Catalytic Conversion of C 1 Molecules on Atomically Precise Metal Nanoclusters

Open AccessCCS ChemistryMINI REVIEW1 Jan 2022Catalytic Conversion of C1 Molecules on Atomically Precise Metal Nanoclusters Dan Yang, Yongnan Sun, Xiao Cai, Weigang Hu, Yihu Dai, Yan Zhu and Yanhui Yang Dan Yang Institute of Advanced Synthesis, School of Chemistry and Molecular Engineering, Nanjing Tech University, Nanjing 211816 School of Chemistry and Chemical Engineering, Nanjing University, Nanjing 210093 , Yongnan Sun School of Chemistry and Chemical Engineering, Nanjing University, Nanjing 210093 , Xiao Cai School of Chemistry and Chemical Engineering, Nanjing University, Nanjing 210093 , Weigang Hu School of Chemistry and Chemical Engineering, Nanjing University, Nanjing 210093 , Yihu Dai Institute of Advanced Synthesis, School of Chemistry and Molecular Engineering, Nanjing Tech University, Nanjing 211816 , Yan Zhu School of Chemistry and Chemical Engineering, Nanjing University, Nanjing 210093 and Yanhui Yang *Corresponding author: E-mail Address: email protected Institute of Advanced Synthesis, School of Chemistry and Molecular Engineering, Nanjing Tech University, Nanjing 211816 https://doi.org/10.31635/ccschem.021.202101188 SectionsAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail Metal nanoclusters with accurate compositions and precise crystalline structures hold remarkable attention in serving as a unique model catalyst for well-defined correlations between structure and catalytic activity. More importantly, these metal nanoclusters exhibit strong quantum confinement effects, which differ from their larger nanoparticles in a number of catalytic reactions. This review focuses on recent advances of atomically precise metal nanoclusters for C1 compound conversion (CO, CO2, CH4, and HCOOH), including thermally-driven catalysis, photocatalysis, and electrocatalysis. The reaction mechanisms are discussed at an atomic- or even electron-level. It is anticipated that the progress in this research area could be extended to catalytic applications of metal nanoclusters in C1 chemistry. Download figure Download PowerPoint Introduction C1 chemistry refers to the conversion of small compounds with only one carbon atom, including carbon monoxide (CO), carbon dioxide (CO2), methane (CH4), methanol (CH3OH), and formic acid (HCOOH), into high value-added chemicals and fuels.1,2 C1 compounds are primarily derived from coal, natural gas, organic wastes, biomass, and so on. With decreasing resources of crude oil and increasingly serious environmental concerns, C1 chemistry attracts widespread attention. Several conversion processes of C1 compounds are shown in Figure 1, presenting the key reactions, such as Fischer-Tropsch synthesis (FTS), water-gas shift (WGS), reverse WGS (RWGS), methanol to olefins (MTO), and oxidative coupling of methane (OCM).3–6 Significant advances have been made, especially for developing C1 conversion processes on the basis of nanoparticle (NP) catalysts (Figure 1). However, it remains a challenge to control selectivity toward desired products and identify reaction intermediates and active sites from C1 molecule conversion processes.6 Recently, various conventional NP catalysts have shown excellent performances for catalytic conversion of C1 molecules.7–11 However, it has the characteristics of multiple reaction pathways and products for most of the C1 conversion processes; a deep understanding on the catalyst behaviors at the atom level is essentially critical and challenging for C1 conversion to further improve the performance. The surfaces of poly-dispersed NPs are usually complicated and often have non-uniform active site distributions, where reaction intermediates are not readily identified or often not detected at all. Furthermore, even slightly different structures, such as only one atom difference, can induce significant changes in catalytic properties.12 Figure 1 | Main sources and conversion pathways of the C1 compounds (CO, CO2, CH4, and CH3OH). MTEG, methanol to ethylene glycol; MTOAH, methane conversion to olefins, aromatics, and hydrogen; NOCM, nonoxidative coupling of methane; POM, partial oxidation of methane; MDA, methane dehydroaromatization. Download figure Download PowerPoint Metal nanoclusters with atomically precise structures and molecular purity afford platforms to unravel the structure-catalytic property correlation as well as the active sites.13,14 It is a novel nanomaterial with a size range of 1–3 nm, protected by organic ligands. With a well-defined core and shell, it is appropriate for exploring the influence of surface structure (or core structure) on catalytic performances by only changing the surface structure with the core structure intact (and vice versa), which is hardly possible to achieve for conventional metal NPs. Therefore, these nanoclusters provide a solid foundation for exploring the catalytic mechanism of small molecule activation, especially for C1 molecules. To date, significant achievements have been made toward atomically precise metal nanoclusters in terms of synthesis, characterization, and structure–reactivity relationship, and relevant progress has been summarized in recent reviews.13–25 However, less information has been provided on the C1 transformation catalyzed by metal nanoclusters. In this review, the research processes of nanoclusters for C1 conversion are highlighted, including thermally-driven reaction processes, such as CO oxidation, CO2 hydrogenation, CO2 organic reaction, and CH4 oxidation, and processes under mild conditions, such as CO2 electrocatalytic reduction, HCOOH electrocatalytic reduction, light-driven CH4, and CO2 conversions (Figure 2). Particularly, the structure-reactivity relationship and reaction mechanisms are discussed in detail. In addition, the challenges and strategies for expanding the horizon to apply metal nanoclusters in C1 molecular conversion are also presented. Figure 2 | Scope and aspects covered in this review. Download figure Download PowerPoint CO Conversion CO is primarily produced from the partial oxidation or incomplete combustion of carbon-containing compounds or via CH4 reforming (Figure 1). Regarding CO conversion and utilization, it is generally divided into two aspects: CO oxidation and syngas FTS.6 The nanoclusters are unsuitable in the latter reaction because of the pyrolytic decomposition. CO oxidation is an important reaction and often employed as a probe reaction to explore the structure–performance relationships, metal size control, surface engineering, ligands, dopants or structure and support that have been identified to greatly affect the catalytic performances of metal nanoclusters. Support effect The uniformly distributed active sites are crucial in retaining the catalytic activity for a heterogeneous catalytic process. The supports benefit good dispersion but also generate active interfaces. In the area of metal nanocluster-based catalysis, Jin et al.26 carried out the pioneer work and investigated the oxidation of CO over Au25(PET)18/MOx (PET = 2-phenylethanethiol and MOx = TiO2, CeO2, and Fe2O3) catalysts, showing that the MOx supports afforded a great influence on the catalytic performances. The Au25(PET)18/TiO2 catalyst remained inactive below 200 °C, whereas the Au25(PET)18/CeO2 and Au25(PET)18/Fe2O3 catalysts exhibited moderate activity with an onset temperature of 50∼60 °C (Figure 3a), implying the substantial roles of support materials for MOx-supported gold catalysts.27,28 Unexpectedly, the Au25(PET)18/CeO2 pretreated in oxygen atmosphere at 150 °C for 1.5 h showed ∼94% CO conversion at 80 °C and ∼100% conversion at 100 °C, attributable to the formation of active surface oxygen species (Figure 3b). Higher pretreatment temperature (250 °C) cannot further increase the catalytic activity. The interface between gold and CeO2 plays a critical role in efficiently facilitating CO oxidation to form CO2. In the case of CeO2-supported Au22/catalysts after thermal pretreatment, the CeO2 rod support obtained more active lattice O and oxygen vacancies than the CeO2 cube support. Therefore, the CeO2 rods served as excellent supports for Au nanoclusters to complete CO oxidation at low temperatures.29 Figure 3 | CO conversion over Au25(PET)18/MOx catalysts pretreated in N2 at room temperature: (a) Effect of supports. (b) Effect of pretreatment. Reaction gas hourly space velocity (GHSV) = 7500 mL·g−1·h−1. Reproduced with permission from ref 26. Copyright 2012 American Chemical Society. Download figure Download PowerPoint Size effect Zhou et al.30 investigated the influence of nanocluster size on catalysis via CO oxidation. They classified the gold nanoclusters into three types: metallic nanoclusters with a size of greater than 2.3 nm, transition nanoclusters with a size of 2.3∼1.7 nm, and nonmetallic nanoclusters with a size smaller than 1.7 nm. The CO conversion over these gold nanoclusters increased with temperature, and the catalytic performance was Au144 < Au333 < Au38 < Au25 (Figures 4a and 4b), indicating that the optimal size existed over the nanoclusters with moderate sizes within a certain range. Figure 4 | (a) The curves of different sized nanoclusters on catalytic CO oxidation. (b) A volcano-like trend of size effect on catalytic CO oxidation. Reproduced with permission from ref 30. Copyright 2016 Springer Nature Limited. Download figure Download PowerPoint Ligand/surface effect Usually, the nanoclusters are capped by ligands; therefore, the geometric modification impact of ligands may either completely block the active metal sites or selectively expose the specific sites for adsorption and conversion of substrates, resulting in the ligand effects in catalytic reactions. The Li group31 explored the ligand effect of CeO2 supported Au25/nanoclusters including Au25(SC2H4Ph)18, Au25(SNap)18 (SNap = thionaphthol), and Au25(PPh3)10(PET)5X2 (PPh3 = triphenylphosphine) in CO oxidation, where the latter two catalysts exhibited low CO conversion of 3.3% and 10.2% at 100 °C, respectively, whereas the former catalyst displayed a remarkably high conversion at identical reaction conditions. Furthermore, the Au25(PET)18, Au25(SNap)18, and Au25(PPh3)10(PET)5X2 catalysts after O2-pretreatment at 100 °C afforded 98.5%, 98%, and 94.6% CO conversion at 150 °C, respectively. The observed results were mainly due to the rich electron density of Au25(SNap)18 and Au25(PPh3)10(PET)5X2 and the large ligand steric hindrance of Au25(PPh3)10(PET)5X2. Importantly, it was inferred that the enhanced activity after the induction period (oxygen pretreatment) was caused by the partial removal of ligands prior to the reaction. Jin et al.32 further explored the ligand effect in CO oxidation. The intact Au25(PET)18 nanocluster on CeO2 rods (fresh catalyst) was unable to adsorb CO. However, CO can be effectively adsorbed and the oxidation reaction activity can be increased dramatically when partially removing the thiolate ligands. Activity increased with increasing O2-pretreatment temperature and then decreased with further temperature increases, while the maximum activity was achieved on the catalysts pretreated at 473∼523 K (Figures 5a and 5b). In situ Fourier transform infrared (FR-IR) spectroscopy and extended X-ray absorption fine structure (EXAFS) results revealed low activity for the nanoclusters pretreated at 423 K, which was the onset temperature for ligand removal. The excellent catalytic performance obtained on the catalyst pretreated at 473 K was owed to the partial removal of the PET ligands, and hence the Au sites of the nanoclusters can be exposed for activating relevant reaction substrates (Figures 5c and 5d). Figure 5 | (a) CO oxidation for Au25(PET)18/CeO2 rod catalyst pretreated in O2 at different temperatures. GHSV: 18,800 cm3·h−1·gcat−1. (b) Arrhenius plots for conversions below 20%. (c) In situ FT-IR spectra and (d) EXAFS spectra of samples pretreated at different temperatures in O2. Reproduced with permission from ref 32. Copyright 2014 American Chemical Society. Download figure Download PowerPoint In addition to the structural control on the core of metal nanoclusters, surface tailoring without influencing the core structure has also been achieved to tune the catalytic performances. The Chen group33 explored the effect of surface structure of nanoclusters on catalytic process using two quasi-isomer Au28(S-c-C6H11)20 (S-c-C6H11 = cyclohexanethiolate) and Au28(SPh-tBu)20 (SPh = benzenethiol and SPh-tBu = p-tert-butylphenylthiophenol) nanoclusters induced from ligands possessing a similar kernel structure but different surface structures. The former nanocluster exhibited a superior catalytic performance on CO oxidation than the latter nanocluster (Figure 6a). It was rationalized that CO was more readily adsorbed and activated on the Au28(S-c-C6H11)20 surface, due to less steric hindrance. Furthermore, the difference in catalytic activity between these two nanoclusters disappeared when completely removing the ligands (Figure 6b). Combining the density functional theory (DFT) calculation results, the catalytic active sites were identified and located on the staple gold. Figure 6 | CO oxidation on CeO2 supported Au28(S-c-C6H11)20 (black profile) and Au28(SPh-tBu)20 (red) catalysts: (a) Pretreated with O2 at 150 °C for 1 h. (b) Pretreated with O2 at 300 °C for 1 h to remove ligands. Reproduced with permission from ref 33. Copyright 2018 American Chemical Society. Download figure Download PowerPoint Jin and co-workers34 investigated the surface effect of Au nanoclusters on CO reaction activity. The results showed that the catalytic activity was Au38(SCH2CH2Ph)24 < Au38(SPh)24 < Au38(o-MBT)24 (o-MBT = o-methyl thiophenol), Au36(S-c-C5H9)24 (S-c-C5H9 = cyclopentanethiol) < Au36(SPh)24 ≈ Au36(SPh-tBu)24 < Au36(2SNap)24 (2SNap = 2-thionaphthol; Figures 7a and 7b), and Au25(SCH2CH2Ph)18 (SCH2CH2Ph = 2-phenylethanethiol) < Au25(2SNap)18 < Au25(1SNap)18 (1SNap =1-thionaphthol), implying that the steric hindrance existed among the ligand, gold, and support, rather than the carbon tails, which was an important factor in the catalytic performance since it prevented the adsorption of CO onto the gold active sites. Moreover, Au38 displayed an excellent catalytic property when it was pretreated at ∼150 °C, while the activity of Au36 cannot be remarkably changed when pretreating in the temperature range of 100∼225 °C (Figures 7c and 7d). It revealed that dissimilar to Au38, Au36 nanoclusters were not impressionable to thermal treatment. Therefore, except for the surface structure, the core structure of nanoclusters also has a critical effect on the catalytic performance. Figure 7 | (a) CO oxidation over CeO2 supported Au38(SCH2CH2Ph)24, Au38(SPh)24, and Au38(o-MBT)24 nanoclusters. (b) CO oxidation over CeO2 supported Au36(S-c-C5H9)24, Au36(SPh)24, Au36(SPh-tBu)24, and Au36(2SNap)24 nanoclusters. (c and d) CO oxidation of (c) Au38(SPh)24 and (d) Au36(SPh)24 pretreated at different temperatures. Reproduced with permission from ref 34. Copyright 2018 American Chemical Society. Download figure Download PowerPoint It is well established that surface atoms with low coordination number on catalysts are regarded as active sites in heterogeneous catalysis. Metal nanostructures protected by ligands involve different local coordination environments, which can affect the catalytic properties. Wu et al.35 first discovered that Au22(L8)6 (L=1,8-bis(diphenylphosphino) octane donate as Au22) nanoclusters can catalyze the oxidation of CO at low temperatures without ligand removal. The fresh sample showed mild activity at room temperature, which was remarkably different from the reported fresh Au25 (Figure 8a), implying that the coordinatively unsaturated (cus) gold atoms (Figure 8b) in the fresh Au22 were active at low-temperature. DFT calculations revealed that the cus gold sites in the fresh Au22 can activate CO as well as O2. Wang et al.36 also reported similar results in that Au23 showed excellent catalytic performance caused by eight uncoordinated gold sites on the metal nanoclusters surface. Figure 8 | (a) CO oxidation curves for different Au22(L8)6-TiO2 samples. (b) The total structure of Au22(L8)6 nanocluster: blue = cus Au, yellow = other Au, grey = C, white = H, pink = P). Reproduced with permission from ref 35. Copyright 2016 American Chemical Society. Download figure Download PowerPoint Doping effect Doping with heteroatoms has been attempted to modify the catalytic performance of metal nanoclusters. Li et al.37 utilized the CeO2-supported CuxAu25-x(PET)18, AgxAu25-x(PET)18, and Au25(PET)18 nanoclusters to explore the doping effect on the catalytic performance via CO oxidation. The results demonstrated that CuxAu25-x(PET)18 had the highest activity, followed by Au25(PET)18 and AgxAu25-x(PET)18 (Figures 9a and 9b),37 indicating that the heteroatoms can indeed change the catalytic activity. Recently, Zhu et al.38 reported a new nanocluster Ag2Au50(PET)36, resulting from two Au25(PET)18 linked with two Ag atoms. The Ag2Au50(PET)36 catalyst was significantly more effective than Au25(PET)18 and inner shell-doped AgxAu25-x(PET)18 catalyst for CO oxidation, which was attributed to a greater content of active sites. In situ FT-IR demonstrated that Ag2Au50(PET)36 contained two kind of active sites, Ag sites and Au sites, while Au25(PET)18 and AgxAu25-x(PET)18 catalysts consisted of only Au sites (Figures 10a and 10b).38 Figure 9 | (a) Catalytic activity of the MxAu25-x(PET)18/CeO2 (M = Cu and Ag), and Au25(PET)18/CeO2 after thermal treatment (120 °C under the reaction gas atmosphere for 1 h). (b) Structure of the Au13@M2Au10(SCH3)15 clusters, where M = Au, Ag, and Cu, pink = two dethiolated metal atoms. Reproduced with permission from ref 37. Copyright 2016 American Chemical Society. Download figure Download PowerPoint Figure 10 | (a) CO oxidation for CeO2 supported Ag2Au50(PET)36, Au25(SR)18, and AgxAu25-x(SR)18. (b) Total structure of Ag2Au50(PET)36. Green/blue = Au, yellow = S, red = Ag. Reproduced with permission from ref 38. Copyright 2020 Wiley-VCH Verlag GmbH (2) some ligands might be removed during the reaction to expose the active sites; (3) it afforded different activity and selectivity from the conventional NP catalysts; (4) CO2 was activated with the assistance of the active hydrogen, not directly activated by metal nanoclusters. Au9(PPh3)8(NO3)3 (denoted as Au9), Au11(PPh3)8Cl2Cl (denoted as Au11), and Au36(TBBT)24 (denoted as Au36; TBBT = p-tert-butylphenylthiophenol) showed different performances from the plasmonic/metallic gold NP catalysts. The activity of nanoclusters was 10–70 than that of Au more to CH4 as the to and Au36 to whereas CH4 was produced on Au NPs with activity. In addition, and protected with different ligands, were to explore the influence of ligand on the catalytic activity and The results revealed that these three nanoclusters showed activity than that of and with the products CH4, and implying that the surface ligands may only influence the catalytic activity, not the it further proved that the catalytic selectivity on the gold sites of the ligands. In situ FT-IR spectra showed that the ligands were partially removed to expose the active sites during the reaction. In addition, a of and can be to and on catalyst, that CH4 was In the case of it was that was via in species. Au36 catalyst, HCOOH was more to be produced via the adsorbed the and species were the intermediates for CH4 and and the species was an important for DFT calculations revealed that CO2 were adsorbed on the nanocluster surface, and hence it was activated by an on the Au sites to form either or species. Au36 the pathway and produced HCOOH with a of pathway with and the pathway was more to be carried out with only and respectively, of the Particularly, and to and form CH4 rather than to form or coupling coupling of with was more on with a low because of the steric hindrance it revealed that the kinetic for the of and the adsorption for species to different catalytic and reaction pathways (Figures Figure | (a) structures of and (b) of CO2 hydrogenation on and by DFT The H, C, S, and Au atoms are shown in and respectively. (c) Catalytic performances of CO2 hydrogenation over the supported and (d) Catalytic results of CO2 hydrogenation over the three gold nanoclusters pretreated to remove ligands. of key reaction intermediates of the toward various products on the three nanoclusters. Reproduced with permission from ref Copyright 2020 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim. Download figure Download PowerPoint structure nanoclusters, including = for Au36(TBBT)24 = for = for and = for were to the catalytic performance of the hydrogenation of CO2. the of these nanoclusters, Au36 the coordination number for Au and the highest molecular molecular to surface sites in the catalytic process. Au36 showed a property with forming whereas mainly afforded while and mainly produced The reported quantum size effect was to be to the results, on the the larger size of nanoclusters to catalytic activity. In addition, the ligand effect was via in situ FT-IR, showing that the ligand could be partially removed with no of nanoclusters under the reaction conditions. In situ FT-IR spectra showed that was from the hydrogenation of toward and intermediates on catalyst and HCOOH was on Au36 catalyst via and catalysts, and species were implying that was produced via coupling of and species. DFT calculations that Au36 afforded adsorption for and intermediates because of larger and coordination Particularly, Au36 was to with rather than it had adsorption for which was to Moreover, and catalysts had remarkably similar with small coordination It was more to with and also had adsorption with which prevented CO and further Additionally, and species were more to on the and clusters, showing the of Therefore, the reaction activity and selectivity were to the and structures of nanoclusters (Figures Figure | (a) structures of nanoclusters = atoms in yellow and Au atoms in blue and and Catalytic activity and selectivity of for CO2 Reaction catalyst % H2O 3 reaction gas °C. and in situ spectra of the reaction intermediates in CO2 hydrogenation over nanoclusters.