Regulating Cu Oxidation State for Electrocatalytic CO2 Conversion into CO with Near-Unity Selectivity via Oxygen Spillover.

Cu-based catalysts are the most intensively studied in the field of electrocatalytic CO2 reduction reaction (CO2RR), demonstrating the capacity to yield diverse C1 and C2+ products albeit with unsatisfactory selectivity. Manipulation of the oxidation state of Cu sites during CO2RR process proves advantageous in modulating the selectivity of productions, but poses a formidable challenge. Here, an oxygen spillover strategy is proposed to enhance the oxidation state of Cu during CO2RR by incorporating the oxygen donor Sb2O4. The Cu-Sb bimetallic oxide catalyst attains a remarkable CO2-to-CO selectivity approaching unity, in stark contrast to the diverse product distribution observed with bare CuO. The exceptional Faradaic efficiency of CO can be maintained across a wide range of potential windows of ≈700 mV in 1 m KOH, and remains independent of the Cu/Sb ratio (ranging from 0.1:1 to 10:1). Correlative calculations and experimental results reveal that oxygen spillover from Sb2O4 to Cu sites maintains the relatively high valence state of Cu during CO2RR, which diminishes the binding strength of *CO, thereby achieving heightened selectivity in CO production. These findings propose the role of oxygen spillover in CO2RR over Cu-based catalysts, and shed light on the rational design of highly selective CO2 reduction catalysts.

Mapping Hydrogen Positions along the Proton Transfer Pathway in an Organic Crystal by Computational X-ray Spectra.

Understanding proton transfer (PT) dynamics in condensed phases is crucial in chemistry. We computed a 2D map of N 1s X-ray photoelectron/absorption spectroscopy (XPS/XAS) for an organic donor-acceptor salt crystal against two varying N-H distances to track proton motions. Our results provide a continuous spectroscopic mapping of O-H···N↔O-··· H+-N processes via hydrogen bonds at both nitrogens, demonstrating the sensitivity of N 1s transient XPS/XAS to hydrogen positions and PT. By reducing the O-H length at N1 by only 0.2 Å, we achieved excellent theory-experiment agreement in both XPS and XAS. Our study highlights the challenge in refining proton positions in experimental crystal structures by periodic geometry optimizations and proposes an alternative scaled snapshot protocol as a more effective approach. This work provides valuable insights into X-ray spectra for correlated PT dynamics in complex crystals, benefiting future experimental studies.

Is Fe the Most Active Site for Fe/N-Doped Graphdiyne?

We performed a systematic study on the activity of pristine, Fe-doped, N-doped, and Fe/N-codoped graphdiyne (GDY) for oxygen reduction reactions (ORRs). We found that the pristine GDY has a high overpotential because of the weak binding of the intermediates. The sp-hybridized N-doped GDY enhances the binding of the intermediates at the adjacent sp-hybridized C site, which greatly enhances its ORR activities with a low overpotential of 0.45 V. On the other hand, on Fe-doped GDY, the binding of the intermediates at the Fe site and its neighboring C sites becomes too strong, while the C site at the second nearest acetylene chain becomes the most active site with an overpotential of 0.43 V. In the case of Fe and N codoping, Fe and the C sites near Fe and N still bind the intermediates too strongly, and the most active site is located at the C with an optimal distance. The binding energy of OH* is an activity descriptor for Fe- and/or N-doped GDY. Based on the machine learning analysis of ΔG(OH*), both the properties of the active center (electronic and geometric properties) and its environment, especially the latter, play important roles in determining its activity. The scaling relation analysis and volcano plot suggest that Fe and N doping enhance the binding of the intermediates to different extents, and the C atom, which is bonded neither to N nor to Fe atom, with an optimal binding strength, becomes the most active site.

What governs the electrocatalytic N2 reduction activity of sp-hybridized boron?

Electrocatalytic nitrogen reduction reactions (NRRs) have attracted intensive scientific attention, and boron in various forms has been shown promising for the activation of N2. In this work, we assessed the NRR activities of sp-hybridized-B (sp-B) doped into graphynes (GYs) using first-principles calculations. Eight inequivalent sp-B sites on five graphynes were considered. We found that boron doping greatly modifies the electronic structures at the active sites. Both the geometric effects and electronic effects play vital roles in the adsorption of the intermediates. Some intermediates prefer the sp-B site while others are bonded to both the sp-B and sp-C sites, which leads to two descriptors: the adsorption energy of the end-on N2 and the side-on N2. The former correlates well with the p-band center of sp-B while the latter correlates well with both the p-band center of sp-C and the formation energy of sp-B-doped GYs. The activity map demonstrates that the limiting potentials of the reactions are very small (-0.57 V to -0.05 V for the eight GYs). The free energy diagrams reveal that the distal path is normally the most favorable pathway, and the reaction may be limited by the adsorption of N2 when its binding free energy is higher than 0.26 eV. All eight B-doped GYs locate near the top of the activity volcano, suggesting that there are very promising candidates for the efficient NRR. This work provides a comprehensive understanding of the NRR activity of sp-B-doped GYs, and it should help guide the design of sp-B-doped catalysts.

First-principles screening of transition metal doped anatase TiO2(101) surfaces for the electrocatalytic nitrogen reduction.

The electrocatalytic nitrogen reduction reaction (eNRR) has been attracting intensive scientific attention as a potential alternative to the industrial Haber-Bosch process for ammonia production. Although many materials have been investigated, optimal catalysts for the reaction remain to be found. In this work, we performed the theoretical screening of 3d-5d transition metal doped anatase TiO2 for the eNRR. The most favorable doping site of each transition metal on the (101) surface was located. We found that the doping of transition metals promotes the formation of oxygen vacancies which are beneficial for the reaction. The scaling relations between the energies of the key intermediates were investigated. Using a machine learning algorithm (SVM), we identified two adsorption modes for the end-on adsorbed *HNN, which exhibited different scaling relations with *NH2. From a two-step process, we screened out several candidates, among which Au and Ta were proposed to be the most efficient dopants. Electronic structure analysis reveals that they can efficiently lower the energy of the intermediates. These results should be helpful for the design of more efficient TiO2-based catalysts for the eNRR.

CO2 Reduction Mechanism on the Cu2 O(110) Surface: A First-Principles Study.

Cu2 O is an attractive catalyst for the selective reduction of CO2 to methanol. However, the mechanism of the reaction and the role of the Cu species in different oxidation states are not well understood yet. In this work, by first-principles calculations, we investigate the mechanism of the reaction on the Cu2 O(110) surface, which is the most selective for methanol, in different degrees of reduction: ideal surface, slightly reduced surface (SRS), and partially reduced surface (PRS). The most favorable reaction pathways on the three surfaces were identified. We found that Cu(I) on the ideal surface is not capable of chemisorbing CO2 , but surface oxygen serves as the active site which selectively converts CO2 to CH3 OH with a limiting potential of -0.77 V. The Cu(0) on the SRS and PRS promotes the adsorption and reduction of CO2 , while the removal of the residue O* becomes potential/rate limiting with a more negative limiting potential than the ideal surface. The SRS is selective to methanol while the PRS becomes selective to methane. The result suggests that the key to high methanol selectivity is to avoid the reduction of Cu(I), which provides a new strategy for the design of more efficient catalysts for selective CO2 reduction to methanol.

Second Sphere Effects Promote Formic Acid Dehydrogenation by a Single-Atom Gold Catalyst Supported on Amino-Substituted Graphdiyne.

Regulating the second sphere of homogeneous molecular catalysts is a common and effective method to boost their catalytic activities, while the second sphere effects have rarely been investigated for heterogeneous single-atom catalysts primarily due to the synthetic challenge for installing functional groups in their second spheres. Benefiting from the well-defined and readily tailorable structure of graphdiyne (GDY), an Au single-atom catalyst on amino-substituted GDY is constructed, where the amino group is located in the second sphere of the Au center. The Au atoms on amino-decorated GDY displayed superior activity for formic acid dehydrogenation compared with those on unfunctionalized GDY. The experimental studies, particularly the proton inventory studies, and theoretical calculations revealed that the amino groups adjacent to an Au atom could serve as proton relays and thus facilitate the protonation of an intermediate Au-H to generate H2 . Our study paves the way to precisely constructing the functional second sphere on single-atom catalysts.

Unifying the Nitrogen Reduction Activity of Anatase and Rutile TiO2 Surfaces.

TiO2 is a model transition metal oxide that has been applied frequently in both photocatalytic and electrocatalytic nitrogen reduction reactions (NRR). However, the phase which is more NRR active still remains a puzzle. This work presents a theoretical study on the NRR activity of the (001), (100), (101), and (110) surfaces of both anatase and rutile TiO2 . We found that perfect surfaces are not active for NRR, while the oxygen vacancy can promote the reaction by providing excess electrons and low-coordinated Ti atoms that enhance the binding of the key intermediate (HNN*). The NRR activity of the eight facets can be unified into a single scaling line. The anatase TiO2 (101) and rutile TiO2 (101) surfaces were found to be the most and the second most active surfaces with a limiting potential of -0.91 V and -0.95 V respectively, suggesting that the TiO2 NRR activity is not very phase-sensitive. For photocatalytic NRR, the results suggest that the anatase TiO2 (101) surface is still the most active facet. We further found that the binding strength of key intermediates scale well with the formation energy of oxygen vacancy, which is determined by the oxygen coordination number and the degree of relaxation of the surface after the creation of oxygen vacancy. This work provides a comprehensive understanding of the activity of TiO2 surfaces. The results should be helpful for the design of more efficient TiO2 -based NRR catalysts.

Visualizing Large Facet-Dependent Electronic Tuning in Monolayer WSe2 on Au Surfaces.

Two-dimensional transition metal dichalcogenides (TMDs) have shown great importance in the development of novel ultrathin optoelectronic devices owing to their exceptional electronic and photonic properties. Effectively tuning their electronic band structures is not only desired in electronics applications but also can facilitate more novel properties. In this work, we demonstrate that large electronic tuning on a WSe2 monolayer can be realized by different facets of a Au-foil substrate, forming in-plane p-n junctions with remarkable built-in electric fields. This facet-dependent tuning effect is directly visualized by using scanning tunneling microscopy and differential conductance (dI/dV) spectroscopy. First-principles calculations reveal that the atomic arrangement of the Au facet effectively changes the interfacial coupling and charge transfer, leading to different magnitudes of charge doping in WSe2. Our study would be beneficial for future customized fabrication of TMD-junction devices via facet-specific construction on the substrate.

Understanding the role of Cl doping in the oxygen evolution reaction on cuprous oxide by DFT.

Designing highly active and earth-abundant oxygen evolution reaction (OER) electrocatalysts for electrochemical water splitting remains a challenge. Recently, Cl-doped Cu2O has emerged as a very promising non-noble-metal electrocatalyst candidate for the OER. However, the mechanism of the OER catalyzed by Cl-doped Cu2O has not been explored systematically. Herein, a comprehensive density functional theory (DFT) study is employed to study the role of Cl doping via comparing the OER on pure and Cl-doped Cu2O surfaces with/without Cu vacancies. Our results reveal that Cl doping increases the adsorption ability of Cu2O(111) by providing an excess electron, while a Cu vacancy decreases its adsorption ability by changing the geometric structure of the adsorption sites and the electronic structures. Cl-Cu2O(111) (η = 0.58 V) and VCu-Cl-Cu2O(111) (η = 0.46 V) have comparable or even better OER activity than those of widely used OER electrocatalysts such as the IrO2 catalyst (η = 0.56 V). It is facile to have a Cu vacancy when Cu2O(111) is doped with Cl because of a large strain introduced by Cl doping. Thus, VCu-Cl-Cu2O(111) should be the most feasible catalyst for the OER catalyzed by Cl-doped Cu2O material. The dual role of Cl doping is that it not only increases the OER activity but also helps to form a Cu vacancy. The results show that Cu2O(111) activity can be greatly enhanced via electronic and geometric structure modulation, which is helpful for the design of more efficient Cu2O-based catalysis.

Single-Site Heterogeneous Organometallic Ir Catalysts Embedded on Graphdiyne: Structural Manipulation Beyond the Carbon Support.

Accurate control over the coordination circumstances of single-atom catalysts (SACs) is decisive to their intrinsic activity. Here, two single-site heterogeneous organometallic catalysts (SHOCs), Cp*Ir-L/GDY (L = OH- and Cl- ; Cp* = pentamethylcyclopentadienyl), with the fine-tuned local coordination and electronic structure of Ir sites, are constructed by anchoring Cp*Ir complexes on graphdiyne (GDY) matrix via a one-pot procedure. The spectroscopic studies and theoretical calculations indicate that the Ir atoms in Cp*Ir-Cl/GDY and Cp*Ir-OH/GDY have a much higher oxidation state than Ir in the SAC Ir/GDY. As a proof-of-principle demonstration, the GDY-supported SHOCs are used for formic acid dehydrogenation, which display a fivefold enhancement of catalytic activity compared with SAC Ir/GDY. The kinetic isotope effect and in situ Fourier-transform infrared studies reveal that the rate-limiting step is the ß-hydride elimination process, and Cp* on the Ir site accelerates the ß-hydride elimination reaction. The GDY-supported SHOCs integrate the merits of both SACs and molecular catalysts, wherein the isolated Ir anchored on GDY echoes with SACs' behavior, and the Cp* ligand enables precise structural and electronic regulation like molecular catalysts. The scheme of SHOCs adds a degree of freedom in accurate regulation of the local structure, the electronic property, and therefore the catalytic performance of single-atom catalysts.

Photo-assisted reductive cleavage and catalytic hydrolysis-mediated persulfate activation by mixed redox-couple-involved CuFeS2 for efficient trichloroethylene oxidation in groundwater.

Persulfate (PS, S2O82-) activation through transition metal sulfides (TMS) has gained increasing attention since it can decompose a wide variety of refractory halogenated organic compounds in groundwater and wastewater. However, the processes of PS activation by TMS and particularly the formation of •OH radical under anoxic and acidic conditions (pH ∼2.8) remain elusive. Herein, by employing mixed redox-couple-involved chalcopyrite (CuFeS2) (150 mg/L) nanoparticles for PS (3.0 mM) activation, 96% of trichloroethylene was degraded within 120 min at pH 6.8 under visible light irradiation. The combination of experimental studies and theoretical calculations suggested that the Cu(I)/Fe(III) mixed redox-couple in CuFeS2 plays a crucial role to activate PS. Cu(I) acted as an electron donor to transfer electron to Fe(III), then Fe(III) served as an electron transfer bridge as well as a catalytic center to further donate this received electron to the O-O bond of PS, thus yielding SO4•- for trichloroethylene oxidation. Moreover, for the first time, •OH radicals were found to form from the catalytic hydrolysis of PS onto CuFeS2 surface, where S2O82- anion was hydrolyzed to yield H2O2 and these ensuing H2O2 were further transformed into •OH radicals via photoelectron-assisted O-O bond cleavage step. Our findings offer valuable insights for understanding the mechanisms of PS activation by redox-couple- involved TMS, which could promote the design of effective activators toward PS decomposition for environmental remediation.

A theoretical study of the role of the non-innocent phenolate ligand of a nickel complex in water oxidation.

Water oxidation is the bottleneck of artificial photosynthesis. A novel nickel phenolate complex with a redox-active ligand has been designed to manage multiple electron transfers during water oxidation (D. Wang and C. O. Bruner, Inorg. Chem., 2017, 56, 13638). However, the mechanism of the reaction is not well understood and verified from a theoretical aspect. Density functional theory calculations were conducted to investigate the mechanism of water oxidation catalyzed by the nickel(II)-phenolate complex. Because only two cyclic voltammogram (CV) peaks were observed and the phenolate ligand is redox-active, the active species was proposed to be NiIII-OH by the experiment. Based on the calculated results, the first CV peak is phenolate ligand-centered and the second peak is a single two-proton-coupled-two-electron process. In addition, the activation barrier of O-O bond formation of NiIII-OH is higher than that of NiIV-2OH by 15.3 kcal mol-1. Thus, the redox-active phenolate ligand does not lower the oxidation state of Ni in the active species to NiIII. The oxidation state of the active species is still NiIV, the same as other Ni complexes for WOCs. As the phenolate ligand and the hydroxyl ligand can act as an internal base, three pathways are compared for O-O bond formation: normal WNA, phenolate-involving single electron transfer (SET)-WNA, and OH-involving SET-WNA. The OH-involving SET-WNA pathway is the most favorable because the hydroxyl ligand is more nucleophilic than the oxygen radical of the phenolate ligand. Based on the experimental observation and theoretical results, the phenolate ligand is not stable and easily oxidized because of the hydrogen at the benzyl position. Thus, WOC candidates should not have the presence of hydrogen at the benzyl position near the active center.

Unlocking bimetallic active sites via a desalination strategy for photocatalytic reduction of atmospheric carbon dioxide.

Ultrathin two-dimensional (2D) metal oxyhalides exhibit outstanding photocatalytic properties with unique electronic and interfacial structures. Compared with monometallic oxyhalides, bimetallic oxyhalides are less explored. In this work, we have developed a novel top-down wet-chemistry desalination approach to remove the alkali-halide salt layer within the complicated precursor bulk structural matrix Pb0.6Bi1.4Cs0.6O2Cl2, and successfully fabricate a new 2D ultrathin bimetallic oxyhalide Pb0.6Bi1.4O2Cl1.4. The unlocked larger surface area, rich bimetallic active sites, and faster carrier dynamics within Pb0.6Bi1.4O2Cl1.4 layers significantly enhance the photocatalytic efficiency for atmospheric CO2 reduction. It outperforms the corresponding parental matrix phase and other state-of-the-art bismuth-based monometallic oxyhalides photocatalysts. This work reports a top-down desalination strategy to engineering ultrathin bimetallic 2D material for photocatalytic atmospheric CO2 reduction, which sheds light on further constructing other ultrathin 2D catalysts for environmental and energy applications from similar complicate structure matrixes.

Identification of M-NH2 -NH2 Intermediate and Rate Determining Step for Nitrogen Reduction with Bioinspired Sulfur-Bonded FeW Catalyst.

The multimetallic sulfur-framework catalytic site of biological nitrogenases allows the efficient conversion of dinitrogen (N2 ) to ammonia (NH3 ) under ambient conditions. Inspired by biological nitrogenases, a bimetallic sulfide material (FeWSx @FeWO4 ) was synthesized as a highly efficient N2 reduction (NRR) catalyst by sulfur substitution of the surface of FeWO4 nanoparticles. Thus prepared FeWSx @FeWO4 catalysts exhibit a relatively high NH3 production rate of 30.2 ug h-1 mg-1cat and a Faraday efficiency of 16.4 % at -0.45 V versus a reversible hydrogen electrode in a flow cell; these results have been confirmed via purified 15 N2 -isotopic labeling experiments. In situ Raman spectra and hydrazine reduction kinetics analysis revealed that the reduction of undissociated hydrazine intermediates (M-NH2 -NH2 ) on the surface of the bimetallic sulfide catalyst is the rate-determing step for the NRR process. Therefore, this work can provide guidance for elucidating the structure-activity relationship of NRR catalysts.

Size-Dependent Activity and Selectivity of Atomic-Level Copper Nanoclusters during CO/CO2 Electroreduction.

As a favorite descriptor, the size effect of Cu-based catalysts has been regularly utilized for activity and selectivity regulation toward CO2 /CO electroreduction reactions (CO2 /CORR). However, little progress has been made in regulating the size of Cu nanoclusters at the atomic level. Herein, the size-gradient Cu catalysts from single atoms (SAs) to subnanometric clusters (SCs, 0.5-1 nm) to nanoclusters (NCs, 1-1.5 nm) on graphdiyne matrix are readily prepared via an acetylenic-bond-directed site-trapping approach. Electrocatalytic measurements show a significant size effect in both the activity and selectivity toward CO2 /CORR. Increasing the size of Cu nanoclusters will improve catalytic activity and selectivity toward C2+ productions in CORR. A high C2+ conversion rate of 312 mA cm-2 with the Faradaic efficiency of 91.2 % are achieved at -1.0 V versus reversible hydrogen electrode (RHE) over Cu NCs. The activity/selectivity-size relations provide a clear understanding of mechanisms in the CO2 /CORR at the atomic level.

Regulating kinetics and thermodynamics of electrochemical nitrogen reduction with metal single-atom catalysts in a pressurized electrolyser.

Using renewable electricity to synthesize ammonia from nitrogen paves a sustainable route to making value-added chemicals but yet requires further advances in electrocatalyst development and device integration. By engineering both electrocatalyst and electrolyzer to simultaneously regulate chemical kinetics and thermodynamic driving forces of the electrocatalytic nitrogen reduction reaction (ENRR), we report herein stereoconfinement-induced densely populated metal single atoms (Rh, Ru, Co) on graphdiyne (GDY) matrix (formulated as M SA/GDY) and realized a boosted ENRR activity in a pressurized reaction system. Remarkably, under the pressurized environment, the hydrogen evolution reaction of M SA/GDY was effectively suppressed and the desired ENRR activity was strongly amplificated. As a result, the pressurized ENRR activity of Rh SA/GDY at 55 atm exhibited a record-high NH3 formation rate of 74.15 µg h-1⋅cm-2, a Faraday efficiency of 20.36%, and a NH3 partial current of 0.35 mA cm-2 at -0.20 V versus reversible hydrogen electrode, which, respectively, displayed 7.3-, 4.9-, and 9.2-fold enhancements compared with those obtained under ambient conditions. Furthermore, a time-independent ammonia yield rate using purified 15N2 confirmed the concrete ammonia electroproduction. Theoretical calculations reveal that the driving force for the formation of end-on N2* on Rh SA/GDY increased by 9.62 kJ/mol under the pressurized conditions, facilitating the ENRR process. We envisage that the cooperative regulations of catalysts and electrochemical devices open up the possibilities for industrially viable electrochemical ammonia production.

First-principles study on the selective hydrogenation of the C[double bond, length as m-dash]O and C[double bond, length as m-dash]C bonds of acrolein on Pt-M-Pt (M = Pt, Cu, Ni, Co) surfaces.

Selective hydrogenation of the C[double bond, length as m-dash]O and C[double bond, length as m-dash]C bonds of acrolein on Pt-M-Pt (M = Pt, Cu, Ni, Co) surfaces has been investigated with first-principles calculations to understand the trends of the activity and selectivity of the reaction. On the pristine Pt(111) surface, the results suggest that the production of allyl alcohol (a product of C[double bond, length as m-dash]O bond hydrogenation) is limited by its desorption, which results in the selective hydrogenation of the C[double bond, length as m-dash]C bond. On the other three bimetallic surfaces, the results show that the desorption of the product is no longer rate-limiting, and the reaction should be selective for the C[double bond, length as m-dash]O bond hydrogenation. Although the calculated trends of activity and selectivity agree well with the experiment, the absolute selectivity predicted on the bimetallic surfaces is in contrast with existing experiments. Therefore, other effects such as the steric effect and reactions at other types of active sites may need to be investigated. On the other hand, the scaling relation analysis shows that the formation free energies of the intermediates, except for H, scale well with that of the adsorbed acrolein. This suggests that modifying the binding of H on the surface may be another dimension for the design of more efficient catalysts for the active and selective hydrogenation of the C[double bond, length as m-dash]O bond of acrolein.

Mixed Redox-Couple-Involved Chalcopyrite Phase CuFeS2 Quantum Dots for Highly Efficient Cr(VI) Removal.

Iron-based nanosized ecomaterials for efficient Cr(VI) removal are of great interest to environmental chemists. Herein, inspired by the "mixed redox-couple" cations involved in the crystal structure and the quantum confinement effects resulting from the particle size, a novel type of iron-based ecomaterial, semiconducting chalcopyrite quantum dots (QDs), was developed and used for Cr(VI) removal. A high removal capacity up to 720 mg/g was achieved under optimal pH conditions, which is superior to those of the state-of-the-art nanomaterials for Cr(VI) removal. The mechanism of Cr(VI) removal was elucidated down to an atomic scale by combining comprehensive characterization techniques with adsorption kinetic experiments and DFT calculations. The experimental results revealed that the material was a good electron donor semiconductor attributed to the existence of "mixed redox couple of Cu(I)-S-Fe(III)" in the crystal structure. With the size-dependent quantum confinement effect and the high surface area, the semiconducting chalcopyrite QDs could effectively remove Cr(VI) from aqueous solution through a syngenetic photocatalytic reduction and adsorption mechanism. This study not only reports the design histogram of the iron-based CuFeS2 QD ecomaterial for efficient Cr(VI) removal but also paves the way for understanding the atomic-scale mechanism behind the syngenetic effects of using the QD semiconducting material for Cr(VI) removal.

First-Principles Study on the Molecular Mechanism of Solar-Driven CO2 Reduction on H-Terminated Si.

Solar-driven conversion of CO2 with H-terminated silicon has recently attracted increasing interest. However, the molecular mechanism of the reaction is still not well understood. A systematic study of the mechanism has been carried out with first-principles calculations. The formation energies of the intermediates are found to be insensitive to the structure of the surface. On the fully H-terminated Si(111) surface, several pathways for the conversion of CO2 into CO at a coordinatively saturated Si site are studied, including the conventional COOH* pathway and the direct insertion of CO2 into Si-H and Si-Si bonds. Although the barrier of the COOH* pathway is lowest among the three pathways, it is higher than that for OH* elimination, which suggests that CO2 should be converted by other types of active site. The reaction at the isolated and dual coordinatively unsaturated (CUS) Si sites, which can be generated by light illumination, heat, and Pd loading, are then studied. The results suggest that the most efficient pathway to convert CO2 is to convert it into CO and O* at an isolated CUS Si site before O* reacts with a terminating H* to form adsorbed OH* and generate new isolated CUS Si sites. Therefore, the CUS Si site catalyzes the reaction until all H* is converted into OH*. The results provide new insight into the mechanism of the reaction and should be helpful for the design of more efficient Si-based catalysts for CO2 conversion.