Copper-Catalyzed Selective Oxidation of Methane by Oxygen: Studies

Aug 12, 2008 - Jong Suk YooJulia SchumannFelix StudtFrank Abild-PedersenJens K. Nørskov. The Journal of Physical Chemistry C 2018 122 (28), 16023- ...
0 downloads 0 Views 196KB Size
13700

J. Phys. Chem. C 2008, 112, 13700–13708

Copper-Catalyzed Selective Oxidation of Methane by Oxygen: Studies on Catalytic Behavior and Functioning Mechanism of CuOx/SBA-15 Yang Li, Dongli An, Qinghong Zhang, and Ye Wang* State Key Laboratory of Physical Chemistry of Solid Surfaces and Department of Chemistry, College of Chemistry and Chemical Engineering, Xiamen UniVersity, Xiamen 361005, P. R. China ReceiVed: March 23, 2008; ReVised Manuscript ReceiVed: June 4, 2008

While copper is the active center of particulate methane monooxygenase in methanotrophic bacteria, there are few studies to utilize synthetic copper catalysts for the selective oxidation of methane by oxygen. In this work, we have found that the copper ions attached on mesoporous silica SBA-15 with high dispersion can catalyze the selective oxidation of methane to formaldehyde by oxygen efficiently. The catalyst with a copper content of 0.008 wt % (Si/Cu ) 13200) exhibits the best catalytic performance, and the specific site rate for formaldehyde formation can reach 5.6 mol (mol Cu)-1 s-1, significantly higher than those reported to date for other catalysts. We have elucidated that the oxidation of methane produces formaldehyde as a major primary product together with a small amount of carbon dioxide, while carbon monoxide is formed mainly via the consecutive oxidation of formaldehyde over our copper-based catalyst. Pulse reaction studies have indicated that methane molecules can react with the lattice oxygen of the catalyst, producing carbon oxides, and CuII in the catalyst is reduced at the same time. Detailed pulse reaction studies combined with EPR characterizations suggest that the reduced copper (probably CuI) sites generated by methane molecules during the reaction account for the activation of molecular oxygen, forming active oxygen species for the selective oxidation of methane to formaldehyde. Introduction Direct conversion of methane to liquid fuels or valuable chemicals has attracted much attention in recent years because of the worldwide demand for the decrease in the dependency on petroleum. The selective oxidation of methane to methanol and formaldehyde in a single catalytic step has been a reaction of particular interest for a long time and is regarded as one of the biggest challenges in catalysis.1 A number of homogeneous and heterogeneous catalysts have been reported for the selective oxidation of methane to organic oxygenates, but none have met the industrial requirements.2 The homogeneous oxidation of methane with use of fuming sulfuric acid catalyzed by HgII salts or PtII complexes could provide methyl bisulphate with high single-pass yields in a long reaction time, but the turnover frequency (TOF) was much lower (∼10-3 s-1).3 Moreover, the consumption of H2SO4 to produce a large amount of SO2 is a big problem. The selective oxidation of methane by oxygen catalyzed by a heterogeneous catalyst is more desirable from the industrial consideration. However, the low productivity of the present catalysts, which are typically supported metal oxides (e.g., MoO3/SiO2 and V2O5/SiO2) and composite metal oxides (e.g., Fe2(MoO3)4 and FePO4), remains the main drawback to their industrial exploitation.2a,e,g,h Moreover, we are still deficient in mechanistic insights for the heterogeneous selective oxidation of methane. In many cases, the lattice oxygen of metal oxides has been proposed for the activation of methane molecules,2a,e,g,h similar to that proposed for the selective oxidation of alkenes or higher alkanes.4 However, a few recent studies have pointed out that the lattice oxygen species may not be responsible for the activation of methane over the MoOx/SiO2 catalyst with high dispersion of molybdenum species.5 * Corresponding author. Phone: +86-592-2186156. Fax: +86-5922183047. E-mail: [email protected].

On the other hand, methane monooxygenases (MMO) catalyze the selective oxidation of CH4 to CH3OH by O2 in methanotrophic bacteria under physiological temperature. Two different forms of MMO, i.e., soluble MMO (sMMO) and particulate MMO (pMMO), are known to exist, and the iron and copper centers dispersed in proteins are believed to be responsible for the selective oxidation of CH4 by O2 in sMMO and pMMO, respectively.6-8 Thus, it would be a promising approach to design and prepare iron- and copper-based heterogeneous catalysts with properly dispersed active sites for the selective oxidation of CH4 by O2. Several research groups have studied iron-based heterogeneous catalysts for the selective oxidation of CH4 to HCHO and found that the highly dispersed iron site accounts for the selective formation of HCHO.2e,h,9-11 The highest specific site rate for HCHO formation based on Fe was ∼1.0 mol (mol Fe) s-1 over the FeOx/SiO2 catalyst prepared by an adsorption-precipitation method.2h On the other hand, few studies have been contributed to the selective oxidation of CH4 by Cu-based catalysts possibly because oxidic copper is well-known for the complete oxidation of CH4.12 In a recent communication, Groothaert et al.13 reported that the chemisorbed oxygen on Cu-ZSM-5, which had been pretreated in O2 at g623 K, could oxidize CH4 to CH3OH at g398 K. However, the reaction could not be operated in a catalytic manner, and the small amount of CH3OH formed by the stoichiometric reaction between the chemisorbed oxygen and CH4 molecules had to be extracted from the surface of catalyst by a water/acetonitrile mixed solvent. Otherwise, the formed CH3OH would undergo further oxidation to CO2 over the Cu-ZSM-5.13 Facile desorption of the formed target product (CH3OH or HCHO) from the solid surface is also a key issue in a heterogeneous catalytic oxidation of CH4. We have clarified that the mesoporous silica, SBA-15, is a superior catalyst support for the selective oxidation of CH4 to HCHO when MoOx, VOx,

10.1021/jp804168y CCC: $40.75  2008 American Chemical Society Published on Web 08/12/2008

Cu-Catalyzed Selective Oxidation of Methane

J. Phys. Chem. C, Vol. 112, No. 35, 2008 13701

TABLE 1: Catalytic Performances of CuOx/SBA-15 Catalysts with Different Cu Contents for Selective Oxidation of CH4 by O2a selectivity (%) catalyst

CH4 conversion (%)

HCHO

CO

CO2

specific site rate of HCHOb [mol (mol Cu)-1 s-1]

SBA-15 0.002 wt % CuOx/SBA-15 0.008 wt % CuOx/SBA-15 0.008 wt % CuOx/SBA-15c 0.03 wt % CuOx/SBA-15 0.08 wt % CuOx/SBA-15 0.3 wt % CuOx/SBA-15 1.0 wt % CuOx/SBA-15 1.5 wt % CuOx/SBA-15

0.30 1.2 3.8 2.3 2.8 2.9 3.3 4.2 4.7

96 71 42 58 44 36 25 19 9

4.0 19 45 32 28 36 35 8 14

0 10 13 10 28 28 40 73 77

5.2 3.1 5.6 0.57 0.17 0.033 0.0095 0.0015

a Reaction conditions: catalyst weight (W), 0.10 g; temperature (T), 898 K; partial pressures (P) of CH4 and O2, 33.8 kPa; total flow rate (F), 120 mL min-1. b Specific site rate was evaluated by the moles of HCHO formed per mole of Cu in the whole catalyst per second (HCHO formed on SBA-15 alone had been subtracted). c W ) 0.05 g.

or FePO4 is used as the active component.14 This is not only because the preparation of highly dispersed active species is possible due to the high surface area of SBA-15 (>500 m2 g-1) but also because the rapid desorption of HCHO becomes facile due to the large mesopores (∼6 nm) and good inertness (without any acidity or basicity) of SBA-15.14 Recently, we investigated the catalytic behaviors of various transition metal ions or oxide clusters (denoted as M, including V, Cr, Mn, Fe, Co, Ni, Cu, Mo, and W) introduced onto SBA-15 with a M/Si molar ratio of 1/13 200, and found that the SBA-15-attached copper catalyst (denoted as CuOx/SBA-15) exhibited the best catalytic performance for HCHO formation.15 This is a significant observation because copper is known to be the active center in pMMO whereas there are few reports on the selective oxidation of CH4 over synthetic Cu-based catalysts. The present paper reports our recent studies on the catalytic behavior and the functioning mechanism of the CuOx/SBA-15 catalyst for the selective oxidation of CH4 to HCHO by O2. Experimental Section Catalyst Preparation. SBA-15 was synthesized by a method described briefly as follows. Triblock copolymer (EO20PO70EO20) and tetraethyl orthosilicate (TEOS), which were used as the template and silicon source, respectively, were dissolved in hydrochloric acid (2.0 mol L-1). The obtained homogeneous mixture was stirred at 313 K for 20 h, and then the resulting milky suspension was further subjected to hydrothermal treatment in an autoclave at 373 K for 24 h. The solid product was recovered by filtration followed by washing with deionized water. SBA-15 was obtained by drying and calcining the obtained powder at 323 K in vacuum and at 923 K in air, respectively. CuOx/SBA-15 was prepared by impregnation of the calcined SBA-15 powder with an ethanolic solution of Cu(NO3)2. After ethanol was evaporated at 343 K, the sample was calcined in air at 923 K for 6 h. Catalytic Reaction. The catalytic reactions were performed on a fixed-bed quartz reactor operated at atmospheric pressure. The catalyst was pretreated in the reactor with a gas flow containing He (20 mL min-1) and O2 (20 mL min-1) for 40 min at 923 K. For the flow-mode reactions, after the catalyst was cooled to the reaction temperature (773-898 K), the reactant gas mixture of CH4 and O2 diluted with He was introduced into the reactor to start the reaction. For the pulsemode reactions, He typically with a flow rate of 80 mL min-1 was used as the carrier gas, and the volume of pulse was typically fixed at 0.63 mL (STP). The products were analyzed by online gas chromatography, and all the lines and values between the exit of the reactor and the gas chromatographs were

heated to 393 K to prevent the condensation of products. CH4 conversion was calculated on a carbon basis from the concentrations of the products detected (i.e., CH3OH, HCHO, CO, and CO2) and the remaining CH4. HCHO selectivity was calculated by the fraction of the amount (in mol) of HCHO in the total amount (in mol) of all the products detected. Catalyst Characterization. X-ray diffraction (XRD) measurements were carried out on a Panalytical X’pert Pro Super X-ray diffractometer with Cu KR radiation (40 kV and 30 mA). N2 physisorption studies were performed at 77 K with a Micromeritics Tristar 3000 surface area and porosimetry analyzer. The sample was pretreated at 573 K in vacuum for 3 h before N2 adsorption. The surface area was calculated by the BET method, and the pore-diameter distribution was evaluated by the BJH method. Electron paramagnetic resonance (EPR) measurements were performed on a Bruker EMX EPR spectrometer at X-band frequency (9.46 GHz). Typically, a fixed amount of sample after pretreatment under different conditions was transferred into a quartz tube instantly, and the quartz tube was set in a quartz Dewar vessel in the EPR cavity. The temperature was set at 100 K throughout the experiment. Results and Discussion Catalytic Behavior of CuOx/SBA-15 Catalyst with Different Cu Contents. Table 1 shows the catalytic performances of the CuOx/SBA-15 catalysts with different Cu contents for the selective oxidation of CH4 by O2 at 898 K. SBA-15 alone gave a low CH4 conversion (0.30%) although HCHO selectivity was high (96%). The presence of a small amount of copper (0.002 wt%, Si/Cu ) 52 900) significantly enhanced CH4 conversion, and a rise in Cu content from 0.002 to 0.008 wt % (Si/Cu ) 13 200) further increased CH4 conversion. However, a further increase in Cu content from 0.008 to 0.3 wt % did not increase or rather decrease CH4 conversions. HCHO selectivity decreased with increasing Cu content. As Cu content rose to g1.0 wt %, CO2 selectivity increased dramatically although CH4 conversion became slightly higher. We further compared HCHO selectivity of catalysts with different Cu contents at different CH4 conversions, which had been regulated by varying reaction temperatures. Figure 1 illustrates that HCHO selectivity decreases with increasing CH4 conversion over all of the catalysts. From Figure 1, it becomes quite clear that the 0.008 wt % CuOx/ SBA-15 exhibits the best catalytic performance for HCHO formation. We have evaluated the specific site rate for HCHO formation by calculating the moles of HCHO formed per mole of Cu in the whole catalyst per second. Note that HCHO produced on SBA-15 alone has been subtracted. As summarized in Table 1,

13702 J. Phys. Chem. C, Vol. 112, No. 35, 2008

Figure 1. HCHO selectivity versus CH4 conversion over CuOx/SBA15 catalysts with different Cu contents. Reaction conditions: W ) 0.10 g, T ) 773-898 K, P(CH4) ) P(O2) ) 33.8 kPa, F ) 120 mL min-1.

Figure 2. XRD patterns for CuOx/SBA-15 catalysts with different Cu contents: (A) low diffraction angles and (B) high diffraction angles.

the specific site rate for HCHO formation also decreased significantly as Cu content exceeded 0.008 wt %. A specific site rate of 5.6 mol (mol Cu)-1 s-1 could be obtained over the 0.008 wt % CuOx/SBA-15 catalyst. This value is higher than the best one achieved over the “single-site” iron (∼1.0 s-1)2h or the “single-site” vanadium (∼0.5 s-1)16 attached on silica or SBA-15 under similar conditions. Therefore, the present CuOx/ SBA-15 catalyst is very active for the selective oxidation of CH4 to HCHO by O2. Characterization of Fresh Catalysts. The ordered mesoporous structure of SBA-15 was characterized by XRD and N2physisorption studies. The diffraction peaks at 2θ of ∼1.0°, 1.7°, and 2.0 °, assigned to the (100), (110), and (200) reflections of hexagonal arrays of the mesoporous structure of SBA-15, were observed for the CuOx/SBA-15 samples with different Cu contents (Figure 2A), indicating that the ordered mesoporous structure was sustained after the incorporation of copper species. All the CuOx/SBA-15 samples exhibited the type-IV isotherms and the type-H1 hysteresis loops, which are typical for mesoporous materials with cylindrical porous channels.17 The surface area and pore volume derived from N2-physisorption decreased slightly with increasing Cu content, and the pore diameter was maintained at 5.6-5.9 nm (Table 2). Thus, it is apparent that the difference in catalytic behavior at different Cu contents (Table 1 and Figure 1) has no relationship with the mesoporous structure but is likely dominated by the state of copper species. XRD patterns at high-diffraction angles (Figure 2B) showed a broad peak at 2θ of ∼22°, which could be attributed to the amorphous framework of SBA-15. When Cu content increased to g1.0 wt %, two weak XRD peaks at 2θ of 35.5° and 38.7°,

Li et al. assignable to the (-111) and (111) reflections of monoclinic CuO, were observed, demonstrating the appearance of crystalline CuO particles. Combining with the catalytic results shown in Table 1, we could conclude that the CuO crystallites predominantly catalyze the formation of CO2. We found that EPR could be used for the characterization of the state of copper in our catalysts with Cu content as low as 0.008 wt %, while other techniques such as XPS, diffuse reflectance UV-vis, Raman, and H2-TPR were all insensitive. Figure 3 shows the EPR spectra of the CuOx/SBA-15 catalysts with Cu contents of 0.008 and 0.03 wt % after the pretreatment in a gas flow containing 50% O2 and 50% He at 923 K, which is the same as that used for catalytic reactions. Four splitting features (mI ) -3/2, -1/2, +1/2, +3/2) were observed in the low-field region for the parallel component due to the hyperfine interaction between the unpaired electron and the nuclear spin of copper (I ) 3/2), whereas the signal for the perpendicular component (g⊥ ) 2.08) was not resolved. These features were typical of Cu2+ cations in axial symmetry,18 and the anisotropic parameters (g| ) 2.41-2.42, A| ) 114-123 G) were in agreement with Cu2+ in octahedral coordination.19 We have estimated the intensity of EPR signals for the samples with different Cu contents, and the normalized EPR signal intensity (the intensity for the 0.008 wt % CuOx/SBA15 sample is normalized to 1.0) for each sample is shown in Figure 4. The signal intensity increased with an increase in Cu content up to 0.3 wt %, and then decreased probably due to the dipole interaction between neighboring paramagnetic species.20 This observation suggests the aggregation of copper species to form larger EPR-insensitive CuOx clusters at higher Cu contents. We have estimated the surface density of Cu atoms by dividing the number of Cu atoms per gram of catalyst (calculated from Cu content) by the specific surface area (with a unit of nm2 g-1), and the Cu content of 0.3 wt % corresponds to a surface density of 0.049 Cu atom per nm2 (Table 2). As shown in Table 1, the selectivity to CO2 increased significantly as Cu content exceeded 0.3 wt %. Thus, it is further confirmed that larger CuOx clusters are responsible for the complete oxidation of CH4 to CO2. However, it should be noted that the best catalytic performance for HCHO formation has been obtained not over the sample with Cu content of 0.3 wt % but over the one with a much lower Cu content (i.e., 0.008 wt %, corresponding to a surface density of 0.0012 Cu atom per nm2). We think that this is possibly because the intensity of EPR signals is not so sensitive to the aggregation of the copper species as the catalytic performance. In other words, we speculate that small oligomeric CuOx species may still contribute to the increase in the intensity of EPR signals but may not be useful in catalytic oxidation of CH4 to HCHO by O2. Thus, it may be concluded that the selective oxidation of CH4 to HCHO requires high dispersion of copper species. Kinetic Studies over the 0.008 wt % CuOx/SBA-15 Catalyst. To gain information about the reaction mechanism, we performed reaction kinetic studies over the 0.008 wt % CuOx/ SBA-15 catalyst, which exhibited the best performance for HCHO formation. From the temperature dependence of catalytic performance (Figure 5A), we can see that, with increasing temperature, the selectivity to HCHO decreases and that to CO increases significantly. CO2 was also formed with a low selectivity even at low temperatures, and the selectivity to CO2 did not change significantly with increasing temperature up to 898 K. The apparent activation energy calculated from the Arrhenius plot (Figure 5B) was 141 kJ mol-1. This value is markedly lower than those (>250 kJ mol-1) for many composite

Cu-Catalyzed Selective Oxidation of Methane

J. Phys. Chem. C, Vol. 112, No. 35, 2008 13703

TABLE 2: Some Physical Properties of CuOx/SBA-15 Catalysts with Different Cu Contents

a

catalyst

surface area (m2 g-1)

Cu surface densitya (Cu atoms nm-2)

pore vol (cm3 g-1)

pore diameter (nm)

SBA-15 0.008 wt % CuOx/SBA-15 0.08 wt % CuOx/SBA-15 0.3 wt % CuOx/SBA-15 1.0 wt % CuOx/SBA-15 1.5 wt % CuOx/SBA-15

645 617 607 581 577 569

0.0012 0.012 0.049 0.16 0.25

0.80 0.80 0.80 0.77 0.73 0.71

5.8 5.8 5.6 5.9 5.6 5.7

Cu surface density was calculated from the Cu content and the surface area.

Figure 3. EPR spectra at 100 K for CuOx/SBA-15 samples with different Cu contents after pretreatment in a gas flow containing 50% O2 and 50% He at 923 K: (a) 0.008 wt % and (b) 0.03 wt %. Figure 5. Temperature dependence of catalytic performance (A) and Arrhenius plot (B) of the 0.008 wt % CuOx/SBA-15 catalyst for selective oxidation of CH4. Reaction conditions: W ) 0.10 g, T ) 773-898 K, P(CH4) ) P(O2) ) 33.8 kPa, F ) 120 mL min-1.

Figure 4. Normalized EPR signal intensity for CuOx/SBA-15 samples with different Cu contents.

metal oxide catalysts such as FeNbBOx and Fe2(MoO4)3,21,22 and is also lower than those (180-230 kJ mol-1) reported for SiO2-supported MoO3 and V2O5 catalysts.23 The lattice oxygen of metal oxides has been proposed for the activation of CH4 over these catalysts.21-26 On the other hand, the activation energy for the present Cu-based catalyst is quite similar to that for the FeOx/SiO2 catalyst containing “isolated” Fe3+ sites.2h Figure 6 illustrates the effect of contact time (expressed as the ratio of catalyst weight to total flow rate, i.e., W/F) on CH4 conversion and product selectivity at 823, 873, and 898 K. At each temperature, CH4 conversion increased almost proportionally to the contact time. The selectivities to HCHO, CO, and CO2 at different contact times and the three temperatures are plotted against CH4 conversion in Figure 6B. It can be seen that the data points for the selectivity to each product versus CH4 conversion at different contact times and different temperatures can be fitted in one curve (dotted line), indicating that the product selectivity is mainly dominated by the CH4 conversion level over the present catalyst. The selectivity to

Figure 6. Dependence of catalytic performance on contact time (W/ F) over the 0.008 wt % CuOx/SBA-15 catalyst at 823, 873, and 898 K. Reaction conditions: W ) 0.01-0.20 g, P(CH4) ) P(O2) ) 33.8 kPa, F ) 120 mL min-1.

HCHO decreased with increasing CH4 conversion, and the decrease in HCHO selectivity corresponded well to the increase in CO selectivity. On the other hand, the selectivity to CO2 only increased slightly with increasing CH4 conversion under the reaction conditions we investigated. Therefore, we can conclude that HCHO and CO2 are the major and the minor primary products, and CO is mainly formed as the secondary product by the consecutive oxidation of HCHO. Extrapolation to zero

13704 J. Phys. Chem. C, Vol. 112, No. 35, 2008

Li et al. TABLE 3: Reaction of CH4 Pulse with the 0.008 wt % CuOx/SBA-15a selectivity (%) pulse No. CH4 converted (µmol) CH4 conv. (%) HCHO CO CO2

Figure 7. Dependence of catalytic performance on partial pressure of CH4 over the 0.008 wt % CuOx/SBA-15 catalyst at 823 and 873 K. Reaction conditions: W ) 0.10 g, P(O2) ) 33.8 kPa, F ) 120 mL min-1.

Figure 8. Dependence of catalytic performance on the partial pressure of O2 over the 0.008 wt % CuOx/SBA-15 catalyst at 823 and 873 K. Reaction conditions: W ) 0.10 g, P(CH4) ) 33.8 kPa, F ) 120 mL min-1.

CH4 conversion provides HCHO and CO2 selectivities of ∼94% and 6%, respectively. The dependences of activity and product selectivity on partial pressures of CH4 and O2 at 823 and 873 K are plotted in Figures 7 and 8. The linear increase in CH4 conversion rate with P(CH4) (Figure 7A) at both temperatures reveals that the reaction order with respect to CH4 was 1.0. Figure 8A suggests that the reaction order with respect to O2 is between 0 and 1, and the evaluation by plotting ln r(CH4) against ln P(O2) (Figure 8C) provides a reaction order of ∼0.25 with respect to O2 at both temperatures. Thus, the rate equation over the ranges of P(CH4) and P(O2) investigated in the present work could be described by,

r(CH4) ) kP(CH4)P(O2)0.25

(1)

The product selectivities also changed with P(CH4) and P(O2) (Figures 7B and 8B); higher P(CH4) and lower P(O2) favored

1 2 3 1b

pretreatment I (standard): O2-containing He gas flow [P(O2) ) 50.7 kPa] at 923 K f He purge 0.019 0.066 0 56 0.0067 0.024 0 51 trace trace 0.0066 0.023 0 79

21

1

pretreatment II: pretreatment I f CH4 flow at 898 K f He purge f O2 pulse (0.63 mL) f He purge 0.024 0.084 33 15

52

1

pretreatment III: pretreatment I f H2 flow at 898 K f He purge f O2 pulse (0.63 mL) f He purge 0.037 0.13 19 9

72

44 49

a

Reaction conditions: 0.008 wt % CuOx/SBA-15, 1.5 g (Cu, ∼1.8 µmol); CH4 pulse, 0.63 mL (STP) (∼28 µmol); T ) 898 K; He carrier gas, 80 mL min-1. CH4 conversion was calculated by the formed products and the unreacted CH4. b 0.008 wt % CuOx/ SBA-15, 0.5 g (Cu, ∼0.6 µmol).

HCHO selectivity, implying that O2 might participate in the consecutive conversion of HCHO to COx. Over MoO3/SiO2 and V2O5/SiO2 catalysts, the oxidation of CH4 was reported to follow first- and zero-order kinetics with respect to CH4 and O2, respectively.23 The Mars-van Krevelen mechanism involving the oxidation of CH4 by the lattice oxygen in metal oxides (especially, the terminal oxygen species, i.e., ModO or VdO) and the rapid replenishment of the lattice oxygen by O2 was proposed in many studies according to the kinetic results.2e,g,h,24-26 The lower O2 reaction order in our case may also suggest that the activation of CH4 by an active oxygen species is the ratedetermining step over the present CuOx/SBA-15 catalyst. Pulse Reaction Studies with the 0.008 wt % CuOx/SBA15. To clarify whether the lattice oxygen is responsible for the conversion of CH4 to HCHO in our case, we performed reactions of CH4 pulse with the 0.008 wt % CuOx/SBA-15 sample after different pretreatments. For comparison, the reactions of CH4 pulse with pure silica SBA-15 after different pretreatments were also examined, and no product could be detected. Table 3 shows that, after pretreatment procedure I (the same with the standard pretreatment used for flow-mode catalytic reactions), only CO and CO2 were formed during the reaction between CH4 pulse and the CuOx/SBA-15. The decrease in the amount of the CuOx/ SBA-15 sample decreased the conversion of CH4, but no HCHO formation was observed, revealing that the lattice oxygen should not be responsible for the conversion of CH4 to HCHO. By assuming that CuII is reduced to CuI by CH4 to form CO and CO2 with the following reactions

CH4 + 6CuO f 3Cu2O + 2H2O + CO CH4 + 8CuO f 4Cu2O + 2H2O + CO2

∆G898K) -489 kJ mol-1 (2) ∆G898K) -645 kJ mol-1 (3)

we estimate that the reduction degree of CuII at 898 K with two CH4 pulses was ∼9.5%. However, the prereduction of the CuOx/SBA-15 sample by CH4 or H2 gas flow followed by O2 pulse treatment (pretreatment procedure II or III in Table 3) provided HCHO during the subsequent reaction with CH4 pulse. CH4 conversion also became higher as compared with that observed in the case of

Cu-Catalyzed Selective Oxidation of Methane

J. Phys. Chem. C, Vol. 112, No. 35, 2008 13705

TABLE 4: Reactions of (CH4 + O2) Pulse and (CH4 + O2) Flow over the 0.008 wt % CuOx/SBA-15a selectivity (%) pretreatment

P(CH4) (kPa)

P(O2) (kPa)

CH4 conv. (%)

(CH4 + O2) pulse pretreatment I pretreatment I pretreatment I pretreatment I pretreatment I f H2 pulse (0.63 mL) f He purge

63.3 33.8 33.8 6.75 6.75

6.33 6.75 33.8 33.8 33.8

0.21 0.30 1.6 1.5 2.2

(CH4 + O2) flow pretreatment I pretreatment I pretreatment I pretreatment I

63.3 33.8 33.8 6.75

6.33 6.75 33.8 33.8

0.61 1.1 3.3 3.9

HCHO

CO

CO2

72 36 28 6.0 20

0 20 39 31 36

28 44 33 63 44

86 78 48 37

4 11 34 39

10 11 18 24

a Reaction conditions: 0.008 wt % CuOx/SBA-15, 0.10 g (Cu, ∼1.8 µmol); T ) 898 K. For pulse-mode reactions: (CH4 + O2) pulse, 0.63 mL (STP), He was also added in the pulse as a balance to regulate P(CH4) and P(O2); He carrier gas, 80 mL min-1. For flow-mode reactions: F ) 80 mL min-1. CH4 conversion was calculated by the formed products and the unreacted CH4.

pretreatment procedure I. This implies that the oxygen species generated by the activation of molecular oxygen on the reduced copper (possibly CuI) site may be responsible for the formation of HCHO. If we assume that all the HCHO molecules formed in Table 3 stem from the newly generated oxygen species while CO and CO2 are due to the reaction of the lattice oxygen, we may roughly estimate that the ratio of the active oxygen species to lattice oxygen is ∼0.008 (prereduction with H2) or ∼0.009 (prereduction with CH4). We further carried out reactions of (CH4 + O2) pulse over the 0.008 wt % CuOx/SBA-15. The presence of O2 in the pulse remarkably increased both CH4 conversion and HCHO selectivity (Table 4). CH4 conversion increased with increasing the ratio of P(O2)/P(CH4) in the pulse up to 1.0, while HCHO selectivity decreased at the same time. A further increase in the ratio of P(O2)/P(CH4) to 5.0 did not increase CH4 conversion but decreased HCHO selectivity markedly. The pretreatment of the CuOx/SBA-15 with H2 pulse followed by He purge enhanced both CH4 conversion and HCHO selectivity significantly. This observation also indicates the importance of the reduced copper site in HCHO formation. However, on the whole, the performances for HCHO formation obtained in the pulse-mode reactions under either CH4-rich or O2-rich conditions are poorer than those obtained in the flow-mode reactions under the same reaction conditions (Table 4). We found significant increases in CH4 conversion and HCHO selectivity with increasing pulse numbers in the repetitive (CH4 + O2) multipulse reactions over the 0.008 wt % CuOx/SBA-15 (Figure 9). At the same time, the selectivities to CO2 and/or CO decreased. This observation demonstrates that there exists an “induction period” during the repetitive reactions of the (CH4 + O2) pulse over the catalyst. The induction period has not been observed in the flow-mode reactions because the initial reaction stage is difficult to follow in the case of flow-mode reactions. Figure 9 also reveals that the “induction period”, i.e., the pulse number required for reaching the steady state, depends on the ratio of P(O2)/P(CH4) in the pulse. The larger P(O2)/P(CH4) ratio caused significant increases in the “induction period”. For example, ∼6 pulses were needed for attaining the steady state under P(CH4) and P(O2) of 63.3 and 6.33 kPa (Figure 9A), whereas ∼15 pulses were required under P(CH4) and P(O2) of 33.8 kPa (Figure 9C). At a P(O2)/P(CH4) ratio of 5.0 (Figure 9D), the steady state could not be attained after repetitive reactions with 30 pulses. Such a trend allows us to speculate that the “induction period” may result from the reduction of the catalyst surface by CH4 molecules in the pulse. To confirm

Figure 9. Changes in CH4 conversion and product selectivities with pulse numbers during the repetitive reactions of (CH4 + O2) pulses over 0.008 wt % CuOx/SBA-15. Reaction conditions: W ) 0.10 g, T ) 898 K, pulse volume ) 0.63 mL (STP), F(He carrier) ) 80 mL min-1.

Figure 10. Effect of the presence of H2 in the pulse on CH4 conversion and product selectivities over the 0.008 wt % CuOx/SBA-15. Reaction conditions: W ) 0.10 g, T ) 898 K, pulse volume ) 0.63 mL (STP), F(He carrier) ) 80 mL min-1. In pulse: P(CH4) ) 6.75 kPa, P(O2) ) 33.8 kPa, P(H2) ) 1.35 kPa.

this speculation, we have investigated the effect of the presence of a small amount of H2 in the pulse. As shown in Figure 10, the steady state could be achieved after reactions with only ∼7 pulses at a P(O2)/P(CH4) ratio of 5.0 if a small amount of H2 was co-fed in the pulse. Moreover, the performance at the steady state in Figure 10 was very close to that obtained in the flowmode reaction at the same P(CH4) and P(O2) (Table 4).

13706 J. Phys. Chem. C, Vol. 112, No. 35, 2008

Figure 11. EPR spectra at 100 K for the 0.008 wt % CuOx/SBA-15 after different treatments or reactions: (a) in 50% O2 and 50% He at 923 K, (b) reduced by H2 at 898 K for 0.5 h, (c) reduced by CH4 at 898 K for 0.5 h, and (d) after repetitive reactions with 30 (CH4 + O2) pulses at 898 K [P(CH4) and P(O2) in the pulse were 63.3 and 6.33 kPa, respectively (see detailed reaction conditions in Figure 9)].

Therefore, we can conclude that the reduced copper sites are responsible for the selective oxidation of CH4 to HCHO by O2. Characterization of Catalysts after Pulse- and Flow-Mode Reactions. To gain further insight into the state of copper after pulse- and flow-mode reactions, we have carried out pseudo in situ EPR measurements. The 0.008 wt % CuOx/SBA-15 samples after different treatments or reactions were instantly transferred to an EPR tube and the tube was set in a quartz Dewar vessel in the EPR cavity. Figure 11 illustrates that, after the reduction by H2 at 898 K, the signals ascribed to Cu2+ ions almost disappeared (curve b). It is well-known that the reduction of Cu2+ will decrease the intensity of EPR signals belonging to Cu2+ because Cu+ (d10) is EPR silent.19,20,27 Sobczak et al.27 once pointed out that the reduction levels of Cu2+ could be approximately estimated on the basis of the decrease in the intensity of the EPR signals of Cu2+. The remarkable decrease in EPR signals after reduction with CH4 (Figure 11, curve c) indicates that CH4 can reduce CuII in the CuOx/SBA-15 at 898 K, in accordance with the results obtained from pulse reaction studies (Table 3). Moreover, Figure 11 further reveals that the EPR signals attributed to Cu2+ also drop significantly after the repetitive (CH4 + O2) multipulse reactions (curve d). This provides further evidence that the “induction period” observed during the (CH4 + O2) multipulse reactions (Figure 9) stems from the gradual reduction of the copper species. We also observed the decrease in the intensity of EPR signals belonging to Cu2+ species after the flow-mode reactions (Figure 12). Moreover, the increase in the ratio of P(CH4)/P(O2) caused the decrease in the intensity of EPR signals, implying that the reduction degree depended on the ratio of P(CH4)/P(O2). These observations indicate that the reduction of CuII also occurs during the flow-mode reactions. CuII may be reduced to CuI and Cu0, and the results described above cannot completely exclude the possibility of Cu0 although Cu0 seems difficult to form under the flow-mode reaction conditions. Since it is known that IR bands of CO adsorbed on CuI and Cu0 sites are different,28 we have performed FT-IR studies of adsorbed CO over the CuOx/SBA-15 catalysts after different in situ treatments. No IR bands due to the adsorbed CO could be observed over the 0.008 wt % CuOx/SBA-15 catalyst after different treatments probably because Cu content is too low. However, useful information could be obtained over the 0.3 wt % CuOx/SBA-15 catalyst. We observed the appear-

Li et al.

Figure 12. EPR spectra at 100 K for the 0.008 wt % CuOx/SBA-15 after different treatments or reactions: (a) in 50% O2 and 50% He at 923 K; (b and c) after flow-mode reactions at 898 K: (b) P(CH4) ) P(O2) ) 33.8 kPa, (c) P(CH4) ) 63.3 kPa and P(O2) ) 6.33 kPa.

SCHEME 1: Reaction Mechanism for Selective Oxidation of CH4 to HCHO over the CuOx/SBA-15 Catalyst

ance of an intense IR band at 2129 cm-1 after the reduction of the 0.3 wt % CuOx/SBA-15 sample by CH4 at 898 K (Figure S1 in the Supporting Information), and this band could be ascribed to CO adsorbed on CuI sites.28 The absence of IR bands at 2105-2120 cm-1 indicated the absence of Cu0.28 The band at 2129 cm-1 could also be observed after the flow-mode reactions but its intensity became much lower (Figure S1 in the Supporting Information). This suggests the generation of CuI during the reaction over this catalyst. It is reasonable to speculate that a part of CuII may undergo similar reductions to CuI but not to Cu0 over the 0.008 wt % CuOx/SBA-15 during the reaction. Reaction Mechanism. We have shown that the copper species attached on the wall surface of SBA-15 with high dispersion are very active for the selective oxidation of CH4 to HCHO by O2. The highest specific site rate for HCHO formation over the present copper catalyst [5.6 mol (mol Cu)-1 s-1] is significantly higher than those over other catalysts reported to date. On the basis of the results obtained from kinetic measurements, pulse reaction studies, and EPR characterizations, we propose a reaction mechanism in Scheme 1 for the CuOx/SBA15-catalyzed selective oxidation of CH4 to HCHO by O2. Pulse reaction studies and EPR characterizations both demonstrate that CH4 can react with the lattice oxygen associated with CuII species, producing CO and CO2, and simultaneously CuII is reduced. Thus, it becomes quite clear that the lattice oxygen is not the active oxygen species for the selective formation of HCHO. In other words, the present copper-based catalyst does not follow the Mars-van Krevelen mechanism as previously proposed for MoO3/SiO2 and V2O5/SiO2,23-26 and for some composite metal oxides such as FeNbBOx,21 Fe2(MoO4)3,22 and FePO4.29 It is of interest to note that, recently, Bell and coworkers5 have pointed out that the lattice oxygen of the isolated molybdate species supported on silica cannot be the active

Cu-Catalyzed Selective Oxidation of Methane

J. Phys. Chem. C, Vol. 112, No. 35, 2008 13707

oxygen, and instead, a molybdenum peroxide species formed by the reaction of O2 with low concentration of reduced molybdate species is likely responsible for the selective oxidation of CH4 to HCHO. Our pulse reaction studies and EPR characterizations suggest that the reduced copper sites (probably CuI as suggested by FT-IR studies) generated by the reduction with CH4 molecules during the reaction are the active sites for the selective oxidation of CH4 to HCHO by O2. This allows us to propose that molecular oxygen is activated on the CuI site to form an active oxygen species (O*), which accounts for the selective conversion of CH4 to HCHO. It would be of interest to compare the functioning mechanism of the present copper-based heterogeneous catalyst with that of the pMMO, in which copper also plays an essential role in the selective oxidation of CH4 by O2.7,8,30 The pMMO was reported to contain mononuclear and dinuclear copper centers,8 but a recent study supported the trinuclear copper model proposed by Chan and co-workers.7,30 The pMMO exploits NADH, a biological reductant, to catalyze the selective oxidation of CH4 as follows,30

CH4 + NADH + H+ + O2 f CH3OH + NAD+ + H2O (4) In the case of trinuclear copper model, Chan and co-workers30b proposed that the reduced trinuclear CuI sites activate molecular oxygen to form a bis(µ3-oxo)CuIICuIICuIII species, capable of inserting oxygen atom facilely into the C-H bonds. On the other hand, through a DFT and QM/MM study, Yoshizawa and Shiota31 suggested possible oxygen species via activation of molecular oxygen on the reduced mononuclear and dinuclear CuI sites. In the case of the mononuclear copper site, CuI reacted with O2, forming CuII-superoxo species (CuII-OO-), which may undergo further conversions to CuII-hydroperoxo and then to CuIII-oxo species. For the dinuclear copper site, the incorporation of oxygen molecules may form a (µ-η2:η2-peroxo)dicopper species, which is then transformed into a bis(µ-oxo)CuIICuIII species. It is argued that both the mononuclear and the dinuclear copper-oxo species are active enough for the selective oxidation of CH4 under physiological conditions.31 In the present work, we have checked the possibility of the existence of Cu2+ pairs by looking at EPR spectra at half-field.32 We have measured the EPR spectra for the 0.008 wt % CuOx/SBA-15 and 0.3 wt % CuOx/SBA-15 samples from low field region, but we cannot obtain clear signals at half the normal magnetic field intensity (Figure S2 in the Supporting Information), which are characteristic of the CuII ion pairs.32 We still cannot distinguish the dispersion state of copper species in our catalysts. However, we have clearly demonstrated that the reduced copper site (most likely CuI) is responsible for the activation of molecular oxygen to form an active oxygen species for the selective conversion of CH4. This is quite similar to the pMMO system. However, in our system, it is not the NADH but the substrate (CH4) itself that works as the reductant to generate the CuI site. The concept of reductive activation of dioxygen molecule is well-known in both homogeneous and heterogeneous selective oxidation of hydrocarbons, but generally, a sacrificial reductant such as metallic iron or zinc powder, carboxylic acid, CO, or H2 is required.2e,33 Thus, the present system using the reactant as the reductant to activate dioxygen is quite significant. Regarding the nature of the active oxygen species in our system, we speculate that the copper-associated oxygen species proposed above for the pMMO model systems such as CuIIIoxo species might also be possible. It has been reported that the CuII-superoxo species formed by the interaction of mono-

nuclear CuI with O2 cannot be responsible for CH4 activation under physiological conditions because the computed activation energy for the H-atom abstraction from CH4 by it is as high as 155 kJ mol-1.31 However, this species may be workable under the reaction conditions used for our catalyst (T ) 773-898 K). Actually, the activation energy for our catalyst (141 kJ mol-1) is quite close to this value. Efforts undertaken to detect the Cusuperoxo or Cu-hydroperoxo species by Raman spectroscopy are not successful at this moment possibly because of the much lower concentration of active species. Further studies on the nature of active oxygen species and the HCHO formation mechanism are planned in our future work because the conversion of CH4 to HCHO should not be a single step. Conclusions We found that the copper ions attached on mesoporous silica SBA-15 with high dispersion were active for the selective oxidation of CH4 to HCHO by O2. The catalyst with a Cu content of 0.008 wt % (corresponding to 0.0012 Cu atom per nm2) exhibited the best catalytic performance, and the specific site rate for HCHO formation reached 5.6 mol (mol Cu)-1 s-1 at 898 K, which was markedly higher than those for other catalysts reported to date. Kinetic studies with the 0.008 wt % CuOx/SBA-15 catalyst showed that HCHO and CO2 were formed as the major and the minor primary products, respectively, while CO was mainly formed via the consecutive oxidation of HCHO. The ratio of HCHO to CO2 was ∼94/6 at the very initial reaction stage (approaching zero CH4 conversion). The apparent activation energy was 141 kJ mol-1, and the reaction orders with respect to CH4 and O2 were 1.0 and 0.25, respectively, suggesting that the activation of the C-H bond was the rate-determining step. We clarified that the lattice oxygen of the catalyst could react with CH4 pulse at reaction temperatures, producing CO and CO2, and CuII was reduced at the same time. On the other hand, the interaction of the prereduced catalyst with O2 pulse led to HCHO formation in the subsequent reaction with CH4 pulse. The reactions of (CH4 + O2) pulse over the CuOx/SBA-15 catalyst could produce HCHO more efficiently. We observed a unique phenomenon that both CH4 conversion and HCHO selectivity increased significantly with pulse numbers in the repetitive (CH4 + O2) multipulse reactions, revealing the existence of an “induction period” in the oxidation of CH4 to HCHO. We found that the “induction period” depended on the ratio of CH4/O2 in the pulse; the higher ratio of CH4/O2 led to the shorter “induction period”. The addition of a small amount of H2 into the pulse could also significantly shorten the “induction period”. EPR characterizations suggested that CuII underwent partial reductions during both the pulse-mode and the flow-mode reactions. We propose that the reduced copper sites (probably CuI) generated by CH4 molecules during the reaction function as the active centers for the activation of O2, forming active oxygen species for the conversion of CH4 to HCHO. Acknowledgment. This work was supported by the NSF of China (20433030, 20625310 and 20773099), the National Basic Program of China (2003CB615803 and 2005CB221408), the Key Scientific Project of Fujian Province (2005HZ01-3), and the Program for New Century Excellent Talents in Fujian Province (to Q.Z.). Supporting Information Available: FT-IR spectra of adsorbed CO and EPR spectra from low-field region. This material is available free of charge via the Internet at http://pubs.acs.org.

13708 J. Phys. Chem. C, Vol. 112, No. 35, 2008 References and Notes (1) (a) Ross, J. Appl. Catal. 1991, 69, N19. (b) Haggin, J. Chem. Eng. News 1993, 71, 27. (c) Roth, J. F. Appl. Catal., A 1994, 113, 131. (2) See reviews: (a) Pitchai, R.; Klier, K. Catal. ReV. 1986, 28, 13. (b) Crabtree, R. H. Chem. ReV. 1995, 95, 987. (c) Wolf, D. Angew. Chem., Int. Ed. 1998, 37, 3351. (d) Lunsford, J. H. Catal. Today 2000, 63, 165. (e) Otsuka, K.; Wang, Y. Appl. Catal., A 2001, 222, 145. (f) Labinger, J. A.; Bercaw, J. E. Nature 2002, 417, 507. (g) Tabata, K.; Teng, Y.; Takemoto, T.; Suzuki, E.; Ban˜ares, M. A.; Pen˜a, M. A.; Fierro, J. L. G. Catal. ReV. -Sci. Eng. 2002, 44, 1. (h) Arena, A.; Parmaliana, A. Acc. Chem. Res. 2003, 36, 867. (3) (a) Periana, R. A.; Taube, D. J.; Evitt, E. R.; Loffler, D. G.; Wentrcek, P. R.; Masuda, T. Science 1993, 259, 340. (b) Periana, R. A.; Taube, D. J.; Gamble, S.; Taube, H.; Satoh, T.; Fujii, H. Science 1998, 280, 560. (4) See for examples: (a) Sokolovskii, V. Catal. ReV. -Sci. Eng. 1990, 32, 1. (b) Grasselli, R. K. Top. Catal. 2002, 21, 79. (c) Grasselli, R. K. Catal. Today 2005, 99, 23. (d) Ueda, W.; Vitry, D.; Kato, T.; Watanabe, N.; Endo, Y. Res. Chem. Intermed. 2006, 32, 217. (5) (a) Ohler, N.; Bell, A. T. J. Phys. Chem. B 2006, 110, 2700. (b) Chempath, S.; Zhang, Y.; Bell, A. T. J. Phys. Chem. C 2007, 111, 1291. (c) Chempath, S.; Bell, A. T. J. Catal. 2007, 247, 119. (6) Rosenzweig, A. C.; Frederick, C. A.; Lippard, S. J.; Nordlund, P. Nature 1993, 366, 537. (7) Chan, S. I.; Chen, K. H. C.; Yu, S. S. F.; Chen, C. L.; Kuo, S. S. J. Biochemistry 2004, 43, 4421. (8) Lieberman, R. L.; Rosenzweig, A. C. Nature 2005, 434, 177. (9) (a) Kobayashi, T.; Nakagawa, K.; Tabata, K.; Haruta, M. J. Chem. Soc. Chem. Commun. 1994, 1609. (b) Kobayashi, T.; Guilhaume, N.; Miki, J.; Kitamura, N.; Haruta, M. Catal. Today 1996, 32, 171. (10) Arena, F.; Gatti, G.; Martra, G.; Coluccia, S.; Stievano, L.; Spadaro, L.; Famulari, P.; Parmaliana, A. J. Catal. 2005, 231, 365. (11) (a) Wang, Y. Res. Chem. Intermed. 2006, 32, 235. (b) Wang, Y.; Yang, W.; Yang, L.; Wang, X.; Zhang, Q. Catal. Today 2006, 117, 156– 162. (12) Park, P. W.; Ledford, J. S. Appl. Catal., B 1998, 15, 221. (13) Groothaert, M. H.; Smeets, P. J.; Sels, B. F.; Jacobs, P. A.; Schoonheydt, R. A. J. Am. Chem. Soc. 2005, 127, 1394. (14) (a) Yang, W.; Wang, X.; Guo, Q.; Zhang, Q.; Wang, Y. New J. Chem. 2003, 27, 1301. (b) Lin, B.; Wang, X.; Guo, Q.; Yang, W.; Zhang, Q.; Wang, Y. Chem. Lett. 2003, 32, 860. (c) Wang, Y.; Wang, X.; Su, Z.; Guo, Q.; Tang, Q.; Zhang, Q.; Wan, H. Catal. Today 2004, 93-95, 155. (15) Li, Y.; Chen, S.; Zhang, Q.; Wang, Y. Chem. Lett. 2006, 35, 572.

Li et al. (16) Ruddy, D. A.; Ohler, N. L.; Bell, A. T.; Don Tilley, T. J. Catal. 2006, 238, 277. (17) Zhao, D.; Feng, J.; Huo, Q.; Melosh, N.; Fredrickson, G. H.; Chmelka, B. F.; Stucky, G. D. Science 1998, 179, 548. (18) Larsen, S. C.; Aylor, A.; Bell, A. T.; Reimer, J. A. J. Phys. Chem. 1994, 98, 11533. (19) Decyk, P. Catal. Today 2006, 114, 142. (20) Amano, F.; Tanaka, T.; Funabiki, T. J. Mol. Catal. A: Chem. 2004, 221, 89. (21) Otsuka, K.; Komatsu, T.; Jinno, K.; Uragami, Y.; Morikawa, A. in Proceedings of the 9th International Congress on Catalysis, vol. 2, The Chemical Institute of Canada, Ottawa, 1988, pp. 915-922 (22) Otsuka, K.; Wang, Y.; Yamanaka, I.; Morikawa, A. J. Chem. Soc. Faraday Trans. 1993, 89, 4225. (23) (a) Spencer, N. D.; Pereira, C. J. AIChE J. 1987, 33, 1808. (b) Spencer, N. D.; Perieira, C. J. J. Catal. 1989, 116, 399. (24) Amiridis, M. D.; Rekoske, J. E.; Dumesic, J. A.; Rudd, D. F.; Spencer, N. D.; Pereira, C. J. AIChE J. 1991, 37, 87. (25) Irusta, S.; Cornaglia, L. M.; Miro´, E. E.; Lombardo, E. A. J. Catal. 1995, 156, 167. (26) Faraldos, M.; Ban˜ares, M. A.; Anderson, J. A.; Hu, H.; Wachs, I. E.; Fierro, J. L. G. J. Catal. 1996, 160, 214. (27) Sobczak, I.; Decyk, P.; Ziolek, M.; Daturi, M.; Lavalley, J. X.; Kevan, L.; Prakash, A. M. J. Catal. 2002, 207, 101. (28) Hadjiivanov, K. I.; Vayssilov, G. N. AdV. Catal. 2002, 47, 307. (29) Wang, Y.; Otsuka, K. J. Catal. 1995, 155, 256. (30) (a) Chan, S. I.; Wang, V. C. C.; Lai, J. C. H.; Yu, S. S. F.; Chen, P. P. Y.; Chen, K. H. C.; Chen, C. L.; Chan, M. K. Angew. Chem, Int. Ed. 2007, 46, 1992. (b) Chen, P. P. Y.; Yang, R. B. G.; Lee, J. C. M.; Chan, S. I. Proc. Natl. Acad. Sci. USA 2007, 104, 14570. (31) Yoshizawa, K.; Shiota, Y. J. Am. Chem. Soc. 2006, 128, 9873. (32) (a) Aboukais, A.; Bennani, A.; Aissi, C. F.; Wrobel, G.; Guelton, M.; Vedrine, J. C. J. Chem. Soc., Faraday Trans. 1992, 88, 615. (b) Aboukais, A.; Bennani, A.; Aissi, C. F.; Guelton, M.; Vedrine, J. C. Chem. Mater. 1992, 4, 977. (33) See for examples: (a) Burton, D.; H, R.; Doller, D. Acc. Chem. Res. 1992, 25, 504. (b) Wang, Y.; Otsuka, K. J. Catal. 1997, 171, 106. (c) Sen, A. Acc. Chem. Res. 1998, 31, 550. (d) Moro-oka, Y. Catal. Today 1998, 45, 3. (e) Ostuka, K.; Yamanaka, I.; Wang, Y. Stud. Surf. Sci. Catal. 1998, 119, 15. (f) Otsuka, K.; Yamanaka, I. Catal. Today 2000, 57, 71. (g) Yamanaka, I. Catal. SurV. Jpn. 2002, 6, 63. (h) Gao, X. H.; Xu, J. Catal. Lett. 2006, 111, 203.

JP804168Y