1. Introduction

Methane, a feedstock for chemicals and the stable structure of regular tetrahedron, its efficient conversion has attracted growing attention from scientific community and industrial circles [1–3]. It is highly desirable but challenge to transform methane into valuable chemicals, because of its high C-H bonding energy (435 kJ/mol), lack of π bonding or unpaired electrons. Hence, the efficient conversion of methane into value-added liquid oxygenates is regarded as a "holy grail" or "dream reaction" in catalysis.

The conversion and utilization of methane mainly rely on two routes [4,5]. One is indirect conversion route, or syngas route, which is the current commercial route for large-scale transformation of methane [6–8]. This indirect route traditionally takes place under harsh conditions (high temperature or high pressure) with tremendous energy consumption, which exerts great pressure on sustainable development. The other is direct conversion route, in which methane could be transformed directly into liquid oxygenates, syngas, hydrocarbons [9–11]. Among various direct conversion routes, partial oxidation of methane (POM) is one of the most attractive routes due to the less energy input and atomic economy. Compared with partial oxidation methane to syngas, the direct conversion of methane to liquid oxygenates is more attractive and challenging for carbon neutralization and sustainable development. The research of POM to liquid oxygenates (POMO) can be traced back to the early 1970s, and Shilov found that methane could be oxidized into methanol with the K₂PtCl₆ as an oxidant at relatively low temperature (120 ℃) [12]. Although much progress has been made to transform methane into value-added liquid oxygenates since then, it still suffers from extensive energy consumption [13–20]. However, the methane conversion with in-situ generated H₂O₂ maybe bring new inspiration [18].

Photocatalysis, as a burgeoning technology, has recently attracted enormous attention owing to it could directly convert solar energy into chemical energy under mild conditions through water splitting, carbon dioxide reduction, environmental remediation and organic transformation [21–58]. It is therefore extremely attractive and meaningful to achieve efficient partial oxidation of methane into liquid oxygenates via photocatalysis. Photocatalytic partial oxidation of methane to liquid oxygenates (PPOMO) has been paid increasing attention since Kaliaguine firstly reported the CH₄ activation to CH₃O⁻ species by hole centers (O⁻) over TiO₂ under UV irradiation in 1978 [59]. CH₃OH [60], HCHO [61–67], HCOOH [68] were subsequently reported as main products from CH₄ over series of metal oxides under light irradiation. These pioneer works primarily validate the feasibility of photocatalytic direct valorization of methane to value-added liquid oxygenates. However, the PPOMO process is usually accompanied by products overoxidation to carbon oxides (CO, CO₂), which would lead to a low liquid oxygenates selectivity. Hence, it is challenging and significant to regulate the reaction systems, such as selection of oxidants, design of photocatalysts, reaction temperature/pressure/time and light source. Recent years, quite a lot of progresses on catalytic CH₄ conversion have been reviewed [69–80]. Ordomsky summarized the major routes in the photocatalytic methane conversion into chemicals and fuels under mild conditions [81]. The design of efficient photocatalysts for methane conversion was reviewed [82–85]. Most of present review focus either on the thermo/electro/photochemical routes of CH₄ utilization or on material design, while the attention paid on the selection of oxidants and reaction route are limited. Oxidants selection is of great importance to product selectivity as well as the overoxidation regulation. Hence, an updated, profound and thorough review on partial oxidation of CH₄ into liquid oxygenates via photocatalysis from the perspective of oxidants selection and activation is of great significance and necessity.

Among the above-mentioned PPOMO processes, the selection of the oxidant plays a crucial role in activity and selectivity. As shown in Fig. 1a, the selected oxidants mainly focus on H₂O (29.1%), H₂O₂ (16.5%), O₂ (48.1%). As illustrated in Fig. 1b, the selected oxidants were activated to various reactive species for subsequent CH₄ activation and/or conversion. In PPOMO, CH₄ was activated mainly by various reactive oxygen species (ROS), such as ^•OH radicals, O⁻ species, ^•O₂⁻ radicals (Fig. 1b). Photocatalytic methane conversion is a redox reaction driven by energetic charge carriers (Fig. 1c). The band gaps and band positions of typical semiconductor photocatalysts and relative potentials of various ROS or intermediates were shown [86].

In this review, we would like to present a comprehensive and systematic summary on the research achievements in photocatalytic partial oxidation of methane to liquid oxygenates from the perspective of the selected oxidants. Meanwhile, the activation of selected oxidants is emphasized and highlighted, and the activation mechanism of methane in PPOMO is summarized. Finally, challenges and perspectives are briefly discussed. We hope this review could provide guidance on how to design the photocatalytic reaction system and how to get a better understanding of reaction mechanism for the development of efficient photocatalytic methane partial oxidation to liquid oxygenates.

2. PPOMO with various oxidants

As mentioned above, the selection of oxidants is important to methane conversion and the overoxidation regulation. In PPOMO process, the most selected oxidants are H₂O, H₂O₂, O₂ (Fig. 1a). According to the oxidant selected, the research progresses of PPOMO are classified into following categories: PPOMO with H₂O, H₂O₂, O₂ as the oxidant, respectively. Recent progress of PPOMO with various oxidants are summarized in the Table 1.

2.1. PPOMO with H₂O as the oxidant

H₂O is not only a mild and clean oxidant, but also an effective inhibitor for product overoxidation due to the easier oxygenates desorption [82]. Therefore, a number of studies was carried out in exploring H₂O as the oxidant for PPOMO. In this PPOMO-H₂O system, the most employed semiconductors photocatalysts are WO₃, and BiVO₄. This section is thus divided into WO₃ systems, BiVO₄ systems and other systems based on the photocatalysts used.

2.1.1. WO₃ systems

WO₃ was first applied to photocatalytic conversion of methane by Taylor et al. in 1997 [87]. The La-doped WO₃ photocatalyst converted methane to methanol and acetic acid at ~94 ℃ and atmospheric pressure under irradiation [88]. They highlighted the important role of ^•OH radicals in both photochemical and photocatalytic processes. In addition, H₂O is considered as a supplier of ^•OH radicals, which is the main species for CH₄ activation. To gain a deeper understanding of the reaction mechanism, a surface fluorinated WO₃ was developed and investigated [89]. After the fluorination treatment, the CH₃OH production decreased (Fig. 2a). It was proposed that the surface adsorbed ^•OH radicals played a dominant role in the PPOMO process (Fig. 2b). They highlighted the importance of surface adsorbed ^•OH radicals for CH₄ activation and conversion. To clarify the role of La in WO₃ for the PPOMO, the activity and surface properties of WO₃/La were evaluated [90]. The CH₃OH production of over WO₃/La was approximately 2 times higher than that of pure WO₃ while the CO₂ production rate was obviously decreased, resulting in a higher CH₃OH selectivity (Fig. 2c). It was claimed that La doping induced the formation of oxygen vacancies which increased the adsorption of H₂O and modification of the basic-acid properties, further enhanced the generation of ^•OH radicals for CH₄ activation. To promote the ^•OH generation, Ag⁺ impregnated WO₃ was developed. Compared to pure WO₃, 5% Ag₂O@WO₃ exhibited a higher CH₃OH yield. It was proposed that the presence of Ag⁺ on WO₃ surface enhanced the formation of hydroxyl radicals by suppressing the charge carrier recombination process through formed heterojunction (Fig. 2d) [91]. The effect of electron scavengers (Fe³⁺, Cu²⁺, Ag⁺) and H₂O₂ species in PPOMO over a mesoporous WO₃ was studied (Figs. 2e-g) [92]. The photocatalytic activity of WO₃ toward CH₃OH production could be enhanced 2.5 and 1.7 times by adding Fe³⁺ (1 mmol/L) and Cu²⁺ (0.1 mmol/L), respectively. However, the addition of Ag⁺ had a negative effect on this PPOMO process, which lead to a higher CO₂ production.

2.1.2. BiVO₄ systems

BiVO₄, an alternative photocatalyst to binary oxides as WO₃ and TiO₂, has emerged as a proper candidate for PPOMO. BiVO₄ was firstly employed to convert methane into liquid oxygenates by Murcia et al. [93]. The BiVO₄ photocatalyst achieved a CH₃OH production rate of 21 µmol g⁻¹ h⁻¹ and a higher selectivity compared to Bi₂WO₆ and Bi₂WO₆/TiO₂-P25 (Fig. 3a). A ^•OH scavenger (nitrite ions, NO₂⁻) was introduced to improve the CH₃OH selectivity for PPOMO (Figs. 3b and c) [94]. In addition, the CH₃OH selectivity is even higher than 90% between 60 and 120 min at a relatively high NO₂⁻ concentration of 1 mmol/L. It was claimed that the effect of NO₂⁻ as a ^•OH scavenger might limit the overoxidation of produced CH₃OH, thereby suppressing the CO₂ formation and improving the CH₃OH selectivity. Crystal facet engineering was also reported to enhance the activity and selectivity of CH₃OH during PPOMO process (Fig. 3d) [95]. Bipyramidal BiVO₄ comprising {102} and {012} surface facets was more active and selective for CH₄ to CH₃OH compared to the platelet BiVO₄ with {001} facets as its top and bottom surface. Compared to the other two photocatalysts, the bipyramid possesses more surface adsorption oxygen species. A CH₃OH production of 111.9 µmol g⁻¹ h⁻¹ with a selectivity of 850% was achieved over the bipyramidal BiVO₄ photocatalyst. It was proposed that the bipyramids combine efficient extraction of holes from the entire surface of the microcrystal and intermediate surface reactivity leading to high ^•OH production from H₂O oxidation and high activity/selectivity for PPOMO. Other Bi-based photocatalysts, such as, Bi₂WO₆, Bi₂WO₆/TiO₂ [93] and Bi/V modified beta zeolite [96], were also developed and reported. The Bi₂WO₆ and Bi₂WO₆/TiO₂ samples exhibited good ability to convert methane, while the Bi₂WO₆, Bi₂WO₆/TiO₂-P25 tend to produce CO₂ and result in a poor selectivity towards CH₃OH. Compared with the pure β-zeolites (HBEA), Bi-V modified HBEA showed a significant decrease of the undesired CO₂ production and an improvement of CH₃OH selectivity (Fig. 3e), which could be attributed to the formed V₂O₅/BiVO₄ heterostructure and the depressed nonoxidative coupling of CH₄ to C₂H₆.

2.1.3. Other systems

It was reported that an ethanol production rate of 106 µmol g⁻¹ h⁻¹ was achieved with in-situ generated H₂O₂ on Cu-0.5/PCN photocatalyst (Figs. 4a-c) [97]. A negligible H₂O₂ production was obtained on Cu-0.5/PCN compared with that of PCN, while a significant increase of ethanol yield was achieved over the Cu-0.5/PCN. The authors claimed that the Cu species not only decomposed the in-situ generated H₂O₂ to form ^•OH, but also acted as the active sites for CH₄ adsorption and activation, which avoided the deep mineralization and further enhanced the photocatalytic anaerobic methane conversion. H₂O was also claimed to be activated to ^•OH radicals via H₂O₂ (H₂O → H₂O₂ → ^•OH) for CH₄ activation on PCN supported W single-atom photocatalyst (W-SA-PCN), SrTiO₃ supported Co particle (Co-SrTiO₃) [98,99]. The W^δ+ was claimed to be active sites for CH₄ activation and ^•OH production. The in-situ IR results indicated the benefit of W-SA PCN for CH₄ adsorption. The interfacial sites between Co particles and SrTiO₃ are reported for in-situ ^•OH generation and CH₄ adsorption, which attributed to enhanced CH₃OH production. Compared with the raw SrTiO₃, 3.6-Co-SrTiO₃ exhibited superior ability of CH₄ adsorption and activation according to the in-situ DRIFT results. Moreover, according to the experiments they carried out using isotopically labeled reactants (D₂O, H₂¹⁸O, or CD₄), it was claimed that the O in the hydroxyl group of the generated CH₃OH originates from water, whereas H₂O and CH₄ contribute 90% and 10% to the H in the hydroxyl group of the generated CH₃OH, respectively. In another system, H₂O was reported to be activated to ^•OOH radicals via H₂O₂ (H₂O → H₂O₂ → ^•OOH) for CH₄ conversion on Pd modified MoO₃ (Pd-MoO₃) when the reaction temperature was 15 ℃ [100]. The proposed mechanism indicated that PdO species on Pd-MoO₃ played an essential role in the suppression of overoxidation. From the corresponding ¹H NMR spectrum of the isotope labeling experiments, they claimed that the CH₃OH product came from CH₄ oxidation. The ceria nanoparticles with oxygen vacancies was reported to evoke photocatalytic methane conversion to C2 liquid oxygenates (Fig. 4d) [101]. Compare with the CeO₂-raw, a higher ethanol production was achieved over CeO₂-x with oxygen vacancies. They proposed that the oxygen vacancies induced H₂O oxidation to ^•OH (H₂O → ^•OH) and further enhanced the photocatalytic activity. RuO_x decorated ZnO/CeO₂ nanorods (RuO_x/ZnO/CeO₂) [102], Co₃O₄ nanoparticles decorated ZnO (Co₃O₄/ZnO) [103], in-plane Z-scheme hetero-ZnO/Fe₂O₃ [104] and Bi₂O₃ [105] were developed for selectively conversion of CH₄ to CH₃OH. For these systems, the selected H₂O oxidant was claimed to be oxidized to ^•OH radicals (H₂O → ^•OH) for CH₄ conversion. The RuOx species was claimed to be a beneficial site for ^•OH generation from H₂O oxidation. It is proposed that Co₃O₄ plays as the active sites of stabilization of ^•CH₃ adsorption and promotion of CH₃OH desorption to suppress overoxidation. The oxygen vacancy in the Bi₂O₃ surface was claimed to be the active site for the adsorption and activation of oxygen. Moreover, the oxygen vacancy could also enhance the adsorption and fixation of CH₄. Generally, ^•OH generation from H₂O oxidation is responsible for CH₄ activation and conversion. Meanwhile, the in-situ generated H₂O₂ from water oxidation reaction have shown potential for efficient PPOMO. Element doping, crystal plane engineering, cocatalyst modification, and heterojunction construction presents as effective strategies for H₂O activation. H₂O is generally considered as a mild and clean oxidant for PPOMO, while endeavors are highly needed to enhance activity and sustainability.

During the process of PPOMO with H₂O as the oxidant, the ^•OH radical generated from water oxidation plays an important role in PPOMO activity and selectivity. The C-H bond of CH₄ is activated by ^•OH radical into ^•CH₃ radical which will be further converted to CH₃OH (Eqs. 1 and 2). Hence, a certain amount or concentration of ^•OH radical is really needed for CH₄ activation and conversion. La doped WO₃, Ag⁺ impregnated WO₃, BiVO₄ with the expose {001} facet were reported to be useful for charge carrier separation and ^•OH radical generation. Cu species, W^δ+, interface between Co particles and SrTiO₃, RuO_x and Ce particle were claimed to be active sites for ^•OH radical generation. However, the excess ^•OH radical would cause the overoxidation of formed CH₃OH and a low PPOMO selectivity (Eq. 3). NO₂⁻ was reported as a ^•OH scavenger for ^•OH regulation. Co₃O₄ was also reported as an active site for overoxidation suppression. In short, the amount or concentration of ^•OH radical should be well regulated to a proper level.

(1)

(2)

(3)

2.2. PPOMO with H₂O₂ as the oxidant

As analyzed and discussed above, ^•OH radicals are crucial to the activation and conversion of methane. H₂O₂, as an efficient ^•OH generator, has been frequently selected for PPOMO [106–110]. The effect of H₂O₂ for PPOMO was investigated [91,92,111] and the content or concentration of added H₂O₂ should be regulated to a relatively appropriate level [112]. Meanwhile, the efficient utilization of added H₂O₂ should be also highly emphasized.

Sun developed and employed the ZnO nanosheets for PPOMO with H₂O₂ as the oxidant [112]. They then investigated the effect of added H₂O₂. The maximal liquid oxygenates production rate of 2.21 mmol g⁻¹ h⁻¹ was achieved when adding 500 µL H₂O₂ in this system (Figs. 5a and b). However, when more H₂O₂ was added, the products overoxidation to CO₂ would take place and the selectivity of liquid oxygenates would decrease due to the excessive H₂O₂. The selected H₂O₂ was activated by photogenerated electrons to ^•OH and ^•OOH radicals, which would be reacted with formed ^•CH₃ to produce liquid oxygenates. The selected H₂O₂ was also claimed to be activated by photogenerated electrons over a fluffy metal free carbon nitride [113], a dye-sensitized TiO₂ [114] and an atomic-scale Pd supported on 2D titania sheet (Pd₁/2DT) [115]. Hu synthesized the fluffy mesoporous graphitic carbon nitride (g-CN) and investigated its performance of PPOMO at a mild condition (35 ℃) in the presence of H₂O₂. To estimate the effective utilization ratio of H₂O₂ for CH₃OH production, the Gain Factor (GF, the ratio of the molecular amount of CH₃OH produced to the molecular amount of H₂O₂ consumed) was introduced. To develop low-energy input route for PPOMO, Wang developed the dye-sensitized TiO₂ (RhB/TiO₂) and obtained a CH₃OH production rate of 143 µmol g⁻¹ h⁻¹ with 94% selectivity under the irradiation of green light (550 nm). The selected H₂O₂ was claimed to be decomposed to ^•OH radicals by photogenerated electrons transferred from RhB to the CB of TiO₂. The oxygen vacancy in Pd₁/2DT was claimed to be not only facilitated the visible light harvesting of 2DT but also stabilized atomic Pd₁ species.

The efficient utilization of the oxidant of H₂O₂ is related closely to the PPOMO performance. As we known, Fenton reaction could effectively utilize H₂O₂ to form sufficient ^•OH radicals for water treatment and pollutants removal [116–119]. Recently, a synergistic photocatalysis–Fenton reaction for efficient PPOMO under ambient conditions was reported by Wang (Fig. 5c) [116]. They found that photocatalysis–Fenton reaction contributed to an excellent performance of 471 µmol g⁻¹ h⁻¹ for selective production of CH₃OH from CH₄ on TiO₂ compared to Fenton reaction, photo-Fenton reaction, photocatalysis reaction. The introduction of Fe²⁺ to this reaction system promotes the activation of H₂O₂ to ^•OH radicals for PPOMO. And the optimal ratio of Fenton reagents H₂O₂/Fe²⁺ was determined to be 20:1. The excess H₂O₂ would react with ^•OH to produce ^•O₂⁻ and further give rise to the overoxidation of CH₃OH.

Besides, a series of Fe species modified catalysts were reported to efficiently utilize H₂O₂ for methane activation and conversion [120–124]. Tang et al. developed the metal oxides anchored on TiO₂ as photocatalysts for PPOMO with H₂O₂ as the oxidant [120]. In this system, the optimal ratio of CH₄: H₂O₂ was determined to be 8.75:1. The excess H₂O₂ would produce more CO₂, which would lead to a low selectivity to CH₃OH. The Fe species not only promoted the activation of H₂O₂, but also enhanced charge transfer and separation, which further improved the CH₃OH selectivity. H₂O₂ was activated by photogenerated electrons on FeO_x and form ^•OH radicals for PPOMO. Visible-light responsive WO₃ was also developed and applied to PPOMO with H₂O₂ as the oxidant [121,122]. They claimed that FeOOH could enhance the activation of H₂O₂ to form ^•OH radicals for CH₄ activation. H₂O₂ is well known for its homolytic cleavage into ^•OH, but this reaction always accompanied by unexpected O₂ generation. To make full use of the ^•OH generated from H₂O₂, a heterogeneous Fenton-type Fe-based catalyst containing Fe-N_x sites and Fe/Fe₃C nanoparticles was fabricated and studied in PPOMO [123]. They claimed that the Fe-N_x in the low spin state provides active sites for H₂O₂ activation and ^•OH generation. The photocatalyst was also claimed to be effective for CH₄ adsorption. Xie developed the Fe clusters anchored on carbon aerogel (0.75FeCA800-4) for PPOMO with a near 100% CH₃OOH selectivity [124]. The positively charged Fe clusters could not only promote the activation of H₂O₂ to ^•OH, but also donate electrons for CH₄ activation. The three-dimensional carbon aerogel was claimed to be favorable to the adsorption of CH₄ and release the products. On the Cu-PHI photocatalyst, the Cu atom was reported to be active sites for the formation of hydroxyl radicals attributed to the adsorption of hydroxyl group.

H₂O₂, as an efficient ^•OH radical supplier, has been well selected and applied to CH₄ activation and conversion. The efficient utilization of H₂O₂ in PPOMO attracts increasing attention whether activated by photogenerated electrons or combined with Fenton-like reaction. On one hand, the efficient activation of selected H₂O₂ could bring sufficient ^•OH radical should be regulated at a proper level of concentration or content. The concentration of added H₂O₂ should be optimized. Single-atom Pd, various Fe species were introduced to be electrons acceptors for efficient H₂O₂ activation and sufficient ^•OH radicals acquisition. On the other hand, the amount or concentration of formed ^•OH radicals should be well regulated at a proper level due to the overoxidation. The loading amount of Fe species or electrons promoter was optimized to regulate the concentration of ^•OH radicals. These progresses and the discussion of H₂O₂ utilization above might provide instructive insights for further study about PPOMO with H₂O₂ as the oxidant.

2.3. PPOMO with O₂ as the oxidant

O₂ is usually employed as the oxidant in varieties of oxidation reactions because of the ability to form reactive oxygen species (ROS) [125]. From the availability, operating safety and the ability to form ROS, O₂ is an attractive and ideal oxidant for efficient PPOMO. As shown in Fig. 1a, O₂ as the oxidant accounts for 48.1% of the published PPOMO articles. It is therefore important to figure out the role played by O₂, and more attention should be paid to the activation of O₂ in PPOMO process. According to the different reaction systems, the related research works could be classified into two categories: Gas phase systems and liquid phase systems.

2.3.1. Gas phase systems

In gas phase systems, methane was reported to be oxidized to CH₃OH [60,126–129], HCHO [61–67], HCOO [68] with O₂ as the oxidant via photocatalysis.

CH₄ conversion to CH₃OH: Brazdil firstly achieved the optimal CH₃OH conversion rate of 6 µmol/h on CuMoO₄ in the presence of O₂ at 100 ℃ under UV irradiation [60]. Mo-containing porous TiO₂ exhibited a higher CH₃OH production than pure TiO₂ in the presence of O₂ UV irradiation, which could be attributed to the broadened light absorption [126]. Recently, Wu et al. reported that g-C₃N₄ decorated Cs_0.33WO₃ (g-C₃N₄@Cs_0.33WO₃) selectively converted low concentration CH₄ into CH₃OH [127]. The selected oxidant of O₂ was activated to ^•O₂⁻ which dominated the C-H bond cleavage and promoted the generation of CH₃O^• intermediate. More photogenerated electrons from Cs_0.33WO₃ protected CH₃O^• from overoxidation and selectively enhanced the conversion ratio from CH₃O^• into CH₃OH (Figs. 6a and b).

A series of zeolites were also developed and applied to PPOMO with O₂ as the oxidant in gas phase systems [128,129]. It was reported that a deep UV photolysis of CH₄ into C1 liquid oxygenates over beta zeolites containing internal silanol groups, and the product selectivity toward C1 liquid oxygenates (CH₃OH, HCHO, HCOOH) was over 95% with the addition of O₂ [128]. Beta-, delaminated ITQ2 and ITQ6, medium-pore ZSM5 zeolites, mesoporous MCM41 silicates, non-porous TiO₂ were employed as the catalysts for converting CH₄ into C1 liquid oxygenates [129]. The UV irradiation induced the homolytic cleavage of surface hydroxyl group in the micropores of beta zeolite, leading to the formation of silyloxyl radicals (≡Si-O^•) that will generate ^•CH₃ from CH₄ (Fig 6c). It was claimed that confinement and porosity increased the proportion of C1 liquid oxygenates adsorbed onto the solid and reduced the contribution of the gas-phase products. And the presence of O₂ in the systems considerably increased the selectivity toward C1 liquid oxygenates and significantly reduced the production of ethane and other alkanes.

CH₄ conversion to HCHO: PPOMO (HCHO) with O₂ as the oxidant in gas phase systems over various catalysts were summarized in Table 2. In the 1990s, Wada developed a series of metal oxides for PPOMO (HCHO) with O₂ as the oxidant in gas phase [61–66]. To investigate the correlation between the activity and the surface state, the effects of the preparation method (ED: evaporation to dryness; IW: incipient wetness; SG: sol-gel) was studied. They claimed that only four-coordinated vanadium oxide surface species rather than crystalline V₂O₅ act as the catalytic active sites for PPOMO. A vanadium-supported SBA-15 was developed and investigated for the PPOMO by Martinez [67]. The isolated four-coordinated V⁵⁺ species was claimed to be the active centers.

During the PPOMO processes with O₂ as the oxidant in gas phase, the photothermal effect should not be ignored. More attention should be paid to how the photothermal effect is involved in the activation of oxygen molecules and how to promote the conversion of methane or improve the selectivity of products.

2.3.2. Liquid phase systems

In liquid phase systems, produced liquid oxygenates are easy to be removed from catalyst surface by the solvent H₂O, which would inhibit the products overoxidation due to the easier oxygenates desorption. The selected O₂ was activated to various ROS, ^•OOH radicals [130–136], ^•OH radicals [137–143], and ^•O₂⁻ radicals [144–155].

O₂ was activated to ^•OOH radicals: Ye reported that photocatalytic CH₄ conversion to liquid oxygenates in H₂O with a productivity of ~12500 and 2540 µmol g⁻¹ h⁻¹ over Au/ZnO and Au-CoO_x-TiO₂, respectively [130,131]. They claimed that molecular O₂ rather than H₂O, is the source of oxygen for direct CH₄ oxidation through the experiments with isotopically labeled O₂ and H₂O. Moreover, they proposed that O₂ was activated to milder oxidative ^•OOH radicals compared to ^•OH radicals which was claimed to be the culprit of products overoxidation (Fig. 7a). The generated ^•OOH radicals will be coupled with ^•CH₃ to generate CH₃OOH and beneficial to avoid the overoxidation. The CoO_x species was introduced as a cocatalyst to lower the potential of photogenerated holes to suppress the formation of ^•OH radicals from H₂O oxidation. The selected O₂ was also reported to be activated to ^•OOH radicals on binary Au-Cu reactive sites decorated ZnO (Au_0.2Cu_0.15-ZnO) [132], defective In₂O₃ supported atomic Pd (Pd-def-In₂O₃) [133], defective TiO₂ supported atomic Pd (Pd-def-TiO₂) [134]. They claimed that Cu species serve as photoinduced electron mediators to promote O₂ activation to ^•OOH, and Au is a hole acceptor to enhance H₂O oxidation to ^•OH through the in-situ EPR and XPS measurements. On the Pd-def-In₂O₃ and Pd-def-TiO₂ photocatalysts, the oxygen vacancy was reported to be the electron acceptor activating O₂ to ^•OOH.

O₂ was activated to ^•OH radicals: Au nanoparticles modified ZnO photocatalysts reported by Tang et al. exhibited good performance with CH₃OH [137]. The selected O₂ was claimed to be activated into ^•OH radicals which will attack the CH₄ molecules or couple with the formed ^•CH₃ to CH₃OH. They claimed that the O-source of CH₃OH was mainly from H₂O oxidation rather than molecule oxygen. The loading amount of Au nanoparticles was reported to be optimized to regulate the concentration of ^•OH radicals and alleviate the product overoxidation. They also developed a quantum-sized bismuth vanadate (q-BiVO₄) for PPOMO [138], which offered a selectivity of 96.6% for CH₃OH or 86.7% for HCHO under optimal conditions. In this system, O₂ was activated to ^•OH radicals for CH₄ activation. Recently, we reported that the efficient conversion of methane with H₂O₂ in-situ generated from oxygen reduction reaction. A record ethanol production rate of 281.6 µmol g⁻¹ h⁻¹ was achieved over nitrogen vacancy modified carbon nitride (RCN-5) at room temperature [140]. Through the investigation of H₂O₂ and ^•OH intermediates, we proposed that O₂ was activated to ^•OH radicals via H₂O₂ (O₂ → H₂O₂ → ^•OH) (Fig. 7b). Moreover, the optimal concentration of in-situ generated H₂O₂ was estimated to be 217.1 µmol/L. Furthermore, an excellent ethanol production was achieved over optimized P-doped g-C₃N₄ (CNP-50) under atmospheric pressure and room temperature [141]. The P-doping enhanced the ^•OH production from O₂ reduction reaction via H₂O₂ (O₂ → H₂O₂ → ^•OH) and further promoted the ethanol production. A sulfone-modified conjugated organic polymers (S-CTTP) was reported to be able to drive O₂ → H₂O₂ → ^•OH for CH₄ activation and conversion to CH₃OH and HCOOH [142]. A mono-iron hydroxyl site immobilized within a metal-organic framework, PMOF-RuFe(OH) was reported to be efficient for CH₄ conversion to CH₃OH with a time yield of 8.81 ± 0.34 mmol g⁻¹ h⁻¹ and 100% selectivity [143]. Compared to PMOF-Ru, PMOF-RuFe(OH) has a hysteretic desorption, suggesting that the mono-Fe^Ⅲ species acts as a binding site for CH₄. Besides, the produced CH₃OH could desorb rapidly from the MOF matrix in the presence of water. The selected O₂ oxidant was claimed to be photo-reduced to ^•OH radicals via in-situ generated H₂O₂ (O₂ → H₂O₂ → ^•OH). ^•OH radicals could be originated from both water oxidation reaction and oxygen reduction reaction. Noble metal modification, vacancy introduction, and the quantum size effect have been reported to be effective ways to enhance the ^•OH radicals acquisition. The nitrogen vacancy modified carbon nitride and P-doped carbon nitride exhibited excellent ability to obtain ^•OH radicals via in-situ generated H₂O₂ for CH₄ activation and conversion. The regulation of the amount of in-situ generated H₂O₂ or ^•OH radicals could be achieved through optimizing the concentration of introduced nitrogen vacancy and the P doping amount.

O₂ was activated to ^•O₂⁻ radicals: Ye developed the Ag-decorated facet-dominated TiO₂ (Ag/TiO_2{001}) to control the overoxidation of PPOMO [143]. Through the operando FT-IR, in-situ ESR, and NMR characterizations, they found that the oxygen vacancies generated on {001} facets could avoid the formation of ^•CH₃ and ^•OH radicals in the following reaction steps (Ti-O₂^• → Ti-OO-Ti and Ti-(OO) → Ti-O^• pairs) to reduce overoxidation (Fig. 7c). The selected O₂ was activated to ^•O₂⁻ radicals which will be stabilized on a metallic cationic site by the oxygen vacancy. A similar speculative mechanism was proposed on the cuboid shaped WO₃ (c-WO₃) by Fan (Fig. 7d) [145]. O₂ was activated to ^•O₂⁻ radicals which will be stabilized on the W⁵⁺ sites for further CH₄ activation and conversion. Based on the XPS results and band structure, the c-WO₃ was claimed to have a high O₂ adsorption capacity. Furthermore, O₂ could renovate the dissipative crystal-O atom in the continuous reaction within 24 h [146]. From the in-situ DRIFTS, a higher signal intensity ascribed to the adsorbed formate species was observed on TiO₂ compared to that of the W-TiO₂. That means the generated oxygenates desorb from the W-TiO₂ easier and a higher oxygenates selectivity. An HCHO production of 8.03 mmol g⁻¹ h⁻¹ was achieved over a bi-phase catalyst with anatase (90%) and rutile (10%) TiO₂ in a "pause-flow" reactor [147]. The selected O₂ oxidant was activated to ^•O₂⁻ radicals which will be further activated to O⁻ species for CH₄ activation (Fig. 7e). Through the in-situ DRIFTS results and XPS spectra, it was claimed that photothermal synergy with a CH₃OH yield of 12.6 mmol g⁻¹ h⁻¹ was reported on Au-Pd/TiO₂ [148]. Based on the comprehensive study of each component, the reaction mechanism was proposed in Fig. 7f. The selected O₂ was activated to ^•O₂⁻ radicals by photo-excited electrons, then transformed into ^•OOH radicals which will be coupled with formed ^•CH₃ radicals to CH₃OOH (O₂ → ^•O₂⁻ → ^•OOH). It was claimed that Au–Pd nanoparticles not only help to transfer photo-generated electrons, but also absorbed visible light to increase the catalyst temperature by 5 ℃. The temperature rise enhanced the driving force for H₂O oxidation into ^•OH, O₂ reduction into ^•O₂⁻, and the reduction of CH₃OOH into CH₃OH. It was also reported that O₂ was activated to ^•OOH radicals via ^•O₂⁻ radicals (O₂ → ^•O₂⁻ → ^•OOH) on polyoxometalates-based photocatalysts (H₄SiMo₁₂O₄₀/TiO₂ [149], CePMo₁₂O₄₀/TiO₂ [150]), W-TiO₂ [151], defective TiO₂ [152], Cu and W codoped TiO₂ (Cu-W-TiO₂) [153], and Pt/WO₃ [154] during the PPOMO process. Besides, the selected O₂ was reported to be activated to ^•O₂⁻ radicals at the adjacent oxygen vacancy site on 2D BiOCl photocatalyst based on the density functional theory calculations, DMPO-^•O₂⁻ ESR test and time-dependent absorption spectra of nitro blue tetrazolium chloride (NBT) oxidation experiments [155]. The oxygen vacancy was also claimed to be effective sites for CH₄ and O₂ adsorption.

O₂ could be activated to ^•OOH radicals, ^•OH radicals and ^•O₂⁻ radicals for CH₄ conversion, and the activation mechanism was studied by various characterizations, such as EPR, FT-IR, NMR as well as isotope labeling experiments. Although much understanding about the activation mechanism of the selected O₂ has been achieved, more efforts should be made to get a deeper insight into how the O₂ involves the PPOMO and reaction pathway of the O₂.

O₂ selected as the oxidant for photocatalytic partial oxidation of methane into liquid oxygenates shows superiority. We therefore would like to present some typical cases on the photocatalyst design in PPOMO with O₂ as the oxidant in liquid phase. The most employed photocatalysts were TiO₂, g-C₃N₄, ZnO, WO₃. Co-catalyst decoration, vacancy and/or single-atom introduction, crystal facet engineering, crystal phase engineering, and doping have been demonstrated to be useful for PPOMO. CoO_x species was adopted to tune the oxidative site of TiO₂ for suppressing the overoxidation of generated CH₃OH [131]. Nitrogen vacancy was introduced into g-C₃N₄ to promoted in-situ H₂O₂ generation for efficient PPOMO [140]. Pd single-atom was used for C-H bond cleavage. The biphase catalyst with anatase (90%) and rutile (10%) TiO₂ was reported to be useful for the formation of intermediate methanol species [147]. The photocatalyst design mainly focuses on the activation of CH₄, the activation of selected O₂, the intermediates adsorption/desorption and the suppression of products overoxidation.

A comparison of PPOMO with H₂O, H₂O₂, O₂ is listed in Table 3, the advantages and disadvantages of different oxidants are shown.

2.4. PPOMO with other oxidants

Besides the oxidants of O₂, H₂O, and H₂O₂, there were some other oxidants (NO [156–158], ClO₂^• [159] and CO [160]) were also employed for PPOMO. Hu reported that the introduction of NO greatly improved the selectivity of CH₃OH production over the V-MCM-41 (acid) photocatalyst [156–158]. It was proposed that the formed ^•CH₃ could easily interact with NO to generate CH₃OH on the acid V-MCM-41. Ohkubo developed a two-phase system comprising perfluorohexane and water for PPOMO with a CH₄ conversion of 99% [159]. The yields of CH₃OH and HCOOH were 14% and 85%, respectively. Moreover, there were no further CO_x products. Skupinski et al. [160] selected CO as the oxidant and reported the direct CH₄ photocarbonylation experiments which gave CH₃CHO in benzene under UV irradiation over rhodium RhCl(CO)(PR₃)₂ complexes (R = alkyl, Ph, or OPh).

3. Activation mechanisms of CH₄

As we know, the CH₄ molecules is difficult to activate, convert and utilize. It is therefore important to make out the activation mechanism of CH₄. Up to now, CH₄ was reported to be activated mainly by ^•OH radicals, O⁻ species (photogenerated holes), ^•O₂⁻ radicals and other species.

3.1. CH₄ was activated by ^•OH radicals

It is generally acknowledged that ^•OH radicals can abstract the H atom from CH₄. The reactive ^•OH radicals were mainly generated from H₂O oxidation, H₂O₂ decomposition, O₂ reduction.

When H₂O was selected as the oxidant, CH₄ can be mainly activated by ^•OH radicals generated from H₂O oxidation. Taylor firstly reported the activation mechanism of CH₄ and preliminary verified by adding the ^•OH generator of H₂O₂ [87]. Similar mechanism of CH₄ activation was proposed on WO₃-based photocatalysts [87–92,111], BiVO₄-based photocatalysts [93–96], CeO₂ [101], Co₃O₄/ZnO [103], and Bi₂O₃ photocatalyst [105]. A slight difference was found on the activation mechanism of CH₄ over Cu/PCN and W-SA-PCN, CH₄ molecules was claimed to be activated by ^•OH radicals from H₂O oxidation via H₂O₂ (Fig. 5c) [97,98].

When H₂O₂ was selected as the oxidant, CH₄ can be mainly activated by ^•OH radicals generated from H₂O₂ activation. H₂O₂ could decompose by UV-irradiation or photogenerated electrons to ^•OH radicals for CH₄ activation (Figs. 6d and e). Focusing on the efficient utilization of H₂O₂ and generation of abundant ^•OH radicals, various reaction systems and photocatalysts were developed. Especially, a series of Fe species decorated photocatalysts (FeOx/TiO₂ [118], FeOOH/Li_0.1WO₃ [119], FeOOH/WO₃ [120], FeN_x/C-R-T [121] and 0.75FeCA800-4 [122]) were developed for coupling with Fenton reaction to obtain an efficient utilization of H₂O₂. Meanwhile, it is worth noting that the concentration of H₂O₂ should be regulated to a proper level to avoid the products overoxidation.

When O₂ was selected as the oxidant in liquid phase (CH₄-O₂-H₂O), CH₄ can be mainly activated by ^•OH radicals generated from H₂O oxidation or O₂ reduction. CH₄ was reported to be activated by ^•OH radicals from O₂ reduction over AuCu-ZnO [132], Pd-def-In₂O₃ [133] based on the in-situ EPR investigation. CH₄ was reported to be activated by ^•OH radicals from O₂ reduction over q-BiVO₄ [138], Au/BP [139]. In addition, it was claimed that CH₄ was activated by ^•OH radicals generated from both H₂O oxidation and O₂ reduction on Au_x/ZnO [137]. According to the study on intermediates of in-situ generated H₂O₂ and ^•OH radicals, we proposed that CH₄ molecules was activated by ^•OH radicals from O₂ reduction via H₂O₂ over nitrogen vacancy-rich graphite carbon nitride [140,141].

3.2. CH₄ was activated by O⁻ species (photogenerated holes)

It was first reported that CH₄ was activated by the hole center of the O⁻ species under γ irradiation during the PPOMO process [59]. The active O⁻ species can be formed by trapping photogenerated hole with lattice oxygen [64]. CH₄ was reported to be activated by O⁻ species on some transition metal oxides, MoO₃ [61,62], vanadium oxide [66,67,156–158], ZnO [63,112,130], TiO₂ [131,143], WO₃ [145]. The O⁻ species dominated activation mechanism of CH₄ is shown in Eq. 4 and Fig. 8a. The O⁻ species could abstract a H atom from CH₄ to form ^•CH₃, which would take part in the subsequent reaction to liquid oxygenates.

(4)

The CH₄ activation was reported to be activated by tetrahedrally coordinated vanadium oxide species (O⁻ species) over vanadium-containing MCM-41 [158]. It was claimed that V⁴⁺ species and ^•CH₃ formed by UV irradiation of the exposed VO₄ species with methane interact easily with NO, resulting in the selective formation of CH₃OH. The charge transfer excited triplet state (V⁴⁺-O⁻)* species, especially the O⁻ hole trap center reacts with CH₄, leading to a break in the C-H bond of CH₄ to form V⁴⁺-OH and ^•CH₃ radicals (Fig. 8b). The *OH species then released from V⁴⁺-OH would react with the ^•CH₃ radicals to form CH₃OH. V₂O₅/SiO₂ [66], mesoporous VO_x/SBA-15 [67], V-MCM-41 mesoporous molecular sieves [156,157] were also reported to be active for CH₄ activation with the tetrahedrally coordinated vanadium oxide species (O⁻ species).

Ye claimed that CH₄ was activated by O⁻ species (photogenerated hole centers) of ZnO rather than ^•OH radicals through a comparison experiment and DFT calculations [130]. The comparison experiment of ^•OH radicals generation using coumarin as probe molecule revealed that cocatalyst/TiO₂ were more efficient than cocatalyst/ZnO, which was inconsistent with the photocatalytic activities. And the DFT calculations showed that the reaction energy barrier of H abstraction of CH₄ by O atoms on ZnO is lower than that of TiO₂, which was in accordance with experimental results of photocatalytic CH₄ oxidation. It was therefore claimed that CH₄ activation over ZnO was dominated by O⁻ species. Specially, two-dimensional oxide semiconductors (ZnO nanosheets) are designed to obtain abundant active O⁻ species for activating C−H bond of CH₄ [112]. The in-situ EPR study verified that lattice O atoms could capture photoexcited holes and generate active O⁻ species, which will abstract H from CH₄ to form ^•CH₃ radicals.

It was reported that CH₄ was activated by photogenerated holes (O⁻ species) from the VB of TiO₂ to ^•CH₃ radicals, which would react with formed ROS to produce liquid oxygenates [116,118,149–153]. On the Au-CoO_x/TiO₂, it was claimed that the photogenerated holes on TiO₂ rather than Co₃O₄ could oxidize CH₄ to ^•CH₃ according to DFT calculations. More specifically, on TiO_{2 {001}} samples, Ti-O^••O-Ti species were claimed to activate CH₄ molecule through the ESR and NMR measurement Fig. 7c) [143]. Under UV irradiation, the surface oxygen (Ti-O-Ti) of TiO_{2 {001}} sample was oxidized by photoinduced holes to form oxygen vacancies (Eq. 5), which can stabilize the formed ^•O₂⁻ (Eq. 6). And the Ti-O^••O-Ti was considered as the active species for CH₄ activation. Details are as following Eqs. 7–9. While on TiO₂₍₁₀₁₎ samples, Ti-O^• species were claimed to activate the CH₄ following the Eqs. 10 and 11. The formed ^•CH₃ and ^•OH were claimed to be unfavorable to liquid oxygenates selectivity.

(5)

(6)

(7)

(8)

(9)

(10)

(11)

It was also reported that the CH₄ was activated by O⁻ species on Au/c-WO₃ according to the isotope characterization, XPS spectra and in-situ ESR test (Fig. 7d) [145]. On the crystal-O exposure site of c-WO₃, surface W⁵⁺ (electron) and O⁻ ion (hole) centers were generated through the charge separation under irradiation. Then the formed O⁻ ion (hole) would engage in H-atom abstraction from CH₄ by being inserted into the C-H bond. The formed surface methoxy (-O-CH₃) would take part in the subsequent reaction for HCHO formation. Moreover, the long-time (24 h) stability for HCHO yield was explained to the timely renovation of dissipative crystal-O atom by O₂. The cocatalyst of Au was claimed to promote high O⁻ ion (hole) formation through the in-situ ESR. On the FeOOH/Li_0.1WO₃ [119] or FeOOH/m-WO₃ [120], CH₄ was claimed to be activated to ^•CH₃ radicals by both photogenerated holes (O⁻ species) and ^•OH radicals. The main product of CH₃OH was generated through the reaction between formed ^•CH₃ radicals and ^•OH radicals.

3.3. CH₄ was activated by other species

It was reported that CH₄ was activated by the ^•O₂⁻ radicals on g-C₃N₄-decorated Cs_0.33WO₃ [127] and oxygen vacancy modified BiOCl [155] under irradiation during the PPOMO process with O₂ as the oxidant. The selected oxidant of O₂ was activated to ^•O₂⁻ radicals, which would subsequently activate CH₄ to form CH₃OH. The difference is the reaction pathway towards CH₃OH. On the g-C₃N₄@Cs_0.33WO₃ photocatalyst, it was claimed that CH₄ was activated to methoxyl radicals, which would be converted to CH₃OH Fig. 6a). On the BiOCl-OV photocatalyst, it was claimed that ^•O₂⁻ radicals attacked CH₄ and initiated the breaking the CH₃-H and the O-O bond together. CH₄ was activated to adsorbed CH₃OH, which would desorb from the catalyst surface easily (Eqs. 12 and 13). However, more direct evidence should be provided for a further confirmation of this ^•O₂⁻ radical dominated CH₄ activation mechanism.

(12)

(13)

It was reported that CH₄ could be activated to ^•CH₃ by silyloxyl radicals (≡Si-O^•) on series of zeolites [128,129]. Under deep UV irradiation (λ < 200 nm), the surface hydroxy group (≡Si-OH) was broken into ≡Si-O^• radicals, which could activate CH₄ to ^•CH₃ by H abstraction. The formed ≡Si-O^• radicals and ^•CH₃ radicals then combined and produced CH₃OH (Fig. 6c and Eqs. 14–17).

(14)

(15)

(16)

(17)

The ClO₂^• was reported to be an efficient oxidant for PPOMO with a 99% conversion [159]. The methane oxygenation is initiated by the photochemical Cl-O bond cleavage of ClO₂^• to generate Cl^• and O₂. The produced Cl^• reacted with CH₄ to form a ^•CH₃, CH₃OH and HCOOH were given by the radical chain reaction (Fig. 8c). The fluorous solvent plays an important role of inhibiting the deactivation of reactive radical species such as Cl^• and ^•CH₃.

4. Conclusions and prospects

Photocatalytic partial oxidation of methane to liquid oxygenates (CH₃OH, C₂H₅OH, HCHO, CH₃CHO, HCOOH, CH₃OOH, etc.) is a sustainable and potential technology towards addressing the increasingly sever energy shortage and environmental deterioration. In this review, we have highlighted research progresses in the PPOMO from the perspective of selected oxidants (including H₂O, H₂O₂, O₂, NO, ClO₂^• and CO). Moreover, the activation of selected oxidant (such as, H₂O₂ and O₂) is also underlined to gain a deeper understanding on reaction mechanism. We then discussed the activation mechanism of CH₄, which was mainly rely on ^•OH radicals and active O⁻ species.

Although remarkable achievements have been made in PPOMO, there are still many challenges to be addressed in exploring efficient reaction systems to its full potential. To make a small step forward and broaden the research scope of the PPOMO, our discussions are as follows:

(1) The reaction systems should be well designed. Temperature, pressure, light source, reaction time as well as the selection of oxidant take an important part in efficient PPOMO. For example, in the CH₄-O₂-H₂O systems, a pressurized reactor was employed to increase the solubility of CH₄ and O₂ in H₂O. The amount of H₂O is significant to the desorption of generated products (CH₃OH, C₂H₅OH, CH₃OOH, etc.) to avoid the overoxidation. Reaction time is also an indispensable factor for the activity and selectivity of PPOMO. A "pause-flow" reactor was also reported for efficient CH₄ conversion. Hence, elaborate design of the reaction systems and precise regulation of each component are exactly needed for achieving an efficient PPOMO with high activity and high selectivity.

(2) The design of novel photocatalysts is in urgent demand. ZnO nanosheets was designed to generate abundant active O⁻ species for activating C-H bond of CH₄. A series of containing Fe species photocatalysts were developed for efficient H₂O₂ utilization to enhance the PPOMO performance. Ag-decorated facet-dominated TiO₂ was reported to efficient and selective photocatalytic CH₄ conversion to CH₃OH with O₂ by controlling overoxidation. Taking the adsorption of substrates into account, the design of photocatalyst surface should be highly heeded. TiO₂, WO₃, ZnO are emerging as efficient metal oxide photocatalysts for CH₄ activation in PPOMO. The acquisition and regulation of the reactive oxygen species could be addressed mainly through the well design of photocatalyst and reaction systems. Crystal facet engineering, Fe or Cu species, single-atom catalyst, vacancy, defects, and three-dimensional structure have been demonstrated to be effective methods to construct active sites for CH₄ adsorption and O₂ adsorption. The solvent may tune the hydrophilic/hydrophobic properties of the photocatalyst surface and the adsorption of intermediate species. Moreover, the metal sites immobilized with MOF was also reported to be efficient for PPOMO. Meanwhile, given the sunlight spectrum distribution and long-term sustainable reaction, visible light-responsive and stable photocatalyst should be developed.

(3) More attention is deserved should be paid on the PPOMO with O₂ as the oxidant in liquid phase due to lower energy input and higher conversion & selectivity. Photocatalytic conversion of methane with H₂O₂ in-situ generated (^•OH radicals) from whether oxygen reduction reaction or water oxidation reaction has shown great potential in this field. Sufficient ^•OH radicals could efficiently activate the CH₄ molecules. However, the excessive ^•OH radicals would like to over oxidize the produced oxygenates and lead to a low oxygenates selectivity. In a word, the acquisition and regulation of in-situ generated H₂O₂ or ^•OH radicals are of great importance have been two important challenges in photocatalytic methane conversion. After obtaining sufficient ^•OH radicals, the amount or concentration of in-situ generated H₂O₂ or ^•OH radicals should be regulated at a proper level.

(4) Advanced characterization technologies are of great significance in understanding activation mechanism of the oxidant as well as CH₄. In-situ EPR was employed to detect the ROS for insights into the mechanism of O₂ activation. Operando FTIR and NMR were carried out to investigate the reactive intermediates for the mechanism of CH₄ activation. Meanwhile, isotope labeling experiments were usually employed to make out the O origination of liquid oxygenates. Specially, to identify the origination of liquid oxygenates products, isotope labeling experiments are necessary in further studies. Therefore, novel characterization technologies should be developed and applied to gain a deeper understanding towards the activation mechanisms of both the selected oxidants and CH₄.

(5) An unprecedentedly high CH₃OH yield of 12.6 mmol g⁻¹ h⁻¹ was achieved on Au-Pd/TiO₂ via a photothermal synergy. The performance of PPOMO process may be enhanced when coupled with thermocatalysis or electrocatalysis or both. Thus, the synergistic effect should be paid more attention.

In summary, photocatalytic partial oxidation of methane to liquid oxygenates is currently attracting enormous attention. Various research explorations have been expected to soar in the next few years. This review summarized the research progresses in PPOMO from the perspective of the selected oxidants to promote this field step forward and provide new understandings or enlightenments.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

We acknowledge support from the National Key R&D Program of China (No. 2021YFA1500800) and National Natural Science Foundation of China (No. 22072106).