Methane to Methanol through Heterogeneous Catalysis and Plasma Catalysis

Abstract

Direct oxidation of methane to methanol (DOMTM) is attractive for the increasing industrial demand of feedstock. In this review, the latest advances in heterogeneous catalysis and plasma catalysis for DOMTM are summarized, with the aim to pinpoint the differences between both, and to provide some insights into their reaction mechanisms, as well as the implications for future development of highly selective catalysts for DOMTM.

1. Introduction

Methane (CH₄), as one of the most important molecules in C1 chemistry, is widely present in natural gas, shale gas, coalbed gas, combustible ice, etc. Noticeably, natural gas, which consists of approximately 70% to 90% CH₄, will be part of the energy system for decades [1]. Unfortunately, most of these CH₄ reserves are located in remote areas, indicating the need of transportation for utilization of CH₄ [1,2]. However, due to a very low boiling point (−161.6 °C at a pressure of 1 atm) and high flammability, compression of CH₄(gas) into CH₄(liquid) for transportation requires huge amounts of energy, making it economically infeasible [3]. In addition, another important issue is the rising global emission of CH₄, mainly as by-product of oil production, and its global warming potential is ca. 30 times that of CO₂ [1]. The International Energy Outlook 2019 (IEO2019) estimated that 140 billion cubic meters (bcm) were flared and 60 bcm released into the atmosphere in 2018, more than the annual LNG (Liquefied Natural Gas) imports of Japan and China combined. This enormous source of emissions accounts for 40% of the total indirect emissions from global oil supply. Therefore, the conversion of CH₄ to value-added chemicals has attracted intensive interests from both academic and industrial communities.

In general, as shown in Figure 1, the conversion of CH₄ into value-added chemicals can be classified into indirect and direct routes [4]. As implemented in industry, indirect routes are, actually, initiated through a steam reforming (SR) and/or auto-thermal reforming (ATR) process to produce syngas (mixture of CO and H₂), and then a variety of products such as olefins, gasoline, and diesel, as well as oxygenates, can be obtained using the well-established technology of Fischer-Tropsch synthesis (FTS) promoted by Fe-based or Co-based catalysts [5,6]. Alternatively, using Cu-Zn-Al-based catalysts, syngas can also be converted into methanol (CH₃OH), which has been used as feedstock to produce light olefins, gasoline and aromatics through industrial technologies of methanol-to-olefins (MTO), methanol-to-gasoline (MTG) and methanol-to-aromatics (MTA) conversion, respectively [6]. Although the above indirect routes are carried out in industry to produce value-added chemicals from CH₄, the syngas production by SR and ATR is energy-intensive and costly, motivating researchers to develop direct routes (not syngas-based) [7].

Direct routes (Figure 1), including CH₄ dehydroaromatization, CH₄ coupling to hydrocarbons (both oxidative and non-oxidative coupling), CH₄ pyrolysis (high temperature pyrolysis and catalytic pyrolysis) and direct oxidation of CH₄ to methanol/formaldehyde, have also been developed and are still being improved [2,3,4,5,6,7,8]. However, CH₄, a molecule with tetrahedral geometry and four equivalent C–H bonds, is inert and difficult to activate and convert. The absence of a dipole moment and a rather small polarizability (2.84 × 10⁻⁴⁰ C²·m²·J⁻¹) imply that CH₄ needs a relatively high local electric field to be polarized and to allow electrophilic or nucleophilic attack [2]. CH₄ exhibits the highest C–H bond strength among all alkanes, with the first bond dissociation energy (BDE) of 493.3 kJ mol⁻¹ (5.1 eV), meaning that CH₄ is the least reactive alkane [2]. Therefore, most of the direct routes must be operated at ultra-high temperature (973~1223 K), except for direct oxidation to methanol/formaldehyde, which can be realized at relatively low temperature (300~700 K), indicating low cost and high feasibility in industry [2,3]. For that reason, direct oxidation has attracted more attention, and in this review, we mainly focus on recent progress of direct oxidation of methane to methanol (DOMTM).

CH₄ + 0.5 O₂ → CH₃OH  ΔH (298K) = −126.4 kJ·mol⁻¹

(1)

DOMTM by O₂ (1) has been considered as a dream reaction in chemical industry and a holy grail in catalytic chemistry [9,10], and it has attracted intensive interests from both academic and industrial communities for more than 100 years. DOMTM has been studied by homogeneous catalysis, in which noble metals (Pt and Pd) are typically used as the central atoms of the complex catalysts, and the reaction is usually carried out in strong acid media (sulfuric and trifluoroacetic acid) [11,12,13]. Alternatively, DOMTM can also be realized by heterogeneous catalysis. In the 1980s, a Mo-based catalyst was developed for CH₄ oxidation. The Mo=O species was considered to be the active site for the oxidation of CH₄ to CH₃OH, and at that time, Mo was considered to be the most active metal catalyst for this reaction [14,15]. However, the biggest drawback of a Mo-based catalyst is that Mo can be easily lost at high temperature through volatilization, which was not conducive for industrial application [16]. Compared with a Mo-based catalyst, a V-based catalyst is more stable, but the CH₄ conversion was too low (less than 10%) [17,18]. At the beginning of this century, inspired by the active sites of double iron and double copper in methane monooxygenase, Fe-based and Cu-based zeolite catalysts have been used in DOMTM [19,20,21,22], as well as supported noble metals, such as Au, Pd and Rh [23,24]. Detail information for the developing process of DOMTM can be obtained in the review paper by Bokhoven [3].

Although many studies about DOMTM have been reported, it has to overcome two challenges, caused by thermodynamics and kinetics, respectively [3,9]. The first is how to improve the CH₃OH selectivity. Thermodynamically, CH₃OH is not the favorable product, as CO and CO₂ are more stable than CH₃OH. Specifically, as shown in Figure 2A, a low temperature (<890 K) favors the production of CO₂ and H₂O, while a high temperature (>890 K) favors CO and H₂. In other words, due to the higher reactivity of CH₃OH than the feedstock CH₄, the catalytic sites, capable of oxidizing CH₄ into CH₃OH, can also further oxidize CH₃OH into CO or CO₂ before CH₃OH desorbs from the catalyst surface. Figure 2B indeed illustrates that CH₃OH and other oxygenates and hydrocarbons are only formed with much lower selectivity.

The second challenge is how to reduce the kinetic energy barrier (Ea) of DOMTM by O₂ at ambient conditions. The Ea of DOMTM by O₂ is much higher than for DOMTM using N₂O or H₂O₂ as oxidants, because both N₂O and H₂O₂ can more easily release an oxygen atom, as the main species to trigger the oxidation of CH₄ to CH₃OH. Therefore, when using O₂ as oxidant, high temperature and high activity catalysts are needed to overcome the Ea of DOMTM, which unfortunately leads to deep oxidation. However, for reaction (1) the entropy is reduced (ΔS < 0) while it is exothermic (ΔH < 0). Therefore, CH₃OH production is favorable at low temperature, because a low temperature can result in a negative ΔG value (i.e., a spontaneous process), leading to a contradiction between the dynamics and thermodynamics of the DOMTM process.

Recently, several innovative studies by heterogeneous catalysis and plasma catalysis have been performed to overcome the above-mentioned challenges, which we will discuss in this review.

2. Heterogeneous Catalysis

Heterogeneous catalysis is more suitable for industrial application than homogeneous catalysis, being more convenient for product and catalyst recovery. A number of excellent reviews have been recently published on DOMTM, adding to the classic reviews in the field [3,4,5,7,8,9]. The collective desire on these articles was emphasized to find a breakthrough, pushing DOMTM technology closer to commercialization. Generally, seeking for efficient catalytic systems with a controllable reaction kinetics process is the key of CH₄ conversion, and the dissociation of the first C–H bond was regarded as the rate-determining step on various CH₄ conversion reactions. Basically, as shown in Figure 3, the mechanism of C–H bond cleavage of CH₄ at low temperature can be mainly classified into two categories [9]. The first mechanism involves H abstraction from CH₄ to form •CH₃ radicals by electrophilic oxygen atoms, while the other mechanism includes the formation of metal-CH₃ (M-C) σ–bond as a reaction intermediate which can directly cleave the C–H bond with concomitant coordination of the CH₃ group [9].

Except seeking advanced catalytic systems that enable efficient C–H activation, the insight obtained in the different roles of oxidants in DOMTM will be also of benefit to the field. On the one hand, strong oxidants (e.g., N₂O and H₂O₂) can easily break the C–H bond, but they will also enhance the over-oxidation of intermediate products adsorbed on the catalyst surface to thermodynamically more stable CO or CO₂, which is not conducive to the generation of CH₃OH or its derivatives. On the other hand, these strong oxidants with better oxidation properties have higher raw material prices, leading to higher economic costs. In this section, we summarize the research progress of DOMTM by different oxidants (

Table 1. Selected experimental results of direct oxidation of methane to methanol (DOMTM) based on different oxidants.

Oxidant	Catalysts	PCH4 (bar)	Temp (°C) a	CH3OH Yield (μmol·gcat−1)	Productivity (mmol /molmetal)	CH3OH Sel. (%)	Ref.
N2O	Fe/ZSM-5	6.6 × 10−4	160	160	/	76	[25]
	Fe/CHA	/	RT	26.8	681	/	[26]
	Cu/MOR	/	150	97	>300	/	[27]
	Cu/SSZ-13	0.3	260	19	/	27	[27]
H2O2	Au-Pd/TiO2	30.5	2	54.6	13.1	45.2	[28]
	Au-Pd colloids	30	≤50	2.8–7.6	/	14–28.3	[23]
	AuPd@ZSM-5	0.48	70	/	91.6	92	[29]
	FeN4/Graphene	18	25	/	/	/	[10]
	Rh/ZrO2	28.5	70	/	1.25	/	[30]
	Cu-Fe/ZSM-5	30.8	50	5.2 × 103	/	88	[31]
O2	Cu/ZSM-5	/	175	8.2	/	98	[32]
	Cu/MOR	/	200	170	0.47	/	[33]
	Cu/SSZ-13	0.1	200	125	0.2	>90	[34]
	Cu/Omega	30	200	200	265	/	[35]
	Co/ZSM-5	1	150	0.3–0.4	/	40–100	[36]
	Ni/ZSM-5	1	175	14.9	/	/	[37]
O2+H2O	CeOx/Cu2O/Cu(111)	2.7 × 10−5	25	/	/	70	[38]
	Ni/CeO2	1.3 × 10−3	177	/	/	<40	[39]
O2+CO	Rh/ZSM-5	20	150	1224	/	6.2	[24]
O2+H2O2	RuCu/NL b	25	50	~1.5 × 103	/	/	[40]
N2O+H2O	Mo/SiO2	/	300	/	16	60	[41]

^a The temperature of the DOMTM process by chemical looping (more information on Section 2.3) is based on the step of the CH₄ reaction for better comparison under the same criteria; ^b NL: Nanolichens.

) in heterogeneous catalysis, with emphasis on the introduction of the catalytic systems with O₂ as oxidant.

2.1. Nitrous Oxide as Oxidant

Nitrous oxide (N₂O) is widely used as oxidant in DOMTM (2), mainly based on Fe-modified zeolites, which is particularly attractive from the perspective of natural catalysts (i.e., methane monooxygenase (MMO) enzymes), activating CH₄ to produce CH₃OH under mild conditions. The unique properties of MMO are related to the Fe-containing sites capable of generating extremely active oxygen species that can insert into non-activated C–H bonds of CH₄ under ambient conditions [42]. Great interest has been paid to Fe-based zeolites, which may contain various iron species, especially activated by N₂O. Early results showed that Fe/ZSM-5 catalyst can decompose N₂O effectively through a Fe/O highly active site on its surface at low temperature (<300 °C) and realize partial oxidation of benzene to phenol [43,44]. This active site on the Fe/ZSM-5 catalyst is called the α-Fe or α-O site, and it was found that DOMTM can be achieved at low temperature [45]. The nature of the active site with exceptional reactivity is difficult to prove spectroscopically mainly due to the presence of inactive spectator iron species.

CH₄ + N₂O → CH₃OH + N₂  ΔH (298K) = −159.0 kJ·mol⁻¹

(2)

The properties of α-oxygen and the state of iron constituting α-sites were studied in many experimental [46,47] and theoretical works [48,49] and discussed in review papers [3,50]. Despite great research efforts have been made, the spectroscopic insight into the structure of actual active site was illustrated only recently on Fe-BEA zeolite [46]. By magnetic circular dichroism (MCD), a mononuclear, high-spin, square planar Fe(ii) site was identified and this high-reactive site was previously regarded as an α-Fe(II) site by Panov and co-workers. Furthermore, the α-O site constrained by the zeolite lattice, considered as the reactive intermediate, is, actually, a mononuclear, high-spin Fe(iv)=O species [46]. From the DR-UV–Vis spectra, an intense band at 40,000 cm⁻¹ and three weak bands at 15,900, 9000, and 5000 cm⁻¹ were observed for the inactivated Fe-BEA (Figure 4a). After N₂O activation and CH₄ reaction, the peak at 16,900 cm⁻¹, along with a weak feature at around 5000 cm⁻¹ present after N₂O activation, disappear upon reaction with CH₄, which corresponds to the process of the α-O site which forms on α-Fe (II) active site. The 5000–13,000 cm⁻¹ region of the CH₄-reacted spectrum overlaps with that of Fe(II)-BEA, indicating that features in this region originate from inactive iron species. Furthermore, the MCD spectra show a band at 15,100 cm⁻¹ that is corresponding to the 15,900 cm⁻¹ absorption band of the α-Fe(II) site from DR-UV–Vis, which is sensitive to a magnetic field and temperature. Additionally, DFT-optimized cluster models with high-energy ligand-field band were predicted by CASPT2, and α-Fe(II) site can be assigned to a high-spin, square planar Fe(II) site with four anionic Si–O–Al ligands (Figure 5), which can be further verified by DFT-calculated Mössbauer parameters.

An Fe-FER (Ferrierite) catalyst was also studied, in which the formation of active oxygen species from N₂O decomposition can react with CH₄ to form CH₃O groups and valuable products at moderate temperatures. The emergence of methoxy group bands in the FTIR spectrum with introduction of CH₄ on Fe-FER catalysts indicates the active surface species leading to the formation of selective products [51]. As shown in Figure 6, the amount of active α-oxygen sites and number of Bronsted acid sites on the Fe-modified zeolites, e.g., Fe/ZSM-5 and Fe/FER, are crucial factors to alter CH₄ conversion and product distribution as well [52]. The Fe/ZSM-5 zeolite, with a large number of strong acid sites, could produce a higher selectivity to unsaturated C₂-C₃ hydrocarbons and dimethyl ether (DME), while the Fe/FER zeolite, with a large number of weak Bronsted acid sites, gave rise to oxygenates such as CH₃OH and DME [52]. The formed coke precursors seem to be more dominant on the Fe/ZSM-5 zeolite with stronger acid sites due to a facile reaction pathway through the MTO reaction. The amount of coke deposition on acid sites and α-oxygen sites was well correlated with the catalytic performance, and thus a possible reaction pathway (Figure 6) was proposed by the authors, based on the product distribution with respect to acid sites and α-oxygen sites [52]. The general consensus of DOMTM reaction pathway over α-O sites follows a radical-based H atom abstraction mechanism (cf. Figure 3), although some steps based on this mechanism are still debated. Briefly, oxygen atoms are introduced by N₂O decomposition before subsequent CH₄ addition. The abstracted hydrogen atom from CH₄ by the α-O results in an Fe^Ⅲ-O-H fragment and a CH₃ radical, and produced CH₃ radical may then either react with a further α-O to form Fe^Ⅲ-O-CH₃ or form Fe^Ⅱ-O(H)-CH₃ with near fragment, which may then desorb and eventually form CH₃OH [48,53].

The remarkable activity of the α-O site can be partially attributed to confinement effects with the zeolite channels [48,54], which are thought to increase the local concentration of the molecular orbitals within the micropores, resulting in increased interactions between confined reactants and enabling unusual transition states to be accessed. Furthermore, the confined molecules within zeolites have also been shown to induce dipoles and multi-poles, potentially strengthen or weaken some C−C bonds, which may lead to modifications of the HOMO/LUMO energy levels.

Periodic structure and cluster modeling of α-O sites in the SSZ-13 zeolite at different levels of theory has been performed and the results showed that the confinement effect of zeolite channels may reduce the energy barrier of DOMTM by over 50% [48]. This confinement effect is mainly electrostatic in nature, which could stabilize all transition states, the reaction intermediate and products. DFT calculations on a periodic system of MO⁺/ZSM-5 zeolite (M=Fe, Co, Ni and Cu) indicate that the confinement effect is attributed to the nanopores of ZSM-5 zeolites, in which interaction between CH₄ and MO⁺ species was confined [55], resulting in a significant destabilization of CH₄ adsorption and further lowering of the activation energy for the C–H bond dissociation. In addition, mono-nuclear α-Fe²⁺ in an extra-lattice site within Fe-beta zeolite (BEA) also showed that the reactive intermediate was a high spin Fe⁴⁺=O species and the confinement of the zeolite lattice promotes the reactivity [46]. Therefore, by preventing geometric distortion, the confined α-O sites can activate CH₄ and cleave the strongest aliphatic C–H bond at room temperature to form CH₃OH.

Variant zeolite topology has thus been considered an important factor in rationally tuning of the active site properties for DOMTM. The single-site α-Fe in the CHA topology zeolite was demonstrated to be active to form a highly reactive α-O, capable of activating CH₄ at room temperature to form CH₃OH, which subsequently desorb from catalyst surface by on-line steaming at 200 °C [26]. It has been found that the topology’s 6MR geometry of CHA zeolite has great effect on the structure, the ligand field, and consequently the spectroscopy of the α-Fe site, by comparing α-Fe in Fe/BEA and Fe/CHA. At higher Fe loading (>0.26 wt%; Si:Al = 12.5), Fe₂O₃ was identified other than α-Fe sites in Fe/CHA zeolite. Therefore, the level of introducing Fe into CHA zeolite is a crucial criterion to improve reaction reactivity with increased active site density [26]. Additionally, indicated by nitrous oxide titration, the method of introducing appropriate extra-framework Al in Fe/MOR catalysts is in favor of increasing iron ions in tetrahedral or octahedral coordination, which can promote the formation of more α-sites in Fe-containing zeolites [56].

The DMTOM process was studied in a so-called quasi-catalytic mode over Fe/ZSM-5 zeolite with N₂O as an oxidant, in which the reaction intermediates or products just migrate from catalytically active centers to other sites of the surface, rather than desorption into the gas phase, mainly due to the activation energy of surface diffusion of adsorbed species is much smaller than desorption energy of the species into the gas phase [25,57]. As shown in Figure 7, the produced CH₃OH over α-O sites moved to neighboring sites without desorption into the gas phase in the quasi-catalytic mode, and the adsorbed species could further be converted to coke or extracted from surface to get methanol and dimethyl ether (DME). The reaction could switch to conventional catalytic process at above 200 °C, with products direct desorption into gas phase. Low CH₄ conversion (0.19%) is indeed a huge limitation for DOMTM reaction, but this approach provides new insight and a possibility to identify the potential intermediates [57].

In addition to Fe-modified zeolites, some other catalysts have been also studied for DOMTM by using N₂O as an oxidant. Liu et al. investigated the catalytic performance of a Mo-based catalyst in partial oxidation of CH₄ to CH₃OH using N₂O as oxidant. However, only 3% CH₄ conversion was achieved with 78% oxygenates selectivity (CH₃OH and HCHO) [58]. Cu/SSZ-13 catalyst activated by N₂O was also reported for DOMTM at low temperatures and ambient pressure, with highest CH₃OH productivity of 55 μmol CH₃OH g⁻¹ h⁻¹, which was more than twice the rate of Cu-mordenite and more than four times the rate of Cu-ZSM-5 [59]. Higher partial pressures of CH₄ and H₂O with lower N₂O partial pressure was thought to further improve CH₃OH selectivity [59]. Recently, the performance of N₂O and O₂ oxidants were compared by DFT calculation on the CoN₃–Graphene catalyst, and the results shows that DOMTM reaction can proceed via a two-step pathway, with N₂O as an oxygen donor [60]. CoN₃–Graphene catalyst exhibited higher catalytic activity for the adsorption of gas reactants, which can be used as an efficient catalyst to fabricate effective C/N catalysts on methane oxidation by N₂O.

N₂O, a relatively stable greenhouse gas (GHGs), mainly from agricultural production and other anthropogenic activities, like waste management, or the combustion of fossil fuels and biomass, which could be correlated to stratospheric ozone destruction by atmospheric photochemical reactions [61]. Indeed, the co-conversion of CH₄ and N₂O is of great interest in reducing the anthropogenic forcing of the climate system, as a win-win strategy for both ozone and climate. However, the utilization of N₂O as an oxidant confronts an insurmountable hurdle associated with its low availability which cannot satisfy the large demand for industrial scale CH₄ valorization.

2.2. Hydrogen Peroxide as an Oxidant

As an important industrial oxidant, hydrogen peroxide (H₂O₂) is widely used in papermaking, sewage treatment, metallurgy, medical and health care, and other fields [62]. The benefit of H₂O₂ oxidant is the clean decomposition to water as an environmentally benign byproduct. Noticeably, H₂O₂ has been considered as a highly efficient oxidant in DOMTM (3). Many studies realized the DOMTM process under milder conditions compared to N₂O oxidant, mainly employing noble metal catalysts.

CH₄ + H₂O₂ → CH₃OH + H₂O  ΔH (298K) = −223.9 kJ·mol⁻¹

(3)

Singly dispersed noble metal atoms anchored on oxides could offer a distinctly different electronic state in contrast to continuously packed metal atoms on the surface of a metal nanoparticle, and thus could exhibit a distinct catalytic activity or/and selectivity [63,64]. Pd₁O₄ single-sites anchored on the internal surface of micropores of ZSM-5 exhibited excellent performance for CH₄ activation [65]. Under the assistance of H₂O₂, CH₄ was partially oxidized into CH₃OH over Pd₁O₄ site at low temperature (50–95 °C), but extra H₂O₂ resulted in further oxidation of CH₃OH to CHOOH. The CH₃OH selectivity remained at 86.4%, while the CH₃OH productivity at 95 °C was about 2.78 molecules per Pd₁O₄ site per second when CuO was used as a co-catalyst. The experiment of varied amounts of Pd on Pd/ZSM-5 catalysts exhibited quite similar yields of the products, although their loadings of Pd were quite different (from 0.01 to 2.0 wt.%), suggesting that the excess Pd, present in the external surface of the zeolite, did not alter the activity, as well as the selectivity of CH₃OH. Furthermore, DFT calculations showed that CH₃OH is the thermodynamically stable product over Pd₁O₄, which was consistent with the experimental result that CH₃OH production was highly favorable compared to the formation of a byproduct, methylperoxide [65]. Kwon and co-workers reported an atomically dispersed Rh/ZrO₂ catalyst using aqueous H₂O₂ as an oxidant for the selective oxidation of CH₄, where the property of Rh active sites significantly affected the CH₄ oxidation [30]. Single atomic Rh could make CH₃OH with the highest productivity, whereas Rh nanoparticles on SiO₂ produced only CO₂ without the formation of C1 oxygenates (Figure 8A). The amount of oxygenated products showed only a little difference up to the fifth cycle (Figure 8B). When replacing the oxidant with O₂, C₂H₆ was observed as main product in gas-phase for direct CH₄ oxidation, as shown in Figure 8C [30]. Additionally, combing EXAFS, XANES, HAADF-STEM/EDS images and CO-adsorption using DRIFTS measurements, the model of single-atom Rh/ZrO₂ catalyst was proposed, based on which DFT calculations were carried out. The results showed that CH₃ intermediates can be energetically stabilized on the catalyst, which was further verified by DRIFTS measurements. The active single-site Rh₁O₅ anchored in microporous aluminosilicates (Rh/ZSM-5) in solution can realize the oxidation of CH₄ to CH₃COOH (acetic acid) and CH₃OH below 150 °C [66]. An isotope experiment confirmed that the C atoms of the methyl and carboxyl groups in CH₃COOH were derived from CH₄ and CO, respectively. Noticeably, high pressure of CO is detrimental to the production of CH₃COOH and finally poisoned the active sites [66]. The conversion of CH₄ to oxygenates on Rh/ZSM-5 may occur via M–CH₃ functionalization. Firstly, CH₄ is activated in the presence of O₂ on isolated Rh⁺ cations under mild conditions to produce Rh–CH₃. The formed Rh–CH₃ can then be functionalized via two independent reaction pathways: oxygen insertion to produce CH₃OH, or CO insertion to produce CH₃COOH. After a hydrolysis step, the whole catalytic cycle can be completed [24]. However, the in-depth mechanism responsible for this activation is still not clear and further investigation is indispensable to guide the design of novel and more efficient CH₄ oxidation catalysts.

Interestingly, Au-Pd nanoparticles (NP) are highly effective catalysts for the direct synthesis of H₂O₂, and the hydroperoxy species (HOO) is effective for the oxidation reaction [67]. Thus, Au-Pd supported nanoparticles (AuPd/TiO₂) are active for the oxidation of CH₄, giving a high selectivity for CH₃OH formation under mild aqueous conditions with H₂O₂ as oxidant [28]. A similar productivity, but with improved CH₃OH selectivity, was observed when using the in situ generated H₂O₂ by adding H₂/O₂ gases mixture, compared to the experiments performed with pre-synthesized H₂O₂. Additionally, both methyl (•CH₃) and hydroxyl (•OH) radicals were observed by electron paramagnetic resonance (EPR) spectroscopy, which suggests that the CH₄ reaction proceeds through a radical mechanism, in contrast to the reaction mechanism previously proposed for CH₄ oxidation using CuFe/ZSM-5, where •CH₃ radicals were not observed [31].

Recently, further experiments showed that colloidal Au-Pd nanoparticles can oxidize CH₄ in aqueous solution at mild temperatures with 92% CH₃OH selectivity, in the presence of both H₂O₂ and O₂ [23]. Different components of the catalyst were studied for the H₂O₂ degradation rates. H₂O₂ degradation rates of bare TiO₂ and unsupported Au-Pd colloidal NPs were low, while the Au-Pd/TiO₂ catalyst exhibited a high rate of H₂O₂ degradation (73%), indicating that either the interfacial sites at the support/metal interface or a change in the morphology of the NP led to the high H₂O₂ degradation rates [67,68,69,70]. The Au-Pd colloid catalyst decomposes H₂O₂ at a much lower rate (38%) but makes substantially more products. Moreover, the addition of 5 bar of O₂ pressure to the reaction resulted in an increased yield of oxygenate products (e.g., CH₃OH, CH₃OOH and HCOOH) compared with the H₂O₂-only reaction. Isotopically labeled oxygen (O₂) as oxidant in the presence of H₂O₂ indicated that the produced CH₃OH incorporated a substantial fraction (70%) of gas-phase O₂ under optimized conditions, which can react with •CH₃ radicals generated via H abstraction by •OH from H₂O₂ as initiation step of CH₄ activation (Figure 9). Additionally, more oxygenated products were formed than the amount of H₂O₂ consumed, suggesting that the controlled breakdown of H₂O₂ activates CH₄, which subsequently incorporates O₂ through a radical process [23]. Titania-supported AuPdCu catalysts are active for the oxidation of CH₄ under mild reaction conditions by using H₂O₂ as oxidant. After depositing Cu together with Au/Pd on the surface of TiO₂ the rate of CH₄ oxidation with addition of H₂O₂ is significantly enhanced. In particular, 2.5% Au 2.5% Pd 1.0% Cu/TiO₂ showed ca. 83% selectivity to CH₃OH [71].

To make CH₄ economically competitive as a source of energy, it is crucial to increase the productivity of CH₃OH and the efficiency of H₂O₂ utility [72]. DOMTM by AuPd@ZSM-5 catalyst (Figure 10) using in situ generated H₂O₂ at mild temperature (70 °C) has been reported, which can prevent H₂O₂ dilution, and thereby keep a high local concentration of H₂O₂ around the AuPd nanoparticles [29]. This hydrophobic zeolite was synthesized by AuPd alloy nanoparticles within aluminosilicate zeolite crystals, where the external surface of the zeolite was modified by organosilanes, as depicted in Figure 10. The hydrophobic sheath appears to allow diffusion of H₂, O₂, and CH₄ to the catalyst active sites, while it hinders the diffusion of generated H₂O₂ from the encapsulated AuPd nanoparticles. As shown in Figure 11A, the AuPd@ZSM-5-C3 catalyst gave a CH₄ conversion of 16.4% with CH₃OH selectivity of 90%, and higher conversions were realized over AuPd@ZSM-5-C6 and AuPd@ZSM-5-C16 catalysts that contained longer organic chains modified on the zeolite crystals. The AuPd@ZSM-5-C16 catalysts exhibited the best performance with 17.3% CH₄ conversion and 92% CH₃OH selectivity, corresponding to CH₃OH productivity up to 91.6 millimoles per gram of AuPd per hour. Additionally, the formation of H₂O₂ from H₂ and O₂ over AuPd@ZSM-5 catalysts was also studied, as shown in Figure 11B. Washing the solid catalysts by cold CH₃OH and tetrahydrofuran (~0 °C), which can liberate the H₂O₂ within the zeolite crystals, 8.1, 68.4, and 78.3 μmol of H₂O₂ was obtained during this process for AuPd@ZSM-5, AuPd@ZSM-5-C3, and AuPd@ZSM-5-C16, respectively. In contrast, the H₂O₂ quantity in the water solution reached 97.9 mmol when the AuPd@ZSM-5 catalyst was used, which further suggests that AuPd@ZSM-5-C16 catalyst can enrich the H₂O₂ within the zeolite crystals. This molecular-fence concept seems interesting and represents a large step toward the application of direct CH₄ oxidation to valuable products. However, the CH₃OH productivity still needs further improvement, and the batchwise operation restricts the commercial application [29].

In addition to the above-mentioned precious metal catalysts, some cheap metal catalysts have also been studied in DOMTM using H₂O₂ as oxidant. Xie and co-workers studied the performance of TiO₂-supported iron catalysts in DOMTM at ambient conditions [73]. It was observed that the H₂O₂ amount was vital for controlling the CH₃OH selectivity. When no H₂O₂ was present, no CH₄ was converted. Increasing the H₂O₂ amount results in a higher CH₄ conversion, but the CH₃OH selectivity decreased. The optimal ratio of H₂O₂ to CH₄ was 0.11, showing a 15% CH₄ conversion and an alcohol selectivity over 97% (CH₃OH selectivity over 90%) [73]. Xu and coworkers studied DOMTM over Cu- and Fe-modified ZSM-5 catalysts in a flow reactor using H₂O₂ as the oxidant under continuous flow operation. Co-impregnation of ZSM-5 with Fe and Cu by chemical vapour impregnation yielded catalysts that showed high CH₃OH selectivity (>92% selectivity, 0.5% conversion) [74]. In another study, Kim et al. reported partial oxidation of CH₄ over Fe/ZSM-5 catalyst using H₂O₂ as oxidant, and found that the total product yield and the amount of H₂O₂ consumed increased with increasing Fe content in the Fe/ZSM-5 catalyst prepared using an ion-exchange method [75]. The effect of Si:Al molar ratio of Cu- and Fe-exchanged zeolites on DOMTM was also studied and the results demonstrated that high CH₃OH production can be realized when catalysts with low Si:Al ratio were used [76]. The Fe-only ZSM-5 catalysts exhibited the highest catalytic activity (total oxygenated products) with HCOOH being the major product, which can be explained by the increased amount of Brønsted acid sites. The presence of Cu aims to maintain high MeOH selectivity by suppressing the production of the deeper oxidation product like CH₃COOH and HCOOH [76].

A graphene-confined single Fe atom catalyst (FeN₄/GN) was screened from a series of 3d metal–N₄ embedded in the lattice structure of graphene nano-sheets, as shown in Figure 12A,B. The unique O–FeN₄–O structure formed in graphene can directly convert CH₄ to C1 oxygenated products (e.g., CH₃OH, CH₃OOH, HCOOH and HOCH₂OOH), with total selectivity around 94% [10]. The CH₄ oxidation reaction was carried out in a specifically designed high-pressure reactor connected with an operando time-of-flight mass spectrometer (TOF-MS) (Figure 12C), which can successfully accomplish the qualitative and quantitative identification of products during reaction and can detect how CH₄ is exactly oxidized stepwise at a more real-time analysis. The CH₃ radical is first converted into CH₃OH and CH₃OOH, and CH₃OH can be further converted to HOCH₂OOH and HCOOH on the O–FeN₄–O site (Figure 12D,E), as illustrated by TOF-MS, ¹³C NMR, and DFT calculations. The intermediate HOCH₂OOH was the first time to be identified as a product in CH₄ oxidation. Comparison studies between the formation energies of O–MN₄–O active sites illustrated that among all O–MN₄–O (M=Cr, Mn, Fe, and Co) structures, O–FeN₄–O has a moderate formation energy and the highest CH₄ activation rate (Figure 12F), corroborating well with the best activity of the FeN₄ site in the experimental studies. Recently, Meysam et al. studied the influence of zeolite acidity on partial oxidation of CH₄ over M-Fe-MFI (M: Ga, Al, B) zeolites [77]. The results indicated that the HCOOH production rate and total formation rate of oxygenated compounds correlated with total acidity. The samples with weaker acidity showed much lower oxygenate productivity and selectivity. Partial oxidation of C₂H₆ to oxygenates using Fe- and Cu-containing ZSM-5 catalysts revealed that the Fe/ZSM-5 catalyst is highly selective for the conversion of C₂H₆ to a range of oxygenates at appreciable levels of conversion under mild conditions [78]. The reaction pathway is more complicated than that previously proposed for the CH₄ oxidation reaction using similar zeolite catalysts.

2.3. Oxygen as Oxidants

As one of the main components in the air, oxygen (O₂) is an inexpensive oxidant, which is conducive to large-scale application from an economic point of view. However, as an oxidant, oxygen has a high reaction energy barrier and requires a high temperature to activate CH₄ to CH₃OH formation (1). In order to avoid excessive oxidation of CH₃OH or its derivatives, researchers proposed a stepwise process (Figure 13) (i.e., stoichiometric chemical looping), which involves three separate steps: (1) Activation of the metal-zeolite catalyst by an oxidant at a relative high temperature (250–500 °C), (2) CH₄ reaction at a relatively low temperature (25–200 °C), and (3) CH₃OH extraction using a solvent or steam at a relatively low temperature (25–200 °C) [3,32,78]. Both O₂ and N₂O can be used as oxidant for the metal-zeolite catalyst activation (Step 1), N₂O could generally operate under lower temperature than O₂, but in this section we mainly focus on the selected catalysts based on O₂ as oxidant. Currently, Cu and Fe exchanged zeolites have been extensively studied, and significant attention was given to elucidating the nature of the active sites. The identity of the catalytic sites (Figure 14) in the active Cu-zeolites was subject of many spectroscopic studies, which mainly include copper monomers, dimers in the form of mono-μ-oxo and bis-μ-oxo dicopper cores, trimers and larger clusters on the basis of powder XRD, EXAFS, Raman, and UV–Vis data [8,79].

Cu/ZSM-5 have been intensively studied in CH₄ partial oxidation, and the bis(μ-oxo)dicopper site, [Cu₂(μ-O)₂]²⁺, was firstly identified as the active site, evidenced from the absorption band at 22,700 cm⁻¹ in UV–Visible spectra [32,80]. Later, the active species in Cu/ZSM-5 catalyst was further ascribed to a bent mono-(μ-oxo) dicupric site, [Cu₂(μ-O)]²⁺. DFT calculation and normal coordinate analysis of symmetric and anti-symmetric vibrations on resonance enhanced Raman spectroscopy provides insight into how the constrained lattice of ZSM-5 restricts the coordination environment of the bound Cu atoms and their spatial orientation in the formation of active sites for DOMTM [81,82]. This mono(μ-oxo)dicopper core can be formed by activation in both N₂O and O₂ atmosphere as evidenced by observation of the UV–Vis NIR band at 22,700 cm⁻¹ [83]. Additionally, the existence of trinuclear active species, [Cu₃(μ-O)₃]²⁺, has been demonstrated and this trinuclear species in the non-frame-work of the zeolite structure is indeed more stable than the binuclear in the MFI zeolite framework indicated by DFT [84]. In general, the composition and varying topology of copper-modified zeolites are thought to have a large effect on the nature of active sites and on the performance in terms of CH₃OH yield. Higher Si:Al ratios with highly dispersed Al atoms on copper-modified zeolites are more likely to support monomeric active sites, whereas lower Si:Al ratios are likely to stabilize multinuclear copper clusters due to the close proximity of Al atoms on zeolites [85].

Cu-MOR is typically observed to produce a higher CH₃OH yield than Cu/ZSM-5, but more than one site is responsible for CH₃OH formation. Cu/MOR has been suggested to possess both binuclear and trinuclear clusters capable of performing CH₄ partial oxidation [84,86,87]. Whereas a binuclear active site in Cu/ZSM-5 is located in the intersection of the two ten-member-rings, Cu-MOR provides two distinct local structures, situated in the eight-member-ring windows of the side pockets, suggested by spectroscopic observations and DFT simulations [88]. The improved performance in reactivity can be ascribed to subtle difference in the ground states of the Cu-O-Cu sites, indicating the participation of the zeolite lattice in the reaction coordinate. In addition, operando X-ray absorption spectroscopy (XAS) and high-energy-resolution fluorescence-detected (HERFD) and XANES spectroscopy provide a novel perspective on the complex nature and dynamics of Cu-species present in the MOR framework, which enabled an accurate quantification of Cu-speciation on zeolites [33]. The results demonstrated that the productivity increases with a slope of exactly 0.5 when the spectroscopically-determined concentration of active Cu is increased, across a series of materials and activation protocols, suggesting the active site on Cu-mordenite for selective CH₄ oxidation is a dicopper site (Figure 15). The proportion of active Cu in Cu-MOR has been quantified by a spectroscopic method and correlated with reaction performance, and the highest CH₃OH yield (170 μmol_MeOH g_cat⁻¹) has been achieved over Cu-MOR with Si:Al = 7 and Cu:Al = 0.18 [33].

Different framework topologies of Cu-zeolites have been studied in stepwise partial oxidation of CH₄ to CH₃OH, and Cu²⁺ ion-exchanged zeolite omega with MAZ type showed the highest yield (86 μmol·g_cat.⁻¹) among these zeolite materials. Further experimental results revealed the highest CH₃OH yield ever reported (i.e., 150 μmol CH₃OH per gram zeolite under 1 bar CH₄, and as high as 200 μmol CH₃OH per gram zeolite under 30 bar CH₄ were achieved on copper-exchanged omega (MAZ) zeolite) [35,86]. The remarkable CH₃OH yield of Cu-omega is attributed to the relatively high density of copper-oxo active species formed on three-dimensional eight-member-ring channels. Additionally, highly dispersed Cu²⁺ ions in the zeolite pores are essential, which depends on the Si:Al ratio and the preparation method. Higher activation temperature leads to the formation of more copper oxo species, whereas too much copper is not favorable for CH₃OH synthesis [86]. Copper-modified small pore zeolites, including SSZ-13, Cu/SSZ-16, Cu/SSZ-39, and Cu/SAPO-34, have been reported for DOMTM, and the improved CH₃OH yield is attributed in part to the high-temperature water vapor extraction protocol [89]. In particularly, Cu/SSZ-13 could produce competitive quantities of CH₃OH per copper atom, with the highest CH₃OH yield of 125 μmol_MeOH g_cat⁻¹ (0.2 mol CH3OH/mol Cu) [34]. Optimal CH₃OH productivity is correlated with optimal reducibility of Cu/SSZ-13, which is highly dependent on the composition, in terms of Si:Al and Cu:Al ratios (Figure 16). Intermediate Si:Al ratios (∼12−15) and high Cu loading (∼0.5) are prone to produce a high population of Z[Cu^IIOH] precursor sites stabilized within an 8MR CHA zeolite, ultimately resulting in high CH₃OH yields. The combination of testing and XAS measurements clearly evidence a positive linear correlation between the CH₃OH productivity and the reducibility of the Cu centers.

The isothermal conditions were proposed to avoid the time-consuming heating and cooling step during the stepwise cycles. Several studies have shown the ability to run this reaction in an isothermal regime using O₂ or NO as oxidant at 473 and 423 K, respectively [90,91,92]. By operated in an isothermal regime, Cu-FAU realized a stable CH₃OH yield of 90 μmol/g and 92% selectivity at 633 K and ambient pressure without suffering long cycling time and temperature swing [92]. Furthermore, increasing the CH₄ pressure to 15 bar results in a CH₃OH yield of 360 μmol/g, which is the highest value achieved over copper-exchanged zeolite in one cycle. The redox properties of Cu^II-oxo species vary with the structure of the sites, exhibiting different reduction temperature for copper-oxo aggregates and isolated sites, which is an important factor for CH₃OH yield and selectivity. High reduction temperature leads to low activity and CH₃OH yield, while a low copper oxo reduction temperature may tend to high activity toward CH₄ conversion, but with low CH₃OH selectivity. This connection between structure and reducibility of the copper-oxo species can be considered for further designing of high-performance catalysts on DOMTM at desired temperature.

Except Cu-exchanged zeolites, some other catalysts (e.g., Fe-, Co- and Ni-exchanged zeolites) have been reported in both spectroscopy and computational studies based on O₂ or N₂O as oxidants. Co-based catalysts for DOMTM were determined in chemical looping mode analogue to the system with Cu-based catalysts [36,93]. Co/ZSM-5 can realize CH₄ oxidation to oxygenates at low temperature by air as oxidant. Two catalyst preparation methods were adopted to tune the catalytic activity and selectivity of the catalytic system. The samples by ion-exchange are selective towards CH₂O, while impregnated samples with more oxidic cobalt species (CoO and Co₃O₄) are favorable for CH₃OH synthesis [94]. Zn-promoted Cu-containing mordenite was investigated using O₂ as oxidant for DOMTM, and it was observed that Zn addition decreased the CH₃OH yield first, followed by a gradual increase upon higher Zn concentration. The promotional effects of Zn addition could be attributed to the fact that is catalytically active by itself, and to promoting O₂ activation by small ZnO particles [95].

An inverse CeO₂/Cu₂O/Cu(111) catalyst was reported that can activate CH₄ to produce C, CH_X fragments, and CO_X species at room temperature. The addition of water had a strong effect on CH₃OH synthesis by DOMTM (Figure 17), mainly due to OH groups formed by the dissociation of water on the catalyst surface, which could adsorb easily on the active sites with special electronic properties, and thus inhibit the CH_X full decomposition [96]. A further study showed that the site blocking effect of water on Ni/CeO₂ catalyst can improve the CH₃OH selectivity, in which the production of CO and CO₂ are inhibited [39]. Another factor that needs to be considered is the effect of metal−support interaction to bind and activate CH₄ and H₂O, which provides a new insight for designing metal/oxide catalysts for DOMTM.

O₂ as oxidant has been intensively studied during the past decades, and many catalysts has been explored in DOMTM under low temperature and atmosphere pressure. Recently, the economic potential and guidance have been reported on DOMTM via a chemical looping (or redox) process by comparison with the industrial performance criteria [97]. Low CH₃OH productivity and long cycle time lead to an overall production rate that is a factor of ∼50 below the industrial threshold; therefore, enhancing productivity and reducing the cycle time are highly recommended for future studies.

2.4. Water as Oxidant or Co-Oxidant

The oxidizing character of water (H₂O) is weak and; therefore, its use has been rarely reported for DOMTM (4), since it is a strong endothermic reaction.

CH₄ + H₂O → CH₃OH + H₂  ΔH (298K) = 262.2 kJ·mol⁻¹

(4)

Van Bokhoven and co-workers proposed a direct stepwise method for DOMTM over a copper-containing zeolite, based on H₂O as an oxidant [98,99]. Firstly, Cu-MOR was activated at 673 K with helium (Figure 18, Top), and then exposed to 7 bars of CH₄, and finally reacting with H₂O at 473 K. Noticeably, both CH₃OH selectivity and yield remained constant over three additional cycles; the CH₃OH productivity stabilized at 0.202 mol_CH3OH/mol_Cu, and the selectivity reached 97%. In this anaerobic process, H₂O as the source of oxygen can regenerate the zeolite active centers and ensures energetically favorable desorption of CH₃OH. An excess of H₂O can stabilize the reaction intermediates, indicated by isotopic labeling and DFT calculation. The reaction pathway predicted by DFT is illustrated in Figure 18, and the proposed mechanism shows that CH₄ oxidation occurs at Cu^II oxide active centers, followed by Cu^I reoxidation by H₂O with the formation of H₂ [98]. Further insight of the active sites structure has been studied on Cu-MOR [86,100]. Monomeric and oligomeric Cu active sites were synthesized by mordenite with varied Si:Al ratio. Copper oligomer species shows a high activity under both aerobic (O₂ as oxidant) and anaerobic (H₂O as oxidant) activation conditions, whereas copper monomer sites produce CH₃OH only in the aerobic process. This difference is most possibly associated with the stabilization effect of the H₂O molecules on active copper sites [86]. Furthermore, Cu(II)-exchanged mordenites (i.e., Cu/H-MOR and Cu/Na-MOR) were compared with varied Si:Al ratio and Cu:Al ratio, and the results exhibited Cu/H-MOR had a higher methanol yield, which contributed to suppression of CO₂ formation from the methoxy species at the final oxidation step indicated by DFT calculation [100]. Additionally, the activity of Cu-MOR is in line with the amount of Brønsted acid sites in the catalysts on DOMTM oxidized by H₂O. Inspired by the study of anaerobic process on methane oxidation on Cu/MOR, other catalyst, e.g., Cu/SSZ-13, was also studied in a continuous flow reactor at ambient pressure and low temperature (≤250 °C) [101]. Undoubtedly, the DOMTM process with H₂O as oxidant demonstrated here is promising from the perspective of cost and local on-site applications, which may contribute to the development of an industrial process for DOMTM. However, this work has also given rise to several discussions among researchers, mainly on several technical comments and replies questioning and defending the thermodynamic feasibility of the proposed mechanisms [98,102,103,104].

Except for H₂O as a sole oxidant, the H₂O-promoted DOMTM process exhibits a big potential to improve the CH₃OH productivity and selectivity [38,39,87,95,105,106,107]. Recently, a CeO₂/Cu₂O/Cu(111) catalyst was reported for which the reaction can be operated using a mixture of CH₄, O₂ and H₂O as feedstock. H₂O participated in the reaction as the actual oxygen provider and enabled direct CH₄ to CH₃OH conversion. Additionally, H₂O can act as a site blocker, where the facile dissociation at the interfacial Ce sites produced the active *OH, which promotes the DOMTM process [38]. Furthermore, H₂O also functions as an extractor in the hydrogenation of *CH₃O or blocks surface sites to preventing its decomposition, thus facilitating the extraction of CH₃OH [39,95].

In this section, we showed the latest progress of different oxidants (N₂O, H₂O₂, O₂, and H₂O) for DOMTM in heterogeneous catalysis. Attempts on how to reduce the kinetic energy barrier (Ea) of DOMTM are key to improve the CH₄ conversion and CH₃OH productivity, by using suitable oxidants. Generally, strong oxidants result in a lower Ea and lower reaction temperature. However, this is not always true, because different catalysts reduce the Ea to different degrees. Furthermore, most Ea values for DOMTM have been obtained by DFT calculations, but they are subject to uncertainties (e.g., due to the used functionals and dispersion corrections). Therefore, unified methods and standards need to be established to compare the role of catalysts in reducing the Ea of DOMTM. Recently, RuCu nano-sheets were reported on which CH₄ can be efficiently converted to CH₃OH and CH₃OOH (methylhydroperoxide) under mild conditions in the presence of O₂ and trace amount of H₂O₂ [40]. The combination of two or more oxidants (e.g., O₂ + H₂O, O₂ + H₂O₂) may be effective in lowering the energy barrier of CH₄ oxidation in DOMTM, avoiding the over-oxidation of CH₃OH. Additionally, the exploration of efficient catalytic system remains a long-pursued target for DOMTM in heterogeneous catalysis.

3. Plasma Catalysis

Plasma catalysis provides an alternative to heterogeneous catalysis, where the catalytic process is complemented by plasma technology to activate the source gas [108,109,110,111]. This combination is often observed to result in a synergy between plasma and catalyst, which is increasingly gaining attention in many fields, such as CO₂ and CH₄ conversion, NO_x decomposition, NH₃ synthesis, H₂O₂ synthesis, Fischer–Tropsch synthesis, volatile organic compounds removal, wastewater treatment and degradation of pesticide residues [110,111,112]. Typically, the function of a catalyst in a given process is to reduce the activation energy barrier for the rate-limiting reaction and regulate the product distribution. This clear-cut function could also apply to plasma catalysis [113]. However, the effects of a catalyst in plasma catalysis are somewhat different than in traditional heterogeneous catalysis. In general, plasma as the “fourth state of matter” consists of many reactive species (i.e., electrons, molecules, radicals, excited species and ions), which could be adsorbed on the catalyst surface. The surface reactions happening in plasma catalysis are more complex than in conventional thermal catalysis and the possible interaction mechanisms between plasma and catalyst are illustrated in Figure 19 [113]. In this section, we mainly focus on the insights obtained by heterogeneous catalysis based on different oxidants for the further development of the emerging field of plasma catalysis for DOMTM.

The kind of plasma used in plasma catalysis is so-called non-thermal plasma (NTP), where the gas temperature remains near room temperature, while the temperature of the electrons is extremely high (usually in the range of 1–10 eV). This electron temperature is sufficient to activate stable gas molecules (e.g., CH₄, CO₂ and O₂). Hence, NTP offers a distinct approach to enable thermodynamically unfavorable chemical reactions to proceed at low temperature by breaking thermodynamic limits [113,114,115,116,117]. The direct oxidation of CH₄ in NTP seems a promising way of forming oxygenates (e.g., CH₃OH or CH₂O). Activated by plasma, the neutral CH₄ and O₂ molecules can form reactive species, and the major dissociation intermediates are the CH₃ radicals and O atoms [118]. As mentioned above, CH₃OH is more reactive than CH₄ and more likely to undergo oxidation; therefore, moderate conditions (i.e., low temperature) are suitable, which can be provided by NTP. There exist various types of plasmas, such as microwave discharges, glow discharges, gliding arc discharges and dielectric barrier discharges (DBD). The latter are particularly suitable for plasma catalysis, because of their simple operation (atmospheric pressure and near room temperature) and simple design, allowing the easy integration of catalysts in the plasma reactor [108,112], Thus plasma-driven DOMTM is mostly based on DBD plasmas, and recent reports have shown the prospect in CH₄ conversion assisted by plasma technology. An overview of the representative results is summarized in

Table 2. Comparison of CH₄ conversion and CH₃OH selectivity in various plasma and plasma catalysis systems.

System	Oxidant	Packed Material	Power (W)	SEI (kJ/L) a	Temp. (K)	CH4 Conv. (%)	CH3OH Sel. (%)	Ref.
Plasma only	N2O	/	0.27–7.7	-	-	5	43	[121]
	N2O	/	6	6	-	15	31	[122]
	N2O	/	6	6	-	<15	28	[123]
	O2	/	5.7	3.4	RT b	1.9	47	[124]
	O2	/	1.7	3.4	RT	7	20	[125]
	O2	/	118	35.4	288	6	19	[126]
	O2	/	200	4.0	353	3	30	[127]
	Air	/	400	11.9	353	15	13.3	[127]
	Air	/	140	28	523	~25	~8	[128]
Plasma catalysis	N2O	Cu-Ni/CeO2	6	6	-	23	36	[122]
	O2	Fe/γ-Al2O3	1.8	3.6	RT	13	36	[118]
	O2	Glass Beads	1.7	5.1	-	15.4	35.4	[129]
	O2	Cu/γ-Al2O3	1.9	3.6	-	9	37	[130]
	O2	Ga/CZA	50	/	-	54.5	22.2	[131]
	O2	Ni/YSZ	80	160	RT	35.3	23.5	[132]
	Air	CuZn/Al2O3	1.7	3.4	-	11	28	[133]
	Air	Fe2O3-CuO /CP	140	28	473	~26	11	[134]
	Air	Fe2O3/CuO/Al2O3	120	24	473	43	3.7	[135]

^a SEI: Specific energy input; ^b RT: Room temperature.

. Noticeably, DOMTM by plasma catalysis mostly can be operated near room temperature. Hence, H₂O₂ is rarely reported as oxidant, due to its high boiling point (150.2 °C). In addition, CO₂ can be used as a soft oxidant for CH₄ conversion in plasma catalysis (i.e., dry reforming of methane (DRM). This is not the focus of our review, as it has been elaborated in recent reviews [119,120].

3.1. Nitrous Oxide as an Oxidant

Co-processing of CH₄ and nitrous oxide into high-value products, like CH₃OH, CH₂O, etc., was attempted in plasma under ambient conditions [121,122,123,136]. In a DBD reactor, it was observed that the mole ratio of CH₄/N₂O significantly affected the conversion and the product selectivity, and the highest selectivity (28%) of CH₃OH could be realized at 5:1 CH₄/N₂O mole ratio, decreasing to 23% and 13%, respectively, for 1: 1 and 1: 5 ratios. A low ratio of CH₄/N₂O (i.e., 1:5) contains more N₂O in the feed, which results in deep oxidation of the formed oxygenates to CO and CO₂ [123]. Additionally, argon carrier gas played an important role in DOMTM for CH₄/N₂O plasma, by energy transfer from excited Ar atoms to the reactant molecules in this DBD system [121]. The formation of small amounts of C₂H₆ and carbonaceous materials was observed at higher input voltages, next to the main products * i.e., CH₃OH, HCHO H₂, CO and CO₂). Under the conditions of 5% CH₄ and 5% N₂O in Ar, 50 cm³/min total flow rate and 7 kV input voltage, the combined yield of CH₃OH and HCHO reached 10 mol%, but a further rise in residence time led to a lower selectivity of CH₃OH and HCHO [121].

In order to improve the selectivity to CH₃OH, CuO(10%)/CeO₂, NiO(10%)/CeO₂ and Cu-Ni(5-5%)/CeO₂ catalysts were packed in a DBD reactor [122]. The catalytic activity followed the order of CuO-NiO/CeO₂ > NiO/CeO₂ > CuO/CeO₂, with the best CH₃OH selectivity of ∼36% on NiO/CeO₂ catalyst. The best performance is highly attributed to the oxygen vacancies (proven by Raman results), and the increasing amount of oxygen vacancies may facilitate oxygen transfer and improve the catalytic activity [122]. Generally, the generation of energetic electrons is considered as initial step in plasma-initiated reactions [110,136]. In the plasma zone, the presence of abundant electrons could active NO₂ and CH₄ into various radicals. Based on molecular modelling results, the energy barrier for the formation of formyl radical (HCO) is much lower than that of other species [136]. Therefore, the authors suggested that HCO as an intermediate could be the major reason for CH₃OH production [136]. Subsequently, further H addition to form CH₃OH was found to have a very low energy barrier. This investigation confirms that DOMTM with NO₂ as oxidant in non-thermal plasma is a simple and effective way for CH₃OH production. However, the competing reaction (Figure 20) producing toxic gas (HCN and nitric acids) is a main challenge, although higher CH₃OH selectivity can be obtained, compared with other oxidants (e.g., O₂ [136]).

3.2. Oxygen or Air as an Oxidant

Molecular oxygen (O₂) could be a perfect oxidant for plasma catalysis, due to its low cost and lack of harmful by-products, so it can meet the industry’s main goals in “green chemistry”. In CH₄/O₂ or CH₄/air plasmas without catalyst, the effects of various parameters (e.g., reactor configuration, feed ratio, applied voltage, input power, residence time, pressure, specified energy input (SEI) and reactor wall temperature) have been extensively studied [124,126,127,137,138,139,140,141,142]. Oxidative products in these studies mainly included CH₃OH, HCHO, HCOOH, CO and CO₂, and a small quantity of C₂⁺ oxygenates (e.g., ethanol, acetic acid and acetone). Noticeably, avoiding the decomposition of the produced CH₃OH or other oxygenates by the plasma is an important issue [119].

Nozaki et al. reported a single step CH₄ conversion into organic oxygenates by a microplasma reactor, and the organic oxygenates reached 5–20% one-pass yield with 70–30% selectivity. Additionally, syngas was found to be the main product with selectivity of 40% [140]. A further study showed that H₂O₂ with concentration of 0.15 wt.% was also detected in the liquid products [141]. Methyl hydroperoxide (CH₃OOH) was also detected by ¹H-NMR, and it was considered to be an important intermediate for plasma-assisted partial oxidation of CH₄ [141]. The decomposition and further oxidation of liquid organic oxygenates cannot be ignored in plasma-assisted DOMTM. Larkin et al. adopted in situ condensation (a cold trap), as well as reduced residence time, to avoid further reaction of the oxygenates [127,138]. This approach can also be found in subsequent studies [140,142,143]. It can be concluded that a longer residence time may drive total combustion of CH₄ to CO_X and H₂O, and there was an optimal residence time for the production of CH₃OH or other oxygenates [139]. Additionally, the specified energy input (SEI) was compared to consider the effects of power density on the DOMTM process, in which a lower SEI value favors the selectivity toward methanol and suppresses the formation of carbon oxides [127], but an opposite experimental result can be found that the selectivity to oxygenated products (e.g., CH₃OH and HCHO) increased on high SEI value, mainly due to numbers of energetic electrons increased in CH₄/O₂ or CH₄/Air plasma [129,130,133].

Coupling of plasma with catalysts is capable of improving CH₃OH selectivity. Indarto et al. found that ZnO/Al₂O₃ catalyst in a DBD reactor doubled CH₃OH selectivity to around 20%, compared with plasma-only conditions [143]. A similar experiment was found on Ni-doped Y-stabilized Zr catalyst, with optimum CH₃OH selectivity of 23% [132]. Chawdhury et al. found that a CH₄/O₂ plasma packed with glass beads provided an optimal CH₃OH selectivity of 35.4% [129]. Further work showed an improved CH₃OH selectivity of 37% when using CuO/γ-Al₂O₃ catalyst (5 wt.% copper loading). Noticeably, a high Cu loading favored HCHO formation, with 20 wt.% Cu loading yielding the highest HCHO selectivity of 34%. Clearly, plasma packed with catalysts can produce significant amounts of oxygenates [130]. Compared with thermal reduction, a Pd/SBA-15 catalyst reduced by NTP showed much better catalytic performance (i.e., 70% oxygenates selectivity and CH₃OH selectivity of 32%) [125]. The authors suggested the Pd-based catalysts reduced by NTP possess better surface characteristics, which can prevent the recombination and favor surface reactions, such as in-situ coupling, cleavage, H-transfer and CO/CO₂ hydrogenation.

The position of the catalyst, embedded in the plasma discharge region, can largely influence the extent of various synergistic effects and thus the conversion rate. Ceramic-supported Pt, Fe₂O₃ and CeO₂ catalysts were located downstream of the discharge zone to promote CH₄/air plasma reaction, and the Fe₂O₃-based catalyst showed the best catalytic activity and the highest stability [128]. A decrease in the molar ratio of CH₄ to air resulted in more O₂ content available to react with the CH₄ molecules, which led to a higher CH₄ conversion. Packing of the catalysts had no significant effect on the CH₄ conversion, but the CH₃OH selectivity was enhanced for all three catalysts. Chen et al., compared in-plasma catalysis (IPC) and post-plasma catalysis (PPC) for CH₄ partial oxidation to CH₃OH [135]. The experimental results showed that IPC presented a better reaction performance, while PPC exhibited a higher stability because of lower carbon deposition. The reactive species (e.g., O₃, O, CH₂ and CH₃) were identified by in situ optical emission spectroscopy (OES) and FT-IR spectroscopy, and more active species were verified in IPC.

It is known that the synergy between catalytic process and plasma activation depends on both the nature of the packing material and the discharge characteristics. The high CH₃OH selectivity (~23%) on Ni/YSZ catalyst can be attributed to the presence of more surface oxygen vacancies on the YSZ surface [132]. Chawdhury et al. compared the catalytic performance of glass beads, Al₂O₃, TiO₂ and CeO₂ catalysts [129]. In contrast to the expectations, glass beads exhibited better performance than the other catalysts, and the authors attributed this better performance to a more uniform distribution of micro discharges and improved field strength on the glass beads [129]. A porous structure of the catalyst surface is beneficial to the formation of micro discharge inside the pores, although modeling predicted that the latter should only occur for pore diameters larger than the so-called Debye length, which is typically several 100 nm [144,145,146]. Furthermore, plasma parameters, e.g., electron temperature and densities of radicals, ions and electrons, show significant enhancement with increasing dielectric constant of the packed material [111,147]. Therefore, it is reasonable to expect that catalysts with high dielectric constant, porous structure and highly efficient active sites may promote CH₃OH formation in CH₄/O₂ plasma catalysis.

Recently, three supported transition metal catalysts (Fe/γ-Al₂O₃, Ni/γ-Al₂O₃ and Cu/γ-Al₂O₃) were compared in plasma-catalytic DOMTM, and the Fe/γ-Al₂O₃ catalyst exhibited the highest CH₃OH selectivity of 36%, while the Cu/γ-Al₂O₃ catalyst improved the selectivity of C₂ oxygenates to 9.4%, mainly due to more acid sites on the surfaces of the Cu catalyst, as revealed by NH₃-TPD experiments [118]. It is noticed that the plasma-catalytic DOMTM reaction involves both plasma gas-phase reactions and plasma-assisted surface reactions on the catalysts [110,118]. In the plasma gas phase, the reactions are initiated through a variety of inelastic collisions between the reactants (CH₄ and O₂) and energetic electrons, generating a lot of reactive species (e. g., CH_x and O) and excited species, which are believed to play a crucial role for CH₃OH production [118]. Combined with the results of optical emission spectroscopy (OES), the plausible reaction pathways in the plasma gas phase were proposed, as illustrated in Figure 21, and the CH₃ radicals, mainly created by electron impact dissociation of CH₄, were considered as the dominant and critical species in this reaction.

The addition of an Fe/γ-Al₂O₃ catalyst remarkably changed the OES intensities, and the main characteristic peaks (CH, H, C₂, CO and O peaks) of the CH₄/O₂ plasma were lowered by the catalyst packing [118]. The authors explained this by the adsorption of reactive species on the catalyst surface. Surface discharges are dominant in the plasma-catalytic reactions, while filamentary microdischarges were limited when packing catalysts in the discharge region [116,148]. For plasma-catalytic surface reactions, the reactions may take place via Langmuir-Hinshelwood (L-H) and Eley-Rideal (E-R) mechanisms, whereas in thermal catalytic reactions, the L-H mechanism dominates [110]. As shown in Figure 22, the radicals (e.g., CH_x, O and OH) can be formed in the plasma gas phase, and they can be directly adsorbed onto the catalyst surface. These chemisorption CH_x species can be further oxidized to form CH_xO by O and OH radicals, via E-R and L-H mechanisms, to speed up the CH₃OH production by stepwise hydrogenation on the Fe/γ-Al₂O₃ catalyst surface [118].

Kinetic modeling of CH₄ conversion by plasma can help to determine the suitable parameters (such as feed gas ratio, residence time, and discharge power) for optimal performance [149,150]. De Bie et al., developed a one-dimensional fluid model to study the chemistry in a CH₄/O₂ and CH₄/CO₂ DBD plasma [149]. The dominant reaction pathways as predicted by the model for CH₄ partial oxidation into CH₃OH and other oxygenates are shown in Figure 23. Densities of the various plasma species as a function of residence time and gas mixing ratio were studied, and the simulation results showed that CH₄/CO₂ plasma favors the formation of H₂, CH₂O, CH₃CHO, and CH₂CO, while the densities of H₂O₂, CH₃OH, C₂H₅OH, CH₃OOH, and C₂H₅OOH were higher in CH₄/O₂ plasma. CO was formed at high density in both gas mixtures. Basically, the modelling results provide a better understanding of the reaction pathways, which is helpful for future experiments to acquire a maximum selectivity/yield with minimum energy consumption.

Next to the plasma chemistry, the reactions at the catalyst surface can also be studied by kinetic modeling. Loenders et al. investigated how different plasma species affect the partial oxidation of CH₄ into CH₃OH and other oxygenates on a Pt(111) surface [151]. In particular they focused on the effect of vibrationally excited CH₄ and O₂, as well as plasma-generated radicals and stable intermediates on the reaction kinetics and mechanisms. The calculation results revealed that vibrational excitation enhances the turnover frequency (TOF) of catalytic CH₄ dissociation and has good potential for improving the selectivities towards CH₃OH, HCOOH, and C₂ hydrocarbons, but the plasma-generated radicals mainly govern the surface chemistry. They enhance the TOFs of CO_x and oxygenates, increase the selectivity towards oxygenates, and make the formation of HCOOH more significant on Pt(111), compared to thermal catalysis. Such modeling is of great interest for obtaining a better understanding of plasma-catalytic DOMTM [151].

3.3. Water as an Oxidant

In a plasma reactor, energetic electrons are capable of dissociating CH₄ together with H₂O into reactive species (i.e., electrons, radicals, excited species and ions), which makes the direct reaction of CH₄ and H₂O to CH₃OH possible under mild conditions. Tsuchiya et al. investigated a CH₄/H₂O plasma under different discharge conditions, such as gas flow rate, gas-mixing ratio, and discharge power, with total gas pressure of 1–10 Torr [152]. Several gaseous organic products, such as C₂H₆, C₂H₄, C₂H₂ and CH₃OH, were detected, and the CH₃OH selectivity was sensitive to the ratio of CH₄/H₂O. Under optimized conditions, the highest CH₃OH selectivity reported was 20%, and CO was detected to be the major product. The possible reaction pathways for DOMTM process in CH₄/H₂O plasma are illustrated in Figure 24.

More insight in the mechanism in CH₄/H₂O plasma could be obtained from ²H, ¹⁷O and ¹⁸O-labeled isotopologues experiment, and the results indicated that O atoms in the ¹D state inserted into CH₄ could be considered as a possible mechanism for CH₃OH production (i.e., CH₄ + O (¹D) → CH₃OH) [153]. The second most abundant isotopologue of CH₃OH was probably formed from a reaction (CH₂OH + H - CH₂DOH) with CH₂OH radical. In addition, mMatrix isolation FTIR experiments revealed that CH₄/H₂O/Ar plasma at 11 K yielded organic molecules, including CH₃OH, CH₂O, CH₂OH, HOCH and HCO, with CH₃OH as the major product.

3.4. Discussion

The previous sections gave an overview of results obtained by plasma-catalytic DOMTM based on different oxidants (i.e., NO_x, O₂ and H₂O). It is clear that the synergistic effects in plasma catalysis are quite complicated. Evidently, the reaction kinetics, product yields and selectivities are influenced by alternative reaction pathways compared to classical heterogeneous catalysis, and there is clear need for a more fundamental understanding of the mutual interactions between plasma and catalyst. As mentioned in the introduction, the selective activation and controlled conversion of CH₄ into CH₃OH are considered the “holy grail” in catalysis, mainly due to the highly stable C–H bond. This activation could be possibly realized by directing energy into C–H bond vibrations, or by creating CH₃ radicals, which is both possible in NTP at room temperature and atmosphere pressure [109,151,154]. These vibrationally or electronically excited species, or plasma radicals, may facilitate certain steps at the catalyst surface (e.g., dissociative adsorption), allowing other catalysts to perform better in the overall catalytic process, as compared to classical thermal catalysis [110,155]. Additionally, the lifetime of plasma-produced reactive species is a key parameter in determining their effectiveness in plasma catalysis. Figure 25 illustrates the various processes occurring in plasma catalysis, in the full range of time scale, from picoseconds to minutes [112]. The lifetime of reactive species in plasma catalysis is largely influenced by the operating conditions. Therefore, various parameters, such as reactor configuration, feed ratio, applied voltage, residence time, pressure and reactor wall temperature, should be fully considered and adjusted for the desired reaction.

The activity of different catalysts inside the plasma region is determined by multiple factors ranging from chemical composition to physical properties. For DOMTM, another important factor is how to avoid deep oxidation of CH₃OH to CO or CO₂. Based on the mechanism of plasma catalysis, the adsorption probability is essential for all surface-mediated reaction mechanisms, namely the Langmuir–Hinshelwood (L-H), Eley–Rideal (E-R) and Mars–van Krevelen (MvK) mechanisms [110]. Recent work on plasma catalysis for HCN production by CH₄/NH₃ NTP corroborated well with the barrierless Eley-Rideak (E-R) reactions between radicals generated in plasma and adsorbed species over the Cu surface, indicated by DFT calculation [156]. As for DOMTM, it is believed that E-R and L-H mechanism may occur in the case of plasma-catalysis on catalyst surfaces, whereas the L-H mechanism dominates in thermal catalytic reactions [118]. These findings mean that once CH₃ radicals are formed, the subsequent reactions will be rapid. That is, the contribution of both E-R and L-H mechanism will be limited by the dissociation rate of CH₄ through electron impact dissociation, implying that advanced plasma systems with fast dissociation rate of CH₄ will promote CH₄ conversion. On the other hand, pre-adsorption of reactive species is necessary for both E-R and L-H mechanisms to produce CH₃OH. This is important in further studies for rational design of catalysts, and the key is to seek advanced catalytic systems which could accelerate the hydrogenation of important intermediates, such as, CH₃O, as well as the desorption of CH₃OH from the catalyst surface, improving CH₃OH selectivity.

4. Outlook and Conclusions

This review paper provides a comprehensive overview about the recent progress in direct oxidation of methane to methanol (DOMTM) based on different oxidants, by both heterogeneous catalysis and plasma catalysis. Compared with the numerous studies in heterogeneous catalysis, much more research should be carried out in the field of plasma catalysis, to improve the CH₄ conversion and target product selectivity. For this purpose, a better understanding of the underlying mechanisms is required, as they are much more complicated than in heterogeneous catalysis. This should be obtained by modeling, as well as by dedicated experiments. Fundamental studies on the interaction between plasma and catalysts are often limited by the lack of available instruments, especially for in situ measurements, which can provide the fundamental information about the behavior of the different species created in the gas-phase and on the catalytic surface.

As shown in Table 2, the reported selectivity of CH₃OH remains low (<40%) and further work needs to improve the plasma-catalytic performance. Innovations in new catalysts and concepts are needed to seek cost-effective, highly active and stable catalysts. A better catalyst selection strategy can be achieved only by combining advanced level simulation on plasma, catalysis and plasma–surface interactions and validate them with dedicated experiments. To achieve this goal, the advantage of plasma should be fully exploited, and at the same time, insights from heterogeneous catalysis (e.g., catalyst combination, reaction combination, active sites design, etc.) can help to further improve the potential of the promising field of plasma catalysis.

In summary, CH₃OH is an important building block for the chemical industry, and DOMTM by several possible oxidants (N₂O, H₂O₂, O₂ and H₂O) could be interesting for this purpose. However, before DOMTM can be exploited commercially, the CH₄ conversion and CH₃OH yield in heterogeneous catalysis must be further improved, especially by a combination of various oxidants, which may effectively lower the energy barrier of CH₄ oxidation, and avoid the over-oxidation of CH₃OH. For plasma-catalytic DOMTM, a much deeper fundamental understanding of the process is required by means of strong interdisciplinary studies. A combination of computer simulations with experiments will be needed for an in-depth understanding of the reaction mechanisms responsible for the synergy between plasma and catalysts. Therefore, future research should focus on a better understanding and rational screening of highly active catalysts.