36, 4357.
17. S. H. Zhou, Z. Ma, H. F. Yin, Z. L. Wu, B. Eichhorn, S. H. Overbury and S. Dai, J. Phys. Chem. C, 2009, 113, 5758.
18. K. Yu, Z. Wu, Q. Zhao, B. Li and Y. Xie, J. Phys. Chem. C, 2008, 112, 2244.
19. J. C. Bauer, D. Mullins, M. Li, Z. Wu, E. A. Payzant, S. H. Overbury and S. Dai, Phys. Chem. Chem. Phys., 2011, 13(7), 2571.
20. X. Liu, A. Wang, L. Li, T. Zhang, C. Y. Mou and J. F. Lee, J. Catal., 2011, 278(2), 288.
21. L. Li, C. Tian, S. H. Chai, A. Binder, S. Brown, G. M. Veith and S. Dai, Catal. Commun., 2014, 46, 234.
22. S. Galvagno and G. Parravano, J. Catal., 1978, 55, 178.
23. C. Mohr, N. Hofmeister, M. Lucas and P. Claus, Chem. Eng. Technol., 2000, 23, 324.
24. P. Schmidt-Winkel, W. W. Lukens, P. D. Yang, D. I. Margolese, J. S. Lettow, J. Y. Ying and G. D. Stucky, Chem. Mater., 2000, 12, 686.
25. (a) H. G. Zhu, Z. Ma, J. C. Clark, Z. W. Pan, S. H. Overbury and S. Dai, Appl. Catal., A, 2007, 326, 89; (b) H. G. Zhu, C. D. Liang, W. F. Yan, S. H. Overbury and S. Dai, J. Phys. Chem. B, 2006, 110, 10842.
26. H. Araki, A. Fukuoka, Y. Sakamoto, S. Inagaki, N. Sugimoto, Y. Fukushima and M. Ichikawa, J. Mol. Catal. A: Chem., 2003, 199, 95.
27. K. Inoue, N. Sasaki, T. Sasadaira, T. Watanabe and T. Nakata, Mater. Trans., 2006, 47(6), 1546.
28. J. C. Bauer, G. M. Veith, L. F. Allard, Y. Oyola, S. H. Overbury and S. Dai, ACS Catal., 2012, 2, 2537.
29. H. Xu, W. Chu, J. Luo and T. Zhang, Chem. Eng. J., 2011, 170, 419.
30. R. Mu, X. Guo, Q. Fu and X. Bao, J. Phys. Chem. C, 2011, 115, 20590.
31. K. I. Hadjiivanov and G. N. Vayssilov, Adv. Catal., 2002, 47, 307.
32. Z. Wu, S. Zhou, H. Zhu, S. Dai and S. H. Overbury, J. Phys. Chem. C, 2009, 113, 3726.
33. C. Tian, S. H. Chai, X. Zhu, Z. Wu, A. Binder, J. C. Bauer, S. Brwon, M. Chi, G. M. Veith, Y. Guo and S. Dai, J. Mater. Chem., 2012, 22, 25227.
34. K. Qian, W. Huang, J. Fang, S. Lv, B. He, Z. Jiang and S. J. Wei, J. Catal., 2008, 255, 269.
35. X. Xu, Q. Fu, X. Guo and X. Bao, ACS Catal., 2013, 3, 1810.
36. H. Yin, Z. Ma, M. Chi and S. Dai, Catal. Today, 2011, 160, 87.
37. G. Abadias, B. Gilles and A. Marty, Appl. Surf. Sci., 2001, 177, 273.
38. S. Zhou, H. Yin, V. Schwartz, Z. Wu, D. Mullins, B. Eichhorn, S. H. Overbury and S. Dai, ChemPhysChem, 2008, 9(17), 2475.
39. D. W. Goodman, Catal. Lett., 2005, 99, 1.
40. G. C. Bond and D. T. Thompson, Gold Bull., 2000, 33, 41.
41. L. Li, S. H. Chai, A. Binder, S. Brown, G. M. Veith and S. Dai, Catal. Lett., 2014, 144, 912.
42. X. Li, S. S. S. Fang, J. Teo, Y. L. Foo, A. Borgna, M. Lin and Z. Zhong, ACS Catal., 2012, 2, 360.
43. A. Knell, P. Barnickel, A. Baiker and A. Wokaun, J. Catal., 1992, 137, 306.
44. Z. Ma, H. F. Yin, S. H. Overbury and S. Dai, Catal. Lett., 2008, 126, 20.

---