Revealing the intrinsic electronic-level principle driving metal-support interaction trends of single-atom Ru with oxides through cooperative orbital coupling

doi:10.21203/rs.3.rs-2442843/v1

Download PDF

Article

Revealing the intrinsic electronic-level principle driving metal-support interaction trends of single-atom Ru with oxides through cooperative orbital coupling

https://doi.org/10.21203/rs.3.rs-2442843/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Supports dominate chemical states and catalytic performances of single-atom catalysts (SACs); however, electronic-scale physicochemical principles intrinsically driving SAC-support interactions and their structure-function relationships are still not coherently clarified nor unified yet. Using atomically-supported Ru on rutile-type oxides MO₂ (TiO₂, SnO₂, and MnO₂) for CO adsorption and oxidation as model systems, here we explore the interaction trends and electronic-level origins of SAC-support interactions. Our results show that the reactivity of surface oxygen atoms overall controls the strengths of Ru₁-MO₂ binding and CO-Ru₁ adsorption, and catalytic activity of CO oxidation. Through combining experimental probing and theoretical calculations, we reveal that these interaction trends can be coherently unified by an electronic-level interaction mode through competitive orbital coupling in CO-Ru₁-O₅-MO₂ systems, in which neighboring bonds are competitive while non-neighboring bonds are cooperative in coupling strength. The interaction mode of orbitals provides new insights for understanding SAC-support interactions and rational design of SACs.

Physical sciences/Chemistry/Catalysis/Heterogeneous catalysis

Physical sciences/Chemistry/Physical chemistry

Physical sciences/Chemistry/Coordination chemistry

Single-atom catalysis

metal-support interaction

support effect

orbital coupling

Ru catalyst

Atomically supported metallic atoms (M₁) have been rising as a new frontier of heterogeneous catalysis featuring in 100% atomic exposures^1–3. Such single-atom catalysts (SACs) can display unique catalytic activity and selectivity in oxidation^4,5, hydrogenation⁶, electrocatalytic⁷, photocatalytic⁸, and biomimetic reactions⁹. These catalytic performances are highly affected or even dominated by supports, and can be commonly tuned through varying coordination atoms and numbers^10,11. In general, SAC-support interactions (SASI) in adsorbate-M₁-(ligand)_n-support systems include three aspects, M₁-support binding, adsorbate-M₁ adsorption, and catalytic performance. Understanding underlying interaction trends and principles is the central scientific issue of SACs, which is critical for better design, synthesis, and optimization of SACs^1,2,12. For example, SAC’s stability on oxide supports is found to be dominated by their bonding strength with surface lattice oxygen atoms (O_sl), and more fundamentally correlates to the oxophilicity of metal atoms and reducibility of supports¹³. SACs’ catalytic activities are usually discussed with charging state or electronic strong metal-support interaction (EMSI), but the trends are often controversy depending on metals and reactions^14,15. Therefore, unifying the physicochemical models and principles driving SASIs at the electronic level is still fundamentally in demand^16–18.

Valence state and d-band model now are two main conceptual tools to discuss SAC’s catalytic mechanisms. Experimentally, valence state has been widely used to explain the activities of adsorbate-M₁ adsorptions and catalytic reactions¹⁹, and can be tuned through H₂ reduction²⁰, EMSI^7,21, and introducing defects^10,22,23. However, valence cannot deeply reveal detailed electronic-level mechanisms like orbital coupling, electron fillings, and catalytic mechanisms^24,25, and thus provides limited information. Theoretically, SAC’s electronic structures and activities are often calculated based on density functional theory (DFT) and interpreted with d-band model^10,26. However, this method is disputable for SACs, because d-band model was developed to calculate chemisorption strength on extended metal surfaces, while SACs are typically stabilized through coordination with nonmetal atoms like O, N, and C²⁷. Now, it is still a vital fundamental issue to explore experimental methods and theoretical models to reveal the electronic-level principles driving SASIs^11,28.

Through combining experimental probing, theoretical calculation, and model analysis, herein we demonstrate an electronic-level SASI principle based on competitive orbital couplings. We design atomically dispersed Ru (Ru₁) on rutile-type MO₂ oxide supports (TiO₂, SnO₂, and β-MnO₂) as SAC models, and use the adsorption and oxidation of CO to probe support-dependent interaction trends of adsorbate-Ru₁ adsorption and catalytic activity. For the CO-Ru₁-(O_sl)₅-MO₂ models, the same lattice structure of RuO₂ with TiO₂, SnO₂, and β-MnO₂ allows Ru atoms to occupy surface metal sites of the supports, which thus can eliminate the effects from different lattice types²⁹. We show that O_sl reactivity (or support reducibility) can overall determine Ru₁-MO₂ interactions, Ru₁ electronic states, and CO-Ru₁/MO₂ adsorption, and can also control the catalytic activity of CO oxidation via participating in catalytic cycles. The splitting and filling degrees of Ru4d orbitals, Ru₁-MO₂ binding strength, CO-Ru₁ adsorption strength, and catalytic activity for CO oxidation can be consistently correlated by a physicochemical model based on competitive orbital couplings²⁴.

Ru₁/MO₂ SAC model preparation and single-atom structure confirmation

Ru₁/MO₂ SACs were prepared following the same protocol to minimize variables (Fig. S1). The supports were first annealed at 450 ^oC to stabilize their surfaces (Fig. S2), and then their aqueous mixtures with RuCl₃ (0.04 wt% Ru) were frozen with liquid N₂ and dried with vacuum freeze-drying. The powder precursors were then calcined in air at 400 ^oC and washed with ammonia water to yield Ru₁/MO₂ SACs (Fig. S3).

Atomic dispersions of Ru were confirmed with aberration-corrected scanning transmission electron microscopy (Ac-STEM) and extended X-ray absorption fine structure (EXAFS)¹¹. The isolated brighter spots in STEM image indicate atomic dispersions of Ru atoms (Fig. 1a). R-space EXAFS results display clear Ru-O but rare Ru-Ru bonds (Fig. 1b, S4), further confirming the atomic dispersions of Ru on MO₂ supports. Mean coordination numbers (CN) of Ru₁ are fitted to be 4.9 on TiO₂, 5.0 on SnO₂, and 5.0 on MnO₂ (Table S1). This indicates that Ru atoms display square-pyramidal coordination configurations on MO₂ supports (Fig. 1c).

Revealing Ru₁-MO₂ interaction trends and electronic-level principles

The primary SASI issue deals with M₁-support interaction mechanisms through surface coordination bonds and the effects on the stabilization and charging states of M₁ atoms. We probe the electronic-level mechanisms of Ru₁-MO₂ interactions with X-ray absorption near-edge structure (XANES) at Ru L₃ edge. In octahedral [MO₆] ligand fields, d orbitals split into π-type t_2g and σ-type e_g states²⁵, and will further split in square-pyramidal [MO₅] configurations like Ru₁/MO₂ SACs (Fig. 1c). L₃-edge spectra result from 2p_3/2→nd dipole electron transition, and thus can probe the splitting degrees and bonding states of valence nd orbitals²⁴. Ru L₃-edge spectra display clear differences among Ru₁/MO₂ SACs (Fig. 1d). In contrast, K-edge XANES data, resulting from 1s→5p electron transitions, do not show significant differences (Fig. S5). The difference indicates that Ru₁/MO₂ SACs display similar bonding states for 5p orbitals, but different for 4d orbitals.

The altered peak positions and widths of L₃ lines reflect varied coupling strengths between Ru4d and O2p orbitals (Fig. 1d)^24,25. Peak position increases following Ru₁/TiO₂ (2,842.5 eV) ≈ Ru₁/SnO₂ (2,842.6 eV) < RuO₂ (2,843.2 eV) < Ru₁/MnO₂ (2,844.1 eV). This trend indicates that Ru₁ shows the highest valence state on MnO₂ and lowest on TiO₂. Meanwhile, peak width also increases following Ru₁/TiO₂ < Ru₁/SnO₂ < RuO₂ < Ru₁/MnO₂, which also indicates different Ru4d-O2p coupling degrees. The clearly split t_2g and e_g states suggest more Ru4d-O2p overlaps in RuO₂ and on MnO₂, while Ru-O coupling is even stronger for Ru₁/MnO₂. In contrast, the narrower L₃ lines on TiO₂ and SnO₂ mean weaker Ru4d-O2p overlaps and more localizations of Ru4d orbitals. The L₃ line of Ru₁/TiO₂ is further narrower than that of Ru₁/SnO₂, suggesting that Ru4d-O2p coupling on TiO₂ are even weaker than on SnO₂. Such different orbital couplings can further lead to varied filling degrees of Ru4d orbitals. The electron densities in Ru4d orbitals increase following Ru₁/MnO₂ < Ru₁/SnO₂ < Ru₁/TiO₂, opposite to the trend of Ru-O coupling strength.

Support effects on Ru₁-MO₂ binding strengths intrinsically result from different O_sl reactivity or oxide reducibility^13,30. We characterized the reducibility of TiO₂, SnO₂, and MnO₂ supports with H₂ temperature-programed reduction (H₂-TPR, Fig. 1e). H₂ uptakes show that the reducibility increases following TiO₂ < SnO₂ < MnO₂, consistent with the trend of Ru₁-MO₂ binding strength. In addition, as oxide reducibility can be measured by the formation energy of oxygen vacancy (E_Vo), we further studied the correlation of Ru₁-MO₂ binding energy with E_Vo through calculations based on density functional theory (DFT) ¹³. Calculated results (Fig. 1f) show that lower E_Vo leads to stronger Ru₁-MO₂ binding with MnO₂, while the weakest with TiO₂. H₂-TPR and calculation results therefore indicate that O_sl reactivity fundamentally determines Ru₁-MO₂ binding strengths. Such varied Ru₁-MO₂ interactions further alter the extension and filling degrees of 4d orbitals and charging states of Ru₁ atoms.

Revealing CO-Ru₁/MO₂ adsorption trends and electronic-level principles

The second SASI issue deals with adsorbate-M₁ adsorption features and mechanisms. We use CO as a model molecule to probe the interaction trend and electronic mechanisms of CO-Ru₁ adsorptions. CO coordinates to M₁ sites through M-C bonds (Fig. 2a), and the bonding strength is synergistically enhanced by the back donations of electrons from d orbitals to CO’s π*-type lowest unoccupied molecular orbitals (LUMO, Fig. 2b)¹⁷. CO-M bonding strength therefore positively correlates to the filling degrees of d orbitals. We characterized CO adsorption on Ru₁/TiO₂ and Ru₁/SnO₂ with diffuse reflection infrared Fourier transformation spectra (DRIFTS) at -140 ^oC (Fig. 2c). No signal could be collected for Ru₁/MnO₂ owing to the dark color of MnO₂. C-O stretching modes locate at 2,128 cm^− 1 for Ru₁/TiO₂ and 2,152 cm^− 1 for Ru₁/SnO₂. This difference means weaker C-O but stronger Ru-C bonds on Ru₁/TiO₂. Therefore, CO adsorbs more strongly on Ru₁/TiO₂ than on Ru₁/SnO₂, further verifying more localization of Ru4d orbitals on TiO₂ than on SnO₂ as revealed from L₃ lines.

We further confirmed CO-Ru₁ interaction trends through DFT calculations (Fig. 2d). Calculated CO-Ru₁ adsorption energies are 0.65 eV on MnO₂, 1.23 eV on SnO₂, and 1.72 eV on TiO₂. The corresponding Ru₁-C bond lengths are 2.04, 2.03, and 1.90 Å. The greatest adsorption energy and least Ru₁-C bond length indicate the strongest adsorption of CO with Ru₁/TiO₂. This calculated Ru₁-CO adsorption trend is consistent with DRIFTS results.

The electronic mechanisms of Ru₁-MO₂ binding and CO-Ru₁ adsorption were further analyzed from the changes in the distribution states of Ru4d orbitals and electron densities. Calculated partial densities of states (PDOS, Fig. 2e) of Ru4d orbitals and CO’s π states show clear overlaps between CO’s HOMOs and Ru4d states. This indicates effective adsorptions of CO by Ru₁/MO₂ SACs. Meanwhile, the different overlapping degrees further indicate different coupling strengths²⁴. Ru₁/TiO₂ displays the greatest d-π overlap, while Ru₁/MnO₂ the least. This means that CO can form stronger coordination bonds with Ru₁/TiO₂. These calculated interaction trends agree with the experimental trends.

The direct effect of CO adsorption is to redistribute the bonding states of Ru4d orbitals, because the nature of forming new bonds and breaking old bonds is to redistribute valence orbitals³¹. In the [MO₅] ligand fields of Ru₁/MO₂ SACs, Ru4d states mainly couple with the 2p orbitals of O_sl atoms, while CO adsorption occurs through overlapping its frontier orbitals with Ru4d orbitals. This interaction polarizes Ru4d states from Fermi levels in Ru₁/MO₂ SACs into surface coordination bonds around 3.5 eV (Fig. 2e)²⁴. This orbital redistribution effect can be further confirmed by the altered charging states of Ru₁ sites¹³. For Ru₁/MO₂ SACs, the electron densities apparently accumulate within Ru-O_sl bonds but deplete on the surfaces, meaning preferential couple of Ru4d to O2p orbitals. After CO adsorption, the electron densities of Ru-O_sl bonds notably decrease for the three Ru₁/MO₂ SACs; instead, CO-Ru bonds show accumulated electron densities. These changes reveal state redistributions of Ru4d orbitals from Ru-O_sl to CO-Ru bonds. In this redistribution-based physicochemical model, CO and O_sl are competitive to overlap Ru4d orbitals. Therefore, CO adsorption strength depends on the redistribution tendency of Ru4d orbitals, which are further confined by Ru₁-MO₂ interactions²⁴.

Revealing catalytic trends of CO oxidation by Ru₁/MO₂ SACs

The third SASI issue deals with how M₁-support and adsorbate-M₁ interactions affect SAC’s catalytic performances. The goal is to reveal the roles of M₁ and supports in catalytic processes. Here we use CO oxidation as a model reaction to probe the catalytic features and trend of Ru₁/MO₂ SACs. Temperature-dependent CO conversions catalyzed by Ru₁/MO₂ and MO₂ supports display three features (Fig. 3a). First, Ru₁ can enhance CO oxidation, particularly on TiO₂ and SnO₂. The onset temperature (T_onset) decreases from 260 to 180 ^oC for TiO₂ and from 240 to 160 ^oC for SnO₂. At 260 ^oC CO conversion increases from 0.0–47.3% for TiO₂ and 3.0–100% for SnO₂. The enhancement effect of Ru₁ on MnO₂ is less significant owing to the high intrinsic activity of MnO₂. Second, Ru₁/MO₂ SACs show support-dependent activities. Both T_onset and T_100% increase as Ru₁/MnO₂ < Ru₁/SnO₂ < Ru₁/TiO₂. This trend means that Ru₁/MnO₂ shows the highest activity while Ru₁/TiO₂ the lowest. Third, the activity trend is opposite to CO-Ru₁ adsorption strength, but consistent with support reducibility. This feature suggests that CO-Ru₁ adsorption strength does not dominate the catalytic activities of Ru₁/MO₂ for CO oxidation, while O_sl reactivity should play critical roles.

We reveal the catalytic mechanisms with DFT calculations. CO oxidation follows Mars-van-Krevelen (MvK) mechanism by SACs (Fig. 3b)^15,17. This approach involves CO adsorption at M₁ sites to yield *CO species (i→ii), combination of *CO with O_sl to release CO₂ and yield V_O (iii→v), adsorption of O₂ and CO to release CO₂ and restore V_O (vi→viii→i)^4,15.

Calculated CO-Ru₁ adsorption energies are − 1.72 eV on TiO₂, -1.22 eV on SnO₂, and − 0.65 eV on MnO₂; however, Ru₁/MnO₂ is the most active for CO oxidation. The strongest CO adsorption on Ru₁/TiO₂ does not lead to the highest activity, thus CO adsorption is not the rate-determining step. This may be because CO adsorption from the gas phase is a spontaneous and rapid process and does not need activation. Instead, the activation of O_sl to release CO₂ should control the whole reaction rates, because the reaction of *CO + O_sl → CO₂ + V_O needs to break multiple O_sl-M bonds.

O_sl-support binding strength intrinsically determines O_sl reactivity and reducibility of oxide supports, and can be measured with the formation energy of V_O (E_Vo)¹³. Calculated E_Vo are − 0.87 eV for TiO₂, -2.13 eV for SnO₂, and − 3.40 eV for MnO₂ when coupling to CO oxidation (v). The activation energies (E_a) to form V_O are 1.21 eV for Ru₁/TiO₂, 0.41 eV for Ru₁/SnO₂, and 0.35 eV for Ru₁/MnO₂ (iv→v). Both the thermodynamic (E_Vo) and kinetic (E_a) energies favor the reaction of *CO with O_sl on Ru₁/MnO₂. Therefore, for CO oxidation, Ru₁ sites mainly capture CO molecules from the gas phase to form *CO, while support reducibility intrinsically determines the catalytic activity owing to the consumption of O_sl through breaking multiple O_sl-M bonds.

Competitive bonding mechanism resulting from orbital normalization

O_sl reactivity therefore can overall control the properties of Ru₁/MO₂ SACs, including Ru₁-MO₂ binding strength, charging state, CO-Ru₁ adsorption, and catalytic activity. Underlying electronic-level principle lies in two groups of competitive bonding interactions in CO-Ru₁-(O_sl)₅-MO₂ systems, Ru₁-(O_sl)₅-MO₂ and CO-Ru₁-(O_sl)₅ (Fig. 4a). The physical nature of such competitions stems from the quantum normalization attribute of electronic orbitals (or wavefunctions)²⁴.

For Ru₁-(O_sl)₅-MO₂ interactions, O2p orbitals simultaneously couple to the valence orbitals of Ru₁ and support’s metal atoms, hence M and Ru₁ are competitive to bond O_sl atoms. Weaker O_sl-M bonds (such as MnO₂) lead to higher O_sl reactivity and oxide reducibility, and further stronger Ru₁-O_sl bonds. The strong Ru₁-O_sl bonds lead to more extended 4d orbitals and delocalized d electrons. Oppositely, stronger O_sl-M bonds (such as TiO₂) decrease O_sl reactivity, oxide reducibility, and Ru₁-O_sl bonds, and further lead to more localized d states and electrons. For CO-Ru₁-(O_sl)₅ interactions, Ru4d orbitals simultaneously couple to O_sl2p orbitals and the frontier orbitals of CO, hence CO and O_sl are competitive to bond Ru atoms. Stronger Ru₁-O_sl bonds (such as Ru₁/MnO₂) can more tightly constrain Ru4d orbitals, and thus reduce its bonding capability with CO. While the less constrained Ru4d orbitals and more localized d electrons of Ru₁/TiO₂ can be more readily redistributed to form stronger CO-Ru₁ coordination bonds.

In this competitive bonding model of SACs, the bonding strengths of neighboring bonds are competitive owing to the quantum normalization attribute of electronic orbitals, such as CO-Ru₁ vs Ru₁-O_sl and Ru₁-O_sl vs O_sl-MO₂. However, non-neighboring bonds are cooperative, such as CO-Ru₁ and O_sl-MO₂ (Fig. 4b). This non-neighboring cooperativity is a type of inductive effect in chemical interactions. In addition, the competitive bonding of O_sl with MO₂ supports and *CO can basically determine the reaction rates of CO oxidation. This is because weaker O_sl-M bonds can promote the combination of *CO with O_sl and the formation of V_O.

In conclusion, we have demonstrated that support reducibility can overall control Ru₁-MO₂ bonding strength, charging state of Ru₁, CO-Ru₁ adsorption strength, and the activity of CO oxidation in CO-Ru₁-(O_sl)₅-MO₂ SAC systems. The electronic-level mechanisms of these interactions can be coherently correlated through a competitive bonding model. In this model, CO-Ru₁ and Ru₁-O_sl bonding strengths depend on the redistribution capabilities of Ru4d and O2p orbitals; however, their redistribution capabilities further inversely depend on the confinement degrees by Ru₁-(O_sl)₅ and O_sl-MO₂ bonds. This competitive bonding model can extend to adsorbate-M₁-(ligand₁)_m(ligand₂)_n-support SAC systems, and provides a general perspective to reveal the interaction trends and structure-function mechanisms of SACs. For specific SAC systems, the critical points to in-depth reveal structure-function relationships lie in clarifying electronic-level adsorption mechanisms and elementary reaction processes.

Synthesis of rutile-type TiO₂ nanorods

TiCl₄ (2.5 mL) was slowly added into ice water (30.0 mL) under vigorous stirring. After being stirred for 15 min, the solution was transferred into a 40-mL Teflon-lined stainless-steel autoclave. Then the solution was heated at 453 K for 14 h in an oven. After cooling down to room temperature, the products were collected by centrifugation and washed with deionized water, and dried at 333 K for 24 h.²⁹

Synthesis of β-MnO₂ nanorods

MnSO₄•H₂O (1.35 g) and (NH₄)₂S₂O₈ (1.83 g) were dissolved into H₂O (30.0 mL) under stirring. After being stirred for 30 minutes, the solution was transferred into a 40-mL Teflon-lined stainless-steel autoclave, and then heated at 413 K for 12 h. After cooling down to room temperature, the products were collected by centrifugation and washed with deionized water, and dried at 333 K for 24 h.³²

Synthesis of Ru₁/MO₂ (M = Mn, Sn, Ti) catalysts

All Ru₁/MO₂ catalysts were prepared following the same approach, in which only the supports were changed. First, the three MO₂ supports were annealed in air at 723 K for 10 h to stabilize their surfaces. Second, the annealed supports (1.00 g) were dispersed in H₂O (30.0 mL) under sonication for 30 min. Then the solutions were mixed with 1.00 mL of RuCl₃ solution (1.00 g∙L^− 1 RuCl₃•xH₂O in 1.00 mol·L^− 1 HCl) under sonication. The mixtures were rapidly frozen with liquid nitrogen, and dried in a freeze drier. Finally, as-obtained powers were calcined at 673 K for 4 h in air, washed with 1 mol·L^− 1 ammonia solution and deionized water to remove residual chloride ions on the surfaces, and dried at 333 K for 24 h.

Characterizations

Power X-ray diffraction (XRD) data were collected on a Bruker D8 diffractometer using Cu Kα radiation (1.5418 Å), which were operated at 40 kV and 40 mA with a scanning rate of 6 degree·min^− 1. AC-STEM was performed on JEM-ARM200F at 300 kV. H₂ temperature-programmed reduction (H₂-TPR) was performed on an Autochem1 II 2920 instrument. Before TPR measurement, the samples were treated with Ar at 573 K for 60 min, and then cooled down to 323 K. The signal was recorded online with a thermal conductivity detector (TCD), as the reactor was heated to 1073 K at a heating rate of 10 K·min^− 1 under 10% H₂/Ar flow.

In-situ diffuse reflectance infrared Fourier transform spectra (DRIFTS) of CO adsorption were performed on Thermo fisher Nicolet 6700 FTIR equipped with an MCT detector and a low-temperature reaction cell. Prior to CO adsorption, the samples were annealed in Ar flow at 573 K for 1 h and then cooled down to 133 K with liquid nitrogen. After collecting the backgrounds, the infrared absorption spectra were collected under different CO pressures with 256 scans at a resolution of 4 cm^− 1.

The X-ray absorption data at the Ru-K edge of the samples were recorded at room temperature in fluorescence mode using a solid-state detector at the 1W1B beamline of the Beijing Synchrotron Radiation Facility, China. Data analysis was performed with Demeter XAS package under a standard protocol. The amplitude reduction factors, S_o² of R edge were obtained from fitting the spectrum of reference samples including Ru assuming the coordination number of the first shell as 12. The X-ray absorption data at the Ru-L₃ edge of the samples were recorded at room temperature in fluorescence mode at 4B7A the beamline of the Beijing Synchrotron Radiation Facility, China.

Catalytic Tests Of Co Oxidation:

CO oxidation reaction was carried out in a fixed-bed reactor made of stainless steel. First, the catalysts (100 mg) were was treated using the reaction gas with V(CO):V(O₂):V(Ar) = 2:18:80 at 573 K for 1 h at a gas hourly space velocity (GHSV) of 30,000 mL·g^− 1·h^− 1. After cooling down to room temperature, the temperature-programmed reactions were performed. The outgassing gas composition were detected by an online gas chromatography (GC) using a TCD detector, which were sampled after stabilization at specific temperatures for 25 min.

Density Functional Theory (DFT) Calculations:

All the calculations were performed with Vienna Ab-initio Simulation Package (VASP) code³³. The projector augmented wave (PAW) method³⁴ was used for treating electron-ion interactions. The valence electrons of Ru, Ti, Sn, Mn, O and C were described by plane-wave basis sets with cutoff energies of 400 eV, while the inner electrons were kept frozen. The exchange-correlation interactions were treated with the Perdew-Burke-Ernzerhof (PBE) functional³⁵. DFT + U method was used for correcting the strong electron-correlation properties of oxides³⁶. The U values are 2.5 eV, 3.5 eV, and 5.1 eV for Ti³⁷, Sn³⁸, and Mn³⁹, respectively.

The slab models of TiO₂(110), SnO₂(110) and MnO₂(110) were cleaved from bulk cells and modeled as periodic slabs with three layers. Vacuum slabs of 15 Å were built between the slabs. Supercells of 5 × 3 were utilized and the bottom two layers were fixed. One of the Ti, Sn, or Mn atoms in the slabs were replaced by Ru atoms to model Ru₁/MO₂ SACs.

The structure models were optimized until the forces were below 0.02 eV·Å⁻¹. For structure relaxations, Monkhorst–Pack (1 × 1 × 1) Γ-centered k-points grid sampling within the Brillouin zones were used for TiO₂ and SnO₂ cells, while k-points of 4 × 4 × 1 were applied for MnO₂ cells. To calculate the density of states (DOS), k-points of 4 × 4 × 1 were used for all cells. Transition states were calculated using the climbing-image nudged elastic band (CI-NEB) methods⁴⁰.

The adsorption energy was calculated by Eq. (1)

E _ads = E_CO−Ru1/MO2 - E_Ru1/MO2 - E_CO (1)

where E_CO−Ru1/MO2, E_Ru1/MO2, and E_CO are the energies of CO adsorbed on Ru₁/MO₂ surfaces, bare Ru₁/MO₂ surfaces, and CO molecule, respectively.

The transition states of CO oxidation were calculated following Mars-van-Krevelen (MvK) mechanism. The formation energy of oxygen vacancy was calculated by Eq. (2):

E _Ovac = 1/2 E_O2 + E_Ru1/MO2−x - E_Ru1/MO2 (2)

where E_O2, E_Ru1/MO2−x and E_Ru1/MO2 are the energies of O₂ molecule, Ru₁/MO₂ surfaces with oxygen vacancies and bare Ru₁/MO₂ surfaces, respectively.

Bader charges were applied for atomic charge computations⁴¹, and charge density differences were represented by Eq. (3):

Δρ = ρ_A+B - ρ_A – ρ_B (3)

where ρ_X is the charge density of species X.

ACKNOWLEDGMENTS

This work was supported by the National Natural Science Foundation of China (21801012). We thank the 4B7A and 1W1B beamlines of the Beijing Synchrotron Radiation Facility (BSRF). We highly appreciate Dr. Yang-Gang Wang for the guide, discussion, and support of DFT calculations.

Competing interests

The authors declare no competing interests.

Author contributions:

G. X. initiated, designed, and supervised this work, analyzed the results and wrote the manuscript. J. Z. prepared the catalysts and performed catalytic experiments. Q. G. performed DFT calculations. Y. L. and W. H. contributed to DRIFTS measurements. G. Z., S. C., L. Z., and L. Z contributed to XAFS measurements and data analysis. X. L. helped with catalytic tests. J. Z. and Q. G. contributed equally.

Data and materials availability: All data are available in the main text or the supplementary materials.

Wang, A. Q., Li, J. & Zhang, T. Heterogeneous single- atom catalysis. Nat. Rev. Chem. 2, 65-81 (2018).
Cui, X. J., Li, W., Ryabchuk, P., Junge, K. & Beller, M. Bridging homogeneous and heterogeneous catalysis by heterogeneous single-metal-site catalysts. Nat. Catal. 1, 385-397 (2018).
Liu, L. C. & Corma, A. Metal Catalysts for Heterogeneous Catalysis: From Single Atoms to Nanoclusters and Nanoparticles. Chem. Rev. 118, 4981-5079 (2018).
Qiao, B. T. et al. Single-atom catalysis of CO oxidation using Pt₁/FeO_x. Nat. Chem. 3, 634-641 (2011).
Beniya, A. & Higashi, S. Towards dense single-atom catalysts for future automotive applications. Nat. Catal. 2, 590-602 (2019).
Ye, L. et al. Ceria Nanocrystals Supporting Pd for Formic Acid Electrocatalytic Oxidation: Prominent Polar Surface Metal Support Interactions. ACS Catalysis 9, 5171-5177 (2019).
Yang, J. R., Li, W. H., Wang, D. S. & Li, Y. D. Electronic Metal-Support Interaction of Single-Atom Catalysts and Applications in Electrocatalysis. Adv. Mater. 32, 2003300 (2020).
Lee, B. H. et al. Reversible and cooperative photoactivation of single-atom Cu/TiO₂ photocatalysts. Nat. Mater. 18, 620-626 (2019).
Ji, S. F. et al. Matching the kinetics of natural enzymes with a single-atom iron nanozyme. Nat. Catal. 4, 407-417 (2021).
Wang, L. et al. Boosting Activity and Stability of Metal Single-Atom Catalysts via Regulation of Coordination Number and Local Composition. J. Am. Chem. Soc. 143, 18854-18858 (2021).
Lang, R. et al. Single-Atom Catalysts Based on the Metal-Oxide Interaction. Chem. Rev. 120, 11986-12043 (2020).
Sarma, B. B. et al. Metal-Specific Reactivity in Single-Atom Catalysts: CO Oxidation on 4d and 5d Transition Metals Atomically Dispersed on MgO. J. Am. Chem. Soc. 142, 14890-14902 (2020).
O'Connor, N. J., Jonayat, A. S. M., Janik, M. J. & Senftle, T. P. Interaction trends between single metal atoms and oxide supports identified with density functional theory and statistical learning. Nat. Catal. 1, 531-539 (2018).
Ye, X. X. et al. Insight of the stability and activity of platinum single atoms on ceria. Nano Res. 12, 1401-1409 (2019).
Nie, L. et al. Activation of surface lattice oxygen in single-atom Pt/CeO₂ for low-temperature CO oxidation. Science 363, 1419-1423 (2019).
Therrien, A. J. et al. An atomic-scale view of single-site Pt catalysis for low-temperature CO oxidation. Nat. Catal. 1, 192-198 (2018).
Hulva, J. et al. Unraveling CO adsorption on model single-atom catalysts. Science 371, 375-379 (2021).
Lu, Y. B. et al. Identification of the active complex for CO oxidation over single-atom Ir-on-MgAl₂O₄ catalysts. Nat. Catal. 2, 149-156 (2019).
Kaden, W. E., Wu, T. P., Kunkel, W. A. & Anderson, S. L. Electronic Structure Controls Reactivity of Size-Selected Pd Clusters Adsorbed on TiO₂ Surfaces. Science 326, 826-829 (2009).
Jeong, H. et al. Controlling the Oxidation State of Pt Single Atoms for Maximizing Catalytic Activity. Angew. Chem. Int. Ed. 59, 20691-20696 (2020).
Li, W. et al. CO₂ Hydrogenation on Unpromoted and M-Promoted Co/TiO₂ Catalysts (M = Zr, K, Cs): Effects of Crystal Phase of Supports and Metal–Support Interaction on Tuning Product Distribution. ACS Catalysis 9, 2739-2751 (2019).
Wan, J. W. et al. Defect Effects on TiO₂ Nanosheets: Stabilizing Single Atomic Site Au and Promoting Catalytic Properties. Adv. Mater. 30, 1705369 (2018).
Pereira-Hernandez, X. I. et al. Tuning Pt-CeO₂ interactions by high-temperature vapor-phase synthesis for improved reducibility of lattice oxygen. Nat. Commun. 10, 1358 (2019).
Xiang, G. L. & Wang, Y. G. Exploring electronic-level principles how size reduction enhances nanomaterial surface reactivity through experimental probing and mathematical modeling. Nano Res. 15, 3812-3817 (2022).
Xiang, G. L. et al. Probing Ligand-Induced Cooperative Orbital Redistribution That Dominates Nanoscale Molecule-Surface Interactions with One-Unit-Thin TiO₂ Nanosheets. Nano Lett. 18, 7809-7815 (2018).
Yao, Y. C. et al. Engineering the electronic structure of single atom Ru sites via compressive strain boosts acidic water oxidation electrocatalysis. Nat. Catal. 2, 304-313 (2019).
Norskov, J. K., Abild-Pedersen, F., Studt, F. & Bligaard, T. Density functional theory in surface chemistry and catalysis. Proc. Natl. Acad. Sci. USA 108, 937-943 (2011).
Fu, Z. M., Yang, B. W. & Wu, R. Q. Understanding the Activity of Single-Atom Catalysis from Frontier Orbitals. Phys. Rev. Lett. 125, 156001 (2020).
Zhou, J. et al. Interfacial compatibility critically controls Ru/TiO₂ metal-support interaction modes in CO₂ hydrogenation. Nat. Commun. 13, 327 (2022).
Jones, J. et al. Thermally stable single-atom platinum-on-ceria catalysts via atom trapping. Science 353, 150-154 (2016).
Chakhalian, J. et al. Orbital reconstruction and covalent bonding at an oxide interface. Science 318, 1114-1117 (2007).
Wang, X. & Li, Y. Selected-Control Hydrothermal Synthesis of α- and β-MnO₂ Single Crystal Nanowires. J. Am. Chem. Soc. 124, 2880-2881 (2002).
Kresse, G. & Hafner, J. Ab initio molecular dynamics for liquid metals. Phys. Rev. B 47, 558-561 (1993).
Blöchl, P. E. Projector augmented-wave method. Phys. Rev. B 50, 17953-17979 (1994).
Perdew, J. P., Burke, K. & Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 77, 3865-3868 (1996).
Anisimov, V. I., Aryasetiawan, F. & Lichtenstein, A. I. First-principles calculations of the electronic structure and spectra of strongly correlated systems: the LDA + U method. J. Phys. Condens. Matter 9, 767-808 (1997).
Morgan, B. J. & Watson, G. W. A DFT+U description of oxygen vacancies at the TiO₂ rutile (110) surface. Surf. Sci. 601, 5034-5041 (2007).
Li, S., Lu, Z., Yang, Z. & Chu, X. The sensing mechanism of Pt-doped SnO₂ surface toward CO: A first-principle study. Sensors and Actuators B 202, 83-92 (2014).
Tompsett, D. A. & Islam, M. S. Surfaces of Rutile MnO₂ Are Electronically Conducting, Whereas the Bulk Material Is Insulating. J. Phys. Chem. C 118, 25009-25015 (2014).
Henkelman, G., Uberuaga, B. P. & Jónsson, H. A climbing image nudged elastic band method for finding saddle points and minimum energy paths. J. Chem. Phys. 113, 9901–9904 (2000).
Bader, R. F. W. A quantum theory of molecular structure and its applications. Chem. Rev. 91, 893-928 (1991).

There is NO Competing Interest.

Ru1COSInatureComm.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Revealing the intrinsic electronic-level principle driving metal-support interaction trends of single-atom Ru with oxides through cooperative orbital coupling

Status:

Version 1

Abstract

Figures

Introduction

Results

Ru₁/MO₂ SAC model preparation and single-atom structure confirmation

Revealing Ru₁-MO₂ interaction trends and electronic-level principles

Revealing CO-Ru₁/MO₂ adsorption trends and electronic-level principles

Revealing catalytic trends of CO oxidation by Ru₁/MO₂ SACs

Discussions

Competitive bonding mechanism resulting from orbital normalization

Conclusion

Methods

Characterizations

Catalytic Tests Of Co Oxidation:

Density Functional Theory (DFT) Calculations:

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1

Revealing the intrinsic electronic-level principle driving metal-support interaction trends of single-atom Ru with oxides through cooperative orbital coupling

Status:

Version 1

Abstract

Figures

Introduction

Results

Ru1/MO2 SAC model preparation and single-atom structure confirmation

Revealing Ru1-MO2 interaction trends and electronic-level principles

Revealing CO-Ru1/MO2 adsorption trends and electronic-level principles

Revealing catalytic trends of CO oxidation by Ru1/MO2 SACs

Discussions

Competitive bonding mechanism resulting from orbital normalization

Conclusion

Methods

Characterizations

Catalytic Tests Of Co Oxidation:

Density Functional Theory (DFT) Calculations:

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1

Ru₁/MO₂ SAC model preparation and single-atom structure confirmation

Revealing Ru₁-MO₂ interaction trends and electronic-level principles

Revealing CO-Ru₁/MO₂ adsorption trends and electronic-level principles

Revealing catalytic trends of CO oxidation by Ru₁/MO₂ SACs