Theoretical Study of Hydrogen Permeation through Mixed NiO–MgO

Article pubs.acs.org/JPCA

Theoretical Study of Hydrogen Permeation through Mixed NiO− MgO Films Supported on Mo(100): Role of the Oxide−Metal Interface Daniel Torres,†,§ Francesc Illas,*,‡ and Ping Liu† †

Center for Functional Nanomaterials, Brookhaven National Laboratory, Upton, New York 11973, United States Departament de Química Física & Institut de Química Teòrica i Computacional (IQTCUB), Universitat de Barcelona, 08028 Barcelona, Spain

‡

ABSTRACT: This work presents a periodic density functional study of the adsorption and permeation of atomic H on Ni-doped MgO oxide thin films supported on a Mo(100) surface. We find that the binding of atomic H is affected by the presence of a metallic support. The chemisorption energies increase considerably when the oxide film is supported. The H permeation through the NiO−MgO oxide was also studied. H migration through the unsupported NiO−MgO oxide is thermodynamically inhibited, while the presence of the metallic Mo makes permeation thermodynamically favored. We attribute the promoting effect to the different character of adsorbed H at the unsupported Ni-doped MgO oxide and at the oxide−Mo interface. In the former case, H forms hydroxyl groups, whereas in the latter case it appears as hydride due to the formation of strong metal−hydrogen bonds. These results illustrate the important role that the oxide−metal interface could play in the mechanism for pure and mixed oxides reduction.

1. INTRODUCTION Metal oxides are one of the most important and widely employed categories of catalysts, and mixed metal−oxides represent a promising type of oxides with potential use as catalytic active phases or support materials. A variety of mixed oxides containing alkali, alkaline, rare earth, or noble metals can be prepared by mixing MgO with other oxides and the final material displays unique catalytic properties, different from each parent material.1−6 To understand the promoting effect of mixed oxides, advanced surface science techniques (e.g., STM and XPS) have been extensively employed, where the oxide have been modeled by means of epitaxial thin films grown on metal surfaces.7−13 For a thick enough film, the electronic structure of the oxide will rest unperturbed due to the metal support. Differently, for thin oxide films, the oxide−metal interface plays a more relevant role, and several experimental and theoretical studies indicated that the presence of a metal substrate affects the surface properties of the oxide in a profound way.14−16 NiO−MgO solid solutions have been widely used as catalysts for methane activations, in particular, dry reforming of methane (CO2 + CH4 → 2CO + 2H2), which combines two of the most problematic greenhouse gases to generate syngas for the synthesis of clean liquid fuels and valuable chemicals.17−20 Both NiO and MgO crystallize in the simple rock-salt structure with almost identical lattice constants, and highly ordered NiO− MgO mixed oxide films can be formed on an appropriate substrate such as Mo(100). The resulting oxide films have the (100) face parallel to the Mo(100) substrate.21,22 Stoichio© XXXX American Chemical Society

metric NiO−MgO phase exhibits low catalytic activity, while a highly active phase forms after a partial reduction under hydrogen. The reduction-produced oxygen vacancies and reduced Ni ions are responsible for the promoted activity. Some of us have recently demonstrated that during oxide intermixing of NiO and MgO, Ni ions antisegregate from the NiO−MgO(100) surface, being preferentially located in the second outermost oxide layer.23 Pacchioni et al.24 have recently shown that an underlying metallic support significantly changes the Ni distribution, favoring the diffusion of Ni toward the metal/oxide interface. However, less attention has been paid for hydrogen permeation, which will impact the location of oxygen vacancies and reduced Ni ions and therefore the overall activity of the material. The hydrogen species used in the preparation of the reduced NiO−MgO samples are mobile and hydrogen permeation through the body of the oxide occurs. In contrast, the diffusion into bulk MgO is, however, very slow, and H atoms are stabilized only in the near-surface region, that is, in the 3−5 outermost oxide layers.25,26 In the case of metal supported NiO−MgO thin film, hydrogen atoms arising from molecular hydrogen dissociation can also penetrate and eventually reach the oxide−metal interface.16 However, the microscopic mechanism of hydrogen permeation through Special Issue: Energetics and Dynamics of Molecules, Solids, and Surfaces - QUITEL 2012 Received: September 4, 2013 Revised: January 21, 2014

A

dx.doi.org/10.1021/jp408872x | J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A

Article

NiO−MgO and the roles of metal support during this process remain incomplete. In the present work, we report periodic density functional theory (DFT) based calculations addressing the interaction of atomic hydrogen with stoichiometric NiO−MgO mixed oxide models with low Ni content (4%). Therefore, the resulting mixed oxide systems can effectively be regarded as Ni-doped MgO. Unsupported and Mo(100) supported films, NiO−MgO and NiO−MgO/Mo(100), respectively, have been considered. Our results indicate that Mo support indirectly affects H binding, which is associated with changes in the Ni distribution on the MgO matrix. Our model calculations also indicate that atomic H permeation through the unsupported NiO−MgO oxide will hardly take place. On the contrary, in the presence of a Mo support, H atoms permeate more easily through the mixed oxide reaching the internal oxide−Mo interface. The origin of the different behavior of H in the unsupported and supported NiO−MgO systems is attributed to the different character of adsorbed hydrogen. In the former case H forms hydroxyl groups, whereas in the latter case it appears as hydride, forming a strong metal−hydrogen bond. Our results highlight the important role that the oxide−metal interface could play in the reduction mechanism of this type of mixed oxide.

Figure 1. Schematic representation of a the NiO−MgO(100) (left panel) and NiO−MgO/Mo(100) (right panel) periodic models used to represent the unsupported and Mo(100) supported mixed oxides. The red spheres represent oxygen atoms, whereas the light blue, dark blue, and green spheres represent the Mg, Mo, and Ni cations, respectively.

2. COMPUTATIONAL METHODS AND MATERIAL MODELS The interaction of atomic hydrogen with unsupported and Mo(100)-supported NiO−MgO systems has been studied using DFT based periodic calculations carried out as implemented in the Vienna ab Initio Simulation Package (VASP).27,28 The calculations have been carried out with PW91 form29 of the generalized gradient approximation (GGA), which provides a good compromise for accuracy versus computational cost and has shown to lead a robust description of all transition metal elements series.30 A caveat regarding the use of GGA is necessary here since it is well-known that this approach fails to properly reproduce the antiferromagnetic, charge-transfer insulator, character of the NiO ground state, a correct description being obtained only when hybrid functionals or GW techniques are used.31 In fact, even the widely used DFT+U has problems. The introduction of the U term for the Ni(3d) opens a gap, but since only the Ni(3d) level are modified, the resulting ground state is of Mott−Hubbard type instead of charge transfer. However, one must realize that NiO−MgO mixtures at the concentration range considered in this work (see below) do not exhibit any magnetic order, and it is enough to ensure that the Ni2+ cations are in a high spin state. The band gap does not represent a problem since it is dominated by MgO and hence large enough. Accordingly, Nidoped MgO should be qualitatively well described with the spin polarized version of the PW91 functional. The electron density was expanded in a plane wave basis set and the effect of the core electrons taken into account through the projector-augmented wave method of Bloch32 as implemented in VASP.33 An energy cutoff of 400 eV ensured planewave convergence, and the Brillouin zone integration was performed on a 7 × 7 × 1 Monkhorst−Pack grid.34 All calculations were carried out using the spin polarized formalism. The convergence criterion for geometry optimization was chosen so that atomic forces were less than 10 meV/Å. A six-layer oxide slab with a (2 × 2) unit cell supported on a three-layer Mo(100) substrate, as shown in Figure 1, was used to represent the MgO/Mo(100) matrix, and one of the Mg

cations was replaced by a Ni cation to model a low Ni content NiO−MgO solid solution. This provides an adequate representation of the Ni0.04Mg0.96O mixture with a low NiO concentration corresponding to one NiO unit per oxide unit cell. In the case of the unsupported film the Ni cation occupies a site in the second layer, whereas in the supported film to study the effect of the interface, the Ni cation is close to the Mo first layer. This is in agreement with previous results concerning the distribution of Ni in the unsupported and supported films and will be described in more detail in the next section.23 The total energy was minimized with the full relaxation of the atomic positions of the mixed oxide and topmost Mo layer in contact with the oxide, whereas the two bottom Mo layers were fixed to their bulk PW91-optimal bulk positions (aMo = 3.165 Å). Atomic H bonding energy, Eb, was computed as E b = Eslab ‐ H − Eslab − E H

(1)

where Eslab‑H and Eslab is the total energy of the oxide substrate with Ni located on its most stable position with and without H, respectively. EH is the total energy of atomic H in gas phase. With the present definition, negative Eb values indicate a stable adsorption. For the bare surfaces, we computed atomic charges following the Atoms in Molecules scheme of charge density decomposition proposed by Bader.35 We also calculated the atomic spin densities that provide an estimate of the magnetic moments. This has been obtained by the projection of spin density into atomic spheres as defined in VASP.

3. RESULTS AND DISCUSSION In the following we describe the interaction and permeation of atomic H through NiO−MgO(100) with and without Mo(100) support. Our results indicate that in the absence of a Mo support Ni atoms prefer to stay in the second topmost layer rather than the top surface layer (Figure 1) and retain the magnetization, being the average magnetic moment and Bader charge of 1.5 μb and +1.19 |e| per Ni atom, respectively.23 This is in consistent with a Ni2+ formal oxidation state with two unpaired electrons mainly located at the Ni site. In the presence B

dx.doi.org/10.1021/jp408872x | J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A

Article

Figure 2. Schematic representation of the NiO−MgO(100) surface and NiO−MgO/Mo(100) surface and interface, showing the different adsorption and interface sites. In the top view of the NiO−MgO/Mo(100) interface, the Mo atoms are not displayed for clarity. The red spheres represent oxygen atoms, whereas the light blue, dark blue, and green spheres represent the Mg, Mo, and Ni cations, respectively.

Table 1. Properties of H Atoms Adsorbed on the Different Sites of Unsupported and Supported NiO−MgO/Mo(100) Surfacea NiO−MgO(100)

NiO−MgO/Mo(100)

site

ONi

OMg

BridgeMg

Mg

HollowMg

OMg

BridgeMg

Mg

HollowMg

Eb (eV) z (Å) dO−H (Å)

−2.71 1.69 0.98

−1.91 1.25 0.99

−1.35 1.75

−1.18 2.06

−1.11 1.52

−2.08 1.26 0.99

−1.55 1.23

−1.24 2.10

−1.07 1.43

a Eb stands for bonding energy defined with respect to the bare substrate and gas phase atomic hydrogen, z for adsorption height defined with respect to the oxide surface plane, and dO−H for the distance of H to the nearest O site.

of the Mo support, Ni ions locate at the oxide−metal interface (Figure 1) and turn into metallic form, with magnetic moment of 0 μb and Bader charge of −0.2 |e|.16 These results are in agreement with available previous studies.24,36 3.1. Atomic Hydrogen on the Surface of NiO− MgO(100) and NiO−MgO/Mo(100). Let us now consider the adsorption of atomic H on the unsupported and Mo(100)supported NiO−MgO(100) substrate. Figure 2 left and middle panels picture the different adsorption sites for atomic H on the surfaces. On the NiO−MgO(100) slab, H atoms adsorb preferentially above O sites, which are directly connected to Ni atoms in the layer below (ONi sites) with Eb = −2.71 eV and an adsorption height z = 1.7 Å. The O−H distance is of ∼1 Å, which is typical for a hydroxyl group. Below we will provide evidence that the resulting species is indeed a surface hydroxyl group with the concomitant reduction of Ni2+ to Ni+. In contrast, the adsorption at Mg sites is energetically less favored by more than 1.5 eV. Considerably different adsorption properties are found on the surface of NiO−MgO/Mo(100) though the preference of H for surface O anions is maintained (Eb = −2.08 eV, Table 1), as one would expect from simple chemical arguments. Because of the absence of metallic Ni sites in the subsurface oxide layer, the most active ONi sites for H adsorption disappears; however, for H at the OMg, BridgeMg, and Mg sites, the presence of the metallic support enhances Eb by 0.06−0.20 eV as compared to those at the same sites of NiO−MgO(100). In order to further confirm the conclusions outlined above regarding the difference in H bonding energy and to obtain additional information about the bonding mechanism of H with the NiO−MgO(100) and NiO−MgO/Mo(100) surfaces, Figure 3 displays appropriate projected density of state (PDOS) plots. The PDOS plots show that, in the absence of the metallic support, H 1s states are hybridized with the d states of Ni (Figure 3, top panel), whereas in the surface of the supported oxide H 1s states are hybridized with the s states of

Figure 3. Density of states (DOS) plots of H adsorbed on an ONi and OMg sites of the NiO−MgO(100) and NiO−MgO(100)/Mo(100) surface, respectively (top and middle panels, respectively), and at the Mg site of the NiO−MgO/Mo(100) interface (bottom panel). DOS projected (PDOS) on the H atom (black line), on the Ni (green line), Mg and O atoms (red line), and the Mo substrate (blue line).

Mg only (Figure 3 middle panel). Hence, the ONi sites in the case of NiO−MgO(100) displays the strongest H binding. On NiO−MgO/Mo(100), the new states of MgO are generated, which enhances the hybridization with H. Therefore, the enhanced H bindings are observed compared to those at the same sites of NiO−MgO(100) (Table 1). Since NiO−MgO is a reducible oxide,37 it is likely that hydroxyl formation is accompanied by the reduction of Ni2+ to Ni+. This is confirmed by the calculated Bader charge for the unsupported NiO−MgO film, which indicates one electron transfer from H to the d orbitals of Ni and leads to the formation of a surface H+ with a Bader charge of +1 |e|. In this case, the Ni cation exhibits a C

dx.doi.org/10.1021/jp408872x | J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A

Article

Table 2. Properties of H Atoms Adsorbed on the Different Sites of the NiO−MgO/Mo(100) Interfacea NiO−MgO/Mo(100) interface Site

OMg

Ni

Mg

HollowMg

HollowNi

BridgeMg

BridgeNi

Eb (eV) z (Å) dinterface (Å) dO−H (Å)

−1.17 0.39 2.15 0.99

−2.31 1.73 2.08

−3.23 1.70 2.03

−3.05 1.31 2.03

−3.09 1.43 1.72

−1.80 0.93 2.04

−2.08 0.33 2.07

a Eb is as in Table 2, z defines the perpendicular distance of H to the average oxide surface plane at the NiO−MgO/Mo interface, dinterface is the average oxide−Mo interface distance, and dO−H is the distance of H to the nearest O site.

reduced Bader charge of +0.64 |e| and a reduced magnetic moment of 1 μb, which is consistent with a Ni1+ species. In the presence of the metallic support, adsorbed H also exhibits a Bader charge of +1 |e| but here the electron has been transferred to the metal conduction band. 3.2. H at the Metal−Oxide Interface of NiO−MgO/ Mo(100). The results for the interaction of atomic H at different sites of the NiO−MgO/Mo(100) interface are compiled in Table 2 where the corresponding adsorption sites are shown in the right panel of Figure 2. Note that the migration of Ni cations to the oxide−Mo interface makes it possible to differentiate two different metallic centers at the oxide layer near the interface (a Ni and a Mg site). Our results indicate that the direct interaction at Mg sites is the most favored (Eb = −3.23 eV) and, in contrast the bindings at the hollow sites (HollowMg,Ni), are only slightly weaker (Eb = −3.05 eV, −3.09 eV). We found very distinct interaction properties for H at Mg sites of the interface and on the surface of NiO− MgO/Mo, though the Mg−H distance at the interface (2.03 Å) is only slightly shorter than at the surface (2.06 Å). The interaction is almost ∼2 eV more favored at the interface than on the surface. At the interface, H atoms at the Mg sites are shared between Mg sites and the Mo substrate. The distance between the H atom and the average plane of the Mo substrate is ∼0.3 Å. The direct interaction with Mo greatly promotes the H binding since the bonding energy of H on the clean Mo(100) surface is quite large (−3.3 eV). In consideration of the most stable adsorption, H binding at the interface is stronger than that on the surface by more than 1 eV, which can act as a driving force for H atoms diffusion through the MgO layers until reaching the interface. Moreover, the H atom at the interface has a negative charge, the calculated Bader charge being of−0.70 |e| is consistent with a charge transfer from the metallic substrate leading to a hydride (H−) species, in accordance with the case of clean Mo(100). Our results also show that hydroxyl-type groups (OMg, BridgeNi, and BridgeMg) are less stable at the metal−oxide interface than the hydride-type groups (Mg, HollowMg, and HollowNi; Table 2). Hence, hydroxyl groups formed during reduction will remain at the surface, and hydride groups dominate the metal−oxide interface. 3.3. H Migration and Permeation on and through NiO−MgO and NiO−MgO/Mo(100). A complete description of the interaction of H with these rather complex mixed-oxide systems requires considering H diffusion through the surface and through the material. The energy profiles describing the diffusion of H on the NiO−MgO(100) and NiO−MgO/ Mo(100) surfaces and the migration through the bulk are given in Figure 4 top and bottom panels, respectively. In each case, the left part of the figure corresponds to surface diffusion from site to site, and the right part, to migration perpendicular to the surface.

Figure 4. Energy profile for the migration through the surface (left side of the panel) and permeation toward the bulk (right side of the panel) of atomic H for unsupported NiO−MgO(100) (top panel) and supported NiO−MgO/Mo(100). Eb was computed with respect to the clean supported oxide with Ni on its most stable arrangement at the interface and gas phase E(H). We considered on surface sites (Oh label) for the surface migration, and interlayer subsurface octahedral (Oh Label) and in-layer sites (In label) for the penetration into the oxide bulk.

The diffusion of H atoms through the surface of unsupported NiO−MgO(100) slab proceeds through direct hopping between two stable oxygen sites (ONi) via a hollow site (HollowMg). Diffusion energies were estimated, in this case, by means of the changes on the binding energy, ΔEdiff ≈ 1.6 eV, which is lower than that corresponding for pure MgO.38 Figure 4 also displays the binding energy of H atoms in subsurface octahedral sites located at successively deeper layers (labeled as Oh in Figure 4). In order to estimate the migration energy barrier between two different octahedral subsurface sites located in two different oxide interlayers, we fixed the coordinates of H in the oxide interlayer (labeled as in-layer D

dx.doi.org/10.1021/jp408872x | J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A

Article

Present Address

sites in Figure 4) and relaxed the whole oxide structure. Our results clearly show that the subsurface adsorption is in general less favored than the on-surface adsorption. The migration of H atoms from the surface toward deeper oxide layers is thermodynamically hindered by ΔEperm ≈ 3.3 eV, with ΔEperm defined as Eb(bulk layers) − Eb(surface). Therefore, under working conditions with temperatures between 400 and 600 K, hydrogen permeation through a perfect NiO− MgO(100) is not likely. This is very similar to the situation for pure MgO, and the results from Oviedo and co-workers suggested that H atoms are stabilized in the near-surface region of the oxide, between the 3−5 layers.25,26 The diffusion of H atoms on the surface of NiO−MgO/ Mo(100) proceeds with a lower energy cost of ∼1 eV, suggesting that the diffusion may occur at lower temperatures (Figure 4). The presence of a metallic support leads to the lacking of strongly adsorbed ONi sites of NiO−MgO(100), and the H binding at the OMg site of NiO−MgO/Mo(100) is 0.63 eV less stable than that at ONi (Table 1). Since the intermediate HollowMg sites for H diffusion between OMg sites are almost unaffected by the presence of Mo(100) (Table 1), this makes ΔEdiff decrease. For H migration from the surface to the inner oxide layers, the presence of the metal−oxide interface is able to lower the effective energy cost by more than 1 eV, favoring the migration of H atoms from the surface toward the interface, ΔEperm being now −1.2 eV (Figure 4). The origin of this different behavior is attributed to the direct interaction of H at the Mg site to Mg and Mo at the oxide−Mo interface forming strong metal−hydrogen bonds. In addition, the metallic Ni atoms at the interface also help indirectly in enhancing the H binding. The bottom panel of Figure 3 displays the PDOS plots for H adsorbed at the most stable site of the metal−oxide interface, indicating that the H 1s level, Mg states, and the Mo, Ni d states, are strongly mixed with H staying as a hydride H−.

§

Physics Department, The New York City College of Technology, The City University of New York, Brooklyn, New York 11201, United States.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was carried out at Brookhaven National Laboratory (BNL) under Contract No. DE-AC02-98CH10886 with the US Department of Energy, Office of Science. The calculations utilized resources at the BNL Center for Functional Nanomaterials (CFN) and the Centre de Supercomputacio de Catalunya (CESCA). Additional support from Spanish MINECO FIS2008-02238 and CTQ2012-30751 research grants and Generalitat de Catalunya grants 2009SGR1041 and XRQTC is accredited. F.I. acknowledges support from the 2009 ICREA Academia award for excellence in research. This work is also a part of COST Actions CM1104.

■

(1) Park, J. B.; Graciani, J.; Evans, J.; Stacchiola, D.; Senanayake, S. D.; Barrio, L.; Liu, P.; Sanz, J. F.; Hrbek, J.; Rodriguez, J. A. Gold, Copper, and Platinum Nanoparticles Dispersed on CeOx/TiO2(110) Surfaces: High Water-Gas Shift Activity and the Nature of the MixedMetal Oxide at the Nanometer Level. J. Am. Chem. Soc. 2010, 132, 356−363. (2) Graciani, J.; Plata, J. J.; Sanz, J. F.; Liu, P.; Rodriguez, J. A. A Theoretical Insight into the Catalytic Effect of a Mixed-Metal Oxide at the Nanometer Level: The Case of the Highly Active Metal/CeOx/ TiO2(110) Catalysts. J. Chem. Phys. 2010, 132, 104703. (3) Yang, F.; Kundu, S.; Vidal, A. B.; Graciani, J.; Ramírez, P. J.; Senanayake, S. D.; Stacchiola, D.; Evans, J.; Liu, P.; Sanz, J. F.; Rodriguez, J. A. Determining the Behavior of RuOx Nanoparticles in Mixed-Metal Oxides: Structural and Catalytic Properties of RuO2/ TiO2(110) Surfaces. Angew. Chem., Int. Ed. 2011, 50, 10198−10102. (4) Yang, F.; Choi, Y.; Agnoli, S.; Liu, P.; Stacchiola, D.; Hrbek, J.; Rodriguez, J. A. CeO2 ↔ CuOx Interactions and the Controlled Assembly of CeO2(111) and CeO2(100) Nanoparticles on an Oxidized Cu(111) Substrate. J. Phys. Chem. C 2011, 115, 23062− 23066. (5) Kundu, S.; Vidal, A. B.; Yang, F.; Ramírez, P. J.; Senanayake, S. D.; Stacchiola, D.; Evans, J.; Liu, P.; Rodriguez, J. A. Special Chemical Properties of RuOx Nanowires in RuOx/TiO2(110): Dissociation of Water and Hydrogen Production. J. Phys. Chem. C 2012, 116, 4767− 4773. (6) Bruix, A.; Rodriguez, J. A.; Ramirez, P. J.; Senanayake, S. D.; Evans, J.; Park, J. B.; Stacchiola, D.; Liu, P.; Hrbek, J.; Illas, F. A New Type of Strong Metal-Support Interaction and the Production of H2 through the Transformation of Water on Pt/CeO2(111) and Pt/ CeOx/TiO2(110) catalysts. J. Am. Chem. Soc. 2012, 134, 8968. (7) Bäumer, M.; Freund, H.-J. Metal Deposits on Well-Ordered Oxide Films. Prog. Surf. Sci. 1999, 61, 127−198. (8) Giordano, L.; Pacchioni, G. Oxide Films at the Nanoscale: New Structures, New Functions, and New Materials. Acc. Chem. Res. 2011, 44, 1244−1252. (9) Schauermann, S.; Nilius, N.; Shaikhutdinov, S.; Freund, H.-J. Nanoparticles for Heterogeneous Catalysis: New Mechanistic Insights. Acc. Chem. Res. 2013, 46, 1673−1681. (10) Gonchar, A.; Risse, T.; Freund, H.-J.; Giordano, L.; Di Valentin, C.; Pacchioni, G. Activation of Oxygen on MgO: O2−• Radical Ion Formation on Thin, Metal-Supported MgO(001) Films. Angew. Chem., Int. Ed. 2011, 50, 2635−2638. (11) Lei, Y.; Lewandowski, M.; Sun, Y. N.; Fujimori, Y.; Martynova, Y.; Groot, I. M. N.; Meyer, R.; Giordano, L.; Pacchioni, G.; Goniakowski, J.; Noguera, C.; Shaikhutdinov, S.; Freund, H. -J. CO

4. CONCLUSIONS In the present work, we employed DFT and provided evidence that the interaction of atomic H with Ni-doped MgO(100) is affected by the presence of a Mo(100) support, where the chemisorptions at O sites (OMg) and Mg sites (BridgeMg and Mg) are more exothermic when the metallic support is present. Our calculations further indicated that the migration of atomic H into NiO−MgO(100) is strongly inhibited both thermodynamically and kinetically with ΔEperm of 3.3 eV. In contrast, the presence of a Mo support facilitates the migration of H atoms from the surface to the internal oxide−Mo interface with ΔEperm of −1.2 eV. The preferential migration though NiO− MgO/Mo(100) is associated with the strong interaction of H atoms with both Mg and Mo sites. Our study suggests that the concentration of the hydride groups should be much larger at the metal−oxide interface than that at the oxide surface, while the hydroxyl groups dominate the surface under the reduction conditions. Hence, the oxide−metal interface is expected to play a key role during oxide reduction. Overall, the thin and thick supported films can exhibit a different chemical behavior, which adds further evidence for the unique properties of supported oxide thin films.

■

REFERENCES

AUTHOR INFORMATION

Corresponding Author

*(F.I.) E-mail: [email protected]. Phone: +34934021229. Fax: +34934021231. E

dx.doi.org/10.1021/jp408872x | J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A

Article

(34) Monkhorst, H.; Pack, J. Special Points for Brillouin-Zone Integrations. Phys. Rev. B 1976, 13, 5188−5192. (35) Bader, R. F. W. A Quantum-Theory of Molecular-Structure and its Applications. Chem. Rev. 1991, 91, 893−928. (36) Stavale, F.; Nilius, N.; Freund, H.-J. Cathodoluminescence of Near-Surface Centres in Cr-Doped Mgo(001) Thin Films Probed By Scanning Tunnelling Microscopy. New J. Phys. 2012, 14, 033006. (37) Chen, H-Y. T.; Giordano, L.; Pacchioni, G. From Heterolytic to Homolytic H2 Dissociation on Nanostructured MgO(001) Films As a Function of the Metal Support. J. Phys. Chem. C 2013, 117, 10623− 10629. (38) Wu, G.; Zhang, J.; Wu, Y.; Li, Q.; Chou, K.; Bao, X. Adsorption and Dissociation of Hydrogen on Mgo Surface: A First-Principles Study. J. Alloys Compd. 2009, 480, 788−793.

plus NO versus CO + O2 Reaction on Monolayer FeO(111) Films on Pt(111). ChemCatChem 2011, 3, 671−674. (12) Shao, X.; Nilius, N.; Myrach, P.; Freund, H.-J.; Martinez, U.; Prada, S.; Giordano, L.; Pacchioni, G. Strain-Induced Formation of Ultrathin Mixed-Oxide Films. Phys. Rev. B 2011, 83, 245407. (13) Pacchioni, G.; Freund, H.-J. Electron Transfer at Oxide Surfaces. The MgO Paradigm: from Defects to Ultrathin Films. Chem. Rev. 2013, 113, 4035−4072. (14) Pacchioni, G. Two-Dimensional Oxides: Multifunctional Materials for Advanced Technologies. Chem.Eur. J. 2012, 18, 10144−10158. (15) Pacchioni, G. Electronic Interactions and Charge Transfers of Metal Atoms and Clusters on Oxide Surfaces. Phys. Chem. Chem. Phys. 2013, 15, 1737−1757. (16) Müller, K.; Torres, D.; Park, J. B.; Liu, P.; Stacchiola, D.; Starr, D. E. NiO−MgO and CoO−MgO Thin-Film Solid Oxide Solutions on a Mo(100) Support: Formation, Reduction, and Influence of the Support. J. Phys. Chem. C 2013, 117, 280−287. (17) Ruckenstein, E.; Hu, Y. H. Carbon Dioxide Reforming of Methane over Nickel Alkaline Earth Metal Oxide catalysts. Appl. Catal., A 1995, 133, 149−161. (18) Kumar, P.; Sun, Y.; Idem, R. O. Nickel-Based Ceria, Zirconia, and Ceria-Zirconia Catalytic Systems for Low-Temperature Carbon Dioxide Reforming of Methane. Energy Fuels 2007, 21, 3113−3123. (19) Cui, Y.; Zhang, H.; Xu, H.; Li, W. The CO2 Feforming of CH4 Over Ni/La2O3/α-Al2O3 Catalysts: The Effect of La2O3 Contents on the Kinetic Performance. Appl. Catal., A 2007, 331, 60−69. (20) Bellido, J. D. A.; De Souza, J. E.; M’Peko, J.-C.; Assaf, E. M. Effect of Adding CaO to ZrO2 Support on Nickel Catalyst Activity in Dry Reforming of Methane. Appl. Catal., A 2009, 358, 215−223. (21) Colbourn, E. A. Computer-Simulation of Defects And Reactions At Oxide Surfaces. Surf. Sci. Rep. 1992, 15, 281−319. (22) Guo, Q.; Xu, C.; Goodman, D. W. Ultrathin Films of Nio On Mgo(100): Studies of the Oxide−Oxide Interface. Langmuir 1998, 14, 1371−1374. (23) Torres, D.; Liu, P. Vacancy-Driven Surface Segregation in NixMg1−xO(100) Solid Solutions from First Principles Calculations. Catal. Lett. 2012, 142, 1211−1217. (24) Prada, S.; Giordano, L.; Pacchioni, G. Li, Al, and Ni Substitutional Doping in Mgo Ultrathin Films on Metals: Work Function Tuning via Charge Compensation. J. Phys. Chem. C 2012, 116, 5781−5786. (25) Oviedo, J.; Calzado, C. J.; Sanz, J. F. Molecular Dynamics Simulations of the MgO(001) Surface Hydroxylation. J. Chem. Phys. 1998, 108, 4219−4225. (26) Gonzalez, R.; Chen, Y.; Tsang, K. L. Diffusion of Deuterium and Hydrogen in Doped and Undoped Mgo Crystals. Phys. Rev. B 1982, 26, 4637−4645. (27) Kresse, G.; Hafner, J. Ab Initio Molecular-Dynamics For LiquidMetals. Phys. Rev. B 1993, 47, 558−561. (28) Kresse, G.; Furthmuller, J. Efficient Iterative Schemes for ab Initio Total-Energy Calculations Using a Plane-Wave Basis Set. Phys. Rev. B 1996, 54, 11169−11186. (29) Perdew, J. P.; Chevary, J. A.; Vosko, S. H.; Jackson, K. A.; Pederson, M. R.; Singh, D.; Fiolhais, C. Atoms, Molecules, Solids, and Surfaces: Applications of the Generalized Gradient Approximation for Exchange and Correlation. Phys. Rev. B 1992, 46, 6671−6687. (30) Janthon, P.; Kozlov, S. M.; Viñes, F.; Limtrakul, J.; Illas, F. Establishing the Accuracy of Broadly Used Density Functionals in Describing Bulk Properties of Transition Metals. J. Chem. Theory Comput. 2013, 9, 1631−1640. (31) de Moreira, I. P.R.; Illas, F.; Martin, R. L. Effect of Fock Exchange on the Electronic Structure and Magnetic Coupling in NiO. Phys. Rev. B 2002, 65, 155102. (32) Blöchl, P. E. Projector Augmented-Wave Method. Phys. Rev. B 1994, 50, 17953−17979. (33) Kresse, G.; Joubert, J. From Ultrasoft Pseudopotentials to the Projector Augmented-Wave Method. Phys. Rev. B 1999, 59, 1758− 1775. F

dx.doi.org/10.1021/jp408872x | J. Phys. Chem. A XXXX, XXX, XXX−XXX