Understanding the structure-performance relationship of cubic In2O3 catalysts for CO2 hydrogenation
Graphical abstract
Introduction
Methanol (CH3OH) synthesis form CO2 hydrogenation has attracted increasing attention for its importance in the CO2 utilization strategy in recent decades. Nobel Laureate George Olah once proposed the “methanol economy” [1], according to which the conversion of CO2 to CH3OH using hydrogen from renewable energy resources may eventually relieve our heavy reliance on fossil fuels. CH3OH synthesis form CO2 hydrogenation involves two key reactions. The main reaction is CH3OH formation from CO2 (CO2 + 3H2 → CH3OH + H2O), which is exothermic (ΔrHɵ: −49.50 kJ mol−1) at the room temperature [2]. The competing side reaction is the reverse water-gas shift (RWGS) reaction (CO2 + H2 → CO + H2O), which is endothermic (ΔrHɵ: +41.20 kJ mol−1) at the room temperature [2,3]. Thus, thermodynamic consideration suggests that the low temperature makes CH3OH formation via CO2 hydrogenation possible.
Cu- and Pd-based catalysts were widely studied for CH3OH synthesis form CO2 hydrogenation, based on a recent analysis of the literature [4]. Under industrially relevant reaction conditions (T = 493−573 K, P = 50−100 bar), the Cu-ZnO-Al2O3 catalyst is used to convert CO2-containing syngas (CO/CO2/H2) to CH3OH. Behrens et al. [5] identified the crucial atomic level structure motif for the Cu-ZnO-Al2O3 catalyst by combined experiments and density functional theory (DFT) calculations, and Cu steps decorated with Zn atoms were identified as the active sites, which were stabilized by surface species and bulk defects. Kattel et al. [6] investigated the active sites of the Cu-ZnO catalyst for CH3OH synthesis form CO2 hydrogenation by combining experimental characterizations, DFT calculations as well as kinetic Monte Carlo simulations. They compared the catalytic activities of the Zn/Cu bimetallic and ZnO/Cu oxide-on-metal model catalysts, and found that the surface Zn on the Zn/Cu catalyst can be transformed to ZnO, and the Zn/Cu reached the same level of catalytic activity as the ZnO/Cu with the same Zn coverage. Thus, their studies suggest the crucial role of the synergy between Cu and ZnO at the interface in this reaction. Although the Cu-ZnO-Al2O3 catalyst is widely adopted for the industrial production of CH3OH from CO2-containing syngas, they have several disadvantages when directly applied to CH3OH synthesis from CO2 hydrogenation. Firstly, water formed during the reaction severely hinders the CO2 hydrogenation reaction at the active site, as it induces sintering of the active phase, and easily deactivates the catalyst [7]. Secondly, in the industrially relevant temperature range of 493−573 K, a large amount of undesired CO is produced as the byproduct from the RWGS reaction on this catalyst, which greatly limits the CH3OH selectivity [8]. Thus, superior water-resistant and RWGS-suppressing catalysts are needed for this reaction.
In recent years, higher CH3OH selectivity in the CO2 hydrogenation reaction was achieved by using indium oxide (In2O3) based catalysts. Two research groups [9,10] investigated the reaction mechanism of CH3OH synthesis from CO2 hydrogenation on the pure In2O3 surface using DFT calculations and reached an agreement that the reaction mechanism of CH3OH formation on the defective In2O3 surface occurred via the formate (HCOO) route, where CO2 was stepwise hydrogenated into the formate intermediate (HCOO*), dioxymethylene intermediate (H2COO*), formaldehyde intermediate (CH2O*), methoxy intermediate (CH3O*), and CH3OH. Martin et al. [11] showed experimentally that the In2O3 catalyst could indeed be used for CO2 hydrogenation to methanol, and methanol selectivity reaches 100 % using the In2O3/ZrO2 mixed-oxide catalyst, which has a remarkable stability under industrially relevant conditions (T = 473−573 K, P = 1−5 MPa, GHSV = 16000−48,000 cm3 h−1gcat-1 and H2/CO2/Ar = 4/1/1.5). Further experimental studies of Sun et al. [12] showed that CO2 conversion increases at increasing temperature and pressure, although the formation rate and selectivity of CH3OH first increase with the temperature, and then reach the maximum at 603 K and decrease thereafter. At T =603 K, P =4 MPa, GHSV = 15,000 cm3 h−1gcat−1 and H2/CO2/N2 = 3/1/1, CO2 conversion and CH3OH selectivity reach 7.13 % and 39.7 %, respectively, over the In2O3 catalyst.
Previously, our research group developed bifunctional catalysts composed of In2O3 (or In2O3/ZrO2 composite oxide) and zeolite (H-form Zeolite Socony Mobile-5 and SAPO-34) for gasoline (C5°-C11°) synthesis and lower olefins (C2=-C4=) synthesis directly from CO2 hydrogenation [13,14]. We showed that the CO2 hydrogenation reaction on the In2O3/ZrO2 composite oxide can proceed at the relatively high reaction temperature of ∼673 K, and doping Zr into In2O3 helps to stabilize the key adsorbates at the Ov site near the Zr dopant [14]. For CH3OH synthesis form CO2 hydrogenation, compared with the In2O3/ZrO2 mixed-oxide catalyst, the pure In2O3 catalyst results in lower CO2 conversion rate as well as inferior CH3OH selectivity at higher reaction temperature. Recent experimental and computational studies further suggest that the product selectivity strongly depends on the crystal phase and surface morphology for CO2 hydrogenation [15,16]. However, due to the presence of multiple distinct Ov sites on the different surface terminations, the precise relationship between the structure of the In2O3 catalyst and its catalytic performance in the CO2 hydrogenation reaction remains to be firmly established, which is vital for the development of more efficient oxide catalysts for this reaction.
In this work, the reaction mechanism and catalytic activities of all the different surface oxygen vacancy (Ov) sites on the stable cubic In2O3 (c-In2O3) surfaces for CH3OH and CO formation from CO2 hydrogenation were thoroughly investigated in order to establish the structure-performance relationship by performing data science analysis using several simple and interpretable classification algorithms. The (111) flat and (110) flat ((111)-f and (110)-f) surfaces, as well as the (110) step ((110)-s) surface of c-In2O3 were considered, which are the most stable surfaces of the most stable phase of In2O3, as shown by recent experiments [17,18], DFT calculations [[19], [20], [21]] and detailed structural characterization of the c-In2O3 catalysts in our recent work [16]. This study shows that data science tools are very useful for learning from a large amount of computational data and can significantly improve our understanding of the structure-performance relationship of complex heterogeneous catalysts.
Section snippets
Computational methods
We performed periodic DFT calculations using the Vienna ab initio simulation package (VASP) [22,23] with the Perdew-Burke-Ernzerhof (PBE) exchange-correlation functional [24] and the projector-augmented wave (PAW) potentials, which account for the valence-core electron interactions [25]. The detailed calculation parameters on this system in this work were identical to those in our previous works. [13,14,16,26,27].
The optimized primitive unit cell based on our previous work [13] was used to
Selection of surface oxygen vacancies
The 12 surface Ov sites from Ov1 to Ov12 on the perfect (111)-f surface (Figure S1b) are classified into four types from high to low Ef,Ov: (1) Ov7, Ov9, Ov11, (2) Ov4, Ov10, Ov12, (3) Ov2, Ov5, Ov6, and (4) Ov1, Ov3, Ov8, according to our previous work [26]. In our initial calculations of the reaction mechanism, we chose the most types of Ov sites with the highest and the lowest Ov formation energies. Thus, the Ov7 site (Figure S1c) with the highest formation energy and the Ov3 site with the
Conclusions
Reaction mechanism and catalytic activities for CH3OH and CO formations from CO2 hydrogenation at all the different surface Ov sites on the stable (111)-flat surface and (110)-f and (110)-s surfaces were thoroughly investigated in order to establish the structure-performance relationship by combining DFT calculations and data science analysis. Both bi−HCOO* hydrogenation to the H2COO* (R2) and H2COO* dissociation to the CH2O* with surface O atom (R3) can be the actual RDS for CH3OH formation,
CRediT authorship contribution statement
Bin Qin: Conceptualization, Formal analysis, Investigation, Visualization, Writing - original draft, Writing - review & editing. Zhimin Zhou: Software, Formal analysis. Shenggang Li: Conceptualization, Resources, Writing - original draft, Writing - review & editing, Supervision, Project administration, Funding acquisition. Peng Gao: Project administration, Funding acquisition.
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.
Acknowledgements
This work was supported by Shell Global Solutions B.V. (CW373032); Strategic Priority Research Program of the Chinese Academy of Sciences (XDA21090204); and Youth Innovation Promotion Association CAS (2018330).
References (34)
- et al.
Insight into methanol synthesis from CO2 hydrogenation on Cu(111): Complex reaction network and the effects of H2O
J. Catal.
(2011) - et al.
Kinetics of deactivation on Cu/ZnO/Al2O3 methanol synthesis catalysts
Appl. Catal. A-Gen.
(2015) - et al.
The role of copper oxidation state in Cu/ZnO/Al2O3 catalysts in CO2 hydrogenation and methanol productivity
Renew. Energ.
(2019) - et al.
Mechanism and microkinetics of methanol synthesis via CO2 hydrogenation on indium oxide
J. Catal.
(2018) - et al.
Hydrogenation of CO2 to methanol over In2O3 catalyst
J. CO2 Util.
(2015) - et al.
Synthesis and characterization of high-quality In2O3 nanobelts via catalyst-free growth using a simple physical vapor deposition at low temperature
Chem. Phys. Lett.
(2004) - et al.
Shape- and phase-controlled synthesis of In2O3 with various morphologies and their gas-sensing properties
Sensor. Actuat. B-Chem.
(2009) - et al.
Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set
Comp. Mater. Sci.
(1996) After oil and gas: methanol economy
Catal. Lett.
(2004)- et al.
Conversion of CO2 to dimethylether and methanol over hybrid catalysts
Chem. Lett.
(1992)
Challenges in the greener production of formates/formic acid, methanol, and DME by heterogeneously catalyzed CO2 hydrogenation processes
Chem. Rev.
The active site of methanol synthesis over Cu/ZnO/Al2O3 industrial catalysts
Science
CATALYSIS Active sites for CO2 hydrogenation to methanol on Cu/ZnO catalysts
Science
Active oxygen vacancy site for methanol synthesis from CO2 hydrogenation on In2O3(110): A DFT study
ACS Catal.
Indium oxide as a superior catalyst for methanol synthesis by CO2 hydrogenation
Angew. Chem. Int. Ed.
Direct conversion of CO2 into liquid fuels with high selectivity over a bifunctional catalyst
Nat. Chem.
Direct production of lower olefins from CO2 conversion via bifunctional catalysis
ACS Catal.
Cited by (10)
Coverage effect of surface oxygen vacancy on In<inf>2</inf>O<inf>3</inf>-catalyzed CO<inf>2</inf> hydrogenation revealed by first principles-based microkinetic simulations
2023, Computational and Theoretical ChemistryDirect carbon dioxide hydrogenation to produce bulk chemicals and liquid fuels via heterogeneous catalysis
2022, Chinese Journal of CatalysisCitation Excerpt :Therefore, the design and development of more efficient In2O3 catalysts are necessary for their industrial use in CO2 hydrogenation reactions. In rationally designing high-performance In2O3 nanocatalysts, we first performed extensive density functional theory (DFT) calculations to understand the reaction mechanism of methanol synthesis and CO formation via the RWGS reaction [32,33]. We found that the cubic In2O3(110) surface can form two distinct CO2 adsorption structures that correspond to the reaction paths of CO2 hydrogenation, which lead to the production of methanol and the release of the byproduct CO.
Mechanistic studies toward the rational design of oxide catalysts for carbon dioxide hydrogenation
2021, Annual Reports in Computational ChemistryCitation Excerpt :The rise of data science and machine learning approaches in recent decades provides important opportunities for the computer-assisted design of heterogeneous catalysts, for example via the recently proposed catalysis informatics approach (105,106). Our limited experience in this research area (69,102) shows that the curation of a sufficient amount of quality experimental or computational data is very costly and time-consuming, whereas the finding of a suitable machine learning algorithm for training is often not the bottleneck. Thus, although high-throughput automated experimentation and computation are important directions for future research (107,108), mechanistic understanding from joint experimental and computational studies remains vital in our search for more efficient heterogeneous catalysts of industrial relevance.