Skip to main content
NCBI home page
JACS Au logoLink to JACS Au
. 2023 Feb 1;3(2):508–515. doi: 10.1021/jacsau.2c00632

Nature of Catalytic Behavior of Cobalt Oxides for CO2 Hydrogenation

Kailang Li †, Xianghong Li †, Lulu Li †, Xin Chang †, Shican Wu †, Chengsheng Yang †, Xiwen Song †, Zhi-Jian Zhao †,*, Jinlong Gong †,‡,§,∥,*
PMCID: PMC9975827  PMID: 36873681

Abstract

graphic file with name au2c00632_0006.jpg

Cobalt oxide (CoOx) catalysts are widely applied in CO2 hydrogenation but suffer from structural evolution during the reaction. This paper describes the complicated structure–performance relationship under reaction conditions. An iterative approach was employed to simulate the reduction process with the help of neural network potential-accelerated molecular dynamics. Based on the reduced models of catalysts, a combined theoretical and experimental study has discovered that CoO(111) provides active sites to break C–O bonds for CH4 production. The analysis of the reaction mechanism indicated that the C–O bond scission of *CH2O species plays a key role in producing CH4. The nature of dissociating C–O bonds is attributed to the stabilization of *O atoms after C–O bond cleavage and the weakening of C–O bond strength by surface-transferred electrons. This work may offer a paradigm to explore the origin of performance over metal oxides in heterogeneous catalysis.

Keywords: CO2 hydrogenation, structure−performance relationship, CoO catalysts, density functional theory, multiscale simulation

Introduction

The efficient utilization of greenhouse gas CO2 has always been of considerable attraction. CO2 hydrogenation is a mature technology, which not only can convert CO2 into high value-added chemicals but also mitigate the greenhouse effect by the consumption of CO2.1−3 Co-based CO2 hydrogenation catalysts are widely applied in methanation,4 methanol production,5 and C–C coupling reactions, i.e., the synthesis of long-chain alkanes6 and higher alcohols.7,8 However, the key factors that affect catalytic performance are still controversial, which are possibly related to, for instance, size dependence,9 metal–support interaction,10 crystal facet dependence,11,12 surface segregation,13 and coverage effect.14

In particular, the structures of Co-based oxide catalysts are complex with variable valence states under reaction conditions.15 Co3O4 is a typical representative that exhibits structure sensitivity during CO2 hydrogenation.16,17 To be more specific, variations in catalyst morphology lead to significant differences in product distribution. Co3O4 nanoparticles display high selectivity for CH4, while Co3O4 nanorods tend to produce CO. In the meantime, Co3O4 is not stable during the reaction. Phase transformation induced by H2 occurs, where Co3O4 is reduced to CoO or even metallic Co. As a result, CoO is believed to be the main active phase during CO2 hydrogenation.18 The complexity of dynamically structural evolution in bulk and surface makes it challenging to interpret the origin of performance over cobalt oxides (CoOx).

Limited by current in situ characterization techniques, it is hard to accurately identify the complex feature sites and perceive the evolution of feature structures. Although density functional theory (DFT) calculations can provide atomic-scale insights into the rational design of catalysts, the expensive computational cost prevents realizing large-scale and long-term simulations, leading to a large gap between simulated models and real catalytic systems. Hopefully, machine learning techniques19,20 enable breaking through the dilemma and surmount the spatial and temporal limitations of DFT by accelerating the time-consuming simulation with an affordable computational cost, pointing out a promising route to describe the dynamic catalyst behavior at the atomic level during the reaction.21,22

Here, neural network potential-based molecular dynamics (NN-MD) simulations were carried out to probe the dynamically structural evolution from Co3O4 to CoO according to experimental reduction conditions. With the application of DFT calculations, kinetic Monte Carlo (kMC) simulations, and experimental studies, the correlation between structure and selectivity was established at the atomic level. We revealed the nature of C–O bond dissociation on feature structures and provided a clear perspective on the reaction mechanism of CoO verified by our experiments. This work can offer a theoretical guidance for rational design of CoOx catalysts in CO2 hydrogenation by a practical multiscale research method for establishing the structure–performance relationship of metal oxides in heterogeneous catalysis.

Results and Discussion

Structural Evolution

In order to understand the dynamically structural evolution of Co3O4 under the reaction condition, NN-MD simulations were employed to simulate the reduction process of the precursor, i.e., Co3O4. The commonly exposed crystalline surfaces of experimentally synthesized nanostructure catalysts, i.e., Co3O4(100), Co3O4(110), and Co3O4(111),18 were chosen as the model surfaces prior to the reduction with a scale of about 1000 atoms in one unit cell. The reduction process was simulated by two iterative steps, which are surface reduction and reconstruction (Figure 1a).23 To begin the reduction process, O vacancy formation free energies (GOv) of surface O atoms under experimental reduction conditions were calculated, and the ones with negative values were removed to mimic surface reduction reactions. Afterward, the partially reduced surface reached equilibrium by 1.5 ns NN-MD simulations with a canonical ensemble (NVT), during which some subsurface O atoms might diffuse to the surface. Then, another round of reduction-reconstruction simulation started, until GOv values of all surface O atoms are greater than 0 eV, implying that the surface cannot be further reduced thermodynamically (Scheme S1 and Figures S1–S9). The feature structures after reduction were consequently confirmed via NN-MD.

Figure 1.

Figure 1

Open in a new tab

(a) Principle of surface reaction-reconstruction to simulate the reduction process of Co3O4. μO represents the chemical potential of O atoms. (b) NN-MD simulations of the reduction process on the Co3O4 model. Color code: blue-Co; red-O.

It is gratifying that the obtained reduced surface model is consistent with the experimental observations that the CoO phase is the main active phase of CoOx catalysts in CO2 hydrogenation.24 From the simulated surface structures where the ratio of Co and O atoms is about 1:1.(Figure 1b), the regular CoO(100) facet is observed after the reduction of Co3O4(100). The reduced Co3O4(110) mainly exposes CoO(110)- and CoO(100)-like structures, accompanied by few defective sites, like grain boundaries and oxygen vacancies. The CoO(111) facet with a small portion of oxygen vacancies and clusters is obtained after the reduction of Co3O4(111). So far, three feature structures after reduction, CoO(100), CoO(110), and CoO(111), have been obtained via our iterative reduction simulations.

Structure–Performance Relationship

On the basis of these surface structures, small slab models were selected for accurate DFT calculations. Over the oxide catalysts, hydroxyl groups (OH) and oxygen vacancies (Ov) may form, induced by the H2 atmosphere. Hence, thermodynamic phase diagrams were calculated with consideration of the possible presence of OH and Ov, and the stable structures were confirmed under the reaction condition (T = 573 K and pH2 = 0.2 atm), which are clean CoO(100), CoO(110) with 1/4 ML OH, and CoO(111) with 7/9 ML OH (Figures S10 and S11).

We further explored the structure–performance relationship of CoO and compared it with the metallic Co that may appear after the reaction.16,25 CO2 activation, as the initial step of CO2 hydrogenation, has an essential impact on subsequent reactions. After adsorption, the CO2 molecule tends to be bent as it acquires electrons from the surface. Straightforwardly, the changes in configurations can visualize the activation ability of different structures. Among them, the most obvious variation of CO2 configuration occurs on CoO(111). The C=O bond length increases from 1.18 to 1.39 Å and ∠OCO decreases from 180 to 122°, indicating that CO2 could be easily activated on CoO(111) (Figure 2a). Analyzed by charge transfer, the CO2 molecule adsorbed on CoO(111) acquires the most electrons (1.11 e) (Figure S16). The transferred electrons fill in the antibonding orbitals of CO2, which can activate the C=O bond and promote the CO2 molecule to be adsorbed in the form of bending rather than linear.26

Figure 2.

Figure 2

Open in a new tab

(a) Bond lengths and bond angles of adsorbed *CO2 on different surfaces. (b) Energy profile for CO hydrogenation and desorption on different surfaces. (c) Energy profile for C–O bond dissociation on different surfaces. (d) Red: The relationship between E(*O) and the C–O bond length; blue: the relationship between E(*O) and -ICOHP of the C–O bond. (e) Relationship between E(*O) and C–O bond dissociation. (f) Electron transfer of adsorbed *O on different surfaces.

The reaction mechanisms of CO2 hydrogenation are mainly divided into the formate pathway, carboxyl pathway, and direct C–O dissociation pathway,27 while it is generally accepted that direct dissociation of the C–O bond is very difficult without the involvement of H.28,29 In the first place, the carboxyl pathway (RWGS reaction + CO hydrogenation) was taken into account, because CO is the reaction product and key reaction intermediate for cobalt oxides,15 and its further hydrogenation is believed as an efficient route that is responsible for the formation of deep reduction products.30 Here, we started with CO hydrogenation to *CHO, as well as its competition step CO desorption, to explore deep reduction capacity (Figure 2b). Surprisingly, our calculated barriers of further hydrogenation are always higher than corresponding CO desorption barriers over all the cobalt oxide module surfaces. Therefore, the CO intermediate is hard to be reduced, which prefers desorption as a product.

CH4 is another main product of cobalt oxides in CO2 hydrogenation besides CO.18,25 As a deep reduction product, CH4 tends to be generated via the formate pathway as analyzed above. The C–O bond dissociation of *CHxO is critical to produce CH4 under the formate pathway, suggesting that *CHxO intermediates serve as key species to determine CH4 selectivity.15,31−33 Since *CHO needs to cross the relatively high barrier to be dissociated (Figure S39), *CH2O and *CH3O species were selected as possible precursors for the C–O bond scission.30 As shown in Figure 2c, for *CH2O and *CH3O species adsorbed on CoO(100) and CoO(110), the reaction is highly endothermic, which is thermodynamically unfavorable for C–O dissociation. In consequence, CoO(100) and CoO(110) need to overcome high barriers to break the C–O bond. On the contrary, CoO(111) is easy to dissociate the C–O bond with low barriers, which are even lower than metallic Co, effectively facilitating CH4 production.

The C–O bond cleavage is usually related to the C–O bond length, the longer C–O bond length is more prone to break the C–O bond.34 Interestingly, the adsorption energies of *O atoms (E(*O)) have a good linear relation with the C–O bond length (Figure 2d), implying that E(*O) may be used to reveal the nature of C–O bond scission. To further quantify the chemical bond strength, we calculated the integrated crystal orbital Hamilton population (ICOHP) of the C–O bond, where the smaller value of -ICOHP means the weaker C–O bond strength. Apparently, the stronger the *O atom is adsorbed on the surface, the smaller the -ICOHP is, indicating easier breaking the C–O bond and thereby promoting the production of CH4. Hence, E(*O) can be a simple descriptor to measure C–O bond scission (Figure 2e).

We further explored essential factors that may lead to the difference in E(*O). The phenomenon can be explained by the analysis of geometric structures (Figure 2f). CoO(111) and Co(111) provide three-fold hollow sites to efficiently stabilize *O atoms after dissociation of *CHxO intermediates. However, *O on CoO(110) and CoO(100) can only be adsorbed on top and bridge sites, resulting in the instability of *O atoms with high barriers for breaking the C–O bond, and thereby inhibiting the formation of CH4. In addition, the electron transfer is equally vital. Compared with other surfaces, the *O atom on CoO(111) acquires more electrons (1.06 e) from the surface. The transferred electrons can fill in the antibonding orbitals of the C–O bond of *CHxO to weaken the C–O bond strength, promoting to break C–O bonds. Therefore, the nature of C–O bond dissociation is attributed to the stabilization of *O atoms after C–O bond cleavage, and the weakening of C–O bond strength via surface-transferred electrons.

Kinetic Investigation

To gain a comprehensive insight into the catalytic mechanism of CoO in CO2 hydrogenation, the whole reaction networks were constructed by the combination of DFT calculations and kMC simulations with CoO(111) as a representative that exhibits wonderful capacities of CO2 activation and C–O bond cleavage. As it is hard to break C–O bonds directly without the assistance of H,28,29 the formate pathway and carboxyl pathway that involves H to assist in dissociating C–O bonds are considered as dominant reaction routes. The reaction network involves 21 different intermediates and 46 elementary steps. The relevant reaction energies and barriers, along with lateral interactions of coadsorption species, are listed in Tables S1 and S2.

As shown in Figure 3a,b, the CO2 hydrogenation reaction is dominated by the production of CH4 through the formate pathway, because the event frequency of *HCOO formation is between 2 and 3 orders of magnitude greater than that of *COOH production. After hydrogenation and dissociation of *HCOO, *CH2O is favorable to be formed without any high barrier to overcome. The formation of CH4 mainly relies on the C–O bond dissociation of *CH2O and *CH3O. The frequency of breaking the C–O bond of *CH2O is almost 4 times higher than that of *CH3O, implying that the task on dissociating C–O bonds is mainly completed via *CH2O (R33, Eact_fwd = 0.72 eV), rather than *CH3O (R34, Eact_fwd = 1.10 eV). Thus, the C–O bond scission of *CH2O becomes the critical step to determine the CH4 selectivity. In addition, as the other product, the main bottleneck for CO to be deeply reduced is that the desorption of *CO (R35, Eact_fwd = 1.51 eV) is more favorable than further hydrogenation or cleavage of *CO to generate *C (R14, Eact_fwd = 3.75 eV), *CHO (R19, Eact_fwd = 1.94 eV) or *COH (R9, Eact_fwd = 2.72 eV), suggesting that CO tends to be the product, which is consistent with the above analysis.

Figure 3.

Figure 3

Open in a new tab

(a) Reaction networks of CoO(111) for CO2 hydrogenation toward CO, CH4, and CH3OH based on DFT calculations. The values are barriers of hydrogenation and C–O dissociation. All values are in eV with ZPE correlation. The most favorable pathway toward CH4 is highlighted with green bold arrows. (b) Event frequency for elementary steps on CoO(111) based on kMC simulations in CO2 hydrogenation.

It is also worth noting that the removal of surface O species (*O and *OH) to form H2O(g) plays a crucial role in the overall reaction rate. The process of *O hydrogenation to form *OH occurs frequently, but only a small portion of *OH on the surface can be further hydrogenated to produce H2O (Eact_fwd = 1.56 eV). Hence, a high coverage of *OH is shown on the surface when the reaction reaches equilibrium (Table 1). The analysis of degree of rate control35 indicates that *OH hydrogenation to H2O controls the rate of the whole reaction (XRC = 0.825), which is the rate-determining step (Table S3).

Table 1. Surface Species Coverage for the Reaction Equilibrium.

species θ (ML)a species θ (ML)a
*CO2 31/3200 *H2COOH 95/3200
*H 92/3200 *CH3O 2/3200
*OH 2328/3200 *CH 4/3200
*HCOO 217/3200 *CH2 10/3200
*H2COO 2/3200 *CH3 14/3200
Open in a new tab
a

θ is the coverage of surface species; ML represents the monolayer.

Experimental Verification

To verify the structure–performance relationship of CoO catalysts in CO2 hydrogenation, we synthesized CoO catalysts with relevant feature structures, matched with theoretical calculation models. Foremost, we prepared Co3O4 samples as the precursors via a hydrothermal method, where nanoparticles expose {111} + {001} facets and nanorods are enclosed with {110} + {001} facets.18 Moreover, {111} and {110} facets take up most of the surface of nanoparticles and nanorods, respectively (Figure S31).36 Though the treatment of H2 at 573 K, the samples display CoO diffraction peaks from X-ray diffraction characterization (Figures S33 and S34). In addition, Co 2p3/2 peaks tested by X-ray photoelectron spectroscopy (XPS) shift from about 779.4 to 780.1 eV. Both of them indicate that Co3O4 is reduced to CoO (Figure S35).37−39 After reduction, CoO catalysts retain nanoparticle and nanorod morphology, while preferential facets are altered. Co3O4(111) is transformed into CoO(111), and Co3O4(110) is converted to CoO(110) (Figure S32), which is consistent with the structural evolution from Co3O4 to CoO via NN-MD simulations.

Afterward, CoO catalysts were tested under the reaction condition (p = 1 atm, T = 573 K and CO2/H2 = 1:3) (Figure 4a). Reaction activity and product distribution change considerably as the morphology of CoO transforms. Upon converting nanoparticles to nanorods, CH4 selectivity decreases from 98 to 23% and CO selectivity increases from 2 to 77%. Meanwhile, CO2 conversion decreases from 46 to 13%. Apparently, CoO nanoparticles are more prone to catalyze the methanation reaction, possessing a stronger capacity to promote C–O bond cleavage with a higher reaction activity. Nevertheless, methanation on CoO nanorods is significantly inhibited. As a consequence, CO is the main product.

Figure 4.

Figure 4

Open in a new tab

Experimental results of the synthesis catalysts. (a) Catalytic performance of CoO-NP and CoO-NR catalysts. Reaction conditions: p = 1 atm, T = 573 K, 0.15 g catalyst, CO2/H2 = 1:3. (b) O 1s XPS spectra of CoO-NP, CoO-NP-used, CoO-NR, and CoO-NR-used. In situ DRIFTS spectra of (c) CoO-NP and (d) CoO-NR. Reaction conditions: p = 1 atm, T = 573 K, CO2/H2 = 1:3. CoO-NP and CoO-NR represent CoO nanoparticles and nanorods after the reduction of Co3O4 by H2; CoO-NP-used and CoO-NR-used represent CoO nanoparticles and nanorods after the CO2 hydrogenation reaction.

Notably, compared with CoO nanorods, a high fraction of hydroxyl O occurs on CoO nanoparticles after the reaction, where the binding energy of O 1s on about 529.5 and 531.5 eV is assigned to lattice O and hydroxyl O on surfaces (Figure 4b).37,44−46 It suggests that the removal of hydroxyl O over CoO(111) exposed on nanoparticles is hard and has a key impact on the overall reaction rate, in agreement with kMC simulations.

To demonstrate the accuracy of our proposed reaction mechanism, in situ DRIFTS was employed to observe surface intermediates (Figure 4c,d). The peaks at 2832, 1588, and 1367 cm–1 are assigned to stretching vibrations of the C–H bond, stretching, and bending vibrations of OCO, respectively, indicating the existence of *HCOO species.40−42 In addition, compared with the weak peak at 1450 cm–1 over CoO nanoparticles, there is an obvious peak at 1458 cm–1 over CoO nanorods. They might correspond to the vibrational mode of *CHx in *CHxO species,43 since the similar peaks are obtained by DFT calculations, which are related to the bending vibration of *CH2 in *CH2O on CoO(111) (1431 cm–1) and CoO(110) (1440 cm–1). The phenomenon indicates the accumulation of *CHxO species on CoO nanorods (CoO(110)), due to the poor capability for C–O bond scission. Inversely, it is easy for CoO nanoparticles (CoO(111)) to break C–O bonds, leading to rapid consumption of *CHxO. As a result, the signal peak of CH4 on 3016 cm–1 can be clearly observed.

Conclusions

In summary, this paper presents a multiscale calculation method to deal with a complex structure–performance relationship over CoOx catalysts. Through employing the surface reduction-reconstruction approach to simulate the reduction process of cobalt oxides via NN-MD, three theoretical models have been obtained with the application of the surface phase diagram, which are clean CoO(100), CoO(110) with 1/4 ML OH, and CoO(111) with 7/9 ML OH. DFT calculations, kMC simulations, and experimental studies were combined to reveal the correlation between the structure and selectivity at the atomic level, wherein CoO(111) contributes to CH4 production and CoO(110) promotes CO formation. We have provided a clear perspective on the reaction mechanism of CoO and found that the C–O bond scission of the *CH2O intermediate plays a key role in determining CH4 selectivity. The nature of dissociating C–O bonds was unveiled, which is attributed to the stabilization of *O atoms after C–O bond cleavage and the weakening of C–O bond strength by surface-transferred electrons. These discoveries could give a deep insight into the origin of performance over CoOx catalysts in CO2 hydrogenation.

Computational and Experimental Methods

NN-MD Simulations

All calculations with NN-MD were performed by LASP, which is generally applied in large-scale simulation for complex chemical systems via NN potential.47 The details of the surface reduction-reconstruction method to simulate the reduction process of Co3O4 can be found in the Supporting Information.

DFT Calculations

DFT calculations were carried out via Vienna ab initio simulation software.48,49 The electron exchange and correlation effects were described by the Perdew–Burke–Ernzerhof functional form of the generalized gradient approximation (GGA).50 In consideration of van der Waals correction for all systems, the DFT-D3 method with Becke–Jonson damping was utilized.51 The projector augmented wave method was used to describe the interaction between atomic cores and electrons.52 To solve the Kohn–Sham equations, the plane-wave basis set was employed with a cutoff energy of 400 eV. In addition, the atomic force convergence criterion of force was set to 0.02 eV/Å. For metal oxides, the DFT + U method was applied to better describe the localized 3d electrons of cobalt in CoO, where an effective U value (Ueff = UCo–JCo = 3.7 eV) was adopted.53−55 More details are described in the Supporting Information.

kMC Simulations

Based on the thermodynamic and kinetic parameters calculated by DFT over CoO(111) (Table S1), the kMC method56 that can simulate the system evolution during the reaction at the molecular level was carried out by software package Zacros57,58 to simulate the CO2 hydrogenation reaction. To perform kMC simulations under experimental reaction conditions, the total pressure and the temperature of our reaction system are set to 1 atm and 573 K, respectively. More details are described in the Supporting Information.

Experimental Details

The CoO catalysts were synthesized via a hydrothermal method. To be more specific, 2.4 g of C4H6O4·Co·4H2O and 60 mL of (CH2OH)2 were added into a three-port round-bottomed flask. The mixture was then heated up to about 433 K. Afterward, 20 mL of 0.5 M K2CO3 solution was added. Under constant stirring and a continuous N2 flow, the slurry was further aged for 0 h and 3 h, respectively. Then, the solid was recovered by centrifugation. We used ultrapurity water and anhydrous ethanol to wash the solid, until the supernatant is neutral (pH = 7). After drying at 343 K overnight, the solid is calcined at 723 K for about 4 h in air. At the moment, as the precursor of CoO, Co3O4 catalysts were obtained. Though the further treatment of H2 at 573 K for 1 h, we finally prepared the CoO catalysts.

The catalytic performance of CoO was tested in a fixed-bed reactor. The reaction condition was set to 573 K and 1 atm. The molar ratio of CO2 and H2 was 1:3. In addition, N2 was the internal standard gas during the reaction. The exhaust streams were analyzed via an online gas chromatograph (GC, Agilent 7890A).

More experimental details are described in the Supporting Information.

Acknowledgments

We acknowledge the National Key R&D Program of China (2021YFA1500704), the National Natural Science Foundation of China (No. 22121004 and U22A20409), the Haihe Laboratory of Sustainable Chemical Transformations, the Program of Introducing Talents of Discipline to Universities (BP0618007), and the XPLORER PRIZE for financial support. We also acknowledge generous computing resources at High Performance Computing Center of Tianjin University.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/jacsau.2c00632.

  • Experimental details and supporting data (PDF)

Author Contributions

K.L. performed the theoretical calculation and the experiment. X.L., C.Y., and X.S. contributed to some experimental test and characterization of the catalysts. L.L., X.C., and S.W. contributed to useful discussion. Z.-J.Z. and J.G. designed and directed the project. Z.-J.Z. and J.G. supervised the project. All the authors contributed to the modification of the manuscript. CRediT: Kailang Li data curation, formal analysis, investigation, writing-original draft, writing-review & editing; Xianghong Li data curation, formal analysis, investigation, methodology, validation, writing-review & editing; Lulu Li data curation, formal analysis, investigation, writing-review & editing; Xin Chang data curation, formal analysis, investigation, writing-review & editing; Shican Wu data curation, formal analysis, investigation, writing-review & editing; Chengsheng Yang formal analysis, investigation, methodology; Xiwen Song data curation, formal analysis, investigation; Zhi-Jian Zhao conceptualization, data curation, formal analysis, funding acquisition, project administration, supervision, validation, writing-original draft, writing-review & editing; Jinlong Gong conceptualization, funding acquisition, investigation, project administration, supervision, validation, writing-original draft, writing-review & editing.

The authors declare no competing financial interest.

Supplementary Material

au2c00632_si_001.pdf (3.8MB, pdf)

References

  1. Jiang X.; Nie X.; Guo X.; Song C.; Chen J. G. Recent Advances in Carbon Dioxide Hydrogenation to Methanol via Heterogeneous Catalysis. Chem. Rev. 2020, 120, 7984–8034. 10.1021/acs.chemrev.9b00723. [DOI] [PubMed] [Google Scholar]
  2. Zhong J.; Yang X.; Wu Z.; Liang B.; Huang Y.; Zhang T. State of the art and perspectives in heterogeneous catalysis of CO2 hydrogenation to methanol. Chem. Soc. Rev. 2020, 49, 1385–1413. 10.1039/C9CS00614A. [DOI] [PubMed] [Google Scholar]
  3. Gao P.; Zhang L.; Li S.; Zhou Z.; Sun Y. Novel Heterogeneous Catalysts for CO2 Hydrogenation to Liquid Fuels. ACS Cent. Sci. 2020, 6, 1657–1670. 10.1021/acscentsci.0c00976. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Li W.; Nie X.; Jiang X.; Zhang A.; Ding F.; Liu M.; Liu Z.; Guo X.; Song C. ZrO2 support imparts superior activity and stability of Co catalysts for CO2 methanation. Appl. Catal. B 2018, 220, 397–408. 10.1016/j.apcatb.2017.08.048. [DOI] [Google Scholar]
  5. Li C.-S.; Melaet G.; Ralston W. T.; An K.; Brooks C.; Ye Y.; Liu Y.-S.; Zhu J.; Guo J.; Alayoglu S.; Somorjai G. A. High-performance hybrid oxide catalyst of manganese and cobalt for low-pressure methanol synthesis. Nat. Commun. 2015, 6, 6538. 10.1038/ncomms7538. [DOI] [PubMed] [Google Scholar]
  6. Xie C.; Chen C.; Yu Y.; Su J.; Li Y.; Somorjai G. A.; Yang P. Tandem Catalysis for CO2 Hydrogenation to C2-C4 Hydrocarbons. Nano Lett. 2017, 17, 3798–3802. 10.1021/acs.nanolett.7b01139. [DOI] [PubMed] [Google Scholar]
  7. Wang L.; Wang L.; Zhang J.; Liu X.; Wang H.; Zhang W.; Yang Q.; Ma J.; Dong X.; Yoo S. J.; Kim J.-G.; Meng X.; Xiao F.-S. Selective Hydrogenation of CO2 to Ethanol over Cobalt Catalysts. Angew. Chem., Int. Ed. 2018, 57, 6104–6108. 10.1002/anie.201800729. [DOI] [PubMed] [Google Scholar]
  8. He Z.; Qian Q.; Ma J.; Meng Q.; Zhou H.; Song J.; Liu Z.; Han B. Water-Enhanced Synthesis of Higher Alcohols from CO2 Hydrogenation over a Pt/Co3O4 Catalyst under Milder Conditions. Angew. Chem., Int. Ed. 2016, 55, 737–741. 10.1002/anie.201507585. [DOI] [PubMed] [Google Scholar]
  9. van Etten M. P. C.; Zijlstra B.; Hensen E. J. M.; Filot I. A. W. Enumerating Active Sites on Metal Nanoparticles: Understanding the Size Dependence of Cobalt Particles for CO Dissociation. ACS Catal. 2021, 11, 8484–8492. 10.1021/acscatal.1c00651. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Parastaev A.; Muravev V.; Osta E. H.; van Hoof A. J. F.; Kimpel T. F.; Kosinov N.; Hensen E. J. M. Boosting CO2 hydrogenation via size-dependent metal-support interactions in cobalt/ceria-based catalysts. Nat. Catal. 2020, 3, 526–533. 10.1038/s41929-020-0459-4. [DOI] [Google Scholar]
  11. Li W.; Nie X.; Yang H.; Wang X.; Polo-Garzon F.; Wu Z.; Zhu J.; Wang J.; Liu Y.; Shi C.; Song C.; Guo X. Crystallographic dependence of CO2 hydrogenation pathways over HCP-Co and FCC-Co catalysts. Appl. Catal. B 2022, 315, 121529 10.1016/j.apcatb.2022.121529. [DOI] [Google Scholar]
  12. Qin C.; Hou B.; Wang J.; Wang Q.; Wang G.; Yu M.; Chen C.; Jia L.; Li D. Crystal-Plane-Dependent Fischer-Tropsch Performance of Cobalt Catalysts. ACS Catal. 2018, 8, 9447–9455. 10.1021/acscatal.8b01333. [DOI] [Google Scholar]
  13. Liu S.; Yang C.; Zha S.; Sharapa D.; Studt F.; Zhao Z.-J.; Gong J. Moderate Surface Segregation Promotes Selective Ethanol Production in CO2 Hydrogenation Reaction over CoCu Catalysts. Angew. Chem., Int. Ed. 2022, 134, e202109027 10.1002/ange.202109027. [DOI] [PubMed] [Google Scholar]
  14. Zijlstra B.; Broos R. J. P.; Chen W.; Oosterbeek H.; Filot I. A. W.; Hensen E. J. M. Coverage Effects in CO Dissociation on Metallic Cobalt Nanoparticles. ACS Catal. 2019, 9, 7365–7372. 10.1021/acscatal.9b01967. [DOI] [Google Scholar]
  15. Fan T.; Liu H.; Shao S.; Gong Y.; Li G.; Tang Z. Cobalt Catalysts Enable Selective Hydrogenation of CO2 toward Diverse Products: Recent Progress and Perspective. J. Phys. Chem. Lett. 2021, 12, 10486–10496. 10.1021/acs.jpclett.1c03043. [DOI] [PubMed] [Google Scholar]
  16. Liu B.; Ouyang B.; Zhang Y.; Lv K.; Li Q.; Ding Y.; Li J. Effects of mesoporous structure and Pt promoter on the activity of Co-based catalysts in low-temperature CO2 hydrogenation for higher alcohol synthesis. J. Catal. 2018, 366, 91–97. 10.1016/j.jcat.2018.07.019. [DOI] [Google Scholar]
  17. Efremova A.; Rajkumar T.; Szamosvölgyi Á.; SÁpi A.; BaÁn K.; Szenti I.; Gómez-Pérez J.; Varga G.; Kiss J.; Halasi G.; Kukovecz Á.; Kónya Z. Complexity of a Co3O4 System under Ambient-Pressure CO2 Methanation: Influence of Bulk and Surface Properties on the Catalytic Performance. J. Phys. Chem. C 2021, 125, 7130–7141. 10.1021/acs.jpcc.0c09717. [DOI] [Google Scholar]
  18. Yang C.; Liu S.; Wang Y.; Song J.; Wang G.; Wang S.; Zhao Z.-J.; Mu R.; Gong J. The Interplay between Structure and Product Selectivity of CO2 Hydrogenation. Angew. Chem., Int. Ed. 2019, 58, 11242–11247. 10.1002/anie.201904649. [DOI] [PubMed] [Google Scholar]
  19. Behler J. Four Generations of High-Dimensional Neural Network Potentials. Chem. Rev. 2021, 121, 10037–10072. 10.1021/acs.chemrev.0c00868. [DOI] [PubMed] [Google Scholar]
  20. Zhang J.; Lei Y. K.; Zhang Z.; Chang J.; Li M.; Han X.; Yang L.; Yang Y. I.; Gao Y. Q. A Perspective on Deep Learning for Molecular Modeling and Simulations. J. Phys. Chem. A 2020, 124, 6745–6763. 10.1021/acs.jpca.0c04473. [DOI] [PubMed] [Google Scholar]
  21. Ma S.; Liu Z.-P. Machine Learning for Atomic Simulation and Activity Prediction in Heterogeneous Catalysis: Current Status and Future. ACS Catal. 2020, 10, 13213–13226. 10.1021/acscatal.0c03472. [DOI] [Google Scholar]
  22. Shi X.; Lin X.; Luo R.; Wu S.; Li L.; Zhao Z.-J.; Gong J. Dynamics of Heterogeneous Catalytic Processes at Operando Conditions. JACS Au 2021, 1, 2100–2120. 10.1021/jacsau.1c00355. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Cheng D.; Zhao Z.-J.; Zhang G.; Yang P.; Li L.; Gao H.; Liu S.; Chang X.; Chen S.; Wang T.; Ozin G. A.; Liu Z.-P.; Gong J. The nature of active sites for carbon dioxide electroreduction over oxide-derived copper catalysts. Nat. Commun. 2021, 12, 395. 10.1038/s41467-020-20615-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. ten Have I. C.; Kromwijk J. J. G.; Monai M.; Ferri D.; Sterk E. B.; Meirer F.; Weckhuysen B. M. Uncovering the reaction mechanism behind CoO as active phase for CO2 hydrogenation. Nat. Commun. 2022, 13, 324. 10.1038/s41467-022-27981-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Wang M.; Zhang G.; Zhu J.; Lia W.; Wang J.; Bian K.; Liu Y.; Ding F.; Song C.; Guo X. Unraveling the Tunable Selectivity on Cobalt Oxide and Metallic Cobalt Sites for CO2 Hydrogenation. Chem. Eng. J. 2022, 446, 137217 10.1016/j.cej.2022.137217. [DOI] [Google Scholar]
  26. Ko J.; Kim B. K.; Han J. W. Density Functional Theory Study for Catalytic Activation and Dissociation of CO2 on Bimetallic Alloy Surfaces. J. Phys. Chem. C 2016, 120, 3438–3447. 10.1021/acs.jpcc.6b00221. [DOI] [Google Scholar]
  27. Xu D.; Wang Y.; Ding M.; Hong X.; Liu G.; Tsang S. C. E. Advances in Higher Alcohol Synthesis from CO2 Hydrogenation. Chem 2021, 7, 849–881. 10.1016/j.chempr.2020.10.019. [DOI] [Google Scholar]
  28. Chen C.; Wang Q.; Wang G.; Hou B.; Jia L.; Li D. Mechanistic Insight into the C2 Hydrocarbons Formation from Syngas on fcc-Co(111) Surface: A DFT Study. J. Phys. Chem. C 2016, 120, 9132–9147. 10.1021/acs.jpcc.5b09634. [DOI] [Google Scholar]
  29. Banerjee A.; van Bavel A. P.; Kuipers H. P. C. E.; Saeys M. CO Activation on Realistic Cobalt Surfaces: Kinetic Role of Hydrogen. ACS Catal. 2017, 7, 5289–5293. 10.1021/acscatal.7b00846. [DOI] [Google Scholar]
  30. Kattel S.; Liu P.; Chen J. G. Tuning Selectivity of CO2 Hydrogenation Reactions at the Metal/Oxide Interface. J. Am. Chem. Soc. 2017, 139, 9739–9754. 10.1021/jacs.7b05362. [DOI] [PubMed] [Google Scholar]
  31. Vogt C.; Monai M.; Kramer G. J.; Weckhuysen B. M. The renaissance of the Sabatier reaction and its applications on Earth and in space. Nat. Catal. 2019, 2, 188–197. 10.1038/s41929-019-0244-4. [DOI] [Google Scholar]
  32. Ye R.-P.; Ding J.; Gong W.; Argyle M. D.; Zhong Q.; Wang Y.; Russell C. K.; Xu Z.; Russell A. G.; Li Q.; Fan M.; Yao Y.-G. CO2 hydrogenation to high-value products via heterogeneous catalysis. Nat. Commun. 2019, 10, 5698. 10.1038/s41467-019-13638-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Roy S.; Cherevotan A.; Peter S. C. Thermochemical CO2 Hydrogenation to Single Carbon Products: Scientific and Technological Challenges. ACS Energy Lett. 2018, 3, 1938–1966. 10.1021/acsenergylett.8b00740. [DOI] [Google Scholar]
  34. Ocampo-Restrepo V. K.; Verga L. G.; Da Silva J. L. F. Ab Initio Study of the C-O Bond Dissociation in CO2 Reduction by Redox and Carboxyl Routes on 3d Transition Metal Systems. J. Phys. Chem. C 2021, 125, 26296–26306. 10.1021/acs.jpcc.1c05468. [DOI] [Google Scholar]
  35. Stegelmann C.; Andreasen A.; Campbell C. T. Degree of Rate Control: How Much the Energies of Intermediates and Transition States Control Rates. J. Am. Chem. Soc. 2009, 131, 8077–8082. 10.1021/ja9000097. [DOI] [PubMed] [Google Scholar]
  36. Xie X.; Li Y.; Liu Z.-Q.; Haruta M.; Shen W. Low-temperature oxidation of CO catalysed by Co3O4 nanorods. Nature 2009, 458, 746–749. 10.1038/nature07877. [DOI] [PubMed] [Google Scholar]
  37. Liotta L. F.; Carlo G. D.; Pantaleo G.; Venezia A. M.; Deganello G. Co3O4/CeO2 composite oxides for methane emissions abatement: Relationship between Co3O4-CeO2 interaction and catalytic activity. Appl. Catal. B 2006, 66, 217–227. 10.1016/j.apcatb.2006.03.018. [DOI] [Google Scholar]
  38. Biesinger M. C.; Payne B. P.; Grosvenor A. P.; Lau L. W. M.; Gerson A. R.; Smart R. S. Resolving surface chemical states in XPS analysis of first row transition metals, oxides and hydroxides: Cr, Mn, Fe, Co and Ni. Appl. Surf. Sci. 2011, 257, 2717–2730. 10.1016/j.apsusc.2010.10.051. [DOI] [Google Scholar]
  39. Chen Z.; Kronawitter C. X.; Koel B. E. Facet-dependent activity and stability of Co3O4 nanocrystals towards the oxygen evolution reaction. Phys. Chem. Chem. Phys. 2015, 17, 29387–29393. 10.1039/C5CP02876K. [DOI] [PubMed] [Google Scholar]
  40. Wu C.; Lin L.; Liu J.; Zhang J.; Zhang F.; Zhou T.; Rui N.; Yao S.; Deng Y.; Yang F.; Xu W.; Luo J.; Zhao Y.; Yan B.; Wen X.-D.; Rodriguez J. A.; Ma D. Inverse ZrO2/Cu as a highly efficient methanol synthesis catalyst from CO2 hydrogenation. Nat. Commun. 2020, 11, 5767. 10.1038/s41467-020-19634-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Wang Y.; Kattel S.; Gao W.; Li K.; Liu P.; Chen J. G.; Wang H. Exploring the ternary interactions in Cu-ZnO-ZrO2 catalysts for efficient CO2 hydrogenation to methanol. Nat. Commun. 2019, 10, 1166. 10.1038/s41467-019-09072-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Kelly T. G.; Stottlemyer A. L.; Ren H.; Chen J. G. Comparison of O–H, C–H, and C–O Bond Scission Sequence of Methanol on Tungsten Carbide Surfaces Modified by Ni, Rh, and Au. J. Phys. Chem. C 2011, 115, 6644–6650. 10.1021/jp112006v. [DOI] [Google Scholar]
  43. Kattel S.; Yu W.; Yang X.; Yan B.; Huang Y.; Wan W.; Liu P.; Chen J. G. CO2 Hydrogenation over Oxide-Supported PtCo Catalysts: The Role of the Oxide Support in Determining the Product Selectivity. Angew. Chem., Int. Ed. 2016, 55, 7968–7973. 10.1002/anie.201601661. [DOI] [PubMed] [Google Scholar]
  44. Zheng Y.; Li K.; Wang H.; Wang Y.; Tian D.; Wei Y.; Zhu X.; Zeng C.; Luo Y. Structure dependence and reaction mechanism of CO oxidation: A model study on macroporous CeO2 and CeO2-ZrO2 catalysts. J. Catal. 2016, 344, 365–377. 10.1016/j.jcat.2016.10.008. [DOI] [Google Scholar]
  45. Konsolakis M.; Ioakimidis Z.; Kraia T.; Marnellos G. E. Hydrogen Production by Ethanol Steam Reforming (ESR) over CeO2 Supported Transition Metal (Fe, Co, Ni, Cu) Catalysts: Insight into the Structure-Activity Relationship. Catalysts 2016, 6, 39. 10.3390/catal6030039. [DOI] [Google Scholar]
  46. Pawlak D. A.; Ito M.; Oku M.; Shimamura K.; Fukuda T. Interpretation of XPS O (1s) in Mixed Oxides Proved on Mixed Perovskite Crystals. J. Phys. Chem. B 2002, 106, 504–507. 10.1021/jp012040a. [DOI] [Google Scholar]
  47. Huang S.-D.; Shang C.; Kang P. L.; Zhang X.-J.; Liu Z.-P. Material discovery by combining stochastic surface walking global optimization with a neural network. Chem. Sci. 2017, 8, 6327–6337. 10.1039/C7SC01459G. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Kresse G.; Furthmüller J. Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set. Comp. Mater. Sci. 1996, 6, 15–50. 10.1016/0927-0256(96)00008-0. [DOI] [PubMed] [Google Scholar]
  49. Kresse G.; Hafner J. Ab initio molecular-dynamics simulation of the liquid-metal-amorphous-semiconductor transition in germanium. Phys. Rev. B 1994, 49, 14251–14269. 10.1103/PhysRevB.49.14251. [DOI] [PubMed] [Google Scholar]
  50. Perdew J. P.; Burke K.; Ernzerhof M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 1996, 77, 3865–3868. 10.1103/PhysRevLett.77.3865. [DOI] [PubMed] [Google Scholar]
  51. Grimme S.; Ehrlich S.; Goerigk L. Effect of the damping function in dispersion corrected density functional theory. J. Comput. Chem. 2011, 32, 1456–1465. 10.1002/jcc.21759. [DOI] [PubMed] [Google Scholar]
  52. Kresse G.; Joubert D. From ultrasoft pseudopotentials to the projector augmented-wave method. Phys. Rev. B 1999, 59, 1758–1775. 10.1103/PhysRevB.59.1758. [DOI] [Google Scholar]
  53. Kulika H. J.; Marzari N. Systematic study of first-row transition-metal diatomic molecules: A self-consistent DFT+U approach. J. Chem. Phys. 2010, 133, 114103. 10.1063/1.3489110. [DOI] [PubMed] [Google Scholar]
  54. Youmbi B. S.; Calvayrac F. Structure of CoO(001) surface from DFT + U calculations. Surf. Sci. 2014, 621, 1–6. 10.1016/j.susc.2013.10.012. [DOI] [Google Scholar]
  55. Wang L.; Maxisch T.; Ceder G. Oxidation energies of transition metal oxides within the GGA+U Framework. Phys. Rev. B 2006, 73, 195107 10.1103/PhysRevB.73.195107. [DOI] [Google Scholar]
  56. Andersen M.; Panosetti C.; Reuter K. A Practical Guide to Surface Kinetic Monte Carlo Simulations. Front. Chem. 2019, 7, 202. 10.3389/fchem.2019.00202. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Stamatakis M.; Vlachos D. G. Unraveling the Complexity of Catalytic Reactions via Kinetic Monte Carlo Simulation: Current Status and Frontiers. ACS Catal. 2012, 2, 2648–2663. 10.1021/cs3005709. [DOI] [Google Scholar]
  58. Stamatakis M.; Vlachos D. G. A graph-theoretical kinetic Monte Carlo framework for on-lattice chemical kinetics. J. Chem. Phys. 2011, 134, 214115. 10.1063/1.3596751. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

au2c00632_si_001.pdf (3.8MB, pdf)

Articles from JACS Au are provided here courtesy of American Chemical Society

RESOURCES