Group Members
Nathan S. Froemming
Introduction
In 1987, Haruta et al. dicovered that gold nanoparticles are highly catalytically active for CO oxidation (CO+½O2→CO2) at temperatures as low as 200 K. This result is quite remarkable considering that gold is the most noble of all transition metals, excluding mercury, and thereby should not be particularly reactive toward any reaction. Haruta's discovery has since served as motivation for a large amount of scientific research, including the experiment perfomed by Goodman et al. shown below. In this experiment, the catalytic activity of titania-supported gold nanoparticles was monitored at 350 K as a function of nanoparticle size, and a peak was observed in CO oxidation reactivity for nanoparticle diameters of approximately 2-5 nm. In the limit of large nanoparticle size, i.e. bulk gold, the reaction has completely ceased.
Interestingly, catalytic CO oxidation over gold nanoparticles requires the presence of moisture to proceed at appreciable rates. Since the {111} facet is the most stable and prevalent configuration of gold nanoparticles, Mullins et al. have examined CO oxidation on Au(111) in the presence of water in an effort to gain insight to how CO oxidation takes place on the surface of gold nanoparticles. This experiment is shown below. In (A), a molecular beam of C16O is impinged upon a 16O-precovered Au(111) surface and only C16O16O is produced. In (B), a molecular beam of C16O is impinged upon a H218O-precovered Au(111) surface and no CO2 is produced. Finally, in (C), a molecular beam of C16O is impinged upon a 16O and H218O-precovered Au(111) surface and both C16O16O and C16O18O are produced. In all cases, the surface coverages of 16O and H218O were 0.18 and 0.10 monolayers, respectively. The red curve in the QMS signal at the left represents the C16O16O production and the blue curve represents C16O18O production as a function of time after the impingement of C16O. The crux of this experiment can be seen in (C), which suggests that water is activated by atomic oxygen on the Au(111) surface.
We have used density functional theory (DFT) to perform calculations of the elementary steps of CO oxidation on Au(111). Understanding how this reaction proceeds on Au(111) may prove to be very useful in understanding it proceeds on supported gold nanoparticles. Studying this reaction may also lend insight into the water gas shift reaction, CO+H2O→CO2+H2, which is a part of the steam reforming process of hydrocarbons commonly used to generate large amounts of H2 (e.g., CH4+H2O→CO+3H2). Lastly, CO is a poison that binds irreversibly to human heme, and so it is a worthwhile endeavor to understand different ways of getting rid of it (the fact that CO is not only a poison but also a byproduct of the reactions that occur in internal combustion engines necessitates the use of catalytic converters in automobile exhaust).
CO oxidation in the absence of water on Au(111)
We have calculated the pathway shown in (A) of the above figure using DFT. First, however, we make a remark about why atomic oxygen is deposited on the Au(111) surface as opposed to molecular oxygen in the experiments shown above. DFT reveals that the barrier for O2 dissociation is 0.77 eV, whereas the binding energy of an O2 molecule is only 0.19 eV. Thus, O2 desorbs before it dissociates on Au(111), hence the experimenters' need to deposit atomic oxygen on the surface when monitoring CO oxidation. Once atomic oxygen has been deposited on the surface and CO is introduced into the system, CO oxidation proceeds with a barrier of 0.25 eV as calculated by DFT. This reaction pathway is shown below. In this pathway, the oxygen atom is bound to an fcc site (the most stable site for O*) and the CO molecule is bound to a nearby bridging site (one of the most stable binding sites for CO*). CO* then migrates to O* and the two react to form CO2. Such a process is active at ~100 K, in reasonable agreement with experimental observations for 77 K (one must remember that these experiments employ the use of a molecular beam of CO, and the kinetic energy of the incoming CO molecules is not taken into account in our DFT calculations).
The role of water in CO oxidation on Au(111)
We now turn to the question of how exactly water participates in the oxidation of CO on Au(111). We will be discussing three important mechanisms, the first of which is shown below. In this mechanism, water hydrogen bonds to the oxygen atom as CO2 is formed. Since a hydrogen atom is not transferred in the reaction (which we will discuss below), this pathway is called Mechanism I: No Hydrogen Transfer. It is important to note the similarities of this mechanism with the one shown above that does not involve water. Close comparison of these two mechanisms reveals that water is simply acting as a spectator molecule. Since the initial state is stabilized by the hydrogen bond between H2O and O, yet the geometry of the transition states is essentially unchanged, the barrier for the reaction shown below is slightly higher than the barrier for the reaction shown above that does not involve water (0.33 eV vs. 0.25 eV).
Hydrogen "scrambling" in the reaction O*+H2O*↔2OH*
In order to address how water is involved in CO oxidation on Au(111) we first used DFT to determine whether OHs could form on the Au(111) surface in the reaction of O* with H2O. This mechanism is shown below. The highest barrier encountered in this barrier is only 0.11 eV, corresponding to a thermal activation temperature of ~45 K. Moreover, the initial state and the final state of the reaction are nearly equal in energy, indicating that this reaction is both fast and reversible on the Au(111) surface.
Can CO react with a single OH molecule on Au(111)?
We have shown that the barrier for CO oxidation is slightly higher when an intact water molecule is present in the reaction, but can CO react with a single OH molecule to form CO2? The reaction of CO with a single OH on Au(111) is shown below. This mechanism is rather complicated. First, a carboxylate intermediate must form, which has a barrier of 0.32 eV. Then, the carboxylate intermediate must undergo a cis-trans isomerization in which the upward-pointing hydrogen atom flips down to point toward the Au(111) surface, which has a barrier of 0.44 eV. Finally, the hydrogen atom must transfer to the surface for CO2 to be liberated, which has a prohibitively high barrier of 0.93 eV. The bottom line is that the reaction of CO with OH is not a possible mechanism invovling H2O (in the form of hydroxyl) that leads to the formation of CO2 on Au(111) at 77 K.
Can COOH react with OH on Au(111)?
Since CO* reacts rather easily with OH* to form COOH* (A-C in previous figure), we cannot discount the possibility of COOH* reacting with some other adsorbate to form CO2. The reaction COOH*+OH*→CO2+H2O* is considered in the figure below. The most important thing to note about this figure is that the barrier for the reaction is 0.28 eV, which is not really any better than the barrier for CO*+O*→CO2 (0.25 eV). Still, the reaction shown below is thermally active at ~110 K, and could be responsible in part for the C16O18O observed by Mullins et al. We shall call the pathway shown below Mechanism II: Early Hydrogen Transfer because it is the second relevant mechanism we have encountered involving water (in the form of hydroxyl) that leads to the formation of CO2 at 77 K, and because hydrogen transfer has occurred before CO oxidation takes place.
Can CO react with two OH molecules on Au(111)?
We will now discuss the last of the three relevant mechanisms involving H2O that lead to the formation of CO2 on Au(111) at 77 K, namely, the reaction CO*+2OH*→CO2+H2O. This mechanism is shown below, and as can be seen in the figure, it is appropriately labeled Mechanism III: Concerted Hydrogen Transfer. Note the complex OCOHOH configuration along the reaction path - hydrogen transfer occurs simultaneously with CO oxidation. The barrier for this reaction is only 0.11 eV, corresponding to a thermal activation temperature of around 45 K. This is the most important mechanism of the three we have discussed since it involves such a low barrier. Ironically, both hydroxyl formation and Mechanism III occur with a barrier of 0.11 eV, both of which are necessary for H2O to participate in CO oxidation at 77 K.
Goldilocks and the three reaction paths
A comparison of the overall energy landscapes of Mechanisms I-III is shown below. It is important to note that each of the three reaction mechanisms have CO, O, and H2O as reactants and CO2 and H2O as products. From this figure it is clear that the reactants in Mechanism III do not get trapped in an intermediate potential well, hence there being such a small barrier for CO oxidation to take place. Mechanism III is entirely consistent with the experimental observation that CO oxidation readily occurs at 77~K on Au(111) reported by Mullins et al.
We have presented a set of DFT calculations that offers insight as to why water acts as a promoter for CO oxidation on Au(111). Water appears to directly react with adsorbed oxygen atoms, leading to the fast and reversible formation of OH groups on the surface. Once surface hydroxyls have formed, CO can react in a variety of ways with OH molecules to form CO2. We have presented three possible mechanisms in which water, possibly in the form of hydroxyl, can be involved in the oxidation CO at low temperatures. One of these reaction mechanisms, Mechanism III, involves the concerted transfer of hydrogen in CO oxidation and is more than likely the reason that CO oxidation is observed to take place on Au(111) at 77 K in the presence of water. A more formal discussion of that which has been described above can be found in the papers linked below.
Carbonate formation and decomposition on Au(111)
A natural extension of the research discussed above is the reaction CO2+½O2→CO3. The energy landscapes for carbonate formation on Au and Ag (110) and (111) surfaces are shown below.
References
L. Xu, D. Mei, and G. Henkelman,
Adaptive kinetic Monte Carlo simulation of methanol decomposition on Cu(100),
J. Chem. Phys. , (submitted, 2009).
R. A. Ojifinni, J. Gong, N. S. Froemming, D. Flaherty, M. Pan, G. Henkelman, and C. B. Mullins,
Carbonate formation and decomposition on atomic oxygen precovered Au(111)
J. Am. Chem. Soc. 130, 11250 (2008).
R. A. Ojifinni, N. S. Froemming, J. Gong, M. Pan, T. S. Kim, J. M. White, G. Henkelman, and C. B. Mullins,
Water-enhanced low-temperature CO oxidation and isotope effects on atomic oxygen covered Au(111)
J. Am. Chem. Soc. 130, 6801 (2008).
G. Henkelman, A. Arnaldsson, and H. Jónsson,
Theoretical calculations of CH4 and H2 associative desorption from Ni(111): Could subsurface hydrogen play an important role?,
J. Chem. Phys. 124, 044706 (2006).
G. Henkelman and H. Jónsson,
Theoretical calculations of dissociative adsorption of CH4 on an Ir(111) Surface,
Phys. Rev. Lett. 86, 664 (2001).