Monolayer iron oxides grown on metal substrates have widely been used as model systems in heterogeneous catalysis. By means of ambient-pressure scanning tunneling microscopy (AP-STM), we studied the in situ oxidation and reduction of FeO(111) grown on Au(111) by oxygen (O2) and carbon monoxide (CO), respectively. Oxygen dislocation lines present on FeO islands are highly active for O2 dissociation. X-ray photoelectron spectroscopy measurements distinctly reveal the reversible oxidation and reduction of FeO islands after sequential exposure to O2 and CO. Our AP-STM results show that excess O atoms can be further incorporated on dislocation lines and react with CO, whereas the CO is not strong enough to reduce the FeO supported on Au(111) that is essential to retain the activity of oxygen dislocation lines.

Ultrathin oxide films on metal substrates show substantial differences in their physical and chemical properties compared with their bulk counterparts.1–3 This phenomenon brings out the concept to design inverse oxide/metal catalysts that benefit from the strong oxide–metal interactions (SOMIs).4,5 Elegant examples include iron oxide,6,7 cobalt oxide,8,9 molybdenum oxide,10 silica,11,12 titania, and ceria13 on well-defined metal substrates.14 Among these, iron oxide is one of the extensively explored systems because of its potential application in water–gas shift reactions15,16 and carbon monoxide (CO) oxidation.17,18 For example, key elements for CO oxidation are the active sites to dissociate oxygen19–21 and to anchor CO, which facilitate the interaction of CO with weakly bonded oxygen to form CO2.22 Many reaction pathways have been proposed based on combined surface science characterization tools and density functional theoretical (DFT) calculations, including the Langmuir–Hinshelwood type23 and Mars–van Krevelen (MvK, creating a surface oxygen vacancy)24,25 mechanisms. It is commonly accepted that the oxide/metal interfacial sites are more active than the terraces. According to the preparation and characterization of a series of FeO(111)/Pt(111) films with different island sizes, Fu et al. addressed that the reactivity of CO oxidation increases as a function of the periphery density. Based on DFT calculations, they found that both coordinatively unsaturated ferrous (CUF) sites at the oxide/metal interface and neighboring Pt atoms are the active centers, with the former one to activate oxygen and Pt to anchor CO.26 The active sites have been directly demonstrated by a high-resolution scanning tunneling microscopy (STM) study.27–29 

Oxygen activation is the crucial step in CO oxidation.19,21,30 A variety of mechanisms were proposed, such as the formation of oxygen vacancy31,32 and the assistance of the hydroxyl group in the presence of water.33–36 Most of these models emphasize the importance of boundaries between oxide and metal supports or metal particles on oxide supports, where the lattice oxygen can react directly with CO. Therefore, the generation of oxygen vacancies on the surface of oxide is one crucial step in CO oxidation. Bilayer FeO films were found to form a trilayer O–Fe–O film, where CO can react with the top-surface oxygen forming CO2 and leaving a surface oxygen vacancy that will be replenished through the reaction with gas-phase oxygen.37 In the absence of oxygen, oxidized FeO2−x on Pt(111) is also active for the oxidation of CO which takes place at the interface between reduced and oxidized phases in the oxide thin film (400–450 K). It is the Pt substrate that promotes the formation of the active interface, while the FeO2−x with extra oxygen is essential for providing lattice oxygen to react with CO.38 However, this catalyst barely can be regenerated even in oxygen rich conditions.

Reaction of CO with O2 can produce carbonates on oxide surfaces,39,40 while the possible reduction of FeOx to Fe by CO generates FeCx.41 These species are regarded as poisons that deactivate the catalysts.42 Preparing appropriate catalysts with high efficiency and stability,43 and optimizing reaction conditions are necessary to avoid deactivation. As the catalytic pathways and active sites depend on the pressure and temperature,44 it is imperative to monitor the atomic and chemical structures of catalysts under reaction conditions.45–47 The developments of ambient-pressure surface science techniques, such as ambient-pressure STM (AP-STM)48,49 and ambient-pressure x-ray photoelectron spectroscopy (AP-XPS),20,21,50 have made it possible to bridge the pressure and temperature gap between traditional surface science studies in ultra-high vacuum (UHV) or low temperature and studies under reaction conditions.51 

Here, we use AP-STM to study the oxidation and reduction of FeO islands supported on Au(111) in O2 and CO, respectively. The fresh FeO(111) surface comprising oxygen dislocation lines shows no obvious activity toward 1.5 mbar CO exposure at room temperature (RT). A large number of O atoms can be incorporated after exposing to 10−2 mbar O2 gas that leads to the disappearance of dislocation lines on the FeO(111) surface. Moreover, these dislocation lines are regenerated in 1.5 mbar CO gas. XPS measurements of Fe 2p and O 1s reveal the reversible oxidation and reduction of FeO islands after exposing to O2 and CO, respectively. Control experiments by sequentially exposing FeO islands to O2 and CO show that the dislocation lines can be regenerated by exposing to 10−2 mbar O2. In the present study, in situ AP-STM results allow us to distinguish the stability of iron oxide and the crucial role of surface O species in determining the reactivity.

Au(111) was cleaned by several cycles of Ar+ sputtering (1.0 keV, 15 min) and annealing (800 K, 10 min). Monolayer FeO islands were prepared using the method reported in Ref. 41 by evaporating iron atoms in 1.5 × 10−7 mbar O2 gas at 300 K, followed by annealing at 600 K for 5 min, and cooling down to 300 K in O2 gas.

Both AP-STM and XPS were attached to the preparation chamber, which allowed for the sample transfer within UHV.

AP-STM measurements were performed at 300 K in the batch mode with a commercial Leiden Probe Reactor STM setup.52 The reactor cell is sealed with an inert fluoroelastomer ring, which allows switching the reactor cell between high-pressure (several bars) and UHV (∼10−8 mbar). All of the STM images were acquired in the constant-current mode with a top-to-down scan direction.53 The scanning parameters are indicated in the figure captions.

Low-temperature UHV STM measurements were performed at 77 K using a SPECS JT-SPM interfaced with a Nanonis controller. The sample was imaged in the constant-current mode using an electrochemically etched W tip.

XPS experiments were performed in a separated UHV chamber at a base pressure better than 2 × 10−10 mbar. The XPS (SPECS) was operated using an Al Kα photon source (1486.6 eV) and hemispherical electron analyzer set to a constant pass energy of 50 eV. Spectra were shifted to align the Au 4f7/2 peak to a binding energy of 84.0 eV. The presented XPS data have a step size of 0.05 eV, and Shirley background subtractions have been applied. All of the spectra were recorded at 300 K.

DFT calculations were carried out using the code Cambridge Sequential Total Energy Package (CASTEP) of Materials Studio (MS).54 A plane-wave cutoff energy of 400 eV is used to describe the electronic wave functions. The Perdew–Burke–Ernzerhof (PBE) functional with the generalized gradient approximation (GGA) was used to deal with the exchange-correlation energies.55 The Brillouin-zone integration was performed by a 4 × 4 × 1 Monkhorst-Pack grid. For the FeO (111) surface, O-terminated surface was selected and modeled with a four-layer 2 × 2 super-cell (four Fe layers and four O layers) with the coverage of selected adsorbates of a 1/4 monolayer. The Hubbard-U method was applied for the 3d-electrons of Fe atoms.56 U = 5 eV was used for Fe ions to explore the correlation effects in 3d orbitals. Spin-polarization was considered for all calculations. The adsorbates and upper two layers are relaxed during all the optimization, while other layers are fixed. A vacuum layer of 15 Å is added along the direction perpendicular to the slab to avoid artificial interactions between the slab and its periodic images. The convergence criteria for configuration optimization are set to the tolerance for SCF, energy, and maximum force, with a maximum displacement of 2.0 × 10−6 eV/atom, 2.0 × 10−5 eV/atom, 0.05 eV/Å, and 2.0 × 10−3 Å.

Au(111) is inert to both the CO and O2 dissociation, and it acts as a support for the formation of the FeO thin film. The coordinatively unsaturated sites of FeO on metal surfaces and the critical role of strong oxide–metal interactions in CO oxidation have been extensively explored.4,6,14 In current studies, we mainly focused on the CO reactions with the activated O atoms on the FeO terrace, so we use FeO(111) only without considering the substrate in our calculations. According to our STM and XPS experiments, the oxygen dissociated along the dislocation lines to generate active oxygen atoms. Therefore, it is reasonable to simplify the model by using an individual O atom on the FeO(111) surface to stress the effect of these pre-adsorbed O atoms on CO oxidation.

The transition states (TSs) are searched by complete Linear Synchronous Transit (LST) and Quadratic Synchronous Transit (QST) methods.57 Furthermore, frequency analysis has been employed to ensure the TS with only one imaginary frequency. TS confirmation is made to ensure that it leads to the desired reactant and product.53 

The adsorption energies on the FeO(111) surface are defined as follows:

Eads=Eadsorbate+EslabEadsorbate/slab,
(1)

where Eadsorbate/slab, Eadsorbate, and Eslab are the total energy of the surface with the adsorbate, the energy of the adsorbate, and the energy of the surface, respectively.

For a reaction such as AB → A + B, the reaction energy (ΔH) and energy barrier (Eb) were calculated on the basis of the following formulas:

ΔH=E(A+B)/surfaceEAB/surface,
(2)
Eb=ETS/surfaceEAB/surface,
(3)

where E(AB)/surface is the total energy of the adsorbed AB, E(A+B)/surface is the total energy of the co-adsorbed A/B on the metal surface, and ETS/surface is the total energy of the transition state on the metal surface.

Monolayer iron oxide was prepared by depositing iron on Au(111) in 10−7 mbar O2, followed by cooling the sample in 10−7 mbar O2 to RT. The coverage can be precisely controlled by tuning the deposition time. Figure 1(a) is a large-scale STM image of 0.8 monolayer (ML) FeO on Au(111), whereby 1 ML corresponds to a fully covered FeO film on Au(111). The FeO island exhibits a height of 1.5 Å along with a modulated Moiré pattern of 3.2 nm due to the lattice mismatch. The triangular loops are O adatom dislocation lines, where the Fe atoms are fourfold coordinated.58 As a result, these lines appear bright in STM topography,22,59 as shown in the enlarged STM image of Fig. 1(b). In addition to FeO(111), other metal oxide thin films on well-defined metal substrates also exhibit such a feature, such as CoO bilayer on Au(111)60 and Pt(111),61 which depends on the oxygen pressure. In our control experiments, the iron oxide prepared without cooling in oxygen exhibits no such lines (Fig. S1 of the supplementary material). As the CUF sites are active for O2 dissociation, we prepared FeO islands at low coverage on which the density of O adatom dislocation lines is much higher than that in the large island, as shown in Fig. 1(c). Figure 1(d) is an enlarged STM image taken from the white square in panel (c) that presents the dislocation lines with matched orientation with respect to the high-symmetric direction on the Au(111) substrate. The island marked by “i” was used as a reference to show the sequential scan in the same area (because of the drift during scan in our AP-STM system). The corresponding line profile selected to cross the pristine FeO island presents a thickness of ∼1.5 Å, while the dislocation lines have a height of ∼0.8 Å (Fig. S2). Introducing 1.5 mbar CO into the reactor cell, neither the FeO islands nor the Au substrate shows any morphology change (dislocation lines), even after extending the exposure time [Figs. 1(e) and 1(f)]. This finding is consistent with previous experimental results that bare FeO is inert to CO oxidation37 under the conditions used in temperature-programmed desorption/reaction (TPD, TPR) studies.42,62 The observed dislocation lines are indicated with the ball sketch in Fig. 1(g). These dislocation lines have been reported previously in the case of lattice displacements due to the formation of four-coordinated metal atoms within metal oxide films supported on Au(111)58 and Pt(111).63 Schematic models in Fig. 1(h) depict the fourfold coordinated Fe atom along the dislocation lines and the perfect FeO island with the threefold coordinated Fe atoms.

FIG. 1.

FeO islands on Au(111). (a) LT-STM image of freshly prepared 0.8 ML FeO(111) on Au(111). The inset is a line-scan profile along the arrow across two domains. (b) Enlarged LT-STM image showing dislocation lines. (c) Freshly prepared FeO(111) with dislocation lines, image taken with AP-STM in UHV. (d) Zoomed image taken from the square in c. The arrows indicate the directions of the dislocation lines with an angle of 60°. (e) Same region during exposure to 1.5 mbar of CO. (f) Large-scale image of FeO islands taken in 1.5 mbar CO. (g) Proposed ball model showing the dislocation lines that are composed of four-fold coordinated Fe atoms. The structure is based on previous experiments and theory.27,64,65 (h) Enlarged model showing the threefold and fourfold coordinated Fe atoms. Scanning parameters for (a) and (b) are 1 V, 100 pA and −50 mV, 200 pA, respectively. The sample bias for all the AP-STM images is −0.2 V.

FIG. 1.

FeO islands on Au(111). (a) LT-STM image of freshly prepared 0.8 ML FeO(111) on Au(111). The inset is a line-scan profile along the arrow across two domains. (b) Enlarged LT-STM image showing dislocation lines. (c) Freshly prepared FeO(111) with dislocation lines, image taken with AP-STM in UHV. (d) Zoomed image taken from the square in c. The arrows indicate the directions of the dislocation lines with an angle of 60°. (e) Same region during exposure to 1.5 mbar of CO. (f) Large-scale image of FeO islands taken in 1.5 mbar CO. (g) Proposed ball model showing the dislocation lines that are composed of four-fold coordinated Fe atoms. The structure is based on previous experiments and theory.27,64,65 (h) Enlarged model showing the threefold and fourfold coordinated Fe atoms. Scanning parameters for (a) and (b) are 1 V, 100 pA and −50 mV, 200 pA, respectively. The sample bias for all the AP-STM images is −0.2 V.

Close modal

After CO exposure, the reactor cell was evacuated to UHV (10−8 mbar), followed by introducing 10−2 mbar O2. Figure 2(a) shows the same scanning area before introducing O2 gas. The dislocation lines on FeO disappear simultaneously after O2 exposure. These islands are referred to as FeOx to distinguish them from the freshly prepared films. Meanwhile, the FeOx islands and Au surface are decorated with bright dots [Fig. 2(b)]. These images were recorded under 10−2 mbar O2 at 300 K. Figures 2(c)–2(e) are sequential scans taken after O2 exposure for 3 min, 8 min, and 14 min, respectively. The white and yellow arrows highlight adsorbates on FeO islands and Au(111) with an average size of ∼0.8 nm (Fig. S3 of the supplementary material), respectively. These species are highly mobile at the initial stage of O2 exposure. After 20 mins, the whole surface was in an equilibrium situation without a further change in the morphology, as shown in Fig. 2(f). Compared to the bilayer FeO before O2 exposure, the FeOx islands are still atomically flat but with an increased height of 4.5 Å, suggesting the incorporation of a layer of O atoms.37Figure 2(f) shows the transformation of the dislocation lines into dark-colored lines on each island after O2 exposure. The whole processes demonstrate the oxygen activation along both the edges of FeO islands and the lines, which have also been detected on cobalt oxide islands comprising dislocation lines.64 The morphology of FeOx islands remained unchanged in UHV after the evacuation of O2 [Fig. 3(a)], which is crucial for ex situ XPS measurements.

FIG. 2.

Disappearance of dislocation lines after O2 exposure. (a) Large-scale STM image of FeO/Au(111) after the evacuation of CO. (b) The same area after introducing 10−2 mbar O2; island marked with “ii” is used as a reference to show the morphology change. (c) After 3 min, all the dislocation lines disappeared; the white and yellow arrows indicate the mobile clusters on FeOx islands and Au surface, respectively. (d) Zoomed-in scan shows the disappearance of dislocation lines on “ii.” (e) Exposure to 10−2 mbar O2 for 14 min. White and yellow arrows indicate the bright spots on the FeOx islands and Au surface, respectively. (f) The FeOx islands reach equilibrium after exposure for 20 min under 10−2 mbar O2. The line-scan profile is taken along the red-dashed line in (f). The sample bias for all the images is −0.1 V.

FIG. 2.

Disappearance of dislocation lines after O2 exposure. (a) Large-scale STM image of FeO/Au(111) after the evacuation of CO. (b) The same area after introducing 10−2 mbar O2; island marked with “ii” is used as a reference to show the morphology change. (c) After 3 min, all the dislocation lines disappeared; the white and yellow arrows indicate the mobile clusters on FeOx islands and Au surface, respectively. (d) Zoomed-in scan shows the disappearance of dislocation lines on “ii.” (e) Exposure to 10−2 mbar O2 for 14 min. White and yellow arrows indicate the bright spots on the FeOx islands and Au surface, respectively. (f) The FeOx islands reach equilibrium after exposure for 20 min under 10−2 mbar O2. The line-scan profile is taken along the red-dashed line in (f). The sample bias for all the images is −0.1 V.

Close modal
FIG. 3.

Regeneration of dislocation lines after CO exposure. (a) Large-scale STM image of FeOx/Au(111) after the evacuation of O2. (b) Image taken from the square in (a) after introducing 1.5 mbar CO. The yellow arrows highlight the reappearance of dislocation lines on the FeOx islands. (c) Zoomed-in scan from the up-right of panel (b) showing the reappearance of dislocation lines (yellow arrows). (d) Sequential scan after CO exposure for 10 min; the yellow arrows show more dislocation lines than those in (c). (e)–(h) Island “iv” is used as a reference to show the generation of mobile bright spots [dashed ellipsoids in (f)–(h)]. (i)–(k). After CO exposure for more than 2 h. The AP-STM image shows the stabilization of the bright spots and the regeneration of dislocation lines. (l) Line-scan profile taken along the yellow arrow in (k). The CO exposure time is imposed at the up-left corner of each panel.

FIG. 3.

Regeneration of dislocation lines after CO exposure. (a) Large-scale STM image of FeOx/Au(111) after the evacuation of O2. (b) Image taken from the square in (a) after introducing 1.5 mbar CO. The yellow arrows highlight the reappearance of dislocation lines on the FeOx islands. (c) Zoomed-in scan from the up-right of panel (b) showing the reappearance of dislocation lines (yellow arrows). (d) Sequential scan after CO exposure for 10 min; the yellow arrows show more dislocation lines than those in (c). (e)–(h) Island “iv” is used as a reference to show the generation of mobile bright spots [dashed ellipsoids in (f)–(h)]. (i)–(k). After CO exposure for more than 2 h. The AP-STM image shows the stabilization of the bright spots and the regeneration of dislocation lines. (l) Line-scan profile taken along the yellow arrow in (k). The CO exposure time is imposed at the up-left corner of each panel.

Close modal

In the next stage, we exposed the iron oxide islands to 1.5 mbar CO at RT by monitoring the morphology changes. The FeOx islands after CO exposure are denoted by R-FeOx. Since there are excess O adatoms on the FeOx islands, oxidation of CO is expected.37Figure 3(a) is a large-scale STM image of the FeOx islands after evacuation of O2. Figures 3(b)–3(d) are sequential scanning images under 1.5 mbar CO. The yellow arrows highlight the regeneration of dislocation lines after CO exposure [Fig. 3(d)]. The smaller the island is, the longer the time required to regenerate the dislocation lines, which can be ascribed to size-dependent structural dynamics,65 as can be seen in Fig. 3(d) after CO exposure for 10 min. Meanwhile, highly mobile bright spots reappear on R-FeOx islands and Au(111), as highlighted by dashed ellipsoids in Figs. 3(e)–3(g). A zoomed-in scan of Fig. 3(h) clearly shows the regeneration of dislocation lines (arrows), while the bright dots still can be detected even after CO exposure for more than 100 min [circles in Fig. 3(i)].

Further increasing the CO exposure time, neither a morphology change on the R-FeOx islands nor a mobile spot has been observed after reaching equilibrium [Fig. 3(j)]. The dislocation lines are clearly visible, while the bright spots are still stabilized on each island, as shown in the zoomed-in scan of Fig. 3(k). Furthermore, compared with the thickness of pristine and oxidized FeOx islands, CO exposure leads to an average height of ∼2.5 Å, indicating the removal of O atoms, as plotted in Fig. 3(l) (Fig. S4 of supplementary material). It is noted that all the images used for line profile measurements were taken under UHV at RT after evacuating the reactor cell with a sample bias of −0.1 V and −0.2 V to minimize the effect of local density of electronic states.

Surface species after AP-STM imaging after each reaction step were interrogated by XPS measurements. As shown in Fig. 4(a), the Fe 2p core levels are split into 2p1/2 and 2p3/2 components due to the spin–orbit coupling. The Fe 2p3/2 core level spectra for elemental Fe on Au(111) and FeO(111)/Au(111) are located at 706.9 eV and 709.62 eV, respectively, which agrees with the literature values.42 The asymmetrically broadened Fe 2p3/2 peak of FeO is due to a satellite excitation of the Fe2+ species.66 The Fe 2p peak shifted to 709.07 eV after exposing to 1.5 mbar CO for 1 hr. Since there is no morphological change in the dislocation lines, the lattice oxygen at the oxide/metal interface may be involved directly in the CO oxidation,67 leading to partial reduction of FeO at the edges. The Fe 2p peak shifted to 710.41 eV after 10−2 mbar O2 exposure.

FIG. 4.

(a) Fe 2p, (b) O 1s, and (c) C 1s XPS spectra acquired (from bottom to top): from fresh iron island on Au(111) (only in a); from fresh FeO/Au(111); after CO exposure only; after O2 exposure only; from first O2 exposure, then evacuating the reactor cell, and finally CO exposure; from alternate O2, CO, and O2 exposure; and from two cycles of O2 and CO exposure, with the reaction conditions shown. The O2 and CO pressure used in each cycle is controlled in the range of 1–2 × 10−2 mbar and 1.5–1.9 mbar, respectively.

FIG. 4.

(a) Fe 2p, (b) O 1s, and (c) C 1s XPS spectra acquired (from bottom to top): from fresh iron island on Au(111) (only in a); from fresh FeO/Au(111); after CO exposure only; after O2 exposure only; from first O2 exposure, then evacuating the reactor cell, and finally CO exposure; from alternate O2, CO, and O2 exposure; and from two cycles of O2 and CO exposure, with the reaction conditions shown. The O2 and CO pressure used in each cycle is controlled in the range of 1–2 × 10−2 mbar and 1.5–1.9 mbar, respectively.

Close modal

One probable reason is the formation of a trilayer O–Fe–O because the binding energy is comparable with FeO2 islands on Pt(111) that the peak position of Fe 2p3/2 shifts to 710.1 eV after O2 exposure.29 However, the peak intensity exhibits a minor increase, which is significantly different from the doubling peak intensity of FeO2 on Pt(111). Furthermore, according to previous XPS studies on the O–M–O trilayer, the oxygen atoms at the upper- and lower-side exhibit a lower and higher binding energy, respectively.29 The transformation of FeO to FeO2 is, therefore, probably due to the intercalation of O atoms at the interface between FeO and the metal substrate. Accordingly, the O 1s should have a new peak at higher binding energy.29 A new peak detected at 530.63 eV is higher than that of pristine FeO, and it is assigned to O atoms incorporated on FeO terraces [Fig. 4(b)]. This assignment is consistent with the AP-STM images where the dislocation lines disappeared along with an increase in the height of FeO islands. It is noteworthy from the XPS results that the peak shape and position of Fe 2p3/2 have a minimal change in each cycle after CO exposure (see details in Sec. II). As the iron oxide islands are stable after evacuating the reactor cell, we proceeded to repeat reaction cycles by sequential O2 and CO exposure. O2 exposure leads to increased intensity of the peak at 710.4 eV, whereas CO exposure has no significant impact on the peaks. The binding energy of the O 1s peak is independent of the oxide phase;29,68 the main peak after different reactions still located at 530.26 eV, as shown in Fig. 4(b).

In contrast to the reduction of FeO/Au(111) in CO at elevated temperature and pressures which is deactivated by significant formation of carbonates and FeCx, FeO islands can be regenerated after CO exposure at 300 K and moderate pressures, where no FeCx (indicated by a characteristic feature at 708.5 eV)41,69 is detected by XPS. As shown in Fig. 4(c), the C1s spectrum points to the absence of carbonates as well (290 eV for CO3).41 As the accumulation of carbon species can severely deactivate iron oxide catalysts, it is essential to either suppress the formation of carbon-containing species or remove these adsorbates effectively during reactions.70 As the reaction temperature and pressure of the reactant significantly influence the surface reactions, precise control of pressures (1.5–1.9 mbar CO) and sequential exposure to reactants would be a probable way to avoid the formation of carbon-containing species.

To correlate the reversible oxidation and reduction of FeO(111) supported on Au(111) with the mechanism of CO oxidation, we performed DFT calculations. The surface structure of FeO(111) is depicted in Fig. 5(a), with the oxygen atoms exposed to vacuum.41 We optimized the adsorption configurations of CO, O, and CO2 on the FeO(111) surface and calculated the adsorption energies at different active sites (Table I). Figures 5(b)–5(d) demonstrated the most stable configurations of CO, O, and CO2 on FeO(111), respectively. Both the CO and O atoms prefer to reside at the face-centered cubic (fcc) hollow sites. The calculated adsorption energy of CO is −1.39 eV, in accordance with previous reports (−1.41 eV).71 Individual O atoms have much higher adsorption energy. In the case of CO2, results indicate its weak physisorbtion on the FeO(111) surface with a negligible value of −0.07 eV, suggesting that it can easily desorb from the surface [Fig. 5(d)]. Meanwhile, the pre-adsorption of oxygen on FeO(111) weakened CO adsorption to −1.16 eV. The decrease in energy has a profound impact on the kinetics of CO oxidation. To examine the catalytic activity toward CO oxidation on different surfaces, we explored the MvK mechanism for the oxidation of CO on the bare and O pre-covered FeO(111) surface.

FIG. 5.

Top and side views: (a) optimized FeO(111) surface and adsorption configurations of (b) CO, (c) O, and (d) CO2 on the FeO(111) surface. Carbon, oxygen, and iron atoms were represented as gray, red, and blue balls.

FIG. 5.

Top and side views: (a) optimized FeO(111) surface and adsorption configurations of (b) CO, (c) O, and (d) CO2 on the FeO(111) surface. Carbon, oxygen, and iron atoms were represented as gray, red, and blue balls.

Close modal
TABLE I.

The adsorption sites, adsorption energies, and key parameters for related species on the FeO(111) surface.

SurfaceSpeciesConfigurationsBond lengths (Å)Adsorption energies (eV)
FeO(111) CO fcc, C-bound dC–O = 1.19 −1.39 
 fcc, O-bound dO–O = 2.42 −2.58 
 CO2 Top, away from the surface N. A. −0.07 
SurfaceSpeciesConfigurationsBond lengths (Å)Adsorption energies (eV)
FeO(111) CO fcc, C-bound dC–O = 1.19 −1.39 
 fcc, O-bound dO–O = 2.42 −2.58 
 CO2 Top, away from the surface N. A. −0.07 

Two different reaction pathways were explored by using the adsorption configuration of CO2 as the final state. Figure 6 demonstrated the calculated potential energy surface for the CO oxidation reaction on different surfaces. Initial states (ISs), transition states (TSs), and final states (FSs) of two different reaction pathways are depicted in each stage, starting from the bare FeO(111) surface which followed the MvK mechanism, as shown in Fig. 6 (top row). The adsorbed CO molecule at the fcc hollow site can easily diffuse into the bridge site (TS1). CO reacts with the nearby lattice oxygen when it is adsorbed on the bridge side (green arrow in TS1), forming chemisorbed CO2. Finally, the CO2 weakly physisorbed above the surface (FS1). The reaction is endothermic (0.39 eV), with an energy barrier of 0.75 eV (Table II). The rate-determining step is extracting lattice oxygen, which is a common barrier for many surface reactions involving lattice oxygen.72,73

FIG. 6.

Potential energy profiles of CO oxidation on FeO(111) surfaces. Theoretical models are corresponding to the initial states (ISs), transition states (TSs), and final states (FSs) in two different reaction pathways. The green arrow in TS1 indicates the lattice oxygen in the FeO island, the black dashed circle represents the missing lattice oxygen that participates in the surface reaction to form CO2, and white dashed circles in the bottom reaction path indicate the adsorbed O atom on FeO. Carbon, oxygen, and iron atoms are represented as gray, red, and blue balls.

FIG. 6.

Potential energy profiles of CO oxidation on FeO(111) surfaces. Theoretical models are corresponding to the initial states (ISs), transition states (TSs), and final states (FSs) in two different reaction pathways. The green arrow in TS1 indicates the lattice oxygen in the FeO island, the black dashed circle represents the missing lattice oxygen that participates in the surface reaction to form CO2, and white dashed circles in the bottom reaction path indicate the adsorbed O atom on FeO. Carbon, oxygen, and iron atoms are represented as gray, red, and blue balls.

Close modal
TABLE II.

Energies of the CO oxidation reaction on different FeO(111) surfaces. E(b)f is the energy barrier of the forward reaction; ΔH is the reaction energy.

SurfacesReactionsTSE(b)f (eV)ΔH (eV)
FeO(111) CO + Olattice → CO2 TS1 0.75 0.39 
O–FeO(111) CO + Oads → CO2 TS2 0.26 0.14 
SurfacesReactionsTSE(b)f (eV)ΔH (eV)
FeO(111) CO + Olattice → CO2 TS1 0.75 0.39 
O–FeO(111) CO + Oads → CO2 TS2 0.26 0.14 

Next, we examined the CO oxidation reaction on an O-decorated FeO(111) surface. In our study, to reduce the computational complexity, we modeled the oxygen pre-covered surface by using an individual O atom.44 The O2 dissociation is the rate-determining step in CO oxidation on types of catalysts.74 Formation of strong C–O bond in CO2 overcompensates the energy cost of removing the O atom from the surface. As demonstrated previously, the formation energy of an oxygen vacancy on FeO2/Pt(111) (1.3 eV) is about half of that on the pristine FeO/Pt(111) (2.8 eV).70 In the current case, the dislocation lines activated O2 on the FeO terrace and incorporated O atoms that would reduce the reaction barrier. As depicted in the bottom row of Fig. 6, a CO molecule diffused from the fcc hollow site to the bridge site, which located closer to the adsorbed O atom (TS2, highlighted by the circle). Subsequently, the CO molecule reacted with the O atom, forming chemisorbed CO2. The reaction barrier for this process was calculated to be 0.26 eV with an endothermic value of 0.14 eV (Table II). Clearly, the FeO surface with pre-adsorbed O atoms is more reactive than the bare surface. Note that, in our model, only the reaction at the flat terraces of the FeO film was considered. The CO interaction at the edges of FeO islands even without oxygen pretreatment, revealed by XPS, can occur at a lower energy cost.75 

In summary, we performed AP-STM and XPS to study the oxidation and reduction of FeO(111) islands supported on the Au(111) surface. The disappearance of dislocation lines upon O2 oxidation resulted in the incorporation of excess O atoms on the oxide islands. Exposing CO gas caused the regeneration of dislocation lines that correlated with the promotion of CO oxidation by surface O atoms. DFT calculations compared the reaction barriers of CO on bare and oxygen modified FeO(111) surfaces, revealing a significant energy barrier drop after introducing adsorbed oxygen atoms. Excess oxygen on metal oxide surfaces could, therefore, be used to activate or regenerate actives sites for oxidation of nanocatalysts.

See the supplementary material for STM images of FeO/Au(111).

Y.J. and Y.Z. contributed equally to this work.

The authors declare no competing financial interest.

The authors acknowledge the financial support from the Natural Science Foundation of Jiangsu Province, Grant No. BK20181297. D.S. was supported by the Integrated Mesoscale Architectures for Sustainable Catalysis (IMASC), an Energy Frontier Research Center funded by the U.S. Department of Energy, Office of Science, Basic Energy Sciences, under Award No. DE SC0012573. This research used resources of the Center for Functional Nanomaterials, which is a U.S. DOE Office of Science Facility, at Brookhaven National Laboratory under Contract No. DE-SC0012704.

1.
A.
Picone
,
M.
Riva
,
A.
Brambilla
,
A.
Calloni
,
G.
Bussetti
,
M.
Finazzi
,
F.
Ciccacci
, and
L.
Duò
, “
Reactive metal-oxide interfaces: A microscopic view
,”
Surf. Sci. Rep.
71
,
32
76
(
2016
).
2.
Q.
Fu
and
T.
Wagner
, “
Interaction of nanostructured metal overlayers with oxide surfaces
,”
Surf. Sci. Rep.
62
,
431
498
(
2017
).
3.
A.
Zavabeti
,
J. Z.
Ou
,
B. J.
Carey
,
N.
Syed
,
R.
Orrell-Trigg
,
E. L. H.
Mayes
,
C.
Xu
,
O.
Kavehei
,
A. P.
O’Mullane
,
R. B.
Kaner
,
K.
Kalantar-zadeh
, and
T.
Daeneke
, “
A liquid metal reaction environment for the room-temperature synthesis of atomically thin metal oxides
,”
Science
358
,
332
335
(
2017
).
4.
J. A.
Rodriguez
,
P.
Liu
,
J.
Graciani
,
S. D.
Senanayake
,
D. C.
Grinter
,
D.
Stacchiola
,
J.
Hrbek
, and
J.
Fernández-Sanz
, “
Inverse oxide/metal catalysts in fundamental studies and practical applications: A perspective of recent developments
,”
J. Phys. Chem. Lett.
7
,
2627
2639
(
2016
).
5.
J.
Zhang
,
H.
Wang
,
L.
Wang
,
S.
Ali
,
C.
Wang
,
L.
Wang
,
X.
Meng
,
B.
Li
,
D.
Su
, and
F.
Xiao
, “
Wet-Chemistry strong metal–support interactions in titania-supported Au catalysts
,”
J. Am. Chem. Soc.
141
,
2975
2983
(
2019
).
6.
G. S.
Parkinson
, “
Iron oxide surfaces
,”
Surf. Sci. Rep.
71
,
272
365
(
2016
).
7.
Q.
Fu
,
F.
Yang
, and
X. H.
Bao
, “
Interface-confined oxide nanostructures for catalytic oxidation reactions
,”
Acc. Chem. Res.
46
,
1692
1701
(
2013
).
8.
M. D.
Santis
,
A.
Buchsbaum
,
P.
Varga
, and
M.
Schmid
, “
Growth of ultrathin cobalt oxide films on Pt(111)
,”
Phys. Rev. B
84
,
125430
(
2011
).
9.
J.
Fester
,
M.
García-Melchor
,
A. S.
Walton
,
M.
Bajdich
,
Z.
Li
,
L.
Lammich
,
A.
Vojvodic
, and
J. V.
Lauritsen
, “
Edge reactivity and water-assisted dissociation on cobalt oxide nanoislands
,”
Nat. Commun.
8
,
14169
(
2017
).
10.
I.
Szenti
,
L.
Bugyi
, and
Z.
Kónya
, “
The promotion of CO dissociation by molybdenum oxide overlayers on rhodium
,”
Surf. Sci.
657
,
1
9
(
2017
).
11.
X.
Zhou
,
G. J. A.
Mannie
,
J.
Yin
,
X.
Yu
,
C. J.
Weststrate
,
X.
Wen
,
K.
Wu
,
Y.
Yang
,
Y.
Li
, and
J. W.
Niemantsverdriet
, “
Iron carbidization on thin-film silica and silicon: A near-ambient-pressure X-ray photoelectron spectroscopy and scanning tunneling microscopy study
,”
ACS Catal.
8
,
7326
7333
(
2018
).
12.
C.
Xu
,
G.
Chen
,
Y.
Zhao
,
P.
Liu
,
X.
Duan
,
L.
Gu
,
G.
Fu
,
Y.
Yuan
, and
N.
Zheng
, “
Interfacing with silica boosts the catalysis of copper
,”
Nat. Commu.
9
,
3367
(
2018
).
13.
J.
Graciani
,
K.
Mudiyanselage
,
F.
Xu
,
A. E.
Baber
,
J.
Evans
,
S. D.
Senanayake
,
D. J.
Stacchiola
,
P.
Liu
,
J.
Hrbek
,
J. F.
Sanz
, and
J. A.
Rodriguez
, “
Highly active copper-ceria and copper-ceria-titania catalysts for methanol synthesis from CO2
,”
Science
345
,
546
550
(
2014
).
14.
J.
IngoFlege
and
D. C.
Grinter
, “
In situ studies of oxide nucleation, growth, and transformation using slow electrons
,”
Prog. Surf. Sci.
93
,
21
45
(
2018
).
15.
C.
Ratnasamy
and
J. P.
Wagner
, “
Water gas shift catalysis
,”
Catal. Rev.
51
,
325
440
(
2009
).
16.
M.
Zhu
,
T. C. R.
Rocha
,
T.
Lunkenbein
,
A.
Knop-Gericke
,
R.
Schlögl
, and
I. E.
Wachs
, “
Promotion mechanisms of iron oxide-based high temperature water-gas shift catalysts by chromium and copper
,”
ACS Catal.
6
,
4455
4464
(
2016
).
17.
I.
Ro
,
I. B.
Aragao
,
J. P.
Chada
,
Y.
Liu
,
K. R.
Rivera-Dones
,
M. R.
Ball
,
D.
Zanchet
,
J. A.
Dumesic
, and
G. W.
Huber
, “
The role of Pt-FexOy interfacial sites for CO oxidation
,”
J. Catal.
358
,
19
26
(
2018
).
18.
H.
Chen
,
Y.
Liu
,
F.
Yang
,
M.
Wei
,
X.
Zhao
,
Y.
Ning
,
Q.
Liu
,
Y.
Zhang
,
Q.
Fu
, and
X.
Bao
, “
Active phase of FeOx/Pt catalysts in low-temperature CO oxidation and preferential oxidation of CO reaction
,”
J. Phys. Chem. C
121
,
10398
10405
(
2017
).
19.
H. J.
Freund
, “
Oxygen activation on oxide surfaces: A perspective at the atomic level
,”
Catal. Today
238
,
2
9
(
2014
).
20.
I. N.
Remediakis
,
N.
Lopez
, and
J. K.
Nørskov
, “
CO oxidation on rutile-supported Au nanoparticle
,”
Angew. Chem., Int. Ed.
44
,
1824
1826
(
2005
).
21.
R. A.
Ojifinni
,
N. S.
Froemming
,
J.
Gong
,
M.
Pan
,
T. S.
Kim
,
J. M.
White
,
G.
Henkelman
, and
C.
Buddie Mullins
, “
Water-enhanced low-temperature CO oxidation and isotope effects on atomic oxygen-covered Au(111)
,”
J. Am. Chem. Soc.
130
,
6801
6812
(
2008
).
22.
Q.
Pan
,
X.
Weng
,
M.
Chen
,
L.
Giordano
,
G.
Pacchioni
,
C.
Noguera
,
J.
Goniakowski
,
S.
Shaikhutdinov
, and
H. J.
Freund
, “
Enhanced CO oxidation on the oxide/metal interface: From ultra-high vacuum to near-atmospheric pressures
,”
ChemCatChem
7
,
2620
2627
(
2015
).
23.
R.
Toyoshima
,
M.
Yoshida
,
Y.
Monya
,
K.
Suzuki
,
K.
Amemiya
,
K.
Mase
,
B. S.
Mun
, and
H.
Kondoh
, “
In situ photoemission observation of catalytic CO oxidation reaction on Pd(110) under near-ambient pressure conditions: Evidence for the Langmuir−Hinshelwood mechanism
,”
J. Phys. Chem. C
117
,
20617
20624
(
2013
).
24.
D.
Widmann
and
R. J.
Behm
, “
Activation of molecular oxygen and the nature of the active oxygen species for CO oxidation on oxide supported Au catalysts
,”
Acc. Chem. Res.
47
,
740
749
(
2014
).
25.
D.
Widmann
and
R. J.
Behm
, “
Dynamic surface composition in a Mars-van Krevelen type reaction: CO oxidation on Au/TiO2
,”
J. Catal.
357
,
263
273
(
2018
).
26.
Q.
Fu
,
W.
Li
,
Y.
Yao
,
H.
Liu
,
H.
Su
,
D.
Ma
,
X.
Gu
,
L.
Chen
,
Z.
Wang
,
H.
Zhang
,
B.
Wang
, and
X.
Bao
, “
Interface-confined ferrous centers for catalytic oxidation
,”
Science
328
,
1141
1144
(
2010
).
27.
H.
Zeuthen
,
W.
Kudernatsch
,
L. R.
Merte
,
L. K.
Ono
,
L.
Lammich
,
F.
Besenbacher
, and
S.
Wendt
, “
Unraveling the edge structures of platinum(111)-supported ultrathin FeO islands: The influence of oxidation state
,”
ACS Nano
9
,
573
583
(
2015
).
28.
W.
Kudernatsch
,
G.
Peng
,
H.
Zeuthen
,
Y.
Bai
,
L. R.
Merte
,
L.
Lammich
,
F.
Besenbacher
,
M.
Mavrikakis
, and
S.
Wendt
, “
Direct visualization of catalytically active sites at the FeO-Pt(111) interface
,”
ACS Nano
9
,
7804
7814
(
2015
).
29.
Q.
Fu
,
Y.
Yao
,
X.
Guo
,
M.
Wei
,
Y.
Ning
,
H.
Liu
,
F.
Yang
,
Z.
Liu
, and
X.
Bao
, “
Reversible structural transformation of FeOx nanostructures on Pt under cycling redox conditions and its effect on oxidation Catalysis
,”
Phys. Chem. Chem. Phys.
15
,
14708
14714
(
2013
).
30.
L. M.
Molina
,
M. D.
Rasmussen
, and
B.
Hammer
, “
Adsorption of O2 and oxidation of CO at Au nanoparticles supported by TiO2(110)
,”
J. Chem. Phys.
120
,
7673
7680
(
2004
).
31.
A. R.
Puigdollers
,
P.
Schlexer
,
S.
Tosoni
, and
G.
Pacchioni
, “
Increasing oxide reducibility: The role of metal/oxide interfaces in the formation of oxygen vacancies
,”
ACS Catal.
7
,
6493
6513
(
2017
).
32.
F.
Esch
,
S.
Fabris
,
L.
Zhou
,
T.
Montini
,
C.
Africh
,
P.
Fornasiero
,
G.
Comelli
, and
R.
Rosei
, “
Electron localization determines defect formation on ceria substrates
,”
Science
309
,
752
(
2005
).
33.
Y.
Jin
,
G.
Sun
,
F.
Xiong
,
L.
Ding
, and
W.
Huang
, “
Water-Activated lattice oxygen in FeO(111) islands for low temperature oxidation of CO at Pt−FeO interface
,”
J. Phys. Chem. C
120
,
9845
9851
(
2016
).
34.
J.
Saavedra
,
H. A.
Doan
,
C. J.
Pursell
,
L. C.
Grabow
, and
B. D.
Chandler
, “
The critical role of water at the gold-titania interface in catalytic CO oxidation
,”
Science
345
,
1599
(
2014
).
35.
M. A.
van Spronsen
,
K. J.
Weststrate
, and
L. B. F.
Juurlink
, “
A comparison of CO oxidation by hydroxyl and atomic oxygen from water on low-coordinated Au atoms
,”
ACS Catal.
6
,
7051
7058
(
2016
).
36.
H.
Jeong
,
J.
Bae
,
J. W.
Han
, and
H.
Lee
, “
Promoting effects of hydrothermal treatment on the activity and durability of Pd/CeO2 catalysts for CO oxidation
,”
ACS Catal.
7
,
7097
7105
(
2017
).
37.
Y.
Sun
,
L.
Giordano
,
J.
Goniakowski
,
M.
Lewandowski
,
Z.
Qin
,
C.
Noguera
,
S.
Shaikhutdinov
,
G.
Pacchioni
, and
H. J.
Freund
, “
The interplay between structure and CO oxidation Catalysis on metal supported ultrathin oxide films
,”
Angew. Chem.
122
,
4418
4421
(
2010
).
38.
K.
Zhang
,
L.
Li
,
S.
Shaikhutdinov
, and
H. J.
Freund
, “
Carbon monoxide oxidation on metal-supported monolayer oxide films: Establishing which interface is active
,”
Angew. Chem., Int. Ed.
57
,
1261
1265
(
2018
).
39.
J.
Saavedra
,
C.
Powell
,
B.
Panthi
,
C. J.
Pursell
, and
B. D.
Chandler
, “
CO oxidation over Au/TiO2 catalyst: Pretreatment effects, catalyst deactivation, and carbonates production
,”
J. Catal.
307
,
37
47
(
2013
).
40.
Y.
Denkwitz
,
Z.
Zhao
,
U.
Hörmann
,
U.
Kaiser
,
V.
Plzak
, and
R. J.
Behm
, “
Stability and deactivation of unconditioned Au/TiO2 catalysts during CO oxidation in a near-stoichiometric and O2-rich reaction atmosphere
,”
J. Catal.
251
,
363
373
(
2007
).
41.
L.
Yu
,
Y.
Liu
,
F.
Yang
,
J.
Evans
,
J. A.
Rodriguez
, and
P.
Liu
, “
CO oxidation on gold-supported iron oxides: New insights into strong oxide-metal interactions
,”
J. Phys. Chem. C
119
,
16614
16622
(
2015
).
42.
W.
Weiss
and
W.
Ranke
, “
Surface chemistry and catalysis on well-defined epitaxial iron-oxide layers
,”
Prog. Surf. Sci.
70
,
1
151
(
2002
).
43.
B.
Qiao
,
A.
Wang
,
X.
Yang
,
L.
Allard
,
Z.
Jiang
,
Y.
Cui
,
J.
Liu
,
J.
Li
, and
T.
Zhang
, “
Single-atom Catalysis of CO oxidation using Pt1/FeOx
,”
Nat. Chem.
3
,
634
641
(
2011
).
44.
Y.
Wang
,
D. C.
Cantu
,
M.
Lee
,
J.
Li
,
V. A.
Glezakou
, and
R.
Rousseau
, “
CO oxidation on Au/TiO2: Condition-dependent active sites and mechanistic pathway
,”
J. Am. Chem. Soc.
138
,
10467
10476
(
2016
).
45.
F.
Tao
,
S.
Dag
,
L.
Wang
,
Z.
Liu
,
D. R.
Butcher
,
H.
Bluhm
,
M.
Salmeron
, and
G. A.
Somorjai
, “
Break-up of stepped platinum catalyst surfaces by high CO coverage
,”
Science
327
,
850
853
(
2010
).
46.
F.
Tao
,
W. T.
Ralston
,
H.
Liu
, and
G. A.
Somorjai
, “
Surface structures of model metal catalysts in reactant gases
,”
J. Phys. Chem. B
122
,
425
431
(
2018
).
47.
M. A.
van Spronsen
,
J. W. M.
Frenken
, and
I. M. N.
Groot
, “
Observing the oxidation of platinum
,”
Nat. Commun.
8
,
429
(
2017
).
48.
F.
Tao
,
D.
Tang
,
M.
Salmeron
, and
G. A.
Somorjai
, “
A new scanning tunneling microscope reactor used for high-pressure and high-temperature Catalysis studies
,”
Rev. Sci. Instrum.
79
,
084101
(
2008
).
49.
F.
Besenbacher
,
P.
Thostrup
, and
M.
Salmeron
, “
The structure and reactivity of surfaces revealed by scanning tunneling microscopy
,”
MRS Bull.
37
,
677
681
(
2012
).
50.
M.
Salmeron
and
R.
Schlögl
, “
Ambient pressure photoelectron spectroscopy: A new tool for surface science and nanotechnology
,”
Surf. Sci. Rep.
63
,
169
199
(
2008
).
51.
J.
Kim
,
M. C.
Noh
,
W. H.
Doh
, and
J. Y.
Park
, “
In situ observation of competitive CO and O2 adsorption on the Pt(111) surface using near-ambient pressure scanning tunneling microscopy
,”
J. Phys. Chem. C
122
,
6246
6254
(
2018
).
52.
V.
Navarro
,
M. A.
van Spronsen
, and
J. W. M.
Frenken
, “
In situ observation of self-assembled hydrocarbon Fischer−Tropsch products on a cobalt catalyst
,”
Nat. Chem.
8
,
929
934
(
2016
).
53.
T.
Niu
,
Z.
Jiang
,
Y.
Zhu
,
G.
Zhou
,
M. A.
van Spronsen
,
S. A.
Tenney
,
J.
Anibal Boscoboinik
, and
D.
Stacchiola
, “
Oxygen-promoted methane activation on copper
,”
J. Phys. Chem. B
122
,
855
863
(
2018
).
54.
M. C.
Payne
,
M. P.
Teter
,
D. C.
Allan
,
T. A.
Arias
, and
J. D.
Joannopoulos
, “
Iterative minimization techniques for ab initio total-energy calculations- molecular-dynamics and conjugate gradients
,”
Rev. Mod. Phys.
64
,
1045
1097
(
1992
).
55.
J. P.
Perdew
,
K.
Burke
, and
Y.
Wang
, “
Generalized gradient approximation for the exchange-correlation hole of a many-electron system
,”
Phys. Rev. B
54
,
16533
(
1996
).
56.
H. J.
Kulik
and
N.
Marzari
, “
Transition-metal dioxides: A case for the intersite term in Hubbard-model functionals
,”
J. Chem. Phys.
134
,
094103
(
2011
).
57.
T. A.
Halgren
and
W. N.
Lipscomb
, “
Synchronous-transit method for determining reaction pathways and locating molecular transition-states
,”
Chem. Phys. Lett.
49
,
225
232
(
1977
).
58.
Y.
Li
,
K. C.
Adamsen
,
L.
Lammich
,
J. V.
Lauritsen
, and
S.
Wendt
, “
Atomic-scale view of the oxidation and reduction of supported ultrathin FeO islands
,”
ACS Nano
13
,
11632
11641
(
2019
).
59.
L. R.
Merte
,
Y.
Bai
,
H.
Zeuthen
,
G.
Peng
,
L.
Lammich
,
F.
Besenbacher
,
M.
Mavrikakis
, and
S.
Wendt
, “
Identification of O-rich structures on platinum (111)-supported ultrathin iron oxide films
,”
Surf. Sci.
652
,
261
268
(
2016
).
60.
A. S.
Walton
,
J.
Fester
,
M.
Bajdich
,
M. A.
Arman
,
J.
Osiecki
,
J.
Knudsen
,
A.
Vojvodic
, and
J. V.
Lauritsen
, “
Interface controlled oxidation states in layered cobalt oxide nanoislands on gold
,”
ACS Nano
9
,
2445
2453
(
2015
).
61.
J.
Fester
,
Z.
Sun
,
J.
Rodríguez-Fernandez
,
A.
Walton
, and
J. V.
Lauritsen
, “
Phase transitions of cobalt oxide bilayers on Au(111) and Pt(111): The role of edge sites and substrate interactions
,”
J. Phys. Chem. B
122
,
561
571
(
2018
).
62.
Y.
Sun
,
Z.
Qin
,
M.
Lewandowski
,
E.
Carrasco
,
M.
Sterrer
,
S.
Shaikhutdinov
, and
H. J.
Freund
, “
Monolayer iron oxide film on platinum promotes low temperature CO oxidation
,”
J. Catal.
266
,
359
368
(
2009
).
63.
X.
Deng
and
H.
Tüysüz
, “
Cobalt-oxide-based materials as water oxidation catalyst: Recent progress and challenges
,”
ACS Catal.
4
,
3701
3714
(
2014
).
64.
J.
Fester
,
M.
Bajdich
,
A. S.
Walton
,
Z.
Sun
,
P. N.
Plessow
,
A.
Vojvodic
, and
J. V.
Lauritsen
, “
Comparative analysis of cobalt oxide nanoisland stability and edge structures on three related noble metal surfaces: Au (111), Pt (111) and Ag (111)
,”
Top. Catal.
60
,
503
512
(
2017
).
65.
Y.
Liu
,
F.
Yang
,
Y.
Zhang
,
J.
Xiao
,
L.
Yu
,
Q.
Liu
,
Y.
Ning
,
Z.
Zhou
,
H.
Chen
,
W.
Huang
,
P.
Liu
, and
X.
Bao
, “
Enhanced oxidation resistance of active nanostructures via dynamic size effect
,”
Nat. Commun.
8
,
14459
(
2017
).
66.
M.
Muhler
,
R.
Schlogl
, and
G.
Ertl
, “
The nature of the iron oxide-based catalyst for dehydrogenation of ethylbenzene to styrene 2. Surface Chemistry of the active phase
,”
J. Catal.
138
,
413
(
1992
).
67.
R.
Bliem
,
J.
van der Hoeven
,
A.
Zavodny
,
O.
Gamba
,
J.
Pavelec
,
P. E.
de Jongh
,
M.
Schmid
,
U.
Diebold
, and
G. S.
Parkinson
, “
An atomic-scale view of CO and H2 oxidation on a Pt/Fe3O4 model catalyst
,”
Angew. Chem., Int. Ed.
54
,
13999
14002
(
2015
).
68.
T.
Fujii
,
F. M. F.
de Groot
,
G. A.
Sawatzky
,
F. C.
Voogt
,
T.
Hibma
, and
K.
Okada
, “
In situ XPS analysis of various iron oxide films grown by NO2-assisted molecular-beam epitaxy
,”
Phys. Rev. B
59
,
3195
3202
(
1999
).
69.
E.
de Smit
,
M. M.
van Schooneveld
,
F.
Cinquini
,
H.
Bluhm
,
P.
Sautet
,
F. M. F.
de Groot
, and
B. M.
Weckhuysen
, “
On the surface chemistry of iron oxides in reactive gas atmospheres
,”
Angew. Chem., Int. Ed.
50
,
1584
1588
(
2011
).
70.
Z.
Duan
and
G.
Henkelman
, “
Calculations of CO oxidation over a Au/TiO2 catalyst: A study of active sites, catalyst deactivation, and moisture effects
,”
ACS Catal.
8
,
1376
1383
(
2018
).
71.
H.
Zhong
,
L.
Wen
,
J.
Li
,
J.
Xu
,
M.
Hua
, and
Z.
Yang
, “
The adsorption behaviors of CO and H2 on FeO surface: A density functional theory study
,”
Powder Technol.
303
,
100
108
(
2016
).
72.
J.
Yoo
,
X.
Rong
,
Y.
Liu
, and
A. M.
Kolpak
, “
Role of lattice oxygen participation in understanding trends in the oxygen evolution reaction on perovskites
,”
ACS Catal.
8
,
4628
4636
(
2018
).
73.
A.
Grimaud
,
O.
Diaz-Morales
,
B.
Han
,
W. T.
Hong
,
Y. L.
Lee
,
L.
Giordano
,
K. A.
Stoerzinger
,
M. T. M.
Koper
, and
S. H.
Yang
, “
Activating lattice oxygen redox reactions in metal oxides to catalyse oxygen evolution
,”
Nat. Chem.
9
,
457
465
(
2017
).
74.
A. D.
Allian
,
K.
Takanabe
,
K. L.
Fujdala
,
X.
Hao
,
T. J.
Truex
,
J.
Cai
,
C.
Buda
,
M.
Neurock
, and
E.
Iglesia
, “
Chemisorption of CO and mechanism of CO oxidation on supported platinum nanoclusters
,”
J. Am. Chem. Soc.
133
,
4498
4517
(
2011
).
75.
H. Y.
Kim
and
G.
Henkelman
, “
CO oxidation at the interface of Au nanoclusters and the stepped-CeO2(111) surface by the Mars–van Krevelen mechanism
,”
J. Phys. Chem. Lett.
4
,
216
221
(
2013
).

Supplementary Material