Dielectric Barrier Discharge Plasma-Assisted Catalytic CO₂ Hydrogenation: Synergy of Catalyst and Plasma

Abstract

CO₂ hydrogenation is an effective way to convert CO₂ to value-added chemicals (e.g., CH₄ and CH₃OH). As a thermal catalytic process, it suffers from dissatisfactory catalytic performances (low conversion/selectivity and poor stability) and high energy input. By utilizing the dielectric barrier discharge (DBD) technology, the catalyst and plasma could generate a synergy, activating the whole process in a mild condition, and enhancing the conversion efficiency of CO₂ and selectivity of targeted product. In this review, a comprehensive summary of the applications of DBD plasma in catalytic CO₂ hydrogenation is provided in detail. Moreover, the state-of-the-art design of the reactor and optimization of reaction parameters are discussed. Furthermore, several mechanisms based on simulations and experiments are provided. In the end, the existing challenges of this hybrid system and corresponding solutions are proposed.

1. Introduction

With economic development, industrialization, and human activity, the continuous increasing emission of CO₂ has led to the increase of global temperature (i.e., global warming) which has an impact on the earth’s ecological environment, such as glacier melting and sea level rise [1,2]. One way to reduce the CO₂ content in the atmosphere is to capture it with adsorbents (e.g., basic solvents and active carbons) [3]. However, major concerns regarding CO₂ capture and storage lie in separation efficiency, operation costs, and long-term stability [4,5]. In comparison, transformation of cheap and abundant CO₂ to value-added products (e.g., CH₄, CH₃OH, C₂–C₄ hydrocarbons) via hydrogenation has drawn tremendous attentions [1,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23]. In some works, CO₂ splitting (dissociation) and reversed water–gas shift (RWGS) are considered a kind of hydrogenation process; however, a reaction involving both C–O bond breaking and C–H bond formation will be mainly covered in this review, such as methanation (Equation (1)) and methanol production (Equation (2)). Notably, CO₂ dissociation and RWGS might be discussed as an intermediate step or side reaction as well.

$${\mathrm{CO}}_2+4{\mathrm{H}}_2\leftrightarrow {\mathrm{CH}}_4+2{\mathrm{H}}_2\mathrm{O} \Delta {\mathrm{H}}_{298\mathrm{K}}=-252.9\mathrm{kJ}\Delta {\mathrm{mol}}^{-1}$$

(1)

$${\mathrm{CO}}_2+3{\mathrm{H}}_2\leftrightarrow {\mathrm{CH}}_3\mathrm{OH}+{\mathrm{H}}_2\mathrm{O} \Delta {\mathrm{H}}_{298\mathrm{K}}=-49.5\mathrm{kJ}\Delta {\mathrm{mol}}^{-1}$$

(2)

The possible reaction mechanisms of various hydrogenations of CO₂ are as below (Figure 1a,b). Obviously, based on their exothermic nature, these processes are thermodynamically favored at low temperatures; however, considering the eight-electron reduction of CO₂ and activation of H₂, a catalyst is necessary for the activation of CO₂ and H₂ to enhance the reaction kinetics [24]. Moreover, intensive conditions are necessarily applied to realize the industrial applications in conventional thermal catalysis. For example, in methanol production, a temperature of 200–300 °C and high pressure of 3–30 MPa are usually required [1,25,26,27,28]. Coupling with non-thermal plasma (NTP), especially dielectric barrier discharge (DBD)-generated plasma can form a synergy where both a mild reaction condition and high selectivity are achieved (Figure 1c). Specifically, the electric field can be enhanced by the catalyst and micro-discharges can be generated in the pores; on the other hand, the catalyst physicochemical properties and reaction pathways could be changed in the presence of plasma and an easier activation of reactant molecules is enabled [29,30,31,32,33,34,35,36]. In the plasma–catalyst hybrid system, a high conversion efficiency and selectivity can be obtained at a relatively milder reaction condition. For example, when Ru/Al₂O₃ was combined with DBD plasma, 12.8% CO₂ conversion and 73% CH₄ selectivity were delivered in methantion of CO₂ at only 25 °C [37]. While in the production of CH₃OH, the combination of DBD plasma and CuO/ZnO/Al₂O₃, the optimal reaction temperature was lowered by 120 °C compared with the catalyst only approach [26]. Several reviews have been reported in plasma-assisted catalytic CO₂ conversions [1,2,6,7,8,29,38,39]; however, a comprehensive and in-depth summary is rarely seen regarding the synergistic effects of catalyst–DBD plasma hybrid system on the catalytic performances and energy efficiency of CO₂ hydrogenation. In the following sections, an overview of the plasma (particularly DBD plasma) and a detailed description of the synergy will be provided in terms of structure–performance relationships. Subsequently, the effects of operation parameters and reactor designs are illustrated, followed by possible reaction mechanisms of CO₂ hydrogenation in this hybrid system. In the end, conclusive remarks and future works are proposed.

2. Overview of Plasma and Catalyst–Plasma Hybrid System

As a chemical mixture of cations, anions, molecules, atoms, excited species, and radicals which interact with each other, plasma is widely utilized in many aspects of material science, microelectronic industry, environmental applications, and medical treatments [42]. Rather than natural plasmas, two man-made plasmas are mainly involved in research, that is, completely ionized plasmas (fusion plasmas) and weakly ionized plasmas (gas discharges) [6]. The gas discharges can be further classified into two types depending on if the plasma is in thermal equilibrium, which are thermal plasmas in local thermodynamic equilibrium (LTE) and non-thermal plasmas (NTP) in non-LTE conditions [6]. Thermal plasmas enjoy many advantages, such as high energy density, high radiation intensity, and high temperature. However, their applications in CO₂ conversions are limited by the lower equilibrium efficiency and conversion than NTP. To create the NTP, a gas-filled reactor is inserted by two electrodes parallel to each other, where an electric field is generated by applying a potential difference. Some gas molecules can be broken into cations and electrons, and the accelerated electrons will collide with the gas molecules on the way to the anode, leading to ionization, excitation, and dissociation. New electrons could be released from the produced ions in previous ionization collisions at the cathode, and these electrons enable further collision and resulting ionization. Owing to this self-sustaining non-thermal plasma, gas molecules would be activated by high-temperature electrons, creating radicals and products [6]. The major merits of NTP in CO₂ conversions include the following points. First, the high-energy electrons created by the plasma in an electromagnetic field can initiate the CO₂ activation even at ambient conditions without the thermal energy input to heat up the reactor or reactants. Second, the operation flexibility (ability to be instantaneously turned on and off) allows the NTP technology to adapt to the intermittent renewable energy [38]. Third, the low cost of plasma reactors without the usage of rare earth materials improves scalability in household devices, on-demand installations, and large-scale plants [43,44]. Among the various discharges applied to create NTP—such as gliding arc discharges, glow discharges, microwave discharges, and DBD [1,38]—DBD is widely studied due to the generation of highly active species and high electron density [39], facile design, and operation in ambient conditions [38]. In the representative DBD plasma reactor (Figure 2a), the grounded electrode in the center is surrounded by the coaxial dielectric tube covered by stainless steel mesh and powered by the high potential. For comparison, a planar DBD reactor is shown in Figure 2b, where the dielectric barrier materials are in contact with the two electrodes at the top and bottom side. Despite the highly active electrons which activate the gas molecules in mild conditions, an uncontrollable recombination of the intermediates occurs, lowering the targeted product selectivity. Even worse, the formed end-products can be further destroyed by the electrons. This scenario becomes more obvious in CO₂ hydrogenation which is more complicated than CO₂ splitting where CO and O₂ dominate the products [6]. Therefore, a catalyst is necessarily coupled in the DBD reactor to realize both a high selectivity and low energy input. In turn, the catalyst lifespan would be prolonged under mild conditions.

In the catalyst-plasma hybrid system, prior to the adsorption on the catalyst surface, the CO₂ has become excited in the presence of DBD plasma, thus promoting the conversion to intermediates with a lower energy barrier than with thermal catalysis alone where a higher temperature is needed to activate the adsorbed ground-state CO₂ molecules [46,47]. Moreover, the surface reactions can be further facilitated due to the involvement of active species excited in the plasma—such as CH, OH, and CO radicals [48,49]. Furthermore, as previously mentioned, the interaction between catalysts and plasma would modify the surface properties of the catalysts (e.g., basicity, adsorption capacity) and the plasma characteristics [50,51,52]. In the following part, an in-depth and comprehensive discussion is provided regarding the applications of the above hybrid system in CO₂ hydrogenation and the corresponding structure–performance relationships.

3. Synergy in Catalyst–Plasma (DBD) Hybrid System

To realize the low-temperature and efficient conversion of CO₂ and high selectivity of targeted products via hydrogenation, the catalyst can be integrated with the DBD plasma reactor to generate a synergy, where the respective properties of active metals, supports, and promoters (additives) can be effectively combined.

3.1. Active Metal and DBD Plasma

When DBD plasma was applied in Pt/In₂O₃ catalytic CO₂ hydrogenation to methanol, a highly dispersed Pt nanoparticle was obtained, delivering a 37% CO₂ conversion and 62.6% methanol selectivity at 30 °C and 1 atm, much higher than that of conversional Pt/In₂O₃ (24.9% and 36.5%, respectively) (

Table 1. Catalytic CO₂ hydrogenation assisted with DBD plasma.

Catalyst	Targeted Product	H2/CO2	CO2 Conversion (%)	Selectivity (%)	Applied Voltage (kV)	Ref.
Pt/In2O3	CH3OH	3	37	62.6	13.3	[53]
Co/CeZrO4	CH4	4	70	100	20	[52]
Co/ZSM-5	C2-C4	3	45	13.7	10	[58]
Ni/Al2O3	CH4	4	60	97	10	[47]
Co/Al2O3	C2+	3	74	46.5	18.5	[59]
Ni/UiO-66	CH4	4	85	99	6.5	[60]
CeNi/Cs–USY	CH4	4	70	95	6	[61]
Ni-CeO2/Al2O3	CH4	4	70	96	7.7	[62]
Ni–La/Na–BETA	CH4	4	85	97	6	[63]
Ni–Fe/LDH	CH4	4	72	99	18	[64]
Ni–Na/CeZrOx	CH4	4	57.5	75.9	15	[65]

). This enhanced performance was mainly attributed to the stronger adsorption of CO₂ by small Pt nanoparticles (2.32 mmol/g) and lower energy barrier to initiate the hydrogenation in the presence of plasma [53]. Besides Pt, Pd is reported to be highly active in methanol production from CO₂ hydrogenation at low pressures [54]. When coupled with DBD plasma at a 30 W discharge power, CO₂ conversion increased from 17.1% to 32.5% due to the accelerated dissociation of gaseous and adsorbed CO₂ by plasma and strengthened adsorption of produced CO intermediate on Pd surface [6,55]. With the increase of Pd loading from 0.69 wt% to 2.29 wt%, the conversion of CO₂ increased by almost 100%. As another representative noble metal, Ru has been proven active for CO₂ hydrogenation in previous studies with HCOO– as the intermediate [56]. Interestingly, Ru³⁺ could be in-situ reduced to Ru metallic phase, which exhibited a highly selective methanation via CO₂ hydrogenation [57]. The combination of DBD plasma and Ru enabled the successive deoxygenation of CO₂ and further conversion to CH₄ [37].

In addition to noble metals, various transition metal catalysts have been widely studied as alternatives considering their low cost and high activity. Cu-based catalysts exhibit superior intrinsic activities towards methanol production from CO₂ hydrogenation [66,67,68,69]. Compared with Pt, Cu delivered a better performance owing to the moderate adsorption of oxygen species and bonding with the intermediates, despite the larger particle size than Pt [70]. When integrated with DBD plasma, a high methanol selectivity of 53.7% was achieved due to the simultaneously promoted adsorption and activation of CO₂ and conversion to CH₃OH. In another work, however, the addition of Cu caused a lower conversion of CO₂ resulting from the increased ionic conductivity and reduced dielectric constant, which enhanced the voltage needed for plasma ignition [71]. By replacing Cu with Mn, the energy efficiency was improved by 116% in DBD reactor (1620 µg/kJ). Moreover, CO₂ conversion was increased by 36% with DBD plasma [72]. Notably, when Cu and Mn were coupled, the side reaction WGS was inhibited and a facilitated formation of carbonate species as a result of CO₂ adsorption was observed on the catalyst surface [73].

Compared with Cu which binds strongly with CO₂ and CO, Co is more selective to produce CH₄ [24,74,75]. At the Cu surface, adsorbed carbonates and formates decomposed to form certain amount of CH₄; however, the main products were still CO derived from the RWGS of CO₂ even with plasma. On the contrary, the CH₄ selectivity was significantly increased from 25% to 88% at the Co surface with the help of DBD plasma (Table 1). In particular, triggered by the plasma electronic impact collisions, the gaseous CO dissociated from CO₂ was re-adsorbed onto the surface of Co metals, which further dissociated to O and C, subsequently reacting with the H atoms simultaneously generated from H₂ molecule decomposition to form CH₄. Owing to the inherent positive contribution of DBD plasma and Co metallic sites, the temperature to reach the maximum CH₄ yield was reduced by 40 °C [52]. In another scenario where the DBD plasma voltage was 13.6 kV, furnace temperature was 250 °C and H₂/CO₂ was 3, CO₂ was selectively converted to CH₄ on Co surface; interestingly, CH_x species formed in plasma were immediately terminated by the H species also produced in plasma, leading to a dominant amount of CH₄ and only a trace amount of longer-chain hydrocarbons [59]. Lan et al. [58] further compared a series of active metals in DBD plasma–catalytic production of CH₄ and lower hydrocarbons via CO₂ hydrogenation. With an input power of 14 W and H₂/CO₂ ratio of 3, Co was proven more active than Cu, Fe, and Mo in terms of the CO₂ conversion and hydrocarbon selectivity. In detail, over 40% conversion of CO₂ was achieved in Co catalyst while those of the counterparts were only less than 20% (Table 1). Moreover, the CH₄ selectivity of Co catalysts reached more than 70% while those of other metals were lower than 20%. These enhanced conversions of CO₂ to hydrocarbons were attributed to the synergistic effects of metals and plasma; specifically, the CO and H species produced in plasma were adsorbed at the metal surface where CO methanation was highly activated, contributing to the formation of CH₄ and CH_x products [76].

Similar to Co, Ni is also proven effective in CH₄ production via CO₂ hydrogenation [58]. When coupled with plasma, over 80% CO₂ conversion and 99% CH₄ selectivity can be achieved at low temperatures [77]. In another situation at 150 °C and 10 kV voltage, the Ni/Al₂O₃–DBD plasma hybrid system realized 60% CO₂ conversion and 97% CH₄ selectivity, increasing by 20 and 5 times in comparison to plasma alone (Table 1) [47]. After the reaction, the Ni size and dispersion was 6.3 nm and 42% in plasma respectively, better than that in thermal catalysis (8.1 nm and 24%) (Figure 3a), which was ascribed to the immediate capture of O radicals from CO₂ dissociation by the active H species to form OH and CHO radicals in plasma (Figure 3b) [37,78]. Rather than the smaller size and higher dispersion, the stabilized metallic Ni phase by excessive electrons and strong reducibility under plasma discharge was more crucial in this case [79]. The electron transfer over the Ni⁰ surface promoted the bond breaking and formation, pushing the forward reaction and enhancing the catalytic conversion and selectivity [79,80]. Besides the metallic Ni effect, the redox property and Lewis basic sites were improved due to the surface electrons, promoting the CO₂ adsorption which is a Lewis acid (Figure 3c) [79,81]. Moreover, the dissociation of CO₂ was facilitated when they were activated by plasma to form vibrational and excited states, intrinsically accelerating the CO production in a lower energy barrier [82]. The improved CO₂ activation could be signaled by the production of CO₂⁺, CO, CHO radicals, and C₂^– species; and the last two species were mainly responsible for the high conversion to CH₄ with surface-bound H at low temperatures in Ni/Al₂O₃ coupled with DBD plasma [83,84,85]. Interestingly, the insignificant enhancement of temperature (170 °C at the outlet and 150 °C at the reaction zone) and zero carbon formation excluded the overheating of bed materials by the plasma (thermal effect) or the resulting hotspots. Thus, the dominant driving force of the excellent activity was the stimulation by the high-intensity microdischarges in-between the reactor and beads, together with a new reaction pathway [86,87].

In a few other studies, however, the Ni particle size effect on the performance and the influence of plasma on the Ni size were emphasized. In Ni/Zeolite catalyst with DBD plasma for CO₂ methanation, the CO₂ conversion was greatly enhanced from 15% to 95% compared with catalyst alone since various reactive species produced by the plasma contributed to the C–O dissociation which determined the reaction rate in CO₂ hydrogenation. Moreover, the highly dispersed and small Ni nanoparticles further increased the conversion efficiencies [88]. To further explore the effect of Ni dispersion and location on the CO₂ conversion and CH₄ selectivity, Ni supported or encapsulated in silicalite-1 with various pore structures were tested in DBD plasma reactor [89]. Notably, those Ni active sites at the external surface of silicalite-1, or embedded in the hierarchical pore structures, exhibited better performances than those in microporous structures, which could be explained by the facilitated diffusion of short-lived reactive species generated in plasma (e.g., radicals and excited atoms/molecules) and high exposure of active centers where the possible deactivation of plasma-induced species was inhibited [90,91]. Apart from the effect of Ni size on the catalytic activity, the presence of DBD plasma relates strongly with the generation of Ni nanoparticles with different sizes. It is reported that less defective and smaller Ni sites were formed when interacting with the plasma [92]. Owing to the activation of DBD plasma, Ni dispersion was improved considerably and the resulting Ni particle size was only 4 nm in average, much smaller than that obtained in thermal catalysis (13 nm) [60]. More interestingly, when plasma was applied in the pre-treatment of Ni nitrate precursors, one-step reduction to Ni⁰ phase was realized in a fast decomposition rate, where the diffusion of Ni species into the MgAl₂O₄ support pores was alleviated, thus obtaining a weak metal–support interaction and better reducibility. Additionally, the nucleation and crystal growth of Ni were modified in the plasma. The formed abundant Ni active sites effectively catalyzed the methanation of CO as an intermediate derived from the formate decomposition [93].

3.2. Support and DBD Plasma

In the presence of DBD plasma, high energy species in abundance can be generated, thus lowering the activation energy and reaction temperature [71]. It is noteworthy that the catalyst itself may change the plasma properties, including micro discharge generation in the pores, discharge type modification, and electric field enhancement [94]. In addition to particle size (a small size usually requires a higher ignition voltage), the intrinsic properties of support materials (e.g., dielectric constant and ionic conductivity) determine the energy of plasma in a way that a high dielectric constant and low conductivity benefit a strong capability to store energy [71]. When Al₂O₃ was introduced to the DBD reactor, the dielectric constant was enhanced owing to the non-conductivity of Al₂O₃, leading to an increased energy of the ionized electrons at a lower breakdown voltage [86,95,96,97]. Benefiting from the denser plasma, reactive radicals and intermediates accelerated the transformation of CO to HCO^* and HCOO^*/H₂CO^*, contributing to the methanol production with a higher TOF [71]. In another scenario where C₂⁺ hydrocarbons were the targeted products, the Al₂O₃ packing in the DBD reactor promoted the chain-growth reaction of CH_x derived from the activation of plasma. Additionally, when the input power increased to 10 W from 4 W, 74% CO₂ conversion and 46.5% C₂⁺ selectivity were delivered at 25 °C (Table 1) [59].

Apart from Al₂O₃, transition metal oxide ZnO is proven effective in CO₂ hydrogenation assisted with DBD plasma. The partially reduced ZnO_x possessed abundant oxygen vacancies, increasing the amount of medium CO₂ activation [98,99]. Meanwhile, the desorption temperature was reduced in the presence of plasma [100]. On the other hand, the addition of ZnO enabled a lower ionic conductivity and higher dielectric constant, thus generating a denser plasma which benefited the CO₂ dissociation via electron impact activation [101].

Different from ZnO, CeZrO₂ is featured with the superior redox property and oxygen storage capacity, acting as a reservoir of CO^* and O^*, which were the dissociation products of CO₂ [88]. Accelerated by the plasma, the decomposition of adsorbed species and recombination with H to form CH₄ were favored at 90 °C, delivering a high CO₂ conversion of 80% and 100% CH₄ selectivity. Moreover, the intrinsic basicity was well maintained after 100 h activity test based on the CO₂-TPD results, suggesting a stable adsorption of CO₂ during the CO₂ methanation catalyzed by Ni/CeZrO₂ and plasma. In comparison, much lower CO₂ conversion (5%) and CH₄ selectivity (0%) were obtained in plasma alone [102]. Parastaev et al. [52] applied temperature-programmed plasma surface reaction method to further study the role of CeZrO₄ in CO₂ hydrogenation, where CO₂ was adsorbed on the catalyst followed by feeding H₂ with a ramping temperature in a tubular DBD reactor. A higher temperature might be detrimental to the CH₄ yield due to RWGS side reaction. For example, more CO but less CH₄ was produced above 275 °C and only a negligible amount of CH₄ was observed at 400 °C [52]. The synergy of CeZrO₄ and plasma was proposed based on the results that CO₂ adsorbed on the CeZrO₄ to form formate and carbonate species; under the plasma-induced electron impact dissociation, CO was generated and subsequently reacted with the reactive H^* species to form CH₄ [103,104]. As previously mentioned, CO₂ adsorption was facilitated at the basic sites of CeZrO_x. Notably, despite no general agreement, low- and medium-strength basic sites are possibly preferred for the methanation compared with strong ones [105]. Interestingly, after pre-treatment in DBD plasma for an excessive duration (i.e., 60 min), the concentration of strong basic sites were increased, which adversely reduced the CO₂ conversion. With a shorter treatment time (i.e., 40 min), 80% CO₂ conversion was delivered at a low input power of 5 W, much smaller than 13 W power for the thermally calcined counterpart [78].

As a porous material, zeolites are a proven efficient support for CO₂ hydrogenation due to the marked influence of the framework on performances [58]. When ZSM-5 was packed on the DBD reactor, the CO₂ conversion was increased to 25% from 8.1% without packing. This enhanced performance was attributed to the synergy between ZSM-5 and DBD plasma. In particular, O and H atoms produced via the plasma-induced electron impact dissociation were anchored on the ZSM-5 surface; meanwhile, the electrons might be trapped in the pores of ZSM-5. Therefore, the lifetime of the active species (e.g., H and O) was considerably prolonged owing to the weak chemisorption, and surface streamers were generated from the trapped electrons acting as the reservoir. In addition to the promoted interaction with the active species for CO₂ hydrogenation, the immediate water removal was realized on the zeolite surface in the plasma, promoting the adsorption/activation of CO₂ and alleviating the negative effect of H₂O on the methanation [61].

Owing to the outstanding adsorption capability of CO₂, Zr-MOF exhibits great potential in CO₂ hydrogenation [57]. Moreover, the high surface area and abundant surface hydroxyl groups benefit the dispersion of active metals [106,107,108]. In the presence of non-thermal plasma, the structure of Zr-MOF is effectively stabilized even under water [109,110]. However, after thermal calcination, the collapse of MOF structure probably occurred. These assumptions could be confirmed by the XPS results that, for the plasma-treated Zr-MOF, only the carboxylate and hydroxylated species in Zr-MOF (UiO-66) appeared without change in the Zr 3d spectra [111]; however, formation of O–Zr⁴⁺ and Zr–O signified the decomposition of the MOF structure under calcination (Figure 4a–d) [112]. The intact MOF support was also reflected by the morphology in TEM image (Figure 4e), ensuring a highly dispersed Ni active site. Furthermore, the hydroxyl groups generated in the plasma-treatment benefited the CO₂ adsorption. Owing to the above merits, the plasma-treated Ni/UiO-66 catalyst exhibited a fairly stable CO₂ conversion (85–90%) over 20 h, outperforming that in thermal catalysis at 380 °C (Figure 4f) (Table 1) [60].

3.3. Additive and DBD Plasma

Featured with the improvement of pore volume and surface area, CeO₂ is widely adopted as an additive in metal-based catalysts [113]. More interestingly, CeO₂ with a high dielectric constant (ɛ_r = 24) is able to change plasma property by adjusting the electron impact reaction kinetics since the charge accumulation and polarization effect enable an enhanced electric field [87]. Thus, a high CO₂ conversion of 70% and CH₄ selectivity of 95% were obtained at 5–6 kV voltage and H₂/CO₂ ratio of 4 (Table 1) [61]. In addition to the modification of plasma, owing to the moderate basic sites, CeO₂ as an additive can adsorb CO₂ to form carbonates, which are further reduced to formates and formaldehydes, subsequently generating CH₄ as the end product [104,114]. When coupled with DBD plasma, CO₂ molecules were first activated to CO₂^*, thus lowering the reaction temperature. According to Figure 5a, both Ni/Al₂O₃ and Ni/CeO₂–Al₂O₃ treated in plasma possessed larger CO₂ desorption peaks than the thermally calcined ones, indicating more abundant basic sites generated by plasma. Moreover, the larger desorption peak between 200 and 350 °C for Ni/CeO₂–Al₂O₃ suggested a higher concentration of medium basic sites, which were mainly responsible for the CO₂ adsorption and activation [115]. More significantly, the methanation of CO as an intermediate of plasma-stimulated CO₂ dissociation preferentially took place at CeO₂ when the loading of CeO₂ was lower than 10% [116]. A higher CeO₂ content would adversely affect the CH₄ yield due to the negative texture effects and unfavorable interaction with basic CO. Benefiting from the synergy of DBD plasma and CeO₂, a high yield of methane (80%) was achieved at 150 °C while only 60% CH₄ was obtained at 400 °C for thermal catalysis [62]. Apart from the modification effect on the surface basicity, the electron transfer environment could be created resulting from the unique redox property of CeO₂, leading to the CO₂ adsorption [117]. In detail, oxygen vacancies could be generated based on the loss of oxygen when exposed to plasma. Subsequently, CO₂ adsorbed onto the surface by fulfilling the vacant sites, facilitating the activation of CO₂ via electron impact dissociation [118]. As shown in Figure 5b, the second reduction peak of Ni/CeO₂–Al₂O₃ treated in plasma signified the partial reduction of Ce⁴⁺ to Ce³⁺ [119]. Importantly, CO₂ preferentially adsorbed onto the oxygen sites next to Ce³⁺ and the activated CO₂ on Ce³⁺ was more easily hydrogenated [10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35], thus delivering a high CO₂ conversion of 63% and CH₄ selectivity of 96% at 250 °C [120]. Again, an appropriate addition of CeO₂ might contribute to the oxidation of surface dissociated carbon to form CO species; however, an excessive amount of CeO₂ possibly initiated the further combination with O to produce CO₂ again, lowering the CO₂ conversion and hydrogenation efficiency [121].

As another rare earth metal oxide, La₂O₃ is also a proven effective additive to enhance the catalytic performance of CO₂ hydrogenation in DBD plasma. When La₂O₃ was incorporated in the Ni/BETA zeolite applied for CO₂ methanation, 85% CO₂ conversion and 97% CH₄ selectivity were obtained below 150 °C, while for the conversional thermal activation, only 80% CO₂ conversion was achieved at 400 °C (Table 1). Moreover, the 15-h stable activity suggested a robust catalyst under DBD plasma [63]. This excellent catalytic performance of Ni/La₂O₃–BETA zeolite was ascribed to the synergy between the plasma and La₂O₃ additive. Specifically, more hydroxyl groups were generated with the introduction of La₂O₃, strengthening the adsorption of CO₂ and preventing the gaseous CO₂ from being dissociated by plasma [122,123]. On the other hand, only carbonates were formed via thermal activation; in comparison, apart from the carbonates, monodentate formates were also generated on the La₂O₃ in the presence of DBD plasma, which were easily hydrogenated to CH_x species, thus leading to a high yield of CH₄ [116,123].

Characterized with the promotional effect on the reducibility and dispersion of Ni, the coupling of Ni with Fe increases the active site concentration of Ni-based catalyst [64]. When Ni–Fe LDH was combined with DBD plasma in CO₂ methanation, 75% CO₂ conversion and 97% CH₄ selectivity were obtained at 250 °C and an input power of 11.7 W (Table 1) [64]. On one hand, the energy utilization efficiency was enhanced in the DBD reactor, thus saving energy and lowering the reaction temperature without heating the reactor [78,105]. On the other hand, both weak and medium basic sites were increased with the addition of Fe, benefiting the CO₂ adsorption. Under plasma, CO₂ dissociated into CO, followed by forming formates and linear carbonyls, which were further hydrogenated into CH₄ [116].

Besides the rare earth and transition metal oxides, alkali metals (e.g., Na and K) are also doped into the Ni-based catalysts owing to their strong basicity. However, they were proven to cause a negative influence on the physicochemical property of Ni/CeZrO_x and reduction of performances in CO₂ methanation assisted with DBD plasma [65]. In detail, owing to the molten salt effect, a larger CeZrO_x particle size with less defects was produced, lowering the cycling rate of Ce³⁺/Ce⁴⁺ and oxygen vacancy concentration, reflected from the higher temperature reduction peaks in TPR profiles (Figure 6a). Additionally, the overly basic sites generated in the presence of Na and K impeded the participation of adsorbed CO₂ in the next-step reaction by formation of bridged and polydentate carbonates [124], as shown in the high temperature desorption of CO₂ in Figure 6b. Furthermore, despite the considerably enhanced dielectric constant in Na- and K-promoted catalysts, the excessively low breakdown voltage and intense ionization accelerated the direct CO₂ splitting into CO rather than further hydrogenation of the intermediates, resulting in a decrease of CH₄ yield and a higher power consumption (Figure 6c). Compared with the undoped Ni/CeZrO_x, the Na- and K-doped samples exhibited a 16–17% reduction of CO₂ conversion, 23–31% decrease of CH₄ selectivity, and 7–8 W increase of power input [65].

Apart from the metal additives, the introduction of Ar gas was proven to be influential for the reaction pathways and CO₂ conversion for Ni/Al₂O₃ catalyst assisted by DBD plasma [55]. In detail, the breakdown voltage was lowered from 3.1 to 2.6 kV with an increased amount of Ar from 0% to 50% due to less attachment of free electrons on CO₂ molecules and reduced dielectric strength of working gas [125]. Moreover, the addition of Ar promoted the Ni–DBD plasma interaction, thus realizing a sufficient contact during the reaction. Furthermore, Ar gas changed the discharge mode ‘partial discharging’ to ‘fully-bridged’ discharge, injecting more power into the reaction instead of being deposited on the dielectric surface, leading to a higher energy efficiency and effective capacitance. As a result, with the Ar feed increasing from 30% to 60%, the CH₄ selectivity was enhanced from 42% to 100% at 150 °C. More interestingly, the generated metastable argon (Ar^*) facilitated the dissociation of CO₂ and H₂ into C, O, and H atoms which further recombined at the catalyst surface to form targeted product CH₄ [55].

4. Operation Parameter

In addition to the catalyst design, the operation parameters considerably affect catalytic performances in CO₂ hydrogenation in catalyst–DBD plasma hybrid system, including the voltage, power, temperature, pressure, feed ratio, and GHSV [26,49,58].

4.1. Voltage/Power/Temperature/SIE (Specific Input Energy)

As usual, a high voltage generates a high power and temperature in the reactor. A series of voltages were applied in the Ni/CeZrO₂ catalyst for plasma catalytic CO₂ methanation (Figure 7) [49]. The highest CO₂ conversion and CH₄ selectivity was achieved at 16 and 15 kV respectively. Correspondingly, the power generated was between 20 and 40 W and the temperature reached the range of 230–270 °C. The enhanced performance at 15–16 kV compared with 14.5 kV was probably due to the promoted CO₂ adsorption and dissociation in a stronger electric field with micro-discharges in the bed materials [126,127,128]. A further increase of the applied voltage (e.g., 18 kV) caused an adverse effect on the performance that an obvious drop of CH₄ was shown from 100% to 82%, which might be related to the higher temperature generated at a higher voltage, benefiting the endothermic RWGS reaction and thus producing more CO instead of CH₄ [104,129]. Similarly, in methanol production by combined Pt/In₂O₃ and DBD plasma, the best performance was obtained at 30 W power; a further increase of the power would reduce the CO₂ conversion and CH₃OH selectivity due to the dissociation of CH₃OH back to CO₂ and H₂O [53]. In another research based on the zeolite catalysts, however, a higher CH₄ selectivity was favored at a higher input power [61]. A possible explanation was that a low power might not be able to further convert the CO to CH₄, while a high power could sufficiently push the methanation process and producing a dominant amount of CH₄. In terms of the production of C₂–C₄ hydrocarbons, a high input power was probably required based on the simultaneous decrease of CH₄ but increase of C₂–C₄ hydrocarbons selectivity [58], which might signify the further conversion of CH₄ to longer-chain products under a stronger electric field and intensified electron collisions. Above all, appropriate selection of the applied voltage (power) based on the specific situation is probably needed to balance between the effective activation and inhibition of side reactions. Notably, SIE (ratio of power to flow rate) increases with the increase of applied voltages at a constant flow rate; thus, in most cases, the SIE and voltages exhibit a similar effect on the performances [130].

4.2. GHSV (Gas Hourly Space Velocity)

As a ratio of the flow rate to the catalyst volume, GHSV is inversely proportional to the contact time [49]. Generally, in conversional thermal catalysis, a lower GHSV increases the CO₂ conversion due to the longer residence time of reactant on the catalyst surface [131]. In the presence of plasma, however, a higher flow rate and smaller catalyst mass enable a higher input power, thus intensifying the ionization process and generation of reactive species [132]. The relationship between GHSV and the temperature was further discussed depending on the magnitude of input power [133]. When the power was low (<10 W), temperature was independent from the GHSV since the plasma discharge was mainly responsible for the heat release; when the power was higher than 10 W, the heat produced from the exothermic reaction would not be ignored and the temperature increased with the increase of GHSV owing to the larger number of CO₂ converted to CH₄ in unit time [133]. However, in terms of the conversion of CO₂ based on the percentage, there was a nearly 10% drop with a 3.5-fold higher GHSV, indicating the effect of contact time. In terms of competitive production of CH₄ and CH₃OH, the yield of CH₃OH was not sensitive to the GHSV but a sharp drop of the CH₄ selectivity was seen with the increase of flow rate [26]. In another case, the production of C₂–C₄ hydrocarbons (rather than CH₄ or CH₃OH) was gradually reduced with the increase of GHSV, indicating the necessity of a relatively longer residence time to ensure the chain-growth reactions [58].

4.3. Feed Ratio

The H₂/CO₂ feed ratio exerts a profound impact on the CO₂ conversion and product selectivity. For Co/ZSM-5 catalyst in DBD plasma, the CO₂ conversion and CH₄ selectivity showed a monotonic increase with the feed ratio increasing from 1 to 4. In comparison, the selectivity of C₂–C₄ hydrocarbons reached the maximum value at a ratio of 3 [58]. It might be reasonable since more H atoms generated in a high feed would convert more CO₂ to CH₄; however, an excessive amount of H₂ may adversely affect the C–C coupling process due to the competitive hydrogenation reaction. As for the methanol production, a relatively higher concentration of CO₂ (above 50% in the mixture of CO₂ and H₂) favored a higher selectivity of CH₃OH against CH₄; on the contrary, 10% CO₂ feed only delivered a low methanol yield (14%), 2.5-fold less than that of CH₄ [26]. This was probably in line with the previous finding that CH₄ was preferentially formed in the presence of abundant H₂.

5. Reactor Configuration

In previous sections, two basic configurations of the DBD reactors—coaxial and plate-to-plate type—have been introduced. In addition, the reactor can also be differentiated based on the location of catalysts. In particular, the reactor where the catalyst was placed downstream was called a ‘two-stage configuration’, while the reactor where the catalyst was loaded inside the discharge zone was called a ‘one-stage configuration’. Obviously, only the long-lived species generated in the plasma could interact with the catalyst in two-stage mode. In one-stage configuration, the catalyst was exposed to more short-life species (e.g., radicals and electrons), leading to a different reaction pathway and performance [49]. In the following section, another few advanced reactor designs will be demonstrated to elaborate the influences on the catalytic performances and energy efficiencies in CO₂ hydrogenation.

As shown in Figure 8a, two types of DBD reactors differed in terms of their operation mode. Compared with the left pseudo-adiabatic one, the right adiabatic reactor was thermally insulated by granular spheres filled in a metal box [133]. Benefiting from the alleviated heat loss, the CO₂ activation was initiated at 5 W, half of the power needed in pseudo-adiabatic mode. Additionally, the maximum CO₂ conversion was achieved at 12.5 W in adiabatic conditions, again half of that in the pseudo-adiabatic conditions. Clearly, the thermal insulation effectively kept the inside temperature thus saving the power input and increasing the discharge transfer rate. At a 75% CO₂ conversion, the 58% energy efficiency (62 kJ/mol) was obtained in adiabatic reactor, 20% higher than that of the counterpart [133].

Different from the thermal insulation strategy to save the energy, SAPO membrane was integrated with the DBD plasma reactor to realize a simultaneous capture and utilization of CO₂ at a high efficiency (Figure 8b) [134]. As a kind of silico aluminophosphate zeolite, SAPO material was manufactured into a highly CO₂-selective membrane considering the appropriate pore size and strong CO₂ affinity [135,136]. When the membrane separation was coupled with the DBD reactor in CO₂ hydrogenation, 91.8% CO₂ capture efficiency and 71.7% CO₂ conversion efficiency were realized with a stable operation over 40 h. Notably, two-membrane system exhibited a 3.7-fold lower flow rate than the single-membrane separator, leading to nearly twice the efficiency for CO₂ capture and a 20% increase of CH₄ selectivity [134].

6. Reaction Mechanism

A deep and clear understanding of the reaction mechanisms benefits the effective control of the operation parameters, catalyst compositions, and reactor designs to ensure an optimized catalytic performance and energy efficiency in CO₂ hydrogenation. Several discussions on the proposed mechanisms will be presented as below.

By simulation with two-dimensional fluid model, CO₂ hydrogenation to CH₃OH was analyzed in terms of the detailed mechanism. Dominant reaction pathways included the electron-impact dissociation of H₂ and CO₂ to produce H, CO, and O radicals, followed by the recombination of H and CO to form key intermediate CHO. After successive hydrogenation by H atoms, CH₃OH was finally generated by the reaction of CH₃O and H, which was deemed as the major route. It is noteworthy that with the increase of H2 feed, the concentration of H₂O⁺, O⁺, and OH⁺ was greatly reduced [137]. Wang et al. [25] further studied the reaction pathways of CH₃OH production over Cu/Al₂O₃ catalyst and DBD plasma. In the plasma-alone system, HCO as a crucial intermediate could be formed by the combination of CO and H; however, the competitive recombination of H and HCO to produce CO and H₂ back would consume CO and lower the H₂CO formation rate [25,138]. In comparison, the catalyst–plasma hybrid system offered multiple reaction routes, such as CO hydrogenation and CO₂/formate hydrogenation [41], thus exhibiting a lower energy barrier and faster reaction kinetics. The necessity of the integration of DBD plasma and the catalysts was emphasized via one-dimensional fluid model [27]. In detail, CO was the only value-added product in plasma-alone system with a CO₂/H₂ feed due to the deficient CH₂ radicals, which was abundant in the CO₂/CH₄ mixture. Thus, a CO₂/H₂ mixed feed might not be suitable as a CO₂/CH₄ combination if other value-added chemicals were the main target, such as CH₃OH. However, the selectivity of certain oxygenates would be enhanced when the plasma was coupled with the catalyst [72].

As for the CH₄ production via CO₂ hydrogenation, one mechanism based on Ni-based catalysts was ever proposed that CO was first produced from the electron-impact dissociation of CO₂ and existed in the form of monodentate formates in gaseous state or adsorbed on the Ni surface. Under the DBD plasma, the formates were converted to linear carbonyls, which were further hydrogenated to generate CH₄ through stepwise recombination with H radicals [64]. The beneficial effect of DBD plasma was proven by Mu et al. [139] that the rate-determining step for CO₂ methanation was the dissociation of adsorbed H₂ and CO₂ on the Ni surface [63]. Assisted by the plasma, the activation energy was significantly decreased from 80 to 29 kJ/mol, thus promoting the facilitated dissociation of CO₂ and H₂ to form reactive CO and H species, which subsequently combined to produce CH_x and CH₄. Benefiting from the rapid O removal, CO₂ dissociation turned out to be irreversible, improving the reaction kinetics and selectivity [49].

7. Conclusive Remarks and Prospects

In this review, an overview is first provided regarding the plasma and catalyst-plasma hybrid system in CO₂ hydrogenation. After that, main contents were placed on the synergy between the catalyst and DBD plasma, including active metals, supports, and additives. Simultaneously, the structure–performance relationships are elucidated in depth. The plasma–catalyst hybrid system affected the reaction route and performance in a synergistic manner that the DBD plasma modified the dispersion of the active site, surface basicity, and oxygen defects; in turn, the plasma property (e.g., input power and ionization energy) could be adjusted by the ionic conductivity and dielectric constants of different catalysts, resulting in various degrees of energy utilization. Subsequently, the influences of operation parameters (e.g., applied voltage, input power, feed ratio, GHSV) and reactor configurations (membrane-assisted DBD reactor, adiabatic reactor, two-stage reactor) on the catalytic performances and energy efficiencies are discussed in detail. In addition, various mechanisms of CO₂ hydrogenation based on two targeted products of CH₄ and CH₃OH are demonstrated. Despite the improvements in plasma–catalytic CO₂ hydrogenation, some issues still exist and the possible solutions are proposed as below.

First, there is still no agreement regarding the synergy between the surface basicity of catalysts and DBD plasma. Advanced characterization techniques (e.g., CO₂-TPD, in-situ FT-IR) and numerical studies based on appropriate models are promising potential solutions.

Second, the general relationship between the input power/applied voltage and the catalytic performances are still under debate. An intensified ionization process might lead to an efficient activation and dissociation of CO₂ while the possible decomposition of the products under this condition is a concern. Depending on the specific catalytic system and targeted product, the desired operation parameters (e.g., input power) are waiting for further explorations via advanced mathematical tools.

Third, as a newly emerging catalysis, plasma catalysis requires tremendous effort to realize its commercialization, by means of enhancing the energy efficiency, exploiting efficient production of H₂ from renewable energy sources, and increasing the selectivity of value-added products based on a thorough investigation of the mechanisms.

Fourth, the origin of the plasma enhancement, at least in some systems, is an important yet unresolved question. Thus, more studies might be needed in the future.