Given that formaldehyde is a major indoor air pollutant, significant efforts have been directed at indoor HCHO removal to meet environmental regulations and human health needs [1-4].Catalytic oxidation is recognized as the most promising HCHO removal technology.Catalysts applied for HCHO oxidation include both supported base metals and noble metals [5, 6].However, base metals require the use of elevated temperatures for HCHO oxidation.For example, the operating temperatures of MoO3-SnO2, MnOx-CeO2 and Ag/CeO2 catalysts are in the range of 100-300 ℃ [7-9].In contrast, noble metal catalysts such as Ru, Pd and Pt can effectively remove HCHO at low temperature [10-12].Among them, supported Pt catalysts have been proven to be the most active [10, 13-15].
A variety of oxide-supported Au catalysts, such as Au/Co3O4-CeO2, Au/ZrO2, and Au/CeO2, have been tested for HCHO catalytic oxidation [16-22].Zhang et al.[20] successfully synthesized three-dimensionally ordered macroporous 3DOM Au/CeO2 catalyst, which was found to exhibit superior catalytic activity as exemplified by 100% HCHO conversion at 75 ℃.The unique structure of 3DOM CeO2 favored high gold dispersions, which was believed to be a key factor in the enhancement of catalytic efficiency for HCHO oxidation.Recently, Ma et al.[17] reported that 50% HCHO conversion was achieved over a mesoporous Au/Co3O4 catalyst at room temperature.It was found that HCHO could be oxidized into formate by Co3+, and could be further transformed into bicarbonate or carbonate species, which then decomposed into CO2 and H2O.
In our previous study, reducible CeO2 and FeOx supported nano-gold catalysts were investigated for catalytic HCHO oxidation, Au/CeO2 catalyst prepared by deposition-precipitation (DP) method using urea as precipitants and Au/FeOx catalyst prepared by co-precipitation and calcined at 200 ℃ showed the best catalytic activity, complete oxidation of HCHO into CO2 and H2O being achieved at room temperature and in humid air [23, 24].Such reducible oxides show strong interaction with supported nano-gold particles, leading to weaken Ce-O and Fe-O band, and increase the reducibility reactivity of these surface active oxygen species towards HCHO oxidation [25-27].
In a recent study, we reported for the first time that γ-Al2O3 supported Au was a very active catalyst for HCHO oxidation at room temperature even in the presence of moisture [28].It is found that although there is no active surface oxygen on γ-Al2O3, surface hydroxyls have the ability to partially oxidize HCHO into formate intermediates, which can be further oxidized into CO2 and H2O by nano-Au [28].This study challenges the traditional idea of supporting noble metals on reducible oxides for HCHO oxidation at room temperature [28].
Herein, under similar conditions, we compare the catalytic properties of Au/CeO2, Au/FeOx, Au/γ-Al2O3, Au/HZSM-5 and Au/SiO2 catalysts for HCHO oxidation at room temperature.The goal of the present study is to provide a clear image on the effect of reducibility of the supports for gold catalysis and the factors that influence the gold catalysts’ activity and stability for HCHO oxidation.
Au/γ-Al2O3, Au/CeO2 and Au/FeOx samples with nominal gold contents of 1 wt% were prepared by DP method with urea as the precipitant.Typically, 2.1 mL of HAuCl4 solution (0.024 mol/L), 0.38 g of urea (urea/Au=125, molar ratio), 1.0 g γ-Al2O3, CeO2 or FeOx (γ-Al2O3 provided by Lanzhou Petrochemical Company Petroleum, CeO2 provided by Tianjin anylink new materials Co., Ltd.and FeOx was prepared according to a co-precipitation method [29]) powders (mesh size > 100) and 150 mL of deionized water were co-added into a three-neck flask mixed adequately in the water bath thermostated at 80 ℃, and the flask was covered with an opaque package to keep off the room light.The suspensions were then stirred vigorously for 8 h (final pH ≈ 7.0), followed by overnight aging at room temperature.The solid was filtered and washed extensively with deionized water until it was free of chloride ions, then dried at 80 ℃ for 8 h and calcined at 300℃ in air for 4 h.
The Au/HZSM-5 (SiO2/Al2O3=500) and Au/SiO2catalysts with nominal Au contents of 1 wt% were prepared by a DP method with NH3·H2O as the precipitant.Typically, an aqueous solution of HAuCl4 (0.024 mol/L), HZSM-5 or SiO2 support (1.0 g) (HZSM-5 SiO2/Al2O3=500, Nankai University, China and SiO2 is provided by Qingdao Haiyang Chemical Co., Ltd) and distilled water (50 mL) were co-added to a three-necked flask and mixed adequately.The slurry was stirred at 60 ℃ for 30 min, then aqueous NH3·H2O solution (1 mol/L) was added under stirring to adjust the pH of the system to about 9, followed by stirring at 60 ℃ for 8 h.After that, the resulting precipitate was filtered and washed with hot distilled water, then dried at 60 ℃ for 6 h and stored in darkness.The solid was then treated at 300 ℃ in reducible atmospheres (10% H2 in N2, 80 mL/min) for 4 h to obtain the final catalysts.
The actual Au content in each sample was determined by inductively coupled plasma-atomic emission spectroscopy (ICP-AES, Optima 2000DV, USA).BET surface area determinations were performed using a nitrogen adsorption apparatus (Micromeritics, Tristar 3000).High-resolution transmission electron microscopy (HRTEM) micrographs were obtained with a Tecnai G2 20 S-TWIN microscope and operated at 200 kV.
Transmission electron microscopy (TEM) images of the catalysts were obtained on a JEOL JEM-2000EX microscope operated at 200 kV.Before observation, the samples were ultrasonically dispersed in ethanol and then a drop of the solution was put onto microgrid carbon polymer supported on copper grid.At least 100 Au nanoparticles were measured to analyze the particle size distribution.
Temperature programmed oxidation (TPO) experiments were used to study the deposits on the all the spent Au/γ-Al2O3, Au/SiO2, Au/HZSM-5, Au/CeO2 and Au/FeOx catalysts, after 360 min of reaction.The catalysts were purged with a dry simulated air stream (100 mL/min) at room temperature for 30 min prior to the experiment.The temperature was heated from room temperature to 400 ℃ at a rate of 10 ℃/min in an air flow and the gas products were analyzed by the on-line COx analyzer and mass spectrometer (MS).No other carbon-containing compounds except CO2 in the products were detected for all the tested catalysts.
The activity of catalysts in HCHO oxidation was carried out in a fixed-bed quartz reactor (i.d.=4 mm).10 mg catalyst powder diluted with 50 mg SiO2 was loaded in the quartz reactor.A feed gas (80 ppm HCHO, 21 vol% O2, 1.6 vol% H2O (RH=50%, 25 ℃), and balanced with N2) was allowed to pass through the catalyst bed at a flow rate of 100 ml/min, resulting in a space velocity (SV) of 600000 ml/(g·h).All the feed gases used in this work were of high-purity grade (99.999%).The gas flow rates were controlled by mass flow controllers.Gaseous HCHO was generated by flowing N2 over paraformaldehyde (99%, Aldrich) in a thermostated bath and the concentration of HCHO was controlled by adjusting the flow rate of N2 and the temperature of the thermostated bath.Gaseous H2O was carried into the gas stream by passing N2 through a bubbler in a water bath at room temperature.The amount of water was controlled by adjusting the flow rate of N2, while keeping the total flow constant.The HCHO/N2 and H2O/N2 streams were then mixed with the main gas stream of O2/N2, leading to a typical feed gas.
Concentrations of CO and CO2 were measured using an online non-dispersive infrared analyzer (SICK-MAIHAK-S710, Germany).In this work, it was not possible to monitor the HCHO concentration directly by Fourier transform-infrared spectroscopy (FT-IR) due to the interfering effects of water.Therefore, HCHO was measured by converting it to CO2 in a homemade HCHO-to-CO2 converter (CuO-MnO2/γ-Al2O3 catalyst) at 300 ℃ and determining the amount of CO2 formed.Several previous reports had pointed out that for the accurate determination of HCHO, the HCHO should first be oxidized to CO2 [30-32].In typical runs, the reaction data were obtained after HCHO oxidation was performed for 40 min in order to achieve the steady state.No other carbon-containing compounds except CO2 in the products were detected for all the tested catalysts.
The conversion of HCHO was calculated as follows:
HCHO conversion (%)=[CO2]out/[HCHO]in × 100%,
where [CO2]out is the CO2 concentration in the product gas stream (vol%) and [HCHO]in is the HCHO concentration in the feed gas (vol%).
Kinetic measurements were conducted in a fixed-bed flow micro-reactor, keeping the HCHO conversion below 20%.The samples were heated in N2 from 30 to 70 ℃, then the gas was switched to the feed gas (80 ppm HCHO, 21 vol% O2, 1.6 vol% H2O and balanced with N2, flow rate 100 mL/min, 5-10 mg sample was used to change the contact time such that the conversion remained < 20%).The gas products were analyzed by the on-line COx analyzer.
CeO2, FeOx, γ-Al2O3, SiO2, and HZSM-5 supported 1 wt% Au catalysts are evaluated for HCHO oxidation at room temperature under humid air and high GHSV of 600000 ml/(g·s), and the results are shown in Fig. 1(a).Au/γ-Al2O3 catalyst exhibits the highest HCHO conversion of ca.88% at the beginning of the reaction, and remains the conversion higher than 78% within 250 min.The conversion decreases to 73% upon 360 min of reaction, which is still the most active catalyst.Though Au/SiO2 catalyst shows a higher HCHO conversion of 79% at its initial, it deactivates fast within the first 100 min, and then decreases at a slower rate.Au/HZSM-5 catalyst exhibits a similar deactivation trend as Au/SiO2 catalyst, but the initial activity is lower, the initial conversion is only ca.44%.As to the CeO2 and FeOx supported Au catalysts, the initial activities under such higher GHSV are quite lower, and they exhibit a gradual decreased activity with time on stream.
Also, the kinetic tests are conducted to further check the supports’ effect on HCHO conversion.Fig. 1(b) shows Arrhenius plots for the rate of HCHO oxidation at the conversion < 20% (the data are collected at the relatively stable stage of 40 min’s reaction at which the reaction temperature range was from 30 to 70 ℃).It is clear that the apparent energy differs from 9 to 14 kJ/mol, and follows the order of Au/γ-Al2O3 < Au/SiO2 < Au/HZSM-5 ≈ Au/CeO2 < Au/FeOx.The quite lower apparent energy indicates the high activities of these supported Au catalysts for HCHO oxidation at room temperature.And the order further accounts for the indirect effect of the reducibility of the supports for gold catalysis in HCHO oxidation reaction.(The H2-TPR profiles of various oxides supported Au catalysts was shown in Fig. 2).
We compare the initial conversions (the data was collected at the ca.10 min) with those after 120 min of reaction for each supported Au catalyst, the results being shown in Fig. 1(c).It’s obvious that fast deactivation were observed over Au/SiO2 and Au/HZSM-5 catalysts.While Au/γ-Al2O3, Au/CeO2 and Au/FeOx catalysts deactivate at much slower rates.From Fig. 1, it suggests that different supports loaded Au catalysts possess different catalytic properties for HCHO oxidation.And there are two types of deactivation observed, fast deactivation for Au/SiO2 and Au/HZSM-5 catalysts and slow deactivation for Au/γ-Al2O3, Au/CeO2 and Au/FeOx catalysts.It is worth raising that Au/γ-Al2O3 catalyst exhibits the highest initial activity as well as a more stable activity for HCHO oxidation.
Then, what is the factor determing the initial activity? And how is the stability being affected? Table 1 summarizes the physical and chemical properties of the gold loaded samples.Au/SiO2 catalyst shows the highest BET surface area, then follows by Au/HZSM-5 and Au/γ-Al2O3 catalysts.CeO2 and FeOx supported Au catalysts have lower surface area, and the lowest one is Au/CeO2 catalyst with surface area of 78 m2/g.ICP results indicate that the gold content is close to the theoretical value but not completely the same.Such a deviation is acceptable in view of the potential experimental error.
Fig. 3 presents several typical TEM images of the supported Au catalysts.It can be seen that Au nanoparticles are highly dispersed on the supports.Estimated according to the TEM images ( > 100 particles measured), the average particle size of gold and the dispersion are also listed in Table 1.Au/SiO2 and Au/γ-Al2O3 catalysts show higher gold dispersions, with an average particle size of 1.6 and 1.8 nm, respectively.Large size distributions are observed over Au/CeO2, Au/HZSM-5 and Au/FeOx catalysts.Among them, the poorest Au dispersion appears over Au/FeOx sample.In spite of its surface area of 187 m2/g, the average Au particle is the largest (4.9 nm), which suggests that the Au dispersion is not only influenced by the surface area of the support, but also the preparation method and the surface interaction between gold and oxide [33, 34].
To study the effect of Au particle size on catalytic activities, the reaction rate is correlated with the average Au particle size, shown in Fig. 4.There is a linear relation between them, given strong evidence of Au particle size determine the catalytic activity.Au/γ-Al2O3 and Au/SiO2 catalysts with the higher Au dispersion and smaller gold particle size render the best catalytic HCHO oxidation activities.Moreover, the effect of hydroxyl groups on the γ-Al2O3 and SiO2 should also be concerned, which could assist the decomposition of HCHO or reaction intermediates.These results suggested that the synergistic effect of the small Au particle size and the hydroxyl groups would lead to the higher initial activities of the Au/SiO2 and Au/γ-Al2O3.However, Au/FeOx catalyst with the larger Au particle size shows the poorest activity.
To study the factors that affect the catalysts’ stability, the spent catalysts upon 360 min of reaction are characterized by TEM and TPO, as shown in Figs. 5 and 6, respectively.For the spent Au/γ-Al2O3 catalyst (Fig. 5(a)), nano-gold particles are still highly dispersed on the support, with an average particle size of ca.2.2 nm, which is a little larger than that of the fresh sample.The similar trends are observed over the spent Au/CeO2 and Au/FeOx catalysts, nano-gold particles are still highly dispersed on the support, with a small increase in gold particle size.However, the images change obviously for the used Au/SiO2 catalyst.In some areas, shown in Fig. 5(b), the particles become crowded together, indicating that the gold particles aggregated obviously over the SiO2 support, even upon the room temperature reaction.Some aggregated gold particles could also be observed over the spent Au/HZSM-5 sample.
Then, why could γ-Al2O3, CeO2 and FeOx supports stabilize nano-gold particles in a better way than SiO2 and HZSM-5 substrates? For the reducible oxides such as CeO2 and FeOx, it is generally regarded that there are strong surface interaction between the gold and the oxides [25, 35].Gold-metal oxide perimeter interface acts as sites for activating the reactants [36, 37].The capacity of highly dispersed small gold particles to weaken the Ce-O and Fe-O band, thus increasing the redox properties of the Au/CeO2 and Au/FeOx catalysts has been studied in many literates as well as our previous works [23-25].Meanwhile, the nano-gold could be anchored with the oxide through-O linkages.That provides the possible way for stabilization of nano-gold particles [38].As to the higher stability of nano-gold over γ-Al2O3 substrate, it is supposed that the enrichment of surface hydroxyl groups act as “ligands” binding the gold species [39-41].To be related to the abundance of surface hydroxyl species of γ-Al2O3, and the results are shown in Fig. 7.However, due to the lack of surface -O and -OH, the binding between nano-gold and SiO2/HZSM-5 is loser, which resulted in the easier aggregation of gold particles.
To study the surface species that deposited on the catalyst surface during the reaction which might result in the coverage of the active sites and deactivation of the catalyst, the spent samples are temperature-programmed oxidized in 21% O2/N2 gases, and the results are shown in Fig. 6.There is no CO(g) detected, except CO2(g) in the products for all the samples.For the spent Au/SiO2 and Au/HZSM-5 samples, with increasing temperature, a sharp peak appears at around 50 and 100 ℃, owing to the release of physical adsorption CO2 on the catalyst surface due to the reaction.And at ca.350 ℃, there is an obvious CO2 peak observed over Au/γ-Al2O3, Au/SiO2 and Au/HZSM-5 catalysts, while the temperature shifts to lower region ( < 300 ℃) over the spent CeO2 and FeOx supported gold catalysts, which should come from the oxidation of the deposited surface species into CO2(g) at elevated temperatures.Based on the literatures and our previous studies, the deposits might be the dioxymethylene (DOM) and formate species, as shown in the Fig. 8 [23, 24, 28, 42, 43].Accordingly, more CO2 generated means more deposits.By calculating the amount of CO2 generated shown in Table 1, it is clear that there are most deposits on Au/CeO2 catalyst, then follows by the Au/FeOx and Au/γ-Al2O3 catalysts, and the deposits on Au/SiO2 catalyst is the least, which correlated well with in-situ DRIFTs results shown in Fig. 8.More formate species formed over Au/CeO2 catalysts, so the more deposits are accumulated on it.
But why does the Au/SiO2 and Au/HZSM-5 samples deactivate at a faster rate, especially at the initial stage of the reaction? From the TEM images of the spent catalysts, it is clear that gold aggregated together obviously for the Au/SiO2 catalyst.In contrast, gold particles keep well dispersed for the spent Au/γ-Al2O3, Au/CeO2 and Au/FeOx catalysts, which only exhibited a gradual deactivation during HCHO oxidation.Such results suggest that the aggregation of gold and the growing gold particle sizes are responsible to the fast deactivation of Au/SiO2 as well as Au/HZSM-5 samples.While the deposition of surface intermediates on the catalyst surface causes the gradual deactivation of gold catalyst.
The catalytic properties of Au/γ-Al2O3, Au/SiO2, Au/HZSM-5, Au/CeO2 and Au/FeOx catalysts for HCHO oxidation were compared at room temperature under high GHSV.The gold particle size is crucial in determining the initial activity, and Au/γ-Al2O3 and Au/SiO2 catalysts with the higher Au dispersion and smaller Au particle size render the best catalytic HCHO oxidation activities, suggesting that the indirect effect of the reducibility of the supports for gold catalysis in HCHO oxidation reaction.TEM and TPO results suggested that the weak interaction between Au and SiO2/HZSM-5 substrate led to the growth of Au particles, and resulted in faster deactivation.Moreover, the accumulation of the surface formate intermediates accounts for the gradually deactivation.The present results suggest that the reducibility of the supports is not the key factor that influences the gold catalysts activities, but affects the stability of the supported Au catalyst for HCHO oxidation.
2016,
Vol. 37 Issue (10): 1729-1737 
PDF
2016,
Vol. 37 Issue (10): 1729-1737 PDF