Low-temperature CO oxidation is valuable as a prototypical reaction for heterogeneous catalysis in fundamental studies [1]. It is also of great importance in practical applications, including automotive emissions control [2] and proton exchange membrane fuel cells [3]. Metal oxides such as copper-manganese (Hopcalite) [4], copper-chromite [5] and cobalt oxides [2, 6] are highly active in CO oxidation, even at ambient temperatures. However, the low thermal stability, undesirable behaviors in cycled transient conditions, and a high susceptibility to be poisoned by sulfur and water [7, 8] severely limit their practical applications. In contrast, noble metals are excellent CO oxidation catalysts with high activity and thermal stability and are widely used in industrial applications; the high cost and limited supply of these noble metals, however, pose significant challenges for sustainable applications [5, 7]. While these noble metals are usually finely divided into nanometer-scale active units and dispersed on high-surface-area materials, their atom efficiency is still very low because only the surface atoms of a nanoparticle are involved in the catalysis. Therefore, it would be highly desirable to maximize the atom efficiency of these supported noble metal catalysts.
Recently, we demonstrated that ferric oxide-supported Pt single-atom catalysts were highly active for CO oxidation [3]. It was subsequently shown by others that Pt [9, 10] and Rh [11] single atoms dispersed on various supports were also highly active, suggesting that the development of single-atom catalysts is a general approach that can maximize the atom efficiency of noble metals [12, 13, 14].
It is generally accepted that oxide-supported Au nanoparticles are the most active catalysts for CO oxidation. However, the activity of single Au atoms in CO oxidation remains elusive. While some theoretical calculations suggested that single Au atoms could be catalytically active for CO oxidation [15], others have suggested CO adsorption could poison Au atoms on FeOx and thus deactivate the catalyst [16]. It has been reported that cationic Au species dispersed on oxide supports showed activity for CO oxidation at room temperature; however, the reported activity was much less than that of the more common small Au nanoparticles [17, 18, 19]. Studies on model catalysts also suggested that when compared with their sub-nanometer-sized cluster counterparts, Au single atoms were much less active [20]. Recently, we found that an FeOx-supported Au single-atom catalyst is not only highly active, but extremely stable for CO oxidation, suggesting a potential application of oxide-supported Au single-atom catalysts [21]. For certain applications, the total oxidation of CO at ambient temperatures is of great practical value. In this work, we report the recent development of single Au atoms supported on Co3O4 as an extremely active catalyst for CO oxidation showing total conversion of CO at room temperature with an extremely low Au loading of only 0.05 wt%.
Co3O4 nanocrystallites were synthesized by co-precipitation of Co(NO3)2 and Na2CO3 at room temperature and then calcined at 400 ℃ for 5 h. The isolated Au atoms at 0.05 wt% Au loading were dispersed onto the Co3O4 by adsorption and calcined at 400 ℃ for 5 h (denoted as Au1/Co3O4). Details of the Co3O4 support and Au1/Co3O4 catalysts preparation procedures are presented in the supporting information (SI).
The X-ray diffraction (XRD) pattern of the synthesized Co3O4 support is presented in Fig. S1 in the SI. It shows typical Co3O4 diffraction peaks. The Brunauer-Emmett-Teller (BET) surface area of the synthesized Co3O4 was 35 m2 g−1. Representative high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) images of the Au1/Co3O4 catalyst are presented in Figs. 1 and S2. Isolated single Au atoms (in the circles) dispersed on the Co3O4 surfaces are clearly visible. The low-magnification images show that no Au particles or clusters can be observed on the catalyst. By analyzing many low and high magnification images of different regions of the catalyst, we have concluded that only single Au atoms are present in the catalyst without the presence of any Au clusters/particles.
CO oxidation was performed at a gas composition of 1 vol% CO + 1 vol% O2 balanced with He. The sample (50 mg) was diluted with 100 mg Al2O3, loaded into a quartz tube reactor, and the reactant gas was allowed to pass through the reactor at a total flow rate of 33.3 ml min−1, giving a space velocity (SV) of 40, 000 ml gcat−1 h−1. Fig. 2 plots the CO conversion by the Au1/Co3O4 catalyst as a function of the reaction temperature. For the first cycle, CO oxidation started at room temperature (20 ℃) and achieved total CO conversion at 120 ℃. However, on the second cycle, the CO total conversion occurred at room temperature, suggesting that the Au1/Co3O4 catalyst was activated during the first cycle of the reaction process. Co3O4 is a highly active catalyst for the oxidation of CO [2, 6, 22]. To show that the observed high activity originated from the isolated, supported Au single atoms rather than the Co3O4 support, we used the pure Co3O4 support under exactly the same reaction conditions. As shown in Fig. S3 in the SI, for the first cycle, the CO oxidation started at 50 ℃ and achieved total CO conversion at 160 ℃, suggesting a much lower activity of the support. In the second cycle, the activity increased as well: 30% CO conversion at 20 ℃ and total conversion of CO at 160 ℃. Clearly, the conditioning of the Co3O4 during the first cycle improved its activity significantly, but the Co3O4 support was still much less active than the Au1/Co3O4 catalyst. This comparative study unambiguously demonstrates that the high activity of the Au1/Co3O4 catalyst originates from the Au single atoms supported on the Co3O4 nanocrystallites. Schuth [23] proposed a definition of high activity for CO oxidation on supported Au catalysts. To be considered a high-activity catalyst, a supported Au catalyst with about 1 wt% Au loading should yield at least 75% CO conversion at a space velocity of 60,000 ml gcat−1 h−1 with 1% CO in the feed gas at 20 ℃. Such a definition corresponds to a rate of about 20 mmolCO gcat−1 h−1, or ~2 molco gAu−1 h−1 for a Au loading of 1 wt%. We estimated the specific rate of our Au1/Co3O4 catalyst by subtracting the contribution of the Co3O4 support at 20 ℃ and obtained a value of ~25 molco gAu−1 h−1, which is about ten times greater than the value in Schuth’s definition.
It has been reported that pretreatment by various atmospheres could enhance the activity of Co3O4 [22, 24, 25, 26]. We therefore investigated the effects of pretreatment by various types of gas atmospheres on the performance of the Au1/Co3O4 catalyst in an attempt to enhance the catalytic performance. The results are presented in Fig. S4. It clearly shows that after treatment at 200 ℃ with different gases (pure He, 5% O2/He and 5% CO/He), the CO conversion increased only slightly with respect to the untreated catalyst, suggesting a slight improvement in performance. However, compared with the observed enhancement of the Au1/Co3O4 catalyst after the first cycle, the improvement due to gas treatment is almost negligible. Therefore, this significant improvement in the activity after the first cycle of the reaction should originate from other factors such as the subtle changes of the catalyst (surface) structure during the reaction process.
In our previous work, we found that for iron oxide-supported Au [27, 28, 29, 30, 31] and other metal [32, 33, 34] catalysts, the as-synthesized, non-calcined catalysts possessed higher activities. To investigate the effects of the calcination on the Au1/Co3O4 catalysts, we further tested the as-synthesized catalyst without any heat treatment (denoted as Au1/Co3O4-UC). As shown in Fig. S5, Au1/Co3O4-UC shows a slightly higher activity at 50 ℃ compared with that of the Au1/Co3O4 catalyst for the first cycle. However, at elevated temperatures, the activity is lower and achieved total CO conversion at the higher temperature of 140 ℃. For the second cycle, it demonstrated a slightly lower activity at 20 ℃ compared with the Au1/Co3O4 catalyst, but still realized total CO conversion at 40 ℃. It is worth noting that during the third cycle, the activity did not change significantly, suggesting a good cycle stability of the conditioned Au1/Co3O4 catalysts. This result shows that calcination at elevated temperatures (e.g., 400 ℃ in this work) did not cause a decrease in activity, suggesting that the Au1/Co3O4 catalyst is probably resistant to calcination.
The long-term stability of the catalyst is important in practical applications. We therefore tested the stability of the Au1/Co3O4-UC sample for CO oxidation at room temperature after a third cycle, to avoid activity saturation. As shown in Fig. 3, the catalyst deactivated rapidly in the initial 400-min experiment, with the CO conversion decreasing from 70% to about 40%, while during the next 400 min, the deactivation slowed, with only a 10% decrease in the CO conversion over this time. Supported gold catalysts have been shown to deactivate during catalytic reactions because of sintering of the Au nanoparticles. However, for CO oxidation at low temperatures, the formation and accumulation of carbonates due to CO2 adsorption is usually the main reason for the deactivation [35]. This is generally reversible [3, 36]. To confirm the deactivation of the Au1/Co3O4-UC was reversible, we performed various in situ treatments to regenerate the catalyst. Fig. 3 shows that after treatment in He at 200 ℃ for 30 min, the activity had almost completely recovered (CO conversion of ~65%), suggesting that the deactivation can be reversed. During heat treatment under He, carbonates have been shown to decompose [3, 36]. Fig. 3 also indicates that treatment under 5% H2/He not only did not regenerate the sample but made it worse, which might be due to the formation of H2O during the H2 reduction treatment. Co3O4 catalysts are very sensitive to the presence of H2O [2]. The subsequent 5% O2/He treatment then recovered the CO conversion to 100%, suggesting that the carbonates could be more efficiently removed in an O2 atmosphere. To confirm the effect of H2O, we further added 2% H2O into the reaction gas. As expected, the CO conversion dropped rapidly, demonstrating the severely negative effects of H2O.
In summary, we have successfully developed a highly active Co3O4-supported Au single-atom catalyst with an extremely low Au loading level of 0.05 wt%. This catalyst completely converted CO at room temperature and yielded a specific rate of ~ 25 molco gAu−1 h−1. The experimentally obtained high activity originated from the isolated Au single atoms supported on the Co3O4 nanocrystallites, although the exact catalytic mechanism is not clear at this stage and further study is warranted. The catalyst is susceptible to the accumulation of CO2, but resistant to sintering. To our knowledge, this is the first report to unambiguously demonstrate that a supported Au catalyst with such a low loading of Au can realize the total conversion of CO at ambient temperatures. The extremely low loading of noble metal is critical to reducing the cost of supported noble metal catalysts and therefore makes them more practical and attractive.
The authors acknowledge the start-up fund of the College of Liberal Arts and Sciences of Arizona State University and the use of facilities in the John M. Cowley Center for High Resolution Electron Microscopy at Arizona State University.
Co3O4 was prepared by a co-precipitation method. Typically, 1 mol L-1 Co(NO3)2 6H2O aqueous solution was dropwise added into suitable amount of 1 mol L-1 Na2CO3 aqueous solution at room temperature until the pH value of the solution decreased to about 8.5. After 2-h stirring and aging at room temperature, respectively, the resultant precipitate was fileted and washed with de-ionized water for several times to remove the residual carbonates and nitrates. Then the recovered precipitate was dried at 60 ℃ for 5 h and further calcined at 400 ℃ for 5 h.
The Au atoms were deposited on the Co3O4 catalyst by a simple adsorption method with 0.05 wt%. 1 g the pre-synthesized Co3O4 powders were dispersed in deionized water with rigorous stirring. Appropriate amount of HAuCl4 (AuCl3 dissolved in diluted hydrochloric acid, corresponding to an Au loading of 0.05 wt%) solutions were added dropwise into the Co3O4 solution under stirring. After continuing stirring for 2 h and followed by aging for 2 h, the solution was filtered and washed with deionized water for several times, and then dried at 60 ℃ for 5 h and calcined at 400 ℃ for 5 h. The synthesized catalyst was denoted as Au1/Co3O4. For comparison, a non-calcined catalyst was also prepared and denoted as Au1/Co3O4-UC.
The actual Au loading was determined by inductively coupled plasma spectrometer (ICP-AES) on an IRIS Intrepid II XSP instrument (Thermo Electron Corporation).
X-ray Diffraction (XRD) patterns were recorded on a PW3040/60 X’ Pert PRO (PANalytical) diffractometer equipped with a Cu Kα radiation source (λ = 0.15432 nm), operating at 40 kV and 40 mA. A continuous mode was used for collecting data in the 2θ range from 20° to 80° at a scanning speed of 5° min-1.
Sub-angstrom resolution high-angle annular dark-field (HAADF) scanning transmission electron microscopy (STEM) images were obtained on a JEM-ARM200F TEM/STEM with a guaranteed resolution of 0.08 nm. Before microscopy examination, the catalyst powders were ultrasonically dispersed in ethanol and then a drop of the solution was put onto a copper grid coated with a thin lacey carbon film.
The catalytic performances of the catalysts for CO oxidation were evaluated in a fixed-bed reactor. Typically, 50 mg of the catalyst diluted with 100 mg Al2O3 was loaded in a straight quartz reactor. After purged with He for 10 min, the feed gas containing 1 vol% CO, 1 vol% O2 and balance He was allowed to pass through the reactor at a flow rate of 33.3 ml/min, corresponding to a space velocity (SV) of 40000 ml g-1cat h-1. The effluent gas compositions were on line analyzed by a gas chromatograph (HP 7890A) equipped with an Agilent SC-ST 80/100 column and a thermal conductivity detector using He as carrier gas. The CO conversion rates were calculated based on the difference between inlet and outlet CO concentrations.
2015, Vol. 36 Issue (9): 1505-1511
PDF (993KB)
2015, Vol. 36 Issue (9): 1505-1511
PDF (993KB)