Recent Advances of Indium Oxide-Based Catalysts for CO2 Hydrogenation to Methanol: Experimental and Theoretical

Methanol synthesis from the hydrogenation of carbon dioxide (CO2) with green H2 has been proven as a promising method for CO2 utilization. Among the various catalysts, indium oxide (In2O3)-based catalysts received tremendous research interest due to the excellent methanol selectivity with appreciable CO2 conversion. Herein, the recent experimental and theoretical studies on In2O3-based catalysts for thermochemical CO2 hydrogenation to methanol were systematically reviewed. It can be found that a variety of steps, such as the synthesis method and pretreatment conditions, were taken to promote the formation of oxygen vacancies on the In2O3 surface, which can inhibit side reactions to ensure the highly selective conversion of CO2 into methanol. The catalytic mechanism involving the formate pathway or carboxyl pathway over In2O3 was comprehensively explored by kinetic studies, in situ and ex situ characterizations, and density functional theory calculations, mostly demonstrating that the formate pathway was extremely significant for methanol production. Additionally, based on the cognition of the In2O3 active site and the reaction path of CO2 hydrogenation over In2O3, strategies were adopted to improve the catalytic performance, including (i) metal doping to enhance the adsorption and dissociation of hydrogen, improve the ability of hydrogen spillover, and form a special metal-In2O3 interface, and (ii) hybrid with other metal oxides to improve the dispersion of In2O3, enhance CO2 adsorption capacity, and stabilize the key intermediates. Lastly, some suggestions in future research were proposed to enhance the catalytic activity of In2O3-based catalysts for methanol production. The present review is helpful for researchers to have an explicit version of the research status of In2O3-based catalysts for CO2 hydrogenation to methanol and the design direction of next-generation catalysts.


Introductions
The greenhouse effect caused by excessive CO 2 emission has seriously threatened the survival of human beings and other organisms [1][2][3][4]. In order to cope with the current grim situation, many countries have established a target timeline to reach the peak of CO 2 emission and achieve carbon neutrality. For example, China has promised to realize a carbon peak by 2030 and carbon neutrality by 2060 [5][6][7][8]. Therefore, CO 2 capture and utilization (CCU) technology has attracted much attention [9][10][11][12][13]. In particular, the use of "green hydrogen" produced with renewable energy to convert waste CO 2 into methanol is not only able to effectively reduce CO 2 emission but also can store renewable energy in liquid fuel, which is an important method to realize resource utilization of CO 2 [14][15][16][17][18][19].

Pure In2O3 Catalyst
The idea of In2O3 as the catalyst for CO2 hydrogenation to methanol stems from its excellent CO2 selectivity in methanol steam reforming (MSR) reactions [56,57]. Based on density functional theory (DFT), Ge et al. [53] predicted the feasibility of CO2 hydrogenation to methanol catalyzed by In2O3 (110) with oxygen vacancies. They proposed that In2O3 would inhibit RWGS reaction, and methanol was the major product on the surface of defective In2O3 (110). As shown in Figure 1, the reaction process obeys the mechanism of periodic generation and annihilation of oxygen vacancies, including adsorption and activation of CO2 on oxygen vacancies, CO2 hydrogenation to form intermediate species, methanol desorption, and regeneration of oxygen vacancies. To confirm the research results of DFT, Liu et al. [38] used commercial In2O3 activated at a high temperature (500 °C) as the catalyst for CO2 hydrogenation. The experimental results demonstrated that methanol yield increased with the increase in reaction pressure; however, due to the limitation of thermodynamics, it increased first and then decreased as the temperature increased. In addition, 2.82% of methanol yield and 3.69 mol h −1 kgcat −1 of methanol production rate were obtained at 330 °C and 4 MPa, which was superior to many other catalysts. In 2016, Pérez-Ramírez et al. [58] revealed that nano In2O3 can efficiently catalyze CO2 hydrogenation to methanol, obtaining more than 0.18 gMeOH h −1 gcat −1 of space-time yield. They also found that compared to Cu/ZnO/Al2O3, the highest methanol yield of In2O3 was achieved at 300 °C, indicating that In2O3 can maintain high methanol selectivity at higher temperatures. Two years later, they reported the mechanism and microkinetics of methanol synthesis from CO2 hydrogenation over In2O3 [59]. The results indicated that the apparent activation energy experimentally determined for CO2 hydrogenation to methanol (103 kJ mol −1 ) was lower than that of the RWGS reaction (117 kJ mol −1 ), which explains the superior methanol selectivity over In2O3. In2O3 (111) was experimentally and theoretically proved to be the most exposed surface termination, indicating CO2 can be activated by oxygen vacancies surrounded by three indium atoms. In addition, the most favorable pathway to methanol comprises three consecutive additions of hydrides and protons, which features CH2OOH* and CH2(OH)2* as intermediates. In 2019, by an operando examination, Müller et al. [60] proved that In2O3-x was the active phase of methanol synthesis, while In 0 led to the deactivation of the catalyst. The active site and catalytic mechanism of CO2 hydrogenation to methanol over defective In2O3 (110). Reproduced with permission from ref. [53]. Copyright 2013 American Chemical Society. Figure 1. The active site and catalytic mechanism of CO 2 hydrogenation to methanol over defective In 2 O 3 (110). Reproduced with permission from ref. [53]. Copyright 2013 American Chemical Society.
Liu et al. [61] prepared an In 2 O 3 catalyst by precipitation method for CO 2 hydrogenation to methanol. The results showed that under the operating conditions (H 2 /CO 2 molar ratio of 4, the volume space velocity of 21,000 cm 3 h −1 g cat −1 , reaction pressure of 5 MPa, and reaction temperature of 300 • C), the CO 2 conversion and methanol space-time yield were 9.4% and 0.335 g MeOH h −1 g cat −1 . Guo et al. [62] investigated the catalytic activity of cubic bixbyite-type indium oxide (c-In 2 O 3 ) and rhombohedral corundum-type indium oxide (r-In 2 O 3 ) in CO 2 hydrogenation to methanol. Due to the impressive reducibility and reactivity, c-In 2 O 3 was higher than r-In 2 O 3 in CO 2 conversion; however, r-In 2 O 3 possessed higher methanol selectivity because of weaker methanol and stronger CO adsorption. Moreover, the in situ DRIFTS experiments revealed that CO 2 could be reduced to CO via redox cycling and hydrogenated to methanol via the formate pathway. In addition, Sun et al. [63] successfully designed an In 2 O 3 nanocatalyst with higher catalytic activity under the guidance of theoretical calculation, which suggested that the hexagonal In 2 O 3 (104) surface had a far superior catalytic performance. As shown in Figure 2, the experimental results also confirmed that compared to cubic In 2 O 3 (c-In 2 O 3 ), a novel hexagonal In 2 O 3 (h-In 2 O 3 -R) with a high proportion of the exposed (104) surface exhibited higher catalytic activity and possessed high stability. Moreover, Li et al. [54] investigated the dissociative adsorption of H 2 during CO 2 hydrogenation over cubic and hexagonal In 2 O 3 by DFT, and they found that the oppositely charged In and O pair sites on the reduced In 2 O 3 surfaces played a significant role in facilitating the heterolytic dissociation of H 2 , which contributed to the formation of anionic hydride around the In sites to promote CO 2 hydrogenation to methanol. Additionally, h-In 2 O 3 (104) surface is considered the best surface for CO 2 hydrogenation to methanol due to the facile formation of the oxygen vacancies at low coverage and the favorable formation of the hydride adsorbate at the In sites.
In the last two years, the research topic of CO 2 hydrogenation to methanol over In 2 O 3 has still attracted considerable interest. Based on the solvothermal method, Wu et al. [64] successfully fabricated mixed-phase indium oxide with controllable cubic and hexagonal phases to enhance catalytic performance in CO 2 hydrogenation to methanol. Due to its enhanced textural properties and oxygen vacancies, mixed-phase c/h-In 2 O 3 catalysts demonstrated higher CO 2 conversion and space-time yield of methanol and kept stable in the reaction. To understand the structure-activity relationship, Nørskov et al. [65] systematically studied the methanol synthesis over In 2 O 3 (111) and In 2 O 3 (110) by combining DFT calculations with microkinetic modeling. The theoretical activity volcano shown in Figure 3 suggested that catalytic activity was closely related to the number of reduced In layers on In 2 O 3 surfaces, specifically, for In 2 O 3 (110), a surface oxygen vacancies between 0.17 and 1 ML (ML: the top layer, from surface to interior) possessed the highest catalytic activity, while for In 2 O 3 (111), the number of oxygen vacancies should be increased to 1~5 ML to obtain the optimal activity. Similarly, Gao et al. [66] revealed the structureperformance relationship of cubic In 2 O 3 catalyst in CO 2 hydrogenation via the study of reaction mechanism and catalytic activities at all the different surface oxygen vacancy sites on stable (111) flat surface, (110) flat surface, and (110) step surface. The conclusion was that the rate-determining step of methanol synthesis for a given oxygen vacancy site can be determined by the stability of H 2 COO* and CH 2 O* intermediates along with the formation energy of the oxygen vacancy sites, and tri-coordinated oxygen vacancy sites were beneficial to the formation of methanol, whereas bi-coordinated oxygen vacancy sites favor CO formation. CO 2 hydrogenation to methanol on indium-terminated In 2 O 3 (100), defective In 2 O 3 (110), and In 2 O 3 (111) surfaces were also deeply investigated by Zhang et al. [67]. It was found that the adsorbed CO 2 was preferable to form HCOO* compared with CO* and COOH* and underwent HCOO*, H 2 CO*, and H 3 CO* intermediates due to the lowest energy barriers. The defective In 2 O 3 (110) was proven to be the optimal surface for CO 2 hydrogenation to methanol, while the indium-terminated In 2 O 3 (100) surface displayed the lowest catalytic activity. In addition, Creaser et al. [39] proposed a kinetic model based on Langmuir-Hinshelwood-Hougen-Watson (LHHW) mechanism for CO 2 hydrogenation to methanol over In 2 O 3 catalyst. The model revealed that RWGS was obviously enhanced at high temperatures, causing methanol synthesis to reverse (methanol steam reforming, MSR). Apparent activation energies for CO 2 hydrogenation to methanol and RWGS were 90 and 110 kJ mol −1 , respectively, over In 2 O 3 derived from the experimental data. The results obtained from these detailed investigations were conducive to the development of reliable reactor and process designs. at high temperatures, causing methanol synthesis to reverse (methanol steam reforming, MSR). Apparent activation energies for CO2 hydrogenation to methanol and RWGS were 90 and 110 kJ mol −1 , respectively, over In2O3 derived from the experimental data. The results obtained from these detailed investigations were conducive to the development of reliable reactor and process designs. Although In2O3 exhibited excellent methanol selectivity in CO2 hydrogenation, the low CO2 conversion limited the methanol yield. Therefore, based on the cognition of the In2O3 active site and the reaction pathway of CO2 hydrogenation over In2O3, two strategies shown in Figure 4 were adopted to enhance the performance of In2O3, including (I) introducing other metal elements into In2O3 and (II) combining In2O3 with other metal oxides. The catalytic performance of In2O3-based catalysts is summarized in Table 1. Although In 2 O 3 exhibited excellent methanol selectivity in CO 2 hydrogenation, the low CO 2 conversion limited the methanol yield. Therefore, based on the cognition of the In 2 O 3 active site and the reaction pathway of CO 2 hydrogenation over In 2 O 3 , two strategies shown in Figure 4 were adopted to enhance the performance of In 2

Metal/In 2 O 3 Composite Catalysts
The abundant oxygen vacancies in In 2 O 3 can adsorb and activate CO 2 , and the periodic generation and annihilation of oxygen vacancies can inhibit the side reactions, leading to the highly selective conversion of CO 2 to methanol. However, the weak hydrogen adsorption and dissociation of In 2 O 3 limit the hydrogenation of carbon species, so CO 2 conversion is very low. Accordingly, the introduction of a noble metal or transition metal (M) could improve CO 2 conversion due to the synergistic catalysis of M and In 2 O 3 . As shown in Figure 5, the H 2 molecule was adsorbed and activated on the M surface to generate H active species (step 1 ) and then combined with lattice oxygen of In 2 O 3 via spillover (step 2 ) to create the oxygen vacancies (step 3 ). CO 2 molecule was adsorbed and activated by the obtained oxygen vacancies (step 4 ) and finally hydrogenated to methanol by combining with H active species (step 5 ).
(M) could improve CO2 conversion due to the synergistic catalysis of M and In2O3. As shown in Figure 5, the H2 molecule was adsorbed and activated on the M surface to generate H active species (step ①) and then combined with lattice oxygen of In2O3 via spillover (step ②) to create the oxygen vacancies (step ③). CO2 molecule was adsorbed and activated by the obtained oxygen vacancies (step ④) and finally hydrogenated to methanol by combining with H active species (step ⑤).

Noble Metal/In 2 O 3 Catalysts
Pd/In 2 O 3 catalyst. Many investigations have been concentrated on Pd/In 2 O 3 catalyst for CO 2 hydrogenation to methanol in recent years. Ge et al. [87] studied methanol synthesis from CO 2 hydrogenation over Pd/In 2 O 3 by the DFT method. They found that the HCOO* route competes with the RWGS route over Pd/In 2 O 3 in the reaction process, and H 2 COO* + H* H 2 CO* + OH* and cis-COOH* + H* CO* + H 2 O* were their rate-limiting steps, respectively. The HCOO* route was the major pathway for methanol synthesis from CO 2 hydrogenation. Moreover, the H adatom activated by the Pd cluster and H 2 O on the In 2 O 3 substrate was extremely significant for the promotion of methanol production, and the adsorbed hydroxyl on the interface of Pd/In 2 O 3 can induce the transformation of the Pd 4 cluster, which caused the change in final hydrogenation step. According to the guidance of the theoretical study, they prepared Pd/In 2 O 3 with high dispersion of Pd nanoparticles by thermal treatment of Pd-peptide composite/In 2 O 3 for methanol synthesis from CO 2 hydrogenation [68]. The prepared catalyst exhibited much more excellent activity than that of pure In 2 O 3 due to the better ability to adsorb and dissociate H 2 for hydrogenation steps and the formation of oxygen vacancies. As a result, such a catalyst was able to demonstrate 20% of CO 2 conversion, 70% of methanol selectivity, and 0.89 g MeOH h −1 g cat −1 of space-time yield (STY), respectively. Huang et al. [69] from our group detailly investigated the effect of strong metalsupport interaction between Pd and In 2 O 3 on the catalytic performance of CO 2 hydrogenation to methanol by adjusting the morphology of In 2 O 3 . The results indicated that the combination of Pd and hollow In 2 O 3 nanotubes derived from MIL-68(In) nanorod was more conducive to the methanol production compared with other morphologies of In 2 O 3 , which was due to more formation of Pd 2+ via electron transfer from Pd to the curved In 2 O 3 (222) to enhance H 2 adsorption and formation of surface oxygen vacancies. In addition, to prevent the formation of the In-Pd bimetallic phase that led to the quick deactivation of the catalyst, our group further developed TCPP(Pd)@MIL-68(In) as precursors to prepare Pd/In 2 O 3 [88]. Compared to PdCl 2 , TCPP(Pd) (metalloporphyrins) can be served as a capping agent for the growth of MIL-68(In) and a shuttle for transporting the Pd 2+ , thereby improving the dispersion of Pd during the process of calcination and reduction, and preventing excessive reduction to form In-Pd bimetallic phase. Both theoretical and experimental results indicated that the prepared Pd/In 2 O 3 possessed excellent thermodynamic selectivity for methanol. For the same purpose of reducing the formation of In-Pd alloy, Zhan et al. [89] from our group adopted rape pollen pretreated by hydrochloric acid as the biological template to fabricate hierarchically structured bio-In 2 O 3 and bio-In 2 O 3 /Pd, as shown in Figure 6. The results suggested that the pollen template with acid etching possessed a hollow cage-like structure and abundant functional groups (viz., -COOH and -NH 2 ) on the surface, which was conducive to the growth of In 2 O 3 with abundant superficial oxygen vacancies. Compared to the sample without acid pretreatment (bio-In 2 O 3 -0/Pd), bio-In 2 O 3 -15/Pd demonstrated a better ability to inhibit the formation of In-Pd alloy due to the more uniform In 2 O 3 spatial distribution to reduce the interaction between Pd and In 2 O 3 . In the following research, our group further developed bifunctional catalyst Pd/In 2 O 3 /H-ZSM-5 for dimethyl ether synthesis from CO 2 hydrogenation, whereby Pd/In 2 O 3 prepared by carbonized alginate templating favored CO 2 hydrogenation into methanol, and H-ZSM-5 favored methanol dehydration into dimethyl ether. Compared to commercial Pd/In 2 O 3 (Com-PdIn), microspherical-confined nano In 2 O 3 possessed more excellent texture properties to disperse the Pd nanoparticles, thus obtaining more than 450 g MeOH kg cat −1 h −1 of STY, whereas Com-PdIn only achieved 50.8450 g MeOH kg cat −1 h −1 of STY [90]. Pérez-Ramírez et al. [91] reported an effective coprecipitation method to incorporate isolated palladium atoms into an In2O3 lattice for forming low-nuclearity palladium clusters, which can overcome the selectivity and stability limitations associated with palladium nanoparticles. Additionally, to disperse the active components highly, Zhang et al. [92] employed the citric acid method to load In2O3 and Pd on SBA-15, respectively. It can be found that oxygen vacancies were promoted with increasing Pd amount. The as-prepared catalyst possessed excellent performance with 12.9% of CO2 conversion, 83.9% of methanol selectivity, and 1.1 × 10 −2 molMeOH h −1 gcat −1 of STY, which was due to the high dispersion of In2O3 and Pd nanoparticles on SBA-15, and the synergetic effect of H2 dissociation on Pd species and CO2 activation on In2O3. Moreover, Wu et al. [70] introduced Mn and Pd into In2O3 to improve the methanol selectivity and CO2 conversion. The results showed that Pd species were highly dispersed on the MnO/In2O3 due to the strong metalsupport interactions, and 1 wt% Pd/MnO/In2O3 exhibited excellent activity (240.6 gMeOH kgcat −1 h −1 of STY) and stability in CO2 hydrogenation.
Pt/In2O3 catalyst. The combination of Pt and In2O3 for CO2 hydrogenation to methanol has also been reported. For instance, Li et al. [71] adopted the coprecipitation method to synthesize Pt/In2O3 and investigated the effect of Pt content on the catalytic performance. They found that as Pt content increased, CO2 conversion increased, whereas meth-n+ Figure 6. Fabrication routes of supported bio-In 2 O 3-x /Pd catalysts. Reproduced with permission from ref. [89]. Copyright 2021 Elsevier.
Pérez-Ramírez et al. [91] reported an effective coprecipitation method to incorporate isolated palladium atoms into an In 2 O 3 lattice for forming low-nuclearity palladium clusters, which can overcome the selectivity and stability limitations associated with palladium nanoparticles. Additionally, to disperse the active components highly, Zhang et al. [92] employed the citric acid method to load In 2 O 3 and Pd on SBA-15, respectively. It can be found that oxygen vacancies were promoted with increasing Pd amount. The as-prepared catalyst possessed excellent performance with 12.9% of CO 2 conversion, 83.9% of methanol selectivity, and 1.1 × 10 −2 mol MeOH h −1 g cat −1 of STY, which was due to the high dispersion of In 2 O 3 and Pd nanoparticles on SBA-15, and the synergetic effect of H 2 dissociation on Pd species and CO 2 activation on In 2 O 3 . Moreover, Wu et al. [70] introduced Mn and Pd into In 2 O 3 to improve the methanol selectivity and CO 2 conversion. The results showed that Pd species were highly dispersed on the MnO/In 2 O 3 due to the strong metal-support interactions, and 1 wt% Pd/MnO/In 2 O 3 exhibited excellent activity (240.6 g MeOH kg cat −1 h −1 of STY) and stability in CO 2 hydrogenation.
Pt/In 2 O 3 catalyst. The combination of Pt and In 2 O 3 for CO 2 hydrogenation to methanol has also been reported. For instance, Li et al. [71] adopted the coprecipitation method to synthesize Pt/In 2 O 3 and investigated the effect of Pt content on the catalytic performance.
They found that as Pt content increased, CO 2 conversion increased, whereas methanol selectivity increased first and then decreased. The highly dispersed Pt n+ was embedded into the In 2 O 3 lattice to promote the formation of oxygen vacancies and contribute to CO 2 activation. In the reaction process, the unstable Pt n+ was reduced to Pt nanoparticle, and the stable Pt n+ kept the high dispersion. Both Pt n+ and Pt can activate H 2 , but the effect on the reaction was quite different; specifically, the highly dispersed Pt n+ was used as the Lewis acid site to promote H 2 dissociation for CO 2 hydrogenation to methanol, while Pt nanoparticles induced the RWGS reaction to decrease the methanol selectivity. Similarly, Liu et al. [61] supported Pt on In 2 O 3 to improve the methanol yield. The results showed that the CO 2 conversion and methanol yield over Pt/In 2 O 3 were 17.3% and 0.542 g MeOH h −1 g cat −1 at 300 • C, respectively (In 2 O 3 : 9.4% and 0.335 g MeOH h −1 g cat −1 ). As compared to In 2 O 3 , Pt/In 2 O 3 possessed more excellent catalytic stability, which was mainly due to the high dispersion of Pt nanoparticles and strong interaction between Pt and In 2 O 3 to inhibit the excessive reduction in In 2 O 3 . In addition, to keep the high dispersion of Pt, Pan et al. [93] synthesized Pt/film/In 2 O 3 catalyst shown in Figure 7 via the cold-plasma/peptide-assembly (CPPA) method. The prepared Pt/film/In 2 O 3 obtained 37.0% of CO 2 conversion and 62.6% of methanol selectivity at 30 • C and 0.1 MPa in a dielectric barrier discharge (DBD) plasma reactor. The film of the catalyst played significant roles in the improvement of catalytic performance, namely inhibiting the agglomeration of Pt nanoparticles and transferring the electrons from the catalyst to CO 2 . The results of this work provided a valuable reference for CO 2 hydrogenation to methanol at room temperature and pressure. Pérez-Ramírez et al. [94] highlighted flame spray pyrolysis as a synthesis platform to assess metal (Pt, Ni, Au, etc.) promotion in In 2 O 3 -based catalysts for CO 2 hydrogenation. Compared to Ni clusters or Au nanoparticles, the atomically dispersed and well-stabilized Pt had a more obvious promoting effect on In 2 O 3 for CO 2 hydrogenation to methanol. Moreover, DFT simulations further revealed that the high concentration of isolated Pt atoms could greatly enhance homolytic H 2 splitting and increase the availability of hydrides for C-H hydrogenation due to the formation In 3 Pt and In 2 Pt 2 ensembles, therefore facilitating methanol production. dispersion of Pt nanoparticles and strong interaction between Pt and In2O3 to inhibit the excessive reduction in In2O3. In addition, to keep the high dispersion of Pt, Pan et al. [93] synthesized Pt/film/In2O3 catalyst shown in Figure 7 via the cold-plasma/peptide-assembly (CPPA) method. The prepared Pt/film/In2O3 obtained 37.0% of CO2 conversion and 62.6% of methanol selectivity at 30 °C and 0.1 MPa in a dielectric barrier discharge (DBD) plasma reactor. The film of the catalyst played significant roles in the improvement of catalytic performance, namely inhibiting the agglomeration of Pt nanoparticles and transferring the electrons from the catalyst to CO2. The results of this work provided a valuable reference for CO2 hydrogenation to methanol at room temperature and pressure. Pérez-Ramírez et al. [94] highlighted flame spray pyrolysis as a synthesis platform to assess metal (Pt, Ni, Au, etc.) promotion in In2O3-based catalysts for CO2 hydrogenation. Compared to Ni clusters or Au nanoparticles, the atomically dispersed and well-stabilized Pt had a more obvious promoting effect on In2O3 for CO2 hydrogenation to methanol. Moreover, DFT simulations further revealed that the high concentration of isolated Pt atoms could greatly enhance homolytic H2 splitting and increase the availability of hydrides for C-H hydrogenation due to the formation In3Pt and In2Pt2 ensembles, therefore facilitating methanol production. Other noble metal/In2O3 catalyst. In addition to Pd and Pt, other noble metals were also introduced into In2O3 to promote catalytic performance. Shrotri et al. [73] found that methanol STY over In2O3-based catalyst can be improved from 0.18 gMeOH h −1 gcat −1 to 1.0 gMeOH h −1 gcat −1 after doping of Rh. This was because, on the one hand, Rh promoted the Other noble metal/In 2 O 3 catalyst. In addition to Pd and Pt, other noble metals were also introduced into In 2 O 3 to promote catalytic performance. Shrotri et al. [73] found that methanol STY over In 2 O 3 -based catalyst can be improved from 0.18 g MeOH h −1 g cat −1 to 1.0 g MeOH h −1 g cat −1 after doping of Rh. This was because, on the one hand, Rh promoted the dissociation of H 2 to lead to the formation of more oxygen vacancies on the In 2 O 3 surface. On the other hand, Rh was related to the production of formate species with a low activation barrier confirmed by DFT. Similarly, Liu et al. [72] also investigated the influence of Rh addition to In 2 O 3 on methanol production from CO 2 hydrogenation. They demonstrated that the existence of Rh can enhance the dissociative adsorption and spillover of hydrogen, which was instrumental in surface oxygen vacancies formation of In 2 O 3 and CO 2 activation, so the STY of 0.5448 g MeOH h −1 g cat −1 over Rh/In 2 O 3 was obtained while it was only 0.3402 g MeOH h −1 g cat −1 over In 2 O 3 . In addition, they also supported Ru [74], Au [75], Ir [76], and Ag [95] on the In 2 O 3 for CO 2 hydrogenation to methanol, and the results indicated that the catalytic activity could be enhanced to a great extent.

Ni/In 2 O 3 catalysts.
Recently, Ni/In 2 O 3 catalysts have also attracted wide attention in methanol production from CO 2 hydrogenation. In 2020, Liu et al. [77] prepared an In 2 O 3supported nickel catalyst (Ni/In 2 O 3 ) by a wet chemical reduction for CO 2 hydrogenation, and the results suggested that the highly dispersed Ni species can be used as active sites for hydrogen dissociation and spillover to contribute to the formation of oxygen vacancies and hydrogenation process. Therefore, the effective synergy of Ni sites and In 2 O 3 support resulted in superior catalytic performance, specifically, 18.47% of CO 2 conversion, more than 54% of methanol selectivity, and 0.55 g MeOH h −1 g cat −1 of STY at 300 • C and 5 MPa. Subsequently, to further understand the superior catalytic performance of Ni/In 2 O 3 , they investigated the synergistic effect of the metal-support interaction and interfacial oxygen vacancies on methanol synthesis via DFT calculation [96]. It was found that the interfacial oxygen vacancies were beneficial for boosting the CO 2 adsorption and charge transfer between the nickel species and indium oxide, synergistically promoting the selectivity of methanol. Simultaneously, among the three reaction pathways examined (formate pathway, CO hydrogenation, and RWGS pathway, respectively), the RWGS pathway was proven to be the most theoretically favored for methanol synthesis from CO 2 hydrogenation over Ni/In 2 O 3 , as shown in Figure 8. In addition to the above research work, they also introduced ZrO 2 into Ni/In 2 O 3 catalyst (Ni/In 2 O 3 -ZrO 2 ) for CO 2 hydrogenation to methanol [97]. The solid solution formed by ZrO 2 and In 2 O 3 can optimize and stabilize the oxygen vacancies of In 2 O 3 to avoid the excessive reduction in the bulk indium oxide, thus possessing a 43.2% increase in STY of methanol. Different from the traditional synthesis method, Hensen et al. [78] combined Ni with In 2 O 3 using flame spray pyrolysis (FSP) synthesis. The obtained NiO-In 2 O 3 catalyst possesses high specific surface areas and block morphology. When NiO loading is 6 wt%,~0.25 g MeOH h −1 g cat −1 of STY can be obtained over the corresponding catalyst at the conditions of 250 • C and 30 bar. The comprehensive characterizations revealed the strong interactions between Ni cations and In 2 O 3 when NiO loading is lower 6 wt%, which contributed to the promotion of surface density of oxygen vacancies. Additionally, DFT calculation suggested that the introduction of Ni species lowered the energy barrier of H 2 dissociation to facilitate hydrogenation of adsorbed CO 2 on oxygen vacancies.
Other metal/In 2 O 3 catalysts. To improve the performance of In 2 O 3 , Qi et al. [79] prepared In x -Co y oxides catalysts for CO 2 hydrogenation to methanol. It was found that the methanation activity catalyzed by Co species was suppressed, and the best catalyst (In 1 -Co 4 ) exhibited nearly five times methanol STY compared to that of pure In 2 O 3 at conditions of 300 • C and 4 MPa. Several in situ and ex situ characterizations suggested that CO 2 hydrogenation over Co species and In x -Co y oxides all followed the formate pathway, and much stronger adsorbed capacity of CO 2 and carbon-containing intermediates on In x -Co y oxides catalyst contributed to a feasible surface C/H ratio, therefore facilitating CH 3 O* to produce methanol instead of being over-hydrogenated to methane. Gascon et al. [80] explored metal-organic framework (MOF) mediated synthetic approaches to prepare a Co 3 O 4 -supported In 2 O 3 catalyst for CO 2 hydrogenation to methanol. Compared to the traditionally coprecipitated In@Co catalytic system, the induction period in the hydrogenation process over MOF-derived In@Co catalyst could be tuned because ZIF-67(Co) support provided better In dopant distribution. In addition, the sequential pyrolysis-calcination steps could promote the formation of mixed-metal carbide (Co 3 InC 0.75 ) to stabilize high In distribution and prevent the formation of large individual oxide domains, thus leading to a faster induction period. The prepared catalyst (used 3In@8Co(300)) showed nanoparticles featuring core-shell morphologies (Co-In oxides shell over Co 3 InC 0.75 core) shown in Figure 9 and could obtain 0.65 g MeOH h −1 g cat −1 of maximum STY with methanol selectivity of 87% at conditions of 250 • C and 50 bar. Additionally, based on ZIF-67(Co), Zhang et al. [98] obtained a Co/C-N catalyst through the pyrolysis method and then mixed it with In 2 O 3 in different methods to prepare In 2 O 3 /Co/C-N for CO 2 hydrogenation to methanol. It was found that the proximity of Co/C-N and In 2 O 3 played a significant role in the synergetic catalysis for methanol synthesis from CO 2 hydrogenation. Moreover, the obvious difference in placement of separate Co/C-N and In 2 O 3 in catalytic performance also indicated CO 2 might be adsorbed and activated on the surface of In 2 O 3 to form carbon intermediates and then were further hydrogenated into methanol or byproducts over Co/C-N surface. Furthermore, the existence of the N element could improve the electron interaction of Co and In 2 O 3 and prevent the sintering of In 2 O 3 particles, thereby increasing the catalytic activity and stability for CO 2 hydrogenation to methanol. [98] obtained a Co/C-N catalyst through the pyrolysis method and then mixed it with In2O3 in different methods to prepare In2O3/Co/C-N for CO2 hydrogenation to methanol. It was found that the proximity of Co/C-N and In2O3 played a significant role in the synergetic catalysis for methanol synthesis from CO2 hydrogenation. Moreover, the obvious difference in placement of separate Co/C-N and In2O3 in catalytic performance also indicated CO2 might be adsorbed and activated on the surface of In2O3 to form carbon intermediates and then were further hydrogenated into methanol or byproducts over Co/C-N surface. Furthermore, the existence of the N element could improve the electron interaction of Co and In2O3 and prevent the sintering of In2O3 particles, thereby increasing the catalytic activity and stability for CO2 hydrogenation to methanol. Additionally, the combination of Cu and In2O3 also can be a good choice to improve the catalytic performance. Wu et al. [81] employed the coprecipitation method to fabricate various CuO-In2O3 and investigated the effect of the Cu:In molar ratio on the physicochemical properties and catalytic activity for methanol synthesis. The prepared catalyst mainly exhibited in the form of Cu11In9 phase and In2O3 at low Cu:In molar ratio (≤1:2) was because, on the one hand, the formation of CuxIny surface species inhibited the RWG reaction on the Cu surface. On the other hand, ZrO2 stabilized the In2O3 and generate additional In-Zr mixed oxide sites for CO2 conversion to methanol.

In2O3/Metal Oxides Composite Catalysts
Combining In2O3 with other metal oxides is also a significant strategy, which can im prove the dispersion of In2O3, increase the content of oxygen vacancies for CO2 adsorption and stabilize the key intermediates to facilitate methanol formation from CO2 hydrogena tion. Supporting In2O3 on ZrO2 is the most common and effective method because th electronic structure effect and crystal lattice mismatching between In2O3 and ZrO2 are ben eficial to CO2 activation for the formation of methanol. The research results by Pérez Ramírez et al. proved that combining In2O3 with ZrO2 can enhance the catalytic activit and stability of CO2 hydrogenation to methanol [58]. On the one hand, the reduced Z centers can attract oxygen atoms from the active phase in the reaction process, therefor increasing oxygen vacancies for CO2 adsorption and activation. On the other hand, ZrO support effectively improved the dispersion of In2O3 nanoparticles. Next, they explore the electronic, geometric, and interfacial phenomena between In2O3 and ZrO2 [86]. Th results suggested that the catalytic performance of mixed In-Zr oxides could not be im proved by coprecipitation, thereby excluding the primary role of electronic parameter The epitaxial growth of In2O3 was permitted on both monoclinic and tetragonal ZrO however, the more obvious lattice mismatching contributes to the lower dispersion o In2O3 on monoclinic ZrO2. Detailed characterizations and kinetic analyses revealed tw major facilitation of monoclinic ZrO2 support for In2O3 performance. One is that the ep taxial alignment of In2O3 on monoclinic ZrO2 ensured the high dispersion of the oxide t prevent sintering. The other is that the less favorable lattice matching between In2O3 an monoclinic ZrO2 produces tensile strain more easily, favoring the formation of oxyge vacancies on In2O3. The strong electronic oxide-support interaction between In2O3 an Additionally, the combination of Cu and In 2 O 3 also can be a good choice to improve the catalytic performance. Wu et al. [81] employed the coprecipitation method to fabricate various CuO-In 2 O 3 and investigated the effect of the Cu:In molar ratio on the physicochemical properties and catalytic activity for methanol synthesis. The prepared catalyst mainly exhibited in the form of Cu 11 In 9 phase and In 2 O 3 at low Cu:In molar ratio (≤1:2) after reduction treatment or in the reaction process, whereas with the increase in Cu content, Cu 7 In 3 phase was continuously weakened, and Cu phase emerged, which resulted in the formation of Cu-Cu 7 In 3 -In 2 O 3 . CuIn(1:2) catalyst obtained maximum methanol STY (5.95 mmol MeOH h −1 g −1 ) at the conditions of 260 • C and 3.0 MPa due to the highest Cu dispersion and the highest surface oxygen vacancies concentration, and the synergistic effect, Cu 7 In 3 phase for H 2 dissociation and In 2 O 3 for CO 2 adsorption, were considered as the major contributions for the efficient catalytic efficiency. The interfacial sites between Cu and metal oxides (In, Zn, and Zr) were tuned by Yu et al. for CO 2 hydrogenation to methanol [99]. The results suggested that the introduction of In 2 O 3 into Cu/ZrO 2 catalyst can increase the methanol formation rate from 52.7 mmol g cat −1 to 60.5 mmol g cat −1 . This was because, on the one hand, the formation of Cu x In y surface species inhibited the RWGS reaction on the Cu surface. On the other hand, ZrO 2 stabilized the In 2 O 3 and generated additional In-Zr mixed oxide sites for CO 2 conversion to methanol.

In 2 O 3 /Metal Oxides Composite Catalysts
Combining In 2 O 3 with other metal oxides is also a significant strategy, which can improve the dispersion of In 2 O 3 , increase the content of oxygen vacancies for CO 2 adsorption, and stabilize the key intermediates to facilitate methanol formation from CO 2 hydrogenation. Supporting In 2 O 3 on ZrO 2 is the most common and effective method because the electronic structure effect and crystal lattice mismatching between In 2 O 3 and ZrO 2 are beneficial to CO 2 activation for the formation of methanol. The research results by Pérez-Ramírez et al. proved that combining In 2 O 3 with ZrO 2 can enhance the catalytic activity and stability of CO 2 hydrogenation to methanol [58]. On the one hand, the reduced Zr centers can attract oxygen atoms from the active phase in the reaction process, therefore increasing oxygen vacancies for CO 2 adsorption and activation. On the other hand, ZrO 2 support effectively improved the dispersion of In 2 O 3 nanoparticles. Next, they explored the electronic, geometric, and interfacial phenomena between In 2 O 3 and ZrO 2 [86]. The results suggested that the catalytic performance of mixed In-Zr oxides could not be improved by coprecipitation, thereby excluding the primary role of electronic parameters. The epitaxial growth of In 2 O 3 was permitted on both monoclinic and tetragonal ZrO 2 ; however, the more obvious lattice mismatching contributes to the lower dispersion of In 2 O 3 on monoclinic ZrO 2 . Detailed characterizations and kinetic analyses revealed two major facilitation of monoclinic ZrO 2 support for In 2 O 3 performance. One is that the epitaxial alignment of In 2 O 3 on monoclinic ZrO 2 ensured the high dispersion of the oxide to prevent sintering. The other is that the less favorable lattice matching between In 2 O 3 and monoclinic ZrO 2 produces tensile strain more easily, favoring the formation of oxygen vacancies on In 2 O 3 . The strong electronic oxide-support interaction between In 2 O 3 and ZrO 2 for CO 2 hydrogenation to methanol was investigated by Gong et al. through quasi-in situ XPS experiments and DFT calculation [82]. Compared to the combination of In 2 O 3 and tetragonal ZrO 2 (In 2 O 3 /t-ZrO 2 ), In 2 O 3 /m-ZrO 2 (m-: monoclinic) exhibits more excellent catalytic performance (CO 2 conversion up to 12.1% with methanol selectivity of 84.6%) due to the stronger interaction to lead to the high dispersion of In-O-In over m-ZrO 2 . Methanol synthesis from CO 2 hydrogenation over In 2 O 3 /m-ZrO 2 follows the formate pathway. It was confirmed that the electron was transferred from m-ZrO 2 to In 2 O 3 to generate electron-rich In 2 O 3 , which can facilitate the dissociation of H 2 and help HCOO* transform into CH 3 O* by hydrogenation. Blum et al. [100] paid important attention to the support effect and surface reconstruction of In 2 O 3 /m-ZrO 2 in the process of CO 2 hydrogenation to methanol. They proposed that the modifying effects of m-ZrO 2 on In 2 O 3 mainly had two aspects: (I) m-ZrO 2 serves as a reservoir for partially reduced Based on their work, they also summarized the reaction mechanism pathway on the bare In 2 O 3 and In 2 O 3 /m-ZrO 2 , as exhibited in Figure 10. Witoon et al. [101] studied the effect of the calcination temperature of ZrO 2 support on the physicochemical properties and catalytic activities of In 2 O 3 /ZrO 2 for converting CO 2 and H 2 into methanol at a high reaction temperature. As the calcination temperature increased (from 600 to 1000 • C), the crystal of ZrO 2 support gradually changed from an amorphous phase to a tetragonal phase. The high calcination temperature of ZrO 2 support can decrease the reduction degree of In 2 O 3 , indicating the better interaction between In 2 O 3 and tetragonal ZrO 2 compared to amorphous ZrO 2 . In addition, the adsorption capacity of prepared In 2 O 3 /ZrO 2 catalysts for CO 2 and H 2 was enhanced with the increase in calcination temperature of ZrO 2 support, which promoted the highly selective conversion of CO 2 and H 2 into methanol instead of methane, whereas it did not have a significant impact on the formation of CO. ture of ZrO2 support can decrease the reduction degree of In2O3, indicating the better in teraction between In2O3 and tetragonal ZrO2 compared to amorphous ZrO2. In additio the adsorption capacity of prepared In2O3/ZrO2 catalysts for CO2 and H2 was enhance with the increase in calcination temperature of ZrO2 support, which promoted the high selective conversion of CO2 and H2 into methanol instead of methane, whereas it did no have a significant impact on the formation of CO. Müller et al. [102] investigated the effect of the ZrO2 phase on the reducibility, loc structure, and catalytic performance of In2O3/ZrO2 for CO2 hydrogenation to methanol b operando X-ray absorption spectroscopy (XAS) and XRD studies. The results suggeste that the amorphous ZrO2 (am-ZrO2) support could not form a solid solution with In2O and led to the rapid reduction in In2O3 to pure In 0 under reaction conditions, therefor suffering deactivation within minutes. For tetragonal ZrO2 (t-ZrO2) support, although Müller et al. [102] investigated the effect of the ZrO 2 phase on the reducibility, local structure, and catalytic performance of In 2 O 3 /ZrO 2 for CO 2 hydrogenation to methanol by operando X-ray absorption spectroscopy (XAS) and XRD studies. The results suggested that the amorphous ZrO 2 (am-ZrO 2 ) support could not form a solid solution with In 2 O 3 , and led to the rapid reduction in In 2 O 3 to pure In 0 under reaction conditions, therefore suffering deactivation within minutes. For tetragonal ZrO 2 (t-ZrO 2 ) support, although it can inhibit the complete reduction of In 2 O 3 into In 0 , the reduction extent was still too great (an average oxidation state of In below +2), resulting in poor catalytic activity. Surprisingly, it was found that the interaction between In 2 O 3 nanoparticles and monoclinic ZrO 2 (m-ZrO 2 ) can impel atomical dispersion of In 2+ /In 3+ into m-ZrO 2 lattice to form solid solution m-ZrO 2 :In, which prevented the over-reduction of In species (an average oxidation state of +2.3) and stabilized the active In-oxygen vacancy (V o )-Zr sites to facilitate CO 2 conversion into methanol. Additionally, the In-V o -Zr sites were vitally more stable toward reduction than In-V o -In sites in bixbyite-type In 2 O 3 , thus exhibiting superior catalytic activity and stability for CO 2 hydrogenation to methanol. Subsequently, they further studied the nature and abundance of sites for the hydrogen dissociation on In 2 O 3 /ZrO 2 -supported catalysts (In 2 O 3 /m-ZrO 2 , In 2 O 3 /t-ZrO 2 , In 2 O 3 /am-ZrO 2 and m-ZrO 2 :In catalysts) in CO 2 hydrogenation to methanol [103]. The results showed that indium hydride species (In-H) and hydroxyl groups (O-H) could be found on the surface of all redox-pretreated catalysts at room temperature when they were exposed to hydrogen, and only a low concentration of hydrogen dissociation sites still existed on the surface of In 2 O 3 /m-ZrO 2 and m-ZrO 2 :In without redox pretreatment. In 2 O 3 /m-ZrO 2(redox) possessed the highest concentration of surface indium sites for heterolytic activation of H 2 , and the obtained In-H species can react with CO 2 to form surface formate species (methanol intermediates) at room temperature, indicating the appreciable reactivity of In-H and carbonates on the m-ZrO 2 support. Additionally, the reduction in hydrogen at 400 • C led to the high dispersion of In into m-ZrO 2 to form a m-ZrO 2 :In solid solution. Hydrogen dissociation in m-ZrO 2 :In solid solution proceeded on In 3+ -O-Zr 4+ sites, obtaining In-H and Zr-OH species.
The preparation method of In 2 O 3 /ZrO 2 also vitally affects the electronic structure effect, thus to optimize the interaction of In 2 O 3 and ZrO 2 , and the surface exposure degree of In 2 O 3 , four different compositing strategies (liquid-phase coprecipitation, precipitationcoating method, ball milling method, and incipient wetness impregnation, respectively) for the synthesis of In 2 O 3 /ZrO 2 were compared by Gao et al. [104]. It was found that the exposure area of In 2 O 3 prepared by the precipitation-coating method was the highest (S In = 6.22 m 2 g −1 ), whereas it was lowest (S In = 1.56 m 2 g −1 ) by the coprecipitation method due to the formation of In 2 O 3 bulk dispersion with ZrO 2 . The dispersion of In 2 O 3 on ZrO 2 can inhibit the over-reduction of In 2 O 3 , and the exposure area of In 2 O 3 was beneficial for CO 2 adsorption and activation. Furthermore, DRIFTS results and DFT calculation demonstrated that the oxygen vacancy defects of In 2 O 3 /ZrO 2 would stabilize the key formate intermediates to facilitate the formation of methanol obeying the carbonate-formate-methoxy pathway, as shown in Figure 11: H 2 was adsorbed on the exposed In 2 O 3 surface (H*), and subsequently generated In-H* and O-H* by hydrogen heterolysis. CO 2 was adsorbed and activated by In-V o -Zr oxygen vacancies to form carbonate species (CO 2 *), and then it combined with the activated In-H* to generate the formate intermediate (HCOO*). Later, HCOO* was further hydrogenated into CH 3 OH via the pathway of HCOO*→H 2 CO*→H 3 CO*→CH 3 OH. Apart from ZrO 2 , Ga 2 O 3 [84], CeO 2 [105], and MnO [106] were also used to combine with In 2 O 3 for converting CO 2 into methanol, and their promotion for In 2 O 3 performance was also associated with the In 2 O 3 dispersion, metal-support interactions, or tuning of basic sites. . Catalytic mechanism diagram for CO2 hydrogenation to methanol over In2O3/ZrO2 catalyst prepared by precipitation-coating method. Reproduced with permission from ref. [104]. Copyright 2022 Elsevier.

Conclusions and Further Directions
In summary, In2O3-based catalysts are promising for the industrial application of thermochemical CO2 hydrogenation to methanol. Various research methods have been adopted to explore the formation process and possible structure of active sites and the reaction mechanism over In2O3-based catalysts for CO2 hydrogenation to methanol. Furthermore, research has been ongoing to further understand the structure-activity relationship and identify the key factors affecting the catalytic performance. It is commonly accepted that methanol synthesis from CO2 hydrogenation over In2O3-based catalysts follows a formate pathway, where CO2 adsorbed on the oxygen vacancy of In2O3 passes through the route of CO2*→HCOO*→H2CO*→H3CO*→CH3OH. The phase state of In2O3 plays a key role in determining CO2 conversion and methanol selectivity, and compared to cubic bixbyite-type In2O3 (c-In2O3), hexagonal In2O3 (h-In2O3) with a high proportion of the exposed (104) surface exhibited the higher catalytic activity and possessed high stability, which is mainly due to the facile formation of the oxygen vacancies at low coverage and the favorable formation of the hydride adsorbate at the In sites on (104) surface. The factors dictating performance improvement of In2O3-based catalysts include (1) the ability for dissociation and spillover of hydrogen, (2) the number of oxygen vacancies for CO2 Figure 11. Catalytic mechanism diagram for CO 2 hydrogenation to methanol over In 2 O 3 /ZrO 2 catalyst prepared by precipitation-coating method. Reproduced with permission from ref. [104]. Copyright 2022 Elsevier.

Conclusions and Further Directions
In summary, In 2 O 3 -based catalysts are promising for the industrial application of thermochemical CO 2 hydrogenation to methanol. Various research methods have been adopted to explore the formation process and possible structure of active sites and the reaction mechanism over In 2 O 3 -based catalysts for CO 2 hydrogenation to methanol. Furthermore, research has been ongoing to further understand the structure-activity relationship and identify the key factors affecting the catalytic performance. It is commonly accepted that methanol synthesis from CO 2 hydrogenation over In 2 O 3 -based catalysts follows a formate pathway, where CO 2 adsorbed on the oxygen vacancy of In 2 O 3 passes through the route of CO 2 *→HCOO*→H 2 CO*→H 3 CO*→CH 3 OH. The phase state of In 2 O 3 plays a key role in determining CO 2 conversion and methanol selectivity, and compared to cubic bixbyite-type In 2 O 3 (c-In 2 O 3 ), hexagonal In 2 O 3 (h-In 2 O 3 ) with a high proportion of the exposed (104) surface exhibited the higher catalytic activity and possessed high stability, which is mainly due to the facile formation of the oxygen vacancies at low coverage and the favorable formation of the hydride adsorbate at the In sites on (104) surface. The factors dictating performance improvement of In 2 O 3 -based catalysts include (1) the ability for dissociation and spillover of hydrogen, (2) the number of oxygen vacancies for CO 2 activation, (3) the dispersion of In 2 O 3 nanostructures, and (4) the stability of key intermediates. Two different strategies, metal doping and hybrid with other metal oxides, respectively, are utilized to optimize the above factors for enhancing the catalytic performance of In 2 O 3 -based catalysts. For the facilitation of dissociation and spillover of hydrogen, the most effective strategy is introducing the metal element (M, M = Pd, Pt, Ni or Co, etc.) into In 2 O 3 . The existence of M nanostructures sharply promotes the dissociative adsorption of hydrogen, therefore being instrumental in enhancing the hydrogenation process and increasing surface oxygen vacancy. On balance, the synergistic catalysis effect of M and In 2 O 3 contributes to the high catalytic performance of M/In 2 O 3 catalysts. As for the improvement of CO 2 adsorption and key intermediates stability, supporting In 2 O 3 on the other metal oxides is considered to be extremely useful, especially the combination of In 2 O 3 with ZrO 2 support. ZrO 2 support is an excellent modifier for In 2 O 3 to promote the concentration of oxygen vacancy, enhance the interaction with CO 2 , and stabilize the key intermediates. The structure-activity relationship of In 2 O 3 /ZrO 2 can be concluded as follows: high surface and dispersion of In 2 O 3 to prevent sintering and strong interaction of In 2 O 3 and ZrO 2 (i.e., solid solution m-ZrO 2 :In) from preventing the over-reduction of In 2 O 3 , generate more active In-oxygen vacancy (V o )-Zr sites for activating CO 2 and stabilizing key formate intermediates, and also form electron-rich In 2 O 3 (electron transfer from ZrO 2 to In 2 O 3 ) to facilitate the dissociation of hydrogen. In addition, the phase state of ZrO 2 support greatly affects the catalytic activity of In 2 O 3 /ZrO 2 , and different from amorphous ZrO 2 and tetragonal ZrO 2 , the interaction between monoclinic ZrO 2 and In 2 O 3 nanoparticles can impel atomical dispersion of In 2+ /In 3+ into m-ZrO 2 lattice to form solid solution m-ZrO 2 :In, which prevented the over-reduction of In species and stabilized the active In-oxygen vacancy-Zr sites to facilitate CO 2 conversion into methanol.
Although the research of In 2 O 3 -based catalysts for CO 2 hydrogenation to methanol has made substantial headway recently, several issues remain to be addressed in future studies. For instance, it is urgent to reveal the evolutionary process of active sites under real reaction conditions, which is extremely crucial to establish a more intuitive and reliable structure-activity relationship for designing In 2 O 3 -based catalysts. In addition, the catalytic mechanism over In 2 O 3 -based catalysts is usually proposed by theoretical study-based DFT calculation at present; however, the validity in practical applications is rather challenging. On the one hand, the microscopic reaction process (molecular level) could not be observed through experiments to verify its validity. On the other hand, the DFT calculation is unable to restore the real experimental conditions (i.e., species of active sites, mass transfer, etc.), therefore resulting in the difference between the theoretical reaction pathway and the actual reaction pathway. In order to obtain the evolutionary process of active sites and valid reaction mechanism, two considerable methods should be highlighted in future studies as follows: (1) making more efforts to analyze and identify the species of key intermediates by comprehensive in situ characterization technology (i.e., in situ DRIFTS, in situ XPS, etc.) and kinetic investigation; (2) combining DFT calculations with other simulation methods (i.e., computational fluid dynamics (CFD), kinetic Monte Carlo (KMC), etc.) to build more realistic models for theoretical study. Furthermore, from the point view of practical application, it is extremely necessary to reveal the deactivation mechanisms and enhance catalytic stability of In 2 O 3 -based catalysts in converting CO 2 into methanol, so more attention should be paid to the issues of sintering and the structural evolution monitored by in situ/operando spectroscopic techniques. Overall, this review mainly summarized the regulation and modification of active sites of In 2 O 3 -based catalysts to facilitate the activation of reactants and stabilization of intermediates in CO 2 hydrogenation, which is conducive to the design of more efficient In 2 O 3 -based catalysts for the highly selective transformation of CO 2 to methanol in future studies, realizing the resource utilization of CO 2 .