Since the 1990s, nanosized gold interacting with metal oxide was reported to be very active for low-temperature oxidation of CO [1, 2]. Such unique catalytic properties were found to be strongly dependent on the specific Au structure, which is also controlled by the supported metal oxide. Although a full mechanism of this catalytic process still needs to be established, careful studies on metal-support interaction may provide mechanistic insights [3]. On the other hand, the exploration for new type of oxide support has attracted considerable attention of many research groups, in aim to enhance the catalytic performance of gold catalyst. Besides inert (SiO2 [4], Al2O3 [5], etc.) and reducible oxides (TiO2 [6], CeO2 [7], Fe2O3 [3], etc.), ZnO [8] and SnO2 [9] have also proved to be effective for deposition of gold.
Zinc tin oxide (Zn2SnO4, ZTO) is an inverse spinel (AB2O4) compound and has shown its unique properties in the applications of transparent conducting oxides [10], anode materials in Li-ion batteries [11], photoelectrical devices [12] and sensors [13]. In synthesis, hydrothermal approach with appropriate parameters (type of base, ratio between base and metal ion, reaction temperature and time, etc.) has been applied to obtain ZTO nanomaterials [14]. However, till now, zinc tin oxide has not been used as a support for preparation of gold catalyst. Therefore, no related observations and understandings have been achieved for the “structure-activity” relationship in the Au-ZTO system. For instance, is ZTO an active or inactive support for gold? What is gold structure on ZTO? Is there any support effect on gold activity for catalytic reaction? To answer these questions, we first need to prepare the Au-ZTO catalysts via a reliable route, e. g. deposition-precipitation [15].
Nowadays, investigations on “structure-activity” relationship have also been dependent upon advanced characterization techniques in catalysis. For oxide-supported gold catalyst, diverse means including X-ray diffraction (XRD) [16], X-ray absorption fine structure (XAFS) [17] and transmission electron microscopy (TEM) [3] have been used to identify the structural evolutions on both active metal (Au) and oxide support for the catalytic CO oxidation. Among them, XAFS, which includes X-ray near edge structure (XANES) and extended X-ray absorption fine structure (EXAFS), is an element-sensitive skill and effective for determinations on both electronic (oxidation state) and short-range (up to 6-8 Å) local structure (coordination number and bond distance) of gold species.
Therefore, in this work, we tried to explore deposition of gold onto the ZTO supports with different textural properties via deposition-precipitation, to apply multiple characterizations, especially XAFS, for revealing the structures of both Au and ZTO, and to investigate “structure-activity” relationship in Au-ZTO system and identify key factors on active gold species.
All the chemicals used in this work were of analytical grade and purchased from Sinopharm Chemical Reagent Co., Ltd. without any further purification.
The ZTO support was synthesized via the hydrothermal synthesis method [14]. ZnCl2 (8 mmol) and SnCl4 (4 mmol) were dissolved in deionized water (80 mL) under vigorous magnetic stirring. N2H4·H2O (molar ratio of N2H4/Zn2+=4/1, 8/1 and 16/1) was then added to the above solution. White precipitates formed immediately, and this solution was kept under stirring for another 30 min. The stock solution was transferred into a Teflon bottle (100 mL), and further tightly sealed in a stainless-steel autoclave. The hydrothermal procedure was carried out in a temperature-controlled electric oven at 180 ℃ for 12 h. The obtained products were thoroughly washed with deionized water dried in a vacuum oven at ca. 80 ℃ overnight and then grounded and sieved (200 mesh). The dried powders were calcined in still air at 400 ℃ for 4 h (ramp rate: 2 ℃/min). The zinc tin oxide supports were donated as ZTO_1 (N2H4·H2O/Zn2+=4/1), ZTO_2 (N2H4·H2O/Zn2+=8/1) and ZTO_3 (N2H4·H2O/Zn2+=16/1) in this work.
The deposition of gold (Au/(Au+Zn2SnO4)=1 at%) was obtained via the deposition-precipitation method [15]. ZTO support (1 g) was suspended in deionized water (50 mL). After stirring for 15 min, ammonium carbonate solution (25 mL, 1 mol/L) was added. HAuCl4·3H2O (12.6 mg) was dissolved in 25 mL deionized water, and then added into the stock solution dropwisely. After stirring and aging at room temperature for 1 h, the as-formed precipitates were gathered by centrifugation and then washed with deionized water at RT. The final product was obtained after the drying (ca. 60 ℃, vacuum, overnight) and calcination (400 ℃, still air, 4 h) processes. The Au-ZTO catalysts were donated as Au_ZTO_1 (N2H4·H2O/Zn2+=4/1), Au_ZTO_2 (N2H4·H2O/Zn2+=8/1) and Au_ZTO_3 (N2H4·H2O/ Zn2+=16/1).
The gold loading concentrations were determined by inductively coupled plasma atomic emission spectroscopy (ICP-AES) on an IRIS Intrepid II XSP instrument (Thermo Electron Corporation).
The powder XRD patterns were recorded on a Burker D8 Advance diffractometer (40 kV, 40 mA) with a scanning rate of 4°/min, using Cu Kα radiation (λ =1.5406 Å). The corresponding XRD patterns were collected from 10° to 90° with a step of 0.02°. The 2q angles were calibrated with a μm-scale Alumina disc. The powder catalyst after grinding was placed inside a quartz-glass sample holder for each test. With the software “LAPOD” of least-squares refinement of cell dimensions from powder data by Cohen’s Method [18, 19].
The nitrogen adsorption-desorption measurements were performed on an ASAP 2020-HD88 analyzer (Micromeritics Co. Ltd.) at -196 ℃. All the tested samples were degassed at 350 ℃ under vacuum ( < 100 mmHg) for over 4 h. The BET specific surface areas (ABET) were calculated from data in the relative pressure range between 0.05 and 0.20. The pore size (rp) distribution was calculated from the desorption branch of the isotherms, based on the BJH method [20].
The TEM and high-resolution TEM (HRTEM) experiments were carried out on a Philips Tecnai G2 F20 instrument at 200 kV. All the tested samples were suspended in ethanol, and then a drop of this dispersed suspension was placed on an ultra-thin (3-5 nm in thickness) carbon film-coated Cu grid. The as-formed sample grid was dried naturally under ambient conditions before inserted into the sample holder. High-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) mode with energy dispersive X-ray analysis (EDAX) was applied to confirm the presence of Au particles.
X-ray photoelectron spectroscopy (XPS) analysis was performed on an Axis Ultra XPS spectrometer (Kratos, U.K.) with 225 W of Al Kα radiation. The C 1s line at 284.8 eV was used to calibrate the binding energies.
The XAFS spectra at Au L2-edge (E0=13734 eV) were performed at BL14W1 beam line of Shanghai Synchrotron Radiation Facility (SSRF) operated at 3.5 GeV under “top-up” mode with a constant current of 220 mA. The XAFS data were recorded under fluorescence mode with a 32-element Ge solid state detector. The energy was calibrated accordingly to the absorption edge of pure Au foil. Athena and Artemis codes were used to extract the data and fit the profiles. For the XANES part, the experimental absorption coefficients as function of energies μ(E) were processed by background subtraction and normalization procedures, and reported as “normalized absorption”. For the extended EXAFS part, the Fourier transformed (FT) data in R space were analyzed by applying metallic Au model for the Au-Au shell. The passive electron factors, S02, were determined by fitting the experimental Au foil data and fixing the Au-Au coordination number (CN) to be 12, and then fixed for further analysis of the measured samples. The parameters describing the electronic properties (e.g., correction to the photoelectron energy origin, E0) and local structure environment including CN, bond distance (R) and Debye Waller (D.W.) factor around the absorbing atoms were allowed to vary during the fit process. The fitted ranges for k and R spaces were selected to be k=2.5-10.5 Å−1 and R=1.7-3.3 Å (k2 weighted), respectively.
The temperature-programmed reduction by hydrogen (H2-TPR) was performed in a Builder PCSA-1000 instrument equipped with a thermal conductivity detector (TCD). The reduction process was carried out in a mixture of 5% H2/Ar (30 mL/min) from room temperature to 800 ℃ (10 ℃/min). The catalysts (ca. 30 mg) were pretreated in pure O2 at 300 ℃ for 30 min before each test.
The catalytic activity for CO oxidation was measured using a continuous flow fixed-bed reactor system. For each test, 50 mg of the catalyst sample powder mixed with about 50 mg of quartz sand was loaded into an iron chamber reactor. A feed stream containing 1 vol% CO and 20 vol% O2 balanced with N2 was allowed to pass through the catalyst sample at a flow rate of 66.7 mL/min, resulting in a space velocity (SV) of 80000 mL gcat-1 h-1). Before each run, the sample was flowed for 30 min with oxidative (20% O2/N2) atmospheres at 300 ℃. For a typical catalytic test, the temperature of Au-ZTO sample was heated from 20 to 280 ℃ (10 ℃/step) and was stabilized at each temperature plateau under the same reaction conditions for 30 min to reach the equilibrium. The compositions of the effluent gases were monitored online by a non-dispersive IR spectroscopy (Gasboard-3500, Wuhan Sifang Company, China). The CO conversion was calculated according to the following equations:
Rate measurements were made in the same gas composition, but at varied space velocities to ensure operation in the kinetic regime (CO conversion < 15%).
Previously, Shin et. al. [14] reported the preparation of zinc tin oxide nanoparticles via coprecipitation by using hydrazine (N2H4·H2O) as precipitating agent. In this work, we used the similar route and focused on the synthetic parameters towards different textures of ZTO.
In experiments, we found that the molar ratio between base and metal ion (N2H4·H2O/Zn2+) determined the nitrogen adsorption-desorption behavior of ZTO. Fig. 1(a) displays that ZTO_1 and ZTO_2 have H3 type of hysteresis loop, typically given by aggregates of particles [21]; while ZTO_3 has H2 type of hysteresis loop, possibly contributed by mesoporous oxide [21]. Table 1 shows that the BET specific surface areas (ABET) of the as-calcined supports are almost identical for ZTO_1 (76 m2/g) and ZTO_2 (79 m2/g), but show clear increase for ZTO_3 (107 m2/g).
Fig. 1(b) exhibits the pore size distribution histograms on different ZTO supports. A single sharp peak appears for ZTO_2; a broad peak is displayed for ZTO_1; while two peaks are exhibited for ZTO_3. Table 1 shows a distinct shift of averaged pore size from 9.6 nm (ZTO_1) or 9.8 nm (ZTO_2) to 5.3 nm (ZTO_3), together with a rapid decrease in pore volume from 0.205 cm3/g (ZTO_1) or 0.223 cm3/g (ZTO_2) to 0.162 cm3/g (ZTO_3). All the above results indicate that the synthetic parameter (N2H4·H2O/Zn2+) plays a crucial role in the textural properties of ZTO.
During the deposition-precipitation step, there was a color change, from white (ZTO) to taupe (Au-ZTO), giving a hint on the complete of gold loading. ICP-AES was conducted to determine the Au concentration in each sample. Table 2 lists that the experimental Au concentrations (0.57-0.59 wt%) are very similar between different supports, and close to the designed value (0.70 wt%).
The catalytic performance of the Au-ZTO catalysts was evaluated for the CO oxidation. The transient profiles in Fig. 2(a) reveal large differences on the “light off” temperatures for the tested samples. The 90% CO conversion temperature (T90) was 137 ℃ for Au_ZTO_2, while 295 ℃ for Au_ZTO_1. However, for Au_ZTO_3, zero conversion was observed before 220 ℃, but a sudden jump appeared from 250 ℃ (9%) to 280 ℃ (100%). The absolute reactivity was lower than the previous reports on Au-ZnO [8] or Au-SnO2 [9], revealing that ZTO may not be a very active support for gold catalyst. Therefore, we mainly focused on the structure-activity relationship in this work. Fig. 2(b) exhibits the Arrhenius-type plots for the CO oxidation reaction rates over the Au-ZTO catalysts. Clearly, Au_ZTO_2 shows the highest reaction rates of 1.8 and 3.2 mmolCO2 g-1 s-1 at 53 and 66 ℃, respectively, while Au_ZTO_1 displays the lower values of 1.3 and 3.0 mmolCO2 g-1 s-1 at 72 and 96 ℃, respectively. However, for Au_ZTO_3, the related reaction rate was the lowest (0.9 mmolCO2 g-1 s-1 at 250 ℃, not shown in Fig. 2(b)). All the above catalytic results on both transient and rate measurements demonstrate the following sequence of CO oxidation reactivity in the Au-ZTO system: Au_ZTO_2 > Au_ZTO_1 » Au_ZTO_3 ZTO_2.
Since Au loaings are the same for all the measured samples, there should be other factors, i.e. Au active sites, support effect, etc., to cause such differences in activity. Here, we first investigated the possible varieties on reaction pathways for these Au-ZTO catalysts. Despite the distinct difference of the CO oxidation reaction rates, the corresponding apparent activation energies from the Arrhenius plots were truly identical (37 kJ/mol) between Au_ZTO_1 and Au_ZTO_2 (Fig. 2(b)), suggesting that the reaction mechanism is the same, while difference being the number of active sites. However, for Au_ZTO_3, only one temperature point was available for the calculation of activation energy, because other CO conversions were out of the range on catalytic reaction even with slight deviation on temperature. Thus, we were unable to obtain the specific value of apparent activation energy for Au_ZTO_3. This also gives a hint that the reaction mechanism for the Au-ZTO catalyst prepared under the highest N2H4·H2O/Zn2+ ratio may be different from that for the other tested samples.
XRD was conducted to determine the crystal structure of the Au-ZTO catalysts. Fig. 3(a) displays a pure cubic (Fd-3m, JCPDS 24-1470) phase with the lattice constants (a) of ca. 8.70 Å (Table 2) for all the fresh samples. After the CO oxidation reaction, the XRD patterns (Fig. 3(b)) were nearly identical for the used catalysts with minor differences in lattice constants only (Table 2). However, no Au diffraction peaks can be observed in both fresh and used Au-ZTO catalysts, due to the presence of ultra-fine gold atoms/clusters or the very low Au concentrations ( < 0.6 wt%).
The TEM and HRTEM measurements were carried out to investigate the morphologies (size and shape) of the Au-ZTO catalysts. Fig. 4(a)-(f) display that all the fresh samples are polyhedron-like nanocrystals with averaged particle sizes of 13-16 nm (Fig. 4(g)-(i) and Table 2). The lattice fringes in HRTEM images (Fig. 4(d)-(f)) clearly verify the cubic Zn2SnO4 phase for the ZTO supports, in good agreement with the XRD results. Fig. 5(a)-(f) exhibit the TEM/HRTEM images of the used catalysts, which are composed by polyhedral particles with averaged sizes of 12-14 nm (Fig. 5(g)-(i) and Table 2). Still, no observable differences between fresh and used Au-ZTO catalysts were detected by TEM/HRTEM, which is well consistent with the XRD data.
The XPS measurement was carried out to investigate the chemical state of Zn and Sn in Au-ZTO. Fig. 6 displays that the Zn 2p and Sn 3d profiles are nearly identical between all the fresh and used samples, indicating the constant oxidation state of Zn and Sn during the CO oxidation reaction. Therefore, if combining the XRD, TEM/HRTEM and XPS characterizations, we can demonstrate that the structure of ZTO support was very stable upon either air-calcination (400 ℃) or reaction conditions (1% CO/20% O2/N2, up to 280 ℃).
Table 3 shows that the binding energies (BEs) of Zn 2p1/2 and Zn 2p3/2 are 1044.4-1044.6 and 1021.3-1021.5 eV, respectively, which are typically for Zn2+; while the BE of Sn 3d3/2 and Sn 3d5/2 are 494.8-494.9 and 486.4-486.5 eV, respectively, which are specifically for Sn4+. From Table 3, we can also find a very minor down-shift (0.1-0.2 eV) of BE as the N2H4·H2O/Zn2+ increases from 4 to 16, revealing a slight reduction of Zn2+ and Sn4+ ions in the ZTO support synthesized by adding more reducing base (N2H4·H2O).
To catch the gold species, we applied HAADF-STEM on the fresh Au-ZTO catalysts. From Fig. 7 we observed some bright spots (marked by arrows) with size of 1-2 nm for all the tested samples. The corresponding EDAX measurements over the same area (data not shown) verified the presence of Au in these particles. So, the missing of gold peaks in XRD is actually due to the low concentration of Au ( < 0.6 wt%). However, because of the high concentration of heavy Sn atoms (z=50), we may fail to identify the large-size ( > 2 nm) gold particles because they were undistinguishable from the ZTO support. Thus, other characterization techniques are required to obtain the statistical results on Au structure.
Unfortunately, the Au 4f profiles were severely overlapped with those of Zn 3p so that we cannot probe the structural information (concentration and oxidation state) by XPS on surface gold species for our Au-ZTO catalysts. Therefore, the XAFS technique was used to investigate the fresh and used catalysts to determine the oxidation state, as well as the local coordination structure of gold. Due to the presence of a large amount of Zn atoms, the fluorescence from Au L3-edge (La line around 9.7 keV) was severely disturbed by that from Zn K edge (Kb line around 9.6 keV). Thus, only the Au L2-edge (Lb line around 11.4 keV), of which signal is much lower than the L3-edge, can be selected to detect gold. So we added a lead cap on top of the detector nose to eliminate the scattering noise from the surrounding environment.
The XANES profiles of Au L2-edge are very similar to that of Au foil in Fig. 8(a), indicating the dominant Au0 species in Au-ZTO before the reaction. For the used samples, the metallic gold state was maintained and no significant differences can be observed (Fig. 8(b)). Thus, the gold metals were actually generated after air-calcination during the deposition-precipitation step and conserved after the catalytic tests. This phenomenon was different from the transformation of Aud+ to Au0 in reducible oxide such as CeO2 for the water-gas shift reaction [22].
Fig. 9 presents the R space EXAFS data (solid lines) on Au-ZTO, as well as the corresponding fitting results (dot lines). We only determined a pure Au-Au bond and the peak-split of 2.5-3.5 Å is due to Ramsauer-Townsend resonance at a single energy in the backscattering amplitude of Au [22]. Table 4 shows that the fitted R value of Au-Au shell is located at 2.83-2.85 Å for all the Au-ZTO samples, whether fresh or used (Fig. 9), and slightly enlarges from Au_ZTO_1/Au_ZTO_2 (2.831-2.836 Å) from Au_ZTO_3 (2.842-2.848 Å). Meanwhile, the fitted averaged coordination number (CN) of Au-Au shell varies in the range of 7.5-10.2 (Table 4), and obviously increases from Au_ZTO_1/Au_ZTO_2 (~8, ca. 2 nm in averaged Au size) to Au_ZTO_3 (~10, 4 nm in averaged Au size) [23]. The longer the bond distance, the higher the coordination number. This was consistent with the previous report on EXAFS of Au [23]. On the other hand, no distinct differences were observed between fresh and used catalysts, revealing the good stability of gold species in Au-ZTO during the CO oxidation reaction.
To investigate the metal-support interaction in the Au-ZTO system, we performed H2-TPR tests on both the Au-ZTO catalysts and the ZTO support. Fig. 10 exhibits that the reduction of ZTO_2 can be divided into two ranges, including one weak low-temperature peak of 300-600 ℃ assigned to the surface reduction of Zn2SnO4 and one strong high-temperature peak above 600 ℃ attributed to the bulk reduction of Zn2SnO4. By the introduction of gold, both surface and bulk reductions shift to lower temperatures, indicating the effect of metal-support interaction between gold and metal oxide [15]. Furthermore, the temperatures of peak-center show the sequence of Au_ZTO_2 (330 ℃; 680 ℃) < Au_ZTO_1 (400 ℃; 740 ℃) < Au_ZTO_3 (450 ℃; 760 ℃) < ZTO_2 (470 ℃; > 800 ℃), inverse to the order of strength for metal-support interaction: Au_ZTO_2 > Au_ZTO_1 > Au_ZTO_3.
The hydrogen consumption amount calculated from H2-TPR provides quantitative information on oxygen species. Typically for Au_ZTO_2, the surface oxygen (reduced at 300-600 ℃) amount was determined as 1170 mmol/g, very close to that for pure ZTO_2 support (1200 mmol/g). It indicates that the addition of gold only weakens the bond strength of M-O (M=Zn or Sn) in the form of Au-Ox-M (x 1), but not contributes more available oxygen species with x » 1. On the other hand, the bulk oxygen (reduced above 600 ℃) was determined as 14500 mmol/g, nearly identical to the theoretic value of 14212 mmol/g for complete reduction of Zn(II) and Sn(IV) ions to metallic Zn(0) and Sn(0) atoms in ZTO. Thus, the majority of oxygen species was unavailable in the CO oxidation tests.
As discussed as above, the support effect of ZTO on the reactivity of Au-ZTO can be demonstrated in Fig. 11. First, according to the nitrogen adsorption-desorption data, the synthetic parameter (N2H4·H2O/Zn2+) distinctly changes the pore structure (volume and size distribution) of ZTO support. Second, on the basis of XAFS results, the textural properties of ZTO effectively control the growth of gold species during deposition-precipitation, resulting in small-size (ca. 2 nm) and large-size ( 4 nm) Au nanocrystals. Third, on the behavior of H2-TPR profiles, the gold size, together with the pore structure of ZTO support, significantly modifies the metal-support interaction between Au and ZTO. Last, as we observed, the strong metal-support interaction benefits the catalytic reactivity on CO oxidation in the Au-ZTO system. Therefore, in this work, gold supported on ZTO with N2H4·H2O/Zn2+=8/1 showed highest pore volume (0.223 cm3/g) and narrowest pore size distribution, and thus resulted in lowest reduction temperatures (surface 330 ℃; bulk 680 ℃) and best activity for the CO oxidation reaction (T90=137 ℃).
Gold catalysts supported on different zinc tin oxide nanoparticles were prepared by the deposition-precipitation method. The reactivity of CO oxidation was significantly dependent on the molar ratio between base and metal ion during the synthesis of Zn2SnO4. On the basis of XRD, TEM/HRTEM and XPS, we found that the crystal structure, shape/size of oxide support and chemical state of Zn/Sn were maintained before and after the reaction. The XANES profiles determined the oxidation state of gold as pure metallic Au0 in both fresh and used catalysts. The EXAFS fitting results indicated the increase in the averaged particle size of gold for N2H4·H2O/Zn2+=16/1 only, corresponding to the rapid decrease in the CO oxidation activity. Furthermore, we can draw a conclusion that the growth of Au nanoparticles, governed by the pore volume and pore size distribution of Zn2SnO4 support, was correlated to the strength of metal-support interaction, which affects the catalytic performance of Au-ZTO.
2016,
Vol. 37 Issue (10): 1702-1711 
PDF
2016,
Vol. 37 Issue (10): 1702-1711 PDF