Rate and Selectivity Enhancements Mediated by OH Radicals in the Oxidative Coupling of Methane Catalyzed by Mn/Na2WO4/SiO2This study was supported by BP as part of the Methane Conversion Cooperative Research Program at the University of California at Berkeley

ion increases the {(k1+ k 0 2)/(k 0 3+k 0 4)} ratios markedly, from 0.03 for surface-mediated pathways to 0.14 for OHmediated pathways (Table 1). H2O also leads to lower k 0 5/k 0 1 ratios (4.3 without H2O; 0.63 for OH-mediated pathways; Table 1), which is consistent with the preferential enhancement of CH4 over C2H4 activation rates by OH-mediated routes. Surfaces oxidize C2H4 (k 0 5) more effectively than CH4 (k 0 2) or C2H6 (k 0 4), in spite of the strong C H bonds in C2H4 (463 kJmol ), apparently because C2H4 is strongly adsorbed. [23] OH-mediated pathways do not involve adsorption and lead to k5/k 0 1 values below unity (0.63). These ratios faithfully reflect the stronger C H bonds in C2H4, without compensation by adsorption energies. In fact, these previously unrecognized OH-mediated pathways are essential to achieving the maximumC2 yields reported (up to 26%), which could not be reached with contributions from surface-mediated pathways only. We have achieved C2+ yields of 26% by replenishing O2 after it is depleted in order to avoid explosive reactant mixtures (Figure 4). These yields match the highest values Figure 2. Plot of incremental differential CH4 conversion rate (obtained from measured differences in rates with and without H2O) vs. P 1=4 O2 P H2O. *: steady state reaction; &: H2O added. The arrows indicate the increase in contact time (0.02 g, 1073 K, unit volume: 550–650 mL, 10.7 kPa CH4, CH4/O2 6:1, 0.4 kPa H2O (when added), 101 kPa total pressure, balance He). Table 1: First-order rate constants (mmolg 1 s 1 kPa ) for the steps shown in Scheme 1 (0.02 g, 1073 K, 10.7 kPa CH4, 1.7 kPa O2, 0.4 kPa C2H6/ C2H4, 0 or 0.4 kPa H2O). Rate constant Surface-mediated OH-mediated k1 0.05 0.16 k2 (k 0 2/k 0 1) 0.01 (0.25) 0.02 (0.11) k3 [b] (k3/k 0 1) 1.7 (33) 1.1 (6.8) k4 [b] (k4/k 0 1) 0.14 (2.7) 0.12 (0.73) k5 [c] (k5/k 0 1) 0.22 (4.3) 0.10 (0.63) (k1+ k 0 2)/(k 0 3+ k 0 4) 0.03 0.14 [a] Calculated from rate differences with and without H2O; [b] calculated from CH4/O2/C2H6(/H2O) mixtures; [c] calculated from CH4/O2/ C2H4(/H2O) mixtures. Figure 3. Ratios of rate constants kRCH4/k R C2H6 for H-abstraction from CH4 relative to C2H6 for various abstracting entities (R) vs. DH for the recombination reaction R + H ! R H at 1073 K. Angewandte Chemie 7691 Angew. Chem. Int. Ed. 2008, 47, 7689 –7693 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.angewandte.org reported previously. A kinetic model based on Scheme 1 and rate constants for surfaceand OH-mediated pathways (Table 1; see the Supporting Information for further details) accurately describes the measured yields and predicts a maximum possible C2 yield of 29%. Figure 4 shows that the rate constants for surface-mediated pathways, which prevail in the absence of H2O, cannot account for observed C2 yields and predict maximum values of only 15%. In summary, this study provides mechanistic evidence that OH-mediated C H bond activation pathways are essential for attaining practical yields and describing the evolution of C2 yields during catalytic reactions. The ability of oxide catalysts to generate equilibrium OH-radical concentrations provides opportunities to exploit pathways mediated by such radicals in related chemistries. Our data provide compelling evidence for the benefits of using more reactive species (or higher temperatures) to weaken the sensitivity of H-abstraction reactions to C H bond energies, a challenge and hurdle that limits the maximum attainable yields of the desired products in most practical applications of oxidation catalysis. Experimental Section SiO2 (Davison chemical, Silica Gel Grade 57) was impregnated with aqueous Mn(NO3)2 (50 wt.%, STREM Chemicals; 2 mLg 1 of SiO2) and the mixture dried in ambient air at 403 K for 5 h. This sample was then impregnated with an aqueous solution of Na2WO4·2H2O (99%, Sigma–Aldrich, 2 mLg 1 of SiO2) to give a sample containing 2 wt.% Mn and 5 wt.% Na2WO4. This sample was dried at 403 K for 5 h and then heated in flowing dry air (Praxair, UHP, 0.167 mLs ) at 1173 K (temperature increase: 0.033 Ks ) for 8 h. The samples were sieved to retain 0.25–0.35-mm aggregates. OCM rates and selectivities were measured in flow or recirculating batch reactors using a U-shaped quartz cell (4 mm I.D.). Samples (0.02 g) were mixed with quartz powder (0.5 g; Fluka, SiO2, 0.25–0.35 mm) and held onto quartz wool. The temperature was maintained with a Watlow controller (Series 982) coupled to a resistively heated furnace and measured with a type K thermocouple set outside the catalyst bed. CH4 (Praxair, 99.999%) and O2 (Praxair, 99.999%) were introduced with He (Praxair, 99.999%) as diluent. In batch experiments, the recirculation loop (275–650 mL) was evacuated to < 0.1 Pa before introducing the reactants, which were circulated with a graphite gear micropump (> 2.5 mLs ). H2O was removed from the reactor loop using a dry ice/acetone trap, which does not condense other products. Reactant and product concentrations were measured with an HP5890 gas chromatograph using a Carbosieve SII packed column (Supelco, 3.2 mmJ2 m) with thermal conductivity detection and a HP-PLOT Q capillary column (Agilent, 0.32 mmJ30 m) with flame ionization detection. Differential rates were obtained from time-derivatives of CH4 concentration profiles vs. time measured in batch reactors after regression to a polynomial fit. Selectivities are reported on a carbon basis as cumulative (integral) values. CD4 (Isotec, 99 atom%-D) and D2O (Cambridge Isotope Laboratories, Inc., 99.9%) were used to measure kinetic isotope effects. Tracer studies used labeled CH4 (Isotec, 99 atom%C) in the presence of C2H6 (Praxair, 99.999%) or C2H4 (Praxair, 99.999%). These isotopic measurements were carried out in a batch recirculating reactor equipped with two HP5890 gas chromatographs, with combined thermal conductivity, flame ionization and mass selective detectors. The latter was connected to an HP-PLOT Q capillary column used for isotopic detection. Received: June 4, 2008 Published online: September 2, 2008 .