English

• The depletion of fossil fuels and climate change are two major problems we face today, so reducing greenhouse gas emissions and finding alternative carbon sources is urgent [1–3]. Converting CO₂ into value-added chemicals [4–7] like methanol (CH₃OH) and formic acid (HCOOH) presents an intriguing approach to both generating renewable energy and mitigating greenhouse gas concentrations [8–11]. Nevertheless, the process of CO₂ hydrogenation poses a significant challenge due to the thermodynamic and kinetic stability of CO₂ [12,13]. To enhance the conversion rate of CO₂ and the selective production of desired products, the development of high-performance catalysts is essential.

  Atomically dispersed catalysts, such as single-atom catalysts (SACs) [14–19] and dual-atom catalysts (DACs) [19–24], have been widely employed in CO₂ reduction reaction (CO₂RR) and have shown variations in catalytic activity and selectivity. For example, Shi et al. [19] reported that Cu/PCN SACs favored the desorption of ^*CO to generate CO products, while InCu/PCN DACs promoted the coupling of two ^*CO intermediates to form ethanol through the synergistic effect of In-Cu. Ren et al. [25] discovered that Ni/Fe-NC DACs demonstrated superior CO₂RR performance for CO production compared to Fe SACs and Ni SACs. The CO selectivity remained above 90% across a wide potential range from −0.5 V to −0.9 V. Wang et al. [26] observed that Fe_2–N-C electrocatalyst achieved an enhanced CO Faradaic efficiency exceeding 80% across broader potential ranges, with a higher turnover frequency and improved durability compared to SAC counterparts. Sun et al. [27] reported that the Faradaic efficiency for CO₂-to-C₂H₄ over Cu₂/NC reached up to ~34.9% at a current density of 33.6 mA/cm², whereas its Cu single-atom counterpart exclusively produced CO without yielding C₂H₄. Despite extensive and ongoing efforts to investigate the application of DACs and SACs in CO₂ conversion, further research is still required to fully understand the catalytic differences between SACs and DACs in CO₂ hydrogenation and to advance the development of efficient catalysts.

  Pt-based catalysts have been widely used in various reactions due to their superior activity [28–30], and Pt SACs and Pt DACs exhibit different catalytic activities. For example, a recent study reported that Pt₂ dimers formed on graphene (Pt₂/graphene) displayed a ~17-fold higher catalytic activity than graphene supported Pt single atoms (Pt₁/graphene) in hydrolytic dehydrogenation of ammonia borane [31]. Inspired by this, we ponder whether Pt₂/graphene outperforms Pt₁/graphene in CO₂ hydrogenation. Clarifying this is beneficial for expanding the potential applications of atomically dispersed catalysts in chemical reactions. Herein, in this work, we investigated the catalytic performance of Pt₁/graphene and Pt₂/graphene for CO₂ hydrogenation by using density functional theory (DFT) calculations and microkinetic analysis. Results indicated that Pt₁/graphene and Pt₂/graphene exhibited significant differences in the catalytic properties for CO₂ hydrogenation. Pt₁/graphene tended to produce HCOOH, whereas Pt₂/graphene facilitated the conversion of CO₂ into CH₃OH. Additionally, the turnover frequencies (TOF) of the top-performing structures among the three configurations of Pt₁/graphene and Pt₂/graphene were predicted to reach as high as 744.48 h^-1 and 789.48 h^-1, respectively, demonstrating their exceptional catalytic activity in CO₂ hydrogenation.

  According to previous studies [31], Pt₁/graphene was synthesized under O₂ environments, leading to the chemisorption of O₂ molecules on Pt atoms and the presence of two O atoms at the interface. In contrast, the synthesis of Pt₂/graphene in O₃ environments resulted in the formation of Pt₂O₆ chain structures, where the O atoms alternate between terminal and bridge positions. The specific structures of Pt₁/graphene and Pt₂/graphene depended on both the size of the C-vacancy and the configuration of the two interfacial O atoms. Based on these findings, we established three distinct models for Pt₁/graphene and Pt₂/graphene, respectively, and investigated their catalytic performance for CO₂ hydrogenation.

  As shown in Fig. 1, three Pt₁/graphene models were identified as Pt_1ⅰ/graphene, Pt_1ⅱ/graphene, and Pt_1ⅲ/graphene. Pt_1ⅰ/graphene featured a single Pt atom supported on the armchair edge of graphene, where the Pt atom was situated within a seven-membered ring composed of one Pt atom, two O atoms, and four C atoms (Fig. 1a). Pt_1ⅱ/graphene showcased a single Pt atom supported on the armchair edge of graphene, where the Pt atom was located within a five-membered ring consisting of one Pt atom, two O atoms, and two C atoms (Fig. 1b). Pt_1ⅲ/graphene exhibited a single Pt atom supported on the zigzag edge of graphene, where the Pt atoms were positioned within a six-membered ring comprising one Pt atom, two O atoms, and two C atoms (Fig. 1c). Pt₂/graphene configurations were constructed based on the corresponding Pt₁/graphene, with the end-adsorbed ^*O₂ molecule being replaced by 2O-Pt-O₂ to form a lattice structure. These three Pt₂/graphene configurations were labeled as Pt_2ⅰ/graphene, Pt_2ⅱ/graphene, and Pt_2ⅲ/graphene, as shown in Figs. 1d-f, respectively.

  Figure 1

  

  Figure 1.  The structures of (a) Pt_1ⅰ/graphene, (b) Pt_1ⅱ/graphene, (c) Pt_1ⅲ/graphene, (d) Pt_2ⅰ/graphene, (e) Pt_2ⅱ/graphene, and (f) Pt_2ⅲ/graphene adsorbing ^*O₂ molecules. The navy blue, grey, and red spheres represented Pt, C, and O atoms, respectively.

  DownLoad: Full-Size Img PowerPoint

  After establishing the catalyst structures, we initially treated them with H₂ to remove the chemisorbed ^*O₂ from the structures. The free energy profiles for the reaction of ^*O₂ with H₂ over Pt_1ⅰ/graphene, Pt_1ⅱ/graphene, Pt_1ⅲ/graphene, Pt_2ⅰ/graphene, Pt_2ⅱ/graphene, and Pt_2ⅲ/graphene were displayed in Figs. S1, S3-S6, and S8 (Supporting information), respectively. For Pt₁/graphene structures, the Pt atoms act as the active sites, while for Pt₂/graphene configurations, the top Pt atoms (named as Pt_top) served as the active sites for the reaction between ^*O₂ and H₂. After the reaction of ^*O₂ with H₂, ^*O₂ in the Pt₁/graphene and Pt₂/graphene structures were removed in the form of H₂O molecules, and the Pt active sites were occupied by two ^*H species (Fig. S9 in Supporting information).

  The two H species on Pt sites may transfer to adjacent O atoms. To illustrate this phenomenon, we chose Pt_1ⅱ/graphene and Pt_2ⅱ/graphene as examples for analyzing the energy barriers associated with H transfer. The calculated values for Pt_1ⅱ/graphene and Pt_2ⅱ/graphene were 1.44 eV (Fig. S10 in Supporting information) and 0.42 eV (Fig. S11 in Supporting information), respectively. The results suggested that the migration of H from Pt sites to O atoms was relatively challenging on Pt_1ⅱ/graphene, whereas it was remarkably straightforward on Pt_2ⅱ/graphene. After the H on the Pt_top sites of Pt_2ⅱ/graphene migrated to O atoms, an H₂ molecule dissociated into two H species and then transferred to the O atoms with an energy barrier of 0.85 eV (Fig. S11).

  Based on the above analysis, we concluded that after Pt₁/graphene with O₂ adsorption was treated with H₂, it ultimately transformed into a 2H-Pt₁/graphene structure, where the Pt site was occupied by two ^*H species, as depicted in Figs. S12a-c (Supporting information). For Pt₂/graphene with O₂ adsorption, after H₂ treatment, it eventually transformed into a 2H-2OH-Pt₂/graphene structure, where the Pt_top site was occupied by two ^*H species, and the bridge oxygen between the two Pt coordinated with the H species, as shown in Figs. S12d-f (Supporting information). Subsequently, we introduced CO₂ molecules to explore the process of CO₂ conversion.

  For 2H-Pt₁/graphene, CO₂ molecules were firstly adsorbed on Pt atoms, and the adsorption free energies of CO₂ on 2H-Pt_1ⅰ/graphene, 2H-Pt_1ⅱ/graphene, and 2H-Pt_1ⅲ/graphene were 0.48, 0.71, and 0.56 eV, respectively. The liner ∠O—C-O on 2H-Pt_1ⅰ/graphene, 2H-Pt_1ⅱ/graphene, and 2H-Pt_1ⅲ/graphene was shrunk to 142.04°, 145.06°, and 139.62°, respectively. The adsorption of CO₂ was followed by hydrogenation to form ^*COOH or ^*HCOO intermediates. The energy barriers for the formation of ^*COOH over 2H-Pt_1ⅰ/graphene, 2H-Pt_1ⅱ/graphene, and 2H-Pt_1ⅲ/graphene were 0.51, 0.50, and 0.39 eV, respectively, which were lower than that for ^*HCOO generation with energy barriers of 1.09, 0.86, and 0.92 for 2H-Pt_1ⅰ/graphene, 2H-Pt_1ⅱ/graphene, and 2H-Pt_1ⅲ/graphene, respectively. The result indicated that ^*CO₂ tended to hydrogenate into ^*COOH species on 2H-Pt₁/graphene.

  The ^*COOH species were unstable and may directly combine with ^*H to produce ^*HCOOH. The energy barriers for hydrogenation of ^*COOH to ^*HCOOH over 2H-Pt_1ⅰ/graphene, 2H-Pt_1ⅱ/graphene, and 2H-Pt_1ⅲ/graphene were 1.90 eV (Fig. S13 in Supporting information), 1.92 eV (Fig. 2), and 1.34 eV (Fig. S14 in Supporting information), respectively, indicating that direct hydrogenation of ^*COOH with ^*H to form ^*HCOOH was unlikely. Instead, it was more likely that an H₂ molecule dissociated into two ^*H atoms firstly, followed by the combination of ^*COOH and ^*H to form ^*HCOOH. Thus, we investigated the dissociation of H₂ molecules in the presence of ^*COOH. For 2H-Pt_1ⅰ/graphene (Fig. S13), the energy barrier for H₂ dissociation was 1.22 eV, which was lower than that for direct hydrogenation of ^*COOH to ^*HCOOH (1.90 eV). A similar phenomenon was observed for 2H-Pt_1ⅱ/graphene and 2H-Pt_1ⅲ/graphene. Thus, we concluded that ^*COOH did not immediately react with ^*H to form ^*HCOOH on three 2H-Pt₁/graphene configurations, but rather dissociated H₂ first, followed by ^*COOH reacting with dissociated ^*H to form ^*HCOOH. The ^*HCOOH was likely to react further with ^*H or to desorb from the Pt site. The hydrogenation of ^*HCOOH to ^*H₂COOH showed a higher energy than ^*HCOOH desorption on three 2H-Pt₁/graphene configurations, suggesting that ^*HCOOH would desorb rather than further hydrogenate.

  Figure 2

  

  Figure 2.  The free energy profiles for CO₂ hydrogenation to HCOOH on 2H-Pt_1ⅱ/graphene. TS represented the transition state. The black route was the most likely reaction pathway. The navy blue, grey, white, and red spheres represented Pt, C, H, and O atoms, respectively. H_v represented one H atom was missing from 2H-Pt_1ⅱ/graphene structure after it binds to ^*CO₂ or ^*COOH intermediate.

  DownLoad: Full-Size Img PowerPoint

  Although three 2H-Pt₁/graphene configurations followed the same pathway to hydrogenate CO₂ into HCOOH, the rate-determining step (RDS) varied structures. For 2H-Pt_1ⅰ/graphene and 2H-Pt_1ⅲ/graphene, the RDS was ^*H₂→2^*H, with energy barriers of 1.22 and 1.02 eV, respectively. In contrast, for 2H-Pt_1ⅱ/graphene, the RDS was ^*COOH+^*H→^*HCOOH, with an energy barrier of 0.66 eV Among the three 2H-Pt₁/graphene configurations, 2H-Pt_1ⅱ/graphene showed the lowest RDS energy barriers for CO₂ hydrogenation to HCOOH.

  For the three 2H-2OH-Pt₂/graphene configurations, Pt₂ dimers acted as the active sites for CO₂ hydrogenation. The CO₂ molecules were physically adsorbed on 2H-2OH-Pt₂/graphene, with adsorption free energies of 0.30, 0.41, and 0.43 eV for 2H-2OH-Pt_2ⅰ/graphene, 2H-2OH-Pt_2ⅱ/graphene, and 2H-2OH-Pt_2ⅲ/graphene, respectively. After adsorption on the surface, CO₂ will undergo gradual hydrogenation to form CH₃OH. The reaction pathways for CO₂ hydrogenation to CH₃OH over 2H-2OH-Pt_2ⅰ/graphene, 2H-2OH-Pt_2ⅱ/graphene, and 2H-2OH-Pt_2ⅲ/graphene were illustrated in Fig. 3, Figs. S16 and S17 (Supporting information), respectively. Considering that the optimal hydrogenation pathway for CO₂ over three 2H-2OH-Pt₂/graphene configurations was similar, we used 2H-2OH-Pt_2ⅰ/graphene as an illustrative example to analyze the detailed reaction pathway.

  Figure 3

  

  Figure 3.  The free energy profiles for the hydrogenation of CO₂ to CH₃OH on 2H-2OH-Pt_2ⅰ/graphene, and the corresponding intermediate structures were shown in Fig. S15 (Supporting information). TS represented the transition state. The black route was the most possible reaction pathway for CO₂ hydrogenation to CH₃OH on 2H-2OH-Pt_2ⅰ/graphene. H_v represented one H atom was missing from 2H-2OH-Pt_2ⅰ/graphene structure after it binds to intermediate. ^*H_top and ^*H_non-top represented H atoms adsorbed on Pt_top sites and Pt_non-top sites, respectively.

  DownLoad: Full-Size Img PowerPoint

  In the reaction process, the adsorption of CO₂ was followed by hydrogenation to form ^*COOH or ^*HCOO intermediates (Fig. 3). The energy barrier to generate ^*COOH on 2H-2OH-Pt_2ⅰ/graphene was 0.71 eV, which was lower than that for ^*HCOO (1.37 eV), indicating that CO₂ preferentially generated ^*COOH. Following the generation of ^*COOH, an H₂ molecule adsorbed on the Pt_top atoms, which then dissociated into two ^*H species. One of the ^*H atoms on the Pt_top atoms migrated to the non-top Pt atoms (named as Pt_non-top) with an energy barrier of 0.04 eV. In this stage, the Pt_non-top atoms adsorbed one H species, while the Pt_top atoms adsorbed two H species and one ^*COOH species. One of the ^*H atoms on the Pt_top atoms combined with ^*COOH to form ^*C(OH)₂ or ^*HCOOH species. The formation of ^*C(OH)₂ exhibited a lower energy barrier (0.86 eV) than ^*HCOOH formation (1.12 eV), suggesting that H preferred to combine with ^*COOH to form ^*C(OH)₂ species.

  When the ^*C(OH)₂ species were formed, the ^*H atom on the Pt_non-top atoms migrated to the Pt_top atoms with a migration barrier of 0.69 eV Subsequently, the ^*H atom on the Pt_top site reacted with ^*C(OH)₂, generating ^*CH(OH)₂ with an energy barrier of 0.33 eV. A second H₂ molecule adsorbed onto the Pt_top sites and dissociated into two ^*H atoms. Afterward, one ^*H atom migrated from the Pt_top sites to the Pt_non-top sites, and the ^*CH(OH)₂ species combined with the ^*H on the Pt_top atoms, breaking the C—O bond and forming an ^*H₂O molecule with an energy barrier of 0.82 eV. However, the ^*H₂O molecule was unstable and desorbed from the Pt_top sites. After desorption of H₂O, one ^*HCOH species and one ^*H species remained on the Pt_top atom. The ^*H on the Pt_non-top atom transferred to the Pt_top sites with an energy barrier of 0.22 eV. One of the ^*H atoms on the Pt_top atom combined with ^*HCOH to form ^*CH₂OH species. A third H₂ molecule physically adsorbed onto the Pt_top sites and dissociated into two H atoms with a low energy barrier of 0.60 eV, and one of the ^*H atoms on the Pt_top sites reacted with ^*CH₂OH to generate ^*CH₃OH products. The RDS of three 2H-2OH-Pt₂/graphene configurations were the dissociation of the first H₂ molecules, with energy barriers of 0.93, 0.87, and 0.87 eV for 2H-2OH-Pt_2ⅰ/graphene, 2H-2OH-Pt_2ⅱ/graphene, and 2H-2OH-Pt_2ⅲ/graphene, respectively. Three Pt₂/graphene configurations exhibited similar RDS energy barriers.

  Based on the above discussion, it was observed that three Pt₁/graphene configurations tended to produce HCOOH, whereas three Pt₂/graphene configurations preferred CH₃OH. To investigate the differences in product selectivity between Pt₂/graphene and Pt₁/graphene, we chose Pt_1ⅱ/graphene and Pt_2ⅰ/graphene as examples to analyze the hydrogenation pathways of the ^*COOH intermediates. As shown in Fig. 4a, the Pt atoms of Pt_1ⅱ/graphene adsorbed three ^*H and one ^*COOH species. The combination of ^*H and ^*COOH on Pt_1ⅱ/graphene resulted in the formation of ^*HCOOH with an energy barrier of 0.66 eV, which was lower than that for ^*C(OH)₂ formation (0.75 eV). The result indicated that ^*COOH would directly react with ^*H to produce ^*HCOOH on Pt_1ⅱ/graphene.

  Figure 4

  

  Figure 4.  The free energy barriers for ^*COOH hydrogenation to ^*HCOOH and ^*C(OH)₂ over (a) Pt_1ⅱ/graphene and (b) Pt_2ⅰ/graphene.

  DownLoad: Full-Size Img PowerPoint

  For Pt_2ⅰ/graphene (Fig. 4b), the Pt_top atoms also adsorbed three ^*H and one ^*COOH species. Nevertheless, ^*H did not directly react with ^*COOH. Instead, it migrated to the Pt_non-top atoms with an energy barrier of 0.04 eV. Subsequently, one of the ^*H atoms on the Pt_top atoms combined with ^*COOH to form ^*HCOOH or ^*C(OH)₂. The formation of ^*C(OH)₂ exhibited a lower energy barrier (0.86 eV) compared to the generation of ^*HCOOH (1.12 eV), suggesting that ^*COOH species on Pt_2ⅰ/graphene were more likely to undergo hydrogenation to generate ^*C(OH)₂ species. Thus, the difference in selectivity between Pt₁/graphene and Pt₂/graphene derived from the migration of ^*H from the Pt_top atoms to the Pt_non-top atoms, altering the pathway of CO₂ hydrogenation.

  The catalytic reactivity of Pt₁/graphene and Pt₂/graphene catalysts for CO₂ hydrogenation processes was further assessed using microkinetic simulations. The reaction conditions were set to be the same as those in previous works [32,33], with specifics detailed in Tables S1 and S2 (Supporting information), as well as computational methods. The microkinetic simulation results for Pt₁/graphene and Pt₂/graphene catalysts were presented in Fig. 5.

  Figure 5

  

  Figure 5.  The results of microkinetic simulations. (a) Turnover frequency (TOF) for HCOOH formation on Pt₁/graphene from the catalytic CO₂ hydrogenation reaction. T = 473.15 K, P = 4.5 MPa and H₂/CO₂ ratio of 3:1. (b) TOF for CH₃OH formation on Pt₂/graphene from the catalytic CO₂ hydrogenation reaction. T = 353.15 K, P = 4.0 MPa and H₂/CO₂ ratio of 3:1.

  DownLoad: Full-Size Img PowerPoint

  For the three Pt₁/graphene catalysts, there is a significant difference in TOF values for the generation of HCOOH (Fig. 5a). Among them, Pt_1ⅱ/graphene exhibited the highest TOF value of 744.48 h^-1, which was comparable to that previously reported value (780.7 h^-1) [33] under identical reaction conditions (T = 473.15 K, P = 4.50 MPa, and H₂/CO₂ ratio of 3:1). To explore the reason for the superior activity of Pt_1ⅱ/graphene compared to Pt_1ⅰ/graphene and Pt_1ⅲ/graphene, we investigated the center of d-orbitals (ɛ_d) of the metal. The ɛ_d for Pt atoms in 2H-Pt_1ⅰ/graphene, 2H-Pt_1ⅱ/graphene and 2H-Pt_1ⅲ/graphene were −1.21, −1.11, and −1.98 eV, respectively (Fig. 6a). Notably, the ɛ_d of 2H-Pt_1ⅱ/graphene was closer to the Fermi level than that of 2H-Pt_1ⅰ/graphene and 2H-Pt_1ⅲ/graphene, which facilitated faster electron transfer and thereby enhancing the catalytic activity for CO₂ hydrogenation to HCOOH.

  Figure 6

  

  Figure 6.  (a) PDOS of Pt d-orbitals and the corresponding d-orbitals center for three 2H-Pt₁/graphene structures. The Fermi level (E_F) was set to 0 eV. The integrated crystal orbital Hamilton population (ICOHP) between Pt_top and Pt_non-top of (b) 2H-2OH-Pt_2ⅰ/graphene, (c) 2H-2OH-Pt_2ⅱ/graphene and (d) 2H-2OH-Pt_2ⅲ/graphene. The distance depicted in the illustration represented the separation between Pt_top and Pt_non-top, measured in Å.

  DownLoad: Full-Size Img PowerPoint

  For three Pt₂/graphene catalysts, the TOF values for the formation of CH₃OH varied significantly (Fig. 5b). Among these catalysts, Pt_2ⅰ/graphene showed the highest TOF value (789.48 h^-1), which was magnitude larger than of a previously reported Ir-based catalyst (0.15 h^-1) [32] under the same reaction conditions (T = 353.15 K, P = 4.0 MPa, and H₂/CO₂ ratio of 3:1). To investigate the underlying cause of the enhanced activity of Pt_2ⅰ/graphene compared to Pt_2ⅱ/graphene and Pt_2ⅲ/graphene, we calculated the distance and the integrated crystal orbital Hamilton population (ICOHP) between Pt_top and Pt_non-top to evaluate their interaction. As shown in Figs. 6b-d, 2H-2OH-Pt_2ⅰ/graphene exhibited a shorter distance between Pt_top and Pt_non-top (2.773 Å) than 2H-2OH-Pt_2ⅱ/graphene (3.105 Å) and 2H-2OH-Pt_2ⅲ/graphene (3.109 Å). Furthermore, the calculated ICOHP value between Pt_top and Pt_non-top for 2H-2OH-Pt_2ⅰ/graphene was −0.13 eV, which was more negative than that for 2H-2OH-Pt_2ⅱ/graphene (−0.03 eV) and 2H-2OH-Pt_2ⅲ/graphene (−0.06 eV). These results indicated that the closer Pt-Pt distance in 2H-2OH-Pt_2ⅰ/graphene enhanced the interaction between two adjacent Pt atoms, thus accelerating the electron transfer in the process of kinetic transformation and improving the catalytic activity [34].

  To assess the stability of 2H-Pt_1ⅱ/graphene and 2H-2OH-Pt_2ⅰ/graphene, we performed ab initio molecular dynamics simulations at 473 K for 2H-Pt_1ⅱ/graphene and at 353 K for 2H-2OH-Pt₂_ⅰ/graphene. The simulations lasted for 6 ps with a time-step of 1 fs. The geometric structures of both 2H-Pt_1ⅱ/graphene and 2H-2OH-Pt_2ⅰ/graphene showed no significant distortion during the simulations, indicating their high thermodynamic stability under reaction conditions (Fig. S18 in Supporting information).

  In conclusion, we investigated the catalytic performance of Pt₁/graphene and Pt₂/graphene for CO₂ hydrogenation, and observed that Pt₁/graphene configurations favored the conversion of CO₂ into HCOOH, whereas CO₂ was hydrogenated into CH₃OH on Pt₂/graphene. The distinct product selectivity between Pt₁/graphene and Pt₂/graphene originated from the synergistic interaction of Pt₂ dimers, which hydrogenated ^*COOH into ^*C(OH)₂ instead of ^*HCOOH, altering the hydrogenation pathway of CO₂. Additionally, Pt_1ⅱ/graphene and Pt_2ⅰ/graphene showed high activity for CO₂ hydrogenation with TOF values of 744.48 h^-1 and 789.48 h^-1, respectively. This work paves the way for manipulating catalytic properties and advancing the mechanistic understanding of heterogeneous catalysis.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  CRediT authorship contribution statement

  Sanmei Wang: Writing – review & editing, Writing – original draft, Software, Methodology, Investigation, Formal analysis, Data curation. Dengxin Yan: Software, Methodology, Data curation. Wenhua Zhang: Writing – review & editing, Supervision, Software, Funding acquisition, Data curation. Liangbing Wang: Writing – review & editing, Supervision, Funding acquisition.

  Acknowledgments

  This work was supported by the National Key Research and Development Program (No. 2022YFA1505800), the National Natural Science Foundation of China (No. 22373092), CAS Project for Young Scientists in Basic Research (No. YSBR-051). China Association for Science and Technology (No. YESS20200031), the Start-up Funding of Central South University (No. 502045005), and Industry-University-Research Cooperation Projects with Zhejiang NHU Co., Ltd., and Ningbo Fengcheng Advanced Energy Materials Research Institute. Wenhua Zhang is supported by USTC Tang Scholarship. The calculations were performed on the Super-computing Center of University of Science and Technology of China (USTC-SCC).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.110611.

  
  
• 1. [1]

     J. Rogelj, M. den Elzen, N. Höhne, et al., Nature 534 (2016) 631–639. doi: 10.1038/nature18307

  2. [2]

     J.D. Shakun, P.U. Clark, F. He, et al., Nature 484 (2012) 49–54. doi: 10.1038/nature10915

  3. [3]

     S. Jin, Z. Hao, K. Zhang, et al., Angew. Chem. Int. Ed. 60 (2021) 20627–20648. doi: 10.1002/anie.202101818

  4. [4]

     L. Li, F. Chen, B. Zhao, Y. Yu, Chin. Chem. Lett. 35 (2024) 109240. doi: 10.1016/j.cclet.2023.109240

  5. [5]

     Q. Cheng, M. Huang, Q. Ye, et al., Chin. Chem. Lett. 35 (2024) 109112. doi: 10.1016/j.cclet.2023.109112

  6. [6]

     H. Zhang, Q. Liu, Z. Shen, Chin. Chem. Lett. 35 (2024) 108607. doi: 10.1016/j.cclet.2023.108607

  7. [7]

     T. Jia, L. Wang, Z. Zhu, et al., Chin. Chem. Lett. 35 (2024) 108692. doi: 10.1016/j.cclet.2023.108692

  8. [8]

     G.A. Olah, G.K.S. Prakash, A. Goeppert, J. Am. Chem. Soc. 133 (2011) 12881–12898. doi: 10.1021/ja202642y

  9. [9]

     X. Duan, J. Xu, Z. Wei, et al., Adv. Mater. 29 (2017) 1701784. doi: 10.1002/adma.201701784

  10. [10]

      R.P. Ye, J. Ding, W. Gong, et al., Nat. Commun. 10 (2019) 5698. doi: 10.1038/s41467-019-13638-9

  11. [11]

      F. Wang, S. Zhang, W. Jing, et al., J. Mater. Sci. Technol. 189 (2024) 146–154. doi: 10.1016/j.jmst.2024.01.002

  12. [12]

      G. Glockler, J. Phys. Chem. 62 (1958) 1049–1054. doi: 10.1021/j150567a006

  13. [13]

      S. Wang, Q. Li, Y. Xin, et al., Nanoscale 15 (2023) 6999–7005. doi: 10.1039/D2NR05874J

  14. [14]

      T. Yang, X. Mao, Y. Zhang, et al., Nat. Commun. 12 (2021) 6022. doi: 10.1038/s41467-021-26316-6

  15. [15]

      K.W. Ting, T. Toyao, S.M.A.H. Siddiki, K.I. Shimizu, ACS Catal. 9 (2019) 3685–3693. doi: 10.1021/acscatal.8b04821

  16. [16]

      Y. Chen, H. Li, W. Zhao, et al., Nat. Commun. 10 (2019) 1885. doi: 10.1038/s41467-019-09918-z

  17. [17]

      Y. Hu, H. Li, Z. Li, et al., Green Chem. 23 (2021) 8754–8794. doi: 10.1039/D1GC02331D

  18. [18]

      T. Liu, W. Zhang, Q. Tan, et al., Chem. Eng. J. 489 (2024) 150286. doi: 10.1016/j.cej.2024.150286

  19. [19]

      H. Shi, H. Wang, Y. Zhou, et al., Angew. Chem. Int. Ed. 61 (2022) e202208904. doi: 10.1002/anie.202208904

  20. [20]

      H. Li, L. Wang, Y. Dai, et al., Nat. Nanotechnol. 13 (2018) 411–417. doi: 10.1038/s41565-018-0089-z

  21. [21]

      Z. Zeng, L.Y. Gan, H. Bin Yang, et al., Nat. Commun. 12 (2021) 4088. doi: 10.1038/s41467-021-24052-5

  22. [22]

      Y. Lou, F. jiang, W. Zhu, et al., Appl. Catal. B: Environ. 291 (2021) 120122. doi: 10.1016/j.apcatb.2021.120122

  23. [23]

      W. Zhang, T. Liu, Q. Tan, et al., ACS Catal. 13 (2023) 3242–3253. doi: 10.1021/acscatal.2c05129

  24. [24]

      X. Wan, Y. Li, Y. Chen, et al., Nat. Commun. 15 (2024) 1273. doi: 10.1038/s41467-024-45516-4

  25. [25]

      W. Ren, X. Tan, W. Yang, et al., Angew. Chem. Int. Ed. 58 (2019) 6972–6976. doi: 10.1002/anie.201901575

  26. [26]

      Y. Wang, B.J. Park, V.K. Paidi, et al., ACS Energy Lett. 7 (2022) 640–649. doi: 10.1021/acsenergylett.1c02446

  27. [27]

      G. Sun, Y. Cao, D. Li, et al., Appl. Catal. A: Gen. 651 (2023) 119025. doi: 10.1016/j.apcata.2023.119025

  28. [28]

      Y. Si, Y. Jiao, M. Wang, et al., Nat. Commun. 15 (2024) 4887. doi: 10.1038/s41467-024-49083-6

  29. [29]

      X. Chen, M. Peng, X. Cai, et al., Nat. Commun. 12 (2021) 2664. doi: 10.1038/s41467-021-22948-w

  30. [30]

      Y. Sun, S. Wang, D. Jiao, et al., Chin. Chem. Lett. 33 (2022) 3987–3992. doi: 10.1016/j.cclet.2021.11.034

  31. [31]

      H. Yan, Y. Lin, H. Wu, et al., Nat. Commun. 8 (2017) 1070. doi: 10.1038/s41467-017-01259-z

  32. [32]

      R. Kanega, N. Onishi, S. Tanaka, et al., J. Am. Chem. Soc. 143 (2021) 1570–1576. doi: 10.1021/jacs.0c11927

  33. [33]

      Z. Wang, Y. Kang, J. Hu, et al., Angew. Chem. Int. Ed. 62 (2023) e202307086. doi: 10.1002/anie.202307086

  34. [34]

      Y. Hu, Z. Li, B. Li, C. Yu, Small 18 (2022) 2203589. doi: 10.1002/smll.202203589

• 

  Figure 1  The structures of (a) Pt_1ⅰ/graphene, (b) Pt_1ⅱ/graphene, (c) Pt_1ⅲ/graphene, (d) Pt_2ⅰ/graphene, (e) Pt_2ⅱ/graphene, and (f) Pt_2ⅲ/graphene adsorbing ^*O₂ molecules. The navy blue, grey, and red spheres represented Pt, C, and O atoms, respectively.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  The free energy profiles for CO₂ hydrogenation to HCOOH on 2H-Pt_1ⅱ/graphene. TS represented the transition state. The black route was the most likely reaction pathway. The navy blue, grey, white, and red spheres represented Pt, C, H, and O atoms, respectively. H_v represented one H atom was missing from 2H-Pt_1ⅱ/graphene structure after it binds to ^*CO₂ or ^*COOH intermediate.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  The free energy profiles for the hydrogenation of CO₂ to CH₃OH on 2H-2OH-Pt_2ⅰ/graphene, and the corresponding intermediate structures were shown in Fig. S15 (Supporting information). TS represented the transition state. The black route was the most possible reaction pathway for CO₂ hydrogenation to CH₃OH on 2H-2OH-Pt_2ⅰ/graphene. H_v represented one H atom was missing from 2H-2OH-Pt_2ⅰ/graphene structure after it binds to intermediate. ^*H_top and ^*H_non-top represented H atoms adsorbed on Pt_top sites and Pt_non-top sites, respectively.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  The free energy barriers for ^*COOH hydrogenation to ^*HCOOH and ^*C(OH)₂ over (a) Pt_1ⅱ/graphene and (b) Pt_2ⅰ/graphene.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  The results of microkinetic simulations. (a) Turnover frequency (TOF) for HCOOH formation on Pt₁/graphene from the catalytic CO₂ hydrogenation reaction. T = 473.15 K, P = 4.5 MPa and H₂/CO₂ ratio of 3:1. (b) TOF for CH₃OH formation on Pt₂/graphene from the catalytic CO₂ hydrogenation reaction. T = 353.15 K, P = 4.0 MPa and H₂/CO₂ ratio of 3:1.

  下载: 全尺寸图片 幻灯片
  

  Figure 6  (a) PDOS of Pt d-orbitals and the corresponding d-orbitals center for three 2H-Pt₁/graphene structures. The Fermi level (E_F) was set to 0 eV. The integrated crystal orbital Hamilton population (ICOHP) between Pt_top and Pt_non-top of (b) 2H-2OH-Pt_2ⅰ/graphene, (c) 2H-2OH-Pt_2ⅱ/graphene and (d) 2H-2OH-Pt_2ⅲ/graphene. The distance depicted in the illustration represented the separation between Pt_top and Pt_non-top, measured in Å.

  下载: 全尺寸图片 幻灯片
•