Electronic communications between active sites on individual metallic nanoparticles in catalysis

Xu, Dongrun; Jin, Yaowei; He, Bowen; Fang, Xue; Chen, Guokang; Qu, Weiye; Xu, Chenxin; Chen, Junxiao; Ma, Zhen; Chen, Liwei; Tang, Xingfu; Liu, Xi; Wei, Guangfeng; Chen, Yaxin

doi:10.1038/s41467-024-52997-w

Download PDF

Article
Open access
Published: 04 October 2024

Electronic communications between active sites on individual metallic nanoparticles in catalysis

Nature Communications volume 15, Article number: 8614 (2024) Cite this article

4377 Accesses
3 Citations
11 Altmetric
Metrics details

Subjects

Abstract

Catalytic activity of metal particles is reported to originate from the appearance of nonmetallic states, but conductive metallic particles, as an electron reservoir, should render electron delivery between reactants more favorably so as to have higher activity. We present that metallic rhodium particle catalysts are highly active in the low-temperature oxidation of carbon monoxide, whereas nonmetallic rhodium clusters or monoatoms on alumina remain catalytically inert. Experimental and theoretical results evidence the presence of electronic communications in between vertex atom active sites of individual metallic particles in the reaction. The electronic communications dramatically lower apparent activation energies via coupling two electrochemical-like half-reactions occurring on different active sites, which enable the metallic particles to show turnover frequencies at least four orders of magnitude higher than the nonmetallic clusters or monoatoms. Similar results are found for other metallic particle catalysts, implying the importance of electronic communications between active sites in heterogeneous catalysis.

A Ta-TaS₂ monolith catalyst with robust and metallic interface for superior hydrogen evolution

Article Open access 18 October 2021

High-entropy single-atom activated carbon catalysts for sustainable oxygen electrocatalysis

Article 30 March 2023

Fast site-to-site electron transfer of high-entropy alloy nanocatalyst driving redox electrocatalysis

Article Open access 28 October 2020

Introduction

A majority of catalytic oxidation reactions often involve the catalyst-mediated electron transfer between reactants^1,2,3,4, in which a catalyst acts as an electron reservoir by accepting electrons from one reactant and then delivering them to other reactant(s)⁵, and thus the electronic states of the catalyst, particularly in terms of nonmetallic states or metallic states, have an important influence on reaction rates^6,7,8,9,10. There exist two prevailing yet diametrically opposed views on the function of the electronic states, emphasizing the importance of either nonmetallic states^6,7 or metallic states^8,9,10 in determining catalytic performance. The fundamental discrepancy between these views lies in whether there exists electronic communications between active sites when a reaction takes place¹¹. Nonmetallic states lead to a lack of electronic communications, so catalytic reactions merely occur locally on individual active sites. In contrast, metallic states can give access to electronic communications for active sites, thus opening up an electron-delocalized mechanism (EDM)¹¹.

In order to test both views, an elaborately designed system including a prototypical reaction and model catalysts is necessary. Carbon monoxide (CO) oxidation is commonly used as a typical probing reaction, yet it is significantly important in industrial catalysis¹². According to the Sabatier principle on medium adsorption strength¹³, metals with not fully unfilled frontier dⁿ (particularly, 5 < n < 10) orbitals are often favorable for effectively adsorbing and activating CO and O₂ molecules^14,15, and thus rhodium (Rh) with an electronic configuration of d⁸s¹ at its ground state should be a suitable candidate¹⁶. As the metal atomicity increases, there often exists a nonmetallic-to-metallic state transition region^{6,17,18,19,20}, and, as typical representatives for metallic and nonmetallic catalysts, two reliable Rh catalyst models should require their atomicity respectively to approach two sides of this transition region. A catalytically inert and electrically insulating support such as Al₂O₃ is also desired to disperse and stabilize Rh clusters or nanoparticles²¹, and to eliminate or rule out the so-called electronic metal-support interaction^22,23.

In this work, we provide direct evidence that electronic communications are present between active sites on individual metallic particles, which enable metallic rhodium particles to possess four orders of magnitude higher turnover frequency than that of nonmetallic particles without such communications. Experimental and theoretical results demonstrate that inter-site electron communications on the metallic particles drive two half-reactions coupling between active sites, thus allowing CO oxidation to proceed along a lower-activation-energy kinetic pathway than the nonmetallic clusters or monoatoms.

Results

Design of metallic and nonmetallic model catalysts

To achieve metallic Rh nanoparticles with desired sizes and well-defined nanofacets²⁴, a colloid preparation method is used to obtain colloidal Rh nanoparticles. These Rh nanoparticles are then deposited onto Al₂O₃ surfaces to gain a metallic Rh_met/Al₂O₃ catalyst. For comparison, a nonmetallic Rh_non/Al₂O₃ catalyst is prepared by a conventional impregnation route (See Supplementary Figs. 1 and 2, Supplementary Table 1, and Methods for details). The metallic and nonmetallic states of Rh_met and Rh_non are verified by the valance band X-ray photoelectron spectra (VB XPS)²⁵. In Fig. 1a, Rh_met/Al₂O₃ shows a significant electronic state density on the Fermi level (E_F), evidencing that the Rh nanoparticles are in metallic states, whereas the electronic state density of Al₂O₃ appears at approximately 3.2 eV below E_F (Supplementary Fig. 3)²⁶. For comparison, the electronic state density nearest to E_F for Rh_non/Al₂O₃ is at −0.6 eV (Fig. 1b), corresponding to nonmetallic states. These results are consistent with the X-ray absorption near-edge structure data (Supplementary Fig. 4).

**Fig. 1: Metallic and nonmetallic model catalysts.**

The electronic states of metal particles are intimately associated with particulate sizes on the (sub)nanometer scale. Fig. 1c, d displays the aberration-corrected scanning transmission electron microscopy (AC-STEM) images of Rh_met/Al₂O₃ and Rh_non/Al₂O₃, respectively. The Rh particles of Rh_met/Al₂O₃ exhibit sizes of 2–3 nm, with an average size of 2.3 nm (Supplementary Fig. 5 and inset in Fig. 1c). A typical Rh nanoparticle together with a corresponding Wulff model (~420 atoms) is shown in Fig. 1e. The Rh particles in this size range are evidenced to possess the metallic behavior (Fig. 1a), consistent with the experimental finding that metal particles with sizes larger than 1.5 nm are in metallic states^6,20. On the other hand, the average size of the Rh clusters in Rh_non/Al₂O₃ is ~0.9 nm (Fig. 1d, f, and Supplementary Fig. 6). Inasmuch as the nonmetallic-to-metallic state transition for some metal particles, including Rh particles, often occurs in a size range of 1–2 nm^6,18,19,20, Rh_met/Al₂O₃ and Rh_non/Al₂O₃ with the Rh sizes on two sides of the state transition region can act as two reliable representatives for metallic and nonmetallic catalysts, respectively.

Owing to the low work function of Rh (5.5 eV)^27,28, O₂ dissociation adsorption energies (E_ad) on Rh particles are large (Supplementary Fig. 7). Therefore, Rh particles are often covered by one surface oxide layer¹⁵. O₂ dissociation adsorption on Rh particles is evidenced for both Rh_met/Al₂O₃ and Rh_non/Al₂O₃, as shown by the Rh-O bonds in Fourier-transformed extended X-ray absorption fine structure (FT EXAFS) spectra (Supplementary Fig. 8) and wavelet transform (WT) EXAFS spectra in Fig. 1g, h, thus leading to the partial positively charged Rh atoms via donating electrons to the adsorbed O species (Supplementary Fig. 9). Likewise, upon CO adsorption, surface Rh atoms will accept electrons from adsorbed CO molecules²⁹. As for one metallic particle, the electron extraction or injection occurring on one active site enables the charges to be redistributed on the entire metallic particle, leading to electronic communications with another entangled active site, whereas such site-site electronic communications do not exist on nonmetallic particles in principle. Hence, Rh_met/Al₂O₃ and Rh_non/Al₂O₃ are reliable model catalysts to test both views aforementioned.

Evidence for inter-site electronic communications

With two model catalysts, we investigate the possibility of electronic communications between active sites on individual metallic particles in CO oxidation. Although Rh_non/Al₂O₃ has much more catalytic sites than Rh_met/Al₂O₃ in virtue of smaller particles exposing more Rh atoms on surfaces, Rh_met/Al₂O₃ shows much higher activity in CO oxidation than Rh_non/Al₂O₃ under identical conditions (Fig. 2a and Supplementary Fig. 10). CO oxidation on Rh_met/Al₂O₃ starts at 60 ^oC and reaches 100% conversion at 130 ^oC, at which Rh_non/Al₂O₃ remains catalytically inactive. Fig. 2b displays apparent activation energies (E_a) extracted from the Arrhenius plots (Supplementary Fig. 11), and the E_a value of Rh_met/Al₂O₃ is ~40 kJ mol⁻¹, in line with the reported value for CO oxidation over Rh nanoparticle catalysts³⁰. A large difference (60 kJ mol⁻¹, or ~0.6 eV) in E_a between Rh_met/Al₂O₃ and Rh_non/Al₂O₃ coincidentally equals minimal energy required for an electron transition from the ground state to E_F of Rh_non/Al₂O₃ (Fig. 1b), implying the validity of the interfacial energy-alignment principle^31,32 that a molecule’s HOMO (highest occupied molecular orbital) level can become pinned to the catalyst’s E_F level on which to exchange the charge. Considering the discrepancy in electronic state between two catalysts, we infer that the high activity and the low E_a of Rh_met/Al₂O₃ are closely related to the metallic states.

**Fig. 2: Evidence for the electronic communications between active sites on the metallic particles.**

One important feature of the metallic states is that the valence band and the conduction band coincide on E_F, where to extract or inject electrons thus becomes easy¹⁴. The formation of the surface oxide layer (due to O₂ + 4e⁻ → 2O²⁻) reflects that electrons extracted from E_F of the metallic Rh particles are energetically favorable (Fig. 1g, and Supplementary Fig. 8). In the O₂ temperature programmed desorption (O₂-TPD) procedure profile of Fig. 2c, a low-temperature O₂ desorption peak at 145 ^oC due to O²⁻ → ½O₂ + 2e⁻ indicates the electron injection onto E_F of the metallic Rh particles, whereas such an electron injection is rather difficult for the nonmetallic Rh clusters judging from the neglected O₂ desorption amount from Rh_non/Al₂O₃. Likewise, a new band ( ~ 2116 cm⁻¹) appears at temperatures higher than 250 ^oC in in situ diffuse reflectance infrared Fourier transform (DRIFT) spectra of Rh_non/Al₂O₃ (Fig. 2d and Supplementary Fig. 12), which is assigned to the C = O vibration of O²⁻-Rh²⁺-CO species (one important reaction intermediates)³³. The existence of O²⁻-Rh²⁺-CO indicates that the reaction between co-adsorbed CO and O²⁻ on individual electrically insulating Rh atoms (CO + O²⁻ → CO₂ + 2e⁻) would not occur, implying that it is difficult to inject electrons to the nonmetallic Rh clusters.

The identification of active sites is an essential prerequisite for studying site-site electronic communications. The oxide layers on the metallic particle nanofacets are evidenced to be a hexagonal O-Rh-O trilayer¹⁵, which are catalytically inactive in low-temperature CO oxidation³⁴, whereas the adsorbed O species from the edge Rh atoms are active and readily desorbed to form under-coordinated sites (Fig. 2c). Subtly, these under-coordinated sites consist of the vertex Rh atoms and the edge Rh atoms not including ending atoms with an atomic ratio of ~1:8 (inset of Fig. 2e, and Supplementary note 1), which can be distinguished by using CO temperature-programmed desorption (CO-TPD) procedure. In Fig. 2e, two CO desorption peaks (denoted as P_L and P_H) are observed with a calculated P_L/P_H ratio in desorption amount of ~1:8, which virtually equals the ratio of the vertex Rh atoms to the edge atoms, and thus P_L and P_H can be attributed to CO desorption from the vertex Rh atoms and the edge atoms, respectively. Obviously, the adsorbed CO molecules on the edge Rh atoms are more stable, and remain catalytically inert in the CO oxidation atmosphere even at temperatures higher than 130 ^oC (Fig. 2f) (see also discussion in Supplementary note 2), at which CO conversion on Rh_met/Al₂O₃ reaches 100% (Fig. 2a). Only the CO molecules adsorbed on the vertex Rh atoms are active at low temperatures, evidencing the vertex Rh atoms as active sites.

Every vertex Rh atom on the well-defined-facet Rh particles has identical geometrical and electronic structures, which allows one to acquire site-site electronic communication information on the basis of macroscopic data measured from these uniform active sites by using site-specific techniques³⁵. The electronic communications originate essentially from the charge redistribution of the charged active sites on the entire metallic particle, and, as for a metallic particle with well-defined structure, such a redistributed charge to other vertex active sites has proven to be more than that to the edge atoms³⁶. To verify whether such electronic communications exist on individual Rh metallic particles, we can monitor the electronic perturbations of the edge atoms by the vertex charged active sites. The DRIFT spectroscopy of CO on noble metal catalysts is site-specific and sensitive to electronic perturbations of the binding sites^33,37 (Fig. 2g). In Fig. 2h, we first collect a DRIFT spectrum of CO-saturated Rh_met/Al₂O₃ at 30 ^oC, and then collect another DRIFT spectrum after heating the sample up to 150 ^oC to make the vertex active sites positively charged after O₂ introduction via substituting the adsorbed CO by surface lattice oxygen (CO + O₂ + 2e⁻ → CO₂ + O²⁻). An observed blue-shift by 11 cm⁻¹ (Fig. 2h) elucidates a charge decrease of the edge atoms (Fig. 2g and Supplementary note 3), reflecting the charge redistribution on the metallic particle. These results evidence that there exist electronic communications between active sites on individual metallic particles. The electronic communications can speed up CO oxidation so as to achieve a high turnover frequency (TOF = 11.2 s⁻¹) at 160 ^oC, at least four orders of magnitude higher than TOF (6.0 × 10⁻⁴s⁻¹) of the nonmetallic clusters or the monoatoms without such electronic communications (Supplementary Fig. 13 and Supplementary note 4).

Mechanisms of electronic-communications-driven reactions

To shed light on reasons why the electronic communications can enhance the reaction rate, we carry out DFT simulations to explore CO oxidation mechanisms over Rh_met/Al₂O₃. A reliable metallic particle Wulff structure containing thirty-eight Rh atoms (see Supplementary Fig. 7 for details) with nanofacets covered by twenty-four O atoms is established, which exposes twenty-four vertex atoms as active sites on surfaces. Noticing that the calculated adsorption free energies of O and CO on the vertex active sites are almost identical, twelve CO molecules and twelve O atoms, as electron donor and acceptor, respectively, are alternately adsorbed on twenty-four vertex active sites (Supplementary Fig. 14), which allows every active site to have an identical chemical potential such that there is no driving force for electron transfer between active sites, and thus we can randomly choose one of the active sites on which to simulate a catalytic cycle with the local reaction mechanism (LRM)¹¹. To simulate the EDM with inter-site electronic communications¹¹, eleven sites of the twelve adsorbed CO were replaced by adsorbed O atoms (Supplementary Fig. 14), on each of which the half-reaction of ½O₂ + 2e⁻ → O²⁻ is assumed to have taken place via extracting two electrons from each active site, thus allowing the metallic particle to be positively charged as the result of the charge redistribution. Therefore, we simulate a catalytic cycle at the reserved CO adsorption site.

The DFT calculation results on LRM and EDM of CO oxidation over Rh_met/Al₂O₃ are shown in Fig. 3a, in which two kinds of the reaction mechanisms follow a same pathway. There are two critical transition states (TSs) corresponding to the O-CO bond formation steps in the CO oxidation catalytic cycle for both cases (Inset of Fig. 3a and Supplementary Fig. 15), which are involved in the electron-injection and electron-extraction processes: (i) CO + O²⁻ → CO₂ + 2e⁻, and (ii) CO + O₂ + 2e⁻ → CO₂ + O²⁻. Owing to the strong intrinsic activation ability of the metallic Rh particles toward O₂¹⁵, the energy required for extracting electrons from the active sites is lower than that for injecting electrons, and the reaction (i) with TS1 is thus the rate-determining step for both cases. In Fig. 3b, the energy barrier (0.57 eV) of the rate-determining step for EDM is lower than that (0.80 eV) for LRM despite sharing the same pathway, and the former agrees satisfactorily with the experimental E_a value (0.42 eV). Thermodynamically, an exothermic process with an energy saving of 0.17 eV for the rate-determining step allows CO oxidation to follow EDM rather than LRM with the endothermic rate-determining step requiring an energy input of 0.27 eV. These results are consistent with the fact that the EDM is commonly present in metallic particle catalysis^4,11,38,39. By coupling two redox half-reactions, the inter-site electronic communications can effectively reduce the energy barrier of the rate-determining step, thus lowering E_a and enhancing the overall reaction rate.

**Fig. 3: Reaction mechanisms and activation of adsorbed CO molecules in reaction.**

We focus on the charge transfer between the active site and the reactants at the rate-determining step (CO + O²⁻ → CO₂ + 2e⁻) for the low-E_a EDM. To produce CO₂ with two C = O bonds^14,40 from CO with a C ≡ O bond (Fig. 3c), the cleavage of one 1π bond and the depletion of two 5σ electrons of CO are required for forming a new C = O bond. At the adsorption state (Fig. 3c), the d → 2π* back-donation⁴¹ by ~1.7e⁻ and the 1π → d donation by ~2.9e⁻ evidence that one CO 1π bond has been cleaved to release one C 2p empty orbital. One neighboring O²⁻ attacks this C 2p empty orbital to energetically favorably form one new C-O π bond, during which the 5σ → d donation⁴¹ by ~1.3e⁻ more occurs at TS1 to subsequently form one new C-O σ bond (Fig. 3d and Supplementary Fig. 16), and these two new bonds constitute one C = O bond of CO₂. As a result, the energy barrier at TS1 originates mainly from the depletion of CO 5σ electrons in order to form the new C-O σ bond. Further d → 2π* back-donation by ~0.8e⁻ at TS1, which partially offsets the incremented positive charge of the adsorbed CO molecule (Fig. 3c and Supplementary Table 2), is conducive to the depletion of 5σ electrons. Although the total amount of the charges transferred from the CO 1π, 5σ, and 4σ orbitals onto the active site at the adsorption state is as large as ~4.1e⁻ (Supplementary Table 2), these electronic charges are rapidly neutralized on this positively charged metallic particle, thus allowing the active site to further successively extract CO 5σ electrons only with a small energy input till the new C-O σ bond forms (Fig. 3b).

Besides, we calculate the average charge of core atoms in metallic Rh particle (Supplementary Table 3). The atoms maintain metallic states during the reaction cycle and have different electronic charges (negatively or positively charged) as a response of the different oxidation-reduction steps on the active sites, which can provide a channel for the electron transfer⁴². Meanwhile, the Rh atoms at nanofacets are covered by O species and these atoms stay in oxidation states (Supplementary Table 3) which should have no influence on the electronic communication during the CO oxidation on Rh_met/Al₂O₃. Thus, the electrons can be transferred between different vertex Rh sites of Rh_met/Al₂O₃ during the oxidation-reduction reaction cycle and the electronic communications couple two half-reactions to reduce the energy barrier.

Universal electronic communications on individual metallic particles

The electronic communications are not unique for metallic Rh particles, and we select the typical top metal (Pt) of a Sabatier volcano plot⁴³ to study the influence of the electronic communications on activity. According to the particle sizes ( > 1.5 nm) required for the nonmetallic-to-metallic state transition^6,20, we support metallic Pt particles or nonmetallic Pt clusters on Al₂O₃ to achieve Pt_met/Al₂O₃ or Pt_non/Al₂O₃, respectively (Supplementary Figs. 1 and 17–19). An average size for the Pt particles on Pt_met/Al₂O₃ or Pt_non/Al₂O₃ is approximately 2.4 or 0.7 nm, respectively (Fig. 4a or 4b). Pt_met/Al₂O₃ shows the metallic states evidenced by the XANES spectra at Pt L₃-edge⁴⁴ and the CO DRIFT spectra^45,46 (Supplementary Figs. 20–21). As expected, Pt_met/Al₂O₃ exhibits much higher activity and lower E_a than Pt_non/Al₂O₃ in CO oxidation (Fig. 4c and inset, Supplementary Figs. 22–23), which suggests that metallic particles can lower the E_a via inter-site electronic communications. Such an inter-site electronic communication on metallic particles appears to be a common phenomenon in other catalytic reactions^{4,11,38,39,47,48,49,50}.

Analogous to two coupled electrochemical half-reactions, electronic communications on individual metallic particles can directly couple two half-reactions occurring on two different active sites in a spatially synchronized way in thermochemical catalytic oxidation⁴. As exemplified by using metallic Rh particle in CO oxidation (the upper half of Fig. 4d), upon being exposed in the reaction atmosphere, metallic Rh particles are often covered with dissociated oxygen via one half reaction: ① ½O₂ + 2e⁻ → O²⁻, which extracts two electrons from the active site I, inducing a charge redistribution on metallic Rh particle. As predicted by DFT calculations (Fig. 3), the charge redistribution energetically favorably triggers the other half reaction occurring on one entangled active site II: CO + O²⁻ → CO₂ + 2e⁻, thus realizing electrons delivery from active site II to active site I. Likewise, electrons will transfer from active site I to active site II, as each active site finishes the other half-reaction to close a catalytic cycle. Similar findings were also reported in heterogeneous catalysis¹¹ or biocatalysis⁵¹, in which the coupling of two half-reactions could accelerate reaction rates to a great extent.

To investigate the driving force of two half-reactions coupling, we shut off channels for the inter-site electronic communications via separating single Rh or Pt atoms by an average distance of more than 3 nm in between on Al₂O₃ surfaces (the down half of Fig. 4d), as shown in the STEM images of Rh₁/Al₂O₃ and Pt₁/Al₂O₃ together with the insets of the atom-atom distance statistics in Fig. 4e or f, respectively (see also Supplementary Figs. 1 and 24–27). The relatively far distance between individual metal monoatoms (active sites) anchoring on the electronically insulating Al₂O₃ surfaces keeps the catalysts nonmetallic (Supplementary Figs. 20 and 28) and hampers two half-reactions coupling in CO oxidation. The E_a values and the TOFs (Fig. 4g, and Supplementary Figs. 29, 30) for Rh₁/Al₂O₃ and Pt₁/Al₂O₃ are in stark contrast with the values for Rh_met/Al₂O₃ and Pt_met/Al₂O₃, as would be expected for CO oxidation catalyzed by Rh_non/Al₂O₃ (Fig. 2b) or Pt_non/Al₂O₃ (Fig. 4c) without inter-site electronic communications. Hence, the electronic communications are the driving force to couple two half-reactions occurring on different active sites, thus lowering the energy barriers. This agrees well with the theoretical simulations (Fig. 3).

The driving force for these processes often originates from the electron-hole excitation, which leads to the generation of so-called “hot” electrons largely via chemisorption⁵². The generation probability and the number of the hot electrons are closely associated with the adsorption energy, and a large adsorption energy is often favorable to generate the large number of the hot electrons, leading to the increasing probability of inter-site electronic communications. As shown in Fig. 3a and Supplementary Fig. 7d, the adsorption energies ( > 1 eV) for CO and O are large enough to generate the reaction-required “hot” electrons¹, thereby drving a smooth cycle by the inter-site electronic communications in CO oxidation.

Roles of metallic states in heterogeneous catalysis

As evidenced above, the metallic states of the particles are an essential prerequisite to realize the inter-site electronic communications with which to couple the two half-reactions. To explore the roles of the metallic states in heterogeneous catalysis, we scrutinize reported E_a values of CO oxidation over various supported metal particles with different sizes, as displayed in Fig. 5a (see also Supplementary Table 4 for more details). As particle sizes increases up to ~2 nm, an energy gap between the unoccupied states and occupied states becomes small, and ultimately a continuous energy level forms where the metallic states appear^6,18,19,20. Correspondingly, the E_a value decreases from ~100 kJ mol⁻¹ to a minimal value (~30 kJ mol⁻¹), and almost remains constant (~0.3 eV) as the size further increases.

To shed light on why the appearance of the metallic states can lower E_a in CO oxidation, we employ a combination of the frontier band orbital model⁵³ and the interfacial energy alignment model^31,32 to describe the energy level between molecular orbitals of adsorbed species and band orbitals of active sites. In Fig. 5b, according to the interfacial energy alignment model^31,32, an adsorbed molecule’s HOMO can become pinned to the E_F level of catalysts (which is also consistent with our DFT calculation). As for the nonmetallic particles with an energy gap between unoccupied states and occupied states, charge exchange between adsorbed species and catalysts requires surmounting energy barriers between HOMO and the lowest unoccupied band orbital (LUBO) or between the lowest unoccupied molecular orbital (LUMO) and the highest occupied band orbital (HOBO). As the metallic states form, HOBO and LUBO are merged on E_F, thus lowering the energy barriers required for the charge exchange^31,53.

The experimental evidence and the related discussion above clearly elucidate the roles of metallic states in lowering E_a in catalytic oxidation: (i) one is to allow HOBO and LUBO to merge on E_F so as to eliminate energy gaps present for nonmetallic catalysts⁵³, thus lowering energy barriers for charge exchanges between HOMO (or LUMO) of adsorbed species and LUBO (or HOBO) of active sites; (ii) the other is to give access to electronic communications for active sites on individual metallic particles as an electron transport, thus coupling two half-reactions occurring on two different active sites in a spatially synchronized way⁴, which lowers energy barriers for charge transfer between reactants. Hence, from the charge transfer view of point, the appearance of the metallic states can accelerate the reactant-catalyst charge exchanges and the catalyst-mediated reactant-reactant charge delivery, which makes CO oxidation follow preferentially a low-E_a electronic-communication-driven half-reactions coupling EDM¹¹.

Discussion

The electron excitation and transfer play a critical role in heterogeneous catalysis. Typical pioneering work includes the electronic excitations by surface reactions, or so-called “exoelectron emission”⁵⁴, and internal electron excitation within a catalysis system, or “chemicurrents”⁵⁵. In particular, Somorjai’s group directly detected chemicurrents in catalytic reactions and found that chemicurrents were well correlated with the turnover rates, reflecting the prevalence of the electron transfer across a wide range of heterogenous reactions⁵². An important work from Ertl’s group is that excited electrons triggered catalytic reactions of CO oxidation with an E_a much smaller than that by phonon excitation⁵⁶, implying that excited electrons by chemi- or physisorption of reactant’s molecule^2,3,4,38 on one active site can be transferred to another entangled active site in reactions upon eligible electron conducting. In this work, we evidenced the presence of the electronic communications between active sites on individual Rh and Pt metallic nanoparticles in catalytic reaction of CO oxidation, and such an electronic communication might be commonly present in heterogeneous catalysis^2,38, which holds for metallic materials such as Au, Pd, Ir, and so forth, catalyzing not only CO oxidation (Fig. 5a), but also for a broad spectrum of catalytic reactions, such as VOCs oxidation reaction^49,57, three-way catalytic reaction⁵⁰, and water-gas shift reaction⁴⁵. Meanwhile, metal nanoparticles are required to bear the metallic states in order to initiate effective electron communications in reactions, and thus the metallic particles with a size as small as possible to expose more metal atoms on surfaces as active sites are favorable for activity enhancement. For further design, after realizing the metallic electronic structures of the metal catalysts, the atomic utilization and green economy of the catalysts may be important which need to be considered in order to achieve the goals of green chemistry. That is, the metal with sizes between 2 ~ 3 nm may be the optimal option but of course there may be some differences between reactions, which should be further investigated in the following work. Hence, this work provides a general strategy to rationally design and effectively develop improved metallic nanomaterials in heterogeneous catalysis.

These results give a unified explanation to disparate and controversial results in activity of single-atom catalysts and metallic particle catalysts. Owing to a lack of inter-site electronic communications, electron transfer for single-atom catalysts with electronically insulating supports is fulfilled merely via single active metal atoms by changing their valence states, which often requires surmounting a considerably high energy barrier to realize the oxidation ↔ reduction state transitions, thus leading to a relatively large E_a. Owing to inter-site electronic communications, metallic particle catalysts often show much higher activity than this kind of single-atom catalysts under identical conditions^47,48,58. However, if suitable reducible supports selected can generate so strong metal-support interactions that only a low energy barrier is required to surmount for the electron transfer between single atoms and supports^23,59, E_a in reactions over this kind of single-atom catalysts could be as low as that over metallic particle catalysts (Fig. 5a), in which case single-atom catalysts could have higher catalytic activity than metallic particles due to more exposed metal atoms as active sites^60,61. Our results could also give a good interpretation for the so-called nanosized effects⁶². As the particle size increases, the appearance of metallic states enables the charge exchange energies to decrease down to a minimum, where a low-E_a-kinetic reaction coupling pathway is triggered via inter-site electronic communications, so catalytic activity increases up to a maximum. As the particle size further increases, catalytic activity often decreases owing to the decreasing number of surface metal atoms acting as active sites. Therefore, this work also has implications for an in-depth understanding of reaction mechanisms.

Methods

Preparation of Rh_met/Al₂O₃, Rh_non/Al₂O₃ and Rh₁/Al₂O₃

The commercial non-reducible γ-Al₂O₃ (average particle size: 10-20 nm, 99.99%, 3AMaterials, China) was calcined at 550 °C for 6 h and then used as the support. The metallic Rh_met/Al₂O₃ containing metallic Rh nanoparticles was synthesized according to the colloid deposition method²⁴. First, an ethylene glycol (EG) solution of NaOH (50 mL, 0.1 mol/L) was added into the EG solution of Rh(NO₃)₃ (50 mL, 5.8 mmol/L) with stirring at 140 °C for 3 h under the protection of Ar atmosphere to obtain Rh particle colloid solution. Then, 1 g γ-Al₂O₃ support was added into 16 mL Rh particle colloid solution under stirring at 80 °C for 3 h, followed by centrifugation with deionized water to remove EG, Na⁺ and NO₃⁻ and dried overnight. Finally, the obtained powder was calcined at 200 °C for 2 h in air. The nonmetallic Rh_non/Al₂O₃ was synthesized by the impregnation method. 1 g γ-Al₂O₃ support was dispersed in deionized water, and 1.91 mL of the (NH₄)₂RhCl₄ solution (2.62 g_Rh/L) was added dropwise. Then the mixed solution was evaporated to dryness at 80 °C. Finally, the powder was dried and calcined at 500 °C for 3 h in air. The theoretical loading of Rh is 0.5 wt.%. Rh₁/Al₂O₃ was prepared as the same process as Rh_non/Al₂O₃ except that the loading of Rh is 0.1 wt.%. The actual loadings of Rh were confirmed by inductively coupled plasma-atomic emission spectroscopy (ICP-AES).

Preparation of Pt_met/Al₂O₃, Pt_non/Al₂O₃ and Pt₁/Al₂O₃

Pt_met/Al₂O₃ was also synthesized by the colloid deposition method. The EG solution of NaOH (50 mL, 0.05 mol/L) was added into an EG solution of H₂PtCl₆ (50 mL, 3.7 mmol/L) with stirring at 140 °C for 3 h under the protection of Ar atmosphere to obtain Pt particle colloid solution. Then, 1 g γ-Al₂O₃ support was added into 13 mL of the prepared Pt particle colloid solution under stirring at 80 °C for 3 h, followed by centrifugation with deionized water to remove EG, Na⁺ and Cl⁻. Finally, the obtained powders were calcined at 200 °C for 2 h in air. Pt_non/Al₂O₃ and Pt₁/Al₂O₃ were synthesized by the impregnation method. 1 g γ-Al₂O₃ support was dispersed in deionized water, and 1.38 and 0.28 mL of the H₂PtCl₆ solution (3.6 g_Pt/L) was added dropwise to prepare Pt_non/Al₂O₃ and Pt₁/Al₂O₃ respectively. Then the mixed solution was evaporated to dryness at 80 °C. Finally, the powders were dried and calcined at 400 °C for 2 h in air. The actual loadings of Pt were measured by ICP-AES.

X-ray diffraction (XRD) patterns

XRD patterns were collected on a D8 advance X-ray diffractometer (Bruker/AXS, Germany) at 40 kV and 40 mA using monochromatized Cu Kα radiation (λ = 1.5405 Å).

Transmission electron microscopy (TEM) and aberration-corrected high-angle annular dark-field scanning transmission electron microscopy (AC-STEM)

TEM experiments were carried out with a JEOL JEM-2100F field-emission gun transmission electron microscope operated at an accelerating voltage of 200 kV and equipped with an ultra-high-resolution pole-piece that provides a point-resolution better than 0.19 nm. Fine powders of the materials were dispersed in ethanol, sonicated, and sprayed on a carbon coated copper grid, and then allowed to air-dry. AC-STEM experiments and energy dispersive X-ray spectroscopy (EDS) elemental mapping were carried out in an aberration-corrected scanning/transmission electron microscope (ThermoFisher Thermis Z) equipped with SuperEDX detector at an accelerating voltage of 300 kV. An aberration-corrected scanning/transmission electron microscope (Hitachi HF5000) was also used for microscopic analysis.

X-ray absorption spectroscopy (XAS)

XAS includes X-ray absorption near-edge structure (XANES) spectra and extended X-ray absorption fine structure (EXAFS) spectra was obtained at beamline BL14W1 of the Shanghai Synchrotron Radiation Facility (SSRF). The data at Rh K-edge and Pt L₃-edge were collected with a fixed exit monochromator using Si(311) and Si(111) crystals, respectively.

X-ray photoelectron spectra (XPS)

XPS were recorded on a Kratos Axis Ultra-DLD system with a charge neutralizer and a 150 W Al (Mono) X-ray gun (1486.6 eV) with a delay-line detector (DLD). Spectra were acquired at normal emission with a passing energy of 40 eV. Charging effects were corrected by adjusting binding energy of C 1 s to 284.8 eV.

Catalytic evaluation

All the samples were pretreated at 300 °C for 30 min in Ar before experiments. CO oxidation was carried out in a fixed-bed quartz reactor. In each activity test, 100 mg catalyst (40 ~ 60 mesh) was used. The total flow rate was 50 mL min⁻¹, including 1.0 vol% CO, 21.0 vol% O₂ and N₂ as balance gas. The CO and CO₂ concentrations were detected by GC7890B (Agilent, USA) with FID and TCD detectors. When determining the reaction TOF, 5 mg Rh_met/Al₂O₃ and 200 mg Rh_non/Al₂O₃ were used and Si₃N₄ (α-phase, 40 ~ 60 mesh, Aldrich) was blended as diluents to keep the gas hourly space velocity (GHSV) consistent.

Temperature-programmed desorption (TPD)

O₂-TPD or CO-TPD profiles were obtained by AutoChem II 2920 (Micromeritics, USA). O₂-TPD experiments were performed in a U-shaped quartz reactor with a TCD as detector. In each test,100 mg sample was pretreated with He gas (50 mL min⁻¹) at 250 °C for 1 h. After cooling down to room temperature, the catalyst was treated with O₂ gas (30 mL min⁻¹) or 5% CO/He (30 mL min⁻¹) for 1 h for O₂-TPD or CO-TPD, respectively. After the adsorption was completed, pure He gas (50 mL min⁻¹) was introduced to remove the weak physical adsorption O₂ or CO on the surface. Finally, the sample was then heated to 300 °C at a rate of 10 °C min⁻¹.

Diffuse-reflectance infrared Fourier-transform (DRIFT) spectra

DRIFT spectra were conducted by accumulating 64 scans at 4 cm⁻¹ resolution from 4000 to 1000 cm⁻¹ on a Nicolet iS50 FTIR spectrometer equipped with a Harrick Scientific DRIFT cell and a mercury-cadmium-telluride MCT/A detector.

For ex situ measurement, samples were pretreated at 300 °C for 1 h in a flow of 50 mL min⁻¹ He to remove adsorbed water and carbon dioxide and then cooled to 30 °C under He flow to obtain a background spectrum which should be deducted from the spectra of samples. Next, 25 mL min⁻¹ 1% CO/He mixed gas was introduced. After the CO saturation, 50 mL min⁻¹ He was used to remove the gas phase CO from cell. Finally, the spectra were obtained at 30 °C. The spectra at 150 ^oC in Fig. 2e and Supplementary Fig. 12 were collected in the O₂/He mixed gas after the temperature increased to 150 ^oC from 30 ^oC.

For temperature-programmed in situ DRIFT spectra, samples were pretreated at 300 °C for 1 h in a flow of 50 mL min⁻¹ He to remove adsorbed water and carbon dioxide and then cooled to 30 °C under He flow. Next, samples were exposed to a flow of 25 mL min⁻¹ mixed gas (1% CO + 20% O₂ in He) and heated to 300 °C at a temperature rate of 5 °C min⁻¹. The experiments were also performed with Al₂O₃ to subtract the signal of gaseous CO from the spectra.

DFT calculation

Density functional theory (DFT) calculations were performed using VASP 6.2.1 packages⁶³ with projected augmented wave (PAW) pseudo-potentials^64,65. The exchange-correlation energy was treated based on the generalized gradient approximation (GGA) by using Perdew–Burke–Ernzerhof (PBE) functional⁶⁶. The plane-wave cutoff energy was set to 400 eV. The DFT-D3(BJ) method of Grimme^67,68 was employed to describe long-range VDW interactions. The Monkhorst–Pack scheme with a k-point separation length of 0.05 Å⁻¹ was utilized for sampling the first Brillion zone⁶⁹. To correct the zero-point energy for the reaction barrier, the vibrational frequency calculations were performed via the finite-difference approach. All atoms were fully relaxed in the calculations. The Quasi-Newton l-BFGS method was used for geometry relaxation until the maximal force on each degree of freedom was less than 0.05 eV Å⁻¹. To derive the free energy reaction profiles, we followed the same approach as our previous work⁷⁰. The standard thermodynamic data⁷¹ were utilized to obtain the temperature and pressure contributions.

Data availability

Source data are provided as a Source Data file. All other data of this study are available from the corresponding author upon reasonable request.

References

Gergen, B., Nienhaus, H., Weinberg, W. H. & McFarland, E. W. Chemically induced electronic excitations at metal surfaces. Science 294, 2521–2523 (2001).
Article ADS CAS PubMed Google Scholar
Novo, C., Funston, A. M. & Mulvaney, P. Direct observation of chemical reactions on single gold nanocrystals using surface plasmon spectroscopy. Nat. Nanotechnol. 3, 598–602 (2008).
Article CAS PubMed Google Scholar
Adams, J. S., Kromer, M. L., Rodríguez-López, J. & Flaherty, D. W. Unifying concepts in electro- and thermocatalysis toward hydrogen peroxide production. J. Am. Chem. Soc. 143, 7940–7957 (2021).
Article CAS PubMed Google Scholar
Ryu, J. et al. Thermochemical aerobic oxidation catalysis in water can be analysed as two coupled electrochemical half-reactions. Nat. Catal. 4, 742–752 (2021).
Article CAS Google Scholar
Gates, B. C. Individual nanoparticles in action. Nat. Nanotechnol. 3, 583–584 (2008).
Article CAS PubMed Google Scholar
Valden, M., Lai, X. & Goodman, D. W. Onset of catalytic activity of gold clusters on titania with the appearance of nonmetallic properties. Science 281, 1647–1650 (1998).
Article ADS CAS PubMed Google Scholar
Fu, Q., Saltsburg, H. & Flytzani-Stephanopoulos, M. Active nonmetallic Au and Pt species on ceria-based water-gas shift catalysts. Science 301, 935–938 (2003).
Article ADS CAS PubMed Google Scholar
Xu, Z. et al. Size-dependent catalytic activity of supported metal clusters. Nature 372, 346–348 (1994).
Article ADS CAS Google Scholar
Kaden, W. E., Wu, T., Kunkel, W. A. & Anderson, S. L. Electronic structure controls reactivity of size-selected Pd clusters adsorbed on TiO₂ surfaces. Science 326, 826–829 (2009).
Article ADS CAS PubMed Google Scholar
Maurer, F. et al. Tracking the formation, fate and consequence for catalytic activity of Pt single sites on CeO₂. Nat. Catal. 3, 824–833 (2020).
Article CAS Google Scholar
Huang, Z. et al. Interplay between remote single-atom active sites triggers speedy catalytic oxidation. Chem. 8, 3008–3017 (2022).
Article CAS Google Scholar
Freund, H.-J., Meijer, G., Scheffler, M., Schlögl, R. & Wolf, M. CO oxidation as a prototypical reaction for heterogeneous processes. Angew. Chem. Int. Ed. 50, 10064–10094 (2011).
Article CAS Google Scholar
Sabatier, P. et al. La Catalyse En Chimie Organique; Librairie Polytechnique, Paris et Liége: Paris. (1920).
Hoffmann, R. A chemical and theoretical way to look at bonding on surfaces. Rev. Mod. Phys. 60, 601 (1988).
Article ADS CAS Google Scholar
Nolte, P. et al. Shape changes of supported Rh nanoparticles during oxidation and reduction cycles. Science 321, 1654–1658 (2008).
Article ADS CAS PubMed Google Scholar
Falsig, H. et al. Trends in the catalytic CO oxidation activity of nanoparticles. Angew. Chem. Int. Ed. 47, 4835–4839 (2008).
Article CAS Google Scholar
Kubo, R. Electronic properties of metallic fine particles. I. J. Phys. Soc. Jpn. 17, 975–986 (1962).
Article ADS CAS Google Scholar
Baetzold, R. C. Electronic properties of metal clusters: size effects. Inorg. Chem. 20, 118–123 (1981).
Article CAS Google Scholar
Li, L. et al. Investigation of catalytic finite-size-effects of platinum metal clusters. J. Phys. Chem. Lett. 4, 222–226 (2013).
Article CAS PubMed Google Scholar
Xu, C., Lai, X., Zajac, G. W. & Goodman, D. W. Scanning tunneling microscopy studies of the TiO₂(110) surface: structure and the nucleation growth of Pd. Phys. Rev. B 56, 13464–13482 (1997).
Article ADS CAS Google Scholar
Hu, S. & Li, W.-X. Sabatier principle of metal-support interaction for design of ultrastable metal nanocatalysts. Science 374, 1360–1365 (2021).
Article ADS CAS PubMed Google Scholar
Bruix, A. et al. A new type of strong metal–support interaction and the production of H₂ through the transformation of water on Pt/CeO₂(111) and Pt/CeO_x/TiO₂(110) catalysts. J. Am. Chem. Soc. 134, 8968–8974 (2012).
Article CAS PubMed Google Scholar
Hu, P. et al. Electronic metal–support interactions in single-atom catalysts. Angew. Chem. Int. Ed. 53, 3418–3421 (2014).
Article CAS Google Scholar
Wang, Y., Ren, J., Deng, K., Gui, L. & Tang, Y. Preparation of tractable platinum, rhodium, and ruthenium nanoclusters with small particle size in organic media. Chem. Mater. 12, 1622–1627 (2000).
Article Google Scholar
Smith, N. V., Wertheim, G. K., Hüfner, S. & Traum, M. M. Photoemission spectra and band structures of d-band metals. IV. X-ray photoemission spectra and densities of states in Rh, Pd, Ag, Ir, Pt, and Au. Phys. Rev. B 10, 3197–3206 (1974).
Article ADS CAS Google Scholar
Filatova, E. O. & Konashuk, A. S. Interpretation of the changing the band gap of Al₂O₃ depending on its crystalline form: connection with different local symmetries. J. Phys. Chem. C. 119, 20755–20761 (2015).
Article CAS Google Scholar
Fu, Q., Wagner, T., Olliges, S. & Carstanjen, H.-D. Metal−oxide interfacial reactions: encapsulation of Pd on TiO₂(110). J. Phys. Chem. B 109, 944–951 (2005).
Article CAS PubMed Google Scholar
Weast, R. C., Astle, M. J., Ed. CRC Handbook of Chemistry and Physics, 63rd ed. (CRC Press, 1982).
Hülsey, M. J. et al. In situ spectroscopy-guided engineering of rhodium single-atom catalysts for CO oxidation. Nat. Commun. 10, 1330 (2019).
Article ADS PubMed PubMed Central Google Scholar
Bunluesin, T., Putna, E. S. & Gorte, R. J. A comparison of CO oxidation on ceria-supported Pt, Pd, and Rh. Catal. Lett. 41, 1–5 (1996).
Article CAS Google Scholar
Hill, I. G., Rajagopal, A., Kahn, A. & Hu, Y. Molecular level alignment at organic semiconductor-metal interfaces. Appl. Phys. Lett. 73, 662–664 (1998).
Article ADS CAS Google Scholar
Greiner, M. T. et al. Universal energy-level alignment of molecules on metal oxides. Nat. Mater. 11, 76–81 (2012).
Article ADS CAS Google Scholar
Wovchko, E. A. & Yates, J. T. Activation of O₂ on a photochemically generated Rh^I site on an Al₂O₃ surface: low-temperature O₂ dissociation and CO oxidation. J. Am. Chem. Soc. 120, 10523–10527 (1998).
Article CAS Google Scholar
Cai, J. et al. Formation of different Rh–O species on Rh(110) and their reaction with CO. ACS Catal. 13, 11–18 (2023).
Article CAS Google Scholar
Wintterlin, J., Völkening, S., Janssens, T. V. W., Zambelli, T. & Ertl, G. Atomic and macroscopic reaction rates of a surface-catalyzed reaction. Science 278, 1931–1934 (1997).
Article ADS CAS PubMed Google Scholar
Stenlid, J. H. & Brinck, T. Extending the σ-hole concept to metals: an electrostatic interpretation of the effects of nanostructure in gold and platinum catalysis. J. Am. Chem. Soc. 139, 11012–11015 (2017).
Article CAS PubMed Google Scholar
Yoon, B. et al. Charging effects on bonding and catalyzed oxidation of CO on Au₈ clusters on MgO. Science 307, 403–407 (2005).
Article ADS CAS PubMed Google Scholar
Zou, N. et al. Cooperative communication within and between single nanocatalysts. Nat. Chem. 10, 607–614 (2018).
Article CAS PubMed Google Scholar
Suchorski, Y. et al. Resolving multifrequential oscillations and nanoscale interfacet communication in single-particle catalysis. Science 372, 1314–1318 (2021).
Article ADS CAS PubMed Google Scholar
Anderson, A. B. Molecular orbital study of the interaction of carbon monoxide and carbon dioxide with copper (100). Surf. Sci. 62, 119–132 (1977).
Article ADS CAS Google Scholar
Doyen, G. & Ertl, G. Theory of carbon monoxide chemisorption on transition metals. Surf. Sci. 43, 197–229 (1974).
Article ADS CAS Google Scholar
Hervier, A., Renzas, J. R., Park, J. Y. & Somorjai, G. A. Hydrogen oxidation-driven hot electron flow detected by catalytic nanodiodes. Nano Lett. 9, 3930–3933 (2009).
Article ADS CAS PubMed Google Scholar
Nørskov, J. K. et al. Origin of the overpotential for oxygen reduction at a fuel-cell cathode. J. Phys. Chem. B 108, 17886–17892 (2004).
Article Google Scholar
Bearden, J. A. & Burr, A. F. Reevaluation of X-ray atomic energy levels. Rev. Mod. Phys. 39, 125–142 (1967).
Article ADS CAS Google Scholar
Ding, K. et al. Identification of active sites in CO oxidation and water-gas shift over supported Pt catalysts. Science 350, 189–192 (2015).
Article ADS CAS PubMed Google Scholar
Niu, B. et al. Modulating the electronic states of Pt nanoparticles on reducible metal–organic frameworks for boosting the oxidation of volatile organic compounds. Environ. Sci. Technol. 58, 4428–4437 (2024).
CAS PubMed Google Scholar
Guo, L.-W. et al. Contributions of distinct gold species to catalytic reactivity for carbon monoxide oxidation. Nat. Commun. 7, 13481 (2016).
Article ADS CAS PubMed PubMed Central Google Scholar
Dessal, C. et al. Dynamics of single Pt atoms on alumina during CO oxidation monitored by operando X-ray and infrared spectroscopies. ACS Catal. 9, 5752–5759 (2019).
Article CAS Google Scholar
Goodman, E. D. et al. Catalyst deactivation via decomposition into single atoms and the role of metal loading. Nat. Catal. 2, 748–755 (2019).
Article CAS Google Scholar
Jeong, H. et al. Highly durable metal ensemble catalysts with full dispersion for automotive applications beyond single-atom catalysts. Nat. Catal. 3, 368–375 (2020).
Article CAS Google Scholar
Reeve, H. A. et al. Enzymes as modular catalysts for redox half-reactions in H₂-powered chemical synthesis: from biology to technology. Biochem. J. 474, 215–230 (2017).
Article CAS PubMed Google Scholar
Park, J. Y., Baker, L. R. & Somorjai, G. A. Role of hot electrons and metal–oxide interfaces in surface chemistry and catalytic reactions. Chem. Rev. 115, 2781–2817 (2015).
Article CAS PubMed Google Scholar
Xu, D. et al. An activity descriptor for perovskite oxides in catalysis. Chem. Catal. 2, 1163–1176 (2022).
Article CAS Google Scholar
Kramer, J. Spitzenzähler und zählrohr bei metallographischen oberflächenuntersuchungen. Z. Phys. 125, 739–756 (1949).
Article ADS CAS Google Scholar
Nienhaus, H. et al. Electron-hole pair creation at Ag and Cu surfaces by adsorption of atomic hydrogen and deuterium. Phys. Rev. Lett. 82, 446–449 (1999).
Article ADS CAS Google Scholar
Bonn, M. et al. Phonon- versus electron-mediated desorption and oxidation of CO on Ru(0001). Science 285, 1042–1045 (1999).
Article CAS PubMed Google Scholar
Liu, J. et al. Benzene abatement catalyzed by ceria-supported platinum nanoparticles and single atoms. Chem. Eng. J. 467, 143407 (2023).
Article ADS CAS Google Scholar
Wang, J., Tan, H., Yu, S. & Zhou, K. Morphological effects of gold clusters on the reactivity of ceria surface oxygen. ACS Catal. 5, 2873–2881 (2015).
Article CAS Google Scholar
Wang, L. et al. Supported rhodium catalysts for ammonia–borane hydrolysis: dependence of the catalytic activity on the highest occupied state of the single rhodium atoms. Angew. Chem. Int. Ed. 56, 4712–4718 (2017).
Article ADS CAS Google Scholar
Qiao, B. et al. Single-atom catalysis of CO oxidation using Pt₁/FeO_x. Nat. Chem. 3, 634–641 (2011).
Article CAS PubMed Google Scholar
DeRita, L. et al. Catalyst architecture for stable single atom dispersion enables site-specific spectroscopic and reactivity measurements of CO adsorbed to Pt atoms, oxidized Pt clusters, and metallic Pt clusters on TiO₂. J. Am. Chem. Soc. 139, 14150–14165 (2017).
Article CAS PubMed Google Scholar
Liu, L. & Corma, A. Metal catalysts for heterogeneous catalysis: from single atoms to nanoclusters and nanoparticles. Chem. Rev. 118, 4981–5079 (2018).
Article CAS PubMed PubMed Central Google Scholar
Kresse, G. & Furthmüller, J. Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set. Comput. Mater. Sci. 6, 15–50 (1996).
Article CAS Google Scholar
Kresse, G. & Joubert, D. From ultrasoft pseudopotentials to the projector augmented-wave method. Phys. Rev. B 59, 1758–1775 (1999).
Article ADS CAS Google Scholar
Blöchl, P. E. Projector augmented-wave method. Phys. Rev. B 50, 17953–17979 (1994).
Article ADS Google Scholar
Perdew, J. P., Burke, K. & Ernzerhof, M. Generalized gradient approximation made simple. Phys. Rev. Lett. 77, 3865–3868 (1996).
Article ADS CAS PubMed Google Scholar
Grimme, S., Antony, J., Ehrlich, S. & Krieg, H. A consistent and accurate ab initio parametrization of density functional dispersion correction (DFT-D) for the 94 elements H-Pu. J. Chem. Phys. 132, 154104 (2010).
Article ADS PubMed Google Scholar
Grimme, S., Ehrlich, S. & Goerigk, L. Effect of the damping function in dispersion corrected density functional theory. J. Comput. Chem. 32, 1456–1465 (2011).
Article CAS PubMed Google Scholar
Monkhorst, H. J. & Pack, J. D. Special points for brillouin-zone integrations. Phys. Rev. B 13, 5188–5192 (1976).
Article ADS MathSciNet Google Scholar
Wei, G.-F., Shang, C. & Liu, Z.-P. Confined platinum nanoparticle in carbon nanotube: structure and oxidation. Phys. Chem. Chem. Phys. 17, 2078–2087 (2015).
Article CAS PubMed Google Scholar
Lide, D. R. CRC Handbook Of Chemistry And Physics. (CRC press, 2004).

Download references

Acknowledgements

This work is supported by National Key R&D Program of China 2022YFB3504100, 2021YFB3500600 (to X.T.), National Natural Science Foundation of China 22276036, 22276037, 22173069, 22072090, 22272106 (to X.T., Z.M., G.W., X.L., and X.L.) and Shanghai Municipal Natural Science Foundation 21ZR1467800 (to G.W.). XAS at Rh K-edge and Pt L₃-edge were performed at beamline BL14W1 of the Shanghai Synchrotron Radiation Facility (SSRF).

Author information

Authors and Affiliations

Department of Environmental Science & Engineering, Fudan University, Shanghai, China
Dongrun Xu, Xue Fang, Weiye Qu, Chenxin Xu, Junxiao Chen, Zhen Ma, Xingfu Tang & Yaxin Chen
Shanghai Key Laboratory of Chemical Assessment and Sustainability, School of Chemical Science and Engineering, Tongji University, Shanghai, China
Yaowei Jin & Guangfeng Wei
School of Chemistry and Chemical, In-situ Center for Physical Sciences, Shanghai Jiao Tong University, Shanghai, China
Bowen He, Guokang Chen, Liwei Chen & Xi Liu
Shanghai Institute of Pollution Control and Ecological Security, Shanghai, China
Zhen Ma & Xingfu Tang
School of Chemistry and Chemical Engineering, Ningxia University, Yinchuan, China
Xi Liu

Authors

Dongrun Xu
View author publications
Search author on:PubMed Google Scholar
Yaowei Jin
View author publications
Search author on:PubMed Google Scholar
Bowen He
View author publications
Search author on:PubMed Google Scholar
Xue Fang
View author publications
Search author on:PubMed Google Scholar
Guokang Chen
View author publications
Search author on:PubMed Google Scholar
Weiye Qu
View author publications
Search author on:PubMed Google Scholar
Chenxin Xu
View author publications
Search author on:PubMed Google Scholar
Junxiao Chen
View author publications
Search author on:PubMed Google Scholar
Zhen Ma
View author publications
Search author on:PubMed Google Scholar
Liwei Chen
View author publications
Search author on:PubMed Google Scholar
Xingfu Tang
View author publications
Search author on:PubMed Google Scholar
Xi Liu
View author publications
Search author on:PubMed Google Scholar
Guangfeng Wei
View author publications
Search author on:PubMed Google Scholar
Yaxin Chen
View author publications
Search author on:PubMed Google Scholar

Contributions

X.T. directed the project; D.X., X.F., C.X., and J.C. synthesized the samples, designed and conducted the experiments; Y.J. and G.W. did the theoretical calculation work; B.H., G.C., L.C., and X.L. provided equipment for electron microscopy experiments; X.T., Y.C., D.X., and W.Q. analyzed the data and interpreted the results. D.X., X.T., Z.M., and Y.C. wrote the manuscript. All authors have given approval to the final version of the manuscript.

Corresponding authors

Correspondence to Xingfu Tang, Xi Liu, Guangfeng Wei or Yaxin Chen.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks Zhimin Ao and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Transparent Peer Review file

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Xu, D., Jin, Y., He, B. et al. Electronic communications between active sites on individual metallic nanoparticles in catalysis. Nat Commun 15, 8614 (2024). https://doi.org/10.1038/s41467-024-52997-w

Download citation

Received: 02 March 2024
Accepted: 27 September 2024
Published: 04 October 2024
DOI: https://doi.org/10.1038/s41467-024-52997-w