The excessive emission of carbon dioxide has caused a series of environmental and ecological problems, such as greenhouse effect and ocean acidification. How to reduce carbon dioxide emissions has become a global concern, and China has put forward the strategic goals of achieving carbon peak and carbon neutrality by 2030 and 2060 respectively.

According to the thermodynamic principle and the research progress in related fields, it is a clean and highly feasible way to utilize CO₂ resources by using the coupling technology of wind, photovoltaic power generation and electrolytic water to obtain H₂, and then synthesize high value-added chemicals or liquid fuels through catalytic hydrogenation of CO₂.It is expected to reduce the emission of carbon dioxide while making it an important carbon resource, and provide an ideal way to solve the problems of environmental pollution and energy shortage^{[1⇓⇓⇓⇓⇓~7]}.

Hydrogenation of CO₂ can produce a series of carbon-containing compounds, such as carbon monoxide, methanol, methane and other C₁ molecules, as well as multi-carbon hydrocarbons, multi-carbon alcohols and other C₂₊ compounds, among which multi-carbon hydrocarbons or multi-carbon alcohols are important chemical raw materials or liquid fuels with high added value^{[8⇓⇓⇓~12]}. For example, the hydrocarbon compounds of C₅₊ can be used as liquid fuels such as gasoline or diesel, while the alcohol compounds of C₅₊ are important raw materials for the production of fine chemicals such as detergents, fragrances, plasticizers or lubricants. At present, the production of multi-hydrocarbon or alcohol mainly uses fossil resources as raw materials, and its production and use are often accompanied by a large number of CO₂ emissions^[13⇓~15].

The important chemical processes involved in the hydrogenation of CO₂ to carbon compounds include the C-C coupling of intermediates such as CO₂ hydrogenation to CH_x,CH_x, hydrogenolysis, dehydrogenation or hydrolysis of alkyl groups, etc. The reaction mechanism or apparent activation energy of the above chemical conversion step is quite different on catalysts with different structures, which is the chemical basis for optimizing the catalyst performance and improving the reaction rate and selectivity of the target product. In general, the carbon-oxygen bond in the CO₂ or intermediate CO molecule has a high chemical inertness, and the C-C coupling reaction has a high energy barrier, so the hydrogenation of CO₂ to multi-carbon compounds on many catalysts needs to be carried out at a high temperature (300 ~ 400 ℃)^[16⇓~18].

In recent years, some complexes of metal oxides and molecular sieves have been found to effectively catalyze the hydrogenation of CO₂ to multi-carbon compounds at high temperatures^[16⇓~18]. For example, the bifunctional catalyst formed by the solid solution of ZnO-ZrO₂ and molecular sieve SAPO-34 prepared by Li can et al. Can effectively catalyze the reaction of CO₂ and H₂ to synthesize light olefins under the reaction conditions of 380 ° C and 2 MPa, and the conversion rate of hydrocarbon products with light olefins (C₂~C₄) as high as 80%,CO₂ is 12.6%^[16]. The stability test results showed that the activity of the catalyst did not decrease significantly within 100 H, and the catalyst had excellent resistance to sulfur poisoning. The iron oxide-molecular sieve composite catalyst Na-Fe₃O₄/HZSM-5 reported by Sun Jian et al. Can catalyze the hydrogenation of CO₂ to hydrocarbons at 320 ° C, the conversion rate of CO₂ is 22%, and the selectivity of the proportion of C₅~C₁₁ hydrocarbon compounds in the product to 78%,CH₄ is only 4%^[17]. The results of stability test showed that the activity of the catalyst decreased by only 6% within 1000 H, and the proportion of C₅₊ compounds in the product was maintained at 67% ± 2%, indicating that the catalyst had good stability under reaction conditions. The composite catalyst In₂O₃/HZSM-5 prepared by Zhong Liangshu et al. Can catalyze the hydrogenation of CO₂ to produce multi-carbon hydrocarbons. At 340 ℃, the conversion rate of CO₂ is 13.1%, and the ratio of multi-carbon hydrocarbons in gasoline components in the product can reach 78.6%,CH₄ with a selectivity of only 1%^[18]. The stability test results showed that the catalyst maintained good stability within 150 H under the reaction conditions of 613 K and 3 MPa, the conversion of CO₂ was stable at about 12%, and the selectivity of C₅₊ hydrocarbon was maintained at about 80%.

Some metal-metal oxide composite catalysts exhibit excellent catalytic performance for the hydrogenation of CO₂ to multi-carbon compounds^[19⇓~21]. Han et al. Prepared a series of metal-metal oxide composite catalysts with good catalytic performance for hydrogenation of CO₂ to multi-carbon compounds by coprecipitation and hydrogen reduction at high temperature (400 ~ 250 ℃)^[19]. The complex-phase metal catalyst prepared by the method and composed of Fe₃O₄ and zero-valent Pt, Ru and Fe can catalyze CO₂ hydrogenation to generate a multi-carbon compound in 1,3-dimethyl-2-imidazolone (DMI) solvent under 160~220℃,20 MPa(H₂/CO₂=2∶3),At 160 ° C, the selectivities of CH₃OH, C₂₊ alcohol, C₁~C₄ hydrocarbon, C₅₊ hydrocarbon and CO in the product were 23.6%, 12.6%, 54.1%, 9.1% and 0.6%, respectively^[19]. The Co₆/MnO_x composite catalyst prepared by them, which is mainly composed of Co⁰, Co²⁺, Mn²⁺ and oxygen, can catalyze the hydrogenation of CO₂ to produce multi-carbon compounds in squalane solvent at 200 C.The selectivity of C₅₊ compounds (C₅~C₂₆ hydrocarbons) in the product reached 53.2%. After 15 H of reaction, the conversion of CO₂ reached 15.3%^[20]. The composite catalyst prepared by them with Pt⁰, Co⁰(hcp, and fcc), Co₃O₄, and CoO as the main components was tested at 140 ° C with H₂O and DMI (1.3-dimethyl-2-imidazolidinone) mixed solvent can catalyze the hydrogenation of CO₂ to generate C₁~C₄ alcohol and C₁~C₄ hydrocarbon, wherein the proportion of the C₂~C₄ multi-carbon alcohol in the alcohol product can reach 82.5%^[21].

Xiao Fengshou et al. Reported a Al₂O₃ supported metal Co-CoO nanocomposite catalyst, in which the average size of Co nanoparticles was 4.6 nm. The catalyst can catalyze the hydrogenation of CO₂ to produce C₁~C₄ alcohol at 140 deg C, the main product is ethanol, the selectivity of the catalyst can reach 92.1%, and the activity of the catalyst for catalyzing the production of ethanol is 0.44 mmol·g^-1·h^-1[22].

Our group has been engaged in the controllable synthesis and performance research of metal nanocluster catalysts for many years^{[23⇓⇓⇓⇓⇓⇓⇓~31]}. Theoretically, lowering the reaction temperature of CO₂ catalytic conversion is beneficial to reduce the CO₂ emission and improve the selectivity of the target product. In recent years, we have made some progress in the development of catalysts for the synthesis of multicarbon metal nanoclusters by CO₂ hydrogenation at low temperature^{[28⇓⇓~31]}. We found that CO₂ hydrogenation to hydrocarbons (C₁~C₂₆) and alcohols (C₁~C₁₀) can be achieved at lower temperatures (40 – 130 ° C) by changing the reaction pathway and employing new nanoscale metal cluster composite catalysts. In this review, we will highlight some of the advances in this field, and discuss the chemical basis for the regulation of product selectivity between C₁ and C₂₊ in the hydrogenation of CO₂ at low temperatures.The progress in the synthesis and structure-activity relationship of Pt-Ru bimetallic nanocluster catalysts for the synthesis of the above multi-carbon compounds by catalytic hydrogenation of CO₂ at low temperature was introduced, and the proposed theory of local charge distribution effect of metal nanocluster catalysts was further elaborated.

Recently, we found that Pt nanoclusters and Ru nanoclusters (Ru-Pt/FeCO₃) supported on the surface of ferrous carbonate can catalyze the hydrogenation of CO₂ to hydrocarbons and alcohols at 40 – 130 ° C^[28]. The electron micrograph of the Ru-Pt/FeCO₃ is shown in Fig. 1. Ru and Pt nanoparticles in the catalyst are well dispersed on the FeCO₃ support, and some Ru and Pt nanoclusters contact to form a Ru-Pt interface^[28]. The mixture of cyclohexane and water was used as a dispersion medium to catalyze the reaction of CO₂ with hydrogen at an initial pressure of 5 MPa (the partial pressure ratio of H₂ to CO₂ was 3),Ru-Pt/FeCO₃.Hydrocarbons (C₁~C₂₆, subscript is the number of carbon atoms in the carbon chain) and alcohols (C₁~C₆) are generated.

Among the products generated from the hydrogenation of CO₂ catalyzed by Ru-Pt/FeCO₃ at 40 ° C, the selectivity of the C₅~C₂₆ and C₂₊ products was 49.6% and 77.7%, respectively, and the selectivity of the methane by-product was only 7.3%^[28]. With the increase of reaction temperature, the selectivity of C₂₊ products decreased, while the selectivity of methane increased. At 130 ° C, the selectivity of C₅~C₂₆ and C₂₊ products in the reaction products was 26.1% and 45.8%, respectively (Fig. 2). The kinetic test results showed that the apparent activation energy of CO₂ hydrogenation to methane and multi-carbon compounds on this catalyst was 79.8 and 50.7 KJ·mol^-1, respectively, implying that the low temperature reaction is beneficial to improve the selectivity of multi-carbon compounds in the products on this catalyst. No CO was detected in the above CO₂ hydrogenation reaction products (CO detection limit of 60 ppm), which is different from many catalytic CO₂ hydrogenation reactions at high temperature reported in the past. The control experiment results showed that the FeCO₃ supported Ru nanocluster catalyst (Ru/FeCO₃) had high activity for CO₂ hydrogenation at 130 ℃, the selectivity of CH₄ in the product was as high as 91.6%, and the content of multi-carbon hydrocarbon (C₂~C₁₀) was very low. However, the catalytic activity of FeCO₃ supported Pt nanoclusters (Pt/FeCO₃) for CO₂ hydrogenation is very low.The selectivity of CH₄ and C₂₊ hydrocarbon (C₂~C₁₄) in the product was 26.1% and 49.6%, respectively.

Isotope tracer method was used to study the reaction pathway of hydrogenation of CO₂ to multi-carbon compounds catalyzed by the above catalysts^[28]. The results show that on this catalyst, the hydrogenation of CO₂ to multi-carbon compounds is achieved by coupling the newly discovered process of hydrogenation of carbonate in ferrous carbonate to multi-carbon compounds with the carbonation reaction of ferrous species formed on the surface of the support (Fig. 3). We believe that the polarization effect of iron ion in ferrous carbonate on carbonate promotes the activation of C — O bond, which is an important factor for the high activity of such catalysts for the catalytic hydrogenation of CO₂ at low temperature. The above mechanism or reaction pathway is different from the CO₂- Fischer-Tropsch or high temperature CO₂ catalytic hydrogenation conversion pathway based on methanol intermediate of many catalysts^[17,18]. Further studies showed that the catalytic performance and mechanism of Pt nanoclusters and Ru nanoclusters (Ru-Pt/CoCO₃) supported on the surface of cobalt carbonate for the hydrogenation of CO₂ to multi-carbon compounds were similar to those of Ru-Pt/FeCO₃, which suggested that the reaction mechanism had a certain universality^[29].

Compared with multi-carbon alkanes, multi-carbon primary alcohols have higher added value. In industrial production, alkene hydrolysis, alkene hydroformylation or alkane oxidation are generally used to prepare higher alcohols, and the preparation of higher primary alcohols by selective catalytic oxidation of alkanes is still an unsolved problem in the field of catalysis^{[32⇓⇓⇓⇓⇓~38]}. Catalytic hydrogenation of CO₂ to higher primary alcohols is a promising process. It is generally believed that the reaction path of CO₂ hydrogenation to higher alcohols is the insertion of intermediate species ^*CO into the alkyl-metal bond on the surface of the catalyst, and then the hydrogenation reaction to higher alcohols. The synthesis of higher alcohols based on this reaction path usually needs to be carried out at high temperature^[39].

In 2018, we reported an iron oxide-supported Ru, Pt bimetallic nanocluster catalyst, which can catalyze CO₂ hydrogenation (starting pressure 3.0 MPa,H₂/CO₂=3∶1) in water at 40 – 130 ° C to produce primary alcohols with carbon numbers ranging from 1 to 8^[30]. We used the ¹⁸O isotope tracer method to study the reaction path to the multi-carbon alcohol in this catalytic process. When Ru-Pt/Fe₃O₄ was used to catalyze the CO₂ hydrogenation reaction in ¹⁸O labeled water, the isotopic ratios of ¹⁸O/¹⁶O in the resulting propanol and butanol were 1.4 and 1.6, respectively (Fig. 4),The ratio of ¹⁸O/¹⁶O in the carbon dioxide gas collected after the reaction is 0.9. If the formation path of the polyalcohol includes the step of CO insertion into the alkyl-metal bond, the isotopic ratio of ¹⁸O/¹⁶O in the produced polyalcohol should not be higher than 0.9, indicating that part of the O in the polyalcohol comes from water. On the other hand, it was found that propylene and CO₂),Ru-Pt/Fe₃O₄ could not catalyze the hydrolysis of olefins to alcohols under similar reaction conditions (130 ℃). The above experimental results show that in this reaction system, the multi-carbon alcohol can be produced through the alkyl-metal bond hydrolysis reaction on the surface of the catalyst, as shown in Figure 4 C^[30]. This new catalytic reaction pathway is different from the previously reported pathway of CO insertion into the alkyl – metal bond at high temperature, enabling the conversion of CO₂ to multi-carbon alcohols at low temperature. The ratio of ¹⁸O/¹⁶O in the water collected after the reaction was 2.1, which was higher than the ratio of ¹⁸O/¹⁶O in the polyalcohol, indicating that part of the O in the polyalcohol may originate from iron oxide or carbon dioxide. The results of subsequent X-ray diffraction experiments showed that part of the Fe₃O₄ in the catalyst was converted to FeCO₃ under the above catalytic reaction conditions, indicating that the carbon chain formation process in the above multi-carbon alcohol formation process was also in accordance with the carbonate hydrogenation-surface iron species carbonation coupling mechanism.

As mentioned above, the catalytic hydrogenation of carbonate in some carbonates can be carried out under mild conditions, and coupling this process with the carbonation reaction of metal ions can catalyze the conversion of CO₂ to CH_x species (^*CH_x) adsorbed on the surface at lower temperatures. Since the process does not pass through CO intermediates and the energy barrier for the conversion between ^*CH₃ and ^*CH₂ is not high, the product selectivity of CO₂ hydrogenation at low temperature mainly depends on the kinetic competition among the following chemical reactions. Carbon – carbon coupling reactions between species such as ^*CH₃+^*H and ^*CH_x are an important set of competing reactions during the hydrogenation of CO₂ at low temperature, and their rate ratios determine the selectivity of the CH₄ and C₂₊ products in the products. In addition, the competition of hydrogenolysis and hydrolysis determines the proportion of multi-carbon hydrocarbons and multi-carbon alcohols in the multi-carbon products after the formation of long carbon chain alkyl by carbon-carbon coupling reaction. By Chan that relative rates of the above major compete reactions, the ratio of the C₁/C₂₊ product and the ratio of alcohol/hydrocarbon in the C₂₊ product can be adjusted (fig. 5).

The rates of these reactions depend not only on the reaction barriers, but also on the adsorption energies of the reactants and products on the catalytic active sites, the steric hindrance of mass transfer, and the reaction conditions.

As mentioned above, the experimental results show that the presence of both Pt and Ru in the above catalysts is beneficial to improve the selectivity of multi-carbon compounds and the carbon chain length of multi-carbon compounds, which enlightens us to improve the selectivity of multi-carbons by adjusting the distribution of Pt and Ru elements in the catalysts. We found that carbon supported Ru and Pt nanoclusters were solvothermally treated in a mixture of CO₂ and H₂ to convert Ru nanoclusters into monometallic complexes, and then small Ru clusters were formed on Pt nanoclusters to prepare Pt nanocluster surface-bonded small Ru cluster catalyst (Ru-co-Pt/C)^[31]. Fig. 6 is a high-resolution electron microscope photograph of the synthesized catalyst and a distribution map of Pt and Ru elements. The extended X-ray absorption fine structure (EXAFS) spectrum (Fig. 7) analysis results show that after the above treatment, the Pt-Pt coordination number in the catalyst increases from 4.4 to 6.6, corresponding to the increase in the size of Pt nanoparticles, while the Ru-Ru coordination number decreases from 4.8 to 1.8, and the Ru-Pt coordination number changes from 0.38 to 5.86, corresponding to the phenomenon that Ru is dispersed on the surface of Pt nanoclusters. X-ray photoelectron spectroscopy (XPS) results showed that both Pt and Ru in the catalyst were in the zero valence state (fig. 7). The catalyst showed excellent catalytic performance in the hydrogenation of CO₂ to multi-carbon compounds. Hydrogenation of CO₂ at 130 ° C produced hydrocarbons of C₁~C₂₂ and alcohols of C₁~C₁₀, with a high selectivity of 90.1% for C₂₊ compounds (Fig. 8 a),It is the highest selectivity of C₂₊ compounds in the synthesis of long carbon chain compounds by hydrogenation of CO₂ reported so far, and no CO is detected in the product (chromatographic detection limit is 60 ppm)^[31]. The hydrocarbon products basically followed the Anderson-Schulz-Flory (ASF) distribution with a chain growth factor (α) of 0.7 (Fig. 8 f), indicating the excellent catalytic performance of Ru-co-Pt/C for CH_x coupling. The results of the catalyst stability test at 130 ° C showed that the yield of the CO₂ hydrogenation product stabilized at 9.7%±1%,C₂₊ and the product selectivity decreased slightly from 90.1% in the first cycle to 84.8% in the third cycle (Fig. 8 C).

The results of the control experiment showed that the carbon-supported Ru metal catalyst (Ru-co-Ru/C) prepared by the same method catalyzed the hydrogenation of CO₂ to C₁~C₁₂ hydrocarbons and C₁~C₂ alcohols, and the product was mainly methane, with a selectivity of 94.3% (Fig. 8 A). After 22 H of reaction, the yield of the hydrogenation product can reach 26. 5%, and the maximum number of carbon atoms in the product molecule can reach 12. The carbon-supported Pt metal nanocluster catalyst (Pt-co-Pt/C) prepared by the same method had low catalytic activity for the hydrogenation of CO₂, and the yield of hydrogenation products was only 0.2% after 22 H of reaction, and the products were hydrocarbons with C₁~C₅ and alcohols with C₁~C₂ (Fig. 8 A). The chain growth factors for the CO₂ hydrogenation reaction were 0.45 and 0.20 on Ru-co-Ru/C and Pt-co-Pt/C, respectively (Fig. 8 d, e).

Carbon-carbon coupling reaction is a key step in the construction of long carbon chain compounds. Reducing the energy barrier of carbon-carbon coupling reaction and increasing the relative rate of carbon-carbon coupling reaction are effective methods to improve the selectivity of multi-carbon compounds. Understanding the relationship between the selectivity of C₂₊ compounds on the above metal nanocluster catalysts and the composition and structure of the catalysts is of great significance for optimizing the structure of the catalysts and improving the catalytic properties. According to the structural characteristics of the above three kinds of catalysts, we established their model catalysts, and studied the catalytic performance of three kinds of metal nanocluster catalysts for CO₂ hydrogenation based on density functional theory (DFT)^[31]. 50-atom Ru and 50-atom Pt nanoclusters were used to simulate the Ru and Pt nanoclusters (1 ~ 2 nm in size) in the actual catalyst, respectively. Meanwhile, the Pt₄₂-Ru₈structural model was used to simulate the Ru-Pt bimetallic nanocluster, in which 3 pairs of Ru dimers (Ru₂) and 2 Ru single atoms were bonded on the Pt nanocluster, respectively (Fig. 9).

The theoretical calculation results (Table 1) based on the above model catalyst show that the energy barrier of the ^*CH₂+^*CH₂ coupling reaction on Ru nanocluster is 1.44 eV,However, the energy barriers of ^*CH₂+^*H and ^*CH₃+^*H are 0.85, respectively, and the energy barriers of 0.77 eV,^*CH₂ coupling reaction are significantly higher than those of hydrogenation reaction.The results show that the ^*CH₂ and ^*CH₃ species on the surface of Ru nanoclusters tend to react with ^*H to form CH₄. The energy barrier of ^*CH₂+^*CH₂ coupling reaction on Pt nanocluster is 0.78 eV, which is slightly lower than that of ^*CH₂+^*H(0.92 eV) and ^*CH₃+^*H(0.81 eV), which is consistent with the experimental results that the selectivity of multi-carbon compounds on Pt metal nanocluster catalyst is higher than that on Ru nanocluster catalyst. We unexpectedly found that the energy barrier for the ^*CH₂+^*CH₂ coupling reaction on the Ru₂ cluster site on the Pt₄₂-Ru₈ nanocluster is only 0.31 eV, which is not only much lower than that on the Ru₅₀ nanocluster,It is also lower than the energy barriers of ^*CH₂+^*H and ^*CH₃+^*H reactions on the active site (0.55 and 0.53 eV, respectively), indicating that ^*CH₂ species tend to couple on the Ru₂ site to form multi-carbon compounds. The above calculation results are basically consistent with the selectivity test results of multi-carbon compounds on the three metal nanocluster catalysts prepared in the previous experiment^[31].

A large number of research results show that the catalytic performance of metal nanocluster catalysts is not only related to the composition, size and morphology of metal nanoclusters, but also significantly affected by the support, ligand or modifier. The structure-function relationship of metal nanocluster catalysts has long been one of the most challenging research topics in the field of catalysis. It is difficult to precisely control the structure of metal nanoclusters by traditional catalyst synthesis methods such as impregnation, which makes it difficult to study the effects of these factors on catalytic performance separately. We have proposed a research strategy of using "Non-protected" metal nanoclusters as structural motifs to assemble heterogeneous catalysts with controllable (consistent) metal nanocluster structures, and then distinguishing the effects of different structural factors on catalytic performance, which has been widely used and has made some important research progress^{[23⇓⇓⇓~27][40⇓⇓~43]}. In addition, the theoretical calculation based on the experimental results has further deepened the understanding of the structure-activity relationship of metal nanocluster catalysts. On the basis of related theoretical and experimental studies, this paper further elaborates our understanding of the structural characteristics of the active sites of metal nanocluster catalysts and the relationship between structure and performance.

The catalytic active sites of metal nanocluster catalysts can be basically divided into two types, one is the surface atoms (or atomic groups) of metal nanoclusters, and the other is the composite catalytic active sites composed of the surface atoms of metal nanoclusters and their neighboring species (such as oxygen holes, ions, ligand atoms, etc.). Composite catalytic active sites can change the adsorption energy of reactants and the energy barrier of chemical reaction by means of coordination polarization and hydrogen bonding, showing synergistic catalytic effect.

The surface atoms of metal nanoclusters can have great differences in coordination unsaturation, arrangement, atomic spacing, charge degree or electronic structure.This makes the activated complexes (transition States) formed with reactants have different structures and properties when they catalyze a specific chemical reaction, resulting in different reaction energy barriers or catalytic activities. Usually, there is a spontaneous charge separation effect on isolated metal nanoclusters, that is, the outer metal atoms are negatively charged and the inner atoms are positively charged. The chemical bonding or electron transfer between the support, ligand or modifier and the metal nanocluster in the catalyst can significantly change the charge distribution of the metal nanocluster and its surface atomic layer, making the heterogeneity of the charge distribution on the scale of the surface atom or subatom (that is, the directionality of the charge distribution) of the metal nanocluster more significant. The above interactions also change the spacing of some atoms. In nanoclusters composed of two metal atoms or clusters, the electron migration between dissimilar metal atoms or clusters will increase the heterogeneity of charge distribution on the atomic or subatomic scale of metal nanoclusters, while changing the spacing of some metal atoms. The influence of supports, ligands or modifiers on the catalytic activity and selectivity of metal nanocluster catalysts is mainly due to the influence of bonding or charge transfer between supports and metal nanoclusters on the atomic or subatomic scale charge distribution and atomic spacing of metal nanoclusters.Or from the formation of composite (cooperative) catalytic active sites with the surface atoms of the metal nanoclusters^[31,44,45]. Hereinafter, this theory is simply refer to as that local charge distribution effect theory of metal nanocluster catalyst.

DFT theoretical calculation result show that that local atomic charge distribution on the surface of the metal nanocluster has an effect on the adsorption energy of ^*CH₂, CH^*, ^*H and other species in the hydrogenation reaction of the CO₂,As well as the C — C bond length in the transition state of the ^*CH₂+^*CH₂ coupling reaction has a significant effect, which further affects the energy barriers of the related reactions, resulting in a significant difference in the catalytic performance of the above catalysts for the carbon-carbon coupling reaction and the alkyl hydrogenolysis reaction in CO₂ hydrogenation.

The Ru dimer in the Pt₄₂-Ru₈ nanocluster is positively charged due to electron transfer to the Pt nanocluster, and the Bader charge on each Ru atom is 0.41 | e |, while the Bader charge of the surface Ru atom on the Ru₅₀ nanocluster is − 0.02 | e |. The Ru-Ru spacing is larger in the above Ru₂ than in the case of the Ru₅₀. The adsorption energy of ^*CH₂ on the Ru₂ site in the Pt₄₂-Ru₈ cluster (− 4.50 eV) is significantly higher than its adsorption energy (-3.42 eV),^*CH₂+^*CH₂ on the Ru site in the Ru₅₀ nanoclusterThe C — C bond length in the transition state of the coupling reaction (2.2626 Å) is smaller than that in the corresponding transition state on the Ru₅₀ nanocluster (2.8989 Å),So that the transition state energy of the ^*CH₂+^*CH₂ coupling reaction at the UNK 1 position is significantly reduced.

The adsorption energy of ^*H on the surface of metal nanoclusters is also an important factor affecting the reaction kinetics of hydrocarbon species hydrogenation. The adsorption energy of ^*H on the negatively charged Pt atom adjacent to the above Ru₂ site (− 3.02 eV) is significantly higher than that on the Ru₂ site (− 2.84 eV),It indicates that H prefers to adsorb on Pt sites, which is unfavorable for ^*CH₃+^*H and ^*CH₂+^*H reactions on Ru₂ sites^[31].

Based on the above analysis and the theory of local charge distribution effect of metal nanocluster catalysts, we can basically understand why Pt₄₂-Ru₈ and Ru₅₀ and their simulated Ru/C and Ru-co-Pt/C catalysts show completely different catalytic selectivity in the hydrogenation of CO₂.

XPS and X-ray absorption near edge structure (XANES) analysis have been widely used to explore the relationship between metal electron binding energy or electron population state and catalytic performance in metal nanocluster catalysts, and have played an important role in the above studies^{[40,46⇓⇓ ~49]}. However, the measurement accuracy of these analytical methods is limited (usually the accuracy of XPS is ± 0.2 eV), and in many cases, they can only effectively provide the average value of the relevant information of some or all metal elements in the sample, or such average value has received more attention. In addition, such analysis methods are difficult to provide information on the charge distribution at the atomic or subatomic scale. In metal nanocluster catalysts, the interaction between the support or modifier and the metal nanoclusters may lead to large differences in the charge distribution state or the charge separation state between the surface atomic layer and the inner core on the atomic or subatomic scale for metal nanoclusters with similar size and atomic arrangement. This makes it difficult in many cases to correlate XPS or XANES results with catalytic performance. Our recent research results show that a representative metal nanocluster catalyst model can be established on the basis of catalyst structure characterization.Theoretical studies on the charge distribution on the surface of metal nanoclusters at the atomic or subatomic scale, the distance between metal atoms, and the synergistic effect of composite catalytic sites in elementary reactions can play an active role in understanding the relationship between the structure and function of metal nanocluster catalysts^[31,44]. For example, the catalytic activity of carbon (MMC) supported Pt nanocluster (d_av=1.4 nm) catalyst modified by Miller amine on its surface for oxygen reduction reaction is much higher than that of carbon supported Pt metal nanocluster catalyst with the same size, but we did not observe the difference in the electronic state of Pt between the two in XPS and XANES test results^[44]. However, the theoretical calculation results show that the bonding interaction between MMC and Pt nanoclusters leads to significant changes in the surface atomic and subatomic charge distribution of Pt nanoclusters, as well as the charge separation between the surface atomic layer and the core. Some Pt atoms with more negative charges in Pt/MMC have lower adsorption energy for ^*OH^-, while the hydrogen bonding interaction between N in MMC and ^*OOH adsorbed on some surface atoms of Pt nanoclusters increases its adsorption strength. The local charge distribution effect of Pt nanoclusters and the existence of composite catalytic active sites in Pt/MMC are the main reasons for its high catalytic activity for oxygen reduction reaction.

The development of catalytic systems for the hydrogenation of CO₂ to multi-carbon compounds is of great significance for reducing carbon emissions and realizing the resource utilization of CO₂. Although the reaction is exothermic, the catalytic systems that can catalyze the hydrogenation of CO₂ to hydrocarbons and alcohols at low temperature are still limited. Carrying out the above catalytic conversion at low temperature is beneficial to reducing the emission of CO₂ and improving the selectivity of the target product. Iron or cobalt carbonate supported RuPt bimetallic catalysts can catalyze the hydrogenation of CO₂ to hydrocarbons and alcohols at 40 ~ 60 ℃, which is mainly due to their unique "carbonate hydrogenation and metal species carbonation coupling reaction pathway". Alcohols can be formed by the hydrolysis of the alkyl-metal bond on the surface of the catalyst. This new pathway is different from the previously reported pathway based on the insertion of CO into the alkyl-metal bond, so that the formation of alcohols can be achieved at low temperature.

The carbon-carbon coupling reaction is the key reaction step in the hydrogenation of CO₂ to form multicarbon compounds, and reducing the energy barrier of the carbon-carbon coupling reaction helps to improve the selectivity of C₂₊ compounds. The Ru-co-Pt/C nanocluster composite catalyst prepared by us can catalyze the hydrogenation of CO₂ with a selectivity of up to 90% for multi-carbon compounds. Experimental and theoretical studies have shown that the small-sized Ru cluster bonded on the Pt nanocluster is a highly active site for catalyzing the ^*CH₂+^*CH₂ coupling reaction, and the energy barriers for the reactions of ^*CH₂+^*H and ^*CH₃+^*H on this site are significantly higher than those for the ^*CH₂+^*CH₂ coupling reaction, making ^*CH₂+^*CH₂ coupling a kinetically dominant reaction. The electron transfer between metals in Ru-Pt bimetallic nanoclusters leads to the positive charge on the Ru site and the negative charge on the Pt atom adjacent to Ru. The charge distribution on the surface of Ru-Pt bimetallic nanoclusters at the atomic scale is an important factor affecting their catalytic performance.

On the basis of previous studies, the theory of active site structure characteristics and local charge distribution effect of metal nanocluster catalysts was further elaborated in this paper. This theory will play an active role in understanding the structure-function relationship of metal nanocluster catalysts and in the design and synthesis of high-performance metal nanocluster-based catalysts. Base on that structural characterization of metal nanoclusters at atomic scale and the experimental result of catalytic performance, a model catalyst is established,It is an important development direction for the future basic research of metal nanocluster-based catalysts to study the arrangement of metal atoms in metal nanocluster catalysts, the charge distribution of metal nanocluster surface atoms or subatomic scale, and the synergistic effect of composite catalytic sites in elementary reactions through theoretical calculation.