Precisely Modulating the Interfacial Oxygen Radicals for Selective Photocatalytic Methane Partial Oxidation to Methanol

ABSTRACT

Selective photocatalytic partial oxidation of methane (CH₄) under ambient conditions offers a sustainable route for direct methanol (CH₃OH) production while reducing greenhouse gases. However, achieving both high yield and selectivity remains challenging due to the uncontrolled generation of reactive oxygen species (ROS), particularly hydroxyl radicals (•OH), which readily overoxidize CH₃OH. Here, we report a PdNi alloy cocatalyst supported on TiO₂ that enables fine‐tuning of interfacial ROS dynamics via electronic and structural modulation. Mechanistic studies reveal that the PdNi interface couples moderate kinetic barriers for *O protonation with thermodynamic enhancement of *OH adsorption via d‐band center modulation. This cooperative effect suppresses the transition of O₂ into reactive •OH species, thereby mitigating CH₃OH overoxidation. As a result, the optimized PdNi_0.49/TiO₂ photocatalyst achieves a CH₃OH yield of 9557.4 µmol g⁻¹ h⁻¹ with 87.5% selectivity using molecular O₂ as the sole oxidant. Beyond regulating ROS dynamics, Ni incorporation also optimizes the desorption behavior of CH₃OH, further contributing to the high selectivity of the target product. This work establishes a rational interfacial design strategy for steering photocatalytic oxidation pathways and advancing selective methane conversion technologies.

Interfacial PdNi alloy cocatalysts enable selective methane‐to‐methanol photooxidation by fine‐tuning ROS kinetics and optimizing methanol desorption. The PdNi interface couples moderate kinetic barriers for *O protonation with thermodynamic enhancement of *OH adsorption via d‐band center modulation. This cooperative effect suppresses the transition of O₂ into reactive •OH species, thereby mitigating CH₃OH overoxidation.

1. Introduction

The direct conversion of methane (CH₄) to value‐added oxygenates under ambient conditions offers a sustainable route to utilize stranded natural gas reserves and mitigate flaring‐derived CO₂ emissions [1, 2, 3, 4]. However, as the simplest alkane, CH₄ possesses a high bond dissociation energy (439 kJ mol⁻¹), rendering it kinetically inert under mild conditions [5]. Conventional thermal processes typically require elevated temperatures (> 400°C) or strong oxidants (e.g., H₂O₂), which compromise energy efficiency and raise environmental concerns [6, 7]. Among various oxygenates, methanol (CH₃OH) is particularly attractive due to its high energy density (15.6 MJ L⁻¹), ease of storage and transport, and versatile applications as a chemical feedstock and hydrogen carrier [8, 9]. Industrially, CH₃OH is produced via a multi‐step syngas‐based approach, involving steam reforming of CH₄ at high temperatures and pressure to generate CO and H₂, followed by catalytic conversion over Cu/ZnO/Al₂O₃ under 5–10 MPa and 200°C–300°C [10, 11]. Nevertheless, the syngas route is energy‐ and capital‐intensive, limiting its deployment in remote sites where methane flaring remains problematic, which drives interest in low‐temperature, single‐step CH₄‐to‐CH₃OH conversion for distributed production and better carbon utilization.

Photocatalysis offers a sustainable alternative by harnessing solar energy to overcome activation barriers and drive uphill reactions under ambient conditions [12, 13, 14]. In semiconductor‐based systems, photoexcited electrons and holes mediate oxygen reduction and water oxidation, respectively, generating reactive oxygen species (ROS) such as hydroxyl radicals (•OH) and hydroperoxyl radicals (•OOH). These ROS initiate CH₄ activation and govern product selectivity through distinct oxidative pathways [15, 16]. Under aerobic conditions, •OH can arise from either hole‐mediated water oxidation or electron‐driven oxygen reduction, while •OOH is primarily formed via single‐electron reduction of oxygen [15, 17]. The reactivity of these species diverges: •OH efficiently abstracts hydrogen from CH₄ to form methyl radicals (•CH₃), which can further react with •OH to yield CH₃OH. However, •OH also readily oxidizes CH₃OH to HCHO and CO₂, leading to undesired deep oxidation [18]. In contrast, •OOH tends to couple with •CH₃ to generate methyl hydroperoxide (CH₃OOH), which may either desorb and decompose to HCHO or remain surface‐bound for selective reduction to CH₃OH [19, 20, 21]. The concurrent generation of multiple ROS and the coexistence of parallel reaction channels often result in overoxidation and complex product distributions, posing a major challenge to achieving high CH₃OH selectivity [22, 23].

Recent studies have highlighted the critical role of dominant ROS species in determining product outcomes during photocatalytic CH₄ oxidation. Catalysts favoring •OH generation typically enhance CH₃OH formation, whereas those dominated by •OOH tend to yield CH₃OOH or HCHO. For example, Au single atoms on In₂O₃ preferentially generate •OOH via end‐on O₂ adsorption [24], while Au nanoparticles promote •OH formation through side‐on adsorption, enabling selective CH₃OH production. Similarly, Au‐Pd alloy nanoparticles on ZnO facilitate O–O bond cleavage in *OOH intermediates [17], accelerating in situ •OH generation and improving CH₃OH yield and selectivity. Ru₁/ZnO single‐atom catalysts exhibit significantly higher CH₃OH selectivity than Ru₁O_x/ZnO [23], attributed to efficient •OOH‐to‐•OH conversion at unsaturated RuO_x sites. Despite these advances, precise regulation of •OH remains underexplored. Insufficient •OH limits CH₃OH productivity, while excessive •OH leads to overoxidation and reduced selectivity. Given that molecular O₂ serves as the primary ROS precursor, tuning the oxygen reduction reaction (ORR) pathway is essential for controlling •OH dynamics. The 3‐electron ORR predominantly yields H₂O and favors CH₃OH formation, whereas the 2‐electron pathway produces H₂O₂ and CH₃OOH [20]. Alloying transition metals with noble metals provides a powerful strategy to steer ORR selectivity by modulating the interaction between reactive intermediates and catalytic sites [25, 26]. Tailoring the 2e⁻/3e⁻ ORR balance through alloy design thus offers a promising route to optimize •OH generation and enhance CH₃OH yield and selectivity.

To address the challenge of uncontrolled ROS generation, we propose an interfacial engineering strategy to modulate oxygen‐derived radical dynamics. A PdNi alloy cocatalyst supported on TiO₂ was constructed to exploit synergistic electronic and structural effects for selective CH₄ oxidation to CH₃OH under ambient conditions. Atomic‐level alloying of Pd and Ni plays a pivotal role in steering product selectivity: PdNi_0.49/TiO₂ achieves a CH₃OH yield of 9557.4 µmol g⁻¹ h⁻¹ with 87.5% selectivity, outperforming both metallic Pd/TiO₂ and Pd‐NiO_x/TiO₂. Mechanistic investigations reveal that Pd/TiO₂ favors the 3‐electron ORR pathway, leading to excessive •OH generation and subsequent CH₃OH overoxidation. Pd‐NiO_x/TiO₂, in contrast, promotes H₂O₂ formation via the 2‐electron route. For PdNi_0.49/TiO₂, Ni incorporation maintains moderate kinetic barriers for *O protonation while simultaneously upshifting the Pd d‐band center to strengthen *OH adsorption. This dual modulation of kinetics and thermodynamics suppresses excessive •OH formation, sustaining an optimal ROS balance that enables efficient CH₄ activation while preserving CH₃OH selectivity. In parallel, Ni incorporation facilitates CH₃OH desorption, which further helps suppress overoxidation and enhances product selectivity.

2. Results and Discussion

2.1. Catalytic Performance of CH₄ Conversion With O₂

Photocatalytic CH₄ oxidation was conducted at room temperature in a stainless‐steel autoclave equipped with top irradiation, under initial conditions of 1.6 MPa CH₄ and 0.4 MPa air. Reactions were performed for 1 h using a 365 nm LED lamp (75 mW cm⁻²). Liquid products (CH₃OH and CH₃OOH) and gaseous CO₂ were quantified by ¹H nuclear magnetic resonance (NMR) and gas chromatography (GC), respectively, while HCHO was determined via a colorimetric method (Figure S1) [27]. Commercial TiO₂ nanoparticles composed of approximately 80% anatase and 20% rutile phases were selected as the support. Noble metal cocatalysts (Au, Ag, Pt, Pd) were deposited via NaBH₄ reduction method to assess their influence on CH₄ oxidation. Pd/TiO₂ synthesized via this method is hereafter referred to as Pd/TiO₂‐NaBH₄, serving as a reference monometallic catalyst. Among the tested cocatalysts, Pd/TiO₂‐NaBH₄ exhibited superior CH₃OH yields and selectivity (Figures S2 and S3), highlighting its potential for selective photocatalysis. To suppress CH₃OH overoxidation, transition metals were introduced via two synthetic routes to modulate surface reactivity and electronic structure (Figure S4). For Pd‐NiO_x/TiO₂, NiO_x was first deposited by the impregnation method, followed by Pd nanoparticle loading via NaBH₄ reduction. In contrast, PdNi/TiO₂ was synthesized via a one‐pot hydrothermal method by co‐introducing Pd(acac)₂ and Ni(acac)₂, enabling in situ alloy formation [28]. In this hydrothermal synthesis, n‐butylamine acted as a bifunctional agent to stabilize particle growth and facilitate Ni^{2 +} reduction, while CO functioned as both a reducing agent and a structural director [29, 30, 31, 32, 33]. Through coordinating with Pd precursors and mediating the nucleation kinetics, CO prevents the formation of ultra‐small clusters and instead promotes well‐dispersed nanoparticles, which is essential for selective CH₃OH production (Figures S5 and S6) [24]. The corresponding monometallic control, Pd/TiO₂ (hereafter referring to the hydrothermally synthesized Pd‐only catalyst), was prepared under identical hydrothermal conditions without Ni precursor. Additional Pd‐MO_x/TiO₂ (M = Fe, Co, Ni, Cu) and PdM/TiO₂ alloy‐type catalysts were prepared via sequential impregnation or one‐pot hydrothermal methods, respectively. Metal loadings were confirmed by inductively coupled plasma–optical emission spectrometry (ICP‐OES, Tables S1 and S2).

Photocatalytic performance of the synthesized catalysts was systematically evaluated (Figure 1a and Table S3). Pristine TiO₂ primarily yielded HCHO (4830.8 µmol g⁻¹ h⁻¹) with low CH₃OH selectivity (14.0%). While transition metal oxides suppressed HCHO and CO₂ formation (Figure S7), they did not significantly enhance CH₃OH selectivity. However, subsequent Pd loading onto these supports (e.g., Pd‐FeO_x/TiO₂, Pd‐CoO_x/TiO₂, Pd‐NiO_x/TiO₂, Pd‐CuO_x/TiO₂) via NaBH₄ reduction method markedly improved CH₃OH selectivity (>85.0%), albeit with reduced yields. Among them, Pd‐NiO_x/TiO₂ achieved the highest CH₃OH yield (2286.3 µmol g⁻¹ h⁻¹) and selectivity (85.9%), underscoring the beneficial role of NiO_x. This trend is consistent with previous theoretical studies [34], which demonstrated that Ni exhibits stronger chemical adsorption capacity and more favorable electronic structure modulation than other transition metals, thereby facilitating C–H bond activation while suppressing overoxidation, leading to enhanced CH₃OH activity and selectivity observed in our system. Catalysts synthesized via a one‐pot hydrothermal method exhibited notably higher CH₃OH yields. For example, Pd/TiO₂ delivered a CH₃OH yield of 4268.5 µmol g⁻¹ h⁻¹ with 65.6% selectivity. In comparison, PdNi_0.49/TiO₂, featuring an alloyed PdNi interface, achieved a CH₃OH yield of 3908.2 µmol g⁻¹ h⁻¹ and selectivity of 91.8%, outperforming both Pd/TiO₂ and Pd‐NiO_x/TiO₂. These results highlight the effectiveness of Pd–Ni alloying in steering reaction pathways toward partial oxidation products.

FIGURE 1.

Performance evaluation of selective photocatalytic oxidation of methane. (a) CH₃OH yield and selectivity over TiO₂ and different metal/TiO₂ samples. (b) Influence of Ni/Pd molar ratio on CH₃OH production. Reaction conditions: 0.625 mg catalyst, 5 mL water, 1.6 MPa CH₄, 0.4 MPa air, 25°C ± 3°C, 1 h. (c) Effect of catalyst concentration on photocatalytic activity. Reaction conditions: 5 mL water, 1.6 MPa CH₄, 0.4 MPa air, 25 ± 3°C, 1 h. (d) Tentatively proposed photocatalytic mechanism over Pd/TiO₂, Pd‐NiO_x/TiO₂ and PdNi/TiO₂. (e) Comparisons with the representative photocatalytic performances on the yield and selectivity of CH₃OH.

The influence of Pd loading was further investigated (Figure S8). Increasing Pd content from 0.1 to 4.0 wt% enhanced CH₃OH yield (1024.9 to 4268.6 µmol g⁻¹ h⁻¹) and suppressed HCHO formation. However, at a Pd loading of 6.0 wt%, the CH₃OH yield decreased despite enhanced selectivity, most likely ascribed to the combined effects of light shielding and accelerated photogenerated electron–hole recombination at high Pd coverage. To optimize the Pd–Ni synergistic effect, PdNi/TiO₂ catalysts with different Ni/Pd molar ratios were prepared (Figure 1b). A Ni/Pd ratio of 0.08 significantly improved CH₃OH selectivity to 91.3% but reduced the yield. Increasing the Ni/Pd ratio to 0.49 enhanced the CH₃OH yield without compromising selectivity. Further increasing the Ni content raised the selectivity to 94.8% but decreased the yield, indicating that Ni suppresses overoxidation but may hinder CH₄ activation at excessive levels. Given this trend, the catalyst with a Ni/Pd molar ratio of 0.49 was selected for further study. The catalyst concentration was optimized to mitigate interparticle light‐shielding (Figure 1c). At an optimal catalyst concentration of 0.05 mg mL⁻¹, CH₃OH yield reached 7630.5 µmol g⁻¹ h⁻¹ with 85.2% selectivity. Lower concentrations led to CH₃OOH accumulation and reduced CH₃OH formation. Increasing system pressure from the initial 2.0 MPa to 3.0 MPa further improved CH₃OH yield to 8576.3 µmol g⁻¹ h⁻¹ while maintaining high selectivity (Figure S9). Enlarging the water volume to 7.5 mL enhanced CH₃OH productivity to 9557.4 µmol g⁻¹ h⁻¹ with 87.5% selectivity (Figure S10). Under a fixed total pressure of 30 bar, air partial pressure was varied to investigate product distribution (Figure S11). A moderate amount of O₂ facilitates CH₄ activation and CH₃OH formation, with the highest CH₃OH yield and selectivity at 6 bar air. Further increasing air pressure caused a significant decline in both CH₃OH yield and selectivity, indicating that excessive O₂ induces CH₃OH overoxidation.

In addition, the apparent quantum efficiency (AQE) under monochromatic 365 nm irradiation was evaluated to be 1.06%. Although lower than some state‐of‐the‐art reports [12, 35, 36], this value primarily reflects reactor geometry rather than intrinsic catalytic activity. Specifically, bottom‐up irradiation in the narrow quartz tube (1.5 cm diameter) is severely shadowed by the magnetic stirring bar, reducing the effective light reaching the catalytic suspension. To further verify the carbon source of the products, isotope labeling experiments with ¹³CH₄ were conducted. The ¹³C NMR spectra displayed a peak at 49 ppm (Figure S12a), which can be assigned to ¹³CH₃OH, confirming that methanol originates from methane photooxidation [37]. Moreover, ¹H NMR analysis revealed the expected splitting of the methyl proton resonance for ¹³CH₃OH (Figure S12b), providing additional evidence that CH₃OH is derived from CH₄ [38].

To rationalize selectivity trends, a mechanistic hypothesis is proposed (Figure 1d). Pd/TiO₂ generates excessive •OH, promoting CH₃OH overoxidation to HCHO. In contrast, Pd‐NiO_x/TiO₂ and PdNi_0.49/TiO₂ effectively suppress •OH accumulation, favoring CH₃OH formation. Compared to reported systems, PdNi_0.49/TiO₂ exhibits competitive performance under optimized conditions (Figure 1e and Table S4). Control experiments confirmed that CH₄ oxidation proceeds exclusively via a photocatalytic pathway, with molecular O₂ as the sole oxygen source (Figure S13). Time‐dependent studies reveal typical kinetic features of photocatalytic methane oxidation (Figure S14). During the initial 20 min, the CH₃OH production rate was high (approximately 5043.7 µmol g⁻¹ h⁻¹) owing to the abundant CH₄ supply and low product concentration. As the reaction progressed to 60 min, the formation rate declined due to product accumulation, while the selectivity toward CH₃OH remained robust, indicating that the CH₃OH concentration had not yet reached the threshold for secondary reactions. The production of HCHO from 60 to 120 min increased rapidly as elevated CH₃OH concentration accelerated its further oxidation, leading to a pronounced decline in CH₃OH selectivity. Beyond 120 min, the production of CH₃OH nearly stopped increasing, whereas HCHO continued to accumulate, further diminishing CH₃OH selectivity. By 240 min, the system reached a quasi‐steady state, where the rate of CH₄ oxidation to CH₃OH was equilibrated with the subsequent oxidation of CH₃OH to HCHO, resulting in negligible net product accumulation. A five‐cycle stability test (Figure S15) demonstrated consistent CH₃OH yield and selectivity, confirming catalyst durability.

2.2. Structural Characterization of Photocatalysts

The morphology of Pd‐NiO_x/TiO₂ and PdNi_0.49/TiO₂ photocatalysts was examined via transmission electron microscopy (TEM). In both cases, uniformly dispersed nanoparticles were observed on the TiO₂ surface (Figure 2a and Figure S16a). High‐resolution TEM (HRTEM) further revealed lattice structure and crystallinity. For Pd‐NiO_x/TiO₂, the support exhibited a lattice spacing of 0.350 nm (Figure S16b), corresponding to the (101) planes of anatase TiO₂ [39]. The deposited nanoparticles displayed lattice fringes with an interplanar spacing of 0.226 nm, assigned to the (111) facet of metallic Pd [40]. High‐angle annular dark field scanning transmission electron microscope (HAADF‐STEM) and energy‐dispersive x‐ray spectroscopy (EDS) mapping confirmed homogeneous spatial distribution of Ni, Ti, and O elements (Figure S16c,d). Additionally, NiO_x/TiO₂ showed uniformly dispersed clusters with an average diameter of ∼2.1 nm and narrow size distribution (Figure S17) [41]. In PdNi_0.49/TiO₂, HRTEM revealed nanoparticles with a lattice spacing of 0.215 nm (Figure 2b), slightly contracted relative to Pd (111), indicating Ni incorporation and alloy formation [42]. EDS mapping confirmed spatial overlap of Pd and Ni signals (Figure 2c,d), validating the alloyed structure. Phase composition was analyzed by x‐ray diffraction (XRD), and all samples exhibited reflections corresponding to mixed‐phase TiO₂ (anatase and rutile, Figures S18 and S19). For Pd/TiO₂ and Pd‐NiO_x/TiO₂, a broad diffraction peak at ∼39.8° was observed, indexed to the (111) planes of face‐centered cubic (fcc) Pd. No distinct Ni‐containing crystalline phases were detected in Pd‐NiO_x/TiO₂, likely due to high dispersion and low content. In PdNi_0.49/TiO₂, the (111) peak exhibited a positive shift relative to Pd/TiO₂, indicating lattice compression from Ni incorporation [43]. This is consistent with the smaller lattice constant of PdNi alloy (a = 3.759 Å) compared to pure Pd (a = 3.890 Å), reflecting substitutional effects and partial lattice distortion.

FIGURE 2.

(a) TEM image, (b) HRTEM image, (c) HAADF‐STEM and (d) EDS‐mapping images of PdNi_0.49/TiO₂ photocatalyst. (e) Normalized Ni K‐edge XANES of PdNi_0.49/TiO₂, Ni foil, and NiO. (f) Fourier transform of the EXAFS spectra of panel (e). (g) WT‐EXAFS contour plots of PdNi_0.49/TiO₂.

X‐ray photoelectron spectroscopy (XPS) was further employed to probe electronic interactions and valence states (Tables S5‐S8). In the Ti 2p spectra (Figure S20a), NiO_x/TiO₂ showed a slight shift to lower binding energy relative to pristine TiO₂, indicating electronic perturbation from NiO_x loading. The Ni 2p spectra in NiO_x/TiO₂ catalyst (Figure S20b) revealed a positive shift in the Ni²⁺ peaks compared to commercial NiO (855.41 and 873.31 eV), suggesting that the nickel species are slightly more cationic than pure NiO and thus represented as NiO_x [41, 44]. The positive shift also implies strong electronic interaction between NiO_x and TiO₂. Pd 3d spectra of Pd/TiO₂‐NaBH₄ (Figure S20c) displayed characteristic peaks for Pd⁰ at 340.12 and 334.87 eV, and for Pd²⁺ at 341.46 eV and 335.93 eV [45]. The existence of Pd²⁺ likely originates from low‐coordinated surface Pd atoms prone to oxidation [46]. Notably, the binding energy of Pd 3d in Pd‐NiO_x/TiO₂ showed a positive shift compared to Pd/TiO₂, suggesting reduced electron density on Pd due to electron withdrawal by adjacent Ni²⁺ species. O 1s spectra (Figure S21) provided further insight into oxygen species. For NiO_x/TiO₂, peaks at 530.07 and 532.01 eV were assigned to lattice oxygen (O_lat) and surface‐adsorbed hydroxyl species (O_ads), respectively. The binding energy of O_lat species lies between that of pristine TiO₂ and commercial NiO (529.37 eV), reflecting a modified electronic environment upon NiO_x incorporation [47].

In contrast, PdNi_0.49/TiO₂ exhibited a negative shift in Pd 3d binding energy (Figure S22), indicating electron donation from Ni to Pd within the alloy. Given the higher electronegativity of Pd (2.20) relative to Ni (1.91), Pd–Ni metallic bonding facilitates electron transfer toward Pd, thereby increasing its electron density. For PdNi_0.49/TiO₂, a small fraction of Ni⁰ was detected by XPS, consistent with partial reduction via Pd–Ni interactions. The surficial Ni exists predominantly in oxidized states, most likely due to surface oxidation upon environmental exposure [48, 49]. No Ni⁰ signal was detected by XPS in Ni/TiO₂ (Figure S23), suggesting that Ni precursor cannot be reduced under identical conditions without Pd(acac)₂. Therefore, PdNi alloy formation proceeds via a noble‐metal‐induced reduction, in which Pd²⁺ is first reduced by CO and subsequently promotes Ni²⁺ reduction [29].

To elucidate the chemical state and electronic structure of Ni species, synchrotron‐based x‐ray absorption near‐edge structure (XANES) and extended x‐ray absorption fine structure (EXAFS) were performed. As depicted in Ni K‐edge XANES spectra (Figure 2e), the absorption edge of PdNi_0.49/TiO₂ lies between Ni foil and NiO, suggesting that Ni primarily exists in a +2 oxidation state with minor metallic Ni⁰ [50]. This is further evidenced by the white line peak (Figure S24), which is slightly lower than NiO and confirms a small fraction of metallic Ni⁰. The Fourier transform EXAFS spectrum of PdNi_0.49/TiO₂ shows a distinct Ni–Pd coordination peak at ≈2.18 Å (Figure 2f), slightly longer than the Ni–Ni bond (≈2.14 Å) in Ni foil [51, 52, 53]. A subtle shift of the Ni‐Ni_oxy peak in PdNi_0.49/TiO₂ sample toward longer distances indicates the incorporation of Pd into the Ni matrix. The presence of Ni–Pd and Ni–Ni_oxy metallic coordination excludes the atomic dispersion of Ni species. Meanwhile, the prominent Ni–O coordination signal observed at ≈1.51 Å rules out the presence of metallic Ni clusters [54]. The observed Ni–O coordination is tentatively attributed to the interfacial interaction between Ni and the TiO₂ support or surface oxidation. More comparison information about the radial distance k‐space resolution was obtained by a two‐dimensional wavelet transform (WT) EXAFS analysis (Figure 2g and Figure S25). The WT contour of PdNi_0.49/TiO₂ at ≈5‐6 Å^{− 1} is centered at ≈1.5 Å, indicative of Ni–O coordination similar to that in the NiO reference. Additional contours at ≈8–9 Å^{− 1} and ≈10–11 Å^{− 1} correspond to Ni–Ni and Ni–Pd metallic coordination, respectively, characteristic of PdNi alloy bonding. The corresponding EXAFS fitting results (Figure S26 and Table S9) further validate the formation of Pd–Ni metallic bonding along with Ni–O species, consistent with the above structural analysis.

PVP and CTAB were employed as stabilizing agents for PdNi_0.49/TiO₂ synthesis. Fourier transform infrared spectroscopy (FTIR) spectra were recorded to monitor any residual organic species (Figure S27). Weak signals at 2918 and 2849 cm^{− 1} (C–H stretching), 1677 cm^{− 1} (C═O stretching), and 1422 cm^{− 1} (C–N stretching or CH₂ bending vibration) indicate trace residual surfactants [55, 56]. However, control experiments with N₂ instead of CH₄ yielded no detectable products, and ¹³CH₄ isotope labeling experiment confirmed CH₃OH originates from CH₄. These results show that trace organic residues are inert with negligible influence on photocatalytic performance. Moreover, XPS spectra before and after reaction revealed no significant changes in the binding energies or peak shapes, confirming PdNi_0.49/TiO₂ maintains good structural and chemical stability during reaction (Figure S28).

2.3. Photophysical and Electrochemical Characterization

To elucidate the factors contributing to the high CH₃OH selectivity of PdNi_0.49/TiO₂, the charge separation efficiency and band structure of catalysts were systematically investigated. Photoluminescence (PL) spectroscopy was first employed to probe charge transfer behavior (Figure 3a). Pristine TiO₂ exhibited strong PL emission, indicative of pronounced charge recombination. Unexpectedly, Pd/TiO₂ showed increased PL intensity, likely due to Pd acting as an electron sink. At high loading levels, Pd nanoparticles accelerate recombination between photogenerated electrons and surface‐accumulated holes, reducing overall charge separation efficiency [57, 58]. PL spectra remained largely unchanged upon Ni modification of Pd/TiO₂ via two distinct approaches, suggesting minimal impact on charge transfer dynamics. Time‐resolved PL measurements revealed slightly shorter carrier lifetimes for Pd/TiO₂ (2.19 ns) and PdNi_0.49/TiO₂ (2.18 ns) compared to pristine TiO₂ (2.60 ns), indicating marginally faster recombination (Figure S29). To further assess charge transport behavior, photocurrent measurement and electrochemical impedance spectroscopy (EIS) were conducted (Figure 3b,c). Pristine TiO₂ exhibited the highest photocurrent density, whereas Pd/TiO₂ and PdNi_0.49/TiO₂ showed lower values, indicative of inferior charge separation efficiency. These results align with the higher overall CH₄ conversion rate observed over bare TiO₂ relative to the modified samples. Accordingly, although Pd‐based catalysts display moderately lower charge separation efficiency and total conversion, their remarkably enhanced CH₃OH selectivity arises from a cocatalyst‐induced shift in the dominant ROS pathway.

FIGURE 3.

Analysis of photocatalytic properties and electronic structures of photocatalysts. (a) Steady‐state photoluminescence spectroscopy. (b) Photocurrent response. (c) Electrochemical impedance spectroscopy. (d) Plots of transformed Kubelka‐Munk function versus photon energy. (e) Mott–Schottky plots at 500 Hz and (f) Band gaps and band structures of TiO₂, Pd/TiO₂, Pd‐NiO_x/TiO₂ and PdNi_0.49/TiO₂.

The band structure of TiO₂, Pd/TiO₂, Pd‐NiO_x/TiO₂, and PdNi_0.49/TiO₂ was examined using UV−vis diffuse reflectance spectroscopy (DRS) and Mott−Schottky analysis. All samples exhibited similar absorption edges near 400 nm (Figure S30). Upon cocatalyst modification, UV absorption weakened while visible‐light absorption was enhanced, attributed to light scattering by Pd nanoparticles and surface plasmon resonance (SPR) effects. Based on Tauc plots, the estimated band gap (E_g) of pristine TiO₂ is ∼3.33 eV (Figure 3d), with Pd‐ and PdNi‐modified samples showing slightly reduced values (∼3.20 eV). Mott–Schottky plots revealed flat‐band potentials (V_fb) of −0.53, −0.65, −0.65, and −0.65 V versus Ag/AgCl for TiO₂, Pd/TiO₂, Pd‐NiO_x/TiO₂, and PdNi_0.49/TiO₂, respectively (Figure 3e). The positive slope confirms the n‐type semiconductor nature of TiO₂, with conduction band edges (E_CB) positioned more negatively than the V_fb [59, 60]. Based on calculated E_CB and valence band (E_VB) positions (Table S10), all catalysts possess sufficient thermodynamic potential to oxidize H₂O into •OH (2.30 V vs. NHE) and reduce O₂ to •O₂ ⁻ (−0.33 V vs. NHE), as illustrated in Figure 3f.

2.4. Mechanism Investigation

Given the critical role of O₂ in CH₄ oxidation, we further examined its reduction pathway mediated by photogenerated electrons. ROS, including •OH and •OOH radicals, was monitored via in situ electron paramagnetic resonance (EPR) using 5,5‐dimethyl‐1‐pyrroline‐N‐oxide (DMPO) as a spin‐trapping agent. Under light irradiation, a characteristic quartet signal (1:2:2:1 splitting, a_N = a_H = 15 G) corresponding to •OH was observed (Figure 4a). Pd/TiO₂ exhibited the strongest •OH signal, followed by PdNi/TiO₂, while pristine TiO₂ and Pd‐NiO_x/TiO₂ showed significantly lower •OH levels. Quantitatively, •OH concentrations over Pd/TiO₂ and PdNi_0.49/TiO₂ were 2.16 and 1.46 times higher than those over Pd‐NiO_x/TiO₂, respectively (Figure 4b). The predominance of CH₃OH as the main product over Pd‐containing catalysts suggests that the coupling of •CH₃ with •OH is the key pathway for CH₃OH formation, whereas the lower •OH levels over Pd‐NiO_x/TiO₂ rationalizes its reduced CH₃OH yield. To validate the EPR results, •OH species were semi‐quantified using coumarin as a probe (Figure 4c) [61]. Consistent with the in situ EPR results, Pd/TiO₂ exhibited the highest fluorescence intensity of 7‐hydroxycoumarin, confirming its superior •OH generation capability.

FIGURE 4.

Evaluation of ROS in photocatalytic systems. (a) DMPO‐OH for monitoring the generation of •OH, (b) intensities of trapped •OH in EPR characterization, (c) fluorescence spectra of 0.1 mM coumarin solution, (d) DMPO‐OOH for monitoring the generation of •OOH, and (e) the kinetic constant of photodegradation of NBT for •OOH detection over TiO₂, Pd/TiO₂, Pd‐NiO_x/TiO₂ and PdNi/TiO₂. (f) Product yields of CH₃OH and HCHO after adding scavengers in the reaction system for trapping h⁺, •OH, e⁻ and •O₂ ⁻, respectively.

In a methanol medium, a sextet signal (a _N = 14.1 G, a _H = 10.0 G) attributed to DMPO‐OOH confirmed the presence of •OOH radicals (Figure 4d). Under illumination, Pd/TiO₂, Pd‐NiO_x/TiO₂, and PdNi_0.49/TiO₂ generated only minimal •OOH, whereas pristine TiO₂ exhibited a pronounced DMPO‐OOH peak. The [DMPO‐OOH]/[DMPO‐OH] ratio for pristine TiO₂ was 34 and 43 times higher than those of Pd/TiO₂ and PdNi_0.49/TiO₂, respectively (Figure S31), indicating that unmodified TiO₂ favors •OOH‐mediated pathways, likely leading to HCHO formation. This shift in ROS profile upon Pd and Ni modification promotes •OH generation and enhances CH₃OH selectivity. Complementary quantification of •OOH was performed via nitro blue tetrazolium (NBT) photodegradation, monitored by UV–vis spectroscopy and fitted with first‐order kinetics (Figure 4e and Figure S32). TiO₂ exhibited the highest rate constant (0.070 min⁻¹), while Pd/TiO₂, Pd‐NiO_x/TiO₂, and PdNi_0.49/TiO₂ showed significantly lower values (0.0075, 0.0047, 0.0069 min⁻¹, respectively), confirming the strong •OOH generation capacity over TiO₂.

To clarify the roles of individual ROS, scavenging experiments were conducted using salicylic acid (•OH), sodium oxalate (h⁺), para‐quinone (•O₂ ⁻), and potassium dichromate (e⁻) (Figure 4f). The presence of •OH, e⁻, and •O₂ ⁻ scavengers completely inhibited the formation of CH₃OH and HCHO, highlighting their crucial roles in CH₄ activation and product generation. Conversely, hole scavenging reduced product yields but did not eliminate them, suggesting h⁺ contributes to CH₃OH formation with limited impact on selectivity. To gain deeper insight into the ORR pathways, the formation of H₂O₂ was further quantified (Figures S33 and S34). Photocatalytic O₂ reduction experiments revealed that H₂O₂ was exclusively detected over Pd‐NiO_x/TiO₂, indicating a dominant 2‐electron ORR pathway. In contrast, Pd/TiO₂ and PdNi_0.49/TiO₂ preferred the 3‐electron pathway, correlating with their higher •OH generation. CH₄ oxidation using H₂O₂ as the oxidant yielded negligible CH₃OH (Figure S35), suggesting that H₂O₂ plays a limited role in ROS generation on the catalyst surface.

In situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) was conducted under CH₄ and air atmospheres to identify reaction intermediates. Under dark conditions with H₂O addition, baseline spectra were recorded (Figure S36). Upon illumination, TiO₂ showed strong CO₂ bands at 2361 and 2335 cm⁻¹ (Figure 5a and Figure S37) [62], along with peaks at 1262 and 1783 cm⁻¹ assigned to adsorbed HCHO* and C═O* species [27], confirming deep oxidation. Pd loading significantly suppressed CO₂ and HCHO signals (Figure 5b and Figure S38), while new bands at 2943 and 2881 cm⁻¹ emerged, corresponding to C–H stretching in *CH₃ and *OCH₃ species [6, 62]. Moreover, the peak at 1485 cm⁻¹ could be assigned to the *CH₃–O group, while the peaks at 1359 cm⁻¹ belonged to the *C–H group of *CH₃ species. Ni incorporation further suppressed CO₂ formation, suggesting enhanced control over oxidation pathways (Figure 5c and Figure S39).

FIGURE 5.

In‐depth investigations of the reaction pathway. In situ DRIFT spectra for photocatalytic CH₄ conversion over (a) TiO₂, (b) Pd/TiO₂ and (c) PdNi_0.49/TiO₂ in the presence of CH₄, O₂ and water vapor. Free‐energy diagram of O₂ reduction over Pd/TiO₂, Pd‐NiO_x/TiO₂, and PdNi/TiO₂. (d) 2‐electron and (e) 3‐electron O₂ reduction. The inset figure shows the adsorption models of intermediates on the surface of catalysts.

Density functional theory (DFT) calculations were performed to elucidate ORR pathway preferences (Figures S40–S42). In the 2‐electron pathway (Figure 5d), the protonation of *OOH to *H₂O₂ was identified as the rate‐determining step, with energy barriers of 0.42 eV for Pd‐NiO_x/TiO₂, 0.97 eV for PdNi/TiO₂, and a notably higher barrier of 1.78 eV for Pd/TiO₂. In the 3‐electron ORR pathway (Figure 5e), the protonation of *O to *OH was rate‐limiting, with barriers of 0.55 eV for Pd/TiO₂, 0.89 eV for PdNi/TiO₂, and 1.07 eV for Pd‐NiO_x/TiO₂. These results indicate distinct pathway preferences among the catalysts. Pd/TiO₂ exhibits a lower barrier for the protonation of *O to *OH, thereby conferring a superior kinetic advantage for *OH formation and increasing the likelihood of subsequent •OH species generation. Conversely, Pd‐NiO_x/TiO₂ possesses a lower barrier for the 2‐electron route and produces detectable H₂O₂, aligning with its suppressed •OH signals and reduced yield. PdNi/TiO₂ displays comparable energy barriers for both pathways, with a slight preference for the 3‐electron route, thereby balancing *OH generation and overoxidation risk. Since the evolution of *OH on the catalyst surface is also regarded as a pivotal step for the generation of •OH, we further evaluated the adsorption behavior of *OH species. The free‐energy diagram (Figure 5e) suggests that PdNi/TiO₂ exhibits stronger adsorption capability for *OH compared with Pd/TiO₂. This trend was further validated by the calculated d‐band center (Figure S43), which shows that Ni incorporation induces an upward shift in the Pd d‐band center. According to the well‐established d‐band theory [63, 64, 65, 66], the upward shift of d‐band center is expected to strengthen *OH adsorption, thereby inhibiting the transition of *OH into highly reactive •OH species. Taken together, these theoretical insights suggest that PdNi/TiO₂ manifests a dual suppressive effect on the generation of •OH species, arising from kinetic and thermodynamic perspectives. This inference is in good agreement with the experimental observations, which revealed a significantly weaker •OH signal for PdNi/TiO₂, thereby providing a plausible mechanistic explanation for the experimental observations.

Beyond the modulation of oxygen reduction pathways, the desorption kinetics of the target product, CH₃OH, also plays a crucial role in determining product selectivity [67]. To investigate this, we calculated the adsorption energies (E _ads) of CH₃OH on different catalyst surfaces. DFT calculations reveal that CH₃OH exhibits strong adsorption on Pd/TiO₂ with an E _ads of –1.474 eV (Figure S44), which extends the surface residence time of CH₃OH and consequently leading to undesired overoxidation. In contrast, PdNi/TiO₂ exhibits a much weaker adsorption energy of –0.400 eV (Figure S45), which facilitates rapid desorption of CH₃OH and effectively suppresses its secondary oxidation to HCHO. These results are consistent with our experimental observation that PdNi/TiO₂ delivers the highest CH₃OH selectivity. The incorporation of Ni not only regulates ROS dynamics through modulation of the ORR pathway but also weakens CH₃OH adsorption, thereby facilitating its desorption and ensuring selective methane oxidation.

Based on these findings, a preliminary mechanism of photocatalytic CH₄ oxidation over TiO₂‐based catalyst is proposed (Scheme 1). On pristine TiO₂, photoexcited electrons reduce molecular O₂ to generate •OOH, which subsequently couples with •CH₃ to form CH₃OOH. The resulting CH₃OOH undergoes spontaneous decomposition to HCHO, with the overall process dominated by an overoxidation pathway. In the case of Pd/TiO₂, photogenerated electrons transfer from TiO₂ to Pd nanoparticles, where adsorbed O₂ is reduced. Pd/TiO₂ enhances •OH generation via the 3‐electron ORR, boosting CH₃OH yield but also increasing overoxidation risk. Pd‐NiO_x/TiO₂ favors the 2‐electron pathway, forming H₂O₂ and suppressing •OH species, resulting in high selectivity but low yield. PdNi_0.49/TiO₂ enables precise modulation of O₂ reduction, where the synergistic effect of regulated kinetic barriers and strengthened *OH adsorption effectively controls •OH availability. This balance facilitates CH₄ activation while suppressing CH₃OH overoxidation. Overall, by tailoring interfacial interactions among Pd, Ni, and TiO₂, the nature and concentration of ROS can be precisely modulated, offering molecular‐level control over CH₄‐to‐CH₃OH conversion.

SCHEME 1.

Schematic illustration of photocatalytic methane conversion over TiO₂, Pd/TiO₂, Pd‐NiO_x/TiO₂, and PdNi_0.49/TiO₂ photocatalysts.

3. Conclusion

In summary, our study presents a rational strategy for enhancing the selective photocatalytic conversion of methane to methanol under ambient conditions by tuning ORR pathways through cocatalyst interface engineering. By alloying Pd with Ni on TiO₂, we constructed a photocatalyst that achieves a high CH₃OH yield of 9557.4 µmol g⁻¹ h⁻¹ and selectivity of 87.5%, outperforming both monometallic and oxide‐modified counterparts. Mechanistic insights reveal that PdNi_0.49/TiO₂ alloy catalyst maintains moderate barriers for the protonation of *O to *OH within the 3‐electron ORR pathway compared with Pd/TiO₂. Meanwhile, Ni incorporation induces an upward shift in the Pd d‐band center, strengthening *OH adsorption on the catalyst surface. This dual effect, arising from both kinetic and thermodynamic contributions, manifests a suppressive influence on •OH generation, which aligns well with the experimentally observed weaker •OH signals for PdNi_0.49/TiO₂. Such regulation of ROS dynamics enables efficient CH₄ activation while mitigating CH₃OH overoxidation. In addition, DFT calculations demonstrate that Ni incorporation weakens CH₃OH adsorption, facilitating its rapid desorption and further ensuring high product selectivity. Overall, this work highlights the critical role of interfacial design and electronic structure modulation in steering ROS‐mediated oxidation pathways, offering a promising approach for sustainable methane valorization via photocatalysis.

Author Contributions

Zhuo Liu: methodology, investigation, data curation, formal analysis, visualization, writing – original draft. Tao Guan: investigation, data curation, visualization. Huanhuan Sun: investigation, visualization, formal analysis. Xuanzhao Lu: data curation, investigation. Pengyu Zhou: investigation, visualization. Yi Liu: software, validation. Yun Chen: resources, funding acquisition. Jun‐Jie Zhu: resources, supervision, project administration. Jian‐Rong Zhang: supervision, writing – review and editing, project administration. Wenlei Zhu: conceptualization, methodology, supervision, project administration, funding acquisition, writing – review and editing, resources.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: anie72771‐sup‐0001‐SuppMat.docx.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.