Comment on “Catalytic Activity of the Rh Surface Oxide: CO Oxidation over Rh(111) under Realistic Conditions”

In their most recent publication, Gustafson et al.1 suggest the importance of a surface oxide phase for high CO2 formation in CO-O2 reactions. However, a close inspection of their data in Figure 2 of their paper shows clearly that the surface oxide phase, formed under O2/CO-rich reaction conditions, contributes negligibly to CO2 formation. At the reaction time of 1200 s (∼230 °C) in Figure 2b, a sudden increase in the CO2 formation rate is apparent while the SXRD feature in Figure 2d, corresponding to the surface oxide phase, is at most barely detectable. Obviously, during the rapid increase in the CO2 formation rate between 170 and 230 °C, the catalytic surface corresponds to reduced (metallic) Rh, not an oxide phase. From reaction time 1200 to ∼2500 s, the SXRD feature of the surface oxide phase intensifies and saturates (Figure 2d) while the CO2 formation rate remains essentially constant (Figure 2b). As shown and discussed extensively in our previous publications,2-6 this plateau region in the reaction rate is due to mass transfer limitations. That is, an increase in the temperature has no effect on the CO2 formation rate, since the rate is now entirely limited by the arrival of CO, the limiting reagent in an oxygen-rich reactant environment (by oxygenrich, we are referring to O2/CO reaction mixtures greater than the stoichiometric ratio). The consequence is that within this plateau region for oxygen-rich reaction mixtures, the nearsurface gas phase composition is essentially pure O2. Obviously, heating Rh in pure oxygen to T ) 230 °C and above will lead to the formation of surface Rh oxide, as is indicated by the SXRD data of Figure 2. The authors conclude that the rate within the plateau of Figure 2b is due to reaction catalyzed by the surface Rh oxide. As pointed out by us previously,7 this is an incorrect conclusion. Within the so-called “oxide” region (plateau region) and with an increase in the temperature, a proportionally smaller fraction of the total surface in the reduced metallic form is required to sustain the CO2 formation rate in this mass transfer limited regime. For example, at T ) 300 °C less than 1% of the surface sites need be in the reduced state to account for the CO2 conversion at the rate measured in the plateau region. As the temperature is reduced in the data of Figure 2 (from ∼2500 s and above), the oxide is maintained until ∼230 °C and the reverse behavior of that seen during the temperature increase is observed. Again, this is simply because the oxide is reduced to the metallic state with the reaction regime following a normal Arrhenius behavior. As for the results shown in Figure 3 (shown below) of the authors’ paper, the correct interpretation is obvious in view of our earlier data previously discussed in detail.2-6 For a stoichiometric reaction mixture (Figure 3c), the rate data display Arrhenius behavior until the rate reaches the mass transfer limit. We note that the mass transfer limit in the present study is considerably restricted (TOF ∼ 25) by the overall pressure (due to the He makeup gas used to bring the overall pressure to 500 mbar) whereas in our previous studies, where no makeup gas was used, the mass transfer limit (turnover frequency (TOF) or CO2 molecules produced per surface site per second) achieves a value of ∼1000. This effect is related solely to the lower total pressure of the reactants, allowing higher rates to be measured. This is important to note in that the ability to “see” the socalled hyperactive states (reaction rate “jump”) found in our previous studies2-5 is facilitated by having a pure reactant boundary layer and a relatively low reactant pressure to minimize mass transfer limitations and thereby maximize the measurable rate. In any case, the authors’ data do, in fact, show a limited “jump” in the rate for oxygen-rich reactant conditions in Figure 3a,b; this is observed where the reaction rate deviates from the Arrhenius red line, prior to the mass transfer “plateau” region. These rapid rises in the rate indicate the approach to the mass transfer limit, where the reactant mixture in the near surface region becomes increasingly rich in oxygen, that is, starved of the rate-limiting reagent carbon monoxide. The CO2 formation rate, in turn, responds to this rapid increase in the O2-CO ratio as predicted by our previous studies2 with the rate rapidly rising to the mass transfer “plateau” region. We emphasize that this behavior, where the reaction rate rises (“jumps”) sharply with temperature, only occurs and is only possible in an oxygen-rich environment and is not exhibited by a stoichiometric or reducing reaction environment, as that found in Figure 3c of the authors’ paper. We also note and reiterate that this “jump” of reaction rate in Figure 3b would correspond to the region of Figure 2d (before ∼1200 s), which shows barely detectable SXRD oxide intensity. * To whom correspondence should be addressed. Tel.: +1 979 845 0214. Fax: +1 979 845 6822. E-mail address: [email protected]. † Current address: School of Chemical Engineering and Bioengineering, P.O. Box 642710, Washington State University, Pullman, WA 99164-2710. Figure 2. From Gustafson et al.1 J. Phys. Chem. C 2010, 114, 22369–22371 22369