Hydrogenation of Carbon Dioxide to Value-Added Liquid Fuels and Aromatics over Fe-Based Catalysts Based on the Fischer–Tropsch Synthesis Route

: Hydrogenation of CO 2 to value-added chemicals and fuels not only effectively alleviates climate change but also reduces over-dependence on fossil fuels. Therefore, much attention has been paid to the chemical conversion of CO 2 to value-added products, such as liquid fuels and aromatics. Recently, efﬁcient catalysts have been developed to face the challenge of the chemical inertness of CO 2 and the difﬁculty of C–C coupling. Considering the lack of a detailed summary on hydrogenation of CO 2 to liquid fuels and aromatics via the Fischer–Tropsch synthesis (FTS) route, we conducted a comprehensive and systematic review of the research progress on the development of efﬁcient catalysts for hydrogenation of CO 2 to liquid fuels and aromatics. In this work, we summarized the factors inﬂuencing the catalytic activity and stability of various catalysts, the strategies for optimizing catalytic performance and product distribution, the effects of reaction conditions on catalytic performance, and possible reaction mechanisms for CO 2 hydrogenation via the FTS route. Furthermore, we also provided an overview of the challenges and opportunities for future research associated with hydrogenation of CO 2 to liquid fuels and aromatics.


Introduction
With the rapid growth of the world's population and the development of the industrial sectors, fossil fuels have made important contributions to economic and social development.However, the overuse of fossil fuels has resulted in large amounts of CO 2 emissions and a series of environmental problems, such as global warming, natural disasters, extreme weather, etc. [1][2][3].In order to reduce CO 2 emissions and to achieve carbon neutrality, it is necessary to adopt low-carbon renewable energy (such as wind and solar energy) and to develop efficient CO 2 capture and utilization (CCU) technologies [4,5].The utilization of renewable energy can effectively reduce CO 2 emissions, but it is currently difficult to be deployed on a large scale due to its characteristics of fluctuation and intermittency under changing natural conditions.By contrast, CCU technologies treat CO 2 as a renewable carbon source and convert it into high value-added chemicals and fuels, which can significantly reduce both CO 2 emissions and the dependence on fossil energy.
Among various CCU technologies, the thermocatalytic conversion of CO 2 has received widespread attention from industry and academia due to its high industrial application potential, as well as adjustable product selectivity.In addition, among the thermocatalytic conversion technologies, CO 2 hydrogenation has been the main focus and has made considerable progress in pilot-scale and large-scale applications [6][7][8][9][10][11], especially for the production of C1 molecules (CH 4 [12][13][14][15], CH 3 OH [16], HCOOH [17][18][19], etc.).Furthermore, long-chain hydrocarbons with higher added value (i.e., liquid fuels and aromatics) have recently received extensive attention due to their wide applications in the transportation and petrochemical industries.Herein, for liquid fuels, they are widely used as transportation fuels (e.g., gasoline, jet fuel, and diesel) and play a significant role in the global energy supply market [11].For aromatics, they are important commodity petrochemicals and are generally identified as the important platform chemicals for the production of solvents, value-added chemicals, and polymers [20,21].Currently, the production of liquid fuels and aromatics heavily depends on petroleum cracking and catalytic reforming, which exacerbates the depletion of fossil fuels and results in considerable CO 2 emissions.In this paper, an alternative strategy for hydrogenation of CO 2 to liquid fuels and aromatics provides a promising and sustainable route for mitigating carbon emissions and producing high value-added fuels and chemicals.
Generally, studies on the hydrogenation of CO 2 to hydrocarbons can be divided into two possible routes: Fischer-Tropsch synthesis (FTS) and methanol-mediated routes [22].For the FTS route, CO intermediates produced from reverse water-gas-shift (RWGS) can be converted to C 2+ hydrocarbons.Furthermore, olefins in the obtained C 2+ hydrocarbons can be further converted into aromatics over the acid sites of zeolites via a series of oligomerization, dehydrogenation, and cyclization reactions.As for the methanol-mediated route, the formation of C 2+ hydrocarbons and aromatics requires methanol intermediate transformation via methanol-to-hydrocarbons (MTH) and methanol-to-aromatics (MTA) routes [21,23].However, the MTH and MTA routes generally require high temperatures of 400-500 • C [20], which result in high CO selectivity due to the endothermic character of RWGS, and this is not conducive to enhancing the yield of high value-added chemicals [24].Therefore, the FTS route is preferred for producing long-chain hydrocarbons and aromatics via CO 2 hydrogenation.However, the production of long-chain hydrocarbons with higher added value still faces many challenges due to the high C-C coupling barrier during the CO 2 hydrogenation process.In recent years, several efficient catalysts with excellent catalytic performance have been applied to convert CO 2 to long-chain hydrocarbons.Considering the lack of reviews on current research progress on the hydrogenation of CO 2 to liquid hydrocarbons (i.e., liquid fuels and aromatics), we conducted a systematic and comprehensive summary of recent advances in the development of highly efficient catalysts for CO 2 hydrogenation.Herein, we mainly focus on the research progress regarding the synthesis of liquid fuels and aromatics via the FTS route.
This work is mainly divided into three sections: (1) Hydrogenation of CO 2 to liquid fuels over modified Fe-based catalysts or multifunctional catalysts combined with Fe-based catalytic components and zeolites.Herein, the effects of alkali promoters, secondary metals, catalyst supports, preparation methods, zeolite pre-treatments, and reaction conditions on the catalytic performance are discussed in detail.(2) Hydrogenation of CO 2 to aromatics over FeO x /zeolite tandem catalysts, where the effects of promoters, catalyst supports, and preparation methods of Fe-based catalysts, the effects of acidic sites, the types of acidity, and the preparation methods of zeolites, and the effects of the reaction conditions on the overall catalytic performance of the tandem catalysts are illustrated.(3) Reaction mechanisms of CO 2 hydrogenation, where the possible reaction mechanisms for hydrogenation of CO 2 to liquid fuels and aromatics via the FTS route are summarized in detail.Moreover, the main contributions of this work are as follows: (1) A comprehensive and systematic review of the research progress on the development of efficient catalysts for the hydrogenation of CO 2 to liquid fuels and aromatics via the FTS route is implemented.(2) The key factors influencing the catalytic activity and stability of various catalysts and the strategies for optimizing the catalytic performance and product distribution, as well as the effects of the reaction conditions on catalytic performance are discussed in detail.(3) The possible reaction mechanisms for the hydrogenation of CO 2 into liquid fuels and aromatics via the FTS route are summarized.(4) An overview of the challenges and opportunities for future research regarding CO 2 hydrogenation to value-added products is proposed, providing very important guidelines for the preparation of highly efficient CO 2 hydrogenation catalysts.

Direct Hydrogenation of CO 2 to Liquid Fuels over Fe-Based Catalysts
Hydrogenation of CO 2 to liquid fuels via the FTS route (Equation ( 1)) usually includes two steps: First, CO 2 is converted to CO by RWGS (Equation ( 2)) and then CO is converted to liquid fuels by FTS (Equation ( 3)).Various catalysts (e.g., Fe-, Co-, and Ru-based catalysts) based on catalysts for CO hydrogenation have been developed due to their excellent selectivity towards long-chain hydrocarbons [25][26][27][28][29].Among them, Ru-based catalysts [27] have the best catalytic performance, however, the high catalyst costs limit their commercial application on a large scale.Instead, Fe-and Co-based catalysts have been widely studied because of their relatively high catalytic activity and low catalyst cost [22,[28][29][30][31][32][33].Typically, Fe-and Co-based catalysts are used as the industrial CO-FTS catalysts.Compared with Fe-based catalysts, Co-based catalysts exhibit higher long-chain hydrocarbon selectivity at low temperatures (<300 • C), but they tend to produce CH 4 at high temperatures (>300 • C).Due to the fact that Co-based catalysts almost have no RWGS activity, a second component (e.g., copper) with RWGS activity is necessary for converting CO 2 to CO [11].In contrast, Fe-based catalysts have both RWGS and FTS activity, with high CO 2 conversion and low CH 4 selectivity within a wide temperature range (300-400 • C) [34][35][36][37][38][39][40] for CO 2 hydrogenation [22].Herein, the recent progress on the catalysts for the hydrogenation of CO 2 to liquid fuels is mainly introduced below, and the specific representative catalysts as well as their catalytic performances are illustrated in Tables 1 and 2.

Strategies for Improving the Catalytic Performance of Fe-Based Catalysts
Generally, in the hydrogenation of CO 2 to liquid fuels based on Fe-based catalysts, Fe 3 O 4 , as the active site for RWGS, is responsible for generating CO intermediates, and iron carbides (Fe x C y ), as the active sites for FTS, play a key role in carbon chain growth [16,72].However, pure Fe-based catalysts usually show low CO 2 conversion and high CH 4 selectivity [73,74].To address this issue, several strategies such as the addition of alkali promoters and secondary metals, the selection of suitable catalyst supports, and the optimization of catalyst preparation methods have been widely investigated to further enhance CO 2 conversion and to improve the selectivity towards long-chain hydrocarbons, which will be summarized below in detail.

Alkali Promoters
The addition of alkali promoters (K, Na, Rb, Cs, etc.) is an effective approach to enhance CO 2 conversion as well as long-chain hydrocarbon selectivity [41,44,47,48,50,56,61,62,75].More specifically, alkali promoters can transfer electrons to the active Fe species, therefore increasing the local electron density around Fe species, which contributes to enhancing Fe-C bonds and weakening C=O bonds [76,77], thus facilitating the formation of iron carbides as well as the carbon chain growth [44].In addition, alkali promoters can enhance CO 2 adsorption while inhibiting H 2 adsorption on the catalyst surface, which leads to a high C/H ratio and favors C-C coupling [41,78].However, there is an optimum loading amount of alkali promoters, considering the catalytic activity, selectivity, and stability [34,36,53,79].As shown in Figure 1, low K (a: 0→5 wt%, b: K/Fe = 0→0.3)significantly facilitates CO 2 conversion and C 5+ hydrocarbon selectivity, accompanied by low CH 4 selectivity.However, further addition of K (a: 5→15 wt%; b: K/Fe = 0.3→1) results in an increase in CO selectivity and also suppresses selectivity towards long-chain hydrocarbons.It is because, on the one hand, excess K covers Fe active sites, which reduces CO 2 adsorption and its subsequent conversion [16]; on the other hand, excess K further enhances Fe-C bonds, which may cause severe carbon deposition due to the difficultly in C desorption [34,36], thus reducing the catalyst stability [80,81].
(1) Cu Promoter The addition of Cu to Fe-based catalysts can generally facilitate the reduction and dispersion of Fe active species.For example, Liu et al. [50] investigated the effects of the Cu content on the catalytic performance of Fe-Cu-K/Al 2 O 3 catalysts and found that with the increase in the Cu content from 0 to 3%, the light olefin (C 2 = -C 4 = ) selectivity decreased from 20.3 to 14.5%, while the C 5+ hydrocarbon selectivity significantly increased from 5.4 to 13.8%.This can be attributed to the strong interaction between Cu and Fe, which promoted the reduction and dispersion of Fe active species, as confirmed by HR-TEM, H 2 -TPR, and CO 2 -TPD analyses [35].Furthermore, DFT calculations by Hwang et al. [48] revealed the synergistic effects of Cu and K for CO 2 hydrogenation over Fe-based catalysts.That is, the Fe-Cu alloy promotes CO 2 activation by reducing the energy required for oxygen removal, and the addition of K causes an increasing carbon binding energy and a decreasing hydrogen binding energy, which facilitates CO formation and C-C coupling over Cu and Fe surfaces, respectively.
In addition, the types of Cu precursors also have important impacts on the catalytic performance of Fe-based catalysts.Choi et al. [52]  (2) Co Promoter The addition of Co with the capability of chain growth to Fe-based catalysts can promote the conversion of CO intermediates and the subsequent selectivity towards longchain hydrocarbons [34,36,37,58].For example, 10Fe0.3K0.53Cocatalysts successfully prepared by Jiang et al. [58] exhibited excellent catalytic performance with C 5+ hydrocarbon selectivity of 47% under CO 2 conversion of 54.6%.Moreover, they also investigated the synergistic effects between Fe and Co active sites on the catalytic performance of Fe-based catalysts, which is related to the proximity between the Fe and Co active sites.The distance between the Fe and Co active sites that changed from the nanoscale to meter scale was simulated by three assembly styles of the 15Fe5K/SiC and 8Co/SiC catalysts: powdermixing, particle-mixing, and layer-by-layer (15Fe5K/SiC and 8Co/SiC catalysts were loaded upstream and downstream, respectively).It was found that as the distance between the Fe and Co active phases increased, the C 5+ hydrocarbon selectivity decreased from 30.5% to 2.6%, corresponding to a sharp increase in CH 4 selectivity from 50.6% to 90.2%.That is, the intimate distance between Fe and Co can generate high CO concentrations on the Co active sites, which favors the formation of C 5+ hydrocarbons via the FTS route.By contrast, the increasing distance between Fe and Co can generate low CO concentrations on the Co active sites, leading to a low C/H ratio, which favors the formation of CH 4 .
A suitable proximity between Fe and Co strongly associated with synergistic effects is essential to obtain high CO 2 conversion and C 5+ hydrocarbon selectivity.However, it is not clear whether the proximity of the Fe-Co mixture is maintained all the time due to the transformation of Fe species during the CO 2 hydrogenation process.Hwang et al. proposed a strategy to stabilize the Fe-Co bimetallic structure [36] in which TMPP cobalt (II) (Co-TMPP) precursors were used as Co-NC supports, and the prepared FeK/Co-NC catalysts had a CO 2 conversion of 51.7% and C 5+ hydrocarbon selectivity of 42.4%.Compared with the FeK/Co-NC catalysts, the FeCoK/NC catalysts prepared with NC supports exhibited a lower C 5+ hydrocarbon selectivity of 28.8%.Furthermore, TEM-EDS mapping was characterized for the FeK/Co-NC and FeCoK/NC catalysts to visualize the degree of difference in the formation of the Fe-Co alloy as shown in Figure 2. FeK/Co-NC accounted for 14-27% Co nanoparticles (NPs), whereas FeCoK/NC has a relatively small amount of Co NPs that varies from 3% to 13% (Figure 2a).The difference in alloying degree was more pronounced.As shown in Figure 2b, after both catalysts underwent the CO 2 hydrogenation reaction for 70 h, and the spent FeK/Co-NC catalyst maintained uniform Fe-Co alloyed structures containing ~20% Co.However, the Fe-Co alloyed structures in the FeCoK/NC catalyst were dealloyed, with Co compositions of 3-60%.This indicates that the Fe-Co alloy in the FeCoK/NC catalysts tended to break away from the alloyed state to Co clusters during the CO 2 hydrogenation process, which reduced the chain growth ability, whereas the Co-NC supports effectively induced the formation of the Fe-Co alloy (Figure 3), which contributed to improving the C 5+ hydrocarbon selectivity.(3) Zn Promoter Similar to Cu, Zn [40,59,61,84] acts as a structural promoter [38,39] in the Fe-based catalysts and can generally facilitate CO 2 adsorption as well as subsequent conversion.This can be attributed to the dispersion of Fe active species caused by the interaction between Zn and Fe.However, too strong an interaction inhibits the reduction of Fe species [59,61], which is not conducive to the production of long-chain hydrocarbons.To deal with this difficulty, strategies of synergistic interactions by combining Zn with Na have been proposed by Zhang et al. [40].The CO 2 and H 2 adsorption properties of the reduced Fe catalysts (pure Fe, Zn-Fe, Na-Fa, and Na-Zn-Fe catalysts) were investigated using CO 2 -TPD (Figure 4a) and H 2 -TPD (Figure 4b), respectively.For CO 2 adsorption, the addition of Zn to Na-Fe catalysts significantly strengthened the α and β peaks [60].As For Na-Fe catalysts, apart from the α and β peaks, a γ peak appeared at around 886 K, which corresponds to the interaction between the CO 2 and alkaline sites increased by Na addition, thus enhancing CO 2 adsorption [60,85].Furthermore, the introduction of Zn to Na-Fe catalysts particularly increased the strong CO 2 adsorption at the γ peak, which can be explained as follows: Zn reduces the particle sizes of Na-Zn-Fe catalysts.As for the H 2 adsorption, the peaks centered at 645-659 K and 773-826 K were attributed to weak H 2 adsorption [49,86] and the dissociative or subsurface H 2 adsorption [87,88] on the catalyst surface, respectively.It can be seen that the addition of Zn significantly enhanced the amounts of dissociative and weak H 2 adsorption, which promoted the activity of CO 2 hydrogenation.In other words, the addition of Zn not only significantly reduces the particle size of the Fe active species but also improves the capacity of H 2 adsorption, which is favorable for CO 2 conversion.In addition, co-modification of the Na promoters significantly contributes to the formation of active Fe 5 C 2 during the CO 2 hydrogenation process, which is conducive to the formation of long-chain hydrocarbons.

Catalyst Supports
Catalyst supports also play a key role in improving the catalytic activity and catalyst stability [89] because the interaction between the supports and Fe species can inhibit the agglomeration of Fe active sites.Currently, traditional supports such as Al 2 O 3 [16], TiO 2 [33,58], SiO 2 [50,57] , and ZrO 2 [44] have been investigated for CO 2 hydrogenation via the FTS route, among which Al 2 O 3 is mainly used as Fe-based catalyst supports.For example, Song et al. investigated the effects of Al 2 O 3 supports on the catalytic performance of FeK/Al 2 O 3 catalysts and found that Al 2 O 3 with hydroxyl groups facilitates the dispersion of Fe particles and the adsorption of CO intermediates [90], which contributed to the formation of small particles of iron carbide and the improved catalytic performance.Furthermore, amorphous AlO x phases, as additional active sites, were recently proposed by Khan et al. [49].Herein, the amorphous AlO x phases, on the one hand, enhance the adsorption of CO 2 and H 2 , thereby improving CO 2 conversion; on the other hand, these phases provide sites for olefin readsorption, which is beneficial for enhancing long-chain hydrocarbon selectivity.
In addition, there is a suitable range for the pore size of Al 2 O 3 supports, considering the catalytic activity and selectivity.Xie et al. reported that a Al 2 O 3 pore size of 7-10 nm was the optimal active size for hydrogenation of CO 2 to hydrocarbons [45].Outside this range, the increase in the pore size of the Al 2 O 3 supports led to an increase in the Fe 2 O 3 particle size, reducing the dispersion of the Fe active sites.However, the smaller pore size of the Al 2 O 3 supports led to too small Fe 2 O 3 particle sizes, which cannot favor the carbon chain growth on the catalyst surface.Therefore, it is important to select a support with a suitable pore size.
In recent years, carbon materials with hydrophobicity and excellent carburized ability have been applied as supports, such as mesoporous carbon [68], carbon nanotubes (CNTs) [65][66][67], and carbon materials based on the pyrolysis of Metal-Organic Frameworks (MOFs).Herein, mesoporous carbon with a mesoporous structure can provide favorable conditions for the rapid diffusion of long-chain hydrocarbons.However, unlike mesoporous carbon, CNTs with tubular structures can provide a high C/H ratio of sites to facilitate carbon chain growth.This is because the outer walls of CNTs naturally possess an electron-rich structure; this structure favors CO 2 adsorption and relatively inhibits H 2 adsorption [65].Therefore, CNTs have a promising application in the field of CO 2 hydrogenation.Moreover, MOFs have been used as effective precursors for the preparation of carbon supports due to their high surface area and adjustable pore size [91,92].Guo et al. first pyrolyzed Fe-MIL-88B into Fe-based catalysts [62], which exhibited a CO 2 conversion rate of 46% but high CH 4 selectivity of 26.6%.They also prepared NC materials containing Zn promoters by pyrolyzing ZIF-8 precursors [61], and their supported Fe catalysts show considerable olefin selectivity of 30%.However, in order to obtain high catalytic activity and long-chain hydrocarbon selectivity, regulation or modification of the structure and composition of MOF-based catalysts need to be further explored.

Preparation Methods
Preparation methods are directly relevant to the distribution of catalyst components, which is crucial to regulate the interaction between Fe active sites and other components (i.e., supports and promoters).Herein, representative catalysts prepared by different preparation methods and their catalytic performance are illustrated in Table 3.
Table 3. Representative catalysts prepared by different preparation methods and their catalytic performance for CO 2 hydrogenation.

Entry Reaction Conditions
Hydrocarbon Distribution (%) Co-precipitation Methods  Currently, Incipient Wetness Impregnation (IWI) methods are commonly used to prepare efficient catalysts [47,48].The IWI methods are effective in uniformly dispersing Fe species and promoters on the supports (Figure 5a-d) [50].Compared with IWI methods, co-precipitation methods can also regulate the particle sizes of Fe species (Figures 5e and 6) and pore sizes of supports by changing the pH values of the reaction solution, stirring speed, and aging time.This provides a feasible pathway to prepare Fe-based catalysts with high activity.In addition, other methods such as Hydrothermal Synthesis Methods (HSM) and Organic Combustion Methods (OCM) have been reported to prepare catalysts [16,40,42,52,59,64,93].HSM are generally conducted under high-temperature and high-pressure conditions.Catalysts prepared by HSM mostly have spinel structures (e.g., ZnFe 2 O 4 [59] catalysts), which can inhibit the sintering of Fe species during the heat-treatment process and improve the stability of the catalysts during the CO 2 hydrogenation process.Recently, OCM have been applied in the CO 2 hydrogenation field.A citric acid-based slurry containing Fe, Mn, and K precursors was ignited to prepare FeMnK catalysts [56] (Equations ( 4) and ( 5)), which exhibited excellent catalytic performance with CO 2 conversion (38.2%) and C 5+ hydrocarbon selectivity (58.4%).The main advantages of OCM are as follows: (1) The vaporization of water consumes the released heat, limiting the rapid rise of temperature and reducing the local sintering of metal oxide particles.(2) The releases of N 2 and CO 2 are conducive for the high-dispersion preparation of catalysts [94].

Tandem Catalysts Combined with Fe-Based Active Components and Zeolites
Generally, the aforementioned Fe-based catalysts designed for CO 2 hydrogenation also co-generate C 2 = -C 4 = , along with the targeted long-chain hydrocarbons.Therefore, tandem catalysts combined with Fe-based active components and zeolites have been adopted to improve the selectivity of long-chain hydrocarbons via further conversion of the C 2 = -C 4 = into long-chain hydrocarbons over the acid sites of zeolites through a series of reactions such as oligomerization, isomerization, and aromatization [95][96][97].The tandem catalysts can not only improve the selectivity towards long-chain hydrocarbons but also increase the research octane number (RON) of liquid fuels.
For the tandem catalysts, the final distribution of the hydrocarbon products heavily depends on the pore structure and acidity of zeolites.For instance, HMCM-22 with the MWW structure and Hβ with the BEA structure benefit the isomerization and alkylation of olefins [98,99], while HZSM-5 with the MFI structure favors the aromatization of olefins.As for the roles of acidity, Wei et al. revealed that HZSM-5 zeolites with strong BAS were conducive to olefin aromatization [100], while HMCM-22 and Hβ zeolites with medium BAS promoted olefin isomerization.Guo et al. indicated that eliminating the surface-strong Brønsted acids of HZSM-5 by using ion-exchange methods (K + , Na + , etc.) is beneficial for improving the selectivity towards C 5+ long-chain hydrocarbons and suppressing the formation of short-chain saturated hydrocarbons [71].In addition, a possible reaction pathway for the hydrogenation of CO 2 into iso-paraffins over a Na-Fe 3 O 4 /HMCM-22 catalyst (Figure 7a) was proposed by Wei et al.In addition to the use of tandem catalysts, the use of a multiple-bed reaction system is another effective way to improve the selectivity towards long-chain hydrocarbons.For example, Noreen et al. designed a tri-bed tandem NaFe + SAPO-11 + HZSM-5 catalytic reaction system for gasoline synthesis via CO 2 hydrogenation (Figure 7b) [70], where the mono-branched hydrocarbons were generated over NaFe + SAPO-11, which undergoes secondary isomerization over HZSM-5, therefore improving the selectivity of multibranched hydrocarbons.In this case, the gasoline selectivity reached 71.7% under a CO 2 conversion rate of 31.2%.Moreover, the produced gasoline had an RON value of 91.7, with the selectivity of mono-and di-branched hydrocarbons of 25.7 and 12.5%, respectively (Figure 8a,b).

Effects of Reaction Conditions on Catalytic Performance
In general, the reaction conditions (the reaction temperature, reaction pressure, space velocity, and H 2 /CO 2 ratio) have significant impacts on the catalytic performance of CO 2 hydrogenation (i.e., CO 2 conversion, selectivity towards targeted hydrocarbons, and catalyst stability).In addition, in the case of the tandem catalysts, the mixing mode of different active components is also an important factor.The relevant content will be discussed below in more detail.

Reaction Temperature
In general, there is an optimum temperature range for CO 2 hydrogenation via the FTS route, that is, too low or too high temperatures are not preferred.This is because the RWGS reaction is endothermic, which favors high temperatures to enhance equilibrium CO 2 conversion, while the FTS reaction is exothermic, which prefers low temperatures to achieve low CH 4 selectivity [101].Therefore, increasing reaction temperatures can improve CO 2 conversion, but this may result in a decrease in the selectivity towards C 5+ long-chain hydrocarbons [102,103].As seen in Figure 9a, as the temperatures increased from 300 to 360 • C, the CO 2 conversion increased by about 20%.However, when the temperature further increased to 380 • C, the CO 2 conversion only slightly increased by 1.2%, but C 5+ long-chain hydrocarbon selectivity significantly decreased by 9.8%.Therefore, it is important to seek an optimum temperature range to obtain optimized catalytic performance for CO 2 hydrogenation by balancing the RWGS and FTS reactions.

Reaction Pressure
Generally, high reaction pressures are beneficial for improving both the CO 2 conversion and the selectivity towards long-chain hydrocarbons.This is because the CO 2 hydrogenation reaction decreases in moles, so high pressure facilitates CO 2 conversion from the thermodynamic point of view.As shown in Figure 9b, as the reaction pressure increased from 2 to 5 MPa, the CO 2 conversion over the K 3 /FeMn 10 Ti 20 catalysts increased from 28.4% to 34.9%.Meanwhile, its chain growth capacity was also enhanced, leading to a decrease in CO selectivity and an increase in the selectivity towards C 5+ long-chain hydrocarbons.This is mainly because of the enlarged amount of surface C* species under high pressure, which benefits the generation of iron carbide active sites for long-chain hydrocarbon synthesis [104].

Space Velocity
In addition to the reaction temperature and pressure, space velocity is also an important factor because it affects the residence time of the reactants and thus the catalytic performance of the catalysts.As shown in Figure 9c, the CO 2 conversion sharply decreased with increasing space velocity, which is because of the insufficient adsorption and activation of CO 2 within a short contact time.Moreover, with the increase in space velocity from 2400 to 120,000 mL•g cat −1 •h −1 , the product distribution shifted to short-chain hydrocarbons.This is mainly because the short contact time between reactants and Fe active sites inhibits the secondary hydrogenation reaction.

H 2 /CO 2 Ratio
CO 2 is chemically very stable; therefore, increasing the active component H 2 can promote CO 2 activation and subsequent conversion.As seen in Figure 9d, the CO 2 conversion increased markedly from 21% to 42% over the Na-Zn-Fe catalysts, along with a significant decrease in the CO selectivity from 30% to 11% as the H 2 /CO 2 ratio increased from 1 to 4, but a H 2 /CO 2 ratio that is too high has an obvious effect on the hydrocarbon distribution.Increasing the H 2 /CO 2 ratio resulted in a decrease in C 5 = -C 12 = selectivity from 41% to 32% (Figure 9e), which was mainly caused by the subsequent hydrogenation of olefins under a high H 2 atmosphere [105,106] considering the situation that the value of O/P decreased from 11 to 8.4 [40].Therefore, an appropriate H 2 /CO 2 ratio should be determined with reference to CO 2 conversion and long-chain hydrocarbon selectivity.

Mixing Mode of Different Active Components
The mixing mode of different active components has significant effects on the catalytic performance for CO 2 hydrogenation.As shown in Figure 9f, the K-Fe/C and K-ZSM-5 catalysts in the granule-mixing mode exhibited an excellent liquid fuel selectivity of 70.1%, where the olefin intermediates formed over the K-Fe/C active sites underwent a series of reactions such as oligomerization, isomerization, and aromatization over the acid sites of K-ZSM-5 due to an appropriate distance between the different active sites.However, compared with the granule-mixing mode, the K-Fe/C and K-ZSM-5 catalysts in the powdermixing mode exhibited lower CO 2 conversion and C 5+ hydrocarbon selectivity.This is because K promoters on the K-Fe/C catalyst poison the acidic sites of zeolites due the close proximity, which lessens the synergistic effects [69].

Hydrogenation of CO 2 to Aromatics over FeO x /Zeolite Tandem Catalysts
Aromatics, as important raw materials, can be used to prepare polymers, medicines, paints, and so on.With the rapid social and economic development, the demands for aromatics are increasing.Generally, the traditional aromatic production process is heavily dependent on fossil energy, which results in considerable CO 2 emissions [107-109].Therefore, hydrogenation of CO 2 to aromatics is regarded as a promising way for the sustainable synthesis of aromatics [108].
The reaction pathway for hydrogenation of CO 2 to aromatics via the FTS route is illustrated in Figure 10, coupled with Fe-based catalysts and zeolites, named as FeO x /zeolite tandem catalysts hereafter.Then, olefin intermediates form over the Fe active sites by RWGS and FTS (Equations ( 2) and ( 6)), and can be further converted into aromatics by aromatization over the acid sites of zeolites (Equation ( 7)) [60,104].As also mentioned above, the Fe-based catalysts generally have both RWGS and FTS activity [110], and after appropriate modification (addition of promoters, introduction of supports, and improvement of preparation methods), they have low CH4 and high olefin selectivity under high temperatures.Therefore, this part focuses on the discussion of zeolite.Herein, the recent progress of the tandem catalysts for the hydrogenation of CO 2 to aromatics is described below in detail, and some representative catalysts as well as their catalytic performances are listed in Table 4.

Strategies for Optimizing Product Distribution of Fe-Based Active Components
As mentioned above, for hydrogenation of CO 2 to aromatics via the FTS route, CO 2 was converted to olefin intermediates by RWGS and FTS over the Fe-based active component, and olefins diffuse to the pores of zeolites for the formation of aromatics [119].Therefore, high olefin intermediate selectivity is a crucial factor to promote the formation of aromatics.However, pure Fe-based catalysts generally have some disadvantages in CO 2 hydrogenation, such as a low yield of olefins and rapid deactivation of Fe active species [120].To address these issues, several strategies such as the addition of metal promoters, the selection of suitable catalyst supports, and the optimization of catalyst preparation methods have been widely investigated to further enhance CO 2 conversion and to improve olefin selectivity, which will be summarized below in detail [121].

Metal Promoters
The addition of alkali metals (Na, K, etc.) can promote CO 2 adsorption and its subsequent conversion [46,122].In addition, the alkali metals also facilitate the reduction of iron oxides and subsequent formation of iron carbides (FTS active sites) [123].As shown in Figure 11a, Ma et al. prepared a Na-Fe 3 O 4 catalyst [111] and found that with the increase in the Na content from 0% to 3%, the CO 2 conversion increased from 15.2% to 26.7% and the C 2 = -C 4 = selectivity increased from 6% to 23%.This is mainly because Na, as an electronic promoter, can transfer electrons to the surface of Fe species, which favors CO 2 adsorption and inhibits H 2 adsorption, therefore generating a high surface C/H ratio and facilitating olefin formation.Then, coupling of the 1Na/Fe catalysts with HZSM-5 (30) can obtain an aromatic selectivity of 63.5% (Table 4, Entry 2).In addition, the addition of transition metals such as Zn [124] and Cu [51] has been widely used for the modification of Fe-based catalysts, which facilitates the reduction and dispersion of Fe active species.For example, Liang et al. investigated the influence of the Zn content on the catalytic performance of the K-nFe/Zn catalysts (n = 0, 1, 3, and 6 refer to the surface molar ratios of Fe to Zn) [96].The results indicated that the appropriate loading of Zn (n = 1, 3) can promote the formation of Fe active sites.However, excessive Zn (n > 3) tends to form large ZnO clusters and/or relatively stable ZnFe 2 O 4 spinels, which are not conducive to the reduction of Fe active species.Moreover, Cu generally has good RWGS activity; therefore, the addition of Cu to Fe-based catalysts may over-promote CO formation.However, as shown in Figure 11b, Song et al. found that as the Cu concentration increased from 0% to 6.25%, CO selectivity decreased from 14% to 3.51%, and the aromatic selectivity increased from 32% to 57.02%.This was mainly attributed to the synergic effects between Cu and Fe species [115].However, as the Cu concentration further increased from 6.25% to 50%, the CO selectivity increased by 11.49% and aromatic selectivity decreased by 26.02%.Therefore, the addition of adequate metal promoters to Fe-based catalysts is crucial for improving CO 2 conversion and aromatic selectivity.

Catalyst Supports
The surface properties and pore structure of supports (e.g., oxides and carbon materials) have significant impacts on the catalytic performance of Fe-based catalysts.For the surface property, Dai et al. prepared a 15Fe-10K/alkaline-Al 2 O 3 catalyst and found that alkaline-Al 2 O 3 supports increased the dispersion of the Fe-K bimetals, which inhibited the agglomeration of Fe and K metal particles during calcination.In addition, alkaline-Al 2 O 3 also promoted CO 2 adsorption and subsequent conversion, which hindered H 2 adsorption.Therefore, the catalyst surface had a low H 2 /CO 2 ratio and promoted light olefin formation [95].The pore structure of the supports is also a crucial factor for the dispersion of Fe active species.Wang et al. prepared honeycomb-structured graphene (HSG)-supported FeK1.5 catalysts (FeK1.5/HSG)[112,125].As shown in Figure 11c, the C 2 = -C 4 = selectivity reached 59% over the FeK1.5/HSGcatalysts.Apart from the promotional effect of the K promoter, the interaction between the HSG supports and Fe active species also promoted olefin formation.Moreover, the FeK1.5/HSGcatalysts coupled with the HZSM-5(50) zeolites had an aromatic selectivity of 68%.This can be explained as follows: the large pore size of HSG promotes the free diffusion of reactants and products, which inhibits the secondary hydrogenation of C 2 = -C 4 = , and therefore a number of light olefins can diffuse to HZSM-5 and be converted to aromatics by aromatization [126].

Preparation Methods
Preparation methods also affect the surface morphology and the property, dispersion, and structure of the Fe active species, resulting in different catalytic performances of Febased catalysts.For example, Cui et al. synthesized ZnFeO x -nNa nanocatalysts by a one-pot method [60].In the process of catalyst preparation, NaOH was used as both a precipitating agent and a promoter source to achieve the preparation and modification of catalysts [24].The samples with various Na concentrations were obtained by controlling the amount of water in the filtration step.The C 2 = -C 4 = selectivity reached 34.8% over the ZnFeO x -nNa catalyst, which was beneficial to obtain an aromatic selectivity of 60% (Table 4, Entry 6).Wang et al. synthesized Fe-MOFs (MIL-88B-Fe) by HSM [113], then prepared Febased catalysts (Fe@C) through the pyrolysis of MIL-88B-Fe, and finally obtained Na-Fe@C catalysts by IWI methods.The transmission electron microscopy (TEM) image indicated that the Na-Fe@C catalyst, encapsulated by a few carbon layers, had a core-shell structure, which benefited good dispersion of Fe active species and suppressed their deactivation.Therefore, reasonable improvement of the preparation methods can enhance the catalytic activity of the catalysts.

Rational Modification of Zeolites
Zeolite is an important part of the tandem catalysts for the hydrogenation of CO 2 to aromatics, which plays a key role in the oligomerization, isomerization, and aromatization of light hydrocarbons.A large number of studies showed that the pore structure, surface acidity, and preparation methods of zeolites have significant effects on the catalytic performance of tandem catalysts [127], which will be summarized below in detail.

2) BAS of Zeolites
For the formation of aromatics, the BAS density of zeolites is a key factor, which is closely related to the Si/Al ratio of zeolites [128].Wei et al. studied the effects of the BAS density of HZSM-5 on the catalytic performance of CO 2 hydrogenation to aromatics [114] and found that with the increase in BAS density from 0 to 356 µmol•g -1 , the aromatic selectivity in total hydrocarbons greatly increased from 1.1% to 44.5%.However, it increased at a decreasing rate, and the aromatic selectivity remained stable when the BAS density exceeded 50 µmol•g -1 , indicating that the promotional effect of BAS density on aromatization is limited.As for hydrocarbon formation excluding aromatics, with the increase in BAS density from 0 to 356 µmol•g −1 , C 2 0 -C 4 0 paraffin selectivity monotonously increased by 22.7%, while C 5+ hydrocarbon selectivity decreased by 9.7%.This is attributed to the cracking of heavy hydrocarbons due to high BAS density.Therefore, a moderate BAS density (i.e., suitable Si/Al ratios) is important for improving the aromatic selectivity and product distribution [118].
(3) Preparation Methods Preparation methods generally have significant impacts on the pore structure and surface acidity of zeolites, therefore affecting the catalytic performance of tandem catalysts.For example, Cui et al. prepared an HZSM-5 zeolite by a seed-assisted method without using organic templates [60].Compared with commercial HZSM-5, the as-prepared HZSM-5 had large amounts of mesopores with a suitable BAS density, which facilitated the rapid diffusion of olefin intermediates and/or aromatics, therefore suppressing the formation of coke and improving the catalyst stability.In addition, Song et al. [115] prepared HZSM-5 by various methods such as phase transfer (pt) [129], dry-gel conversion (dg) [130], and hydrothermal synthesis (hy) [131] and found that the highest BTX aromatic selectivity of 54.18% was obtained over 6.25Cu-Fe 2 O 3 /HZSM-5-pt catalysts, as seen in Figure 13.This is because HZSM-5-pt has abundant mesopores of 5-12 nm, which are favorable for the formation of BTX.Therefore, the formation of aromatics can be effectively regulated by adjusting the pore structure and surface acidity through proper preparation methods.Therefore, as mentioned above, the types, preparation methods, and BAS density affect the acidity of zeolites.For hydrogenation of CO 2 to aromatics via the FTS route, olefin aromatization is the typical acid-catalyzed reaction.The proper acidity and density can provide active sites for cracking, oligomerization, aromatization, and isomerization of olefin intermediates.However, excessive intensity and density of acid sites can lead to low paraffin formation and/or carbon deposition, which is not conductive to generating aromatics and preparing highly stable catalysts [132].

Strategies for Regulating the Pore Structure and Surface Acidity
The modification of HZSM-5 is mainly to mediate surface acidity, regulate the pore structure, and improve carbon accumulation resistance.In recent years, post-treatment methods such as alkali treatments [132], cationic modification [127], and SiO 2 coating [104] have been widely used to obtain mesopores and suitable BAS density and to passivate the external BAS of HZSM-5. (

1) Alkali Treatments
In general, the as-purchased commercial HZSM-5 only contains micropores, which hinders the rapid mass transfer inside the internal pores of HZSM-5 and causes its easy deactivation [133].However, HZSM-5 under alkali treatment is beneficial for generating a large number of mesopores and tuning its acidity by redistributing the framework of Al [134].In general, NaOH solution is used for alkali treatment of HZSM-5.For example, Wang et al. investigated the effects of alkali treatment on the catalytic performance of Na-Fe@C/HZSM-5 tandem catalysts and found that as the concentration of the NaOH solution increased from 0 M to 0.2 M, the aromatic selectivity increased from 30.9% to 50.2%.This is due to the synergistic effects between the appropriate amount of BAS and the hierarchical structure, promoting the formation of aromatics.In addition, the hierarchical structure is conducive to increasing the mass-transfer efficiency, thus inhibiting carbon deposition and extending the lifetime of the catalysts [113].In addition, Zhang et al. studied the acidity of alkali-treated HZSM-5 [135] and found that the weak, strong, and total acidity increased after alkali treatment, as shown in Figure 14a.This is due to the fact that the alkali treatment selectively removes the framework of Si, but the Al remains.Compared with commercial HZSM-5, the Lewis acid site (LAS) density of alkali-treated HZSM-5 increased at the cost of BAS density reduction, as shown in Figure 14b.Furthermore, the ZSM-5(50) after alkali treatment had a higher BTEX yield and better stability than commercial HZSM-5(50).In more detail, when the TOS ranged from 0 h to 48 h, the BTEX yield over ZSM-5(50) in the alkali treatment decreased by 5.1%, whereas that over commercial HZSM-5(50) decreased by 15.7%.This can be mainly attributed to the improved total acidity, the increased L/B ratio, and the large amounts of mesopores generated after alkali treatment.(2) Modification of HZSM-5 Modification of HZSM-5 via cation substitution (Ga, P, Zn, etc.) is a common approach for tuning its surface acidity, which can be realized by using impregnation or ion exchange methods.More specifically, Dai et al. prepared P-modified HZSM-5 by IWI methods and studied the catalytic performance of Fe-K/alkaline-Al 2 O 3 and P/ZSM-5 with different P loadings [95].They found that as the P loadings increased from 0 wt% to 0.8 wt%, the peak of B acid decreased (Figure 15a), and the aromatic selectivity increased by 2.5%.This can be explained as follows: the P modification increases the amount of medium-strength acid sites, and a moderate P loading is important for the improvement of the acidity and density of zeolites, which can promote the formation of aromatics [136], which promoted the aromatization of ethylene to BTEX aromatics [137].Moreover, a Zn loading of 0.5 wt% significantly improved the yield of BTEX aromatics (Figure 15b,c), whereas when the Zn loading further increased to 2 wt%, it had negative effects on catalyst stability.Meanwhile, compared with Zn and P, Ga-modified HZSM-5 has a stronger dehydrogenation capacity, which favors paraffin dehydrogenation and subsequent aromatization [138][139][140].The Zn doping level represents the mass ratio of Zn/Fe, including 0.5%, 1%, and 2%; P-parents; AT-Alkali treatment.Reprinted with permission from Zhang et al. [136].Copyright (2019) Elsevier.
(3) SiO 2 Coating Among various aromatics, BTEX aromatics have a higher added value and are widely used as basic raw materials in the current petrochemical industry.Many studies have shown that BTEX aromatics formed on the internal surface acid sites can be further converted to their isomers or even heavy aromatics on the external surface acid sites of HZSM-5 [138].For example, p-xylene (PX) can be converted to o-xylene (OX) and m-xylene (MX) by isomerization [141,142].Therefore, in order to enhance the BTEX selectivity, it is necessary to passivate the external BAS of HZSM-5, thus inhibiting the side reactions such as isomerization and over-aromatization [143], which can be realized by coating SiO 2 on the external surface of HZSM-5.For example, Sibi et al. prepared a Na-FeAlO x/ Zn-HZSM-5(12.5)@SiO 2 tandem catalyst, in which SiO 2 coating over the external surface of HZSM-5 was conducted by using the chemical liquid phase deposition (CLD) method [116,144].The results indicated that SiO 2 coating passivated the external BAS of HZSM-5 and successfully inhibited the alkylation of BTEX aromatics.In addition, the content of SiO 2 has significant impacts on BTEX aromatic selectivity, which can be controlled by adjusting the number of silanization cycles.After one silanization cycle, the BTEX aromatics significantly increased from 5% to 22.8%.However, on the contrary, over silanization (three cycles) resulted in pore blockage and subsequently hindered the diffusion of reaction intermediates into the micropores and inhibited the formation of BTEX aromatics.Therefore, a suitable SiO 2 coating content is crucial to promote the BTEX formation of aromatics.

Effect of the Reaction Conditions on the Catalytic Performance
For hydrogenation of CO 2 to aromatics, reaction conditions, including the reaction temperature, reaction pressure, space velocity, H 2 /CO 2 ratio, and the mass ratio of different active components and mixing mode, significantly affect the catalytic performances (CO 2 conversion, CO selectivity, aromatic selectivity, BTEX selectivity, etc.).The specific content will be introduced below in detail.

Reaction Temperature
In general, hydrogenation of CO 2 to aromatics has a proper range of reaction temperatures.That is, too low and too high temperatures are not preferred [111].More specifically, too low temperature is not conducive to CO 2 activation [145], whereas too high temperature may result in the hydrocracking of long-chain hydrocarbons to light paraffins.For example, Song et al. investigated effects of the reaction temperature on the CO 2 conversion and product distribution over 6.25Cu-Fe 2 O 3 /HZSM-5-c catalysts [115].As shown in Figure 16a, when the reaction temperature increased from 280 • C to 320 • C, the aromatic selectivity increased by 32.46%, whereas as the temperature further increased to 360 • C, the aromatic selectivity decreased by 27.61%, with an increase in CH 4 and C 2 0 -C 4 0 selectivity of 10 and 16%, respectively.Therefore, a suitable reaction temperature is favorable for improving the product distribution.

Reaction Pressure
As mentioned above, during the hydrogenation of CO 2 to aromatics, the olefin intermediates formed on Fe-based catalysts diffuse to the pores of HZSM-5, which are subsequently converted to aromatics by oligomerization, cyclization, dehydrogenation, and aromatization [104].During the process of aromatization, the oligomerization and cyclization reactions decrease in moles, whereas the dehydrogenation reaction increases in moles.Therefore, the effect of the reaction pressure on the hydrogenation of CO 2 to aromatics is rather complex [146].For example, Wang et al. prepared FeK1.5/HSG|HZSM-5(50) catalysts and investigated the effects of pressure on the aromatic selectivity and space time yield of the aromatics (STY aro ) [53].As shown in Figure 16b, the aromatic selectivity and STY aro increased significantly by 22 and 7.4% when the pressure increased from 1.0 MPa to 2.0 MPa.It can be explained that the increase in pressure benefits the oligomerization and cyclization reactions, therefore promoting the formation of aromatics.However, on the contrary, as the pressure further increased to 3 MPa, aromatic selectivity and STY aro decreased by 12 and 1.9%.Generally, dehydrogenation is a rate-limiting step during the process of olefin aromatization [147].Therefore, too high pressure is not conducive to dehydrogenation and the formation of aromatics, indicating that a suitable pressure is important for improving aromatic selectivity and STY aro .

Space Velocity
In addition to the reaction temperature and pressure, space velocity is also a vital factor, which determines the contact time between reactants and catalysts [128].Sibi et al. prepared a Na-FeAlO x /Zn-HZSM-5(12.5)@SiO 2 catalyst and studied the effects of space velocity on aromatics and BTEX selectivity in detail [116].When the space velocity increased from 2000 mL•g cat −1 •h −1 to 4000 mL•g cat −1 •h −1 , the aromatic and BTEX selectivity increased by 5% and 6%, respectively, which is probably because a long residence time results in the secondary hydrogenation of olefins.However, when the space velocity continued to increase to 6000 mL•g cat -1 •h -1 , the aromatic and BTEX selectivity decreased by 33.8 and 18.7%, respectively.It can be explained by the fact that the reaction intermediates formed at the Fe-based active sites cannot effectively diffuse into the pores of HZSM-5 in time due to the short residence time.Therefore, a moderate space velocity is essential for improving aromatic and BTEX selectivity.

H 2 /CO 2 Ratio
As mentioned above, the addition of active H 2 benefits CO 2 activation and its subsequent conversion and affects the secondary hydrogenation of olefins [105].As shown in Figure 17, as the H 2 /CO 2 ratio increased from 2 to 6, the CO 2 conversion increased by 30% and the CO selectivity decreased by 7.5% over the ZnFeO x -4.25Na/S-HZSM-5 catalyst.This is because abundant H 2 promotes the hydrogenation of CO 2 reactants and/or CO intermediates [106], but a H 2 /CO 2 ratio that is too high has an obvious effect on aromatic selectivity.More specially, increasing the H 2 /CO 2 ratio from 2 to 6 resulted in a decrease in aromatic selectivity from 67.2% to 47.4%, which is due to the secondary hydrogenation of olefins under high H 2 pressure [60].Therefore, an appropriate H 2 /CO 2 ratio is important to balance the activation capacity of CO 2 and product selectivity.

Mass Ratio and Mixing Mode of Different Active Components
The mass ratio of different active components (zeolites/Fe oxides) and mixing mode have significant effects on the catalytic performance [64,112].For the mass ratio, a high mass ratio is favorable for improving aromatic and PX selectivity, whereas too high mass ratios may result in isomerization and alkylation of PX.Gao et al. prepared a 2.83Na-FeMn (90/10)/HZSM-5@S1-S catalyst and studied the effects of different mass ratios on the aromatic and PX selectivity.When the mass ratio increased from 1 to 2, the aromatic and PX selectivity increased by 7.7 and 3.5%, respectively, while, as the mass ratio further increased to 3, the PX selectivity decreased by 0.6% [117].Therefore, an appropriate mass ratio can effectively promote the generation of high-value aromatics.
For mixing mode (i.e., powder-mixing, granule-mixing, granule-mixing with quartz sand, and dual bed modes) [148,149], as shown in Figure 18, the 6.25Cu-Fe 2 O 3 /HZSM-5-c tandem catalyst in the granule-mixing mode showed high CO 2 conversion (57.3%) and aromatic selectivity (56.61%), whereas the powder-mixing mode had relatively lower CO 2 conversion (44%) and aromatic selectivity (30%).This may be due to the strong interaction between the acid sites of HZSM-5 and the active sites of the 6.25Cu-Fe 2 O 3 surface for the powder-mixing mode.More specially, the acid sites of HZSM-5 poison the alkali sites on the 6.25Cu-Fe 2 O 3 surface, reducing the surface basicity and acidity of the 6.25Cu-Fe 2 O 3 and HZSM-5, respectively, therefore suppressing CO 2 hydrogenation and olefin aromatization [115,150].In addition, a further increase in the distance between the two active components by the dual bed mixing mode led to the deteriorated catalytic performance with CO 2 conversion (37%) and aromatic selectivity (32%).This can be attributed to the distance that hinders the intermediate transfer and weakens the thermodynamical driving force.Therefore, an appropriate mixing mode of Fe-based catalysts and HZSM-5 can exert effective synergistic effects, which contribute to improving the CO 2 conversion and aromatic selectivity [148].

Reaction Mechanism 4.1. Mechanistic Study of the Hydrogenation of CO 2 to Liquid Fuels
The relationship between active sites and mechanisms remains controversial despite that many excellent catalysts have been developed for the hydrogenation of CO 2 to liquid fuels.According to the most-accepted reaction mechanism [151,152], the mechanism for RWGS can be identified as either the decomposition of HOCO* species or the direct cleavage of C=O bonds to generate CO* [153].During the FTS process, the main steps include the formation of CH x * monomers (x = 0-3) with hydrogen-assisted insertion into dissociated CO* [154], the initiation of chain growth through the coupling of CH x *, and the termination of chain growth through further hydrogenation, dehydrogenation, or CO insertion (Figure 19), leading to the formation of paraffins, olefins, and oxygenates.Secondly, two possible mechanisms are proposed for the carbon chain growth based on the FTS reaction [158]: the "carbide" mechanism [159] and the "CO insertion" mechanism [160,161] (Figure 21).The "carbide" mechanism involves the formation of CH x * species through CO* dissociation and subsequent chain growth by CH x * species insertion.First, the CH x * species produced from CO* dissociation initiate the entire reaction, and then carbon chain growth is achieved by the insertion of subsequent CH x * species, for which the CH* and CH 2 * are identified as key CH x * species, whereas the "CO insertion" mechanism realizes carbon chain growth by the direct insertion of undissociated CO into CH x * species.Herein, CH x * species initiate the whole reaction, and subsequent CO* is inserted into the generated CH x * species, and then the initial C 2 H x * species are formed through the C-O bonds cleavage of C 2 oxygenic intermediates, and further carbon chain growth continues by continuously inserting CO* into the C n H x * species [160,162].According to DFT calculations, Fe-Cu bimetallic catalysts generally follow the "carbide" mechanism [157], and Co-based catalysts generally follow the "CO insertion" mechanism [163,164].Therefore, both the "carbide" and "CO insertion" mechanisms are applicable to the hydrogenation of CO 2 to liquid fuels, which depends on the types and properties of the used catalysts.

Mechanistic Study of the Hydrogenation of CO 2 to Aromatics
The hydrogenation of CO 2 to aromatics via the FTS route mainly uses tandem catalysts composed of Fe-based active components and HZSM-5.In general, Fe 3 O 4 and Fe x C y are considered active sites for RWGS and FTS, respectively, while aromatization of olefin intermediates mainly occurs on the acid sites of HZSM-5.For example, Gascon et al. [118] proposed a reaction pathway for the hydrogenation of CO 2 to aromatics over Fe 2 O 3 @KO 2 /HZSM-5 tandem catalysts, as illustrated in (Figure 22).In more detail, as shown in Figure 22a, CO 2 is initially converted to CO via RWGS over the Fe 3 O 4 active sites.Next, a portion of as-formed CO is directly converted into hydrocarbons over the Fe x C y active sites, and the remaining portion of CO diffuses into the confined pores of HZSM-5 to form surface-formate species, which are subsequently converted to surface-methoxy species (SMS) in situ via hydrogenation [165].Then, SMS is further converted into ethylene via a series of reactions including CO insertion, hydrogenation, and deprotonation as seen in Figure 22b.Finally, the olefin intermediates formed via the aforementioned two pathways are converted into aromatics by oligomerization, cyclization, hydrogen transfer, and aromatization on HZSM-5 as shown (Figure 22c).
Sibi et al. used situ operando diffuse reflectance infrared Fourier-transform spectroscopy (DRIFTS) experiments to obtain a plausible reaction for the formation of ethylene species over HZSM-5(12.5)zeolite [116].As the pressure increased from 1 to 3 MPa, CO 2adsorbed species (b-CO 3 2-[166], HCO 3 - [167,168], and -OH), L-CO, water, and CH 4 were observed.Moreover, peaks associated with ethylene vibrations (3125 cm −1 , 3079 cm −1 , 2989 cm −1 , 2048 cm −1 , 1889 cm −1 , and 1445 cm −1 ) were also observed.This suggests that the surface -OH groups of pristine H-ZSM5(12.5)promote the RWGS and C-C coupling reactions to a certain degree.In detail, the silanol groups and the extra framework aluminum (EFAL) species could be regarded as the adsorption sites of CO 2 .Over the EFAL site, bicarbonate species can be converted to formate and dissociate to generate CO* and *OH.Then, the hydrogenation of CO* with the hydrogen from the adsorbed H species in HZSM-5(12.5)produces CH x species, and subsequent C-C coupling could finally produce ethylene.However, Wang et.al.reported that the synergistic effects between BAS and LAS also play a vital role in enhancing the formation of aromatics.Taking the aromatization of cyclohexene as an example [113], the cyclohexene is able to form carbenium ions on the LAS by dehydrogenation.Then, the hydride combines with the acidic proton in the BAS, releasing H 2 .Meanwhile, the cyclic carbocation undergoes further dehydrogenation, leading to the formation of cyclohexene and the recovery of BAS.After this process is repeated, aromatics are finally produced.In addition, Song et al. found that the synthesis of aromatics involving dehydrogenation and cyclization of light olefins resulted in abundant H species [169].If these H species cannot be consumed or diffuse out immediately, then the high concentration of H present in the system will inhibit the aromatization of olefins [170].These authors proposed a "H recycling" mechanism, as shown in Figure 23 [115], in which the H species formed during the aromatization of olefins on the acid sites of HZSM-5 can diffuse to the surface of the Fe active sites and react with the absorbed CO 2 on the oxygen vacancy.In other words, Fe active species act as "H acceptors" to hold H species that originate from HZSM-5.Meanwhile, CO 2 acts as a "H consumer" to promote the formation and subsequent aromatization of olefins.

Conclusions and Perspective
To mitigate climate change and achieve carbon neutrality, the hydrogenation of CO 2 to liquid fuels and aromatics is considered to be a promising and sustainable technology that can effectively reduce CO 2 emissions and convert waste CO 2 into high value-added products.In this review, we have critically and comprehensively summarized the recent research progress of the development of efficient Fe-based catalysts for the hydrogenation of CO 2 to liquid fuels and aromatics via the FTS route.
For the hydrogenation of CO 2 to liquid fuels, the alkali promoters, secondary metals, catalyst supports, and preparation methods have significant effects on the catalytic performances.In more detail, alkali promoters can effectively enhance the capacity of carbon chain growth and improve olefin selectivity.The addition of secondary metals can form specific bimetallic structures, which promote the formation and stability of Fe active species, thus facilitating carbon chain growth via C-C coupling.Catalyst supports with a high specific surface area generally benefit the dispersion of Fe species; however, the interaction between Fe active components and supports should be properly regulated because too strong a metal-support interaction is not conducive to the catalytic performance.The preparation methods can effectively regulate the physical properties of the catalysts that are closely associated with the efficient production of liquid fuels.In addition, tandem catalysts using the acidity and porous properties of zeolites can further enhance the selectivity and octane number of liquid fuels.
The formation and subsequent aromatization of olefin intermediates are critical for the hydrogenation of CO 2 to aromatics over the FeO x /zeolites tandem catalysts, where zeolites play a key role in aromatization.Herein, HZSM-5 is preferred due to the suitable surface acidity and pore structure; however, it is easy to form heavy aromatics and deactivate due to carbon deposition.To address these issues, a series of strategies, such as alkali treatment, cationic modification, and SiO 2 coating, have been applied to regulate the surface acidity and pore structure of HZSM-5.More specially, alkali treatment regulates the BAS and LAS density by redistributing the framework of Al and generates a hierarchical pore structure, which promotes the formation of aromatics and catalyst stability.Moreover, cationic modification (Zn, P, Ga, etc.) can increase the LAS density, which facilitates direct dehydrogenation, therefore promoting the formation of olefin intermediates and targeted aromatics.In addition, SiO 2 coating passivates the external BAS of HZSM-5 and effectively inhibits the alkylation and isomerization of aromatics.
However, there are a variety of challenges in the preparation of catalysts for CO 2 hydrogenation to liquid fuels and aromatics.We should pay sufficient attention to catalyst design in order to achieve a balance of synergistic effects between alkali promoters and secondary metals towards liquid fuel synthesis or Fe-based active components and the acid sites of zeolites towards aromatic synthesis through the combination of preparation methods from multiple fields and disciplines.Apart from developing efficient catalysts for suitable reaction conditions, we should also focus on the optimization or design of new reactors to facilitate the CO 2 conversion and remove the by-product, i.e., water [11].For example, the membrane reactor with selective permeation membranes can selectively filter water [171], thereby breaking the reaction equilibrium and promoting the forward process of CO 2 hydrogenation, thus greatly improving the CO 2 conversion and long-chain hydrocarbon (i.e., liquid fuels and aromatics) selectivity.In addition, with the development of computational tools relevant for catalysis, it will become an important research direction to further study the reaction mechanisms combining DFT calculation results and existing experimental data and to predict the relationship between the properties (the particle sizes and distribution of Fe-based catalyst active components, the pore sizes of supports) and catalytic performances (activity, selectivity, and stability) of catalysts.In addition to the various challenges discussed above, H 2 production technologies have also received considerable attention.Generally, water electrolysis using renewable energies is considered a potential and sustainable technology to produce green H 2 , whereas the technology still faces challenges in terms of technical maturity and cost.However, with the development of efficient water electrolysis technologies and large-scale deployment of renewable energies, it is expected that CO 2 hydrogenation technologies for value-added chemicals and fuels using green H 2 will become more feasible and competitive.

Figure 19 .
Figure 19.Scheme of CO 2 -modified FTS-based catalytic mechanism.Reprinted with permission from Ye et al. [155].Copyright (2019) Springer Nature.The formation of CH x * monomers and the growth of the carbon chain[156] directly affect the hydrogenation of CO 2 to liquid fuels.Firstly, the formation of CH x * monomers is related to the composition of catalysts.For example, Nie et al. studied CO 2 hydrogenation on the surfaces of pure Fe and bimetallic Fe-Cu catalysts with DFT calculations and identified favorable pathways through the HCOO* intermediates (CO 2 * → HCOO* → HCOOH* → HCO* → HCOH* → CH*) rather than the formation of CO* (CO 2 * → CO* → HCO* → HCOH* → CH*).This is attributed to the bimetallic synergetic effects of Fe-Cu, in which the addition of Cu reduces the barrier for C-C coupling and facilitates carbon chain growth (Figure20)[157].Therefore, the addition of secondary metals not only enhances CO 2 conversion but also changes the pathway of CO 2 hydrogenation.

Figure 20 .
Figure 20.Reaction networks examined to identify energetically favorable C 1 species from CO 2 hydrogenation on the Fe (100) and Cu-Fe (100) surface.Activation barriers are given in eV (the networks connected with red arrows represent the preferred path for CO 2 conversion to CH*).Reprinted with permission from Nie et al. [157].Copyright (2017) American Chemical Society.

Figure 21 .
Figure 21.Schematic illustration of the "carbide" and "CO insertion" mechanisms proposed in CO-FTS for hydrocarbon synthesis.Reprinted with permission from Nie et al. [20].Copyright (2019) published Elsevier.Reprinted with permission from van Santen et al. [158].Copyright (2013) The Royal Society of Chemistry.

Figure 23 .
Figure 23.Schematic illustration of the "H recycling" mechanism for CO 2 hydrogenation to aromatics between oxide and zeolite.Reprinted with permission from Song et al. [115].Copyright (2020) American Chemical Society.

Table 1 .
Representative catalysts and their performance for the direct hydrogenation of CO 2 to liquid fuels.
a Meaning paraffins(o) and olefins (=) in C 2-C 4, respectively , b Olefins/paraffins ratio of C 2 -C 4 products, c Including a small amount of alcohol, d C 1 -C 4 , e Containing 0.03% sodium by mass fraction, f Gas volume was calculated based on the ideal gas law, g N.A.:Not available, h Oxy = oxygen compounds (MeOH + Me 2 O).

Table 2 .
Representative catalysts and their performance for direct hydrogenation of CO 2 to iso-paraffins.
successfully prepared Fe-Cu catalysts derived from delafossite-CuFeO 2 precursors.Compared with pure Fe 2 O 3 , CuO-Fe 2 O 3 mixture, and spinel CuFe 2 O 4 catalysts, the Fe-Cu catalysts exhibited the optimum performance with CH 4 and C 5+ hydrocarbon selectivity of 64.9 and 2.1%, respectively.It is because CuFeO 2 with the intermediate oxidation state of Cu + has a stronger reduction ability than Cu 2+ , which is conducive to the rapid reduction and carburization of Fe species to form Fe 5 C 2 active species.

Table 4 .
Representative catalysts and their performance for direct hydrogenation of CO 2 to aromatics.